United States      Office of Air Quality      EPA-450/3-83-016b
           tnvironmental Protection  Planning and Standards     June 1988
           A9encV        Research Triangle Park NC 27711
           Air
<»EfiA     Benzene Emissions     Draft
          from Coke              EIS
          By-Product
          Recovery Plants-
          Background
          Information for
          Revised Proposed
          Standards

-------
                 ERRATA FOR COKE BY-PRODUCT RECOVERY PLANT
      BACKGROUND INFORMATION DOCUMENT FOR REVISED PROPOSED STANDARDS

     This background information document (BID) responds to comments on the
1984 proposal and also serves as the basis for reproposal of a revised
standard based on EPA's response to the court decision noted on page 1-1 of
this BID.  However,  readers of this document should note that while this
BID refers to "the revised proposed standard" on several pages, EPA is
proposing a total of four different regulatory approaches that would result
in different revised proposed standards.  References in this BID to the
"revised proposed standard" and associated impact data pertain to
Approaches A and B presented in the preamble.   All information on the
revised proposed standards under Approaches C and D is presented in the
preamble.

-------
                              EPA-450/3-83-016b
      Benzene Emissions from
Coke By-Product Recovery Plants-
      Background Information
 for Revised Proposed Standards
             Emission Standards Division
                        ft^?*!*"*	
                                         ^Sfency
         U.S. ENVIRONMENTAL PROTECTION AGENCY
             Office of Air and Radiation
         Office of Air Quality Planning and Standards
         Research Triangle Park, North Carolina 27711

                 June 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, for a fee, from the National Technical Information Services, 5285 Port Royal Road,
Springfield, Virginia 22161.

-------
                       ENVIRONMENTAL PROTECTION AGENCY

                            Background Information
                                  and Draft
                        Environmental Impact Statement
                              sed Proposed Standards for
                             y-Product Recovery Plants
                                 Prepared by:
 /
Jack R. Farmer                                                      (Date)
Director, Emission Standards Division
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina  27711

1.  The revised proposed national emission standards would limit emissions
    of benzene from existing and new coke by-product recovery plants.   The
    revised proposed standards would implement Section 112 of the Clean Air
    Act and are based on the Administrator's determination of June 8,  1977
    (42 FR 29332), that benzene presents a significant risk to human health
    as a result of air emissions from one or more stationary source
    categories and is therefore a hazardous air pollutant.  The EPA
    Regions III, IV, and V are particularly affected because most plants
    are located in these areas.

2.  Copies of this document have been sent to the following Federal
    Departments:  Labor, Health and Human Services, Defense, Transportation,
    Agriculture, Commerce, Interior, and Energy; the National Science
    Foundation; the Council on Environmental Quality; State and Territorial
    Air Pollution Program Administrators; EPA Regional  Administrators; Local
    Air Pollution Control Officials; Office of Management and Budget;  and
    other interested parties.

3.  For additional  information contact:

    Mr. Gilbert H. Wood
    Emission Standards Division (MD-13)
    U.S. Environmental  Protection Agency
    Research Triangle Park, North Carolina  27711
    Telephone:  (919) 541-5625

4.  Copies of this document may be obtained from:

    U.S. Environmental  Protection Agency Library (MD-35)
    Research Triangle Park, North Carolina  27711
    Telephone:  (919) 541-2777

    National  Technical  Information Service
    5285 Port Royal  Road
    Springfield, Virginia  22161
    Telephone:  (703) 487-4650

                                       iii

-------
                             TABLE OF CONTENTS


                                                                        Pa^e

 Tables	     viii

 1.   SUMMARY	                          l  l
     1.1   SUMMARY  OF  CHANGES  SINCE 1984  PROPOSAL!  .'.'.'.'	      ill
     1.2   SUMMARY  OF  IMPACTS  OF  REVISED  PROPOSED ACTION  ....''.'      1-2
          1.2.1  Environmental and Energy  Impacts  of
                Revised  Proposed  Action	      1_2
          1.2.2  Health Risk  Impacts  of  Revised Proposed Action  .'  .'      1-3
          1.2.3  Cost  and Economic Impacts of  Revised Proposed
                Action	      1_3
          1.2.4  Other Considerations  	 .**'.''*'      1-4
                1.2.4.1  Irreversible and Irretrievable
                         Commitment	      1_4
                1.2.4.2  Environmental and Energy Impacts
                         of  Delayed Standards  	      1-4
                1.2.4.3  Urban  and Community  Impacts 	      1-5

2.  SUMMARY OF PUBLIC COMMENTS  	     ?-l

3.  SELECTION OF SOURCE CATEGORY  	                           3 i
    3.1  SELECTION OF SOURCE CATEGORY	     3"i
    3.2  REGULATION OF MERCHANT PLANTS	     31
    3.3  EXCLUSION OF FORM-COKE PLANTS 	  "'*''!!     3-2

4.  SELECTION OF REVISED  PROPOSED STANDARDS	                    41
    4.1  SELECTION OF LEVEL OF CONTROL	    	     4 }
    4.2  REGULATORY DEFINITIONS OF FOUNDRY AND FURNACE  	
         BY-PRODUCT PLANTS  	     4_2

5.  EMISSION  CONTROL  TECHNOLOGY   . .                                    c i
    5.1  DEMONSTRATION OF CONTROL  TECHNOLOGY '. '.	     51
    5.2  SAFETY, DESIGN,  AND  OPERATION OF  POSITIVE-PRESSURE*  '  '  '
         CONTROL SYSTEM  	                                  5 3
    5.3  SAFETY, DESIGN,  AND  OPERATION OF  NEGATIVE-PRESSURE*    *  '
         CONTROL SYSTEM	                                  ,- o
    5.4  MONITORING FOR CARBON MONOXIDE   . .  ' !	     Tin
    5.5  SUMP  CONTROLS	      	     jf {V
    5.6  OPERATIONAL  PROBLEM  FROM  PLUGGED*VENTs'oR VALVES*  .'  .'  .'  "     i-12

-------
                          TABLE  OF  CONTENTS  (con.)

                                                                       Page

    5.7  CONTROLS FOR BENZENE  STORAGE  TANKS   	      5-12
    5.8  DETERMINATION OF CONTROL EFFICIENCIES  	      5-13
    5.9  GAS BLANKETING VERSUS WASH-OIL  SCRUBBERS   	      5-14
    5.10 FINAL COOLERS AND NAPHTHALENE PROCESSING   	      5-15

6.  ENVIRONMENTAL IMPACTS  	      6-1
    6.1  DATA BASE FOR ENVIRONMENTAL IMPACTS	      6-1
    6.2  FOUNDRY PLANT EMISSION  FACTORS   	      6-4
    6.3  MODEL COKE PLANTS	      6-7
    5.4  EMISSION FACTORS FOR  TAR-RELATED SOURCES   	      6-9
    6.5  METHODOLOGY FOR EMISSION  FACTORS  	      6-12
    6.6  VOC BENEFITS FOR OZONE  REDUCTION	      6-13

7.  COST IMPACT	      7-l
    7.1  REVISIONS TO COST ANALYSIS	      7-1
    7.2  REVISIONS TO PRODUCT  RECOVERY CREDITS  	      7-2
    7.3  ECONOMIES OF SCALE FOR  SMALL PLANTS 	      7-4

8.  ECONOMIC IMPACT	      8-1
    8.1  REGULATORY BASELINE 	      8-1
    8.2  SELECTION OF DOLLAR YEAR	      8-2
    8.3  POTENTIAL ECONOMIC IMPACT 	      8-2
    8.4  ESTIMATED EMPLOYMENT  IMPACT 	      8-3
    8.5  IMPORT TRENDS	      8-4
    8.6  ECONOMIC IMPACT ON SMALL  PLANTS 	      8-4
    8.7  PRICE  IMPACTS	      8-6
    8.8  ECONOMIC IMPACTS ON FOUNDRY PLANTS  	      8-7

9.  QUANTITATIVE RISK ASSESSMENT 	      9-1
    9.1  USE OF MODEL FOR HEALTH RISK ESTIMATES	      9-1
    9.2  SELECTION OF RISK MODEL	      9-3
    9.3  UNIT RISK ESTIMATE	      9-5
    9.4  DERIVATION OF UNIT RISK ESTIMATE	      9-7
    9.5  COMPARATIVE RISK FROM GASOLINE MARKETING  	      9-8
    9.6  COMPARATIVE RISKS FROM OTHER SOURCES  	      9-9
    9.7  SELECTION OF BENZENE VS.  POM	      9-10
    9.8  CONSIDERATION OF OTHER HEALTH  EFFECTS 	      9-10
    9.9  ANCILLARY COMMENTS   	      9-12

10. EQUIPMENT LEAK DETECTION AND REPAIR	      10-1
    10.1   DETERMINE  EMISSIONS OVER BACKGROUND LEVELS  	      10-1
    10.2   COMPLIANCE WITH LEAK  DETECTION AND REPAIR PROGRAM  .  . .      10-1
    10.3   DEFINITION OF  EQUIPMENT LEAK	      10-2
    10.4   ON-LINE  VALVE  REPAIR  	     10-5
    10.5   EQUIPMENT  LEAK REPAIR PERIOD  	     10-6
    10.6   DELAY OF  REPAIR	     10-8
    10.7   ALTERNATIVE  STANDARD  FOR VALVES   	     10-8
    10.8   EXEMPTION  FOR  DIFFICULT-TO-MONITOR VALVES   	     10-9
    10.9   ALTERNATIVE  STANDARD  FOR OPEN-ENDED VALVES  OR LINES   . .     10-11

                                      vi

-------
                          TABLE OF CONTENTS  (con.)
                                                                       Page
 11. RECORDKEEPING AND REPORTING  	                  jj.j
    11.1  ALTERNATIVE MONITORING AND RECORDKEEPING  ..!!!'*'    11-1
    11.2  RETENTION PERIOD FOR RECORDS AND REPORTS  ....            11-1
    11.3  ENFORCEMENT BASED ON RECORDS AND REPORTS  FOR
          EQUIPMENT LEAKS	    u_2

 12. MISCELLANEOUS	                        12 l
    12.1  ALTERNATIVE MEANS OF EMISSION LIMITATION  ] .* '.	    12-1
    12.2  DEFINITION OF TAR DECANTER	               *      i2 i
    12.3  DEFINITION OF EXHAUSTER	       	    12 2
    12.4  WAIVER REQUESTS	   	    12_2
    12.5  NEED FOR ADDITIONAL ENFORCEMENT RESOURCES  !!!"*"'    12-3
    12.6  SELECTION OF FORMAT	               * * '    12 3
    12.7  LIGHT-OIL SUMP CONTROL EFFICIENCY  .....'!.'.' .* .' .'    !2-5

APPENDIX A  ENVIRONMENTAL IMPACT ANALYSIS  	                   A-l
APPENDIX B  COST IMPACT ANALYSIS 	       ....
APPENDIX C  ECONOMIC IMPACT ANALYSIS 	       	     r 1
APPENDIX D  HEALTH RISK IMPACT ANALYSIS	'. '.	     o_

-------
                                   TABLES

Number                                                                Zi3§.

  2-1      List of Commenters  on  1984  Proposed  National  Emission
           Standards for Coke  By-Product Recovery  Plants 	       2-2
                                     viii

-------
                             1.  SUMMARY

     On  June  6,  1984, the Environmental Protection Agency (EPA) proposed
 national  emission  standards  for benzene emissions from coke by-product
 recovery  plants  (49 FR 23522) under the authority of Section 112 of the
 Clean Air Act  (CAA).  Public comments were requested on the proposal in
 the  Federal  Register, and the comment period was extended, by request, to
 October 17,  1984 (49 FR 33904).  The 20 commenters were composed mainly
 of affected  companies and industry trade associations.  Also commenting
 were various State and county air pollution control  or environmental
 health  departments and one environmental  group.  The comments that were
 submitted, along with responses to these comments, are summarized in this
 document.  The EPA reconsidered the proposed standards in light of the
 court decision in Natural  Resources Defense Council, Inc. v.  EPA,
 824  F.2d  1146 (D.C. Cir.,  July 28, 1987)  and reproposed the standards  in
 June 1988.  The summary of comments and responses serves as the basis  for
 the  revisions made to the  standard between proposal  and reproposal.
 1.1  SUMMARY OF CHANGES SINCE 1984 PROPOSAL
     Since the 1984 proposal, the data base has been revised  to reflect
 the  industry operating status as of November 1984 (shortly  after  the
 close of the comment period).  Based on comments received on  the  1984
 proposal,  EPA revised  the  estimated nationwide impacts  of control
 (including baseline) for furnace and foundry  coke producers separately.
The Administrator used the  revised environmental, health,  cost, and
economic impacts for his reconsideration.
     One major change  since  the  1984 proposal  is the revised  proposal  of  a
zero  emission limit for  naphthalene processing operations,  final  coolers,
and final-cooler cooling towers  at  plants  producing  furnace coke.  The
revised  proposal  is based on  wash-oil  final coolers.  Another major  change
is that  proposed standards for control  of  storage tanks  containing

                                  1-1

-------
light-oil, benzene-toluene-xylene  (BTX) mixtures,  benzene, or excess
ammonia-liquor at  furnace  and  foundry  plants  have  been eliminated  from the
reproposed standards.
     Other changes to  the  standards  have  been made for clarifying  purposes.
The definition of  "coke by-product  recovery plant" has been  revised specif-
ically to exclude  form-coke  plants.  New  definitions  for  "furnace" and
"foundry" coke and coke by-product  recovery plant  also have  been added to
distinguish these  industry segments  in terms  of the volatile content  of
coke produced, the length  of the coking cycle and  the percent of each
type of coke produced  annually.  New definitions for  "exhauster" and  "tar
decanter" also have been added.  For gas-blanketed process vessels and
light-oil sumps, monitoring  provisions have been added to ensure that there
are no leaks from  the  access hatches and  covers upon  reclosure after  usage.
The regulation also has been revised to directly cross reference the
provisions of 40 CFR 61, Subpart V  for equipment leak requirements.   The
EPA also proposes  to amend Subpart  V where necessary  for  clarification of
the cross referencing.

1.2  SUMMARY OF IMPACTS OF REVISED  PROPOSED ACTION
1.2.1  Environmental and Energy  Impacts of Revised Proposed  Action
     The environmental  and energy  impacts of  the revised  proposed  standards
are discussed in Chapter 6 of  this  background information document (BID).
The estimated environmental  impacts  have  been revised since  the 1984
proposal to update the operating status of the  industry to November 1984.
These changes are  discussed  in Chapter 6, "Environmental  Impacts."  Revised
environmental impact tables  and  emission  factors are  presented in
Appendix A.
     Implementing  the  revised  proposed standards would reduce nationwide
benzene emissions  at 44 furnace  and  foundry plants from the  baseline  level
of about 26,000 megagrams/year (Mg/yr) to about 2,000 Mg/yr, a 93-percent
reduction.  Nationwide emissions from  coke by-product recovery plants of
total volatile organic compounds  (VOC) including benzene  also would be
reduced from the baseline  estimated level of  171,000  Mg/yr to about
6,000 Mg/yr, a 96-percent  reduction.  Assuming  recovery of 21.3 liters of
gas/min/Mg of coke/day, the  revised proposed  standards would result in a
national energy savings of approximately  1,800  TJ/yr  from recovered coke

                                  1-2

-------
 oven gas.  Impact calculations for energy requirements and coke oven gas
 recovery estimates are shown in Appendix A.

 1.2.2  Health Risk Impacts of Revised Proposed Action
      The quantitative risk assessment conducted for the proposed  standards  is
 discussed in Appendix E of the proposal  BID  (Benzene Emissions  from  Coke
 By-Product Recovery Plants - Background  Information for Proposed  Standards.
 EPA-450/3-83-016a); further information  is in  the  preamble to the proposal
 (49 FR 23525).   The risk assessment  has  been revised since the  1984  proposal
 to update the current industry operating status and to  incorporate adjusted
 emission factors.   Other changes  in  the  risk assessment  since the 1984
 proposal  include a revised benzene unit  risk estimate  (URE), which is 17
 percent higher,  and an  increase in the exposure modeling radius to 50
 kilometers  (km).   Further information regarding these changes is  provided in
 Chapter 9,  "Quantitative Risk  Assessment," and  in Appendix D.
      Annual  leukemia  incidence associated with  baseline benzene emissions
 at 44 plants  is  estimated at 3 cases/yr.  Implementation of the revised
 proposed  standards  would reduce the  estimated incidence to 0.2 case/yr.
 The maximum  individual  lifetime risk  (MIR) at the baseline is predicted to
          •^
 be 6  x  10"  .  The  revised  proposed standards are expected to reduce the MIR
 to approximately 4  x  10'4  (about 4 in 10,000).

 1.2.3   Cost and Economic  Impacts of Revised Proposed Action
      Control costs  for model by-product recovery plants are discussed in
 Chapter 7 of the BID  for the proposed standards.  This  analysis  has been
 updated since the June 1984 proposal  to reflect the industry operating status
 as  of November 1984.  Other changes include the adjustment  of certain cost
 functions and the modification of light-oil/fuel recovery credits, as applied
to  plants that practice the flaring of excess coke  oven gas.   These changes
are discussed in Chapter 7, "Cost  Impact."  The revised cost analysis is
presented in Appendix B.
     Based on the revised analysis,  the estimated national  capital  cost  of
the revised proposed standards  for the 44 plants is  estimated at about
$84 million  over baseline costs (1984 dollars).   The total  annualized  cost is
estimated at $16 million/yr.
                                 1-3

-------
     The nationwide economic  impact  of  the  proposed standards  is analyzed in
Chapter 9 of the proposal  BID.   This analysis, which  also  has  been  revised
and updated since the 1984 proposal, is discussed  in  Chapter 8 of this
document; further information is presented  in Appendix  C.  Based on the
revised economic analysis, the  price of foundry  and foundry coke is projected
to increase by less than 1 percent  above baseline  values.  The Agency does
not expect closures as a direct result  of the  revised proposed standards.
However, many furnace and foundry plants are presently  in  marginal  economic
condition or operating at a loss, and the Agency recognizes that the
standards could be a factor that would  trigger  closure  decisions at some  of
these plants.
1.2.4  Other Considerations
     1.2.4.1  Irreversible and Irretrievable Commitment.   As discussed  in
Chapter 7 of the BID for the proposed standards, the  control options  do  not
involve a tradeoff between short-term environmental  gains  at the expense  of
long-term environmental losses.  An increased  cyanide (HCN) concentration in
wastewater is expected if indirect final cooling is  used instead of direct
final cooling.  Measured HCN air emission and  calculations based on
once-through cooling water indicate that about 200 g/Mg of coke  could be
added to wastewater for treatment.  However, this increase is  not  anticipated
to cause problems  for compliance with effluent regulations.
     The control options do not  result in irreversible and irretrievable
commitment of resources.  As a  result of the control  options,  resources such
as light aromatic  hydrocarbons  are  recovered,  and emissions from the
majority of  affected  sources are reduced substantially or eliminated.
     1.2.4.2  Environmental and  Energy  Impact's of Delayed Standards.   The
environmental and  energy  impacts of delayed standards are discussed in
Chapter 7  of the BID  for  the proposed  standards.  Although delayed
promulgation  of the  revised  proposed standards would not  impact current
levels  of  water pollution  or solid  waste,  such  a delay would result in
benzene emissions  from  furnace  and  foundry  plants remaining at the baseline
nationwide level of  nearly 26,000 Mg/yr.   Total emissions of benzene and
other  VOC  also  would remain  at  their baseline level  of about 171,000 Mg/yr.
No net nationwide  savings in energy use would be  achieved as  a result of

                                   1-4

-------
recovered coke oven gas if implementation of the revised proposed  standards
were delayed.

     1.2.4.3  Urban and Community Impacts.   The beneficial  urban and
community impacts of the revised proposed standards  include major
reductions in benzene emissions at plant  sites, many of  which  are  located
near highly populated areas.   This emission  reduction would reduce
substantially the health risk associated  with operation  of  coke by-product
recovery plants.   An added benefit to urban  and community areas is the  VOC
emission reduction for ozone  nonattainment  areas.
     The urban and community  economic impacts associated with  the  revised
proposed standards are discussed in Section  9.3.4  of the BID.  As  indicated
in this analysis, closure of  plants now in marginal  economic condition
because of market conditions  could occur  with resulting  community  impacts.
However, no closures are expected as a direct result of  these  revised
proposed standards.
                                 1-5

-------
                    2.  SUMMARY OF PUBLIC COMMENTS

     A total of 20 comments on the proposed standards  and  the BID  for  the
1984 proposed standards were received.   A list  of commenters, their
affiliations, and EPA docket number assigned to their  correspondence is
given in Table 2-1.
     For the purpose of orderly presentation, the comments have been
categorized under the following topics:
     Chapter 3
     Chapter 4
     Chapter 5
     Chapter 6
     Chapter 7
     Chapter 8
     Chapter 9
     Chapter 10
     Chapter 11
     Chapter  12
     Appendix A
    Appendix B
    Appendix C
    Appendix CW
 Selection  of Source  Category
 Selection  of Final Standards
 Emission Control Technology
 Environmental  Impacts
 Cost  Impact
 Economic Impact
 Quantitative Risk Assessment
 Equipment Leak Detection and Repair
 Recordkeeping and Reporting
 Miscellaneous
 Environmental Impact  Analysis
Cost Impact Analysis
Economic Impact Analysis
Health Risk Impact  Analysis
                               2-1

-------
        TABLE  2-1.   LIST OF COMMENTERS ON PROPOSED NATIONAL EMISSION
            STANDARDS  FOR COKE BY-PRODUCT RECOVERY PLANTS
Docket item number3
Commenter and affiliation
      IV-D-1
      IV-D-2
      IV-D-3
      IV-D-4
      IV-D-5
       IV-D-6
       IV-D-7
       IV-D-8
Ronald J. Chleboski, Deputy Director
Air Pollution Control  Bureau
Allegheny County Health Department
Pittsburgh, Pennsylvania  15201

George P. Ferreri, Director
Air Management Administration
Maryland Department of Health
  and Mental Hygiene
Baltimore, Maryland  21201

Alfred C. Little
Environmental Engineer
FMC Corporation
2000 Market Street
Philadelphia, Pennsylvania  19103

Danny L. Lewis
Assistant Plant Manager
Empire Coke Company
Birmingham, Alabama  35259

Daniel J. Goodwin, Manager
Division of Air Pollution Control
Illinois Environmental Protection
  Agency
2200 Churchill Road
Springfield,  Illinois  62706

Glen C.  Tenley, Vice President
Koppers  Company,  Inc.
1201 Koppers  Building
Pittsburgh,  Pennsylvania   15219

James R.  Zwikl
Director of  Environmental  Control
Shenango Incorporated
Neville  Island
Pittsburgh,  Pennsylvania   15225

D.  C. Miller, Resident Manager
Phosphorus Chemical Division
FMC Corporation
Box 431
Kemmerer, Wyoming  83101
 a Footnote at  end  of  table.
                         (continued)
                                 2-2

-------
                           TABLE  2-1  (continued)
Docket item number3
 Commenter and affiliation
      IV-D-9
      IV-D-10
      IV-D-11
      IV-D-12
     IV-D-13
     IV-D-14
     IV-D-15
 Donald C. Lang
 Director, Air and Water Control
 Inland Steel Company
 Indiana Harbor Works
 3210 Watling Street
 East Chicago, Indiana  46312

 Lucian M. Ferguson
 Executive Vice President
 American Coke and Coal  Chemicals
   Institute
 1800 M Street, N.W.
 Washington, DC  20036

 Lecil  M.  Colburn
 Jim Walter Corporation
 P.O. Box  22601
 1500 North Dale  Mabry
 Tampa,  Florida  33622

 R.  Wade Kohlmann
 Environmental  Engineer
 Citizens  Gas  and Coke Utility
 2020 North Meridan Street
 Indianapolis,  Indiana 46202-1306

 David D.  Doniger
 Senior  Staff Attorney
 Natural Resources Defense Council,
  Inc.
 1350 New  York Avenue, N.W.,
  Suite 300
 Washington, DC   20005

 Neil Jay King, Esq.
 Wilmer, Cutler & Pickering
 1666 K Street, N.W.
 Washington, DC  20006

 David M. Anderson, Director
 Environmental and Governmental
  Programs
Bethlehem Steel Corporation
Bethlehem, Pennsylvania   18016
  >otnote at end of table.
                                                             (continued)
                               2-3

-------
                           TABLE 2-1  (continued)
Docket item number3
Commenter and affiliation
      IV-D-16
      IV-D-17
      IV-D-18
      IV-D-19
      IV-D-33b
       and
      IV-D-34
Terry McGuire, Chief
Technical  Support Division
California Air Resources Board
1102 Q Street
P.O. Box 2815
Sacramento, California  95812

Moyer B. Edwards
Director,  Environmental  Control
Alabama By-Products Corporation
First National-Southern  National
  Building
P.O. Box 10246
Birmingham, Alabama  35202

Neil Jay King, Esq.
Wilmer, Cutler & Pickering
1666 K Street, N.W.
Washington, DC 20006

Barbara Patala, Acting Chairman
Committee on Environmental Matters
National Science Foundation
Washington, DC 20550

Michael A. Hanson
USS
208  South LaSalle Street
Chicago, Illinois  60604
 a  The  docket  number for this project is A-79-16.  Dockets are on file at
   EPA  Headquarters in Washington, DC, and at the Office of Air Quality
   Planning  and Standards (OAQPS) in Durham, North Carolina.

 b  Letters numbered IV-D-20 to IV-D-32 are correspondence regarding
   extension of the comment period, development of regulatory definitions
   for  furnace and foundry coke, and responses to information requests and
   are  not comments on the 1984 proposed standards or BIO.

                                 2-4

-------
                    3.   SELECTION  OF  SOURCE CATEGORY

 3.1   SELECTION  OF  SOURCE  CATEGORY
      Comment:   Commenters  IV-D-6,  IV-D-10, IV-D-12, IV-D-14, and IV-D-17
 question  the  selection  of  coke by-product recovery plants as a source
 category  for  regulation based on the benzene health risk estimates pre-
 dicted  at proposal  in 1984.  The commenters contend:  (1) the scientific
 basis of  the  health  risk estimates is not sufficient without verification
 by monitoring and  an epidemiologic study of an exposed community, (2) the
 benzene health  risk  is  low compared to other common risks or risks from
 other benzene source categories, and (3) the benzene health risk is less
 significant than estimated because of the exaggerated exposure assumptions
 applied to the  risk model.

    Response:   Specific responses are contained in Chapter 9 to the
 commenter's concerns regarding the methodology and assumptions applied to
 the quantitative risk assessment for coke by-product recovery plants.  The
 uncertainties and assumptions associated with the quantitative health risk
 assessment also are discussed in the preamble to the proposed rules
 (49 FR  23525),  in the preamble to the revised proposed rules, and are not
 repeated  here.  As discussed in the preamble  to the revised proposal, EPA
 determined that control  of this source  category is warranted to protect the
 public  health with an ample margin of safety.
 3.2  REGULATION OF MERCHANT PLANTS
     Comment:   Commenters  IV-D-4,  IV-D-6,  IV-D-7,  IV-D-10,  IV-D-11,
 IV-D-12, and IV-D-17 oppose the regulation of merchant plants.  The
commenters argue that merchant  plants generate  fewer emissions  compared
to larger furnace plants (or  other benzene source  categories)  and pose
little or no health risk.   The  commenters also  allege  that  the  estimated
costs  per merchant  plant,  the cost  per  incident  of leukemia,  and  the
                                3-1

-------
 overall economic impacts are higher than predicted and would adversely
 impact this industry segment.  The commenters also believe that the
 merchant plant segment was not represented properly in the BID for the
 1984 proposed standards.

     Response:  In considering the commenters' concerns, the data base
 has been revised since the 1984 proposal to indicate the environmental,
 health, and cost impacts of controls separately for furnace and foundry
 plants.  Merchant plants generally fall under the foundry plant industry
 segment.  As discussed in response to comment 6.2, emission factor
 adjustments have been made to account for the lower emission rates
 characteristic of foundry plants.   Environmental  impact estimates
 for foundry plants are shown in Appendix A.  As discussed in the preamble
 of the revised proposal, EPA determined that control  is warranted to
 protect the public health with an  ample margin of safety.
     The EPA does not agree that  foundry plants were represented
 improperly in the BID for the proposed standards.  The small-sized model
 plant (1,000 Mg/day of coke)  remains representative of sites in this
 industry segment—both in terms of capacities and processes practiced.
 Additionally,  the preproposal  economic analysis showed the impacts of
 control  on furnace and foundry plant industry segments.
 3.3  EXCLUSION OF FORM-COKE PLANTS
     Comment:   One commenter  in two comments (IV-D-3  and IV-D-8)  requests
 that the regulation be clarified to exclude form-coke plants.   In
 support, the commenter cites  separate conversations with EPA personnel
who stated that the 1984 proposed  standards were  not  intended  to include
 form-coke plants because the  process does  not result  in significant
 benzene emissions.

     Response:   In response to the commenter's concerns, the definition
of "coke by-product recovery  plant" under  Section 61.131 of the 1984
proposed standards has  been revised to exclude form-coke plants.   As
discussed in correspondence to the commenter on this  subject (Docket Item
 IV-C-10),  this  exclusion was  not made because of  the  absence of
significant  benzene emissions  from the form-coke  process.   Data are
insufficient to draw this conclusion,  although EPA would not expect

                                3-2

-------
 significant emissions based on a review of process description
 information.
     The EPA's major reason for excluding the form-coke process is that
 the form-coke process is too different from the coke by-product recovery
 process to apply the standards development study.  For example, only one
 form-coke plant currently is in operation.  This plant does not recover
 by-products.  Also, the form-coke plant has a fluidized bed process.
 Consequently,  potential  by-product  materials are different  in chemical
composition.  Because of the difference in chemical  composition,  the
process (and control) equipment also is different from equipment  (and
controls)  found at  plants using the conventional  coking process.
                               3-3

-------
                4.  SELECTION OF REVISED PROPOSED STANDARDS
  4.1   SELECTION OF LEVEL OF CONTROL
       Comment:  Many commenters indicate that the 1984 proposed standard
  should be either more stringent or less stringent.  For example,
  commenter IV-D-14 supports control levels of 90 percent on process
  vessels, tar storage tanks, and tar intercepting sumps, and commenter
  IV-D-13 recommends levels of control  that provide 100 percent benzene
  control regardless of costs.   Commenter IV-D-13 supports the most
  effective emission reduction  techniques for equipment leaks, storage
  tanks, and selection  of wash-oil  final  coolers  over tar-bottom final
  coolers.   Selection of wash-oil  final  coolers  (or similar,  equivalent
  systems)  also is  recommended  by  commenters  IV-D-5,  IV-D-9,  and IV-D-15.
 Conversely,  many  foundry coke  producers argue that  the  economic  impacts
 of the standard as proposed would  affect their  plants adversely.

      Response:  On July  28, 1987,  the United States Court of Appeals for
 the District  of Columbia Circuit handed down an _en bane decision in
 Natural Resources Defense Council. Inc.. v. EPA, 824 F.2d 1146
 (D.C.  dr.,  1987), hereafter referred to as "Vinyl Chloride", a case
 concerning the emission standard EPA set under Section 112 of the Clean
 Air Act for vinyl  chloride.  The Administrator reconsidered the proposed
 benzene standard for coke by-product recovery plants in light of the
 VlnyJ.  Chloride opinion.  For his reconsideration, the Administrator used
 the revised estimates  of nationwide emissions,  health risks, cost,  and
 economic impacts.   These estimates  were revised  after the 1984 proposal
 based on the consideration of  comments  received  on the proposal,  on
 information collected  from industry and  other  sources,  and  on additional
technical  and cost  analyses.   The specific  details  of  these  revisions  are
described  in  Chapters  6 through  9 of this document.
                                4-1

-------
     Because the Administrator  followed  a  new  policy  approach  in  the
reconsideration, his  selection  of  the  level  of the  standard  is being
published for comment in the  Federal Register  in  a  supplemental notice of
proposed rulemaking.   The difference between the  level of the  revised
proposed standard and the original  proposal  can be  found in  Chapter 1 of
this document.   The Administrator's policy and the  rationale for  his
decision, as well as  the legal  framework from  the Vinyl Chloride  opinion,
are described in the  preamble to the supplemental proposal.

4.2  REGULATORY DEFINITIONS OF  FOUNDRY AND FURNACE  BY-PRODUCT  PLANTS
     The control option chosen  by  EPA  for  the  revised proposed standards
would require different levels  of  control  for  final coolers  and
associated cooling towers at  furnace plants  than  at foundry  plants.  This
choice necessitated the development of definitions  of foundry  and furnace
coke and coke by-product recovery  plants for the  regulation.   The EPA
contacted the two industry trade associations, the  American  Iron  and
Steel Institute (AISI)  and the  American  Coke and  Coal  Chemicals Institute
(ACCCI)  to obtain additional  technical information  regarding these
definitions.  The related letters  and  telephone communications can be
found in Docket A-79-16.
     The resulting definition of foundry coke  is  coke that is  produced
from raw materials with an average of  less than 26  percent volatile
material by weight per charge/push cycle and that is  subject to a coking
period of 24 hours or more.   When  defining foundry  coke by-product
recovery plant, EPA recognized  that plants that predominantly  produce
foundry coke are typically merchant (non-captive  plants).  Because of
the fluctuating demand for foundry coke, some  of  these plants  also fill
some orders for furnace coke.  The EPA does  not intend that  these be
classified as furnace by-product plants, since they mainly produce
foundry coke.  However, as the  percentage  of foundry  coke increases,
there is a corresponding increase  in benzene emissions.  One reason is
that furnace coke production  is estimated  to yield  larger quantities of
benzene emissions per megagram  of  coke produced than  foundry coke. Also,
typically more furnace coke  can be produced  from  the  same coke oven
battery in a given period of  time  than foundry coke.   The EPA  judged that
                                 4-2

-------
a  reasonable consideration of these two factors would be to define
foundry plants as producing up to 25 percent furnace coke.
     Furnace coke is defined as any coke that is not foundry coke;
similarly a furnace coke by-product recovery plant is one that is not a
foundry coke by-product recovery plant.  These definitions avoid any
potential for coke production and by-product plants that are neither
furnace nor foundry.
     There are a few independent firms that make close to 50 percent
furnace and 50 percent foundry coke that would be considered furnace
coke by-product plants for this reproposed benzene regulation.   The
Agency does not believe that it is necessary to develop a special
category to examine every particular situation when developing  national
regulations.  However, the economic analysis used company-specific
financial  data to the extent possible and modeled these firms  as  being
merchant plants,  rather than captive to steel  companies.   The  analysis
shows no significant adverse economic affects  on these companies  with
control  alternatives that included wash-oil  final  coolers proposed to
control  final-coolers and cooling  towers at  furnace plants.
                                4-3

-------
                   5.   EMISSION CONTROL TECHNOLOGY

5.1  DEMONSTRATION OF CONTROL TECHNOLOGY
     Comment:  Two commenters (IV-D-10 and IV-D-12) claim that gas-
blanketing controls are no longer demonstrated and, consequently,  are
unproven.  The commenters cite closure of the Armco-Houston plant  and
claim that the controls are not demonstrated elsewhere.   One commenter
adds that the firm previously designing and constructing the controls no
longer participates in that business, implying that a lack of design  and
engineering services impairs "demonstration" of the Armco-Houston  system.
Also, commenters IV-D-7 and IV-D-14 allege that EPA conclusions  regarding
the system's safety are based on the limited experience  at Armco-Houston
and other plants.

     Response:  The EPA disagrees with these commenters.   Not only does
Armco-Houston's closure have no effect on the successful  use of  gas-
blanketing controls at this plant for the 4-year period  prior to closure,
but gas-blanketing systems currently are used at four other plant  sites.
     The systems used at other plants are described in Chapter 4 of the
BID for the 1984 proposed standards and in the preamble  to the proposal
in 49 FR 23530 (see also Docket Items II-B-45, II-B-46,  and II-B-47).
Gas blanketing has been used since 1960 in Plant A  at Bethlehem  Steel,
Sparrows Point.  In Plant B, the gas-blanketing system installed during
1954 was replaced during 1978 as part of the conversion  to a wash-oil
final-cooler system.   In Plants A and B, coke-oven  gas from the  wash-oil
scrubbers is used to blanket wash-oil  decanters, circulation tanks,
collecting tanks, and wastewater storage tanks.  Gas  blanketing  also  has
been used since 1960  at the Republic Steel-Cleveland  Coke  Plant  No. 1.
Updated in 1978, the  system currently is applied to wash-oil  decanters,
circulating tanks, rectifier separators, primary and  secondary light-oil
                                 5-1

-------
separators, condensers,  and final-cooler circulating  tanks.   In  Coke
Plant No. 2, clean coke-oven gas  from the battery  underfire  system  is
applied to primary and secondary  light-oil  separators,  rectifier
separators, and wash-oil  circulation  tanks.   At  the four  Bethlehem  and
Republic Steel  sites,  gas-blanketing  systems  were  installed  initially to
prevent oxidation and  sludge formation in light-oil plant  lines  and
equipment.
     At Armco-Houston, four gas-blanketing techniques were applied  to
light-oil and tar separation equipment.   The  system incorporated
blanketing from the gas  holder for  light-oil  recovery vessels, gas
blanketing from the collecting main for  tar decanters and  a  flushing
liquor-collecting tank,  negative  pressure venting  of  tar-collecting tanks
to the primary coolers,  and gas blanketing from  the wash-oil  final  cooler
(i.e., a slip stream of  wash oil  containing naphthalene is removed  and
routed to a wash-oil  decanter tank).
     The Armco-Houston system was installed between 1976  and  1977
according to an emission  control  agreement with  the Texas  Air Control
Board (TACB).  Prior to  1977, natural  gas had been used to underfire the
ovens; the coke oven gas  was flared with no by-product  recovery.
Although the plant had been scheduled  for shutdown in 1976, TACB agreed
to continued operation with installation of emission  controls.   The
system was operated for  4 years with  no  significant problems  until  the
plant closed in March  1981.  The  closure was  the result of economic
conditions, not failure  of the control  system.   Although  their shutdown
is unfortunate, it does  not detract from the  proven effectiveness or
viability of the emission control systems employed.   Thus, EPA does not
consider that the closure in any  way  affects  demonstration of the
controls or'application  of the system at other plants.
     One commenter mentions that  Koppers1  Engineering Construction
Division (which designed  and constructed the  Armco system) no longer
engages in that line of  business.   According  to  the commenter, this
impairs the "demonstration" of the  system.  The  EPA disagrees.   This
company's business decision has no  relevance  on  whether the  system  has
been demonstrated.  Other major engineering design and  construction firms
are available for this service.  In particular,  Oravo/Still  Corporation
has designed and installed a positive-pressure gas-blanketing system in
                                 5-2

-------
 an existing European coke by-product plant.   The system uses  clean
 coke oven gas (at about 1 inch [2.54 cm]  of  water positive  pressure)  to
 blanket a variety of storage tanks  and  process  vessels.   There  are  no
 domestic installations  of this system at  present.   However, Dravo/Still
 has had discussions  with at  least one U.S. coke plant  operator  about  such
 a system for the operator's  plant.

 5.2  SAFETY, DESIGN,  AND OPERATION  OF POSITIVE-PRESSURE  CONTROL SYSTEM
      Comment:   Commenters IV-D-3, IV-D-4,  IV-D-6,  IV-D-7, IV-D-14,
 IV-D-17, and IV-D-34  argue that  gas-blanketing  systems,  although appro-
 priate and  cost-effective for  some  plants, should  not  be mandatory  at all
 sites because  of safety,  design, and  operational  concerns.  One commenter
 states that in some  existing plants,  redesign of  the process operations
 and installation of  new equipment will  be necessary for  gas-blanketing
 systems to  work  safely  and effectively.  Without  these changes, the
 commenter questions the safety of positive-pressure blanketing systems,
 contending  that  leaks from older pieces of equipment that are difficult
 to  seal  effectively  (e.g., tar decanters and tar storage tanks) present a
 potential explosion or  fire hazard.   One of the commenters submitted a
 qualitative  comparative  study of the safety of gas blanketing for one of
 their plants.  The report  concluded that gas blanketing would involve a
 significant  increase in  risk to operating personnel and the surrounding
 community.   Other commenters argue that  leaks from covers, gaskets,  and
 connections  in the piping  system pose an explosion danger that is
 aggravated by the large number of sources, the presence of electrical
 equipment, and the vehicular traffic in  areas where blanketing systems
 would be  installed.  Two commenters  add  that  the probability of leaks
 (and the  associated safety hazard)  increases  with the  long pipe runs
 needed at some sites  to connect the  sources to the system.   Other
 operational  concerns  cited by the commenters  include the possibility of
 naphthalene clogging  in  cold  climates if steam or electrical  power for
 heated lines were lost and the chance of product contamination  (benzene
or light-oil) from the sulfur content of the  coke-oven  gases.

     Response:   The safety of recommended  control  systems should always
be considered,  and a  system considered inherently unsafe  would  not be
selected by  EPA as a  viable control  technique.   As discussed above in
                                 5-3

-------
response to comment 5.1,  gas  blanketing  has  been demonstrated as safe and
effective during an operating period  of  more than 24 years  (1960 to 1984)
at four plant sites in  addition  to  Armco-Houston.   In fact, in direct
contradiction to the commenters1  statements, EPA considers  that the
proposed system will  improve  the safety  level  found in uncontrolled
by-product plant environments.   The reasons  for this conclusion are
explained below.
     Leaks in a negative-pressure system are discussed in response to
comment 5.3, but the AISI  argues that even leaks in a positive-pressure
system may allow oxygen infiltration, causing  tank  vapors to reach
explosive limits and creating a  potential safety hazard.  The commenter
then cites preamble text  in 49 FR 23530  to support  this  assertion.  As
shown below, however, the  preamble  statement in 49  FR 23530 clearly
refers to the safety and  operational  advantages of  blanketing from the
gas holder, not to the  possibility  of explosion because  of  oxygen
infiltration:
     One advantage of blanketing with clean  coke oven gas
     from the gas holder  is the  elimination  of oxidation
     reactions between  oxygen in the  air and organic materials
     in the vessels. These reactions often  result  in a  sludge
     that may pose fouling and plugging  problems in lines and
     process equipment.  In addition, oxygen infiltration can
     cause tank vapors  to  reach  the explosive  limits of  vapor
     when tanks are periodically emptied or  when significant
     cooling takes place.  Applying a positive pressure  blanket
     would eliminate oxygen infiltration and maintain the vapor
     space in the tank  above  its upper explosive limit [emphasis
     added].
     The AISI also contends that "the low positive  pressure of the
proposed system is insufficient  to  alleviate explosive conditions if
leaks occur."  The standards  do  not dictate  an overall pressure level for
system operation.  The  system installed  may  be based on  positive or
negative pressure or on a  combination of the two.   The pressure
maintained will vary by necessity according  to the  type  of  source and
location of the connections to the  system (i.e., at the  main or the gas
holder) and overall process design.
                                 5-4

-------
      If, as the commenter asserts,  leaks  in the system occur or  the
 positive-pressure blanket fails,  the possibility of  an explosive
 atmosphere forming certainly is no  greater than the  possibility  under
 current plant conditions.  At most  uncontrolled plant  sites,  explosive
 conditions are now present.   Liquid organics  float on  the  surface  of open
 sumps and trenches and leak  from  equipment components  and  piping systems
 throughout the plant.   Organic vapors  also are  released  from  "breathing"
 tanks as air enters venting  systems or through  holes in  the  covers.  The
 breathing loss is recognized particularly  at  the light-oil condenser
 vent, where a continuous  steam purge may operate.  In  EPA's judgment,
 enclosing these sources and  ducting the emissions back to  the process via
 a closed positive-pressure system will  reduce substantially the  explosion
 hazard that now exists.   The EPA does  recognize that some  sources  at
 existing plants such as tar  decanters  and  tar tanks may  be in poor
 condition and will  require upgrading to accept  gas blanketing.   The
 necessary modifications for  typical  plants, however, have  been reflected
 in the cost estimates.
      The Agency also reviewed  the qualitative assessment submitted by the
 commenter to support the  contention  that gas  blanketing would involve a
 significant increase in risk to operating  personnel  and the surrounding
 community.   However, EPA  does  not believe that such a conclusion can be
 drawn  from  the  assessment for  several  reasons.  First,  the assessment is
 qualitative;  it  does not  draw  quantitative conclusions  as to the
 frequency of  a  major system  failure.   In the comparative risk assessment,
 probability  ratings were  assigned to various hazards  within the plant.
 For example,  for explosion potential under current plant conditions,  a
 probablity  rating of "D" which means "likely to  occur 1 time every  10
years" was  assigned.  With coke gas  blanketing,  the  explosion potential
was reduced to  "C" which means "likely to  occur  every 100 years."
However, with gas blanketing, higher ratings were assigned  to the
potential for explosion propagation, on-site safety,  and  financial  loss.
These types of ratings  were assigned to various  plant operations  and  to
various control scenarios.  The results were weighted and combined  to
provide a relative qualitative rating that  may be used  in evaluating
options in terms of economics and  safety.

                                5-5

-------
     Moreover, EPA does not believe the conclusions  in  the  report  are
warranted for the following reasons:   (1)  the report  did  not  utilize a
gas-blanketing design for the plant on which  to base  a  quantitative
comparision; without a specific design;  it is not  possible  to evaluate
safety features that could be engineered into the  system, (2)  the  assess-
ment was based on a review of the existing conditions in  the  plant,
without consideration of the substantial  upgrading of the coke plant
equipment that would be necessary to accommodate installation of a gas-
blanketing system, and (3) the report did  not provide any basis or
criteria for assigning the probability ratings or  consequence categories
that are reported.  After reviewing the assessment, EPA remains convinced
that the upgrading of equipment needed to  accommodate gas blanketing,
together with the installation of a control system that is  well-designed
with safety features included and that is  well  operated and maintained
will improve existing safety conditions at the sites.
     The EPA recognizes that leaks in a blanketing system will  occur
occasionally because of the gradual  deterioration  of  sealing  materials.
The prompt repair of these leaks, as  required by the  standards, not only
ensures proper operation and maintenance of the system  but  also promotes
safety by eliminating the leak sources.  With application of  a diligent
leak detection and repair program, the blanketing  system will  not  become
a "network of leaks," as asserted by  one commenter.   In fact,  if the
system is allowed to deteriorate, the owner or operator will  likely be
found in violation of the standards.
     Other commenters allege that leaks of pressurized  gas  from the
blanketing system will create a potential  explosion hazard  around
associated process equipment and that this hazard  is  aggravated by the
large number of sources, coupled with the  presence of electrical
equipment and vehicular traffic in gas-blanketed areas. The  EPA's review
of the safety aspects of the proposed system  does  not support this
contention.  Hydrogen and methane are the  major components  of coke-oven
gas, accounting for 69 to 97 percent  of the emission  stream.   According
to National Fire Code (NFC) guidelines,  these lighter-than-air gases
seldom produce hazardous mixtures (i.e.,  presenting a fire  or explosion
danger) in the zones where most electrical connections  are  made.
                                 5-6

-------
 Although special precautions such as explosion-proof electrical
 components may be required where light oil  or benzene is stored,  this
 equipment should be already in place at plants where the NFC or  plant
 safety codes have required their installation.  In addition, the  authors
 of the NFC guidelines state that, in their  experience,  it  generally  has
 not been necessary to classify as hazardous "locations  that  are
 adequately ventilated where flammable substances  are contained in
 suitable, well-maintained, closed piping systems  which  include only  the
 pipes, valves, fittings,  flanges, and meters." The NFC  recommends a
 common-sense safety approach.   The guidelines  encourage  using a
 positive-pressure system,  avoiding contact  with electrical equipment or
 using only intrinsically  safe  electrical  systems  with low  power needs, or
 applying a general  purpose enclosure to isolate the leak area (Docket
 Item II-C-132).
      Two commenters  assert that  the  safety  problem  increases with the
 long pipe runs needed  in  some  cases  to  connect the  sources to the system.
 Long pipe runs for  coke-oven gas  already  exist in many plants because the
 gas is used  as fuel  in other areas of the steel plant.  The EPA contends
 that a long  pipe  run associated with  a  coke-oven  gas-blanketing system
 poses  no  more  risk than even longer  pipe  runs  for transporting the
 coke-oven gas  throughout the plant.
     Prior to  proposal of  the  standard  in 1984, EPA thoroughly evaluated
 the safety aspects of gas-blanketing systems.  This review included
 visits  to each of the five plant  sites with blanketing systems to discuss
 safety  and operating problems with plant personnel.  As discussed in  the
 preamble  in 49 FR 23530, no safety or operation problems were reported
 that minimal,  routine maintenance would not  resolve (Docket Items
 II-B-45,  II-B-46, and II-B-47).  Appropriate safety features  also  were
 evaluated by an independent consultant (Docket Item II-B-49).  At  the
 time of the 1984 proposal, the  system included such features  as flame
 arrestors; an atmospheric  vent  on the collecting main or gas  holder to
 relieve excess pressure; three-way valves to lower the possibility  of
operator error; and steam-traced lines with  drip points,  condensate
traps, and steam-out connections (coupled with  an  annual  maintenance
check) to reduce potential  plugging problems.   Since the  1984

                                 5-7

-------
proposal, additional  features  have  been added such as water drains and
overflow connections  for  tar tanks  and liquid level sampling/gauging
instrumentation with  vapor-tight  seals.  Assuming each system is properly
operated and maintained after  installation, EPA considers that the
positive-pressure system  is a  safe  and effective control technique and
that leaks (if repaired as  required)  do not present the fire or explosion
hazard described by the commenters.
     The EPA agrees that  a  loss of  steam or electrical power for heated
lines may cause naphthalene clogging  in cold climates.  Unless a backup
power supply sufficient for the entire plant is available, EPA assumes
that such a power loss would affect most plant operations and probably
would result in a shutdown  until  power was restored.  Unfortunately, EPA
is aware of no other  reasonable approach capable of overcoming the
effects of cold climates.
     Nitrogen or natural  gas are  two  other possibilities for substitutes
to coke oven gas.  In fact, as described in Appendix B, the use of
nitrogen was costed by EPA  for blanketing benzene storage tanks because
of the possibility of contamination.  Factors relating to the selection
of blanketing gases for particular  types of sources are discussed in the
preamble to the 1984  proposed  standard at 49 FR 23530.  The revised
proposed standards do not dictate the type of blanketing gas to be used,
however.  Thus, nitrogen, natural gas, dirty or clean coke oven gas, or
any other gas can be  used as a blanketing medium for any of the affected
sources.
5.3  SAFETY, DESIGN,  AND  OPERATION  OF NEGATIVE-PRESSURE CONTROL SYSTEM
     Comment:  Commenters IV-D-4, IV-D-6,  IV-D-7, IV-D-14 and IV-D-17
argue that, in negative-pressure  systems,  air infiltration resulting from
ineffective sealing of older  vessels, operator error, or equipment
failure also craates  a potential  explosion or fire hazard.  For example,
failure.to close overflow pipes  during  filling or pumping out of
dehydrators could cause  air infiltration  in the collecting main.  Failure
of the control system when  a  light-oil tank car is loaded from the
storage tank could cause  the  vacuum relief valve to  function, creating  an
explosive atmosphere  in  the storage tank.  Failure of both the control
system and the vacuum relief  valve  could  cause a tank to collapse while

                                 5-8

-------
 emptying or to rupture while filling, causing a light-oil  spill  and
 possibly fire.  Commenter IV-D-14 also believes that  use of a
 negative-pressure gas-blanketing system requires additional  controls
 because of the potential  explosion hazard.   Specifically,  the  commenter
 states that continuous monitoring of the explosive  hazard  would  be
 necessary at three or four locations in the  gas distribution system.
 Also, an increase in oxygen concentration would require  such additional
 measures as automatic nitrogen  dilution with nitrogen  or enrichment with
 natural  gas to keep the coke-oven gas mixture below the  lower  explosive
 limit (or above the upper explosive  limit).

      Response:   The standards (and associated costs) are based on the use
 of a positive-pressure system because preproposal comments questioned the
 safety of the negative-pressure  system recommended  initially.  Although
 the use  or construction of  a negative-pressure  system  is not precluded by
 the regulation  in  any  way,  EPA encourages companies to install  safety
 equipment as  necessary in accordance  with their  historical safety
 policies  and  the  system's characteristics.
      Also recommended  is  the installation of  equipment included in the
 costs  for the positive-pressure  system  intended to alleviate many of the
 operating concerns  cited  by the  commenters (see  response to comment 5.2).
 For example,  operator  failure (on  a negative-pressure system) to close
 overflow  pipes during  filling or pumping out of dehydrators can be
 avoided by  installing  an overflow  pipe with  a liquid seal.   The potential
 for operator  error  also can be reduced by installing three-way  valves  so
 that tanks  are vented  at all times, either to the blanketing system or to
 the  atmosphere.
     The  commenters also point to light-oil  tank loading  operations where
 a control system failure (or control  system  failure  concurrent  with
 failure of a vacuum-relief valve) could lead  to an explosion  hazard.   If
 a storage tank is uncontrolled  (i.e., open to the atmosphere) as  in the
 current situation at most  by-product  plants,  such a  loading operation
would tend to draw vapors  back into the tank.  If a  tank  is controlled by
a negative-pressure system,  failure of the control system would cause the
vacuum-relief valve to function,  permitting vapors to be  drawn  into the
                                 5-9

-------
tank.  Therefore,  EPA considers  that  negative-pressure  system  failure
under the scenario suggested by  the commenters  presents  no more  danger
than similar situations encountered in  the  current  uncontrolled  plant
environment.  Failure of the control  system implies a pressure swing
within the system.  Concurrent failure  of a storage tank control  system
and vacuum-relief valves would cause  vacuum-relief  valves on other  parts
of the coke-oven gas system to function,  drawing oxygen from other
points.  Provisions for proper operation  and maintenance of  relief  valves
are included in the standard, however,  to minimize  the  potential  for  such
a failure.
5.4  MONITORING FOR CARBON MONOXIDE
     Comment:  Commenter IV-D-14 states that overpressurization  of  a
positive pressure system poses an explosive and occupational  hazard
because of  the carbon monoxide (CO)  released.  The presence  of CO
increases costs for additional monitoring and employee  training because
CO  hazards  currently do not exist. Similarly, commenter IV-D-6 states
that additional employees would be necessary for explosive  conditions
monitoring  or that hydrocarbon detection monitors should be  required on
every  (emphasis added  by commenter) piece of gas-blanketed  equipment.

     Response:  Coke plant  operators have stated that pressure control in
the collecting main  and gas  holder is  inherently reliable because large
pressure  fluctuations  can  cause serious  operating  and safety  difficulties
in  the operation  of  the coke-oven batteries  and the by-product plant.
Collecting  main  pressure  is  controlled by  an Askania valve at a  few
millimeters of water pressure,  and the pressure is  often watched and
 adjusted  manually if necessary.   Similarly,  the pressure in the  gas
 holder is also  carefully  controlled.   Overpressurization is prevented by
 bleeder or  pressure  relief valves and  water seals.
      No costs  were added  to the recommended gas-blanketing controls  for
 CO monitoring  because the existing and demonstrated systems,  installed at
 other coke  plants, did not have such provisions.   Therefore,  the
 monitoring question appears to  be one  of company  policy and site-specific
 conditions.  The revised proposed regulations  would not require CO
 monitoring, but EPA encourages  companies to follow their practice  of

                                  5-10

-------
 safety reviews and implementation of precautions based on each company's
 historical experience, its policy, and the site's characteristics.
      One additional point to consider is that a CO hazard from coke oven
 gas would not be unique to blanketed vessels.  Coke oven gas is handled
 in many parts of the coke plant,  which indicates that  a significant
 portion of the facility may currently pose a CO hazard.  For example,
 leaks of coke oven gas routinely  occur around the battery proper from
 lids, offtakes,  doors, charging,  and the collecting main.  Piping for
 coke oven gas winds through the plant, and the gas is  treated in enclosed
 vessels such as  primary coolers,  direct-water coolers,  and  scrubbers.
 The gas also is  piped to the battery underfiring system and  is  used in
 other parts  of the steel  plant.   The gas-blanketed equipment  is  required
 to be enclosed and sealed and,  consequently,  should not  be more  prone to
 leaks than other  equipment  that handles  coke  oven  gas.   If a  company's
 current policy requires  detectors  and  monitors  for every  point that
 contains  coke-oven gas,  then  consistent  application  of  safety policy
 would require them for blanketed  vessels.
 5.5   SUMP  CONTROLS
      Comment:  Two commenters (IV-D-4  and  IV-D-17)  believe that  covering
 and  sealing  sumps  create  a  fire or explosion  hazard  from  concentrated
 fumes  because no gas or steam can be used  for purging.   One commenter
 states  that the purpose of  leaving open sumps and trenches is to prevent
 such  a  hazard, and  at  his plant tramp  steam is discharged routinely  into
 sumps  and trenches  to  reduce the possibility of fire.
     Response:  Two points are relevant in response to  this comment:  (1)
 steam purging  increases emissions  of and exposure to hazardous organic
 compounds, and (2) alternatives  exist to detect and correct hazardous
 conditions.
     Steam purging strips organic  compounds from the sump and can be
 especially efficient at removing volatile compounds such as  benzene.
 Most sumps are installed below grade; consequently, workers  and  others  in
the plant can be  exposed to locally high  concentrations  of these organic
compounds at  ground level  from an  uncontrolled sump, especially  with a
purge gas.
                                 5-11

-------
     The current  practice  of  discharging  tramp  steam  to  an  open  sump
already poses a hazard  if  concentrations  are  high  enough to be explosive.
In addition,  the  steam  purging  may  create the movement of explosive vapor
from the sump to  ground level if a  surge  or  slug of organic material
accidentally  entered the sump during purging.  The EPA's costs include  an
air-tight seal  and a vent  to  the atmosphere  for safety.   The operator may
choose other  measures to increase safety, such  as  including a flame
arrester on the vent or installing  detectors  for explosive  conditions.
Other alternatives are  replacing the sump with  an  above-grade closed tank
that may be easier to keep air-tight, or  separating organic compounds
upstream of the sump so the sump will not contain  explosive gases.
The solution  to the commenter's question  will depend  on  each site's
specific conditions and each  company's policy.
5.6  OPERATIONAL  PROBLEM FROM PLUGGED VENTS  OR  VALVES
     Comment:  Commenter IV-D-17 suggests that  mechanical vents  and
pressure relief valves may be fouled easily,  resulting in ruptured  tanks.
The commenter adds that many ruptured tanks  occur  as  the result  of
plugged valves that were supposed to relieve pressures.
     Response:  The EPA recognizes that plugged vents or valves  pose
an operational problem and potential safety   hazard if not repaired.   For
this reason, the  revised proposed regulation requires an annual
maintenance  check for abnormalities  such as   plugs, sticking valves,  and
clogged or improperly operating condensate traps.   A first attempt  at
repair  of  any  defect must be made within 5 days, with any necessary
repairs made within 15 days of  inspection.   The regulation  requires that
records containing  a brief description of any  abnormalities, the repairs
made,  and  the  dates of  repair be maintained  for a minimum of 2 years.
Although the regulation requires a maintenance inspection only once a
year,  plant  owners  or  operators may  want to  consider performing this
maintenance  check more  frequently,  such  as in  conjunction with the
semiannual leak  inspection.

5.7   CONTROLS  FOR BENZENE STORAGE TANKS
      Comment:  Commenter  IV-D-13 requests that EPA determine whether any
benzene storage  tanks  at  by-product  plants are equipped with shingle

                                 5-12

-------
 seals.  If so, the commenter recommends that the regulation require any
 shingle seals to be replaced with continuous seals.   In support,  the
 commenter cites the Federal  Register notice of withdrawal  for benzene
 storage tanks (49 FR 8386,  March 6,  1984).   The notice states in  part
 that about 12 percent of existing benzene storage tanks in the chemical
 and petroleum industries have shingle seals, which are far less effective
 than continuous seals.

      Response:  The shingle  and  continuous  seals to  which  the commenter
 refers are the seals on  floating roofs  in tanks.   Many of  the tanks  in
 by-product plants are horizontal  or  an  older riveted design.   The  EPA
 does not know of any benzene storage tanks  with  floating roofs in
 by-product plants.   The  controls  EPA has  analyzed for storage tanks  are
 wash-oil  scrubbers  and gas blanketing.  These  controls  are  applicable to
 horizontal  tanks and would not require  major tank modification (unless  a
 tank is  in extremely poor condition).   The  revised proposed standard  does
 not require control  of these tanks,  however.

 5.8  DETERMINATION  OF CONTROL EFFICIENCIES
      Comment:   Commenter IV-D-9 asks  how  efficiencies of 90,  95, and  98
 percent  are determined under the  standard.

      Response:   The  90-percent control  efficiency applicable  to wash-oil
 scrubber controls is based on design  calculations.  A full  description of
 the methodology  and  design parameters is  contained in Docket  Items
 II-B-51 and  IV-J-1;  a summary description is provided in Chapter 4 of the
 BID for the  proposed standards.  A 95-percent control efficiency for the
 tar decanter was derived by  adjusting the control efficiency  for
 enclosure and  gas blanketing  (98 percent)  downward to account for
 uncontrolled emissions from the approximately 13 percent of the liquid
 surface of the decanter that must remain open to allow clearance for the
 sludge conveyor.  A 98-percent control efficiency has been  established
 for  gas-blanketing systems  and sealed enclosures (e.g., the light-oil
 sump).  As discussed in the  preamble  to the  1984 proposed  rule in
49 FR 23529, the theoretical  efficiency of source enclosure with gas
blanketing approaches 100 percent. However, this efficiency cannot be
expected to be maintained continuously for the service life of the

                                 5-13

-------
equipment because of the eventual  deterioration  of  seals  and  sealing
materials.  Because deterioration  of piping,  seals,  or  sealing  materials
can occasionally result in leaks,  engineering judgment  was  applied  to
reduce the overall  control efficiency to 98 percent.
     Installation of the specified equipment  demonstrates compliance
with the standard for these sources.  In other words, these control
efficiencies are assumed to be achieved if the required equipment is
applied.  However,  design calculations and verifying test data  will be
needed if the owner or operator wishes to apply  for permission  to use an
alternative means of emission limitation.
5.9  GAS BLANKETING VERSUS WASH-OIL SCRUBBERS
     Comment:  Commenter IV-D-14 recommends that the standards  permit the
use of a 90-percent efficient control device (e.g., a wash-oil  scrubber)
in lieu of gas blanketing on process vessels, tar storage tanks,  and
tar-intercepting sumps.  The commenter argues that  use  of the wash-oil
scrubber would provide essentially the same health  benefit  as gas
blanketing.  Specifically, the commenter suggests that  the  control
efficiency of blanketing at an older plant may be lower than  98 percent
because of more likely leakage and downtime, and a  wash-oil scrubber  may
achieve higher than 90-percent control.

      Response:  The control efficiency of gas blanketing theoretically
is 100 percent.  For conservative comparisons with  other  controls,  this
efficiency has been reduced to the value of 98 percent  to account for
occasional leakage from seals or sealing materials.  Leak detection and
repair requirements are included in the gas-blanketing  standards to
ensure that 98 percent control or greater is maintained through proper
operation and maintenance of the equipment.  Thus,  EPA  does not expect
well-designed, well-operated, and well-maintained gas-blanketing systems
to achieve less than 98 percent control efficiency.  Although it is
acknowledged in the BIO for the 1984 proposed standards  (page 4-28a)  that
an efficiency higher than 90 percent  (e.g., 95 percent  or greater)
theoretically may  be achieved, the  parameters have been developed to
ensure that  all plants using this technique could achieve 90 percent
control  continuously.  Thus, at proposal, EPA considered gas blanketing
                                 5-14

-------
 compared to wash-oil scrubbing on a common basis of conservative
 estimates of control efficiencies.  Similarly, for the revised proposal,
 the Agency believes that a common basis of representative, if somewhat
 conservative, control efficiencies should be applied.  If the two control
 techniques are compared on this basis, wash-oil  scrubbers are less
 effective than gas blanketing and may be more costly from a nationwide
 perspective.

 5.10  FINAL COOLERS AND NAPHTHALENE PROCESSING
      Comment:  Commenter IV-D-15 requests that the standards  allow use of
 a technology for controlling naphthalene processing and  final-cooler
 cooling tower emissions that the commenter claims  to be  more  effective
 than the tar-bottom final  cooler on which the proposed standard  was
 based.   This system,  recently patented by his company, eliminates  cooling
 tower emissions  through indirect heat  exchangers and reduces  (but  does
 not eliminate) emissions  from naphthalene processing by  enclosing  the
 separator and froth flotation units.   The commenter estimates  benzene
 emissions from the processing of naphthalene  skimmings at  a maximum  of
 20 grams (g)  of  benzene/Mg of coke.  The  commenter  claims  that this
 system  achieves  a  95-percent  benzene emission  reduction  from baseline
 compared to  81 percent  for tar-bottom  final coolers  and has lower
 capital, operating, and energy  requirements than do  wash-oil final
 coolers.  According to the commenter,  use  of  a single liquid phase
 (water)  prevents the problem  of water  and  oil emulsion found in wash-oil
 final coolers.
       Commenter IV-D-9 states that the regulation is unclear regarding
 the use  of alternatives to the proposed zero emission limit for
 naphthalene processing and use of the tar-bottom final cooler.  The
 commenter's company proposes to convert an existing direct-water final
 cooler to a closed-loop recirculated-water final  cooling  system.   This
 system would use flushing liquor to cool coke-oven  gas and heat exchanger
 "closed-to-the-atmosphere" mode of operation.   The  commenter believes that
this system, in conjunction with gas blanketing the tanks used to hold  the
flushing liquor that contains naphthalene, would  comply with the  zero
emission limit for  naphthalene processing in a cost-effective  manner.
                                 5-15

-------
The installed cost of this  system is  estimated  at  $8  million  compared  to
$12.5 million for the tar-bottom final-cooling  system shown in  the  1984
proposal BID.
     Response:  Both commenters are suggesting  the use of  alternatives to
tar-bottom final coolers for reducing benzene emissions from  final-cool ing
operations at plants.  Commenter IV-D-15 proposes  a system that will
eliminate final-cooler cooling tower emissions, but that will  allow some
emissions from naphthalene processing.  Commenter  IV-D-9 proposes a system
that eliminates final-cooler cooling tower emissions  and also absorbs
naphthalene in tar, thus avoiding the need for  physical separation  and
processing of naphthalene and its attendant emissions.  These comments
plus cost estimates included in the comment letter of commenter IV-D-14
for tar-bottom and wash-oil final coolers led EPA  to further  investigate
technical and cost data for final-cooler control technologies.  Appendix B
contains a discussion for revisions to the control cost estimates  since
the 1984 proposal in addition to the revised cost  estimates for final
cooling.
      Estimated capital and operating costs were requested for the
proprietary  indirect cooling system proposed by commenter IV-D-15.
Information  also  was requested  on the degree to which  the proprietary
indirect technology  had been demonstrated and  limitations on its
applicability.  When asked  about wash-oil final-cooler costs in commenter
IV-D-141s  letter, the commenter responded instead with  capital cost
estimates  for two indirect  cooling schemes proposed  for application to an
existing U.S.  coke  by-product  plant.  Technical information on various
 indirect cooling  schemes was  provided to  EPA by Dravo/Still, an engi-
 neering firm that designs  coke by-product plants  and associated control
 systems.
       The term "indirect"  is  used  in  two  contexts  when discussing  final
 cooling.  One context  refers  to cooling of the coke-oven  gas where there
 is no direct contact between  the cooling  fluid and the coke-oven gas.   In
 the other context,  there  is direct contact of  the cooling fluid with  the
 coke-oven gas, but  the cooling fluid itself  is cooled indirectly.  Both
 types of indirect final cooling eliminate benzene emissions  from the
 final-cooler cooling tower that result  when  direct contact water is cooled

                                  5-16

-------
 atmospherically.  By the above definition,  wash-oil  coolers,  one of the
 control  options considered in the 1984 proposal  BIO  as  more  stringent  than
 tar-bottom final coolers,  is an indirect  final  cooler.
       All  the indirect final-cool ing schemes  must  in some  way deal  with
 naphthalene remaining in the coke-oven gas  just  prior to light-oil
 recovery.   The ways in which naphthalene  is condensed,  absorbed,  dissolved
 or otherwise removed from  the coke-oven gas prior  to or during indirect
 final  cooling yield varied potentials  for benzene  emissions  from further
 handling/processing of naphthalene-containing liquids.
       The  proprietary indirect  cooling technology  proposed by commenter
 IV-D-15, when used to replace a direct-water  final cooler, generates a
 liquid stream containing naphthalene that is  processed  in the same  way  as
 the liquid stream from a direct-water  final cooler.   The commenter  has
 suggested  enclosing the froth flotation/gravity  separation equipment to
 reduce benzene emissions,  but he makes  no suggestion  with respect to melt
 pit/drying tank  emission control.  The  commenter's proposal would reduce
 final-cooler benzene  emissions  by 70 g/Mg of  coke  over  that required by
 the 1984 proposed  standard,  and  it would  reduce emissions from
 naphthalene-processing  from  107 g/Mg of-coke  to 28.8  g/Mg (assuming a
 90-percent  efficient  wash-oil scrubber  could  be applied to the emissions
 from the flotation/separation enclosure).    Because the  1984 proposed
 standard would have  required  zero benzene emissions from these
 naphthalene-processing  operations, this scheme would not comply with the
 proposed or  revised proposed  regulation.  If the technology recommended by
 commenter  IV-D-15 is  applied to a plant with a tar-bottom final cooler or
 mixer-settler  (wherein  naphthalene is absorbed into the  tar), then the
 technology would comply with the regulation  as proposed  in  1984.
      Commenter  IV-D-15 submitted information  that  indicated  the indirect
 cooling technology was tested for a 13-week  period  using full-scale  plant
 equipment.   This pilot demonstration  program yielded  enough data to
 permit a design of full-scale installation for a  plant approximately the
 size of Model Plant 3 in the 1984 proposal BID.   As of August 1985,  a
 full-scale  installation of  this technology was completed in a  Canadian
 steel plant.  The Canadian  plant has  an existing  tar-bottom final  cooler,
thus eliminating the need for processing and handling of naphthalene.

                                 5-17

-------
However, the indirect  final  cooler  has  not yet been operated for an
extended period.   Capital  and  annualized cost estimates generated from
data submitted by commenter  IV-0-15 indicate that the costs for their
proprietary technology are in  the  range of cost  estimates for tar
mixer-settler/tar-bottom final-cooler equipment.
      The indirect final-cooling technology  suggested by commenter  IV-D-9
avoids physical separation and processing  of naphthalene by absorbing
naphthalene in tar circulated  with the flushing  liquor  through the  final
cooler.  Provided the tanks used to accumulate the  naphthalene-containing
flushing liquor are gas blanketed  or closed  to the  atmosphere, this
final-cooling scheme would be  equivalent  to  the  originally  proposed
gas-blanketed tar mixer-settler for naphthalene  operations  emission
control.   It would exceed the mixer-settler's  performance with  respect  to
final-cooler emissions.   Information supplied  by Oravo/Still  indicates
that  this  type of indirect final-cooler system is in use at the  LTV
Aliquippa  plant  (whether  gas blanketing of accumulating tanks is in use
is  not  known).   The capital cost estimate provided by commenter IV-D-9
for his  plant  is  about  70 percent  higher than EPA's current estimate for
a wash-oil  final  cooler applied to that plant size.
       Dravo/Still provided  information about two other potential indirect
 final cooling  schemes.  One scheme uses warm wash-oil absorption to
 remove naphthalene  from the coke-oven  gas stream in the first stage of
 the final  cooler.   The second  stage  of the  final cooler uses water to
 cool  the coke-oven  gas.  This  cooling  water is  itself cooled in an
 indirect wet surface  air  cooler.   The  warm  wash  oil, containing naphtha-
 lene, is sent to the  light-oil  still  equipment  or  to a naphthalene
 stripper to separate  naphthalene  from the wash  oil.  Naphthalene vapors
 are  returned to the coke-oven gas suction main  upstream of the  primary
 coolers.  This recirculation  system for  naphthalene  leads  ultimately to
 excess naphthalene being accumulated in  the recovered  tar.   The warm
 wash-oil  absorption system is in  use at  the Armco-Middletown plant.
 However,  the Armco plant uses an  atmospheric  cooling tower for  the cooling
 water  rather than an indirect cooler.  The  system as described  by
  Dravo/Still eliminates benzene emissions from both naphthalene  processing
  and  the final-cooler cooling tower.  A capital  cost estimate for  this

                                  5-18

-------
 indirect cooling technology  applied to an existing U.S. plant was provided
 to EPA by commenter IV-D-14.  That estimate indicated the capital cost was
 about the same as EPA's  estimate  shown in this document for a wash-oil
 final cooler applied to  that  plant size.
       The other indirect final-cooler scheme discussed by Dravo/Still
 avoids direct contact between the cooling liquid and the coke-oven gas.
 Indirect cooling is achieved  in a cross-tube cooler with water flowing
 through the  tubes and gas  flowing outside the tubes.  To prevent naphtha-
 lene fouling of the heat exchanger surface, tar is injected into the
 cooler on the gas side where  it mixes with condensing water and keeps
 naphthalene  in suspension.  The water-tar-naphthalene mixture withdrawn
 from the cooler is  recycled to the collecting main.  As in the above
 system,  excess naphthalene leaves the by-product plant in the recovered
 tar.   The cooling water  has not contacted the gas stream, so it may be
 cooled in an atmospheric cooling  tower without generating benzene
 emissions.   This  system  also eliminates benzene emissions from both
 naphthalene  processing and final  cooling.  According to Dravo-Still,  this
 type  of  indirect  final cooler is  in use at a Dofasco plant in Hamilton,
 Ontario.   Commenter  IV-D-14 provided a capital  cost estimate for this
 system applied  to an  existing U.S. plant.  That estimate indicated  that
 the  capital  cost would be about 28 percent higher than the current  EPA
 estimate  for  a  wash-oil final cooler applied to the same plant size.   One
 reason that  the cost  of this indirect  final-cooler system may be higher
 than  the  one  described above is that  it  is difficult  to make use of
 existing  plant  equipment in retrofitting  the latter system.
       Based on the information and data  presented above,  all  of the
 indirect  final-cooling schemes except  that of  commenter IV-D-15 achieve
 equivalent benzene emission control  for  naphthalene processing and
 handling operations when compared  to the  tar-bottom final  cooler or tar
 mixer-settler required by the proposed  regulation.   All  of the described
 indirect final-cooling schemes produce greater  final-cooler  benzene
 emission reductions than would be  achieved through  installation  of
 tar-bottom final coolers  or tar  mixer-settlers.
     Based on the revised environmental,  risk,  and  cost  data,  the
Administrator has decided to  repropose wash-oil final-cooler technology

                                5-19

-------
as the basis for the naphthalene-processing standard  for  furnace  plants.
Wash-oil  and tar-bottom final  coolers  have been  applied at  several  plants.
However,  any system (including the ones  discussed  here) that  meets  the
zero emission limit for naphthalene processing could  be used  in foundry
plants, and any system that  meets  both the zero  emission  limit for
naphthalene processing and for final coolers (and  associated  cooling
towers) could be used at furnace coke  plants.
                                 5-20

-------
                       6.   ENVIRONMENTAL  IMPACTS

 6.1  DATA BASE  FOR ENVIRONMENTAL  IMPACTS
      Comment:   Commenters  IV-D-9  and  IV-D-14  state that the health  risk
 from by-product plants is  less  significant than  projected at proposal
 because nationwide benzene  emission estimates are overestimated as  a
 result of the effect  of plant closures and reduced battery capacities.
 One commenter estimates nationwide benzene emissions to be 21,800 Mg/yr
 compared to  the 24,100 Mg/yr estimate in the preamble (49 FR 23525).

      Response:   The interim status of the estimated environmental impacts
 was acknowledged in the preamble to the 1984 proposed standards in
 49  FR 23524.  As stated, the impacts were calculated initially from a
 data base of 55  plants.  Industry data and information from the U.S.
 Department of Energy  (DOE) received prior to proposal indicated that 13
 of  the  55 plants  had  been closed.  Information was not available,
 however,  to determine whether all reported closures were permanent.
 Consequently, the  preamble presented environmental  impacts based on  42
 plants  and stated  that the impacts and calculations in the BID  would be
 revised  following  proposal.
      The  data base has been revised since the 1984 proposal  in  several
 respects.  Information regarding permanent  closures,  changes  in battery
 capacities, and  changes or corrections in site-specific  operating
 processes have been applied to  reflect the  industry  operating status as
 of November 1984.  These data were supplied  by individual  companies  and
 by the two major industry  trade  associations—the American  Iron and  Steel
 Institute (AISI) and the American  Coke and Coal Chemicals  Institute
 (ACCCI).  As  discussed below in  response  to comment 6.2, adjustments to
emission factors also  have  been  made since the 1984 proposal to indicate
lower emission rates  from  sources  at plants producing  foundry (or  furnace

                                6-1

-------
and foundry)  coke.   These  revisions  have  been made to account for the
combination of lower light-oil yields  and lower benzene concentrations
for foundry coke plants  compared  to  concentrations for furnace coke
plants.   For this reason,  the  data base also has  been segregated to show
separately as well  as combined the environmental  impacts of control
options  on furnace  and foundry plant industry segments.
     Tables A-1 and A-2  (Appendix A)  reveal a potential total of 44
furnace  and foundry plants with  a combined operating capacity of 50.9
million  Mg/yr of coke.  Of the 44 plants, 30 produce furnace coke, and  14
(mainly  merchant plants) produce  foundry  (or furnace and foundry) coke.
Of the 30 furnace plants,  6 are  on cold-idle as of November 1984.  These
plants have been identified as follows:   (1) LTV  Steel--Thomas, Alabama;
(2) LTV  Steel—E. Chicago, Indiana;  (3) U.S. Steel—Fairless Hills,
Pennsylvania; (4) U.S. Steel—Lorain,  Ohio;  (5) U.S. Steel--Fairfield,
Alabama; and (6) Wei rton Steel—Brown's  Island, West Virginia.  Also,  1
of the 14 foundry plants (Alabama By-Products—Keystone, Alabama) is  on
cold-idle as of November 1984  .   Because  information is  insufficient  to
predict whether these temporary  closures  will become permanent, these
seven plants have not been deleted  from the data  base used to estimate
the environmental impacts of the revised  proposed standards.  The
deletion of six furnace plants from  the data  base would  reduce  the
operating capacity of this industry  segment from  about  45.8 million  Mg/yr
of coke to about 39.2 million Mg/yr  of coke nationwide.   Foundry  plant
operating capacity would be reduced  by about  8  percent  (from  about  5.1
million Mg/yr to about 4.7 million  Mg/yr) if the  cold-idle plant  were
excluded  from the data base for this industry  segment.
     Tables A-3  and A-4 (Appendix A) display the  operating processes
practiced  at  each furnace or  foundry site, as  reported  by the individual
plants  (Docket  Items  IV-D-1 through IV-D-18).   Based on these data,
half  (15  of  30)  of  the  furnace plants practice naphthalene handling and
processing,  a major  source of benzene emissions.   Direct-water final
coolers  and  tar-bottom  final  coolers are  used at 16 and 4 furnace plants,
 respectively;  5  furnace plants use wash-oil final coolers.  Although tar
 recovery  sources (e.g., tar decanters, dewatering, sumps, and storage)
 are  found at  all 30  sites, as are most light-oil  plant sources (light-oil
                                  6-2

-------
 decanters,  sumps, storage, and wash-oil circulation tanks), BTX storage
 is  practiced at only 10 sites, and benzene is stored only at 4 sites.
     Table  A-4 indicates that naphthalene handling and processing also is
 practiced at half (7) of the 14 foundry plant sites.  Reported data show
 direct-water final coolers at seven plants, tar-bottom final coolers at
 two plants, and no wash-oil final coolers in use.  Although tar recovery
 sources are present at each site, light-oil storage is found at 9 of the
 14  sites.   Benzene and BTX are stored at one plant.
     Tables 7-1 through 7-6 of the BID for the 1984 proposed standards
 have been revised to show the updated estimated environmental  impacts.
 These tables display the estimated baseline nationwide benzene emissions
 and process capacity data for sources at the 30 furnace plants, the 14
 foundry plants, and the total industry combined.   Comparable data for
 total VOC emissions (benzene and other VOC) also are shown.  Based on
 these data, estimated nationwide benzene emissions from the industry
 total nearly 26,000 Mg/yr;  VOC emissions are about 171,000 Mg/yr.
     The effects control  options have on reducing benzene and total  VOC
 emissions also are shown in Appendix A.  Implementing the revised
 proposed standards would reduce overall benzene emissions from furnace
 and foundry coke producers  from approximately 26,000 Mg/yr to about
 2,000 Mg/yr, a reduction of about 93 percent.  Nationwide VOC emissions
 also from these sources would be reduced from approximately 171,000  Mg/yr
 to  about 6,000 Mg/yr.
     The revised data  base  and foundry plant emission factors  have little
 effect on the impacts  or benefits of other environmental  considerations
 associated with the final  standards,  such  as energy requirements,  water
 pollution, solid waste disposal,  and noise or odor levels.   As  discussed
 in  the preamble to the 1984 proposal  notice (49  FR 23525),  a nominal
 increase in electrical  or  steam requirements at  furnace  plants  could
 occur if gas blanketing piping were  heated to prevent vapors from
 condensing or freezing in vent lines.   Tables A-ll and A-12  show energy
 use and coke-oven  gas  recovery estimates  for model  furnace  and  foundry
 plants.
     Although no water pollution  problems  are associated  with  recycling
benzene vapors,  implementing  the  revised proposed  standards  could  result
in an increased  HCN  concentration  at  plants  using  indirect  final cooling.
                                 6-3

-------
As discussed in the BIO  for  the  1984  proposed  standards  (page 7-7), HCN
is emitted currently from the  final-cooler  cooling tower at some plants
by air stripping of wastewater.   Measured HCN  air emissions and
calculations based on once-through  cooling  water indicate that about 200
g/Mg of coke could be added  to wastewater for  treatment, if indirect
cooling rather than direct cooling  were used (Docket  Item II-B-30).  The
actual amount of additional  HCN  in  the wastewater could  depend on  cooling
water temperature, degree of recycle  practiced,  and  numerous  other
factors.
     As suggested by the commenters,  the effects of  reduced operating
capacities and revised emission  factors have been taken  into  account  in
the updated risk assessment.  Further information  regarding the  calcu-
lation of revised emission estimates  for furnace and foundry  plants
is discussed below  in response to comment 6.2.

6.2   FOUNDRY PLANT  EMISSION FACTORS
      Comment:  Commenters IV-D-6, IV-D-7, IV-D-10,  IV-D-11,  and  IV-D-12
claim that  operating dissimilarities result in fewer emissions compared
to emissions from  furnace coke plants.  The commenters state  that foundry
plants generate  fewer emissions because of:   (1) the use of less volatile
coal  in their  feed  (21 to 22  percent volatile matter in foundry  blends
versus 28 to 30  percent  volatile matter in  furnace coal  blends), and (2)
the  use of  longer  coking  cycles  (28 to 30  hours for foundry coke versus
14  to 16  hours for furnace  coke).  In  support, one commenter also states
that the  percentage of  benzene  in  light oil at his plant is 55 to 60
 percent,  considerably less  than  the  70-percent example shown in Table 3-6
 of  the BID  for the 1984  proposed standards.   Another commenter maintains
 that merchant  plants generate fewer  emissions than furnace plants not
 only because of different operating  practices but also because of the
 relative  size  of the industry segment  compared to furnace coke plants.

      Response:  In response to  the public  comments  received  on this
 issue, EPA has reviewed available  information and data  to determine
 whether the development of separate  emission  factors  for foundry  and
 furnace coke production is warranted..  Based  on  results of this  review,
                                  6-4

-------
 EPA  agrees with the commenters' contention that benzene emissions from a
 foundry coke by-product plant would be expected to be less than the
 emissions from a furnace coke by-product plant of similar capacity.
 Because no emission measurements were performed in foundry coke plants
 during the 1979 to 1980 sampling survey, appropriate emission factor
 adjustments have been made based on available data for light-oil yields.
     Foundry coke is produced from a coal mixture that generally contains
 less volatile matter than the mixtures used to produce furnace coke.  The
 ACCCI comments suggest that typical furnace-coke coal mixes contain 28 to
 30 percent volatile matter and foundry-coke coal mixes contain 21 to 22
 percent volatile matter.  This statement is confirmed, in part, by data
 contained in one primary reference source on the coking of high- and
 low-volatile coals (Docket Item IV-J-5) that show light-oil yields
 as significantly lower (less than half) for the low-volatile coals.
 However, definitive data on light-oil  yields published by the DOE show
 that, over a 4-year period, the light-oil yields in merchant coke by-
 product plants (mostly foundry coke producers) averaged about 66 percent
 of those in furnace plants on a per-ton-of-coal-charged basis (Docket
 Items II-I-43, II-I-50, IV-J-2, IV-J-3, and IV-J-4).   These yields are
 shown in Table A-13 of Appendix A.  Table A-13 also provides data on the
 relative yields of tar and coke-oven gas in merchant  coke plants compared
 to furnace coke plants.  The data displayed in Table  A-13 represent the
 principal  basis for the technique used to adjust the  proposed emission
 factors for foundry coke producers.
     Based on a review of data contained in another coke-making reference
 source (Docket Item II-I-2),  EPA also  agrees with commenters who suggest
 that the lower coking temperatures associated with foundry coke produc-
 tion compared to furnace coke production (for the same coal) would lead
 to production of less by-product benzene.  In support, one merchant plant
 commenter  indicated that the  light oil  from foundry coking contains 55 to
 60 percent benzene compared to the 70  percent assumed in  the 1984 pro-
 posal BID  (Docket  Item IV-D-7).   Based on an informal  poll  of some member
 companies, ACCCI  provided an  average estimate of 63.5 percent for foundry
producers  (Docket  Item IV-D-7).   For furnace coke production, however, a
benzene content for light oil  of 70 percent is still  considered appro-
priate.
                                 6-5

-------
     Separate emission  factors  for  foundry  plant sources have been
developed by applying correction  factors  to the emission factors ini-
tially proposed for both  furnace  and  foundry  plants.  These changes do
not affect the revised  emission factors  as  applied to furnace plants.
The computations of correction  factors are  shown in Table A-14 of Appen-
dix A; the final emission factors for furnace and foundry plants are
shown in Table A-15.
     For plants that produce only foundry coke, benzene emission factors
for light-oil recovery  plant sources  (i.e., wash-oil decanters, wash-oil
circulation tanks, light-oil condensers,  and  light-oil  sumps) have  been
adjusted by a correction  factor of 0.54.   This adjustment factor combines
the effects of lower light-oil  yields,  lower  benzene concentrations  in
the light oil, and different coal-to-coke ratios.  Physically, the  re-
duced emission estimates  may be viewed as a result of  lower benzene
throughput in the foundry coke by-product plants.
     For sources treating or handling water that  has contacted the  coke
oven gas (i.e., flushing liquor circulation tank,  excess  ammonia-liquor
storage tank, direct-water final-cooler cooling tower,  tar-bottom  final-
cooler cooling tower, and naphthalene handling/processing), benzene
emissions are expected to be proportional to the  ratio of benzene  in the
coke oven gas  (i.e., partial pressure and partitioning between the  liquid
and gas).  The  light-oil-to-coke-oven-gas ratios  in  Table A-14  are  indi-
cative of the partitioning.  These ratios are multiplied  by  relative
benzene  concentrations in the  light oil  to yield  a correction factor of
0.73  for the above  sources  in  foundry coke plants.
      Emissions  from storing or processing liquids containing  tar (i.e.,
tar decanters,  tar-intercepting  sumps, tar storage tanks, and tar-
dewatering tanks)  also are  expected to be proportional  to the ratio of
benzene  in coke oven gas.   In  addition, the  relative yield of tar (i.e.,
the amount  of  tar  exposed to the benzene)  is  expected to affect  the
partitioning of benzene  between  the tar and  the gas.  Therefore, the
correction  factor  applied reflects both the  relative quantity of benzene
produced,  and  the  proportion of  that benzene transferred to the tar,
ultimately  available for dissolution in the  sources.  Combining the tar
yield,  light-oil-in-gas  ratio, benzene concentration in light oil, and

                                  6-6

-------
 coal-to-coke  ratio  factors  has produced a correction factor of 0.47 for
 the above  source emissions  in foundry coke plants.
      Equipment  leaks  from fugitive emission sources (i.e., pumps, valves,
 exhausters, sampling  connection systems, open-ended valves and lines, and
 flanges  or other connectors in benzene service) are expected to emit ben-
 zene emissions  proportional to the benzene concentration in the fluids
 handled.   The correction factor applied to emission estimates for foundry
 coke plants is  based  on the estimated benzene content of the light oil at
 a  foundry  coke  plant.  When the estimate supplied by ACCCI is used, the
 correction factor is  0.91.  Table A-15 indicates the revised uncontrolled
 emission factors for  furnace and foundry plants.  Table A-16 shows the
 derivation of revised foundry plant benzene fugitive emission rates from
 VOC emission factors.
      Emission estimates incorporating revised foundry plant emission
 factors and other data base revisions are discussed above in response to
 comment 6.1.  The revised emission estimates were incorporated into the
 estimated  impacts on which the selection of the revised proposed standard
 is  based.  The  selection of the revised proposed standard is discussed in
 the Federal Register preamble for the revised proposal.
 6.3   MODEL COKE PLANTS
      Comment:    Commenter IV-D-4 argues that benzene emission estimates
 for  model  coke  plants are not representative of emissions from an actual
 small plant (coke capacity of 440 Mg/day).   The commenter estimates
 uncontrolled emissions from a medium-sized  model plant  (4,000  Mg/day of
 coke) at 1,080  Mg/yr; with the 1984 proposed controls  (and assuming 89
 percent recovery),  remaining uncontrolled  emissions of  120 Mg/yr would
 result.  This  estimate excludes  certain  emission sources  (e.g.,
 direct-water or tar-bottom final-cooler  tower,  tar-dewatering  tanks,  and
 benzene or BTX  storage tanks).   The commenter compares  120 Mg/yr to 70
 Mg/yr (uncontrolled) for the small  plant.   In summary,  the commenter
 argues that small  merchant  plants  should  not  be regulated because  of  the
 low emission  level.

     Response:   In  essence,  the  commenter argues that small  merchant
plants should  not  be regulated because of the low benzene emission  levels

                                6-7

-------
compared to estimated  emissions  from  medium-sized model  plants.   In
support, the commenter suggests  that  his  calculations  show  uncontrolled
emissions at his plant site as  less than  emissions  after control  at  a
larger site.  The EPA  believes  that the commenter has  misconstrued the
purpose of model plants and their  role in the  EPA decisionmaking  process.
First, EPA's decision  to regulate  is  not  based on model  plant  emission
estimates and their level  compared to larger model  facilities.  The  EPA's
decision regarding to  what level to control  foundry plants  is  discussed
further in the preamble to the  revised proposed standard.
     The purpose of constructing model plants  is to portray typical
facilities in terms of size and processes representative of the industry.
However, the nationwide emission estimates have been based  on  site-
specific process data  rather than  by  model plant extrapolation.   In  the
preproposal analysis,  model plant  size parameters were selected based  on
the approximate distribution of actual plant capacities as  a function  of
coke capacity.  This distribution  indicated that 25 of 55 plants  produce
between 300 and 2,000  Mg/day of coke, accounting for 17 percent of
domestic capacity.  Consequently,  a  small model plant  was defined as a
facility producing 1,000 Mg/day of coke,  slightly  less than the midpoint
of the actual range.
     The industry operating data updated  after proposal indicate  foundry
plant capacities at existing sites ranging from 130,000 to 617,000  Mg/yr.
The commenter says his plant's  capacity  is 440 Mg/day.  When converted to
an annual basis (approximately  161,000 Mg/yr), the  commenter's plant
remains well within this size range.   The EPA considers that the  Model
Plant 1  (small) size range remains representative  of small  plant
operations.  In terms of actual onsite processes,  further discussions
with the commenter revealed that the commenter's plant does operate a
direct-water final cooler  (the commenter's statements indicated that no
direct-water final cooler was present at  the plant).  Also present  are a
light-oil plant, fugitive emission sources, and most tar separation
vessels.  Benzene and BTX storage are not practiced.  Again, although the
process  parameters identified with Model  Plant 1 may not reflect  all
actual  operations at the commenter's site, they typify small plant
processes  on the whole.
                                 6-8

-------
      Other changes to the data base also have been made since the 1984
 proposal to reflect more closely foundry plant operations.  These changes
 include lower emission factors and recovery credits.  The EPA believes
 that these adjustments better reflect differences between furnace and
 foundry plants.  The changes would tend to lower the site-specific
 estimate suggested by the commenter.   However, the inclusion of a
 direct-water final cooler also would  need to be considered by the
 commenter when recalculating actual  benzene emission estimates for this
 site.

 6.4  EMISSION FACTORS FOR TAR-RELATED SOURCES
      Comment:   Commenters IV-D-7,  IV-D-10,  and IV-D-12  question the data
 base used to estimate emission factors and  their resulting industry wide
 applicability  for tar decanters, tar  dewatering,  and flushing liquor
 circulation  tanks.

      Response:   The  commenters argue,  in  essence,  that  test  data  for
 certain sources  are  not  sufficient to  take  into  account variance  in
 emissions  because of  differences in the methods  of operation  and  other
 factors.   The  EPA certainly  agrees that differences  in  methods  of
 operation, operating  parameters, and design  features are evident  from
 plant to  plant and will  influence actual  emissions from each  source.
 During  development of the estimated emission factors, these variations
 have  been taken  into  account to the extent possible  by averaging appli-
 cable measurements to obtain a factor  representative of a  "typical"
 source.  Also, the emission factors have  been adjusted since the 1984
 proposal for improved applicability to plants producing foundry coke.
     Specifically, the commenters state that data (12 tests) supporting
 the tar-decanter emission factor are not sufficient for industry wide
 application because emissions are sensitive to variability in gas-liquid
 separator residence time and optional  heating.  As discussed in Chapter 3
 of the BID for the 1984 proposed standards (page 3-10),  typical residence
 times are about 10 minutes for liquor  and  about 40 hours for tar.   Op-
 tional heating tends  to increase the total benzene emitted even though
 the concentration of  benzene per unit  volume of emissions  may be reduced.
The degree of separation  achieved is highly  variable  because of coal  type
and differences between  plants. As stated above,  adjustments  have been
                                6-9

-------
made since the 1984 proposal  to account  for  differences  between  furnace
and foundry plants.
     The tar-decanter emission  factor (applicable  to  furnace  plants)  is
based on three measurements  for each of  four vents (12 total)  at two  tar
decanters at two plants (Bethlehem Steel—Burns  Harbor,  Indiana, and
Bethlehem Steel—Bethlehem,  Pennsylvania).   The  tests spanned  a  flow  rate
range of 50 to 275 std ft^/min.  The benzene emission rate  measured at
the Pennsylvania steel plant was 1.2 kg/hr  (Docket Item  II-A-22).  This
decanter was one of two for  a coke battery.   Emissions from the  two
decanters were assumed to be twice the emissions from the single
decanter, or 2.4 kg/hr.  The corresponding  benzene emission factor for
this decanter was calculated as 84.7 g/Mg coke.   One  of  three tar
decanters was tested at the  steel plant  in  Indiana (Docket  Item  II-A-25),
where the average benzene emission rate  from three vents on the  decanter
was 4.4 kg/hr.  The corresponding emissions for  three decanters  at this
Indiana plant are 13.3 kg/hr, which yields  a benzene  emission factor  of
69.6 g/Mg of coke.  The average benzene  emission factor  from  these two
plants was 77.2 g/Mg of coke.  Consequently, the emission factor was
designated as 77 g of benzene/Mg of coke.  The EPA considers  this data
base of 12 measurements adequate to estimate the average level of
emissions from typical decanter vessels  under varying conditions.
     The commenters also maintain that the tar-dewatering emission factor
should not be applied industry wide because emissions depend  on  the
method of operation.  In support, the commenters point  to the "unex-
plained" variations in the range of emission factors  for this source
(9.5 to 41 g/Mg of coke).
     Emissions from tar-dewatering tanks were evaluated  at three plants
(see Docket  Items  II-A-26, II-A-27, and II-A-28).  Three measurements
were made for each of two vents  at one plant; one measurement was made at
the  second plant.  At the third  plant, one test was made at the tar
storage tank where dewatering was performed.  The EPA considers that
these measurements, as averaged, are sufficient to provide a   reasonable
estimate of  emissions from a typical source.
     The extent and effect of  the variation in  dewatering emission
factors have  been  discussed  in  the BID for the  1984  proposed  standards
 (page 3-16),  which states as follows:
                                 6-10

-------
      The emissions  data  for  tar  dewatering at the first plant showed
      higher emissions  from the west  tank  (3.2 kg/hr) than from the east
      tank (1.1  kg/hr).   These tanks  are operated in series rather than in
      parallel,  and  the wet tar enters the west tar dehydrator first.
      Consequently,  the emissions  from the west tar dehydrator are
      expected to  be higher than emissions from the east tar dehydrator.
      The daily  benzene emission  rates from the two tar-dewatering tanks
      at  this first  plant were 27  and 76 kg, respectively.  Daily benzene
      emissions  from tar  dewatering at the second plant were 43 kg.  The
      tar is dewatered in storage  at the third plant, where benzene
      emissions  were 24 kg/day.  The benzene emission factors from these
      three  plants were 41, 9.5, and 12.9 g/Mg of coke, respectively.
      These  were averaged to  obtain a benzene emission factor for tar
      dewatering of  21 g/Mg of coke.
      The tar-dewatering tanks contained tar with 200 to 2,000 ppm benzene
      in  the liquid.  Tar, as collected from the flushing liquor and the
      primary cooler, can contain  greater than 0.2 percent benzene or
      2,000  ppm  at a  rate of 40 kg/Mg of coke produced.  The maximum
      potential   for  benzene loss from tar dewatering and storage
      calculated from these values is greater than 2,000 ppm at a  rate of
      40  kg/Mg of coke produced.  The maximum potential  for benzene loss
      from tar dewatering and storage calculated from these values is
      greater than 80 g/Mg of coke.  The benzene emissions  from tar
      dewatering and storage probably will  be less than 80  g/Mg of coke
      and will depend on the method of operating these  processes.
      The commenters also question the adequacy of test  data  from  the
primary cooler  condensate tank  as the basis  for the  flushing  liquor
circulation tank emission factor  and its resulting  applicability  industry-
wide.  The emission factor for  the flushing  liquor  circulation tank
(9 g/Mg of coke) was obtained from one  test  in which emissions from  a
primary cooler  condensate tank  were measured  (Docket Item  II-A-13).   This
tank was assumed to be  similar  to a flushing  liquor  circulation tank
because both vessels function to  hold liquor taken  from the gas stream
during early stages of  gas  processing.   Although  it  is  desirable  to  have
more than one test measurement  as the basis  of the estimated  factor,
engineering judgment suggests that the measurement is  a  reasonable value
                                 6-11

-------
for emissions from flushing liquor  circulation  tanks.   The  EPA  agrees
with the commenter,  however,  that emissions  will  vary  necessarily  de-
pending on the number and geometry  of  tanks,  the  number of  vents,  and
other factors.

6.5  METHODOLOGY FOR EMISSION FACTORS
     Comment:  Commenter IV-D-33 contends  that  the  EPA VOC  and  benzene
emission factors are not applicable to the sources  at  his site  and also
that VOC emissions should be calculated using a different methodology.

     Response:  The EPA developed emission factors  to  obtain  an estimate
of the nationwide emissions of benzene and VOC  from by-product  plant
process operations.   The EPA is aware  that site-specific factors could
cause the actual emissions from a particular facility  to vary from the
estimates based on EPA emission factors.  In fact,  the commenter's
estimate of benzene and VOC emissions  from his  facility using his
alternate set of emission factors is within 20  to 30 percent  of the
emissions estimated when using EPA  emission factors.  This  difference  is
within the range of uncertainty for the emission  factors.
     The commenter also proposed an alternate methodology for developing
VOC emission factors using EPA test data that would, in general, would
tend to make them lower.  The commenter states  that EPA overestimates  VOC
emissions by assuming that all components  of light oil will volatilize  to
the same extent as benzene.  The EPA agrees that  this  may lead to an
overestimate of VOC emissions; however, the EPA believes the commenter's
approach underestimates VOC because not all  components of the light-oil
vapor were explicitly measured in the  test program whose results were  used
to develop the VOC emission factors.  In any case, the cumulative
difference between emissions estimates for the commenter's  plant using the
commenter's methodology were shown  to  be within the range of uncertainty
for the emission factors, as noted above.

6.6  VOC BENEFITS FOR OZONE REDUCTION
     Comment:  Commenter IV-D-2 supports the proposed standard for
by-product plants, particularly when applied to a plant site located in
his State.   This commenter believes that the estimated emission
                                 6-12

-------
reductions are realistic and provide the added benefit  of helping  his
State reduce the VOC inventory in the Baltimore ozone nonattainment
area.

     Response:  The EPA thanks the commenter for his  support.   The
estimated VOC emission reductions also will  benefit the  country as a
whole in reducing ozone formation.
                               6-13

-------
                          7.  COST IMPACT

7.1  REVISIONS TO COST ANALYSIS
     Comment:  Commenters IV-D-9 and IV-D-14 argue that the capital  costs
of the proposed equipment are $50 million to $100 million or more,
compared to the estimated cost of $23.8 million.  According to these
commenters, the true costs exceed model plant estimates by 50 to 100
percent at some facilities.  In support, the commenters cite the
following major factors contributing to 1984 EPA estimates:  (1) low
estimates of unit material costs and construction expenses, (2)  site-
specific factors such as equipment conditions and pipeline length,  (3)
EPA's reliance on cost-estimating references rather than experience  and
price quotations from local suppliers and contractors,  (4) the dollar
year of the estimates (1982), and (5) additional costs  for work  in
hazardous areas requiring special safety precautions.   One commenter
provides for EPA review an example of these points using estimates
prepared by National Steel, Armco, and by United Engineers for a
Bethlehem Steel plant.  Another commenter (IV-D-33 and  IV-D-34)
submitted information on cost estimates for controls at his plant; the
commenter contends that the capital  costs would be higher than the model
unit costs in the 1984 proposal  BID.
            «•
     Response:   To consider the commenters'  concerns, EPA conducted  a
detailed review of the United Engineers'  estimate for the Bethlehem  plant
of Bethlehem Steel,  Bethlehem Steel's estimate for their Sparrows Point
plant,  and the  Armco and National  Steel  cost data;  EPA  also had  a cost
estimate prepared by C.  R. S. Sirrine,  Inc., a third-party design
engineering firm.   The details of EPA's final  analysis  are shown in
Appendix B.  Included in the review was a site visit to the Bethlehem
plant to resolve questions regarding  equipment locations,  and  the sources
subject  to the  proposed  emission controls,  and to obtain examples of
                                 7-1

-------
site-specific conditions  pertinent  to  the  development  of  revised unit
cost factors.
     As shown in Appendix B,  the  revised cost  analysis  includes higher
unit costs for most materials,  which affects the  costs  estimates for most
sources.  The revised unit costs  were  composed from the data  received in
the comments and the cost data  developed by Sirrine.   The  revised analy-
sis also includes costs  for sealing all sources,  installation of roofs on
certain storage tanks, more pipe  supports, pressure/vacuum relief valves
for sealed sources, and  adjustments to unit cost  factors  for  work in
hazardous areas requiring special  safety precautions.
     The commenters1 criticism  of EPA  for  reliance on  cost-estimating
references is valid.  The EPA agrees that  it is desirable  to  base cost
estimates on previous experience  and site-specific factors.   However, in
the absence of abundant  experience, engineering and construction firms
use those same references to develop cost  estimates.   Backup  information
requested by EPA to support the commenters1 cost  estimates indicated that
such was the case.  Also, the preference for site-specific information
must be compromised somewhat when attempting to develop within  schedule
and budgetary restraints  nationwide cost impacts  for 44 plants.  The EPA
acknowledges that costs  for particular plants  may be higher or  lower than
EPA estimates, depending on site-specific  conditions.   However, the
revised cost analysis addresses concerns cited by the  commenters and the
costs are reasonable estimates  of the  industry wide cost  of controls.
     Using the revised cost analysis,  the  nationwide capital  cost of the
revised proposed standards is approximately $84 million (1984 dollars),
compared to estimated capital costs of about $24  million  (1982  dollars)
af proposal in 1984.  Of the $84 million,  approximately half  is due to
the inclusion of wash-oil final coolers, the cost of which was  not
reflected in the 1984 proposal  estimate.
7.2  REVISIONS TO PRODUCT RECOVERY CREDITS
     Comment:  Commenters IV-D-6 and  IV-D-14  state that the value of
potential product recovery credits has been  overestimated.  In  parti-
cular,  commenter IV-D-6 states  that the value  of  light oil for  small
plants  is overstated.  One commenter  explains  that the assumption that
the recovered product can be used as  plant fuel or sold is not  valid

                                 7-2

-------
because, when the production of coke oven gas is greater than demand for
potential fuel consumption, the excess gas is flared.  One commenter
states that for one plant (Lackawanna) no product recovery credit can be
assumed and for another plant (Bethlehem) the credit should be reduced.
The excess gas is flared at Bethlehem, and Lackawanna now has no
steelmaking operation creating fuel demand.

     Response:  The EPA essentially agrees with the commenters that the
value of potential product recovery credits was overestimated in the 1984
proposal.  As discussed further in response to comment 6.2, foundry
plants produce less light oil than larger furnace plants do.   This
difference in production quantity (reflected in new emission  factors for
foundry plants) has been taken into account in the computation of revised
fuel value and light-oil recovery credits.
     In response to the commenters1 concerns, a telephone survey of seven
(three furnace and four foundry) plants was conducted to determine the
extent of flaring excess coke-oven gases (IV-E-9).  Briefly,  EPA found
that this flaring is not generally practiced except as a last resort.  Of
the five foundry plants surveyed, only one (Empire Coke) flares gas
continuously.  Of the remaining four foundry plants, two do not flare
excess gas at all, and two plants flare only seldomly.  Of the three
furnace plants surveyed, one plant never flares, one plant flares only in
emergency situations, and one plant flares occasionally in periods of low
demand.  Consequently, the adjusted credits have been applied to most,
but not all,  plants.  No fuel credits were applied to the Lackawanna
plant of Bethlehem Steel and Empire Coke for the reasons cited by the
commenters.
     The value of potential  product recovery credits also has been
adjusted since that time to reflect 1983 data published by the DOE
(Docket Item  IV-J-4).   Based on  these data (Table A-ll), the  credit for
light oil  has been decreased from $0.33/kg to $0.27/kg light  oil.   The
fuel value recovery credit for coke oven gas also has been adjusted
downwards--$0.14/kg coke oven gas compared to $0.15/kg estimated in the
1984 proposal.
                                 7-3

-------
7.3  ECONOMIES OF SCALE  FOR  SMALL  PLANTS
     Comment:   Commenters  IV-D-4,  IV-D-6,  IV-D-7,  and  IV-D-17 maintain
that small  plants should not be  regulated  because  of the disproportionate
cost impact resulting from the lack  of economies of scale  compared to
moderate or large plants,  coupled  with higher per-unit control  costs.
One commenter notes that,  although control  costs for small  plants are the
same as for medium-sized plants, the costs in relation to  production are
200 to 400 percent higher; another commenter indicates that small plant
costs are 900 percent higher than  for medium-sized plant costs.  The
commenters point to the use  of a cost model  based  on a moderate to large
plant with a number of economies of scale  in terms of  the  number of
control units per ton of production.  According to one commenter, this  is
reflected in Section 8.1.5 of the  BID for  the 1984 proposed standards,
where actual costs are compared to estimated costs for two large plants
with economies of scale.
     Response:  As described in the response to comment 7.1 and Appendix
B,  the capital and annualized cost estimates for control  of benzene
emissions have been  revised.  The  basis for estimating these revised
costs are the three  original model plants, sized at 1,000, 4,000,  and
9,000 Mg/day  of  coke.   New  capital and annualized costs of control  were
estimated for these  plants, then cost functions (equations relating  cost
to  plant production  capacity) were developed for each  process.   These
cost functions do  provide for economies of scale,  and  they adequately
represent the costs  of  control  from the smallest foundry coke by-product
plant to the  largest furnace  coke by-product plant.
                                  7-4

-------
                        8.  ECONOMIC IMPACT
8.1  REGULATORY BASELINE
     Comment:  Commenters IV-D-6 and IV-D-7 state that the economic
analysis for the proposed standards fails to consider the true state of
the coking industry at baseline and that the economic impact will  have an
adverse effect on the industry.  In support, one commenter notes that the
analysis does not take into account the plant closures and capacity
reductions that have occurred since 1980.  Both commenters also note that
the baseline does not include the cost of other environmental  regulations
incurred by 1983.  New regulations include final iron and steel effluent
guidelines, National Pollutant Discharge Elimination System (NPDES) per-
mit upgrading, State implementation plan (SIP) compliance rules (includ-
ing reasonably available control technology [RACT], lowest achievable
emission rate [LAER], and new source review of coke plant rebuilding),
and the pending coke-oven battery NESHAP.

     Response:  At the time the original analysis was conducted, the
information from published and unpublished sources was current.  A
reanalysis has been conducted (see Appendix C) that utilizes data  on
plants and capacity in existence in November 1984.  Financial  data and
production data used in baseline estimates are from the available
published and unpublished sources as of 1984.   A discussion of industry
trends as of 1984 is provided in Section C.I.6 of Appendix C.
     The baseline of the reanalysis assumes companies meet regulations
existing in 1984, including  OSHA rules  for coke oven emissions;  State
regulations related to desulfurization, pushing, coal  handling,  coke
handling,  quench tower,  and  battery stack controls;  and Best Practicable
Technology (BPT)  and Best Available Technology (BAT)  water regulations.
All of these were due to be  in effect  by 1983  at the latest.   Other
regulations that are pending or have not reached the deadline  date for
                                 8-1

-------
compliance are not  likely  to  be  a  part  of  1984  production costs for
firms, or they will  have little  effect  on  those costs.

8.2  SELECTION OF DOLLAR YEAR
     Comment:   Commenter IV-D-7  suggests that the  economic  impact
analysis should be in 1986 dollars because the  project  schedule places
promulgation closer to regulation  in 1986.
     Response:  Selection  of  the year for  dollar values in  analysis  is
somewhat irrelevant because conversions may be  made  for any current  or
past year based on gross national  product  (GNP) implicit price deflators.
The important values are the  baseline data from which  regulatory  impacts
are determined.  For these values, current information  (in  1984 dollars)
was used to produce realistic results in the reanalysis.
     Projection to future year-dollars is  difficult  primarily because of
the confounding effects of inflation.  Prediction  of inflation  rates is
beyond the scope of this analysis.  The 1984 dollar  values  in the
reanalysis are best updated for future timeframes  when those years are
current  so that GNP implicit price deflators accounting for actual
inflation may be used.

8.3   POTENTIAL ECONOMIC IMPACT
      Comment:  Commenter  IV-D-14  states that the economic impacts of the
proposed standards are  more  severe than estimated and will  have an
adverse  economic impact on the  industry.   In support, the commenter cites
examples from a  recent  Price Waterhouse "Steel  Segment" survey for the
period  1979 through  the third quarter  of  1983  to illustrate the overall
financial  condition  of  steel companies.   The following major factors are
cited:   (1) the  steel industry  is depressed and suffers capital formation
problems;  (2) the  period  analyzed shows a rising debt-to-equity ratio,
with  declining  stockholder equity;  (3) investment exceeded cash from
operations;  and  (4)  the industry  experienced $6 billion in losses between
 1982  and 1983.
      Response:   The measure  of  severity of impacts  is  best made  relative
 to some reference value rather  than from  the standpoint of  absolute
 values.  In the reanalysis,  capital  costs of compliance are  compared to
                                  8-2

-------
 average annual net investment averaged over the period from 1979 to 1983
 (converted to 1984 dollars) for individual companies.  Table C-25 in
 Appendix C of this document shows these comparisons for furnace coke.
 For furnace coke plants, capital costs of compliance for Regulatory
 Alternative II range from 0 to 3 percent of net investments.  For
 Regulatory Alternative III, these costs range from 0 to 5 percent of net
 investments.  The regulatory alternatives are outlined in Table C-l of
 Appendix C.
     The industry trends noted by the commenter are discussed in Section
 C.I.6 of Appendix C.  Companies have made adjustments through mergers,
 acquisitions, and creative financing measures to generate investment
 funds.  The fact that, as the commenter states, investment exceeded cash
 from operations indicates that capital is available for investment even
 for firms sustaining losses.
     Although the industry is having some capital  difficulty, the burden
 of regulation will  differ from firm to firm.   The  net investment analysis
 indicates that in no case will the cost of regulations be a significant
 burden.

 8.4  ESTIMATED EMPLOYMENT IMPACT
     Comment:   Commenter IV-D-7 questions the estimated employment
 impacts of the proposed standards.  The commenter  suggests that the
 estimates should include total plant employment because by-product plants
 cannot be separated.   This commenter employs  36 people in his by-product
 operation, but he employs a total  of 268 persons in his coke plant.

     Response:  The commenter's argument is  answered in Tables  C-23  and
 C-28 in Appendix C, which show the employment effects of the regulatory
 alternatives in the furnace and foundry coke  plants, respectively.   These
 are industry totals.   For the furnace coke sector,  neither regulatory
 alternative results in a loss of more than 0.5 percent of baseline  jobs
 at furnace coke plants for the entire industry.  This is not a  sub-
 stantial  loss, and  it  should be weighed against  the benefits.
     For the foundry  coke sector,  employment  impacts are calculated  for
two scenarios.  Scenario A assumes that foundry  coke producers  do  not
compete with imports  in the domestic market,  and Scenario B  assumes  they
do.  Under Scenario A,  the regulatory alternatives  result in job losses
                                 8-3

-------
that are less than 1.0  percent  of  baseline  foundry  coke employment.
Under Scenario B,  employment  losses  for  the industry  are less than
3.2 percent of baseline.   Again, these losses  are not  large.
     It is possible that  unemployment will  not  occur  as a  result of the
regulations for two reasons.  First, workers may be reallocated within
the industry to perform other tasks  because of  labor  contracts or other
constraints.  Second, there are potential employment  gains  from the
regulations such as labor to  operate and maintain control  equipment.
This labor is included  in the cost analysis, but it is not  evaluated in
terms of added jobs. These gains  may offset estimated job  losses.
8.5  IMPORT TRENDS
    Comment:  Commenter IV-D-7  states that  the  regulation  will increase
the trend of importing  coke.  The  commenter cites Table 9-1 of the
1984 proposal BID, which  shows  a growing coke-importing trend since
1974, and Table 9-2, which shows a decrease in  domestic production.
     Response:  Data up through 1983 indicate  that  imports  have been
decreasing since 1979  (see Table C-2 in  Appendix C).   Trends in the steel
industry away from coke-using processes  and toward  decreased steel
production overall are  the most likely sources  of this decrease.
     The reanalysis indicates that the regulatory alternatives result  in
a slight reversal  of this trend.   Table  C-22 in Appendix C  shows that
furnace coke imports will increase by 9,000 Mg/yr under Regulatory
Alternative II and by 25,000  Mg/yr under Regulatory Alternative III.
These represent increases from  the baseline of 0.23 percent and 0.64
percent, respectively.   These are  negligible changes  in imports.
8.6  ECONOMIC IMPACT ON SMALL PLANTS
     Comment:  Commenter IV-D-6 maintains that  the  economic impact
assumptions for integrated,  captive producers  compared to  small merchant
foundry plants are dissimiliar; these differences should result in
separate regulations.   The commenter states that  "foundry  producers,
unlike captive producers, cannot distribute costs among operations,
cannot adjust the price of coke oven gas or light-oil  used elsewhere  in
the facility, and cannot increase  the price of other  by-products."  The
                                 8-4

-------
 additional costs to foundry plants result in a direct  increase in  product
 price, which may give advantages to foreign competitors.
      Commenter IV-D-7 argues that, for the same reasons,  small  plants
 will incur a disproportionate economic impact.  This commenter also  cites
 Table 9-40 of the economic analysis,  which estimates a coke  price
 increase ranging from 6.4  to 15.4 percent for small plants to  comply with
 baseline.

      Response:   A distinction is made between furnace  and foundry  plants
 in the BID analysis and  in the reanalysis for the  revised proposal.  Most
 furnace coke producers are captive, and most  foundry producers  are
 merchant.   This  distinction allows the analysis  to examine impacts
 separately.
      The differences  between furnace  and  foundry producers expressed by
 the commenter do not  necessarily result in  a  worsened  competitive
 situation  for foundry firms with respect  to other firms in the  foundry
 industry.   In the reanalysis,  no foundry  batteries become uneconomic
 (candidates  for  closure) under either  regulatory alternative.   This
 implies  that  industry impacts  of regulation will not be concentrated on
 any one  plant sufficiently  to  force it out of  business.
      Tables C-24  and  C-29  in Appendix  C show  the capital costs  of com-
 pliance  of the regulatory  alternatives for furnace and foundry  producers.
 For both regulatory alternatives,  the  foundry  coke producers' share of
 total  capital costs of compliance  is less than 16 percent.  For indivi-
 dual  foundry  coke-producing  firms, Table C-30  in Appendix C  shows that
 the capital costs of  the regulatory alternatives amount to no more than
 11  percent of net investment in  for firms for which data were available.
 This  is not substantially higher than the maximum share of capital  costs
 of  net investment for furnace coke producers (see Table C-25  of Appendix
 C).   Furnace coke producers face additional pressures  because of the
 difficulties being experienced in the steel industry, which  composes  the
market for furnace coke.
     The influence of imports in the foundry coke industry is accounted
for under Scenario B of the reanalysis.  A worst case bound  is  assumed  so
that quantity reductions  in domestic production are assumed to  be offset
by quantity increases in  imported coke sold domestically.  The  changes
are 61,000  Mg/yr  under Regulatory Alternative  II,  and 94,000  Mg/yr  under
                                 8-5

-------
Regulatory Alternative III.   These represent  2.1  percent  and  3.2  percent
of foundry coke demand,  respectively.   Advantages gained  by foreign
competitors because of the regulatory  alternatives are  small.
     In the reanalysis,  price impacts  under Scenario  A  for  foundry  coke
producers are $0.99/Mg for Regulatory  Alternative II  and  $1.46/Mg for
Regulatory Alternative III.   These represent  0.58 percent and
0.86 percent increases from baseline (see Table C-27  in Appendix  C).
Under Scenario B, no price impacts will  result.  No significant  impacts
are projected for foundry coke producers because of these price  changes.

8.7  PRICE IMPACTS
     Comment:  Commenter IV-D-11 states that  the economic analysis  is
inaccurate in predicting the increased price  of coke for  merchant plants.
This cornmenter estimates an increase in the price of coke at  his  plant of
$1.38/Mg versus $0.24/Mg estimated at  proposal  in 1984  (49 FR 23525).
This estimate is based on the commenter's estimate of the cost of
compliance at his facility (capital costs of $1.8 million versus  average
cost of $408,500 cited in the 1984 proposal BID; annualized costs of
$80,000 versus $70,500 cited in the 1984 BID).   The commenter notes also
that his capacity is 681 Mg/day rather than 1,362 Mg/day.
     Response:  The determination of changes in the price of coke must be
made on a market basis rather than a pi ant-by-plant basis.  The price
changes are  due to shifts in the entire supply curve, as  well as the
effects of the marginal plant at equilibrium for the entire market.  The
economic impact model uses this basis for its computation.
     The reanalysis calculates capital costs of compliance, annualized
compliance costs, and price changes based on capacity information avail-
able in November 1984.  Capital costs of compliance for furnace and
foundry plants are given  in Tables C-24 and C-29 of Appendix C,  respect-
ively.  Annualized compliance costs are shown  in Table C-31 for  furnace
coke producers and Table  C-32 for  foundry coke producers.  Tables C-21
and C-27  show  price effects  of the  regulatory  alternatives on furnace and
foundry  coke producers.   These costs differ  from  engineering estimates
because  of the calculation  of costs based on batteries with
                                  8-6

-------
 marginal  cost of production  at  or  below  price,  rather than  all
 batteries.
      For  furnace coke,  the average capital  cost  per  plant is
 approximately $1.0  million for  Regulatory Alternative II and $1.7 million
 for Regulatory Alternative III.  The  average  annualized cost per plant  is
 $87,500/yr  for Regulatory Alternative II and  $310,000/yr for Regulatory
 Alternative III.  Price increases  are $0.13/Mg  (a 0.12-percent increase)
 for Regulatory Alternative II and  $0.36/Mg  (a 0.33-percent  increase) for
 Regulatory  Alternative  III.
      For  foundry  coke,  the average capital  cost  per  plant is
 approximately $636,000  for Regulatory Alternative II and $1.1 million for
 Regulatory  Alternative  III.  Average  annualized  cost per plant is
 $118,000/yr and  $264,000/yr  for Regulatory  Alternative II and Regulatory
 Alternative III,  respectively.  The price increase associated with
 Regulatory  Alternative  II for foundry coke  is $0.99/Mg (a 0.58-percent
 increase  from baseline), and, for  Regulatory Alternative III, the price
 increases by  $1.46/Mg (a 0.86-percent  increase from baseline).
      The  average  values may  not reflect actual costs for individual
 plants. They  serve  as indicators of the neighborhood of costs a plant
 may  be expected to  face in complying  with the regulatory alternatives.
 8.8   ECONOMIC  IMPACTS ON FOUNDRY PLANTS
      Comment:  Commenters IV-D-10  and  IV-D-14 state that merchant plants
 should not  be  regulated because of the adverse economic  impact  on the 13
 plants comprising this  industry segment.   The commenters disagree that
 no merchant plant will close as a  result  of the 1984 proposed standards.
 One  commenter predicts the closure of three entire merchant  plants
 because of the estimated costs of compliance.   An added  impact  of these
 closures is the metal casting industry's  dependence on  1.4 million tons
 of foundry coke production.

     Response:  The  estimated annual compliance costs for  foundry plants
 computed in the reanalysis  are presented  in  Table C-32 of Appendix C,
 and the estimated capital  compliance costs  are shown  in  Table C-29.
Average annual plant compliance  costs  are $118,000/yr for Regulatory
Alternative II and $264,000/yr for  Regulatory  Alternative  III.  Average

                                 8-7

-------
capital costs of compliance for foundry coke plants are $636,000 for
Regulatory Alternative II and $1.1 million for Regulatory Alternative
III.
     In terms of net investments for companies, capital costs of com-
pliance are relatively small.  For firms for which data are available,
capital costs amount to no more than 11 percent of net investment for
either regulatory alternative (see Table C-30 of Appendix C).  This does
not imply an excessive capital  burden because of the regulatory
alternatives.
     In the reanalysis, two scenarios for the foundry coke industry are
evaluated.  Under Scenario A, foundry coke producers are assumed to
supply all of the domestic coke market so that supply shifts induced by
the regulatory alternatives result in slightly higher prices and
slightly reduced production (see Table C-27).  In all  cases, changes in
price and quantity produced are less than 1.0 percent of baseline
values.
     Scenario B assumes that foundry coke producers must compete with
foreign producers in the domestic coke markets.  As a worst case, foreign
coke is assumed to be available at a price equal  to baseline, and that
price is assumed to remain constant regardless of changes in the domestic
market.  Furthermore, imported  coke is assumed to be a perfect substitute
for domestic coke so that, for  any reduction in domestic production,
consumers will purchase amounts of imported coke equal  to the reduction.
Under this scenario, there is no price change because of the regulatory
alternatives.  The quantity changes shown in Table C-27 indicate that
domestic production will  decrease by 61,000 Mg/yr under Regulatory
Alternative II and by 94,000 Mg/yr under Regulatory Alternative III.
Import increases by these amounts reflect 2.1 percent and 3.2 percent of
domestic demand, respectively.
     Under either scenario, the metal  casting industry is unlikely to
suffer.  Under Scenario A, if price and quantity changes do occur, they
will not be substantial.   Under Scenario B, domestic coke reductions will
be offset by increased availability of imported coke.
     Even if the entire shortfall  in domestic production is compensated
by increased imports, domestic  foundry coke producers are unlikely to be
                                 8-8

-------
significantly impacted.   No closures  from  the  regulatory alternatives are
predicted under either scenario.   Other  impacts  such as employment are
unlikely to be substantial, as  shown  in  Table  C-28 of Appendix C.
                               8-9

-------
                    9.  QUANTITATIVE RISK ASSESSMENT
 9.1  USE OF MODEL FOR HEALTH RISK ESTIMATES
      Comment:  One commenter (IV-D-14)  states  that  EPA's  prediction  of
 the leukemia risk to the community is  overstated  because  of  the  linear,
 nonthreshold extrapolation model.   Other commenters suggest  that,  by
 mathematically predicting benzene exposures in the  vicinity  of the coke
 by-product recovery facilities  and consequential  risks, EPA  may  be
 estimating values that  really do  not exist  (IV-D-6, IV-D-10,  IV-D-12).
 These commenters  suggest that EPA (1) monitor  benzene near these
 facilities to verify the model  and (2)  conduct  epidemiologic  studies of
 the communities surrounding  the facilities.

      Response:  Because  a  specific environmental  carcinogen is likely to
 be  responsible for  at most a small  fraction of  a  community's  overall
 cancer  incidence  and because the  general population is exposed to  a
 complex mixture of  potentially  toxic agents, it is  currently  not possible
 to  directly  link  actual  human cancers with ambient  air exposure to
 chemicals  such as benzene.  Today's epidemiologic techniques  are not
 sensitive  enough  to  measure the association.  Direct measurement of
 health  effects or estimation of a  causal relationship to chemical
 exposure through  community health  studies usually is not possible due to
 the limited statistical  sensitivity of such studies  and  the presence  of  a
 large number of potentially confounding variables (e.g., general  health
 status, occupational exposure,  smoking,  diet,  migration, age, etc.).
 Therefore, EPA must rely largely upon  mathematical modeling techniques to
 estimate human health risks.   These techniques, collectively  termed
 "quantitative risk assessment,"  are the  means  whereby the  risk of adverse
health effects from exposure  to  benzene  in  the  ambient environment  can be
estimated mathematically; effects  found  at  higher  occupational exposure
levels are extrapolated  to lower concentrations characteristic of  human
                                9-1

-------
exposure in the vicinity  of  industrial sources of benzene.  The analysis
estimates the risk  of cancer at  various  levels of exposure.  A unit risk
estimate for benzene  is derived  from the dose-response relationship
observed in the occupational  studies.  The unit  risk estimate represents
the cancer risk for an individual  exposed to  a unit concentration of a
carcinogen [e.g.,  1 part  per million (ppm)] for  a lifetime.
     Although EPA  agrees  that the  linear, nonthreshold model is
conservative in nature and would tend  to provide a  reasonable upper bound
to the statistical  range, the Agency does not agree that the assumptions
upon which it is based are unreasonable  or that  the results of its use
are exaggerated.  The dose-response mathematical model with low dose
linearity is used  by EPA  because it has  the best, albeit limited,
scientific basis of any of the various extrapolation models currently
available.  The EPA has described  the  scientific suppositions underlying
the preference of the linear, nonthreshold model over  other mathematical
models  (Water Criteria Documents Availability, 45 FR 79319, November 28,
1980).   In this notice EPA stated:
      There is really no  scientific basis  for any mathematical
      extrapolation model which relates carcinogen  exposure to cancer
      risks at the extremely low levels of  concentration that must be
      dealt with in evaluating the environmental hazards.   For practical
      reasons, such low levels of risk cannot be measured  directly either
      using animal experiments or epidemiologic  studies.   We must,
      therefore, depend on our current understanding of  the mechanisms  of
      carcinogens  for guidance as to which risk  model  to use.  At the
      present time, the dominant view of the carcinogenic  process
      involves the concept that most agents which  cause  cancer  also  cause
      irreversible damage to DNA. This position  is  reflected  by  the  fact
      that  a  very  large  proportion of agents which  cause cancer  are  also
      mutagenic.   There  is  reason to expect that the quanta!  type of
      biological response that  is characteristic of mutagenesis  is
      associated with a  linear  non-threshold dose-response relationship.
      Indeed,  there  is substantial evidence  from mutagenesis  studies with
      both ionizing  radiation and with  a wide variety of chemicals  that
      this type of dose-response  relationship is also consistent with the

                                  9-2

-------
       relatively few epidemiological studies of cancer responses to
       specific agents that contain enough information to make the
       evaluation possible (e.g., radiation-induced leukemia,  breast and
       thyroid cancer, skin cancer induced by aflatoxin in the diet).
       There is also some evidence from animal  experiments that is
       consistent with the linear non-threshold hypothesis (e.g.,  liver
       tumors induced in mice by 2-acetylaminof1uorene in the  large  scale
       ED01 study at the National  Center for  Toxicological  Research,  and
       initiation stage of the the two-stage  carcinogenesis model  in  the
       rat liver and mouse skin)  (45  FR  79359).
       With regard to the need for epidemiologic  study of the  population
 residing in the vicinity of  the  coke-oven  by-product  recovery  plants, it
 must be  kept in mind that current methods  are  not  sufficiently  sensitive
 to detect a causal  association  between  chronic,  low-level benzene expo-
 sure and cancer.   Such  studies  are complicated by  a number of  potentially
 confounding factors.   These  factors  include  genetic diversity,  population
 changes  and mobility,  lack of consolidated medical records, lack of
 historical  benzene  exposure  data  over each individual's  lifetime,
 community exposure  to other  carcinogens besides benzene, and the latency
 period of cancer.
      In  the evaluation of benzene emissions from coke oven by-product
 recovery  plants under Section 112 of the CAA, EPA has followed a policy
 in which  the nature and  relative magnitude of health hazards are a
 primary  consideration.   In the absence of scientific certainty,
 regulatory  decisions must be made on the basis of the best information
 available.  In the case of benzene, EPA has evaluated the potential
 adverse effects associated with human exposure based on the best
 scientific  information currently available.  For benzene, this is
 represented by the occupational  epidemiologic studies  demonstrating  a
 causal association between exposure and  leukemia.
 9.2  SELECTION OF RISK MODEL
     Comment:  One commenter  (IV-D-13)  suggests  that,  in  using  HEM,  and
not the Industrial  Source Complex  (ISC)  model,  to estimate  annual  average
ground level concentrations of benzene around  coke-oven by-product
recovery  plants, EPA has  underestimated  exposure to the population living
                                9-3

-------
near those facilities.   The  commenter  alleges that EPA has admitted the
model underestimated exposure  from  200 to  300 percent in the benzene
fugitive emission rulemaking.   Therefore,  the commenter states that risk
to the most exposed individuals should be  much  higher.  On the other
hand, commenter IV-D-14 states that EPA's  assumption  in the item that
individuals are exposed to the maximum annual ground-level concentration
of benzene for 24 hours/day, 365 days/year for  70 years are unrealistic
assumptions that lead to exaggerated risk  calculations.
     Response:  Commenter IV-D-13 (NRDC)  raised these same concerns in a
petition to the Administrator of EPA to reconsider four final benzene
decisions as published in the Federal  Register  (49 FR 23478,
June 6, 1984).  The EPA responded to these concerns  in EPA's  response to
the NRDC petition (50 FR 34144, August 23, 1985).  The EPA  reviewed
NRDC's concerns about correcting the alleged bias  in  the  assessment used
in evaluating the benzene fugitive emission standard. In order  to test
the sensitivity of the regulatory decisions to  changes in the exposure
assessment, EPA recalculated the exposure assessment  used in  the benzene
fugitive emission decision by increasing the ambient  concentrations and,
therefore, exposure by 300 percent.  A factor of 300  percent  was used
because it is the upper limit to the alleged underestimation  of  exposure
based on the  analysis presented in Appendix C of the Benzene  Fugitive
Emissions  Background Information for Promulgated Standards  and  detailed
in Docket  A-79-27,  Item IV-B-18.  After doing so,  EPA concluded  that  the
standard would  not  change based on the new exposure  assessment.   More-
over, the  HEM does  not always predict lower concentrations  than  the  ISC;
it  is dependent  on  the source-specific assumptions.   In  addition, EPA
does not  know whether the ISC would be a  better predictor of ambient
benzene concentrations around  coke by-product plants.  Because of these
considerations,  EPA decided that this additional analysis for coke
by-product plants  was  not warranted.
      The EPA recognizes  that  the  assumption of continuous exposure to
the maximum annual  concentration over a 70-year lifetime may tend to
 overestimate the maximum  individual lifetime risk.   In addition,  for coke
 by-product plants,  the assumption  that the  plants operate continuously at
 full capacity for  70 years  is likely  to overestimate the risk.  On the
                                  9-4

-------
  other  hand,  some assumptions may tend to underestimate the risk.  For
  example, there may be more susceptible subgroups than the population from
  which  the unit risk estimate is derived.  Such susceptibility can differ
  with infirmity, age, genetic composition, or immune-incompetancy.  For
  these  individuals, the cancer risk may be underestimated.  The model
  assumes terrain is flat; for plants in complex terrain where the surroun-
  ding topography is at higher elevation than the emission sources, the
  model  may possibly underestimate maximum annual concentration.   There-
  fore,  overall the Agency believes the leukemia risk estimates are
  plausible, if conservative.
  9.3  UNIT RISK ESTIMATE
      Comment:  Commenter IV-D-13 contends that the  benzene  unit  risk
  estimate used in  the 1984 proposal  has  not  been updated  since 1981 and,
 therefore,  did not  take into  consideration  recently published scientific
 reports on  benzene  carcinogenicity.   The  commenter  maintains  that such an
 update  would increase the unit  risk  estimate  15 times.  Therefore, EPA is
 underestimating risk to  the population  residing near coke by-product
 recovery plants.

      Response:  On  October  17,  1984, the  commenter  (NRDC) petitioned the
 Administrator of the EPA  to reconsider four final decisions regarding
 benzene emissions as published in a Federal Register notice on June 6,
 1984  (49 FR  23478).  Of central  relevance to the petition was the con-
 tention that  the health risk assessment relied upon in June was outdated
 and that the  risk estimate should be revised to reflect the most current
 literature on benzene carcinogenesis.  The EPA agreed to a current review
 of the  published literature and reevaluated the unit risk estimate for
 benzene accordingly.  The methodology for the evaluating of the unit  risk
 estimate is described in a document titled Interim Quantitative Cancer
 Risk Estimates Due to Inhalation of Benzene (Docket  OAQPS 79-3[l]
 VIII-A-4) and is summarized in EPA's response to the NRDC  petition
 (50 FR 34144, August 8,  1985).   In the revaluation  of  the unit  risk
 estimate, EPA pooled the leukemia responses  observed in  the  retrospective
epidemiologic studies  of rubber  hydrochloride  workers exposed  to  benzene
 (Rinsky  et al. 1981) and  chemical  manufacturing workers  exposed to
benzene  (Ott  et  al.  1978),  then  EPA  computed  a  geometric mean  of  each
                                 9-5

-------
point risk estimate.   The  data were  aggregated to encompass a range of
plausible risks observed by  independent  investigators of benzene exposure
in different occupational  settings.  The leukemia incidence observed in a
third epidemiologic study  (Wong  et al. 1983)  of  benzene exposure in
chemical  manufacturing was used  as a comparison  to the computed risk
estimates of the pooled studies.  The  resulting  ratio between these two
sets of data was used to adjust  the  computed  mean estimate.  Based on
these calculations, the unit risk estimate  (the  probability of an
individual contracting leukemia  after  a  lifetime exposure to a benzene
concentration of one part  benzene per  million parts  air) was revised
upwards from 0.0223/ppm (6.9 x  1CT6  per  ug/m3) to 0.026/ppm (8.0 x 1CT6
per pg/m3).  The revised estimate represents  a 17-percent increase in the
estimate used in the June  1984  decisions.
     The significant gap between EPA1 s revised risk  estimate (a
17-percent increase) and the fifteenfold increase  recommended by NRDC
results from a major policy  difference on the appropriate use of animal
versus human data.  The increase advocated by NRDC  is  obtained by  relying
exclusively on the incidence of preputial gland  tumors  of male B6C3F mice
Although the results of an animal  bioassays have been  considered in the
Agency's  revaluation, EPA believes  that the unit  risk  estimate  for
inhalation of benzene is appropriately based on  the principal epidemio-
logic  studies because these studies  are of recognized  quality  and  have
the  greatest relevance in  the estimation of health  risks for the general
population.  Well-conducted epidemiologic studies provide  direct evidence
of  a causal link between the chronic exposure to benzene and  leukemia.
This direct evidence  precludes the  biological uncertainties inherent  in
extrapolating  animal  data to humans.  Given the wide range  of  levels  of
benzene  exposures  and durations of  exposure, the epidemiologic  studies
showed a  threefold to twentyfold increase in risk of leukemia  above
individuals  not  exposed to  benzene.   These findings present unequivocal
evidence  that  chronic  inhalation of benzene  causes leukemia in humans and
therefore falls  within the  criteria of  EPA's current guidelines for
carcinogen risk  assessment  (51  FR 33992, September 24, 1986).   Although a
clear  dose-response  association between carcinoma and benzene exposure
was demonstrated in  rodent  bioassays, the EPA believes that human data,
                                  9-6

-------
 when available,  should be the  principal  factor  in  the derivation of  a unit
 cancer risk estimate.   In the  case  of  benzene,  EPA believes that the
 animal data are  appropriately  used  qualitatively to buttress the
 conclusion regarding benzene's  carcinogenicity.
 9.4  DERIVATION  OF  UNIT RISK ESTIMATE
      Comment:  One  commenter (IV-D-14) expresses the opinion that the
 benzene unit risk estimate overstates  the true  risk by at least one order
 of magnitude.  Moreover,  a minor adjustment of  17  percent in the unit risk
 estimate published  in  the June  6, 1984,  Federal Register notice as
 response to public  comments on  the  listing of benzene (49 FR 23478) did
 not adequately respond to the criticisms made during the maleic anhydride
 proceeding.  According to the commenter, the principal criticism not
 addressed concerned  the inclusion of the Ott et al. 1978 study in the
 derivation of the unit risk estimate.  The commenter maintains that the
 study should not have  been  used because the leukemia incidence was small,
 and there was a  likelihood  of exposure to other chemicals.  In addition,
 the commenter feels  that  EPA inappropriately reclassified one of the
 deaths in the Ott study  as myelogenous leukemia even though the cause of
 death on  the death certificate was  listed as pneumonia.

      Response:  The  EPA has previously responded to these concerns in the
 response  to  public comments concerning the regulation  of benzene as a
 hazardous  air pollutant (49 FR 23478, June 6,  1984).  Although  EPA does
 not  view  the Ott et  al. (1978)  study, taken alone,  as  conclusive evidence
 of  an  association between low level  (2 to 9 ppm) occupational  exposure to
 benzene and  leukemia, the Agency believes that  this work,  combined  with
 other  findings in the published benzene health  literature,  serves  to
 reinforce the public health concerns regarding  benzene exposure.   Ott  et
 al.  observed 3 cases of leukemia in  a cohort of 594 chemical workers  when
 only 0.8 case was expected.  This  represents an  excess risk  of  leukemia  of
 3.75.  The EPA does  not believe that omitting  from  the study the
 individual who suffered from leukemia but died  of pneumonia would be  an
 appropriate change.   In view of the  recognized  causal  relationship  between
 benzene and nonlymphatic leukemias,  EPA believes that a  case of
myelogenous leukemia such as this, if documented, should not be ignored.

                                9-7

-------
     The EPA does  not  view  the  extent  of  confounding exposures in the Ott
et al. study as severe.   The  authors did  exclude  from their analysis
persons known to have  been  exposed  to  levels  of arsenicals, vinyl
chloride, and asbestos,  all of  which have been associated with human
cancer.  This exclusion  eliminated  53  persons from  consideration
including one leukemia victim.   The remaining substances, which include
the suspect carcinogen vinylidene chloride,  have  not been shown to  be
associated with a leukemia  risk in  either man or  animals.  Thus,
inclusion of such exposed individuals  in  the cohort would not  be  likely
to affect the target endpoint for benzene exposure  (leukemia)  in  terms of
increased risk.
9.5  COMPARATIVE RISK FROM GASOLINE MARKETING
     Comment:  Several commenters argue that benzene emissions from
sources other than coke-oven  by-product recovery  plants present  a greater
risk to exposed populations and, therefore,  should  warrant the full
resources of EPA (IV-D-10, IV-D-12, IV-D-17).  They argue that gasoline
service stations and other segments of the gasoline marketing industry
present far  greater risk to residents  living near those facilities than
do coke-oven by-product  recovery plants.
      Response:  The EPA  agrees that there are sources  of benzene
emissions  into the ambient air other than coke oven by-product recovery
plants.  The EPA has evaluated many of the industrial  sources of benzene
(49  FR 23558,  June 6, 1984).   In addition, the EPA has concluded an
extensive  analysis of benzene  and  gasoline vapor emissions  from the
gasoline marketing industry, such  as  service stations, vehicle
 refueling  operations, bulk plants, and bulk  terminals.  On August  19,
1987, the  EPA  Administrator  issued a  notice  of proposed rulemaking to
control  refueling emissions  from gasoline-fueled light duty vehicles, and
to control  the volatility  of gasoline (52 FR 31162).  These proposed
 standards  will  help  protect  the  general  public from the risk  of  cancer
 due to exposure to  benzene,  a  component  of  gasoline vapor, and to
 evaporative gasoline as a  whole.  This proposal  is estimated  to  reduce
 benzene emissions  from  gasoline  refueling by about 90  percent from
 uncontrolled levels.

                                  9-8

-------
      As described in the preamble to the revised proposed  standard,  the
 Administrator determined that control  of benzene emissions from  coke
 by-product recovery plants is warranted to protect  the  public  health with
 an ample margin of safety.

 9.6  COMPARATIVE RISK FROM OTHER SOURCES
      Comment:  A commenter (IV-D-14)  states  that the  risk  to benzene
 exposure from coke by-product plants  does  not  seem  high when compared to
 other risks that are accepted as commonplace in  society.   The  commenter
 suggests that the average leukemia  risk  for  the  entire population
 exposed to benzene emissions  from these  facilities  is 7 x  10~8 (or 7  in
 100,000,000).  Examples  of commonly  accepted risk were given,  e.g.,
 smoking one pack of cigarettes  per day  is  a  risk of cancer of  5 x lO'3.

      Response:   The EPA  does  not average the maximum individual lifetime
 cancer  risk calculations,  but it assumes an  aggregate of risk to the
 population residing within  a  radius of 50  km around coke by-product
 plants.   Aggregate risk  is  a  summation of  all the risks to people
 estimated  to  be  living within the 50 km radius of the facility.  The
 aggregate  risk  is  expressed as  incidences of cancer among all  of the
 exposed  population  after  70 years of continuous exposure to predicted
 ambient  concentrations of benzene emitted from the facilities;  for
 convenience,  the total is often  divided by 70 and expressed as  cancer
 cases per year.  On the other hand,  maximum lifetime risk  reflects the
 probability of contracting cancer to those individuals exposed
 continuously  to the estimated maximum ambient air concentration of
 benzene  for 70 years.  The nationwide risk  to the exposed population
 residing near coke by-product recovery plants due to the plant  emissions
 has been calculated to be about 3 cases of  leukemia  per year.   The
maximum  lifetime risk is  estimated to be 6  x  10~3.
      The EPA does recognize that most  human  activities and events
involve some degree of inevitable risk.   However,  the Administrator has
judged that quantitative  estimates of  the risk  from  other activities
should not be used as quantitative benchmarks for making decisions on
hazardous air pollutants.
                                9-9

-------
9.7  SELECTION OF BENZENE  VS.  POM
     Comment:   One commenter (IV-D-7)  suggests  that  polycyclic organic
matter (POM) compounds result  in a  higher  health  risk  than  benzene
emissions, and that EPA has  not chosen to  regulate POM emissions.
     Response:  The Agency examined the information  regarding benzene and
POM as two different cases.   The POM decision  by  EPA was  based on several
factors, including the great uncertainty as  to  the magnitude of  the  cancer
risk to the public, the fact that  many POM source categories are being
controlled under programs  to attain and maintain  the national ambient air
quality standard (NAAQS) for particulate matter,  and difficulties in
devising control programs  for source categories not  well-regulated  (e.g.,
existing woodstoves, forest fires, and agricultural  burning).  The  EPA
concluded that a more appropriate  regulatory strategy  would be to regulate
specific POM  source categories (e.g., coke oven emissions,  new woodstoves,
and diesel cars  and trucks).
9.8  CONSIDERATION OF OTHER HEALTH EFFECTS
     Comment:  Commenter  IV-D-13 states that EPA's health impact analysis
based on  "cost-benefit" is  flawed because:  (1) the analysis includes only
one of  benzene's hazardous  effects  (leukemia), (2) EPA has ignored  data
showing public  health danger  greater  in degree and broader in kind  than
included  in the  risk  assessment, and  (3) the assessment makes no attempt to
account for concurrent  control of other suspected carcinogens (e.g.,
toluene and xylenes).
      Response:   The EPA does  recognize there are other health effects
associated  with  human exposure to  benzene.  These effects are summarized
in a  recent  review of the health literature by the Occupational  Safety
and Health  Administration (OSHA)  (52  FR 34460, September 11, 1987).  The
toxic effects of benzene  on the hematopoietic  system  include not only
myelogenous leukemia, but also aplastic anemia.  Aplastic anemia is a
 rare, and often fatal,  disorder characterized  by cytopenia in the
 peripheral  blood and  in the bone  marrow.  Aplastic  anemia  is known  to
 progress  into leukemia, and both  diseases are  thought to occur  as a
 result of a common pathogenic mechanism.  Benzene has been associated
 with chromosomal damage in  the circulating  lymphocytes of  exposed

                                  9-10

-------
 workers.   Many  cytogenetic  agents  are known to cause cancer in humans,
 e.g., vinyl  chloride,  arsenic,  and  ionizing radiation.  Therefore, the
 chromosomal  aberrations  seen  in workers exposed to benzene should be
 regarded  as  a serious  consequence  of exposure.  Benzene exposure has also
 been causally linked with multiple  myeloma, various forms of lymphoma,
 and other types  of  cancer.  However, most of the observed cancers were
 not suitable for quantitative risk  assessment because of statistical
 deficiencies in  the observed data,  i.e., the cancer incidence did not
 achieve statistical significance, the relative risk of the specific
 cancer could not be numerically quantified, or exposure only to benzene
 was not identified  in  the studies.  On the other hand, the causal
 association  between benzene exposure and leukemia is a strong statistical
 association, and it provides the most appropriate basis for estimating
 the population risk of cancer through the use of quantitative risk
 assessment.  The EPA does recognize, however,  that the exclusion  of other
 rates of  disease associated with benzene exposure may potentially under-
 estimate  the risk,  but EPA resorted to using those studies having the
 highest degree of statistical  confidence,  demonstrating a strong  associa-
 tion  between leukemia and human exposure,  in the derivation of an estimate
 of  carcinogenic  potency.
      A commenter also pointed  out that EPA's analysis of population  risk
 from  coke by-product recovery  plant emissions  was  weakened by  not
 including other  carcinogens, e.g., toluene and xylenes,  in the evaluation.
 The Agency has reviewed the scientific literature  regarding the
 carcinogenicity  of toluene and xylenes,  and  has  determined that there is
 insufficient evidence to classify the carcinogenic potential of these
 compounds  [Health and Environmental  Effects  Profile for  Xylenes (o-,m-,p-),
 EPA Environmental Criteria and Assessment  Office,  Docket  A-79-16,  Docket
 Item No.  IV-A-7  and  Assessment of Toluene  as a  Potentially Toxic  Air
 Pollutant, 49 FR 22195, Docket A-79-16,  Docket  Item No.  IV-I-4],   The EPA
 has reasonably good  data characterizing  the  magnitude of  benzene  emissions
 from the  source  category.  Other specific  pollutants  that  may  be
carcinogenic to  humans  have  not  been evaluated  in  the emissions.
Simultaneous  exposure to several  chemical  carcinogens is  a  frequent
occurrence in the environment,  and  EPA is  committed to toxicological
research in the  health  risks posed  to exposure of  complex mixtures.   The
                                9-11

-------
ability to predict how the total  mixture  of  toxicants may  interact must
be based on an understanding  of  the  biological  mechanisms  involved in
such interactions.  With  regard  to toluene,  EPA has  reviewed  and evaluated
the current information on health effects, and  has determined that ambient
air concentrations of toluene do not pose a  significant  risk  to public
health and that it is not currently  necessary to regulate  toluene under
the Clean Air Act (49 FR  22195,  May  25,  1984).   In the public notice it
was made clear that EPA is aware of  additional  animal testing that is
under way to investigate  the  potential  carcinogenicity of  toluene, and
that further assessment and review of toluene will occur upon completion
of these studies.  The potential noncarcinogenic health  effects associated
with xylenes are currently under evaluation, and EPA has not  yet  reached  a
decision on this pollutant.  The EPA in the  reproposal of  this standard  is
only focusing on the emission of benzene.  However,  EPA  believes  that
control and reduction of  benzene emissions from coke by-product  recovery
plants will have the added benefit  of controlling and reducing other VOC's
that may also be present  in the emissions.

9.9  ANCILLARY COMMENTS
     Comment:  As an attachment to  their comments on the proposed
regulation, commenter IV-D-13 (NRDC) included their  petition  to EPA  for
reconsideration of four final benzene decisions.  These  benzene decisions
were the withdrawal of proposed national emission standards for benzene
storage vessels, maleic anhydride plants, and ethylbenzene/styrene  plants
(49 FR 23478, June 6, 1984), and the promulgation of standards for
benzene fugitive emissions (49 FR 23512 June 6, 1984).   To complete  the
record for this  rulemaking,  commenter IV-D-18 submitted  supplemental
comments  on behalf of AISI.  They comprise the  responses to the NROC
petition  that were submitted to EPA by the American Petroleum Institute
(API)  and  the Chemical Manufacturers Association (CMA).

      Response:   The  EPA  responded to the NRDC's petition for
reconsideration  on August  23, 1985  (50 FR 34144).   Included  in that
notice were EPA's  responses  to  API's and CMA's  sumittals.  Therefore,
EPA's  response  is  not  repeated  here.
                                 9-12

-------
                10.   EQUIPMENT LEAK DETECTION AND REPAIR
 10.1   DETERMINE EMISSIONS OVER BACKGROUND LEVELS
      Comment:   Commenter IV-D-9 asks, "What is the background level for
 proposed  standards  of 500 ppm above background?"

      Response:   Section 4.3.2 of Method 21 (48 FR 37598, August 18, 1983,
 Docket  Item  IV-I-1)  describes the procedure for determining the presence
 of  emissions over background levels.  Accordingly, the local ambient con-
 centration around the source (i.e., background) is determined by moving
 the probe inlet  randomly upwind and downwind at a distance of 1 to 2
 meters  (m) from  the  source.  If an interference exists with this deter-
 mination because of  a nearby emission or leak, the local ambient concen-
 tration may be  determined at distances closer to the source [but not
 closer than 25 centimeters (cm)].  The probe inlet is then moved to the
 surface of the  source to determine the concentration.  (This procedure is
 described in Section 4.3.1 of the Method.)  The difference between these
 concentrations determines whether there are no detectable emissions
 (i.e., no more than 500 ppm above background).

 10.2  COMPLIANCE WITH LEAK DETECTION AND REPAIR PROGRAM
     Comment:  Commenter IV-D-16  recommends  that  the regulations state
 specifically that a leak (a reading over 10,000 ppm)  is a violation when
 documented during a compliance  inspection.   According to the commenter,
 the 1984 proposed rule provides  no assurance  that  a  component  is actually
 inspected, reported, or  repaired  because this  information could  easily be
 fabricated.   Also,  enforcement  action  is unlikely  because EPA  must  prove
that inspection, reporting,  or  recordkeeping  requirements were not  met.
According  to the commenter,  only  such  a  direct  enforcement  mechanism will
provide incentive for diligent, reliable inspections;  without  this
                                10-1

-------
change, the commenter considers  the  recordkeeping  and  reporting provi-
sions to be only industry  "self-enforcement"  rules.

     Response:   Sealings and  packings  inherently leak;  only the use of
leakless equipment can prevent occasional  leakage.   Because an occasional
leak cannot be  prevented without the use  of  leakless equipment, EPA cannot
accept the commenter's suggestion that a  leak  (a reading  over 10,000  ppm)
should be considered a violation when  documented during a compliance
inspection.  Instead, the  compliance burden  has been placed on the owner
or operator to  repair leaks as soon  as possible after  their detection.
     The commenter asserts that  enforcement  is unlikely because it must
be proven that  recordkeeping  and reporting requirements were not  met  or
that the leaking component was not repaired.   The  EPA  disagrees.  The
regulation states that compliance will be determined by review of
records, reports, performance test results,  or inspections.  By comparing
records and reports of plant  performance  to  the actual  sources during an
onsite inspection, enforcement personnel  will  be able  to  detect
unrepaired sources, unsubstantiated records  regarding  delayed  repair,
falsified records, and a  lack of records  or  reports.  Under these
standards, the records and reports (or lack  thereof) provide usable
evidence of a violation,  and enforcement  action  is likely. Although  the
recordkeeping and reporting requirements, coupled  with onsite  inspections,
are the only measures to determine compliance, EPA believes these
provisions are adequate to ensure diligent monitoring  and repair  of  leaks
by plant personnel and effective enforcement by  EPA.

10.3   DEFINITION OF EQUIPMENT LEAK
      Comment:  Commenter IV-D-13 requests that EPA reconsider  changing
the  definition of an  equipment  "leak"  from 10,000  parts per million
volume (ppmv) to 1,000 ppmv or to the highest level at which  EPA  can
demonstrate, with data, that directed maintenance  does not result in net
emission  reductions.  The  commenter remarks that emissions from equipment
leaking at  rates below 10,000 ppmv are substantial:  about 13  percent of
total  emissions  from  pumps, 2 percent of total emissions  from valves in
                                10-2

-------
 gas service,  16  percent  of emissions from valves in liquid service, and
 16 percent  of total  emissions  from compressors, according to the BID for
 proposed  national  emission standards for benzene fugitive emissions
 (EPA-450/3-80-032a).  The commenter NRDC states that data from a study on
 "directed maintenance"*  summarized in the BID for the proposed new source
 performance standards (NSPS) for equipment leaks at petroleum refineries
 (Docket  Item  II-A-43) contradict its position that a lower definition
 would  not reduce emissions.  In this study, EPA tested the performance of
 both undirected  and  directed maintenance on valves with initial  leak
 rates  less  than  10,000 ppmv.   The EPA found that with directed maintenance
 there  was a net  reduction of 85.2 percent emissions.

     Response:  The  EPA's rationale for selecting the 10,000-ppmv leak
 definition  has been  discussed  in the promulgation BID's for VOC  fugitive
 emissions,  in the  proposal preamble for this rule, and in the rulemakings
 for the synthetic  organic chemicals maufacturing industry (SOCMI)
 (Docket Item  IV-A-2), petroleum refinery fugitive emissions (Docket Item
 IV-A-3),  and benzene fugitive emissions (Docket Item IV-A-1).
     The  key criterion for selecting a leak definition is the mass  emis-
 sion reduction demonstrated to be achievable.   The EPA has not concluded
 that a lower leak  definition is demonstrated.   A net increase in mass
 emissions might result if higher concentration levels  result  from
 attempts  to repair a valve with a screening value between 1,000  and
 10,000 ppmv.  Although many leaks can be repaired successfully at
 concentrations less than 1,000 ppmv,  even one  valve  repair failure  may
 offset many successful  valve  repairs.   Most  data on  leak  repair  effective-
 ness have applied  10,000 ppm as the leak definition  and therefore do  not
 indicate the effectiveness of repair for leak  definitions between 1,000
 and  10,000 ppm.  Even though  data between these  values  are available,
 they are  not sufficient  to support  a  leak definition below 10,000 ppm.
 As the  commenter noted,  although there  is some evidence that  directed
 In  directed maintenance" efforts,  the  tightening  of  the  packing  is
r2Hu.£re^m?lmi:lltarieoVs1^ ard '!  continued  only  to the  extent that  it
reduces emissions.   In contrast,  "undirected" repai r means  repairs such
as tightening valve packings without simultaneously monitoring the result
to determine if the repair is increasing or  decreasing emissions.
                                10-3

-------
maintenance is  more  effective,  available data are insufficient to serve
as a basis for  requiring  directed maintenance for all sources.
     A leak definition  is an  indicator of whether a source is emitting
benzene in quantities  large enough to warrant repairs.  Certainly, a leak
definition of 10,000 ppmv accomplishes this goal.  About 10 percent of
all valves (leaking  and nonleaking) contribute about 90 percent or more
of the emissions from valves.   At a leak definition of 10,000 ppmv,
approximately 90 percent  or more of the leaking valves would be detected,
based on testing in  refineries  and chemical plants (Docket A-80-44,
Docket Items II-A-30 and  II-A-34).  Most seals on pumps and exhausters
leak to a certain extent  while  operating normally, compared to valves
that generally  have  no leakage.  When the seal wears over time, the
concentration and emission rate  increase.  Properly designed, installed,
and operated seals have low instrument meter  readings, and seals that
have worn out or failed have  readings generally greater than 10,000 ppmv.
Over 90 percent of emissions  from exhauster seals and pump seals in light
liquid service  are from sources  with instrument readings greater than or
equal to 10,000 ppmv.
     The EPA believes that there is only a small potential emission
reduction for sources having  benzene concentrations between 1,000 and
10,000 ppmv.  Therefore,  using  a lower  leak definition would not increase
emission reductions significantly, even if EPA judged that repair was
effective for leaks of 1,000  ppmv.   In the proposal BID for the petroleum
refinery fugitive emissions  NSPS (Docket  Item II-A-43, p. 4-8), there is
a  comparison of the percentage  of total mass  emissions affected by
selecting a 10,000-ppmv leak  definition over  a 1,000-ppmv leak definition.
These percentages represent  maximum theoretical emission reductions that
could be expected if the  sources were  instantaneously  repaired to a zero
leak rate and no new leaks occurred.   For  pump seals  in liquid service  and
compressor seals (similar to exhausters  in coke by-product plants), the
estimated decrease is only 6 to 7  percent; for valves  in gas  service, it
is only  1 percent.  This small  potential  decrease  in  emissions may  be
offset by attempting to  repair sources  with  low  leaks.
                                10-4

-------
      In  summary, EPA does not disagree with NRDC that additional emission
 reductions potentially could be achieved by reducing the leak definition
 from  10,000 to 1,000 ppmv.  However, EPA has concluded that 10,000 ppmv
 is  a  demonstrated and effective leak definition (i.e., there are large
 enough emissions that repair can be accomplished with reasonable costs),
 but has  not concluded that 1,000 ppmv is a demonstrated leak definition.
 Until EPA has adequate data to support the repair potential associated
 with  leak definitions such as 1,000 ppmv, EPA is selecting the clearly
 demonstrated leak definition of 10,000 ppmv instead of a lower level.
 10.4  ON-LINE VALVE REPAIR
      Comment:  Commenter IV-D-13 refers to the 1984 proposal BIO
 discussion indicating that on-line repair of valves by drilling into the
 valve housing and injecting a sealing compound is growing in acceptance,
 especially because of safety concerns.  The commenter states that this
 discussion means the practice has been demonstrated and should be
 required in the final standards.
     Response:  The EPA does not agree that acknowledging a promising
 repair method must be interpreted as meaning "demonstrated" within the
 context of the CAA, or that acknowledgment alone constitutes sufficient
 justification for a regulatory requirement.  The 1984 proposal  BIO does
 state on page 4-52 that drilling into the valve housing and injecting  a
 sealing compound is a practice "growing in acceptance" for the on-line
 repair of valves.  Although the term "growing in acceptance" can be
 interpreted to mean that  the practice has been reported as one repair
 method, the phrase also implies reluctance by plant  owners and operators
 to  use the technique.  This hesitancy would be due,  in part, to factors
 such as the type and location  of the valve or the  nature  of the leak.
 For example,  plant personnel may prefer to tighten the packing gland
 rather than drill  into the housing  of a critical  valve containing  a
 potentially explosive mixture.  Or,  as  discussed  in the preamble to the
 1984 proposed  rules  (49 FR 23533),  the valve  (or the leak)  may  require
 removal  or isolation.   Also, this  repair  approach cannot  be used on
control  valves  or other block  valves  that are  frequently  operated  because
the valve would  then be destroyed and  must  be  replaced.   Because of
                                10-5

-------
uncertainties regarding the applicability of this  method to the different
types of valves and varying repair conditions,  this technique cannot  be
considered fully demonstrated at this time.
     Also, the long-term practicability and  cost effectiveness of this
method are unknown.  Depending on the valve  and other factors, this
approach may be no more than a temporary repair until  the next unit
shutdown.  Without such information,  the technique cannot be (and was
not) evaluated as the basis of the standards and a potential  regulatory
requirement.
     Even if the practice were fully  demonstrated  and its long-term
practicability and costs were known and deemed  superior,  an ensuing
regulatory requirement still might not be appropriate.   The leak
detection and repair program places the regulatory burden on plant owners
and operators to detect and repair leaks as  they occur.   Unless a shut-
down is required, all  valves must be  repaired.   A  repair period of 5 to
15 days has been provided to allow plant owners or operators the  flexi-
bility necessary for efficient handling of repair  tasks  while main-
taining an effective emission reduction.  To provide this flexibility,
the standards do not dictate any single repair  method—only the repair;
delays are allowed only under limited circumstances.  If any plant owner
or operator chooses to apply this method, it is certainly not precluded
under the standards.  To require this method for all valves,  however,
would be premature and unwarranted.
10.5  EQUIPMENT LEAK REPAIR PERIOD
     Comment:  Commenter IV-D-13 recommends  that the repair period for
equipment leaks be 24 hours for the first attempt  (rather than 5  days,  as
proposed), with completion within 5 days as  opposed to 15 days.  The
commenter suggests that the shorter timeframe is adequate because moni-
toring personnel should be accompanied by workers  prepared to fix any
leak upon detection or immediately afterwards.

     Response:  The EPA's justification for  proposing the 5-day first-
attempt-at-repair interval and the 15-day repair period  for pumps,
valves, and exhausters was described  in the  preamble to  the 1984
proposed rule at 49 FR 23541.
                                10-6

-------
      The  selected  repair intervals provide maximum effectiveness of the
 leak  detection  and  repair program by requiring expeditious emission
 reduction, while allowing the owner or operator the time to maintain a
 reasonable overall  maintenance schedule for the plant.
      During development of the standards already promulgated for equip-
 ment  leaks in refineries and chemical plants, EPA personnel made a con-
 certed  effort to investigate and gain knowledge of the industry main-
 tenance practices.  In EPA's technical  judgment, an initial attempt at
 repair  within 5 days is ample for all simple field repairs.  A 24-hour
 period  following leak detection is often not long enough to allow main-
 tenance personnel to identify the cause of a leak and then to attempt
 repair.   Although plants could schedule repair personnel to accompany the
 monitoring team in  advance of monitoring, emergency situations or criti-
 cal equipment problems could easily postpone these arrangements.  Al-
 though  some or perhaps even most repairs can be made within 24 hours, it
 is not  practical to require an attempt  to repair all equipment within 24
 hours.  The EPA has not been able to distinguish between equipment that
 could and could not always be repaired  within 24 hours.   In addition,
 with the  commenter's approach, repair crews would spend  much of their
 time on an inspection with few needed repairs.   The costs of this
 approach  have not been estimated by EPA because it is not practical.
 Furthermore,  the owner or operator has  an incentive to repair leaks as
 quickly as possible to prevent additional  product losses.
     A  15-day repair interval  provides  time for isolating leaking
 equipment for other than simple field repairs.   A 5-day  interval, as
 suggested by  NRDC,  however,  could cause scheduling problems in  repairing
 valves that are not conducive  to simple field repair and that  may require
 removal  from  the process for repair.  A 15-day  interval  provides the
 owner or operator with  enough  time for  determining precisely  which  spare
 parts are needed and sufficient  time  for  reasonably  scheduling  repair.
 In addition,  a 15-day  repair interval allows  more efficient handling  of
more complex  repair tasks  while  maintaining an  effective reduction  in
equipment  leaks. Again, the owner or operator  has  an  incentive  to  repair
leaks promptly.
                                10-7

-------
     The commenter's suggestion  that  leaks  can  be  detected  and  repaired
within a shorter timeframe if repair  workers  accompany  monitoring  per-
sonnel may be helpful  for plants able to make such arrangements.

10.6  DELAY OF REPAIR
     Comment:  Commenter IV-D-13 recommends that the  proposed provisions
for delay of repair beyond a unit shutdown  be tightened to  prevent
abuses.  The ccmmenter suggests  that  it  is  possible under the proposed
rules to claim lack of equipment in stock as  a  reason to delay  when
"there was in fact plenty of time to  anticipate stock needs."
     Response:  The delay of repair provisions  included in  the  standards
is necessary to ensure technical achievability  and reasonable costs.
Delay of repair beyond a unit shutdown is not allowed for any types  of
equipment other than valves.  Spare parts for valves  (e.g., packing  gland
bolts and valve packing materials) can be stocked  so  all leaks  that
cannot be repaired without a process  unit shutdown can  be repaired
during the shutdown.  In a few instances, however, the  entire valve
assembly may require replacement.  The standards address this situation
by allowing delay of repair beyond a  process  unit  shutdown  only if the
owner or operator can demonstrate that a sufficient stock of spare valve
parts has been maintained and that the supplies had been depleted.   If  an
owner or operator has sufficient time to obtain a  piece of  equipment, he
or she could not reasonably claim a delay of  repair as  a result of lack
of equipment.
10.7  ALTERNATIVE STANDARD FOR VALVES
     Comment:  Commenter IV-D-13 states  that  the proposed alternative
performance standard for leaking valves  should  be  1 percent rather than  2
percent.  According to the commenter, the allowance of  2 percent leaking
valves will result in an average leak rate well over  1  percent.  The
commenter believes it inappropriate for EPA and the public  to bear all
the risk of statistical sampling error.

     Response:  The alternative standards for valves  were  provided for
owners and operators of units exhibiting low leak  frequencies because the
cost effectiveness of monthly/quarterly leak  detection  and  repair becomes
                                10-8

-------
 unreasonable at low leak frequencies.   The 2-percent  limit  is  intended  to
 be used as an upper limit for determining compliance  with the  alternative
 standards.  If a process unit is subject to and exceeds  the  2-percent
 limit, the unit does not comply with the standard  and is subject  to
 enforcement actions.  The EPA believes  that enforcement  action should be
 taken when noncompliance is  supported by the facts.   Thus, because the
 2-percent limit accounts for the uncertainty in setting  this numerical
 emission limit, EPA can proceed with enforcement action  clearly supported
 by the facts.   Although there is a  regulatory difference between  a
 2-percent and  a 1-percent limit, there  is  no significant practical
 difference to  either plant owners and operators or to EPA between limits
 of 1 percent or 2  percent of valves  leaking.   An owner or operator of a
 process unit would implement the same control  measures to comply with the
 alternative valve  standard whether the  limit  were set  at 1 or  2 percent.
 The NRDC implies that  the 2-percent  limit  is  set in industry's favor; in
 a  practical  sense,  however,  there is little  difference in terms of
 numbers of valves  leaking when  maximum  limits  and averages are compared.
 For example, a  typical  process  unit with about 105 valves in service is
 allowed to have  no  more  than  2  valves leaking  out of the control at the
 2-percent  maximum  limit.  A  1-percent limit would allow no more than one
 valve  leaking.  The  work  practices and equipment used to achieve a rate
 of  2 valves  leaking  out  of 105  valves in a process  unit at  any  one time
 are the  ones that would  be used to achieve a 1-percent limit.
 10.8    EXEMPTION FOR DIFFICULT-TO-MONITOR VALVES
     Comment:  Commenter  IV-D-13 states  that the exemption  for  difficult-
 to-monitor valves is not warranted.   Valves above 2 m, according to  the
 commenter, can be reached by  a sampling  probe on a  boom or by a mobile
 "cherry picker."

     Response:   The EPA disagrees that the exemption  for  difficult-to-
monitor valves  is unreasonable.   The  intent of the  standards  is to
monitor valves  that can be reached with  the portable  ladders  or with
existing supports such as platforms  and  fixed ladders.  A valve  only may
be  exempted from monthly monitoring,  provided:   (1) the plant owner or
operator demonstrates that the valve  cannot be monitored  without

                                10-9

-------
elevating monitoring personnel  more  than  2  m  above  a  support surface,
(2) the valve is in an existing process unit,  and  (3) the  plant owner or
operator follows a written plan requiring monitoring  at  least once per
year.
     The EPA compared the cost  effectiveness  of  scaffolding to annual,
quarterly, and monthly monitoring  of difficult-to-monitor  valves in petro-
leum refineries (see Docket Item IV-B-4).   Based on this analysis,
EPA found the costs of using scaffolding  for  annual monitoring of benzene
emissions from difficult-to-monitor  valves  to  be reasonable compared to
similar costs for monthly and quarterly programs.   These costs were
estimated as the base cost for  monitoring and  maintenance  for readily
accessible valves plus the additional  labor cost for  scaffolding.  No
purchase cost of scaffolding was included because the plant was assumed
to have purchased this equipment for maintenance.   However, the previous
purchase of a sampling probe on a  boom or a mobile  cherry  picker cannot
be assumed.  Consequently, these purchase costs  would result in even
higher costs for each difficult-to-monitor  valve.   Some  valves may be
located in plant areas that are not  accessible for  repair  work using a
mobile cherry picker or a sampling probe  on a  boom.
     Other cost and technical problems are  associated with use of a
mobile cherry picker or sampling probe on a boom for  monitoring.  In
general, few leaking difficult-to-monitor valves are  expected at a
typical by-product plant.  Although  some  valves  may be  located in groups
(e.g., elevated pipe racks), others  may be  scattered  throughout the
plant.  The additional labor required for driving,  scheduling, and
transporting the vehicle from valve  to valve  would  further increase the
costs previously discussed.
     The EPA considers impractical NRDC's suggestion  for use of a
sampling probe on a boom because it  lacks the  precision  necessary for
effective monitoring.  The monitoring team  would not  be  able to move the
probe around the leaking valve  stem  or as close  as  possible to other
potential leak interfaces, as required by the  standard.  Considering the
high cost and the technical infeasibility,  EPA considers that no benefits
would be achieved by this approach.
                                10-10

-------
 10.9  ALTERNATIVE STANDARD FOR  OPEN-ENDED  VALVES  OR  LINES
      Comment:   Cornmenter IV-D-14  suggests  that  an alternative  standard of
 no detectable  leaks (10,000 ppm)  be  considered  for open-ended  valves or
 lines in lieu  of the  proposed equipment  requirement  of a cap or plug.
 This alternative, coupled  with  monthly monitoring, would satisfy the EPA
 goal of leak prevention.

      Response:   The standards would  require open-ended valves  and lines
 to be equipped  with a cap,  plug,  blind flange,  or a  second valve depend-
 ing on the  individual  application.   If a second valve is used, the  up-
 stream valve must be  cleared first before the downstream valve is closed
 to prevent  process  fluid from being trapped between the valves.  The
 standards would also  allow  a bleed valve or line  in a double block and
 bleed system to remain open when  the line between the two block valves is
 vented.  The bleed  valve must be  capped, however, when not opened.  This
 provision is intended to avoid  plugging out-of-service bleed valves in a
 block and bleed system.  These equipment and operational  requirements
 will  reduce  uncontrolled benzene  and VOC emissions from open-ended valves
 or lines by  100 percent.
      The commenter  suggests an alternative standard of no detectable
 leaks, with  applicable leak detection and repair  (LDAR)  requirements.
 Application  of  a  cap, plug, blind flange, or second valve is the  only
 effective method  available for reducing or eliminating emissions  from
 open-ended valves or lines.  In EPA's judgment,  this  equipment  still  would
 be  necessary to meet the repair requirements of  the LDAR  program,  even
 with  a leak definition of 10,000 ppm.  However,  plant owners or operators
 would continue to bear the additional  cost  of  monthly monitoring.
     The LDAR program, with a leak definition  of 10,000 ppm,  should  not
 be  confused with a no  detectable emission limit.  Plants  subject to  a
 no  detectable emission limit would be required to  conduct  an annual
 performance test for each open-ended  valve  and line.   The  plant would be
 out of compliance if emissions  from any  of  the sources exceeded 500  ppm
 above background, as measured by Reference  Method  21.  Again, use of  a
cap, plug,  blind flange,  or second valve  still would  be needed  to ensure
compliance.   Additional  costs also should be anticipated  for'the record-
keeping and  reporting  requirements associated  with performance  testing.
                                10-11

-------
Although this approach does not  seem reasonable  because  it  would  require
the same controls at additional  cost,  the  owner  or  operator could apply
to use this method as an alternative means of compliance with  the
standard.
                                 10-12

-------
                    11.   RECORDKEEPING AND REPORTING
 11.1  ALTERNATIVE  MONITORING AND RECORDKEEPING
      Comment:   Commenter IV-D-9 asks if monitoring and recordkeeping
 requirements  can be  modified if a technology better than that required by
 the  standard  is used.

      Response:  Section  61.136 of the standards describes the procedures
 for  obtaining EPA  approval of alternative means of emission reduction
 that are equivalent  to or better than the equipment, design, operational,
 or work practice standards required by the standard.  Provisions are
 included that allow  EPA  to include requirements necessary to ensure
 proper operation and maintenance.  Consequently, if an owner or operator
 applies for use of an alternative means of emission limitation,  EPA would
 consider requiring monitoring,  recordkeepi ng, and reporting requirements
 appropriate for the  alternative on a case-by-case basis.

 11.2  RETENTION PERIOD FOR RECORDS AND -REPORTS
     Comment:  Commenter IV-D-13 argues  that records and  reports should
 be maintained permanently (or for a  minimum  of 5 years) because  of the
 availability of automated data  systems.   If  audits  or inspections occur
 only once every 1 or 2 years,  it is  important to have  available  complete
 records for more than 2 years.

     Response:  The Office of Management  and Budget  (OMB)  implementation
of the Paperwork Reduction Act  of  1980  (PL-511)  specifies  3  years  as  a
limit beyond which  it becomes burdensome  for plant  owners  and operators
to keep records  other than health, medical,  or tax  records.  The  EPA
selected  the 2-year period based on  considerable enforcement experience.
The  2-year  limit, although less  than that allowed by OMB,  applies  to
significantly  detailed  plant records that would  help enforcement personnel
                                11-1

-------
assess compliance with the standards.   The  EPA considers  the  burden
associated with these records to be  reasonable for the  2-year period.
However, EPA does not agree with the  commenter that,  if EPA audits a
plant less frequently than once  every  1 or  2 years, EPA would not be able
to ensure compliance with the standard.  Once  every 2 years is frequent
enough to review and determine compliance for  most owners or  operators
affected by the standard.  Thus, it would not  be  necessary for a plant to
keep records longer than 2 years.  For these reasons, EPA believes that
it is not necessary to require that owners  and operators  retain records
longer than a 2-year period.   Permanent retention by  automated data
systems was not considered necessary  for effective enforcement.

11.3  ENFORCEMENT BASED ON RECORDS AND REPORTS FOR EQUIPMENT  LEAKS
     Comment:  Commenter IV-D-13 states that the  recordkeeping and
reporting requirements are not strong  enough for effective enforcement.
In support, the commenter cites  the failure of the proposed rules to
require identifying tags for  leaking equipment to facilitate  identifi-
cation of "repeat offenders"  and the  failure of the rules to  require
reporting of the specific identity of  leaking  equipment—only totals.

     Response:   Tagging is not specifically required  by the standard, but
the standard does require some form of weatherproof and readily visible
identification  that would enable plant personnel or EPA inspectors to
locate leaking  sources readily.   Tagging is a  useful  method of identifi-
cation that has been used in  leak detection and repair  programs.  Any
form of identification is acceptable,  however,  as long  as it  is weather-
proof and readily visible.  For  example, a  process unit may have a system
of identifying  markings on valves and  a diagram that  is available to
allow easy location of the marked valves.   This type  of identification
system works as effectively as tagging and  is  often used by chemical
plants and petroleum refineries. To  require tagging  would be unneces-
sarily restrictive if an owner or operator  can identify leaking equipment
just as effectively by other  means.
                                11-2

-------
                           12.  MISCELLANEOUS
 12.1  ALTERNATIVE MEANS OF EMISSION LIMITATION
    Comment:  Commenter IV-D-14 recommends revised requirements for
 collection and verification of test data to demonstrate equivalence of an
 alternative means of emission limitation.  In general, the commenter
 suggests permitting demonstration of equivalence based on design and
 engineering data, with verification after the implementation of controls.
 This approach would solve the timing problem encountered in collecting
 and verifying data before permission is granted because actual  data may
 not be available until after controls are installed and relevant data
 from other facilities may not be available.

     Response:  The 1984 proposed regulation provided the plant owner or
 operator the opportunity to offer a unique approach to demonstrate the
 equivalency of any means of alternative emission limitation to  the
 standard.  If an owner or operator could demonstrate sufficiently the
 equivalency based on design and engineering data,  EPA will  consider that
 approach acceptable.

 12.2  DEFINITION OF TAR DECANTER
     Comment:   Commenter IV-D-14 recommends a revised definition of "tar
 decanter."  The commenter argues that EPA assumes  98-percent control
 efficiency on  tar-intercepting sumps and 95-percent control  for decanters
 because sumps  separate light  tars while decanters  separate  heavy tars and
 sludge.   However, some sump units separate light and heavy  tars, requir-
 ing a sludge conveyor similiar to that used by the decanter.  Because of
the conveyor,  the sump cannot be endorsed for 98-percent  control.   The
commenter recommends a revised definition of tar decanter to include  "any
vessel,  tank,  or other type control  that functions to separate  heavy  tar
and sludge from flushing liquor."
                                  12-1

-------
     Response:   The  EPA  agrees with the commenter's suggestion that the
definition of "tar decanter"  be  clarified.   In  response, the revised
proposed regulation  contains  the following definition:
          "Tar decanter" means any vessel, tank, or other type of
          container  that functions to  separate  heavy tar and sludge from
          flushing liquor by  means of  gravity,  heat, or chemical emulsion
          breakers.   A tar decanter may also be known  as a
          flushing-liquor decanter.

12.3  DEFINITION OF  EXHAUSTER
     Comment:  Commenter IV-D-9  asks to what extent upstream and down-
stream of the rotors does the definition  of  "exhauster" extend?

     Response:  In response to the  commenter's  question, EPA has revised
Section 61.131 of the proposed standards  to  include the following
definition for "exhauster":
          "Exhauster" means a fan located between  the  inlet  gas  flange
          and outlet gas flange  of  the coke  oven  gas  line that  provides
          motive power for coke  oven  gases.

12.4  WAIVER REQUESTS
     Comment:  Commenter IV-D-14 recommends  that  the  standard  allow any
waiver  request submitted within 90  days to be granted on  an  interim basis
until final  determination  is made.   The commenter indicates  that many
waiver  requests will be made and suggests that it is  unlikely  that all
waivers can  be submitted and reviewed by EPA within 90 days  of the
effective date.  Without such a provision, operators  will  be in a
technical state of  noncompliance until the final  determination can be
made.
     Response:  The CAA  clearly states in Section 112(c)(l)(8) that an
existing  source shall comply with the standard within 90 days of the
effective date  unless the  source is operating  under a waiver of
compliance.   Section 112 makes  the granting  of a waiver contingent upon
EPA's  finding "that such  period is necessary for the  installation of the
waiver to assure  that the  health of persons  will be protected from
 imminent  endangerment."  Granting a waiver  before making these findings

                                  12-2

-------
 would  be  inconsistent with the statute.  Thus, EPA has not included the
 commenter's  recommendation.  The owner or operator of a source should
 submit the waiver application as soon as practicable to allow time for
 the Agency to make a determination within the 90-day period after the
 effective date.  One should note that the owner or operator should take
 advantage of the time between reproposal and promulgation to prepare
 significant portions of a plan for achieving compliance.  In addition,
 the source should continue to take all possible steps toward achieving
 compliance while the Agency is evaluating the waiver application.
 12.5   NEED FOR ADDITIONAL ENFORCEMENT RESOURCES
     Comment:  Commenter IV-D-1 requests that the proposed standard be
 simplified to reduce the enforcement resources needed to ensure
 compliance.  According to this commenter, additional  enforcement
 resources will be necessary or a reduction in enforcement activities in
 other  areas will be required.

     Response:  The commenter did not describe specifically the
 provisions of the regulation he considers would require resource-
 intensive enforcement.  The regulation inherently has many aspects
 because by-product plants comprise several  sources with different
 applicable control techniques.  However, EPA has designed the reporting
 requirements to be as simple as possible, while also  providing
 enforcement personnel indications of potential  noncompliance.
 12.6  SELECTION OF FORMAT
     Comment:  Commenter IV-D-17  states that the regulations  do not make
 clear why some requirements are expressed as equipment  standards  while
 others are expressed  as  emission  limits.   The commenter asks  specifically
why different standards  are applied for different process  sources,  such
 as tar decanters,  tar dewatering,  and the naphthalene sump (e.g.,
 naphthalene processing).

     Response:   The type of standard (e.g.,  emission, equipment,  work
practice,  design,  or  operational)  depends not on  the  function of  the
source, as implied by the commenter,  but  on  the  control  technique
selected  as  the basis of the  standard.
                                  12-3

-------
     Section 112  of  the  CAA  requires that an emission standard be estab-
lished unless such a standard  is  not feasible to prescribe or enforce.
"Not feasible to  prescribe or  enforce" means that the pollutant cannot be
emitted through a conveyance designed and constructed to emit or capture
the pollutant or  that measurement methodology is not practicable to apply
because of technological  or  economic limitations.   If an emission
standard is not feasible to  prescribe or enforce, one of the other types
of standards (including  any  combination) can be applied.
     Gas blankecing  has  been selected as the basis  of the standards for
both tar decanters  and tar-dewatering vessels.  In  the  original preamble
discussion in 49  FR  23528, EPA explained why an emission standard, such
as a zero emission  limit, was  not feasible  for gas-blanketing systems.
Such a standard could not be achieved on a  continuous basis because,
after installation  of the system, vapor  leaks occur occasionally because
of the gradual deterioration of sealing  materials,  even when proper
operation and maintenance procedures  are applied.   Fugitive emissions
also may be released from openings such  as  access  hatches and sealing
ports.  These fugitive emissions cannot  be  eliminated because the
openings are necessary for  proper operations  and maintenance of the
sources.  An emission standard, it was  argued, would be infeasible to
prescribe or enforce not only  because it could  not  be achieved on  a
continuous  basis (and thus  was not appropriate),  but because these  vapor
leaks and fugitive emissions could not  be  emitted  through  a  conveyance
designed and operated to emit  or capture the emissions.  Therefore,  mass
emissions could  not  be measured.  For these reasons, an equipment
standard rather  than an emission standard  (i.e.,  limit) was  developed  for
gas-blanketed  sources.
      The commenter  also questions why different  standards (e.g.,  zero
emissions)  have  been established for naphthalene sumps  (processing).   In
this  case,  a  process modification requiring the collection of naphthalene
in tar  (for foundry  coke plants) or wash-oil  (for furnace coke plants)
was selected  as  the basis of the revised proposed standards.   Collecting
naphthalene in tar  (or wash oil) would eliminate naphthalene-processing
operations  (including naphthalene sumps) and the emissions that result
                                   12-4

-------
 from separating  naphthalene  from  the  hot well of  a direct-water final
 cooler.   Because these  emissions  and  emission sources can be eliminated by
 such a modification,  a  zero  emission  limit  is considered feasible to
 prescribe and  enforce and  has  been  selected as the format for the revised
 proposed  standards  for  naphthalene-processing.

 12.7  LIGHT-OIL  SUMP  CONTROL EFFICIENCY
      Comment:  Commenter IV-D-14  states that the  98-percent control
 efficiency assigned to  light-oil  sumps is unsupported and should not be
 used as the basis for qualifying  an alternative means of emission limi-
 tation.   The commenter  recommends instead application of semiannual
 monitoring to  determine that there are no detectable emissions.

      Response:   The 98-percent control efficiency assigned to light-oil
 sumps is  legitimately supported on engineering judgment.  As discussed in
 the  preamble to  the 1984 proposed rules in 49 FR 23537-23539, the control
 efficiency of  source  enclosure is theoretically 100 percent.  However,
 eventual  deterioration of the gasket seal  (of the cover) may result in
 occasional  leaks, even with proper operation and maintenance.  Because mass
 emissions  from these  leaks cannot reasonably be measured, EPA con-
 servatively  judged the control  to obtain a 98-percent emission reduction.
 The  98-percent efficiency for the light-oil  sump is consistent with the
 98-percent  efficiency assigned to gas-blanketing systems.
      The  semiannual  monitoring provisions  do not constitute the control
 itself.   Rather, semiannual monitoring of  the light-oil  sump cover and
 gas-blanketed sources is required to ensure  proper operation and main-
 tenance (O&M) of the sealed enclosures, i.e.,  to locate  and  repair any
 leaks that may have developed in the control system.   Thus,  the commenter1s
 recommendation is to allow any  alternative control technique provided  it
 uses the same O&M procedures.  However, the  equivalency  of  an alternative
 control to the standard  must  be based  on the emission reduction  achieved by
 the control itself.   Then,  the  provisions  necessary for  ensuring  proper
would have to be  determined specifically for the  alternative control
 system.  Without  further information,  EPA  believes that  the  98-percent
 value is  the best estimate  available for comparing the efficiency  of an
alternative control  system.
                                  12-5

-------
          Appendix A



Environmental  Impact Analysis

-------
 I
to
                                                       TABLE A-l.   FURNACE COKE UV-PROOUCT RECOVERY PLANTS:
                                                                COKE-OVEN AND PLANT  CAPACITY STATUS
                                                                           (1,000 Mg/yr)
No.
lb
2
3


4
5
6

7

8

9



10b

11 ,
12b

13

14b
15



16

17

18

19b






Plant
LTV Steel, Thomas, AL
New Boston, Portsmouth, OH
Wheeling-Pitt, Monessen, PA


Lone Star Steel, Lone Star, TX
LTV Steel, So. Chicago, IL
National Steel, Granite
City, IL

lnter)ake, Chicago, IL

LTV Steel, Gadsden, AL

Rouge Steel, Dearborn, MI



U.S. Steel, Fairless Hills, PA

LTV Steel , Warren, OH
LTV Steel, E. Chicago, IN

Arroco Inc. , Ashland, KY

Weirton Steel, Brown's Is., WV
U.S. Steel, Provo, UT



LTV Steel, Aliquippa, PA

Bethlehem Steel ,
Lack a wanna, NY

National Steel, Detroit, MI

U.S. Steel, Lorain, OH






Battery
no.
1
1
1A
IB
2
C
2
A
B
C
1
2
2
3
A
Ax
B
Dx
1
2
4
4
9
3
4
1
1
2
3
4
Al
A5
7
8
9
4
5
D
G
H
I
J
K
L
Battery
capacity
315
364
195
195
100
507
563
285
285
298
291
291
379
379
256
57
312
153
458
458
945
432
516
349
614
1,097
290
290
290
290
604
614
382
382
528
473
924
218
218
218
218
208
208
208
Status*
2
0
2
0
0
0
0
0
0
3
0
o
0
0
0
0
0
0
2
2
0
2
2
0
0
2
0
0
0
0
0
o
0
0
0
0
0
2
2
2
2
2
2
2
No. of
ovens
65
70
37
37
19
70
60
45
45
47
50
50
65
65
45
10
55
27
82
82
85
75
87
76
70
87
63
63
63
63
106
56
76
76
73
78
85
59
59
59
59
59
59
59
Online
0
364
295


507
563
570

582

758

778



0

945
0

963

0
1,160



1,218

1,292

1,397

0


•



Hot
idle
0
0
0


0
0
0

0

0

0



0

0
0

0

0
0



0

0

0

0






Cold
idle
315
0
195


0
0
0

0

0

0



916

0
948

0

1.U97
0



0

0

0

1,496






Under
u
u
0


0
0
298

0

0

0



0

0
0

0

0
0



0

0

0

0






Existing
plant
J15
364
490


b(J7
563
570

582

758

778



916

945
948

963

1,097
1,160



1,218

1,292

1,397

1,496






                   Footnotes  at  end  of  table.
                                                                                                                                   (continued)

-------
                                                  TWJLt A-l.   (continued)
                                                                                                                 " ExfbtTny
No.
20



i\ '





22


23
24




25&



26



27








28






Battery
Plant no.
Wheel iny- Pitt, t. Steubenvil le, WV 1
2
3
8
LTV Steel, Cleveland, OH 1
2
3
4
6
7
Arraco Inc. , Middletown, OH 1
2
4
Bethlehem Steel, Burns 1
Harbor, IN 2
LTV Steel, Pittsburgh, PA PI
P2
P3N
P3S
P4
U.S. Steel, Fairfield, AL 2
5
6
9
Bethlehem Steel, Bethlehem, PA A
2
3
5
Bethlehem Steel , Sparrows 1
Pt. , I'D 2
3
4
5
6
11
12
A
Inland Steel, E. Chicago, IN 6
7
8
9
10
11
C
Battery
capacity
199
199
215
896
274
274
274
274
332
332
664
664
448
895
895
340
340
340
340
432
818
320
320
364
809
522
522
400
273
263
273
273
273
273
365
365
1,148
278
372
395
395
450
995
830
Status-*
0
0
u
U
0
0
0
2
0
0
0
0
0
u
0
0
0
0
0
0
2
2
2
2
0
0
0
0
1
1
2
2
2
2
0
0
0
0
0
0
0
0
0
2
No. of
ovens
47
47
bl
79
bl
bl
51
51
63
63
57
57
76
82
82
59
59
59
59
59
57
77
77
63
80
102
102
80
63
60
63
63
63
63
65 *
65
80
65
87
87
87
51
69
56
Hot Cold unuer plant
Online idle idle construction total
l,bU9 0 U u l,bl)9



1,486 0 274 u l,7t>U





1,776 0 U U 1,776


1,790 000 1,790
1,792 000 1,792




0 0 1,822 0 1,822



2,253 000 2,253



1,878 536 1,092 0 3,506








2,885 0 830 0 3,715





t rnnf l nllpd )
Footnotes at end of table.

-------
                                                            TABLE A-l.  (continued)
i
Cn
Battery
No. Plant no.
29 U.S. Steel, Gary, IN 1
2
3
5
7
13
15
16
30 U.S. Steel, Clairton, PA 1
2
3
7
8
9
15
19
20
21
22
B
Total (30 plants)
(24 plants)
Battery
capacity
843
995
995
279
279
279
279
279
296
296
296
296
296
296
302
535
535
535
535
1,076
42,102
39,508
Status*
2
0
0
1
1
1
0
1
1
1
1
1
1
1
1
0
0
0
0
0

No. of
ovens
85
57
57
77
77
77
77
77
64
64
64
64
64
64
61
87
87
86
87
75
7,100
5,935
Existing
Hot Cold Under plant
Online idle idle construction total
2,269 1,116 843 0 4,228







3,216 2,078 0 0 5,294











31,646 3,730 9,828 298 45,804
31,646 3,730 3,234 298 39,210
         Note:   Data current  as  of  November  1984.


        j> Status:  0 =  online;  1 = hot  idle; 2  = cold idle; and  3  = under construction.
        ° Denotes cold  idle  plants.

-------
                                    TABLE A-2.  FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS:  COKE-OVEN AND PLANT CAPACITY STATUS

                                                                         (1,000 Mg/yr)
 i
cr>
No.
1

2

3
4

5

6
7


8b

9


10


11

12




13


14


Battery
Plant no.
Chattanooga Coke,
Chattanooga, TN
IN Gas, Terre Haute, IN

Koppers, Toledo, OH
Empire Coke, Holt, AL

Koppers, Erie, PA

Tonawanda, Buffalo, NY
Carondolet, St. Louis, NO


AL Byproducts, Keystone, PA

Citizens Gas,
Indianapolis, IN

Jim Walters, Birmingham, AL


Shenango, Pittsburgh, PA

Koppers, Woodward, AL




AL Byproducts, Tarrant, AL


Detroit Coke, Detroit, MI
Total (14 plants)
Total (13 plants)
1
2
1
2
C
1
2
A
B
1
1
2
3
3
4
E
H
I
3
4
5
1
4

2A
28
4
5
A
5
6
1
31
29
Battery
capacity
71
59
66
66
157
107
54
82
125
299
142
64
124
201
201
93
79
305
125
125
249
322
199
149
97
97
145
75
353
113
117
617
5,078
4,676
No. of
Status3 ovens
0
0
0
0
0
0
0
0
0
0
0
0
0
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0


24
20
30
30
57
20
40
23
35
60
40
18
35
55
55
47
41
72
30
30
60
56
35
60
38
40
58
30
78
25
29
70
1,341
1,231
Online
130

132

157
161

207

299
330


0

477


499


521

563




470


617
4,563
4,563
Hot
idle
0

0

0
0

0

0
0


0

0


0


0

0




113


0
113
113
Existing
Cold Under plant
idle construction total
0

0

0
0

0

0
0


402

0


0


0

0




0


0
402
0
0

0

0
0

0

0
0


0

0


0


0

0




0


0
0
0
130

132

157
161

207

299
330


402

477


499


521

563




583


617
5,078
• 4,676
                   Note:   Data current  as of November 1984.

                   aStatus:   0 = online;  1 = hot idle;  2  =  cold  idle; and 3 = under construction.

                   bUenotes  cold idle plants.

-------
                    TABLE A-3.  FURNACE COKE UY-PROOUCT PLANT OPERATIN3 PROCESSES
No.
1
2
3
4
5
fa
7
B
9
10
11
12
13
14
15
16
17

18
19
20

21
22
23

24
25
26
27
28
29
30
Tar Tar Tar
Plant decanter dewatering storage
LTV Steel, Thomas, AL 1
New Boston, Portsmouth, OH 1
Wheeling-Pitt, Monessen, PA 1
Lone Star Steel, Lone Star, TX 1
LTV Steel, So. Chicago, IL 1
National Steel, Granite 1
City, IL
Interlake, Chicago, IL 1
LTV Steel, Gadsden, AL 1
Rouge Steel, Dearborn, MI I
U.S. Steel, Fairless Hills, PA 1
LTV Steel, Warren, OH 1
LTV Steel, E. Chicago, IL 1
Armco Inc., Ashland, KY I
Weirton Steel, Browns Island, WV 1
U.S. Steel, Provo, UT
LTV Steel, Aliquippa, PA
Bethlehem Steel,
Lackawanna, NY
National Steel, Detroit, MI
U.S. Steel, Lorain, OH
Wheeling-Pitt,
E. Steubenville, WV
LTV Steel, Cleveland, OH
Armco Inc. , Middletown, OH
Uethlehem Steel, Burns
Harbor, IN
LTV Steel, Pittsburgh, PA
U.S. Steel, Fairfield, A'L
Bethlehem Steel, Bethlehem, PA
Bethlehem Stee) , Sparrows
Pt., MO
Inland Steel, E. Chicago, IN
U.S. Steel, Gary, IN
U.S. Steel, Clairton, PA
1
1
1

1
1
I

1
1
1

1
1
I
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
1

1
1
1

1
1
1

1
1
1
1
1
1
0
1
1
I
1
1
1
1
1
1
1
1
1
1
1
1
1
1

1
1
1

1
1
1

1
1
1
1
1
1
1
Excess-
ammonia
1 iquor
storage
1
1
1
1
1
1
1
1

1

1
1
1

1
1
1

1
1
1
1
1
1
1
Flushing-
liquor
cjrc. tank
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

1
1
1

1
1
1

1
1
1
1
1
1
1
Tar-
interc.
sump
1
1
1
1
1
1
1
1
1
1
1
- 1
1
1

1
1

1

1

1
1
1

1
1
1
1
1
1
1
Light-
oil
storage
1
1
1
1
1
1
1
1
1
1
1
1
1
i
i
1
I

.
1
1

.
.
g

1

1
1
1
1
1
BTX
storage
0
0
0
1
1
1
0
0
1
0
0
1
0

0
1

n
u
o

i
i
i
i
n
u
o
Q
1
0
Q
Q
0
Benzene
storage
0
0
0
0
1
0
0
0
0
0
0
0
0

0
(I
u


0





Q

Q
1
n
u
Q
1
Naphthalene processiny/nandl iny
Denver Naphth. Naphth.
flo. unit melt, nit HrU t*ntQ
1
0
0
1
U
1
1
1
1
1
0
0
0
1
u
0
1
1
-J
J
1
1
n
u




1
1
1
1
u


0
l
0
0
1
u
1
1
1
1
1
u
0
0
1
0
0
1
i






u

1
1
1
1
0

u
1
1
u
1
u
0
I
0
1
I
1
1
u
u
u
1
o
u


1
1
u

u
0
u



0

u
u
Total
           30
                    28
                               30
                                        30
                                                   30
                                                              30
                                                                       29
                                                                               10
                                                                                                   15
                                                                                                               Ib
                                                                                                                     (continued)

-------
                                                                      TABLE A-3.   (continued)
oo
Nn.
1
2
3
4
5
6

7
8
9
10
11
12
13
14
15
16
17

18
19
20

21
22
23

24
25
26
27

28
29
30

Direct-
water
final
Plant cooler
LTV Steel, Thomas, AL
New Boston, Portsmouth, OH
Wheeling-Pitt, Monessen, PA
Lone Star Steel, Lone Star, TX
LTV Steel, So. Chicago, IL
National Steel, Granite
City, IL
Interlake, Chicago, IL
LTV Steel, Gadsden, AL
Rouge Steel, Dearborn, MI
U.S. Steel, Fairless Hills, PA
LTV Steel, Warren, OH
LTV Steel, E. Chicago, IN
Armco Inc., Ashland, KY
Weirton Steel, Browns Island, WV
U.S. Steel, Provo, UT
LTV Steel, Aliquippa, PA
Bethlehem Steel,
Lackawanna, NY
National Steel, Detroit, MI
U.S. Steel, Lorain, OH
Wheeling-Pitt,
E. Steubenville, WV
LTV Steel, Cleveland, OH
Armco Inc. , Middletown, OH
Bethlehem Steel, Burns
Harbor, IN
LTV Steel, Pittsburgh, PA
U.S. Steel, Fairfield, AL
Bethlehem Steel, Bethlehem, PA
Bethlehem Steel, Sparrows
Pt., MD
Inland Steel, E. Chicago, IN
U.S. Steel, Gary, IN
U.S. Steel, Clairton, PA
Total
1
0
0
1
0
1

1
1
1
1
0
1
0
0
0
0
1

1
1
0

0
1
0

1
1
1
0

0
1
Q
16
Tar-
bottom
final
cooler
0
1
1
0
0
0

0
0
0
0
0
0
0
0
1
0
0

0
0
0

0
0
0

0
0
0
0

1
0
0
4
Wash-
oil Light-
final oil
cooler sump
0
0
Q
0
0
0

0
0
0
0
1
0
0
1
0
0
0

0
0
1

1
0
0

0
0
0
1

0
0
0
5
1
1
. 1
1
1
1

1
1
1
1
1
1
1
1
1
1
1

1
1
1

1
1
0

1
1
1
1

1
1
1
29
Light-
oil decanter/
condenser Wash-oil
vent decanter
1
1
1
1
1
1

1
1
1
1
1
1
1
1
1
1
1

1
1
1

1
1
0

1
1
1
1

1
1
1
29
1
1
1
1
1
1

1
1
1
1
1
1
1
1
1
1
1

1
1
1

1
1
0

1
1
1
1

1
1
1
29
Wash-oil
circ. tank
1
1
1
1
1
1

1
1
1
1
1
1
1
1
1
1
1

1
1
1

1
1
0

1
1
1
1

1
1
1
29
Equipment
leaks
1
1
1
1
1
1

1
1
1
1
1
1
1
1
1
1
1

1
1
1

1
1
0

1
1
1
1

1
1
1
29
                        Note:   Data  current as of  November  1984.

-------
                                               TABLE A-4.  FOUNDRY COKE BY-PRODUCT PLANT OPERATINS PROCESSES
 No.
10
11
12
13
14
            Plant
                                                                 ammonia  Flushing-
                                                                                                                     Naphthalene processing/h
                                                                                                                     "enverNaphth.—  N
                                                  -        •»>     ii\fuu
                                    decanter dewaterinq storage  stora
                                                                                                storage storage  storaqe
                                                                                                                                      melt pit  dry tanks
 Chattanooga  Coke,
   Chattanooga,  TN
 IN Gas,  Terre Haute,  IN
 Koppers,  Toledo, OH
 Empire Coke, Holt, AL
 Koppers,  Erie,  PA
 Tonawanda, Buffalo, NY
 Carondolet, St.  Louis, MO
 AL Byproducts,  Keystone, PA
 Citizens Gas,
   Indianapolis,  IN
Jim Walters, Birmingham, AL
Shenango, Pittsburgh,  PA
Koppers,  Woodward,  AL
AL Byproducts,  Tarrant, AL
Detroit Coke, Detroit, MI
                                                                                                                                                continued)

-------
                                                           TABLE A-4.  (continued)
3=>

t—>
O
No.
1

2
3
4
5
6
7
8
9

10
11
12
13
14

Wash-oil
final
Plant cooler
Chattanooga Coke,
Chattanooga, TN
IN Gas, Terre Haute, IN
Koppers, Toledo, OH
, Empire Coke, Holt, AL
Koppers, Erie, PA
Tonawanda, Buffalo, NY
Carondolet, St. Louis, MO
AL Byproducts, Keystone, PA
Citizens Gas,
Indianapolis, IN
Jim Walters, Birmingham, AL
Shenango, Pittsburgh, PA
Koppers, Woodward, AL
AL Byproducts, Tarrant, AL
Detroit Coke, Detroit, MI'
Total
0

0
0
0
0
0
0
0
0

0
0
0
0
0
0
Direct-
water
final
cooler
1

1
0
1
0
0
0
1
1

1
0
0
1
0
7
Tar-
bottom
final
cooler
0

0
1
0
0
0
0
0
0

0
0
1
0
0
2
Light-
oil
sump
1

1
0
1
0
0
0
1
0

1
1
1
1
0
8
Light-
oil decanter/
condenser Wash-oil
vent decanter
1

1
0
1
0
0
0
1
0

1
1
1
1
0
8
1

1
0
1
0
0
0
1
0

1
1
1
1
0
8
Wash-oil
circ. tank
1

1
0
1
0
0
0
1
0

1
1
1
1
0
8
Equipment
leaks
1

1
0
1
0
0
0
1
0

1
1
1
1
0
8
           Note:  Data current as  of  November 1984.

-------
TABLE A-5.   ESTIMATED NATIONWIDE  BASELINE  BENZENE  EMISSIONS
 FROM FURNACE AND FOUNDRY  COKE  BY-PRODUCT  RECOVERY PLANTS
Source
1. Direct-water final-cooler
cool iny tower
2. Tar-bottom final-cooler
cooling tower
3. Naphthalene processing/
handling
4. Light-oil decanter/
condenser vent
b. Tar-intercepting sump
6. Tar decanter
7. Tar dewatering
8. Tar storage
y. Light-oil sump
1U. Light-oil storage
11. BTX storage
12. Benzene storage
13. Flushing liquor
circulation tank
14. Excess-ammonia liquor
storage tank
lb. Wash-oil decanter
16. Wash-oil circulation tank
Footnotes at end of tables.

affected
plants3
16
4
Ib
29
30
30
28
30
29
29
10
4
30
30
29
29
Furnace plants
Capacity,15
1,000 Mg/yr
21,430
b,729
19,664
44,014
45,804
45,804
39,947
45,804
44,014
44,014
11,544
10.523
45,804
45,804
44,014
44,014
Nationwide
emissions,
Mg/yr
5,786
401
2,103
3,456
4,351
3,527
839
550
660
242
54
bl
412
412
154
154

No. of
affected
plants8
7
2
7
9
14
14
13
14
9
9
2
1
14
14
9
9
Foundry plants
Capacity, b
1,000 Mg/yr
2,384
720
2,384
3,290
6,078
5,078
4,917
5,078
3,290
3,290
901
402
5,078
b,078
3,290
3,290
Furnace and foundry
Nationwide
emissions,
Mg/yr
470
37
186
158
227
184
49
29
27
10
3
1
33
33
7
7
No. of
affected
plants*
23
6
22
38
44
44
41
44
38
38
12
b
44
44
3«
38
Capacity,15
1,000 Mg/yr
23,814
6,449
22,U3b
47.3U4
50,882
511,882
44,864
50,882
47,304
47.3U4
12,445
lu,92b
6U.B82
bu,882
47.JU4
47,304
plants
Nationwide
emissions,
My/yr
6,2bb
4J«
2,289
J.614
4,b78
3,711
687
b/8
687
ZbJ
b?
62
446
44b
161
Ibl
                                                                                              (continued^

-------
                                                                TABLE  A-5.   (continued)
Furnace plants
No. of
affected
Source plants3
17.
18.
19.
20.
21.
22.

Pump seals
Valves
Pressure-relief devices
Exhausters
Sampling connection systems
Open-ended lines
Total (rounded)
29
29
29
29
29
29

Capacity, b
1,000 Mg/yr
44,014
44,014
44,014
44,014
44,014
44,014

Nationwide No. of
emissions, affected
Mg/yr plants3
370 9
249 9
168 9
17 9
33 9
11 9
24,000
Foundry plants
Capacity, b
1,000 Mg/yr
3,290
3,290
3,290
3,290
3,290
3,290

Furnace and foundry plants
Nationwide No. of Nationwide
emissions, affected Capacity,5 emissions,
Mg/yr plants3 l.UOO My/yr My/yr
101
68
46
5
9
3
1,700
J8 47,304 471
3« 47,304 317
38 47.3U4 214
38 47,304 2i!
38 47,304 42
38 47,304 14
2b,7UU
Note:  Oata current as of November 1984.

aNumber of plants having this source (out of a total  of 30  furnace plants,  14  foundry  plants,  or  44  furnace  and  foundry  plants  combined;  includes  plants
 currently on cold idle.

Capacity of plants with this source.

-------
TABLE A-6.   ESTIMATED NATIONWIDE BASELINE VOCa EMISSIONS
FROM FURNACE AND FOUNDRY COKE BV-PRODUCT RECOVERY PLANTS
Furnace plants
TT: 	 ~ 	 • 	 •

i.
2.
3.
4.
b.
I
L 6.
co
7.
8.
9.
10.
11.
12.
13.
14.
lb.
16.
Font ri
Source
Direct-water final-cooler
(Doling tower
Tar-bottom final-cooler
cooling tower
Naphthalene processing/
handling
Light-oil decanter/
condenser vent
Tar intercepting sump
Tar decanter
Tar dewatering
Tar storage
Light-oil sump
Light-oil storage
BTX storage
Benzene storage
Flushing-liquor
circulation tank
Excess-ammonia liquor
storage tank
Wash-oil decanter
Wash oil circulation tank
ifttp*; Af pnd nf Fahlac
no. or
affected
plants'5
16
4
lb
29
30
30
28
30
29
29
10
4
30
30
29
29
Capacity,0
1,000 Mg/yr
21,430
5,729
19,6b4
44,014
45,«04
45,804
39,947
45,804
44,014
44,014
11,544
10,523
45,804
45,804
44,014
44.U14

Nationwide
emissions,
Mg/yr
90,842
6,302
3,302
4,031
9,252
7,512
19,654
12.S71
942
347
77
61
591
591
219
219

No. of
affected
plants0
7
2
7
9
14
14
13
14
9
9
2
1
14
14
9
9
Foundry plants
Capacity,0
1,000 Mg/yr
2,384
720
2,384
3,290
5,078
5,078
4,917
5,078
3,290
3,290
901
402
b,07«
5,078
3,290
3,290
Furnace and foundry
Nationwide
emissions,
Mg/yr
7,377
578
292
226
482
391
1,137
671
38
15
4
1
48
48
10
10
No. of
affected
plants'3
23
6
22
38
44
44
41
44
38
38
12
b
44
44
38
38
Capacity,0
1,000 Mg/yr
23,814
6,449
22,038
47,304
50,882
bO,882
44,864
bO,882
47.3U4
17,304
12,44b
10,92b
50,882
bU,882
4/.304
47,304
plants
Nationwide
emissions,
Mg/yr
98,219
6.88U
3,b94
5,lb/
9,73b
7,903
20,791
13,542
9811
362
82
62
639
639
229
229
                                                                                            (continued)

-------
                                                                 TABLE  A-6.   (continued)
Furnace plants


17.
18.
19.
20.
21.
22.


Source
Pump seals
Valves
Pressure-relief devices
Exhausters
Sampling connection systems
Open-ended lines
Total (rounded)
No. of
affected
plants'1
29
29
29
29
29
29

(
Capacity,0
1,000 Mg/yr
44,014
44,014
44,014
44,014
44,014
44,014

Nationwide
emissions,
Mg/yr
528
355
240
25
47
16
59,200
No. of
affected
plants'1
9
9
9
9
9
9

Foundry plants

Capacity,0
1,000 Mg/yr
3,290
3,290
3,290
3,290
3,290
3,290

Furnace and foundry plants
Nationwide
emissions,
Mg/yr
1S9
107
73
8
14
5
11,700
No. of
affected
plants'1
38
38
38
38
3«
38


Capacity,0
l.OOU Mg/yr
47,304
47,304
47,304
47,304
47,304
47,304

Nationwide
emissions,
My/yr
bb7
463
312
3J
bl
21
17U.900
Note:  Data current as of November 1984.

aBenzene and other VOC.

bNuraber of plants having this source out of a total  of  30  furnace plants,  14  foundry plants, or 44 furnace and foundry plants combined;
 includes plants currently on cold idle.

cCapacity of plants with this source.

-------
                                                         T/tBLE A-7.   CONTROL  OPTION IMPACTS FOR FURNACE PLANTS
3=-

H-"
in
Emission source
Final cooler
cooling tower and
naphthalene
processing/
handling
Tar decanter, tar-
intercepting
sump, and
flushing-liquor
circulation tank6
Tar storage tanks
and tar-dewatering
tanks6
Light-oil
condenser, light-
oil decanter,
wash-oil decanter
and wash-oil
circulation tanks
Excess-ammonia
liquor storage
tank6
Light-oil storage
tanks and BTX
storage tanks6
Benzene storage
tanks
Light-oil sump

Pumps
Control option/
efficiency, %
Baseline0:
Tar-bottom final cooler 81
Wash-oil final cooler 100
Baseline0:
Gas blanketing 98d


Baseline0:
Gas blanketing 98
Baseline0:
Wash-oil scrubber 90
Gas blanketing 98


Baseline0:
Gas blanketing 98
Baseline0:
Gas blanketing 98
Baseline0:
Wash-oil scrubber 90
Nj gas blanketing 98
Baseline0:
Cover 98
Baseline0:
Quarterly inspections 71
Monthly inspections 83
Dual mechanical seals 100
Footnotes at end of table. " " "™ "
Controlled Controlled
emissions, incidence,
**9/yr, lives/yr
Benzene/VOC^
8,290
1,900
0
8,290
270


1,390
30
3,760
380
80


410
8
300
6
60
6
1
660
10
370
110
60
0
100,000
29,400
0
17,400
600


32 ,500
600
5,370
550
120


590
11
420
9
60
6
1
940
20
530
150
90
0
0.78
0.16
0.00
0.99
0.02


0.16
0.003
0.39
O.*040
0.009


0.047
0.001
0.039
0.001
0.0063
0.0006
0.0001
0.079
0.001
0.044
0.013
0.008
0.000
Annual
costs,'5
1984 $/yr
152,600
12,871,380
(1,062,260)


1,735,070

348,320
504,530



849,370

1,119,040

120,570
134,040

468,470

29,800
35,780
1,067,250
Benzene cost VOCa cost
effectiveness, b.f effectiveness
1984 $/Hq 1984 I/Mo
aver./incre.
20 20
1,550 6,690
(130) (130)


1,280 1,280

100 100
140 520



2,100 2,100

3,860 3,860

2,200 2,200
2,240 2,760

720 720

110 110
120 130
2,890 16,690
aver./incre.
2 2
130 430
(60) (60)


50 50

70 70
100 360



1,470 1.470

2,700 2,700

2,200 2,200
2,240 2,760

500 500

80 80
81 90
2,020 11,780

-------
                                                                     TABLE A-7.   (continued)
CT>
Emission source
Valves



Exhausters




Pressure-relief
devices

Sampling
connection systems
Open-ended lines

Naphthalene
processing/
handlings
Control option/
efficiency, %
Baseline0:
Quarterly inspections
Monthly inspections
Sealed-bellows valves
Basel inec:
Quarterly inspections
Monthly inspections
Degassing reservoir
vents
Basel inec:
Quarterly inspections
Monthly inspections
Rupture disc system
Baseline0:
Closed-purge sampling
Baseline0:
Cap or plug
Baseline0:
Mixer-settler

Control led
emissions,
Mg/yr
Benzene/VOCa

63
73
100

55
64
100


44
52
100

100

100

100

250
90
70
0
20
8
6
0

170
90
80
0
30
0
10
0
2.100
0

350
130
100
0
30
10
9
0

240
130
no
0
50
0
20
0
3,300
0

dont'rolled'
incidence, Annual
lives/yr costs, &
1984 $/yr
0.030
0.011
0.009
0.000
0.0021
0.0011
0.0008
0.0000

0.020
0.011
0.009
0.000
0.0039
0.0000
0.0013
0.0000
0.29
0.00


(36,760)
(19,760)
4,324,320

13,870
30,160
437,320


(30,940)
(26,990)
153,820

40,770

7,110

1,603,960

Benzene cost VOCa cost
effectiveness,11.* effectiveness,^
1984 $/Mg 1984 $/My
aver./incre.

(240)
(110)
17,380

1,450
2,740
25,130


(410)
(310)
920

1,250

640

760


(240)
720
62,780

1,450
11,230
68,820


(410)
290
2,270

1,250

640

760

aver./incre.

(160)
(80)
12,170

1,020
1,920
17,590


(290)
(210)
640

870

450

490


(160)
500
43,950

1,020
7,870
44,640


(290)
200
1,590

870

450

490

              Note:  Data current as of November 1984.

              aVOC estimates include benzene.

              ^Parentheses denote cost savings.

              cBaseline numbers represent relatively uncontrolled levels.

              d95 % efficiency for tar decanter.

              ^ash-oil scrubbers are more costly and less effective than gas blanketing for these sources.

              ^''Average" means compared to baseline; "incremental" means compared to the next less stringent control option.

              9The mixer-settler control option for naphthalene processing and handling is shown
               separately to address a comment on new indirect cooling technology that would
               not necessarily control naphthalene processing emissions.

-------
                                                                   TABLE A-8.   CONTROL OPTION IMPACTS FOR FOUNDRY PLANTS
 I
I—"
^•g
Emission source
Final cooler
cooling tower
and naphthalene
processing/
handling
Tar decanter, tar-
intercepting
sump, and
flushing-liquor
circulation
tank6
Tar storage
tanks and tar-
dewatering tanks6
Light-oil
condenser, light-
oil decanter,
wash-oil
decanter
and wash-oil
circulation
tanks6
Excess-ammonia
liquor storage
tank6
Liyht-oil
storage tanks
and BTX storage
tanks6
Benzene storage
tanks6
Light-oil sump

Pumps



Control option/
efficiency, %
Baseline0:
Tar-bottom final cooler 81
Wash-oil final cooler 100


Baseline"-:
Gas blanketing 93d




Baseline0:
Gas blanketing 98

Baseline0:
Gas blanketing 98






Baseline0:
Gas blanketing 98

Baseline0:
Gas blanketing 98


Baseline0:
«2 gas blanketing 98
Baseline0:
Cover 98
Baseline0:
quarterly inspections 71
Monthly inspections 83
Dual mechanical seals 100
Controlled
emissions,
Mg/yr
Controlled
Incidence,
Hves/yr
Benzene/VOCa
690
160
0


440
10




80
2

170
3






30
0.7

10
0.3


1.0
U.I
30
0.5
100
30
20
0
8,250
2,490
0


920
30




1.810
40

250
5






50
1

20
0.4


1.0
0.1
40
0.8
160
50
30
0
0.077
0.012
0.000


0.070
0.003




0.012
0.001

0.023
0.0005






0.0049
0.0001

0.0021
0.00004


0.00041
0.00001
0.004
0.000
0.019
0.006
0.003
0.000
Benzene cost VOCa cost
Annual effectiveness, ">f effectiveness,0^
costs, b 1984 $/Mg 1984 $/Mg
1984 $/yr Aver./incre. Aver./1ncre.

305,550 570 570 50 50
2,831,680 4,090 15,900 340 1,000



410,390 960 960 460 460





397,510 5,260 5,260 220 220


238,200 1,420 1,420 990 990







277,630 8,490 8.490 5,920 5,920


251,140 19,520 19,500 13,640 13,640



14,360 11,640 11,640 11,640 11,640

44,290 1,700 1,700 1,190 1,190

9,360 130 130 80 80
11,270 134 150 90 100
323,960 3,200 18,500 2,030 11,730
                                                                                                                                                (continued)

-------
                                                                             TABLE  A-8.   (continued)
3=>
 i
i—"
CO
Emission source
Valves



Exhausters




Pressure-relief
devices


Sampling
connection
systems
Open-ended lines

Naphthalene
processing/
handlings
Control option/
efficiency, %
Baseline0:
Quarterly inspections
Monthly inspections
Sealed-bellows valves
Baseline0:
Quarterly inspections
Monthly inspections
Degassing reservoir
vents
Baseline0:
Quarterly inspections
Monthly inspections
Rupture disc system
Baseline0:
Closed-purge sampling

Baseline0:
Cap or plug
Baseline0:
Mixer-settler

Controlled
emissions,
Mg/yr
Controlled
incidence,
lives/yr
Benzene/VOCd

63
73
100

55
64
100


44
52
100

100


100

100

70
30
20
0
4.9
2.2
1.8
0.0

50
30
20
0
9
0

3
0
190
0

110
40
30
0
7.7
3.4
2.8
0.0

70
40
30
0
14
0

5
0
290
0

0.013
0.005
0.004
0.000
0.00091
0.00040
0.00030
0.00000

0.0085
0.0047
0.0040
0.0000
0.0017
0.0000

0.00056
0.00000
0.034
0.000

Annual
costs, b
1984 $/yr

(10
(5
1,310

4
9
135


(9
(8
47

12


2

453


,920)
,740)
,400

,310
,360
,720


,320)
,110)
,010

,380


,220

,080

Benzene cost
effectiveness, b>
1984 $/Mg
Aver.

(260)
(120)
19,250

1,590
3,010
27,640


(450)
(330)
1,020

1,390


730

2,430

/incre.

(260)
800
69,530

1,590
12,330
70,200


(450)
320
2,520

1,390


730

2,430

VOCa cost
' effectiveness, b«f
1984 $/Mg
Aver,

(160)
(70)
12,220

1,010
1,910
17,540


(290)
(210)
650

880


460

1,550

./incre.

(160)
510
44,170

1,010
7,900
44,490


(290)
200
1,600

880


460

1,550

                       Note:  Oata current as of November 1984.

                       a VOC estimates include benzene.

                       b Parentheses denote cost savings.

                       0 Baseline numbers represent relatively  uncontrolled  levels.

                       d 95% efficiency for tar decanter.

                       e Wash-oil scrubbers are more costly and less  effective  than  gas  blanketing  for  these  sources.

                       f "Average" means compared to baseline;  "incremental" means compared  to  the  next less  stringent  control option.

                       9 The mixer-settler control option for naphthalene  processing and handling is  shown
                         separately to address a comment on new indirect cooling  technology  that would
                         not necessarily control naphthalene processing emissions.

-------
                                                      TABLE A-9.  EFFECT OF CONTROL OPTIONS ON REDUCING BENZENE EMISSIONS AT FURNACE

                                                                        AND FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS
 I
I—»
•.o
Furnace Plants

1.
2.
3.
4.
5.
6.
7.
8.
y.
10.
11.

12.
13.
14.
15.
Source
All sources
Final-cooler
cooling tower
Light-oil decanter/
condenser vent
Tar-Intercepting
sump
Tar decanter
Tar dewaterlng
Tar storage
Light-oil sump
Light-oil storage
BTX storage
Benzene storage

Flushing-1 iquor
circulation tank
Excess-ammonia-
liquor storage
Wash-oil decanter
Mash-oil
circulation tank
Control option
No national emission standard
1. Tar-bottom final cooler
2. Wash-oil final cooler
1. Wash-oil scrubber
2. Coke-oven gas blanketing
Coke-oven gas blanketing
Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
1. Coke-oven gas blanketing
Sealed cover
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Gas blanketing
Coke-oven gas blanketing
Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
No. of
affected
plants3
30
16
20
29
29
30
30
28
28
30
30
29
29
29
10
10
4
4
30
30
29
29
29
29
national
benzene
emissions,
Mg/yr
24,000
8,290
8,290
3,456
3,456
4,351
3,527
839
839
550
550
660
242
242
54
54
61
61
412
412
154
154
154
154
Controlled
benzene
emissions,
Mg/yr
24,000
1,901
0
356
78
87
176
84
17
55
11
13
24
5
7
1
6
1
8
8
15
3
15
3
Foundry Plants
National Controlled
No. of benzene benzene
affected emissions, emissions,
plants3 Mg/yr Mg/yr
14
7
9
9
9
14
14
13
13
14
14
9
9
9
2
2
1
1
14
14
9
9
9
9
1,690
693
693
158
158
227
184
49
49
29
29
27
10
10
3
3
1
1
33
33
7
7
7
7
1,690
159
0
16
3
5
9
5
1
3
0.6
0.5
1
0.2
0.3
0.1

0.03
0.7
0.7
6.1
0.1
6.1
0.1
Furnace and Foundrv Plant-;
No. of
affected
plants3
44
23
29
38
38
44
44
41
41
44
44
38
38
38
12
12

5
44
44
38
38
38
38
National
benzene
emissions,
Mg/yr
25,900
8,983
8,983
3,614
3,614
4,578
3,711
887
887
578
578
687
253
253
57
57

62
446
446
161
161
161
161
Controlled
benzene
emissions,
Mg/yr
25,900
2,060
0
372
82
92
186
89
18
58
12
14
26
5
7
1

6
1
9
9
21
3
21
3
Footnote at end of table. ~ 	
                                                                                                                                                               (continued)

-------
                                                                                  TABLE A-9.  (continued)
 I
rv»
o
Furnace Plants

16.


17.


18.


19.


20.

21.

Source
Pump seals


Valves


Pressure- relief
devices

Exhausters


Sampling connection
systems
Open-ended lines


1.
2.
3.
1.
2.
3.
1.
2.
3.
1.
2.
3.
Control option
Quarterly inspection
Monthly inspection
Dual mechanical seals
Quarterly inspection
Monthly inspection
Sealed-bellows valves
Quarterly inspection
Monthly inspection
Rupture disc
Quarterly inspection
Monthly inspection
Rupture disc
Closed purge sampling

Cap


or plug
Total (rounded)
No. of
affected
plants9
29
29
29
29
29
29
29
29
29
29
29
29
29

29

National
benzene
emissions,
Mg/yr
370
370
370
249
249
249
168
168
168
17
17
17
33

a
24,000
Controlled
benzene
emissions,
Mg/yr
108
62
0
93
69
0
93
80
0
8
6
0
0

0

Foundry Plants
No. of
affected
plants3
9
9
9
9
9
9
9
9
9
9
9
9
9

9

National
benzene
emissions,
Mg/yr
101
101
101
68
68
68
46
46
46
5
5
6
9

3
1,700
Control led
benzene
emissions,
Mg/yr
29
17
0
25
19
0
26
22
0
2
2
0
0

0

Furnace and Foundry Plants
No. of
affected
plants4
38
38
38
38
38
38
38
38
38
38
38
38
38

38

National
benzene
emissions,
Mg/yr
471
471
471
317
317
317
214
214
214
22
22
22
42

14
2b,700
Control led
benzene
emissions,
Mg/yr
1J7
79
0
118
8b
U
119
1U1
U
10
8
U
0

U

                Note:  Data current as of  November  1984.


                aNuraber of plants  having this  source  out of a total of 30 furnace plants, 14 foundry plants, or 44 furnace and foundry plants combined;
                includes plants currently  on  cold  idle.

-------
                                               TABLE  A-10.   EFFECT  OF  BENZENE CONTROL OPTIONS ON REDUCING V0a
                                              EMISSIONS AT  FURNACE  AND FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS
Furnace plants

1.
2.

3.

4.

5.
6.

7.

8.
9.

10.

11.

12.

13.


14.

IS.

Source
All sources
Final-cooler
cooling tower
Light-oil decanter/
condenser vent
Tar-intercepting
sump
Tar decanter
Tar dewatering

Tar storage

Light-oil sump
Light-oil storage

BTX storage

Benzene storage

Flushing-liquor
circulation tank
Excess-ammonia
liquor storage
tank
Wash-oil decanter

Wash-oil
circulation tank
Control option
No national emission standard
1. Tar-bottom final cooler
2. Wash-oil final cooler
1. Wash-oil scrubber
2. Coke-oven gas blanketing
Coke-oven gas blanketing

Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
Sealed cover
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Gas blanketing
Coke-oven gas blanketing

1. Wash-oil scrubber
2. Coke-oven gas blanketing

1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
No. of
affected
plants
30
16
20
29
29
30

30
28
28
30
30
29
29
29
10
10
4
4
30

30
30

29
29
29
29
VOC
nationwide
emissions,
Mg/yr
159,200
100,446
100,446
4,931
4,931
9,252

7,512
19,654
19,654
12,871
12,871
942
347
347
77
77
61
61
591

591
591

219
219
219
219
Controlled
VOC
emissions,
Mg/yr
159,200
29,400
0
493
112
185

376
1,965
393
1,287
257
19
35
7
10
2
6
1
12

59
12

22
5
22
5
Foundry plants
No. of
affected
plants
14
7
9
9
9
14

14
13
13
14
14
9
9
9
2
2
1
1
14

14
14

9
9
9
9
VOC
nationwide
emissions,
Mg/yr
11,700
8,248
8,248
226
226
482

391
1,137
1,137
671
671
38
15
15
4
4
1
1
48

48
48

10
10
10
10
Controlled
VOC
emissions,
Mg/yr
11,700
2,943

23
5
10

20
114
23
67
13.4
0.8
1
0.3
0.4
0.1
0.1
0.03
1

5
1

1
0.2
1
0.2
Furnace and foundry plants
VOC Controlled
No. of nationwide VOC
affected emissions, emissions,
plants Mg/yr Mg/yr
44
23
29
38
38
44

44
41
41
44
44
38
38
38
12
12
5
5
44

44
44

38
38
38
38
170,900
108,694
108,694
5,157
5,157
9,735

7,903
20,791
20,791
13,542
13,542
980
362
362
82
82
62
62
639

639
639

229
229
229
229
170,900
32,343
0
516
116
195

395
2,079
416
1,354
271
20
36
7
10
2
6
1
13

64
13

23
5
23
5
Footnotes at end of table.
                                                                                                                                                (continued)

-------
                                                                           TABLE A-10.  (continued)
3=-

r>o
Furnace plants



16.


17.


18.


19.




Source
Pump seals


Valves


Pressure relief
devices

Exhausters





1.
2.
3.
1.
2.
3.
1.
2.
3.
1.
2.
3.


Control option
Quarterly inspection
Monthly inspection
Dual mechanical seals
Quarterly inspection
Monthly inspection
Sealed-bellows valves
Quarterly inspection
Monthly inspection
Rupture disc
Quarterly inspection
Monthly inspection
Rupture disc

(to. of
affected
plants
29
29
29
29
29
29
29
29
29
29
29
29
VOC
nationwide
emissions,
Mg/yr
528
528
528
355
355
355
240
240
240
25
25
25
Controlled
VOC
emissions,
Mg/yr
154
88
0
132
99
0
133
114
0
11
9
0
Foundry plants

tto. of
affected
plants
9
9
9
9
9
9
9
9
9
9
9
9
TGC
nationwide
emissions,
Mg/yr
159
159
159
107
107
107
73
73
73
8
8
8
Control led
VOC
emissions,
Mg/yr
46
27
0
40
30
0
40
34
0
3
3
0
Furnace

to. of
affected
plants
38
38
38
38
38
38
38
38
38
38
38
38
and foundry plants
• voc 	
nationwide
emissions,
Mg/yr
b87
687
687
463
463
463
312
312
312
33
33
33
Control led
VOC
emissions,
My/yr
2UO
115
0
172
129
0
173
148
0
15
12
0
         20.  Sampling connection  Closed purge sampling
               systems


         21.  Open-ended lines     Cap or plug
29
29
           47
           16
                                              14
                                                                     38
                                                                                                                                         38
                                                                                61
                                                                                                                                                    21
         Note:   Data current as of November 1984.


         aBenzene and other VOC.


         bNumber of plants having this source out of a total  of 30 furnace plants,  14  foundry plants,  or 44 furnace and foundry plants combined-
          includes plants currently on cold idle.                                                                                              '

-------
J=

to
                                      TABLE A-ll.   ENERGY  USE  AT MODEL BY-PRODUCT PLANTSa
                               User
Steam,
Mg/yr
                                                              Furnace  plants'5
 Gas  blanketing

   Tar  decanter,  tar-intercepting sump,      350
     and  flushing liquor circulation tank

   Tar  dewatering, tar storage               440

   Excess ammonia-liquor storage tank        126

   Condenser, light-oil decanter, wash-oil   174
    decanter, and circulation tank


Wash-oil scrubber

  Excess ammonia liquor storage tank         24

  Benzene storage tank


Final cooler

  Tar-bottom final cooler

  Wash-oil  final  cooler
                                                         380

                                                       1.210
j| For information on derivation
D 4,000 Mg coke/day.
c 1,000 Mg coke/day.
                                                         Electricity,
                                                          MWh/yr
                                                                          0.4

                                                                          0.9
                                                            98

                                                         1,330
                                                                          steam,
                                                                          Mg/yr
                                                                               Foundry Plants0
             Electricity,
               MWh/yr
127

303
                                               26

                                              333
                                                estimates,  see  Docket  A-79-16,  item IV-B-11

-------
             TABLE A-12.  EMISSIONS OF COKE-OVEN GAS FROM SELECTED
            FURNACE AND FOUNDRY COKE-OVEN BY-PRODUCT PLANT SOURCES
                           Furnace plant emissions,     Foundry plant emissions,
      Source                * gas/min/Mg coke/day        t gas/min/Mg coke/day
Tar decanter                        10.0                          7.5

Light-oil condenser                  0.18                         0.14

Tar dehydrator                       2.9                          2.2

Tar storage                          2.8                          2.1
                                     A-24

-------
                                         TABLE A-13.  YIELDS—FOUNORy VS. FURNACE COKE PLANTS
Year
1976
1977
1978
1979
1980
1981
1982
i
ro 1983
en
Coal-to-coke ratio
Merch. a Furn.
1.35 1.46

1.31 1.47
1.34 1.47
1.34 1.46

1.32 1.47

1.29 1.46
Average 1.325 1.465
ratios--
merch./furn.
™i5f™ Ifyr« ?' J? y f ' i , Gas yield, Light oil /gas cone. , Light oil/tar cone. .
gal/ton of coal gal/ton of coal 1,000 ft3/ton coal gal light oil/1.000 ft3 gal/gal
Merch. Furn. Merch. Furn. Merch. Furn. Merch Furii Merch Fi.m
l'61 2-58 5.26 7.77 9.21 11.02 0.18 0.23 0.32

!-77 2-9 5.52 7.78 9.23 11.2 0.19 0.26 0.32
1.82 2.51 5.94 7.86 9.03 11.04 0.2 0.23 0.31
1.82 2.67 5.97 8.27 8.94 11.14 0.2 0.24 0.3



8.08 10.37
0.33

0.37
0.32
(J.J2




1.77 2.665 5.6725 7.952 9.1025 10.954 0.1925 0.24 0.3125 0.335
•664 -713 .831 0.802 0.933
Merchant coke plants are assumed to be  the  same as  foundry coke plants.

-------
            TABLE A-14.  CORRECTION FACTOR COMPUTATION FOR FOUNDRY
                        COKE BY-PRODUCT RECOVERY PLANTS
Source
 Concentration
   adjustment
Volume  (throughput)
    adjustment
  Total
correction
Light oil plant  Benzene in
                   light oil
               (63.5/70)=0.907
                      Light oil yield
                        coke basis
                      x(0.664)(l.325/1.465)
                              0.54
Water contact
  with coke
  oven gas
Benzene in light oil
Light oil in
  coke oven gas
 (0.907x0.802)        x
                                                                   0.73
Tar sources
Benzene in light oil
Light oil  in coke
  oven gas
 (0.907x0.802)
  Tar yield
  coke basis

x(0.713)(l.325/1.465)  =
                                                                   0.47
Equip, leaks
Benzene in light oil
    (0.907)
                                                                   0.91
                                    A-26

-------
               TABLE A-15.   UNCONTROLLED  BENZENE  EMISSIONS FACTORS
                 FOR FURNACE AND  FOUNDRY  COKE  BY-PRODUCT PLANTS
Source
Cooling tower
Direct-water
Tar-bottom
Naphthalene separation
and processing
Light-oil condenser vent
Tar intercepting sump
Tar dewatering
Tar decanter
Tar storage
Light-oil sump
Light-oil storage
BTX storage
Benzene storage
Flushing-! iquor
circulation tank
Excess-ammoni a
liquor tank
Wash-oil decanter
Wash-oil circulation tank
Pump seals
Valves
Pressure-relief devices
Exhausters
Sample connections
Open-ended lines
Furnace plant
emission factors
g benzene/Mg coke

270
70
107

89
90
21
77
12
15
5.8
5.8
5.8
9

9

3.8
3.8
a
a
a
a
a
a
Foundry Plant
emission factors
g benzene/Mg coke

197
51
79

48
45
9.9
36
5.6
8.1
3.1
3.1
3.1
6.6

6.6

2.1
2.1
a
a
a
a
a
a
Emission factors are not  related  to  coke  production capacity and are listed in
 Table A-16.
                                  A-27

-------
                                      TABLE A-16.  BENZENE EMISSION FACTORS FOR EQUIPMENT LEAKS
ro
00

Valves
Pumps
Exhausters
Pressure relief
devices
Sampling
connections
Open-ended lines
Percent of
sources
leaking
initially
11
24
35
d
d
d
VOC emission
factor,
kg/source
day
0.26
2.7
1.2
3.9
0.36
0.055
Furnace plant
Benzene emission factors,
kg benzene/source day
Plant Aa
0.18
1.9
0.28C
2.7
0.25
0.038
Plant B&
0.22
2.3
0.28C
3.4
0.31
0.047
Foundry plant
benzene emission factors
kg benzene/source day
Plant A*
0.16
1.7
0.25
2.5
0.23
0.035
Plant Bb
0.2U
2.1
0.25
3.1
0.28
0.043
          aPlant A recovers light oil.  The amount of benzene in the light oil is assumed to be 70 percent at furnace
           plants and 63.5 percent at foundry plants.
                 B recovers refined benzene.  The amount of benzene averaged over the light oil and refined benzene
           is assumed to be 86 percent at furnace plants and 78 percent at foundry plants.

          C23.5% benzene in nonmethane hydrocarbon.

          dThis type of information would not be appropriate for relief valve
           overpressure, sampling connections, and open-ended lines.

-------
    Appendix B
Cost Impact Analysis

-------
                    APPENDIX  B:  COST  IMPACT ANALYSIS

  B.I   DEVELOPMENT  OF  REVISED CONTROL  COST ESTIMATES
       In  response  to  comments received during the public comment period
  for the  proposed  coke-oven by-product plant National Emission Standards
  for Hazardous Air  Pollutants (NESHAP), an in-depth review of the benzene
  control  cost estimates was undertaken.  The nature of the comments
  received touched virtually all  aspects of the cost-estimating metho-
  dology.  This report [Appendix B] documents the various elements of the
  review process, the revisions made to the cost-estimating methodology,
  and the  resulting changes in nationwide  cost  impacts.
      The general theme of the comments received was  that control  costs
  for the industry had been underestimated,  and  therefore the  cost  and
 economic impacts were understated.   The  American Iron  and  Steel  Institute
  (AISI) supplied cost factors and  cost estimates for  particular  portions
 of the proposed control  systems on  the basis of data supplied by member
 companies.   The information  supplied  ranged from cost  factors for indivi-
 dual  components  of the  control systems to a cost  estimate  for the pro-
 posed  controls  applied  to  an entire plant.  Because the  differences
 between the  cost information  received  and U.S.  Environmental Protection
 Agency (EPA)  cost  estimates was large  in some cases, a  complete review
 was undertaken.
     To begin the  review,  additional information supporting the cost data
 provided  by AISI was  requested.  At least three member companies had
 contributed specific cost  data.  However, this step alone was not
 expected  to provide sufficient explanation for all the  differences
 between EPA and  commenters1 cost  data.  Simultaneously, plans were
 developed to have a firm not  previously involved in the cost-estimating
efforts prepare a set of unit cost factors for gas-blanketing systems.
These new factors then could  be  compared  with  those  used by EPA  and  those
                                 B-3

-------
supplied by the commenters.   To have  the  unit  cost  factors  reflect
conditions imposed by a real  plant  situation,  Bethlehem  Steel  was asked
to allow EPA and/or their contractors to  visit the  Bethlehem,  Pennsyl-
vania, plant and use the conditions existing there  as  the basis  for  unit
costs.  A second purpose of  the visit to  the Bethlehem plant was to
generate a cost estimate for applying positive-pressure  gas blanketing to
all sources covered by the proposed regulation for  comparison  with the
Bethlehem plant cost estimate contained in  the AISI  comments.
     The CRS Si mine was retained to  develop the  unit  cost  factors and
overall cost estimate for the specific case of the  Bethlehem plant (thus
avoiding any potential conflict of  interest).   The  company  was selected
because its staff has not done much work  with  steel  plants  and parti-
cularly coke-oven facilities.  However, they have considerable experience
with petroleum refinery and  petrochemical  plant engineering and cost
estimating (these plants handle similar materials,  e.g., oils, tars,
explosive mixtures).  The EPA, through Research Triangle Institute (RTI),
supplied copies of the proposed regulation, the background  information
document (BID) for the proposed regulation, and estimates of the gaseous
emission rates from the various by-product  plant  sources to CRS Sirrine.
Bethlehem Steel provided plant drawings,  estimated  pumping  rates, and
processed vessel size information to  EPA  for CRS  Sirrine as requested.
Bethlehem Steel also provided information on in-plant  restrictions on
welding and safety matters.
     The direction given to CRS Sirrine was to develop cost estimates  for
safe positive-pressure gas-blanketing controls applied to the  various
groups of sources within the Bethlehem plant.   In doing  this,  they were
instructed to make use of existing  connection  points and existing support
structures for piping, where possible, and to  use pipe routings, tank
roofing, and vessel closure methods that  would tend to minimize costs.
They were also informed of the nature of  the  compounds in the  blanketing
gas and the vapor space over the process  liquids, as well as  the need  to
avoid  condensation and subsequent  plugging that might  result  in the
gas-blanketing pipelines.
     The plant visit  for cost estimation  purposes was  made  on  February 19
and 20,  1985.  Representatives from RTI,  CRS  Sirrine,  Bethlehem Steel,
                                 B-4

-------
 and United Engineers (the engineering firm supplying the cost estimate
 contained in AISI's comments) were present at the Bethlehem plant during
 this effort.  The CRS Sirn'ne developed the required cost estimates and
 gave the results to EPA and RTI in the form of a report titled Benzene
 Emissions Control Estimate (Docket Item IV-J-8).  The report provided an
 overall cost estimate for positive-pressure gas-blanketing systems
 applied to six groups of sources in the plant, as well  as light-oil  sump
 covers and roof installation for tar dewatering tanks and an excess
 ammonia-liquor storage  tank.   The wide range  of pipe sizes,  valves,  and
 other piping hardware used in the control  systems estimate provided  the
 desired unit cost information.
      National  Steel, Armco,  and Bethlehem  Steel  (including United
 Engineers)  contributed  additional  background  information  in  response  to
 EPA's  request  for more  details  for the cost comment  evaluation.
 Ultimately,  much  of  the  industry-contributed  data were  used  in the
 development  of  the  revised cost estimates.
 B.2 COMPARISON OF UNIT  COST  FACTORS
     One  of  the major findings  from the Bethlehem plant cost  study was
 that,  in  general, the unit cost  factors for piping and  piping hardware
 should  be increased.  The  unit  cost factors for  piping  and piping hard-
 ware developed for CRS Sirrine's estimate were higher than those used in
 the BID for the proposed standards, and they were more in the range of
 the unit  cost factors contained in the comments  received.  A principal
 factor  contributing  to the difference was labor cost for installation of
 the piping.  The  revised factors are presented later in this report.
 They are basically a composite of the Si mine  and industry-supplied
 data.
     The use of the Bethlehem plant for the cost study provided a basis
 for estimating the costs resulting from roof additions to tanks not
 currently covered.  There was no provision  for this cost element  in the
 proposal BID estimates;  this cost has been  added to the  control  cost
estimates for tar dewatering and excess ammonia-liquor storage tanks.
Another result  from the  Bethlehem cost study was the  addition of  pipe
supports for the minimum gas-blanketing cost cases.  The proposal  BID
estimates included pipe  supports only  in the maximum  cost  cases.   Pipe
                                 B-5

-------
support costs also were  added  to  the  wash-oil  scrubbers based on data
provided In Industry comments.  The operating  labor  costs  for wash-oil
scrubbers were Increased as  a  result  of  the  higher hourly  labor rates
developed during the review.   Costs for  sealing  all  process vessels and
Installing pressure/vacuum relief valves also  were added for both
gas-blanketing and wash-oil  scrubber  cases.
     A review of the United  Engineers' control cost  estimates for  final -
cooler cooling towers (prepared  for Bethlehem  Steel  and submitted  in the
AISI comments) suggested some  appropriate revisions  to the EPA cost
estimates for those controls.  The proposal  BID  estimate for the tar-
bottom mixer-settler included  no  allowance  for piping to and from  the  new
equipment.  In some plants this  may be a significant cost  if the new
equipment cannot be located  immediately  adjacent to  the existing final
cooler equipment.  Piping costs  were  added  for the  revised cost esti-
mates.  After considering the  number  of  pumps  and vessels  required in  the
tar bottom mixer-settler installation,  source  operating labor also was
added to the annualized costs.
     The United Engineers' cost  estimate for wash-oil  final coolers
indicated that some use could  be  made of existing direct-water  final-
cooler equipment in the conversion to a  wash-oil cooler scheme.  The
capital costs estimated by United Engineers for  Bethlehem  Steel's
Bethlehem plant were significantly lower than  the proposal BID  cost
estimates for a wash-oil final cooler installation.   In their  comments on
the proposed  regulation, AISI  said one  of the  member companies  had a  1981
budgetary cost estimate for a  wash-oil  final cooler of $7.4 million  (1984
dollars) installed in a 2,51)0 Mg/day  plant, much higher than  EPA esti-
mates.  The  EPA  requested further details about  this cost  estimate.
Rather  than  supplying the requested information, the company  supplied
newer  costs  estimates for other types of final coolers.   One  estimate was
for a  tar-bottom final  cooler, and the  other two estimates were for
indirect  cooling schemes  that would achieve final cooler  and  naphthalene
processing emission  reductions equivalent to wash-oil  final  coolers.
      One  indirect  scheme  uses warm wash-oil absorption of naphthalene
upstream  of  the  final cooler with final  cooling  provided  by direct water
contact with gas.   However, the  direct  contact water is cooled indirectly
                                 B-6

-------
 in a heat exchanger thereby avoiding atmospheric  emissions  from  the
 direct contact water.   The capital  cost  for  this  system  was  estimated  at
 $1.5 million for a 1,600 Mg/day  plant.
      The other indirect scheme uses a cross-tube  cooler  to  cool  the gas.
 Tar is injected into the gas-side  of the  cooler to  keep  condensed  naph-
 thalene in solution.  The condensed water and tar-containing naphthalene
 are returned to the collecting main.   The cooling fluid  is water flowing
 through tubes of the heat exchanger,  never coming into direct contact
 with the coke oven gas.   The  capital  cost of this system was estimated at
 $2.0 million for a 1,600 Mg/day  plant.
      According to Dravo/Still, an equipment  vendor  and design firm, the
 latter indirect cooling  scheme is in  use  at Dofasco in Hamilton, Ontario.
 Warm wash-oil  absorption of naphthalene is in use at Armco's Middletown
 plant;  however, the  direct water used to  cool the gas at Armco is cooled
 in an atmospheric cooling tower  rather being cooled indirectly.
      The cost  of the warm wash-oil  absorption with indirect cooling of
 the cooling  water is about the same  (scaled to equivalent plant  size)  as
 the United Engineers' estimate for  retrofitting a wash-oil  final  cooler
 to the  Bethlehem plant.   The cost estimate for indirect cooling  with the
 cross-tube cooler and naphthalene absorption by tar is about 25  percent
 higher  (when  scaled to equivalent plant size).   The United  Engineers'
 estimate  was  selected as  the basis for the revised wash-oil  final cooler
 costs.   Dravo/Still indicated the cross-tube cooler is more  expensive
 because  reuse  of  existing equipment is difficult  with this  final  cooler
 scheme.
      One  other  unit cost  related  to final  coolers  was -revised.*  The cost
 of  makeup wash  oil for wash-oil  final cooler systems was  estimated  in  the
 proposal  BID at $0.11/kg.  This  unit cost  was increased  for  the  current
 cost estimates to $0.34/kg or $283/m3'on  the  basis of information pro-
 vided by  Shenango Inc.   The quantity of makeup wash  oil  required,  how-
ever, was revised downward by a significant amount from the  proposal BID.
Comparison of the old makeup wash-oil  estimates to potential  losses  of
wash oil indicated that the proposal  BID  substantially overestimated the
 required makeup wash oil  for final  coolers.
                                 B-7

-------
B.3  COMPARISON OF WHOLE  PLANT ESTIMATE
     The changes to the  unit  cost-estimating  factors  cited above do  not
account for all the cost  differences  between  EPA  cost estimates and  those
submitted by AISI for the Bethlehem plant.  Study of  the  Bethlehem plant
cost estimates contained  in the AISI  comments revealed that  some of  the
sources included in their control  systems were  not required  to be
controlled by the proposed regulation.   This  was  partially attributable
to questions about source definitions in the  proposed regulation.  As  a
result, the source definition for excess ammonia-liquor  storage tanks  has
been changed.  Elimination of costs for  those sources not requiring
control reduces the gap  between the two  estimates.
     Assumed pipe sizes  for gas-blanketing  control systems was another
area of difference between the EPA and AISI estimates for the Bethlehem
plant.  In general, the  pip~ sizes indicated  in the industry cost
estimate exceeded the sizes assumed in the  EPA design for gas-blanketing
control systems.  The system design presented by Sirrine in  their  study
of the Bethlehem plant generally used smaller gas blanket pipe diameters
than either EPA or industry estimates.  Sirrine argued  that  the  small  gas
flow rates expected to and from the various process vessels  only  required
smaller pipe sizes.  According to Sirrine,  the more uniform  heating
achievable by  heat tracing the smaller pipe sizes and the higher  flow
rates  that would occur in  smaller pipe sizes  would reduce the  likelihood
of condensation and resultant plugging rather than increase  it.   The EPA
ultimately was  influenced  more by the fact  that the proposal BID
estimates were  based on  pipe  sizes used in existing gas-blanketing
systems.  On the whole,  the pipe sizes for the revised  gas-blanketing
cost estimates  were neither increased or decreased.  This fact  explains
some  of  the  remaining differences between EPA and industry estimates.

B.4   OTHER  COST ELEMENT  REVISIONS
      Armco  and Bethlehem Steel  submitted cost estimates  for wash-oil
scrubbers  applied  to  specific sources in some of  their plants.   Compari-
son  of their data  with the proposal  BID data for  wash-oil scrubbers
suggested  a  more  appropriate  way to  estimate costs for scrubber
applications.   The proposal  BID estimates  used scrubber  shell  area  as
the  basis  for scrubber  capital  cost,  with  shell  area estimated roughly
                                 B-8

-------
 from the expected gas flows  to be  treated.   The  industry  cost estimates
 appeared to be directly proportional  to  the  number  of  sources to  be
 treated rather than the size of the  sources.   For the  majority of
 emissions sources for which  wash-oil  scrubbers were  costed, the gas flow
 rate is intermittent.  The highest flow  would  tend  to  occur during
 pumping into the  tank,  when  flow rate would  be controlled by the  pumping
 rate.   Pump capacity is not  expected  to  vary in  direct proportion to tank
 size,  but rather  over a narrower range,  because  a particular pump
 capacity could handle a range  of tank capacities by  varying pumping size.
 For these reasons,  we agree  that the  number  of sources is a more
 appropriate basis for the estimate.   The wash-oil scrubber unit costs
 were revised.   The  result of this change is that the costs for wash-oil
 scrubbers applied to medium  and  large plants have increased compared to
 the estimates  at  proposal.
      Plan drawings  of by-product plant facilities submitted in support of
 the AISI  comments suggested  another revision to the cost estimates.  The
               t
 drawings  indicated  that  the  number of light-oil, benzene-toluene-
 xylene  (BTX),  and benzene storage tanks for plants of specific size were
 generally higher  than was assumed in the model  plants used to develop the
 proposal  BID cost-estimating equations.  The number of these tanks in the
 model facilities  was, therefore, increased.  This revision when combined
 with the  one described  above has increased the  costs estimates for all
 plants  recovering light  oil.   Table B-l lists the revised cost factors
 used in developing the  revised capital and operating cost estimates.
 B.5  EXTENSION OF UNIT COST FACTORS TO PLANT COST ESTIMATES
     The  unit cost factors provided in Tables B-l and B-2 were extended
 into full  plant cost estimates in the same manner as presented in  Chapter
 8 of the  proposal  BID.  Piping distances  and numbers of process equipment
 were specified for each model plant'size  and each group of emitting
 sources.  To reflect the variation  that  is typical  from plant  to  plant,
 minimum and maximum values were specified for piping distances  and other
 control  equipment  elements.   Minimum  and  maximum  values also were
 specified for certain of the  unit cost factors  such  as  pipe  supports and
wash oil scrubbers .  For each model  plant,  minimum  and maximum capital
costs were estimated by multiplying the equipment element  numbers  by the
                                 B-9

-------
                   TABLE B-l.  REVISED CAPITAL COST FACTORS
   Pipe + fittings.

    diameter-i n
                Cost/unit1
ripe + r i it i ng, sredur tracea,
           & insulated,

   cost/unit*  (1984$)
I $j
4 !
3 !
2 !
1 !
LOO/ft
565/ft
50/ft
>40/ft
22/ft
15/ft
70
50
30
20
<
<
(
(
L45/ft
.UO/ft
>83/ft
>72/ft
>54/ft
>46/ft
153
130
109

Valves (3-way lubricated plug valves)
   Diameter-i n
            Cost/unit

PI

Pi

8
6
4
2
1
ug valves
8
6
pe supports
t
Minimum case -
Maximum case -
Tar decanter clean

Minimum case -
Maximum case -
<
{
i
i
<
(

I

$7/ft
$30/ft
53,000
52^500
1*000
5 700
i 500
5 200

51,600
5 900


, cover, and seal
$5/ft2 ,
$30.5/ft2
Tar sumps clean, cover, and

seal
   Minimum case - $10.5/ft?
   Maximum case - $44.5/fti::
Hot tap
   Sin.  -
   12 in. -
53,800
57,600
Footnote at end of table.
                                                     (conti nued)
                                  B-10

-------
                           TABLE B-l.  (continued)
Tank sealing
   Flushing liquor circulation - $l,400/tank
   Tar tanks - $l,400/tank
Tank roof (including tank cleanout)
   Tar dewatering tank - $46.5/ft^      9
   Excess ammonia-liquor tank - $46.5/ft'::
Vessel  sealing
   Ammonia-liquor area - $3,000/unit
   Light-oil area - $l,500/unit
Nitrogen blanketing site preparation
   Large plant (assume Model  Plant 3 size)  - $30,000/plant
Flame arresters
   6 in - $2,000/arrestor
   4 in - 51,000/arrestor
Pressure/vacuum relief valves
   6 in - $1.300
   4 in - $8(30
   3 in - $660
Pressure reducing valve
   Valve - $2,000
Wash-oil scrubber pump
   Pump - $3,900
Wash-oil scrubber instrumentation
   Flow, temperature,  and pressure - $2,500
Light-oil sump cover
   Minimum case - $30.5/ft2
   Maximum case - $164/ft^

  Parenthetical numbers refer to light-oil  plant  cost  considering  restricted
  construction conditions, e.g., no welding.
                                     B-ll

-------
                  TABLE 8-2.  REVISED ANNUALIZED COST ITEMS
                  Item                                  Cost (1984$)
  Benzene credit, as fuel9                             $0.14/kg
  Benzene credit, recovered9                           $0.39/kg
  Light-oil  credit9                                    $U.27/kg
  Capital recovery (20 yr @ 6.2%)                      8.86% of capital
  Electricityb                                         $0.05/kWh
  SteamC                                              $18.30/Mg
  Cooling water0                                       $U.06/m3
  Wash oild                                             $283/m3
  Operating  labor (including plant                   $39.69/h
     overhead @ 80%)e

aDerived from Quarterly Coal Report, Energy Information Administration,
 DOE/EIA-0121, January-March 1984, table A-16.
bEscalated from proposal  BID (1982$) 20 percent (electrical rates
 generally escalated more rapidly  than overall  rate of inflation).
cEscalated from proposal  BID (1982$) by 4 percent.
^Based on $1.02/gallon + freight (assumed $0.05/gal) per telecon with
 James Zwikl, Snenango Inc., August 9, 1985.
eBased on $22.05 per hour from United Engineers estimates in Bethlehem
 Steel comments on the proposed regulations,  and 80 percent plant
 overhead rate.
                                      B-12

-------
 unit cost factors.  This procedure resulted in a set of minimum and
 maximum capital cost estimates for each group of emission sources within
 each model plant size.  These capital cost estimates are shown in Tables
 B-3 through B-18.
      Average capital cost estimates were computed for each group of
 emission sources and each model  plant size by averaging the  minimum and
 coke by-product plants in the industry,  equations were developed that
 estimated the capita"  costs for  each  group of emission sources as a
 function of coke production capacity.  An  equation  best fitting the
 capital  cost estimates for each  emission source  group and  for  the three
 model  plant sizes was  obtained by  performing  linear or curvilinear
 regression analyses  on the  estimates.  Nationwide capital  cost  estimates
 was generated by using the  equations  to  estimate  the  average capital cost
 for each plant.
     Minimum and maximum  annualized costs  were estimated for each  group
 of emission  sources  and each  model plant size using the unit cost  factors
 in Table  B-2  and the capital  costs estimated  by the above  procedure.  A
 set of equations for estimating annualized costs for each  emission  source
 group as  a  function of plant  coke production capacity was  generated by
 the same  procedure described  for capital cost estimates.
     Capital  recovery charges were computed on the basis of a 20-yr
 equipment  lifetime at 6.2 percent interest.  The equipment lifetime was
 increased from that used in the proposal BID cost estimates to  make the
 equipment lifetime assumption more compatible with typical  lifetimes for
 coke-oven plant equipment in general.   The  6.2 percent interest assump-
 tion is estimated to be the real  (net  of inflation)  cost of capital to
 the coke industry.
    In the process of estimating  nationwide annualized costs, credits  for
 recovery of benzene and/or light  oil were applied  to all  plants except
those few specifically  identified  as not  being able  to benefit  from
 recovery.  The annualized  costs are  shown in  Tables  B-3 through  B-18.
                                B-13

-------
                                        TABLE B-3.   COSTS  FUR GAS BLANKETING OF TAR DECANTER, TAR-INTERCEPTING SUMP,
                                                           AND FLUSHING-LIQUOR CIRCULATION TANK
                                                                (All Costs In 1984 Dollars)
DO
I

Cost element
Pressure taps
20-cm (8-in) pipe, m
(ft)
7.6-cm (3-in) pipe, m
(ft)
Pipe supports, m
(ft)
Valvesb
20-cm (8-in) plug valve
Clean, cover, seal decante
m2
(ft2)
Clean, cover, seal sump,
m2
(ft2)
Seal flushing liquor tanks
Capital cost0
Total capital costd
Model
Minimum
1
61
(200)
46
(ISO)
107
(350)
4
1
r,
149
(1,600)

3.0
(32)
1
72,600
102,300
plant 1
Maximum
1
122
(400)
91
(300)
213
(700)
4
1

149
(1,600)

3.0
(32)
1
172,800
243,700
Model
Minimum
1
91
(300)
46
(150)
137
(450)
6
1

223
(2,400)

23
(250)
2
103,100
145,300
plant 2
Maximum
1
244
(800)
91
(300)
335
(1,100)
6
1

223
(2,400)

23
(250)
2
285,900
403,200
Model
Minimum
1
152
(500)
91
(300)
244
(800)
10
1

446
(4,800)

46
(500)
3
176,600
248,900
plant 3
Maximum
1
366
(1,200)
183
(600)
549
(1,800)
10
1

446
(4,800)

46
(500)
3
487,500
687,300
Cost pe
Minimum
3,800
476
(145)
236
(72)
23
(7)
3,800
1,600

53.8
(5)

713
(105)
1,400


ir unit3
Maximum





98.4
(30)



328
30.5

479
(44.5)



                    Footnotes at end of table.
(continued)

-------
                                                         TABLE B-3.   (continued)
CO
I
Model plant 1
                                                               Model plant 2
                                          Model  plant 3
           Cost  element
                                        Minimum   Maximum    Minimum   Maximum    Minimum   Maximum
Annualized cost
  Maintenance, overhead (9%)e   9,210
  Utilities1"                    1,970
  Taxes, insurance (4%)         4,090
  Capital  recovery (8.86%)9      9,070
  Total  annualized cost        24,300
21,900
 3,940
 9,750
21.600
57,200
13,100
 2,740
 5,810
12.900
34,500
36,300
 7,010
16,100
35.700
95,100
                                         22,400    61,900
                                          4,710    10,900
                                          9,960    27,500
                                         22,000    60.900
                                         59,100   161,200
                                                                   values arc
          b 3-way valves, 15 cm (6 in)-$2,500;  and pressure/vacuum relief valves,  15 cm (6 in)-$l,300.
          c Capital cost includes subcontractor overhead and profit and contractor material  markup.
            Total capital cost includes construction fee, contingency,  engineering, and startup  (41%).
          e Maintenance and overhead are 5% and 4% of total  capital  cost, respectively.
          f Steam at 18.3/Mg.
          9 Capital recovery factor for 20-yr lifetime at 6.2%.
                                                                                                   Cost  per  unita
                                                                                                Minimum  Maximum

-------
cn
t—»
CTl
                                                              TAliLt U-4.  COSTS  FOR WASH-OIL  VENf SCRUBBER FOK  MR  OECANTER,  TAR-INTERCEPTING SUMP
                                                                                       ANO  FLUSHING-LIQUOR CIRCULATION TANK
                                                                                             (All Costs  in 1984 Dollars)


                                                                               H'odel 'pl'an't"l '""*   Hod'el"'pTa'n't''2 '"''"   'Ho'deT'p'l'a'n't'~3'"""       C'o's't''u'er'TTnYt'^'"
Cost element
Scrubber vessels
15.2-cm (6-in) vent pipe,b

7.6-cm (3-in) vent pipe to
sump,0 m
(ft)
2.5-cm (1-in) wash-oil
supply, m
(ft)
5.1-cm (2-in) wash oil
drain, d m
(ft)
Valves6
Seal flushing-liquor tanks
Clean, cover, and seal tar
decanter, w?
(ft2)
Clean, cover, and seal tar
sumps, in2
(ft2)
Pump
Instrumentation^
Capital costy
Total capital costh
Minimum
4,
m 46
(ft) (150)

46
(150)

61
(200)

61
(200)
4
1

149
(1,600)

2.9
(32)
1
1
68,11)0
96,100
Maximum
4
46
(150)

91
(300)

152
(500)

152
(500)
4
1

149
(1,600)

17 2.97
(32)
1
1
146,900
207, inu
Minimum
6
76
(250)

46
(150)

91
(300)

91
(300)
6
2

223
(2,400)

23.2
(250)
1
1
100,500
141,700
Maximum
6
76
(250)

91
(300)

610
(2,000)

610
(2,000)
6
2

223
(2,400)

23.2
(250)
1
1
270,700
381,700
Minimum
10
122
(400)

91
(300)

122
(400)

122
(400)
10
3

446
(4,800)

46.5
(500)
2
2
170,600
240,500
Maximum
10
122
(400)

183
(600)

762
(2,500)

762
(2,500)
10
3

446
(4,800)

46.5
(500)
2
2
451.700
6 36, BOO
Minimum
1,000
351
(107)

259
(79)

49.2
(15)

95.1
(29)
3,800
1,400

53.8
(5)

113
(10.5)
3,900
2,51)0


Maximum
2,000














328.3
(30.5)

479
(44.5)





-------
                                                                           TABLE B-4.   (continued)
CXI
i

Cost element
Annual i zed costs
Maintenance, overhead (9%)'
UtilitiesJ
Taxes, insurance (4%)
Operating labor^
Capital recovery (8.86*)1
Total annual 1 zed cost

Model
Minimum

8,650
1,360
3,840
7,240
8,510
29,600

plant 1
Maximum

18,600
1,790
8,280
7,240
1£,300
54,300

Model
Minimun

12,800
2,110
5,670
7,240
12,600
40,300

plant 2
i Maximum

34,300
2,970
15,300
7,240
33*800
93,600

Model
Minimum

21,600
3,700
9,620
14,500
21.300
70,800

Plant 3 Cost per unit3
Maximum Minimum Maximum

57,300
4.040
25,500
14,500
56.400
157,700

                                                     low values are shown as maximum and
b Pipe, steam traced @ $328/m or $100/ft and pipe supports 9 $23/m or $7/ft.
c Pipe @ $236/m or $72/ft and pipe supports 0 $23/m or $7/ft.
d Pipe e $72.20/m or $22/ft and pipe supports @ $23/m or $7/ft.
e 3-way valves, 15 cm (6-in) - $2,500 and pressure/vacuum release  valves,  15 cm  (6  in)  - $1,300.
f Includes flowmeter with alarm, pressure gauge,  and temperature gauge.
9 Capital  cost includes subcontractor overhead and  profit  and contractor  or  material markup.
n Total capital cost includes construction fee, contingency, engineering,  and startup (41%).
' Maintenance and overhead  are 5% and 4% of total capital  cost,  respectively.
J Electricity at $0.05/kWh.
k For 30 min/day/scrubber system at  $39.69/h.
1  Capital  recovery factor for 20-yr  lifetime  at 6.2%.
                                                                                                                                          *****

-------
                                                                TABLE  B-5.   COSTS  FOR  GAS  BLANKETING AMMONIA LIQUOR  STORAGE  TANKS
                                                                                   (All  Costs  in 1984 Dollars)
TO
I
oo

Cost element
15.2-cm (6-1n) vent pipe, ra
(ft)
Valvesb
lb.2-cm (6-in) plug valve
Pipe supports, m
(ft)
Seal tanks
Tank roofs, of
(ft2)
Capital cost0
Total capital costd
Annuall zed costs
Maintenance, overhead (9%)e
Utilitiesf
Taxes, insurance (4%)
Capital recovery (8.86*)9
Total annual i zed cost
Model
Minimum
46
(150)
1
1
46
(150)
1
0
(0)
23,800
33,500

3,010
870
1,340
2,970
8,190
plant 1
Maximum
152
(500)
1
1
152
(500)
1
49.3
(531)
97,400
137,300

12,300
2,890
5,490
12,200
32,900
Model
Minimum
61
(200)
3
1
61
(200)
3
0
(0)
42,700
60,200

5,420
1,150
2,410
5,330
14,300
plant 2
Maximum
183
(600)
3
1
183
(600)
3
74.7
(804)
136,700
192,700

17,300
3,460
7,710
17.100
45,600
Model
Minimum
91
(300)
6
1
91
(300)
6
0
(0)
73,800
104,100

9,370
1,730
4,160
9,220
24,500
plant 3
Maximum
305
(1,000)
6
1
305
(1,000)
6
169
(1,816)
256,100
361,200

32,500
5,770
14,400
32,000
84,700
Cost per unita
Minimum Maximum
328
(100)
3,800
900
23 98.4
(7) (30)
3,000
501
(46.5)








                                       a Where a  range of unit costs was used, the high and low values are shown as maximum and minimum;  where  only  a  single
                                        value was used, it is shown in the minimum column.


                                       b 3-way valves, 15.2 cm (6 in)-$2,500; and pressure/vacuum relief valves, 15.2 cm (6 in)-$l,300.


                                       c Capital  cost includes subcontractor overhead and profit and contractor material markup.


                                       d Total capital cost includes construction fee, contingency, engineering, and startup (41%).


                                       e Maintenance and overhead are 5% and 4% of total capital cost, respectively.


                                       f Steam at $18.3/My.

                                       9 Capital  recovery factor for a 20-yr lifetime at 6.2%.

-------
UD
                               TABLE B-6.  COSTS FOR  WASH-OIL  VENT  SCRUBBER  FOR AMMONIA LIQUOR  STORAGE TANKS
                                                        (All Costs  in  1984 Dollars)
Model plant 1
Cost element Minimum
Scrubber vessels
7.6-cm (3-in) vent pipeb, m
(ft)
2.5-cm (1-in) wash-oil line, m
(ft)
5.1-cm (2-in.) wash-oil
drain,0 m
(ft)
Valvesd
Pumps
Instrumentation6
Seal tanks
Tank roofs, m2
(ft2)
Capital cost^ 14
Total capital cost9 20
I
9.1
(30)
30.5
(100)

30.5
(100)
1
0
1
1
0
(0)
,600
,600
Maximum
1
9.1
(30)
152
(500)

152
(500)
1
2
1
1
49.3
(531)
65,700
92,700
Model plant 2
Minimum
3
46
(150)
61
(200)

61
(200)
3
0
1
3
0
(0)
39,200
55,300
Maximum
3
46
(150)
152
(500)

152
(500)
3
2
1
3
74.7
(804)
100,600
141,900
Model plant 3
Minimum
6
91
(300)
122
(400)

122
(400)
6
0
1
6
0
(0)
76,000
107,100
Maximum
6
91
(300)
305
(1,000)

305
(1,000)
6
3
1
6
169
(1,816)
204,500
288,400
Cost per unit3
Minimum Maximum
1,000 2,000
259
(79)
49.2
(15)

95.1
(29)
1,360
3,900
1,300
3,000
501
(46.5)


          Footnotes  at end of table.
                                                                                                                (continued)

-------
                                                             TABLE  B-6.   (continued)
CO
i
IV)
o
Model plant 1
Cost element
Annual i zed costs
Maintenance, overhead (9%)n
Utilities1
Taxes, insurance (4%)
Operating labor J
Capital recovery (8.86%)k
Total annual ized cost
Minimum

1,860
92
825
7,240
1,830
11,800
Maximum

8,340
92
3,710
7,240
8,210
27,600
Model plant 2
Minimum

4,980
451
2,210
7,240
4,900
19,800
Maximum

12,700
451
5,680
7,240
12,600
38,700
Model plant 3 Cost per unit3
Minimum

9,640
902
4,280
7,240
9,490
31,600
Maximum Minimum Maximum

26,000
902
11,500
7,240
25,600
71,200
Where a range of unit costs was used,  the  high  and  low  values  are  shown  as maximum  and minimum; where only  a single
value was used, it is shown in the minimum column.
Pipe 9 $236/m or $72/ft and pipe supports  & $23/m or  $7/ft.
Pipe & $72.1/m or $22/ft and pipe supports & $23/m  or $7/ft.
3-way valves, 7.6 cm (3 in)-$700; and  pressure/vacuum relief valves,  7.6 cm  (3  in)-$660.
Includes flowmeter with alarm, pressure gauge,  and  temperature gauge.
Capital cost includes subcontractor overhead and profit and contractor material markup.
Total capital cost includes construction fee, contingency, engineering,  and  startup (41%).
Maintenance and overhead are 5% and 4% of  total capital  cost,  respectively.
Electricity at $0.05/kWh.
For 30 min/day at $39.69/h.
Capital recovery factor for 20-yr lifetime at 6.2%.

-------
I
ro
                                            TABLE B-7.  COSTS FOR GAS BLANKETING OF LIGHT-OIL CONDENSER,  LIGHT-OIL  DECANTER
                                                                 WASH-OIL DECANTER, AND CIRCULATION TANK
                                                                       (All  Costs in 1984 Dollars)

Cost element
Pressure tap
10- to 15-cm (4- to 6-1n)
pipe,b m
(ft)
Plug valve, 15 cm (6 in)
Valves0
Seal vessels
Flame arrestors
Capital costd
Total capital cost6
Annuali zed costs
Maintenance, overhead (9*)f
UtilitiesS
Taxes, insurance (4%)
Capital recovery (8.86%)n
Total annual ized cost
Model
Minimum
1

61
(200)
1
6
6
6
60,000
84,500

7,600
1,100
3,400
7.500
19,500
plant 1
Maximum
1

183
(600)
1
6
6
6
118,900
167,600

15,100
3,200
6,700
14,800
39,800
Model
Minimum
1

122
(400)
1
8
8
8
98,000
138,200

12,400
2,100
5,500
12,200
32,300
plant 2
Maximum
1

244
(800)
1
8
8
8
156,900
221,200

19,900
4,200
8,800
19,600
52,600
Model
Minimum
1

183
(600)
1
13
13
13
149,000
210,000

18,900
3,200
8,400
18,600
49,100
plant 3 Cost per unit3
Maximum Minimum Maximum
1 3,800

305 483.1
(1,000) (147.25)
1 900
13 1,800
13 1,500
13 1,000
207,900
293,100

26,400
5,300
11,700
26,000
69,300
vatue wasaut!d0fun!<  T'S was "sed! the h1^ and >°«
value was used,  it  is  shown in the minimum column.
                                                                                         are shown as  maximum and minimum;  where  only  a  single
                         ta Assumes  75* of pipe is 15-cm (6-in)  header and 25% is  10-cm (3-in)  vent  lines.

                         c 3-way valves, 1U.2 cm (4 in)-$l,000;  and  pressure/vacuum relief  valves,  10.2  cm (4  in)-$800.

                         d Capital  cost includes subcontractor  overhead  and  profit  and contractor material  markup.

                         e Total capital cost includes construction  fee,  contingency,  engineering,  and startup  (41%).

                         f Maintenance and overhead are 5*  and 4%  of total capital  cost,  respectively.

                         9 Steam at $18.3/My.
                         h Capital  recovery factor for 20-yr  lifetime at  5.2%.

-------
CO
I
ro
IN3
                                        TABLE B-8   COSTS OF WASH-OIL VENT SCRUBBER FOR LIGHT-OIL CONDENSER, LIGHT-OIL
                                                    DECANTERS, WASH-OIL DECANTERS, AND CIRCULATION TANKS
                                                                 (All Costs in 1984 Dollars)

Cost element
Scrubber vessels
10.2-cm (4-in) vent pipe,b

2.5-cm (1-in) wash-oil
supply pipe, m
(ft)
5.1-cm (2-in) wash-oil
drain pipe,c m
(ft)
Valvesd
Seal vessels
Pump
Instrumentation6
Capital costf
Total capital cost9
Model p
Minimum
6
m 110
(ft) (360)

30.5
(100)

30.5
(100)
6
6
0
1
83,300
117,500
lant 1
Maximum
6
110
(360)

122
(400)

122
(400)
6
6
2
1
114,200
161,100
Model j)
Minimum
8
146
(480)

91.4
(300)

91.4
(300)
8
8
0
1
119,800
168,900
lant 2
Maximum
8
146
(480)

244
(800)

244
(800)
8
8
2
1
164,100
231,300
Model j
Minimum
13
238
(780)

122
(400)

122
(400)
13
13
0
2
190,600
268,700
jlant 3
Maximum
13
238
(780)

305
(1,000)

305
(1,000)
13
13
4
2
253,400
357,200
Cost per unit3
Minimum Maximum
1,000 2,000
449.5
(137)

65.6
(20)

121.4
(37)
1,800
1,500
3,900
2,500


                      Footnotes at  end  of  table.
                                                                                                                      (continued)

-------
                                               TABLE B-8.   (continued)








co
i
ro
CO

Cost element
Annuali zed costs
Maintenance, overhead (9%)n
Utilities1
Taxes, insurance (4%)
Operating labor J
Capital recovery (8.86%)k
Total annuali zed cost
a Where a range of unit costs
Model
Minimum

10,600
1,400
7,700
7,240
10,400
34,300
was used,
plant 1
Maximum

14,500
1,400
6,440
7,240
14.300
43,900
the high
Model plant 2
Minimum

15,200
1,920
6,750
7,240
15,000
46,100
Maximum

20,800
1,920
9,250
7,240
20,500
59,700
and low values are shown
Model plant 3 Cost oer unit3
Minimum

24,200
3,170
10,700
14,500
13,800
76,400
as maxim
Maximum Minimum Maximum

32,200
3,170
14,300
14,500
31,700
95,800
lum and minimum: where onlv a sinalp
  value was used, it is shown in the minimum column.
b Pipe, steam traced 
-------
                                                   TABLE B-9.  COSTS FOR GAS BLANKETING OF LIGHT-OIL AND BTX STORAGE TANKS
                                                                         (All Costs in 1984 Dollars)
ro
Cost element
10- to 15-cm (4- to 6-in)
p1pe,b m
(ft)
Pipe supports, m
(ft)
Seal tanks
Flame arresters
Capital costd
Total capital cost6
Annual 1 zed costs
Maintenance, overhead (9%)f
UtilitiesS
Taxes, insurance (4%)
Capital recovery (8.86%)h
Total annual i zed cost
Model
Minimum
49
(160)
4
49
(160)
4
4
41,900
59,100

5.32U
850
2,360
5.230
U.800
plant 1
Maximum
183
(600)
4
183
(600)
4
4
123,600
174,200

15,700
3,170
6,970
15.400
41,300
Model plant 2
Minimum
61
(200)
9
61
(200)
9
9
69,600
98,100

8,830
1,060
3,920
8,690
22,500
Maximum
259
(850)
9
259
(850)
9
9
189,400
267,000

24,000
4,490
10,700
23.700
62,900
Model plant 3 Cost per unit3
Minimum
122
(400)
15
122
(400)
15
15
126,200
177,900

16,000
2,110
7,120
15,800
41,000
Maximum Minimum Maximum
335 483.1
(1,100) (147.25)
15 1,800
335 23 98.4
(1,100) (7) (30)
15 1,500
15 1,000
259,500
365.900

32,900
6,810
14,600
32,400
85,800
                             a Where a ranye of unit costs was used,  the high  and  low  values  are  shown  as maximum and minimum; where only a single
                               value was used, it is shown in the minimum column.

                             b Assumes 75% of pipe is 15-cm (6-in) header and  25%  is 10-cm (3-in)  vent  lines.

                             c 3-way valves, 10.2 cm (4 in) - $1.000; and pressure/vacuum relief  valves, 10.2  cm. (4 in) - $800.

                             d Capital cost includes subcontractor overhead and  profit and contractor markup.

                             e Total capital cost includes construction fee, contingency, engineering,  and  startup  (41%).

                             f Maintenance and overhead are 5% and 4% of total capital cost,  respectively.

                             y Steam at $18.3/Mg.

                             n Capital recovery factor for 2U-yr lifetime at 6.2%.

-------
en
                                 TABLE B-10.  COSTS OF WASH-OIL VENT SCRUBBER FOR LIGHT-OIL AND BTX STORAGE TANKS
                                                            (All  Costs In 1984 Dollars)

Cost element
Scrubber vessels
10-cm (4-in) vent pipeb,

2.5-cm (1-in) wash-oil
line, m
(ft)
b.l-cm (2-in) wash-oil
drain,0 m
(ft)
Pumps
Valvesd
Vessel sealing
Instrumentation6
Capital costf
Total capital cost9
Model
Minimum
4
m 61
(ft) (200)

30.5
(100)

30.5
(100)
0
4
4
1
52,800
74,400
plant 1
Maximum
4
61
(200)

183
(600)

183
(600)
2
4
4
1
93,100
131,300
Model
Minimum
9
137
(450)

30.5
(100)

30.5
(100)
0
9
9
1
108,600
153,100
plant 2
Maximum
9
137
(450)

213
(700)

213
(700)
2
9
9 .
1
159,600
225,000
Model
Minimum
15
229
(750)

61
(200)

61
(200)
0
15
15
2
183,700
258,900
plant 3
Maximum
15
229
(750)

244
(800)

244
(800)
4
15
15
2
248,500
350,300
Cost per unit3
Minimum Maximum
1,000 2,000
449.5
(137)


(20)


(37)
3,900
1,800
1,500
2,500


              Footnotes at end of table.
                                                                                                                   (continued)

-------
                                                             TABLE B-10.  (continued)
Model plant 1
Model plant 2
Model plant 3
                                                                                                                   Cost per unit3
rv>
cr>
Cost element
Annual i zed costs
Maintenance, overhead (9%)h
Utilities1
Taxes, insurance (4%)
Operating laborJ
Capital recovery (8.86%)k
Total annual i zed cost
Minimum

6,700
794
2,980
7,240
6,600
24,300
Maximum

11,800
794
5,250
7,240
11,600
36,700
Minimum

13,800
1,800
6,120
7,240
13,600
42,500
Maximum

20,200
1,800
9,000
7,240
19,900
58,200
Minimum

23,300
2,980
10,400
14,500
22,900
74,100
Maximum Minimum Maximum

31,500
2,980
14,000
14,500
31,000
94,000
                a Where  a  range  of  unit  costs was used, the high and low values are shown as maximum and minimum; where only a single

                 value  was  used, it  is  shown in the minimum column.


                D Pipe 0 $426.5/m or  $130/ft and pipe supports @ $23/m or $7/ft.


                c Pipe B $98.4/m or $30/ft  and pipe supports & $23/m or $7/ft.


                d 3-way  valves,  10.2  cm  (4  in) - $1,000; pressure/vacuum relief valves, 10.2 cm (4-in) - $800.


                e Includes flowmeter  with alarm, pressure gauge, and temperature gauge.


                f Capital  cost includes  subcontractor overhead and profit and contractor material markup.


                9 Total  capital  cost  includes construction fee, contingency, engineering, and startup (41%).


                n Maintenance  and overhead  are 5% and 4% of total capital cost, respectively.


                1 Steam  at $18.3/Mg and  electricity at $0.05/kWh.


                J For 30 min/day/scrubber system at $39.69/h.


                k Capital  recovery  factor for 20-yr lifetime at 6.2%.

-------
                                             TABLE B-ll.  COSTS FOR GAS BLANKETING OF TAR COLLECTING, STORAGE, AND OEWATERING TANKS
                                                                           (All Costs In 1984 Dollars)
OT

PO

Cost element
15-cm (6-in) pipe, m
(ft)
Seal tanks
Tank roofs-dewaterlng, m2
(ft2)
Pipe supports, m
(ft)
Valves6
Capital cost0
Total capital costd
Annual! zed costs
Maintenance, overhead (9%)e
Utilitiesf
Taxes, insurance (4%)
Capital recovery (8.86%)9
Total annualized cost
Model
Minimum
61
(200)
5
0
(0)
61
(200)
5
47,400
66,800

6,020
1,150
2,670
5,920
15,800
plant 1
Maximum
152
(500)
5
49
(531)
152
(500)
5
115,700
163,100

14,700
2,880
6,530
14,500
38,500
Model
Minimum
91
(300)
10
0
(0)
91
(300)
10
84.100
118,600

10,700
1,730
4,740
10.500
27,700
plant 2
Maximum
762
(2,500)
10
75
(804)
762
(2.500)
10
414,400
584,300

52,600
14.400
23,400
51,800
142,100
Model
Minimum
122
(400)
16
0
(0)
122
(400)
16
126,000
177,700

16,000
2,310
7,110
15,700
41,100
Plant 3 Cost per unit*
Maximum Minimum Maximum
914 328.1
(3,000) (100)
16 1,400
169 500.5
(1.816) (46.5)
914 23 98.4
(3.000) (7) (30)
16 3,800
557,600
786,300

70,800
17,300
31,500
69.700
189,200
                              b  From  Table B-l, 3-way vaives, 15 cm (6 in) - $2.500; pressure/vacuum relief valves. 15 cm (6 in) - $1,300

                              c  Capital cost includes subcontractors overhead and profit and contractor material markup.

                              d  Total capital cost includes construction fee, contingency, engineering, and startup (41%).

                              e  Maintenance and overhead are 5t and 4% of total  capital cost, respectively.

                              f  Steam at $18.3/My.

                              y  Capital recovery factor for 2U-yr lifetime at 6.2%.

-------
                                              TABLE li-12.  COSTS OK WASH-OIL VENT SCRUBBER  FOR  TAR  COLLECTING,  STORAGE,  AND OEWATEKING TANKS

                                                                                (All  Costs  in 1984  Dollars)
00
I
oo

Cost element
Scrubber, heat exchanger,
separator
lb.2-cm (fa-in) vent
pipe,b m
(ft)
10.2-cm (4-1 nO wastewater
pipe, m
(ft)
2.5-cm (1-m) wash oil supply
pipe, m
(ft)
5.1-cm (2-in) wash oil
drain pipe,c m
(ft)
Seal tank
Tank roofs, dewatering, or
(ft2)
Pump
Valves'1
Valves and level control
Instrumentatione
Capital costf
Total capital cost^
Model
Minimum

62,700

122
(400)

30.5
(100)

30.5
(100)

30.5
(100)
5
0
(0)
1
5
1
1
159, 701)
225,200
plant 1
Maximum

62,700

122
(400)

61
(200)

152
(500)

152
(500)
5
49.3
(531)
1
5
1
1
210,300
2 96,500
Model
Minimum

144,000

244
(800)

30.5
(100)

91.4
(300)

91.4
(300)
10
0
(0)
1
10
1
1
318,600
449,200
plant 2
Maximum

144.000

244
(800)

61
(200)

640
(2,100)

640
(2,100)
10
74.7
(804)
1
10
1
1
443,500
625,300
Model
Minimum

234,000

390
(1,280)

30.5
(100)

122
(400)

122
(400)
16
0
(0)
2
16
2
2
511,100
720,600 1
plant 3
Maximum

234,000

390
(1.280)

61
(200)

853
(2,800)

853
(2,800)
16
169
(1,816)
2
16
2
2
709,400
,000,300
Cost per unit8
Minimum Maximum



351
(107)

272.3
(33)

49.2
(15)

95.1
(29)
1,400
500.5
(46.5)
11,000
3,800
2,000
2,500


                                   Footnotes at end of table.
                                                                                                                                      (continued)

-------
                                                                  TABLE  B-12.   (Continued)
f\>

Cost element
Annual ized costs
Maintenance, overhead (9%)h
Utilities1
Taxes, insurance
Operating laborJ
Capital recovery1*
Total annualized cost
Model f
Minimum

20,300
9,890
9,000
7,240
20,000
66,400
slant 1
Maximum

26,700
10,300
11,900
7,240
26,300
82,300
Model |
Minimum

40,400
33,700
18,000
7,240
,39,800
139,200
Jlant 2
Maximum

56,300
34,100
25,000
7,240
55,400
178,100
Model j
Minimum

64,900
72,400
28,800
14,500
63,800
244,400
2.1 ant 3 Cost per unit3
Maximum Minimum Maximum

90,000
72,800
40,000
14,500
88,600
305,900
                                                                        low values are shown as
                   b Pipe, steam traced 9 $328/m or $100/ft and pipe supports 0 $23/m or $7/ft.
                   c Pipe 9 $72.20/m or $22/ft and pipe supports 9 $23/m or $7/ft.
                   d 3-way valves, 15 cm (6 in)-$2,500, pressure/vacuum release valves,  15 cm (6  in)-$l,300.
                   e Includes flowmeter with alarm, pressure gauge,  and temperature gauge.
                   f Capital  cost includes  subcontractor overhead and profit  and contractor material  markup.
                   9 Total capital  cost includes  construction  fee, contingency,  engineering,  and  startup  (41%).
                   h Maintenance and overhead are 5% and 4% of total  capital  cost,  respectively.
                   1  Steam at $18.3/Mg  and  electricity  at $0.05/kWh.
                   J  For 30 min/day/scrubber system at  $39.69/h.
                   k  Capital  recovery factor for  20-yr  lifetime  at 6.2%.

-------
                                                       TABLE  B-13.   COSTS  FOR  COVERING  LIGHT-OIL  SUMP

                                                                 (All  Costs  in 1984  Dollars)
co
i
oo
o

Cost element
Clean, cover, and seal, m2
(ft2)
7.6-cm (3-in) vent pipe, ra
(ft)
Capital costb
Total capital costc
Annualized costs
Maintenance, overhead (9%)d
Taxes, insurance (4%)
Capital recovery (8.86%)e
Total annualized cost
Model p
Minimum
3.3
(36)
4.6
(15)
1,700
2,390

215
96
212
523
>lant 1
Maximum
20.9
(225)
4.6
(15)
37,500
52,900

4,760
2,120
4,690
11,600
Model
Minimum
3.3
(36)
4.6
(15)
1,700
2,390

215
96
212
523
plant 2
Maximum
93
(1,000)
4.6
(15)
164,600
232,100

20,900
9,280
20,600
50,700
Model [
Minimum
6.7
(72)
9.1
(30)
3,400
4,790

431
192
424
1,050
>lant 3 Cost per unit3
Maximum Minimum Maximum
186 328 1,765
(2,000) (30.5) (164)
9.1 131
(30) (40)
329.200
464,200

41,800
18,600
41,100
101,500
                     a  Where a  range  of  unit  costs was used, the high and low values are shown as maximum and minimum; where only a single

                       value was used, it  is  shown in the minimum column.


                     0  Capital  cost includes  subcontractor overhead and profit and contractor material markup.


                     c  Total capital  cost  includes construction fee, contingency, engineering, and startup (41%).


                     d  Maintenance and overhead  are 5% and 4% of total capital cost, respectively.


                     e  Capital  recovery  factor for 20-yr lifetime at 6.2%.

-------
                                           TABLE  B-14.   COSTS  FOR  NITROGEN  OR  NATURAL  GAS  BLANKETING  OF  PURE  BENZENE  STORAGE  TANKS
                                                                       (All  Costs  in 1984  Dollars)
CO

OJ

Cost element
2.5-cm (1-in) gas supply, ^ m
(ft)
7.6-cm (3-in) vent pipe, m
(ft)
Pressure controller
Pressure reducers
Site preparation
10.2-cm (4-in) flame arrestors
Valves0
Tank sealing
Pipe supports, m
(ft)
Capital costsd
Total capital costs6
Annuali zed costs
Maintenance, overhead (9S)f
UtilitiesQ
Taxes, insurance (4%)
Operating laborh
Capital recovery (a.86%)1
Total annual i zed cost
Model p
Minimum
12.2
(40)
15.2
(50)
1
2
0
1
1
1
15.2
(50)
16,100
22,700

2,040
0
910
0
2,010
4,960
lant 1
Maximum
30.5
(100)
91.4
(300)
1
2
8,000
1
1
1
91.4
(300)
46,900
66,100

5,950
1,700
2,640
7,240
5.860
23,400
Model p
Minimum
30.5
(100)
30.5
(100)
1
2
0
3
3
3
30.5
(100)
28,100
39,600

3,560
0
1,580
0
3.510
8.650
ilant 2
Maximum
91.4
(300)
152
(500)
1
2
18,400
3
3
3
152
(500)
86,200
121,500

10,900
6,700
4,860
7,240
10.800
40,500
Model
Minimum
61
(200)
61
(200)
1
2
0
7
7
7
61
(200)
51,500
72,600

6,540
0
2,910
0
6,440
15,900
jjlant 3 Cost per unit3
Maximum Minimum Maximum
213 (88.6)
(700) (27)
244 164
(800) (50)
1 4,400
2 2,000
30,000
7 1 ,000
7 1,360
7 1,400
244 23 98.4
(800) (7) (30)
147,600
208,100

18,700
15,000
8,330
7,240
18,400
67,700
                         Footnotes at end of table.
                                                                                                                                     (continued)

-------
                                                                TABLE B-14.   (continued)
                 a Where a range of unit costs was  used, the  high  and low  values are shown as maximum and minimum; where only a single
                   value was used, it is shown in the minimum column.

                 5 Includes pipe supports dt $23/m  ($7/ft).

                 c 3-way valves, 7.6 cm (3 in)-$700;  and pressure/vacuum relief valves, 7.6 cm  (3 in)-$660.

                 d Capital cost includes subcontractor overhead  and  profit and contractor material markup.

                 e Total capital cost includes construction fee, contingency, engineering, and  startup  (41%).

                 f Maintenance and overhead are 5%  and 4%  of  total capital cost, respectively.

                 9 Nitrogen at $0.27/m3 (0.76/100 ft3).  Includes  rental of 5.7-m3  (1,500-gal)  liquid nitrogen storage tank, vaporizer,
co                 and gas usage.  Some plants are  assumed to have a nitrogen source and others must purchase nitrogen.
CO
1X3               n For 30 min/day at $39.69/h when  liquid  nitrogen is used.

                 i Capital recovery factor for 20-yr  lifetime at 6.2%.

-------
                                                  TABLE B-lb.  COSTS OF HASH-OIL

                                                                      (All Costs
VENT SCRUBBER FOR BENZENE STORAGE TANKS
in 1984 Dollars)
co
 i
co
CO
Model plant 1
Cost element
Scrubber vessels

2.b-cm (1-in) wash-oil line,

5.1-cm (2-in.) wash-oil drain,

Pump

10-cm (4-in) vent pipe0, m
(ft)
Valvesd

10.2-cm (4-in) flame arrestors

Tank sealing

Instrumentation6

Capital costf
Total capita) costsS
Annuali zed costs
Maintenance, overhead (9%)n
Utilities1
Taxes, insurance (4%)
Operating laborJ
Capital recovery (16.3%)k
Total annual ized cost
Minimum
1

Maximum
i
i
m 30.5 183
(ft) (100)
b m 30.5
(ft) (100)
o

15.2
50
i
i
1

1

1

17,300
24,300

2,190
6
973
7,240
2,160
12,600
(600)
183
(600)
i
A
15.2
50


1

1

1
i
50,700
71,400

6,430
6
2,860
7,240
6.330
22,900
Model plant 2
Minimum


30.5
(100)
30.5
(100)


45.7
150

3






35,400
49,800

4,490
19
1,990
7,240
4,420
18,200
Maximum

3
213
(700)
213
(700)

2
45.7
(150)

3





1
80,400
113,300

10,200
19
4,530
7,240
10,040
32,000
Model
Minimum

7
61
(200)
61
(200)

0
107
(350)

7

7

7

1
77,300
108,900

9,800
43
4,360
7,240
9.650
31,100
plant 3
Maximum

7
244
(800)
244
(800)

2
107
(350)

7

7

7

1
126,300
178,000

16,000
43
7,120
7,240
15,800
46,200
Cost per unit3
Minimum Maximum

1,000 2,000
65.6
(20)
121.4
(37)

3,900
252.6
(77)

1,800

1,000

1,400

2,500









                       Footnotes at end of table.
                                                                                                                                  (continued)

-------
                                                TABLE  B-lb.   (continued)
a Where a range of unit costs  was  used,  the  high  and  low  values  are  shown  as maximum  and minimum; where  only  a  single
  value was used, it is shown  in the  minimum column.

b Pipe & 98.4/m or $30/ft and  pipe supports  @ $23/m or  $7/ft.

c Pipe e $229.6/m or $70/ft and  pipe  supports (? $23/m or  $7/ft.

d 3-way valves, 10.2 cm (4 in)-$l,000; and pressure/vacuum  relief  valves,  10.2 cm  (4  in)-$800.

e Includes flowmeter with alarm, pressure gauge,  and  temperature gauge.

f Capital cost includes subcontractor overhead and profit and contractor or material  markup.

9 Total capital cost includes  construction fee, contingency, engineering,  and startup (41%).

n Maintenance and overhead are 5%  and 4% of  capital,  respectively.

1 Electricity at $0.05/kWh.

J For 30 min/day at $39.69/h.

k Capital recovery factor for  20-yr lifetime at 6.2%.

-------
                                                                   TABLE 8-16.  COSTS FOR TAR BOTTOM FINAL COOLER
                                                                             (All Costs in 1984 Dollars)
DO
 i
GO
cn

Cost element
Tank-separator, decanter tank,
wastewater tank, and tar tank
Pumps - tar transfer, water skimmer
Piping and valves
Site preparation, modify existing
cooler, miscellaneous
Instrumentation (6.75% of equipment)
Electrical (10.5% of equipment)
Capital cost3
Total capital costb
Annual i zed cost
Maintenance, overhead (9%)c
Utilities'1
Taxes, insurance (4%)
Operating labor6
Capital recovery (8.86%)f
Total annual ized cost
Model t
Minimum

61,700
16,300
26,800

31,700
7,070
11,000
154.500
217,800

19,600
2,200
8,710
22,000
19.300
71,800
>lant 1
Maximum

61,700
16,300
82,800

31,700
10,800
16,900
220.100
310,400

27,900
5,040
12,400
22,000
27,500
94,900
Model p
Minimum

141,700
26,600
80,100

72,800
16,800
26,100
363,900
513,200

46,200
7,590
20.500
22,000
45.500
141,800
lant 2
Maximum

141,700
26,600
247,700

72,800
28,100
43.700
560,500
790,200

71,100
16,000
31,600
22,000
70,000
210,800
Model
Minimum

230,500
37,800
152,000

118,400
28,400
44,100
611.100
861,700

77,600
16,200
34,500
22 .000
76,300
226,600
plant 3
Maximum

230.500
37,800
470,000

118,400
49,800
77,500
984.000
1,387,400

124,900
32,300
55,500
22,000
122.900
357,500
                                a  Capital cost includes subcontractor overhead and profit and contractor material markup.


                                  Total capital cost includes construction fee, contingency, engineering, and startup (4U).
                                                  i\
                                c  Maintenance and overhead are 5% and 4% of total capital cost, respectively.

                                d  Steam at $18.3/My and electricity at $0.05 kWh.

                                6  For 1.5 h/day at $39.6y/h.


                                f  Capital recovery factor for 20-yr lifetime at 6.2%.

-------
                                                                     TABLE B-17.   COSTS  FOR  WASH-OIL FINAL COOLER
                                                                              (All  Costs in  1984  Dollars)
CD
CO
en
Cost element

Final cooler
Instrumentation (6.75% of final
cooler capital cost)
Electrical (10.5% of final cooler
capital cost)
Capital cost3
Total capital costb
Annuali zed costs
Maintenance, overhead (9%)c
Utilitiesd
Makeup wash oil6
Taxes, insurance (4%)
Operating laborf
Capital recovery (8.86%)9
Total annual i zed cost
Model plant 1

708,000
47,800
74,300
830,100
1,171,000

105,400
46,200
7,600
46,800
69.000
103,700
378,700
Model plant 2

1,627,000
109,800
170,800
1,908,000
2,690,000

242,100
184,900
30,500
107,600
69,000
238,300
872,400
Model plant 3

2,646,000
178,600
277,800
3,102,000
4,374,000

393,700
416,100
68,500
175,000
69,000
387 ,600
1,509,900
                                 a Capital  cost includes  subcontractor overhead and profit and contractor material markup.

                                 D Total  capital  cost  includes construction fee, contingency, engineering, and startup  (41%).

                                 c Maintenance and overhead are 5% and 4% of total capital cost, respectively.

                                 d Steam  at $18.3/Mg and  electricity at $0.05/kWh.

                                 e Estimated at $U.34/kg  ($1.07/gal); based on losses to wastewater and light-oil crude residue.

                                 f For 4.8  h/day at $39.69/h.

                                 9 Capital  recovery factor for 20-yr lifetime at 6.2%.

-------
                                                      TABLE B-18.  COSTS OF MIXER-SETTLER FOR NAPHTHALENE PROCESSING AND HANDLING
                                                                              (All Costs in 1984 Dollars)
CO
Model plant 1
Cost element
Tank-separator, decanter tank,
wastewater tank, and tar tank
Pumps - tar transfer, water skimmer
Piping and valves
Site preparation, modify existing
cooler, miscellaneous
Instrumentation (6.75% of equipment)
Electrical (10.5% of equipment)
Capital cost3
Total capital cost15
Annuali zed cost
Maintenance, overhead (9%)c
Utilitiesd
Taxes, insurance (4Z)
Operating labor6
Capital recovery (8.86%)f
Total annual i zed cost
Minimum
61,700
16,300
26,800
31,700
7,070
11,000
154,500
217,800

19,600
2,200
8,710
22,000
19,300
71,800
Maximum
61,700
16,300
82,800
31,700
10,800
16,900
220.100
310,400

27,900
5,040
12,400
22,000
27 ,500
94,900
Model plant 2
Minimum
141,700
26,600
80,100
72,800
16,800
26,100
363.900
513,200

46,200
7,590
20.500
22,000
45,500
141,800
Maximum
141,700
26,600
247,700
72,800
28.100
43,700
560.500
790,200

71,100
16,000
31,600
22,000
70,000
210,800
Model
Minimum
230,500
37,800
152,000
118,400
28,400
44.100
611.100
861.700

77,600
16,200
34,500
22,000
76,300
226,600
plant 3
Maximum
230,500
37,800
470,000
118,400
49,800
77,500
984,000
1,387,400

124,900
32,300
55,500
22,000
122,900
357,500
                                 a Capital cost includes  subcontract  or  overhead  and  profit  and contractor  or material  markup.

                                 b Total  capital  cost  includes  construction  fee,  contingency, engineering,  and  startup  (41%).

                                 c Maintenance and overhead  are 5%  and 45£  of total capital cost,  respectively.

                                 d Steam at $18.3/My and  electricity  at  $0.05  kMh.

                                 e For 1.5 h/day  at $39.69/h.

                                 f Capital recovery factor for  20-yr  lifetime  at  6.2%

-------
         Appendix C
Economic Impact Analysis

-------
                                 APPENDIX C
                               ECONOMIC IMPACT

      This appendix addresses  the economic impacts  of the regulatory alterna-
 tives for coke-oven by-product plants.   It provides  an updated version of
 Chapter 9 of the background information document (BID),  Benzene Emissions
 from Coke By-Product Recovery Plants.   This appendix includes  revised
 estimates of the economic  impacts  of  the regulatory  alternatives  and more
 recent information on the  state of the coke industry.  Where possible,  data
 are  updated  to  1984.
      Section C.I presents  a profile of the coke  industry.   Section  C.2
 contains  a reanalysis of the  impacts  of the regulatory alternatives.   These
 alternatives are outlined  in  Table C-l.   These impacts are  measured against
 the  baseline state of control  for  all  sources.   Section  C.3 presents  poten-
 tial  socioeconomic and inflationary impacts.
 C.I   INDUSTRY PROFILE
 C.I.I   Introduction
      Coke  production  is a part  of  Standard  Industrial  Code  (SIC) 3312--
 Blast  Furnaces and Steel Mills.  Coke  is  principally used in the production
 of steel and ferrous  foundry products, which are also  part  of the output  of
 SIC 3312.  Thus, coke  is both produced and principally consumed within
 SIC 3312.  Furthermore, many producers of  furnace coke are  fully integrated
 iron- and  steel-producing companies.  Any  regulation on coke production is
 expected to  have some  impact on the entire blast furnaces and steel mills
 industry with special emphasis on coke producers.
     This profile has two purposes:  (1) to provide the reader with a broad
overview of the industry and (2) to lend support to an economic analysis by
assessing the appropriateness of various economic models to analyze the
                                   C-3

-------
                 TABLE C-l.   COKE BY-PRODUCT PLANT CONTROL OPTIONS'
                                              Control  option
     Emission source
Regulatory Alternative    Regulatory Alternative
         II                       III
Direct water final cooler
Tar bottom final cooler
Tar decanter, flushing-
 liquor circulation tank,
 tar-intercepting sump
Tar storage tanks and
 dewatering tanks
Light-oil decanter-
 condenser, wash-oil
 circulation tank,
 wash-oil decanter
Excess ammonia-liquor
 storage tanks
Light-oil tanks and BTX
 storage tanks
Benzene storage tanks
Light-oil sump
Pump seal leaks
Valve  leaks
Exhauster leaks
Pressure relief device
 leaks
Sampling connection system
 leaks
Open-ended  line leaks
Tar bottom final cooler
Coke-oven gas-
 blanketing system
Coke-oven gas-
 blanketing system
Coke-oven gas-
 blanketing system
Coke-oven gas-
 blanketing system
Coke-oven gas-
 blanketing system
Wash-oil scrubber
Cover
Monthly inspection
Monthly inspection
Quarterly inspection
Rupture disc  system

Closed-purge  system

Cap  or plug
Wash-oil final cooler
Wash-oil final cooler
Coke-oven gas-
 blanketing system
Coke-oven gas-
 blanketing system
Coke-oven gas-
 blanketing system
Coke-oven gas-
 blanketing system
Coke-oven gas-
 blanketing system
Wash-oil scrubber
Cover
Monthly  inspection
Monthly  inspection
Quarterly inspection
Rupture  disc  system

Closed-purge  system

Cap  or plug
 aThese  regulatory  alternative control options differ from the proposed
  regulations.
                                        C-4

-------
 industry.   Further, the profile provides some of the data necessary to the
 analysis itself.
      The industry profile comprises six major sections.   The remainder of
 this introduction, which constitutes the first section,  provides a brief,
 descriptive,  and largely qualitative look at the industry.   The remaining
 five sections of the profile conform with a particular model  of industrial
 organizational  analysis.   This  model maintains that an industry can be
 characterized by its basic conditions,  market structure,  market conduct,
 and market performance.
      The basic  conditions in the industry,  discussed in  the  second and
 third sections  of this  profile,  are believed to be  major determinants  of
 the prevailing  market structure.   Most  important of these basic conditions
 are supply conditions,  which are largely technological  in nature,  and
 demand conditions,  which  are determined by  the attributes of  the products
 themselves.
      The market structure and market conduct of the blast furnaces  and
 steel  mills industry are  examined  in the  fourth section.   Issues  addressed
 include geographic  concentration,  firm  concentration,  integration,  and
 barriers to entry.   Market  structure is  believed  to  have  a major  influence
 on  the  conduct  of market  participants.  Market  conduct is the price and
 nonprice behavior of sellers.  Of  particular  interest  is  the degree to
 which  the  industry  pricing  behavior  can be approximated by the  competitive
 pricing model,  the  monopoly  pricing  model, or  some model  of imperfect
 competition.
     The fifth  section of the industry profile  addresses market perform-
 ance.   The historical record of the  industry's  financial performance is
 examined, with  some  emphasis on its comparison with other industries.   The
 sixth section of the  industry profile presents a discussion of  industry
 trends  for the  coke  and steel sectors.  The seventh section discusses
market  behavior.
     C.I.1.1  Definition of the Coke Industry.  Coke production is a part
of SIC 3312—Blast Furnaces and Steel Mills, which includes establishments
that produce coke and those that primarily manufacture hot metal, pig iron,
silvery pig iron,  and ferroalloys from iron ore and iron and steel scrap.
                                    C-5

-------
Establishments that produce steel  from pig iron,  iron scrap,  and steel
scrap and establishments that produce basic shapes such as plates,  sheets,
and bars by hot rolling the iron and steel also are included  in SIC 3312.1
The total value of shipments from SIC 3312 in 1982 was $36,931,900,0002 and
an approximate value for total coke production in 1982 was $3,220,Oil,000,3
or less than 10 percent of the total value of shipments.
     Coke is produced in two types of plants:  merchant and captive.
Merchant plants produce coke to be sold on the open market, and many are
owned by chemical or other companies. The majority of coke plants in the
United States are captive plants that are vertically integrated with iron
and steel companies and use coke in the production of pig iron.  At the end
of 1984, 15 plants were merchant and 36 were captive, and merchant plants
accounted for only 12 percent of total coke production.4 5  For the economic
analysis, it is assumed that more than one plant may exist at a single
location.
     C.I.1.2  Brief History of the Coke Industry in the Overall Economy.
Traditionally, the value of coke produced in the United States has con-
stituted less than 1 percent of the gross national product (GNP).6 7
During most of the 1950's, coke production was about 0.30 percent of GNP,
and during the 1960's and until the mid-1970's, coke production was only
about 0.20 percent or less of GNP.  However, in 1974, coke production as a
percent of GNP rose to above 0.30 percent.  This trend continued for the
next 2 years.  By 1982, coke production was about 0.1 percent of GNP.3 8
     Previously, U.S. coke exports had been greater than imports, but that
trend has fluctuated.  The values of all  U.S.  imports and exports and U.S.
coke imports and exports are shown in Table C-2.  From 1950 to 1972, coke
exports were much greater than coke  imports, but after 1973, this trend was
reversed.   In 1982 and 1983, exports again exceeded imports.   Data for the
second quarter of 1984 indicate that coke imports are again on the rise.
Imports  for the  first two quarters of 1984 totaled 247,604 megagrams (Mg)
compared to 6,874 Mg for the  same period  in  1983, and to 32,000 Mg for all
of 1983.16  Exports for the  first two quarters of 1984 and 1983 were
307,540  Mg, and  300,283 Mg,  respectively, and  they were 603,288 Mg for all
of 1983.1S
                                     C-6

-------
                                TABLE C-2.   COKE INDUSTRY FOREIGN TRADE3 9 10 »» 12 13 »«
Year
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
=====
a,.
Total U.S. imports,
IO9 $a
8.9
11.0
10.7
10.9
10.2
11.4
12.6
13.0
12.8
15.2
14.7
14.7
16.4
17.1
18.7
21.4
25.5
26.8
33.2
36.0
39.9
45.6
55.8
70.5
103.7
98.0
124.0
151.9
176.0
212.0
249.7
265.1
243.9
258.0
Coke imports
for consumption,
IO6 $a
5.3
1.9
4.5
1.7
1.3
1.4
1.5
1.5
1.6
1.4
1.5
1.5
1.9
2.0
1.5
1.4
1.8
1.7
1.9
3.4
3.5
5.0
4.6
39.3
193.2
156.5
111.1.
137. 9^
410.9°
340. l£
52'°b e
9'2h
1.9b
Coke imports
as a share of
total imports,
0.06
0.02
0.04
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.06
0.19
0.16
0.09
0.09
0.23
0.16
0.02
0.02
0.004
0.0007
U.S. exports,
IO9 $
10.3
15.0
15.2
15.8
15.1
15.5
19.1
20.9
17.9
17.6
20.6
21.0
21.7
23.3
26.5
27.5
30.3
31.5
34.6
38.0
42.5
43.5
49.4
71.4
98.3
107.1
114.7
120.8
142.1
184.5
224.2
237.0
212.3
200.5
Coke exports,
IO6 $a
6.2
17.7
13.7
9.3
6.2
8.2
11.5
14.4
7.1
8.7
6.9
8.2
7.4
8.3
10.1
16.3
23.4
16.5
18.6
38.5
78.9
44.8
30.7
33.1
43.6
74.7
66. 7b c
68.9b'd
15.0°
12-9?
11. 3*
13. 8^
8.3b
Coke exports
as a share of
total exports,
0.06
0.12
0.09
0.06
0.04
0.05
0.06
0.07
0.04
0.05
0.03
0.04
0.03
0.04
0.04
0.06
0.08
0.05
0.05
0.10
0.19
0.10
0.06
0.05
0.04
0.07
0.06
0.06
0.05
0.01
0.005
0.005
0.01
0.004
 See Product SIC (331210) in References 11-13.
 Defined as "Pitch coke, coke of coal,  lignite,  or peat."
eDefined as "Coal coke,  calcined and not calcined "
 Cumulative through November 1981.   Annual  cumulative value not available.

-------
     The same pattern applies to the percentages of coke imports and exports
within total U.S. imports and exports.   From 1950 to 1972, coke exports
were a larger percentage of total U.S.  exports than coke imports were of
total U.S.  imports.   Again, from 1973 to 1981, this trend reversed, and
coke imports were a larger proportion of total U.S. imports than coke
exports were of total U.S. exports.   Percentage shares of exports were
greater than imports in 1982 and 1983.
     U.S. coke production always has been a substantial portion of world
coke production.  This share has decreased during the past 30 years, as
indicated in Table C-3.  From 1950 to 1977, world coke production generally
increased while U.S. coke production decreased.  This trend explains the
decline in the U.S.  percentage of world coke production.
     C.I.1.3  Size of the Iron and Steel Industry.  The value of shipments
of SIC 3312 has increased since 1960.  There have been a few fluctuations
in this growth; for example, as shown in Table C-4, the 1965 value of
shipments of SIC 3312 was the highest value between 1960 and 1972.  Since
1972, the value of shipments has remained around $30 million, with the
highest value being $35 million (1972 dollars) in 1974.  After reaching
another peak of $34 million (1972 dollars), the value of shipments declined
to a 23-year low of about $18 million (1972 dollars).  This result
reflected conditions in the steel industry.   In 1982, the steel industry
sustained record financial losses close to $3.2 billion (1982 dollars).23
In 1983, an additional $3.6 billion was lost.24
     For SIC 3312, Table  C-5 shows the value  added by manufacture, the
total number of  employees, and the value added per employee.  Current and
constant (1972)  dollar figures are included.  Both the total value added by
manufacture and  the value  added per employee  peaked  in 1974, the  same year
in which the value  of  shipments  for this industry  was the highest.  The
increasing  value added per employee might  indicate that this industry is
changing to a more  capital-intensive production process.  This  aspect is
discussed  in Section C.I.6.
C.I.2   Production
      C.I.2.1  Product  Description.  Two types of  coke  are produced:   furnace
coke and foundry coke.   Furnace  coke is used  as a fuel  in blast furnaces;
                                     C-8

-------
               TABLE C-3.   COKE PRODUCTION IN THE WORLD6 17 1S
Year
World production,
     106 Mg
U.S.  production,
    ID6 Mg
                                                             U.S.  production
                                                              as a share of
                                                            world production,
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978b
1979°
182.3
204.1
208.9
225.6
211.5
242.3
256.8
266.1
255.0
260.4
279.7
272.0
272.9
281.7
298.5
310.3
310.4
303.9
315.8
335.8
350.5
342.7
340.5
365.8
367.4
363.3
367.2
373.5
364.7
341.0
65.9
71.9
62.0
71.5
54.4
68.3
67.6
69.0
48.6
50.7
51.9
46.9
47.1
49.3
56.4
60.7
61.2
58.6
57.8
58.8
60.3
52.1
54.9
58.4
55.9
51.9
52.9
48.5
44.5
48.0
36.1
35.2
29.7
31.7
25.7
28.2
26.3
25.9
19.1
19.5
18.6
17.2
17.3
17.5
18.9
19.6
19.7
19.3
18.3
17.5
17.2
15.2
16.1
16.0
15.2
14.3
14.4
13.0
12.2
14.1
Oven and beehive coke combined.

Information on world coke production not available after 1979.
                                    C-9

-------
TABLE C-4.   VALUE OF SHIPMENTS, SIC 33122 19 20 21 22
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
Current dollars,
106
15,738.8
14,873.3
15,571.6
16,418.0
18,840.1
20,841.7
21,193.9
19,620.6
21,161.1
22,299.0
21,501.6
21,971.3
23,946.7
30,365.5
41,671.7
35,659.8
39,684.1
41,897.8
49,055.4
55,695.8
50,303.9
57,472.9
36,931.9
1972 Dollars,
106
22,981.7
21,468.4
22,071.7
22,933.4
25,914.9
28,043.2
27,610.6
24,829.9
25,628.1
25,713.8
23,535.0
22,882.0
23,946.7
28,700.9
35,917.7
28,038.8
29,643.8
29,645.4
32,879.0
34,358.9
28,244.8
29,473.3
17,884.7
                        C-10

-------
TABLE C-5.  VALUE ADDED, SIC 33122 19 20 21 22
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
Value added by
Current dollars
106
6,844.4
6,546.3
6,620.9
7,506.4
8,479.6
9,379.8
9,643.6
8,910.1
9,275.8
9,853.2
9,350.5
9,563.1
10,304.7
12,769.4
17,425.8
13,356.2
14,755.5
15,021.4
19,085.7
21,039.0
18,632.2
20,100.2
manufacture
1972 dollars,
106
9,965.6
9,449.0
9,384.7
10,485.3
11,663.8
12,620.8
12,563.3
11,275.8
11,233.9
11,362.1
10.234.8
9,959.5
10,304.7
12,069.4
15,019.7
10,501.8
11,022.3
10,628.6
12,792.0
12,979.0
10,461.6
10,307.8
Employees,
103
550.0
503.4
502.2
500.5
532.9
565.4
559.4
533.1
533.1
537.7
526.5
482.2
469.1
502.1
518.0
451.3
451.9
441.4
443.5
451.2
402.9
390.3
Value added
per employee —
1972 dollars,
103
18.1
18.8
18.7
20.9
21.9
22.3
22.5
21.2
21.1
21.1
19.4
20.7
22.0
24.0
29.0
23.3
24.4
24.1
28.8
28.8
26.0
26.4
                   C-ll

-------
foundry coke is used as a fuel  in the cupolas of foundries.   Coke also is
used for other miscellaneous processes such as residential  and commercial
heating.  In 1983, only 3 percent of all  coke used in the United States was
used for these miscellaneous purposes, 92 percent was used in blast furnaces,
and the remaining 5 percent was used in foundries.25  Time-series data for
the percent of total U.S. consumption attributable to each use from 1950 to
1980 are shown in Figure C-l.
     C.I.2.2  Production Technology.  Coke is typically produced from coal
in a regenerative type of oven called the by-product oven.   The type of
coal used in coke production and the length of time the coal is heated
(coking time) determine the end use of the coke.  Both furnace and foundry
coke usually are obtained from the carbonization of a mixture of high- and
low-volatile coals.  Generally, furnace coke is obtained from a coal mix of
10 to 30 percent low-volatile coal and is coked an average of 18 hours, and
foundry coke is obtained from a mix of 50 percent or more low-volatile coal
and is coked an average of 30 hours.
     The first by-product oven in the United States was built in 1892 to
produce coke and to obtain ammonia to be used in the production of soda
ash.  In such ovens, the by-products of carbonization (such as ammonia,
tar, and gas) are collected  instead of being emitted into the atmosphere as
they were  in the older, beehive ovens.
     The total amount of coke that can be produced each year is restricted
by the  number of ovens in operation for that year, and not all ovens are in
operation  all of the time.   Oven operators try  to avoid closing down a
group of ovens for  any reason because of the time and energy lost while the
ovens cool and reheat and because of the oven deterioration that results
from cooling and reheating.  However, it is estimated that, at any time,
approximately 5 to  10 percent of existing coke-oven capacity is out of
service  for rebuilding or repair.28   In a report written for the Department
of Commerce, Father William  T. Hogan estimated  the potential annual maximum
capacity of U.S.  oven coke plants as of July 31, 1979.29  Hogan assumed
that almost 10 percent of his estimate of total capacity would be out of
service at any given time; therefore, he subtracted the out-of-service
capacity from total capacity to  obtain maximum  annual capacity.  The actual
                                    C-12

-------
o
I
         O
         Ul
         C/l
         D
         Ul
         X
         O
         O
         u.
         O
         U)
         O
         tc
         ui
         a.
92



00



QB



86



84
                              A
                     /
                     »
    /FURNACE
12
10
- > OTHER USES
               v>v
                                                                           A
                                                                     \  /  FURNACE
                                                                          FOUNDRY
              J—i	l
                                                                              OTHER

                                                                              USES
             50   62   54   56   68   60  62   64   66   68  70  72  74   76   70   00
                                          YEAR
                      Figure C-1. Uses of oven coke as percents of total coke consumption.6-26-27

-------
number of ovens that are out of service in a given year varies greatly.   In
December 1983, 112 of 6,978 ovens,  or 1.6 percent, were being rebuilt or
repaired, and annual capacity totalled 35,575,000 Mg.30  In November 1984,
1,756 of 8,204 ovens, or 21.4 percent, were out of service, and annual
capacity totalled 51,180,000 Mg.5  Table C-6 presents the data for November
1984.
     In actuality, ovens that are removed from service and placed on "hot
idle" status are those likely to be returned to production in the short
term.  Ovens that are placed on "cold idle" status are less likely to be
returned to service and, historically, have not been returned to service.
The capacity of these ovens is included in a plant's total capacity for
bookkeeping purposes even though the ovens may be scheduled for demoli-
tion.31
     Within the limits of the number of ovens available for coking, both
furnace and foundry coke production levels vary.  Some ovens that produce
furnace coke can be switched to produce foundry coke by changing the coal
mix and increasing the coking time.  Furthermore, some ovens that produce
foundry coke could be changed to produce furnace coke by changing the coal
mix and decreasing the coking time.  Also, some variation in the combina-
tion of flue temperature and coking time is possible for either type of
coke.  A shorter coking time results in greater potential annual produc-
tion.
     C.I.2.3  Factors of Production.  Table C-7 provides a typical labor
and materials cost breakdown for furnace coke production.  Coal is the
major material input in the production of coke.  In 1979, greater than 61
percent of the coal received by coke plants was from mines that were company
owned or affiliated.33  In this same year, 14 States shipped some coal to
coke plants outside their borders.34  Of the coal received by domestic coke
plants, over 81 percent came from West Virginia, Kentucky, Pennsylvania,
and Virginia.34  Any potential adverse impact on the coke  industry probably
will have some impact in these States.  A total of 33.6 million Mg of
bituminous coal was carbonized in 1983.3S
     Table C-8 shows employment in the by-product coke industry from 1950
to 1970 and the percentage of total SIC 3312 employees in  the by-product
coke industry.  This table shows decreasing employment in  the by-product
                                     C-14

-------
         TABLE C-6.  MAXIMUM ANNUAL CAPACITY OF OVEN COKE PLANTS
                  IN THE UNITED STATES IN NOVEMBER 1984s

In existence
Furnace plants
Foundry plants
Total
Out of service3
Furnace plants
Foundry plants
Total
In operation
Furnace plants
Foundry plants
Total
Number of
batteries

105
35
140

(25)
(2)
(27)

80
33
113
Number of
ovens

6,638
1,566
8,204

(1,646)
(110)
(1,756)

4,992
1,456
6,448
Capacity,
Mg

44,810,000
6,370,000
51,180,000

(9,828,000)
(402,000)
(10,230,000)

34,982,000
5,968,000
40,950,000
Batteries and ovens down for rebuilding and repair,  or on cold idle prior
to permanent closure.

Defined as "online" or "on hot idle."
                                  C-15

-------
     TABLE 07.   TYPICAL COST BREAKDOWNS:   FURNACE COKE PRODUCTION AND
                    HOT METAL (BLAST FURNACE) PRODUCTION32

Furnace coke production	Percent of cost

  Labor and materials

    Coking coal                                                      77.1
    Coal transportation                                              9.4
    Labor (operation and maintenance)                                6.6
    Maintenance materials                                            6.9

    Total labor and material costs                                 100.0

Hot metal production

  Charge metal lies
    Iron ore
    Agglomerates
    Scrap
  Fuel inputs
    Coke
    Fuel oil
  Limestone fluxes
  Direct labor
  Maintenance
  General expenses
    Total labor and material costs                                 100.0
                                    C-16

-------
            TABLE C-8.   EMPLOYMENT IN THE BY-PRODUCT COKE INDUSTRY36
Year
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971a
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
Number of employees
20,942
22,058
21,919
21,011
17,944
19,595
19,318
19,203
15,654
15,865
15,779
13,106
12,723
12,696
13,021
14,003
13,745
13,662
14,136
13,617
13,997
11,955
11,127
11,121
11,207
12,109
11,047
10,196
10,578
10,477
9,673
8,846
6,778
Percentage of all
employees in SIC 3312
NA
NA
NA
NA
NA
NA
NA
NA
3.06
3.13
2.87
2.60
2.53
2.54
2.44
2.48
2.46
2.56
2.65
2.53
2.66
2.48
2.37
2.21
2.16
2.68
2.44
2.31
2.38
2.32.
2.40
2.27
2.28
NA = Not applicable.
a
 Figures for 1971-1982 are estimates.  See text for more detail
                                    C-17

-------
coke industry.   A similar decline in employment has occurred in SIC 3312.
Unfortunately,  employment data for the by-product coke industry are not
available after 1970; however, these figures can be estimated by regressing
employment in the by-product coke industry on total iron and steel  industry
employment and on the ratio of coke used in steel production.*  These
estimates are also shown in Table C-8.
C.I.3  Demand and Supply Conditions
     Domestic consumption of coke from 1950 to 1980 is graphed in Figure C-2.
In the early 1950's, the amount of coke consumption was fairly large; an
average of 65 million Mg was consumed annually between 1950 and 1958.  The
late 1950's and early 1960's showed a sharp decrease in coke consumption,
with an average of only 48 million Mg consumed annually.  Domestic con-
sumption of coke increased during the mid-19601s to mid-19701s to an annual
figure of 57 million Mg, but it did not reach the 1950 to 1957 level.  The
late 1970's showed another slump in coke consumption.
     The variation in coke consumption shown in Figure C-2 has both cyclic
and trend components.  The demand for coke is derived from demands for iron
and steel products, and these demands are sensitive to the performance of
the overall economy.  Cycles in coke demand are linked to cycles in aggre-
gate demand or cycles in demand for particular products such as automobiles.
     The trend component in coke consumption results from changes in blast
furnace production techniques.  Coke is used as a fuel in blast furnaces,
but it is not the only fuel that can be used.  Coke-oven gas, fuel oil, tar
and pitch, natural gas, and blast furnace gas have all been used as supple-
ments to coke in heating the blast furnaces.  The increased use of these
supplemental fuels over the past 20 years has caused the amount of coke
used per ton of pig iron produced (the coke rate) to decrease.  Other
causes of the decline in coke rate are increased use of oxygen in the blast
furnaces and use of higher metallic content ores.  Table C-9 shows U.S. pig
iron production, coke consumed in the production of pig iron, and the coke
rate for 1950 to 1983.  (Data limitations make it difficult to calculate
the foundry coke rate in cupola production.)
^Regressions performed by Research Triangle Institute (RTI) in 1980 and 1985.
                                    C-18

-------
o
i
              70
              65
o
o
u.
O
to
5
              60
          g  55
          tu
O
              50
              45
              35
                  _L
              JL
_L
                                             JL
                                        J-
                      J_
4-
                                                                                   JL
                  50   52   54   56    58   60    62
                                             64   66

                                             YEAR
                                                                                    J_
                                68    70   72   74    76   78   80
                                     Figure C-2. U.S. apparent consumption of coke.6-26

-------
TABLE C-9.   COKE RATE3 18 2S 37 38
Year
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
Pig iron production,
103 Mg
58,514
63,756
55,618
67,906
52,570
69,717
68,067
71,128
51,851
54,622
60,329
58,834
59,546
65,173
77,527
80,021
82,815
78,744
80,529
. 86,186
82,820
73,829
80,628
91,915
86,616
72,322
79,788
73,931
79,552
Coke used in
blast furnaces,
103 Mg
51,403
55,362
49,386
58,880
46,861
60,675
58,279
60,861
42,898
44,107
46,462
42,855
42,298
44,596
51,076
53,576
54,653
51,300
51,399
55,065
54,754
48,269
50,214
54,791
51,154
44,375
47,678
44,292
47,889
Coke rate
0.86
0.87
0.89
0.87
0.89
0.87
0.86
0.86
0.83
0.81
0.77
0.73
0.71
0.68
0.66
0.67
0.66
0.65
0.64
0.64
0.66
0.65
0.62
0.60
0.59
0.61
0.60
0.60
0.60
                                          (continued)
                C-20

-------
                            TABLE C-9 (continued)
                                             Coke  used  in
            Pig iron production,             blast  furnaces,
Year              103 Mg                        103  Mg             Coke  rate
1979
1980
1981
1982
1983
78,926
62,325
66,951
39,282
46,267
45,862
37,583
37,832
21,918
25,009
0.58
0.60
0.56
0.56
0.54
                                   e-2i

-------
     Recently, there has been some concern about the ability of the United
States' coke-making capacity to support domestic steel  production—the
major source of coke demand.   The study conducted by Hogan and Koelble of
the Industrial Economics Research Institute at Fordham University indicates
that, in 1978, U.S. production of coke was 14.1 percent below domestic
consumption.39  Imports increased dramatically in that same year.  Hogan
and Koelble attributed this decline in coke production to the abandonment
of coke ovens for environmental reasons and predicted a severe coke shortage
by 1982.40  This prediction was disputed in a Merrill Lynch Institutional
Report by Charles Bradford.41  The Bradford report attributed the lack of
adequate U.S. coke production in 1978 to two factors:  (1) a coal miner's
strike, which caused the drawing down of stocks of coke when they should
have been increasing, and (2) the premature closing because of U.S. Envi-
ronmental Protection Agency (EPA) regulation of some coke ovens that
normally would have been replaced before they were closed.41  The Bradford
report stated that a survey of U.S. steel producers revealed that all of
the major steel producers were or soon would be self-sufficient with regard
to coke-making capacity.42  The Bradford explanation of 1978 coke imports
seems more reasonable because 1979 coke imports decreased about 1.6 million
Mg compared to the 1978 level.
     The following values describe the situation in the 1980s with respect
to production, imports, and apparent consumption of coke (thousand mega-
grams).16
     Year     Production     Imports     Consumption     Distributor Stock
                                                               7,009
                                                               5,556
                                                               7,141
                                                               4,024
                                                               2,776
1980
1981
1982
1983
1984a
41,851
38,815
25,506
23,413
14,446
598
478
109
32
248
37,447
39,975
23,384
27,080
14,886
     aTwo quarters of 1984

     Production is less than apparent consumption in 1981, 1983, and 1984.
For each of these years, stocks and imports more than accommodate the
shortfall.  Coke producers were operating at 80 percent of total capacity
                                     C-22

-------
  in November 1984.5  Thus,  it is  unlikely that major shortages  will  develop
  in the near future.
  C.I.4  Market Structure
       Market power,  the degree  to which  an individual  producer  or  groups  of
  producers  can control market price,  is  of particular  economic  importance.
  Market structure  is  an important determinant  of market power.   Pricing
  behavior is relevant to the  choice of the methodology used  in  assessing  the
  potential  impacts  of new regulations.   It is  important to determine  if the
  competitive pricing  model  (price equal  to  marginal  cost) adequately  des-
  cribes  pricing behavior for  coke producers.
       Any analysis of market  structure must consider the characteristics of
  the industry.  This  analysis addresses  the number of  firms producing coke;
  the concentration of production  in specific firms; the degree of inte-
  gration in  coke production; the  availability of substitutes for coke; and
  the availability of  substitutes  for the commodities for which coke is an
  input to production.   Also, some information on past pricing in the coke
  industry is presented.   These topics  will be considered together with
 financial  performance (Section  C.I.5) and trends  (Section C.I.6) in asses-
 sing market behavior (Section C.I.7).
      C.I.4.1  Concentration Characteristics and Number of Firms.   This
 section describes  various  concentration  measures  that  can be computed for
 the furnace and foundry coke  industries.   Normally,  concentration  ratios
 are used as an indication of  the  existence of  market power.   Although
 concentration ratios  are a  useful tool for describing  industry  structure,
 concentration should  not be used  as an exclusive measure  of  market power.
 Many other  factors  (e.g., availability of substitutes, product  homogeneity,
 ease of market entry) determine a firm's  ability to  control  market price.
     As  of  November 1984, 23  companies operated by-product coke ovens.5 43
 Twelve  companies are  integrated iron and  steel producers; 11 companies are
 merchant firms.  These companies  owned and operated a  total of 51 coke
 plants; 36  of  these plants were captive and 15 of them were merchant.  A
 list of these  companies,  their plant locations, the major uses of coke at
 each plant,   and plant coke capacities  is given in  Table C-10.  A plant site
may include more than one  complete plant.
                                   C-23

-------
                                                 TABLE C-10.  COKE PLANTS IN THE UNITED STATES, November 1984s
o
 i
no
Company name
Armco, Inc.
Bethlehem Steel Corp.
Rouge Steel
Inland Steel Co.
Interlace, Inc.
The LTV Steel Corp.
Lone Star Steel Co.d
National Steel Corp.e
Weir ton Steel Corp.
New Boston Coke Corp.
Plant location
Ashland, KY h
Middletown, OH (2)°
Bethlehem, PA
Burns Harbor, IN
Lackawanna, NY
Sparrows Point, MO
Dearborn, HI
E. Chicago, IN (3)
Chicago, IL
Aliquippa, PAC
Cleveland, OH (2)
E. Chicago, IN
Gadsden, AL
Pittsburgh, PA
S. Chicago, IL
Thomas, AL
Warren, OH
Lone Star, TX
Granite City, IL
Detroit, MI
Brown's Island, WV
Portsmouth, OH
Classification
of plant
Captive
Captive
Captive
Captive
Merchant
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Major uses of coke
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Coke capacity,
103 Mg/yr
963
1,776
2,253
1,790
1,292
3,506
778
3,715
582
1,218
1,760
948
758
1,792
563
315
945
507
868
1,397
1,097
364
                        Footnotes  at end of table.
                                                                                                                          (continued)

-------
                                                                    TABLE  C-10   (continued)
o
en
Company name
U.S. Steel Corp.





Wheeling-Pittsburgh9
Steel Corp.
Jim Walter Corp.
Koppers Co. , Inc.



Shenango, Inc.
Alabama By-Products
Corp.

Carondelet Coke Corp.

Chattanooga Coke and.
Chemical Co. , Inc.
Citizens Gas and Coke
Utility
Plant location
Clairton, PA (4)
Fairfield, AL
Fairless Hills, PA
Gary, IN
Loral n, OH
Provo, UT
E. Steubenville, WV
Monessen, PA
Birmingham, AL
Erie, PA

Toledo, OH
Woodward, AL
Neville Island, PA
Tarrant, AL

Keystone, PA
St. Louis, MO

Chattanooga, TN
Indianapolis, IN
Classification
of plant
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Merchant
Merchant

Merchant
Merchant
Merchant
Merchant

Merchant
Merchant

Merchant
Merchant
Major uses of coke3 	
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace, foundry
Foundry, other
industrial
Foundry
Blast furnace, foundry
Blast furnace, foundry
Foundry, .
other industrial"
Foundry
Foundry,
other industrial
Foundry,
other industrial
Foundry
Coke capacity
103 Mg/yr
	 S 	 a/ J 	
5,294
1,822
916
4,228
1,496
1,160
1,509
490
499
207

157
563
521
583

402
330

130
477
                       Footnotes at end of table
                                                                                                                          (continued)

-------
                                                                    TABLE C-10  (continued)
o
i
ro
en
Company name
Detroit Coke Corp.
Empire Coke Co.
Indiana Gas and
Chemical Corp.
Tonawanda Coke Corp.
Plant location
Detroit, MI
Holt, AL
Terre Haute, IN

Buffalo, NY
Classification
of plant
Merchant
Merchant
Merchant

Merchant
Major uses of coke
Foundry
Foundry
Foundry,
other industrial
Foundry
Coke capacity
103 Mg/yr
617
161
132

299
aAn end use is considered a major use if it is  at least 20 percent of the plant's  total  distribution of coke.
bNumbers in parentheses indicate the number of  plants  at that location.   If no number is indicated,  only one
 plant exists at that location.
CLTV announced its intention in May 1985 to reduce production of steel  at the Aliquippa, Pennsylvania,  plant.
 The plant may convert to a cold-idle status eventually.
^Northwest Industries, Inc., the parent company of Lone Star Steel, announced in April  1985  its  merger  with
 Farley Industries.
eA merger between National Intergroup,  Inc., the parent company of National Steel  Corp., and Bergen  Brunswig
 Corp. fell through in April 1985, 2 weeks before its  scheduled date.   Some market consultants feel  that
 National Intergroup, Inc., is now a potential  target  for corporate raiders.
fMcLouth Steel Corp., the parent company of New Boston Coke Corp., is operating under Chapter 11 filed  in 1981.
gWheeling-Pittsburgh Steel Corp. filed for Chapter 11  in April 1985.
Residential and commercial heating included in other  industrial category.
^Chattanooga Coke and Chemical Co., Inc., is operating under Chapter 11 filed in March 1984.

-------
       Reported capacities in Table C-10 are maximum,  nominal  figures,  which
  do not include any allowance for outage like  that determined for the  overall
  industry in  Table  C-6.   All  but  one  of the largest plants  are captive,  and
  most  of the  merchant  plants  have very  small capacities.  Furnace coke
  production is concentrated  in  captive  plants.  Virtually all  of  the coke
  used  in foundries  and in other industries  was produced by  merchant plants.
  If coke plant sites were ranked  according  to capacity, the top 5 plant
  sites  and top 10 plant sites would have 37.1 percent and 54.6 percent of
  total  coke capacity,  respectively.
      By-product coke  plants are  concentrated in the States bordering  on the
  Ohio River, probably  because of  the coal in that area.   Pennsylvania  contains
  12 plants, and Ohio and  Indiana  each have 8 plants.5
      Table C-ll divides  the United States into 11 coke-consuming and  coke-
 producing regions and shows the amount of coke produced in each  region and
 the locations of coke consumption in 1977.   Most of the regions produce the
 bulk of the coke they consume;  only three regions produced  less than  80
 percent of their own consumption, and only one produced more  than it  needed
 for its own consumption.   Transportation of coke  across long  distances is
 avoided whenever possible to reduce breakage of  the product into  smaller,
 less valuable pieces and  to  minimize  freight charges.46
      The concentration of production  or capacity  in specific  firms may have
 economic importance.   Table  C-12  presents  the  percent of total  capacity
 owned  by the  largest 4 (of 23)  firms.   The  four-firm concentration ratio
 for the coke  industry  has increased over the years.  In 1959,  the four-firm
 concentration  ratio  was 53.5  (the top four  firms owned 53.5 percent of
 total capacity)4?;  in  1984 it was  69.9  percent.  Consolidation of the
 industry through mergers,  acquisitions, and closures has encouraged this
 trend.
     In  the preceding  discussion,   furnace and foundry coke production  are
 considered jointly.   However, each existing coke battery may be considered
 a furnace or foundry coke producer, based on the battery's primary use.
 Separate capacity-based concentration ratios for the two types of coke are
calculated based on  this allocation.   The 1984  four-firm concentration
ratio for furnace coke is  75.4;  the 1984 four-firm ratio for foundry  coke
is 65.2.5
                                    C-27

-------
                                                        TABLE  C-ll.   INTERREGIONAL COKE SHIPMENTS IN 1977<5
                                                                         (103 megagraras)
 i
l\3
00
Consuming region
Producing region
Alabama 2
California,
Colorado, Utah
Maryland, New York
Illinois
Indiana
Kentucky, Missouri,
Tennessee, Texas
Michigan
Minnestoa, Wisconsin
Ohio
Pennsylvania
Virgina,
West Virginia
TOTAL 2
AL
,228

0
0
0
0

14
0
0
0
9

0
,251
CA, CO
UT
27

2,668
3 4
0
5

18
0
6
4
0

0
2,731 4
MD,
NY
10

0
,392
0
0

0
0
0
0
51

0
,453
IL
81

0
123
1,424
69

15
0
269
138
1,241

0
3,360
IN
112

0
0
0
7,594

5
7
70
366
134

0
8,288
KY, MO,
TN, TX
465

0
22
0
35

928
1
1
379
3

8
1,842
MI
195

0
88
0
97

125
2,639
61
260
52

412
3,929
MN,
WI
7

0
0
0
11

0
0
158
0
0

0
176
OH
114

0
6
0
62

13
6
5
6,356
1,370

0
7,932
PA
1

0
8
0
3

0
0
1
2
10,257

214
10,556
VA.
wv
51

0
0
0
0

20
0
0
12
3

2,465
2,551
Total
3,361

2,668
4,642
1,424
7,876

1,138
2,653
571
7,517
13,120

3,099
48,069

-------
TABLE C-12.   PERCENT OF COKE CAPACITY OWNED BY TOP FIRMS
                    NOVEMBER 19845
Firm
U.S. Steel, Inc.
Bethlehem Steel Corp.
The LTV Steel Corp.
Inland Steel Co.
Sum of largest four firms
Capacity,
103 Mg
14,916
8,841
8,299
3,715
35,771
Percent of
total capacity
29.14
17.27
16.22
7.26
69.89
                     C-29

-------
     Concentration in the steel  industry has economic relevance because a
large fraction of all furnace coke is produced by integrated iron and steel
companies.   Historically, the eight largest steel producers have been
responsible for approximately 75 percent of industry production.  However,
from 1950 to 1976, the share of production attributable to the top four
firms declined from 62 percent to 53 percent.48  In 1981,  the seven largest
steel companies produced about 70 percent of steel made in the United
States.49
     In summary, concentration exists in the production of both types of
coke and in steel production.  However, the concentration probably is not
sufficient to guarantee market power, and many companies are involved in
the production of both coke and steel products.  Other factors must be
considered in any final assessment of market power.
     C.I.4.2  Integration Characteristics.  When one firm carries out
activities that are at separate stages of the same productive process,
especially activities that might otherwise be performed by separate firms,
that firm is said to be vertically integrated.  Through vertical integra-
tion, the firm substitutes intrafirm transfers for purchases from suppliers
and/or sales to distributors.  A firm may seek to supply its own materials
inputs to ensure a stable supply schedule or to protect itself  from monopo-
listic suppliers.  The firm may seek to fabricate further or distribute its
own products to maintain greater control over the consuming markets or to
lessen the chance of being shut out of the market by large buyers or middle-
men.  Therefore, the presence of vertical integration may constitute a
firm's attempt to control costs or ensure input  supplies.  Vertical integra-
tion does  not guarantee market power (control over market price).
     Many  coke-producing firms, especially furnace coke producers, are
vertically integrated  enterprises.  As previously mentioned, 36 of the
existing coke plants are captive,  i.e., they are  connected with blast
furnaces and/or  steel  mills.  In addition, many  coke firms own  coal mines,
and  greater than  61.0  percent of the coal used in  ovens was  from captive
mines  in 1979.33  Assurance  of coal  supply to  coke production  and coke
supply to  pig  iron production appears  to  be the  motivation behind such
integration.
                                  C-30

-------
      One implication of vertical integration is that much of the furnace
 coke used in the United States never enters the open market—it is consumed
 by the producing company.   Accordingly, the impact analysis for furnace
 coke (Section C.2.2) uses an implied price for furnace coke based on its
 value in producing steel products,  which are transferred on the open market.
      C.I.4.3  Substitutes.   Substitutes for a given commodity reduce the
 potential  for market power in production of the commodity.   The substitu-
 tion of other inputs for coke in blast furnaces is somewhat limited, but
 not totally unfeasible.   In addition,  electric arc furnaces,  which do not
 require coke, are becoming increasingly important in steel  production.   The
 trend toward electric arc  furnaces  and minimills has eased  entry into the
 iron and steel  industry, which in turn reduces market power.
      Imported coke also  can be substituted for domestically produced coke.
 In fact,  although U.S.  iron and steel  producers prefer to rely  on  domestic
 sources of  coke,  coke imports have  increased most recently.   If the  cost of
 domestic  coke increased  substantially  compared to the cost  of imported
 coke,  U.S.  iron and steel  producers  might  attempt to increase imports even
 more.   Correspondingly,  if  costs  of  imported coke are reduced because of
 improved foreign  technology and productivity,  reductions in foreign  labor
 cost,  or other reasons,  imports might  become more desirable.
      Furthermore,  substitutes  exist  for  the  final  products  (iron and  steel)
 to which coke is  an  input.   Increases  in the price of coke and  the result-
 ing increases in  the  price  of  iron and steel  products  can lead  to some
 substitution  of other materials for  iron and steel, which also  reduces
 market  power  in the production  of coke.  Analogous substitutions for foundry
 coke are possible, and cupola production of  ferrous products, which uses
 foundry coke, has competition from electric  arc furnaces that do not use
 coke.   Hence, there is a technological substitute for foundry coke in the
 manufacture of ferrous products.  Furthermore, imported foundry coke can be
 substituted for domestic foundry production.  In conclusion, some substitu-
 tion for coke is possible in the manufacture of both steel  and ferrous
products.
     C.I.4.4  Pricing History.  As previously indicated, a significant
portion of all U.S.  coke production  is not traded on the market.  However,
                                 C-31

-------
the U.S.  Bureau of Mines and the Energy Information Administration collect
annual data on coke production and consumption and give the quantity and
the total value of coke consumed by producing industries,  sold on the open
market, and imported.   Dividing total  value by quantity yields an average
price for each of these categories.   Time-series data on these three average
values are given in Table C-13.  (Furnace and foundry coke are combined in
these figures.)
     Also shown in Table C-13 are data on the average value of coal that is
carbonized in coke ovens.  An examination of coke and coal prices reveals
that increases in coal prices generally coincide with increases in coke
prices.  In fact, only 3 years show an increase in the price of coal that
was not accompanied by an increase in the price of the two categories of
coke.  Although it is impossible to conclude from this trend that individual
firms have market power, it indicates that the industry can pass through
some increases in costs.
     C.I.4.5  Market Structure Summary.  Although there is no perfect
method for measuring the extent of market power, the preceding sections
addressed four characteristics used to measure the potential for market
power—concentration, integration, substitution, and historical price
trends.  Concentration statistics indicated that some potential for market
power exists  in the coke industry, yet these statistics are not conclusive
proof.   Similarly, vertical integration in the steel industry is not con-
clusive  in identifying the presence of market power because vertical inte-
gration  is a  method of controlling the cost and ensuring the quality and
supply of  inputs.  Finally, the possibility of substitution represents a
strong argument against  the existence of extensive market power in  the
coke-making  industry.
C.I.5  Financial  Performance
      Financial data on the coke-producing  firms or their parent firms,
including  captive and merchant furnace and foundry producers, are  shown  in
Table  C-14.   Firms for which  data are  not  available are noted.
      Ten companies show  negative  earnings  before  interest and taxes.  Of
these, nine  are  furnace  coke  producers, whose earnings  reflect the  disas-
trous  years  for  the  steel  industry.  As mentioned,  in  1983, steel  firms  had
                                  C-32

-------
        TABLE C-13.
COMPARISON OF COAL PRICES AND DOMESTIC AND IMPORTED
       COKE PRICES6 50 51 52 53
       Average value of  Average value of  Average value of
       coal carbonizeg fa  oven coke used    oven coke sold   Average value of
       in coke ovens, '   by producers,     commercially,    imported coke, >c
                               $/Mg              $/Mg              $/Mg
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
9.56
9.85
10.17
10.19
9.92
9.74
10.31
10.92
10.90
10.89
10.90
10.79
10.86
10.46
10.23
10.48
10.78
11.05
11.03
11.49
13.46
15.43
17.34
20.19
40.22
48.73
48.68
50.99
57.37
55.88
62.09
69.29
71.62
65.36
14.26
14.50
15.11
15.36
17.33
17.90
19.39
19.98
19.82
19.16
19.92
19.12
19.53
18.88
19.17
17.89
18.40
18.58
19.57
21.54
30.30
32.86
35.76
41.34
82.32
92.84
93.83
90.57
105.79
117.39
123.42
125.52
131.24
124.57
14.54
15.72
17.63
17.96
18.95
18.52
20.27
21.51
21.90
23.03
22.32
23.30
23.36
23.24
22.85
23.90
24.49
24.99
24.25
27.01
33.04
41.29
44.87
47.31
72.47
96.61
104.01
111.95
118.03
107.54
113.24
124. 34
126.24
124.67
13.34
13.17
15.96
12.02
11.98
12.26
12.38
14.43
14.25
12.89
13.06
13.44
14.42
14.78
16.10
16.95
20.60
20.41
22.31
21.36
25.46
31.93
27.70
40.16
60.14
94.84
93.35

— ~
94.32
87.12
89.59
84.61
61.45
 Both furnace and foundry coke and the coals used to produce furnace and
 foundry coke are included in these figures.

 Market value at the oven (current dollars).
r»
""General customs value as reported by the U.S.  Department of Commerce
 (current dollars).
                                  C-33

-------
   TABLE C-14.   FINANCIAL  INFORMATION ON COKE-PRODUCING FIRMS, 1983
                  (million 1983 dollars)3 S4 5S 5S
Company name
Armco, Inc.
Bethlehem Steel Corp.
Ford Motor Co.
(Rouge Steel)
Inland Steel Co.
Interlake, Inc.
The LTV Steel Corp.
McLouth Steel Corp.1'-"
(New Boston Coke Corp.)
National Intergroup, Inc.
(National Steel Corp.)
Northwest Industries"1
(Lone Star Steel Co.)
Shenango Furnace Co. , Inc. '
(Shenango, Inc.)
U.S. Steel Corp.
Weirton Steel Corp.p
Wheeling-Pittsburgh Steel Corp.q
Jim Walter Corp."
Koppers Co. , Inc.
Alabama Byproducts Corp.
r
Carondelet Coke Corp.

Chattanooga Coke and Chemicals
Co., Inc.
Citizens Gas and Coke Utility
r t
Detroit Coke Corp. '
Indiana Gas and Chemical Corp.

McWane, Inc. (Empire Coke Co.)
f*
Tonawanda Coke Corp.
Net sales
4,165
4,898
44,455

3,046
835
4,578
11

2,993

1,608

145

16,869
1,000
772
2,025
1,566
229
k


17

316
k

64
k

k

EBITC
(526)
(239)
2,166

(177)
38
(252)
(0.09)

(177)

(104)

k

(1,208)
k
(72)
113
42
k
k

k


k
k

(0.2)
k

k

Cash flow
70
1299
5,542

106g
67
(164)9
k

161

1549

k

l,563g
k
(70)9
159
1749
192
k

k


91
k

k
k

k

Footnotes at end of  table.
                             C-34

-------
TABLE  C-14 (continued)
Company name
Armco, Inc.
Bethlehem Steel Corp.
Ford Motor Co.
Inland Steel Co.
Interlake, Inc.
The LTV Steel Corp.
McLouth Steel Corp.1'^'
(New Boston Coke Corp. )
National Intergroup, Inc.
(National Steel Corp.)
Northwest Industries
(Lone Star Steel Co.)
Shenango Furnace Co., Inc.n>0
(Shenango, Inc.)
U.S. Steel Corp.
Weirton Steel Corp.p
Wheeling-Pittsburgh Steel Corp.q
Jim Walter Corp. n
Koppers Co. , Inc."
Alabama Byproducts Corp.
Carondelet Coke Corp.1"
Chattanooga Coke and Chemicals
Co., Inc.5
Citizens Gas and Coke Utility
Detroit Coke Corp.r>t
Indiana Gas and Chemical Corp.
McWane, Inc. (Empire Coke Co.)
Tonawanda Coke Corp.1"
Annual
interest
expense
154
104
567
63
12
171
k
62
63
k
1,074
k
58
140
26
6
k
k
6
k
k
0.5
k
Total
assets
3,609
4,457
23,869
2,626
674
4,406
11
2,649
1,811
k
19,314
357
1,241
2,609
1,175
243
k
k
343
28
36
112
k
Long-
term
debt
832
1,134
2,713
788
116
1,560
3
606
451
10
7,164
149
514
1,151
233
52
k
k
145
19
0
21
k
Tangiblef
net worth
1,213
1,088
7,545
1,118
314
985
(4)
875
530
73
4,570
k
247
717
554
243
k
k
136
(3)
22
74
k
Footnotes at end of table. , i.. ""
        C-35

-------
TABLE C-14 (continued)
b
Company name
Armco, Inc.
Bethlehem Steel Corp.
Ford Motor Co.
(Rouge Steel)
Inland Steel Co.
Interlake, Inc.
The LTV Steel Corp.
McLouth Steel CorpJ'J
(New Boston Coke Corp.)
National Intergroup, Inc.
(National Steel Corp.)
Northwest Industries
(Lone Star Steel Co.)
Shenango Furnace Co., Inc.n>0
(Shenango, Inc.)
U.S. Steel Corp.
Weirton Steel Corp.p
Wheeling-Pittsburgh Steel Corp.q
Jim Walter Corp."
Koppers Co. , Inc."
Alabama Byproducts Corp.
Carondelet Coke Corp.r
Chattanooga Coke and Chemicals
Co., Inc.5
Citizens Gas and Coke Utility
Detroit Coke Corp.r>t
Indiana Gas and Chemical Corp.
McWane, Inc. (Empire Coke Co.)
Tonawanda Coke Corp. r
Footnotes at end of table.
Net working
capital
563
271
503
233
203
538
(10)
252
338
16
789
147
102
136
282
50
k
k
20
1
2
48
k

Current
assets
1,576
1,259
10,819
789
378
1,848
2
875
762
43
4,298
332
343
1,594
527
73
k
k
82
13
13
58
k

Current
liabilities
1,013
988
10,316
556
175
1,310
12
623
424
27
3,509
185
241
1,458
245
23
k
k
62
12
11
10
k
(continued)
            C-36

-------
                           TABLE C-14 (continued)
  Values  in  parentheses  represent negative  numbers.

  Parent  firms  of furnace  coke  producers  are  listed  first,  followed by
  parent  firms  of foundry  coke  producers.   Subsidiaries  are listed in
  parentheses below parent companies.

  EBIT  =  earnings before interest and  taxes.

  Cash  flow  = operating  income  +  depreciation  -  interest expenses  - taxes.

  Net working capital =  current assets  -  current liabilities.

  Tangible net  worth = equity - intangible  assets.

9Received income tax credit in 1983.   Income  tax represented  as zero in
  cash  flow  calculation.

  McLouth Steel Corp. has  debtor-in-possession status.   The parent company
  filed for  bankruptcy in  1981  and  filed  a  petition  for  reorganization in
 December 1984.   Financial information listed is  for the subsidiary.

JFigures are interim values reported for first  11 months of 1984.
  Converted  to  1983  dollars using GNP implicit price deflator.
l(
  Information not available.

  A merger between National Intergroup, Inc.,  the parent company of National
  Steel Corp.,  and Bergen  Brunswig Corp. fell  through in April 1985,  2
  weeks before  its scheduled date.  Some market  consultants  feel that
  National Intergroup, Inc., is now a potential  target for  corporate  raiders,

  Northwest  Industries,  Inc., the parent company of Lone Star Steel,
  announced  in April 1985  its merger with Farley Industries.

  Producer of both furnace and  foundry coke.

  Financial   information  listed  applies to subsidiary rather  than parent
  company.

pEmployees  formally took control  in January 1984.  All  figures are interim
 values reported for first 3 months of 1984.   Conversion to 1983 dollars
 using GNP  implicit price  deflator.

qWheeling-Pittsburgh Steel Corp.  filed for Chapter 11 in April 1985.
"\
 Owned by James D. Crane.   Financial  information denied.

 Chattanooga Coke and Chemicals Co.,  Inc.  has debtor-in-possession status.
 The company filed for arrangement under  Chapter 11  in  March 1984.

 Latest information available  is  for 1982.   Conversion  to 1983 dollars
 using  GNP  implicit price  deflator.
m
s
                                 C-37

-------
financial losses totalling $3.6 billion.   The balance of steel  trade favored
imports by a 14 to 1 import-export ratio.   Imports totalled 20.5 percent of
apparent supply in 1983.58
     Two integrated steel  producers exhibit negative cash flows, while a
third has negative calculated working capital, as financial resources have
dwindled with the recession.   Two companies, one furnace coke producer and
one foundry coke producer, are operating under bankruptcy status.
     From the financial data in Table C-14, three ratios have been calculated
(Table C-15).  The first,  a liquidity ratio, is a measure of a firm's
ability to meet its current obligations as they are due.  A liquidity ratio
above 1.0 indicates that the firm is able to pay its current debts with its
current assets; the higher the ratio, the bigger the difference between
current obligations and the firm's ability to meet them.  All of the coke-
producing firms have liquidity ratios between 1.0 and 4.0, with the excep-
tions of McLouth Steel (0.17) and McWane, Inc. (5.80).  These figures are
consistent with liquidity ratios for firms in a wide variety of manufactur-
ing industries.
     The second ratio, a coverage ratio, gives an indication of the firm's
ability to meet its interest payments.  A high ratio indicates that the
firm is more likely to be able to meet interest payments on its loans.
This ratio also can be used to determine the ability of a  firm to obtain
more loans.  The coverage ratio of the coke-producing firms ranged from 0.8
to 3.9.  Seven firms for which information was available had negative
coverage ratios because of negative EBIT values.  The positive ratios are
comparable to  the coverage evidenced in most manufacturing industries.  The
poor performance of those firms with negative ratios may be a result of
problems in  the steel  industry.  However, many firms continue to make
investments  funded through mergers, joint ventures, and other means.
     The last  of the ratios, a leverage ratio, indicates the relationship
between  the  capital contributed by creditors  and  that contributed by the
owners.  Leverage magnifies returns to owners.  Aggressive use  of debt
increases  the  chance of default and bankruptcy.   The  chance of  larger
returns  must be balanced with  the  increased  risk  of such actions.  The
leverage ratio indicates  the vulnerability  of the firm  to  downward business
                                   C-38

-------
              TABLE C-15.   FINANCIAL RATIOS FOR COKE-PRODUCING FIRMS
Company name3 Liquidity ratio
Armco, Inc.
Bethlehem Steel Corp.
Ford Motor Co.
(Rouge Steel)
Inland Steel Co.
Interlake, Inc.
The LTV Corp.
McLouth Steel Corp.
(New Boston Coke
Corp.)
National Intergroup, Inc.
(National Steel Corp.)
Northwest Industries
(Lone Star Steel Co.)
Shenango Furnace Co. ,
Inc. (Shenango, Inc.)9
U.S. Steel Corp.
Weirton Steel Corp.
Wheeling-Pittsburgh
Steel Corp.
Jim Walter Corp.9
Koppers Co. , Inc.9
Alabama Byproducts Corp.
Citizens Gas and Coke
Utility
.Detroit Coke Corp.
Indiana Gas and
Chemical Corp.
1.56
1.27
1.05
1.42
2.16
1.41
0.17
1.40
1.80
1.59
1.22
1.79
1.42
1.09
2.15
3.17
1.32
1.08
1.18
Coverage ratio0 Leverage ratio
-3.42e 1.52
-2.306 1.95
3.82 1.73
-0.256 1.20
3.08 0.93
-1.476 2.91
f -3.75e
-2.85e 1.40
-1.666 1.65
f 0.51
1.12 2.34
f f
-1.256 3.06
0.81 3.64
1.59 0.86
f 0.35
f 1.52
f -10. 3e
f 0.50
Footnotes at end of table.
(continued)
                                  C-39

-------
                           TABLE C-15 (continued)
Company name
            a         Liquidity ratio*3    Coverage ratio0   Leverage ratioc

McWane, Inc.                 5.80                                0.42
  (Empire Coke Co.).

aParent firms of furnace coke producers are listed first, followed by
 parent firms of foundry coke producers.  Subsidiaries are listed in
 parentheses below parent companies.  No ratios were calculated for
 Carondelet Coke Corp., Chattanooga Coke and Chemicals Co., Inc., and
 Tonawanda Coke Corp. because of a lack of information.
b, .   ....    ..      Current assets
 Liquidity ratio = current liabilities

Cn          . .      	EBIT	
 Coverage ratio = Annual interest expense   *

d,           ..    Total liabilities
 Leverage ratio = Tang1b1e net worth

Negative values are not meaningful.

 Information not available.
^Produces both furnace and foundry coke.
                                  C-40

-------
 cycles.  Also, a high ratio reveals a low future debt capacity,  i.e. addi-
 tions to debt in the future are less likely.  The firms with coke-making
 capacity had leverage ratios that ranged from 0.3 to 3.7.  Six companies
 had ratios below 1.0, while one firm experienced a negative ratio.  These
 figures highlight the poor financial condition of many firms in the coke
 industry.   Currently, firms with coke-making capacity are engaged in sub-
 stantial amounts of debt financing,  while continuing to make investments.
      Another measure of financial  performance is the rate of return on
 equity.   Studies of the iron and steel  industry show low rates of return on
 equity.   In an analysis performed by Temple, Barker, and Sloane,  Inc.
 (TBS), the real  (net of inflation) rate  of return in the steel  industry was
 estimated  to be  0.2 percent for the  period 1970 to 1980.   The TBS analysis
 projected  a rate of return on  equity of  1.0  percent for 1980 to 1990.59
 These estimates  of  historical  and  projected  return on equity compare very
 poorly with estimates of the required  return on investment in the steel
 industry.   A difference  between realized and required returns implies  that
 equity financing of capital  expenditures may be difficult.
      As  noted, low,rates of  return on equity affect  common stock  prices  and
 have  implications for future investment  financing,  including  environmental
 control  expenditures.  The  following data represent  total  pollution  abatement
 capital  expenditures  (PACE)  as  a percentage  of  new capital  expenditures
 (NCE)  for SIC 3312.2  60  61 62
               Year           Percentage  PACE of NCE
               1975                   20.25
               1976                   20.92
               1977                   22.95
               1978                   22.20
               1979                   25.57
               1980                   20.11
               1981                   15.75
               1982                   12.32
PACE as a percentage of NCE peaked at 25.57 percent in 1979, after having
been fairly steady throughout the latter part of the 1970's.  The trend is
the 1980's  show a PACE as a declining percentage of NCE.   This decrease may
reflect the capital  availability restrictions experienced by the steel
industry during this period.
                                  C-41

-------
     For the steel industry, issuing new stock to raise investment capital
is unlikely under current circumstances.  If environmental  and other control
investments cannot be financed through new equity, another source of funds
must be found.   Increased debt is one potential source.  However, firms
with coke-making capacity already have incurred substantial amounts of
debt.  The TBS analysis concluded that to avoid deterioration in its finan-
cial condition, the steel industry is likely to reduce expenditures to
modernize productive facilities rather than increase its external financ-
ing.63
     The steel industry has had to resort to more creative forms of financ-
ing to provide funds for modernization of facilities.  This upgrading is a
key to gaining and maintaining a competitive position with respect to
imports.  Cash flow for the industry has been below capital requirements
for the past two decades.  Mergers, joint ventures, shared production
arrangements, abandonment of uneconomic facilities, and the sales of assets
are  likely to continue being sources of capital.64  Funds advanced by
customers, with repayment geared to earnings, have been used for equipment
modernization.65
C.I.6   Industry Trends
     The demand for coke is derived from the demand for steel produced by
processes that utilize coke.  Hence, a  description of  steel industry trends
in  technological  development and production is a  useful indicator of future
coke production and coke capacity  requirements.
     As mentioned, there has been  a technological  shift, which  is expected
to  continue, toward labor-saving technology.  Trends  in modernization are
away from open-hearth  furnace production and toward electric arc furnaces
and basic oxygen  furnaces.  In 1960, these processes  accounted  for 88.2  per-
cent,  9.5 percent, and 3.3  percent of U.S. production,  respectively, while
in  1982 these  values were 8.2 percent,  31.1 percent,  and 60.7 percent.66
In  1985, basic oxygen  furnaces are expected to account for 61.5 percent  of
steel  production, with electric  furnaces contributing 34.0 percent or more,
and open hearth  furnaces declining to 6.1  percent or  less.67  Electric  arc
and basic  oxygen  furnaces  represent reductions in production time, as well
as  shifts  to less expensive inputs.67   The increased  use of  these  types  of
furnaces will  result  in  some  decrease in demand  for  coke.
                                  C-42

-------
      Other changes have  improved  industry productivity, quality of yields,
 and energy efficiency.   In-ladle  processes  (performed after the melting
 furnace stage) include inert gas  stirring and vacuum treatments.68  These
 techniques yield higher  quality steel.
      The use of continuous casters, which convert molten steel directly
 into shapes ready for rolling, has increased from 18 percent of production
 in the late 1970's to 35 percent  in 1984.69  Yield of finished product per
 ton of raw steel may be  boosted to 95 percent from the current 76 percent
 by use of this process.   For each ton of finished steel  produced using this
 technology, 15 percent to 20 percent less raw steel  is required, while
 40 percent to 50 percent less energy is needed.67  The impact of these
 technological  developments on the coke industry is unknown.   Any effects
 will  be through productivity improvements in the steel  industry.
      Technological  trends have reduced steel use per unit  of output of
 durable goods.   Since the 1970's,  the decline in consumption of steel  per
 dollar of GNP  has  averaged 4 percent annually,  with  continued  decline
 expected.70  Increases  in economic growth are predicted  to  offset  this
 effect,  resulting  in  an  increase  in steel  use to 95  million  Mg by  1988,
 with  domestic  shipments  representing a 5-percent average annual  rate
 increase  over  the  1983 to 1988 period.65   Projections by the U.S.  Bureau  of
 Mines  predict  U.S.  raw steel  demand will  be  138  million Mg  in  1990, and  164
 million Mg  in  2000.71
     Steelmaking capacity utilization  recently has been low, averaging
 47.3 percent in 1982  and  55.4 percent  in  1983.72   In 1984, this  rate rose
 to 82 percent  in April before dropping to 57  percent in September.73
 Capacity utilization  is an  important measure  of  industry performance
 because of high fixed costs for the  industry.  The larger the volume of
 production, the smaller the cost per unit of  steel produced.  For the  steel
 industry, the break-even  point for operations is at approximately 65 per-
 cent of capability, although this  figure is highly dependent on prices.73
 This means that steel  companies have been operating at losses for several
years.
     The steel  industry has responded to this financial  difficulty by
permanently reducing capacity.  Since 1983, this reduction has been more
than 10 percent, with  perhaps another 5-percent cut necessary.73  From
                                 C-43

-------
122.5 million Mg in 1984, capacity is likely to be trimmed to 109 million
Mg by the late 1980's.73  However, the U.S.Bureau of Mines predicts U.S.
production of raw steel will rise to 113 million Mg in 1990,  and to
132 million Mg in 2000, under assumptions of slow growth in the rate of
production and increases in demand.71  Changes in capacity utilization
affect coke production only to the extent that coke is an input to the
steel production process.  Reductions in steel production, coupled with
shifts to noncoke energy inputs could greatly reduce demand for coke.
     The emergence of minimi 11s to supply regional demand for steel has had
an impact on the operation of the larger integrated steel mills.  Mini
mills now account for approximately 20 percent of U.S. steel  production, at
a cost per ton of installed capacity about 75 percent less than for inte-
grated plants.  The use of electric arc furnaces in minimi 11s may result in
dramatic reductions in coke demand if such mills claim 40 percent of the
steel market by 2000, as some predict.74
     The combination of the factors described in this section indicate that
coke consumption is destined to continue declining.  Technological  improve-
ments are likely to result  in an  input shift away from coke, while  reduced
capacity in the integrated  steel  industry signals a decrease in amounts of
coke needed for blast  furnace steel production.
C.I.7  Market Behavior:  Conclusions
     Market structure,  financial  performance, and potential growth  influence
the choice of a methodology to describe supply  responses  in the coke-making
industry.  Although some characteristics of this  industry indicate  a poten-
tial for market power,  other characteristics  belie  it.
     Some concentration exists in coke-making capacity and steel  produc-
tion; however, many firms  produce coke  and  iron and steel products.  Vertical
integration  is  substantial;  however,  integration appears  to  result primarily
from a  desire for  increased certainty in the  supply of critical  inputs.
Furthermore,  substitution  through alternative technologies and  coke imports
is  feasible,  and  some  substitutes for the  industry's  final products (iron
and steel)  are  available.   In  any industry,  the potential  for  substitution
is  a major  factor leading  to competitive pricing.   Certainly,  the financial
                                  C-44

-------
 profile  of coke-making  firms  is  not  indicative  of monopoly  profits.   Pros-
 pects  for  industry  growth  are limited.   An  individual  firm  must  actively
 compete  with  other  firms  in the  industry to improve  its  profit position,  or
 even to  remain  viable.
     No  industry  matches  the  textbook  definition  of  perfect competition.
 The  important issue is  whether the competitive  model  satisfactorily  captures
 major  behavioral  responses of firms  in the  industry.   Based on the factors
 outlined in this  section,  the competitive pricing model  adequately describes
 supply responses  for coke-making firms.
 C.2.   ECONOMIC  IMPACT OF  REGULATORY  ALTERNATIVES
 C.2.1  Summary
     Economic impacts are  projected  for the baseline  and for each regula-
 tory alternative.   Furnace and foundry coke impacts  are  examined separately
 because  their production costs and markets  differ.   In the  reanalysis, all
 cost and price  impacts  are in second-quarter 1984 dollars.
     All costs  and  prices  used in calculations  were  originally in third-
 quarter  1979  dollars, except  prices  of  steel, furnace coke,  and foundry
 coke,  which were  in  1983 dollars.  Conversions  to the 1984  values were made
 by multiplying  the  1979 values by 1.362,  the ratio of 1984  second-quarter
 GNP implicit  price  deflator to the 1979  GNP  implicit price  deflator.14 75
 The 1983 values were  converted by multiplying by  1.032,  the  ratio of the
 producer price  index  for second-quarter  1984 to the same  index for 1983.76
     When  measured  on a per-unit of  output  basis,  the costs  of meeting
 baseline regulations  for foundry coke plants tend  to be  greater than those
 for furnace coke plants for two  reasons.  First,  some economies of scale
 are present for some  of the controls.  Foundry plants tend  to be smaller
 than furnace plants;  thus, they  have higher  per-unit control costs.   Second,
 for a  given battery,  foundry  coke output will be  less than  furnace coke
 output because foundry coke coking time  is about  two-thirds  longer than
 furnace coke coking time.
     Regulatory Alternative II has annualized costs of $4.8 million above
baseline for furnace and foundry coke producers combined.  Regulatory
Alternative II requires capital expenditures of $45 million above baseline
for furnace and foundry coke producers combined.  Regulatory Alternative  III
                                  C-45

-------
would result in annualized costs of $15.3 million and capital costs of
$80 million over baseline for the combined furnace and foundry coke sectors.
The values are the same whether import competition is assumed for foundry
coke producers (Scenario B) or not (Scenario A).   These costs differ from
engineering estimates because of the calculation of costs based on batteries
with marginal cost of production below price, rather than all batteries.
     Price impacts are estimated under the empirically supported hypothesis
that furnace coke demand is responsive to higher coke prices.  Foundry coke
demand also is assumed to respond to price.   Regulatory Alternative II
would have impacts of $0.13/Mg (0.12 percent change) on the price of furnace
coke, and $0.99/Mg (0.58 percent change) on the price of foundry coke under
Scenario A (1984 dollars).  Under Scenario B, there are no price effects.
Regulatory Alternative III would result in furnace coke price increases of
$0.36/Mg (0.33 percent) and $1.46/Mg (0.86 percent) price increase for
foundry coke under Scenario A, and $0.00/Mg change under Scenario B.
     Regulatory Alternatives II and III would have less than a 1.0-percent
impact on the production of either furnace or foundry coke under Scenario A.
Under Scenario B, Regulatory Alternative II would decrease foundry coke
production by 2.1 percent, while Regulatory Alternative III would result in
a  3.2-percent reduction in foundry coke production.  There are 14 furnace
coke batteries that  currently appear uneconomic.  There are  no uneconomic
foundry coke batteries.  Regulatory Alternative  II does not  force any more
batteries  into the uneconomic production region.  Regulatory Alternative III
results in one additional  furnace coke battery being pushed  into the uneco-
nomic region.
C.2.2  Methodology
     The  following approach  focuses on the  long-run  adjustment process  of
furnace and  foundry  coke  producers to the higher costs of coke production
that the  regulatory  alternatives will create.  These long-run adjustments
involve  investment and shutdown decisions.   Short-run  adjustments,  such as
altering  coking  times, to  meet  the fluctuations  in  the demand for  coke  are
not the  subject  of this  analysis.
      Because of  differences  in  production costs  and markets, furnace  and
 foundry  coke producers are modeled  separately.   Both are  assumed to behave
                                  C-46

-------
 as if they were competitive industries selling coke in a market.  This
 assumption is somewhat more realistic for foundry than for furnace coke
 producers because most furnace coke is produced in plants captive to the
 steel industry.   However, interfirm and intrafirm shipments of coke are not
 uncommon, as can be inferred from Table C-ll.  A plant-by-plant review of
 the coke industry by Hogan and Koelble also confirmed the existence of such
 exchanges.77
      A set of programmed models has been developed to produce intraindustry
 and interindustry estimates of the economic impacts of the alternative
 regulations.   The models are applied to both furnace and foundry coke,  and
 the sectors included are coke,  steel,  and ferrous  foundries.   The rest of
 the economy is incorporated into  the interindustry portion of the analysis.
      The analytical  approach incorporates a production cost model  of the
 coke  industry, based on engineering data,  and an econometric model  of the
 steel  industry.   The interrelationships  of these models  for furnace  coke
 are shown in  Figure  O3.   The  upper portion of Figure  C-3 encompasses the
 supply side impacts  of  the regulatory  alternatives;  the  lower portion con-
 tains  the demand  side impacts.  In  the synthesis step,  the  two sides  are
 brought  together,  and the equilibrium  price  and quantity  relationships  are
 determined.   An analogous diagram for  foundry coke would  substitute  ferrous
 foundry  products  for steel.  The methodology is described further in  the
 following subsections.
     C.2.2.1  Supply Side.   A production cost model that  incorporates
 technical  relationships and  engineering cost estimates is used with plant-
 specific  information to compute separate industry supply  functions, with
 and without additional controls.78  Supply functions are  estimated on*a
year-by-year  basis for furnace and  foundry coke plants projected to be in
 existence between 1984 and 1995.  Both coke  production costs and the costs
 that plants incur to meet existing  environmental regulations are computed
to estimate the industry supply curve before any additional controls are
applied.   Estimates of the costs of control  for compliance with the regula-
tory alternatives are used to compute the projected upward shifts in that
supply function.   All costs are in 1983 dollars, converted to 1984 dollars
for this reanalysis.
                                 C-47

-------
 i
-t»
oo
                   COS IS Of  fOKC
                    I'ltomici ION
                   cnsrs OF EXISTING
                     ENVIRONMENTAL
                      HECULAIIOHS
                     cosis or
                   111 GUI AlOltV
                   AL1EKNATIVES
                   FOREIGN DEMAND
                    I OR US SI EEL
                                               EX I SUNG PI ANT
                                                 INVENTORY
                                                 NEW PLANT
                                               CONFIGURATIONS
                          FOREIGN f.OKE
                             SUPPLY
                          DOMESTIC
                         -  COKE
                           SUPPLY
FOREIGN DEMAND
  FOR US COKE
                    DOMESTIC DEMAND
                      FOII US STEEL -
                                                 DEMAND FOR
                                                  US STEEL
                                              SYNTHESIS   .IMPACTS  	IMPACTS
                                            " bWVTORAL   ON  COKE        ON FINAL
                                             ASSUMPTIONS   AND STEEL   OCMAJO PRICES
                             DEMAND FOR
                               US COKE
                                                        Figure C-3. Economic impact model.

-------
       This  approach  provides  a method  of  estimating  the  industry  supply
  curve for  coke, which  shows  the alternative  coke  quantities  that will be
  placed on  the market at alternative prices.  When the supply curve  is
  considered  in conjunction with the demand curve,  an equilibrium  price and
  coke  output  rate can be projected.  Supply curve  shifts caused by the
  regulatory alternatives can  be developed from the compliance cost estimates
  made  by the  engineering contractor.  These new supply functions,  along with
  the demand curve, then can be used to compute the equilibrium price and
  output rate  under each regulatory alternative.
       C. 2.2.1.1  Data base.   PI ant-by-plant data on more than 60  variables
  for furnace  and foundry coke plants in existence  in 1979 were compiled from
  government publications, industry contacts, and previous studies  of the
 coke  industry.   The data base was sent to the American Iron and Steel
  Institute (AISI),  which submitted it to their members for verification,
 corrections, and additions,79 and to the American Coke and Coal  Chemicals
 Institute  (ACCCI).   The data were adjusted  to account for the 1984 plant
 inventory  in the reanalysis.   Capacity,  number  of ovens, and status  (hot
 idle,  cold idle, under  construction,  or online)  were updated for each
 battery.5
      C.2.2.1.2   Output  relationships.   For  a  given battery,  the  full capac-
 ity output of coke,  measured in megagrams per year (Mg/yr),  is dependent on
 the nominal  coal charge (megagrams  of  coal  per charge) per oven,  the number
 of ovens,  and the effective  gross coking  time (net coking  time plus  decar-
 bonization  time).  The  following values for effective  gross  coking time
 were used  except where  plant-specific  values  were  available.78
                                       Furnace          Foundry
                                       coke              coke
              Wet coal                  is h              30 h
              Preheated  coal            13 h              24 h
 An  age-specific outage  rate that varies from 4 to  10 percent  is assumed to
 allow  for normal maintenance  and repair.  Thus,  the model assumes  some
 increase in such costs as a plant's age.
     The quantities of by-products produced are estimated from engineering
 relationships.  These quantities  depend on the amount of coal carbonized,
percentage  of coal  volatile matter,  coking time,  and  configuration of the
                                 C-49

-------
by-product facility at a plant.   The by-products included in the model  are
coke breeze, coke-oven gas, tar,  crude light oil,  benzene-toluene-xylene
(BTX), ammonium sulfate, anhydrous ammonia, elemental  sulfur, sodium phe-
nol ate, benzene, toluene, xylene, naphthalene, and solvent naphtha.   All
plants are assumed to produce breeze and coke-oven gas.
     C.2.2.1.3  Operating costs.   The major costs  of operation for a coke
plant are expenditures for coal,  labor, utilities, and chemicals.   The
activities within the coke plant  were allocated to 5 production and 10
environmental control cost centers (Figure C-4) to facilitate the develop-
ment of the operating cost estimates.
     Coal  is the major operating  cost item in coke production.  Plant-
specific estimates of the delivered price of coal  were developed by identi-
fying the mine that supplies each plant and estimating transportation costs
from the mine to the plant.  When it was not known which coal mine supplied
a particular plant, it was assumed that the coal came from the nearest
mines supplying coal of the same  volatile matter and ash content as that
used by the plant.  Transportation cost estimates  were based on the dis-
tances traveled and the transport mode (barge or rail) employed.
     Maintenance labor and supervision requirements were estimated for  69
jobs within the coke plant.  Primary variables that determine the number of
maintenance labor and supervision man-hours needed include type of plant
(merchant or captive), number of battery units, number of plants at a site,
size of by-product plant, type of coal charge (wet or preheated), and coke
production.   The labor rates used for captive plants were $23.21/h for
supervisory positions and $21.38/h for production labor.  For merchant
plants, rates of $21.52/h and $19.61/h were assumed.  These values represent
numbers used in the 1979 analysis and scaled by the GNP implicit price
deflator to 1984 dollars for the reanalysis.
     The major utilities at a coke plant are steam, electricity, water, and
other  fuels.  Utility requirements were estimated from the data on the
plant  configuration and output rates for coke and the by-products.  The
prices used for the utilities are $7.41/103 Ib steam; $0.037/kWh electric-
ity; $0.22/103 gal cooling water; and $3.76/106 Btu underfire gas.  These
values are the original 1979 figures scaled to 1984 dollars by the GNP
implicit price deflator for the  reanalysis.
                                 C-50

-------
o
I
en
                                                                                                             '          •  IWIM im~wi< I
                                                                                                             I OSLO*!!?* '  '    «•">•«'   I
                                                                        >-|   iMAinii  ;
                                                                         I UiLUSUlJ J
                                                                         &*•••••••—«ww*
                                                                 Figure C-4.  Coke plant cost centers.

-------
     C.2.2.1.4  Capital  costs.   Although no net additions to industry
coke-making capacity are anticipated during the 1984 to 1995 period,  a
number of producers had plans to rebuild or replace existing batteries in
1979.  In 1984, three new batteries had been constructed and one was  under
construction.5  Such actions alter the long-run industry supply curve
because the new batteries typically have lower operating costs per unit of
output than the batteries they replace and, most important, their capital
costs will be reflected in the new supply curve.
     The capital cost breakdown for new plants is shown in Table C-16.  For
such plants, the major capital cost items are the battery, quench tower,
quench car, pusher machine, larry car, door machine and coke guide, by-
product plant, coal-handling system, and coke-handling system.  A 60-oven
battery is assumed.  Pipeline charging can increase the coke-making capacity
of a given oven by about 25 percent by reducing gross coking time.  Conse-
quently, the per-unit operating cost is reduced.  The capital costs show
economies of scale, i.e., larger plants have smaller per-unit-of-capacity
capital costs.  The capital cost per unit of capacity is higher for pipe-
line-charged batteries than for conventionally charged batteries.
     Periodically, batteries must undergo major rehabilitation or rebuild-
ing  because of performance deterioration.  The costs of pad-up rebuilds
will vary from site to site depending on battery maintenance, past operat-
ing  practices, and other factors.  Average estimates of the cost of rebuild-
ing  were developed for this study and are shown in a report by PEDCo  Envi-
ronmental,  Inc.81  The economic life of coke-making facilities is subject
to considerable variation depending on past maintenance and operating
practices,  which also affect current operating  costs.  For this study,  25
years was used as  the average preferred life of a  new coke-making facility;
however, many  batteries  are operated for 35 to  40  years.   If  35 to 40 years
is a more reasonable battery  lifetime,  use of  a 25-year  lifetime will
result  in some overestimation of  the annual costs  of new  or  rebuilt  facili-
ties.   However, firms probably  will not plan or expect to  wait 35 to
40 years  to recoup an  investment  in coke-making capacity.
     C.2.2.1.5 Environmental costs.   Plant-specific estimates of the
installed capital  and  operating costs  for  current  environmental  regulations
                                  C-52

-------
	TABLE C-16.  ESTIMATED CAPITAL COSTS OF NEW PLANTS80

                                   Conventionally            Pipeline
                                   charged battery        charged battery

	4-ma       6-ma        4-ma       6-ma

Capacity (10s Mg/yr)              450        720         560        900

Capital costs by element
  (106 1979 dollars)

  Coke battery                     34.20      48.90        64.60      83 70
  Quench tower with baffles         2.45       2.85         2 45       2*85
  Quench car and pushing
    emissions control                6.58       7 92         6 58       7 92
  Pusher machine                    2.50       3.20         2*40       3*20
  Air-conditioned  larry car         1.72       2.28         0 00       o'oo
  Door machine and coke guide        1.80       2 10         1 80       2 10
  By-product plant                 32.50      39.75        35.'76      43*74
  Coal-handling system              18.20      23.60        20 62      26*70
  Coke-handling system               6.85       8.80         7 74      lo'oo
  Offsltes                           1.60       1.80         1.69       L91

    Tota1	$108.40   $141.20     $143.74   $182.12

In  the  production  cost model, new  foundry batteries were  assumed to  be
4-m batteries  and  new  furnace batteries were  assumed to be  6-m batteries.
                                C-53

-------
and the regulatory alternatives under consideration in this study were
incorporated in the model.   In the reanalysis,  the current regulations are
assumed to include workplace standards (Occupational  Safety and Health
Administration [OSHA]), water quality regulations (best practicable tech-
nology [BPT] and best available technology [BAT]), and State implementation
plan (SIP) requirements.  Compliance expenses incurred for all plants in
the data base for each of the current regulations assumed baseline control
costs were estimated.  Costs to comply with OSHA and BPT water requirements
under the Federal Water Pollution Control Act were assumed incurred by
1981.  Costs for all other baseline environmental regulations were assumed
to be incurred by 1983.
     The scatter diagrams in Figures C-5 and C-6 show estimates from the
coke supply model of average total cost of production in 1984, including
environmental costs, for all furnace and foundry coke plants.  A weak,
inverse relationship between the average cost of production and the size of
the plant is evident in Figures C-5 and C-6.  However, a number of other
factors create variability in the average cost of production across coke
plants.  The most important of these factors are the delivered price of
coal, the age of the plant, and the by-products  recovered.
     C.2.2.1.6  Coke supply function—existing facilities.  The operating
and capital cost functions were used to estimate the cost of production,
including relevant  environmental costs, for all  plants in the data base.
This cost does not  include a return on investment for existing facilities.
The  capital costs for  these facilities already have been incurred and do
not  affect  operating decisions.
     Capital costs  that have not yet been  incurred are annualized at 6.2
percent, which is estimated to be the real  (net  of inflation) cost of
capital  for the  coke industry.  (This percentage is an after-tax estimate.)
This figure, which  was estimated from data  on the  capital  structure  for
publicly owned steel companies, has  been  used in this  study  as the minimum
acceptable  rate  of  return  on  new  facilities.82
     The regulatory alternatives  for coke-oven by-product  plants  involve
control  equipment  that is  not  affixed to  batteries.   Accordingly,  the
equipment is  not affected  by  battery age  or size (height)  of the  battery
                                  C-54

-------
en
01
1 •.'».'
14O
5
^ 1.30-
c
o
3 1 2O -
O
O.
H-'
o
o
too -
90 -
o.:
a '

1 Q D
a a
a aD
a
qj -ft n °
n o a cp a a a
D
D D n
a
~~i — i 	 1 	 1 	 1 	 1 	 1 — § •
? 0.4 0.6 0.8 1.0 1.2 iT"^* * 1.8 ' 2.0 ^^2
                                                Production (1,000,000 Mg/Yr)
                                Figure C-5. Estimated average cost of furnace coke production as a function

                                                     of plant production, 1984.

-------
o




'o>
2
•\
4ft
C
g
-M
u
3
O
i-
*&-
O
-M
W
O
O


1 00 -j

180 -
170 -



1<50 -

150 -

140 -



1.30 -
120 -
110-
	 	 •• • • 	
a
p
I
i
1

i
n p
D a


a
B a




a
— - -| | | i i I i I
                         .0
150.0           250.0          350.0

          Production (1000 Mg/Yr)
                                                                                       450.0
                               Figure C-6. Estimated average cost of foundry coke production as a function
                                                   of plant production, 1984.

-------
 replacement.  The capital costs of the regulatory alternatives are annual-
 ized over the life of the control equipment (20 years).  This action is
 tantamount to assuming either that all by-product plants have a remaining
 life of at least 20 years or that the control  equipment is salvageable.
      The supply function for each plant is estimated as follows:   the
 average cost of production is computed for each battery in the plant; these
 batteries are arranged in increasing order of  their average costs of pro-
 duction, and the output for each battery is accumulated to produce a stepped
 marginal cost function for the plant; plant overhead costs are averaged for
 all  relevant plant output rates; and average total  costs are computed for
 each output rate by summing the average costs  for plant overhead  and the
 battery.  Each plant's supply function is  the  portion of the marginal cost
 function above the average total cost function.   For existing plants where
 the  average total  cost exceeds marginal  cost over the entire range of
 output,  the supply function is the  point on the  plant's average total cost
 function represented  by capacity output (after allowing for outages).  .The
 aggregate long-run supply function  for all  currently existing coke plants
 and  batteries  is obtained by  horizontally  summing the supply function for
 each plant.  The 1984 industry marginal  cost (supply)  curves  for  existing
 furnace  and foundry coke  plants  are  presented  in  Figures  C-7  and  C-8,
 respectively.
      C. 2.2.1.7  Coke  supply function—new  facilities.   The  cost of coke
 production  for new furnace  and  foundry batteries  was  estimated from  the
 engineering cost model, assuming the  new model  plants  described previously.
 These costs include the normal  return on investment  and allowances  for
 depreciation and corporate  income taxes.  When expressed on a per-unit
 basis, these costs are the minimum price at which it  is attractive to build
 new  facilities.
     C.2.2.2  Demand Side.  The  demand for coke is derived from the demand
 for products that use coke as an input to production—primarily steel and
 ferrous foundry products.   A demand function for furnace coke was  derived
by econometrically modeling the  impacts of changes in furnace coke produc-
tion costs on the steel industry.83
                                 C-57

-------
o
en
00



o>
4A
W
O
o





zuu -
190
180 -
1 70 -
1 .50 -
150 -
MO -
1.30 -

1 20 -
110-
100 -
rtn -
0.





/""
i
/
_^ ^ 	 -^
_^£:^~~~~~~~~
i i i ii i i i i i i i | i
0 i.O 8.0 12.0 16.O' 20.O 24.0 28.0
                                              Production (1 .OOO.OOO MgAr)
                                  Morglnol  cost                •	Average
                                 Figure C-7. Marginal and average cost functions for furnace coke, 1984.

-------
en
10
o
(J
 190 -T—





 180





 170





 16O





 150





14O





13O -





120 -
                  no


                                   T   r—i	1	1	r
                      0.2   0.4   0.6    0.8    1.0   1.2    1^4  '  L^  1^8 ^O^T^


                                           Production (1,000.000 M^/Vr)
                                Marglnol cost                	  Avorag« cost
                               Figure C-8. Marginal and average cost functions for foundry coke, 1984.

-------
     The econometric model  of the steel  industry has two sectors:   steel
and coke.   The steel sector includes domestic steel  supply,  steel  imports
and exports, and steel consumption (steel  supply plus imports minus exports).
Similarly, the model of the coke sector consists of domestic coke supply,
domestic coke demand, and coke imports and exports.   The two sectors are
linked by a derived coke demand function,  which includes as  variables steel
production, steel price, and quantities and prices of other  inputs to steel
production.  The domestic supply of steel  is assumed equal to domestic
demand for U.S. steel plus world demand for U.S. steel minus U.S.  import
demand.
     Both linear and nonlinear specifications were used to estimate the
steel-sector model.  Two-stage least squares was used to estimate the
different components of the steel sector.   Visual inspections of the corre-
lation matrix and a plot of the dependent variable versus the residuals
indicated no multicollinearity or heteroscedasticity problems.  The Durbin-
Watson statistic showed no evidence of autocorrelation.
     The econometric estimation of the coke sector was complicated by the
small  share of total domestic production that is traded in the market.  The
fact that very little coke is actually sold creates concern over the
reported price of coke.  Therefore, estimates of the implied price of coke
were developed, based on the value of coke in steel making, and used in the
estimation of  elasticities.84 8S  Estimates of elasticities for coke and
steel  functions are presented in Table C-17.  Actual prices for coke pro-
duced  and  used internally by the producing companies were used in the
reanalysis.
     An attempt was made to derive a demand function for  foundry coke in an
analogous  manner.   However, the  relevant coefficient estimates were not
statistically  significant at a reasonable level.  A direct estimation of
the demand  function,  based on the prices of foundry coke, foundry coke
substitutes, and complementary inputs, also was  attempted.  Unfortunately,
the data  necessary  to properly estimate the demand  function were not  readily
available  from published sources.  Accordingly,  the elasticity of demand
for foundry  coke was  estimated based on the theoretical relationship  between
the production function for  foundry products  and the  derived  demand  function
                                   C-60

-------
       TABLE C-17.   ESTIMATES  OF  ELASTICITIES  OF STEEL  AND  COKE MARKETS

                                                  Point         Interval
	                             estimate        estimate3

1.    Percent change in  furnace coke  demand        -1.29b             —c
      for  1  percent  change  in  the price  of
      furnace coke

2.    Percent change in  foundry coke  demand        -1.03C             --c
      for  1  percent  change  in  the price  of
      foundry coke

3.    Percent change  in  import demand for 1        1.88       (-1.68, 5.44)
      percent change  in  the price of  furnace
      coke

4.    Percent  change  in  price of steel for         O.llc            —c
      1 percent change in the price of
      furnace  coke

5.    Percent  change in  steel demand for          -1.86d      (-0.54  -3 18)
      percent  change in the price of steel
                              t
6.    Percent  change in steel imports for          1.51d      (0.51, 2.51)
      1 percent change in the price of steel

Note:   Estimates are based on  the empirical  analysis using annual  data for
       the years 1950-1977 with  a structural  econometric model  of  steel  and
       coke markets.

 Interval  estimates  are  based  on  95 percent  confidence  level.

 Derived from the production function for steel, and input  cost shares.

 Calculation based  on the theoretical relationship between  input demand
 elasticity and input cost share  in the  production of outputs.   Accord-
 ingly,  no interval  is provided.

Significantly different from  zero  at 1  percent  level of  statistical
significance.
                                 C-61

-------
for inputs to foundry production.   This elasticity calculation is based on
a 3-year average of the cost share of foundry coke in foundry production.
This estimate is presented in Table C-17.   This elasticity assumes U.S.
demand for foundry coke is supplied entirely from domestic production
(Scenario A).
     Another scenario is that imported foundry coke competes with the
domestic product in the open market (Scenario B).  A simplifying assumption
is that they are perfect substitutes in the production processes that
utilize foundry coke.  In this case, a reduction in U.S.  supply is compen-
sated by imports, so that price need not rise if the quantity of imports
purchased is increased.  Both scenarios are examined in the reanalysis.
     C.2.2.3  Synthesis.  Separate linear functions were fit to the furnace
and foundry coke marginal cost values depicted in Figures C-7 and C-8.  As
illustrated in Figures C-9 and C-10, each supply function is used with the
demand function for the appropriate type of coke to compute the initial
equilibrium price-quantity values (Px and Qj. in  Figures C-9 and C-10).  In
                     i
the case where imports are not perfect substitutes for domestically produced
coke, the supply function is reestimated for each regulatory alternative
(S2 in Figure C-9), and the new equilibrium price-quantity values (P2  and
Q2  in Figure C-9) are predicted.
     The case where  imports compete with domestically produced foundry coke
is  shown in  Figure C-10.  As in Figure C-9, the  supply curve shifts backward
to  S2.  However, because  imports  are available,  no change in price and
quantity need be experienced by the consumer.  Although domestic production
is  reduced by Qi-Q2, the  share of the market supplied by  imports  increases
by  this same amount.  The new equilibrium price  and quantity for  domestic
coke are Pj  and Q2 in  Figure C-10.
     C.2.2.4 Economic  Impact Variables.  Table  C-18 shows the specific
economic variables for  which impacts are estimated.  The  methodology  pre-
sented previously was  designed to provide  industry-level  estimates of these
impacts.  The conventional  demand and  supply partial equilibrium  model of  a
competitive  market was  chosen  for this analysis  because  it was believed to
represent the  key  characteristics of the coke  market and  many  of  the  impacts
of  interest  can  be estimated  readily from  this model.
                                  C-62

-------
S/Mg
    P2




    PI
                                                                     S2
                                                                     SI
                                                           10 J Mg /Tr
               Figure C-9. Coke supply and demand without import competition.
                                 C-63

-------
$/Mg
    PI
                                                                   S2
                                                                    SI
                          Q2
Ql
                                                          10   Mg / Yr
                 Rgure C-10. Coke supply and demand with import competition.
                                   C-64

-------
        TABLE C-18.  ECONOMIC IMPACT VARIABLES AND AFFECTED SECTORS
Variable
Price
Output
Profits
Costs
Plant closures/openings
Capital requirements
Factor employment
Labor
Metallurgical coal
Imports

Furnace
coke
X
X
X
X
X
X
X
X
X
Sector
Foundry
coke Steel
X X
X X
X
X

X
X
X
Xa X

Final
demand
X







Impacted under Scenario B.
                              C-65

-------
     Figure Oil represents the markets for furnace coke and for foundry
coke which is free of import competition (Scenario A).   Figure C-12 describes
the market for foundry coke that must compete with imported coke, which is
assumed to be a perfect substitute for domestic coke (Scenario B).   In
Figure C-ll, D represents the derived demand for coke.   The line S^ repre-
sents the baseline supply curve for coke.   The equilibrium price and quantity
are Pt and Ql5 respectively.  The area Ci+g+h is the total cost of coke
production, b+c1+c2+e+g+h is the total revenue, and b+c2+e represents
before-tax profits.  The total cost of coke production (Ci+g+h) can be
divided into costs incurred to produce coke per se and the costs being
incurred to meet baseline environmental regulations.
     The regulatory alternatives will increase the cost of coke production
by shifting the supply function to S2.  This is not a parallel shift because
of the small magnitude of changes and the continued production by uneconomic
firms.  Given the demand and supply functions as drawn in Figure C-ll,
higher costs of production will lead to higher prices.   A production decrease
as shown in Figure C-ll would cause price to rise to P2 and the quantity
demanded to fall to Q2.  The actual costs to the producer of the regulatory
alternative are c2+d-cls and profits before income taxes are a+b+Ci.
     Costs to consumers are represented by a+d+f, the amount that consumers
paid to purchase the amount (Qi-Q2) at price P! before the regulation, but
now must pay price P2 to purchase.
     In Figure C-12, D represents derived demand for coke and Sx represents
the baseline supply curve for coke, with P± and Qx representing equilibrium
price and  quantity
     As in Figure  C-ll, area Cj+g+h is the total cost of coke production,
including  expenses incurred to meet baseline environmental regulations.
Area b+Ci+c2+e+g+h is the total revenue, and b+c2+e is before-tax profits.
     The regulatory alternatives  shift the supply  function to S2.   As
explained, this is not a parallel shift.  In this  scenario, price does  not
rise even  though domestic production  is reduced.   Instead, because  imported
coke  is assumed to be a perfect substitute for  domestic coke  and because
imported coke  is assumed to be available at price  Pj, domestic  consumers
purchase more  imports and  less domestic coke.   The  results are  that domestic
                                  C-66

-------
$/Mg
     P2

     PI
                                                                          S2
SI
                                                               10   Mg/?r
                 Rgure C-11. Coke demand and supply with and without regulatory
                          alternatives, without import competition.
                                 C-67

-------
$/Mg
                                                                          S2
     PI
                                     Q2
Ql
                                                               10   Mg /Yr
                     Rgure C-12. Coke demand and supply with and without
                        regulatory alternatives, with import competition.
                                     C-68

-------
 production decreases to Q2, imports increase by Q^Qs, and price remains at
 PI, as shown in Figure C-12.
      The costs of the regulation to the producer are Ca'Cj.  Total revenue
 is b+C!+c2+h, and production costs are h+c2.  Profits before income taxes
 are b+Ci.  There are no costs to consumers because they are able to purchase
 quantity Qi at price Px as they were before the regulation.
 C.2.3  Furnace Coke Impacts
      As described in Section C.2.2 of this analysis, the furnace coke
 industry has been modeled as a competitive industry supplying coke to the
 steel  industry.   This definition implies the existence of interfirm and
 intrafirm shipments of coke.   However,  no allowance has been made for coke
 transportation costs, although coal  transportation costs  are included in
 the cost of coke production estimates.   Coke plants and their associated
 steel  mills are  typically clustered  together.   As  noted in Section  C.I.4.1,
 most coke is  consumed within  the region where it is produced.   Hence,
 transportation across great distances  is uncommon.   Therefore,  the  omission
 of coke transport  costs  should not greatly influence the  calculations.
     The  baseline  values  for  1983, presented in  Table C-19,  are actual  data
 for 1983,  except for coke  prices, which are  calculated  by  the model.  The
 values  for  1983  are  assumed to  reflect  full  compliance  with  applicable  SIP
 and OSHA  air  quality regulations  and water quality  regulations.   The  coke
 supply  model  was used  to compute  the price of furnace coke,  costs,  revenues,
 and profits,  given these actual  values.  Coal consumption  and employment
 projections were made  using current coal-  and labor-output ratios.  The
 supply  function was  reestimated assuming control levels being practiced  in
 1984 for all  emission sources.  This estimation was  used to determine the
 impacts of moving from baseline industry control levels to alternative
 regulations control  for all sources.
     Table C-20 presents total costs incurred by companies in SIC 3312 in
meeting environmental regulations up to 1983.  These costs represent indus-
try efforts to achieve baseline compliance.  Expenditures are segmented  by
type of pollutant treated.   Total cost for abatement increased throughout
the late 1970's and peaked at $956.5  million (1972 dollars) in 1979.
Expenditures declined slightly in 1980 and 1981 and dropped to an 8-year
low of $549.8 million (1972 dollars)  in 1982.
                                 C-69

-------
          TABLE C-19.   BASELINE VALUES FOR ECONOMIC
         IMPACT ANALYSIS—FURNACE COKE,  198337 86 87

                                            Baseline values

Coke market
  Price (1983 $/Mg)                                106.25b
  Production (103 Mg)                            20,462
  Consumption (10s Mg)                          24,380
  Imports (103 Mg)c,                             3,918
  Employment (jobs)                              6,236
  Coal consumption (103 Mg)                     29,787

Steel market
  Price (1983 $/Mg)                                319.97
  Production (10s Mg)                            76,763
  Consumption (103 Mg)                          75,710
  Imports (103 Mg)                              15,486
  Employment (jobs)                            295,000

aBaseline assumes companies meet existing regulations
 including OSHA (coke-oven emissions); State regulations on
 desulfurization, pushing, coal handling, coke handling,
 quench tower, and battery stack controls; and BPT and BAT
 water regulations.
 Calculated.  Market price for furnace coke was $123.51 in
 the fourth quarter of 1983.
GCalculated.  Imports  = Consumption - production.
 Calculated.  Furnace  coke employment = Employment in
 byproduct coke industry x proportion of coke production
 represented by furnace coke sector.
Represented by employment in SIC 3312 (Blast furnaces and
 steel mills).
                            C-70

-------
                            TABLE 020.   POLLUTION ABATEMENT EXPENDITURES FOR SIC 331262
o
—I
Abatement expenditures.
Year
1975
1976
1977
1978
1979
1980
1981
1982
Air
pollution
477.
606.
675.
709.
932.
925.
940.
666.
3
9
1
1
6
4
9
8
Total cost = Capital
K
Implicit
1972 val
clncludes
T M T ~J
price
ues.
Water
pollution0
306.8
309.5
385.8
424.9
511.6
494.9
512.6
408.3
expenditure
deflator for the

payments to

government
106 current $
Solid. Recovered cost,
waste 106 current $
43.
51.
85.
111.
99.
127.
153.
92.
1
4
5
2
5
9
0
4
18.1
17.8
18.1
15.1
1.0
18.8
22.3
22.0
+ operating costs - recovered
nonfarm

business

units for public
sector used to

sewage use.
Total cost,a
106 current $


1,
1,
1,
1,
1,
1,
809.
950.
128.
230.
542.
529.
584.
145.
cost summed
convert


1
0
3
1
7
4.
2
5
for all
current dol




Total cost,
106 1972 $
647.
722.
812.
824.
956.
856.
807.
549.
3
0
4
2
5
5
9
8
pollutants.
lars to





        Includes payments to government units for solid waste collection and disposal.

-------
     C.2.3.1  Price Effects.   The price of furnace coke is assumed to be
established in a competitive market.   In the basic model  of a competitive
market, the interaction of supply and demand determine the equilibrium
price.  This price is dependent on the costs of production of the marginal
producer and the value of the product to the marginal buyer.   The marginal
producer is the producer who is willing to supply the commodity at the
market price because he is just covering all his costs at that price.  The
marginal buyer is just willing to pay the market price.  Other buyers who
value the product more still pay only the market price.
     Estimates of the demand and supply functions for furnace coke are
necessary to develop projections of the equilibrium price for furnace coke
with and without increased control.  The supply of furnace coke as shown
previously would be shifted by the regulatory alternatives.  The demand for
furnace coke has been econometrically estimated and found to be responsive
to price changes.  The estimated elasticity of demand for furnace coke is
-1.3.  This responsiveness reflects the substitution of other fuels  for
coke  in blast furnaces; the substitution of other inputs, primarily  scrap,
for pig iron in steelmaking; and the substitution of other commodities for
steel throughout the economy.
      Higher prices for coke will increase the cost of  steel production
unless there is a perfect substitution between coke and other inputs to
steelmaking.  In that case, the consumption of coke would decrease to zero.
If substitutions for coke in steelmaking were not possible (i.e., input
proportions were fixed), the steel price  increase would be the percentage
change  in  coke price times the  share that coke represents  in the  cost of
steelmaking (10.7 percent) times the base price of steel.
      Table C-21 presents the furnace coke and steel  price  impacts of the
regulatory alternatives.  The proposed  regulatory alternatives raise coke
prices  only slightly:   0.12 percent  for Alternative  II, and  0.33  percent  for
Alternative III.
      C.2.3.2   Production and Consumption  Effects.  The estimated  demand  and
supply relationships  for coke  are  used  to project the  production  and con-
sumption  effects  of  the regulatory alternatives.  As shown in Table  C-22,
the  changes  in  coke  production  and consumption  are fairly small  for  the  two
                                  C-72

-------
  TABLE C-21.   PRICE EFFECTS OF REGULATORY ALTERNATIVES-
                    FURNACE COKE, 1984a

Regulatory                         Coke,            steel,
Alternative	               $/Mg              $/Mg


    11                             0.13             0.04
                                  (0.12)           (0.01)

    111                            0.36             0.12
	(0.33)	(0.04)

 Values in parentheses  are percentage changes from baseline.
                          C-73

-------
                  TABLE C-22.   PRODUCTION  AND CONSUMPTION EFFECTS OF REGULATORY ALTERNATIVES-
                                              FURNACE COKE, 1984a
Regulatory
Alternative
II
III
Coke market, 103 Mg/yr
Production Consumption
-32 -23
(-0.16) (-0.09)
-90 -65
(-0.44) (-0.26)

Imports
9
(0.23)
25
(0.64)
Steel market, 103 Mg/yr
Production Consumption Imports
-21 -18 3
(-0.03) (-0.02) (0.02)
-60 -51 9
(-0.08) (-0.07) (0.06)
a
 'Values in parentheses are percentage changes  from baseline.

-------
 regulatory alternatives.  For Alternative II, changes in production and
 consumption are less than 0.2 percent.  For Alternative III, the quantity
 changes are less than 0.5 percent.
      Imported coke is a close substitute for domestically produced coke.
 Imported coke is not a perfect substitute because coke quality deteriorates
 during transit and contractual arrangements between buyers and sellers are
 not costless.   However, increases in the costs of production for domestic
 plants will increase the incentive to import coke.
      The projected increases in coke imports are reported in Table C-21.
 Imports increase by 0.23 percent under Alternative  II and 0.64 percent
 under Alternative III.   As illustrated below,  coke  imports increased sig-
 nificantly since 1972,  but peaked in 1979 and  began a marked decline.
                          Year       Imports,  103 Mq
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
168
978
3,211
1,650
1,189
1,659
5,191
3,605
598
478
109
32
     The  increase in imports in the 1970's is believed to be the result of
a coal strike in the United States during 1978 combined with depressed
conditions in the market for steel in the countries exporting coke to the
United States.  Accordingly, future importation at a high level may depend
on future market conditions for steel in other countries.   In any case, the
change in coke imports projected for all the regulatory alternatives is
small.
     C.2.3.3  Coal  Consumption and Employment Effects.   Any reductions in
coke and steel production are expected to cause reductions in the use of
the factors that produce them,   The major inputs to coke production are
coal  and labor.   Labor is also an important input in coal  mining.
                                 C-75

-------
     The coal consumption and employment implications of the projected
reductions in coal, coke, and steel  production are shown in Table C-23.
For Regulatory Alternative II, coal  consumption and employment impacts are
less than 0.2 percent, while for Regulatory Alternative III, impacts are
less than 0.5 percent.  These values were developed assuming constant coal-
and labor-output ratios.   The employment impacts shown do not include the
estimated increases in employment caused by the regulatory alternatives.
Therefore, the employment impacts represent maximum values.
     C.2.3.4  Financial Effects.  The aggregate capital costs of the regu-
latory alternatives are summarized in Table C-24.   Capital costs also have
been summed across member plants to determine the cost to each coke-
producing company of meeting alternative regulations.  The total capital
costs by company may be used to produce percentages that express the rela-
tion between total capital cost and the annual average net capital invest-
ment of the company and the annual cash flow of the company.  This analysis
is presented to give some insight into the distribution of the financial
effects across coke-producing firms.
     Total capital cost as a percentage of average annual net investment  is
an indicator of whether the usual sources of investment capital available
to the firm will be sufficient  to finance the additional capital costs
caused by the regulatory alternatives.  The larger this percentage, the
greater the probability that  investment needed to comply with the regula-
tory alternatives would significantly reduce investment in other areas.
This percentage provides some insights regarding the degree to which firms
will be able to finance the controls required to meet  the  regulatory alter-
natives without a  serious impact on their financial position.
     Total capital cost as a  percentage of cash flow provides similar
information.  Cash flow data  accounts for revenues, operating costs, depre-
ciation,  expenditures  on dividends, interest expenses, and taxes.  Thus,  it
is a more realistic measure of  the  funds available to  the  firm.  However,
this index may  not be  consistent across firms because  depreciation account-
ing varies across  firms.  As  with the net investment  ratio, the  larger  the
ratio,  the greater the probability  that cash flow will be  diverted  from
other  sources to  finance compliance expenditures.
                                  C-76

-------
          TABLE  C-23.   COAL  CONSUMPTION  AND  EMPLOYMENT  EFFECTS  OF
              REGULATORY  ALTERNATIVES-FURNACE  COKE, 1984S
Regulatory
Alternatives
II
III
Coal
consumption
for coke,
103 Mg/yr
-47
(-0.16)
-131
(-0.44)
Employment, jobs
Coalc
mining
-13
(-0.01)
-37
(-0.02)
Coke Steel -
plant making
-10 -83
(-0.16) (-0.03)
-27 -230
(-0.43) (-0.08)
Values in parentheses are percentage changes from baseline.

Employment impacts are based on input-output relationships and production
impacts.   Impacts on coke plant employment do not include jobs created bv
the relevant controls.

Annual labor productivity in coal  mining is estimated as 3,515 Mg per year
                               C-77

-------
         TABLE C-24.   INDUSTRY CAPITAL REQUIREMENTS  OF  REGULATORY
                     ALTERNATIVES—FURNACE COKE,  1984

                                                Capital  costs
Regulatory                                      of regulations,3
Alternative                                       106 1984  $

    57                                                -


    HI                                                68


Calculated for all  plants in operation in 1984.44 46
                                 C-78

-------
      Financial analysis is necessarily restricted to companies for which
 financial data are accessible.  Therefore, financial analysis cannot be
 conducted for some privately owned companies for which reporting has been
 restricted.   These companies are usually the smallest in a given industry,
 and they probably experience higher per unit costs of regulation and higher
 costs for securing financing than do larger companies.
      A further complication of financial  analysis is that many coke-produc-
 ing companies are wholly owned subsidiaries of larger,  highly diversified
 corporations.  Financial data are available for the parent corporations
 only.   Analysis of these data will  probably lead to the conclusion that the
 parent companies have ample resources to  finance additional  capital  costs.
 However,  the extent to which these  corporations will  make such investments,
 or will  cease some coke operations  in favor of other investment opportuni-
 ties evidencing higher rates of return, cannot be determined without knowl-
 edge of  the  required return on investment specific to the firm and the
 other investment opportunities that  exist for the firm.
      Table C-25 provides the capital  costs  as a percentage of average
 annual net investment by company  for each regulatory  alternative.  The
 average annual  net investment was calculated  from financial  records  for
 each company by averaging  net investment  data (in constant 1983 dollars)
 for 1979  to  1983.   These values are  converted to  1984 dollars  using  an
 implicit  GNP price deflator.   In some  instances,  less than 5 years of data
 were available.   In the  case  of subsidiaries,  net investment  of parent
 companies was used.   The regulatory  alternatives  impose capital costs as
 percentages  of  average annual  net investments between 0 and 5  percent.
     Table C-26 shows capital  cost as a percentage of cash flow for  firms
 for  which information was available.  Cash flow data were  derived from
 Table C-14.   The values  for 1983 were converted to 1984 dollars using an
 implicit GNP price deflator.  As for net  investment, in the case of subsid-
 iaries, cash flow of parent companies was  used.  Capital costs as percent-
 ages of cash flows range from 0 to 8 percent for the regulatory alternatives.
     The leverage ratios presented in Table C-15 indicate that coke-produc-
 ing firms are engaged in a substantial amount of external  financing.   These
firms may be  reticent (or unable)  to borrow more heavily,  especially in the
                                 C-79

-------
o
I
00
o
                   TABLE C-25   CAPITAL COSTS OF COMPLIANCE AS A PERCENTAGE OF NET INVESTMENT--
                                           FURNACE COKE PRODUCERS,  1984d
                                                	Furnace coke producer
       Regulatory	      II~      ~
       Alternative         ABC       DEFGH
II
III
1
2
1
2
0
0
1
2
3
5
1
1
2
3
1
3
1
1
1
2
0
1
1
2
       Average annual net income calculated from company profiles in Moody's Industrial Manual  Moody s
        Investor Service, New York, 1984, and Standard New York Stock Exchange Reports, Standard and Poor
        Corp., New York, 1984.  (Calculations were made on a constant 1983 dollar basis and converted to

        1984 values using a GNP implicit price deflator).

       bData on annual investment are not available for two companies.

-------
           TABLE C-26.  CAPITAL COST AS A PERCENTAGE OF ANNUAL
                CASH FLOW—FURNACE COKE PRODUCERS, 1984a
Reaulatorv
Alternative
II
III
Furnace coke producer
A
4
8
B
5
8
C
0
0
D
2
5
E
2
4
FC G
2
4
H
1
1
I
0
1
JC K
0
1
L
0
1
Cash flow data is from Table C-14 in Appendix C.  Values are converted to
1984 dollars using a GNP implicit price deflator.

There are three furnace coke producers for which cash flow data were not
available.

This company had negative cash flow.
                                C-81

-------
current economic climate for steel.   Furthermore,  financing capital  expendi-
tures by issuing additional  common stock would tend to dilute existing
stockholder equity.   Considering the low historical return on investment in
the industry, this dilution probably would be unacceptable.  An analysis of
the iron and steel industry undertaken by TBS,63 addressed the question of
external financing with regard to water pollution control expenditures.
This analysis concludes that to avoid deterioration in its financial condi-
tion, the industry is likely to reduce expenditures to modernize production
facilities rather than increase its external financing.
     In summary, the capital costs of the regulatory alternatives are in
the tens of millions of dollars range.  However, these amounts do not
appear unduly burdensome when compared with normal investment expenditures
or cash flow for companies for which data are available.
     C.2.3.5  Battery and Plant Closures.  Uneconomic batteries are those
that have marginal costs of operation greater than the price of coke.
Theoretically, these batteries are candidates for closure.  There are 14
batteries operating in the uneconomic portion of the supply curve (above
the point where price!ine intersects the supply curve) under baseline
conditions.  They are owned by 10 companies and are located in 11 plants.
Regulatory Alternative II does not add any companies or batteries to this
list.  Regulatory Alternative III forces one more battery  owned by  an
additional company into the category of uneconomic batteries.  It is impor-
tant to note, however, that this does not necessarily  imply that the regu-
lation would cause the closure of this battery.
     The decision to close a battery  is more complicated than the basic
closure rule would indicate; this is particularly  true  for integrated  iron
and  steel producers.  Continued access to profits  from  continued steel
production  is a  key factor  in the closure decision for  a captive battery.
Before  closing  or idling a  coke battery, managers  would want to  know where
they would  get  coke on  a reliable basis  in  order  to continue making steel.
The  obvious  sources to  be  investigated  include  other  plants within  the  same
company, other  companies,  and  foreign  suppliers.   As  noted in Section  C.I,
some interregional and  international  movement  of  coke occurs.
                                    C-82

-------
      Obtaining coke from offsite sources introduces two potential complica-
 tions:  the cost of transporting coke and the certainty of the coke supply.
 Obtaining coke from a nearby source might be the most profitable alternative
 to transporting coke.   If coke must be shipped over long distances, onsite
 production at a cost above the projected market price might be more profit-
 able.
      If coke must be purchased, certainty of supply is a complication.
 Steel producers prefer to have captive sources of coke to safeguard against
 supply interruptions,  and they may be willing to pay a premium for this
 security.   Producing at a cost above market price would involve such a
 premium.   Five of the  fourteen uneconomic batteries under baseline compli-
 ance produce coke at marginal  costs that are less than 5 percent above the
 market price.   Five percent does not appear to be an excessive premium to
 pay for certainty of supply.
      Several  other factors  could affect  a particular plant's  decision  to
 close a battery.   These  factors relate to the relationship  of coke quality
 to the type  of steel commodities produced,  the existence of captive  coal
 mines,  the costs  of closing a  battery  and potential  costs of  restarting it
 in the future,  and  required control  and  other expenditures.
      An alternative to closure  for  a financially  troubled company  is to
 file  for Chapter  11 bankruptcy.   This  option  allows  firms to  continue
 operating coke  plants under a restructured debt payment  schedule.  Of  firms
 owning  the 14  uneconomic batteries  under  baseline, one has  filed for
 Chapter 11, and another is  expected  to file  in the future if  the steel
 industry continues  to sustain large  losses.88  This  action may  improve  a
 firm's  competitive  situation.   McLouth Steel  Corporation, which filed  for
 bankruptcy in 1982, has modernized equipment  and reduced overhead to enable
 it to capture market share  from  larger companies.
     The developed demand model uses a single coke price, which represents
 an average quality of coke used to produce a weighted average mix of steel
 products.   If high production costs for a particular battery are associated
with a higher than average quality of coke, continued production might be
 justified.   Production  also would be justified if the firm produces only
the most highly valued  steel products.
                                 C-83

-------
     Some coke-producing firms also own coal  mines and may wish to secure
continued access to profits from coal  mining.   Because profits in the coal
sector may be subject to less effective taxation because of depletion
allowances, these profits may be extremely attractive.
     Furthermore, an integrated iron and steel  producer must consider the
question of necessary expenditures for its entire steel plant.  If the
steel facility is old or if substantial additional expenditures will  be
necessary to comply with regulations on other parts of the facility,  then
closure is more likely.
     Closure decisions are so specific to individual situations and mana-
gers' perceptions regarding their future costs and revenues that exact pro-
jections of closures should be viewed with caution.  In the current market
for steel, it is difficult to say whether uneconomic batteries will be
closed.   Of companies owning uneconomic batteries, three have cut back
capacity by closing batteries, although it is unknown whether they are
those projected as candidates for closure.88
C.2.4  Foundry Coke Impacts
     Oven coke other than furnace coke represents less than 15 percent of
U.S. coke production.  The majority of it is used as a fuel in the cupolas
of foundries.  The remainder is used for a variety of purposes, especially
for heating.
     Values of various foundry coke variables in the absence of the regu-
latory alternatives are presented in Table C-27.  These values assume base-
line compliance with the regulations listed in the footnote to the table.
     C.2.4.1  Price and Production Effects.  In developing the estimates of
price and quantity impacts, a vertical, nonparallel shift caused by each
regulatory alternative has been projected in the linear estimate of the
foundry coke supply function generated under the regulatory baseline.  This
shift is used in conjunction with an estimated elasticity demand for foundry
coke of -1.03 and is designated Scenario A.  Under this scenario, domestic-
ally produced foundry coke does not compete with imported coke.  The
reanalysis also estimates the effects of alternative regulations assuming
that foundry coke producers must compete with imports in open market sales.
In this case, consumers of foundry coke may purchase imported coke as a
                                  C-84

-------
          TABLE C-27.  BASELINE VALUES FOR ECONOMIC
             IMPACT ANALYSIS—FOUNDRY COKE, 1983

       Coke market                          Baseline values3
Price (1983 $/Mg)
Production (103 Mg)
Consumption (103 Mg)
Employment (jobs)c
Coal consumption (103 Mg)
169.
2,951
2,938
542
3,809
58b
 Baseline assumes companies meet existing regulations
 including OSHA (coke-oven emissions); State regulations on
 desulfurization, pushing, coal handling, coke handling,
 quench tower, and battery stack controls; and BPT and BAT
 water regulations.

 Calculated.   Market price for foundry coke was $149.66 in
 the fourth quarter of 1983.
£
 Calculated.   Foundry coke employment = Employment in
 byproduct coke industry x proportion of coke production
 represented by foundry coke  sector.
                            C-85

-------
perfect substitute for domestically produced coke.   The price of imports is
assumed to be constant.   As regulations cause less  domestic coke to be
produced, its price relative to imported coke rises.   Consumers are able to
purchase as much coke as before at the same price,  but a larger proportion
of sales is made up of imports.  Thus, there are quantity effects for
domestic producers, but no price effects.   This shift is designated
Scenario B.  Impacts are assessed for both scenarios.  Impacts for
Scenario B represent the maximum effect of import substitution in the
foundry coke market.
     The projected price and quantity effects are shown on Table C-28.
Under Scenario A, both price and quantity impacts are less than 0.7 percent
of baseline for Alternative II and less than 0.9 percent of baseline for
Alternative III.  Under Scenario B, there are no price impacts.  Quantity
impacts are 2.1 percent of baseline for Alternative II and 3.2 percent of
baseline for Alternative III.
     C.2.4.2  Coal Consumption and Employment Effects.  Any reductions in
foundry coke production are expected to cause reductions in the use of the
factors that produce the foundry coke.  The major inputs to foundry coke
production are coal and labor.  Labor is also an important input in coal
mining.
     The coal consumption and employment implications of the projected
reductions in coke production are shown in Table C-29 for both scenarios.
These values were developed assuming constant coal- and labor-output ratios.
The employment impacts shown do not include any employment increases caused
by the regulatory alternatives.  Consequently, the  employment impacts
represent maximum values.
     Under both scenarios and for all regulatory alternatives, effects on
employment in the coal mining industry are negligible.  Under Scenario A,
neither Alternative II nor III results in coal consumption impacts or
employment impacts in coke plants greater than a 1.2 percent change from
baseline.  Under Scenario B, coal consumption is reduced by 2.0 percent for
Alternative II, and 3.2 percent for Alternative III.   Employment in coke
plants is reduced by about the same percentages from baseline.
     C.2.4.3  Financial Effects.  The aggregate capital costs of the regu-
latory alternatives are summarized in Table C-30.  The capital requirements
                                  C-86

-------
    TABLE C-28.  PRICE AND QUANTITY EFFECTS OF REGULATORY ALTERNATIVES
                            FOUNDRY COKE, 1984a

Regulatory                   Coke price impact,       Coke quantity impact,
Alternative                      1984 $/Mg                  io3 Mg/yr

Scenario A

    11                             0.99                       -18
                                  (0.58)                      (-0.61)

    "I                            1.46                       -26
                                  (0-86)                      (-0.88)


Scenario B

  11                               0.00                       -61b
                                  (0.00)                      (-2.07)

  111                               0.00                       -94b
          	(0.00)	(-3.18)b

 Values in parentheses  are  percentage  changes  from  baseline.
 Coke  consumption  is  not  affected  due  to imports.   Imports under  Scenario  B
 are equal in  magnitude and of  opposite sign to quantity  impacts.  Under
 Scenario  A, imports  are  zero.
                                C-87

-------
    TABLE C-29.   COAL CONSUMPTION AND EMPLOYMENT EFFECTS OF REGULATORY
                     ALTERNATIVES—FOUNDRY COKE, 1984a	


                                       Coa1..         Employment (jobs)b
                                    consumption     	c—2	^	—
Regulatory                           for coke,          Coal       Coke
Alternative                          103 Mg/yr         mining      plant

Scenario A
    II                                -23             -6     ,     -3
                                      (-0.78)          (0.00)a    (-0.55)

    III                               -34             -9     ,     -5
                                      (-1.16)          (O.OOr    (-0.92)

Scenario B
    H                                -78            -22         -11
                                      (-2.05)         (-0.01)     (-2.03)

    III                              -121            -34         -17
                                      (-3.18)         (-0.02)     (-3.14)

aValues  in parentheses are percentage changes  from baseline.

bEmployment impacts are based on input-output relationships and production
  impacts.  Impacts on coke plant employment do  not include jobs created
  by the  relevant controls.

°Annual  labor productivity in coal mining is estimated as  3,515 Mg/yr per
  job.

  Zero  due  to rounding.
                                  C-88

-------
         TABLE C-30.  INDUSTRY CAPITAL REQUIREMENTS OF REGULATORY
                     ALTERNATIVES—FOUNDRY COKE, 1984

                                                   Capital costs
Regulatory                                        of regulations.
Alternative                                          10§ 1984 $
    -             —                                     _


    III                                                  12


 Calculated for all  plants in operation in 1984.48 50
                                 C-89

-------
to meet Regulatory Alternatives II and III for the foundry coke industry
are $7 million and $12 million, respectively.   Capital costs also have been
summed across member plants to determine the cost to each company of meet-
ing alternative regulations.   These company capital costs, along with
firm-specific financial data, are used to produce the same financial percent-
ages as described above for furnace coke and total capital cost as a percent-
age of net capital investment and of annual cash flow.  Financial data are
not available for many of the foundry coke producers that are privately
held companies.   Therefore, percentages for these companies are not included
in the analysis.
     Capital costs as percentages of average annual net investment for the
foundry coke producers are provided in Table C-31.  The costs of moving
from baseline to a regulatory alternative are never more than 11 percent of
the average annual net investment.  Foundry coke production plants operate
at a significantly lower production rate for the same level of investment
as compared with furnace coke production rates.   This is due to the longer
coking time for foundry coke.  Furthermore, in looking at the available
data on the age of the batteries used in the production processes within
each plant, there appears to be a correlation between the age of the battery
used and the level of compliance costs facing the firm.   The data suggest
that the foundry coke producing plants that are facing the highest pending
compliance costs are operating with batteries installed between 1919 and
1946.   Conversely, the foundry coke producers that are facing the lowest
pending compliance costs are operating, for the most part, with batteries
installed between 1950 and 1979.
                                                           tr
     Table C-32 provides capital cost as a percentage of annual cash flow.
The regulatory alternatives result in capital  costs no greater than 2 per-
cent of cash flow for foundry coke producers for which information is
available.
     Firms would use internal financing, additional equity financing, debt
financing, or perhaps some of the methods mentioned in Section C.I.5, to
make these capital expenditures.  Because many of the foundry plants are
owned by private corporations, data are insufficient to assess the eventual
sources of capital that these firms will use.   Therefore, only qualitative
statements can be made concerning the impacts of financing regulatory
                                  C-90

-------
     TABLE C-31.  CAPITAL COSTS OF COMPLIANCE AS A PERCENTAGE OF NET
                INVESTMENT—FOUNDRY COKE PRODUCERS, 1984a
Regulatory
Alternative
II
III


AA
4
11
Foundry coke producers

BB
0
1


CC
0
1
Average annual net investment calculated from company profiles in Moody's
Industrial Manual, Moody1s Investor Service, New York, 1984; Standard
New York Stock Exchange Reports, Standard and Poor's Corp., New York,
1984; and Dun and Bradstreet Financial Profiles, 1985.  (Calculations
were made on a constant 1983 dollar basis and converted to 1984 dollars
using a GNP implicit price deflator.)

There are eight foundry coke producers for which annual investment data
are not available.
                                C-91

-------
  TABLE C-32.   CAPITAL COST AS A PERCENTAGE OF ANNUAL CASH
             FLOW—FOUNDRY COKE PRODUCERS, 1984a

                                   Foundry coke producers
Regulatory Alternative          AA       BB       CC       DD

          II                     1001

          III                    1111

aCash flow data are from Table C-14 in Appendix C.   Values
 are converted to 1984 dollars using a GNP implicit price
 deflator.

 There are six foundry coke producers for which cash flow
 data are not available.
                            C-92

-------
 investments.   Any  internal  financing would  reduce  return  on  equity  by
 directly  reducing  dividends  or  by  reducing  productive  capital  expenditures.
 Debt  financing may reduce the return on  equity  by  increasing the  cost  of
 debt  financing.  Financing  regulatory  capital requirements using  new common
 stock issues will  have  a tendency  to dilute present  owner's  equity.  This
 dilution  could be  substantial.  As  an  alternative, one foundry coke firm
 entered bankruptcy status.   No  information  on the  competitive  status of
 this  firm is available.  In  the firms  for which data are  available, the
 capital requirements of the  regulatory alternative do  not appear  unduly
 burdensome.
      C.2.4.4   Battery and Plant Closures.   The  decision rule used to indi-
 cate  closure candidates among furnace  batteries also is used for  foundry
 batteries.  Any foundry battery for which the marginal cost  of operation is
 greater than the price  of foundry coke is an uneconomic battery.  According
 to this criterion  and assuming  baseline  control, no  batteries  that were in
 operation in 1984  are uneconomic to operate.  The  regulatory alternatives
 do not cause any batteries to move  into  the uneconomic region  under either
 scenario.
 C.3.  POTENTIAL SOCIOECONOMIC AND INFLATIONARY IMPACTS
 C.3.1 Compliance  Costs
      The  estimated total annualized costs to coke producers  for compliance
 with  the  regulatory alternatives are shown  in Tables C-33 and  C-34.  Costs
 for furnace and foundry producers are  differentiated because of differences
 in coke prices and control costs per unit of output.   The costs are for all
 plants in  operation in  1984 afe calculated.
      As shown  in Table  C-33, in 1984,  Regulatory Alternative II would
 result in  compliance costs of $3.5 million per year  for furnace coke pro-
 ducers.   Regulatory Alternative III would result in compliance costs of
 $12.4 million per year  for furnace coke producers.
      Compliance costs for foundry coke producers is shown in Table C-34.
 For Regulatory Alternative II, this cost is $1.3 million per year.  For
 Regulatory Alternative  III,  compliance cost is $2.9 million per year.
Combined compliance cost for furnace and foundry coke producers is
$4.8 million per year for Regulatory Alternative II.   For Regulatory
Alternative III, this cost is $15.3 million per year.
                                  C-93

-------
    TABLE C-33.   COMPLIANCE COSTS OF REGULATORY
       ALTERNATIVES—FURNACE COKE PRODUCERS,
                       1984

                                   Compliance
Regulatory Alternative         cost, 106 1984 $/yr

          II                           3.5

          III                         12.4
                        C-94

-------
 TABLE C-34.   COMPLIANCE  COSTS  OF  REGULATORY  ALTERNATIVES-
	FOUNDRY COKE  PRODUCERS,  1984
     Regulatory                           Compliance
     Alternative                     cost,  106  1984 $/yr
         n                                   i~j
         Ill                                  2.9
                          C-95

-------
C.3.3  Balance of Trade
     Recent trends indicate that imports are decreasing.   Imposition of the
regulatory alternatives is expected to slightly reverse this trend.   Some
increase in steel imports is possible also.   However, because steel  price
increases caused by coke price increases are projected to be quite small,
any increase in imports caused by the regulatory alternatives should be
minor.   Moreover, trade regulations covering steel imports may mitigate
such increases.
     In the aggregate, it appears unlikely that these regulatory alter-
natives would significantly affect the U.S.  balance of trade position,
given the small share of international trade represented by coke imports.
C.3.4  Community Impacts
     Furnace and foundry coke and steel production facilities are in Penn-
sylvania, Indiana, Ohio, Maryland, New York, Michigan, Illinois, Alabama,
Utah, Kentucky, Tennessee, Texas, Missouri,  Wisconsin, and West Virginia.
Closure of coke facilities, if they occur, could have impacts on commun-
                                    t
ities in these States.  The regulatory alternatives are not necessarily
projected to result in closures.  Potential  production decreases should not
be sufficient to generate significant community impacts.   However, further
compliance with proposed regulations could result in additional battery and
plant closures and the resulting community impacts.
C.3.5  Small Business Impacts
     The Regulatory Flexibility Act (RFA) requires consideration of the
potential impacts of proposed regulations on small "entities."  A regula-
tory flexibility analysis is required for regulations that have a "signifi-
cant economic impact on a substantial number of small entities."  For the
NESHAP for coke oven by-product plants, small entities can be defined as
small furnace and foundry coke firms.  This section addresses the require-
ments that relate to the economic aspects of the RFA.  Steps necessary for
determination of applicability of the RFA are:
          Identification of small firms impacted by the NESHAP
          Estimation of the economic impact of the NESHAP on these
          small firms.
                                  C-96

-------
     The guidelines for conducting a regulatory flexibility analysis define
a small business as "any business concern which is independently owned and
operated and not dominant in  its field as defined by the Small Business
Administration Regulations under Section 3 of the Small Business Act."  The
Small Business Administration (SBA) defines small firms in terms of employ-
ment.  Firms owning coke ovens are included in SIC 3312, which also
includes blast furnaces, steel works, and rolling mills.  The SBA has
determined that any firm that is in SIC 3312 and employs fewer than 1,000
workers will be considered small under the Small Business Act.
     Table C-35 shows employment data for all U.S. firms that operate
by-product coke ovens.  Six firms in the list--9, 14, 16, 19, 20, and
23--can be designated as small based on SBA definitions.  Because the
standard being proposed is a NESHAP and all existing and new plants will be
expected by law to comply, all plants of the small firms not currently in
compliance could be adversely impacted.  A "substantial number" of small
business is defined as "more than 20 percent of these entities."  This rule
implies that at least two firms be impacted to qualify as a "substantial
number."
     After the affected small firms are identified, the guidelines for the
RFA require an estimate of the degree of economic impact.   Four criteria
are applied in assessing whether significant economic impact occurs from
the regulation.  The first criterion determines whether annual compliance
costs increase average total production costs of small entities by more
than 5 percent.  None of the small  firms identified was found to have an
average cost increase greater than 5 percent.
     A second criterion compares compliance costs as a percentage of sales
for small entities with the same percentage for large entities.   If the
result for small  entities is at least 10 percentage points higher than for
large firms, the impact is considered significant.   Based on net sales data
in Table C-14 and compliance cost data in Tables C-31 and C-32,  one small
company is significantly impacted.   It should be noted that sales data are
not available for all  small  entities.
     The third criterion for assessing significant impact is whether capital
costs of compliance represent a "significant"  portion of capital  available
to small  entities.   The criterion recommends  examining internal  cash flow
                                  C-97

-------
     TABLE C-35.   EMPLOYMENT DATA FOR
   U.S.  FIRMS OPERATING COKE OVENS, 1983

Company                        Employment
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
48,071
52,800
28,700
9,107
163,356
37,300
16,000
32,000
210
1,300
2,400
19,200
7,300
102
7,512
164
1,245
98,722
150
150
14,518
2,200
200
                   C-98

-------
 in addition to external sources of financing.  Small, privately owned  firms
 often do not report their annual investment expenditures, so that determi-
 nation of capital availability is impossible.  No financial data could be
 located for the small coke-producing firms previously identified.
      The final criterion is whether the requirements of the regulation are
 likely to result in closures of small entities.  None of the small firms
 identified is projected to close as a result of the regulatory alternatives.
      The regulatory alternatives are unlikely to result in adverse economic
 impact on a "substantial number" of small  entities (as defined by  SBA).
 Based on the four criteria used by EPA for which assessment may be made,
 one firm may be "significantly impacted"  under the second criterion.   No
 significantly impacted firms were  identified under the other rules.
 C.3.6  Energy
      The regulatory alternatives do  not have any significant direct  energy
 impacts.   Although  some  indirect impacts are possible,  they  are likely  to
 be  minor in  nature.
      Indirect  impacts  could  include  the substitution  of  fossil  fuels  for
 coke  in  blast  furnaces or an  increase  in use  of electric  furnaces, further
 reducing the coke rate.  Some  reduction is  projected  to occur  in any  case,
 but technological limits govern the degree  to which the coke rate can be
 reduced.  Furthermore, projected coke price  increases are minor when  com-
 pared to recent and projected  fossil fuel price increases.  Of  course,  if
 imports increase, fuel will be needed to transport them.  Furthermore,  if
 imports replace domestic coke production,  excess coke oven gas, some  of
which currently is used in other parts of the steel plant, may  be replaced
by other fuels.  But if steel production decreases, there will  be some
reduction in fuel consumption.
C.4  REFERENCES
 1.   Executive Office of the Predident, Office of Management and Budget
     Standard Industrial  Classification Manual.   U.S.  Government Printina
     Office.   Washington, DC.   1972.   p. 145.
 2.   Bureau of the Census.   1982 Census of  Manufactures.   Preliminary
     Report,  Industry Series.   Blast  Furnaces and Steel  Mills (Industry
     ?J2);«a!?C82~I~33A~1(P)-   Wash1n9ton,  DC.   U.S.  Department of  Commerce.
     July 1984.   p.  3.
                                  C-99

-------
 3.   Office of Coal, Nuclear, Electric, and Alternate Fuels, Energy Infor-
     mation Administration.   Quarterly Coal Report, October - December
     1982.   Appendix A.   Publication No.   DOE/EIA-0121(82/4Q).   Washington,
     DC.   U.S. Department of Energy.  April 1983.  p. 54-55.

 4.   Office of Coal  and Electric Power Statistics, Energy Information
     Administration.  Energy Data Reports:   Coke and Coal Chemicals in
     1979.   Publication No.  DOE/EIA 0121(79).   Washington, DC.   U.S.
     Department of Energy.   October 31, 1980.   p. 33-37.

 5.   Research Triangle Institute.  Coke Battery Survey and Information from
     the American Coke and Coal Chemicals Institute and the American Iron
     and Steel Institute.  Unpublished data.   1984.

 6.   Bureau of Mines.  Minerals Yearbook.  Washington, DC.  U.S. Department
     of the Interior.  1950-1978.  (Table and page numbers vary because of
     reclassifications.)

 7.   Bureau of the Census.   Statistical Abstract of the United States:
     1978.   Table 710.  99th Edition.   Washington, DC.  U.S. Department of
     Commerce.  1978.  p. 441.

 8.   Bureau of the Census.   Statistical Abstract of the United States:
     1984.   Table 735.  104th Edition.  Washington, DC.  U.S. Department of
     Commerce.  1984.  p. 449.

 9.   Reference 7.  Table 1511.   p.  874.

10.   Bureau of the Census.   Historical Statistics of the United States:
     Colonial Times to 1970.  Part 2.   Washington, DC.  U.S. Department of
     Commerce.  1975.  p. 904-905.

11.   Bureau of the Census.   U.S. Department of Commerce.  U.S.  Exports—
     Domestic Merchandise,  SIC-Based Products by World Areas.  FT610/
     Annual.  Washington, DC.  U.S. Department of Commerce.  1978-1984.

12.   Bureau of the Census.   U.S. Imports—Consumption and General,  SIC-
     Based Products by World Areas.  FT210/Annual.  Washington, DC.  U.S.
     Department of Commerce.  1978-1981, 1983-1984.

13.   Bureau of the Census.   U.S. General Imports—Schedule A, Commodities
     by Country.  FT135/November 1981.  Washington, DC.  U.S. Department of
     Commerce.  1982.  p. 46.

14.   Bureau of Economic Analysis.  Survey of Current Business.   64:
     S16-S17.  October 1984.

15.   Reference 8.  Table 1471.   p.  832.

16.   Office of Coal, Nuclear, Electric, and Alternate Fuels, Energy Infor-
     mation Administration.  Quarterly Coal Report, April-June 1984.
     Appendix A.  Publication No. DOE/EIA-0212(84.2Q).  Washington, DC.
     U.S. Department of  Energy.  October 1984.  p. 54,  61.


                                 C-100

-------
 17.  Office of  Energy  Data Operation,  Energy  Information Administration.
      Energy Data  Reports:  Coke  and  Coal  Chemicals  in  1978.   Washington,
      DC.  U.S.  Department of  Energy.   1980.   p.  3-5.

 18.  Office of  Coal, Nuclear, Electric and Alternate Fuels,  Energy Informa-
      tion Admistration.  Coke and Coal Chemicals  in 1980.   Publication  No
      DOE/EIA 0120(80).  Washington,  DC.   U.S. Department of Energy
      November 1981.  p. 4.

 19.  Reference  8.  Table 799.  p. 486.

 20.  Bureau of  the Census, U.S.  Department of Commerce.   1977 Census  of
      Manufactures.  Washington,  DC.   U.S. Government Printing Office
      1979.   p.   2.

 21.  Bureau of  the Census, U.S.  Department of Commerce.   1972 Census  of
      Manufactures.  Washington,  DC.   U.S. Government Printing Office
      1976.   p.  33A-6.

 22.  Bureau of the Census.   Statistical Abstract of the  United States-
      1980.   Table 724.   101st Edition.   Washington, DC.  U.S. Department  of
      Commerce.   1980.   p.  437.

 23.  Bureau of  Industrial  Economics.   1984 U.S.  Industrial Outlook.  25th
      Edition.   Washington,  DC.   U.S.  Department of Commerce.  January 1984
      p.  18-3.

 24.  Standard and Poor's,  Inc.   Steel and Heavy Machinery—Basic Analysis
      Industrial  Surveys.   152(43):Sec 1,  p.  515.   New York.  October 1984.

 25.  Office  of  Coal, Nuclear,  Electric,  and  Alternate  Fuels, Energy Infor-
      mation  Administration.   Quarterly  Coal  Report,  October - December
      1983.   Appendix A.  Publication  No.  DOE/EIA-0121(83/4Q).   Washington,
      DC.  U.S.  Department  of  Energy.  April  1984.   p. 55.

 26.   Reference  4.   p. 3-5.

 27.   Reference  18.  p.  3-5.

 28.   Industrial  Economics Research Institute.  Analysis of the  U.S.  Metal-
      lurgical Coke  Industry.   Fordham University.  October 1979,   p.  41.

 29.   Reference 28.  p.   40.

 30.   Office of Coal, Nuclear, Electric, and Alternate Fuels,  Energy Infor-
     mation Administration.  Quarterly Coal Report, January - March 1984
     Appendix A.  Publication No. DOE/EIA-0121(84/1Q).   Washington   DC
     U.S. Department of Energy.  July 1984.  p. 62.

31.  Telecon.  Peterson, J., U.S. Steel Corporation, with Lohr, L.,
     Research Triangle Institute, December 10, 1984.  Operational status of
     cold idle coke batteries.
                                 C-101

-------
32.   Kerrigan, T. J.  Influences upon the Future International Demand  and
     Supply for Coke.  Ph.D. dissertation.  New York.  Fordham University.
     1977.   p. 14, 114.

33.   Reference 4.  p. 1.

34.   Reference 4.  p. 13.

35.   Reference 25.  p. 54.

36.   Emissions Standards and Engineering Division.  Standard Support and
     Environmental Impact Statement:  Standards of Performance for  Coke
     Oven Batteries.  Research Triangle Park, NC.  Environmental  Protection
     Agency.  May 1976.   p. 3-7.

37.   Bureau of Mines.  Mineral Commodity Summaries.  Washington,  DC.   U.S.
     Department of the Interior.  1984.  p. 78.

38.   Merrill Lynch, Pierce, Fenner, and Smith, Inc.  The Outlook  for Metal-
     lurgical Coal and Coke.  Institutional Report.  1980.  p. 3.

39.   Reference 28.  p. i.

40.   Reference 28.  p. ii-iii.

41.   Reference 38.  p. 1.

42.   Reference 38.  p. 5.

43.   PEDCo Environmental, Inc.  Technical Approach for a Coke Production
     Cost Model.  1979.   p. 39-50.

44.   Reference 18.  p. 33-35.

45.   Office of Energy Data and Interpretation, Energy Information Adminis-
     tration.  Energy Data Reports:  Distribution of Oven  and Beehive  Coke
     and Breeze.  Washington, DC.  U.S. Department of Energy.  April 10,
     1979.   p. 7-8.

46.   The Politics of Coke.  33 Metal Producing.  March 1980.  p.  49-51.

47.   DeCarlo, J. A., and M. M. Otero.  Coke Plants in the  United  States  on
     December 31, 1959.   Washington, DC.  Bureau of Mines,  U.S. Department
     of the Interior.  1960.  p. 4-10.

48.   Bureau of Economics, Federal Trade Commission.  Staff Report on the
     United States Steel Industry and  Its International Rivals:   Trends  and
     Factors Determining International Competitiveness.  Washington, DC.
     U.S. Government Printing Office.  November 1977.  p.  53.
                                C-102

-------
 49.  Schottman,  F. J.   Iron  and  Steel.   Mineral  Commodity Profiles, 1983.
      Washington, DC.   Bureau of  Mines,  U.S.  Department of the Interior
      1983.  p. 4-5.

 50.  Reference 18.  p.  3-4,  8.

 51.  Reference 3.  p.  53-55,  58.

 52.  Reference 25.  p.  53-55, 58.

 53.  Reference 4.  p.  3-4, 8.

 54.  Moody1s Investors' Service.  Moody1s Industrial Manual.   Vol.  1  and 2
      New York.   1984.

 55.  Moody1s Investors' SErvice.  Moody1s OTC Manual.   New York.   1984.

 56.  Standard and Poor's Corp.  Standard and Poor's New York  Stock  Exchanqe
      Stock Reports.   New York.  1984.

 57.  Dun and Bradstreet, Inc.  File reports printed for use of Research
      Triangle Institute.  January 31, 1985.

 58.  Reference  24.   p.  S16.

 59.  Temple,  Barker,  and Sloane,  Inc.   An Economic Analysis of Proposed
      Effluent  Limitations Guidelines, New Source Performance Standards  and
      Pretreatment Standards  for  the Iron and Steel Manufacturing Point
      Source Category.   Exhibit 6.   Lexington, MA.   December 1980.

 60.   Bureau of  the Census.  Annual  Survey of Manufactures.  Washington  DC
      U.S.  Department  of Commerce.   1975  - 1982.

 61.   Bureau of  the Census.  1977  Census of  Manufactures.   Vol  II
      Industry Statistics.  Part  2,  SIC Major Groups 27-34.  Washington  DC
      U.S.  Department of Commerce.   August 1981.   p.  33A-15.

 62.   Bureau of the Census.  Pollution Abatement  Costs and  Expenditures
      Current Industrial  Reports.  MA-200(year).  Washington,  DC.   U  S
      Department of Commerce.   1975  -  1983.

 63.   Reference 59.  p.  VI-4.

 64.   Reference 24.  p.  S28-S29.

65.   Reference 23.  p.  18-5.

66.  Reference 8.   Table 1408.  p. 788.
                                  C-103

-------
67.   Reference 24.  p. S22-S25.

68.   Reference 49.  p. 3.

69.   Reference 24.  p. S21.

70.   Reference 23.  p. 18-4.

71.   Reference 49.  p. 14.

72.   Reference 24.  p. S3.

73.   Reference 24.  p. S19-S20.

74.   Reference 25.  p. SI.

75.   Office of the President.  Economic  Report  of  the  President,  February
     1984.  Washington, DC.  U.S. Government  Printing  Office.   1984.
     p. 224.

76.   Reference 14.  p. S-6.

77.   Reference 28.  p. 63-111.

78.   Reference 43.  p. 1-69.

79.   Letter from  Young, E.  F., American  Iron  and Steel  Institute, to
     Pratopos, J., U.S. Environmental  Protection Agency.   May 16, 1980.

80.   Reference 43.  p. 5.

81.   Reference 43.  p. 85.

82.   Research Triangle Institute.   Economic  Impact of  NSPS Regulations on
     Coke Oven Battery Stacks.   Research Triangle  Park,  NC.   May 1980.
     p. 8-46, 8-47.

83.   Research Triangle Institute.   An  Econometric  Model  of the U.S.  Steel
     Industry.   Research  Triangle Park,  NC.   March 1981.

84.   Ramachandran, V.  The Economics of  Farm  Tractorization in India.
     Ph.D.  dissertation.   Raleigh,  NC.   North Carolina State University.
     1979.

85.   Heckman, J.  J.   Shadow Price,  Market Wages,  and Labor Supply.  Econo-
     metrica.  42:679-694.   July 1974.

86.   Reference 24.  S16,  S27.
                                   C-104

-------
87.  Reference 25.  p. 55-57, 61.

88.  Symonds, W.  C., and G. L. Miles.    It's  Every  Man  for Himself in the
     Steel Business.  Business Week.  (2897):76,78.   June  3,  1985.
                                 C-105

-------
         Appendix D





Health Risk Impact Analysis

-------
                     APPENDIX  D:  HEALTH RISK  IMPACT ANALYSIS
  D.I  REVISED  INPUT DATA

      The health risk impact analysis was repeated after proposal  of the
  standards because of revisions to the baseline data regarding plant caoacities
  and emission factors for foundry plants (as described in Chapter  6 of thi
  document).  In addition, EPA made more accurate estimates of the  latitudes and
  longitudes of plant locations by examining U.S. Geological  Survey
  topographical  maps.  These revised input data are shown on  Tables D-l and


 nf wal!!eJiPA 3lSKKdevel°Pfid stack Parameters  for modeling the control  option
 of wash-oil  scrubbers  that would achieve 90-percent  emission reduction;  they
 onHnnfl0"  ^\*~2', The  Sta(* Paramete^ for baseline  and  other  control
 options have not  changed since the analysis performed  before proposal.

      The^unit  risk estimate has  been  revised  since proposal  to  U.026/ppm
 This  revision  is  described  in  Chapter  9  of this  document.   The  revised
 estimate  was used  in the  updated  calculation  of  incidence and maximun  lifetime
 D.2 METHODOLOGY

     The Human Exposure Model was used to generate revised risk estimatps  as
 was used for the preproposal analysis.  This methodology was described in"
 Appendix E of the proposal BID.                              j^Lrioea in

     After the computer modeling was completed, the estimates for the three
 furnace and three foundry plants with the highest values for maximum lifetime
 risk were examined further.  A detailed check was made to determine  whether
 the location of the most exposed individual  was realistically placed bv the
 computer.  A review of the U.S. Geological  Survey maps revealed that  at  four
of these six plants,  there appeared  to be no possible  residential  sites where
^computer placed the most exposed individuals.   Therefore,  the  points  of

                                       baS6l1ne and controlled  «s£s if these
                                   D-3

-------
TABLE 0-1.  FURNACE COKE BV-PRODUCT  RECOVERY PLANTS LOCATION AND BENZENE EMISSIONS (kg/yr) FOR REGULATORY BASELINE
K. ........ * Coke production
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

Nn
Plant
LTV Steel, Thomas, AL
New Boston, Portsmouth, OH
Wheeling-Pitt, Monessen, PA
Lone Star Steel, Lone Star, TX
LTV Steel, So. Chicago, IL
National Steel, Granite City, IL
Interlake, Chicago, IL
LTV Steel, Gadsden, AL
Rouge Steel, Oearborn, MI
U.S. Steel, Fairless Hills, PA
LTV Steel, Warren, OH
LTV Steel, E. Chicago, IN
Annco Inc., Ashland, KY
Weirton Steel, Brown's Is., WV
U.S. Steel, Provo, UT
LTV Steel, Aliquippa, PA
Bethlehem Steel, Lackawanna, NY
National Steel, Detroit, MI
U.S. Steel, Lorain, OH
Wheeling-Pitt, E. Steubenville, WV
LTV Steel, Cleveland, OH
Arraco Inc., Middletown, OH
Bethlehem Steel, Burns Harbor, IN
LTV Steel, Pittsburgh, PA
U.S. Steel, Fairfield, AL
Bethlehem Steel, Bethlehem, PA
Bethlehem Steel, Sparrows Pt., MD
Inland Steel, E. Chicago, IN
U.S. Steel , Gary, IN
U.S. Steel, Clairton, PA
Totals
tp- n*t* rnrrpnt as of November 198
Latitude
33°32'47"
38°44'57"
40°09'46"
32°54'59"
41°41'29"
38°41'40"
41°39'22"
34°00'46"
42°18'19"
40°09'28"
41039'48"
38°30'07"
40°24'58"
40°18'43"
40°37'16"
42°49'20"
42°15'16"
41°2b'56"
40°20'36"
41°28'26"
39°29'45"
41°37'41"
40°25'34"
33°29'22"
40"36'51"
39°13'10"
4r37'53"
41°36'55"
40°18'04"

4.
Longitude
86°50'13"
82°56'01"
79°53'47"
94°42'57"
87"32'50"
90°07'42"
87°37'32"
86°02'38"
83°09'40"
74°44'32"
80°48'40"
87°26'42"
82°40'08"
80"35'16"
80°14'24"
83°07'43"
82°07 '50"
80°36 '25"
81°39'55"
84°23'15"
87°10'20"
79°57'47"
86°55'32"
75°21'13"
76°29'19"
87°27'15"
87°20'03"
79°52'21"


1,000 Mg/yr
315
364
490
507
563
570
582
758
778
916
945
948
963
1,097
1,160
1,218
1,292
1,397
1.496
1,509
1,760
1,776
1,790
1,792
1,822
2,253
3,506
3,715
4,228
6,294
45,804

decanter
2.43E+04
2.80E+04
3.77E+04
3.90E+04
4.34E+04
4.39E+04
4.48E+04
5.84E+04
5.99E+04
7.05E+04
7.28E+04
7.3UE+04
7.42E+04
8.45E+04
8.93E+04
9.38E+04
9.95E+04
1.08E+05
1.1SE+05
1.16E+05
1.36E+OS
1.37E+05
1.38E+05
1.38E+05
1.40E+05
1.73E+05
2.70E+05
2.86E+05
3.26E+05
4.08E+05
3.53E+06

Tar
storage
3.78E+03
5.88E+03
6.U8E+03
6.76E+03
6.84E+03
6.98E+03
9.10E+03
9.34E+03
1.10E+04
1.13E+04
1.14E+04
1.16E+04
1.32E+04
1.39E+04
1.46E+04
1.68E+04
1.80E+04
1.81E+04
2.11E+04
2*15E+04
2.19E+04
4i21E+04
4.46E+04
5.07E+04
6.35E+04
5.50E+05

Excess
ainiion i a -
liq. tank
2.84E+03
3.28E+03
4.56E+03
5.07E+03
5.13E+03
5.24E+03
6.82E+03
7.00E+03
8.24E+03
8.51E+03
8.53E+03
8.67E+03
9.87EK13
1.04E+04
U16E+04
1.26E+04
K58E+04
1.60E+04
1.61E+04
1.61E+04
1.64E+04
2.03E+04
3.16E+04
3.34E+04
3.81E+04
4.76E+04
4.12E+05

Light-
oil
storage
1.83E+03
2.11E+03
2.84E+03
2.94E+03
3.27E+03
3.31E+03
3.38E+03
4.40E+03
5!31E«)3
5.48E+03
5.50E+03
5.59E+03
6.36E+03
7]49E+03
8.10E+03
K02E+04
1.03E+04
O.OOE+00
1.04E+04
1 .06E+04
2!o3E+04
2.15E+04
2.42E+05

Light-
oil
sump
4.73E+03
5.46E+03
7.35E+03
7.61E+03
8.45E+03
8.55E+03
8.73E+03
1.14E+04
1.17E+04
1.37E+04
1.42E+04
1.42E+04
1.44E+04
l!?4E+04
1.83E+04
2J10E+04
2.24E+04
2.26E+04
2.64E+04
2.66E+04
O.OOE+00
2.69E+04
2.73E+04
3.38E+04
5.26E+04
5.57E+04
?!94E+04
6.60E+05

Light-oil
cond. vent
2.80E+04
3.24E+04
4.36E+04
4.51E+04
5!o?E+04
5.18E+04
6.75E+04
6.92E+04
s!44E+04
8.57E+04
9.76E+04
1.03E+05
1.08E+05
1.15E+05
1.24E+05
1.33E+05
1.34E+05
1.57E+05
1.58E+05
O.OOE+00
1.59E+05
1.62E+05
2.01E+05
3.12E+05
3.31E+05
3.76E+05
9.42E+03
3.46E+06

Wash-oil
decanter
1.20E+03
1.38E+03
1.86E+03
1.93E+03
2.14E+03
2.17E+03
2.21E+03
2.88E+03
2.96E+03
3.48E+03
3.59E+03
3.60E+03
3.66E+03
4.17E+03
4.41E+03
4.63E+03
4.91E+03
5.31E+03
5.68E+03
5.73E+03
6.69E+03
6.75E+03
O.OOE+00
6.81E+03
6.92E+03
8.56E+03
2.66E+02
1.41E+04
1.61E+04
2.01E+04
1.54E+05

Wash-oil
circ.
tank
1.20E+03
1.38E+03
1.86E+03
1.93E+03
2.14E+03
2.17E+03
2.21E+03
2.88E+03
2.96E+03
3.48E+03
3.59E+03
3.60E+03
3.66E+03
4.17E+03
4.41E+03
4.63E+03
4.91E+03
5.31E+03
5.68E+Q3
5.73E+03
6.69E+03
6.75E+03
O.OOE+OQ
6.81E+03
6.92E+03
8.56E+03
2.66E+02
1.41E+04
1.61E+04
2.01E+04
1.54E+05
(continued)

-------
                                                                              TABLE  11-1.   (continued)
O
 I
cn


Plant
No.
1
2
3
4
b
6
7
8
y
10
11
12
13
14
Ib
16
17
18 .
19
20
21
22
23
24
25
26
27
28
29
30
Total



Leaks
2.46E+04
2.46E+04
2.46E+04
2.46E+04
5.74E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
5.74E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
O.OOE+00
2.46E+04
2.46E+04
2.46E+04
b.74E+04
2.46E+04
2.4bE+04
b.74E+04
8.4bE+Ob



Tar
dewatering
6.62E+03
7.64E+03
1.03EH)4
1.06E+04
O.OOE+00
1.20E+04
1.22E+04
1.59E+04
1.63E+04
1.92E+04
1.98E+04
1.99EHH
2.02EtU4
2. JOE +04
2.44E+04
2.56E+04
2.71E+04
2.93E+04
3.14E+U4
3.17E+04
3.70E+04
3.73E+04
3.76E+04
3./6E+04
3.83E+04
4.73E+04
7.36E+04
7.80E+04
8.88E+04
O.OOE+UO
8.39E+05


Flushing-
liquor
circ. tank
2.84E+03
3.28E+03
4.41E+03
4.56E+03
5.07Et03
5.13E+03
5.24E+U3
6.82E+03
7.00Et03
8.24E+03
8.51E«-03
«.53EH>3
8.67E+03
9.37E+03
1.04E+04
1.10E+04
1.16E+04
1.26E+04
1.3bEt04
1.36E+04
1.58E+04
1.60E+04
1.61E+04
1.61E+04
1.64E+04
2.03E+04
3.16E+U4
3.34E+04
3.81Et04
4.76E+04
4.12E<-05



Benzene
storage
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
3.27E+03
O.OOE+00
O.OOE+UO
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
6.73E+03
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
2.03E+04
O.OOE+00
O.OOE+00
3.07E+04
6.10E+04



Denver
flo unit
2.74E+04
O.OOE+00
O.OOE+00
4.41E+04
O.OOE+00
4.95E+04
5.06E+04
6.59E+04
6.76E+04
7.96E+04
O.OOE+00
8.24E+04
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
1.12E+05
1.21E+05
1.30E+05
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
1.56E+05
1.58E+05
1.96E+05
O.OOE+00
O.OOE+00
3.67E+05
O.OOE+00
1.71E+06



Naphth.
melt pit
6.30E+03
O.OOE+00
O.OOE+00
1.01E+04
O.OOE+00
1.14E+04
1.16E+04
1.52E+04
1.56E+04
1.83E+04
O.OOE+00
1.90E+04
O.OOE+00
O.OOE+00
O.OOE+00
O.OUE+00
2.58E+04
2.79E+04
2.99E+04
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
3.58E+04
3.64E+04
4.51E+04
O.OOE+00
O.OOE+00
8.46E+04
O.OOE+00
3.93E+Ob



Naphth.
dry tank
3.15E+01
O.OOE+00
O.OOE+00
5.07E+01
O.OOE+00
5.70E+01
5.82E+01
7.58E+01
7.78E+01
9.16E+01
O.OOE+00
9.48E+01
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
1.29E+02
1.40E+02
1.50E+02
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
1.79E+02
1.82E+02
2.25E+02
O.OOE+00
O.OOE+00
4.23E+02
O.OOE+00
1.97E+03



TBFC
O.OOE+00
2.55E+04
3.43E+04
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
8.12E+04
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
2.60E+05
O.OOE+00
O.OOE+00
4.01E+05



UWFC
8.51E+04
O.OOE+00
O.OOE+00
1.37E+05
O.OOE+00
1.54E+05
1.57E+05
2.05E+05
2.10E+05
2.47E+05
O.OOE+00
2.56EK)5
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
3.49E+05
3.77E+05
4.04E+05
O.OOE+00
O.OOE+00
4.80E+05
O.OOE+00
4.84E+05
4.92E+05
6.08E+05
O.OOE+00
O.OOE+00
1.14E+06
O.OOE+00
5.79E+06



BTX
storage
O.OOE+00
O.OOE+00
O.OOE+00
2.94E+03
3.27E+03
3.31E+03
O.OOE+00
O.OOE+00
4.51E+03
O.OOE+00
O.OOE+00
5.50E+03
O.OOE+00
6.36E+03
O.OOE+00
O.OOE+00
7.49E+03
O.OOE+00
O.OOE+00
O.OOE+00
1.02E+04
1.03E+04
O.OOE+00
O.OOE+00
O.OOE+00
2.51E+02
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
5.41E+04



Tar inc.
suinp
2.99E+04
3.46E+04
4.66E+04
4.82E+04
b.35E+04
5.42E+04
5.53E+04
7.20E+04
7.39E+04
8.70E+04
8.98E+04
9.01E+04
9.15E+04
1.04E+05
1.10E+05
1.16E+05
1.23E+05
1.33E+05
1.42E+05
1.43E+05
1.67E+05
1.69E+05
1.70E+05
1.70E+05
1.73E+05
2.14E+05
3.33E+05
3.53E+05
4.02E+05
5.03E+05
4.35E+06
	 -


Total Benzene
By Plant
2.51E+05
1.74E+05
2.26E+05
3.91E+05
2.44E+05
4.37E+05
4.42E+05
5.68E+05
5.87E+05
6.82E+05
3.46E+05
7.10E+05
3.52E+05
4.04EK)5
5.40E+05
4.39E+05
9.59E+05
1.03006
1.10E+06
5.38E+Ob
6.34E+05
1.12E+06
3.99E+05
1.31E+06
1.33E+06
1.63EI-06
1.26E+06
1.55E+06
3.06E+06
1.32E+06
2.40E+07

               Data  current  as  of  November  1984.

-------
TABLE l)-2.  FOUNDRY CUKE BY-PRODUCT  RECOVERY  PLANTS  LOCATION  AND  BENZENE  EMISSIONS  (kg/yr)  FOR  REGULATORY  BASELINE


No.
1.
I:
&
• 7.
8.
9.
10.
11.
12.
13.
14.

Nnf


Plant
Chattanooga Coke, Chattanooga, TN
IN Gas Terre Haute IN
Koppers, Toledo, OH
Empire Coke, Holt, AL

Carondolet, St. Louis, MO
AL Byproducts, Keystone, PA
Citizens lias, Indianapolis, IN
Jim Walters, Birmingham, AL
Shenango, Pittsburgh, PA
Koppers, Woodward, AL
AL Byproducts, Tarrant, AL
Detroit Coke, Detroit, MI
Totals
p- lists rnrrpnt as of November 19£


Latitude
3b002'16"
39°26'48"
41°40'10"
33° 14 '25"
42°08'43"
42°58'56N
38° 32 '08"
40°05'12"
39°45'16"
33°33't7b
40°28'491I~
33°26'13"
33°34'57"
420H'19"

14.
C

Longitude
85018'U"
87"23'47"
83029'31"
87°3U'n"
80001'32"
78°56'19"
90° 16 '05"
7b°18'59"
86°06'49"
86°48'38"
80°03'34"
86°57'50"
86°46'47"
83°09'16"


)oke production
capaci ty ,
1,000 Mg/yr
130
132
157
161
207
299
330
402
477
499
521
563
583
617
5,078


Tar
decanter
4.70E+03
4.78E+03
5.68E+03
5.83E+03
7.49E+03
1.08E+04
1.19E+04
1.45E+04
1.73E+04
1.81E+04
2.04E+04
2.11E+04
1.84E+05


Tar
storage
7.33E+02
7.44E+02
8.85E+02
9.08E+02
1.17E+03
1.69E+03
1.86E+03
2.27E+03
2.69E+03
2.81E+03
2.94E+03
3.18E+03
3.29E+03
3.48E+03
2.86E+04

Excess
ammonia
liq. tank
8.54E+02
8.67E+02
1.03E+03
1.06E+03
1.36E+03
1.96E+03
2.17E+03
2.64E+03
3.13E+03
3.28E+03
3.42E+03
3.70E+03
3.83E+03
3.34E+04

Light-
oil
storage
4.07E+02
4.13E+02
O.OOE+00
b.04E+02
O.OOE+00
9.36E+02
O.OOE+00
1.26E+03
O.OOE+00
1.56E+03
1.63E+03
1.76E+03
1.83E+03
O.OOE+00
1.03E+04

Light-
oil
sump
1.05E+03
1.07E+03
O.OOE+00
1.30E+03
O.OOE+00
2.42E+03
O.OOE+00
3.26E+03
O.OOE+OQ
4.04E+03
4.22E+03
4.56E+03
4.72E+03
O.OOE+00
2.66E+04


Light-oil
cond. vent
6.25E+03
6.34E+03
O.OOE+00
7.74E+03
O.OOE+00
1.44E+04
O.OOE+00
1.93E+04
O.OOE+QO
2.40E+04
2.50E+Q4
2.71E+04
2.80E+04
O.OOE+00
1.58E+05


Wash-oil
decanter
2.67E+02
2.71E+02
O.OOE+00
3.30E+02
O.OOE+00
6.14E+02
O.OOE+00
8.25E+02
O.OOE+00
1.02E+03
1.07E+03
1.16E+03
1.20E+03
O.OOE+00
6.75E+03

Wash-oi 1
circ.
tank
2.67E+02
2.71E+02
O.OOE+00
3.30E+02
O.OOE+00
6.14E+02
O.OOE+00
8.25E+02
O.OOE+00
1.02E+03
1.07E+03
1.16E+03
1.20E+03
O.OOE+00
6.75E+03
(continued)

-------
                                                                      TABLE D-2.  (continued)
— ^-=-^"'

Plant
no.
1
2
3
4
5
6
7
a
9
1U
11
12
13
14
Total


Leaks
2.24E+U4
2.24E+04
O.UOE+00
2.24E+04
U.OOE+UO
2.24E+04
O.UOE+UO
b.22E+04
O.OOE+00
2.24E+04
2.24E+04
2.24E+04
2.24E+04
O.OUE+UO
2.31E+U5


Tar
dewatering
1.28E+03
1.30E+U3
1.55E+03
O.OOE+UU
2.U4E+03
2.9bE+03
3.26E+03
3.97E+03
4.71E+U3
4.93E+U3
5.14E^03
5.56E+03
5.75E+U3
6.09E+03
4.85E+04

Flushing-
Hquor
circ. tank
8.54E+02
8.67Et02
1.03E+03
1.06E+03
1.36E+U3
1.96E<-03
2.17Et03
2.64E>U3
3.13E+03
3.28E+U3
3.42E+03
3.70E+03
3.83E+03
4.05E+03
3.34E+04


Benzene
storage
U.OOE+00
O.OOE+00
O.OOE+00
O.UOE+00
O.OOE+00
O.OOE+00
O.OOE+UO
1.26E+03
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
1.26E+03


Denver
flo unit
8.25E+03
8.37E+03
O.OOE+00
1.02E+04
O.OOE+00
O.OOE+00
O.OOE+00
2.55E+04
3.03E+04
3.17E+04
O.OOE+00
O.OOE+00
3.70E+04
O.OOE+00
l.^lE+Ob


Naphth.
melt pit
1.90E+03
1.93E+03
O.OOE+00
2.3bE+03
O.OOE+00
O.OOE+00
O.OOE+00
b.87E+03
6.96E+03
7.29E+03
O.OOE+00
O.OOE+00
8.51E+03
O.OOE+00
3.48E+04


Naphth.
dry tank
9.49E+00
9.64E+00
O.OOE+00
1.18E+01
O.OOE+00
O.UOE+OU
O.OOE+00
•2.93E+01
3.48E+01
3.64E+01
O.OOE+00
O.UOE+00
4.26E+01
O.OOE+00
1.74E+02


TBFC
O.OOE+OU
O.OOE+00
8.02E+03
O.OOE+00
O.OOE+00
O.OOE+00
O.UOE+OU
O.OOE+00
O.OUE+OU
O.OOE+00
O.OOE+UO
2.88E+04
O.OOE+00
O.OOE+00
3.6HE+04


UWFC
2.b6E+04
2.60E+U4
O.UUE+00
3.17E+04
O.OOE+00
O.UOE+UO
O.UOE+UO
7.92E+04
9.4UE+04
9.84E+U4
U.UUE+UO
0;UOE+00
l.lbE+Ub
U.UOE+00
4.7UE+Ob


BTX
storage
O.UUE+OU
O.UOE+UO
O.OOE+00
O.UUE+OU
U.OOE+OU
U.OOE+UO
U.UUE+UU
1.26E+U3
U.UOE+OU
l.bbE+u3
U.UUE+UU
O.OUE+UO
O.OUE+OU
O.OOE+UU
2.82E+03


Tar inc.
b.8UE+03
b.89E+U3
7.01E+03
7.19E+U3
9.24E+U3
1.34E+U4
1.47E+04
1.79E+U4
2.13E+U4
2.23E+U4
2.J3E+04
2.blE+U4
2.6UE+U4
2.7bE+U4
2.27E+Ub


Total Benzene
by plant
U.U6E+U4
S.lbE+04
2.b2E+04
9.29E+04
2.2VE+U4
7.41E+U4
J.tilE+U4
2.3bE+Ob
1.84E+Ub
2.4WE+Ub
l.l<^E+Ub
1.48E+OS
2.H4E+03
b.7bE+U4
l.b9E+Ub
Note:  Data current as of November 1984.

-------
 TABLE D-3.  PARAMETERS FOR 90% EMISSION REDUCTION OPTION  (WASH-OIL  SCRUBBER)


                Scrubber   Vertical  cross-   Diameter    Stack  gas   Stack  gas
                 height,     sectional        of vent,    velocity,  temperature,
  Source           ma        area, m2a           m         m/sb        °KC
Storage tanks
  for Light oil,
  8TX, or benzene   4.3

Excess ammonia-
  1iquor tank      10.1

Tar storage        12.3

Tar dewatering
  tanks            12.3

Tar decanter,
  tar intercepting
  sump, and
  flushing-liquor
  circulation
  tankd             4.6

Light-oil condenser
  vent, wash-oil
  decanter, wash-
  oil circulation
  tank6             4.3
 23.6


101.9

240.0


240.0
 29.2
0.191


0.191

0.191


0.191
0.191
0.46


0.46

1.149


1.149
1.149
305


305

311


311
311
 23.6
0.191
1.149
305
aAssumed wash-oil  scrubber attached to the side  of the  source  and  is the  same
 height and same vertical  cross-sectional  area

bStack gas velocity derived from costing  design  flow  of 0.013  m-^/s  for  storage
 tanks containing light oil, BTX, or benzene  and the  excess  ammonia liquor
 tank, and from costing design flow of 0.03 m3/s for  other sources, which have
 greater emissions.

cStack gas temperature is  the temperature of  cool  wash  oil for sources  with
 cool gases and is slightly higher (38 °C) for sources  with  hot  gases.

^Assumes this group of sources all  vented to  one scrubber attached  to the
 flushing-liquor circulation tank.

eAssumes this group of sources all  vented to  one scrubber attached  to the
 wash-oil circulation tank.
                                    D-8

-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO. ' 2.
EPA-450/3-83-0165
4. TITLE AND SUBTITLE
Benzene Emissions from Coke By-Product Recovery
Plants - Background Information for Revised 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
DAA for Air Quality Planning and Standards
Office of Air and Radiation
U.S., Environmental Protection Agency
Research Triangle Park, North Carolina 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
June 1988
6. PERFORMING C ^GANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
  National emission  standards to control emissions of benzene  from new and existing
  coke by-product   recovery plants are being promulgated  under Section 112 of the
  Clean Air Act.   This  document contains summaries of public comments, EPA responses,
  and a discussion of differences between the proposed and  revised proposed standard.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Air pollution Steel industry
Pollution control
National emission standards
Industrial processes
Coke by-product recovery
Hazardous air pollutants
Benzene
18. DISTRIBUTION STATEMENT
Unlimited

b.lDENTIFIERS/OPEN ENDED TERMS
Air Pollution Control
Benzene
Stationary Sources
19. SECURITY CLASS (This Report/
Unclassified
20. SECURITY CLASS (This page)
Unclassified
c. COSATI Field/Group
13B
21. NO. OF PAGES
255
22. PRICE
' EPA Form 2220-1 (R«v. 4-77)   PREVIOUS EDITION js OBSOLETE

-------