United States Office of Air Quality EPA-450/3-83-016b
tnvironmental Protection Planning and Standards June 1988
A9encV Research Triangle Park NC 27711
Air
<»EfiA Benzene Emissions Draft
from Coke EIS
By-Product
Recovery Plants-
Background
Information for
Revised Proposed
Standards
-------
ERRATA FOR COKE BY-PRODUCT RECOVERY PLANT
BACKGROUND INFORMATION DOCUMENT FOR REVISED PROPOSED STANDARDS
This background information document (BID) responds to comments on the
1984 proposal and also serves as the basis for reproposal of a revised
standard based on EPA's response to the court decision noted on page 1-1 of
this BID. However, readers of this document should note that while this
BID refers to "the revised proposed standard" on several pages, EPA is
proposing a total of four different regulatory approaches that would result
in different revised proposed standards. References in this BID to the
"revised proposed standard" and associated impact data pertain to
Approaches A and B presented in the preamble. All information on the
revised proposed standards under Approaches C and D is presented in the
preamble.
-------
EPA-450/3-83-016b
Benzene Emissions from
Coke By-Product Recovery Plants-
Background Information
for Revised Proposed Standards
Emission Standards Division
ft^?*!*"*
^Sfency
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
June 1988
-------
This report has been reviewed by the Emission Standards Division of the Office of Air Quality Planning and
Standards, EPA, and approved for publication. Mention of trade names or commercial products is not
intended to constitute endorsement or recommendation for use. Copies of this report are available through
the Library Services Office (MD-35), U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina 27711; or, for a fee, from the National Technical Information Services, 5285 Port Royal Road,
Springfield, Virginia 22161.
-------
ENVIRONMENTAL PROTECTION AGENCY
Background Information
and Draft
Environmental Impact Statement
sed Proposed Standards for
y-Product Recovery Plants
Prepared by:
/
Jack R. Farmer (Date)
Director, Emission Standards Division
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
1. The revised proposed national emission standards would limit emissions
of benzene from existing and new coke by-product recovery plants. The
revised proposed standards would implement Section 112 of the Clean Air
Act and are based on the Administrator's determination of June 8, 1977
(42 FR 29332), that benzene presents a significant risk to human health
as a result of air emissions from one or more stationary source
categories and is therefore a hazardous air pollutant. The EPA
Regions III, IV, and V are particularly affected because most plants
are located in these areas.
2. Copies of this document have been sent to the following Federal
Departments: Labor, Health and Human Services, Defense, Transportation,
Agriculture, Commerce, Interior, and Energy; the National Science
Foundation; the Council on Environmental Quality; State and Territorial
Air Pollution Program Administrators; EPA Regional Administrators; Local
Air Pollution Control Officials; Office of Management and Budget; and
other interested parties.
3. For additional information contact:
Mr. Gilbert H. Wood
Emission Standards Division (MD-13)
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
Telephone: (919) 541-5625
4. Copies of this document may be obtained from:
U.S. Environmental Protection Agency Library (MD-35)
Research Triangle Park, North Carolina 27711
Telephone: (919) 541-2777
National Technical Information Service
5285 Port Royal Road
Springfield, Virginia 22161
Telephone: (703) 487-4650
iii
-------
TABLE OF CONTENTS
Pa^e
Tables viii
1. SUMMARY l l
1.1 SUMMARY OF CHANGES SINCE 1984 PROPOSAL! .'.'.'.' ill
1.2 SUMMARY OF IMPACTS OF REVISED PROPOSED ACTION ....''.' 1-2
1.2.1 Environmental and Energy Impacts of
Revised Proposed Action 1_2
1.2.2 Health Risk Impacts of Revised Proposed Action .' .' 1-3
1.2.3 Cost and Economic Impacts of Revised Proposed
Action 1_3
1.2.4 Other Considerations .**'.''*' 1-4
1.2.4.1 Irreversible and Irretrievable
Commitment 1_4
1.2.4.2 Environmental and Energy Impacts
of Delayed Standards 1-4
1.2.4.3 Urban and Community Impacts 1-5
2. SUMMARY OF PUBLIC COMMENTS ?-l
3. SELECTION OF SOURCE CATEGORY 3 i
3.1 SELECTION OF SOURCE CATEGORY 3"i
3.2 REGULATION OF MERCHANT PLANTS 31
3.3 EXCLUSION OF FORM-COKE PLANTS "'*''!! 3-2
4. SELECTION OF REVISED PROPOSED STANDARDS 41
4.1 SELECTION OF LEVEL OF CONTROL 4 }
4.2 REGULATORY DEFINITIONS OF FOUNDRY AND FURNACE
BY-PRODUCT PLANTS 4_2
5. EMISSION CONTROL TECHNOLOGY . . c i
5.1 DEMONSTRATION OF CONTROL TECHNOLOGY '. '. 51
5.2 SAFETY, DESIGN, AND OPERATION OF POSITIVE-PRESSURE* ' ' '
CONTROL SYSTEM 5 3
5.3 SAFETY, DESIGN, AND OPERATION OF NEGATIVE-PRESSURE* * '
CONTROL SYSTEM ,- o
5.4 MONITORING FOR CARBON MONOXIDE . . ' ! Tin
5.5 SUMP CONTROLS jf {V
5.6 OPERATIONAL PROBLEM FROM PLUGGED*VENTs'oR VALVES* .' .' .' " i-12
-------
TABLE OF CONTENTS (con.)
Page
5.7 CONTROLS FOR BENZENE STORAGE TANKS 5-12
5.8 DETERMINATION OF CONTROL EFFICIENCIES 5-13
5.9 GAS BLANKETING VERSUS WASH-OIL SCRUBBERS 5-14
5.10 FINAL COOLERS AND NAPHTHALENE PROCESSING 5-15
6. ENVIRONMENTAL IMPACTS 6-1
6.1 DATA BASE FOR ENVIRONMENTAL IMPACTS 6-1
6.2 FOUNDRY PLANT EMISSION FACTORS 6-4
6.3 MODEL COKE PLANTS 6-7
5.4 EMISSION FACTORS FOR TAR-RELATED SOURCES 6-9
6.5 METHODOLOGY FOR EMISSION FACTORS 6-12
6.6 VOC BENEFITS FOR OZONE REDUCTION 6-13
7. COST IMPACT 7-l
7.1 REVISIONS TO COST ANALYSIS 7-1
7.2 REVISIONS TO PRODUCT RECOVERY CREDITS 7-2
7.3 ECONOMIES OF SCALE FOR SMALL PLANTS 7-4
8. ECONOMIC IMPACT 8-1
8.1 REGULATORY BASELINE 8-1
8.2 SELECTION OF DOLLAR YEAR 8-2
8.3 POTENTIAL ECONOMIC IMPACT 8-2
8.4 ESTIMATED EMPLOYMENT IMPACT 8-3
8.5 IMPORT TRENDS 8-4
8.6 ECONOMIC IMPACT ON SMALL PLANTS 8-4
8.7 PRICE IMPACTS 8-6
8.8 ECONOMIC IMPACTS ON FOUNDRY PLANTS 8-7
9. QUANTITATIVE RISK ASSESSMENT 9-1
9.1 USE OF MODEL FOR HEALTH RISK ESTIMATES 9-1
9.2 SELECTION OF RISK MODEL 9-3
9.3 UNIT RISK ESTIMATE 9-5
9.4 DERIVATION OF UNIT RISK ESTIMATE 9-7
9.5 COMPARATIVE RISK FROM GASOLINE MARKETING 9-8
9.6 COMPARATIVE RISKS FROM OTHER SOURCES 9-9
9.7 SELECTION OF BENZENE VS. POM 9-10
9.8 CONSIDERATION OF OTHER HEALTH EFFECTS 9-10
9.9 ANCILLARY COMMENTS 9-12
10. EQUIPMENT LEAK DETECTION AND REPAIR 10-1
10.1 DETERMINE EMISSIONS OVER BACKGROUND LEVELS 10-1
10.2 COMPLIANCE WITH LEAK DETECTION AND REPAIR PROGRAM . . . 10-1
10.3 DEFINITION OF EQUIPMENT LEAK 10-2
10.4 ON-LINE VALVE REPAIR 10-5
10.5 EQUIPMENT LEAK REPAIR PERIOD 10-6
10.6 DELAY OF REPAIR 10-8
10.7 ALTERNATIVE STANDARD FOR VALVES 10-8
10.8 EXEMPTION FOR DIFFICULT-TO-MONITOR VALVES 10-9
10.9 ALTERNATIVE STANDARD FOR OPEN-ENDED VALVES OR LINES . . 10-11
vi
-------
TABLE OF CONTENTS (con.)
Page
11. RECORDKEEPING AND REPORTING jj.j
11.1 ALTERNATIVE MONITORING AND RECORDKEEPING ..!!!'*' 11-1
11.2 RETENTION PERIOD FOR RECORDS AND REPORTS .... 11-1
11.3 ENFORCEMENT BASED ON RECORDS AND REPORTS FOR
EQUIPMENT LEAKS u_2
12. MISCELLANEOUS 12 l
12.1 ALTERNATIVE MEANS OF EMISSION LIMITATION ] .* '. 12-1
12.2 DEFINITION OF TAR DECANTER * i2 i
12.3 DEFINITION OF EXHAUSTER 12 2
12.4 WAIVER REQUESTS 12_2
12.5 NEED FOR ADDITIONAL ENFORCEMENT RESOURCES !!!"*"' 12-3
12.6 SELECTION OF FORMAT * * ' 12 3
12.7 LIGHT-OIL SUMP CONTROL EFFICIENCY .....'!.'.' .* .' .' !2-5
APPENDIX A ENVIRONMENTAL IMPACT ANALYSIS A-l
APPENDIX B COST IMPACT ANALYSIS ....
APPENDIX C ECONOMIC IMPACT ANALYSIS r 1
APPENDIX D HEALTH RISK IMPACT ANALYSIS '. '. o_
-------
TABLES
Number Zi3§.
2-1 List of Commenters on 1984 Proposed National Emission
Standards for Coke By-Product Recovery Plants 2-2
viii
-------
1. SUMMARY
On June 6, 1984, the Environmental Protection Agency (EPA) proposed
national emission standards for benzene emissions from coke by-product
recovery plants (49 FR 23522) under the authority of Section 112 of the
Clean Air Act (CAA). Public comments were requested on the proposal in
the Federal Register, and the comment period was extended, by request, to
October 17, 1984 (49 FR 33904). The 20 commenters were composed mainly
of affected companies and industry trade associations. Also commenting
were various State and county air pollution control or environmental
health departments and one environmental group. The comments that were
submitted, along with responses to these comments, are summarized in this
document. The EPA reconsidered the proposed standards in light of the
court decision in Natural Resources Defense Council, Inc. v. EPA,
824 F.2d 1146 (D.C. Cir., July 28, 1987) and reproposed the standards in
June 1988. The summary of comments and responses serves as the basis for
the revisions made to the standard between proposal and reproposal.
1.1 SUMMARY OF CHANGES SINCE 1984 PROPOSAL
Since the 1984 proposal, the data base has been revised to reflect
the industry operating status as of November 1984 (shortly after the
close of the comment period). Based on comments received on the 1984
proposal, EPA revised the estimated nationwide impacts of control
(including baseline) for furnace and foundry coke producers separately.
The Administrator used the revised environmental, health, cost, and
economic impacts for his reconsideration.
One major change since the 1984 proposal is the revised proposal of a
zero emission limit for naphthalene processing operations, final coolers,
and final-cooler cooling towers at plants producing furnace coke. The
revised proposal is based on wash-oil final coolers. Another major change
is that proposed standards for control of storage tanks containing
1-1
-------
light-oil, benzene-toluene-xylene (BTX) mixtures, benzene, or excess
ammonia-liquor at furnace and foundry plants have been eliminated from the
reproposed standards.
Other changes to the standards have been made for clarifying purposes.
The definition of "coke by-product recovery plant" has been revised specif-
ically to exclude form-coke plants. New definitions for "furnace" and
"foundry" coke and coke by-product recovery plant also have been added to
distinguish these industry segments in terms of the volatile content of
coke produced, the length of the coking cycle and the percent of each
type of coke produced annually. New definitions for "exhauster" and "tar
decanter" also have been added. For gas-blanketed process vessels and
light-oil sumps, monitoring provisions have been added to ensure that there
are no leaks from the access hatches and covers upon reclosure after usage.
The regulation also has been revised to directly cross reference the
provisions of 40 CFR 61, Subpart V for equipment leak requirements. The
EPA also proposes to amend Subpart V where necessary for clarification of
the cross referencing.
1.2 SUMMARY OF IMPACTS OF REVISED PROPOSED ACTION
1.2.1 Environmental and Energy Impacts of Revised Proposed Action
The environmental and energy impacts of the revised proposed standards
are discussed in Chapter 6 of this background information document (BID).
The estimated environmental impacts have been revised since the 1984
proposal to update the operating status of the industry to November 1984.
These changes are discussed in Chapter 6, "Environmental Impacts." Revised
environmental impact tables and emission factors are presented in
Appendix A.
Implementing the revised proposed standards would reduce nationwide
benzene emissions at 44 furnace and foundry plants from the baseline level
of about 26,000 megagrams/year (Mg/yr) to about 2,000 Mg/yr, a 93-percent
reduction. Nationwide emissions from coke by-product recovery plants of
total volatile organic compounds (VOC) including benzene also would be
reduced from the baseline estimated level of 171,000 Mg/yr to about
6,000 Mg/yr, a 96-percent reduction. Assuming recovery of 21.3 liters of
gas/min/Mg of coke/day, the revised proposed standards would result in a
national energy savings of approximately 1,800 TJ/yr from recovered coke
1-2
-------
oven gas. Impact calculations for energy requirements and coke oven gas
recovery estimates are shown in Appendix A.
1.2.2 Health Risk Impacts of Revised Proposed Action
The quantitative risk assessment conducted for the proposed standards is
discussed in Appendix E of the proposal BID (Benzene Emissions from Coke
By-Product Recovery Plants - Background Information for Proposed Standards.
EPA-450/3-83-016a); further information is in the preamble to the proposal
(49 FR 23525). The risk assessment has been revised since the 1984 proposal
to update the current industry operating status and to incorporate adjusted
emission factors. Other changes in the risk assessment since the 1984
proposal include a revised benzene unit risk estimate (URE), which is 17
percent higher, and an increase in the exposure modeling radius to 50
kilometers (km). Further information regarding these changes is provided in
Chapter 9, "Quantitative Risk Assessment," and in Appendix D.
Annual leukemia incidence associated with baseline benzene emissions
at 44 plants is estimated at 3 cases/yr. Implementation of the revised
proposed standards would reduce the estimated incidence to 0.2 case/yr.
The maximum individual lifetime risk (MIR) at the baseline is predicted to
•^
be 6 x 10" . The revised proposed standards are expected to reduce the MIR
to approximately 4 x 10'4 (about 4 in 10,000).
1.2.3 Cost and Economic Impacts of Revised Proposed Action
Control costs for model by-product recovery plants are discussed in
Chapter 7 of the BID for the proposed standards. This analysis has been
updated since the June 1984 proposal to reflect the industry operating status
as of November 1984. Other changes include the adjustment of certain cost
functions and the modification of light-oil/fuel recovery credits, as applied
to plants that practice the flaring of excess coke oven gas. These changes
are discussed in Chapter 7, "Cost Impact." The revised cost analysis is
presented in Appendix B.
Based on the revised analysis, the estimated national capital cost of
the revised proposed standards for the 44 plants is estimated at about
$84 million over baseline costs (1984 dollars). The total annualized cost is
estimated at $16 million/yr.
1-3
-------
The nationwide economic impact of the proposed standards is analyzed in
Chapter 9 of the proposal BID. This analysis, which also has been revised
and updated since the 1984 proposal, is discussed in Chapter 8 of this
document; further information is presented in Appendix C. Based on the
revised economic analysis, the price of foundry and foundry coke is projected
to increase by less than 1 percent above baseline values. The Agency does
not expect closures as a direct result of the revised proposed standards.
However, many furnace and foundry plants are presently in marginal economic
condition or operating at a loss, and the Agency recognizes that the
standards could be a factor that would trigger closure decisions at some of
these plants.
1.2.4 Other Considerations
1.2.4.1 Irreversible and Irretrievable Commitment. As discussed in
Chapter 7 of the BID for the proposed standards, the control options do not
involve a tradeoff between short-term environmental gains at the expense of
long-term environmental losses. An increased cyanide (HCN) concentration in
wastewater is expected if indirect final cooling is used instead of direct
final cooling. Measured HCN air emission and calculations based on
once-through cooling water indicate that about 200 g/Mg of coke could be
added to wastewater for treatment. However, this increase is not anticipated
to cause problems for compliance with effluent regulations.
The control options do not result in irreversible and irretrievable
commitment of resources. As a result of the control options, resources such
as light aromatic hydrocarbons are recovered, and emissions from the
majority of affected sources are reduced substantially or eliminated.
1.2.4.2 Environmental and Energy Impact's of Delayed Standards. The
environmental and energy impacts of delayed standards are discussed in
Chapter 7 of the BID for the proposed standards. Although delayed
promulgation of the revised proposed standards would not impact current
levels of water pollution or solid waste, such a delay would result in
benzene emissions from furnace and foundry plants remaining at the baseline
nationwide level of nearly 26,000 Mg/yr. Total emissions of benzene and
other VOC also would remain at their baseline level of about 171,000 Mg/yr.
No net nationwide savings in energy use would be achieved as a result of
1-4
-------
recovered coke oven gas if implementation of the revised proposed standards
were delayed.
1.2.4.3 Urban and Community Impacts. The beneficial urban and
community impacts of the revised proposed standards include major
reductions in benzene emissions at plant sites, many of which are located
near highly populated areas. This emission reduction would reduce
substantially the health risk associated with operation of coke by-product
recovery plants. An added benefit to urban and community areas is the VOC
emission reduction for ozone nonattainment areas.
The urban and community economic impacts associated with the revised
proposed standards are discussed in Section 9.3.4 of the BID. As indicated
in this analysis, closure of plants now in marginal economic condition
because of market conditions could occur with resulting community impacts.
However, no closures are expected as a direct result of these revised
proposed standards.
1-5
-------
2. SUMMARY OF PUBLIC COMMENTS
A total of 20 comments on the proposed standards and the BID for the
1984 proposed standards were received. A list of commenters, their
affiliations, and EPA docket number assigned to their correspondence is
given in Table 2-1.
For the purpose of orderly presentation, the comments have been
categorized under the following topics:
Chapter 3
Chapter 4
Chapter 5
Chapter 6
Chapter 7
Chapter 8
Chapter 9
Chapter 10
Chapter 11
Chapter 12
Appendix A
Appendix B
Appendix C
Appendix CW
Selection of Source Category
Selection of Final Standards
Emission Control Technology
Environmental Impacts
Cost Impact
Economic Impact
Quantitative Risk Assessment
Equipment Leak Detection and Repair
Recordkeeping and Reporting
Miscellaneous
Environmental Impact Analysis
Cost Impact Analysis
Economic Impact Analysis
Health Risk Impact Analysis
2-1
-------
TABLE 2-1. LIST OF COMMENTERS ON PROPOSED NATIONAL EMISSION
STANDARDS FOR COKE BY-PRODUCT RECOVERY PLANTS
Docket item number3
Commenter and affiliation
IV-D-1
IV-D-2
IV-D-3
IV-D-4
IV-D-5
IV-D-6
IV-D-7
IV-D-8
Ronald J. Chleboski, Deputy Director
Air Pollution Control Bureau
Allegheny County Health Department
Pittsburgh, Pennsylvania 15201
George P. Ferreri, Director
Air Management Administration
Maryland Department of Health
and Mental Hygiene
Baltimore, Maryland 21201
Alfred C. Little
Environmental Engineer
FMC Corporation
2000 Market Street
Philadelphia, Pennsylvania 19103
Danny L. Lewis
Assistant Plant Manager
Empire Coke Company
Birmingham, Alabama 35259
Daniel J. Goodwin, Manager
Division of Air Pollution Control
Illinois Environmental Protection
Agency
2200 Churchill Road
Springfield, Illinois 62706
Glen C. Tenley, Vice President
Koppers Company, Inc.
1201 Koppers Building
Pittsburgh, Pennsylvania 15219
James R. Zwikl
Director of Environmental Control
Shenango Incorporated
Neville Island
Pittsburgh, Pennsylvania 15225
D. C. Miller, Resident Manager
Phosphorus Chemical Division
FMC Corporation
Box 431
Kemmerer, Wyoming 83101
a Footnote at end of table.
(continued)
2-2
-------
TABLE 2-1 (continued)
Docket item number3
Commenter and affiliation
IV-D-9
IV-D-10
IV-D-11
IV-D-12
IV-D-13
IV-D-14
IV-D-15
Donald C. Lang
Director, Air and Water Control
Inland Steel Company
Indiana Harbor Works
3210 Watling Street
East Chicago, Indiana 46312
Lucian M. Ferguson
Executive Vice President
American Coke and Coal Chemicals
Institute
1800 M Street, N.W.
Washington, DC 20036
Lecil M. Colburn
Jim Walter Corporation
P.O. Box 22601
1500 North Dale Mabry
Tampa, Florida 33622
R. Wade Kohlmann
Environmental Engineer
Citizens Gas and Coke Utility
2020 North Meridan Street
Indianapolis, Indiana 46202-1306
David D. Doniger
Senior Staff Attorney
Natural Resources Defense Council,
Inc.
1350 New York Avenue, N.W.,
Suite 300
Washington, DC 20005
Neil Jay King, Esq.
Wilmer, Cutler & Pickering
1666 K Street, N.W.
Washington, DC 20006
David M. Anderson, Director
Environmental and Governmental
Programs
Bethlehem Steel Corporation
Bethlehem, Pennsylvania 18016
>otnote at end of table.
(continued)
2-3
-------
TABLE 2-1 (continued)
Docket item number3
Commenter and affiliation
IV-D-16
IV-D-17
IV-D-18
IV-D-19
IV-D-33b
and
IV-D-34
Terry McGuire, Chief
Technical Support Division
California Air Resources Board
1102 Q Street
P.O. Box 2815
Sacramento, California 95812
Moyer B. Edwards
Director, Environmental Control
Alabama By-Products Corporation
First National-Southern National
Building
P.O. Box 10246
Birmingham, Alabama 35202
Neil Jay King, Esq.
Wilmer, Cutler & Pickering
1666 K Street, N.W.
Washington, DC 20006
Barbara Patala, Acting Chairman
Committee on Environmental Matters
National Science Foundation
Washington, DC 20550
Michael A. Hanson
USS
208 South LaSalle Street
Chicago, Illinois 60604
a The docket number for this project is A-79-16. Dockets are on file at
EPA Headquarters in Washington, DC, and at the Office of Air Quality
Planning and Standards (OAQPS) in Durham, North Carolina.
b Letters numbered IV-D-20 to IV-D-32 are correspondence regarding
extension of the comment period, development of regulatory definitions
for furnace and foundry coke, and responses to information requests and
are not comments on the 1984 proposed standards or BIO.
2-4
-------
3. SELECTION OF SOURCE CATEGORY
3.1 SELECTION OF SOURCE CATEGORY
Comment: Commenters IV-D-6, IV-D-10, IV-D-12, IV-D-14, and IV-D-17
question the selection of coke by-product recovery plants as a source
category for regulation based on the benzene health risk estimates pre-
dicted at proposal in 1984. The commenters contend: (1) the scientific
basis of the health risk estimates is not sufficient without verification
by monitoring and an epidemiologic study of an exposed community, (2) the
benzene health risk is low compared to other common risks or risks from
other benzene source categories, and (3) the benzene health risk is less
significant than estimated because of the exaggerated exposure assumptions
applied to the risk model.
Response: Specific responses are contained in Chapter 9 to the
commenter's concerns regarding the methodology and assumptions applied to
the quantitative risk assessment for coke by-product recovery plants. The
uncertainties and assumptions associated with the quantitative health risk
assessment also are discussed in the preamble to the proposed rules
(49 FR 23525), in the preamble to the revised proposed rules, and are not
repeated here. As discussed in the preamble to the revised proposal, EPA
determined that control of this source category is warranted to protect the
public health with an ample margin of safety.
3.2 REGULATION OF MERCHANT PLANTS
Comment: Commenters IV-D-4, IV-D-6, IV-D-7, IV-D-10, IV-D-11,
IV-D-12, and IV-D-17 oppose the regulation of merchant plants. The
commenters argue that merchant plants generate fewer emissions compared
to larger furnace plants (or other benzene source categories) and pose
little or no health risk. The commenters also allege that the estimated
costs per merchant plant, the cost per incident of leukemia, and the
3-1
-------
overall economic impacts are higher than predicted and would adversely
impact this industry segment. The commenters also believe that the
merchant plant segment was not represented properly in the BID for the
1984 proposed standards.
Response: In considering the commenters' concerns, the data base
has been revised since the 1984 proposal to indicate the environmental,
health, and cost impacts of controls separately for furnace and foundry
plants. Merchant plants generally fall under the foundry plant industry
segment. As discussed in response to comment 6.2, emission factor
adjustments have been made to account for the lower emission rates
characteristic of foundry plants. Environmental impact estimates
for foundry plants are shown in Appendix A. As discussed in the preamble
of the revised proposal, EPA determined that control is warranted to
protect the public health with an ample margin of safety.
The EPA does not agree that foundry plants were represented
improperly in the BID for the proposed standards. The small-sized model
plant (1,000 Mg/day of coke) remains representative of sites in this
industry segment—both in terms of capacities and processes practiced.
Additionally, the preproposal economic analysis showed the impacts of
control on furnace and foundry plant industry segments.
3.3 EXCLUSION OF FORM-COKE PLANTS
Comment: One commenter in two comments (IV-D-3 and IV-D-8) requests
that the regulation be clarified to exclude form-coke plants. In
support, the commenter cites separate conversations with EPA personnel
who stated that the 1984 proposed standards were not intended to include
form-coke plants because the process does not result in significant
benzene emissions.
Response: In response to the commenter's concerns, the definition
of "coke by-product recovery plant" under Section 61.131 of the 1984
proposed standards has been revised to exclude form-coke plants. As
discussed in correspondence to the commenter on this subject (Docket Item
IV-C-10), this exclusion was not made because of the absence of
significant benzene emissions from the form-coke process. Data are
insufficient to draw this conclusion, although EPA would not expect
3-2
-------
significant emissions based on a review of process description
information.
The EPA's major reason for excluding the form-coke process is that
the form-coke process is too different from the coke by-product recovery
process to apply the standards development study. For example, only one
form-coke plant currently is in operation. This plant does not recover
by-products. Also, the form-coke plant has a fluidized bed process.
Consequently, potential by-product materials are different in chemical
composition. Because of the difference in chemical composition, the
process (and control) equipment also is different from equipment (and
controls) found at plants using the conventional coking process.
3-3
-------
4. SELECTION OF REVISED PROPOSED STANDARDS
4.1 SELECTION OF LEVEL OF CONTROL
Comment: Many commenters indicate that the 1984 proposed standard
should be either more stringent or less stringent. For example,
commenter IV-D-14 supports control levels of 90 percent on process
vessels, tar storage tanks, and tar intercepting sumps, and commenter
IV-D-13 recommends levels of control that provide 100 percent benzene
control regardless of costs. Commenter IV-D-13 supports the most
effective emission reduction techniques for equipment leaks, storage
tanks, and selection of wash-oil final coolers over tar-bottom final
coolers. Selection of wash-oil final coolers (or similar, equivalent
systems) also is recommended by commenters IV-D-5, IV-D-9, and IV-D-15.
Conversely, many foundry coke producers argue that the economic impacts
of the standard as proposed would affect their plants adversely.
Response: On July 28, 1987, the United States Court of Appeals for
the District of Columbia Circuit handed down an _en bane decision in
Natural Resources Defense Council. Inc.. v. EPA, 824 F.2d 1146
(D.C. dr., 1987), hereafter referred to as "Vinyl Chloride", a case
concerning the emission standard EPA set under Section 112 of the Clean
Air Act for vinyl chloride. The Administrator reconsidered the proposed
benzene standard for coke by-product recovery plants in light of the
VlnyJ. Chloride opinion. For his reconsideration, the Administrator used
the revised estimates of nationwide emissions, health risks, cost, and
economic impacts. These estimates were revised after the 1984 proposal
based on the consideration of comments received on the proposal, on
information collected from industry and other sources, and on additional
technical and cost analyses. The specific details of these revisions are
described in Chapters 6 through 9 of this document.
4-1
-------
Because the Administrator followed a new policy approach in the
reconsideration, his selection of the level of the standard is being
published for comment in the Federal Register in a supplemental notice of
proposed rulemaking. The difference between the level of the revised
proposed standard and the original proposal can be found in Chapter 1 of
this document. The Administrator's policy and the rationale for his
decision, as well as the legal framework from the Vinyl Chloride opinion,
are described in the preamble to the supplemental proposal.
4.2 REGULATORY DEFINITIONS OF FOUNDRY AND FURNACE BY-PRODUCT PLANTS
The control option chosen by EPA for the revised proposed standards
would require different levels of control for final coolers and
associated cooling towers at furnace plants than at foundry plants. This
choice necessitated the development of definitions of foundry and furnace
coke and coke by-product recovery plants for the regulation. The EPA
contacted the two industry trade associations, the American Iron and
Steel Institute (AISI) and the American Coke and Coal Chemicals Institute
(ACCCI) to obtain additional technical information regarding these
definitions. The related letters and telephone communications can be
found in Docket A-79-16.
The resulting definition of foundry coke is coke that is produced
from raw materials with an average of less than 26 percent volatile
material by weight per charge/push cycle and that is subject to a coking
period of 24 hours or more. When defining foundry coke by-product
recovery plant, EPA recognized that plants that predominantly produce
foundry coke are typically merchant (non-captive plants). Because of
the fluctuating demand for foundry coke, some of these plants also fill
some orders for furnace coke. The EPA does not intend that these be
classified as furnace by-product plants, since they mainly produce
foundry coke. However, as the percentage of foundry coke increases,
there is a corresponding increase in benzene emissions. One reason is
that furnace coke production is estimated to yield larger quantities of
benzene emissions per megagram of coke produced than foundry coke. Also,
typically more furnace coke can be produced from the same coke oven
battery in a given period of time than foundry coke. The EPA judged that
4-2
-------
a reasonable consideration of these two factors would be to define
foundry plants as producing up to 25 percent furnace coke.
Furnace coke is defined as any coke that is not foundry coke;
similarly a furnace coke by-product recovery plant is one that is not a
foundry coke by-product recovery plant. These definitions avoid any
potential for coke production and by-product plants that are neither
furnace nor foundry.
There are a few independent firms that make close to 50 percent
furnace and 50 percent foundry coke that would be considered furnace
coke by-product plants for this reproposed benzene regulation. The
Agency does not believe that it is necessary to develop a special
category to examine every particular situation when developing national
regulations. However, the economic analysis used company-specific
financial data to the extent possible and modeled these firms as being
merchant plants, rather than captive to steel companies. The analysis
shows no significant adverse economic affects on these companies with
control alternatives that included wash-oil final coolers proposed to
control final-coolers and cooling towers at furnace plants.
4-3
-------
5. EMISSION CONTROL TECHNOLOGY
5.1 DEMONSTRATION OF CONTROL TECHNOLOGY
Comment: Two commenters (IV-D-10 and IV-D-12) claim that gas-
blanketing controls are no longer demonstrated and, consequently, are
unproven. The commenters cite closure of the Armco-Houston plant and
claim that the controls are not demonstrated elsewhere. One commenter
adds that the firm previously designing and constructing the controls no
longer participates in that business, implying that a lack of design and
engineering services impairs "demonstration" of the Armco-Houston system.
Also, commenters IV-D-7 and IV-D-14 allege that EPA conclusions regarding
the system's safety are based on the limited experience at Armco-Houston
and other plants.
Response: The EPA disagrees with these commenters. Not only does
Armco-Houston's closure have no effect on the successful use of gas-
blanketing controls at this plant for the 4-year period prior to closure,
but gas-blanketing systems currently are used at four other plant sites.
The systems used at other plants are described in Chapter 4 of the
BID for the 1984 proposed standards and in the preamble to the proposal
in 49 FR 23530 (see also Docket Items II-B-45, II-B-46, and II-B-47).
Gas blanketing has been used since 1960 in Plant A at Bethlehem Steel,
Sparrows Point. In Plant B, the gas-blanketing system installed during
1954 was replaced during 1978 as part of the conversion to a wash-oil
final-cooler system. In Plants A and B, coke-oven gas from the wash-oil
scrubbers is used to blanket wash-oil decanters, circulation tanks,
collecting tanks, and wastewater storage tanks. Gas blanketing also has
been used since 1960 at the Republic Steel-Cleveland Coke Plant No. 1.
Updated in 1978, the system currently is applied to wash-oil decanters,
circulating tanks, rectifier separators, primary and secondary light-oil
5-1
-------
separators, condensers, and final-cooler circulating tanks. In Coke
Plant No. 2, clean coke-oven gas from the battery underfire system is
applied to primary and secondary light-oil separators, rectifier
separators, and wash-oil circulation tanks. At the four Bethlehem and
Republic Steel sites, gas-blanketing systems were installed initially to
prevent oxidation and sludge formation in light-oil plant lines and
equipment.
At Armco-Houston, four gas-blanketing techniques were applied to
light-oil and tar separation equipment. The system incorporated
blanketing from the gas holder for light-oil recovery vessels, gas
blanketing from the collecting main for tar decanters and a flushing
liquor-collecting tank, negative pressure venting of tar-collecting tanks
to the primary coolers, and gas blanketing from the wash-oil final cooler
(i.e., a slip stream of wash oil containing naphthalene is removed and
routed to a wash-oil decanter tank).
The Armco-Houston system was installed between 1976 and 1977
according to an emission control agreement with the Texas Air Control
Board (TACB). Prior to 1977, natural gas had been used to underfire the
ovens; the coke oven gas was flared with no by-product recovery.
Although the plant had been scheduled for shutdown in 1976, TACB agreed
to continued operation with installation of emission controls. The
system was operated for 4 years with no significant problems until the
plant closed in March 1981. The closure was the result of economic
conditions, not failure of the control system. Although their shutdown
is unfortunate, it does not detract from the proven effectiveness or
viability of the emission control systems employed. Thus, EPA does not
consider that the closure in any way affects demonstration of the
controls or'application of the system at other plants.
One commenter mentions that Koppers1 Engineering Construction
Division (which designed and constructed the Armco system) no longer
engages in that line of business. According to the commenter, this
impairs the "demonstration" of the system. The EPA disagrees. This
company's business decision has no relevance on whether the system has
been demonstrated. Other major engineering design and construction firms
are available for this service. In particular, Oravo/Still Corporation
has designed and installed a positive-pressure gas-blanketing system in
5-2
-------
an existing European coke by-product plant. The system uses clean
coke oven gas (at about 1 inch [2.54 cm] of water positive pressure) to
blanket a variety of storage tanks and process vessels. There are no
domestic installations of this system at present. However, Dravo/Still
has had discussions with at least one U.S. coke plant operator about such
a system for the operator's plant.
5.2 SAFETY, DESIGN, AND OPERATION OF POSITIVE-PRESSURE CONTROL SYSTEM
Comment: Commenters IV-D-3, IV-D-4, IV-D-6, IV-D-7, IV-D-14,
IV-D-17, and IV-D-34 argue that gas-blanketing systems, although appro-
priate and cost-effective for some plants, should not be mandatory at all
sites because of safety, design, and operational concerns. One commenter
states that in some existing plants, redesign of the process operations
and installation of new equipment will be necessary for gas-blanketing
systems to work safely and effectively. Without these changes, the
commenter questions the safety of positive-pressure blanketing systems,
contending that leaks from older pieces of equipment that are difficult
to seal effectively (e.g., tar decanters and tar storage tanks) present a
potential explosion or fire hazard. One of the commenters submitted a
qualitative comparative study of the safety of gas blanketing for one of
their plants. The report concluded that gas blanketing would involve a
significant increase in risk to operating personnel and the surrounding
community. Other commenters argue that leaks from covers, gaskets, and
connections in the piping system pose an explosion danger that is
aggravated by the large number of sources, the presence of electrical
equipment, and the vehicular traffic in areas where blanketing systems
would be installed. Two commenters add that the probability of leaks
(and the associated safety hazard) increases with the long pipe runs
needed at some sites to connect the sources to the system. Other
operational concerns cited by the commenters include the possibility of
naphthalene clogging in cold climates if steam or electrical power for
heated lines were lost and the chance of product contamination (benzene
or light-oil) from the sulfur content of the coke-oven gases.
Response: The safety of recommended control systems should always
be considered, and a system considered inherently unsafe would not be
selected by EPA as a viable control technique. As discussed above in
5-3
-------
response to comment 5.1, gas blanketing has been demonstrated as safe and
effective during an operating period of more than 24 years (1960 to 1984)
at four plant sites in addition to Armco-Houston. In fact, in direct
contradiction to the commenters1 statements, EPA considers that the
proposed system will improve the safety level found in uncontrolled
by-product plant environments. The reasons for this conclusion are
explained below.
Leaks in a negative-pressure system are discussed in response to
comment 5.3, but the AISI argues that even leaks in a positive-pressure
system may allow oxygen infiltration, causing tank vapors to reach
explosive limits and creating a potential safety hazard. The commenter
then cites preamble text in 49 FR 23530 to support this assertion. As
shown below, however, the preamble statement in 49 FR 23530 clearly
refers to the safety and operational advantages of blanketing from the
gas holder, not to the possibility of explosion because of oxygen
infiltration:
One advantage of blanketing with clean coke oven gas
from the gas holder is the elimination of oxidation
reactions between oxygen in the air and organic materials
in the vessels. These reactions often result in a sludge
that may pose fouling and plugging problems in lines and
process equipment. In addition, oxygen infiltration can
cause tank vapors to reach the explosive limits of vapor
when tanks are periodically emptied or when significant
cooling takes place. Applying a positive pressure blanket
would eliminate oxygen infiltration and maintain the vapor
space in the tank above its upper explosive limit [emphasis
added].
The AISI also contends that "the low positive pressure of the
proposed system is insufficient to alleviate explosive conditions if
leaks occur." The standards do not dictate an overall pressure level for
system operation. The system installed may be based on positive or
negative pressure or on a combination of the two. The pressure
maintained will vary by necessity according to the type of source and
location of the connections to the system (i.e., at the main or the gas
holder) and overall process design.
5-4
-------
If, as the commenter asserts, leaks in the system occur or the
positive-pressure blanket fails, the possibility of an explosive
atmosphere forming certainly is no greater than the possibility under
current plant conditions. At most uncontrolled plant sites, explosive
conditions are now present. Liquid organics float on the surface of open
sumps and trenches and leak from equipment components and piping systems
throughout the plant. Organic vapors also are released from "breathing"
tanks as air enters venting systems or through holes in the covers. The
breathing loss is recognized particularly at the light-oil condenser
vent, where a continuous steam purge may operate. In EPA's judgment,
enclosing these sources and ducting the emissions back to the process via
a closed positive-pressure system will reduce substantially the explosion
hazard that now exists. The EPA does recognize that some sources at
existing plants such as tar decanters and tar tanks may be in poor
condition and will require upgrading to accept gas blanketing. The
necessary modifications for typical plants, however, have been reflected
in the cost estimates.
The Agency also reviewed the qualitative assessment submitted by the
commenter to support the contention that gas blanketing would involve a
significant increase in risk to operating personnel and the surrounding
community. However, EPA does not believe that such a conclusion can be
drawn from the assessment for several reasons. First, the assessment is
qualitative; it does not draw quantitative conclusions as to the
frequency of a major system failure. In the comparative risk assessment,
probability ratings were assigned to various hazards within the plant.
For example, for explosion potential under current plant conditions, a
probablity rating of "D" which means "likely to occur 1 time every 10
years" was assigned. With coke gas blanketing, the explosion potential
was reduced to "C" which means "likely to occur every 100 years."
However, with gas blanketing, higher ratings were assigned to the
potential for explosion propagation, on-site safety, and financial loss.
These types of ratings were assigned to various plant operations and to
various control scenarios. The results were weighted and combined to
provide a relative qualitative rating that may be used in evaluating
options in terms of economics and safety.
5-5
-------
Moreover, EPA does not believe the conclusions in the report are
warranted for the following reasons: (1) the report did not utilize a
gas-blanketing design for the plant on which to base a quantitative
comparision; without a specific design; it is not possible to evaluate
safety features that could be engineered into the system, (2) the assess-
ment was based on a review of the existing conditions in the plant,
without consideration of the substantial upgrading of the coke plant
equipment that would be necessary to accommodate installation of a gas-
blanketing system, and (3) the report did not provide any basis or
criteria for assigning the probability ratings or consequence categories
that are reported. After reviewing the assessment, EPA remains convinced
that the upgrading of equipment needed to accommodate gas blanketing,
together with the installation of a control system that is well-designed
with safety features included and that is well operated and maintained
will improve existing safety conditions at the sites.
The EPA recognizes that leaks in a blanketing system will occur
occasionally because of the gradual deterioration of sealing materials.
The prompt repair of these leaks, as required by the standards, not only
ensures proper operation and maintenance of the system but also promotes
safety by eliminating the leak sources. With application of a diligent
leak detection and repair program, the blanketing system will not become
a "network of leaks," as asserted by one commenter. In fact, if the
system is allowed to deteriorate, the owner or operator will likely be
found in violation of the standards.
Other commenters allege that leaks of pressurized gas from the
blanketing system will create a potential explosion hazard around
associated process equipment and that this hazard is aggravated by the
large number of sources, coupled with the presence of electrical
equipment and vehicular traffic in gas-blanketed areas. The EPA's review
of the safety aspects of the proposed system does not support this
contention. Hydrogen and methane are the major components of coke-oven
gas, accounting for 69 to 97 percent of the emission stream. According
to National Fire Code (NFC) guidelines, these lighter-than-air gases
seldom produce hazardous mixtures (i.e., presenting a fire or explosion
danger) in the zones where most electrical connections are made.
5-6
-------
Although special precautions such as explosion-proof electrical
components may be required where light oil or benzene is stored, this
equipment should be already in place at plants where the NFC or plant
safety codes have required their installation. In addition, the authors
of the NFC guidelines state that, in their experience, it generally has
not been necessary to classify as hazardous "locations that are
adequately ventilated where flammable substances are contained in
suitable, well-maintained, closed piping systems which include only the
pipes, valves, fittings, flanges, and meters." The NFC recommends a
common-sense safety approach. The guidelines encourage using a
positive-pressure system, avoiding contact with electrical equipment or
using only intrinsically safe electrical systems with low power needs, or
applying a general purpose enclosure to isolate the leak area (Docket
Item II-C-132).
Two commenters assert that the safety problem increases with the
long pipe runs needed in some cases to connect the sources to the system.
Long pipe runs for coke-oven gas already exist in many plants because the
gas is used as fuel in other areas of the steel plant. The EPA contends
that a long pipe run associated with a coke-oven gas-blanketing system
poses no more risk than even longer pipe runs for transporting the
coke-oven gas throughout the plant.
Prior to proposal of the standard in 1984, EPA thoroughly evaluated
the safety aspects of gas-blanketing systems. This review included
visits to each of the five plant sites with blanketing systems to discuss
safety and operating problems with plant personnel. As discussed in the
preamble in 49 FR 23530, no safety or operation problems were reported
that minimal, routine maintenance would not resolve (Docket Items
II-B-45, II-B-46, and II-B-47). Appropriate safety features also were
evaluated by an independent consultant (Docket Item II-B-49). At the
time of the 1984 proposal, the system included such features as flame
arrestors; an atmospheric vent on the collecting main or gas holder to
relieve excess pressure; three-way valves to lower the possibility of
operator error; and steam-traced lines with drip points, condensate
traps, and steam-out connections (coupled with an annual maintenance
check) to reduce potential plugging problems. Since the 1984
5-7
-------
proposal, additional features have been added such as water drains and
overflow connections for tar tanks and liquid level sampling/gauging
instrumentation with vapor-tight seals. Assuming each system is properly
operated and maintained after installation, EPA considers that the
positive-pressure system is a safe and effective control technique and
that leaks (if repaired as required) do not present the fire or explosion
hazard described by the commenters.
The EPA agrees that a loss of steam or electrical power for heated
lines may cause naphthalene clogging in cold climates. Unless a backup
power supply sufficient for the entire plant is available, EPA assumes
that such a power loss would affect most plant operations and probably
would result in a shutdown until power was restored. Unfortunately, EPA
is aware of no other reasonable approach capable of overcoming the
effects of cold climates.
Nitrogen or natural gas are two other possibilities for substitutes
to coke oven gas. In fact, as described in Appendix B, the use of
nitrogen was costed by EPA for blanketing benzene storage tanks because
of the possibility of contamination. Factors relating to the selection
of blanketing gases for particular types of sources are discussed in the
preamble to the 1984 proposed standard at 49 FR 23530. The revised
proposed standards do not dictate the type of blanketing gas to be used,
however. Thus, nitrogen, natural gas, dirty or clean coke oven gas, or
any other gas can be used as a blanketing medium for any of the affected
sources.
5.3 SAFETY, DESIGN, AND OPERATION OF NEGATIVE-PRESSURE CONTROL SYSTEM
Comment: Commenters IV-D-4, IV-D-6, IV-D-7, IV-D-14 and IV-D-17
argue that, in negative-pressure systems, air infiltration resulting from
ineffective sealing of older vessels, operator error, or equipment
failure also craates a potential explosion or fire hazard. For example,
failure.to close overflow pipes during filling or pumping out of
dehydrators could cause air infiltration in the collecting main. Failure
of the control system when a light-oil tank car is loaded from the
storage tank could cause the vacuum relief valve to function, creating an
explosive atmosphere in the storage tank. Failure of both the control
system and the vacuum relief valve could cause a tank to collapse while
5-8
-------
emptying or to rupture while filling, causing a light-oil spill and
possibly fire. Commenter IV-D-14 also believes that use of a
negative-pressure gas-blanketing system requires additional controls
because of the potential explosion hazard. Specifically, the commenter
states that continuous monitoring of the explosive hazard would be
necessary at three or four locations in the gas distribution system.
Also, an increase in oxygen concentration would require such additional
measures as automatic nitrogen dilution with nitrogen or enrichment with
natural gas to keep the coke-oven gas mixture below the lower explosive
limit (or above the upper explosive limit).
Response: The standards (and associated costs) are based on the use
of a positive-pressure system because preproposal comments questioned the
safety of the negative-pressure system recommended initially. Although
the use or construction of a negative-pressure system is not precluded by
the regulation in any way, EPA encourages companies to install safety
equipment as necessary in accordance with their historical safety
policies and the system's characteristics.
Also recommended is the installation of equipment included in the
costs for the positive-pressure system intended to alleviate many of the
operating concerns cited by the commenters (see response to comment 5.2).
For example, operator failure (on a negative-pressure system) to close
overflow pipes during filling or pumping out of dehydrators can be
avoided by installing an overflow pipe with a liquid seal. The potential
for operator error also can be reduced by installing three-way valves so
that tanks are vented at all times, either to the blanketing system or to
the atmosphere.
The commenters also point to light-oil tank loading operations where
a control system failure (or control system failure concurrent with
failure of a vacuum-relief valve) could lead to an explosion hazard. If
a storage tank is uncontrolled (i.e., open to the atmosphere) as in the
current situation at most by-product plants, such a loading operation
would tend to draw vapors back into the tank. If a tank is controlled by
a negative-pressure system, failure of the control system would cause the
vacuum-relief valve to function, permitting vapors to be drawn into the
5-9
-------
tank. Therefore, EPA considers that negative-pressure system failure
under the scenario suggested by the commenters presents no more danger
than similar situations encountered in the current uncontrolled plant
environment. Failure of the control system implies a pressure swing
within the system. Concurrent failure of a storage tank control system
and vacuum-relief valves would cause vacuum-relief valves on other parts
of the coke-oven gas system to function, drawing oxygen from other
points. Provisions for proper operation and maintenance of relief valves
are included in the standard, however, to minimize the potential for such
a failure.
5.4 MONITORING FOR CARBON MONOXIDE
Comment: Commenter IV-D-14 states that overpressurization of a
positive pressure system poses an explosive and occupational hazard
because of the carbon monoxide (CO) released. The presence of CO
increases costs for additional monitoring and employee training because
CO hazards currently do not exist. Similarly, commenter IV-D-6 states
that additional employees would be necessary for explosive conditions
monitoring or that hydrocarbon detection monitors should be required on
every (emphasis added by commenter) piece of gas-blanketed equipment.
Response: Coke plant operators have stated that pressure control in
the collecting main and gas holder is inherently reliable because large
pressure fluctuations can cause serious operating and safety difficulties
in the operation of the coke-oven batteries and the by-product plant.
Collecting main pressure is controlled by an Askania valve at a few
millimeters of water pressure, and the pressure is often watched and
adjusted manually if necessary. Similarly, the pressure in the gas
holder is also carefully controlled. Overpressurization is prevented by
bleeder or pressure relief valves and water seals.
No costs were added to the recommended gas-blanketing controls for
CO monitoring because the existing and demonstrated systems, installed at
other coke plants, did not have such provisions. Therefore, the
monitoring question appears to be one of company policy and site-specific
conditions. The revised proposed regulations would not require CO
monitoring, but EPA encourages companies to follow their practice of
5-10
-------
safety reviews and implementation of precautions based on each company's
historical experience, its policy, and the site's characteristics.
One additional point to consider is that a CO hazard from coke oven
gas would not be unique to blanketed vessels. Coke oven gas is handled
in many parts of the coke plant, which indicates that a significant
portion of the facility may currently pose a CO hazard. For example,
leaks of coke oven gas routinely occur around the battery proper from
lids, offtakes, doors, charging, and the collecting main. Piping for
coke oven gas winds through the plant, and the gas is treated in enclosed
vessels such as primary coolers, direct-water coolers, and scrubbers.
The gas also is piped to the battery underfiring system and is used in
other parts of the steel plant. The gas-blanketed equipment is required
to be enclosed and sealed and, consequently, should not be more prone to
leaks than other equipment that handles coke oven gas. If a company's
current policy requires detectors and monitors for every point that
contains coke-oven gas, then consistent application of safety policy
would require them for blanketed vessels.
5.5 SUMP CONTROLS
Comment: Two commenters (IV-D-4 and IV-D-17) believe that covering
and sealing sumps create a fire or explosion hazard from concentrated
fumes because no gas or steam can be used for purging. One commenter
states that the purpose of leaving open sumps and trenches is to prevent
such a hazard, and at his plant tramp steam is discharged routinely into
sumps and trenches to reduce the possibility of fire.
Response: Two points are relevant in response to this comment: (1)
steam purging increases emissions of and exposure to hazardous organic
compounds, and (2) alternatives exist to detect and correct hazardous
conditions.
Steam purging strips organic compounds from the sump and can be
especially efficient at removing volatile compounds such as benzene.
Most sumps are installed below grade; consequently, workers and others in
the plant can be exposed to locally high concentrations of these organic
compounds at ground level from an uncontrolled sump, especially with a
purge gas.
5-11
-------
The current practice of discharging tramp steam to an open sump
already poses a hazard if concentrations are high enough to be explosive.
In addition, the steam purging may create the movement of explosive vapor
from the sump to ground level if a surge or slug of organic material
accidentally entered the sump during purging. The EPA's costs include an
air-tight seal and a vent to the atmosphere for safety. The operator may
choose other measures to increase safety, such as including a flame
arrester on the vent or installing detectors for explosive conditions.
Other alternatives are replacing the sump with an above-grade closed tank
that may be easier to keep air-tight, or separating organic compounds
upstream of the sump so the sump will not contain explosive gases.
The solution to the commenter's question will depend on each site's
specific conditions and each company's policy.
5.6 OPERATIONAL PROBLEM FROM PLUGGED VENTS OR VALVES
Comment: Commenter IV-D-17 suggests that mechanical vents and
pressure relief valves may be fouled easily, resulting in ruptured tanks.
The commenter adds that many ruptured tanks occur as the result of
plugged valves that were supposed to relieve pressures.
Response: The EPA recognizes that plugged vents or valves pose
an operational problem and potential safety hazard if not repaired. For
this reason, the revised proposed regulation requires an annual
maintenance check for abnormalities such as plugs, sticking valves, and
clogged or improperly operating condensate traps. A first attempt at
repair of any defect must be made within 5 days, with any necessary
repairs made within 15 days of inspection. The regulation requires that
records containing a brief description of any abnormalities, the repairs
made, and the dates of repair be maintained for a minimum of 2 years.
Although the regulation requires a maintenance inspection only once a
year, plant owners or operators may want to consider performing this
maintenance check more frequently, such as in conjunction with the
semiannual leak inspection.
5.7 CONTROLS FOR BENZENE STORAGE TANKS
Comment: Commenter IV-D-13 requests that EPA determine whether any
benzene storage tanks at by-product plants are equipped with shingle
5-12
-------
seals. If so, the commenter recommends that the regulation require any
shingle seals to be replaced with continuous seals. In support, the
commenter cites the Federal Register notice of withdrawal for benzene
storage tanks (49 FR 8386, March 6, 1984). The notice states in part
that about 12 percent of existing benzene storage tanks in the chemical
and petroleum industries have shingle seals, which are far less effective
than continuous seals.
Response: The shingle and continuous seals to which the commenter
refers are the seals on floating roofs in tanks. Many of the tanks in
by-product plants are horizontal or an older riveted design. The EPA
does not know of any benzene storage tanks with floating roofs in
by-product plants. The controls EPA has analyzed for storage tanks are
wash-oil scrubbers and gas blanketing. These controls are applicable to
horizontal tanks and would not require major tank modification (unless a
tank is in extremely poor condition). The revised proposed standard does
not require control of these tanks, however.
5.8 DETERMINATION OF CONTROL EFFICIENCIES
Comment: Commenter IV-D-9 asks how efficiencies of 90, 95, and 98
percent are determined under the standard.
Response: The 90-percent control efficiency applicable to wash-oil
scrubber controls is based on design calculations. A full description of
the methodology and design parameters is contained in Docket Items
II-B-51 and IV-J-1; a summary description is provided in Chapter 4 of the
BID for the proposed standards. A 95-percent control efficiency for the
tar decanter was derived by adjusting the control efficiency for
enclosure and gas blanketing (98 percent) downward to account for
uncontrolled emissions from the approximately 13 percent of the liquid
surface of the decanter that must remain open to allow clearance for the
sludge conveyor. A 98-percent control efficiency has been established
for gas-blanketing systems and sealed enclosures (e.g., the light-oil
sump). As discussed in the preamble to the 1984 proposed rule in
49 FR 23529, the theoretical efficiency of source enclosure with gas
blanketing approaches 100 percent. However, this efficiency cannot be
expected to be maintained continuously for the service life of the
5-13
-------
equipment because of the eventual deterioration of seals and sealing
materials. Because deterioration of piping, seals, or sealing materials
can occasionally result in leaks, engineering judgment was applied to
reduce the overall control efficiency to 98 percent.
Installation of the specified equipment demonstrates compliance
with the standard for these sources. In other words, these control
efficiencies are assumed to be achieved if the required equipment is
applied. However, design calculations and verifying test data will be
needed if the owner or operator wishes to apply for permission to use an
alternative means of emission limitation.
5.9 GAS BLANKETING VERSUS WASH-OIL SCRUBBERS
Comment: Commenter IV-D-14 recommends that the standards permit the
use of a 90-percent efficient control device (e.g., a wash-oil scrubber)
in lieu of gas blanketing on process vessels, tar storage tanks, and
tar-intercepting sumps. The commenter argues that use of the wash-oil
scrubber would provide essentially the same health benefit as gas
blanketing. Specifically, the commenter suggests that the control
efficiency of blanketing at an older plant may be lower than 98 percent
because of more likely leakage and downtime, and a wash-oil scrubber may
achieve higher than 90-percent control.
Response: The control efficiency of gas blanketing theoretically
is 100 percent. For conservative comparisons with other controls, this
efficiency has been reduced to the value of 98 percent to account for
occasional leakage from seals or sealing materials. Leak detection and
repair requirements are included in the gas-blanketing standards to
ensure that 98 percent control or greater is maintained through proper
operation and maintenance of the equipment. Thus, EPA does not expect
well-designed, well-operated, and well-maintained gas-blanketing systems
to achieve less than 98 percent control efficiency. Although it is
acknowledged in the BIO for the 1984 proposed standards (page 4-28a) that
an efficiency higher than 90 percent (e.g., 95 percent or greater)
theoretically may be achieved, the parameters have been developed to
ensure that all plants using this technique could achieve 90 percent
control continuously. Thus, at proposal, EPA considered gas blanketing
5-14
-------
compared to wash-oil scrubbing on a common basis of conservative
estimates of control efficiencies. Similarly, for the revised proposal,
the Agency believes that a common basis of representative, if somewhat
conservative, control efficiencies should be applied. If the two control
techniques are compared on this basis, wash-oil scrubbers are less
effective than gas blanketing and may be more costly from a nationwide
perspective.
5.10 FINAL COOLERS AND NAPHTHALENE PROCESSING
Comment: Commenter IV-D-15 requests that the standards allow use of
a technology for controlling naphthalene processing and final-cooler
cooling tower emissions that the commenter claims to be more effective
than the tar-bottom final cooler on which the proposed standard was
based. This system, recently patented by his company, eliminates cooling
tower emissions through indirect heat exchangers and reduces (but does
not eliminate) emissions from naphthalene processing by enclosing the
separator and froth flotation units. The commenter estimates benzene
emissions from the processing of naphthalene skimmings at a maximum of
20 grams (g) of benzene/Mg of coke. The commenter claims that this
system achieves a 95-percent benzene emission reduction from baseline
compared to 81 percent for tar-bottom final coolers and has lower
capital, operating, and energy requirements than do wash-oil final
coolers. According to the commenter, use of a single liquid phase
(water) prevents the problem of water and oil emulsion found in wash-oil
final coolers.
Commenter IV-D-9 states that the regulation is unclear regarding
the use of alternatives to the proposed zero emission limit for
naphthalene processing and use of the tar-bottom final cooler. The
commenter's company proposes to convert an existing direct-water final
cooler to a closed-loop recirculated-water final cooling system. This
system would use flushing liquor to cool coke-oven gas and heat exchanger
"closed-to-the-atmosphere" mode of operation. The commenter believes that
this system, in conjunction with gas blanketing the tanks used to hold the
flushing liquor that contains naphthalene, would comply with the zero
emission limit for naphthalene processing in a cost-effective manner.
5-15
-------
The installed cost of this system is estimated at $8 million compared to
$12.5 million for the tar-bottom final-cooling system shown in the 1984
proposal BID.
Response: Both commenters are suggesting the use of alternatives to
tar-bottom final coolers for reducing benzene emissions from final-cool ing
operations at plants. Commenter IV-D-15 proposes a system that will
eliminate final-cooler cooling tower emissions, but that will allow some
emissions from naphthalene processing. Commenter IV-D-9 proposes a system
that eliminates final-cooler cooling tower emissions and also absorbs
naphthalene in tar, thus avoiding the need for physical separation and
processing of naphthalene and its attendant emissions. These comments
plus cost estimates included in the comment letter of commenter IV-D-14
for tar-bottom and wash-oil final coolers led EPA to further investigate
technical and cost data for final-cooler control technologies. Appendix B
contains a discussion for revisions to the control cost estimates since
the 1984 proposal in addition to the revised cost estimates for final
cooling.
Estimated capital and operating costs were requested for the
proprietary indirect cooling system proposed by commenter IV-D-15.
Information also was requested on the degree to which the proprietary
indirect technology had been demonstrated and limitations on its
applicability. When asked about wash-oil final-cooler costs in commenter
IV-D-141s letter, the commenter responded instead with capital cost
estimates for two indirect cooling schemes proposed for application to an
existing U.S. coke by-product plant. Technical information on various
indirect cooling schemes was provided to EPA by Dravo/Still, an engi-
neering firm that designs coke by-product plants and associated control
systems.
The term "indirect" is used in two contexts when discussing final
cooling. One context refers to cooling of the coke-oven gas where there
is no direct contact between the cooling fluid and the coke-oven gas. In
the other context, there is direct contact of the cooling fluid with the
coke-oven gas, but the cooling fluid itself is cooled indirectly. Both
types of indirect final cooling eliminate benzene emissions from the
final-cooler cooling tower that result when direct contact water is cooled
5-16
-------
atmospherically. By the above definition, wash-oil coolers, one of the
control options considered in the 1984 proposal BIO as more stringent than
tar-bottom final coolers, is an indirect final cooler.
All the indirect final-cool ing schemes must in some way deal with
naphthalene remaining in the coke-oven gas just prior to light-oil
recovery. The ways in which naphthalene is condensed, absorbed, dissolved
or otherwise removed from the coke-oven gas prior to or during indirect
final cooling yield varied potentials for benzene emissions from further
handling/processing of naphthalene-containing liquids.
The proprietary indirect cooling technology proposed by commenter
IV-D-15, when used to replace a direct-water final cooler, generates a
liquid stream containing naphthalene that is processed in the same way as
the liquid stream from a direct-water final cooler. The commenter has
suggested enclosing the froth flotation/gravity separation equipment to
reduce benzene emissions, but he makes no suggestion with respect to melt
pit/drying tank emission control. The commenter's proposal would reduce
final-cooler benzene emissions by 70 g/Mg of coke over that required by
the 1984 proposed standard, and it would reduce emissions from
naphthalene-processing from 107 g/Mg of-coke to 28.8 g/Mg (assuming a
90-percent efficient wash-oil scrubber could be applied to the emissions
from the flotation/separation enclosure). Because the 1984 proposed
standard would have required zero benzene emissions from these
naphthalene-processing operations, this scheme would not comply with the
proposed or revised proposed regulation. If the technology recommended by
commenter IV-D-15 is applied to a plant with a tar-bottom final cooler or
mixer-settler (wherein naphthalene is absorbed into the tar), then the
technology would comply with the regulation as proposed in 1984.
Commenter IV-D-15 submitted information that indicated the indirect
cooling technology was tested for a 13-week period using full-scale plant
equipment. This pilot demonstration program yielded enough data to
permit a design of full-scale installation for a plant approximately the
size of Model Plant 3 in the 1984 proposal BID. As of August 1985, a
full-scale installation of this technology was completed in a Canadian
steel plant. The Canadian plant has an existing tar-bottom final cooler,
thus eliminating the need for processing and handling of naphthalene.
5-17
-------
However, the indirect final cooler has not yet been operated for an
extended period. Capital and annualized cost estimates generated from
data submitted by commenter IV-0-15 indicate that the costs for their
proprietary technology are in the range of cost estimates for tar
mixer-settler/tar-bottom final-cooler equipment.
The indirect final-cooling technology suggested by commenter IV-D-9
avoids physical separation and processing of naphthalene by absorbing
naphthalene in tar circulated with the flushing liquor through the final
cooler. Provided the tanks used to accumulate the naphthalene-containing
flushing liquor are gas blanketed or closed to the atmosphere, this
final-cooling scheme would be equivalent to the originally proposed
gas-blanketed tar mixer-settler for naphthalene operations emission
control. It would exceed the mixer-settler's performance with respect to
final-cooler emissions. Information supplied by Oravo/Still indicates
that this type of indirect final-cooler system is in use at the LTV
Aliquippa plant (whether gas blanketing of accumulating tanks is in use
is not known). The capital cost estimate provided by commenter IV-D-9
for his plant is about 70 percent higher than EPA's current estimate for
a wash-oil final cooler applied to that plant size.
Dravo/Still provided information about two other potential indirect
final cooling schemes. One scheme uses warm wash-oil absorption to
remove naphthalene from the coke-oven gas stream in the first stage of
the final cooler. The second stage of the final cooler uses water to
cool the coke-oven gas. This cooling water is itself cooled in an
indirect wet surface air cooler. The warm wash oil, containing naphtha-
lene, is sent to the light-oil still equipment or to a naphthalene
stripper to separate naphthalene from the wash oil. Naphthalene vapors
are returned to the coke-oven gas suction main upstream of the primary
coolers. This recirculation system for naphthalene leads ultimately to
excess naphthalene being accumulated in the recovered tar. The warm
wash-oil absorption system is in use at the Armco-Middletown plant.
However, the Armco plant uses an atmospheric cooling tower for the cooling
water rather than an indirect cooler. The system as described by
Dravo/Still eliminates benzene emissions from both naphthalene processing
and the final-cooler cooling tower. A capital cost estimate for this
5-18
-------
indirect cooling technology applied to an existing U.S. plant was provided
to EPA by commenter IV-D-14. That estimate indicated the capital cost was
about the same as EPA's estimate shown in this document for a wash-oil
final cooler applied to that plant size.
The other indirect final-cooler scheme discussed by Dravo/Still
avoids direct contact between the cooling liquid and the coke-oven gas.
Indirect cooling is achieved in a cross-tube cooler with water flowing
through the tubes and gas flowing outside the tubes. To prevent naphtha-
lene fouling of the heat exchanger surface, tar is injected into the
cooler on the gas side where it mixes with condensing water and keeps
naphthalene in suspension. The water-tar-naphthalene mixture withdrawn
from the cooler is recycled to the collecting main. As in the above
system, excess naphthalene leaves the by-product plant in the recovered
tar. The cooling water has not contacted the gas stream, so it may be
cooled in an atmospheric cooling tower without generating benzene
emissions. This system also eliminates benzene emissions from both
naphthalene processing and final cooling. According to Dravo-Still, this
type of indirect final cooler is in use at a Dofasco plant in Hamilton,
Ontario. Commenter IV-D-14 provided a capital cost estimate for this
system applied to an existing U.S. plant. That estimate indicated that
the capital cost would be about 28 percent higher than the current EPA
estimate for a wash-oil final cooler applied to the same plant size. One
reason that the cost of this indirect final-cooler system may be higher
than the one described above is that it is difficult to make use of
existing plant equipment in retrofitting the latter system.
Based on the information and data presented above, all of the
indirect final-cooling schemes except that of commenter IV-D-15 achieve
equivalent benzene emission control for naphthalene processing and
handling operations when compared to the tar-bottom final cooler or tar
mixer-settler required by the proposed regulation. All of the described
indirect final-cooling schemes produce greater final-cooler benzene
emission reductions than would be achieved through installation of
tar-bottom final coolers or tar mixer-settlers.
Based on the revised environmental, risk, and cost data, the
Administrator has decided to repropose wash-oil final-cooler technology
5-19
-------
as the basis for the naphthalene-processing standard for furnace plants.
Wash-oil and tar-bottom final coolers have been applied at several plants.
However, any system (including the ones discussed here) that meets the
zero emission limit for naphthalene processing could be used in foundry
plants, and any system that meets both the zero emission limit for
naphthalene processing and for final coolers (and associated cooling
towers) could be used at furnace coke plants.
5-20
-------
6. ENVIRONMENTAL IMPACTS
6.1 DATA BASE FOR ENVIRONMENTAL IMPACTS
Comment: Commenters IV-D-9 and IV-D-14 state that the health risk
from by-product plants is less significant than projected at proposal
because nationwide benzene emission estimates are overestimated as a
result of the effect of plant closures and reduced battery capacities.
One commenter estimates nationwide benzene emissions to be 21,800 Mg/yr
compared to the 24,100 Mg/yr estimate in the preamble (49 FR 23525).
Response: The interim status of the estimated environmental impacts
was acknowledged in the preamble to the 1984 proposed standards in
49 FR 23524. As stated, the impacts were calculated initially from a
data base of 55 plants. Industry data and information from the U.S.
Department of Energy (DOE) received prior to proposal indicated that 13
of the 55 plants had been closed. Information was not available,
however, to determine whether all reported closures were permanent.
Consequently, the preamble presented environmental impacts based on 42
plants and stated that the impacts and calculations in the BID would be
revised following proposal.
The data base has been revised since the 1984 proposal in several
respects. Information regarding permanent closures, changes in battery
capacities, and changes or corrections in site-specific operating
processes have been applied to reflect the industry operating status as
of November 1984. These data were supplied by individual companies and
by the two major industry trade associations—the American Iron and Steel
Institute (AISI) and the American Coke and Coal Chemicals Institute
(ACCCI). As discussed below in response to comment 6.2, adjustments to
emission factors also have been made since the 1984 proposal to indicate
lower emission rates from sources at plants producing foundry (or furnace
6-1
-------
and foundry) coke. These revisions have been made to account for the
combination of lower light-oil yields and lower benzene concentrations
for foundry coke plants compared to concentrations for furnace coke
plants. For this reason, the data base also has been segregated to show
separately as well as combined the environmental impacts of control
options on furnace and foundry plant industry segments.
Tables A-1 and A-2 (Appendix A) reveal a potential total of 44
furnace and foundry plants with a combined operating capacity of 50.9
million Mg/yr of coke. Of the 44 plants, 30 produce furnace coke, and 14
(mainly merchant plants) produce foundry (or furnace and foundry) coke.
Of the 30 furnace plants, 6 are on cold-idle as of November 1984. These
plants have been identified as follows: (1) LTV Steel--Thomas, Alabama;
(2) LTV Steel—E. Chicago, Indiana; (3) U.S. Steel—Fairless Hills,
Pennsylvania; (4) U.S. Steel—Lorain, Ohio; (5) U.S. Steel--Fairfield,
Alabama; and (6) Wei rton Steel—Brown's Island, West Virginia. Also, 1
of the 14 foundry plants (Alabama By-Products—Keystone, Alabama) is on
cold-idle as of November 1984 . Because information is insufficient to
predict whether these temporary closures will become permanent, these
seven plants have not been deleted from the data base used to estimate
the environmental impacts of the revised proposed standards. The
deletion of six furnace plants from the data base would reduce the
operating capacity of this industry segment from about 45.8 million Mg/yr
of coke to about 39.2 million Mg/yr of coke nationwide. Foundry plant
operating capacity would be reduced by about 8 percent (from about 5.1
million Mg/yr to about 4.7 million Mg/yr) if the cold-idle plant were
excluded from the data base for this industry segment.
Tables A-3 and A-4 (Appendix A) display the operating processes
practiced at each furnace or foundry site, as reported by the individual
plants (Docket Items IV-D-1 through IV-D-18). Based on these data,
half (15 of 30) of the furnace plants practice naphthalene handling and
processing, a major source of benzene emissions. Direct-water final
coolers and tar-bottom final coolers are used at 16 and 4 furnace plants,
respectively; 5 furnace plants use wash-oil final coolers. Although tar
recovery sources (e.g., tar decanters, dewatering, sumps, and storage)
are found at all 30 sites, as are most light-oil plant sources (light-oil
6-2
-------
decanters, sumps, storage, and wash-oil circulation tanks), BTX storage
is practiced at only 10 sites, and benzene is stored only at 4 sites.
Table A-4 indicates that naphthalene handling and processing also is
practiced at half (7) of the 14 foundry plant sites. Reported data show
direct-water final coolers at seven plants, tar-bottom final coolers at
two plants, and no wash-oil final coolers in use. Although tar recovery
sources are present at each site, light-oil storage is found at 9 of the
14 sites. Benzene and BTX are stored at one plant.
Tables 7-1 through 7-6 of the BID for the 1984 proposed standards
have been revised to show the updated estimated environmental impacts.
These tables display the estimated baseline nationwide benzene emissions
and process capacity data for sources at the 30 furnace plants, the 14
foundry plants, and the total industry combined. Comparable data for
total VOC emissions (benzene and other VOC) also are shown. Based on
these data, estimated nationwide benzene emissions from the industry
total nearly 26,000 Mg/yr; VOC emissions are about 171,000 Mg/yr.
The effects control options have on reducing benzene and total VOC
emissions also are shown in Appendix A. Implementing the revised
proposed standards would reduce overall benzene emissions from furnace
and foundry coke producers from approximately 26,000 Mg/yr to about
2,000 Mg/yr, a reduction of about 93 percent. Nationwide VOC emissions
also from these sources would be reduced from approximately 171,000 Mg/yr
to about 6,000 Mg/yr.
The revised data base and foundry plant emission factors have little
effect on the impacts or benefits of other environmental considerations
associated with the final standards, such as energy requirements, water
pollution, solid waste disposal, and noise or odor levels. As discussed
in the preamble to the 1984 proposal notice (49 FR 23525), a nominal
increase in electrical or steam requirements at furnace plants could
occur if gas blanketing piping were heated to prevent vapors from
condensing or freezing in vent lines. Tables A-ll and A-12 show energy
use and coke-oven gas recovery estimates for model furnace and foundry
plants.
Although no water pollution problems are associated with recycling
benzene vapors, implementing the revised proposed standards could result
in an increased HCN concentration at plants using indirect final cooling.
6-3
-------
As discussed in the BIO for the 1984 proposed standards (page 7-7), HCN
is emitted currently from the final-cooler cooling tower at some plants
by air stripping of wastewater. Measured HCN air emissions and
calculations based on once-through cooling water indicate that about 200
g/Mg of coke could be added to wastewater for treatment, if indirect
cooling rather than direct cooling were used (Docket Item II-B-30). The
actual amount of additional HCN in the wastewater could depend on cooling
water temperature, degree of recycle practiced, and numerous other
factors.
As suggested by the commenters, the effects of reduced operating
capacities and revised emission factors have been taken into account in
the updated risk assessment. Further information regarding the calcu-
lation of revised emission estimates for furnace and foundry plants
is discussed below in response to comment 6.2.
6.2 FOUNDRY PLANT EMISSION FACTORS
Comment: Commenters IV-D-6, IV-D-7, IV-D-10, IV-D-11, and IV-D-12
claim that operating dissimilarities result in fewer emissions compared
to emissions from furnace coke plants. The commenters state that foundry
plants generate fewer emissions because of: (1) the use of less volatile
coal in their feed (21 to 22 percent volatile matter in foundry blends
versus 28 to 30 percent volatile matter in furnace coal blends), and (2)
the use of longer coking cycles (28 to 30 hours for foundry coke versus
14 to 16 hours for furnace coke). In support, one commenter also states
that the percentage of benzene in light oil at his plant is 55 to 60
percent, considerably less than the 70-percent example shown in Table 3-6
of the BID for the 1984 proposed standards. Another commenter maintains
that merchant plants generate fewer emissions than furnace plants not
only because of different operating practices but also because of the
relative size of the industry segment compared to furnace coke plants.
Response: In response to the public comments received on this
issue, EPA has reviewed available information and data to determine
whether the development of separate emission factors for foundry and
furnace coke production is warranted.. Based on results of this review,
6-4
-------
EPA agrees with the commenters' contention that benzene emissions from a
foundry coke by-product plant would be expected to be less than the
emissions from a furnace coke by-product plant of similar capacity.
Because no emission measurements were performed in foundry coke plants
during the 1979 to 1980 sampling survey, appropriate emission factor
adjustments have been made based on available data for light-oil yields.
Foundry coke is produced from a coal mixture that generally contains
less volatile matter than the mixtures used to produce furnace coke. The
ACCCI comments suggest that typical furnace-coke coal mixes contain 28 to
30 percent volatile matter and foundry-coke coal mixes contain 21 to 22
percent volatile matter. This statement is confirmed, in part, by data
contained in one primary reference source on the coking of high- and
low-volatile coals (Docket Item IV-J-5) that show light-oil yields
as significantly lower (less than half) for the low-volatile coals.
However, definitive data on light-oil yields published by the DOE show
that, over a 4-year period, the light-oil yields in merchant coke by-
product plants (mostly foundry coke producers) averaged about 66 percent
of those in furnace plants on a per-ton-of-coal-charged basis (Docket
Items II-I-43, II-I-50, IV-J-2, IV-J-3, and IV-J-4). These yields are
shown in Table A-13 of Appendix A. Table A-13 also provides data on the
relative yields of tar and coke-oven gas in merchant coke plants compared
to furnace coke plants. The data displayed in Table A-13 represent the
principal basis for the technique used to adjust the proposed emission
factors for foundry coke producers.
Based on a review of data contained in another coke-making reference
source (Docket Item II-I-2), EPA also agrees with commenters who suggest
that the lower coking temperatures associated with foundry coke produc-
tion compared to furnace coke production (for the same coal) would lead
to production of less by-product benzene. In support, one merchant plant
commenter indicated that the light oil from foundry coking contains 55 to
60 percent benzene compared to the 70 percent assumed in the 1984 pro-
posal BID (Docket Item IV-D-7). Based on an informal poll of some member
companies, ACCCI provided an average estimate of 63.5 percent for foundry
producers (Docket Item IV-D-7). For furnace coke production, however, a
benzene content for light oil of 70 percent is still considered appro-
priate.
6-5
-------
Separate emission factors for foundry plant sources have been
developed by applying correction factors to the emission factors ini-
tially proposed for both furnace and foundry plants. These changes do
not affect the revised emission factors as applied to furnace plants.
The computations of correction factors are shown in Table A-14 of Appen-
dix A; the final emission factors for furnace and foundry plants are
shown in Table A-15.
For plants that produce only foundry coke, benzene emission factors
for light-oil recovery plant sources (i.e., wash-oil decanters, wash-oil
circulation tanks, light-oil condensers, and light-oil sumps) have been
adjusted by a correction factor of 0.54. This adjustment factor combines
the effects of lower light-oil yields, lower benzene concentrations in
the light oil, and different coal-to-coke ratios. Physically, the re-
duced emission estimates may be viewed as a result of lower benzene
throughput in the foundry coke by-product plants.
For sources treating or handling water that has contacted the coke
oven gas (i.e., flushing liquor circulation tank, excess ammonia-liquor
storage tank, direct-water final-cooler cooling tower, tar-bottom final-
cooler cooling tower, and naphthalene handling/processing), benzene
emissions are expected to be proportional to the ratio of benzene in the
coke oven gas (i.e., partial pressure and partitioning between the liquid
and gas). The light-oil-to-coke-oven-gas ratios in Table A-14 are indi-
cative of the partitioning. These ratios are multiplied by relative
benzene concentrations in the light oil to yield a correction factor of
0.73 for the above sources in foundry coke plants.
Emissions from storing or processing liquids containing tar (i.e.,
tar decanters, tar-intercepting sumps, tar storage tanks, and tar-
dewatering tanks) also are expected to be proportional to the ratio of
benzene in coke oven gas. In addition, the relative yield of tar (i.e.,
the amount of tar exposed to the benzene) is expected to affect the
partitioning of benzene between the tar and the gas. Therefore, the
correction factor applied reflects both the relative quantity of benzene
produced, and the proportion of that benzene transferred to the tar,
ultimately available for dissolution in the sources. Combining the tar
yield, light-oil-in-gas ratio, benzene concentration in light oil, and
6-6
-------
coal-to-coke ratio factors has produced a correction factor of 0.47 for
the above source emissions in foundry coke plants.
Equipment leaks from fugitive emission sources (i.e., pumps, valves,
exhausters, sampling connection systems, open-ended valves and lines, and
flanges or other connectors in benzene service) are expected to emit ben-
zene emissions proportional to the benzene concentration in the fluids
handled. The correction factor applied to emission estimates for foundry
coke plants is based on the estimated benzene content of the light oil at
a foundry coke plant. When the estimate supplied by ACCCI is used, the
correction factor is 0.91. Table A-15 indicates the revised uncontrolled
emission factors for furnace and foundry plants. Table A-16 shows the
derivation of revised foundry plant benzene fugitive emission rates from
VOC emission factors.
Emission estimates incorporating revised foundry plant emission
factors and other data base revisions are discussed above in response to
comment 6.1. The revised emission estimates were incorporated into the
estimated impacts on which the selection of the revised proposed standard
is based. The selection of the revised proposed standard is discussed in
the Federal Register preamble for the revised proposal.
6.3 MODEL COKE PLANTS
Comment: Commenter IV-D-4 argues that benzene emission estimates
for model coke plants are not representative of emissions from an actual
small plant (coke capacity of 440 Mg/day). The commenter estimates
uncontrolled emissions from a medium-sized model plant (4,000 Mg/day of
coke) at 1,080 Mg/yr; with the 1984 proposed controls (and assuming 89
percent recovery), remaining uncontrolled emissions of 120 Mg/yr would
result. This estimate excludes certain emission sources (e.g.,
direct-water or tar-bottom final-cooler tower, tar-dewatering tanks, and
benzene or BTX storage tanks). The commenter compares 120 Mg/yr to 70
Mg/yr (uncontrolled) for the small plant. In summary, the commenter
argues that small merchant plants should not be regulated because of the
low emission level.
Response: In essence, the commenter argues that small merchant
plants should not be regulated because of the low benzene emission levels
6-7
-------
compared to estimated emissions from medium-sized model plants. In
support, the commenter suggests that his calculations show uncontrolled
emissions at his plant site as less than emissions after control at a
larger site. The EPA believes that the commenter has misconstrued the
purpose of model plants and their role in the EPA decisionmaking process.
First, EPA's decision to regulate is not based on model plant emission
estimates and their level compared to larger model facilities. The EPA's
decision regarding to what level to control foundry plants is discussed
further in the preamble to the revised proposed standard.
The purpose of constructing model plants is to portray typical
facilities in terms of size and processes representative of the industry.
However, the nationwide emission estimates have been based on site-
specific process data rather than by model plant extrapolation. In the
preproposal analysis, model plant size parameters were selected based on
the approximate distribution of actual plant capacities as a function of
coke capacity. This distribution indicated that 25 of 55 plants produce
between 300 and 2,000 Mg/day of coke, accounting for 17 percent of
domestic capacity. Consequently, a small model plant was defined as a
facility producing 1,000 Mg/day of coke, slightly less than the midpoint
of the actual range.
The industry operating data updated after proposal indicate foundry
plant capacities at existing sites ranging from 130,000 to 617,000 Mg/yr.
The commenter says his plant's capacity is 440 Mg/day. When converted to
an annual basis (approximately 161,000 Mg/yr), the commenter's plant
remains well within this size range. The EPA considers that the Model
Plant 1 (small) size range remains representative of small plant
operations. In terms of actual onsite processes, further discussions
with the commenter revealed that the commenter's plant does operate a
direct-water final cooler (the commenter's statements indicated that no
direct-water final cooler was present at the plant). Also present are a
light-oil plant, fugitive emission sources, and most tar separation
vessels. Benzene and BTX storage are not practiced. Again, although the
process parameters identified with Model Plant 1 may not reflect all
actual operations at the commenter's site, they typify small plant
processes on the whole.
6-8
-------
Other changes to the data base also have been made since the 1984
proposal to reflect more closely foundry plant operations. These changes
include lower emission factors and recovery credits. The EPA believes
that these adjustments better reflect differences between furnace and
foundry plants. The changes would tend to lower the site-specific
estimate suggested by the commenter. However, the inclusion of a
direct-water final cooler also would need to be considered by the
commenter when recalculating actual benzene emission estimates for this
site.
6.4 EMISSION FACTORS FOR TAR-RELATED SOURCES
Comment: Commenters IV-D-7, IV-D-10, and IV-D-12 question the data
base used to estimate emission factors and their resulting industry wide
applicability for tar decanters, tar dewatering, and flushing liquor
circulation tanks.
Response: The commenters argue, in essence, that test data for
certain sources are not sufficient to take into account variance in
emissions because of differences in the methods of operation and other
factors. The EPA certainly agrees that differences in methods of
operation, operating parameters, and design features are evident from
plant to plant and will influence actual emissions from each source.
During development of the estimated emission factors, these variations
have been taken into account to the extent possible by averaging appli-
cable measurements to obtain a factor representative of a "typical"
source. Also, the emission factors have been adjusted since the 1984
proposal for improved applicability to plants producing foundry coke.
Specifically, the commenters state that data (12 tests) supporting
the tar-decanter emission factor are not sufficient for industry wide
application because emissions are sensitive to variability in gas-liquid
separator residence time and optional heating. As discussed in Chapter 3
of the BID for the 1984 proposed standards (page 3-10), typical residence
times are about 10 minutes for liquor and about 40 hours for tar. Op-
tional heating tends to increase the total benzene emitted even though
the concentration of benzene per unit volume of emissions may be reduced.
The degree of separation achieved is highly variable because of coal type
and differences between plants. As stated above, adjustments have been
6-9
-------
made since the 1984 proposal to account for differences between furnace
and foundry plants.
The tar-decanter emission factor (applicable to furnace plants) is
based on three measurements for each of four vents (12 total) at two tar
decanters at two plants (Bethlehem Steel—Burns Harbor, Indiana, and
Bethlehem Steel—Bethlehem, Pennsylvania). The tests spanned a flow rate
range of 50 to 275 std ft^/min. The benzene emission rate measured at
the Pennsylvania steel plant was 1.2 kg/hr (Docket Item II-A-22). This
decanter was one of two for a coke battery. Emissions from the two
decanters were assumed to be twice the emissions from the single
decanter, or 2.4 kg/hr. The corresponding benzene emission factor for
this decanter was calculated as 84.7 g/Mg coke. One of three tar
decanters was tested at the steel plant in Indiana (Docket Item II-A-25),
where the average benzene emission rate from three vents on the decanter
was 4.4 kg/hr. The corresponding emissions for three decanters at this
Indiana plant are 13.3 kg/hr, which yields a benzene emission factor of
69.6 g/Mg of coke. The average benzene emission factor from these two
plants was 77.2 g/Mg of coke. Consequently, the emission factor was
designated as 77 g of benzene/Mg of coke. The EPA considers this data
base of 12 measurements adequate to estimate the average level of
emissions from typical decanter vessels under varying conditions.
The commenters also maintain that the tar-dewatering emission factor
should not be applied industry wide because emissions depend on the
method of operation. In support, the commenters point to the "unex-
plained" variations in the range of emission factors for this source
(9.5 to 41 g/Mg of coke).
Emissions from tar-dewatering tanks were evaluated at three plants
(see Docket Items II-A-26, II-A-27, and II-A-28). Three measurements
were made for each of two vents at one plant; one measurement was made at
the second plant. At the third plant, one test was made at the tar
storage tank where dewatering was performed. The EPA considers that
these measurements, as averaged, are sufficient to provide a reasonable
estimate of emissions from a typical source.
The extent and effect of the variation in dewatering emission
factors have been discussed in the BID for the 1984 proposed standards
(page 3-16), which states as follows:
6-10
-------
The emissions data for tar dewatering at the first plant showed
higher emissions from the west tank (3.2 kg/hr) than from the east
tank (1.1 kg/hr). These tanks are operated in series rather than in
parallel, and the wet tar enters the west tar dehydrator first.
Consequently, the emissions from the west tar dehydrator are
expected to be higher than emissions from the east tar dehydrator.
The daily benzene emission rates from the two tar-dewatering tanks
at this first plant were 27 and 76 kg, respectively. Daily benzene
emissions from tar dewatering at the second plant were 43 kg. The
tar is dewatered in storage at the third plant, where benzene
emissions were 24 kg/day. The benzene emission factors from these
three plants were 41, 9.5, and 12.9 g/Mg of coke, respectively.
These were averaged to obtain a benzene emission factor for tar
dewatering of 21 g/Mg of coke.
The tar-dewatering tanks contained tar with 200 to 2,000 ppm benzene
in the liquid. Tar, as collected from the flushing liquor and the
primary cooler, can contain greater than 0.2 percent benzene or
2,000 ppm at a rate of 40 kg/Mg of coke produced. The maximum
potential for benzene loss from tar dewatering and storage
calculated from these values is greater than 2,000 ppm at a rate of
40 kg/Mg of coke produced. The maximum potential for benzene loss
from tar dewatering and storage calculated from these values is
greater than 80 g/Mg of coke. The benzene emissions from tar
dewatering and storage probably will be less than 80 g/Mg of coke
and will depend on the method of operating these processes.
The commenters also question the adequacy of test data from the
primary cooler condensate tank as the basis for the flushing liquor
circulation tank emission factor and its resulting applicability industry-
wide. The emission factor for the flushing liquor circulation tank
(9 g/Mg of coke) was obtained from one test in which emissions from a
primary cooler condensate tank were measured (Docket Item II-A-13). This
tank was assumed to be similar to a flushing liquor circulation tank
because both vessels function to hold liquor taken from the gas stream
during early stages of gas processing. Although it is desirable to have
more than one test measurement as the basis of the estimated factor,
engineering judgment suggests that the measurement is a reasonable value
6-11
-------
for emissions from flushing liquor circulation tanks. The EPA agrees
with the commenter, however, that emissions will vary necessarily de-
pending on the number and geometry of tanks, the number of vents, and
other factors.
6.5 METHODOLOGY FOR EMISSION FACTORS
Comment: Commenter IV-D-33 contends that the EPA VOC and benzene
emission factors are not applicable to the sources at his site and also
that VOC emissions should be calculated using a different methodology.
Response: The EPA developed emission factors to obtain an estimate
of the nationwide emissions of benzene and VOC from by-product plant
process operations. The EPA is aware that site-specific factors could
cause the actual emissions from a particular facility to vary from the
estimates based on EPA emission factors. In fact, the commenter's
estimate of benzene and VOC emissions from his facility using his
alternate set of emission factors is within 20 to 30 percent of the
emissions estimated when using EPA emission factors. This difference is
within the range of uncertainty for the emission factors.
The commenter also proposed an alternate methodology for developing
VOC emission factors using EPA test data that would, in general, would
tend to make them lower. The commenter states that EPA overestimates VOC
emissions by assuming that all components of light oil will volatilize to
the same extent as benzene. The EPA agrees that this may lead to an
overestimate of VOC emissions; however, the EPA believes the commenter's
approach underestimates VOC because not all components of the light-oil
vapor were explicitly measured in the test program whose results were used
to develop the VOC emission factors. In any case, the cumulative
difference between emissions estimates for the commenter's plant using the
commenter's methodology were shown to be within the range of uncertainty
for the emission factors, as noted above.
6.6 VOC BENEFITS FOR OZONE REDUCTION
Comment: Commenter IV-D-2 supports the proposed standard for
by-product plants, particularly when applied to a plant site located in
his State. This commenter believes that the estimated emission
6-12
-------
reductions are realistic and provide the added benefit of helping his
State reduce the VOC inventory in the Baltimore ozone nonattainment
area.
Response: The EPA thanks the commenter for his support. The
estimated VOC emission reductions also will benefit the country as a
whole in reducing ozone formation.
6-13
-------
7. COST IMPACT
7.1 REVISIONS TO COST ANALYSIS
Comment: Commenters IV-D-9 and IV-D-14 argue that the capital costs
of the proposed equipment are $50 million to $100 million or more,
compared to the estimated cost of $23.8 million. According to these
commenters, the true costs exceed model plant estimates by 50 to 100
percent at some facilities. In support, the commenters cite the
following major factors contributing to 1984 EPA estimates: (1) low
estimates of unit material costs and construction expenses, (2) site-
specific factors such as equipment conditions and pipeline length, (3)
EPA's reliance on cost-estimating references rather than experience and
price quotations from local suppliers and contractors, (4) the dollar
year of the estimates (1982), and (5) additional costs for work in
hazardous areas requiring special safety precautions. One commenter
provides for EPA review an example of these points using estimates
prepared by National Steel, Armco, and by United Engineers for a
Bethlehem Steel plant. Another commenter (IV-D-33 and IV-D-34)
submitted information on cost estimates for controls at his plant; the
commenter contends that the capital costs would be higher than the model
unit costs in the 1984 proposal BID.
«•
Response: To consider the commenters' concerns, EPA conducted a
detailed review of the United Engineers' estimate for the Bethlehem plant
of Bethlehem Steel, Bethlehem Steel's estimate for their Sparrows Point
plant, and the Armco and National Steel cost data; EPA also had a cost
estimate prepared by C. R. S. Sirrine, Inc., a third-party design
engineering firm. The details of EPA's final analysis are shown in
Appendix B. Included in the review was a site visit to the Bethlehem
plant to resolve questions regarding equipment locations, and the sources
subject to the proposed emission controls, and to obtain examples of
7-1
-------
site-specific conditions pertinent to the development of revised unit
cost factors.
As shown in Appendix B, the revised cost analysis includes higher
unit costs for most materials, which affects the costs estimates for most
sources. The revised unit costs were composed from the data received in
the comments and the cost data developed by Sirrine. The revised analy-
sis also includes costs for sealing all sources, installation of roofs on
certain storage tanks, more pipe supports, pressure/vacuum relief valves
for sealed sources, and adjustments to unit cost factors for work in
hazardous areas requiring special safety precautions.
The commenters1 criticism of EPA for reliance on cost-estimating
references is valid. The EPA agrees that it is desirable to base cost
estimates on previous experience and site-specific factors. However, in
the absence of abundant experience, engineering and construction firms
use those same references to develop cost estimates. Backup information
requested by EPA to support the commenters1 cost estimates indicated that
such was the case. Also, the preference for site-specific information
must be compromised somewhat when attempting to develop within schedule
and budgetary restraints nationwide cost impacts for 44 plants. The EPA
acknowledges that costs for particular plants may be higher or lower than
EPA estimates, depending on site-specific conditions. However, the
revised cost analysis addresses concerns cited by the commenters and the
costs are reasonable estimates of the industry wide cost of controls.
Using the revised cost analysis, the nationwide capital cost of the
revised proposed standards is approximately $84 million (1984 dollars),
compared to estimated capital costs of about $24 million (1982 dollars)
af proposal in 1984. Of the $84 million, approximately half is due to
the inclusion of wash-oil final coolers, the cost of which was not
reflected in the 1984 proposal estimate.
7.2 REVISIONS TO PRODUCT RECOVERY CREDITS
Comment: Commenters IV-D-6 and IV-D-14 state that the value of
potential product recovery credits has been overestimated. In parti-
cular, commenter IV-D-6 states that the value of light oil for small
plants is overstated. One commenter explains that the assumption that
the recovered product can be used as plant fuel or sold is not valid
7-2
-------
because, when the production of coke oven gas is greater than demand for
potential fuel consumption, the excess gas is flared. One commenter
states that for one plant (Lackawanna) no product recovery credit can be
assumed and for another plant (Bethlehem) the credit should be reduced.
The excess gas is flared at Bethlehem, and Lackawanna now has no
steelmaking operation creating fuel demand.
Response: The EPA essentially agrees with the commenters that the
value of potential product recovery credits was overestimated in the 1984
proposal. As discussed further in response to comment 6.2, foundry
plants produce less light oil than larger furnace plants do. This
difference in production quantity (reflected in new emission factors for
foundry plants) has been taken into account in the computation of revised
fuel value and light-oil recovery credits.
In response to the commenters1 concerns, a telephone survey of seven
(three furnace and four foundry) plants was conducted to determine the
extent of flaring excess coke-oven gases (IV-E-9). Briefly, EPA found
that this flaring is not generally practiced except as a last resort. Of
the five foundry plants surveyed, only one (Empire Coke) flares gas
continuously. Of the remaining four foundry plants, two do not flare
excess gas at all, and two plants flare only seldomly. Of the three
furnace plants surveyed, one plant never flares, one plant flares only in
emergency situations, and one plant flares occasionally in periods of low
demand. Consequently, the adjusted credits have been applied to most,
but not all, plants. No fuel credits were applied to the Lackawanna
plant of Bethlehem Steel and Empire Coke for the reasons cited by the
commenters.
The value of potential product recovery credits also has been
adjusted since that time to reflect 1983 data published by the DOE
(Docket Item IV-J-4). Based on these data (Table A-ll), the credit for
light oil has been decreased from $0.33/kg to $0.27/kg light oil. The
fuel value recovery credit for coke oven gas also has been adjusted
downwards--$0.14/kg coke oven gas compared to $0.15/kg estimated in the
1984 proposal.
7-3
-------
7.3 ECONOMIES OF SCALE FOR SMALL PLANTS
Comment: Commenters IV-D-4, IV-D-6, IV-D-7, and IV-D-17 maintain
that small plants should not be regulated because of the disproportionate
cost impact resulting from the lack of economies of scale compared to
moderate or large plants, coupled with higher per-unit control costs.
One commenter notes that, although control costs for small plants are the
same as for medium-sized plants, the costs in relation to production are
200 to 400 percent higher; another commenter indicates that small plant
costs are 900 percent higher than for medium-sized plant costs. The
commenters point to the use of a cost model based on a moderate to large
plant with a number of economies of scale in terms of the number of
control units per ton of production. According to one commenter, this is
reflected in Section 8.1.5 of the BID for the 1984 proposed standards,
where actual costs are compared to estimated costs for two large plants
with economies of scale.
Response: As described in the response to comment 7.1 and Appendix
B, the capital and annualized cost estimates for control of benzene
emissions have been revised. The basis for estimating these revised
costs are the three original model plants, sized at 1,000, 4,000, and
9,000 Mg/day of coke. New capital and annualized costs of control were
estimated for these plants, then cost functions (equations relating cost
to plant production capacity) were developed for each process. These
cost functions do provide for economies of scale, and they adequately
represent the costs of control from the smallest foundry coke by-product
plant to the largest furnace coke by-product plant.
7-4
-------
8. ECONOMIC IMPACT
8.1 REGULATORY BASELINE
Comment: Commenters IV-D-6 and IV-D-7 state that the economic
analysis for the proposed standards fails to consider the true state of
the coking industry at baseline and that the economic impact will have an
adverse effect on the industry. In support, one commenter notes that the
analysis does not take into account the plant closures and capacity
reductions that have occurred since 1980. Both commenters also note that
the baseline does not include the cost of other environmental regulations
incurred by 1983. New regulations include final iron and steel effluent
guidelines, National Pollutant Discharge Elimination System (NPDES) per-
mit upgrading, State implementation plan (SIP) compliance rules (includ-
ing reasonably available control technology [RACT], lowest achievable
emission rate [LAER], and new source review of coke plant rebuilding),
and the pending coke-oven battery NESHAP.
Response: At the time the original analysis was conducted, the
information from published and unpublished sources was current. A
reanalysis has been conducted (see Appendix C) that utilizes data on
plants and capacity in existence in November 1984. Financial data and
production data used in baseline estimates are from the available
published and unpublished sources as of 1984. A discussion of industry
trends as of 1984 is provided in Section C.I.6 of Appendix C.
The baseline of the reanalysis assumes companies meet regulations
existing in 1984, including OSHA rules for coke oven emissions; State
regulations related to desulfurization, pushing, coal handling, coke
handling, quench tower, and battery stack controls; and Best Practicable
Technology (BPT) and Best Available Technology (BAT) water regulations.
All of these were due to be in effect by 1983 at the latest. Other
regulations that are pending or have not reached the deadline date for
8-1
-------
compliance are not likely to be a part of 1984 production costs for
firms, or they will have little effect on those costs.
8.2 SELECTION OF DOLLAR YEAR
Comment: Commenter IV-D-7 suggests that the economic impact
analysis should be in 1986 dollars because the project schedule places
promulgation closer to regulation in 1986.
Response: Selection of the year for dollar values in analysis is
somewhat irrelevant because conversions may be made for any current or
past year based on gross national product (GNP) implicit price deflators.
The important values are the baseline data from which regulatory impacts
are determined. For these values, current information (in 1984 dollars)
was used to produce realistic results in the reanalysis.
Projection to future year-dollars is difficult primarily because of
the confounding effects of inflation. Prediction of inflation rates is
beyond the scope of this analysis. The 1984 dollar values in the
reanalysis are best updated for future timeframes when those years are
current so that GNP implicit price deflators accounting for actual
inflation may be used.
8.3 POTENTIAL ECONOMIC IMPACT
Comment: Commenter IV-D-14 states that the economic impacts of the
proposed standards are more severe than estimated and will have an
adverse economic impact on the industry. In support, the commenter cites
examples from a recent Price Waterhouse "Steel Segment" survey for the
period 1979 through the third quarter of 1983 to illustrate the overall
financial condition of steel companies. The following major factors are
cited: (1) the steel industry is depressed and suffers capital formation
problems; (2) the period analyzed shows a rising debt-to-equity ratio,
with declining stockholder equity; (3) investment exceeded cash from
operations; and (4) the industry experienced $6 billion in losses between
1982 and 1983.
Response: The measure of severity of impacts is best made relative
to some reference value rather than from the standpoint of absolute
values. In the reanalysis, capital costs of compliance are compared to
8-2
-------
average annual net investment averaged over the period from 1979 to 1983
(converted to 1984 dollars) for individual companies. Table C-25 in
Appendix C of this document shows these comparisons for furnace coke.
For furnace coke plants, capital costs of compliance for Regulatory
Alternative II range from 0 to 3 percent of net investments. For
Regulatory Alternative III, these costs range from 0 to 5 percent of net
investments. The regulatory alternatives are outlined in Table C-l of
Appendix C.
The industry trends noted by the commenter are discussed in Section
C.I.6 of Appendix C. Companies have made adjustments through mergers,
acquisitions, and creative financing measures to generate investment
funds. The fact that, as the commenter states, investment exceeded cash
from operations indicates that capital is available for investment even
for firms sustaining losses.
Although the industry is having some capital difficulty, the burden
of regulation will differ from firm to firm. The net investment analysis
indicates that in no case will the cost of regulations be a significant
burden.
8.4 ESTIMATED EMPLOYMENT IMPACT
Comment: Commenter IV-D-7 questions the estimated employment
impacts of the proposed standards. The commenter suggests that the
estimates should include total plant employment because by-product plants
cannot be separated. This commenter employs 36 people in his by-product
operation, but he employs a total of 268 persons in his coke plant.
Response: The commenter's argument is answered in Tables C-23 and
C-28 in Appendix C, which show the employment effects of the regulatory
alternatives in the furnace and foundry coke plants, respectively. These
are industry totals. For the furnace coke sector, neither regulatory
alternative results in a loss of more than 0.5 percent of baseline jobs
at furnace coke plants for the entire industry. This is not a sub-
stantial loss, and it should be weighed against the benefits.
For the foundry coke sector, employment impacts are calculated for
two scenarios. Scenario A assumes that foundry coke producers do not
compete with imports in the domestic market, and Scenario B assumes they
do. Under Scenario A, the regulatory alternatives result in job losses
8-3
-------
that are less than 1.0 percent of baseline foundry coke employment.
Under Scenario B, employment losses for the industry are less than
3.2 percent of baseline. Again, these losses are not large.
It is possible that unemployment will not occur as a result of the
regulations for two reasons. First, workers may be reallocated within
the industry to perform other tasks because of labor contracts or other
constraints. Second, there are potential employment gains from the
regulations such as labor to operate and maintain control equipment.
This labor is included in the cost analysis, but it is not evaluated in
terms of added jobs. These gains may offset estimated job losses.
8.5 IMPORT TRENDS
Comment: Commenter IV-D-7 states that the regulation will increase
the trend of importing coke. The commenter cites Table 9-1 of the
1984 proposal BID, which shows a growing coke-importing trend since
1974, and Table 9-2, which shows a decrease in domestic production.
Response: Data up through 1983 indicate that imports have been
decreasing since 1979 (see Table C-2 in Appendix C). Trends in the steel
industry away from coke-using processes and toward decreased steel
production overall are the most likely sources of this decrease.
The reanalysis indicates that the regulatory alternatives result in
a slight reversal of this trend. Table C-22 in Appendix C shows that
furnace coke imports will increase by 9,000 Mg/yr under Regulatory
Alternative II and by 25,000 Mg/yr under Regulatory Alternative III.
These represent increases from the baseline of 0.23 percent and 0.64
percent, respectively. These are negligible changes in imports.
8.6 ECONOMIC IMPACT ON SMALL PLANTS
Comment: Commenter IV-D-6 maintains that the economic impact
assumptions for integrated, captive producers compared to small merchant
foundry plants are dissimiliar; these differences should result in
separate regulations. The commenter states that "foundry producers,
unlike captive producers, cannot distribute costs among operations,
cannot adjust the price of coke oven gas or light-oil used elsewhere in
the facility, and cannot increase the price of other by-products." The
8-4
-------
additional costs to foundry plants result in a direct increase in product
price, which may give advantages to foreign competitors.
Commenter IV-D-7 argues that, for the same reasons, small plants
will incur a disproportionate economic impact. This commenter also cites
Table 9-40 of the economic analysis, which estimates a coke price
increase ranging from 6.4 to 15.4 percent for small plants to comply with
baseline.
Response: A distinction is made between furnace and foundry plants
in the BID analysis and in the reanalysis for the revised proposal. Most
furnace coke producers are captive, and most foundry producers are
merchant. This distinction allows the analysis to examine impacts
separately.
The differences between furnace and foundry producers expressed by
the commenter do not necessarily result in a worsened competitive
situation for foundry firms with respect to other firms in the foundry
industry. In the reanalysis, no foundry batteries become uneconomic
(candidates for closure) under either regulatory alternative. This
implies that industry impacts of regulation will not be concentrated on
any one plant sufficiently to force it out of business.
Tables C-24 and C-29 in Appendix C show the capital costs of com-
pliance of the regulatory alternatives for furnace and foundry producers.
For both regulatory alternatives, the foundry coke producers' share of
total capital costs of compliance is less than 16 percent. For indivi-
dual foundry coke-producing firms, Table C-30 in Appendix C shows that
the capital costs of the regulatory alternatives amount to no more than
11 percent of net investment in for firms for which data were available.
This is not substantially higher than the maximum share of capital costs
of net investment for furnace coke producers (see Table C-25 of Appendix
C). Furnace coke producers face additional pressures because of the
difficulties being experienced in the steel industry, which composes the
market for furnace coke.
The influence of imports in the foundry coke industry is accounted
for under Scenario B of the reanalysis. A worst case bound is assumed so
that quantity reductions in domestic production are assumed to be offset
by quantity increases in imported coke sold domestically. The changes
are 61,000 Mg/yr under Regulatory Alternative II, and 94,000 Mg/yr under
8-5
-------
Regulatory Alternative III. These represent 2.1 percent and 3.2 percent
of foundry coke demand, respectively. Advantages gained by foreign
competitors because of the regulatory alternatives are small.
In the reanalysis, price impacts under Scenario A for foundry coke
producers are $0.99/Mg for Regulatory Alternative II and $1.46/Mg for
Regulatory Alternative III. These represent 0.58 percent and
0.86 percent increases from baseline (see Table C-27 in Appendix C).
Under Scenario B, no price impacts will result. No significant impacts
are projected for foundry coke producers because of these price changes.
8.7 PRICE IMPACTS
Comment: Commenter IV-D-11 states that the economic analysis is
inaccurate in predicting the increased price of coke for merchant plants.
This cornmenter estimates an increase in the price of coke at his plant of
$1.38/Mg versus $0.24/Mg estimated at proposal in 1984 (49 FR 23525).
This estimate is based on the commenter's estimate of the cost of
compliance at his facility (capital costs of $1.8 million versus average
cost of $408,500 cited in the 1984 proposal BID; annualized costs of
$80,000 versus $70,500 cited in the 1984 BID). The commenter notes also
that his capacity is 681 Mg/day rather than 1,362 Mg/day.
Response: The determination of changes in the price of coke must be
made on a market basis rather than a pi ant-by-plant basis. The price
changes are due to shifts in the entire supply curve, as well as the
effects of the marginal plant at equilibrium for the entire market. The
economic impact model uses this basis for its computation.
The reanalysis calculates capital costs of compliance, annualized
compliance costs, and price changes based on capacity information avail-
able in November 1984. Capital costs of compliance for furnace and
foundry plants are given in Tables C-24 and C-29 of Appendix C, respect-
ively. Annualized compliance costs are shown in Table C-31 for furnace
coke producers and Table C-32 for foundry coke producers. Tables C-21
and C-27 show price effects of the regulatory alternatives on furnace and
foundry coke producers. These costs differ from engineering estimates
because of the calculation of costs based on batteries with
8-6
-------
marginal cost of production at or below price, rather than all
batteries.
For furnace coke, the average capital cost per plant is
approximately $1.0 million for Regulatory Alternative II and $1.7 million
for Regulatory Alternative III. The average annualized cost per plant is
$87,500/yr for Regulatory Alternative II and $310,000/yr for Regulatory
Alternative III. Price increases are $0.13/Mg (a 0.12-percent increase)
for Regulatory Alternative II and $0.36/Mg (a 0.33-percent increase) for
Regulatory Alternative III.
For foundry coke, the average capital cost per plant is
approximately $636,000 for Regulatory Alternative II and $1.1 million for
Regulatory Alternative III. Average annualized cost per plant is
$118,000/yr and $264,000/yr for Regulatory Alternative II and Regulatory
Alternative III, respectively. The price increase associated with
Regulatory Alternative II for foundry coke is $0.99/Mg (a 0.58-percent
increase from baseline), and, for Regulatory Alternative III, the price
increases by $1.46/Mg (a 0.86-percent increase from baseline).
The average values may not reflect actual costs for individual
plants. They serve as indicators of the neighborhood of costs a plant
may be expected to face in complying with the regulatory alternatives.
8.8 ECONOMIC IMPACTS ON FOUNDRY PLANTS
Comment: Commenters IV-D-10 and IV-D-14 state that merchant plants
should not be regulated because of the adverse economic impact on the 13
plants comprising this industry segment. The commenters disagree that
no merchant plant will close as a result of the 1984 proposed standards.
One commenter predicts the closure of three entire merchant plants
because of the estimated costs of compliance. An added impact of these
closures is the metal casting industry's dependence on 1.4 million tons
of foundry coke production.
Response: The estimated annual compliance costs for foundry plants
computed in the reanalysis are presented in Table C-32 of Appendix C,
and the estimated capital compliance costs are shown in Table C-29.
Average annual plant compliance costs are $118,000/yr for Regulatory
Alternative II and $264,000/yr for Regulatory Alternative III. Average
8-7
-------
capital costs of compliance for foundry coke plants are $636,000 for
Regulatory Alternative II and $1.1 million for Regulatory Alternative
III.
In terms of net investments for companies, capital costs of com-
pliance are relatively small. For firms for which data are available,
capital costs amount to no more than 11 percent of net investment for
either regulatory alternative (see Table C-30 of Appendix C). This does
not imply an excessive capital burden because of the regulatory
alternatives.
In the reanalysis, two scenarios for the foundry coke industry are
evaluated. Under Scenario A, foundry coke producers are assumed to
supply all of the domestic coke market so that supply shifts induced by
the regulatory alternatives result in slightly higher prices and
slightly reduced production (see Table C-27). In all cases, changes in
price and quantity produced are less than 1.0 percent of baseline
values.
Scenario B assumes that foundry coke producers must compete with
foreign producers in the domestic coke markets. As a worst case, foreign
coke is assumed to be available at a price equal to baseline, and that
price is assumed to remain constant regardless of changes in the domestic
market. Furthermore, imported coke is assumed to be a perfect substitute
for domestic coke so that, for any reduction in domestic production,
consumers will purchase amounts of imported coke equal to the reduction.
Under this scenario, there is no price change because of the regulatory
alternatives. The quantity changes shown in Table C-27 indicate that
domestic production will decrease by 61,000 Mg/yr under Regulatory
Alternative II and by 94,000 Mg/yr under Regulatory Alternative III.
Import increases by these amounts reflect 2.1 percent and 3.2 percent of
domestic demand, respectively.
Under either scenario, the metal casting industry is unlikely to
suffer. Under Scenario A, if price and quantity changes do occur, they
will not be substantial. Under Scenario B, domestic coke reductions will
be offset by increased availability of imported coke.
Even if the entire shortfall in domestic production is compensated
by increased imports, domestic foundry coke producers are unlikely to be
8-8
-------
significantly impacted. No closures from the regulatory alternatives are
predicted under either scenario. Other impacts such as employment are
unlikely to be substantial, as shown in Table C-28 of Appendix C.
8-9
-------
9. QUANTITATIVE RISK ASSESSMENT
9.1 USE OF MODEL FOR HEALTH RISK ESTIMATES
Comment: One commenter (IV-D-14) states that EPA's prediction of
the leukemia risk to the community is overstated because of the linear,
nonthreshold extrapolation model. Other commenters suggest that, by
mathematically predicting benzene exposures in the vicinity of the coke
by-product recovery facilities and consequential risks, EPA may be
estimating values that really do not exist (IV-D-6, IV-D-10, IV-D-12).
These commenters suggest that EPA (1) monitor benzene near these
facilities to verify the model and (2) conduct epidemiologic studies of
the communities surrounding the facilities.
Response: Because a specific environmental carcinogen is likely to
be responsible for at most a small fraction of a community's overall
cancer incidence and because the general population is exposed to a
complex mixture of potentially toxic agents, it is currently not possible
to directly link actual human cancers with ambient air exposure to
chemicals such as benzene. Today's epidemiologic techniques are not
sensitive enough to measure the association. Direct measurement of
health effects or estimation of a causal relationship to chemical
exposure through community health studies usually is not possible due to
the limited statistical sensitivity of such studies and the presence of a
large number of potentially confounding variables (e.g., general health
status, occupational exposure, smoking, diet, migration, age, etc.).
Therefore, EPA must rely largely upon mathematical modeling techniques to
estimate human health risks. These techniques, collectively termed
"quantitative risk assessment," are the means whereby the risk of adverse
health effects from exposure to benzene in the ambient environment can be
estimated mathematically; effects found at higher occupational exposure
levels are extrapolated to lower concentrations characteristic of human
9-1
-------
exposure in the vicinity of industrial sources of benzene. The analysis
estimates the risk of cancer at various levels of exposure. A unit risk
estimate for benzene is derived from the dose-response relationship
observed in the occupational studies. The unit risk estimate represents
the cancer risk for an individual exposed to a unit concentration of a
carcinogen [e.g., 1 part per million (ppm)] for a lifetime.
Although EPA agrees that the linear, nonthreshold model is
conservative in nature and would tend to provide a reasonable upper bound
to the statistical range, the Agency does not agree that the assumptions
upon which it is based are unreasonable or that the results of its use
are exaggerated. The dose-response mathematical model with low dose
linearity is used by EPA because it has the best, albeit limited,
scientific basis of any of the various extrapolation models currently
available. The EPA has described the scientific suppositions underlying
the preference of the linear, nonthreshold model over other mathematical
models (Water Criteria Documents Availability, 45 FR 79319, November 28,
1980). In this notice EPA stated:
There is really no scientific basis for any mathematical
extrapolation model which relates carcinogen exposure to cancer
risks at the extremely low levels of concentration that must be
dealt with in evaluating the environmental hazards. For practical
reasons, such low levels of risk cannot be measured directly either
using animal experiments or epidemiologic studies. We must,
therefore, depend on our current understanding of the mechanisms of
carcinogens for guidance as to which risk model to use. At the
present time, the dominant view of the carcinogenic process
involves the concept that most agents which cause cancer also cause
irreversible damage to DNA. This position is reflected by the fact
that a very large proportion of agents which cause cancer are also
mutagenic. There is reason to expect that the quanta! type of
biological response that is characteristic of mutagenesis is
associated with a linear non-threshold dose-response relationship.
Indeed, there is substantial evidence from mutagenesis studies with
both ionizing radiation and with a wide variety of chemicals that
this type of dose-response relationship is also consistent with the
9-2
-------
relatively few epidemiological studies of cancer responses to
specific agents that contain enough information to make the
evaluation possible (e.g., radiation-induced leukemia, breast and
thyroid cancer, skin cancer induced by aflatoxin in the diet).
There is also some evidence from animal experiments that is
consistent with the linear non-threshold hypothesis (e.g., liver
tumors induced in mice by 2-acetylaminof1uorene in the large scale
ED01 study at the National Center for Toxicological Research, and
initiation stage of the the two-stage carcinogenesis model in the
rat liver and mouse skin) (45 FR 79359).
With regard to the need for epidemiologic study of the population
residing in the vicinity of the coke-oven by-product recovery plants, it
must be kept in mind that current methods are not sufficiently sensitive
to detect a causal association between chronic, low-level benzene expo-
sure and cancer. Such studies are complicated by a number of potentially
confounding factors. These factors include genetic diversity, population
changes and mobility, lack of consolidated medical records, lack of
historical benzene exposure data over each individual's lifetime,
community exposure to other carcinogens besides benzene, and the latency
period of cancer.
In the evaluation of benzene emissions from coke oven by-product
recovery plants under Section 112 of the CAA, EPA has followed a policy
in which the nature and relative magnitude of health hazards are a
primary consideration. In the absence of scientific certainty,
regulatory decisions must be made on the basis of the best information
available. In the case of benzene, EPA has evaluated the potential
adverse effects associated with human exposure based on the best
scientific information currently available. For benzene, this is
represented by the occupational epidemiologic studies demonstrating a
causal association between exposure and leukemia.
9.2 SELECTION OF RISK MODEL
Comment: One commenter (IV-D-13) suggests that, in using HEM, and
not the Industrial Source Complex (ISC) model, to estimate annual average
ground level concentrations of benzene around coke-oven by-product
recovery plants, EPA has underestimated exposure to the population living
9-3
-------
near those facilities. The commenter alleges that EPA has admitted the
model underestimated exposure from 200 to 300 percent in the benzene
fugitive emission rulemaking. Therefore, the commenter states that risk
to the most exposed individuals should be much higher. On the other
hand, commenter IV-D-14 states that EPA's assumption in the item that
individuals are exposed to the maximum annual ground-level concentration
of benzene for 24 hours/day, 365 days/year for 70 years are unrealistic
assumptions that lead to exaggerated risk calculations.
Response: Commenter IV-D-13 (NRDC) raised these same concerns in a
petition to the Administrator of EPA to reconsider four final benzene
decisions as published in the Federal Register (49 FR 23478,
June 6, 1984). The EPA responded to these concerns in EPA's response to
the NRDC petition (50 FR 34144, August 23, 1985). The EPA reviewed
NRDC's concerns about correcting the alleged bias in the assessment used
in evaluating the benzene fugitive emission standard. In order to test
the sensitivity of the regulatory decisions to changes in the exposure
assessment, EPA recalculated the exposure assessment used in the benzene
fugitive emission decision by increasing the ambient concentrations and,
therefore, exposure by 300 percent. A factor of 300 percent was used
because it is the upper limit to the alleged underestimation of exposure
based on the analysis presented in Appendix C of the Benzene Fugitive
Emissions Background Information for Promulgated Standards and detailed
in Docket A-79-27, Item IV-B-18. After doing so, EPA concluded that the
standard would not change based on the new exposure assessment. More-
over, the HEM does not always predict lower concentrations than the ISC;
it is dependent on the source-specific assumptions. In addition, EPA
does not know whether the ISC would be a better predictor of ambient
benzene concentrations around coke by-product plants. Because of these
considerations, EPA decided that this additional analysis for coke
by-product plants was not warranted.
The EPA recognizes that the assumption of continuous exposure to
the maximum annual concentration over a 70-year lifetime may tend to
overestimate the maximum individual lifetime risk. In addition, for coke
by-product plants, the assumption that the plants operate continuously at
full capacity for 70 years is likely to overestimate the risk. On the
9-4
-------
other hand, some assumptions may tend to underestimate the risk. For
example, there may be more susceptible subgroups than the population from
which the unit risk estimate is derived. Such susceptibility can differ
with infirmity, age, genetic composition, or immune-incompetancy. For
these individuals, the cancer risk may be underestimated. The model
assumes terrain is flat; for plants in complex terrain where the surroun-
ding topography is at higher elevation than the emission sources, the
model may possibly underestimate maximum annual concentration. There-
fore, overall the Agency believes the leukemia risk estimates are
plausible, if conservative.
9.3 UNIT RISK ESTIMATE
Comment: Commenter IV-D-13 contends that the benzene unit risk
estimate used in the 1984 proposal has not been updated since 1981 and,
therefore, did not take into consideration recently published scientific
reports on benzene carcinogenicity. The commenter maintains that such an
update would increase the unit risk estimate 15 times. Therefore, EPA is
underestimating risk to the population residing near coke by-product
recovery plants.
Response: On October 17, 1984, the commenter (NRDC) petitioned the
Administrator of the EPA to reconsider four final decisions regarding
benzene emissions as published in a Federal Register notice on June 6,
1984 (49 FR 23478). Of central relevance to the petition was the con-
tention that the health risk assessment relied upon in June was outdated
and that the risk estimate should be revised to reflect the most current
literature on benzene carcinogenesis. The EPA agreed to a current review
of the published literature and reevaluated the unit risk estimate for
benzene accordingly. The methodology for the evaluating of the unit risk
estimate is described in a document titled Interim Quantitative Cancer
Risk Estimates Due to Inhalation of Benzene (Docket OAQPS 79-3[l]
VIII-A-4) and is summarized in EPA's response to the NRDC petition
(50 FR 34144, August 8, 1985). In the revaluation of the unit risk
estimate, EPA pooled the leukemia responses observed in the retrospective
epidemiologic studies of rubber hydrochloride workers exposed to benzene
(Rinsky et al. 1981) and chemical manufacturing workers exposed to
benzene (Ott et al. 1978), then EPA computed a geometric mean of each
9-5
-------
point risk estimate. The data were aggregated to encompass a range of
plausible risks observed by independent investigators of benzene exposure
in different occupational settings. The leukemia incidence observed in a
third epidemiologic study (Wong et al. 1983) of benzene exposure in
chemical manufacturing was used as a comparison to the computed risk
estimates of the pooled studies. The resulting ratio between these two
sets of data was used to adjust the computed mean estimate. Based on
these calculations, the unit risk estimate (the probability of an
individual contracting leukemia after a lifetime exposure to a benzene
concentration of one part benzene per million parts air) was revised
upwards from 0.0223/ppm (6.9 x 1CT6 per ug/m3) to 0.026/ppm (8.0 x 1CT6
per pg/m3). The revised estimate represents a 17-percent increase in the
estimate used in the June 1984 decisions.
The significant gap between EPA1 s revised risk estimate (a
17-percent increase) and the fifteenfold increase recommended by NRDC
results from a major policy difference on the appropriate use of animal
versus human data. The increase advocated by NRDC is obtained by relying
exclusively on the incidence of preputial gland tumors of male B6C3F mice
Although the results of an animal bioassays have been considered in the
Agency's revaluation, EPA believes that the unit risk estimate for
inhalation of benzene is appropriately based on the principal epidemio-
logic studies because these studies are of recognized quality and have
the greatest relevance in the estimation of health risks for the general
population. Well-conducted epidemiologic studies provide direct evidence
of a causal link between the chronic exposure to benzene and leukemia.
This direct evidence precludes the biological uncertainties inherent in
extrapolating animal data to humans. Given the wide range of levels of
benzene exposures and durations of exposure, the epidemiologic studies
showed a threefold to twentyfold increase in risk of leukemia above
individuals not exposed to benzene. These findings present unequivocal
evidence that chronic inhalation of benzene causes leukemia in humans and
therefore falls within the criteria of EPA's current guidelines for
carcinogen risk assessment (51 FR 33992, September 24, 1986). Although a
clear dose-response association between carcinoma and benzene exposure
was demonstrated in rodent bioassays, the EPA believes that human data,
9-6
-------
when available, should be the principal factor in the derivation of a unit
cancer risk estimate. In the case of benzene, EPA believes that the
animal data are appropriately used qualitatively to buttress the
conclusion regarding benzene's carcinogenicity.
9.4 DERIVATION OF UNIT RISK ESTIMATE
Comment: One commenter (IV-D-14) expresses the opinion that the
benzene unit risk estimate overstates the true risk by at least one order
of magnitude. Moreover, a minor adjustment of 17 percent in the unit risk
estimate published in the June 6, 1984, Federal Register notice as
response to public comments on the listing of benzene (49 FR 23478) did
not adequately respond to the criticisms made during the maleic anhydride
proceeding. According to the commenter, the principal criticism not
addressed concerned the inclusion of the Ott et al. 1978 study in the
derivation of the unit risk estimate. The commenter maintains that the
study should not have been used because the leukemia incidence was small,
and there was a likelihood of exposure to other chemicals. In addition,
the commenter feels that EPA inappropriately reclassified one of the
deaths in the Ott study as myelogenous leukemia even though the cause of
death on the death certificate was listed as pneumonia.
Response: The EPA has previously responded to these concerns in the
response to public comments concerning the regulation of benzene as a
hazardous air pollutant (49 FR 23478, June 6, 1984). Although EPA does
not view the Ott et al. (1978) study, taken alone, as conclusive evidence
of an association between low level (2 to 9 ppm) occupational exposure to
benzene and leukemia, the Agency believes that this work, combined with
other findings in the published benzene health literature, serves to
reinforce the public health concerns regarding benzene exposure. Ott et
al. observed 3 cases of leukemia in a cohort of 594 chemical workers when
only 0.8 case was expected. This represents an excess risk of leukemia of
3.75. The EPA does not believe that omitting from the study the
individual who suffered from leukemia but died of pneumonia would be an
appropriate change. In view of the recognized causal relationship between
benzene and nonlymphatic leukemias, EPA believes that a case of
myelogenous leukemia such as this, if documented, should not be ignored.
9-7
-------
The EPA does not view the extent of confounding exposures in the Ott
et al. study as severe. The authors did exclude from their analysis
persons known to have been exposed to levels of arsenicals, vinyl
chloride, and asbestos, all of which have been associated with human
cancer. This exclusion eliminated 53 persons from consideration
including one leukemia victim. The remaining substances, which include
the suspect carcinogen vinylidene chloride, have not been shown to be
associated with a leukemia risk in either man or animals. Thus,
inclusion of such exposed individuals in the cohort would not be likely
to affect the target endpoint for benzene exposure (leukemia) in terms of
increased risk.
9.5 COMPARATIVE RISK FROM GASOLINE MARKETING
Comment: Several commenters argue that benzene emissions from
sources other than coke-oven by-product recovery plants present a greater
risk to exposed populations and, therefore, should warrant the full
resources of EPA (IV-D-10, IV-D-12, IV-D-17). They argue that gasoline
service stations and other segments of the gasoline marketing industry
present far greater risk to residents living near those facilities than
do coke-oven by-product recovery plants.
Response: The EPA agrees that there are sources of benzene
emissions into the ambient air other than coke oven by-product recovery
plants. The EPA has evaluated many of the industrial sources of benzene
(49 FR 23558, June 6, 1984). In addition, the EPA has concluded an
extensive analysis of benzene and gasoline vapor emissions from the
gasoline marketing industry, such as service stations, vehicle
refueling operations, bulk plants, and bulk terminals. On August 19,
1987, the EPA Administrator issued a notice of proposed rulemaking to
control refueling emissions from gasoline-fueled light duty vehicles, and
to control the volatility of gasoline (52 FR 31162). These proposed
standards will help protect the general public from the risk of cancer
due to exposure to benzene, a component of gasoline vapor, and to
evaporative gasoline as a whole. This proposal is estimated to reduce
benzene emissions from gasoline refueling by about 90 percent from
uncontrolled levels.
9-8
-------
As described in the preamble to the revised proposed standard, the
Administrator determined that control of benzene emissions from coke
by-product recovery plants is warranted to protect the public health with
an ample margin of safety.
9.6 COMPARATIVE RISK FROM OTHER SOURCES
Comment: A commenter (IV-D-14) states that the risk to benzene
exposure from coke by-product plants does not seem high when compared to
other risks that are accepted as commonplace in society. The commenter
suggests that the average leukemia risk for the entire population
exposed to benzene emissions from these facilities is 7 x 10~8 (or 7 in
100,000,000). Examples of commonly accepted risk were given, e.g.,
smoking one pack of cigarettes per day is a risk of cancer of 5 x lO'3.
Response: The EPA does not average the maximum individual lifetime
cancer risk calculations, but it assumes an aggregate of risk to the
population residing within a radius of 50 km around coke by-product
plants. Aggregate risk is a summation of all the risks to people
estimated to be living within the 50 km radius of the facility. The
aggregate risk is expressed as incidences of cancer among all of the
exposed population after 70 years of continuous exposure to predicted
ambient concentrations of benzene emitted from the facilities; for
convenience, the total is often divided by 70 and expressed as cancer
cases per year. On the other hand, maximum lifetime risk reflects the
probability of contracting cancer to those individuals exposed
continuously to the estimated maximum ambient air concentration of
benzene for 70 years. The nationwide risk to the exposed population
residing near coke by-product recovery plants due to the plant emissions
has been calculated to be about 3 cases of leukemia per year. The
maximum lifetime risk is estimated to be 6 x 10~3.
The EPA does recognize that most human activities and events
involve some degree of inevitable risk. However, the Administrator has
judged that quantitative estimates of the risk from other activities
should not be used as quantitative benchmarks for making decisions on
hazardous air pollutants.
9-9
-------
9.7 SELECTION OF BENZENE VS. POM
Comment: One commenter (IV-D-7) suggests that polycyclic organic
matter (POM) compounds result in a higher health risk than benzene
emissions, and that EPA has not chosen to regulate POM emissions.
Response: The Agency examined the information regarding benzene and
POM as two different cases. The POM decision by EPA was based on several
factors, including the great uncertainty as to the magnitude of the cancer
risk to the public, the fact that many POM source categories are being
controlled under programs to attain and maintain the national ambient air
quality standard (NAAQS) for particulate matter, and difficulties in
devising control programs for source categories not well-regulated (e.g.,
existing woodstoves, forest fires, and agricultural burning). The EPA
concluded that a more appropriate regulatory strategy would be to regulate
specific POM source categories (e.g., coke oven emissions, new woodstoves,
and diesel cars and trucks).
9.8 CONSIDERATION OF OTHER HEALTH EFFECTS
Comment: Commenter IV-D-13 states that EPA's health impact analysis
based on "cost-benefit" is flawed because: (1) the analysis includes only
one of benzene's hazardous effects (leukemia), (2) EPA has ignored data
showing public health danger greater in degree and broader in kind than
included in the risk assessment, and (3) the assessment makes no attempt to
account for concurrent control of other suspected carcinogens (e.g.,
toluene and xylenes).
Response: The EPA does recognize there are other health effects
associated with human exposure to benzene. These effects are summarized
in a recent review of the health literature by the Occupational Safety
and Health Administration (OSHA) (52 FR 34460, September 11, 1987). The
toxic effects of benzene on the hematopoietic system include not only
myelogenous leukemia, but also aplastic anemia. Aplastic anemia is a
rare, and often fatal, disorder characterized by cytopenia in the
peripheral blood and in the bone marrow. Aplastic anemia is known to
progress into leukemia, and both diseases are thought to occur as a
result of a common pathogenic mechanism. Benzene has been associated
with chromosomal damage in the circulating lymphocytes of exposed
9-10
-------
workers. Many cytogenetic agents are known to cause cancer in humans,
e.g., vinyl chloride, arsenic, and ionizing radiation. Therefore, the
chromosomal aberrations seen in workers exposed to benzene should be
regarded as a serious consequence of exposure. Benzene exposure has also
been causally linked with multiple myeloma, various forms of lymphoma,
and other types of cancer. However, most of the observed cancers were
not suitable for quantitative risk assessment because of statistical
deficiencies in the observed data, i.e., the cancer incidence did not
achieve statistical significance, the relative risk of the specific
cancer could not be numerically quantified, or exposure only to benzene
was not identified in the studies. On the other hand, the causal
association between benzene exposure and leukemia is a strong statistical
association, and it provides the most appropriate basis for estimating
the population risk of cancer through the use of quantitative risk
assessment. The EPA does recognize, however, that the exclusion of other
rates of disease associated with benzene exposure may potentially under-
estimate the risk, but EPA resorted to using those studies having the
highest degree of statistical confidence, demonstrating a strong associa-
tion between leukemia and human exposure, in the derivation of an estimate
of carcinogenic potency.
A commenter also pointed out that EPA's analysis of population risk
from coke by-product recovery plant emissions was weakened by not
including other carcinogens, e.g., toluene and xylenes, in the evaluation.
The Agency has reviewed the scientific literature regarding the
carcinogenicity of toluene and xylenes, and has determined that there is
insufficient evidence to classify the carcinogenic potential of these
compounds [Health and Environmental Effects Profile for Xylenes (o-,m-,p-),
EPA Environmental Criteria and Assessment Office, Docket A-79-16, Docket
Item No. IV-A-7 and Assessment of Toluene as a Potentially Toxic Air
Pollutant, 49 FR 22195, Docket A-79-16, Docket Item No. IV-I-4], The EPA
has reasonably good data characterizing the magnitude of benzene emissions
from the source category. Other specific pollutants that may be
carcinogenic to humans have not been evaluated in the emissions.
Simultaneous exposure to several chemical carcinogens is a frequent
occurrence in the environment, and EPA is committed to toxicological
research in the health risks posed to exposure of complex mixtures. The
9-11
-------
ability to predict how the total mixture of toxicants may interact must
be based on an understanding of the biological mechanisms involved in
such interactions. With regard to toluene, EPA has reviewed and evaluated
the current information on health effects, and has determined that ambient
air concentrations of toluene do not pose a significant risk to public
health and that it is not currently necessary to regulate toluene under
the Clean Air Act (49 FR 22195, May 25, 1984). In the public notice it
was made clear that EPA is aware of additional animal testing that is
under way to investigate the potential carcinogenicity of toluene, and
that further assessment and review of toluene will occur upon completion
of these studies. The potential noncarcinogenic health effects associated
with xylenes are currently under evaluation, and EPA has not yet reached a
decision on this pollutant. The EPA in the reproposal of this standard is
only focusing on the emission of benzene. However, EPA believes that
control and reduction of benzene emissions from coke by-product recovery
plants will have the added benefit of controlling and reducing other VOC's
that may also be present in the emissions.
9.9 ANCILLARY COMMENTS
Comment: As an attachment to their comments on the proposed
regulation, commenter IV-D-13 (NRDC) included their petition to EPA for
reconsideration of four final benzene decisions. These benzene decisions
were the withdrawal of proposed national emission standards for benzene
storage vessels, maleic anhydride plants, and ethylbenzene/styrene plants
(49 FR 23478, June 6, 1984), and the promulgation of standards for
benzene fugitive emissions (49 FR 23512 June 6, 1984). To complete the
record for this rulemaking, commenter IV-D-18 submitted supplemental
comments on behalf of AISI. They comprise the responses to the NROC
petition that were submitted to EPA by the American Petroleum Institute
(API) and the Chemical Manufacturers Association (CMA).
Response: The EPA responded to the NRDC's petition for
reconsideration on August 23, 1985 (50 FR 34144). Included in that
notice were EPA's responses to API's and CMA's sumittals. Therefore,
EPA's response is not repeated here.
9-12
-------
10. EQUIPMENT LEAK DETECTION AND REPAIR
10.1 DETERMINE EMISSIONS OVER BACKGROUND LEVELS
Comment: Commenter IV-D-9 asks, "What is the background level for
proposed standards of 500 ppm above background?"
Response: Section 4.3.2 of Method 21 (48 FR 37598, August 18, 1983,
Docket Item IV-I-1) describes the procedure for determining the presence
of emissions over background levels. Accordingly, the local ambient con-
centration around the source (i.e., background) is determined by moving
the probe inlet randomly upwind and downwind at a distance of 1 to 2
meters (m) from the source. If an interference exists with this deter-
mination because of a nearby emission or leak, the local ambient concen-
tration may be determined at distances closer to the source [but not
closer than 25 centimeters (cm)]. The probe inlet is then moved to the
surface of the source to determine the concentration. (This procedure is
described in Section 4.3.1 of the Method.) The difference between these
concentrations determines whether there are no detectable emissions
(i.e., no more than 500 ppm above background).
10.2 COMPLIANCE WITH LEAK DETECTION AND REPAIR PROGRAM
Comment: Commenter IV-D-16 recommends that the regulations state
specifically that a leak (a reading over 10,000 ppm) is a violation when
documented during a compliance inspection. According to the commenter,
the 1984 proposed rule provides no assurance that a component is actually
inspected, reported, or repaired because this information could easily be
fabricated. Also, enforcement action is unlikely because EPA must prove
that inspection, reporting, or recordkeeping requirements were not met.
According to the commenter, only such a direct enforcement mechanism will
provide incentive for diligent, reliable inspections; without this
10-1
-------
change, the commenter considers the recordkeeping and reporting provi-
sions to be only industry "self-enforcement" rules.
Response: Sealings and packings inherently leak; only the use of
leakless equipment can prevent occasional leakage. Because an occasional
leak cannot be prevented without the use of leakless equipment, EPA cannot
accept the commenter's suggestion that a leak (a reading over 10,000 ppm)
should be considered a violation when documented during a compliance
inspection. Instead, the compliance burden has been placed on the owner
or operator to repair leaks as soon as possible after their detection.
The commenter asserts that enforcement is unlikely because it must
be proven that recordkeeping and reporting requirements were not met or
that the leaking component was not repaired. The EPA disagrees. The
regulation states that compliance will be determined by review of
records, reports, performance test results, or inspections. By comparing
records and reports of plant performance to the actual sources during an
onsite inspection, enforcement personnel will be able to detect
unrepaired sources, unsubstantiated records regarding delayed repair,
falsified records, and a lack of records or reports. Under these
standards, the records and reports (or lack thereof) provide usable
evidence of a violation, and enforcement action is likely. Although the
recordkeeping and reporting requirements, coupled with onsite inspections,
are the only measures to determine compliance, EPA believes these
provisions are adequate to ensure diligent monitoring and repair of leaks
by plant personnel and effective enforcement by EPA.
10.3 DEFINITION OF EQUIPMENT LEAK
Comment: Commenter IV-D-13 requests that EPA reconsider changing
the definition of an equipment "leak" from 10,000 parts per million
volume (ppmv) to 1,000 ppmv or to the highest level at which EPA can
demonstrate, with data, that directed maintenance does not result in net
emission reductions. The commenter remarks that emissions from equipment
leaking at rates below 10,000 ppmv are substantial: about 13 percent of
total emissions from pumps, 2 percent of total emissions from valves in
10-2
-------
gas service, 16 percent of emissions from valves in liquid service, and
16 percent of total emissions from compressors, according to the BID for
proposed national emission standards for benzene fugitive emissions
(EPA-450/3-80-032a). The commenter NRDC states that data from a study on
"directed maintenance"* summarized in the BID for the proposed new source
performance standards (NSPS) for equipment leaks at petroleum refineries
(Docket Item II-A-43) contradict its position that a lower definition
would not reduce emissions. In this study, EPA tested the performance of
both undirected and directed maintenance on valves with initial leak
rates less than 10,000 ppmv. The EPA found that with directed maintenance
there was a net reduction of 85.2 percent emissions.
Response: The EPA's rationale for selecting the 10,000-ppmv leak
definition has been discussed in the promulgation BID's for VOC fugitive
emissions, in the proposal preamble for this rule, and in the rulemakings
for the synthetic organic chemicals maufacturing industry (SOCMI)
(Docket Item IV-A-2), petroleum refinery fugitive emissions (Docket Item
IV-A-3), and benzene fugitive emissions (Docket Item IV-A-1).
The key criterion for selecting a leak definition is the mass emis-
sion reduction demonstrated to be achievable. The EPA has not concluded
that a lower leak definition is demonstrated. A net increase in mass
emissions might result if higher concentration levels result from
attempts to repair a valve with a screening value between 1,000 and
10,000 ppmv. Although many leaks can be repaired successfully at
concentrations less than 1,000 ppmv, even one valve repair failure may
offset many successful valve repairs. Most data on leak repair effective-
ness have applied 10,000 ppm as the leak definition and therefore do not
indicate the effectiveness of repair for leak definitions between 1,000
and 10,000 ppm. Even though data between these values are available,
they are not sufficient to support a leak definition below 10,000 ppm.
As the commenter noted, although there is some evidence that directed
In directed maintenance" efforts, the tightening of the packing is
r2Hu.£re^m?lmi:lltarieoVs1^ ard '! continued only to the extent that it
reduces emissions. In contrast, "undirected" repai r means repairs such
as tightening valve packings without simultaneously monitoring the result
to determine if the repair is increasing or decreasing emissions.
10-3
-------
maintenance is more effective, available data are insufficient to serve
as a basis for requiring directed maintenance for all sources.
A leak definition is an indicator of whether a source is emitting
benzene in quantities large enough to warrant repairs. Certainly, a leak
definition of 10,000 ppmv accomplishes this goal. About 10 percent of
all valves (leaking and nonleaking) contribute about 90 percent or more
of the emissions from valves. At a leak definition of 10,000 ppmv,
approximately 90 percent or more of the leaking valves would be detected,
based on testing in refineries and chemical plants (Docket A-80-44,
Docket Items II-A-30 and II-A-34). Most seals on pumps and exhausters
leak to a certain extent while operating normally, compared to valves
that generally have no leakage. When the seal wears over time, the
concentration and emission rate increase. Properly designed, installed,
and operated seals have low instrument meter readings, and seals that
have worn out or failed have readings generally greater than 10,000 ppmv.
Over 90 percent of emissions from exhauster seals and pump seals in light
liquid service are from sources with instrument readings greater than or
equal to 10,000 ppmv.
The EPA believes that there is only a small potential emission
reduction for sources having benzene concentrations between 1,000 and
10,000 ppmv. Therefore, using a lower leak definition would not increase
emission reductions significantly, even if EPA judged that repair was
effective for leaks of 1,000 ppmv. In the proposal BID for the petroleum
refinery fugitive emissions NSPS (Docket Item II-A-43, p. 4-8), there is
a comparison of the percentage of total mass emissions affected by
selecting a 10,000-ppmv leak definition over a 1,000-ppmv leak definition.
These percentages represent maximum theoretical emission reductions that
could be expected if the sources were instantaneously repaired to a zero
leak rate and no new leaks occurred. For pump seals in liquid service and
compressor seals (similar to exhausters in coke by-product plants), the
estimated decrease is only 6 to 7 percent; for valves in gas service, it
is only 1 percent. This small potential decrease in emissions may be
offset by attempting to repair sources with low leaks.
10-4
-------
In summary, EPA does not disagree with NRDC that additional emission
reductions potentially could be achieved by reducing the leak definition
from 10,000 to 1,000 ppmv. However, EPA has concluded that 10,000 ppmv
is a demonstrated and effective leak definition (i.e., there are large
enough emissions that repair can be accomplished with reasonable costs),
but has not concluded that 1,000 ppmv is a demonstrated leak definition.
Until EPA has adequate data to support the repair potential associated
with leak definitions such as 1,000 ppmv, EPA is selecting the clearly
demonstrated leak definition of 10,000 ppmv instead of a lower level.
10.4 ON-LINE VALVE REPAIR
Comment: Commenter IV-D-13 refers to the 1984 proposal BIO
discussion indicating that on-line repair of valves by drilling into the
valve housing and injecting a sealing compound is growing in acceptance,
especially because of safety concerns. The commenter states that this
discussion means the practice has been demonstrated and should be
required in the final standards.
Response: The EPA does not agree that acknowledging a promising
repair method must be interpreted as meaning "demonstrated" within the
context of the CAA, or that acknowledgment alone constitutes sufficient
justification for a regulatory requirement. The 1984 proposal BIO does
state on page 4-52 that drilling into the valve housing and injecting a
sealing compound is a practice "growing in acceptance" for the on-line
repair of valves. Although the term "growing in acceptance" can be
interpreted to mean that the practice has been reported as one repair
method, the phrase also implies reluctance by plant owners and operators
to use the technique. This hesitancy would be due, in part, to factors
such as the type and location of the valve or the nature of the leak.
For example, plant personnel may prefer to tighten the packing gland
rather than drill into the housing of a critical valve containing a
potentially explosive mixture. Or, as discussed in the preamble to the
1984 proposed rules (49 FR 23533), the valve (or the leak) may require
removal or isolation. Also, this repair approach cannot be used on
control valves or other block valves that are frequently operated because
the valve would then be destroyed and must be replaced. Because of
10-5
-------
uncertainties regarding the applicability of this method to the different
types of valves and varying repair conditions, this technique cannot be
considered fully demonstrated at this time.
Also, the long-term practicability and cost effectiveness of this
method are unknown. Depending on the valve and other factors, this
approach may be no more than a temporary repair until the next unit
shutdown. Without such information, the technique cannot be (and was
not) evaluated as the basis of the standards and a potential regulatory
requirement.
Even if the practice were fully demonstrated and its long-term
practicability and costs were known and deemed superior, an ensuing
regulatory requirement still might not be appropriate. The leak
detection and repair program places the regulatory burden on plant owners
and operators to detect and repair leaks as they occur. Unless a shut-
down is required, all valves must be repaired. A repair period of 5 to
15 days has been provided to allow plant owners or operators the flexi-
bility necessary for efficient handling of repair tasks while main-
taining an effective emission reduction. To provide this flexibility,
the standards do not dictate any single repair method—only the repair;
delays are allowed only under limited circumstances. If any plant owner
or operator chooses to apply this method, it is certainly not precluded
under the standards. To require this method for all valves, however,
would be premature and unwarranted.
10.5 EQUIPMENT LEAK REPAIR PERIOD
Comment: Commenter IV-D-13 recommends that the repair period for
equipment leaks be 24 hours for the first attempt (rather than 5 days, as
proposed), with completion within 5 days as opposed to 15 days. The
commenter suggests that the shorter timeframe is adequate because moni-
toring personnel should be accompanied by workers prepared to fix any
leak upon detection or immediately afterwards.
Response: The EPA's justification for proposing the 5-day first-
attempt-at-repair interval and the 15-day repair period for pumps,
valves, and exhausters was described in the preamble to the 1984
proposed rule at 49 FR 23541.
10-6
-------
The selected repair intervals provide maximum effectiveness of the
leak detection and repair program by requiring expeditious emission
reduction, while allowing the owner or operator the time to maintain a
reasonable overall maintenance schedule for the plant.
During development of the standards already promulgated for equip-
ment leaks in refineries and chemical plants, EPA personnel made a con-
certed effort to investigate and gain knowledge of the industry main-
tenance practices. In EPA's technical judgment, an initial attempt at
repair within 5 days is ample for all simple field repairs. A 24-hour
period following leak detection is often not long enough to allow main-
tenance personnel to identify the cause of a leak and then to attempt
repair. Although plants could schedule repair personnel to accompany the
monitoring team in advance of monitoring, emergency situations or criti-
cal equipment problems could easily postpone these arrangements. Al-
though some or perhaps even most repairs can be made within 24 hours, it
is not practical to require an attempt to repair all equipment within 24
hours. The EPA has not been able to distinguish between equipment that
could and could not always be repaired within 24 hours. In addition,
with the commenter's approach, repair crews would spend much of their
time on an inspection with few needed repairs. The costs of this
approach have not been estimated by EPA because it is not practical.
Furthermore, the owner or operator has an incentive to repair leaks as
quickly as possible to prevent additional product losses.
A 15-day repair interval provides time for isolating leaking
equipment for other than simple field repairs. A 5-day interval, as
suggested by NRDC, however, could cause scheduling problems in repairing
valves that are not conducive to simple field repair and that may require
removal from the process for repair. A 15-day interval provides the
owner or operator with enough time for determining precisely which spare
parts are needed and sufficient time for reasonably scheduling repair.
In addition, a 15-day repair interval allows more efficient handling of
more complex repair tasks while maintaining an effective reduction in
equipment leaks. Again, the owner or operator has an incentive to repair
leaks promptly.
10-7
-------
The commenter's suggestion that leaks can be detected and repaired
within a shorter timeframe if repair workers accompany monitoring per-
sonnel may be helpful for plants able to make such arrangements.
10.6 DELAY OF REPAIR
Comment: Commenter IV-D-13 recommends that the proposed provisions
for delay of repair beyond a unit shutdown be tightened to prevent
abuses. The ccmmenter suggests that it is possible under the proposed
rules to claim lack of equipment in stock as a reason to delay when
"there was in fact plenty of time to anticipate stock needs."
Response: The delay of repair provisions included in the standards
is necessary to ensure technical achievability and reasonable costs.
Delay of repair beyond a unit shutdown is not allowed for any types of
equipment other than valves. Spare parts for valves (e.g., packing gland
bolts and valve packing materials) can be stocked so all leaks that
cannot be repaired without a process unit shutdown can be repaired
during the shutdown. In a few instances, however, the entire valve
assembly may require replacement. The standards address this situation
by allowing delay of repair beyond a process unit shutdown only if the
owner or operator can demonstrate that a sufficient stock of spare valve
parts has been maintained and that the supplies had been depleted. If an
owner or operator has sufficient time to obtain a piece of equipment, he
or she could not reasonably claim a delay of repair as a result of lack
of equipment.
10.7 ALTERNATIVE STANDARD FOR VALVES
Comment: Commenter IV-D-13 states that the proposed alternative
performance standard for leaking valves should be 1 percent rather than 2
percent. According to the commenter, the allowance of 2 percent leaking
valves will result in an average leak rate well over 1 percent. The
commenter believes it inappropriate for EPA and the public to bear all
the risk of statistical sampling error.
Response: The alternative standards for valves were provided for
owners and operators of units exhibiting low leak frequencies because the
cost effectiveness of monthly/quarterly leak detection and repair becomes
10-8
-------
unreasonable at low leak frequencies. The 2-percent limit is intended to
be used as an upper limit for determining compliance with the alternative
standards. If a process unit is subject to and exceeds the 2-percent
limit, the unit does not comply with the standard and is subject to
enforcement actions. The EPA believes that enforcement action should be
taken when noncompliance is supported by the facts. Thus, because the
2-percent limit accounts for the uncertainty in setting this numerical
emission limit, EPA can proceed with enforcement action clearly supported
by the facts. Although there is a regulatory difference between a
2-percent and a 1-percent limit, there is no significant practical
difference to either plant owners and operators or to EPA between limits
of 1 percent or 2 percent of valves leaking. An owner or operator of a
process unit would implement the same control measures to comply with the
alternative valve standard whether the limit were set at 1 or 2 percent.
The NRDC implies that the 2-percent limit is set in industry's favor; in
a practical sense, however, there is little difference in terms of
numbers of valves leaking when maximum limits and averages are compared.
For example, a typical process unit with about 105 valves in service is
allowed to have no more than 2 valves leaking out of the control at the
2-percent maximum limit. A 1-percent limit would allow no more than one
valve leaking. The work practices and equipment used to achieve a rate
of 2 valves leaking out of 105 valves in a process unit at any one time
are the ones that would be used to achieve a 1-percent limit.
10.8 EXEMPTION FOR DIFFICULT-TO-MONITOR VALVES
Comment: Commenter IV-D-13 states that the exemption for difficult-
to-monitor valves is not warranted. Valves above 2 m, according to the
commenter, can be reached by a sampling probe on a boom or by a mobile
"cherry picker."
Response: The EPA disagrees that the exemption for difficult-to-
monitor valves is unreasonable. The intent of the standards is to
monitor valves that can be reached with the portable ladders or with
existing supports such as platforms and fixed ladders. A valve only may
be exempted from monthly monitoring, provided: (1) the plant owner or
operator demonstrates that the valve cannot be monitored without
10-9
-------
elevating monitoring personnel more than 2 m above a support surface,
(2) the valve is in an existing process unit, and (3) the plant owner or
operator follows a written plan requiring monitoring at least once per
year.
The EPA compared the cost effectiveness of scaffolding to annual,
quarterly, and monthly monitoring of difficult-to-monitor valves in petro-
leum refineries (see Docket Item IV-B-4). Based on this analysis,
EPA found the costs of using scaffolding for annual monitoring of benzene
emissions from difficult-to-monitor valves to be reasonable compared to
similar costs for monthly and quarterly programs. These costs were
estimated as the base cost for monitoring and maintenance for readily
accessible valves plus the additional labor cost for scaffolding. No
purchase cost of scaffolding was included because the plant was assumed
to have purchased this equipment for maintenance. However, the previous
purchase of a sampling probe on a boom or a mobile cherry picker cannot
be assumed. Consequently, these purchase costs would result in even
higher costs for each difficult-to-monitor valve. Some valves may be
located in plant areas that are not accessible for repair work using a
mobile cherry picker or a sampling probe on a boom.
Other cost and technical problems are associated with use of a
mobile cherry picker or sampling probe on a boom for monitoring. In
general, few leaking difficult-to-monitor valves are expected at a
typical by-product plant. Although some valves may be located in groups
(e.g., elevated pipe racks), others may be scattered throughout the
plant. The additional labor required for driving, scheduling, and
transporting the vehicle from valve to valve would further increase the
costs previously discussed.
The EPA considers impractical NRDC's suggestion for use of a
sampling probe on a boom because it lacks the precision necessary for
effective monitoring. The monitoring team would not be able to move the
probe around the leaking valve stem or as close as possible to other
potential leak interfaces, as required by the standard. Considering the
high cost and the technical infeasibility, EPA considers that no benefits
would be achieved by this approach.
10-10
-------
10.9 ALTERNATIVE STANDARD FOR OPEN-ENDED VALVES OR LINES
Comment: Cornmenter IV-D-14 suggests that an alternative standard of
no detectable leaks (10,000 ppm) be considered for open-ended valves or
lines in lieu of the proposed equipment requirement of a cap or plug.
This alternative, coupled with monthly monitoring, would satisfy the EPA
goal of leak prevention.
Response: The standards would require open-ended valves and lines
to be equipped with a cap, plug, blind flange, or a second valve depend-
ing on the individual application. If a second valve is used, the up-
stream valve must be cleared first before the downstream valve is closed
to prevent process fluid from being trapped between the valves. The
standards would also allow a bleed valve or line in a double block and
bleed system to remain open when the line between the two block valves is
vented. The bleed valve must be capped, however, when not opened. This
provision is intended to avoid plugging out-of-service bleed valves in a
block and bleed system. These equipment and operational requirements
will reduce uncontrolled benzene and VOC emissions from open-ended valves
or lines by 100 percent.
The commenter suggests an alternative standard of no detectable
leaks, with applicable leak detection and repair (LDAR) requirements.
Application of a cap, plug, blind flange, or second valve is the only
effective method available for reducing or eliminating emissions from
open-ended valves or lines. In EPA's judgment, this equipment still would
be necessary to meet the repair requirements of the LDAR program, even
with a leak definition of 10,000 ppm. However, plant owners or operators
would continue to bear the additional cost of monthly monitoring.
The LDAR program, with a leak definition of 10,000 ppm, should not
be confused with a no detectable emission limit. Plants subject to a
no detectable emission limit would be required to conduct an annual
performance test for each open-ended valve and line. The plant would be
out of compliance if emissions from any of the sources exceeded 500 ppm
above background, as measured by Reference Method 21. Again, use of a
cap, plug, blind flange, or second valve still would be needed to ensure
compliance. Additional costs also should be anticipated for'the record-
keeping and reporting requirements associated with performance testing.
10-11
-------
Although this approach does not seem reasonable because it would require
the same controls at additional cost, the owner or operator could apply
to use this method as an alternative means of compliance with the
standard.
10-12
-------
11. RECORDKEEPING AND REPORTING
11.1 ALTERNATIVE MONITORING AND RECORDKEEPING
Comment: Commenter IV-D-9 asks if monitoring and recordkeeping
requirements can be modified if a technology better than that required by
the standard is used.
Response: Section 61.136 of the standards describes the procedures
for obtaining EPA approval of alternative means of emission reduction
that are equivalent to or better than the equipment, design, operational,
or work practice standards required by the standard. Provisions are
included that allow EPA to include requirements necessary to ensure
proper operation and maintenance. Consequently, if an owner or operator
applies for use of an alternative means of emission limitation, EPA would
consider requiring monitoring, recordkeepi ng, and reporting requirements
appropriate for the alternative on a case-by-case basis.
11.2 RETENTION PERIOD FOR RECORDS AND -REPORTS
Comment: Commenter IV-D-13 argues that records and reports should
be maintained permanently (or for a minimum of 5 years) because of the
availability of automated data systems. If audits or inspections occur
only once every 1 or 2 years, it is important to have available complete
records for more than 2 years.
Response: The Office of Management and Budget (OMB) implementation
of the Paperwork Reduction Act of 1980 (PL-511) specifies 3 years as a
limit beyond which it becomes burdensome for plant owners and operators
to keep records other than health, medical, or tax records. The EPA
selected the 2-year period based on considerable enforcement experience.
The 2-year limit, although less than that allowed by OMB, applies to
significantly detailed plant records that would help enforcement personnel
11-1
-------
assess compliance with the standards. The EPA considers the burden
associated with these records to be reasonable for the 2-year period.
However, EPA does not agree with the commenter that, if EPA audits a
plant less frequently than once every 1 or 2 years, EPA would not be able
to ensure compliance with the standard. Once every 2 years is frequent
enough to review and determine compliance for most owners or operators
affected by the standard. Thus, it would not be necessary for a plant to
keep records longer than 2 years. For these reasons, EPA believes that
it is not necessary to require that owners and operators retain records
longer than a 2-year period. Permanent retention by automated data
systems was not considered necessary for effective enforcement.
11.3 ENFORCEMENT BASED ON RECORDS AND REPORTS FOR EQUIPMENT LEAKS
Comment: Commenter IV-D-13 states that the recordkeeping and
reporting requirements are not strong enough for effective enforcement.
In support, the commenter cites the failure of the proposed rules to
require identifying tags for leaking equipment to facilitate identifi-
cation of "repeat offenders" and the failure of the rules to require
reporting of the specific identity of leaking equipment—only totals.
Response: Tagging is not specifically required by the standard, but
the standard does require some form of weatherproof and readily visible
identification that would enable plant personnel or EPA inspectors to
locate leaking sources readily. Tagging is a useful method of identifi-
cation that has been used in leak detection and repair programs. Any
form of identification is acceptable, however, as long as it is weather-
proof and readily visible. For example, a process unit may have a system
of identifying markings on valves and a diagram that is available to
allow easy location of the marked valves. This type of identification
system works as effectively as tagging and is often used by chemical
plants and petroleum refineries. To require tagging would be unneces-
sarily restrictive if an owner or operator can identify leaking equipment
just as effectively by other means.
11-2
-------
12. MISCELLANEOUS
12.1 ALTERNATIVE MEANS OF EMISSION LIMITATION
Comment: Commenter IV-D-14 recommends revised requirements for
collection and verification of test data to demonstrate equivalence of an
alternative means of emission limitation. In general, the commenter
suggests permitting demonstration of equivalence based on design and
engineering data, with verification after the implementation of controls.
This approach would solve the timing problem encountered in collecting
and verifying data before permission is granted because actual data may
not be available until after controls are installed and relevant data
from other facilities may not be available.
Response: The 1984 proposed regulation provided the plant owner or
operator the opportunity to offer a unique approach to demonstrate the
equivalency of any means of alternative emission limitation to the
standard. If an owner or operator could demonstrate sufficiently the
equivalency based on design and engineering data, EPA will consider that
approach acceptable.
12.2 DEFINITION OF TAR DECANTER
Comment: Commenter IV-D-14 recommends a revised definition of "tar
decanter." The commenter argues that EPA assumes 98-percent control
efficiency on tar-intercepting sumps and 95-percent control for decanters
because sumps separate light tars while decanters separate heavy tars and
sludge. However, some sump units separate light and heavy tars, requir-
ing a sludge conveyor similiar to that used by the decanter. Because of
the conveyor, the sump cannot be endorsed for 98-percent control. The
commenter recommends a revised definition of tar decanter to include "any
vessel, tank, or other type control that functions to separate heavy tar
and sludge from flushing liquor."
12-1
-------
Response: The EPA agrees with the commenter's suggestion that the
definition of "tar decanter" be clarified. In response, the revised
proposed regulation contains the following definition:
"Tar decanter" means any vessel, tank, or other type of
container that functions to separate heavy tar and sludge from
flushing liquor by means of gravity, heat, or chemical emulsion
breakers. A tar decanter may also be known as a
flushing-liquor decanter.
12.3 DEFINITION OF EXHAUSTER
Comment: Commenter IV-D-9 asks to what extent upstream and down-
stream of the rotors does the definition of "exhauster" extend?
Response: In response to the commenter's question, EPA has revised
Section 61.131 of the proposed standards to include the following
definition for "exhauster":
"Exhauster" means a fan located between the inlet gas flange
and outlet gas flange of the coke oven gas line that provides
motive power for coke oven gases.
12.4 WAIVER REQUESTS
Comment: Commenter IV-D-14 recommends that the standard allow any
waiver request submitted within 90 days to be granted on an interim basis
until final determination is made. The commenter indicates that many
waiver requests will be made and suggests that it is unlikely that all
waivers can be submitted and reviewed by EPA within 90 days of the
effective date. Without such a provision, operators will be in a
technical state of noncompliance until the final determination can be
made.
Response: The CAA clearly states in Section 112(c)(l)(8) that an
existing source shall comply with the standard within 90 days of the
effective date unless the source is operating under a waiver of
compliance. Section 112 makes the granting of a waiver contingent upon
EPA's finding "that such period is necessary for the installation of the
waiver to assure that the health of persons will be protected from
imminent endangerment." Granting a waiver before making these findings
12-2
-------
would be inconsistent with the statute. Thus, EPA has not included the
commenter's recommendation. The owner or operator of a source should
submit the waiver application as soon as practicable to allow time for
the Agency to make a determination within the 90-day period after the
effective date. One should note that the owner or operator should take
advantage of the time between reproposal and promulgation to prepare
significant portions of a plan for achieving compliance. In addition,
the source should continue to take all possible steps toward achieving
compliance while the Agency is evaluating the waiver application.
12.5 NEED FOR ADDITIONAL ENFORCEMENT RESOURCES
Comment: Commenter IV-D-1 requests that the proposed standard be
simplified to reduce the enforcement resources needed to ensure
compliance. According to this commenter, additional enforcement
resources will be necessary or a reduction in enforcement activities in
other areas will be required.
Response: The commenter did not describe specifically the
provisions of the regulation he considers would require resource-
intensive enforcement. The regulation inherently has many aspects
because by-product plants comprise several sources with different
applicable control techniques. However, EPA has designed the reporting
requirements to be as simple as possible, while also providing
enforcement personnel indications of potential noncompliance.
12.6 SELECTION OF FORMAT
Comment: Commenter IV-D-17 states that the regulations do not make
clear why some requirements are expressed as equipment standards while
others are expressed as emission limits. The commenter asks specifically
why different standards are applied for different process sources, such
as tar decanters, tar dewatering, and the naphthalene sump (e.g.,
naphthalene processing).
Response: The type of standard (e.g., emission, equipment, work
practice, design, or operational) depends not on the function of the
source, as implied by the commenter, but on the control technique
selected as the basis of the standard.
12-3
-------
Section 112 of the CAA requires that an emission standard be estab-
lished unless such a standard is not feasible to prescribe or enforce.
"Not feasible to prescribe or enforce" means that the pollutant cannot be
emitted through a conveyance designed and constructed to emit or capture
the pollutant or that measurement methodology is not practicable to apply
because of technological or economic limitations. If an emission
standard is not feasible to prescribe or enforce, one of the other types
of standards (including any combination) can be applied.
Gas blankecing has been selected as the basis of the standards for
both tar decanters and tar-dewatering vessels. In the original preamble
discussion in 49 FR 23528, EPA explained why an emission standard, such
as a zero emission limit, was not feasible for gas-blanketing systems.
Such a standard could not be achieved on a continuous basis because,
after installation of the system, vapor leaks occur occasionally because
of the gradual deterioration of sealing materials, even when proper
operation and maintenance procedures are applied. Fugitive emissions
also may be released from openings such as access hatches and sealing
ports. These fugitive emissions cannot be eliminated because the
openings are necessary for proper operations and maintenance of the
sources. An emission standard, it was argued, would be infeasible to
prescribe or enforce not only because it could not be achieved on a
continuous basis (and thus was not appropriate), but because these vapor
leaks and fugitive emissions could not be emitted through a conveyance
designed and operated to emit or capture the emissions. Therefore, mass
emissions could not be measured. For these reasons, an equipment
standard rather than an emission standard (i.e., limit) was developed for
gas-blanketed sources.
The commenter also questions why different standards (e.g., zero
emissions) have been established for naphthalene sumps (processing). In
this case, a process modification requiring the collection of naphthalene
in tar (for foundry coke plants) or wash-oil (for furnace coke plants)
was selected as the basis of the revised proposed standards. Collecting
naphthalene in tar (or wash oil) would eliminate naphthalene-processing
operations (including naphthalene sumps) and the emissions that result
12-4
-------
from separating naphthalene from the hot well of a direct-water final
cooler. Because these emissions and emission sources can be eliminated by
such a modification, a zero emission limit is considered feasible to
prescribe and enforce and has been selected as the format for the revised
proposed standards for naphthalene-processing.
12.7 LIGHT-OIL SUMP CONTROL EFFICIENCY
Comment: Commenter IV-D-14 states that the 98-percent control
efficiency assigned to light-oil sumps is unsupported and should not be
used as the basis for qualifying an alternative means of emission limi-
tation. The commenter recommends instead application of semiannual
monitoring to determine that there are no detectable emissions.
Response: The 98-percent control efficiency assigned to light-oil
sumps is legitimately supported on engineering judgment. As discussed in
the preamble to the 1984 proposed rules in 49 FR 23537-23539, the control
efficiency of source enclosure is theoretically 100 percent. However,
eventual deterioration of the gasket seal (of the cover) may result in
occasional leaks, even with proper operation and maintenance. Because mass
emissions from these leaks cannot reasonably be measured, EPA con-
servatively judged the control to obtain a 98-percent emission reduction.
The 98-percent efficiency for the light-oil sump is consistent with the
98-percent efficiency assigned to gas-blanketing systems.
The semiannual monitoring provisions do not constitute the control
itself. Rather, semiannual monitoring of the light-oil sump cover and
gas-blanketed sources is required to ensure proper operation and main-
tenance (O&M) of the sealed enclosures, i.e., to locate and repair any
leaks that may have developed in the control system. Thus, the commenter1s
recommendation is to allow any alternative control technique provided it
uses the same O&M procedures. However, the equivalency of an alternative
control to the standard must be based on the emission reduction achieved by
the control itself. Then, the provisions necessary for ensuring proper
would have to be determined specifically for the alternative control
system. Without further information, EPA believes that the 98-percent
value is the best estimate available for comparing the efficiency of an
alternative control system.
12-5
-------
Appendix A
Environmental Impact Analysis
-------
I
to
TABLE A-l. FURNACE COKE UV-PROOUCT RECOVERY PLANTS:
COKE-OVEN AND PLANT CAPACITY STATUS
(1,000 Mg/yr)
No.
lb
2
3
4
5
6
7
8
9
10b
11 ,
12b
13
14b
15
16
17
18
19b
Plant
LTV Steel, Thomas, AL
New Boston, Portsmouth, OH
Wheeling-Pitt, Monessen, PA
Lone Star Steel, Lone Star, TX
LTV Steel, So. Chicago, IL
National Steel, Granite
City, IL
lnter)ake, Chicago, IL
LTV Steel, Gadsden, AL
Rouge Steel, Dearborn, MI
U.S. Steel, Fairless Hills, PA
LTV Steel , Warren, OH
LTV Steel, E. Chicago, IN
Arroco Inc. , Ashland, KY
Weirton Steel, Brown's Is., WV
U.S. Steel, Provo, UT
LTV Steel, Aliquippa, PA
Bethlehem Steel ,
Lack a wanna, NY
National Steel, Detroit, MI
U.S. Steel, Lorain, OH
Battery
no.
1
1
1A
IB
2
C
2
A
B
C
1
2
2
3
A
Ax
B
Dx
1
2
4
4
9
3
4
1
1
2
3
4
Al
A5
7
8
9
4
5
D
G
H
I
J
K
L
Battery
capacity
315
364
195
195
100
507
563
285
285
298
291
291
379
379
256
57
312
153
458
458
945
432
516
349
614
1,097
290
290
290
290
604
614
382
382
528
473
924
218
218
218
218
208
208
208
Status*
2
0
2
0
0
0
0
0
0
3
0
o
0
0
0
0
0
0
2
2
0
2
2
0
0
2
0
0
0
0
0
o
0
0
0
0
0
2
2
2
2
2
2
2
No. of
ovens
65
70
37
37
19
70
60
45
45
47
50
50
65
65
45
10
55
27
82
82
85
75
87
76
70
87
63
63
63
63
106
56
76
76
73
78
85
59
59
59
59
59
59
59
Online
0
364
295
507
563
570
582
758
778
0
945
0
963
0
1,160
1,218
1,292
1,397
0
•
Hot
idle
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Cold
idle
315
0
195
0
0
0
0
0
0
916
0
948
0
1.U97
0
0
0
0
1,496
Under
u
u
0
0
0
298
0
0
0
0
0
0
0
0
0
0
0
0
0
Existing
plant
J15
364
490
b(J7
563
570
582
758
778
916
945
948
963
1,097
1,160
1,218
1,292
1,397
1,496
Footnotes at end of table.
(continued)
-------
TWJLt A-l. (continued)
" ExfbtTny
No.
20
i\ '
22
23
24
25&
26
27
28
Battery
Plant no.
Wheel iny- Pitt, t. Steubenvil le, WV 1
2
3
8
LTV Steel, Cleveland, OH 1
2
3
4
6
7
Arraco Inc. , Middletown, OH 1
2
4
Bethlehem Steel, Burns 1
Harbor, IN 2
LTV Steel, Pittsburgh, PA PI
P2
P3N
P3S
P4
U.S. Steel, Fairfield, AL 2
5
6
9
Bethlehem Steel, Bethlehem, PA A
2
3
5
Bethlehem Steel , Sparrows 1
Pt. , I'D 2
3
4
5
6
11
12
A
Inland Steel, E. Chicago, IN 6
7
8
9
10
11
C
Battery
capacity
199
199
215
896
274
274
274
274
332
332
664
664
448
895
895
340
340
340
340
432
818
320
320
364
809
522
522
400
273
263
273
273
273
273
365
365
1,148
278
372
395
395
450
995
830
Status-*
0
0
u
U
0
0
0
2
0
0
0
0
0
u
0
0
0
0
0
0
2
2
2
2
0
0
0
0
1
1
2
2
2
2
0
0
0
0
0
0
0
0
0
2
No. of
ovens
47
47
bl
79
bl
bl
51
51
63
63
57
57
76
82
82
59
59
59
59
59
57
77
77
63
80
102
102
80
63
60
63
63
63
63
65 *
65
80
65
87
87
87
51
69
56
Hot Cold unuer plant
Online idle idle construction total
l,bU9 0 U u l,bl)9
1,486 0 274 u l,7t>U
1,776 0 U U 1,776
1,790 000 1,790
1,792 000 1,792
0 0 1,822 0 1,822
2,253 000 2,253
1,878 536 1,092 0 3,506
2,885 0 830 0 3,715
t rnnf l nllpd )
Footnotes at end of table.
-------
TABLE A-l. (continued)
i
Cn
Battery
No. Plant no.
29 U.S. Steel, Gary, IN 1
2
3
5
7
13
15
16
30 U.S. Steel, Clairton, PA 1
2
3
7
8
9
15
19
20
21
22
B
Total (30 plants)
(24 plants)
Battery
capacity
843
995
995
279
279
279
279
279
296
296
296
296
296
296
302
535
535
535
535
1,076
42,102
39,508
Status*
2
0
0
1
1
1
0
1
1
1
1
1
1
1
1
0
0
0
0
0
No. of
ovens
85
57
57
77
77
77
77
77
64
64
64
64
64
64
61
87
87
86
87
75
7,100
5,935
Existing
Hot Cold Under plant
Online idle idle construction total
2,269 1,116 843 0 4,228
3,216 2,078 0 0 5,294
31,646 3,730 9,828 298 45,804
31,646 3,730 3,234 298 39,210
Note: Data current as of November 1984.
j> Status: 0 = online; 1 = hot idle; 2 = cold idle; and 3 = under construction.
° Denotes cold idle plants.
-------
TABLE A-2. FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS: COKE-OVEN AND PLANT CAPACITY STATUS
(1,000 Mg/yr)
i
cr>
No.
1
2
3
4
5
6
7
8b
9
10
11
12
13
14
Battery
Plant no.
Chattanooga Coke,
Chattanooga, TN
IN Gas, Terre Haute, IN
Koppers, Toledo, OH
Empire Coke, Holt, AL
Koppers, Erie, PA
Tonawanda, Buffalo, NY
Carondolet, St. Louis, NO
AL Byproducts, Keystone, PA
Citizens Gas,
Indianapolis, IN
Jim Walters, Birmingham, AL
Shenango, Pittsburgh, PA
Koppers, Woodward, AL
AL Byproducts, Tarrant, AL
Detroit Coke, Detroit, MI
Total (14 plants)
Total (13 plants)
1
2
1
2
C
1
2
A
B
1
1
2
3
3
4
E
H
I
3
4
5
1
4
2A
28
4
5
A
5
6
1
31
29
Battery
capacity
71
59
66
66
157
107
54
82
125
299
142
64
124
201
201
93
79
305
125
125
249
322
199
149
97
97
145
75
353
113
117
617
5,078
4,676
No. of
Status3 ovens
0
0
0
0
0
0
0
0
0
0
0
0
0
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
24
20
30
30
57
20
40
23
35
60
40
18
35
55
55
47
41
72
30
30
60
56
35
60
38
40
58
30
78
25
29
70
1,341
1,231
Online
130
132
157
161
207
299
330
0
477
499
521
563
470
617
4,563
4,563
Hot
idle
0
0
0
0
0
0
0
0
0
0
0
0
113
0
113
113
Existing
Cold Under plant
idle construction total
0
0
0
0
0
0
0
402
0
0
0
0
0
0
402
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
130
132
157
161
207
299
330
402
477
499
521
563
583
617
5,078
• 4,676
Note: Data current as of November 1984.
aStatus: 0 = online; 1 = hot idle; 2 = cold idle; and 3 = under construction.
bUenotes cold idle plants.
-------
TABLE A-3. FURNACE COKE UY-PROOUCT PLANT OPERATIN3 PROCESSES
No.
1
2
3
4
5
fa
7
B
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Tar Tar Tar
Plant decanter dewatering storage
LTV Steel, Thomas, AL 1
New Boston, Portsmouth, OH 1
Wheeling-Pitt, Monessen, PA 1
Lone Star Steel, Lone Star, TX 1
LTV Steel, So. Chicago, IL 1
National Steel, Granite 1
City, IL
Interlake, Chicago, IL 1
LTV Steel, Gadsden, AL 1
Rouge Steel, Dearborn, MI I
U.S. Steel, Fairless Hills, PA 1
LTV Steel, Warren, OH 1
LTV Steel, E. Chicago, IL 1
Armco Inc., Ashland, KY I
Weirton Steel, Browns Island, WV 1
U.S. Steel, Provo, UT
LTV Steel, Aliquippa, PA
Bethlehem Steel,
Lackawanna, NY
National Steel, Detroit, MI
U.S. Steel, Lorain, OH
Wheeling-Pitt,
E. Steubenville, WV
LTV Steel, Cleveland, OH
Armco Inc. , Middletown, OH
Uethlehem Steel, Burns
Harbor, IN
LTV Steel, Pittsburgh, PA
U.S. Steel, Fairfield, A'L
Bethlehem Steel, Bethlehem, PA
Bethlehem Stee) , Sparrows
Pt., MO
Inland Steel, E. Chicago, IN
U.S. Steel, Gary, IN
U.S. Steel, Clairton, PA
1
1
1
1
1
I
1
1
1
1
1
I
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
I
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Excess-
ammonia
1 iquor
storage
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Flushing-
liquor
cjrc. tank
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Tar-
interc.
sump
1
1
1
1
1
1
1
1
1
1
1
- 1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Light-
oil
storage
1
1
1
1
1
1
1
1
1
1
1
1
1
i
i
1
I
.
1
1
.
.
g
1
1
1
1
1
1
BTX
storage
0
0
0
1
1
1
0
0
1
0
0
1
0
0
1
n
u
o
i
i
i
i
n
u
o
Q
1
0
Q
Q
0
Benzene
storage
0
0
0
0
1
0
0
0
0
0
0
0
0
0
(I
u
0
Q
Q
1
n
u
Q
1
Naphthalene processiny/nandl iny
Denver Naphth. Naphth.
flo. unit melt, nit HrU t*ntQ
1
0
0
1
U
1
1
1
1
1
0
0
0
1
u
0
1
1
-J
J
1
1
n
u
1
1
1
1
u
0
l
0
0
1
u
1
1
1
1
1
u
0
0
1
0
0
1
i
u
1
1
1
1
0
u
1
1
u
1
u
0
I
0
1
I
1
1
u
u
u
1
o
u
1
1
u
u
0
u
0
u
u
Total
30
28
30
30
30
30
29
10
15
Ib
(continued)
-------
TABLE A-3. (continued)
oo
Nn.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Direct-
water
final
Plant cooler
LTV Steel, Thomas, AL
New Boston, Portsmouth, OH
Wheeling-Pitt, Monessen, PA
Lone Star Steel, Lone Star, TX
LTV Steel, So. Chicago, IL
National Steel, Granite
City, IL
Interlake, Chicago, IL
LTV Steel, Gadsden, AL
Rouge Steel, Dearborn, MI
U.S. Steel, Fairless Hills, PA
LTV Steel, Warren, OH
LTV Steel, E. Chicago, IN
Armco Inc., Ashland, KY
Weirton Steel, Browns Island, WV
U.S. Steel, Provo, UT
LTV Steel, Aliquippa, PA
Bethlehem Steel,
Lackawanna, NY
National Steel, Detroit, MI
U.S. Steel, Lorain, OH
Wheeling-Pitt,
E. Steubenville, WV
LTV Steel, Cleveland, OH
Armco Inc. , Middletown, OH
Bethlehem Steel, Burns
Harbor, IN
LTV Steel, Pittsburgh, PA
U.S. Steel, Fairfield, AL
Bethlehem Steel, Bethlehem, PA
Bethlehem Steel, Sparrows
Pt., MD
Inland Steel, E. Chicago, IN
U.S. Steel, Gary, IN
U.S. Steel, Clairton, PA
Total
1
0
0
1
0
1
1
1
1
1
0
1
0
0
0
0
1
1
1
0
0
1
0
1
1
1
0
0
1
Q
16
Tar-
bottom
final
cooler
0
1
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
4
Wash-
oil Light-
final oil
cooler sump
0
0
Q
0
0
0
0
0
0
0
1
0
0
1
0
0
0
0
0
1
1
0
0
0
0
0
1
0
0
0
5
1
1
. 1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
29
Light-
oil decanter/
condenser Wash-oil
vent decanter
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
29
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
29
Wash-oil
circ. tank
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
29
Equipment
leaks
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
29
Note: Data current as of November 1984.
-------
TABLE A-4. FOUNDRY COKE BY-PRODUCT PLANT OPERATINS PROCESSES
No.
10
11
12
13
14
Plant
ammonia Flushing-
Naphthalene processing/h
"enverNaphth.— N
- •»> ii\fuu
decanter dewaterinq storage stora
storage storage storaqe
melt pit dry tanks
Chattanooga Coke,
Chattanooga, TN
IN Gas, Terre Haute, IN
Koppers, Toledo, OH
Empire Coke, Holt, AL
Koppers, Erie, PA
Tonawanda, Buffalo, NY
Carondolet, St. Louis, MO
AL Byproducts, Keystone, PA
Citizens Gas,
Indianapolis, IN
Jim Walters, Birmingham, AL
Shenango, Pittsburgh, PA
Koppers, Woodward, AL
AL Byproducts, Tarrant, AL
Detroit Coke, Detroit, MI
continued)
-------
TABLE A-4. (continued)
3=>
t—>
O
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Wash-oil
final
Plant cooler
Chattanooga Coke,
Chattanooga, TN
IN Gas, Terre Haute, IN
Koppers, Toledo, OH
, Empire Coke, Holt, AL
Koppers, Erie, PA
Tonawanda, Buffalo, NY
Carondolet, St. Louis, MO
AL Byproducts, Keystone, PA
Citizens Gas,
Indianapolis, IN
Jim Walters, Birmingham, AL
Shenango, Pittsburgh, PA
Koppers, Woodward, AL
AL Byproducts, Tarrant, AL
Detroit Coke, Detroit, MI'
Total
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Direct-
water
final
cooler
1
1
0
1
0
0
0
1
1
1
0
0
1
0
7
Tar-
bottom
final
cooler
0
0
1
0
0
0
0
0
0
0
0
1
0
0
2
Light-
oil
sump
1
1
0
1
0
0
0
1
0
1
1
1
1
0
8
Light-
oil decanter/
condenser Wash-oil
vent decanter
1
1
0
1
0
0
0
1
0
1
1
1
1
0
8
1
1
0
1
0
0
0
1
0
1
1
1
1
0
8
Wash-oil
circ. tank
1
1
0
1
0
0
0
1
0
1
1
1
1
0
8
Equipment
leaks
1
1
0
1
0
0
0
1
0
1
1
1
1
0
8
Note: Data current as of November 1984.
-------
TABLE A-5. ESTIMATED NATIONWIDE BASELINE BENZENE EMISSIONS
FROM FURNACE AND FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS
Source
1. Direct-water final-cooler
cool iny tower
2. Tar-bottom final-cooler
cooling tower
3. Naphthalene processing/
handling
4. Light-oil decanter/
condenser vent
b. Tar-intercepting sump
6. Tar decanter
7. Tar dewatering
8. Tar storage
y. Light-oil sump
1U. Light-oil storage
11. BTX storage
12. Benzene storage
13. Flushing liquor
circulation tank
14. Excess-ammonia liquor
storage tank
lb. Wash-oil decanter
16. Wash-oil circulation tank
Footnotes at end of tables.
affected
plants3
16
4
Ib
29
30
30
28
30
29
29
10
4
30
30
29
29
Furnace plants
Capacity,15
1,000 Mg/yr
21,430
b,729
19,664
44,014
45,804
45,804
39,947
45,804
44,014
44,014
11,544
10.523
45,804
45,804
44,014
44,014
Nationwide
emissions,
Mg/yr
5,786
401
2,103
3,456
4,351
3,527
839
550
660
242
54
bl
412
412
154
154
No. of
affected
plants8
7
2
7
9
14
14
13
14
9
9
2
1
14
14
9
9
Foundry plants
Capacity, b
1,000 Mg/yr
2,384
720
2,384
3,290
6,078
5,078
4,917
5,078
3,290
3,290
901
402
5,078
b,078
3,290
3,290
Furnace and foundry
Nationwide
emissions,
Mg/yr
470
37
186
158
227
184
49
29
27
10
3
1
33
33
7
7
No. of
affected
plants*
23
6
22
38
44
44
41
44
38
38
12
b
44
44
3«
38
Capacity,15
1,000 Mg/yr
23,814
6,449
22,U3b
47.3U4
50,882
511,882
44,864
50,882
47,304
47.3U4
12,445
lu,92b
6U.B82
bu,882
47.JU4
47,304
plants
Nationwide
emissions,
My/yr
6,2bb
4J«
2,289
J.614
4,b78
3,711
687
b/8
687
ZbJ
b?
62
446
44b
161
Ibl
(continued^
-------
TABLE A-5. (continued)
Furnace plants
No. of
affected
Source plants3
17.
18.
19.
20.
21.
22.
Pump seals
Valves
Pressure-relief devices
Exhausters
Sampling connection systems
Open-ended lines
Total (rounded)
29
29
29
29
29
29
Capacity, b
1,000 Mg/yr
44,014
44,014
44,014
44,014
44,014
44,014
Nationwide No. of
emissions, affected
Mg/yr plants3
370 9
249 9
168 9
17 9
33 9
11 9
24,000
Foundry plants
Capacity, b
1,000 Mg/yr
3,290
3,290
3,290
3,290
3,290
3,290
Furnace and foundry plants
Nationwide No. of Nationwide
emissions, affected Capacity,5 emissions,
Mg/yr plants3 l.UOO My/yr My/yr
101
68
46
5
9
3
1,700
J8 47,304 471
3« 47,304 317
38 47.3U4 214
38 47,304 2i!
38 47,304 42
38 47,304 14
2b,7UU
Note: Oata current as of November 1984.
aNumber of plants having this source (out of a total of 30 furnace plants, 14 foundry plants, or 44 furnace and foundry plants combined; includes plants
currently on cold idle.
Capacity of plants with this source.
-------
TABLE A-6. ESTIMATED NATIONWIDE BASELINE VOCa EMISSIONS
FROM FURNACE AND FOUNDRY COKE BV-PRODUCT RECOVERY PLANTS
Furnace plants
TT: ~ • •
i.
2.
3.
4.
b.
I
L 6.
co
7.
8.
9.
10.
11.
12.
13.
14.
lb.
16.
Font ri
Source
Direct-water final-cooler
(Doling tower
Tar-bottom final-cooler
cooling tower
Naphthalene processing/
handling
Light-oil decanter/
condenser vent
Tar intercepting sump
Tar decanter
Tar dewatering
Tar storage
Light-oil sump
Light-oil storage
BTX storage
Benzene storage
Flushing-liquor
circulation tank
Excess-ammonia liquor
storage tank
Wash-oil decanter
Wash oil circulation tank
ifttp*; Af pnd nf Fahlac
no. or
affected
plants'5
16
4
lb
29
30
30
28
30
29
29
10
4
30
30
29
29
Capacity,0
1,000 Mg/yr
21,430
5,729
19,6b4
44,014
45,«04
45,804
39,947
45,804
44,014
44,014
11,544
10,523
45,804
45,804
44,014
44.U14
Nationwide
emissions,
Mg/yr
90,842
6,302
3,302
4,031
9,252
7,512
19,654
12.S71
942
347
77
61
591
591
219
219
No. of
affected
plants0
7
2
7
9
14
14
13
14
9
9
2
1
14
14
9
9
Foundry plants
Capacity,0
1,000 Mg/yr
2,384
720
2,384
3,290
5,078
5,078
4,917
5,078
3,290
3,290
901
402
b,07«
5,078
3,290
3,290
Furnace and foundry
Nationwide
emissions,
Mg/yr
7,377
578
292
226
482
391
1,137
671
38
15
4
1
48
48
10
10
No. of
affected
plants'3
23
6
22
38
44
44
41
44
38
38
12
b
44
44
38
38
Capacity,0
1,000 Mg/yr
23,814
6,449
22,038
47,304
50,882
bO,882
44,864
bO,882
47.3U4
17,304
12,44b
10,92b
50,882
bU,882
4/.304
47,304
plants
Nationwide
emissions,
Mg/yr
98,219
6.88U
3,b94
5,lb/
9,73b
7,903
20,791
13,542
9811
362
82
62
639
639
229
229
(continued)
-------
TABLE A-6. (continued)
Furnace plants
17.
18.
19.
20.
21.
22.
Source
Pump seals
Valves
Pressure-relief devices
Exhausters
Sampling connection systems
Open-ended lines
Total (rounded)
No. of
affected
plants'1
29
29
29
29
29
29
(
Capacity,0
1,000 Mg/yr
44,014
44,014
44,014
44,014
44,014
44,014
Nationwide
emissions,
Mg/yr
528
355
240
25
47
16
59,200
No. of
affected
plants'1
9
9
9
9
9
9
Foundry plants
Capacity,0
1,000 Mg/yr
3,290
3,290
3,290
3,290
3,290
3,290
Furnace and foundry plants
Nationwide
emissions,
Mg/yr
1S9
107
73
8
14
5
11,700
No. of
affected
plants'1
38
38
38
38
3«
38
Capacity,0
l.OOU Mg/yr
47,304
47,304
47,304
47,304
47,304
47,304
Nationwide
emissions,
My/yr
bb7
463
312
3J
bl
21
17U.900
Note: Data current as of November 1984.
aBenzene and other VOC.
bNuraber of plants having this source out of a total of 30 furnace plants, 14 foundry plants, or 44 furnace and foundry plants combined;
includes plants currently on cold idle.
cCapacity of plants with this source.
-------
T/tBLE A-7. CONTROL OPTION IMPACTS FOR FURNACE PLANTS
3=-
H-"
in
Emission source
Final cooler
cooling tower and
naphthalene
processing/
handling
Tar decanter, tar-
intercepting
sump, and
flushing-liquor
circulation tank6
Tar storage tanks
and tar-dewatering
tanks6
Light-oil
condenser, light-
oil decanter,
wash-oil decanter
and wash-oil
circulation tanks
Excess-ammonia
liquor storage
tank6
Light-oil storage
tanks and BTX
storage tanks6
Benzene storage
tanks
Light-oil sump
Pumps
Control option/
efficiency, %
Baseline0:
Tar-bottom final cooler 81
Wash-oil final cooler 100
Baseline0:
Gas blanketing 98d
Baseline0:
Gas blanketing 98
Baseline0:
Wash-oil scrubber 90
Gas blanketing 98
Baseline0:
Gas blanketing 98
Baseline0:
Gas blanketing 98
Baseline0:
Wash-oil scrubber 90
Nj gas blanketing 98
Baseline0:
Cover 98
Baseline0:
Quarterly inspections 71
Monthly inspections 83
Dual mechanical seals 100
Footnotes at end of table. " " "™ "
Controlled Controlled
emissions, incidence,
**9/yr, lives/yr
Benzene/VOC^
8,290
1,900
0
8,290
270
1,390
30
3,760
380
80
410
8
300
6
60
6
1
660
10
370
110
60
0
100,000
29,400
0
17,400
600
32 ,500
600
5,370
550
120
590
11
420
9
60
6
1
940
20
530
150
90
0
0.78
0.16
0.00
0.99
0.02
0.16
0.003
0.39
O.*040
0.009
0.047
0.001
0.039
0.001
0.0063
0.0006
0.0001
0.079
0.001
0.044
0.013
0.008
0.000
Annual
costs,'5
1984 $/yr
152,600
12,871,380
(1,062,260)
1,735,070
348,320
504,530
849,370
1,119,040
120,570
134,040
468,470
29,800
35,780
1,067,250
Benzene cost VOCa cost
effectiveness, b.f effectiveness
1984 $/Hq 1984 I/Mo
aver./incre.
20 20
1,550 6,690
(130) (130)
1,280 1,280
100 100
140 520
2,100 2,100
3,860 3,860
2,200 2,200
2,240 2,760
720 720
110 110
120 130
2,890 16,690
aver./incre.
2 2
130 430
(60) (60)
50 50
70 70
100 360
1,470 1.470
2,700 2,700
2,200 2,200
2,240 2,760
500 500
80 80
81 90
2,020 11,780
-------
TABLE A-7. (continued)
CT>
Emission source
Valves
Exhausters
Pressure-relief
devices
Sampling
connection systems
Open-ended lines
Naphthalene
processing/
handlings
Control option/
efficiency, %
Baseline0:
Quarterly inspections
Monthly inspections
Sealed-bellows valves
Basel inec:
Quarterly inspections
Monthly inspections
Degassing reservoir
vents
Basel inec:
Quarterly inspections
Monthly inspections
Rupture disc system
Baseline0:
Closed-purge sampling
Baseline0:
Cap or plug
Baseline0:
Mixer-settler
Control led
emissions,
Mg/yr
Benzene/VOCa
63
73
100
55
64
100
44
52
100
100
100
100
250
90
70
0
20
8
6
0
170
90
80
0
30
0
10
0
2.100
0
350
130
100
0
30
10
9
0
240
130
no
0
50
0
20
0
3,300
0
dont'rolled'
incidence, Annual
lives/yr costs, &
1984 $/yr
0.030
0.011
0.009
0.000
0.0021
0.0011
0.0008
0.0000
0.020
0.011
0.009
0.000
0.0039
0.0000
0.0013
0.0000
0.29
0.00
(36,760)
(19,760)
4,324,320
13,870
30,160
437,320
(30,940)
(26,990)
153,820
40,770
7,110
1,603,960
Benzene cost VOCa cost
effectiveness,11.* effectiveness,^
1984 $/Mg 1984 $/My
aver./incre.
(240)
(110)
17,380
1,450
2,740
25,130
(410)
(310)
920
1,250
640
760
(240)
720
62,780
1,450
11,230
68,820
(410)
290
2,270
1,250
640
760
aver./incre.
(160)
(80)
12,170
1,020
1,920
17,590
(290)
(210)
640
870
450
490
(160)
500
43,950
1,020
7,870
44,640
(290)
200
1,590
870
450
490
Note: Data current as of November 1984.
aVOC estimates include benzene.
^Parentheses denote cost savings.
cBaseline numbers represent relatively uncontrolled levels.
d95 % efficiency for tar decanter.
^ash-oil scrubbers are more costly and less effective than gas blanketing for these sources.
^''Average" means compared to baseline; "incremental" means compared to the next less stringent control option.
9The mixer-settler control option for naphthalene processing and handling is shown
separately to address a comment on new indirect cooling technology that would
not necessarily control naphthalene processing emissions.
-------
TABLE A-8. CONTROL OPTION IMPACTS FOR FOUNDRY PLANTS
I
I—"
^•g
Emission source
Final cooler
cooling tower
and naphthalene
processing/
handling
Tar decanter, tar-
intercepting
sump, and
flushing-liquor
circulation
tank6
Tar storage
tanks and tar-
dewatering tanks6
Light-oil
condenser, light-
oil decanter,
wash-oil
decanter
and wash-oil
circulation
tanks6
Excess-ammonia
liquor storage
tank6
Liyht-oil
storage tanks
and BTX storage
tanks6
Benzene storage
tanks6
Light-oil sump
Pumps
Control option/
efficiency, %
Baseline0:
Tar-bottom final cooler 81
Wash-oil final cooler 100
Baseline"-:
Gas blanketing 93d
Baseline0:
Gas blanketing 98
Baseline0:
Gas blanketing 98
Baseline0:
Gas blanketing 98
Baseline0:
Gas blanketing 98
Baseline0:
«2 gas blanketing 98
Baseline0:
Cover 98
Baseline0:
quarterly inspections 71
Monthly inspections 83
Dual mechanical seals 100
Controlled
emissions,
Mg/yr
Controlled
Incidence,
Hves/yr
Benzene/VOCa
690
160
0
440
10
80
2
170
3
30
0.7
10
0.3
1.0
U.I
30
0.5
100
30
20
0
8,250
2,490
0
920
30
1.810
40
250
5
50
1
20
0.4
1.0
0.1
40
0.8
160
50
30
0
0.077
0.012
0.000
0.070
0.003
0.012
0.001
0.023
0.0005
0.0049
0.0001
0.0021
0.00004
0.00041
0.00001
0.004
0.000
0.019
0.006
0.003
0.000
Benzene cost VOCa cost
Annual effectiveness, ">f effectiveness,0^
costs, b 1984 $/Mg 1984 $/Mg
1984 $/yr Aver./incre. Aver./1ncre.
305,550 570 570 50 50
2,831,680 4,090 15,900 340 1,000
410,390 960 960 460 460
397,510 5,260 5,260 220 220
238,200 1,420 1,420 990 990
277,630 8,490 8.490 5,920 5,920
251,140 19,520 19,500 13,640 13,640
14,360 11,640 11,640 11,640 11,640
44,290 1,700 1,700 1,190 1,190
9,360 130 130 80 80
11,270 134 150 90 100
323,960 3,200 18,500 2,030 11,730
(continued)
-------
TABLE A-8. (continued)
3=>
i
i—"
CO
Emission source
Valves
Exhausters
Pressure-relief
devices
Sampling
connection
systems
Open-ended lines
Naphthalene
processing/
handlings
Control option/
efficiency, %
Baseline0:
Quarterly inspections
Monthly inspections
Sealed-bellows valves
Baseline0:
Quarterly inspections
Monthly inspections
Degassing reservoir
vents
Baseline0:
Quarterly inspections
Monthly inspections
Rupture disc system
Baseline0:
Closed-purge sampling
Baseline0:
Cap or plug
Baseline0:
Mixer-settler
Controlled
emissions,
Mg/yr
Controlled
incidence,
lives/yr
Benzene/VOCd
63
73
100
55
64
100
44
52
100
100
100
100
70
30
20
0
4.9
2.2
1.8
0.0
50
30
20
0
9
0
3
0
190
0
110
40
30
0
7.7
3.4
2.8
0.0
70
40
30
0
14
0
5
0
290
0
0.013
0.005
0.004
0.000
0.00091
0.00040
0.00030
0.00000
0.0085
0.0047
0.0040
0.0000
0.0017
0.0000
0.00056
0.00000
0.034
0.000
Annual
costs, b
1984 $/yr
(10
(5
1,310
4
9
135
(9
(8
47
12
2
453
,920)
,740)
,400
,310
,360
,720
,320)
,110)
,010
,380
,220
,080
Benzene cost
effectiveness, b>
1984 $/Mg
Aver.
(260)
(120)
19,250
1,590
3,010
27,640
(450)
(330)
1,020
1,390
730
2,430
/incre.
(260)
800
69,530
1,590
12,330
70,200
(450)
320
2,520
1,390
730
2,430
VOCa cost
' effectiveness, b«f
1984 $/Mg
Aver,
(160)
(70)
12,220
1,010
1,910
17,540
(290)
(210)
650
880
460
1,550
./incre.
(160)
510
44,170
1,010
7,900
44,490
(290)
200
1,600
880
460
1,550
Note: Oata current as of November 1984.
a VOC estimates include benzene.
b Parentheses denote cost savings.
0 Baseline numbers represent relatively uncontrolled levels.
d 95% efficiency for tar decanter.
e Wash-oil scrubbers are more costly and less effective than gas blanketing for these sources.
f "Average" means compared to baseline; "incremental" means compared to the next less stringent control option.
9 The mixer-settler control option for naphthalene processing and handling is shown
separately to address a comment on new indirect cooling technology that would
not necessarily control naphthalene processing emissions.
-------
TABLE A-9. EFFECT OF CONTROL OPTIONS ON REDUCING BENZENE EMISSIONS AT FURNACE
AND FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS
I
I—»
•.o
Furnace Plants
1.
2.
3.
4.
5.
6.
7.
8.
y.
10.
11.
12.
13.
14.
15.
Source
All sources
Final-cooler
cooling tower
Light-oil decanter/
condenser vent
Tar-Intercepting
sump
Tar decanter
Tar dewaterlng
Tar storage
Light-oil sump
Light-oil storage
BTX storage
Benzene storage
Flushing-1 iquor
circulation tank
Excess-ammonia-
liquor storage
Wash-oil decanter
Mash-oil
circulation tank
Control option
No national emission standard
1. Tar-bottom final cooler
2. Wash-oil final cooler
1. Wash-oil scrubber
2. Coke-oven gas blanketing
Coke-oven gas blanketing
Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
1. Coke-oven gas blanketing
Sealed cover
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Gas blanketing
Coke-oven gas blanketing
Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
No. of
affected
plants3
30
16
20
29
29
30
30
28
28
30
30
29
29
29
10
10
4
4
30
30
29
29
29
29
national
benzene
emissions,
Mg/yr
24,000
8,290
8,290
3,456
3,456
4,351
3,527
839
839
550
550
660
242
242
54
54
61
61
412
412
154
154
154
154
Controlled
benzene
emissions,
Mg/yr
24,000
1,901
0
356
78
87
176
84
17
55
11
13
24
5
7
1
6
1
8
8
15
3
15
3
Foundry Plants
National Controlled
No. of benzene benzene
affected emissions, emissions,
plants3 Mg/yr Mg/yr
14
7
9
9
9
14
14
13
13
14
14
9
9
9
2
2
1
1
14
14
9
9
9
9
1,690
693
693
158
158
227
184
49
49
29
29
27
10
10
3
3
1
1
33
33
7
7
7
7
1,690
159
0
16
3
5
9
5
1
3
0.6
0.5
1
0.2
0.3
0.1
0.03
0.7
0.7
6.1
0.1
6.1
0.1
Furnace and Foundrv Plant-;
No. of
affected
plants3
44
23
29
38
38
44
44
41
41
44
44
38
38
38
12
12
5
44
44
38
38
38
38
National
benzene
emissions,
Mg/yr
25,900
8,983
8,983
3,614
3,614
4,578
3,711
887
887
578
578
687
253
253
57
57
62
446
446
161
161
161
161
Controlled
benzene
emissions,
Mg/yr
25,900
2,060
0
372
82
92
186
89
18
58
12
14
26
5
7
1
6
1
9
9
21
3
21
3
Footnote at end of table. ~
(continued)
-------
TABLE A-9. (continued)
I
rv»
o
Furnace Plants
16.
17.
18.
19.
20.
21.
Source
Pump seals
Valves
Pressure- relief
devices
Exhausters
Sampling connection
systems
Open-ended lines
1.
2.
3.
1.
2.
3.
1.
2.
3.
1.
2.
3.
Control option
Quarterly inspection
Monthly inspection
Dual mechanical seals
Quarterly inspection
Monthly inspection
Sealed-bellows valves
Quarterly inspection
Monthly inspection
Rupture disc
Quarterly inspection
Monthly inspection
Rupture disc
Closed purge sampling
Cap
or plug
Total (rounded)
No. of
affected
plants9
29
29
29
29
29
29
29
29
29
29
29
29
29
29
National
benzene
emissions,
Mg/yr
370
370
370
249
249
249
168
168
168
17
17
17
33
a
24,000
Controlled
benzene
emissions,
Mg/yr
108
62
0
93
69
0
93
80
0
8
6
0
0
0
Foundry Plants
No. of
affected
plants3
9
9
9
9
9
9
9
9
9
9
9
9
9
9
National
benzene
emissions,
Mg/yr
101
101
101
68
68
68
46
46
46
5
5
6
9
3
1,700
Control led
benzene
emissions,
Mg/yr
29
17
0
25
19
0
26
22
0
2
2
0
0
0
Furnace and Foundry Plants
No. of
affected
plants4
38
38
38
38
38
38
38
38
38
38
38
38
38
38
National
benzene
emissions,
Mg/yr
471
471
471
317
317
317
214
214
214
22
22
22
42
14
2b,700
Control led
benzene
emissions,
Mg/yr
1J7
79
0
118
8b
U
119
1U1
U
10
8
U
0
U
Note: Data current as of November 1984.
aNuraber of plants having this source out of a total of 30 furnace plants, 14 foundry plants, or 44 furnace and foundry plants combined;
includes plants currently on cold idle.
-------
TABLE A-10. EFFECT OF BENZENE CONTROL OPTIONS ON REDUCING V0a
EMISSIONS AT FURNACE AND FOUNDRY COKE BY-PRODUCT RECOVERY PLANTS
Furnace plants
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
IS.
Source
All sources
Final-cooler
cooling tower
Light-oil decanter/
condenser vent
Tar-intercepting
sump
Tar decanter
Tar dewatering
Tar storage
Light-oil sump
Light-oil storage
BTX storage
Benzene storage
Flushing-liquor
circulation tank
Excess-ammonia
liquor storage
tank
Wash-oil decanter
Wash-oil
circulation tank
Control option
No national emission standard
1. Tar-bottom final cooler
2. Wash-oil final cooler
1. Wash-oil scrubber
2. Coke-oven gas blanketing
Coke-oven gas blanketing
Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
Sealed cover
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Gas blanketing
Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
1. Wash-oil scrubber
2. Coke-oven gas blanketing
No. of
affected
plants
30
16
20
29
29
30
30
28
28
30
30
29
29
29
10
10
4
4
30
30
30
29
29
29
29
VOC
nationwide
emissions,
Mg/yr
159,200
100,446
100,446
4,931
4,931
9,252
7,512
19,654
19,654
12,871
12,871
942
347
347
77
77
61
61
591
591
591
219
219
219
219
Controlled
VOC
emissions,
Mg/yr
159,200
29,400
0
493
112
185
376
1,965
393
1,287
257
19
35
7
10
2
6
1
12
59
12
22
5
22
5
Foundry plants
No. of
affected
plants
14
7
9
9
9
14
14
13
13
14
14
9
9
9
2
2
1
1
14
14
14
9
9
9
9
VOC
nationwide
emissions,
Mg/yr
11,700
8,248
8,248
226
226
482
391
1,137
1,137
671
671
38
15
15
4
4
1
1
48
48
48
10
10
10
10
Controlled
VOC
emissions,
Mg/yr
11,700
2,943
23
5
10
20
114
23
67
13.4
0.8
1
0.3
0.4
0.1
0.1
0.03
1
5
1
1
0.2
1
0.2
Furnace and foundry plants
VOC Controlled
No. of nationwide VOC
affected emissions, emissions,
plants Mg/yr Mg/yr
44
23
29
38
38
44
44
41
41
44
44
38
38
38
12
12
5
5
44
44
44
38
38
38
38
170,900
108,694
108,694
5,157
5,157
9,735
7,903
20,791
20,791
13,542
13,542
980
362
362
82
82
62
62
639
639
639
229
229
229
229
170,900
32,343
0
516
116
195
395
2,079
416
1,354
271
20
36
7
10
2
6
1
13
64
13
23
5
23
5
Footnotes at end of table.
(continued)
-------
TABLE A-10. (continued)
3=-
r>o
Furnace plants
16.
17.
18.
19.
Source
Pump seals
Valves
Pressure relief
devices
Exhausters
1.
2.
3.
1.
2.
3.
1.
2.
3.
1.
2.
3.
Control option
Quarterly inspection
Monthly inspection
Dual mechanical seals
Quarterly inspection
Monthly inspection
Sealed-bellows valves
Quarterly inspection
Monthly inspection
Rupture disc
Quarterly inspection
Monthly inspection
Rupture disc
(to. of
affected
plants
29
29
29
29
29
29
29
29
29
29
29
29
VOC
nationwide
emissions,
Mg/yr
528
528
528
355
355
355
240
240
240
25
25
25
Controlled
VOC
emissions,
Mg/yr
154
88
0
132
99
0
133
114
0
11
9
0
Foundry plants
tto. of
affected
plants
9
9
9
9
9
9
9
9
9
9
9
9
TGC
nationwide
emissions,
Mg/yr
159
159
159
107
107
107
73
73
73
8
8
8
Control led
VOC
emissions,
Mg/yr
46
27
0
40
30
0
40
34
0
3
3
0
Furnace
to. of
affected
plants
38
38
38
38
38
38
38
38
38
38
38
38
and foundry plants
• voc
nationwide
emissions,
Mg/yr
b87
687
687
463
463
463
312
312
312
33
33
33
Control led
VOC
emissions,
My/yr
2UO
115
0
172
129
0
173
148
0
15
12
0
20. Sampling connection Closed purge sampling
systems
21. Open-ended lines Cap or plug
29
29
47
16
14
38
38
61
21
Note: Data current as of November 1984.
aBenzene and other VOC.
bNumber of plants having this source out of a total of 30 furnace plants, 14 foundry plants, or 44 furnace and foundry plants combined-
includes plants currently on cold idle. '
-------
J=
to
TABLE A-ll. ENERGY USE AT MODEL BY-PRODUCT PLANTSa
User
Steam,
Mg/yr
Furnace plants'5
Gas blanketing
Tar decanter, tar-intercepting sump, 350
and flushing liquor circulation tank
Tar dewatering, tar storage 440
Excess ammonia-liquor storage tank 126
Condenser, light-oil decanter, wash-oil 174
decanter, and circulation tank
Wash-oil scrubber
Excess ammonia liquor storage tank 24
Benzene storage tank
Final cooler
Tar-bottom final cooler
Wash-oil final cooler
380
1.210
j| For information on derivation
D 4,000 Mg coke/day.
c 1,000 Mg coke/day.
Electricity,
MWh/yr
0.4
0.9
98
1,330
steam,
Mg/yr
Foundry Plants0
Electricity,
MWh/yr
127
303
26
333
estimates, see Docket A-79-16, item IV-B-11
-------
TABLE A-12. EMISSIONS OF COKE-OVEN GAS FROM SELECTED
FURNACE AND FOUNDRY COKE-OVEN BY-PRODUCT PLANT SOURCES
Furnace plant emissions, Foundry plant emissions,
Source * gas/min/Mg coke/day t gas/min/Mg coke/day
Tar decanter 10.0 7.5
Light-oil condenser 0.18 0.14
Tar dehydrator 2.9 2.2
Tar storage 2.8 2.1
A-24
-------
TABLE A-13. YIELDS—FOUNORy VS. FURNACE COKE PLANTS
Year
1976
1977
1978
1979
1980
1981
1982
i
ro 1983
en
Coal-to-coke ratio
Merch. a Furn.
1.35 1.46
1.31 1.47
1.34 1.47
1.34 1.46
1.32 1.47
1.29 1.46
Average 1.325 1.465
ratios--
merch./furn.
™i5f™ Ifyr« ?' J? y f ' i , Gas yield, Light oil /gas cone. , Light oil/tar cone. .
gal/ton of coal gal/ton of coal 1,000 ft3/ton coal gal light oil/1.000 ft3 gal/gal
Merch. Furn. Merch. Furn. Merch. Furn. Merch Furii Merch Fi.m
l'61 2-58 5.26 7.77 9.21 11.02 0.18 0.23 0.32
!-77 2-9 5.52 7.78 9.23 11.2 0.19 0.26 0.32
1.82 2.51 5.94 7.86 9.03 11.04 0.2 0.23 0.31
1.82 2.67 5.97 8.27 8.94 11.14 0.2 0.24 0.3
8.08 10.37
0.33
0.37
0.32
(J.J2
1.77 2.665 5.6725 7.952 9.1025 10.954 0.1925 0.24 0.3125 0.335
•664 -713 .831 0.802 0.933
Merchant coke plants are assumed to be the same as foundry coke plants.
-------
TABLE A-14. CORRECTION FACTOR COMPUTATION FOR FOUNDRY
COKE BY-PRODUCT RECOVERY PLANTS
Source
Concentration
adjustment
Volume (throughput)
adjustment
Total
correction
Light oil plant Benzene in
light oil
(63.5/70)=0.907
Light oil yield
coke basis
x(0.664)(l.325/1.465)
0.54
Water contact
with coke
oven gas
Benzene in light oil
Light oil in
coke oven gas
(0.907x0.802) x
0.73
Tar sources
Benzene in light oil
Light oil in coke
oven gas
(0.907x0.802)
Tar yield
coke basis
x(0.713)(l.325/1.465) =
0.47
Equip, leaks
Benzene in light oil
(0.907)
0.91
A-26
-------
TABLE A-15. UNCONTROLLED BENZENE EMISSIONS FACTORS
FOR FURNACE AND FOUNDRY COKE BY-PRODUCT PLANTS
Source
Cooling tower
Direct-water
Tar-bottom
Naphthalene separation
and processing
Light-oil condenser vent
Tar intercepting sump
Tar dewatering
Tar decanter
Tar storage
Light-oil sump
Light-oil storage
BTX storage
Benzene storage
Flushing-! iquor
circulation tank
Excess-ammoni a
liquor tank
Wash-oil decanter
Wash-oil circulation tank
Pump seals
Valves
Pressure-relief devices
Exhausters
Sample connections
Open-ended lines
Furnace plant
emission factors
g benzene/Mg coke
270
70
107
89
90
21
77
12
15
5.8
5.8
5.8
9
9
3.8
3.8
a
a
a
a
a
a
Foundry Plant
emission factors
g benzene/Mg coke
197
51
79
48
45
9.9
36
5.6
8.1
3.1
3.1
3.1
6.6
6.6
2.1
2.1
a
a
a
a
a
a
Emission factors are not related to coke production capacity and are listed in
Table A-16.
A-27
-------
TABLE A-16. BENZENE EMISSION FACTORS FOR EQUIPMENT LEAKS
ro
00
Valves
Pumps
Exhausters
Pressure relief
devices
Sampling
connections
Open-ended lines
Percent of
sources
leaking
initially
11
24
35
d
d
d
VOC emission
factor,
kg/source
day
0.26
2.7
1.2
3.9
0.36
0.055
Furnace plant
Benzene emission factors,
kg benzene/source day
Plant Aa
0.18
1.9
0.28C
2.7
0.25
0.038
Plant B&
0.22
2.3
0.28C
3.4
0.31
0.047
Foundry plant
benzene emission factors
kg benzene/source day
Plant A*
0.16
1.7
0.25
2.5
0.23
0.035
Plant Bb
0.2U
2.1
0.25
3.1
0.28
0.043
aPlant A recovers light oil. The amount of benzene in the light oil is assumed to be 70 percent at furnace
plants and 63.5 percent at foundry plants.
B recovers refined benzene. The amount of benzene averaged over the light oil and refined benzene
is assumed to be 86 percent at furnace plants and 78 percent at foundry plants.
C23.5% benzene in nonmethane hydrocarbon.
dThis type of information would not be appropriate for relief valve
overpressure, sampling connections, and open-ended lines.
-------
Appendix B
Cost Impact Analysis
-------
APPENDIX B: COST IMPACT ANALYSIS
B.I DEVELOPMENT OF REVISED CONTROL COST ESTIMATES
In response to comments received during the public comment period
for the proposed coke-oven by-product plant National Emission Standards
for Hazardous Air Pollutants (NESHAP), an in-depth review of the benzene
control cost estimates was undertaken. The nature of the comments
received touched virtually all aspects of the cost-estimating metho-
dology. This report [Appendix B] documents the various elements of the
review process, the revisions made to the cost-estimating methodology,
and the resulting changes in nationwide cost impacts.
The general theme of the comments received was that control costs
for the industry had been underestimated, and therefore the cost and
economic impacts were understated. The American Iron and Steel Institute
(AISI) supplied cost factors and cost estimates for particular portions
of the proposed control systems on the basis of data supplied by member
companies. The information supplied ranged from cost factors for indivi-
dual components of the control systems to a cost estimate for the pro-
posed controls applied to an entire plant. Because the differences
between the cost information received and U.S. Environmental Protection
Agency (EPA) cost estimates was large in some cases, a complete review
was undertaken.
To begin the review, additional information supporting the cost data
provided by AISI was requested. At least three member companies had
contributed specific cost data. However, this step alone was not
expected to provide sufficient explanation for all the differences
between EPA and commenters1 cost data. Simultaneously, plans were
developed to have a firm not previously involved in the cost-estimating
efforts prepare a set of unit cost factors for gas-blanketing systems.
These new factors then could be compared with those used by EPA and those
B-3
-------
supplied by the commenters. To have the unit cost factors reflect
conditions imposed by a real plant situation, Bethlehem Steel was asked
to allow EPA and/or their contractors to visit the Bethlehem, Pennsyl-
vania, plant and use the conditions existing there as the basis for unit
costs. A second purpose of the visit to the Bethlehem plant was to
generate a cost estimate for applying positive-pressure gas blanketing to
all sources covered by the proposed regulation for comparison with the
Bethlehem plant cost estimate contained in the AISI comments.
The CRS Si mine was retained to develop the unit cost factors and
overall cost estimate for the specific case of the Bethlehem plant (thus
avoiding any potential conflict of interest). The company was selected
because its staff has not done much work with steel plants and parti-
cularly coke-oven facilities. However, they have considerable experience
with petroleum refinery and petrochemical plant engineering and cost
estimating (these plants handle similar materials, e.g., oils, tars,
explosive mixtures). The EPA, through Research Triangle Institute (RTI),
supplied copies of the proposed regulation, the background information
document (BID) for the proposed regulation, and estimates of the gaseous
emission rates from the various by-product plant sources to CRS Sirrine.
Bethlehem Steel provided plant drawings, estimated pumping rates, and
processed vessel size information to EPA for CRS Sirrine as requested.
Bethlehem Steel also provided information on in-plant restrictions on
welding and safety matters.
The direction given to CRS Sirrine was to develop cost estimates for
safe positive-pressure gas-blanketing controls applied to the various
groups of sources within the Bethlehem plant. In doing this, they were
instructed to make use of existing connection points and existing support
structures for piping, where possible, and to use pipe routings, tank
roofing, and vessel closure methods that would tend to minimize costs.
They were also informed of the nature of the compounds in the blanketing
gas and the vapor space over the process liquids, as well as the need to
avoid condensation and subsequent plugging that might result in the
gas-blanketing pipelines.
The plant visit for cost estimation purposes was made on February 19
and 20, 1985. Representatives from RTI, CRS Sirrine, Bethlehem Steel,
B-4
-------
and United Engineers (the engineering firm supplying the cost estimate
contained in AISI's comments) were present at the Bethlehem plant during
this effort. The CRS Sirn'ne developed the required cost estimates and
gave the results to EPA and RTI in the form of a report titled Benzene
Emissions Control Estimate (Docket Item IV-J-8). The report provided an
overall cost estimate for positive-pressure gas-blanketing systems
applied to six groups of sources in the plant, as well as light-oil sump
covers and roof installation for tar dewatering tanks and an excess
ammonia-liquor storage tank. The wide range of pipe sizes, valves, and
other piping hardware used in the control systems estimate provided the
desired unit cost information.
National Steel, Armco, and Bethlehem Steel (including United
Engineers) contributed additional background information in response to
EPA's request for more details for the cost comment evaluation.
Ultimately, much of the industry-contributed data were used in the
development of the revised cost estimates.
B.2 COMPARISON OF UNIT COST FACTORS
One of the major findings from the Bethlehem plant cost study was
that, in general, the unit cost factors for piping and piping hardware
should be increased. The unit cost factors for piping and piping hard-
ware developed for CRS Sirrine's estimate were higher than those used in
the BID for the proposed standards, and they were more in the range of
the unit cost factors contained in the comments received. A principal
factor contributing to the difference was labor cost for installation of
the piping. The revised factors are presented later in this report.
They are basically a composite of the Si mine and industry-supplied
data.
The use of the Bethlehem plant for the cost study provided a basis
for estimating the costs resulting from roof additions to tanks not
currently covered. There was no provision for this cost element in the
proposal BID estimates; this cost has been added to the control cost
estimates for tar dewatering and excess ammonia-liquor storage tanks.
Another result from the Bethlehem cost study was the addition of pipe
supports for the minimum gas-blanketing cost cases. The proposal BID
estimates included pipe supports only in the maximum cost cases. Pipe
B-5
-------
support costs also were added to the wash-oil scrubbers based on data
provided In Industry comments. The operating labor costs for wash-oil
scrubbers were Increased as a result of the higher hourly labor rates
developed during the review. Costs for sealing all process vessels and
Installing pressure/vacuum relief valves also were added for both
gas-blanketing and wash-oil scrubber cases.
A review of the United Engineers' control cost estimates for final -
cooler cooling towers (prepared for Bethlehem Steel and submitted in the
AISI comments) suggested some appropriate revisions to the EPA cost
estimates for those controls. The proposal BID estimate for the tar-
bottom mixer-settler included no allowance for piping to and from the new
equipment. In some plants this may be a significant cost if the new
equipment cannot be located immediately adjacent to the existing final
cooler equipment. Piping costs were added for the revised cost esti-
mates. After considering the number of pumps and vessels required in the
tar bottom mixer-settler installation, source operating labor also was
added to the annualized costs.
The United Engineers' cost estimate for wash-oil final coolers
indicated that some use could be made of existing direct-water final-
cooler equipment in the conversion to a wash-oil cooler scheme. The
capital costs estimated by United Engineers for Bethlehem Steel's
Bethlehem plant were significantly lower than the proposal BID cost
estimates for a wash-oil final cooler installation. In their comments on
the proposed regulation, AISI said one of the member companies had a 1981
budgetary cost estimate for a wash-oil final cooler of $7.4 million (1984
dollars) installed in a 2,51)0 Mg/day plant, much higher than EPA esti-
mates. The EPA requested further details about this cost estimate.
Rather than supplying the requested information, the company supplied
newer costs estimates for other types of final coolers. One estimate was
for a tar-bottom final cooler, and the other two estimates were for
indirect cooling schemes that would achieve final cooler and naphthalene
processing emission reductions equivalent to wash-oil final coolers.
One indirect scheme uses warm wash-oil absorption of naphthalene
upstream of the final cooler with final cooling provided by direct water
contact with gas. However, the direct contact water is cooled indirectly
B-6
-------
in a heat exchanger thereby avoiding atmospheric emissions from the
direct contact water. The capital cost for this system was estimated at
$1.5 million for a 1,600 Mg/day plant.
The other indirect scheme uses a cross-tube cooler to cool the gas.
Tar is injected into the gas-side of the cooler to keep condensed naph-
thalene in solution. The condensed water and tar-containing naphthalene
are returned to the collecting main. The cooling fluid is water flowing
through tubes of the heat exchanger, never coming into direct contact
with the coke oven gas. The capital cost of this system was estimated at
$2.0 million for a 1,600 Mg/day plant.
According to Dravo/Still, an equipment vendor and design firm, the
latter indirect cooling scheme is in use at Dofasco in Hamilton, Ontario.
Warm wash-oil absorption of naphthalene is in use at Armco's Middletown
plant; however, the direct water used to cool the gas at Armco is cooled
in an atmospheric cooling tower rather being cooled indirectly.
The cost of the warm wash-oil absorption with indirect cooling of
the cooling water is about the same (scaled to equivalent plant size) as
the United Engineers' estimate for retrofitting a wash-oil final cooler
to the Bethlehem plant. The cost estimate for indirect cooling with the
cross-tube cooler and naphthalene absorption by tar is about 25 percent
higher (when scaled to equivalent plant size). The United Engineers'
estimate was selected as the basis for the revised wash-oil final cooler
costs. Dravo/Still indicated the cross-tube cooler is more expensive
because reuse of existing equipment is difficult with this final cooler
scheme.
One other unit cost related to final coolers was -revised.* The cost
of makeup wash oil for wash-oil final cooler systems was estimated in the
proposal BID at $0.11/kg. This unit cost was increased for the current
cost estimates to $0.34/kg or $283/m3'on the basis of information pro-
vided by Shenango Inc. The quantity of makeup wash oil required, how-
ever, was revised downward by a significant amount from the proposal BID.
Comparison of the old makeup wash-oil estimates to potential losses of
wash oil indicated that the proposal BID substantially overestimated the
required makeup wash oil for final coolers.
B-7
-------
B.3 COMPARISON OF WHOLE PLANT ESTIMATE
The changes to the unit cost-estimating factors cited above do not
account for all the cost differences between EPA cost estimates and those
submitted by AISI for the Bethlehem plant. Study of the Bethlehem plant
cost estimates contained in the AISI comments revealed that some of the
sources included in their control systems were not required to be
controlled by the proposed regulation. This was partially attributable
to questions about source definitions in the proposed regulation. As a
result, the source definition for excess ammonia-liquor storage tanks has
been changed. Elimination of costs for those sources not requiring
control reduces the gap between the two estimates.
Assumed pipe sizes for gas-blanketing control systems was another
area of difference between the EPA and AISI estimates for the Bethlehem
plant. In general, the pip~ sizes indicated in the industry cost
estimate exceeded the sizes assumed in the EPA design for gas-blanketing
control systems. The system design presented by Sirrine in their study
of the Bethlehem plant generally used smaller gas blanket pipe diameters
than either EPA or industry estimates. Sirrine argued that the small gas
flow rates expected to and from the various process vessels only required
smaller pipe sizes. According to Sirrine, the more uniform heating
achievable by heat tracing the smaller pipe sizes and the higher flow
rates that would occur in smaller pipe sizes would reduce the likelihood
of condensation and resultant plugging rather than increase it. The EPA
ultimately was influenced more by the fact that the proposal BID
estimates were based on pipe sizes used in existing gas-blanketing
systems. On the whole, the pipe sizes for the revised gas-blanketing
cost estimates were neither increased or decreased. This fact explains
some of the remaining differences between EPA and industry estimates.
B.4 OTHER COST ELEMENT REVISIONS
Armco and Bethlehem Steel submitted cost estimates for wash-oil
scrubbers applied to specific sources in some of their plants. Compari-
son of their data with the proposal BID data for wash-oil scrubbers
suggested a more appropriate way to estimate costs for scrubber
applications. The proposal BID estimates used scrubber shell area as
the basis for scrubber capital cost, with shell area estimated roughly
B-8
-------
from the expected gas flows to be treated. The industry cost estimates
appeared to be directly proportional to the number of sources to be
treated rather than the size of the sources. For the majority of
emissions sources for which wash-oil scrubbers were costed, the gas flow
rate is intermittent. The highest flow would tend to occur during
pumping into the tank, when flow rate would be controlled by the pumping
rate. Pump capacity is not expected to vary in direct proportion to tank
size, but rather over a narrower range, because a particular pump
capacity could handle a range of tank capacities by varying pumping size.
For these reasons, we agree that the number of sources is a more
appropriate basis for the estimate. The wash-oil scrubber unit costs
were revised. The result of this change is that the costs for wash-oil
scrubbers applied to medium and large plants have increased compared to
the estimates at proposal.
Plan drawings of by-product plant facilities submitted in support of
the AISI comments suggested another revision to the cost estimates. The
t
drawings indicated that the number of light-oil, benzene-toluene-
xylene (BTX), and benzene storage tanks for plants of specific size were
generally higher than was assumed in the model plants used to develop the
proposal BID cost-estimating equations. The number of these tanks in the
model facilities was, therefore, increased. This revision when combined
with the one described above has increased the costs estimates for all
plants recovering light oil. Table B-l lists the revised cost factors
used in developing the revised capital and operating cost estimates.
B.5 EXTENSION OF UNIT COST FACTORS TO PLANT COST ESTIMATES
The unit cost factors provided in Tables B-l and B-2 were extended
into full plant cost estimates in the same manner as presented in Chapter
8 of the proposal BID. Piping distances and numbers of process equipment
were specified for each model plant'size and each group of emitting
sources. To reflect the variation that is typical from plant to plant,
minimum and maximum values were specified for piping distances and other
control equipment elements. Minimum and maximum values also were
specified for certain of the unit cost factors such as pipe supports and
wash oil scrubbers . For each model plant, minimum and maximum capital
costs were estimated by multiplying the equipment element numbers by the
B-9
-------
TABLE B-l. REVISED CAPITAL COST FACTORS
Pipe + fittings.
diameter-i n
Cost/unit1
ripe + r i it i ng, sredur tracea,
& insulated,
cost/unit* (1984$)
I $j
4 !
3 !
2 !
1 !
LOO/ft
565/ft
50/ft
>40/ft
22/ft
15/ft
70
50
30
20
<
<
(
(
L45/ft
.UO/ft
>83/ft
>72/ft
>54/ft
>46/ft
153
130
109
Valves (3-way lubricated plug valves)
Diameter-i n
Cost/unit
PI
Pi
8
6
4
2
1
ug valves
8
6
pe supports
t
Minimum case -
Maximum case -
Tar decanter clean
Minimum case -
Maximum case -
<
{
i
i
<
(
I
$7/ft
$30/ft
53,000
52^500
1*000
5 700
i 500
5 200
51,600
5 900
, cover, and seal
$5/ft2 ,
$30.5/ft2
Tar sumps clean, cover, and
seal
Minimum case - $10.5/ft?
Maximum case - $44.5/fti::
Hot tap
Sin. -
12 in. -
53,800
57,600
Footnote at end of table.
(conti nued)
B-10
-------
TABLE B-l. (continued)
Tank sealing
Flushing liquor circulation - $l,400/tank
Tar tanks - $l,400/tank
Tank roof (including tank cleanout)
Tar dewatering tank - $46.5/ft^ 9
Excess ammonia-liquor tank - $46.5/ft'::
Vessel sealing
Ammonia-liquor area - $3,000/unit
Light-oil area - $l,500/unit
Nitrogen blanketing site preparation
Large plant (assume Model Plant 3 size) - $30,000/plant
Flame arresters
6 in - $2,000/arrestor
4 in - 51,000/arrestor
Pressure/vacuum relief valves
6 in - $1.300
4 in - $8(30
3 in - $660
Pressure reducing valve
Valve - $2,000
Wash-oil scrubber pump
Pump - $3,900
Wash-oil scrubber instrumentation
Flow, temperature, and pressure - $2,500
Light-oil sump cover
Minimum case - $30.5/ft2
Maximum case - $164/ft^
Parenthetical numbers refer to light-oil plant cost considering restricted
construction conditions, e.g., no welding.
B-ll
-------
TABLE 8-2. REVISED ANNUALIZED COST ITEMS
Item Cost (1984$)
Benzene credit, as fuel9 $0.14/kg
Benzene credit, recovered9 $0.39/kg
Light-oil credit9 $U.27/kg
Capital recovery (20 yr @ 6.2%) 8.86% of capital
Electricityb $0.05/kWh
SteamC $18.30/Mg
Cooling water0 $U.06/m3
Wash oild $283/m3
Operating labor (including plant $39.69/h
overhead @ 80%)e
aDerived from Quarterly Coal Report, Energy Information Administration,
DOE/EIA-0121, January-March 1984, table A-16.
bEscalated from proposal BID (1982$) 20 percent (electrical rates
generally escalated more rapidly than overall rate of inflation).
cEscalated from proposal BID (1982$) by 4 percent.
^Based on $1.02/gallon + freight (assumed $0.05/gal) per telecon with
James Zwikl, Snenango Inc., August 9, 1985.
eBased on $22.05 per hour from United Engineers estimates in Bethlehem
Steel comments on the proposed regulations, and 80 percent plant
overhead rate.
B-12
-------
unit cost factors. This procedure resulted in a set of minimum and
maximum capital cost estimates for each group of emission sources within
each model plant size. These capital cost estimates are shown in Tables
B-3 through B-18.
Average capital cost estimates were computed for each group of
emission sources and each model plant size by averaging the minimum and
coke by-product plants in the industry, equations were developed that
estimated the capita" costs for each group of emission sources as a
function of coke production capacity. An equation best fitting the
capital cost estimates for each emission source group and for the three
model plant sizes was obtained by performing linear or curvilinear
regression analyses on the estimates. Nationwide capital cost estimates
was generated by using the equations to estimate the average capital cost
for each plant.
Minimum and maximum annualized costs were estimated for each group
of emission sources and each model plant size using the unit cost factors
in Table B-2 and the capital costs estimated by the above procedure. A
set of equations for estimating annualized costs for each emission source
group as a function of plant coke production capacity was generated by
the same procedure described for capital cost estimates.
Capital recovery charges were computed on the basis of a 20-yr
equipment lifetime at 6.2 percent interest. The equipment lifetime was
increased from that used in the proposal BID cost estimates to make the
equipment lifetime assumption more compatible with typical lifetimes for
coke-oven plant equipment in general. The 6.2 percent interest assump-
tion is estimated to be the real (net of inflation) cost of capital to
the coke industry.
In the process of estimating nationwide annualized costs, credits for
recovery of benzene and/or light oil were applied to all plants except
those few specifically identified as not being able to benefit from
recovery. The annualized costs are shown in Tables B-3 through B-18.
B-13
-------
TABLE B-3. COSTS FUR GAS BLANKETING OF TAR DECANTER, TAR-INTERCEPTING SUMP,
AND FLUSHING-LIQUOR CIRCULATION TANK
(All Costs In 1984 Dollars)
DO
I
Cost element
Pressure taps
20-cm (8-in) pipe, m
(ft)
7.6-cm (3-in) pipe, m
(ft)
Pipe supports, m
(ft)
Valvesb
20-cm (8-in) plug valve
Clean, cover, seal decante
m2
(ft2)
Clean, cover, seal sump,
m2
(ft2)
Seal flushing liquor tanks
Capital cost0
Total capital costd
Model
Minimum
1
61
(200)
46
(ISO)
107
(350)
4
1
r,
149
(1,600)
3.0
(32)
1
72,600
102,300
plant 1
Maximum
1
122
(400)
91
(300)
213
(700)
4
1
149
(1,600)
3.0
(32)
1
172,800
243,700
Model
Minimum
1
91
(300)
46
(150)
137
(450)
6
1
223
(2,400)
23
(250)
2
103,100
145,300
plant 2
Maximum
1
244
(800)
91
(300)
335
(1,100)
6
1
223
(2,400)
23
(250)
2
285,900
403,200
Model
Minimum
1
152
(500)
91
(300)
244
(800)
10
1
446
(4,800)
46
(500)
3
176,600
248,900
plant 3
Maximum
1
366
(1,200)
183
(600)
549
(1,800)
10
1
446
(4,800)
46
(500)
3
487,500
687,300
Cost pe
Minimum
3,800
476
(145)
236
(72)
23
(7)
3,800
1,600
53.8
(5)
713
(105)
1,400
ir unit3
Maximum
98.4
(30)
328
30.5
479
(44.5)
Footnotes at end of table.
(continued)
-------
TABLE B-3. (continued)
CO
I
Model plant 1
Model plant 2
Model plant 3
Cost element
Minimum Maximum Minimum Maximum Minimum Maximum
Annualized cost
Maintenance, overhead (9%)e 9,210
Utilities1" 1,970
Taxes, insurance (4%) 4,090
Capital recovery (8.86%)9 9,070
Total annualized cost 24,300
21,900
3,940
9,750
21.600
57,200
13,100
2,740
5,810
12.900
34,500
36,300
7,010
16,100
35.700
95,100
22,400 61,900
4,710 10,900
9,960 27,500
22,000 60.900
59,100 161,200
values arc
b 3-way valves, 15 cm (6 in)-$2,500; and pressure/vacuum relief valves, 15 cm (6 in)-$l,300.
c Capital cost includes subcontractor overhead and profit and contractor material markup.
Total capital cost includes construction fee, contingency, engineering, and startup (41%).
e Maintenance and overhead are 5% and 4% of total capital cost, respectively.
f Steam at 18.3/Mg.
9 Capital recovery factor for 20-yr lifetime at 6.2%.
Cost per unita
Minimum Maximum
-------
cn
t—»
CTl
TAliLt U-4. COSTS FOR WASH-OIL VENf SCRUBBER FOK MR OECANTER, TAR-INTERCEPTING SUMP
ANO FLUSHING-LIQUOR CIRCULATION TANK
(All Costs in 1984 Dollars)
H'odel 'pl'an't"l '""* Hod'el"'pTa'n't''2 '"''" 'Ho'deT'p'l'a'n't'~3'""" C'o's't''u'er'TTnYt'^'"
Cost element
Scrubber vessels
15.2-cm (6-in) vent pipe,b
7.6-cm (3-in) vent pipe to
sump,0 m
(ft)
2.5-cm (1-in) wash-oil
supply, m
(ft)
5.1-cm (2-in) wash oil
drain, d m
(ft)
Valves6
Seal flushing-liquor tanks
Clean, cover, and seal tar
decanter, w?
(ft2)
Clean, cover, and seal tar
sumps, in2
(ft2)
Pump
Instrumentation^
Capital costy
Total capital costh
Minimum
4,
m 46
(ft) (150)
46
(150)
61
(200)
61
(200)
4
1
149
(1,600)
2.9
(32)
1
1
68,11)0
96,100
Maximum
4
46
(150)
91
(300)
152
(500)
152
(500)
4
1
149
(1,600)
17 2.97
(32)
1
1
146,900
207, inu
Minimum
6
76
(250)
46
(150)
91
(300)
91
(300)
6
2
223
(2,400)
23.2
(250)
1
1
100,500
141,700
Maximum
6
76
(250)
91
(300)
610
(2,000)
610
(2,000)
6
2
223
(2,400)
23.2
(250)
1
1
270,700
381,700
Minimum
10
122
(400)
91
(300)
122
(400)
122
(400)
10
3
446
(4,800)
46.5
(500)
2
2
170,600
240,500
Maximum
10
122
(400)
183
(600)
762
(2,500)
762
(2,500)
10
3
446
(4,800)
46.5
(500)
2
2
451.700
6 36, BOO
Minimum
1,000
351
(107)
259
(79)
49.2
(15)
95.1
(29)
3,800
1,400
53.8
(5)
113
(10.5)
3,900
2,51)0
Maximum
2,000
328.3
(30.5)
479
(44.5)
-------
TABLE B-4. (continued)
CXI
i
Cost element
Annual i zed costs
Maintenance, overhead (9%)'
UtilitiesJ
Taxes, insurance (4%)
Operating labor^
Capital recovery (8.86*)1
Total annual 1 zed cost
Model
Minimum
8,650
1,360
3,840
7,240
8,510
29,600
plant 1
Maximum
18,600
1,790
8,280
7,240
1£,300
54,300
Model
Minimun
12,800
2,110
5,670
7,240
12,600
40,300
plant 2
i Maximum
34,300
2,970
15,300
7,240
33*800
93,600
Model
Minimum
21,600
3,700
9,620
14,500
21.300
70,800
Plant 3 Cost per unit3
Maximum Minimum Maximum
57,300
4.040
25,500
14,500
56.400
157,700
low values are shown as maximum and
b Pipe, steam traced @ $328/m or $100/ft and pipe supports 9 $23/m or $7/ft.
c Pipe @ $236/m or $72/ft and pipe supports 0 $23/m or $7/ft.
d Pipe e $72.20/m or $22/ft and pipe supports @ $23/m or $7/ft.
e 3-way valves, 15 cm (6-in) - $2,500 and pressure/vacuum release valves, 15 cm (6 in) - $1,300.
f Includes flowmeter with alarm, pressure gauge, and temperature gauge.
9 Capital cost includes subcontractor overhead and profit and contractor or material markup.
n Total capital cost includes construction fee, contingency, engineering, and startup (41%).
' Maintenance and overhead are 5% and 4% of total capital cost, respectively.
J Electricity at $0.05/kWh.
k For 30 min/day/scrubber system at $39.69/h.
1 Capital recovery factor for 20-yr lifetime at 6.2%.
*****
-------
TABLE B-5. COSTS FOR GAS BLANKETING AMMONIA LIQUOR STORAGE TANKS
(All Costs in 1984 Dollars)
TO
I
oo
Cost element
15.2-cm (6-1n) vent pipe, ra
(ft)
Valvesb
lb.2-cm (6-in) plug valve
Pipe supports, m
(ft)
Seal tanks
Tank roofs, of
(ft2)
Capital cost0
Total capital costd
Annuall zed costs
Maintenance, overhead (9%)e
Utilitiesf
Taxes, insurance (4%)
Capital recovery (8.86*)9
Total annual i zed cost
Model
Minimum
46
(150)
1
1
46
(150)
1
0
(0)
23,800
33,500
3,010
870
1,340
2,970
8,190
plant 1
Maximum
152
(500)
1
1
152
(500)
1
49.3
(531)
97,400
137,300
12,300
2,890
5,490
12,200
32,900
Model
Minimum
61
(200)
3
1
61
(200)
3
0
(0)
42,700
60,200
5,420
1,150
2,410
5,330
14,300
plant 2
Maximum
183
(600)
3
1
183
(600)
3
74.7
(804)
136,700
192,700
17,300
3,460
7,710
17.100
45,600
Model
Minimum
91
(300)
6
1
91
(300)
6
0
(0)
73,800
104,100
9,370
1,730
4,160
9,220
24,500
plant 3
Maximum
305
(1,000)
6
1
305
(1,000)
6
169
(1,816)
256,100
361,200
32,500
5,770
14,400
32,000
84,700
Cost per unita
Minimum Maximum
328
(100)
3,800
900
23 98.4
(7) (30)
3,000
501
(46.5)
a Where a range of unit costs was used, the high and low values are shown as maximum and minimum; where only a single
value was used, it is shown in the minimum column.
b 3-way valves, 15.2 cm (6 in)-$2,500; and pressure/vacuum relief valves, 15.2 cm (6 in)-$l,300.
c Capital cost includes subcontractor overhead and profit and contractor material markup.
d Total capital cost includes construction fee, contingency, engineering, and startup (41%).
e Maintenance and overhead are 5% and 4% of total capital cost, respectively.
f Steam at $18.3/My.
9 Capital recovery factor for a 20-yr lifetime at 6.2%.
-------
UD
TABLE B-6. COSTS FOR WASH-OIL VENT SCRUBBER FOR AMMONIA LIQUOR STORAGE TANKS
(All Costs in 1984 Dollars)
Model plant 1
Cost element Minimum
Scrubber vessels
7.6-cm (3-in) vent pipeb, m
(ft)
2.5-cm (1-in) wash-oil line, m
(ft)
5.1-cm (2-in.) wash-oil
drain,0 m
(ft)
Valvesd
Pumps
Instrumentation6
Seal tanks
Tank roofs, m2
(ft2)
Capital cost^ 14
Total capital cost9 20
I
9.1
(30)
30.5
(100)
30.5
(100)
1
0
1
1
0
(0)
,600
,600
Maximum
1
9.1
(30)
152
(500)
152
(500)
1
2
1
1
49.3
(531)
65,700
92,700
Model plant 2
Minimum
3
46
(150)
61
(200)
61
(200)
3
0
1
3
0
(0)
39,200
55,300
Maximum
3
46
(150)
152
(500)
152
(500)
3
2
1
3
74.7
(804)
100,600
141,900
Model plant 3
Minimum
6
91
(300)
122
(400)
122
(400)
6
0
1
6
0
(0)
76,000
107,100
Maximum
6
91
(300)
305
(1,000)
305
(1,000)
6
3
1
6
169
(1,816)
204,500
288,400
Cost per unit3
Minimum Maximum
1,000 2,000
259
(79)
49.2
(15)
95.1
(29)
1,360
3,900
1,300
3,000
501
(46.5)
Footnotes at end of table.
(continued)
-------
TABLE B-6. (continued)
CO
i
IV)
o
Model plant 1
Cost element
Annual i zed costs
Maintenance, overhead (9%)n
Utilities1
Taxes, insurance (4%)
Operating labor J
Capital recovery (8.86%)k
Total annual ized cost
Minimum
1,860
92
825
7,240
1,830
11,800
Maximum
8,340
92
3,710
7,240
8,210
27,600
Model plant 2
Minimum
4,980
451
2,210
7,240
4,900
19,800
Maximum
12,700
451
5,680
7,240
12,600
38,700
Model plant 3 Cost per unit3
Minimum
9,640
902
4,280
7,240
9,490
31,600
Maximum Minimum Maximum
26,000
902
11,500
7,240
25,600
71,200
Where a range of unit costs was used, the high and low values are shown as maximum and minimum; where only a single
value was used, it is shown in the minimum column.
Pipe 9 $236/m or $72/ft and pipe supports & $23/m or $7/ft.
Pipe & $72.1/m or $22/ft and pipe supports & $23/m or $7/ft.
3-way valves, 7.6 cm (3 in)-$700; and pressure/vacuum relief valves, 7.6 cm (3 in)-$660.
Includes flowmeter with alarm, pressure gauge, and temperature gauge.
Capital cost includes subcontractor overhead and profit and contractor material markup.
Total capital cost includes construction fee, contingency, engineering, and startup (41%).
Maintenance and overhead are 5% and 4% of total capital cost, respectively.
Electricity at $0.05/kWh.
For 30 min/day at $39.69/h.
Capital recovery factor for 20-yr lifetime at 6.2%.
-------
I
ro
TABLE B-7. COSTS FOR GAS BLANKETING OF LIGHT-OIL CONDENSER, LIGHT-OIL DECANTER
WASH-OIL DECANTER, AND CIRCULATION TANK
(All Costs in 1984 Dollars)
Cost element
Pressure tap
10- to 15-cm (4- to 6-1n)
pipe,b m
(ft)
Plug valve, 15 cm (6 in)
Valves0
Seal vessels
Flame arrestors
Capital costd
Total capital cost6
Annuali zed costs
Maintenance, overhead (9*)f
UtilitiesS
Taxes, insurance (4%)
Capital recovery (8.86%)n
Total annual ized cost
Model
Minimum
1
61
(200)
1
6
6
6
60,000
84,500
7,600
1,100
3,400
7.500
19,500
plant 1
Maximum
1
183
(600)
1
6
6
6
118,900
167,600
15,100
3,200
6,700
14,800
39,800
Model
Minimum
1
122
(400)
1
8
8
8
98,000
138,200
12,400
2,100
5,500
12,200
32,300
plant 2
Maximum
1
244
(800)
1
8
8
8
156,900
221,200
19,900
4,200
8,800
19,600
52,600
Model
Minimum
1
183
(600)
1
13
13
13
149,000
210,000
18,900
3,200
8,400
18,600
49,100
plant 3 Cost per unit3
Maximum Minimum Maximum
1 3,800
305 483.1
(1,000) (147.25)
1 900
13 1,800
13 1,500
13 1,000
207,900
293,100
26,400
5,300
11,700
26,000
69,300
vatue wasaut!d0fun!< T'S was "sed! the h1^ and >°«
value was used, it is shown in the minimum column.
are shown as maximum and minimum; where only a single
ta Assumes 75* of pipe is 15-cm (6-in) header and 25% is 10-cm (3-in) vent lines.
c 3-way valves, 1U.2 cm (4 in)-$l,000; and pressure/vacuum relief valves, 10.2 cm (4 in)-$800.
d Capital cost includes subcontractor overhead and profit and contractor material markup.
e Total capital cost includes construction fee, contingency, engineering, and startup (41%).
f Maintenance and overhead are 5* and 4% of total capital cost, respectively.
9 Steam at $18.3/My.
h Capital recovery factor for 20-yr lifetime at 5.2%.
-------
CO
I
ro
IN3
TABLE B-8 COSTS OF WASH-OIL VENT SCRUBBER FOR LIGHT-OIL CONDENSER, LIGHT-OIL
DECANTERS, WASH-OIL DECANTERS, AND CIRCULATION TANKS
(All Costs in 1984 Dollars)
Cost element
Scrubber vessels
10.2-cm (4-in) vent pipe,b
2.5-cm (1-in) wash-oil
supply pipe, m
(ft)
5.1-cm (2-in) wash-oil
drain pipe,c m
(ft)
Valvesd
Seal vessels
Pump
Instrumentation6
Capital costf
Total capital cost9
Model p
Minimum
6
m 110
(ft) (360)
30.5
(100)
30.5
(100)
6
6
0
1
83,300
117,500
lant 1
Maximum
6
110
(360)
122
(400)
122
(400)
6
6
2
1
114,200
161,100
Model j)
Minimum
8
146
(480)
91.4
(300)
91.4
(300)
8
8
0
1
119,800
168,900
lant 2
Maximum
8
146
(480)
244
(800)
244
(800)
8
8
2
1
164,100
231,300
Model j
Minimum
13
238
(780)
122
(400)
122
(400)
13
13
0
2
190,600
268,700
jlant 3
Maximum
13
238
(780)
305
(1,000)
305
(1,000)
13
13
4
2
253,400
357,200
Cost per unit3
Minimum Maximum
1,000 2,000
449.5
(137)
65.6
(20)
121.4
(37)
1,800
1,500
3,900
2,500
Footnotes at end of table.
(continued)
-------
TABLE B-8. (continued)
co
i
ro
CO
Cost element
Annuali zed costs
Maintenance, overhead (9%)n
Utilities1
Taxes, insurance (4%)
Operating labor J
Capital recovery (8.86%)k
Total annuali zed cost
a Where a range of unit costs
Model
Minimum
10,600
1,400
7,700
7,240
10,400
34,300
was used,
plant 1
Maximum
14,500
1,400
6,440
7,240
14.300
43,900
the high
Model plant 2
Minimum
15,200
1,920
6,750
7,240
15,000
46,100
Maximum
20,800
1,920
9,250
7,240
20,500
59,700
and low values are shown
Model plant 3 Cost oer unit3
Minimum
24,200
3,170
10,700
14,500
13,800
76,400
as maxim
Maximum Minimum Maximum
32,200
3,170
14,300
14,500
31,700
95,800
lum and minimum: where onlv a sinalp
value was used, it is shown in the minimum column.
b Pipe, steam traced
-------
TABLE B-9. COSTS FOR GAS BLANKETING OF LIGHT-OIL AND BTX STORAGE TANKS
(All Costs in 1984 Dollars)
ro
Cost element
10- to 15-cm (4- to 6-in)
p1pe,b m
(ft)
Pipe supports, m
(ft)
Seal tanks
Flame arresters
Capital costd
Total capital cost6
Annual 1 zed costs
Maintenance, overhead (9%)f
UtilitiesS
Taxes, insurance (4%)
Capital recovery (8.86%)h
Total annual i zed cost
Model
Minimum
49
(160)
4
49
(160)
4
4
41,900
59,100
5.32U
850
2,360
5.230
U.800
plant 1
Maximum
183
(600)
4
183
(600)
4
4
123,600
174,200
15,700
3,170
6,970
15.400
41,300
Model plant 2
Minimum
61
(200)
9
61
(200)
9
9
69,600
98,100
8,830
1,060
3,920
8,690
22,500
Maximum
259
(850)
9
259
(850)
9
9
189,400
267,000
24,000
4,490
10,700
23.700
62,900
Model plant 3 Cost per unit3
Minimum
122
(400)
15
122
(400)
15
15
126,200
177,900
16,000
2,110
7,120
15,800
41,000
Maximum Minimum Maximum
335 483.1
(1,100) (147.25)
15 1,800
335 23 98.4
(1,100) (7) (30)
15 1,500
15 1,000
259,500
365.900
32,900
6,810
14,600
32,400
85,800
a Where a ranye of unit costs was used, the high and low values are shown as maximum and minimum; where only a single
value was used, it is shown in the minimum column.
b Assumes 75% of pipe is 15-cm (6-in) header and 25% is 10-cm (3-in) vent lines.
c 3-way valves, 10.2 cm (4 in) - $1.000; and pressure/vacuum relief valves, 10.2 cm. (4 in) - $800.
d Capital cost includes subcontractor overhead and profit and contractor markup.
e Total capital cost includes construction fee, contingency, engineering, and startup (41%).
f Maintenance and overhead are 5% and 4% of total capital cost, respectively.
y Steam at $18.3/Mg.
n Capital recovery factor for 2U-yr lifetime at 6.2%.
-------
en
TABLE B-10. COSTS OF WASH-OIL VENT SCRUBBER FOR LIGHT-OIL AND BTX STORAGE TANKS
(All Costs In 1984 Dollars)
Cost element
Scrubber vessels
10-cm (4-in) vent pipeb,
2.5-cm (1-in) wash-oil
line, m
(ft)
b.l-cm (2-in) wash-oil
drain,0 m
(ft)
Pumps
Valvesd
Vessel sealing
Instrumentation6
Capital costf
Total capital cost9
Model
Minimum
4
m 61
(ft) (200)
30.5
(100)
30.5
(100)
0
4
4
1
52,800
74,400
plant 1
Maximum
4
61
(200)
183
(600)
183
(600)
2
4
4
1
93,100
131,300
Model
Minimum
9
137
(450)
30.5
(100)
30.5
(100)
0
9
9
1
108,600
153,100
plant 2
Maximum
9
137
(450)
213
(700)
213
(700)
2
9
9 .
1
159,600
225,000
Model
Minimum
15
229
(750)
61
(200)
61
(200)
0
15
15
2
183,700
258,900
plant 3
Maximum
15
229
(750)
244
(800)
244
(800)
4
15
15
2
248,500
350,300
Cost per unit3
Minimum Maximum
1,000 2,000
449.5
(137)
(20)
(37)
3,900
1,800
1,500
2,500
Footnotes at end of table.
(continued)
-------
TABLE B-10. (continued)
Model plant 1
Model plant 2
Model plant 3
Cost per unit3
rv>
cr>
Cost element
Annual i zed costs
Maintenance, overhead (9%)h
Utilities1
Taxes, insurance (4%)
Operating laborJ
Capital recovery (8.86%)k
Total annual i zed cost
Minimum
6,700
794
2,980
7,240
6,600
24,300
Maximum
11,800
794
5,250
7,240
11,600
36,700
Minimum
13,800
1,800
6,120
7,240
13,600
42,500
Maximum
20,200
1,800
9,000
7,240
19,900
58,200
Minimum
23,300
2,980
10,400
14,500
22,900
74,100
Maximum Minimum Maximum
31,500
2,980
14,000
14,500
31,000
94,000
a Where a range of unit costs was used, the high and low values are shown as maximum and minimum; where only a single
value was used, it is shown in the minimum column.
D Pipe 0 $426.5/m or $130/ft and pipe supports @ $23/m or $7/ft.
c Pipe B $98.4/m or $30/ft and pipe supports & $23/m or $7/ft.
d 3-way valves, 10.2 cm (4 in) - $1,000; pressure/vacuum relief valves, 10.2 cm (4-in) - $800.
e Includes flowmeter with alarm, pressure gauge, and temperature gauge.
f Capital cost includes subcontractor overhead and profit and contractor material markup.
9 Total capital cost includes construction fee, contingency, engineering, and startup (41%).
n Maintenance and overhead are 5% and 4% of total capital cost, respectively.
1 Steam at $18.3/Mg and electricity at $0.05/kWh.
J For 30 min/day/scrubber system at $39.69/h.
k Capital recovery factor for 20-yr lifetime at 6.2%.
-------
TABLE B-ll. COSTS FOR GAS BLANKETING OF TAR COLLECTING, STORAGE, AND OEWATERING TANKS
(All Costs In 1984 Dollars)
OT
PO
Cost element
15-cm (6-in) pipe, m
(ft)
Seal tanks
Tank roofs-dewaterlng, m2
(ft2)
Pipe supports, m
(ft)
Valves6
Capital cost0
Total capital costd
Annual! zed costs
Maintenance, overhead (9%)e
Utilitiesf
Taxes, insurance (4%)
Capital recovery (8.86%)9
Total annualized cost
Model
Minimum
61
(200)
5
0
(0)
61
(200)
5
47,400
66,800
6,020
1,150
2,670
5,920
15,800
plant 1
Maximum
152
(500)
5
49
(531)
152
(500)
5
115,700
163,100
14,700
2,880
6,530
14,500
38,500
Model
Minimum
91
(300)
10
0
(0)
91
(300)
10
84.100
118,600
10,700
1,730
4,740
10.500
27,700
plant 2
Maximum
762
(2,500)
10
75
(804)
762
(2.500)
10
414,400
584,300
52,600
14.400
23,400
51,800
142,100
Model
Minimum
122
(400)
16
0
(0)
122
(400)
16
126,000
177,700
16,000
2,310
7,110
15,700
41,100
Plant 3 Cost per unit*
Maximum Minimum Maximum
914 328.1
(3,000) (100)
16 1,400
169 500.5
(1.816) (46.5)
914 23 98.4
(3.000) (7) (30)
16 3,800
557,600
786,300
70,800
17,300
31,500
69.700
189,200
b From Table B-l, 3-way vaives, 15 cm (6 in) - $2.500; pressure/vacuum relief valves. 15 cm (6 in) - $1,300
c Capital cost includes subcontractors overhead and profit and contractor material markup.
d Total capital cost includes construction fee, contingency, engineering, and startup (41%).
e Maintenance and overhead are 5t and 4% of total capital cost, respectively.
f Steam at $18.3/My.
y Capital recovery factor for 2U-yr lifetime at 6.2%.
-------
TABLE li-12. COSTS OK WASH-OIL VENT SCRUBBER FOR TAR COLLECTING, STORAGE, AND OEWATEKING TANKS
(All Costs in 1984 Dollars)
00
I
oo
Cost element
Scrubber, heat exchanger,
separator
lb.2-cm (fa-in) vent
pipe,b m
(ft)
10.2-cm (4-1 nO wastewater
pipe, m
(ft)
2.5-cm (1-m) wash oil supply
pipe, m
(ft)
5.1-cm (2-in) wash oil
drain pipe,c m
(ft)
Seal tank
Tank roofs, dewatering, or
(ft2)
Pump
Valves'1
Valves and level control
Instrumentatione
Capital costf
Total capital cost^
Model
Minimum
62,700
122
(400)
30.5
(100)
30.5
(100)
30.5
(100)
5
0
(0)
1
5
1
1
159, 701)
225,200
plant 1
Maximum
62,700
122
(400)
61
(200)
152
(500)
152
(500)
5
49.3
(531)
1
5
1
1
210,300
2 96,500
Model
Minimum
144,000
244
(800)
30.5
(100)
91.4
(300)
91.4
(300)
10
0
(0)
1
10
1
1
318,600
449,200
plant 2
Maximum
144.000
244
(800)
61
(200)
640
(2,100)
640
(2,100)
10
74.7
(804)
1
10
1
1
443,500
625,300
Model
Minimum
234,000
390
(1,280)
30.5
(100)
122
(400)
122
(400)
16
0
(0)
2
16
2
2
511,100
720,600 1
plant 3
Maximum
234,000
390
(1.280)
61
(200)
853
(2,800)
853
(2,800)
16
169
(1,816)
2
16
2
2
709,400
,000,300
Cost per unit8
Minimum Maximum
351
(107)
272.3
(33)
49.2
(15)
95.1
(29)
1,400
500.5
(46.5)
11,000
3,800
2,000
2,500
Footnotes at end of table.
(continued)
-------
TABLE B-12. (Continued)
f\>
Cost element
Annual ized costs
Maintenance, overhead (9%)h
Utilities1
Taxes, insurance
Operating laborJ
Capital recovery1*
Total annualized cost
Model f
Minimum
20,300
9,890
9,000
7,240
20,000
66,400
slant 1
Maximum
26,700
10,300
11,900
7,240
26,300
82,300
Model |
Minimum
40,400
33,700
18,000
7,240
,39,800
139,200
Jlant 2
Maximum
56,300
34,100
25,000
7,240
55,400
178,100
Model j
Minimum
64,900
72,400
28,800
14,500
63,800
244,400
2.1 ant 3 Cost per unit3
Maximum Minimum Maximum
90,000
72,800
40,000
14,500
88,600
305,900
low values are shown as
b Pipe, steam traced 9 $328/m or $100/ft and pipe supports 0 $23/m or $7/ft.
c Pipe 9 $72.20/m or $22/ft and pipe supports 9 $23/m or $7/ft.
d 3-way valves, 15 cm (6 in)-$2,500, pressure/vacuum release valves, 15 cm (6 in)-$l,300.
e Includes flowmeter with alarm, pressure gauge, and temperature gauge.
f Capital cost includes subcontractor overhead and profit and contractor material markup.
9 Total capital cost includes construction fee, contingency, engineering, and startup (41%).
h Maintenance and overhead are 5% and 4% of total capital cost, respectively.
1 Steam at $18.3/Mg and electricity at $0.05/kWh.
J For 30 min/day/scrubber system at $39.69/h.
k Capital recovery factor for 20-yr lifetime at 6.2%.
-------
TABLE B-13. COSTS FOR COVERING LIGHT-OIL SUMP
(All Costs in 1984 Dollars)
co
i
oo
o
Cost element
Clean, cover, and seal, m2
(ft2)
7.6-cm (3-in) vent pipe, ra
(ft)
Capital costb
Total capital costc
Annualized costs
Maintenance, overhead (9%)d
Taxes, insurance (4%)
Capital recovery (8.86%)e
Total annualized cost
Model p
Minimum
3.3
(36)
4.6
(15)
1,700
2,390
215
96
212
523
>lant 1
Maximum
20.9
(225)
4.6
(15)
37,500
52,900
4,760
2,120
4,690
11,600
Model
Minimum
3.3
(36)
4.6
(15)
1,700
2,390
215
96
212
523
plant 2
Maximum
93
(1,000)
4.6
(15)
164,600
232,100
20,900
9,280
20,600
50,700
Model [
Minimum
6.7
(72)
9.1
(30)
3,400
4,790
431
192
424
1,050
>lant 3 Cost per unit3
Maximum Minimum Maximum
186 328 1,765
(2,000) (30.5) (164)
9.1 131
(30) (40)
329.200
464,200
41,800
18,600
41,100
101,500
a Where a range of unit costs was used, the high and low values are shown as maximum and minimum; where only a single
value was used, it is shown in the minimum column.
0 Capital cost includes subcontractor overhead and profit and contractor material markup.
c Total capital cost includes construction fee, contingency, engineering, and startup (41%).
d Maintenance and overhead are 5% and 4% of total capital cost, respectively.
e Capital recovery factor for 20-yr lifetime at 6.2%.
-------
TABLE B-14. COSTS FOR NITROGEN OR NATURAL GAS BLANKETING OF PURE BENZENE STORAGE TANKS
(All Costs in 1984 Dollars)
CO
OJ
Cost element
2.5-cm (1-in) gas supply, ^ m
(ft)
7.6-cm (3-in) vent pipe, m
(ft)
Pressure controller
Pressure reducers
Site preparation
10.2-cm (4-in) flame arrestors
Valves0
Tank sealing
Pipe supports, m
(ft)
Capital costsd
Total capital costs6
Annuali zed costs
Maintenance, overhead (9S)f
UtilitiesQ
Taxes, insurance (4%)
Operating laborh
Capital recovery (a.86%)1
Total annual i zed cost
Model p
Minimum
12.2
(40)
15.2
(50)
1
2
0
1
1
1
15.2
(50)
16,100
22,700
2,040
0
910
0
2,010
4,960
lant 1
Maximum
30.5
(100)
91.4
(300)
1
2
8,000
1
1
1
91.4
(300)
46,900
66,100
5,950
1,700
2,640
7,240
5.860
23,400
Model p
Minimum
30.5
(100)
30.5
(100)
1
2
0
3
3
3
30.5
(100)
28,100
39,600
3,560
0
1,580
0
3.510
8.650
ilant 2
Maximum
91.4
(300)
152
(500)
1
2
18,400
3
3
3
152
(500)
86,200
121,500
10,900
6,700
4,860
7,240
10.800
40,500
Model
Minimum
61
(200)
61
(200)
1
2
0
7
7
7
61
(200)
51,500
72,600
6,540
0
2,910
0
6,440
15,900
jjlant 3 Cost per unit3
Maximum Minimum Maximum
213 (88.6)
(700) (27)
244 164
(800) (50)
1 4,400
2 2,000
30,000
7 1 ,000
7 1,360
7 1,400
244 23 98.4
(800) (7) (30)
147,600
208,100
18,700
15,000
8,330
7,240
18,400
67,700
Footnotes at end of table.
(continued)
-------
TABLE B-14. (continued)
a Where a range of unit costs was used, the high and low values are shown as maximum and minimum; where only a single
value was used, it is shown in the minimum column.
5 Includes pipe supports dt $23/m ($7/ft).
c 3-way valves, 7.6 cm (3 in)-$700; and pressure/vacuum relief valves, 7.6 cm (3 in)-$660.
d Capital cost includes subcontractor overhead and profit and contractor material markup.
e Total capital cost includes construction fee, contingency, engineering, and startup (41%).
f Maintenance and overhead are 5% and 4% of total capital cost, respectively.
9 Nitrogen at $0.27/m3 (0.76/100 ft3). Includes rental of 5.7-m3 (1,500-gal) liquid nitrogen storage tank, vaporizer,
co and gas usage. Some plants are assumed to have a nitrogen source and others must purchase nitrogen.
CO
1X3 n For 30 min/day at $39.69/h when liquid nitrogen is used.
i Capital recovery factor for 20-yr lifetime at 6.2%.
-------
TABLE B-lb. COSTS OF HASH-OIL
(All Costs
VENT SCRUBBER FOR BENZENE STORAGE TANKS
in 1984 Dollars)
co
i
co
CO
Model plant 1
Cost element
Scrubber vessels
2.b-cm (1-in) wash-oil line,
5.1-cm (2-in.) wash-oil drain,
Pump
10-cm (4-in) vent pipe0, m
(ft)
Valvesd
10.2-cm (4-in) flame arrestors
Tank sealing
Instrumentation6
Capital costf
Total capita) costsS
Annuali zed costs
Maintenance, overhead (9%)n
Utilities1
Taxes, insurance (4%)
Operating laborJ
Capital recovery (16.3%)k
Total annual ized cost
Minimum
1
Maximum
i
i
m 30.5 183
(ft) (100)
b m 30.5
(ft) (100)
o
15.2
50
i
i
1
1
1
17,300
24,300
2,190
6
973
7,240
2,160
12,600
(600)
183
(600)
i
A
15.2
50
1
1
1
i
50,700
71,400
6,430
6
2,860
7,240
6.330
22,900
Model plant 2
Minimum
30.5
(100)
30.5
(100)
45.7
150
3
35,400
49,800
4,490
19
1,990
7,240
4,420
18,200
Maximum
3
213
(700)
213
(700)
2
45.7
(150)
3
1
80,400
113,300
10,200
19
4,530
7,240
10,040
32,000
Model
Minimum
7
61
(200)
61
(200)
0
107
(350)
7
7
7
1
77,300
108,900
9,800
43
4,360
7,240
9.650
31,100
plant 3
Maximum
7
244
(800)
244
(800)
2
107
(350)
7
7
7
1
126,300
178,000
16,000
43
7,120
7,240
15,800
46,200
Cost per unit3
Minimum Maximum
1,000 2,000
65.6
(20)
121.4
(37)
3,900
252.6
(77)
1,800
1,000
1,400
2,500
Footnotes at end of table.
(continued)
-------
TABLE B-lb. (continued)
a Where a range of unit costs was used, the high and low values are shown as maximum and minimum; where only a single
value was used, it is shown in the minimum column.
b Pipe & 98.4/m or $30/ft and pipe supports @ $23/m or $7/ft.
c Pipe e $229.6/m or $70/ft and pipe supports (? $23/m or $7/ft.
d 3-way valves, 10.2 cm (4 in)-$l,000; and pressure/vacuum relief valves, 10.2 cm (4 in)-$800.
e Includes flowmeter with alarm, pressure gauge, and temperature gauge.
f Capital cost includes subcontractor overhead and profit and contractor or material markup.
9 Total capital cost includes construction fee, contingency, engineering, and startup (41%).
n Maintenance and overhead are 5% and 4% of capital, respectively.
1 Electricity at $0.05/kWh.
J For 30 min/day at $39.69/h.
k Capital recovery factor for 20-yr lifetime at 6.2%.
-------
TABLE 8-16. COSTS FOR TAR BOTTOM FINAL COOLER
(All Costs in 1984 Dollars)
DO
i
GO
cn
Cost element
Tank-separator, decanter tank,
wastewater tank, and tar tank
Pumps - tar transfer, water skimmer
Piping and valves
Site preparation, modify existing
cooler, miscellaneous
Instrumentation (6.75% of equipment)
Electrical (10.5% of equipment)
Capital cost3
Total capital costb
Annual i zed cost
Maintenance, overhead (9%)c
Utilities'1
Taxes, insurance (4%)
Operating labor6
Capital recovery (8.86%)f
Total annual ized cost
Model t
Minimum
61,700
16,300
26,800
31,700
7,070
11,000
154.500
217,800
19,600
2,200
8,710
22,000
19.300
71,800
>lant 1
Maximum
61,700
16,300
82,800
31,700
10,800
16,900
220.100
310,400
27,900
5,040
12,400
22,000
27,500
94,900
Model p
Minimum
141,700
26,600
80,100
72,800
16,800
26,100
363,900
513,200
46,200
7,590
20.500
22,000
45.500
141,800
lant 2
Maximum
141,700
26,600
247,700
72,800
28,100
43.700
560,500
790,200
71,100
16,000
31,600
22,000
70,000
210,800
Model
Minimum
230,500
37,800
152,000
118,400
28,400
44,100
611.100
861,700
77,600
16,200
34,500
22 .000
76,300
226,600
plant 3
Maximum
230.500
37,800
470,000
118,400
49,800
77,500
984.000
1,387,400
124,900
32,300
55,500
22,000
122.900
357,500
a Capital cost includes subcontractor overhead and profit and contractor material markup.
Total capital cost includes construction fee, contingency, engineering, and startup (4U).
i\
c Maintenance and overhead are 5% and 4% of total capital cost, respectively.
d Steam at $18.3/My and electricity at $0.05 kWh.
6 For 1.5 h/day at $39.6y/h.
f Capital recovery factor for 20-yr lifetime at 6.2%.
-------
TABLE B-17. COSTS FOR WASH-OIL FINAL COOLER
(All Costs in 1984 Dollars)
CD
CO
en
Cost element
Final cooler
Instrumentation (6.75% of final
cooler capital cost)
Electrical (10.5% of final cooler
capital cost)
Capital cost3
Total capital costb
Annuali zed costs
Maintenance, overhead (9%)c
Utilitiesd
Makeup wash oil6
Taxes, insurance (4%)
Operating laborf
Capital recovery (8.86%)9
Total annual i zed cost
Model plant 1
708,000
47,800
74,300
830,100
1,171,000
105,400
46,200
7,600
46,800
69.000
103,700
378,700
Model plant 2
1,627,000
109,800
170,800
1,908,000
2,690,000
242,100
184,900
30,500
107,600
69,000
238,300
872,400
Model plant 3
2,646,000
178,600
277,800
3,102,000
4,374,000
393,700
416,100
68,500
175,000
69,000
387 ,600
1,509,900
a Capital cost includes subcontractor overhead and profit and contractor material markup.
D Total capital cost includes construction fee, contingency, engineering, and startup (41%).
c Maintenance and overhead are 5% and 4% of total capital cost, respectively.
d Steam at $18.3/Mg and electricity at $0.05/kWh.
e Estimated at $U.34/kg ($1.07/gal); based on losses to wastewater and light-oil crude residue.
f For 4.8 h/day at $39.69/h.
9 Capital recovery factor for 20-yr lifetime at 6.2%.
-------
TABLE B-18. COSTS OF MIXER-SETTLER FOR NAPHTHALENE PROCESSING AND HANDLING
(All Costs in 1984 Dollars)
CO
Model plant 1
Cost element
Tank-separator, decanter tank,
wastewater tank, and tar tank
Pumps - tar transfer, water skimmer
Piping and valves
Site preparation, modify existing
cooler, miscellaneous
Instrumentation (6.75% of equipment)
Electrical (10.5% of equipment)
Capital cost3
Total capital cost15
Annuali zed cost
Maintenance, overhead (9%)c
Utilitiesd
Taxes, insurance (4Z)
Operating labor6
Capital recovery (8.86%)f
Total annual i zed cost
Minimum
61,700
16,300
26,800
31,700
7,070
11,000
154,500
217,800
19,600
2,200
8,710
22,000
19,300
71,800
Maximum
61,700
16,300
82,800
31,700
10,800
16,900
220.100
310,400
27,900
5,040
12,400
22,000
27 ,500
94,900
Model plant 2
Minimum
141,700
26,600
80,100
72,800
16,800
26,100
363.900
513,200
46,200
7,590
20.500
22,000
45,500
141,800
Maximum
141,700
26,600
247,700
72,800
28.100
43,700
560.500
790,200
71,100
16,000
31,600
22,000
70,000
210,800
Model
Minimum
230,500
37,800
152,000
118,400
28,400
44.100
611.100
861.700
77,600
16,200
34,500
22,000
76,300
226,600
plant 3
Maximum
230,500
37,800
470,000
118,400
49,800
77,500
984,000
1,387,400
124,900
32,300
55,500
22,000
122,900
357,500
a Capital cost includes subcontract or overhead and profit and contractor or material markup.
b Total capital cost includes construction fee, contingency, engineering, and startup (41%).
c Maintenance and overhead are 5% and 45£ of total capital cost, respectively.
d Steam at $18.3/My and electricity at $0.05 kMh.
e For 1.5 h/day at $39.69/h.
f Capital recovery factor for 20-yr lifetime at 6.2%
-------
Appendix C
Economic Impact Analysis
-------
APPENDIX C
ECONOMIC IMPACT
This appendix addresses the economic impacts of the regulatory alterna-
tives for coke-oven by-product plants. It provides an updated version of
Chapter 9 of the background information document (BID), Benzene Emissions
from Coke By-Product Recovery Plants. This appendix includes revised
estimates of the economic impacts of the regulatory alternatives and more
recent information on the state of the coke industry. Where possible, data
are updated to 1984.
Section C.I presents a profile of the coke industry. Section C.2
contains a reanalysis of the impacts of the regulatory alternatives. These
alternatives are outlined in Table C-l. These impacts are measured against
the baseline state of control for all sources. Section C.3 presents poten-
tial socioeconomic and inflationary impacts.
C.I INDUSTRY PROFILE
C.I.I Introduction
Coke production is a part of Standard Industrial Code (SIC) 3312--
Blast Furnaces and Steel Mills. Coke is principally used in the production
of steel and ferrous foundry products, which are also part of the output of
SIC 3312. Thus, coke is both produced and principally consumed within
SIC 3312. Furthermore, many producers of furnace coke are fully integrated
iron- and steel-producing companies. Any regulation on coke production is
expected to have some impact on the entire blast furnaces and steel mills
industry with special emphasis on coke producers.
This profile has two purposes: (1) to provide the reader with a broad
overview of the industry and (2) to lend support to an economic analysis by
assessing the appropriateness of various economic models to analyze the
C-3
-------
TABLE C-l. COKE BY-PRODUCT PLANT CONTROL OPTIONS'
Control option
Emission source
Regulatory Alternative Regulatory Alternative
II III
Direct water final cooler
Tar bottom final cooler
Tar decanter, flushing-
liquor circulation tank,
tar-intercepting sump
Tar storage tanks and
dewatering tanks
Light-oil decanter-
condenser, wash-oil
circulation tank,
wash-oil decanter
Excess ammonia-liquor
storage tanks
Light-oil tanks and BTX
storage tanks
Benzene storage tanks
Light-oil sump
Pump seal leaks
Valve leaks
Exhauster leaks
Pressure relief device
leaks
Sampling connection system
leaks
Open-ended line leaks
Tar bottom final cooler
Coke-oven gas-
blanketing system
Coke-oven gas-
blanketing system
Coke-oven gas-
blanketing system
Coke-oven gas-
blanketing system
Coke-oven gas-
blanketing system
Wash-oil scrubber
Cover
Monthly inspection
Monthly inspection
Quarterly inspection
Rupture disc system
Closed-purge system
Cap or plug
Wash-oil final cooler
Wash-oil final cooler
Coke-oven gas-
blanketing system
Coke-oven gas-
blanketing system
Coke-oven gas-
blanketing system
Coke-oven gas-
blanketing system
Coke-oven gas-
blanketing system
Wash-oil scrubber
Cover
Monthly inspection
Monthly inspection
Quarterly inspection
Rupture disc system
Closed-purge system
Cap or plug
aThese regulatory alternative control options differ from the proposed
regulations.
C-4
-------
industry. Further, the profile provides some of the data necessary to the
analysis itself.
The industry profile comprises six major sections. The remainder of
this introduction, which constitutes the first section, provides a brief,
descriptive, and largely qualitative look at the industry. The remaining
five sections of the profile conform with a particular model of industrial
organizational analysis. This model maintains that an industry can be
characterized by its basic conditions, market structure, market conduct,
and market performance.
The basic conditions in the industry, discussed in the second and
third sections of this profile, are believed to be major determinants of
the prevailing market structure. Most important of these basic conditions
are supply conditions, which are largely technological in nature, and
demand conditions, which are determined by the attributes of the products
themselves.
The market structure and market conduct of the blast furnaces and
steel mills industry are examined in the fourth section. Issues addressed
include geographic concentration, firm concentration, integration, and
barriers to entry. Market structure is believed to have a major influence
on the conduct of market participants. Market conduct is the price and
nonprice behavior of sellers. Of particular interest is the degree to
which the industry pricing behavior can be approximated by the competitive
pricing model, the monopoly pricing model, or some model of imperfect
competition.
The fifth section of the industry profile addresses market perform-
ance. The historical record of the industry's financial performance is
examined, with some emphasis on its comparison with other industries. The
sixth section of the industry profile presents a discussion of industry
trends for the coke and steel sectors. The seventh section discusses
market behavior.
C.I.1.1 Definition of the Coke Industry. Coke production is a part
of SIC 3312—Blast Furnaces and Steel Mills, which includes establishments
that produce coke and those that primarily manufacture hot metal, pig iron,
silvery pig iron, and ferroalloys from iron ore and iron and steel scrap.
C-5
-------
Establishments that produce steel from pig iron, iron scrap, and steel
scrap and establishments that produce basic shapes such as plates, sheets,
and bars by hot rolling the iron and steel also are included in SIC 3312.1
The total value of shipments from SIC 3312 in 1982 was $36,931,900,0002 and
an approximate value for total coke production in 1982 was $3,220,Oil,000,3
or less than 10 percent of the total value of shipments.
Coke is produced in two types of plants: merchant and captive.
Merchant plants produce coke to be sold on the open market, and many are
owned by chemical or other companies. The majority of coke plants in the
United States are captive plants that are vertically integrated with iron
and steel companies and use coke in the production of pig iron. At the end
of 1984, 15 plants were merchant and 36 were captive, and merchant plants
accounted for only 12 percent of total coke production.4 5 For the economic
analysis, it is assumed that more than one plant may exist at a single
location.
C.I.1.2 Brief History of the Coke Industry in the Overall Economy.
Traditionally, the value of coke produced in the United States has con-
stituted less than 1 percent of the gross national product (GNP).6 7
During most of the 1950's, coke production was about 0.30 percent of GNP,
and during the 1960's and until the mid-1970's, coke production was only
about 0.20 percent or less of GNP. However, in 1974, coke production as a
percent of GNP rose to above 0.30 percent. This trend continued for the
next 2 years. By 1982, coke production was about 0.1 percent of GNP.3 8
Previously, U.S. coke exports had been greater than imports, but that
trend has fluctuated. The values of all U.S. imports and exports and U.S.
coke imports and exports are shown in Table C-2. From 1950 to 1972, coke
exports were much greater than coke imports, but after 1973, this trend was
reversed. In 1982 and 1983, exports again exceeded imports. Data for the
second quarter of 1984 indicate that coke imports are again on the rise.
Imports for the first two quarters of 1984 totaled 247,604 megagrams (Mg)
compared to 6,874 Mg for the same period in 1983, and to 32,000 Mg for all
of 1983.16 Exports for the first two quarters of 1984 and 1983 were
307,540 Mg, and 300,283 Mg, respectively, and they were 603,288 Mg for all
of 1983.1S
C-6
-------
TABLE C-2. COKE INDUSTRY FOREIGN TRADE3 9 10 »» 12 13 »«
Year
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
=====
a,.
Total U.S. imports,
IO9 $a
8.9
11.0
10.7
10.9
10.2
11.4
12.6
13.0
12.8
15.2
14.7
14.7
16.4
17.1
18.7
21.4
25.5
26.8
33.2
36.0
39.9
45.6
55.8
70.5
103.7
98.0
124.0
151.9
176.0
212.0
249.7
265.1
243.9
258.0
Coke imports
for consumption,
IO6 $a
5.3
1.9
4.5
1.7
1.3
1.4
1.5
1.5
1.6
1.4
1.5
1.5
1.9
2.0
1.5
1.4
1.8
1.7
1.9
3.4
3.5
5.0
4.6
39.3
193.2
156.5
111.1.
137. 9^
410.9°
340. l£
52'°b e
9'2h
1.9b
Coke imports
as a share of
total imports,
0.06
0.02
0.04
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.06
0.19
0.16
0.09
0.09
0.23
0.16
0.02
0.02
0.004
0.0007
U.S. exports,
IO9 $
10.3
15.0
15.2
15.8
15.1
15.5
19.1
20.9
17.9
17.6
20.6
21.0
21.7
23.3
26.5
27.5
30.3
31.5
34.6
38.0
42.5
43.5
49.4
71.4
98.3
107.1
114.7
120.8
142.1
184.5
224.2
237.0
212.3
200.5
Coke exports,
IO6 $a
6.2
17.7
13.7
9.3
6.2
8.2
11.5
14.4
7.1
8.7
6.9
8.2
7.4
8.3
10.1
16.3
23.4
16.5
18.6
38.5
78.9
44.8
30.7
33.1
43.6
74.7
66. 7b c
68.9b'd
15.0°
12-9?
11. 3*
13. 8^
8.3b
Coke exports
as a share of
total exports,
0.06
0.12
0.09
0.06
0.04
0.05
0.06
0.07
0.04
0.05
0.03
0.04
0.03
0.04
0.04
0.06
0.08
0.05
0.05
0.10
0.19
0.10
0.06
0.05
0.04
0.07
0.06
0.06
0.05
0.01
0.005
0.005
0.01
0.004
See Product SIC (331210) in References 11-13.
Defined as "Pitch coke, coke of coal, lignite, or peat."
eDefined as "Coal coke, calcined and not calcined "
Cumulative through November 1981. Annual cumulative value not available.
-------
The same pattern applies to the percentages of coke imports and exports
within total U.S. imports and exports. From 1950 to 1972, coke exports
were a larger percentage of total U.S. exports than coke imports were of
total U.S. imports. Again, from 1973 to 1981, this trend reversed, and
coke imports were a larger proportion of total U.S. imports than coke
exports were of total U.S. exports. Percentage shares of exports were
greater than imports in 1982 and 1983.
U.S. coke production always has been a substantial portion of world
coke production. This share has decreased during the past 30 years, as
indicated in Table C-3. From 1950 to 1977, world coke production generally
increased while U.S. coke production decreased. This trend explains the
decline in the U.S. percentage of world coke production.
C.I.1.3 Size of the Iron and Steel Industry. The value of shipments
of SIC 3312 has increased since 1960. There have been a few fluctuations
in this growth; for example, as shown in Table C-4, the 1965 value of
shipments of SIC 3312 was the highest value between 1960 and 1972. Since
1972, the value of shipments has remained around $30 million, with the
highest value being $35 million (1972 dollars) in 1974. After reaching
another peak of $34 million (1972 dollars), the value of shipments declined
to a 23-year low of about $18 million (1972 dollars). This result
reflected conditions in the steel industry. In 1982, the steel industry
sustained record financial losses close to $3.2 billion (1982 dollars).23
In 1983, an additional $3.6 billion was lost.24
For SIC 3312, Table C-5 shows the value added by manufacture, the
total number of employees, and the value added per employee. Current and
constant (1972) dollar figures are included. Both the total value added by
manufacture and the value added per employee peaked in 1974, the same year
in which the value of shipments for this industry was the highest. The
increasing value added per employee might indicate that this industry is
changing to a more capital-intensive production process. This aspect is
discussed in Section C.I.6.
C.I.2 Production
C.I.2.1 Product Description. Two types of coke are produced: furnace
coke and foundry coke. Furnace coke is used as a fuel in blast furnaces;
C-8
-------
TABLE C-3. COKE PRODUCTION IN THE WORLD6 17 1S
Year
World production,
106 Mg
U.S. production,
ID6 Mg
U.S. production
as a share of
world production,
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978b
1979°
182.3
204.1
208.9
225.6
211.5
242.3
256.8
266.1
255.0
260.4
279.7
272.0
272.9
281.7
298.5
310.3
310.4
303.9
315.8
335.8
350.5
342.7
340.5
365.8
367.4
363.3
367.2
373.5
364.7
341.0
65.9
71.9
62.0
71.5
54.4
68.3
67.6
69.0
48.6
50.7
51.9
46.9
47.1
49.3
56.4
60.7
61.2
58.6
57.8
58.8
60.3
52.1
54.9
58.4
55.9
51.9
52.9
48.5
44.5
48.0
36.1
35.2
29.7
31.7
25.7
28.2
26.3
25.9
19.1
19.5
18.6
17.2
17.3
17.5
18.9
19.6
19.7
19.3
18.3
17.5
17.2
15.2
16.1
16.0
15.2
14.3
14.4
13.0
12.2
14.1
Oven and beehive coke combined.
Information on world coke production not available after 1979.
C-9
-------
TABLE C-4. VALUE OF SHIPMENTS, SIC 33122 19 20 21 22
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
Current dollars,
106
15,738.8
14,873.3
15,571.6
16,418.0
18,840.1
20,841.7
21,193.9
19,620.6
21,161.1
22,299.0
21,501.6
21,971.3
23,946.7
30,365.5
41,671.7
35,659.8
39,684.1
41,897.8
49,055.4
55,695.8
50,303.9
57,472.9
36,931.9
1972 Dollars,
106
22,981.7
21,468.4
22,071.7
22,933.4
25,914.9
28,043.2
27,610.6
24,829.9
25,628.1
25,713.8
23,535.0
22,882.0
23,946.7
28,700.9
35,917.7
28,038.8
29,643.8
29,645.4
32,879.0
34,358.9
28,244.8
29,473.3
17,884.7
C-10
-------
TABLE C-5. VALUE ADDED, SIC 33122 19 20 21 22
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
Value added by
Current dollars
106
6,844.4
6,546.3
6,620.9
7,506.4
8,479.6
9,379.8
9,643.6
8,910.1
9,275.8
9,853.2
9,350.5
9,563.1
10,304.7
12,769.4
17,425.8
13,356.2
14,755.5
15,021.4
19,085.7
21,039.0
18,632.2
20,100.2
manufacture
1972 dollars,
106
9,965.6
9,449.0
9,384.7
10,485.3
11,663.8
12,620.8
12,563.3
11,275.8
11,233.9
11,362.1
10.234.8
9,959.5
10,304.7
12,069.4
15,019.7
10,501.8
11,022.3
10,628.6
12,792.0
12,979.0
10,461.6
10,307.8
Employees,
103
550.0
503.4
502.2
500.5
532.9
565.4
559.4
533.1
533.1
537.7
526.5
482.2
469.1
502.1
518.0
451.3
451.9
441.4
443.5
451.2
402.9
390.3
Value added
per employee —
1972 dollars,
103
18.1
18.8
18.7
20.9
21.9
22.3
22.5
21.2
21.1
21.1
19.4
20.7
22.0
24.0
29.0
23.3
24.4
24.1
28.8
28.8
26.0
26.4
C-ll
-------
foundry coke is used as a fuel in the cupolas of foundries. Coke also is
used for other miscellaneous processes such as residential and commercial
heating. In 1983, only 3 percent of all coke used in the United States was
used for these miscellaneous purposes, 92 percent was used in blast furnaces,
and the remaining 5 percent was used in foundries.25 Time-series data for
the percent of total U.S. consumption attributable to each use from 1950 to
1980 are shown in Figure C-l.
C.I.2.2 Production Technology. Coke is typically produced from coal
in a regenerative type of oven called the by-product oven. The type of
coal used in coke production and the length of time the coal is heated
(coking time) determine the end use of the coke. Both furnace and foundry
coke usually are obtained from the carbonization of a mixture of high- and
low-volatile coals. Generally, furnace coke is obtained from a coal mix of
10 to 30 percent low-volatile coal and is coked an average of 18 hours, and
foundry coke is obtained from a mix of 50 percent or more low-volatile coal
and is coked an average of 30 hours.
The first by-product oven in the United States was built in 1892 to
produce coke and to obtain ammonia to be used in the production of soda
ash. In such ovens, the by-products of carbonization (such as ammonia,
tar, and gas) are collected instead of being emitted into the atmosphere as
they were in the older, beehive ovens.
The total amount of coke that can be produced each year is restricted
by the number of ovens in operation for that year, and not all ovens are in
operation all of the time. Oven operators try to avoid closing down a
group of ovens for any reason because of the time and energy lost while the
ovens cool and reheat and because of the oven deterioration that results
from cooling and reheating. However, it is estimated that, at any time,
approximately 5 to 10 percent of existing coke-oven capacity is out of
service for rebuilding or repair.28 In a report written for the Department
of Commerce, Father William T. Hogan estimated the potential annual maximum
capacity of U.S. oven coke plants as of July 31, 1979.29 Hogan assumed
that almost 10 percent of his estimate of total capacity would be out of
service at any given time; therefore, he subtracted the out-of-service
capacity from total capacity to obtain maximum annual capacity. The actual
C-12
-------
o
I
O
Ul
C/l
D
Ul
X
O
O
u.
O
U)
O
tc
ui
a.
92
00
QB
86
84
A
/
»
/FURNACE
12
10
- > OTHER USES
v>v
A
\ / FURNACE
FOUNDRY
J—i l
OTHER
USES
50 62 54 56 68 60 62 64 66 68 70 72 74 76 70 00
YEAR
Figure C-1. Uses of oven coke as percents of total coke consumption.6-26-27
-------
number of ovens that are out of service in a given year varies greatly. In
December 1983, 112 of 6,978 ovens, or 1.6 percent, were being rebuilt or
repaired, and annual capacity totalled 35,575,000 Mg.30 In November 1984,
1,756 of 8,204 ovens, or 21.4 percent, were out of service, and annual
capacity totalled 51,180,000 Mg.5 Table C-6 presents the data for November
1984.
In actuality, ovens that are removed from service and placed on "hot
idle" status are those likely to be returned to production in the short
term. Ovens that are placed on "cold idle" status are less likely to be
returned to service and, historically, have not been returned to service.
The capacity of these ovens is included in a plant's total capacity for
bookkeeping purposes even though the ovens may be scheduled for demoli-
tion.31
Within the limits of the number of ovens available for coking, both
furnace and foundry coke production levels vary. Some ovens that produce
furnace coke can be switched to produce foundry coke by changing the coal
mix and increasing the coking time. Furthermore, some ovens that produce
foundry coke could be changed to produce furnace coke by changing the coal
mix and decreasing the coking time. Also, some variation in the combina-
tion of flue temperature and coking time is possible for either type of
coke. A shorter coking time results in greater potential annual produc-
tion.
C.I.2.3 Factors of Production. Table C-7 provides a typical labor
and materials cost breakdown for furnace coke production. Coal is the
major material input in the production of coke. In 1979, greater than 61
percent of the coal received by coke plants was from mines that were company
owned or affiliated.33 In this same year, 14 States shipped some coal to
coke plants outside their borders.34 Of the coal received by domestic coke
plants, over 81 percent came from West Virginia, Kentucky, Pennsylvania,
and Virginia.34 Any potential adverse impact on the coke industry probably
will have some impact in these States. A total of 33.6 million Mg of
bituminous coal was carbonized in 1983.3S
Table C-8 shows employment in the by-product coke industry from 1950
to 1970 and the percentage of total SIC 3312 employees in the by-product
coke industry. This table shows decreasing employment in the by-product
C-14
-------
TABLE C-6. MAXIMUM ANNUAL CAPACITY OF OVEN COKE PLANTS
IN THE UNITED STATES IN NOVEMBER 1984s
In existence
Furnace plants
Foundry plants
Total
Out of service3
Furnace plants
Foundry plants
Total
In operation
Furnace plants
Foundry plants
Total
Number of
batteries
105
35
140
(25)
(2)
(27)
80
33
113
Number of
ovens
6,638
1,566
8,204
(1,646)
(110)
(1,756)
4,992
1,456
6,448
Capacity,
Mg
44,810,000
6,370,000
51,180,000
(9,828,000)
(402,000)
(10,230,000)
34,982,000
5,968,000
40,950,000
Batteries and ovens down for rebuilding and repair, or on cold idle prior
to permanent closure.
Defined as "online" or "on hot idle."
C-15
-------
TABLE 07. TYPICAL COST BREAKDOWNS: FURNACE COKE PRODUCTION AND
HOT METAL (BLAST FURNACE) PRODUCTION32
Furnace coke production Percent of cost
Labor and materials
Coking coal 77.1
Coal transportation 9.4
Labor (operation and maintenance) 6.6
Maintenance materials 6.9
Total labor and material costs 100.0
Hot metal production
Charge metal lies
Iron ore
Agglomerates
Scrap
Fuel inputs
Coke
Fuel oil
Limestone fluxes
Direct labor
Maintenance
General expenses
Total labor and material costs 100.0
C-16
-------
TABLE C-8. EMPLOYMENT IN THE BY-PRODUCT COKE INDUSTRY36
Year
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971a
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
Number of employees
20,942
22,058
21,919
21,011
17,944
19,595
19,318
19,203
15,654
15,865
15,779
13,106
12,723
12,696
13,021
14,003
13,745
13,662
14,136
13,617
13,997
11,955
11,127
11,121
11,207
12,109
11,047
10,196
10,578
10,477
9,673
8,846
6,778
Percentage of all
employees in SIC 3312
NA
NA
NA
NA
NA
NA
NA
NA
3.06
3.13
2.87
2.60
2.53
2.54
2.44
2.48
2.46
2.56
2.65
2.53
2.66
2.48
2.37
2.21
2.16
2.68
2.44
2.31
2.38
2.32.
2.40
2.27
2.28
NA = Not applicable.
a
Figures for 1971-1982 are estimates. See text for more detail
C-17
-------
coke industry. A similar decline in employment has occurred in SIC 3312.
Unfortunately, employment data for the by-product coke industry are not
available after 1970; however, these figures can be estimated by regressing
employment in the by-product coke industry on total iron and steel industry
employment and on the ratio of coke used in steel production.* These
estimates are also shown in Table C-8.
C.I.3 Demand and Supply Conditions
Domestic consumption of coke from 1950 to 1980 is graphed in Figure C-2.
In the early 1950's, the amount of coke consumption was fairly large; an
average of 65 million Mg was consumed annually between 1950 and 1958. The
late 1950's and early 1960's showed a sharp decrease in coke consumption,
with an average of only 48 million Mg consumed annually. Domestic con-
sumption of coke increased during the mid-19601s to mid-19701s to an annual
figure of 57 million Mg, but it did not reach the 1950 to 1957 level. The
late 1970's showed another slump in coke consumption.
The variation in coke consumption shown in Figure C-2 has both cyclic
and trend components. The demand for coke is derived from demands for iron
and steel products, and these demands are sensitive to the performance of
the overall economy. Cycles in coke demand are linked to cycles in aggre-
gate demand or cycles in demand for particular products such as automobiles.
The trend component in coke consumption results from changes in blast
furnace production techniques. Coke is used as a fuel in blast furnaces,
but it is not the only fuel that can be used. Coke-oven gas, fuel oil, tar
and pitch, natural gas, and blast furnace gas have all been used as supple-
ments to coke in heating the blast furnaces. The increased use of these
supplemental fuels over the past 20 years has caused the amount of coke
used per ton of pig iron produced (the coke rate) to decrease. Other
causes of the decline in coke rate are increased use of oxygen in the blast
furnaces and use of higher metallic content ores. Table C-9 shows U.S. pig
iron production, coke consumed in the production of pig iron, and the coke
rate for 1950 to 1983. (Data limitations make it difficult to calculate
the foundry coke rate in cupola production.)
^Regressions performed by Research Triangle Institute (RTI) in 1980 and 1985.
C-18
-------
o
i
70
65
o
o
u.
O
to
5
60
g 55
tu
O
50
45
35
_L
JL
_L
JL
J-
J_
4-
JL
50 52 54 56 58 60 62
64 66
YEAR
J_
68 70 72 74 76 78 80
Figure C-2. U.S. apparent consumption of coke.6-26
-------
TABLE C-9. COKE RATE3 18 2S 37 38
Year
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
Pig iron production,
103 Mg
58,514
63,756
55,618
67,906
52,570
69,717
68,067
71,128
51,851
54,622
60,329
58,834
59,546
65,173
77,527
80,021
82,815
78,744
80,529
. 86,186
82,820
73,829
80,628
91,915
86,616
72,322
79,788
73,931
79,552
Coke used in
blast furnaces,
103 Mg
51,403
55,362
49,386
58,880
46,861
60,675
58,279
60,861
42,898
44,107
46,462
42,855
42,298
44,596
51,076
53,576
54,653
51,300
51,399
55,065
54,754
48,269
50,214
54,791
51,154
44,375
47,678
44,292
47,889
Coke rate
0.86
0.87
0.89
0.87
0.89
0.87
0.86
0.86
0.83
0.81
0.77
0.73
0.71
0.68
0.66
0.67
0.66
0.65
0.64
0.64
0.66
0.65
0.62
0.60
0.59
0.61
0.60
0.60
0.60
(continued)
C-20
-------
TABLE C-9 (continued)
Coke used in
Pig iron production, blast furnaces,
Year 103 Mg 103 Mg Coke rate
1979
1980
1981
1982
1983
78,926
62,325
66,951
39,282
46,267
45,862
37,583
37,832
21,918
25,009
0.58
0.60
0.56
0.56
0.54
e-2i
-------
Recently, there has been some concern about the ability of the United
States' coke-making capacity to support domestic steel production—the
major source of coke demand. The study conducted by Hogan and Koelble of
the Industrial Economics Research Institute at Fordham University indicates
that, in 1978, U.S. production of coke was 14.1 percent below domestic
consumption.39 Imports increased dramatically in that same year. Hogan
and Koelble attributed this decline in coke production to the abandonment
of coke ovens for environmental reasons and predicted a severe coke shortage
by 1982.40 This prediction was disputed in a Merrill Lynch Institutional
Report by Charles Bradford.41 The Bradford report attributed the lack of
adequate U.S. coke production in 1978 to two factors: (1) a coal miner's
strike, which caused the drawing down of stocks of coke when they should
have been increasing, and (2) the premature closing because of U.S. Envi-
ronmental Protection Agency (EPA) regulation of some coke ovens that
normally would have been replaced before they were closed.41 The Bradford
report stated that a survey of U.S. steel producers revealed that all of
the major steel producers were or soon would be self-sufficient with regard
to coke-making capacity.42 The Bradford explanation of 1978 coke imports
seems more reasonable because 1979 coke imports decreased about 1.6 million
Mg compared to the 1978 level.
The following values describe the situation in the 1980s with respect
to production, imports, and apparent consumption of coke (thousand mega-
grams).16
Year Production Imports Consumption Distributor Stock
7,009
5,556
7,141
4,024
2,776
1980
1981
1982
1983
1984a
41,851
38,815
25,506
23,413
14,446
598
478
109
32
248
37,447
39,975
23,384
27,080
14,886
aTwo quarters of 1984
Production is less than apparent consumption in 1981, 1983, and 1984.
For each of these years, stocks and imports more than accommodate the
shortfall. Coke producers were operating at 80 percent of total capacity
C-22
-------
in November 1984.5 Thus, it is unlikely that major shortages will develop
in the near future.
C.I.4 Market Structure
Market power, the degree to which an individual producer or groups of
producers can control market price, is of particular economic importance.
Market structure is an important determinant of market power. Pricing
behavior is relevant to the choice of the methodology used in assessing the
potential impacts of new regulations. It is important to determine if the
competitive pricing model (price equal to marginal cost) adequately des-
cribes pricing behavior for coke producers.
Any analysis of market structure must consider the characteristics of
the industry. This analysis addresses the number of firms producing coke;
the concentration of production in specific firms; the degree of inte-
gration in coke production; the availability of substitutes for coke; and
the availability of substitutes for the commodities for which coke is an
input to production. Also, some information on past pricing in the coke
industry is presented. These topics will be considered together with
financial performance (Section C.I.5) and trends (Section C.I.6) in asses-
sing market behavior (Section C.I.7).
C.I.4.1 Concentration Characteristics and Number of Firms. This
section describes various concentration measures that can be computed for
the furnace and foundry coke industries. Normally, concentration ratios
are used as an indication of the existence of market power. Although
concentration ratios are a useful tool for describing industry structure,
concentration should not be used as an exclusive measure of market power.
Many other factors (e.g., availability of substitutes, product homogeneity,
ease of market entry) determine a firm's ability to control market price.
As of November 1984, 23 companies operated by-product coke ovens.5 43
Twelve companies are integrated iron and steel producers; 11 companies are
merchant firms. These companies owned and operated a total of 51 coke
plants; 36 of these plants were captive and 15 of them were merchant. A
list of these companies, their plant locations, the major uses of coke at
each plant, and plant coke capacities is given in Table C-10. A plant site
may include more than one complete plant.
C-23
-------
TABLE C-10. COKE PLANTS IN THE UNITED STATES, November 1984s
o
i
no
Company name
Armco, Inc.
Bethlehem Steel Corp.
Rouge Steel
Inland Steel Co.
Interlace, Inc.
The LTV Steel Corp.
Lone Star Steel Co.d
National Steel Corp.e
Weir ton Steel Corp.
New Boston Coke Corp.
Plant location
Ashland, KY h
Middletown, OH (2)°
Bethlehem, PA
Burns Harbor, IN
Lackawanna, NY
Sparrows Point, MO
Dearborn, HI
E. Chicago, IN (3)
Chicago, IL
Aliquippa, PAC
Cleveland, OH (2)
E. Chicago, IN
Gadsden, AL
Pittsburgh, PA
S. Chicago, IL
Thomas, AL
Warren, OH
Lone Star, TX
Granite City, IL
Detroit, MI
Brown's Island, WV
Portsmouth, OH
Classification
of plant
Captive
Captive
Captive
Captive
Merchant
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Major uses of coke
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Coke capacity,
103 Mg/yr
963
1,776
2,253
1,790
1,292
3,506
778
3,715
582
1,218
1,760
948
758
1,792
563
315
945
507
868
1,397
1,097
364
Footnotes at end of table.
(continued)
-------
TABLE C-10 (continued)
o
en
Company name
U.S. Steel Corp.
Wheeling-Pittsburgh9
Steel Corp.
Jim Walter Corp.
Koppers Co. , Inc.
Shenango, Inc.
Alabama By-Products
Corp.
Carondelet Coke Corp.
Chattanooga Coke and.
Chemical Co. , Inc.
Citizens Gas and Coke
Utility
Plant location
Clairton, PA (4)
Fairfield, AL
Fairless Hills, PA
Gary, IN
Loral n, OH
Provo, UT
E. Steubenville, WV
Monessen, PA
Birmingham, AL
Erie, PA
Toledo, OH
Woodward, AL
Neville Island, PA
Tarrant, AL
Keystone, PA
St. Louis, MO
Chattanooga, TN
Indianapolis, IN
Classification
of plant
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Merchant
Merchant
Merchant
Merchant
Merchant
Merchant
Merchant
Merchant
Merchant
Merchant
Major uses of coke3
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace, foundry
Foundry, other
industrial
Foundry
Blast furnace, foundry
Blast furnace, foundry
Foundry, .
other industrial"
Foundry
Foundry,
other industrial
Foundry,
other industrial
Foundry
Coke capacity
103 Mg/yr
S a/ J
5,294
1,822
916
4,228
1,496
1,160
1,509
490
499
207
157
563
521
583
402
330
130
477
Footnotes at end of table
(continued)
-------
TABLE C-10 (continued)
o
i
ro
en
Company name
Detroit Coke Corp.
Empire Coke Co.
Indiana Gas and
Chemical Corp.
Tonawanda Coke Corp.
Plant location
Detroit, MI
Holt, AL
Terre Haute, IN
Buffalo, NY
Classification
of plant
Merchant
Merchant
Merchant
Merchant
Major uses of coke
Foundry
Foundry
Foundry,
other industrial
Foundry
Coke capacity
103 Mg/yr
617
161
132
299
aAn end use is considered a major use if it is at least 20 percent of the plant's total distribution of coke.
bNumbers in parentheses indicate the number of plants at that location. If no number is indicated, only one
plant exists at that location.
CLTV announced its intention in May 1985 to reduce production of steel at the Aliquippa, Pennsylvania, plant.
The plant may convert to a cold-idle status eventually.
^Northwest Industries, Inc., the parent company of Lone Star Steel, announced in April 1985 its merger with
Farley Industries.
eA merger between National Intergroup, Inc., the parent company of National Steel Corp., and Bergen Brunswig
Corp. fell through in April 1985, 2 weeks before its scheduled date. Some market consultants feel that
National Intergroup, Inc., is now a potential target for corporate raiders.
fMcLouth Steel Corp., the parent company of New Boston Coke Corp., is operating under Chapter 11 filed in 1981.
gWheeling-Pittsburgh Steel Corp. filed for Chapter 11 in April 1985.
Residential and commercial heating included in other industrial category.
^Chattanooga Coke and Chemical Co., Inc., is operating under Chapter 11 filed in March 1984.
-------
Reported capacities in Table C-10 are maximum, nominal figures, which
do not include any allowance for outage like that determined for the overall
industry in Table C-6. All but one of the largest plants are captive, and
most of the merchant plants have very small capacities. Furnace coke
production is concentrated in captive plants. Virtually all of the coke
used in foundries and in other industries was produced by merchant plants.
If coke plant sites were ranked according to capacity, the top 5 plant
sites and top 10 plant sites would have 37.1 percent and 54.6 percent of
total coke capacity, respectively.
By-product coke plants are concentrated in the States bordering on the
Ohio River, probably because of the coal in that area. Pennsylvania contains
12 plants, and Ohio and Indiana each have 8 plants.5
Table C-ll divides the United States into 11 coke-consuming and coke-
producing regions and shows the amount of coke produced in each region and
the locations of coke consumption in 1977. Most of the regions produce the
bulk of the coke they consume; only three regions produced less than 80
percent of their own consumption, and only one produced more than it needed
for its own consumption. Transportation of coke across long distances is
avoided whenever possible to reduce breakage of the product into smaller,
less valuable pieces and to minimize freight charges.46
The concentration of production or capacity in specific firms may have
economic importance. Table C-12 presents the percent of total capacity
owned by the largest 4 (of 23) firms. The four-firm concentration ratio
for the coke industry has increased over the years. In 1959, the four-firm
concentration ratio was 53.5 (the top four firms owned 53.5 percent of
total capacity)4?; in 1984 it was 69.9 percent. Consolidation of the
industry through mergers, acquisitions, and closures has encouraged this
trend.
In the preceding discussion, furnace and foundry coke production are
considered jointly. However, each existing coke battery may be considered
a furnace or foundry coke producer, based on the battery's primary use.
Separate capacity-based concentration ratios for the two types of coke are
calculated based on this allocation. The 1984 four-firm concentration
ratio for furnace coke is 75.4; the 1984 four-firm ratio for foundry coke
is 65.2.5
C-27
-------
TABLE C-ll. INTERREGIONAL COKE SHIPMENTS IN 1977<5
(103 megagraras)
i
l\3
00
Consuming region
Producing region
Alabama 2
California,
Colorado, Utah
Maryland, New York
Illinois
Indiana
Kentucky, Missouri,
Tennessee, Texas
Michigan
Minnestoa, Wisconsin
Ohio
Pennsylvania
Virgina,
West Virginia
TOTAL 2
AL
,228
0
0
0
0
14
0
0
0
9
0
,251
CA, CO
UT
27
2,668
3 4
0
5
18
0
6
4
0
0
2,731 4
MD,
NY
10
0
,392
0
0
0
0
0
0
51
0
,453
IL
81
0
123
1,424
69
15
0
269
138
1,241
0
3,360
IN
112
0
0
0
7,594
5
7
70
366
134
0
8,288
KY, MO,
TN, TX
465
0
22
0
35
928
1
1
379
3
8
1,842
MI
195
0
88
0
97
125
2,639
61
260
52
412
3,929
MN,
WI
7
0
0
0
11
0
0
158
0
0
0
176
OH
114
0
6
0
62
13
6
5
6,356
1,370
0
7,932
PA
1
0
8
0
3
0
0
1
2
10,257
214
10,556
VA.
wv
51
0
0
0
0
20
0
0
12
3
2,465
2,551
Total
3,361
2,668
4,642
1,424
7,876
1,138
2,653
571
7,517
13,120
3,099
48,069
-------
TABLE C-12. PERCENT OF COKE CAPACITY OWNED BY TOP FIRMS
NOVEMBER 19845
Firm
U.S. Steel, Inc.
Bethlehem Steel Corp.
The LTV Steel Corp.
Inland Steel Co.
Sum of largest four firms
Capacity,
103 Mg
14,916
8,841
8,299
3,715
35,771
Percent of
total capacity
29.14
17.27
16.22
7.26
69.89
C-29
-------
Concentration in the steel industry has economic relevance because a
large fraction of all furnace coke is produced by integrated iron and steel
companies. Historically, the eight largest steel producers have been
responsible for approximately 75 percent of industry production. However,
from 1950 to 1976, the share of production attributable to the top four
firms declined from 62 percent to 53 percent.48 In 1981, the seven largest
steel companies produced about 70 percent of steel made in the United
States.49
In summary, concentration exists in the production of both types of
coke and in steel production. However, the concentration probably is not
sufficient to guarantee market power, and many companies are involved in
the production of both coke and steel products. Other factors must be
considered in any final assessment of market power.
C.I.4.2 Integration Characteristics. When one firm carries out
activities that are at separate stages of the same productive process,
especially activities that might otherwise be performed by separate firms,
that firm is said to be vertically integrated. Through vertical integra-
tion, the firm substitutes intrafirm transfers for purchases from suppliers
and/or sales to distributors. A firm may seek to supply its own materials
inputs to ensure a stable supply schedule or to protect itself from monopo-
listic suppliers. The firm may seek to fabricate further or distribute its
own products to maintain greater control over the consuming markets or to
lessen the chance of being shut out of the market by large buyers or middle-
men. Therefore, the presence of vertical integration may constitute a
firm's attempt to control costs or ensure input supplies. Vertical integra-
tion does not guarantee market power (control over market price).
Many coke-producing firms, especially furnace coke producers, are
vertically integrated enterprises. As previously mentioned, 36 of the
existing coke plants are captive, i.e., they are connected with blast
furnaces and/or steel mills. In addition, many coke firms own coal mines,
and greater than 61.0 percent of the coal used in ovens was from captive
mines in 1979.33 Assurance of coal supply to coke production and coke
supply to pig iron production appears to be the motivation behind such
integration.
C-30
-------
One implication of vertical integration is that much of the furnace
coke used in the United States never enters the open market—it is consumed
by the producing company. Accordingly, the impact analysis for furnace
coke (Section C.2.2) uses an implied price for furnace coke based on its
value in producing steel products, which are transferred on the open market.
C.I.4.3 Substitutes. Substitutes for a given commodity reduce the
potential for market power in production of the commodity. The substitu-
tion of other inputs for coke in blast furnaces is somewhat limited, but
not totally unfeasible. In addition, electric arc furnaces, which do not
require coke, are becoming increasingly important in steel production. The
trend toward electric arc furnaces and minimills has eased entry into the
iron and steel industry, which in turn reduces market power.
Imported coke also can be substituted for domestically produced coke.
In fact, although U.S. iron and steel producers prefer to rely on domestic
sources of coke, coke imports have increased most recently. If the cost of
domestic coke increased substantially compared to the cost of imported
coke, U.S. iron and steel producers might attempt to increase imports even
more. Correspondingly, if costs of imported coke are reduced because of
improved foreign technology and productivity, reductions in foreign labor
cost, or other reasons, imports might become more desirable.
Furthermore, substitutes exist for the final products (iron and steel)
to which coke is an input. Increases in the price of coke and the result-
ing increases in the price of iron and steel products can lead to some
substitution of other materials for iron and steel, which also reduces
market power in the production of coke. Analogous substitutions for foundry
coke are possible, and cupola production of ferrous products, which uses
foundry coke, has competition from electric arc furnaces that do not use
coke. Hence, there is a technological substitute for foundry coke in the
manufacture of ferrous products. Furthermore, imported foundry coke can be
substituted for domestic foundry production. In conclusion, some substitu-
tion for coke is possible in the manufacture of both steel and ferrous
products.
C.I.4.4 Pricing History. As previously indicated, a significant
portion of all U.S. coke production is not traded on the market. However,
C-31
-------
the U.S. Bureau of Mines and the Energy Information Administration collect
annual data on coke production and consumption and give the quantity and
the total value of coke consumed by producing industries, sold on the open
market, and imported. Dividing total value by quantity yields an average
price for each of these categories. Time-series data on these three average
values are given in Table C-13. (Furnace and foundry coke are combined in
these figures.)
Also shown in Table C-13 are data on the average value of coal that is
carbonized in coke ovens. An examination of coke and coal prices reveals
that increases in coal prices generally coincide with increases in coke
prices. In fact, only 3 years show an increase in the price of coal that
was not accompanied by an increase in the price of the two categories of
coke. Although it is impossible to conclude from this trend that individual
firms have market power, it indicates that the industry can pass through
some increases in costs.
C.I.4.5 Market Structure Summary. Although there is no perfect
method for measuring the extent of market power, the preceding sections
addressed four characteristics used to measure the potential for market
power—concentration, integration, substitution, and historical price
trends. Concentration statistics indicated that some potential for market
power exists in the coke industry, yet these statistics are not conclusive
proof. Similarly, vertical integration in the steel industry is not con-
clusive in identifying the presence of market power because vertical inte-
gration is a method of controlling the cost and ensuring the quality and
supply of inputs. Finally, the possibility of substitution represents a
strong argument against the existence of extensive market power in the
coke-making industry.
C.I.5 Financial Performance
Financial data on the coke-producing firms or their parent firms,
including captive and merchant furnace and foundry producers, are shown in
Table C-14. Firms for which data are not available are noted.
Ten companies show negative earnings before interest and taxes. Of
these, nine are furnace coke producers, whose earnings reflect the disas-
trous years for the steel industry. As mentioned, in 1983, steel firms had
C-32
-------
TABLE C-13.
COMPARISON OF COAL PRICES AND DOMESTIC AND IMPORTED
COKE PRICES6 50 51 52 53
Average value of Average value of Average value of
coal carbonizeg fa oven coke used oven coke sold Average value of
in coke ovens, ' by producers, commercially, imported coke, >c
$/Mg $/Mg $/Mg
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
9.56
9.85
10.17
10.19
9.92
9.74
10.31
10.92
10.90
10.89
10.90
10.79
10.86
10.46
10.23
10.48
10.78
11.05
11.03
11.49
13.46
15.43
17.34
20.19
40.22
48.73
48.68
50.99
57.37
55.88
62.09
69.29
71.62
65.36
14.26
14.50
15.11
15.36
17.33
17.90
19.39
19.98
19.82
19.16
19.92
19.12
19.53
18.88
19.17
17.89
18.40
18.58
19.57
21.54
30.30
32.86
35.76
41.34
82.32
92.84
93.83
90.57
105.79
117.39
123.42
125.52
131.24
124.57
14.54
15.72
17.63
17.96
18.95
18.52
20.27
21.51
21.90
23.03
22.32
23.30
23.36
23.24
22.85
23.90
24.49
24.99
24.25
27.01
33.04
41.29
44.87
47.31
72.47
96.61
104.01
111.95
118.03
107.54
113.24
124. 34
126.24
124.67
13.34
13.17
15.96
12.02
11.98
12.26
12.38
14.43
14.25
12.89
13.06
13.44
14.42
14.78
16.10
16.95
20.60
20.41
22.31
21.36
25.46
31.93
27.70
40.16
60.14
94.84
93.35
— ~
94.32
87.12
89.59
84.61
61.45
Both furnace and foundry coke and the coals used to produce furnace and
foundry coke are included in these figures.
Market value at the oven (current dollars).
r»
""General customs value as reported by the U.S. Department of Commerce
(current dollars).
C-33
-------
TABLE C-14. FINANCIAL INFORMATION ON COKE-PRODUCING FIRMS, 1983
(million 1983 dollars)3 S4 5S 5S
Company name
Armco, Inc.
Bethlehem Steel Corp.
Ford Motor Co.
(Rouge Steel)
Inland Steel Co.
Interlake, Inc.
The LTV Steel Corp.
McLouth Steel Corp.1'-"
(New Boston Coke Corp.)
National Intergroup, Inc.
(National Steel Corp.)
Northwest Industries"1
(Lone Star Steel Co.)
Shenango Furnace Co. , Inc. '
(Shenango, Inc.)
U.S. Steel Corp.
Weirton Steel Corp.p
Wheeling-Pittsburgh Steel Corp.q
Jim Walter Corp."
Koppers Co. , Inc.
Alabama Byproducts Corp.
r
Carondelet Coke Corp.
Chattanooga Coke and Chemicals
Co., Inc.
Citizens Gas and Coke Utility
r t
Detroit Coke Corp. '
Indiana Gas and Chemical Corp.
McWane, Inc. (Empire Coke Co.)
f*
Tonawanda Coke Corp.
Net sales
4,165
4,898
44,455
3,046
835
4,578
11
2,993
1,608
145
16,869
1,000
772
2,025
1,566
229
k
17
316
k
64
k
k
EBITC
(526)
(239)
2,166
(177)
38
(252)
(0.09)
(177)
(104)
k
(1,208)
k
(72)
113
42
k
k
k
k
k
(0.2)
k
k
Cash flow
70
1299
5,542
106g
67
(164)9
k
161
1549
k
l,563g
k
(70)9
159
1749
192
k
k
91
k
k
k
k
Footnotes at end of table.
C-34
-------
TABLE C-14 (continued)
Company name
Armco, Inc.
Bethlehem Steel Corp.
Ford Motor Co.
Inland Steel Co.
Interlake, Inc.
The LTV Steel Corp.
McLouth Steel Corp.1'^'
(New Boston Coke Corp. )
National Intergroup, Inc.
(National Steel Corp.)
Northwest Industries
(Lone Star Steel Co.)
Shenango Furnace Co., Inc.n>0
(Shenango, Inc.)
U.S. Steel Corp.
Weirton Steel Corp.p
Wheeling-Pittsburgh Steel Corp.q
Jim Walter Corp. n
Koppers Co. , Inc."
Alabama Byproducts Corp.
Carondelet Coke Corp.1"
Chattanooga Coke and Chemicals
Co., Inc.5
Citizens Gas and Coke Utility
Detroit Coke Corp.r>t
Indiana Gas and Chemical Corp.
McWane, Inc. (Empire Coke Co.)
Tonawanda Coke Corp.1"
Annual
interest
expense
154
104
567
63
12
171
k
62
63
k
1,074
k
58
140
26
6
k
k
6
k
k
0.5
k
Total
assets
3,609
4,457
23,869
2,626
674
4,406
11
2,649
1,811
k
19,314
357
1,241
2,609
1,175
243
k
k
343
28
36
112
k
Long-
term
debt
832
1,134
2,713
788
116
1,560
3
606
451
10
7,164
149
514
1,151
233
52
k
k
145
19
0
21
k
Tangiblef
net worth
1,213
1,088
7,545
1,118
314
985
(4)
875
530
73
4,570
k
247
717
554
243
k
k
136
(3)
22
74
k
Footnotes at end of table. , i.. ""
C-35
-------
TABLE C-14 (continued)
b
Company name
Armco, Inc.
Bethlehem Steel Corp.
Ford Motor Co.
(Rouge Steel)
Inland Steel Co.
Interlake, Inc.
The LTV Steel Corp.
McLouth Steel CorpJ'J
(New Boston Coke Corp.)
National Intergroup, Inc.
(National Steel Corp.)
Northwest Industries
(Lone Star Steel Co.)
Shenango Furnace Co., Inc.n>0
(Shenango, Inc.)
U.S. Steel Corp.
Weirton Steel Corp.p
Wheeling-Pittsburgh Steel Corp.q
Jim Walter Corp."
Koppers Co. , Inc."
Alabama Byproducts Corp.
Carondelet Coke Corp.r
Chattanooga Coke and Chemicals
Co., Inc.5
Citizens Gas and Coke Utility
Detroit Coke Corp.r>t
Indiana Gas and Chemical Corp.
McWane, Inc. (Empire Coke Co.)
Tonawanda Coke Corp. r
Footnotes at end of table.
Net working
capital
563
271
503
233
203
538
(10)
252
338
16
789
147
102
136
282
50
k
k
20
1
2
48
k
Current
assets
1,576
1,259
10,819
789
378
1,848
2
875
762
43
4,298
332
343
1,594
527
73
k
k
82
13
13
58
k
Current
liabilities
1,013
988
10,316
556
175
1,310
12
623
424
27
3,509
185
241
1,458
245
23
k
k
62
12
11
10
k
(continued)
C-36
-------
TABLE C-14 (continued)
Values in parentheses represent negative numbers.
Parent firms of furnace coke producers are listed first, followed by
parent firms of foundry coke producers. Subsidiaries are listed in
parentheses below parent companies.
EBIT = earnings before interest and taxes.
Cash flow = operating income + depreciation - interest expenses - taxes.
Net working capital = current assets - current liabilities.
Tangible net worth = equity - intangible assets.
9Received income tax credit in 1983. Income tax represented as zero in
cash flow calculation.
McLouth Steel Corp. has debtor-in-possession status. The parent company
filed for bankruptcy in 1981 and filed a petition for reorganization in
December 1984. Financial information listed is for the subsidiary.
JFigures are interim values reported for first 11 months of 1984.
Converted to 1983 dollars using GNP implicit price deflator.
l(
Information not available.
A merger between National Intergroup, Inc., the parent company of National
Steel Corp., and Bergen Brunswig Corp. fell through in April 1985, 2
weeks before its scheduled date. Some market consultants feel that
National Intergroup, Inc., is now a potential target for corporate raiders,
Northwest Industries, Inc., the parent company of Lone Star Steel,
announced in April 1985 its merger with Farley Industries.
Producer of both furnace and foundry coke.
Financial information listed applies to subsidiary rather than parent
company.
pEmployees formally took control in January 1984. All figures are interim
values reported for first 3 months of 1984. Conversion to 1983 dollars
using GNP implicit price deflator.
qWheeling-Pittsburgh Steel Corp. filed for Chapter 11 in April 1985.
"\
Owned by James D. Crane. Financial information denied.
Chattanooga Coke and Chemicals Co., Inc. has debtor-in-possession status.
The company filed for arrangement under Chapter 11 in March 1984.
Latest information available is for 1982. Conversion to 1983 dollars
using GNP implicit price deflator.
m
s
C-37
-------
financial losses totalling $3.6 billion. The balance of steel trade favored
imports by a 14 to 1 import-export ratio. Imports totalled 20.5 percent of
apparent supply in 1983.58
Two integrated steel producers exhibit negative cash flows, while a
third has negative calculated working capital, as financial resources have
dwindled with the recession. Two companies, one furnace coke producer and
one foundry coke producer, are operating under bankruptcy status.
From the financial data in Table C-14, three ratios have been calculated
(Table C-15). The first, a liquidity ratio, is a measure of a firm's
ability to meet its current obligations as they are due. A liquidity ratio
above 1.0 indicates that the firm is able to pay its current debts with its
current assets; the higher the ratio, the bigger the difference between
current obligations and the firm's ability to meet them. All of the coke-
producing firms have liquidity ratios between 1.0 and 4.0, with the excep-
tions of McLouth Steel (0.17) and McWane, Inc. (5.80). These figures are
consistent with liquidity ratios for firms in a wide variety of manufactur-
ing industries.
The second ratio, a coverage ratio, gives an indication of the firm's
ability to meet its interest payments. A high ratio indicates that the
firm is more likely to be able to meet interest payments on its loans.
This ratio also can be used to determine the ability of a firm to obtain
more loans. The coverage ratio of the coke-producing firms ranged from 0.8
to 3.9. Seven firms for which information was available had negative
coverage ratios because of negative EBIT values. The positive ratios are
comparable to the coverage evidenced in most manufacturing industries. The
poor performance of those firms with negative ratios may be a result of
problems in the steel industry. However, many firms continue to make
investments funded through mergers, joint ventures, and other means.
The last of the ratios, a leverage ratio, indicates the relationship
between the capital contributed by creditors and that contributed by the
owners. Leverage magnifies returns to owners. Aggressive use of debt
increases the chance of default and bankruptcy. The chance of larger
returns must be balanced with the increased risk of such actions. The
leverage ratio indicates the vulnerability of the firm to downward business
C-38
-------
TABLE C-15. FINANCIAL RATIOS FOR COKE-PRODUCING FIRMS
Company name3 Liquidity ratio
Armco, Inc.
Bethlehem Steel Corp.
Ford Motor Co.
(Rouge Steel)
Inland Steel Co.
Interlake, Inc.
The LTV Corp.
McLouth Steel Corp.
(New Boston Coke
Corp.)
National Intergroup, Inc.
(National Steel Corp.)
Northwest Industries
(Lone Star Steel Co.)
Shenango Furnace Co. ,
Inc. (Shenango, Inc.)9
U.S. Steel Corp.
Weirton Steel Corp.
Wheeling-Pittsburgh
Steel Corp.
Jim Walter Corp.9
Koppers Co. , Inc.9
Alabama Byproducts Corp.
Citizens Gas and Coke
Utility
.Detroit Coke Corp.
Indiana Gas and
Chemical Corp.
1.56
1.27
1.05
1.42
2.16
1.41
0.17
1.40
1.80
1.59
1.22
1.79
1.42
1.09
2.15
3.17
1.32
1.08
1.18
Coverage ratio0 Leverage ratio
-3.42e 1.52
-2.306 1.95
3.82 1.73
-0.256 1.20
3.08 0.93
-1.476 2.91
f -3.75e
-2.85e 1.40
-1.666 1.65
f 0.51
1.12 2.34
f f
-1.256 3.06
0.81 3.64
1.59 0.86
f 0.35
f 1.52
f -10. 3e
f 0.50
Footnotes at end of table.
(continued)
C-39
-------
TABLE C-15 (continued)
Company name
a Liquidity ratio*3 Coverage ratio0 Leverage ratioc
McWane, Inc. 5.80 0.42
(Empire Coke Co.).
aParent firms of furnace coke producers are listed first, followed by
parent firms of foundry coke producers. Subsidiaries are listed in
parentheses below parent companies. No ratios were calculated for
Carondelet Coke Corp., Chattanooga Coke and Chemicals Co., Inc., and
Tonawanda Coke Corp. because of a lack of information.
b, . .... .. Current assets
Liquidity ratio = current liabilities
Cn . . EBIT
Coverage ratio = Annual interest expense *
d, .. Total liabilities
Leverage ratio = Tang1b1e net worth
Negative values are not meaningful.
Information not available.
^Produces both furnace and foundry coke.
C-40
-------
cycles. Also, a high ratio reveals a low future debt capacity, i.e. addi-
tions to debt in the future are less likely. The firms with coke-making
capacity had leverage ratios that ranged from 0.3 to 3.7. Six companies
had ratios below 1.0, while one firm experienced a negative ratio. These
figures highlight the poor financial condition of many firms in the coke
industry. Currently, firms with coke-making capacity are engaged in sub-
stantial amounts of debt financing, while continuing to make investments.
Another measure of financial performance is the rate of return on
equity. Studies of the iron and steel industry show low rates of return on
equity. In an analysis performed by Temple, Barker, and Sloane, Inc.
(TBS), the real (net of inflation) rate of return in the steel industry was
estimated to be 0.2 percent for the period 1970 to 1980. The TBS analysis
projected a rate of return on equity of 1.0 percent for 1980 to 1990.59
These estimates of historical and projected return on equity compare very
poorly with estimates of the required return on investment in the steel
industry. A difference between realized and required returns implies that
equity financing of capital expenditures may be difficult.
As noted, low,rates of return on equity affect common stock prices and
have implications for future investment financing, including environmental
control expenditures. The following data represent total pollution abatement
capital expenditures (PACE) as a percentage of new capital expenditures
(NCE) for SIC 3312.2 60 61 62
Year Percentage PACE of NCE
1975 20.25
1976 20.92
1977 22.95
1978 22.20
1979 25.57
1980 20.11
1981 15.75
1982 12.32
PACE as a percentage of NCE peaked at 25.57 percent in 1979, after having
been fairly steady throughout the latter part of the 1970's. The trend is
the 1980's show a PACE as a declining percentage of NCE. This decrease may
reflect the capital availability restrictions experienced by the steel
industry during this period.
C-41
-------
For the steel industry, issuing new stock to raise investment capital
is unlikely under current circumstances. If environmental and other control
investments cannot be financed through new equity, another source of funds
must be found. Increased debt is one potential source. However, firms
with coke-making capacity already have incurred substantial amounts of
debt. The TBS analysis concluded that to avoid deterioration in its finan-
cial condition, the steel industry is likely to reduce expenditures to
modernize productive facilities rather than increase its external financ-
ing.63
The steel industry has had to resort to more creative forms of financ-
ing to provide funds for modernization of facilities. This upgrading is a
key to gaining and maintaining a competitive position with respect to
imports. Cash flow for the industry has been below capital requirements
for the past two decades. Mergers, joint ventures, shared production
arrangements, abandonment of uneconomic facilities, and the sales of assets
are likely to continue being sources of capital.64 Funds advanced by
customers, with repayment geared to earnings, have been used for equipment
modernization.65
C.I.6 Industry Trends
The demand for coke is derived from the demand for steel produced by
processes that utilize coke. Hence, a description of steel industry trends
in technological development and production is a useful indicator of future
coke production and coke capacity requirements.
As mentioned, there has been a technological shift, which is expected
to continue, toward labor-saving technology. Trends in modernization are
away from open-hearth furnace production and toward electric arc furnaces
and basic oxygen furnaces. In 1960, these processes accounted for 88.2 per-
cent, 9.5 percent, and 3.3 percent of U.S. production, respectively, while
in 1982 these values were 8.2 percent, 31.1 percent, and 60.7 percent.66
In 1985, basic oxygen furnaces are expected to account for 61.5 percent of
steel production, with electric furnaces contributing 34.0 percent or more,
and open hearth furnaces declining to 6.1 percent or less.67 Electric arc
and basic oxygen furnaces represent reductions in production time, as well
as shifts to less expensive inputs.67 The increased use of these types of
furnaces will result in some decrease in demand for coke.
C-42
-------
Other changes have improved industry productivity, quality of yields,
and energy efficiency. In-ladle processes (performed after the melting
furnace stage) include inert gas stirring and vacuum treatments.68 These
techniques yield higher quality steel.
The use of continuous casters, which convert molten steel directly
into shapes ready for rolling, has increased from 18 percent of production
in the late 1970's to 35 percent in 1984.69 Yield of finished product per
ton of raw steel may be boosted to 95 percent from the current 76 percent
by use of this process. For each ton of finished steel produced using this
technology, 15 percent to 20 percent less raw steel is required, while
40 percent to 50 percent less energy is needed.67 The impact of these
technological developments on the coke industry is unknown. Any effects
will be through productivity improvements in the steel industry.
Technological trends have reduced steel use per unit of output of
durable goods. Since the 1970's, the decline in consumption of steel per
dollar of GNP has averaged 4 percent annually, with continued decline
expected.70 Increases in economic growth are predicted to offset this
effect, resulting in an increase in steel use to 95 million Mg by 1988,
with domestic shipments representing a 5-percent average annual rate
increase over the 1983 to 1988 period.65 Projections by the U.S. Bureau of
Mines predict U.S. raw steel demand will be 138 million Mg in 1990, and 164
million Mg in 2000.71
Steelmaking capacity utilization recently has been low, averaging
47.3 percent in 1982 and 55.4 percent in 1983.72 In 1984, this rate rose
to 82 percent in April before dropping to 57 percent in September.73
Capacity utilization is an important measure of industry performance
because of high fixed costs for the industry. The larger the volume of
production, the smaller the cost per unit of steel produced. For the steel
industry, the break-even point for operations is at approximately 65 per-
cent of capability, although this figure is highly dependent on prices.73
This means that steel companies have been operating at losses for several
years.
The steel industry has responded to this financial difficulty by
permanently reducing capacity. Since 1983, this reduction has been more
than 10 percent, with perhaps another 5-percent cut necessary.73 From
C-43
-------
122.5 million Mg in 1984, capacity is likely to be trimmed to 109 million
Mg by the late 1980's.73 However, the U.S.Bureau of Mines predicts U.S.
production of raw steel will rise to 113 million Mg in 1990, and to
132 million Mg in 2000, under assumptions of slow growth in the rate of
production and increases in demand.71 Changes in capacity utilization
affect coke production only to the extent that coke is an input to the
steel production process. Reductions in steel production, coupled with
shifts to noncoke energy inputs could greatly reduce demand for coke.
The emergence of minimi 11s to supply regional demand for steel has had
an impact on the operation of the larger integrated steel mills. Mini
mills now account for approximately 20 percent of U.S. steel production, at
a cost per ton of installed capacity about 75 percent less than for inte-
grated plants. The use of electric arc furnaces in minimi 11s may result in
dramatic reductions in coke demand if such mills claim 40 percent of the
steel market by 2000, as some predict.74
The combination of the factors described in this section indicate that
coke consumption is destined to continue declining. Technological improve-
ments are likely to result in an input shift away from coke, while reduced
capacity in the integrated steel industry signals a decrease in amounts of
coke needed for blast furnace steel production.
C.I.7 Market Behavior: Conclusions
Market structure, financial performance, and potential growth influence
the choice of a methodology to describe supply responses in the coke-making
industry. Although some characteristics of this industry indicate a poten-
tial for market power, other characteristics belie it.
Some concentration exists in coke-making capacity and steel produc-
tion; however, many firms produce coke and iron and steel products. Vertical
integration is substantial; however, integration appears to result primarily
from a desire for increased certainty in the supply of critical inputs.
Furthermore, substitution through alternative technologies and coke imports
is feasible, and some substitutes for the industry's final products (iron
and steel) are available. In any industry, the potential for substitution
is a major factor leading to competitive pricing. Certainly, the financial
C-44
-------
profile of coke-making firms is not indicative of monopoly profits. Pros-
pects for industry growth are limited. An individual firm must actively
compete with other firms in the industry to improve its profit position, or
even to remain viable.
No industry matches the textbook definition of perfect competition.
The important issue is whether the competitive model satisfactorily captures
major behavioral responses of firms in the industry. Based on the factors
outlined in this section, the competitive pricing model adequately describes
supply responses for coke-making firms.
C.2. ECONOMIC IMPACT OF REGULATORY ALTERNATIVES
C.2.1 Summary
Economic impacts are projected for the baseline and for each regula-
tory alternative. Furnace and foundry coke impacts are examined separately
because their production costs and markets differ. In the reanalysis, all
cost and price impacts are in second-quarter 1984 dollars.
All costs and prices used in calculations were originally in third-
quarter 1979 dollars, except prices of steel, furnace coke, and foundry
coke, which were in 1983 dollars. Conversions to the 1984 values were made
by multiplying the 1979 values by 1.362, the ratio of 1984 second-quarter
GNP implicit price deflator to the 1979 GNP implicit price deflator.14 75
The 1983 values were converted by multiplying by 1.032, the ratio of the
producer price index for second-quarter 1984 to the same index for 1983.76
When measured on a per-unit of output basis, the costs of meeting
baseline regulations for foundry coke plants tend to be greater than those
for furnace coke plants for two reasons. First, some economies of scale
are present for some of the controls. Foundry plants tend to be smaller
than furnace plants; thus, they have higher per-unit control costs. Second,
for a given battery, foundry coke output will be less than furnace coke
output because foundry coke coking time is about two-thirds longer than
furnace coke coking time.
Regulatory Alternative II has annualized costs of $4.8 million above
baseline for furnace and foundry coke producers combined. Regulatory
Alternative II requires capital expenditures of $45 million above baseline
for furnace and foundry coke producers combined. Regulatory Alternative III
C-45
-------
would result in annualized costs of $15.3 million and capital costs of
$80 million over baseline for the combined furnace and foundry coke sectors.
The values are the same whether import competition is assumed for foundry
coke producers (Scenario B) or not (Scenario A). These costs differ from
engineering estimates because of the calculation of costs based on batteries
with marginal cost of production below price, rather than all batteries.
Price impacts are estimated under the empirically supported hypothesis
that furnace coke demand is responsive to higher coke prices. Foundry coke
demand also is assumed to respond to price. Regulatory Alternative II
would have impacts of $0.13/Mg (0.12 percent change) on the price of furnace
coke, and $0.99/Mg (0.58 percent change) on the price of foundry coke under
Scenario A (1984 dollars). Under Scenario B, there are no price effects.
Regulatory Alternative III would result in furnace coke price increases of
$0.36/Mg (0.33 percent) and $1.46/Mg (0.86 percent) price increase for
foundry coke under Scenario A, and $0.00/Mg change under Scenario B.
Regulatory Alternatives II and III would have less than a 1.0-percent
impact on the production of either furnace or foundry coke under Scenario A.
Under Scenario B, Regulatory Alternative II would decrease foundry coke
production by 2.1 percent, while Regulatory Alternative III would result in
a 3.2-percent reduction in foundry coke production. There are 14 furnace
coke batteries that currently appear uneconomic. There are no uneconomic
foundry coke batteries. Regulatory Alternative II does not force any more
batteries into the uneconomic production region. Regulatory Alternative III
results in one additional furnace coke battery being pushed into the uneco-
nomic region.
C.2.2 Methodology
The following approach focuses on the long-run adjustment process of
furnace and foundry coke producers to the higher costs of coke production
that the regulatory alternatives will create. These long-run adjustments
involve investment and shutdown decisions. Short-run adjustments, such as
altering coking times, to meet the fluctuations in the demand for coke are
not the subject of this analysis.
Because of differences in production costs and markets, furnace and
foundry coke producers are modeled separately. Both are assumed to behave
C-46
-------
as if they were competitive industries selling coke in a market. This
assumption is somewhat more realistic for foundry than for furnace coke
producers because most furnace coke is produced in plants captive to the
steel industry. However, interfirm and intrafirm shipments of coke are not
uncommon, as can be inferred from Table C-ll. A plant-by-plant review of
the coke industry by Hogan and Koelble also confirmed the existence of such
exchanges.77
A set of programmed models has been developed to produce intraindustry
and interindustry estimates of the economic impacts of the alternative
regulations. The models are applied to both furnace and foundry coke, and
the sectors included are coke, steel, and ferrous foundries. The rest of
the economy is incorporated into the interindustry portion of the analysis.
The analytical approach incorporates a production cost model of the
coke industry, based on engineering data, and an econometric model of the
steel industry. The interrelationships of these models for furnace coke
are shown in Figure O3. The upper portion of Figure C-3 encompasses the
supply side impacts of the regulatory alternatives; the lower portion con-
tains the demand side impacts. In the synthesis step, the two sides are
brought together, and the equilibrium price and quantity relationships are
determined. An analogous diagram for foundry coke would substitute ferrous
foundry products for steel. The methodology is described further in the
following subsections.
C.2.2.1 Supply Side. A production cost model that incorporates
technical relationships and engineering cost estimates is used with plant-
specific information to compute separate industry supply functions, with
and without additional controls.78 Supply functions are estimated on*a
year-by-year basis for furnace and foundry coke plants projected to be in
existence between 1984 and 1995. Both coke production costs and the costs
that plants incur to meet existing environmental regulations are computed
to estimate the industry supply curve before any additional controls are
applied. Estimates of the costs of control for compliance with the regula-
tory alternatives are used to compute the projected upward shifts in that
supply function. All costs are in 1983 dollars, converted to 1984 dollars
for this reanalysis.
C-47
-------
i
-t»
oo
COS IS Of fOKC
I'ltomici ION
cnsrs OF EXISTING
ENVIRONMENTAL
HECULAIIOHS
cosis or
111 GUI AlOltV
AL1EKNATIVES
FOREIGN DEMAND
I OR US SI EEL
EX I SUNG PI ANT
INVENTORY
NEW PLANT
CONFIGURATIONS
FOREIGN f.OKE
SUPPLY
DOMESTIC
- COKE
SUPPLY
FOREIGN DEMAND
FOR US COKE
DOMESTIC DEMAND
FOII US STEEL -
DEMAND FOR
US STEEL
SYNTHESIS .IMPACTS IMPACTS
" bWVTORAL ON COKE ON FINAL
ASSUMPTIONS AND STEEL OCMAJO PRICES
DEMAND FOR
US COKE
Figure C-3. Economic impact model.
-------
This approach provides a method of estimating the industry supply
curve for coke, which shows the alternative coke quantities that will be
placed on the market at alternative prices. When the supply curve is
considered in conjunction with the demand curve, an equilibrium price and
coke output rate can be projected. Supply curve shifts caused by the
regulatory alternatives can be developed from the compliance cost estimates
made by the engineering contractor. These new supply functions, along with
the demand curve, then can be used to compute the equilibrium price and
output rate under each regulatory alternative.
C. 2.2.1.1 Data base. PI ant-by-plant data on more than 60 variables
for furnace and foundry coke plants in existence in 1979 were compiled from
government publications, industry contacts, and previous studies of the
coke industry. The data base was sent to the American Iron and Steel
Institute (AISI), which submitted it to their members for verification,
corrections, and additions,79 and to the American Coke and Coal Chemicals
Institute (ACCCI). The data were adjusted to account for the 1984 plant
inventory in the reanalysis. Capacity, number of ovens, and status (hot
idle, cold idle, under construction, or online) were updated for each
battery.5
C.2.2.1.2 Output relationships. For a given battery, the full capac-
ity output of coke, measured in megagrams per year (Mg/yr), is dependent on
the nominal coal charge (megagrams of coal per charge) per oven, the number
of ovens, and the effective gross coking time (net coking time plus decar-
bonization time). The following values for effective gross coking time
were used except where plant-specific values were available.78
Furnace Foundry
coke coke
Wet coal is h 30 h
Preheated coal 13 h 24 h
An age-specific outage rate that varies from 4 to 10 percent is assumed to
allow for normal maintenance and repair. Thus, the model assumes some
increase in such costs as a plant's age.
The quantities of by-products produced are estimated from engineering
relationships. These quantities depend on the amount of coal carbonized,
percentage of coal volatile matter, coking time, and configuration of the
C-49
-------
by-product facility at a plant. The by-products included in the model are
coke breeze, coke-oven gas, tar, crude light oil, benzene-toluene-xylene
(BTX), ammonium sulfate, anhydrous ammonia, elemental sulfur, sodium phe-
nol ate, benzene, toluene, xylene, naphthalene, and solvent naphtha. All
plants are assumed to produce breeze and coke-oven gas.
C.2.2.1.3 Operating costs. The major costs of operation for a coke
plant are expenditures for coal, labor, utilities, and chemicals. The
activities within the coke plant were allocated to 5 production and 10
environmental control cost centers (Figure C-4) to facilitate the develop-
ment of the operating cost estimates.
Coal is the major operating cost item in coke production. Plant-
specific estimates of the delivered price of coal were developed by identi-
fying the mine that supplies each plant and estimating transportation costs
from the mine to the plant. When it was not known which coal mine supplied
a particular plant, it was assumed that the coal came from the nearest
mines supplying coal of the same volatile matter and ash content as that
used by the plant. Transportation cost estimates were based on the dis-
tances traveled and the transport mode (barge or rail) employed.
Maintenance labor and supervision requirements were estimated for 69
jobs within the coke plant. Primary variables that determine the number of
maintenance labor and supervision man-hours needed include type of plant
(merchant or captive), number of battery units, number of plants at a site,
size of by-product plant, type of coal charge (wet or preheated), and coke
production. The labor rates used for captive plants were $23.21/h for
supervisory positions and $21.38/h for production labor. For merchant
plants, rates of $21.52/h and $19.61/h were assumed. These values represent
numbers used in the 1979 analysis and scaled by the GNP implicit price
deflator to 1984 dollars for the reanalysis.
The major utilities at a coke plant are steam, electricity, water, and
other fuels. Utility requirements were estimated from the data on the
plant configuration and output rates for coke and the by-products. The
prices used for the utilities are $7.41/103 Ib steam; $0.037/kWh electric-
ity; $0.22/103 gal cooling water; and $3.76/106 Btu underfire gas. These
values are the original 1979 figures scaled to 1984 dollars by the GNP
implicit price deflator for the reanalysis.
C-50
-------
o
I
en
' • IWIM im~wi< I
I OSLO*!!?* ' ' «•">•«' I
>-| iMAinii ;
I UiLUSUlJ J
&*•••••••—«ww*
Figure C-4. Coke plant cost centers.
-------
C.2.2.1.4 Capital costs. Although no net additions to industry
coke-making capacity are anticipated during the 1984 to 1995 period, a
number of producers had plans to rebuild or replace existing batteries in
1979. In 1984, three new batteries had been constructed and one was under
construction.5 Such actions alter the long-run industry supply curve
because the new batteries typically have lower operating costs per unit of
output than the batteries they replace and, most important, their capital
costs will be reflected in the new supply curve.
The capital cost breakdown for new plants is shown in Table C-16. For
such plants, the major capital cost items are the battery, quench tower,
quench car, pusher machine, larry car, door machine and coke guide, by-
product plant, coal-handling system, and coke-handling system. A 60-oven
battery is assumed. Pipeline charging can increase the coke-making capacity
of a given oven by about 25 percent by reducing gross coking time. Conse-
quently, the per-unit operating cost is reduced. The capital costs show
economies of scale, i.e., larger plants have smaller per-unit-of-capacity
capital costs. The capital cost per unit of capacity is higher for pipe-
line-charged batteries than for conventionally charged batteries.
Periodically, batteries must undergo major rehabilitation or rebuild-
ing because of performance deterioration. The costs of pad-up rebuilds
will vary from site to site depending on battery maintenance, past operat-
ing practices, and other factors. Average estimates of the cost of rebuild-
ing were developed for this study and are shown in a report by PEDCo Envi-
ronmental, Inc.81 The economic life of coke-making facilities is subject
to considerable variation depending on past maintenance and operating
practices, which also affect current operating costs. For this study, 25
years was used as the average preferred life of a new coke-making facility;
however, many batteries are operated for 35 to 40 years. If 35 to 40 years
is a more reasonable battery lifetime, use of a 25-year lifetime will
result in some overestimation of the annual costs of new or rebuilt facili-
ties. However, firms probably will not plan or expect to wait 35 to
40 years to recoup an investment in coke-making capacity.
C.2.2.1.5 Environmental costs. Plant-specific estimates of the
installed capital and operating costs for current environmental regulations
C-52
-------
TABLE C-16. ESTIMATED CAPITAL COSTS OF NEW PLANTS80
Conventionally Pipeline
charged battery charged battery
4-ma 6-ma 4-ma 6-ma
Capacity (10s Mg/yr) 450 720 560 900
Capital costs by element
(106 1979 dollars)
Coke battery 34.20 48.90 64.60 83 70
Quench tower with baffles 2.45 2.85 2 45 2*85
Quench car and pushing
emissions control 6.58 7 92 6 58 7 92
Pusher machine 2.50 3.20 2*40 3*20
Air-conditioned larry car 1.72 2.28 0 00 o'oo
Door machine and coke guide 1.80 2 10 1 80 2 10
By-product plant 32.50 39.75 35.'76 43*74
Coal-handling system 18.20 23.60 20 62 26*70
Coke-handling system 6.85 8.80 7 74 lo'oo
Offsltes 1.60 1.80 1.69 L91
Tota1 $108.40 $141.20 $143.74 $182.12
In the production cost model, new foundry batteries were assumed to be
4-m batteries and new furnace batteries were assumed to be 6-m batteries.
C-53
-------
and the regulatory alternatives under consideration in this study were
incorporated in the model. In the reanalysis, the current regulations are
assumed to include workplace standards (Occupational Safety and Health
Administration [OSHA]), water quality regulations (best practicable tech-
nology [BPT] and best available technology [BAT]), and State implementation
plan (SIP) requirements. Compliance expenses incurred for all plants in
the data base for each of the current regulations assumed baseline control
costs were estimated. Costs to comply with OSHA and BPT water requirements
under the Federal Water Pollution Control Act were assumed incurred by
1981. Costs for all other baseline environmental regulations were assumed
to be incurred by 1983.
The scatter diagrams in Figures C-5 and C-6 show estimates from the
coke supply model of average total cost of production in 1984, including
environmental costs, for all furnace and foundry coke plants. A weak,
inverse relationship between the average cost of production and the size of
the plant is evident in Figures C-5 and C-6. However, a number of other
factors create variability in the average cost of production across coke
plants. The most important of these factors are the delivered price of
coal, the age of the plant, and the by-products recovered.
C.2.2.1.6 Coke supply function—existing facilities. The operating
and capital cost functions were used to estimate the cost of production,
including relevant environmental costs, for all plants in the data base.
This cost does not include a return on investment for existing facilities.
The capital costs for these facilities already have been incurred and do
not affect operating decisions.
Capital costs that have not yet been incurred are annualized at 6.2
percent, which is estimated to be the real (net of inflation) cost of
capital for the coke industry. (This percentage is an after-tax estimate.)
This figure, which was estimated from data on the capital structure for
publicly owned steel companies, has been used in this study as the minimum
acceptable rate of return on new facilities.82
The regulatory alternatives for coke-oven by-product plants involve
control equipment that is not affixed to batteries. Accordingly, the
equipment is not affected by battery age or size (height) of the battery
C-54
-------
en
01
1 •.'».'
14O
5
^ 1.30-
c
o
3 1 2O -
O
O.
H-'
o
o
too -
90 -
o.:
a '
1 Q D
a a
a aD
a
qj -ft n °
n o a cp a a a
D
D D n
a
~~i — i 1 1 1 1 1 — § •
? 0.4 0.6 0.8 1.0 1.2 iT"^* * 1.8 ' 2.0 ^^2
Production (1,000,000 Mg/Yr)
Figure C-5. Estimated average cost of furnace coke production as a function
of plant production, 1984.
-------
o
'o>
2
•\
4ft
C
g
-M
u
3
O
i-
*&-
O
-M
W
O
O
1 00 -j
180 -
170 -
1<50 -
150 -
140 -
1.30 -
120 -
110-
•• • •
a
p
I
i
1
i
n p
D a
a
B a
a
— - -| | | i i I i I
.0
150.0 250.0 350.0
Production (1000 Mg/Yr)
450.0
Figure C-6. Estimated average cost of foundry coke production as a function
of plant production, 1984.
-------
replacement. The capital costs of the regulatory alternatives are annual-
ized over the life of the control equipment (20 years). This action is
tantamount to assuming either that all by-product plants have a remaining
life of at least 20 years or that the control equipment is salvageable.
The supply function for each plant is estimated as follows: the
average cost of production is computed for each battery in the plant; these
batteries are arranged in increasing order of their average costs of pro-
duction, and the output for each battery is accumulated to produce a stepped
marginal cost function for the plant; plant overhead costs are averaged for
all relevant plant output rates; and average total costs are computed for
each output rate by summing the average costs for plant overhead and the
battery. Each plant's supply function is the portion of the marginal cost
function above the average total cost function. For existing plants where
the average total cost exceeds marginal cost over the entire range of
output, the supply function is the point on the plant's average total cost
function represented by capacity output (after allowing for outages). .The
aggregate long-run supply function for all currently existing coke plants
and batteries is obtained by horizontally summing the supply function for
each plant. The 1984 industry marginal cost (supply) curves for existing
furnace and foundry coke plants are presented in Figures C-7 and C-8,
respectively.
C. 2.2.1.7 Coke supply function—new facilities. The cost of coke
production for new furnace and foundry batteries was estimated from the
engineering cost model, assuming the new model plants described previously.
These costs include the normal return on investment and allowances for
depreciation and corporate income taxes. When expressed on a per-unit
basis, these costs are the minimum price at which it is attractive to build
new facilities.
C.2.2.2 Demand Side. The demand for coke is derived from the demand
for products that use coke as an input to production—primarily steel and
ferrous foundry products. A demand function for furnace coke was derived
by econometrically modeling the impacts of changes in furnace coke produc-
tion costs on the steel industry.83
C-57
-------
o
en
00
o>
4A
W
O
o
zuu -
190
180 -
1 70 -
1 .50 -
150 -
MO -
1.30 -
1 20 -
110-
100 -
rtn -
0.
/""
i
/
_^ ^ -^
_^£:^~~~~~~~~
i i i ii i i i i i i i | i
0 i.O 8.0 12.0 16.O' 20.O 24.0 28.0
Production (1 .OOO.OOO MgAr)
Morglnol cost • Average
Figure C-7. Marginal and average cost functions for furnace coke, 1984.
-------
en
10
o
(J
190 -T—
180
170
16O
150
14O
13O -
120 -
no
T r—i 1 1 r
0.2 0.4 0.6 0.8 1.0 1.2 1^4 ' L^ 1^8 ^O^T^
Production (1,000.000 M^/Vr)
Marglnol cost Avorag« cost
Figure C-8. Marginal and average cost functions for foundry coke, 1984.
-------
The econometric model of the steel industry has two sectors: steel
and coke. The steel sector includes domestic steel supply, steel imports
and exports, and steel consumption (steel supply plus imports minus exports).
Similarly, the model of the coke sector consists of domestic coke supply,
domestic coke demand, and coke imports and exports. The two sectors are
linked by a derived coke demand function, which includes as variables steel
production, steel price, and quantities and prices of other inputs to steel
production. The domestic supply of steel is assumed equal to domestic
demand for U.S. steel plus world demand for U.S. steel minus U.S. import
demand.
Both linear and nonlinear specifications were used to estimate the
steel-sector model. Two-stage least squares was used to estimate the
different components of the steel sector. Visual inspections of the corre-
lation matrix and a plot of the dependent variable versus the residuals
indicated no multicollinearity or heteroscedasticity problems. The Durbin-
Watson statistic showed no evidence of autocorrelation.
The econometric estimation of the coke sector was complicated by the
small share of total domestic production that is traded in the market. The
fact that very little coke is actually sold creates concern over the
reported price of coke. Therefore, estimates of the implied price of coke
were developed, based on the value of coke in steel making, and used in the
estimation of elasticities.84 8S Estimates of elasticities for coke and
steel functions are presented in Table C-17. Actual prices for coke pro-
duced and used internally by the producing companies were used in the
reanalysis.
An attempt was made to derive a demand function for foundry coke in an
analogous manner. However, the relevant coefficient estimates were not
statistically significant at a reasonable level. A direct estimation of
the demand function, based on the prices of foundry coke, foundry coke
substitutes, and complementary inputs, also was attempted. Unfortunately,
the data necessary to properly estimate the demand function were not readily
available from published sources. Accordingly, the elasticity of demand
for foundry coke was estimated based on the theoretical relationship between
the production function for foundry products and the derived demand function
C-60
-------
TABLE C-17. ESTIMATES OF ELASTICITIES OF STEEL AND COKE MARKETS
Point Interval
estimate estimate3
1. Percent change in furnace coke demand -1.29b —c
for 1 percent change in the price of
furnace coke
2. Percent change in foundry coke demand -1.03C --c
for 1 percent change in the price of
foundry coke
3. Percent change in import demand for 1 1.88 (-1.68, 5.44)
percent change in the price of furnace
coke
4. Percent change in price of steel for O.llc —c
1 percent change in the price of
furnace coke
5. Percent change in steel demand for -1.86d (-0.54 -3 18)
percent change in the price of steel
t
6. Percent change in steel imports for 1.51d (0.51, 2.51)
1 percent change in the price of steel
Note: Estimates are based on the empirical analysis using annual data for
the years 1950-1977 with a structural econometric model of steel and
coke markets.
Interval estimates are based on 95 percent confidence level.
Derived from the production function for steel, and input cost shares.
Calculation based on the theoretical relationship between input demand
elasticity and input cost share in the production of outputs. Accord-
ingly, no interval is provided.
Significantly different from zero at 1 percent level of statistical
significance.
C-61
-------
for inputs to foundry production. This elasticity calculation is based on
a 3-year average of the cost share of foundry coke in foundry production.
This estimate is presented in Table C-17. This elasticity assumes U.S.
demand for foundry coke is supplied entirely from domestic production
(Scenario A).
Another scenario is that imported foundry coke competes with the
domestic product in the open market (Scenario B). A simplifying assumption
is that they are perfect substitutes in the production processes that
utilize foundry coke. In this case, a reduction in U.S. supply is compen-
sated by imports, so that price need not rise if the quantity of imports
purchased is increased. Both scenarios are examined in the reanalysis.
C.2.2.3 Synthesis. Separate linear functions were fit to the furnace
and foundry coke marginal cost values depicted in Figures C-7 and C-8. As
illustrated in Figures C-9 and C-10, each supply function is used with the
demand function for the appropriate type of coke to compute the initial
equilibrium price-quantity values (Px and Qj. in Figures C-9 and C-10). In
i
the case where imports are not perfect substitutes for domestically produced
coke, the supply function is reestimated for each regulatory alternative
(S2 in Figure C-9), and the new equilibrium price-quantity values (P2 and
Q2 in Figure C-9) are predicted.
The case where imports compete with domestically produced foundry coke
is shown in Figure C-10. As in Figure C-9, the supply curve shifts backward
to S2. However, because imports are available, no change in price and
quantity need be experienced by the consumer. Although domestic production
is reduced by Qi-Q2, the share of the market supplied by imports increases
by this same amount. The new equilibrium price and quantity for domestic
coke are Pj and Q2 in Figure C-10.
C.2.2.4 Economic Impact Variables. Table C-18 shows the specific
economic variables for which impacts are estimated. The methodology pre-
sented previously was designed to provide industry-level estimates of these
impacts. The conventional demand and supply partial equilibrium model of a
competitive market was chosen for this analysis because it was believed to
represent the key characteristics of the coke market and many of the impacts
of interest can be estimated readily from this model.
C-62
-------
S/Mg
P2
PI
S2
SI
10 J Mg /Tr
Figure C-9. Coke supply and demand without import competition.
C-63
-------
$/Mg
PI
S2
SI
Q2
Ql
10 Mg / Yr
Rgure C-10. Coke supply and demand with import competition.
C-64
-------
TABLE C-18. ECONOMIC IMPACT VARIABLES AND AFFECTED SECTORS
Variable
Price
Output
Profits
Costs
Plant closures/openings
Capital requirements
Factor employment
Labor
Metallurgical coal
Imports
Furnace
coke
X
X
X
X
X
X
X
X
X
Sector
Foundry
coke Steel
X X
X X
X
X
X
X
X
Xa X
Final
demand
X
Impacted under Scenario B.
C-65
-------
Figure Oil represents the markets for furnace coke and for foundry
coke which is free of import competition (Scenario A). Figure C-12 describes
the market for foundry coke that must compete with imported coke, which is
assumed to be a perfect substitute for domestic coke (Scenario B). In
Figure C-ll, D represents the derived demand for coke. The line S^ repre-
sents the baseline supply curve for coke. The equilibrium price and quantity
are Pt and Ql5 respectively. The area Ci+g+h is the total cost of coke
production, b+c1+c2+e+g+h is the total revenue, and b+c2+e represents
before-tax profits. The total cost of coke production (Ci+g+h) can be
divided into costs incurred to produce coke per se and the costs being
incurred to meet baseline environmental regulations.
The regulatory alternatives will increase the cost of coke production
by shifting the supply function to S2. This is not a parallel shift because
of the small magnitude of changes and the continued production by uneconomic
firms. Given the demand and supply functions as drawn in Figure C-ll,
higher costs of production will lead to higher prices. A production decrease
as shown in Figure C-ll would cause price to rise to P2 and the quantity
demanded to fall to Q2. The actual costs to the producer of the regulatory
alternative are c2+d-cls and profits before income taxes are a+b+Ci.
Costs to consumers are represented by a+d+f, the amount that consumers
paid to purchase the amount (Qi-Q2) at price P! before the regulation, but
now must pay price P2 to purchase.
In Figure C-12, D represents derived demand for coke and Sx represents
the baseline supply curve for coke, with P± and Qx representing equilibrium
price and quantity
As in Figure C-ll, area Cj+g+h is the total cost of coke production,
including expenses incurred to meet baseline environmental regulations.
Area b+Ci+c2+e+g+h is the total revenue, and b+c2+e is before-tax profits.
The regulatory alternatives shift the supply function to S2. As
explained, this is not a parallel shift. In this scenario, price does not
rise even though domestic production is reduced. Instead, because imported
coke is assumed to be a perfect substitute for domestic coke and because
imported coke is assumed to be available at price Pj, domestic consumers
purchase more imports and less domestic coke. The results are that domestic
C-66
-------
$/Mg
P2
PI
S2
SI
10 Mg/?r
Rgure C-11. Coke demand and supply with and without regulatory
alternatives, without import competition.
C-67
-------
$/Mg
S2
PI
Q2
Ql
10 Mg /Yr
Rgure C-12. Coke demand and supply with and without
regulatory alternatives, with import competition.
C-68
-------
production decreases to Q2, imports increase by Q^Qs, and price remains at
PI, as shown in Figure C-12.
The costs of the regulation to the producer are Ca'Cj. Total revenue
is b+C!+c2+h, and production costs are h+c2. Profits before income taxes
are b+Ci. There are no costs to consumers because they are able to purchase
quantity Qi at price Px as they were before the regulation.
C.2.3 Furnace Coke Impacts
As described in Section C.2.2 of this analysis, the furnace coke
industry has been modeled as a competitive industry supplying coke to the
steel industry. This definition implies the existence of interfirm and
intrafirm shipments of coke. However, no allowance has been made for coke
transportation costs, although coal transportation costs are included in
the cost of coke production estimates. Coke plants and their associated
steel mills are typically clustered together. As noted in Section C.I.4.1,
most coke is consumed within the region where it is produced. Hence,
transportation across great distances is uncommon. Therefore, the omission
of coke transport costs should not greatly influence the calculations.
The baseline values for 1983, presented in Table C-19, are actual data
for 1983, except for coke prices, which are calculated by the model. The
values for 1983 are assumed to reflect full compliance with applicable SIP
and OSHA air quality regulations and water quality regulations. The coke
supply model was used to compute the price of furnace coke, costs, revenues,
and profits, given these actual values. Coal consumption and employment
projections were made using current coal- and labor-output ratios. The
supply function was reestimated assuming control levels being practiced in
1984 for all emission sources. This estimation was used to determine the
impacts of moving from baseline industry control levels to alternative
regulations control for all sources.
Table C-20 presents total costs incurred by companies in SIC 3312 in
meeting environmental regulations up to 1983. These costs represent indus-
try efforts to achieve baseline compliance. Expenditures are segmented by
type of pollutant treated. Total cost for abatement increased throughout
the late 1970's and peaked at $956.5 million (1972 dollars) in 1979.
Expenditures declined slightly in 1980 and 1981 and dropped to an 8-year
low of $549.8 million (1972 dollars) in 1982.
C-69
-------
TABLE C-19. BASELINE VALUES FOR ECONOMIC
IMPACT ANALYSIS—FURNACE COKE, 198337 86 87
Baseline values
Coke market
Price (1983 $/Mg) 106.25b
Production (103 Mg) 20,462
Consumption (10s Mg) 24,380
Imports (103 Mg)c, 3,918
Employment (jobs) 6,236
Coal consumption (103 Mg) 29,787
Steel market
Price (1983 $/Mg) 319.97
Production (10s Mg) 76,763
Consumption (103 Mg) 75,710
Imports (103 Mg) 15,486
Employment (jobs) 295,000
aBaseline assumes companies meet existing regulations
including OSHA (coke-oven emissions); State regulations on
desulfurization, pushing, coal handling, coke handling,
quench tower, and battery stack controls; and BPT and BAT
water regulations.
Calculated. Market price for furnace coke was $123.51 in
the fourth quarter of 1983.
GCalculated. Imports = Consumption - production.
Calculated. Furnace coke employment = Employment in
byproduct coke industry x proportion of coke production
represented by furnace coke sector.
Represented by employment in SIC 3312 (Blast furnaces and
steel mills).
C-70
-------
TABLE 020. POLLUTION ABATEMENT EXPENDITURES FOR SIC 331262
o
—I
Abatement expenditures.
Year
1975
1976
1977
1978
1979
1980
1981
1982
Air
pollution
477.
606.
675.
709.
932.
925.
940.
666.
3
9
1
1
6
4
9
8
Total cost = Capital
K
Implicit
1972 val
clncludes
T M T ~J
price
ues.
Water
pollution0
306.8
309.5
385.8
424.9
511.6
494.9
512.6
408.3
expenditure
deflator for the
payments to
government
106 current $
Solid. Recovered cost,
waste 106 current $
43.
51.
85.
111.
99.
127.
153.
92.
1
4
5
2
5
9
0
4
18.1
17.8
18.1
15.1
1.0
18.8
22.3
22.0
+ operating costs - recovered
nonfarm
business
units for public
sector used to
sewage use.
Total cost,a
106 current $
1,
1,
1,
1,
1,
1,
809.
950.
128.
230.
542.
529.
584.
145.
cost summed
convert
1
0
3
1
7
4.
2
5
for all
current dol
Total cost,
106 1972 $
647.
722.
812.
824.
956.
856.
807.
549.
3
0
4
2
5
5
9
8
pollutants.
lars to
Includes payments to government units for solid waste collection and disposal.
-------
C.2.3.1 Price Effects. The price of furnace coke is assumed to be
established in a competitive market. In the basic model of a competitive
market, the interaction of supply and demand determine the equilibrium
price. This price is dependent on the costs of production of the marginal
producer and the value of the product to the marginal buyer. The marginal
producer is the producer who is willing to supply the commodity at the
market price because he is just covering all his costs at that price. The
marginal buyer is just willing to pay the market price. Other buyers who
value the product more still pay only the market price.
Estimates of the demand and supply functions for furnace coke are
necessary to develop projections of the equilibrium price for furnace coke
with and without increased control. The supply of furnace coke as shown
previously would be shifted by the regulatory alternatives. The demand for
furnace coke has been econometrically estimated and found to be responsive
to price changes. The estimated elasticity of demand for furnace coke is
-1.3. This responsiveness reflects the substitution of other fuels for
coke in blast furnaces; the substitution of other inputs, primarily scrap,
for pig iron in steelmaking; and the substitution of other commodities for
steel throughout the economy.
Higher prices for coke will increase the cost of steel production
unless there is a perfect substitution between coke and other inputs to
steelmaking. In that case, the consumption of coke would decrease to zero.
If substitutions for coke in steelmaking were not possible (i.e., input
proportions were fixed), the steel price increase would be the percentage
change in coke price times the share that coke represents in the cost of
steelmaking (10.7 percent) times the base price of steel.
Table C-21 presents the furnace coke and steel price impacts of the
regulatory alternatives. The proposed regulatory alternatives raise coke
prices only slightly: 0.12 percent for Alternative II, and 0.33 percent for
Alternative III.
C.2.3.2 Production and Consumption Effects. The estimated demand and
supply relationships for coke are used to project the production and con-
sumption effects of the regulatory alternatives. As shown in Table C-22,
the changes in coke production and consumption are fairly small for the two
C-72
-------
TABLE C-21. PRICE EFFECTS OF REGULATORY ALTERNATIVES-
FURNACE COKE, 1984a
Regulatory Coke, steel,
Alternative $/Mg $/Mg
11 0.13 0.04
(0.12) (0.01)
111 0.36 0.12
(0.33) (0.04)
Values in parentheses are percentage changes from baseline.
C-73
-------
TABLE C-22. PRODUCTION AND CONSUMPTION EFFECTS OF REGULATORY ALTERNATIVES-
FURNACE COKE, 1984a
Regulatory
Alternative
II
III
Coke market, 103 Mg/yr
Production Consumption
-32 -23
(-0.16) (-0.09)
-90 -65
(-0.44) (-0.26)
Imports
9
(0.23)
25
(0.64)
Steel market, 103 Mg/yr
Production Consumption Imports
-21 -18 3
(-0.03) (-0.02) (0.02)
-60 -51 9
(-0.08) (-0.07) (0.06)
a
'Values in parentheses are percentage changes from baseline.
-------
regulatory alternatives. For Alternative II, changes in production and
consumption are less than 0.2 percent. For Alternative III, the quantity
changes are less than 0.5 percent.
Imported coke is a close substitute for domestically produced coke.
Imported coke is not a perfect substitute because coke quality deteriorates
during transit and contractual arrangements between buyers and sellers are
not costless. However, increases in the costs of production for domestic
plants will increase the incentive to import coke.
The projected increases in coke imports are reported in Table C-21.
Imports increase by 0.23 percent under Alternative II and 0.64 percent
under Alternative III. As illustrated below, coke imports increased sig-
nificantly since 1972, but peaked in 1979 and began a marked decline.
Year Imports, 103 Mq
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
168
978
3,211
1,650
1,189
1,659
5,191
3,605
598
478
109
32
The increase in imports in the 1970's is believed to be the result of
a coal strike in the United States during 1978 combined with depressed
conditions in the market for steel in the countries exporting coke to the
United States. Accordingly, future importation at a high level may depend
on future market conditions for steel in other countries. In any case, the
change in coke imports projected for all the regulatory alternatives is
small.
C.2.3.3 Coal Consumption and Employment Effects. Any reductions in
coke and steel production are expected to cause reductions in the use of
the factors that produce them, The major inputs to coke production are
coal and labor. Labor is also an important input in coal mining.
C-75
-------
The coal consumption and employment implications of the projected
reductions in coal, coke, and steel production are shown in Table C-23.
For Regulatory Alternative II, coal consumption and employment impacts are
less than 0.2 percent, while for Regulatory Alternative III, impacts are
less than 0.5 percent. These values were developed assuming constant coal-
and labor-output ratios. The employment impacts shown do not include the
estimated increases in employment caused by the regulatory alternatives.
Therefore, the employment impacts represent maximum values.
C.2.3.4 Financial Effects. The aggregate capital costs of the regu-
latory alternatives are summarized in Table C-24. Capital costs also have
been summed across member plants to determine the cost to each coke-
producing company of meeting alternative regulations. The total capital
costs by company may be used to produce percentages that express the rela-
tion between total capital cost and the annual average net capital invest-
ment of the company and the annual cash flow of the company. This analysis
is presented to give some insight into the distribution of the financial
effects across coke-producing firms.
Total capital cost as a percentage of average annual net investment is
an indicator of whether the usual sources of investment capital available
to the firm will be sufficient to finance the additional capital costs
caused by the regulatory alternatives. The larger this percentage, the
greater the probability that investment needed to comply with the regula-
tory alternatives would significantly reduce investment in other areas.
This percentage provides some insights regarding the degree to which firms
will be able to finance the controls required to meet the regulatory alter-
natives without a serious impact on their financial position.
Total capital cost as a percentage of cash flow provides similar
information. Cash flow data accounts for revenues, operating costs, depre-
ciation, expenditures on dividends, interest expenses, and taxes. Thus, it
is a more realistic measure of the funds available to the firm. However,
this index may not be consistent across firms because depreciation account-
ing varies across firms. As with the net investment ratio, the larger the
ratio, the greater the probability that cash flow will be diverted from
other sources to finance compliance expenditures.
C-76
-------
TABLE C-23. COAL CONSUMPTION AND EMPLOYMENT EFFECTS OF
REGULATORY ALTERNATIVES-FURNACE COKE, 1984S
Regulatory
Alternatives
II
III
Coal
consumption
for coke,
103 Mg/yr
-47
(-0.16)
-131
(-0.44)
Employment, jobs
Coalc
mining
-13
(-0.01)
-37
(-0.02)
Coke Steel -
plant making
-10 -83
(-0.16) (-0.03)
-27 -230
(-0.43) (-0.08)
Values in parentheses are percentage changes from baseline.
Employment impacts are based on input-output relationships and production
impacts. Impacts on coke plant employment do not include jobs created bv
the relevant controls.
Annual labor productivity in coal mining is estimated as 3,515 Mg per year
C-77
-------
TABLE C-24. INDUSTRY CAPITAL REQUIREMENTS OF REGULATORY
ALTERNATIVES—FURNACE COKE, 1984
Capital costs
Regulatory of regulations,3
Alternative 106 1984 $
57 -
HI 68
Calculated for all plants in operation in 1984.44 46
C-78
-------
Financial analysis is necessarily restricted to companies for which
financial data are accessible. Therefore, financial analysis cannot be
conducted for some privately owned companies for which reporting has been
restricted. These companies are usually the smallest in a given industry,
and they probably experience higher per unit costs of regulation and higher
costs for securing financing than do larger companies.
A further complication of financial analysis is that many coke-produc-
ing companies are wholly owned subsidiaries of larger, highly diversified
corporations. Financial data are available for the parent corporations
only. Analysis of these data will probably lead to the conclusion that the
parent companies have ample resources to finance additional capital costs.
However, the extent to which these corporations will make such investments,
or will cease some coke operations in favor of other investment opportuni-
ties evidencing higher rates of return, cannot be determined without knowl-
edge of the required return on investment specific to the firm and the
other investment opportunities that exist for the firm.
Table C-25 provides the capital costs as a percentage of average
annual net investment by company for each regulatory alternative. The
average annual net investment was calculated from financial records for
each company by averaging net investment data (in constant 1983 dollars)
for 1979 to 1983. These values are converted to 1984 dollars using an
implicit GNP price deflator. In some instances, less than 5 years of data
were available. In the case of subsidiaries, net investment of parent
companies was used. The regulatory alternatives impose capital costs as
percentages of average annual net investments between 0 and 5 percent.
Table C-26 shows capital cost as a percentage of cash flow for firms
for which information was available. Cash flow data were derived from
Table C-14. The values for 1983 were converted to 1984 dollars using an
implicit GNP price deflator. As for net investment, in the case of subsid-
iaries, cash flow of parent companies was used. Capital costs as percent-
ages of cash flows range from 0 to 8 percent for the regulatory alternatives.
The leverage ratios presented in Table C-15 indicate that coke-produc-
ing firms are engaged in a substantial amount of external financing. These
firms may be reticent (or unable) to borrow more heavily, especially in the
C-79
-------
o
I
00
o
TABLE C-25 CAPITAL COSTS OF COMPLIANCE AS A PERCENTAGE OF NET INVESTMENT--
FURNACE COKE PRODUCERS, 1984d
Furnace coke producer
Regulatory II~ ~
Alternative ABC DEFGH
II
III
1
2
1
2
0
0
1
2
3
5
1
1
2
3
1
3
1
1
1
2
0
1
1
2
Average annual net income calculated from company profiles in Moody's Industrial Manual Moody s
Investor Service, New York, 1984, and Standard New York Stock Exchange Reports, Standard and Poor
Corp., New York, 1984. (Calculations were made on a constant 1983 dollar basis and converted to
1984 values using a GNP implicit price deflator).
bData on annual investment are not available for two companies.
-------
TABLE C-26. CAPITAL COST AS A PERCENTAGE OF ANNUAL
CASH FLOW—FURNACE COKE PRODUCERS, 1984a
Reaulatorv
Alternative
II
III
Furnace coke producer
A
4
8
B
5
8
C
0
0
D
2
5
E
2
4
FC G
2
4
H
1
1
I
0
1
JC K
0
1
L
0
1
Cash flow data is from Table C-14 in Appendix C. Values are converted to
1984 dollars using a GNP implicit price deflator.
There are three furnace coke producers for which cash flow data were not
available.
This company had negative cash flow.
C-81
-------
current economic climate for steel. Furthermore, financing capital expendi-
tures by issuing additional common stock would tend to dilute existing
stockholder equity. Considering the low historical return on investment in
the industry, this dilution probably would be unacceptable. An analysis of
the iron and steel industry undertaken by TBS,63 addressed the question of
external financing with regard to water pollution control expenditures.
This analysis concludes that to avoid deterioration in its financial condi-
tion, the industry is likely to reduce expenditures to modernize production
facilities rather than increase its external financing.
In summary, the capital costs of the regulatory alternatives are in
the tens of millions of dollars range. However, these amounts do not
appear unduly burdensome when compared with normal investment expenditures
or cash flow for companies for which data are available.
C.2.3.5 Battery and Plant Closures. Uneconomic batteries are those
that have marginal costs of operation greater than the price of coke.
Theoretically, these batteries are candidates for closure. There are 14
batteries operating in the uneconomic portion of the supply curve (above
the point where price!ine intersects the supply curve) under baseline
conditions. They are owned by 10 companies and are located in 11 plants.
Regulatory Alternative II does not add any companies or batteries to this
list. Regulatory Alternative III forces one more battery owned by an
additional company into the category of uneconomic batteries. It is impor-
tant to note, however, that this does not necessarily imply that the regu-
lation would cause the closure of this battery.
The decision to close a battery is more complicated than the basic
closure rule would indicate; this is particularly true for integrated iron
and steel producers. Continued access to profits from continued steel
production is a key factor in the closure decision for a captive battery.
Before closing or idling a coke battery, managers would want to know where
they would get coke on a reliable basis in order to continue making steel.
The obvious sources to be investigated include other plants within the same
company, other companies, and foreign suppliers. As noted in Section C.I,
some interregional and international movement of coke occurs.
C-82
-------
Obtaining coke from offsite sources introduces two potential complica-
tions: the cost of transporting coke and the certainty of the coke supply.
Obtaining coke from a nearby source might be the most profitable alternative
to transporting coke. If coke must be shipped over long distances, onsite
production at a cost above the projected market price might be more profit-
able.
If coke must be purchased, certainty of supply is a complication.
Steel producers prefer to have captive sources of coke to safeguard against
supply interruptions, and they may be willing to pay a premium for this
security. Producing at a cost above market price would involve such a
premium. Five of the fourteen uneconomic batteries under baseline compli-
ance produce coke at marginal costs that are less than 5 percent above the
market price. Five percent does not appear to be an excessive premium to
pay for certainty of supply.
Several other factors could affect a particular plant's decision to
close a battery. These factors relate to the relationship of coke quality
to the type of steel commodities produced, the existence of captive coal
mines, the costs of closing a battery and potential costs of restarting it
in the future, and required control and other expenditures.
An alternative to closure for a financially troubled company is to
file for Chapter 11 bankruptcy. This option allows firms to continue
operating coke plants under a restructured debt payment schedule. Of firms
owning the 14 uneconomic batteries under baseline, one has filed for
Chapter 11, and another is expected to file in the future if the steel
industry continues to sustain large losses.88 This action may improve a
firm's competitive situation. McLouth Steel Corporation, which filed for
bankruptcy in 1982, has modernized equipment and reduced overhead to enable
it to capture market share from larger companies.
The developed demand model uses a single coke price, which represents
an average quality of coke used to produce a weighted average mix of steel
products. If high production costs for a particular battery are associated
with a higher than average quality of coke, continued production might be
justified. Production also would be justified if the firm produces only
the most highly valued steel products.
C-83
-------
Some coke-producing firms also own coal mines and may wish to secure
continued access to profits from coal mining. Because profits in the coal
sector may be subject to less effective taxation because of depletion
allowances, these profits may be extremely attractive.
Furthermore, an integrated iron and steel producer must consider the
question of necessary expenditures for its entire steel plant. If the
steel facility is old or if substantial additional expenditures will be
necessary to comply with regulations on other parts of the facility, then
closure is more likely.
Closure decisions are so specific to individual situations and mana-
gers' perceptions regarding their future costs and revenues that exact pro-
jections of closures should be viewed with caution. In the current market
for steel, it is difficult to say whether uneconomic batteries will be
closed. Of companies owning uneconomic batteries, three have cut back
capacity by closing batteries, although it is unknown whether they are
those projected as candidates for closure.88
C.2.4 Foundry Coke Impacts
Oven coke other than furnace coke represents less than 15 percent of
U.S. coke production. The majority of it is used as a fuel in the cupolas
of foundries. The remainder is used for a variety of purposes, especially
for heating.
Values of various foundry coke variables in the absence of the regu-
latory alternatives are presented in Table C-27. These values assume base-
line compliance with the regulations listed in the footnote to the table.
C.2.4.1 Price and Production Effects. In developing the estimates of
price and quantity impacts, a vertical, nonparallel shift caused by each
regulatory alternative has been projected in the linear estimate of the
foundry coke supply function generated under the regulatory baseline. This
shift is used in conjunction with an estimated elasticity demand for foundry
coke of -1.03 and is designated Scenario A. Under this scenario, domestic-
ally produced foundry coke does not compete with imported coke. The
reanalysis also estimates the effects of alternative regulations assuming
that foundry coke producers must compete with imports in open market sales.
In this case, consumers of foundry coke may purchase imported coke as a
C-84
-------
TABLE C-27. BASELINE VALUES FOR ECONOMIC
IMPACT ANALYSIS—FOUNDRY COKE, 1983
Coke market Baseline values3
Price (1983 $/Mg)
Production (103 Mg)
Consumption (103 Mg)
Employment (jobs)c
Coal consumption (103 Mg)
169.
2,951
2,938
542
3,809
58b
Baseline assumes companies meet existing regulations
including OSHA (coke-oven emissions); State regulations on
desulfurization, pushing, coal handling, coke handling,
quench tower, and battery stack controls; and BPT and BAT
water regulations.
Calculated. Market price for foundry coke was $149.66 in
the fourth quarter of 1983.
£
Calculated. Foundry coke employment = Employment in
byproduct coke industry x proportion of coke production
represented by foundry coke sector.
C-85
-------
perfect substitute for domestically produced coke. The price of imports is
assumed to be constant. As regulations cause less domestic coke to be
produced, its price relative to imported coke rises. Consumers are able to
purchase as much coke as before at the same price, but a larger proportion
of sales is made up of imports. Thus, there are quantity effects for
domestic producers, but no price effects. This shift is designated
Scenario B. Impacts are assessed for both scenarios. Impacts for
Scenario B represent the maximum effect of import substitution in the
foundry coke market.
The projected price and quantity effects are shown on Table C-28.
Under Scenario A, both price and quantity impacts are less than 0.7 percent
of baseline for Alternative II and less than 0.9 percent of baseline for
Alternative III. Under Scenario B, there are no price impacts. Quantity
impacts are 2.1 percent of baseline for Alternative II and 3.2 percent of
baseline for Alternative III.
C.2.4.2 Coal Consumption and Employment Effects. Any reductions in
foundry coke production are expected to cause reductions in the use of the
factors that produce the foundry coke. The major inputs to foundry coke
production are coal and labor. Labor is also an important input in coal
mining.
The coal consumption and employment implications of the projected
reductions in coke production are shown in Table C-29 for both scenarios.
These values were developed assuming constant coal- and labor-output ratios.
The employment impacts shown do not include any employment increases caused
by the regulatory alternatives. Consequently, the employment impacts
represent maximum values.
Under both scenarios and for all regulatory alternatives, effects on
employment in the coal mining industry are negligible. Under Scenario A,
neither Alternative II nor III results in coal consumption impacts or
employment impacts in coke plants greater than a 1.2 percent change from
baseline. Under Scenario B, coal consumption is reduced by 2.0 percent for
Alternative II, and 3.2 percent for Alternative III. Employment in coke
plants is reduced by about the same percentages from baseline.
C.2.4.3 Financial Effects. The aggregate capital costs of the regu-
latory alternatives are summarized in Table C-30. The capital requirements
C-86
-------
TABLE C-28. PRICE AND QUANTITY EFFECTS OF REGULATORY ALTERNATIVES
FOUNDRY COKE, 1984a
Regulatory Coke price impact, Coke quantity impact,
Alternative 1984 $/Mg io3 Mg/yr
Scenario A
11 0.99 -18
(0.58) (-0.61)
"I 1.46 -26
(0-86) (-0.88)
Scenario B
11 0.00 -61b
(0.00) (-2.07)
111 0.00 -94b
(0.00) (-3.18)b
Values in parentheses are percentage changes from baseline.
Coke consumption is not affected due to imports. Imports under Scenario B
are equal in magnitude and of opposite sign to quantity impacts. Under
Scenario A, imports are zero.
C-87
-------
TABLE C-29. COAL CONSUMPTION AND EMPLOYMENT EFFECTS OF REGULATORY
ALTERNATIVES—FOUNDRY COKE, 1984a
Coa1.. Employment (jobs)b
consumption c—2 ^ —
Regulatory for coke, Coal Coke
Alternative 103 Mg/yr mining plant
Scenario A
II -23 -6 , -3
(-0.78) (0.00)a (-0.55)
III -34 -9 , -5
(-1.16) (O.OOr (-0.92)
Scenario B
H -78 -22 -11
(-2.05) (-0.01) (-2.03)
III -121 -34 -17
(-3.18) (-0.02) (-3.14)
aValues in parentheses are percentage changes from baseline.
bEmployment impacts are based on input-output relationships and production
impacts. Impacts on coke plant employment do not include jobs created
by the relevant controls.
°Annual labor productivity in coal mining is estimated as 3,515 Mg/yr per
job.
Zero due to rounding.
C-88
-------
TABLE C-30. INDUSTRY CAPITAL REQUIREMENTS OF REGULATORY
ALTERNATIVES—FOUNDRY COKE, 1984
Capital costs
Regulatory of regulations.
Alternative 10§ 1984 $
- — _
III 12
Calculated for all plants in operation in 1984.48 50
C-89
-------
to meet Regulatory Alternatives II and III for the foundry coke industry
are $7 million and $12 million, respectively. Capital costs also have been
summed across member plants to determine the cost to each company of meet-
ing alternative regulations. These company capital costs, along with
firm-specific financial data, are used to produce the same financial percent-
ages as described above for furnace coke and total capital cost as a percent-
age of net capital investment and of annual cash flow. Financial data are
not available for many of the foundry coke producers that are privately
held companies. Therefore, percentages for these companies are not included
in the analysis.
Capital costs as percentages of average annual net investment for the
foundry coke producers are provided in Table C-31. The costs of moving
from baseline to a regulatory alternative are never more than 11 percent of
the average annual net investment. Foundry coke production plants operate
at a significantly lower production rate for the same level of investment
as compared with furnace coke production rates. This is due to the longer
coking time for foundry coke. Furthermore, in looking at the available
data on the age of the batteries used in the production processes within
each plant, there appears to be a correlation between the age of the battery
used and the level of compliance costs facing the firm. The data suggest
that the foundry coke producing plants that are facing the highest pending
compliance costs are operating with batteries installed between 1919 and
1946. Conversely, the foundry coke producers that are facing the lowest
pending compliance costs are operating, for the most part, with batteries
installed between 1950 and 1979.
tr
Table C-32 provides capital cost as a percentage of annual cash flow.
The regulatory alternatives result in capital costs no greater than 2 per-
cent of cash flow for foundry coke producers for which information is
available.
Firms would use internal financing, additional equity financing, debt
financing, or perhaps some of the methods mentioned in Section C.I.5, to
make these capital expenditures. Because many of the foundry plants are
owned by private corporations, data are insufficient to assess the eventual
sources of capital that these firms will use. Therefore, only qualitative
statements can be made concerning the impacts of financing regulatory
C-90
-------
TABLE C-31. CAPITAL COSTS OF COMPLIANCE AS A PERCENTAGE OF NET
INVESTMENT—FOUNDRY COKE PRODUCERS, 1984a
Regulatory
Alternative
II
III
AA
4
11
Foundry coke producers
BB
0
1
CC
0
1
Average annual net investment calculated from company profiles in Moody's
Industrial Manual, Moody1s Investor Service, New York, 1984; Standard
New York Stock Exchange Reports, Standard and Poor's Corp., New York,
1984; and Dun and Bradstreet Financial Profiles, 1985. (Calculations
were made on a constant 1983 dollar basis and converted to 1984 dollars
using a GNP implicit price deflator.)
There are eight foundry coke producers for which annual investment data
are not available.
C-91
-------
TABLE C-32. CAPITAL COST AS A PERCENTAGE OF ANNUAL CASH
FLOW—FOUNDRY COKE PRODUCERS, 1984a
Foundry coke producers
Regulatory Alternative AA BB CC DD
II 1001
III 1111
aCash flow data are from Table C-14 in Appendix C. Values
are converted to 1984 dollars using a GNP implicit price
deflator.
There are six foundry coke producers for which cash flow
data are not available.
C-92
-------
investments. Any internal financing would reduce return on equity by
directly reducing dividends or by reducing productive capital expenditures.
Debt financing may reduce the return on equity by increasing the cost of
debt financing. Financing regulatory capital requirements using new common
stock issues will have a tendency to dilute present owner's equity. This
dilution could be substantial. As an alternative, one foundry coke firm
entered bankruptcy status. No information on the competitive status of
this firm is available. In the firms for which data are available, the
capital requirements of the regulatory alternative do not appear unduly
burdensome.
C.2.4.4 Battery and Plant Closures. The decision rule used to indi-
cate closure candidates among furnace batteries also is used for foundry
batteries. Any foundry battery for which the marginal cost of operation is
greater than the price of foundry coke is an uneconomic battery. According
to this criterion and assuming baseline control, no batteries that were in
operation in 1984 are uneconomic to operate. The regulatory alternatives
do not cause any batteries to move into the uneconomic region under either
scenario.
C.3. POTENTIAL SOCIOECONOMIC AND INFLATIONARY IMPACTS
C.3.1 Compliance Costs
The estimated total annualized costs to coke producers for compliance
with the regulatory alternatives are shown in Tables C-33 and C-34. Costs
for furnace and foundry producers are differentiated because of differences
in coke prices and control costs per unit of output. The costs are for all
plants in operation in 1984 afe calculated.
As shown in Table C-33, in 1984, Regulatory Alternative II would
result in compliance costs of $3.5 million per year for furnace coke pro-
ducers. Regulatory Alternative III would result in compliance costs of
$12.4 million per year for furnace coke producers.
Compliance costs for foundry coke producers is shown in Table C-34.
For Regulatory Alternative II, this cost is $1.3 million per year. For
Regulatory Alternative III, compliance cost is $2.9 million per year.
Combined compliance cost for furnace and foundry coke producers is
$4.8 million per year for Regulatory Alternative II. For Regulatory
Alternative III, this cost is $15.3 million per year.
C-93
-------
TABLE C-33. COMPLIANCE COSTS OF REGULATORY
ALTERNATIVES—FURNACE COKE PRODUCERS,
1984
Compliance
Regulatory Alternative cost, 106 1984 $/yr
II 3.5
III 12.4
C-94
-------
TABLE C-34. COMPLIANCE COSTS OF REGULATORY ALTERNATIVES-
FOUNDRY COKE PRODUCERS, 1984
Regulatory Compliance
Alternative cost, 106 1984 $/yr
n i~j
Ill 2.9
C-95
-------
C.3.3 Balance of Trade
Recent trends indicate that imports are decreasing. Imposition of the
regulatory alternatives is expected to slightly reverse this trend. Some
increase in steel imports is possible also. However, because steel price
increases caused by coke price increases are projected to be quite small,
any increase in imports caused by the regulatory alternatives should be
minor. Moreover, trade regulations covering steel imports may mitigate
such increases.
In the aggregate, it appears unlikely that these regulatory alter-
natives would significantly affect the U.S. balance of trade position,
given the small share of international trade represented by coke imports.
C.3.4 Community Impacts
Furnace and foundry coke and steel production facilities are in Penn-
sylvania, Indiana, Ohio, Maryland, New York, Michigan, Illinois, Alabama,
Utah, Kentucky, Tennessee, Texas, Missouri, Wisconsin, and West Virginia.
Closure of coke facilities, if they occur, could have impacts on commun-
t
ities in these States. The regulatory alternatives are not necessarily
projected to result in closures. Potential production decreases should not
be sufficient to generate significant community impacts. However, further
compliance with proposed regulations could result in additional battery and
plant closures and the resulting community impacts.
C.3.5 Small Business Impacts
The Regulatory Flexibility Act (RFA) requires consideration of the
potential impacts of proposed regulations on small "entities." A regula-
tory flexibility analysis is required for regulations that have a "signifi-
cant economic impact on a substantial number of small entities." For the
NESHAP for coke oven by-product plants, small entities can be defined as
small furnace and foundry coke firms. This section addresses the require-
ments that relate to the economic aspects of the RFA. Steps necessary for
determination of applicability of the RFA are:
Identification of small firms impacted by the NESHAP
Estimation of the economic impact of the NESHAP on these
small firms.
C-96
-------
The guidelines for conducting a regulatory flexibility analysis define
a small business as "any business concern which is independently owned and
operated and not dominant in its field as defined by the Small Business
Administration Regulations under Section 3 of the Small Business Act." The
Small Business Administration (SBA) defines small firms in terms of employ-
ment. Firms owning coke ovens are included in SIC 3312, which also
includes blast furnaces, steel works, and rolling mills. The SBA has
determined that any firm that is in SIC 3312 and employs fewer than 1,000
workers will be considered small under the Small Business Act.
Table C-35 shows employment data for all U.S. firms that operate
by-product coke ovens. Six firms in the list--9, 14, 16, 19, 20, and
23--can be designated as small based on SBA definitions. Because the
standard being proposed is a NESHAP and all existing and new plants will be
expected by law to comply, all plants of the small firms not currently in
compliance could be adversely impacted. A "substantial number" of small
business is defined as "more than 20 percent of these entities." This rule
implies that at least two firms be impacted to qualify as a "substantial
number."
After the affected small firms are identified, the guidelines for the
RFA require an estimate of the degree of economic impact. Four criteria
are applied in assessing whether significant economic impact occurs from
the regulation. The first criterion determines whether annual compliance
costs increase average total production costs of small entities by more
than 5 percent. None of the small firms identified was found to have an
average cost increase greater than 5 percent.
A second criterion compares compliance costs as a percentage of sales
for small entities with the same percentage for large entities. If the
result for small entities is at least 10 percentage points higher than for
large firms, the impact is considered significant. Based on net sales data
in Table C-14 and compliance cost data in Tables C-31 and C-32, one small
company is significantly impacted. It should be noted that sales data are
not available for all small entities.
The third criterion for assessing significant impact is whether capital
costs of compliance represent a "significant" portion of capital available
to small entities. The criterion recommends examining internal cash flow
C-97
-------
TABLE C-35. EMPLOYMENT DATA FOR
U.S. FIRMS OPERATING COKE OVENS, 1983
Company Employment
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
48,071
52,800
28,700
9,107
163,356
37,300
16,000
32,000
210
1,300
2,400
19,200
7,300
102
7,512
164
1,245
98,722
150
150
14,518
2,200
200
C-98
-------
in addition to external sources of financing. Small, privately owned firms
often do not report their annual investment expenditures, so that determi-
nation of capital availability is impossible. No financial data could be
located for the small coke-producing firms previously identified.
The final criterion is whether the requirements of the regulation are
likely to result in closures of small entities. None of the small firms
identified is projected to close as a result of the regulatory alternatives.
The regulatory alternatives are unlikely to result in adverse economic
impact on a "substantial number" of small entities (as defined by SBA).
Based on the four criteria used by EPA for which assessment may be made,
one firm may be "significantly impacted" under the second criterion. No
significantly impacted firms were identified under the other rules.
C.3.6 Energy
The regulatory alternatives do not have any significant direct energy
impacts. Although some indirect impacts are possible, they are likely to
be minor in nature.
Indirect impacts could include the substitution of fossil fuels for
coke in blast furnaces or an increase in use of electric furnaces, further
reducing the coke rate. Some reduction is projected to occur in any case,
but technological limits govern the degree to which the coke rate can be
reduced. Furthermore, projected coke price increases are minor when com-
pared to recent and projected fossil fuel price increases. Of course, if
imports increase, fuel will be needed to transport them. Furthermore, if
imports replace domestic coke production, excess coke oven gas, some of
which currently is used in other parts of the steel plant, may be replaced
by other fuels. But if steel production decreases, there will be some
reduction in fuel consumption.
C.4 REFERENCES
1. Executive Office of the Predident, Office of Management and Budget
Standard Industrial Classification Manual. U.S. Government Printina
Office. Washington, DC. 1972. p. 145.
2. Bureau of the Census. 1982 Census of Manufactures. Preliminary
Report, Industry Series. Blast Furnaces and Steel Mills (Industry
?J2);«a!?C82~I~33A~1(P)- Wash1n9ton, DC. U.S. Department of Commerce.
July 1984. p. 3.
C-99
-------
3. Office of Coal, Nuclear, Electric, and Alternate Fuels, Energy Infor-
mation Administration. Quarterly Coal Report, October - December
1982. Appendix A. Publication No. DOE/EIA-0121(82/4Q). Washington,
DC. U.S. Department of Energy. April 1983. p. 54-55.
4. Office of Coal and Electric Power Statistics, Energy Information
Administration. Energy Data Reports: Coke and Coal Chemicals in
1979. Publication No. DOE/EIA 0121(79). Washington, DC. U.S.
Department of Energy. October 31, 1980. p. 33-37.
5. Research Triangle Institute. Coke Battery Survey and Information from
the American Coke and Coal Chemicals Institute and the American Iron
and Steel Institute. Unpublished data. 1984.
6. Bureau of Mines. Minerals Yearbook. Washington, DC. U.S. Department
of the Interior. 1950-1978. (Table and page numbers vary because of
reclassifications.)
7. Bureau of the Census. Statistical Abstract of the United States:
1978. Table 710. 99th Edition. Washington, DC. U.S. Department of
Commerce. 1978. p. 441.
8. Bureau of the Census. Statistical Abstract of the United States:
1984. Table 735. 104th Edition. Washington, DC. U.S. Department of
Commerce. 1984. p. 449.
9. Reference 7. Table 1511. p. 874.
10. Bureau of the Census. Historical Statistics of the United States:
Colonial Times to 1970. Part 2. Washington, DC. U.S. Department of
Commerce. 1975. p. 904-905.
11. Bureau of the Census. U.S. Department of Commerce. U.S. Exports—
Domestic Merchandise, SIC-Based Products by World Areas. FT610/
Annual. Washington, DC. U.S. Department of Commerce. 1978-1984.
12. Bureau of the Census. U.S. Imports—Consumption and General, SIC-
Based Products by World Areas. FT210/Annual. Washington, DC. U.S.
Department of Commerce. 1978-1981, 1983-1984.
13. Bureau of the Census. U.S. General Imports—Schedule A, Commodities
by Country. FT135/November 1981. Washington, DC. U.S. Department of
Commerce. 1982. p. 46.
14. Bureau of Economic Analysis. Survey of Current Business. 64:
S16-S17. October 1984.
15. Reference 8. Table 1471. p. 832.
16. Office of Coal, Nuclear, Electric, and Alternate Fuels, Energy Infor-
mation Administration. Quarterly Coal Report, April-June 1984.
Appendix A. Publication No. DOE/EIA-0212(84.2Q). Washington, DC.
U.S. Department of Energy. October 1984. p. 54, 61.
C-100
-------
17. Office of Energy Data Operation, Energy Information Administration.
Energy Data Reports: Coke and Coal Chemicals in 1978. Washington,
DC. U.S. Department of Energy. 1980. p. 3-5.
18. Office of Coal, Nuclear, Electric and Alternate Fuels, Energy Informa-
tion Admistration. Coke and Coal Chemicals in 1980. Publication No
DOE/EIA 0120(80). Washington, DC. U.S. Department of Energy
November 1981. p. 4.
19. Reference 8. Table 799. p. 486.
20. Bureau of the Census, U.S. Department of Commerce. 1977 Census of
Manufactures. Washington, DC. U.S. Government Printing Office
1979. p. 2.
21. Bureau of the Census, U.S. Department of Commerce. 1972 Census of
Manufactures. Washington, DC. U.S. Government Printing Office
1976. p. 33A-6.
22. Bureau of the Census. Statistical Abstract of the United States-
1980. Table 724. 101st Edition. Washington, DC. U.S. Department of
Commerce. 1980. p. 437.
23. Bureau of Industrial Economics. 1984 U.S. Industrial Outlook. 25th
Edition. Washington, DC. U.S. Department of Commerce. January 1984
p. 18-3.
24. Standard and Poor's, Inc. Steel and Heavy Machinery—Basic Analysis
Industrial Surveys. 152(43):Sec 1, p. 515. New York. October 1984.
25. Office of Coal, Nuclear, Electric, and Alternate Fuels, Energy Infor-
mation Administration. Quarterly Coal Report, October - December
1983. Appendix A. Publication No. DOE/EIA-0121(83/4Q). Washington,
DC. U.S. Department of Energy. April 1984. p. 55.
26. Reference 4. p. 3-5.
27. Reference 18. p. 3-5.
28. Industrial Economics Research Institute. Analysis of the U.S. Metal-
lurgical Coke Industry. Fordham University. October 1979, p. 41.
29. Reference 28. p. 40.
30. Office of Coal, Nuclear, Electric, and Alternate Fuels, Energy Infor-
mation Administration. Quarterly Coal Report, January - March 1984
Appendix A. Publication No. DOE/EIA-0121(84/1Q). Washington DC
U.S. Department of Energy. July 1984. p. 62.
31. Telecon. Peterson, J., U.S. Steel Corporation, with Lohr, L.,
Research Triangle Institute, December 10, 1984. Operational status of
cold idle coke batteries.
C-101
-------
32. Kerrigan, T. J. Influences upon the Future International Demand and
Supply for Coke. Ph.D. dissertation. New York. Fordham University.
1977. p. 14, 114.
33. Reference 4. p. 1.
34. Reference 4. p. 13.
35. Reference 25. p. 54.
36. Emissions Standards and Engineering Division. Standard Support and
Environmental Impact Statement: Standards of Performance for Coke
Oven Batteries. Research Triangle Park, NC. Environmental Protection
Agency. May 1976. p. 3-7.
37. Bureau of Mines. Mineral Commodity Summaries. Washington, DC. U.S.
Department of the Interior. 1984. p. 78.
38. Merrill Lynch, Pierce, Fenner, and Smith, Inc. The Outlook for Metal-
lurgical Coal and Coke. Institutional Report. 1980. p. 3.
39. Reference 28. p. i.
40. Reference 28. p. ii-iii.
41. Reference 38. p. 1.
42. Reference 38. p. 5.
43. PEDCo Environmental, Inc. Technical Approach for a Coke Production
Cost Model. 1979. p. 39-50.
44. Reference 18. p. 33-35.
45. Office of Energy Data and Interpretation, Energy Information Adminis-
tration. Energy Data Reports: Distribution of Oven and Beehive Coke
and Breeze. Washington, DC. U.S. Department of Energy. April 10,
1979. p. 7-8.
46. The Politics of Coke. 33 Metal Producing. March 1980. p. 49-51.
47. DeCarlo, J. A., and M. M. Otero. Coke Plants in the United States on
December 31, 1959. Washington, DC. Bureau of Mines, U.S. Department
of the Interior. 1960. p. 4-10.
48. Bureau of Economics, Federal Trade Commission. Staff Report on the
United States Steel Industry and Its International Rivals: Trends and
Factors Determining International Competitiveness. Washington, DC.
U.S. Government Printing Office. November 1977. p. 53.
C-102
-------
49. Schottman, F. J. Iron and Steel. Mineral Commodity Profiles, 1983.
Washington, DC. Bureau of Mines, U.S. Department of the Interior
1983. p. 4-5.
50. Reference 18. p. 3-4, 8.
51. Reference 3. p. 53-55, 58.
52. Reference 25. p. 53-55, 58.
53. Reference 4. p. 3-4, 8.
54. Moody1s Investors' Service. Moody1s Industrial Manual. Vol. 1 and 2
New York. 1984.
55. Moody1s Investors' SErvice. Moody1s OTC Manual. New York. 1984.
56. Standard and Poor's Corp. Standard and Poor's New York Stock Exchanqe
Stock Reports. New York. 1984.
57. Dun and Bradstreet, Inc. File reports printed for use of Research
Triangle Institute. January 31, 1985.
58. Reference 24. p. S16.
59. Temple, Barker, and Sloane, Inc. An Economic Analysis of Proposed
Effluent Limitations Guidelines, New Source Performance Standards and
Pretreatment Standards for the Iron and Steel Manufacturing Point
Source Category. Exhibit 6. Lexington, MA. December 1980.
60. Bureau of the Census. Annual Survey of Manufactures. Washington DC
U.S. Department of Commerce. 1975 - 1982.
61. Bureau of the Census. 1977 Census of Manufactures. Vol II
Industry Statistics. Part 2, SIC Major Groups 27-34. Washington DC
U.S. Department of Commerce. August 1981. p. 33A-15.
62. Bureau of the Census. Pollution Abatement Costs and Expenditures
Current Industrial Reports. MA-200(year). Washington, DC. U S
Department of Commerce. 1975 - 1983.
63. Reference 59. p. VI-4.
64. Reference 24. p. S28-S29.
65. Reference 23. p. 18-5.
66. Reference 8. Table 1408. p. 788.
C-103
-------
67. Reference 24. p. S22-S25.
68. Reference 49. p. 3.
69. Reference 24. p. S21.
70. Reference 23. p. 18-4.
71. Reference 49. p. 14.
72. Reference 24. p. S3.
73. Reference 24. p. S19-S20.
74. Reference 25. p. SI.
75. Office of the President. Economic Report of the President, February
1984. Washington, DC. U.S. Government Printing Office. 1984.
p. 224.
76. Reference 14. p. S-6.
77. Reference 28. p. 63-111.
78. Reference 43. p. 1-69.
79. Letter from Young, E. F., American Iron and Steel Institute, to
Pratopos, J., U.S. Environmental Protection Agency. May 16, 1980.
80. Reference 43. p. 5.
81. Reference 43. p. 85.
82. Research Triangle Institute. Economic Impact of NSPS Regulations on
Coke Oven Battery Stacks. Research Triangle Park, NC. May 1980.
p. 8-46, 8-47.
83. Research Triangle Institute. An Econometric Model of the U.S. Steel
Industry. Research Triangle Park, NC. March 1981.
84. Ramachandran, V. The Economics of Farm Tractorization in India.
Ph.D. dissertation. Raleigh, NC. North Carolina State University.
1979.
85. Heckman, J. J. Shadow Price, Market Wages, and Labor Supply. Econo-
metrica. 42:679-694. July 1974.
86. Reference 24. S16, S27.
C-104
-------
87. Reference 25. p. 55-57, 61.
88. Symonds, W. C., and G. L. Miles. It's Every Man for Himself in the
Steel Business. Business Week. (2897):76,78. June 3, 1985.
C-105
-------
Appendix D
Health Risk Impact Analysis
-------
APPENDIX D: HEALTH RISK IMPACT ANALYSIS
D.I REVISED INPUT DATA
The health risk impact analysis was repeated after proposal of the
standards because of revisions to the baseline data regarding plant caoacities
and emission factors for foundry plants (as described in Chapter 6 of thi
document). In addition, EPA made more accurate estimates of the latitudes and
longitudes of plant locations by examining U.S. Geological Survey
topographical maps. These revised input data are shown on Tables D-l and
nf wal!!eJiPA 3lSKKdevel°Pfid stack Parameters for modeling the control option
of wash-oil scrubbers that would achieve 90-percent emission reduction; they
onHnnfl0" ^\*~2', The Sta(* Paramete^ for baseline and other control
options have not changed since the analysis performed before proposal.
The^unit risk estimate has been revised since proposal to U.026/ppm
This revision is described in Chapter 9 of this document. The revised
estimate was used in the updated calculation of incidence and maximun lifetime
D.2 METHODOLOGY
The Human Exposure Model was used to generate revised risk estimatps as
was used for the preproposal analysis. This methodology was described in"
Appendix E of the proposal BID. j^Lrioea in
After the computer modeling was completed, the estimates for the three
furnace and three foundry plants with the highest values for maximum lifetime
risk were examined further. A detailed check was made to determine whether
the location of the most exposed individual was realistically placed bv the
computer. A review of the U.S. Geological Survey maps revealed that at four
of these six plants, there appeared to be no possible residential sites where
^computer placed the most exposed individuals. Therefore, the points of
baS6l1ne and controlled «s£s if these
D-3
-------
TABLE 0-1. FURNACE COKE BV-PRODUCT RECOVERY PLANTS LOCATION AND BENZENE EMISSIONS (kg/yr) FOR REGULATORY BASELINE
K. ........ * Coke production
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Nn
Plant
LTV Steel, Thomas, AL
New Boston, Portsmouth, OH
Wheeling-Pitt, Monessen, PA
Lone Star Steel, Lone Star, TX
LTV Steel, So. Chicago, IL
National Steel, Granite City, IL
Interlake, Chicago, IL
LTV Steel, Gadsden, AL
Rouge Steel, Oearborn, MI
U.S. Steel, Fairless Hills, PA
LTV Steel, Warren, OH
LTV Steel, E. Chicago, IN
Annco Inc., Ashland, KY
Weirton Steel, Brown's Is., WV
U.S. Steel, Provo, UT
LTV Steel, Aliquippa, PA
Bethlehem Steel, Lackawanna, NY
National Steel, Detroit, MI
U.S. Steel, Lorain, OH
Wheeling-Pitt, E. Steubenville, WV
LTV Steel, Cleveland, OH
Arraco Inc., Middletown, OH
Bethlehem Steel, Burns Harbor, IN
LTV Steel, Pittsburgh, PA
U.S. Steel, Fairfield, AL
Bethlehem Steel, Bethlehem, PA
Bethlehem Steel, Sparrows Pt., MD
Inland Steel, E. Chicago, IN
U.S. Steel , Gary, IN
U.S. Steel, Clairton, PA
Totals
tp- n*t* rnrrpnt as of November 198
Latitude
33°32'47"
38°44'57"
40°09'46"
32°54'59"
41°41'29"
38°41'40"
41°39'22"
34°00'46"
42°18'19"
40°09'28"
41039'48"
38°30'07"
40°24'58"
40°18'43"
40°37'16"
42°49'20"
42°15'16"
41°2b'56"
40°20'36"
41°28'26"
39°29'45"
41°37'41"
40°25'34"
33°29'22"
40"36'51"
39°13'10"
4r37'53"
41°36'55"
40°18'04"
4.
Longitude
86°50'13"
82°56'01"
79°53'47"
94°42'57"
87"32'50"
90°07'42"
87°37'32"
86°02'38"
83°09'40"
74°44'32"
80°48'40"
87°26'42"
82°40'08"
80"35'16"
80°14'24"
83°07'43"
82°07 '50"
80°36 '25"
81°39'55"
84°23'15"
87°10'20"
79°57'47"
86°55'32"
75°21'13"
76°29'19"
87°27'15"
87°20'03"
79°52'21"
1,000 Mg/yr
315
364
490
507
563
570
582
758
778
916
945
948
963
1,097
1,160
1,218
1,292
1,397
1.496
1,509
1,760
1,776
1,790
1,792
1,822
2,253
3,506
3,715
4,228
6,294
45,804
decanter
2.43E+04
2.80E+04
3.77E+04
3.90E+04
4.34E+04
4.39E+04
4.48E+04
5.84E+04
5.99E+04
7.05E+04
7.28E+04
7.3UE+04
7.42E+04
8.45E+04
8.93E+04
9.38E+04
9.95E+04
1.08E+05
1.1SE+05
1.16E+05
1.36E+OS
1.37E+05
1.38E+05
1.38E+05
1.40E+05
1.73E+05
2.70E+05
2.86E+05
3.26E+05
4.08E+05
3.53E+06
Tar
storage
3.78E+03
5.88E+03
6.U8E+03
6.76E+03
6.84E+03
6.98E+03
9.10E+03
9.34E+03
1.10E+04
1.13E+04
1.14E+04
1.16E+04
1.32E+04
1.39E+04
1.46E+04
1.68E+04
1.80E+04
1.81E+04
2.11E+04
2*15E+04
2.19E+04
4i21E+04
4.46E+04
5.07E+04
6.35E+04
5.50E+05
Excess
ainiion i a -
liq. tank
2.84E+03
3.28E+03
4.56E+03
5.07E+03
5.13E+03
5.24E+03
6.82E+03
7.00E+03
8.24E+03
8.51E+03
8.53E+03
8.67E+03
9.87EK13
1.04E+04
U16E+04
1.26E+04
K58E+04
1.60E+04
1.61E+04
1.61E+04
1.64E+04
2.03E+04
3.16E+04
3.34E+04
3.81E+04
4.76E+04
4.12E+05
Light-
oil
storage
1.83E+03
2.11E+03
2.84E+03
2.94E+03
3.27E+03
3.31E+03
3.38E+03
4.40E+03
5!31E«)3
5.48E+03
5.50E+03
5.59E+03
6.36E+03
7]49E+03
8.10E+03
K02E+04
1.03E+04
O.OOE+00
1.04E+04
1 .06E+04
2!o3E+04
2.15E+04
2.42E+05
Light-
oil
sump
4.73E+03
5.46E+03
7.35E+03
7.61E+03
8.45E+03
8.55E+03
8.73E+03
1.14E+04
1.17E+04
1.37E+04
1.42E+04
1.42E+04
1.44E+04
l!?4E+04
1.83E+04
2J10E+04
2.24E+04
2.26E+04
2.64E+04
2.66E+04
O.OOE+00
2.69E+04
2.73E+04
3.38E+04
5.26E+04
5.57E+04
?!94E+04
6.60E+05
Light-oil
cond. vent
2.80E+04
3.24E+04
4.36E+04
4.51E+04
5!o?E+04
5.18E+04
6.75E+04
6.92E+04
s!44E+04
8.57E+04
9.76E+04
1.03E+05
1.08E+05
1.15E+05
1.24E+05
1.33E+05
1.34E+05
1.57E+05
1.58E+05
O.OOE+00
1.59E+05
1.62E+05
2.01E+05
3.12E+05
3.31E+05
3.76E+05
9.42E+03
3.46E+06
Wash-oil
decanter
1.20E+03
1.38E+03
1.86E+03
1.93E+03
2.14E+03
2.17E+03
2.21E+03
2.88E+03
2.96E+03
3.48E+03
3.59E+03
3.60E+03
3.66E+03
4.17E+03
4.41E+03
4.63E+03
4.91E+03
5.31E+03
5.68E+03
5.73E+03
6.69E+03
6.75E+03
O.OOE+00
6.81E+03
6.92E+03
8.56E+03
2.66E+02
1.41E+04
1.61E+04
2.01E+04
1.54E+05
Wash-oil
circ.
tank
1.20E+03
1.38E+03
1.86E+03
1.93E+03
2.14E+03
2.17E+03
2.21E+03
2.88E+03
2.96E+03
3.48E+03
3.59E+03
3.60E+03
3.66E+03
4.17E+03
4.41E+03
4.63E+03
4.91E+03
5.31E+03
5.68E+Q3
5.73E+03
6.69E+03
6.75E+03
O.OOE+OQ
6.81E+03
6.92E+03
8.56E+03
2.66E+02
1.41E+04
1.61E+04
2.01E+04
1.54E+05
(continued)
-------
TABLE 11-1. (continued)
O
I
cn
Plant
No.
1
2
3
4
b
6
7
8
y
10
11
12
13
14
Ib
16
17
18 .
19
20
21
22
23
24
25
26
27
28
29
30
Total
Leaks
2.46E+04
2.46E+04
2.46E+04
2.46E+04
5.74E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
5.74E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
2.46E+04
O.OOE+00
2.46E+04
2.46E+04
2.46E+04
b.74E+04
2.46E+04
2.4bE+04
b.74E+04
8.4bE+Ob
Tar
dewatering
6.62E+03
7.64E+03
1.03EH)4
1.06E+04
O.OOE+00
1.20E+04
1.22E+04
1.59E+04
1.63E+04
1.92E+04
1.98E+04
1.99EHH
2.02EtU4
2. JOE +04
2.44E+04
2.56E+04
2.71E+04
2.93E+04
3.14E+U4
3.17E+04
3.70E+04
3.73E+04
3.76E+04
3./6E+04
3.83E+04
4.73E+04
7.36E+04
7.80E+04
8.88E+04
O.OOE+UO
8.39E+05
Flushing-
liquor
circ. tank
2.84E+03
3.28E+03
4.41E+03
4.56E+03
5.07Et03
5.13E+03
5.24E+U3
6.82E+03
7.00Et03
8.24E+03
8.51E«-03
«.53EH>3
8.67E+03
9.37E+03
1.04E+04
1.10E+04
1.16E+04
1.26E+04
1.3bEt04
1.36E+04
1.58E+04
1.60E+04
1.61E+04
1.61E+04
1.64E+04
2.03E+04
3.16E+U4
3.34E+04
3.81Et04
4.76E+04
4.12E<-05
Benzene
storage
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
3.27E+03
O.OOE+00
O.OOE+UO
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
6.73E+03
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
2.03E+04
O.OOE+00
O.OOE+00
3.07E+04
6.10E+04
Denver
flo unit
2.74E+04
O.OOE+00
O.OOE+00
4.41E+04
O.OOE+00
4.95E+04
5.06E+04
6.59E+04
6.76E+04
7.96E+04
O.OOE+00
8.24E+04
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
1.12E+05
1.21E+05
1.30E+05
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
1.56E+05
1.58E+05
1.96E+05
O.OOE+00
O.OOE+00
3.67E+05
O.OOE+00
1.71E+06
Naphth.
melt pit
6.30E+03
O.OOE+00
O.OOE+00
1.01E+04
O.OOE+00
1.14E+04
1.16E+04
1.52E+04
1.56E+04
1.83E+04
O.OOE+00
1.90E+04
O.OOE+00
O.OOE+00
O.OOE+00
O.OUE+00
2.58E+04
2.79E+04
2.99E+04
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
3.58E+04
3.64E+04
4.51E+04
O.OOE+00
O.OOE+00
8.46E+04
O.OOE+00
3.93E+Ob
Naphth.
dry tank
3.15E+01
O.OOE+00
O.OOE+00
5.07E+01
O.OOE+00
5.70E+01
5.82E+01
7.58E+01
7.78E+01
9.16E+01
O.OOE+00
9.48E+01
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
1.29E+02
1.40E+02
1.50E+02
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
1.79E+02
1.82E+02
2.25E+02
O.OOE+00
O.OOE+00
4.23E+02
O.OOE+00
1.97E+03
TBFC
O.OOE+00
2.55E+04
3.43E+04
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
8.12E+04
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
2.60E+05
O.OOE+00
O.OOE+00
4.01E+05
UWFC
8.51E+04
O.OOE+00
O.OOE+00
1.37E+05
O.OOE+00
1.54E+05
1.57E+05
2.05E+05
2.10E+05
2.47E+05
O.OOE+00
2.56EK)5
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
3.49E+05
3.77E+05
4.04E+05
O.OOE+00
O.OOE+00
4.80E+05
O.OOE+00
4.84E+05
4.92E+05
6.08E+05
O.OOE+00
O.OOE+00
1.14E+06
O.OOE+00
5.79E+06
BTX
storage
O.OOE+00
O.OOE+00
O.OOE+00
2.94E+03
3.27E+03
3.31E+03
O.OOE+00
O.OOE+00
4.51E+03
O.OOE+00
O.OOE+00
5.50E+03
O.OOE+00
6.36E+03
O.OOE+00
O.OOE+00
7.49E+03
O.OOE+00
O.OOE+00
O.OOE+00
1.02E+04
1.03E+04
O.OOE+00
O.OOE+00
O.OOE+00
2.51E+02
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
5.41E+04
Tar inc.
suinp
2.99E+04
3.46E+04
4.66E+04
4.82E+04
b.35E+04
5.42E+04
5.53E+04
7.20E+04
7.39E+04
8.70E+04
8.98E+04
9.01E+04
9.15E+04
1.04E+05
1.10E+05
1.16E+05
1.23E+05
1.33E+05
1.42E+05
1.43E+05
1.67E+05
1.69E+05
1.70E+05
1.70E+05
1.73E+05
2.14E+05
3.33E+05
3.53E+05
4.02E+05
5.03E+05
4.35E+06
-
Total Benzene
By Plant
2.51E+05
1.74E+05
2.26E+05
3.91E+05
2.44E+05
4.37E+05
4.42E+05
5.68E+05
5.87E+05
6.82E+05
3.46E+05
7.10E+05
3.52E+05
4.04EK)5
5.40E+05
4.39E+05
9.59E+05
1.03006
1.10E+06
5.38E+Ob
6.34E+05
1.12E+06
3.99E+05
1.31E+06
1.33E+06
1.63EI-06
1.26E+06
1.55E+06
3.06E+06
1.32E+06
2.40E+07
Data current as of November 1984.
-------
TABLE l)-2. FOUNDRY CUKE BY-PRODUCT RECOVERY PLANTS LOCATION AND BENZENE EMISSIONS (kg/yr) FOR REGULATORY BASELINE
No.
1.
I:
&
• 7.
8.
9.
10.
11.
12.
13.
14.
Nnf
Plant
Chattanooga Coke, Chattanooga, TN
IN Gas Terre Haute IN
Koppers, Toledo, OH
Empire Coke, Holt, AL
Carondolet, St. Louis, MO
AL Byproducts, Keystone, PA
Citizens lias, Indianapolis, IN
Jim Walters, Birmingham, AL
Shenango, Pittsburgh, PA
Koppers, Woodward, AL
AL Byproducts, Tarrant, AL
Detroit Coke, Detroit, MI
Totals
p- lists rnrrpnt as of November 19£
Latitude
3b002'16"
39°26'48"
41°40'10"
33° 14 '25"
42°08'43"
42°58'56N
38° 32 '08"
40°05'12"
39°45'16"
33°33't7b
40°28'491I~
33°26'13"
33°34'57"
420H'19"
14.
C
Longitude
85018'U"
87"23'47"
83029'31"
87°3U'n"
80001'32"
78°56'19"
90° 16 '05"
7b°18'59"
86°06'49"
86°48'38"
80°03'34"
86°57'50"
86°46'47"
83°09'16"
)oke production
capaci ty ,
1,000 Mg/yr
130
132
157
161
207
299
330
402
477
499
521
563
583
617
5,078
Tar
decanter
4.70E+03
4.78E+03
5.68E+03
5.83E+03
7.49E+03
1.08E+04
1.19E+04
1.45E+04
1.73E+04
1.81E+04
2.04E+04
2.11E+04
1.84E+05
Tar
storage
7.33E+02
7.44E+02
8.85E+02
9.08E+02
1.17E+03
1.69E+03
1.86E+03
2.27E+03
2.69E+03
2.81E+03
2.94E+03
3.18E+03
3.29E+03
3.48E+03
2.86E+04
Excess
ammonia
liq. tank
8.54E+02
8.67E+02
1.03E+03
1.06E+03
1.36E+03
1.96E+03
2.17E+03
2.64E+03
3.13E+03
3.28E+03
3.42E+03
3.70E+03
3.83E+03
3.34E+04
Light-
oil
storage
4.07E+02
4.13E+02
O.OOE+00
b.04E+02
O.OOE+00
9.36E+02
O.OOE+00
1.26E+03
O.OOE+00
1.56E+03
1.63E+03
1.76E+03
1.83E+03
O.OOE+00
1.03E+04
Light-
oil
sump
1.05E+03
1.07E+03
O.OOE+00
1.30E+03
O.OOE+00
2.42E+03
O.OOE+00
3.26E+03
O.OOE+OQ
4.04E+03
4.22E+03
4.56E+03
4.72E+03
O.OOE+00
2.66E+04
Light-oil
cond. vent
6.25E+03
6.34E+03
O.OOE+00
7.74E+03
O.OOE+00
1.44E+04
O.OOE+00
1.93E+04
O.OOE+QO
2.40E+04
2.50E+Q4
2.71E+04
2.80E+04
O.OOE+00
1.58E+05
Wash-oil
decanter
2.67E+02
2.71E+02
O.OOE+00
3.30E+02
O.OOE+00
6.14E+02
O.OOE+00
8.25E+02
O.OOE+00
1.02E+03
1.07E+03
1.16E+03
1.20E+03
O.OOE+00
6.75E+03
Wash-oi 1
circ.
tank
2.67E+02
2.71E+02
O.OOE+00
3.30E+02
O.OOE+00
6.14E+02
O.OOE+00
8.25E+02
O.OOE+00
1.02E+03
1.07E+03
1.16E+03
1.20E+03
O.OOE+00
6.75E+03
(continued)
-------
TABLE D-2. (continued)
— ^-=-^"'
Plant
no.
1
2
3
4
5
6
7
a
9
1U
11
12
13
14
Total
Leaks
2.24E+U4
2.24E+04
O.UOE+00
2.24E+04
U.OOE+UO
2.24E+04
O.UOE+UO
b.22E+04
O.OOE+00
2.24E+04
2.24E+04
2.24E+04
2.24E+04
O.OUE+UO
2.31E+U5
Tar
dewatering
1.28E+03
1.30E+U3
1.55E+03
O.OOE+UU
2.U4E+03
2.9bE+03
3.26E+03
3.97E+03
4.71E+U3
4.93E+U3
5.14E^03
5.56E+03
5.75E+U3
6.09E+03
4.85E+04
Flushing-
Hquor
circ. tank
8.54E+02
8.67Et02
1.03E+03
1.06E+03
1.36E+U3
1.96E<-03
2.17Et03
2.64E>U3
3.13E+03
3.28E+U3
3.42E+03
3.70E+03
3.83E+03
4.05E+03
3.34E+04
Benzene
storage
U.OOE+00
O.OOE+00
O.OOE+00
O.UOE+00
O.OOE+00
O.OOE+00
O.OOE+UO
1.26E+03
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
1.26E+03
Denver
flo unit
8.25E+03
8.37E+03
O.OOE+00
1.02E+04
O.OOE+00
O.OOE+00
O.OOE+00
2.55E+04
3.03E+04
3.17E+04
O.OOE+00
O.OOE+00
3.70E+04
O.OOE+00
l.^lE+Ob
Naphth.
melt pit
1.90E+03
1.93E+03
O.OOE+00
2.3bE+03
O.OOE+00
O.OOE+00
O.OOE+00
b.87E+03
6.96E+03
7.29E+03
O.OOE+00
O.OOE+00
8.51E+03
O.OOE+00
3.48E+04
Naphth.
dry tank
9.49E+00
9.64E+00
O.OOE+00
1.18E+01
O.OOE+00
O.UOE+OU
O.OOE+00
•2.93E+01
3.48E+01
3.64E+01
O.OOE+00
O.UOE+00
4.26E+01
O.OOE+00
1.74E+02
TBFC
O.OOE+OU
O.OOE+00
8.02E+03
O.OOE+00
O.OOE+00
O.OOE+00
O.UOE+OU
O.OOE+00
O.OUE+OU
O.OOE+00
O.OOE+UO
2.88E+04
O.OOE+00
O.OOE+00
3.6HE+04
UWFC
2.b6E+04
2.60E+U4
O.UUE+00
3.17E+04
O.OOE+00
O.UOE+UO
O.UOE+UO
7.92E+04
9.4UE+04
9.84E+U4
U.UUE+UO
0;UOE+00
l.lbE+Ub
U.UOE+00
4.7UE+Ob
BTX
storage
O.UUE+OU
O.UOE+UO
O.OOE+00
O.UUE+OU
U.OOE+OU
U.OOE+UO
U.UUE+UU
1.26E+U3
U.UOE+OU
l.bbE+u3
U.UUE+UU
O.OUE+UO
O.OUE+OU
O.OOE+UU
2.82E+03
Tar inc.
b.8UE+03
b.89E+U3
7.01E+03
7.19E+U3
9.24E+U3
1.34E+U4
1.47E+04
1.79E+U4
2.13E+U4
2.23E+U4
2.J3E+04
2.blE+U4
2.6UE+U4
2.7bE+U4
2.27E+Ub
Total Benzene
by plant
U.U6E+U4
S.lbE+04
2.b2E+04
9.29E+04
2.2VE+U4
7.41E+U4
J.tilE+U4
2.3bE+Ob
1.84E+Ub
2.4WE+Ub
l.l<^E+Ub
1.48E+OS
2.H4E+03
b.7bE+U4
l.b9E+Ub
Note: Data current as of November 1984.
-------
TABLE D-3. PARAMETERS FOR 90% EMISSION REDUCTION OPTION (WASH-OIL SCRUBBER)
Scrubber Vertical cross- Diameter Stack gas Stack gas
height, sectional of vent, velocity, temperature,
Source ma area, m2a m m/sb °KC
Storage tanks
for Light oil,
8TX, or benzene 4.3
Excess ammonia-
1iquor tank 10.1
Tar storage 12.3
Tar dewatering
tanks 12.3
Tar decanter,
tar intercepting
sump, and
flushing-liquor
circulation
tankd 4.6
Light-oil condenser
vent, wash-oil
decanter, wash-
oil circulation
tank6 4.3
23.6
101.9
240.0
240.0
29.2
0.191
0.191
0.191
0.191
0.191
0.46
0.46
1.149
1.149
1.149
305
305
311
311
311
23.6
0.191
1.149
305
aAssumed wash-oil scrubber attached to the side of the source and is the same
height and same vertical cross-sectional area
bStack gas velocity derived from costing design flow of 0.013 m-^/s for storage
tanks containing light oil, BTX, or benzene and the excess ammonia liquor
tank, and from costing design flow of 0.03 m3/s for other sources, which have
greater emissions.
cStack gas temperature is the temperature of cool wash oil for sources with
cool gases and is slightly higher (38 °C) for sources with hot gases.
^Assumes this group of sources all vented to one scrubber attached to the
flushing-liquor circulation tank.
eAssumes this group of sources all vented to one scrubber attached to the
wash-oil circulation tank.
D-8
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO. ' 2.
EPA-450/3-83-0165
4. TITLE AND SUBTITLE
Benzene Emissions from Coke By-Product Recovery
Plants - Background Information for Revised Proposed
Standards
7. AUTHOR(S)
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
12. SPONSORING AGENCY NAME AND ADDRESS
DAA for Air Quality Planning and Standards
Office of Air and Radiation
U.S., Environmental Protection Agency
Research Triangle Park, North Carolina 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
June 1988
6. PERFORMING C ^GANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
National emission standards to control emissions of benzene from new and existing
coke by-product recovery plants are being promulgated under Section 112 of the
Clean Air Act. This document contains summaries of public comments, EPA responses,
and a discussion of differences between the proposed and revised proposed standard.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Air pollution Steel industry
Pollution control
National emission standards
Industrial processes
Coke by-product recovery
Hazardous air pollutants
Benzene
18. DISTRIBUTION STATEMENT
Unlimited
b.lDENTIFIERS/OPEN ENDED TERMS
Air Pollution Control
Benzene
Stationary Sources
19. SECURITY CLASS (This Report/
Unclassified
20. SECURITY CLASS (This page)
Unclassified
c. COSATI Field/Group
13B
21. NO. OF PAGES
255
22. PRICE
' EPA Form 2220-1 (R«v. 4-77) PREVIOUS EDITION js OBSOLETE
------- |