-? /---;! :;-;r^ w;v^-' ^yS^v ^ «f v%"£::^
:. >\ /-'.t ,'-> i -',;'fe < , ^ v>, /-<'/A /v,,.e/' - .-,; '(^:, ^. ^>."• ^ ,>'< • ^--,^
-^^'•.^'
'^-^•^ , f ;.'i^ , ' \ ',>--*, >V^ ' ^ 'r<^CI
^•^^.-ff^iaijlilllp^'
:: -jfe ilfcS 'j^a^tj^taiwb*'; ~J & i ^ir-i-x.'.,.,..
:-.-ckj -.^^^/ ]*' s v^t^*-la.'i,A~' * - -i /=^^.. -* y^, ^_- --^> 7" va^ ^t^ ' , V\ -^ ( -. \ ^
|j0 s~\ ' rS * .-. ,^' -^ ,' f •^. ^-" S.__ - ^
^ r/ V o , N^- %> / ; .-I, ^ ^-
i^ ->:-
r\\: -7
-'•\V' ft ^f\l v-'\ >/ -,>,<) ^ -.^
. •; Hv^^^^ii^^-^^-"^^'^' -?^- v>-^-r^ ->'-\v bv^\J/^
-A''' '&. s*3[si*'-\>iCj * ^r'^\>, 'A^v'"^T-#'>-Ai'.*"--r ^ -.. ^ ~^/^/"<^ ^
^ V' - - • - >x i -• ^~-T '-/;\J. - L ^ i " "r
1 ' ! ~- '^ ~^" --1 ^' rt,^ sS^-v'Ti/^' '-' • . v/>."77 is.' • ^J*'>./-'
\ Ak "v \
0 p
^>M^S^fe^ ^.c^^s
-------
Draft:
Regulatory Impact Analysis
for
National Emissions Standards for
Hazardous Air Pollutants
for
By-Product Coke Oven
Charging, Door Leaks, and Topside Leaks
Emission Standards Division
U.S. Environmental Protection Agency
Office of Air and Radiation
Office of Air Quality Planning and Standards
MD-13, Research Triangle Park, N.C. 27711
November 1992
li.S. Envi.-onrnentii! Protection Agency
Ri-gion 5, Library (PL-12J)
77 West Jackson Boulevard, 12th Floor
Cnicago, IL 60604-3b90
-------
(Disclaimer)
This report has been reviewed by the Emissions 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, N.C. 27711; or from
National Technical Information Services, 5285 Port Royal Road, Springfield, Virginia
22161.
u
-------
EXECUTIVE SUMMARY
The Environmental Protection Agency (EPA) has developed a National Emissions
Standard for Hazardous Air Pollutants (NESHAP) regulations for emissions from by-product
coke oven batteries, both new and existing, under the authority of section 112(b)(l)(A) of the
Clean Air Act Amendments of 1990. This decision is based on evidence from EPA and state
agency studies that coke oven batteries release air pollutants that have adverse effects on both
public health and welfare.
Air pollutants from coke oven batteries are emitted from charging operations, door
leaks, topside leaks, and process upsets. The emissions contain benzene soluble organics
(BSO) and other carcinogens such as beryllium, benzene, and arsenic. As a result, coke
oven emissions were classified as a Class I carcinogen and listed as a hazardous air pollutant
under the Clean Air Act in September 1984* with new levels of control being specified in
the Clean Air Act, as amended in November 1990.
Because coke is used as an input both in traditional steel-making and in the production
of iron castings in the cupolas of foundries, the market for coke actually comprises two
distinctly separate markets: the market for furnace coke (used to produce steel), and the
market for foundry coke (used to produce iron castings). Where possible, the impacts
associated with each type of coke product is analyzed.
During the development of the regulation, the EPA entered into regulatory
negotiations with industry, labor unions, states, and environmental groups. The resulting
requirements for the control of emissions from coke oven batteries differs somewhat from the
emission limits mandated by the CAAA. However, due to time constraints and the ever-
changing aspects of a regulatory negotiation, the following analyses are based on the
emission limits set by the CAAA.
The standard requires compliance with maximum achievable control technology
(MACT) by 1993 and lowest achievable emission rate (LAER) by 1998. The EPA's best
estimate of the total annual cost of compliance with MACT is $25 million, while LAER is
anticipated to cost $46 million annually. It should be noted that the emission reductions and
costs associated with LAER include those incurred by achieving MACT in 1993. The annual
cost of controlling emissions from process upsets with the use of flares is $2.8 million
annually.
Research Triangle Institute. Economic Analysis of Air Pollution Regulation: By-Product
Coke Ovens. Final report prepared for the U.S. Environmental Protection Agency,
OAQPS. Research Triangle Park, N.C. Publication No. RTI/4853-33 DR. August
1991. pp. 1-1.
111
-------
Economic impacts associated with the standard are minimal with small market
changes in price and quantity. The price of furnace coke is expected to increase under
MACT by 0.22 percent (with a quantity decrease of 0.66 percent) and by C.68 percent under
LAER (with a quantity decrease of 2.13 percent). In the foundry coke sector, price may
increase by 0.80 percent under MACT (with a quantity decrease of 1.08 percent). The price
increase under LAER could reach 2.53 percent (with a quantity decrease of 2.6 percent). As
with the cost estimates, the impacts associated with LAER include the impacts incurred under
MACT. Only two batteries are expected to close under LAER for furnace coke, and at
most, one battery would close in the foundry coke market. This assumes that the decrease in
quantity produced is completely absorbed by these batteries, rather than all batteries sharing
the burden of decreased production. Additionally, the small business impact is expected to
be minimal. While some of the small firms will experience adverse impacts, two of these
firms are expected to experience increased profits as a result of the regulation.
The benefits of reduced coke oven emissions are quantified for three benefit
categories: morbidity, mortality, and household soiling. Time and resource constraints
preclude quantification of all potential benefits of the regulation. These benefits are
discussed qualitatively in Chapter 7. The quantified benefits associated with MACT range
from $2.8 million to $18.0 million annually,'while the benefits of LAER range from $3.3
million to $20.7 million annually. The monetary value of emission reductions for flares
range from $4.2 million to 26.5 million annually.
Measuring net benefits in a benefit-cost analysis is one way of determining the
efficiency of a regulation. Because the benefits presented in Chapter 7 are quantified for
only three benefit categories, the EPA cannot conclude that the benefits outweigh the costs of
the regulation. Another method of coming to an efficient resolution to an externality is
through negotiation. As mentioned previously, the EPA entered into regulatory negotiations
with the affected parties to come to an efficient solution. The outcome of the negotiation is
presented in Chapter 4.
-------
CONTENTS
Chapter Page
Executive Summary iii
List of Tables x
List of Figures xii
Acronyms, Definitions, and Conversions xiii
1 Background 1-1
1.1 The Coking Process 1-1
1.2 Legal History 1-3
1.3 Statutory Provisions 1-3
1.4 Executive Order 12291 1-4
1.5 Guide to References 1-5
2 Need for Regulation 2-1
2.1 The Pollution Problem 2-1
2.2 Need for Regulation 2-2
2.2.1 Market Failure 2-2
2.2.2 Harmful Effects of Coke Oven Air Emissions 2-4
2.3 Consequences 2-5
2.3.1 If The Regulation Is Met 2-5
2.3.2 If The Regulation Is Not Met 2-6
3 Control Techniques 3-1
3.1 Technology for the Control of Emissions from Charging 3-1
3.1.1 Stage Charging 3-1
3.2 Technology for the Control of Door Leaks 3-5
3.2.1 Oven Door Seal Technology 3-6
3.2.2 Modern Hard Seals 3-7
3.2.3 Other Hard Seals 3-8
-------
3.2.4 Saturn Doors 3-9
3.3 Other Control Techniques 3-10
3.3.1 Operating and Maintenance Procedures 3-10
3.3.2 Effects of Process Variables 3-11
3.3.3 Startup, Shutdown, Upsets, and Breakdowns 3-11
3.3.4 Door Controls for Tall Ovens 3-12
3.3.5 Thompson Non-Recovery Coke Ovens 3-12
3.4 Technology for the Control of Topside Leaks 3-13
3.3.1 Description of Technology 3-13
3.3.2 Performance of Topside Leaks Control 3-14
3.5 Flares 3-14
3.5.1 Applicability 3-15
3.5.2 Efficiency : 3-16
3.5.3 Types of Flares .....3-17
4 Regulatory Options 4-1
4.1 Introduction 4-1
4.2 Limits in the Clean Air Act 4-1
4.2.1 Requirements for MACT, LAER, and Residual Risk 4-2
4.2.2 Timing of Coke Oven Provisions in the CAAA 4-2
4.3 Regulatory Negotiation 4-3
4.3.1 Source Categories 4-3
4.3.2 Emission Limitations Set by MACT and LAER 4-4
4.3.3 Process Upsets 4-6
4.3.4 Work Practices 4-6
4.4 Economic Incentives: A Market-Based Approach 4-6
5 Cost Analysis 5-1
5.1 Introduction 5-1
5.1.1 Approach 5-1
5.1.2 Issues and Assumptions 5-2
5.2 MACT Costs 5-5
*
vi
-------
5.2.1 Lids and Offtakes 5-5
5.2.2 Charging 5-6
5.2.3 Doors 5-6
5.2.4 Monitoring 5-7
5.3 LAER Costs 5-7
5.4 Process Upsets 5-11
6 Economic Impacts 6-1
6.1 Introduction 6-1
6.2 Domestic Demand for Furnace Coke 6-1
6.2.1 Quantity of Raw Steel Produced 6-1
6.2.2 Trends in Steel-Making Technology 6-4
6.2.3 Domestic Consumption of Furnace Coke 6-6
6.2.4 Demand for Foundry Coke 6-6
6.3 Coke Production 6-8
6.3.1 Coke Production Process 6-9
6.3.2 Profile of Coke Producers 6-9
6.3.3 Historical Coke Production Trends 6-13
6.3.4 Inputs in Coke Production 6-14
6.4 Baseline Conditions in the Markets for Coke 6-15
6.4.1 Markets for Furnace and Foundry Coke 6-16
6.4.2 The Baseline: Projected Conditions 6-17
6.4.3 Baseline Costs of Coke Production 6-23
6.5 Costs of the Regulation 6-24
6.5.1 Components of Compliance Costs , 6-24
6.5.2 Calculating Compliance Costs 6-27
6.5.3 Availability of Resources 6-29
6.6 Economic Impacts 6-30
6.6.1 Qualitative Analysis of Expected Impacts 6-30
6.6.2 Economic Impacts Estimation Model 6-31
6.6.3 Implementing the Impacts Estimation Procedure 6-36
i
vii
-------
6.6.4 Results of the Estimation Model 6-38
6.6.5 Evaluation of Small Business Impacts 6-46
6.6.6 Indirect Impacts 6-48
6.7 Conclusions ....6-51
7 Benefits Analysis 7-1
7.1 Introduction 7-1
7.2 Pollutants 7-1
7.3 Environmental Benefits 7-2
7.4 Methodology 7-2
7.4.1 Identification of Potential Benefit Categories 7-2
7.4.2 Identification of Concentration-Response Functions Appropriate
for Benefits Estimation 7-5
7.4.3 Development of Benefit Estimates 7-5
7.4.4 Aggregation To Total Incremental Benefits 7-7
7.5 Data 7-7
7.5.1 Air Quality Data 7-7
7.5.2 Demographic Information 7-8
7.6 Exposure Assessment 7-8
7.7 Benefits Estimation 7-9
7.7.1 Mortality Due to BSO Exposure 7-9
7.7.2 Acute Morbidity Due to Particulate Matter Exposure 7-10
7.7.3 Household Soiling 7-11
7.8 Findings 7-12
7.9 Process Upsets 7-16
7.10 Non Quantified Benefits ; 7-18
7.10.1 Cancer Mortality of Workers 7-18
7.10.2 Mortality and Morbidity Due to Exposure to Non-BSO
or Non-PM Related Compounds 7-18
7.10.3 Adverse Environmental Impacts 7-18
vui
-------
8 Benefit/Cost Analysis 8-1
8.1 Introduction 8-1
8.2 Comparison of Quantified Benefits and Costs for MACT & LAER 8-1
8.3 Comparison of Quantified Benefits and Costs for Flare Control 8-3
8.4 Rationale for the Proposed Regulatory Action 8-4
8.5 Conclusions 8-4
Appendix A - Cost Components for Each Battery A-l
Appendix B - Cost Estimates for Each Battery B-l
IX
-------
TABLES
4.1 Proposed Limits for Existing By-Product Batteries 4-5
5.1 Capital Cost Elements for Category E 5-9
5.2 Capital Cost Elements for Category F - Rebuilds 5-10
6.1 Quantity of Steel Produced Domestically, Imports, Exports of Steel 6-4
6.2 Furnace Coke Production in Captive Batteries, Exports, Imports,
and Apparent Consumption 6-7
6.3 Coke Producers in the United States 6-11
6.4 History of U.S. Coke Production 6-14
6.5 Data Used in Forecasting Furnace Coke Consumption 6-19
6.6 Regression Analysis of Furnace Coke Consumption in
Captive Facilities 6-20
6.7 Projected Production of Coke-Using Steel and Furnace Coke
Consumption: 1990 through 2010 6-20
6.8 Data Used in Forecasting Merchant Coke Production 6-21
6.9 Regression Analysis of Merchant Coke Production 6-22
6.10 Projected Production of Castings and Foundry Coke:
1990 through 2020 6-24
6.11 Parameter Estimates for the Furnace and Foundry Coke
Impacts Model, 1993 Baseline 6-34
6.12 Baseline Prices and Quantities in Affected Markets, 1993 6-39
6.13 Percentage Changes in Prices and Quantities in Affected
Markets Under MACT Minimum Control Level, 1993 6-39
6.14 Changes in Prices and Quantities in Affected Markets
Under MACT Minimum Control Level 6-40
6.15 New Prices and Quantities in Affected Markets Under MACT
Minimum Control Level, 1993 6-41
6.16 Baseline Process and Quantities in Affected Markets, 1998 6-42
6.17 Percentage Changes in Prices and Quantities in Affected Markets
Under LAER Control Level 6-43
6.18 Changes in Prices and Quantities in Affected Markets Under
LAER Control Level, 1998 6-44
-------
6.19 New Prices and Quantities in Affected Markets Under
LAER Control Level, 1998 6-45
7.1 Composition of Coke Oven Gas 7-3
7.2 Potential Physical Effects Categories 7-4
7.3 Effects of Baseline Level of Emissions 7-13
7.4 Estimated Monetary Value of Damages due to Baseline Levels of Emissions 7-13
7.5 Estimated Monetary Value of Compliance with MACT Requirements 7-14
7.6 Estimated Monetary Value of Compliance with LAER Requirements 7-15
7.7 Estimated Monetary Value of Emission Reductions for Flares 7-17
8.1 Total Annual Costs, Benefits, and Net Benefits of
MACT and LAER Requirements 8-2
8.2 Total Annual Costs, Benefits, and Net Benefits for Flare Requirements 8-3
B.I Summary of MACT Cost , B-2
B.2 Summary of LAER Costs B-4
XI
-------
FIGURES
1.1 A Typical Coke Oven Battery 1-2
6.1 Total Steel Production and Consumption in the U.S., 1970-1990 6-2
6.2 Share of Steel Production by Technology 6-4
6.3 Effects of Decreased Steel Production and Changing Steel-Making and
Iron-Making Technologies on the Market for Furnace Coke 6-5
6.4 Example of Truncated Normal Distribution 6-27
6.5 Effect of Control Options on Market for Coke 6-32
6.6 Markets and Commodities in the Furnace Sector of the Coke Ovens NESHAP
Economic Impacts Model 6-33
7.1 Benefits Estimation Process 7-6
Xll
-------
ACRONYMS, DEFINITIONS, AND CONVERSIONS
ACRONYMS
ACCCI
AISA
BG/EDs
BID
BSO
CAA
CAAA
CFR
CPI-U
EPA
FR
HAP
HEM
HIS
H2S
IARC
ISCLT
LAER
MACT
NESHAP
OAQPS
OSHA
American Coke and Coal Chemicals Institute
American Iron and Steel Institute
Block Group/Enumeration Districts
Background Information Document
Benzene Soluble Organics
Clean Air Act
Clean Air Act Amendments of 1990
Code of Federal Regulations
Consumer Price Index - All Urban Consumers
Environmental Protection Agency
Federal Register
Hazardous Air Pollutant
Human Exposure Model
Health Interview Survey
Hydrogen Sulfide
International Agency for Research on Cancer
Industrial Source Complex Model - Long Term
Lowest Achievable Emission Rate
Maximum Achievable Control Technology
National Emission Standards for Hazardous Air Pollutants
Office of Air Quality Planning and Standards
Occupational Safety and Health Administration
Xlll
-------
PLD
PLL
PLO
RAD
RCRA
RFA
RIA
RRAD
SHEAR
SIC
SIP
SO2
TSP
VOC
WLD
Percent Leaking Door
Percent Leaking Lids
Percent Leaking Offtakes
Restricted Activity Days
Resource Conservation and Recovery Act
Regulatory Flexibility Act; also, Regulatory Flexibility Analysis
Regulatory Impact Analysis
Respiratory-Related Restricted Activity Days
Systems Applications Human Exposure and Risk Model
Standard Industrial Classification
State Implementation Plan
Sulfur Dioxide
Total Suspended Particulate
Volatile Organic Compound
Work Loss Days
§112
1991 $
Annual Costs
Title HI
ECONOMIC. REGULATORY. AND SCIENTIFIC TERMS
Section of Title III in the CAAA that requires EPA to
promulgate regulations establishing emission standards for new
and existing sources of HAP emissions for By Product for Coke
Ovens
Constant (real) dollars at their fourth quarter 1991 value
Annualized capital plus annual operating costs
Micogram (10"6 gram)
The third title of the CAAA that lists the 189 HAP's to be
controlled with MACT, as well as the the control of major and
area sources.
xiv
-------
UNITS AND CONVERSIONS
This report uses metric units, some of which may not be familiar to all readers. The
EPA is required by Congress to use metric measurements. The following is a short guide to
the units and their conversions.
Conversions
To Approximate
As
Multiply by
Mg (megagram)
scm (standard
cubic meter)
MJ (megajoule)
MW (megawatt)
kg (kilogram)
Ton (2,000 Ib)
scf(standard
cubic foot)
Btu (British
thermal unit)
Btu/second
Ib (pound)
1.1
35.3
949
949
2.2
XV
-------
CHAPTER 1
BACKGROUND
1.1 THE COKING PROCESS
Iron and steel are refined metals used for making several
various products. In a series of processes, refined iron ore is
manipulated to produce iron metals, which are then used in the
production of steel. Coke is a chief fuel used in blast furnaces
for the conversion of iron ore into iron. Coke is also used by a
number of other industries, principally iron foundries,
nonferrous smelters, and chemical plants.
Coke is a metallurgical coal that has been baked into a
charcoal-like substance that burns more evenly and has more
structural strength than coal. It is produced in a coke oven by
driving off the volatile compounds in the coal, leaving a strong
residue that contains a high percentage of carbon and relatively
few impurities. The particular mix of high- and low-volatile
coals used, and the length of time the coal is heated determine
the type of coke produced. Furnace coke, used as a fuel in blast
furnaces, is produced by baking a coal mix of 10 to 30 percent
low-volatile coal for 15 to 18 hours. Foundry coke, used as a
fuel in the cupolas of foundries, is produced by baking a mix of
50 percent or more low-volatile coal for 25 to 30 hours1.
The coking procedure is performed in ovens that are
constructed in groups with common side walls, called batteries.
A typical coke oven battery is shown in Figure l-l. During the
coking process, coal is fed into the coke oven battery (charged)
through ports at the top of the oven, which are then covered with
lids. The coal is then heated in the absence of air in specially
designed refractory chambers. Volatile material is driven off in
the form of raw coke oven gas and then piped through an offtake
system (for distillation and separation), where valuable by-
products such as phenols, naphthalene, benzene, toluene, and
ammonium sulfate are recovered as part of the production process.
1-1
-------
The cleaned gas is used to underfire the coke ovens and for fuel
elsewhere in the plant2.
Volatile materials
to by-produca ptim
Figure 1.1*
A Typical Coke Oven Battery
1 Air Pollution Training Institute. Air Pollution Control
Systems for Selected Industries. U.S. Environmental Protection
Agency. Research Triangle Park, N.C. Course No. SI:431,
Publication No. EPA 450/2-82-006. June 1983. pp. 10-4.
1-2
-------
Coke oven emissions contain benzene soluble organics (BSO)
and other carcinogens such as beryllium, benzene, and arsenic.
In a coke battery, emissions occur from a number of locations or
operations. These include the following:
charging operations
topside leaks'*
door leaks
pushing operations
quenching operations, and
coke battery combustion stacks.
As a result of the pollutants contained in these emission
sources, coke oven emissions were classified as a Class I
carcinogen and listed as a hazardous air pollutant under the
Clean Air Act in September 19843. The proposed NESHAP
concentrates on the control of emissions from charging
operations, doors, and topsides (lids and offtakes). The
remaining emission points are on the ten year bin source category
list and will be considered for regulation at a later date.
1.2 LEGAL HISTORY
EPA first addressed coke ovens in the late 1970s. A
standard was proposed in 1987, but it was held in abeyance due to
the anticipated requirements of the CAAA. The new regulations
are required by Title II of the Act, and Title III provides for a
reduction in adverse effects of hazardous air pollutants from new
and existing sources. Under §112 of the CAAA, EPA is required to
set allowable emission limits for coke oven doors, lids, removals
(offtakes), and seconds of charging4.
1.3 STATUTORY PROVISIONS
The Clean Air Act requires standards of maximum achievable
control technology (MACT) for existing sources, lowest achievable
emissions rate (LAER) for existing sources, MACT for new sources,
and work practices. When considering limits for MACT for
existing sources, the CAAA specify that these standards are to
require at a minimum that coke oven emissions not exceed 8
percent leaking doors (PLD), l percent leaking lids (PLL), 5
percent leaking offtakes (PLO), and 16 seconds of visible
emissions per charge. In establishing the standards, the use of
luting compounds to prevent door leaks and the use of nonrecovery
technologies as the basis for standards for new sources have been
evaluated. Existing coke oven batteries are to comply with the
b Topside leaks refers to leaks from lids and offtakes.
1-3
-------
proposed standard by December 31, 1995, and new batteries will
comply with MACT for new sources upon start-up3.
Section 112(d)(8) also requires promulgation of work
practice regulations for new and existing coke oven batteries.
Existing batteries must comply with the work practice regulations
by November 15, 1993. The CAAA specify that the work practice
regulations require, as appropriate, the use of luting compounds,
if the EPA determines they are an effective means of controlling
leaks, as well as door and jam cleaning practices.
Section 112(f) also requires EPA to promulgate residual risk
standards in the year 2000. Coke oven batteries would be
required to comply with these limits by December 31, 2003.
Section 112 (f) permits an owner or operator of a coke oven
battery to defer meeting the residual risk limit until the year
2020 provided that the following requirements are met6:
• By November 15, 1993, batteries must not exceed 8
PLD, 1 PLL, 5 PLO, and 16 seconds of visible
emissions per charge.
• By January 1, 1998, the batteries must meet the
LAER standard that is defined for a coke oven
battery that is rebuilt or replacements at a coke
oven plant for an existing battery, or any
subsequent revision of LAER. The Act requires
that these limits may be no less stringent that 3
PLD for doors less that 6 meters tall, and 5 PLD
for doors 6 meters or taller; 1 PLL; 4 PLO; and,
16 seconds of visible emissions per charge. An
exclusion may be considered for emissions from
doors during a period after the closing of self-
sealing oven doors or the total mass emissions
equivalent.
• By January 1, 2000, the owner or operator must
make available to the surrounding community the
results of any risk assessment performed by the
EPA to determine the appropriate level of residual
risk standard.
1.4 EXECUTIVE ORDER 12291
On February 17, 1981, President Reagan issued Executive
Order 12291, which requires the EPA to prepare Regulatory Impact
Analyses (RIA's) for all "major rules"7. A "major rule" consists
of any regulation that is likely to result in an annual effect on
the economy of $100 million or more, a major increase in costs or
prices, or significant adverse effects on employment,
competition, investments, productivity, innovation, or on the
1-4
-------
ability of the United States-based enterprises to compete with
foreign-based enterprises in domestic or export markets8-9. The
EPA considers the regulation for Coke Oven Emissions to be major
and thus is issuing this RIA.
Along with requiring an analysis of benefits and costs,
Executive Order 12291 specifies that EPA, to the extent allowed
by the Clean Air Act (CAA) and court orders, demonstrate: (1)
that the benefits of the Coke Oven regulation will outweigh the
costs, and (2) that the maximum level of net benefits will be
reached. This document reviews the need for the regulation
(chapter 2), control techniques (chapter 3), regulatory options
(chapter 4), costs of control (chapter 5), economic impacts
(chapter 6), benefits of the regulation (chapter 7) , and a.
comparison of the benefits and costs associated with the
regulation (chapter 8).
All analyses presented in this RIA are based on the emission
limits set by the CAAA. During the development of the NESHAP for
Coke Ovens, the EPA entered into regulatory negotiations with
industry. Because the level of control recommended by the
regulatory negotiation committee was just recently reached, time
and resource constraints limit the ability of re-evaluating the
analyses of costs, benefits, and economic impacts based on the
suggested levels of control.
In the analysis of costs associated with the emission limits
set by the CAAA, all estimates were accumulated on a plant-by-
plant basis and reviewed extensively by the Regulatory
Negotiation Cost Work Group. To provide a summary of these costs
and to avoid exposing information deemed confidential by
industry, the overall costs to a plant of complying with MACT and
LAER are presented as total capital and annual costs.
Monetized estimates of benefits for three of the health and
welfare components are presented in Chapter 7. These benefit
categories include: mortality, morbidity, and household soiling.
Data, time, and resource limitations preclude a quantitative
analysis for all potential benefit categories. These benefits
are qualitatively discussed in Chapter 7.
1.5 GUIDE TO REFERENCES
The composition of this RIA is mostly a summary of research
reports, analyses, correspondence, minutes of various meetings
and hearings, policy directives, legal notices, laws,
regulations, and other documents related to Coke Oven operations.
The principal references are listed in the back of the chapter on
the subject of interest to you. Consult these references, as
well as the preambles that accompany proposal of the regulation
for Coke Ovens in the Federal Register, for more details.
1-5
-------
for Coke Ovens in the Federal Register, for more details.
References are held in public dockets and are available for
inspection and copying. For more information on the docket,
contact:
Air Docket
Maildrop: LE-131
Waterside Mall, Rm. M-1500
401 M. Street, SW
Washington, D.C. 20460
Hours: 8:00 a.m. to 3:30 p.m.
Phone: (202) 382-7549
Refer to Docket #A-79-l5 for all material relating to this
regulation.
1-6
-------
REFERENCES
1. Research Triangle Institute. Economic Analysis of Air
Pollution Regulations: By-Product Coke Ovens. Final Report
prepared for U.S. Environmental Protection Agency, OAQPS.
Research Triangle Park, N.C. Publication No. RTI/5153-9 FR.
June 1992. pp. 1-1.
2. Reference 1, pp. 1-1.
3. Reference 1, pp. 1-1.
4. U.S. Federal Register. Volume 57, Number 10, Notices.
January 15, 1992. pp. 1730.
5. Reference 4, pp. 1730-1731.
6. Reference 4, pp. 1732.
7. U.S. Office of the President. Federal Regulation, Executive
Order 12291. February 17, 1981.
8. U.S. Environmental Protection Agency; Office of Policy,
Planning, and Evaluation; Office of Policy Analysis.
Guidelines for Performing Regulatory Impact Analyses.
Publication No. EPA 1230-01-84-003. December 1983 with
updates through March 1988. pp. 3.
9. U.S. Office of Management and Budget. Regulatory Impact
Analysis Guidance. Appendix V of Regulatory Program of the
United States Government. April 1, 1988 through March 31,
1989.
1-7
-------
CHAPTER 2
NEED FOR THE REGULATION
2.1 THE POLLUTION PROBLEM
Section 112 of the Clean Air Act defines a hazardous air
pollutant as one which "in the judgement of the Administrator
causes, or contributes to, air pollution which may reasonably be
anticipated to result in an increase in mortality or an increase
in serious irreversible, or incapacitating reversible illness"1.
The toxic constituents of coke oven emissions include both gases
and respirable particulate matter with a varying chemical
composition including BSO, and other carcinogens such as
beryllium, benzene, and arsenic2. There is, therefore, concern
over the potential health risks caused by long-term exposure to
particulate matter and gases contained in coke oven emissions.
As a result, the EPA listed coke oven emissions as a HAP
under §112(b)(1)(A) of the Act on September 18, 1984. The
passage of the Clean Air Act Amendments in November 1990
reinforced the list of HAPs. Section 112(d)(8) of the amendments
specifies the actions to be taken to implement Emission Standards
for Coke Ovens. The listing is supported by occupational
exposure studies of coke oven workers that show statistically
significant excess mortality from cancers of the respiratory
tract, kidney, prostate, and all cancer sites combined. Based on
the entire set of health studies, the Administrator concluded
that coke oven emissions present a significant risk of cancer to
the public. In addition, coke oven emissions in the presence of
an air inversion also may be related to episodes of asthma,
bronchitis, and other acute respiratory conditions in both
children and adults3.
Because compliance with the proposed coke oven emission
regulations would be achieved through the containment of
emissions in the oven rather than through the collection and
removal of emissions by a pollution control device, the current
level of water effluent discharges, solid waste, energy use, and
2-1
-------
noise would be unchanged4. Subsection 2.2.3 provides more detail
on the health risks of the pollutants of coke oven emissions.
2.2 NEED FOR REGULATION
2.2.1 Market Failure
The U.S. Office of Management and Budget (OMB) directs
regulatory agencies to demonstrate the need for a major rule.
The regulatory impact analysis must show that market failure
exists and that it cannot be resolved by measures other than
Federal regulation5. Market failures are categorized by OMB as
externalities, natural monopolies, or inadequate information.
The following paragraphs address the three categories of market
failure. Chapter 4 discusses the regulatory options and the
requirements for a federal regulation under the CAAA of 1990.
2.2.1.1 Air Pollution as an Externality. Air pollution
is an example of a negative externality. This means that, in the
absence of government regulation, the decisions of generators of
air pollution do not fully reflect the costs associated with the
pollution. For a coke oven battery operator, pollution from coke
oven emissions is a by-product that can be ignored or disposed of
cheaply by venting it to the atmosphere. Other than his concerns
for meeting OSHA requirements, the coke oven battery operator
does not fully realize the social cost of the pollution created,
and do not "internalize" the damage caused by emissions. This
damage is born by society, and the receptors - the people who are
adversely affected by the pollution - are not able to collect
compensation to offset their costs. They cannot collect
compensation because the adverse effects, like increased risks of
morbidity and mortality, are by and large, non-market goods (that
is, goods that are not explicitly and routinely traded in
organized free markets).
Consider an example. It may be somewhat unreal, but it
illustrates why air pollution is a market externality. A young
man estimates that over his remaining lifetime he has a risk of
getting cancer of, let's say, 4 chances in 10. A new coke oven
battery is being constructed in his neighborhood, and he
pessimistically calculates that the added pollution to his own
environment will boost his odds of getting cancer to, say, 5
chances in 10. He walks up to the owners of the coke oven
battery and offers to "sell his exposure" to the air pollution
generated by the coke oven battery for a bargain basement price
of just $5 a day. For his efforts he gets no more than a laugh.
What's wrong? Most young men either would be unwilling to even
consider such a transaction, or, if they were willing, they would
not know enough about their futures and about the effects of the
pollution to set such a precise price. Furthermore, even if they
are willing and did have a price, they would not have any good
way of coming to terms with the coke oven battery owners. The
2-2
-------
owners have little incentive to pay anything to the receptors of
the air pollution. Little incentive, that is, unless government
requires them to pay or to reduce the pollution.
How would it help to force coke oven batteries either to
compensate the people suffering the consequences of the
pollution, or simply to reduce the pollution? Where there are
negative externalities like air pollution, the market price of
goods and services does not fully reflect the costs borne by
receptors of air pollution, generated in the course of producing
the goods and services. Government regulation can be used to
improve the situation. The proposed NESHAP will increase the
cost of production of coke to include not only the actual
production costs, but also the cost of damages to the atmosphere
created from the pollution generated by coke production. This
increase in cost will force coke oven battery owners and
operators to reduce the amount of air pollution they emit. With
the NESHAP in effect, what coke oven battery owners and operators
must spend to operate may more closely approximate the full
social costs of coke production. This does not, however, imply
that total control is most efficient since coke producers may
still choose to emit pollutants even if they are required to pay
all social costs of production. Overall, if we could internalize
all negative externalities in the country - including, of course,
those from coking - society's allocation of resources would be
improved.
2.2.1.2 Natural Monopoly. Another cause for government
intervention to bring about a socially optimal allocation of
resources is when a natural monopoly exists. When there are
relatively few firms in an industry, due to some barrier to entry
(i.e. heavy up-front capital needed to enter), these firms have a
majority of the power to influence prices and quantities of the
good in the market. The monopolistic power that naturally occurs
in this type of market does not provide the competitive market
checks and balances that are needed to ensure the best
utilization of society's resources.
Because the majority of coke production (with the exception
of foundry coke) is produced and used internally, it is difficult
to evaluate a market for coke directly. Instead, we must look at
trends in related markets, like steel, to estimate the market for
coke. The steel industry, in general, is a relatively
competitive market. Final products of steel are imported and
exported, creating competition for market share among producers
domestically and internationally. In recent years, however, the
steel industry in the United States has experienced a decline in
output. Technical changes in steel manufacturing in the near
future will result in a decreased quantity of coke required per
ton of steel produced6. In 1980, there were 60 plants in
operation with 195 batteries. During the period from 1980 to
1991, the number of plants fell to 32, and the number of
2-3
-------
operating batteries declined by more than 50 percent, to 90
batteries7 (additional closures are expected during the
development of the NESHAP). This decreasing demand has limited
the number of firms in the industry, but this does not indicate
that a natural monopoly exists.
Overall, the lack of competition in the steel industry (and
coke industry) is not a problem. Therefore, the proposed NESHAP
is not designed to address this problem.
2.2.1.3 Inadequate Information. The third category of
potential market failure that sometimes is used to justify
government regulation is inadequate information. As stated in
the Guidelines to Performing Regulatory Impact Analysis8, tlie
optimum level of information is not necessarily the maximum
possible amount, because information, like other goods, should
not be produced when the costs of doing so exceed the benefits.
It would certainly be costly to have each individual in the
country to search for information on the emissions of coke ovens
by travelling to each facility "and gathering data. Although the
amount of coke oven emissions currently placed in the atmosphere
and its toxicity is not open information to the general public,
the flow of information on control techniques and work practices
is adequate and does not create a category of market failure
under inadequate information. This information, while not
provided by the producers of coke oven emissions, may be supplied
by news media, consumer and environmental groups, public health
agencies and similar services.
Regulatory intervention to address and information problem
will not be undertaken for this NESHAP due to the lack of
substantial reason to believe that private incentives to provide
information are seriously inadequate. The fact that industry
representatives and environmental groups are willing to negotiate
throughout this regulatory process shows a fairly free-flowing
rate of information.
2.2.2 Harmful Effects of Coke oven Air
The toxic constituents of coke oven emissions include both
gases and respirable particulate matter with a varying chemical
composition. Historically, the greatest attention has been
focused on toxic effects of the benzene soluble organic (BSO)
portion of particulate matter emitted from coke ovens, because
this fraction includes compounds that are known animal
carcinogens. In addition, beryllium, benzene, and arsenic are
known carcinogens emitted from coke ovens. There is also concern
over the potential health risks caused by long-term exposure to
trace metals (e.g. cadmium, chromium, lead, and nickel) and gases
2-4
-------
(e.g. hydrogen sulfide, carbon monoxide, nitric oxide and sulfur
dioxide) contained in the coke oven emissions9.
As mentioned earlier, the listing of coke oven emissions as
a hazardous air pollutant under §112 of the Act is supported by
occupational exposure studies of coke oven workers that show
statistically significant excess mortality from cancers of the
respiratory tract, kidney, and prostate, and all cancer sites
combined. There is sufficient evidence for carcinogenicity in
humans and experimental animals for the International Agency for
Research on Cancer (IARC) to classify coke oven emissions as
Category 1, meaning that this mixture is carcinogenic to
humans10.
2.3 CONSEQUENCES OF REGULATION
2.3.1 Consequences if EPA's Emission Reduction Objectives are
Met
2.3.1.1 Allocation of Resources. There will be improved
allocation of resources associated with coke production.
Specifically, more of the costs" of the harmful effects of coke
production will be internalized by coke oven plants. This, in
turn, will affect consumers' decisions on whether, where, how,
and how much of a product produced using coke (i.e. iron and
steel products) to use. To the extent these newly-internalized
costs are then passed along to these consumers, and to the extent
these people are free to buy as much or as little of the products
of coke as they wish, they will purchase less (relative to their
purchases of other competing services). If this same process of
internalizing negative externalities occurs throughout the entire
coke producing industry, an economically optimal situation is
approached. This is the situation when the marginal cost of the
resources devoted to coke production equals the marginal value of
the products to the people who are using the products produced
using coke. There are many "ifs" in this chain of events. It is
easy to cite situations where the air pollution control costs
will not ripple through as suggested here and effect decisions by
the consumers of coke-produced products. Nevertheless, in the
aggregate and in the long run, the proposed NESHAP will move
society toward this economically optimal situation.
2.3.1.2 Emissions Reductions and Air Quality. Under the
proposed standard, it is estimated that emissions of coke oven
air pollutants will be reduced by 66 percent per year under MACT
and by 84 percent per year under LAER. For more information on
this topic, refer to Chapter 7 on the benefits of the regulation.
Air quality will improve, however, this analysis does not
translate emission reductions into ambient air quality
improvements.
2.3.1.3 Costs, Benefits, and Economic Impacts. The
2-5
-------
national annual cost of emission control under MACT will increase
by approximately $25 to $33 million by 1995. The estimated
annual cost of emission control under LAER is $46 to $57 million
by 1998. Expected benefits include reduced risks for mortality,
morbidity, and other adverse health and welfare effects from
lower levels of VOC emission reductions (Reference: Chapters 7
and 8). The resulting NESHAP will create relatively small market
changes in price and quantity, and result in , at most, two
battery closures. Additionally, the impacts on small coke
producers is minimal (Reference: Chapter 6).
2.3.1.4 Water Quality, Solid Waste, and Energy Impacts.
As previously mentioned, because compliance with the proposed
coke oven emission regulations would be achieved through the
containment of emissions in the oven rather than through the
collections and removal of emissions by a pollution control
device, the current level of water effluent discharges, solid
waste, and noise would be unchanged11. A minor increase in
energy use may occur from the coke by-product removal. This
impact, however, is insignificant and is not considered in the
analyses that follow.
2.3.1.5 Technological Innovation. Section 112 of the
CAA regulations serve to disseminate both pollution control and
coke oven battery technology, and to stimulate further
technological development. Coke oven facility constructors have
the freedom to seek the most economical way to comply with
standards. The proposed NESHAP may promote the sharing of
technology with other countries, and probably will open new
directions of research in coke production technology.
2.3.1.6 State Regulation and New Source Review. State
regulatory programs will be strengthened. The NESHAP will be
delegated to the states for enforcement. Assuming states do not
pull resources from other programs to handle their enlarged
responsibilities, there will be a natural strengthening of state
air pollution control staffs. Recognition that the NESHAP is
effectively reducing emissions will expedite the state process of
reviewing applications for new coke oven batteries and issuing
permits for their construction and operation. There will be less
controversy involved. Finally, state regulations will be
uniform, and the disadvantages of the piecemeal approach to
emission regulation will be avoided.
2.3.2 Consequences if EPA1a Emission Reduction Objectives are
not Met
The most obvious consequence of failure to meet EPA's
emission reduction objectives would be emissions reductions and
benefits that are not as large as EPA is projecting. However,
costs are not likely to be as large either. Whether it is
noncompliance from ignorance or error, or from willful intent, or
2-6
-------
simply slow compliance due to owners and/or operators exercising
legal delays, poor compliance can save some resources money.
Unless states respond by pouring more resources into enforcement,
then poor compliance could bring with it smaller aggregate
nationwide control costs. EPA has not included an allowance for
poor compliance in its estimates of emissions reductions. This
is because the potential effects of poor compliance are expected
to be minor.
If the emission control devices degraded rapidly over time
or in some other way did not function as expected, there could be
a misallocation of resources. This situation is very unlikely
because the NESHAP is based on demonstrated technology. Other
ways the regulation could fail are conceivable.
2-7
-------
REFERENCES
1. U.S. Congress. The Clean Air Act of 1970, § 112.
2. Research Triangle Institute. Economic Analysis of Air
Pollution Regulations: By-Product Coke Ovens. Final Report
prepared for U.S. Environmental Protection Agency, OAQPS. RTI
Project No. 5153-9 FR. June 1992. pp. 4.1-4.5.
3. U.S. Congress. The Clean Air Act, as amended. Title III,
§112. November 15, 1990.
4. Federal Register. Volume 52, No. 78; 40 CFR Part 61. U.S.
Environmental Protection Agency. April 23, 1987. pp. 13586-
13686.
5. U.S. Office of Management and Budget. Regulatory Impact
Analysis Guidance. Appendix V of Regulatory Program of the
United States Government. April 1, 1988 through March 31,
1989.
6. Reference 2, pp. 2-1.
7. Reference 2, pp. 3-7.
8. U.S. Environmental Protection Agency; Office of Policy,
Planning, and Evaluation; Office of Policy Analysis.
Guidelines for Performing Regulatory Impact Analyses.
Publication No. EPA 1230-01-84-003. December 1983 with
updates through march 1988. pp.
9. Reference 4, pp. 13587.
10. Reference 4, pp. 13587.
11. Reference 4, pp. 13596
2-8
-------
CHAPTER 3
CONTROL TECHNIQUES
This chapter discusses the technology for control of
emissions from wet-coal charging, oven door leaks, and topside
leaks during coking.
3.1 TECHNOLOGY FOR THE CONTROL OF EMISSIONS FROM CHARGING
Charging practices have been altered by past efforts of
regulatory agencies and coke oven operators to reduce emissions.
In the past, the most common procedure was to isolate the gas-
collection system from the oven and charge the coal into the red-
hot ovens simultaneously through three to five charging holes in
the top of the oven. When the moist coal entered the hot oven,
it displaced the air. This displacement and the immediate
gasification of moisture and volatile components of the coal
caused the oven pressure to rise sharply. Because the gas-
collection system was blocked off, the only escape for the smoke,
hydrocarbons, gases, and steam was to the atmosphere through any
opening. Techniques to control these emissions are discussed in
this section.
Investigations have revealed that successful control of
these emissions is often more dependent on adherence to specified
work practices and operating procedures than on the design of
charging equipment. Consequently, the following discussion will
concentrate on these procedures as well as required equipment
modifications. The one control system (or procedure) used at
present is stage charging.
3.1.1 Stage Charging
Stage charging was first developed in England in the 1950's
and was more recently applied in the United States. Although
there are necessary equipment requirements, these do not include
conventional air pollution control devices such fabric filters,
electrostatic precipitators, and scrubbers.
Stage charging is the ordered pouring of coal into the oven
so that, regardless of coal flow, a consistent exit space is
maintained for gas. The larry car hoppers are discharged in an
3-1
-------
ordered sequence so that an open tunnelhead is maintained at the
top of the oven until the last hopper is discharged. Emissions
are effectively contained in the ovens and collecting mains by
steam aspiration, and they are exhausted through the regular gas
handling equipment.1 Successful stage charging is dependent on
many factors such as equipment design, maintenance, and operating
procedures. Each factor is significant; failure in only one area
can negate all the money and effort expended in the other areas.
3.1.1.1 Description of Stage Charging. Stage charging
is primarily an operating technique based on a predetermined
sequence for simultaneously charging coal from one or two larry
car hoppers into the incandescent ovens. The ovens are
maintained under a slight negative pressure by applying steam
aspiration in the goosenecks of the offtake. The assembly that
comprises the standpipe and gooseneck is often called the offtake
or ascension pipe. The stage charging technique is uncomplicated
but requires close attention to detail. The most important
aspects of stage charging are good aspiration and the operating
crew's strict adherence to specific charging and leveling
practices.2
Perhaps the most important ingredient of a good stage charge
is adequate aspiration from the oven. In essence, a slightly
negative pressure must be maintained at every open charging port
throughout the entire charge. Steam is the common aspiration
fluid, although it has been reported that liquor sprays can
achieve stronger aspiration.3 Other advantages reported for the
liquor sprays are less frequent cleaning of nozzles and less
steam, condensate, and heat in the effluent from the mains.
However, aspiration efficiency of steam and liquor sprays is
sensitive to the flow and pressure maintained at the nozzles.
Steam pressures and nozzle sizes necessary for adequate
aspiration vary from plant to plant depending on factors such as
offtake design and amount of air leakage into the oven. To
ensure adequate aspiration, the stream nozzles require frequent
inspection and cleaning.
The rate at which gases can be aspirated from an oven must
be limited. Too high a rate can pull an excessive amount of coal
dust into the collecting mains. Therefore, aspiration systems
are carefully designed to provide just enough draft on the oven
to prevent emissions during the charging cycle.
Timely removal and replacement of the lids on the charging
ports is crucial to good stage charging. For this, both manual
or automated methods can be used. The most common manual method
uses hydraulically operated electromagnets.4 Either method can
be effective, although the automated system may reduce the
incidence of improper alignment of the lids. Any mechanical
3-2
-------
system must allow the lids to be moved individually.
Another necessary aid used to minimize emissions is a seal
which closes the opening between the leveler bar and the chuck
door when the bar is in the oven. As previously discussed, any
air that enters the oven thwarts the aspiration system. A seal,
commonly referred to as a leveler boot, can be used throughout
the charge to preclude any air intrusion at the chuck door. To
operate without air leakage, the leveler bar is extended to the
chuck door opening the seal before initiating the charge. The
opening remains sealed until the leveler bar is retracted at the
end of the charge.
Even with an adequate flow of the aspirating fluid through
clean nozzles, the system may fail if the standpipes and
goosenecks do not provide a clear passage for the gases to flow
from the oven. During the coking cycle, carbon deposits reduce
the cross-sectional area of the standpipes and goosenecks. Well-
controlled stage charging can be achieved only if these deposits
are removed before every charge as a routine part of the charging
process.
Carbon deposits on the roof of the oven must be removed
before every charge to ensure unimpeded gas flow across the top
of the oven. This roof carbon is removed by blades that are
mounted on the top of the pusher ram and that scrape the carbon
off as the coke is pushed. Sometimes high-pressure air jets
mounted at the top of the ram are used to provide better
cleaning. Even if all other forms of emission control are
operating perfectly, failure to remove excessive roof carbon may
result in poor emission control.5
Another essential factor to good emission control is to
ensure that the entire length of the oven is under vacuum. On a
new plant, this condition can best be achieved with two mains,
standpipes, and aspiration systems. One main is placed at each
end of the ovens. If the coal blocks the open space at the top
of the oven each side of the blockage will remain under vacuum.
An obvious design consideration is the amount of coal in
each hopper of the larry car. This amount is predetermined so
that the coal dumped from all but the last hopper will peak below
the "coal line." The amount of coal placed in each hopper
depends on the size of the oven, the number of charging holes,
and the angle of repose assumed by the coal. This angle of
repose is influenced by the bulk density of the coal, oil
additives, moisture content, and the particle size to which the
coal is ground. The amount placed in each hopper varies from
plant to plant, depending on the three factors mentioned earlier.
Just before charging the coal, the larry car must be
3-3
-------
accurately aligned over the charging holes. Poor alignment can
result in spillage of coal on the top of the battery where the
coal is heated and excess emissions are produced. A poor fit
between the drop sleeves and charging hole permits excessive air
leakage into the oven. Air drawn into the oven may overpower the
aspiration system. When the gas volume exceeds the capability of
the aspiration system, pressure will rise elsewhere in the oven,
and emissions may escape to the atmosphere.
Stage charging achieves a marked reduction in emissions by
aspiration of air into the by-products system from points where
otherwise pollutant gases and smoke are emitted. The aspiration
steam and the inspired air significantly affect the by-product
plant. Furthermore, aspiration can cause coal fines to be
entrained into the by-product system. The steam increases the
wastewater load, the air increases the density and reduces the
heating value of the coke oven gas, and the coal fines tend to
reduce the acceptability of the coal^tar .pitch to the
manufacturers of carbon or synthetic graphite electrodes, a major
end use. Careful attention to the selection of nozzle dimensions
and steam pressure, coupled with adherence to procedures which
avoid needless air leakage into"the system, can usually decrease
the problem of coal fines. The primary cooler system usually
removes tar "sludge" containing the coarser coal particles.6
3.1.1.2 Optimizing Stage Charging. Optimizing stage
charging generally is a function of the performance of the
battery workers, particularly the battery top workers. A
detailed, written procedure, an effective training program, and
coordination of the battery top worker's activity are required.
A few of the important worker job functions are listed below:
* Inspection and cleaning of the gooseneck.
* Prompt lid replacement.
* Prompt luting of lids.
* Turning the aspiration system on and off.
* Observing the position of drop sleeves.
* Spotting the larry car.
* Notifying the larry car operator if the car is improperly
spotted.
* Continuing a program of cleaning and maintenance.
* Consistently following operating procedures.
3-4
-------
A study of a specific battery's operation would aid in
developing an optimum written procedure and effective training
program. In addition, such a study may reveal minor equipment
modifications that are peculiar to that specific battery and may
dramatically improve control. For example, CF&I discovered that
altering the steam nozzles improved aspiration for their
particular case. The company also discovered that constructing a
platform for inspecting goosenecks improved the gooseneck
inspection and cleaning procedure. Worker coordination,
training, and communication were also improved.7
3.2 TECHNOLOGY FOR THE CONTROL OF DOOR LEAKS
Because charging takes only a few minutes, while coking
continues for many hours, it might seem that door leaks are a
much more serious and pervasive problem than charging emissions.
However, outward leakage from any given door (2 per oven in a
battery of perhaps 50 ovens) may occur near the beginning of the
cycle and inward leakage may occur later. Inward leakage of air
affects the utility of the coke oven gas; therefore, coke oven
operators were concerned about door leakage before most
regulatory agencies were formed.
Control techniques for coke oven door emissions may be
separated into four basic categories:
* Oven door seal technology,
* Pressure differential devices,
* Hoods and sheds over doors, and
* Operating and maintenance procedures.
The first category relies on the principle of producing a
resistance to the flow of gases out of the coke oven. This
resistance may be produced by a metal-to-metal (or hard) seal, a
resilient soft seal, or a luted seal. Small cracks and defects
in the seal permit pollutants to escape from the coke oven early
in the cycle. The magnitude of the oven door seal leak is
determined by three factors: (1) size of the opening, (2) the
pressure drop between the oven and the atmosphere, and (3) the
composition of the emission.
The effectiveness of a pressure differential control device
depends on the ability of the device to reduce or reverse the
pressure differential across any defects in the door seal. These
systems either provide a channel to permit gases evolved at the
bottom of the oven to escape to the collecting main or provide
external pressure on the seal through the use of steam or inert
gases.
3-5
-------
Oven door emissions to the atmosphere also can be reduced by
collection of the leaking gases and particulates and subsequent
removal of these pollutants from the air stream. A suction hood
above each door with a wet electrostatic precipitator for fume
removal is an example of this type of system.
Other control techniques rely on operating and maintenance
procedures rather than only hardware. Operating procedures for
emission reduction could include changes in the oven cycle times
and temperatures, the amount and placement of each charge, and
any adjustments of the end-door while the oven is on line.
Maintenance procedures include routine inspection, replacement,
and repair of control devices and doors.
Performance of the control techniques for doors is expressed
in terms of visible emissions as percent leaking doors (PLD).
The number of doors leaking is divided by the total number of
doors and then multiplied by 100 to obtain the PLD. The chuck
door is not counted as a separate door but is considered as part
of the pusherside door.
3.2.1 Oven Door Seal Technology
Oven door seals can be divided into three subsets: hard
seals, soft seals, and luted seals. Hard seals contain the oven
gas by pressing a metallic strip against the oven jamb. To
obtain uniform pressure, the metallic strip has adjustable
screws, springs, or cams. Soft seals are resilient and they
permit the seal to conform to the shape of the door jambs to seal
in the gas. Luting is a water-based dispersion of clay and other
materials which flow to seal the door. The oven heat evaporates
the water and the luting composition dries in position. A
combination of hard and soft seals sometimes is used in pressure-
differential devices such as the prechamber door.
Hard seals rely on the principle of self-sealing. Emissions
that contain steam, volatile oils, and tars pass through small
defects in the sealing surface. The tars condense and seal the
small openings after the steam content is reduced. The time
required for self-sealing varies with oven pressures and gap size
in the door.
Hard seals are commonly used in the production of
metallurgical coke. The major types of industrial seals used in
the United States are the Koppers and the Wilputte seals, named
for their manufacturers.
Luting, which is one of the oldest coke oven control
technologies, is used commercially on many foundry batteries and
on a few furnace batteries. Hand-luted doors are sealed by
trowelling a luting mixture into a V-shaped opening between the
metal door frame and a roll formed steel shape (door jamb) on the
3-6
-------
end of the oven. The luting is a mixture of clay, coke breeze,
and water, which dries and seals the gap between the door jamb
and the door. The carbonaceous material is added to reduce
shrinkage and cracking upon curing.
Some advantages of luting are that there are no leaks when
it is properly applied, and luted doors are less costly to
maintain than self-sealing doors. However, the luting mixture is
applied manually and workers may be exposed to fumes during the 5
minutes required to lute a door. Another concern is that the
luting can crack because of the mild combustion explosion from
the coal first entering the oven. Reluting may be required after
charging to avoid uncontrolled fires if cracks develop or the
luting is jarred loose. The removal and disposal of luting
material does not represent a major problem because it is
recyclable.
Luting is particularly interesting because it represents an
emission control technology that has the potential for
eliminating almost all door leaks. Unfortunately, the lack of
fully developed luting formulations and the absence of proven
luted door technology for high production rate metallurgical
coking have limited widespread adoption. Although luting has
been used for foundry coking with 30-hour cycle times, the faster
cycle time for metallurgical coking (18 hr) would require
additional manpower, new equipment, and solutions to material
handling problems.
3.2.2 Modern Hard Seals
Currently available major emission control techniques are
based on hard seals. Self-sealing doors with hard seals
invariably have a small clearance between the sealing surface and
the jamb, and tar in the escaping gas ultimately plugs these
small gaps. The time needed to plug the gap depends on the size
of the gap, the temperature of the seal, the pressure in the
oven, and other factors. Leaks cannot be prevented solely by
metal-to-metal contact without plastic (irreversible) deformation
of the metal.
A few coke oven doors have been observed to be completely
free of visible emissions during the entire process.8 One
hundred percent control of visible emissions can be obtained with
Koppers doors when new seals, well-adjusted doors, and relatively
straight and clean jambs are used. However, the performance
often begins to deteriorate in less than 6 months.
Effective sealing is inhibited by several factors, including
distortion and damage to jambs, doors, sealing strips, and
adjusting hardware.' Most of the components of the oven's end-
door assembly are tightly constrained; consequently, when the
3-7
-------
assembly is heated, stresses result because gross distortions are
prevented. Thermal cycling under these constrained conditions
causes thermal warping of the metal components. Occasional
temperature excursions and fires from leaking doors also cause
warping, because plastic deformation occurs at temperatures of
500°c or higher. Although no metal-to-metal contact without
plastic deformation will prevent leaks, this very deformation can
lead to long-term problems in maintaining close tolerances on
sealing surfaces.
3.2.3 Other Hard Seals
There are several other types of hard seals. One is the
Ikio seal (oven door). The Ikio oven door, a Japanese
technology, is similar to a modified Wilputte oven door.
However, rather than the knife edge holder being welded to the
door diaphragm or sealing plate as in the Wilputte door, the Ikio
oven door has a flexible sealing plate which is welded directly
onto the knife edges. Another difference is that the Ikio door
has springs which can be moved forward and backward up to 15 mm
(0.6 in) and are positioned 300 mm (12 in) apart along the sides
of the oven door, unlike the adjustable screws used with the
Wilputte door. This unique construction is claimed to seal in
the gas completely.
A second type of hard seal is the Battelle seal, developed
by the Battelle Columbus Laboratories in association with AISI
and the EPA. The seal is retrofittable both to Wilputte and
Koppers enclosure systems. The concept behind this seal is to
provide a seal which is highly flexible in the direction
perpendicular to the face of the jamb. The design stress levels
are below the allowable stress for the high-temperature material.
If these objectives are achieved, the seal will conform to a
badly distorted jamb without taking a permanent set at the normal
operating temperatures.
A third type of hard seal has been used commercially in
Japan since the mid-1970*s. This gas-sealed door technology as
practiced at the Kamiashi Works includes use of a luting mixture
on the exterior sealing surface edge and use of gas pressure to
prevent leakage.10 The gas used is nontoxic, nonexplosive,
noncombustible, and generally available to all coke plants.
Details of the seal design, gas type, and gas distribution system
are confidential.
In practice, the gas supply to each door is turned on after
the push is complete and the door has been reinstalled. Gas is
injected in the oven chamber at 200 kPa (30 psi).11 While the
oven is charging, workers rap the knock-type seal to force the
sealing edge into residual tar deposits and effect a partial
seal. As points of leakage are identified, a luting mixture
3-8
-------
composed of mortar and coke breeze is spread along the exterior
edge of the seal. The amount of luting mixture applied is very
small in comparison to the amount used in hand-luted door
practice. The worker can hold the supply of luting mixture in a
pan that fits the palm of his hand. Reluting is practiced as
necessary to prevent obvious leaks.
3.2.4 Saturn Doors
The Saturn door is a flexible door technology that includes
several features different from those of conventional doors and
seals:
* Flexibility: The door and doorplug are constructed in
two or three sections, which allows the doors to flex
(either in the concave or convex direction) more than the
rigid conventional doors to conform to warped jambs with an
outward or inward bow.
* The seal: The door seal is constructed of InconelR,
which is a durable, heat-resistant alloy that is flexible
and easily repaired. The seal is mounted on a flexible
diaphragm plate.
* Leaf springs: Pressure is maintained on the seal by
leaf springs that provide continuous force around the door
perimeter instead of the point loading of plungers (spaced
at 10 to 20 cm [4 to 8 in] intervals) used on conventional
doors.
A major advantage of the Saturn door is in seal maintenance
because the seal can be relatively easily repaired at the plant
by replacing a damaged seal section or filling in place.
Conventional doors generally are sent to outside contractors for
seal replacement and repairs, and these repairs are usually more
extensive and expensive. Another advantage is that the door can
be more easily adjusted to conform to the deflections of warped
jambs. The use of leaf springs instead of numerous spring-loaded
plungers also improves seal performance by providing a uniform
and continuous pressure on the seal against the jamb. This
eliminates the need for manual adjustment of numerous spring-
loaded plungers, which may result in too much or too little
pressure on the seal.12
As of 1988, the only complete set of Saturn doors in place
was on a 6-meter battery with 60 ovens. (Single experimental
doors were being evaluated in trials at several other coke
plants.) This battery had several structural design problems and
also had a leakage problem from warped jambs. The design
problems were being remedied, and all of the jambs on the battery
were being replaced. The design problems and warped jambs were
•
3-9
-------
unrelated to the installation or indicated emission performance
of the Saturn doors. Fifty-eight observations of door leaks
conducted on the doors showed an overall average of 4.5 PLD and
an upper 95 percent confidence level of 7 PLD (based on a three-
run average) .
While the use of Saturn door technology could require the
replacement of existing doors with the new flexible doors, no
major modifications to existing door machines are expected. The
capital cost (as of 1988) for a new door for short ovens ranged
from about $10,000 to $12,000 per door and about $160,000 each
for a door or jamb cleaner. For a typical 60-oven battery, the
capital cost for the doors was about $1.3 million, and the cost
of door and jamb cleaners added about $0.64 million for a total
of about $2 million per short battery. For a tall (6 meter)
battery with 60 ovens, the capital cost rose to $3.2 million, $2
million for the doors and $1.2 million for the cleaners.13
3.3 OTHER CONTROL TECHNIQUES
There are several other techniques for the control of
emissions from door leaks. Emphasis is given here on work
practice controls, fume collection, design modifications,
promising technology that is in the trial stage, and proven
technology that is receiving greater notice.
3.3.1 Operating and Maintenance Procedures
The avoidance of leakage greatly depends on good operating
practices and maintenance. One factor in emissions control is
provision of the best possible work environment for the
operators. In one reference (Graham and Kirk) , good operating
and maintenance practices were advocated: "In some ways it is
the easiest method, probably the least expensive, the most
effective, and justifiably on many occasions 'the best practical
means' of controlling the pollution."14
Good operation requires the removal of deposits from the
sealing edge and the jamb. Cleaning is perhaps the most
burdensome task of the coking process, and workers tend to
overlook the hard-to-reach sections on high oven doors. The task
of manual cleaning is more difficult on the taller ovens than on
the short ovens, and there is a greater tendency to not clean the
oven thoroughly.
For a long time, coke plant operators have recognized that
large, unchecked leaks may cause fires and damage the seal and
buckstays beyond repair. The cleaning of the metal surfaces
which come into contact should emphasize the removal of
encrustations and particulate deposits which can cause gaps that
allow leaks.
3-10
-------
Cleaning the doors and jambs to the bare metal does not
always provide additional benefits. It has been reported that
excessive leakage always resulted when the operators scraped the
jambs to bare metal.15 With Wilputte doors, a thin film of tar
should remain after cleaning to aid in sealing gaps between the
knife edge and jamb.
The seal on hard coke oven doors should be maintained so
that they meet a maximum gap specification. Typical gap
specifications are 0.005 to 0.008 cm (0.002 to 0.003 in). The
use of temperature-resistant materials and seals designed to
provide uniform sealing pressures can reduce much of the
maintenance effort necessary to meet the specifications.
3.3.2 Effects of Process Variables
Oven pressure is a process variable which can be used to
moderate emissions because flow through small leaks is
proportional to the pressure differential. A disadvantage of
using low overhead pressure as a control technique is that low
oven pressures have been reported to cause severe damage to the
oven wall brickwork.16 Also, oxygen that is introduced into the
oven at the later stages of the coking cycle by the low pressures
causes soot formation that can block the ascension pipes and
lower the quality of the coke oven gas.
Temperature effects of the process variables damage the door
components and increase emissions. The major drawback to
lowering the coke to minimize thermal damage is the resulting
increase in cycle time and decrease in coking capacity.
Decreasing the coke temperature from 1,000°C to 800°C could
conceivably reduce capacity by one-half. The temperature change
could also alter the composition of the by-product gas.
3.3.3 Startup, Shutdown, Upsets, and Breakdowns
The emission rate from coke oven doors during startup and
shutdown operations is expected to be lower than during normal
operations. One reason for the lower emission rate is that the
initial heatup or shutdown of a battery requires 5 to 7 weeks;
therefore, shutdowns are very infrequent and are undertaken only
as a last resort. Also, the coking rate is much slower during a
heatup or shutdown because of lower coking temperatures. This
slow coking rate results in a slower evolution of gases, lower
oven pressures, and consequently fewer and smaller leaks. An
oven is probably in better condition during an initial heatup,
because repairs are usually performed during a shutdown.
Process upsets that affect oven pressures have a significant
impact on door emissions. For example, the pressure regulating
valve, usually located in the crossover main between the
3-11
-------
collecting main and suction main, controls the collecting main
pressure and consequently affects oven pressure. If this valve
malfunctioned, the oven pressure could increase and cause an
increase in door leaks. Dirty standpipes and goosenecks may plug
during the coking cycle and cause excessive pressures to build up
in the oven. This plugging problem is remedied by regular,
periodic cleaning. The pressure in the bottom of the oven may
also be increased if the door's gas channel or vented plug fouls
because of accumulation of carbon deposits. Upsets like this may
be avoided by regular inspection, maintenance, and cleaning.
Another factor that may cause a process upset is the
introduction of a high-moisture or high-volatile coal into the
oven. The rapid evolution of these additional volatiles at the
beginning of the coking cycle may increase the oven pressure and
initially increase leaks and door sealing time.
3.3.4 Door Controls for Tall Ovens
The metal-to-metal seal technology that was previously
discussed applied to tall (6-meter) ovens. However, a review of
the available data indicates that these tall ovens do not control
door leaks as well as the smaller ovens. Among the explanations
for this phenomenon is the increased potential for leak
occurrence (by a factor of two) because of the larger door
perimeter where the sealing edge must contact the jamb. In
addition, the oven pressure at the bottom of a 6-meter door is
greater than on smaller doors for batteries operating at the same
collecting main pressure. Also, when dry coal is charged, the
resulting pressure surges may increase leaks.
3.3.5 Thompson Non-Recovery Coke ovens
This non-recovery design was developed by B. Ray Thompson
more than 30 years ago. The process has been operational for
over 20 years. The Thompson design is used at only one location
in the United States, the Jewell Coke and Coal Company facility
in Vansant, Virginia, similar to by-product coke ovens, Thompson
non-recovery coke ovens use heat to produce coke. The pushing
and quenching of the coke is also similar to the same steps at
by-product coke ovens. Thompson coke ovens, however are not
designed to recover the chemical by-products from the raw coke
oven gas as is done at most coke plants. Instead, the raw coke
oven gas is burned to provide the energy for coking, and excess
heat is used to generate steam. The coking cycle is 24 hours
rather than 18 hours which is typical for by-product ovens.
During coking, the raw coke oven gas is removed from the
ovens by a natural draft (exhausters are not required), which
maintains negative pressure in the ovens. This is the exact
opposite of how a by-product oven operates. The non-recovery
ovens have two basic emission points, the vents to the atmosphere
3-12
-------
and the doors during charging. No lids or offtakes are present.
There are no leaking door emissions. No pollution controls are
needed on the stack at the plant; however, ovens constructed in
nonattainment areas may need to consider controls for sulfur
dioxide and particulate matter.
The Thompson non-recovery coke oven does not have to be
pushed and charged in a special sequence as is done at by-product
batteries. Instead, the ovens are pushed sequentially, and the
charging is performed in one shift, which results in lower
manpower requirements. This type of coke oven is not affected
by this NESHAP.
3.4 TECHNOLOGY FOR THE CONTROL OF TOPSIDE LEAKS (CHARGING PORT
LIDS AND STANDPIPES)
3.4.1 Description of Technology
>
Topside leaks occur around the rims of charging port and
standpipe lids; standpipes can also leak at their bases or
through other cracks. These leaks are primarily controlled by
proper maintenance and operating procedures which include:
* Replacement of warped lids,
* Cleaning carbon deposits or other obstructions from the
mating surfaces of lids or their seals,
* Patching or replacing cracked standpipes,
* Sealing lids after a charge or whenever necessary with
a luting mixture, and
* Sealing cracks at the base of a standpipe with the same
luting compounds.
Luting mixtures are generally prepared by plant personnel
according to formulas developed by each plant. The consistency
(or thickness) of the mixture is adjusted to suit different
applications. Charging port lids are relatively horizontal;
therefore, a thinner mixture can be used to seal them. Standpipe
lids come in a variety of positions; those that are not
horizontal require a thicker mixture to prevent runoff. Careful
application of lute is necessary to prevent a buildup of residue
which can cause standpipes to burn out. The buildup must be
removed from sealing surfaces when lids are opened, to prevent
poor sealing when the lids are closed again.
Some equipment designs may reduce the effort required to
keep leaks sealed. Heavier lids or better sealing edges may
reduce leaks. Automatic lid lifters can rotate charging-hole
lids after they are sealed to provide a better seal. Even with
3-13
-------
such equipment, manual effort will still be required to seal
leaks.
Because there are many places where leaks can develop,
keeping all charging lid and standpipe leaks sealed is a
continuous job. In essence, success in controlling these
emissions is directly related to the amount of manpower and the
dedication of the employees. The number of topside workers
required for effective emission control depends on several
factors, such as the job assignment, number of ovens, cycle time,
and extent of automation. In general, a battery may have a work
force of 4 lidsmen if automatic lid lifters are used or 8 lidsmen
if the lid lifting is performed manually.17 For some batteries,
the larry car operator or helper seals standpipe caps; on other
batteries the lidsmen perform this function.
3.4.2 Performance for Topside Leaks control
Mass emission measurements are not available to indicate
emission control performance for methods of reducing topside
leaks. However, measurement of visible emissions is possible and
provides a good indicator of performance for all emissions.
Emissions are controlled by sealing the leaks or plugging the
holes. If the hole is plugged so that fine particles that make
the emissions visible cannot escape, then all emissions,
including the carcinogens, are controlled effectively. The
emissions from topside leaks for an entire battery are measured
by counting the number of leaks that are visible and expressing
this number as a percentage of total potential leaks. Each
charging port and each standpipe is considered to be a potential
source of one leak.
As explained in the previous section, the primary technique
used to reduce the number of topside leaks is luting. With this
simple technique of sealing holes that allow coke oven emissions
to escape to the atmosphere, it is plausible that the number of
leaks depends on the effort applied to luting. Greater effort
can be achieved by making luting a prime responsibility of
topside workers. Additional manpower may be required to carry
out this responsibility. The emission control performance
increases proportionally with the diligence of workers in
watching for leaks and promptly sealing them.
Data on offtake leaks at Kaiser Steel's Fontana, California
batteries show the effect of additional manpower at one plant.
This plant noted 35 to 56 percent fewer leaks for the seven
batteries over a 3 month period when one employee per battery per
shift was responsible for luting topside leaks and tending to lid
removal and replacement.18 The level of effort provided was
sufficient to lower the average PLO below the local requirement
of 10 PLO.
3-14
-------
Leaks in battery mains on a well maintained battery will
occur infrequently. If battery mains are closely watched by
plant operators, preventive maintenance and prompt repair of
leaks will allow them to be maintained without leaks.
3.5 Flares
Flaring is an open combustion process in which the oxygen
necessary for combustion is provided by the air around the flame.
The organic compounds to be combusted are piped to a remote,
usually elevated, location and burned in an open flame in the
open air using a specially designed burner tip, auxiliary fuel,
and sometimes steam or air to promote mixing for nearly complete
(98 percent minimum) destruction of combustibles. Good
combustion in a flare is governed by flame temperature, residence
time of organic species in the combustion zone, turbulent mixing
of the organic species to complete the oxidation reaction, and
the amount of oxygen available for free radical formation.
Combustion is complete if all combustibles (i.e., VOC's) are
converted to CO2 and water, while incomplete combustion results
in some of the VOC's being unaltered or converted to other
organic compounds such as aldehydes or acids.
Flares are generally categorized in two ways: 1) by. the
height of the flare tip (i.e., ground-level or elevated), and 2)
by the method of enhancing mixing at the flare tip (i.e., steam-
assisted, air-assisted, pressure-assisted, or unassisted).
Elevating the flare can prevent potentially dangerous conditions
at ground level where the open flame is located near a process
unit. Further, the products of combustion can be dispersed above
working areas to reduce the effects of noise, heat radiation,
smoke, and objectionable odors.
In most flares, combustion occurs by means of a diffusion
flame. A diffusion flame is one in which air diffuses across the
boundary of the fuel/combustion product stream toward the center
of the fuel flow, forming the envelope of a combustible gas
mixture around a core of fuel gas. This mixture, on ignition,
establishes a stable flame zone around the gas core above the
burner tip. This inner gas core is heated by diffusion of hot
combustion products from the flame zone.
Cracking can occur with the formation of small hot particles
of carbon that give the flame its characteristic luminosity.19 If
there is an oxygen deficiency and if the carbon particles are
cooled to below their ignition temperature, smoking occurs. In
large diffusion flames, combustion product vortices can form
around burning portions of the gas and shut off the supply of
oxygen. This localized instability causes flame flickering,
which can be accompanied by soot formation.
3-15
-------
3.5.1 Applicability
Flares can be dedicated to almost any VOC stream, and can
handle fluctuations in VOC concentration, flow rate, heating
value, and inerts content. Flaring is appropriate for
continuous, batch, and variable flow vent stream applications.
Some streams, such as those containing halogenated or
sulfur-containing compounds, are usually not flared because they
corrode the flare tip or cause formation of secondary pollutants
(such as acid gases or sulfur dioxide). If these vent types are
to be controlled by combustion, thermal incineration, followed by
scrubbing to remove the acid gases, is the preferred method.20
The majority of chemical have existing flare systems
designed to relieve emergency process upsets that require release
of large volumes of gas. Often, large diameter flares designed
to handle emergency releases are also used to control continuous
vent streams from various process operations. Typically in
refineries, many vent streams are combined in a common gas header
to fuel boilers and process heaters. However, excess gases,
fluctuations in flow rate in the fuel gas line, and emergency
releases are sometimes sent to a flare.
3.5.2 Efficiency
Five factors affecting flare combustion efficiency are vent
gas flammability, auto-ignition temperature, heat content of the
vent stream, density, and flame zone mixing.
The flammability limits of the vent stream influence
ignition stability and flame extinction. Flammability limits are
the stoichiometric composition limits (maximum and minimum) of an
oxygen-fuel mixture that will burn indefinitely at given
conditions of temperature and pressure without further ignition.
In other words, gases must be within their flammability limits to
burn. If these limits are narrow, the interior of the flame may
have insufficient air for the mixture to burn. Fuels, such as
hydrogen, with wide limits of flammability are therefore easier
to combust.
The auto-ignition temperature of a vent stream affects
combustion because gas mixtures must be at a sufficient
temperature and concentration to burn. A gas with a low auto-
ignition temperature will ignite more easily than a gas with a
high auto-ignition temperature.
The heat content of the vent stream is a measure of the heat
available from the combustion of the VOC in the vent stream. The
heat content of the vent stream affects the flame structure and
stability. A gas with a lower heat content produces a cooler
flame that does not favor combustion kinetics and is more easily
3-16
-------
extinguished. The lower flame temperature will also reduce
buoyant forces, which reduces mixing.
The density of the vent stream also affects the structure
and stability of the flame through the effect on buoyancy and
mixing. By design, the velocity in many flares is very low;
therefore, most of the flame structure is developed through
buoyant forces as a result of combustion. Lighter gases
therefore tend to burn better. In addition to burner tip design,
the density also affects the minimum purge gas required to
prevent flashback, with lighter gases requiring more purge.21
Poor mixing at the flare tip or poor flare maintenance can
cause smoking (particulate matter release). Vent streams with
high carbon-to-hydrogen ratios (> 0.35) have a greater tendency
to smoke and require better mixing to burn smokelessly.22 For
this reason, one generic steam-to-vent-stream ratio is not
appropriate for all vent streams. The steam required depends on
the vent stream carbon-to-hydrogen ratio. A high ratio requires
more steam to prevent a smoking flare.
The efficiency of a flare in reducing VOC emissions can be
variable. For example, smoking flares are far less efficient
than properly operated and maintained flares. Flares have been
shown to have high VOC destruction efficiencies, under proper
operating conditions. Up to 99.7 percent combustion efficiency
can be achieved.
3.5.3 Types of Flares
3.5.3.1 Steam-Assisted Flares
Steam-assisted flares are single burner tips, elevated above
ground level for safety reasons, that burn the vented gas in
essentially a diffusion flame. They reportedly account for the
majority of the flames installed in most industries today.23 To
ensure an adequate air supply and good mixing, this type of flare
system injects steam into the combustion zone to promote
turbulence for mixing and to induce air into the flame.
3.5.3.2 Air-Assisted Flares
Air-assisted flares use forced air to provide the combustion
air and the mixing required for smokeless operation. These
flares are built with a spider-shaped burner (with many small gas
orifices) located inside but near the top of a steel cylinder two
feet or more in diameter. Combustion air is provided by a fan in
the bottom of the cylinder, and the amount of combustion air can
be varied by varying the fan speed. The primary advantage air-
assisted flares provide is that they can be used in the absence
of steam.
3-17
-------
3.5.3.3 Non-Assisted Flares
The non-assisted flare is just a flare tip without any
auxiliary provision for enhancing the mixing of air into its
flame. Its use is limited essentially to gas streams that have a
low heat content and a low carbon/hydrogen ratio that burn
readily without producing smoke.24 These streams require less air
for complete combustion, have lower combustion temperatures that
minimize cracking reactions, and are more resistant to cracking.
3.5.3.4 Pressure-Assisted Flares
This type of flare use vent stream pressure to promote
mixing at the burner tip. If sufficient vent stream pressure is
available, these flares can be applied to streams previously
requiring steam or air assist for smokeless operation. Pressure-
assisted flares generally have the burner arrangement at ground
level, and consequently, must be located in a remote area of the
plant where there is plenty of space available. They have
multiple burner heads that are staged to operate based on the
quantity of gas being released. The size, design, number, and
group arrangement of the burner heads depend on the vent gas
characteristics.
3.5.3.5 Enclosed Ground Flares
The burner heads of an enclosed flare are inside a shell
that is insulated. This shell reduces noise, luminosity, and
heat radiation and provides wind protection. A high nozzle
pressure drop is usually adequate to provide the mixing necessary
for smokeless operation and air or steam assist is not required.
In this context, enclosed flares can be considered a special
class of pressure-assisted or non-assisted flares. Enclosed
flares are always at ground level.
Enclosed flares generally have less capacity than open
flares and are used to combust continuous, constant flow vent
streams, although reliable and efficient operation can be
attained over a wide range of design capacity. Stable combustion
can be obtained with lower heat content vent gases than is
possible with open flare designs, probably due to their isolation
from wind effects.25
3-18
-------
REFERENCES
1. Clark, F.M. Stage Charging on a Single Collector Main
Battery: A Total System Concept. Paper Presented at the
1973 Iron and Steelmaking Conference.
2. Coke Oven Emissions from Wet-Coal Charged By-Product Coke
Oven Batteries-Background Information for Proposed
Standards. U.S. Environmental Protection Agency, Office of
Air Quality Planning and Standards, Research Triangle Park,
NC, 27711. EPA-450/3-85-028a. April 1987. p. 4-2.
3. Barnes, T.M., H.W. Lownie, Jr., and J. Varga, Jr. Control
of Coke Oven Emissions. American Iron and Steel Institute.
December 31, 1973. p. 24-25.
4. Reference 2. p. 4-6.
5. Reference 2. p. 4-9.
6. VanOsdell, D.W., et al. Research Triangle Institute trip
report to U.S. Steel, Fairfield, Conn. October 27, 1977.
7. Oliver, J.F. and J.T. Lane. Control of Visible Emissions at
CF&I's Coke Plant - Pueblo, Colorado. Journal of the Air
Pollution Control Association. 29 (9): September 1979.pp.
920-925.
8. Lownie, Jr. , H.W. and A.O. Hoffman. A Research Approach
to Coke-Oven End-Closure Problems. Ironmaking Proceedings,
3_5, 109(1976).
9. Reference 8.
10. Telecon. Jablin, R. with Feiser, Art, A.F. Industries.
July 10, 1978.
11. Reference 10.
12. U.S. Environmental Protection Agency. National Emission
Standards for Hazardous Air Pollutants; Coke Oven Emissions
from Wet-Coal Charged By-Product Coke Oven Batteries. 53 FR
251, December 30, 1988, Washington, DC. U.S. Government
Printing Office. January 1989.
13. Reference 12.
14. Graham, J.P. and B.P. Kirk. Problems of Coke-Oven
Air-Pollution Control. The Metals Society. London.
p. 82-100.
3-19
-------
15. Proposed Interim Report on Development of Concepts for
Improving Coke-Oven Door Seals. Summary of Task 4: Pre-
Engineering Analysis Evaluations and Recommendations.
Battelle Columbus Laboratories. March 1978.
16. Lownie, H.W., et al. study of Concepts for Minimizing
Emissions from Coke-Oven Door Seals. EPA-690/2-75-064.
July 1975.
17. Reference 2. p. 4-88.
18. Draft of Standards Support and Environmental Impact
Statement Volume I: Proposed National Emission Standard By-
Product Coke Oven Wet Coal Charging and Topside Leaks.
Emission Standards and Engineering Division. U.S. EPA. p.
4-61.
19. U.S. Environmental Protection Agency. Office of Air Quality
Planning and standards. Research Triangle Park, North
Carolina. OAQPS Control Cost Manual, Fourth Edition.
January 1990. p. 7-5.
20. Reactor Processes in SOCMI — Background Information for
Proposed Standards. U.S. Environmental Protection Agency.
Office of Air Quality Planning and Standards. Research
Triangle Park, North Carolina. Preliminary Draft. EPA
Publication No. 450/3-90-016a. June 1990.
21. Letter from David Shore (Flaregas Corp., Spring Valley,
NY) to William M. Vatavuk (U.S. EPA, OAQPS, RTP, NC).
October 3, 1990.
22. Reference 20.
23. Guide for Pressure-Relieving and Depressuring Systems.
Refining Department, API Recommended Practice 521, Second
Edition. September 1982.
24. Reference 21.
25. Kalcevic, V. (IT Enviroscience). "Control Device
Evaluation Flares and the Use of Emissions as Fuels,"
Organic Chemical Manufacturing Volume 4. U.S. Environmental
Protection Agency. Research Triangle Park, North Carolina.
EPA Publication No. 450/3-80-026. December 1980. Report 4.
3-20
-------
CHAPTER 4
REGULATORY OPTIONS
4.1 INTRODUCTION
The EPA listed coke oven emissions as a hazardous air
pollutant on September 18, 1984. This listing decision was
followed by proposal of a NESHAP for the control of coke oven
emissions from wet-coal charged batteries.1 These proposed
standards were not promulgated because Congress revisited the
issue during development and passage of the CAAA of 1990. The
CAAA establish specific requirements for the development of
regulations governing coke oven emissions. Under §112(d)(8), EPA
must promulgate standards based on maximum achievable control
technology (MACT) for new and existing coke oven batteries by
December 31, 1992. In addition, §112(i)(8) instructs EPA to
promulgate standards that reflect the lowest achievable emissions
rate (LAER), as defined by §171 of the Act, by December 31, 1992.
If the LAER standard is not promulgated, the Act specifies limits
that will automatically go into effect January l, 1998. The EPA
must also promulgate work practice regulations and residual risk
standards.
EPA's option for regulation is to set MACT and LAER
standards at least as stringent as the limits outlined in the
CAAA. The first section of this chapter describes these limits.
In the development of the NESHAP for coke ovens, the EPA entered
into regulatory negotiations with industry, labor unions, states,
and environmental groups to develop a satisfactory regulation
that is more stringent than the limits outlined in the CAAA. The
second section of this chapter discusses some of the history and
the results of the regulatory negotiations.
4.2 LIMITS IN CLEAN AIR ACT
Emission reductions for coke oven batteries are to be phased
in over a twenty-seven year period. This period spans from 1993
4-1
-------
to 2020 with the implementation of MACT, LAER, and residual risk
standards.
4.2.1 Requirements for MACT. LAER. and Residual Risk standards
By definition in the CAA, MACT is the best demonstrated
control technology or practices used by the coke oven industry.
New sources that must meet MACT requirements must meet emission-
reduction standards as strict as those achieved by the best-
controlled similar sources. Existing sources must at least meet
the average emission limitation achieved by the best performing
12 percent of existing sources. For coke ovens, the CAAA require
at a minimum that batteries will not exceed 8 PLD, 1 PLL, 5 PLO,
and 16 seconds of visible emissions per charge.2
LAER, as defined by §171 of the CAAA, is the rate of
emissions for any source which reflects:3
(1) the most stringent emission reduction which is
contained in a SIP, unless the source demonstrates
such levels are unachievable, or
(2) the most stringent level of control achieved in
practice by such a category (or class) of source.
The default limitations set by LAER are to reflect at least 3
PLD, 1 PLL, 4 PLO, and 16 seconds of visible emissions per
charge.
Within six years after the date of enactment of the CAAA,
the EPA is to investigate and report to Congress the methods of
calculating any remaining risk to public health from emissions of
coke oven batteries subject to regulation. The EPA must also
report the public health significance of such remaining risks,
and the available technology and methods for the reduction of
such risks along with an estimation of the costs associated with
further reductions. In addition, an evaluation is to be
conducted on the actual health effects with respect to persons
living in the vicinity of sources, and provide recommendations as
to legislation regarding such remaining risk.4 From this report,
residual risk standards will be implemented.
4-2.2 Timing of Coke Oven Provisions in the CAAA
Technology-based emission standards for coke oven batteries
must be established by December 31, 1992. Coke oven owners and
operators have two options for meeting their Clean Air Act
obligations. The first requires coke oven batteries to meet MACT
requirements by the end of 1995 and residual risk emission
standards by 2003. Owner/operators of coke ovens seeking an
extension of the residual risk requirements must meet a special 8
PLD standard by 1993. By 1998, these coke ovens must meet the
4-2
-------
standard of LAER set by EPA. Any coke oven that does not meet
this standard in 1998 must meet the residual risk standard by
2003 or cease operations. Those coke ovens that meet the 1993
and 1998 standards will be granted the residual risk extension to
2020s.
4.3 REGULATORY NEGOTIATION
During the spring and summer of 1991, EPA met with
representatives of the industry, labor unions, states, and
environmental groups to discuss available data to be used as the
basis of the new regulations. A workshop format was used to
explore and clarify the varying viewpoints. Following these
informal discussions, EPA announced its intention to establish a
committee to negotiate a new approach for the control of coke
oven emissions and conducted formal meetings and informal
workshops over the next several months to identify and resolve
the many issues associated with the regulation.
At the final negotiating session, the Committee members
conceptually resolved all outstanding major issues and decided to
reach final agreement after reviewing and concurring on the draft
preamble and regulation describing in detail the scope,
application, and impacts.
4.3.1 Source Categories
A foundry battery is defined as a battery that is not owned
or operated by an integrated steel producer, and had an annual
capacity less than 1.25 million megagrams per year. The
Committee agreed that the standard for door leaks at foundry coke
producers should be slightly less stringent than the LAER door
leak standard for other coke oven batteries.
A new source is a stationary source for which construction
commences after the date of proposal. Any coke oven battery for
which construction is begun at a plant site (where no batteries
previously existed) after the date of proposal in the Federal
Register would be subject to the emission limitations for new
sources included in the proposed standards. This type of
construction is termed "greenfield" construction. In addition,
the construction of a new battery or the reconstruction of an
existing battery that results in an expansion in capacity of an
existing coke plant would subject such a battery to the emission
limitations for new sources. The emission limitations for new
sources are based on the emission control performance achieved by
nonrecovery coke oven batteries.
Batteries that are completely reconstructed on the same site
as an existing battery without an increase in the coke plant
capacity are called "padup rebuilds", and new batteries that
replace existing batteries without increasing plant capacity are
4-3
-------
called "brownfield" construction. Padup rebuild and brovmfield
batteries are considered existing and are subject to the LAER
limits for as long as the battery is on the extension track.
Distinctions were made for door leaks on short batteries
(batteries with ovens less than 6 m in height) and tall batteries
(batteries with ovens 6 m or more in height) with a slightly less
stringent standard for coke oven doors on tall batteries because
they are more difficult to control.
4.3.2 Emission Limitations Set bv MACT & LAER
The Committee agreed to use the recently-collected data
from self monitoring and state or local agency inspections to
assess control levels that have been achieved and to develop the
emission limits. The Committee also concluded that the rolling
30-day limit should be based on an upper confidence level of 99.7
percent. In addition, agreement was reached that the emission
limits effective in November 1993 for batteries on the risk
extension track would be converted to 30-day average limits at
the 99.7 percent confidence level. Although several options for
the level of control of coke oven emissions were considered
during the months of the regulatory negotiations, the limitations
described in Table 4-1 are the resulting levels agreed upon by
the regulatory negotiation committee.
MACT limits. For existing by-product batteries not seeking
a compliance date extension, the limits to be met by December 31,
1995 and January 1, 2003 are in the first two columns of Table
4-1.
MACT for new source limitations are based on levels achieved
by nonrecovery batteries with 0.0 PLD, 0.0 PLL, 0.0 PLO, and 34
seconds per charge.
LAER limits. The negotiated limits for batteries seeking a
compliance date extension require that leaks must follow the last
three columns of Table 4-1 with compliance dates of November 15,
1993, January 1, 1998, January 1, 2010, and then residual risk
standards in 2020.
The Act provides that at any time prior to January 1, 1998,
an owner or operator may elect to comply with residual risk
standards by the required date rather than comply with the LAER
and revised LAER standards and compliance dates. However, the
owner or operator would be legally bound to comply with the
residual risk standards as of January 1, 2003. If EPA has not
promulgated industrywide residual risk standards by that time,
the Agency must promulgate residual risk standards for those
batteries that choose to meet residual risk standards by 2003.
(Work to develop the residual risk standards has not yet begun.)
4-4
-------
TABLE 4-1. PROPOSED LIMITS FOR EXISTING
BY-PRODUCT BATTERIES
MACT LIMITS
LIDSPLL
OFFTAKES PLO
CHARGING (Jog)
s/charge
DOORS PLD
TALL
SHORT
FOUNDRY
1231/95
0.6
3.0
12
6.0
5.5
5.5
Beyond 2003
(must meet
residual risk)
0.6
3.0
12
5.5
5.0
5.0
LAER EXTENSION TRACK
11/15/93
(CAA
Limits)
0.83
4.2
12
7.0
7.0
7.0
1/1/98
0.4
2.5
12
4.3
3.8
4.3
1/1/10
0.4
2.5
12
4.0
3.3
4.0
4-5
-------
4.3.3 Process Upsets
During periods of short process upsets or when a
catastrophic failure occurs (such as an exhauster malfunction, or
electrical failure), raw coke oven gas is vented directly to the
atmosphere. Such an event can release tons of organic compounds
in a short period of time. Ignitors can be installed on the coke
oven batteries to flare the gas when it is bypassed. Combustion
in a flare destroys the organic compounds in the gas and also
converts highly-toxic hydrogen sulfide (H2S) to less toxic sulfur
dioxide (SO2) .6
Only 16 of the 82 batteries currently in this industry
control these emissions. Although information from a
manufacturer indicates flares can achieve 99.5 percent control, a
conservative estimate of 95 percent control is expected from the
installation of flares on coke oven batteries. The regulatory
negotiation committee agreed to install flares on currently
uncontrolled batteries.
4.3.4 Work Practices
The proposed work practice standards would require the owner
or operator of an existing or new coke oven battery to develop a
written plan describing emission control work practices to be
implemented for each battery. The plan, required by November 15,
1993, must include provisions for training and procedures for
controlling emissions from coke oven doors, charging operations,
topside port lids, and offtake system(s). Compliance with such
work practices is November 15, 1993.
4.4 ECONOMIC INCENTIVES: A MARKET-BASED APPROACH TO THE CONTROL
OF COKE OVEN EMISSIONS
An alternative approach to the control of coke oven
emissions could be established using economic incentive
strategies. When designed properly, these market based
approaches act to harness the marketplace to work for the
environment. Such strategies influence, rather than dictate
producer and consumer behavior, in order to achieve environmental
goals. Such environmental goals are achieved with the most
flexibility and at the least cost to society.
Several types or categories of economic incentive strategies
exist, including fees, subsidies, and emissions trading. Fee
programs establish and collect a fee on emissions, providing a
direct economic incentive for emitters to decrease emissions to
the point where the cost of abating emissions equals the fee.
Similarly, subsidy programs provide a direct incentive for
emitters to decrease emissions by providing subsidy payments for
4-6
-------
emission reductions beyond some baseline. Emissions trading
allows sources with low abatement cost alternatives to trade or
sell emission allowances to higher abatement cost alternatives so
that the cost of meeting a given total level of abatement is
minimized7.
Legal constraints imposed by Title III of the CAAA severely
limit the usefulness of economic incentive strategies for
reducing HAP emissions. Because Title III requires the
implementation of MACT, sources have little or no choice as to
the type or level of control they implement, except perhaps if
the source controls beyond the requirements set for doors, lids
and offtakes. In the development of the coke ovens NESHAP, the
regulatory negotiations acted as a means for industry to express
the desired level of control. Plus, the costs associated with
compliance with the default levels of control were based on input
by each battery operator. Thus some flexibility in control
techniques has been considered in the analysis of costs.
However, this form of providing flexibility to the industry does
not constitute an incentive to reduce pollution. Overall, the
applicability of economic incentive programs for the coke oven
NESHAP is very limited.
4-7
-------
REFERENCES
1. U.S. Federal Register. Volume 52, No. 78;40 CFR Part 61.
U.S. Environmental Protection Agency. April 23, 1987.
pg. 13586.
2. U.S. Congress. The Clean Air Act, as Amended. Title III,
Section 112(d)(8)(C). November 1990.
3. Reference 2, Section 171(3).
4. Reference 2, Section 112(f).
5. Commerce Clearinghouse, Inc. Clean Air Act - Law and
Explanation. Chicago, Illinois. Prepared for the U.S.
Environmental Protection Agency. Research Triangle Park, N.C.
pp.35-39.
6. Research Triangle Institute. Preliminary Assessment of
Bypassed Coke Oven Gas: Evaluation of Emissions and Controls.
Prepared for U.S. Environmental Protection Agency. March,
1992.
7. U.S. Environmental Protection Agency; Office of Air Quality
Planning and Standards. Draft Regulatory Impact Analysis for
the NESHAP for Source Categories: Organic Hazardous Air
Pollutants from Synthetic Organic Chemical Manufacturing
Industry and Seven Other Source Categories. Publication No.
EPA-450/3-92-009. February 1992. pp. 5-11, 5-12.
4-8
-------
CHAPTER 5
COST ANALYSIS
5.1 INTRODUCTION
This chapter summarizes the cost analysis that was performed
to estimate the cost of the NESHAP for coke oven emissions.
Additional details are provided in the appendices. Appendix A is
a battery-by-battery listing of the major cost elements, which
were identified in many cases by representatives from the coke
plant. Appendix B lists the estimated costs for each battery.
The cost estimates are based on" achieving the "default" emission
limits given in the Clean Air Act amendments.
5.1.1 Approach
The approach used for the cost analysis was developed with
the assistance of a Cost Work Group that was formed by the
Regulatory Negotiation Committee for the coke oven NESHAP.1 The
Work Group consisted of representatives from EPA, individual coke
plants, the American Iron and Steel Institute (AISI), the
American Coke and Coal Chemicals Institute (ACCCI), the United
Steelworkers of America, environmental groups, and consultants
for the industry and union.
Based on suggestions from the"Work Group, a draft approach
was written and costs were estimated for each battery using a
generic approach.2 Several categories of batteries were
identified based on the current emission control performance,
extent of repairs needed, and other factors. Cost functions were
developed for each class of battery, and each of the 82 batteries
was assigned to one of the different cost classifications. The
cost functions were then applied to each battery to estimate
costs.
A major difficulty in performing an accurate cost analysis
is that no one knows exactly what each coke oven battery must do
(in terms of additional labor, new equipment, and repairs) to
meet the upcoming coke oven regulation. There is much variation
among existing batteries in terms of battery condition, remaining
5-1
-------
useful-life, and current emission control capability.
Consequently, the generic or model battery approach may r-ssult in
an inaccurate estimate (either too high or too low) for a
specific battery. However, the approach provides a range of
costs for different types of batteries from the various cost
categories. In addition, the total nationwide cost estimate
should be reasonable if the proper distribution of battery types
is used in the generic approach, even if every battery is not
assigned to the proper cost category.
To improve the generic approach, preliminary estimates
generated by EPA were sent to the individual coke plants by the
trade associations for review and comment. After reviewing the
estimates provided for their batteries, many individual plants
provided alternative estimates of the additional expenses that
they anticipated in the next few years. Other companies
commented that the generic approach provided a reasonable
estimate for their plant, and a few companies offered no
additional comments. The comments received on the cost estimates
from individual companies were summarized and discussed in a
meeting of the Cost Work Group on March 9, 1992.3 In addition,
revisions were discussed for the generic approach that was used
to estimate costs for batteries that did not provide specific
cost data. Many specific recommendations were incorporated into
the cost estimates presented in this analysis.
The cost report was revised based on the extensive review
and comments received from the industry representatives. The
site-specific cost data were included, and revisions were made to
the generic approach. A final cost report was prepared and
distributed to members of the Cost Work Group.4 The cost report
is the primary reference for the information presented in this
chapter.
There is uncertainty as to the stringency of the regulation,
including the numerical emission limits, the format of the
standard, and how it will be enforced (because the negotiations
are still underway). Consequently, the Work Group concluded that
the cost analysis at this point should be based on the emission
limits written into the Clean Air Act Amendments of 1990 as the
least stringent that can be promulgated for maximum achievable
control technology (MACT) and lowest achievable emission rate
(LAER).
5.1.2 Issues and Assumptions
The individuals who provided battery-specific estimates of
costs cautioned that their estimates were preliminary at this
point, and that no commitment had been made to implement the
capital or other improvements that were described. For some of
5-2
-------
these estimates, there was difficulty in separating items
directly attributable to the NESHAP from those attributable to
other factors. For example, some of the planned expenditures may
be attributable to a current regulation that is already as
stringent as the anticipated NESHAP. In other cases, extensive
repairs planned for batteries that are old and in poor condition
may have been needed in the near future to continue operating,
even if the NESHAP were not in place. Another issue is the cost
of ancillary equipment that might be replaced if extensive
repairs are undertaken. A discussion is provided below for these
issues and how they are handled in this analysis.
For this analysis, two costs estimates were developed to
characterize the potential costs attributable to the NESHAP. The
first estimate is EPA's best estimate of the costs that could be
attributable to the standard, and the second estimate includes
essentially all of the items identified by the company as
potentially attributable to the standard. Major differences in
the estimates are described below:
(a) The EPA estimate does not include the costs to meet
current regulations that may be as stringent as the
NESHAP. For example, several batteries have current
regulations (State implementation plans or consent
decrees) that limit percent leaking lids and offtakes to
1 and 5 percent, respectively, based on any single
observation. This analysis assumes that the MACT limits
are 1 percent leaking lids and 5 percent leaking
offtakes based on the average of three runs. For the
industry estimate, these costs are included as provided
by the company.
(b) The EPA estimate does not include the cost of through-
wall repairs because these repairs are required for
other reasons, including the proper operation of the
battery and to meet emission limits for combustion
stacks. These costs are included in the industry
estimate.
(c) Some batteries estimated millions of dollars in costs
for extensive repairs to meet either the November 1993
limits or the MACT limits (December 1995). In addition,
some of these batteries are projected to be completely
rebuilt to meet LAER. The EPA cost estimate includes a
portion of the rebuild cost. (The portion of the
rebuild cost attributed to the standard is determined in
the economic impact analysis, which considers factors
such as the remaining useful life of the battery and the
declining demand for coke.) The industry estimate
includes both the extensive repairs followed by the
5-3
-------
rebuild in a few years.
(d) No additional costs are estimated for the Acme Steel
batteries to meet MACT for door leaks in the EPA
estimate. A total of 21 observations over 7 years shows
exemplary control for percent leaking doors. Cost
estimates provided by the company are included in the
industry estimate.
(e) One company estimated that jambs will be replaced every
8 years on one of their batteries because of their past
history and because of the NESHAP. For the industry
estimate, this cost was annualized over 8 years. For
the EPA estimate, a 20-year lifetime was used for the
annualization.
One issue discussed by the Work Group concerned whether the
NESHAP might "trigger" an early rebuild of some batteries. The
LAER standard increases the stringency of the standard for
percent leaking doors, which could require some extensive repairs
on batteries that are not in good condition. These batteries may
require extensive repairs to their entire end-closure system,
including brickwork, jambs, doors, and seals. If extensive
repairs are required for one part of the battery, the owner may
choose to rebuild the entire battery for economic reasons. The
issue is how much, if any, of the rebuild cost is attributable to
the NESHAP. The cost of a complete rebuild is several times more
costly than extensive repairs to improve door leak control. For
this analysis, the cost of rebuilding batteries is apportioned to
the NESHAP based on the estimated remaining life of the battery.
Consequently, this represents the cost of rebuilding a battery
earlier than it would otherwise have been rebuilt. The
apportionment of rebuild cost and estimates of nationwide costs
are provided in the economic impacts analysis, which also
considers the declining demand for coke and other factors.
Another issue is related to equipment for other operations
that might be upgraded as part of the rebuild. The estimated
rebuild cost that was used included upgraded or new equipment
used to operate the battery, primarily for items that might be
related to emission control. For example, the costs included
door machines, door and jamb cleaners, larry cars, pusher
machines, and door spotting devices. The rebuild cost did not
include extensive repairs to or replacement of ancillary
operations, such as the by-product recovery plant, coke and coal
handling equipment, and wastewater treatment.
Each coke battery has a choice of two compliance "tracks",
and the choice of tracks may have a significant effect on costs.
For example, a battery nearing the end of its useful life might
5-4
-------
be a candidate for the MACT route, which requires meeting MACT by
December 31, 1995, and then meeting a risk standard by 2003.
This battery would not have to meet the more stringent LAER
standard in 1998, and a decision to rebuild or shut down might be
postponed until 2003. For this analysis, only two coke batteries
were identified by the company as probable candidates for the
"MACT only" track. Other plants are still evaluating options,
including a choice to meet the requirements of both regulatory
tracks until the requirements of a risk-based standard are known.
This analysis takes a somewhat worst-case approach by assuming
that all batteries (with the 2 exceptions identified) will incur
the MACT costs by November 1993 and all will incur the LAER costs
by 1998.
Costs were not included for Bethlehem Steel's plant at
Sparrows Point, MD. This plant shut down in 1991, and the
company has not announced any plans to rebuild the batteries.
Costs also are not included for the batteries at Inland Steel.
The company has announced that Battery 11 will be shut down in
1992, and the remaining batteries will be closed before the end
of 1994. There are no announced plans to rebuild these
batteries.
5.2 MACT COSTS
The costs associated with MACT are given in Appendix B for
each battery and show that the total capital cost ranges from $66
to $100 million. The total annualized cost is estimated to range
from $25 to $33 million per year. The following sections
summarize the basic components of cost used in the estimates.
5.2.1 Lids and Offtakes
Control costs associated with the NESHAP are estimated for
batteries with current emission limits greater than 1 percent
leaking lids (PLL) and 5 percent leaking offtakes (PLO). These
current regulations are enforced as limits that are not to be
exceeded for any single observation. A total of 74 batteries
have emission limits greater than 1 PLL, and approximately 45
batteries have emission limits greater than 5 PLO.5
The estimated cost for batteries with current limits higher
than 1 PLL and 5 PLO is based on adding one additional person per
battery unit per shift to locate and seal leaking lids and
offtakes. Labor costs are estimated as $25/hour, including
benefits, or approximately $220,000 per year for each battery
unit.6'7 If improvement in control is required for only one of
the two emission points, the increased labor is estimated as 0.5
persons per shift.
5-5
-------
5.2.2 Charging
The Cost Work Group suggested that MACT for charging may
require certain single main batteries (those without jumper
pipes) to make extensive modifications to the larry car and
charging operation with a cost on the order of $600,000. A
review of available data (based primarily on a 1979 report)8
indicated that approximately 44 out of 82 batteries have a single
collecting main. However, most of these batteries already have
jumper pipes to perform stage charging properly and to meet
current State regulations for charging.
Single main batteries with the least stringent current
limits for charging were investigated more closely. A total of
20 single main batteries have current limits that are higher than
25 seconds per charge. Five of these batteries are in the
rebuild category, and the previous approach assumed that these
batteries would install upgraded or new larry cars at that time.
Reviewers pointed out that these batteries that might be rebuilt
would likely incur costs to meet the MACT charging standard
before rebuilding the battery to meet LAER limits at a later
date. Consequently, the cost analysis was revised to include
estimates for upgrading the charging system to meet MACT, even if
the battery is expected to be rebuilt before 1998.9
5.2.3 Doors
The additional cost for the control of percent leaking doors
(PLD) to meet the MACT limits is based on the cost of increased
cycling of doors through the door repair shop. This cost is
estimated as $440/oven per year for an improvement of 1 PLD.10
However, this cost function is used only to estimate the cost to
improve to an average of 5.7 PLD (which is the long-term average
associated with a 3-run limit of 8 PLD). Capital improvements
may be required for some batteries to achieve much lower levels
for PLD. Several individual plants identified specific capital
expenditures they expected to make to meet MACT. When specific
information was provided by the company on capital expenditures
for MACT, these costs were included in the analysis.
The improvement in control that would be required was
estimated two ways. For the first case, the current performance
was based on visible emissions data available for each battery.
For the second case, each battery was assumed to be meeting its
current regulation 95 percent of the time. The two approaches
yielded essentially the same estimates for nationwide costs, even
though the estimates for specific batteries were different. The
annualized cost was estimated as $3.4 million per year based on
the performance data and $3.2 million per year based on the
current regulation (i.e., assuming that batteries are meeting
5-6
-------
their current regulations 95 percent of the time).11
5.2.4 Monitoring
For visible emission monitoring, the analysis assumes that
the cost of monitoring (above that currently performed) will
include 4 hours per battery per day at $25 per hour. Many
batteries currently perform daily visible emissions monitoring
and may incur no additional monitoring cost. The total of 4
hours includes 0.5 hours to inspect for leaking doors, lids, and
offtakes; 2.5 hours to observe 5 consecutive charges; and 1 hour
of travel time for the inspector.
5.3 LAER COSTS
LAER cost estimates are given for each battery in
Appendix B. The total capital cost, excluding complete rebuilds,
is estimated to range from $150 to $240 million. The total
annualized cost ranges from $46 to $57 million per year, again
excluding the annualized cost of rebuilds. The capital cost of
rebuilds that may be triggered early was estimated as $709
million, and a portion of this cost will be attributed to the
NESHAP in the economic analysis.
The cost estimates for LAER are based on a more stringent
door standard. In addition, the control costs incurred under
MACT for lids, offtakes, and charging will be continued under
LAER. For this analysis, several categories were developed to
estimate the costs to meet LAER limits for batteries that did not
provide specific information. These categories are described
below12. Site-specific cost information was received for most
coke oven batteries from company representatives. Consequently,
the generic categories described in this section were used only
when the company representatives did not provide alternative
estimates for their specific batteries. Appendix A provides a
battery-by-battery listing of the specific cost components
supplied by the industry, and for those cases when site-specific
information was not given, the generic category assigned to the
battery is given.
Category A - Batteries with hand-luted doors; Assume that
additional control is provided for these batteries by adding an
additional person per battery unit per shift to aid in locating
and luting leaks. The estimated labor cost at $25 per hour (for
24 hours per day and 365 days per year) is $220,000/yr.
Category B - Batteries Currently Subject to a Standard of
5 PLD; Several batteries are currently subject to a standard of
5 PLD, based on any single observation and usually excluding 2
leaks. This standard has been applied to new or rebuilt
5-7
-------
batteries; consequently, these batteries are not expected- to
require major rebuilding or new doors and jambs. The cost
estimate for these batteries assumes an additional person per
battery unit per shift to locate and seal leaks (at $220,OuO/yr)
and the material cost of sodium silicate ($56,000/yr) for a total
cost of $276,000/yr.
Category C - Batteries Consistently Averaging 2 to 4 PLD;
Batteries that are currently averaging 2 to 4 PLD are expected to
require only a marginal improvement in control to meet LAER
limits. The cost estimate for these batteries assumes an
additional person per battery unit per shift to locate and seal
leaks (at $220,000/yr) and the material cost of sodium silicate
($56,000/yr) for a total cost of $276,000/yr.
Category D - Batteries Averaging 4 to 6 PLD; Assume that
these batteries will require an additional person for each
battery unit per shift to locate and seal leaks and that sodium
silicate will be used as a supplemental sealant ($276,000/yr as
described above). Also include automatic door and jamb cleaners
on the coke side of short batteries (.$700,000). Most tall
batteries have automatic cleaners. Assume that half of the tall
batteries will rebuild their existing cleaners ($400,000) and the
other half will install new cleaners ($1.4 million) on both sides
of the battery with a midrange cost of $0.9 million.
Category E - Batteries Averaging 7 to 10 PLD; Assume that
batteries in this category will require extensive repairs or a
partial rebuild to achieve the LAER limit for doors. The cost
estimate includes jamb and end flue repairs, new doors, 10
percent spare doors, new jambs, automatic door and jamb cleaners
on the coke side of short batteries and both sides of tall
batteries, and a spotting device for improved door placement.
Also assume that these batteries will require an additional
person for each battery unit per shift to locate and seal leaks,
and that sodium silicate will be used.
Commenters noted that some batteries may choose to perform
through-wall repairs when the end flue repairs are made. If
through-wall repairs are needed, they would probably be performed
to maintain the proper operation of the battery and to meet
existing,regulations for the battery's combustion stack.
However, industry commenters have argued that when repairs or
modifications are made to meet the LAER limits, plants may also
perform through-wall repairs at the same time. Consequently, the
LAER limits may trigger these repairs earlier than they normally
would be performed. For this analysis, assume that half of the
batteries will perform through-wall repairs at a cost of $12.5
million per battery (from the Struthers1 cost report)13. Capital
costs are given in Table 5-1 for the case without through-wall
5-8
-------
repairs. Through-wall repairs will be distributed across all
batteries in this category at $6.3 million each and will be
included in the industry cost estimate. The cost of through-wall
repairs will be annualized over a 10-year lifetime.
TABLE 5-1. CAPITAL COST ELEMENTS FOR CATEGORY E
Item Unit Cost for 4 m Cost for 6 m
Jamb and end flue oven $62,400 $119,000
repairs
New doors (installed)
Spare doors (10%)
New jambs (installed)
Jamb cleaner3
Door cleaner3
Spotting device
oven
oven
oven
each
each
each
24,600
2,200
23,000
450,000
250,000
50,000
28,600
2,600
27,000
550,000
350,000
50,000
Assume that new door cleaners ($250,000) and new jamb
cleaners ($450,000) will be installed on the cokeside of
short batteries (less than 6 meters). Tall batteries (6
meters) generally already have automatic cleaners. Assume
that half of the tall batteries will install new cleaners
on both the pusher side and coke side ($1.4 million), and
assume that the other half will have their existing
cleaners rebuilt ($400,000) with a midrange cost of
$900,000.
• Cost for short batteries = $750,000 + 112,000 (no. ovens)
• Cost for tall batteries = $950,000 + 177,000 (no. ovens).
Category F - Batteries Averaging over 10 PLD; Assume that
batteries averaging over 10 PLD must be rebuilt earlier than
planned because of the 1998 LAER limit. Include the cost of a
pad up rebuild, automatic door and jamb cleaners on the coke side
(both sides for tall batteries), spotting devices, and new or
refurbished larry car, pusher machine, and door machines. For
this analysis assume that half of these machines will be
refurbished and that the other half will be replaced by new
equipment. For tall batteries that already have automatic
5-9
-------
cleaners, assume that half will be refurbished and half replaced
by new cleaners. Also assume that these batteries will require
an additional person for each battery unit per shift to locate
and seal leaks with sodium silicate ($276,000/yr). Also include
$10 million for repairing or refurbishing the by-product plant.
The capital cost of a padup rebuild is given in Table 5-2.
TABLE 5-2. CAPITAL COST ELEMENTS FOR CATEGORY F - REBUILD
Item Unit Cost for 4 m Cost for 6 m
Padup rebuild
Jamb c leaner (s)a
Door cleaner (s)a
* Spotting device
Larry car
Pusher
Door machines
Repairs to by-product
plant
oven
each
each
two
each
each
two
plant
$790,000
450,000
250,000
100,000
1,150,000
2,250,000
2,200,000
10,000,000
$1,056,000
550,000
350,000
100,000
1,800,000
2,800,000
2,800,000
10,000,000
Assume that new door and jamb cleaners will be installed
on the cokeside of short batteries (less than 6 meters).
Tall batteries (6 meters) generally already have automatic
cleaners. Assume that half of the tall batteries will
install new cleaners on both the pusher side and coke
side, and assume that the other half will have their
existing cleaners rebuilt.
Assume that half of the batteries will install new door
machines, larry cars, and pusher machines. Assume that
the other half will rebuild existing equipment. Use the
midrange cost.
0 Cost for the by-product plant will be used only in the
study of capital availability.
• Cost for short batteries:
$6,400,000 + 790,000 (no. ovens)
• Cost for tall batteries:
$8,400,000 + 1,056,000 (no. ovens)
5-10
-------
5.4 PROCESS UPSETS
The venting of raw coke oven gas to the atmosphere from
process upsets seldom occurs. However, the magnitude of effects
from the emissions from a process failure could potentially be
significant due to the amount released in a short period of time.
Currently, twenty percent of existing batteries have installed
flares to control emissions from process upsets.
Although the data on the bypassing of raw coke oven gas are
limited, annual emissions and the cost of control are estimated
from data obtained from EPA Region III and from the Allegheny
County (Pennsylvania) Air Pollution Control Agency on the
frequency, duration, and mass emissions associated with these
bypass events. Data on three coke oven plants (19 batteries) in
the Pittsburgh, PA. area indicate the average battery vented for
a total of 4.1 hours per year (hrs/yr).15 The more serious
episodes of venting for 37 hours (on average) are believed to
occur infrequently. For this analysis, venting from a serious
malfunction is assumed to occur once every 10 years, which yields
an annual rate of 3.7 hrs/yr. Adding the infrequent (but long)
events to the average obtained from Allegheny County for short
episodes yields an annual venting rate of 7.8 hrs/yr per battery.
It is estimated from this that current nationwide bypass BSO
emissions are 470 Mg/yr, which results in a level that is higher
than the baseline level of emissions from charging and leaks from
doors, lids, and offtakes. When MACT and LAER limits are applied
(using the limits written into the CAA), the bypassed emissions
dwarf the emissions from the NESHAP sources.16
ChemTech consultants estimated the installed cost of a flare
system for a coke oven battery is $100,000 to $200,000 per flare,
with two flares per battery (one on each end of the battery).
The upper end of the range is for some batteries that may require
additional structural support to install the flare system or to
provide the ducting required to carry the gas to the flare. The
company suggested a midrange value of $150,000 per flare or
$300,000 per battery as a reasonable estimate.17
Operating costs are minimal. The gas used for the pilot
flame is negligible, and the only labor requirement is to steam
clean the pilot system once per week. The labor cost, based on
one hour per flare (two hours per battery) and $25 per hour, is
$2,600 per year. The life of a flare system is estimated as at
least 10 years and probably closer to 20 years. A midrange value
of 15 years is used for the cost analysis.
5-11
-------
As a result, it is estimated that for the 66 batteries
required to install flares, the total capital cost is $20
million. Total annual cost is estimated to be 2.8 million per
year.
5-12
-------
REFERENCES
1. More information on the Regulatory Negotiation
Committee for coke ovens can be found in 57 FR 1730
(January 15, 1992); 57 FR 4025 (February 3, 1992);
57 FR 5267 (February 13, 1992); and 57 FR 6830
(February 28, 1992).
2. Revised Approach for the Coke Oven NESHAP Cost
Analysis. Draft distributed for review and comment.
February 11, 1992.
3. Coke Oven Cost Work Group Teleconference. Memo from
D. Bell, EPA, to Cost Work Group. March 3, 1992.
4. Cost Analysis for the Coke Oven NESHAP. Transmitted by
Amanda Agnew, EPA, to Cost. Work Group. April 20, 1992.
34 pages.
5. Reference 4, p. 27-28.
6. Cost Implications for the Coking Industry to Meet the
Clean Air Act Amendments of 1990. Prepared by
Struthers Corporation for the American Iron and Steel
Institute and the American Coke and Coal Chemicals
Institute. November 1991. Section X of Appendices.
7. Forbes Magazine. January 8, 1990. p.40-42.
8. Technical Approach for a Coke Production Cost Model.
PEDCo Environmental, Inc. EPA Contract No. 68-02-3071.
December 1979, p. 38-50.
9. Reference 4, p. 6.
10. Reference 4, p. 7, 17.
11. Reference 4, p. 7-8.
12. Reference 4, p. 8-12.
13. Reference 6, Figure 2.
14. Reference 4, p. 12-16.
5-13
-------
15. Research Triangle Institute. Preliminary Assessment of
Bypassed Coke Oven Gas: Evaluation of Emissions and
Controls. Prepared for U.S. Environmental Protection
Agency. March, 1992. pg. 4.
16. Reference 15, pp. 1-2.
17. Reference 15, pg. 5.
5-14
-------
CHAPTER 6
ECONOMIC IMPACTS
6.1 INTRODUCTION
Analyzing the economic impacts of the control levels
specified by the CAAA requires describing the baseline conditions
in the market for coke. Then, changes in the markets for coke
that result from the regulation can be quantified, and their
impacts estimated. To accomplish this, we first profile the
market for coke at baseline (in. the absence of the emission
standard). Then, we use a mathematical simulation model to
estimate the changes in those markets resulting from the costs of
complying with the emission standard.
As mentioned previously, coke is used as an input both in
traditional steel-making and in the production of iron castings
in the cupolas of foundries. Because the physical properties
required of coke used to produce steel differ from those required
of coke used to produce iron castings, the market for coke
actually comprises two distinctly separate markets: the market
for furnace coke, used to produce steel, and the market for
foundry coke, used to produce iron castings. Thus, conditions in
the steel industry determine the demand for furnace coke, while
conditions in the iron castings industry determine the demand for
foundry coke.
6.2 THE DOMESTIC DEMAND FOR FURNACE COKE, DERIVED FROM THE
DEMAND FOR STEEL
Because the demand for furnace coke depends on the quantity
of steel produced, recent developments in the steel industry have
had a profound impact on the demand for furnace coke. Overall
U.S. steel output has declined in the past 20 years, and
technical changes in steel manufacturing will result in a
decreased quantity of furnace coke required per ton of steel
produced. The following sections describe trends in the quantity
of raw steel produced in the U.S. and trends in the types of
steel-making technologies employed.
6-1
-------
6.2.1
Quantity of Raw Steel Produced
Trends for U.S. steel production and consumption for the
period 1979-1989 are presented in Figure 6-1. Although in some
years, consumption and production increased from the previous
year's level, the general trend is downward. Table 6-1 shows
trends in steel production, consumption, and trade. The decline
of the last 20 years reflects in part the reduction in steel
intensity typical of developed economies, which are less involved
in infrastructure construction. The change in production and
consumption was also due to a fundamental change in the pattern
of steel consumption in the U.S. For example, steel use in
automobile production has fallen because of substitution of other
materials, especially plastics and aluminum, for steel, and
because of the general reduction in size of American cars. Also
over the past 20 years, supply-side factors have resulted in U.S.
steel producers having production costs that are high relative to
those of their foreign competitors. The forecast for the
domestic steel industry predicts a slight decline in production.
Total coke-using steel production is predicted to be 52.8 million
Mg in 1995 and 52.7 million Mg "in 2000.1 Because furnace coke
consumption depends on coke-using steel production, furnace coke
consumption will likely also decline in the future.
Millions of Mg
160 -r
1970
1975
1980
1985
••- PRODUCTION
CONSUMPTION ••• TREND
Source: American Iron and Steel Institute (AISI).
Reports. 1973, 1974, 1978, 1983, 1984, 1989.
Annual Statistical
Figure 6-1. Total steel Production and Consumption in the U.S., 1970-1989.
6-2
-------
6.2.2 Trends in Steel-Making Technology
Historically, vertical integration ensured that steel mills
had a reliable supply of all the raw materials necessary to their
production. Until the 1930s, the traditional method of steel-
making, which starts with iron ore, was the only method. The
steel-making firms, therefore, depended on adequate supplies of
iron ore, coal for coking, and limestone for fluxing to guarantee
adequate supplies of pig iron or "hot metal," without which their
steel-making could not go forward. Changes in the mix of
technologies used to produce steel, particularly the increasing
use of electric air furnaces (EAFs), have reduced the need for
such integration. Figure 6-2 shows the changing shares of total
steel production over the last 20 years for the three main steel-
making technologies: open hearth furnaces (OHF), basic oxygen
furnaces (EOF), and EAFs.
The OHF was the dominant furnace type for nearly a century,
but it has now virtually disappeared, replaced in integrated
steel manufacturing by the BOF. Pig iron is melted and refined
in these furnaces and transformed into steel. The molten"steel
is formed into ingots, which are in turn shaped in the primary
mill into blooms, billets, or slabs, or put through a continuous
casting process that forms them directly into these semi-finished
forms. Finally, in the rolling mills, the steel is finished and
turned into sheets, wire, rods, or structural forms, for example.
Much of the recent investment in the integrated steel sector has
been the introduction of continuous casting facilities into the
plants.
Another steel-making technology, the EAF, uses scrap steel
and electricity as inputs and does not require coke, iron ore, or
limestone; therefore the steel-maker does not depend on these raw
material supplies. Although EAFs have been in existence since
the 1930s, rapid improvements in design during the 1960s
increased their profitability. EAFs do not require the massive
investment that integrated traditional plants do and have,
therefore, given rise to "minimills," which have much smaller
average capacity than traditional steel mills. EAFs have
accounted for a growing share of total steel production (see
Figure 6-2). The quantity of domestic raw steel produced using
this technology increased steadily from 1970 to 1989, during a
period when total domestic raw steel production was falling.
Thus, their share of domestic raw steel production has increased.
In 1970, EAFs accounted for 15.3 percent of domestic raw steel
production. EAFs produced 27.9 percent of domestic raw steel in
1980 and 35.9 percent in 1989.
Because of continuing technological developments, the steel-
making technologies that do not require coke are expected to
continue to increase their share of production. New EAF
technologies are being developed that can produce a wider variety
of steels at lower costs than conventional EAFs, and several new
6-3
-------
TABLE 6-1. QUANTITY OP STEEL PRODUCED DOMESTICALLY, IMPORTS, AND
EXPORTS OF STEEL (103 Mg)
Year
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
Quantity of
Steel
Produced
Domestically
119,558
109,494
121,128
137,090
132,473
106,038
116,364
113,939
124,574
123,946
101,668
109,844
67,797
76,924
84,115
80,235
74,186
81,046
90,840
89,040
Quantity of
Steel
Imported
12,149
16,640
16,074
13,773
14,518
10,920
12,986
17,552
19,214
15,925
14,086
18,089
15,148
15,518
23,785
22,051
* 18,811
18,558
18,992
15,746
Quantity of
Steel
Exported
6,412
2,570
2,612
3,684
5,303
2,685
2,413
1,821
2,202
2,562
3,728
2,640
1,675
1,090
891
.847
845
1,026
1,881
4,162
Apparent
Supply of
Steel
125,295
123,564
134,590
147,179
141,688
114,273
126,937
329,670
141,585
137,309
112,026
125,293
81,270
91,352
107,009
101,439
92,152
98,578
107,951
100,624
Source: AISI. Annual Statistical Report.
1984,1989.
1973, 1974, 1978, 1983,
Percent
100
'$.',, 5 '• .
, Basic Oxygen v;
"
tr-a—cr~»-a
•«••«__
90 •'-
80 ••
70 ..
60 ...
50 .3
40 ••
30 ••
20 4-
10
1970 1975 1980 1985
Source: AISI. "Annual Statistical Reports. 1973-1989.
Figure 6-2. Share of Steel Production by Technology
Open Hearth
1990
6-4
-------
EAFs are under construction. In addition, the AISI, along with
the Department of Energy, is developing a direct steel-making
process that involves pre-reduction of iron ore pellets. The
Japanese are developing a similar process.2 These processes will
eliminate the intermediate step of using a blast furnace to
produce iron and thus will not require coke.
The industry is also investigating technical changes that
reduce the quantity of coke used per ton of iron produced in
blast furnaces, so that even when blast furnaces are still used,
furnace coke consumption is reduced. Such technical changes
include direct injection of pulverized or granulated coal into
blast furnaces.3
Figure 6-3 depicts the effects of these trends on the
consumption of furnace coke. As the domestic demand for furnace
coke decreases from D, to D2, the quantity produced, Q, falls to
Q2. At the same time, the price per megagraro of coke falls from
P, to P2.
S/Mg Coke
Mg of Coke/t
Figure 6-3. Effects of Decreased Steel Production and Changing Steel-Making and
Iron-Making Technologies on the Market for Furnace Coke
In addition to domestic demand for furnace coke, another
source of demand for domestically produced furnace coke is demand
from outside the U.S. Coke exports are small relative to
domestic consumption. As shown in Table 6-2, export consumption
of furnace coke declined over the period from 1970 to 1989,
averaging 1.3 million Mg in the 1970s and 0.98 million Mg during
the 1980s. In 1989, furnace coke exports were 0.99 million Mg.*
6-5
-------
6.2.3 Domestic consumption of Furnace Coke
Domestic consumption of furnace coke is approximated by the
"apparent consumption" of furnace coke. Apparent consumption is
computed by subtracting exports of furnace coke from domestic
production, then adding imports of furnace coke. This
calculation yields the amount of furnace coke produced or
imported during a given year available for use in the U.S. (It
may differ from actual use because some of the coke produced may
go into inventory, while some of the coke consumed may come out
of inventory.) Historical data on furnace coke production are
not available; instead, data on coke production by integrated
("captive") coke producers are used. A relatively small quantity
of furnace coke (less than 3 million Mg in 1992) is produced by
merchant producers—independent businesses that specialize in
coke production. Merchant coke producers sell their coke rather
than using it on site. For this analysis, data on the apparent
consumption of furnace coke are computed using captive coke
production rather than total furnace coke production as the
starting point. Thus, our estimate of apparent furnace coke
consumption, shown in Table 6-2, underestimates total apparent
furnace coke consumption by the amount of merchant furnace coke
production.
6.2.4 The Demand for Foundry Coke Derived from the Demand for
Iron Castings
Approximately 90 percent of all manufactured goods and
almost all industrial machines currently produced in the U.S.
contain some type of cast metal.3 Castings with very different
physical properties can be produced by varying the composition of
the melted metal, varying the temperature to which the metal is
heated, and varying the cooling time for the castings produced.6
Three types of furnaces are commonly used to melt the metals
used to produce iron castings: cupolas, EAFs, and induction
furnaces. Of these three furnace types, only cupolas, which
resemble miniature blast furnaces, require using foundry coke.
Foundry coke is used in the initial stage of casting production
to melt scrap iron and steel into a liquid that can be poured
into prepared molds and allowed to cool and harden.
According to Robert Eppich, Vice President of Technology of
the American Foundrymen's Society, only gray iron, ductile iron,
and malleable iron castings ever require coke as an input.
Because malleable iron castings make up less than 4 percent of
the total tonnage of these types of castings produced each year
and most of the molten metal used for malleable iron castings is
melted in induction furnaces, demand for foundry coke is
essentially a function of the demand for gray and ductile iron
castings. Approximately 60 percent of the gray and ductile iron
castings produced nationwide are produced with metals melted in
6-6
-------
cupolas using foundry coke as an input.7 Although this analysis
assumes a constant ratio of foundry coke production to gray and
TABLE 6-2. FURNACE COKE PRODUCTION IN CAPTIVE BATTERIES, EXPORTS,
IMPORTS, AND APPARENT CONSUMPTION (10JMg)
Year
1970*
1971'
1972'
1973'
1974'
1975*
1976'
1977k
1978b
1979b
X 7 / 7
1 QQftC
iyou
1981°
1982C
19.83C
1984C
1985C
1986C
1987C
1988C
1989C
1990
Captive Coke
Production
54,308
46,452
49,298
52,932
50,573
47,071
48,514
44,899
40,755
43,630
38,090
35,366
23,067
20,505
24,355
22,886
20,228
22,260
25,942
26,550
22,916
Exports
2,285
1,372
1,120
1,268
1,162
1,157
1,195
1,127
630
1,309
1,883
1,064
903
605
950
1,020
913
522
994
986
520
Imports
139
158
168
980
3,218
1,654
1,192
1,663
5,202
3,613
599
479
109
32
529
525
299
838
2,444
2,101
695
Apparent
Consumption
52,162
45,238
48,346
52,644
52,629
47,568
48,511
45,435
45,327
45,934
36,806
34,781
22,273
19,932
23,934
22,391
19,614
22,576
27,392
27,665
23,091
'The quantities of coke production, exports, imports, and consumption are
calculated from the values reported in U.S. Department of the Interior,
Bureau of Mines. Mineral Industry Surveys. Washington, DC. 1970-1976.
"The quantities of coke production, exports, imports, and consumption are
calculated from the values reported in U.S. Department of Energy, Energy
Information Administration. Coke and Coal Chemicals Report. Washington,
DC. 1977-1980.
"The quantities of coke production, exports, imports, and consumption are
calculated from the values reported in U.S. Department of Energy, Energy
Information Administration. Quarterly Coal Report. Washington, DC.
1982-1990.
ductile iron castings production, some observers believe that
competitive forces and stringent environmental regulations are
encouraging foundries to increase their use of electric induction
ovens, which are believed to be more flexible, more energy
efficient, and cleaner than cupolas.*
Demand for iron castings depends on demand for the finished
products that contain them. This condition is particularly true
of castings sold to the automotive, machinery manufacturing, and
transportation industries. Demand in these industries is greatly
affected by fluctuations in industrial activity, business and
6-7
-------
consumer spending, and the strength of the dollar.9 Increased
substitution of alternative materials for iron castings also
influences total demand for castings. For example, in the market
for automobiles, an 8 percent reduction in iron and steel
castings per vehicle was predicted to occur between 1989 and
1992.10
The quantity of foundry coke consumed each year in the U.S.
is small relative to the quantity of furnace coke consumed. (In
1990, U.S. foundry coke consumption was less than 4 percent of
total domestic furnace coke consumption.) Foundry coke is a
specialized of breakage during transportation and other quality
concerns, such as the higher sulphur content of foreign-made
coke, importing foundry coke is not practical." Although some
foundry coke may be exported from the U.S. to Canada and Mexico,12
foundry coke exports cannot be reliably distinguished from
furnace coke exports in the records kept by the U.S. Department
of Commerce. For this reason this analysis assumed all
internationally traded coke is furnace coke.
Unlike the integrated steel firms producing furnace coke,
merchant coke producers sell their coke rather than using it on
site. They are much smaller businesses than the integrated steel
producers that produce the majority of furnace coke. The average
battery capacity of foundry coke producers is only 152 Mg per
year, much less than the average furnace coke battery capacity of
378 Mg per year.
Foundries producing ferrous castings in the U.S. suffered
serious cutbacks in production during the early 1980s as a result
of the combined effects of a slumping U.S. economy, fierce
international competition, and higher costs of production brought
on by stricter environmental regulations.13 Total production of
ferrous castings in 1986 fell to about half the level recorded in
1978. Almost half of the foundries, representing over a third of
U.S. production capacity operating in 1980, closed down during
that period.
6.3 COKE PRODUCTION
Regulation of coke oven emissions required in the Clean Air
Act Amendments will affect the costs of coke production. Both
fuel, so substitution opportunities are severely limited.
Because captive coke and merchant coke producers will be affected
by the candidate NESHAP regulating coke oven emissions. To
analyze the impacts of the NESHAP, the following sections
identify and characterize coke producers, analyze coke production
trends, and evaluate the variable inputs to the production
process.
6-8
-------
6.3.1 The Coke Production Process
A small positive back-pressure maintained on the oven
prevents air from leaking into the ovens during coking. During
coking, raw coke oven gas is removed through an offtake system
composed of standpipes and other piping to a central collecting
main. By-products such as coal tar, ammonia, and light oil
(benzene, toluene, xylene) are removed from the coke oven gas in
the by-product recovery plant. The cleaned coke oven gas is used
to underfire the coke ovens, and excess gas is used in other
parts of the plant. At the end of the coking cycle, doors on
each end of the oven are removed, and the coke is pushed from the
oven into a special railroad car (quench car). The incandescent
coke is carried to a quench tower where it is deluged with water
for cooling.14
Air pollutants may be emitted from several sources during
this process. First, if the oven doors warp or fail to seal
properly, they may leak. Second, if the lids over the charging
ports do not seal properly, emissions may occur there. Finally,
the offtake system that recovers the coke oven gas may leak. The
candidate NESHAP specifies the maximum allowable percentages of
each of these sources at a plant permitted to release emissions.
6.3.2 Profile of Coke Producers
As noted in Section 6.2, both the producers and consumers of
coke are part of SIC code 3312, Blast Furnaces and Steel Mills.
Almost all of the furnace coke produced in integrated steel
plants is used on site. Blast furnace operators typically prefer
coke that is sized between 0.75 and 3 inches. Smaller coke
fragments, called breeze, are used either as fuel in steel mill
boiler houses to generate steam or to assist in ore agglomeration
as a fuel in the sintering process.15 Captive furnace coke
plants, part of vertically integrated steel mills, produce most
of the domestically produced coke. Merchant coke producers
produce a small percentage of furnace coke; however, they produce
all of the foundry coke produced in the U.S.
6.3.2.1 Captive Coke Producers. Almost 90 percent of
domestically produced blast furnace coke is produced by captive
furnace coke plants that are part of vertically integrated steel
mills. These steel mills include most, if not all, of the
following:
• plants that produce coke,
• plants that process iron ore,
• plants that combine the coke and iron ore to produce pig
iron,
• plants that use the pig iron to produce raw steel, and
6-9
-------
• plants that use the raw steel to produce semi-finished
and finished shapes.
6.3.2.2 Merchant Coke Producers About 11.5 percent of
domestic blast furnace production capacity is in the merchant
sector. Merchant coke producers are smaller, independent
companies that rely solely on the sale of coke and coke by-
products to generate revenue. Merchant furnace coke producers
sell their coke to the large steel companies and sell the by-
products of coking to coal chemical refineries. One merchant
furnace producer, Jewel Coke and Coal, uses a nonrecovery coking
process and relies entirely on coke sales for survival. This
NESHAP applies only to by-product coke facilities; therefore, the
non-recovery production process used by Jewel Coke and Coal is
not affected.
All of the foundry coke produced in the U.S. is made by
merchant producers. Foundries require coke that is sized at 3
inches or greater. Only about 70 to 75 percent of the coke
produced by merchant foundry producers meets this requirement and
is sold to iron foundries. The remaining 25 to 30 percent of
coke produced by foundry producers is sized between 0.75 and 3
inches and is sold at reduced prices to producers of sugar,
mineral wool, and other non-metallurgic commodities.16 Merchant
coke producers have no use for coke breeze.
Although much of it never reaches a foundry cupola, we refer
to all coke products by merchant foundry producers as foundry
coke. The two columns labeled "Foundry Coke" and "Other Coke" in
Table 6-3 should therefore be added together to represent what we
refer to as foundry coke production.
Table 6-3 lists the plants in operation in 1992, their
location, batteries, coke type, and production capacity. These
data reveal systematic differences between captive and merchant
coke plants. Merchant coke plants tend to be smaller than
captive coke plants; they tend to have a smaller number of
batteries per plant; and the individual batteries tend to be
smaller than those of captive coke producers. For example, three
of the 11 merchant coke plants have only one battery, and the
largest number of merchant batteries at a single plant is five.
Merchant coke plants have an average of 2.45 batteries per plant,
with an average annual production capacity of 169,000 Mg per
battery. Captive coke plants, on the other hand, average 3.1
batteries per plant, with an average production capacity of
340,000 Mg per year. The largest captive producer in 1991 was
the U.S. Steel (USX) furnace coke production site in Clairton,
Pennsylvania, which has 12 batteries and a capacity of
approximately 4,030,000 Mg of coke per year.
6-10
-------
TABLE 6-3.
COKE PRODUCERS IN THE UNITED STATES
Plant Naae
MERCHANT PRODUCERS
ABC
Citizens Gas
Empire
Erie
Jewell Coke
and Coal*
Koppers
New Boston
Shenango
Sloss
Toledo
Tonawanda
TOTAL MERCHANT CAPACITY
CAPTIVE PRODUCERS
Acme Steel
Armco
Armco
Bethlehem
Steel
Bethlehem Steel
Bethlehem Steel
Geneva Steel
Address
Tarrant , AL
Indianapolis, IN
Holt, AL
Erie, PA
Vasant, VA
-
Woodward , AL
Portsmouth, OH
Pittsburgh, PA
Birmingham, AL
Toledo, OH
Buffalo, NY
Chicago, IL
Middleton, OH
Ashland, KY
Bethlehem, PA
Burns Harbon, IN
Lackawana , NY
Provo, UT
Blast
Furnace
Battery
Capacity
(103 Kg)
245
221
153
111
153
161
128
252
55
341
301
151
111
111
1,953
226
226
466
466
316
305
528
200
200
591
809
664
340
340
223
120
211
188
Foundry
Battery
Capacity
(10s Kg)
317
63
73
83
75
82
70
35
58
89
97
119
127
188
1,476
Other
Battery
Capacity
(103 Kg)
128
25
29
34
31
28
14
19
29
48
52
62
499
(continued)
6-11
-------
TABLE 6-3. COKE PRODUCERS IN THE UNITED STATES (CONTINUED)
Plant Ni
ui« Address
Blast
Furnace Foundry
Battery Battery
Capacity Capacity
<103 Mg) <103 Kg)
Other
Battery
Capacity
(10J Mg)
CAPTIVE PRODUCERS (continued)
Gulf States
Steel
Inland Steel
LTV
LTV
LTV
LTV
National Steel
National Steel
Sharon Steel
U.S. Steel
U.S. Steel
Wheeling-
Pittsburgh
Total Captive
Total Capacity
Gadsden, AL
E. Chicago, IN
Pittsburgh, PA
S. Chicago, IL
Warren, OH
Cleveland, OH
Granite City, IL
Detroit, MI
Monessen, PA
Clairton, PA
Gary, IN
Steubenville, WV
Capacity
268
268
131
223
224
281
482
256
256
289
289
382
558
447
248
248
250
250
795
212
105
760
260
260
260
260
260
260
270
270
270
450
450
700
680
250
250
151
151
163
782
20,038
21,991 1,476
499
•All of Jewell Coke and Coal's batteries use a non-recovery coking process.
These batteries will not be affected by the candidate NESHAP.
6-12
-------
6.3.3 Historical Coke Production Trends
As already mentioned, the number of coke plants, the number
of batteries, and the total amount of coke produced have
decreased over the last decade. In 1980, 60 coke plant locations
had a total of 195 batteries. This total included 20 merchant
coke plant locations with 47 batteries and 40 furnace coke plant
locations with 148 batteries. Total production for these
batteries in 1980 was approximately 48 million Mg of coke.
During the period from 1980 to February of 1992, the number
of plants fell to 30, and the number of operating batteries
present at coke plants declined by more than 50 percent, to 86
batteries. During this period, at least one new coke battery was
constructed and several batteries were modified. Total coke
production, shown in Table 6-4, fell from about 48 million Mg in
1979 to about 26.3 million Mg in 1990. Production generally
declined from 1980 to 1986, then rebounded slightly between 1986
and 1989. At its lowest level, in 1986, coke production fell to
23.2 million Mg, less than 50 percent of 1980 coke production.
Coke production gradually increased from 1986 to 1989 to about 30
million Mg, approximately 62 percent of 1980 coke production,
before falling once more in 1990 to 26.3 million Mg per year. As
shown in the list of coke producers, the most recent information
indicates that the total capacity of batteries present in
February 1992 is 23.9 million Mg.
6-13
-------
TABLE 6-4. HISTORY OF U.S. COKE PRODUCTION (103 Mg)
Year
1970'
1971*
1972'
1973*
1974'
1975"
1976'
1977b
1978b
1979b
u. s * y
1980b
1981°
1982C
1983C
1984C
1985C
1986C
1987C
1988C
1989"
1990C
Merchant
Plant
5,377
5,061
5,115
4,792
4,642
4,290
3,966
3,336
3,097
4,500
3,848
3,531
2,492
2,957
3,427
3,160
2,990
3,228
3,516
3,464
3,400
Captive
Plant
54,308
46,452
49,298
52,932
50,573
47,071
48,514
44,899
40,755
43,630
38,090
35,366
23,067
20,505
24,355
22,886
20,228
22,260
25,942
26,550
22,916
Total
Production
59,685
51,513
54,413
57,724
55,215
51,361
52,480
48,235
43,852
48,130
41,938
38,897
25,559
23,462
27,782
26,046
23,218
25,488
29,458
30,014
26,316
The quantities of coke production are calculated from the values reported
in U.S. Department of the Interior, Bureau of Mines. Mineral Industry
Surveys. Washington, DC. 1970-1976.
"The quantities of coke production are calculated from the values reported
in U.S. Department of Energy, Energy Information Administration. Coke and
Coal Chemicals Report. Washington, DC. 1977-1980.
The quantities of coke production are calculated from the values reported
in U.S. Department of Energy, Energy Information Administration. Quarterly
Coal Report. Washington, DC. 1982-1990.
6.3.4 Inputs in Coke Production
Production of coke requires only a few variable inputs:
high- and low-volatile coals, energy, and labor. These are
combined with capital equipment (the coke oven batteries) whose
quantity is fixed in the short run. The marginal cost of
producing a megagram of coke, therefore, is the marginal cost of
the coal, energy, and labor required to produce it. The baseline
marginal cost of the ovens themselves is zero, because they are
already in place. This analysis concentrates on two inputs, coal
and labor.
6.3.4.1 Coal. Two principal markets exist for U.S. coal:
• the market for boiler fuel—coal that is burned for its
energy content, and
6-14
-------
• the market for "metallurgical coal"—coal that is
pyrolyzed and reduced to coke for use by the iron and
steel industry.
This analysis evaluates the market for metallurgical coal.
The quantity of metallurgical coal traded has declined as the
quantity of coke produced and consumed has declined. Domestic
consumption of metallurgical coal is now half of the 1977 level
and currently accounts for only 5 percent of total U.S. coal
consumption. 7 The average price of coal receipts at coke plants
in 1989 was $52.57 per Mg.18
6.3.4.2 Labor. Unfortunately, employment data for the
by-product coke sector are not available. Because coke-making is
only a part of the Blast Furnace and Steel Mill industry,
employment in the by-product coke industry is only a portion of
the total industry employment. However, Harry Kokkinis, an
industry analyst with Locker Associates, and Monty Stuart,
Supervisor of Environmental Affairs and Technical Programs for
Bethlehem Steel, provided estimates of the number of workers
employed in coke production on a per-battery basis within the
steel industry.19-2" Because the batteries for which the estimates
were made were of different sizes, with larger batteries
requiring fewer man-hours per unit of output, our estimate of 1.2
man-hours per Mg of coke produced, which is a weighted average of
the two estimates, might overstate employment levels in larger
facilities and understate them for smaller batteries. Using this
ratio of 1.2 man-hours of labor per megagram of coke production
and assuming that the industry average of 41 hours of labor per
worker per week applies to workers engaged in coke production, we
estimate total employment in captive furnace coke production in
1992 in the U.S. to be 11,278 workers.
Our estimate of employment in coke production for the
merchant sector relies on 1990 survey data gathered by the
American Coke and Coal Chemicals Institute (ACCCI), a trade
association of merchant coke producers.21 ACCCI represents
merchant furnace coke producers as well as all U.S. foundry coke
producers. According to the information supplied by ACCCI, the
27 coke batteries at 11 merchant plants producing coke employed
2,530 production workers. This figure corresponds to 1.586 man-
hours per Mg of coke produced in the merchant sector.
6.4 BASELINE CONDITIONS IN THE MARKETS FOR FURNACE AND FOUNDRY
COKE
The producers and consumers of coke interact to define the
baseline conditions in the markets for furnace and foundry coke.
Together, they determine the price and quantity of coke consumed.
The baseline conditions form the setting in which the impacts are
projected resulting from the emissions standards.
6-15
-------
6.4.1 The Markets for Furnace and Foundry coke
Both furnace and foundry coke are intermediate goods. That
is, they are produced to be used as inputs into the production of
other goods. Furnace coke is used to produce iron in blast
furnaces, which in turn is used to make steel. Foundry coke is
used to melt iron for castings. Thus, the demand for the two
types of coke is derived from the demand for the goods they are
used to produce.
On the supply side of the market for furnace coke are
domestic suppliers and imports of furnace coke. As discussed
previously, two very different types of U.S. companies produce
furnace coke: integrated steel producers and merchant coke
producers. Foundry coke is supplied exclusively by merchant
producers. Because importing foundry coke is not practical,
domestic merchant producers are the only suppliers of foundry
coke.
As described earlier, most furnace coke is produced by the
same facilities that use it to produce iron. These producers
are, in general, vertically integrated steel mills that combine
coke production, iron ore processing, iron production, and steel
production and finishing. Historical data on the quantity of
furnace coke produced by such captive coke plants are available
from the Department of Energy's Quarterly Coal Report. The total
furnace coke production estimate published in the Quarterly Coal
Report actually represents integrated firms' furnace coke
production and does not include furnace coke production by
merchant coke producers. The estimate of total furnace coke
production in 1990 is, therefore, the sum of furnace coke
produced by coke batteries within integrated steel companies and
the estimated quantity of furnace coke produced by merchant
furnace coke producers. This estimate of total furnace coke
production is 24.4 million Mg.
Although much of the furnace coke produced in the U.S. is
consumed by different branches of the same large corporations
that produce it and is, therefore, not for sale, the National
Energy Information Center within the Department of Energy has
years 1949 through 1985 from a wide range of previously published
sources. The Producer Price Index for furnace coke (SIC code
3212/11111) was used to update this time series, producing an
estimate of $90.44 per Mg for the equilibrium price of furnace
coke in 1990.
Foundry coke, unlike most furnace coke, is produced for sale
by independently owned and operated companies for which coke
production represents the primary revenue source. The quantity
of foundry coke produced in 1990 is defined for this analysis as
the quantity of coke produced by merchant coke producers
specializing in foundry coke production in 1990. A quantity of
1.92 M Mg is estimated by subtracting the estimated amount of
6-16
-------
merchant foundry coke production in 1990 from the total merchant
coke production.
The equilibrium published price of foundry coke in 1990 was
estimated by comparing recent and historical price information
obtained from several foundry coke producers and traders22'27 with
published price indices from the Department of Commerce. The
prices for foundry coke produced by these facilities ranged from
$160.50 to $161.00 per Mg for 1990. This analysis selects
$160.50 as the 1990 published price of foundry coke. This price
represents an official asking price; the actual prices realized
may be $15.00 to $20.00 lower than producers' published prices.28
6.4.2 The Baseline: Projected Conditions in the Markets for
Coke
The emissions standards evaluated here will take effect
during the period 1990 to 2000. To assess their impacts, several
projections of yearly furnace coke consumption and yearly foundry
coke production were developed to represent estimated baseline
conditions. Each of these baseline scenarios reflects different
assumptions regarding future production levels of coke-using
steel and gray and ductile iron castings, in the absence of the
regulation.
The year 1990 is the latest year for which historical data
on coke and steel production are available. To model the baseline
market conditions for coke production that follow, these data
were combined with projections of coke-using steel production,
furnace coke consumption, and foundry coke production. In
addition, adjustments were made to reflect changes in coke
production capacity that have occurred since 1990.
Table 6-3, discussed previously, shows batteries present at
plants producing coke in February 1992 and identifies, for each,
the type of coke produced and its production capacity. We
updated this information for this analysis and verified the
capacities with coke experts and the facilities themselves.
Sixty-nine furnace coke batteries at 24 plants had a production
capacity of 21.9 million Mg of coke per year, while 14 foundry
coke batteries at 8 plants had an annual production capacity of
1.5 million Mg of foundry coke and 0.5 million Mg of other coke,
for a total annual coke production capacity of 23.9 million Mg.
(One battery at Citizens' Coke in Indianapolis, Indiana, is
counted twice in these figures, because it produces both furnace
and foundry coke.)
Baseline aggregate data are available for 1990, but analysis
of regulatory impacts should include only those plants and
batteries still operational in February 1992. Several coke
batteries were shut down during 1990 and 1991. For the analysis,
the capacity of batteries present in 1992 was summed and compared
with estimated coke production. For the furnace sector, capacity
6-17
-------
is less than estimated coke production for 1990.* Therefore
furnace coke production is set equal to the summed battery
capacities, with the assumption that imports of coke will
increase sufficiently so that estimated coke consumption will
remain unchanged. Foundry coke production in 1990 was only
slightly higher than the 1992 summed capacity of batteries
producing foundry coke. Therefore to estimate each foundry
battery's output, each battery is assumed to produce enough so
that their summed production just equals estimated foundry
production for 1990.
6.4.2.1 1993 and 1998 Furnace and Foundry Coke Production
Projections as Baseline
In the furnace sector, the relationship between historical
captive coke consumption within the steel industry and coke-using
steel production was estimated, similarly, using time-series
data and regression analysis we estimated a statistically
significant relationship between historical merchant coke
production and coke-using gray and ductile iron castings
production. Then, the coefficients determined by our regression
analysis of these historical relationships, along with
projections made by others of future steel and castings
production levels, combine to estimate future captive and
merchant coke production levels. To estimate future furnace and
foundry coke production levels based on these projections of
future captive and merchant coke production, we subtracted the
estimated production of the three merchant furnace coke producers
from the coke production projected for merchant coke production
and added it to the projected coke production for the captive
sector.
6.4.2.2.1 Projected Furnace Coke Consumption. The baseline
scenario projects furnace coke consumption based on a forecast of
coke-using steel production, using a relationship estimated using
regression analysis. Using time-series data for the years 1970
to 1989, coke consumption was estimated to be a linear function
of the production of coke-using steel and the ratio of coke
consumption to coke-using steel. The values of these variables
for the years 1970 to 1989 appear in Table 6-5.
* For simplicity, the Citizens' Coke battery that produces both
furnace and foundry coke is assumed to produce only furnace coke.
Data on capacity from the facility indicate that this battery has
a capacity of 245 Mg of furnace coke and 82 Mg of foundry coke.
Foundry coke takes approximately 1.5 times as long per cycle to
produce as furnace coke. Thus, we estimate that the battery's
furnace coke capacity is 245 + (1.5)*82 = 368 Mg of furnace coke.
6-18
-------
TABLE 6-5. DATA USED IN FORECASTING FURNACE COKE CONSUMPTION
Year
Coke Consumption
(OOP Mg)
Coke-Using Steel
Production
(OOP Mg)
Ratio of Captive Furnace
Coke Production to Coke-
Using Steel Production
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
52,162
45,238
48,346
52,644
52,629
47,568
48,511
45,435
45,327
45,934
36,806
34,781
22,273
19,932
23,934
22,391
19,614
22,576
27,392
27,665
101,229
90,456
99,564
111,855
106,410
85,420
93,989
88,592
95,267
93,104
73,335
78,803
46,745
52,728
. 55,598
53,012
46,559
50,147
57,344
57,082
0.568
0.556
0.537
0.513
0.538
0.607
0.558
0.551
0.508
0.542
0.554
0.486
0.530
0.434
0.492
0.482
0.486
0.515
0.539
0.545
The regression was based on the equation
where
COKECONS = j80+ j8,STEEL + /82RATIO
COKECONS = the quantity of furnace coke consumed in
captive plants
STEEL = the quantity of coke-using (EOF and OH)
steel produced
RATIO = the ratio of coke consumption to coke-
using steel production
/30 represents the intercept term; /81 and /32 are the
coefficients for the explanatory variables. Correcting for
serial correlation was necessary. Table 6-6 displays the
regression results. All of the explanatory variables are
significantly different from zero and have the expected signs.
Furthermore, the account for nearly all of the variation in coke
production.
Captive coke consumption is estimated to be 27.0 million Mg
in 1993 and 26.3 million Mg in 1998 (see Table 6-7). This
forecast is consistent with the historical trend.
6-19
-------
TABLE 6-6. REGRESSION ANALYSIS OP FURNACE COKE CONSUMPTION IN
CAPTIVE FACILITIES
Variable Coefficient T-Statistic
Intercept
STEEL
RATIO
-35,205
0.4937
65,078
-16.539*
54.232'
17.294
Adjusted R2 0.9951
•Denotes significance at the 1 percent level.
TABLE 6-7. PROJECTED PRODUCTION OF COKE-USING STEEL AND FURNACE
COKE CONSUMPTION; 1990 THROUGH 2010
Projected
Production of Projected Total Furnace
Coke-Using Steel Coke Consumption
Year (103 Mg) (103 Mg)
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
55,635
54,918
54,202
53,487
52,773
52,751
52,730
52,708
52,686
52,665
52,337
52,090
51,805
51,521
51,239
50,959
50,680
50,403
50,127
49,853
27,730
27,376
27,022
26,669
26,317
26,306
26,295
26,284
26,274
26,263
26,121
25,979
25,839
25,699
25,559
25,421
25,283
25,146
25,010
24,875
6-20
-------
6.4.2.2.2 Projected Foundry Coke Production. The model
projects foundry coke production as a function of gray and
ductile iron castings production. The model assumes that foundry
coke is not exported or imported, and that foundry coke added to
inventory exactly equals coke consumed out of inventory.
Approximately 60 percent of all gray and ductile iron
castings are made with iron melted in cupolas, for which foundry
coke is an important input. We estimated the baseline quantities
of foundry coke for the years 1991 through 2000 using regression
analysis. Using time-series data for the years 1970 to 1989,
merchant coke production was estimated to be a linear function of
time, the production of gray and ductile iron castings, and the
ratio of coke consumption to coke-using castings. The values of
these variables for the years 1970 to 1989 appear in Table 6-8.
TABLE 6-8: DATA USED IN FORECASTING MERCHANT COKE PRODUCTION
Year
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
Merchant
Coke
Production
<10J Mg)
5,377
5,061
5,115
4,792
4,642
4,290
3,966
3,336
3,097
4,500
3,848
3,531
2,492
2,957
3,427
3,160
2,990
3,228
3,516
3,464
Total Gray and
Ductile Iron
Castings
Shipment
(10* Mg)
12.67
13.29
13.91
15.50
14.24
11.28
12.88
13.74
14.65
14.00
10.73
10.73
7.46
8.41
9.56
9.11
7.57
7.83
8.41
6.81
Time
Trend
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Ratio of Merchant Coke
Production to Coke-
Using Castings
Shipments
(%)
385.72
346.17
334.31
281.06
296.42
345.69
279.89
220.78
192.12
292.21
326.10
299.24
303.53
319.68
325.76
315.37
358.94
374.91
380.11
462.48
6-21
-------
The regression was based on the equation
MERCPROD = j30+ |8,CASTINGS + 02TIME + &RATIO
where
MERCPROD - the quantity of merchant coke produced
CASTINGS = the quantity of coke-using (gray and ductile)
iron castings produced
TIME = a linear time trend
RATIO = the ratio of merchant coke consumption to
coke-using castings production
£„ represents the intercept term; /?,, /32, and #3 are the
coefficients for the explanatory variables. We used the ordinary
least squares technique; correcting for serial correlation was
not necessary. We present the regression results in Table 6-9.
Each of the explanatory variables is significant and has the
expected sign. As was the case in the model of the furnace
sector, the explanatory variables account for nearly all of the
variation in coke production.
TABLE 6-9. REGRESSION ANALYSIS OF MERCHANT COKE PRODUCTION
Variable Coefficient T-Statistic
Intercept
CASTINGS
TIME
RATIO
Adjusted R2
Durbin-Watson
F-statistic
-2,326.15
246,370
-42.225
11.158
0.9656
1.606
178.911
-3.621'
8.861'
-3.410'
13.610'
'Denotes significance at the 1 percent level.
The analysis examines two very different projections of
future gray and ductile iron castings production. One of these
projections, obtained from the Department of Commerce, forecasts
a reversal of the historical downward trend in castings shipments
during the period between 1990 and 2000. This forecast, which
predicts a 76 percent increase in gray and ductile iron castings
shipments over recorded 1989 levels by the year 2000, is based on
the belief that iron castings exports will increase markedly over
the next decade because of a persistently weak dollar and free-
trade agreements with Canada and Mexico. The other forecast,
6-22
-------
prepared for the years 1991 to 2001 by Stratecasts, projects a 30
percent increase in shipments of U.S. gray and ductile iron
castings over the recorded 1989 level by the year 2000.b
After consulting industry experts (many of whom consider
even the Stratecasts forecast overly optimistic), the more
moderate Stratecasts forecast of castings production was selected
to develop our baseline scenario for the foundry sector and
extended to the year 2010 using the average annual projected rate
of change in castings production. The ratio of merchant coke
consumption to coke-using iron castings production was assumed to
remain constant at the 1989 level. Forecasts of the explanatory
variables for the years 1990 through 2010, combined with the
regression coefficients, allow a prediction of merchant coke
production for those years. Table 6-10 shows the forecast values
for coke-using castings production and total merchant coke
production over the period 1990 through 2010.
As indicated earlier, the projection of foundry coke
production was derived by subtracting the estimated production of
the three merchant furnace coke producers from the projected
total merchant coke production." Historical data are available
only for merchant coke production.
6.4.3 Baseline Costa of Coke Production, 1990
Using the estimated 1990 baseline unit production cost (or
average total cost [ATC]) for each battery present in 1991 as
presented in Chapter 5, each site's batteries were grouped in
order of increasing average variable cost per megagram of coke
produced. This organization reflects the assumption that if a
plant were going to reduce production, output would be reduced at
the highest unit-cost battery first, then the next highest, and
so forth. We computed total plant production costs and
calculated the marginal costs (MCs) by computing the additional
cost per megagram of coke produced resulting from the operation
of each battery. Each battery's baseline "supply price" was then
determined.
Using baseline supply price for each battery in each market,
baseline supply functions for the furnace and foundry coke
markets were then constructed by ranking the batteries in each
industry in order of increasing supply price. These supply
functions were statistically smoothed using econometric methods.
b This forecast was provided by Citizen Gas and Coke Utility,
a client of Stratecasts, Inc. It is used here with permission from
Kenneth Kirgin, president of Stratecasts, Inc.
6-23
-------
TABLE 6-10.
PROJECTED PRODUCTION OP CASTINGS AND FOUNDRY COKE:
1990 THROUGH 2020
Year
Projected Gray
and Ductile Iron
Castings
Shipments
(103 Kg)
Projected
Production of
Merchange Coke
(10J Kg)
Projected Production
of Foundry Coke
(103 Mg)
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
7,530.00
8,160.91
8,808.18
9,512.73
9,161.82
8,422.73
8,832.73
9,552.73
9,160.91
8,920.91
8,348.18
8,452.33
8,560.25
8,672.05
8,787.85
8,907.77
9,031.93
9,160.46
9,293.49
9,431.15
3,946
4,075
4,208
4,357
4,219
3,977
4,046
4,198
4,050
3,943
3,745
3,731
3,718
3,706
3,696
3,686
3,677
3,670
3,664
3,659
1,452
1,581
1,714
1,863
1,725
1,482
1,552
1,704
1,556
1,449
1,251
1,237
1,224
1,212
1,202
1,192
1,183
1,176
1,170
1,165
6.5 COSTS OF THE REGULATION
6.5.1 Components of Compliance Costs
The development of compliance costs for both the MACT and
LAER control levels are discussed in Chapter 5. As with the cost
estimates this analysis assumes the level of control specified in
the 1990 Amendments. With the cooperation of the facilities
providing current performance and current emissions limits, EPA
defined the actions that would be required at each battery to
achieve the specified limits and estimated corresponding battery-
specific MACT compliance costs.
The activities required to meet the LAER standard range from
using additional labor to locate and seal leaks to rebuilding
batteries from the pad up. The net cost of making the capital
investments associated with these rebuilds earlier than would
otherwise have been necessary is attributed to the regulation.
Therefore, battery life expectancy is an integral part of
computing the compliance costs for batteries estimated to require
rebuilding. We describe our formula for estimating battery life
expectancy later in this section.
6-24
-------
6.5.1.1 Expected Rebuild Pate.The first step in computing
this element of the battery's compliance cost is estimating when
each battery may be expected to be rebuilt in the absence of the
regulation.
Coke batteries do not have an infinite life span. The
expected rebuild date depends on the age of the battery, the
maintenance history of the battery, its initial design, and many
other factors. These factors are difficult to quantify;
therefore, the expected rebuild date is modeled as a function of
battery age only.
Engineers typically estimate the expected life of a battery
to be between 20 and 40 years. In the absence of LAER, all
batteries presently operating would have been rebuilt either
before or after 1998. Properly assessing the additional costs of
the proposed regulation for each battery requires estimating the
expected rebuild date of existing batteries. Then, the net cost
of rebuilding the battery earlier than would otherwise have been
required is attributed to the regulation. In addition, the costs
of rebuilding should not be attributed to the LAER regulation for
any batteries that are expected to rebuild before 1998 for
reasons unrelated to the regulation.
The expected rebuild date for an existing battery is based
on the assumption that battery life expectancy is normally
distributed across the population. Comments received from
industry and from Mr. Bindu Madhava at Davy/Still Otto, a battery
building firm, indicate that 6-meter batteries built before 1980
had structural problems that resulted in their having a shorter
lifespan than other batteries.30 Based on these comments, we
split the distribution of batteries in existence in 1992 into two
groups: 6-meter batteries built before 1980 and all others.
Statistical properties of a normal distribution allow for
estimating the expected rebuild date if the mean and standard
deviation are known. The mean and standard deviation of current
battery age distributions, and engineering estimates of battery
life,31 suggest a mean age at which batteries would be rebuilt of
18 years for the old 6-meter batteries and 35 years for all other
batteries. The standard deviations of the distributions of
battery life expectancy were assumed to be the same as the
standard deviations of the distributions of present battery age:
4 years for the old 6-meter batteries and 13 years for all other
batteries.
After determining the assumed characteristics of the
distributions of battery life expectancy, we computed the
expected rebuild date of each existing battery. As with human
life expectancy, the life expectancy of an existing battery is
not the mean of the distribution of battery life expectancy.
Rather, it depends on the characteristics of the distribution and
6-25
-------
on the current age of the existing battery. The older a battery
currently is, the older it is expected to be when it is rebuilt.
Thus, the expected rebuild date for each battery was
estimated statistically* using a technique described by Johnson
and Kotz.32 The estimation is conditional on the present age of
the battery. For example, if an analysis required predicting the
life expectancy of a brand new battery, the mean life expectancy
provides the most reasonable estimation because no additional
information is available. However, if an analysis required
predicting the life expectancy for a battery that is presently 28
years old, an age greater than the mean of the distribution would
be predicted, because that battery has already survived for many
years. Thus, the expected rebuild date is conditional on the
battery's current age.
Figure 6-4 highlights this example. A normal distribution
with a mean of 35 and a standard deviation of 13 illustrates the
distribution of life expectancy of most coke batteries. The "x"
represents a battery that is presently 28 years old. Any
estimate of the life expectancy of that battery would be bounded
by its present age. The bold line in Figure 6-4 indicates the
portion of the normal curve to be evaluated. Thus, the expected
lifespan of a battery is based on a truncated normal
distribution. For this particular battery, the expected lifespan
is 41 years. It would then be expected to be rebuilt in 13
years, or in 2004.
After estimating each battery's expected rebuild date, we
compared it with 1998 to determine the number of years in which
that battery would have to be rebuilt prematurely as a result of
the LAER standard. The cost of LAER assigned to the regulation
is only the net cost of rebuilding early.
The computation is as follows. The expected value of X in a
single-sided truncated normal distribution is
*(£•)
E[X] = m + I ° '
•» »
where
m « the mean of the distribution
A * the lower truncation point (i.e., current age of battery)
0 = the standard deviation of the distribution
X = the probability density function
6-26
-------
Frequency
M = 35 S - 13 Life Expectancy
Figure 6-4. Example of Truncated Normal Distribution
6.5.2 Calculating Compliance Costa
Using the data on capital and operating costs combined with
costs associated with early rebuilds allowed calculation of
compliance costs. For the MACT scenario, the capital portion of
the cost was annualized using a 10 percent discount rate. The
number of years that these costs are discounted depends on the
type of cost. The equipment associated with using sodium
silicate was annualized over 10 years; the new doors and jambs
and the automatic cleaners were annualized over 20 years. We
computed the annualized value using a capital recovery factor,
which spreads out the capital payment over that time span. Any
annual operating costs were then added to the annualized value of
the capital cost.
Computing the compliance costs associated with LAER was more
complex because the cost of rebuilding that would have been
incurred in the absence of the regulation was subtracted from the
full cost of the regulation. This task was accomplished by
computing the present value of the payment stream in the year in
which the battery would have been rebuilt. Then the present
value of that sum was computed for 1998 and subtracted from the
capital cost of the regulation that would be incurred in 1998.
An example helps illustrate this computation. Suppose that
a particular battery is 15 years old in year 0, and its life
expectancy is 35 years. Further suppose that the capital cost of
rebuilding is $25 million. Without the regulation, this battery
would have been rebuilt in year 20. Therefore, the annualized
value of the capital costs over the 15 years between year 20 and
year 35 would have been incurred without the regulation.
6-27
-------
Annualizing the rebuild cost of $25 million over 35 years at 10
percent yields an annual value of $2.4 million. The present
value of the stream of payments of $2.4 million for years 20 to
35 is computed using a present value interest factor for an
annuity (PVIFA).* This value amounts to $18.25 million and
represents the value in year 20 of that stream of annualized
costs. The present value of $18.25 million in year 0 (1998) is
$2.7 million, which is computed using a present value interest
factor (PVIF) .b The $2.7 million represents the present value in
year 0 of the cost that would have been incurred without the
regulation. The $2.7 million is subtracted from the $25 million
to compute the net cost of rebuilding the battery in 1998. The
result is $22.3 million, which represents the capital cost of the
regulation for this battery.
After computing the net capital cost of rebuild for LAER,
the net rebuild cost, annualized over 35 years at 10 percent, was
added to this amount to the other annualized components of LAER,
plus recurrent elements of the MACT costs, to compute the total
annual cost of LAER.
After computing total annual compliance costs for each
battery under MACT and LAER, we computed unit compliance costs
for each battery by dividing the total annual compliance costs
for the battery by the battery's estimated output in 1991. Each
battery's unit cost of compliance was added to its baseline
marginal cost of coke production to compute the battery's
marginal cost of coke production, or supply price, with the
regulation in effect.
The following section describes the results of our analysis
based on our primary scenario, described as follows:
The formula for computing the PVIFA is
1 -
(1
where
r is the interest rate, and
N is the number of years.
b The formula for computing the PVIF is
(1 + r)-"
where
r is the interest rate
N is the number of years
6-28
-------
• Furnace coke consumption will vary over the period 1991
to 2010 as a result of variations in the production of
steel.
• Foundry coke production and consumption will vary over
the period 1991 to 2010 as a result of changes in the
production of gray and ductile iron castings.
• Batteries' expected rebuild date depends on their current
age and type of battery.
• Sufficient resources will be available to permit
facilities to make whatever changes are required to
achieve the emissions control levels.
• Compliance costs of MACT and LAER reflect EPA's cost
estimates.
6.5.3 Availability of Resources
One of the assumptions of our analysis is that the companies
will be able to obtain whatever resources are required to enable
them to meet the control levels. Most of the costs associated
with meeting the minimum MACT limits are operating costs; we do
not anticipate a resource availability problem for most
facilities meeting MACT. However, this may not be the case for
alterations that facilities may choose to implement to meet LAER.
The resources that may provide constraints to companies' repairs
and rebuilds are:
• bricks for rebuilds,
• expert labor for repairs and rebuilds, and
• funding to finance capital expenditures.
We estimate that ten batteries operating in 1992 will be
rebuilt before 1998. Information from engineering firms that
specialize in coke-oven construction indicates that bricks for
rebuilding a given battery are specialized, take 6 to 9 months to
make, and must be imported. The one American brick manufacturer
makes bricks only for repairs at present. Mr. Bindu Madhava of
Davy/Still Otto estimated that the maximum number of batteries
that can be rebuilt worldwide annually is five, and that a
maximum of five batteries in the U.S. could be rebuilt before
1998.33 Mr. Jim Eriser of ICF-Kaiser Engineering agreed that a
maximum of five U.S. batteries could be rebuilt by 1998 and added
that the total length of time for a rebuild, from design through
"first push," is 2 to 3 years.34
In addition to the possible shortage of brick, the
availability of engineering experts and specialized brick-layers
may present problems. Several firms specialize in battery
6-29
-------
rebuilds, but they prefer to work on only one or two batteries at
a time.35
Finally, capital availability is questionable: will
companies be able to obtain financing to make the capital
expenditures they feel will be necessary to achieve the control
levels? EPA has addressed this question in a separate study.
Preliminary results from that analysis indicate that the larger,
integrated producers will be able to obtain capital funding. Of
the merchant coke producers, only Indianapolis Coke (Citizens'
Gas and Coke Utility) provided data for a capital availability
study. Their data indicated that capital would be available for
their repairs. Data for the smaller, independent merchant coke
producers are not publicly available, so no analysis was done for
them. Therefore, we do not know whether capital availability
will be a problem for them. For the companies for which data are
available, however, the EPA draft analysis indicated that capital
would be available to make needed repairs.36
6.6 ECONOMIC IMPACTS
Imposing emission controls- on coke ovens will increase the
cost of producing coke. This regulation, in turn, will affect
the markets for factors of production used to produce coke and
for products that use coke as a factor of production. In
addition, the markets for the by-products of the coke-making
process will be affected. This section discusses qualitatively
the effects anticipated as a result of the emissions controls,
describes the model used to estimate the quantitative impacts of
the emissions controls, and finally discusses the impacts
projected by the model.
6.6.1 Qualitative Analysis of Expected Impacts
Economic theory allows us to predict the qualitative changes
that will result in the market for coke and other affected
markets as a result of the new control emissions, which will
increase the cost of producing coke. The additional costs of
production include:
• the cost of additional labor required to monitor
and repair leaks,
• the cost of any materials employed in repairing
the leaks, and
• the annualized cost of any capital equipment that must be
installed or replaced to comply with the emission
standard.
Each regulated facility's average total cost of production, or
cost per megagram of coke produced,-will increase. This
6-30
-------
increase, in turn, will raise the supply price each facility will
accept for the coke produced by its batteries. As shown in
Figure 6-5, this increase is reflected in an upward shift in the
market supply curves in the furnace coke and foundry coke
markets. Because the industry supply curve shifts upward, the
equilibrium market price for coke also increases, and the
equilibrium market quantity decreases relative to the situation
without the emissions standard.
The changes in the market for coke resulting from the
regulation give rise to changes in related markets. The
increased price of coke, for example, increases the cost of
producing steel, which shifts upward the supply curve for steel
produced using coke. Because the supply curve for this type of
steel has increased, the equilibrium price of coke-using steel
will increase, and the equilibrium quantity of steel will
decrease. The demand for EAF steel will increase, as some steel
users substitute EAF steel for coke-using steel, which is now
relatively more expensive. In addition, the increased cost of
domestically produced steel is expected to result in increased
imports of steel.
Similarly, because the quantity of coke produced is expected
to decrease as a result of the regulation, demand for the coal
and labor used to produce it will decrease. This result is
reflected in a downward shift in the demand curves for these
inputs. Because the demand for the inputs has decreased, their
equilibrium prices and quantities are expected to fall as a
result of the regulation. Because coke-making represents a small
share of the total markets for coal and labor, large changes in
the prices of these inputs are not expected. In the model used
to simulate these market responses, the prices of coal and labor
are assumed to be unaffected by the regulation.
6.6.2 The Economic Impacts Estimation Model
The impacts of the NESHAP were estimated using a model that
simulates its effects on the markets for labor, coal, coke
(including imports and exports of coke), and steel (including
imports and exports of steel). In this model, separate sectors
represent furnace coke and foundry coke production. In the
furnace sector, production relationships are specified that
relate furnace coke consumption to steel production. Other
expressions describe the relationships between domestically
produced steel and coke and imports and exports of those
commodities. Additional expressions describe the relationship
between steel produced using coke and steel produced using an
EAF, which are assumed to be close but not perfect substitutes
for one another. Figure 6-6 depicts the markets simulated by the
furnace sector model and shows the commodities for which changed
prices and quantities are estimated. The foundry sector of the
model is simpler, containing expressions for quantities of coal
6-31
-------
$/Mg Coke
Unit Compliance Cost
Q2 Qi
Mg of Coke/t
Si
S2
shows supply of coke without regulation. PI and Qi
show market equilibrium price and quantity of coke
without the regulation.
shows supply of coke with regulation. 82 is vertically
greater than Si by the amount of the unit compliance
cost for each battery. ?2 and Q2 show equilibrium
price and quantity for coke with the regulation (a
single control option shown).
Figure 6-5. Effect of Control Option* on Market for Coke
6-32
-------
Coal-
Labor
Domestic
Furnace Coke
Production
Coke
Exports
Domestically
Produced Coke
Blast Furnace
Steel Production
.(BOF, OH)
Imported
Coke
Blast Furnace
Steel
Domestic
Coke-Using
Steel
Production
Domestically
Produced Coke-Using
Steel
Electric Arc
Furnace Steel
U.S. Steel
Consumption
Steel
Imports
v
Steel
Exports
Figure 6-6. Marketa and Commodities in the Furnace Sector of the Coke Ovens
NESHAP Economic Impacts Model
6-33
-------
and labor used to produce foundry coke, the demand for foundry
coke, and the supply of foundry coke.
The analysis of the standards given in the 1990 Amendments
to the Clean Air Act used this model, together with projected
baseline conditions in the affected industries, to quantitatively
evaluate the relative impacts of the estimated compliance costs
resulting from the emissions controls.
We provide the parameter estimates required for the impact
estimation model in Table 6-11. Some of these parameters were
adopted directly from an earlier study performed to analyze the
effects of the emissions controls contained in the 1987 NESHAP
proposal. These parameters include elasticities of demand, which
are parameter estimates resulting from regression analyses in an
econometric analysis of the steel industry.37 Conversations with
foundry producers suggest that the elasticity of demand for
TABLE 6-11. PARAMETER ESTIMATES FOR THE FURNACE AND FOUNDRY COKE IMPACTS
MODEL, 1993 BASELINE
Parameter Value
Demand Elasticities
Furnace coke imports 9.93
Steel exports -1.69
Steel imports 1.51
Steel consumption -1.86
Foundry coke exports 0.00
Foundry coke consumption -1.03
Coke steel, price of EAF steel 0.6
EAF steel, price of coke steel 1.0
Supply Elasticities
Furnace coke 0.77
Foundry coke 2.03
Other steel inputs 1.00
EAF steel 1.20
Substitution Elasticities
Furnace coke, other steel inputs 0.10
Cost Shares
Furnace coke in coke-using steel 8.41%
Other steel inputs in coke-using steel 91.59%
Market Shares
Furnace coke production/furnace coke consumption 79.69%
Furnace coke imports/furnace coke consumption 22.23%
Furnace coke exports/furnace coke consumption 1.92%
Coke-using steel in steel production 61.71%
EAF steel in steel production 38.29%
Steel production/steel consumption 87.03%
Steel imports/steel consumption 17.48%
Steel exports/steel consumption 4.50%
Foundry coke product ion/foundry consumption 100.00%
foundry coke estimated for the 1987 analysis may be too low.
This elasticity measures the percentage increase (decrease) in
the quantity of foundry coke demanded in response to a 1 percent
6-34
-------
decrease (increase) in the price of foundry coke. A higher
elasticity estimate would mean that foundry producers are less
able to pass cost increases along to customers; a given increase
in cost would result in a larger decrease in quantity of coke
produced and a smaller increase in its price. Attempts to re-
estimate the elasticity econometrically failed to produce
statistically significant results, so the old elasticity estimate
was used.
Other parameters such as cost shares, output shares, or
market shares were re-calculated based on the 1989 conditions in
the affected markets. Still other parameters, such as the cross-
price elasticities of demand for coke-using steel and EAF steel
and the elasticity of substitution in production of coke-using
steel between coke and other steel inputs, were assumed, based on
expectations of these relationships.
Estimating the elasticities of supply in the two coke
markets requires first constructing supply curves in those
markets by sorting the batteries by increasing supply price,
reflecting the assumption that they would be taken out of
production, if at all, in order'of "highest cost first." This
sorting process enabled construction of industry supply curves
for furnace and foundry coke. These supply curves relate the
supply price of coke to the total quantity of coke that would be
supplied in the market at each price. Finally, log-linear
regressions computed the elasticities of supply in the furnace
coke and foundry coke markets. The coefficient b in the
following equation is the elasticity of supply:
ln(Q coke) = a + b • In (P coke) + u;
where
Q coke = the cumulative quantity of coke produced,
P coke = the supply price of coke including
compliance costs,
Uj = an error term reflecting the unexplained
variation in the quantity of coke produced.
The model uses these parameters and the percentage increase
in cost resulting from the controls to calculate percentage
changes in the prices and quantities of the commodities shown in
Figure 6-6 that result from each control level. The direct
result of implementing the controls is to increase the supply
price at which a given quantity of coke will be offered by each
battery. The increase in the supply price resulting from the
regulation, estimated by computing the average unit compliance
cost as a percentage of the average baseline supply price,
provides the exogenous shock that motivates the model.
6-35
-------
6.6.3 Implementing the Impacts Estimation Procedure
The model described above, combined with our projected
baseline, described above, determines the estimated impacts of
the MACT and LAER standards. In general, facilities under
analysis are expected to choose to meet the minimum MACT
standards in 1993 and to implement LAER controls in 1998. Thus,
a projected 1993 baseline was constructed against which the
impacts of the MACT controls can be measured.
6.6.3.1 Estimating the Impacts of MACT in 1993. First,
1993 coke production in each of the sectors was estimated. We
projected total furnace coke consumption as a function of coke-
using steel production. Projected furnace coke consumption in
1993 exceeds furnace coke capacity currently present. For the
furnace sector, therefore, all batteries still present in 1993
are assumed to produce at capacity. (One facility informed EPA
that a furnace coke battery would be closed during 1992.
Consequently, this battery was omitted from the analysis.) We
projected total merchant coke production as a function of gray
and ductile iron castings production, then projected total
foundry coke production by subtracting the 1992 quantity of
furnace coke produced by merchant facilities. Each foundry
battery's coke production was projected by multiplying its
capacity by a scaling factor so that the sum of all the foundry
batteries' output exactly equaled projected foundry coke
production. The scaling factor used was the ratio of total
projected foundry coke production in 1993 to the summed
capacities of foundry coke batteries projected to be present in
1993.
Then, assuming that production relationships remained
unchanged, coal and labor use in 1993 was estimated based on 1993
projected coke production. Furnace coke imports were adjusted to
make up the difference between coke production and coke
consumption plus coke exports. We assumed furnace coke exports
were constant. Because this analysis assumed that the process of
producing coke-using steel remains unchanged, the quantity of
other steel inputs used to produce coke-using steel (iron ore was
used as a proxy for all other inputs used) was adjusted to
maintain its 1990 relationship to coke-using steel production.
Coke-using steel production was changed to reflect the projected
quantity in 1993. EAF steel production was assumed to remain
constant at 1990 levels, as were imports and exports of steel and
all baseline prices.
Each battery's unit cost (cost per megagram) of complying
with MACT controls was computed by dividing the battery's
annualized cost of compliance by the battery's estimated output.
Adding the battery unit cost of compliance to the baseline supply
price yielded the battery's supply price of coke with the
controls in place.
6-36
-------
The model computes the changes in certain prices and
quantities, relative to the baseline conditions in those markets.
These relative changes result from an exogenous change in one of
the interrelated markets. In this case, compliance with the
emissions controls will increase the cost of supplying coke,
resulting in an upward shift in the supply curves for both
furnace and foundry coke (see Figure 6-5). The model uses this
upward shift from the compliance cost analysis to quantify the
exogenous shock to which the affected markets respond.
The supply-shift parameter that measures the percentage
increase in coke production costs resulting from the standard was
calculated by dividing the average cost of compliance by the
average baseline cost. These parameters were entered into the
impacts estimation model, and the model simulated the response of
affected markets to this exogenous shock. The market interaction
of supply and demand in the coke markets results in higher prices
for both furnace and foundry coke and a smaller quantity of coke
produced. The quantity of coke-using steel produced falls,
because its production is now more costly, and its price rises.
The quantity of EAF steel produced increases, because some steel
users substitute EAF steel for coke-using steel. Imports of both
coke and steel are expected to increase, because they are now
relatively less expensive than their domestically produced
counterparts. Finally, the markets for the inputs used to
produce coke and for the steel whose production uses coke will
also be affected, and the model predicts decreases in their
quantities as well.
6.6.3.2 Estimating the Impacts of LAER in 1998.
Estimating the impacts of the LAER controls assumed to be adopted
by all coke producers in 1998 first involves estimating the
baseline conditions in 1998. The projected baseline quantities
of furnace coke consumed, foundry coke, and coke-using steel
produced in 1998 were used to determine the 1998 baseline in the
absence of the regulation. EPA has been informed that four
batteries will close in 1994; these batteries were included in
the production estimate for the 1993 analysis but are deleted for
the 1998 analysis.
As was the case for the 1993 baseline, the summed capacities
of the furnace batteries expected to be present in 1998 are less
than projected furnace coke consumption; therefore, all furnace
coke batteries present in 1998 are assumed to produce at
capacity. Their summed capacity determines estimated furnace
coke production at baseline. Furnace coke imports are projected
to make up the difference between furnace coke consumption and
furnace coke production plus exports. Estimated foundry coke
production exceeds projected foundry coke production in 1998.
Therefore, we scaled each foundry battery's production again so
that the summed outputs of all batteries exactly equal total
projected production.
6-37
-------
Then, the LAER unit compliance costs were added to the
baseline supply price for the batteries present in each sector,
the data were sorted by this resulting supply price with
compliance cost, and elasticities of supply and exogenous shift
parameters were computed for each cost scenario. (See
Section 6.6.2 for a more detailed description of this process.)
The resulting parameters, entered into the impacts
estimation model, result in the same types of decreases in
quantities and increases in prices as occurred under MACT. But,
unlike MACT, several batteries are estimated to close as a result
of the LAER standard.
6.6.4 Results of the Market Impact Estimation Model
As described above, the model uses a shift parameter that
measures the percentage increase in the marginal supply price of
coke in each market resulting from complying with the emissions
controls. This supply-shift parameter provides an exogenous
shock to the interrelated markets for coke, coal, labor, and
steel. The model uses a system of simultaneous equations to
estimate the resulting percentage changes in the equilibrium
prices and quantities in those markets.
6.6.4.1 impacts of MACT Standard, 1993. Table 6-12
depicts the baseline conditions projected for 1993. These
conditions reflect a moderate decline in steel and coke
production during the period 1989-1993. The economic impacts
assessment model produces impact measures that are described in
terms of percentage changes in the prices and quantities traded
in the affected markets. These percentage changes can be
combined with baseline information about these markets to predict
the absolute changes in prices and quantities and the levels of
prices and quantities that would result with the control levels
in place. Finally, the estimated changes in quantity of coke
produced can be used to determine whether any batteries may be
shut down as a result of the regulation.
6.6.4.1.1 Percentage Changes in Quantities and Prices*
Table 6-13 shows the percentage changes in the prices and
quantities in affected markets resulting from the MACT control
level. The changes in the equilibrium prices and quantities are
expected to be small. Furnace coke production is predicted to
decline by 0.66 percent, while the price of coke increases by
0.22 percent. Furnace coke imports are projected to increase by
nearly 2.21 percent, so that overall furnace coke consumption
decreases by only 0.03 percent. The production of coke-using
steel is estimated to fall by 0.01 percent and its price to
increase by 0.01 percent. Overall, domestic steel production and
consumption are projected to fall by at most 0.01 percent.
In the foundry sector, changes are also expected to be
small. Foundry coke production is expected to fall by 1.08
6-38
-------
TABLE 6-12. BASELINE PRICES AND QUANTITIES IN AFFECTED MARKETS, 1993
Price* Quantity"
Furnace Coke Sector
Coal
Labor
Furnace coke produced
Furnace coke imported
Furnace coke exported
Furnace coke consumed
Other steel inputs (iron ore)
Domestic coke-using steel produced
Electric arc furnace steel produced
Total domestic steel produced
Steel imported
Steel exported
Steel consumed
54.08
16.21
90.88
90.88
90.88
90.88
32.78
538.76
538.76
538.76
538.76
538.76
538.76
28,194
12,121
21,535
6,007
520
27,022
73,067
54,202
33,628
87,830
17,637
4,546
100,921
Foundry Coke Sector
Coal
Labor
Foundry coke produced
Foundry coke consumed
51.79
16.21
160.50
160.50
4,048
1,261
1,714
1,714
'Units are 1991 $/Mg except for labor. Labor's price is in 1991 $/hour.
''Units are 103 Mg except for labor. Labor is measured in jobs.
TABLE 6-13. PERCENTAGE CHANGES IN PRICES AND QUANTITIES IN AFFECTED MARKETS
UNDER MACT MINIMUM CONTROL LEVEL, 1993
Price
Quantity
Furnace Coke Sector
Coal
Labor
Furnace coke produced
Furnace coke imported
Furnace coke exported
Furnace coke consumed
Other steel inputs (iron ore)
Domestic coke-using steel produced
Electric arc furnace steel produced
Total domestic steel produced
Steel imported
Steel exported
Steel consumed
Foundry Coke Sector
Coal
Labor
Foundry coke produced
Foundry coke consumed
0.00
0.00
0.22
-0.01
0.01
0.00
0.01
0.00
0.00
1.05
-0.66
0.66
0.66
2.21
0.00
-0.03
0.01
0.01
0.00
0.01
0.01
0.01
0.00
1.08
1.08
1.08
percent. With no change in foundry exports, this corresponds to
a decrease in domestic foundry consumption of 1.05 percent. As
noted above, we suspect that the estimated elasticity of demand
6-39
-------
is too low. If in fact the elasticity of demand is -2 instead of
-1.03, the percentage changes in quantity of foundry coke
produced and consumed are somewhat larger: -1.59 percent. The
price of foundry coke, on the other hand, would increase by only
0.80 percent under these conditions.
Thus, the MACT controls projected to take effect in 1993 are
expected to reduce the quantity of coke produced by at most
approximately one percent. Relative impacts on steel are even
smaller.
6.6.4.1.2 Absolute Changes in Quantities and Prices.
Table 6-14 shows the absolute changes in the quantities and
prices in the affected markets, computed by combining the
percentage changes shown in Table 6-13 and the baseline prices
and quantities shown in Table 6-12. Under MACT, furnace coke
production is projected to decrease by about 142,000 Mg. Imports
of furnace coke increase by 133,000 Mg, so consumption of furnace
coke decreases by only 9,000 Mg. The price of furnace coke is
projected to increase by $0.20 per Mg. Iron ore use is projected
to decrease by 8,000 Mg, and its price to decline by $0.06 per
Mg. As a result of these changes, relatively small changes also
occur in the market for steel. The price of steel increases by
TABLE 6-14. CHANGES IN PRICES AND QUANTITIES IN AFFECTED MARKETS UNDER MACT
MINIMUM CONTROL LEVEL, 1993
Price*
Quantity"
Furnace Coke Sector
Coal
Labor
Furnace coke produced
Furnace coke imported
Furnace coke exported
Furnace coke consumed
Other steel inputs (iron ore)
Domestic coke-using steel produced
Electric arc furnace steel produced
Total domestic steel produced
Steel imported
Steel exported
Steel consumed
Foundry Coke Sector
Coal
Labor
Foundry coke produced
Foundry coke consumed
0.00
0.00
0.20
-0.06
0.04
0.01
0.03
0.00
0.00
1.68
-186
-80
-142
133
0
-9
-8
-7
1
-6
2
-0
-4
-44
-14
-19
-19
'Units are 1991 $/Mg except for labor. Labor's price is in 1991 $/hour.
''Units are 103 Mg except for labor. Labor is measured in jobs.
about $0.03 per Mg, and steel production using pig iron from
blast furnaces is expected to decrease by 7,000 Mg. Electric arc
steel production increases by 1,000 Mg, so total domestic steel
production decreases by 6,000 Mg. Exports decrease slightly and
6-40
-------
imports increase slightly.
decrease by 4,000 Mg.
Steel consumption is projected to
Foundry coke incurs relatively small impacts under MACT.
Production and consumption of foundry coke decrease by about
19,000 Mg, while prices for foundry coke increase by $1.71 per
Mg. If the elasticity of demand is -2 instead of -1.03, the
quantity produced and consumed would decrease by 27,000 Mg, while
the price would increase by $1.28 per Mg.
6.6.4.1.3 New Prices and Quantities. Table 6-15 shows the
resulting prices and quantities in each of the affected markets
under MACT, computed by combining the absolute changes in
Table 6-14 with the baseline quantities and prices shown in
Table 6-12.
TABLE 6-15.
NEW PRICES AND QUANTITIES IN AFFECTED MARKETS UNDER MACT
MINIMUM CONTROL LEVEL, 1993
Price*
Quantity11
Furnace Coke Sector
Coal
Labor
Furnace coke produced
Furnace coke imported
Furnace coke exported
Furnace coke consumed
Other steel inputs (iron ore)
Domestic coke-using steel produced
Electric arc furnace steel produced
Total domestic steel produced
Steel imported
Steel exported
Steel consumed
Foundry Coke Sector
Coal
Labor
Foundry coke produced
Foundry coke consumed
54.08
16.21
91.08
32.72
538.80
538.77
538.79
51.79
16.21
162.18
28,008
12,041
21,393
6,140
520
27,013
73,059
54,195
33,629
87,824
17,639
4,546
100,917
4,004
1,247
1,695
1,695
•Units are 1991 $/Mg except for labor. Labor's price is in 1991 $/hour.
••Units are 103 Mg except for labor. Labor is measured in jobs.
Under MACT, furnace coke production falls to 21.4 million
Mg, and the price increases to $91.08 per Mg. Furnace coke
imports increase to 6.1 million Mg, and furnace coke consumption
falls to 27.0 million Mg. Steel production using pig iron
decreases to 54.2 million Mg, and total steel production
decreases to 87.8 million Mg. The price of steel remains
approximately $539 per Mg.
In the foundry coke market, under MACT, the quantity of coke
produced falls to 1.70 million Mg, while the price increases to
$162.18 per Mg. Foundry coke consumption falls to 1.70 million
6-41
-------
Mg. If the elasticity of demand were -2 instead of -1.03, the
quantity of foundry coke produced and consumed would decrease to
1.69 million Mg, and its price would increase to approximately
$161.78 per Mg.
6.6.4.2 Impacts of LAER standard, 1998. The baseline
against which the impacts of LAER were measured was constructed
as described in Section 6.6.3.2. Table 6-16 shows this 1998
TABLE 6-16. BASELINE PRICES AND QUANTITIES IN AFFECTED MARKETS, 1998
Price*
Quantity"
Furnace Coke Sector
Coal
Labor
Furnace coke produced
Furnace coke imported
Furnace coke exported
Furnace coke consumed
Other steel inputs (iron ore)
Domestic coke-using steel produced
Electric arc furnace steel produced
Total domestic steel produced
Steel imported
Steel exported
Steel consumed
Foundry Coke Sector
Coal
Labor
Foundry coke produced
Foundry coke consumed
54.08
16.21
91.08
91.08
91.08
91.08
32.72
538.80
538.77
538.79
538.79
538.79
538.79
51.79
16.21
162.18
162.18
27,070
11,637
20,676
6,128
520
26,284
70,694
52,708
33,629
86,337
17,636
4,456
99,427
4,084
1,254
1,703
1,704
•Units are 1991 $/Mg except for labor. Labor's price is in 1991 $/hour.
"Units are 103 Mg except for labor. Labor is measured in jobs.
baseline. Then, adding unit compliance costs to baseline supply
price and sorted the data by the resulting supply price with
compliance costs produced supply curves with the LAER standard in
place. Exogenous shift parameters were computed, and these
parameters were input into the model. The model then simulated
the response to the implementation of LAER controls in 1998 in
the interrelated markets for coke and steel.
6.6.4.2.1 Percentage Changes in Quantities and Prices.
Table 6-17 shows the percentage changes in the prices and
quantities in affected markets resulting from the LAER control
level. In the furnace coke sector, we estimate that the quantity
of coke produced will decrease by 2.13 percent. This decrease is
6-42
-------
TABLE 6-17. PERCENTAGE CHANGES IN PRICES AND QUANTITIES IN AFFECTED
MARKETS UNDER LAER CONTROL LEVEL, 1998
Price" Quantity"
Furnace Coke Sector
Coal
Labor
Furnace coke produced
Furnace coke imported
Furnace coke exported
Furnace coke consumed
Other steel inputs (iron ore)
Domestic coke-using steel produced
Electric arc furnace steel produced
Total domestic steel produced
Steel imported
Steel exported
Steel consumed
0.00
0.00
0.00
0.68
-0.04
0.02
0.01
0.02
-2.13
-2.13
-2.13
6.74
0.00
-0.11
-0.04
-0.04
0.01
-0.02
0.03
-0.03
-0.01
Foundry Coke Sector
Coal . 0.00-2.60
Labor 0.00 -2.60
Foundry coke produced . 2.53 -2.60
Foundry coke consumed -2 . 60
accompanied by a 0.68 percent increase in the price of furnace
coke. Because imports of furnace coke are projected to increase
by 6.74 percent, the overall quantity of furnace coke consumed
falls by only 0.11 percent. A small decrease in the quantity of
coke-using steel produced (0.04 percent) combines with small
increases in EAF steel production (0.01 percent) and steel
imports (0.03 percent) and results in a 0.01 percent decrease in
steel consumed in the U.S.
Relative impacts in the foundry coke sector are larger.
Foundry coke production and consumption are projected to fall by
2.60 percent in the main scenario and price would increase by
2.53 percent. If the elasticity of demand for foundry coke is
actually -2 instead of -1.03, the quantity of foundry coke
produced and consumed would decrease by 3.84 percent. In that
case, the price would increase by only 1.92 percent.
6.6.4.2.2. Absolute Changes in Quantities and Prices.
Table 6-18 shows the absolute changes in the quantities and
prices in the affected markets attributable to LAER, computed by
combining the percentage changes shown in Table 6-17 and the
baseline prices and quantities shown in Table 6-16. Furnace coke
production is projected to fall by 441,000 Mg. Imports of
furnace coke are projected to increase by 413,000 Mg, so furnace
coke consumption decreases by only 28,000 Mg. The price of
furnace coke is projected to increase by $0.62 per Mg. Small
decreases in the production of coke-using steel and total steel
combine with small increases in EAF production and steel imports
6-43
-------
TABLE 6-18.
CHANGES IN PRICES AND QUANTITIES IN AFFECTED MARKETS UNDER LAER
CONTROL LEVEL, 1998
Price1
Quantity"
Furnace Coke Sector
Coal
Labor
Furnace coke produced
Furnace coke imported
Furnace coke exported
Furnace coke consumed
Other steel inputs (iron ore)
Domestic coke-using steel produced
Electric arc furnace steel produced
Total domestic steel produced
Steel imported
Steel exported
Steel consumed
Foundry Coke Sector
Coal
Labor
Foundry coke produced
Foundry coke consumed
0.00
0.00
0.62
-0.19
0.13
0.04
0.10
0.00
0.00
4.10
-577
-248
-441
413
0
-28
-25
-22
3
-19
5
-1
-12
-106
-33
-44
-44
•Units are 1991 $/Mg except for labor. Labor's price is in 1991 $/hour.
"Units are 103 Mg except for labor. Labor is measured in jobs.
to result in a projected decrease in steel consumption of 12,000
Mg. The price of various types of steel is projected to increase
by $0.04 to $0.13 per Mg.
Production and consumption of foundry coke decrease by about
44,000 Mg in our main scenario, while prices for foundry coke
increase by $4.10 per Mg. If, on the other hand, the elasticity
of demand for foundry coke is -2 instead of -1.03, foundry coke
production and consumption would decrease by 65,000 Mg, and the
price would increase by only $3.10 per Mg.
6.6.4.3 New Prices and Quantities. Table 6-19 shows the
resulting prices and quantities in each of the affected markets
under each of the control levels, computed by combining the
absolute changes in Table 6-18 with the baseline quantities and
prices shown in Table 6-16.
Production of furnace coke is projected to fall to 20.2
million Mg. Imports of furnace coke are expected to increase to
6.5 million Mg, so that furnace coke consumption falls to 26.3
million Mg. The price of furnace coke is expected to increase to
$91.70 per Mg. Only small adjustments are projected to take
place in the markets for steel. Coke-using steel produced
decreases to 52.7 million Mg, while EAF production increases to
6-44
-------
TABLE 6-19.
NEW PRICES AND QUANTITIES IN AFFECTED MARKETS UNDER LAER
CONTROL LEVEL, 1998
Price*
Quantity"
Furnace Coke Sector
Coal
Labor
Furnace coke produced
Furnace coke imported
Furnace coke exported
Furnace coke consumed
Other steel inputs (iron ore)
. Domestic coke-using steel produced
Electric arc furnace steel produced
Total domestic steel produced
Steel imported
Steel exported
Steel consumed
Foundry Coke Sector
Coal
Labor
Foundry coke produced
Foundry coke consumed
54.08
16.21
91.70
32.53
538.93
538.81
538.89
51.79
16.21
166.28
26,492
11,389
20,235
6,541
520
26,256
70,669
52,686
33,632
86,318
17,641
4,545
99,415
3,977
1,221
1,660
1,660
•Units are 1991 $/Mg except for labor. Labor's price is in 1991 $/hour.
""Units are 103 Mg except for labor. Labor is measured in jobs.
33.6 million Mg. Total domestic steel production decreases to
86.3 million Mg, while steel consumption in the U.S. falls to
99.4 million Mg. The new prices of different types of steel
range from $538.81 to $538.93 per Mg.
In the foundry coke market, production and consumption fall
to 1.66 million Mg, while the price increases to $166.28 per Mg.
Under the scenario where the elasticity of demand is higher, the
quantity of foundry coke falls to 1.64 million Mg, while the
price increases to only $164.88 per Mg.
6.6.4.4 Battery Closures. The decreases in coke
production that result under the MACT standard and under both
cost scenarios for the LAER standard may result in the closure of
some coke batteries. This analysis measures closures in terms of
the average-sized battery in each sector. That is, the change in
production is compared with the size of the average battery in
each sector. If the change in production exceeded 50 percent of
the output of an "average" battery, one battery closure would
occur. In so doing, we assumed that one battery would make all
the adjustment in quantity. Possibly the adjustment in quantity
would be spread over many operating batteries, and no batteries
would close.
Under MACT, the reduction in coke production can be
accomplished by reducing the output of a single average furnace
coke battery and a single foundry coke battery, with no batteries
6-45
-------
being required to close. Under LAER, two furnace batteries are
projected to close. If the elasticity of demand for foundry coke
is -1.03, no batteries would close. The projected decrease in
foundry coke production could be accompanied by reducing the
output of one average foundry battery. If, on the other hand,
the elasticity of demand is -1.75 or greater (in absolute value),
one foundry battery is projected to close.
This projection must be interpreted in the context of the
projected decline in furnace coke consumption and foundry
production at baseline. These adjustments are projected to
result from the expected decrease in furnace coke consumption and
foundry coke production in the absence of this regulation. Over
the period from 1990 to 2010, furnace coke consumption is
projected to decline by 2.86 million Mg. During the same period,
foundry coke production is projected to decline by 0.29 million
Mg, resulting in the closure of batteries. EPA has been
informed by industry that five batteries at one facility will be
shut down by 1998. Two other batteries at two facilities are not
expected to meet LAER and may close by 2003. These closures,
which have not been attributed by their owners to this
regulation, far outnumber the closures attributed to the NESHAP
in this analysis.
6.6.5 Evaluation of Small Business Impacts
The Clean Air Act requires that economic impacts on small
businesses be investigated and that significantly adverse impacts
be mitigated. The first step in such an investigation is to
identify small businesses owning affected facilities. The Small
Business Administration's definition of "small business" for each
SIC Code is given in 13 CFR Part 121. For SIC code 3312, Blast
Furnaces and Steel Mills, small firms are those with less than
1000 employees. Data were obtained about numbers of employees
from Dun and Bradstreet's DUNS Market Identifiers38 and from
telephone conversations with some of the affected facilities.3*45
Of the companies owning coke plants in operation in 1991, only
four are small. Three of these are foundry coke producers:
• Erie Coke,
• Toledo Coke, and
• Tonawanda Coke.
The other small company, New Boston Coke, is a furnace
producer. All are merchant coke producers. The three foundry
coke producers are all owned by one individual; but even if they
are treated as one firm, it would qualify as a small firm.
6-46
-------
One measure of a "significant" impact is whether compliance
costs exceed 1 percent of baseline production costs. At the
minimum MACT control level, estimated cost of compliance exceeds
1 percent of estimated baseline production costs for two of the
four small companies. Under LAER, compliance costs exceed
1 percent of baseline costs for all four. For one of the small
companies, estimated LAER compliance costs exceed 10 percent of
estimated baseline costs of production. This criterion,
therefore, indicates that significant impacts will occur as a
result of the candidate NESHAP.
This conclusion must be interpreted in the context of the
effect of the regulation on the firms' revenues. One of the four
small firms is estimated to be unprofitable at baseline. It is
projected to become more unprofitable in the MACT and LAER
analyses. The other three small businesses are believed to be
profitable at baseline. They continue to be profitable under
both the MACT and LAER scenarios, although one that is thought to
be extremely profitable at baseline is expected to be somewhat
less profitable with the controls in place. The other two small
businesses are expected to increase their profitability under
LAER.
The market prices for furnace and foundry coke are projected
to increase as a result of the regulation. This will result in
higher revenues for the facilities that continue in production
with the regulation in place. The firms that own these
facilities will suffer adverse impacts only if their increased
revenues do not exceed the increased production costs arising
from complying with the emissions controls. Under MACT, three of
the small firms are expected to incur compliance costs exceeding
the increase in revenue they are projected to receive. Under
LAER, two small businesses are expected to incur compliance costs
exceeding the increase in revenues predicted by the model.
Although one firm is expected to lose approximately $1.2 million
more under LAER relative to its baseline positions, it will
continue to be profitable under LAER. However, the other small
firm is expected to lose $1.6 million and may become more
unprofitable under LAER than it is currently.
The other two small firms are expected to experience profits
that increase by $650,000 and $800,000 as a result of the
regulation. This result occurs because market price is projected
to increase by more than their costs of control. These firms are
projected to be more profitable with the controls in place than
at baseline, so we do not believe they will be adversely affected
by the controls. Because only a small absolute number of small
firms (two) is projected to incur significant adverse impacts as
a result of the standard, a formal Regulatory Flexibility
Analysis is not required.
6-47
-------
6.6.6 Indirect Impacts
In addition to the impacts projected by the model, there are
additional impacts that may result from the regulation. Such
impacts include the effects of decreased coke production on the
markets for the by-products of coke ovens, impacts on communities
of plant closures, and additional costs incurred by companies
when facilities are closed. Although it is beyond the scope of
this study to analyze these impacts quantitatively, they are
discussed qualitatively in this section.
6.6.6.1 Impacts on By-Product Markets. The projected
decreases in coke production are expected to be associated with
decreased production of coke by-products. The by-products of
carbonizing coal to make coke include gas, gas liquor, light oils
and tar. By-product yields are a function of the volatility of
the coals carbonized. Because the production of foundry coke
requires using a less volatile coal mixture than does the
production of furnace coke, carbonizing coal for foundry coke
production yields more coke (foundry coke and "other" coke) and
fewer by-products than for furnace coke production.
Carbonization of 1 Mg of dry coal to produce furnace (foundry)
coke yields approximately 0.7 Mg (0.8 Mg) of coke, 12,100
(10,450) cubic feet of gas, 35 (15) gallons of gas liquor, 4
(2.5) gallons of light oils, and 9 (5) gallons of coal tar.46-47
By-product yields for both types of coke production will vary
with the different recovery methods and different carbonization
temperatures.
About 40 percent of the coke gas produced is used to
underfire the coke ovens that produce it. In integrated steel
plants virtually all of the excess gas produced is used for fuel
in other steps of the steel-making process. Foundry coke
producers generally use what they can of the excess gas, and
attempt to sell the remainder. Several foundry coke producers
simply burn their excess gas to dispose of it. Coke gas is not
as volatile as natural gas. It typically only has about 500 Btu
per cubic foot compared to 1,000 Btu per cubic foot for natural
gas.49 Those coke producers that use or sell their excess gas
could significantly affect local energy market conditions if they
stopped producing the coke gas that currently provides for their
own or others' energy needs. Other down-stream products that use
coke gas distillates include synthetic resins, adhesives, anti-
freeze, insecticides, explosives, Pharmaceuticals, and
fertilizers.50
Gas liquor is used in the production of fertilizers,
explosives, and pyridine tar bases, which are inputs to the
production of many products including dyestuffs, latex adhesives,
Pharmaceuticals, solvents, and many others.51
Light oils derived from carbonizing coal are used to produce
hundreds of products ranging from aspirin, flavorings,
6-48
-------
detergents, and deodorants to poison gas, herbicides,
photographic chemicals, and saccharin.52 Light oils produced in
coking make up about 10 percent of the U.S. market supply.5-'
The by-product markets that would be most significantly
affected by a reduction in the amount of coke produced in by-
product recovery plants are the market for coal tar and its down-
stream products. Coal carbonization is the only available source
of coal tar. Coal tar is used to make creosote oils, the most
widely used and efficient products for preserving wood.54 Coal
tar is also refined to produce tars and pitch used for roofing,
road surfacing, sidewalk composition, waterproofing, insulation,
and many other uses. It is distilled to extract other chemicals
such as naphthalene that is used to produce products ranging from
mothballs and insect repellents to solvents, detergents, fabric
dyes, inks, and paints. 5
6.6.6.2 Community Impacts. When a facility is closed, the
workers employed there generally become unemployed. Under MACT,
we project that employment will fall by 96 workers at furnace and
foundry coke facilities. No facility closures are projected, so
these changes in employment may be interpreted as temporary
layoffs associated with reductions in output at facilities that
remain open. Under LAER, we project the closure of two furnace
batteries, with an associated reduction in employment of 279
workers. (Information from industry indicates that as many as
500 workers may be displaced by the closing of two batteries.)
In the foundry sector, we project the closure of at most one
battery. In our main scenario, there are no closures, but output
reductions are associated with a decrease in employment of 37
workers. In our alternative scenario, however, output falls by
more than 50 percent of the capacity of the average foundry
battery; this might result in a closure. Our model predicts a
decrease in employment of only 59 workers, but if a battery were
closed, that might result in the displacement of as much as 250
workers.
When workers become unemployed, their incomes fall and they
spend less money in the local community. This situation in turn
reduces the incomes of others in the community, who in turn spend
less. If a large enough number of workers are displaced due to
the original closure, additional workers in the community may
lose their jobs. Thus, the immediate impact of the regulation is
to reduce employment and income by the amount of the displaced
coke-oven workers. This reduction will generally result in
further reductions elsewhere in the community. For industries
that produce final goods, multipliers are computed by the
Department of Commerce that estimate the ultimate change in
employment and income that will result from the original decrease
in employment due to the facility closure. Because coke is an
intermediate good, we are unable to use these multipliers to
estimate the total change in employment and income that will
result. Although we are unable to quantify it, we recognize that
6-49
-------
the total change in 'community employment and income will be
somewhat higher than the direct change resulting from the
closure.
6.6.6.3 Additional Costs of Closure. When a battery shuts
down, the facility avoids the costs associated with coke
production at that battery. They also avoid the costs associated
with achieving any regulations that apply to the battery. The
facility also incurs costs associated with shutting down the
battery. These costs include personnel costs, the costs of
cleaning up the site, and the costs of foregone sales of coke by-
products .
Estimates of employee-related costs of coke battery closures
were provided by Harry Kokkinis of Locker Associates and Monty
Stuart of Bethlehem Steel. Mr. Kokkinis defined employee costs
as the costs of paying early retirement benefits, retiree health
insurance, and supplemental unemployment benefits to displaced
workers. Assuming that half of the workers who would lose their
jobs due to a premature battery.closure would be over the age of
55 and therefore eligible for early retirement, Mr. Kokkinis
estimated the total cost per job lost due to a battery closing
earlier than expected to be about $60,000. Mr. Stuart estimated
the total cost per employee of permanently closing any Bethlehem
facility to be $100,000. This figure is the combined cost of
cash outlays to the employee pension fund and to the individual
employee required as the result of closing a facility
permanently.
Mr. Stuart also provided an estimate of the cost of the site
remediation activities for a typical by-product recovery coke
plant required to comply with RCRA standards for continuing
releases and corrective action in the event of the permanent
closure of a hypothetical steel mill. The estimate was taken
from a study conducted for AISI by REMCOR, Inc., in 1989. The
size, in terms of production capacity,-of the hypothetical coke
plant for which costs were estimated was 1,090,909 Mg of furnace
coke per year. The REMCOR estimate of the total capital cost of
site remediation was $15,100,000. The REMCOR estimate for annual
operating and maintenance costs during the clean-up period was
$210,000 per year (all costs were estimated in 1989 dollars).
Our estimate of the revenues foregone because of foregone
by-product sales resulting from a reduction in coke production
depends on
• the unit prices of the individual by-products,
• the amount that production is reduced, and
• the type of coke facing the reduction in production.
6-50
-------
The thermal value of 1,000 cubic foot of coke gas is
approximately 0.5 decatherms. The going rate for 1 decatherm of
energy in January, 1992, was $2.60. The price of coal tar is
approximately $0.45 per gallon, and the price of light oils is
about $0.32 per gallon. We do not believe that any net cost
accrues to foregone sales of ammonia, the primary by-product
derived from gas liquor, because the cost of properly disposing
of this by-product is as high as the sale price. Based on the
estimated by-product yields from carbonization of 1 Mg of coal to
produce furnace (foundry) coke presented in Section 6.6.6.1, we
estimate the foregone revenue per megagram reduction of furnace
(foundry) coke produced to be $21.58 ($16.64).56
This amount represents the gross revenue loss per megagram
reduction in coke produced rather than the net revenues loss.
The processing and marketing of by-products result in costs,
which are also foregone when by-product production decreases. We
do not have estimates of these by-product production costs, so we
cannot compare the decrease in revenues to the decrease in costs.
It is likely, however, that net revenue from by-product sales
will decrease due to decreases .in coke production.
6.7 CONCLUSIONS
The aggregate impacts of this regulation on affected
markets, even at the more stringent control level, are not very
large. No batteries are expected to close as a result of MACT,
and only two or three are expected to close as a result of LAER.
This impact must be evaluated in the context of current
conditions in the coke industry. Although coke production has
recovered slightly in the last few years, a general downward
trend in coke production and coke consumption by steel-makers,
accompanied by the shutting down of coke batteries, is evident.
(More than half the batteries in existence in 1979 had shut down
by 1991.) Many batteries presently operating are old and will
need to be replaced within the next 10 years or so, unless
technology can be developed to extend their lives beyond 35 years
(40 percent of the batteries in North America fall into this
category, according to the International Iron and Steel
Institute) ,57 A declining quantity of coke consumption is
projected over the period 1990 through 2000, due largely to a
decline in projected coke-using steel production. This analysis
assumes that the ratio of coke consumption to coke-using steel
production will remain constant throughout the period. For
several reasons, this may not be the case.
The shortage of investment capital, the considerable cost of
building coke ovens, and the increasing cost of meeting
antipollution laws are
some of the reasons why the industry throughout the
world is developing and evaluating alternative
methods to reduce coke requirements and possibly
6-51
-------
eliminate the need entirely. These developments
include the use of pulverized or granulated coal
injection into blast furnaces, with or without
oxygen injection, and direct iron ore smelting with
coal. Other developments are formed coke that uses
less costly noncoking coals and heat-recovery coke-
making.58
Indeed, Inland Steel is expected to install a new nonrecovery
process for coke-making with cogeneration of electricity that is
expected to have much lower emissions than conventional by-
product recovery coke-making. In Europe, producers are
developing a smelting/reduction process that eliminates the need
for coke and provides a degree of continuity between the iron-
making and steel-making production processes.59 Other influences
suggesting lower coke demands include the direct reduction iron-
making and the direct steel-making methods discussed in Section
6.2.
Thus, technological developments in the steel industry
suggest that the demand for coke, relative to other steel-making
inputs, may fall over time, as coke-saving technical changes are
adopted by steel producers. If trends involving the substitution
of other materials, such as plastics and ceramics, for steel
continue, the demand for steel may continue to decrease,
resulting in even lower demand for coke than projected.
Therefore, the regulation may at most accelerate the
reductions in coke production that are occurring as a result of
the changing conditions in the markets for iron and steel.
Quantitatively, the impact of the regulation will probably be
small relative to the influence of technical and economic changes
presently occurring in the industry. Only two or three batteries
are predicted to cease production as a result of the LAER
standard, and only two small firms are projected to incur
significant adverse economic impacts.
6-52
-------
REFERENCES
1. DRI/McGraw-Hill Steel Industry Review, Long-Term Steel
Industry Outlook and Iron and Steel Castings Shipments,
Fourth Quarter 1991. Obtained from Charles L. Bell, U.S.
Department of Commerce, Office of Basic Industries.
2. Labee, C.L., and N.L. Samways. Developments in the Iron and
Steel Industry, U.S. and Canada - 1990. Iron and Steel
Engineer, February 1991, p. D-22.
3. Reference 2.
4. U.S. Department of Energy, Energy Information Administration
Quarterly Coal Report. 1989.
5. Douglas, John. Electrifying the Foundry Fire. EPRI
Journal, October/November 1991.
6. U.S. Steel. The Making, Shaping, and Treating of Steel, 9th
Edition. Pittsburgh. 1971.
7. Telecon. Eppich, Robert, American Foundryman's Society,
with Robert Cushman, Research Triangle Institute. January 8
and 9, 1992.
8. Reference 5.
9. Duggan, Brian. U.S. Industrial Outlook 1991—Metals.
September 1990.
10. Duggan, Brian. U.S. Industrial Outlook 1990—Ferrous
MetaIs. September 1989.
11. Reference 7.
12. Telecon. Engle, Mark, American Coke and Coal Chemicals
Institute, Washington, DC, with Cushman, Robert, Research
Triangle Institute. January 15, 1992.
13. Reference 7.
14. U.S. Steel. The Making, Shaping, and Treating of Steel,
10th Edition. Pittsburgh, Association of Iron and Steel
Engineers. 1985.
15. Reference 6.
6-53
-------
16. Telecon. Dusel, Martin, Indianapolis Coke (Citizens' Gas
and Coke Utility) with Cushman, Robert, Research Triangle
Institute. March 3, 1992.
17. U.S. Department of Commerce. U.S. Industrial Outlook, Coal.
Washington, DC. 1990.
18. U.S. Department of Energy, Energy Information Agency
Quarterly Coal Report. 1990.
19. Telecon. Kokkinis, Harry, Locker Associates with Cushman,
Robert, Research Triangle Institute. January 17, 1992.
20. Stuart, L.M. Bethlehem Steel Corporation. Letter to Robert
Cushman, Research Triangle Institute, describing costs of
coke battery closure. January 20, 1992.
21. American Coke and Coal Chemicals Institute (ACCCI),
Employment Survey, Table 2. Provided by David Ailor, ACCCI,
1992.
22. Telecon. Brown, M., ABC Goke, with Sarmiento, Rhythm,
Research Triangle Institute. March 6, 1991.
23. Telecon. Erie Coke, Erie, PA, with Sarmiento, Rhythm,
Research Triangle Institute. March 6 and March 11, 1991.
24. Telecon. Landsell, M., Sloss Industries, Birmingham, AL,
with Sarmiento, Rhythm, Research Triangle Institute. March
6 and March 8, 1991.
25. Telecon. Citizens' Gas, Indianapolis, IN, with Sarmiento,
Rhythm, Research Triangle Institute. March 11, 1991.
26. Frazier, Richard M., Hickman, Williams, and Company.
Written communication with Robert Cushman, Research Triangle
Institute, detailing historical Indianapolis foundry coke
price increases. December 31, 1991.
27. Pearson, John M., ABC Coke, Birmingham, AL. Written
communication with Robert Cushman, Research Triangle
Institute, including an 11-year history of published prices
and estimated actual foundry coke prices for ABC Coke.
January 31, 1992.
28. Reference 27.
29. Reference 1.
30. Telecon. Madhava, Bindu, Davy/Still Otto, with Cushman,
Robert, Research Triangle Institute. January 20 and 22,
1992.
6-54
-------
31. PEDCo Environmental, Inc. Revised Technical Approach for a
Coke Production Cost Model. Prepared for U.S. Environmental
Protection Agency. January 1991.
32. Johnson, Norman L., and Samuel Kotz. Continuous Univariate
Distributions, Vol. 1. New York, John Wiley & Sons, 1970.
33. Reference 30..
34. Telecon. Eriser, Jim, ICF-Kaiser Engineering with Cushman,
Robert, Research Triangle Institute. January 20, 1992.
35. Reference 34.
36. U.S. Environmental Protection Agency/Industrial Economics,
Inc. Coke Ovens Capital Availability Analysis. Draft
Memorandum. March 6, 1992.
37. Ramachandran, V. An Econometric Model of the U.S. Steel
Industry. Draft Report. Research Triangle Institute.
Research Triangle Park, NC.. April 1981.
38. Dun and Bradstreet. DUNS Market Identifiers. February
1991. Online computerized database.
39. Telecon. Ailor, D., and Engle, M., American Coke and Coal
Chemicals Institute, with Heller, Katherine, Sarmiento,
Rhythm, and Curtis-Powell, Sandra, March 15, 1991.
40. Telecon. Hall, D., Koppers Industries, Woodward, AL, with
Sarmiento, Rhythm, Research Triangle Institute. March 5,
March 6, and March 11, 1991.
41. Reference 24.
42. Telecon. Webb, K., Terre Haute Coke, Terre Haute, IN, with
Sarmiento, Rhythm, Research Triangle Institute. March 6,
1991.
43. Telecon. Seals, M., Empire Coke, Holt, AL, with Sarmmiento
Rhythm, Research Triangle Institute. March 8, 1991.
44. Reference 23.
45. Reference 22.
46. Reference 16.
47. Reference 6.
48. Reference 16.
49. Reference 16.
6-55
-------
50. Koppers Industries. Products Derived from Coal and Coal
Products Tree. Information Pamphlets provided by D.->n Sweet.
1992.
51. Reference 50.
52. Reference 50.
53. Telecon. Lung, Henry, U.S. Steel (USX), with Cushman,
Robert, Research Triangle Institute. March 18, 1992.
54. Telecon. Sweet, Don, Koppers Industries, with Cushman,
Robert, Research Triangle Institute. March 13, 1992.
55. Reference 50.
56. Reference 16.
57. Reference 2.
58. Reference 2.
59. Reference 2.
6-56
-------
CHAPTER 7
BENEFITS ANALYSIS
7.1 INTRODUCTION
This chapter presents analyses of the potential benefits
associated with the National Emission Standards for Coke Oven
Batteries. Benefits represent the improvement in society's well-
being as a result of improved air quality. Total benefits are
comprised of both use and non-use values. Use values are the
values associated with an individual's desire to avoid his or her
own exposure to an environmental risk; non-use values are values
an individual may have for lowering the level of risk that is
unrelated to his or her own exposure. This chapter presents
monetized estimates of benefits for parts of three of the health
and welfare components of use value associated with various
levels of coke oven emission reductions: mortality, morbidity,
and household soiling. The estimated benefits represent the
incremental improvement from a baseline level of coke oven
emissions that reflects current operating practice to compliance
with the CAAA default limits of emission control. Data, time and
resource limitations preclude a quantitative analysis for all
potential benefit categories, therefore total benefits are
understated. However, other components of total benefits are
qualitatively discussed.
7.2 POLLUTANTS
Coke production results in the release of chemically-complex
emissions to the atmosphere. Coke oven emissions are comprised
of both gases and respirable particulate matter. The particulate
fraction contains polycyclic organic matter, aromatic compounds
(e.g. benzene, toluene), trace metals (e.g. arsenic, lead,
chromium), and inorganic gases (e.g. nitric oxide, carbon
monoxide, hydrogen sulfide), some compounds of which are known
human carcinogens. The condensed particulates are within the
respirable size range.
Table 7-1 presents the relative composition of coke oven gas
based on an analysis conducted at USX Clairton.1 Per ton of
coal, almost 60% by weight of the coke oven gas is water vapor
7-1
-------
while approximately 37% of the coke oven gas is composed of
carbon monoxide, heavy hydrocarbons, benzene, hydrogen, carbon
dioxide, ethylene, and ethane. The remaining 3% of coke oven gas
is composed of compounds such as napthalene, hydrogen sulfide,
hydrogen cyanide, ammonia, toluene, and xylene. It should be
noted that this analysis does not cover the trace elements that
are simultaneously emitted such as arsenic, beryllium, cadmium,
chromium, cobalt, iron, lead, nickel, selenium and mercury.
7.3 ENVIRONMENTAL BENEFITS
The release of hazardous compounds to the atmosphere as a
result of coke production can adversely impact human health and
welfare. Coke oven emissions may accelerate the onset of
mortality and increase the incidence of acute and chronic
morbidity. Reductions in coke oven emissions as a result of
regulation will improve ambient air quality and in turn reduce
the incidence of adverse health and welfare effects. The health
and welfare improvements are the direct benefits of these
environmental regulations.
7.4 METHODOLOGY
Ideally, the estimation of potential economic benefits would
be accomplished using data, assumptions, and modeling techniques
specifically developed for the analytic objective. However, time
and resource constraints prevent this approach. Therefore,
benefit estimates are based upon existing studies that may be
applied to the health and welfare impacts from coke oven
emissions. The following discussion outlines the steps involved
in this benefit assessment.
7.4.1 Identification of Potential Benefit Categories
Table 7-2 illustrates the range of potential physical
effects categories that may result from coke oven emission
control strategies. For this analysis, benefits resulting from
the control of coke oven emissions can be derived for both
benzene soluble organics (BSO)(i.e., the organic component of
coke oven emissions that is soluble in benzene) and particulate
matter (PM).
7-2
-------
TABLE 7-1
COMPOSITION OF COKE OVEN GAS
COMPOUND
Water Vapor
Methane
Carbon Monoxide
Heavy Hydrocarbons ;
Hydrogen
Benzene
Carbon Dioxide •
Ethylene
Ethane
Napthalene
Hydrogen Sulfide
Ammonia
Nitrogen
Propylene
Butene *
Hydrogen Cyanide
Toluene
Carbon
Propane
Butane
Pentene
Tar Acids
Tar Bases
Solvents
Tar Abs Oil
Acetylene
Xylene
% OF COKE OVEN GAS
57.5
14.3
5.8
4.2
4.1
2.7
2.5
2.1
1.4
.8
.8
.8
.8
.4
.4
.3
.2
.2
.1
.1
.1
.1
.1
.1
.1
.05
.02
Coke oven gas analysis conducted at USX Clairton.' Analysis does not consider trace elements that are also emitted. These are
the following: arsenic, beryllium, cadmium, chromium, cobalt, iron, lead, nickel, selenium, mercury.
7-3
-------
TABLE 7-2.
POTENTIAL PHYSICAL EFFECTS CATEGORIES FOR
COKE OVENS NESHAP
Human Health Effects
Mortality Due to Chronic Exposure
Mortality Due to Acute Exposure
Morbidity Due to Chronic Exposure
Morbidity Due to Acute Exposure
Reduced Activity Days
Human Welfare Effects
Worker Productivity Losses
Odors
Non-Human Biolocrical Effects
Agriculture
Forestry
Recreational/Commercial Fishing
Ecosystem
Soilina & Materials Damacre
Residential/Commercial/ Industrial
Facilities
Miscellaneous Materials
BSO
*
PM
*
*
* - Quantitatively estimated for this analysis,
7-4
-------
7.4.2 Identification of Concentration-Response Functions
Appropriate for Benefit Estimation
All identified studies are screened on the basis of several
criteria, the most important of which are analytic quality and
potential for extrapolation of estimates for benefit analysis
(e.g. requisite air quality data available). As a result of this
screening analysis, monetized estimates of benefits are provided
for three benefit categories.
Emission controls will reduce ambient concentrations of the
hazardous compounds contained in coke oven emissions. This
analysis quantitatively estimates mortality benefits from
reductions in ambient BSO concentrations (used as a surrogate
measure for the carcinogenic component of coke oven emissions)
and acute morbidity and household soiling benefits from
reductions in ambient particulate matter concentrations. It is
not possible, however, to quantify the other potential benefit
categories presented in Table 7-2. For example, BSO and
particulate coke oven emissions may have additional adverse
health and welfare impacts (e.g., ecosystem, forestry,
agricultural, and recreational and commercial fishing effects)
for which there is no dose-response data.
There may be additional benefits that result from the
reduction of hazardous compounds not covered by BSO or
particulate matter benefits. For instance, some of the inorganic
gases such as ammonia and hydrogen sulfide that escape during the
coking process may contribute to an odor problem in the vicinity
of coke plants. Additionally, trace metal emissions (e.g., lead,
nickel, arsenic, and cadmium) may enter the aquatic environment
via atmospheric deposition, locally impacting aquatic ecosystems.
7.4.3 Development of Benefit Estimates
Benefit estimates are developed using the identified
concentration-response functions and the air quality improvements
predicted for each regulatory option. Figure 7-1 illustrates the
process of benefit estimate development.
The first step is to identify the magnitude of the ambient
air quality improvement that is estimated to occur in each area
and year. This is the improvement achieved due to implementation
of a particular coke oven emission control measure, relative to a
baseline situation that reflects existing controls.
The selected concentration-response functions are then used
to determine the health and welfare improvements that may occur
as a result of the improvement in ambient air quality for each
area and year for which there is air quality improvement.
7-5
-------
STEP1
STEP 2
STEP 3
STEP 4
STEPS
Identify air quality improvement
in area i at year t
Estimate health or welfare
improvement in area i at year t
3a
3b
Estimate economic value of health
or welfare improvement at area i at year t
Aggregate over t to obtain
discounted present values
Aggregate over i to obtain
regional totals
Figure 7-1. Benefits estimation process for an individual study.
7-6
-------
Consideration of mitigating and averting behavior is
important in determining the extent of population exposure to a
hazardous compound and in turn estimating health effect
incidence. For example, over time individuals may choose to move
away from areas surrounding coke oven plants to escape the
adverse effects of coke oven emissions. To the extent that this
is the case, comparison of population data over time may capture
this migration effect.
The third step is to impute an economic value to the
estimated changes in health and welfare (step 3a). It is also
possible to estimate economic values directly from the air
quality improvement for some classes of benefits (step 3b) such
as household soiling.
Benefit estimates are next discounted over a specified time
period such that results are in present value terms and then
summed across geographic area according to regulatory
alternative.
7.4.4 Aggregation to Total Incremental Benefits
Benefit estimates for each of the effects categories are
summed by regulatory alternative. These benefit estimates should
represent total benefits associated with achieving each of the
proposed levels of control versus present new and existing source
control requirements. Benefit estimates reflecting the
incremental change in emissions achieved through each of the
alternative emission standards are required inputs to correctly
analyze the total incremental benefits and costs of each
regulatory alternative.
7.5 DATA
7.5.1 Air Quality Data
The Industrial Source Complex Model-Long Term (ISCLT) was
used to estimate ambient concentrations of BSO around 28 coke
oven plants. The ISCLT model was set to the "regulatory mode"
and the source type was set to "volume source" in an urban
setting with flat terrain2. Each plant was assumed to have just
one oven 10 meters on a side and 5 meters high. An initial plume
width was set based on width and height of a single oven. All
emissions were assumed to exit from this one structure. In this
study, the ISCLT model calculated long-term concentrations which
were based on several years of meteorological observations. The
output, concentration profiles in a polar grid coordinate system
centered at each plant, were designed for direct input into the
exposure model. The polar grid extended from 200 meters out to
50 km for 16 different radial directions.
7-7
-------
Particulate matter concentration profiles were obtained by
extrapolation from ambient BSO concentrations. Ambient BSO
concentrations were multiplied by the ratio of total suspended
particulates to BSO (2.3) to yield ambient TSP concentrations3.
7.5.2 Demographic Information
Public exposure was estimated using data from the 1980
population census. However, enough data from the 1990 census
were available to derive correction factors to adjust exposure
results to reflect population changes between 1980 and 1990. The
adjustment process had little impact on the benefit estimates
because there was only a slight increase in population in the
counties affected by the coke oven emissions. In 1980 the
population in the 159 affected counties was 34 million; by 1990
the population had increased by only 40,000.
The morbidity benefit estimates require two additional
pieces of information: household income and the percentage of
population that is employed. Household income was obtained from
the 1990 Census. Because the 1990 Census labor participation
rate has not been released, the" Commerce Department's OBERS
projections of employment growth, combined with 1980 Census and
1986 Survey of Current Population data were used as a basis for
extrapolating the labor participation rate by county4.
All monetary values in this chapter are in 1991 dollars.
The Consumer Price Index - All Urban Consumers (CPI-U) was used
to adjust prices where needed.
7.6 EXPOSURE ASSESSMENT
The Human Exposure Model (HEM) was used to derive
quantitative expressions of public exposure to ambient
concentrations of BSO and TSP. The BSO and TSP concentration
profiles obtained from the. ISCLT model were used as input data
(see section 7.5). Because county-specific data was required to
run the PM morbidity and household soiling benefits models, a
specialized version of the HEM, called the Systems Applications
Human Exposure and Risk Model (SHEAR) option, was selected for
this study5. The HEM-SHEAR model contains the 1980 U.S. census
data by block group/enumeration districts (BG/EDs). As described
above, 1980 population was adjusted to 1990 population through
multiplication by a population growth factor. All people grouped
by BG/ED were defined as living at the approximate center of the
BG/ED area. Average exposure concentrations for BSO and TSP were
calculated by estimating the BSO and TSP concentrations at the
center of the BG/ED area and multiplying that concentration by
the number of people in that BG/ED. These products of people and
pollutant concentrations were summed over all BG/EDs in each
county to yield total exposure to both BSO and TSP.
*
7-8
-------
7.7 QUANTIFIED BENEFIT ESTIMATION
The following section outlines the studies selected to
estimate the benefits of coke oven emission standards. The
section also discusses the application of these studies to the
analysis and provides benefit estimates for mortality, morbidity
and household soiling. Benefits have been calculated in 1991
dollars.
7.7.1 Mortality Due to BSO Exposure
7.7.1.1 Study Selection. According to EPA's "Carcinogen
Assessment of Coke Oven Emissions", coke oven emissions are
carcinogenic to humans. Due to the complex chemical composition
of coke oven emissions, the benzene soluble organics (BSO)
fraction is used as a surrogate measurement for coke oven
emissions when conducting exposure analysis and cancer risk
assessment. EPA has calculated a unit risk estimate for BSO that
defines the relationship between exposure to coke oven emissions
and respiratory cancer risk (assumed to be linear). The unit
risk estimate is the lifetime risk associated with a lifetime (70
year) exposure to an average unit concentration of the pollutant.
The unit risk factor was derived from human epidemiological data
collected from coke oven and steel workers exposed to coke oven
emissions. Positive epidemiologic studies are generally treated
as the most conclusive evidence for a particular disease.
Additionally, the unit risk factor reflects an upper bound
estimate of cancer risk.
7.7.1.2 Application. The HEM-SHEAR model was used to
estimate cancer risk estimates associated with each coke oven
plant. BSO risk estimates were calculated by multiplying the BSO
exposure products (obtained from the exposure assessment) by the
unit risk estimate and dividing by 70. The unit risk estimate
for BSO is 6.2 x 10"* per 1 /xg/m3. Multiplying the exposure
products by the unit risk estimate provides the number of cancer
cases expected over a lifetime. Thus, division of lifetime
cancer cases by 70 years yields an estimate of the number of
cancer cases per year.
7.7.1.3 Valuation. The value of a change in mortality
rates is measured through the value of a change in mortality
risk. The value of risk reduction has been estimated in studies6
that examine the wage premia required by workers to accept risky
jobs. These studies reveal that workers require approximately an
additional $100 to $700 annually (1983 $) to accept an additional
mortality risk of 1 x 10"*. This is equivalent to $1.6 to $8.5
million per statistical life saved (1986 $).
In this analysis, mortality benefits are presented in 1991
dollars as low, mean, and high estimates reflecting values of
7-9
-------
$1.6 million, $4.4 million, and $8.5 million (1986 $) per
statistical life saved respectively.
7.7.2 Acute Morbidity due to Particulate Matter (PM) Exposure
7.7.2.1 Study Selection. A cross-sectional
microepidemiological study by Ostro7 was used to estimate the
particulate matter benefits associated with each of the coke oven
emission control options. The Ostro study focused on broad
health end points that relate work loss and restricted activity
days due to respiratory disease to fine particulate
concentrations. The PM Staff Paper* states that this study
provides strong qualitative support for a relationship between
current PM levels and restricted activity in adults.
Ostro analyzed six years of individual data from the
National Center for Health Statistics Health Interview Survey
(HIS) to examine the relationship between air pollution and
morbidity. Ostro's sample included all adults age 18 to 65 from
49 metropolitan areas for which pollution data and HIS sample
data were available. The sample contained approximately 12,OOO
adults for each of the six years from 1976 through 1981.
Three measures of morbidity were used in Ostro's analysis:
work loss days (WLD), restricted activity days (RAD), and
respiratory-related restricted activity days (RRAD). Information
on these morbidity measures was obtained in response to a survey
question asking the individual how many days did illness in the
previous two weeks prevent him/her from working or participating
in his/her usual activities.
The concentration-response functions estimated by Ostro
regressed a measure of the individual's acute WLD, RAD, or RRAD
against a measure of ambient particulate matter concentration,
the individual's personal and economic characteristics, and
temperature. The measure of particulate matter was a two-week
average lagged to represent the two-week exposure period prior to
the study period. In addition to the PM measure, the independent
variables included the individual's age, sex, race, education,
family income, marital status, existence of a chronic condition,
quarter of the survey, and average two-week minimum temperature.
The concentration-response functions that included WLD as the
dependent variable controlled for paid sick leave and whether the
individual worked in a blue or white collar job. In addition to
the basic set of variables, the RAD and RRAD concentration-
response functions included a variable reflecting whether or not
the restricted activity occurred on work time versus leisure
time.
7.7.2.2 Application. For this analysis, the
concentration-response function for TSP and RRAD's was used.
Average ambient total suspended particulate concentrations and
7-10
-------
the exposed population on the county level obtained from HEM are
used to calculate the baseline number of work loss days, reduced
activity days, and direct medical expenses resulting from current
levels of coke oven emissions. This study cannot be used to
estimate benefits for a reduction in pain and suffering prior and
subsequent to the receipt of medical care, therefore these values
may underestimate morbidity benefits.
7.7.2.3 Valuation. In this analysis, the cost-of-
illness approach is used to approximate willingness to pay to
avoid morbidity effects. A reduction in the number of work loss
days is valued at the average daily wage rate. A reduced
activity day is valued at one-half the average daily wage rate.
Direct medical expenditures are calculated for acute respiratory
conditions only. The range of benefit estimates presented
reflects variability around the regression coefficient of the
dose-response equation (mean + 2 standard deviations).
+
7.7.3 Household Soiling
7.7.3.1 Study Selection. . The 24 Standard Metropolitan
Statistical Areas longitudinal study by Mathtech9 was used to
measure household soiling benefits resulting from reductions in
the particulate component of coke oven emissions. In this model,
benefits are estimated directly from air quality changes without
first measuring physical damage. Adverse effects of pollutants
are reflected in changes in market demand and supply
relationships. Household behavior in terms of soiling
perception, cleaning activity, and expenditures to maintain a
given degree of cleanliness is estimated econometrically.
7.7.3.2 Application. The inputs to the household soiling
model are 1990 economic and socio-demographic data such as
product prices and household income. Background TSP
concentrations were obtained from EPA's National Air Quality and
Emissions Trend Report. 198810 which reports annual arithmetic
mean PM10 levels by city. PM10 values were divided by 0.55 to
recover the equivalent annual arithmetic mean TSP. Baseline
levels of household soiling due to current levels of coke oven
emissions were calculated from the model. These benefits
represent soiling occurring in the household sector only and do
not cover the commercial, industrial, and governmental sectors.
Therefore, this analysis may underestimate the total benefits of
reduced soiling. The range of benefit estimates presented
reflects variability around the regression coefficient of the
econometric function (mean + 2 standard deviations).
7-11
-------
7.8 FINDINGS FOR MACT AND LAER COMPLIANCE
The coke oven NESHAP will result in the control of emissions
from doors, topside and charging at 28 coke oven plants.
Exposure modeling results of current levels of coke oven
emissions reveal that 23 million people are exposed to the toxic
emissions released from these facilities. The cancer risk
assessment indicates that there are approximately 1.8 excess
cancer cases per year due to baseline levels of coke oven
emissions. Similarly, current levels of coke oven emissions are
estimated to annually contribute 2,200 additional work loss days,
13,000 additional reduced activity days, and an extra $103,000
(1991 $) in direct medical expenses. Table 7-3 summarizes these
statistics.
These damages resulting from current levels of coke oven
emissions can be monetized. As presented in Table 7-4, the
monetary value of estimated mortality, morbidity, and household
soiling effects in the absence of the proposed coke oven NESHAP
range from $3.4 million to $21.4 million (1991 $) annually.
Compliance with the MACT requirements (66% control from
baseline) results in benefits of $2.8 million to $18.0 million
that are annualized for the five year time period 1993 - 1998 at
a 10 percent interest rate. This benefit estimate considers all
batteries that must meet MACT for the LAER track and the two
Armco batteries that are expected to follow the MACT track. See
Table 7-5 for a summary of these results.
Compliance with the LAER requirements (90% control from
baseline inclusive of 66% MACT control) results in annualized
benefits fo $3.3 million to $20.7 million for the 15 year time
period 1998 - 2013. See Table 7-6 for a summary of these
results.
7-12
-------
TABLE 7-3.
EFFECTS OF BASELINE LEVELS OF COKE OVEN EMISSIONS
FROM DOORS, TOPSIDE, AND CHARGING
Number of Coke Oven Plants Subject to NESHAP
Number of People currently exposed*
Maximum Individual Mortality Risk
Expected Annual Excess Mortality
Number of Work Loss Days
Number of Reduced Activity Days / Year
Annual Direct Medical Expenses
28
23 Million
1.6 x 10*2
1.8
2,200
13,000
$103,000 (1991$)
NOTE:
'Underestimate because some counties are impacted by more than
one coke oven plant. The exposure model calculates exposure
levels at the sub-county level. In order to avoid double
counting, "Number of People Currently Exposed" is the sum of the
largest population in each county that is affected by any one
plant. This underestimates the total exposed population in any
county where one plant affects certain people, and other plants
affect additional people.
TABLE 7-4. ESTIMATED MONETARY VALUE OF DAMAGES DUE TO BASELINE
LEVELS OF COKE OVEN EMISSIONS FROM DOORS, TOPSIDE, AND CHARGING
Lower Bound
Mean Estimate
ANNUAL VALUE (1991 $)
Excess Mortality
Excess Morbidity
Soiling
TOTAL
$3.3 Million
$11,000
$47,000
$3.4 Million
$9.1 Million
$1.6 Million
$223,000
$10.9 Million
Upper Bound
$17.6 Million
$3.4 Million
$440,000
$21.4 Million
7-13
-------
TABLE 7-5. ESTIMATED MONETARY VALUE OF COMPLIANCE WITH
MACT REQUIREMENTS*
(66% Control of Baseline Emissions)
Lower Bound
Mean Estimate
Upper Bound
ANNUAL VALUE (1991 $)
Excess Mortality
Excess Morbidity
Soiling
TOTAL
$2.8 Million
$8,700
$39,000
$2.8 Million
$7.6 Million
$1.3 Million
$186,000
$9.1 Million
$14.7 Million
$2.9 Million
$366,000
$18.0 Million
* Due to incomplete coverage of effects categories, monetized
benefits may be understated.
Note:
Annualized benefits for time period 1993 - 1998.
Analysis considers 2 Armco batteries reported to follow MACT
track.
7-14
-------
TABLE 7-6. ESTIMATED MONETARY VALUE OF COMPLIANCE WITH
LAER REQUIREMENTS*
(90% Control of Baseline Emissions)
Lower Bound
Mean Estimate
ANNUAL VALUE (1991 $)
Excess Mortality
Excess Morbidity
Soiling
TOTAL
$3.2 Million
$10,000
$45,000
$3.3 Million
$8.8 Million
$1.5 Million
$216,000
$10.5 Million
Upper Bound
$17.0 Million
$3.3 Million
$424,000
$20.7 Million
* Due to incomplete coverage of effects categories, monetized
benefits may be understated.
Note:
Annualized benefits for time period 1998 - 2013.
Analysis excludes 2 Armco batteries reported to follow MACT
track.
7-15
-------
7.9 PROCESS UPSETS
As stated in Chapter 4, in addition to reducing emissions to
meet MACT and LAER, compliance with the Coke Oven NESHAP will
require plants to install flares to control emissions for process
upsets. While the occurrence of process upsets is considered
infrequent, the effects of the emissions can be significant in
that the amount of pollutants released in a short period of time
could result in acute exposures to toxic compounds including
known human carcinogens.
It was estimated in Chapter 5 that 470 Mg of BSO is emitted
nationwide each year from bypassed coke oven gases during process
upsets. In addition to BSO, many other pollutants not included
in this total would be controlled by the use of flares. These
pollutants include low molecular weight volatile organics (e.g.,
benzene, toluene) and inorganic gases (e.g., carbon monoxide,
sulfur dioxide) some of which are extremely toxic (e.g., hydrogen
sulfide, ammonia).
It is possible to estimate- some of the benefits of coke oven
emission reductions achieved by flares by transferring the
monetized benefit per Mg value determined from the coke oven
analysis described earlier in this chapter. This benefit value
represents the per unit value of decreased mortality from
exposure to BSO and decreased morbidity and household soiling
from particulate matter exposure. Other benefit categories are
not captured by this value.
For this analysis, the Clairton and Gary US Steel facilities
were excluded as flares have already been installed at these
batteries. By dividing total monetized benefits for the
remaining 26 facilities by total BSO emission reductions for
those 26 facilities, an average estimate of benefits per Mg of
BSO reduced can be obtained ($8,298 to $52,477 per Mg BSO (1991
$)) . The installation of flares to control process upset
releases is assumed to be 95 percent efficient. The benefit per
Mg range is transferred and applied to the flares scenario and
then multiplied by 95% control efficiency [($8,298 to $52,477) *
470 Mg of BSO emitted annually from process upsets * .95]. The
total benefit range is annualized over the 15 year analytic time
period to be consistent with the cost analysis. Total benefits
range from $4.2 million to $26.5 million (1991 $). Table 7-7
summarizes these results.
7-16
-------
TABLE 7-7. ESTIMATED MONETARY VALUE OF EMISSION
REDUCTIONS FOR FLARES*
(95% Control of Baseline Emissions)
Lower Bound
Mean Estimate
Upper Bound
ANNUAL VALUE (1991 $)
TOTAL
I $4.2 Million
$13.2 Million $26.5 Million
* Due to incomplete coverage of effects categories, monetized
benefits may be understated.
Note;
Monetized benefit estimates annualized over 15 year time
period 1993 - 2008.
7-17
-------
7.10 NON-QUANTIFIED BENEFITS
As described in the preceding sections, the quantitative
component of this benefit assessment covers cancer mortality due
to coke oven emission exposure and acute morbidity and household
soiling from particulate matter exposure. However, there are
many other potential benefit categories that are not covered by
this analysis because of significant data gaps. The following
discussion qualitatively describes some of these other benefit
categories that may apply to the coke oven regulation.
7.10.1 Cancer Mortality of Coke Oven Workers
Absent from this analysis is the benefit of mortality risk
reduction for coke oven workers. These individuals would
typically experience the highest exposures to coke oven emissions
due to working proximity to coke oven battery doors and lids.
Emission reductions achieved through this regulation would in
turn decrease worker mortality risk if all other variables remain
constant.
7.10.2 Non-cancer Mortality and Morbidity Due to Exposure to
Non-BSO or Non-PM Related Compounds
Benzene soluble organics and the particulate matter
components of coke oven emissions do not include all of the toxic
volatile organics, all of the inorganic gases, or all of the
trace metals that are also emitted from the coking process. Many
of these compounds have associated adverse non-cancer health
effects. For instance, carbon monoxide (CO) is 5.8% of coke oven
gas. CO brings about oxygen deficiency in the blood thus having
detrimental effects on the cardiovascular, central nervous,
pulmonary, and other body systems. Exposures to high
concentrations of CO could lead to death. Ammonia is a serious
skin, eye and membrane irritant. Ammonia releases resulted in
the second greatest number of injuries and deaths of all
chemicals in the acute hazard database. Hydrogen sulfide is
fatal at high concentrations, causing pulmonary edema and eye and
respiratory tract irritation. This compound is the leading cause
of accidental death in the workplace. Adverse health effects
from exposure to toxic metals ranges from brain and kidney damage
(mercury) to teratogenic effects (lead).
7.10.3 Adverse Environmental Impacts
There is scant information specifically on the environmental
impacts of coke oven emissions. However, there is data available
indirectly linking coke oven emissions with adverse environmental
effects such as information regarding the transport and
deposition of hazardous air pollutants from areas in which coke
ovens operate. For instance, atmospheric loading is estimated to
account for approximately 80 - 90% of all pollutant inputs to the
7-18
-------
upper Great Lakes, an area considered relatively pristine and
with few major sources of toxics. Similarly, short-range
atmospheric deposition is thought to be responsible for 90 - 99%
of lead inputs to the mid-lower Chesapeake Bay. Both of these
areas contain coke and steel production facilities that may be
contributing to toxic inputs to the two water bodies.11
A number of the chemical constituents of coke oven emissions
have a high potential to bioconcentrate and are also quite
persistent in .the environment.12 Ambient concentrations of these
compounds may be directly toxic to organisms while atmospheric
deposition of these compounds to land may have direct impacts on
individual terrestrial organisms and ecosystems as a whole.
It has also been shown that atmospheric deposition of
hazardous air pollutants, some of which are found in coke oven
emissions, contributes to adverse aquatic ecosystem effects.
Through the process of biomagnification, persistent compounds
accumulate in toxic concentrations in the tissues of species high
on the food chain leading to adverse impacts on wildlife and,
through subsequent ingestion of contaminated fish or waterfowl,
adverse health effects in humans. These adverse effects may pose
significant impacts on recreational and commercial fishing
industries.13
Although there is indirect evidence that coke oven emissions
are hazardous to the environment, it is unclear the extent to
which coke oven emissions contribute to adverse environmental
impacts and the degree to which this regulation will mitigate
these effects.
7-19
-------
REFERENCES
1. Memorandum from Elizabeth M. Ackerman, U.S. EPA Region III,
to Amanda Agnew, U.S. EPA, OAQPS, April 8, 1991.
2. U.S. Environmental Protection Agency. Industrial Source
Complex Dispersion Model User's Guide, Second Edition
(revised), Volume I. EPA-450/4-88-002a. December 1987.
3. Memorandum from Marvin Branscome, Research Triangle
Institute, to Amanda Agnew, U.S. EPA, OAQPS, March 27, 1992.
4. U.S. Department of Commerce, Bureau of Economic Analysis.
1985 OBERS BEA Regional Projections. 1985.
5. U.S. Environmental Protection Agency. User's Manual for the
Human Exposure Model (HEM)". EPA-450/5-86-001. June 1986.
6. Fisher, A., Chestnut, L. and Violette, D. "The Value of
Reducing Risks of Death: A Note on New Evidence", Journal of
Policy Analysis and Management 8(1), 1989, pp. 88-100.
7. Ostro, Bart D. "Air Pollution and Morbidity Revisited: A
Specification Test." Journal of Environmental Economics and
Management 14(1), 1987, pp. 87-98.
8. U.S. Environmental Protection Agency. Office of Air Quality
Planning and Standards. Review of the National Ambient Air
Quality Standards for Particulate Matter: Assessment of
Scientific and Technical Information. OAQPS Staff Paper.
EPA-450/5-82-001. Research Triangle Park, NC. January
1982.
9. MathTech, Inc. Benefits Analysis of Alternative Secondary
National Ambient Air Quality Standards for Sulfur Dioxide
and Total Suspended Particulates. Final report prepared for
U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards, Research Triangle Park, NC. 1982.
10. U.S. Environmental Protection Agency. National Air Quality
and Emissions Trends Report, 1988. EPA 450/4-4-90-002,
March, 1990.
7-20
-------
11. ICF, Inc. Atmospheric Deposition of Toxic Chemicals to
Surface Waters: Identification and Summary of Recent
Literature. Draft Report. Prepared for Office of Air
Quality Planning and Standards, U.S. EPA, Research Triangle
Park, NC August 1991.
12. ICF, Inc. Focus Chemicals for the Clean Air Act Amendments
Great Waters Study. Draft Report. Prepared for Office of
Air Quality Planning and Standards, U.S. EPA, Research
Triangle Park, NC. August 1991.
13. Environment Canada. Toxic Chemicals in the Great Lakes and
Associated Effects. Toronto, Ontario. March 1991.
7-21
-------
CHAPTER 8
BENEFIT-COST ANALYSIS
8.1 INTRODUCTION
This chapter presents a comparison of the incremental
benefits and incremental costs of the coke oven MACT and LAER
limits mandated in the CAAA and the flares requirement.
Additionally, the chapter discusses the rationale for the
proposed regulatory action.
In the course of internalizing the air pollution
externality, air quality regulations bring about a reallocation
of resources within the economy that affect society's well-being.
The cost of reducing coke oven emissions is reflected in the
production, distribution, and consumption of products affected by
the proposed Coke Oven NESHAP. This additional cost is in
contrast to the improvement in society's well-being from a
cleaner environment and concomitant reductions in adverse health
and welfare effects. Benefit-cost analysis is one vehicle that
provides a consistent framework for the evaluation of the
economic effects of alternative regulatory policies.
The efficiency criterion is one measure of the desirability
of the resource allocation (the optimum level of pollution
abatement) resulting from the'coke oven NESHAP. An allocatively
efficient regulation maximizes the positive net benefits to
society, which yields the optimal level of pollution control.
For this decision rule, the preferred regulatory option would be
the alternative that maximizes positive net benefits. However,
in addition to economic considerations, other factors, such as
political or statutory constraints, play a role in the decision
making process. Therefore the allocatively efficient regulatory
option is not always selected.
8.2 COMPARISON OF QUANTIFIED BENEFITS AND COSTS FOR COKE OVEN
DOORS, TOPSIDE, AND CHARGING MACT AND LAER REQUIREMENTS
Table 8-1 presents net quantified benefits for the MACT and
LAER requirements mandated by the CAAA assuming 64 batteries
follow the LAER track and two batteries follow the MACT track. A
range of benefit estimates is used that reflects the lower bound,
8-1
-------
mean estimate, and upper bound annual values as presented in
Chapter 7. The cost data used are EPA's annualized point
estimates for both the MACT ($25 million) and LAER ($46 million)
compliance tracks. Monetized benefits presented in the table do
not include the impact of the non-quantified benefits discussed
in Chapter 7.
The nationwide annual costs minus quantified benefits of
implementing the MACT standards for doors, topside and charging
are -$7.0 to -$22.2 million. The nationwide annual costs minus
quantified benefits of implementing the doors, topside, and
charging LAER standard are -$25.3 million to -$42.7 million.
Recognize that incomplete coverage of benefits precludes a
definitive statement on the allocative efficiency aspects of
these two requirements.
TABLE 8-1. TOTAL ANNUAL COSTS, BENEFITS*, AND
NET BENEFITS OF COKE OVEN MACT AND LAER REQUIREMENTS
(1991 $)
ANNUAL COST
ANNUAL BENEFIT*
Low
Mean
High
QUANTIFIED BENEFITS
MINUS COST
Low
Mean
High
MACT
$25 million
$2.8 million
$9.1 million
$18.0 million
($22.2 million)
($15.9 million)
($7.0 million)
LAER
$46 million
$3.3 million
$10.5 million
$20.7 million
($42.7 million).
($35.5 million)
($25.3 million)
* Quantified benefits only; benefits coverage is incomplete.
8-2
-------
8.3 COMPARISON OF QUANTIFIED BENEFITS AND COSTS FOR COKE OVEN
FLARE CONTROL
The nationwide cost for installation, operation and
maintenance of flares for coke oven batteries at 26 facilities is
estimated to be $2.8 million annually (1991 $). Net benefits
(quantified benefits minus cost) for the coke oven flare
requirement are positive. They range from $1.4 million to $23.7
million annually (1991 $). This suggests an improvement in
society's well-being on allocative efficiency grounds despite
incomplete quantification of the benefits. See Table 8-2 for a
summary of these results.
TABLE 8-2. TOTAL ANNUAL COSTS, BENEFITS*, AND NET BENEFITS FOR
COKE OVEN FLARE REQUIREMENTS
(1991 $)
ANNUAL COST
ANNUAL BENEFIT*
Low
Mean
High
QUANTIFIED BENEFIT
MINUS COST*
Low
Mean
High
FLARES - 95% CONTROL
$2.8 million
$4.2 million
$13.2 million
$26.5 million
$1.4 million
$10.4 million
$23.7 million
* Quantified benefits only; benefits coverage is incomplete.
Note; Costs and benefits annualized over 15 year time period
1993 - 2008.
8-3
-------
8.4 RATIONALE FOR THE PROPOSED REGULATORY ACTION
As discussed in Chapter 2, there are alternative ways to
internalize a negative externality. Economists acting as third
parties could assess the benefits and costs of regulations in an
attempt to identify the optimum level of pollution control for
the regulation. Alternatively, parties directly affected by the
negative externality could negotiate and develop the regulation
to reduce air pollution.
This more direct approach of negotiation is sometimes
preferred on economic efficiency grounds. This is especially the
case when identification of affected parties is not difficult and
the transactions costs of their participation in the negotiations
is modest.
The EPA entered into regulatory negotiations with industry,
environmental groups, and state agencies in the development of
the proposed coke oven regulation. The CAAA set default limits
of 3 PLD, 1 PLL, 4 PLO, and 16 seconds charging, and required for
the MACT track, MACT compliance by 1995 or for the LAER track
MACT compliance by 1993 and LAER compliance by 1998.
Regulatory negotiations resulted in decisions on control levels
that differ somewhat from the default limits for doors, topside,
and charging (See Chapter 4).
8.5 CONCLUSIONS
Estimates of potential quantified benefits, costs, and net
benefits for the Coke Oven doors, topside, and charging MACT and
LAER requirements are presented in Table 8-1. The upper bound of
quantified benefits do not outweigh the costs associated with
this component of the regulation. Because the benefits were
quantified for only three benefit categories, EPA cannot conclude
that net benefits are necessarily negative (net cost). The net
cost as reflected in this analysis could be reduced, however, if
some portion of the non-quantified benefits presented in Chapter
7 were quantified. For example, reductions in adverse human
welfare effects (worker productivity losses and odors) and non-
human biological effects (agriculture, forestry,
recreational/commercial fishing and ecosystem impacts) could
potentially yield higher benefit estimates if quantified.
There are other potential benefit categories associated with
the particulate matter component of coke oven emissions beyond
acute morbidity and household soiling such as chronic morbidity,
industrial and commercial sector soiling and materials damage,
and ecosystem impacts.
There are other compounds that are emitted by coke ovens,
yet are not covered by BSO or PM, that are highly toxic to humans
8-4
-------
and the environment. Due to lack of data, the benefits of
controlling compounds such as hydrogen sulfide, benzene, ammonia,
and the trace metals have not been included in this analysis.
Estimates of quantified benefits, costs, and net benefits
for compliance with the flares requirement are presented in Table
8-2. Annual benefits outweigh annual costs by a factor of 1.5 to
9.5.
Major conclusions of this analysis:
• With incomplete coverage of the benefits, the
allocative efficiency aspects of the MACT and LAER
requirements are indeterminant:
complete coverage of benefits (i.e.,
quantification of all benefit categories) may or
may not result in positive net benefits
The ratios of monetized benefits to the costs for
doors, topside, -and charging MACT requirements
range from .11:1 to .72:1.
The ratios of monetized benefits to the costs for
doors, topside, and charging LAER requirements
range from .07:1 to .45:1.
• Despite incomplete coverage of the benefits, compliance
with the coke oven flare requirement results in
allocative efficiency gains.
Net benefits for the flares requirement are
positive.
The ratios of monetized benefits to estimated
costs for flares range from 2:1 to 10:1.
8-5
-------
APPENDIX A
COST COMPONENTS FOR EACH BATTERY
-------
APPENDIX A.
COSTS COMPONENTS FOR EACH BATTERY
This appendix identifies cost components provided by
specific plants through the American Iron and Steel Institute and
the American Coke and Coal Chemicals Institute. For plants that
did not provide alternative information, the generic category
used is identified for each battery. This information is
documented in the report for the Cost Work Group.14
1. ABC Coke; The company plans to install a spare larry
car for Battery A. (This cost was not attributed to the NESHAP.)
For LAER, Battery A was placed in Category E and Batteries 5 and
6 were placed in Category A.
2. Acme Steel; To meet MACT, the plant plans to install
new doors, a door cleaner, improved door spotting equipment, new
offtake caps, larry car modifications, and additional training
with a total estimated cost of $3.9 million. If the company
chooses to meet the LAER limits; they estimate an additional cost
of $0.6 million for jamb cleaners and improved back pressure
controls.
[Note: Door leak data for this plant show that the two
batteries have averaged 0.6 and 0.8 PLD from 21 observations over
7 years. Consequently, this plant may already be able to meet
the MACT and LAER limits for percent leaking doors.]
3. Armco. Middletown, OH; The company plans to meet MACT
in 1995 for Battery 3 at a cost of $19 million. This cost
includes the installation of modified doors, replacement of one-
third of the jambs and buckstays, ceramic welding behind jambs,
anti-hourglassing modifications, resetting one-third of the
charging hole castings, new standpipe valves, a new larry car,
rebuild of existing larry car for use as a spare, upgrade pusher,
new door machine, rebuild of existing door machine, new quench
car, new backpressure controls, and by-product plant
modifications. Batteries 1 and 2 were placed in Category F
(rebuild) for LAER.
4. Armco. Ashland. KY; The company plans to meet MACT in
1995 for Battery 3 at a cost of $17.5 million. This cost
includes the installation of new doors, new jambs, replacement of
12 buckstays, anti-hourglassing modifications, reset charging
hole castings, new and rebuilt larry cars, upgrade pusher, new
and rebuilt door machines, new quench car, door cleaner, and new
steam aspiration. To meet LAER with Battery 3, the company plans
A-l
-------
to spend $16.5 million over the next few years for new doors,
jambs, upgraded or new equipment, and repairs. An additional $30
to $35 million will probably be needed for through-wall and end
flue repairs around the year 2000.
5. Bethlehem Steel, Bethlehem, PA: MACT costs for Battery
A include new drop sleeves ($200,000) and new slip joints for the
offtakes ($800,000). MACT costs for Batteries 2 and 3 include
new slip joints ($1 million) and the conversion of the pusher
side self-sealing doors to hand-luted doors ($4 million). For
LAER, 120 new jambs will be installed on Battery A ($3 million).
Batteries 2 and 3 will probably need to be rebuilt to meet LAER.
[NOTE: The costs for drop sleeves and slip joints may be
associated with meeting the current regulations for offtakes and
charging, which are as stringent as the default MACT limits.
These costs are included in the industry estimate and are not
included in the EPA estimate as directly attributable to the
NESHAP.]
6. Bethlehem Steel. Burns Harbor. IN; The costs for these
two batteries to meet MACT includes replacement/repair of
standpipes ($2.4 million), larry car modifications ($300,000),
gooseneck cleaners ($200,000), new jamb cleaners and improved
door cleaners ($2 million), jamb replacement ($1.1 million), new
door seals ($2 million), buckstay replacement on Battery 2 ($6.3
million), and additional door repair capability/equipment ($3.5
million). These batteries will probably be rebuilt to meet LAER.
[NOTE: The MACT capital investment must be recovered in 5
years (the capital recovery factor is 0.264 at 10 percent), which
results in an annualized cost of $5 million/yr for this plant to
meet MACT. In addition, this plant may be rebuilt to meet LAER.
Only the rebuild cost is used in the EPA estimate and both costs
are used in the industry estimate.]
7. Bethlehem Steel, Lackawanna. NY: To meet MACT, these
two batteries will require larry car modifications ($750,000),
new door seals ($500,000), new door plugs ($1.6 million), and
improved back pressure control ($500,000). To meet LAER, new
doors, new jambs, and automatic door and jamb cleaners may be
required.
8. Citizens Gas and Coke; Batteries E and H may require
larry car modifications to meet MACT for charging. Battery 1 is
placed in Category B and Batteries E and H are placed in Category
A for LAER costs. An automatic door cleaner is planned for the
coke side of Batteries E and H.
A-2
-------
9. Empire Coke: The costs to meet MACT include an
additional person per shift, modifications to the larry car
($750,000), exhauster ($200,000), and an additional person for
monitoring. For LAER, an additional person per shift would be
required.
10. Erie Coke; These batteries are assigned to Category A
for LAER costs. In addition, 3 to 4 walls must be repaired at a
cost of $25,000 to $50,000 each.
11. Geneva Steel; These batteries are assigned to
Category C for LAER costs.
12. Gulf States steel; To meet MACT, the company expects
to install a new larry car ($4 million), repair/replace doors
($1.5 million), and hire 5 new people. For LAER, these batteries
were assigned to Category F.
13. Inland Steel; Batteries 6, 7, 9, and 10 will shut down
in 1994, and Battery 11 will shut down in 1992. No costs are
attributed to the NESHAP.
*
14. Koppers; The estimated cost to meet MACT includes
$400,000/yr for operating labor, $200,000/yr for door maintenance
labor, $100,000 for improved back pressure control, and $350,000
for updating equipment. These batteries are placed in Category A
for LAER costs.
15. LTV, Cleveland, OH; Battery 6 is assigned to Category
E and Battery 7 is assigned to Category F for LAER costs.
16. LTV. Pittsburgh, PA; These batteries are assigned to
Category E for LAER costs. The cost of coke side door and jamb
cleaners should be estimated as $5 to 7 million because of site-
specific conditions (insufficient structural support for the
cleaners).
17. LTV, Chicago; No costs are estimated for this battery
to meet LAER (it is among the best-performing tall batteries for
the control of door leaks).
18. LTV. Warren, OH; This battery is assigned to Category
D for LAER costs.
19. National Steel, Ecorse. MI; This battery is assigned
to Category B for LAER costs.
20. National steel. Granite city. IL; Battery A is
assigned to Category D and Battery B is assigned to Category B
for LAER costs.
A-3
-------
21. New Boston Coke; To meet MACT, the company expects to
replace offtakes ($1.4 million), install a new steam system and
jumper pipes ($2 million), and install new doors and jambs
($2 million). The costs to meet LAER include a new jamb cleaner
($450,000), door cleaner ($250,000), spotting device ($50,000),
and jamb/end flue repairs ($2 million).
22. Sharon Steel; These batteries are assigned to
Category C for LAER costs.
23. Shenanqo; One person per shift is added for topside
leaks for Battery 4. Battery 1 is assigned to Category B and
Battery 4 is assigned to Category A for LAER costs.
24. Sloss Industries; These batteries are assigned to
Category D for LAER costs.
25. Toledo Coke; This battery is assigned to Category F
for LAER costs.
26. Tonawanda Coke; This battery is assigned to Category B
for LAER costs.
27. USX, Clairton. PA; For LAER, include automatic door
and jamb cleaners on the coke side, the replacement of 50 percent
of the jambs, and additional labor for the use of a supplemental
sealant for Batteries 1, 2, 3, 7, 8, 9, and 19. For Batteries
13, 14, 15, 20, and B, assign Category B. In addition, include
coke side door and jamb cleaners on Batteries 13, 14, 15, and 20.
28. USX, Gary, IN; Batteries 5 and 7 are assigned to
Category D and Batteries 2 and 3 are assigned to Category E for
LAER costs.
29. Wheeling-Pittsburgh; To meet MACT for doors on
Batteries 1, 2, and 3, add 2 new door machines and automatic door
cleaners. Also include automatic lid replacement for the last
lid. To estimate LAER costs, assign Batteries 1, 2, and 3 to
Category D. For Battery 8, include replacement of jambs and
rehabilitation of doors at $18 million every 8 years as LAER
costs. The company plans to spend $10.5 million to replace half
of the jambs and to install water-sealed standpipe caps as part
of their current operation. The cost for jambs is annualized
over 8 years for the industry estimate and over 20 years for the
EPA estimate.
30. Bethlehem Steel, Sparrows Point. MD; These batteries
are shut down and are not included in the cost analysis.
A-4
-------
APPENDIX B
COST ESTIMATES FOR EACH BATTERY
-------
TABLE B-1. SUMMARY OF MACT COSTS (1991 DOLLARS)
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
NUMBER A
OF P
PLANT BATTERY BATTER LJ
UNITS I
ABC Coke, Tanant, AL
Acme Steel, Chicago, IL
Armco Inc., Middetown, OH
Atmoo Inc. Ashland, KY
Bethlehem Steel, Bethlehem, PA
Bethlehem Steel, Bums Harbor, IN
Bethlehem Steel, Lackawanna, NY
Citizens Gas, Indianapolis, IN
Empire Coke, Holt AL
Erie Coke, Erie, PA
Geneva Steel, Provo, UT
Gulf States Steel, Gadsden, AL
Inland Steel, East Chicago, IN
Koppers, Woodwaid, AL
LTV Steel, Cleveland, OH
A
5
6
1
2
1
2
3
3
4
A
2
3
1
2
7
8
E
H
1
1
2
A
B
1
2
3
4
2
3
6
7
9
10
11
1
2A
2B
4
5
6
7
1.0
0.5
0.5
0.5
0.5
0.5
0.5
1.0
1.0
1.0
1.0
0.5
0.5
1.0
1.0
1.0
1.0
0.5
0.5
1.0
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
1.0
1.0
.0
.0
.0
.0
.0
1.0
0.5
0.5
0.5
0.5
1.0
1.0
VERAGE
ERCENT
EAKING
XDORS
3.6
6.2
6.2
0.6
0.8
15.5
15.5
8.7
18.0
16.0
5.7
5.2
5.1
8.6
9.7
13.4
13.8
3.0
3.1
4.4
6.3
5.9
7.6
7.6
3.6
2.4
2.8
3.6
13.3
7.3
5.6
8.9
9.1
1Z1
8.7
1.4
0.0
1.7
1.4
0.6
6.5
112
MACT COSTS -EPA
OPRTNG
Z6E+05
1.5E+05
1.5E+05
1.5E+05
1.5E+05
3.9E+05
3.9E+05
3.6E+05
6.7E+05
4.6E+O5
1.5E+05
9.1E+04
9.1E+04
3.6E+05
4.0E+05
5.1E+05
5.3E+05
1.5E+05
1.5E-t05
Z6E+05
1.5E+05
1.5E+05
1.1E+05
1.2E+05
9.1E+O4
9.1E+04
9.1E+04
9.1E+04
1.7E+05
1.7E+05
CAPITA
O.OE+00
O.OE+OO
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
1.9E+07
1.8E+07
O.OE+00
O.OE+00
ZOE+06
ZOE+06
O.OE+00
O.OE+00
1.7E+06
1.7E+06
4.3E+05
4.3E+05
O.OE+00
7.8E+05
7.8E+05
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
Z8E+06
2.8E+06
TOTAL
ANNUAL
Z6E+05
1.5E+05
1.5E+05
1.5E+05
1.5E+05
3.9E+05
3.9E+05
3.4E+06
3.5E+06
4.6E+05
1.5E+05
62E+05
6.2E+05
3.6E+05
4.0E+05
7.1E+05
72E+05
2.0E+05
2.0E+05
Z6E+05
2.4E+05
2.4E+05
1.1E+05
12E+05
9.1E+04
9.1E+04
9.1E+04
9.1E+04
4.9E+05
4.9E+05
MACT COSTS INDUSTRY
OPRTNG
Z6E+05
1.5E+05
1.5E+O5
1.5E+O5
1.5E+05
Z7E+05
2.7E+05
3.2E+05
Z6E+O5
1.5E+05
Z1E+O5
1.7E+05
1.7E+05
3.8E+05
3.8E+05
Z6E+05
Z6E+O5
Z2E+05
Z1E+05
2.6E+05
1.5E+05
1.5E+05
1.1E+05
12E+05
9.1E+04
9.1E+04
9.1E+04
9.1E+04
1.7E+05
1.7E+05
CAPITA
3.1E+06
O.OE+OO
O.OE+OO
ZOE+O6
2.0E+06
O.OE+00
O.OE+00
1.9E+07
1.8E+07
O.OE+00
1.0E+06
2.5E+06
2.5E+06
6.3E+06
1J2E+07
1.7E+06
1.7E+06
4.3E+05
4.3E+05
O.OE+00
7.8E+O5
7.8E+05
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
2.8E+06
2.8E+06
TOTAL
ANNUAL
6.2E+05
1.5E+05
1.5E+05
3.8E+05
3.8E+05
Z7E+05
Z7E+05
3.4E+06
3.1E+06
1.5E+05
3.3E+05
8.3E+O5
8.3E+05
ZOE+06
3.4E+06
4.5E+OS
4.5E+05
2.7E+05
2.6E+05
2.6E+05
24E+05
Z4E+05
1.1E+05
1^E+05
9.1E+04
9.1E+04
9.1E+04
9.1E+O4
4.9E+05
4.9E+05
PLAN TO SHUT DOWN
Z7E+05
Z1E+05
Z7E+05
Z7E-tO5
Z7E-t05
iSE+OS
41E-tO5
9.0E+04
9.0E+04
3.9E+05
3.9E+OS
3.9E+05
O.OE+00
OOE+00
2.8E+05
22E+05
3.1E+05
3.1E+O5
3.1E+05
2.8E+05
4.1E+05
Z7E+05
Z1E+05
Z7E+05
Z7E+05
Z7E+05
3.1E+05
31E+O5
9.0E+04
9.0E+04
3.9E+05
3.9E+05
3.9E+05
O.OE+00
O.OE+00
2.8E+O5
Z2E+O5
3.1E+05
3.1E+O5
3.1E+05
3.1E+05
3.1E+05
B-2
-------
TABLE B-1. SUMMARY OF MACT COSTS (1991 DOLLARS)
NUMBER AVERAGE
No.
16
17
18
19
20
21
22
23
24
25
26
27
28
29
OF PERCENT
PLANT BATTERY BATTER LEAKING
UNITS DOORS
LTV Steel, Pittsburgh, PA
LTV Steel, Chicago, IL
LTV Steel, Waron, OH
National Steel, Ecorse, Ml
National Steel, Granite City, IL
New Boston, Portsmouth, OH
Sharon Steel, Monessen, PA
•• •
Shenango, Pittsburgh, PA
Sloss Industries, Biimingham, AL
Toledo Coke, Toledo, OH
Tonawanda, Buffalo, NY
USX, Clairton, PA
USX, Gary, IN
Wheeling-Pitt, East Steubenville,
WV
TOTALS
P1
P2
P3N
P3S
P4
2
4
5
A
B
1
1B
2
1
4
3
4
5
C
1
1
2
3
7
8
9
13
14
15
19
20
B
2
3
5
7
1
2
3
8 .
82
0.5
0.5
0.5
0.5
1.0
1.0
1.0
1.0
0.5
0.5
1.0
0.5
0.5
1.0
1.0
0.5
0.5
1.0
1.0
1.0
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
1.0
1.0
1.0
1.0
1.0
0.5
0.5
0.7
0.7
0.7
1.0
61.0
5.7
7.1
7.0
7.8
6.9
3.4
4.9
Z8
4.9
5.4
3.0
3.1
5.4
7.1
5.5
4.5
1.9
9.7
2.2
5.8
6.3
7.0
4.8
6.8
7.3
3.1
3.2
3.5
5.9
3.9
3.0
9.0
9.4
6.0
5.1
4.5
5.7
4.6
6.4
MACT COSTS -EPA
OPRTNG
9.1E+04
1.3E405
1.2E+05
1.5E+05
1.9E+O5
1.5E+Q5
1.5E+05
3.7E+04
1.5E+05
3.7E+04
2.6E+05
9.1E+04
9.1E+04
3.7E+04
Z8E+05
1.5E+05
1.5E+05
2.6E+05
3.6E-fO5
2.6E+05
1.1E+05
1.3E+05
1.5E+05
1.1E+05
1.4E+05
1.5E+05
3.7E+04
3.7E+04
3.7E+04
1.5E+05
3.7E+04
3.7E+O4
3.4E+05
3.5E+05
1.6E+05
1.5E-t05
1.8E+05
1.8E405
1.8E+05
Z8E+05
1.5E+07
CAPITA
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
5.4E+06
O.OE+00
O.OE+00
O.OE+00
O.OE+00
3.0E+O5
3.0E+05
6.0E+05
6.0E+05
6.0E+05
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
1.6E+06
1.6E+06
1.6E+OS
O.OE+00
6.6E+07
TOTAL
ANNUAL
9.1E+04
1.3E+O5
1.2E+05
1.5E+05
1.9E+05
1.5E+05
1.5E+05
3.7E+04
1.5E+05
3.7E+O4
8.9E+05
9.1E+04
9.1E+O4
3.7E+04
2.8E+05
1.8E+05
1.8E+05
3.3E+05
4.3E+05
3.3E+05
1.1E+05
1.3E+05
1.5E+05
1.1E+05
1.4E+05
1.5E+05
3.7E+04
3.7E+04
3.7E+04
1.5E+05
3.7E+04
3.7E+04
3.4E+05
3.5E+05
1.6E+05
1.5E+05
3.8E+05
3.8E+O5
3.8E+05
Z8E+05
2.5E+07
MACT COSTS - INDUSTRY
OPRTNG
1.4E+05
1.4E+O5
1.4E+O5
1.4E+05
Z1E+O5
1.5E+05
Z1E+05
3.7E+O4
1.5E+05
3.7E+O4
3.1E+05
1.2E+05
1.1E+05
3.7E+04
2.8E+05
2.0E+05
2.0E+05
3.7E+05
3.0E+05
2.6E+06
1.6E+O5
1.6E+O5
1.6E+05
1.6E+05
1.6E+O5
1.6E+05
3.7E+04
3.7E+O4
3.7E+O4
2.1E+05
3.7E+O4
3.7E+O4
3.4E+05
3.4E+05
2.6E+05
2.6E+O5
2iE+05
Z2E+05
23E+05
3.2E+05
1.5E+07
CAPITA
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
5.4E+06
O.OE+00
O.OE+00
O.OE+00
O.OE+00
3.0E+05
3.0E+O5
6.0E+05
6.0E+05
6.0E+O5
O.OE+00
O.OE+00
O.OE+00
O.OE+OO
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
1.6E+O6
1.6E+06
1.6E+O6
1.1E+07
1.0E+O8
TOTAL
ANNUAL
1.4E+06
1.4E+O5
1.4E+05
1.4E+05
Z1E+O5
1.5E+05
2.1E+05
3.7E+04
1.5E+05
3.7E+04
9.5E+05
1.2E+05
1.1E+05
3.7E+04
2.8E+05
2.4E+05
Z4E+05
4.4E+05
3.7E+05
3.3E+05
1.6E+05
1.6E+05
1.6E+05
1.6E+05
1.6E+05
1.6E+05
3.7E+04
3.7E+04
3.7E+04
Z1E+05
3.7E+04
3.7E+04
3.4E+05
3.4E+05
Z6E+05
Z6E+05
4.1E+05
4.1E+05
4ZE+05
2£E+06
3.3E+07
B-3
-------
TABLE B-2. SUMMARY OF LAER COSTS (1991 DOLLARS)
LAER COSTS
EPA INDUSTRY PADUP BYPRODUC TOTAL
No. PLANT
1 ABC Coke, Tanant, AL
2 Acme Steel, Chicago, IL
BATTERY
A
5
6
1
2
CAPITAL
1.5E+07
O.OE+00
O.OE+00
3.0E+05
3.0E+05
TOTAL
ANNUAL
2.3E+06
Z6E+05
2.6E+05
1.7E+05
1.7E+05
CAPITA
Z1E+07
O.OE+00
O.OE+00
3.0E+O5
3.0E+O5
REBUILD PLANT
TOTAL TRIGGERED REPAIRS
ANNUAL EARLY
3.1E+06
Z6E+05
2.6E+05
4.0E+05
4.0E+O5
CAPITAL
COST
ESTIMATE
Z1E+07
O.OE+00
O.OE+00
3.0E+05
3.0E+O5
3 Armco Inc., MkJdetown, OH
1 O.OE+00 Z8E+05 O.OE+00 Z8E+05 6.4E+O7
2 O.OE+00 2.8E+05 O.OE+00 2.8E+05 6.4E+07
3 NOT APPLICABLE
3.3E+06 6.8E+07
3.3E+06 6.8E+07
3.3E+O6 3.3E+06
4
5
6
7
8
9
10
11
12
13
14
15
Aimco Inc., Ashland, KY
Bethlehem Steel, Bethlehem, PA
Bethlehem Steel, Bums Hartsor, IN
Bethlehem Steel, Lackawanna, NY
Citizens Gas, Indianapolis, IN
Empire Coke, Holt AL
Erie Coke, Erie, PA
Geneva Steel, Pnovo, UT
Gulf States Steel, Gadsden, AL
Inland Steel, East Chicago, IN
Koppere, Woodwaid, AL
LTV Steel, Ctewland, OH
3
4
A
2
3
1
2
7
8
E
H
1
1
2
A
B
1
2
3
4
2
3
6
7
9
10
11
1
2A
2B
4
5
6
7
NOT APPLICABLE
1.7E+07 2.4E+06
3.0E+06
O.OE+00
O.OE+OO
O.OE+00
O.OE+00
4.3E+06
4.3E+06
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
1.0E+05
1.0E+05
O.OE+00
O.OE+00
o.oE+oo
0.06+00
O.OE+00
O.OE+00
7.7E+05
2.3E+05
2.3E+05
5.3E+05
5.3E+05
1.2E+06
1.2E+06
3.1E+05
3.1E+05
5.3E+05
3.5E+05
3.5E+05
22E+05
2.2E+05
2.3E+05
2.3E+05
Z3E+05
2.3E+OS
7.6E+05
7.6E+05
4.9E+07
3.0E+06
O.OE+00
O.OE+00
O.OE+00
O.OE+00
4.3E+06
4.3E+06
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
1.0E+05
1.0E+05
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
5.0E+06
8.9E+05
Z3E+05 8.4E+07
^3E+05 8.4E+07
5.3E+05 9.5E+07
5.3E+05 9.5E+07
12E+06
1.2E+06
3.1E+05
3.1E+05
5.3E+05
3.5E+05
3.5E+05
22E+05
2^E+05
2.3E+05
Z3E+05
Z3E+05
2.3E+05
7.6E+05 5.8E+07
7.6E+05 5.8E+07
3.3E+06
3.3E+06
3.3E+06
5.0E+06
5.0E+06
5.0E+06
5.0E+06
4.9E+07
6.3E+06
8.7E+07
8.7E+07
1.0E+08
1.0E+08
4.3E+06
4.3E+06
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
1.0E+05
1.0E+05
O.OE+00
O.OE+00
O.OE+00
O.OE+00
6.3E+07
6.3E+07
PLAN TO
SHUT
DOWN
IN 1994
PLAN TO SHUT DOWN IN 1992.
O.OE+00
0.06+00
0.06+00
O.OE+00
0.06+00
7.86+06
0.06+00
5.0E+05
3.3E+05
42E+05
4^E+O5
42E+05
1.4E+06
5.3E+05
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
1.4E+07
O.OE+00
5.0E+05
3.3E+05
42E+05
42E+05
42E+05
ZOE+06
5.3E+05 5.6E+07
5.0E+06
5.0E+06
O.OE+00
O.OE+00
O.OE+00
O.OE+00
O.OE+00
1.9E+07
6.1E+07
B-4
-------
TABLE B-2. SUMMARY OF LAER COSTS (1991 DOLLARS)
LAER COSTS
EPA INDUSTRY PADUP BYPRODUC
No.
16
17
18
19
20
21
22
23
24
25
26
27
28
29
PLANT BATTERY
LTV Steel, Pittsburgh, PA
LTV Steel, Chicago, IL
LTV Steel, Warren, OH
National Steel, Ecorse, Ml
National Steel, Granite City, IL
New Boston, Portsmouth, OH
Sharon Steel, Monessen, PA
Shenango, Pittsburgh, PA
Sloss Industries, Birmingham, AL
Toledo Coke, Toledo, OH
Tonawanda, Buffalo, NY
USX dairton, PA
USX, Gary, IN
Wheeling-Pitt, East Steubenvifle,
WV
TOTALS
P1
P2
P3N
P3S
P4
2
4
5
A
B
1
18
2
1
4
3
4
5
C
1
1
2
3
7
8
9
13
14
15
19
20
B
2
3
5
7
1
2
3
8
82
CAPITAL
7.6E+06
7.6E+06
7.6E+06
7.6E+06
9.8E+06
O.OE+00
7.0E+05
O.OE+00
7.0E+05
O.OE+OO
2.8E+06
O.OE+00
O.OE+00
O.OE+00
O.OE+00
3.5E+05
3.5E+05
7.0E+05
O.OE+00
O.OE+00
1.2E+06
1.2E+06
1.2E+06
1.2E+06
1.2E+06
1.2E+06
4.7E+05
4.7E+05
4.7E+05
1.7E+06
7.0E+05
O.OE+00
1.1E+07
1.1E+07
3.5E+05
3.5E+05
INCLUDED
IN
MACT
1.8E+07
1.49E+08
TOTAL
ANNUAL
.1E+06
.1E+06
.1E+06
.1E+06
.6E+06
1.5E+05
5.0E+05
3.1E+05
3.7E+05
1.7E+05
1.5E+06
2.3E+05
2.3E+05
3.1E+05
4.7E+05
3.6E+05
3.6E+05
6.8E+05
6.0E+05
6.0E+05
4.4E+05
4.4E+05
4.4E+05
4.4E+05
4.4E+05
4.4E+05
Z8E+05
Z8E+05
2.8E+05
6.2E+05
3.9E+05
3.1E+05
1.8E+06
1.8E+06
3.2E+05
3.2E+05
5.6E+05
5.6E+05
5.6E+05
2.6E+06
4.6E+07
CAPITA
1.4E+07
1.4E+07
1.4E+07
1.4E+07
1.6E+07
O.OE+00
7.0E+05
O.OE+00
7.0E+05
O.OE+00
Z8E+06
O.OE+00
O.OE+00
O.OE+00
O.OE+00
3.5E+05
3.5E+05
7.0E+05
O.OE+00
O.OE+00
1.2E+06
1.2E+06
1.2E+06
1.2E+06
1.2E+06
1.2E+06
4.7E+05
4.7E+05
4.7E+05
1.7E+06
7.0E+05
O.OE+00
1.7E+07
1.7E+07
3.5E+05
3.5E+05
O.OE+00
O.OE+00
O.OE+00
1.8E+07
2.4E+08
REBUILD PLANT
TOTAL TRIGGERED REPAIRS
ANNUAL EARLY
1.6E+06
1.6E+06
1.6E+06
1.6E+06
Z1E+06
1.5E+05
5.0E+05
3.1E+05
3.7E+05
1.7E+05
1.5E+06
2.3E+05
^3E+05
3.1E+06
4.7E+05
3.6E+05
3.6E+05
6.8E+05
6.0E+05 5.1E+07 1.0E+07
6.0E+05
4.4E+05
4.4E+05
4.4E+05
4.4E+05
4.4E+05
4.4E+05
Z8E+05
28E+05
Z8E+05
62E+05
3.9E+05
3.1E+05
Z3E+06
Z3E+06
32E+05
32E+05
5.6E+05
5.6E+05
5.6E+O5
5.8E+06
5.7E+07 7.09E+08 6.00E+07
TOTAL
CAPITAL
COST
ESTIMATE
1.4E+07
1.4E+07
1.4E+07
1.4E+07
1.6E+07
O.OE+00
7.0E+05
7.0E+05
O.OE+00
Z8E+06
O.OE+00
O.OE+00
O.OE+00
O.OE+00
3.5E+05
3.5E+05
7.0E+05
6.1E+07
O.OE+00
1.2E+06
12E+06
12E+06
12E+06
12E+06
12E+06
4.7E+05
4.7E+05
4.7E+05
1.7E+06
7.0E+05
O.OE+00
1.7E+07
1.7E+07
3.5E+06
3.5E+05
O.OE+00
O.OE+00
O.OE+00
1.8E+07
1.01E+09
B-5
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA 453/D-92-014
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Regulatory Impact Analysis of National Emission
Standards for Hazardous Air Pollutants for By-Product
Coke Oven Charging, Door Leaks, and Topside Leaks
5. REPORT DATE
November 1992
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Director
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
DRAFT
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Under the authority of the 1990 Clean Air Act Amendments, a National Emissions
Standard for Hazardous Air Pollutants (NESHAP) is proposed to control emissions from
By-product Coke Oven Charging, door leaks, and topside leaks. Because the EPA consider
the regulation for By-product Coke Oven batteries to be a "major" rule, the attached
Regulatory Impact Analysis was prepared to fulfill the requirements of E012291. This
document reviews the need for regulation, control techniques, regulatory options, costs
of control, economic impacts, benefits of the regulation, and compares benefits and
costs associated with the regulation.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATl Field/Group
Air Pollution
Coke Oven
Hazardous Air Pollutant
Emission Controls
Economic Impact
Benefit
Cost
13B
8. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport!
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (R»». 4-77) PREVIOUS EDITION is OBSOLETE
-------
INSTRUCTIONS
1. REPORT NUMBER
Insert the EPA report number as it appears on the cover of the publication.
2. LEAVE BLANK
3. RECIPIENTS ACCESSION NUMBER
Reserved for use by each report recipient.
4. TITLE AND SUBTITLE
Title should indicate clearly and briefly the subject coverage of the report, and be displayed prominently. Sol subtitle, if used, in smaller
type or otherwise subordinate it to main title. When a report is prepared in more than one volume, repeal the primary title, add volume
number and include subtitle for the specific title.
5. REPORT DATE
Each report shall carry a date indicating at least month and year. Indicate the hasis on which il was selected (e.g.. Jatc of issue, date of
approval, date of preparation, etc.).
6. PERFORMING ORGANIZATION CODE
Leave blank.
7. AUTHOR(S)
Give namc(s) in conventional order (John R. Doc. J. Robert Doc. etc.). List author's affiliation if it differs from (ho performing organi-
zation.
8. PERFORMING ORGANIZATION REPORT NUMBER
Insert if performing organization wishes to assign this number.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Give name, street, city, state, and ZIP code. List no more than two levels of an organizational hircarchy.
10. PROGRAM ELEMENT NUMBER
Use the program element number under which the report was prepared. Subordinate numbers ni.iy be included in parentheses.
11. CONTRACT/GRANT NUMBER
Insert contract or grant number under which report was prepared.
12. SPONSORING AGENCY NAME AND ADDRESS
Include ZIP code.
13. TYPE OF REPORT AND PERIOD COVERED
Indicate interim final, etc., and if applicable, dates covered.
14. SPONSORING AGbNCY CODE
Insert appropriate code.
15. SUPPLEMENTARY NOTES
Enter information not included elsewhere but useful, such as: Prepared in cooperation with. Translation of, I'reseiilecl al conlcicmc nl'.
To be published in, Supersedes, Supplements, etc.
16. ABSTRACT
Include a brief (200 words or less) factual summary of the most significant information contained in ilie report. II I he report ioniums a
significant bibliography or literature survey, mention it here.
17. KEY WORDS AND DOCUMENT ANALYSIS
(a) DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper aullion/.cd terms that identity the major
concept of the research and are sufficiently specific and precise to be used us index entries for cataloging.
(b) IDENTIFIERS AND OPEN-ENDED TERMS - Use identifiers for project nanus, code names, equipment designators, etc. Use open-
ended terms written in descriptor form for those subjects for which no descriptor exists.
(c) COSATI I IELD GROUP - l-'ield and group assignments are to be taken from the 1965 COSA11 Subject Category List. Since the ma-
jority of documents are multidisciplinary in nature, the Primary F-ield/Group assignmcnt(s) will be specific discipline, area ol human
endeavor, or type of physical object. The application!s) will be cross-referenced with secondary I ieid/(,roup assignments that will follow.
the primary postmg(s)
18. DISTRIBUTION STATEMENT
Denote releasability to the public or limitation for reasons other than security for example "Release Unlimited." Cite any availability to
the public, with address and price.
19.&20. SECURITY CLASSIFICATION
DO NOT submit classified reports to the National Technical Information service.
21. NUMBER OF PAGES
Insert the total number of pages, including this one and unnumbered pages, but exclude distribution list, if any.
22. PRICE
Insert the price set by the National Technical Information Service or the Government Printing Office, it known.
EPA Form 2220-1 (Rev. 4-77) (Reverse)
------- |