EPA-340/1 -83-022
 Wet Scrubber Inspection
„,_.„.....,..'• .   .  ,  arid...-..:J.7-._w.-~r~
                               ,> * -^.
      Evaluation

                  by

            Engineering-Science
             501 Willard Street
             Durham, NC 27701
           Contract No. 68-01-6312
           Work Assignment No. 33
        EPA Project Officer: John R. Busik
       EPA Project Manager: Kirk E. Foster
                Prepared for

     U.S. ENVIRONMENTAL PROTECTION AGENCY
      Office of Air Quality Planning and Standards
       Stationary Source Compliance Division
            Washington, D.C. 20460
              September 1983

-------
                                DISCLAIMER
     _..       ^ ,  ,     •••'••• " : :'. v    ;••-j  ;   ; ,:,;iC   ;; , f.;+   u.     	  ,•
     This report was prepared by Engineering-Science for-the Stationary
Source Compliance Division of the U.S.  Environmental Protection Agency in
partial fulfillment of Contract,.^$&&}-&&& TftsK^  .The cmtents of
this report are reproduced herein as received from the contractor.  The
opinion, findings, a^ppn^]^^^,^pr^a8|i;iare}^hflSfts-olf3^*i3B authors and
not necessarily those of 1J.S. Environmental Protection Agency.  The mention
of product names does not cpnsti^uteilendons.ement;,^ ttie ,U.\S> Environmental
Protection Agency.  Copies of th'is" report are available through the
Library Services Office (MD-35), U.S. Environmental Protection Agency
Research Triangle Park, N. C. 27711, or, for a fee, from National Technical
Information Services, 5285 Port Royal Road, Springfield, Virginia 22161.
                                 ii

-------
                             TABLE  OF  CONTENTS
      List  of Figures
      List  of Tables
      Glossary
 1.0   INTRODUCTION
Page

  vi
vi i i
  ix
          -Wet Scrubber Sys-tem'-OOTrtttOneh'ts'
          2.2.2  Dependencei" "of'Scrubber Performance on'
                 Particle Size                                      2-12
          2.2.3  Factors That Modify Particle Size                  2-15

     2.3  Pressure Drop as a Performance Parameter                  2-17

          2.3.1  Definition of Pressure Drop                        2-17
          2.3.2  Relationship Between Penetration and
                 Pressure Drop                                      2-20

     2'.4  Plume and Breeching Opacities as Performance Parameters   2-35

     2.5  Liquid-to-Gas Ratio as Performance Parameters             2-37

     2.6  Effect of Liquor Surface Tension                          2-39

     2.7  Condensation and Evaporation Effects on Performance       2-40

3.0  FACTORS AFFECTING SCRUBBER SYSTEM RELIABILITY                  3-1

     3.1  Corrosion and Erosion of Scrubber Shell                   3-1

     3.2  Erosion and Pluggage of Spray Nozzles                     3-3

     3.3  Fans                                                      3.5

     3.4  Pumps                                                     3_6

     3.5  Piping and Valves                                         3-7

     3.6  Ducts                                                     3_7
                                   i i i

-------
                       TABLE OF CONTENTS (Continued)
                                                                    Page
 4.0  SCRUBBER INSPECTION AND PERFORMANCE EVALUATION

      4.1   Baseline Performance Evaluation Technique

           4.1.1    File Review
           4.1.2    Evaluation of Plume Opacity
           4.1.3    Fan  Evaluation
           4.1.4    Demister Pressure Drop
           4.1.5    Scrubber Pressure Drop;,<•.',;•'  AT !"7 I fAT-, DHA:-
           4.1.6    Liquid-Gas Distribution
    „ „    4.1.7    Liquor Quality
          '4.1.8    Nozzle Pluggage
    MI    4.1.9    Bypass Duct Damper    '2TJIAH3 QHITOOHZ3JaUOflT
           4.1.10   Presaturator and  Ductwork
           4.1.11   Process Conditions

      4.£   /Application  Of Ddacr i M<^ rei'rOr.ila.r^t- JVdi Uid'irui:   	 •
           Technique to Specific  Types of Scrubbers

           4.2.1  Preformed Spray Scrubbers
           4.2.2  Packed Bed Scrubbers                  ,
           4.2.3  Moving Bed Scrubbers
           4.2.4  Tray-type Scrubbers
           4.2.5  Gas-atomized  Scrubbers

     4.3   Application  of Baseline Performance Evaluation
           Technique to  Specific  Industries

           4.3.1  Sludge  Lime Kilns  at Kraft P.ulp Mills
           4.3.2  Asphalt Batch Plants
           4.3.3  Grey  Iron  Foundries
           4.3.4  Coal-fired Boilers
           4.3.5  Municipal  Incinerators
           4.3.6  Basic  Oxygen  Furnaces

5.0  EVALUATION OF OPERATION AND  MAINTENANCE  PROCEDURES

     5.1   Startup and Shutdown Procedures

     5.2  Routine Maintenance

          5.2.1  Fans
          5.2.2  Pumps
          5.2.3  Valves
          5.2.4  Nozzles
          5.2.5  Recirculation Tank
          5.2.6  Dampers
          5.2.7  Instruments
          5.2.8  Motors
          5.2.9  Demisters
 4-1

 4-2

 4-4
 4-5
 4-6
 4-6
 4-16
 4-16
 4-17
 4-18
 4-19
 4-19

 4-21
 4-23
 4-25
 4-27
 4-29
4-38

4-39
4-40
4-44
4-46
4-47
4-48

5-1

5-1

5-3

5-3
5-4
5-4
5-4
5-5
5-5
5-5
5-6
5-6
                                    iv

-------
                 TABLE OF CONTENTS  (Continued)
5.3  Elements of Model Record and Report  System


     5.3.1  Scrubber Operation Record  Elements
     5.3.2  Process Record Elements
6.0  REFERENCES


APPENDIX A  PENETRATION


APPENDIX B  BASIC STATISTICAL METHODS


APPENDIX C  DATA TABLES
   •:r_&

APPENDIX D  TROUBLESHOOTING CHARTS
                                  ^qtWitjouCJ aasq^B   6.1
                                    tine "f0.l6"iy^6a9"fcl  OJ.
                                             jasoo1"^  H.
.r
,j>
,i>
.&
Page


 5-6


 5-7
 5-10


 6-1


 A-l


 B-l


 C-l


 D-l

-------
                             LIST OF FIGURES
Figure No.
                                                                   Page
2-1
2-2a
2-2b
2-3a
2-3b
2-4a
2-4b
2-5
2-9
2-10
2-11
2-12

2-13
2-14

2-15

2-16

2-17

2-18

2-19

2-20

2-21

2-22

2-23


2-24

2-25

2-26
Example wet scrubber system.                                 2-
Negative pressure system.                                    2-
Positive pressure system.                                    2-
Once-through liquor system.                                  2-
Recirculated liquor system.                         v         2-
Scrubber system with evaporative cooler. '.'Jj>j^  to  jr;»ii,s''  ;i;&s42-
Scrubber system with presaturator.       '"    "   "            2-
Neutralization system.                                       2-
Hypothetical curve illustrating general rela-Wo^sh^p huinsv  e2-

Fractional efficiency curves for venturi scrubteek'saoa  rTUj-nsv2-
Predicted relationship betweeff9^e!fS?l V^e^etfcstn'icmeanjd rsorqyl?-
particulate mass mean diameter,  ."'-^rhnv bed  be>bf-a  ffwroyj
Parallel scrubber trains.                                    2-
Parallel scrubber trains with unequal  resistance.            12-
Converging and diverging venturi angles.                     2-
Contact power curve for a cyclone scrubber controlling       2-
      talc dust.
      Contact power curve for
      Contact power curve for
      furnace fume.
      Contact power curve for
      lack of independence of
                        venturi scrubbers on open hearths.
                        scrubbers controlling ferrosilicon
                                              a pilot scale
                                              slip stream.
                                              venturi scrubbers
serving three Q-BOP furnaces (plotted using data of Kemner
and Mcllvaine).
Opacity vs. emissions data for a venturi scrubber on
a coal-fired boiler slip stream.
Predicted effect of liquid-to-gas ratio on penetration
as calculated using the venturi scrubber model.
Beneficial effect of condensation.
2-
2-

2-
                        venturi scrubbers illustrating
                        the liquid-to-gas ratio.
Relationship between pressure drop and emissions for
flooded disc scrubbers on lime kilns.
Replotted Walker and Hall data; 90% confidence interval
is denoted by shaded area.
Pressure drop versus emissions for wet scrubbers serving
coal-fired utility boilers.
Replotted data of Kashdan and Ranade using only scrubbers
90% confidence interval is denoted by shaded area.
Pressure drop versus emissions data for venturi scrubbers
serving asphalt plants.
Pressure drop versus emissions data for-venturi scrubbers    2-
serving coal thermal dryers.
Pressure drop versus emissions data for
venturi scrubber on a coal-fired boiler
Pressure drop versus emissions data for
                                                                   o_
                                                                   2-
                                                                   2-
                                                                   2-
                                                                   2-
2-

2-
                                                                   2-
                                                                   2_
                                                                   2-
-2
-4
•4
-5
•5   :

-8"
•9
-12
 da-*
-13
•K-£
  8-f
-19
•19
-20
•24

•25
•26

 26

 27

 28

 29

 30

 31

 32

 33

 34


 36

 37

 41
3-1  Areas of a venturi scrubber vulnerable to erosion.
                                                                   3-2
                                    vi

-------
                       LIST OF FIGURES (Continued)
Figure No.
                                                                   Page
4-1
4-2
4-3
4-4a
4-4b
4-5
4-6a
   Ji j,
4-6b
4-8
4-9
4-10
4-11
4-12
4-13
4-14
4-15
4-16
4-17
4-18
4-19
4-20

A-l
 A-2
 A-3
 A-4
      Counterflow inspection sequence.
      Preferred static pressure measurement locations.
      Aspiration effect error in static pressure measurement,
      Measurement of static pressure using a copper tube.  .
      .Measurement of static pressure using a pitot tube.  ,
      Measurement port seal.
      Poor approach, for measuring pressure drop across
     -,;a venturi escRubber:*si  is'isnsg gnf•tBi^aurrr.evtuD
      Preferred approach for m.e^surijitg/^i^SjS
                            'nujnsv 'fctf asviuo xprmorTts
•Sventuri scrubbed^-
•STypical .,	._   .
 Typical packed bed  scrubber.    ;  -—-  ~:,  — ^ -z-:~ r—;;•:;-;.!••'-
 Typical packing materials.
 Typical moving bed  scrubber.
 Tray-type  scrubber.
 Types  of stages.
 Typical orifice type  gas-atomized scrubber.
 Typical orifice type  gas-atomized scrubber.
 Venturi scrubber with long  divergent  section.
 Flooded disc  type  venturi  scrubber.
 Adjustable venturi  throat  mechanism.
 Variable rod  venturi  scrubber.
 Pressure drop collection efficiency  relationship.
 Gas  flow rate and  CO  content  during  a blow.

 Penetration as related to control device inlet
 concentration, solids collection, and resulting emissions;
 in this  case, the  penetration equals 0.01.
 Conversion from collection efficiency to transfer units.
 Conversion from penetration to transfer units.
 Conversion from collection efficiency to penetration.
                                                               4-3
                                                               4-8
                                                               4-9
                                                               4-10
                                                               4-10
                                                               4-11'
                                                               4-12
 C-l   Gas densities as a function- of temperature and pressure.
                                                                     4-24
                                                                     4-26
                                                                     4-28
                                                                     4-28
                                                                     4-30
                                                                     4-31
                                                                     4-33
                                                                     4-36
                                                                     4-37
                                                                     4-38
                                                                     4-45
                                                                     4-49

                                                                     A-l
                                                                     A-3
                                                                     A-4
                                                                     A-5

                                                                     C-3
                                     vi i

-------
LIST OF TABLES
                   M"
                   o,

                   0
Table  No.

2-1    Effects  of Recirculation  Liquor  Solids
2-2    Predicted Effect  of  Surface  Tension  on
       the  Penetration through a Venturi  Scrubber

4-1    Performance  Evaluation Parameters
4-2    Routine  Inspection Points
4-3    Fan  Data Temperature Correction
4-4    Major Types  of Wet Scrubbers
4-5    Preformed Scrubber Evaluatclon^Data^,. ^..•.,
4-6    Packed Tower Evaluation P.a.ta  .   , .  ,.
4-7    Turbulent Contact Absorb^'l^i/IWfcn'tfeta
4-8    Tray-type ScrabfoBf fEyaflniBarMasan rtetgn'rnnuO =
4-9    Orifice Scrubber  Evaluation  Data
4-10   Evaluation Data for  Venturi" Scrubbers '     , v •   -  .,-,..
4-11   Typical Operating Characteristics  of  Scrubbers on
       Rotary Lime  Kilns of Kraft Pulp Mills
4-12   Typical Operating Characteristics  of  Asphalt Plant
       Particulate  Wet Scrubbrs
4-13   Typical Operating Characteristics  of  Particulate Scrubbers
       on Grey Iron Cupolas
4-14   Typical Operating Characteristics  of  Particulate Wet
       Scrubbers on Coal-fired Boilers
4-15   Typical Operating Characteristics  of Municipal Incinerator
       Venturi Scrubbers
4-16   Typical Operating Characteristics  of  Venturi Scrubbers
       Serving Basic Oxygen Furnaces

5-1    Scrubber Operation Data
5-2    Process Condition Data
5-3    Daily Scrubber Operation  Log
5-4   Weekly Scrubber Ancillary Equipment Performance Log
5-5    Monthly Scrubber Ancillary Equipment Performance Log
5-6   Semi-Annual  Scrubber Ancillary Equipment Performance Log

C-l    Viscosities of Air at 1 Atm Pressure
C-2   Gas Density at Saturated Temperature and Pressure

D-l   Troubleshooting Chart Based on Parameter Measurements
D-2   Troubleshooting Chart Based on Scrubber Condition and
      Performance
                                     2-16
                                     2-39
4-1
4-3
4-15
4-20
4-22
4-25
4-26
4-29
4-30
4-34
4-40

4-42

4-45

4-46

4-48

4-50
                                     5-8
                                     5-10
                                     5-12
                                     5-13
                                     5-14
                                     5-16

                                     C-l
                                     C-2

                                     D-2
                                     D-5
    vi i i

-------
                                  GLOSSARY*


 ADIABATIC SATURATION.1  A process by means of which an air or gas stream is
      saturated with water vapor without adding or subtracting heat from the
      system.

 AERODYNAMIC DIAMETER.  The diameter of a unit density sphere having the .
      same aerodynamic properties as an actual particle.  It is related
      to the physical diameter according to the equation below:


                            "pa- Vc..>v2.  ;.  :'   v;;   V-,     -
            where:  dpa = aerodynamic diameter
 ,.,.  .                dp  = physic^&idtameter, 6vj  -isciauioc; benno'tsi1^
 2S~P>                                          -
  -                  p
 8S-£                 C  = Cunningha'm s^hheeftd'alh -fatct'OTog sq^-ye-iT   8-4
 0£-£     •                     "   scfeCI  nofjeufsv.l  Ts
-------
CYCLONE.  A device in which the velocity of an inlet gas stream is trans-
     formed into a confined vortex from which inertia! forces tend to drive
     particles to the wall.

DAMPER. 2  An adjustable plate installed in a duct to regulate gas flow.

DEHUMIDIFY.1  Reduction of water vapor content of a gas stream.

DEMISTER.  A mechanical device used to remove entrained water droplets from
     a scrubbed gas stream.

DENSITY.2  The ratio of the mass of a specimen of a substance to the volume
     of the specimen.  The mass of a unit volume of a substance.
DIFFUSION (AEROSOL).
Random motion of particles caused by repeated colli-
           rjDsffos  Toacms  To  ad^'ft- ..... .MOIW)33,OXMI
        fts  to  eb ha. .9/11 zSwswt J»3crts6  ne  rbrriw
                     aoszci ..... fts to e   a. .9/   zwsw  »3crts6  ne  rrw  .  •   •
                   Force acting on a parti,cl£,^ef^c±i,narmpyemenFt due to a
               .    ..       i*.      ti* "•*."' > > * — * -'* *r''»"-W <. "" C C v •* •u/ *%  V «J f L ' 'Jf *Ji U
     vapor condensation gradient, resultant of differences in molecular
     impacts on opposite sides of a particle.

DRAFT.3-  A gas flow resulting from the pressure difference between the
     incinerator, or any component part, and the atmosphere,  which moves
     the products of combustion from the incinerator to the atmosphere.
     (1) Natural draft:  The negative pressure created by the difference
     in density between the hot flue gases and the atmosphere.   (2)
     Induced draft:  the negative pressure created by the vacuum action
     of a fan or blower between the incinerator and the stack.   (3)
     Forced draft:  the positive pressure created by the fan or blower,
     which supplies the primary or secondary air.

DRAG FORCE.  Resistance of a viscous medium due to relative motion of a
     fluid and object.

DUST.2  Solid particles less than 100 micrometers created by the attrition
     of larger particles.

DUST LOADING.2  The weight of solid parti cul ate suspended in an airstream
     (gas), usually expressed in terms of grains per cubic foot, grams
     per cubic meter, or pounds per thousand pounds of gas.

EXCESS AIR.1  Air supplied for combustion in excess of that theoretically
     required for complete combustion; usually expressed as percentage of
     theoretical air (130% excess air).

FEEDSTOCK.1  Starting material used in a process.  Can be raw material or
     an intermediate product that will undergo additional  procesing.

GRAVITY, SPECIFIC.2  The ratio of the mass of a unit volume of a substance
     to the mass of the same volume of a standard substance at a standard
     temperature.  Water is usually the standard substance.   For gases,
     dry air at the same temperature and pressure as the gas is often the
     standard substance.

-------
GRID.1  A stationary support or retainer for a bed of packing in a packed
     bed scrubber.

HEADER.1  A pipe used to supply and distribute liquid to downstream outlets.

HUMIDITY, ABSOLUTE.2  The weight of water vapor carried by a unit weight
     of dry air or gas.

HUMIDITY, RELATIVE.2  The ratio of the absolute humidity in a gas to the
     absolute humidity of a saturated gas at the same temperature.

HYDROPHIL1C MATERIAL.  Particulate matter that adsorbs moisture.

INERTIA.  Mbme;rrtum^ tencleriCy" to" remairf frf a ¥fxedJrdirec%Ton,' 'proportional
     to mass and velocity.                                  ,     .    , ;

INTERCEPTION.  A type of aerosol collection re?I%1t&'a3fef fMpac¥icPrf1,0fa
   .  which an aeroso.l. impacts ins side of an obstacle because of reduced
&  oj-  ^o^^f^^ro^^^^^^TH^^^^sq  &  "o enrj-DB  eofoi  .aiaaflQHqoiaunia
   isfuosfom n'r  aeons-talari}  TO  jn'stfuest  .jn-sfbeig noMeartabftoD  locifsv
ISOKINETIC SAMPLING.  Matchingr:the''g¥s %eY6cTty%t'^h¥%¥aip1!'inV'-pWb1e
     entrance to the gas velocity of the localized gas stream to collect a
     representative particle size distribution.

LIQUOR.1  A solution of dissolved substance in a liquid (as opposed to a
     slurry, in which the materials are insoluble).

LOG-NORMAL DISTRIBUTION.  A series of points that can be defined by a geo-
     metric mean value and a geometric standard deviation.

MEAN FREE PATH.  The average distance between successive collisions of gas
     molecules; related to molecular size and number per unit volume.

OPACITYo  Measure of the fraction of light attenuated by suspended particu-
     late.      . ,      .   .                                    '

PARTICLE.  Small discrete mass  of solid or liquid matter.

PARTICLE SIZE.  An expression  for the size of liquid or solid particle.

PARTICULATE MATTER.  As related to control technology, any material except
     uncombined water that exists as a solid or liquid in the atmosphere
     or in a gas stream as measured by a  standard (reference) method at
     specified conditions.  The standard method of measurement and the
     specified conditions should be implied in or included with the
     particulate matter definition.

PARTICULATE MATTER, ARTIFACT.   Particulate matter formed by one or more
     chemical reactions within  the sampling train.

PENETRATION.  Fraction of suspended particulate that passes through a
     collection device.

pH.1   A measure of acidity-alkalinity of  a solution; determined by
     calculating the negative  logarithm of the hydrogen ion concentration.
                                     xi

-------
  POLYDISPERSITY.  A particle size distribution consisting of different
       size particles.

  PRESSURE, STATIC.  The pressure exerted in all  directions by a fluid-
       measured in a direction normal  to the direction of flow.

  PRESSURE, JOTAL.  The algebraic sum  of the velocity pressure and  the
       static pressure.

  PRESSURE, VELOCITY.   The kinetic pressure in the  direction  of  gas  flow.

  PRIME COAT (PRIMER). 1  A first coat  of paint applied to inhibit corrosion
       or to improve adherence of the  next  coat.

  QUENCH. 1  Cooling of  hot gases by rapid evaporation  of  water.

  REYNOLDS NUMBER,  FLUID.   A  dimensionless  quantity in  fluids  to describe the
             °J nWfW                         enrnssID z&3  FeH^ubni   .S
 ni-«»«, r.~         -SYei taupuA eS-l no rJso t rdul  .anneT lo. ^eaaofS bns
 REYNOLDS NUMBER  PARTICLE.' A dimensionless quantity in aerosol science to
       describe the ratio of inertia! to viscous forces relative to the parti-


 SATURATED GAS.l  A mixture of gas and vapor to which no additional vapor
      can be added, at specified conditions.  Partial pressure of vapor is
      equal to vapor pressure of the liquid at the gas-vapor mixture tempera-


 SIZE DISTRIBUTION.  Distribution of particles of different sizes within a
      matrix of aerosols;  numbers of particles of specified sizes or size
      ranges, usually in micrometers.

 SLURRY.1  A mixture of liquid and finely  divided insoluble solid materials.

 SMOKE.  Small  gasborne particles resulting from incomplete combustion-
      particles consist predominantly of carbon  and  other  combustible ma-
      terial;  present in sufficient quantity  to  be observable  independently
      OT  otner solids.

 SPECIFIC GRAVITY.1   The ratio  between the  density of a substance at a
      given  temperature and  the density  of  water at  4°C.

 SPRAY NOZZLE.1  A device used for  the controlled introduction of scrubbing
      liquid at predetermined rates,  distribution patterns, pressures  and
      droplet sizes.                                                 '

 STOKES NUMBER.  Descriptive of the particle collection potential of a
      specific system;  the ratio  of particle-stopping distance to the
      distance a particle must travel to be captured.

STREAMLINE.  The visualized path of a fluid in motion.

TEMPERATURE, ABSOLUTE.2  Temperature expressed in degrees  above absolute zero.

TERMINAL SETTLING VELOCITY.   The steady-state speed  of  a  falling particle
     after the equilibration of gravitation,  drag,  and  buoyant forces  has
     OCCUPPGQ •              =                                       '
                                   xii

-------
VAPOR.  The gaseous form of substances  that are  normally in the solid or
     liquid state and whose states can  be  changed either by increasing
     the pressure or by decreasing the  temperature.

WET/DRY LINE.1  The interface of  hot, dry  particulate-laden gas and
     cooling or scrubbing liquid, at which an  accumulation of solids
     can occur.
                                 GLOSSARY
                                REFERENCES
                                         Wet Scrubber Terminology.

                                                           ,83SMU!/i
 Industrial  Gas Cleaning Institute.
 Publication  WS-1, July 1975.
fdrioasb oJ  abruR  nr  ^r^nsup aaesnor.ansmrb A
 Industrial  Gas Cleaning Insti.tu-teT.cft FWdame'ht
 and Glossary of Terms.  Publication F-2, August 1972.

      Ot Av'-jsfqt 3<3.TKn gUCl-e'tV. 07 i5rt'"3fU fQ f
                                                              „.,,,„
                                                             »
                                    xm

-------
     ysiqz bne .atsddutaa
989r(j- aonra atsddinoa bss
                  rorJurfcq
bsMosq .atsddinoz sq^-sJefq ,at9ddutDa:'
       o* nsv'rg ar e^srtqraB  ,?t9ddutoa
          beau sq^ nommoa. Jaora'sri* STB

-------
               '             1.0  INTRODUCTION

     The Stationary Source Compliance Division of the U.S.  Environmental
Protection Agency has sponsored the development of this manual.   The
purpose of the manual is to assist EPA personnel, State and local  agency
personnel, and operating personnel in the routine evaluation of  wet
scrubbers in order that abnormal operating conditions contributing to
excessive emissions can be identified and rectified as rapidly as possible.
It covers the inspection and evaluation of particulate wet scrubbers
installed at stationary sources.  And its scope includes gas-atomized
scrubbers, pi ate-type scrubbers, packed tower scrubbers, and spray tower
scrubbers.  Emphasis is given to the gas-atomized scrubbers since these
are the most common type used for large air pollution control systems.
     This manual  presents specific evaluation techniques for identifying
and assessing wet scrubber operating problems.  It is not intended to be
an exhaustive survey of operation and maintenance (O&M) practices for wet
scrubbers or a summary of design  principles; this material is already
available in commercial literature and in numerous EPA publications.
     In  the past, regulatory agencies and operators alike have used
emission correlations based on  wet scrubber  performance parameters as an
indirect measure of  performance.  Generally, this practice has meant simply
evaluating the pressure drop and  comparing it with pressure drops generally
experienced  in the  specific industry.  Among the numerous problems encoun-
tered  with this  approach  is plant-to-plant particle size .differences.  In
many  scrubber  applications, emissions  sometimes  vary  substatially without
a major  change in  the  observed pressure  drop.  Thus,  an evaluation approach
based  only on  pressure  drop is not adequate.   In  addition,  simply recording
the  values indicated by the differential  pressure  gauges  installed on wet
scrubber systems can lead to erroneous conclusions,  as  they  are  sometimes
not  operating  correctly.
                                    1-1

-------
     In the evaluation approach proposed in this manual, current performance
parameters are evaluated against their site-specific baseline values for
the wet scrubber being evaluated.  The influence of important variables
or parameters that cannot be measured, such as the degree of maldistribution
during liquid-gas contact and the particle size distribution, is thereby
taken into account.  An additional  noteworthy aspect of the proposed
approach is that parameters beside  pressure drop must be considered.  The
available literature has been reviewed to identify those variables  having
a significant impact on scrubber performance.
                                  '«Kj-0'3t?iS
                                                    ........ gff
                                  1-2

-------
       2.0  BASIS FOR SELECTION OF  SCRUBBER PARAMETERS  USED  IN
               .           PERFORMANCE EVALUATION
     Routine evaluation of particulate wet scrubbers  depends  on  the
assessment of scrubber operating variables that are both  meaningful  and
measurable.  Description of such a performance evaluation approach must
start with the identification and understanding of the functions of
scrubber system, components and a comprehension of the significant operating
variables.  These are discussed In- tire "Subsequent .sections.   '  - •    -
2.1  WET SCRUBBER SYSTEM COMPONENTS
     Because operating problems affecting performance of wet scrubbers
originate not only in the scrubber vessel, but also in the process  equip-
ment and in the independent components of the scrubber system, the scrubber
vessel must be evaluated as part of the larger system into which it is
integrated, and not as an isolated piece of equipment.  A wet scrubber
system is composed of a large number of individual components all of
which must work properly even when process conditions vary.   A simplified
flowchart of one wet scrubber system is shown in Figure 2-1.  Individual
components which make up this particular system are listed below:
                    1.  Scrubber vessel
                    2.  Demister
                    3.  Fan
                    4.  Recirculation  tank
                    5.  Recirculation  pumps
                    6.  Presaturator
                    7.  Bypass duct and dampers
                    8.  Alkaline  additive  system
                    9.  Clarifier
                    10.  Vacumn filter
                    11.  Purge and make-up  systems
                    12.  Stack
                    13.  Flow monitors
                    14.  pH monitors
                    15.  Static  pressure monitors
 There is considerable diversity  in  the design of  wet  scrubber systems
 due to the different  control  requirements,  different  water availabilities
 and qualities,  and process  related  factors.
                                    2-1

-------
                                                           I
                                                           to
                                                           s-
                                                           a>
                                                           3
                                                           &.
                                                           O
                                                           OJ

                                                           "o.


                                                           
-------
      Gas movement through the wet scrubber system is maintained by the
 use of a fan  (except for a few systems in which the process equipment
 pressure is sufficient to move the gas).  If the process equipment and
 the scrubber  vessel are located before the fan, as shown in Figure 2-2a,
 then the entire system operates at pressures lower than ambient.  This is
 termed "negative" static presure (static pressure is simply the pressure
 exerted by a  gas in all directions, measured normal to the direction of
 flow, if any).  With the fan in this position, ambient air will leak into
 the ductwork  and scrubber if there are any holes, open access hatches, or
 weld gaps.           ' •
      In Figjire 2-2b,,i;the fan,, is located between| the pnacgss equipment arid-

theiscrubSer^Vessel > Thil
        JQ TO       \

                              __  ..    w~__- — —~ — - — — -— — --—-"^—--—
                             typ.e of  sEftjbber  system operates at positive
U>

            (greater  thaji  ^gt&ierttj!.
.1
                                                          in this type of
                                                           for fan placement
^system^contaminated  gas will
  is  a  push-pull  arrangement with  fans  placed before and after the scrubber.
  This  arrangement  is  often used when the  scrubber has been added to an
  existing  control  system.
       The  pressure drop across the  scrubber vessel is an  important opera-
  ting  variable  since  it is often  related  to the  effectiveness of particulate
  capture.   The  pressure drop  is simply the mathematical difference between
  the static pressures before  and  after the scrubber.  Both of the examples
  shown in  Figure 2-2  have a pressure drop of 14.5 inches  water.
       The  liquor flow,, through the  scrubber can  either be "once-through1'
  or  recalculated.   Most wet scrubber systems have a recircu'iating liquor
  system to reduce  the volume  of water  needed,  and to reduce  the cost  of
  treating  the scrubber effluent liquor.   Both  liquor systems are illustrated
  in  Figure 2-3.
       Obviously the recirculating liquor  circuit is more  complicated  than
  the once-through  system.  The addditional components required include the
  recirculation  tank,  the  purge stream  controls,  and the makeup stream
  controls.  The recirculating system is prone  to buildup  of  solids and
  corrosive agents, while  the  once-through system is free  of  these potential
  problems  as long  as  there  is a  sufficient supply of high quality water.
  The advantages of the recirculating liquor  system  include lower operating
  cost (in  most situations)  and the  opportunity to neutralize and treat the
  liquor prior to entry to the scrubber vessel.
                                     2-3

-------
PROCESS
SOURCE
                                       FAN
                                               P=I0.5
                               AMBIENT STATIC PRESSURE=
                       VESSEL
          Figure 2-2a.  Negative pressure system.
PROCESS
SOURCE
FAN
                 SCRUBBER
                  VESSEL
          Figure 2-2b.  Positive pressure system.
                           2-4

-------
               SUPPLY
                PUMP
            FRESH
            LIQUOR
              GAS-
                                GAS, 50,000 ACFM
                                      @  130° F
                           _J
                       TREAMENT AND
                           DISPOSAL
                                              33FIU08
       Figure 2-3a.  Once-through liquor system.
                          t GAS, 5O.OOO ACFM
                                  @130° F
     GAS $
30,000 ACFM
   @3006F
O
           I
        _J
               FLOWMETER
                            RECIRCULATION
                               TANK
       RECIRCULATION
           PUMP
                                      MAKE-UP STREAM
                    PURGE STREAM
       Figure 2-3b.  Recirculating  liquor system.
                        2-5

-------
      A useful measure of the quantity of liquor being used by the scrubber
 system is the liquid-to-gas ratio which is normally abbreviated  as L/6.
 This ratio is usually (but not always) defined as the total  volume (in
 gallons) of liquor entering the scrubber divided by the outlet gas flow
 rate (in thousands of ACFM).   The outlet gas flow rate is  used as  the
 basis of the parameter since  this easier to measure than the inlet gas
 flow rate.  For the scrubber  illustrated in Figure 2-3a, the liquid-to-gas
 ratio is 750 gallons per minutetidiyjjded £y( 50;X; 10^,ACF per^jnijiute^or 15.
      As indicated earlier,  a  recirculating liquor system must have  some
      ftf ^mpa^¥iTuW/%tbfage^%^h2as^tFe«' re^t^Wn ^WnMlo^in
 £1gtfre" 2-3bv~¥^drfd^ b^rhlfps'^a to\r sfetfeloh*%%nhfeHb1;W Wf W B
 sdhjKbkr^esse^
 location  for the additionof  neutralizing  agents,  surfactants, and/or
 anti-foaming agents.   Often the pond  or tank  is  the site of makeup water
 addition  and/or  drawing  off of  a purge  stream.
      Because of  the  buildup of  solids  in the  scrubbing  liquor and the need
 to  minimize  the  concentration of chlorides  and dissolved compounds, a
 portion of the recirculation  liquor is  often drawn off  for disposal or
 further treatment.  This  purge  stream usually consists of 2% to 5% of the
 total recirculation  stream  rate.  In scrubber systems without a purge
 stream, all  of the liquor within  the system is replaced on a regular
 basis.  Depending on the  frequency of replacement, the liquor quality
 varies considerably.
     To maintain the desired recirculation flow rate, sufficient  liquor
 is  added to  (1) account for that removed in the purge stream and  (2) make
 up  the liquor lost to evaporation and that lost with the wet sludge (if
 any).  This makeup stream is usually fresh process water, well water,  or
municipal water.  The makeup stream can be added at a number of points  in
the scrubber system with the most common point of addition  being  the
recirculation tank or pond.
     If the temperature of the gas stream entering the scrubber is  very
hot (greater than 300° F), the gas stream is often cooled.prior to  entry
                                2-6

-------
 to the scrubber.   Cooling  the  gas  stream  protects  the scrubber vessel
 materials  of construction,  especially  the corrosion and abrasion resistant
 linings.   If the  cooling is done by  the evaporation of liquor, the increased
 gas stream humidity  also enhances  the  particle capture mechanisms.
      The gas stream  can be cooled  in an evaporative cooler or in a pre-
 saturator  or quencher.  The latter is  simply  a small chamber immediately
 before the scrubber  in  which clean liquor is  sprayed.  Because of the
 short residence time within the  presaturator  or  quencher, many of the
 droplets do not have time  to evaporate completely; therefore, more than
r^tejn» ,%iaJ1£i£y pfi'l*w?r~3s::
 a .W WWfSSQtf^^rt^
 inear the^top.  The,use  of  .nozzles  optimizes;droplet evaporation,.duetto,:
 U'UL-M 33 C£?U Iff JW i,^ v-''"'•"* ''"** . ***** '•'- 	 ;     	" '  " •
 smaller water droplet sizes created.  At  least 85% of the liquor injected
 into the evaporative coolers evaporates;  therefore, the effluent liquor
 stream is quite small.   The locations  of  evaporative coolers and presatur-
 ators in a wet scrubber system are shown  in Figures 2-4a and b.
      Alkaline material  is  often added  to  the  scrubber liquor circuit in
 order to maintain the pH  in a  range in which  corrosion  is not a problem.
 The materials most frequently  used include, limestone,  dolomite, soda
 ash, and sodium hydroxide.  The rate of  feed  of  the  alkaline material  is
 controlled by a pH meter within the scrubber  system, as  shown in Figure
 2-5.  Another means of adding  a neutralizing  material  is  illustrated in
 Figure 2-1.  When lime is used, a slaker is also required.
      The liquor treatment system reduces the  total solids content  of the
 liquor prior to return of the liquor to  the recirculation circuit.   Com-
 mon  equipment  for this purpose includes   a clarifier  and a  rotary  vacumn
 filter.  Flocculant is added to the clarifier to improve  the  settling
 characteristics of  the solids.  The reduction in solids content reduces
 the  potential  for pump impeller erosion, nozzle erosion,  and  pipe  pluggage.
 The  clear  effluent  from the clarifier is returned to the main  liquor
 circuit either at the  recirculation tank or at the scrubber itself.
                                    2-7

-------
  Figure 2-4a.  Scrubber system with evaporative cooler.
PROCESS
SOURCE   PRESATURATOR
t
                              SCRUBBER
                                                       FAN
      PRESATURATOR
        LIQUID      RECIRCULATOR
                       TANK
    Figure 2-45.  Scrubber system with presaturator.
                      2-8

-------
                           GAS STREAM
                                                               ALKALINE
                                                               MATERIAL
                                                                  BIN
                                                    MIXING TANK
 RECIRCULATOR
     PUMP
RECtRCULATION
    TANK
                  PUMP
                                     PURGE
                    Figure 2-5.  Neutralization system.

     Many scrubber systems incorporate  a bypass duct tp protect the
process equipment in the  case  of a major upset and the scrubber in the
event of the loss of liquid flow.  A  set of dampers are used to prevent
inadvertent leakage of untreated gas  through the duct.
     Flow of liquor through the  system  is maintained by pumps.  The most
common type of pump in wet scrubber systems is the centrifugal pump.  If
the liquor contains abrasive or  corrosive materials, the pumps are fitted
with protective liners.
2.2  IMPORTANCE OF PARTICLE SIZE DISTRIBUTION
     The performance of all  types of  particulate wet scrubbers is highly
dependent on the inlet particle  size  distribution.  The importance of
the particle size distribution is a result of the strong particle size
dependence of the fundametal particle capture mechanisms used in scrubbers,
namely, inertia!  impaction and to a lesser extent, Brownian diffusion.
                                  2-9

-------
     Unfortunately, there are substantial differences in particle size
distribution on a pi ant-by-plant basis.  The particle size distribution
can also have temporal variations at a specific scrubber system under
different process operating conditions or due to process upsets. In some
cases, the operating conditions of the scrubber itself can modify the
inlet particle size distribution.  Particle size is difficult to measure
during routine evaluations and, therefore, less direct indications of a
shift in particle size must be used.
2.2.1  Particle Capture Mechanisms
     The;Rey; step .invol ved";ih :wet scrubbing,:is the capture,of "thetparticle«
into eithersliquidiidropTetsfiTsheetSiVor'jetSii The principal phys,icaT^3DOS"t one
mechanism used in commercially available systems isMnertial impaction.cr ::.'r&s1T
Other physical phenomena aiding capture include:  Brownian diffusion,
diffusiophoresis, and thermophoresis.  Electrostatic attraction may
contribute to capture in certain cases such as venturi scrubbers where
liquid atomization can result in static electrical charge buildup.  These
physical capture mechanisms are briefly reviewed to illustrate the sensitive
relationship between scrubber performance and particle size.
     2.2.1.1 Inertia! Impaction.  Entrained particles have a much greater
mass and, therefore, a much greater inertia when in motion,  than the sur-
rounding gas.  As a gas stream approaches an obstacle, the gas molecules
pass on either side on it, leaving the particle propelled toward the
obstacle by its inertia.
     The typical droplet sizes in wet scrubbers range from 100 to 800
micrometers,^>3 while the particle sizes of most interest are in the
0.1 to 10 micrometer range.  There is normally a large difference in the
initial relative velocity between the droplet and the particle.  The
effectiveness of impaction is proportional  to the Stokes Number
defined in Equation 2-1.
                 KT =
C V "
 18 v. DC
                                                         Equation 2-1
                                   2-10

-------
where:
          K! = Stokes Number, dimensionless
           C = Cunningham Slip Correction Factor, dimensionless
          d  = particle diameter, mA (g/cnr)0'5
                                     o
          p  = particle density, g/cnr
           V = relative velocity between particle and droplet
          Dc = diameter of droplet or collector water layer
 _  -^ -.-    M = P,as JLl^-c4^-t^--v^
     The Stokes Number is proportional to the particle diameter squared.
This relationship means that as the particle diameter increases, the ease
of impaction increases substantially.  Impaction-4s~ directly-r-e~l-ated™to-    '-°Ji
the pafJjie'1'eqderisity,swh.Jqb3cainyvaryptfroiTiijQ;2 g/
-------
 of  the  distance from the surface leads to an imbalance in the molecular
 collisions  on  a particle close to the surface of the droplet.  This
 imbalance causes the particle to migrate toward the site of the condensa-
 tion.   The  opposite  effect occurs when the droplet is undergoing evaporation.
     2.2.1.4 Thermophoresis.   When there is a distinct temperature gradient,
 there is an imbalance between the forces exerted on the sides of a particle
 by  the  molecular collisons.   The molecules on the higher temperature side
 have a  greate.r  ye,lpcity^a,nd thereby transfer more mqmen.tum to the particle
                         . "~g '   " • "  '- * *e  '—i'*  *(>  -n jH'MtV, *f"7MM'•*  t~} \ . f t^j ,j> * , j i AU ,si i I •
when colliding  than  those molecules on the colder side.  The  result is a
net movement toward  the  colder temperatures.
     jon z~  norjsijsnsq no norauttfo nernwoiS	-TO *aa't^9. fel'O'f^snad sriT   ,9sl~a
                                            fe Rjaytt-iclas'&igei  airfl rfr J'nsisqqB
                                       ^^^V.f^W^^^^^.
etration curve as a  function  of  the  particle  size  generally appears  as
shown in Figure 2-6.  The peak penetration  corresponds  to the 0.2 to 0.5
?'1?/^":'-,:F0 fll .'W
           1.0
         DIFFUSION
           MORE
         EFFECTIVE
                                I I I 11
                                    I  i  i i 11
      O
         o.io
      UJ
      •z.
      UJ
      a.
          O.OI
                                   IMF-ACTION MORE
                                     EFFECTIVE
                     I    i   i  i  i i  11
            O.I    0.2      0.5    1.0    234  56789IO
                        PARTICLE DIAMETER,
   Figure 2-6.
 Hypothetical  curve  illustrating general  relationship
    between  penetration  and particle size.
                                   2-12

-------
micrometer size range, in which inertia! impaction is relatively ineffective
and the particles are too large for significant diffusional capture.
     The emissions from a given scrubber system depend primarily on the
quantity of parti cul ate matter in the size range from 0.1 to 5 micrometers
and the penetration level achievable for the particle sizes.  The latter
is generally considered to be a function of the total power input into
the scrubber and the type of scrubber itself.
   '"'•  tlie:|iartiCle: s!i'z^':d1ependehce frf venturi scr:ubbers: is il 1 ustrated' in
Figure 2-7.  The highest penetration is for particles of 0.5 micrometers in
                              ,.  .."nu.^'tawriaJ  'teDiOD srtj tnswp;? tnjamevqm tan
size.  The beneficial effect of Brbwnia'n diffusion on penetration is  not
apparent in this ft^e^fDMosjK^^                                    s>i%tfi#
determine mass concentrations as the total mass with diameters less than a
     ^ -.-,.. -, e, -.     »pt  • j-f. T-. •         _          _, ......... :.„.-.,.,-   . . '.< ..... '*.
            ''"han 0.2

                                               If the Affect of Brownian
diffusion had been included, the curve shown in Figure 2-7 would begin
          1.0
     §
     cc
     Lu
     Q
     5
     or
     5
     UJ
     0.
     UJ
     o
p
         0.01
1 — r~n MIII
T	1  I  I  I I II-
                                            '    '   i  ' * '  ' i
                                    1.0
                         PARTICLE DIAMETER,
                                        10
      Figure 2-7.   Fractional  efficiency curve for venturi  scrubbers.7
                                    2-13

-------
to  resemble the curve presented in Figure 2-6.  A number of similar
particle  size-penetration curves are presented in references 3, 4, and 5.
     On a  given scrubber system a shift in the particle size distribution
in  the gas stream has a major impact on the performance.  The can be
illustrated by use of the venturi scrubber model described in references
6 through  9.  For a hypothetical set of operating conditions, the model was
used to calculate the penetration.  The mass mean particle size was
varied from 5 to 20 micrometers and a standard-deviatri on of,2v5-.was,-used
                            ^       „.-.-,
in  all three cases.  The results are shown in Figure 2-8.  The predicted
                      ibyr a-.fjacfcorr o,fi 3s,whenf«thaj sjzBijdlfetnSbiifclm shifts
                     5          10          15          20
                   GEOMETRIC MASS MEAN DIAMETER, microns

Figure 2-8.  Predicted relationship between overall penetration
                 and particulate mass mean diameter.
                                   2-14

-------
2.2.3  Factors That Modify Particle Size
     Numerous process-related operating factors and upset conditions
affect the particle size distribution at the scrubber inlet.  Thus a wet
scrubber system cannot be considered as a separate entity operating
independently of the process it serves.
     2.2.3.1 Process Operational Change.  Any change that modifies the
quantity of particulate and vaporous materials evolved from the process
can cause a decrease; in the particle size distribution.  For example, an
increase in the sodium content of the lime sludge feed to a lime kiln at
a, kraf t Kpulp: mi 3 1 woul du subs tan ti a! Ty; i"n:cheases the Jqu DtSM *ov.£i
micrometer size particulate matter
the
                                      ! '           I           '••
plants can lead to a shift in the quantity of organic material  volatilized
and thereby increase the quantity of of very fine hydrocarbon aerosol
particulate which must be removed.
     The indications of a change in the process operation may include
some of the following:  a major increase or decrease in the gas tempera-
ture entering the scrubber, a change in the types of raw materials (to
the extent this can be determined), an increase in the residual opacity
of the plume downstream from the scrubber (bluish white material), and a
change in the fuel characteristics.  As many symptoms as possible should
be considered together to reliably identify process changes as responsible
for a particular shift in the particle size disbribution.
     2.2.3.2 Evaporative Release.  Many commercial wet scrubbers are
preceded by a quench tower which serves to cool the gas stream to the
125° to 200°F temperature range prior to its entry into the scrubber.
This is necessary to protect scrubber internal parts and may also improve
particle capture.
     The liquor quality used in the quench tower (or presaturator) varies
from source to source depending on the types of spray nozzles used in the
quench tower and the need to recycle liquor to minimize treatment costs.
The total solids content can sometimes be as high as 15% by weight.
Since a major portion of the liquor injected into the quench tower and/or
presaturator evaporates, some of the solids entering as suspended and
dissolved solids in the liquor are released as particulate matter.
                                   2-15

-------
     Kalika has presented some calculated effects of evaporative release
based on conditions observed at an incinerator scrubber.10 The results
are provided in Table 2-1.  In preparing these data, he assumed an inciner-
ator exit temperature of 1600°F with a humidity of 0.08 pound water per
pound of dry air.  He also assumed that 85% of the quench water droplets
evaporate to dryness.


           TABLE 2-1.  EFFECT OF RECIRCULATION LIQUOR SOLIDS10
Solids
in Recirculated
Water, %
»•
2
5
10
Water
Recirculated
to Quench, Gpm
0
22
25
26
Outlet
Dust Loading,
•Lb./Min.
0.032
0.114
0.259
0.512
Apparent
Scrubber
Efficiency, %
98.0
92.9
83.9
68.2
     As the solids content of the liquor is increased to 10% by weight, the
emissions from the scrubber are predicted to increase by more than an
order-of magnitude.  These data provide only a rough guide to the potential
effect; the actual effect will depend on the droplet size distribution
in the quench tower, the gas temperature at the quench tower inlet, the
liquid-to-gas ratio at the quench tower, the residence time within the
quench tower, and the manner in which the droplets shatter during the later
stages of evaporation.  A droplet must be completely evaporated before
the suspended and dissolved particulate matter are released.
     At the present time, the extent to which evaporative release con-
tributes to additional particulate matter emissions is not well  documented.
Considerable study will be necessary to determine the evaporative con-
ditions within a quench tower and the effect of the released material on
the particle size distribution.
     Indirect indications .of potential  or actual  evaporative release in-
clude:  high turbidity liquor going to the quench tower, frequent pluggage
problems with the quench tower nozzles, high gas inlet temperatures, and
high residual opacity.
                                   2-16

-------
     2.2.3.3 Vapor Condensation.  Some industrial  sources generate organic
and/or metallic vapor in addition to partial!ate matter.   If these vaporous
materials begin to condense in a homogeneous manner while passing through
the scrubber, very little of this material  will  be'captured.  To the
extent possible, the condensation process should be initiated as far
upstream of the scrubber as possible to enhance  heterogenous condensation
on the surfaces of existing aerosols.
     Symptoms of increased vapor condensation aerosol  include a change in
the process operating temperature, a change in the use of raw materials
and/or corrosion inhibitors, and increased residual opacity plume downwind
of the stack exit.
2.3  PRESSURE DROP AS A PERFORMANCE PARAMETER
     The static pressure drop of the gas stream  passing through a particu-
late wet scrubber is used extensively by both operators and inspectors to
evaluate scrubber performance.   In some cases,  conclusions regarding
adequacy of performance have been based solely on these data.  The uses and
limits of pressure drop as a performance parameter are examined in this
section.
2.3.1  Definition of Pressure Drop
     Pressure drop is the reduction in potential energy (through conversion
to kinetic energy) of the gas stream as it passes from one point to
another.  It results from frictional losses on the ductwork and scrubber
internal components, from acceleration of the gas and  liquid streams,
from the atomization of the liquor, and from any changes  in elevation of
the gas and liquid streams.
     The scrubber pressure drop is simply the arithmetic  difference
batween the static pressures of the gas stream at the  inlet and at the
outlet of the collector.  If there are two collectors  in  series, the
total pressure drop across the system is the sum of• the two individual
pressure drops.  The additive nature of static pressure drop (P^)  is
illustrated in Equation 2-2.
Pt =
               P2
Equation 2-2
                                   2-17

-------
where:
              Pt = total pressure drop of system
              Pj[ = pressure drop across unit 1
              ?2 = pressure drop across unit 2
              Pn = pressure drop across unit n
The same concept applies to a single scrubber with multiple trays.   The
pressure drop across the scrubber is the sum of the pressure drops  across
each tray plus the pressure drop across the demister.
     If there are two scrubbers in parallel on a given effluent stream, as
shown in Figure 2-9, the pressure drop across the system is equivalent to
the pressure drop across either unit.  In other words, the pressure drop
along each path must be identical. The equivalent nature of static  pressure
along parallel paths is illustrated in Equation 2-3.
     Pt = P2 - Pi = P4 -
Equation 2-3
where:
              P-t = total pressure drop of system
              ?l = static pressure at inlet of path A
              ?2 - static pressure at outlet of path A
              ?3 = static pressure at inlet of path B
              P4 = static pressure at outlet of path B
     While  the  pressure drops across each path must be identical, the
 pressure drops  across each  scrubber may not be identical.  If one of the
 paths  has longer ductwork,  more elbows, or a flow restriction such as a
 damper or solids deposit, then the pressure drop of the scrubber on this
 path will be  less than that in the other path.  This difference is illus-
 trated in Figure 2-10.
     With certain types of  venturi scrubbers, there is static pressure
 recovery in the divergent section; in other words, the pressure drop
 across the  throat is greater than the pressure drop across the entire
 venturi section (convergent section - throat - divergent  section).  This
 occurs in units where the divergent section expansion angle is equal to
                                    2-18

-------
                  Figure 2-9.  Parallel scrubber trains.
Figure 2-10.  Parallel scrubber trains with unequal duct resistance.

or less than 7.5 degrees as shown in Figure 2-11.  With units of this
type, boundry layer separation probably does not occur.H  Pressure
recovery can also occur in scrubbers with a slightly greater expansion
angle; however, the extent of the gain would be slight.  Most commercial
designs have a divergent angle ($ )  of greater than 25° and many are  in
the 35° to 45° range.  A few types do not even include a divergent
section to the venturi; pressure recovery in these would be negligible.
     In order to properly define the pressure drop, it is very important
to specify the locations where the "inlet" and "outlet" measurements  are
made.  A sketch of the system showing the scrubber, all major internal
parts, and flow restrictions of the  inlet and outlet ductwork is helpful
in avoiding misinterpretation of the pressure drop value.
                                   2-19

-------
          Figure 2-11.  Converging and diverging venturi  angles.

2.3.2  Relationship Between Penetration and Pressure Drop
     There is a useful relationship between the collection efficiency of
a particulate wet scrubber and the pressure drop of the gas stream passing
through it.  For this reason, operators have been using pressure  drop as
an indicator of scrubber performance for a long time.
     Tn 1956, Kemack and Lapple proposed that penetration is a function
of the total energy consumption within the scrubber, not just the energy
loss represented by the gas phase pressure drop.12  This theory was
expanded and stated in its present form by Semrau in I960.13  (See also
references 14-16.)  Referred to as the Contact Power Theory, it has
enjoyed broad acceptance by both operators and regulatory agencies.
Since much of the available data are presented in the Contact Power format,
the basis of this theory is briefly examined.
     According  to the Contact Power Theory, scrubber efficiency is directly
related to the  total  energy consumption of the system.  The power can be
expended by the gas stream, the liquid stream, by a mechanical rotor, or
by a combination of all three.  Other variables  such as collector size,
scrubber design characteristics,  liquid-to-gas ratio, liquor surface
tension, and gas velocity  are assumed to  have no independent effect on
the  scrubber performance.  Accordingly, it  should be possible to  quantify
the  emissions  from  a  specific wet scrubber  system using only those para-
meters with  an effect on  the  power input.
                                    2-20

-------
     The Contact Power Theory has also been used to compile  industry-by-
industry mass emission-contact power curves.   In this section,  the adequacy
of these unit specific and the industry wide applications of the theory
are evaluated .
     The Contact Power equations commonly presented in the literature are
reproduced below.  These apply to all particulate wet scrubbers.*
     PQ = 0.158 AP
     PL = 0.583(QL/QG)
     PT = PG + PL
      H = 1 - e
      P = 1 - ( i?/100)
                                                  Equation 2-6
                                                  Equation 2-7
                                                  Equation 2-8
                                                  Equation 2-9
                                                  Equation 2-10
                                                  Equation 2-11
where:
 AP
   P
   77
a,
  Pt
  PL
  PQ
                 gas phase pressure drop, inches w.c.
                 penetration
                 collection efficiency, percent
                 number of transfer units, dimension! ess
                 empirical constants, dimension! ess
                 total power input, hp/1000 ACF
                 power input from liquid stream, hp/1000 ACF
                 power input from gas stream, hp/1000 ACF
                 liquid-to-gas  ratio, gal/1000 ACF
 The constants  a and  0 are functions of the particle size and particle
 characteristics.   Equations 2-10 and 2-11 simply indicate the relationship
 between transfer units, efficiency, and penetration.
      The derivation of these equations, which was only briefly described
 by Semrau,14 was apparently based on a mechanical energy balance  across
 the scrubber.   To  fully evaluate the adequacy of these equations  and the
 means to apply the Contact Power approach,  the  derivation was reconstructed.
 It is presented below.  The starting point  is the basic Bernoulli  Equation
 for incompressible flow.
 *An additional term  for  shaft  power would be  required if a mechanical
  rotor were included  in  the scrubber.
                                    2-21

-------
     P! - P21       g F       1      faiVi2  - a?Vo2]
     _L	£ G  + — Zi  -  Z2  G  + P-i	±J-
       PG   J       ScL1     2J     I    2 gc     J
                                                  Wp -
                                L  +
                                       V  2
                                       PVP
                                        2V2
                                          2 gc
                                              L  +
     G = 0
 Equation 2-12


*) 2gc Equation 2-14 2-22


-------
                                               p"p
                                                   2gc
                                                           Equation 2-15
                                                           Equation 2-16
     PT  =
                                                           Equation 2-17
Since the liquor pressure is zero psig,  AP|_ is equivalent to P|_.
     PT     APG
                           (L/G)
Equation 2-18
      The left hand term, Py/G is identical to the contact power form
of horsepower per 1000 ACF. To convert the gas phase energy term,
 PG/PG» to tne fom of Equation 2-6, the gas density at 70°F,  14.7 psia
must be inserted.  This reconstructed derivation is intended to demonstrate
a very important point, namely that total energy input is a function of
gas density.  Too often Equation 2-6 has been used without questioning
the applicability of the "constant," 0.158.  The actual  gas density can
vary substantially as a function of the inlet gas temperature, the scrub-
bing liquor inlet temperature, the liquid-to-gas ratio,  and the outlet
static pressure^
     As stated earlier, the form of the Bernoulli Equation used in the
derivation of the Contact Power equation applies only to incompressible
flow.  Since it is conceivable that the gas density varies more than 10%
while passing through some types of scrubbers, these equations would not
                                   2-23

-------
       be strictly applicable.17  Gas density data  at various humidities,  tem-
       peratures, and static pressures are presented in the Appendix A to
       illustrate the degree of variability possible.
            The presssure drop term in Equation  2-18 applies only to the throat
       in a venturi  scrubber.  If there is some  degree of pressure recovery in the
       divergent section of the venturi, the observed pressure drop will be lower
       than that used in this equation.
            Despite  the theoretical limits of the Contact Power Theory, there
       appears to be a number of cases in which  it  provides reasonable cor-
       relations between performance and energy  input.  A number of correlations
       originally compiled by Semrau are provided in Figures 2-12, 2-13, 2-14,
       and 2-15.13  The gas phase pressure drop  is  probably the dominant energy
       loss in these correlations.
    6.0

    4.0
   .3.0

t   2.0
       6
       z
       cr
       K
       CO
       u.
       o
       DC
       UJ
       m
       S
 1.0
0.8
0.6

0.4
0.3

0.2
          SPRAY NOZZLE
             LOCATION
          O    CYCLONE
          •    CYCLONE
          A  INLET DUCT
          	I
            SPRAY
          DIRECTION
   DOWNWARD OVER INLET
HORIZONTAL, ACROSS INLET
     COUNTER CURRENT
                        I   i  i  i  i  i i
    _L
J_
111!
                   0.3  O.4   0.6 0.8 1.0       2.0   3.0 4.0
                       CONTACTING POWER, HP/(IOOOcuft/min)
                                                       6.0  8.0
Figure 2-12.  Contact power curve for  a  cyclone scrubber controlling talc dust.
                                        2-24

-------
      All  of these  data  are  from very early studies and were obtained with
 test methods that  are not as accurate as current methods.   However, the
 data do support the  contention that emissions corelations  can be  devel-
 oped for a group of  similar sources using only the power input and  possibly
 just the pressure  drop.
      Figure 2-15 is  presented specifically to demonstrate that the  liquid-
 to-gas ratio does  not have  any independent effect.13  Semrau qualifies
 the statement, however,  by  suggesting that some effect may occur  at very
 low liquid-to-gas  ratios of 1 to 3 gallons per 1000 ACFM.18  No data
 in this range was  apparently available when the data in Figure 2-15 was
 originally compiled.                           ,
       8
       7
    tr
     u.
     o
     OL
     UJ
     00

                                      AND MELTING
                      SCRAP MELTING, HOT METAL WORKING
                      WORKING
                           2          3      45   6   7   8  9 10
                       CONTACTING POWER, HP/dOOO cu ft/sec)
Figure 2-13.   Contact  power curve for venturi scrubbers on  open  hearths.13
                                    2-25

-------
    3.0
  H-
 Z

 ff
 Z

 u
    2.0
   0.8

   0.6
I
£ 0.4

O 03
£C  '

i 0.2
 z     —
    01
           I I I
-D
                      I     I   I   I  I  I  I I
                                        IITIIIFT
                                           cfia
                                • VENTURI SCRUBBER REF.

                                D CYCLONIC  SPRAY SCRUBBER
                   I	I
                          I  I  I I I I
                                                       l   tit
      0.06   0.1      02  0.3 04   0.6 0.8 1.0      2.0   3.0 4.0  6.0 8.0 100

                    CONTACTING POWER, HP/(IOOO cu ft/min)

   Figure 2-14.  Contact  power curve scrubbers controlling  ferrosilicon
                     furnace fume.13
  IO.O

<  40
DC
H-
DC
   2.0
   1.0
                           I   I   I  I  M I
                                                          I   I  I  I  I  L
                 I      I    I   I  i  I I  I I
                                                    LEGEND

                                                   (L/G,gal/IOOOft3)
                                                     olO

                                                     A 15

                                                     •20

                                                I      I    1   I   I  I  I I
                         4    6   8  10       20
                     EFFECTIVE FRICTION LOSS, inches
                                                          40   60  80
  Figure  2-15.
              Contact  power curve for venturi scrubbers  illustrating
              lack of  independent effect  of the liquid-to-gas ratio. 18
                                   2-26

-------
     Walker and Hall19 have compiled a contact power correlation for
flooded disc venturi scrubbers serving four commercial  lime kilns.  Data
from a pilot plant were also included in the data set.   The resulting
emissions-power input correlation presented in Figure 2-16 suggests a
linear relationship  between the pressure drop (the only significant energy
input is the gas  phase) and the measured emissions.  To further  evaluate
its usefulness, the  curve  has been replotted (see Figure 2-17)  into a
more conventional format and the confidence interval prepared (using the
procedures presented in Appendix B).
                                            T	1	1	1—
                                            •  PLANT A
                                            A  PLANTS
                                            A  PLANTC
                                            O  PLANTD
                                             •  PILOT PLANT
   20
Q.
O
O  10
LU
ir
Z>
en
UJ
£   5
               T	1
                        A A
                                            j.
                                                           J	L
          0.02
                            0.05     0.10      0.20
                              EMISSIONS, grains/SCF
0.50
  Figure 2-16.
                    Relationship between  pressure drop and emissions for
                      flooded disc scrubbers  on lime kilns.
                                    2-27

-------
     It is readily apparent in the modified curve of Figure 2-17 that
there is considerable scatter in this correlation.   For  example, at a
pressure drop of 9 inches, the emissions can vary from 0.075 grain per.
ACF to as high as 0.20 grain per ACF.  Obviously, the degree of scatter
would preclude application of this type of curve  for the purposes of
evaluating compliance with an emission standard.   During the preparation
of this curve, it was not possible to determine whether  individual site-
specific correlations would have less scatter, because the raw data was
not included in reference 19.
         o.io
      co" 0.08
      g 0.06
      tn
      | 0.04
      LU
          0.01
Figure.2-17.
 I      2        3    10     20  30
          PRESSURE DROP, inches

Replotted Walker and Hall  data;  90% confidence interval is
      denoted by shaded area.
     Kashdan and Ranade have summarized  emissions data and contact power
levels (uncorrected for saturated  gas  density) for pulverized coal and
lignite-fired  utility boilers  20.   The  modest degree of scatter in
Figure 2-18 is due in part to the  inclusion of data from various types of
scrubber systems incuding the Krebs  Preformed scrubber, the UOP Turbulent
                                  2-28

-------
Contact Absorber,  and several major types of venturi scrubbers.  Considering
the diversity of scrubber  designs and the major differences in the fuel
characteristics, it is remarkable that any correlation is apparent.
                                2        3      4    56789
                 THEORETICAL POWER CONSUMPTION, hp/1000 ACFM

Figure 2-18.  Pressure drop  versus emission for wet scrubbers serving coal-
                   fired utility boilers. 20

     A subset of Kashdan and Ranade's data was used to prepare a contact
power curve for power plants using only  venturi-type scrubbers (see
Figure 2-19).  It is apparent that the scatter is substantially reduced.
The range of variables for this subset of data is presented along with
the curve to demonstrate that even with  this small data group, there is
considerable variation.   Additional  testing and evaluation is necessary
to confirm that a curve representing all types of coal-fired boilers with
venturi scrubbers can be prepared.   Data from Kansas City Power's LaCygne
Station are considerbly higher than  the  other points.
                                   2-29

-------
       _
      o
      v>
.050
.045
.040

.035

.030
      I
      co~ .025,
      O
      CO
      CO
      5 .020
      tu
         .015
                     5       10      15      20      25
                          PRESSURE DROP, inches
Figure 2-19.   Replotted  data of Kashdan and Rariade using  only  scrubbers;
                   90% confidence interval is denoted by  shaded  area.

     Genoble,  et  al. have collected a number of emission-contact power
correlations,  two  of which are shown in Figures 2-20 and  2-21.I6  In both
of these cases,  there is considerable scatter in the data.  The  scatter
in the data set  for the  coal preparation plant thermal  dryers  has been
partially attributed to  differences introduced by the various  organizations
conducting these tests.
                                  2-30

-------
                  .20
                  .
                
                cn"
                z
                g
                en
                en
                LU
                  .02
                                       10        20
                            PRESSURE  DROP, inches
 Figure 2-20.
Pressure drop versus emissions  data  for  venturi scrubbers
    serving asphalt plants.16
     The data scatter in the asphalt concrete plant curve, the thermal
dryer curve, and many of the curves  presented earlier, is evidence of the
weakness of industry wide correlations.  These correlations inherently
cannot take into account the numerous important site-specific variables
which include, but are not limited to, particle size distribution, particle
size distribution, particle surface  characteristics, liquid-to-gas ratio,
gas-liquor maldistribution, and liquor droplet size.  Correlations prepared
on a site-specific basis should be less  vulnerable to these variables and
hence, should be more useful  for evaluating  scrubber performance.
                              2-31

-------
            .08
            .07
            .06
          Q
            .04
          to" .03
          z
          o
          en
          CO
          UJ
            .02
             .01
                10
20      30   40  50 607080
PRESSURE DROP, inches
 Figure 2-21.  Pressure drop versus emissions  data  for venturi scrubbers
                   serving coal  thermal  dryers.16

     Site-specific data are very limited since few  operating wet scrubber
systems have been tested enough  times  for an emissions - contact power
correlation to be prepared.  Available data are usually limited to pilot
scale equipment.
     Pilot plant wet scrubber data for the Minnesota Power and Light
system have been reported by Johnson and also  by Nixon and Johnson.21
The scrubber tested handled a 3200 ACFM  slip stream from the Clay Boswell
Station, Unit No. 3.  Emission tests conducted during periods when two
different coal  supplies were being burned are  summarized in Figure 2-22.
The data for the scrubber are very consistent,  despite the slight differ-
ences apparently introduced by the coal  supplies.
                                     2-32

-------
o
    0.04
0.03
o
en

in
c

s
o>

cT

5
<

3
?  0.02
<
cc
O.Oi
                                          MCKAY COAL
               ROSEBUD COAL
                     _L
                             _L
                                            _L
                                                    -L
                     5           IO          15         2O


                        VENTURI  PRESSURE DROP, inches water
                                                                25
     Figure 2-22.   Relationship between pressure drop  and emissions for a pilot

                    scale  venturi scrubber on a coal-fired boiler slip stream.



          A contact power  relationship for a full  scale Q-Basic .Oxygen Process


     (Q-BOP) furnace is  presented in Figure 2-23.   It  was prepared from a


     tabulated data set  provided by Kemner and Mcllvaine.22  The data are for


     three identical vessels controlled by adjustable  throat Baumco venturies


     in series with venturi prequenchers.


          The values shown on  Figure 2-23 represent the average pressure drops


     during the period of  5 to 11 minutes into the oxygen blow.  The pressure
                                       2-33

-------
drop variation during cycle is due to the rate of gas evolution.  The
pressure drop values have not been corrected for gas density  since the
necessary temperature data were not included.
     The correlation between the average pressure drop and  the outlet
dust loading has very little scatter.  This may be partially  due to the
identical design and operating conditions of the three vessels.  A site-
specific relationship of this type would be adequate for an emissions
correlation.  The relationship, however, wuold probably not be applicable
to other basic oxygen furnaces because of the  significant differences
made by the various types of hooding and process conditions on the particle
characteristics.
     Since performance is affected by so many  variables, from a number of
different scrubbers should not be grouped into a single correlation.  It
would be preferable to use the contact power approach  only on a site-
specific basis.   Even in this case,  it would be necessary to confirm that
there have been  no major operating changes that would  invalidate the mass
emission-power correlation.                                  •,
     o
     u.
     o
       .04
     1 .03
     s
     o>
     CD"
     §.02
     CD
     CD
                  J	I	L
              60     62     64     66     68      70
                PRESSURE  DROP, inches (minutes 5-11)
 Figure  2-23.
Pressure drop versus emissions  data  for venturi scrubbers
  serving three Q-BOP furnaces  (  plotted using data of
       Kemner and Mcllvaine).22
                                  2-34

-------
2.4  PLUME AND BREECHING OPACITIES AS PERFORMANCE PARAMETERS
     Regulatory agencies and control equipment operators routinely use
plume opacity as an indicator of the particulate matter removal  performance.
In the case of wet scrubbers, however, the use of opacity has been limited
by the practical problems created by the condensation of water droplets
i n the piume.
     Theoretically, opacity could be a meaningful operating indicator
since particles that scatter visible light most effectively are in the
range of 0.2 to 2 micrometers and this coincides with the peak penetration
range for the particles treated in  a wet scrubber system (as was shown in
Figure 2-6.)  Thus, as the penetration through the scrubber system in-
creases  in this important size range, the opacity should also increase
substantially.
     Mass emission-opacity correlations have rarely been developed based
on manual observations.  This is  partially due to the lack of sufficient
emission test data and partially  due  to the difficulty  in making these
observations.   For example,  in compiling  these correlations it is impor-
tant that the observational  path  length through  the plume remain the  same.
This is  possible  only when the condensation of water droplets occurs  at
a  point  downstream of the stack.  Under these somewhat  unusual conditions,
 it is  possible  to make  the observation of opacity at a  point directly
 above  the stack,  where  tne path length is known.  Observation of the
 residual opacity  of  a plume  downwind  of the stack is not as useful since
 the path length at this point can be  highly variable due to meteorological
 conditions  and  dilution of  the  plume  becomes  a  factor.
      As  early as  1963,  an extractive  instrument was  tested  as a means to
 evaluate the stack opacity without interference from water  condensation.23
 The device  consisted of a photocell mounted in  a flow-through gas cell.
 The source of light  was a  simple  flashlight bulb.  The sample was extract-
 ed through  a heated  sample  line to prevent condensation, and  the  instrument
 was calibrated using neutral density filters.   While successful performance
 was claimed, it is doubtful  that this instrument could satisfy  the  specifi-
 cation  requirements  of present day transmissometers.
                                    2-35

-------
       In some cases, a conventional  transmissometer  has been successfully
  utilized as a continuous monitor.  Very often  condensation of water drop-
  lets downsteam of the scrubber does  not occur  in  the ductwork.  Therefore,
  a transmissometer installed in these locations should perform satisfactor-
  ily.  A study by Nixon and Johnson at the Minnesota  Power and Light plant
  using a pilot scale scrubber system  included the  evaluation of the plume
  opacity as a function of the mass concentration in the effluent.21  The
  opacity was determined using a Lear  Siegler RM41  transmissometer mounted
  along a 6'6" section of the outlet duct.   As seen in Figure 2-24, the study
  showed a linear relationship between the  mass  concentration and the opacity.
  This supports the logical  presumption that opacity should be a useful  indi-
  cator of mass emissions when the opacity  can be accurately determined.
      More recent test work has been  done  using a heated extractive
  sample line similar in concept to this  early instrument.24  The opacity
 monitor in this case, however, is a  Nephleometer which measures light
  scatter.  Test work done at a kraft  pulp  mill  recovery boiler equipped
    30
    20
Q x
w fc
05 O
GQ UI


ll'l
^8_  8

      6
      5
o to
0. O
                I
                          I
I
I
I
      .005    .010      .015     .020     .025     .030
                   OUTLET GRAIN LOADING, grains/SCFD
                                                           .035
                                 .040
  Figure  2-24.  Opacity versus emissions data for a venturi  scrubber on a
                         coal-fired boiler slip stream.2^
                                   2-36

-------
with a venturi  scrubber was  encouraging.  With certain refinements, this
approach may soon provide  a  convenient means to continuously monitor the
performance of wet scrubbers.
2.5  LIQUID-TO-GAS RATIOS  AS PERFORMANCE PARAMETERS
     The liquid-to-gas  ratio has  a  significant impact on the penetration
through a wet scrubber.  However, it  is questionable whether the effect of
the liquid-to-gas ratio1needs  to  be considered independently of the
pressure drop,  because,  for  most  types of wet scrubbers, the pressure
drop is linearly related to  the liquid-to-gas ratio.7»8
     The possible effect of  the liquid-to-gas ratio has been examined
using the venturi scrubber analytical model discussed earlier.  The pene-
tration was calculated  for a number of combinations of gas velocities and
liquid-to-gas ratios.   Then  the penetration and corresponding pressure
drop data was plotted using  the liquid-to-gas ratio as a parameter.  The
results are illustrated in Figure 2-25.
     O.IO
    O.O8
    0.06
  ui
  Q_
    0.04
    0.02
               LIQUID-TO-GAS RATIO,
                 gals/IOOOACF
                20     4O     60     80     100
                     PRESSURE DROP, centimeters
                                          120
Figure 2-25.
Predicted effect of liquid-to-gas ratio on penetration as
   calculated using the  venturi  scrubber model.
                                   2-37

-------
     Based on the model predictions, the optimum liquid-to-gas ratio is
between 8 and 12 gallons per thousand ACFM.  At this rate,  the penetration
is lowest at a given pressure drop as illustrated by the dotted line at
40 centimeters of water.  There is only a slight difference in performance
as the liquid-to-gas ratio varies over the range of 6 to 20 gallons  per
thousand ACFM.  The model suggests that there is essentially no independent
effect over this range, which is the most common operating  range of  com-
mercial wet scrubber systems.
     At very low liquid-to-gas ratios, the venturi  scrubber model  predicts  .
that there is an effect which is independent of the pressure drop.   The
4 gallon per thousand ACFM curve is significantly higher than the other
liquid-to-gas ratio curves.  This prediction is consistent  with the
observations of Semrau and others during the testing of operating systems.^
At low liquid-to-gas ratios, it would be necessary  to monitor the liquid
flow rate since a change of plus or minus 1 gallon  per thousand ACFM
would have a major impact on scrubber performance.
     As normally used, the liquid-to-gas ratio applies to the entire scrub-
ber.  In other words, the liquid-to-gas ratio is the total  recirculation
liquor flow divided by the total scrubber outlet gas flow rate.  Unfortu-
nately, the distribution of the liquor across the gas stream is not  entirely
uniform; therefore, there can be significant variation in the local  liquid-
to-gas ratios.  This variation can have a significant impact on the  penetra-
tion without a proportional impact on the pressure  drop. It is very diffi-
cult to identify the maldistribution problems.
     Poor gas-liquid distribution is often the result of improper treat-
ment of the recirculation liquor.  Excessive suspended solids can lead to
pluggage of the spray nozzles, the main liquor pipes, and the strainer
before the pump.  In some spray tower scrubbers, as many as half of  the
nozzles have been found to be plugged by suspended  solids in the liquor.
In venturi scrubbers and traytype scrubbers, scaling of the scrubber
internals can also result in serious gas-liquor maldistribution.   To
minimize scaling, it is important to maintain a proper pH (generally
below 10) and a proper purge/blowdown rate.
     Buildup of material in the wet-dry interface can lead  to maldistri-
bution, too.  To prevent buildup of material in this location, the gas
approach must be wetted.
                                      2-38

-------
2.6  EFFECT OF LIQUOR SURFACE TENSION
     The surface tension of the scrubbing liquor can potentially affect
the penetration rate whenever impaction on droplets is an important unit
mechanism.  This is true for gas atomized scrubbers such as Venturis and
orifice units, impingement plate and sieve plate scrubbers, mechanically
aided scrubbers, and spray tower scrubbers.  Low surface tension reduces
the liquor droplet size and improves coalescence (the inclusion of the
particle into the droplet).  In one study by Hesketh, it was found that
lowering the surface tension from 50-60 dynes per centimeter to 10 dynes
per centimeter reduced the penetration by 50%.25
     The impact of reduced droplet size can be predicted using the ven-
turi scrubber analytical model  developed by Calvert et alJ The data
presented in Table 2-2 was computed by Woffinden, Markowski, and Ensor^^
using the Calvert Model and the conditions of an 8000 cm/sec throat
velocity and a monodisperse aerosol of 1 /imA.  The model  suggests no
improvement if the droplet diameter without surfactant is below approxi-
mately 50 micrometers.  This is not a commercially important case since
the large majority of nozzles used in wet scrubbers produce sprays in
the 200 to 800 micrometer range. 27,28
  TABLE 2-2.  PREDICTED  EFFECT OF SURFACE TENSION (a)  ON THE PENETRATION
                       THROUGH A VENTURI SCRUBBER 26
Droplet
a= 72 dynes/cm
205
144
51
Diameter
a= 30 dynes/cm
144
102
36
Penetration
a= 72 dynes/cm a= 30
0.020
6.020
0.020
dynes/cm
0.011
0.013
0.021
     The effect of improved particle coalescence into a droplet is  diffi-
cult to estimate.  Presumably this would benefit sources with marginally
wettable particulate ;matter.
     Liquor surface tension will  vary according to the quantities of mater-
ials absorbed from the gas stream, the quantities of surfactant added,  and
                                   2-39

-------
the  rates of the purge and makeup streams.  Chemicals added to the liquor
to improve the rate of settling of solids in the scrubber clarifier and
recirculation tank can have an impact opposite to that for the surfactant.
The  influence of foaming suppressants on the surface tension of the liquor
is not known.  Due to the potential variability of the surface tension and
the  moderate effect this may have on scrubber performance, the quantities
of all surfactants, flocculants, and anti-foam agents should be recorded
during each stack test and during subsequent evaluation periods.
     If it is necessary to quantify the surface tension, ASTM Method
D 1590-60 can be used.29  This technique is based on the ring method of
measurement and has a precision of 0.3 dyne per centimeter.29  Most com-
mercial laboratories have the necessary equipment for this test procedure.
2.7  CONDENSATION AND EVAPORATION EFFECTS ON PERFORMANCE
     Impaction can be enhanced by the condensation of water vapor on the
surfaces of particles and, to a lesser degree, by the condensation of water
vapor on existing water droplets.  Evaporation of water vapor from existing
droplets has an adverse effect on scrubber performance.
     The beneficial effect of condensation is primarily due to the in-
creased mass present on a particle.  The increased inertia of the particle
yields higher impaction efficiency.  A secondary benefit results from
the mass gradiant which exists whenever condensation is occurring.  The
imbalance in the momentum imparted by the water vapor molecules on each
side of the particle results in a modest movement toward the surface of
the condensation site.30  This phenomenon is termed diffusiophoresis.  The
negative effect of evaporation is the opposite of that described above.
     The quantity of water available for the inadvertent condensation is
usually quite limited.   This can be estimated by use of psychrometric
chart (actual  physical  conditions are dependent on the prevailing gas
pressures).   It is usually in the range of 0.01 to 0.15 pound of water
per pound of dry air.
     Calvert and Jhaveri  have presented information indicating that the
quantities necessary to achieve substantial  improvement in scrubber per-
formance are in the range of 0.2 to 1.0 pound of water per pound of dry
air.31  This is illustrated in Figure 2-26.
                                   2-40

-------
          0.4
          0.2
          O.I
                          _L
I
I
                         O.I    0.2   O.3    0.4   0.5
               GRAMS H20 CONDENSED PER GRAM DRY AIR


            Figure 2-26.  Beneficial  effect  of  condensation.31

     There are few commercially operating  scrubbers which have a quantity
of water vapor condensing that is  this  large.   In  order to induce this phe-
nomenon, it is possible  to evaporate  a  large quantity of water in a high
temperature zone of the  process duct  work  or to inject large quantities of
low pressure steam.  Cooling  of the inlet  liquor stream would also favor
condensation.  Lemon has presented operating data  on a novel scrubber system
which used artificial means to induce condensation.32  This system was also
able to achieve compliance with all applicable  requirements at a pressure
drop of only 20 inches w.c.,  while the  earlier  scrubber on this source re-
quired a pressure drop of 40  inches w.c. for a  similar penetration rate.
     At the moderate water vapor levels present in most effluent streams,
the presence of condensation  would have little  effect.  The plume appear-
ance would provide an initial  indication of  a dramatic change in the water
vapor content of the efflue'nt stream. Other parameters which would be
useful indicators include the gas  stream inlet temperature and the
liquor stream inlet temperature.
                                   2-41

-------

-------
           3.0  FACTORS AFFECTING SCRUBBER SYSTEM RELIABILITY

     The performance of a scrubber system is dependent on  the  operating
characteristics discussed in Section 2.0 and on the integrity  of the
scrubber internals.  The latter is addressed in this section.   Failure
of these components can lead to extended downtime of the system or
operation in temporary noncompliance.  Both the operator and the regu-
latory agency inspector have an interest in minimizing these problems
by identifying the emergence of factors which threaten scrubber compo-
nents and ultimately degrade scrubber performance.
3.1  CORROSION AND EROSION OF SCRUBBER SHELL
     The susceptibility of a scrubber to erosion is a direct function
of the gas velocities within the entry ductwork and the scrubber shell.
The venturi scrubber throat is most prone to these problems due to the
extreme gas velocities of 20,000 to 40,000 feet per minute at  this
point.  Other areas of common erosion failure are illustrated  in
Figure 3-1.
     Regulatory agency inspectors usually do not have the opportunity
or safety equipment to perform internal inspections of control  equipment,
therefore, they must use the less direct, external symptoms of erosion.
This includes the obvious holes worn through the shell at critical
high velocity points or points of gas stream turning.  In extreme cases,
these holes can lead to severe inleakage of ambient air and a  reduction
of the quantity of gas pulled from the process..  The extent of the air
infiltration can be quantified using pi tot tubes before and after the
collector (and accounting for absorption and condensation/evaporation
within the scrubber).
     A valuable sign of potential erosion problems is the quantity of
suspended solids in the liquor.  The potential for erosion is  directly
related to the percent suspended solids.  The performance of the clari-
fier and/or the settling pond can also have an impact on the susceptibil-
ity to erosion in that the large, abrasive particles should settle out.
The greater the recirculation flow rate relative to the make-up and purge
flows, the greater the potential for build-up of high suspended solids.
                                  3-1

-------
    Figure 3-1.  Areas of a venturi  scrubber vulnerable  to  erosion.

Also, systems which purge only once  per day or once per  week  can  develop
high suspended solids levels late in the operating period.  A rough
check on the quantity and character  of the suspended solids can be made
by obtaining a sample of the liquor  leaving the scrubber sump and
observing the initial turbidity and  the rate of settling.
     Due to the absorption of corrosive materials  such as sulfur  dioxide,
sulfur trioxide, hydrogen chloride,  and hydrogen fluoride,  corrosion of
the scrubber shell and ancillary equipment is a common problem.   It is
                                  3-2

-------
important to maintain the pH of the liquor well  above the levels  at which
carbon steel (the most common material  used in scrubber construction)  is
attacked.  A pH of 6 or greater is usually satisfactory.   An appropriate
pH is usually maintained using alkaline additives such as soda  ash, lime
or limestone.  The performance of the alkaline additive delivery  system
and the process control instrumentation are very important in minimizing
short term low pH excursions.  The operation of the pH monitor  used to
control the rate of additive injection should be checked on at  least a
daily basis unless long term operating experience justifies less  frequent
inspection.  A portable pH meter or pH paper can be used for this check.
The sample should be collected at the sump of the scrubber since  this is
often the point of minimum pH.
     The recirculation rate is an important factor in determining the
extent to which halogenated compounds are building up in the scrubbing
liquor.  Relatively small levels of chlorides and fluorides can attack
many types of materials, especially the 300 series stainless steels.
     One convenient means to monitor potential corrosion problems is to
prepare small coupons  (small circular samples) of the various materials
used throughout the scrubber system.  These are placed in racks which
can be mounted at various locations in the scrubber.  During every
outage these are visually inspected for pitting and cracking and are
weighed for material loss.  This information provides an early indication
of developing corrosion problems.
3.2  EROSION AND PLUGGAGE OF SPRAY NOZZLES
     Spray  nozzles  are extremely susceptible to erosion  and pluggage
problems due to the high velocities of the liquid stream and due to the
suspended  solids within the  stream.  The most common  types of nozzles
in use include the  hollow cone and the full cone.  The latter is particu-
larly  prone to pluggage due  to the presence of  an internal spinner vane.
The  vane is  installed  to achieve the full  cone  spray  pattern which is
necessary  for  distribution  of  liquor on a  packed bed.
     Damage to  the  nozzles  can sometimes be determined by observing the
spray  angle while  the  nozzles  are  operating at  normal  line pressures.
                                    3-3

-------
(This can usually be done safely while the system is  down  and  only  the
pumps are operating.) If the spray angle is considerably greater  than pre-
viously, then it is very possible that the nozzle orifice  has  been  enlarged
by erosion.  If the spray angle has decreased or if a distinct spray pat-
tern is no longer achieved, it is likely that the nozzle is  partially or com-
pletely plugged.  Both conditions lead to severe gas-liquid  maldistribution.
     Erosion problems on the tangential  inlet to the  entrainment  separa-
                          \
tor are common to many types of scrubbers  because of the  concentration
of large, abrasive particulates in the outer part of  the inlet duct.  The
venturi-rod type scrubber is also prone to erosion of the  rods within the
throat, hence these must be checked on a routine basis.
     Venturi scrubbers in which the throat is oriented vertically may have
erosion/abrasion failure at the bottom of the divergent section as  the gas
stream turns 90° to enter the entrainment separator.   One  means of  minimiz-
ing this problem is to include, below the venturi throat,  a  flooded elbow
containing 6" to 9" of trapped liquor.
     Erosion can best be identified during a routine  internal  inspection
of the scrubber shell, entrainment separator, entry ductwork,  and other
such components.  The erosion of corrosion protective linings  such  as
rubber or synthetic coatings is particularly important since gaps and
crevices in the lining can lead to rapid corrosion underneath  the lining.
The entire lining can be defeated by erosion of a small  area in an  ero-
sion prone zone of the scrubber.
     The visual check of spray angle usually cannot be done  by a  regula-
tory agency inspector or an operator while the entire system is operating.
Some indirect indications of the nozzle condition can be used  in  lieu of
these observations.  If a nozzle has plugged significantly,  the pipe
leading to that nozzle in the scrubber begins to cool.  Comparision of
the skin temperature of one pipe against that of the  pipes leading  to
other nozzles is a reasonably reliable indicator of pluggage.   It is
common to find a skin temperature 5 to 10 degrees below that of adjacent
pipes.  Pluggage of a nozzle will also be indicated by an  increase  in
the line pressure near the nozzle.  Erosion of the nozzle  orifice yields
the opposite conditions.  Unfortunately, line pressure is  difficult to
interpret since it is strongly dependent on the liquid flow rate.  The
                                   3-4

-------
relationship between pressure and flow rate (for a nozzle in original
condition) is presented in Equation 4-1.28
          _Ll_
           t-2
                                                           Equation 3-1
where:
                LI
                l_2
                P!
                ?2
= liquid flow rate at pressure 1
= liquid flow rate at pressure 2
— rvv»Q e* o 11 v»a 1
  pressure 1
  pressure 2
     During a routine inspection it is difficult to determine whether a
change in line pressure is due to a change in the internal condition of
the nozzle or simply due to a change in the liquid flow rate.
     If possible, all nozzles for venturi scrubbers should have external
rod-out capability.  This is usually not feasible for spray tower scrubbers
and the nozzles used to clean demisters.  In these cases, an automatic
flush system may be of benefit.  If a pond is used for settling, it is ad-
visible that it have several zones separated by weirs and the pick-up for
the pump be near the water's surface to prevent entrainment of silt.
3.3  FANS
     Fans can either be installed before or after the scrubber.  Only
radial blade designs are used ahead of the scrubber due to the large
quantities of suspended particulate in the gas stream and the fact that
other fan wheel designs are prone to fan wheel particulate buildup.
Fans installed following the scrubber should be the radial blade type.
     The principal problems of fans on wet scrubber systems include fan
wheel erosion, fan wheel buildup, and bearing failure.33  Corrosion is
also a common problem.  Both fan wheel erosion and fan wheel buildup
will eventually result in increased vibration which is usually audibly
and visually detectable.  For systems prone to this problem, vibration
sensors are advisable, enabling a source to have the fan automatically
tripped if the vibration reaches undesirable levels.  If during a routine
inspection of a scrubber system, excessive fan vibration is noticed,
extreme caution and immediate action is necessary.  Disintegrating fans
can fling metal parts over a wide area.
                                   3-5

-------
     Fan bearings should be inspected on a frequent basis for oil  level,
oil color, oil temperature, and vibration.  Bearings found to be inade-
quately lubricated may require a greater frequency of lubrication  or the
installation of a forced lubrication system.  Excessive bearing wear may
be caused by a higher than originally specified fan operating temperature
or by misalignment of the bearing mountings.
     Fan belts should be checked periodically for wear.  At the time of
installation, belts should be checked for proper tension.  Loose belts
can cause the entire dust collector system to malfunction.  Out-of-line
sheaves will destroy belts; uneven wear on grooves of sheaves and  the
surfaces of the belts indicates misalignment.
     Fan vibration can sometimes be caused by air flow factors and,  in
these cases, can be eliminated by adjusting the inlet or outlet dampers,
modifying the inlet or outlet dampers, modifying the inlet and/or  outlet
ductwork, or changing the sheaves.  The latter will have a direct  affect
on the gas flow rate and will result in a change in the tip speed  (which
should never exceed the manufacturer's recommended rate).
     Operating problems experienced by fans downstream of scrubbers  are
generally only symptoms of more fundamental problems with the particle
capture and/or entrainment separation.
3.4  PUMPS
     The two types of pumps in service on wet scrubber systems are the
centrifugal and positive displacement pumps.  The centrifugal pump is
the most common.
     Accelerated wear of the pump impeller occurs at high suspended  sol-
ids levels.  As is the case with spray nozzles, the best way to minimize
scrubber downtime due to pump malfunctions, is to minimize the total
quantity of suspended solids and especially the larger suspended particles
(greater than 25 micrometers).  A strainer ahead of the pump will  improve
the quality of liquor passing through and thereby minimize wear.
     Another common problem with pumps is leakage through the packing
glands; pumps must be repacked whenever necessary.  This can be mini-
mized through the use of doub.le seals with a clean water, pressurizing
line.  All bearings on the pumps and the pump couplings should be  lubricat-
ed periodically in accordance with the manufacturer's instructions.

                                   3-6

-------
3.5  PIPING AND VALVES
     All piping is subject to erosion, especially at elbows.  The rate
of erosion is a function of the flow velocity, the quantity of suspended
solids, the particle size distribution of the suspended solids, and the
presence of corrosive agents.  The proper materials of construction
must be selected in order to minimize this problem.
     Valves are also subject to erosion.  Those which are operated in
either a fully open or closed position are affected much less than those
use.d in a partially open mode for flow throttling purposes.
     In order to reduce pluggage of lines during outages, the piping
should be completely drained.  This requires that each line have a modest
slope so that a pocket of liquid cannot remain after draining.  The drain
should be at the lowest point of each piping system.  Draining is also
necessary to prevent the freezing of lines during outages.
     During routine inspections, one should be aware of the piping design
with regard to drainage capability.  If it appears inadequate, and the
systems runs in a cyclic nature, occasional  pluggage and/or freezing prob-
lems should be anticipated.   Periods of noncompliance can occur initially
after start-up of the system and prior to the rectification of a piping
freezing/pluggage problem.  In such cases, the pump discharge pressure
will  be high, the nozzle pressure will  be low, and the gas exit tempera-
ture from the scrubber will  be high. The gas phase pressure drop across
the scrubber will  be low due to the lack of typical  liquor flow rates.
3.6  DUCTS
     The most common problems with the ducts leading to a scrubber include
dust buildup, erosion/abrasion, flex failure, and failure of expansion
joints.   The last three of these conditions  result in air infiltration
and consequently,  a reduction in the quantity of gas pulled from the
process  equipment.   The extent of air infiltration can be quantified by
preparing an oxygen profile  at various points along the duct from the
source to the scrubber.      .   ,
     Dust buildup occurs in  low gas velocity zones of the duct work.
This  can occur wherever the  duct work is oversize for the effluent gas
flow rate and during periods of operation at reduced process rates.   Clean
                                   3-7

-------
out ports should be located at convenient locations along the ductwork '
to facilitate occassional removal  of accumulated material.   Without this
cleaning the duct work can collapse (unless the supports have been  de-
signed to take the combined weight of the duct and the deposits).
                                   3-8

-------
           4.0  SCRUBBER INSPECTION AND PERFORMANCE  EVALUATION


     The purpose of a wet scrubber performance evaluation  is  to  determine

whether the penetration has increased or decreased since the  previous  eval-

uation and to identify any conditions which threaten the proper  operation
of the scrubber system.  The performance evaluation  must incorporate the

operating variables and factors presented earlier in Sections 2.0  and  3.0.
A brief summary of these are presented below in Table 4-1.

             TABLE 4-1.  PERFORMANCE EVALUATION PARAMETERS
                         Penetration Variables

              Inlet and Outlet Static Pressure
              Demister Pressure Drop
              Bypass Duct Pressure Drop
              Gas Inlet and Outlet Temperatures
              Liquid Inlet and Outlet Temperatures
              Gas Inlet 02 Levels
              Gas Inlet COg Levels
              Liquid Suspended Solids and Dissolved Solids
              Liquid Surface Tension
              Bypass Duct Temperature Profile
              Rate of Liquor Evaporation and Condensation
              Condensible Vapor Levels in Inlet Gas
              Particle Size Distribution
              Absolute Humidity of Inlet Gas

                           Operating Factors

              Liquor pH
              Liquor Suspended and Dissolved Solids
              Liquor Supply Line Pressure
              Nozzle Spray Pattern            ,
              Fan Vibration
              Integrity of Ductwork	

The list of variables is long and some of them  cannot  be measured  or
observed directly.  The performance evaluation  approach must  include  some

indirect means to consider these difficult-to-measure  factors.

     The routine performance evaluation is structured  to aquire  only  the

minimum data necessary to detect a change in wet scrubber performance.
When abnormal conditions are found, the remainder of the measurements and
other inspection procedures are performed to determine the  nature  of  the
specific problem.

                                   4-1

-------
 4.1   BASELINE PERFORMANCE EVALUATION TECHNIQUE
       The  routine evaluation of wet scrubbers must be done in a logical
 sequence  in order to obtain all the necessary information and to minimize
 the time  requirements.  The technique recommended is the Baseline Inspection
 Technique.34  This approach is based on the assumption that the operating
 characteristics and performance of each control system is unique.  The
 approach  is designed to control for the myriad of process variables and
 control device design factors, any one of which can singly or collectively
 influence performance.                                 ...  ,  .
      To ensure the accuracy of the performance evaluation, the Baseline
 Inspection Technique utilizes a comparison of present operating conditions
 with historical  baseline levels for the unit.   Each variable which  has
 shifted significantly is considered a "symptom"  of possible  operating  prob-
 lems.   While compiling the operating  data, the general  condition  of the
 wet scrubber is  observed to  identify  problems  which could  affect  future
 reliability of the  system.   The baseline data  are  compiled during periods
 when the wet scrubber is operating well,  preferably during a compliance
 or acceptable  stack  test.
     The evaluation  is  accomplished by  proceeding  in a  counterflow manner
 through the entire system; the stack  and  plume opacities are observed  first,
 and then the scrubber system is  inspected  component  by  component  proceeding
 backwards  through the system in  the direction opposite  to  the flow of  the
 gas stream.  This is  illustrated in Figure 4-1.
     The time consuming  inspection  of process equipment and  operating  con-
 ditions  is done by the agency  inspector only when, deemed-necessary based on
 data collected during inspection of the scrubber and observation of the stack
 opacity.   The advantages of this approach  include the following:  (1) poten-
 tially confidential  process information is obtained only when absolutely
 necessary,  (2) the inconveniecnce to the operator is minimized, and (3) on-
 site time  is minimized.  Since the inspection data are organized in a coherent
manner as  the inspection proceeds,  the effort can gradually be focused on
only the problems of direct interest.   In other words, the intensity of the
evaluation can be minimized and many of the measurements can  be avoided.
This is a particularly important point considering  the extensive list of
operating variables  and reliability factors presented earlier.   The  typical
order of inspection  for a routine evaluation is presented  in  Table 4-2.
                                   4-2

-------
                                        CONTINUOUS
                                          MONITOR
            WASTES
                             CONCENTRATED
                              POLLUTANTS
        Figure 4-1.  Counterflow inspection sequence.
            TABLE 4-2.   ROUTINE INSPECTION"POINTS
1,  Stack Discharge

2.  Fan
3.  Demi ster
4.  Scrubber
5.  Bypass Duct
6.  Ductwork

7.  Process
Plume Opacity at the Stack Discharge
Presence or Absence of Entrainment
Presence or Absence of Condensed Water
Presence or Absence of "Mud" Lip
Vibration (Qualitative Check)
Pressure Drop
Pressure Drop
pH
Liquor Line Skin Temperatures
Turbidity
Continuous or Interim'ttant Purge
Use of Anti-Foaming Agents and Surfactants
Damper Position
Holes and Leaks
Integrity of Expansion Joints
Hood Capture Effectiveness (Qualitative)
                             4-3

-------
     Depending on the size of the scrubber system and the convenience of
the measurement ports, the routine inspection should take between 1  and 4
hours.  This is a reasonable time commitment considering that this type
of evaluation can aid in the early identification of operating problems
that might ultimately result in extended periods of noncompliance and
expensive repairs.
4.1.1  File Review
     Prior to inspecting a scrubber system, an agency inspector should
review the agency files to familiarize himself with all  pertinent infor-
mation on the particular scrubber, types of emissions,  and operating
history.  The following items should be checked:
     1.  Pending compliance schedules and variances
     2.  Construction and/or operating permits pertaining
         to source processes
     3.  Past conditions of noncompliance
     4.  Malfunctions reports
     5.  History of abnormal operations
     The inspector should know what regulations are applicable to the
facility; and if he has any questions, he should consult the  agency
administration or legal  staff prior to conducting the inspection.   It is
important to know what data are useful in confirming compliance and  what
evidence is necessary to document possible violations of each of the
applicable regulations.
     The inspector should also refer to plant layout drawings to fami-
liarize himself with emission point locations, to draw flow diagrams,
and to prepare the inspection report.  If available in the files,  the
inspector should note the facility's personal  safety equipment requirements.
     Based on his review of agency files, the inspector  should schedule an
inspection time when plant processes will be operating at representative
conditions.  The scheduling of a time to visit plants having  batch opera-
tions or other irregular operating schedules (e.g., seasonal)  is especially
important.  The files should always be reviewed briefly  before entry to
the plant so that important plant characteristics will  be more easily
remembered.
                                   4-4

-------
4.1.2  Evaluation of Plume Opacity
     The initial  step in any scrubber inspection is to observe the visible
emissions exiting the stack under representative operating conditions and
under weather conditions which permit opacity observations.  The visible
emission observation, while more difficult than on other control  devices
because of steam plume interferences, can provide a good indication of
the scrubber's performance.
     In order to compare the opacity against baseline levels, the opacity
must be observed at the point of stack discharge where the pathlength
is fixed.  Very often this is not possible due to the presence of light
scattering condensed water droplets.  If this is the case, the opacity
of the residual plume should be observed after the point of water droplet
dissipation.  The latter provides only a very rough check on the overall
performance of the scrubber.
     Experienced observers are able to distinguish between a plume
obscured by water droplets and a plume free of water droplets.  Parti-
culate matter in a plume usually has a bluish-white or brownish-white
color.  Water vapor, on the other hand, is usually a brilliant white color.
A plume containing water droplets also appears to have more texture and
is more wispy in nature than a plume free of water droplets.  To confirm
the difference between the plume being observed and a water droplet
plume, it is usually possible to simultaneously note the behavior of a
nearby steam vent plume.
     After the plume observation is completed, ,the discharge from any
bypass stacks (if separate) should be observed.  A general check should
also be made for fugitive emissions from process sources.
     During these initial observations, a check should be made for any
entrained droplet fallout from the plume.  The presence of dried spots on
adjacent equipment and in the surrounding area is an indication of demister
or entrainment separator problems.  When the entrainment of solids is
severe, a lip of mud forms around the stack mouth and the  stack is
discolored several feet down from the opening.
                                   4-5

-------
 4.1.3   Fan  Evaluation
     The location  of the  fan  relative  to  the  scrubber system should be
 ascertained.   If the fan  is ahead  of the  scrubber, then the remainder of
 the  scrubber  system and ductwork is under positive pressure.  Leaks in
 the  gas  handling system or an open measurement port will allow the gas to
 escape  into the areas  immediately  around  the  scrubber.  Personnel conduct-
 ing  the  performance evaluation should  avoid any partially confined areas
 where leaking gas  could result in  high concentrations of toxic agents.
     If  the fan is located downstream of  the  scrubber, then the shell
 and  the  inlet ductwork is under negative  pressure.  In this case, any
 leaks or open measurement ports will result in inleakage of ambient air.
 This presents less of a personnel exposure problem, however, it should
 not  be assumed that pockets of toxic contaminants are not present in
 some partially confined areas.  Fans on the downstream side are vulnerable
 to buildup of material  on the fan wheel.  This can lead to imbalance of
 the wheel, particularly if the fan tip speed  is high.  If this remains
 uncorrected,  severe vibration and eventual disintegration of the fan is
 possible.  Personnel evaluating scrubber operation should leave the vi-
 cinity of a "severly vibrating fan immediately; operators of such systems
 should take immediate action to take the fan off line.
     The  inspector should ask about the recent operating history of the fan
 with respect  to particulate buildup, vibration, bearing failure, or cor-
 rosion.  These are indications of potential  carryover from the demister.
 4.1.4  Demister Pressure Drop  "
     If there  is entrained solids fallout from .the plume or vibration  of
 the induced draft fan,  the operation of the demister should be checked.
The upstream and downstream static pressures should be measured and com-
 pared against  baseline levels.  A significant increase in the  pressure
 drop usually means partial pluggage of the demister,  while pressure drop
 values much lower than  baseline values can signal  that there are openings
 through the demister beds.
     It is generally advisable to have several static pressure  ports
around the scrubber shell  both before  and after the demister.   In  this way
the upstream and downstream static pressures can  be averaged.
                                   4-6

-------
     The measurements may first be made using a portable slack  tube type
manometer or a Minihelic® differential  pressure gauge.   Due to  the  low
pressure drops involved, however, it may be necessary to repeat the
measurements using an inclined manometer or, at least,  a low range  Magna-
helic® gauge.  The Magnahelic® and Minihelic® gauges should be  calibrated
regularly.35
     The presence or absence of a demister flush system should  be noted.
This often consists of an array of spray nozzles above and/or below the
demister.  Use of the sprays on a once per shift and a once per day basis
substantially reduces the potential for demister pluggage.   The agency
inspector should inquire about the cleaning system operating schedule.
     If an inspector is being conducted when the scrubber is down,  it may
be possible to visually check for buildup on the demister.   (Care is
necessary to avoid exposure to toxic gases which may be trapped within
the scrubber vessel.)
4.1.5  Scrubber Pressure Drop
     4.1.5.1  Measurement of Pressure Drop.  The penetration of parti-
culate matter through the scrubber is partially a function of the pressure
drop.  Thus, measurement of the pressure drop is an essential step in the
routine performance evaluation.  Unfortunately, the permanent on-site
monitors are not always reliable because of the potential for line pluggage.
For this reason, it is usually advisable to measure the pressure drop
using portable gauges.
     Locating proper measurement ports is the first step in measuring the
pressure drop.  The fundamental rule in locating and using measurement
ports is only those that are safely accessible should be used.   Access to
ports in some scrubber systems may involve close contact with hot ducts,
unsafe climbing, and/or exposures to fugitive gas leaks.  Heroic efforts
should not be taken to reach ports which have been incorrectly  located.
A  second rule is ports must be in a location relatively free of inter-
ference  from sludge and entrained water droplets.  Recommended  locations
for ports in tray-type scrubber and venturi scrubber systems are shown in
Figure 4-2.
                                   4-7

-------
     It is important that the fan and the fan dampers jiot be included in
the system between the two measurement ports.  Inclusion of the demister
(or cyclonic entrainment separator in venturi scrubbers) is also not
ideal; however,  it must often be included because  the region between the
venturi throat and the demister has a high flow rate with high concen-
trations of entrained water droplets and other materials, making measure-
ments of static  pressure difficult.  Inclusion of  the demister usually
adds only 0.5 to 2 inche's of pressure drop (if there are no deposits),
only a small fraction of the venturi throat pressure drop.
                    GAS INLET
                                                         PREFERRED
                                                         OUTLET PORT
  PREFERRED
  INLET PORT
                                      \\\\\\\\\\\
                                     DEMISTER
                                                   *-«—*- PURGE
       Figue 4-2.   Preferred  static pressure measurement locations.
                                  4-8

-------
     Once the appropriate measurement ports have been located, the
measurement work can begin.  However, before opening the port, the
operating pressure within the duct or scrubber shell should be
considered; it,can either be roughly estimated or previous records and
measurements can be reviewed.  If the system operates under positive
pressure, the opening of the port can result in the unintentional
fumigation of the area, in which case everyone present should be prepared
with the necessary personal protective equipment.  If the system operates
under negative pressure, then it is necessary to be aware of the pos-
sibility that air infiltration through the port can result in measurement
errors by the aspiration effect.  This is illustrated in Figure 4-3.
NEGATIVE
PRESSURE
DUCT

-20"

                                 -25"
                               AIR/INFILTRATION
                                            COPPER PROBE
                                         RUBBER STOPPER
  Figure 4-3. , Aspiration effect error in static pressure measurement.

     Suction induced by air infiltration around the probe leads to
measured static pressures which are less (more negative) than actual.
The magnitude of the error can be as great as 10 to 15 inches w.c. for
some venturi scrubbers.  Even tray-type units operating in the 6 to 12
inch pressure drop range, can have aspiration errors of 1 to 2 inches.
One way to minimize the aspiration effect in the measurement of static
pressure is to use a probe inserted well into the duct or scrubber
shell.  This is illustrated in Figures 4-4a and b.
                                  4-9

-------
                              DUCT WALL
                                   COPPER TUBE
                                       RUBBER STOPPER
                                                  TUBING
                                       I"*"!   pr-i	*U
                                         ELECTRICAL
                                         BONDING WIRE
                                       GROUNDING CLAMP
                               GROUNDING TAB
Figure  4-4a.  Measurement of static pressure using a copper tube.
                             -DUCT
                                 S TYPE PITOT TUBE
                                       RUBBER STOPPER
                                                  TUBING
                                                   •J
                                              -GROUNDING
                                               CLAMP
                                         ELECTRICAL
                                         BONDING WIRE
                                      -GROUNDING CLAMP
                              -GROUNDING TAB
 Figure 4-4b.  Measurement of static pressure using a pitot  tube.
                             4-10

-------
     Figure 4-4a illustrates the use of a 1/4" O.D. copper tube, while
Figure 4-4b illustrates use of the pitot tube.  Both approaches are very
effective in avoiding problems with aspiration; however, both involve the
risk of water droplet and solids impaction on the probe.  The measurement
port must usually be at least 3/4" in diameter to use the S-type pitot
tube.
     The higher the negative static pressure is at a given location, the
smaller the measurement port should be in order to ensure that the port
can be effectively sealed during measurement.  If it is necessary to use
an oversized port, care, must be taken to insure a good seal.   The use of
gloves is not recommended since they can be sucked into the gas stream.
On the same account, sampling probes can also be lost.
     One quite useful  assembly for oversized ports is illustrated in
Figure 4-5.  This consists of a 1/4" O.D. copper tube which is inserted
through a #4 rubber stopper and then a commercial  sanding disk (with a
1/4" O.D. hole).  The sanding disk serves as an effective flange to
prevent air infiltration and the rubber stopper holds the copper tube.
This assembly has been used with no difficulties on measurement ports
serving ducts at up to 100" negative pressure.  The sanding disk should
be at least 1" greater in diameter than the port on which it  is used.
                                      SANDING DISK
                                       COPPER TUBE
                                        RUBBER STOPPER
                          )UCT WALL
                   Figure 4-5.   Measurement port seal.

                                   4-11

-------
     In measuring the static pressure drop across  a scrubber,  it  is
good practice to measure the upstream and downstream static pressures  one
at a time.  First of all, this guarantees that errors will  not occur due
to pluggage of one of the ports.  In the case of wet scrubbers, it  is
possible (and even likely) that a port will  plug minutes  after it has
been rodded-out.  Therefore, hooking up two separate lines  to  one gauge
as shown in Figure 4-6a is not recommended.   It is preferable  to  make
the measurements one at-;a time as illustrated in Figure 4-6b.   In this
case, a zero reading clearly indicates a port which was not originally
cleaned out or which has become blocked again.  Furthermore,  less tubing
is necessary when making the measurements one at a time.  Since most of
the ports are some distance apart, the connecting  tubing  necessary  to
measure pressure drop directly could represent a safety hazard; it  could
also pass next to a hot gas duct which would cause it to  collapse or burn.
Figure 4-6a.
Poor approach for measuring pressure drop  across  a  venturi
                    scrubber.              -
Figure 4-6b.
Preferred approach for measuring pressure drop  across  a
             venturi  scrubber.
                                   4-12

-------
     Before opening any static pressure port,  it is important to confirm
that the port is not connected to any type of  electrical  transducer.   In
some systems, transducers are an integral  part of the overall  process con-
trol system; the opening of the port can result in a low pressure signal
which will trigger the tripping off-line of the fan and process equipment.
If there is any doubt about the existence of a transducer somewhere along
the static pressure line, it should be assumed that one exists.  Preferably
there should be separate, isolated static pressure ports installed strictly
for the purposes of routine evaluaton using portable gauges.
     When recording the static pressure data,  the location of each measure-
ment should be carefully noted.  The data has  no meaning whatsoever with-
out the accompanying data concerning measurement location. A sketch of
the system, showing the ports, the scrubber, the demister, dampers,
bypass ducts, and fan(s) should be included in the notes.  As discussed in
the next section, the temperature of the gas stream should be measured at
the same points.
     The pressure drop measurements should be  conducted over  a moderate
time period at each location to determine if there are any short term
temporal variations.  These can be caused by cyclic process operating
conditions or by "seeking" type adjustable throat mechanisms  (for venturi
scrubbers).
     4.1.5.2  Measurement of Temperature.   Immediately after  measuring the
static pressure at the inlet and outlet ports, the gas stream temperature
should be recorded at the same points.  These  values can then be used to
convert the static pressures back to a standard gas density.   (See Sec-
tion 2.3 for a discussion of the relationship  between pressure drop,  gas
density, and energy consumption.)  In making these temperature measurements,
caution must be exercised to avoid (1) air infiltration at the port which
causes cooling of the measurement probe and (2) water droplet impaction
on the probe which also causes cooling.
     The temperature at the scrubber inlet can also be used as an indica-
tion of potential operating problems.  If it is greater than  300° F and
the gas stream contains a high concentration of condensible metallic or
organic vapors, then submicrometer particles can condense after the gas
has passed through the zone of optimal inertia! impaction. If it is over
                                   4-13

-------
300° F in combination with a high solids content liquor,  then  "regeneration"
can result from the droplets evaporating to dryness.
     4.1.5.3  Analysis of Pressure Drop.  Low static  pressure  drop  across
a wet scrubber, coupled with a high stack or residual  opacity  suggests
that the penetration rate has increased relative to the baseline  level.
The logical starting point when confronted with these conditions  is to
carefully inspect the entire liquid pumping and distribution system.
The reason for this is that pressure drop is a direct function of the
liquid-to-gas ratio at any specific gas velocity.
     The most common reason for low liquid flowrate is nozzle  or  pipe plug-
gage.  This may be indicated by an increase in the pump discharge pressure
(from the baseline levels).  Other problem symptoms include low pipe skin
temperatures and high suspended solids in the recirculation line.  If pos-
sible, the operator should rod out each nozzle and determine if there is a
change in the pressure drop and in the pump discharge pressure.
     A second reason for the low static pressure drop may be a decrease
in the gas velocity through the system.  In the case  of adjustable  throat
Venturis, the movement of the throat mechanism (damper, vanes, plump bob)
can change the throat velocity by a factor of 3.  The relationship  between
the gas velocity and throat velocity is linear, therefore, the pressure
drop can also vary by a factor of 3.  In some designs, it is possible to
check for relative movement of the throat mechanism.   If the wet  scrubber
is a tray-type, spray tower, or simple venturi, then  throat velocity adjust.
ments cannot be made independently of the gas flow rate.   In these  cases,
it may be advisable to measure the inlet gas flaw rate to determine if
the wet scrubber system is operating at less than design flow  rates;
most scrubbers are less efficient under these conditions since particle
impaction is less effective.
     The recommended technique for the measurement of gas flow rate up-
stream of a wet scrubber is the Pitot tube (EPA Reference Method  2).  Due
to the large quantities of suspended material upstream of the  scrubber, it
is generally necessary to use the S-type probe which  is less prone  to
pluggage.  It should be noted that wet scrubbers are often used on  sources
which interim"ttantly or continuously generate effluent gases in the
                                 4-14

-------
explosive range.  Accordingly,  any probes inserted upstream of  the  scrubber
should be carefully grounded.   A less accurate,  but safer approach  is  to
measure the gas flow rate downstream of the scrubber.   The decreased
accuracy results from evaporation of some of the liquor,  and condensation
in the scrubber.  Nevertheless, the downstream value is generally close
to the upstream value (after correction for gas density differences).
     Another indication of the gas flow rate can be obtained indirectly
by using the fan motor current.  A decrease in the corrected fan motor
current usually accompanies a decrease in the gas flow rate.  Correction
of the motor current to standard conditions can be accomplished by  multi-
plying the fan current at actual conditions by the appropriate factor
presented in Table 4-3.  The data in Table 4-3 is accurate only when  the
fan inlet static pressure is not less than -20 inches w.c.
     Other problems which can lead to low pressure drop all involve the
pump and the liquid distribution system.  Pump impeller wear and pump
seal leakage can both contribute to a decrease in the liquid flow rate.
At the earliest opportunity, the operator should inspect the pump  packing,
the pump impeller, and strainers mounted before the pump.  The position
of any flow control valves should also be checked.
               TABLE 4-3.  FAN  DATA TEMPERATURE CORRECTION9
Temp
°F
20
40
60
80
100
120
140
160
180
200
220
240
260
280
300
^From
tion

Factor
0.91
0.94
0.98
1.92
1.06
1.09
1.13
1.17
1.21
1.25
1.28
1.32
1.36
1.40
1.43
"Basic Energy/Envi
Temp
°F
320
340
360
380
400'
420
440
460
480
500
520
540
560
580
600
ronment Analysis",

Factor
1.47
1.51
1.55
1.59
1.62
1.66
1.70
• 1.74
1.77
1.81
1.85
1.89
1.92
1.96
2.00
NAP A informa
series 67, by C. Heath, August 1978.
                                    4-15

-------
 4.1.6  Liquid-Gas  Distribution
     Liquid-gas distribution is, unfortunately, a difficult factor to con-
 sider  during  a routine evaluation.  Usually, the only conclusive evidence
 of poor distribution  (also called liquor maldistribution) is the observation
 of severe  scaling and deposits inside the unit.  In extreme cases, there
 may be a decrease in the pressure drop across the scrubber or across an
 individual tray,  at a constant gas flow rate.
     Wet scrubber systems with moderate to high suspended solids levels
 should be  checked to the extent possible for maldistribution problems.
 Partially  or  completely plugged nozzles are a major cause of poor gas-
 liquid distribution which is particularly disruptive in venturi scrubbers.
 One external  check for plugged nozzles which can be made is to compare
 the skin temperatures of the various liquor supply lines going to the
 scrubber.  If one of these lines is plugged, the flow of hot recirculated
 liquor through it will be substantially reduced and its skin temperature
 will begin to fall.  A difference of 5 to 10 degrees F between two supply
 lines  often indicates a plugged line.  (The temperature difference is not
 as  great as might be expected since conduction of heat along the pipe
 continues despite the plug.)  Supply line pressure can also provide
 evidence of pluggage.
 4.1.7  Liquor Quality
     4.1.7.1  pH Measurement and Evaluation.  The pH of the recirculated
 liquor should be measured during each routine evaluation.   This can be
 done using color  indicating pH paper or a portable pH meter.   The pH paper
 is  adequate assuming it is not necessary to have highly accurate data
 and the  scrubber liquor will  not attack the paper.   Highly colored liquor,
 liquor with high levels of foam,  liquor with high levels of colloidal
matter,  and oxidizing liquor all  preclude the use of the pH paper.
     The liquor sample should be drawn from a region of the system
where its pH is at the lowest value.   This  is commonly at  the  scrubber
 sump effluent point,  prior to the discharge into the recirculation  tank.
Another sampling point of interest is the collected  liquor entrainment
at the region of the demister.   This  sample can  be  acquired using the
probe developed and discussed by  Schifftner.36
                                   4-16

-------
     A battery powered portable pH meter is ideal  for pH  evaluation
of wet scrubber systems.   These normally have an  accuracy of-_+ 0.2 pH
units and should be standardized using at least two  buffers.
     4.1.7.2  Solids Content.   The liquor turbidity  should be quali-
tatively analyzed.   If the turbidity is high, nozzle erosion and/or line
piuggage could occur in the system.   The rate by  which the material
settles to the bottom of the sampling flask should also be considered,
since it is an indirect indication of the particle size.   Larger  parti-
culate material is  more abrasive.
     If the turbidity appears  high,  the inspector should  evaluate the
purge rates.  A reduction in this  purge rate can  lead to  an increase in
the total solids content.  The settling pond (if  any) should also be
inspected to ensure there is minimum total  solids at the  pump intake.
     The presence of a persistent, high opacity plume at  apparently ade-
quate pressure drop levels can be  due to evaporative regeneration of
particulate in the  presaturator or the quench tower.  Samples of  the
liquor fed to the chamber should be obtained over a  representative period
of time; these should be analyzed  for both suspended solids and dissolved
solids.  Similar samples should be obtained of liquor drained from the
chamber which did not evaporate.  The difference  in  the total solids con-
tent of the feed and discharge streams will give  a rough  indication of the
loss of particulate to the gas stream.  This is only approximate  since
some of the solids  are lost as entrained water droplets from the  chamber,
and it is possible  that some particulate from the inlet gas stream is
captured in the chamber discharge  stream.
     4.1.7.3  Surface Tension.  If the opacity of the discharge has
increased without any other apparent changes, then one possible cause
is a change in the surface tension.  This change  could affect both the
droplet size distribution and the effectiveness of impaction.  The surface
tension can be measured using ASTM Method D-450.   The quantity of anti-
fearning agents, surfactants and flocculants should be recorded, if known.
4.1.8  Nozzle PIuggage
     Nozzle piuggage should be suspected when there  is increased  opacity
and one or more liquor feed lines with lowered skin  temperatures. If  it
                                   4-17

-------
 is impossible to rod the nozzles  out during  operation, then it must be
 done during the next outage.   At  this time,  each  nozzle  should be visually
 inspected for erosion and pluggage;  if damaged, the  nozzles should be
 replaced.   In some  cases,  it  will  be possible  to  open'an access hatch and
 observe the operation of the  nozzles (with no  gas flowing through the
 system).   If a nozzle is partially plugged,  the spray angle and pattern
 will  be severely distorted.   If it is completely  plugged, at best, only a
 trickle of liquor will  be  observed.   A high  powered  light (explosion
 proof)  is  useful  for inspection of the nozzles from  the access hatch.
 4.1.9   Bypass Duct  Damper
     The damper position of the bypass duct  should be checked.  If there is
 any  indication  of substantial  leakage  across the duct, follow up testing
 is necessary.
 4.1.10  Presaturator and Ductwork-
     Presaturator (or quench)  water  quality  is particularly important;  a
 sample  of  this  liquor should be obtained and checked.  Turbidity of this
 liquor  should be  very low, in  fact,  it  should resemble drinking water
 quality.
     The general  physical condition of the ductwork leading from the process
 to the  scrubber should be briefly checked.  Fugitive leaks  from positive
 pressure systems  are a direct bypass of the scrubber system.   Leaks  into
 negative systems  can reduce the quantity of gas pulled from the source  and
 thereby affect capture and result in fugitive emissions  at  the process.
The locations of  any obvious holes or gaps/tears  in expansion  joints
should be noted on a sketch.
     Air infiltration along the inlet duct can be  evaluated  by determin-
ing the 02 levels (combustion sources only)  at various points  along  the
duct.  An increase in the 02 levels between two measurement  points provides
clear indication of the location  of the leakage.   A material balance can
be used to estimate the quantity.
     Air infiltration along the ductwork can  also  be  detected  by compari-
son of the gas inlet temperature  to the baseline value.   A substantial
decrease is indicative of air infiltration.
                                   4-18

-------
4.1.11  Process Conditions
     As discussed earlier, the process operating conditions and raw material
characteristics can have a major impact on scrubber operating conditions
including, but not limited to:  the particle size distribution, the gas
flow rate, the particle surface characteristics, and the concentration of
condensible vapors.  Nonetheless, it is easier to identify the emergence of
problems by observing the stack and performing a routine evaluation of the
scrubber, than by first evaluating the process equipment.  Thus, the evalua-
tion of process conditions should be used primarily as a follow-up to
problems identified by inspection of the air pollution control equipment.
     On a routine basis, the evaluation should be limited to ascertaining
the overall process operating rate and determining the adequacy of hood cap-
ture (if applicable).  The hood capture can usually be evaluated visually
from a safe vantage point; an alternative method is to use the hood static
pressure as a qualitative means to estimate the hood gas flow rate.  Compari-
son of the present hood static pressure against the baseline period hood sta-
tic pressure provides an estimate of any shifts in hood capture efficiency.
     An increase in the vapor concentration (material  which may condense
in the scrubber to form particulate matter) of the gas stream can be indi-
cated by an increase in the stack opacity as compared to baseline values
and perhaps by a change in the process operating conditions.  The presence
of vaporous material can be confirmed by an extractive stack test using a
front filter (unheated) and a set of impingers immersed in ice.  Note that
this does not yield a quantitative measurement, only an indication of
the pretense of vaporous material.
4.2  APPLICATION OF BASELINE PERFORMANCE EVALUATION TECHNIQUE TO SPECIFIC
                            TYPES OF SCRUBBERS
     There is substantial diversity in the design of particulate wet
scrubbers.  The various types of scrubbers serve different applications,
are vulnerable to different operating problems, have different main-
tenance requirements, and have different particulate removal capabi-
lities.  To a certain extent, the means of evaluating performance is
also different.  Some of the unique factors which must be considered
while evaluating performance for common types of scrubbers are described
in this section.  It is by no means an exhaustive summary of inspection
techniques, rather it is intended to illustrate use of the evaluation
approach presented earlier.
                                   4-19

-------
     Six major categories of participate wet scrubbers are presented in
Table 4-4.  These are listed in the approximate order of both increasing
effectiveness and increasing operating cost.  The preformed spray scrubbers

range from simple "homemade" units for small scale material transfer
applications to very large, sophisticated systems used in flue gas desul-
furization systems.  Packed tower scrubbers are generally used for combina-
tion particulate and gaseous contaminant removal.  They are rarely used
solely for particulate control since there are more economical and simple
approaches of equal or greater effectiveness.  Moving bed scrubbers were
developed as an alternative to packed beds for applications where the gas

stream contains sticky particulate which could clog a packed bed.  This  is
a comparatively new approach, having been only introduced during the last
20 years.  There are numerous tray-type scrubbers almost all of which offer

               TABLE 4-4.  MAJOR TYPES OF WET SCRUBBERS3>b
  Category
   Common
 Application
   Specific Type
Preformed Spray
Packed Bed
Moving Bed


Tray-type
 Impingement Plate

Mechanically Aided

Gas-atomized
Material Handling
Asphalt
Phosphate Fertilizer
Primary Aluminum
Coal-fired Boilers
Spray Towers
Cyclonic Spray Towers
Vane-type Cyclonic Towers
Multiple-tube Cyclones

Standard Packed Bed
Fiber-bed
Cross-flow
Grid-packed

Turbulent Contact Absorbers
Municipal Incinerators   Perforated Plate
Material Handling

Foundries
Coal-fired boilers
Asphalt
Steel Mills
Sewage Sludge
  Incinerators
Wet Fans

Standard Venturis
Variable Throat Venturis
Orifice Scrubbers
   List not intended to be all inclusive.
   Adapted from a similar table presented in Reference 37.
                                   4-20

-------
moderate to high removal efficiencies at low pressure drops.  They are
currently used in a very wide set of applications.  The principal  category
of scrubber for high efficiency collection of submicron particulate is
the gas-atomized type, the most common of which is the venturi scrubber.
4.2.1  Preformed Spray Scrubbers
     The preformed spray scrubber is the simplest type of wet scrubber
and it has the lowest overall particulate collection capability.   It
usually consists of a vertical contact chamber with an array of spray
nozzles as shown in Figure 4-7.  The particuTate laden gas stream enters
near the bottom and passes upward past the spray headers.  The nozzles
may consist of sophisticated liquid atomizing nozzles or may simply be
small holes drilled into the spray header.
                                         Clean gas
                                                 Dirty gas
              Figure 4-7.  Typical preformed spray scrubber.
          (Courtesy of EPA's Air Pollution Training Institute.)
     This type of scrubbing system is often difficult to evaluate due to
the lack of instrumentation and accessibility.  Rarely do preformed spray
scrubbers include continuous recording pH meters and recirculation liquor
flow meters.  There is usually a pump discharge pressure gauge but rarely
pressure gauges on the individual pipes leading to the nozzle.
     The types of data which are useful for evaluating the performance are
listed in Table 4-5.  Emphasis is given to analyses of the recirculation
liquor quality since nozzle erosion and pluggage can be major problems
                                   4-21

-------
for preformed spray scrubbers.  Both problems have a direct and significant
impact on the particulate collection efficiency by changing the spray
coverage and average droplet sizes.  During a routine inspection,  the
turbidity of the recirculation should be observed (classified as "light",
"moderate", and heavy").  If the turbidity is moderate or heavy, the
condition of the spray nozzles should be evaluated to the extent possible.
Pluggage of some of the nozzles will be indicated by an increase in the
supply pipe pressure.  Erosion of the nozzle will yield the opposite
effect on the supply pipe pressure gauge.
     Many of these systems operate on an intermittent basis.   The condition
of the nozzles can be checked while the scrubber is off-line by operating
the recirculation pump for a short period of time and observing each
nozzle with a high powered flashlight (explosion proof).  Pluggage or
erosion of the nozzle will result in a distorted spray angle.  This check
should only be done if there is an appropriate access hatch and if the
scrubber vessel is purged of any toxic gaseous and particulate.
     If there is any indication of liquor solids related operating prob-
lems, the condition of the pond (if any) should be checked.  The pond
should have sufficient capacity to allow suspended materials to settle;
multiple zones separated by weirs will permit maximum settling.  The pump
intake line should be located well above the elevation in the pond where
silt and collected solids can be entrained in the liquor.

             TABLE 4-5.  PREFORMED SCRUBBER EVALUATION DATA
Data Baseline
Recirculation Liquor Turbidity
(Light, Moderate Heavy)
Recirculation Liquor Total Solids
Recirculation Liquor Suspended Solids
Recirculation Liquor pH
Spray Nozzle Operating Pressure
Physical Condition of Shell
Scrubber Inlet 02 or C0£ Levels
(Combustion Sources Only)
Quantity of Surfactants Used
Quantity of Anti -Foaming Agents Used
Liquor Flow Rate
X

X
X
X
X
X
X

X
X
X
Routi ne
X



X
X
X


X
X
X
                                   4-22

-------
     It is important to measure the pH of spray tower scrubbers since many
of these are constructed of carbon steel.  If the pH of the recirculation
liquor falls below 6, corrosion is probable.  If the pH increases above 10,
precipitation of calcium and magnesium compounds can result in scaling of
the scrubber and pluggage of the pipes and nozzles.
     The liquor flow rate to the scrubber is an important operating
variable.  The on-site monitor should be checked.  If one is not available,
the flow rate should be estimated at the point where the liquor returns
to the pond or the recirculation tank.  One parameter which is not of
interest is the pressure drop across the scrubber.  This is not directly
related to the effectiveness of the particle capture and the magnitudes
of the pressure drops are small.
4.2.2  Packed Bed Scrubbers
     Packed bed scrubbers (also called packed tower scrubbers) have
primarily been used for gas absorption or for gas cooling.  Both functions
are facilitated by the large liquid surface exposed to the gas stream as
the liquid flows downward over the packing material.  A typical packed
bed scrubber is illustrated in Figure 4-8.  Another common type of packed
bed system is a cross flow scrubber.  It is particularly advantageous
when a liquid mist is being collected since this facilitates drainage of
the accumulated material from the bed.
                                          Mist eliminator
                                             •;•>. Dirty gas
                Figure 4-8.  Typical packed bed scrubber.
           (Courtesy of EPA's Air Pollution Training Institute.)
                                 4-23

-------
     The packing materials used in packed beds can consist of crushed
granite, raschig rings, pall rings tellerettes, and  intalox saddles.
These are illustrated in Figure 4-9  (with the exception of the gravel).
All of the packing materials are usually the size range of 0.5 to 3
inches (1.25-7.5 centimeters), and the packing is usually randomly orien-
ted in the bed.  Some commercial units utilize several beds in series,
each with a separate liquor supply and distribution  system.
                        t
     The types of data most useful for the evaluation of packed bed
scrubbers (particulate removal only) are listed in Table 4-6.  As in the
case with the preformed scrubbers, the primary emphasis is the liquor
quality.  One of the major operating problems of packed bed scrubbers is
pluggage of the bed due to deposition and/or precipitation of solids.*38
The flow rate of liquor is often insufficient to remove the accumulated
material.  These solids can cause increased static pressure drop across
the unit, reduced gas flow through the control system, and channeling of
the gas stream through only a part of the bed.
                                          Berl saddle
                                          Intalox saddle
                         Pall ring
                                           Tellerette
                  Figure 4-9.  Typical packing materials.
          (Courtesy of EPA's Air Pollution Training Institute.)
                                   4-24

-------
             TABLE 4-6.  PACKED BED SCRUBBER EVALUATION  DATA
Data
Recirculation Turbidity
(Light, Moderate, Heavy)
Recirculation Liquor Total Solids
Recirculation Liquor pH
Recirculation Liquor Nozzle Pressures
Recirculation Liquor Flow Rate
Inlet and Outlet Gas Temperatures
Apparent Shell Corrosion and Erosion
Baseline
X

X
X
X
X
X
X
Routine
X

X
X
X
X

X
     The turbidity of the recirculation liquor should be low.   The  condi-
tion of the packing should be checked if the turbidity is moderate  to
heavy.  This can usually be done by opening the access hatches  above and
below each bed (the scrubber must be off-line).
     Typical liquid-to-gas ratios for packed bed scrubbers are  1  to 6  gal-
lons per 1000 ACF (0.1 to 0.5 liters per AM3).39  A decrease  in the liquor
flow rate relative to the baseline period may indicate a reduction  in  the
particulate removal rate and a major decrease in the gaseous  removal
efficiency.  The pH is measured primarily to ensure that corrosion  is  not
occurring.
4.2.3  Moving Bed Scrubbers
     Moving bed scrubbers were developed primarily as an alternative to
packed bed scrubbers in applications having very sticky particulate
material and/or high levels of gaseous contaminants.  The turbulent
motion of the lightweight packing material  allows self cleaning of  the
deposited material without sacrificing the good gas transfer  capability
common to packed bed scrubbers.  Principal  applications of this type of
scrubber include primary aluminum Soderberg pptlines and coal-fired
boilers (for flue gas desulfurization).  A simple schematic of  a  moving
bed scrubber is shown in Figure 4-10.  The trade name commonly  used is
Turbulent Contact Absorber (TCA).
     As shown in Figure 4-10, the gas enters at the lower side  of the  unit
after passing through a presaturator chamber (optional).  The pollutant
laden gas stream passes upward through a series of beds each  of which  is
10% to 25% full of the lightweight packing.  The liquor is introduced  at
                                   4-25

-------
                                    Clean gas
                                               Mist eliminator
                Figure 4-10.  Typical moving bed scrubber.
          (Courtesy of EPA's Air Pollution Training Institute.)

the top through a set of nozzles and passes through the beds counter-
current to the gas flow.  The bed is fluidized by the moving gas stream
and this results in the formation of liquid droplets and sheets in the
turbulent zone of the bed.  Often one or more beds of packing are operated
dry (above the point of liquid injection) in order to serve as an initial
demister.  A chevron demister or equivalent serves as the main entrainment
separator.
     Evaluation data for moving bed scrubbers are presented in Table 4-7.
Unlike the previous types of scrubbers discussed, the liquor turbidity is
not critical to the performance of the scrubber.  Therefore, an increase
or decrease in the liquor turbidity would not be cause for major concern,
this would, however, eventually degrade the liquor distribution nozzles at
the top of the scrubber.

          TABLE 4-7.  TURBULENT CONTACT ABSORBER EVALUATION DATA
Data
Static Pressure Drops Across Each
Recirculation Liquor pH
Recirculation Liquor Flow Rate
Spray Nozzle Header Pressure
Physical Condition of Shell
Baseline
Stage x
X
X
X
X
Routine
X
X

X
X
                                  4-26

-------
     The pressure drop the entire scrubber would be measured by  measuring
the static pressure at the scrubber inlet, and the pressure drop prior  to
the demister.  The procedures for measuring static pressure discussed
earlier should be used.  Typical pressure drops are 4 to 20 inches  of
water.  The pressure drop is proportional to the gas velocity, the  rate
of liquor addition, and to the number of beds in series.
     The recirculation liquor flow rate needs to be confirmed only  if
there has been a decrease in the pressure drops or an increase in the
plume residual opacity.
4.2.4  Tray-type Scrubbers
     A tray-type tower consists of a vertical shell with one or  more
plates or stages mounted horizontally.  The liquor is introduced at the
top through a simple delivery pipe and flows onto the top stage. The
height of liquor on this stage is maintained at 0.5 to 3 inches  (1.3 to
7.6 centimeters) by a downcomer on the side opposite the inlet pipe.
While the liquor is passing across the stage it is exposed to a  relatively
high velocity gas stream which passes up through numerous holes  in  the
stage.  A sketch of a tray-type scrubber is shown in Figure 4-11.  Two
common stage designs are illustrated in Figure 4-12.
     Penetration of particles greater than 1 jimA is low in most  tray type
scrubbers.  The degree of control for submicron particles is partially
dependent on the number of stages, the type of stage, the size of the
holes in the stage, and the design gas velocities through the holes.  The
static pressure drop is several inches of water per stage.40,41
     As with almost all types of particulate wet scrubbers, tray-type
units are vulnerable to corrosion and pluggage.  Both corrosion  and pluggage
have an impact on the particulate collection efficiency in addition to
their obvious effect on the scrubber's useful life.
     The static pressure drop across the scrubber is directly related to
the particulate removal effectiveness, therefore, this should be measured
during each inspection.  Measurement posts should be located before the
first tray, between each tray, and after the last tray before the gas
stream enters the demister section.  The liquor quality is important,
                                   4-27

-------
                         Clean gas
          Figure 4-11.  Tray-type scrubber.
(Courtesy of EPA's Air Pollution Training Institute.)
                          Sieve  Plate
                                             Impingement Plate
            Figure 4-12.   Types  of stages.
(Courtesy of EPA's Air Pollution Training Institute.)
                         4-28

-------
especially for the impingement tray type scrubbers.   Due to the extremely
small holes in the tray (approximately 3/16ths  of  an  inch), suspended
solids concentrations of more than 1% by weight can lead to rapid pluggage.
Obviously, the turbidity of the recirculation  liquor  should be very  low.
     If the opacity of the unit has increased,  it  may be necessary to
evaluate the condition of the trays.  This  can  be  done by  opening access
hatches above and below each of the trays.   A  build-up of  materials  in
part of the tray can lead to the channeling problem.   Partial pluggage
of the holes leads to an increased pressure drop.  Evaluation data for
tray-type scrubbers are presented in Table  4-8.

              TABLE 4-8.  TRAY-TYPE SCRUBBER EVALUATION DATA
Data
Pressure Drop Across Each Stage/Tray
Recirculation Liquor Turbidity
(Light, Moderate, Heavy)
Recirculation Liquor Suspended Solids
Recirculation Liquor pH
Physical Condition of Shell
Physical Condition of Trays
Baseline
X
X
X
X
X
X
Routine
X
X
X
X
4.2.5  Gas-atomized Scrubbers
     Gas-atomized scrubbers are the most energy intensive and efficient
type of particulate wet scrubber now  in large scale commercial  utiliz-
ation.  Regardless of the specific design, they all  operate by accelera-
ting the gas stream to high velocities, usually in the range of 10,000
to 40,000 feet per minute (3,050 to 12,2000 meters per minute).   The
design differences are found in the method of injecting the scrubber
liquid into the gas stream at the converging section of the scrubber.
Here, during the period before the atomized liquid reaches the same
velocity as the accelerated gas stream, the droplets serve as inertial
impaction targets for the entrained particles.
     4.2.5.1  Orifice scrubbers.  The common denominator in orifice
scrubber design is that the constricted area which serves as the "throat"
is partially flooded by a stationary pool of liquor.  Atomization and
                                    4-29

-------
acceleration of the  liquor  occurs as the gas stream passes  through
the constricted area with typical liquid-to-gas ratios up to 20 gallons
per 1000 ACFM.   Evaluation  data  requirements are provided in Table  4-9.

               TABLE 4-9.   ORIFICE SCRUBBER EVALUATION DATA
Data
Pressure Drop
Liquor pH
Liquor Level
Physical Condition of Shell and
Internal Components
Baseline
X
X
X
X

Routine
X
X
X
X

     One commercial  type of orifice scrubber, a Western Precipitation
Tubulaire Gas Scrubber,  is  illustrated in Figure 4-13.  The particulate
laden gas stream passes  down through the central delivery tube and then
makes a 180 degree turn  after it  passes around the central cone.  The
baffle to the left of the inlet tube serves as a demister.  It is very
difficult to evaluate the liquid-to-gas ratio with this kind of device,
since the liquor flows by gravity back to the main sump without the need
for a recirculation pump and piping  (some systems have recycle pumps);
nevertheless, the liquor level control is very important for the proper  -
operation of the scrubber.
                       GAS INLET
                                                            GAS
                                                            OUTLET
   DRAIN
                                        SCRUBBING LIQUOR
       Figure 4-13.  Typical  orifice type  gas-atomized scrubber.
                                   4-30
                                                                                 .

-------
     A second common type of orifice  scrubber is illustrated in Figure 4-14,
The contaminated gas stream enters  from the top of the collector and flows
down the rectangular inlet channel.   The  gas turns near the bottom and
enters a set of venturi  nozzles  inclined  slightly upward.  The liquid
level is controlled to a height  that  ensures that the desirable amount of
liquid is entrained and  atomized.   There  are a number of venturi  nozzles
arranged in parallel;  particulate impact ion on the droplets occurs primar-
ily in the throat of these nozzles  due to their high relative velocites
in this region.  The entrained droplets are later removed in the series
of stationary baffles  located downstream of the nozzles.
                                                  GAS OUTLET
                                                            BAFFLES
       VENTURI
       LIQUOR
       DRAG
       CHAIN
      Figure 4-14.  Typical orifice type gas-atomized scrubber.43
                                  4-31

-------
     As shown in Figure 4-14, a drag conveyor at the bottom of the sump  re-
moves the accumulated solids.  As with the previously mentioned type of
orifice scrubber, there is no recirculation line and pump since the liq-
uor is retained inside the collector.  For this reason,  it is  difficult
to independently evaluate the liquid-to-gas ratio.
     Since the fan is often mounted directly above the collector,  care
must be taken in the measurement of the pressure drop across the unit.
The inlet static pressure may be taken on the inlet duct or anywhere
along the inlet chamber portion of the scrubber.  The outlet pressure
must be taken before the gas enters the inlet to the fan.  For this
reason, the static pressure port must be located along the collector
wall in the general vicinity of that illustrated in Figure 4-14.  Loca-
tion of the port below the first baffle is not recommended due to  the
high quantity of sludge and liquor which impacts on the wall and could
cause pluggage.
     One of the principal  problems of orifice scrubbers  is maintaining
proper liquid levels.  Sufficient make-up water must be added  to allow
for evaporation, carry-over of droplets, and loss of liquid in the sludge.
Corros'ion and erosion of the constricted area or venturi nozzles are
also common problems and can lead to substantially reduced collection
efficiency.  Accordingly,  these nozzles often are simply bolted in to
facilitate replacement.^
     4.2.5.2  Venturi scrubbers.  This type of gas-atomized scrubber is
generally used in applications requiring large scale systems.   There are
numerous varieties available with the principal differences being  in:
     1.  Methods for varying the throat area to adjust for changes in gas
         flow rate.
     2.  Methods for injecting the liquor ahead of the throat.
     3.  Presence or absence of a diverging section.
     4.  Design of the entrained liquor droplet separator following
         the throat.
     A standard fixed throat venturi scrubber is illustrated in Figure
4-15.  The particulate laden gas stream enters the converging  section of
the venturi which usually  has an angle of approximately  25°.  The  liquor
                                   4-32

-------
   is injected  in  the converging section and is atomized near the inlet to
   the throat.   Immediately following the converging  section, the gas
   stream enters a diverging section (the throat having a negligible length
   in this case).  In the diverging section of the  throat of the unit
   shown in Figure 4-15 the angle is less than 15°; here the gas is gradually
   decelerated  prior to entry to the cyclonic chamber.  As shown in the
   plan view sketch, the gas stream enters the entrainment separator tangen-
   tial ly.  Some commercial units incorporate an additional demister tray
   to further remove the entrained droplets.
            THROAT
  LIQUOR INLET
CONVERGENT SECTION—^
      DIVERGENT SECTION
                                                      GAS OUTLET
                                                     CYCLONIC DEMISTER
SUMP
         Figure 4-15.  Venturi  scrubber with long divergent section.

       Data necessary for the evaluation of venturi scrubbers is presented
   in  Table 4-10.  The pressure  drop is  particularly important since this
   is  related to the impaction effectiveness.  As described earlier, the
   pressures must be converted back  to a standard gas density, if the  inlet
   and outlet gas densities have changed substantially since the baseline
   period.
                                     4-33

-------
            TABLE 4-10.  EVALUATION DATA FOR VENTURI SCRUBBERS
Data
Pressure Drop
Recirculation Liquor Flow Rate
Recirculation Liquor Total Solids
Recirculation Line Skin Temperatures
Presaturator Total Solids
Presaturator Turbidity
(Light, Moderate, Heavy)
Inlet and Outlet Gas Temperatures
Physical Condition of Shell
Quantity of Surfactants Used
Quantity of Flocculants Used
Quantity of Anti -Foaming Agents Used
Recirculation Liquor pH
Baseline
X
X
X
X
X
X

X
X
X
X
X
X
Routine
X
X
X
X

X

X
X



X
     Due to the gradual slope of the divergent section in this unit,  the
kinetic energy of the gas/liquor stream is partially converted back to po-
tential energy*  As a result, the gas stream static pressure will  increase
moderately as the gas passes through the divergent section.  The pressure
drop of direct interest to the inertial impaction phenomenon is the pres-
sure drop between points 1 and 2 in Figure 4-15.  The actual pressure
drop measured in many commercial systems is representative of the pressure
drop between points 1 and 3.  The latter value will be lower according
to the extent of static pressure recovery.
     A variation of the classical fixed throat design is the annular  ring
throat. The flange separating the converging and diverging sections has
a small extension lip which supports a ring insert.  The area of the
opening in this insert determines the maximum throat velocity at a given
gas flow rate, and thereby controls the collection efficiency.  This
insert can be changed and will vary the collection efficiency.45 On
some commercial units it is difficult to determine if such a change
has been made or even that the scrubber has an insert.  Erosion of an
insert over the time can lead to a reduction in gas velocities, and a
corresponding reduction in the particulate control efficiency.  Erosion
of the insert or installation of a different insert can often be de-
tected by comparing the present pressure drop at a given gas flow rate
and liquor flow rate with baseline data.  One generally accepted rela-
tionship for the pressure drop is shown in Equation 5-1.
                                   4-34

-------
                      AP = 0.001 Vt2 (L/6)
Equation 5-1
where:
             AP   =  pressure drop, centimeters of water
               2
             Vj.   =  throat velocity, cm/sec
             L/G  =  liquid-to-gas ratio, (m3/g)/(m'3/g)
     In a common type of adjustable venturi,. a hand cranked assembly
moves a center cone up or down to change the annular throat area.  The
length of the throat also varies slightly as the cone moves.  The liquor
is injected from a set of four tangential pipes (no nozzles) and from a
center deluge pipe oriented directly above the center cone.  A flooded
elbow (a depression of 6" to 9" filled with liquid) is used at the bottom
of the venturi section to absorb the impact of the turning gas stream.
This reduces erosion of the elbow.
     Another variation of the annular area variable throat venturi is the
flooded disc scrubber.  As shown in Figure 4-16, the center assembly
includes a large flat disc mounted horizontally at a point just upstream
of the constricted area.  The liquor passes up through the center column
and out over the disc.  As the gas stream is accelerated around the
edges of the disc, the liquor is entrained and atomized.  Particulate
impaction on the atomized liquor occurs directly downstream of the
disc.  The gas velocity can be controlled by slight adjustments in
the elevation of the flooded disc with relation to the constricted area.46
There is no divergent section per se, therefore, no substantial pressure
recovery would be expected downstream of this type of collector.
     One variable throat design common to many commercial units is depicted
in Figure 4-17.  In this design, two damper blades are used to vary the
open area in the rectangular throat and thus control the gas velocity
through the throat.  The position of the damper blades are controlled
either manually or by a differential pressure sensor and controller.  The
liquor can be injected by a number of methods including the deluge nozzles
with deflectors shown in Figure 4-17.  Other designs have a single variable
                                   4-35

-------
damper on the throat with  a  side mounted  controller.   In  these units,  the
liquor spray is directed towards the  section  opposite  the damper.
     The dampers are in the  high velocity zone, therefore,  erosion  can
be a problem.  The scrubber  may gradually lose the  capability  to achieve
the baseline pressure drops.  To diagnose this problem an internal
inspection may be necessary  when the  scrubber is  off-line and  purged
out (agency inspectors should jTot^ enter equipment).
                                      clean gas out
                      dirty gas in
                  water in
                                            slurry but
           Figure  4-16.   Flooded disc 'type venturi  scrubber.
                     (Courtesy  of Research  Cottrell.)
                                   4-36

-------
    GAS INLET-
   LIQUOR  INLET
  THROAT DAMPERS
    GAS OUTLET	

            Figure 4-17.  Adjustable venturi throat mechanism.

     A variable rod-type, adjustable venturi scrubber is illustrated  in
Figure 4-18.  In this design, the throat is comprised of the multiple
slots between the rows of horizontal rods.  The gas velocity can  be mod-
ified by  changing the diameter  of the rods, the spacing of the  rods,  and
the spacing between adjacent layers of the rods.47  This type of  scrubber
is conceptually similar to the  annular ring type discussed earlier, except
that changes in the throat area can be made while the unit is on-line.
However,  it is difficult for regulatory agency personnel to independently
determine that the rod arrangement has been changed.  Erosion and corro-
sion of the rods are potential  problems.  The pressure drop can provide
a good indication of these conditions.
                                  4-37

-------
               Dirty gas
                  « ^^ »
                   •^"i  Water sprays
               Figure 4-18.  Variable rod venturi  scrubber.

4.3  APPLICATION OF BASELINE PERFORMANCE EVALUATION TECHNIQUE TO SPECIFIC
                                INDUSTRIES
     This section describes a number of wet scrubbers applications and
the problems unique to each application to a specific source category.
It is not an exhaustive summary of all applications of the various types
of wet scrubbers.  However, each application has been included because it
illustrates conditions and potential problems that must be considered
when evaluating the performance of the wet scrubbers on many different
processes.
                                   4-38

-------
4.3.1  Sludge Lime Kilns at Kraft Pulp Mills
     Calcination of the sludge produced in the kraft pulping process  is
necessary to reuse the lime.  The sludge, which is normally 55% to 65%
solids, is fed to a rotary kiln which is usually fired by either oil  or
natural gas.48,49,50 Typically rotary kilns are 8 to 13 feet in diameter
and 100 to 400 feet long,49 and have exit temperatures in the range of
300° to 500° F.48  A venturi scrubber having either a fixed or variable
throat is the most common type of air pollution control device used on
rotary kilns.  Some plants have previously used impingement plate scrub-
bers, but the Venturis have gradually gained the major share of the market
because of their lower capital cost and higher efficiency.49
     Typical static pressure drops for the venturi scrubbers range from
10 to 30 inches of water; however, the large majority of the systems  op-
erate in the much narrower range of 15 to 20 inches of water.48 The
liquid-to-gas ratios vary somewhat depending on the design of the equipment,
but a representative ratio would fall in the range of 10 to 25 gallons
per 1000 ACF.49 Due to the alkaline nature of the entrained dust, the pH
of the recirculating liquor is high, often above 11.  Consequently,
scaling has been a problem in some units
     The mass loadings of the inlet gas stream from the rotary kiln are
relatively high at 3 to 20 grains per ACF.48,49  The large majority of
these particles are calcium oxide dust particles which have a fairly  large
geometric mean size.  However, the gas stream may also contain a small
small quantity of sodium oxide material which evolves as a very small dia-
meter fume.50*51  During the periods when substantial quantities of
submicron sodium oxide particulate are present, the mass emissions can
be higher than the baseline period even though the pressure drop has  not
changed.  This illustrates an important point:  A wet scrubber cannot be
evaluated independent of basic process operational checks.  In this case,
the amount of sodium material evolved is strongly dependent on the quantity
of sodium in the limestone mud feed, which varies from 0.1% to 2.5% by
weight.  The inspector should evaluate the operation of the mud washers
if there is any question concerning the emissions from the scrubber
system.
                                   4-39

-------
      Compared with other applications of wet scrubbers,  the  quantity  of
 suspended solids in the recirculation liquor of lime  kiln  scrubbers can
 be high.  The solids usually range from 10% to 30%  by weight,49  and plug-
 gage of the inlet lines to the venturi  scrubber is  a  frequent problem.
      Venturi scrubbers on lime kilns  are rarely equipped with presatura-
 tion chambers, thus, the inlet gas temperatures to  the throat are usually
 greater than 300°F, causing some liquor evaporation.   This can have an
 adverse effect on particle collection and will  increase  the  gas  volume
 handled by the induced draft fan.
      The typical  operating characteristics of  particulate wet scrubbers
 used on rotary lime kilns are summarized in Table 4-11.
       TABLE  4-11.
TYPICAL OPERATING CHARACTERISTICS OF SCRUBBERS ON
    ROTARY LIME KILNS AT KRAFT PULP MILLS50.5!
Parameter
Liquid-To-Gas
Liquor Solids
Pressure Drop,

Ratio, Gal
Content, %
Inches H2

s/1000
by wei
0

ACF
ght

Scrubber
Impingement
4-5
1-2
5-7
Type
Venturi
10-25
10-30
10-30
     the principal problem of the impingement plate scrubber as applied
to rotary line kilns, is the pluggage of the very small holes in the
plates.  For this reason, the solids content of the recirculation liquor
must be maintained at less than 2% by weight.  The venturi scrubber on
the other hand, is less expensive to purchase and can be operated at a
higher solids level, therefore, use of the venturi facilitates the over-
all plant-wide water and lime balances.  The venturi operates at a higher
pressure drop and achieves greater particulate removal  efficiencies at
these pressure drops.  The venturi scrubber is the type used on systems
installed in the last 10 years.
4.3.2  Asphalt Batch Plants
     There are several  major processes used to make asphalt concrete,  the
two most common being the hot mix batch plants and the  newer drum mix
plants.  There are substantial  differences in the characteristics of the
                                   4-40

-------
particulate matter from these two processes.  In both cases,  subtle
process changes change the particle size distribution substantially and
may even alter the wettability of the materials.
     In the conventional hot mix plant, the various grades of aggregate
necessary to produce the types of asphalt are stored in open  piles or
bins near the dryer.  These are fed from the bins at a controlled rate
using weigh belt feeders.  The accumulated material passes into the top of
the dryer where it is exposed to high temperature flue gas.  The aggregate
is dried as it passes through the dryer countercurrent to the gas stream.
A single burner, usually fired by either gas or oil, is used  to provide
the necessary heat.
     The dried aggregate is then transferred to the top of the tower where
it is mixed with heated asphalt purchased from local refineries.  The ag-
gregate and asphalt are mixed in the pug mill and discharged  to delivery
trucks.  The flue gas from this process passes through a cyclone for re-
moval of the potentially useful aggregate dust, and then goes to the
scrubber.  Most newer plants, being subject to the Standards  of Perfor-
mance fpr New Sources, use a venturi type scrubber; however,  numerous
older stationary plants use either a simple spray tower or a  cenrifugal
washer.  In some cases, the spray tower scrubbers in-use have either
been designed or extensively modified by the operators.
     In the drum mix process, the aggregate drying and mixing with asphalt
are both accomplished in the rotary drum.  The asphalt is injected at a
point midway down the dryer where the gas temperature should be suffi-
ciently low to prevent significant volatilization of the asphalt components,
However, the injection point should not be too far removed from the
firing end, or the gas temperature will be too low to permit proper
coating of the aggregate and the resulting product will not meet customer
specifications.
     The particulate matter emitted from drum mix plants is significantly
different from that of conventional hot mix plants.  Emissions from drum
mix plants, especially if the  injection point is too close to the flame
zone, will include a portion of condensed hydrocarbon particles.
These can be significantly smaller in size than the entrained aggregate
itself and the material may be less wettable due to the organic nature
                                   4-41

-------
 of  the  emissions.  A similar problem can result when a drum mix plant
 is  used for the recycling of asphalt.  In this case, the old road bed
 material is ground and added through a chute near the middle of the drum.
 Volatilization of the asphalt materials can lead to a very difficult to
 control, partially organic particulate matter.52
     When there is partial volatilization of the asphalt binder, the wet
 scrubber baseline data may be misleading.  Due to the large number of
 very small particles, the emissions may be substantially higher than
 during  the baseline period even though there has been no significant
 change  in the observed pressure drop or the liquid to gas ratio.  If the
 plume residual opacity is high and there has been no significant change
 in  the  scrubber operating parameters, the inspector should inquire about
 changes in the asphalt binder injection.
     Typical operating conditions for the wet scrubbers used in asphalt
 plants  are listed in the following Table 4-12.   The range pf pressure
 drops is quite large.  This is partially due to the variable quantities
 of organic material  volatilized (drum mix plants) and the variable particle
 size distribution resulting from the drying of the aggregate itself (drum
mix and hot mix batch).   The latter is due to variations in the friability
 of the  aggregate which can breakdown in certain plants to form  a high con-
centration of fines  and the presence of clay fines in certain aggregates.
Both act to substantially increase the amount of small  diameter dusts
entering the scrubber.   The process and raw material  variables  discussed

       TABLE 4-12.   TYPICAL OPERATING CHACTERISTICS OF ASPHALT  PLANT
                      PARTICIPATE WET SCRUBBERS53,54,55
Parameter
Liquid-to-Gas Ratio,
Gals/1000 ACF
Pressure Drop, Inches H20
Inlet Gas Temperature, °F
Hot Mix
Spray Tower Scrubbers and
Centrifugal Scrubbers
3-10
2-6
325-425

Venturi
Scrubbers
5-10
15-25
. 325-425
Drum Mix
Venturi
Scrubbers
5-10
15-30*
225-350
*Asphalt Concrete NSPS Background Document (EPA) recommends a venturi
scrubber with a pressure drop of at least 20 inches of water.
                                   4-42

-------
above underscore the importance of evaluating each system on a site-
specific basis using the baseline ;data.  The data presented in Table 4-12
is not intended to be applied as performance criteria for a specific
asphalt concrete scrubber.
     Regardless of the type of asphalt plant, the size of the facility is
usually quite small.  The range of gas volumes is 15,000 to 50,000 SCFM,54>55
well below that for wet scrubbers in other applications.  Due to the size
of the plants and the intermittent operating schedules, there is rarely
an engineering staff available to assist the operator in diagnosing
operating problems.  Furthermore, many of the wet scrubber systems at
these plants have no instrumentation for determining pressures, liquid
flow rates, and gas temperatures.  Another problem common to asphalt
concrete plants is reentrainment.
     Many of the plants construct a make-shift pond for settling of the
solids collected.  In some cases, inadequate ponds or inadequate pump
intake designs result in excessive solids pick-up.  This can lead to
pluggage and erosion in the scrubber.  Ponds sho.uld be separated by several
overflows in order to improve settling.  General  design criteria for
settling ponds are listed below..                 _
                 Number of zones =3
                 Volume          =1/2 day circulating pump capacity
                 Depth           = 6 feet (minimum)
                 Width           = twice the length
Additional  design criteria are discussed in reference 56.
     Scrubbers used on hot mix plants burning natural  gas  have few problems
with corrosion;  however,  those burning oil  with  a high sulfur content can
have low pH problems, especially if the recirculation rate is high.   Corro-
sion of the scrubber can  also be caused by mildly acidic ground waters in
certain locations  and by acidic components in the aggregates being  dried.
It is important to check  the liquor pH frequently and neutralize with
hydrated lime or caustic  soda, if necessary.
                                   4-43

-------
 4.3.3   Grey  Iron  Foundries
     A  grey  iron  foundry  cupola  is  a  firebrick, refractory lined cylin-
 drical  furnace  used  for the melting of  scrap to produce grey iron.  The
 cupola  is  alternately charged with  coke, limestone and soda ash, and
 scrap.  The  effluent gases pass  through the top of the cupola and exit
 at temperatures as high as 1800  °F.   The entrained particulate matter is
 partially  composed of incompletely  combusted coke and entrained charge
 materials.   Because  of the high  temperatures involved, there is a sub-
 stantial quantity of submicron fume and condensed organic particles.  The
 gas stream also contains  very high  concentrations of carbon monoxide (CO).
 The concentration of CO can vary from as low as several percent to as
 high as 25%  depending on  the iron to  coke ratio charged and the degree of
 maldistribution of the material  across the cross section of the cupola.57
 The extremely high negative static  pressures present near the I.D. fan,
 can cause  severe  air inleakage.  If this inleakage results in inadequate
 exhaust of the  cupola, then some discharge may be evident from the cupola
 top.  Fugitive  leaks of exhaust gases during periods of positive pressure
 operation can pose a risk to inspection personnel  because of the high
 concentrations  of CO and  particulate matter present in the gas stream.
     Due to  the batch type operation  of cupolas, their effluent conditions
 are not constant.  Thus,  stack test data taken at one point in time may
 not be representative of  other stages of a heat.  Significant differences
 in  the particle size distribution, the carbon dioxide/carbon monoxide
 ratio, and the  gas temperature may occur making it very difficult to
 prepare a baseline data set.   The cupola conditions must be well  docu-
mented in order to make the baseline data reasonable.
     Venturi scrubbers are the most frequently used for control  of cupola
 emissions.  These are necessary due to the small particle sizes  involved.
Often the pressure drop of the scrubber must exceed 40 inches w.c.  in order
 to  achieve an adequate removal  efficiency.   In some cases the required
 pressure drop exceeds 120 inches of water.   At the typical  pressure
drops of cupola venturi  scrubbers,  the outlet static pressures (between
the  scrubber and the fan)  are in the range of -25  inches  to -125  inches.
                                   4-44

-------
Care must be taken to avoid very  large  errors in the measurement of the
pressure drop due to the aspiration  effect discussed earlier.
     As shown in Figure 4-19 the  effluent concentration from the scrubber
can be a function of the scrap  quality.  The increased emissions with
oily scrap is probably due to increased  concentrations of organic vapor
which condense in the scrubber  to form  a very small, difficult to wet
particulate material.  Therefore, pressure drop is not acceptable as a
single indicator of performance.   As in  other scrubber applications,
the process operation and material  characteristics must be taken into
account.
         cc.
         UJ
         o
         z
         o
         o
         en
         o
         t-
         Ul
         _J
fe
o»
               0.4
   0.3
   0.2
    O.I
              O.O5
        'CLEAN SCRAP
                  0   10  20  3O  40  5O  6O  70  8O  90  100
                  SCRUBBING PRESSURE DROP, inches-water gauge
     Figure 4-19.  Pressure drop collection  efficiency  relationship.

     Typical operating conditions for the wet  scrubber  used on cupolas are
listed in Table 4-13.  It appears that basic cupola  design characteristics
do not have a significant effect on emissions  or specific melting  rates.
The use of briquettes increases emissions.  The blast rate also  has a
significant effect causing greater emissions and smaller particle  sizes
as the blast rate increases.
                                   4-45

-------
    TABLE 4-13.  TYPICAL OPERATING CHARACTERISTICS OF PARTICULATE WET
                          SCRUBBERS ON GREY IRON CUPOLAS58
      Parameter
 Values
 Liquid-to-Gas Ratio, Gals/1000 ACF
 Pressure Drop, Inches H20
 Gas Inlet Temperature, °F
 Liquor Solids Content. % by weight
 10-25
 25-45
160-200
  1-10
 4.3.4  Coal-fired Boilers
      Particulate wet scrubbers used on coal-fired boilers are generally
 found in conjunction with S02 absorption systems.  The most common
 types of particulate scrubbers in this application include variable
 throat Venturis (e.g. plumb bob, venturi rod) and moving bed scrubbers.
 This type of installation is large, with gas flows exceeding 106 ACFM;
 typical operating conditions are presented in Table 4-14.  In most cases,
 there are several  scrubbing trains in series so that changes in  boiler
•load can be accommodated.  Due to the high cost of such systems, there
 is almost always reasonable instrumentation including pH monitors, static
 pressure gauges, and liquor flow rate meters.

 TABLE 4-14.  TYPICAL OPERATING CHARACTERISTICS OF PARTICULATE WET SCRUBBERS
                               ON COAL-FIRED BOILERS59*60
Parameter
Liquid-to-Gas Ratio, Gals/1000 ACF
Pressure Drop, Inches H20
Gas Inlet Temperture, °F
Liquor pH
Venturi
(all types)
16-30
3-20
280-450
6-8
Moving
Bed
50
10-15
280-350
6-8
     The particle size  distribution and mass loadings from a given boiler
are a function of the fuel quality and the boiler firing conditions (such
as excess air rates).   Compared with other types of sources, the gas
                                   4-46

-------
quantities and the particle size characteristics are relatively stable.
Thus, scrubbers at pulverized coal-fired boilers are suitable candidates
for the development of relatively simple baseline performance curves.
     Two of the principal problems reported from these types of scrubbers
are corrosion and scaling, both of which are partially dependent on the  pH
of the recirculation liquor.  Operation of the particulate scrubber at
pH levels below 6 can lead to severe corrosion of carbon steel components.
Scaling results from precipitation of calcium and magnesium compounds
which normally happens at pH levels of 9 and above.  One factor of impor-
tance in this phenomenon seems to be the degree of oxidation of sulfite
to sulfate in the liquor.
     Recent water quality requirements have resulted in greater recircu-
lation rates than previously would have been considered.  This, in turn,
has  increased  (1) the risk of corrosion due to the concentration of chlo-
rides and other corrosive agents, (2) the  risk of erosion due to the
higher solids  levels, and  (3) the risk of  pluggage.
     High opacity conditions have been observed at a number  of wet scrubbers.
which"appear to be  operating satisfactorily.  The light scattering aerosol
formation is not well understood, however,  it may be due to  condensation
of vaporous material  formed  in  the  boiler.  The most likely  compound  is
sulfuric  acid  vapor.61
4.3.5  Municipal  Incinerators
      During the  last ten years,  the number of municipal incinerators  in
 operation has  been  gradually decreasing  due to  the  increased use  of
 landfills.   Nevertheless/ a  number  of wet  scrubber  systems  are used
 for  particulate  control  on the incinerators still  in  operation.   The
 types  of scrubbers  commonly found in this  application  include impingement
 plate scrubbers  and various types of Venturis.   Typical operating charac-
 teristics for these venturi  systems are presented in  Table  4-15.
      One of the major problems of wet scrubber  operation  as applied to
 municipal incinerators  is control of the condensible,  partially oxidized
 organic matter which forms when the incinerator operating temperature is
 too low.  This material condenses while passing through the scrubber.
 The final particle sizes are too small for effective impaction in the
                                    4-47

-------
 scrubber,  and therefore,  collection  efficiency  is  low.  Few scrubbers
 have been  able to  achieve the  NSPS limits because  of this problem.
 The residual  plume from these  units  is  generally a bluish white.  Due to
 the high chloride  content of the material charged, corrosion of the
 scrubbing  systems  applied to incinerators can be severe.
  TABLE 4-15.
TYPICAL OPERATING CHARACTERISTICS OF MUNICIPAL  INCINERATOR
              VENTURI SCRUBBERS62,63,64
          Parameter
                                       Venturi  Scrubber
  Liquid-to-Gas Ratio, Gals/1000
  Pressure Drop,  Inches F^O
  Inlet Gas Temperature, °F
  Liquor Solids Content, % by weight
                                             4-8
                                            15  - 60
                                           300  - 500
                                             1  - 10
     As with other applications of scrubbers, some process operational
data must be evaluated to determine if there has been a significant shift
in the baseline conditions.  The incinerator operating temperature and
the charging rate should be checked at a minimum.  It may also be necessary
to determine if there has been a change in the types of materials charged
to the incinerator.
4.3.6  Basic Oxygen Furnaces
     Venturi scrubbers have commonly been used to control  the fine parti-
culate emissions from basic oxygen furnaces (BOF's).  This application  is
of interest due to the very high static pressure drops involved and the
cyclic nature of the gas flow rate over the 30 to 45 minute operating
cycle; the latter raises a question concerning the definition of pressure
drop.
     The basic oxygen furnace is a batch reactor in which  up to 350 tons
of pig iron and scrap are converted to steel  by the oxidation of the
carbon, phosphorus, silicon, and magnesium impurities.  The materials
charged include a minimum of 70% molten iron,  10% to 30% scrap, and
limited quantities of flux materials such as  lime and fluorspar.65  The
offgas from the BOF consists of 75% to 90% carbon monoxide and the gas
temperature is approximately 3000°F before dilution .
                                   4-48

-------
     The gas evolution rate varies substantially  during  the oxygen blow
period.  An example gas flow rate versus time  curve  is presented in Figure
4-20.  Since the pressure drop across  a venturi scrubber is a function of
the gas flow rate, the pressure drop will  also have  major variations
during the production cycle.  To be explicit,  the time must be specified
along with pressure drop value.  The typical operating conditions of
venturi scrubbers and quenchers serving BOF's  are presented in Table 4-16.
                                       10-MINUTE RECOVERY
                                67% VOLUME (215 Btu/cf) = 533,594 scf
                                               OFF-GAS FLOW RATE
                                               % CO
                                         10    12    14    16    18
        Figure 4-20.  Gas flow rate and CO content during a blow.22

      The particulate matter entrained in the gas stream is very fine with
 85% to 95% of the mass smaller than 1 micrometer in diameter and as much
 as 20% of the mass less than 0.5 micrometers.66  The venturi scrubbers
 are operated at very high pressure drops to achieve reasonable control
 efficiencies.  The typical range is 40 to 70 inches of water.22»65  Due
 to these high negative static pressures, pressure drop measurements must
                                    4-49

-------
be made carefully.  The particle size distribution generated by the
process and consequently the pressure drop necessary, is a function of
the raw material charge characteristics, the rate of oxygen blow, the
type and quantity of ladle additions, and the condition of the refractory.
     A quencher is used to decrease the gas temperature prior to entry to
the venturi scrubber.  As much as 80% of the particulate matter is removed
in the quencher, therefore, the pressure drop across this collector is also
of interest.
      TABLE 4-16.
TYPICAL OPERATING CONDITIONS OF VENTURI SCRUBBERS
     SERVING BASIC OXYGEN FURNACES22*65.66
Parameter
Liquid to Gas Ratio,
Gallons/1000 ACF
Pressure Drop, Inches 1^0
Gas Inlet Temperature, °F
Venturi Scrubber
25-50
50-70
120-150
Quencher
Variable
5-10
3000
                                  4-50

-------
        5.0  EVALUATION OF OPERATION AND MAINTENANCE PROCEDURES
     This section addresses basic operation and maintenance procedures
for wet scrubber systems.  Since there is considerable diversity in
the design of scrubber systems and in applications, there is no one
set of operation and maintenance procedures which is adequate in every
case.  Instead, it is necessary to tailor the procedures to the
specific site involved;  this is an evolutionary process in which the
initial procedures are modified in response to the specific operating
problems experienced at the site.
     The procedures presented in this section represent a skeleton
of the initial operation and maintenance requirements for a wet
scrubber system.  For some systems, these procedures include steps ,
which are unnecessary and for some systems, important maintenance
items may not be included.  Therefore, regulatory agency personnel
should not conclude that the following procedures necessarily consti-
tute a complete description of good operating practice.
5.1  STARTUP AND SHUTDOWN PROCEDURES
     Proper startup and shutdown procedures are important in preventing
(1) damage to the corrosion and abrasion resistant lining in the scrubber
(if used), (2) overheating of the fan motor,  and (3) damage to the pump
impeller and motor.  Each scrubber manufacturer provides instructions
for proper startup and shutdown of its system which should be strictly
followed.  To illustrate the major elements,  a condensed version of
these procedures is provided in the two lists that follow.
Routine Startup
  o  Inspect piping and valves and open or close valves to permit proper
     flow of fluid from the system to recycle tank, pumps,  and return lines.
  o  Inspect ductwork system to assure that all  ductwork leading to the
     scrubber is intact and that damper blades move with damper control
     activation.  Flow of exhaust gas to scrubber should not be initialed
     until liquid circuit has been fully started.   Inspect and test in-
     struments to be sure they are operating  properly.
  o  Make sure that the unit is level  (especially important for tray-
     type scrubbers).
  o  Turn on all electrical  switches for controls and instruments.
                                  5-1

-------
  o  Fill system with scrubbing liquid by opening plant water lines
     to recycle tank, allow it to fill to proper level, and then
     start recycle pump.
  o  Open liquid line to quench/presaturator sections and adjust to
     proper flow.
  o  Open damper to direct exhaust gas flow to venturi  scrubber and
     start the fan.
     Prior to the startup of a wet scrubber, a complete internal  inspec-
tion should be conducted.  A partial  list of the items  to inspect in-
clude:  the condition of the spray nozzles; the condition of the supply
lines to the scrubber; presence of internal deposits; presence of worn
or eroded linings; presence of partially plugged or warped demisters;
and presence of corrosion on the interior trays, downcomers, and other
components.  For some types of scrubbers, it may be advisable to visually
confirm that the inlet liquor flow is sufficient to completely wet the
venturi convergent section to avoid any wet-dry interface problems.
     During startup, the bypass duct (if any) should be gradually closed
as the inlet damper is gradually opened.  If the fan inlet damper is opened
too quickly, it is possible to overload the fan motor.
     After operating conditions have stabilized, a complete set of readings
should be taken and compared against the baseline operating levels.  It
is advisable to conduct a routine evaluation, as discussed earlier, to
confirm that the system is operating properly.
     It may become necessary to adjust the liquid flow  rates to maintain
the recycle tank level.  This requires balancing the water make-up rate,
the purge rate, the sludge removal  rate, and the recycle rate.
Routine shutdown
  o  Turn off the fan and close the inlet or outlet damper.  Open the
     bypass damper to ensure that the scrubber is isolated from any
     process gas flow.
  o  Continue pump operation and all  internal water sprays until  the
     scrubber system has cooled.
  o  If the system will be exposed to cold weather conditions,  or if
     an extended shutdown is anticipated, all pipes should be drained.
                                5-2

-------
  o  An  internal inspection should be conducted to determine if there are
     eroded or plugged nozzles, worn (corrosion or abrasion resistant)
     liners, damaged demisters, or other problems which warrant
     .maintenance attention prior to startup.
  o  All deposits which could accelerate corrosion of the scrubber
     should be washed out.
     During shutdown, it is particularly important not to expose the scrub-
ber system to any high temperature excursions.  For this reason the pumps
and internal sprays should be operated until well after the scrubber has
been isolated from the process.  It is also important to continue the
alkali additions to the system long enough to prevent low pH excursions
after shutdown.
5.2  ROUTINE MAINTENANCE
     This section presents general  information concerning the operation
and maintenance of wet scrubber components; specific maintenance require-
ments vary substantially depending on the design of the scrubber and its
application.
5.2.1  Fans
     The types of operating problems experienced with fans depend on
whether the fan unit is before or after the scrubber.  If the fan is in
the upstream position where the particulate mass concentration is high,
erosion of the fan wheel  is of concern.   Due to the high gas temper-
atures prior to a scrubber, it may also be necessary to cool the fan
bearings.  If the fan is in the downstream position, it is vulnerable to
the buildup of solids and corrosion.
     Regardless of location,  all  fan wheels should be visually inspected
on a regular basis.   A typical  frequency is monthly, although in some appli-
cations weekly inspection is  warranted.   If the fan vibration increases
noticeably, the fan wheel  should be inspected as soon as possible.
     Fan bearings require frequent inspection;  they should be checked
for oil  level,  oil  color,  and operating temperature.   Bearing wear  can
cause vibration problems.   The bearing  housing and mountings should
also be checked on a weekly basis.
                                   5-3

-------
     On belt-driven fans, the belt tension should be checked on  a  weekly
basis.  Loose belts are prone to a distinctive squealing which suggests
an unintentional reduction of several  hundred R.P.M. and a decrease in
the gas flow delivered by the fan.  The rotational  speed of the  fan can
be checked using a manual tachometer,  a phototachometer, or a strobo-
tachometer.  Use of these instruments  is discussed in the paper  by Richards
and Segall.35  Fans should never be operated at tip speeds greater than
those recommended by the fan manufacturer.
5.2.2  Pumps
     The most common pump related problems include erosion of the  impellers,
erosion and chipping of the liner, and leakage of the pump glands.  All
pump bearings should be lubricated regularly and inspected on a  frequent
basis.  The actual schedule depends somewhat on the severity of  the
application, with more frequent attention required if the total  solids
content of the liquor is high and/or the pH is low.  High wear pump
parts should be kept on hand since delivery times are sometimes  long.  If
possible a spare pump should be available.
5.2.3  Valves
     The operational status of all control valves should be checked on
at least a weekly basis.  They are vulnerable to sticking, pluggage, and
erosion.  Air activated valves are also subject to freezing if the com-
pressed air supply is not properly dried.
5.2.4  Nozzles
     The frequency of nozzle inspection varies with the type of  scrubber
and the quantity of suspended solids in the recirculation liquor.   If
the suspended solids level is greater than several percent, erosion
and/or pluggage are probable.  As an initial check, the nozzle operating
pressures should be recorded on a daily basis.  An increase or decrease
in the operating pressure without a corresponding change in the  liquor
flow rate strongly suggests that the nozzle condition has deteoriated.
If pluggage is suspected due to an increase in the nozzle pressure,
the nozzles.should be rodded out  (if possible) and then the presssure
rechecked.
                                   5-4

-------
     When the scrubber is down, nozzle conditions can be evaluated
by an internal inspection or by observing the spray angle while the liquor
recirculation system is operating.  A high intensity (explosion proof)
flashlight is useful for checking the individual  spray nozzles.  It is
generally advisable to keep a set of replacement  nozzles when  they are
used in severe applications.
5.2.5  Recirculation Tank
     The gradual accumulation of sludge at the bottom of the recirculation
tank is undesirable.  Whenever the system is down for maintenance,  the
quantity of the deposits should be checked either by draining  the  system
or simply by the use of a rod.  Draining the tank is preferable since
it is possible to check for corrosion and erosion when the tank is  empty.
     If a mixer is used in the recirculation system, the condition  of the
impeller should be checked on at least a monthly  basis.
5.2.6  Dampers                    .   -               •   '
     It is advisable to exercise all dampers at least once a month  to
confirm that they have not "frozen" in the open position.  The damper
drive mechanism and bearings should be lubricated in accordance with the
manufacturers specifications.
     If a purge air blower is necessary to keep the damper seating
surfaces clean, this should be checked on a daily basis to ensure  that
the blower is operating.
5.2.7  Instruments
     Instrumentation systems are particularly prone to operating problems
and therefore should be checked on a daily basis.  The status  of a  pH
control system can be checked using either a portable pH meter or  pH
paper.  It is important to take these measurements at approximately the
same location as the pH probe of the instrumentation system, since  pH
can vary substantially throughout a wet scrubber  system.  Deviations
from the baseline level and/or a difference in the two readings suggest
either a problem with the installed pH meter or failure of the alkaline
additive delivery system.  The most common pH meter problems are scaling
scaling of the probe and breakage of the probe.
                                   5-5

-------
5.2.8  Motors
     The motor bearings and the motor starting switches should be lubri-
cated on a routine basis.  The manufacturers'  guidelines should be followed.
All contacts, cooling fans, and other components should be inspected at
least once per year.
5.2.9  Demisters
                         \
     All chevron and meslr pad type demisters should be inspected on at
least a weekly basis to prevent excessive solids buildup.   Gradual  scaling
of the demister will lead to an increase in local  gas velocities through
the open areas and this in turn will  result in entrainment.  Presence of
solids on the demister can be determined by using a high intensity
(explosion proof) flashlight.  The condition of the demister flush nozzles
should be evaluated at the same time.  Any deposits should be removed
as soon as possible using either the  existing spray system or an external
hose.
5.3  ELEMENTS OF MODEL RECORD AND REPORT SYSTEM
     A comprehensive recordkeeping system addressing the operation, main-
tenance, and performance of the wet scrubbers at a particular source can
be of valuable assistance to both the control  agency inspector and the
plant manager.  The information contained in such a system can provide
the inspector with a preliminary indication of the overall condition and
performance of the control  equipment.  It can also direct the inspector's
attention to specific problem areas that may require special  attention
during the inspection.
     The records will allow the operator to ascertain the effectiveness
of his operation and maintenance program, as well  as determine the extent
of deterioration of control system components.  The latter benefit can,
in a well-maintained recordkeeping system, minimize the time that the pro-
cess and control equipment is out of  service due to repairs.   In addition,
significant cost savings can be realized by diagnosing and repairing
major equipment failures before they  occur.
     For either the inspector or the  plant manager to benefit from a
recordkeeping and report program, there are several important aspects
of the use of the program that should be realized.  For the program to
                                   5-6

-------
be effective, recordkeeping should routinely executed on a daily basis.
Some advance warning is usually given prior to equipment failure; how-
ever, the time between the initial indication of a problem and the re-
sultant failure can be very short.  The earlier a problem is discovered
and rectified, the greater the reduction in repair time and cost.
     A recordkeeping program should never be considered an entirely inde-
pendent system on which decisions are based.  The data contained in the
records can, for many reasons, be inaccurate or misleading.  The major
cause of such problems is that the data are obtained from faulty instru-
ments.  Thus, agency inspectors and plant operators should always conduct
a physical inspection of the equipment to verify the accuracy of instru-
ment data.  Quality assurance is another important part of such a program,
particularly in ensuring that the data are that useful.
     The recommended recordkeeping practices that follow are intended to
provide an overview example of elements of a model record and report pro-
gram.  The items that are discussed are not for a particular scrubber, but
rather encompass many different scrubber types and applications.  Each
recordkeeping system should be tailored for the specific scrubber configu-
ration and the process to which ft is applied.  In tailoring a recordkeeping
system, applicable elements from this model program should be included,
as well as any others which may apply to the specific situation.
5.3.1  Scrubber Operation Record Elements
     There are certain data common to all scrubber types which should be
included in any scrubber recordkeeping system.  These data elements are
listed in Table 5-1.  These routinely measured parameter values, compared
to the baseline values obtained during a compliance test (or the initial
performance evaluation test), can provide a very good indication of the
performance of a scrubber.  In addition, examination of these parameters
over time can aid in the detection of component deterioration in the
scrubber system.
     It is recommended that whenever possible, the scrubber operation data
be obtained using portable instruments.  Each tap hole through which a
measurement is made should be cleaned prior to every measurement.  Because
the plugging of tap holes occurs so frequently, cleaning of the holes is
                                   5-7

-------
                  TABLE 5-1.  SCRUBBER OPERATION DATA
                  Inlet Gas Temperature
                  Outlet Gas Temperature
                  Total Static Pressure Drop
                  Static Pressure Drop of Mist Eliminator
                  Liquor Feed Rate
                  Liquor pH
                  Water Makeup Rate
                  Fan Current
                  Fan RPM
                  Fan Gas Inlet Temperature
                  Nozzle Pressure
                  Pump Discharge Pressure
                  Recycle Bleed Rate
                  Chemical Addition Rate
                  Liquor Solids Concentration
important to ensure that a partially or completely plugged hole does not
result in an erroneous measurement.  This, in fact, is one of the reasons
that portable instruments should be used rather than fixed gauges.   Too
often a reading from a fixed gauge will be recorded without checks  to see
that the gauge's tap hole is not plugged.  Regardless of whether the instru-
ments are fixed or portable, each must be calibrated at intervals which
are at least as frequent as the manufacturers' specifications.
     The measurement of the inlet and outlet gas temperatures can provide
an indication of problems in the heat exchange mechanics of the scrubber.
For instance, abnormally high outlet gas temperatures suggest that  the
scrubber liquor flow is below normal or the gas flow rate is well above
design value.  In either case, the liquor-gas contact within the scrubber
is less than optimum, which results in a lower particulate removal  effici-
ency.  A low outlet gas temperature can indicate a low gas flow rate
through the scrubber or an inleakage of ambient air.
     Pressure drop measurements can confirm problem areas identified
by the temperature measurements.  They can also indicate other potential
problems.  Low pressure drops across the scrubber are indicative of little
or no liquor flow to the scrubber and low gas flow rates through the
scrubber.  In ventuni scrubbers, a low pressure drop can be caused  by
the venturi throat being out of adjustment, while the cause in a tray-
type scrubber may be the collapse of the impingement tray.  A high
                                   5-8

-------
pressure drop on a packed bed scrubber is usually the result  of  the  bed
becoming plugged.  Pressure drop increases are associated with unexpected
increases in gas or liquor flow rate.
     A determination of the gas flow rate through the scrubber can be
made using information collected from fan measurements.   It  is very
important to know if the scrubber is operating in the flow rate  range
for which it was designed.  If the necessary baseline data were  collected
during a previous performance test, the fan current  can  be used  to estimate
the present flow rate.
     The measurement of the scrubbing liquor pH can  provide  insight
into operational problems, as well as indicate potential  damage  to the
scrubber structure.  A low pH reading usually indicates  loss  of  additive
feed or low liquor flow rates.  Since these rate measurements are included
in the routinely collected scrubber data, the cause  of the problem can
be easily determined.  Low pH measurements should serve  as a  warning
that,corrosion damage to the scrubber shell may have occurred, and thus,
a visual, internal inspection of the scrubber should be  conducted.
     Low nozzle operating pressure is usually a result of low liquor
flow rate or erosion of the nozzles.  If nozzle erosion  has  occurred,
the droplet size emitted by the nozzles will be larger than  the  design
size.  Also, low liquor flow will produce altered droplet size distribu-
tions.  Either situation will result in decreased collection  efficiency.
Since the liquor flow rate is a routinely measured parameter, an increase
in the flow reading should prompt a visual inspection of the  nozzles for
erosion; at this time, the valves should also be examined for erosion.
A decrease in the flow can indicate pump impeller wear or liquor line
pluggage.  The recycle bleed rate measurements can show  bleed line
pluggage when the flow is reduced and valve wear when the flow increases.
     The various combinations of pump motor current  and  discharge pressure
can be used to identify several probable problems.  For  instance, a.de-
crease in both water pressure and amperage draw can  mean that there  are
nozzles missing, significant pump wear, or suction line  plugging.  Other
combinations suggest nozzle or -spray bar plugging and manifold or spray
bar leakage due to holes.  The pump motor current can also be used to
determine if a decrease in scrubber liquid flow is caused by  an  equipment
                                   5-9

-------
 maintenance problem such  as  a  plugged  line,  by  a worn pump  impeller, or
 in fact,  may only  be a result  of  an  improperly  calibrated flow meter.
      The  remaining scrubber  operation  data which should be  monitored are
 liquor solids concentration  and chemical  addition  rates.  Monitoring the
 solids concentration can  ensure that the  desired collection efficiency
 is maintained and  scaling is minimized.   It  also helps reduce abrasion
 in the liquor contact  points throughout the  piping and scrubber system.
 In scrubbers which use chemical additives to control the liquor pH or the
 concentration of dissolved solids, measurement  of the rate  of addition
 is important in reducing  the problems  created by corrosion  and scaling.
 5.3.2  Process Record  Elements
      In addition to  the scrubber  data, there are certain process data that
 should also be routinely  recorded.  These data  are shown in Table 5-2.
 Since baseline scrubber data are  used  as standards on which to judge sub-
 sequently measured scrubber parameters, it is equally important to know
 how the daily process  data compare to  the baseline values.  Process vari-
 ations in feed types and  rates can effect the collection efficiency of
 scrubbers by  altering  the  character of the inlet particulate gas stream.
 Of particular importance  in this  regard is the particle size distribution
 at the inlet, a parameter  on which greatly affects scrubber performance.

                   TABLE 5-2.  PROCESS CONDITION DATA
                        Process Feed Rate
                        Percent Load Capacity
                        Process Feed Descriptor
     The ancillary equipment in a scrubber system is also often responsi-
ble for many of the operational problems.  This equipment includes the
fans, pumps, motors, ductwork, piping, valves, clarifiers, and instrumen-
tation.  Failure of any of these components can be disastrous not  only to
the performance, but also to the longevity of the entire scrubber  system.
     In order to facilitate the recording of the scrubber parameters  pre-
viously discussed, a suggested record format is presented in Table 5-3.
                                   5-10

-------
The daily, monthly, and semi-annual performance logs for the scrubber's
ancillary equipment are provided in Tables 5-4, 5-5, and 5-6, respectively.
As with any other operation and maintenance record system,  space is pro-
vided to explain any corrective actions which were initiated to resolve
any abnormal observed values.  It is highly recommended that all data be
collected on a routine basis and compiled in a notebook used only for
recordkeeping purposes.
                                   5-11

-------
                   LL.
                                  O
                   GO
                                  ou
                                  O
                                  0)

                                  c_
                                  o
                                 o
 tD
 O
                   LU

                   1
 s
 Ul
 Ou
 O
 LU
 CO
 CO
                                 Ol
 •a
  0)

  S_
  CD
  C/)
O
CO
  OJ
CO

in

LU

CD
                                O)
 fd
 CO
                                                                      ns
                                                                      c:
                                                        Q.
                                                        O
                                                        £_
                                                                      UJ
                                              0)
                 UJ
 t_
 Ol
•»-»
 Ol

 ro
 c_
 CO
D_

 c
 o
                               CD
                               Q.
                               o
 CL)
 £_
 3
•*->
 ro

 CL)
 CL

 CL)
 CO
 ro
CD
CD    i—
CL)
£_
3
4-)
ro
t_
QJ
Q.
F
CL)
1—

CO
CO
CD

4_>
CD
i~~
3
0

0)
£_
3
in
to
CL)
<_
, CL.

O
•r-
40
re
-M
oo

r—
ro
o
1—
Q.
O
£_
Q

CL)
£_
3
CO
to
CL)
c_
Q-

C_
CL)
ja
ja
3
C_
O
00
•^
^v^
Q.
O
£_
Q

CL)
{_
3
to
CO
0)
£_
Q-

O
•n"
4->
ro
4J
OO
ry

C
O
•r—

ro
r^l
3
O
£_
• p—
O
CL)
Qi

J_
O
i 3
cr
_l













m
cu

e
O
3
cr
	 i
                                                 CL)
                                                 ^
                                                 (O
                                                 CD

                                   'r-    •!-     n3
Nozzle Pressure
                                                                                                         0)
                                                                                                         (O
                                                                                                        o;
                                                                                                        CL)
                                                                                                        en
                                                                         3
                                                                         0-
                                                                                                                O)
                                                                                                                fO
                                                                                                               o;
Add
Chemica
                                                                                                                     01
                                                                            CD
                                                                            c
                                                                           •!"•

                                                                            (O
                                                                            o
occulant
•»->
 C
 ro
4-J
 o
 ro
4-
 c_
 3
oo
                                                                                                                                          £_
                                                                                                                                          3
                                                                                                           £_
                                                                                                           O
                                                                                                           3
                                                                                                           cr
                                                                             5-12

-------


u_
1 — 1
IE
00

O3
O
1
_I
UJ
o
121
eC
s:
0
(i
rv 1 1 1










UJ b-'
Q. <:
Q
I—
2:
UJ
^Z"
•CL.
Q_
h-H
^3
o-
UJ
^~
— 1 O
_1 2:






1 — 1
0 h-
2: 1— 1
^f* ' —yr
^•J* <^—
UJ
CQ
CQ
o;
O
CO
UJ
UJ
•
1
ir>
UiJ
_i
CQ
1—







































UJ
s:
e^
^
i—
^^

•r—
-M
O

t_
£_
O
o






(O
£_
o
c:
ja



£_
O

L«
O)
4^
QJ
E
(O
^.
fO
D_

C
o
•i—
03
L_
O)
OL
0


































































































































































































































..






















































































UJ
CO |—
CD CO
2: CO >^
i— i t/> t£3 CO
ct: CT) 2: . c
«£ c: I-H uj o
UJ OJ -i- C£. O3 -r-
CQ t- £_ eC Q; +J
rs ro co uj ^> to
2: 4-> Ol CD CQ Q. £_
«C to CQ c: 2: caj
u_ s_ o >— i co co ctoaiO-
Ci — £_O)£-'i- Qi i — _^C O •«- >— I O
Q o a) o Q- o -t-> «=c ojroo i — +-> =t
2:-i->i — E+J(OLU >d)T-O 03 <*-
•=C 4-»aJotuoocQ a) 	 i +•> S <-> a; o
(O i c_5 ^— ^" (p— t fQ >r— 1 1 i
cos- c_o_ i— £_ a_ £_ a_ a>
2: JDI — i — i — cj2 s: i — ooou_ — i a. • o co > ci- 	 i c^ uj
5-13

-------
CD

O
O
o
U.
a:
ui
a.
uj
LU
oi
«C
O
UI
OQ
CO



1

CO
in

in
ca
<:


u.



•— «
3C
CO






UI







a







o








\-
=


































tJ-1
5
^
r-
_i
a.

0
•r™*
CJ

E
to
c_
(O
Ou

C
0
+j
(O
cu
o.
o






































































































































































































































































































































































































































en en
c c

+J CO £
iZ =C iZ
2C
O) O O)
CO UI CO
O c o en c:
— 1 UI CU OJ 	 1 C O
>• in E T-T-
•>«•!- CCO«> X 4->
co o; o; £_ c: cncDco cu • CU CQ O CU 'i— CO *+— CU
CQco a; o c _l o oi wT (Ocoui ro
^liCS O.OT- OO-iil — JiiCUQ- CUjKi
*ZZ. CO ^ E t- CO r— C_ I — CO O CO O y CO CO
«=C CU Di CU O. x: -r— Q. O CU^> CUXej; COCU
Ll- _l Q t—COOOCO 2! —1 Q _IUI Q Ul_l
CU
13
c.
•r™
-p
c:
o
o







































                                                      5-14

-------
1 1

U-
1 — 1
*~^ "T~
•o oo
CL)
3
f"»
+J
C
O
o
— '

o
_l
UJ
o









2: uj
s: «=c
fV I— I
O
.1 1

UJ
Q.

1—
^
UJ
Q-
1 — I








o- o
UJ Z
"^— |L_J
?- (—
Qi 1—4
«=c 2:
_l
1 — 1
•z.
«=c
C£
UJ
CO
CO
Qi
O
.00
>-
_i
rn
^^*
0

•
LO
I
LO
UJ
	 1
CO
•=C

LU
1—
•z.
. 	 1
o.






















o
•r—
(J
eC

O)
O
0)

^_
0
o





(O
E

O
c.
-Q





ID
O










C
o
-!3
-o
o
C_5
£_
0
O)
1 ^
cu
(O
£_
(C
Q.
C
O
Q.
C3


























































































































































































































































































































































-:




















































I







' o;, a.-
UJ 3
oo . ea • -a ' -
Q ' ' . 02 r— . . '
M ° 0£ '3 ' .
	 1 - C t — 1 1 i 1 (Sy
> — i Oi =C ^ n ni
Q- 01£_2: eC OO)
OO 3 IO CD i t
a; cc: i— cu en >- c
1-iJ OleC Q-3 'Z. CD UJ >> Q - i — CO
'-' =: CQ <: - c l 01 ' o; O0(0o-i-
U_ ••— O) CD -i- c: =3 ' rd ' ca • *i— >r- in
"— ' Ol>- i — i — - OO • Ol «r- OO i. t-(/lO
C^ Olc=C IMMUJ Ol_i£C/0 3S:O)(O£_
•^C 3 Q; Nl N] CL. H3 (TJ UJ O t— l J_> £_ £_
— 1 i — Q- OO i— i , — OJ QiO et (T3J2O
0 Q- 00 -Z. - -Z. Q- Q-_l Q- CC S S ^ -0
5-15

-------






t
t— 1
nr
CO

§
_j

UJ
o
z:

o-
UJ
>-
3







•-J *
I— I O
o 2:

CO
CD
cm
O
CO
|
2:
•a:
I
1— <
UJ
CO


•
CO
1
to

UJ
_J
CD
•a;
r-










































UJ
«*T^
^"

h-
z:
^c
_j
cu

c:
o
•r—
O
•"•^

CU
'•p
0
cu
£_
£_
O
0





t
re
£_
o
c
jQ




r_
re
E
o
•z.








c
0
*r™
•p

•5
c
o
o

c_
o

c_
cu
4->
CD
fO
f^r
(0
o.
,_
o
•r-
+J
(O
c_
cu
o.
CD





















































































































































































































































































































































































































*












































































V

































































































































































CO
on
UJ

*~^
Q
in UJ co
en o; en
c: c: ;
•i- a; -r-

4J LU P
•r- CD T-
en LI_ en en u_ en
c: c: 03 c c
•i- CU •!- T- CU -r-
t- co £- CO t- co £-
O O O CD O • O O
o 002: o o • o
CO — 1 CO I— I CO — 1 CO
•> «> "•=£ •> "CO «
cncoco cnuj en COCD en
CO CCD-i^ CCQ C .^2: CCO
CD -i- 2= O -r- C T- O i— i T- _l
S cuts S £cup £i -?Jjcu£ < culpuj
Qi O •!- et CJ O -i- «C -P -r- O O LU O -r- CO
•^ C O- 1 1 1 c Q- "T" ro ri c CO C d.
UJ fG CQ •» fO O O re " re rv*
iv^ t— « CO t_ •* »r-- " £— CO rv* f A 1 1 t
re £— o. ^^ re c. CD £_ c_ re _^ CD re c_ cu c.
-^~ QJ re ^> re cu re ^^ ^> ^3 QJ 
-------
-a
 a>
                oo
 o

4J
 O
<

 O)
 o
O
cs
o
 o
 O)
 £_
 £_
 O
o
Qi
O
                UJ
                «=c
                a
LU
                              ns
                              O
o-
UJ
C£.
UJ
CO
03
                              £_
                              O
o
oo
                              c:
                              O
ca
•=c
                UJ
                <;

                O-
                              O
                             o
                              £_
                              O)
                              (O
                              (O
                             Q.
 re
 c_
 O)
 a.
o
00
I—1



o
UJ


UJ
>
>—I
o;
                                       UJ
                                       a.
                                                         O
                                                         _1
                                                         OQ
                                                         oo
                                                         UJ
                                                         Q
                                                         CO
                                                         CD
           CXL

 C   -(-)    UJ
 o   c    en
•r-      s   • o;
 "3   c:    uj
 (-   o>   o.
 O)   -I-    S
 CL  I—    <:
O   =C    Q
(O
(LI
                              o>
                              CD
                              (O
                             ^
                              03
                              O)
                                             00
                                             o;
                                             UJ
                                             I—
                                             UJ
                                                                                 (O
o
o
                                                                 5-17

-------

-------
                            6.0  REFERENCES
1.  Steiner, B. A. and Thompson, B. J.  Wet Scrubbing Experience for
    Steel Mill Applications.  Second EPA Fine Particle Scrubber Symposium.
    U.S. Environmental Protection Agency, Research Triangle Park, N.C.
    Publication EPA 600/2-77-193.  pages 5-24.  September 1977.

2.  Bete Fog Nozzle Inc.  Catalog 81A.  1981.

3.  Calvert, S., J. Goldschmid, D. Leith, and D. Mehta.  Wet Scrubber
    System Study, Volume I:  Scrubber Handbook.  U.S. Environmental  Pro-
    tection Agency; Research Triangle Park, N.C.  Publication EPA-R2-72-118a,
    pages 5-89 to 5-93.  August 1972.

4. ' Calvert, S.; Barbarika, H. F.; and Monahan, G.M.  Evaluation of Three
    Industrial Particulate Scrubbers.  U.S. Environmental Protection Agency,-
    Research Triangle Park.  Publication EPA-600/2-78-032, February 1978.

5.  Calvert, S., H. Barbarika and 6. Monahan.  American Air Filter Kinpactor
    10x56 Venturi Scrubber Evaluation.  U.S. Environmental Protection Agency,
    Research Triangle Park, N.C.  Publication EPA-600/277-209b.  November 1977.

6.  Calvert, S.   "Engineering Design of Fine Particle Scrubbers."  Journal
    of  the Air Pollution Control Associates, Vol. 24, No. 10.  pages 929-934.
    October  1974.

7.  Yung, S., S.  Calvert and H. F. Barbarika.  Venturi Scrubber Performance
    Model.   U. S. Environmental Protection Agency.  Publication 600/2-77-172.
    August  1977.

8.  Yung, S., Calvert,  S., Barbarika, H., and Sparks, .L.   "Venturi Scrubber
    Performance Model"  Environmental Science and Technology,  Volume 12.
    Number  4.  pages  456-459.   April  1978.

9.  Yung, S.,  S.  Calvert,  and H. Barbarika.   Venturi  Scrubber  Performance
    Model.   U  .S. Environmental Protection Agency,  Cincinnati, Ohio.
    Publication  No.  EPA-600/2-77-172.   August  1977.

10.  Kalika,  P. W.   How  Water Recirculation and  Steam  Plumes  Influence
    Scrubber Design."  Chemical Engineering.   July  28, 1969.   pages  133-138.

11.  McCabe,  W.  L.  and J. C. Smith.   Unit  Operation  of Chemical  Engineering.
    McGraw  Hill,   page  116.   1967.

12.   Lapple,  C.  E.  and H. J. Kemack.   "Performance of  Wet Dust-Scrubber".
    Chemical Engineering Progress.   Volume  51.   pages 110-121.  1955.

13.   Semrau,  K.  T.  "Correlation of Dust Scrubber Efficiencies."   Journal of
     the Air Pollution Control.   Volume  10.   pages 200-207.   1961.

14.   Semrau, K. T.  "Dust  Scrubber Design  - A Critique on the State  of the
     Art."  Journal  of the  Air Pollution Control  Association.  Volume 16.
     pages 587-5-94.   1967.
                                   6-1

-------
15.
16.
17.
18.
      Semrau, K. T., C. L. Witham, and W. W. Kerlin.  Energy Utilization by
      Wet Scrubbers.  U. S. Environmental Protection Agency, Cincinnati, Ohio.
      Publication No. EPA-600/2-77-234.  November 1977.

      Engineering-Science, Inc.  Scrubber Emissions Correlation Final  Report
      to U.S. Environmental Protection Agency.   Contract  No. 68-01-4146
      Task Order 49.  May 1979.
      Perry, R. H. and C. H. Chilton, Editors.
      Fifth Edition.  1973.
                                               Chemical  Engineers'  Handbook,
      Semrau,  K. T., C. T. Witham,  and W.  W.  Kerlin.   Energy  Utilization  by
      Wet Scrubbers.  U. S. Environmental  Protection  Agency,  Cincinnati,  Ohio,
      Publication No. EPA-600/2-77-234.   November 1970.

 19.   Walker,  A. B.  and R. M.  Hall.   "Operating  Experience with a Flooded
      Disc Scrubber.  A New Variable  Throat Orifice Contactor."  Journal  of
      the Air  Pollution Control  Association.   Volume  18.  pages 319-323.
      May 1968.

 20.   Ranade,  M. B.  and E. R.  Kashdan.   Second Symposium on the Transfer
      and Utilization of Particulate  Control  Technology.  U.  S. Environmental
      Protection Agency,  Cincinnati,  Ohio.  Publication 600/9-80-039a.
      September  1980.   pages 583-560.

 21.   Nixon, D., and C.  Johnson.  "Particulate Removal and Opacity Using a
      Wet Venturi  Scrubber - The Minnesota Power  and Light Experience."
      Second Symposium on  the  Transfer and Utilization of Particulate
      Control  Technology.   U.S. Environmental Protection Agency, Reseach
      Triangle Park,  N.C.   Publication EPA-600/9-80-039a.  September 1980.

 22.   Kemner,  W.  F.  and  R.  W.  Mcllvaine.  Review  of Venturi  Scrubber
      Performance  on  Q-BOP  Vessel C at the Fairfield Works of the United
      States Steel Corporation, Birmingham, Alabama.  Final  Report to U.S.
      Environmental  Protection Agency.  Contract 68-01-4147,  Task 03.
      February 1979.

 23.   Yocom, T.  "Problems  in Judging Plume Opacity - A Simple Device for
      measuring Opacity of Wet Plumes."  Journal,of the Air  Pollution Control
     Association.  Volume 13.  pages 36-39.   1963

 24.   Nadar, J. S. and Conner, W. D.  "Impact  of Sulfuric Acid on  Plume
     Opacity."  Presented at the Symposium on Transfer and Utilization of
     Particulate Control Technology.   Denver, Colorado.   July 23-28, 1978.

25.  Hesketh,  H.  "Atomization and  Cloud Behavior in  Wet  Scrubbers."  Pre-
     sented at the Symposium in Control  of Fine  Particulate  Emissions
     from Industrial Sources.   San  Francisco, California.  January  15-18,
     1974, pages 455-478.

26.  Woffinden,  C. J., Markawski, C.  R.  and D. S. Ensor.  "Effects  of
     Surface Tension on Particle Removal." Symposium on the  Transfer  and
     Utilization of  Particulate Control  Technology.   U.S. Environmental
     Protection  Agency.  Publication  EPA 600/7-79-044c.  pages  179-192
                                  6-2

-------
27.  Spraying Systems Co.  Pollution Abatement  Manual.  ,No  Date.

28.  Referenced, page 6.

29.  American Society of Testing Methods.   Annual  Book  of ASTM Standards,
     Part 31.  Water.  "Standard Test Method for Surface  Tension  of Water."
     1982.'                                        ••.;-'

30.  Sparks, L. and M. Pilot.  "Effect of  Diffusiophoresis  on Particle Collec-
     tion by Wet Scrubbers."  Atmospheric  Environment - Volume 4.  pages  651-
     660.  1970.

31.  Calvert, S. and N. Jhaveri.  Flux Force/Condensation  Scrubbing.
     Journal of the Air Pollution Control  Association.   Volume 24.   pages
     946-951.  October 1974.

32.  Lemon, E.  "Wet Scrubbing Experiencee with Fine Borax  Dust."  Second EPA
     Fine Particle Scrubber Symposium.  U.S. Environmental  Protection Agency,
     Research Triangle'Park, N.C.  Publication 600/2-77-193 pages 25-34.
     September 1977.

33.  Czuchra, P. A.  "Operation and Maintenance of a Particulate Scrubber
     System's Ancillary Components."  Presented at the U.S. EPA Environmental
     Research Information Center Seminar on Operation and Maintenance of
     Air Pollution Equipment for Particulate Control.  Atlanta, Georgia.
     April 1979.

34.  Richards, J.  Baseline  Inspection Procedures (Draft) Final Report to
     U.S. Environmental  Protection Agency.  Contract 68-01-6312, Task 30.
     March 1983.

35.  Richards, J and R.  Segall.  The Use of Portable Instruments.  Final
     Report to U.  S. Environmental Protection Agency.  Contract 68-02-6312,
     Task 60.  June  1982.

36.  Schifftner, K.  C.   How  to Check Scrubber Entrainment.  Pollution
     Engineering,  July  1982, pages 38-39.

37.  U.S. Environmental  Protection Agency, Wet Scrubbers", Section included
     in  Control  Techniques  for  Particulate Emissions from Stationary
     Sources  - Volume  1.  Report No. EPA-450/3-81-005a,  page  4.5-2.
     September  1982.

38.  Gilbert, W.,  "Troubleshooting Wet Scrubbers."   Chemical  Engineering,
     October 24,  1977.   pages  140-144.

39.  Reference  37, page 4.5-32.

40.  Reference  37, page 4.5-34.

41.  W.  W.  Sly  Inc.  Bulletin lOM-2-66.  No date.

42.  Western Precipitation,  Inc. Type  "T" Tubulaire  Scrubber.  No  Bulletin
      Number.  No Date,
                                    6-3

-------
 43.  Carborundum  Inc.  Process Gas Scrubber System.  Bulletin 900/12-74/2M.

 44.  Pangborn Corporation.  The Pangborn Ventrijet.  Bulletin 920 B.  No Date.

 45.  Ducon.  Venturi Scrubber Type WO.  Bulletin W-9079.  1975.

 46.  Research-Cottrell.  Flooded-Disc Wet Scrubbers, Bulletin RC-975.  No Date.

 47.  Riley Engineering, Inc.  A33 Venturi-Rod Scrubber.  Bulletin 110.
     October, 1975.

 48.  Ekono, Inc.  Environmental Pollution Control, Pulp and Paper Industries,
     Part I-Air.  U.S. Environmental Protection Agency, Research Triangle
     Park, N.C.   Publication 625/7-76-01.  October 1976.

 49.  Environmental Science and Engineering, Inc.  Field Surveillance and
     Enforcement  Guide:  Wood Pulping Industry.  U.S. Environmental  Protec-
     tion Agency, Research Triangle Park, N.C.  Publication 450/3-75-027.
     March 1975.

 50.  U.S. Environmental Protection Agency.  Atmospheric Emission for the
     Pulp and Paper Manufacturing Industry.  Publication EPA-450/1-73-002.
     September 1973.

 51.  PEDCo Environmental, Inc.  Operation and Maintenance of Particulate
     Control Devices in Kraft Pulp Mill  and Crushed Stone Industries.
     U.S. Environmental Protection Agency, Research Triangle Park,  N.C.
     Publication 600/2-78-210.  October 1978.

 52.  Scheroenan, J. A., and L. V. Binz.   "Controlling Air Pollution  While
     Recycling Asphalt Pavements Through a Drum Mix Plant."  Presented at
     the Canadian Technical  Asphalt Association Meeting.  November  1979.

 53.  Patankar, W., and K. E. Foster.  "Evaluation of the Drum-Mix Process
     for Asphalt Concrete Manufacturing."  Presented at the Seminar  on
     Asphalt Industries Environmental  Solutions.  January 1978.

 54.  JACA Corporation.  Model  Operation  and Maintenance Guidelines  for
     Asphalt Concrete Plants.   U. S. Environmental  Protection Agency,
     Washington, D.C.  Final Report.  Contract No.  68-01-4135, Task  44.

 55.  U. S. Environmental  Protection Agency.  "Asphaltic Concrete Plants."
     In Compilation of Air Pollution Emission Factors.   U.S.  Publication
     AP-42,  Supplement 12.   April  1981.

 56.  National  Asphalt Paving Association.  The Maintenance and Operation
     of Exhaust Systems in  the Hot Mix Plant.   Information Series 52
     (second edition) andd  52A (combined volumes).   1975.

57.  American  Foundryman's  Society.   Cupola Handbook.   4th Edition.   1975.
58.  Reference 3, pages 7-44.
                                   6-4

-------
59.  Ranade, M. B. and-E.  R.  Kashdan.   "Design  Guidelines  for  an Optimum
     Scrubber System."  Second Symposium on  the Transfer and Utilization
     of Particulate Control  Technology.   U.  S.  Environmental Protection
     Agency Cincinnati, Ohio.  Publication No.  EPA-600/9-80-039a.  September
     1980.  pages 538-560.

60.  U.S. Environmental Protection Agency.   Control  Techniques for Particu-
     late Emission for Stationary Sources, Volume  II.   U.S.  Environmental
     Protection Agency, Research Triangle Park, N.C.  Publication  EPA
     450/3-81-0056.  page  9.2-8.  September  1982.

61.  Reference 24, page 3.

62.  Reference 32, page 9.3-6.

63.  Mcllvaine Co.  The Mcllvaine Scrubber Manual.   "Commercial and  Industrial
     Refuse Incineration".  Chapter IX.   pages  178.0-179.4.  November  1975.

64.  Hopper, J. G.  Incinerator Enforcement  Manual.   Final Report  to U.S.
     Environmental Protection Agency.   Contract No.  68-01-3173, Task 18,
     January 1977.

65.  Carpenter, B. H., D.  W. Van Osdell, D.  W.  Coy;  and R. Jablin.
     "Pollution Effects of Abnormal Operations  in  Iron and Steel Making -
     Volume II.  Sintering,  Manual of  Practice."  U. S. Environmental
     Protection Agency.  Publication 600/2-78-1186.   June  1978.

66.  Reference 60.  page 9.8-56 to 9.8-67.
                                   6-5

-------

-------
                               APPENDIX A
                               PENETRATION
     Throughout this  report  the  term "penetration"  is used frequently.
This is a convenient  means to express  the  quantity  of emissions leaving
a scrubber.  Penetration is  related  to the collection efficiency of the
scrubber as shown in  Equation A-l.
   X  = 1 — E/100
Equation A-l
        Where:   X = penetration,  dimensionless
                E = collection efficiency  (percent),  dimensionless
Penetration is  also related to the number  of transfer units  as  shown  in
Equation A-2.
   Nt = Ln(l/X)                                             Equation  A-2
Conversion charts for converting  from one  to the other of  these terms are
provided.  The  concept of penetration is illustrated  in Figure  A-l, below.
                                              C Emissions
 A Inlet
 100 l/Hr.
                                  B Solids, 99 #/Hr
Figure A-l.  Penetration as related to control  device inlet concentration,
             solids collection, and resulting emissions;  in this  case,
             the penetration equals 0.01.
                                   A-l

-------
The penetration is simply the participate emissions divided by the inlet
loading of participate (C/A) or 0.01, while the efficiency is the inlet
loading minus the participate emissions divided by one hundred [(A-C)/100]
or 0.99.  The penetration is a more descriptive parameter at the high effi-
ciencies typical of present scrubber systems and is a much more convenient
expression to use when referring to multiple sets of control  systems, since
the penetration of a system of control  devices is related to the individual
penetration values (see equations A-3 and A-4 below).
     Parallel  Collectors
        Xt = Xj + $2 + ...  +Xn
     Series of Collectors
        Xt = (Xi)(X2)  ...  (Xn)
          Where:   Xt = total  penetration  of the  system
                  Xi = penetration  of Unit  1
                  X£ = penetration  of Unit  2
                  Xn = penetration  of Unit  n
Equation A-3
Equation A-4
                                  A-2

-------
5  -
4  -
1
0
              20
40         60
COLLECTION EFICIENCY
100
Figure A-2.  Conversion from collection  efficiency to transfer units.


                                 A-3

-------
      0.2         0.1         0.6        0.8         1.0
Figure A-3.  Conversion from penetration  to  transfer  units,
                            A-4

-------
T
                                                   PT = IT QQLLECIIQN_EFFICiENCY
      0,0
                                 40     '     60         80
                                  COLLECTION EFICIENCY
          Figure A-4.  Conversion from collection efficiency to penetrati
on.
                                         A-5

-------

-------
                              APPENDIX  B
                      BASIC  STATISTICAL METHODS
1.   Plotting Emission  Correlation  Lines

    a.   Step 1  -  Plot  Data  in  any  one  of the  following forms:

        (1)   linear
        (2)   semi-log
        (3)   log-log

        Select  the  curve  with  linear characteristics.

    b.   Compute statistical parameters using  table below.

        Point          x           x2       y          y2

          1
          2
          3
          4
          5
                                                         xy
   c.
   d.
                     IX'=
                                       Sy=
                   m = z (x-x)2 = zx2 - (_JL*)2
                                          n

                   p = z (x-7) (y-y) = zxy - _fex)fcy)
            k = s(y-y)2 = zy2 - _(zj/)2
                                  n

            IT = zx/n

            7 = zy/n

Calculate equation for linear regression  line  using form:

            y = 7 + JP_ (x-x)
                    m

Calculate correlation coefficient:

            r =     P
                      V
                            B-l

-------
e.  Calculate confidence interval  for linear regression  line.

    (1)  Calculate sum of squares  of deviations  of y  from
         regression line:
         z = k -
                 m
    (2)   Calculate residual  mean  square:


         So2=-i   -
               n-2
                'n-2

    (3)   Select  confidence  interval  level desired, then look up value
         for t using  the  following table.
           n-2
90%
95%
98%
.'99%
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
6.31
2.92
2.35
2.13
2.02
1.94
1.89
1.86
1.83
1.81
1.80-
1.78
1.77
1.76
1.75
1.74
1.74
1.73
1.73
1.72
12.71
4.30
3.18
2.78
2.57
2.45
2.37
2.31
2.26
2.23
2.20
2.18
2.16
2.15
2.13
2.12
2.11
2.10
2.09
2.09
'31.82
6.97
4.54
3.75
3.37
3.14
3.00
2.89
2.82
2.76
2.72
2.68
2.65
2.62
2.60
2.58
2.57
2.55
2.51
2.53
63.66
9.93
5.84
4.60
4.63
3.71
3.50
3.36
3.25
3.17
3.11
3.06
3.01
2.98
2.95
2.92
2.90
2.88
2.86
2.83
   (4)  Calculate interval values using the equations below.

          yupper = yo + (t)(s)  /H. + (x^E)2l
                                         m
          Blower = y  - (t)(s)  1(1 + (x-x)2)
                              B-2

-------
                      APPENDIX C
                     DATA TABLES
   TABLE C-l.  VISCOSITIES OF AIR AT 1 ATM PRESSURE
Temp°F
Centipoise
Ib/ft-sec
120
140
160
180
200
250
300
350
400
450
500
600
700
800
0.0192
0.0195
0.0198
0.0201
0.0206
0.0218
0.0229
0.0238
0.0250
0.0260
0.0270
0.0285
0.0305
0.0320
1.29xlO-5
1.31xlO-5
1.33x10-5
1.35x10-5
1.38x10-5
1.46x10-5
1.54x10-5
1.60x10-5
1.68x10-5
1.75x10-5
1.81x10-5
1.92x10-5
2.05x10-5
2.15x10-5
Source:   Chemical Engineers Handbook, page 3-211.
                        C-l

-------
 TABLE C-2.  GAS DENSITY AT SATURATED  TEMPERATURE  AND  PRESSURE
 Degrees, F
-0"
  Pressure,  inches w.c.

-20"     -40"     -60"
-80"
120
130
140
150
160
170
180
190
0.0655
0.0634
0.0613





0.0621
0.0601
0.0580
0.0556
0.0531
0.0502
0.0470
0.0434
0.0587
0.0568
0.0547
0.0524
0.0499
0.0471
0.0439
0.0404
0.0554
0.0535
0.0515
0.0492
0.0468
0.0440
0.0409
0.0374
0.0520
0.0502
0.0482
0.0461
0.0436
0.0409
0.0378
0.0344
Source:   Environmental  Elements,  Psychrometric Tables  for Wet
         Scrubber System Design  Information,  Series DIS-10-003,
         June 1979.
                             C-2

-------
      0,0700
      0,0600
     0,0500
     0,0400
     0,0300 —
                 I
I
_L
                120    130   140  150    160   170   180   190



                               SATURATION TEMPERATURE, °F






Figure C-l.  Gas densities as a function of temperature  and  pressure.
                                 C-3

-------

-------
      APPENDIX D
TROUBLESHOOTING CHARTS
         D-l

-------




















r-
10
A
vo
in
CO
I—

1
LU
CO
LU

LU
l«
|

f*g*
<£
cu

J^
O

O
LU
CO

CO

H"
g
3:
o

(3
P»*|
HH
|— .
g
=c
CO
LU

oa
o
1—
1
a
LU

00
•^
i—




















•
o at
c« 4.? cn
o at •> t-
«»- to re
•» at x:
_ii to > o
o at i — «/)
o r— re T-
r— N > -a
N 1
* * ^3 *^O ^^
C 4-> C Ol E
O 3 C 3
•r- Q. •> Ot O.
j ,> | ^ QT) *">
O 3 C: O T3
>i—
CL-r- i— 4-3
E CX-P 4->
3 00
Q.-O at t_ •
at s- j= to
NX cn e | ^ QJ
o cn o t- >
at 3 o at i—
x: r— c: > re
O CL-r- O >
to



at
4J

c_
3
o
^M
tf—

•o
•r—
3
cr

r—

at -i-
to t-
3 3
(O 4->
O C
at
5>

3
0

>J
c.
at
>
.
(0


at
c_
3
4->

re *r~ re
r** ^— ^.
o o at
X O CL
at o E
i at
re to
at 3 TJ
x: T-J c
TJ re
j*: re
o at
at -o 4->
x: c re
o re t-
-0*
at
3
4^
re
£_
at
CL
F
at
^_>

i ^
at
r—
CI
l^n

C_
at

re
i
cn
E=

^_
O
O
0
^-
cn
•i^
"T*

j^


























I
to
t-
Ol
CL
O

j *
c—
at
F
CL
•r-
3
cr
at

E

O)
£_
CO
Q.
3

XX
o
O)
x:
o
o

f_
o
at
t.
3
[ ^
CO
t-
Ol
CL
E
at
4-»

i *
at

c
•i—
£_
O)
f*
f^V
3
C_
O
CO
^
cn
•r—
ar

o



















































•
£Z
O
•p



















•
o
r—
<»-

to
re
cn

CO
CO
at
0
x
01















































t-
Ol
o
r—
t~t

VX
O
at
x:
o
re


















at
re
£_

^
o
i«_
*i
c.
•r—
re

^
o


re










CL
0
£_
T3

at
£_
3
CO
CO
at

a.

3
O













to
Ol
1^
N
N
O
C
CL
E
3
CL

VX
O
Ol
O
•°


t











QJ
4-^
CO


O
^~
<4—

T3
•r—
3
CT
i —

^
O


XI









c:
0
•r—

O
at
CO

£_
at

r~\
3
£_
0
CO
























| ^
O
O)
CL
CO
C
M
o











c*~
o
•1—
1 ^
o
at


cn
c
•1—
c
to
Ol
r—
O

O)
•a
o
t_
LU

O













































»
to
at

•i—


c
re
at
0
T3



















-o
01
cn
cn
3
^_
CL

CO
C_
at
4->
Ol
s:

-a













































e
at
o
i —
ri

NX
O
at
x:
o
re

















ai

£

^
o
r—
t[
£_
•r-
re
r~
cn
«i —
ar

re









CL
0
t_
T3

Ol
e
3
to
CO
Ol
t_
CL

^*
cn
2C























[ ^
o
Ol
CL
CO
c:
t — i
ja

















CO

o
3
-o (-
at
g~ r^
•r— o
3
cn £_
c o
•r- CO
cn
cn £_
3 0

51

ja













































t_
at
o
^~
o

NX
O
at
0
re


















Ol
4->
c.

^
o
P_
*i
t_
re

3
O
1

re










Q.
o
£_
-a

Ol

3
CO
CO
Ol
t
CL

3
O













to
at
i—
N
N
O
C
CL
E
3
CL

NX
O
at
o
•°














at
•4-*
fC
£.

O
^~
M —

T3J
•i—
rs
cr
•r-
r-™

3
o
	 i

X!









, — .
(_
0

re
c:
•r«
E

r—
O)

^_J
CO
•P—
F
^7^























j ^
O
0)
CL
CO
c:
i — i
0


















-o
O)
re
o
o
r—
CO
•r—
T3
re
•i—
-a
O)
s:

o













































t_
01
o
r— •
o

sx
O
01
x:
0
re

















Ol
i *
re
£_

3
O
r~—
t) 	
£_
•r-
re
x:
cn
•r—
"T"

re









CL
o
t
^Q

Ol
£_
3
to
CO
Ol
£_
CL

t~
cn
'£













to
01
r—
IM
N
0
C
CL
E
3
CL

NX
O
Ol
0
XI














Ol
4_>
re

3
0
^—

T3
• r—
3
CT

•~
x:
cn
• r-
"T"

XI












































r"
re
at
o
"••^
^_>
o
at
CL
to
cl
I-H
0

























cn
c~
•1—
cn
cn
O
r—
O

o




























•
-a
—
4_>
C
0
o









e .
•r-
re
•o
***^
p
o
at
CL
to
c
i— i
T3

























cn
c
•a
o
o
r—
U_

•o



























-------





r^.
vo
•V
^o
LO
CO
I—
2=
LU
s:
LU
tt.
CO

C i —
CU 03
O CU >
03 O
i — 3 C
CL T3 CU
CU CU Q.
0£. C£ O

03 jQ O

*











CO
CU i — C7>
,co ro C
3 CU •!- -O
03 «/) Cn 
0 N i—
03 O 03
0. Z =»

03 J2 O



E CU
O £_
4-^ 3
Q. CO CO
E NX (/) CU
>, 03 CU CO
CO 
r— CU
r— r— T3
CU NI CU
CL N CU
E 0 Q.
I-H -Z. CO

03 J3 O



^
[ ^
03 -a
t_ cu
S CU -C
o t. to
1 	 3 •!-
t- CO C
CO •!-
CL CU E
E t- -r-
3 CL T3
Q-
£_
CU
03

. £_
« CU
to i —
CU •—
Q. CU
•t- a.
cu a. £
O -1-
03 j^
i— o -a
CL CU C
CU JC 03
o; cj>

T3 CU








o^
c
•r—
CL
C^ 'I-
c: o_

\* ^-
U T-
03
CL C
o
CU •!-

•i— O
-4-5 3
o c_
CU 4->
<»- CO

Q 0

-a a>



















-a
cu
O5
03
E -
03
-o

CU
O
03
i—
a.
cu
£_

T3
d
03
CO
^j^ _!_'.%
0 t-
O) 03
J= Q.
C_3
CU
C
•r™
r—

-o
O)
CU
r—
jQ

C.
03
CU
i—
O

-o
1=
03

4_>
O
CU
Q.
CO
C
1 — 1
CU
r—
03


•o
CU
CU
f—
o

CU
o
03
r-™
Q_
O)
£_

•a
c
03

vX
O
CU
JC
o


CU
o
03
r—
CL
CU
£_

£_
0

C
03
CU

O

-o
C CO
03 CU
r—
-^ N
O N
CU O
-G C
CJ>

CU
O
03

Q.
O)
^.

i_
0

g
•r-H
03
CL
CU
£_

T3
g~
03

^: CL
o E
CU 3
x: Q.
0

!_
O


03
>

-a
cu
c:
cu
CL
O

>^
^_
4->
O
CU
c_
t_
o
0
c
•r~

CU
c_
03
-C
o
to

•^5

CL
E
3
CL

Tjj
CU

[ %
4->
0
c_
J=

t_
CU
>
o






















•
CU

r—
(O


                                                                          c
                                                                          o
                                                                          o
           03
                    03
                          03   03
                                                 03
CO
LU
o;
co
=c
LU
CO
o
t_
CU

CU

NJ
N
O
           £_
           O
         cu
         CD
         03
         CD
 CU
 in
 3
 03
 O
 CU
-Q  CO
    CU

 O  N
1-
 o

 CU
 CO

 03
 O  CO
 CU  Q-
jf^  £--

 S  Q-
 o
                                                <«-
03      i—
         Q.

         CU
           a>
r—  IM   r—  O)
'
    C      •!-   S_
         £- 4->   O
         O  03   3
         31-   CT
         O- CU   -i-
             Q.   i—
                               £_
                               O T3
                               3 CU
                               O" CD
                              •r- O>
           03       •!-
                              r— 3   r—  O
                                  r—       C
                               o a.   o  T-
                    03    03    03
                                                 03 '


CU
^_3
03
£_
s
O
<^ CU
CO
CU 03
F— CU
O t_
>> o
o c:
O) ••-
o;
c:
0
^_>
^)
ZS
•o
OJ
^.
^
0
r— «
<4-

T3f
O)
CU

CO
CU
CO
03
CU
1_
O
c:
•r—

o
r—~
4-

T3
O)
CU

CO
CO
03

4->
CU
i —
4->
3
O
-a
C*"
03

1 ^
CU'

f~
1 — 1
03
3
cr
CU

CO
CU
1_
3
4->
03
5_
CU
Q.
E
CU
_l->

                                                 CU
                                                 t_
                                                 3
                                                 CO
                                                 CO
                                                 CU
                                                 £_
                                                 CL

                                                 CU
                                                I—
                                                 tsl
                                                 N
                                                 o
                                                 a


                                                 o
  D-3

-------
                                re
t-.
10
  «l
10
in
 co
 UJ
 o;
 o
JQ
 re

 d)
 O)
                      co  o
 o)    a)
 c  01 >
•i-  c:  t_
4-> •!-  3
4J 4J  O
 O) 4->
 to  O)  C
     to  re
 (_    lj_
 O) 4->
' CL re   •
 E  o  to
 re  t—  >

 I  4->  C
 c     o
 re -I- T-

 O)
co
                                               re  a>
                                               O 4-3
      0)

      0 O
                                               t/) •!—
                                               3 T3
                                               •r-5-o
                                               •a  re
 UJ
                       fO
 UJ


 UJ
o

o
UJ
•S
CO

a; QJ
«C to
:c 3
«j re
0
CO
z:
c
o
o

CO
f—
z
^
UJ

^^>
CO

UJ






0)
g £
•— 3:
<*- o

t- c
o
o
o
CO
UJ
CQ
o
cc
                      O   v-
                                                  a>
                   r—         t-

                   t-   0) ^
                   O   •!- •—
                   3   o  re
                   cr  •!-  o
                   •r-   
                     •M
                      re
                      x
                      ai
     a>
     <_



    I
     O)
     Q.
                      CD   CU
                   O
                   £_
                  •i—
                   O
                   a>
                   c_
                   O
                   3
                                                                    D-4

-------











r-.
IO

vo
tr>
UJ
o

e^
s:
a:

U_
Qi
UJ
a.

Q

•=c
o
1 — 1
r-
1 — 1
Q
O


a:
UJ
CO
CO

o;
0
CO

z:
o

Q
UJ
CO

CO

1—

 CO
4-00 •> £-
CU 00 03
j>^ t— cu x:
O N > O
O N r— CO
_J O 03 -r-
c > -o
1
• -a -a Q.
c p c: cu E
O 3 03 C 3
•r- CL CU CL
P .p CO CL
O 3 C O -O
< o ••- cu
CL >>r—
CUT- r— p
E CL-P 4->
3 00
CL"O CU £-
cu t- x: •
.i*: co t_ .p cu
o co o t- >
CU 3 O CU r—
x: i — c > re)
O CL-I- O >
.
03 .






cu
-P
03
C_
3
O

l! 	

£_
O

cr
•i—
CU r- .
00 1
3 CO
03 C
O ••-

XI
3
£_
O
OO

5
O
	 1


03







CU
o
c
03
E
O
ll
*^—
C_
CU
CL
E CO
O C
-P -i-
Q. C
E 03
>, CU
OO r—
o
f_
o
0
Q-

t_
O
3
cr •-.

r— [ \
1 03
CO £_
c: 3
•r- 0

X5 4-
3
t- CO
O 03
OO CO

3 3
0 O


£_ ••>
o cu
4- -P
03
-i^ S-
o 3
CU 0

0 

OO 00 o £_ O 03 O. 0 £_ -pj I CU 3 CO oo cu CL 3 0 _J XI •*^ cu •o "(_ J • ** CO CU £- i— •p cu x> 03 i — t- i — -p /^ O O3 •r- S_ O >— -P t- 03 c x: 0 0 -P c: 0 i 3 CU •P t— t- 03 XI O O 03 C_ •!— CU f* C_ > -P 03 •r- 1 > -P CU 03 i— -a t- XJ CU cu 03 co • CL-F- 03 J* O t- E CO C 03 03 -i- •i- > -a -o i— CU •r- N 03 >, 4- r- 03 • 00 C >"> C 03 -P O -r- •r- • •> 4J 03 CU 03 O r- 3 •r— XI O" 4- O •i- t_ -a T3 CL C O 03 E CU x: cu c •.- •i— 4-> OO -P 0 03 CU CU £_ £_ i — CU £- O CL O T- O O -P 4- O 03 i— i -p CL . O 1 •r— C_ I ^ OO •i- C T3 CO •r— CU OO IM CU •r- T3 OO t_ £- CU O X2 CO 3 C £_ •r- O T3 CO 03 O T3 r— C o •P ^i ^s co a?-p 3 f^* «r— ' -o i— c: -F- ] > ^5 Q CU -r- 03 i— -P CL C 3 03 i — i XI O 0 C 03 1 4- c cu E > cu | ^ CO OO c >> •i— OO J * •p " CU CO oo c: . •!— t_ 4^ CU -P CL CU E co 03 "O -P 1 03 C 0 03 t- 4- XT •P .*: i O •!- cu c- x: 3 0 P . -a 3 O ^_ 4- oo 03 cn O) •r— CO OO cu o X UJ "0 • cu ^_ 3 . 0 c~ 03 4- • OO > c: o •r— •P 03 £_ cu CL O 1 CO c c: c •r- 0-1-4- •r- <— 0 4- +J 03 •r- -r- O CU • oo oo oo t- 00 O 3 O >> CL4- '3 03 E H-I CU cr £_ O -P •!- CL O 03 r— OO « CO ^a t_ -r- co c: o O -P c 3 CU 3 OO -r- o x: cr cu x: "O O T- > co x: i — c: 3 00 -i- i — 03 • >, 4- S -a 03 » O) t- 00 3= ^i i— CL C- CL O i — OO 3 1 CU 03 03 x: -P 4- o o C_3 CO O O i — . cu .p c cu E c •1— 03 < -p cu -a cu \>* o o i — o ^> f r— O i— ^^ 03 03 •r- £- p O3 t- CL 03 CU O- CO CU E 0 CU -P CU i— x: xi t- +-> o cu t- P P CL 03 3 3 0 4- I-H D^ cu c 03 • x: c: c: co 00 -r- 3 03 1— -O £- 4- C T3 03 T3 t- "O C O 03 3 -P O 03 •> -a c oo •i- C- 4-> E 3 3 T- 0 x: r— cu OO CU t- . 03 .^ — , •a >, CU r- CO C CO O 3 i — OO CL i- O c -P •r- 03 03 c: t- -r- -0 E •c— £_ i— O O •P 03 CU C C •t- 03 E > •i— r— t— CU 03 •r- CU -P C- £= 3 CU -p CO co c: •r- 03 0 +J ^~ •**—' 03 •P c: cu £ d •r— 03 £_ -P C cu o 3 CT |n CO 03 CO 1 •p CO 03 x: x UJ c •r- 03 C_ •a cu •p 03 CO •i— t_ £_ •r— >> CO 3 0 3 C •i— • -P CU c CL 0 ••- O CL t- co CU 03 c: N OJ -i- >> f^T ^y; r^ 4- 0. 0 03 I-H 03 x: c: • co O OO CO CL O >> £= 03 E C 03 -r- 03 E 03 CO oo oo t- c: CU COT3 •!- • t. o c E O T3 T- T3 03 3 ECO crx: 03 03 4- •r- p O i — O 4- oo t^. £- £-0 03 4- •> CU 4- >s E CU £- r— 0 £- CO O Jsri O O I- 3 0 C 3 03 CT-i- 3 cr CL-i- 3 O -r- oo r- cr-O.r- . XI >, t— x> •r- *~^ CO t- 00 O 0 3 CL cr •"•-^ .p- r— o> C C •r-* *i~~ ^: i ^ i % "o c t- a> 4- CO 03 t_ CU cn •P != 03 -r- 3 E 03 CU O > 4- •r~ CO O 00 -P CU O CU X 3 LU T3 ^ X) • 13 -• •P C O o D-5


-------











r-.
10
•*
to
in
UJ
o
z>
^c
J-
on
o
U.
cc
UJ
CL.

O
^^
«c
to cn
c c
O •!-
•r- _c 1 cn
4-> to to c:
•i— 3 re _i£ -i—
"O i — Or— O i —
C q- r- Q) re
0 C- -i- XT 0
0 Q) 0 5 O to
t- o
>, O Q- O> f_
18 E O • O
t. c 0.4-
e: CL o> • re 3
O to to t_ E "O >>
•i- 1 => 3 t- r— f_
•«J cn o o T- ^j
0 C -C 4- 3 t/)
«C 'I— • t_ _Q •!—
J= C C_ 0) E
to l_ <1) CL OJ OJ
3 0) CL +J J=
!—•»-> cn re o
4- 4-> to C C_ •
IO "O •!— OJ £_ to
JK! CL 0 C •— 0 ^->
o -i- re o> 3 c
Q) 73 C_ OL) O CT OJ
.G C 03 i— O T- Cn
o re CL o re i — re


i \
re
o
• c_

o -t->
•r~ 1

•r- t—
10 3
O 4->
CL c:
0)
t- >
O)
CL d)
E r—
ro ^^
TD re
•r—
C t-
re re •

c

O O £1
O) d) (L)
jc Jr CL
o o o
 O
 HH
 H-
 H-I
 O

 o
 o
 UJ
 CO
 CO
 O
 o
I
o
CD
z:
»— i

o
o
UJ

CO

o
                                                         r—     CX OJ  -(->
                                                         i—     E •—  re
                                                         •r-     re .a  o
                                                         3     -o  re  t-
                                                             O     •!- JZ.  i
                                                         t- 4->  C  £_ ^J 3
                                                         a>     re  re     o
                                                         CL=   4-  > ^J t—
                                                         E  a>         o 4-
                                                         re j«j  o jc  D
                                                         •a  o  to 4-> jc t_


                                                         re =   -o        cr
                                                         4-     re 73   • T-
                                                            ^:  o  a)  cnr—

                                                         ^=  3  II  o T- -o
                                                         +-> .o  ID  d)  c: c
                                                            c  > j=  ,            O £_
                                                         i— O)  +->  CU     O

                                                         re T-  a>     re re
                                                         E >  > -P  o s-
                                                         t- o  a> to t_ a>
                                                         O £_  t- 3 .C CL
                                                         z: CL  CL E -(-» O
en
c
•r-
d
O) to
CL
0
I
O)
r—
N
N
O
C
"^^.
t.
O)
c:
• r—
re
£_
4->
to

^

»^-
•a
o
s:
. O)
r—
O
.£1

0)
n^
N
M
O
C

4->
re
g~
-t^

o
to

o
•r-
.{_3
re
£_
ameter
•r—
T3

ai
.c
4^

O)
O
•r—
3
4->

-l->
to
re
0)
r—

•»->
re

0)
i_
re





•
to
cn
c
•r—
C
0)
CL
O

C.
(U
C
•r—
re
£_
^_>
to

4 —
O
•
c:
O)
E
CL
•r—
3
cr
0)

c
O
•r—
P
re
£.
re
CL
(1)
to

SX
U
O)
J=
C_3
-a
re
o

•)->
to
3
•a

a>

•r—
to
to
0)
O
X

£_
4->


to
re
cn

c:
•r—

en
c:
•r—
C
O
•1—
I)
o

3
4-
r—
re


O)

•r-
P^
1
ai
cn

3
Q.
3
r—
4-

•a
c"
re

to
-a
re
O)


>,
(3
c_
CL
to

QJ
>
O
E
O)
OL

•
to
O)

M
N
o
C

-a
cr
re

cn
C

Q.

CL

>,
re

Q.
to
•I—
o

to
0
£_
£_
O
o

4-


cn
c:
•i—*
CL
•i—
CL

O)
O
re
r—
CL
O)
C£

















•
t ^
C

£_
re
CL
CL
re
 0)
 to
 =1
 re
o
        o
        t_
        o
               re
               c
 0)
 
 


                            (O
                            to
                            ai
                            o
                            x
 t- eu
 
JD -r-
JD t/)
 3 tO
 £_ <1)

 O O
 tO X
    0)


 §0
                                                re
                                                                         re

                                                                         to
                                                                                o
                                                                                o
                                                                                to
                                                                                en
                                                                                c:
 CD
 CL
 O

 a>
r—
 N
 N
 O
                                                                                                 en
                                                                                                 o
                                                                                                 o
                                                                                                           cn
                     £_

                     O
                     3
                     cr
                                                                                          O
 to  £_
•a  CL
•i—  to

 o  c
OO •!-
                                                                                                                       >,
 O) OJ
•a ja

 t- E
 o re
    a>
 <1) £-
i— -P
 re 
 o
 to -o

 O) 3
 Q. cr

CX J—
O
UJ

CO
       o
       4->
       Ci

       I
                                                re
                                                o

                                                £_
                                                a>

                                                o

                                                £_
                                                O
                                                       re
                                                       u.
                                                                 to
                                                                 OJ


                                                                 N
                                                                 N
                                                                 O
                                                                 £_
                                                                 Q.
                                                                 (O

                                                                •a
                                                                 a>
                                                                 cn
                                                                 cn
                                                                            D-6

-------














r-.
10

10
in
UJ
o

 p— en
C i-
 O.
Q. 3
•r- -0 £_
3 O •
cr cu <»- en
 c
C -r- -T3 -r-
o c to c c
•r- o in re o
P •!- CD •!-
o -p o E -P
 E
^4 ^t to
o o en cu
cu cu e

O O •!- i—
re







o
o
•p

C i-
O 0
•i- 3
p cr
re T-
CU C- r-

3 C" >^
re cu re
o o t-
C Q.
O in
O
c
in T-
•o
•r— f~
.' i— en
O •!-
CO -C


re





£_
re
cu


cu
• r—
N
E N
0 0
•p cr
Q.
E cu
CO -r-
in
in
CU
O
X
UJ



£_
0 1
3 C
cr O s-
•r- -I- re
i— tO CU
re s
E t-
o jo c
t_ re T-
<4_
i — in
in i — en
cu re c:
> 1 * «r—
•r- in C
in C T-
re T- r—
£_
O f [ *
re o c
re
cu E -P
> re m •
§.
re
£_
Q.
to

C
•r—

in
cu
>
•i—
in
re
£_
-Q ,
•=C


JD



























~^j
re
• 0

c:
cu -p
E in
Q. 3

3
cr 0)
cu >
•(•«
c: to
O in
•r- CU
4-> 0
re X
c- cu
re
CL I-
cu o
in <»-

XX NX
0 0
cu cu

0 0
o




^,
_E C
•P 0
•i— »r—
2 in
0
C 4-
0 CU
•^.
•P en
re c
C •!-
•i- in
-Q =5
E re
0 0
u

c 
•i- cu

"^T", «^-
d. t/J
03
s t—
0 -Q
_i re


o




























i cu E
cu > re
en o cu
t. E t-
3 CU -P
CXQ; m

t_ 5_
O • O
M- en 3
£= cr
^3 »^— »r™
C C i—
re o
•i- E
E -P o
re o «-
cu c <»-
t- 3
4-> <4- in
in r— cu
re >
in E-r-
re in
en cu re
'c t-
C: -r- J3
•i- r— re




*


















c:
o
•r—
in
o
£_
£_
O
o






























^_>
c:
as ~a •
•P -O C
«/» -si o
•r— »r-
IO 4J
cu • re
£_ in o
1 CU •!-
c c n-
O O -r-
•r- N T3
in o
re s- E
t- re
.a cu 3:
re s Q.

r- C £_
•— -r- O
re  re en
•r- C
c: in E -r-
o in re c
•i- cu cu o
•P O £- •!-
re x -p 4->
t_  -P en
• a; c in t_
J= CU 3 3
o E -o ca-
re






c
o
•r—
) ^
re

3
o

•^
o

•i—
c to
O in
•r- O)
•P 0
re x
c- cu
re
CL £-
CU 0
in <*-

VX XX
0 0
cu cu
J= JZ
0 0
^














C. •
0

crt
O
c_
flj
""i«»^.
C
O
•r—
t/)
O
^.
L.
0
o


JD

























•r—
re
VX p^
1 1— 1—
cu re re
en +j.
t- -o in
3 T3 C
Cu^C *"~*

£_
O • •

re o re
•r- O
E -p-i-
re o 4-
cu c: -r-
£- 3 T3
-P 
CU 3 O
c cr re

r^ r— CL

p 4- p
c ••- o
re c:
-P in
in cu cu
•r- C «_
in o re
CU N
i- in
i c- c:
c re o
O CU "i—
•r~ 2 1 *
in I re
re j= o
t_ en-i-
JD T- <4-
re j= T-





























































TD
c
re

CL
O
t_ •
TD •
1 -P
CU C
t- •!-
3 O
in Q.
in
cu £=
t. en
Q.-I-
in
•p cu
re T3
o
£- 0

-P

NX O
O 3
CU T3
J= CU
O t- .
O









>^
1 ^
lr_
0
o

"cu


in
re
en

cu
>
• r—


0
O




















































































D-7

-------










f^»
«£>
«i
«S
to
UJ
O
z

s:
o;
o
u_
cn
UJ
o.

/-*i
^r?
•=5
£_
0

! £_
•r-
03
Q.

•r- C_
4->  v^
O
C CD
o x:
<_>
CO
CL
03 •
[ ^ r>
3
t_ "XD
O i—
IO -r-
C 3
CD -Ct
CO
CO
\y -^3
0 T-
>-P O
i^ *^
f^ \s 40
CL O •!-
3 CD 10
co x: o
1 0 CL
«_
•I— \S
03 • O
CO CD
-P t- x:
C CD O
CD -P
E r—
3-r- •
£-  •
i — CO C
x: i— o
en 03 -r-
3 c: P
0 t_ 03
t_ CD S-
O P CD
x: c a.
1— -r- 0
o
O
O
O
UJ
CO
CO

cc
o
CO
LU
CO
«c
CO

fe
 £_  03 03
                                                                                                                         T3

                                                                                                                         P
                                                                                                                                 O
                                                                                                                                 O
                                                                                                 M-  O)
                                                                                       03 i—i  > CD  03 >
                                                                                       C
                                                                                       £_
                                                                                       CD
                                                                                                                CD "O  CL
                                                                                              (O
                                                                                       4-J  C JQ

                                                                                       CD  03
                                                                                           £_
                                                                                       CD -P  O

                                                                                       03  CD
                                                                                       E  >
                                                                                                     a)  03  en
                                                                                                            c: i— •
                                                                                                           •r- 03 .
                                                                            cn4->
                                                                            c x:
                                                                            T- en
                                                                                                         03
                                                                        e:  03
                                                                        03  CL >,
                                                                        E     r—
                                                                            £_ CD
                                                                        £-  O
                                                                                                     COOPOM--i—
                                                                                              -p  S_
                                                                                                            O 4J
                                                                                                            CD 03 .
                                                                                                               to
                                                                                                               CO TD
                                                                                              COT303PCL3OCDC
                                                                                                     CD
                                                                                                            CO 4-»  CD  O  03
                                                                                              •i- 4-  t-  e  cox:  Xi
                                                                                              U_'O03  3-r-  03O  CDCn
               03
 
-------
o

•p
o
o;
o
Q-


Q
                       o
                       i.
                       0
                       o
                       o
                       cu
o
                            CU
      cu
      o
      re

      CL
      cu
     o;
             co
             i —
             re

             £_
             cu
             4->
             re
                                   cu
                                   o
                                   re
                                  •—
                                   CL
                                   CU
                                  OL
                                        CD
                                        -P
                                        01
                                         £_

                                         CU
                                 re

                                 cu
                                         CO
                                         re
            -P

            cu

            -P
            re
            cu

            •P


            co

            re
            JC
            o
                                                c:
                                                re
                                                cu
                                                             CO
                                                             co
 CL

 CL

 cu

 c
 re

o
o
cu
     cu
     o
     re
                                                                      CL
                                                                      cu
                                                               o   a:
o
cu
                                                            O


CO
cu
T3
re
r—
O

C
re
i-
c:
re
cu

o
<4-
0

c:
0

•P
re
t_
cu
CL
O
j,,.
0
CU

o




t_
r_
re
CL
cu
o;
                                   re
                                                re
                                                       to   .a
                                                                                   o
                                                                                          re
                                                                                                            re
o
o
a;
CQ
CO
o
00
LU
CO

CQ

I—
C£.

3=
o

CD
O
O
CO
LU
        re
        o
      c:
      o
                     re

               c:    o
               O   ••-
              •i-    t-
              •P    cu
               re   -p
               N    cu
              •i-   -a

              're    co
               t_    c
              •p   -i-
               ^    c=
               CU   -r-
               C   I—
     •<->
     C
     CU
     -p
     c:
     o
CO   O
                                   re
                                   •r—
                                   t_
                                   CU
                                   4->
                                   re
                       cu

                       re
                       3
                       cr
                       cu
                      •a
                       re
                            cu
      o
      cu
ds
so
ed
                      I—I   CL,     I—I
                           +1   O
                            re   cu
                            CL   «-

                            o   j=
                            o   co
                            c:   -r-
                       re
                                    re
treatment
ca
                          cu

                          o
             a>
             CL
             o
             £_
             CL
                                                re
content
                                  O
                                  co
 c:
 o
•r-
•p
 o
 re
 :_
+j
 c
 o
 o

 c:
 o
•t—
 to
 c:
 re
 CL
 x
 cu
                                       CL co
                                       3 T3
                                       s_  c:
                                       f-i  CU
                                      =C -Q
                                          CU
                                          CO
                                    ^    re
                                     c:    E
                                     cu    re   c:
                                     E   -o   o
                                                                    CO
                                                                    CU
                                                                   -o
                                                                    re
                                                                    c
                                                                    o

                                                                    CL
                                                                    Z3
                                                                   •a
stribution
ze
e
                                                                                         •r-     o
            CO   CO
           •i- .  cr
                                                                        re   c-   •!-
                                                                        co   re    >
                                                                       •r-   cu    re
                                                                       s:   CQ   o
           4->     re
           re     •!—
                                                                        re
                                                                                   o
                              cu

                              re
                                                                                          re
               c_
               re
               CL

               cu
               co

               re
                                                                                                 re
                                                                                                            CO
                                                                                                            CU
                                                                                                            en
                                                                                                            en
                                                                                                            3
                                                                                                            CO
                                                                                                            CU
                                                                                                            CL
re

to
cu

N
Nl
o
                                                                                                                 CU
                                          re

                                          cu
                                          CL
                                          o
                                          c
                                          to
                                          CL
                                          E
                                          3
                                         Q-
 OJ
  I
 Q
 CO
 •=£
         O
        +J
         Q.
                       CO
                       O
                       O
                       o
                                    CO
                                    O
                                   LU
                                                        CO
                                  CO
                                  CO
                                  3
                           CO    r—
                           £=      CL


                          r—      CU
                           re      CL
                           o     -r-
                          CO     Q-
                                                                        re
                                                                        cu
                                                                        cu
                                                                        CO

                                                                        o
                                                                        c:

                                                                        CL
                                                                                                  c:
                                                                                                  o
                                                                                                  CO  T3
                                                                                                  re  o
                                                                    •r-      re
                                                                                          re
                                                                                          £_
                                                                                          re
                                                                                              £_
                                                                                              cu
                                                                                          CU J3
                                                                                          > JO
                                                                                         •r-  3
                                                                                          CO  £-
                                                                                          CO  O
                                                                                          CU  CO
                                                                                          o
                                                                                          X  c:
                                                                                         LU  O
                                                                                                            O
                                                                                                            3
                                                                                                            cr
        *U.S.  GOVERNMENT PRINTING OFFICE:  1984—739-002/13,503
                                                                                 D-9

-------

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
  REPORT NO.
    EPA 340/1-83-022
                                                            3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
        Scrubber Inspection and Evaluation Manual
                                 5. REPORT DATE

                                    September 1983
                                                            6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
                                                            8. PERFORMING ORGANIZATION REPORT NO,
    John R. Richards  and  Robin R. Segal!
9. PERFORMING ORGANIZATION NAME AND ADDRESS
                                                            10. PROGRAM ELEMENT NO.
    Engineering-Science
    501  Wlllard Street
    Durham, NC 27701
                                  11. CONTRACT/GRANT NO.
                                                                68-01-6312
12. SPONSORING AGENCY NAME AND ADDRESS
   US  Environmental Protection Agency
   Office of Air Quality Planning and Standards
   Washington, D. C. 20460
                                                            13. TYPE OF REPORT AND PERIOD COVERED
                                  14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES

   EPA  Project Officer;
Kirk Foster,  Stationary Source  Compliance Division
16. ABSTRACT
   This  report concerns  the inspection and  evaluation of performance of paniculate
   wet scrubbers installed  at stationary sources to identify and  rectify operating
   conditions contributing  to excessive emissions.   It includes discussions of
   gas-atomized scrubbers,  plate-type scrubbers, packed tower  scrubbers, and spray
   tower scrubbers.  The evaluation approach  proposed utilizes comparisons of
   present performance parameters and conditions with site-specific performance
   established previously under a controlled  set of conditions.
7.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.IDENTIFIERS/OPEN ENDED TERMS  C. COS AT I Field/Group
   Air  Pollution
   Wet  Scrubber
   Particulates
   Operation and Maintenance
18. DISTRIBUTION STATEMENT

   Release Unlimited
                    19. SECURITY CLASS (ThisReport)

                      Unclassified	
21. NO. OF PAGES

  162	
                                              20. SECURITY CLASS (Thispage)

                                                Unclassified
                                                                          22. PRICE
EPA Form 2220-1 (Rev. 4-77)   PREVIOUS EDITION is OBSOLETE

-------

-------