ENVIRONMENTAL  PROTECTION  AGENCY
                                                                  DRAFT
                            Background  Information
                           -     and  Final                         JUN 2 8 1985
                        Environmental  Impact  Statement
                     for Coke By-Product  Recovery  Plants

                                 Prepared by:
Jack R. Farmer                                                     (Date)
Director, Emission Standards  and  Engineering  Division
U.S. Environmental Protection Agency
Research Triangle Park,  NC  27711

1.  The promulgated national  emission  standards  would  limit  emissions of
    benzene from existing and new coke by-product  recovery plants.  The
    promulgated standards 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.  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 and Engineering Division  (MD-13)
    U.S. Environmental Protection Agency
    Research Triangle Park,  North Carolina  27711
    Telephone:  (919) 541-5578

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
                                     1-1

-------
                                  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  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 19 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 summary of  comments
and responses serves as the basis for the revisions made to  the  standard
between proposal and promulgation.
1.1  SUMMARY OF CHANGES SINCE PROPOSAL
              [Subject to change  due to  undecided  policy issues]
     Since proposal, the data base has been revised in  several  respects  to
reflect current industry operating status and  the  differences in the impacts
of control options on furnace as  opposed to foundry plants.   Separate
consideration of environmental, health,  cost,  and  economic  impacts for  the two
industry segments has resulted in several changes  to  the control  options
selected as the basis of the proposed standards.
     The major change since proposal applicable to  furnace  plants  is the
requirement for gas blanketing controls  rather  than wash-oil  scrubbers  for
excess  ammonia-liquor storage.  Also, proposed  controls  for  light-oil  and
benzene-toluene-xylene (BTX) storage at  furnace plants  have  been eliminated
from the final standards.  For foundry plants,  the  proposed  control
requirements for storage tanks containing benzene,  light-oil, BTX, or  excess
ammonia-liquor have been eliminated  from the final  rule.   These  issues  are
discussed in response to comment  5.1.
     The final standards require  installation  of  the  controls proposed  for the
major sources of benzene emissions at furnace  and  foundry  plant  sites.   As
proposed, installation of tar-bottom final  cooler  technology is  required for
control of the most significant benzene  emission  source  in  either
industry segment—uncontrolled naphthalene  handling and  processing operations.
Gas blanketing coupled with leak  detection  and  repair requirements also
                                     1-2

-------
remains the basis of the standards for  affected  process  vessels,  including
tar-intercepting sumps.  No changes  have occurred  in  proposed  control
requirements for other sources.
     Other changes to the standards  have been made  for clarifying  purposes.
The definition of "coke by-product recovery  plant"  has been  revised
specifically to exclude form-coke plants.  New definitions  for "furnace"  and
"foundry" plants also have been added to distinguish  these  industry  segments
in terms of the type of coke produced.  For  example,  furnace plants  produce
furnace coke, while foundry plants may  produce foundry coke  or furnace  and
foundry coke.
     Other clarifying changes include elimination  of  the  definition  for
"light-oil storage tank", addition of a new  definition for  "exhauster",  and
revision of the proposed definitions for "excess-ammonia  liquor  storage  tank"
and "tar decanter".  The revised definition  of "tar decanter"  and  the  new
definition for "exhauster" are discussed in  response  to  comments  3.2 and
3.3, respectively.  The final regulations  define  "excess  ammonia-liquor
storage tank" as follows:
     "Any tank, reservoir, or other  type container  used  to  collect or
      store a flushing-liquor solution  prior to  ammonia  or  phenol
      recovery, whichever process comes first."
The intent of this definitional change  is  to clarify  that applicable sources
upstream of ammonia recovery or upstream of  phenol  recovery  are  considered
excess ammonia liquor-storage tanks, but not sources  located between ammonia
and phenol recovery process operations.

1.2  SUMMARY OF IMPACTS OF PROMULGATED  ACTION

1.2.1  Alternatives to Promulgated Action
     The regulatory alternatives are discussed in  Chapter 6 of the background
information document (BID) for the proposed  standards  (EPA-450/3-83-016a, May
1984).  These regulatory alternatives reflect the  different  levels of  emission
control from which one is selected as the  basis  of  the proposed  standards.
These alternatives remain the same for  the promulgated standards.  Although
certain source-specific changes have been  made since  proposal  (see Section 1.1
above) the final standards are based on application of the  best  available
technology (BAT).  The approach used to select BAT  is  discussed  in the
preamble to the proposed rules at 49 FR 23533.   Also  examined  was  an
alternative requiring a level of control more stringent  than BAT.  However,
                                     1-3

-------
this higher level of control was not selected because the  additional  risk
reduction achieved was small in relation to the additional  costs.   As
described in §61.134 of the final standards, an owner or operator  may  apply
for permission to use alternative control methods that  provide  an  emission
reduction equivalent to or greater than BAT controls.

1.2.2  Environmental and Energy Impacts of Promulgated  Action
     The environmental and energy impacts of the proposed  standards  are
discussed in Chapter 7 of the BIO.  The estimated environmental  impacts  have
been revised since proposal to reflect the current operating  status  of the
industry.  These changes are discussed in Chapter 7 of  this document,  entitled
"Environmental Impacts."  Revised environmental impact  tables and  emission
factor data are presented in Appendix A.
     Implementing the final standards would reduce nationwide benzene  emissions
at 44 furnace and foundry plants from their current level  of  about 25,900  Mg/yr
to about 3,000 Mg/yr, an 88-percent reduction.  Total uncontrolled nationwide
emissions of benzene and other volatile organic compounds  (VOC's)  also would be
reduced from their current estimated level of 170,900 Mg/yr to  about  35,000
Mg/yr, an 80-percent reduction.  Assuming recovery of 21.3 a of gas/min/Mg of
coke/day, the final standards would result in a national energy savings  of
approximately 15.0 TJ/yr from recovered coke oven gas at furnace plants  and
about 11.3 TJ/yr at foundry plants.  A net nationwide energy  savings  of  about
1,800 TJ/yr would be expected.  Impact calculations for energy  requirements  and
coke oven gas recovery estimates are shown in Tables A-9 and  A-10.

1.2.3  Cost and Economic Impacts of Promulgated Action
     Control costs for model by-product recovery plants are discussed  in Chapter
8 of the BID for the proposed standards.  This analysis has been updated since
proposal to reflect the current industry operating status.  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 8 of  this
document, entitled "Cost Analysis."  The revised cost analysis  is  presented  in
Appendix B.
     Based on the revised analysis, estimated national  capital  cost  of the final
standards for the 44 plants are estimated at about $49.^ million over  baseli.ne
costs (1984 dollars).  Of these costs, approximately $41.8 million and $7.6
                                     1-4

-------
million are attributable to furnace and foundry plants, respectively.  Total
annualized costs for the 44 plants are approximately $4.2 million/yr.
Annualized costs for furnace plants account for an estimated $2.8 million/yr
compared to $1.4 million/yr for the 14 foundry plants.  The overall  cost
effectiveness of the final  regulation is about $180/Mg of benzene removed or
about $31/Mg of total VOC removed.  The cost effectiveness of controls for
furnace plants is about $130/Mg of benzene removed or $22/Mg of total VOC
removed; the estimated cost effectiveness for foundry plants is about $980/Mg
of benzene removed or about $164/Mg of total VOC removed.
     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 proposal, is  discussed in Chapter 9 of this document;  further
information is presented in Appendix C.  Based on the revised economic analysis,
the price of foundry coke is projected to increase by about $0.99/Mg, while the
price of furnace coke is projected to increase by about $0.13/Mg.  For both
furnace and foundry coke, the price increases are less than 1 percent above
baseline values.  Although  many furnace and foundry plants are in marginal
economic condition, the revised analysis predicts no closures as a result of the
final standards.

1.2.4  Health Risk Impacts  of Promulgated Action
     The quantitative risk  assessment conducted for the proposed standards is
discussed in Appendix E of  the proposal BID; further information is  in the
preamble to the proposal at 49 FR 23525.  The risk assessment has been revised
since proposal to reflect the current industry operating status and  adjusted
emission factors.  Other changes include the application of a new benzene unit
risk factor, which is 17 percent higher.  Further information regarding these
changes is provided in Chapter 10 of this document, "Quantitative Risk
Assessment" and in Appendix D.
     Annual leukemia incidence associated with uncontrolled benzene  emissions at
44 plants is estimated at 2.9 cases/yr.  Implementation of the final  standards
would achieve a total incidence reduction of 2.6 cases/yr.  Incidence reduction
for the furnace plant industry segment is estimated at 2.4 cases/yr;  incidence
reduction at foundry plants is about 0.2 cases/yr.  The maximum lifetime risk
(MLR) at the current level  of control is predicted at 6.3 x 10~3 for  furnace
plants and 1.1 x 10-3 for foundry plants.  The final standards are expected to

                                     1-5

-------
reduce the MLR for furnace and foundry plant industry segments to approximately
5.3 x 10~4 and 7.5 x 10~5, respectively.
1.2.5  Other Considerations
   •  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.  However, an increased cyanide
(HCN) concentration in wastewater should be 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.
     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.

     1.2.5.2  Environmental and  Energy Impacts 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 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
current nationwide level of 25,900 Mg/yr.  Total uncontrolled emissions of
benzene and other VOC's also would remain at their current level of about
170,900 Mg/yr.  No net nationwide savings in energy use would be achieved as
a result of recovered coke oven  gas if implementation of the standards  were
delayed.

     1.2.5.3  Urban and Community Impacts.  The beneficial urban and community
impacts of the final 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 proposed
standards are discussed in Section 9.3.4 of the BID.  As indicated in this
analysis, compliance with existing and other proposed environmental regulations
may result in closure of plants  in marginal economic condition, with resulting

                                     1-6

-------
community impacts.  However, no closures are predicted as a  result  of  these
promulgated standards.
                                     1-7

-------
                         2.  SUMMARY OF PUBLIC COMMENTS
     A total of ly letters commenting on the proposed standards and the
background information document (BID) for the proposed standards were received.
A list of commenters, their affiliations, and the U.S. Environmental Protection
Agency (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
     Chapter 13
     Appendix A
     Appendix B
     Appendix C
     Appendix D
Selection of Source Category
Legal and Policy Considerations
Selection of Standards
Emission Control Technology
Environmental  Impact
Cost Impact
Economic Impact
Quantitative Risk Assessment
Equipment Leak Detection and Repair
Reporting and Recordkeeping
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
IV-D-1
IV-D-2
IV-D-3
IV-0-4
IV-D-5
IV-D-6
IV-D-7
IV-D-8
Commenter and affiliation

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
Footnote on last page of table,
                       (continued)
                                      2-2

-------
                             TABLE 2-1 (continued)
Docket item number9

IV-D-9
IV-D-10
IV-D-11
IV-D-12
IV-0-13
IV-D-14
IV-D-15
Commenter and affiliation

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
300 North Lee Street, Suite 306
Alexandria, Virginia  22314

Lecil  M. Col burn
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 N. 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, D.C.  20005

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

David M. Anderson, Director
Environmental and Governmental
  Programs
Bethlehem Steel Corporation
Bethlehem, Pennsylvania  18016
Footnote on last page of table.
                       (continued)
                                      2-3

-------
                             TABLE 2-1 (continued)
Docket item number9                         Commenter and affiliation

IV-D-16                                     Terry McGuire, Chief
                                            Technical Support Division
                                            California Air Resources Board
                                            1102 Q Street
                                            P.O. Box 2815
                                            Sacramento, California  95812

IV-D-17                                     Moyer B. Edwards
                                            Director, Environmental Control
                                            Alabama By-Products Corporation
                                            First National-Southern National
                                              Building
                                            P.O. Box 10246
                                            Birmingham, Alabama  35202

IV-D-18                                     Neil Jay King, Esq.
                                            Wilmer, Cutler & Pickering
                                            1666 K Street, N.W.
                                            Washington, D.C.  20006

IV-D-19                                     Barbara Patala, Acting Chairman
                                            Committee on Environmental Matters
                                            National Science Foundation
                                            Washington, D.C.  20550
aThe docket number for this project is A-79-16.  Dockets are on file at EPA
 Headquarters in Washington, D.C., and at the Office of Air Quality Planning and
 Standards (QAQPS) in Durham, North Carolina.


                                      2-4

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

3.1  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
     predicted at proposal.  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 to the commenter's  concerns  regarding  the
     methodology and assumptions applied to the quantitative  risk  assessment  for
     coke by-product recovery plants are contained  in  Chapter 10.  This response
     is intended to provide a summary update of the  revised risk  impacts  in
     support of EPA's decision to select coke by-product recovery  plants  as  a
     source category for regulation.
          As at proposal, the Human Exposure Model  (HEM) was  used  to estimate
     the average annual benzene concentrations to which persons within 50 km  of
     44 coke by-product recovery plants are exposed.   Based on  a  revised  unit
     risk factor, leukemia incidence associated with uncontrolled  benzene
     emissions of nearly 26,000 Mg/yr is estimated  at  2.9 cases/yr.  Of the
     estimated 2.9 cases/yr, furnace plants and foundry plants  account for
     approximately 2.6 and 0.2 cases/yr, respectively. As a result of  exposure
     to these benzene concentrations, the maximum lifetime risk for  the most
     exposed population is estimated at 6.3 x 10~3  for furnace  plants  and 1.1  x
     10~3 for foundry plants.
          These revised risk estimates differ little from risk  estimates
     predicted at proposal.  These incidence and maximum lifetime  risk estimates
     are considerably higher than for other stationary sources  for which
     regulations already have been promulgated.  Controls are available to
     reduce benzene emissions from by-product recovery plants at  a reasonable
     cost.  Based on the magnitude of benzene exposures associated with
     emissions from this source category, the resulting estimated  maximum
                                      3-1

-------
     lifetime risk and leukemia incidence in  the exposed population  (including
     consideration of the associated uncertainties),  and the availability of
     control  techniques to reduce these emissions,  the Administrator has
     determined that coke by-product recovery plants  constitute a  significant
     emission source category for which national  emission standards  are
     warranted.
          The uncertainties and assumptions  associated with  the quantitative
     health risk assessment are discussed in  the preamble to the proposed rules
     (49 FR 23525) and are not repeated here.  However, as discussed at 49 FR
     23526, EPA considers that the unit risk  factor for benzene and  the
     evaluation of public exposure represent  plausible, if not  conservative,
     estimates of actual  conditions.  In this context, EPA believes  that  such
     estimates of the health hazard can and  should  play an important role in the
     regulation of hazardous air pollutants.

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

     Response:  In considering the commenter's concerns, the data  base has been
     revised  since proposal to indicate the  environmental, health,  cost,  and
     economic impacts of  the control options  separately for  furnace  and foundry
     plants.   Merchant plants generally fall  under  the foundry  plant industry
     segment.  As discussed in response to comment  7.2, emission factor
     adjustments have been made to account for the  lower emission  rates
     characteristic of foundry plants.  However, nationwide  uncontrolled  benzene
     emissions from the 14 foundry plants are estimated at about 1,700 Mg/yr
     even with reduced emission factors.  Consequently, EPA  considers  that
     benzene  emissions from the foundry plants constitute a  significant source
     of benzene emissions on their own merits.  The EPA does not agree that
                                      3-2

-------
     these uncontrolled emissions pose little or no health  threat.   Annual
     incidence associated with foundry plants is estimated  at  0.2  cases/yr.
          Neither do we agree that foundry plants have been  represented  properly
     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.
          The revisions made to the standards since proposal,  as applicable  to
     foundry plants, are discussed in response to comment 5.1.   The final
     standards would reduce nationwide benzene emissions  at  foundry plants  by
     about 84 percent, from approximately 1,700 Mg/yr to  about  270  Mg/yr.
     Annual incidence associated with uncontrolled emissions would  be  reduced
     from about 0.24 cases/yr to about 0.02 cases/yr.  The  maximum  lifetime  risk
     would be reduced from about 1.1 x 10~3 under current controls  to  about
     7.5 x 10"5.  Capital and annualized costs (estimated at $7.6  million and
     $1.4 million/yr, respectively) are considered reasonable; the  economic
     analysis does not predict adverse economic impacts resulting  from
     implementation of these standards in terms of closures  or  price  increases.
     The EPA considers that the standards will achieve a  significant  reduction
     in emissions and associated health risks without imposing  costs  that
     outweigh the public health benefits achieved.

3.3  Comment:  One commenter in two comments (IV-D-3  and  IV-D-8) requests that
     the final regulation be clarified to exclude form-coke  plants.  In  support,
     the commenter cites separate conversations with  EPA  personnel  who stated
     that the 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 §61.131 of the  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 significant emissions based on  review  of  process
     description information.
                                      3-3

-------
     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.
                                 3-4

-------
                      4.  LEGAL AND POLICY CONSIDERATIONS

4.1  Comment:  Commenter IV-D-13 argues that EPA has misconstrued  Section  112
     of the CAA in the Agency's best available technology  (BAT) approach.   The
     commenter states that Section 112 does not allow rejection of more
     effective controls due to cost effectiveness or cost-benefit; Section  112
     requires application of controls providing the best emission  reduction to
     provide "an ample margin of safety to protect the public health."

     Response:  The commenter argues that EPA could not lawfully consider  costs
     in setting standards under Section 112.  The EPA disagrees.
          Neither Section 112 nor its legislative history  states that costs may
     not be considered.  On the contrary, the legislative  history  indicates
     that costs may be taken into account.  Section 112 was enacted  in  1970.
     All of the bills Congress considered provided that EPA could  consider
     feasibility in setting standards for hazardous air pollutants.  For
     example, the Senate Report notes:
           The Committee recognizes that some of these hazardous
           pollutants . . . are present in nearly all raw  materials  . .  .
           Recognizing that complete control . . . may not be ...
           practicable, the Committee has provided the Secretary with
           authority to differentiate among categories ....
     S. Rep. No. 1196, 91st Cong., 2d Sess. 20 (1970), reprinted in
     Congressional  Research Service, A Legislative History of the  Clean  Air
     Act Amendments of 1970 at 420.
          There are sound policy reasons for interpreting  Section  112 to permit
     consideration of cost and feasibility when carcinogens are concerned.   As
     EPA has consistently observed in Section 112 rulemakings, there is  no
     direct evidence that air pollutants cause cancer, i.e., no studies  show an
     increased incidence of cancer in humans due to exposure to particular
     pollutants in the ambient air.  Instead, regulation of airborne
     carcinogens is based on studies of workers exposed in the workplace and
     animals exposed in laboratory experiments.  However,  the workplace  and
     laboratory exposures are much higher, generally many  orders of  magnitude
                                     4-1

-------
higher, than the levels of these carcinogens  found  in  the  ambient  air.   It
is only by extrapolating from the higher  exposures  to  ambient  levels  that
the Agency concludes that the air pollution due  to  these carcinogens
threatens public health within the meaning of  Section  112.   This
extrapolation is based on the "no threshold"  assumption, i.e.,
that any exposure to a carcinogen presents some  risk,  although  the risk
becomes increasingly small as the exposure does.  Under the  no  threshold
assumption, emission standards for carcinogens under Section 112  could
prevent all risks only by preventing all  exposures  and all emissions.
     However, EPA does not believe that Congress  intended  Section  112
standards to prevent all risks.  Section  112(b)(l)(B)  simply requires  that
standards be "at the level which in [EPA's] judgment provides  an  ample
margin of safety to protect the public health  from  such hazardous  air
pollutant."  First, the requirement for an ample  margin of  "safety"  does
not imply eliminating all risk.  For example,  the Occupational  Safety  and
Health Act requires standards that "provide safe  or healthful  employment
and places of employment."  As the Supreme Court  held  in reviewing an  OSHA
benzene standard,"  . .  . safe is not the  equivalent of  'risk-free.'"
Industrial Union Dept., AFL-CIO v. American Petroleum  Institute.
448 U.S. 607, 642 (1980).
     Second, Congress did not contemplate that Section 112 would  require
the elimination of  all risk from, and all emissions of, carcinogens.
Virtually every basic industry in our society—chemicals,  petroleum,
electric power, and metals—emits one or  more  of  the six carcinogens  listed
under Section 112 (i.e., asbestos, vinyl  chloride,  benzene,  inorganic
arsenic, radionuclides, and coke oven emissions).   In most cases,  these
emissions cannot be completely eliminated.  Therefore, standards
prohibiting all emissions (and eliminating all risk) would effectively  shut
down the Nation's basic industry.  The legislative  history of  Section  112
makes clear that Congress had no such draconian  results in mind.   See
Proposed Policy for Airborne Carcinogens, 44  FR  58642, 58659 -58661
(October 10, 1979).
     In these circumstances, EPA has established  Section 112 standards  for
carcinogens at levels that reflect demonstrated,  effective control  systems.
This reduces, but does not eliminate, emissions,  exposure, and  risk.   It

                                4-2

-------
necessarily involves considering feasibility and cost.   First,  a  control
system cannot be considered demonstrated or effective  unless  it is
feasible.  Second, whether a system is feasible depends  on  its  cost;  almost
any degree of emission limitation is theoretically  possible:  whether it  is
feasible depends on the reasonableness of  its cost.   Indeed,  NRDC tacitly
accepts much of this analysis, since the agency agrees that the standards
may be based on "available" control systems, and whether a  system is
available must depend on consideration of  feasibility  and cost.
     The EPA has consistently interpreted  Section 112  to permit the Agency
to consider cost and feasibility, at least when standards are set for
carcinogens.  The EPA began implementing Section 112  on  March 31, 1971
(only 3 months after its enactment), when  it listed three hazardous air
pollutants, including the carcinogen asbestos (36 FR  5931).  The  EPA
thereupon proposed and promulgated asbestos emission  standards  that,  in
every case, were based on identified and feasible control techniques  (36  FR
23239, December 7, 1971; 38 FR 8820, April 6, 1973):
     EPA considered the possibility of banning production,
     processing, and use of asbestos or banning all emissions for
     asbestos into the atmosphere, but rejected these  approaches. .  .
     Either approach would result in the prohibition  of  many
     activities which are extremely important ....  For  example,
     demolition of any building containing asbestos fire proofing
     or insulating materials would have to be prohibited .  .  .
38 FR 8820, col. 2.  The EPA has continued consistently  to  base Section 112
standards for carcinogens on consideration of, among  other  things,  cost and
feasibility.  .[See amendments to standards for asbestos,  39 FR  38064
(October 25, 1974); 40 FR 48299 (October 14, 1975); standards for vinyl
chloride, 40 FR 59532 (December 24, 1975); 41 FR 46560  (October 21, 1976);
proposed amendments to standards for vinyl chloride,  42  FR  29005  (June  7,
1977); amendments to standards for asbestos, 43 FR  26372 (June  19,  1978);
standards for benzene, 46 FR 1165 (January 5, 1981);  49  FR  23498  (June  6,
1984); proposed standards for benzene, 49  FR 23522  (June 6, 1984);
45 FR 26660 (April 18, 1980); 45 FR 83448  (December 18,  1980);  45 FR  83952
(December 19, 1980); standards for radionuclides, 48  FR  15076 (April  6,
1983); 49 FR 43906 (October 31, 1984); proposed standards for inorganic
arsenic, 48 FR 33112 (July 20, 1983).]
                                4-3

-------
           This  consistent  and long-standing  view of  the  Agency  charged with
      implementing  Section  112 is  entitled to substantial  deference.
      E.I,  du  Pont  de  Nemours  & Co.  v.  Collins,  432 U.S.  46,  54-55 (1977);
      Power Reactor Co.  v.  Electricians.  367  U.S. 396,  408 (1961).
           Moreover, when  Congress reenacted  the Act  in  1977,  it was  well  aware
      of EPA's settled interpretation  of  Section 112.  By  that time,  the
      standards  for asbestos  and vinyl  chloride  had been  proposed, promulgated,
      and litigated.   Indeed,  the  principal amendment to  Section 112  was the
      explicit authorization  of design, equipment,  work  practice,  and operational
      standards  (Section  112(e)),  which were  the heart  of  the  asbestos  standards.
      The legislative  history  states that "[t]his limited  provision would fully
      authorize  the present EPA regulations governing asbestos."  S.  Rep. No.
      127,  95th  Cong., 1st  Sess. 44 (1977) (the  enacted  Amendments adopted  the
      Senate bill without  significant  change. H.R. Rep.  No.  564,  95th  Cong.,  1st
      Sess. 131-132 (1977)  (Conference  Report).   Congress1  reenactment  of Section
      112 with full  Knowledge  of EPA's  interpretation further  shows that
      interpretation to  be  reasonable.  North Haven Board  of  Education  v. Bell,
      	U.S. 	,  	 102  S. Ct.  1912, 1925  (1982);  NLRB  v.  Bell  Aerospace
      Cp_.,  416 U.S. 267,  274-275 (1974),  and  cases  cited  therein.
           The EPA  disagrees  with  NRDC's  claim that there  is  a general  legal
      principle  forbidding  consideration  of cost under  any statute using the
      phrase "margin of  safety."  For  example, in National  Association  of
      Demolition Contractors  v. Costle, 565 F.2d 748  (D.C. Cir.  1977),  the  Court
      upheld EPA's  action  in  setting asbestos standards  based  on feasible control
      systems:
                 NRDC  argues  that  the  Administrator's statutory
            mandate to protect the public health with "an  ample
            margin  of  safety"  is inconsistent with  his  decision
            to use  the "best  available  control methods  ....
            We disagree.
                 Protection of the public with "an  ample  margin  of
            safety" may  necessitate  use of different  control
            measures  ....
565 F. 2d at 753.
                                      4-4

-------
         The cases cited by NRDC,  Lead Industries  Association  v.  EPA,  647  F.2d
     1130 (O.C. Cir. 1980), cert,  den. 449 U.S.  1042 (1980);  American  Petroleum
     Institute v. Costle.  665 F.2d 1176 (D.C.  Cir. 1981)  cert, den.  102 S. Ct.
     1737 (1982); Hercules. Inc.  v.  EPA,  598 F.2d  91 (D.C.  Cir. 1978),  interpret
     only two statutory provisions,  Section  109  of the  CAA  and Section  307 of
     the Clean Water Act.   Those  cases do not  set  forth a general  principle that
     a statute using the phrase "margin of safety" forbids  consideration of
     feasibility and cost.  On the contrary, those cases  were  based  on  detailed
     examination of Sections 109  and 307, each taken as a whole,  along  with
     their legislative history, Agency interpretation,  and  the facts involved.
     In all  these respects, the cases  are entirely different  from the  regulation
     of carcinogens under  Section  112.

4.2  Comment:  Commenter IV-D-13  states that EPA's new  alternative approach
     (considering in one step the  before-and-after control  risks,  economic,
     and societal costs) would be  even less  protective  of public  health than
     basing  standards on cost effectiveness  or cost benefit.   The commenter
     states  that the new approach  would provide  less protection to persons in
     our region versus another, with different degrees  of protection depending
     on the  density of the surrounding population.

     Response:  [To be completed  upon  policy decisions]
                                     4-5

-------
[NOTE:   Because policy issues are still  undecided,  this  chapter is subject to
   change.  Therefore, do not focus on the exact wording of the responses.]

                         5.  SELECTION OF STANDARDS

5.1   Comment:   Commenter IV-D-14 recommends  that the final  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  states that use of the
      wash-oil  scrubber would provide 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
      due to more likely leakage  and downtime, while  a wash-oil scrubber  may
      achieve higher than 90 percent control.  If 50  percent of the industry
      uses gas  blanketing and 50  percent uses a wash-oil  scrubber control
      system (for sources required by the proposal  to use blanketing),
      differential emissions of 392 Mg/yr (or 1.8 percent of the baseline)
      would occur.  According to  the commenter, these additional  emissions
      would result in a health risk of 0.028  case/yr  (one every 36 years).

      Response:  Commenter IV-D-14 recommends that  wash-oil  scrubbers be
      selected  rather than gas blanketing as  the basis of the  standards for
      certain sources where similar control efficiencies might be achieved if
      gas blanketing were not to  work as well as expected and  wash-oil
      scrubbers were to work better than expected.   However, EPA believes that
      it is more reasonable to consider  the technological feasibility and
      impacts of control options  for representative scenarios  for the industry
      rather than unlikely, hypothetical circumstances.
           The  control efficiency of gas blanketing theoretically is 100  per-
      cent.  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, the EPA does  not expect
      well-designed, -operated, and -maintained gas blanketing systems to
      achieve less than 98 percent control efficiency.
                                     5-1

-------
     The engineering design parameters for wash-oil  scrubbers were
developed for potential application to light-oil, BTX, benzene, and
excess ammonia-liquor storage tanks.  While it is acknowledged in the
BID for the 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  we considered gas blanketing compared to wash-oil
scrubbing on a common basis of conservative estimates  of control
efficiencies.  Similarly, now at promulgation, we believe that a  common
basis of representative, if somewhat conservative, control efficiencies
should be applied.  To compare one system on a conservative basis and
the other on a liberal basis would not be reasonable and would be like
comparing apples and oranges.  Thus, the commenter's claim that the
incremental incidence is about 0.028 case/yr is not  a  reasonable
comparison.
     As discussed in the preamble to the proposed rules at 49 FR  23533,
the selection of best available technology (BAT) has been based on
consideration of the estimated costs of control and  the emission
reduction achieved.  Since proposal, however, the estimated costs and
emission reductions have been revised in several respects but still are
in keeping with the concept of representative, if somewhat conservative,
values.  These revised data are shown for furnace and  foundry plants in
Tables 5-1 and 5-2, respectively.  Consideration of  these revised data
has resulted in certain changes in the control options selected as the
basis of the final standards, particularly in regards  to the storage
tank sources cited by the commenter.
     As shown in Table 5-1, the average cost effectiveness of benzene
removal by gas blanketing or by wash-oil scrubber controls for light-oil
and BTX storage tanks at furnace plants appear unreasonably high
nationwide (about $3,850/Mg and $5,050/Mg, respectively).  These  costs,
in our judgment, do not override the benzene emission  reduction
achieved, even when the benefits of total VOC controls are considered.
The average cost effectiveness for total VOC control ranges between
$2,700/Mg and $3,500/Mg.  Consequently, the final regulations do  not
require controls for light oil or BTX storage tanks.  However, this does
                               5-2

-------
                                           TABLE  5-1.   FURNACE  COKE  NATIONWIDE  EMISSION  REDUCTIONS  AND  COST  EFFECTIVENESS
Benzene
Emission source
1. Final cooler cooling tower

2. Tar decanter tar-intercept-
ing sump, and flushing-b
liquor circulation tank
3. Tar storage tanks .
and tar-dewatering tanks
4. Light-oil condenser, light-
oil decanter, wash-oil
decanter, and wash-oil
circulation tanks
5. Excess-ammonia liquor
storage tank
6. Light-oil storage
u, tanks and BTXfa
i storage tanks
u>
7. Benzene storage tanks


8. Light-oil sump
9. Pumps



10. Valves



11. Exhausters

.

12. Pressure-relief devices



13. Sampling connection systems
14. Open-ended lines
Uncontrol led
emissions
Control option (Mg/yr)
1. Use tar-bottom final cooler
2. Use wash-oil final cooler
Coke oven gas blanketing system


Coke oven gas blanketing
system
Wash-oil scrubber
Coke oven gas blanketing system


Coke oven gas blanketing
system
Coke oven gas blanketing
system


1. Wash-oil scrubber
2. Nitrogen gas blanketing
system
Cover
a. Quarterly inspections
b. Monthly inspections
c. Equip with dual
mechanical seals
a. Quarterly inspections
b. Monthly inspections
c. Equip with sealed-bellows
valves
a. Quarterly inspections
b. Monthly inspections
c. Equip with degassing
reservoir vents
a. Quarterly inspections
b. Monthly inspections
c. Equip with rupture disc
system
Closed-purge sampling
Cap or plug
8,480
8,480
8,290


1,389

3,764
3,764


412

309



61
61

660
370
370
370

249
249
249

17.4
17.4
17.4

168
168
168

33
11
Emission
reductions
(Mg/yr)
6,580
8,480
8,014


1,361

3,378
3,682


404

290



55
60

647
262
308
370

156
180
249

9.6
11.0
17.4

75
88
168

33
11
Average
cost
effec-
tiveness
($/Mg)
(20)
1,262
(122)


1,287

103
137


2,113
•
3,856



1,992
2,246

735
114
116
2,888

(236)
(110)
17,381

1,450
2,737
25,133

(415)
(306)
917

1,247
642
Incremental
cost
effec-
tiveness
($/Mg)
(20)
5.699
(122)


1.287

103
520


2,113

3,856



1,992
5,106

735
114
131
16,690

(236)
724
62,776

1,450
11,232
63,818

(415)
291
2,273

1.247
642
voca
Uncontrolled Emission
emissions reductions
(Mg/yr) (Mg/yr)
100,744
100,744
17,355


32,525

5,369
5,369


591

443



61
61

942
528
528
528

355
355
355

24.9
24.9
24.9

240
240
240

47
16
70,896
100,744
16,789


31,900

4,824
5,248


579

416



55
60

923
374
440
528

223
256
355

13.7
15.8
24.9

107
126
240

47
16
Average Incremental
cost cost
effec- effec-
tiveness tiveness
($/Mg) ($/Mg)
(2)
106
(58)


55

72
96


1,492

2,693



1,992
2,246

515
80
81
2,022

(165)
(77)
12,166

1,015
1,916
17,593

(290)
(214)
642

873
449
(2)
363
(58)


55

72
365


1,492

2,693



1,992
5,106

515
80
92
11,683

(165)
507
43,943

1,015
7,862
44,673

(290)
203
1,591

873
449
 VOC reduction includes benzene.
 Wash-oil scrubbers are more costly and less effective than gas blanketing for these
Note:   (1) parentheses denote cost savings;  (2) data current as of November 1984.
sources.

-------
not preclude the establishment of  controls  by  State  and  local  concerns
where site specific conditions warrant  regulatory  action.   W.ash-oil
scrubber controls for benzene storage tanks  at  furnace plants,  however,
are not affected by this decision  because the  cost effectiveness  for
both benzene and total VOC removed  are  reasonable  for this  industry
segment.
     The basis of the standards for  excess  ammonia-liquor  storage at
furnace plants also has been revised.   At proposal,  the  wash-oil
scrubber was selected as best available technology (BAT) because  the
incremental cost effectiveness of  gas blanketing was  relatively high and
the additional VOC emission reduction did not  convince EPA  that the
higher costs were warranted.  However,  the  data shown on Table  5-1 have
resulted in reconsideration of EPA's decision  at proposal.   These data
indicate that gas blanketing would  provide  a higher  emission reduction
at a lower cost per Mg of benzene  (or total  VOC) removed.   For  this
reason, gas blanketing controls have been selected as the  basis of the
final standards for excess ammonia-liquor storage tanks  at  furnace
plants.
     As shown on Table 5-2, nationwide  uncontrolled  benzene emissions
from light-oil storage and BTX at  8  foundry  plants are estimated  at  13
Mg/yr; an additional 1.3 Mg/yr is  contributed  from benzene  storage
practiced at 1 foundry plant.  The  estimated annual  incidence  associated
with these emissions is about 0.04  cases/yr.   The  average  costs per
megagram of benzene emission reduction  by gas  blanketing or wash-oil
scrubber controls (ranging from $19,500/Mg  to  $28,800/Mg)  are  in  our
judgment unreasonable from a nationwide perspective.  The  average cost
effectiveness in terms of total VOC  are similarly  unreasonable  and do
not convince EPA that the supplemental  benefits of VOC control  warrant
the high costs.  Because imposition  of  controls on these sources  would
outweigh the estimated nationwide  public health benefit  expected, the
final standards do not include control  requirements  for  light-oil, BTX,
and benzene storage sources at foundry  plants.  This  decision,  however,
does not preclude the establishment  of  controls by local or State air
pollution control agencies if warranted by  site-specific considerations
or other factors.
                               5-4

-------
                                           TABLE  5-2.   FOUNDRY COKE NATIONWIDE EMISSION REDUCTIONS AND COST EFFECTIVENESS
                                                                                       Benzene
                                                                                                                                           VOC°
         Emission source
          Control option
                                                        Average
                                                         cost
                             Uncontrolled  Emission     effec-
                              emissions    reductions  tiveness
                                (Mg/yr)     (Mg/yr)     ($/Mg)
                                Incremental                            Average   Incremental
                                   cost                                 cost       cost
                                  effec-    Uncontrolled  Emission     effec-     effec-
                                 tiveness    emissions    reductions  tiveness   tiveness
                                  ($/Mg)      (Mg/yr)      (Mg/yr)     ($/Mg)      ($/Mg)
 1.  Final  cooler cooling  tower

 2.  Tar decanter tar-intercept-
    ing sump,  and flushing-
    liquor circulation tank
 3.  Tar storage tanks        .
    and tar-dewatering tanks

 4.  Light-oil  condenser,
    light-oil  decanter,
    wash-oil  decanter,
    and wash-oil    .
    circulation tanks
1.
2.
Use tar-bottom final cooler
Use wash-oil final cooler
Coke oven gas blanketing system
Coke oven gas blanketing
 system

Coke oven gas blanketing system
693
693

444
                                   78


                                  172
534
693

430
            76.4


             168
  386
3,814

  949
          5,261


          1,416
   386
15,361

   949
            5,261


            1,416
8,247
8,247

  921
             1,808


               246
 VOC reduction includes benzene.

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

Note:  (1) parentheses denote cost savings;  (2)  data current as  of  November 1984.
                                                  sources.
5,752
8,247

  891
              1,772


                240
 36
320

464
             225


             992
 36
978

464
            225


            992
5.

6.
Ul
I
Ln
7.


8.
9.



10.



11.



12.



13.
14.
Excess-ammonia b
liquor storage tank
Light-oil storage
tanks and BTX.
storage tanks
Benzene storage tanks


Light-oil sump
Pumps



Valves



Exhausters



Pressure-relief devices



Sampling connection systems
Open-ended lines
Coke oven gas blanketing
system
Coke oven gas blanketing
system

1. Wash-oil scrubber
2. Nitrogen gas blanketing
system
Cover
a. Quarterly inspections
b. Monthly inspections
c. Equip with dual
mechanical seals
a. Quarterly inspections
b. Monthly inspections
c. Equip with sealed-bellows
valves
a. Quarterly inspections
b. Monthly inspections
c. Equip with degassing
reservoir vents
a. Quarterly inspections
b. Monthly inspections
c. Equip with rupture disc
system
Closed-purge sampling
Cap or plug
33

13


1.3
1.3

27
101
101
101

68
68
68

4.9
4.9
4.9

46
46
46

9
3
32.3

12.7


1.1
1.2

26.5
72
84
101

43
49
68

2.7
3.1
4.9

20
24
46

9
3
8,505

19,523


27,139
28,775

1,717
131
134
3,204

(256)
(117)
19,251

1,593
3,008
27,619

(455)
(335)
1,021

1,385
731
8,505

19,523


27,139
47,180

1,717
131
153
18,474

(256)
802
69,534

1,593
12,343
70,130

(455)
323
2,524

1,385
731
48

19


1.3
1.3

38
159
159
159

107
107
107

7.7
7.7
7.7

73
73
73

14
5
47

18.6


1.1
1.2

37.2
113
132
159

67
77
107

4.3
4.9
7.7

33
39
73

14
5
5,916

12,940


27,139
28,775

1,203
83
85
2,035

(162)
(74)
12,225

1,012
1,910
17,538

(289)
(213)
648

879
464
5,916

12,940


27,139
47,180

1,203
83
97
11,731

(162)
509
44,154

1,012
7,838
44,532

(289)
205
1,603

879
464

-------
           Control  requirements for excess ammonia-liquor storage tanks at
      foundry plants also have been eliminated from the final standards.  The
      average cost  per megagram of benzene emission reduction exceeds $8,500;
      the cost effectiveness for total VOC ranges between $5,900/Mg and
      $6,200/Mg.  Annual  incidence associated with nationwide uncontrolled
      benzene emissions of 33 Mg/yr is estimated at 0.005 cases/yr.  As with
      light-oil, BTX, and benzene storage, the costs of achieving the emission
      reduction outweigh  the public health benefits expected on a nationwide
      basis.

                       [Response subject to policy decisions]
           As discussed in the preamble to the proposed standards, owners or
      operators at  furnace or foundry plants may apply for the use of
      alternative control  techniques that provide an emission reduction
      equivalent to or better than the controls selected as BAT.  If
      site-specific parameters allow the design and operation of wash-oil
      scrubbers that can  be demonstrated to achieve at least 98 percent
      control for process vessels, tar storage tanks, tar intercepting sumps,
      or excess ammonia liquor storage tanks, this control strategy certainly
      is not precluded.
5.2   Comment:  Commenter IV-D-13 supports proposed measures and practices
      that provide  or approach 100 percent benzene emission control but
      recommends that more effective options (rejected for some sources due to
      cost considerations) be selected because of the additional emission
      reduction achieved.  The commenter cites specifically the additional
      benzene emission reduction of 2,130 Mg/yr achievable with wash-oil final
      coolers compared to tar-bottom final coolers.  Other recommendations
      include gas blanketing of excess-ammonia liquor, light-oil, BTX, and
      benzene storage tanks; dual  mechanical seals for pumps and sealed
      bellows valves; and degassing reservoir vents for exhausters.  When
      combined with wash-oil final cooler controls, this equipment would
      provide an overall  benzene emission reduction of 2,367 Mg/yr and would
      increase the  total  emission reduction associated with the standard from
      89 to 98 percent.
                                     5-6

-------
      Response:  The EPA agrees that certain control techniques capable  of
      providing a higher emission reduction compared to BAT were  rejected in
      some cases as the basis of the standards due to cost considerations.
      The validity of considering costs in the selection of BAT for  national
      emission standards is further discussed in response to comment 4.1.
           At proposal, EPA selected tar bottom final cooler controls  as BAT
      over wash-oil cooling even though the latter technology provided a
      significantly higher emission reduction.  However, the cost  of this
      higher emission reduction, if required nationwide, was predicted to
      result in a capital  availability problem for the industry.

      (Insert final cooler/naphthalene processing decision for furnace plants)

           The EPA's decision not to require leakless equipment for  pumps,
      valves, and exhausters remains unchanged.  For both furnace  and  foundry
      plants, the costs of these controls are clearly unreasonable.  The
      technical feasibility of requiring these controls industrywide is
      questionable.  At furnace plants, the combined effect of requiring these
      equipment would yield an additional benzene emission reduction of
      approximately 135 Mg/yr.  These benefits would be achieved,  however,  at
      incremental cost-effectiveness values ranging from about $16,700 to over
      $63,800/Mg benzene removed.  The supplemental benefits of total  VOC
      control, when compared to the incremental costs range from  $ll,700/Mg to
      $44,700/Mg of total  VOC removed and do not convince EPA that this  control
      strategy is warranted.  The costs per megagram of benzene (and total  VOC)
      removal for leakless equipment at foundry plants are even higher and
      clearly not warranted for the incremental risk reduction achieved.  The
      EPA considers that the BAT controls selected yield the highest emission
      and risk reductions  achievable without imposing costs that  exceed  the
      public health benefits expected.

5.3   Comment:  Commenter IV-D-5 recommends that wash-oil final coolers
      be required rather than tar-bottom final coolers because of  the
      resulting reduction  in benzene emission risk, coupled with  the volatile
      organic compound  (VOC) emission reduction.  At one plant, the  commenter
      states, the benzene emission reduction would reduce the maximum  lifetime
                                     5-7

-------
      risk (MLR) from 3.3 x 10"^ (with direct-water  final  cooling)  to  3.7  x
      10'4 (with tar-bottom final cooler controls) to  4.7  x  10'5  (with
      wash-oil final coolers).  At another plant, the  same control  scenario
      would reduce the MLR from 6.9 x 1CT4 to 7.8 x  10'5 to  9.8 x 10'6.  The
      commenter calculates that wash-oil final coolers  at  these plants would
      reduce the MLR for tar-bottom final coolers by a  factor  of  7.9.
           The commenter also points to the supplemental benefit  of VOC
      emission reduction.  At one plant, total VOC emissions from naphthalene
      processing would be reduced from 3,133 Mg/yr to  595  Mg/yr (with  a
      tar-bottom final  cooler) to 0 Mg/yr with wash-oil final  coolers.   VOC
      emissions at the other plant would be reduced  from 2,274 Mg/yr to  432
      Mg/yr under the proposed standards and also would be eliminated  by
      wash-oil final cooler controls.  At both plants,  VOC emissions after
      application of the proposed controls still would  be  greater than  100
      ton/yr and would remain a major source.  The commenter adds that the
      estimated cost effectiveness of $350/Mg for wash-oil final  cooler  VOC
      control  compares favorably with the cost effectiveness for  most
      reasonably available control technology (RACT) categories.

      Response:    [insert final cooler/naphthalene  processing decision]

5.4   Comment:  Commenter IV-D-15 requests that the  standards  allow use  of new,
      more effective technology as an alternative control  technique for
      naphthalene processing and cooling tower emissions.  This system,
      recently patented by his company, eliminates cooling tower  emissions
      through  indirect heat exchangers and reduces emissions from naphthalene
      separation with enclosed separator and froth flotation units. The
      commenter estimates benzene emissions from the processing of  naphthalene
      skimmings at a maximum of 20 g of benzene/Mg of coke.  The  system
      achieves a 95-percent benzene emission reduction  from  baseline compared
      to 81 percent for tar-bottom final cooling and has lower capital,
      operating, and energy requirements than wash-oil  systems.   According to
      this commenter, use of a single liquid phase (water) prevents the  problem
      of water and oil  emulsion found in wash-oil final cooler systems.

        Response:      [Insert final  cooler/naphthalene processing  decision]
                                     5-8

-------
5.5   Comment: Commenter IV-D-9 states that the regulation is unclear  regarding
      the use of alternatives to the proposed "zero" emissions 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
      exchangers to cool the recirculated liquor in lieu of open cooling
      towers.  The system also would utilize tar addition to absorb naphthalene
      in a "closed-to-the-atmosphere" mode of operation.  The commenter
      believes that this system, in conjunction with gas blanketing, would
      comply with the "zero" emissions limit for naphthalene processing in a
      cost-effective manner.  The installed cost of this system is estimated at
      $8 million compared to $12.5 million for the tar-bottom final cooling
      system.

       Response:       [Insert final cooler/naphthalene processing decision]
                                     5-9

-------
                        6.   EMISSION CONTROL TECHNOLOGY

6.1  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 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
     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.

                                      6-1

-------
     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, we do 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 (who 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.  We disagree.  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, Dravo/Stills Corporation has
designed and installed a positive pressure gas blanketing system in an
existing European coke by-product plant.  The system uses clean coke oven
gas (at about 1 inch 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/Stills has had discussions
with at least one U.S. coke plant operator about such a system for their
plant.
                                 6-2

-------
6.2  Comment:  Commenters IV-D-4, IV-D-6,  IV-D-7,  IV-D-14, and  IV-D-17 argue
     that gas blanketing systems, although appropriate and cost  effective  for
     some plants, should not be mandatory  at all sites due to 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.  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.  One commenter adds 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
     response to comment 6.1, gas blanketing has been demonstrated as safe  and
     effective during an operating period  of over  24 years (1960-1984) at  four
     plant sites in addition to Armco-Houston.  In fact, in direct
     contradiction to the commenters1 statements,  we consider that the
     proposed system will improve the safety level found in uncontrolled
     by-product plant environments.  The reasons for our conclusions are
     explained below.
          The American Iron and Steel Institute (AISI) argues that 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

                                      6-3

-------
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  due
to 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.
     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 our judgment,
enclosing these sources and ducting the emissions back  to the process  via
a closed positive-pressure system will  reduce substantially  the  explosion
                                 6-4

-------
hazard that now exists.  We do 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, however, have been  reflected  in  the  cost  estimates.
     We also recognize that leaks  in a blanketing system will  occur
occasionally due to 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.  Our  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.
Although special precautions such  as explosion-proof  electrical
components may be required where light oil or  benzene  is stored, this
equipment should already be in place at plants  where  the NFC  or  plant
safety codes have required their installation.  In  addition,  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
                                 6-5

-------
intrinsically safe electrical systems with low power needs, or applying a
general purpose enclosure to isolate the leak area (Docket Item
II-C-132).
     One commenter asserts 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.  We contend
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 throughtout the plant.
     Prior to proposal  of the standard, we 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
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 proposal, additional features such as
water drains and overflow connections for tar tanks  and liquid level
sampling/gauging instrumentation with vapor-tight seals have been added.
Assuming each system is properly operated and maintained after
installation, we consider 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.
     We agree 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, we assume that such
a power loss would affect most plant operations and  probably would result
in a shutdown until power was restored.  Unfortunately, we are aware  of
                                 6-6

-------
     no other reasonable approach capable of overcoming the effects  of  cold
     climates.
          If a plant operator is concerned that coke oven gas would
     contaminate the product handled by a gas-blanketed vessel,  nitrogen  or
     natural gas could be used as a substitute.  As discussed in 49  FR  23530,
     nitrogen is recommended for blanketing benzene storage tanks because  of
     the possibility of contamination.  Emissions could be routed to the  coke
     oven gas main and burned in the gas combustion system or routed to the
     gas main before light-oil removal and recovered in the light-oil plant.
     In comparison, we anticipate no contamination problem.  The use of
     undesulfurized gas as the blanketing medium for light-oil storage  tanks
     has been demonstrated in at least one plant with no reported safety,
     operational, or contamination problems (49 FR 23530).

6.3  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 creates 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
     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 due to 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 (COG) mixture below the lower explosive limit (or  above
     the upper explosive limit).
                                      6-7

-------
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 6.2).
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
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.
                                 6-8

-------
6.4  Comment:  Commenter IV-D-14 states that overpressurization of  a positive
     pressure system poses an explosive and occupational hazard due to the
     carbon monoxide (CO) released.  The presence of CO increases costs  for
     additional  monitoring and employee training because CO hazards do not
     exist presently.  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  mm  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 regulations would not require the CO monitoring, but we
     encourage companies to follow their practice of safety reviews and
     implementation of precautions based on the 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.  Coke oven
     piping winds through the plant, and the gas is treated in enclosed
     vessels such as primary coolers, direct water coolers, and scrubbers.
     The gas is also piped to the battery underfiring system and is used  in
     other parts of the steel plant.  If. all of these locations are subject  to

                                      6-9

-------
     the stated monitoring by the company, then the consistent  application  of
     policy would dictate similar monitoring at gas-blanketed vessels.
          The regulation does not require explosive condition monitoring
     suggested by the commenter because the existing and  demonstrated  systems
     installed by other companies had no such monitoring.   Consequently,  the
     costs of such monitoring was not included.  However,  less  frequent
     monitoring is included in the cost estimate and is required  for  leaks  and
     inspection of equipment.  The gas-blanketed equipment  is required to be
     enclosed and sealed, and consequently, should not be  more  prone  to  leaks
     than other by-product equipment that handles coke oven  gas.   If  the
     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.
6.5  Comment:  Two commenters (IV-0-4 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 this 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.
          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

                                      6-10

-------
     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
     arrestor 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 that the  sump will not contain explosive  gases.
     The solution to the commenter's question will depend upon the sites'
     specific conditions and the company's policy.

6.6  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  standards recognize that  plugged  vents or  valves pose  an
     operational  problem and potential  safety hazard  if not repaired.   For
     this reason, the 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.   Section 61.136 of the proposed  regulation would  require  that
     records containing a brief description of  any abnormalities,  the  repairs
     made, and the dates of repair be  maintained  for  a  minimum of  5 years.
     Similar provisions are included in the regulation  in the  event a  system
     blockage is  detected.  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 biannual  leak inspection.

6.7  Comment:   Commenter IV-D-13 requests  that  EPA determine whether any
     benzene storage tanks at by-product  plants are equipped with  shingle
     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.   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  s'eals.
                                      6-11

-------
     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.  We do not
     know of any benzene storage tanks with floating roofs in by-product
     plants.  The controls we have analyzed for storage tanks, and which would
     most likely be installed, 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 standard for benzene storage tanks would require a control
     system designed and operated to reduce emissions by 90 percent.  In the
     event that a coke by-product plant owner chose to use an internal
     floating roof as a control, a continuous seal  would be necessary  to meet
     the 90 percent design standard.

6.8  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 on pages 4-28 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  (98 percent) downward to account for uncontrolled emissions
     from the approximately 13 percent of  the liquid surface  of the sump 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 proposed rule  at  49 FR 23529, the theoretical
     efficiency of source enclosure approached 100 percent.  However,  this
     efficiency cannot be expected to  be maintained continuously for the
     service life of the equipment due to  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.
                                      6-12

-------
     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.
                                 6-13

-------
                           7.  ENVIRONMENTAL IMPACTS

7.1  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 due to 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 proposed standards in 49 FR 23524.  As
     stated, the impacts were calculated initially from a data base of 55
     plantsi  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 background information  document (BID) would be revised
     following proposal .
          The  data base has been revised since 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 current industry operating status.  These  data,
     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)—are displayed in Appendix A
     (Tables A-l and A-2).  As discussed below in response to comment  7.2,
     adjustments to emission factors also have been made since proposal to
     indicate  lower emission rates from sources  at plants producing foundry (or
     furnace 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 clearly the
     environmental  impacts of control  options on furnace and foundry plant
     industry segments as well as on furnace and foundry plants combined.
                                      7-1

-------
     Tables A-l and A-2 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, while 14 (mainly merchant
plants) produce foundry or furnace.and foundry coke.  Of the 30  furnace
plants, 6 currently are on cold idle.  These plants have been identified  as
follows:  (l).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) Weirton
Steel—Brown's Island, West Virginia.  Also, 1 of the 14 foundry plants
(Alabama By-Products—Keystone, AL) is currently on cold idle.   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 final
standards.  The deletion of 6 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 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 currently 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, over half (16 of 30) of the  furnace plants continue to 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 cooling
technology.  Although tar separation sources (e.g., tar decanters,
dewatering, sumps,  and storage) are found at all 30 sites,.as are most
light-oil plant sources (light-oil decanters, sumps, storage, and wash-oil
circulation tanks), BTX storage is practiced at only 10 sites while 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 7 plants; tar-bottom final  coolers at
2 plants, and no wash-oil  final coolers in use.  Although tar separation

                                 7-2

-------
sources are present at each site, light-oil storage  is  found  at  8  of  the 14
sites.  Benzene and BTX are stored at 1 plant.
     Tables 7-1 through 7-6 of the BID for the proposed  standards  have  been
revised to show the estimated environmental impacts  of  the  final standards.
Table A-5 displays the estimated nationwide benzene  emissions  and  process
capacity data for affected sources at the 30 furnace  plants and  at the  14
foundry plants; total industry estimates for furnace  and  foundry plants
combined also are provided.  Table A-6 shows comparable  data  for total  VOC
emissions (benzene and other volatile organic compounds  [VOC's]).
     As Tables A-5 and A-6 show, estimated nationwide benzene  emissions
from the 30 furnace plants total approximately 24,200 Mg/yr compared  to
about 1,700 Mg/yr from the 14 foundry plants.  Estimated  VOC  emissions  from
the 30 furnace plants total approximately 159,200 Mg/yr;  VOC  baseline
emissions from foundry plants are estimated at 11,700 Mg/yr.
     The affects control options have on reducing benzene and  total VOC
emissions from furnace plants, foundry plants, and furnace  and foundry
plants combined are shown  in Tables A-7 and A-8.  Estimated emission
reductions also are shown  by source.  Implementing the  final  standards
would reduce nationwide benzene emissions from furnace  plants  to about
2,800 Mg/yr, an emission reduction of about 88 percent  from-the  current
baseline level of approximately 24,200 Mg/yr.  Total  uncontrolled  VOC
emissions would be reduced to about 31,900 Mg/yr, an  80-percent  reduction
from about 159,200 Mg/yr.  Nationwide benzene emissions  from  foundry  plants
would be reduced to about  270 Mg/yr, an 84-percent reduction  from  the
estimated baseline level of 1,700 Mg/yr.  A VOC emission  reduction of 78
percent also would be achieved with emissions reduced from  11,700  Mg/yr  to
about 3,150 Mg/yr.
     Implementing the final standards would reduce overall  benzene
emissions from furnace and foundry coke producers from  approximately
25,900 Mg/yr to about 3,000 Mg/yr, a reduction of about  88  percent.
Nationwide VOC emissions also from these sources would  be reduced  by  about
80 percent, from approximately 170,900 Mg/yr to about 35,000 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

                                 7-3

-------
     pollution,  solid waste disposal,  and noise or odor levels.  As discussed in
     the preamble to the proposal  notice in  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-9 and A-10 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  final  standards could result  in an
     increased cyanide (HCN) concentration at plants  using indirect final
     cooling.  As discussed in  the BID for the proposal  standards on page  7-7,
     HCN is emitted presently 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  cyanide 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  calculation of
     revised emission estimates for furnace  and foundry plants is discussed
     below in response to comment  7.2.

          [Subject to change due to policy decisions  on the final  standard]

7.2  Comment:  Commenters IV-D-6,  IV-D-7, IV-D-10, IV-D-11, and IV-D-12
     oppose the  regulation of small  merchant  or foundry plants because
     operating dissimilarities  result  in fewer emissions compared to emissions
     from furnace coke plants.   The commenters state  that foundry plants
     generate fewer emissions due  to:  -(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
                                      7-4

-------
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 proposed standards.  Another commenter maintains that  merchant
plants generate fewer emissions than furnace plants  not only due  to
different operating practices but also due to the relative size of the
industry segment compared to furnace coke plants.

Response:  In response to the public comments received on this  issue, we
have 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, we  agree  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  1979-1980
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 while 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	) 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 U.S. Department of
Energy 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-ll of Appendix A.  Table A-ll 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-ll
represent the principal basis for the technique  used to adjust  the proposed
emission factors for foundry coke producers.
                                 7-5

-------
     Based on a review of data  contained  in  another  cokemaking  reference
source (Docket Item II-I-2), we  also  agree with  commenters  who  suggest  that
the lower coking temperatures associated  with  foundry  coke  production
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 BID  at
proposal  (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
light-oil benzene content of 70  percent is still considered appropriate.
     Separate emission factors  for foundry plant sources  have  been
developed by applying correction factors  to  the  emission  factors  initially
proposed for both furnace and foundry plants.  These changes do not affect
the current emission factors as  applied to furnace plants.  The
computations of correction factors are shown in  Table  A-12 of  Appendix  A;
the final emission factors for  furnace and foundry plants are  shown in
Table A-13.
     For plants that produce only foundry coke,  benzene emission  factors
for product storage (i.e., BTX,  light oil, and benzene) and light-oil
recovery plant sources (i.e., wash-oil decanters, wash-oil  circulations
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  reduced 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-12  are
indicative of the partitioning.  These ratios  are multiplied by relative

                                 7-6

-------
     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 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
     benzene 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 foundry coke
     plant light oil.  When the estimate supplied by ACCCI is used, the
     correction factor is 0.91.  Table A-13 indicates the final uncontrolled
     emission factors for furnace and foundry plants.  Table A-14 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 7.1.  The effect of reduced foundry plant emission rates on health
     risk estimates is discussed further in Chapter 10 of this document.
     Additional discussion of revised environmental, cost, economic, and health
     risk impacts on foundry plant sources is presented in Chapter 5, "Selection
     of Standards."
7.3  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
                                      7-7

-------
emissions from a medium-sized model  plant  (4,000 Mg/day  of  coke)
at 1,082.7 Mg/yr; with the proposed  controls  (and  assuming  89  percent
recovery), remaining uncontrolled  emissions of  119.1 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  119.1   Mg/yr to  71.7 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, Empire Coke argues that  small merchant  plants  should
not be regulated because of the low  benzene emission levels 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.  We
believe that the commenter has  misconstrued the purpose  of  model  plants  and
their role in the EPA decision-making process.  First, EPA's decision to
regulate is not based on model  plant  emission estimates  and their  level
compared to larger model  facilities.  Uncontrolled nationwide  benzene
emissions from foundry plants are  currently estimated at  about 1,700  Mg/yr.
Although considerably lower than emissions from furnace  plants, these
foundry plant emissions constitute a  significant source  on  their  own
merits.  However, other factors such  as  the potential risk  reduction
achieved and cost impact of controls  also are important  factors that  are
considered on a source-by-source basis in the decision-making  process.   The
EPA's decision regarding controls  for foundry plant sources is discussed
further in response to comment  5.1.
     The purpose of constructing model plants is to portray typical
facilities in terms of size and processes representative  of the industry.
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 Mg/day  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.
                                 7-8

-------
          Since proposal,  the data base has  been  revised to reflect current
     industry operating status based on site-specific information.  These
     revised data indicate foundry plant capacities  at existing sites ranging
     from 130,000 to 617,000 Mg/yr.  Empire  says  its capacity is 440 Mg/day.
     When converted to an  annual  basis  (approximately 161,000 Mg/yr), Empire
     remains well within this size range. We  consider that the Model  Plant 1
     (small) size range remains representative of small  plant operations.  In
     terms of actual on-site processes, further discussions with the commenter
     revealed that Empire  does (emphasis added) operate a direct-water final
     cooler (the Empire comments  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 Empire site,
     they typify small plant processes  on the  whole.
          Other changes to the data base also  have been made since proposal  to
     reflect more closely  foundry plant operations.   These changes include lower
     emission factors and  recovery credits.   Also, nationwide emission estimates
     have been made based  on site-specific process data, rather than by model
     plant extrapolation.   We believe 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  would  also need to be considered
     by the commenter when recalculating actual benzene emission estimates for
     this site.
7.4  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 due
     to differences in method of  operation and other factors.  We certainly
     agree that difference in methods of operation,  operating parameters, and
     design features are evident  from plant  to plant and will  influence actual

                                     7-9

-------
emissions from each source.  During development of the estimated emission
factors, these variations have been taken into account to the extent
possible by averaging applicable measurements to obtain a factor
representative of a "typical" source.  Additionally, the emission factors
have been adjusted since proposal for improved applicability to plants
producing foundry coke (or furnace and foundry coke).
     Specifically, the commenters state that data (12 tests) supporting the
tar decanter emission factor are not sufficient for industrywide
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 proposed standards (page 3-10), typical residence times
are about 10 minutes for liquor and about 40 hours for tar.  Optional
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 due to coal type and
difference between plants.  As stated above, adjustments have been made
since proposal to account for differences between furnace coal  types used
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-26), where the average benzene
emission rate from 3 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 at this
plant.  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 per megagram of coke.  We consider this data base of 12

                                 7-10

-------
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 industrywide because emissions depend on the method
of operation.  In support, the commenters point to the "unexplained"
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 3 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
has been discussed in the BID for the proposed standards (page 3-16), which
states as follows:
     "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
                                 7-11

-------
     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
     industrywide.   The emission  factor  for the  flushing  liquor circulation tank
     (9 g/Mg  of coke)  was obtained from  one test where 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.  While 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  for emissions  from
     flushing liquor  circulation  tanks.  We agree with the commenter, however,
     that  emissions will  vary necessarily  depending  on the number and geometry
     of tanks, the  number of vents, and  other  factors.

7.5  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  reductions are
     realistic and  provide the added benefit  of  helping this State  reduce  the
     VOC inventory  in  the Baltimore ozone  nonattainment area.
     Response:  We  thank  the commenter for his support.  It  may be  of interest
     to note  the following revisions to  the previous emission estimates.
     Implementating the standards is expected  to reduce total VOC emissions at
     the plant in the  commenter's State  by about 400,600  Mg/yr.
                                      7-12

-------
                               8.  COST IMPACT

8.1    Comment:  Commenters IV-D-9 and IV-D-14 argue that the capital  costs  of
       the proposed equipment are $50 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 the 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/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.
       Response:  The cost impact analysis has been revised since  proposal to
       incorporate answers to the concerns cited by the commenters.  The final
       analysis, details of which are shown in Appendix B, indicates nationwide
       capital costs of approximately $49 million (1984 dollars) compared to
       estimated capital costs of about $24 million (1982 dollars) estimated  at
       proposal.
            In revising the analysis, to consider the commenter's  concerns,  EPA
       conducted a detailed review of the United Engineers estimate for the
       Bethlehem plant of Bethlehem Steel, the Bethlehem Steel estimate for
       their Sparrows Point plant, the Armco and National Steel cost data, and
       secured the assistance of C. R. S. Sirrine, Inc., another design
       engineering firm to assist in the development of revised unit costs.
       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 emissions controls, and to obtain examples of 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
                                      8-1

-------
       the comments and the cost data developed by Si mine.  The  revised
       analysis 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 commenter's 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 commenters cost estimates  indicated that
       such was the case.  Also, the preference for  site-specific  information
       must be compromised somewhat when attempting  to develop an  estimate of
       nationwide cost impacts for 44 plants within  schedule and  budgetary
       restraints.  The EPA believes that the revised cost analyses  accommodates
      .the commenters1 concerns.
8.2    Comment:  Commenters IV-D-6 and IV-D-14 state that the value  of potential
       product recovery credits has been overestimated.  In particular,
       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 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 steel making
       operation creating fuel  demand.
       Response:  We essentially agree with the commenters in that the value of
       potential product recovery credits was overestimated at proposal.   As
       discussed further in response to comment 7.2, foundry plants  produce  less
       light oil than larger furnace plants.  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.

                                      8-2

-------
            In response to the commenters' concerns, a telephone  survey  of  7
       (3 furnace and 4 foundry) plants was conducted to determine, the extent  of
       flaring excess coke oven gases.  (IV-E-	).  Briefly, we  found that this
       flaring is not generally practiced except as a last resort.  Of the  5
       foundry plants surveyed, only 1  (Empire Coke) flares gas continuously.
       Of the unremaining 4 foundry plants, 2 plants do not flare excess  gas at
       all while 2 plants flare only seldomly.  Of the 3 furnace  plants
       surveyed, 1 plant never flares;  1 plant flares only in emergency
       situations; and 1 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 U.S.
       Department of Energy (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 at proposal.

8.3    Comment:  Commenter IV-D-14 maintains that the cost per megagram  of
       benzene reduction is more than estimated because of underestimated
       capital costs and overestimated  product recovery credits.

       Response:  We agree with the commenter in that the cost per megagram of
       benzene reduction is closely related to capital costs and  product
       recovery credits.  Revisions to  these factors are discussed above  in
       response to comments 8.2 and 8.3, respectively.  Revised estimates of the
       cost per megagram of benzene reduction are included in Appendix B  (see
       also Tables 5-1 and 5-2).  Further discussion of these estimates
       (regarding EPA's regulatory decisions) can be found in Chapter 5,
       entitled "Selection of Standards".

8.4    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 due to the lack of economies of scale compared to moderate  or
       large plants, coupled with higher per-unit control costs.  One commenter

                                      8-3

-------
       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 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  8.1  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, and 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 adequately
       represent the costs of control  from the smallest  foundry  coke by-product
       plant to the largest furnace coke by-product plant.
8.5    Comment:  Commenter IV-D-4 states that the amount of benzene  recovered
       by a small plant (capacity of 440 Mg/day of  coke) is insufficient  to pay
       for monitoring and inspections  or to justify installation and maintenance
       costs.  In support, the commenter estimates  that  with a capacity 11
       percent of a medium-sized plant's capacity and an assumed recovery  of 89
       percent, about 63.3 Mg/yr of recovered product would be achieved.   At
       current prices, the value of recovered product would be about $15,000/yr.

       Response:  As described in response to comment 8.3, several revisions
       have been made in the cost analysis regarding capital expenditures  and
       product recovery values.  These, and other changes, have  eliminated the
       savings in annualized costs for by-product controls estimated at
       proposal.  For the foundry plant industry segment,  the  total  annualized
       cost of the final  standards is  about $1.4 million/yr (after application
       of light-oil  and fuel  recovery  credits, where applicable).  However, no
                                       8-4

-------
       foundry plant closures are predicted as a result of these costs (See
       Appendix C).  Annualized costs for Empire Coke, the commenter's plant,
       are estimated at $171,000/yr.  These costs are considered reasonable in
       light of the reduction in emissions and health risk expected.
8.6    Comment:  Commenter IV-D-9 states that the annual  operating costs  of
       implementing the proposed programs outlined in the standards and
       operating the equipment are estimated at $444,000 for this plant (Inland,
       East Chicago).

       Response:  The revised cost analysis indicates total  annualized mid-range
       costs of $500,000/yr  for the Inland plant before application of fuel  and
       light oil credits; net annualized costs are estimated at $99,000/yr.
       This cost can be compared to annual  costs of about $2.8 million/yr for
       the furnace plant industry segment.

8.7    Comment:  Commenter IV-D-11 states that merchant plants should not be
       regulated because the  BID inaccurately portrays the cost per incident at
       his facility.  The commenter cites the following BID data used to
       calculate a cost per  incidence of $47.9 million:  (1) plant capacity of
       1,362 Mg/day; (2) incidence rate of $396,000 to $421,000, with an  average
       of $408,500; and (4)  annualized cost of $69,000 to $72,000, with an
       average of $70,500.  The commenter corrects plant  capacity to 681  Mg/yr
       and adjusts the incident rate to correspond proportionately (i.e., 0.005
       case/yr).  The commenter then applies the following facility cost
       estimates:  (1) capital cost of $1,804 million; and (2) annualized cost
       of $80,000.  Based on  these data, the commenter estimates the cost per
       incident for his plant at $188.4 million.

       Response:  The approach used by the commenter to estimate the cost-per-
       leukemia incident at  his plant is to extrapolate model  plant and
       nationwide data according to plant capacity.  This approach does not
       reflect the health risk and incidence predicted to occur within the
       exposed population surrounding the plant site (estimated at 80,600
       persons within 50 km  of the site).  To calculate the cost-per-incident,
       the estimated annualized cost of control for the plant is divided  by the
       estimated annual incidence.  Based on current data, the annualized cost

                                      8-5

-------
of control (including light-oil but not total VOC recovery credits)  is
estimated at $131,000/yr.  Therefore, the estimated cost-per-incident  is
about $4 million.  Further information regarding EPA's decision to
regulate the foundry plant industry segment  (including merchant plants)
is provided in Chapter 5.
                               8-6

-------
                             9.0  ECONOMIC IMPACT

9.1  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 base-
     line 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) permit upgrading,
     State Implementation Plan (SIP) compliance rules (including reasonably
     available control technology [RACT], lowest achievable emission rate
     [LAER], and new source review of coke plant rebuilding), and the pending
     coke oven battery national emission standard for hazardous air pollutants
     (NESHAP).
     Response:  At the time the original analysis was conducted, the informa-
     tion from published and unpublished sources was current.  A reanalysis
     has been conducted (see Appendix C) which utilizes data on plants and
     capacity in existence in November 1984.  Financial data and.production
     data used in baseline are from the latest available published and un-
     published sources.  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 existing
     regulations including OSHA (coke oven emissions); state regulations
     related to desulfurization, pushing, coal handling, coke handling, quench
     tower, and battery stack controls; and BPT and BAT water regulations.
     All of these were due to be in effect by 1983 at latest.  Other regula-
     tions which are pending or which have not reached the deadline date  for
     compliance are not likely to be a part of current production costs for
     firms, or will have little effect on those costs.
9.2  Comment:  Commenter IV-D-7 suggests that the economic impact analysis
     should be in 1986 dollars because the project schedule places promulga-
     tion and implementation of the regulation in 1986.
                                       9-1

-------
     Response:   Selection of the year for dollar values in analysis is some-
     what irrelevant,  since conversions may be made for any current or past
     year based on 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 due to the
     confounding effects of inflation.   Prediction of inflation rates is
     beyond the scope of this analysis.  The 1984 dollar values in the re-
     analysis are best updated for future time frames when those years are
     current, so that GNP implicit price deflators accounting for actual
     inflation may be used.
9.3  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 prob-
     lems; (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 average
     annual net investment averaged over the period from 1979 to 1983 (con-
     verted to 1984 dollars) for individual companies.  Table C-25 in Appendix C
     shows these comparisons for furnace coke.  For furnace coke plants,
     capital  cost 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

                                      9-2

-------
     mergers, acquisitions,  and creative financing measures to generate in-
     vestment 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.
          Though 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.
9.4  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 sepa-
     rated.  This commenter employs 36 people in his by-product operation, but
     a total  of 268 persons in the coke plant.
     Response:  The commenter1s argument is answered in Tables C-23 and C-28
     in Appendix C, which show the employment effects of the regulatory alter-
     natives in the furnace and foundry coke plants, respectively.  These are
     industry totals.   For the furnace coke sector, neither regulatory alterna-
     tive 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 substantial
     loss, and should be weighed against the benefits.   ~        •     -   ---=^s=^=
          For the foundry coke sector, employment impacts are calculated for
     two scenarios.  Scenario A assumes foundry coke producers do not compete
     with imports in the domestic market, while Scenario B assumes they do.
     Under Scenario A, the regulatory alternatives result in job losses which
     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 due to 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 not evaluated in terms
     of added jobs.  These gains may offset estimated job losses.
                                      9-3

-------
9.5  Comment:   Commenter IV-D-7 states that the regulation will increase the
     trend of importing coke.  The commenter cites Table 9-1 of the BID, which
     shows a growing coke importing trend since 1974, and Table 9-2, which
     shows a decrease in domestic production.
     Response:   The most current data indicates that imports have been decreas-
     ing 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 Alter-
     native II and by 25,000 Mg/yr under Regulatory Alternative III.  These
     represent increases from baseline of 0.23 percent and 0.64 percent,
     respectively.  These are negligible changes in imports.
          For foundry coke,  only under Scenario B, where import substitution
     is assumed, do imports  appear.  They are increased by the same amount
     that domestic production is reduced - 61,000 Mg/yr for Regulatory Alter-
     native II, and 94,000 Mg/yr for Regulatory Alternative III.  These
     represent 2.1 percent and 3.2 percent of demand, respectively.
9.6  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 "... 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
     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 BID for the proposed standards, which estimates a coke
     price increase ranging  from 6.4 to 15.4 percent for small plants to
     comply with baseline.
                                       9-4

-------
Response:   A distinction is made between furnace and foundry plants in
the BID analysis and in the reanalysis.   Most furnace coke producers are
captive, while 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 situ-
ation for foundry firms with respect to other firms in the foundry industry.
In the reanalysis, no batteries become uneconomic (candidates for closure)
under either regulatory alternative.  This implies that industry impacts
of regulations 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 for furnace and foundry producers of the regulations.  For both
regulatory alternatives, foundry coke producers' share of total capital
costs of compliance are less than 16 percent.  For individual foundry
coke producing firms, Table C-30 in Appendix C shows that the capital
cost of the regulatory alternatives are no more than 11 percent of net
investment for firms for which data was available.   This is not substan-
tially 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 due to the difficulties being experi-
enced in the steel industry which composes the market for furnace coke.
Differences in regulatory treatment of foundry and furnace coke producers
are not warranted.
     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 Regulatory Alternative III.  These represent 2.1 percent and
3.2 percent of foundry coke demand, respectively.  Advantages gained by
foreign competitors due to 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

                                 9-5

-------
     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 due to these price changes.
9.7  Comment:  Commenter IV-D-11 states that the economic analysis is inaccu-
     rate in predicting the increased price of coke for merchant plants.  This
     commenter estimates an increase in the price of coke at his plant of
     $1.38/Mg versus $0.24/Mg estimated at proposal (49 FR 23525).  This
     estimate is based on the commenter1s estimate of the cost of compliance
     at his facility (capital costs of $1.8 million versus average cost of
     $408,500 cited in BID; annualized costs of $80,000 versus $70,500 cited
     in 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 plant by plant basis.   The price changes
     are due to shifts in the entire supply curve, and 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 cost 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, respec-
     tively.  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
     due to the calculation of costs based on batteries with 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.
                                      9-6

-------
          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),
     while 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.
9.8  Comment:  Commenters IV-D-10 and IV-F-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 proposed or final standards.
     One commenter predicts  the closure of three entire merchant plants due to
     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 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 are 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 due to 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
                                      9-7

-------
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 due to the regulatory
alternatives.  The quantity changes shown in Table C-27 indicate that
domestic production will decrease by 61,000 Mg/yr under Regulatory Alter-
nativ 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
significantly impacted.  No closures from the regulatory alternatives
predicted under either scenario.  Other impacts such as employment are
unlikely to be substantial, as shown in Table C-28 of Appendix C.
                                 9-8

-------
                       10.  QUANTITATIVE RISK ASSESSMENT

10.1  Comment:  One Commenter IV-D-14 stated that EPA's prediction  of  the
      leukemia risk to the community is overstated because if  reflects  an
      extremely conservative linear, non-threshold extrapolation model  and
      exposure assumptions.  The commenter also stated that EPA's assumption
      that individuals are exposed to the maximum annual ground level
      concentration of benzene for 14 hours/day, 365 days/year for  70  years are
      unrealistic assumptions, and lead to exaggerated risk calculations.  Other
      comments suggested 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 suggested that the  EPA monitor
      benzene near these facilities to verify the model, and conduct
      epidemiologic studies of the communities surrounding the facilities.

      Response:  Since a specific environmental carcinogen is  likely to  be
      responsible for at most a small fraction of a community's overall  cancer
      incidence, and since 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.  Therefore EPA must rely largely upon
      mathematical modeling techniques to estimate human health risks.   These
      techniques, termed "quantitative risk assessment," are means  whereby the
      risk of adverse health effects from exposure to benzene  in the ambient
      environment can be estimated mathematically by extrapolating  effects found
      at higher occupational exposure levels to lower concentrations
      characteristic of human exposure in the vicinity of industrial sources  of
      benzene.  The analysis estimates the risk of cancer at various levels of
      exposure.  A unit risk factor for benzene is derived from the
      dose-response relationship observed in the occupational  studies.   The unit
      risk factor represents the cancer risk for an individual exposed  to a unit
      concentration of a carcinogen  (e.g., 1 ppm) for a lifetime.
           While EPA agrees that the linear, non-threshold model is conservative
      in nature and would tend to provide a reasonable upper bound  to  the

                                     10-1

-------
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 the 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, non-threshold model over other mathematical  models  (Water
Criteria Documents Availability, 45 FR 79319, November  28, 1980).   In this
notice the 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
     characterstic 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  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-acetylaminofluorene in the large scale ED01 study at  the National
     Center for Toxicological  Research, and initiation stage of the
                               10-2

-------
           the two-stage cardnogenisis 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 methodologies  are  not sufficiently
      sensitive to detect a casual  association between  chronic,  low-level
      benzene exposure and cancer.  While such studies  have  revealed an
      association between occupational exposure to  ambient  benzene  and  leukemia,
      such associations may not be  confounding factors.  These  include  genetic
      diversity, population changes and mobility, lack  of  consolidated medical
      records, lack of historical benzene exposure  data over each  individual's
      lifetime, public 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 Clean  Air  Act,  the  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 epidemiologic studies of occupationally exposed
      population.
           The basic assumptions implicit in the risk  assessment methodology are
      that all exposure occurs at people's residences,  that  people  stay  at the
      same location for 70 years, that the ambient  air  concentrations  and  the
      emissions which cause these concentrations persist for 70 years,  and that
      the concentrations are the same inside and outside the residences.  From
      this it can be seen that public exposure is based on  a hypothetical
      premise.  It is not known whether this results  in an  overestimation  or an
      underestimation of public exposure.

10.2  Comment:  One commenter (IV-D-13) suggested that  in  using the  Human
      Exposure Model  (HEM), and not the Industrial  Source  Complex model, to
      estimate annual average ground level concentrations  of benzene around  coke
      oven by-product recovery plants, EPA has underestimated exposure to  the

                                     10-3

-------
population living near those  facilities.  The  commenter alleges that EPA
has admitted the model underestimated  exposure 200  to  300 percent in the
benzene fugitive emissions  rulemaking.  Therefore,  the commenter states
that risk to the most exposed  individuals should  be much  higher.

Response:  These same concerns were  raised  by  the Natural  Resources
Defense Council  (NRDC) in a petition to the  Administrator of EPA to
reconsider four  final benzene  decisions as  published in the Federal
Register  (49 FR  235478, June  6,  1984).  The  EPA  responded to these
concerns  in EPA's response  to  the  NRDC petition  (50 FR 	,
August 8, 1985).  The EPA reviewed NRDC's concerns  about  correcting' the
alleged bias in  the  assessment used  in evaluating the  benzene fugitive
emissions standard.  Specifically, EPA reviewed  NRDC's concerns about
estimating exposure  to benzene around  plants using  or  producing benzene.
In doing  so, EPA decided to recalculate the  exposure assessment used in
the benzene fugitive emissions 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 analysix 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, and, therefore, the  alleged underestimation of exposure was
not centrally relevant to establishing a standard.   A  factor of 2 or 3 is
well within the  bounds of modeling uncertainty.   The uncertainty would
have to result in a  computational  change to  the  estiamated exposure of
about one order  of magnitude to  significantly  change the  estimated risk.
     The  EPA has developed  the HEM for use  in  computing both exposure and
risk calculations to pollutants  emitted from industrial  sources.  The HEM
is a general  model  containing  both (1) an atmospheric  dispersion model,
with included meteorological data, and (2)  a population distribution
estimate  based on Bureau of Census data.  The  input data  needed to operate
the HEM are source data, e.g., plant location,  heihgt  of  emission release
point, and volumetric rate  of  release temperature of the  combustion
off-gases.  Based on the source  data, the HEM  estimates the magnitude and
distribution of  ambient air concentrations  of  the pollutant in  the

                               10-4

-------
      vicinity of the source.  The model is programmed to estimate  these
      concentrations for a specific set of points within a  radial, distance of 50
      kilometers from the source.  The HEM numerically combines  the distribution
      of pollutant concentrations and people to produce quantitative expressions
      of public exposure to the pollutant.

10.3  Comment:  Commenter IV-D-13 contends that the benzene  unit  risk  factor
      used in the 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  oven by-product
      recovery plants.

      Response:  On October 17, 1984, the Natural Resources  Defense Council
      (NRDC) petitioned the Administrator of the EPA to reconsider  four final
      decisions regarding benzene emissions as published in  a  Federal  Register
      notice June 6, 1984 (49 FR 23478).  Of central relevance to the  petition
      was the contention 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 re-evaluated  the unit risk  factor
      for benzene accordingly.  The methodology for the evaluating  of  the unit
      risk estimate is described in a document entitled Interim  Quantitated
      Cancer Risk Estimates Due to Inhalation of Benzene (Docket  OAQPS 79-3(1),
      VIII-A-4) and is summarized in EPA's response to the  NRDC  petition  (50
      FR	, August 8, 1985).  In the re-evaluation  of  the unit risk
      estimate, the 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), and computed  a  geometric
      mean of each point risk estimate.  The data were aggregated in order 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

                                     10-5

-------
      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  factor  (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 0.222/ppm (6.7 x  10~6  per  yg/m3)  to
      0.026/ppm (8.0 x 10"6 per yg/m3).  The revised  estimate  represents  a 17
      percent increase in the estimate used in the June, 1984  decisions.
          The significant gap between EPA's revised risk estimate  (a  17%
      increase) and the 15 fold 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 86C3F mice  of an  animal
      bioassays have been considered in the Agency's  revaluation,  EPA believes
      that the unit risk estimate for inhalation of benzen is  appropriately
      based upon the principal epidemiologic studies  since these studies  are of
      recognized quality and have the greatest relevance in  the  estimation of
      health risks for the general population.  Well-conducted epidemiological
      studies provide direct evidence of a casual 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 benzen exposures and durations of
      exposure, the epidemiologic studies showed a 3-fold  to 20-fold  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
      proposed guidelines for carcinogen risk assessment (49 FR  46294).
      Although a clear dose-response association between carcinoma  and benzene
      exposure was demonstrated in rodent bioassays,  the EPA believes  that
      human data, 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.
10.4  Comment:  One commenter expressed the opinion (IV-D-14)  that  the benzene
      unit risk factor overstates the true risk by at least  one  order  of

                                     10-6

-------
magnitude.  Moreover a minor adjustment  of  7  percent  in  the unit risk
factor published in the June 6,  1984  Federal  Register notice as response
to public comments  (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 factor.
The commenter maintained 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 felt  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).   While the 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  exopsure 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 three cases  of leukemia in  a  cohort  of  594 chemical
workers when only 0.8 cases were 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.
     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

                               10-7

-------
      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.

10.5  Comment:  Several commenters argued 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
      the EPA (IV-D-10, IV-D-12, IV-D-17).  They argued 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 is currently evaluating  benzene  emissions
      from the gasoline marketing industry (e.g., service stations, bulk plants,
      bulk terminals, and vehicle refueling operations) (49 FR 31709).  The
      preliminary results of this evaluation entitled Evaluation of Air
      Pollution  Regulatory Strategies for Gasoline Marketing Industry
      (EPA-450/3-84-012a, July, 1984) were released for public comment in  August
      1984. Benzene is a major industrial chemical ranking  among the  top fifteen
      with a U.S. production volume of almost 4.5 million megagrams (or  5
      million tons) in 1981.  In addition to industrially produced benzene,
      roughly an equal amount is found in gasoline.  The  vast  majority of
      benzene is derived from petroleum, with a smaller percentage produced  as a
      by-product of coke ovens.  Regardless of the magnitude of  these  other
      benzene sources, coke oven by-product recovery plants by themselves  emit
      about 26,000 megagrams per year benzene, and this has been estimated to
      result in  2.9 cancer incidences each year within the  population  exposed to
      emissions  from these plants.  Moreover, the lifetime  risk  to the most
      exposed individual residing in the vicinity of these  plants  is  a
      probability of leukemia of 6.3 x 10~3, and this is  considered as a
      probability above the prevailing cancer rate in the non-exposed
      population.  The EPA continues to believe that benzene emissions from
      these sources can be anticipated to pose significant  health  risks to the
      exposed population and that national emission standards  are  warranted.
                                     10-8

-------
10.6  Comment:  A commenter (IV-D-14) stated 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
      suggested 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 10~3.

      Response:  The EPA does not average the maximum  risk  calculations,  but
      assumes an aggregate of risk to the population residing  within  50
      kilometers of 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 exposure  to ambient
      concentrations of benzene emitted from the facilities,  and  for
      convenience, it is often divided by 70 and expressed  as  cancer  incidences
      per year.  Individual  risk, on the other hand, is  expressed as  "maximum
      lifetime risk" and reflects the probability of getting  caner  if  one were
      exposed continuously to the estimated maximum ambient air concentration  of
      benzene for 70 years.   The aggregate risk to the exposed  population
      residing near furnace  coke by-product recovery plants has been  calculated
      to be 2.6 and the aggregate risk to populations  residing  near  foundry coke
      by-product recovery plant has been estimated to  be 0.24  cases  of cancer
      per year.  Therefore,  in total almost 3 cases of cancer  per year are
      expected to occur in the population exposed to benzene  emissions from
      facilities, and this is directly attributed to the plant  emisisons  and  not
      other sources of benzene.
           The probability of contracting cancer to the  most  exposed  individual
      living in the vicinity of coke by-product facilities  has  been  estimated  to
      be about 6 in 1,000.  The EPA regards this level of risk  to be
      significant.  This level of risk is above the prevailing  risk  of cancer
      which is currently regarded to be a probability  of 1  in  4 over  one's
      lifetime.
           EPA does recognize that most human activities and  events  involve some
      degree of inevitable risk.  Risk may either be imposed  on the  individual
      (involuntary), or an individual may elect to accept a certain  level of

                                     10-9

-------
      risk as a result of acertain action such as cigarette  smoking.   This
      latter risk is considered a voluntary  imposition  of  risk.   In  the case of
      benzene exposure from the by-product recovery  facilities the  risk of
      cancer is imposed on the individual without consent  nor  acceptance.   The
      ambient air concentrations of benzene  emanating from these  facilities
      increases the individual's probability or odds of  getting cancer, as  well
      as adds to the prevailing cancer incidence.  The  current overall  lifetime
      cancer incidence rate is approximately 250,000 cases per 1,000,000 people.
      Thus benzene exposure to the population residing  near  coke  by-product
      recovery plants increases this nationwide rate of  cancer by three.

10.7  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 EPA has determined that the public  health risk  of ambient
      air exposure to benzene emitted from coke by-product recovery  plants  to be
      significant.  Therefore, the Administrator has decided to take  measures to
      reduce the atmospheric release of benzene from these sources.   The EPA has
      previously commented on the proopsal not to regulate POM as a  generic
    •  designation of pollutants under The CAA in a previous  Federal  Register
      notice (49 FR 5580, February 13, 1984).  The POM  decision does  not
      preclude the EPA from regulating specific carcinogenic compounds,
      especially if risks have been determined to be significant.
10.8  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:                  [To be inserted]
                                     10-10

-------
                    11.  EQUIPMENT LEAK DETECTION AND REPAIR

11.1  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
      concentration 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 from the source.  If an interference exists with this determination
      due to a nearby emission or leak, the local ambient concentration may be
      determined at distances closer to the source (but not closer than 25 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).

11.2  Comment:  Commenter IV-D-16 recommends that the regulation 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 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 change,
      the commenter considers the recordkeeping and reporting provisions only
      industry "self-enforcement" rules.
      Response:  Sealings and packings inherently leak; only the use of leakless
      equipment can prevent occasional leakage.  As discussed in response to
      comment 5.2,  EPA has determined that it would be unreasonable to require
      leakless equipment nationwide.  Because an occasional leak cannot be
      prevented, we cannot accept the commenter1s 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
                                     11-1

-------
      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.  We disagree.  The  regulation
      (Section 61.132-1) states that compliance will be determined by  review of
      records, reports, performance test results, and inspections  (emphasis
      added).  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,  we believe
      these provisions are adequate to ensure diligent monitoring  and  repair of
      leaks by plant personnel and effective enforcement by EPA.

11.3  Comment:  Commenter IV-D-13 requests that EPA reconsider changing the
      definition of an equipment "leak" from 10,000 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 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
      lower leak definition would reduce benzene emissions by  an  additional 85.2
      percent when coupled with directed maintenance programs  (already used by
      some companies).  The Natural  Resources Defense Council  (NRDC) states that
      data on "directed maintenance"* available to EPA contradict  its  position
      that a lower definition would not reduce emissions.
     *In "directed maintenance" efforts, the tightening of the packing  is
     monitored simultaneously and is continued only to the extent that  it
     reduces emissions.  In contrast, "undirected" repair means repairs such  as
     tightening valve packings without simultaneously monitoring the  result to
     determine if the repair is increasing or decreasing emissions.
                                     11-2

-------
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  synthetic
organic chemicals manufacturing 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
emission 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.  While many leaks can be successfully repaired  at  concentrations
less than 1,000 ppmv, even one valve repair failure  would  offset many
successful valve repairs.  Most data on leak repair  effectiveness  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.  While data between these values  are available,  they  are  not
sufficient to support a leak definition below  10,000  ppm.  As  the
commenter noted, while there is some evidence  that directed  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  which
generally have no leakage.  When the seal wears over  time, the
concentration and emission rate increases.  Properly  designed,  installed,
and operated seals have low instrument meter readings,  while 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
                               11-3

-------
      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.
           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, while 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), EPA 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 like 1,000 ppmv,
      EPA is selecting the clearly demonstrated leak definition of 10,000 ppmv
      instead of lower level.

11.4  Comment:  Commenter IV-D-13 refers to the proposal BID discussion
      indicating that on-line repair of valves by drilling into the valve
      housing and injecting a sealing compound is growing in acceptance,
      especially due to 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
                                     11-4

-------
the CAA, or that acknowledgment alone constitutes  sufficient  justification
for a regulatory requirement.  The BID does  state  on  page  4-5.2  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
housing of a critical valve containing a potentially  explosive  mixture.
Or, as discussed in the preamble to  the proposed rules of  49  FR 23533, the
valve (or the leak) may require removal or  isolation.  Also,  this  repair
approach cannot be used to control valves or other block valves that  are
frequently operated because the valve would  then be destroyed and  must be
replaced.  Because of 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 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 shutdown  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  flexibility necessary for
efficient handling of repair tasks while maintaining  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 chooses to  apply  this method,  it is
certainly not precluded under the standards.  To require this method  for
                               11-5

-------
      all  valves, however, would be premature and unwarranted.

11.5  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 time frame is  adequate  because
      monitoring 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  proposed rule
      at 49 FR 23541.
           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.
           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 maintenance 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 critical equipment  problems
      could easily postpone these arrangements.  Although 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 NRDC'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
                                     11-6

-------
      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.
           The commenter's suggestion that leaks can be detected and  repaired
      within a shorter time frame if repair workers accompany monitoring
      personnel may be helpful for plants able to make such arrangements.
11.6  Comment:  Commenter IV-D-13 recommends that the proposed  provisions for
      delay of repair beyond a unit shutdown be tightened to prevent  abuses.
      The commenter 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 could not
      reasonably claim a delay of repair due to lack of equipment.
11.7  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
                                     11-7

-------
      and operators of units exhibiting low leak frequencies because the  cost
      effectiveness of monthly/quarterly leak detection and  repair  becomes
      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.  While 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. .
11.8  Comment:  Commenter IV-D-13 states that the exemption  for difficult-to-
      monitor valves is not warranted.  Valves above 2 meters, according  to the
      commenter, can be reached by a sampling probe on a boom or  by a mobile
      "cherry picker."
      Response:  We disagree 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 elevating monitoring  personnel more
                                     11-8

-------
      than 2 meters 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
      petroleum refineries (see Docket Item IV-B-4).  Based on this analysis, we
      found the costs of using scaffolding for annual monitoring of benzene
      emissions from difficult-to-monitor valves 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 by use of 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 shown above.
           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.

11.9  Comment:  Commenter 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.

                                     11-9

-------
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  depending on
the individual application.   If a second  valve  is  used, the upstream 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 at a reasonable cost ($730/Mg benzene  and  $460/Mg  VOC  at
foundry plants; $640/Mg benzene and $450/Mg  VOC  at  furnace  plants.
     The commenter suggests an alternative standard of no detectable
leaks, with applicable 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 our
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 emissions 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
recordkeeping and reporting requirements  associated with  performance
testing.  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.
                               11-10

-------
                        12.  RECORDKEEPING AND REPORTING

12.1  Comment:  Commenter IV-D-9 asks if monitoring and recordkeeping
      requirements can be modified if a technology better than BAT is used.

      Response:   Section 61.134 of the standards describes the procedures  for
      obtaining  EPA approval of alternative means of emission reduction that are
      equivalent to or better than BAT.  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, we would consider requiring
      monitoring, recordkeeping, and reporting requirements appropriate for the
      alternative on a case-by-case basis.
12.2  Comment:  Commenter IV-D-13 argues that records and reports should be
      maintained permanently (or for a minimum of 5 years) due to 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, while less than that allowed by OMB, applies to significantly
      detailed plant records that would help enforcement personnel 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 NRDC 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.  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.
                                     12-1

-------
12.3  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 identification  of
      "repeat offenders" and the failure of the rules to require  reporting  of
      the specific identity of leaking equipment—only totals.

      Response:  The recordkeeping and reporting requirements have been
      developed to provide sufficient compliance information to  enforcement
      personnel, through the most practicable, cost-effective approach for  the
      industry.  We do not consider, as the commenter suggests, that  not
      requiring identifying tags on "repeat offenders" or reporting  the  specific
      identity of leaking equipment will  impair effective enforcement of the
      standards.
           The standards require that each leaking source be identified.   If a
      tag is used for this purpose, it may be removed after repair.   The use of
      additional tags to identify sources with a tendency to leak may be
      confusing because only leaking sources are actually tagged.  This  approach
      also recognizes the temporary nature of equipment tags, which  tend to
      become lost or illegible in by-product plant environments.  Consequently,
      a more permanent identification tag for sources with a "leak history" even
      after repair may not be practical.
           Neither do we recognize the enforcement benefit gained by  requiring
      the reporting of the specific identity of leaking equipment.   For
      enforcement, the information of primary concern is whether  all  leaks
      detected, including those from "repeat offenders," have been repaired.
      The standards require reporting of the number of leaks detected and  the
      number of leaks not repaired for each type of source.  This information
      is adequate for enforcement personnel to determine if the  leak  detection
      and repair program is operating adequately and whether an onsite
      inspection may be required.  Enforcement personnel can ascertain the
      specific identity of leaking equipment and other detailed  information by
      reviewing the compliance records available at the plant site or by
      requesting "leak history"'data under the authority of Section  114 of  the
      CAA.
                                     12-2

-------
                               13.  MISCELLANEOUS

13.1   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 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.
13.2   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,
       requiring 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."

       Response:  We agree with the commenter1s suggestion that the definitions
       of "tar decanter" be clarified.  In response, the final regulations
       contain the following definition:

                 "Tar decanter" means any vessel, tank, or other type
                 container that functions to separate heavy tar and sludge from
                 flushing liquor by means of gravity, heat, or chemical emulsion
                 breakers.  Any sump, vessel, tank or ether type container from

                                      13-1

-------
                 which tar is removed continuously by means of  a  chain  or  drag
                 conveyor is a tar decanter.  A tar decanter may  also be known
                 as a flushing-liquor decanter.

13.3   Comment:  Commenter IV-D-9 asks to what extent upstream  and downstream of
       the rotors does the definition of "exhauster" extend?

       Response:  In response to the commenter's question, we have revised
       §61.131 of the 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.

13.4  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) (1) (B) 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 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
      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 proposal 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
                                      13-2

-------
      compliance while the Agency is  evaluating the waiver application.

13.5  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  non-compliance.  For  example,  for leak detection  and repair
      programs for pumps, valves, and exhausters, the reports focus  on the
      number of  leaks detected and the  number  of leaks not  repaired.  For
      wash-oil  scrubbers, the reports denote periods when operating  parameters
      are outside design ranges, indicating that the source may be out of
      compliance.

13.6  Comment:   Commenter IV-0-17 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.
           Section 112 of the CAA requires that an emission standard be
      established 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
                                      13-3

-------
      standard is not feasible to prescribe or enforce, one of the  other  types
      of standards (including any combination) can be applied.
           Gas blanketing has been selected as the basis of the  standards  for
      both tar decanters and tar-dewatering vessels.  In the preamble  discussion
      in 49 FR 23528, EPA explained why an emission standard, such  as  a zero
      emissions 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 due to 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 operation 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 (or an alternative medium, such, as wash oil) was selected as  the
      basis of the proposed standards.  Collecting naphthalene in tar  (or  wash
      oil) would eliminate naphthalene processing operations (including
      naphthalene sumps) and the emissions that result 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 emissions limit is considered feasible to prescribe
      and enforce and has been selected as the format for the naphthalene
      processing standards.
13.7  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 limitation.  The
                                      13-4

-------
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 proposed rules in  49 FR 23537-23539, the control
efficiency of source enclosure is  theoretically 100 percent.  However,
eventual deterioration of the gasket seal  (or the cover) may result in
occasional leaks, even with proper operation and maintenance.  Because
mass emissions from these leaks cannot reasonably be measured, EPA
conservatively 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
maintenance (O&M) of the sealed enclosures, i.e., to locate and repair any
leaks that may have developed in the control system.  Thus, the
commenter's 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 O&M would have to be determined specifically for the
alternative control system.  Without further information, we believe that
the 98-percent value is the best estimate available for comparing the
efficiency of ah alternative control system.
                                13-5

-------
          APPENDIX A
ENVIRONMENTAL IMPACT ANALYSIS

-------
                                    Table A-l.  FURNACE COKE  BY-PRODUCT  RECOVERY  PLANTS:
                                             COKE OVEN AND  PLANT  CAPACITY  STATUS*
                                                        (1,000 Mg/yr)
No.
*1
2
3


4
5
6


7

8

9



no

11
*12

13

*14
15



16

17


18

*19






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, Dearborn, MI



U.S. Steel, Fairless Hills, PA

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

Armco 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






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
0
0
0
0
0
0
0
2
2
0
2
2
0
0
2
0
0
0
0
0
0
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

1,097
0



0

0


0

1,496






Under
construction
0
0
0


0
0
298


0

0

0



0

0
0

0

0
0



0

0


0

0






Existing
plant
total
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






footnotes on last page of table.
(continued)

-------
                                    Table A-l.   FURNACE COKE BY-PRODUCT RECOVERY PLANTS:
                                                COKE OVEN AND PLANT CAPACITY
                                              STATUSa (1,000 Mg/yr) (Continued)
No.
20



21





22


23

24




*25



26



27








28






Battery
Plant no.
Wheeling-Pitt, E. Steubenville, WV 1
2
3
8
LTV Steel, Cleveland, OH 1
2
3
4
6
7
Armco 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., MD 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
0
0
0
0
0
2 ,
0
0
0
0
0
0
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
51
79
51
51
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
Existing
Hot Cold Under plant
Online idle idle construction total
,1,509 000 1,509



1,486 0 274 0 1,760





1,776 000 1,776


1,790 000 1,790

1,792 0 00 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






Footnotes on last page of table.
(continued)

-------
                                    Table A-l.  FURNACE COKE BY-PRODUCT RECOVERY PLANTS:
                                COKE OVEN AND PLANT CAPACITY STATUS (1,000 Mg/yr) (Continued)
No. Plant
29 U.S. Steel, Gary, IN






30 U.S. Steel , Clairton










TOTALS
* Denotes cold idle plants
STATUS:
0 = online 2 =
1 = hot idle 3 =
Battery
no.
1
2
3
5
7
13
15
16
, PA 1
2
3
7
8
9
15
19
20
21
22
B
(30 plants)
(24 plants)
•

cold idle
under construction
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.

-------
                 Table A-2.  FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS:  COKE OVEN AND PLANT CAPACITY  STATUS
                                                       (1,000 My/yr)


No.
1

2

3
4

5

6
7


8

9


10


11

12




13


14



Plant
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
TOTALS

Battery
no.
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

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


Status
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
4

No. of
ovens
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


Online
130

132

157
161

207

299
330


6

477


499


521

563




470


617
4,563

Hot
idle
0

0

0 '
0

0

0
0


0

0


0


0

0




113


0
113

Cold
idle
0

0

0
0

0

0
0


402

0


0


0

0




0


0
402

Under
construction
0

0

0
0

0

0
0


0

0


0


0

0




0


0
0
Existing
plant
total
130

132

157
161

207

299
330


402

477


499


521

563




583


617
5,078
STATUS
0 =
online 2 = cold
idle








1 = hot idle                3 = under construction
Note:  data current as of November 1984.

-------
                                               Table A-3.  FURNACE COKE BY-PRODUCT PLANT OPERATING PROCESSES
No.
Plant
                             Excess-
                             ammonia  Flushing-   Tar-
  Tar       Tar       Tar    liquor   liquor      interc.
decanter dewatering storage  storage  circ. tank   sump
                   Light-
                    oil      BTX   Benzene
                  storage  storage  storage
                                  Naphthalene processing/handling
                                  Denver      Naphth.    Naphth.
                                  flo  unit     melt pit  dry tanks
1
2
3
4
5
6

7
8
9
10
11
• 12
13
14
15
16
17

18
19
p> 20
1
Ul 21
22
23

24
25
26
27

28
29
30
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, IL
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
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
1
1
1
1
0
0

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
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 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 1
1 1

1 1
1 1
1 ..1

1 1
1 1
1 1
1 1

1 1 .
1 1
1 1
1
I 1
1
1
I 1
1

1
1
1
1
1
1
1
1
1
1
1

1
1
L 1

1 1
1 1
L 0

1 1
L 1
1 1
t 1

1 1
1 1
1 1
0
0
0
1
1
1

0
0
1
0
0
1
0
0
0
0
1

0
0
0

1
1
0

0
0
1
0

0
0
0
0
0
0
0
1
0

0
0
0
0
0
0
0
0
0
0
0

0
0
0

0
0
0

0
0
0
1

0
0
1
1
0
0
1
0
1

1
1
1
1
0
0
0
1
0
0
1

1
1
0

0
1
0

1
1
1
0

0
1
0
1
0
0
1
0
1

1
1
1
1
0
0
0
1
0
0
1

1
1
0

0
1
0

1
1
1
0

0
1
0
1
0
0
1
0
1

1
1
1
1
0
0
0
1
0
0
1

1
1
0

0
1
0

1
1
1
0

0
1
0
                           TOTALS
                    30
            28
                                                          30
                                                                   30
30
30
                                                                                30
                                                                                                          10
16
16
16
                                                                                                                                                 (Continued)

-------
Table A-3.  FURNACE COKE BY-PRODUCT PLANT OPERATING PROCESSES  (Continued)
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
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, Ml
U.S. Steel , Fairless Hills, PA
LTV Steel, Warren, OH
LTV Steel , E. Chicago, IN
Annco 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
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
0
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
Wash-
oil
final
cooler
0
0
0
0
0
0

0
0
0
0
1
0
0
1
0
0
0

0
0
1

1
0
0

0
0
0
0

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

1
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

1
1
1

1
1
0

1
1
1
1

1
1
1
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
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
    TOTALS   16
30
30
                                                            30
                                  30
                                                                                    30

-------
Table A-4.  FOUNDRY COKE BY-PRODUCT  PLANT  OPERATING PROCESSES
No.
1

2
3
4
5
6
7
8
9

10
11
12
13
14

Tar
Plant decanter
Chattanooga Coke, 1
Chattanooga, TN
IN Gas, Terre Haute, IN 1
Koppers, Toledo, OH 1
Empire Coke, Holt, AL 1
Koppers, Erie, PA 1
Tonawanda, Buffalo, NY 1
Carondolet, St. Louis, HO 1
AL Byproducts, Keystone, PA 1
Citizens Gas, 1
Indianapolis, IN
Jim Walters, Birmingham, AL 1
Shenango, Pittsburgh, PA 1
Koppers, Woodward, AL 1
AL Byproducts, Tarrant, AL 1
Detroit Coke, Detroit, MI 1
TOTALS 14
Tar Tar
dewatering storage
1

1
1
0
1
1
1
1
1

1
1
1
1
1
13
1

1
1
1
1
1
1
1
1

1
1
1
1
1
14
Excess-
ammonia Flushing- Tar- Light-
liquor liquor interc. oil
storage circ. tank sump storage
1111

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
0
1
0
1
0
1
0

1
1
1
1
1110
14 14 14 9
BTX
storage
0

0
0
0
0
0
0
1
0

0
0
0
0
0
1
Benzene
storage
0

0
0
0
0
0
0
1
0

0
0
0
0
0
1
Naphthalene
Denver
flo. unit
1

1
0
1
0
0
0
1
1

1
0
0
1
0
7
processing/handling
Naphth. Naphth.
melt pit dry tanks
1

1
0
1
0
0
0
1
1

1
0
0
1
0
7
1

1
0
1
0
0
0
1
1

1
0
0
1
0
7

-------
                               Table A-4.  FOUNDRY COKE BY-PRODUCT PLANT OPERATING PROCESSES (Continued)



No.

Wash-oil
final
Plant cooler
Direct-
water
final
cooler
Tar-
bottom
final
cooler

Light-
oil
sump
Light-



oil decanter/
condenser
vent
Wash-oil
decanter
Wash-oil
circ. tank
Equipment
leaks
    1    Chattanooga Coke,
           Chattanooga, TN
    2    IN Gas, Terre Haute, IN
    3    Koppers, Toledo, OH
    4    Empire Coke, Holt, AL
    5    Koppers, Erie, PA
    6    Tonawanda, Buffalo, NY
    7    Carondolet, St. Louis, MO
    8    AL Byproducts, Keystone, PA
    9    Citizens Gas,
           Indianapolis, IN
   10  .  Jim Walters, Birmingham, AL
   11    Shenango, Pittsburgh, PA
   12    Koppers, Woodward, AL
^ 13    AL Byproducts, Tarrant, AL
 I  14    Detroit Coke, Detroit, MI
CO
                               TOTALS
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
0
0
1
1
1
0
0
1
0
0
1
0
0
0
0
0
0
0
0
1
0
0
1
0
1
0
0
0
1
0
1
1
1
1
0
1
0
1
0
0
0
1
0

1
1
1
1
0

-------
   Table A-5.  ESTIMATED NATIONWIDE BENZENE EMISSIONS
FROM FURNACE AND FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS
Furnace plants
Source
1. Direct-water final-cooler
cooling tower
2. Tar-bottom final-cooler
cooling tower
3. Naphthalene processing/
handl ing
4. Light-oil decanter/
condenser vent
5. Tar-intercepting sump
6. Tar decanter
7. Tar dewatering
8. Tar storage
9. Light-oil sump
10. Light-oil storage
11. BTX storage
12. Benzene storage
13. Flushing liquor
circulation tank
14. Excess-ammonia liquor
storage tank
15. Wash-oil decanter
16. Wash-oil circulation tank
No. of
affected
plants3
16
4
16
29
30
30
28
30
29
29
10
4
30
30
29
29
Capacity6
(1,000 Mg/yr)
21,430
5,729
21,430
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
2,293
3,456
4,351
3,527
839
550
660
242
67
61
412
412
154
154
No. of
affected
plants3
7
2
,7
9
14
14
13
14
9
9
2
1
14
14
9
9
Foundry plants
Capacity
(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
5,078
5,078
3,290
3,290
Furnace and foundry plants
Nationwide
emissions
(Mg/yr) .
470
37
186
158
227
184
49
29
27
10
3
1
33
33
7
7
No. of
affected
plants3
23
6
23
38
44
44
41
44
38
38
12
5
44
44
38
38
Capacity
(1,000 Mg/yr)
23,814
6,449
23,814
47,304
50,882
50,882
44,864
50,882
47,304
47,304
12,445
10,925
50,882
50,882
47,304
47., 304
Nationwide
emissions
(Mg/yr)
6,256
438
2,479
3,614
4,578
3,711
887
578
687
253
70
62
446
446
161
161
(Continued)

-------
                                                       Table A-5.  ESTIMATED NATIONWIDE BENZENE EMISSIONS
                                              FROM FURNACE AND FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS (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

Capacity6
(1,000 Mg/yr)
44,014
44,014
44,014
44,014
44,014
44,014

Nationwide
emissions
(Mg/yr)
370
249
168
17
33
11
24,200
No. of
affected
plants3
9
9
9
9
9
9

Foundry plants
Capacity
(1,000 Mg/yr)
3,290
3,290
3,290
3,290
3,290
3,290

Nationwide
emissions
(Mg/yr)
101
68
46
5
9
3
1,700
Furnace and foundry plants
No. of
affected
plants3
38
38
38
38
38
38

Capacity
(1,000 Mg/yr)
47,304
47,304
47,304
47,304
47,304
47,304

Nationwide
emissions
(Mg/yr)
471
317
214
22
42
14
25,900
 I
M
o  lumber of plants having this source (30 furnace plants, 14 foundry plants, or 44 furnace and foundry plants  combined.

    bCapacity of plants with this source.

-------
    Table A-6.  ESTIMATED NATIONWIDE VOC* EMISSIONS
FROM FURNACE AND FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS

1.
2.
3.
4.
5.
1 6.
h~*
•- 7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Source
Direct-water final-cooler
cooling 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

No. of
affected
plantsb
16
4
16
29
30
30
28
30
29
29
10
4
30
30
29
29
Furnace plants
Capacity0
(1,000 Mg/yr)
21,430
5,729
21,430
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)
90,842
6,302
3,600
4,931
. 9,252
7,512
19,654
12,871
942
347
96
61
591
591
219
219

No. of
affected
plantsb
7
2
7
9
14
14
13
14
9
9
2
1
14
14
9
9
Foundry plants
Capacity0
(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
5,078
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
. plantsb
23
6
23
38
44
44
41
44
38
38
12
5
44
44
38
38
Capacity0
(1,000 Mg/yr)
23,814
6,449
23,814
47,304
50,882
50,882
44,864
50,882
47.304
47,304
12,445
10,925
50,882
50,882
47,304
47,304
plants
Nationwide
emissions
(Mg/yr)
98,219
6.880
3,893
5,157
9,735
7,903
20,791
13,542
980
362
100
62
639
639
229
229
(continued)

-------
                                                     Table A-6.  ESTIMATED NATIONWIDE VOC^ EMISSIONS
                                           FROM FURNACE AND FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS (Continued)



No. of
affected
Source plants'3
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

Furnace plants
Capacity.0
(1,000 Mg/yr)
44,014
44,014
44,014
44,014
44,014
44,014

Foundry plants
Nationwide
emissions
(Mg/yr)
528
355
240
25
47
16
159,200
No. of
affected
plants'5
9
9
9 •
9
9
9

Capacity0
(1,000 Mg/yr)
3,
3,
3,
3,
3,
3,

290
290
290
290
290
290

Nationwide
emissions
(Mg/yr)
159
107
73
8
14
5
11,700
Furnace and foundry plants
No. of
affected
plants6
38
38
38
38
38
38

Nationwide
Capacity0 emissions
(1,000 Mg/yr) (Mg/yr)
47,304
47,304
47,304
47,304
47,304
47,304

687
463
312
33
61
21
170,900
aBenzene and other VOC.

bNumber of plants having this source (30 furnace plants, 14 foundry plants, or 44 furnace and foundry  plants  combined).

°Capacity of plants with this source.

-------
Table A-7.  THE EFFECT OF CONTROL OPTIONS ON REDUCING BENZENE 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.
15.

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
Wash-oil decanter
Wash-oil
circulation tank
No. of
affected
Control option plants3
No national emission standard.
1. Tar-bottom final cooler
2. Wash-oil final cooler
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
Coke oven gas blanketing
Coke oven gas blanketing

30
16
20
29

30

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

30
30
29
29

National
benzene
emissions
(Mg/yr)
24,200
8,480
8,480
3,456

4,351

3,527
839
839
550
550
660
242
242
67
67
61
61
412

412
412
154
154

Controlled
benzene
emissions
(Mg/yr)
24,200
1,901
0
78

87

176
84
17
55
11
13
24
5
7
1
6
1
8

41
8
3
3

FOUNDRY PLANTS
No. of
affected
plants3
14
7
9
9

14

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

14
14
9
9

National
benzene
emissions
(Mg/yr)
1,690
693
693
158

227

184
49
49
29
29
27
10
10
3
3
1
1
33

33
33
7
7

FURNACE AND FOUNDRY PLANTS
Controlled
benzene No. of
emissions affected
(Mg/yr) plants3
1,690
159
0
3

5

9
5
1
3
0.6
0.5
1
0.2
0.3
0.1
0.1
0.03
0.7

3
0.7
0.1
0.1

44
23
29
38

44

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

44
44
38
38

National
benzene
emissions
(Mg/yr)
25,900
9,173
9,173
3,614

4,578

3,711
887
887
578
578
687
253
253
70
70
62
62
446

446
446
161
161

Controlled
benzene
emissions
(Mg/yr)
25,900
2,060
0
82

92

186
89
18
58
12
14
26
5
7
1
6
1
9

45
9
3
3

                                                                                                             (continued)

-------
                           Table A-7.  THE EFFECT OF CONTROL OPTIONS ON REDUCING BENZENE EMISSIONS AT FURNACE
                                                   AND FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS (Continued)
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
No. of
affected
plants3
29
29
29
29
29
29
29
29
29
29
29
29
Nationwide Controlled
benzene benzene
emissions
(Mg/yr)
370
370
370
249
249
249
168
168
168
17
17
17
emissions
(Mg/yr)
108
62
0
93
- 69
0
93
80
0
8
6
0
Foundry plants
No. of
affected
plants3
9
9
9
9
9
9
9
9
9
9
9
9
Nationwide Controlled
benzene benzene
emissions
(Mg/yr)
101
101
101
68
68
68
46
46
46
5
5
5
emissions
(Mg/yr)
29
17
0
25
19
0
26
22
0
2
2
0
Furnace and foundry plants
No. of
affected
plants3
38
38
38
38
38
38
38
38
38
38
38
38
Nationwide Controlled
benzene benzene
emissions
(Mg/yr)
471
471
471
317
317
317
214
214
214
22
22
22
emissions
(Mg/yr)
137
79
0
118
88
0
119
101
0
10
8
0
20. Sampling connection  Closed purge sampling
      systems

21. Open-ended lines     Cap or plug
                                                           29
                                                           29
33
11
38
38
42
                                                                                                                                            14
aNumber of plants having the source (30 furnace plants, 14 foundry plants, or 44 furnace and foundry plants combined;
 includes plants currently on cold idle).

-------
                                        Table A-8.  THE EFFECT OF BENZENE CONTROL OPTIONS ON REDUCING VOCa
                                         EMISSIONS AT FURNACE AND FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS
Furnace plants

1.
2.

3.

4.

5.
Jf ** •
1
}_J
01 7.

8.
9.

10.

11.

12.

13.


14.
15.
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
Control option
No national emission standard
1. Tar-bottom final cooler
2. Wash-oil final cooler
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

Coke oven gas blanketing
Coke oven gas blanketing
No. of
affected
plants
30
16
20
29

30

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

30
30

29
29
VOC
nationwide
emissions
(Mg/yr)
159,200
100,744
100,744
4,931

9,252

7,512
19,654
19,654
12,871
12,871
942
347
347
96
96
61
61
591

591
591

219
219
Controlled
VOC
emissions
(Mg/yr)
159,200
29,875
0
112

185.

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

59
12

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

14

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

14
14

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

482

391
1.137
1,137
671
671
38
15
15
4
4
1
1
48

48
48

10
10
Controlled
VOC
emissions
(Mg/yr)
11,700
2,943
0
5

10

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

5
1

0.2
0.2
Furnace and foundry plants
No. of
affected
plants
44
23
29
38-

44

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

44
44

38
38
VOC
nationwide
emissions
(Mg/yr)
170,900
108,992
108,992
5,157

9,735

7,903
20,791
20,791
13,542
13,542
980
362
362
100
100
62
62
639

639
639

229
229
Controlled
VOC
emissions
(Mg/yr)
170,900
32,367
0
116

195

395
2,079
416
1,354
271
20
36
17
10
2
6
1
13

64
13

5
5
circulation tank

-------
                                             Table A-8.  THE EFFECT OF BENZENE CONTROL OPTIONS ON REDUCING VOCa
                                        EMISSIONS AT FURNACE AND FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS (Continued)
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
No. of
affected
plantsb
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
No. of
affected
plants'1
9
9
9
9
9
9
9
9
9
9
9
9
VOC
nationwide
emissions
(Mg/yr)
159
159
159
107
107
107
73
73
73
8
8
8
Controlled
VOC
emissions
(Mg/yr)
46
27
0
40
30
0
40
34
0
3
3
0
Furnace and foundry plants
No. of
affected
plants'3
38
38
38
38
38
38
38
38
38
38
38
38
VOC
nationwide
emissions
(Mg/yr)
687
687
687
463
463
463
312
312
312
33
33
33
Controlled
VOC
emissions
(Mg/yr)
200
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
aBenzene and other VOC's.

bNumber of plants having this source (30 furnace plants, 14 foundry plants, or 44 furnace and foundry plants combined).

-------
                                  Table A-9.  Energy Use at Model By-Product  Plants
 I
I-1
~J
User
Gas blanketing
Tar decanter, tar-intercepting sump,
and flushing-liquor circulation tank
Tar dewatering, tar storage
Excess ammonia-liquor storage tank
Condenser, light-oil decanter, wash-oil
decanter, and circulation tank
Wash-oil scrubber
Excess ammonia liquor storage tank
Benzen storage tank
Final cooler
Tar-bottom final cooler
Wash-oil final cooler
Furnace plants3
Steam Electricity
(Mg/yr) (MWh/yr)
350
440
126
174
24 0.4
0.9
380 98
1,210 1,330
Foundry Plants'1
Steam Electricity
(Mg/yr) (MWh/yr)
162
111
116
127 26
303 333
            4,000 Mg  coke/day.

            1,000 Mg  coke/day.

-------
             Table A-10.  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)       (a  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-18

-------
Table A-ll.  Yields--Foundry vs. Furnace  Coke  Plants
Year
1976
1977
1978
1979
1980
1981
1982
1983
Average
Ratios-
merch/furn
Coal to
Merch.
1.35

1.31
1.34
1.34

1.32
1.29
1.325
Coke Ratio
Furn.
1.46

1.47
1.47
1.46

1.47
1.46
1.465
Light Oil
gal/ton of
Merch.
1.67

1.77
1.82
1.82



1.77
Yield
coal
Furn.
2.58

2.9
2.51
2.67



2.665
.664
Tar
gal/ton
Merch.
5.26

5.52
5.94
5.97



5.6725
Yield
of coal
Furn.
7.77

7.78
7.86
8.27 .


8.08
7.952
.713
Gas Yield
1,000 ft3/ton coal
Merch. Furn.
9.21

9.23
9.03
8.94



9.1025
11.02

11.2
11.04
11.14


10.37
10.954
.831
LO/Gas
gal LO/1
Merch.
.18
— -
.19
.2
.2



.1925
Cone.
,000 ft3
Furn.
.23

.26
.23
.24



.24
.802
LO/Tar Cone.
gal/gal
Merch. Furn.
.32 .33

.32 .37
.31 .32 •
.3 .32



.3125 .335
.933

-------
            Table A-12.  Correction Factor Computation for  Foundry
                        Coke By-Product Recovery Plants
Source
 Concentration
   Adjustment
Volume (throughput)
    Adjustment
  Total
Correction
LO Plant
  Benz.  in LO

(63.5/70)=0.907
     LO Yield
    Coke Basis
x(0.664)(l.325/1.465)
                                                                  0.54
Water Contact
with CO gas
  Benz.  in LO
 LO in CO gas
 (0.907x0.802)
                                                                  0.73
Tar Sources
  Benz.  in LO
 LO in CO gas
 (0.907x0.802)
    Tar Yield
    Coke Basis
x(0.713)(l.325/1.465)
                                                                  0.47
Equip. Leaks
  Benz.  in LO
    (0.907)
                                                                  0.91
                                    A-20

-------
              Table A-13.  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-liquor
circulation tank
Excess-ammonia
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
a^n)||sion factors are not related to coke  production  capacity and are listed in

-------
                            Table A-14.  Benzene Emission Factors for Equipment Leaks
Furnace Plant Foundry PI
Benzene emission factors Benzene emissio
Percent of VOC emission (kg benzene/source day) (kg benzene/sou
sources
leaking
initial ly
Valves 11
Pumps 24
Exhausters 35
Pressure relief ^
devices
Sampling d
connections
Open-ended lines d
1
i^J aPlant A recovers light oil.
factor Plant A"
(kg/source
day)
0.26 0.18
2.7 1.9
1.2 0.28C
3.9 2.7
0.36 0.25
0.055 0.038
For furnace and foundry plants it
Plant Bu Plant A,a
0.22 0.16
2.3 1.7
0.28C 0.25
3.4 2.5
0.31 0.23
0.047 0.035
assumes 70 percent and 63.5 percent
ant
n factors
rce day)
Plant B,b
0.20
2.1
0.25
3.1
.28
.043

bPlant B recovers refined benzene.  For furnace and foundry plants, it assumes 86  percent  and  78 percent benzene
 in light oil and refined benzene, respectively.

C23.5 percent benzene in nonmethane hydrocarbon.

''This 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

Development of Revised Control Cost Estimates
Introduction
     In response to comments received during the public comment period for the
proposed Coke Oven By-product Plant 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 methodology.  This report
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 bases of data supplied by member companies.  The information
supplied ranged from cost factors for individual  components of the control
systems to a cost estimate for the proposed controls applied to an entire
plant.  Because the differences between the cost information received and the
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
commenters 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 could then be compared
with those used by EPA and those 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,
Pennsylvania 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.
                                     B-l

-------
     CRS Sirrine 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 they have
not done much work with steel plants and particularly 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).  EPA,, through RTI,
supplied copies of the proposed regulation, the background information
document, 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 process 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
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, and 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, and United
Engineers (engineering firm supplying the cost estimate contained in AISI's
comments) were present at the Bethlehem plant during this effort.  CRS Sirrine
developed the required cost estimates and gave the results to EPA and RTI in
the form of a report entitled "Benzene Emissions Control  Estimate".  (Docket
Item	).  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 dewaterinrg
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.
                                     B-2

-------
     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.
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 hardware developed for
CRS Sirrine estimate were higher than those used in the BID for the proposed
standards, and 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 Sirrine 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 presently
covered.  There was no provision for this cost element in the 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 BID estimates included pipe supports only in the maximum cost
cases.  Pipe support costs were also 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 were also added for both gas
blanketing and wash-oil scrubber cases.
     A review of the United Engineer's 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 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 estimates.
                                     B-3

-------
     The United Engineer's 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.  (THIS DISCUSSION WILL BE
COMPLETED AFTER RESOLUTION OF THE FOOTNOTE ON WASH-OIL FINAL COOLER COSTS)

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 pipe 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 Si mine in their study of the Bethlehem plant
generally used smaller gas blanket pipe diameters than either the 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 increasing it.  EPA ultimately was influenced more by the fact
that the original  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.
                                     8-4

-------
Other Cost Element Revisions
     Armco and Bethlehem Steel submitted cost estimates for wash-oil scrubbers
applied to specific sources in some of their plants.  Comparison of their data
with the BID data for wash-oil scrubbers suggested a more appropriate way to
estimate costs for scrubber applications.  The BID estimates used scrubber
shell area as the basis for scrubber capital costs, with shell area estimated
roughly from the expected gas flows to be treated.  The industry cost
estimates were based more on the number of sources to be treated rather than
size of the sources.  Given the passive nature of the scrubber operations and
other reasons, we agree that the number of sources is a more appropriate basis
for the estimates.  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-size 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 drawings
indicated that the number of light oil, BTX, and benzene storage tanks for
plants of specific size were generally higher than was assumed in the model
plants used to develop the 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.

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 were also 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 unit cost factors.  This procedure
resulted in a set of minimum and maximum capital  cost estimates for each group
                                     B-5

-------
                   Table B-l.  REVISED CAPITAL COST  FACTORS
   Pipe + Fittings
    Diameter-in
Cost/unit*
Pipe + Fitting, Steam traced,
           & Insulated
            Cost/unit*
8
6
4
3
2
1
$100/ft
$65/ft
$50/ft (70)
$40/ft (50)
$22/ft (30)
$15/ft (20)
$145/ft
$100/ft (153)
$83/ft (130)
$72/ft (109)
$54/ft
$46/ ft
Valves (3-way lubricated plug valves)

   Diameter-in           Cost/unit

        8                  $3,000
        6                  $2,500
        4                  $1,000
        3                  $  700
        2                  $  500
        1                  $  200

Plug Valves

        8                  $1,600
        6                  $  900

Pipe Supports

   Minimum case - $7/ft.
   Maximum case - $30/ft.

Tar decanter clean, cover, and seal

   Minimum case - $5/ft^
   Maximum case - $30.5/ft2

Tar sumps clean, cover, and seal

   Minimum case - $10.5/ft2
   Maximum case - $44.5/ft2

Hot tap

   Eight inch -  $3,800
   Twelve inch - $7,600
                                     B-6
                                                                  (Continued)

-------
             Table B-l.  REVISED CAPITAL COST FACTORS  (Continued)


Tank sealing

   Flushing liquor circulation - $l,400/tank
   Tar tanks - $l,400/tank

Tank roof (including tank clean out)

   Tar dewatering tank - $46.5/ft2
   Excess ammonia-liquor tank - $46.5/ft2

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

   Six inch - $2,000/arrestor
   Four inch - $l,000/arrestor

Pressure/vacuum relief valves

   Six inch - $1,300
   Four inch - $800
   Three inch - $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/ft2
Note:  Data current as of November  1984.
  Parenthetical numbers refer to light oil plant cost  considering  restricted
  construction conditions, e.g. no  welding.


                                     B-7

-------
                 Table B-2.  REVISED ANNUALIZED  COST  ITEMS'
                 Item
   Cost
 Benzene credit, as fuel
 Benzene credit, recovered
 Light-oil credit
 Capital recovery (20 years @ 6.2%)
 Electricity
 Steam
 Cooling water
 Operating labor (including plant
    overhead @ 80%)
 $0.14/kg
 $0.39/kg
 $0.27/kg
 8.86% of capital
 $0.05/kWh
$18.30/Mg
 $0.06/m3
$39.69/h
Data current as of November 1984.
                                    B-8

-------
of emission sources within each model  plant  size.   These  capital  cost
estimates are shown in Tables  B-3 through  B-17.
     Average capital cost estimates  were computed  for  each  group of emission
sources and each model plant size by  averaging the minimum  and maximum cost
estimates.  To apply these cost estimates  to each  of the  coke by-product
plants in the industry, equations were developed that  estimated the capital
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 were 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 were generated by the same
procedure described for capital cost  estimates.  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-17.
                                      B-9

-------
w
 I
                          Table B-3.   COSTS FOR GAS BLANKETING OF TAR DECANTER, TAR-INTERCEPTING SUMP,
                                              AND FLUSHING-LIQUOR CIRCULATION TANK
                                                   (All  Costs in 1984 Dollars)
Cost element
Pressure taps
20-cm (8-in.)

7.6-cm (3-in.

Pipe supports

Valvesb
20-cm (8-in.)
Clean, cover,
m2
(ft2)
Clean, cover,
m2
(ft2)
Seal flushing
Capital costc
Total capital


pipe, m
(ft)
) pipe, m
(ft)
, m
(ft)

plug valve
seal decanter


seal sump,


liquor tanks

Model
Minimum
1
61
(200)
46
(150)
107
(350)
4
1
,
149
(1,600)

3.0
(32)
1
72,600
costd 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
Minimum
3,800
476
(145)
236
(72)
23
(7)
3,800
1,600

53.
(5)

713
(105)
1,400


per unit3
Maximum





98.4
(30)



8 328
30.5

479
(44.5)



                                                                                                         (Continued)

-------
                                                     Table  B-3.   (Continued)
CO
 I
Cost element
Annuali zed cost


Model
Minimum

Maintenance, overhead (9%)e
Utilities'
Taxes, insurance
Capital recovery
Total annuali zed

(4%)
(8.86%)9
cost

9
1
4
9
24

,210
,970
,090
,070
,300
Plant
1
Maximum

21,
3,
9,
21,
57,

900
940
750
600
200
Model
Minimum

13,100
2,740
5,810
12,900
34,500
Plant 2
Maximum

36,300
7,010
16,100
35,700
95,100
Model
Minimum

22,400
4,710
9,960
22,000
59,100
Plant
3
Maximum

61,
10,
27,
60,
161,

900
900
500
900
200
Cost per unit3
Minimum Maximum






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 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.
d Total capital cost includes construction fee, contingency, engineering,  and startup (41 percent).
e Maintenance and overhead are 5 and 4 percent of total capital  cost,  respectively.
f Steam at 18.3/Mg.
9 Capital recovery factor for 20-year lifetime at 6.2 percent.

-------
                           Table 13-4.  COSTS FOR WASH-OIL  VENT  SCRUBBER  FOR  TAR  DECANTER,  TAR-INTERCEPTING SUMP,
                                                   AND FLUSHING-LIQUOR CIRCULATION  TANK
                                                         (All Costs  in  1984 Dollars)
DO
 I
M
NJ
Cost element
Scrubber vessels
15.2-cm (6-in.) vent pipeb,

7.6-cm (3-in.) vent pipe to
sumpc, m
(ft)
2.5-cm (1-in.) wash oil
supply, in
(ft)
5.1-cm (2-in.) wash oil
draind, in
(ft)
Valves6
Seal flushing-liquor tanks
Clean, cover, and seal tar
decanter, m2
(ft2)
Clean, cover, and seal tar
sumps, m2
(ft2)
Pump
Instrumentation1"
Capital costQ
Total capital costh
Model
Minimum
4
m 46
(ft) (150)

46
(150)

61
(200)

61
(200)
4
1

149
(1,600)

2.
(32)
1
1
68,100
96,100
Plant 1
Maximum
4
46
(150)

91
(300)

152
(500)

152
(500)
4
1

149
(1,600)

97 2.97
(32)
1
1
146,900
207,100
Model
Minimum
6
76
(250)

46
(150)

91
(300)

91
(300)
6
2

223
(2,400)

23.2
(250)
1
1
100,500
141,700
Plant 2
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
Model
Minimum
10
122
(400)

91
(300)

122
(400)

122
(400)
10
3

446
(4,800)

46.
(500)
2
2
170,600
240,500
Plant 3
Maximum
10
122
(400)

183
(600)

762
(2.500)

762
(2,500)
10
3

446
(4,800)

5 46.5
(500)
2
2
451.700
636,800
Cost per unit3
Minimum
1,000.
351
(107)

259
(79)

49.
(15)

95.
(29)
3,800
1,400

53.
(5)

113
(10.
3,900
2,500


Maximum
2,000






2


1




8 328.3
(30.5)

479
5) (44.5)




                                                                                                                    (Continued)

-------
                                                           Table B-4.  (Continued)
                                                                  Model  Plant 2
                                                                         Model  Plant  3
           Cost element
  Model Plant 1      	    	
Minimum   Maximum    Minimum   Maximum    Minimum   Maximum
  Cost per unita
Minimum   Maximum
DO
M
U>
           Annual ized  costs
Maintenance, overhead (9%)i
UtilitiesJ
Taxes, insurance (4%)
Operating labor''
Capital recovery (8. 86%)!
Total annual ized cost
8,650
1,360
3,840
7,240
8,510
29,600
18,600
1,790
8,280
7,240
18,300
54,300
12,800
2,110
5,670
7,240
12,600
40,300
34,300
2,970
15,300
7,240
33,800
93,600
21,600
3,700
9,620
14,500
21,300
70,800
57,300
4,040
25,500
14,500
56,400
157,700
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, steam traced & $328/m or $100/ft  and pipe  supports  @  $23/m or $7/ft.
Pipe @ $236/m or $72/ft and pipe  supports &  $23/m or  $7/ft.
Pipe @ $72.20/m or $22/ft and pipe supports  @  $23/m or  $7/ft.
3-way valves, 15 cm (6-in.) - $2,500  and pressure/vacuum  release valves, 15 cm (6 in.)  - $1,300.
Includes flowmeter with alarm, pressure gauge, and  temperature gauge.
Capital cost includes subcontractor overhead and profit and  contractor or material  markup.
Total capital cost includes construction fee,  contingency,  engineering,  and startup (41 percent).
Maintenance and overhead are 5 and 4  percent of  total capital  cost, respectively.
Electricity at $0.05/kWh.
For  30 min/day/scrubber system at  $39.69/hr.
Capital recovery factor for 20-year lifetime at  6.2 percent.

-------
                          Table B-5.
COSTS FOR GAS BLANKETING AMMONIA LIQUOR STORAGE TANKS
       (All  Costs in 1984 Dollars)
Cost element
15.2-cm (6-in.) vent pipe, ra
(ft)
Valvesb
15.2-cm (6-in.) plug valve
Pipe supports, ra
(ft)
Seal tanks
Tank roofs, m?
(ft2)
Capital costc
Total Capital costd
Annual ized costs
Maintenance, overhead (9%)e
Utilitiesf
Taxes, insurance (4%)
Capital recovery (8.86%)9
Total annualized 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 Plant 3
Minimum
91
(300)
6
1
91
(300)
6
0
(0)
73,800
104,100

9,370
1,730
4,160
9,220
24,500
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 unit3
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 cnst includes subcontractor overhead and profit and contractor material markup.
d Total capital cost includes construction fee, contingency, engineering, and  startup (41 percent).
e Maintenance and overhead are 5 and 4 percent of total capital cost,  respectively.
f Steam at $18.3/Mg.
9 Capital recovery factor for a 20-year lifetime at 6.2 percent.

-------
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
'drain0, m
(ft)
Valvesd
Pumps
Instrumentation6
Seal tanks
Tank roofs, m^
ro 
-------
Table B-6.  (Continued)










w
1
en



Model Plant 1 Model Plant 2 Model Plant 3
Cost element Minimum Maximum Minimum Maximum Minimum Maximum
Annual i zed costs
Maintenance, overhead (9%)h 1,860 8,340 4,980 12,700 9,640 26,000
Utilities1 92 92 451 451 902 902
Taxes, insurance (4%) 825 3,710 2,210 5,680 4,280 11,500
Operating laborJ 7,240 7,240 7,240 7,240 7,240 7,240
Capital recovery (8.86%)k 1,830 8,210 4,900 12,600 9,490 25,600
Total annualized cost 11,800 27,600 19,800 38,700 31,600 71,200
a Where a range of unit costs was used, the high and low values are shown as maximum and minimum;
value was used, it is shown in the minimum column.
b Pipe @ $236/m or $72/ft and pipe supports @ $23/m or $7/ft.
c Pipe @ $72.1/m or $22/ft and pipe supports @ $23/m or $7/ft.
d 3-way valves, 7.6 cm (3 in.)-$700; and pressure/vacuum relief valves, 7.6 cm (3 in.)-$660.
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 percent)
n Maintenance and overhead are 5 and 4 percent of total capital cost, respectively.
1 Electricity at $0.05/kWh.
J For 30 min/day at $39.69/hr.
k Capital recovery factor for 20-year lifetime at 6.2 percent.
Cost per unit3
Minimum Maximum





where only a single











-------
                  Table B-7.  COSTS FOR GAS BLANKETING OF LIGHT-OIL CONDENSER, LIGHT-OIL DECANTER,
                                       WASH-OIL DECANTER, AND CIRCULATION TAN*
                                             (All Costs in 1984 Dollars)









Cost element
Pressure tap
10- to 15-cm (4-
pipe*5, m
(ft)
Plug valve, 15 cm
Valvesc
Seal vessels
Flame arrestors
Capital cost^

to 6-in.)


(6 in.)




Total capital cost6

CO
1
I-1
-j



Annuali zed costs

Maintenance, overhead (9%)^
Utilities9
Taxes, insurance
Capital recovery
Total annual ized

(4%)
(8.86%)"
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







98
138

12
2
5
12
32
1
122
(400)
1
8
8
8
,000
,200

,400
,100
,500
,200
,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







149
210

18
3
8
18
49
1
183
(600)
1
13
13
13
,000
,000

,900
,200
,400
,600
,100
Plant 3 Cost per unit3
Maximum Minimum Maximum


(1




207
293

26
5
11
26
69
1 3,800
305 483.1
,000) (147.25)
1 900
13 1,800
13 1,500
13 1,000
,900
,100

,400
.300
,700
,000
,300
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 Assumes 75 percent of pipe is 15-cm  (6-in.) header and 25 percent is  10-cm (3-in.)  vent  lines.
c 3-way valves, 10.2 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  percent).
f Maintenance and overhead are 5 and 4 percent of total capital cost,  respectively.
9 Steam at $18.3/Mg.
n Capital recovery factor for 20-year  lifetime at 6.2  percent.

-------
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)
Model Plant 1
Cost element

Scrubber vessels
10.2-cm (4-in

2.5-cm (1-in.
supply pipe

5.1-cm (2-in.
drain pipec

Valvesd
Seal Vessels
Pump
.) vent pipeb,

) wash-oil
, m
(ft)
) wash-oil
, m
(ft)



Instrumentation6
dj Capital cost^
,_, Total capital
co

cost9

Minimum
6
m 110
(ft) (360)

30.5
(100)

30.5
(100)
6
6
0
1
83,300
117,500

Maximum
6
110
(360)

122
(400)

122
(400)
6
6
2
1
114,200
161,100

Model Plant 2
Minimum
8
146
(480)

91.4
(300)

91.4
(300)
8
8
0
1
119,800
168,900

Maximum
8
146
(480)

244
(800)

244
(800)
8
8
2
1
164,100
231,300

Model
Minimum
13
238
(780)

122
(400)

122
(400)
13
13
0
2
190,600
268,700

Plant 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



(Continued)

-------
                                               Table B-8.   (Continued)
Model Plant 1
Cost element
Annual ized costs
Minimum

Maintenance, overhead (9%)n
Utilities
Taxes, insurance
Operating labor1
Capital recovery
Total annualized

(4%)

(8.86%)J
cost

10
1
7
7
10
34

,600
,400
,700
,240
,400
,300
Maximum

14
1
6
7
14
43

,500
,400
,440
,240
,300
,900
Model Plant 2
Minimum

15
1
6
7
15
46

,200
,920
,750
,240
,000
,100
Maximum

20,800
1,920
9,250
7,240
20,500
59,700
Model Plant 3 Cost per unit3
Minimum

24
3
10
14
23
76

,200
,170
,700
,500
,800
,400
Maximum Minimum Maximum

32,200
3,170
14,300
14,500
31,700
95,800
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, steam traced P $426.5/m or  $130/ft and  pipe  supports  @ $23/m or  $7/ft.
c Pipe P $98.4/m or $30/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.
' Capital cost includes subcontractor overhead  and profit and contractor material  markup.
9 Total capital cost includes construction fee,  contingency,  engineering,  and startup (41 percent).
h Maintenance and overhead are 5 and 4 percent  of total  capital  cost,  respectively.
1 For  30 min/day/scrubber system at $39.69/hr.
J Capital recovery factor for 20-year lifetime  at 6.2 percent.

-------
                           Table B-9.  COSTS FOR GAS BLANKETING OF LIGHT-OIL AND BTX STORAGE  TANKS
                                                 (All Costs in 1984 Dollars)
td
 I
Cost element
10- to 15-cm (4- to 6-in.)
pipeb, m
(ft)
Pipe supports, m
(ft)
Seal tanks
Flame arrestors
Capital costd
Total capital cost6
Annual i zed costs
Maintenance, overhead (9%)^
Utilities9
Taxes, insurance (4%)
Capital recovery (8.86%)n
Total annuali zed cost
Model
Minimum
49
(160)
4
49
(160)
4
4
41,900
59,100

5,320
850
2,360
5,230
13,800
Plant 1
Maximum







123
174

15
3
6
15
41
183
(600)
4
183
(600)
4
4
,600
,200

,700
,170
,970
,400
,300
Model
Minimum
61
(200)
9
61
(200)
9
9
69,600
98,100

8,830
1,060
3,920
8,690
22,500
Plant
2
Maximum







189
267

24
4
10
23
62
259
(850)
9
259
(850)
9
9
,400
,000

,000
,490
,700
,700
,900
Model
Minimum







126
177

16
2
7
15
41
122
(400)
15
122
(400)
15
15
,200
,900

,000
,110
,120
,800
,000
Plant
3
Maximum

(1


(1


259
365

32
5
14
32
85
335
,100)
15
335
,100)
15
15
,500
,900

,900
,810
,600
,400
,800
Cost per unit3
Minimum Maximum
483.1
(147.25)
1,800
23 98.4
(7) (30)
1,500
1,000








    a Where a range of unit costs was used, the high and low  values  are  shown  as  maximum and minimum; where ony a single
      value was used, it is shown in the minimum column.
    b Assumes 75 percent of pipe is 15-cm  (6-in.) header and  25 percent  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 percent).
    f Maintenance and overhead are 5 and 4 percent of total capital  cost,  respectively.
    9 Steam at $18.3/Mg.
    n Capital recovery factor for 20-year  lifetime at 6.2  percent.

-------
Table B-10.
COSTS OF WASH-OIL VENT SCRUBBER FOR LIGHT-OIL AND BTX  STORAGE  TANKS
              (All Costs in 1984 Dollars)
Model Plant 1
Cost element
Scrubber vessels
10-cm (4-in.) vent pipe*3, m
(ft)
2.5-cm (1-in.) wash-oil
line, m
(ft)
5.1-cm (2-in.) wash-oil
drain0, m
(ft)
Pumps
Valvesd
Vessel sealing
Instrumentation6
Capital costf
Total capital cost9
Minimum
4
61
(200)

30.5
(100)

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

183
(600)

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

30.5
(100)

30.5
(100)
0
9
9
1
108,600
153,100
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 unita
Minimum Maximum
1,000 2,000
449.5 .
(137)


(20)


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


                                                                                    (Continued)

-------
                                                     Table  B-10.   (Continued)
da
 I
to
Model Plant 1
Cost element
Annual ized costs
Minimum

Maintenance, overhead (9%)n
Utilities1
Taxes, insurance
Operating laborJ
Capital recovery
Total annual ized

(4%)

(8.86%)k
cost

6

2
7
6
24

,700
794
,980
,240
,600
,300
Maximum

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

13,800
1,800
6,120
7,240
13,600
42,500
2
Maximum

20
1
9
7
19
58

,200
,800
,000
,240
,900
,200
Model Plant
Minimum

23,
2,
10,
14,
22,
74,

300
980
400
500
900
100
3 Cost per unit3
Maximum Minimum Maximum

31
2
14
14
31
94

,500
,980
,000
,500
,000
,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.
b Pipe @ $426.5/m or $130/ft and pipe supports & $23/m or  $7/ft.
c Pipe @ $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 percent).
n Maintenance and overhead are 5 and 4 percent 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-year lifetime at 6.2 percent.

-------
               Table B-ll.  COSTS FOR GAS BLANKETING OF TAR  COLLECTING,  STORAGE,  AND  DEWATERING TANKS
                                              (All Costs in  1984  Dollars)
                                                      Model  Plant  2
                                                                            Model  Plant  3
Cost element
  Model Plant 1
Minimum   Maximum    Minimum   Maximum    Minimum   Maximum
  Cost per unit3
Minimum   Maximum




dd
I
M
U>

15-cm (6-in.) pipe, m
(ft)
Seal tanks
Tank roofs-dewatering, m^
(ft2)
Pipe supports, m
(ft)
Valvesb
Capital costc
Total capital costd
Annualized costs
Maintenance, overhead (9%)e
Utilitiesf
Taxes, insurance (4%)
Capital recovery (8.86%)9
Total annualized cost
61
(200)
5
0
(0)
61
(200)
5
47,400
66,800
6,020
1,150
2,670
5,920
15,800
152
(500)
5
49
(531)
152
(500)

115,
163.
14,
2,
6,
14,
38,
5
700
100
700
880
530
500
500

84
118
10
1
4
10
27
91
(300)
10
0
(0)
91
(300)
10
.100
,600
,700
,730
,740
,500
,700
(2
(2

414
584
52
14
23
51
142
762
,500)
10
75
(804)
762
,500)
10
,400
,300
,600
,400
,400
,800
,100
122
(400)
16
0
(0)
122
(400)
16
126,000
177,700
16,000
2,310
7,110
15.700
41,100
(3
(1
(3

557
786
70
17
31
69
189
914 328.1
,000) (100)
16 1,400
169 500.5
,816) (46.5)
914 23 98.4
,000) (7) (30)
16 3,800
,600
,300
,800
,300
,500
,700
,200
a Where a range of unit costs was used,  the high  and  low  valves  are  shown  as  maximum and minimum; where only a single
  valve was used, it  is shown in the minimum column.
b From Table B-l, 3-way valves,  15 cm  (6 in.)  -  $2,500  and  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 percent).
e Maintenance and overhead are  5 and 4 percent of total capital  cost,  respectively.
f Steam at $18.3/Mg.
9 Capital recovery factor for 20-year  lifetime at 6.2 percent.

-------
Table B-12.
COSTS OF WASH-OIL VENT SCRUBBER FOR TAR COLLECTING, STORAGE, AND DEWATERING  TANKS
                     (All  Costs in 1984 Dollars)

Model
Cost element Minimum
Scrubber, heat exchanger,
seperator 62
15.2-cm (6-in.) vent
pipe*3, m
(ft)
10.2-cm (4-in.) wastewater
pipe, m
(ft)
2.5-cm (1-in.) wash oil supply
pipe, m
(ft)
5.1-cm (2-in.) wash oil
drain pipe0, m
(ft)
Seal tank
Tank roofs, dewatering, m2
(ft2)
Pump
Valvesd
Valves and level control
Instrumentation6
Capital costf 159
Total capital cost9 225

,700

122
(400)

30.5
(100)

30.5
(100)

30.5
(100)
5
0
(0)
1
5
1
1
,700
,200
Plant 1
Maximum

62,700

122
(400)

61
(200)

152
(500)

152
(500)
5
49.3
(531)
1
5
1
1
210,300
296,500
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 Plant 3
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
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 unit3
Minimum Maximum



351
(107)

272.3
(83)

49.2
(15)

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


                                                                                         (Continued)

-------
                                                        Table B-12.   (Continued)
00
 I
K)
Ul
Model Plant 1
Cost element
Annuali zed costs
Maintenance, overhead (9%)n
Utilities
Taxes, insurance
Operating labor'
Capital recoveryJ
Total annual i zed cost
Minimum

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

26,700
10,300
11,900
7,240
26,300
82,300
Model Plant 2
Minimum

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

56,300
34,100
25,000
7,240
55,400
178,100
Model Plant 3 Cost per unit8
Minimum

64,900
•72,400
28,800
14,500
63,800
244,400
Maximum Minimum Maximum

90,000
72,800
40,000
14,500 i
88,600
305,900
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, steam traced @ $328/m or $100/ft  and pipe supports  @ $23/m  or  $7/ft.
c Pipe P $72.20/fn or $22/ft and pipe supports  @  $23/m or  $7/ft.
d 3-way valves, 15 cm  (6  in.)-$2,500 and  pressure/vacuum  release valves,  15 cm (6 1n.)-$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 percent).
n Maintenance and overhead are 5 and 4  percent of total capital  cost,  respectively.
' For 30 min/day/scrubber  system at $39.69/hr.
J Capital recovery factor  for 20-year lifetime at 6.2 percent.

-------
                                       Table B-13.  COSTS FOR COVERING LIGHT-OIL SUMP
                                                 (All Costs in 1984 Dollars)
Model Plant 1
Cost element
Clean, cover, and seal, n>2
(ft2)
7.6-cm (3-in.) vent pipe, m
(ft)
Capital costb
Total capital costc
Annual i zed costs
Maintenance, overhead (9%)^
Taxes, insurance (4%)
Capital recovery (8.86%)e
Total annual i zed cost
._ a Where a range of unit costs
1 value was used, it is shown
Minimum
3.3
(36)
4.6
(15)
1,700
2,390
215
96
212
523
was used,
in the mir
Maximum




37
52
4
2
4
11
20.9
(225)
4.6
(15)
,500
,900
,760
,120
,690
,600
the high and
limum column.
Model
Minimum
3.3
(36)
4.6
(15)
1,700
2,390
215
96
212
523
Plant 2
Model Plant 3
Maximum Minimum

(1


164
232
20
9
20
50
low values
93
,000)
4.6
(15)
,600
,100
,900
,280
,600
,700
are shown
6.7
(72)
9.1
(30)
3,400
4,790
431
192
424
1,050
Maximum

(2


329
464
41
18
ii
101
as maximum
186
,000)
9.1
(30)
.200
,200
,800
,600
.100
,500
and minimum;
Cost per unit3
Minimum Maximum
328 1,765
(30.5) (164)
131
(40)





where only a single
en  b Capital  cost includes subcontractor overhead and profit and contractor material markup.
    c Total  capital  cost includes construction fee, contingency, engineering, and startup  (41  percent).
    d Maintenance and overhead are 5 and 4 percent of total capital cost, respectively.
    e Capital  recovery factor for 20-year lifetime at 6.2 percent.

-------
                     Table B-14.
COSTS FOR NITROGEN OR NATURAL GAS BLANKETING OF PURE BENZENE  STORAGE TANKS
               (All Costs in 1984 Dollars)
CB
Model Plant 1
Cost element Minimum Maximum
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
Valvesb
Tank sealing
Pipe supports, m
(ft)
Capital costs0
Total capital costsd
Annualized costs
Maintenance, overhead (9%)e
Utilities'"
Taxes, insurance (4%)
Capital recovery (8.86%)9
Total annual i zed cost
Model Plant 2 Model
Minimum Maximum Minimum
30.5
(100)
61
(200)
1
2
0
7
7
7
61
(200)
48,100
67,800

6,110
—
2,710
6,000
14.800
Plant 3
Maximum
91.4
(300)
244
(800)
1
2
1
7
7
7
244
(800)
137,400
190,000

17,100
15,000
7,600
16,800
56,500
Cost per unit3
Minimum Maximum
65.6
(20)
164
(50)
4,400
2,000
30,000
1,000
1,360
1,400
23 98.4
(7) (30)

i






    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, 7.6 cm (3 in.)-$700; and  pressure/vacuum  relief  valves,  7.6 cm  (3  in.)-$660.
    c Capital cost includes subcontractor overhead and profit and contractor material  markup.
    d Total capital cost includes construction fee, contingency, engineering,  and startup (41  percent).
    e Maintenance and overhead are 5 and 4   percent of total  capital  cost,  respectively.
    f Nitrogen at $0.27/m3 (0.76/100 ft3).   Includes  rental of  5.7-m3 (1,500-gal) liquid  nitrogen  storage  tank,
      vaporizer, and gas usage.  Some plants  are assumed  to have a nitrogen  source  and others  must  purchase  nitrogen.
    9 Capital recovery factor for 20-year lifetime at 6.2 percent.

-------
                         Table  B-15.
   COSTS OF WASH-OIL VENT SCRUBBER FOR BENZENE  STORAGE  TANKS
          (All Costs in 1984 Dollars)
Cost element
  Model Plant 1
Minimum   Maximum
  Model Plant 2
Minimum   Maximum
  Model  Plant 3
Minimum   Maximum
 Cost per unit3
Minimum  Maximum
Scrubber vessels
2.5-cm (1-in.) wash-oil line, m
(ft)
5.1-cm (2-in.) wash-oil drainb, m
(ft)
Pump
10-cm (4-in.) vent pipec, m
(ft)
Valvesd
10.2-cm (4-in.) flame arrestors
Tank sealing
Instrumentation6
Capital costf
Total capital costs9
Annual i zed costs
Maintenance, overhead (9%)n
Utilities1
Taxes, insurance (4%)
Operating labor J
Capital recovery (16.3%)k
Total annual i zed cost
7
61
(200)
61
(200)
0
45.7
(150)
7
7
7
1
"61,900
87,200

7,850
43
3,490
7,240
7,730
26,400
7
244
(800)
244
(800)
2
45.7
(150)
7
7
7
1
110,900
156,300

14,100
43
6,250
7,240
13,800
41,400
1,000 2,000
65.6
(20)
121.4
(37)
3,900
252.6
(77)
1,800
1,000
1,400
2,500









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.
b Pipe @ 98.4/m or $30/ft and pipe supports @ $23/m or  $7/ft.
c Pipe @ $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  percent).
h Maintenance and overhead are 5 and 4 percent of capital, respectively.
1 Electricity at $0.05/kWh.
J For 30 min/day at $39.69/h.
k Capital  recovery factor for 20-year lifetime at 6.2 percent.
                                                        B-28

-------
                                            Table B-16.   COSTS FOR TAR BOTTOM FINAL COOLER
                                                      (All  Costs in 1984 Dollars)
co
 I
M
VO
Model Plant 1
Cost element
Tanks-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 ized cost
Maintenance, overhead (9%)c
Utilities
Taxes, insurance (4%)
Capital recovery (8.86%)d
Total annual ized cost
Minimum

61,700
16,300
26,800

31,700
7,070
11,000
154,500
217,800

19,600
2,200
8,710
19,300
49,800
Maximum

61,700
16,300
82,800

31,700
10,800
16,900
220,100
310,400

27,900
5,040
12,400
27,500
72,900
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
45,500
119,800
Maximum

141,700
26,600
247,700

72,800
28,100
43,700
560,500
790,200

71,100
16,000
31,600
70,000
188,800
Model Plant 3
Minimum

230,500
37,800
152,000

118,400
28,400
44,100
611,100
861,700

77,600
16,200
34,500
76,300
204,600
Maximum

230,500
37,800
470,000

118,400
49,800
77,500
984,000
1,387,400

124,960
32,300
55,500
122,900
335,500
         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 percent).

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

         d  Capital  recovery factor for 20-year lifetime at 6.2 percent.

-------
                                          Table B-17.   COSTS FOR INDIRECT/WASH-OIL FINAL COOLER
                                                       (All  Costs in 1984 Dollars)
tt)
 I
U)
o
Model Plant 1
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
Annual i zed costs
Maintenance, overhead (9%)c
Utilities
Taxes, insurance (4%)
Operating labor
Capital recovery (8.86%)d
Total annuali zed cost
Minimum
166,700

11,300

17,500
195,500
246,300

22.200
3,620
9,850
69,000
21,800
126,500
Maximum
708,000

47,800

74,300
830,100
1,170,500.

105,300
22,200
46,800
69,000
103,700
347,100
Model Plant 2
Minimum
340,500

23,000

35,800
399,200
503,000

45,300
11,800
20,100
69,000
44,600
190,800
Maximum
1,627,000

109,800

170,800
1,907,700
2,689,800

242,100
88,800
107,600
69,000
238,300
745,800
Model Plant 3
Minimum
542,400

36,600

57,000
636,000
801,300

72,100
24,200
32,100
69,000
71,000
268,400
Maximum
2,646,000

178,600

277,800
3,102,400
4,374,400

393,700
199,700
175,000
69,000
387,600
1,225,000
          a Capital cost includes subcontractor overhead and profit and contractor material markup.
          b Total capital cost includes construction fee, contingency, engineering, and startup  (41  percent).
          c Maintenance and overhead are 5 and 4 percent of total capital cost, respectively.
          d Capital recovery factor for 20-year lifetime at 6.2 percent.

-------
       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 versio'n of
Chapter 9 of the Background Information Document, 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-2

-------
                 TABLE C-l.   COKE BY-PRODUCT PLANT CONTROL OPTIONS'
                                              Control option
     Emission source
Regulatory Alternative
         II
Regulatory Alternative
        III
Direct water final cooler

Tar bottom final cooler

Tar decanter, flusing-
 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
 These regulatory alternative control  options differ from the proposed
 regulations.
                                     C-3

-------
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-4

-------
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 are also included in SIC 3312.x
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 $3,220,011,000,3
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 which 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  were  603,288 Mg for  all  of
1983.16
                                  C-5

-------
                                                 1ABLE C-2.  COKE INDUSTRY FOREIGN TRADE
                                                                                          9 10 II  12 1.1 14 15
O
 I
tn
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
Total U.S. imports
(10° $)"
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
(106 $)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. 9b
410. 9^
340. lb
52'°b e

9.2
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
(10° $>a
1U.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
(10C $)*
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
66.7.
71.9b'c.
68.9b'd
15. Ob
12.9°
11.3°
13. 8b
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
                   Current dollars.
                  bSee Product SIC (331210) in References 11-13.
                   Defined as "Pitch coke. Coke of coal, lignite, or peat."
                   Defined 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 has always 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 twenty-three 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-7

-------
              TABLE C-3.   COKE PRODUCTION IN THE WORLD6 17 18
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° .
World production3
(106 Mg)
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
U.S. production
(106 Mg)
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
U.S. production
as a share of
world production
(%)
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-8

-------
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-9

-------
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-10

-------
foundry coke is used as a fuel in the cupolas of foundries.  Coke  is also
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 are usually 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-ll

-------
o
I
          Q
          ui
          (/>
          D

          UJ
          *
          O
          O
          2
          ui
          O
          DC
          UJ
          a.
94



92



90



88



86



04
                                  A
    ,--•
_  /
   /
  /FURNACE
             12
10
   - \ OTHER USES
                         \"\
                                                                                        A
                                                                                       /  FURNACE
                                                                                       FOUNDRY
                                                                                           USES
                                                                            I
                                                                   I
                                                                                      I
                50   52   04   56  58   60   62   64   66   68   70   72   74


                                                  YEAR
                                                                  76   70   00
                          Figure C-1. Uses of oven coke as percents of total coke consumption.6-26-27

-------
number of ovens which 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 capac-
ity totalled 51,180,000 Mg5.  Table C-6 presents the data for November,
1984.
     In actuality, ovens which are removed from service and placed on "hot
idle" status are those likely to be returned to production in the short
term.  Ovens which 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 demolition.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 megagrams
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-13

-------
          TABLE  C-6.   MAXIMUM ANNUAL  CAPACITY OF OVEN COKE  PLANTS
                  IN  THE  UNITED  STATES  IN  NOVEMBER, 19845

In existence
Furnace plants
Foundry plants
Total
Out of service
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-14

-------
     TABLE C-7.   TYPICAL COST BREAKDOWNS:   FURNACE COKE PRODUCTION AND
                    HOT METAL (BLAST FURNACE) PRODUCTION32

Furnace coke production

     Labor and materials                                      Percent of cost

     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                                          Percent of cost

     Charge metal lies                                               42.5
          Iron ore                                                  (6.3)
          Agglomerates                                             (33.3)
          Scrap                                                     (2.9)
     Fuel inputs                                                     44.8
          Coke                                                     (41.8)
          Fuel oil                                                  (3.0)
     Limestone fluxes                                                0.7
     Direct labor                                                    7.6
     Maintenance              •       '                                 1.5
     General expenses                                                2.9

     Total labor  and material costs                                100.0
                                 C-15

-------
           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.



aFigures for 1971-1982 are estimates.   See text for more detail
                                 C-16

-------
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 megagrams 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 megagrams consumed annually.
Domestic consumption of coke increased during the mid-I960's to mid-1970's
to an annual figure of 57 million megagrams 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 in 1980 and 1985.
                                  C-17

-------
o
I
co
o
01
s
2
O
               70 ,-
               65
            O
            O
            u.
            O  60
            CD
               50
               45
               35
                    50   52   54    56   58    60   62
                                              64   66
                                             YEAR
68   70   72   74    76   78   80
                                       Figure C-2. U.S. apparent consumption of coke.6-26

-------
TABLE C-9.  COKE RATE3 18 25 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-19

-------
                            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
                                 C-20

-------
     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.  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 EPA regula-
tion 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 megagrams 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   *
         *Two quarters of 1984

     Production is less than apparent consumption in 1981,  1983,  and 1984.
For each of these years, stocks and imports more than accomodate  the short-
fall.   Coke producers were operating at 80 percent of total  capacity in
November, 1984.5  Thus, it is unlikely that major shortages  will  develop in
the near future.
                                  C-21
1980
1981
1982
1983
1984*
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

-------
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.   While concen-
tration ratios are a useful tool for describing industry structure, concen-
tration 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.
     Reported capacities in Table C-10 are maximum, nominal  figures,  which
do not include any allowance for outage like that determined for the  overall
                                  C-22

-------
TABLE C-10.   COKE PLANTS IN THE UNITED STATES,  November 1984s 44
Company name
Armco, Inc.
Bethlehem Steel Corp.
Rouge Steel
Inland Steel Co.
o
iv) Interlake, Inc.
GJ '
The LTV Steel Corp.
Lone Star Steel Co.d
National Steel Corp.6
Weirton Steel Corp.
New Boston Coke Corp.
Plant location
Ashland, KY
Middletown, OH (2)
Bethlehem, PA
Burns Harbor, IN
Lackawanna, NY
Sparrows Point, MO
Dearborn, MI
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









(continued)

-------
TABLE C-10  (continued)
Company name
U.S. Steel Corp.
Wheel i ng-Pi ttsburgh9
Steel Corp.
Jim Walter Corp.
0
i\i 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
Lorain, 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 coke
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)
5,294
1,822
916
4,228
1,496
1,160
1,509
490
499
207
157
563
521
583
402
330
130
477
                                                           (continued)

-------
                                            TABLE C-10   (continued)
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 coke3
Foundry
Foundry
Foundry,
other industrial
Foundry
Coke capacity
(103 Mg/yr)
617
161
132
299
 An end use is considered a major  use  if  it  is  at  least 20 percent of  the  plant's  total  distribution of coke.

 Numbers in parentheses indicate the number  of  plants  at  that  location.  If  no  number is indicated,  only one
 plant exists at that location.

 LTV 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.

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

 McLouth Steel Corp., the parent company  of  New Boston Coke Corp., is  operating under Chapter 11 filed in 1981.

9Wheel ing-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.

-------
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 fo.undr.ies. and in other industries was produced by merchant plants.
If coke plant sites were ranked according to capacity, the top five plant
sites and top ten 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 four (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)47;  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-26

-------
                                 TABLE C-ll.   INTERREGIONAL COKE SHIPMENTS IN 197745
                                                 (103 megagrams)
o
Consuming region
Producing region
Alabama 2
Cal ifornia,
Colorado, Utah
Maryland, New York
11 linois
Indiana
Kentucky, Missouri,
Tennessee, Texas
Michigan
Minnestoa, Wisconsin
Ohio
Pennsylvania
Virgina,
West Virginia
AL
,228

0
0
0
0

14
0
0
0
9

0
CA, UT
CO
27

2,668
3 4
0
5

18
0
6
4
0

0
MD,
NY
10

0
,392
0
0

0
0
0
0
51

0
IL
81

0
123
1,424
69

15
0
269
138
1,241

0
IN
112

0
0
0
7,594

5
7
70
366
134

0
KY, MO,
TN, TX
465

0
22
0
35

928
1
1
379
3

8
MI
195

0
88
0
97

125
2,639
61
260
52

412
MN, WI
7

0
0
0
11

0
0
158
0
0

0
OH
114

0
6
0
62

13
6
5
6,356
1,370

0
PN
1

0
8
0
3

0
0
1
2
10,257

214
VA,
WV
51

0
0
0
0

20
0
0
12
3

2,465
Total
3,361

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

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

3,099
  TOTAL
2,251   2,731  4,453  3,360  8,288     1,842    3,929    176   7,932  10,556   2,551  48,069

-------
TABLE C-12.   PERCENT OF COKE CAPACITY OWNED BY TOP FIRMS
                  (NOVEMBER, 1984)5
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-28

-------
     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 is probably 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-29

-------
     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 mini-mills has eased entry into the
iron and steel industry, which in turn reduces market power.
     Imported coke can also 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 due to 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.   Analagous 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-30

-------
the 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 conclu-
sive in identifying the presence of market power because vertical  integration
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 is 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-31

-------
      TABLE  C-13.   COMPARISON  OF  COAL PRICES AND DOMESTIC AND IMPORTED
                          COKE PRICES6 50  51 52  53
       Average  value  of
       coal  carbonized.
       in  coke  ovens  '
            ($/Mg)
Average value of  Average value of
                   oven coke sold   Average value of
                    commercially     imported coke  '
oven coke used
 by producers
      ($/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).

"General  customs value  as  reported  by the  Department  of Commerce (current
 do!lars).
                                 C-32

-------
TABLE C-14.   FINANCIAL INFORMATION ON COKE-PRODUCING FIRMS, 1983
               (million 1983  dollars)3 54 ss 5S 37
Company name
Armco, Inc.
Bethlehem Steel Corp.
Ford Motor Co.
(Rouge Steel)
Inland Steel Co.
Interlake, Inc.
The LTV Steel Corp.
McLouth Steel Corp.1lJ
(New Boston Coke Corp. )
National Intergroup, Inc.
(National Steel Corp.)
Northwest Industries111
(Lone Star Steel Co.)
Shenango Furnace Co. , Inc. '"
(Shenango, Inc.)
U.S. Steel Corp.
Weirton Steel Corp.13
Wheeling-Pittsburgh Steel Corp.q
Jim Walter Corp."
Koppers Co. , Inc. n
Alabama Byproducts Corp.
Carondelet Coke Corp.r
Chattanooga Coke and Chemicals
Co., Inc.
Citizens Gas and Coke Utility
Detroit Coke Corp. r>t
Indiana Gas and Chemical Corp.
McWane, Inc. (Empire Coke Co.)
Tonawanda Coke Corp. '
Net sales
4,165
4,398
44,455
3,046
835
4,578
11
2,993
1,603
145
16,369
1,000
772
2,025
1,566
229
k
17
316
*
54
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
:<
Cash flowd
70
1299
5,542
106s
67 '
(164)9
k
161
1549
k
1,5639
k
(70)9
159
1749
192
k
k
91
k
!<
k
k
                                                         (continued)
                              C-33

-------
.  TABLE C-14  (continued)
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 >J
(New Boston Coke Corp. )
National Intergroup, Inc.
(National Steel Corp. )
Northwest Industries
(Lone Star Steel Co. )
Shenango Furnace Co., Inc.n'°
(Shenango, Inc. )
U.S. Steel Corp.
Weirton Steel Corp.P
Wheeling-Pittsburgh Steel Corp.q.
Jim Walter Corp. n
Koppers Co. , Inc.n
Alabama Byproducts Corp.
Carondelet Coke Corp. r
Chattanooga Coke and Chemicals
Co., Inc.
Citizens Gas and Coke Utility
Detroit Coke Corp.r>t
Indiana Gas and Chemical Corp.
McWane, Inc. (Empire Coke Co.)
Tonawanda Coke Corp.r
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-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>J
(New Boston Coke Corp. )
National Intergroup, Inc.
(National Steel Corp.)
Northwest Industries
(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.n
Alabama Byproducts Corp.
Carondelet Coke Corp.1"
Chattanooga Coke and Chemicals
Co., Inc.
Citizens Gas and Coke Utility
Detroit Coke Corp.1"'1
Indiana Gas and Chemical Corp.
McWane, Inc. (Empire Coke Co.)
Tonawanda Coke Corp. r
Annual
interest
expense
• 154
104
567
68
12
171
k
62
63
k
1,074
"
53
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
2,509
1,175
243
k
k
343
28
36
112
!<
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
IS
0
21
:<
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
(continued)
         C-35

-------
                          TABLE C-14 (continued)
aVa1ues 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.
CEBIT = earnings before interest and taxes.
 Cash flow = operating income + depreciation - interest expenses - taxes.
a
 Net working capital = current assets - current liabilities.
 Tangble. net worth = equity - intangible assets.
^Received 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.
^Figures are interim values reported for first eleven months of 1984.
 Converted to 1983 dollars using GNP implicit price deflator.
\,
 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, two
 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.
^Employees formally took control in January, 1984.   All figures are interim
 values reported for first three months of 1984.   Conversion to 1983
 dollars using GNP implicit price deflator.
Wheeling-Pittsburgh Steel Corp. filed for Chapter 11 in April  1985.
rOwned by James  D.  Crane.   Financial information denied.
sChattanooga 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 inplicit price deflator.
                                 C-36

-------
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 can also 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 due to  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 due to 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-37

-------
TABLE C-15.   FINANCIAL RATIOS FOR COKE-PRODUCING FIRMS
Company name3 Liquidity ratio Coverage 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.)^
U.S. Steel Corp.
Weirton Steel Corp.
Wheeling-Pittsburgh
Steel Corp.
Jim Walter Corp.^
Koppers Co. , Inc. ^
Alabama Byproducts Corp.
Citizens Gas and Coke
Utility
Detroit Coke Corp.
Indiana Gas and
Chemical Corp.
1.56 -3.42s
1.27 -2.30s
1.05 3.82
1.42 -0.25s
2.16 3.08
1.41 -1.47s
0.17 f
1.40 -2.85s
1.80 -1.66s
1.59 f
1.22 ' 1.12
1.79 f
1.42 -1.25s
1.09 0.81
2.15 1.59
3.17 f
1.32 f
1.08 f
1.18 f
Leverage ratio
1.52
1.95
1.73
1.20
0.93
2.91
-3.75s
1.40
1.65
0.51
2.34
f
3.06
3.64
0.86
0.35
1.52
-10.3s
0.50
                                                  (continued)
                    C-38

-------
                           TABLE C-15 (continued)
Company name          Liquidity ratio     Coverage ratio    Leverage ratio

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

 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.   No ratios were calculated for
 Carondelet Coke Corp.,  Chattanooga Coke and Chemicals Co., Inc. and
 Tonawanda Coke Corp.  due to lack of information.

"Liquidity ratio =   Current assets
   M     3         Current liabilities

          ratl'° =
                  Annual  interest expense
eNegative values are not meaningful.

 Information not available.

Produces both furnace and foundry coke.
                                 C-39

-------
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 one, 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 substan-
tial 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 1970s.   The trend is
the 1980s shows 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-40

-------
     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
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, description of steel  industry trends
in technological development and production are useful  indicators of future
coke production and coke capacity requirements.
     As mentioned, there has been a technological  shift toward labor-saving
technology which is expected to continue.   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 percent,
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-41

-------
     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 1970s 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 1970s, the decline in consumption of steel per
dollar of gross national  product 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
megagrams by 1988, with domestic shipments representing a 5 percent average
annual rate increase over the 1983-1988 period.65  Projections by the
Bureau of Mines predict U.S.  raw steel demand will be 138 million megagrams
in 1990, and 164 million megagrams in 2000.71
     Steelmaking capacity utilization has recently 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 due to
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 breakeven point for operations is at  approximately 65 percent of capa-
bility, though 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-42

-------
122.5 million megagrams in 1984, capacity is likely to be trimmed to
109 million megagrams by the late 1980s.73  However, the Bureau of Mines
predicts U.S. production of raw steel will rise to 113 million megagrams in
1990, and to 132 million megagrams 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 non-coke energy inputs could greatly reduce demand
for coke.
     The emergence of mini mills 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 mini mills may result
in dramatic reductions in coke demand if the minis 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-43

-------
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 or not 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-44

-------
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 due to 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 is also 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-45

-------
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 pi ant-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 C-3.   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-46

-------
     COS IS Of COKE
      PRODUCT ION
    COSTS OF EXISTING
      ENVIRONMENTAL-
       REGULATIONS
      COSTS OF
    REGULATORY
    ALTERNATIVES
                           FOREIGN COKE
                              SUPPLY
                                EXISTING PLANT
                                    INVENTORY
                                  NEW PLANT
                                CONFIGURATIONS
                           DOMESTIC
                             COKE
                            SUPPLY
o
I
-Pi
--.I
FOREIGN DEMAND
  FOR US COKE
    FOREIGN DEMAND
     FOR US STEEL "
    DOMESTIC DEMAND
      FOR US.STEEL—'
                                  DEMAND FOR
                                   US STEEL
                                                SYNTHESIS
                                              "  BEHAVIORAL"
                                              ASSUMPTIONS
 IMPACTS
"ON COKE
 AND STEEL
   IMPACTS
   ON FINAL
DCMA.ND 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, can then 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 over 60 variables for fur-
nace and foundry coke plants in existence in 1979 were compiled from govern-
ment publications, industry contacts, and previous studies of the coke
industry.  The data base was sent to the American Iron and Steel Institute,
which submitted it to their members for verification, corrections, and
additions,79 and to the American Coke and Coal Chemicals Institute.   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 construc-
tion, 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, 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 decarboni-
zation 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                 18 hours        30 hours
             Preheated coal           13 hours        24 hours
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 plants 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-48

-------
by-product facility at a plant.  The by-products included in the model are
coke breeze, coke oven gas, tar, crude light oil, BTX, ammonium sulfate,
anhydrous ammonia, elemental sulfur, sodium phenolate, 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 five production and ten
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-49

-------
                                                      cosi ctniin i . mm MKUAI
o
i
en
O



D.t. CA«
OWI«





COAl HOISI 1
| cotimon en
1 1 111

MM J •"
1 l^N HKIKC ""
COAl MlOCt — 111



S
1
1

|

















                                                    j-	.	1 CHUSHtH U.	

                                            1   »--i—.___._______.______»__


                                             =»*    I-..LL	   ~-
I tg^j «"Ji?J! i   i  cosi_cL»U«-l I
-              _,....

11 QUO
SIIIUMO IAIUS
	 1 PHINOl
1 	 ' 1
SOOlUt >H|MOIA
rod IUOIHH rioci
/AltOllA\ AMOIIA (IWjl
\?S

— —
" cost CINHI s



HASH OK
IINIOIIHO
MASH Oil

I UOII Oil | t 	 —
| BIMKIHG | |
1—"
1
                                                                                                                   eiiiiiM


                                                                                                                   SOLVtNl NAPHIHA
                                                                                                                   >mM o« immniDiAii
                                                                                                                  .UfHij>lLJjiU|	
                                                        -*
                                                        | COSI Ct«U» >  j
                                                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 importantly, 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 pipeline-
charged batteries than for conventionally charged batteries.
     Periodically, batteries must undergo major rehabilitation or rebuilding
because of performance deterioration.   The costs  of pad-up rebuilds will
vary from site to site depending on battery maintenance, past operating
practices, and other factors.   Average estimates  of the cost of rebuilding
were developed for this study and are shown in a  report by PEDCo.81  The
economic life of coke-making facilities is subject to considerable variation
depending upon 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 facilities.   However, firms will probably
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-51

-------
          TABLE C-16.   ESTIMATED CAPITAL COSTS OF NEW PLANTS80
                                   Conventionally
                                   charged battery
   Pipeline
charged battery

Capacity (103 Mg/yr)
4-metera
450
6-metera
720
4-meter
560
6-metera
900
Capital costs by element
  (106 1979 dollars)
Coke battery
Quench tower with baffles
Quench car and pushing
emissions control
Pusher machine
Air-conditioned larry car
Door machine and coke guide
By-product plant
Coal -hand! ing system
Coke-handling system
Off sites
Total
34.20
2.45

6.58
2.50
1.72
1.80
32.50
18.20
6.85
1.60
$108.40
48.90
2.85

7.92
3.20
2.28
2.10
39.75
23.60
8.80
1.80
$141. 20
64.60
2.45

6.58
2.40
0
1.80
35.76
20.62
7.74
1.69
$143.74
83.70
2.85

7.92
3.20
0
2.10
43.74
26.70
10.00
1.91
$182.12
 In the production cost model, new foundry batteries were assumed to be
 4-meter batteries and new furnace batteries were assumed to be 6-meter
 batteries.
                                  C-52

-------
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 technol-
ogy [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 have already 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-53

-------
I ru.i -
140 -
I 1-30-
c
o
'-4J
3 1 20 -
T?
O
0 I-
(^ Q-
° 1 1 0 -
o
0

1 00 -

90 -
D

D D

D n
D C1D


D
CI
* a 'B °
a a a j-p a a
a

D D n

D
Mi)
i i i i i i i i ii i i i i i I I I i (
0.2     0.4     0.6     0.8     1.0     1.2     1.4     1.6     1.8    2.0
                          Production (1,000,000 Mg/Yr)
       Figure C-5. Estimated average cost of furnace coke production as a function of plant production, 1984.
2.2

-------
i yu -
180 -
'o> 1 70 -
\
** 1 60 -
c
o
-M
o _3 1 50 -
(Jl \.f
en fj
Q.
•5 1 40 -
O
0 1 .30 -
120 -
1 i n -
a
a

a D
a

a
" a

a
50.0
150,0            250,0            350.0
           Production (1000  Mg/Yr)
4-50.0
       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 produc-
tion 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-56

-------
i
tn



*' \
\
M
o
0





zuu -
190 -
180 -
170 -
160 -
150 -
1 40 -
1.30 -

120 -
110-
100 -
on
«7 v,1
0.





.-'"*
f'
1
/
/•
^__ 	 	 .. 	 — •-" — •" 	 ^ 	 	
^^-^- ~ ~~
1 1 1 1 1 1 1 1 1 1 1 1 1 1
0 4.0 8,0 12,0 16.0 20.0 24.0 28.0
                                      Production (1,000,000
                         M-cirglnol cost                 • --  Averoge cost
                              Figure C-7. Marginal and average cost functions for furnace coke, 1984.

-------
I
en
CD
   o
   o
        180 -
        170 -
        160 -
150 -
        140-
        1 30 -
        120 -
        110
         /
                    S

                 ]   i    i   i
            0.2    0.4    0.6
                          \   i    i
                        0.8    1.0
 I   T   III^  T^

1.2    1.4    1.6    1.8
2.0    2.2    2.4
                                   Production (1,000,000 Mg/Vr)
                       Marginal  cost                -  Average  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 multicol linearity 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 steelmaking,  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, was also attempted.   Unfortunately,
the data necessary to properly estimate the demand function was 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-59

-------
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 which
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 Qi in Figures C-9 and C-10).  In
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.   Though 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 Pt and Q2 in Figure C-10.
     C.2.2.4  Economic Impact Variables.   Table C-18 shows the specific
economic variabl-es 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 readily estimated from this model.
                                  C-60

-------
      TABLE C-17.   ESTIMATES OF ELASTICITIES OF STEEL AND COKE MARKETS

                                                 Point         Interval
                                                estimate       estimate

1.    Percent change in furnace coke demand       -1.29
     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 steel demand for 1           -1.86       (-0.54, -3.18)
     percent change in the price of steel

     Percent change in steel  imports for
     1 percent change in the  price of steel
6.    Percent change in steel  imports for          1.51       (0.51,  2.51)
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

-------
S/Mg
    P2




    PI
                                    Q2     Ql
                                                                   S2
                                                                    SI
                                                           10   Mg  / Yr
                Figure C-9. Coke supply and demand without import competition.
                                     C-62

-------
$/Mg
                                                                    S2
                                                                    SI
    PI
                          Q2
Ql
                                                           10   Mg / Yr
                 Figure C-10. Coke supply and demand with import competition.
                                   C-63

-------
        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-64

-------
     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 which must compete with imported coke which is
assumed to be a perfect substitute for domestic coke (Scenario B).  In
Figure Oil, D represents the derived demand for coke.  The line Sx repre-
sents the baseline supply curve for coke.   The equilibrium price and quantity
are Pt and Q^ respectively.  The area cl+g+h is the total cost of coke
production, b+cl+c2+e+g+h is the total revenue, and b+c2+e represents
before-tax profits.   The total cost of coke production (cl+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 due to
the small magnitude of changes and the continued production by uneconomic
firms.  Given the demand and supply functions as drawn in Figure Oil,
higher costs of production will lead to higher prices.  A production decrease
as shown in Figure Oil would cause price to rise to P_ and the quantity
demanded to fall to Q«.   The actual costs to the producer of the regulatory
alternative are c^+d-c,, and profits before income taxes are a+b+c,.
     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 P~ to purchase.
     In Figure C-12, D represents derived demand for coke and S,  represents
the baseline supply curve for coke, with P, and Q, representing equilibrium
price and quantity
     As in Figure Oil,  area c,+g+h is the total cost of coke production,
including expenses incurred to meet baseline environmental regulations.
Area b+c,+c2+e+g+h is the total revenue, and b+c2+e is before-tax profits.
     The regulatory alternatives shift the supply function to $„.   As
explained, this is not a parallel  shift.   In this scenario,  price does not
rise even though domestic production is reduced.   Instead, since imported
coke is assumed to be a perfect substitute for domestic coke, and since
imported coke is assumed to be available at price P,,  domestic  consumers
purchase more imports and less domestic coke.   The results are  that domestic
                                  C-65

-------
$/Mg
     P2


     PI
                                                 Q2       Ql
                                                                          S2
SI
                                                                10   Mg/Tr
                  Figure C-11. Coke demand and supply with and without regulatory
                            alternatives, without import competition.
                                     C-66

-------
S/Mg
                                                                           S2
     PI
                                      Q2
Ql
                                                                10   Mg /Yr
                      Figure C-12. Coke demand and supply with and without
                         regulatory alternatives, with import competition.
                                        C-67

-------
production decreases to Q2, imports increase by Q-i'Qo' anc^ price remains at
P,,  as shown in Figure C-12.
     The costs of the regulation to the producer are c^-c,.  Total revenue
is b+c,+c2+h, and production costs are h+c^.  Profits before income taxes
are  b+c,.   There are no costs to consumers because they are able to purchase
quantity Q, at price P, 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 1970s and peaked at $956.5 million (1972 dollars) in 1979.   Expend-
itures declined slightly in 1980 and 1981, and dropped to an eight-year low
of $549.8 million (1972 dollars) in 1982.

                                  C-68

-------
          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 (103 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 (103 Mg)                          "76,763  .
  Consumption (103 Mg)                          75,710
  Imports (103 Mg)                              15,486
  Employment (jobs)                            295,000

 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 furnace coke was $123.51 in
 the fourth quarter of 1983.
cCalculated.  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-69

-------
                    TABLE C-20.   POLLUTION  ABATEMENT  EXPENDITURES  FOR  SIC  331262
Abatement expenditures (106 current $)
Year
1975
1976
1977
1978
1979
1980
1981
1982
Air
pollution
477.3
606.9
675.1
709.1
932.6
925.4
940.9
666.8
Water
pollution0
306.8
309.5
385.8
424.9
511.6
494.9
512.6
408.3
Solid.
waste
43.1
51.4
85.5
111.2
99.5
127.9
153.0
92.4
Recovered Cost
(106 current $)
18.1
17.8
18.1
15.1
1.0
18.8
22.3
22.0
Total Cost3
(106 current $)
809.1
950.0
1,128.3
1,230.1
1,542.7
,1,529.4
1,584.2
1,145.5
Total cost6
(106 1972 $)
647.3
722.0
812.4
824.2
956.5
856.5
807.9
549.8
 Total  cost = Capital  expenditure +  operating costs  -  recovered  cost  summed for all  pollutants.
 Implicit price deflator for the nonfarm business  sector used  to convert  current dollars  to
 1972 values.
"Includes payments  to  government units  for  public  sewage use.
 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 steel making; 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-71

-------
  TABLE C-21.   PRICE EFFECTS OF REGULATORY ALTERNATIVES-
                    FURNACE COKE, 1984a

Regulatory                         Coke            Steel
Alternative                       ($/Mg)           ($/Mg)


    II                             0.13             0.04
                                  (0.12)           (0.01)

    III                            0.36             0.12
                                  (0.33)           (0.04)

 Values in parentheses are percentage changes from baseline.
                            C-72

-------
                 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)
Values i.n 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 signif-
icantly since 1972, but peaked in 1979 and began a marked decline.   •
                         Year       Imports (103 Mg)
                         1972            168
                         1973            978
                         1974          3,211
                         1975          1,650
                         1976          1,189
                         1977          1,659
                         1978          5,191
                         1979          3,605
                         1980            598
                         1981            478
                         1982 .           109
                         1983             32
     The increase in imports in the 1970s is believed to be the result of a
coal  strike in the United States during 1978 combined with depressed condi-
tions in the market for steel in the countries exporting coke to the United
States.  Accordingly, future importation at a high level may depend upon
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-74

-------
     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 regula-
tory alternatives are summarized in Table C-24.  Capital costs have also
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 relation between
total capital cost and the annual average net capital investment of the
company and the annual cash flow of the company.   This analysis is pre-
sented 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 regulatory
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 alternatives
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-75

-------
          TABLE C-23.   COAL CONSUMPTION AND EMPLOYMENT EFFECTS OF
               REGULATORY ALTERNATIVES—FURNACE COKE, 1984
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)
aValues 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 by
 the relevant controls.

 Annual labor productivity in coal  mining is estimated as 3,515 megagrams
 per year per job.
                                 C-76

-------
         TABLE C-24.   INDUSTRY CAPITAL REQUIREMENTS OF REGULATORY
                     ALTERNATIVES—FURNACE COKE, 1984

                                                Capital costs
Regulatory                                      of regulations
Alternative                                     (106 1984 $)

    II                                                 38

    III                                                68


Calculated for all plants in operation in 1984.44 46
                                 C-77

-------
     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 was 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 subsidi-
aries, cash flow of parent companies was used.  'Capital  costs as percentages
of cash flows range from 0 to 8 percent for the regulatory alternatives.
     The leverage ratios presented in Table C-15 indicate that coke-producing
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-78

-------
                  TABLE C-25.   CAPITAL COSTS OF COMPLIANCE AS A PERCENTAGE OF NET INVESTMENT-
                                          FURNACE COKE PRODUCERS, 1984
Regulatory
Alternative
II
III
Average annual net
Investor Service,
Furnace coke producer
A
1
2
B
1
2
income
New York
C
0
0
calculated
, 1984, and
D
1
2
E
3
5
from company
Standard New
F
I
1
prof i
York
G
2
3
les in
Stock
H
1
3
Moody1 s
Exchange
I
1
1
Industrial
Reports,
J
1
2
Manual ,
Standard
K L
0 1
1 2
Moody 's
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.

       Data on annual investment are not available for two companies.
o
to

-------
            TABLE C-26.   CAPITAL COST AS A PERCENTAGE OF ANNUAL
                 CASH FLOW—FURNACE COKE PRODUCERS, 1984
Regulatory
Alternative
II
III
Furnace Coke Producers

A
4
8

B
5
8

C
0
0

D
2
5

E Fc
2
4

G
2
4

H
1
1

I JC
0
1

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 GNF implicit price deflator.

 There are three furnace coke producers for which  cash flow data was not
 available.

cThis company had negative cash flow.
                                  C-80

-------
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 would probably be unacceptable.  An analysis of
the iron and steel industry undertaken by Temple, Barker, and SToane, Inc. ,63
addressed the question of external financing with regard to water pollution
control expenditures.   This analysis concludes that to avoid deterioration
in its financial condition, the industry is likely to reduce expenditures
to modernize production facilities rather than increase its external financ-
ing.   In fact, as described in Section 1.5, capital expenditures for pollu-
tion abatement as a percentage of new capital expenditure are decreasing.
     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 priceline 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 regula-
tion 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-81

-------
     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 fourteen 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 would also be justified if the firm produces only
the most highly valued steel products.
                                  C-82

-------
     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 managers'
perceptions regarding their future costs and revenues that exact projections
of closures should be viewed with caution.  In the current market for
steel, it is difficult to say whether or not uneconomic batteries will be
closed.  Of companies owning uneconomic batteries, three have cut back
capacity by closing batteries, though 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, domesti-
cally 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-83

-------
       TABLE C-27.   1983 BASELINE VALUES FOR ECONOMIC
            IMPACT ANALYSIS—FOUNDRY COKE, 198395

                                            Baseline values

Coke market

  Price (1983 $/Mg)                               169. 58b
  Production (103 Mg)                           2,951
  Consumption (103 Mg)                          2,938
  Employment (jobs)                               542
  Coal  Consumption (103 Mg)                     3,809

 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-84

-------
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 regula-
tory alternatives are summarized in Table C-30.   The  capital  requirements
                                  C-85

-------
    TABLE C-28.   PRICE AND QUANTITY EFFECTS OF REGULATORY ALTERNATIVES
                            FOUNDRY COKE, 1984

Regulatory                    Coke price impact        Coke quantity impact
Alternative                     (1984 $/Mg)                (103 Mg/yr)

Scenario A
    II                             0.99                       -18
                                  (0.58)                      (-0.61)

    III                            1.46                       -26
                                  (0.86)                      (-0.88)
Scenario B
  II                               0.00                       -61b
                                  (0.00)                      (-2.07)

  III                              0.00                       -94b
                                  (0.00)                      (-3.18)°

 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-86

-------
    TABLE C-29.   COAL CONSUMPTION AND EMPLOYMENT EFFECTS OF REGULATORY
                     ALTERNATIVES—FOUNDRY COKE, 1984a
Regulatory
Alternative
Scenario A
II
III
Coal
consumption
for coke
(103 Mg/yr)

-23
(-0.78)
-34
(-1.16)
Employment (jobs)
Coal Coke
mining plant

-6 , -3
(O.OOr (-0.55)
-9 , -5
(0.00)a (-0.92)
Scenario B

    II                                -78            -22         -11
                                      (-2.05)         (-0.01)      (-2.03)

    III                              -121            -34         -17
                                      (-3.18)         (-0.02)      (-3.14)

 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
 by the relevant controls.

 Annual labor productivity in coal  mining is estimated as  3,515 megagrams
 per year per job.

 Zero due to rounding.
                                  C-87

-------
         TABLE C-30.   INDUSTRY CAPITAL REQUIREMENTS OF REGULATORY
                     ALTERNATIVES--FOUNDRY COKE,  1984

                                                   Capital  costs.
Regulatory                                        of regulations3
Alternative                                         (106 1984 $)

    II                                                    7

    III                                                  12

Calculated for all  plants in operation in 1984.48 50
                                  C-88

-------
to meet Regulatory Alternatives II and III for the foundry coke industry
are $7 million and $12 million, respectively.  Capital costs have also 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, total capital cost as a percentage
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 which are facing the highest pending
compliance costs are operating with batteries which were installed between
1919 and 1946.  Conversely, the foundry coke producers which are facing the
lowest pending compliance costs are operating, for the most part, with
batteries installed between 1950 and 1979.
     Table C-32 provides capital cost as a percentage of annual cash flow.
The regulatory alternatives result in capital costs no greater than 2 percent
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.  Since 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-89

-------
     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, Moody's Investor Service, New York, 1984, Standard
New York Stock Exchange Reports, Standard and Poor 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-90

-------
 TABLE C-32.  CAPITAL COST AS A PERCENTAGE OF ANNUAL CASH
            FLOW—FOUNDRY COKE PRODUCERS, 1984a
Foundry coke producers
Regulatory alternative
II
III
AA
1
1
BB
0
1
CC
0
1
DO
1
2
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 six foundry coke producers for which cash flow
data was not available.
                            C-91

-------
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 is also 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 are 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-92

-------
 TABLE C-33.  CAPITAL COST AS A PERCENTAGE OF ANNUAL CASH
            FLOW—FOUNDRY COKE PRODUCERS, 1984a
Foundry coke producers
Regulatory alternative
II
III
AA
1
I
BB
0
1
CC
0
1
DD
1
2
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 six foundry coke producers for which cash flow
data was not available.
                            C-93

-------
TABLE C-34.   COMPLIANCE COSTS OF REGULATORY ALTERNATIVES-
               FOUNDRY COKE PRODUCERS, 1984

    Regulatory                          Compliance
    Alternative                    cost (106 1984 $/yr)

        II                                  1.3

        III                                  2.9
                            C-94

-------
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, since 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
Pennsylvania, 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
communities 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 regulatory
flexibility analysis is required for regulations which have a "significant
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.   The Environmental Protection Agency Office
of Standards and Regulations recently drafted a memo on RFA compliance.89
This section addresses the requirements  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, and
          Estimation of the economic impact of the NESHAP on these
          small firms.
                                 C-95

-------
     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 less 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."89  This
rule implies that at least two firms be impacted to qualify as a "substan-
tial 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 standard 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-96

-------
     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-97

-------
in addition to external sources of financing.   Small, privately owned firms
often do not report their annual investment expenditures, so that determina-
tion 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 the U.S. Environmental Protection Agency
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 compared
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 is
currently 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.  Standard Industrial Classification Manual, 1972.  Executive Office of
     the President, Office of Management and Budget.   Washington, DC.   U.S.
     Government Printing Office.  1972.  p. 145.
 2.  1982 Census of Manufactures.   Preliminary Report, Industry Series.
     Blast Furnaces and Steel Mills (Industry 3312).   Bureau of the Census,
     U.S. Department of Commerce.   Washington, DC.  MC82-I-33A-1(P).
     July 1984.  p. 3.
                                 C-98

-------
 3.   Quarterly Coal Report, October - December 1982.  Appendix A.  Office
     of Coal, Nuclear, Electric, and Alternate Fuels, Energy  Information
     Administration, U.S. Department of Energy.  Washington,  DC.  DOE/EIA-
     012K82/4Q).  April 1983.  p. 54-55.

 4.   Energy Data Reports:  Coke and Coal Chemicals  in 1979.   Office of Coal
     and Electric Power Statistics, Energy Information Administration, U.S.
     Department of Energy.  Washington, DC.  DOE/EIA 0121(79).  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.   Minerals Yearbook.  Bureau of Mines, U.S.  Department of  the Interior.
     Washington, DC. _1950-1978.  (Table and page numbers vary because of
     reelassifications.)

 7.   Statistical Abstract of the United States:  1978.   Table 710.  99th
     Edition.  Bureau of the Census, U.S. Department of Commerce.
     Washington, DC.  1978.  p. 441.

 8.   Statistical Abstract of the United States:  1984.   Table 735.  104th
     Edition.  Bureau of the Census, U.S. Department of Commerce.
     Washington, DC.  1984.  p. 449.

 9.   Reference 7.  Table 1511.  p. 874.

10.   Historical Statistics of the United States:   Colonial Times to 1970.
     Part 2.   Bureau of the Census, U.S. Department of Commerce.  Washington,
     DC.  1975.  p. 904-905.

11.   U.S. Exports—Domestic Merchandise, SIC-Based  Products by World Areas.
     Bureau of the Census, U.S. Department of Commerce.   Washington, DC.
     FT610/Annual.   1978-1984.

12.   U.S. Imports—Consumption and General, SIC-Based Products by World
     Areas.   Bureau of the Census, U.S. Department  of Commerce.  Washington,
     DC.  FT210/Annual.  1978-1981, 1983-1984.

13.   U.S. General Imports—Schedule A, Commodities  by Country.  Bureau of
     the Census, U.S.  Department of Commerce.   Washington, DC.
     FT135/November 1981.  1982.  p. 46.

14.   Survey of Current Business.  Vol. 64, No.  10.  Bureau of Economic
     Analysis, U.S. Department of Commerce.  Washington, DC.   October 1984.
     p. S16-S17.

15.   Reference 8.  Table 1471.  p. 832.
                                   C-99

-------
16.   Quarterly Coal Report, April-June 1984.  Appendix A.  Office of  Coal,
     Nuclear, Electric, and Alternate Fuels, Energy Information Administra-
     tion, U.S.  Department of Energy.  Washington, DC.  DOE/EIA-0121(84/2Q).
     October 1984.   p.  54, 61.

17.   Energy Data Reports:   Coke and Coal Chemicals in 1978.  Office of
     Energy Data Operation, Energy Information Administration, U.S. Depart-
     ment of Energy.   Washington, DC.  1980.  p. 3-5.

18.   Coke and Coal  Chemicals in 1980.  Office of Coal, Nuclear, Electric
     and Alternate Fuels.   Energy Information Administration, U.S. Depart-
     ment of Energy.   Washington, DC.  DOE/EIA 0120(80).  November 1981.
     p.  4.

19.   Reference 8.   Table 799.  p. 486.

20.   1977 Census of Manufactures.  Bureau of the Census, U.S. Department of
     Commerce, U.S. Government Printing Office.  Washington, DC.  1979.
     p.  2.

21.   1972 Census of Manufactures.  Bureau of the Census, U.S. Department of
     Commerce, U.S. Government Printing Office.  Washington, DC.  1976.
     p.  33A-6.

22.   Statistical Abstract of the United States:  1980.  Table 724.  101st
     Edition.  Bureau of the Census, U.S.  Department of Commerce.
     Washington, DC.  1980.  p. 437.

23.   1984 U.S. Industrial  Outlook.   25th Edition.  Bureau of Industrial
     Economics,  U.S.  Department of Commerce.  Washington, DC.  January
     1984.  p. 18-3.

24.   Standard and Poor's,  Inc.  Steel and Heavy Machinery-Basic Analysis.
     Industry Surveys.   Vol. 152, No. 43,  Sec.  1.  New York.  October 1984.
     p.  S15.

25.   Quarterly Coal Report, October - December 1983.  Appendix A.  Office
     of Coal, Nuclear,  Electric, and Alternate Fuels, Energy Information
     Administration,  U.S.  Department of Energy.  Washington, DC.
     DOE/EIA-0121(83/4Q).   April 1984.  p. 55.

26.   Reference 4.   p. 3-5.

27.   Reference 18.   p.  3-5.

28.   Analysis of the U.S.  Metallurgical Coke Industry.  Industrial Economics
     Research Institute.  Fordham University.  October 1979.  p. 41.

29.   Reference 28.   p.  40.
                                 C-100

-------
30.  Quarterly Coal Report, January - March 1984.  Appendix  A.   Office of
     Coal, Nuclear, Electric, and Alternate Fuels, Energy  Information
     Administration, U.S. Department of Energy.  Washington,  DC.   DOE/EIA-
     0121(84/1Q).  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.

32.  Kerrigan, Thomas J.  Influences upon the Future International Demand
     and Supply for Coke.   Ph.D. dissertation.   Fordham  University.  New
     York.  1977.  p. 14, 114.

33.  Reference 4.  p. 1.

34.  Reference 4.  p. 13.

35.  Reference 25.  p. 54.

36.  Standard Support and Environmental Impact Statement:  Standards of
     Performance for Coke Oven Batteries.  Emissions Standards and Engineer-
     ing Division, Environmental Protection Agency.  May 1976.   p. 3-7.

37.  Mineral Commodity Summaries, 1984.  Bureau  of Mines,  U.S. Department
     of the Interior.  Washington, DC.  1984.  p. 78.

38.  The Outlook for Metallurgical Coal and Coke.  Institutional  Report.
     Merrill Lynch, Pierce, Fenner, and Smith, Inc.  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.  Energy Data Reports:   Distribution of Oven  and Beehive  Coke  and Breeze.
     Office of Energy Data  and Interpretation, Energy Information Adminis-
     tration, U.S. Department of Energy.   Washington, DC.  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.  Bureau of Mines, U.S. Department  of the  Interior.
     Washington,  DC.  1960.  p.  4-10.
                                 C-101

-------
48.   Staff Report on the United States Steel Industry and  Its  International
     Rivals:   Trends and Factors Determining International Competitiveness.
     Bureau of Economics, Federal Trade Commission, U.S. Government  Printing
     Office.   Washington, DC.  November 1977.  p. 53.

49.   Schottman, Frederick J.  Iron and Steel.  Mineral Commodity  Profiles,
     1983.  Bureau of Mines, U.S. Department of the Interior.  Washington,
     DC.  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.   Moody's Investors' Service.  Moody's OTC Manual.  New York.   1984.

56.   Standard and Poor's Corp.  Standard and Poor's New York Stock Exchange
     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, Massachusetts.   December 1980.

60.   Annual Survey of Manufactures.  Bureau of the Census, U.S. Department
     of Commerce.   Washington, DC.   1975 - 1982.

61.   1977 Census of Manufactures.  Vol. II.   Industry Statistics.  Part 2,
     SIC Major Groups 27-34.  Bureau of the Census, U.S. Department  of
     Commerce.   Washington, DC.   August 1981.  p. 33A-15.

62.   Pollution Abatement Costs and Expenditures.  Current  Industrial  Reports.
     Bureau of the Census, U.S.  Department of Commerce.  Washington,  DC.
     MA-200(year).  1975 - 1983.

63.   Reference 59.  p.  VI-4.

64.   Reference 24.  p.  S28-S29.
                                 C-102

-------
65.   Reference 23.  p. 18-5.

66.   Reference 8.  Table 1408.  p. 788.

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.   Economic Report of the President,  February  1984.   U.S.  Government
     Printing Office.  Washington, DC.  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.   North Carolina State University.   Raleigh, NC.
     1979.

85.   Heckman, James J.  Shadow Price, Market  Wages,  and  Labor Supply.
     Econometrica.  42:679-694.  July 1974.

86.   Reference 24.  p. S16, S27.
                                 C-103

-------
87.   Reference 25.- p.  55-57,61.

88.   Symonds, William C., and Gregory L. Miles.  It's Every Man for Himself
     in the Steel Business.   Business Week.  (2897):76,78.  June 3, 1985.

89.   Memorandum from Smith,  C.  Ronald, Office of Standards and Regulations,
     U.S.  Environmental Protection Agency, to Steering Committee Representa-
     tives, EPA.   April 12,  1983.  pp. 1-2 and Attachment 1.  Guidelines
     for compliance with Regulatory Flexibility Act.
                                C-104

-------
         APPENDIX D
HEALTH RISK IMPACT ANALYSIS

-------
                           Table 0-1.  FURNACE COKE BY-PRODUCT RECOVERY PLANTS AND SOURCES OF BENZENE  EMISSIONS  FOR  REGULATORY BASELINE
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

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, Dearborn, MI
U.S. Steel, Fairless Hills, PA
LTV Steel , Warren, OH
LTV Steel , E. Chicago, IL
Armco Inc., Ashland, KY
Weirton Steel, Brown's Is., WV
U.S. Steel, Provo, IL
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
TOTALS
Latitude
33°32'47U
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"
41°13'13"
41°39'48"
38° 30 '07"
40°24'58"
40°18'43"
40°37'16"
42°49'20"
42°15'16"
41°26'56"
40°20'36"
41°28'26"
39°29'45"
41°37'41"
40°25'34"
33° 29 '22"
40°36'51"
39°13'10"
41°39'53"
41°36'55"
40° 18 '04"

Coke production
capacity Tar
Longitude (1,000 Mg/yr) Decanter
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"
74044.32-
80°48'40"
87°26'42"
82°40'08"
80°35'16"
111°44'31"
80°14'24"
78°51'H"
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"

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
5,294
45,804
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
7.30E+04
7.42E+04
8.45E+04
8.93E+04
9.38E+04
9.95E+04
1.08E+05
1.15E+05
1.16E+05
1.36E+05
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
4.37E+03
5.88E+03
6.08E+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
1.55E+04
1.68E+04
1.80E+04
1.81E+04
2.11E+04
2.13E+04
2.15E+04
2.15E+04
2.19E+04
2.70E+04
4.21E+04
4.46E+04
5.07E+04
6.35E+04
5.50E+05
Excess
Ammoni a
Liq. Tank
2.84E+03
3.28E+03
4.41E+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
9.87E+03
1.04E+04
1.10E+04
1.16E+04
1.26E+04
1.35E+04
1.36E+04
1.58E+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
4.51E+03
5.31E+03
5.48E+03
5.50E+03
5.59E+03
6.36E+03
6.73E+03
7.06E+03
7.49E+03
8.10E+03
8.68E+03
8.75E+03
1.02E+04
1.03E+04
O.OOE+00
1.04E+04
1.06E+04
2.61E+04
2.03E+04
2.15E+04
2.45E+04
3.07E+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
1.65E+04
1.74E+04
1.83E+04
1.94E+04
2.10E+04
2.24E+04
2.26E+04
2.64E+04
2.66E+04
O.OOE+04
2.69E+04
2.73E+04
3.38E+04
5.26E+04
5.57E+04
6.34E+04
7.94E+04
6.60E+05
Light -oil
Cond. Vent
2.80E+04
3.24E+04
4.36E+04
4.51E+04
5.01E+04
5.07E+04
5.18E+04
6.75E+04
6.92E+04
8.15E+04
8.41E+04
8.44E+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+05
1.59E+05
1.62E+05
2.01E+05
3.12E+05
3.31E+05
3.76E+05
9.42E+05
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+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
Note:  Data current as of November 1984.

-------
                              Table D-l.  FURNACE COKE BY-PRODUCT RECOVERY PLANTS AND SOURCES OF BENZENE EMISSIONS FOR REGULATORY BASELINE  (Continued)
D
 I
NJ
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+04
2.46E+04
2.46E+04
2.46E+04
5.74E+04
2.46E+04
2.46E+04
5.74E+04
8.45E+05
Tar
Dewatering
6.62E+03
7.64E+03
1.03E+04
1.06E+04
O.OOE+00
1.20E+04
1.22E+04
1.59E+04
1.63E+04
1.92E+04
1.98E+04
1.99E+04
2.02E+04
2.30E+04
2.44E+04
2.56E+04
2.71E+04
2.93E+04
3.14E+04
3.17E+04
3.70E+04
3.73E+04
3.76E+04
3.76E+04
3.83E+04
4.73E+04
7.36E+04
7.80E+04
8.88E+04
O.OOE+00
8.39E+05
Flushing
Circ"°Tank
2.84E+03
3.28E+03
4.41E+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
9.87E+03
1.04E+04
1.10E+04
1.16E+04
1.26E+04
1.35E+04
1.36E+04
1.58E+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
Benzene
Storage
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
3.27E+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
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
1.54E+05
O.OOE+00
1.56E+05
1.58E+05
1.96E+05
O.OOE+00
O.OOE+00
3.67E+05
O.OOE+00
1.86E+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.OOE+00
2.58E+04
2.79E+04 ,
2.99E+04
O.OOE+00
O.OOE+00
3.55E+04
O.OOE+00
3.58E+04
3.64E+04
4.51E+04
O.OOE+00
O.OOE+00
8.46E+04
O.OOE+00
4.29E+05
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
1.78E+02
O.OOE+00
1.79E+02
1.82E+02
2.25E+02
O.OOE+00
O.OOE+00
4.23E+02
O.OOE+00
2.14E+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
DWFC
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
2.56E+05
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
1.31E+04
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
6.70E+04
Tar Inc.
Sump
2.99E+04
3.46E+04
4.66E+04
4.82E+04
5.35E+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
(kg/yr)
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
4.04E+05
5.40E+05
4.39E+05
9.59E+05
1.03E+06
1.10E+06
5.38E+05
6.34E+05
1.31E+06
3.99E+05
1.31E+06
1.33E+06
1.64E+06
1.25E+06
1.55E+06
3.06E+06
1.32E+06
2.42E+07

-------
                                      TABLE 0-2.  FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS AND SOURCES OF BENZENE EMISSIONS FOR REGULATORY BASELINE
D
U>
No.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.

Plant
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
TOTALS
Latitude
35°02'16"
39°26'48"
41°40'10"
33°14'25"
42°08'43"
42°58'56"
38°32'08"
40°05'12"
39°45'16"
33°33'40"
40°28'49"
33°26'13"
33°34'57"
42°iri9"

Coke production Excess
capacity Tar Tar Ammonia
Longitude (1,000 Mg/yr) Decanter Storage Liq. Tank
85°18'11"
87°23'47"
83°29'31"
87°30'11"
80°01 ' 32"
78°56'19"
90°16'05"
75°18'59"
86°06'49"
86048'38"
80°03'34"
86°57'50"
86°46'50"
83°09'16"

130
132
157
161
207
299
330
402
477
499
521
563
583
617
5078
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
1.89E+04
2.04E+04
2.11E+04
2.23E+04
1.84E+05
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
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
4.05E+03
3.34E+04
Light-
Oil
Storage
4.07E+02
4.13E+02
O.OOE+00
5.04E+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+03
1.30E+03
O.OOE+00
2.42E+03
O.OOE+00
3.26E+03
O.OOE+00
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+00
2.40E+04
2.50E+04
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-oil
f1r£-
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
           aBenzene emissions are in kilograms/yr.
           Note:   Data current as of November 1984.

-------
                                TABLE  D-2.   FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS AND SOURCES OF BENZENE  EMISSIONS  FOR  REGULATORY BASELINE (Continued)
Leaks
2.24E+04
2.24E+04
O.OOE+00
2.24E+04
O.OOE+00
2.24E+04
O.OOE+00
5.22E+04
O.OOE+00
2.24E+04
2.24E+04
2.24E+04
2.24E+04
O.OOE+00
2.31E+05
Tar
Dewatering
1.28E+03
1.30E+03
1.55E+03
O.OOE+00
2.04E+03
2.95E+03
3.26E+03
3.97E+03
4.71E+03
4.93E+03
5.14E+03
5.56E+03
5.75E+03
6.09E+03
4.85E+04
Flushing
liquor
Circ. 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
4.05E+03
3.34E+04
Benzene
Storage
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
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
1.51E+05
Naphth.
Melt Pit
1.90E+03
1.93E+03
O.OOE+00
2.35E+03
O.OOE+00
O.OOE+00
O.OOE+00
5.87E+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.OOE+00
O.OOE+00
2.93E+01
3.48E+01
3.64E+01
O.OOE+00
O.OOE+00
4.26E+01
O.OOE+00
1.74E+02
TBFC
O.OOE+00
O.OOE+00
8.02E+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
2.88E+04
O.OOE+00
O.OOE+00
3.68E+04
DWFC
2.56E+04
2.60E+04
O.OOE+00
3.17E+04
O.OOE+00
O.OOE+00
O.OOE+00
7.92E+04
9.40E+04
9.84E+04
O.OOE+00
O.OOE+00
1.15E+05
O.OOE+00
4.70E+05
BTX
Storage
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
1.26E+03
O.OOE+00
1.56E+03
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
2.82E+03
Tar Inc.
Sump
5.80E+03
5.89E+03
7.01E+03
7.19E+03
9.24E+03
1.34E+04
1.47E+04
1.79E+04
2.13E+04
2.23E+04
2.33E+04
2.51E+04
2.60E+04
2.75E+04
2.27E+05
TOTAL BENZENE
BY PLANT
(kg/yr)
8.06E+04
8.15E+04
2.52E+04
9.29E+04
2.27E+04
7.41E+04
3.61E+04
2.35E+05
1.84E+05
2.48E+05
1.12E+05
1.48E+05
2.84E+05
• 6.76E+05
1.69E+06
D
 I

-------
 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          (m)a     Area (m2)a         (m)       (m/s)b       (°K)c
Storage tanks
for Light Oil,
BTX, or Benzene   4.3

Excess Ammonia
Liquor 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
Tanke             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

^Stack 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 m^/s for other sources, which have
 greater emissins.

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-5

-------