EPA-450/3-87-01 Oa
Chromium Emissions from
Comfort Cooling Towers —
  Background Information
  for Proposed Standards
        Emissions Standards Division
    U.S.ENVIRONMFNTAL PROTECTION AGENCY
         Office <_ Air And Radiation
    -Office Of Air Quality Planning And Standards
    Research Triangle Park, North Carolina 27711

             March 1988

-------
This report has been reviewed by the Emission Standards Division of the Office of Air Quality Planning and
Standards, EPA, and approved for publication. Mention of trade names or commercial  products is not
intended to constitute endorsement or recommendation for use. Copies of this report are available through
the Library Services Office (MD-35), U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina 27711, or from National Technical  Information Services, 5285 Port Royal Road  Sorinafield
Virginia 22161.                                                                    •   f  a    .

-------
                       ENVIRONMENTAL PROTECTION AGENCY

                           Background Information
                                  and Draft
                       Environmental Impact Statement
             for Chromium Emissions From Comfort Cooling Towers
                    1 <	}
                            Prepared by:
  ack R.  Farmer
 Jirector,  Emission Standards  Division
 U.  S. Environmental  Protection  Agency
 Research Triangle Park,  N.C.  27711

 !•   T{je  proposed  standards of performance would eliminate hexavalent
     chromium  emissions from existing  and new comfort coolinq towers bv
     ES5J51 lng*  5f!  ule  ?f i2exavalent chromium in these towers.  Under
     Section 6 of  the Toxic Substances Control Act, EPA is authorized to
     impose regulatory controls  if the Agency finds that there is a
     reasonable basis to  conclude that tfie manufacture, processing,
     distribution  in  rnnปซ^o  use, Or disposal of a chemical substance
                              in unreasonable risk of injury to human
2.  Copies of this document have been sent to the following Federal

    ?iEnrtS^itLabor> Health a?d. Human ^vices, DefenS, T?anlporta-
    tion, Agriculture, Commerce, Interior, and Energy; the National
    Science Foundation; the Council on Environmental Quality; members of
    Aซnr5Sfnand/frri?ฐ^alDA]r f91lซt1on Program Administrators! the
    Association of Local Air Pollution Control Officials; EPA Regional
    Administrators; and other interested parties.             Regional

3.  The comment peri9d for review of this document is 60 days from the
    date of publication of the proposed standard in the Federal  Register
    yQiQ?0^IiBlJJQ0r Ms- ,"?ebb2! Stackhouse may be contacSo^aT1    9
    (919) 541-5258 regarding the date of the comment period.

4.  For additional information contact:

    Mr. Al Vervaert
    Industrial  Studies Branch  (MD-13)
    U. S. Environmental  Protection  Agency
Research Triangle Park, N.C.
Telephone:  (919) 541-5602
                                g
                                7
                                  27711
5.  Copies of this  document  may  be  obtained from:
U. S. EPA Library (MD-35)
Research Triangle Park,  N.C.
                                 27711
    National  Technical  Information Service
    5285 Port Royal  Road
    Springfield, Va.  22161
                                   ill

-------

-------
                             TABLE  OF  CONTENTS

                                                                      Page

 LIST OF FIGURES	  vi i i

 LIST OF TABLES	_    1x

 CHAPTER 1.   SUMMARY........	   ^

 CHAPTER 2.   REGULATORY  AUTHORITY	   2-l

 CHAPTER 3.   COMFORT COOLING  TOWERS	.'.	   3_j_

             3.1  GENERAL	               3 i
             3.2  DEFINITION  OF  SOURCE  CATEGORY	  "*	   31
             3.3  COMFORT  COOLING SYSTEM  COMPONENTS	'.'.'.'"   3-2
                  3.3.1  Cooling Tower	....!!!***   3-2
                  3.3.2  HVAC and Refrigeration Equipment	'.'.'.   3-6
             3.4  CHEMICAL TREATMENT PROGRAMS......................     3-9
                  3.4.1  Purpose	**['   3.9
                  3.4.2  Corrosion	.....!  3-10
                  3.4.3  Scaling and Fouling	  3_17
                  3.4.4  Microbiological  Control	  "  3_ig
             3.5  MAINTENANCE REQUIREMENTS	         3.to
             3.6  FORMATION OF DRIFT..	           3_20
             3.7  REFERENCES  FOR CHAPTER  3		]'.'.'.'.  3-22

CHAPTER  4.   MODEL COMFORT COOLING TOWER  SYSTEMS	   4-1

             4.1  MODEL TOWER PARAMETERS	         4.1
                  4.1.1  Building Size....		."'.'m   4.1
                  4.1.2  Cooling Requirements	„	*   4.2
                  4.1.3  Recirculation Rate	   4.2
                  4.1.4  Evaporation Rate..	   4.2
                  4.1.5  Slowdown Rate	ซ	  *   4.3
                  4.1.6  Hexavalent Chromium Emissions.....	'.'.   4.3
                  4.1.7  Stack Parameters	      4_4
                  4.1.8  Maintenance	     *   4.5
             4.2   MODEL HEAT EXCHANGER PARAMETERS	     "   4_6
             4.3   BASELINE EMISSIONS	         4_6
             4.4   REFERENCES FOR CHAPTER 4	'.]'.  4.23

CHAPTER 5.   EMISSION CONTROL TECHNIQUES..	   5.1

             5.1   CONTROL TECHNIQUES	    5_1
                  5.1.1  Nonchromiurn-Based Treatment Programs....!!*   5-1
                  5.1.2  High-Efficiency Drift Eliminators	   5-7
             5.2  REFERENCES FOR CHAPTER 5			\."  5.13

CHAPTER 6.  REGULATORY CONTROL  ALTERNATIVES		   6-1

-------
                       TABLE OF CONTENTS  (continued)
                                                                      Page
 CHAPTER 7.   COSTS ........................ ..... . ................ ....    7_i

             7.1  COSTS  OF  EMISSION  CONTROL  TECHNIQUES.." ..........     7-1
                 7.1.1   Nonchromi urn-Based Treatment  Programs .......    7-1
                 7.1.2   Higher Efficiency Drift  Eliminators ........    7-3
             7.2  FIVE-YEAR SALES AND  COSTS  PROJECTIONS .............    7.4
             7.3  REFERENCES FOR CHAPTER  7 ................. . ....... .  7.10

 CHAPTER 8.   ECONOMIC  IMPACTS .................... • ...................  - 8_i

             8.1  SUMMARY ........... . ................                  R i
             8.2  INDUSTRY  PROFILE ....................... '.'.'.'.'.'.'.'.'.'.'.   8-1
                 8.2.1  Size of Chromium-Based Corrosion Inhibitor"
                           Market .................... . . .............   o-l
                 8.2.2  Distribution of  Cooling Towers by
                           Building Size ....... ! ......... . ..........   8-2
                 8.2.3  Commercial Building Rental Rates... ........   8-2
             8.3  CONTROL COSTS ............ . ......... ........          8-3
             8.4  ECONOMIC  IMPACTS ...... .. .............. ... ........   1 4
             8.5  REFERENCES  FOR CHAPTER 8 ................. .'..'.'.'.'.'.'.  8-15

CHAPTERS.   OTHER IMPACTS., ............................... „ ........   9.!

            9.1  IMPACTS RESULTING FROM PROHIBITING CHROMIUM .......   9-1
            9.2  IMPACTS RESULTING FROM HIGHER EFFICIENCY DRIFT
                   ELIMINATOR RETROFITS .................... .          9.3
            9.3  REFERENCES FOR CHAPTER 9 .............. ........'.'.'.'.   9-3

APPENDIX A  EVOLUTION OF THE BACKGROUND INFORMATIONN DOCUMENT ......  A-l

APPENDIX B  PUBLIC CANCER RISKS FROM THE EMISSIONS OF HEXAVALENT
              CHROMIUM FROM COMFORT COOLING  TOWERS ..... ..;.;...' ...   B-l

            B.I  INTRODUCTION .............................            R  i
                 B.I.I  Overview ............................... *ฐ*|   B-!
                 B.I. 2  The Relationship  of  Exposure  to  Cancer
                        Risk ...... .............................       B-l
                 B.I. 3  Public  Exposure ..........................  *   B-3
                 B.I. 4  Public  Cancer  Risks.... .................      B-4
            B.2  THE UNIT RISK  ESTIMATE FOR  HEXAVALENT CHROMIUM...."   B-5
            B.3  QUANTITATIVE EXPRESSIONS OF PUBLIC EXPOSURE. ......   B-6
                 B.3.1  Model Selection and  Description .........      B-6
                 B.3. 2  Input Data and  Results ____ . ........           B-9
            B.4  ANALYTICAL UNCERTAINTIES APPLICABLE  TO THE   .....
                  CALCULATIONS OF PUBLIC HEALTH RISKS CONTAINED
                  IN  THIS  APPENDIX.. ......................          B_10
                 B.4.1  The Unit Risk  Estimate ...............    **  R-10
                 B.4. 2  Public  Exposure .....................    ""  R-17
            B.5  REFERENCES FOR APPENDIX  B .............. ..    .....  B-18
                                   vi

-------
                       TABLE OF CONTENTS  (continued)
APPENDIX  C   EMISSION  TEST  DATA	:	./..    c_i

             C.I   DESCRIPTION OF TESTS	      C-2
                  C.I.I   Department of  Energy,  Gaseous  Diffusion""
                           Plant, Paducah, Kentucky	    C-2
                  C.I.2   National Bureau of Standards,
                           Gaithersburg, Maryland	     C-5
                  C.I.3   Exxon Refinery, Ethylene  Production,
                           Baytown, Texas	         r_7
                  C.I.4   Exxon Refinery, Lube Oil  Production,
                           Baytown, Texas..		        r  10
             C.2   SUMMARY OF TEST DATA	.."...."..".*."  C-13

APPENDIX D   CHROMIUM  DISCHARGE REGULATIONS	   D-l

             D.I   CHROMIUM  DISCHARGE REGULATIONS	..                 n 1
             D.2   REFERENCES FOR APPENDIX D	!".."""."."i.*."!   oil

APPENDIX E   CALCULATION OF ANNUAL WEIGHTED AVERAGE COST OF
              CHEMICAL TREATMENT PROGRAMSv		.   E-l

APPENDIX F  SAMPLE CALCULATIONS OF FLOWRATES, TOWER PARAMETERS,
              AND HEXAVALENT CHROMIUM (Cr^)  EMISSION RATEs!.......   F-l

            F.I  COOLING TOWER CAPACITY...	             F 1
            F.2  RECIRCULATION RATE	      	   F 1
            F.3  EVAPORATION RATE	   F 1
            F.4  SLOWDOWN RATE.		           	   F?
            F.5  AIRFLOW RATE	   F"o
            F.6  STACK DIAMETER	        	   F ?
            F.7  RECALCULATED EXIT AIR VELOCITY	!."!!"	   i-2
            F.8  DISTRIBUTION OF CHROMIUM-USING CCT's...       	   F 2
            F.9  HEXAVALENT CHROMIUM  EMISSIONS  RATE	.'.	   F-3
                                   vii

-------
                               LIST OF FIGURES
                                                                      Page
 Figure 3-1   Internals of crossflow and counterflow cooling
                towers	   3_4
 Figure 3-2   Details of splashfill and filmfill	   3_5
 Figure 3-3   Operating principles of various drift eliminators	   3-7
 Figure 3-4   Cooling systems served by comfort cooling towers	   3-9
 Figure 3-5   Corrosion mechanism on carbon steel  surface...	  3-12
.Figure C-l   Tower C-637-2A at Department of Energy Gaseous
                Diffusion Plant	  c-15
 Figure C-2   Cooling tower at NBS facility in Gaithersburg,
                Maryland	  C-16
 Figure C-3   Tower 68 at Exxon-Baytown refinery	  C-17
 Figure C-4   Tower 84 at Exxon-Baytown refinery	  C-18
                                   viii

-------
                                LIST  OF  TABLES
 TABLE 3-1

 TABLE 3-2

 TABLE 3-3

 TABLE 3-4

 TABLE 4-1


 TABLE 4-2

 TABLE 4-3

 TABLE 4-4

 TABLE 4-5

 TABLE 4-6a


 TABLE 4-6b


 TABLE 4-7a



 TABLE 4-7b



 TABLE  4-8


 TABLE  7-1

 TABLE  7-2


TABLE 7-3


TABLE 7-4
 CORROSION RATE OF METALS....	   3_13

 ANODIC AND CATHODIC CORROSION  INHIBITORS	   3_15

 CHROMATE-BASED CORROSION INHIBITORS	   3_16

 EFFECT OF EVAPORATION ON DROPLET SIZE	   ^-22

 MODEL COMFORT COOLING TOWERS AND HOURLY BASELINE
   EMISSIONS	;   4_g

 EMISSION FACTORS FROM EPA-SPONSORED TESTS	  4_1Q

 STACK PARAMETERS FOR MODEL CCT1 s.	  4_n

 NUMBER OF MODEL CCT's	  4_12

 DISTRIBUTION OF CHROMIUM-USING CCT's  PER STATE.	  4-13

 LOWER- AND UPPER-BOUND ESTIMATES OF Cr+6 EMISSIONS
   FROM INDIVIDUAL MODEL TOWERS PER STATE (METRIC)	  4-14

 LOWER- AND UPPER-BOUND ESTIMATES OF Cr+6 EMISSIONS
   FROM INDIVIDUAL MODEL TOWERS PER STATE (ENGLISH)....  4-16

 LOWER- AND UPPER-BOUND ESTIMATES OF TOTAL Cr+6
   EMISSION PER  STATE AND ANNUAL Cr+6  EMISSIONS
   PER PERSON  (METRIC)	   4_18

 LOWER- AND UPPER-BOUND  ESTIMATES OF TOTAL Cr+6
   EMISSIONS PER STATE AND ANNUAL Cr+6 EMISSIONS
   PER PERSON  (ENGLISH)	   4_20

 ANNUAL INCIDENCE  OF CANCER AS MODELED BY EPA's HUMAN
   EXPOSURE MODEL	   4-22

 COSTS  AND  COST EFFECTIVENESS FOR NO-CHROMIUM OPTION...   7-5

 COSTS  AND  COST EFFECTIVENESS FOR HIGH-EFFICIENCY
  DRIFT ELIMINATOR RETROFITS ON CHROMIUM-USING CCT's..   7-7

 5-YEAR SALES PROJECTIONS OF CCT's FOR  NEW
  INSTALLATIONS	   7_g

ANNUALIZED COSTS OF NONCHROMATE AND HIGH-EFFICIENCY
  DRIFT ELIMINATOR CONTROL TECHNIQUES		   7.9
                                    1x

-------
                         LIST OF TABLES  (continued)
 TABLE 8-1    ESTIMATED 1986 MARKET SHARE OF MAJOR SUPPLIERS
                OF CORROSION INHIBITORS FOR CCT'S	  8-6

 TABLE 8-2    NUMBER OF BUILDINGS BY TYPE AND SIZE	  8-7

 TABLE 8-3    NUMBER OF BUILDINGS WITH CENTRAL COOLING SYSTEMS
                BY TYPE AND SIZE	  8-8

 TABLE 8-4    NUMBER OF BUILDINGS WITH CCT'S BY TYPE AND SIZE	  8-9

 TABLE 8-5    NUMBER OF CCT'S USING CHROMATES BY BUILDING
                TYPE AND SIZE	  8-10

 TABLE 8-6    INCREMENTAL COST PER UNIT AREA FOR USING NONCHROMATE
                CORROSION INHIBITORS	  8-11

 TABLE 8-7    INCREMENTAL COST PER UNIT AREA FOR RETROFITTING
                DRIFT ELIMINATORS	  8-12

 TABLE 8-8    TOTAL INCREMENTAL COST OF USING NONCHROMATE
                CORROSION INHIBITORS BY BUILDING TYPE  AND SIZE	   8-13

 TABLE 8-9    TOTAL ANNUALIZED  COST OF  RETROFITTING  DRIFT
                ELIMINATORS BY  BUILDING TYPE  AND SIZE	   8-14

 TABLE A-l    EVOLUTION  OF THE  BACKGROUND  INFORMATION  DOCUMENT	   A-2

 TABLE B-l    HEXAVALENT CHROMIUM  EMISSION RATE  PER  PERSON AND
                ANNUAL INCIDENCE BY  STATE	   B-10

 TABLE B-2    LOCATIONS  USED  IN ESTIMATING INDIVIDUAL RISK	   B-ll

 TABLE  B-3    MODEL  PLANT PARAMETERS	„	  B-14

 TABLE  B-4   MAXIMUM CONCENTRATIONS PREDICTED	  B-15

 TABLE  C-l    SUMMARY OF OPERATING PARAMETERS AND METEOROLOGICAL
               DATA DURING TESTS AT DEPARTMENT OF ENERGY, GASEOUS
               DIFFUSION PLANT, PADUCAH, KENTUCKY	  C-19

TABLE C-2    SUMMARY OF OPERATING PARAMETERS AND METEOROLOGICAL
               DATA DURING TESTS AT NATIONAL BUREAU  OF STANDARDS,
               GAITHERSBURG, MARYLAND	  C-21

TABLE C-3    SUMMARY OF OPERATING PARAMETERS AND METEOROLOGICAL
               DATA DURING TESTS OF TOWER 68 AT EXXON REFINERY,
               BAYTOWN, TEXAS	            C-22

-------
TABLE C-4
                        LIST OF TABLES (continued)
SUMMARY OF OPERATING PARAMETERS AND METEOROLOGICAL
  DATA DURING TESTS OF TOWER 84 AT EXXON REFINERY
TABLE C-5
TABLE C-6
TABLE C-7
TABLE C-8
TABLE C-9
SUMMARY OF EMISSION TEST RESULTS-DEPARTMENT OF
ENERGY GASEOUS DIFFUSION PLANT, PADUCAH, KENTUCKY...
SUMMARY OF EMISSION TEST RESULTS-NATIONAL BUREAU OF
STANDARDS, GAITHERSBURG, MARYLAND 	
SUMMARY OF EMISSION TEST RESULTS FOR TOWER 68 AT
EXXON REFINERY, BAYTOWN, TEXAS 	
SUMMARY OF EMISSION TEST RESULTS FOR TOWER 84 AT
EXXON REFINERY, BAYTOWN, TEXAS 	
SUMMARY OF SENSITIVE PAPER DRIFT MEASUREMENTS
C-24
C-26
C-28
r ?Q
u— tJ
r_^n
C-31

-------

-------
                                1.  SUMMARY

     Regulations for chromium  (Cr)  emissions  from  comfort  cooling  towers
 (CCT's)  are being developed  because hexavalent chromium  (Cr+s),  the  form
 of Cr used in CCT's, is considered  to be a potent  carcinogen.  Nationwide
 annual Cr emissions from CCT's are  estimated  to  be about 7.2 to
 206 Megagrams (16,000 to 453,000 pounds) (based  on an average  nationw-Me—
 utilization rate of 46 percent).  These emissions  result-in a  health risk
 estimate of 4 to 112 cancer  cases per year.   This  range reflects the
 lower- and upper-bound emission estimates and the  upper-bound  unit risk
 factor.
     Two regulatory alternatives were evaluated  for CCT's:  nonchromium
 option and a high-efficiency drift  eliminator option (HEDE).   The least
 burdensome regulatory alternative is a standard  that prohibits the use of
 chromium-based chemicals in CCT's and the sales  of these chemicals for use
 in CCT's.  Such a standard can reduce the risk by  100 percent, and this
 benefit can be achieved without operating problems or unreasonable
 costs.
     A standard requiring retrofit  of existing CCT's with HEDE1s would be
 less effective (85 percent reduction in risk, or less).   In addition,
 implementation of an HEDE standard would be impossible at some existing
 sites because of tower configuration.   Where retrofit would be possible or
 in new CCT's, an HEDE standard would be much more complex than switching
 to nonchromium treatment.   An annual inspection of the drift eliminator
for proper sealing with the tower structure and an initial  certification
program to verify that selected drift  eliminators can achieve  the required
emission rate would be required.   This inspection and certification
process would be more costly for  operators  than switching to nonchromium
treatment programs.   Furthermore,  the  certification process may take
several  years to complete.

                                    1-1

-------
      Comfort cooling towers are used extensively as components of heating,
 ventilation, and air conditioning (HVAC) or refrigeration systems.  The
 HVAC systems are installed in hospitals; hotels; shopping malls; and
 office,  educational, and  other commercial buildings.  Refrigeration
 systems  are used for ice  skating rinks, cold storage (food)  warehouses,
 and other commercial operations.
      Hexavalent chromium  is used in CCT water to inhibit corrosion in the
 heat exchanger of the HVAC or refrigeration equipment and in the
 connecting piping.   Hexavalent chromium has long provided the best
 corrosion protection at a reasonable cost.   However, environmental
 concerns (primarily wastewater discharges)  have prompted many operators to
 switch to nonchromate water treatments.  Currently,  an estimated 75 to_
 90  percent of the 250,000 CCT's nationwide  use nonchrorriiunrtreatment
 programs or no water treatment.   Although the'remaining CCT's do not
 operate  under special  conditions that require the  use of Cr+s,  many
 operators are reluctant to change from a treatment procedure with which
 they are familiar to one  that will  require  more attention to achieve the
 same results.
      A regulation eliminating Cr+6  emissions  from CCT's  is most
 effectively accomplished  under Section 6 of the Toxic  Substances  Control
 Act  (TSCA).   Under  TSCA,  both owner/operator  use of  Cr+6  in  CCT's  and
 vendor sales  of  Cr+s  for  use  in  CCT's  could be  prohibited.   Under  the
 Clean Air Act  (CAA),  the  thrust  of  the  rule would be on  compliance by the
 owners and  operators  of the CCT's.  The  population of distributors (20+
 major and approximately 400 total)  is much  smaller than  the  population of
 CCT's (approximately 250,000)  that wouldhave to be  inspected to ensure
 implementation of the standard.  Thus, TSCA is  the best  authority for
 regulating  the use of Cr+s chemicals  in  CCT's.   By prohibiting the sales
 of the chemicals  and also making the chemical sales companies liable, it
 is expected that  the burden of ensuring compliance with the regulation
would be reduced  for both  EPA and owners/operators of CCT's.
                                    1-2

-------
                          2.  REGULATORY AUTHORITY

      Regulation of chromium emissions from comfort cooling towers (CCT's)
 can be accomplished under either the Toxic Substances Control Act (TSCA)
 or the Clean Air Act (CAA).  Under Section 112 of the CAA, it would not be
 possible to  prohibit the sales  of a chemical.   However,  it would be
 possible to  prohibit indirectly the use of chromium chemicals in CCT's by
 setting a zero  emission  standard.   Under the CAA,  the recordkeeping
 requirements extend to "any person who  owns or operates  any emission
 source or who is subject to any requirements of this  chapter."   This would
 allow EPA to require the CCT owners and operators  to  maintain records,
 but,  because the source  category is so  large,  enforcement  would  be
 difficult.   The  thrust of regulations under the CAA would  be  on  compliance
 with  the  emission  limitations by the CCT owners and operators.
      The  TSCA provides EPA  with  broad authority to assess  and regulate
 chemical  substances  in the  environment, in  the  workplace,  and in
 commercial products.  Under Section 6(a) of TSCA, EPA is authorized  to
 impose  regulatory controls  if the Agency finds  that there  is a reasonable
 basis  to  conclude that the  manufacture, processing, distribution in
 commerce, use, or disposal  of a chemical substance presents or will
 present an unreasonable risk of  injury to human health or the
 environment.   To determine whether a risk is unreasonable, EPA balances
 the probability that harm will occur from the chemical substance under
consideration against the social and economic costs to society of placing
restrictions  on the substance.
     If the EPA Administrator determines that  an unreasonable risk exists,
one or more of several regulatory measures may  be applied to the  xtent
necessary to  protect adequately  against  the risk.  Those  measures include
prohibiting or limiting  the manufacture, processing,  or distribution in
                                   2-1

-------
commerce; labeling; prohibiting or otherwise regulating any manner or
method of commercial use or disposal; requiring the revision of quality
control procedures; and requiring that chemical manufacturers notify the
public of unreasonable risks associated with a chemical substance.  The
EPA Administrator is required by TSCA to apoly the least burdensome
requirement(s) to protect adequately against the risks.
                                   2-2

-------
                          3.   COMFORT COOLING TOWERS

 3.1  GENERAL
      This section provides a description of  the source category and the
 major users-of-comfort cooling towers (CCT's); the cooling tower system;
 heating, ventilating, and air conditioning  (HVAC) and refrigeration
 systems served by the cooling towers; cooling water chemical treatment
 programs; the mechanism by which hexavalent chromium (Cr+6) is emitted,
 and the emission control techniques.
 3.2  DEFINITION  OF SOURCE CATEGORY
      There are two broad categories of cooling towers,  comfort and
 industrial.   Comfort cooling towers are  used to maintain a specified
 environment  or refrigeration system.  Industrial  process coo Ting towers
 are used to  control  the temperatures of  process fluids  in industrial
 production units.   The EPA is evaluating industrial  towers separately
 because  the  cooling  system conditions under which  industrial  towers
 operate  may  make controlling  corrosion in an industrial  tower more complex
 and difficult.  The  conditions that distinguish these towers and that make
 controlling  Cr+6 emissions  from CCT's less  difficult than controlling Cr+6
 emissions  from industrial  towers  include a  lower potencial  for corrosion
 because  lower water  temperatures  typically  are encountered; a lower
 utilization  rate; the  extensive use  of heat exchangers constructed of
 copper, which is less  susceptible to  corrosive attack; little opportunity
 for upsets in the chemical treatment  programs from contamination of the
 cooling water; and a Tower potential  for significant economic loss should
 alternative chemicals not perform satisfactorily.   Usually only one heat
 exchanger is integrated into a single cooling tower system in CCT systems
whereas a single  industrial tower may support many heat  exchangers.
                                    3-1

-------
      Comfort cooling towers are used in all States  in the U.S.,  primarily
 in urban areas.  Major users of CCT's with HVAC systems  include  hospitals,
 hotels, educational facilities, office buildings, and shopping malls.
 Refrigeration systems that may operate with CCT's include ice skating
 rinks, cold storage (food) warehouses, and other commercial operations.
 Estimates from the two largest manufacturers of cooling towers (each
 represents a 40 percent market share) indicate that the nationwide
 population of CCT's is between 200,000 and 300,000.J'2 -An estimate
 slightly lower than 200,000 was provided by a smaller, regional
 manufacturer.   The 300,000 estimate was provided by the manufacturer that
 appears to have the higher percentage of the CCT market.  Water treatment
 vendors estimated that 10 to 25 percent of CCT's use chromium-based water
                     n  6
 treatment chemicals.  ~   For analysis purposes, it is assumed that the
 nationwide population  of CCT's  is  250,000  units and that 15  percent of
 CCT's  (about 37,500)  use chromium-based water  treatment  chemicals.
 Hexavalent chromium use in CCT's appears to be distributed randomly across
 the  country.
 3.3  COMFORT COOLING SYSTEM COMPONENTS
 3.3.1   Cooling  Tower
     Cooling  towers are  devices that  cool warm water by  contacting  it with
 ambient air  that  is drawn  or forced through the tower.   For most  cooling
 towers,  about 80  percent of the cooling occurs  from  evaporation of  water
 as the  air flowing through  the tower contacts water  cascading from  the top
 to the  bottom of  the tower.   Most tower systems are designed with
 recirculating water systems to conserve water resources or reduce costs of
 purchasing water.  Typically, the CCT is an open recirculating system, but
 some are closed recirculating systems.  The major cooling tower components
 include the fan(s), fill material,  water distribution deck or header,
 drift eliminator, structural frame, and cold water basin.  Other
 components that affect tower operation include the pumps  and  pipes
 necessary to circulate the cooling  water through the cooling  tower'and
 heat exchanger loops.
     Cooling towers are designed with mechanically induced-,  mechanically
forced-, or natural-draft airflow.   Induced draft is provided  by  a
propeller-type fan located in the stack  at  the  top  of the tower.
                                   3-2

-------
 Forced-draft towers are usually smaller than induced-draft towers and have
 either centrifugal fans located at the base of the tower, which  is
 constructed as a plenum to provide positive-pressure airflow through the
 fill material, or axial fans located on the side of the tower.  Natural-
 draft airflow, which is not used in CCT's, relies on buoyancy created by
 temperature differences between the air in the tower and the atmosphere.
 When the cooling demands are minimal and the air temperature is low
 enough, water can be circulated through the tower and cooled sufficiently
 without using the fans.  In these instances, a natural  draft is created in
 a mechanical-draft tower.   In addition, the direction of the airflow
 through a mechanical-draft tower is either crossflow or counterflow.
 Crossflow refers to horizontal  airflow through  the fill, and counterflow
 refers to upward vertical  airflow.   Schematics  of counterflow and
 crossflow cooling towers are presented in  Figure  3-1.   Comfort  cooling
 towers typically are  designed with  crossflow air  configuration.
      Fill  material  is used to maintain an  even  distribution  of  water
 across the horizontal plane of  the  tower and  to create  as much  water
 surface as practical  to enhance evaporation and sensible heat exchange.  A
 large water-to-air  interface  is provided by either a  large number of water
 droplets  or many  thin vertical  sheets  (or  tubes)  of water.  The two  types
 of fill,  splash  and film,  are shown  in  Figure 3-2.  Splash fill is
 constructed as successive  layers of  staggered impact surfaces.  Small
 droplets  are formed as  warm water falls through the fill and splashes off
 each  layer.  Film fill  is  constructed of sheets of material in a
 "honeycomb" configuration.  The fill is oriented  such that water enters
 the open end of the honeycomb and flows vertically in sheets along the
 surface of  the fill material.  Typically, fill materials are wood,
 polyvinyl  chloride (PVC), polystyrene, polypropylene, asbestos cement
 board, or  cellulose.
     The distribution of the warm water over the fill  material is
 accomplished in one of two ways depending on the type of tower.   In
crossflow towers, there is  a water distribution  deck  above  the fill
material at the top of the  tower.  The floor of  this  deck contains gravity
flow nozzles,  and the water level in the deck controls  the  rate  of water
flow onto the fill.   In  counterflow  towers, the  water  distribution system
                                   3-3

-------
                             AIR
                           OUTLET
             WATER
             INLET
FAN

 WATER
 INLET
                             DRIFT
                         ELIMINATORS
             WATER OUTLET
                   MECHANICAL  DRAFT
                   CROSS-FLOW  TOWER
                           AIR
                         OUTLET
          WATER  OUTLET

                 MECHANICAL  DRAFT

                 COUNTER-FLOW  TOWER
Figure 3-1.  Internals of  crossflow and counterflow cooling towers,
       (Reprinted from Drift Technology for Cooling Towers,
                    The Marley Company,  1973)
                              3-4

-------
                   SPLASH FILL
                           c.~ป I
                     FILM FILL
Figure  3-2.  Details of  splash  fill  and film fill.
(Reprinted  from  Custodis-Cottrell product  brochure)
                         3-5

-------
 1s constructed  of a series  of header pipes connected to pressure flow
 nozzles  placed  above the  fill  material.   In both systems, the nozzles are
 arranged such that even water distribution over the fill  material  is
 provided.
     Water  droplets and the dissolved solids they contain that are
 entrained in the  air and  emitted  from the CCT stack are referred to as
 drift.   (Drift  formation  is discussed in  Section 3.6.)   Drift eliminators
 can be installed  at the exit of the  fill  sections to reduce  the drift in
 the exiting airflow.  The drift removal efficiency of a drift eliminator
 is a function of  its  design.   Figure 3-3  presents schematics  of the three
 major drift eliminator designs:   blade-type,  waveform,  and cellular.   A
 fourth drift eliminator design, herringbone,  is  similar to the bl'ade-type
 except that the blades in one  row are offset  from the blades  in the next
 row.  Typically,  herringbone and  blade-type units are the least efficient,
 waveform units are moderately  efficient,  and  cellular units are the most
 efficient.  However,  some cellular designs may be less  efficient than  some
 waveform designs, and some  waveform  designs may  be  less efficient than
 some herringbone  designs.   Drift  eliminators  are  constructed of wood,  PVC,
 metal, asbestos-cement, polystyrene,  or cellulose.  Currently,  the
 material most often specified  is  PVC.  Drift  eliminators installed  in
 towers built in the last  several years are more  likely to be higher
 efficiency waveform or cellular units, but a  large number of older towers
 still have  lower  efficiency herringbone and waveform eliminators.
     The structural frame of cooling towers can be wood, concrete,
 masonry, steel, and combinations of these materials.  The cold water
 basins (reservoirs) typically are located directly below the fill material
 at the base of the cooling tower.   Basin size is affected by the size of
 the tower and by the necessity to accommodate any expected short-term
 fluctuations in the water volume of the system.
 3.3.2  HVAC and Refrigeration Equipment
     The components of an HVAC or a refrigeration system include the
cooling distribution system, a heat rejection system, and the
refrigeration machine, commonly referred to as a chiller system. The
cooling distribution system consists  of the air handling units, and  the
heat rejection system consists of  the cooling tower.  The  chiller system
                                   3-6

-------
                                   BLADE-TYPE
                                   ELIMINATOR
                                    WAVEFORM
                                    ELIMINATOR
                                                   Wood Lath
                                                   Bladss
                                                  Extruded
                                                  favo Form
                                    CELLULAR
                                    ELIMINATOR
Figure 3-3.  Operating principles of various drift  eliminators
                (Reprinted  from Reference No.  7)
                              3-7

-------
can be 'either a compression-cycle or  an  absorption  type  (see
Figure 3-4).  In both systems, cooling water  is  required  to condense  the
refrigerant vapor  (i.e., Freonฎ  in a  compression-cycle system  and  water in
most absorption-type systems).   (Note:   Water-ammonia compression  systems
also have been used with ammonia as the  refrigerant.  However,  in  HVAC
systems, water/lithium bromide compression  systems  predominate.)   In  the
absorption system, cooling water also is required for the absorber.   Water
is supplied from a cooling tower at about 29ฐC (85ฐF), and it  returns to
the tower at about 35ฐC (95ฐF).  The  condenser typically  is constructed of
copper tubes and a cast steel shell.  The condensed refrigerant passes to
the evaporator where warm water from  the air  handling units transfers heat
that evaporates the refrigerant, thereby chilling the water.   The  chilled
                                               A Q
water is pumped back to the air handling units.  ป
     The required tower size for a given air  conditioning load depends on
the type of chiller system used.  Compression-type equipment is designed
for a temperature difference of 5.6ฐC (10ฐF)  across the tower  and  a
recirculating rate of 11.6 liters per minute  (a/min) (3 gallons per minute
[gal/min]) per ton of air conditioning.  Absorption-type equipment is
designed for a temperature difference of 8.3ฐC (15ฐF) across the tower and
a recirculating rate of 15.1 z/min (4 gal/min) per ton of air condi-
tioning.  A ton of air conditioning is defined as 3,514 Watts  (W)
(12,000 Btu/hour [h]).  Compression-type equipment will reject 4,392 W
(15,000 Btu/h) to the cooling tower because for each ton of air
conditioning, an additional 878 W (3,000 Btu/h) is required to perform the
work needed to compress the refrigerant.  When absorption-type equipment
is used, 8,785 W (30,000 Btu/h) are rejected at the cooling tower for each
ton of air conditioning because 5,270 W  (18,000 Btu/h)  are required to
drive the process.  Water is evaporated  at the rate of 7.2 a (1.9 gal) and
14.0 i (3.7 gal) per hour per ton of  cooling in the cooling towers used
with compression and absorption systems, respectively.
3.4  CHEMICAL TREATMENT PROGRAMS
3.4.1  Purpose
     Chemicals are added to the recirculating cooling water to inhibit the
corrosive effects of the water, to control  the rate of  scaling and
fouling, and to control the growth of micro-orga"hisms in  both  the cooling
tower and the heat exchangers.  As evaporation occurs during  cooling,  the

                                    3-8

-------
0.3E
>- uu
H- t—
 I ซ/ป

o 
ป—I
I—O
0. Z
Qฃ 1-4
O —J
(
                                                  :"S
                     LU Z
                     (X. >—>
                     Q_ _J
                     S O
                     00
                     o c_>
                                               •a-
                                                 i
                                                (U

                                                3
                                                cn
3-9

-------
chemical constituents of the water become concentrated.  A percentage of
the recirculating water is  intentionally discharged  (blowdown) to maintain
acceptable concentrations of susnended and dissolved  solids.  Also, as
water cascades through the  tower, some is entrained and emitted from the
stack as drift.  Fresh water is added to make up for  the losses resulting
from evaporation, blowdown, and drift.
     Typical water treatment program chemicals include (1) a corrosion
inhibitor, (2) an antiscalant, (3) an antifoulant, (4) a dispersant, (5) a
surfactant, (6) a biocide,  and (7) an acid and/or caustic soda for pH
control.  Chromium-based chemicals are corrosion inhibitors.  Other
chemicals fall into one or  more categories, and many  combinations of
various chemicals are used.  The quality of the cooling tower water supply
directly affects the type and quantity of chemicals required to maintain
satisfactory protection.  The three problems—corrosion, scaling and
fouling, and microbiological growth—and the chemicals used to control
them are discussed later in this section.
     Major water chemistry  parameters that affect the selection of
chemical treatment programs include pH, calcium hardness (calcium ion
concentration), alkalinity  (bicarbonate, carbonate, and hydroxide ions),
chloride, sulfate, silica, dissolved solids (conductivity), and suspended
solids.  Water quality also directly affects the number of cycles of
concentration that can be maintained.  The number of cycles of concen-
tration is defined as the ratio of conductivity or calcium hardness of  the
recirculating water to that of the makeup water.   The maximum level  of
either parameter established is based on the chemical treatment program
and the acceptable rates of corrosion and scaling.
3.4.2  Corrosion.
     3.4.2.1  Description.  Corrosion is the oxidation of a metal  by some
oxidizing agent in the environment.   The area over which metal  is  oxidized
(corroded) is called the anode; the area over which the oxidizing  agent is
reduced is called the cathode.   Many metals contain both anodic and
cathodic areas.  As corrosion proceeds,  electrons flow through  the metal
from the anode to the cathode.   In water, an electrochemical  cell  is
formed as cations migrate toward the cathode and  anions move  toward  the
anode.  The water is the conducting  fluid or electrolyte.   The  flow  of
                                   3-10

-------
 electrons through the metal (external circuit)  is the corrosion current,
 which is limited by the rate at which electrons are accepted by the
 oxidizing agent at the cathode.  Figure 3-5 illustrates the mechanisms for
 corrosion in a single metal.  Galvanic corrosion occurs when two different
 metals are in physical contact.  Depending on their relative character-
 istics, one of the metals will function as an anode and the other as a
 cathode, and the anodic metal  will corrode.  Conditions that increase the
 conductivity of the water (e.g., high dissolved solids content or high
 temperature) increase the rate of corrosion.
      Several types of corrosion occur in cooling water systems including
 general  etching, pitting, tuberculation, and crevice corrosion.  The least
 harmful  is the general etching that occurs uniformly over the surface of
 the metal.   Pitting is the formation of small  holes  from corrosive attack
 at the metal surfaces, and tuberculation refers to the corrosion  products
 cap that forms over a pit.  Crevice corrosion  occurs where metallic and
 nonmetallic materials contact.   Corrosion  can  be retarded, but  not totally
 prevented,  and the rate of corrosion that  is acceptable  varies  among
 systems.   Inhibitors  are added  primarily to protect  the  heat  transfer
 surfaces, which are the most critical metal  components  in  the system, from
 corrosion.   The terms and  corrosion rates  presented  in Table  3-1 are
 generally used to  describe the  severity  of carbon  steel  and copper
 corrosion.
     Chemicals used in a recirculating water system are  either  anodic or
 cathodic corrosion  inhibitors,  and  the mechanisms by which they protect
 the metal are  passivation,  precipitation,  and/or adsorption.  Anodic
 corrosion inhibitors  function by providing a barrier film or deposit at
 anodic areas that prevents  the electrolytic half-cell reaction from
 occurring.  Cathodic  inhibitors prevent the other half-cell reaction from
 occurring by providing a barrier film or deposit at cathodic areas.
 Table 3-2 lists various anodic and cathodic inhibitors.  The barrier
 created by passivation is an oxide that forms on the  metal  surface, and
 the barrier created by precipitation is an insoluble  precipitate that
 coats the metal surface.  Generally, anodic inhibitors are passivators,
 and cathodic inhibitors are precipitators.   Molecules that  have  polar
properties provide a barrier by adsorbing on the entire metal  surface.
Adsorption inhibitors are usually organic compounds.10'12

                                   3-11

-------
                        WATER (ELECTROLY FEJ
ANODIC REACTIONS
   Fe * fe
   Fe^+ZOH" * Fe(OH)2 (soluble form)
   Fe(OH)2+H20-4o2 * 2 Fe(OH)3 (insoluble red-brown deposit)
CATHODIC REACTIONS
   2e"+2H'r * H.
               gas
              - H20+OH'
        Figure 3-5.  Corrosion mechanism on carbon steel  surface,
                    (Reprinted from Reference No.  10)
                                  3-12

-------
            TABLE 3-1.  CORROSION RATES OF METALS
                                                  12
Description
       Corrosion rates, um/yr (mil/vr)	
                                 Copper alloy
Carbon steel
Negligible
Mild
Moderate
Severe
 <25.4-50.8
  50.8-127.0 (2-5)
127.0-254.0 (5-10)
      >254.0
                           <2.54
                   3.81-5.08 (0.15-0.2)
                   5.08-8.89 (0.2-0.35)
                      12.7-25.4 (0.5-1)
                            3-13

-------
      3.4.2.2   Chromium-Based  Inhibitors.   Chromates historically have
provided  the  best protection  against corrosion for the money and effort.
Very  little monitoring  and  control  of concentration are required.   Even if
the inhibitor feed is temporarily interrupted, the existing film will
continue  to provide protection  for  several  days.
      Chromate concentration typically must  be  maintained above  200  parts
per million (ppm)  if the  chromates  are used alone  in the recirculating
water.  However,  chromates  (which are anodic inhibitors)  typically  are
used  in combination with  cathodic inhibitors.   In  these combinations, less
chromate  provides  the same  corrosion protection that is provided by high
concentrations of  chromat.e  alone.   Vendors  provide many chromate-based
combinations  for  use in CCT's;  and  definitions of  low-,  very  low-, and
ultra-low-chromate concentrations vary.   In all cases,  the  chromate
concentration is  less than  30 ppm.   For the purposes of this  document,  all
of these  treatment program  classifications  will be considered low-
chromate.
     The  chemicals most commonly  added to chromate-based  formulations are
zinc and  phosphate;  but organic compounds,  polysilicates, and molybdates
also have been used.  (The  same chemicals are  used  in treatment programs
without chromates  and are discussed  in greater detail in  Section 5.1.1.)
Table 3-3 provides  concentrations and typical  operating conditions of some
of the formulations  discussed in  various publications-, which generally
focus on  industrial  tower systems.10"12  However, at those facilities that
use chromates  in the CCT's, the same treatment program  is used in both the
CCT's and the  industrial process cooling systems.  Chromate is an
excellent copper corrosion  inhibitor, but many publications indicate that
it is common to add  the organic triazole inhibitors to low-chromate
formulations that  are used  in systems with copper heat transfer  surfaces;
this was confirmed by some of the EPA studies of chemical treatment
programs used  at CCT's (case studies).13'17  However, this does  not  mean
that triazoles are more effective or even as effective as chromate at
protecting copper from corrosion.  The addition of  triazoles also protects
copper su- "aces from reactions with phosphonate dispersants.
     The cathodic half-cell  reaction controls the rate of corrosion  at the
anode.  Thus,  for a given cathodic reaction rate,  the same mass  of the
anode will corrode.  If the cathodic reaction is not controlled  by

                                   3-14

-------
     TABLE  3-2.  ANODIC AND  CATHODIC  CORROSION  INHIBITORS
Anodic
                                              Cathodic
Chromate
Molybdate
Orthophosphate
Nitrites
Orthosilicate
Polyphosphate
Zinc
Polysilicate
                            3-15

-------
          TABLE 3-3.  CHROMATE-BASED  CORROSION  INHIBITORS
                                                          10_12  18
 Combination3
Concentration, ppm   Operating conditions
Chromate/zinc
Chromate/orthophosphate
5-20/2.5-10
20-25/3-3.5
5-10/10
pH 6.5-7.0
pH 7.0-7.5
pH 6.2-6.8
Chromate/phosphate/zincc

Chromate/polyphosphate/zinc
Chromate/zinc/phosphpnate


Chromate/phosphonate

Chromate/phosphonate/d i spersant

Chromate/di spersant

Chromate/polysi1icate


Chromate/zi nc/di spersant

Chromate/molybdate
15-25/2-5/2-5

10-30/3-5/3-5

20-25/5-10/2.5-3.0

5/10/unk

15-25/2-4/3-5
2-3/2-3/5-10

5-10/3-5

5-15/2-6/2-6

3-5/30

5-10/5-10


10-20/1-2/1.5-10

10-30/1-5 '
 pH  6.0-7.0

 pH  6.5-7.0
  CaH  100-600 ppmd
 pH  6.0-6.5
  CaH  <400-600 ppm
 pH  6.5-7.2

 pH  6.5-7.5
 pH  6.5-7.0

 Not specified

 pH  7.5-8.5

 pH  7.5-8.5

 pH >7.5
 Si <10 ppm

pH 7.0-9.0

pH >7.5
aln all combinations except chromate/polysilicate, the organic triazole
 corrosion inhibitors should be included at 1 to 10 ppm when the system
.contains copper.
DThe components of some combinations can be formulated differently for
 different applications.
JjPolyphosphate and/or orthophosphate.
aCaH is calcium hardness.
                                   3-16

-------
 cathodic  inhibitors or triazoles  and  if  only  small  spots  of  the  anodic
 surface remain unprotected, pitting may  occur.  Although  chromate  is  an
 excellent inhibitor, the  likelihood that small  areas will  remain
 unprotected is greater with lower chromate treatments than it  is with
 higher chromate treatments.
 3.4.3  Scaling and Fouling
      3-4.3.1  Description.  Scale formation occurs when dissolved solids
 and gases-in cooling water reach their limit of solubility and precipitate
 out onto piping and heat transfer surfaces.  Fouling occurs when deposits
 of dirt, leaves, and/or floes of insoluble salts or hydrous oxides from
 corrosion agglomerate in the heat exchanger tubes.  Scaling reduces the
 heat transfer capacity of heat exchangers, and fouling hinders the flow of
 water through heat exchangers.  These conditions also contribute to
 pitting-type corrosion by creating corrosion cells and preventing the
 corrosion inhibitor from contacting the surface of the metal.  Calcium
 carbonate is the most common scale found  in cooling water  systems,  but
.calcium sulfate and calcium phosphate  also can be  formed in many
 systems.   All  three scales become .less soluble,  and,  therefore, more
 likely to precipitate at  higher temperatures.   Calcium  sulfate  is more
 likely to precipitate  at  lower pH  and  the other  two  scales are  more likely
 to  precipitate at  higher  pH.   Control  of  scaling and  fouling  depends on
 the control  of deposition  onto surfaces.   The  deposition can  be affected
 by  changing  the solubility of  scale-related salts, reducing the
 crystalline  growth  capacity of scale-related salts, and dispersing
 constituents that form fouli.-^-related  flocculations.
      3.4.3.2   Antisealants  and  AntifoulantS
      Chemical  compounds that are commonly  used and are the most effective
 in  controlling the  rate of  scaling include polyphosphates and
 phosphonates.  These compounds  reduce the  crystalline growth capacity of
 calcium salts.  Certain phosphonate compounds affect the solubility of the
 calcium salts  and reduce the formation of  scale.  Polymeric dispersants
 (with a molecular weight less than 20,000) reduce the potential  for
 fouling.10-12
     The phosphonates typically are added  to chromate-zinc  formulations as
 an alternative to phosphate, and scaling is better  controlled  than it  is
                                   3-17

-------
with  phosphate.   Also,  the amount of chromate can be reduced since the
phosphonate  system can  be operated at a slightly higher pH.
      The  powerful  oxidizing potential of chlorine can promote corrosion of
copper  if phosphonates  also are present in the system.   The addition of
benzotriazole  (or other azoles) and dispersants can minimize this
effect.   Phosphonates are also subject to biological  oxidation,  which
results in the release  of orthophosphate ions that can  cause fouling as
well  as reduction  of the recommended concentration of phosphonate.
However,  this  fouling problem  is much less severe than  that produced in
systems treated with polyphosphates.12
3.4.4 Microbiological  Control
      Three types of microorganisms are found  in cooling tower water
systems:   bacteria, fungi,  and  algae.   Bacteria are dispersed in the
water, fungi invade wood components,  and algae  attach to surfaces  in the
tower.  Slime  produced  by bacteria can coat and aggregate debris on  heat
exchanger surfaces, thereby reducing  the efficiency of  heat transfer.
Biological deposits on  metal surfaces  also can  accelerate pitting
corrosion.  Fungi  can cause decay  of wood  either by surface attack (soft
rot)  or internal attack of  the  cellulose (white rot).   Algal  growth  can
coat  the  fill material  and  reduce  the  effectiveness of  the  water droplet
formation  and, thus, the effectiveness  of  heat  transfer  in  the tower.
      Microbiocides can  be classified as  oxidizing agents, enzyme poisons,
organic chemical compounds,  and miscellaneous compounds.  The oxidizing
agents include chlorine,  bromine,  and  iodine.   Enzyme poisons include
methylene  bisthiocyanate, acrolein, and  heavy metals (e.g., copper
sulfate, copper citrate, tin, phenylmercuric acetate', methyl mercury).
Acrolein and the heavy metals are  not widely used and are not known to be
used  in CCT's.  Organic  compounds  normally require high dosage rates  and
include dodecylquamidine hydrochloride and quaternary ammonia salts.   Most
microbiocides used for  treating cooling water are included  in the
categories above, but dithiocarbamates are a class of miscellaneous
compounds  that also are effective microbiocides.  However,  they  reduce
chromate and, thus, cannot be used in chromate-treated systems.10'12
                                   3-18

-------
      Organic chemical compounds which either  hydrolyze  to  relatively
 nontoxic forms or can be detoxified are also  used as micobiocides.
 Hydrolyzable materials include 2,2-dibromo-3-nitrilopropionamide,
 chlorinated cyanurates, and halogenated hydantoins.  Chemicals which are
 both hydrolyzable and detoxifiable are methylene bis-thiocyanate and
 bromonitrostyrene.  Isothiazolin is a widely used biocide that can be
 detoxified.10*18
 3.5  MAINTENANCE REQUIREMENTS
      The maintenance requirements for CCT's and associated heat exchangers
 are affected by the effectiveness of the chemical treatment programs.   The
 effectiveness of the corrosion inhibitor is important to the life
 expectancy of metal  surfaces exposed to the recirculating water in both
 the tower and heat exchanger.   Excessive corrosion will  cause premature
 replacement of equipment  and additional  maintenance.   Metal structural
 components of the tower such as steel  columns  and beams, connector plates
 and bolts, piping and  pumps, valves,  and controller equipment corrode  from
 exposure to the cooling tower  water.   However, the most  critical  corrosion
 occurs  in the heat exchangers. Many  CCT operators have  Eddy-current tests
 conducted periodically (every  1 to  5 years)  to detect  heat  exchanger tubes
 in  danger of failing,  and these tubes  are either  replaced or
         1925
 plugged.   ~    Some  loss of  efficiency or cooling capacity  because of
 plugged  tubes  can usually be tolerated because  cooling systems are often
 designed  with  a margin of safety.
     Scaling occurs  on the surfaces of CCT components and heat
 exchangers.  Scaling in the  CCT can reduce tower heat rejection capacity
 by  interfering  with  splash or  film fill water distribution and the
 formation of water droplets.   Airflow characteristics also can be altered
 when airflow passages in the fill and drift eliminator become blocked;
 this increases  the pressure drop across the system and reduces cooling
 efficiency.  The  tower fill  and drift eliminator surfaces typically are
 inspected annually and cleaned if necessary.  Winter is the most common
 time to perform the work since the demand, '.if any, on  the system is
minimal.  Scaling and fouling in heat exchanger tubes  reduce the heat
transfer capacity because  of the low conductivity  of  the  crystalline  film
created on the tube surfaces and flow restrictions that result.   The  scale
                                   3-19

-------
 and  foulants  can  be  removed  physically by a process called rodding, by
 water  blasting, or by  flushing with  acids.   Physical  cleaning methods
 require  that  the  heat  exchanger be taken  out of service until the
 maintenance is completed.   In  CCT systems,  the  heat exchangers are
 routinely  inspected"  and  cleaned,  if  necessary,  each winter when the system
 demand is  low.  Chemical cleaning can  take  place while  the system is  in
 service, but  the  effectiveness of this method depends upon the level  of
 scaling  or fouling.  If  strong acids are  to be  used, the heat exchanger
 will be  taken out of service to protect ancillary cooling water system
 components.   Increases in  scaling and  fouling can result from changes  or
 variations in the water  treatment program,  which may also increase the
 level  of maintenance required.
 3.6  FORMATION OF DRIFT
     Water droplets  are  formed  as water splashes down through  the  fill
 material and  from the shearing  action  of  the airflow along  the water
 surfaces within the  tower.   Water droplets  and  the  dissolved  solids they
 contain that  become  entrained  in  the air  and are emitted  from the  stack
 are referred  to as drift.  The  amount  of  drift  is a function of both the
 water  and  air flow rates through  the tower, but the airflow rate has the
                                  27 28
 largest impact on the drift  rate.  *    The  rate of water  flow through the
 fill material is referred to as "water  loading," and the  typical range of
 water  loading in cooling towers is 81.5 to  204  liters per minute per
 square meter  (a/min/m2)  (2 to 5 gallons per minute per square foot
           2   8
 [gal/min/ft ]).   In some CCT's,  however, the water loading may be as high
 as 610 a/min/m2 (15 gal/min/ft2).27
     The velocity of the airflow  in the fill typically is 91.4 to
 213 meters per minute (m/min)  (300 to 700 feet per minute (ft/mini).  At
91.4 m/min (300 ft/min),  a 370-micrometer (urn)  (14.6 thousands-of-an-inch
 [mils]) water droplet will  become entrained in the airflow.  At 213 m/min
 (700 ft/min), an 800-pm  (31.5-mils) water droplet will  become entrained in
 the airflow.     A drift eliminator manufacturer indicated that the drift
 rates are  highest when the air velocity is at either end of the range.
Most towers are designed  with an airflow rate that produces a drift rate
 as near to the minimum as is practical.  In addition,  better drift
eliminators expand the range of airflow rates that produce minimum drift
                                   3-20

-------
 rates and they reduce the effect of substantially higher or lower airflow
                         2 fl
 rates on the drift rate.
      All droplets evaporate to some extent while they are entrained.
 Table 3-4 presents the sizes that various droplets will attain after
 3 seconds in environments of 80 and 90 percent relative humidity at 26.7ฐC
 (80ฐF).  These conditions probably are comparable with those in a CCT.
 The final droplet sizes presented in Table 3-4 indicate that droplets that
 are created at 30 urn (1.2 mils) and below will undergo a significant
 reduction in size as they evaporate and cause concentration of the solids
 in those droplets.
      Data from EPA-sponsored emissions tests indicate that a number  of
 droplets larger than 1,000 ym (39.4 mils)  were emitted from the stack.  As
 discussed above,  entrainment of these  droplets in the airflow  through the
 fill  would not be expected.   However,  the  air velocity may not be  uniform
 across  the face of the  fill.   The variation  is minimized  in  the most
 effective tower designs,  but,  in  many  towers,  the variation  is
 significant.   Thus,  droplets  larger than 370 wm and 800 urn can  be
 entrained in  the  higher velocity  portions  of the  air  stream.  Also,
 because  of the angularity of  flow out  of the last pass of the eliminator,
 the air  velocity  leaving the  eliminators is  higher than that through the
 fill.  Another possible explanation for larger  droplets than expected
 being emitted  is  that the large droplets are created  in the stack.  The
 sides of  the  stack, the fan blades, and other support structures in the
 stack are  constantly bombarded with drift droplets.  Eventually, the
 surface tension is overcome and large droplets fall off.  Because of the
 high velocity  (average of about 457 m/min  [1,500 ft/min])  in the stack,
 larger droplets can be entrained  in the stack than can be entrained in the
 fill.  Also, large droplets may be created from smaller droplets impacting
with each other because of the turbulence created  by the fan.
                                   3-21

-------
             TABLE 3-4.  EFFECT OF EVAPORATION ON DROPLET SIZE

Original
droplet
size,
vim (mils)
500 (19.69)
300 (11.81)
100 (3.94)
50 (1.97)
30 (1.2)

Particle size
at dryness.
pm (mils)*
36.0 (1.4)
21.6 (0.85)
7.2 (0.28)
3.6 (0.14)
2.2 (0.09)

Droplet size,
80 percent
relative
humidity
499.3 (19.66)
298.9 (11.77)
96.6 (3.80)
42.7 (1.68)
15.0 (0.59)

um (mils)5
90 percent
relative
humidity
499.7 (19.67)
299.5 (11.79)
98.4 (3.87)
46.7 (1.84)
24.2 (0.95)
Solids
concen-
tion, jjpm3
80 percent
relative
humidity
1,004
1,007
1,109
1,605
8,000
aAssumes total3dissolved solids content of droplets is 1,000 ug/ml
 (0.0624 Ib/ft ) and that the dissolved solids are primarily calcium
 carbonate (35 percent),'magnesium carbonate (48 percent), and sodium
 carbonate (17 percent).  Also assumes that the specific gravity of
 resulting dry particulate is the same as the weighted average of the
.specific gravity of the three major components.
 Assumes an evaporation time of 3 seconds and 26.7ฐC (80ฐF) dry bulb
 temperature.  See Reference 29 for the equation used to calculate the
 droplet size.
                                   3-22

-------
 3.7  REFERENCES FOR CHAPTER 3

  1.  Telecon.  P. Bellin, MRI, with I. Kuharic, Marley Cooling Towers.
      July 5, 1985.  Comfort cooling tower population.

  2.  Telecon.  M. Upchurch, MRI, with J. Carroll, BAC-Pritchard.
      September 3, 1986.  Comfort cooling tower population.

  3.  Telecon.  M. Upchurch, MRI, with L. Cornwell, Star Cooling Towers of
      Houston.  September 3, 1986.  Comfort cooling tower population.

  4.  Telecon.  C. Green, MRI,  with R.  Kellogg, Jr., Dubois Chemical
      Division—CHEMED Corp.  March 29, 1985.   Chromium-using comfort
      cooling towers.

  5.  Telecon.  C. Green, MRI,  with P.  Thomas,  Drew Chemical  Company,
      April  2, 1985.   Chromium-using comfort cooling towers.

  6.  Telecon.  C. Green, MRI,  with J.  Lee,  Olin Water Services.   April  2
      1985.   Chromium-using  comfort cooling  towers.

  7.  Kelly,  G.  M.  A  System-Efficient  Approach to  Cooling  Tower  Energy
      Modifications.   Cooling Tower Institute Technical  Paper
      No.  TP-85-18.  New Orleans,  Louisiana.  January 1985.

  8.   Chemical Engineers'  Handbook.   5th  Edition.   Perry and  Chilton,
      eds.  New  York,  McGraw Hill.   1973.

  9.   Mechanical Engineer's  Handbook.   8th Edition.   T. Baumesiter, Editor-
      in-Chief.  New York, McGraw  Hill.   1978.   pp.  19-11.

10.   Handbook of  Industrial Water Conditioning.  8th Edition.  Betz
      Laboratories.  Trevose, Pennsylvania.  1980.

11.  Strauss, S.  and  Puckorius, P.  Cooling-Water Treatment for Control of
     Scaling, Fouling, Corrosion.  Power Magazine.  June 1984.
     pp. 5-1  -  5-24.

12.  McCoy, J. W.  The Chemical Treatment of Cooling Water.  New York
     Chemical Publishing Company, Inc.  1983.

13.  R. Gibson, MRI, to R. Myers, EPA:ISB.  Trip report:  North Carolina
     State University, Raleigh, North Carolina, on June 17, 1986.

14.  D. Randall, MRI,  to R.  Myers, EPA:ISB.  Trip report:   Crabtree Valley
     Mall, Raleigh, North Carolina, on July 10, 1986.

15.  C. Mumma, and M.  Putnam,  MRI, to R.  Myers, EPA:ISB.  Trip report-
     Duke University,  Durham,  North Carolina, on July 3,  1986.

16.  Letter and attachments.  Jennings, T.,  Occidental  Chemical
     Corporation,  to  Farmer, J.,  EPA:ESED.   August  1, 1986.   Response to"
     Section 114 information request.
                                   3-23

-------
 17.   Letter  and  attachments.   Richardson,  M.,  Pfizer  Inc.,  to  Farmer,  J.,
      EPA:ESED.   September  29,  1986.   Response  to  Section  114 information
      request.

 18.   Letter.  Macht, W. A., Betz Entec, to Randall, 0., MRI.   April  10,
      1987.   Comment on draft Regulatory Imoacts Analysis.

 19.   C. Mumrna, MRI, to R.  Myers, EPA:ISB.   Trip report:   University  of
      North Carolina at Chapel  Hill, Chapel  Hill,  North Carolina, on
      July 2, 1986.

 20.   D. Randall, MRI, to R. Myers, EPA:ISB.  Trip report:  Greenbrier
      Mall, Norfolk, Virginia,  on July 15,  1986.

 21.   D. Randall, MRI, to R. Myers, EPA:ISB.  Trip report:  Humana Bayside
      Hospital, Norfolk, Virginia, on July  17,  1986.

 22.   D. Randall, MRI, to R. Myers, EPAiISB.  Trip report:  Old Dominion
      University, Norfolk, Virginia, on July 16, 1986.

 23.   D. Randall, MRI, to R. Myers, EPA:ISB.  Trip report:  Omni
      International Hotel, Norfolk, Virginia, on July 15, 1986.

 24.   D. Randall, MRI, to R. Myers, EPA:ISB.  Trip report:  Sovran Bank,
      Norfolk, Virginia, on July 15, 1986.

 25.   R. Gibson, MRI, to R. Myers, EPA:ISB.  Trip report:  North Carolina
     State University, Raleigh, North Carolina, on July 22,  1986.

 26.  Baker, D.  Cooling Tower Performance.  New York,  Chemical  Publishing
     Company.  1984.

27.  Telecon.  D. Randall, MRI, with  T.  Depalma,  Custodis-Cottrell.
     June 4, 1987.  Water loading in  comfort cooling towers.

28.  Letter.  Holmberg, J. D.,  The Marley  Cooling  Tower Company,  to
     Crowder, J., EPA:ESED.  February 24,  1987.  Comment on  Draft
     Regulatory Impacts Analysis.

29.  Chemical Engineers'  Handbook.  3rd  Edition.   John  H.  Perry,  ed.  New
     York,  McGraw-Hill.   1950.   p.  806.
                                   3-24

-------
                   4.  MODEL COMFORT COOLING TOWER SYSTEMS

      This chapter provides a discussion of six model comfort cooling
 towers (CCT's) that were developed to represent the CCT population
 nationwide.  It has been assumed that the sizes and types of towers are
 evenly distributed among all of the building types in the source
 category.  The parameters used to estimate the nationwide economic and
 health risk impacts attributable to regulatory alternatives for
 controlling hexavalent chromium (Cr+6) emissions are presented.  Specific
 items discussed include factors related to cooling towers, refrigeration
 equipment/heat exchangers,  chemical  treatment programs,  and baseline
 emissions.
 4.1  MODEL  TOWER PARAMETERS
      Model  tower design parameters  and associated  assumptions  are
 summarized  in  Table  4-1 and are discussed  in  further detail  below.   Sample
 calculations are presented  in Appendix F.
 4.1.1  Building Size
      Data on the  number of  commercial  buildings  in the U.S.  were  obtained
 from  a  Department of Energy (DOE) study.1  Total floorspace  was used to
 distribute the  buildings  into seven categories:  buildings with 465 square
 meters  (m2) (5,000 square feet  [ft2])  or less of floorspace, those between
 465 and 929 m2  (5,001 and 10,000 ft2), those between 929 and 2,323 m2
 (10,001 and 25,000 ft2), those between 2,323 and 4,645 m2 (25,001 and
 50,000 ft ), those between 4,645 and 9,290 m2 (50,001 and 100,000 ft2),
 those between 9,290 and 18,581 m2 (100,001 and 200,000 ft2), and those
with more than  18,581 m2 (200,000  c2).  An average building size was
determined for each size range.   These building sizes  were  then used as
model  buildings (to determine the cooling requirements  for  model  CCT's),
which  also are presented in  Table 4-1.   A model  for buildings under 465 m2
                                   4-1

-------
 (5,000 ft ) was not developed because it is not cost effective for
 buildings of that size to have a CCT.
 4.1.2  Cooling Requirements
      Table 4-1 also presents the cooling requirements for the model
 towers.  The required cooling capacity of central  HVAC systems for large
 buildings is based on the amount of floorspace in  the building.  Central
 HVAC cooling capacities typically range from about 107 watts (W)/m2
 (34 Btu/h/ft2) for office buildings to 158 W/m2 (50 Btu/h/ft2)  for  •
 buildings such as shopping malls that accommodate  a greater concentration
 of people or such as hospitals that have heat-generating equipment.2'3
 However,  the capacity of the cooling tower must be about 25 percent
 greater than the capacity of the HVAC system to account for the heat added
 to compress the refrigerant.   Thus, the cooling requirement for CCT's
 ranges  from 134 W/m2 (43 Btu/h/ft2) to 198 W/m2 (63 Btu/h/ft2).   The
 cooling requirements for the model  towers  are  based on  142  W/m2
 (45 Btu/h/ft ),  a value near the low end of the range,  because  office
 buildings are by far the most prevalent type of commercial  building.
 4.1.3   Recirculation Rate
     To determine the recirculation rate,  a heat balance  based  on  the
 defined cooling  requirement  for  the tower was  performed using
 Equation  1.*   It was assumed  that the  cooling  range  is  5.6ฐC  (10ฐF) (a
 typical temperature  range derived from case studies  and vendor
 information).5"13
 where:
        Recirculation rate
                                          Heat  duty
                             (Heat  capacity  of  water)(AT  of  water)
(1)
     Recirculation rate [=] n/min (gal/min)
     Heat duty [ป] W (Btu/min)
     Heat capacity of water = 4.18 J/gปฐC (1.0 Btu/1bปฐF)
     AT of water - 5.6ฐC (10ฐF)
4.1.4  Evaporation Rate
     Evaporation rates based on an empirical relationship were calculated
using Equation 2.llf
                                   4-2

-------
       Evaporation rate =  (0.00085)(Recirculation rate)(AT of water)    (2)
                                          ! '-
 where:
      Evaporation rate [=]  a/min (gal/min),
 4.1.5 Slowdown  Rate
      Slowdown  rates  depend on cycles of concentration as shown  in
 Equation  3.    Cycles of concentration can vary widely with  the  treatment
 program used and the makeup water quality.  Cycles  of concentration
 ranging from 2 to 8  have been reported.   A uniform  assumption of 5 cycles
 of concentration (the midpoint of the range)  was used for all six model
 towers.   The assumption is supported by  several CCT operators.6'11'13
               Slowdown rate
                                 	Evaporation rate
                                (Cycles of concentration-1)
(3)
4.1.6  Hexavalent Chromium Emissions
     Data on chromate concentrations in CCT water treatment programs vary
from less than 1 ppm to 20 ppm chromate.9'15'16  Although a large amount
of data has shown that the average concentration of chromate  in  industrial
process cooling towers is 13 ppm, there are not sufficient data  on CCT
systems to justify using a concentration other than the midpoint of the
observed range.17"25  Therefore, the chromate concentration in the
recirculating water for the model CCT's is assumed to be 10 ppm  (10 ppm
chromate is equal to 4.48 ppm Cr"*"6).
     Four chromium emission factors developed from EPA-sponsored tests
performed on two industrial cooling towers equipped with lower efficiency
                                                    These emission factors
drift eliminators are presented in Table 4-2.26'27
                                                                     .+6
relate chromium emissions to the chromium concentration in the
recirculation water in units of milligrams of CrrS emitted per ppm Cr" in
the recirculating water per liter of recirculating water (mg Cr+s/ppm
Cr+ /j. H20) (pounds of Cr+6 emitted per ppm of Cr+s in the recirculating
water per gallon of recirculating water [Ib Cr+6/Ppm Cr+6/gal H2OJ).
Because the product of the recirculating water flow rate and the Cr"1"6
concentration in the water is the recirculating Cr"1"6 flow rate, the
emission factor also can be expressed as mg Cr+s emitted/mg Cr"1"6
recirculating (Ib Cr+6/lb Cr"1"6).  The recirculation rates,  chromate
                                   4-3

-------
 concentrations, and cooling ranges are lower for CCT's'than for industrial
 towers.   However, because the emission factors are independent of these
 parameters,  they should be applicable to CCT's as well  as to industrial
 towers.   The lowest emission factor for an individual  riser cell  was
 6.6 x  HP5 mg,Cr+6/mg Cr+6 (6.6 x 1(T5 Ib Cr+6/lb Cr+6),  and the  highest
 was 1.874 x  10"3 mg Cr+6/mg Cr+6 (1.874 x 1CT3 Ib Cr+6/lb Cr+6).   The
 lowest emission factor is incorrect because some of the Cr*6 in the
 samples  was  retained on the walls of the beakers used  to  concentrate the
 samples.  Because of this phenomenon,  the aliquots that were analyzed were
 not representative of the original  sample,  and the resulting Cr*6  concen-
 trations  are lower than the actual  Cr+s  concentrations.   Thus,  the lowest
 emission  factor should be higher than  6.6 x 10"5 mg Cr+6/mg  Cr+6,  but it
 is  not expected that ฃhe_am.ount of  Cr+6  retained on the walls of the
 beakers accounts for most of the difference between the lowest  and highest
 emission  factors.   The wide differences  between  emission  factors indicate
 that emissions  may vary substantially  with  time  for an  individual  tower  as
 well as from tower to tower.  Because  the national  average Cr+6 emission
 cannot be estimated  accurately  with  the  available  data, CCT  emissions have
 been calculated  using both  emission  factors  to provide an emissions rate
 range for each model  tower.   Thus, the emission  rate range for each model
 tower was calculated  by multiplying  the  emission factors by the model
 recirculation rate  in  liters  per minute  and  the  Cr+6 concentration of
 4.48 ppm as  shown  by  Equation 4.
where:
Cr+  emissions = K • CCr+sปR
                                                                       (4)
     Cr+G emissions [=1 mg Cr+6/hour (Ib Cr+6/hour)
     K = Cr+S emission factor, mg Cr+6/mg Cr+s (Ib Cr+6/lb Cr+s)
     C(-r+6 = concentration of Cr+s in cooling water, ppm.
             (For the model towers, this equals 4.48 ppm)
     R = recirculation rate of cooling water, liters/hour  (gal/hour)
             (1 liter of water equals 1,000,000 mg of water)
             (1 gallon of water equals 8.33 Ib of water)
                                   4-4

-------
 4.1.7  Stack Parameters
      Table'4-3 lists the stack parameters developed to perform dispersion
 modeling on all model CCT's.  To define these parameters, it was necessary
 to estimate the airflow rates through each model tower.  A leading CCT
 manufacturer indicated that fans and stacks are designed to maintain stack
 velocities between 365 and 670 meters per minute (m/min) (1,200 and
 2,200 feet per minute [ft/min]).28  Also, the liquid-to-gas ratio (L/G)
 for an individual tower can vary from 0.5 to 2.5:1.  The typical L/G ratio
 for a CCT of 1.5:1 was used to calculate the airflow rate for each model
       28
 tower.    Next, the airflow rate and a velocity of 520 m/min
 (1,700 ft/min)  were used to calculate model  stack diameters.   Finally,
 because standard diameter fans are not available in fractions of a foot,
 the calculated  diameter was rounded to the nearest whole number.   The
 stack velocity  was then recalculated, and, in each case,  it  was within  the
 design range of 365 to 670 m/min (1,200 to 2,200 ft/min).  The range of
 recalculated velocities was 424 to 556 m/min (1,390 to 1,824  ft/min).
      Actual  stack heights (distance from the base of the  tower to  the top
 of  the stack) for the larger towers were determined from  manufacturer's
 product literature,  and the heights of the smallest towers were
 extrapolated from the available information.   Comfort  cooling  towers are
 most  commonly mounted at ground level  near a building  or on the roof of
 the building they serve.   Table 4-3 presents  stack  heights corresponding
 to  both ground-mounted  and  roof-mounted  CCT's.  An  average number of
 floors  for each model  building  was  determined  from  DOE data, and it was
 assumed  that each  floor  adds  3.7 m  (12 ft) to  the building height.1
     As  mentioned  above,  heat transfer occurs  because of evaporation (heat
 of  vaporization)  and  sensible heat  differences.  The dry bulb temperature
 and the  relative  humidity of the ambient air determine both the relative
 effect of each type of heat transfer and the temperature of the air in the
 stack.   For the typical case in which heat of vaporization accounts for
 80 percent of the  heat transfer and sensible heat accounts for 20 percent
 of the heat transfer, the dry bulb temperature of the air in the stack
would be sevev ;1 degrees (Fahrenheit) warmer than the dry bulb temperature
of the ambient air.  A CCT is most likely to operate during the spring,
 summer, and fall.  Therefore, the mean nationwide dry bulb temperature of
                                    4-5

-------
the air 1n the stack for these three seasons of the year  should be used  in
the dispersion modeling; it has been assumed that 27ฐC  (80ฐF) reasonably
approximates this value.
4.1.8  Maintenance
     The towers are assumed to be shut down for at least  2 weeks in
winter.  At this time, the nozzles can be cleaned, broken slats can be
repaired, drift eliminator spaces can be cleared, and any other routine
repairs can be performed.
4.2  MODEL HEAT EXCHANGER PARAMETERS
     Information from the case studies has indicated that compression-type
refrigeration units are used in nearly all CCT applications and that the
heat exchangers have copper tubes and a cast steel shell. ~  '       Thus,
these features have been assumed in the development of model equipment.
     Because the CCT operators contacted during the case studies and
limited corrosion data indicate that it is possible to obtain equivalent
corrosion protection with nonchromate treatment programs, it has been
assumed for the model system that equipment life is not affected by the
choice of corrosion inhibitor.6ป7t10,11,13,29-32  The case studies also
have confirmed that systems using nonchromates do not require more
frequent heat exchanger maintenance than when chromates are used.
Therefore, it has been assumed that maintenance on the heat exchanger will
be performed annually in both cases.
4.3  BASELINE EMISSIONS
     From information supplied by industry vendors,  it was assumed for
purposes of analysis that there are approximately 250,000 CCT's nationwide
and that 15 percent (approximately 37,500) of all  CCT's use
chromates.  ~38  To estimate nationwide emissions, these 37,500 CCT's were
apportioned among the model plants.   The DOE study contains data
concerning the number of buildings that have central  cooling in each  of
the building size categories.1  However, not all  of  these cooling  systems
use CCT's.  General HVAC design principles indicate  that the larger the
building, the more likely that CCT's would be chosen  for HVAC cooling.
Also, as the square footage exceeds  9,290 m2 (100,000 ft2),  HVAC design
criteria make CCT's the overwhelming choice for central  cooling.   However,
little data are available concerning the actual  distribution of CCT's over
                                    4-6

-------
 the various size ranges.  Thus, the following size distribution was based
 on engineering judgments:  5 percent of buildings between 465 and 929 m2
 (5,001 and 10,000 ft ) were assumed to have CCT's, 25 percent in the 929
 to 2,323 m2 (10,001 to 25,000 ft2) range, 40 percent in the 2,323 to
 4,645 m2 (25,001 to 50,000 ft2) range, 60 percent in the 4,645 to 9,290 m2
 (50,001 to 100,000 ft2) range, 90 percent in the 9,290 to 18,581 m2
 (100,001 to 200,000 ft2) range, and 95 percent over 18,581 m2
 (200,000 ft ).  The data do not indicate that the percentage of CCT's
 using chromates varies with CCT size.   Therefore, it was assumed that
 15 percent of the CCT's in each size range use chromates.   The number of
 buildings with CCT's in each size range and the estimated  number of model
 CCT's using chromates are presented in Table 4-4.
      The Cr+  emissions per State were estimated from the  number of model
 towers 1n the State and from the Cr+6  emissions for individual  model
 towers in each State.   Table 4-5 presents  the number of  model  towers
 assigned to each  State as determined by the population of  each  State from
 the 1980 census and assuming that the  proportion of each model  tower  size
 is  constant among the  States.   The Cr+6 emissions for individual model
 towers in each State are dependent upon the utilization  rate for the
 State.   For the purposes of  analysis,  the  utilization rate  is the
 percentage  of  the number of  days  that  the  fan operates annually.  The
 utilization rate  depends on  the climate at the CCT site  and the building
 use;  but for an approximation of  the rate,  it was  assumed that the fan is
 not used on days  when the average  temperature is  below 15.6ฐC (60ฐF).39
 This  rough  approximation incorporates two  errors that may offset each
 other:   it  is  known  that some towers do operate when the temperature is
 much  less than  15.6ฐC  (60ฐF) and that they do not  necessarily operate for
 24 hours  on days  when the  temperature exceeds 15.6ฐC  (60ฐF).
 Meteorological  data were  examined  to determine the percentage of days that
 the average temperature  is above'15.6ฐC (60ฐF) in each State.39  Table 4-6
 presents the percentages  for each  State., which range from 0 percent in
Alaska to 100 percent in Hawaii.  The average nationwide utilization rate
of 46 percent was determined by multiplying each State utilization rate by
the number of towers in the State and then dividing the sum of the
products by the total number of towers  nationwide.
                                   4-7

-------
     The annual Cr+6 emissions  from  individual model  towers  in  each  State
were estimated by calculating annual emissions from the  hourly  emissions
presented in Table 4-1 and multiplying by the utilization  rate  for the
Stste.  The estimated individual emission rates for the  six  model tower
sizes in each State were then multiplied by the respective number of CCT's
of that model size in the State to obtain the Statewide  Cr*6 emissions
from each model tower size.  The total Cr+s emission  rate  for each State
was determined by summing the Statewide Cr"1"6 emissions for each of the
model tower sizes.  By summing  the emissions for all  States, nationwide
Cr*6 emissions were estimated to be  7.2 to 206 Mg/yr  (8  to 227 tons/yr).
The average emission rate per person for each State was  estimated by
dividing the total Cr+6 emission rate for the State by the population of
the State.  Table 4-6 presents  the annual Cr+s emissions from individual
model towers in each State.  Table 4-7 presents the Statewide and
nationwide annual Cr*6 emission rates per model tower size and the average
annual Cr"*"6 emissions per person for each State.
     Risk analysis was performed using the area source model contained in
                                 I. A
EPA's Human Exposure Model (HEM).    (A description of the HEM is provided
in Appendix B).  Using this modeling technique with the  lower-bound  and
upper-bound emission estimates, the incidence of cancer from Cr+s.
emissions from CCT's is estimated to be from 4 to 112 cases per year.
Table 4-8 presents the nationwide risk for each model  and the entire
category.
                                    4-8

-------

o
oo
00
LU
BASELINE
>j
|
z
00
Qฃ
1
O
o
8
Oฃ
0
LL.
0
LU
0
o
r-H
1
LU
OQ
IS



E
•r—
JZ
O

c
•>>
"(O
a
^^
c
'e
0*
CU
3
u-


cu .
•ง:
2:

emissions
awn ner tower.
-c
o
OQ

R
LU
•^ e
1 tower Reclra
requirements latlo
cu
•O CT
o c
*o
c^
^n

a
'cu
•o
o
^
"8
ฐ
•^ ..
^^
L f.
•*+*
w
m
*->
ซ

CO
$.
4J
(U
2
M
O
.C
3
33
ซ^x
*^
E
1
O
4->

*^ LO o .LO o vTป o T or** oco
CMVO f^-CTV COCO LOLO OO LO •
i— 1 CM CM LO VO • i— 1 • CO • Cn LO
1 • 1 • 1 i— 1 ซCM "LO ,ซi-4
^•OLOOCMI i— II CMI VO 1
• 1* • 1 CM cn 1 OO 1 VO 1 cn
^ซ P** cn i"""4 ^T* T* <*O CD 1*^ f^ CM
งCM o tr o oo i-i T LO
O • • ซ CM •
O • O O O O
co ^r *r o LO CD LO oo 1-1 rr ^r
to i— i i— i co vor*s oo CM vo 10 COP^
OO i-*O CMO *Ti-H CTlCM C7>(^
1
00 LO 'O' O O CM LO VO
O LO LO CM VO OO ^" •— ( ^i—l OC3ป
CMO <<3'i— 1 OCM OปLO COO r-ปO
	 ' ^- ' i-H--' i-HNป^ CO-H ^-4 CO
VO LO ^-i-4 OO OCM O "ST OCM
^r vo co ^" LO co co o CM cn CD ^
CJซ— ' LO i— ( CMCO CM VO LO i— 1 OOVO
ปH CM 'S' i— 1 CO CO
f^ cri r^* I~H co c*^
CM LO CO LO Oป CM
i— 1 CM ^- LO
1—1
88 88 88 88 88 88
^roo i— i *f . o\ o oo oo oo
LOLO r-.r^ CMCT> CMLO o vo coo
O^CM oo oo^" oo i— i ior-ป co co
co CM P^ *^r vo co o r*^ cr* co CM
i— 1 CO <—l LO LO OD
•>—• ^-* ซ~ ' ^-t
coo oo LOO ^ro coo voo
lvป'3- VOCM OLO CM. rrvo CMCT* coco voo
!"•ป i— 1 LO COVO VOVO CMCM rซ.LO
	 ' i-l CO VO r-H CO CO O
*— < CM co ^r to vo
I]



o o o"*-^
C4 CM C^O
^ ^ a: eg
t-l— J- =
QJ (O QJ i^^
4*^ O^ -^^ to
••--ปป -^- o>
r-+0 r- ^ป
— + ^-o
-—. J- —-4-
00. 0 E
a. a.
Q& F*ฐ
Si '^ปtj •>ป. i.
—  J- ;— j_ .a
!t- o - o r-
^^ O
3 Dy DO
+f Si El
OQ O O
VO i— 1 *T i-4
LO vo x r>. x
•ซ• o oo co ^r
	 O O ~H VO
0 LO 0 LO
 —
—. 'C. " "
--^•u_ o $_ j_
U- o o O O
o LO i— 4^ X
LO OO • V(_ 
CM r-i r-H || || .,_ .^.
II ป— C E E
M u o +-> c cu 5
CU C_) -i- rซ 0
CD i- o 01 +j ai .,— = =
l-3VO4->tOj=4_>3 3
3 -4-ป • C V. (O -i- .1-
-MrtSLOCU-l-Ji — i-E E
fOS- ECfO-MO O
i-CUIICUCU4->Cl- 1-
CUQ. i-oocu^: >-
Q-ECU-i-C4->3O^C
VI CU-UCCTU-MOT3 -0
C4->  ja ja
Q-^-l_>.MOO(J.|->J_3 Q.
=t 33IOOO_IO-3 5"
4-9

-------
           TABLE 4-2.  EMISSION FACTORS FROM EPA-SPONSORED TESTS

                                                         Emission factor
Plant/location
Department of Energy,
Paducah, Ky.
Exxon, Bay town, Tex.
Mean (Exxon, Bay town) b
Tower/
fan cells
1/7,8
1/9,10
68/1,2,3,4
68/5
Flow type
Crossflow
Crossflow
Counterflow
Crossflow
mg Cr+6/mg Cr+6 or
Ib Cr+6/lb Cr+6
0.000075
0.000066
0.001063
0.001874
0.00147
aThese emission factors are incorrect because some of the Cr+  in the
 samples was retained on the wajls of the beakers used to concentrate the
 samples.  Thus, the actual Cr   emission factors are higher than those
 Indicated in this table.  This phenomenon also occurred with the samples
 in other tests, but the beakers were subsequently rinsed with aqua
 regia.  Both the concentrated sample and the rinse were then
 reanalyzed.  The amount of Cr   recovered increased by an average of 3.1
.times using this rinse procedure.
DData obtained in the test at the Department of Energy have been omitted
 from the mean because of the situation described above.  The mean is
 presented because crossflow and counterflow CCT's are evenly distributed.
                                   4-10

-------
TABLE 4-3.  STACK PARAMETERS FOR MODEL CCT's
,.


Model
tower
1
2
3
4
5
6


Stack
diameter,
m (ft)
0.6 (2)
0.9 (3)
1.2 (4)
1.8 (6)
2.4 (8)
4.6 (15)
Groind-
mourted
stack
height,
m (ft)
1.5 (5)
2.4 (8)
3.0 (10)
4.9 (12)
6.4 (16)
7.6 (25)
Roof-
mounted
stack
height,
m (ft)
8.8 (29)
9.8 (32)
14.0 (46)
19.5 (64)
24.7 (81)
44.2 (145)


Exit
velocity,
m/s (ft/s)
8.2 (27.0)
7.9 (26.0)
10.4 (34.2)
8.5 (27.8)
9.4 (31.0)
8.2 (26.9)


Exit
temperature,
ฐC (ฐF)
27 (80)
27 (80)
27 (80)
27 (80)
27 (80)
27 (80)
                    4-11

-------
        I
                    QJ
                    cn
        C  O
        0)
        o  cu
        s-  cn'Z
             o o
                     -o
           t/t  CD-4-J C
          -   c  ซ ซ
          I— ซr-  S VI
          CJ  CO  O 3
          CJ  3  1. 0
                    CO
                    
LO •
LO
CO
t— I

g
CO
CO

o
T— 1
LO

O
o
•
CM


g
6
.—I
CO

LO
l-s.
•
CO
CM

O
LO
O
LO
CM
ฃi
             O Si-
             -M -   CJ
CQ
       in
       cn    i—
       c ^. ia g
                 - in
      f- •=• OJ

      CO
o
             
                 CO
                                 CO
                                 CO
                       *ST   CO    VO
                         •     •     •

                       O   VO    CM
                       CM   >—I    r-1
                                          LOl
CO
r-.
CM
LO
CM
CO
r^
O
CO
10
CO
r-.
ซa-
CO
LO
CO
                                                             CO
                                 LO
                                       LO    LO    LO    LO
                           LO
                                 LO
                                 CM
                                   o
                                   cn
                   to
                   cn
          r>.
          co
                                 T    co
                                 co    CM
                                 co    t— i
                                   LO
                                   co
                   LO
                   CM
O
^*
CM
f^.
"•^
0
CM
f*^
LO
r-H
O
LO
^Pซ
^o"
CO
O O
cn o
cn co

                                   oo
                                   CO
                                   CO
                                    A
                                   CM
                                   LO
                                                        4-12

-------
           TABLE  4-5.  DISTRIBUTION  OF CHROMIUM-USING  CCT's  PER  STATE
STATE
ALABAMA
ALASKA (it
ARIZONA
UWMAi
CALIFORNIA
COLORADO
CONNECTICUT
DOMME
aoRin
GEWBIB
HAWAII
IDPM
ILLINOIS
INDIAN)
IOUA
KANSAS
KENTUCKY
LOUISIANA
MAINE
MARYLAND
MASSAOtttl IS
moiiam
MINNESOTA
MISSISSIPPI
MISSOURI
mDM
NEBRASKA
NEVPJM
NEB HAMPSHIRE
NEH JERSEY
NEUKIICn
*n vow
NORTH CAROLINA
NORTH OMWHA
OHIO
OKLAHBB
OflESON
PEWGYLVWIA
SWBE ISLAND
SOUTH CAROLINA
SOUTH DAKOTA
TENNESSEE
TBAS
UTAH
<ฃBOff
VIRSINIA
UASHINBTCN
ซST VIRBINIA
WISCONSIN
•Y3IIN6
•OSH1N6TON, DC
TOTflL FOR U.S. (U
Met*ป<"ปrn 1 nn i f-;i 1 Ha+a
POPULATION
4,030,000
448,000
3,010,000
2,400,000
24,800,000
3,110,000
3,130,000
610,000
10,600,000
5,710,000
1,020,000
1,020,000
11,500,000
5,570,000
2,940,000
2,410,000
3,790,000
4,360,000
1,170,000
4,330,000
5,730,000
9,400,000
4,160,000
2,600,000
5,010,000
826,000
1,600,000
891,000
981,000
7,450,000
1,390,000
17,400,000
6,110,000
667,000
10,300,000
3,170,000
2.300,000
11,300.000
753,000
3,280,000
700,000
4,790,000
15,100,000
1,530,000
530,000
5,590,000
4,310,000
2,000.000
4,320,000
514.000
508,000
233, ISO. 000
f i-ปrป A 1 -sซ~L> -ป

1
48
0
36
29
296
37
37
7
127
63
12
12
137
66
35
29
45
52
14
52
69
112
SO
31
60
to
19
11
12
39
17
208
73
8
129
33
33
141
11
39
3
57
130
19
6
67
51
24
58
6
7
5733


-
217
0
162
129
1332
167
168
33
570
307
55
55
618
299
158
129
204
234
63
233
311
505
224
140
369
44
86
48
53
400
75
935
328
36
530
170
ISO
634
51
176
38
257
311
35
28
300
232
107
259
i3
33
-.2525


• 3
133
0
: 99
79
: 817
102
103
20
349
133
34 .
34
379
184
97
79
125
144
39
143
190
310
137
36
165
. 27
S3
29
, 32
245
46
173
201
22
356
104
92
389
31
108
23
158

52
' 17
:34
142

159
17
JO
7580

ซ...
4
109
0
81
65
670
34
85
16
286
154 -
28
23
311
151
79
65
102
US .
32
117
156
254
112
70
135
22
43
24
27
201
38
470
165
18
292
36
76
319
36
39
19
129
•08
43

151
116
54
:30
.4
is
5300


5
32
0
61
49
503
63
63
12
215
116
21
21
233
113
60
49
77
38
24
38
117
191
34
53
102
17
32
13
20
151
28
353
124
14
219
64
57
239
19
66

37
206
' 32

,,,
37

53

12
>723

=C^B:^KMM
6
62
0

37
379
43

•3
162
87
16
16
176
35
45
37
58
67
18
66
38
144

4Q
77
13
24
14
13
114
21
266
93
10
165
43
ป3
130

50


531

3
35

,,
"4
Q
3
^=^=^
    — _._=.—  —.._  , „. ,,
 temperature exceeds  fiO'F.

Excludes population  of Alaska.
    indicate that,  on average,  there are no days  when me
Thus,  Alaska was not considered to have chromium  emissions
                                        4-13

-------




o
ป— 4
1—4
t— 4
QC
It
v%
o
I— <
CO
CO
HH
ซ0 1—4
•f 0ฃ
i. (—
CJLU
O
LU
Ujfe
iซ
I— LU

LU
CO
o ce
2SLU
1
LU LU
a. o
Q
1
r™%
3

*
at
^o
•
LU
*""













u
X.
c
ง
1
E
t-
i
in
c
C
as
U!
1
O
4-
ซ-
O








1
roซ
M
=> +



a
+
fl
+•
cr

lO




n


ป




tO






N







ซ.





>
•
1
i
o g T — p-
— g to in P-
— g p-ao to
o> O T o co
inotOT CN
IO 1 IO IO to
i g i i i
co g o in to
T g o\ g P-
too PปO in
CM • — CM —
• CD • • •

to o om T
ing r- to P-
• • • • •
~ฐ 2 — 2
t g i t i
p- g to to to
T o to to at
— O ooot P-
T • IO tO IO
O OOO
งg CNO to
gtoro to
ot o m to T
m o in in in
i i i i i
CN g o g — in p-
CN g — — o
NO CNCN CN
ooooo
ddddd




ot o m to T
in in in in


10
in c
: ro c tn o
.* O c —
in N 10 •—
— — t- t_ ~ oo CM
— in into T
to to toot to
ddddd
— to to — to
— CM CMT CM
— — — to CM
ooooo
• * • * *
ooooo




ot to to ^tt ot
CN to toco in
-
1
3
O-^CD
ฐO •(— L. (O 
— O OO O
OOOOO



in — to to eo
CM - G •a
* 10 C
in o •— 10
10 3 tn •
C C 3 — l_
10 u o 10 ia
COCO CNO CO
in in — — f
O O P- in in
77777
i i i i i
O O CN to O
P- P- tO CN Ot
o o o' — d
CN CN CO to T
to to p- inCM
OO 00 CO P- OO
in in p- p- to
to 10 in -co
o o oo i CM
CM CN top- in
CM CM CM — CM
odd ซo
O
IO IO tO O T
CNCN OปO CO
CN CN — T CN
tOIOOtOt CN
to to CM in T
1 1 1 1 1
• • * • •
OOOOO


to to to p- T
— — Ol T —
ao ao m CN to
— — — IO CM
1 1 1 1 1
1 1 1 1 1
OO to in in
T T IO T —
000*00



tO to — Ot to
ot o in to CN
P- P- OOOt Ot
P- P- IO to Ot
oo d — o
in in — — at
P- Pป T Ot T
CM o o in L.
o o> in tn 3
to — CD .— o
in jr c tn tn
+
 (O co d)
•^ u 
3

0
[o



































4-i4

-------
 (U
 3
 <0


V.
      i
     101ซ
     N
     - O
       
r>. CM oo ao o
rป a\ Pป in —
vo ot to ao —
aOOt rซ O O>
CM 10100 in
IO — CM CM IO
1 1 1 1 1
vo — O T as
rป to >o rป ซr
in ot to o vo
— f 001— CM
— 000 —

^ ^ r- CM 10
r* o to to in
r- o\ co co r-
rป in r— in r>
O ซr r- vo —
T rป — o T
vo — ซr CM r-
01 vo r-ป to T
r* — CM CM —
•oo'o^
O O
VO VO V IO O
vo in vo CM o
10 — CM CM V
T to ot to ot
'in CM to 10 in
i i i l i
in • TT CM f- Ot
CM in ot rป to
— o o o —
i i i i i
Ot — in in —
^ Ot CM r- ot
T — IO CM T
OOOOO
OOOOO

VOIO to — O
r- O *T Ot CM
oo in CM in ซr
in CM T 10 vo
OOOOO
"•• 00 Ot *Q VO
N O — — CM
OOOOO
OOOOO



in CM to to in





c
(0 -o —
•— c —
C IO O
IO — l_
10 > in 10
งc"ฐ
.c O vn Q) .C
IO Ol C T3 -f-
—  .c o
OO o. at v>
oo — r- in in
in to in oo vo
aoto •ป vo —
OT IO OO CM
ooซฎ2
l l l i i
^f ot in 10 ot
r>- — o T in
O r- in vo to
i*- O 10 vo in
O — — o O

CM O vo O in
vo ot to rป o
OO f*ป Is* 00 GO
mot in — ot
VO Ot CM VO *T
ii — ii
O in i ot t**
CM — oo r- in
10 in CM — rป
CM IO *T CM —
OO -O 0
o
10 ot rซ o Ot
CM IO CM — VO
CM to •ป CM —
too to — m
to in vo to CM
OlO T Oป VO
r* t^ toot oo
— rป CM o oo
— — CM — O
OOOOO

voo'om o
iป- ซr o r- p-
— m r- o r-
oo rป rr p- to
— CM IO— —
i i i i i
OOCM — in
*r rป CM O ao
vo ot CM  -t- 
oo CM oo in in —
10 to to oo vo 10
vo r- vo vo — 10
in — in 
00 Ov OO OO OO P*
IO Ot IO — Ot Ot
oo to ao vo T ot
1 1 1 1 1 1
CM vo CM ot r- in
in o in rป in —
ot ^r ot *~ r^ in
CM — CM CM — IO
oooooo

CO IO OO — VO IO
CM — CM CM — to
CMOCM — mo
T CM T IO CM in
1 1 1 1 1 1
Ol Ov Ot Ot VO IO
co o co o> co r~
•ป• r~ T o ao rป
— 0— — 0—
oooooo

T vo T in o O
10 — to t^ r~ ซr
— O — o rป in
to — 10 rป to r-
CM — CM — — CM
1 1 II II
in oo in .— in o
— ao •— o ao r-
ง10 oo vo T ot
OOOOO
OOOOOO

10 in 10^ vo 10
CM CM CM CM O —
OOOOO—
1 1 1 t 1 1
Ot VO O> 00 OO VO
•ป vo T in o —
IO — IO CM CM ^~
oooooo
OOOOOO

O vo O IO O O
in — in to r~ ^
T CM T fO CM in
??????
oooooo
oooooo



CN O  c = c
CD.C 4- O E ^
L. ift in m o m
— 10 0) — >. 10
IO
in
to

in
f
H-
10
in
10
c
u.
0
vo
in
•o
u
X
o
o
3
IO
Q.
1
c
IO
IO
 c
IO O
c in
O in
4-"!
o
t.
DO
o iu
u >
3 i:
- O
IO
^ u
— -o
•o —
^ tn
71 O
o u

5 i_
jl

: S
                                                         4-15

-------




i
a:
co
o
1— 1
C/1
s:^
10 *"
"*j- — '
, UJ
U- ,__,
uu ฐ-
= ^
G3 5"^
2 —1

= 1
a _j
^2
•
J3
VO
UJ
g







u
X
h
CJ
1
1
in
in
O
*




i
ซ3l
N
—
•—
ZJ H
<
j,
f
+
U





n



*

o


N


*
5
a
>
i
CM O P- VO CO
O O CM *r O
at o to m CM
1 0 1 1 1
^O to 10 —
ง0 Otto CM
o ot *r in
p^ • in to in
CM CMCM CM
Ot O CM to P-
in o ซr ปr to
to o to co ao
*r o CM Pป va
— O in 10 to
at -co ao 03

in o P^ at in
งO OCM 03
o CM v at
tOOCMCM —
7o 77 7
CMO OVCO —
— O OIP* CM
tO O CM IO CM
• f"t • • •
O OOO
ซr O Pป — to
*r O ao o P>-
vo O Pป o in
— O O
O .
T co oo m o
O ao 03 in to
in pป pป to o
7777?
CM in in to CM
too O to co
in va va vo o
o O o — —
OOOOO
Pป vo IO at IO
in — — TT in
ot ot ot to —
VO P- Pป — rr
OO O CM —
1 1 1 1 1
m ot at CM oo
CM CM CM Pป T
OOOOO
OOOOO
ot to to ot at
CM to to oo in
•a +- t_ 10 10
(O U IO "O —
L. - c •o
.* 10 C
in u — 10
IO 3 U) (D —
in *- — c >.
C C 3 •— l_
VO vo CM CM V
— — T O OO
03 03 — pป in
CM CM Ot — to
^ ^T CO Ot vo
1 1 1 1 1
vo vo vo ซr at
Ot Ot O CO ^
in in p- 03 ao
in in to r* ot
งงงงง
CM CM IO vo CO
in in p* at <ซ•
•9- T CM in 03
i I i I i
— — Ot T O
— — •
in  X I
O g> • Si s.
+- 4- O
X X I. U —
IU 1) O O J=
'Z.'Z.'Z'Z.O
S
c
+-
J















4-16

-------
•a
(U
o
u
1 '











u
V.
JC
g


!
u
a
> a.
in
ง


3
>
CO O CO to CM
O ro CM — o
to to ro oo Pป
T eo ro CM —
CM O CM T ON
p* ro in ^* pป
1 I i I I
— O CM to T
CM r> ro ON CO
in co O T — T
ro m o — —
co ro to tn ON
OOOOO


to O CM in ro
in in to *r in
co O in CM ON
ON — toroo
— in COP- ro
— 1 1 1 —
1 CO ONO 1
— ON T CO CM
 — co ro
•™ ON ro •— r**
OO — O r- O
CM — eg — 10
i t i i i
— CM in in CM
JNCM — O CO
3 O OO —
OOOOO

ro ป*• in to ro
in — in — in
ON in ro ON —
eg in ON r— ซr
— O oo —
i i i i i
tO T ON ON CO
in ON CM r* ot
S — ro eg T
O OO O
OOOOO



ซr ro ONIO ON
in CM ro ro in






•o
c
10 TD —
— C —
e 10 O

t > in 10
i ^~ "••• CJ
1 S *•
-c o in m .c
 rr — ro
ซr to co ซป ro
1 i i I i
to O T — in
ON ro rป in —
in to r- to eo
in ro ON T —
— eg eg — —

OOOOO
OOOOO
CM O CM ^ O
m o rป to o
ซr eg r>ซ ro —
— eg eg — —
i i i t i
^r GO ro ^r ^r
— T to o r-
— r* r^ eo oo
in r* ON ^r ro
OOOOO


in r-ป eg to ON
^r Pป to O 03
eg ON oo 03 T
ro O ON GO in
r— — ro to in
1 — *- • 1 1
ง1 1 ro •ป
ON ine>i in
in o eg T ON
eg ON ON eg —
O '"^OO
O O
eg in' — rr co
p. — o T in
O P-* in to ro -
O O tor-- O
*r to r*. ro ro
i i I I i
— CO T tO ON
— ro ON eg to
T — to ro o
— eg eg — —
OOOOO

co eg roto —
co v — ซr eg
— O eo — o
t— to eg to ro
_  -r- 01
 •>
co r- co rป to ro
in ro in ON ซr ON
ro GO ro in in O
to to to — ro r-
in eg in t ro to
1 1 i i i
ON eg ON — in o
T in T in — ro
oo T ao to co to
ON ON ON T — IO
— p — — — eg

o o o o o o
o o o o o o
03 O O3 T O O
•S- CO T to O O
co oo oo ro — eg
— i — — — eg
1 ON t 1 1 1
ON ON ON ^ ^ CO
O O O O r— *r
in ro in 03 GO r*ป
to • to T rorป
O O O O O


— — — to ON f.
CM ON CM O 03 rป
CM ro CM co *r ON
ro *r ro co in o
ON *T ON to in —
1 i i i i —
ro ro ro ro ปT I
co to co eg in ON
eg in eg rr ON o
ro — ro eg — ON
O O O o' O "2
O
— to — T co in
O 03 o t in —
O eg o to ro r-
— TT — r-. o O
in eg in ro ro to
i i 1 i I i
to in to to ON co
ON m ON eg to ro
rซ co rป- ro o —
— O — — — eg
O O O O O O

in r>ป m to — eg
co ซr co — o o
— O — to ro to
eg — eg — — eg
1 1 1 1 1 1
O r-. O ON ON r-
rซป to r^ to in —
rป ro r. in T ON
OOOO OO
O O OO O O

in co in to r* *r
P* ON P* IO ON ON
O r-. o  c c c

O1.C 4- O E ^
L. m tn tn o ป IO
=> 3 3 3 3 3
II O
in
(O
^^
in
3
! —
IO
in
IO
c



O
to
in
•a
o>
u
X
0)
C-

IO
0)
a.
ฃ
 m
> c

.— IT3
C
— O
IO
IO 
-------
                                   i  CD o  — o  —
                                                                          ffl     *i     ป*
                                    a  ฐ a a  g     s  s s  5
                                    J5     -* m  CD     r-  c> m  —
                                    V     /-T c\T   *     —  -T     2"
                                                            I    t   I   I    I
                                                                                     -•  oj cu oj      cu
                           lilt

                      P  ซ  ซ2 2  S
CU  CU  — tp  CU     —  OJ  CU —  CU
                                                                                 iป-  f-n  m  •- •ซ>      cu  a* in  •*  —*     3Q
                                                                                 en  •*  r-ป  in —      iTto-vfai     so  <3ป
                                                                                                        -^"  en" in
                                          I   I   I    I
                                         o oj o  ea
                                                                      g  =
       OS
      LU
  U. Q-
 LU O
3E <^>
*—I *-H
OO t-Lj

"*
9s-

                                     Oj O O  ^  C*-     in O^  •*•
                                  i   I   I   i   i    i       i   i   i    I   i       i    I   I   t    I       i   i    I   I
                                  t   ff% ^  tf|  *•  03     co  p1* ^  *  •&     tn  ••* *o  P*>  ^     •* ^  •**  i"
                                  iro<5ปn*?irf     ojVcu^ซa     V  n ซ  Q  (rt     '*AiS'*1
                                  i   m     cu  —  e>     —'  —     m-ซ-     —•     Cart-*     —• cu  ^*
                                                                                                                                      I    I   I       I   I
                  i       i      i  a      5  S a     a s  ^  is
                  ป       '      •                 -r                ^r
                                                                         r*-     eo  0 *•  ซ4  •*     "^  '*'  ~* ^  ***J      01 ^  o  r*-
                                                                                                                                                                  ,'?
                                                                                                                                                                  tn
                                                                                                                                                                  ซo
                                     i   i    i    i   i       i   i    i    i   t       i   i   i    i    i       i   i    i   i    t       i   i    i    i   t

                                    AjcStfifuaa     aao*--cJcy     o>cuu3cuo     rfif5rซปcu2     22^1ซ2ปoj
                                                   p*|      \& ^  (vj  ••• —•      *o ft t*~  c\j  o     ^  0  ••• •*•  *5
                                                   ^      — -—      i*ซ \a      —•     co  ^  cu     ~ป  cu  tn     r*i
                      i   i    i   i    >
                      o co  m in  o
                      cu iri  ro At  o
                                    s
                                                                                     i   i    i   i       i    i   i   i    i
                                                                                    •* o>  -3>  —     vn  AI  -*
 as
t^.
  l
CO
  i*5
                                                                                                                                                       ^  fvr 03  i
                                                                                                                                                           * *
                                                                                                                                                     i  O>  OJ P-
                                                                                                                                                     :  rn  cu -
                                            O^5      •* -•  *o  rป *a      pi i^ซ  ro  f^ cu     **!  ^ ^ S  13     ^^^JG^I
                                            Mf*>o      —• *•ป  vu  trป cu      fu cu  —•  o* *•     OJff>;fi^)1O     •*  gj  in *ป  y>
                                            CU !M |=.      _ —.      -* ^-      ซ     u3CU>ซ     -*Cun     OJ     OjrO*-^—*Cxi
                                 I   I   I    I   I   1       I   I    I   I   t        I
                                 JU3G^GBV      OU3  CT> Oj —ป      U3
                                 j  d  o  r^ ซi o^      V^oojtn      V—---diri      -^rปcu-i
                                                                                                       i   i   r    i    i       i    i   i   i    i
                                                                                          mo      tn-ซf*--*o     ino^cn-*
                                                                                                                            co  "*) in  wa  o5

                                i tn  •*  ซ3
   i   t    i
  oj in  o
                                                                                                                                                   d  d d  d  —
                        s      I
                                           I/I  >^3  ^     ITS  m *wi ff*  ^     o  —•  c*j oj  a      cu cu  ut  ^*5      ^  ^o ff*  o>  cu     tf'SJ'^lC**^1
                                           Intriin     Kjcnrnrawl     OCU^"-*'f*i      •* *  ปfl  CU  ^      M  **1 CU  tfl  ^*     ^urnf*1CU-*

                                                                                                          i  I     1      11  a  S
                                                                                                          as3~;ce      S5zS
                                                                                                          i^PSSS      a ss  sr  =
                                                                                                                                                          a  *
                                                                                        4-18

-------
             f   ui  C
                —.  w
                fO  Uป

                S i
 •o
 O)
 o
 u
 ซ3
r*.
OQ
                a  a
                                     oj  oj oj     —  ,0 iu
                                  2
                               11111
                                                       sss3     s  s  a =  g     is s?  g  r g  s
                                                       —•• fu  ซ.xi -ซ•     cu  n  •ป• oj  —^     j  cu —  f-i
                                            Sj     " S?  rT "^  ""o     in oj  ro  .ป• u?      r*4'yi?"*aiifi     '*ป  !
                                            --     M ซj  ซ a  ^-     53^SSj      SiSSiSSin     31
                                     $*J  ~"  s

                                                                        —  vo -^  ro
                                 ปTJ /^     ro



                              l   r   t    :   i

                             QO  •• —•  ITJ O

                             f^  CO y5  u-j cd
                             —  ซa o     n
                                                                                                                      .
                        H S 2 3    ^ S =  S ป     fS  ^ a  3 ป  2     ซ
                                                ™                                     n
                                                  s =     S
              ;         'Xiujm^l     n-inotr,     ปซoซa>ro     r-
-  s ig  ~ s     s  " s  ป si     rf s  ซ ^  -     a" ~  si rf  - ป     *!

                                                                                                                   s
                                                                                               ,    ,   ,   ,
                            "J1^^"^*     t*^";^1^     <\jm.xi p.i-1      inriomai
                                                               ป     *  S g  s a     S  ซ ซ  5 a  a     ป
                                  ^":0:     oป  - ป  o *     u-,mo,_a>     o 'in -v  in o>  m  -    i

                                  ft  - jg     S  * S3  -^ ซj     -  s  S "  =     ซ• in irf  a? o  -    3
                            J Ifl  I *=  2     ^ *"  g  -  •"     - S y  *' •=ป     o  m r-i  2 S  3     5
II
U -t
X 
-------





LU
ฃ
CO
OH
LU
Q.
CO
O-— '

CO 00
CO t->
s: cs
LU 2T
LU
"i
_i a:

fฃ Q.
go.
LU
u. a.
o
CO
co z
LU O
LMM H"M
5co
3E CO
ป— i i— i
• 	 .5"
r™ z
CO LU
*•*
S>3
3D
O _j

ง1
= o
pi^ ""t


ci
3J
0
•ซJ

*
f5
LU
|
ง i
t—






























t_
^>
A
s
o
MI
•
ซ•
*a
t,
u





















"5

5 ง
ซ 5

o
-
"5
^





















ฃ
I
















!
2
i



fs.
>
•*ป
xf

ฃ 5
in
ง >
•
01





U3





ID







ซ*




C")




jffip--p-M5 r-tซ-^"^ci ^^cutn^ซ c- iii *^ — r-~
, i , 	 , , , , , , , ,..,1,1,11, , , . , ,
SSSjSg SSSSS R2SS3S SSCIS SSS35C SSSSS3
8oaas a-a8a , = !,25 8ass!S gSS-5 S8Rja8(
o P- in ^ p-)ซr 52 "* S0'"*'  -^ Oj Oj r*l in •ซ• n m OJ 11  P? •* 5? R ฃJ i^SS^S KoPJS^ ฐ 2 S *" g
•ซ• (U — " —
	 - 	

Jฐali sง5|i sงsp SKgSS 5I5I^-5SงSฃ

1 1 1 1 1 III!) I I I 1 t 1(111 1 1 1 1 1 1 I I 1 1
ggOgjCDgj cj t- V m —" -:^njffirn ;r"jฃt"^J ^"Kjr^? ^ ^ ^ ^ Jg
* IM ^ * •• * — • k ~*
-oggg ss=5S f^gg* iS5*g SaSsS 3=S*g
in" f*J —T —I -Z — —

i i i i i i t i i i i i i i r i i t i i i i i i i i i i i i
soasa as-ss =ซsaH Za R ซ a a ป t s a ,~--".s
goSgg gji^lS i3gs= gsgssg sagas sssss
rn rtj -T

ป0ซ\1<=. O.ftl-0-L ''''o 4 ' J, ' ' ' ' d. ' ' 1 4 ' ' ' '
SS ซ a 2 ^ • 2 ~ SB R 2 "• 5 S = 2 2 55 <" 2 2g = 2S ซ - ซ •
iSill
4-20

-------
                                         ฃ  ง
                                         O  Wl '
                                     ซ  *n  gj o

                                     1 .ฃ  S o-
                                        sa   .

                                        S  i
                                        ,2  S
                          -o
                          0)
                          c
                          o
                          u
                         CD
                                           r  S
  !2S
                                                      in V  r^ rn  in    ซ3  i
                                                                                 vn  o ™~     o  fi  *o *^j' vฃ     0*1- in  cr^  i\j  ^ m
f*?  •<• f*ป  t~- cfi     ojr*ปtfamcu     W  —•  ^  o t?5
                                                                                                                    — -^  — cxi
                                                                                                                              •
                                                                                                                       eu  oj cปj
                                                                                                                       —  — • nj
                                                      -^ ^  vo"    03     r-T -T o>     r-i



                                                      t   i   i   i   i       i   i   r   i   i       i
                                                      1 .   .   .   . l .       ..*..     ,ซ   *  "^  *   *       ....'..     ?
                                                         in >ฃ     r-i     JT     V^     -I


                                                          i   i    i   i       i   i    i   i    i
                                                                                                                   p-  — . -*•  cr> c-
    —  —•     _            —                  vu  ~*








fJJ^S —  S     Sj^^SjCo     Siro"<งi?*S
                                                                                                                                    2     a
   r*i  —     oj     —*    tvi    —ซ         —• in


    i    i   t   t      i   t   i    i   i      (ii
                                                                                                     ซ-
                                                                                                     o  -*
                                                                                                     A4
o> .—


x 
-------
 TABLE 4-8.  ANNUAL INCIDENCE OF CANCER AS MODELED
           BY  EPArs  HUMAN  EXPOSURE MODELa
Model tower
Range of annual
   Incidence
  of cancer,
  cases/yrD
     1
     2
     3
     4
     5
     6
TOTAL
  0.03-0.76
   0.26-7.4
  0.37-10.6
  0.56-15.9
  0.83-23.6
   1.9-54.2
      4-112
 See Appendix B for the methodology of risk
.modeling.
 This range reflects lower- and upper-bound
 emission estimates and the upper-bound unit risk
 factor.
                       4-22

-------
 4.4  REFERENCES FOR CHAPTER 4
  1.  Characteristics of Commercial Buildings 1983.  U.S. Department of
      Energy.   Report No. DOE/EIA-0246(83).  July 1985.
  2.  Telecon:   R.  Gibson, MRI, with B. Taylor, Envirotek, Inc.  April 11,
      1986.   Design cooling loads.
  3.  Telecon:   M.  Upchurch, MRI,  with T.  Ellison,  Newcomb and Company.
      October  2, 1986.   Heat load  design ranges.
  4.  Handbook  of Industrial Water Conditioning.  8th Edition.  Betz
      Laboratories, Trevose, Pennsylvania.   1980.
  5.  D.  Randall, MRI,  to R. Myers, EPA:ISB.   Trip  report:  Crabtree Valley
      Mall,  Raleigh,  North Carolina,  on July  10,  1986.
  6.  C.  Mumrna,  MRI,  to R. Myers,  EPArlSB.  Trip  report:   Duke University,
      Durham, North Carolina,  on July 3,  1986.
  7.  D.  Randall, MRI,  to R. Myers, EPA:ISB.   Trip  report:   Greenbrier
      Mall,  Norfolk,  Virginia,  on  July 15,  1986.
  8.   D.  Randall, MRI,  to R.  Myers, EPA:ISB.   Trip  report:   Omni
      International Hotel,  Norfolk, Virginia,  on July 16,  1986.
  9.   D.  Randall, MRI,  to R.  Myers, EPA:ISB.   Trip  report:   Sovran Bank,
      Norfolk, Virginia,  on  July 15,  1986.
10.   R.  Gibson,  MRI, to  R.  Myers,  EPA:ISB.  Trip report:  North Carolina
      State  University, Raleigh, North  Carolina, on July 22, 1986.
11.   C.  Mumrna, MRI,  to R. Myers,  EPA:ISB.  Trip report:  Wake Medical
      Center, Raleigh,  North Carolina,  on July 8, 1986.
12.   R.  Gibson,  MRI, to  R. Myers,  EPA:ISB.  Trip report:  North Carolina
      State  Department of Administration, Raleigh, North Carolina, on
      July 8, 1986.
13.  M.  Putnam and R. Gibson, MRI, to  R. Myers, EPA:ISB.  Trip report:
     NASA Langley Research Center, Hampton, Virginia, on September 4
      1986.  Draft.
14.  Chemical  Engineers' Handbook.  6th edition.   Perry and Chilton
     eds.  McGraw Hill, New York.   1984.  :
15.  Letter and attachments..  Campbell, R., Union Oil Company  of
     California, .to Farmer, J., EPA:ESED.   July 11, 1986.   Response  to
     Section 114 information request.
16.  Letter and attachments.  Arvidson, P., BASF  Corporation  to
     Farmer, J., EPA:ESED.  July 31, 1986.,  Response  to Section 114
     information request.
                                   4-23

-------
 17.  Letter and attachments.  Mayer, A., Chemical Manufacturers
      Association, to Cuffe, S., EPA:ISB.  September 27, 1986.  Summary of
      CMA member survey on corrosion inhibitors used in process cooling
      towers including average circulating water ppm.

 18.  Letter and attachments.  Evans, R., AMOCO Oil Company, to Myers  R
      EPA:ISB.  June 14, 1985.  Response to Section 114 information
      request.

 19.  Letter and attachments.  Simmons, R., ARCO Petroleum Products
      Company, to Farmer, J., EPA:ESED.  June 3, 1985.   Response to
      Section 114 information request.

 20.  Letter and attachments.  Parker,  F., Chevron U.S.A.,  to Farmer,  J.,
      EPArESED.   May 16 and 31, 1985.  Response to Section  114 information
      request.

 21.  Letter and attachments.  Johnson, J., Exxon Company,  U.S.A.,  to
      Farmer, J., EPArESED.  May 20,  1985.  Response to Section 114
      information request.

 22.  Letter and attachments.  Williams,  J.,  Gulf Oil Products Company,  to
      Farmer, J., EPA:ESED.  May 20 and July  26, 1985.   Response to
      Section 114 information request.

 23.   Letter and attachments.  Kienle,  R., Shell Oil  Company,  to
      Farmer, J., EPA:ESED.  May 22,  1985. Response  to Section 114
      information request.

 24.   Letter and attachments.   Hawes, R.,  Mobil  011  Corporation,  to
      Farmer,  J.,  EPA:ESED.   May 20,  1985. Response  to Section 114
      information request.

 25.   Letter and attachments.   Cox, R., Texaco  U.S.A.,  to Farmer, J.,
      EPA:ESED.   May 24,  1985.   Response to Section  114 information
      request.

 26.   Emission Test Report:   Department of Energy Paducah Gaseous Diffusion
      Plant,  Paducah, Kentucky.   Prepared  for U. S. Environmental
      Protection  Agency,  Research Triangle Park, North  Carolina.
      October  1986.  EMB  Report  S5-CCT-2.

 27.   Emissions  Test Report:  Exxon Company Petroleum Refinery, Baytown,
     Texas.  Prepared for  U. S.  Environmental Protection Agency, Research
     Triangle Park, North  Carolina.  November 1986.  EMB Report 85-CCT-3.

 28.  Telecon:   B. Nicholson, MRI, with I. Kuharic, Marley Cooling
     Towers.  October 15,  1986.  Liquid-to-gas ratios and stack
     velocities.

29.  C. Mumma, MRI, to R. Myers, EPA:ISB.  Trip report:  University of
     North Carolina, Chapel Hill, North Carolina.  July 2,  1986.
                                   4-24

-------
30.  0. Randall, MRI, to R. Myers,  EPA:ISB.  Trip  report:  Old Dominion
     University, Norfolk, Virginia.  July  16,  1986.

31.  D. Randall, MRI, to R. Myers,  EPA:ISB.  Trip  report:  Humana Bayside
     Hospital, Virginia Beach, Virginia.   July 17, 1986.

32.  Telecon.  P. Bellin, MRI, with A. Watkins, North Carolina Department
     of Administration.  June 25, 1986.  Water treatment programs at State
     facilities.

33.  Telecon.  P. Bellin, MRI, with I. Kuharic, Marley Cooling Towers.
     July 5, 1985.  Comfort cooling tower  population.

34.  Telecon.  M. Upchurch, MRI, with J. Carroll, BAC-Pritchard.
     September 3, 1986.  Comfort cooling tower population.

35.  Telecon.  M. Upchurch, MRI, with L. Cornwell, Star Cooling Towers of
     Houston.  September 3, 1986.  Comfort cooling tower population.

36.  Telecon.  C. Green, MRI, with R. Kellogg,  Jr., Dubois Chemical
     Division—CHEMED Corp.  March 29, 1985.  Sale and cost of chemical
     treatments.

37.  Telecon.  C. Green, MRI, with P. Thomas, Drew Chemical Company.
     April 2, 1985.  Use of chromium in comfort cooling towers.

38.  Telecon:  C. Green, MRI, with J. Lee,  Olin Water Services.   April  2,
     1985.  Concentrations of chromate and  nonchromate chemicals.

39.  Memorandum.  R. Myers, EPA:ISB, to R.  Gibson,  MRI.   October  15,
     1986.  Meteorological  data.

40.  Memorandum.  M. Dusetzina,  EPA:PAB,  to Comfort Cooling Tower Project
     Files,   April  20, 1987.   Chromium cooling  tower  risk  assessment
     results.
                                   4-25

-------

-------
                      5.  EMISSION CONTROL TECHNIQUES

 5.1   CONTROL  TECHNIQUES
      Techniques  for  reducing  hexavalent chromium (Cr+s)  emissions  from
 comfort cooling  towers (CCT's)  include  eliminating  chromium-based
 chemicals  from the water  treatment programs  and  retrofitting  towers  with
 higher efficiency drift eliminators.
 5.1.1  Nonchromiurn-Based  Treatment Programs
     Most  nonchromate treatment  formulations contain a combination of
 cathodic and  anodic  inhibitors along with antiscalants and dispersants.
 With all nonchromate inhibitors, monitoring  and  control  are critical.   It
 is necessary  to  control carefully the recommended pH, cycles  of
 concentration, and inhibitor  concentrations.  Most  vendors recommend that
 inhibitor  feed,  blowdown, and pH be controlled automatically.  Also,
 microbiological  control is more difficult with nonchromates because
 phosphates are nutrients  and  some of the other inhibitors are not
 compatible with  chlorine.  The compounds used in the nonchromate programs
 are essentially  the same  as those used  in low-chromate programs described
 in Section 3.4.2.2.  Nonchromate programs have become more successful as
 operators have become more experienced with automatic controllers used to
 regulate inhibitor feed and blowdown and as vendors have developed better
 products, especially antiscalants and dispersants.1  Typical  formulations
used in CCT's are based on phosphates, molybdates, and organics.  These,
 and less common programs,  are discussed below.      '
   •  5.1.1.1  Phosphates.   Combinations of polyphosphates and or^ho-
phosphates can be used alone at concentrations  of 10 to 30 ppm.   However,
 it is more common to add  phosphonate,  polymeric dispersants,  or  both with
the phosphates to reduce  scaling.  With these combinations,  effective
corrosion control can be  achieved.
                                    5-1

-------
      One successful formulation consists of a combination of
 orthophosphate, polyphosphate, phosphonate, and a dispersant.  Virtua-lly
 no calcium carbonate and calcium phosphate scale result if conscientious
 control is maintained at all times and calcium phosphate fouling is
 controlled with the dispersant.  It is important that the water have a
 calcium hardness within the range of about 100 to 600 ppm9 and pH should
 be maintained between about 7 and 8.  When the pH exceeds 8.5, scaling may
 become a problem; if the system experiences a low pH excursion, the
 recovery is much slower than with chromate.  At low pH (<6.5), rapid
 corrosion will  occur.  Chlorination up to 1 ppm is acceptable, as is the
 use of most nonoxiding biocides; but because of the nutrient value  of the
 polyphosphates, additional  biocides are required.   Azoles should be added
 1f the system contains any  copper alloy materials  because phosphate does
 not adequately  protect copper from corrosion.
      5.1.1.2 Molybdates.   Molybdates  are not  a common program primarily
 because of their cost.  However, they  are commonly blended with other
 Inhibitors,  dispersants, and antiscalants in a-treatment  program.   For
 example,  a combination of up to 15 ppm molybdates  with an  azole and
 phosphate can provide better corrosion protection  than is  achieved  by the
 other nonchromate treatment programs.   The  pH  should  be maintained  in the
 range of  7.5 to 8.0  because destruction of  the metal  oxide film occurs at
 lower pH  levels.  At higher pH  levels,  scale is more  difficult  to control,
 and  its presence  will  interfere with film formation and repair.  For  best
 results,  it  is  recommended  that a  passivating pretreatment with molybdate
 be conducted  for  up  to 1 week at two to five times the maintenance
 dosage.   Low levels  of molybdate may permit pitting unless a small  amount
 of zinc is also  included in  the  formulation.  Short-term pH excursions
 reportedly are  not harmful;  but  scaling,  fouling, and biological growth
 must  be controlled carefully because deposition will  interfere with film
 formations.  Automatic  pH control and blowdown are strongly recommended.
 Low water velocity should be avoided because it can cause fouling and
 deposition that damage  the film.
      5.1.1.3  Organics.  A number of all-organic formulations can be used
 as corrosion inhibitors.  Modified lignins and tannins and/or polyamines,
phosphonides, phosphonium compounds, and heterocyclic nitrogen compounds
                                    5-2

-------
 have been used as the primary corrosion inhibitor component.  The most
 common heterocyclic nitrogen compounds are the azoles, which are excellent
 copper corrosion inhibitors.  Azoles protect copper by repairing defects
 (penetration and areas of erosion) that occur in the naturally formed
 protective film of cupric oxide.  They also act synergistically with the
 natural film of calcium carbonate precipitate.  Although azoles protect
 copper, another inhibitor also is required to protect the carbon steel
 heat exchanger shell and the system piping.  Tolyltriazole is one of the
 best of the organic compounds, but it is also the most difficult to
 formulate.  Another effective organic compound is 2-mercaptobenzotriazole,
 but chlorine oxidizes it to the inactive disulfide.   These treatments are
 more costly than chromate,  and the protective films  decay faster than
 those produced by chromate.  The total  organic formulation is required in
 concentrations of 50 to  150 ppm, and  pH should be maintained in the range
 of 7.5 to  8.5.  Microbiological  growth  must be strictly controlled  to
 avoid interference with  film formation.  The alkaline  conditions  that
 organic treatments require  also  decrease the toxicity  of  some biocides
 (i.e.,  chlorine and methylene bisthiocyanate), which results  in greater
 consumption  of biocides.  Effective protection with  little  pitting  can be
 attained if  pH,  alkalinity,  blowdown, and  inhibitor concentrations  are all
 controlled as  recommended.   The  best results  can  be achieved  when the
 system  is  controlled  automatically.
      5.1.1.4  Zinc.   To be most  effective,  zinc must be used  with
 combinations of phosphates,  phosphonates, organics, or polymeric
 dispersants.   When  zinc is used  alone, the  recommended pH can vary between
 6.8 and 8.5  depending on the materials, water quality, and temperature.
 At higher  pH,  zinc  hydroxides are  insoluble.  However, the use of various
 organic phosphorus compounds and polymeric dispersants, and the use of
 automatic  blowdown and pH control  help prevent a fouling problem.
     When  the calcium content of the water  is low, zinc is a beneficial
 addition to  the orthophosphate-phosphate-phosphonate  treatment program,
especially in all-ferrous systems.  If the system contains copper, as do
CCT's, azoles are required.   Excessive orthophosphate,  however, can
precipitate zinc and create  a fouling problem.  An azole-zinc-phosphate
formulation also is effective.
                                    5-3

-------
      5.1.1.5  No Corrosion Inhibitor.   A treatment program without a
 corrosion Inhibitor is  not common.   Only scale and fouling control agents
 are  added,  and pH is controlled (kept  high but varies depending on
 temperature).   The intent is to use calcium carbonate scale to prevent
 corrosion but  also to avoid scale buildup from reaching the point where it
 would hamper heat transfer.  Therefore,  makeup water must have calcium
 hardness, and  the amount  in the recirculating water must be monitored
 closely.   Automatic controllers help prevent problems.   If antiscalants
'and  dispersants are not added,  the  number of cycles of  concentration would
 have to decrease to prevent deposition,  and blowdown would have to be
 increased.   Scaling would still probably occur,  and the maintenance effort
 would increase.
      5.1.1.6  Others.  Nitrites and orthosilicates are  two anodic
 inhibitors  that are often used  in closed systems,  but they are rarely  used
 in open systems.   Nitrites are  required  in very  high concentrations, are
 attacked  by oxidizing agents and certain bacteria, are  toxic to animal
 life,  and are  only about  two-thirds as effective as chromates.
 Orthosilicates also must  be used in high dosages,  are slow to take effect,
 and  are not as effective  as the other nonchrdmate  inhibitors.
      5.1.1.7  Chromate  Versus Ngnchromate Treatment Programs.   Actual
 experience  with CCT's at  NASA's Langley  Research Center indicates that
 equal  or  better corrosion protection is  provided when the  control
 parameters  of  a phosphate-based program  are carefully monitored and
 maintained.  The  NASA facility  has  been  operating  light industrial  and
 CGT's  on  phosphate-based  programs for the past 2 years.   Prior  to this
 period, chromate  programs were  used at concentration levels  in  the range  •
 of 5 to 10  ppm.   The NASA facility  does  not have a target  corrosion  rate
 for  CCT systems,  but the  life expectancy of the  heat exchangers  is  about
 30 years.   Carbon steel corrosion coupons have been  used  to  evaluate the
 effectiveness  of  the phosphate-based program.  Preliminary  information
 indicates that a  corrosion rate of  about 50.8 urn yr  (2  mil/yr)  was
 achieved  in  the 3 to 4  months ending in  February 1986.  More  recent  data
 are  scarce,  but early indications are that  the current  corrosion  rate is
 even slower.   Because the corrosion rate of copper  is about  an  order of
 magnitude less than the corrosion rate of carbon steel, the  copper
                                    5-4

-------
 condenser tubes may be corroding at about 5.1 pm yr (0.2 mil/yr).2
 Because the NASA personnel  are conscientiously attending to the program,
 the corrosion control  results are excellent.   Data from the chromate-based
 treatment are not available,  but the operators do not believe there has
 been an increase in the corrosion rate since  the switch to nonchromates.
 Also,  there has been no change in the preventive maintenance schedule of
 general  maintenance at 1-year intervals;  cleaning,  if necessary,  every
 other  year; and overhaul  every 5 years.
      Nonchromates also have been used successfully  in CCT's when  the
 makeup water is of poor quality (high chlorides content,  high calcium
 concentration,  and/or  high  conductivity).   The total  dissolved solids
 concentration in the Chesapeake, Virginia,  water often is as high  as
 3,000  ppm (conductivity greater than approximately  3,000  ymhos) during the
 summer,  but nonchromates  are  used in CCT's  because  publicly owned
 treatment works discourage  the use  of chromates.3   For example, the  CCT at
 the  Greenbrier  Mall  in Chesapeake,  Virginia,  has  been  treated  with
 nonchromates  for 5 years.   The chemical feed  rate,  blowdown,  and pH  are
 controlled  automatically  to maintain operating  parameters within vendor
 recommended control  ranges.   During  the summer,  however,  the maximum
 recommended conductivity  often is exceeded  even though  the  blowdown  rate
 is maximized.   However, each  winter  the chiller tubes  are inspected,  and
 they typically  are found  to be quite  clean  and the amount of general
 corrosion  is  acceptable/
     At most  sites where chromates have been discontinued in favor of
 nonchromates, quantitative costs and maintenance requirements have not
 been available.   In general,  it  has been claimed that equivalent
 protection from corrosion has  been achieved.  However, some comments have
 been received that mention  increased chemical  cost;  increased monitoring;
 necessity for automatic controllers  (especially when the water quality is
poor); increased number and frequency of analyses; and increased use of
biocides, antiseal ants, and dispersants.
     5.1.1.8  Health Effects.   The amount of information in the scientific
 literature on the health effects of the substitutes  for chromium in CCT's
 is limited for some of the pollutants, especially specific triazole
compounds, polyphosphates, and polyacrylates.   However, the information
                                   5-5

-------
 available Indicates that the health risk is less (and, in most cases, much
 less)  from exposure to the substitutes than to chromium.5
      Hexavalent chromium is a very potent lung carcinogen when the
 exposure route is by inhalation.   The EPA currently believes that a level
 cannot be identified below which  there is no increased risk of cancer.
 Therefore,  any exposure is associated with some finite increased health
 risk.   In addition to cancer, exposure to Cr+6 at concentrations as low as
 about  100 pg/m  (as low as 1 ug/m3 for chromic acid)  has  been associated
 with a number  of adverse respiratory effects that increase in severity
 with increasing exposure duration and concentration.   Many Cr+6 compounds
 are strong  oxidizers,  which explains their toxic properties and their role
 as an  irritant.
     Studies of the substitutes have not  clearly shown any of them to be
 potentially carcinogenic.   Although  the triazoles are mutagenic in several
 assays,  this alone is  not  enough  to  consider the compounds to be
 carcinogenic.   Animal  studies were inconclusive.  One animal  study with
 molybdenum  salts  reported  an increased  number of lung tumors  among
 Strain A mice  exposed  by intraperitoneal  injection.   That  particular  test
 system has  not  been  demonstrated  to  be  predictive of  cancer,  and,
 therefore,  the.results  should be  given  little  weight.
     For noncancer health  effects  of  the  substitutes,  the  triazoles appear
 to be the most toxic.  At  high concentrations, a  number of  toxic effects
 have been reported  including  respiratory  and neurological  symptoms,
 chronic liver morphological  and functional changes, and alteration of
 female reproductive function.  Benzotriazole is considered  to be a severe
 acute respiratory  hazard as defined by the Federal Hazardous Substances
Act because of one study that reported an LCSO value  (concentration
required to kill 50 percent of the test animals) below 2,000 ug/m3
 (1,910 ug/m  for benzotriazole).    Information from a Russian study
suggests a value of 10 ug/m3 as that level below which no adverse health
effects would be observed.  The maximum hourly value of benzotriazole
predicted from CCT emissions is estimated to be in the range of 0.008 to
0.2 ug/m .   Although the risks are difficult to predict accurately, it is
unlikely that anyone exposed to this concentration would suffer
significant adverse health effects.
                                   5-6

-------
     Molybdenum compounds appear to be the next most toxic of  the
alternatives.  The primary health effects of molybdenum compounds  are
acute and respiratory system changes and anemia.  The threshold  limit
value (TLY) for molybdenum compounds is 5 mg/m3, a concentration that
industrial hygienists believe workers may be exposed to continuously for
8 hours per day without adverse health effects.  The maximum hourly value
                                                a value one thousand times
 estimated  for molybdates  from CCT's is 5 yg/nf
 lower than the TLV.6
      There is little  information available on polyacrylic acid;  however
 for acrylic acid  monomer  and  methyl aerylate,  the TLV's are 30 mg/m3 and
 35  mg/m ,  respectively.   The  maximum hourly value estimated for
 polyacrylate from CCT's is  0.5 yg/m3,  about four orders of magnitude
 lower.   Polyphosphates and phosphate  esters  are rapidly hydrolyzed to
 orthophosphates (some are insoluble),  and  toxic  effects in animals  are
 rarely  reported.   Some diphosphonates  that are relatively nontoxic  are
 used  as pharmacologic agents  for diagnosis and treatment of bone
 disease.                                  ;
 5.1.2  High-Efficiency Drift  Eliminators
      5.1.2.1   Design and  Performance of Drift Eliminators.   Historically,
 the purpose of drift reduction  has been to alleviate nuisance  problems
 related to deposition of  drift  on nearby buildings or  on  personal property
 such  as automobiles.  Installation of drift eliminators  is  the most  common
 way to  reduce  drift.  Splash fill towers inherently produce  the most
 drift.
      Drift eliminators are  designed with very low pressure drops compared
 to other water-based air pollution control equipment and rely primarily
 upon  the impaction of water droplets on drift eliminator surfaces to
 reduce  the concentration of drift from the exit air of cooling towers.
 The blades of drift eliminators are configured to require directional
 changes in the airflow such that the momentum of water droplets will carry
 the droplets out of the airflow stream and onto the blade surfaces.   The
major differences  among drift eliminators that affect efficiency  are the
 number of directional  airflow changes,  the; spacing between the blade
 surfaces, the angle of directional change,' and the ability to return the
collected water to a quiescent area of the plenum.
                                   5-7

-------
      Figure 3-3 shows sketches of various types of dri'ft eliminator
 designs.   Lower efficiency drift eliminators include herringbone, some
 waveform  (sinusoidal), and some cellular designs.   Herringbone designs are
 constructed to create two or three major directional changes in the air.
 The blades are sloped in opposing directions in a  manner that provides
 drainage  of the accumulated drift into the fill area.   The blades
 typically are constructed of wood, but other materials  (e.g., metal)  are
 also used.  Waveform designs are configured in a sinusoidal  wave pattern
 such that two major directional  changes in the airflow  are created.   The
 sinusoidal blades  are constructed of asbestos cement board or PVC
 material.   Cellular designs are  configured with thinner blades in a
 honeycomb pattern.   The airflow  directional  changes  are usually greater in
 number  but less abrupt than in herringbone and waveform styles.   Closer
 spacing of the airflow passages  in the cellular design  reduces the
 distance  a droplet  must travel across  the stream to  impact on the
 surface.   Asbestos  cement board  and  PVC are  typically used as the
 construction material  for low-efficiency drift eliminators.   Also,
 drainage  of the collected water  to prevent reentrainment  is  not  a design
 criteria.
      Higher efficiency drift eliminators include some cellular and
 sinusoidal  designs.   The  cellular designs  that achieve  the higher
 efficiencies  use complex  configurations  and  airflow  passages  that contain
 numerous,  closely constructed  airflow  changes.  .Thin materials of
 construction  are used  to  reduce  the  area of  blockage to the airflow and
 minimize the  pressure  drop  that  is created by  the eliminator.  The higher
 efficiency  sinusoidal  designs  are  spaced closer together than  are the
 lower efficiency designs,  and  the  exit  is  configured with  a tip  for
 draining captured water that normally  is partially reentrained in the
 airflow.  Typically, drainage of water  into  a  quiescent area of the tower
 or back into  the fill  is  a major design  consideration of higher efficiency
 drift eliminators.
     For a given tower design, factors that affect  the performance of the
 drift eliminators include airflow velocity through  the eliminator passages
 and the particle size distribution of the drift entering the eliminator.
The pressure drop across the eliminator is a direct result of the velocity
                                    5-8

-------
of  the  airflow  through  the  passages.   The  drift  size distribution  is
affected  by  both  the  size of  droplet  created  in  the  tower and  the
evaporation  rate  that occurs.   The  rate  of evaporation  directly  affects
the droplet  size  reduction  that results  as water evaporates  from the
surface of the  droplet  when H  is traveling through  the fill zone  of  the
tower.  Reentrainment of drift  because of  water  drainage problems
resulting from  the eliminator design  can reduce  performance.   The  shearing
action  of air on  water  draining off a drift eliminator  can result  in  the
formation of droplets that  can  be reentrained.   A drift eliminator design
that  incorporates a structural  modification to direct water  to a quiescent
area  of the  tower can result  in lower drift rates.   The air  velocity
through gaps between  the drift  eliminator  panels and the walls,
distribution deck, other panels, structural members, and the basin will be
greater than the  velocity through the drift eliminator  because the gaps
are areas of decreased  flow resistance.  Thus, the drift emissions will be
greater from towers with gaps than  from  sealed towers.
      Drift is often expressed as the  percentage of the  recirculating water
flow  rate that  is emitted.  A major drift  eliminator manufacturer uses a
drift rate of 0.008 percent to  distinguish  between higher and lower
efficiency drift  eliminators.    This manufacturer further states that
higher  efficiency drift eliminators can  achieve drift rates as low as
0.001 percent of  the  recirculating water flow rate and  lower efficiency
drift eliminators can achieve drift rates,as high as 0.02 percent of the
recirculating water flow rate.7"9  Another way to express drift is  as the
mass of droplets emitted per mass of airflow.   Test results have indicated
that the drift  loss from lower efficiency drift eliminators (herringbone)
averages 200 to 350 ppm and that higher efficiency drift eliminators  can
reduce the drift loss to about 20 ppm.10   The  drift rates presented as a
percentage of the recirculating rate do not necessarily  correlate directly
with the drift rates reported  as a concentration  because the  rates  were
probably determined in tests with different drift eliminators and with
different liquid-to-gas  ratios.
     Drift eliminator manufacturers  indicate that higher efficiency drift
eliminators  can remove up to 80 to 90  percent  of  the  amount of  drift
discharged from lower efficiency drift eliminators.   These drift
                                    5-9

-------
 eliminator efficiencies  are based  on pilot-scale cooling tower tests using
 test methods  that have not  been verified  by EPA.  Emission tests of drift
 eliminators performed by EPA at three facilities indicate that the average
 nationwide control  achieved by  higher efficiency drift eliminators is
 80  percent more  effective than  that  aciieved by lower efficiency drift
 eliminators.   Of the two cooling towers tested  by EPA with higher
 efficiency drift eliminators, the  average emissions  for the tower with the
 highest  average  emissions were  0.3 x 10"3 milligram  of Cr"1"6 per milligram
 of  Cr*6  recirculating (mg Cr+6/mg  Cr+s) (0.3 x  10~3  pound of Cr+6 emitted
 per pound  of  Cr+s recirculating [Ib  Cr+6/lb Cr"1"6]).   As indicated in
 Section  4.1.6, EPA  emission tests  of the  two cooling towers with lower
 efficiency herringbone drift eliminators  provide an  estimate of average
 national emissions  of 1.5 x 10"3 mg  Cr+6/mg Cr"1"6 (1.5 x 10~3 Ib Cr+6/lb
 Cr+6).
      The individual isokinetic  emission test runs  used to determine  the
 average emissions for both  higher  and  lower efficiency drift eliminators
 were highly .variable, which  indicates  uncertainty  in the  calculated
 difference between  the two  average levels  of emissions.   Additional
 testing is planned  to determine  the  cause  of this  variability and  to
 provide additional  data  for  estimating the  incremental  effectiveness  of
 higher efficiency drift  eliminators.
      In addition to isokinetic emission tests to determine  Cr"1"6  emissions
 from four  cooling towers, tests by the sensitive paper  technique were
 conducted  by  EPA to determine the mass emissions and  size  spectra of water
 droplets greater than about  30 urn in diameter.  Tests by  the  absorbent
 paper technique were conducted to determine  the emissions of  Cr+6 in
 droplets greater than about  30 urn in diameter.  The  results  of  these tests
 consistently  indicated that greater than 90 percent of  the  Cr+6 emissions
were contained in drift droplets smaller than 30 pm  in diameter.  It has
 been reported that the collection efficiency of older design drift
eliminators for droplets  or particles less than 50 pm (2.0 mils) is
relatively insignificant.11
     The size of the drift droplets determine, in part, whether the drift
droplets are dispersed in the air or deposited on the ground upon leaving
the cooling tower.  After leaving the exhaust plume of the cooling tower,
                                   5-10

-------
all droplets fall  at their  terminal  settling  velocity.   In  addition,
evaporation of the droplets will  increase  after  they  leave  the  exhaust
plume.   In the atmosphere,  droplets  and  particles  smaller than  about  10  ym
act as a gas.  As  indicated in Table 3-4,  for 80 percent relative  humidity
and 26.7ฐC (80ฐF)  (close to expected worst-case  meteorological
conditions), a 30-ym droplet will evaporate to 15  ym  in about 3 seconds.
By the same procedure, a 30-ym droplet will evaporate to 10 ym  in  about
3.5 seconds.  In half this  time,  it  will evaporate to 22 urn.  In addition,
the settling velocity of a  droplet can be  calculated from Equation  I.12

where,      V = velocity-,- -cm/ s   ~
        .   Pp = density of droplet, 1.0 g/cm3
           Pg = density of air, 1.163 x 10'3 g/cm3
            g = acceleration of gravity, 981 cm/s2
           dp = diameter of droplet, cm
           yg = viscosity of air, 1.983 x 10"" g/cm/s
The settling rate of a 22-wm droplet is about 1.3 cm/s (0.51 in./s); this
also is the average settling rate of a 30-ym droplet evaporating to
10 ym.  Thus, the droplet would fall less than 5 cm in 3 seconds.  This
distance is insignificant compared to the release height of drift from a
CCT (especially if the CCT is mounted on the roof of a building).  Thus,
because EPA-sponsored emission tests indicated that more than 90 percent
of the Cr+  is contained in droplets less than 30 urn, it is reasonable to
assume that more than 90 percent of the Cr+6 emitted from these towers
will remain airborne to be dispersed over a large area.
     5.1.2.2  Retrofit Problems with Drift Eliminators.   Vendors have
indicated  that most existing CCT's have lower efficiency drift
eliminators.  One cooling tower manufacturer with 40 percent of the market
indicates  that all its towers have drift eliminators and its older towers
do not have the higher efficiency drift eliminators that the company uses
in newer installations.  Other manufacturers do  not routinely install
higher efficiency eliminators. I3f L1+   Thus, a large market exists for
retrofits  of high-efficiency drift eliminators.
                                   5-11

-------
      Cooling tower manufacturers have indicated that 50 percent of all
 CCT's are counterflow towers and 50 percent are crossflow towers.
 Manufacturers also have indicated that retrofit can be accomplished by
 replacing the drift eliminator in all counterflow towers ?nd in about
 50  percent of the crossflow towers.  Although it is not possible to
 replace the drift eliminator in the remaining 50 percent of crossflow
 towers because the eliminators are molded integrally with the fill  sheets,
 a second drift eliminator could be installed above the existing units in
 about 90 percent  of these towers.   This procedure would increase the
 electricity requirements  by about  15 percent if the same airflow must be
 maintained.  Also, extensive electrical  and mechanical  modifications
 including installation of a larger motor may be required by 50  to
 75  percent of,the towers  to accommodate the increased electrical  load.
 The remaining 10  percent  of the towers utilizing drift  eliminators  that
 are molded integrally with  the fill  sheets  (2.5 percent of  all  towers)
 would have to be  rebuilt  because they do not have the necessary clearance
 to  have a second  drift eliminator  installed.llfป15
      Rather than  retrofit high-efficiency drift eliminators,  it is  likely
 that  the owners of the towers  that would incur  higher electricity costs
 and possibly  require  extensive modification or  rebuilding (25 percent o.f
 all CCT's)  would  switch to  nonchromate treatment.   Some  towers  already
 have  high-efficiency  drift  eliminators.   However,  because a drift
 eliminator retrofit regulation that  specifies a Cr+6  emission limit has
 not been developed, it  is not  known what percentage of existing towers
would  be required  to  retrofit.   Thus,  it has been  assumed that  all towers
 are affected  and that  for the  37,500  CCT's  using  chromate treatment
programs,  the retrofit control  technique would  consist of drift eliminat?r
replacement for 75 percent  of  the  towers and a  switch to nonchromate
treatment  for 25 percent of the  towers.
     Retrofits on  towers  in systems that are over designed for the actual
heat load  and that have more than one  tower or more than one cell per
tower could be performed any time during the year.  However, to avoid
inconvenience to the user,  systems that are operating at capacity could be
retrofit only when weather conditions do not require tower use.   This time
could be as short as a few weeks in the southern United States.
                                   5-12

-------
      Prior  to  implementation of  a regulation requiring the retrofit of

 higher  efficiency  drift  eliminators,  those drift eliminators would have to

 be certified as  higher efficiency using  a''-'standard  test method to be

 developed by the Agency.   Furthermore, proper installation and sealing are

 critical to the  operation  of any drift eliminator.   Any maintenance

 performed on the tower that  requires  drift eliminator  removal  or

 adjustment also  may result in improper sealing.   Thus,  periodic

 inspections by owners and  enforcement personnel  would  be required to
 ensure  compliance.

 5.2   REFERENCES

 1.   Telecon.  P.  Bell in,  MRI, with A. Watkins,  North  Carolina Department
      of Administration.  June 25,  1986.  Water treatment programs at State
      facilities.

 2.   Simon, D.   Survey of  Nonchromate Cooling Water  Corrosion
      Inhibitors.   Materials  Technology Institute.  1980.

 3.   Telecon.  R.  Gibson,  MRI, with G. Aydlett,  Hampton  Roads  Sanitation
      District, Hampton Roads, Virginia.  July 1,  1986.   Chromium
      discharges  into POTW.

 4.   D. Randall, MRI, to R. Myers, EPA:ISB.  Trip report:  Greenbrier
     Mall, Norfolk, Virginia, on July 15, 1986.

 5.   Review of Health Effects of Substitutes for Chromates in Comfort
      Cooling Water Treatment  Programs.  Environmental Criteria and
     Assessment Office.   U. S. Environmental Protection Agency.
     March 1987.

 6.  Memorandum from K.  Blanchard, EPA:PAB, to W. Victery, EPA:ECAO.
     January 15, 1987.  Exposure estimates for hexavalent chromium from
     comfort cooling towers.               !

 7.  Kelly,  G.  M.  A System-Efficient Approach to Cooling Tower Energy
     Modifications.  Cooling  Tower Institute Technical  Paper No. TP-85-18.
     New Orleans, Louisiana.   January 1985.

 8.  Telecon:   C. Clark,  MRI, with J.  Holmberg,  Marley  Cooling Tower
     Company.   April  2,  1985.  Drift eliminator efficiency.

 9.  Telecon:   P. Bell in, MRI, with J. Holmberg,  Marley Cooling Tower
     Company.   July 19,  1985.  Drift eliminator efficiency.

10.  Baker,  D.   Cooling  Tower Performance.,  New York, Chemical  Publishing
     Company.   1984.
                                   5-13

-------
11.  Foster, P. M., et al.  Droplet Behavior and Collection By Counterflow
     Cooling Tower Eliminators.  Atmospheric Environment.  8:349.  1974.
12.  Handbook of Air Pollution Technology.  Calvert and Englund, eds.
     York.  John Wiley and Sons.  1984.

13.  Telecon:  P. Bell in, MRI, with J. Carroll, Baltimore Aircoil
     Company.  June 5, 1986.  Drift eliminator installations.

14.  Telecon:  P. Bellin, MRI, with V. Bruning, Marley Cooling Tower
     Company.  June 25, 1986.  Drift eliminator installations and
     retrofits.

15.  Telecon.  D. Randall, MRI, with V. Bruning, Marley Cooling Tower
     Company.  November 18, 1986.  Drift eliminator retrofit.
New
                                   5-14

-------
                    6.  REGULATORY CONTROL ALTERNATIVES

      This  section  discusses  the  regulatory control  alternatives available
 for  reducing  or  eliminating  the  risks  from hexavalent chromium (Cr+s)
 emissions  from comfort cooling towers  (CCT's).   The alternatives available
 include  eliminating chromium-based treatment  programs and  retrofitting
 CCT's with high-efficiency drift eliminators.
      Prohibiting the use of  Cr+6-based water  treatment chemicals in. CCT's
 has  been shown to  be a feasible  control technique.   Alternative treatment
 programs are  as effective as chromium at  inhibiting corrosion  if used  as
 recommended and are readily  available.  The majority of CCT's  use
 nonchromium chemicals and a  few  CCT's use no corrosion inhibitors (about
 85 percent).  As discussed in Section 5.1.1.7, present information
 indicates  that these systems do  not experience significant adverse  impacts
 from  corrosion.  Also, cooling tower and water treatment vendors  have
 stated that switching from chromium to nonchromium  corrosion inhibitors is
 not expected  to cause any operation or maintenance  problems with  the
 chiller  components (heat exchanger) of the cooling  system.  Automated
 monitoring and control systems are readily available  to control chemical
 feed  (including corrosion inhibitor, biocides, antiscalants, and
 dispersants), makeup, and blowdown.  in addition, this option results  in
 the maximum level of environmental  protection because Cr+6 emissions are
 completely eliminated.
     Retrofitting existing CCT's with higher efficiency drift eliminators
 is not possible at all sites  and would be much more complex than  switching
 to nonchromates at most of the remaining sites.   As discussed in
Section 5.1.2, retrofitting a CCT with a high-efficiency drift eliminator
would achieve at most an 80 percent reduction in  airborne  emissions of
Cr+  from the CCT.   However,  because 25 percent of existing CCT's would be
                                    6-1

-------
 expected  to  switch  to  nonchromate  water treatment programs,  a high-
 efficiency drift  eliminator  retrofit  regulatory control  alternative would
 reduce  the nationwide  emissions  of Cr+6 by up  to 85  percent  below
 baseline.
     At the  present time,  there  is no standard high-efficiency drift
 eliminator design;  at  least  three  manufacturers produce  what they claim
 are  higher efficiency  drift  eliminators,  but the designs and physical
 parameters of  each  are different.   In addition,  it is unclear how emission
 rates'or  efficiencies  of these drift  eliminators are related to  their
 design.   Therefore,  the information is  not available currently to specify
 design  requirements  for a  high-efficiency drift eliminator.
     To implement an effective standard based  on the use of  high-
 efficiency drift eliminators, some combination of equipment  specifica-
 tions,  performance testing,  random inspection,  and/or product
 certification would  be necessary in addition to  an annual inspection of
 the  drift  eliminator.  The Agency  also  would need to develop  a reference
 test method, develop a standard set of  test conditions for certification
 testing related to actual operating conditions  of a variety of types of
 cooling towers, and  identify critical installation criteria to be
 specified  in the installation instructions of  the drift  eliminator.  It is
 unlikely that one set  of installation criteria could be  developed to
 include all possible situations.  All installations (250,000 CCT's) would
 have to indicate the type of chemical treatment  they use, and the sources
 using chromates (about 37,500) would be subject  to the testing and
 inspection process.  Thus, prohibiting the use of chromium would make
 enforcement more manageable than implementation  and enforcement of a
 standard based on installation of higher efficiency drift eliminators.
     Standards that prohibit chromium chemical  use in CCT's  could be
 implemented under the Toxic Substances Control  Act (TSCA) or the Clean  Air
Act  (CAA).  Section 6 of TSCA would provide the authority not only to
prohibit the use of chromium-based  chemicals in CCT's but also to prohibit
the  sales of such chemicals for use in CCT's.   Under  Section 112 of the
CAA,  it would not be possible to prohibit the  sales of  a chemical.
However, it would be possible to prohibit the  use of  chromium-based
chemicals in CCT's by setting a zero emissions  limit  under the CAA.  While
                                   6-2

-------
TSCA and the CAA could require equally stringent control levels, TSCA
would provide a better mechanism for enforcement.  Prohibiting the
distribution or sales of chromium-based chemicals for use in CCT's and
monitoring chemical treatment vendor (20+ major and approximately
400 total) sales and recordkeeping would be far less difficult for both
regulatory agencies and sources than inspecting and permitting
37,500 individual users.  Any complaint at a particular site could be
traced to the vendor for determination of compliance.
                                   6-3

-------

-------
                                  7.   COSTS

 7.1  COSTS OF EMISSION CONTROL TECHNIQUES
      This section presents the costs of the regulatory alternatives for
 controlling hexavalent chromium (Cr+6) emissions from comfort cooling
 towers  (CCT's).   The first option is to use nonchromiurn-based chemicals in
 CCT's,  and the second option is to install higher efficiency drift
 eliminators.
 7.1.1   Nonchromiurn-Based  Treatment Programs
      Prohibiting  the use  of chromates in CCT's  would require a switch to
 other more costly corrosion control  chemicals such as phosphates  and
 molybdates.   Vendors have indicated  that chromium-based water treatment
 programs  typically cost between $88  and  $132/M  kg ($40 and  $60/M  Ib)  of
 blowdown.   It has been assumed  that  the  higher  unit cost  is representative
 of  the  baseline cost to CCT operators because they purchase smaller
 quantities of chemicals than industrial  users.  Vendors also  estimated
 that alternative  treatment  programs  (without always  specifying  the type  of
 chemical)  cost from  10 to 360 percent more  than chromium-based
 programs.1'7   Ranges  of costs were provided because water quality and the
 size of the CCT affect the  cost of the treatment  program.   Typically, the
 smaller the CCT,  the  smaller the difference in total cost because the cost
 of vendor  service  is  greater than the chemical  cost.  Also, the cost
 difference will be less for systems with good water quality.  The
 phosphate-based programs typically cost  10 to 100 percent more than
 chromium-based treatments, and molybdate-based  treatments are the most
 costly alternative at up to 360 percent over the cost of  chromium-based
 treatments.  Phosphate-based and molybdate-based chemical  treatments  that
cost 100 percent and 360 percent more than baseline chromate-based
treatment, respectively,  were selected to ensure that the economic impacts
                                    7-1

-------
 were not underestimated.  Thus, phosphate-based and molybdate-based
 chemical treatment unit costs are $264 and $608/M kg ($120 and $276/M 15)
 of  blowdown,  respectively.  The cost of service has not been included in
 the unit cost because it has been assumed that the service requirements
 are similar for a 'jiven CCT regardless of the type of chemical treatment
 program being used.   Cost data from several  operating CCT's that are
 larger  than 400 tons  vary but the average cost is higher than the selected
 unit cost. ~     Also, vendor-estimated costs of organic- and molybdate-
 based treatments for  a given set of CCT operating conditions appear to be
 higher  than the selected unit costs.11  However, both sets of cost data
 include the cost of service.
      To obtain the best results with alternative chemical  programs, the
 chemical  feed  and blowdown must be controlled automatically based on
 conductivity of the water, and pH must be controlled  within narrow ranges
 that  are typically recommended by vendors.   The purchase and installation
 cost  of these  automatic feed  and monitoring  systems would  be about
                1 9
 $500  per tower.-   Although many of the CCT's using chromium-based
 programs  may have  automatic feed systems,  all  towers  currently using
 chromate  are assumed  to need  these systems if they were  switched  to an
 alternative treatment.   This  assumption was  made so that costs would not
 be  underestimated.
      The  percentages  of the CCT market  that  use  the various  alternative
 treatment chemicals are not known,  but  phosphate  programs  are the  most
 prevalent and molybdate use is  rare.   It  is  expected  that  the CCT  systems
 using chromates will  switch to  these two alternatives in the same  relative
 proportion.  Thus, phosphate-based  treatment  programs, which can be used
 in  almost all  situations and  are the least costly alternative, were used
 to  develop the  individual  and nationwide annualized costs  incurred  by
 switching to nonchromiurn-based  chemical treatments.
     The capital cost of automatic feed and monitoring equipment was
 annualized over 15 years, which  is the estimated useful  life of the : jed
equipment.  An  interest rate of  10 percent, which represents the rate of
return considered reasonable for industry to expect in the current
economy, was  used in the calculation of the annualized cost.  The total
annualized cost of phosphate-based treatment was estimated by summing the
                                    7-2

-------
 annualized cost of the feed equipment and the annual incremental cost of
 phosphate-based treatment over chromate-based treatment;  Table 7-1
 presents the total chemical costs and the cost effectiveness of
 eliminating chromium use in CCT's.  The approximately 37,500 CCT's using
 chromium-based treatment would incur the costs of switching to
 nonchromiurn-based treatment, and the total annualized cost would be about
 $9.4 million.  Based on this cost and the reduction in the incidence of
 lung cancer to zero, the cost effectiveness of eliminating chromium-based
 treatment programs ranges from $0.083 million to $2.4 million per cancer
 case avoided.
 7.1.2  Higher Efficiency Drift Eliminators
      The second control technique is to retrofit all chromate-using CCT's
 with higher efficiency drift eliminators.  As described in Section 5.1.2,
 approximately 75 percent of existing CCT's can be retrofitted.   The
 remaining 25 percent of the towers would require extensive modifications
 or need to be totally rebuilt to accommodate a higher efficiency drift
 eliminator.  Detailed estimates of the increased electrical  and structural
 costs have not been presented because it has been assumed  that  rather than
 incur this prohibitive cost, this 25 percent of  all  CCT's  would likely
 switch to nonchromiurn treatment programs.
      Included in the cost of the control  technique are  the costs of the
 drift eliminator retrofits,  annual  inspections,  and  the drift eliminator
 certification program.   The  drift eliminator certification process  was
 estimated to cost industry about $2.3 million, and  it was  assumed  that the
 total  cost would be  passed on to the CCT  owner/operators that use
 chromates.     Both  the  certification and  retrofit  costs were annualized
 over 15 years,  and  the  interest rate used  in the  calculation was
 10  percent.
      The Cr+  emissions  obtained  with'low-efficiency drift eliminators are
 the  baseline emissions.   As  mentioned  in  Section  5.1.2,  it  is assumed that
 drift  from  CCT's  can  be  reduced by up  to 80  percent by  installing high-
 efficiency  drift  eliminators.
      Table  7-2  presents  the  annualized costs  and the cost effectiveness  of
 the  drift  eliminator  retrofit  control  technique for individual model tower
•units  and "nationwide.  The nationwide  annualized costs of the control
                                    7-3

-------
technique are about  $22 million.   The  cancer  incidence  is  reduced  by  3.4
to 96 cases per year,  and  the  nationwide  cost-effectiveness  ranges from
$0.23 million to $6.5  million  per  cancer  case  avoided.
7.2  FIVE-YEAR SALES AND COSTS PROJECTIONS
     Comfort cooling tower manufacturers  have  estimated that about
20,800 CCT's (including evaporative condensers and fluid coolers)  will be
sold in the next 5 years for use in new applications.llfป1S  Sales  in  the
first 2 years will be  slower because of the present building slowdown in
many parts of the nation.  In  the  following 3 years,  sales growth  will
increase gradually.  Table 7-3 presents the total annual sales, and
Table 7-4 presents the annualized  costs for both control techniques.  The
cost to switch 15 percent of the towers to phosphate-based treatment would
cost about $780,000, and requiring high-efficiency drift eliminator
retrofits would cost about $1.4 million.
                                   7-4

-------













rd

0
ป-H
|—
Q_
O
IE
=3
1—4
O
0ฃ
a:
cj
0

Of
o
u_


LU

til
UJ
S
t>
LU
U.
LU-
LU

H™
O
^j



%
t^_
^^

o
ฐ
•
*™1
1


LU
— 1
CO
r—



















j_ 41 4)
"ฐ UJ •*- •*" "-
41 S ro ro 4)
S 1 S .C S 3
— ffl ฐ Q. 5 O
— I. in U 4-
10 o _ O -C v,
3 C .ฃ .C 0 (.
C — "J Q. >.
< 1* 4>ซ
Eฐ2



-•ป•••ฃ 1
"ฐ S -K
in i>ป ^^
9 S >•
s-t^


4) „_
•t "S 4-" o c
2 *o e oi x

_ +; jr — 5
i ^iPlI
"o4" **



4> * *-
^ ^_ 4- 4>
ztjl
o o ^ ^~

tt-fcwป




ฐ ซ 4-" 0 c
, ^_ ^^ ^^ ™
(A 2 dJ -i^t ^^ Q
8 *^' fa ^* *"* ^
^_ S 2^2 2
"~ "o. *- M -ฐ
C 4* ^^
^3
u
*^ o -2


_, c "^
s| ^
^^ U ^*
*^




^ ^;
C *^
X ^
o —

X
O u
CD \
05





tf) (Q tl O ป
O S QJ Q

ฑ | ซ s 0
^ 03 ^~ ~—

O +"



i

4> 4)
0 0
^ 4—


r* *o m m
— to co in





00 vO — fO
— ro r*






^O ^O tQ \rt
f>4 04 04 01

O O O O
^3 *O ^O ^O






S 2? ฐ ฐ
fO r-^ r* —








S 04 ฐ 04
^ *™ ป~ *"

*o *o ^o ^o
Ol O4 O4 O4





rป to in in
— to oo in





ง000
o o o
co 
•— in

in in
O "
o o
O\ o
- r-
•— in
ro \o
o
Ol •
r-


0 0
to to









in to


jr
•*-
^
-Q L.
at >
0 < 0
> S 4-
4) t- 4)
in O) E
O 4-
i_i- ro
O Q. 4>
'4- 4-
4~ C
O E 4)
04- tn
ro ro
, <- 1

^3 4—
3 — ro
U U Q.
c *— in
"'I ?
O JZ
+m F- O.
g ซ r
•*• "o
in o
41 ' 4-
€4) in
Q. O
>- • U
•o 4- in
C I. —
ro j: ro ro
U 4> 4-
4- ro >. c
in 4) o
O in E
O t- — 4)
0 (-
— ••- l_ O
ro 4) c
'i 1 ฐ —
4) in •Q ro
ฃ 4) 3
U 4) N C
-C — e
ro ro
3 4> 3 4)
C C. C J=
c ro c 4-
10 ro
in -o
41 4" in C
f in •— ro
4- 0
O 4- 4-
L. C C
— Q. Q.
in > — —
— U 3 3
4* cr cr
in in 4> dj
ro — "O -o
L. ro 4) 41
Ol O 4) 4>
O — x- *.
i. e
OLฃ u u
4- 4) -1- 4-
c 4- ro ro
4) E E
^ ^ O O
ro •(- 3 3
4> ro ro
U 4-
4- ro l_ •*-
COO
4) +- —
SI -t-
4- 13 O in
4) O O
0 Ii2 "
U) ^3 •*—
-C Ul •*- 4) C
O IO O N 4)
ro — S
 4)  < O o
^3 U
7-5

-------



















,-.
cu
3
C
4J
O
u
*""*
•
ซ— 4
1
r-.
y
ซ-I
CQ
S
i^



.













O m
••• c
in in
งin

B
> V
4--0
O 1
t_
•I™ O
in o.
O in
tfป1n
o in
g% K
| i
••-•o
u c
ซ p
H- O
ซ•• A
* J
"In *
5 ^





S
i






•o •

ID
3
C
C -
J0
"3
ฃJ
p^


H

•o
^— IO O
IO 4"
C IO O
•< o u

i
7






l_
S
=
*

1




t.
a
u
u
X
f^




u
c
ง
i



ง
*t
—
=i
—
~
3
3
U>
2
g
c
a
n
i


0
K
/l
3
U
J
3
3
a

)
i
fl
•* M
"c;P


*^

o|
10
fl
** J>
%o
--2




Jjj
Ss
5 ง
IO




c u
o ^*
u o
3 in
•0 IO
o o
L.

•ง
"5
0
15
z

*- >A
o."1
$ o
u
u
o
1
t
^
ft
13 ^"
!<"

i_
CU
li
in m ซr *r r~ 10
— oo in v 10 to
• O O O O O
ป* o o o o o
f^.

• g> 
^ CM •- •- •— O

ซ~ t*ป o 10 in CM
P" IO ซ IO — O
•*  f^ ^*% JK
o o o o o
Sin in 10 co r-
— to oo — 10
r*ซ o* ^ป *s* *- oo
rป ซ• to CM CM —



ซ *r  ซ CM
• r-ป o in 10 *r
O i — — CM in
1 
-------




1 —

LU
o
a:

^~"
UJ

0

T- {jg
eg 2;
ป— 1 1— i
ac oo
M_
Qฃ 1
21
hL. ซ••••?
I-H
V) O
LU Qฃ
2: a:
LU <_>

H-I 2;
CJ) ^"^
LU
U.

LU
1—


O

a
z


h-
S
Nซ^


CM
1

LLJ
	 i
CO
•
w Hf ^^
— ฐ*ป


4)

fH

ฃ "5 o
-H
L?




•>
_ y**>

S **• 1^ o
O  ปr co M ^
— V CM CM — — ซT








JJ in O in in col to
oป <3> to CM f pป a
O ro r- rซ in lo| .-
CM o in ^3* rซi CM! co
1 CM












— — — — _ _









888888
ir\ tr\ m in trt ir\



o o o o o o o
ง*O OO CD f*t *"~ ^*
QO vo xo •— IA  <\n oo
1 CN



i


in r^ ^ oo *— m







O in o in m o
CTV in GO co co oo
fO in co t^\ ro ^h
— rsi ^







4-
c
H)
o> u
JD L.
CO. 4-
u in u
CM C E
-o .— +.+-
C 4- to x-
4) IO 3 4} •
c io in 4- o
OS t- O 10 c

IO L. 2 O. O — "
.CO t- O
T3 O 4- -4- ฃ —
0) O u **-
4- a ซ c -*.
3 in ^ 4- o 0>
.O 3 4- in SI
— On
t- 4- "O O O O)
4- IO — 4- —
in -c — -o .c
— 4- 3 0) ฃ
w -Q ^ O ^
-^ m — 4-4-
in 4- ซ- — — —
— c 10 as
IB c 3 in
4- E IO C 13
tn 4- jz c -o ซ
O 10 4- ID — 4-
0 ID 34-
U t- TJ O —
in 4- ID c 3-ซ-
— -C. (O O
ซC T3 4— tn ^
H- O IO ป - 4-
ป> L. C )— 4)


4- 4) CO IO IO O)^
in 4- vป e o c
111 IO O *~ •— X)
4- s tn L. ป- 4. —
4- t- '" O ฃ ~ O
0 J= 1- C t- X 3
3 OO O I1J (U
xi u c o m
งO) I]} •
C -C O T) n h-
O — O 4- C 4- O
tn 10 (o o
O 3 H) JC ซ.
4- O C O *•
•o _w 2 *" ฐ ฐ
C H- S '-H C 4-
10 o 4- tn o ID c
O in a) o 4)
>. p — Q. U 0
4- g O — in IB 1_
— ปJB ~ in Q'
o r- t— t_ ••- CM in
10 fo o O rป
**- 4- • 4-
• IB • IO *•ป 4- ID 4-
x: O 4- t. — CO 4-.C
<31_ O E H> 0 4-
— 4- •*- 4- — E C
IปIOOIO— 4-TJ OT3
i_ o c — c*
^Z'c'i^u— a^o
C4-O — — IO 3O4-

4)4> T3<1)C tn4-3
• o 0.4- tn c 3>- — IO-4- C4-4)
^ซ_g OO— SO Oco
— ,T3
jC — 3 yl u (U
t_COU) OlB^oD* —
OOO — ^J)UiO>^z
•4- — o ฃ +- - Oc^UO

•1 	 ODJ=2OOJ '-^=
tn •— O ซC — T3l_
4~W*1'O 4-4-1 C-4-C O
tn tn s — in u — s -o
EOOSJ=Oa>ol_ —
— in a. 4- in sc-oo —
h-ซiiOL.<;Ot/)CQiCL(B
0 .O O T3 IB
7-7

-------




















T3
CD
c
o
u
^ป^

*
CM
1
!*ป.
LU
5






















*4
V
u
t
c
a
:
L
a
ซ*•
**•
c
+•
1
ฃ
•*•
U]
M
a
c
a
+•
c
c
+-
ฃ











"a
ฃ


e
cu
j)
C (.
X
aj o
•o •ป- +-
 in
• c
O
i in
i in

' a>
• •o
) C
i 3
1
O
) 0.
Q.
in
ง
—
.
a> +• X
a o •
O -0 2
c a>
_ I.

e 10
-ST"!
a 0 >-
g o
o S 5
ป .2.2
I- v^o
= t- <0
Q. Uป •(-
a> o 10
L. 0 C

 .tl o

"~


\ u
>> X
\ o
- u
3 a
3 f •
f in — to to to
Oป T CM •— O3 ซป
• • • • O C
o o o o •
— -~ ^ -^ o O
T o T in
o in CN — — r-
• • • • o* T
— O O O O O
o o'

ง00000
o o o o c
o o o o c
 1 — O
o\ to CN r^ป co (j*
to — o* tn *— 03




fO CN — — CN —
oo o in 
-------
    TABLE 7-3.   5-YEAR SALES PROJECTION OF CCT's FOR NEW INSTALLATIONS
                                                                       13
                  Year
Comfort cooling
tower sales, No.
                   1987

                   1988

                   1989
                   1990

                   1991
      3,700

      3,400

      4,300

      4,600

      4,800
   TABLE 7-4.  ANNUALIZED COSTS OF NONCHROMATE AND HIGH-EFFICIENCY DRIFT
                       ELIMINATOR CONTROL TECHNIQUES

Model
tower
No.
1
2
3
4
5
6


No. of
CCT'sa
1,540
6,930
4,240
3,500
2,620
1,980
20,800

No. usinq
chromates
230
1,040
640
520
390
300
3,120

Annual i zed cost of
phosphate treatment,
$/yr/nat i onwi dec
18,900
106,000
95,500
115,300
145,900
295,. '00
776,900
Annual i zed
cost of inspection,
and certification,
$/yr/nationwided
93,200
424,800
267,900
229,800
187,200
187,400
1,390,000 '
aAssumes that towers built in the future will  be distributed in the same
 proportion as the existing towers.
^Assumes 15 percent use chromates.
^Assumes the same annualized cost per tower presented in Table 7-1.
 Assumes 75 percent of the towers are initially constructed with HEDE's
 and 25 percent are switched to nonchromates.   Also assumes the same
 annualized inspection and certification costs per tower that are
 presented in Table 7-2.
                                   7-9

-------
 7.3  REFERENCES FOR CHAPTER 7

  1.  Telecon:   R.  Hathaway, MRI, with G. Sexton, Betz Laboratories.
      November  8, 1984.   Costs of cooling tower water treatment
      chemicals.

  2.  Telecon:   C.  Green,  MRI, with J. Nygren,  Chemtreat,  Inc.   March 14,
      1985.   Cost of cooling tower water treatment chemicals.

  3.  Telecon:   C.  Green,  MRI, with R. Kellogg,  Jr.,  Dubois  Chemical
      Division—CHEMED Corp.  March 29,  1985.   Cost of cooling  tower  water
      treatment chemicals.

  4.   Telecon:   C.  Green,  MRI, with R. Holtkamp,  Hercules, Inc.  April  2,
      1985.   Cost of cooling tower water treatment chemicals.

  5.   Telecon:   R.  Gibson, MRI, with B.  Dowdle, Mogul  Division of Dexter
      Corp.   July 30, 1986.   Cost of cooling tower water treatment
      chemicals.

  6.   Telecon:   R.  Gibson, MRI, with D.  Pratt, Betz Laboratories.
      October 28, 1986.  Percentage cost increase  of nonchromate treatment
      programs.

  7.   Telecon:   J.  Glanville,  MRI,  with  F. Foster,  Unichem International.
      April 4,  1985.  Cost of  cooling  tower water  treatment chemicals.

  8.   D. Randall, MRI, to R. Myers,  EPA:ISB.  Trip  report:  Greenbrier
      Mall, Norfolk, Virginia, on July 15, 1986.

  9.   Telecon:  M.  Upchurch, MRI, with ARCHEM, Inc.  July 29, 1986.   Cost
      of cooling tower water treatment chemicals.

10.   R. Gibson, MRI, to R. Myers,  EPA:ISB.  Trip report:   North Carolina
     State University, Raleigh, North Carolina, on July 22,  1986.

11.   Letter and attachments.  Water treatment vendor, to  Gibson, R.,
     MRI.  October 9, 1986.  Cost of cooling tower water  treatment
     chemicals.

12.  R. Gibson, MRI, to R. Myers, EPA:ISB.  Trip report:   North  Carolina
     State University, Raleigh, North Carolina,  on June 17,  1986.

13.  Memorandum from M.  Putnam, MRI, to R. Myers, EPA:ISB.   December  12,
     1986.  Drift'eliminator certification cost  estimate.

14.  Telecon:  D. Randall, MRI, with B.  C-ites,  Baltimore Aircoil.
     November 7, 1986.   Projection of  total  comfort cooling  tower sales
     for 1987 through 1991.                               *
                                   7-10

-------
15.  Letter.  Holmberg, J. D., The Marley Cooling Tower Company, to
     Crowder, J., EPA:ESED.  February 24* 1987.  Comment on draft
     Regulatory Impacts Analysis.
                                  7-11

-------

-------
                            8.   ECONOMIC IMPACTS

 8.1   SUMMARY
      This chapter discusses  the economic consequences  to  suppliers  of
 chromate corrosion  inhibitors and commercial building  owners  and  tenants
 of controlling hexavalent chromium  (Cr+s) emissions  from  comfort  cooling
 towers  (CCT's).
      Estimates of the numbers of buildings by size and type that  use
 chromates in CCT's  also are  presented along with estimates of the
 incremental costs to replace chromates with alternative corrosion
 inhibitors or to install high-efficiency drift eliminators (HEDE's).
 Control costs are expressed  per unit area to evaluate potential building
 rent  increases and  are multiplied by the numbers of  buildings by  type and
 size  to estimate the total nationwide cost of control.
      This analysis  has determined that the economic  consequences  of
 eliminating Cr+ -based chemical  use in CCT'S are insignificant because-
 rent  increases for  affected buildings would amount to less than
 0.1 percent, and total nationwide annualized costs would not exceed
 $9.4 million.  The economic consequences of retrofitting CCT's with HEDE's
 also  are insignificant because rent increases for affected buildings would
 amount to less than 1.0 percent  and the total nationwide annualized costs
would be about $22 million.
8.2   INDUSTRY PROFILE
8.2.1  Size of Chromium-Based Corrosion Inhibitor Market
     Sodium dichromate is  the major chromium product  used  to  control
corrosion in CCT's;  however,  only  7 percent  of  all  sodium  dichromate
consumed in the U.S. is estimated  to be used  in  corrosion  control
applications,  and less than  5 percent of that amount  is used  in
CCT's. ป   Thus,  it  is estimated that less than  0.35  percent  of  the  total
sodium dichromate consumed  in the  U.S.  is used  in CCT's.

                                    8-1

-------
      The total  annual  value of corrosion inhibitor chemicals used in CCT
 applications  is about  $85.1 million.   Of this,  chromate-based products
 account for approximately  $6.9 million;  nonchromate substitutes account
 for the remaining  $78.2 million.   These  estimates  have been developed
 based on chromate  and  nonchromate  corrosion  inhibitor  costs and the
 cooling tower population data  presented  in Chapters 4  and  7,  and the
 calculations  used  to derive the annual cost  estimates  are  presented in
 Appendix E.   The major suppliers of corrosion inhibitors for  use in CCT's,
 light industry,  and utilities  and  average market share each supplier holds
 are presented in Table 8-1.
 8.2.2  Distribution of Cooling  Towers  by Building  Size
      As discussed  in Section 4.1.1., a Department  of Energy (DOE)  survey
 distributed total  U.S. commercial  building floorspace  into  categories
 based on size.   The category of 465 square meter (m2)  (5,000  square  foot
    2
 [ft 1)  or less  has been excluded from this analysis because industry
 sources  indicate it is unlikely that buildings that size would  use  cooling
 towers  as part of  central cooling  systems.
      Using the DOE data, the number of buildings in each size class  having
 a central cooling'system and the percentage believed to have CCT's were
 estimated.  These estimates are presented in Section 4.3.   Finally,
 assuming  that 15 percent of all CCT's are chromate-using and that
 chromate-using towers are distributed evenly among all  sizes and building
 types as  classified by DOE, the number of buildings with CCT's that use
 chromates was estimated.   The  total  number of buildings, the number of
 buildings with central  cooling  systems, the number of buildings with
 CCT's, and the number of CCT's using chromates are presented in Tables 8-2
 to 8-5, respectively.
 8.2.3  Commercial Building Rental  Rates
      If building owners and operators  were required to  use  alternative
corrosion inhibitors,  the operating costs for any building  currently using
chromates are  likely to increase.   To  estimate the  significance of such
cost  increases,  the economic analysis  in  the  following  sections compares
the costs for  two control  options  to average  annual rents for the affected
buildings on  a unit area basis.
                                   8-2

-------
      Data collected  by the Urban Land Institute indicate that the national
 average  rental  fate  for commercial  buildings including all  utilities and
 taxes is approximately $129/m2/yr ($12/ft2/yr).6  Although  rates vary
 according to  building  type and  location,  the lowest rates observed are
 those for large department stores.   The  national average rental  rate for
 such  stores is  slightly over  $22/m2  ($2/ft2).   At the other end  of the
 scale, rental rates  for fast  food restaurants  and similar facilities run
 close to $215/m2  ($20/ft2).   The wide range  of rates is  attributable to
 the level  of  facilities required for the  rented space.   For example,
 restaurants and similar facilities require services such as plumbing,
 other utilities,  and solid waste disposal; department stores are
 essentially large open spaces that require relatively low levels of  such
 services.
 8.3   CONTROL  COSTS
      The costs  to control  Cr+6 emissions  from  CCT's are  presented for  two
 control  alternatives:   (1) completely  prohibiting  the use of chromates,
 thus  requiring  a  shift  to  more costly  nonchromates,  and  (2)  requiring  the
 installation of HEDE's.  The costs of  each of  these alternatives  are
 presented  by building  size and type  and were developed based on Table  4-4
 and the  DOE survey of  buildings  by size and type.
      Costs for  nonchromate substitutes were presented in  Table 7-1.  These
 costs are  based on a phosphate corrosion  inhibitor  system, which  is  the
 most  prevalent  substitute  for chromates.  Table  7-1  also  shows the annual
 cost of  using chromates and phosphates in individual  towers  as well  as the
 incremental cost of switching from chromates to  phosphates.    In addition
 to the incremental costs, CCT users may also need to purchase an  automatic
 feed  system that controls chemical feed and blowdown.  The purchase  and
 installation costs of such systems also are shown in Table 7-1.  These
 costs were annualized using a capital recovery factor that assumes a
 15-year  life,  which is the estimated useful  life of the feed equipment,
 and a 10-percent  interest rate.   Note that the costs presented in
Table 7-1 represent purchase costs for the chromate corrosion inhibitors
and feed  system and do not include the costs  of technical services
required  to ensure the proper operation of the CCT's.  However, the costs
of such technical services are believed to.be largely unaffected  by the
choice of corrosion inhibitors.

                                   8-3

-------
      The  capital  and  annualized  costs  to retrofit drift eliminators were
 presented in  Table 7-2  and  they  ranged from $390 to $4,980 and from $51 to
 $655,  respectively.   Capital  costs  are expressed on an annual  basis
 through a capital  recovery  factor that approximates annual  depreciation
 over  a 15-year  life and 10-percent  interest rate.   In  addition to the cost
 to retrofit the drift eliminators,  CCT operators will  incur costs to have
 the towers inspected  and  certified.  These  costs are presented in
 Table  7-2.  Inspection  costs  are estimated  to  be $500  per year per tower,
 and certification  costs are estimated  to be $2.3 million industrywide.
 The certification  costs are assumed to be incurred  once during the life of
 each tower and  are, therefore, annualized by use of a  capital  recovery
 factor.   The  annualized certification  cost  distributed  over all  chromate-
 using  towers  is $11.00  per  tower per year.
     Approximately 75 percent of chromate-using  CCT's  can be retrofitted
 with HEDE's.  The  remaining 25 percent  would require extensive
 rebuilding.   Therefore, it was assumed  that  CCT  operators would  switch  to
 nonchromate corrosion inhibitors rather than incur  the  high cost  of
 rebuilding the towers.  Costs per unit  area  for  each control alternative
 are shown  in  Tables 8-6 and 8-7.  As shown  in Table 8-6, the costs for
 substituting  nonchromates are relatively  low.  The  two  smallest building
 size classes  would incur costs of $0.12/m2  ^$0.01/ft2)  by changing to
 nonchromate corrosion inhibitors.  The  remaining four size classes would
 face even smaller  increases.  As shown  in Table 8-7, the two smallest
 building size classes would face costs  of $0.84/m2  ($0.08/ft2)  and $0.4/mz
 ($0.04/ft2) to retrofit with HEDE's.
     In the following section, the control costs are used to estimate the
 total annual  cost for each control  alternative by building type and
 size.  Percentage rent increases are also calculated to approximate the
 extent to which the occupants of such buildings might be affected by a
 regulation to control  Cr+s emissions from CCT's.
8.4  ECONOMIC IMPACTS
     The change to nonchromate corrosion inhibitors is  not  expected to
 have a significant economic impact  on owners or tenants of  affected
properties.  While the previously noted average annual  rental rate is
            2         9
about $129/m   ($12/ft ), the incremental annualized control  costs noted
                                   8-4

-------
above would add less than $0.12/m2  ($0.01/ft2) if all costs were passed
through in the form of increased rental rates.  Such costs would represent
a rent increase of about 0.1 percent.  Even those properties that are
rented at the lowest rate of $22/m2 ($2/ft2) would incur rent  increases of
less than 1 percent if nonchromate  alternatives are used.
     The installation of HEDE's also is unlikely to have significant
economic impacts on the owners of CCT's or their tenants.  Those owners
who retrofit their towers rather than switch to nonchromiurn-based
inhibitors would face control costs of $0.84/m2 ($0.08/ft2).  Assuming a
      22                         '
$129/m  ($12/ft ) rental rate, rents would increase by less than 1 percent
if costs are fully passed through to the tenant.  Properties in the
smallest size class which are rented at $22/m2 ($2/ft2) would incur rental
increases of 4 percent under the drift eliminator alternative.  However,
it is unlikely that these properties would retrofit the drift eliminators
but would switch to nonchromate corrosion inhibitors.
     The total nationwide costs of substituting nonchromate corrosion
inhibitors for chromates is estimated to be under $9.4 million.  The use
of HEDE's would entai;  nationwide costs of about $22 million.   Such
estimates are detailed in Tables 8-8 and 8-9 and have been made based on
the incremental annualized control  costs presented in Tables 7-1 and 7-2
and the building counts by type and size presented in Table 8-5.
     Finally, it is not expected that prohibiting the use of
chromium-based corrosion inhibitors would have a significant effect upon
either the chromate suppliers or the water treatment service companies
that provide chromate corrosion inhibitors.   This conclusion is based upon
the insignificant use of sodium dichromate in CCT's  relative to all  other
uses (i.e.,  less than 0.35 percent)  mentioned in  Section  8.2.1  above and
the fact that most specialty chemical  companies supply  nonchromate  as well
as chromate  corrosion inhibitors.
                                   8-5

-------
 TABLE 8-1.   MAJOR SUPPLIERS  OF  CORROSION  INHIBITORS  FOR  CCT'S
                LIGHT INDUSTRY,  AND  UTILITIESa
                                                     Average
                                                  market  share,
 Company                                              percent
NALCO Chemical Company
Mogul Division—Dexter Corp.
Betz Laboratories, Inc.
Drew Chemical Company
Dearborn Chemical Division—CHEMED Corp.
Calgon Corp.b
Others
19.2
14.4
13.3
 9.8
 9.8
 7.7
25.8
TOTAL
                                                    100.0
^Includes chromates as well  as nonchromates.
 Calgon indicated that they  do not sell  chemicals  for  use  in
 CCT's.
                             8-6

-------
 CO


 a
UJ

0.
CO


CO
03


U.

O
         :S!
        CN ^ป
        oT
C^sl

 I

CO
ca
            o

          So
          in  •
        riOO
        ซr in —
        •ป

       3 o
ฃ~s
         i
           0)
           Q.
           O)
           c


           •o
           m
                 oooo
                    งo    o
                    o    o
                                          o o
                                          o o
                                          oo

                                           8
                                                 o

                                                 8

                                                 o*
                                                 in
                o
                o
                      O
                      o
                      o
                          OO
                          oo
                          oo
                           i
                                    i
                                                 CM
                                                     x
                        t-    o
                        a    —
                       O
                        —  o>

                        
-------
LU



CO
 LU
 Q.
 CQ
CO

CO
 g
 LU
 CO
CO
        CM O
       28s.
 i
CO

LU

CO
<:
           rg
          L. (.
          O ioซOKios
                               iri
               in •— 
          c
                                             S
                            U)
                            • _o  — o    —    +-  ซ


             
-------
         So
e
CO
 o
 o
 OS
 O
         CM —

         oT
         CN  O
        oo<
        ซ-o
CQ


U_
O


0ฃ


03
 I
00
CO
orปinoป  — KIKS



                   CM  m        —  ซr  KI — iปi
                                                        O
                                                        in
                                                        CM
                   — in     rป — r-Ki  — rปrom
                   CO V     O
                      CM     fO    IO  in
                            pป  o>  
-------
a

H-
CD
 CD
 OQ
 CO
ง

5

CD
3=
MM
C/y
CJ
u_
o
CD
to
 I
CO

UJ
       ra
       <•
       CM
       4.
       u_
        in
        (U

        a
               a
           VO O
           esj a
           i-- in
           cn o
           PO OO
           CM CM
           ^H CO
         "TO
         CM CTป
         cvj cn
          in o
          O tn
            I VO
             cn
          0 O
          vo cvj
             IT)
             ai
             a.
           f
           •o
             3
            aa
                      CM
                            VO
                            co
                                                    i— io
                                                                 cn
                                                                             vo
                                                                                     CM
                                                                                   vo
                                                                                          ง
                                                                                          IT)
                                                                                           cn
                                                                                          o
                                                                                          VO
                                                                                          in

                                                                                          cn
                                                                                          o
                                                                                          cn
                                                                                          o
                                                                                          o
                                                                                          cn

                                                                                          vo"
                                                                                         0
                                                                                         CO
                                                                                         VO
                                                                                                 c

                                                                                                 o

                                                                                                 IS*
 
u
x
at
                                                   oo    r-.
                                                     ซ     •
                                                   CM    --4
                                                                       vo
                                                                                   vo
                     OO
                           CM
                                 CO
                                              OD
                                                          CO
                                                                CM
                                                    oo
                                                                      VO
                                                                      ^-
                                                                      CM
                                                                             CM
                                                                                         O
                                                                                         cn
                                                                                         in

                                                                                         CM
                                                                                        O
                                                                                        CO
                                                                                         CVJ
                                 (/>
                                 >    O    r—

                    r-    -i-    id
                                       O)
                     
                                             o
                                                   VI
                                                   ai
                                                   o
                                                   at
                                                  oo
                                                   -t->


                                                   u
                                                                                                ai

                                                                                                o

                                                                                                O)

                                                                                                
                                                                                               J3

                                                                                                OJ
                                                                                                0)
                                                         f-    ••-    a
                                                                             
-------
TABLE 8-6.  INCREMENTAL COST PER UNIT AREA FOR USING NONCHROMATE
                      CORROSION INHIBITORS
                         (1986 Dollars)

Model
tower
1
2
3
4
5
6
Model
building
m (ft2)
673 (7,240)
1,460 (15,720)
3,405 (36,650)
6,224 (66,990)
12,338 (132,800)
37,626 (405,000)
Total annual
substitution
cost,
$/yr/towera
83
102
151
221
372
1,001

Incremental
fost,
$/m ($/ft )
0.12 (0.01)
0.07 (0.01)
0.04 (<0.01)
0.04 (<0.01)
0.03 (<0.01)
0.03 (O.01)
  Cost includes annualized capital  cost of automatic feed
  system and annual  incremental  cost of phosphate-based
  treatment over chromate-based  treatment.
                              8-11

-------
TABLE 8-7.  INCREMENTAL COST PER UNIT AREA FOR RETROFITTING
                     DRIFT ELIMINATORS
                       (1986 Dollars)



Model
tower
1
2
3
4
5
6


Model
building
i/ (ft2)
673 (7,240)
1,460 (15,720)
3,405 (36,650)
6,224 (66,990)
12,338 (132,800)
37,626 (405,000)
Annual
certifica-
tion and
inspection
costs,
$/tower
511
511
511
511
511
.511


Annual 1 zed
retrofit
cost,
$/tower
51
73
116
182
314
655


Total
annual
cost,
$/tower
562
584
627
693
825
1,166
„


Incremental
cost, $/m
($/fO
0.84 (0.08)
0.40 (0.04)
0.18 (0.02)
0.11 (0.01)
0.07 (0.01)
0.03 (<0.01)
                          8-12

-------
 LU
 IVJ
 CO
 0.
 o
CO


CO

co

O
           CT
           C

   *o
   CO
    
             10. 0
             CO f)
                OO
            O O
            ^.CM
            ^T ^^
               %  *ป
                LO
              r> o

              o CM
00

00

LU


03
                0>
                a.
                CT>
                CQ
                               co
                         O    C7>
                           •      •
                         OO    00    O
                                                  T— i    LO
                                                  CO
                                                                     CO
                                                                                 LO
                                                                                       LO
                                                                                 LO
                                                                                       LO
                        ซ— 1    OO
                        ป— ICO
                               00
                              CO
                                           o   ^


                                           CM   V3
                                                        CM
                                                        CM
                                                              CM
                                                              CM
                                                        CMCO
                                                        CM    *->
                                     CO
                                                                           CM
                                                                          CM
                                                                    COCO
                                                                          CM
                                                                    O>    ซ— <
                                                                                O    ^H



                                                                                CO    i-H
                                                                                 CM
                                                                                       00
                        oo
                                     cr>
                                           co    oo    CM    LO    CM    o    co    co
                                                       CM    ซ-H    t— (
                        CT>
                              <ฃ>
                                                 CO
                                                                    00
                                                                                IO
                                                       vo    oo
                                                                          CM
                              03
                              C
                              O
                                     (/I
                                     a>
                                     s-
                                     O)
                                     V)
                                     O)
                                           OJ
                                           n3
                                           U
                                                        i-
                                                        O)
                                                       01
                                                 CJ)
                                           a;     o
                                                                                              LO
                                                                                              CO
                                                                                              oo

                                                                                              CTl
                                                                                              VO
                                                                                              VO
                                                                                              LO
                                                                                                ซ
                                                                                              oo
                                                                                              CO
                                                                                              LO
                                                                                              P-.
                                                                                             CM
                                                                                             o
                                                                                             vo
                                                                                             oo
                                                                                             l^s.
                                                                                             CM
                                                                                             00
                                                                                             CM
C

o

t/1
                                                                                                     U
                                                                                                     ns
                                                                                                     O)
                                                                                                     O)
                                                                                                     o
                                                                                                     X
                                                                                                    
                                                                                                    T3
                                                                                                    01
                                                                                                    o
                                                                                                    s_

                                                                                                    C
                                                                                                    01
                                                                                                    O)
                                                                                                    O)
                                                                                                   .a
                                                    8-13

-------
 CO
 Q.
 >-
                                  CO
                                                      CM    r-4
                                                                   OO    ^-4
                                                                                       to
                                                                                              oo
 00
 CO
 —I Vt
 UJ L.
 IJU I—
 •-* O
 Cฃ O
 a
    to
 Cฃ> CO
   :cn
O c/)
ฃX "O
J— C
UJ n3
ce: ui

U. o
CO
o
a
UJ
            CO
   *   ft
 CM CM
    OO
    O
CM cr>
CM O"ป
  ป  m
to to
    to
u> o
    10
    to
     <•ป
ro to
    oo
 O CM
   r*.
  ป   *
  I LO
                 o
                 *T
                 CM
) ซ3-
) cn
J cn
J CM

CO
to


CM
00
cn


t— i
CM
cn


rซป to
.CMOOOOLO
"CMt— ir-ป.cnLoooi— i
3 •— < o to to rv. rป. o
1 CO •-> r-4 LO ซT ^H

>cnTcocnoooซT
r-t lf> 
                                                                                                            cn  •
                                                                                                            •r-  t/l
                                                                                             J=  i-
                                                                                             +j  cn
                                                                                             •r-  O
                                                                                              3E  S_
                                to
                                              CO
                                                     CM
                                                           CM
                                                           LO
                                                                         OO    CM
                                                                         CM    oo
                                                                         CM    tO
                                                                                      LO
 CO
 cn
 CM
                                                                    O
                                                                    o
                                                                    cn
                         CO
-)
*
3
3
cn
i*-.
oo
oo
~
CO
LO
oo
CM
i— 4
LO
LO OO
T i— '
cn *a*
•-H 00
oo
*
cn
CM
oo
LO cn
ซ~H r**ซ
cn ->     co

 o     -p
 ™3      ^3
™C(      C3
                                              (U
                                              4-> O
                                                        0)  ,
                                                       •—     ซ
                                                        3  1- E
                                                        O  ->  C  CU
                                                      •r- O J=
                                                       X C 4->
                                                       O)
                                                          O  •"
                                                       O) -i-) T3
                                                      -ฃ=     0)
                                                      •ป-> J= T3
                                                          U  C
                                                      t- 4->  3
                                                      O •<-  O
              c
              (U 13
              O i—
              S- 3
              a> o
              CX 3
                                                                                                                  c
                                                                                                                  d)
                                                                                                                  OJ
                                                                                                   O)
                                                                                            LO 4->  fO
                                                                                            r*.  c ^
                                                                                                01
                                                                                            01  (J   LO  3
                                                                                           
-------
8.5  REFERENCES FOR CHAPTER 8

1.  Chromium Products.  Chemical Economics Handbook.  SRI International,
    Inc., Menlo Park, California,  p. 734.2000 C.

2.  Memorandum from JACA Corp., to D. Gillette, EPA:SASD.  Status report
    on the development of an economic impact analysis for cooling tower
    sources of chromium emissions.  December 23, 1985.

3.  Corrosion Inhibitors Market.  Frost and Sullivan, Inc.  1983.
    p. 132.                              :

4.  Memorandum from C. Hester, MRI, to R. Myers, EPA:ISB.  Technical
    Report—Cooling Towers.  September 27, 1985.  p. 3.

5.  Characteristics of Commercial Buildings 1983.  U.S.  Department of
    Energy.  Energy Information Administration.  DOE/EIA-0246(83).
    July 1985.

6.  Dollars and Cents of Shopping Centers:  1984.  Urban Land Institute.
    pp. 30-35.

7.  Telecon.   P.  Bell in, MRI,  with G. Schweitzer, Calgon Corporation.
    May 22, 1985.   Chromium use as a corrosion inhibitor in  CCT's.
                                  8-15

-------

-------
                             9.  OTHER IMPACTS

     This chapter discusses  the other  impacts  attributable  to  each  of  the
regulatory alternatives.
9.1  IMPACTS RESULTING FROM  PROHIBITING  CHROMIUM
     Prohibiting chromium use  in comfort cooling towers  (CCT's) would
eliminate air pollution and  health risks'TrWchromium emissions from
CCT's.  It is estimated that the use of  nonchromium treatment  chemicals
would increase by about 18 percent (assuming 15 percent  of  the CCT's
currently use chromium).  None of the alternative treatments (phosphates,
triazoles, or molybdates) are believed to be carcinogenic.
     The effects of prohibiting chromium use in CCT's on bacteria such as
Legionella pneumophila also have been investigated.1  Chromium is used in
cooling towers to control corrosion and  not for microorganism  control, and
there is no definitive work currently to substantiate that  chromium is
biocidal against Legionella pneumophila.  Usually when a problem with the
bacteria arises, high concentrations of  chlorine are used to reduce the
growth of the bacteria.
     Because some nonchromium treatment programs cannot use chlorine to
control  microorganisms, the Agency was concerned that the alternative
chemical programs may not provide comparable control of Legionella
pneumophila.   All cooling towers that have been associated with outbreaks
of Legionnaires disease were not following a good biocidal regimen using
an EPA-approved biocide at the dosages recommended by the manufacturer.l
     Solid waste disposal  would not be affected significantly because
currently there are no known CCT facilities that are treating wastewater
discharges for chromium onsite.  However, many industrial and research
facilities use chromium in both CCT's and industrial cooling towers  and
treat solid waste onsite.   Thus,  prohibiting chromium use in CCT's would
                                   9-1

-------
 reduce  slightly the amount of solid waste containing chromium.   In cases
 where sewage  treatment  plants are receiving chromium in  quantitites large
 enough  to  treat,  it is  likely that sources other than CCT's  contribute
 most of the chromium; thus,  the  effect  of reducing  chromium  in  CCT's  would
 be negligible.
     Water pollution from  CCT discharges  of chromium would be completely
 eliminated.   As chromate users switch to  nonchromate programs,  the
 discharges of the  nonchromates would increase.   The percentage  of  CCT's
 currently  using each of the  nonchromate treatments  is not known, but  it is
 expected that the  37,500 CCT's using chromates would switch  in  the same
 proportion.   Thus,  discharges of  each nonchromate chemical would increase
 by a maximum  of about 18 percent  (15 percent/85  percent  = 18 percent) if
 these compounds are not also  used-iS5~an additive to chromate treatments.
 Because phosphates,  the most  popular alternative treatment, are included
 in many chromate treatments,  phosphate  discharges would  increase by less
 than 18 percent.   In addition, even  if  all  of the plants switched  to
 phosphate  treatments and none of  the previously  used  chromate treatments
 included phosphate  in the  formulation,  the  nationwide  increase  in
 phosphorus discharges to sewage treatment plants would be less than
 0.1 percent.
     Worker exposure to Cr+6  would be completely eliminated.  At the
 present time, worker exposure to  Cr+6 is expected to  primarily involve
 dermal contact.  Little inhalation exposure  is expected because the
 chromate is expected to remain in solution as a dissociated salt.   A
worst-case scenario  for dermal exposure to the chromate solutions would be
 for an operator who  is not wearing gloves to open a valve connected to a
 drum of solution, fill  a pail with solution, and then pour the contents of
 the pail directly into the tower basin.   Assuming a concentration of
chromate in the solution of 5 to 65 percent, dermal  exposures to chromate
could range from 65 mg/d (1.4 x 10'" Ib/d) to 2,500 mg/d  (55 x lO"1" Ib/d)
 if gloves are not worn.   However, the trend in recent years  has  been to
 switch from manual to automated control  of the feed pumps.   Thus,  less
frequent exposure would  occur during operation of the feed  pumps.
Potential exposure to chromate during sampling is negligible  due to the
low concentration of chromate in  the water.  If manual control  is used,
                                    9-2

-------
 the maximum potential worker  population  exposed  is 75,000 based on
 estimates of 37,500 cooling towers  that  use  chromates and one to two
 operators per site.
 9.2   IMPACTS RESULTING FROM HIGHER  EFFICIENCY  DRIFT E'LIMINATOR RETROFITS
      Retrofit of higher efficiency  drift eliminators may  reduce the  air
 pollution and health risk from chromium  by 85  percent.  Water pollution
 would not be reduced from the present discharge  because the  cycles of
 concentration, chromium concentration, and blowdown rate  would not
 change.
     New higher efficiency cellular drift eliminators have been designed
 with pressure drops lower than the  older wood  or asbestos-cement
 eliminators.  Therefore, replacement of  an existing lower efficiency drift
 eliminator with a higher efficiency drift eliminator would result in  	
 decreased power consumption.  A manufacturer has indicated that .the
 horsepower savings can be as much as 35  percent when a herringbone drift
 eliminator is replaced with the most efficient cellular drift  eliminator
 in a counterflow tower.  The savings in  a crossflow tower would  not be as
 great.
9.3  REFERENCES FOR CHAPTER 9
 1.  Memorandum from R.  Myers,  EPA:ISB, to Comfort Cooling Tower Project
    Files.   October 16, 1986.   Summary of telephone conversation with B.
    Davis,  Center for Disease Control, concerning Legionnaire's disease.
2.  Memorandum from D.  Randall,  MRI, to D. Stackhouse, EPArSOB.
    September 3,  1987.   Calculations of phosphorus discharges from model
    CCT's and nationwide discharges  of phosphorus to sewage treatment
    plants.
3.  Kelly,  G.  M.   A System-Efficient Approach to  Cooling  Tower Energy
    Modifications.   Cooling  Tower Institute  Technical Paper
    No.  TP-85-18.   New  Orleans,  Louisiana.   January 1985.
                                   9-3

-------

-------
                                APPENDIX A.
              EVOLUTION  OF THE  BACKGROUND INFORMATION DOCUMENT

     The purpose of this  study was to develop  a  basis for  supporting
proposed chromium emission standards for CCT's.  To  accomplish the
objectives of this program, technical data were  acquired on the following
aspects of chemical treatment programs and CCT's:  (1) formulations,
effects, and costs of water chemical treatments; (2) the release of
hexavalent chromium emissions into the atmosphere by CCT's; and (3) the
types and costs of demonstrated control technologies.  The bulk of the
information was gathered  from the following sources:
     1.  Technical literature;
     2.  State, regional, and local air pollution control agencies;
     3.  Site visits and  case studies;
     4.  Industry representatives; and
     5.  Equipment vendors.
Significant events relating to the evolution of the background information
document are itemized in Table A-l.
                                   A-l

-------
             TABLE  A-l.    EVOLUTION  OF  THE  BACKGROUND  INFORMATION  DOCUMENT
 Date
             Coapany,  consultant, or agency/location
                                                                                   Nature of action
 02/07/85    Mogul  Division, Dexter Corp., Chagrin Falls, Ohio
 02/21/85    Harley Cooling Tower Coapany, Mission, Kans.
 04/16/85    Mobil  Oil Corp., Torrance. Calif.
 04/16/85    ARCO Petroleum Refinery, Carson, Calif.
 04/17/85    Chevron U.S.A., Richmond. Calif.

 04/18/85    Aaoco  Oil Coapany, Chicago, 111.
             Atlantic Richfield Company, Los Angeles.  Calif.
             Chevron U.S.A., Inc., San Francisco, Calif.
             Exxon  Coapany, U.S.A., Houston. Tex.
             Gulf Oil Products Coapany, Houston, Tex.
             Mobil  Oil Corp.. Fairfax, Va.
             Phillips Petroleum Coapany. Bartlesville.  Okla.
             Shell  Oil Coapany, Houston, Tex.
             Texaco, Inc., Houston, Tex.
 04/23/85    Araco,  Inc.. Middletown, Ohio
             Bethlehe* Steel Conpany, Bethlehem. Pa.
             Inland  Steel Corp., East Chicago, Ind.
             LTV Steel Coapany,  Cleveland,  Ohio
             Lone Star Steel Coapany, Lone Star, Tex.
            McLouth Steel Products Corp.,  Trenton, Mich.
            National Steel Corp.,  Pittsburgh, Pa.
            U.S. Steel Corp.,  Pittsburgh,  Pa.
            Wheel ing-Pittsburgh Steel  Corp., Wheeling, W. Va.
 05/08/85    Chemical Manufacturers Association, Washington, O.C.
09/01/85    U. S. EnvironHntal  Protection Agency
01/15/86    Association of Building  Owners and Management

05/13/86    U. S. Environmental  Protection Agency

05/22/86    National Bureau of Standards, Gaithersburg, Md.
06/04/86    Union Oil Coapany, Los Angeles, Calif.
06/12/86    Chevron U.S.A., San  Francisco. Calif.
            Shell Oil Coapany. Houston, Tex.
06/16/86    Hooker Industrial  and Specialty Chemicals,
              Niagara Falls, N.Y.
            Inmont Corp..  Clifton, N.J.
06/17/86    North Carolina State University, Raleigh, N.C.

06/20/86    Interlake, Inc., Oak Brook, 111.
            Kaiser Aluminum &  Chemical Corp., Oakland, Calif.
            LTV Corp., Cleveland, Ohio
06/23/86-   U. S. Environmental  Protection Agency
06/27/86
07/02/86    University of  North  Carolina at Chapel Hill,
              Chapel Hill,  N.C.
07/03/86    Duke University, Durham, N.C.
 Visit to water chemicals  vendor.
 Visit to cooling tower manufacturer.
.Visit to petroleum refinery.
 Visit to petroleum refinery.
 Visit to petroleum refinery.

 Section 114 information request.
 Section 114  information request
 Request for information about cooling water and
   corrosion Inhibitor use from member chemical
   manufacturing plants.
 Technical Report—Cool ing Towers
 Requesting ABOM participation in cooling towers study.

 Start Action Request for Development of Accelerated
   NESHAP—Chromium Emissions  From Cooling Towers
 Site visit.
 Section 114 information  request.
 Section 114 information  request.

 Section 114 information  request.
Case study.

Section 114 information request.


Emission tests  at Department of Energy Gaseous
  Diffusion  Plant, Paducan, Ky.
Case study.

Case study.
07/08/86

07/10/86
07/15/86

07/16/86

07/17/86
Wake Medical Center, Raleigh. N.C.
Peoples Security Insurance, Durham, N.C.
Crabtree Valley Mall, Raleigh, N.C.
Greenbrier Mall, Norfolk, '/a.
Sovran Bank. Norfolk, Va.
Oani International Hotel, Norfolk. Va.
Old Dominion University, Norfolk, Va.
Humana Bayside Hospital, Norfolk, Va.
Case study.
. Case study.
Case study.
Casa study.

Case study.

Case study.
                                                                                                              (continued)
                                                          A-2

-------
                                           TABLE  A-l.    (continued)
 Date
             Coapany, consultant, or agency/location
                                                                                    Nature of action
 07/22/86
 07/29/86

 08/12/86
 08/21/86
08/18/86-
08/22/86

09/01/86-
09/05/86
09/15/86
09/17/86

09/22/86
 North Carolina State University, Raleigh, N.C.
 North Carolina State Department of Administration,
  Raleigh, N.C.
 Industry representatives. Associations, and Concerned
  Individuals
ARCHEM, Inc.. Virginia Beach, Va.
Anderson Cheisical Company, Litchfield, Minn.
Anderson Chemical Company, Macon, Ga.
Aqua-Chen, Inc., Raleigh, N.C.
Betz Laboratories, Trevose, Pa.
Calgon Corp.. Pittsburgh. Pa.
Cheatreat. Inc.. Ashland, Va.
Dearborn Chemical Division—CHEMED Corp.
Drew Chemical Company, Boonton,  N.J.
Dubois Chemical Division—CHEMEO Corp..
Hercules, Inc., Wilmington, Del.
Industrial Maintenance Corp., Raleigh,  N.C.
Mogul Division—Dexter Corp., Charlotte,  N.C.
Nalco Chemical Company, Oak Brook, 111.
Olin Water Services, Inc., Overland Park, Kans.
Unichem International,-Inc., Hobbs, N.  Hex.
Water Chemist, Inc., Los Angeles, Calif.
Water Chemistry, Inc., Norfolk,  Va.
Water Services, Inc.,  Knoxville,  Tenn.
U. S. Environmental  Protection Agency
U. S. Environmental  Protection Agency

Federal Register
National Air Pollution Control  Techniques Advisory
  Committee
U. S. Environmental  Protection Agency
10/03/86    Working group


12/08/86    Working Group

02/02/87    Working Group

02/23/87    U.  S.  Environmental  Protection Agency


04/28/87    Steering Committee
 Followup to case study.
 Visit to State water treatment chemicals
   purchasing agency.
 iNotice of September 17-19, 1986, meeting of National
   Air Pollution Control  Techniques Advisory Committee
   and draft of Federal Register Notice of Solicitation
   of Information.
 Requesting information on comfort cooling
   tower population and chemical treatment
   program technical  and  cost data.
 Emission tests at National Bureau .of
  Standards, Gaithersburg, Md.

 Emission tests at Exxon Company Petroleum
  Refinery, Baytown, Tex.
 Notice of Solicitation of Information.
 Meeting.

 Press release concerning comfort cooling tower study
  and soliciting information on aspects of regulating
  chromium use in cooling towers.
 Meeting to discuss status of project and appropriate
  authority.

 Meeting to discuss draft Regulatory Impacts Analysis,
  Preamble, and Regulation
 Mai lout of draft Regulatory Impacts Analysis, Preamble,
  and Regulation.
 Document title changed from Regulatory Impacts Analysis
  to Background Information Document.

Mai lout of draft Background Information Document
                                                          A-3

-------

-------
          APPENDIX B.  PUBLIC CANCER RISKS FROM THE EMISSIONS OF
              HEXAVALENT CHROMIUM FROM COMFORT COOLING TOWERS

B.I   INTRODUCTION
B.I.I.  Overview
      The quantitative expressions of public cancer risks  presented in this
appendix are based on (1) a dose-response model that  numerically  relates
the degree of exposure to airborne hexavalent chromium (Cr+6) to  the  risk
of getting lung cancer and (2) numerical expressions  of public  exposure  to
ambient air concentrations of Cr"1"6 estimated to be caused by emissions
from  comfort cooling towers (CCT's).  Each of these factors is  discussed
briefly below, and details are provided in the following sections  of  this
appendix.
B.I.2  The Relationship of Exposure to Cancer Risk
      The relationship of exposure to the risk of getting lung cancer  is
derived from epidemiclogleal  studies in occupational  settings rather  than
from  studies of excess cancer incidence among the public.  The
epidemiological methods that have successfully revealed associations
between occupational exposure and cancer for substances such as asbestos,
benzene, vinyl chloride, and  ionizing radiation as well as for chromium
are not easily applied to the public sector with its  increased number of
confounding variables, much more diverse and mobile exposed population,
lack of consolidated medical  records,  and almost total absence of
historical  exposure data.  Given such  uncertainties,  EPA  considers it
improbable  that any association,  short of very large increases in cancer,
can be verified in the general  population with any reasonable  certainty by
an epidemiological  study.  Furthermore,  as noted by the National Academy
of Sciences (NAS),  ". .  . when  there  is  exposure to a  material,  we are not
starting at an origin of zero cancers.   Nor  are  we starting  at an origin
                                   B-l

-------
 of zero carcinogenic agents in our environment.  Thus, it is likely that
 any carcinogenic agent added to the environment will act by a particular
 mechanism on a particular cell population that is already being acted on
 by the same mechanism to induce cancers."1  In discussing experimental
 dose-response curves, MAS observed that most information on carcinogenesis
 is derived from studies of ionizing radiation with experimental  animals
 and with humans which indicate a linear no-threshold dose-response
 relationship at low doses.   They added  that although some evidence exists
 for thresholds in some animal  tissues,  by and large, thresholds  have not
 been established for most tissues.   The NAS concluded that establishing
 such low-dose thresholds "...  would require massive,  expensive,  and
 impractical  experiments ..."  and  recognized that the  U.S.  population
 M.  .  .  is  a  large,  diverse,  and  genetically heterogeneous group  exposed to
 a  large variety of  toxic agents."   This fact,  coupled with the known
 genetic variability to carcinogensis and the predisposition  of some
 individuals  to  some form of  cancer, makes  it extremely  difficult,  if not
 impossible,  to  identify a threshold.
      For these  reasons,  EPA  has  taken the  position,  which  is shared  by
 other Federal regulatory agencies,  that  in  the absence  of  sound  scientific
 evidence to  the contrary, carcinogens should be considered to pose some
 cancer  risk  at  any  exposure  level.  This no-threshold presumption  is based
 on  the  view  that as  little as one molecule of a carcinogenic substance may
 be  sufficient to transform a normal cell into a cancer cell.  Evidence  is
 available from  both the  human and animal health literature that cancers
 may arise from  a single transformed cell.  Mutation research with ionizing
 radiation in cell cultures indicates that such a transformation can occur
 as  the  result of interaction with as little as a single cluster of ion
 pairs.   In reviewing the available data  regarding carcinogenicity,  EPA
 found no compelling scientific reason to abandon the no-threshold
 presumption for Cr+s.
     In developing the exposure-risk relationship for Cr"1"6, EPA has
 assumed that a linear no-threshold relationship _
-------
 and levels of public exposure is  the same as that between cancer risks and
 levels  of occupational  exposure.   The EPA believes that this assumption is
 reasonable for public health  protection  in light of presently available
 information.   The  exposure-risk relationship used by EPA represents a
 plausible uppe--limit risk  estimate  in the sense that the risk is probably
 not higher than the  calculated level  but  could  be lower.
      The  numerical constant that  defines  the exposure-risk relationship l
 used by EPA in its analysis of carcinogens is called the  unit risk
 estimate.   The unit  risk estimate for an  air pollutant is defined as  the
 lifetime  cancer risk  occurring in a  hypothetical  population in which  all
 individuals are exposed continuously  from birth  throughout their  lifetimes
 (about  70  years) to  a concentration of one ug/m3  of  the agent in  the  air
 which they breathe.   Unit risk estimates  are used for two purposes:
 (1)  to  compare the carcinogenic potency of several agents with each other
 and  (2) to give a crude indication of the  public  health risk  which might
 be  associated  with estimated air  exposure  to  these agents.  The
 comparative potency of different  agents is more reliable  when  the
 comparison is  based on studies of like populations and on  the  same route
 of exposure, preferably inhalation.
     The Health Assessment  Document for Chromium  (HAD) (EPA 600/8-83-014F)
 contains the derivation of  the unit risk number.2  The HAD notes that
 although there  are many epidemiologic studies demonstrating that chromium
 is a potential  human carcinogen,  few provide adequate exposure data for
 use  in risk estimation purposes.   It is not clear from the epidemiological
 studies whether only hexavalent or both trivalent and Cr+6 are responsible
 for the increased cancer risk.  Because Cr+s compounds have generally
yielded positive results in animal bioassays and mutagenicity studies and
 trivalent  (Cr+ ) generally have not,  EPA  has taken the position that Cr+s
 is the form responsible for the carcinogenic response.  However, this
position may change pending results of research  currently underway.
B.I.3  Public Exposure
     The unit risk  estimate is only one of the factors needed  to produce
quantitative expressions of public health  risks.   Another  factor needed is
a numerical expression of  public exposure, i.e.,  the  numbers of people
exposed  to the various concentrations  of Cr+s.  The difficulty of  defining
                                   B-3

-------
 public exposure was noted by the National Task Force on Environmental
 Cancer and Health and Lung Disease in its 5th Annual Report to Congress in
 1982.    The Task Force reported that ".  . .a large proportion of the
 American population works some distance  away from their homes and
 experience different types of pollution  in their homes, on the way to and
 from work, and in the workplace.   Also,  the American population is quite
 mobile,  and many people move every few years."  They also noted the
 necessity and  difficulty of dealing with long-term exposures because of
 "...  the long latent period required for the development and expression
 of neoplasia [cancer]  .  .  .  ."
     The EPA's numerical  expression of public exposure  is based on two
 estimates.  The first  is  an  estimate of  the magnitude and location of
 long-term average ambient  air concentrations of Cr+s in the vicinity of
 emitting sources,  which  is based on dispersion modeling using  long-term
 estimates of source  emissions and  meteorological  conditions.   The  second
 is  an estimate of the  number and distribution of  people living in  the
 vicinity of  emitting sources based  on  Bureau  of Census  data which
 "locates"  people  by population centroids  in  block group or  enumeration
 district (BG/ED)  areas.  The people  and concentrations  are  combined  to
 produce  numerical  expressions of public exposure  by  an  approximating
 technique  contained in a computerized model.  The methodology  is described
 1n  B.3 below.
 B.I.4  Public  Cancer Risks
     By  combining  numerical  expressions of public exposure with the unit
 risk estimate, two types of  numerical expressions of public cancer risks
 are produced.  The first, called individual risk, relates to the person or
 persons  estimated to live  in the area of  highest concentration as
 estimated by the dispersion model.   Individual risk is expressed as
 "maximum lifetime risk."  As used here, the word "maximum" does not mean
 the greatest possible risk of cancer to the public.   It  is based only on
 the maximum exposure estimated by the procedure used. The second,  called
 aggregate risk, is a summation of all the risks to people estimated to be
 living  within the vicinity (usually within 50 kilometers) of a  source and
 is customarily summed for all the sources in a particular category.  The
aggregate risk is expressed as incidence  of cancer among all of the
                                   B-4

-------
 exposed  population after 70 years of exposure; for statistical
 convenience,  it is often divided by 70 and expressed as annual cancer
 incidence.  These calculations are described in more detail in B.4 below.
 B.2   THE UNIT RISK ESTIMATE FOR HFXAVALENT CHROMIUM
      The following discussion is summarized from a more detailed
 description of the Agency's Cr+s unit risk estimate in the HAD mentioned
 above.    The  model  used  to  estimate risk  is linear with age-specific
 incidence being a function  of the background incidence, age of the
 individual, and dose  to  which the person  is exposed.   The theory relating
 the maximum likelihood and  nonlinear least square estimation was used to
 estimate the  key parameters in the model.   Calculating the unit risk also
 required estimating the  probability of  surviving and  relied upon U.S.
 vital statistics.
      The unit  risk estimate for  Cr"1"6  was based  on the Mancuso  (1975)  data
 in which a cohort of  332 white male workers  who were  employed  in a
 chromate plant  between 1931 and  1937  were  followed  to 1974."   In his
 study, Mancuso  reported  lung  cancer death  rates by  levels  of exposure to
 soluble,  insoluble, and total  chromium concentrations.  Because  only  lung
 cancer mortality for  total  chromium exposure was  reported  by age  group,
 EPA's Carcinogen Assessment Group used only the  dose-response data for
 total chromium  to estimate the carcinogenic potency of Cr+s.  Although the
 use of dose-response data for  total chromium results  in an underestimation
 of the potency of Cr+6, the effect of this underestimation is
 approximately compensated for by other factors that may overestimate the
 risk such as the failure of the author to correct for smoking.
     The unit risk estimate calculated for Cr+s based on the Mancuso study
 is 1.2xlO~ .   This means that  if a person is continuously exposed for
 70 years to 1 pg/m , the probability of getting lung cancer would not
 likely exceed 1.2 chances in 100.  There are numerous uncertainties
concerning this estimate.  The effects of  age,  sex, race,  and general
health of the  sensitivity of responses to  Cr+s  exposure are unknown.
Because of the unavailability of sufficient data to correct for these
factors,  the .impact of these factors cannot be  addressed in this
assessment.
                                   B-5

-------
 B.3  QUANTITATIVE EXPRESSIONS OF PUBLIC EXPOSURE
      The EPA's Human Exposure Model (HEM) is a general model capable of
 producing quantitative expressions of public exposure to ambient air
 concentrations of pollutants emitted from stationary sources.  The HEM
 contains (1)  an atmospheric dispersion model, which included
 meteorological data, and (2) a population distribution estimate based on
 Bureau  of Census data.   The only input data needed to operate this model
 are source data, e.g.,  plant location, height of the emission release
 point,  and temperature  of the off-gases.   Based on the source data,  the
 model estimates the  magnitude and distribution of ambient air
 concentrations of the pollutant in the vicinity of the source.   The  model
 1s  programmed to estimate these concentrations within a radial  distance of
 50  kilometers from the  source.   If other  radial  distances are preferred,
 an  override feature  allows the  user to select the distance desired.   The
 selection of  50 kilometers as the programmed  distance is  based  on  modeling
 considerations,  not  on  health effects  criteria or EPA policy.   The
 dispersion model  contained in HEM is reasonably  accurate  within
 50  kilometers.   If the  user wishes  to  use  a dispersion  model  other than
 the one  contained  in HEM to estimate ambient  air concentrations in the
 vicinity of a source, HEM can accept the concentrations if  they are  put
 into an  appropriate  format.   It  also is possible to evaluate  the effect
 particle deposition  near the  stack  has on  the  ambient air concentrations
 of  the pollutant.  A detailed description  of  the  HEM  can be found  in
 Reference  5.
     Based  on  the  radial  distance specified, HEM  combines numerically the
 distributions  of pollutant  concentrations  and people to produce
 quantitative  expressions  of public  exposure to the pollutant.   The HEM
 allows for  estimates  to  be made for both point sources and area sources.
 B.3.1  Model Selection and  Description
     The area  source model that is contained in HEM was selected to assess
 the carcinogenic risks from CCT's for several  reasons:  (1) the nationwide
 population of CCT's  is estimated to be about 250,000; of these, about
 37,500 are estimated to use chromium-based water treatment chemicals,
 (2)  the specific locations of CCT's are largely unknown, (3) CCT's  are
most likely to be located in urban areas,  which the area source model is
                                    B-6

-------
well suited to address,  and  (4)  the majority of Cr+6  emitted from the
CCT's  is expected to  remain  airborne  and  be  dispersed over a large area.
As discussed  in Section  5.1.2.1,  the  majority of the  droplets in the plume
are expected  to evaporate rapidly because they are  smaller than 30 ym.
Thus,  it was  not necessary to  include deposition in the modeling effort.
     The area source  model is  used for those sources  which cannot be
specified in  detail.  The allocation  of emissions from such sources  must
be inferred by relating  the  source to a correlated  parameter such as
population, motor vehicles,  etc.   For these  sources,  the dispersion  of
emissions is  then modeled by a simplified  dispersion  algorithm  to estimate
concentration patterns.  The Gifford  urban area  dispersion algorithm
(Hanna and Gifford, 1973) has proved  to be a simple but physically
realistic model capable  of estimating atmospheric pollutant concentrations
caused by area source emissions in cities.   The  basic  Hanna-Gifford
equation is given as:
                                 X = CQQ/U
(1)
where X is air pollutant concentration, QQ is the effective emissions rate
per unit area, and U is the wind speed.  The parameter C, generally
referred to as the Gifford coefficient, is a weak function of the city
size; it may be taken to be approximately constant.  Theoretically, the
parameter C is given by:
                        C = (2)1/2-X1-b/[a(l-b)]-1
(2)
where X is the distance from a receptor point to the upwind edge of the
area source.  The constants a and b are defined by the vertical
atmospheric diffusion length, Oz = axb.  Values of a and b for different
atmospheric dispersion conditions have been discussed by Pasquill  (1970,
1971).  The parameter C can be estimated for various combinations  of the
stability factors a and b and by assuming that X equals half the city size
(Hanna, 1978).  For example, 213 would be an appropriate value of  C for a
city with a land area of 400 km2 under Pasquill Class D stability  (where
a = 0.15 and b = 0.75).  Specific values of the parameter C have been
                                   B-7

-------
 empirically estimated by Hanna and Gifford for a large number of U.S.
 cities based on a large quantity of air quality data, average annual
 emissions,  and meteorological conditions.  The mean value of C has been
 found to equal 2-25,  with a standard deviation roughly half that
 magnitude.   This value of the parameter C has been recommended for use in
 evaluating  an area source by EPA if removal  and decay processes may be
 neglected.   Estimates of the parameter C were calculated by using
 Equation 2  and by assuming Pasquill Class D  stability as the average long-
 term  meteorological  condition.
      The application of the Gifford approach within HEM has been modified
 to  provide  variation of atmospheric concentration  across a modeling region
 in  proportion to the local  emission rate per unit  area.   This approach
 provides a  higher degree of resolution of concentration  patterns than does
 the single  urban box approach but  does not address  the details of
 pollutant advection  and dispersion that are  treated by grid dispersion
 models.
      In  the  present  approach, box  model  (Gifford model)  dispersion  results
 are simply  scaled  at each BG/ED  by the ratio  of the density of emissions
 per surface  area at  the BG/ED to the regional mean  emission density.
      Options  in  the  AREA code provide  for  varying or nonvarying  (from
 district to district) emission rates.   Emissions that  vary  with  BG/ED,
which was used for CCT's,.. are scaled by  the population density of the
 BG/ED.  This  is  to address  pollutant-emitting activities that uniform
fractions of  the population are expected to be engaged in at any given
time.  Examples of such  activities  are motor vehicle usage  and operation
of  home furnaces.
     The basic Hanna-Gifford equation  (1)  shows that the concentration is
 inversely proportional  to the wind  speed.  In HEM,  each wind speed in the
stability array  (STAR)  set  is used.3  The STAR matrix is summed over wind
direction and stability class to give the freer  icy of occurrence of each
 Stability array refers to meteorological  data usually collected at
 airports.  These data consist of frequency distributions of wind speed
 and direction and atmospheric stability.
                                   B-8

-------
 speed.   The concentration is computed as the sum of the frequency-weighted
 concentrations  for each wind speed.
      The population data in the area model  require estimates of population
 data  density.   This information was  not available when the model was
 developed.   The only data taken from Bureau of Census were the location
 (UTM  coordinates of centroid)  and  population of each 8G/ED.  In the
 absence  of  information  on the  area of eac.h  BG/ED, arbitrary estimates  are
 made  for each BG/ED.
      Estimates  of BG/ED areas  require dividing the analysis region  into a
 grid  of  Cartesian cells.   The  size and number of cells are chosen to
 produce  a grid  mesh that  is as fine  as resources permit.   Ideally,  grid
 cells should be much  smaller than  the distance between the closest  BG/ED
 centroids,  but  much coarser resolution may  be acceptable,  depending on
 analysis  goals.
      The  "best  guess" area  of  any  BG/ED  is  defined  here to be  the sum of
 the areas of grid  cells for which  the centroid  of the  BG/ED  in question is
 the closest centroid.   If more  than  one  BG/ED  centroid  falls within a grid
 cell, the cell  area is  divided  among  the districts  so  identified.   The
 cells are scanned  in square  "spirals"  about each  centroid, with  cells
 "belonging" to  their centroid of origin until  spirals overlap.   Specific
 radius tests resolve the  "ownership"  of cells  in  overlapped portions of
 spirals.
 B.3.2  Input Data  and Results
     To facilitate  area source modeling and conserve computer resources,
 given that population exposure and cancer risks are proportional to  the
 chemical's carcinogenic potency and emissions, the area source model was
 previously run for each U.S. county and aggregated to the State  level for
 a unit emission rate of 1 kilogram per person and a cancer potency of
 1.00 x 10"  (lifetime probability of  cancer per ug/m3 of the modeled
pollutant).   This run provides  for each county and State an estimated unit
annual cancer incidence that may be scaled  by the actual potency factor
for Cr+  and the Cr+  emission  rate by State to obtain specific area
source estimates of risk for the CCT  category.
                                  ' B-9

-------
    cancer incidence
     (State-CCT's)
                      + 6
 cancer incidence   Cr  potency.
(State-model  run)       0.1
emission rate/
person (State)
     1 kg
      Table B-l lists population,  annual  incidence (prescaling),  the
 emission rate (person x 10~5 grams),  and estimated annual  incidence for
 each State.   (Chapter 4 provides  a discussion of the emission rate
 calculations).   Alaska was  assumed  to have no  CCT's  that emit Cr+6.   The
 total  nationwide annual  incidence for Cr+6 was estimated to  range from 4
 to 112 per year.  This range  reflects lower- and  upper-bound emissions
 estimates.   (For more  detail,  see Chapter  4.)
     To estimate maximum lifetime risk,  the largest  model plant was
 assumed to release all  the  Cr+6  at  ground  level (1.5 m)  (see Table B-3).
 Ground-level release was specified  since concentrations  are  inversely
 proportional to  release  height.   This plant was then placed  in 50 large
 cities,  one  in each State plus Washington,  D.C. (Alaska  was  excluded  from
 the analysis).   The highest maximum lifetime risk ranged from 2.3 x KT6
 to 6.6  x 10"s.   Table  B-2 shows  the cities  used to estimate  maximum
 lifetime risk.   The maximum lifetime risk  is calculated  by multiplying the
 Cr+6 unit risk factor  (1.2 x 10'2)  by the highest concentration to which
 any person is predicted  to be exposed  (1.9  x 10"" to 5.5 x 10"3 yg/m3
 annual  average by  State).  Table  B-3  shows  the model CCT emission
 parameters.  Table B-4 shows the  range of maximum concentrations to which
 people  are predicted to  be.exposed  by State.
 B.4  ANALYTICAL  UNCERTAINTIES APPLICABLE TO THE CALCULATIONS OF PUBLIC
     HEALTH RISKS CONTAINED IN THIS APPENDIX
 B.4.1  The Unit  Risk Estimate
     The procedure used  to develop  the unit risk estimate is described in
 Reference 2.   The model used and  its application to epidemiological  data
 have been the subjects of substantial comment  by health scientists.   The
uncertainties are too complex to be  summarized  in  this  appendix.
     The unit risk estimate used in  this analysis  applies only to  lung
cancer.  Other health effects are possible; these  include respiratory
tract irritation and hypersensitivity, i.e., asthmatic-like  symptoms.   No
numerical expressions of risks relevant to  these health effects  is
included in this analysis.
                                   B-10

-------
        TABLE  B-l.  HEXAVALENT CHROMIUM EMISSION RATE  PER PERSON  AND
                           ANNUAL INCIDENCE .BY STATE


Area  source analysis (by county)
Emission rate = 1 kg/person/yr
Unit  potency =0.1
Cr ฐ  potency =0.012
State
A 1 abama
Alaska
Arizona
Arkansas
Ca 1 i f orn i a
Co 1 orado
Connecticut
Delaware
D.C.
Florida
Georgia
Hawa i !
Idaho
1 1 1 inois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Ma i ne
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carol ina
Nor^h Dakota
Ohio
Ok 1 ahoma
Oregon
Pennsylvania
Rhode 1 s 1 and
South Carol ina
South Dakota
Tennessee
Texas
Utah
Vermont
V i rg i n i a
Wash i ngton
West Virginia
Wisconsin
Wyoming
Total annual incidence
Population
4,030,000
448,000
3,010,000
2,400,000
24,800,000
3,110,000
3,130,000
610,000
608,000
10,600,000
5,710,000
1,020,000
. 1 ,020,000
11,500,000
5,570,000
2,940,000
2,410,000
3,790,000
436,000
1,170,000
4,330,000
5,780,000
9,400,000
4,160,000
2,600,000
5,010,000
. 826,000
1 ,600,000
891 ,000
981,000
7,450,000
1,390,000
17,400,000
6,110,000
667,000
10,800,000
3,170,000
2,300,000
11,800,000
953,000
3,280,000
700,000
4,790,000
15,100,000
1,580,000
530,000
5,590,000
4,310,000
2,000,000
4,820,000
514,000

Unit annual
incidence
7,800
73
12,000
23,000
160,000
14,000
11,000
13,000
15,000
41 ,000
1 1 ,000
3,200
1,300
82,000
1 1 ,000
2,800
4,300
5,200
17,000
750
34,000
26,000
29,000
8,800
2,800
12,000
700
3,900
5,000
960
52,000
2,800
290,000
1 1 ,000
430
25 ,000
5,300
8,200
60,000
2,500
5,500
380
8,800
28,000
5,900
250
19,000
14,000
1,900
8,200
410

Annual Cr*6 .
emissions
per person,
kgxIoVyr
4.01-113.8
0.0-0.0
3.74-106.1
3.80-108.0
3.67-104.1
1.97-55.9
2.24-63.6
2.24-63.6
6.04-171.6
4.01-113.8
6.79-192.8
1.43-40.5
2.85-81.0
2.85-81.0
2.58-73.3
2.85-81.0
2.85-81.0
4.41-125.3
1.43-40.5
3.12-88.7
2.24-63.6
2.24-63.6
1.97-55.9
4.01-113.8
2.85-81.0
1.70-48.2
2.58-73.3
2.65-75.2
1.83-52.1
2.85-81.0
2.65-75.2
2.24-63.6
3.60-102.2
1.70-48.2
2.65-75.2
3.67-104.1
1 .56-44.3
2.65-75.2
2.24-63.6
4.01-113.8
2.24-63.6
3.40-96.4
4.28-121 .5
2.11-59.8
1.70-48.2
2.85-81.0
1.36-38.6
2.85-81 .0
2.11-59.8
1.70-48.2
3.40-96.4

Cr+6
annual
incidence
0.038-1 .1
0
0.054-1 .5
0.11-3.0
0.70-20
0.033-0.94
0.030-0.84
0.035-0.99
0.30-8.4
0.053-1.5
0.026-0.74
0.002-0.063
0.28-8.0
0.038-1 .1
0.009-0.25
0.015-0.42
0.018-0.51
0.090-2.6
0.001-0.036
0.13-3.6
0.070-2.0
0.078-2.2
0.021-0.59
0.014-0.38
0.041-1.2
0.001-0.041
0.012-0.34
0.016-0.45
0.002-0.060
0.18-5.1
0.009-0.25
0.78-22
0.048-1 .3"
0.001-0.025
0.080-2.3
0.023-0.66
0.015-0.44
0. i9-5.4
0.007-0.19
0.026-0.75
0.001-0.029
0.036-1 .0
0. 14-4. 1
0.015-0.42
0.001-0.015
0.065-1 .3
0.023-0.65
0.007-0. 18
0.021-0.59
0.001-0.024
0.061-1 .7
4-112
                                      B-ll

-------
TABLE B-2.  LOCATIONS USED IN ESTIMATING INDIVIDUAL RISK
Location
Alabama, Birmingham
Arizona, Phoenix
Arkansas, Little Rock
California, Los Angeles
Colorado, Denver
Connecticut, Hartford
Delaware, Dover
Washington, D.C.
Florida, Miami
Georgia, Savannah
Hawaii, Honolulu
Idaho, Boise
Illinois, Chicago
Indiana, Indianapolis
Iowa, Des Moines
Kansas, Kansas City
Kentucky, Louisville
Louisiana, Baton Rouge
Maine, Augusta
Maryland, Baltimore
Massachusetts, Boston
Michigan, Detroit
Minnesotta,- Duluth
Mississippi, Jackson
Missouri, Kansas City
Montana, Helena
Nebraska, Omaha
Nevada, Las Vegas
New Hampshire, Concord
New Jersey, Trenton
New Mexico, Albuquerque
New York, New York City
North Carolina, Charlotte
North Dakota, Bismarck
Ohio, Cincinnati
Latitude

Degrees Minutes
33
33
34
34
39
41
39
38
25
32
21
43
41
39
40
39
38
30
44
39
42
42
46
32
39
46
41
36
43
40
35
40
35
46
39
31
27
44
03
43
46
10
54
46
04
19
37
53
46
22
07
16
23
19
17
21
20
47
18
05
36
16
11
12
13
05
43
14
48
06
Lonqitude

Degrees Minutes
86
112
92
118
105
72
75
77
80
81
157
116
87
86
91
94
85
91
69
76
71
83
92
90
94
112 ,
95
115
71
74
106
24
80
100
92
50
05
15
15
01
41
32
01
12
05
52
13
38
09
26
36
45
11
47
36
04
03
06
12
35
01
57
08
32
45
40
01
50
47
56
                                                     (continued)
                         B-12

-------
                          TABLE B-2.   (continued)
Location
                                    Latitude
Degrees
Minutes
                            Longitude
                                                       Degrees
Minutes
Oklahoma, Oklahoma City
Oregon, Portland
Pennsylvania, Philadelphia
Rhode Island, Providence
South Carolina, Columbia
South Dakota, Sioux Falls
Tennessee, Nashville
Texas, Houston
Utah, Salt Lake City
Vermont, Montpelier
Virginia, Richmond
Washington, Seattle
West Virginia, Charleston
Wisconsin, Milwaukee
Wyoming, Cheyenne
35
43
39
41
34
43
36
29
40
44
37
47
38
43
41
28
39
57
50
00
32
09
46
46
16
32
36
21
02
08
97
70
75
71 ,
81
96
86
95
111
72
77
122
81
87
104
32
17
07
25
03
44
48
22
53
35
28
20
38
54
49
                                  B-13

-------
                    TABLE B-3.  MODEL PLANT PARAMETERS3
Emission
rate, kg/yr
 Release      Dia-          Exit
height, m   meter, m   velocity,  m/s
                                                                 Building
                                                                   cross
                           Exit    sectionaj
                        temp.,  ฐK  area, m b
2.13-60.6
  7.5
4.6
8.2
300
                                                                  2,240
"The area source model was used to estimate annual incidence.
 The building cross-sectional area was used to calculate the maximum
 indivdual risk.
                                   B-14

-------
          TABLE B-4.  MAXIMUM CONCENTRATIONS PREDICTED
 State
Concentration!,
 yg/rn  x  10
 Oklahoma
 Vermont
 California
 Missouri
 Kansas

 North Dakota
 Tennessee
 Idaho
 Florida
 Utah

 Nebraska
 Colorado
 Massachusetts
 South"Dakota
 Minnesota

 Texas
 New Mexico
 Wyomi ng
 Maryland
 Arizona

 Washington, D.C,
 New Hampshire
 Nevada
Michigan
 Kentucky

Washington
 Illinois
Montana
 Iowa
West Virginia

Wisconsin
New York
Pennsylvania
Rhode Island
Oregon
   0.19-5.5
   0.16-4.6
   0.16-4.5
   0.15-4.1
   0.15-4.1

   0.13-3.8
   0.13-3.8
   0.13-3.8
   0.13-3.8
   0.13-3.7

   0.12-3.4
   0.12-3.4
   0.11-3.2
   0.11-3.2
   0.11-3.0

   0.11-3.0
   0.11-3.0
   0.10-2.9
   0.10-2.8
   0.10-2.8

   0.10-2.8
   0.09-2.6
   0.09-2.5
   0.09-2.5
   0.09-2.5

   0.09-2.5
   0.08-2.3
   0.08-2.3
   0.08-2.3
   0.08-2.2

  0.07-2.1
  0.07-2.0
  0.07-2.0
  0.07-2.0
  0.07-1.9
                                                   (continued)
                             B-15

-------
                    TABLE B-4.  (continued)
State
Concentration,
ug/m  x  10"
New Jersey
Arkansas
New Mexico
Pennsylvania
Louisiana

Rhode Island
Ohio
Indiana
Alabama
Massachusetts

Georgia
North Carolina
Del aware
South Carolina
Mississippi
   0.8-22.7
   0.8-22.7
   0.8-22.7
   0.8-22.7
   0.7-20.6

   0.7-20.6
   0.7-20.6
   0.7-20.6
   0.7-20.6
   0.7-20.6

   0.7-18.6
   0.7-18.6
   0.7-18.6
   0.7-18.6
   0.6-16.5
                            8-16

-------
B.4.2  Public Exposure
     B.4.2.1  General.  The basic assumptions implicit in the methodology
are that all exposure occurs at people's residences, that people stay at
the same location for 70 years, that the ambient air concentrations and
the emissions which cause these concentrations persist for 70 years, and
that the concentrations are the same inside and outside the residences.
From this, it can be seen that public exposure is based on a hypothetical
rather than a realistic premise.  It is not known whether this results in
an overestimation or an underestimation of public exposure.
     B.4.2.2  The Public.  The following are relevant to the public as
dealt with in this analysis:
     1.  Studies show that all people are not equally susceptible to
cancer.  There is no numerical recognition of the "most susceptible"
subset of the population exposed.
     2.  Studies indicate that whether or not exposure to a particular
carcinogen results in cancer may be affected by the person's exposure to
other substances.  The public's exposure to other substances is not
numerically considered.
     3.  Some members of the public included in this analysis are likely
to be exposed to Cr"1"6 in the air in the workplace, and workplace air
concentrations of a pollutant are customarily much higher than the
concentrations found in the ambient, or public air.  Workplace exposures
are not numerically approximated.
     4.  Studies show that there is normally a long latent period between
exposure and the onset of lung cancer.  This has not been numerically
recognized.
     5.  The people dealt with in the analysis are not located by actual
residences.  As explained previously, they are "located"  in the Bureau of
Census data for 1980 by population centroids of census districts.
Further, the locations of these centroids have not been changed to reflect
the 1980 census.  The effect is that the actual  locations of residences '
with respect to the estimated ambient air concentrations  is not known and
that the relative locations used in the exposure model have changed since
the 1970 census.
                                   B-17

-------
      6.   Many  people  dealt with 1n this analysis are subject to exposure
 to  ambient  air concentrations  of Cr*6  where they travel  and shop (as in
 downtown-areas and  suburban shopping centers),  where they congregate (as
 1n  public parks,  sports  stadiums,  and  schoolyards),  and  where they work
 outside  (as mailmen,  milkmen,  and  construction  workers).  These types of
 exposures are  not numerically  dealt with.
      B.4.2.3   The Ambient  Air  Concentrations.   The following are relevant
 to  the estimated  ambient air concentrations of  Cr+B  used in this
 analysis:
      1.   Flat  terrain was  assumed  in the dispersion  model.   Concentrations
 much  higher than  those estimated would  result if emissions  impact on
 elevated  terrain  or tall buildings  near a plant.
      2.   The estimated concentrations do not account for the additive
 impact of emissions from plants  located close to one another.
      3.   The increase in concentrations that could result from
 reentrainment  of  Cr+6-bearing  dust  from areas such as city  streets, dirt
 roads, and  vacant lots is  not  considered.
      4.  Meteorological  data specific to plant  sites are  not used  in the
 dispersion model.-   As explained, HEM uses the meteorological data  from the
 STAR  station nearest  the plant site.  Site-specific meteorological data
 could result in significantly different estimates, e.g.,  the estimates of
 where the higher concentrations  occur.
      5.  With  few exceptions, the Cr+6  emission rates are based on
 engineering estimates rather than on emission tests.   See Chapter 4 for
 details.
 B.5  REFERENCES FOR APPENDIX B
1.


2.

3.
National Academy of Sciences.  "Arsenic" Committee on Medical and
Biological Effects of Environmental Pollutants, Washinqton, D.C.
1977.  Docket No. OAQPS 79-8 II-A-3.
Health Assessment Document for Chromium.
No. EPA-600/8-83-014F.  August 1984.
Publication
Environmental Cancer and Heart and Lung Disease.   Fifth Annual  Report
to Congress by the Task Force on Environmental  Cancer and Health  and
Lung Disease.  August 1982.  .
                                   B-18

-------
Mancuso, T. F.  Consideration of Chromium As An Industrial
Carcinogen.  Proceedings of the International Conference on Heavy
Metals in the Environment.  Hutchinson, T. C., ed.  Institute for
Environmental Studies.  Toronto, 1975.  pp. 343-356.
User's Manual for the Human Exposure Model (HEM).
No. EPA-450/5-86-001.  June 1986.
Publication
                              8-19

-------

-------
                     APPENDIX C.  SUMMARY OF TEST DATA

     This appendix presents the results of three EPA-conducted  tests  for
hexavalent chromium  (Cr+6) emissions from industrial process  cooling
towers  (IPCT's) and  one EPA-conducted test of Cr+s emissions  from  a
comfort cooling tower  (CCT).  Emission data from IPCT's can be  used to
represent emissions  from CCT's because design parameters  that affect  the
emission rate are similar for both tower types.  The test data  were
considered in developing emission factors and in quantifying  the
performance of high-efficiency drift eliminators (HEDE's) versus low-
efficiency drift eliminators (LEDE's) in Chapter 4.  The emission  data
include mass emissions and particle size distributions.  For  each  test
series, Section C.I  presents descriptions of the physical and operating
parameters of the cooling tower and of the water treatment program.  The
test results are tabulated in Section C.2.
     Three test methods were used to quantify emissions of Cr+s from
cooling towers, one  isokinetic method and two methods that rely on water
droplet impaction.  The EPA isokinetic test method utilizes a Method 13
sampling train with  the exception that the filter is made of Teflonฎ and a
propeller anemometer is used in place of the pitot tube.  The collected
samples were analyzed for total  chromium by Neutron Activation Analysis or
graphite furnace atomic adsorption after concentrating the liquid to
25 milliliters.  Because the isokinetic sampling probe does not alter the
airflow approaching the probe nozzle,  all  emissions are collected by this
method.  The total chromium in the cooling water was 99 percent Cr+s;
therefore, it was assumed that the total  emissions  are Cr"1"6.
     The sensitive paper test method utilizes  the collection  of water
droplets by inertial  impaction onto a chemically treated paper.   The water
droplets, which turn the paper blue,  are  examined optically with a
                                   C-l

-------
 microscope to. quantify the number of each size water droplet impacting the
 paper.   The results of these counts are totaled to quantify the emissions
 of water droplets from the cooling tower.
      The absorbent test method uses the same collection method as the
 sensitive test method except that absorbent paper (Whatman 541 filter
 paper)  replaces the chemically treated sensitive paper.  The absorbent
 papers  were analyzed for chromium by Neutron Activation Analysis or
 graphite furnace atomic adsorption to quantify emissions.
      Both the  sensitive paper method and  the absorbent paper method alter
 the airflow approaching the collection media.   Although the inertia of the
 larger  droplets approaching these devices would cause the  droplets  to
 continue in a  straight line and,  therefore,  impact on the  surface of the
 paper,  the smaller droplets tend  to follow the streamlines around the
 sampling device.   Given the typical  air velocities of the  cooling tower
 stack- and the  size of the  collection device,  less  than 50  percent of the
 droplets smaller  than 30 micrometers in diameter would impact  the surface
 of  the  paper.   Because of  this  phenomenon,  the sensitive paper  analyses
 include a correction factor for different  size droplets.
 C.I  DESCRIPTION  OF TESTS
 C.I.I   Department of Energy. Gaseous  Diffusion Plant,  Paducah.  Kentucky
      C.I.1.1   Process  Description.   The Department of  Energy facility  at
 Paducah,  Kentucky,  is  operated  by Martin Marietta  Energy Systems, Inc.
 This  facility  enriches  uranium  in the U235 isotope using a gaseous
 diffusion  (cascade)  process.  The diffusion process involves pressure-
 induced  flow of the  uranium  hexafluoride  (UF6) process gas through
microporous barriers.  The  heat of compression is removed from the process
gas by  thermosyphon  refrigerant systems to control the operating
temperature.  The refrigerant is vaporized in process gas coolers and  is
transferred to water-cooled heat exchangers where it is condensed before
 it returns to the gas coolers.  Recirculating cooling water is pumped from
a basin  to the process condensers and- returned to the cooling towers where
waste process heat  is rejected to the atmosphere.  Indirect cooling  of the
UF6 is used for safety and reliability considerations.
     The process cooling tower system consists of two towers that are
designated C-637-2A and C-637-2B.   A sketch of the C-637-2A and C-637-2B
                                   C-2

-------
 system  is  shown  in  Figure  C-l.   The C-637-2A tower was selected for source
 testing; this  tower is  a seven-cell Marley crossflow design with two fans
 per  riser  cell and  is equipped  with both LEDE's (herringbone) and HEDE's
 (Thermatec Spectra).  Riser  cell  Nos.  1  through 5 are equipped with LEDE's
 and  redwood splash  fill.   The HEDE  riser cell  Nos.  6 and 7 contain
 polyvinyl  chloride  (PVC) splash fill.
     The tower was  originally constructed in the early 1950's with redwood
 splash  fill  and  herringbone  drift eliminators  in all  the riser cells.
 Riser cell  Nos.  6 and 7 were recently  modified  by the installation of  the
 PVC  splash fill  and Thermatec Spectra  drift  eliminators.   The water
 systems of towers C-637-2A and  C-637-2B  are  served  by a common pumphouse
 that has a total nominal capacity of 605,670 liters per minute (i/min)
 (160,000-geitlons per minute  [gal/min]):   six pumps  rated  for  75,709 fc/min
 (20,000 gal/min) each and  four  pumps rated for  37,854 Ji/min
 (10,000 gal/min) each.  Each of the tower systems  is  constructed  with  a
 water basin  having  a capacity of  15.9  million liters  (4.2  million
 gallons).  Makeup water from the  Ohio  River  is  softened and clarified  and
 then supplied through a 76.2-centimeter  (30-inch) pipeline to  the
 pumphouse.
     Two 152.4-centimeter  (60-inch) cooling  water supply and return  loops
 ("G" and "H" on Figure C-l) are used to recirculate the tower water
 through the  process building.  The  return lines of each loop are  connected
 by a "crossover" pipeline that allows water  to be directed to either the
 C-637-2A or  2B tower for cooling.  Another "crossover" pipeline
 interconnects the process cooling water supply lines.  The recirculating
water enters the tower after the flow is spli't into seven branches  (riser
 pipes)  that  serve each of the seven riser cells.  The flow frqm each of
 the riser pipes is split and conveyed into the water distribution decks of
each of the  two fan cells.
     The water distribution decks are located directly above the splash
fill sections of the fan cells  and equipped with gravity flow nozzles for
even distribution of the recirculating  water in a cascade over the fill
material.   Propeller fans measuring  6.7 meters (m) (22 feet [ft])  in
diameter that are located  in the stack  of each cell  provide 17,273 cubic
meters  per min (m3/min)  (610,000 cubic  feet per min [ft3/min])  of induced
horizontal  airflow through  the  fill  sections.

                                    C-3

-------
       Sodium bichromate with a target concentration of 18 to 20 parts per
  million (ppm) is added to the recirculating cooling water to inhibit
  corrosion in the heat exchangers.  Chromate additions are made manually,
  and the chromate levels are measured dail>.  A chlorine residual of
  0.5 ppm is the target concentration for providing control of biological
  organism levels in the recirculating water.  Chlorine is continuously
  Injected into the system at a constant flow rate.  The pH of the water is
  monitored continuously by a pH probe and meter.   Additions of sulfuric
  acid are controlled manually to maintain the 6.0 to 6.1  target pH range.
  The calcium hardness is maintained at concentrations between 350 and
  500 ppm 1n the recirculating water by controlling the blowdown rate.
       C.I.1.2  Operating Conditions During Testing.   The  C-637-2A cooling
—tower operating parameters  that were monitored throughout the  test period
  were the fan motor amperage,  pump outlet pressures,  total  water flow,
  basin water temperature,  return water temperature,  chlorine  addition rate,
  makeup water flow rate,  pH,  wet well  temperature, and  blowdown  rate.
  Meteorological  data were  obtained from the  National Weather  Service  (NWS)
  at  the Paducah  Airport  for each  day  that  tests were performed and  included
  hourly observations  of  dry bulb  temperature, dew point, wind speed,  and
  wind  direction.   Table  C-l is  a  summary of  the cooling tower operating
  parameters  and meteorological  data recorded and obtained during the  test
  period.
      The cooling  tower was not operating at the recirculating water  design
  capacity during the tests due to  low process cooling demands.  It was
  necessary to  increase the water flow rates of the riser cells being tested
 to between 90 and 100 percent of design capacity (30,564 to 33,959 a/min
  [8,074 to 8,971 gal/min], respectively) by directing some of the
 recirculating water in the riser cells not being  tested to the riser cells
 that were being tested.  This was accomplished by partially closing the
 isolation valve for the riser cells not being tested.   Additionally,  the
 distribution of the riser cell water to each of  the  fan cells was  balanced
 by adjusting the individual  flow control  valves on each fan cell  until  the
 depth of water appeared to be equal  in the distribution decks.   The
 blowdown rate was maintained  at zero throughout the  test  period  to
 minimize the loss of sodium  bromide that  was added to  the  recirculating
 water as a tracer chemical.
                                    C-4

-------
      On the day prior to the first test series, the ^circulating water
 flow rates on riser cell Nos.  4 and 7 were adjusted while the waterflow
 rates were measured.  Waterflow rates were' established at 32,176 2,/min
 (8,500 gal/min) and 32,555 a/min (8,600 gal/min)  for riser cell Nos. 4 and
 7,  respectively.   A waterflow  measurement on riser pipe No. 7-concurrent
 with the first test series indicated that the flow was at 85 percent of
 capacity or 28,390 a/min (7,500 gal/min).  The reason for this variation
 is  not known,  but  there  may have been a leak in the pitot tube during the
 pretest flow rate  measurement.   Inspection of the  drift eliminator in fan
 cell  No.  13 indicated the  presence of a significant water leak from the
 distribution deck  into the tower on the fan side of the drift eliminator
 section.   The  first test on riser cell  No.  7 was invalidated because the
 waterflow rate was less  than 90 percent of  the design flow rate and
 because of the water leak  into  the tower on the fan side of the drift
 eliminator.  The tests on  riser cell  No.  7  were successfully repeated
 after  the pitot tube was repaired  and  a broken redwood  plank in the side
 of  the water distribution  deck  was  replaced.   The  remaining  tests  on riser
 cell Nos.  4, 5, and  6 were  completed under  acceptable conditions with
 respect to  the  test  plan and Cooling Tower  Institute guidelines.
 C.I.2   National Bureau of Standards. Gaithersburq. Maryland
     C.I.2.1   Process  Description
     The  National  Bureau of Standards  (NBS)  is a Federal government
 research  facility  near Gaithersburg, Maryland.  On the grounds are  seven
 laboratory/office  buildings with a total floor area of 58,066 square
 meters  (m2)  (625,000  square feet  [ft2]) and a number of support buildings
 with a  floor area of  62,711 m2   (675,000 ft2).  Comfort cooling and cooling
 for laboratory  processes (lasers, ovens, etc.) are  both provided by a
 four-cell Marley tower located  near the western boundary of the
 facility.  The tower was installed in the early 1960's.
     A  sketch of the cooling tower system is provided in Figure C-2.  The
 tower  is a crossflow design with redwood splash fill and one fan per
cell.   Propeller fans measuring 6.7 m (22 ft) in diameter are located in
the stack of each cell.  In 1985, the tower was retrofitted with
high-efficiency Munters D-15 drift eliminators.
                                   C-5

-------
      The capacity of the water basin is about 1.893xl06 a, (500,000 gal).
 Four pumps each rated for 33,312 ซ,/min (8,800 gal/mi n) circulate the water
 to the chillers.  The water from the chillers is combined and returned to
 the tower through a 106.7-centimeter (42-inch) riser pipe.  />bove the
 tower, the flow is split into four branches and distributed to each of the
 cells.  The water distribution decks are located directly above the fill
 and are equipped with gravity flow nozzles.  In winter, heated water is
 sprayed up into the rear of the tower to prevent icing conditions.
      A solution of molybdate and polyacrylate is used to inhibit corrosion
 in the heat exchangers.   The target concentration of molybdate in the
 recirculating water is about 15 ppm.   Conductivity and pH are monitored
 continuously, and blowdown  occurs automatically when the conductivity
 reaches 1,800 micromhos  (ymhos).   Blowdown  averages about 227,126 liters
 per day (a/d) (60,000 gallons  per day [gal/d])  in summer and  about
 7,571 a/d  (2,000 gal/d)  in  winter.
      Makeup water is  provided  by  the  City of  Gaithersburg.  The
 conductivity is  generally about  300 umhos,  but  after heavy  rains  and  after
 salt has been applied  to  the roads  in the winter,  the  conductivity
 increases.   Makeup  requirements average  about  1.136xl06  a/d
 (300,000 gal/d)  in  summer and  about 208,200 a/d  (55,000  gal/d) in
 winter.  Most of the water  has first  been used for  once-through cooling  of
 oil  and air  compressors.
      Biological  growth is controlled  by manually  adding  24.6  a (6.5 gal)
 of  a solution containing disodium cyanodithiocarbamate  (7.35  percent) and
 potassium methyldithiocarbamate (10.15 percent) once a week.
      C.I.2.2  Operating Conditions  During Testing.  Eight test series were
 conducted.  The  cooling tower operating parameters  that were monitored
 during each test series included the recirculating water temperatures into
 and out of the chiller, recirculating water flow rate, daily blowdown and
water makeup, wind speed, and wind direction.   Meteorological  data were
 also obtained from the NWS at Washington National Airport.
     The design water flow was achieved on each of the test days, but one
chiller was not operated; water simply circulated through it.   The low
ambient temperature and low demand during test series 5, 6,  and 7
necessitated turning off  a second chiller and  one fan.   Table  C-2 is a
                                   C-6

-------
 summary of the cooling tower operating parameters and meteorological data
 recorded during the test period.
      It was determined from the estimated system volume that the addition
 of about 90.7 kilograms (kg) (200 pounds [lb]) of crystalline sodium
 dichromate would result in a Cr"1"6 concentration of slightly over 15 pprr in
 the recirculating water.  This amount of sodium dichromate was added on
 the day before the first test, and lesser amounts were added on following
 days to replenish the estimated losses via blowdown and drift.  To
 determine the actual Cr+s concentration, water samples were taken during
 each test series and later analyzed for Cr+6.  Sodium bromide also was
 added to the recirculating water for evaluation of bromide as a surrogate
 for chromium in drift emissions testing.
      A pretest walk-through of the tower was conducted on Tuesday,
 August 19.   Inspection of the drift eliminators revealed a number of water
 leaks into  the fan side of the eliminator sections.   This was most
 significant in the first cell,  but in no case did  the airflow appear to be
 shearing droplets away from the water stream.  Inspection of the  water
 flow along  the outside of the tower revealed an unequal  distribution that
 was most pronounced on the windiest days.  The strongest winds were
•evident on  Wednesday,  August 20, when the anemometer  mounted atop a  nearby
 building indicated gusts of up  to  22.5 kilometers  per hour (km/h)
 (14 miles per hour [mph]).   On  the tower  itself, an anemometer indicated
 22.5 km/h (14 mph),  and the NWS reported winds  of  16.1 to  24.1 km/h  (10 to
 15  mph)  for that  day.   In  no instance,  however, was drift  observed from
 the sides of the  tower.  All  tests were completed under  acceptable
 conditions  with respect  to  the  test plan and  Cooling  Tower  Institute
 guidelines.
 C-1-3   Exxon  Refinery,  Ethylene Production, Baytown, Texas
     C.I.3.1   Process  Description.  Tower No. 68 provides cooling for the
 catalytic light end units, which recover ethylene and other  light end
 products.   The tower handles a constant heat  load 24 hours per day.
 Figure C-3  is  a sketch of tower No. 68.  This tower consists of four
 counterflow cells  and one Marley crossflow cell.  Each cell has one
 single-speed fan and redwood herringbone drift eliminators.  The
 counterflow section has redwood splash fill  and is served by two  risers
                                    C-7

-------
 that distribute the water over the fill through a manifold and pressure
 spray nozzles.  The crossflow section has plastic splash fill and is
 served by one riser that supplies a water distribution deck equipped with
 gravity flow nozzles.   Two pumps circulate water from the northern end of
 the common basin to the process heat exchangers, and a third pump is on
 standby.  Slowdown is  withdrawn from the system before the water is
 returned to the tower.  Makeup water from the San Jacinto River is
 supplied through a 10.2-centimeter (4-inch)  pipeline to the basin.  The
 fans are 5.5 m (18 ft) in diameter in the counterflow cells and 7.3  m
 (24 ft)  in diameter in the crossflow cell.
      The corrosion inhibitor is a chromate/zinc formulation that is
 supplied by Betz.   The target concentration  of chromate in the
 recirculating water is 10 to 15 ppm.   The solution is added automatically
 at  a rate that is  set  manually.   Dispersant  is added in the same manner.
 A free chlorine residual  of 0.2 to 0.5 ppm is  the  target for control  of
 microbiological  growth.   Chlorine gas  is  injected  into  a side stream  of
 the makeup water and added  to the southern end of  the basin.   The pH  of
 the water 1s  monitored continuously, but  it  is not used  as  an automatic
 controller.   When  pH exceeds  the critical  control  range  of  6.0 to 9.0,  it
 must be  corrected  by manually adding acid  or caustic  soda.   Slowdown  is
 dictated  by the  conductivity,  which should not  exceed 1,500 umhos.
      C.I.3.2   Operating Conditions During Testing.  The operating
 parameters  that  were monitored throughout the  test period included fan
 motor amperage,  pump outlet pressure,  hot water line pressure, water flow
 in  each riser, temperature  in  each riser, basin water temperature, pH,
 conductivity, wind speed and direction, wet bulb temperature, and dry bulb
 temperature.   In addition, the makeup  flow rate was measured and the
 blowdown was estimated concurrently with the fourth test series.
Table C-3  is a summary of the cooling tower operating parameters and
meteorological data recorded during the test  period.
     On the day prior to the first test, the  recirculating water flow
rates were measured.  The flow in the crossflow cell  was about 20 percent
greater than the flow in each of the counterflow cells.   However, because
the pump outlet pressures and fan amperages were constant and within
design specifications,  no changes were made to the air or water flow  rates
for the test.
                                   C-8

-------
      The drift eliminator on one side of the crossflow cell was determined
 to be in good  condition based on the visual  inspection.  The drift
 eliminators  in the counterflow cells could not be inspected.  However, it
 appeared that  a similar quantity of drift was emitted from each stack
 although the amount may have been slightly less from cell  No.  1.  The
 quantity of  steam  rising from cell  No.  1 also appeared to  be slightly less
 than  that from the other cells.   Some of the nozzles in the distribution
 deck  on  cell No. 5 were plugged,  and a  few of the redwood  slats in the
 lower sections of  the counterflow cells were broken; but the overall
 condition of the tower  was reasonably good.
      Water meters  are not  installed  on  the makeup and  total  blowdown
 lines.   To estimate these flows,  alternative methods were  attempted.
 During the fourth  test  series, a  meter  was connected to the  prlfsure"taps
 on an existing orifice  plate  in the  makeup line.   This  indicated an
 average  flow of about 1,060  a/min (280  gal/min) over the 6  hours of
 monitoring (greater in  the afternoon  than  in  the  morning) but did  not
 include  the  56.8 to 75.1 a/min (15 to 20 gal/min)  diverted for  chlorine
 injection  or the amount  leaking through a  valve into the system from a
 nearby tower (No.  58), which  is treated with  a phosphate inhibitor from
 Calgon.  The Betz  representative  used the  phosphate concentration  in the
 recirculating  water of tower No.  68 to calculate  a gain of about
 94.6  a/min (25  gal/min).  Later work by Exxon confirmed that this estimate
 was correct.
     To estimate the tower No. 68 blowdown, the flow was diverted to a
 208-liter  (55-gallon) drum.  The amount of time required to fill the drum
 a  couple of times was recorded.  This estimated flow rate was within
 20 percent of the estimate calculated by the Betz representative based on
 cycles of concentration and an estimate of evaporation.
     Water temperatures also  are not monitored by online equipment.
Therefore, fittings were attached to taps on the three risers and the  hot
water return line itself.  Mercury-in-glass thermometers were used  to
record the temperature.   The  basin temperature was determined about 5  feet
from the basin  wall below cell Nos.  1, 2, and 5.  A mercury-in-glass
thermometer was placed in a perforated"can that-was attached to  a length
of conduit.  With this method, it was not possible to determine  the actual
                                   C-9

-------
 temperature drop in each cell, but the average basin temperature  in all
 three locations was the same.
      Two sources of meteorological data were available:  one station set
 up at the tower and one maintained by Exxon refinery personnel less than a
 mile from the tower.  Both stations indicated that the wind direction was
 from the southeast, and very few directional changes deviated more than
 45 degrees from the southeast.  Both average and peak wind speeds,
 however, were considerably higher at the tower station.  The differences
 may have been the result of instrument calibration differences or they may
 have been caused by a slight tunneling effect created at the tower station
 where the wind had to pass between the cooling tower and a cryogenic
 process  column (and other shorter equipment) 27.4 to 36.6 m (30 to
 40 yards)  downwind of the station.  Gusts  rarely exceeded- 24-.1 km/h
 (15 mph),  and drift was never visible  from the sides of the crossflow
 tower.   The ambient temperature  also varied between the stations.   The
 actual temperature is probably that obtained at the tower site since the
 several  thermometers that were used recorded the  same  levels.
      Three days  prior to  the first test, the Exxon  process  personnel
 responsible for  the tower disconnected the  chlorine injection  line to
 preclude any possible adverse  health effects  on test personnel.  Chlorine
 will  also  react  with most  hydrocarbons.  Thus,  a decrease in the free
 chlorine residual  concentration  (normally determined once per shift) is
 the best indicator  of  a process fluid leak  into the water.  Alternatively,
 gas traps  on  the hot water return  line, visual  inspection of the surface
 of  the water  in  the  basin and  the  distribution deck of cell No. 5, and the
 chromate concentration were used to confirm that the process heat
 exchangers were  not  leaking.  The  chromate concentration, as determined by
 the operators each shift, was essentially constant and within the desired
 control range during the testing period.  The Betz analysis on Tuesday
 agreed with that of the operators.  The pH and conductivity were also
within control ranges.
 C.I.4  Exxon Refinery, Lube Oil Production, Baytown, Texas
     C.I.4.1  Process Description.  Cooling for the vacuum distillation
unit for lube oil is provided by  tower  No.  84.  Although the tower  is
operating at less than design capacity,  it  handles a constant heat  load
                                   C-10

-------
 24 hours per day.   Figure C-4 is a sketch of tower No. 84.  The tower is a
 Marley counterflow design with four riser cells and four fan cells.  Each
 fan cell has one 6.7-m (22-ft) diameter constant-speed fan.  The average
 measured airflow per fan ranged from 222 to 287 dry standard cubic meters
 per minute (dsm3/min)  (470,000 to 609,000 dry standard cubic feet per
 minute [dscfm]).   Each cell  is equipped with PVC film fill and a high-
 efficiency Marley  XCEL-15 drift eliminator.  Water is distributed over the
 fill  through a manifold and  spray nozzles.   Two pumps circulate the water
 from the basin extension at  the south end of the tower through the process
 heat exchangers.   A recent potassium retention time study determined that
 the system volume  was  about  2.082xl06 i (550,000 gal)  of  water.
      Slowdown is designed to be controlled  by the conductivity of the
 recirculating water.   At certain set points,  a valve  is actuated in a line
 off the  main hot water return.   Most of the makeup water  is  supplied
 through  a 15.2-centimeter (6-inch)  pipe to  the basin  extension,  but part
 of  it is diverted  continuously  into  five  smaller lines.   The  inhibitor,
 dispersant,  chlorine,  sulfuric  acid,  and  caustic soda  are injected  into
 the smaller  lines  automatically.
      The corrosion  inhibitor  is  a chromate/zinc  formulation in a  7:1  ratio
 that  is  supplied by Nalco.  The  target  chromate  concentration  in  the
 recirculating water is 8  to 12 ppm.  The  solution  is injected  into  one of
 the small makeup lines for a  specific fraction of every 10-minute
 interval.  The on/off  time fraction can be changed by entering new  values
 into  the computer memory.  The dispersant is injected into another makeup
 line  in  an identical manner.   Acid and caustic are injected based on pH
 set points within the control range of 6.8 to 7.5.  Chlorine gas is
 injected  continuously at a rate controlled by a free chlorine residual
monitor  that is generally set to keep the concentration in the range of
0.3 to 0.5 ppm.  The conductivity of the makeup water is about 150 umhos,
and the  control range for the number of cycles is 6 to 8.
     C.I.4.2  Operating Conditions During Testing.  The operating
parameters monitored throughout the test period were fan motor amperage,
pump outlet pressures,  cold water line pressure, water flow in each riser,
temperature in three of the risers,  basin temperature,  temperature in pump
inlet lines, pH, conductivity, wind  speed and direction, and  dry  bulb
                                   C-ll

-------
 temperature.  The computerized system that monitors inlet and outlet
 temperatures and the makeup, blowdown, and recirculating water flow rates
 was not calibrated correctly at the start of the test.  With the exception
 of the blowdown, attempts at calibration were not successful.  These
 problems are not considered to affect the amount of drift, and only the
 makeup and blowdown could not be monitored directly by the test
 personnel.  Table C-4 is a summary of the cooling tower operating
 parameters and meteorological data recorded during the test period.
      On the day prior to the first test series,  the water flow rates in
 each riser were measured.   The flows  in Risers A and B were about
 15 percent less than the flows in Risers C and D.   The total  flow was
 25 percent greater than  the tower design and 20  percent greater than the
 pump ratings.   From the  pump head pressure and the manufacturer's pump
 curves,  it was calculated  that the flow should be  about 77,980 z/min
 (20,600  gal/min).   The measured  rate  was about 10  percent greater than
 this calculated rate.  As  scale  and fouling increase,  and with additional
 process  heat loads,  the  head  pressure will  increase  slightly  and  cause a
 decrease in the flow rate.  The  conditions  as measured  (and with  all the
 fans running)  represented  normal  operation.  Therefore,  no attempt  was
 made to  equalize the flow  in  the  risers  or  to reduce the  overall  flow  to
 the  design  rate.
      The drift  eliminators could  be inspected through a porthole  in the
 fan  stack below the  fan.  The drift eliminator in Cell A  is assumed to
 have at  least one defect because entrained droplets were observed
 periodically in the  same area of the  stack.  The other drift eliminators
 appeared to be  in good condition.  The water distribution through the fill
was  even although it did cascade along some vertical beams at a greater
rate  than along others.
     The quantity of blowdown was not easily determined because the
conductivity control was not working and the valves in the line were
closed.  Also,  recirculating water can be withdrawn from the system in  the
process area for general  ground cleaning purposes.   The operators,
however, indicated that they had not been using any of this water on the
test days.  Finally, a water balance on the process side of the overhead
vacuum condensers indicated an excess  of about 189  a/min (50  gal/min).
                                   C-12

-------
This is approximately the amount that the Nalco representative calculated
for the blowdown based on the cycles of concentration and an estimate of
the evaporation loss.
     The rec^'rculating water temperature was measured with mercury-in-
glass thermometers in fittings attached to taps in three of the risers.
The basin temperature was determined with a mercury-in-glass thermometer
at the intersection of the main basin and the basin extension.  The
temperatures indicated by gauges on the lines to the pumps were also
recorded; they were always 2 degrees lower than the thermometer reading.
     Meteorological data were available both at the tower site and from
the Exxon meteorological station almost a mile away.  The wind direction
continued to be steady from the southeast, and the wind speeds were higher
on the chart recorder at the tower station.  At this site, there were no
obstructions around the station except for the tower itself.
     The operator log of the chromate concentration in the recirculating
water was constant at the upper limit of the control range over the 2-day
test period.  The concentration agreed with that obtained by the Nalco
representative on August 29.  The pH, conductivity, and free chlorine
residual were also within the control ranges.
C.2  SUMMARY OF TEST DATA
     The results of the EPA isokinetic and the absorbent paper emission
tests at the Department of Energy, Gaseous Diffusion Plant in Paducah,
Kentucky, are summarized in Table C-5.  For the tower tested, each riser
supplies water to two fan cells.  Stack emissions were sampled from fan
cell Nos. 7 through 14 (riser cell Nos. 4 through 7).  For most tests,
half of the sample was collected from each of the fan stacks corresponding
to a riser cell.  All data for the isokinetic emission tests are reported
in Table C-5 as being greater than the value presented because only about
one third of the chromium was transferred with the liquid to the vial  used
for analysis after the concentration of the sample.  The balance of the
chromium remained in the beaker used to evaporate the water from the
sample and required rinsing with aqua regia to solubilize the chromium for
analysis.  This method of rinsing was not accomplished with the beakers
used to concentrate samples Collected at the Department of Energy, Gaseous
Diffusion Plant in Paducah, Kentucky.
                                   C-13

-------
     The results of the Method 13 and absorbent paper emission tests at
NBS in Gaithersburg, Maryland, are summarized in Table C-6.
     The results of the Method 13 and absorbent paper emission tests at
Tower 68 at the Exxon refinery in Baytown, Texas, are summarized in
Table C-7.  Although there are three riser cells and five fan stacks,
Individual tests were conducted on each fan stack.
     The results of the Method 13, absorbent paper, and ion exchange
emission tests at Tower 84 at the Exxon refinery in Baytown, Texas, are
summarized in Table C-8.
     Sensitive paper drift measurements at all  four test sites are
summarized in Table C-9.
                                  C-14

-------
s
Z
UJ
i
u.
3=
a
1/1

= | 3 ซซ.ซ.ซ ฃ ฃ 3
ซ , 5 2 || ง ง5S
I ซ S | ซ ป 5
aj i, Vdป
0*2"!
s *
• ;<<-ป ^ O
5 — u i *
s g s a ซ 3 ^
0- -- — >-u 39 •
* < '-*• "" V Z CM
uj w a < — 5 ,
* i * 3 Z " S
i -j -j cc 3 ^ \& •
I 5 I 5 o o cj,
- < i= S z z

_
aฑ
O

v^J

Q.
OB



-i
•-^
•H^
— r
(ฃ}


a.
3
Q.
,

o!
- — -i
"--- r
^^ I
X^N
v^^ '
n !
u!

Q. Qฃ
• 0

ce
UJ
1



,
;o
;o

v=x


Jj
=9
>

o
o
p]
p














•••














r
I





•^
CM
-*-
<
-*-
VJ3








—*•
—*-
-^r-
— ^




_
o
z
-u
=3 S






t














Pvj
J^
'— J






















^>
ฃ





.^
u
Qฃ
Qฃ
















j*
l^




.

o





ac
UJ
s
I
z
o;
^
^ '






s-O
"^
\













LU
O
ง
u
>•
a.
^



1

!

J


a.
3 a:
UJ UJ
^ 3





mnia ,,H,,




















c
ous Diffusion P
QJ
ซS
CD
>,
O)
c
LU
O
O j_ป
z j;
o 3^
i 1
vi. Q.
tf> 
-------
                                                 id
                                                 c/i


                                                 01
                                                 A3

                                                 CD
                                                 GO
                                                 •W

                                                 id
                                                 
C-16

-------
01   uj
          C3 U_  ซ   IO Qฃ
          <   3 O1   UJ O1
          „ U. O -J -J (— _l _I
          oeo—i—i—iuj_i_i
          LU   l^ LiJ L^kl 3P I^J i^
          2  •OCCJCJ
-------
           r-.


           CM O
                CS4

                CM
ฃ
00
            •S ฃ
            (Q *•

            WO
  UJ QC I—



Ss3a   3
UJ CD U. U. OO LU
>-Z     0ฃ0
  •-. a: u. LU

uj  "—i
i— ae < o
                                                                           a>
                                                                            c



                                                                           •M
                                                                           CD
                                                                            I



                                                                           I

                                                                           X
                                                                           •tf
                                                                           OO
                                                                           CJ
                                                                           (U
                                                 ^
                                   C-18

-------

^
LU
OS ฐ
\ND METEOROLOGICAL DATA
SION PLANT, PADUCAH, KE
SUMMARY OF OPERATING PARAMETERS i
EPARTMENT OF ENERGY, GASEOUS DIFFU
T— 1
1
UJ
CO
I—
C CO
Of
*J 3
l/t
H-
8
i_ r-.
Of
o
•*-• a:
8
01
o
*•ป 3B
S
V
i"
Of
H-
Of
V
wป
Of
U 
fc-
o
t.
01
1
rtl
Q_
QQ 1 f*. ^? e? *"""* f*l *"" * *O ^— •• .^^ .— * ^^
'x*>^ c*i >•*• to W i i i CM rt S to S S !^ i^ ฃ2 Si! i(1** 5? 'T* *^ 5S *^ j5ซ to CM — * ^ to irt
&<•> ~S.a2 SS ^cn""tS ^ Lrt 1 I S ^ f^ . J^_ป 7^" jr~ ^^.- *T* *"^
^•^*^^"^ CH ^ซ *o en t to  CM CO CM CM CM  V **> cd —I i
ปO CO CM U>
ฃs • - ~* ซn en ^r ^^COCMCO r^Sto *•* <\7 ***• en m rp g m •— 'ซi ป—cn en
in *7* in ป-^ ri
•sj. *-*— ป ปo5?^Si 98S^! i I Jo S Si JS i ! *28 *7" - **inr^Siu>Sir^!ฃ*|
CM^C 0lPS;<>J.el OCMซJtO ^^ 1 *t CM l-i. ซ IS." -J 1 10 2 ^J S Q ?
tO 1 ''^ *"•* O W "•• PM
t
Jfl" . 	 „ (wj" r"" ^. ^ ^, rn>— ' CM-^— ^.^ OT, — -TT- — • rปJ CVj .30 2-
1 ' t^l \Nj Lfl ,•*!
3^ -— -
j| 1 - ^ ~
l-si s| ' . t- I 1-
35 s ฃ * =
•Si g • fc a2,= J= = c
-•2g. ง., ^ ^ ^ I^cocno-^c^^ ^ "^ * ฐ- ^ = = =
"ป 'a.S^^ui.t.i. S — "" ~" "" "" 3 5 ?S=zt0?'?'?!
ฐ yซ. a V 11 
-------
                                 cvjcg i


                                   o"
                     r*t ft ^?
             O



             O
"O
 oป
 I
CJ
UJ
      ~3
      fe-
            .e

           oป
                     S3
                    to c
                    0
                    s   ฐ
                      I  
— • cn CM

cn c*To r>s. cTTco en*
CD T* CM -* I CM T*
CM cn i cn ^ CM CM
I •— *O i CO 1 r*
to — * cn —CM — •
. CM . .
•T CO CM
•— CM CM


cn CM*O cn c\Tco en*
cn 7 2 jn ? CM 7
cn "^ '10 co ซ-^^ซ C^
CM (Nl ^
•^ cn CM



o> co o ซ-4 m m tn*
• I CO • CO • IO
CM tO ^* *^ 1 CO 1
-ซป— i cncM;=4 ซr
i es ป co i to
cn r*T i*C
CM --



cn aTo cn -?to o"
• i en • co • (O
CM to o CO i tn i
r*. in CM -— O •*-
* ฐa a


^cfgCMgog-
2S2.'7>Rjcoi2(i
• o • •
CM in cn
•-ซ CM —4
tO oTo CM O*— O*
• ' g • 01 . f*ป
*-.-*-' i m co CM to
i o i co i ซo


^- ^M CO
en ^i

^ ซ ^s
. i S *ฐ. 5 "^ S
'•M in o o ' — i

co F> en 2,*r S-
CO ^33 Q{

,^
^3 D
a. -n • i.


03 ^
J ง b 1
T," 5 I "
a ,- H?.5 ?
wป T3 c — ' g — '
at a.


5 5.5  **-
0 0
•fo 4^
a e
|Q U
.rfe
4^ a.
Is
2 tl
1^
(A
C 01
(U (U
fef
V* "O
^1

S|
•5 4J
O) J=


U U
3 tt

Of Of
> -C
4rf
a/ -a
j<

a* ^
'i t
s *
O 4^
"^ 5
01 —
ซJ OJ
•^ a.
^ ซ.
i'
— 0
.a
S *O
^3



J3 ' ~
                                                             C-20

-------





J-
oo
oo
UJ
^

C3
1—4
a:
0
sl

™" ^^ป
_l Gฃ
ii
2 <2
(— ^
si t
szs
O
in in
to ro



^
o 10
O oo
r- to
r- vo
10 ro



งfO
oo
r- 10
r- vo


^
o oo
O CN
O> t
r-*io*
IO fO

O CN



sง


^
O 03
O O
VO vo"
to ro





c.

Recirculating
water flow
r i sen , d/m i
(gal/min)
ฃ
o> '
*

~~
*
^
.?
*
—


.M-
ฃi ฃ
O) O
.C —



*ป^
O) O
IO


ฃ
O)

f
13)
JZ
—
.. <
ฃ
O)
5

c
0)
•sr

•^
<





tn
U.
1
si
— in
^— ^^


88
j^ "*




||
o -^



ll
SK



8ฐ
0*0*
ON m

2o
go
0,0,,
งs



si
" .
10 10*



ll
!o2







Blowdown, d/c
(gal/d)
__^
งo
o
VOCM*
"- lO
VO —


si
VOCM*
IO -ป




go
o o
VOQ
r- CM
^^


oo
8,1
vo o
•"


If
vo!n
ON CM

งo
O
10*^*
voin
ON CM



^r IQ
f** to
_^*-'


1ง
O \o
KS
" *~'
V—





•o
erf ^.
13
Q.—
3 IO
•••
0
c
O
IO

**-
0
c
0
IO

o
c.
o

to




in
 i oo
O —OO *-*
in rป
CM CM

• r^ ซco
in i r- i
CM in CM o
i r— i oo
ON >-r- ^
IO vO
CM CM


— O\T >f\
• r** • oo
tO 1 Ol 1
CM 00 CN T
I r- I 03
IO —ON ^
* •
in oo
CM CM

VO OOOl ?
in I co i
CM rซ- CM to
1 I^ 1 03
O —JO *-*
in 03
CM CM

P- O O vo
• co • oo
VO 1 O 1
CM ON IO in
i rป i oo
VO ON
CM CM


in I ON i
CM in CM —
| p*. | ฃQ
Ol ^CM —
to • P^
CM CM


 1 >
 *D
V- -H T3 IO T3
10 O C (0
•H a LU _J
IO

to o o
00 ^0



. in ^^
03 pป
— *ฐ




! |S



1 go
1 O —
— *ฐ



1 g in
1 O •
"•„ป
VO
CM

1 O in
1 S *
to
^
CM

! 82
—


I O o
r-. ^



01
O
^1


Water chemisti
PH
Conduct! vit'
Biocide add
a (gai)

O in o
• I 10
oo o to
I — I
0 0
CM


r- — o
p^T to
T 2 '
0 ~ฐ
CM



O in o
• I 01
00 O 1



O in o
• i
CO O
0 *~



r- — in
r^"" ฐ
72o"
O en

CM 1 1
72!?
o




-------













ป
UJ ^
*jgr


So:
— 4
{iJo:
UJ -j.
*r^ C3
7*S X
2r x
2* LiJ

งงฃ
1— ฐฐ
^2
<^0
is
3ง: i— i
* CD
*** ^^
tJ> H-4
. Z3
LxJ ^^
CQ
<
I—











CN
•
jj

U)
*""
o
U)
4.
tf>
,o
H*




•-
ฃ
UJ
CD
"c
(D
in

in
ฃ
r~*





+.
in
ฉ
+r
t_
a.






















D
Ii
3
3

 o> CN oo oo CN
oo eo r^













inOOOOO O O 03
09 at iป at CN eo oo CN
























O)
at
O) o.^
at •>•
a, u — '

• 3O
 a.33 0)
OOOOO EQ-Q.-t--J--
3 OK]
a. xs
^^ *^ ^^
ininincN~l !o I^-I-'CNCN
S2S77vlT 2S- '-"So""
i v i in o —o i — —
CN • fO . o O l~- I I
oo at oo at •-• ^^ oo ot o
•ป-ปCN ~^Ot *-* *^ • CN O
*" ro **" ฐซ ฐ
• *O> • VI • — —
at at >oo oo
CN CNOO ro ro
1 1 ro 1 1
09 rO 1 09 09
• • in • •
r^ GO • f*ป r*ป
CN CN rป- ro ro
ro












—
^^
C
ฃ t.
— •+—
to ._
O) c

c c
•— •— aj
e e c
\ -x.-. u> -ft
o? — — O O g
(O | ^ il Q.
U. * O>c S So.
X ^0 3-ui =L
* — C >ป "^i "* /t^ {=
4) — •- t. >.>. ฃc a.
C S -^ -H— 4- .— Q.
jn ^- — o o?._ > c '> o •
\\ V E — •  •*- +_j34_
C C C-H U L.U * |"g || 1^1
"Jซ"J*ซ>ซ'ซ'3T)(_ CIO ? a. c3u- o
X i CQ3 O

$
3
!
^
Q
o
%"•ป

















































C-22

-------





























^•^
•o
O)
3
C

t ^
"^
0
>-*

CO
1
*J
LU
_J
ซC





























CM


in
t_
in

+™

t_









—
O
z
in
10
i

in
•2









4.
ID
4"
i

































'arameter
*•*•
^** .— . it
ON Q O ^ O IO ON O O C ^ ^ O VO
•2*CO^ CM* • CM 3 ป — OOOO
p><^ . CMITI "• i o ? i. Iz

— _ ~-* So ^O"
• CM =5 ~" in ""
ฃ• <>; ซ ,c ฐ.
*? CM" 7 <=
>0 *? rO 1
.J <0 .J ฐ1

Q — CO
10 ^








u
Sgง CMO?
i ^ y^ i •— >o
^1^. ซ8*. .
s vl si S
1 1
•* O CM
• ซ ' •
1
.
in
CM






















x.

la
O)
^^
_ c
5 o

ฐ* 4-
ID

ID in *-.
4- j-^ LL.
ro i_ LU c o
1-0) ซ O ^
X ^^ X
•o o x o
ID 4---^OOUJ'-ปOซ
ID -C ปO • JZ 1O
,., -4-4- Q.IO 4- O-tO •
f! ID E i -• ™~  C • O — 4) 4) -1 	 4) 4) 4-
2 ~ — J>— _- 4) 0)4)1- 014)L.
c 4- — -+-.CE 4- oa. — 4-Oa. — -t-
10 0!OI003iniO — in-OC— U5T3C
3^ฃE — 4)E O 4)O 4)
1- "O — O ID O — O I-T3T3— (_T313-—
^x S?J5ฃ!5SiJ= S.= .5f8.f=i
CQ.OSOU.OOO 4-3S<4--33<
9 0)0)
S ZE
i
in
Q)
4-
01
4-
c
4)

4>
• -C
"U 4"
o
— 01
1. S
a. u
3
in
ID 4)
4- -a
10
4> E
•— 4-
"c i
0)
4) L.
f 4)
+- I
a> in
•™ ^s
(. 4)
3 E
u

ID ID
S 4)

4-
O ID
c m
3
4) ID
1- U
4) 4>
X A
in c
4- 3
S i
10
3 O
in • c
ID in
4) 4) in
4) 4) —
in in
10 4- O
4) S2
A 4- in
4)
O O 0)
c u in
-* 4)
in 4-
4- in
O •ป- 4)
C O 4-
U) 4) O)
— 4- C
1O —
to ^ t_
3
i_ e -o
4) O
V) 4)
— -O i_
a; 4) 3
E 4-
l_ t- SJ
001-
H- V4- 4)
t- a.
4) 4) E
t- Q. 4)

4- >•
10 — 4-
t- C C
4) O 4)
a. —
= in J3
4) — S
*• in 10
>.
S — E
3 IO 3
ECS
— IO —
c c •
'i o 'i ^
-o .—
4> C 4) t-
f 4) J= 4)
1— > 1— a.
0 J3 0
C-23

-------







a
-J <:
-
LU o5
3ฃ

a >-
z a:
<: LU
OS Lu
H- Qi
LU
ง O
ce x
ซC X
O- LU

CDH-
^^ ^P
)"*4
J— ซ3"
gOT
LU ****
f\ 1 1 i
o 3:
U. H-
>• O
Cฃ
งฃ
LU
•
T งฃ
O H-1
a:
iti —i
09ฐ
<:
i—






M
5
in
ID
in

+•
in
t—


—
5
in
_ซ
a
in
•ป-
n
h-







in
o
u
a.























5
i
a
a

ง7777
in i i i i
\ CO CO CO CO
Ch 1 1 1 1
Oc3fec3fe3







ง7777
'X CO CO OO GO
O\ 1 1 1 1
Ooofefe3









^o
y^
O
^s
0








c
1
10
O)

c
'g
oj

1
*^
U

a.
e
IO
a)
O)
IO
< CD O Q L.
0)
i- e i_ i_ Q.
 ' ' • • •

m ' ' IO IO IO
Si-- rป i i i
^i CN ^ft ^h CK
• • •
*o *o *o
m m fo

(0
ซ22งIง -. IK
^CM M 7 7 7 T =o
cocoinmin ao ^ i
010000^^^ ฐ
eg MCO coco
co co r-' p~' r-'
^ rป i i i
cs CM m in in
• • •
IO IO fO














_,




a)
c ._.

-i o
^** ^-
10 'e
0) ง
=!ซ
E \ — yi
\ - _ 0 6
<=* -o ( ฃ Q.
< ca en c = o.
S co aT""* ฐ ^ . ""
Q. a. ra ~ e" ^c $,
EE I_ S 4- -i 	
33 X u> •- 1. 10
a.a.ซ:oo soซ._ >Oc
O 0 (. u i. — ซ1 'Z~c "*
*-+-IBlUCD-ป-Cฃ: OOI-
c ininm so 30
._ CD 
3.—
• 0 E
>- 3 Q.
+~ ^ CL

PH
Conduct iv
Makeup coi
Chromate,
C-24

-------

























^

4)
U
0)
U)
4-
W
*~




^™
•
O
^
U)
O 5J1 — — 00
' 1 ^ ' "^ '
^g KI ~2 2
ปr eg • — "*

ฃ vl 2* o'
1 1 10
03 O 1
— ~ ฐ
•T co .
s


































c
o

T? "*"

\ 4-
<ฐ !2~" i2
o) c • c ป
^•" 4) *^ O "^
— * X
3ป O O X 0
^J 4- O ป LU O ป

Q. 5 +?* fl Sฐ ^ * ^^ *
Q. E (O 4- 4-4- ^^ 4_
* "O S C l_"O E C 1_
U5 4)Eซ T3 JฃO4) -^ O 1)
— C Q. >. 03 — — Q. — — Q.
D — O.4- 03 — 03 4) 4-— 03 03 4-
C -C 4- 4- 4- O O. — 4- O O.™ 4-
n) oraomio— in-oc— oi-ac
E303EO 03O 4)
5 4>ucut_occ^occ5
*^Xt-.C Q>-J=03 — — E43 — — &
c O.U. 0 OO 04-33 <4-- 3 3 <
ซ i J








O Z
(- Q.
43

U) C 4-
4- 4) in
C • E 43
งU> 4) 4-
4) *
— 3 4)
i- i_ tn .c
3 4) 10 4-
tf) U> 4)
10 E J=
4> 4- O)
E  3
4> U) O
4)4- 3 1-
u> (Of
3 •D U 4-
-Q 4) c X
ซ X T3
C O —
X ป- c s
O O .*
c -o
* .

4- O U C
(O M- 03 O
43 4) E in
O. CL 4) (0
E 1- S
03 >•
H 	 V 03
C C l_
C O 03 3
ro — E l_
^ t/3 (O 4)
>• a.
a- E E
3 ITS 3 03
JS.S+-
C C 4-
'I o 'i S
"O •—
43 C 0> J3
J= D -C E
I- > h- <
3 -O o -a
C-25

-------




o
=3
u.
U.

h"H
o
CO
LU

i
CD
Cฃ
LU
LU
Ll-
CS
I
r^
LU >-
Qi =3
a. =:
LU LU
a ;*:
i
i ซ
to n:
=3 =3
10 a
LU <:
Oป Qu

t— "
to H-
LU Z
Z CU
1—4
LU
It
O
~
Qฃ









-.






















„_
1
d


































S
e
 O CM ฐ| CM CM CM
*o ep ~* cn in*~* en co
CM O ~*r CM



;o CM o ซn ScM*m ซ*•
ca CM CM . ifjo F*ป 10
u*?cM"~*cn in*^ cn *ปC
CM~ w x;^ CM

10 in a tn SCM in  a.
— cn *-ป 'JHt-e • s
** C flป 3* "X. •'Ob. ^
j3j  oio -o u
II =!MM~ u 3
*— • in CM in co CM
*r "^ CM M
-s -^


ai" ~ si" ss
to en ^r ^
in in






no% • ^-fซ* CM o
^5 "*S


*^ VO* CO CO S* CM  ^
— f^ cn f* cn in
" 2S: 22
•w ป— *



*^ —1 *-S
So *7 S S 55 S
^ r^ U^(^ — ' * ^O "- '
*—• *^ C 6 5

3 o *Nป e

t. . .
1 3 " 2 I
1 '5 1 J ซ,"
H- I U_ U.
zi =s
A . A A
A


งa sf
S E






CO r^ cn 00*
CM •"• CM in
A . A A
O •"—•
A


f^. ^ (O C?
A . A •
O •*
A A

en CM ~* 
A A

co cn* to ST
<•ป. O CM <*)
•w CM ^j1 cn
A 0- * i.




Ilil
--S * -
A A A
•w ซ^"




en o" cn c?
 ja ^— IA
•a — ,,((
^s. •— vl

-e si
^ *^ U (
s o _•< s J
o-2 -* 5 -^
= 11 1:
ง3; -51
co 'a 7a ^
*- 0 UJ *- %
g ,0 |
CM ^
^ rซ
•v o



CM f







co o"
m m
• — • CM


ซ O*
Sin
. *"*
-^ CM

CO O*
CM rn
CO CO




o o"
— in






in ^4
^r to


•v in











3
i) s-
- > f
3 C 1
3 1
5 5

1
C
o
!•—
(rt
is
i II

1 ฃ
II ^


•o 'c a_
j ^ *aป 2
o ฃ ^ "O O/
1 at *-
II *"* "** *
ll 2 w u
- - ^
f I*
u *u .
>'53i
•— i- -5 "o.
> ซ ป 1

o aj
v a> >4_ wt
CL */* o at
ggซ5
• 5ซ.S

"fS
irt -O fl
llfi
tn C 2 Oi
^ *j cl —
<^ o — ป *.
="ฃ.!ง
" ^~y
S ifc
a; SiT^
5 JC 3
M O u
c ฐ I „,

5- i1^

"" U 3 5
c c
C O O 3*
s i i a
^— hU ?O CQ
C-26

-------














O)
c

ฃ

.
in
UJ

^™j
•-



























 O *T S^* CM CM
CM CM • s; ซo r-. en
CM 3 JfJ ^ CM
& -

f lsi IIP


ssงซ. Igss
CM —ซ ฐ* 2**^ SQ *""*
 e
 ns *j -a trt a> tit.
C a  c/i •— i cj cau ui
*1J rn
O O
1O S*CO P-. ^ C3 CM*
^es en" co"
cs ^.

S^r CM cs .— —* en
C3 . cz r-. o en

"ฃ. -S.
—* fปป cn en aT ^r CM*
en as . CM T tn -w

^r fป^ CM" co"
"* Sง *"" $

co^Ti^ ou? mC*
en o • co i*^ *-4 en
^ซ K CM r*&
^r c^ ซO* V
•-* cn ~* ~*
tn in

— . R en CM en
in * • - -
^H CM ^^ i

> ^ -o
'-) — "3

i- - c i 3
u g a ^ 2
3 <-ป -^. S S
-•US -a ji
a> a. c
s -a/ 
A


o ^ *- oT
cn o oo cn
CM -^ .— ' rn
A • A A
O "w'
A

si ss

CM r-4 CM U>
A A A A

in to en ^r
en -v co CM
en ซ-ซ CM to
A ea A •^-

•o



m !2T c
e *-^ 3
T3 — ""
cn """* -^
•^.5
c ^
•o o ^ =
T 1 •;; 1
"O O i/t
o c .— _
I ฐ " 3
in m
f* to


si"


in en
cn o
CM



en in"
cn o
CM


— S?
-H ^>
— i CM


^H CM*
in — *
>—


en cn
— * CM
'*"'


tn ~-t

ฐa
X



-3 . — .
H^

"~l—
ฃ1 ^"
O -^
^2".
measured by
Emissions,
C-27

-------

a
1

*ซ
C3
cd
LU
H-
>—l
a
•i^r
|
1 1 ^
^^

13
2
a:
=>
CQ
1— 4

L^f
^**j

t
ฃ
J_j

5
LU
QC
LU

3s
0
(X)
oo
ป— 4

LU
u.
o

^^
e:
g?
=s

trt
CJ
LU
03
*—











a
K




















S


a



CM
C

C


CM
00





•a




M





5
















in
4-
C
It)
4-
Q
งm o to ooป eg o
O CM • tOKl • —
x •• — o\  * • to t—
51- se -
CO
ปOO\OT OT — o
CM O TO^ •—
CO ""*
to oo — pin o o
co — CM • ego _ m
*Sk •ซ •— C3 *^CO • •
O CM O • ซO Pป
CN ^~ •"" ** Ol to ซ~
"*s^ IO s*^ •
CO


toinomOinoo
CO T CM • m-
CO




งrป o 01 Or- o o
T eg • ^CM O O
oป CM o • • d i*-
*— *~ ^— ^* G\ to ^~

CO


to o\o — op o o
CO in CM O r^CM O O*

CO



C

•5 oi.=
*• X 8
c ซ .0
CD  i_ • —
— a. a)
S c i_ c
- 0 30
• O — in —
13 .- j- m +-
O 4- (O -U 13
— IO — (_ t_
t_ i_ n a. 4- E
O tj^- c a.
OQ.UL.CO DO.
— oi-t- u1! 'd c • s
4- C uo 4)
— H 	 — me 4. •_
jj-jjfei"o— SS1? i In
= Q (/>— o cao in
0 0
to — co to 10 — to
in co • — CN to p-
eg rป to • • • •
•^ o to cj* to
eg — — co
rป to

*-* 03 •-*
r- in co CM in co r-ซ
• - CO IO CM T
eg r* eg *o * ป
— • o to oปio
CM I-. — O
'-'

to co rป — 10 to to
T in • to o to o\
eg r-eg • • • •
ป-• Oซ tO CO to
— co — in
tO to
T o oป r^ — T in
. .01 f. CO CMO\
T to • to eg to to
CM r-CM - • ซ •
•••• ซ T cor- .
— co —in
to to
T in CO T CN tOO>
• • ro oป in eg —
to ^ft • r^ to r*ป ov
eg rป to • ซ ป •
•^ — o\ o —
eg to eg to
i— r-


CM T r> in co T —
in rป • eo O> o rป
eg r- to - • • •
^ O r>- o to
eg ro eg O
^* *"^



o> eg to o — o co
• -to co eg Pป T
pป eg • to rป in T
CM CO tO - • • •
•^ co o> rป o
— in — eg
IO to


in r-ป o o o co co
• -T CO O O> —
to at • to rป •' ^^




4) ^.
3 •*-
— •-- U >J
O E in -O

U ~- 13
^-* >* (O 43
u- .a -~ c i_
* . — 3
— t- c E in
c — x .13
O 4> E 10 u
' U X S S
• 'D E 13 -Si ซ
4) a. c

+- 4) 4- +- in
IO l_ H3 03 in
1-31- l_ —
4) 4- g
a. 01 I x 4>
E — O O
4> 0 — — 10
1 — 2i U. u. •ป•
O
CO to CO —
T O — T
T — in —
o -ฃ


in T to in
r- to T r-
to — T a\
O ^


rป pป o T
T in 10 in
eg — eg to

CO 00 — —
to T 10 o>
in eg to to
O —

in in in —
co CM in co
CM — to r—
o" "^




in — TO
r-ป in o\ to
in CM to in
O — *




o to eg in
CM co in o>
^* ^~ ^r o\




to in o eg
O to in to
to CM to T
• *
O —




,—k
•<- -a
U 4>
in i_
•o 3
i- 10
a) 4)

to *-* in
E J= c
Ji x o
-0 J3 .-
O *— ' yi
j= i ซ
C (31
3 O Jฃ =
0 •- 3

X IO J) E
L. C O
tO -1- O L.
— C — ^
14) m u
•o u in
O = — —
J= 0 S 10
-1- O I4J 4-
aj o
2: h-
CO to OA CM
TO* •— T
T — in —
O •ฃ


in T to to
rป to T r~-
to — T a\
O **


o\ rป eg co
T in to in
tO^— Cl T
eg — CM to

O\ CO CM —
to T to o>
in eg to to
O —

in in in eg
co eg in co
eg — 10 rป
o




to e7 in e7
pป in o> 10
in eg to in
O — *




eo oo — co
o to eg in
eg co in ch
T — T O*




to in o eg
O to in 10
to CM to T
o —





c 'S
u t_
in 3
•o in
x 10
L. 4)
O) E

^— 'Jl
^ ^ งซ
in jo —
X -*^ !n
O) .—
D ฃ E
O ซ X 4)
— C 0)
X O s^ s

to r . .1
1 (- CO
•o-i- 0 u
O C — Si
c. 4) in u
4- u in
0) C — —
s cS 5 2
11 ฃ
in to I
CM r-
— eg 1


CN CO
. 10
to —



to co
u> —
eo -~-

o eg
oป ~*^

o o<
^r oo




to o
-5




T CO
\o




CO ^




















4)
4.
U
a
a.
U)
4>
u>
4)

4)
in
4-
tn
4-
C

.
4)
E
Iฑ
_

4>



3 -C
O X
— J3
x ^
4) J=
a. x
5 en

4_ k
C U1
o c
i3 O
O in
in in
a —
(0 E
LU
ol
4)
•^
"*~
in
•*~
5

^
o"
3
4)
f-
—
a •
C-28

-------


oo
X
UJ
.
o

"^
oo
00
1—4
^5!r
L* J

LL_
O

Qฃ
^^

2:
i^
OO
.
P-*
1
UJ

CQ
1—
















IO
c
a
















fO
u
0
VI
t

X •• — TT CM^ul O •
ON "~ — —




vo r^ o f- O— tr\ oo
09 — CM • ON— CM in
X •• — ON CMin • • -
— ^ ON • • o rป
— — tOto to
ON — —



vo r- o O o!o in ป>•
03 — CM • iON CM VO
x ^ *" 2 in*"* ฐ" •*•
ON *" "~ —3 m







C
x =
^j -2^ "**

= . .8
ajo) co
o 4- —
u to c
— a. d>
S c c c
- O 3 O
• O — in —
•o — +- tn +-
O 4- to d) to
— (O — t_ C.
<- V. 3 0. -t- S

d)O.Ot_+- O S S- C • g
4- C 1) d)X +- O I- to
- 4- IO E O 	 —* L.4- I in
: Q oo vv> — cj Q3o in
> rS
r- to CM r- — vo oo
• • KI in ON ON CM
Or- • in in to oo
to 03 ^r ^ • * •
• — ' r- CM in o : ON r-
O r- • ^ oO 03 r-
to co ^* ป ^ * •ซ
— ' oo eo r- ao
CM CM
— —

ON — o ^r ON in o
• • ON o ^ tn ON
ao ^ • vo — o P-
CM 00 (0 • ซ . .
*"•* ON ON ON ON
IO —
IO IO
— •—








d) *^
E E
— r* 7i
O E in
> •*. 13
O — T3
-^ >^ to d)
u. S3 ~— e i_
' '~ 3
-^ 4- C E •/*
C — \ - '
CJ d) ฃ to
0^ E I ~
* Q S "O Ul
DO. c
t- * * . 0
3 * d) 
S — 0 0
d) O — — vo
t— S u. u- 4.
O .
oo — ao to
CM — o o in ON
-ON • ซ
00 • 03 00
to vo in CM
••" —


ao in in in
CM O r- O
in to CM vo
CM —
; O




to to oo to
00 ON CM VO
ON CM in —
o —




to o in to
— rป — vo
r~J.ฐ"~ซฐ.
• ao • ^ ON
to — •


vo Or* c7
in oo in ON
to to o —
f O> in in
T — CM in






^
^*- T3
u d;
en i_
•a 3
x in
<- tc
CT d]
^^ E
to r~* tn
E -C c
•S ^ ฐ
Q) ' ^, Hป

0 ~ ^ 1"
— C O)
x O ^ E
'•~ 3
•0 4- . , ._
— :O in S
I i_ c -o
V 4- O L.
c d) in tj
t- u in
fc — —
o s n
CJ Ul 4-
o ,2
co — ooro
CM — tr —
vo r> in vo
•— O CM in



03 in in —
O 
4!
u
<0
a.
in
0)
u
in
0)
d)
U)
4-
V>
0)
4-
1?
s

T oT
•CM
ON IO
— ~~














3 !c
O X
X —

L.
d) J=
c. X
iO O>
Q. V*


c in
D c
n o
i. —
O tn
in in
Q —
to E
UJ
a
L.
1
C
~
U
^
I-
0)
X
0
O)
^
8
u


•—

4-
C
•U
•a


O
2
3
d)
-C
h—
3
C-29

-------



ฃ
X
LLJ

2ฃ
^-c
ฃ
-
o;
u**T

*JL.
S


•
CO

O

LU
CD

a
3
CM
ซc
CO
1
3
j;
OO
3
















in
m
*~o*ซo oo
c)J -SIS'*1




tooor-Oom oo
OOIOCM • T pg  O — in —
•o — *- tn -i-
O 4- ซJ v>— O 03O in
at 10
O 0
^-* ^^ .— ,
rป r> •— o CM in to
to CM • T o CM In
to Oi in • • ป •
— T — to Ol
— — — 10
in T

T to in 01 to 01 oo
• • rป in to oo —
CM o • oo O to r-
IO Ol T • ซ • •
•^ in o ซr oo

in in

in T T r- Oi Or*
to CM • in o — to in in oo
— oo — to
in in



tooo — o in — rป
• • oo CM r- r* CM
CM O • rป O CM Oi
to Oi T • * • *
— oo — r- 01
— to — o
to to
tO tO Oi O 03 Ol to
• • rป t~ to r- —
CMO • O — tOO
tO Ol T • • • •
— ' 00 OO to Ol
— to — oo
to in

O Oi ao 01 in to T
• • to in in CM to
— r- • to ซr T 01
tooo T ซ...
"~ — CM 2|—
to in
— in CM in in oo S
• • Oi o O r- in
to — • to — oi CM
to Ol T * * • *
— rป — in T
to in






o — -o
•— >- IO ID
Lu ฃ> ~— c u
— t- e • E in
c — \ 10
O 'U E ro (U ซ
• ON E S
i-ro tn
3> E -0 tn
ma. c
(- • - . O
3 - 0) 0) —
+- at +- 4- , in
to L. 10 10 in
L. 3 (_ U —
01 •*- E
a. in x x 3t
E •- 0 0
4U O ~~ ^ to
I— Z U. Lu +•
0
^^
r^tooo*— t**Ttoto
CM — Ol O inCM(Nin
Oi •— ao ao oiCMOioo
• CM * • • • * •
•— ซoi— •— inoi*—
— in CM — ^ CM

CMTOCM T CM •— to
to — — to tOO— to
CMOCMT CM — CM T
• O
o -i


totoinT in^roo-
r*ปtotoo i — toto —
to- tooo to — to oo
0 ^ o ~




in p~ 03 co toOr^CM
tooo— in oitototo
torปTin oiOir-CM
oo to oo oo oo to eo*oT
^^ ^i*



m *o ^h *f 10 ^3 CD ^o
OO^CMO oo\mo
CMOOCMtn CNOCNin
O *-" * "-•
• o
o ^

Oซ~O^O* fMtnป-CM
in^oo— in co ov  • •>
o 3 S C.
^^

gjo^.— CM in — CM
OOOCM^T or-*^r*-
to i^ ^r ^^ ^* ^ tp ^h
i 2

CMO^OO intoo*CM
^*o^**r ior^voo
O^^O CM ro o r-. CM *r
^"•*TO* ^— •q-O*
^^

^^
O ^*-
"O u>
"S "O
<- \
Ol tซ
^o S
ro ~* m — ^^. m
E .e c x to ^ c
in \ o E \ o
•o a — to ji ^ .-
N — tn — "o — ',n
01 *-* tn i \ *-* en
E — "O 5) ._
3 ^= E O ^ E

— CO) •"- c O)
^ ~ ". I - •! *. !
^"* fc ^o 'j) c
It- C O -O L. C O
•o -t- o L. •<- o c:
O C — C T3 c — .C
f>lt in uiu^t m u
+- u in i_ o in
!EO E lOinO E (O
O LU •ป- IO O LU •*-
>• o at o
J3 t- S 1-

II ~ —
ป- ro —
in ^™*


00 to 1 ~~
to* 2 ' J,
T ^^



II 03 P
rr *-^




to c7 i T
0= ' ฑ
in •^




II Ol Jo
II • Oi
•** CM CM
IO "^
"^


s'B -


\l \l

r~* — ii
co'- ' -L
"""



•o
o L
•— 3
x ui

(- (U
a E~ in = j= ji
il c O o c
O i_ x o
•o .— ฃi -y •—
4) in 
-------
TABLE C-9.  SUMMARY OF SENSITIVE PAPER DRIFT MEASUREMENTS
Site/location
Department of Energy, Paducah, Ky.



National Bureau of Standards, Gaithersburg, Md.
!


Exxon Refinery, Baytown, Tex. (Tower 68)




Exxon Refinery, Baytown, Tex. (Tower 84)


i
Cell
4
5
6
7
A
B
C
D
1
2
3
4
5
A
B
C
D
Sensitive
paper drift
rate percent
of recirculation
0.0083
0.0093
0.0009
0.0003
0.0002
0.0004
0.0001
0.0001
0.0047
- 0.0103
0.0072
0.0040
0.0045
0.0009
0.0006
0.0005
0.0008
                            C-31

-------

-------
                                APPENDIX D.
D.I  CHROMIUM DISCHARGE REGULATIONS
     Most States have wastewater discharge regulations that  limit the
amount of chromium that may be discharged into publicly owned treatment
works or to surface waters from any type of source.1  Although some of
these State regulations are fairly stringent, none prohibit  the discharge
of chromium-laden wastewater.  No State regulations directly affect air
emissions of chromium from cooling towers although there are States that
have ambient air quality standards for chromium (e.g., Maine) or hazardous
air pollutant regulations for some chromium compounds (e.g.,
Connecticut).2'3                          •        '
     At the present time, no information has been found on chromium
environmental regulations in countries other than the U.S.
D.2  REFERENCES FOR APPENDIX D
1.


2.

3.
Memorandum from M. Upchurch, MRI, to Comfort Cooling Tower Project
Files.  August 4, 1986.  State water effluent regulations for chromium
discharge.

State Air Laws.  Environment Reporter.  Bureau of National Affairs
Inc., Washington, D.C.   Volume 2.  p.  396:0105.  January 9, 1987.
State Air Laws.  Environment Reporter.•  Bureau of National Affairs
Inc., Washington, D.C.   Volume 1.  pp. 331:0534-0538.   October 24,
1986.
                                   0-1

-------

-------
                               APPENDIX E.
     The following data were used to estimate the total annual market
value of corrosion inhibitor chemicals sold for use in CCT's.  Annual
weighted average chromate costs for one tower are estimated as follows:
Chromates
Building
size, m
673
1,460
3,405
6,224
12,338
37,626

No. of CCT's,
thousands3
18.55
83.50
51.20
42.00
31.50
23.75
250.5
Percent of
total CCT's
7.4
33.3
20.4
16.8
12.6
9.5

Annual weighted average nonchromate costs
as follows:
Nonchromates
Building
size, m
673
1,460
3,405
6,224
12,338
37,626

aTable 4-4.
DTable 7-1.


No. of CCT's,
thousandsa
18.55
83.50
51.20
42.00
31.50
23.75
250.5




Percent of
total CCT's
7.4.
33.3
20.4
16.8
12.6
9.5



Annual
chromate
costs0
17
36
85
155
306
935
TOTAL
for one tower are

-
Annual
phosphate
costs0
33
72
169
309
613
1,869
TOTAL


Weighted
average
cost
1
12
17
26
38
89
= $183
estimated


Weighted
average
cost
2
24
35
52
77
177
= $367


                                   E-l

-------
     The total market value  of  chromate  and  nonchromate  corrosion
inhibitors used annually  in  CCT's  is  estimated  by  solving  the  following
equation:
            TV = 37,580 ($183) + 213,020 ($367) = $85.1 million
where:
          TV = Total annual  value  of  chromate and  nonchromate  corrosion
               inhibitors used  in  comfort cooling  systems
      37,580 = Estimated  number of chromate-based  comfort  cooling towers
     213,020 = Estimated  number of nonchromate-based comfort cooling
               towers
        $183 = Average annual cost of using chromates per  comfort cooling
               tower
        $367 = Average annual cost of using nonchromates per comfort
               cooling tower.

     Note that the above costs represent only the cost to purchase the
corrosion inhibitor chemicals themselves and do not include the cost of
technical services that may be required of specialty chemical companies.
However, the costs of such technical  services are not expected to increase
significantly as substitutes for chromates are used.
                                   E-2

-------
                                APPENDIX  F.
          SAMPLE CALCULATIONS OF FLOW RATES, TOWER PARAMETERS, AND
                 HEXAVALENT CHROMIUM  (Cr+6) EMISSION RATES

      The  calculations  presented in  this Appendix  correspond with the
 discussion  in  Chapter  4.   Equations presented  in  Chapter 4  are not
 repeated  in this Appendix.  Calculations  of flow  rates  and  tower
 parameters  are presented for model  tower  No. 1  and  calculations  of Cr"1"6
 emission  rates are presented for Alabama.   English  units were used in the
 calculations in Chapter 4  and the results were  converted to metric
 units.  The results shown  in Chapter  4 also are shown in this Appendix for
 consistency.   However, the calculations in this Appendix may yield
 slightly  different results than those shown because rounded  results  from
 previous  equations and the results  converted to metric units  in  Chapter 4
 are used  in the calculations.
 F.I  COOLING TOWER CAPACITY
     The  size of model building No. 1 is 673 m2 and the  cooling
 requirement is  142 W/m .  Thus, the cooling tower capacity must be at
 least,
                (673 m2)(142 W/m2)  = 95,400 W = 95,400 J/s
F.2  REC'IRCULATION RATE
          Recirculation rate =
     (95.40.0 J/s) (60 s/min)
(4.18 j/g/oc)(5.6ฐC)(1,000 g/a)
                             = 246 ii/min
F.3  EVAPORATION RATE
       Evaporation rate = (0.00085/ฐF)(246 a/min)(5.50C)(1.8ฐF/ฐC)

                        =2.08 a/min
                                   F-l

-------
F.4  SLOWDOWN RATE
                       Slowdown rate =
                                        2.08 a/min
                                    =0.53 a/min
F.5  AIRFLOW RATE
     The airflow mass rate, G, is 1.5 times the water recirculating
rate.  The density of saturated air at 26.7ฐC (80ฐF) is 1,162 g/m3.
              A1rf ,ow rate =
                          = 141 m /min
F.6  STACK DIAMETER
     Typical stack airflow velocity, V, is 520 m/min and area of the stack
is given by A = ud /4 = 6/V.
            Stack diameter  =   <4H141 *3/"n-n)(min/520 m)
                     1/2
                           = 0.59 m = 1.9 ft
Because equipment specifications are typically in English units, the size
of the stack was rounded up to 2.0 ft (0.6 m)*
F.7  RECALCULATED EXIT AIR VELOCITY

                      V = ฃ = (141 m /min)(min/60 s)
                          A       (ir)(0.6 m)2/4
                            = 8.2 m/s
F.8  DISTRIBUTION OF CHROMIUM-USING CCT'S
     The ratio of the population of Alabama to that of the United States
is used to represent the percentage of all  towers nationwide  that are
located in Alabama.
                  Percentage of all
                  towers that are
                  located in Alabama
=  3,893,046
  226,147,597 '
=1.72 percent
                                   F-2

-------
      The total number of model tower No. 1's nationwide is 2,780.  Thus,

 the number of model tower No. 1's 1n Alabama is,



             r0.0172 tower in Alabama^- 70n .

             I   tower nationwide	K2ซ780 towers nationwide)



             = 48 towers in Alabama


 F.9  HEXAVALENT CHROMIUM EMISSIONS RATE
                   , g

      The hourly Cr   emission rates are based  on the lowest and highest


 emission factors obtained from EPA-sponsored emissions tests of industrial


 towers  using low-efficiency drift eliminators.   The lowest value was


 obtained at DOE-Paducah and the highest value  was obtained at Exxon-

 Baytown.
Lower-bound Or

emission rate
                                      0.000066 mq Cr
                        (ppm  Cr    recirculating)(2,  H20  recirculating)
                                                                    •]
                        246  9.  H20
                       rrecirculatinQif4.48  ppm Cr+6]r,n   .  , ,

                       1     min      JlrPCirrtjIal-mnH60 min/hl
                                   recirculating'
                       4.4 mg Cr+6/h
                  +6
    Upper-bound Cr  _

    emission rate
                                 0.001874 mq CrH
                       (ppm Cr   recirculating) (a  H20 recirculating)
                                                                •]
                       246 a  H.O

                     '' rrecircuiatingir4.48  ppm
                                                         .  .. ,
                                                       min/h]
                     = 124 mg Cr+6/h




Estimates of Cr+s emissions from model tower No. 1 in Alabama are based on


the tower utilization factor for the State.  In Alabama, it is estimated

that CCT's are operated 59 percent of the time.
                                    F-3

-------
                 +6
                                +6
                      [4*4 m9 Cr/h][8,760 h/yr] [0.59] [kg/1,000,000 mg]
                                  +6
                    = 0.0226 kg Cr/yr

                                +6
                    = [124 mg Cr  /hi[8,760 h/yr][0.59][kg/1,000,000  mg]

                    = 0.6420 kg Cr*6/yr
Estimates of the total Cr+6 emissions from all model tower No. 1's in
Alabama is based on the emission per tower and the number of towers in the
State.
   eSissionU?ater   = [0'0226 k9 Cr+6/yr/tower No. 1][48 tower No. 1's]

                    ^ 1.08 kg Cr+6/yr
= [0*6420
                                +6
                    = 30.8 kg Cr  /yr
                                               No. 1][48 tower No.  1's]
                                   F-4

-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing!
1. REPORT NO. • 2.
EPA 450/3-87-010a
4. TITLE AND SUBTITLE
Chromium Emissions from Comfort Cooling Towers--
Background Information for Proposed Standards
7. AUTHOR(S)
9. PERFORMING ORGANIZATION NAME AND ADDRESS
• Office of Air Quality Planning and Standards
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
12. SPONSORING AGENCY NAME AND ADDRESS
Director of Air Quality Planning and Standards
Office of Air, Noise, and Radiation
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
March 1988
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3817
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES 	
       Standards of performance for the control  of  hexavalent chromium emissions from
  comfort cooling towers are being proposed under authority of Section 6 of the Toxic
  Substances Control Act.  These standards would apply to existing and new comfort
  cooling towers.  This document contains background  information and environmental and
  economic impact assessments of the regulatory  alternatives considered in developing
  the proposed standards.
1T- KEY WORDS AN-^ DOCUMENT ANALYSIS
a DESCRIPTORS
Air pollution
Pollution control
Comfort cooling towers
Hexavalent chromium
HVAC and refrigeration systems
Corrosion inhibitors
Unlimited 1
i
b. IDENTIFIERS/OPEN HNDED TERMS
Air pollution control
19. SECURITY CLASS I Hits Report,
Unclassified
20. SECURITY CLASS iThis pa%e/
Unclassified
<:. COSATl Held/Group

21. NO. OF PAGES
194
22. PRICE
EPA Form 2220-! (Rev. 4-77)
                     = RฃVIOUS SOIT1ON IS OBSOLETE

-------

-------