United States      Office of Air Quality       EPA-450/3-85-028a
           Environmental Protection  Planning and Standards      April 1987
           Agency        Research Triangle Park NC 27711

           Air
&ERI\     Coke Oven            Draft
           Emissions from       EIS
           Wet-Coal Charged
           By-Product Coke
           Oven Batteries—
           Background
           Information for
           Proposed Standards

-------
                                    EPA-450/3-85 028a
    Coke Oven Emissions from
          Wet-Coal Charged
By-Product Coke Oven Batteries-
    Background Information for
         Proposed Standards
          Emission Standards and Engineering Division
                  U.S. Environmental Protection Apncy

                  5^fiS £-'*
                  Chicago. IL  60604-3590
           U.S Environmental Protection Agency
              Office of Air and Radiation
          Office of Air Quality Planning and Standards
          Research Triangle Park, North Carolina 27711
                  April 1987

-------
This report has been reviewed by the Emission Standards and Engineering 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 NC 27711, or
from National Technical Information Services, 5285 Port Royal Road, Springfield VA 22161

-------
                       ENVIRONMENTAL PROTECTION  AGENCY

                            Background Information
                                  and Draft
                        Environmental Impact Statement
                    for Coke Oven Emissions from Wet-Coal
                    Charged J3v-Product Coke Oven Batteries
Jack R. Farmer    ^                             (   (Date)
Director, Emission Standards and Engineering Division
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711

1.  The proposed national  emission standards would limit  coke  oven  emissions
    from existing and new wet-coal  charged by-product  coke  oven  batteries.
    The proposed standards implement Section 112 of the Clean  Air Act  and
    are based on the Administrator's determination of  September  18,  1984
    (49 FR 36560) that coke oven emissions are a hazardous  air pollutant.
    EPA Regions III, IV, and V are particularly affected  because most  coke
    oven batteries are located in these regions.

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

3.  The comment period for review of this document is  75  days.  Mr.  Doug
    Bell, Standards Development Branch, telephone (919) 541-5568, may  be
    contacted regarding the date of the comment period.

4.  For additional information contact:

    Dr. James U. Crowder
    Industrial Studies Branch (MD-13)
    U.S. Environmental Protection Agency
    Research Triangle Park, NC  27711
    Telephone:  (919) 541-5596

5.  Copies of this document may be obtained from:

    U.S. EPA Library (MD-36)
    Research Triangle Park, NC  27711

    National Technical Information Service
    5285 Port Royal Road
    Springfield, VA  22161
                                     m

-------
                             TABLE OF CONTENTS
Tables	viii

Figures	xii

1.   SUMMARY	   1-1
     1.1  REGULATORY ALTERNATIVES 	   1-1
     1.2  ENVIRONMENTAL IMPACT	1-2
     1.3  ECONOMIC IMPACT	   1-2

2.   INTRODUCTION	2-1

3.   THE BY-PRODUCT COKE INDUSTRY	3-1
     3.1  GENERAL	3-1
     3.2  BY-PRODUCT COKING PROCESS 	   3-6
          3.2.1  Coal Preparation and Charging	3-6
          3.2.2  Thermal Distillation 	   3-9
          3.2.3  By-Product Collection	3-14
     3.3  COKE OVEN CHARGING, TOPSIDE, AND DOOR EMISSIONS	3-16
          3.3.1  Discussion of Visible and Mass Emissions	3-19
          3.3.2  Wet-Coal Charging Emissions	3-22
               3.3.2.1  Emission Sources and Pollutants 	   3-24
               3.3.2.2  Emission Estimates for Uncontrolled
                        Charging	3-24
               3.3.2.3  Emission Estimates for Controlled
                        Charging	   3-27
               3.3.2.4  Factors Affecting Emissions from
                        Charging	3-33
          3.3.3  Coke Oven Doors and Their Emissions	3-36
               3.3.3.1  Facility Description	3-36
               3.3.3.2  Emission Sources and Pollutants 	   3-43
               3.3.3.3  Emissions from Poorly Controlled Door
                        Leaks	3-44
               3.3.3.4  Emissions from Well-Controlled Door
                        Leaks	3-46
               3.3.3.5  Factors Affecting Emissions from Coke
                        Oven Doors	3-50
                    3.3.3.5.1  Oven pressure	3-50
                    3.3.3.5.2  Temperature effects	3-52
                    3.3.3.5.3  Miscellaneous factors affecting
                               emissions	3-54

-------
                           CONTENTS (continued)

                                                                       Page
          3.3.4  Topside Leaks and Their Emissions	3-55
     3.4  BASELINE REGULATIONS	3-58
     3.5  REFERENCES	3-65

4.    EMISSION CONTROL TECHNIQUES	4-1
     4.1  TECHNOLOGY FOR THE CONTROL OF EMISSIONS FROM CHARGING ...   4-1
          4.1.1  Stage Charging	4-1
               4.1.1.1  Description of Stage Charging 	   4-2
               4.1.1.2  Equipment and Engineering Requirements.  .  .  .   4-17
               4.1.1.3  Optimizing Stage Charging 	   4-20
               4.1.1.4  Performance of Optimized Stage Charging .  .  .   4-22
          4.1.2  Sequential Charging	4-35
               4.1.2.1  Description 	   4-35
               4.1.2.2  Performance of Sequential Charging	4-40
          4.1.3  Scrubber Systems Mounted on Larry Cars 	   4-42
               4.1.3.1  Description 	   4-42
               4.1.3.2  Performance of Scrubber  Systems Mounted on
                        Larry Cars	4-45
     4.2  TECHNOLOGY FOR THE CONTROL OF DOOR LEAKS	4-48
          4.2.1  Traditional Oven Door Seal  Technology	4-49
               4.2.1.1  Koppers Door	4-50
               4.2.1.2  Wilputte Door	4-52
               4.2.1.3  Wolff Self-Sealing Door  	   4-54
               4.2.1.4  Hand-Luting 	   4-54
          4.2.2  Modern Metal-to-Metal Seals	4-56
               4.2.2.1  Modified Koppers Door 	   4-58
               4.2.2.2  Modified Wilputte Seals  	   4-61
               4.2.2.3  Ikio Seals	4-63
               4.2.2.4  Battelle Seal (EPA-AISI)	4-65
               4.2.2.5  Gas Seal, Nippon Steel  Company	4-67
               4.2.2.6  Prechamber Doors	4-68
          4.2.3  Other Control Techniques 	   4-70
               4.2.3.1  Maintenance and Operating Procedures	4-70
               4.2.3.2  Effects of Process Variables	4-71
               4.2.3.3  Soft Seals for Control  of Chuck Door
                        Emissions	4-72
               4.2.3.4  Hoods for Oven Door Fume Collection	4-73
               4. 2.. 3.5  Cokeside Sheds	4-77
               4.2.3.6  Vented Plug	4-78
               4.2.3.7  Jamb Design	4-80
               4.2.3.8  Heat-Settable Sealant 	   4-80
               4.2.3.9  Inboard Luting	4-80
          4.2.4  Startup, Shutdown, Upsets,  and  Breakdowns	4-80
          4.2.5  Door Controls for Tall Ovens	4-82
          4.2.6  Summary of Door Leak Control Performance	4-83
     4.3  TECHNOLOGY FOR THE CONTROL OF TOPSIDE  LEAKS (CHARGING
          PORT LIDS AND STANDPIPES)	4-86
          4.3.1  Description	4-86
          4.3.2  Performance of Control for Topside Leaks 	   4-89
     4.4  REFERENCES	4-97

-------
                           CONTENTS (continued)
5.    MODIFICATIONS	   5-1
     5.1  BACKGROUND	5-1
     5.2  POSSIBLE MODIFICATIONS  	   5-1
          5.2.1  Process Modifications  	   5-2
               5.2.1.1  Dry-Coal Charging 	   5-2
               5.2.1.2  Recycle of Waste  	   5-2
               5.2.1.3  Green Coke	5-3
          5.2.2  Equipment Modifications  	   5-3
               5.2.2.1  Steam or Liquor Aspiration  	   5-5
     5.3  REFERENCES	5-5

6.    MODEL PLANTS AND REGULATORY ALTERNATIVES 	   6-1
     6.1  MODEL BATTERIES	6-1
     6.2  DEVELOPMENT OF REGULATORY ALTERNATIVES	6-3
          6.2.1  Regulatory Alternatives	6-3
               6.2.1.1  Regulatory Alternative I	   6-3
               6.2.1.2  Regulatory Alternative II 	   6-5
               6.2.1.3  Regulatory Alternative III	   6-7
          6.2.2  Alternative Control Technologies 	   6-8
     6.3  REFERENCES	   6-9

7.    ENVIRONMENTAL IMPACT 	   7-1
     7.1  EMISSION RATE ESTIMATES	7-1
     7.2  EMISSION ESTIMATES FOR MODEL BATTERIES  	   7-7
     7.3  ESTIMATES OF NATIONWIDE EMISSIONS 	   7-14

8.    COSTS	8-1
     8.1  INTRODUCTION	8-1
     8.2  COST COMPONENTS	8-2
     8.3  COST ESTIMATES	8-6
          8.3.1  Charging Control Costs 	   8-6
          8.3.2  Door Leak Control Costs	8-19
          8.3.3  Topside Leak Control  Costs	8-30
     8.4  MODEL BATTERY AND NATIONWIDE COSTS  	   8-39
          8.4.1  Model Battery Costs  	   8-39
          8.4.2  Nationwide Costs 	   8-43
     8.5  OTHER COST CONSIDERATIONS	8-45
     8.6  REFERENCES	8-45

9.    ECONOMIC IMPACT	9-1
     9.1  INDUSTRY PROFILE	9-1
          9.1.1  Introduction	9-1
               9.1.1.1  Definition of  the Coke Industry	9-4
               9.1.1.2  Brief History  of the Coke Industry in the
                        Overall Economy 	   9-4
               9.1.1.3  Size of the Iron and Steel Industry	9-6

-------
                      CONTENTS (continued)
     9.1.2  Production	9-10
          9.1.2.1  Production Description	9-10
          9.1.2.2  Production Technology 	   9-10
          9.1.2.3  Factors of Production 	   9-12
     9.1.3  Demand and Supply Conditions	9-15
     9.1.4  Market Structure 	   9-21
          9.1.4.1  Concentration Characteristics  and Number
                   of Firms	9-21
          9.1.4.2  Integration Characteristics  	   9-28
          9.1.4.3  Substitutes 	   9-29
          9.1.4.4  Pricing History 	   9-30
          9.1.4.5  Market Structure Summary	9-30
     9.1.5  Financial Performance	9-30
     9.1.6  Industry Trends	9-40
     9.1.7  Market Behavior:   Conclusions	9-42
9.2  ECONOMIC IMPACT OF REGULATORY ALTERNATIVE  	   9-43
     9.2.1  Summary	9-43
     9.2.2  Methodology	9-44
          9.2.2.1  Supply Side	9-45
               9.2.2.1.1  Data base	9-47
               9.2.2.1.2  Output relationships  	   9-47
               9.2.2.1.3  Operating costs	9-47
               9.2.2.1.4  Capital costs	9-48
               9.2.2.1.5  Environmental costs	9-50
               9.2.2.1.6  Coke supply function—existing
                          facilities	9-52
               9.2.2.1.7  Coke supply function—new
                          facilities	9-55
          9.2.2.2  Demand Side	   9-55
          9.2.2.3  Synthesis 	   9-58
          9.2.2.4  Economic Impact Variables 	   9-62
     9.2.3  Furnace Coke Impacts	   9-66
          9.2.3.1  Price Effects 	   9-69
          9.2.3.2  Production and Consumption Effects	9-69
          9.2.3.3  Coal Consumption and Employment Effects .  .  .   9-72
          9.2.3.4  Financial Effects 	   9-74
          9.2..3.5  Battery and Plant Closures	9-79
     9.2.4  Foundary Coke Impacts	   9-81
          9.2.4.1  Price and Production Effects  	   9-81
          9.2.4.2  Coal Consumption and Employment Effects .  .  .   9-83
          9.2.4.3  Financial Effects 	   9-83
          9.2.4.4  Battery and Plant Closures	9-90
9.3  POTENTIAL SOCIOECONOMIC AND INFLATIONARY IMPACTS	9-90
     9.3.1  Compliance Costs 	   9-90
     9.3.2  Balance of Trade	   9-90
     9.3.3  Community Impacts  	   9-93
                               vn

-------
                           CONTENTS (continued)

                                                                       Page

          9.3.4  Small Business Impacts 	   9-93
          9.3.5  Energy	9-96
     9.4  REFERENCES	9-96

APPENDIX A  EVOLUTION OF THE PROPOSED STANDARDS 	   A-l
APPENDIX B  INDEX TO ENVIRONMENTAL IMPACT CONSIDERATIONS	B-l
APPENDIX C  EMISSION SOURCE TEST DATA	C-l
APPENDIX D  EMISSION MEASUREMENT AND CONTINUOUS MONITORING	D-l
APPENDIX E  COKE OVEN EMISSIONS RISK ASSESSMENT FOR WET-COAL
            CHARGED COKE OVEN BATTERIES	E-l

-------
                                  TABLES

Number                                                                 Page

1-1       Assessment of Environmental and Economic Impacts for
          Each Regulatory Alternative Considered	1-3

3-1       Production and Consumption History of Coke in the
          United States 	   3-3
3-2       Geographical Location and Production of Coke in the United
          States	   3-4
3-3       Coke and Coal Chemicals Produced by U.S. Coke Plants in
          1980	   3-5
3-4       BSO Emission Factors for Poorly Controlled Coke Oven
          Sources	3-18
3-5       Gaseous Pollutants from Charging	3-25
3-6       Estimates of Controlled Charging Emissions	3-32
3-7       Types of Ovens in Current Use	3-38
3-8       Major Components of Coke Oven Gas	3-43
3-9       Minor Components of Coke Oven Gas	3-44
3-10      POM Pollutants from Coke Oven Door Leaks	3-45
3-11      Cokeside Shed Test Results	3-47
3-12      Exponential Model for Door Leak Emissions	3-49
3-13      The Effect of Temperature on the Relative Coking
          Time	3-54
3-14      Topside Leak Emission Test	3-57
3-15      Summary of Baseline Regulations for Charging	3-59
3-16      Summary of Baseline Regulations for Door Leaks	3-59
3-17      Summary of Baseline Regulations for Lid Leaks 	   3-60
3-18      Summary of Baseline Regulations for Offtake Leaks 	   3-60
3-19      Baseline Visible Emission Limits  	   3-62

4-1       Equipment Requirements for Stage Charging 	   4-19
4-2       Summary of Visible Emission Data on Optimized
          Stage Charging	4-23
4-3       Summary of Visible Emission Data on Basic Stage
          Charging	4-25
4-4       Statistical Results from Charging Data  	   4-28
4-5       Comparison of Arithmetic Average and Log Average	4-30
4-6       Visible Emission Data for the Bethlehem Plants
          at Johnstown	4-32
4-7       Particulate Emissions from Scrubbers Mounted on
          Larry Cars	4-47

-------
                            TABLES (continued)

Number                                                                 Page

4-8       Standard Koppers Door at Kaiser Steel 	  4-52
4-9       Percent Leaking Doors for Batteries at U.S. Steel,
          Clairton	4-61
4-10      Summary of Door Leak Data	4-84
4-11      Confidence Levels for Door Leak Data	4-85
4-12      Visible Emission Data on Topside Leaks from Charging
          Port Lids, U.S. Steel, Fairfield	4-91
4-13      Visible Emission Data on Topside Leaks from Charging
          Port Lids, U.S. Steel, Clairton	4-92
4-14      Visible Emission Data on Offtake Leaks, Kaiser Steel,
          Fontana	4-93
4-15      Summary of Lid Leak Data	4-94
4-16      Summary of Offtake Leak Data	4-96

6-1       Model Batteries	   6-2
6-2       Summary of Regulatory Alternatives 	   6-4

7-1       Charging Baseline Limits and Estimated Emission Rates  .  .   7-3
7-2       Door Leak Baseline Limits and Estimated Emission Rates .  .   7-5
7-3       Lid and Offtake Leak Baseline Limits and Estimated
          Emission Rates 	   7-6
7-4       Demonstrated Performance Levels and Estimated Emission
          Rates for Regulatory Alternative II	   7-8
7-5       Emission Limits and Estimated Emission Rates for
          Regulatory Alternative III 	   7-9
7-6       Model Battery Characteristics  	   7-10
7-7       Emission Estimates for the Model Batteries for the
          Baseline and Regulatory Alternative II 	   7-12
7-8       Regulatory Alternative III Emission Estimates for the
          Model Batteries	   7-13
7-9       Battery Characteristics  	   7-15
7-10      Estimated Average Visible Emissions  	   7-18
7-11      Baseline Emission Estimates  	   7-22
7-12      Summary of Nationwide BSD Emission Estimates 	   7-25

8-1       Cost Data from Empire Coke	   8-3
8-2       Cost Data from Bethlehem Steel	   8-4
8-3       Cost Data from Armco, Inc	   8-7
8-4       Armco Costs for Two New 6-Meter Batteries	   8-9
8-5       Indices for Updating Costs 	   8-10
8-6       Charging Costs at Empire Coke	   8-11
8-7       Charging Costs for Armco's Wilputte Batteries  	   8-13
8-8       Charging Costs for Armco's Tall Batteries  	   8-14
8-9       Charging Costs for Bethlehem Steel 	   8-15
8-10      Charging Costs for U.S.  Steel at Clairton	   8-17
8-11      Charging Costs Summary 	   8-18
8-12      Door Control  Costs for Empire Coke	   8-20

-------
                            TABLES (continued)
Number
Page
8-13      Door Control Costs for Armco's Wilputte Battery  	   8-21
8-14      Door Control Costs for Armco's 6-M Batteries	   8-22
8-15      Door Control Costs for Bethlehem Steel 	   8-24
8-16      Door Control Costs for U.S.  Steel at Clairton	   8-25
8-17      Door Costs Summary	   8-26
8-18      Door Costs and Performance	   8-28
8-19      Topside Leak Control  Costs for Empire Coke	   8-31
8-20      Topside Leak Control  Costs for Armco, Inc	   8-32
8-21      Topside Leak Control  Costs for Bethlehem Steel 	   8-33
8-22      Topside Leak Control  Costs for U.S. Steel at Clairton  .   .   8-34
8-23      Lid Leak Control Costs	   8-36
8-24      Offtake Leak Control  Costs	   8-38
8-25      Model Battery Costs for Regulatory Alternative II  ....   8-40

9-1       Coke Oven Control Option:   Regulatory Alternative II  ...  9-2
9-2       Coke Industry Foreign Trade 	  9-6
9-3       Coke Production in the World	9-7
9-4       Value of Shipments, SIC 3312	9-8
9-5       Value Added, SIC 3312	9-9
9-6       Maximum Annual Capacity of Oven Coke Plants in
          the United States in November, 1984	9-13
9-7       Typical Cost Breakdowns:   Furnace Coke Production
          and Hot Metal (Blast Furnace) Production	9-14
9-8       Employment in the By-Product Coke Industry	9-16
9-9       Coke Rate	  9-18
9-10      Coke Plants in the United States, November 1984	9-23
9-11      Interregional Coke Shipments in 1977	9-26
9-12      Percent of Coke Capacity Owned by Top Firms
          (January 1980)	9~27
9-13      Comparison of Coal Prices and Domestic and Imported
          Coke Prices	  9-31
9-14      Financial Information on Coke-Producing  Firms, 1983 ....  9-32
9-15      Financial Ratios for Coke-Producing Firms 	  9-37
9-16      Estimated Capital Costs of New Plants 	  9-51
9-17      Estimate of Elasticities of Steel and Coke Markets	9-59
9-18      Economic Impact Variables and Affected Sectors	9-63
9-19      Baseline Values for Economic Impact
          Analysis--Furnace Coke, 1983  	  9-67
9-20      Pollution Abatement Expenditures for SIC 3312	9-68
9-21      Price Effects of Regulatory Alternative  II—Furnace
          Coke, 1984	9-70
9-22      Production and Consumption Effects of Regulatory
          Alternative II--Furnace Coke, 1984	9-71
9-23      Coal Consumption and Employment Effects  of Regulatory
          Alternative II--Furnace Coke, 1984	9-73
9-24      Industry Capital Requirements of Regulatory
          Alternative II--Furnace Coke, 1984	9-75

-------
                            TABLES (continued)

Number                                                                 Page

9-25      Capital Cost of Compliance as a Percentage of Net
          Investment Regulatory Alternative II--Furnace
          Coke Producers, 1984	9-77
9-26      Capital Cost as a Percentage of Annual  Cash Flow
          Regulatory Alternative II--Furnace Coke Producers,  1984 .  .   9-78
9-27      1983 Baseline Values for Economic Impact
          Analysis—Foundry Coke, 1983	9-82
9-28      Price and Quantity Effects of Regulatory
          Alternative II—Foundry Coke, 1984	9-84
9-29      Coal Consumption and Employment Effects of Regulatory
          Alternative II—Foundry Coke, 1984	9-85
9-30      Industry Capital Requirements of Regulatory
          Alternative II—Foundry Coke, 1984	9-86
9-31      Capital Cost of Compliance as a Percentage of Net
          Investment Regulatory Alternative II—Foundry
          Coke Producers, 1984	   9-88
9-32      Capital Cost as a Percentage of Annual  Cash
          Flow Regulatory Alternative II--Foundry Coke
          Producers, 1984	9-89
9-33      Compliance Costs of Regulatory Alternative II—
          Furnace Coke Producers, 1984	9-91
9-34      Compliance Costs of Regulatory Alternative II—
          Foundry Coke Producers, 1984	9-92
9-35      Employment Data for U.S.  Firms Operating
          Coke Ovens, 1983	   9-95
                                   XI 1

-------
                                    FIGURES

Number                                                                 Page

3-1       Schematic diagram of by-product coke battery	3-7
3-2       Flow sheet showing the major steps in the by-product
          coking process	3-8
3-3       Charging and leveling operations	3-10
3-4       Transverse and longitudinal  sections through Koppers-
          Becker underjet-fired low-differential combination
          by-product coke oven	3-12
3-5       General perspective cut-away section of Koppers-Becker
          combination ovens with gun-flue heating facilities	3-13
3-6       Schematic representation of the pushing operation 	   3-15
3-7       Schematic of coke oven topside	3-23
3-8       Comparison of results between recording only visible
          emissions greater than 20 percent opacity and recording
          all visible emissions 	   3-28
3-9       Comparison of charging emissions models 	   3-31
3-10      Estimates of controlled charging emissions  	   3-34
3-11      Uncontrolled charge 	   3-35
3-12      Vertical cross section of coke oven door on pusher
          side of oven	3-37
3-13      Outside elevation of a pusherside coke oven door	3-39
3-14      Vertical cross section of a pusherside coke oven door .  .  .   3-40
3-15      Horizontal cross section of a coke oven door	3-41
3-16      The effect of time on the internal oven pressure	3-51

4-1       Steps in stage charging coal into a coke oven with four
          charging holes	4-3
4-2       A typical aspiration system operating in the return
          bend of the gas collection duct	4-5
4-3       Aspiration capacity vs. nozzle pressure for single
          standpipe	4-7
4-4       "Still-ERIN" charging system	4-8
4-5       Air flow gooseneck area open in test ovens	4-10
4-6       Aspiration of gases from a coke oven into the gas
          collecting mains  	   4-11
4-7       Variation in the time required to charge coal	4-14
4-8       Jumper pipe	4-16
4-9       A "puff" of emissions during stage charging 	   4-36
4-10      AISI/EPA larry car coke oven charging system	4-38
4-11      Flow of gases in the AISI/EPA sequential charging
          system	4-39
                                   XI 11

-------
                            FIGURES (continued)

Number                                                                 Page

4-12      Representative scrubber system on a larry car 	   4-43
4-13      Details of a shrouded drop sleeve	4-44
4-14      Koppers door	4-51
4-15      Wilputte door	4-53
4-16      Wolff self-sealing door design	4-55
4-17      Cross-section of modified Wilputte door 	   4-62
4-18      Ikio door	4-64
4-19      Battelle seal design	4-66
4-20      Wolff prechamber door	4-69
4-21      Coke oven door hood that assists in controlling pushing
          emissions	4-74
4-22      Channel for the passage of gas formed by hinged plates on
          the door	4-75
4-23      Channel for the passage of gas formed by a door hinged
          to a buckstay	4-76
4-24      Vented plug on pusherside door	4-79
4-25      Slotted jamb design	4-81

9-1       Uses of oven coke as percents of total coke consumption .  .   9-11
9-2       United States apparent consumption of coke	9-17
9-3       Economic impact model 	   9-46
9-4       Coke plant cost centers	   9-49
9-5       Estimated average cost of furnace coke production as a
          function of plant production, 1984	9-53
9-6       Estimated average cost of foundry coke production as a
          function of plant production, 1984	9-54
9-7       Marginal and average cost functions for furnace
          coke, 1984	   9-56
9-8       Marginal and average cost functions for foundry
          coke, 1984	9-57
9-9       Coke supply and demand without import competition 	   9-60
9-10      Coke demand and supply with import competition	9-61
9-11      Economic impact variables without import competition  .  .  .   9-64
9-12      Economic impact variables with import competition 	   9-65
                                   xiv

-------
                                1.   SUMMARY


1.1  REGULATORY ALTERNATIVES

     The regulatory alternatives for wet-coal charging are:

     I -  No national emission standard (regulation by existing
          State Implementation Plans (SIPs)/Occupational Safety and
          Health Administration (OSHA)/consent decree limits).
          Existing emission limits range from 11 to 32 seconds per
          charge.

    II -  For all  new and existing batteries, a maximum of 16 seconds
          of visible emissions per charge (log average* of 10 charges)
          and not more than I charge in 10 that exceeds 45 seconds
          of emissions.

   Ill -  For all  new and existing batteries, a maximum of 11 seconds
          of visible emissions per charge (average of 10 charges).

     Regulatory Alternative I requires full stage charging,  and Regula-

tory Alternative II requires the optimization of stage charging.   It

is not known how all existing batteries could meet the emission limit

of Regulatory Alternative III.

     The regulatory alternatives for door leaks are:
     I -  No national emission standard (regulation by existing SIP/
          OSHA/consent decree limits).  Emission limits range from 4
          to 16 percent leaking doors (PLD).

    II -  Ten PLD for all new and existing batteries.

   Ill -  Five PLD for all new and existing batteries.

     Regulatory Alternative I requires a door leak control program of

inspecting,  cleaning, replacing, or repairing damaged parts  and routine
-Log average = e* - 1 where y = 1n 
-------
maintenance of doors and seals.  Regulatory Alternative II may require
modified doors and seals for all self-sealing doors and additional
manpower for luted doors.  The control requirements for all batteries
to meet the limits of Regulatory Alternative III have not been identified.
     The regulatory alternatives for topside leaks are:
     I -  No national emission standard (regulation by existing
          SIP/OSHA/consent decree limits).  Existing emission
          limits range from 1 to 5 percent leaking lids (PLL) and
          from 4 to 10 percent leaking offtakes (PLO).
    II -  Three PLL and six PLO.
   Ill -  One PLL and four PLO.
     Regulatory Alternative II may require additional manpower for
luting and equipment modifications for some batteries.  It is not
known how all existing batteries could meet the emission limits of
Regulatory Alternative III for topside leaks.
1.2  ENVIRONMENTAL IMPACT
     Table 1-1 summarizes the environmental impacts of the regulatory
alternatives.  The regulatory baseline (Regulatory Alternative I) of
SIP, consent decrees, and OSHA regulations will control benzene soluble
organics (BSO) emissions to a level of 719 Mg/yr (126 to 1,310 Mg/yr).
Regulatory Alternative II for the combined sources reduces nationwide
BSO emissions to 420 Mg/yr (75 to 765 Mg/yr).  Regulatory Alternative III
for the combined sources would reduce BSO emissions to 101 Mg/yr  (21
to 180 Mg/yr).
     None of the regulatory alternatives beyond the current baseline
would have an impact on water pollution, the generation of solid
waste, or energy usage.
1.3  ECONOMIC IMPACT
     Cost functions were derived from data supplied by industry.
These cost functions were applied to each battery to estimate the cost
of improved emission control.  Individual cost estimates were summed
to estimate nationwide costs.
                                 1-2

-------








re
o
LU
Qi
O
LI-
CO
1—
O

z >-
LU as:
o
U_ H-
	 1
f— ~~^
Z C3
LU LU
y rv*
CO
LU
CO
CO
^£


^_j
1
I— 1

LU
— 1
CO
«^
I—
















-P
O) U
to ro
•r- Q.
o £
O5 U
i. ro
01 a.
c e
LU T-

01
-P
ioro
3s 
>
•r—
ro
£_
4^
to
•r"
C
•r—
TJ
«SC












O*
tj
i-
3
O
CO







O 0 0

iH O C3
1








O O O









CM O O
1






«*• CM CM







F— 1
i — r I — F
F— I 1 — 1 1— (

CU CU O!
> > >
-p -p -p
ro ro TO
c c c
S- i. J_
CU O) O)
-p -p -p
 •**> ^)
J-, t_ S_
000
-p -p -p
TO TO TO
333
cn O) cn
OJ O> O>
Q; as: ex



01
c
»r—
O^
S-
TO
x:
u

r^
TO
O
U
1
.p
cy
3=



o o o o o

O O CD O O









o o o o o









o o o o o







^f CM CM «*• CM
~f* ~t~ "4* ~f" "h






1 — 1
I — 1 F— 1 1 — I
h— 1 1 — 1 1 — F 1 — 1 1 — 1

CU O> CU Ol O)
> > > > >
+J 4J +J -P -P
TO TO TO TO TO
c c c c c
i- s- i. i, i~
CU CU CD O) O)
+J -|_5 +J -p 4.)
 ^*^ ^^ 5*^ ?^>
ft- 5— S— S— S—
O O 0 O O
4J 4J +J +J ^->
TO TO TO (& TO
333 33
CTJ CD CO O} CX)
O> O> O> CU O
O^ r>^ rv rv^ rv







to
NX
TO
to 01
vX r—
TO
O> CU
r— -a

S- to
o a.
O 0
Q 1—



0

O









O









O







CM







i — F
I — I
I — 1

Ol
>
4_j
TO
C
i.
0)
-p

t^
o
-p
TO
-j
O)
CU
as:






















































-p
u
TO
-P O.-P
<-> E <->
03 '<— TO
a. a.
e o> e
• r- -p «r-
TO
r— J_ 01
i— Ol CD
TO "O ^-
CO S — J

CM OO 'S-









•p -p
U( 1
^y
TO TO
Q.-P Q.
e u E
•r- TO -i-
Q.
r- e 01
ro -i- -P i—
-P -i-  re -I-
ro -i— to a,  OJ
OJ T3 O O>
E CD «C Z Z
J-
01 4- 1 0 i-H
-P

cn . .
c >-
O LU
1 **^*
ro
1-3

-------
     Regulatory Alternative II for charging may require a total  annual -
ized expenditure of $4.7 million per year.   No significant capital
expenditures are expected because most of the equipment items needed
to control charging emissions are in place as a result of existing
regulations.  Regulatory Alternative II for doors may require a
nationwide capital investment of $6.1 million and a total annualized
cost of $5.4 million per year.  This cost estimate assumes that a
significant number of doors and seals will  be modified for improved
emission control.   Regulatory Alternative II for topside leaks may
require a nationwide capital investment of $5.2 million if many
batteries replace existing offtake systems.  Total annualized costs
are estimated as $1.6 million per year for control of lid leaks and
$7.6 million per year for control of offtake leaks.
     Regulatory Alternative III would apply lowest achievable emission
rates (LAER) to all batteries.  LAER limits have currently been applied
to only a few new and rebuilt batteries.   If all existing batteries
were required to meet LAER limits, many of these batteries would
likely have to be rebuilt to operate as new batteries.  The cost of
rebuilding is on the order of millions of dollars per battery, and
there are an estimated 134 existing batteries.  A finite cost estimate
for Regulatory Alternative III could not be derived; however, forcing
numerous battery closures and rebuilds would have a severe and adverse
economic impact on the industry.
                                 1-4

-------
                           2.   INTRODUCTION

     EPA announced a decision to list coke oven emissions as a hazardous
air pollutant under Section 112 of the Clean Air Act on September 18,
1984 (49 FR 36560).  As a result, standards for regulating coke oven
emissions are under development.   The standard-setting process involves
the identification of sources of coke oven emissions and options for
controlling them.   The selection of a standard is based on an evaluation
of the cost, economic, environmental (air and nonair), and health
impacts of the control options.  This document provides the background
information necessary for this evaluation of coke oven emissions from
wet-coal charging, doors, lids, and offtakes on by-product coke oven
batteries.
                                 2-1

-------
                        3.   THE BY-PRODUCT COKE INDUSTRY

3.1  GENERAL
     Coke is one of the basic materials used in blast furnaces for the
conversion of iron ore into iron.   The major portion (92 percent in
1983) of the coke produced in the United States is used for this purpose.
Most of the iron is subsequently processed into steel, and an adequate
supply of coke is necessary to ensure a continuing steel supply.  Coke
also is used by a number of other industries, principally iron foundries,
nonferrous smelters, and chemical  plants.
     Coke is produced in the United States by two methods:   the original
beehive process and the contemporary by-product recovery or slot oven
process.  Approximately 99 percent of U.S. annual coke production is
produced by the slot oven process.   This conversion of coal to coke is
performed in long, narrow slot ovens, which are designed to permit separa-
tion and recovery of the volatile materials (by-products) evolved from
coal during the coking process.
     In 1975 it was estimated that in the United States there were 62
operating by-product coke plants,  which consisted of 231 batteries contain-
ing 13,324 ovens.1  However, by 1984 only 43 by-product coke plants with
134 batteries remained in existence and only 36 plants were operating.2
The industry has two sectors, and plants are classified generally as
"furnace" and "merchant."  Furnace plants are owned by or affiliated
with iron- and steel-producing companies that produce coke primarily
for consumption in their own blast furnaces, although they also engage
in some intercompany sales among steel firms with excesses or deficits
in coke capacity.   In 1984 there were 28 furnace plants, which accounted
for roughly 92 percent of the total coke production.
                                  3-1

-------
     Independent merchant plants that produce coke for sale on the open
market are typically owned by chemical or coal firms.   The 15 merchant
plants in existence in 1984 accounted for about 8 percent of the total
coke produced.  These firms sell most of their products to other firms
engaged in blast furnace, foundry, and nonferrous smelting operations.
Chemical companies have entered the coke industry to obtain the by-product
hydrocarbon gases that are released when coal is converted into coke;
coal firms have entered the coke industry as a form of downstream vertical
integration.
     In 1983, 23.5 million megagrams of coke were produced in slot ovens
in the United States.  This production was less than the 1982 production
level of 25.6 million megagrams and was 40 percent less than the 1981
production level of 38.9 million megagrams.3  In 1983 furnace plants
produced 20.5 million megagrams of coke and merchant plants produced 3.0
million megagrams of coke.  Principal markets for merchant coke were
blast furnaces not associated with integrated coke-producing facilities,
independent gray-iron foundries, nonferrous smelters, chemical plants,
and affiliated foundries.  A production and consumption history of coke
in the United States since 1970 is presented in Table 3-1.
     Although coke was produced in 14 States in 1983, 59 percent of the
production occurred in three eastern States:  Pennsylvania, Ohio, and
Indiana.2  Indiana, with 5.5 million megagrams of output, was the leading
coke-producing State and accounted for 23 percent of U.S. coke production.
Pennsylvania produced 5.3 million megagrams of coke while Ohio produced
3.0 million megagrams of coke.  The relative amounts of coke produced in
Pennsylvania declined dramatically in 1980 to 1982.  The geographical
distribution of coke oven facilities reflects the locations of coal
deposits and steelmaking facilities.  Table 3-2 shows the geographical
distribution of coke production in the United States from 1972 through
1983.2 3 4
     The yield of coke from coal, approximately 69 percent, has remained
nearly constant during the past decade.3 4  This production does not  in-
clude breeze, the undersize coke that results from the crushing and
screening of the coke after it is removed from the oven.  Although not
completely standardized, the term breeze is generally applied to coke
                                   3-2

-------
c
co in
+->•!-•!- 3
C -P E O S-
CU CU U O °r— fO
U D) 3 $- > Q)
S- C T3 f- ,
O) ro O S-
o- x: s_ a.
u a.
•o
a> a*
-P O 0
in (0 c 3
flrt f* ^ °n
ry L_ ^J t^
i— J- i- O
CO 3 •!- S-
rO <*- Q.
OO
LLJ
1 —
1— r—
oo ro
Q O
UJ h-
r-
i— i XI
Z C
Z3 O
•i— U)
LLJ -P 0)
x Q i- in
i— e •— a> o
3^_ f— ^
I"™ .i. ti
Z tfl < -P S-
i— i c: o 3
o a.
LLJ (J
O 
O CO)
u_ o o - £_
Q£ C 3
O i-H 4-
00
I— (
3=
U)
*y +}
o  E o s-
<2C CU CD O O *r~ rO
U CT> 3 £- > CU
Z S- C T3 >
o a» ro o s-
I-H o_ x: i- a.
r- U O.
a
o
a: >—
Q- CO
"O -P
0) O
U h-
rH 3
1 T3
co o
i- CU
LLJ a u -P
— 1 rO C
GO CU c ro
< -^ 1- r—
I— o 3 a.
u u_
c
CU
0 C
ro -P
. x: c
oo U ro
S- r—
Z3 CU Q.
s:


£-
fO
0)
>-


cn
CO
i





CO
CM
CO
CO



^.
CO
J.^
LO






cn
m
^j.






LO
I-*
CM
LO




^"
r-H
O







LO

CM






to
LO
en
LO




cn
r-t

^j-
LO




f^»
co

LO




O
en
rH


o
i-H
rH
1




CM
rH
,-j.
r^.



CO

rH
LO






rH
I--
,3.






CM
r^
to
*^*




LO
r-H
O







1^.

CO
rH





rH
•*•
rH
LO




to
OO

to
^~




LO
O

LO




rH
cn
rH


<*
cn
•4-





0
rH
rH
CO



^
^~
•=1"
LO






CO
cn
^j.






«^
LO
cn
^f




to
rH
O







10

LO
^





CTl
CO
rj.
LO




0
CM

cn
^j-




rH
rH

LO




CM
r-.
a>
rH


CM CM
CO LO
rH 1
+




rH Cn
co cn
rH tD
cn oo



tO LO
(X) rH
cn co
LO LO






CO CO
LO rH
«5j- LO






CO CM
0 0
LO 00
in in



&***
cn r*>
CT> rH
O CO







rH CO

to «*•
+ 1





0 rH
IO rH
r- LO
in LO




CM l-»
CO «3-

CM O
LO LO




00 "SJ-
r^ to

*3- «3-




oo «d-
cn cn
rH rH


r--
to
r-t
1




00
CM




rH
f*-
00







CM
<*"
^.






cn
CM
^j.
^*




co
to
H







0

f»»
1





LO
CM
rH
LO




r*^
cn

to
^-




CO
CM

^j.




LO
cn
rH


00 'l-
00 IO
+ 1





co in
oo r>.
00 CO
r^- r*^



in fv
in o
rH Cn
m <•






00 00
p*** ci
«d- in






r-. cn
r~- cn
to co
*"J* ^t*




CO O
rH Cn
rH rH







CM O

CM CO
^ 1





r- co
CO rH
CM OO
LO «!fr




CM O
.
«^




r**^.
r-t
LO







rH

CD






LO

CO
^J-




to
to

o
^^




cn
o

co




CO
cn
i-H


CO CD
O rH
1 CM
1




CO CO
cn co
CO CM
r^ to



00 00
to N-
cn o
*tf- ^J-






CM CO
1 — CM
co co






to m
cn in
LO r-
*t CO




rH 0
to to
CO O







o cn

0 CM
rH rH
+ 1




CO <3-
rH Cn
00 rH





co en
to o

CO CO
.
f~"
to
CO




00

O







CM

^
1





O
cn
CO
CO




LO
to

,3.
CO




LO
CM

^j-




i-H
CO
cn
rH


CO
rH
•*J
1




CO
CM
cn
CO



in
co
«^-
CM






cn
CO
CM






to
cn
rH
CM




rH
rH
O







CO

,3.
CO
1




to
in
LO
CM




rx^
o

CO
CM




cn
^^

CM




CM
00
en
rH


oo
r^
rH
•+•




CM
to
^f



^
co
i>^
CM






rH
CO
CM






to
0
LO
CM


-Q

c
m
cr
CO i—
CO r-
cn 
-------





re
•*
IN
V)
I—
t-
0
LU
1— 1
Z
LU
or
z
h-H

UJ
o
o
1 1
o
z
o
[^ , |
I—
o
*n
0
o
a:
a.

Q
Z
*3C

z
o
1— 1
r-
O
O
— J

1

3t



a.



*T~
o

Z 1-1
s: 3:





£



Q >-
s: z




*>
O X.


f\ •%
>- z
:*: 1—




z
t— 1





_J
1— 1




o
o
r—
#. *~^
-
CO rH
, .
^~ in
m in
CM CM
. •
CO CO


^- oo
•3- <•
rH r*


0 0

00 00
r-
r^ oo

d o



CO O
co co


en en
, *
•>*• in






en r**
. .
r-l rH





00 CM
. .
00 00




en r~
, *
rH rH





r** o
, .
CM ro




o*> to
. .
^f ^*



CM «t
r*^. r**1
o^ o*>
rH rH
^.
,
CM
in
rH
.
OO


1^
rH


U3
.
^
in
UD

CD



rH
00


00
*
LO






CM
•
rH





in
.
fs^




LT)
.
rH





Q^
t
CM




CM
.
^I



1^
P**,
a*
iH
LO

^"
*
rH
.
OO


00
rH
rH


O

r-
in
IT)

o



00
CM


oo

oo





*«D
CD
,
O





^^
,
f^.




OO
•
rH





^-
•
CM




CO
.
CO



00

CD
rH
o to

CM in
<* CM
00

CM


rH ID
r-i «^
rH


oo in
. .
l£> «f

1 1
1 1




00 CO
CM CM


t-^.
• -Q
OO





en

, t
rH O





CO 00
, ,
rx. ^~




rH rH
. .
rH rH





^>
• X3
CM




rH (jD
, .
00 rH



O CM
00 00
en en
rH rH
in

CO
CM

f>



CO
m


0

CO

i
i




oo
r-i



f*t







O
.
rH





ID
,
LT)




CM
•
rH






-Q





CM
.
p-l



CO
00

rH



























s~\
in
'C
o
•p

u
•r—
i.
4^
0)
E


w
E
2
o>
(0
D5
0)
E

4—
0

I/)
c
o
•r—
•—
• r—
E

c
•r—

O)
•P
•r-
-P
C
re
cr

r^.
r—
^
re .
























re
re
•a
^j
c
re

Q.

r—
re

•o
•f—
^
•r*
•a
c
••-
^_
0

01
i-
3
in
O
r—
U
in

T3

"O
0

re

o

T3
a>
x:

•P
•rr
5
re
+j
re
o
a
3-4

-------
that will pass through a half-inch screen.  In 1980,  coke plants produced
2.7 million megagrams of breeze, which is equivalent to 64 kg/Mg of coal
carbonized.3  Because of its small size, breeze is not suitable for use
in ferrous blast furnaces.   However, it is used for the sintering of
iron-bearing dust and fine ores, boiler fuel, and for other industrial
purposes.
     After separation and recovery, the by-products that evolve during
the coking process are used within the facility or are marketed.  Typical
products and by-products from the production of coke during 1980 are
presented in Table 3-3.3
     Approximately 92 percent (25.1 million megagrams) of the coke
distributed by U.S. producers in 1983 was shipped to blast furnace
plants, 5 percent went to foundries, and the remainder went to other
industrial plants.3  Apparent consumption (or total consumption) of coke
in the United States equaled 27.4 million megagrams in 1983 and 24.4 million
megagrams in 1982.  Apparent consumption includes domestic production
plus imports, minus exports, plus or minus any net change in stocks.
Apparent consumption, linked closely with iron demand, increased from
1975 to 1976, corresponding to an 8.8 percent increase in blast furnace
iron production.  As iron production declined in 1982, apparent consump-
tion decreased by 83 percent from the 1981 level.  Both apparent consump-
tion of coke and blast furnace iron production increased in 1983.
     Consumption of coke in iron furnaces has decreased partly because
of decreases in the coke rate for these furnaces.  The coke rate is the
ratio between the coke consumed in blast furnaces and blast furnace output.
Although the output of pig iron and ferroalloys from blast furnaces has

           TABLE 3-3.  COKE AND COAL CHEMICALS PRODUCED BY U.S.
                           COKE PLANTS IN 19803
              Product                               Total production
     Coke                                       41.9 million megagrams
     Breeze                                      2.7 million megagrams
     Crude tar                                    2,020 million liters
     Crude light oil                              603 million liters
     Ammonia (sulfate equivalent)                0.36 million megagrams
     Coke oven gas                              20 billion cubic meters
                                  3-5

-------
increased significantly during the past 2 decades, the blast furnace coke
rate has declined.   In 1983, an average of only 540 kg of coke was required
to produce 1 Mg of pig iron and ferroalloys, compared with 860.0 kg in
1956.5  Various factors have contributed to this reduction, but the primary
causes are higher grade ore burdens and the use of increased quantities
of supplemental fuels.
     Supplemental fuels, which consist primarily of fuel oil, tar, and
natural gas, can reduce the coke requirements for the ore reduction proc-
ess.  The use of supplemental fuels has increased significantly in the past
10 years, and in 1975 it resulted in a decreased need for blast furnace
coke of approximately 3.6 million megagrams.5  The application of supple-
mental fuels to replace coke used in blast furnaces is inherently limited
because coke is the principal agent in reducing iron ore.  Therefore, even
though the supplemental fuels are cheaper than coke on a heat-content
basis, they will not be able to replace coke in this vital application.  In
the near future, however, these fuels may help to offset the deficit in
projected coke production levels.
3.2  BY-PRODUCT COKING PROCESS
     Figure 3-1 illustrates the major process equipment in a schematic
diagram of a by-product coke battery.  Note that the coke side is the side
where the coke is quenched and dumped, and the pusher side is the side from
which the pushing ram operates.  A flow sheet is provided in Figure 3-2 to
give an overview of the process from coal charging to by-product recovery.
This operation will be discussed in greater detail in three major subproc-
esses:  coal preparation and charging, thermal distillation and pushing,
and by-product recovery.
3.2.1  Coal Preparation and Charging
     The coal that is charged to the by-product coke ovens is usually a
blend of two or more low, medium,  or high volatile coals that are generally
low in sulfur and ash.  Blending is required to control the properties of
the resulting coke, to optimize the quality and quantity of by-products,
and to avoid the expansion exhibited by types of coal that may cause exces-
sive pressure on the oven walls during the coking process.
     Coal is usually received on railroad cars or barges.  Conveyor belts
transfer the coal as needed from the barges or from a coal storage pile to
                                 3-6

-------
                              CO
                               I
                               o>
3-7

-------
3-8

-------
mixing bins where the various types of coal are stored.   The coal is trans-
ferred from the mixing bins to the crusher where it is pulverized to a
preselected size between 0.15 and 3.2 mm.  The desired size depends on the
response of the coal to coking reactions and the ultimate coke strength
that is required.  For example, low volatile coals coke more readily if the
particle size is small, and smaller particles are reported to give greater
strength to the coke.
     The pulverized coal is then mixed and blended, and sometimes water and
oil are added to control the bulk density of the mixture.  The prepared
coal mixture is transported to the coal storage bunkers on the coke oven
batteries as shown in Figure 3-1.6  A weighed amount or volume of prepared
coal is discharged from the bunker into a larry car, a vehicle which is
driven by electric motors and travels the length of the battery top on a
wide gauge railroad track.   The larry car is positioned over the empty, hot
oven, the lids on the charging ports are removed, and the coal is discharged
from the hoppers of the larry car through discharge chutes.  The flow rate
from the hoppers to the oven may be controlled by gravity, a rotary table,
or screw feeders.  To prevent gases from escaping during charging, a steam-
jet aspirator is used in most plants to draw gases from the space above the
charged coal into the collecting main.7
     Peaks of coal will form directly under the charging ports as the oven
is filled.   As shown in Figure 3-3, these peaks are leveled by a steel bar
that is cantilevered from the pusher machine through an opening called the
chuck door on the pusher side of the battery.  This leveling process pro-
vides a clear vapor space and exit tunnel for the gases that evolve during
coking to flow to the standpipes and aids in the uniform coking of the
coal.7  After filling, the chuck door and the topside charging ports are
closed; the latter may be sealed with a wet clay mixture called luting.
3.2.2  Thermal Distillation
     The thermal distillation of coal to separate volatile and nonvolatile
components takes place in coke ovens that are grouped in batteries.  A
battery consists of 20 to 100 adjacent ovens with common side walls which
are made of high quality silica and other types of refractory brick and
contain integral flues.   Typically, the individual slot ovens are 11 to
16.8 m long, 0.35 to 0.5 m wide, and 3.0 to 6.7 m high.   The heating systems
fall into two general classes:   underjet and gun-flue.  In the underjet
                                  3-9

-------
    SEPARATE
    CHARGING
     LARRY
                                     TOPSIDE PORT
                                                 PUSHER
   A.  THE CHARGING LARRY, WITH HOPPERS CONTAINING MEASURED AMOUNTS OF COAL, IS IN POSI-
      TION  OVER CHARGING HOLES FROM WHICH COVERS HAVE BEEN REMOVED.  THE PUSHER  HAS
      BEEN  MOVED INTO POSITION.
                                          CHUCK DOOR
      THE COAL FROM THE LARRY HOPPERS HAS DROPPED INTO THE OVEN CHAMBER, FORMING
      PEAKED PILES.
LEVELING BAR
               I— i  '.   j r^    H
                -   — ~- • 'y
                     COAL
                              -..1
    .   ,,   ,                        .       V                *

   C.  THE LEVELING DOOR AT THE TOP OF THE OVEN DOOR ON THE PUSHER SIDE HAS BEEN OPENED,
      AND THE LEVELING BAR ON THE PUSHER HAS BEEN MOVED BACK AND FORTH ACROSS THE
      PEAKED COAL PILES TO LEVEL THEM.  THE BAR  NEXT IS  WITHDRAWN FROM THE OVEN,THE
      LEVELING  DOOR AND CHARGING HOLES ARE CLOSED, AND THE COKING OPERATION BEGINS.



Figure 3-3.  Charging and leveling  operations,  (copyright 1971 by

                   United States Steel Corporation)
                                  3-10

-------
heating system, illustrated in Figure 3-4, the flue gas is introduced into
each flue from piping in the basement of the battery.  The gas flow to each
flue can be metered and controlled.  The gun-flue heating system, shown in
Figure 3-5, introduces the gas through a horizontal gas duct extending the
length of each wall slightly below the oven floorline.  Short ducts lead
upward to a nozzle brick at the bottom of each of the vertical flues.7
     Heat for the coking operation is provided by a regenerative combustion
system located below the ovens.  Because the combustion flue gas contains a
significant amount of process heat, two heat regenerators are used for
recovery.  These regenerators are located below each oven, one for combus-
tion air and one for the combustion waste gas, and the flow is alternated
between the two at about 30 min intervals.  The slot ovens operate like
chemical retorts in that they are both batch operated, fitted with exhaust
flues (standpipes), and function without the addition of any reagent.
     The operation of each oven in the battery is cyclic, but the batteries
usually contain a sufficiently large number of ovens (an average of 57) so
that the yield of by-products is essentially continuous.  The individual
ovens are charged and discharged at approximately equal time intervals
during the coking cycle.  The resultant constant flow of evolved gases from
all the ovens in a battery helps to maintain a balance of pressure in the
flues, collecting main, and stack.  All of the ovens are fired continuously
at a constant rate, irrespective of a particular oven's stage in the coking
cycle.  If damage to the refractory occurs in inaccessible locations,
through overheating or expansion of coal, repairs may be extremely diffi-
cult.  A cooldown takes from 5 to 7 weeks, so a battery shutdown is under-
taken only as a last alternative.
     After the ovens are filled, coking proceeds for 15 to 18 hr to produce
blast furnace coke and 25 to 30 hr to produce foundry coke.  The coking
time is determined by the coal mixture, moisture content of the coal, rate
of underfiring, and the desired properties of the coke.  The coking tempera-
tures generally range from 900° to 1,100° C and are kept on the high side
of the range to produce blast furnace coke.   Air is prevented from leaking
into the ovens by maintaining a positive back pressure of about 10 mm
water.  The gases and hydrocarbons that are evolved during the thermal
distillation are removed through the offtake main and sent to the by-product
plant for recovery.

                                  3-11

-------
L_ _      TRANSVERSE SKTIOM   	
[T/iWtf HEATlNS FLUES MO UMOERJfT SAS-DUCTi
     Figure 3-4.  Transverse and longitudinal sections through Koppers-Becker underjet-fired
                 low-differential combination by-product coke oven, (copyright 1971 by
                 United States Steel Corporation)
                                             3-12

-------
Figure 3-5.  General perspective cut-away section of Koppers-Becker combination ovens
            with gun-flue heating facilities. 1. Oven chamber.  2. Vertical combustion
            flues.  3. Horizontal flues.  4. Cross-over flues. 5. Regenerators.  6. Oven
            sole flues.  7. Gas and air connections to waste-gas flue.  8. Waste-gas flues.
            9. Gas ducts for coke-oven gas. 10. Oven gas main. 11. Blast-furnace gas
            main.  12.  Charging holes, (copyright 1971 by United States Steel Cor-
            poration)
                                   3-13

-------
     At the end of the coking cycle,  doors at both ends  of the oven are
removed and the incandescent coke is  pushed out the coke side of the oven
by a ram which is extended from the pusher machine.   This operation is
illustrated in Figure 3-6.   The coke  is pushed through a coke guide into a
special railroad car, called a quench car, which traverses the coke side of
the battery.   The quench car carries  the coke to the end of the battery to
a quench tower where it is deluged with water so that it will not continue
to burn after being exposed to air.  The quenched coke is discharged onto
an inclined "coke wharf" to allow excess water to drain  and to cool the
coke to a reasonable handling temperature.
     Gates along the lower edge of the wharf control the rate of coke
falling on a conveyor belt which carries it to the crushing and screening
system.  The coke is then crushed and screened to obtain the optimum size
for the particular blast furnace operation in which it is to be used.7  The
undersize coke generated by the crushing and screening operations is used
in other steel plant processes, stockpiled, or sold.
3.2.3  By-Product Collection
     Gases evolved during coking leave the coke oven through the stand-
pipes, pass into goosenecks, and travel through a damper valve to the gas
collection main which directs them to the by-product plant.   These gases
account for 20 to 35 percent by weight of the initial coal charge and are
composed of water vapor, tar, light oils, heavy hydrocarbons, and other
chemical compounds.
     The raw coke oven gas exits at temperatures estimated at 760° to
870° C and is shock cooled by spraying recycled "flushing liquor" in the
gooseneck.  This spray cools the gas to 80° to 100° C, precipitates tar,
condenses various vapors, and serves as the carrying medium for the con-
densed compounds.  These products are separated from the liquor in a decan-
ter (Figure 3-2) and are subsequently processed to yield tar and tar deriva-
tives. 7
     The gas is then passed either to a final tar extractor or an electro-
static precipitator for additional tar removal.  When the gas leaves the
tar extractor, it carries three-fourths of the ammonia and 95 percent of
the light oil originally present when leaving the oven.
     The ammonia is recovered either as an aqueous  solution by water absorp-
tion or as ammonium sulfate salt.  Ammonium sulfate is crystallized in a

                                  3-14

-------
 COKE GUIDED
 QUENCHING
    CAR \
PUSHER
          COKING OF THE COAL ORIGINALLY CHARGED INTO THE OVEN HAS BEEN COMPLETED (IN ABOUT
          18 HOURS) AND THE OVEN IS READY TO BE "PUSHED.* THE OVEN DOORS ARE REMOVED FROM
          EACH END, AND THE PUSHER, COKE GUIDE AND QUENCHING CAR ARE MOVED INTO POSITION.

                                          RAM OF PUSHER
                                                             I 4
          THE RAM OF THE PUSHER ADVANCES TO PUSH THE INCANDESCENT COKE OUT OF THE OVEN,
          THROUGH THE COKE GUIDE AND  INTO THE QUENCHING CAR.
Figure 3-6. Schematic representation of the pushing operation,  (copyright
                  1971  by United States Steel Corporation)
                                      3-15

-------
saturator which contains a solution of 5 to 10 percent sulfuric acid and is
removed by an air injector or centrifugal  pump.   The salt is dried in a
centrifuge and packaged.
     The gas leaving the saturator at about 60° C is taken to final  coolers
or condensers, where it is typically cooled with water to approximately
24° C.   During this cooling, some naphthalene separates and is carried
along with the wastewater and recovered.  The remaining gas is passed into
a light oil or benzol scrubber, over which is circulated a heavy petroleum
fraction called wash oil or a coal-tar oil which serves as the absorbent
medium.  The oil is sprayed in the top of the packed absorption tower while
the gas flows up through the tower.  The wash oil absorbs about 2 to 3 per-
cent of its weight, of light oil, with a removal efficiency of about 95 per-
cent of the light oil vapor in the gas.  The rich wash oil is passed to a
countercurrent steam stripping column.  The steam and light oil vapors pass
upward from the still through a heat exchanger to a condenser and water
separator.  The light oil may be sold as crude or processed to recover
benzene, toluene, xylene, and solvent naphtha.7
     After tar, ammonia, and light oil removal, the gas undergoes a final
desulfurization process at some coke plants before being used as fuel.  The
coke oven gas has a rather high heating value, on the order of 20 MJ/Nm3
(550 Btu/stdft3).  Typically, 35 to 40 percent of the gas is returned to
fuel the coke oven combustion system, and the remainder is used for other
heating needs.
3.3  COKE OVEN CHARGING, TOPSIDE, AND DOOR EMISSIONS
     Coke oven emissions consist of a yellow-brown gas which contains over
10,000 compounds as gases, condensible vapors, and particulates.  The
components that primarily concern public health include benzene and the
other  known or suspected carcinogens belonging to a class of compounds
termed polycyclic organic matter (POM).  POM, which condenses on fine
particulates at ambient temperatures, consists of compounds with two or
more fused rings.  There are thousands of POM compounds which vary widely
in physical and chemical characteristics.   These potential pollutants are
sometimes reported as benzene soluble organics (BSD) or by the quantity of
a specific surrogate compound, such as benzo(a)pyrene (BaP).  BSO is composed
                                  3-16

-------
of many compounds, some of which are not POM.   Specific compounds found in
coke oven emissions are discussed in the following sections for each emis-
sion point.
     Coal tar, which is one of the constituents of coke oven gas, is an
example of the many chemical compounds found in coke oven emissions.   Coal
tar consists of aromatic hydrocarbons that are characterized by a ring-type
chemical structure and contain compounds such as benzene, naphthalene,
anthracene,  and their related homologues.  The analysis of coal tar samples
from both foundry and metallurgical coke facilities revealed an average BaP
content of 0.51 percent with a range of 0.23 to 0.87 percent.8
     Emission factors and emission rates for BSD from the four coke oven
emission points are summarized in Table 3-4.  A range of emission rates is
provided because of the difficulty in sampling the fugitive emissions and
the differences in reported values.  These values are order of magnitude
estimates for poorly-controlled batteries and are discussed in detail in
the following sections.  For this section, a typical battery is defined as
having the following features:  62 ovens, 16.3 Mg of coal per oven, 18-hour
coking cycle, double collecting main, and three lids per oven.  The BSO
emission rate for charging is based on a sample of uncontrolled charging
emissions with about 3 to 5 min of visible emissions per charge.  The lower
emission factor for doors (0.13 kg/Mg of coal) is based on the typical
battery with 30 percent of the doors leaking at an average rate of 0.2 kilo-
gram of BSO per hour.  The higher emission factor for doors (0.25 kg/Mg of
coal) is based on actual measurements of door leaks collected by cokeside
sheds.  The emission factors for lid and offtake leaks are based on the
typical battery with 20 to 40 percent of the lids and offtakes leaking at
the minimum and maximum BSO rate.  Emission estimates for well-controlled
batteries are discussed in later sections for each emission point.
     With the exception of charging, an uncontrolled condition is difficult
to define for these sources because routine operations involve some degree
of control.   The emission factors for doors represent a range of about 30
to 75 percent leaking doors, while the emission factors for lids and offtakes
represent about 20 to 40 percent leaking lids and offtakes.  Any emission
estimate for a specific battery should consider the number of leaking
doors, lids, or offtakes and the range of emission rates given in Table 3-4.
                               3-17

-------
 TABLE 3-4.   BSD EMISSION FACTORS FOR POORLY CONTROLLED COKE OVEN SOURCES
Source             BSD (kg/Mg of coal)            BSD emission rate
Charging3         0.059 to 0.5510 1J          0.2 to 1.5 kg/mi n of emissions
Doorsb            0.13 to 0.2510 ll 12        0.2 to 0.7 kg/hr per leaking
                                                door10 12
Lidsc             0.002 to 0.03               0.0033 to 0.021 kg/hr per
                                                leaking lid13
Offtakes0         0.002 to 0.02               0.0033 to 0.021 kg/hr per
                                              leaking offtake13
aFor 3 to 5 minutes of uncontrolled emissions.
 For 30 to 75 percent leaking doors.
cFor 20 to 40 percent leaking lids or offtakes.

A sample calculation is given below for a battery of 62 ovens with 30 per-
cent of the doors leaking at an average rate of 0.2 kilogram of BSO per
hour per leaking door.
     Sample calculation:
     Number of leaking doors = 62 ovens x  Q °°rs x 0.3 (fraction leaking)
                             = 37

     kg of BSO per year = 37 doors x  -S_ x 8^6_p_hr = 5
For 16.3 Mg of coal per oven and an 18-hour cycle time, the coal usage
would be:
     Mg of coal per year = 16'3    * coal x U. x          = 492,000
     Emission factor = 64,800 -f 492,000 = 0.13 kg of BSO/Mg of coal
     Data on benzo(a)pyrene (BaP) for the different emission points vary by
orders of magnitude for different emission tests.  This variability may be
caused by time into the coking cycle, temperature, type of coal, analytical
techniques, or other differences between batteries.   However, data that
reflect the overall BaP content of coke oven emissions are available from
                                 3-18

-------
samples collected on the topside of batteries with both personnel  dosimeters
and ambient air samplers.
     A study of coke oven emissions by Suta1 reported that the BaP was
generally 1 percent of BSD in the analysis of 1,440 airborne samples; this
percentage was confirmed in two additional studies which included 14 and
12 samples.  A statistical analysis of the data from these additional
studies showed that the assumption of 1 percent of BaP in the BSD could not
be rejected at the 0.05 significance level.1
3.3.1  Discussion of Visible and Mass Emissions
     The emission of pollutants is generally characterized by both the
concentration and the flow rate of the pollutant stream.  This analysis of
emissions is difficult to apply to coke oven doors, lids, offtakes, and
charging.  The rates of emissions for doors are highly dependent both on
the time into the coking cycle and the gap size of the metal-to-metal seal.
The rates of emissions for lids and offtakes are dependent on worker prac-
tice in applying luting mixtures, pressure fluctuations, and the gap size
of the emission point.  Charging emission rates are a function of time,
pressure fluctuations, and gap size around the drop sleeves and the charg-
ing ports.  The concentration of pollutants is also expected to vary with
time, and there may be a variation of the concentration of POM from battery
to battery caused by operating conditions and the coal type or blending
practices.  Even if the leaks were well characterized in terms of the size
and length of the gap, there would be potential difficulties in assessing
the flow rate of the pollutant.  Measurement of the concentration of the
POM above the coke battery is difficult because the concentrations are
transient.  The monitored particulate concentration would be a function of
the location of the sampler, the existing wind conditions, and other emis-
sions that may interfere.
     The collection and measurement of fugitive coke oven emissions is
further complicated by the fact that the gases which are emitted from the
oven condense on metal surfaces present in any collection system used for
the sampling device.  These tars even condense on the hot oven jambs.  This
condensation can lead to erroneous results when the gases are carried
through ducts before they reach the sampling device.
                                  3-19

-------
     Mass emission measurements are not practical;  therefore,  regulatory
agencies and coke oven operators have devised visible emission techniques
to assess coke oven emission control.  To measure the emissions from coke
oven doors, lids, and offtakes, a method has been used whereby the emis-
sions are characterized by percent leaking doors (PLD), percent leaking
lids (PLL), and percent leaking offtakes (PLO).   Charging emissions are
characterized by the total time that visible emissions occur during charg-
ing of the oven.  The coke oven has two main doors, one on the pusher side
and the other on the coke side, three to five charging ports,  and one or
two offtakes connecting the oven with the collector main.  The pusherside
door has a smaller chuck door for the leveler bar at the top of the door.
     The coke batteries are inspected and the number of doors, lids, and
offtakes from which any emissions are observable is reported.   These emis-
sions are generally in the form of yellow-brown smoke.  Although some of
the luting produces a white, condensed water smoke upon drying, this is not
considered as a  leak.  The percent leaking can be calculated by dividing
the number of leaking doors, lids, or offtakes by the total number of each
on the battery's operating ovens.
     To measure  visible emissions from charging, an observer records the
length of time that visible emissions occur from the larry car and charging
ports of an oven.  These emissions may be continuous or  intermittent, but
only the time during which visible emissions occur is totaled.  This method
has advantages over an opacity reading which is a subjective evaluation by
the observer on  the level or concentration of the emission.
     There are several advantages to using seconds per charge, PLD, PLL,
and PLO as measures of emission control.  Two advantages are that no special
equipment is needed for the measurement and frequent use of these measure-
ment methods can help identify control problems such as  deviations from
proper work practices and procedures.
     Correlating visible emissions to mass emission rates is difficult for
several reasons.  Capture and measurement of mass emissions are complicated
by the fugitive  and transient nature of the visible emissions.  Also,
multiple sources of emissions are possible for each emission point.  The
mass emission rate for charging depends on the size and  number of gaps,
oven pressure, and the length of time that the emissions occur.  If the
                                  3-20

-------
time of visible emissions from a specific oven (with a given pressure
profile and specific number and sizes of gaps) is reduced, the total emis-
sions from that oven would be reduced.  Seconds of visible emissions are
reduced by controlling the number and size of gaps and by removing the
emissions through the steam aspiration system, which affects the oven
pressure during the charge.  A reduction in mass emissions is expected from
a reduction in visible emissions because control of visible emissions is
accomplished by controlling those factors which influence mass emissions.
     The control of percent leaking doors is also accomplished by control-
ling those factors that directly affect mass emissions.  Most doors are
self-sealing, i.e., gaps are plugged in stages by the condensation of tar
from the escaping coke oven gas.  Doors have the greatest tendency to leak
immediately after charging when the oven pressure and gap size are greatest.
Reports from U.S. Steel's research division14 state that the average PLD is
linearly proportional to the average sealing time:

                              PLD = =p x 100                     (1)
where:
     PLD = percent leaking doors,
      T  = average sealing time of each door (hours), and
      T  = gross coking time (hours).
     For example, an average sealing time of 1.8 hours after charging would
be necessary for a battery to maintain 10 PLD for an 18-hour coking cycle.
Because of the linear relationship between PLD and T , a reduction in the
average sealing time will decrease the percent leaking doors by a propor-
tional amount.   If the average door sealing time for a battery is reduced,
one would expect a reduction in total mass emissions because the average
door leaks for a shorter time.   From the previously discussed example, a
battery on an 18-hour cycle would have an average sealing time of 3.6 hours
if 20 percent of the doors were leaking.
     The U.S.  Steel report also states that sealing time is proportional to
oven pressure and gap size.   The proportionality is with the square root of
the pressure drop or linearly with the cross sectional area of the gap.14
Mass emissions  depend on oven pressure,  gap size, and number of gaps;
therefore,  reducing either the oven pressure or gap size reduces the mass
emissions,  sealing time,  and percent leaking doors.
                                  3-21

-------
     A closer examination of the door sealing mechanism reveals why percent
leaking doors is not necessarily linearly proportional  to the mass emission
rate.   A battery with poorly maintained doors may have  door seals with gaps
that are too large to seal (for example, from "hits"  when the seal is
nicked or damaged from improper door handling) or the seal  may have many
gaps from hardened deposits as a result of poor door  cleaning.   Then these
poorly maintained doors (as measured by percent leaking doors) would have a
higher emission rate of pollutants in kilograms per hour per leaking door
than a better maintained battery with few "hits" and  fewer gaps.  The
implication is that a shorter average sealing time results in a lower PLD;
therefore, the gaps are smaller (i.e., they seal more quickly) and the
emission rate of kilograms per hour per leaking door  is less.
     Visible emissions for topside leaks are expressed as the percent of
leaking lids or offtakes; this is directly related to the number of leaking
lids or offtakes.  If the number of leaking lids and  offtakes is reduced by
sealing the emission point, then the mass emission rate would be reduced.
3.3.2  Wet-Coal Charging Emissions
     When coal is introduced into the incandescent oven, the large volume
of steam, gases, and smoke that forms is forced from  the oven by the increas-
ing pressure of the expanding gas volume within the confines of the oven.
The steam is the result of the naturally occurring moisture in the coal and
any water added during the grinding operations.  Gases are generated by the
volatilization and reactions of the volatile components of the coal and any
oil that may have been added to the coal.  The smoke  is a combination of
fumes evolved with these gases and any particulate that may be entrained by
them.   As the expanding gas volume increases pressure in an oven, emissions
will flow at high velocity from any openings.  Figure 3-7 depicts a coke
oven top side with the larry car in charging position.   Emissions can occur
from the three to five charging ports, the feed hoppers, the one or two
ascension pipes, the standpipe caps, and the collecting main.  The charging
ports in an oven are open for only a relatively short period for each
charge; however, ovens are charged sequentially and the collective effect
is a major source of emissions.
                                  3-22

-------




CL
<
O
UJ
O.
a.
O
ft


UJ
Q.
O.
Z
O
V)
UJ
O
V)
Z
0

f-
u
LU
_l
_l
O
O
I






Ul
Q
co
ex.
O
V)
                                                                                         a
                                                                                         2
                                                                                         a>
                                                                                        Si
                                                                                         8
                                                                                        H-
                                                                                         o
                                                                                         u
                                                                                         01
                                                                                        I".
                                                                                        CO

                                                                                         1
                                                                                         D)
                                     3-23

-------
     3.3.2.1  Emission Sources and Pollutants.   Particles emitted during
the charging cycle have been identified as coke balls,  pyrolitic carbon,
high-temperature coke, char, coal, mineral matter,  and  fly ash.15 16
Particulate samples collected at one plant showed an average of 57 percent
tar.9  For the purpose of this discussion, it is assumed that the quantity
of tar is approximately equal to the quantity of BSD.   Analysis of the tar
from portions of the samples showed from 260 to 18,000  ppm (by weight)
benzpyrene, which includes 3,4 benzpyrene (benz(a)pyrene).   Other polycyclic
organic compounds identified as being present in the samples were:
          benz(c)phenanthrene,
          benz(a)anthracene,
          a benzfluoranthrene isomer, and
          cholanthrene.
These compounds are known or suspected to be carcinogenic.   The fact that
carcinogenic compounds are emitted from coke ovens has  been well document-
ed.1 17
     The distribution of particle sizes of the particulate emissions shows
two distinct size groups.9  The finer particles (47 percent of the total)
have a mass mean diameter of 8.5 urn, and the larger particles have a mass
mean diameter of 235 pm.  The tar portion of the particulate was found
primarily with the finer particles.
     Table 3-5 shows the results of an analysis of gaseous pollutants meas-
ured at one plant.9  Another source separates the gaseous emissions into
the following categories:   nitrogen, nitrogen oxides, methane, hydrogen,
carbon monoxide, carbon dioxide, polynuclear aromatic hydrocarbons, and
coal tar pitch volatiles.18
     3.3.2.2  Emission Estimates for Uncontrolled Charging.  For this type
of fugitive emissions source, collection of representative emission samples
is extremely difficult and, consequently, very little data on mass rates
are available.  Few quantitative measurements for POM or BSD are available.
However, a compilation of available data on particulate emissions and the
results of one test that shows the percent BSD in charging emissions are
available.  Estimated emission factors in the literature vary by at least
one order of magnitude, and the accuracy of the emission factor should be
judged according to this variation.
                                  3-24

-------
               TABLE 3-5.  GASEOUS POLLUTANTS FROM CHARGING9
Pollutant
Total hydrocarbons
Carbon dioxide
Carbon monoxide
Nitrogen oxides
Sulfur dioxide
Hydrogen sulfide
Methane
Ammonia
Phenol
Cyanide
Standard cubic
meters
per charge
0.96
0.83
0.50
0.0031






Standard
cubic feet
per charge3
34
29
17
0.11






Maximum
concentration
(ppm)




232
42
4
130
31
16
aFourteen megagrams (15 tons) of coal per charge.

     With the assistance of the American Iron and Steel Institute, EPA has
compiled and analyzed data on participate emissions from iron and steel
mills to assist in the definition of participate emission factors for each
process.  This study suggests an emission factor for uncontrolled charging
of .25 to .75 gram of particulate per kilogram (0.5 to 1.5 Ib/ton) of coal
charged.19
     A test was conducted for EPA by Mitre Corporation at J&L Steel, Pitts-
burgh, to compare the American Iron and Steel Institute (AISI) larry car
with a conventional Wilputte larry car.9  Samples were collected by putting
enclosures around the Wilputte larry car drop sleeves and evacuating the
emissions through a stack where they could be sampled.   The samples were
collected with an Anderson impactor or a cyclone and glass filter, followed
by impingers in an ice bath.   Isokinetic conditions could not be maintained
because of a high variability in the flow of emissions.  The reported
particulate measurements represented composite samples from different
emission points.   The particulate catch averaged 815 g/charge, or about
0.05 gram of particulate per kilogram of coal (0.1 Ib/ton) from tests of
10 charges with an average sampling time of 3.5 min.
                                  3-25

-------
     Reliable results for total  organic emissions were not obtained from
this test, but the percent BSD and some indication of the BaP content were
obtained.   The average amount of BSO measured (excluding the impinger
catch) was 57 percent.  The impingers averaged 96 percent of the mass
collected in the front of the sampling trains and contained an average of
60 percent BSO.9
     The particulate emission factors were combined with these results to
calculate a BSO emission factor.  The particulate emissions are based on
particulate captured by the filter and do not reflect the BSO collected in
the impingers.
     Total BSO = BSO in particulate + BSO in impingers
               = 0.57 x particulate + (0.96)(0.60) x particulate      (2)
               = 1.1 x particulate
The estimate of BSO emissions from the test at J&L Steel would be 1.1 x
815, or about 900 g/charge from 3.5 minutes of visible emissions (0.055 gram
of BSO per kilogram of coal).
     Data on organic emissions from Russian and Czech coke plants have also
been reported.20  The values reported as aromatic hydrocarbons or tar range
from 0.0005 to 0.1 g/kg of coal (0.001 to 0.2 Ib/ton of coal).  Mass emis-
sions of particulate reported with these values were far below the 0.25 to
0.75 g/kg of coal particulate emission factors discussed previously.  For
this reason, the BSO emission factor calculated above (0.055 to 0.55 g/kg
of coal) is judged to be more realistic than the values reported from the
Russian and Czech plants.
     The data on BaP emissions from which the percent BSO was obtained are
not sufficient to allow calculation of an emission factor.  Only selected
portions of some samples were analyzed for BaP.  The results varied con-
siderably; there were values up to 3,000 and one value at 18,000 ug/g of
BSO reported.9  For 12 of 19 samples, the BaP concentration was below
detectable limits which varied from 220 to 1,400 ug/g of BSO depending on
the sample size.  If the portion of the sample analyzed is assumed to be
representative of the entire sample, the six positive results (excluding
                                              -4           -4
the one high value) are equivalent to 2.8 x 10   to 17 x 10   grams of BaP
                                      -4           -4
per kilogram of coal charged (5.6 x 10   to 33 x 10   Ib/ton).
                                  3-26

-------
     3.3.2.3  Emission Estimates for Controlled Charging.   The previous
section discussed emission factors for uncontrolled or poorly-controlled
charges which generally have dense clouds of emissions.   During observation
of charges controlled by the stage charging operating procedure, EPA observ-
ers noticed that on good charges (a small duration of visible emissions),
the emissions were generally small wisps or puffs which drifted from around
the drop sleeves on the larry car.  For charges where the duration was
longer, the emissions changed to clouds of smoke which escaped to the
atmosphere with higher velocities.  Generally, the longer the duration, the
more large clouds and fewer wisps were seen.  The mass in 40 seconds of
wisps is probably about twice that of 20 seconds of wisps; consequently,
40 seconds of mixed wisps and clouds must be more than twice the mass of
20 seconds of only wisps.  Therefore, assuming a linear reduction from
uncontrolled emissions based on the time of visible emissions would over-
estimate the mass in puffs or wisps from a well-controlled charge.
     For charging, the frequency of wisps of visible emissions can be
compared to the frequency of heavier emissions by relating the time when
any visible emission was observed to the time when emissions exceeded
20-percent opacity.   (Opacity is a crude indicator of the concentration of
pollutants.)  Figure 3-8 illustrates the differences in the results obtained
when only emissions greater than 20-percent opacity are recorded in compari-
son to recording all visible emissions.   Little difference occurs in the
two methods for charges where high levels of emissions are recorded (greater
than 25 seconds in duration where the difference between the methods is
less than 5 percent).   However, for charges where less than 25 seconds are
recorded for all visible emissions, the difference between the two methods
is very significant.  For example, the variance between the two methods is
21 percent for 15 seconds of emissions,  but the variance increases to
47 percent for 5 seconds of emissions.
     A mathematical  model was developed to estimate controlled charging
emissions from the uncontrolled emission factors.21  The model assumes:
          Turbulent flow of the emissions out of the oven,
          A cyclic variation that can be represented by a sine wave and
          occurs in the mass rate of gas generated during charging,
                                  3-27

-------
           1     I     T
                       i     i     r
                                                            o
                                                            00
                                                                       0)
                                                                       a
                                                                       c
                                                                       CO
                                                                       •B
                                                           o
                                                           VO
                                                            o
                                                            Lf>
                                                                O
                                                                O
                                                                O)
                                  o

                                  •_J

                                  Lul
                                                            w> JO
                                                            Jo
                                                            <*> "53
                                                            E£
                                                            o> co

                                                            li
                                                            >«

                                                            11
                                                            o en
                                                                        81
                                                                        0>  w>
                                                                        *~ '
                                                              ?
 I     1
                                                                       I
                                                                       51
                                                                       * 8
                                                                        C -
                                                                       .8:
                                                                        i— i
                                                                        a;

                                                                        o
                                                                       o

                                                                       00
                                                                       CO
                                                                        £
                                                                        o>
o
o
o    o     o     o    o    o    o
OO    P"**     VO     t-D    ^*    C^    
-------
          A constant rate of aspiration, and
          Similar battery conditions, such as the size of the opening
          through which the pollutant escapes.
This model predicts that mass emissions are proportional to the square of
the time of observed emissions.  The model contains many limitations and
assumptions which may not apply to a specific battery; however, the model
is used only as a tool to provide a reasonable upper bound on estimates of
controlled charging emissions.
     The discussion of controlled charging emissions stated that an exponen-
tial reduction in mass emissions with respect to seconds of visible emissions
occurs when the nature of the emissions change as control improves.  Poorly
controlled charges have been characterized as dense clouds of emissions
that escape at relatively high rates from several different emission points;
however, visible emission readings represent only the total time that
emissions occur and do not account for differences in the density of emis-
sions, the mass emission rate, or the number of emission points.  When
emissions are controlled on the order of 30 seconds or less, the emissions
have been characterized as wisps or puffs that are emitted through fewer
emission points.  Figure 3-8 provides additional support for this observa-
tion.  For roughly 30 seconds or more of emissions, essentially all of the
visible emissions have an opacity of 20 percent or more.  For visible
emissions that are less than 30 seconds per charge, the percent of total
emissions greater than 20-percent opacity decreases as the seconds of
emissions decrease.  Therefore, improved control results in both a decrease
in duration and a decrease in density (or mass rate), which suggests an
exponential reduction in mass with respect to seconds of emissions.
     Because the only data available were for uncontrolled charges, the
mathematical model was applied to these single-point estimates of mass
emissions.  An uncontrolled charge cannot be defined precisely in terms of
visible emissions; however, most observers would agree that visible emissions
occur on the order of minutes per charge throughout the 4- to 6-minute
charging period.   Two crude estimates are available for uncontrolled charges:
7,500 g/charge for an estimated 5 minutes of emissions and 900 g/charge
from sampling emissions for 3.5 minutes.   The exponential model (exponent = 2)
was applied to each of these data points to give the two exponential curves,

                                  3-29

-------
labeled Exponential  A and B,  in Figure 3-9.   The model  results are tabulated
in Table 3-6.   The two exponential  models predict that  8 to 33 grams of BSO
escape when 20 seconds of visible emissions  are observed.
     The model labelled Exponential A is recommended as an upper bound
estimate to estimate charging emissions controlled in the range of 11 to 30
seconds per charge.   This approach is designed to avoid grossly overestimat-
ing the mass emissions in a well-control led  charge that would result if a
linear model was used to extrapolate from the uncontrolled data points.  In
addition, a linear reduction in mass with respect to duration would not
account for the observed transition in the nature of charging emissions as
control improves.
     Little data are available to establish  an absolute minimum estimate
for small wisps of emissions from a single drop sleeve  with one small gap.
Data are available from tests of topside leaks with 1-  to 2-meter plumes
originating from a single gap.13  Extrapolation of charging emissions from
the topside test results is difficult because the sampling was conducted
1 hour after charging and differences may be expected in emission composi-
tion, oven pressure, and coal temperature.  In addition, charging emissions
may escape from more than one gap or more than one drop sleeve.  The industry
has argued and our visible emissions data base has confirmed that even the
best-controlled batteries occasionally have  a poorly controlled charge.
Because many of these factors tend to increase the estimate of controlled
charging emissions, the topside leak test of a 1- to 2-meter plume should
provide a reasonable absolute minimum estimate.
     Although the transition in mass emissions from uncontrolled to con-
trolled charging is apparently exponential,  additional  decreases in con-
trolled emissions may have a linear relationship with the visible emission
time.  For example, a linear relationship would be expected when the time
of visible emissions is the cumulative total from a series of puffs that
are similar in size and duration.  The minimum estimate previously discussed
represents a linear model in which the measured emission rate is constant
and the decrease in mass results only from a decrease in the duration of
the emissions.  The results of this attempt  to place reasonable bounds on
estimates of well-controlled charges (11 to  30 seconds  per charge) is
                                  3-30

-------
6,000
5,400 -
                   60
90     120    150    180     210
   Seconds of Visible Emissions
240    270    300
                 Figure 3-9.  Comparison of charging emissions models.
                                  3-31

-------
         _Q
          00
           s-
           rO
 CU
 O1
 s_
 to

 u
\
o
CO
CQ

 O)
CO
Z
o
I—I
CO

I—I
s;
LU


•z.
t—I
fj
o
LU
O
OL
I—
Z
o
o
 co
 LU
CO
LU




CD


CO


LU
_l
oa
         ro
            E
       T3  CU

        O  CU
        U i—

       CO -r-
           in
                 oooooooo
                 oooooooo
                 oooooooo
                  oooooooco
                  «*oocoiar^-r^
                  00 CM  CM CO CT> 00
Is-. LO CO rH















C
o
•r-
+J
•r- O LO O LT> O
in CO CM CM rH rH
C
(0
S-
1—



in
c
0
• r—
in
IS)
E
CU

o

in
CU
4_)
3
C
E
LO
T3
C
TO

»»
C
>
0

i.
CU
Q.
, —
to
o
u

M-
O
in
E

s-
D)
m
D)
CU

tO
rH
J_
fO
O
U

*+—
o
E
10
5-
D)
10
D)
*-

in
X^
0)

CD
O
.
O
\^f/

S-
3
O
p—
_L_
S-
cu
Q.

O
CO
on
i,_
o
E
2
O)
o
r—
•r-

CM
O
*
O


-------
illustrated in Figure 3-10.  The upper bound for these controlled estimates
is Exponential A and the absolute minimum is the linear "wisp" model.  The
wide range between these two bounds reflects the uncertainty in estimates
of controlled charging emissions.
     The discussion of emission estimates for well-controlled charges
focused on the range of about 11 to 30 seconds per charge and the use of an
exponential upper bound and a linear "wisp" model for the lower bound.  The
basis for these estimates  is the judgment that the density and mass flow
rate of the emissions decrease as control improves.   However, mass emission
reductions outside of the transition region may be linearly related.  The
plausibility of a linear model for a very low level  of emissions was dis-
cussed.  A linear model may also be appropriate for higher levels of emis-
sions for various degrees of poor control.  For example, consider a battery
with an average control performance of 60 seconds per charge that is improved
to 30 seconds per charge.  If the oven pressure, gap sizes, and number of
emission points are similar for the two cases, the mass flow rates of
emissions for the two cases are likely to be similar.   Then the reduction in
mass is directly proportional to duration and yields a linear 50-percent
reduction for the example.
     The results for a linear model applied to the two uncontrolled estimates
are shown in Table 3-6 and Figure 3-9 for comparison with the exponential
models.  Results for the two linear extrapolations are not presented for
visible emissions less than 30 seconds in duration because of the expected
exponential reduction from a decrease in density and flow rate.   Conversely,
extrapolations of the "wisp" linear model are not presented for visible
emissions greater than 30 seconds because of the expected exponential
increase in emissions as they change in nature from wisps to dense emissions.
The linear models may be more appropriate for estimating emissions and
emission reductions for those batteries that are not well controlled and
that have occasionally dense emissions of long duration.
     3.3.2.4  Factors Affecting Emissions from Charging.  Uncontrolled
emissions from charging cause large visible clouds of smoke and fumes.
Figure 3-11 shows a typical uncontrolled charge; occasionally the visible
emissions almost completely obscure the larry car.   If the emissions are
uncontrolled,  they may exist from the time that the  coal starts to flow
                                  3-33

-------
   80  r-
    70
    60 -
    50  -
0)
O)
CO
I40
QQ

OS
    30  -
    20  -
    10  -
                  5          10          15         20          20
                                Seconds of Emissions


                  Figure 3-10.  Estimates of controlled charging emissions.
                                  3-34

-------
Figure 3-11. Uncontrolled charge.
             3-35

-------
into the oven until the charging port lids are replaced—a period of about
3 to 5 min.
     Many variables can potentially influence the quantity of emissions
during charging.   (The effect of these variables on the control  of emis-
sions is discussed in Section 4.3.1.)  Data are not available to quantify
the effects of these variables; however, the possibilities can be discussed.
A higher volatile content of the coal or the addition of more oil to the
coal would probably result in a greater volume of gases and fume (more
emissions) generated in the oven as these components decompose.   More
moisture in the coal will also add to the volume of gases (as it is converted
to steam) in the oven.  This extra volume tends to force emissions into the
atmosphere more rapidly.  On the other hand, high moisture may retard the
rate at which the temperature of the coal increases, because evaporation of
the water absorbs a considerable amount of heat.  At lower temperatures,
the coal volatiles and oil will decompose more slowly; this results in
slower gas generation and less emissions.
     The fineness of grind, which determines the size of the coal particles,
may be another factor that affects the emission rate.  Finely ground coals
should have a greater tendency to be entrained in the effluent gases and
carried out of the oven than coarser ground coals.  Also, the finer coal
may volatilize faster and generate more gas in a short time.
3.3.3  Coke Oven Doors and Their Emissions
     Coke oven doors are both functionally and structurally important parts
of a coke battery.  During the coking cycle, the door (1) minimizes heat
loss, (2) seals the oven to prevent gas loss in the early part of the cycle
and prevent air infiltration in the latter part of the cycle, (3) holds the
coal mass in the oven during charging, and (4) allows the operator to level
the coal charge in the oven.22  Figure 3-12 is a vertical cross-section
sketch of a door in place on an oven.  The buckstay and jamb provide the
structural support that maintains the integrity of the oven brick across
the width of the battery.  This section will describe the coke oven doors
in more detail and discuss the pollutants that are emitted from coke oven
door leaks.
     3.3.3.1  Facility Description.  Coke is produced in slot ovens that
are typically 11 to 16.8 m long, 3 to 6.7 m high, and 0.35 to 0.5 m in
                                  3-36

-------
              GASES REMOVED VIA THE
              ASCENSION PIPE TO THE
                COLLECTOR MAIN
 DOOR HANGER
 S-SHAPED SEAL
  CHUCK  DOOR
    A-CASTING


    TOP LATCH


    BRICK  PLUG

  DOOR FRAME
LIFTING LUGS
  E
  
-------
width.  The most recently built batteries are the large size with ovens 6 m
in height.  Each slot oven in a by-product coke battery has two doors which
serve as end closures for the oven.   The door is constructed with a special
refractory plug on a cast iron or fabricated steel frame that is 3 to 6 m
long.  The typical door may weigh from 3.6 to 7.3 Mg.   Figure 3-13 shows an
outside view of a coke oven door.  More detail is provided in Figure 3-14
which illustrates a vertical cross section of a door in place on the oven.
     Two basic types of seals are used to control the leakage of pollutants
from gaps between the door and the door jamb.  The most common type is a
self-sealing door which has a seal in the form of a flexible S-shaped metal
band or rigid knife edge (Figure 3-15) around the periphery of the door.
It is termed "self-seal ing" because it relies on gaps being sealed by the
condensation of tars in the escaping gas.  Other doors are sealed (hand-
luted) by troweling a luting mixture into a V-shaped opening between the
metal door frame and the roll-formed steel shape on the end of the ovens.
Koppers and Wilputte are the two major manufacturers of self-sealing doors
in the United States.  A distribution of batteries by manufacturer is pro-
vided in Table 3-7 from a recent listing of 222 batteries.24
     The oven's door jamb is a complex, rigid, one-piece shape made of gray
cast  iron or ductile cast iron.  The door jamb is attached to the end of
the oven by connecting it to a bolting plate that is held in place behind
the buckstay, or by bolting the jamb behind the buckstay.  The buckstay is
a wide flange beam that is flush with the brickwork on each vertical side
of the oven and provides the required support (compression) for the oven
refractory and regenerator walls.

                 TABLE 3-7.  TYPES OF OVENS  IN CURRENT USE24
Type
Wilputte
Koppers
Koppers-Becker
Other
Percent
of batteries
25.2
22.5
44.1
8.2
                                  3-38

-------
       LATCH HOOK
SEAL ADJUSTING SCREW
                                          LEVELER OR "CHUCK" DOOR
                                           UPPER LATCH
                                           OVEN NUMBER
                                           PLATE BRACKET
                                           LOWER LATCH
    Figure 3-13. Outside elevation of a pusherside coke oven door.23
                         3-39

-------
LEVELER OR "CHUCK" DOOR
       SPRINGLOADED
       LOCKING BAR
              ooon JAMB
                                                   REFRACTORY PLUG
        Figure 3-14. Vertical cross section of a pusherside coke oven door.
                                                                  23
                                  3-40

-------
                                  CO
                                  «M_

                                   o
                                   •§

                                   g
                                   §
                                   0)

                                   *
                                   u
                                   (0
                                   o
                                   2
                                   o

                                   I

                                   I
                                   ^
                                   o
                                   X

                                   in
                                   CO
                                   a>
                                   05
3-41

-------
     There are two types of door latch mechanisms.   The first is the gravity
type which consists of bars which drop or swing into wedge-shaped latch
hooks.   The second is the screw type (Figure 3-15)  in which a screw forces
the door against the jamb or compresses a spring, which in turn forces the
door against the jamb.   The gravity type latch is generally found on older
batteries, whereas the screw latch is found on newer batteries.
     Door extracting machines are operated on rails on both the pusher side
and coke side of the battery.   On the pusher side,  the door extractor may
be a part of the pusher machine or a separate, self-propelled device.   The
door machines remove and hold the door during the pushing operation, and
then they reattach the doors after pushing, before  the empty oven is re-
charged.  They are either electrically and/or hydraulically operated from
an elevated cab that contains all the controls. The head of the extractor
on door machines for the screw type latching contains a mechanism for
latching the doors and compressing the loading springs.  This mechanism is
motor driven and usually cuts off at a preset load  by means of a current
limiter to the drive motor.  Door removal occurs at a relatively slow speed
and is easily controlled to allow accurate door alignment and to avoid
damaging the door's sealing edge.7
     The pusherside door has a small door at the top called the leveler or
chuck door (Figures 3-13 and 3-14).  A leveling bar enters through the
chuck door to level peaks of coal formed during charging, thereby clearing
the space at the top of the oven to allow gases to  escape to the collecting
mains.   The chuck door is also an emission source  for coke oven gas.
     To start the coking cycle, the oven is charged with coal from the top
of the oven by a larry car.  A leveling bar from the pusher car enters the
oven through the chuck door and levels the peaks of coal that form under
the charging holes.  At the end of the coking cycle, the doors on both
sides are removed, and the coke is pushed with a ram from the pusher side
out of the oven into a quench car on the coke side.  Usually, the door and
jamb are cleaned, and the doors are replaced and adjusted.  At batteries
that use hand luting, the door is sealed with a luting mixture before
charging.
                                  3-42

-------
     3.3.3.2  Emission Sources and Pollutants.   Door leak emissions are
fugitive in nature and can occur at any point on the perimeter of the door
where a gap exists between the door and the door jamb.   Another emission
source is the small leveler or chuck door on the pusherside door.
     The gas that escapes from coke oven doors contains many chemical
compounds.   Hydrogen and methane are the major constituents, as listed in
Table 3-8.   Some of the minor constituents are listed in Table 3-9.25  The
components listed in these tables are averages; the constituents of door
leak emissions vary widely and depend upon the time into the coking cycle.
     The fugitive nature of door leaks has posed collection and sampling
problems that were discussed in detail in Section 3.3.1.  EPA attempted to
measure emissions from a coke oven door by positioning a collection hood
over a leaking door.26  Their data showed a mass emissions rate of 0.06 kilo-
gram of particulate per hour from a single door during the first hour of
                                                        -5
the coking cycle.  This rate dropped to less than 6 x 10   kg/hr during the
latter portion of the cycle.  The flow rate of gas from the door was 99 m3/hr
initially and dropped to 39 to 54 m3/hr for the rest of the cycle.  Interpre-
tation of other analytical results was difficult because of the accumula-
tion of deposits on the hood and ductwork of the sampling system.  However,
Battelle's analyses showed the presence of several compounds known to be
carcinogenic.26  Specific POMs which have been identified in door leak
emissions are listed in Table 3-10.27

               TABLE 3-8.  MAJOR COMPONENTS OF COKE OVEN GAS25
Compound
H2
CH4
CO
C02
N2
02
C H
m n
H2S
NH3
Percent volume
46.5 -
22.6 -
4.0 -
1.1 -
0.7 -
0.2 -
1.7 -

0.5 -
1.3 -
65
32.1
10.2
2.8
8.5
0.8
5.2

4.5
9.0
                                  3-43

-------
               TABLE 3-9.   MINOR COMPONENTS  OF  COKE  OVEN GAS25
                                             Concentration
                    Component                   (g/m3)

                      HCN                     0.1 -  4.0
                     Dust                     1.8 -  36
                      BaP                     0.2 -  0.6
                    Benzene                  21.4 -  35.8
                    Toluene                   1.5 -  3.0
     3.3.3.3  Emissions from Poorly-Controlled Door Leaks.   Probably the
most reliable coke oven door data are those gathered on BSD emissions from
cokeside sheds.   A cokeside shed is a large hood which extends over the
entire coke side of the battery to capture both pushing and cokeside door
emissions.   Available cokeside shed test results are generally representa-
tive of high levels of percent leaking doors (i.e., levels greater than
30 PLD).
     In May 1977, EPA tested Wisconsin Steel's shed that covered the coke
side of 45 5-meter ovens.10  Both the inlet and outlet of the wet electro-
static precipitator (WESP) were sampled during four 10-hour tests.  Sam-
pling was discontinued during pushing so that the data would only reflect
emissions from doors.  BSD emissions during the test averaged 6 kg/hr from
an average of 31 leaking doors (70 PLD).  Emission factors for BSD and POM
were estimated as 0.25 g/kg of coal and 7.0 x 10   g/kg of coal, respec-
tively.10  POM species which were identified during the tests are given in
Table 3-10 with their uncontrolled emission rates.  BaP emissions from the
Wisconsin Steel  test averaged 0.043 g/hr.27 29  The same testing program
measured 0.165 to 0.212 gram of particulate per kilogram of coal.30  Benzene
analyses were performed on samples taken from three of the sampling runs.
Benzene emissions varied from 42 to 84 g/Mg of coal with an emission rate
of about 2 to 4 kg/hr.29
     A similar test was conducted at Armco, Inc. in Houston, Texas, in
October 1979.12 31  The Armco shed encloses the coke side of 62 4-meter
                                  3-44

-------
           TABLE 3-10.   POM POLLUTANTS FROM COKE OVEN DOOR LEAKS27
                           (cokeside doors only)
  POM species
                                      Carcinogenicity
Emission rate
   (mg/hr)
Naphthalene
Fluoranthrene
Pyrene
                                                                 387
                                                                 428
                                                                 184
Benz(c)phenanthrene
Chrysene
Benz(a)anthracene
                                             +++
                                              ±
                                              +
      17
     124
     114
7,12-Dimethylbenz(a)anthracene
Benz fluoranthrenes
Benz(a)pyrene
                                             ++++
                                              ++
                                             +++
       I
     154
      43
Benz(e)pyrene
Cholanthrene
Indeno (l,2,3-cd)pyrene
                                                                  95
                                                                  <0.04
                                                                  46
Dibenz(a,h)anthracene
Dibenz acridines
Dibenz(c,g)carbazole
                                             +++
      48
      <0.04
      <0.04
Dibenz pyrenes
3-Methyl cholanthrene
                                             +++
                ++
                 +
                 ±
                         Strongly carcinogenic
                         Carcinogenic
                         Uncertain or weakly carcinogenic
                         Not carcinogenic
      43
      <0.04
 Reported carcinogenicity by Public Health Service28 where
                                  3-45

-------
ovens.   Three tests conducted during nonpush periods  measured 6.8 to 13
kilograms of BSO per hour from 10 to 24 leaking doors (16 to 39 PLD).12
BaP emissions ranged from 25 to 32 g/hr at the inlet  to the WESP.   Benzene
emissions ranged from 0.7 to 2.1 kg/hr.31
     Bethlehem Steel sampled the emissions from their Burns Harbor shed on
Battery I.10  The shed is 122 m (400 ft) long and covers the coke side of
82 6-meter ovens.  During these tests,  BSO emissions  during nonpush periods
averaged 3.9 kg/hr;10 the number of doors  leaking was not reported.   Beth-
lehem Steel also sampled emissions from temporary stacks mounted on pusher-
side doors at their Burns Harbor plant.  A total of 14 samples were collected
at four doors that were completely enclosed between buckstays.   Toluene
soluble organics averaged 0.22 kg/hr for each door.10
     The shed test data, summarized in  Table 3-11, reveal a range of BSO
emissions of approximately 0.2 to 0.7 kilogram of BSO per hour per leaking
door for tests where the number of leaking doors was  recorded.   The number
of doors leaking at the Burns Harbor test was not reported, but the range
per leaking door compares favorably with the other data for a level of 50
percent leaking doors.  (An EPA inspection in March 1975 showed a level of
27 to 69 percent leaking doors.32)  The results in Table 3-11 represent a
range of 10 to 35 actual leaking doors.  In terms of percent leaking doors,
a range of 16 to 78 percent is represented.  The limited data in Table 3-11
show no direct relationship between mass emissions and either the number of
door leaks or the percent of doors leaking for different tests.  However,
the Armco shed data show that the overall  mass emission rate increased when
the number and percent of doors leaking increased.
     3.3.3.4  Emissions from Well-Control led Door Leaks.  The emission rate
from two cokeside shed tests averaged 0.2 to 0.7 kilogram of BSO per hour
per leaking door.  However, the results from leaking doors covered by
cokeside sheds would overestimate the emissions from batteries with well-
controlled doors.  Larger door leaks are generally observed under sheds,
because the emissions are usually removed by a pollution control device
and, consequently, the incentive to prevent leaks is not as great.  Poor
lighting may also lead to more leaks because of the increased difficulty  in
inspecting, cleaning, and handling the doors.
                                  3-46

-------
                  TABLE 3-11.   COKESIDE  SHED  TEST  RESULTS

Wisconsin



Average
ARMCO, Inc


Average
Bethlehem
Bethlehem
Kilograms
of BSO
Test per hour
Steel Shed30 7.0
5.9
5.4
6.0
6.1
. Shed12 6.8
11
13
10.3
Steel, Burns Harbor10 3.9
3.9
Steel, hoods10 0.22C
Percent
leaking
doors
73
78
60
69
70
16
31
39
29
50b
--
Number
of
leaking
doors
33
35
27
31
32
10
19
24
18
	 a
41b
1
Kilograms
of BSO
per hour- pet-
leaking door
0.21
0.17
0.20
0.19
0.19
0.68
0.59
0.55
0.58
	 a
0.10b
0.22C
 The number of leaking doors  was  not  reported.
""Assumes  that 50 percent of the 82 doors  were  leaking.
"This value is for toluene solubles from  a  pusherside  hood  over the  door.
                                 3-47

-------
     A linear relationship between PLD and mass emissions from cokeside
shed tests may be inappropriate to estimate emissions from well-controlled
doors.   Visible emissions from a typical  door leaking under a shed are
heavier than the visible emissions from a typical  door on a well-controlled
battery.   To account for the change in plume size  as the total number of
leaking doors decreases, a mathematical model was  developed to estimate
controlled door leak emissions from the cokeside shed test results.33  The
theoretical model relates reductions in mass emissions to reductions in PLD
by considering average sealing time, oven pressures, gap sizes, and gas
characteristics.  The model is developed according to the previously dis-
cussed relationship between average sealing time and PLD.  The limitations
of the model include limited data on oven pressure profiles and the assump-
tion of typical sizes and numbers of gaps.  The model results predict that
mass emissions are roughly proportional to the 2.5 power of PLD down to
levels of approximately 10 PLD.  Below 10 PLD, the model becomes more
linear; for example, the projected exponent is 1.6 at about 5 PLD.33
     The model was applied to the Wisconsin Steel  and Armco, Inc.  shed
tests for average levels of 6.1 kilograms of BSO per hour at 70 PLD and
10.3 kilograms of BSO per hour at 29 PLD, respectively.   The results,
summarized in Table 3-12, are used to estimate emissions from well-controlled
doors.
     The emission rates in Table 3-12 can be compared with the previously
discussed topside leak tests to determine if the range of estimates for
door leaks is reasonable.  Most doors leak during  the initial portion of
the coking cycle when oven pressures and gap sizes are greatest; the doors
seal later in the cycle.  As the doors seal, the emission rate is reduced
and the plume size decreases.  The emission rate of interest is the average
amount of emissions from the end of charging to the time when no emissions
are visible.  The minimum average rate may be approximated by the mass
emissions from a narrow 1- to 2-meter plume from a single gap.  Three tests
were conducted during the second hour of coking to collect and measure the
mass emissions in a 1- to 2-meter plume originating from a single hole in a
charging port lid.13  The test results averaged 0.021 kilogram of BSO per
hour and ranged from 0.012 to 0.035 kilogram of BSO per hour; this compares
favorably with the exponential model results extrapolated from the Wisconsin
                                  3-48

-------
          TABLE 3-12.   EXPONENTIAL MODEL FOR DOOR LEAK EMISSIONS3

                             Wisconsin Steel               Armco,  Inc.
   Percent                        test30                    test12
leaking doors              (kg/hr/leaking door)         (kg/hr/leaking  door)
70
29
20
15
12
10
0.19b
0.052
0.030
0.019
0.014
0.010
—
0.58b
0.33
0.21
0.15
0.12
 For comparison,  the measured BSO rate for a 1-  to  2-meter  plume  from  one
 gap was 0.021 kg/hr.13
 Average results  from the cokeside shed tests.
                                 3-49

-------
Steel test (0.01 to 0.03 kg/hr for 10 to 20 PLD).   Therefore,  this exponen-
tial model should provide a reasonable lower bound for estimating emissions
from wel1-controlled doors.   The comparison also indicates that the exponen-
tial model results for the Wisconsin Steel  test in Table 3-12  may under-
estimate emissions from doors that have more than one gap between the door
seal and jamb.   The model results for the Armco, Inc.  shed tests should
provide a reasonable upper bound which would apply to controlled door leaks
originating from either a very large gap or from multiple gaps on the same
door.
     3.3.3.5  Factors Affecting Emissions from Coke Oven Doors.
     3.3.3.5.1  Oven pressure.  A mixture of coals is introduced into a
slotted by-product oven and heated to 900° to 1,100° C (1,650° to 2,000° F).
Inside the oven, the bituminous coal initially develops a pressure caused
by the destructive distillation of the coal structure and the  release of
volatiles.  This initial pressure diminishes with time because the coke
develops fissures and channels along the oven wall for the gas to escape.34
Figure 3-16 demonstrates the initial rapid decrease in pressure followed by
a slower second stage pressure reduction.  After approximately 1 hr the
pressure is 5 mm of water.  The emissions from coke oven doors follow this
general trend of leaking more severely at the beginning than at the end of
the coking cycle because the pressure is relatively low after approximately
30 to 60 min.
     Coke oven collector mains are operated with a positive pressure--
usually 6 to 10 mm of water.  The purpose of this elevated pressure is to
assure that the oven remains above atmospheric pressure at all times and
locations.  If the pressure becomes negative at any time or place, a leak
will produce infiltration of air with subsequent damage to the ovens.
     Consider a 4-meter oven having a standpipe with a height of 1 m and
coking at 980° C (1,800° F).  The draft effect under these conditions is
slightly greater than 4 mm of water.  Thus, if the pressure in the collector
main were exactly atmospheric pressure, the pressure at the bottom of the
oven would be -4 mm.  In this particular oven, a positive pressure of 6 to
7 mm in the collector main would assure that the pressure at the bottom of
the oven would never go below 2 to 3 mm.  For higher ovens and standpipes
                                  3-50

-------
                                                                           (O


                                                                           CO

                                                                           £

                                                                           o>
o

-------
and for higher coking temperatures,  the collector main pressure would have
to be correspondingly higher.   Thus, oven height and coking temperature
establish one set of variables for predetermining collector main pressure.
     Another consideration in regard to oven pressures relates to the
timing in the coke oven cycle.  At the end of the cycle,  coke temperatures
are high, leading to maximum stack effect.   Gas evolution is low and resis-
tance to gas flow is negligible.   Therefore, at the end of the coking
cycle, the pressure at the bottom of the oven is at its lowest point.
     At the beginning of the coking cycle, gas temperatures are low, lead-
ing to low stack effect, and gas evolution is high, leading to resistance
to gas flow.  Therefore, at the beginning of the cycle the pressure rises
in the oven.  The rise in pressure is nominal in the gas space above the
charge.  However, at the bottom of the oven, the gas which evolves must
pass through the entire coal mass in order to reach the gas space leading
to high resistance and high pressures.
     Invariably the major portion of the leak from a coke oven door appears
to be at the top of the door and in some cases, this observation is correct.
However, in most cases, the major portion of the leak occurs at the bottom
of the door because of the higher pressure in that area.   The rising gases
along the outside of the door, rather than the position of the leak, give
the impression that gases are being emitted principally from the top of the
door.
     As tars start to condense on the jamb, the gaps will close in stages,
usually starting from the bottom of the door.  This reaction is partially
caused by the decrease in temperature from the top to the bottom of the
jamb where tars first begin to condense and plug gaps.  In many instances,
the top gaps never seal.35
     3.3.3.5.2  Temperature effects.  Coal is a complex mixture of organic
compounds, and decomposition of these compounds forms gases, liquids, and
solid organic compounds when heated in the absence of air.  The rate of
heating and the temperature are important in determining the composition of
the by-product compounds.  The products of coking are formed in the follow-
ing three stages:
                                  3-52

-------
     1.   Primary products of coal below 700° C (1,778° F) include
          water, carbon oxides, hydrogen sulfide, hydroaromatic com-
          pounds, paraffins, olefins, phenolic, and nitrogen com-
          pounds.7  With temperatures less than 500° C (1,270° F)
          there is a loss of side chains and small rings are split
          off.  Between 500° and 600° C (1,270° and 1,524° F), the
          aromatic lamellae condense.  Between 600° and 700° C (1,524°
          and 1,800° F) there is formation of polycyclic aromatics
          that can be present in coke oven emissions.36

     2.   Above 700° C in the hot oven walls, in the hot coke, and in
          the free space in the oven, the primary decomposition pro-
          ducts have secondary reactions involving synthesis and
          degradation.  Hydrogen is evolved with the formation of
          aromatic hydrocarbons and methane.7  The coal residue has
          larger aromatic clusters above 800° C (2,032° F).

     3.   Hydrogen is progressively removed from the residue; this
          increases the size of the aromatic clusters and produces a
          hard coke.7

     The diverse complex reactions that produce the coke oven gas are
temperature dependent.  Temperature can be controlled by adjusting the flue
temperature, thus providing a possible control variable for the composition

of the by-product gases.   Moderate changes in the flue temperature exert
little relative influence on coke properties, provided that coking rates

are constant.  Increases in the coking time are reported to increase the
shatter resistance, cause the hardness to decrease approximately 10 percent,
and cause the stability to increase approximately 10 percent.37  Additional

investigations confirmed these trends, but the coal type may have had an
influence on these results.
     Heat is transferred to the coal mix from the hot refractory bricks by
radiation, conduction, and convection.  At the high temperatures of coking,
heat transfer by radiation is expected to dominate.  Heat transfer by
radiation is proportional to the difference in the fourth power of the
temperature, and the coking time has been shown to be inversely related to
the fourth power of the inner wall temperature.37

     Based on a coking rate that is proportional to the fourth power of the
temperature, the relative coking hours were estimated for a number of tem-

peratures.  The capacity of the battery can be calculated from the coking

time.   This approach does not include the time between pushes and charges.
                                  3-53

-------
Decreasing the temperature from 1,000° to 800° C (1,832° to 1,472° F) would
reduce the coke capacity of a battery by one-half.   Some of these calcula-
tion results are presented in Table 3-13.  These calculations are rough
approximations and do not take into account other factors, such as oven
width and coal type, that affect coking time.
     The temperature also affects the performance of doors and end compo-
nents.  Each coke oven is in cyclic operation and all of its components are
subjected to thermal gradients and thermal cycling.   Most of the components
are restrained, so warpage and distortion of the end components occur.
Gaps, and their resulting leaks, occur from thermal  effects when the jamb
bows inward or hourglasses more than the seal  can accommodate or when the
seal edge warps because of thermal buckling.  Thermal damage to the end
components of the oven can also occur from flames caused by ignition of
leaking gases.  Exposure of the oven ends to rain increases the thermal
gradient through the end components and can increase the tendency for
thermal distortion.35
     3.3.3.5.3  Miscellaneous factors affecting emissions.  The accumula-
tion of deposits on the sealing surface of the jambs increases emissions
after the deposits harden.  The deposits may improve sealing when they are
still soft and the sealing edge can cut through the tar.  The deposits

     TABLE 3-13.  THE EFFECT OF TEMPERATURE ON THE RELATIVE COKING TIME
Oven wal 1
0 C
1,000
950
900
850
800
750
700
temperature
0 F
1,832
1,742
1,652
1,562
1,472
1,382
1,292
Coking times
(hours)
16.0
18.78
22.19
26.4
32.7
38.4
46.8
Percent of
capacity at 1,000° C
100.0
85.2
72.1
60.6
50.5
41.7
34.1
                                  3-54

-------
harden after continuous exposure to the normal jamb temperatures and turn
into an adhering form of hard carbon.  The deposit can then prevent uniform
contact of the seal edge with the door jamb and leave gaps where leaks
occur.35
     The time into the coking cycle also affects the door emissions.
Immediately after charging, the major portion of escaping gases and vapors
is steam.  The high steam content in addition to higher pressure result in
more severe emissions at the beginning of the coking cycle.  More of the
higher boiling temperature tars are removed later and they may seal gaps by
condensing and plugging the passages.35  The length of the coking cycle
also affects emissions.  The longer cycle time for producing foundry coke
(30 hr) as compared to that for metallurgical coke production (18 hr) re-
sults in a lower internal pressure and requires a lower coking temperature.
     The age of the battery and the maintenance program also affect the
leakage rate from the coke oven doors.  Doors, door seals, and jambs re-
quire periodic inspection, cleaning, and repair.  Damage is caused by
fires, temperature excursions, thermal distortion, buckling, or door hand-
ling equipment so that the door seal may not contact the jamb sealing
surface in all places.  One may expect that an older battery could have
more emission problems caused by expansion and shifting of brick, warping
of metal parts, and looser tolerances, particularly on old door handling
equipment.  Proper and precise control of the door handling equipment is
necessary to align the door correctly to accomplish the desired seal con-
tact.   If the equipment is worn badly with excessive tolerances, the seal,
door,  or jamb may be damaged or misaligned.   An uneven or low sealing
force, which is caused by a poor latching method, damaged equipment, or
inadequate adjustment and occurs on the door during mounting, leaves gaps
and increases emissions.
3.3.4  Topside Leaks and Their Emissions
     After a charge, the charging port lids must be replaced and the stand-
pipe lids must be closed to isolate the oven from the atmosphere.  Leaks
caused by improperly seated or distorted lids permit emissions to reach the
atmosphere.   The bodies of the standpipes can also leak.   Cracks or a bro-
ken seal at the bottom of the standpipe are the most common causes of leaks
                                  3-55

-------
in the standpipe body.   Topside emissions  are  expected  to  be  similar  in
composition to door leaks because both are composed of  coke oven  gas  that
is evolved during coking.
     An emission test was conducted at U.S.  Steel,  Clairton's Battery 1  by
EPA in August 1978.13  During the second hour  of coking, samples  were
collected from a vent pipe on a charging port  lid.   The leak  rate was
adjusted to yield small leaks with a 0.3-meter (1-foot) visible plume and
large leaks with a 1- to 2-meter (3- to 6-foot) visible plume.  The results,
listed in Table 3-14, show a range of 0.0017 to 0.0053  kg/hr  for  a small
leak, with an average rate of 0.0033 kg/hr.  Emissions  from the large leak
ranged from 0.012 to 0.035 kg/hr, with an average rate  of  0.021 kg/hr.   The
analysis for BaP showed that 1.4 to 1.8 percent of the  BSO was BaP.  The
emission rates for POMs ranged from 0.002 kg/hr for the small leak to
0.013 kg/hr for the large leak.
     The emission rate of 0.0033 kg/hr is a reasonable  lower  bound to
estimate the mass emissions in a small lid leak.  The emission rate of
0.021 kg/hr should be a reasonable upper bound to estimate emissions from a
badly leaking offtake with a 1- to 2-meter plume.  This range would under-
estimate emissions from multiple large leaks on the same lid  or offtake.
     Other estimates of topside leak emissions were based  on  a comparison
between the length of the sealing edge on lids and standpipe  caps and the
length of door seals.  This method was used as an approximation in the
absence of mass emissions data.  These approximations are  not presented in
this report because the topside tests previously discussed provide a more
accurate basis for estimates of mass emissions.
     The major factor affecting topside emissions is worker practice in
applying and cleaning luting mixtures to the lids and offtakes.  The luting
mixtures are prepared by the plant personnel and different consistencies
have been recognized as being essential for good sealing.   For example,
luting mixtures used for the standpipe cap are more viscous than  mixtures
for the lids.
     Another factor affecting topside emissions is pressure  fluctuation
from aspiration in the collector mains.  If these fluctuations are uncon-
trolled, they could result in charging lids or standpipe caps being unseated
and requiring topside workers to reseal the lids or caps.
                                  3-56

-------
                 TABLE 3-14.   TOPSIDE LEAK EMISSION TEST13
Leak size3
Large
Average
BSO (kg/hr) BaP (kg/hr)
0.002b
0.017
0.035
0.012 0.00022C
0.021
Small                              0.0017

                                   0.0029
                                   0.0053                        0.000072C
  Average                          0.0033
aA large leak yields a 1- to 2-meter (3- to 6-foot) visible plume.
 A small leak yields a 0.3-meter (1-foot) visible plume.

 Experimental run:   vent was partially plugged and flow was restricted.

CThese values are 1.4 to 1.8 percent of the BSO.
                                 3-57

-------
3.4  BASELINE REGULATIONS
     Current consent decrees and State Implementation Plans  require
varying levels of control for existing batteries.   In addition,  Occupational
Safety and Health Administration (OSHA) regulations require  equipment
and work practice controls for coke oven emissions to reduce worker
exposure, but do not set a performance level  in terms of visible emissions.
This section summarizes existing regulations  for wet-coal  charging, door
leaks, and topside leaks.
     Existing regulations vary from battery to battery with  the  most
stringent limits applied to new or rebuilt batteries.  Most  of the
existing batteries have current limits that follow the guidance  of EPA's
Reasonably Available Control Technology (RACT) with visible  emission
limits of about 25 seconds per charge, 10 to  12 PLD, 3 PLL,  and  10 PLO.
The limits associated with Lowest Achievable  Emission Rate (LAER) have
been applied to new and rebuilt batteries with visible emission  limits
of 11 seconds per charge, 5 PLD, 1 PLL, and 4 PLO.  The current  limits
in effect are collectively called the regulatory baseline and result
from State regulations and negotiated consent decrees.
     Current regulations are summarized in Tables 3-15 through 3-18.
For charging, approximately 125 batteries (93 percent of the total) have
existing regulations of 25 seconds per charge or less.  Approximately 38
of these batteries are given an exclusion for the highest observation in
a total of 20 in calculating control performance.   Approximately 14
batteries (10 percent of the total) are required to meet LAER limits of
11 to 12 seconds per charge.
     The regulations for door leaks in Table  3-16 indicate that 16 PLD
is the most lenient baseline limit and 4 PLD  is the most stringent.
Approximately 104 batteries have limits of 10 PLD or less; however, many
of these batteries are given exclusions of two to four doors.  A total
of 7 batteries have LAER limits of 5 PLD or less.
     The baseline limits for lid leaks range  from 1 to 5 PLL.  Approxi-
mately 84 batteries have limits of 3 PLL or less and 6 batteries have
the LAER limit of 1 PLL.  The limits for offtake leaks range from 4 to
10 PLO.  A total of 13 batteries have current limits equal to the LAER
limit of 4 PLO.
                               3-58

-------
       TABLE 3-15.  SUMMARY OF BASELINE REGULATIONS FOR CHARGING
Charging limit
(average seconds/charge)
32
30 (-1/20)3
25 (-1/20)3
25
20
19
12
11
Number of batteries
8
1
38
9
11
53
2
12
Percent of total
6.0
0.7
28.4
6.7
8.2
39.5
1.5
9.0
 Excludes highest observation in 20.
             TABLE 3-16.  SUMMARY OF BASELINE REGULATIONS
                            FOR DOOR LEAKS
Door leak limit
(percent leaking)
16
15
12
10 (-4)a
10 (-2)b
10
5
4
Number of batteries
2
21
7
20
50
27
6
1
Percent of total
1.5
15.7
5.2
14.9
37.3
20.2
4.5
0.7
aExcludes 4 doors.

 Excludes 2 doors.
                                  3-59

-------
TABLE 3-17.  SUMMARY OF BASELINE REGULATIONS
                FOR LID LEAKS
Lid leak limit
(percent leaking)
5
4
3
2
1
Number of batteries
44
6
28
50
6
Percent of total
32.8
4.5
20.9
37.3
4.5
TABLE 3-18.  SUMMARY OF BASELINE REGULATIONS
              FOR OFFTAKE LEAKS
Offtake leak limit
(percent leaking)
10
5
4
Number of batteries
84
37
13
Percent of total
62.7
27.6
9.7
                     3-60

-------
     A battery-by-battery listing of status and baseline regulations
compiled during the fourth quarter of 1984 is given in Table 3-19.   The
status was described as operating, hot idle (no production but the
battery is kept hot),  and cold idle (the battery has been cooled down
but could be restarted if needed).  Table 3-19 includes 43 plants and
134 total batteries; however, only 93 batteries at 36 plants were operating
in late 1984.
     OSHA is presently enforcing a set of regulations that are intended
to protect workers' health by limiting personnel exposure to emissions
from coke-producing operations.   The OSHA standard limits worker exposure
to 150 micrograms of benzene soluble fraction of total particulate
matter per cubic meter based on a time-averaged exposure and also mandates
equipment controls and work practices.   The OSHA requirements for charging
include:38
          Stage or sequential charging;
          Double main or jumper pipe;
          Written procedure;
          Adequate aspiration;
          Inspection and cleaning of goosenecks, standpipes, roof carbon
          buildup, steam nozzles, and liquor sprays;
          Larry car modifications; and
          Leveler bar seals.
For door leak controls, OSHA requires:38
          Written procedures,
          Inspection and cleaning,
          Door repair facilities,
          Adequate spare doors,  and
          Chuck door gaskets.
The OSHA requirements for topside leaks are:38
          Regular inspection, cleaning, repair, or replacement of
          equipment; and
          Prevention of miscellaneous topside emissions.
                                  3-61

-------
TABLE 3-19. BASELINE VISIBLE EMISSION LIMITS
No.
1


2

3




4

5
6


7



8


9

10
11

12


13

14




15





16

Plant
AL Byproducts, Tarrant, AL


Byire Coke, Holt, AL

toppers, Woodward, AL




LTV Steel, Gadsden,AL

LTV Steel, Thomas.AL
Jim Walters, Birmingham, AL


U.S. Steel, Fairfield, AL



National Steel, Granite City, IL


Interlake, Chicago, IL

LTV Steel .So.Chicago, IL
Bethlehem Steel, Bums Harbor, IN

Citizens Gas, Indianapolis, IN


IN Gas, Terre Haute, IN

Inland Steel, E. Chicago, IN




U.S. Steel, Gary, IN





LTV Steel, E.Chicago, IN

Battery
No.
A
5
6
1
2
1
2A
26
4
5
2
3
1
3
4
5
2
5
6
9
A
8
C
1
2
2
1
2
E
H
I
1
2
6
7
8
9
10
1
5
7
13
15
16
4
9
Status
0
1
0
0
0
0
0
0
0
0
0
0
2
0
0
0
2
2
2
2
0
0
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
1
1
1
0
1
2
2
Baseline Visible Emission
s/chg
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
11
19
19
11
11
11
11
25(1/20)
25(1/20)
11
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
PLO
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
10
15
15
5
10
10
10
10
10
10
10(-4)
10(-4)
10(-4)
10(-4)
10(-4)
10(-4)
10(-4)
10(-4)
10(-4)
10(-4)
10(-4)
1fl(-4)
10(-4)
10(-4)
10(-4)
10(-4)
10(-4)
10(-4)
10(-4)
10H)
PLL
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
2
5
5
1
5
5
5
5
5
5
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
Limits
PLO
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
5
10
10
4
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
                                                             (continued)
                                  3-62

-------
  TABLE 3-19. (continued)
No.
17

18








19



20

21


22
23


24
25


26
27
28





29






30

Plant
Arrrco Inc., Ashland, KY

Bethlehem Steel, Sparrows Pt., HO








Rouge Steel, Dearborn, HI



National Steel, Detroit, HI

Carandolet, St. Louis, HO


Tonawanda, Buffalo, NY
Bethlehem Steel, Lackawama, NY


LTV Steel, Warren.OH
Armco Inc., Middletown,OH


New Boston, Portsmouth, OH
toppers, Toledo, CH
LTV Steel, Cleveland.*





U.S. Steel, Lorain, OH






AL 8yproducts,Keystone,PA

Battery
No.
3
4
1
2
3
4
5
6
11
12
A
A
Ax
B
Ox
4
5
1
2
3
1
7
8
9
4
1
2
4
1
C
1
2
3
4
6
7
D
6
H
I
J
K
L
3
4
Status
0
0
1
1
2
2
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
2
2
2
2
2
2
2
2
2
Baseline Visible Emission
s/chg
25
11
32
32
32
32
32
32
32
32
11
20
20
20
20
20
20
20
20
20
30(1/20)
25
25
25
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
12
25(1/20)
25(1/20)
25(1/20)
20
20
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
25(1/20)
19
19
PLD
10
10
10
10
10
10
10
10
10
10
5
10
10
10
10
10
12
15
15
15
10
10
10
10
10
16
16
1Q(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
PLL
5
5
3
3
3
3
3
3
3
3
1
4
4
4
4
4
4
2
2
2
2
2
2
2
5
5
5
5
5
5
2
5
5
5
5
5
5
5
5
5
5
5
5
2
2
Limits
PLO
10
10
10
10
10
10
10
10
10
10
4
4
4
4
4
4
4
10
10
10
10
10
10
10
10
10
10
10
10
10
5
10
10
10
5
5
10
10
10
10
10
10
10
5
5
                           (continued)
3-63

-------
                                      TABLE 3-19.  (continued)
No.
31



32
33




34

35

36











37

38


39

40
41



42
43



Plant
Bethlehem Steel, Bethlehem, PA



LTV Steel, Aliquippa.PA
LTV Steel, Pittsburgh, PA




Koppers, Erie, PA

Shenango, Pittsburgh, PA

U.S. Steel, Clairton, PA











U.S. Steel, Fairless Hills, PA

Wheeling-Pitt, Monessen, PA


Chattanooga Coke, Chattanooga, TN

Lone Star Steel, Lone Star, TX
U.S. Steel, Provo, UT



Weirton Steel, Brown's Is.,WV
Wheeling-Pitt, E.Steubenville,WV



Battery
No.
A
2
3
5
A1
P1
P2
P3N
P3S
P4
A
8
1
4
1
2
3
7
8
9
15
19
20
21
22
8
1
2
1A
18
2
1
2
C
1
2
3
4
1
1
2
3
8
Status
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
0
0
0
0
0
2
2
2
0
0
0
0
0
0
0
0
0
2
0
0
0
0
Baseline
s/chg
19
19
19
19
19
19
19
19
19
19
19
19
11
19
19
19
19
19
19
19
11
19
11
19
19
11
19
19
19
19
19
19
19
12
19
19
19
19
25
25
25
25
25
Visible
PLO
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
W-2)
10(-2)
10(-2)
10(-2)
10(-2)
5
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
5
10(-2)
5
10(-2)
10(-2)
5
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
10
10
4
10
10
10
10
10(-2)
10(-2)
10(-2)
10(-2)
10(-2)
Emission
PLL
2
2
2
2
2
2
2
2
2
2
2
2
1
2
2
2
2
2
2
2
1
2
1
2
2
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
Limits
PLO
5
5
5
5
5
5
5
5
5
5
5
5
4
5
5
5
5
5
5
5
4
5
4
5
5
4
5
5
5
5
5
10
10
4
5
5
5
5
10
10
10
10
10
Status
0=online
1=hot idle
2=cold idle
3=under construction
(1/20)=excludes  highest  in  20.
(-2)=excludes 2  doors.
(-4)=excludes 4  doors.
                                3-64

-------
3.5  REFERENCES
 1.  Suta, B. E. Human Population Exposure to Coke Oven Atmospheric  Emissions.
     EPA Contract Nos. 68-01-4314 and 68-02-2835.  Revised May 1979.  p.  3,
     37, and 38.

 2.  Coke Producers in the United States in 1977.  U.S. Department of
     Energy, Energy Information Administration.  Washington, D.C. 1978.

 3.  Coke and Coal Chemicals.  U.S. Department of Energy, Energy  Informa-
     tion Administration.  Washington, D.C.  Energy data reports.  Decem-
     ber 1976, 1977, 1978, 1980, 1983.

 4.  Sheridan, E. T.  Coke and Coal Chemicals.  Minerals Yearbook.   U.S.
     Department of the Interior, Bureau of Mines.  1970-1975.

 5.  Sheridan, E. T.  Supply and Demand for United States Coking  Coals  and
     Metallurgical Coke.  U.S. Department of the Interior, Bureau of Mines.
     1976.  p. 15-18.

 6.  Kuljuian, N. J.  By-Product Coke Battery Compliance Evaluation.  EPA
     Contract No. 68-02-1321.  1.  Task 13.  June 1975.

 7.  McGannon, H. E. (ed.).  The Making, Shaping, and Treating of Steel.
     U.S. Steel Corporation.  Pittsburgh, PA. 1971.  p. 104-177.

 8.  Letter from Sparacino, C.,  Research Triangle Institute, to  Coy, D.,
     Research Triangle Institute.  January 12, 1979.  BaP Analysis—Coal
     Tar Samples A, B, and C.

 9.  Bee, R. W., et al.  Coke Oven Charging Emission Control Test Program--
     Volume I & II.  Office of Research and Development, U.S. Environmental
     Protection Agency.  Environmental Protection Technology Series.
     Publication No. EPA-650/2-74-062.  July 1974.

10.  Trenholm, A. R., L. L. Beck, and R. V. Hendriks.  Hazardous  Organic
     Emissions From Slot Type Coke Ovens.  U.S. Environmental Protection
     Agency.  1978.

11.  Kemner, W. F.  Cost Effectiveness Model for Pollution Control at
     Coking Facilities.  Publication No. EPA-600/2-79-185.  August 1979.
     p.  24, 30, 31.

12.  Benzene Soluble Organics Study - Coke Oven Door Leaks (Draft).  Clayton
     Environmental Consultants, Inc.  EPA Contract No. 68-02-2817.   EMB
     Report No. 79-CKO-22.   December 1979.

13.  Hartman, M.  W.  Emission Test Report - U.S. Steel Corporation,  Clairton,
     Pennsylvania.  TRW Environmental Engineering Division.  EMB  Report
     No. 78-CKO-13.  July 1980.   p. 3.
                                 3-65

-------
14.   Giunta, J.S.   U.S.  Steel Corporation Technical Report on Door Research
     Program.   Submitted to EPA, Division of Stationary Source Enforcement.
     p.  10, B-13.

15.   Herrick,  R.  A.  and L.  G. Benedict.   A Microscopic Classification of
     Settled Particles Found in the Vicinity of a Coke-Making Operation.
     Journal of the Air Pollution Control Association.  19:325-328.  May 1969.

16.   Stolz, J.  H.   Coke Charging Pollution Control Demonstration.  American
     Iron and Steel  Institute and Office of Research and Development, U.S.
     Environmental Protection Agency.   Environmental Protection Technology
     Series.  Publication No. EPA-650/2-74-022.  March 1974.  302 p.

17.   Department of Health,  Education,  and Welfare, National Institute for
     Occupational  Safety and Health.  Criteria for a Recommended Stan-
     dard—Occupational  Exposure to Coke Oven Emissions.  1973.

18.   Smith, W.  M.   Evaluation of Coke Oven Emissions.  Yearbook of the
     American Iron and Steel Institute.   1970.  p. 163-179.

19.   Cuscino,  T.  A.   Particulate Emission Factors Applicable to the  Iron
     and Steel  Industry.  Publication No. EPA-450/4-79-028.  September
     1979.   p.  54.

20.   White, et al. Atmospheric Emissions of Coking Operations—A Review.

21.   Allen, C.  C.   Estimation of Charging Emissions for By-Product Coke
     Ovens.  EPA Contract No. 68-02-3056.  RTI No. 1736/2/02.  March 1980.

22.   Giunta, J.  S.  Door Emissions Control Technology.  (Presented at IISI
     Meeting,  Fourth Working Session:  Air Pollution Control Related  to
     Coking Operations.   Chicago, IL.   June 11, 1979).

23.   Schultz,  L.  D.   Control of Coke Oven Door Emissions.  Allied Chemical
     Corporation.   (Presented at 84th National Meeting of AIChE.  Atlanta,
     GA.  February 26-March 1, 1978).

24.   Baldwin,  V.  H.  and D.  W. Coy.  Study to Develop Retrofit  Information
     and Other Data for Use in Setting Standards for Coke Oven Emissions.
     EPA Contract No.  68-02-2612.  Task 39.  March 1978.

25.   Mabey, W.  R.   Identity and Chemical and Physical Properties of  Com-
     pounds in Coke Oven Emissions.  EPA Contract No. 68-01-4314.  Septem-
     ber 1977.

26.   Barrett,  R.  E., et al.  Sampling and Analysis of Coke-Oven Door Emis-
     sions.  Publication No. EPA-600/2-77-213.  October 1977.

27.   Barrett,  R.  E.  and P.  R. Webb.  Effectiveness of a Wet Electrostatic
     Precipitator For Controlling POM Emissions From Coke Oven Door  Leakage.
     (Presented at 71st Meeting of APCA.  Houston, TX.  June 25-29,  1978).
                                 3-66

-------
28.  Participate Polycyclic Organic Matter.  National Academy of Sciences.
     Washington, D.C.  1972.  p. 5-12.

29.  Webb, P. R. and R. E. Barrett.  Stack Emission Sampling At Wisconsin
     Steel Company Coke Oven Plant, Chicago, Illinois.  Report No.  77-CKO-ll.
     EPA Contract No. 68-02-1409.  Task 50.  November 4, 1977.  21  p.

30.  Air Pollution Emission Test, Final Report, Wisconsin Steel.  Clayton
     Environmental Consultants.  Report No. 77-CKO-ll-A.  EPA Contract  No.
     68-02-2817.  May 1978.

31.  Coke Oven Emission Testing - Armco Steel Corporation, Houston,  Texas
     (Draft).  TRW Environmental Engineering Division.  EPA Contract
     No. 68-02-2812.   (Test conducted October 1979).

32.  Source Testing of a Stationary Cokeside Enclosure—Bethlehem Steel
     Corporation, Chesterton, Indiana.  Clayton Environmental Consultants,
     Inc.  EPA Contract No. 68-02-1408, Task 10.  February 9, 1977.   p.  18.

33.  Allen, C. C.  A Model to Estimate Hazardous Emissions from Coke Oven
     Doors.  EPA Contract No. 68-02-3056.  RTI No. 1736/2/01.  March 1980.

34.  Stanley, R. W.  Coke Oven Door System:  Field Data.  U.S. Steel.
     (Presented at 84th Meeting of AIChE.  Atlanta, GA.  February 26-
     March 1, 1978).

35.  Lownie, H.  W., et al.  Study of Concepts for Minimizing Emissions  from
     Coke-Oven Door Seals.  Publication No. EPA-650/2-75-064.  July 1975.

36.  Graham, J.  P. and B.  P. Kirk.   Problems of Coke-Oven Air Pollution
     Control.  The Metals Society.   London SWIY 5DB.  p. 82-100.

37.  Lowry, H. H. (ed.).   Chemistry of Coal Utilization.  Supplementary
     Volume.  John Wiley and Sons,  Inc.  NY.  1963.

38.  Allen, C. C., et al.   Comparison of OSHA's and Potential EPA Regula-
     tion For Coke Oven Batteries.   EPA Contract No. 68-02-3056.  RTI No.
     1736/2-02S.  Revised May 1980.
                                  3-67

-------
                     4.  EMISSION CONTROL TECHNIQUES

     This chapter discusses the technology for control of emissions from
wet-coal charging, oven door leaks, and topside leaks during coking.
4.1  TECHNOLOGY FOR THE CONTROL OF EMISSIONS FROM CHARGING
     Charging practices have been changed by the current efforts of regula-
tory agencies and coke oven operators to reduce emissions.   Previously, 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 operating procedures than
on the design of charging equipment.   Consequently, the following discussion
will concentrate on these procedures as well as the required equipment
modifications.   These control systems (or procedures) are stage charging,
sequential charging, and wet scrubbers mounted on larry cars.
4.1.1  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 certain
equipment requirements, these do not include conventional air pollution
control  devices such as fabric filters and scrubbers.
                                  4-1

-------
     Stage charging is the ordered pouring of coal  into the oven so that an
exit space for gas is maintained consistently regardless of coal flow.   The
larry car hoppers are discharged in an ordered sequence so that an open
tunnel head is maintained at the top of the oven until  the last hopper is
discharged.   Emissions are effectively contained in the ovens and collect-
ing 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.
     4.1.1.1  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 apply-
ing steam aspiration in the goosenecks of the offtake.  The assembly that
comprises the standpipe and gooseneck is often called the offtake or ascen-
sion pipe.  The stage charging technique is uncomplicated but requires
close attention to detail.  Figure 4-1 illustrates the stage charging
technique and presents the case with four charging holes and aspiration at
two offtakes.  Where there is only one active offtake, only one charging
port will be open at a time, to avoid pressure imbalance.2  The most impor-
tant aspects of stage charging are good aspiration and the operating crew's
strict adherence to specific charging and leveling practices.  These prac-
tices, as observed by EPA at United States Steel Corporation at Clairton,3
are as follows:
     Step 1.   After the gooseneck (the connection between the top of the
standpipe and the collecting main) is cleaned, the loaded larry car is
positioned over the oven to be charged as shown in Figure 4-1.  The drop
sleeves (or boots) under the hoppers are lowered around the previously
opened charging holes, the damper between the offtake and main is opened,
and the steam for aspiration in the gooseneck is turned on.  At this time,
coal charging can begin.
     Step 2.   Coal is simultaneously discharged into the oven from Hoppers
1 and 4.   (At some plants, these hoppers are emptied one at a time.)  When
these hoppers are empty, their slide (or shear) gates are closed, the drop
sleeves are raised, and the charging hole lids are replaced.  (Some plants

                                  4-2

-------
                 ~rJ-.   STAGE CHARGING
                  POSITIONING CAR PRIOR TO CHARGING
                   Noil niyino. tmouiiu ol cul m koppttt
                                                      ^rNs-X^^S : - J
                                                     «teC»'-Ai:iTXf^.U.-.>5r.Vs:y'£i -1---J
ff=frrn
\  COM I»«»U j-i       /"
                                                     DISCHARGE N*.11 ( HOPPERS
                        STAGE CHARGING
                      DISCHARGE No 3 HOPPER
                                                       DISCHARGE No.2 HOPPER
                                                       AND LEVELING OF COAL
Figure 4-1.  Steps in stage charging coal into a coke oven with four charging holes.
                                   4-3

-------
do not raise drop sleeves or replace lids until  all  the hoppers are empty.)
The discharge of these two hoppers is illustrated in Figure 4-1.   The
height of the coal does not extend above a designated coal  line for the
oven.  This limit ensures that the open space at the top of the oven remains
clear so that the effect of the aspiration is not blocked from any portion
of the oven.
     Step 3.  Hopper 3 is discharged as shown in Figure 4-1.  Again, the
peak level of the coal is below the designated coal  line for the oven.
Immediately after this hopper is emptied, its slide  gate is closed, the
drop sleeve is raised, and the charging hole lid is  replaced.
     Step 4.  Hopper 2 is the last to be discharged.  The coal from this
hopper will peak to the top of the oven (thereby blocking the open space)
and move up into the charging hole as shown in Figure 4-1.   After the coal
flow stops, the chuck door is opened, and the leveler bar is inserted to
level the coal (and reestablish the open space).  This process also permits
the  last bit of coal in the hopper to flow into the  oven.  A leveler boot,
which is one of the equipment requirements for stage charging, is used to
seal the opening between the chuck door and leveler bar.  When Hopper 2 is
empty, its slide gate is closed, the drop sleeve is  raised, and the charging
hole lid is replaced.  The leveler bar then makes about two more passes to
assure complete leveling of the coal charge.  (This  procedure also assures
enough open space at the top of the oven for gases to pass to the collecting
mains during the coking cycle.)  The leveler bar is  retracted, the chuck
door is closed, and the aspiration steam is turned off.
     All of the batteries at U.S. Steel, Clairton have double mains.  At
the U.S. Steel, Fairfield batteries with single mains, a similar procedure
is followed except a jumper pipe joining the oven on charge to an adjacent
oven must be connected before the charge starts.1
     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.
Figure 4-2 shows a typical aspiration system.4  Steam is the common aspira-
tion fluid, although it has been reported that liquor sprays can achieve
stronger aspiration.5  Other advantages reported for the liquor sprays are
less frequent cleaning of nozzles and less steam, condensate, and heat in
                                  4-4

-------
                                  O

                                  •a


U
o
! :

i
o.
iJ &•
I !
*^ "^
flj *
5 5
<
« ^_i 	
^

x
o
U Jt
H
u w
u J3
C'
ac
IJ



01
>
0
* ^ (uojjpj-jdsv)
" J y .^- 3JBJQ paanpui
•„ 0 ll T / 	
0 3 ,
w 5
00 —
C
•^ *a£
c. n
M4 O
c. a
r^»
!
i










__^_



                                  « tS
                                  o  c
                                  E.2
U
r^
w
u
**
N
|g
O
z


*•/•' .',' i

; ~l
s

-O k>
2
w o
3 4)


C
•«
                                  a
                                  C/l
                                  
-------
the effluent from the mains.   However,  aspiration efficiency of steam or
liquor sprays is sensitive to the flow and pressure maintained at the
nozzles.   Some examples of steam pressures used are 700 to 800 kPa (90 to
115 lb/in2) at U.S.  Steel, Fairfield,6 900 to 1,100 kPa (120 to 150 lb/in2)
at CF&I Steel Corporation,7 and 850 to 1,000 kPa (110 to 130 lb/in2) at
U.S.  Steel, Gary.8  These plants use steam nozzles with inside diameters of
1.4 cm (9/16 in),9 0.8 cm (5/16 in),10 and 1.9 cm (3/4 in),11 respectively.
Figure 4-312 shows the effect of the steam pressure and nozzle size on
aspiration capacity.   The 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 steam 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 collect-
ing mains.  The coal  dust in the main creates potential problems with the
recovery of by-products, especially tar. Therefore, aspiration systems are
carefully designed to provide just enough draft on the oven to prevent
emissions during the charging cycle.  Extra openings at the oven being
charged must be avoided.
     Timely removal and replacement of the lids on the charging ports is
crucial to good stage charging; manual or automated methods can be used.
(The most common manual method uses hydraulically operated electromagnets.)
Either method can be effective, although the automated system may reduce
the incidence of improper alignment of the lids.  Any mechanical system
must allow the lids to be moved individually.
     Typically, shrouds or sleeves (moveable tubes which connect the larry
car hopper to the charge hole through which the coal falls) are used to
prevent air filtration at open charging ports.  However, another system
which was observed by EPA is much more effective.13  At the Ruhrkold,
Zollverein Plant in Essen, Germany, the "Still-ERIN" charging car has small
discharge hoppers mounted below the normal larry car hoppers (Figure 4-4).
An extension from the bottom of each discharge hopper is tapered at the
lower end.  This taper mates with the charging hole to provide a very
effective seal which prevents air infiltration or escape of emissions.  To
minimize escape of emissions through the hoppers, a level controller in the
                                  4-6

-------
 1200



 1100




 1000



I 900
-.800
>•
£ 700
                              NOZZLE PRESSURE, kg/cm2, gauge

                                 3456
  600
K
   500




   400



   300



   200
                                                                                           30
                                                                                           25  1
                                                                                              <«•>
                                                                                               E

                                                                                               tl
                                                                                           20
                                                                                               u
                                                                                               o

                                                                                           15  <
                                                                                           10
                  20
                                40            60           80

                                    NOZZLE PRESSURE, Ib/in2, gauge
100
120
         Figure 4-3.  Aspiration capacity vs nozzle pressure for  single standpipe.12
                                              4-7

-------
LEVEL MEASURING INSTRUMENT
         BUTTERFLY VALVE
                                                         CHARGING HOPPER
                                                        DISCHARGE HOPPER
                                                       TAPERED DISCHARGE RING
               OVEN TOP
                                                      CHARGING HOLE
                       Figure 4-4. "Still-ERIN" charging system.13
                                  4-8

-------
discharge hopper closes a butterfly valve to leave a "head" of coal in each
discharge hopper after its coal is discharged.   The small amount of coal
remaining on top of the valve helps seal the hoppers from the oven; the
butterfly valves are not expected to be airtight.  Each discharge hopper
operates independently.
     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 to seat 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.  Figure 4-512 shows the effect of carbon deposits in the
gooseneck on aspiration capacity.
     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.  Figure 4-6
shows the gas flow pattern from the oven.  The restriction to this flow is
increased if the roof of the oven is lowered toward the coal by a heavy
build-up of carbon.   Typically, this roof carbon is removed by blades which
are mounted on the top of the pusher ram and scrape the carbon off as the
coke is pushed.  High-pressure air jets mounted at the top of the ram are
sometimes used to provide better cleaning.  Even with all other factors at
their best, failure to remove excessive roof carbon may result in poor
emission control.
     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 condi-
tion can best be achieved with two mains, standpipes, and aspireition systems
                                  4-9

-------

                                      (M


                                       in
                                  c
                                  01
                              o   <->
                                 O
                                 _l

                                 u.

                                 CC
                                 5

                              <*  X

                                       I
'V3UV
      4-10

-------
                                           M
                                           '(5

                                           ?
                                           '€
                                           _»
                                           8
                                           W)
                                           a
                                           O
                                           4->


                                           I
                                           O
                                           0)
                                           a
                                           E
                                           o
                                           co
                                           c
                                           o
                                          1
                                          '5.
                                          (6
                                           O)
                                          LE
4-11

-------
(one at each end of the ovens) as shown in Figure 4-6.   If the coal  blocks
the open space at the top of the oven (e.g.,  when coal  is charged from the
last hopper just before leveling) each side of the blockage will  remain
under vacuum.   (Alternate techniques to provide aspiration to both sides of
an oven that has only one standpipe are discussed later.)
     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 just 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 is influenced by the bulk density of the coal, oil additives,
the moisture content, and the particle size to which the coal is ground.
For one domestic plant, 33.5 percent of the total charge is placed in both
Hopper 1 and Hopper 4, 20 percent is placed in Hopper 2, and 13 percent is
placed in Hopper 3.  These percentages vary from plant to plant,  depending
on the previously mentioned factors.  A given plant may have to alter its
scheme occasionally as coal properties change.  The heights of the coal
peaks should be checked periodically.
     Just before the coal is charged, the larry car must be 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.
     Although independently charging the coal from each hopper gives the
most assurance of successful emission control, an alternative technique has
been witnessed at U.S. Steel, Gary where the two outside hoppers are charged
simultaneously.8  The remaining hoppers are then charged one at a time.
The distance between the end charging ports on four- or five-hole ovens
allows only minimal interference with formation of the coal pile at one end
of the oven by coal entering the other end of the oven.  (This minimal inter-
ference may not occur on an oven with only three charging holes.)  However,
the rate at which gases are aspirated from the oven may need to be increased.
                                  4-12

-------
The larger quantity of coal entering the oven for the same length of time
will result in a higher rate of gas generation.   At the National Coal Board
Avenue Plant in Sheffield, England, this technique and charging each hopper
one at a time were both observed.14  There was no apparent difference in
the degree of control achieved, perhaps because the flow of aspirating
steam was more than adequate.  This tradeoff between consumption of aspirat-
ing steam and the duration of the charging process must be resolved by the
owner.
     A production benefit of charging the two outside hoppers simultaneously
is that it speeds up the charging operations.  Figure 4-7 shows the differ-
ence in time to use stage or conventional charging procedures.  Although
the data are limited and charging time may vary from plant to plant, the
data indicate that conventional charging has a comparative advantage over
stage charging in terms of reduced charging time. However, at domestic
plants, the charging operation is not typically the limiting factor in
attaining higher production rates on the smaller 3- or 4-meter ovens.
Pushing, door cleaning, or quenching operations are more often the limiting
factors; consequently, the benefit of reduced charging time may not be
significant.  However, the increased time required for stage charging may
be significant for the taller 5- or 6-meter batteries containing over 70
ovens that are serviced by only one larry car.
     Gravity is commonly used to induce coal flow in stage charging;  how-
ever, mechanical feeders are also used.  A variable speed drive on the
feeders would allow the charging rate to be controlled.
     Leveling begins before the last hopper is emptied and the last lid is
replaced.   During leveling of the coal charge, leakage must be controlled
at the chuck door where the leveler bar enters the oven.   Air induced
through an open chuck door can completely overpower the aspiration on that
end of the oven.   (A charge was observed at the Sheffield, England, plant14
where the chuck door was inadvertently left open.  All of the air pollution
control benefits of the stage charging used at this plant were lost.)  In
addition to maintaining the normal  closing of the chuck door, the leveler
boot sealing system previously discussed is available to minimize air
intrusion when the leveler bar is inserted.
                                  4-13

-------
  4.0
2*3.5 —
c
1-(
B
  1.0
to
G
•*•* "> 5
60 ^ »3


-------
     The value of aspiration from both ends of the oven has been emphasized.
Use of double mains is the best way to achieve this type of aspiration and,
for construction of new batteries, this approach will be used most.   However,
there is an alternative method of drawing a vacuum at both ends of an oven.
In addition to the normal aspiration obtained from an oven with a single
main, a jumper pipe can be placed between the end of the oven opposite the
main and a neighboring oven (ordinarily two ovens away).  Aspiration applied
to the neighboring oven will effect a vacuum on the oven being charged.  To
provide maximum suction through the jumper, the neighboring oven should be
near the end of its coking cycle when the least amount of gas is evolving.
One or two ovens at one end of the battery may not be served by this proce-
dure, but permanent jumpers may be installed to overcome this problem.
     As presented in a recent report,15 several alternative designs for a
jumper are listed below.
          Connect the jumper between "smoke holes" (extra holes placed in
          the top of each oven opposite the main) on the oven being charged
          and the neighboring oven.
          Connect the jumper to a charging hole and discontinue its use for
          charging.  Usually the third hole from the main (in a four-hole
          oven) is used.   In this design and in the following designs, the
          other end of the jumper duct connects to the corresponding charg-
          ing hole on a neighboring oven.
          Manifold two or more charging holes via the shrouds or hoods
          around the holes and connect the jumper to this manifold.
          Connect the jumper to the shroud around a single charging hole
          (usually the farthest hole from the main).
The latter three alternatives can be used on the last oven in a battery
only if the jumper is designed to extend from either end of the larry car.
     The batteries at U.S. Steel, Fairfield, which EPA visited, use the
third alternative on a battery containing ovens with four charging holes.15
The first and fourth drop sleeves of the larry car are manifolded and the
jumper is attached at right angles to the manifold at the drop sleeve
farthest from the main.   A jumper can be seen on the right side of Figure
4-8.   At this plant, the  jumper is connected after the drop sleeves  are
lowered and aspiration is turned on at the oven to be charged and at the
neighboring or "jumper"  oven.   The jumper is connected to the fourth charg-
                                  4-15

-------
Figure 4-8.  Jumper pipe.
         4-16

-------
ing hole on the second oven from the one being charged and is not removed
until all lids have been replaced on the oven being charged.   Other than
connection of the jumper, the stage charging procedures are similar to
those on a double-main battery.
     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 leakage of air into the system, can
usually decrease the problem of coal fines.  The primary cooler system
usually removes tar "sludge" containing the coarser coal particles.16
     The increased density and reduced heating value of the coke oven gas,
resulting from aspiration, affect burner operation on some ovens.  Good
coke oven gas has a heat content of 21 MJ/m3 (550 Btu/ft3).  One plant
reports that the heat content of their gas was 15 MJ/m3 (400 Btu/ft3).17
For burners designed with little tolerance, problems in maintaining oven
temperature may occur.  Careful  operation and use of smoke boots on the
leveler bar (leveler boot) and drop sleeves to seal openings to the atmos-
phere will make these problems manageable.
     4.1.1.2  Equipment and Engineering Requirements.  Some engineering
problems are encountered when existing batteries are retrofitted to stage
charging.  These problems are not easily categorized because coke oven bat-
teries have been traditionally constructed on an individual basis with few
attempts at standardization or repetition of construction details.  One of
the most difficult problems encountered is batteries with inadequate clear-
ance between the coal  bunkers and the battery top; this prevents the simple
modification of the hopper size  essential to stage charging.   However,
other battery modifications can  be made, though they may be costly.
                                  4-17

-------
     A list of major equipment modifications is given in Table 4-1.   Al-
though larry car modifications and aspiration equipment are required at
most batteries, not all batteries will  have to make each modification or
addition listed in Table 4-1.   In one survey of 122 batteries at 30 plants,
no one took the position that they could not modify or retrofit for stage
charging.  The unanimous verdict of all  sources contacted was that retrofit-
ting for stage charging is a sometimes  difficult engineering feat, which
will be limited only by economic considerations.17
     Some examples of retrofit problems  identified in that same study are
discussed in this section.  As mentioned previously, the most difficult
equipment modification may be associated with the coal bunker.  The larger
coal capacity in the outer hoppers of the larry car is usually accomplished
by increasing the height of the hoppers, but the larry car must move under
the bunker to be filled.  In some cases, it is not possible to design a
modified larry car that will fit under  the bunker.  To solve this problem,
the bottom structures of some bunkers have been modified to obtain the
necessary vertical clearance.   If this  method is not feasible, then the
solution is to jack up the bunker—a more drastic and costly procedure.17
     To obtain the necessary volume in  the outer hoppers, the cross section
can be changed to an oval, rectangular,  or "bathtub" shape.   Another
technique is to increase the volume and  supply screw feeders.17
     Many older larry cars are equipped  with unsuitable mechanisms for drop
sleeves.  Such mechanisms operate the drop sleeves of all hoppers either
partially or entirely in common.  To achieve effective stage charging, the
operators of the larry car must be able  to control each drop sleeve independ-
ently.  The former system can be modified as desired, without replacing the
larry car.
     Each specific battery must solve its own particular problems with
respect to coal flow which will vary with coal type and climate.  Charging
options include feeding by gravity, turntable, or screw.  The addition of
oil to the coal to stabilize the bulk density and use of vibrators in the
hoppers are other options.
     An increase in aspiration to draw the smoke into the gas collecting
system is often required.  This increase is accomplished by increasing the
pressure of the steam and/or the size of the steam nozzles.  Frequently the
                                  4-18

-------
           TABLE 4-1.   EQUIPMENT REQUIREMENTS FOR STAGE CHARGING

Larry car modifications

     Independently operated drop sleeves
     Independent hopper control of coal flow
     Hoppers with sufficient excess capacity to accept redistribution
       of the coal load
     Automatic lid lifters or access to the charging holes without
       moving the car
     Adjustments on the larry car hoppers to assure accurate coal
       volumes

Aspiration equipment

     Double drafting with either a second collecting main or a jumper
       pipe for single collecting main batteries
     Steam headers, lines, and nozzles
     Mechanical gooseneck cleaner
     Decarbonizing air and scraper
     Leveler bar seal
     For some batteries, increased boiler capacity, standpipe
       straightening,  and increased exhauster capacity (to handle  the
       increased load caused by aspiration) are necessary

Battery modifications (not required for all batteries)

     Modify coal bunker to accommodate new larry car
     Reset or replace lid rings
     Repave top and realign larry car tracks
                                  4-19

-------
increased aspiration requires larger steam headers and occasionally it
requires new boilers.   Adequate aspiration also requires an appropriate
type of gooseneck and damper.  The gooseneck may have to be replaced,  and
the older Corliss valves will usually have to be replaced.   This valve
design restricts aspiration and is easily rendered inoperable by the carbon
and tar build-up that falls into it when the goosenecks are cleaned.17
     Another retrofit consideration is the possibility of adding a second
collecting main instead of using a jumper pipe for double drafting.  A
second collecting main might provide slightly superior aspiration because
the jumper pipe would be connected to an oven that still contains coke.  In
1978 only two batteries had been retrofitted with second mains.   According
to Koppers and Wilputte, the two major coke oven builders,  the construction
problems may be very difficult.  Both of these retrofitted batteries had
existing brickwork which was readily adapted for a second offtake.  Some
ovens do not have this feature because of a flue which crosses over the top
of the oven and makes the installation of a second offtake difficult.18
     The second main requires structural support and has to be fitted into
the space occupied by the larry car hot rails and other equipment.  Supply-
ing the flushing liquor to the second main is another consideration.
Because the collecting and crossover mains carry both liquid and gas,  the
liquid must be drained from the main and around the end of the battery to
merge with the existing crossover mains.  The demand for a double supply of
flushing liquor may require additional pumps and a new supply system.
Extensive and elaborate control systems may also be required to maintain
the delicate pressure balance needed to prevent gas from recirculating in
the ovens.18
     4.1.1.3  Optimizing Stage Charging.  The basic operating techniques
and equipment requirements for stage charging have been discussed in detail
in previous sections.   Each specific battery may need to incorporate other
changes to "fine tune" their performance at stage charging.  Battery condi-
tions, worker attitude, and training vary at each location.  An engineering
study and evaluation at each individual battery or coke plant may be neces-
sary to determine the required minor improvements.  At CF&I, such a study
resulted in improved written procedures, better worker training, and a few
minor additional equipment changes to demonstrate exemplary control of
                                  4-20

-------
charging emissions.19  U.S. Steel also conducted an extensive engineering
study at their Clairton batteries to optimize emission control from stage
charging.12
     The performance of the battery top workers is critical to achievement
of exemplary control of charging emissions.  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.
     A study of a specific battery's operation would aid in developing an
optimum written procedure and effective training program.  In addition, the
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.19
     After the engineering study and the physical  changes that were made to
allow stage charging, U.S.  Steel at Clairton implemented extensive training,
observation, and monitoring programs.   Each member of the charging crew was
individually trained in the stage charging procedure, and following the
training session, each crew member was observed to determine if additional
training was required.   Additional  personnel  were  hired to monitor visible
emissions,  to investigate unsatisfactory charges,  and to identify and
correct the causes of poor charging performance.   U.S. Steel  attempted to
increase motivation and improve working relationships through the use of
                                  4-21

-------
awards and a crew/foreman concept.   U.S.  Steel  awarded employees who exhib-
ited superior performance in their smoke  abatement efforts.   The crew/fore-
man concept consisted of assigning the foreman  and crew members to the same
5 working days.   An additional supervisor was added to each line of operat-
ing batteries to provide increased supervision  for the topside operation.
     4.1.1.4  Performance of Optimized Stage Charging.  Mass emissions from
stage charged ovens have not been measured because of the complexities and
expense that such measurements would entail.  Consequently, the performance
of emission controls for charging must be judged on the basis of visible
emissions.  EPA gathered data by measuring the  cumulative length of time
that emissions were visible during charging.  This method is described in
Appendix D.
     The batteries at CF&I and U.S.  Steel, Clairton have implemented the
equipment modifications and operating procedure for stage charging.  In
addition, the technique was optimized by conducting an engineering study
and by implementing the items described in the  preceding section.
     Visible emission data for these two batteries are summarized in
Table 4-2.  As emphasized in earlier sections,  the degree of control
achieved with stage charging depends on operating procedures and many other
factors.   The data show that the level of stage charging control varies at
U.S. Steel, Clairton where all batteries are physically equipped to use
stage charging.   The batteries at CF&I and U.S. Steel, Clairton have under-
gone stage charging optimization, which includes an engineering study,
detailed training and operating procedures, visible emission observers who
provide feed-back to supervisory personnel, and minor equipment changes.
The cause of the variability could not be isolated and it appears to result
from normal battery-to-battery variations.  Subtle differences in construc-
tion features, equipment, or worker effort could contribute to the vari-
ability.   Another possible influence that is difficult to identify or
measure is the battery operator's incentive to  control emissions at a level
that is lower than the legally required level.   For example, the Clairton
batteries in the data base, except for Battery 20, have an existing regula-
tion of 75 seconds of emissions for four charges, or about 19 s/charge.
Table 4-2 indicates that the average performance of each of the Clairton
batteries is much less than 19 s/charge.
                                  4-22

-------
  TABLE 4-2.   SUMMARY OF VISIBLE EMISSION DATA ON OPTIMIZED STAGE CHARGING
Company
CF&I, Pueblo7
U.S. Steel, Clairton20













Battery
B
I
2
3
7
8
9
10
11
16
17
19
20
21
22
Date
7/77
2/80
8/79
2/80
2/80
2/80
2/80
2/80
2/80
8/79
8/79
2/80
8/79
2/80
8/79
Average
(s/
charge)
5.4
8.6
9.0
7.8
7.6
8.9
8.4
8.8
11. 6a
1.0a
1.2a
6.8
12.0
13.1
6.6
Range
(s/
charge)
0-25
2-82
2-23
3-15
3-12
3-16
3-20
2-46
3-29
0-13
0-10
2-14
2-127
1-96
1-41
Number of
charges
92
20
55
16
16
16
16
16
16
28
32
16
65
18
62
Run-of-the-mine (nonpulverized) coal was charged.
                                4-23

-------
     The batteries of CF&I and U.S.  Steel, Clairton have double collecting
mains and are less than 5 meters in height.   Because batteries vary within
the industry, these batteries do not represent all  of the potentially
important construction features that may affect emissions.   According to
available data, no other plants have implemented a detailed program of
optimization like that reported by CF&I and U.S. Steel, Clairton.   However,
batteries that had implemented the basic stage charging equipment and
operating procedure were identified.
     The data for batteries that have implemented the basic aspects of
stage charging are summarized in Table 4-3.   The battery at Bethlehem,
Burns Harbor is 6 meters in height with a double collecting main, four
charging holes, mechanical feed, and magnetic lid lifters.   The batteries
at Jones & Laugh!in (J&L), Pittsburgh have short ovens (less than 5 meters)
with three charging holes, a single collecting main, a jumper pipe, and
mechanical and gravity feed for charging.  The battery at Lone Star Steel
is a tall battery (5 meters in height) that was constructed in 1979 with a
double collecting main.  The battery at National Steel, Weirton is 6 meters
in height with a double collecting main, mechanical feed, and automatic lid
lifters.  The batteries at Shenango, Inc. have short ovens with three
charging holes, a double collecting main, gravity feed, and an extended
30-hour operating cycle.  The batteries at U.S. Steel, Fairfield are short
batteries with a single collecting main, gravity feed, and four charging
holes.  Battery 9 at U.S. Steel, Fairfield was recently rebuilt in 1979 and
it has a mechanical gooseneck cleaner.  Battery 1 at U.S. Steel, Gary is a
6-meter battery which was retrofitted with a second collecting main.  This
battery has  four charging holes, magnetic lid lifters, and mechanical feed
for  charging.  The batteries in Tables 4-2 and 4-3 represent the following
industry variations which may affect emissions.
          Koppers, Koppers-Becker, and Wilputte ovens.
          Ovens that are 3 to 6 m in height.
          Single and double collecting mains.
          Three or four charging holes.
          Older (after 1951), rebuilt, and new batteries.
          Gravity and mechanical feed.
                                  4-24

-------
   TABLE  4-3.   SUMMARY  OF  VISIBLE  EMISSION  DATA  ON  BASIC  STAGE  CHARGING
Company
Bethlehem
Harbor21

Steel , Burns

J&L, Pittsburgh22


Lone Star


Steel23
National Steel, Weirton24

Shenango,

U.S. Steel



U.S. Steel


Neville Island25

, Fairfield



, Gary8 "

Battery
2a

P2

P4
Cb
la

3
4
6


9
la

Date
5/80

12/78
8/79
8/79
10/79
8/81
6/82
10/80
10/80
9/7526
12/7626
6/7927
6/7927
6/76
11/77
Average
(s/
charge)
10.

9.
5.
6.
7.
13.
10.
6.
6.
8.
9.
10.
5.
10.
14.
8

5
4
2
4
5
2
1
6
0
9
0
4
5
6
Range
(s/ Number of
charge) charges
1 -

4 -
0 -
2 -
2 -
3 -
2 -
1 -
3 -
3 -
1 -
2 -
1 -
0 -
0 -
43

29
14
11
15
43
28
23
11
24
185
34
38
44
150
25

8
9
10
30
21
20
20
20
15
91
20
50
41
65
 A 6-meter battery.
3A new 5-meter battery.
                                4-25

-------
          Manual and automatic lid lifters.
          Different geographical  locations and times of the year.
          Variety of coking coal  sources.
          Coal sizes include nonpulverized and 65 to 81 percent less than
          0.3 cm (0.125 in).
A discussion of these features and possible effects on the control of
charging emissions follows the data analysis.
     The charging data were collected by EPA personnel or their contractors
during official EPA inspections.   Most of the EPA inspections were conducted
to determine the battery's compliance status in relation to existing State
regulations and consent decrees;  a few inspections were conducted to docu-
ment emission control performance only.   Preference was given to data
collected during official EPA inspections because operating conditions,
equipment failures, and deviations from good work practice were generally
noted on raw data sheets or in written reports.  Also, during the enforce-
ment inspection, it is assumed that the source is trying to achieve good
performance.
     All of the data presented in Tables 4-2 and 4-3 were taken under
conditions which were considered normal for the batteries that were in-
spected.  The only data points that were deleted from the summary in
Tables 4-2 and 4-3 were those where the observer reported interference,
usually from fugitive emissions from other sources.  Those data points
where the observer noted equipment failure or  lapses in good work practice
are included  in the data base because these lapses and failures have been
observed in the normal operation of the best controlled batteries.  Descrip-
tive information on each battery and a complete data listing are given in
Appendix C with observer comments on individual measurements.
     A statistical analysis of the data revealed the following:
         There  is no significant difference in the performance of single-
         and double-main batteries,
         There  is a significant difference in  the performance of short and
         tall batteries,
         The data are not  normally distributed, and
         A  logarithmic (log) transformation yields normally distributed
         data.
                                 4-26

-------
The data are not normally distributed primarily because of occasional  high
readings of seconds of visible emissions.   A logarithmic transformation of
In (X + 1), where X = seconds per charge,  was used to allow the application
of normal statistics.  The results of the statistical analysis are summar-
ized in Table 4-4.   Log averages were calculated with the equation:
                                             v
                              Log average = e -1
where:
              Y = In (X1 + 1) + In (X2 + 1) +.  .  .  In (Xp + 1)
                                       _

              X = seconds of visible emissions, and
              n = number of observations.
     A 95-percent confidence level was calculated for each battery,  based
on the log average of 10 charges in an inspection.   The 95-percent confi-
dence level means that the battery should be able to attain that level of
control 95 percent of the time.   If an emission limit is based on a given
battery's performance at the 95-percent confidence level, then there is a
5 percent probability of citing another battery in violation when it is
controlling emissions as effectively as the given battery.
     A significant difference was found in the performance and variance
components for short and tall batteries.  Possible explanations for this
difference are discussed later in terms of physical characteristics of the
two types of batteries.  The variance used to estimate the confidence level
for short batteries (less than 5 meters in height) was a pooled variance
from all of the short batteries; similarly, a pooled variance for tall
batteries (5 meters or greater in height) was calculated from the tall
battery data.  The variance components included the variance between charges
within days, observers within charges, and days.   The major variance compo-
nent was between charges, and the smallest variance component (less than 10
percent) was the variance between observers within charges.
     The best performance in terms of log average were U.S. Steel's Batter-
ies 16 and 17 at Clairton, but these batteries were using run-of-the-mine
coal which is not routinely used by the industry.   The best performance
(95-percent log average) for short batteries representing retrofitted,

                                  4-27

-------
             TABLE 4-4.   STATISTICAL  RESULTS  FROM  CHARGING  DATA
Company
U.S. Steel, Clairton3
U.S. Steel, Fairfield
CF&I, Pueblo3
U.S. Steel, Clairton3
Shenango, Neville Island
J&L, Pittsburgh
U.S. Steel, Clairton3
Shenango, Neville Island
U.S. Steel, Fairfield
U.S. Steel, Clairton3


Lone Star Steel
Arithmetic Log 95-percent level
average average (log average,
Battery (s/charge) (s/charge) s/charge)
16
17
9
B
1
22
3
P2
P4
10
19
4
6
3
7
21
9
cb
Bethlehem Steel, Burns Harbor 2
U.S. Steel, Clairton3

U.S. Steel, Gary
National Steel, Weirton
U.S. Steel, Clairton3
These batteries used the
2
8
20
lb
lb
11
optimization
1.0
1.2
5.4
5.4
8.6
6.6
6.1
7.3
6.2
8.8
6.8
6.6
10.0
7.8
7.6
13.1
8.4
7.4
10.8
9.0
8.9
12.0
13.0
11.9
11.6
of stage
0.5
0.8
4.0
4.5
4.8
5.1
5.1
5.6
5.6
6.0
6.1
6.2
6.8
6.9
7.3
7.3
7.5
6.6
6.9
7.7
8.1
8.1
8.4
9.0
10.1
charging described
1.2
1.7
6.7
7.4
7.9
8.3
8.4
9.1
9.1
9.8
9.9
10.0
10.9
11.2
11.6
11.7
12.0
12.0
12.5
12.4
13.0
13.0
15.0
16.1
16.1
in Sec-
 tion 4.1.1.3.
3Tall (5-  or 6-meter)  batteries.
                                  4-28

-------
rebuilt, single-main, and double-main batteries is a log average of 8
s/charge for 10 charges.  These representative batteries are Battery 9
at Fairfield, Battery B at Pueblo, and Battery 1 at Clairton.   The
representative tall batteries in the data base are Battery C at Lone
Star Steel with a log average of 12 s/charge, Battery 2 at Burns Harbor
with a log average of 12.5 s/charge, and Battery 1 at Gary with a log
average of 15 s/charge at the 95-percent confidence level.  The worst of
the batteries in the data base was Clairton1s Battery 11 which has a log
average of 16 s/charge at the 95-percent confidence level.  The statistical
analysis also showed that no more than one charge in ten exceeded 45
seconds at the 95-percent confidence level.
     The 95-percent level log average is difficult to compare with
existing regulations which are expressed in terms of arithmetic averages.
In all cases, the log average is smaller than or equal to the arithmetic
average.  A crude comparison can be made in Table 4-4 which shows that
the largest difference between log and arithmetic averages is about 5 to
6 s/charge for the data base.  A statistical analysis that compared log
averages to arithmetic averages from the charging data base indicated
that a log average of 16 s/charge was roughly equal to an arithmetic
average of 20 s/charge.
     Simulated data sets are given in Table 4-5 to demonstrate some of
the potential differences between log and arithmetic averages.  The data
base for charging shows that occasional  high readings are obtained from
a charging inspection because of equipment failure or operator error
that is a part of normal operation.   Examples 2, 3, and 4 show that one
or two high readings can increase the arithmetic average without signifi-
cantly increasing the log average.
     Battery age is one of the factors that may affect emission control.
Older batteries may present more difficult control problems because of
shifting of brick, alignment problems, and general wear on equipment.
The normal life of a battery is 25 to 30 years before a major rehabilitation,
although some batteries can be operated  longer than normal if they
undergo a partial  rebuild or minor repairs.  However, Table 4-4 indicates
that the performance of some older batteries can equal or exceed that of
some of the newer existing batteries.   For example, Batteries 1, 2, and
                                 4-29

-------
TABLE 4-5.  COMPARISON OF ARITHMETIC AVERAGE AND LOG AVERAGE











Arithmetic
average =
Log
average =

1
16
16
16
16
16
16
16
16
16
16
16.0
16.0
Example
2
6
16
16
16
16
16
16
16
16
40
17.4
16.0
(s/charge)
3
10
10
12
12
16
16
16
16
16
68
19.2
16.0

4
8
8
10
10
12
12
14
14
72
74
23.4
16.0
                         4-30

-------
3 at Clairton were built in 1955 and are scheduled to be rebuilt or
replaced in 1986.  Batteries 7, 8, 9, 10, and 11 are tentatively scheduled
to be abandoned in 1983 and replaced by new batteries.   Batteries 16 and
17 were built in 1951 and may be abandoned in 1982.  Batteries 3 and 4
at Shenango are older batteries which were operational  in 1950 and 1952
and have furnace walls that were rebuilt between 1974 and 1975.  In the
near future, Battery 3 may be rebuilt from the pad up.
     Representatives of the steel industry have commented on several
process variables that potentially affect the performance of stage
charging.  Two of these variables are coal moisture content and the
degree of pulverization.  (The potential effect of these variables on
uncontrolled emissions is discussed in Chapter 3.)  The exact magnitude
of the effects of these variables cannot be determined because they are
masked by the effects of many other factors which influence the performance
of stage charging.  However, available evidence indicates that the
effect of these variables may be small relative to other factors such as
equipment configuration, work practice, worker effort,  or steam aspiration.
     EPA attempted to delineate the possible effect of coal moisture
content and the degree of pulverization by measuring emissions at two
plants with similar control systems and procedures where coals with
different moisture content and fineness were used. Table 4-6 presents
these data.28  Bethlehem Steel's Rosedale plant actually had less emissions
than the Bethlehem Steel plant at Franklin, although the Rosedale plant
charged drier and slightly finer coal than the Franklin plant (5.5-percent
moisture compared to 7.8-percent moisture and 79.5 percent less than 0.3
cm (0.125 in) in size versus 76.8 percent).  This result is opposite to
the expected trend of increased emissions when drier and finer coal is
used.
     In addition, a distinct difference is evident in the degree of
control achieved on different batteries.  Although the data are not
conclusive,  they indicate that if moisture content or degree of pulveriza-
tion has an effect, it is overshadowed by other process or operating
variables which more significantly affect the performance of stage
charging.
                                 4-31

-------
  TABLE 4-6.   VISIBLE EMISSION DATA FOR THE BETHLEHEM PLANTS AT JOHNSTOWN328
Plant and date
Franklin
09/24/74

Rosedalec
09/25/74

Battery
number
17
18
15
16
Number
of charges
observed
3
33
16
15
Duration
Average
142
73
74
34
of visible emissions
(seconds)
Range
125 - 172
15 - 217
17 - 167
11 - 78
 The data at these plants were recorded by two to four observers.   Complete
 results are presented in Appendix C.

 Coal analysis:   7.8 percent moisture;  76.8 percent less than 0.3 cm
 (0.125 in) in diameter.

GCoal analysis:   5.5 percent moisture;  79.5 percent less than 0.3 cm
 (0.125 in) in diameter.
                                 4-32

-------
     Experience at the National Coal Board Avenue Plant in Sheffield,
England, supports the contention that neither the moisture content nor
the fineness of the coal has a significant effect on the performance of
stage charging.  Personnel at the plant stated that they have charged
coal with a moisture content as low as 1 percent and fineness as high as
80 percent less than 0.3 cm (0.125 in) without a noticeable difference
in emissions from stage charging.13
     EPA could not find any trend in the industry toward use of finer-
ground or lower-moisture coals.  Engineering firms that construct new
coke oven batteries indicated that their customers have not told them to
design for finer coals or lower moisture contents on future installations.
The data in Tables 4-2 and 4-3 represent a range of coal fines from 65
to 81 percent less than 0.3 cm (0.125 in).   The other extreme of pulveriza-
tion would be the charging of run-of-the-mine (nonpulverized) coal.
Results from U.S. Steel Batteries 16 and 17 at Clairton show that nonpul-
verized coal has no detectable adverse effect on charging emissions.
These batteries provided the lowest readings of visible emissions from
charging.
     The use of tall ovens may cause more charging emissions.  Most new
batteries are constructed with 6-meter ovens with a volume of approximately
40 m3 (1,400 ft3), or about twice the volume of 3- to 4-meter ovens.
Because of this larger coal capacity, about twice the volume of gas will
be displaced from the oven during charging, and a larger volume of
gasification products will be released because more coal contacts more
hot refractory.  In addition, tall ovens have a higher thermal head,
i.e., the rising gases in the oven achieve a higher velocity because of
a greater stack effect.
     Tall ovens also require a longer charge time which increases the
potential for emissions.   The smaller ovens have three to four charging
ports which are 0.35 to 0.41 m (14 to 16 in) in diameter; 6-meter ovens
have charging ports that are generally 0.46 m (18 in) in diameter.   Al-
though twice the volume of coal is handled in the 6-meter ovens, the
cross sectional area for coal flow is only about 25 percent greater
(comparing a 16-inch diameter to an 18-inch diameter for the same number
of ports per oven).   For example,  CF&I charges 4-meter ovens with a
                                 4-33

-------
mechanical feed in about 4 min; U.S.  Steel,  Fairfield Battery 6 (4-meter)
charges with gravity feed in about 5 min;  and U.S.  Steel,  Gary Battery 1
(6-meter) with mechanical feed requires about 6 min.   The  factors men-
tioned in the preceding paragraphs indicate that a double  collecting
main may be necessary for tall ovens to minimize the potential for
overloading the aspiration system.
     The data analysis and information in Table 4-4 indicate that several
short batteries controlled charging emissions better than  the two tall
batteries at Lone Star Steel and U.S.  Steel, Gary.   However, the perform-
ance of Battery 1 (6-meter) at U.S.  Steel, Gary and Battery C (5-meter)
at Lone Star Steel is better than the worst performance of the short
batteries with optimized stage charging at U.S. Steel, Clairton.
     Use of single or double collecting mains is another factor that may
affect emissions.  Double mains may provide more aspiration for removal
of the potential pollutants to the by-product recovery system.  The
batteries at J&L, Pittsburgh and U.S.  Steel, Fairfield use a single
collecting main with a jumper pipe connected to an adjacent oven.  The
data in Table 4-4 and the statistical  analysis indicate that the control
performance of single-main batteries is not significantly  different than
that of double-main batteries.
     Several other variations among batteries were examined and no
significant differences  in control performance were found.  For example,
the batteries listed in Table 4-4 represent a variety of coking coal
sources from Alabama, Colorado, Indiana, Pennsylvania, and Texas.
Batteries with three or  four charging holes, gravity or mechanical feed
for charging, manual or  automatic lid lifters, and operation during the
summer and winter are also included in the list.  These factors may have
an unquantified effect on emission control.   However, the  data  indicate
that any adverse effects from these variations are apparently controlled
by implementing the recommended equipment changes, operating procedures,
and optimization.
     All of the previous discussion in this section is based on the
observation and measurement of visible emissions.  This limitation was
necessary because of the absence of data on mass emissions from stage
charging for varying seconds of visible emissions.   Some explanation of
                                 4-34

-------
the relation between mass emissions and the duration of visible emissions
is necessary.  During observation of stage charging, EPA observers
noticed that on good charges (a small duration of visible emissions),
the emissions were generally small wisps or puffs which drifted out from
around the drop sleeves on the larry car.  (Figure 4-9 shows a typical
puff.)  For charges where the duration was higher, the character of the
emissions changed to clouds of smoke which escaped to the atmosphere
with higher velocities.  Generally, the length of the duration and the
number of larger clouds increased proportionately.  When the length of
the duration increased, fewer wisps were seen.  The mass in 40 seconds
of wisps is probably about twice that of 20 seconds of wisps; therefore,
40 seconds of mixed wisps and clouds must be more than twice the mass of
20 seconds of only wisps.
4.1.2  Sequential Charging
     4.1.2.1  Description.  Stage and sequential charging are often
confused.  The procedure called stage charging in the United States is
called sequential charging in England where the procedure was first de-
veloped.  The following definitions were the basis for this document,
and they correspond to the most common usage in the United States.
          Stage charging—a procedure where a maximum of two hoppers are
          discharging at the same time (as described in Section 4.1.1).
          Sequential charging—a procedure where the first hoppers are
          still discharging when subsequent hoppers begin dumping.  (This
          faster charging procedure uses an automatic sequence to start the
          discharge of coal from the hoppers.)
     In both procedures, the oven is under aspiration.   All the considera-
tions presented for stage charging in the previous section apply to
sequential charging.  Although the difference between the two procedures
seems quite small, it may be very significant in effective control of
emissions.
     The sequential charging procedure is designed to shorten charging
time and to provide a level coal  charge, thereby reducing the opportunity
for blocking the free space at the top of the oven.   On an oven with
four charging holes, the two outside hoppers usually are discharged
simultaneously, the third hopper  begins discharging 10 to 20 s after the
two outside hoppers, and the last hopper discharges 10 to 20 s later.
                                 4-3b

-------
Figure 4-9.  A "puff" of emissions during stage charging.
                      4-36

-------
     The American Iron and Steel Institute (AISI) and EPA jointly developed
and demonstrated one automated sequential charging system, which is
commonly referred to as the AISI/EPA larry car.  The three primary
components of the system are:  (1) automatic sequencing of operations,
(2) an air-conditioned cab (to protect the larry car operator from
emissions), and (3) a single-spot coke-charging machine.  A schematic
illustration of the system is shown in Figure 4-10.29  This system was
developed for a battery of ovens which have only one main (and no jumper)
and three charging holes.  The flow of gases from the oven during charging
is illustrated in Figure 4-11.
     Initially, the AISI/EPA larry car was designed for fully automatic
operation.  However, many operations were restored to manual operation
because of problems with reliability and to permit more flexibility in
the charging sequence.  Features of the larry car include a remote-controlled
standpipe damper and steam valves and mechanized lid lifters.   The
following description shows the sequence of operation.   (The letters in
parentheses refer to Figure 4-10.)
     a.  The larry car hoppers are each filled with a specified volume
of coal.
     b.  The car is moved to the oven to be charged (X) and positioned
using a visual spotting target.   The larryman uses a pushbutton to
operate the gooseneck cleaner (E).
     c.  At about the same time, the pusher-machine operator removes the
pusherside door of the oven to be pushed (Y) and then positions the
pusher for single-spot pushing and leveling.
     d.  The leveler door is opened on the oven (X), and the pusher
operator advances the leveler bar (B) until the smoke-seal (J) is in
position against the door opening.  He then notifies the larryman over
the voice communication system that charging can begin.
     e.  The "Steam-On" pushbutton is actuated to aspirate the oven (X);
it turns on aspirating steam, opens the damper, and closes the standpipe
cap.
     f.  The drop sleeves (H) are lowered over the open charging holes
with individual pushbuttons.
     g.  The butterfly valve (K) on each hopper is opened with individual
pushbuttons.

                                  4-37

-------
                Figure 4-10. AISI/EPA larry car coke oven charging system.
                                                              29
B.  Leveler bar
C.  UHF alignment device
D.  Ascension-pipe damper actuator
E.  Gooseneck cleaner
F.  Ascension-pipe actuator
G.  Lid lifters
H.  Feed hopper drop sleeves
I.  Leveler-door operator
J.  Leveler-door smoke  seal
K.  Butterfly valves
L.  Controlled environment  cab
M.  Coal hopper
K.  Oven to be charged
Y.  Oven to be pushed
Z.  Oven to be dampered off
                                       4-38

-------
                                                      ASCENSION

                                                      MPE 6 ELBOW
:J
•;j     i
L—A	
              r •'         ' ^\.     s.;?-
   Figure 4-11.  Flow of gases in the AISI/EPA sequential charging system.
                                                                       29
                                    4-39

-------
     h.  The operator manually stops the coal flow from Hoppers 1 and 2
when their indicator lights show that 75 percent of their coal  has been
charged.  Hopper 3 is permitted to empty completely.
     i.  The larryman notifies the pusher operator to start leveling
when the first "75-percent coal level" light turns on, provided that
indicators show that coal flow has started from all hoppers.   If coal
flow has not started from a hopper, he initiates leveling after the 75
percent coal charge level is indicated on all three hoppers.
     j.  As soon as the pusherman notifies the larryman that leveling
has started, the butterfly valves on Hoppers 1 and 2 are operated to
complete the coal charge.  As the hoppers empty, bottom-level detectors
automatically close the respective butterfly valves.
     k.  The larryman then notifies the pusherman that all three hoppers
are empty and the pusherman makes the final leveler passes before retract-
ing the bar and closing the leveler door.  The pusherman tells the
larryman when the leveler door is closed so that sequential relidding
(1-2-3) of charging holes can be completed.
     1.  The pusher operator retracts the leveler arm while the aspirating
steam  is still on.  He then moves the pusher car to the next oven to be
pushed.  The aspirating steam remains on the oven just charged to protect
the larry car which extends over it when spotted for the operation
below.
     m.  The larry car operator then moves to the next oven to be charged,
closes  the damper in the main, and removes the lids of the oven previously
pushed  to decarbonize the charging holes.
     n.  The car then returns to the coal bin, and the lidman turns the
aspirating steam off of the oven (X) which was just charged.
     One significant aspect of the AISI/EPA larry car is the amount of
automation used.  Automation should reduce the chance of operator error
and give more consistent operation from charge to charge.  This error
reduction should result in better emission control, which is heavily
dependent on adherence to a specified procedure for both sequential and
stage  charging.  As yet, total automation has not proved to be reliable.
     4.1.2.2  Performance of Sequential Charging.  Mass emissions have
been measured at one installation which uses sequential charging30 and
                                  4-40

-------
the AISI/EPA Tarry car described in Section 4.1.2.  The average result
of the tests is 120 grams of participate per charge31 (0.26 pound per
charge or 0.017 pound per ton of coal charged for a 15-ton charge of
coal).  About half of the emissions were benzene soluble organics.  To
make these measurements, "emission guides" were fabricated to fit around
the drop sleeves and channel the emissions to a duct of defined cross
section.  Samples were collected with EPA Method 5 sampling equipment at
a single point in each duct.  Because of the adverse sampling condi-
tions, the results of these tests are expected to be a rough estimate of
the level of system performance.  The duration of visible emissions was
also measured.   The average duration for the longest period of observation
(236 charges over 2 months) was 174 seconds.32
     However, these results do not represent the true potential of
sequential charging because they were obtained on a battery with a
single main and no jumper pipe.  After a jumper pipe was installed,
measurements showed that the average duration of emissions for 15 charges
was only 22 seconds.33 These data dramatically show the necessity of
two-way drafting of the oven to achieve a high level of performance.
Use of the Still-ERIN discharge hoppers, the effects of coal moisture
content and degree of pulverization, and use of tall ovens (discussed in
Secton 4.1.1.4) are important to sequential charging.
     One additional factor may be significant in consistently achieving
good charges with the sequential technique.  A disruption of coal flow
from one hopper will change the pattern of coal piles in the oven, thus
increasing the potential for blockage in the open space at the top of
the oven.  If the flow of coal from one hopper stops, the coal from the
remaining hoppers will fill part of the oven volume intended for the
coal that stopped flowing.   When the flow of this coal resumes, it will
cause blockage by overfilling the oven below the charging hole it enters.
When a portion of the oven is blocked from the aspiration system, emissions
will increase.   This situation does not occur with stage charging because
a problem with coal flow in a hopper can be corrected before the remaining
hoppers are discharged.
     Coal flow problems are more likely to occur when gravity feed is
used,  and mechanical feeders should at least partially overcome these
                                  4-41

-------
problems.   Typical problems with coal  flow are bridging of the coal  at
the outlet of a gravity feed hopper or, less frequently, failure of a
mechanical feeder.
4.1.3  Scrubber Systems Mounted on Larry Cars
     4.1.3.1  Description.   Another system used to control emissions
from charging coke ovens is a scrubber which captures emissions contained
by the sleeves or shrouds at the charging ports.   The whole system is
mounted on the larry car as shown on Figure 4-12.
     Operation of this system typically consists of the following steps.
     a.  The scrubber pumps and the draft fan are started.
     b.  The larry car moves into position over the uncovered charging
ports.
     c.  The shrouded drop sleeves are lowered over the charging ports.
(In one system the shrouds do not seat directly onto the charging ports
but are supported by projections on the base of each shroud.   This
method permits intrusion of combustion air that oxidizes emissions from
the coke oven.)
     d.  The discharge valves on the coal hoppers are opened in a predeter-
mined sequence which may vary from plant to plant.  As pressure increases
within the oven, the gases and particulates are forced out, but they are
contained by the draft on the shrouds and directed to the scrubber.
Some of the hydrocarbons are burned with the intrusion air at the base
of the shrouds to prevent explosions in the ducts or scrubber.  A detailed
drawing of a shrouded drop sleeve is provided in Figure 4-13.34
     e.  When a hopper is empty, the discharge gate is closed to prevent
the escape of emissions through the hopper.  A small amount of coal can
be retained in the hopper to provide a more effective seal.
     f.  As each hopper empties, the shrouds and drop sleeves are raised
and the charging port lids are replaced by magnetic lid-handling equipment.
     When the larry car returns to the coal storage bunkers to receive
another load of coal, it may also discharge the dirty scrubber water and
refill the scrubber water storage tanks.  The frequency of water replacement
is determined by the capacity of the storage tanks on the larry car.
     The number and arrangement of the scrubbers mounted on the larry
cars may vary.  Early attempts to place scrubber systems on larry cars
                                  4-42

-------
                                                     Exhaust
Figure 4-12. Representative scrubber system on a larry car.
                          4-43

-------
                   Coal
 Combustion ——
   Gases
Igniter
        Coke.Oven
            Gases
T]C-.Air  Ports
    Gas Port
  Coke Oven
    Gases
 Figure 4-13.  Details of a shrouded drop sleeve.34
                 4-44

-------
were not satisfactory because under-powered scrubbers could not effectively
handle the volumes of gas and particulates or ineffective igniters did
not provide consistent ignition of the gases.  Newer systems are designed
to use high-energy venturi scrubbers and improved igniters.  In some
cases, electronic igniters (large spark plugs about 25 cm in length)
have been used effectively.  Most of these newer systems are operating
in Japan and Europe.
     Some significant disadvantages of early larry cars with integral
scrubbing systems were reported in a Battelle study:35
          "They are relatively costly [and] complex ...",
          Their operating history shows "progressive severe maintenance
          problems and increasing breakdown rates,"
          "Intensive ignition ...  [of the gases from the coke oven] ... is
          essential," and
          They "create a substantial problem in the disposal of polluted
          exhaust waters."
     Further reduction of emissions is achieved by systems that discharge
the gases from the scrubbers on the larry car to a second scrubber
located at a fixed site.36  These systems use a quick-connect fitting to
a manifold which extends the length of the battery.
     4.1.3.2.  Performance of Scrubber Systems Mounted on Larry Cars.
Systems that use a scrubber mounted on a larry car have two emission
sources:  the area around the drop sleeves where poor capture will allow
emissions to escape and the effluent from the scrubber.
     Although data are not available to show the extent of either the
mass or visible emissions that escape at the drop sleeves, the capture
can be judged relative to other systems.   Observations of 12 installations
in five countries indicate that "on no occasion did any observer see any
scrubber car perform even one completely smokeless charge."37  Emissions
during leveling were the most noticeable.   Emissions from the drop
sleeves probably will  be greater than those from stage charging.
     When this system is used on an oven with low aspiration, emissions
escape when the drop sleeves are raised to allow replacement of the
charging port lids.   Because there are larger openings to the atmosphere
below the drop sleeves through which emissions are evacuated, the system's
                                  4-45

-------
effectiveness in capturing emissions will  be drastically reduced.
     Though no data on carcinogenic emissions are available,  particulate
emissions in the scrubber effluent have been measured at two  plants in
Germany and are presented in Table 4-7.13   The Ruhrkold Zollverein plant
in Essen, Germany, uses a low-energy Krupp scrubber with rotating screen
discs in the fan housing, followed by a cyclonic mist eliminator.   The
type of igniter used is not known.   Measurements were made on two similar
cars at this plant.  The Osterfield plant  in Essen, Germany,  uses five
low-energy centrifugal dust collectors (one for each charging hole)
followed by mist eliminators.   Gas burners located before the scrubbers
are used to ignite the gases.
     Both of these scrubbers have a low-energy input.  Newer  installations
have used venturi scrubbers with a higher  energy input.  Although no data
on Venturis were gathered for this report, it is generally accepted that
lower emissions can be attained with the venturi scrubbers.   This low
emission level was noted during the survey of the 12 installations previously
mentioned.37
     An important aspect of scrubber performance in collecting charging
emissions is the scrubber's overwhelming dependence on the ignition and com-
bustion of gases before they enter the scrubber.  Although combustion of the
gases is needed to reduce the risk of an explosion in the ducts that lead to
the scrubber, it also has a profound effect on the efficiency of the scrubber.
Along with the gases, the fine fume and smoke that are emitted during
charging are burned.  If they are not burned, these types of emissions will
pass through the scrubber (particularly the low-energy scrubbers) almost
completely intact.  The previously mentioned study of these systems concluded
that performance is "completely dependent upon the degree of ignition at-
tained and maintained."37
     In conclusion, three primary variables that affect the performance of
scrubber systems in controlling emissions  from charging are:
          The type of scrubber and energy input used,
          The consistency with which ignition of the gases is maintained,
          and
          The amount of suction used to capture emissions in the drop
          sleeves.
                                  4-46

-------
    TABLE 4-7.   PARTICIPATE EMISSIONS FROM SCRUBBERS MOUNTED ON LARRY CARS13
Scrubber performance
   Ruhrkold Zollverein plant
    Larry car       Larry car     Osterfeld
    number 1        number 4        plant
Emissions

  Grams (pounds)/charge

  Grams/megagram (pounds/
    ton) of coal charged

  Milligrams/standard cubic
    meter (grains/standard
    cubic foot)

Collection efficiency,
  percent

Gas flow, standard cubic
  meters/hour (standard
  cubic feet/minute)

Test method
  201 (0.44)        83 (0.18)      150 (0.33)

    7.7 (0.017)      3.2 (0.007)     3.7 (0.008)
  738 (0.32)
   95
585 (0.26)
 96
167 (0.07)
 97
7,200 (4,200)    3,700 (2,200)  19,400 (11,400)
   N/AC
   N/AC
 VDI 2302
  (8/74)
 The test method is not known but it is probably the same as that used for
 the Osterfeld Plant.
                                  4-47

-------
     One observation38 of the performance of a second scrubber in series
with a scrubber on the larry car indicated that the second scrubber did not
noticeably increase the efficiency of the system.
4.2  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 (two 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.  Various systems to control  leaks have been tested
and improved.
     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 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 leak is determined by the
size of the opening, the pressure drop between the oven and the atmosphere,
and 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 collect-
ing main or provide external pressure on the seal  through the use of steam
or inert gases.
     Oven door emissions also can be reduced by collection of the leaking
gases and particulates and subsequent removal of these pollutants from the
                                  4-48

-------
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.
     Where data are available, the performance of the control techniques
for doors is included with the descriptions.  Generally, this performance
is expressed in terms of visible emissions as percent leaking doors
(PLD). A discussion of the test procedure for measurement of PLD is
included in Appendix D.   The number of doors leaking is divided by the
total number of doors; the chuck door is not counted as a separate door
but is considered as part of the pusherside door.
4.2.1  Traditional 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.39  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.   One estimate of the
sealing time is 1 hr if the gaps are less than 0.02 cm (0.009 in).40
One-hour sealing times are considered satisfactory for meeting some of
the current (1979) State emission standards, because the average leak
rate would be only 5.5 PLD on one battery side with an 18-hour cycle
time.41
                                  4-49

-------
     Metal-to-metal seals are commonly used in the production of metallur-
gical coke.   The major types of industrial  seals used in the United States
are the Koppers and the Wilputte seals, which are named for their manu-
facturers.
     4.2.1.1  Koppers Door.   The Koppers door contains an S-shaped seal
which is flexible and presses against the jamb.   Springs reinforce and
assist the sealing ring to provide uniform contact between the seal edge
and jamb surface.  The sealing ring is bolted to the door frame, and the
spring boxes are an integral part of the door frame.  Door limit stops are
provided to prevent over-stressing the sealing ring.  A diagram of the
Koppers door is shown in Figure 4-14.42
     The sealing strips and springs are made of 304 stainless steel which
is annealed during initial use.  The yield strength is reduced in the
annealed state, and 304 stainless steel cannot be heat treated to increase
hardness and strength.  The coefficient of thermal expansion of 304 stain-
less steel is greater than that of the conventional door frames to which
the material is bolted.  At temperatures in excess of 500° C (932° F) the
creep rate of 304 stainless steel is high and deformation is permanent.
Eventually, these deformations will destroy the seal's ability to maintain
a firm contact with the jamb.43
     The many elements involved in maintaining the effectiveness of a
metal-to-metal seal for limiting door emissions have been compared to the
links in a chain, because any one missing element will cause the overall
system to fail.40  Metal-to-metal seal performance is influenced by the
seal adjustment, new door installation on the oven, and door handling,
which includes spotting, placement, latching, and removal.  Door and jamb
cleaning, door leakage troubleshooting and correction, latch lubrication,
and quality control of replacement parts are important aspects in maintain-
ing emission control.  Therefore, maintenance procedures are extremely
important in providing continuous good emission control with any metal-to-
metal seal.
     When a Koppers seal is new and installed on clean, straight jambs, it
may control emissions in the range of 0 to 10 PLD;43 44 however, this
performance may deteriorate after only 6 months.  Probably 40 to 50 percent
of the standard doors will eventually leak.40 45  This range represents the
performance in the absence of good operating and maintenance procedures.

                                  4-50

-------
                           o
                           o
                           •a
                           §

                           B
                               S
                              1

                               c
                              O)
4-51

-------
     When the standard seal is used with good operating and maintenance
practices, an average level of about 10 PLD may be obtained.   Table 4-8
shows data collected at Kaiser Steel Batteries F and G which use the stand-
ard Koppers door with maintenance procedures.  These data indicate a range
of 0 to 20.5 PLD.

              TABLE 4-8.   STANDARD KOPPERS DOOR AT KAISER STEEL
Date
5/79

9/79

Battery
F
G
F
G
Average PLD
4.2
3.8
11.0
9.2
Range
0 -
0 -
2.3 -
2.4 -
of PLD
7.8
7.8
16.6
20.5
Reference
46
46
47
47
     4.2.1.2  Wilputte Door.  The Wilputte coke oven door is the other
major emission control technology used in the United States.  Wilputte
batteries are equipped with self-sealing doors with a fixed edge seal
design (Figure 4-15).  The seal that contacts the jamb is a knife edge
which can be adjusted by screws.  The knife edge is typically a 316 stain-
less steel strip with round edges, and it fits into the slot of the knife
edge holder.  The edges are installed in sections.   The knife edge holder
is welded to the door diaphragm or sealing plate, which is 0.6 cm (0.25 in)
thick.  The welds are continuous on the inside, and tack welds are used on
the outside.  Oval point knife edge adjusting screws for the sides of the
door are supported by U-shaped clips welded onto the door frame.  These
adjusting screws in the top and bottom sectors of the doors are run through
a square nut welded to the underside of a flange that is integral with the
top and bottom door frame castings.  Each adjusting screw has a lock nut.
Latch springs develop the force to hold the oven door against the jamb.
The typical force of the knife edge against the jamb is 150 N/cm (80 Ib/in).
     A recent program of maintenance and repair of Wilputte doors at one
plant has reduced emissions from 45 PLD to an average of approximately 7
PLD.45  The elements of the emissions control program included replacement
of damaged components, adjustment of oven doors, cleaning, inspection, and
maintenance of oven doors, adequate spare doors, good operational practices,
                                  4-52

-------
  Ul
  _l<
%CDO   a,


|S*   1
m<*.°   K
iiS   o

S££   §
                                                       o
                                                       o
                                                            o
                                                       IT)   a-

                                                       5   -


                                                           O5
                                                           Q.
                        4-53

-------
and monitoring.   Also, the door design was improved by adding extra adjust-
ing screws.   Without the maintenance program,  the door's performance is
similar to that of the Koppers door.   An EPA inspection of Battery P3N at
J&L, Pittsburgh showed a range of 21 to 38 PLD and an average of 27 PLD
from nine observations.22
     4.2.1.3  Wolff Self-Sealing Door.   The Wolff cast iron,  self-sealing
door contains a simple strip which is forced flush against a  jamb with a
cammed lock plate (Figure 4-16).   The Wolff door was introduced in 1950  and
is predominantly used in Europe.   More than 85,000 Wolff doors and jambs are
in operation in the world.  Initially,  the door contained screw-type latches;
however, later modifications included a pneumatically operated steel bellows
for spring-type latches.  A later modification of this door includes a
gravity latch whereby the weight of the door is transferred to the sealing
edge and becomes the sealing force, replacing the screw and the pneumatic
bellows as the latching mechanism.42
     Sealing of the Wolff door is achieved by a steel straight-edge held in
position against the jamb by a cam.  The sealing edge is adjusted by hammer-
ing lightly on its back edge to contour it to the oven jamb.   The sealing
strip cuts through the tar deposit on the jamb to form a seal.
     The door is equipped with mushroom cap guides on both sides for ease
in replacing the door on the jamb;42 this placement reduces the damage to
the sealing edge.  The sealing edges are also equipped with backup bolts to
further force the sealing edge to conform to the jamb over the short span
of the oven width at the top and bottom of the door.
     The Wolff seal is a simple, sturdy design that is easy to clean.48
Advancement of the sealing strip compensates for thermal deflection of the
door.   Cammed lock plates keep the sealing strip in position  and secure  it
against displacement.   The sealing strip is pressed against the oven jamb
with a latching force of 100 N/cm.
     4.2.1.4  Hand-Luting.  Luting, which is one of the oldest coke oven
control technologies, is used commercially on many foundry batteries and on
a few metallurgical coking 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 end of the oven.
The luting is a mixture of clay, coke breeze, and water, which dries and
                                  4-54

-------
                Door  Body
                Curved  Lock  Plate
                   Wolff Sealing  Strip
Figure 4-16. Wolff self-sealing door design.48
                   4-55

-------
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 can be recycled.
     Luting is particularly interesting because it represents an emission
control technology that has the potential for eliminating almost all door
leaks.  However,  the lack of fully developed luting formulations and the
absence of proven luted door technology for high production rate metallur-
gical 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 equip-
ment, and solutions to material handling problems.
     Emission data for batteries with hand-luted doors show a wide range of
emission control.  Three plants averaged 2, 5, and 13 PLD with ranges of 0
to 10, 0 to 17, and 5 to 19 PLD, respectively.49 50  A conscientious program
of reluting doors as necessary would be required to maintain this low level.
4.2.2  Modern Metal-to-Metal Seals
     Currently available major emission control techniques are based on
metal-to-metal seals.  Self-sealing doors with metal-to-metal seals in-
variably 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 Battelle investigation39 of the self-sealing
mechanism found that even small gaps were not sealed during the first 30
minutes after the coal was charged.  Most of the early flow from the coke
ovens was steam-containing volatile oils and tar.  The condensation of tars
sealed the small  openings only after the steam content was reduced.

                                  4-56

-------
     A few coke oven doors have been observed to be completely free of
visible emissions during the entire process.39  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.43
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.39  Most
of the components of the oven's end-door assembly are tightly constrained;
consequently, when the assembly is heated, stresses result because gross
distortions are prevented.  Thermal cycling under these constrained condi-
tions 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.51
     One of the major causes of poor sealing is the accumulation of irrever-
sible deformations of the jamb.51  The constrained jamb is heated by conduc-
tion through the bricks between the jamb and the end flue.52  The permanent
deformations that result appear both in the plane of the jamb (hourglassing)
and normal to the plane (bowing).   An investigation of jamb distortion
found significant warping on all jambs measured, and warping probably has
occurred in various degrees on most of the 25,000 or more cast jambs in
service.39  One technique to minimize jamb deformations is to allow the
jamb to bow inward to relieve the stress and then to bolt the jamb to the
backplate-buckstay arrangement.53  When this technique was used, the jambs
were straight, but fires subsequently warped the jambs.
     Stress relaxation of the metal seals also causes gaps between the seal
and the jamb.   These gaps form when the seal permanently distorts to relieve
stress forces or when the sealing edge deflects backward because of thermal
buckling (usually above the upper latch).  The flatness of the sealing edge
also changes because of stress relaxation and distortion.53  In this work,
some periodic up and down effect was noted with 0.094 cm (0.037 in) and
0.064 cm (0.025 in) relative gaps.   The problem of maintaining a uniform
seal  contact is compounded by changes in the contact pressure on the seal.
                                  4-57

-------
Uneven stress distributions may be caused by the forces from adjusting
screws and spring plunger assemblies or from a variation in the forces of
the adjusting and latching springs as the temperature distribution changes.
These factors tend to increase the gap size which increases both the time
required to seal the gap and the amount of gas escaping through the gap.
Although flexible seals are more capable of conforming to irregular jamb
contours than fixed seals, both types are subject to distortion.  Door
leaks that are caused by this seal distortion may be minimized by repairing
or replacing the seal when the gap size exceeds a specified limit.  Some
plant operators say the gap between the seal and jamb must be less than
0.02 cm,40 while others require a gap size that is less than 0.005 cm41 or
0.008 cm.51
     Although the distortion and damage of the jamb and sealing strip is of
primary importance, the structural behavior and interaction of all elements
involved in sealing accounts for the sealing performance of the system as a
whole.51  When carbon and tar build up on the sealing surfaces, plastic
deformation of the components will not provide an acceptable seal.  Hardened
deposits of tar can change the sealing surface so that the gaps depend on
the exact placement of the door.  If the jamb is not uniform, different
placements can impose different stress patterns on the door seals.  To
avoid irreversible distortion, the seal must not be stressed beyond the
elastic limit.  Once contact is made at the latch levels of the door, it is
useless to increase the latch force, and the adjustable feature of the door
seal must be used.
     The possibility of permanent distortion increases with the rigidity of
the end-closure components.  Any reduction  in rigidity would promote posi-
tive results.  Any decrease of temperatures, thermal gradients, and material
buildup behind the jamb will result in improved dimensional stability of
the end-closure system.51
     4.2.2.1  Modified Koppers Door.  The major innovations used to upgrade
the standard Koppers door technology are a  stronger, more temperature-
resistant seal, a more temperature-resistant plunger spring, and stop guide
blocks.40 44  This technology can be applied to Wilputte doors as well as
to Koppers doors.
                                  4-58

-------
     U.S. Steel has implemented the following modifications on their Koppers
doors.  The old seals were replaced with new seals of a special alloy
(NiCuTi) which is more resistant to permanent distortion at elevated temper-
atures.  The stop blocks prevent distortion of the seal beyond the elastic
limit.
     Springs in the Koppers doors were replaced with more temperature-
resistant springs.  This new material is less subject to stress relaxation
and greatly extends the useful lifetime of the sealing assembly.  U.S.
Steel added a corner spring assembly which reduced leaks from the square
corners.
     Four stops were added to each door.  The stop guides are located
behind the knife edge to prevent excessive stress on the sealing edge.  The
guides also assist in the positioning of the door.  Operator training in
combination with the use of the stops is important.   Plug spotting is used
to center the door; then the guides are used.  The latch legs nick the
sealing edge if the door is improperly positioned.  Nicks can now be easily
repaired, because the stops prevent excessive damage.  Different position-
ing of the stop blocks is required for different types of door machines.
Some door machines arc the door into place while other machines push the
door straight in.  Three out of four stops must be flush with each jamb or
the door is not considered to be properly in place.
     Old cast-iron jambs were replaced with new nodular cast-iron jambs.  A
special bolt added to the buckstays permits adjustment to a more favorable
jamb bow.  Latches were equipped with low friction bearings so that repro-
ducible pressures can be applied to the seals.
     Although regular Koppers doors can seal initially, U.S.  Steel believes
this ability is lost over time.   U.S. Steel's modifications of the door
seals produce a more uniform sealing, and the seal has a longer service
life expectancy.
     Specifications have been set for the manufacture of the door parts and
these have improved the part quality and emission control.   Design speci-
fications, assembly guidelines,  installation and adjustment procedures,
maintenance procedures, and troubleshooting techniques are  significant
parts of the overall  door-sealing system.   Door leaks are monitored over
                                 4-59

-------
several cycles, because small  carbon fragments can interfere with sealing.
Maintenance on improperly sealing doors consists of a sequence of three
procedures.  After proper identification procedures,  a problem door is
systematically examined, adjusted, and repaired as needed.   If the door
does not properly seal after several cycles,  changing the adjustment usually
increases the problem.  The identification of problem doors (repeaters) is
an important aspect of the overall technology, and doors are replaced when
they leak repeatedly.
     Repair of the door involves replacing the plug,  testing the springs,
repairing nicks on the seal, or replacing the seal.  An important aspect of
the maintenance procedure is testing the door sealing assembly and comparing
the measurements with specifications that have been developed for door
sealing.  The door leak problem has been reduced by adherence to these
specifications.  Maintenance especially emphasizes seals, nicks, and align-
ment.   If the seal is out of specification after repair, it is replaced.
     Motivation of the oven workers is essential to a successful emissions
control program.  Achieving this motivation involves a substantial amount
of management skills.
     The previously mentioned oven door technology has been in use since
January 1978.  As currently practiced, this technology is not expected to
reduce  the emissions to significantly less than 5 PLD.  This apparent
limitation is consistent with the concept of self-sealing by tar generation
during  the initial stages of the coking cycle.  (Doors seal after 5 percent
of the  coking cycle.)  Data from an EPA inspection in August 1979 (provided
in Table 4-9) show a range of 1 to 5 PLD for six of the Clairton batteries
with modified doors.20
     CF&I uses similar technology on the Koppers doors at their Pueblo,
Colorado, coke batteries.19  All the doors were equipped with new NiCuTi
seals which can be easily repaired or welded.  The oven door springs were
changed from chrome vanadium steel 6150 to chrome silicon steel SAE 9254.
This change improved static loading properties and increased creep resistance.
Rebuilding techniques for doors were refined to include "blueprinting"--
repairing the seal components to design specifications.  Operation and work
practice techniques were improved to provide more thorough inspection,
                                 4-60

-------
 TABLE 4-9.  PERCENT LEAKING DOORS FOR BATTERIES AT U.S. STEEL, CLAIRTON20
Battery
1
2
19
20
21
22
Observations
7
7
5
9
5
6
Percent leaking doors
Average
0.9
2.3
4.0
4.7
3.2
3.4
Range
0 -
0 -
3.5 -
2.3 -
1.8 -
2.3 -
2.3
5.5
4.8
8.7
5.9
5.4
 Collected over a period of 4 to 7 days in August 1979.

cleaning, placement, adjustment, and repair of doors.  This program at CF&I
has reduced door leaks from the 15 to 20 PLD level in 1976 to an average of
about 5 PLD in 1978 and 1979.19
     4.2.2.2  Modified Wilputte Seals.54 55  Jones and Laugh!in Steel (J&L)
believes that accurate door seals provide the most effective control of
emissions.  At their Pittsburgh Works, Wilputte doors have been converted
to a Koppers type flexible seal (Figure 4-17).   The two major areas of seal
modification are the materials of construction of the sealing edge and the
use of temperature-resistant materials in the springs.
     Conventional sealing edges are distorted during high temperature use;
however, the new NiCuTi seals are temperature resistant and can be removed
while the plug is on the door (not possible with the old design).  Guiding
the door to replace the seal in a reproducible fashion is also required for
consistent sealing.   Guides were added to position the door top-to-bottom
and side-to-side for repeated seal positioning.   The springs were changed
to a special design for high-temperature service and location of the springs
was carefully chosen to improve sealing.
     Proper maintenance practices are essential  for this control technology,
and operator attention is a critical factor.   Under J&L's maintenance
practice,  the jamb deflection is measured and jambs are replaced as neces-
sary to meet the existing standards.  The limits of bowing and hourglassing
are 1.9 cm (0.75 in) and 1.3 cm (0.5 in),  respectively.
                                 4-61

-------
SPRING LOADED
PLUNGER ASSEMBLY
            BIG HOLE PLUG
            K\NX\\\'V\ \ \ \ \\
               Figure 4-17. Cross-section of modified Wilputte door.54
                              4-62

-------
     Maintenance is more stringent for rehabilitated seals.   According to
leakage history, seals have been changed from one to twelve times a year.
To change the seals, the leaking door is taken out of service, sandblasted,
and tested for warping.  The seals are replaced and the springs checked for
compression.  The rehabilitated door is then placed on the oven and adjusted.
The clearance is specified as 0.008 cm (0.003 in) around the door.   The
performance of the door is checked over several cycles.  Adjustments are
not performed on doors in service because extensive efforts to adjust
leaking doors have been unsuccessful.  A door must be removed from service
when it will not seal.
     The conversion from a standard Wilputte seal to a flexible seal is
expensive.  J&L's estimate of labor and materials required is $8,000 for a
pusherside door and $7,000 for a cokeside door.  Rebuilding the doors from
the conventional to the new design may require three times as long as re-
building with a conventional Wilputte design, but after implementing the
new design, the seal change is faster on the modified doors because the
plugs do not need to be removed.55
     Changing Wilputte doors to the modified design has reduced leaks from
between 40 and 60 percent to 10 percent.55  This rate is less than that of
rehabilitated Wilputte batteries which were only able to achieve a 25
percent leakage rate using conventional technology.  In August 1979, an EPA
inspection of J&L's P2 battery in Pittsburgh showed even better short-term
results.  A range of 2.5 to 3.4 PLD was observed with an average of 3.2
PLD.22
     4.2.2.3  Ikio Seals.56   The Ikio oven door is similar to a modifica-
tion of the 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 (Figure 4-18).  The sealing plate is 304
stainless steel, 1.5 cm (0.06 in) thick, and is positioned between the main
body and the brick side of the oven door.  The sealing strip is a flat
ductile steel bar, 8.0 mm (0.315 in) thick, welded to the sides of the
sealing plate.   Whereas the Wilputte door has adjustable screws which
maintain the relative position of the sealing edge, the Ikio door has
                                 4-63

-------
                             FLEXIBLE SEALING PLATE
                                       Detail of Su»p**«t«i Brick Holder.
Figure 4-18.  Ikiodoor.56
         4-64

-------
springs which can be moved 10 mm (0.4 in) forward and 15 mm (0.6 in) back-
ward and are positioned 300 mm (12 in) apart along the sides of the oven
door.  These springs provide uniform pressure for the sealing plate and
knife edge against the oven frame.  This unique construction is claimed to
seal in the gas completely.
     The spacing of 40 mm (1.6 in) between the main body and the brick side
of the oven door reduces heat transmission from the oven.  The reduction in
temperature differential minimizes the stress on the oven door and warping
is reduced.  Measurements obtained on a 6.8-meter (22-foot) oven door
indicate that the oven door plate operates at substantially lower tempera-
tures than conventional oven doors and that there is also a lower temperature
difference between the door plate and door frame.
     The reduction in strains of the oven door is claimed to reduce the
amount of travel required for the knife edge to conform to the jamb contour,
and  this reduces the adjustments necessary in the springs.  Constant atten-
tion is not required; therefore, labor costs are reduced.  The insulation
of the oven by the new door seal reduces the overall heat radiation from
the  ovens and provides improvement in working conditions and productivity,
according to the manufacturer's claims.56
     Conventional doors can be replaced with the Ikio oven door.  The manu-
facturer prefers to sell a new door and seal with a new jamb, which is
expected to provide better emission control.  The Ikio enclosures are re-
ported to be performing well at some plant operations.57 58
     An inspection was conducted by EPA to observe Ikio doors in operation
on the tall ovens of Battery 5 at National Steel's Zug Island coke plant.
The  6-meter doors were leaking at a rate of 38 to 55 PLD.47  Plant personnel
indicated that the seals were not adjusted on a regular basis.  Other plant
personnel felt that there was an unresolved design problem related to
domestic coal mixes.59  Based on this plant visit, it could not be concluded
that Ikio doors represented an exemplary control device.
     4.2.2.4  Battene Seal (EPA-AISI).  The Battelle Columbus Laboratories,
in association with AISI and the U.S.  EPA, has developed a retrofittable
seal.60 61  The material of construction is a heat-treated, high-nickel
              ®
alloy—Inconel  X-750, and the seal is a continuous spring which is rein-
forced if necessary by continuous springs (Figure 4-19).  The retrofitted

                                 4-65

-------
                                     jamb
Door Frame
                               Scale:
                                   L
                                            Inches
                Figure 4-19.  Battelle seal design.61
                             4-66

-------
seal and spacers will be installed to align the bow of the door with the
bow of the jamb.  The seal is retrofittable both to Wilputte and Koppers
enclosure systems.   This seal is less dependent on point loading than are
seals with individual spring units.  The Battelle concept 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.   This project has now entered the demonstra-
tion phase, and no conclusive results have been reported.
     4.2.2.5  Gas Seal, Nippon Steel Company.   The gas seal has been commer-
cially used for 2 to 3 years at Nagoya, Tobata, and Kamaishi Works of
Nippon Steel.  In the case of Kamaishi Works,  the seal was retrofitted to
existing batteries.  The 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.62  The self-
sealing doors used on the observed batteries at Kamaishi are Otto doors
with Wolff "knock-type" seals.   The Japanese claim that the details of the
seal design, gas type, and gas distribution system are business confidential;
therefore, this information is not presented in this report.  The gas used
is nontoxic, nonexplosive, noncombustible, and generally available to all
coke plants.
     In practice, the gas supply to each individual 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).62  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 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 the hand.
Reluting is practiced as necessary to prevent  obvious leaks.  Gas usage on
any particular door need be continued only for a portion of the coking
cycle.
                                 4-67

-------
     Performance observations during a 3-day period revealed that the gas
seal technology produced leakage rates in the range of 0 to 10 PLD with
collecting main pressures of about 6 mm of water.   These performance levels
are about equivalent to the best technology currently in use in the United
States but they are not clearly better than this technology.
     To retrofit this technology, doors and jambs must be in good condition.
The vendor claims that if several spare doors are available, there is no
need to curtail production during the retrofit period.  The maintenance
program was claimed to involve rigorous cleaning of the doors and jambs;
however, such cleaning was not observed during the plant visit.  The esti-
mated capital cost to install the technology is about $1,000 per door.62
     4.2.2.6  Prechamber Doors.  The Wolff steam prechamber door uses a
double seal with steam introduced between the two seals to control door
leaks (Figure 4-20).48  The inner seal (System 1) is accomplished by metal-
to-metal contact, and the outer pliable seal (System 2) is silicone rubber.
The metal seal is made of ordinary carbon steel with a tensile strength of
500 to 600 N/mm2 and a width of 8 mm for 6- and 7-meter doors and 6 mm for
4- and 5-meter doors.  Every fifth screw around the periphery of the door
provides a pressure adjustment for the inner seal.  The prechamber has a
volume of 830 liters on a 7.1-meter door.  Steam is injected into the
prechamber and is regulated by a 2-millimeter orifice.63  Emissions from
the oven are reduced because of the steam pressure; the outer seal prevents
the escape of any leaks through the inner seal.
     With respect to retrofit, original door frames can be reused, but the
other door components must be replaced.  Conventional door machines can
handle the door installation and removal.  One problem area is the 9- to
10-month life of the silicone rubber seal; a new material with an improved
life would make this technology more attractive.63
     Five doors at a German coke battery were fitted with the steam pre-
chamber.  These doors were in operation for 2 years with no problems other
than replacement of the soft seal.  Complete control (i.e., no leaks) was
reported for the five test doors.58  However, these doors have been removed
from service and replaced with conventional doors.  These doors are no
longer used because of the experience of soft seal replacement and the
absence of regulatory pressure to reduce emissions further.63
                                 4-68

-------
                                      Locking  Mechanism
Latch  Hook
                                      Prechamber
                                      Sealing System  2
Sealing  System  1
                      V	  1
Walt  Protection Plate
                                      Curved  Lock  Plate
                                      Prechamber  Hood
Buckstay
                                      Door  Plug
                  Figure 4-20. Wolff prechamber door.64
                             4-69

-------
     An advantage of this control  technology is the reduction in tar de-
posits because of the steam pressure.64  Moreover,  air is prevented from
entering the oven, and this eliminates soot formed  by combustion in the
oven.   A longer life is expected for the door seal  and the oven.
     One disadvantage is that space is needed within the buckstay for the
chamber structure.  Retrofitting existing ovens requires space which older
batteries may not provide.   One possible disadvantage is that steam can re-
act with carbon at elevated temperatures and produce extra coke oven gas at
the expense of coke yield.   The door cost is 40 to  50 percent higher than a
standard coke oven door.
     Another modification of the gas chamber is the metal-to-metal seal
which contains a gas chamber for coke oven gas at above atmospheric pressure.
The current development status is a patent which has been granted.65  One
of the claims of the patent is that double protection is provided against
leakage of gas to the atmosphere.   The purge gas system is claimed to
collect particulate matter from the chamber.
4.2.3  Other Control Techniques
     Techniques that are discussed in this section do not fall into the
previously discussed category of metal-to-metal seals.  Emphasis is given
here on work practice controls, fume collection, design modifications, and
promising technology that is in the patent or trial stage.
     4.2.3.1  Maintenance and Operating Procedures.  The avoidance of leak-
age greatly depends on good operating practices and maintenance.  One
factor in emissions control is provision of the best possible work environ-
ment for the operators.  In one reference, good maintenance and operating
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."66
     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.67
                                 4-70

-------
     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.
     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.43  With Wilputte
doors, it is impractical and undesirable to clean the sealing surfaces to
bare metal; a thin film of tar should remain after cleaning to aid in
sealing gaps between the knife edge and jamb.45
     The seal on metal-to-metal coke oven doors should be maintained so
that they meet a maximum gap specification.  Seal distortions should be
corrected by filing or the seal should be replaced to meet specifications.
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.
     Door-machine operation should emphasize smoothness and continuity of
movement to minimize stress on the sealing structure, and the door machine
should be carefully maintained.  Variables such as the force applied to the
latch and the lubrication pressure are important in obtaining uniform door
placement.  In the U.S. Steel paper by Stanley,41 an investigation of latch
forces yielded values that ranged between 30,000 and 170,000 N (6,800 to
39,000 Ib) per latch.  Screw torques also varied between 270 and 1,150
newton-meters (200 to 850 foot-pounds).  U.S. Steel found it necessary to
establish a practice for providing uniform latch forces, as follows:
          Standardize drive motors and torque converters,
          Pretest the torque motors in the shop to obtain the proper torque,
          and
          Set the amperage in the field at the desired level.
     4.2.3.2  Effects of Process Variables.  The 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
                                 4-71

-------
have been reported to cause severe damage to the oven wall  brickwork.61
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 compo-
nents and increase emissions.  The major drawback to lowering the coke
temperature 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.
     4.2.3.3  Soft Seals for Control of Chuck Door Emissions.  The perform-
ances of three major types of soft seals have been reported.  These three
types are graphite impregnated with carbon,54 asbestos impregnated with
graphite,54 and the Escolator chuck door gasket.  Generally the soft seals
do not provide long service life because oxidation or compliance loss
causes them to fail.  Any tar that builds up on the seals is subject to
hardening.  New asbestos-gasket material was reported to be in use for 3h
weeks with no leaks.68  The gasket was adjusted in place to eliminate
leaks.
     The Escolator chuck door gasket is composed of a compressible and de-
formable sealing member between outer layers of steel foil.   The major dis-
advantage of this chuck door gasket is that it does not continue to work
effectively.55 69  One of the principal reasons for the short life of the
Escolator chuck door gasket is that the load which is transmitted through
the  sealing edge to the sealing member is limited to the width of the
sealing edge, about 3 mm (0.125 in).  A new modification of this chuck door
gasket is provided by using a perforated steel strip as one layer of the
sealing member.70  The life of the modified gasket is estimated at twice
that of the unmodified gasket.  Another disadvantage of the Escolator chuck
door gasket was that it was fragile and difficult to handle and install.
The modified seal is claimed to be easier to use.
     Graphite-impregnated asbestos chuck door seals have been reported
effective for 2 month's use on batteries with about 40 chuck doors.  No
emission problem was reported on these chuck doors.55
                                 4-72

-------
     4.2.3.4  Hoods for Oven Door Fume Collection.  A technology that is in
the patent stage concerns a separate emission collection hood positioned
above each coke oven door.  Fans remove the emissions from the doors through
manifolds.  When the door is on the oven, the valve is slightly open, but
during the push, the valve is fully open to remove pushing emissions (Figure
4-21).71
     This technology is claimed to reduce emissions when the doors are
either in place or removed.  The emissions can be vented to a recovery
process.  The collection hood also may provide for sealing between a coke
oven jamb and a coke guide during pushing.  A variation of this system is
used with the Ministerstein pushing emission control system at the DOFASCO
Plant in Hamilton, Ontario.  Another variation is used on the pusherside
doors at Armco's Houston, Texas, coke battery.
     Another type of hood for the coke oven is provided by a hinged plate
on the door.  The hinged plate forms a channel between the door and the
buckstay (Figure 4-22).72  A suction conduit is located above these channels.
An alternate method is use of an air curtain along the buckstay.  This hood
technology provides for the removal of the escaping gases to eliminate the
emission of carcinogenic substances.  Capital investment in a fume-cleaning
system is also required for this technology.
     A third hood device is created by a door which is hinged to the buckstay
and creates a chimney to direct gases to a hood above the oven door (Figure
4-23).73  The chimneys are required because high winds deflect emissions
from the hoods over the door.74  This technology has been tested on 10
doors.  The capital and operating costs, as well as the energy requirements,
are expected to be similar to the other two hood systems.
     For hood collection and recovery systems, both the capital and operat-
ing costs are expected to be significant.  The nature of the operating cost
will depend upon the air treatment used.  The recovery of the emissions
would involve generation of solid waste and tar.  These wastes are expected
to be hazardous and may require special treatment before disposal; however,
they could be used for burning or recycle through the coke ovens if they
are in a suitable form.  One additional disadvantage of hoods is that cross
winds disrupt emission collection and may make some hood designs completely
ineffective.
                                 4-73

-------






















1
1
©!!<

1 1
1 1



^
\
/
(
1 i
I
)
/
l

1
/ \
(
\ (
1
t


0


/
r*
1
1
Y



1 i



7
1
1 1 |
I
IP
H
ra
1
.ii


$ t




















,1






1
II





I
1




i
B
in
,-:,


>ii®!
''•'
i1
i,
'

y'!
ii
i
i
i
i
i
f i
|I
ii
i.
:!
n
i
i
i'
ii
ii








ii



"ii





I
1

fn
h
jj


1 —

il'
















1
|

i
I
,ll


I
' I


01
r
r-


(
•








[


.

r
O1
1
1















I
I ;
]
j

l

H
! !
1 1
1 1
1 1
1
i













1
1
i
(
1
L

(

I


/
1
1
1













•- —














.
1



i
1
1
1
=3
n
V
                                           n
Figure 4-21. Coke oven door hood that assists in controlling
                  pushing emissions.71

                         4-74

-------
                Exterior View
               Cross Section View
Figure 4-22. Channel for the passage of gas formed by
            hinged plates on the door.72
                   4-75

-------
                                               CO
                                               ^.

                                                I
                                                •8
                                                3
                                                 to
                                                 o
                                                 i
                                                •o
                                                 (O
                                                •§
                                                 I
                                                •*-
                                                 o
                                                 ft
                                                 
-------
     4.2.3.5  Cokeside Sheds.  Cokeside sheds are similar in principle to
individual door hoods, but they enclose the whole oven door area within a
single structure.  A shed can be used to capture door leak emissions, but
its primary purpose is the control of pushing emissions.  Therefore, there
are no shed installations covering both the pusherside and cokeside doors.
Sheds can be constructed on the pusher side; however, this method would be
more complicated because the pusher machine needs more horizontal clearance
for the pusher ram and leveler bar.
     Obviously, shed exhaust must be vented to a fume cleaner just as door
hood exhaust is vented and this leads to the disposal problems previously
cited.  A further possible, and perhaps more significant, disadvantage to
using sheds is the potential for increased worker exposure to carcinogenic
emissions contained by the sheds.  Also, poor lighting conditions within
the sheds and the presence of the sheds make it more likely that good door
maintenance will not be performed unless there is a serious risk of door
fires.  Depending on the control device attached to the shed exhaust, the
net change in emissions may not be nearly as good as that resulting from an
effective door rehabilitation and maintenance program.
     One control device in present use is a wet electrostatic precipitator
(WESP).  A WESP has been used to remove BaP and other polycyclic organic
material from the gas collected by a cokeside shed at Wisconsin Steel's
battery.  Four tests for POM were conducted on the inlet and outlet of the
WESP when coke was not being pushed.75  The testing showed a removal effi-
ciency of over 93 percent for all POM except naphthalene, with an average
efficiency of 95.6 percent.  BaP was reduced from an inlet concentration of
141 ng/m3 to 1.82 ng/m3, an efficiency of 98.6 percent.   The overall POM
removal efficiency, including naphthalene, was 69 percent.76
                             ®
     This WESP was a MikroPul  type used to remove pollutants from pushing
emissions and fugitive emissions from 45 cokeside doors.  The precipitator
collector consists of two parallel units, each designed for a flow rate of
50 m3/s (100,000 actual ft3/min).   The design inlet and outlet temperatures
are 73° C and 43° C, respectively.   Each precipitator is a two-field unit
which provides for about 4.9 m of electrical treatment length.   From each
precipitator, the gas flow passes to an induced-draft fan.   The two fan
                                 4-77

-------
outlets are connected to a common 2.4-meter (8-foot)  diameter stack which
is 24 m (80 ft) high.   A common water-treatment  system including pH control
tanks and recirculating pumps is located under the precipitators.   The
design water flow rate is about 0.04 m3/s (600 gal/min).76
     A test was conducted on Armco's Houston,  Texas,  cokeside shed during
nonpush periods.   The data showed an emission  rate of 11.2 to 13.1 kg
benzene soluble organics (BSD) per hour during normal production with a
removal efficiency by WESP of 43 to 52 percent.77  Although the WESP design
parameters are essentially the same as those described previously (same
manufacturer), one of the two parallel units was inoperative because of
corrosion problems; therefore,  a lower efficiency was observed.78  An
average of 34 percent of the cokeside doors were leaking during these two
test runs, and the BSD emission factor was about 0.25 kg of BSD per megagram
of coal charged.77
     4.2.3.6  Vented Plug.54 55  At the time this chapter was written
(1979), the vented plug had been in use on commercial coke batteries for
about 1 year.  The vented plug provides a gas  escape from the bottom to the
top of the oven to reduce the pressures which build up at the base of the
oven door, thereby reducing leaks.  In practice, emissions from oven doors
were reduced by 5 to 10 percent when initially the emissions were in the 40
to 60 PLD range.55  However, there has been no report of the effectiveness
of the vented plug when initial leakage was in the 10 PLD range.
     The vented plug is an encouraging technology because it can supplement
seal technology and requires no energy input for its performance.   A sketch
of the vented plug is presented in Figure 4-24.   The large hole in the plug
has been less subject to plugging than vents along the sides of the coke
oven doors.  There is some variability in the cleaning needed for the
vented plug.  Although one company reported no required cleaning and no
heat damage,55 another company reported fouling and plugging.69  The differ-
ence in performance of the two vented plugs could reflect process differences
in the batteries.  The type of coal used and the amount of air entering the
process could account for differences in the carbon formation in the plug.
One disadvantage of vented plugs is that they are more expensive and tend
to break more easily than conventional plugs.
                                  4-78

-------
Chuck Door
                                   Gasses Vented to
                                   Top of Oven
Latch Bar
                                      Large Vent Area
                                      at Top of  Plug


                                       Chimney or
                                       Vent Area
                                       Thru Plug
                                         Gasses
                                         Entering Plug
              Figure 4-24. Vented plug on pusherside door.55
                          4-79

-------
     4.2.3.7  Jamb Design.   A recent development is  a novel  jamb design
(Figure 4-25) which can be straightened while it is  on the battery.   The
jamb is designed to reduce stress creep.   Theoretically,  the design  allows
the jamb to operate at lower temperatures.   Twenty-five of these jambs  were
in operation in August 1978.
     One possible advantage of this jamb is that it  may have a longer
life-time than conventional jambs.   The economics depend upon the replace-
ment rate, because the modified jamb is approximately four times as  expen-
sive as conventional jambs.  However, the jamb fabrication may be somewhat
difficult in large scale production.  There has been no evidence of  any
emissions reductions from the use of the modified jamb.55
     4.2.3.8  Heat-Settable Sealant.79  A heat-settable sealant, claimed to
be an improvement over traditional  luting, is formed with starch, water,
and clay slurry and injected into a channel in the door seal.  Some  small-
scale trials and plant trials of one or two doors on a battery have  been
attempted with this emission control technique.  The process is claimed to
have low manpower and operating cost and to eliminate fumes and fires.
Better heat distribution is obtained in the oven because of the reduction
in leakage.  Door warping  is minimized and accurate leveling of doors is
not as critical.  Cleaning of the door and jamb are eliminated, because the
sealant does not stick to the oven.  These advantages are claims and have
not been demonstrated in full-scale operations.
     4.2.3.9  Inboard-Luting.  One technology that has been proposed to re-
duce oven emissions is use of luting in addition to current metal-to-metal
seals.43  The advantages of this technique are that it has the potential
for eliminating all emissions and can be used on old batteries with defec-
tive seals.  Seal and jamb damage caused by fires and high temperatures is
reduced.  Air leakage into the oven in the later stage of the coking process
would be reduced.
4.2.4  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
                                 4-80

-------
                                           €/»
                                           .*
                                           o
                                           o
    0)
                                                  (0
                                                  o
                                                 55

                                                 U)
                                                  £
                                                  §>
4-81

-------
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 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.
Periodic cleaning remedies this plugging problem.  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.   These upsets
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 the steam content of the gas.
Both of these factors  tend to increase leaks and door sealing time.
     In the previous sections, the details of existing control systems,
their susceptibility to breakdown, and the avoidability of breakdown are
discussed in detail for each control technique.  Data are unavailable on
the frequency of occurrence and the  increase in emissions resulting from
these upsets and breakdowns.
4.2.5  Door Controls for Tall Ovens
     The metal-to-metal seal technology that was previously discussed
applies 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.
     There are several possible explanations for this control  level.  The
potential for  leak occurrence is  almost doubled  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
                                 4-82

-------
doors for batteries operating at the same collecting main pressure.   Also,
when dry coal is charged, the resulting pressure surges may increase leaks.
Nine of the 10 batteries presently charging dry coal have tall ovens.   Door
leak data were obtained for three tall batteries.   National Steel's Battery 1
at Weirton averaged 2.9 PLD over 4 days of observations with a range of 1.2
to 4.8 PLD.24  For U.S. Steel's Battery 2 at Fairfield, seven observations
over 4 days averaged 7.2 PLD with a range of 2.6 to 12.7 PLD.27  Lone Star's
new 5-meter battery averaged 0.4 PLD with a range of 0 to 1.4 PLD from 21
observations.23  The Lone Star data do not include chuck door leaks.
4.2.6  Summary of Door Leak Control Performance
     A review of EPA data on door leak control performance revealed that
the best control of door leaks is provided by the modified seal technology
coupled with a leak control program of routine inspection, cleaning, and
repair.  This technology has been implemented on the batteries at CF&I and
U.S. Steel, Clairton.  The data for the three CF&I batteries and 14 U.S.
Steel, Clairton batteries are summarized in Table 4-10.  A wide range of
battery types is included in the data base.  The types of original door
construction include Koppers, Koppers-Becker, and Wilputte.  The battery
age ranges from 1951 to newly rebuilt, and batteries that are scheduled for
rebuilt or replacement in the next 2 to 6 years are included.
     The test method for door leaks states that PLD is calculated by divid-
ing the number of leaking doors by the total number of doors on operating
ovens.  The test method also recommends an attempt to observe doors that
were previously blocked from view; however, the total number of doors is
always used in the denominator of the PLD calculation.  Some of the data
summarized in Table 4-10 were originally reported as the number of leaking
doors divided by the number of doors observed.  These data were converted
to be consistent with the proposed test method summarized in Appendix D.
In most cases, the orginally reported PLD changed by only 0.1 to 0.2 percent.
     No tall batteries that had implemented the modified seal technology
were identified.  Because tall batteries may have a more difficult control
problem, the industry was surveyed to determine how door leaks from tall
batteries were controlled.   U.S. Steel Battery 2 at Fairfield is a 6-meter
battery that has accomplished effective control by a door leak control
                                   4-83

-------
                     TABLE  4-10.   SUMMARY  OF  DOOR  LEAK DATA0
Company
U.S. Steel,
Clairton20



National Steel,
Weirton24
U.S. Steel,
Clairton20
CF&I, Pueblo80
U.S. Steel,
Clairton20

U.S. Steel,
Fairfield27
CF&I, Pueblo80
U.S. Steel,
Clairton20


CF&I, Pueblo80
U.S. Steel,
Clairton20

U.S. Steel,
Fairfield27
U.S. Steel,
Clairton20

Battery

3
1
2
7
i
lb

21
C

22
19

9C
B

20
16
10
D

8
17
i
2b

11
9
Date

2/80
8/79
8/79
2/80

6/82

8/79
3/78

8/79
8/79

6/79
3/78

8/79
2/80
2/80
3/78

2/80
2/80

6/79

2/80
2/80
Average
PLD

0.8
1.0
2.2
2.9

2.9

3.0
3.2

3.5
3.9

3.9
4.3

4.5
5.0
5.8
5.9

6.1
6.2

7.2

8.9
9.4
Range Number of
PLD observations

0 -
0 -
0 -
0 -

1.2 -

1.1 -
1.1 -

2.3 -
3.4 -

2.7 -
0.8 -

2.3 -
4.2 -
4.8 -
1.1 -

5.5 -
4.9 -

3.0 -

5.5 -
3.9 -

1.6
2.3
5.5
5.5

4.8

5.7
6.4

5.2
4.6

4.8
9.2

7.5
6.7
7.1
16.1

6.3
8.2

13.0

14.1
16.4

4
7
7
4

5

5
27

6
5

6
28

9
4
2
27

4
4

7

3
4
 The batteries at CF&I  and
 modified seal technology.
3 A 6-meter battery without
"A rebuilt battery without
U.S.  Steel, Clairton have implemented the

the modified seal technology.
the modified seal technology.
                                 4-84

-------
program of monitoring, cleaning, and repair or replacement of damaged
seals.   Battery 2 has a typical spring-plunger knife edge seal, an auto-
matic door cleaner supplemented by manual cleaning, and an accurate door
positioning system.   Although this 6-meter battery does not use the
modified seal technology, the baseline requirements of a door leak
control program have been implemented.  This battery was added to the
data base to represent the control performance that could be attained by
a tall  battery.  The control performance of Battery 2 supports the
observation that many short batteries can control door leaks better than
tall batteries.  However, 7 PLD is better than the control performance
of two U.S. Steel, Clairton short batteries that have implemented the
modified seal technology.
     Battery 9 at U.S. Steel, Fairfield was examined to determine if a
recently rebuilt battery (1979) could control emissions better than an
older battery with the modified seal technology.   The control technique
for this battery is also a door leak control program that includes
proper door placement, routine cleaning and repair, and a detailed
procedure for cleaning the door jamb and seal.  The performance of this
rebuilt battery is not obviously better than that of the older U.S. Steel,
Clairton batteries that have been retrofitted with the modified seal
technology.
     The statistical analysis of the door leak data revealed that a
Poisson distribution was applicable.  The 95-percent confidence level
was estimated for various average PLDs, based on a minimum of three
inspections.  The average PLD and the 95-percent level associated with
this average are listed in Table 4-11.

              TABLE 4-11.  CONFIDENCE LEVELS FOR DOOR LEAK DATA
Average PLD
4
6
7
9
95-percent confidence level3
7
8
10
12
 Based on three inspections.  For a battery of 50 ovens with 100 doors, at
 least 300 individual door observations are made.
                                 4-85

-------
     Ten of the U.S.  Steel and CF&I batteries averaged 4 PLD or less
with a 95-percent confidence level  of 7 PLD.   A total  of 18 of the 20
batteries listed in Table 4-10 averaged 7 PLD or less  with a 95-percent
confidence level of 10 PLD.   Two of the U.S.  Steel  batteries (Batteries 9
and 11) averaged around 9 PLD during the EPA inspection.   Self-monitoring
data for Battery 9 were available for 685 daily observations over a
period of 25 months of operation after the EPA inspection in February
1980.   These data averaged 5.3 PLD  (excluding 2 doors) or 6.9 PLD with
no exclusions.   Battery 11 remained in operation for 5 months following
the EPA inspection and 130 observations revealed an average of 4.1 PLD
(excluding 2 doors') or 5.7 with no  exclusions.  The long-term data for
Batteries 9 and 11 indicated that door leak emission control improved to
levels better than 7 PLD.  Consequently, an average performance of 7 PLD
(maximum) can be obtained with U.S. Steel's door control technology with
an upper 95-percent confidence level of 10 PLD.
4.3  TECHNOLOGY FOR THE CONTROL OF  TOPSIDE LEAKS (CHARGING PORT LIDS AND
     STANDPIPES)
4.3.1  Description
     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 seats,
          Patching or replacing cracked standpipes,
          Sealing lids after a charge or whenever necessary with a
          slurry mixture of clay, coal, and other materials (commonly
          called lute), and
          Sealing cracks at the base of a standpipe with the same slurry
          mixture.
     Luting mixtures are generally prepared by plant personnel according
to formulas developed by each plant.  The consistency (thickness) of the
mixture is adjusted to suit different applications.  Charging port lids
are relatively horizontal; therefore, a thinner mixture can be used to
                                 4-86

-------
seal them.   Standpipe lids come in a variety of positions; those that
are not horizontal require a thicker mixture to prevent runoff.   In one
study,81 at least three surveyed plants reported that adjusting the
mixture consistency allowed successful use of luting mixtures on standpipe
lids.   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 plants report that residue from lute interferes with the use
of magnetic lid lifters.  This interference could occur if the luting
mixture was applied carelessly or was covering the lid or if the mixture
residue was allowed to build up.   Plant personnel from two plants contacted
in one study81 reported that they have no problem with lute residue, but
they were aware that residue could accumulate without proper attention
to operations.  The actions taken to avoid the problem were careful
application of the lute and cleaning the residue from the lids.
     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 seated and
provide a better seal.  Even with such equipment, manual effort will
still  be required to seal leaks.
     After closely examining their program for controlling topside
leaks, CF&I made several changes in their emission control methods.  One
method used increased supervisor attention to emphasize that oven top
workers use additional care and effort to replace and set standpipe caps
and charging lids properly.   However, the configuration of standpipe
caps and goosenecks was not conducive to luting, and numerous caps that
were not making an adequate mechanical seal had to be hit into place
with a bar.  The caps were placed in a lathe and the diameter was reduced
19 mm (0.75 in) to allow the cap to be self-positioning when dropped
into place.  The modified cap seated lower in relation to the gooseneck;
this allowed luting material to be held in position and effectively
sealed the caps.19
     CF&I decided "that a better overall luting practice would be the
most expeditious manner to achieve the lowest top emissions."  An improved
                                4-87

-------
luting slurry was developed and luting tanks  were added to  the larry
cars to carry the material  on board.   The luting tanks on the larry cars
make the luting slurry readily available for  the larry car  operator to
seal the standpipe caps and for the lidman to seal  the charging port
lids.19  An alternative to  carrying the luting tank on the  larry car
would be to place pails of  the luting slurry  at convenient  locations on
the battery top.   This reduces the worker effort to lute or relute leaks
because the material is readily available and the worker would not be
required to walk the length of the battery to obtain the necessary
luting mixture.
     Another improvement made at CF&I was in  the control of collecting
main backpressure because pressure surges of  10 to 20 mm of water can
break the seal on luted lids and caps.  Improved backpressure control
was obtained by altering the procedure for turning on the steam aspiration
and by modifying the pressure control system  for the collecting main.
Another pressure control problem was created  when the steam aspiration
was inadvertently left on after an oven was charged.  Each steam aspiration
system was fitted with a pipe nipple having a small hole and a directional
check valve installed between the nipple and  aspiration jet.   If the
steam aspiration system were inadvertently left on, the steam flow from
the small hole would be visible and audible and would indicate a procedural
deviation.19
     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 upon 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.  For some batteries,
the larry car operator or helper seals standpipe caps; on other batteries
the lidsman performs this function.  After increasing the topside work
force to improve emission control, Kaiser Steel used 48 lidsmen for 7
batteries with 45 ovens each.  Kaiser used manual lid lifting, but only
                                  4-88

-------
7 "Hdsmen per battery were required because the batteries were operated
as 3 units with overlapping job responsibilities on the individual
batteries.  The lidsmen job duties included:82
     1.    Inspect every gooseneck, standpipe and charging steam jet prior
          to charging the oven, and clean as necessary.  Routinely clean
          each of these areas on oven marked "cleaners" (approximately
          every fifth oven).
     2.    Ignite standpipe gas emissions if oven is smoking prior to the
          push.  If the emissions cannot be ignited or leaks from the gas
          main cannot be blocked with charging steam, reseal the standpipe
          cap.
     3.    Work with the larry operator to charge each oven according to the
          published staged charging procedure.
     4.    After charging is complete, lute all charging hole lids,  standpipe
          caps, and standpipe bases.   Relute as necessary to maintain
          smokeless oven tops.
4.3.2  Performance of Control for Topside Leaks
     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 (charging port lids and offtakes) 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.  (This method of data collection is described in
more detail in Appendix D.)
     As discussed in Section 4.3.1, the primary technique used to reduce
the number of topside leaks is luting (sealing the leaks).  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
                                 4-89

-------
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.   Theoretically,
the percentage of topside leaks measured by the method cited previously
can be reduced to near zero if all leaks are sealed as soon as they
appear.
     Data on percent leaks for charging port lids are presented in
Tables 4-12 and 4-13.   The data were gathered at U.S. Steel, Fairfield
and U.S.  Steel, Clairton.   The data in Table 4-12 for two batteries at
Fairfield show that the average for any test never exceeded 0.5 percent.
Out of 221 traverses,  the highest single reading was only 3.2 percent.
During 6 months of data gathering, the highest monthly averages for
seven batteries at Clairton ranged from 0.3 to 0.9 percent.  Out of
1,033 traverses on the seven batteries, the highest single reading was
only 3.5 percent.  This consistently high level of control is attributable
to the effort used to seal leaks.  The occasional higher reading shown
by the data can be prevented by more diligent sealing of leaks as an
increase in the number of leaks is observed by the plant operators.
     Data on offtake leaks at Kaiser Steel's Fontana, California batteries86
show the effect of additional manpower at one plant.  Table 4-14 presents
data for 3 months when one employee per battery per shift was responsible
for luting topside leaks and tending to lid removal and replacement.
Below these data are data for 6 months after one additional employee was
provided to help with luting and tending the lids.  For all seven batteries
at Fontana, there was substantial improvement in control  of offtake
leaks after the second employee was hired.  The improvement ranged from
35 to 56 percent fewer leaks for the seven batteries.  The level of
effort provided was sufficient to lower the average PLO below the local
requirement of 10 PLO.  The percentage of individual readings with a
value of 5 percent leaks or less is also shown on Table 4-14.  From
these data, it can be seen that the number of higher single readings is
also reduced by an increase in manpower to seal leaks.  The remaining
higher single readings can be controlled by frequent scrutiny of potential
sources of leaks and prompt sealing of the leaks when they appear.
                                 4-90

-------
      TABLE 4-12.  VISIBLE EMISSION DATA ON TOPSIDE LEAKS FROM CHARGING
                  PORT LIDS, U.S. STEEL, FAIRFIELD15 83 84
Date Battery
July, 1974

September, 1975

April, 1976

December, 1976

5
6
5
6
5
6
5
6
Number of
traverses
34
15
21
23
30
28
36
34
Percent leaks .
observed per traverse
a Range
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0 -
2.9
2.9
0.3
0.3
1.0
1.0
3.2
0.6
Average
0.3
0.3
0.01
0.04
0.1
0.1
0.5
0.2
A traverse is one recording of leaks from an entire battery.
Percentage of total potential leaks (308 for U.S. Steel, Fairfield).
                                4-9i

-------
TABLE  4-13.  VISIBLE EMISSION DATA  ON TOPSIDE  LEAKS FROM CHARGING PORT  LIDS
                             U.S. STEEL, CLAIRTON85
Date
September. 1977





October, 1977






November, 1977






December, 1977






January, 1978






February, 1978






Battery
12A
15
16
17
19
21
22
12A
15
16
17
19
21
22
12A
15
16
17
19
21
22
12A
15
16
17
19
21
22
12A
15
16
17
19
21
22
12A
15
16
17
19
21
22
Number ofa
Traverses
26
26
25
25
26
26
26
26
29
29
28
29
28
28
21
26
26
26
23
24
25
1«
25
26
25
22
26
27
25
14
20
20
20
22
21
16
27
27
27
27
27
27
Percent
Observed Per
Rama
0 - 0.6
0 - 1.2
0 - 1.2
"j - 0.3
0 - 1.1
0 - 1.5
0 - 1.1
0 - 0.6
0 - 0.8
0 - 0.3
0 - 0.8
0 - 1.1
0 - 0.9
0 - 1.1
0 - 0.5
0 - 1.2
0 - 1.2
0 - 0.8
0 - 0.5
0 - 0.6
0 - 0.6
0 - 0.9
0 - 0.4
0 - 0.8
0 - 0.4
0 - 0.9
0 - 1.2
0 - 0.9
0 - 3.0
0 - 1.2
0 - 1.2
0 - 2.0
0 - 2.6
0 - 3.5
0 - 2.3
0 - 1.8
0 - 2.9
0 - 1.2
0 - 3.3
0 - 1.7
0 - 3.2
0 - 0.9
Leaks
Traverse*5
Average
0.1
0.2
0.2
0.2
0.3
0.3
0.5
0.1
0.2
0.1
0.1
0.2
0.2
0.3
0.1
0.1
0.1
0.1
0.1
0.2
0.1
0.1
0.0
0.1
0.1
O.I
0.3
0.2
0.3
0.2
0.3
0.4
0.4
0.9
0.7
0.7
0.8
0.3
0.5
0.3
0.6
0.3
     a A traverse is-one recording of leaks frox an entire battery.

     b Percentage of total  potential leaks.
                                  4-92

-------













tf\
w
00


Q£
LU

to
1— 1
jr
LU
LU
i
CO
to
t_l
>

•si"
i-l
1
*!"
LU
	 1
CO

aj
4J
2
JC
*
0)
,0
in
^£
T)
a
c
a
£.













































o








u.









LU









O







 Ol
«t *0
O «
O) *-.
*» c o
c: f- («
at -a m u»
u «o at
U 0) *> r-
ai *. «o
cu
i
t.
a* a>
5 5?
O 1/1
CD t-
*•» C O
C f t/i
a> "a m w*
O fO 0)
t. at *J •—
0) 1- *O
a.

i
i_
a* U
** c o
C— at
v *a vo «
u >a a>
t. 01 *J t—
41 t- 10
0.
U
ii> a>
> 01
< 10
*•
O u<
a> b
*> C 0
C i- i/l
a) *D in IA
u 10 a
I. V «J •—
v b n
a.
i.
a> aj
> o
"< /O
tt_
O i/>
cn t-
** c o
CNT v>
a) *n in (A
u n» a>
I_ OJ 4J r-
« fc. (0
Ck.
1
(_
 CU
> cn
<: as
O vi
cn c.
*> C 0
C •»- to
a* T3 m wi
o «o o
t. a> 4-> >~
01 U «O
cw

1
t-
a> cj
< S*





41
*J
10
O



k
«D Of
C?S
W|
^J E
\


CM rO rH
CM CM CM




«-« WO CM
(s! r^ co


iO i0 CO
ui m CM



cn m o
^ in r**


O CM 10
^ ^ CO





en vo r»»
m in CM


2 CM CO
VD CO




^- <• o
^- V cn



id ««r cn
t-l — ^ CM

o v o
a s *




PO en *r
in o r-»



—4 O CO
in <• rn

«D cn t^
m i« m



•— t m -«r
uo -^ *r



r^
f** f*^
cn r*» r*.
-Hr-.cn
cn »-(
i x 8
5- 5 ^





r-t




eg
CM




10
p^


r^
*
c
•?
I


u
a.
«c






m r-4 r— • 10 ^ »o
r-> co co <• u> co




cn <«• ^- rH m *r
cn cn CM m ^- PO

*
gio r^ to cn cn
r** co co f* ^r



co *r co r- co co
^r cn «M c4 rn **•


in *n r*. cn to o
cn cn cn co co o
r-t




cn o us o u> CM
r-4 CM r-t CM CM CM


So h* co o iO
CT> o> cn o co
iH



co CM cn CO CO rn
tH CM rH r-J rH CM



to CM co *r cn r^
in r-. u3 cn *r m

rs. co rH m co cn
*d* cn ^ ai m <-




10 to o *o r-* r*
i** co o cn m m
r-t



cn r-t CM r-t CM <•
CNJ CM rH CSJ ^ -^

CO rH i— t CM rH r—
h- cn cn co r^ m



f* CM O en ««• rH
CM CM CM CM cn ^

r*. r^ r*.
cn r^ f"^ f**
t-* •-• r»« cn cn
P-B Cn r-H rH
r«* cn s- rH
r*. rH 0) £- (-
cn .c f- a; ai
•-i *> E ft" -a .a
«fl » 3 «J O Of Of
•^ tn a. ^-» > u
3 3 cu u o a;
-3 
c
^o
c
01
4-»
?
a
?
jj
3
r~

fc.
O
•*-
4>
JQ
M
C
o
Q.
in
>t
C
Of
*j
4^
2
U
Q)
a.

**
H-
£
V)
t.
S.
^


x
i
1
01
ai
4J
(O
e

X
o
1
1 ^
1")
4-93

-------
TABLE 4-15.   SUMMARY OF LID LEAK DATA
Company Battery
U.S. Steel, Clairton20 88 1


2


3


7


8


9


10


11


12

12A
13
14
16


Date
8/78
8/79
2/80
8/78
8/79
2/80
8/78
8/79
2/80
8/78
8/79
2/80
8/78
8/79
2/80
8/78
8/79
2/80
8/78
8/79
2/80
8/78
8/79
2/80
8/78
8/79
8/78
8/78
8/78
8/78
8/79
2/80
Average
PLL
0.3
0.2
0.2
0.2
0.4
0.3
1.4
0.5
0.3
0.9
1.6
0.4
0.3
0.1
0.3
0.9
0.8
0.4
0.5
0.7
0
0.5
0
0.3
0.5
0
0.6
0.4
0.3
0.2
0.1
0.1
Number
of
Range obser-
PLL vations
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0.4 -
1.2 -
o -
0 -
0 -
o -
0 -
0.4 -
0 -
0 -
0.4 -
-
o -
-
o -
o -
-
o -
o -
o -
o -
o -
o -
0.8
0.4
0.4
0.8
2
1.2
4.7
2
1.2
2.0
2.4
0.8
0.8
0.4
0.8
2.0
1.2
0.8
1.2
1.2

1.6

0.4
1.6

2.6
1.2
1.6
0.4
0.4
0.4
9
5
4
9
5
4
9
5
4
8
3
4
9
3
4
9
3
4
9
3
3
9
3
3
9
3
9
9
9
9
8
3
                                              (continued)
             4-94

-------
TABLE 4-15.   (continued)
Company Battery
17


19


20

21


22


U.S. Steel, Fairfield27 2
6
9
Kaiser, Fontana46 F
G
Date
8/78
8/79
2/80
8/78
8/79
2/80
8/79
2/80
8/78
8/79
2/80
8/78
8/79
2/80
6/79
6/79
6/79
5/79
5/79
Average
PLL
0.2
0.2
0.1
0.4
1.2
0.4
0.9
0.7
0.2
0.5
0.4
0.2
0.9
0.2
0.1
1.8
0
0
0.2
Number
of
Range obser-
PLL vations
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0 -
o -
0 -
-
-
-
o -
1.2
0.8
0.4
1.1
3.5
0.9
1.8
1.5
0.6
0.9
0.6
0.6
1.8
0.9
0.5



1.7
9
7
3
9
6
5
7
5
9
6
4
9
7
5
6
1
6
8
8
      4-95

-------
TABLE 4-16.   SUMMARY OF OFFTAKE LEAK DATA
Company
U.S. Steel, Clairton20













U.S. Steel, Fairfield27
Battery
1
2
3
7
8
9
10
11
16
17
19
20
21
22
9
Date
8/79
8/79
8/79
2/80
2/80
2/80
2/80
2/80
8/79
8/79
9/81
9/81
9/81
9/81
6/79
Average
PLO
0.7
0.8
0.9
1.5
1.7
2.3
0.6
3.4
0.8
1.8
1.5
1.4
3.0
3.5
0.3
Number
of otr
Range serva-
PLO tions
0 -
0 -
0 -
0.8 -
0 -
0.8 -
0 -
0 -
0 -
0.9 -
0 -
0.6 -
0 -
0.6 -
0 -
3.5
2.5
2.5
2.5
3.6
4.4
0.9
6.8
2.6
3.6
3.1
1.9
8.5
9.0
1.7
7
7
7
4
4
4
3
3
7
7
5
5
5
5
6
              4-96

-------
     A summary of the data collected by EPA for lid leaks is given in
Table 4-15.  All of the U.S. Steel, Clairton batteries are included in
the data base and reveal a range of averages from 0 to 1.6 PLL.  Fourteen
batteries  in the data base averaged 0.2 PLL or less during at least one
inspection, and the highest average for all 23 batteries is 1.8 PLL.
     The statistical analysis concluded that a Poisson distribution was
applicable for both lid and offtake leaks.  The performance of batteries
averaging  0.2 PLL yields a 95-percent confidence level of 1 PLL, when
averaged over three runs.  The highest average observed at U.S. Steel,
Clairton during EPA inspections (Battery 7 in August 1979) was 1.6 PLL
with a 95-percent confidence level of 2.7 PLL when averaged over three
runs.  The highest average listed in Table 4-15 is 1.8 PLL with a 95-percent
confidence level of 2.9 PLL, averaged over three runs.
     A summary of data collected by EPA for offtake leaks is given in
Table 4-16.  The inspection data for four batteries at U.S. Steel,
Clairton in February 1980 are not included in the data base; Batteries
19 and 20 were experiencing problems with gooseneck failure during the
inspection, and Batteries 21 and 22 were adversely affected by offtake
leakage at the slipjoint.87  Investigation and correction of these
problems was necessary to enable these batteries to meet their existing
consent decree requirements of 5 PLO.   These four batteries were reinspected
in September 1981 after the offtake leak problem had been corrected and
show levels of 1.4 to 3.5 PLO.   Battery 9 at Fairfield was included to
indicate the potential control  performance of a rebuilt battery.
     Six of the batteries in the data base averaged 0.9 PLO or less with
a 95-percent confidence level of 4 PLO when averaged over three runs.
The highest value was 3.5 PLO (Battery 22 at Clairton) with a 95-percent
confidence level of 6 PLO when averaged over three runs.
     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.
4.4  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.
                                  4-97

-------
2.    Barnes, T.  M.,  et al.  Control of Coke Oven Emissions, Ironmaking, and
     Steelmaking.   19/b.   p.  16b.

3.    Iversen, R. E.   EPA trip report of a visit to U.S. Steel, Clairton.
     February 7, 1975.

4.    Coke Ovens Designed for Increased Steam Aspiration to Curtail Effluents.
     Industrial  Heating.   39:1098-1100.  June 1972.

5.    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.

6.    Coke Battery Survey - Procedures Description and  Data Presentation.
     U.S. Environmental Protection Agency.  Publication Number
     EPA-330/1-77-012.   December 1977.  p. 35 and D-17 to D-22.

7.    Raw data sheets for an emission test at CF&I, July 26-30, 1977.

8.    Coke Battery Data Presentation.  U.S. Steel, Gary, Indiana.  EPA
     Contract No.  68-02-2814, Work Assignment 6.  June 12, 1979.  114 p.

9.    Reference 6,  p.  21.

10.  Letter from Winkley, J.  C. CF&I, to Dickstein,  I. L., EPA,  Region VIII.
     July 29, 1977.

11.  Jones, L. G.   EPA trip report of a visit to U.S.  Steel, Gary.  September 1,
     1976.

12.  Munson, J.  G.,  et al.  Emission Control in Coking Operations by  Use  of
     Stage Charging.   Journal of the Air Pollution Control Association,
     24:1059-1062.   1974.

13.  Trenholm, A.  R.   International Trip Report.  February 18, 1976.

14.  Reference 5,  p.  30-31.

15.  Iversen, R. E.   EPA trip report of a visit to U.S. Steel, Fairfield,
     December 4, 1974.

16.  VanOsdell, D.  W., et al.  Research Triangle Institute trip  report to
     U.S. Steel, Fairfield.  October 27, 1977.

17.  Baldwin, V. H.  and D. W. Coy.  Study to Develop Retrofit  Information
     and Other Data for Use in Setting Standards for Coke Oven Emissions.
     Prepared by the Research Triangle Institute for the  Environmental
     Protection Agency.  EPA Contract Number 68-02-2612.  Task  39.  March,
     1978.  112 p..

18.  Phelps, R.  G.   Manager of Primary Technology, Inland Steel  Company.
     Comments on Proposed National Standards: By-Product  Coke  Oven Wet-Coal
     Charging and Topside Leaks.  July 26, 1978.


                                  4-98

-------
19.   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.  p. 920-925.

20.   Data collected by GCA Corporation and EPA Region III at U.S. Steel,
     Clairton.   August 1979, February 1980, and September 1981.

21.   Data collected by GCA Corporation and EPA Region III at Bethlehem
     Steel,  Burns Harbor.  May 1980.

22.   Raw data sheets on file at EPA Region III for J&L, Pittsburgh emission
     test.  Inspections in December 1978 and August 1979.

23.   Data collected at Lone Star Steel for Texas Air Control Board by
     Radian Corporation.  December 26, 1979.

24.   Data collected by GCA Corporation and EPA Region III at National Steel,
     Weirton.  August 1981 and June 1982.

25.   Raw data sheets on file at EPA Region III for Shenango, Incorporated,
     Neville Island, Pennsylvania.  Inspection in October, 1980.

26.   Coke Battery Survey:  Procedures, Description, and Data Presentation,
     U.S. Steel, Fairfield.  EPA-330/1-77-012.  December 1977.

27.   Data collected by GCA Corporation and EPA Region III at U.S. Steel,
     Fairfield.  June 1979.

28.   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 Enginering Division.
     U.S. EPA.  p.  4-39 to 4-42.  June 1978.

29.   Stolz,  J.  H.   Coal Charging Pollution Control Demonstration.
     EPA-650/2-74-022.  March 1974.

30.   Bee, R. E., et al.  Coke Oven Charging Emission Control Test Program  -
     Volume I.   163 pp.  July 1974.

31.   Reference 30, p.  111-117.

32.   Reference 29, p.  67.

33.   Reference 29, p.  6.

34.   Balla,  D.  A.  and G. E. Wieland.  Performance of Gas-Cleaning System on
     Coke Oven Larry Car at Burns Harbor.  Iron and Steel Engineer.
     December 1970.   p. 79.

35.   Reference 5,  p. 27.

36.   Reference 5,  p. 28-29.
                                 4-99

-------
37.   Reference 5, p. 26-27.

38.   Reference 5, p. 28.

39.   Lownie, Jr., H. W. and A. 0. Hoffman.  A Research Approach to Coke-Oven
     End-Closure Problems.  Ironmaking Proceedings, 35, 109(1976).

40.   Giunta, J. S.   U.S. Steel Development of Coke-Oven-Door System
     Technology.   Paper 12d.  AIChE Meeting, Atlanta.  February 26-March 1
     1978.

41.   Stanley, R.  W.   Coke Oven Door System:  Field Data.  Paper 12b.  AIChE
     Meeting, Atlanta.   February 26-March 1, 1978.

42.   McGannon, H. E. (ed.).  The Making, Shaping, and Treating of Steel.
     9th Ed.  U.S.  Steel Corporation.  Pittsburgh.  1971.

43.   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.

44.   Trip Report, U.S.  Steel.  Research Triangle  Institute.  September 8,
     1978.

45.   Schultz, L.  D.   Control of Coke Oven Door Emissions.  Allied Chemical
     Co., Semet-Solvay Division.  Paper 12c.  AIChE Meeting, Atlanta.
     February 26-March 1, 1978.

46.   Data collected by GCA Corporation and EPA Region III at Kaiser  Steel.
     May 1979.

47.   Clayton Environmental Consultants, Inc.  Coke Oven Emissions: Visible
     Observations of Coke Oven Door Leaks.  EMB Report No. 79-CKO-23.  p. 3.
     November 1979.

48.   Wolff, G. Jr.   Wolff Prechamber Door.  Advertising Literature.
     Ferrostaal Overseas Corporation.

49.   Data obtained from Jefferson County Department of Health.  Birmingham,
     Alabama.  1978.

50.   Reference 47, Koppers in Erie, Pennsylvania,  p. 1-3.

51.   Hopper, Allen T. and E. N. Kaznetsov.  Summary Report on Task-I
     Mathematical Modeling and Analysis of End-Closure Systems.  EPA
     Contract No. 68-02-2173.

52.   Hopper, Allen T., R. L. Paul, and D. N. Gideon.  Summary Working Paper
     on Task II  Physical Modeling and Laboratory  Experiments.  EPA Contract
     No. 68-02-2173.  January 10, 1978.
                                 4-100

-------
53.   Hoffman, A. E.  Report on Official Travel.  November 3,  1977.

54.   Barchfeld, F.  J., C. C. Gerding, and J. M. Stoll.  Coke  Oven Door
     Emission Control Technology.  Paper 12a.  AIChE Meeting, Atlanta.
     February 26-March 1, 1978.

55.   Trip Report, J&L Steel.  Research Triangle Institute.  August  22,  1978.

56.   Here's A Good News For You.  Advertising product  literature.   Iron
     Works Company,  Ltd.  1979.

57.   Sealant Interlake Asbestos.  Battelle  Columbus Laboratories.   Task 3
     Progress Narrative.  July 1977.

58.   Letter from Reichhardt, Heinz A. to McCrillis, Robert  C.,  Metallurgical
     Processes  (MD-62), EPA.  June 13, 1977.

59.   Klotz, W.  L.  Research Triangle Institute trip report  to National
     Steel's Zug Island plant.  September 18, 1979.

60.   Phelps, R. G.    AISI-EPA-Battelle Coke  Oven Door Sealing  Program.   JAPCA.
     29, p. 908, September 1979.

61.   Lownie, H. W.,  et al.  Study of Concepts for Minimizing  Emissions  from
     Coke-Oven  Door  Seals.  EPA-690/2-75-064.  July 1975.

62.   Telecon.   Jablin, R. with Feiser, Art, A. F. Industries.   July 10,
     1978.

63.   Coy, D. W.  Meeting with Freidhelm Haaf and Werner Schmitz,  Fried.
     Krupp GMBH, Oberhausen Works.  West Germany.  March 15,  1979.

64.   Letter from Reichhardt, Heinz A. to McCrillis, Robert  C.,  Metallurgical
     Processes  (MD-62), EPA.  July 1977.

65.   Acheren, Joseph Van and Linwood G. Tucker.  U.S.  Patent  4,067,778.
     January 10, 1978.

66.   Graham, J. P.  and B. P. Kirk.  Problems of Coke-Oven Air-Pollution
     Control.   The Metals Society.  London, p. 82-100.

67.   Lowry, H.  H. (ed.).  Chemistry of Coal Utilization.  Supplementary
     volume.  John Wiley and Sons, Incorporated.  New  York.   1963.

68.   Hoffman, A. L.  Interlake Trip Report.  June 1, 1977.

69.   Trip Report to  U.S. Steel.  Research Triangle Institute.   September 8,
     1978.

70.   Apparatus  for Sealing a Coking Chamber.  U.S. Patent 4,033,827.

71.   Oven Door  Fume Collection System.   U.S. Patent 3,933,595.
                                4-101

-------
72.   Coking Oven.  Hartung, Kuhn, and Company.  Dusseldorf, Germany.
     U.S. Patent 3,957,591.

73.   Control of Coke Oven Door Emissions.  Interlake Incorporated, Chicago,
     Illinois.   U.S. Patent 3,926,740.

74.   Telecon.  Jablin, R. with Armour, Frank, Interlake Steel.  June 30,  1978.

75.   Air Pollution Emission Test, Final Report, Wisconsin Steel Co.  Report
     No. 77-CKO-ll-A.  EPA Contract No. 68-02-2817.  May 1978.

76.   Barrett, R. E. et al.   Effectiveness of a Wet Electrostatic  Precipitator
     for Controlling POM Emissions From Coke Oven Door Leakage.   APCA Meeting,
     Houston, Texas.  June 1978.

77.   Benzene Soluble Organics Study - Coke Oven Door Leaks.   Clayton
     Environmental Consultants, Inc.  EMB Report No. 70-CKO-22.   December
     1979.

78.   Branscome, M.  Research Triangle Institute trip report to Armco, Inc. ,
     Houston, Texas.  October 23, 1979.

79.   Calderon, Albert.  Coke Oven Door with Heat Settable Sealants.  U.S.
     Patent 3,875,018.

80.   Data collected by National Enforcement Investigation Center,  Denver,
     Colorado at CF&I.  Inspections in February and March, 1978.

81.   Reference 17, pp. 26-77.

82.   Kaiser Steel Corporation.  Coke Oven Emission Control Plan.   May 1977.

83.   Reference 6, pp. D-28 to D-29.

84.   Reference 28, p. 4-56.

85.   Reference 28, p. 4-57.

86.   Reference 28, p. 4-61.

87.   Letter from Stromness, N. R. , U. S. Steel Corporation, to Maslany,  T.,
     EPA.  October 17, 1980.

88.   Data on file at EPA Region III for U.S. Steel, Clairton.  Inspections
     in August 1978.
                                 4-102

-------
                           5.   MODIFICATIONS

5.1  BACKGROUND
     The purpose of this chapter is to discuss coke plant changes that
could be deemed "modifications."  A national emission standard for
hazardous air pollutants (NESHAP) is applicable to both existing and
new sources of the hazardous pollutant to which the NESHAP applies.   A
new source is defined under 40 CFR 61.02 as one that commenced construc-
tion or modification after the proposal date of the applicable NESHAP.
An existing source is defined as any stationary source that is not a
new source.  An owner or operator of an existing source planning to
modify that source must apply to the U.S. Environmental Protection
Agency (EPA) Administrator for approval prior to beginning the
modification (see 40 CFR 61.07).
     A modification is defined under 40 CFR 61.02 as
     any physical change in, or change in the method of operation
     of, a stationary source which increases the amount of any
     hazardous air pollutant emitted by such source or which
     results in the emission of any hazardous air pollutant not
     previously emitted, except that:
     (1)  Routine maintenance, repair, and replacement shall not
          be considered physical changes, and
     (2)  The following shall not be considered a change in the
          method of operation:
          (i)  An increase in the production rate, if such
               increase does not exceed the operating design
               capacity of the stationary source;
          (ii) An increase in hours of operation.1
5.2  POSSIBLE MODIFICATIONS
     Equipment or process modifications the by-product coke industry
theoretically could make to the existing facility are numerous.
However, several factors greatly limit the determination of whether a
modification qualifies as a modification under Section 61.02.   First,
the existence of Occupational Safety and Health Administration (OSHA)
                                  5-1

-------
and State Implementation Plan (SIP) regulations establishes a regula-
tory baseline that the by-product coke industry is striving to meet.
Consequently, many modifications a plant would undertake on the exist-
ing facility are designed to comply with the existing regulation by
decreasing the hazardous air pollutants emitted to the atmosphere.
Second, the definition in Section 61.02 excludes modifications for
routine maintenance, repair, and replacement of equipment.   Once these
basic restrictions on possible modifications have been incorporated,
the actual number of Section 61.02 modifications becomes quite small.
5.2.1  Process Modifications
     The by-product coke industry is a mature industry with a mode of
operation that has been developed by over 50 years of experience.  It
is unlikely that a new coking process will be commercially available
within the next 10 years.  Some companies have experimented with using
inferior coking coals by either coal briquetting or formed coke processes,
but large-scale commercial use of these is not expected in the near
future.2  Thus, any process modifications will be within the process
description explained in Chapter 3.
     5.2.1.1  Dry-Coal Charging.  A modification to the charging
operation would be the use of dry-coal charging.  Dry charging may
reduce emissions from the charging operation at the battery topside,
but this reduction may be offset by an increase associated with the
use of a coal preheater.  Currently, 11 batteries in the United States
use dry charging.2  However, because EPA currently is developing
separate air emission standards for dry charging, any modification
that would use this process will be regulated by those specific standards.
     5.2.1.2  Recycle of Waste.  Because of the rapid volatilization
of polycyclic organic material (POM), the addition of flushing-liquour
sludge and/or coal tar to the coal mix for the coke ovens could increase
the amount of POM released during charging.  Data are not available to
quantify the effect of this modification.  (See Subsection 3.3.1.2.).
Currently, one plant has indicated the possibility of adding flushing-
liquor sludge to the coal mix.  In the future, plants may incorporate
this process for the ultimate disposal of various sludges that no
                                  5-2

-------
longer can be disposed of in a landfill because of the Resource Conser-
vation and Recovery Act (RCRA).   This process change might be considered
a modification if it increased pollutant levels.
     5.2.1.3  Green Coke.   The production of green coke is undesirable
and affects emissions during pushing.  Several operational modifications
to the coke ovens could increase the occurrence of green coke.   Decreasing
the coking time, overfilling the ovens, and lowering the heating value
of the underfiring gas to the battery all could contribute to the coke
not being fully carbonized.   Green coke is not a desired product, so
intentional process modifications to produce it are unlikely.  However,
any changes that increase the frequency of green coke production might
be considered a modification.
5.2.2  Equipment Modifications
     The recent trend in the by-product coke industry has been to make
equipment modifications that decrease the pollutants from the affected
facility in an attempt to meet existing standards.  Combined with the
definition of modification that excludes routine maintenance, repair,
and replacement of equipment, essentially all rational equipment
changes probably will not be considered modifications.  Therefore, the
following probably would not be classified as Section 61.02 modifica-
tions:
          Change in hopper size and independent control of coal flow
          and independently operated drop sleeves on the larry car.
          Addition of automatic lid lifters or access to the charging
          holes without moving the larry car.
          Addition of human-guided mechanical cleaners for goosenecks.
          Addition of jumper pipes that move with the larry car on
          batteries with only one collecting main.
          Increase of clearance between the coal  bunkers and the
          battery to allow for volumetric hoopers.
          Complete repaving of the battery top and replacement of lid
          rings and lids.
          Replacement of gooseneck and pipes.
                                  5-3

-------
          Replacement of Corliss  valves with  Pullman dampers.

          Addition of second  collecting mains to  batteries with
          single collecting mains.

          Addition of mechanical  scrapers  and/or  decarbonization  air
          on the pusher ram.

          Installation of leveler bar smoke boot.

          Blueprinting of doors  to restore the original  specifications.

          Replacement of seals on doors as needed.

          Replacement of doors as needed.

          Replacement of refractory in oven walls where  necessary.

          Modification of self-seal ing Koppers doors by  adding stop
          blocks and replacing plunger springs with a  more temperature-
          resistant alloy.

          Replacement of cast iron jambs with ductile  iron jambs.

          Replacement of original door seals  with NiCuTi alloy seals.

          Modification of self-sealing Wilputte doors  by adding stop
          blocks, replacing plunger springs with  a more  temperature-
          resistant alloy,  and proving guide  blocks.

          Modification of hand-luted doors by enlarging  the  door plug
          and replacing the jambs with ones that  more  easily accommodate
          luting.

If a plant has incorporated and  then considers discontinuing any of
the above modifications, a modification could occur if the plant does
not install a device or practice that would produce an equal or lesser

amount of emissions than the  discontinued  modifications.
     Any discontinuance of a  control or control technique without

offsetting the increased emissions by implementing an  alternate control

technique would be considered a  modification.  For example,  if a plant

discontinued the use of a cokeside shed or door hoods  with a wet

electrostatic precipitator without offsetting increased  emissions by

decreasing door leaks, the change might be termed a modification.
                                5-4

-------
     5.2.2.1  Steam or Liquor Aspiration.  Increasing the aspiration
at the gooseneck to facilitate smokeless stage charging may inadvertantly
affect the operation of the by-product recovery plant and the underfiring
of the battery.3  The by-product recovery plant may not be able to
accommodate the additional load of gas or smoke, and additional oxygen
may change the nature of the coal tar chemicals to a type that the
recovery plant was not designed to process.   The oven underfiring
could be affected because the burners may not operate efficiently with
the lower heating value gas that results from dilution by too much
aspirated air.  Oven temperatures will be lower and green coke will be
produced.  An optimum aspiration range exists for individual plants
above and below which operation causes either charging problems or the
above problems, which in turn could increase emissions.4  Operation
outside this optimum range might be considered a modification.
5.3  REFERENCES
1.   U.S. Environmental Protection Agency.  Code of Federal Regulations.
     Title 40, Chapter I.  Subchapter A, Section 61.02.
2.   Hogan, W. T., and F. T. Koebkle.   Analysis of the U.S. Metallurgical
     Coke Industry.  Industrial Economics Research Institute, Fordham
     University.  (Prepared for U.S. Department of Commerce.)  EDA
     Project 99-26-09886-10.  October 1979.
3.   Baldwin, V. H., and D. W. Coy.  Study to Develop Retrofit Information
     and Other Data for Use in Setting Standards for Coke Oven Emissions.
     Research Triangle Institute.  Research Triangle Park, N.C.
     (Prepared for IERL, EPA.)  EPA Contract No. 68-02-2612, Task 39.
     March 1979.
4.   Barnes, T. M., H.  W. Lownie, Jr., and J.  Varga, Jr.  Control of
     Coke Oven Emissions.  Ironmaking and Steelmaking (Quarterly).
     3:157-186.  1975.
                                5-5

-------
            6.   MODEL BATTERIES AND REGULATORY ALTERNATIVES

     This chapter defines model batteries and alternative ways that
EPA can regulate fugitive emissions from wet-coal  charging, doors,
lids, and offtakes of coke ovens.   Regulatory alternatives are defined
for these various courses of action and are characterized in terms of
visible emission levels.
6.1  MODEL BATTERIES
     Model batteries are parametric descriptions of both the types of
batteries that exist and those which EPA judges may be constructed,
modified, or reconstructed.   Three model batteries which are repre-
sentative of major types of coke ovens are defined for the purpose of
estimating the emissions from coke ovens and the cost of control
technology.   These model  batteries are presented in Table 6-1.  The
battery type and dimensions are important in describing the emissions
and the emission control  systems.   For example, a battery with a
single collecting main would require installation of either a jumper
pipe or a second collecting main to obtain adequate oven aspiration
for wet-coal charging.  The oven height is important because the
pressure at the bottom of the doors on the 6-meter ovens is expected
to be greater than that on the smaller oven doors; therefore, door
leak control may be more difficult on 6-meter ovens.   Model Battery 3
not only represents the most recently constructed existing batteries,
it also represents the type most likely to be constructed in the
future (new sources).  The environmental, economic, and energy impact
analyses of each regulatory alternative are conducted by applying the
regulatory alternatives to the model batteries (Chapters 7 and 8).
                                  6-1

-------
                          TABLE 6-1.   MODEL BATTERIES

Type of coke
Cycle time (hr)
Ovens per battery
Door height (m)
Oven width (m)

1
Foundry
30
23-80 (36)
3-5 (3.5)
0.44
Model battery3
2
Metallurgical
18
15-118 (62)
3-5 (4.0)
0.44

3
Metallurgical
18
56-87 (71)
(6.0)
0.47
Battery dimensions:
  Length  (m)
  Width (m)
  Height5 (m)
Coal (Mg/oven)
Coalb (103 Mg/yr)
Cokeb (103 Mg/yr)
25.5-88.8 (40)
12.7
8.1-10.0 (8.6)
14.5
91-340 (152)
68-240 (106)
16.7-134.5 (70.7)
12.7
8.1-10.9 (9.1)
16.3
100-910 (492)
73-640 (344)
76.6-119.2 (97.3)
14.6
(11.1)
31.8
910-1,300 (1,100)
640-910 (770)
 Model Batteries 1, 2, and 3 all represent existing batteries; in addition,
 Battery 3 represents new batteries.   For charging, Battery 1 has a single
 collecting main, and Batteries 2 and 3 have double collecting mains.
""Numbers in parentheses are averages.
                                     6-2

-------
6.2  DEVELOPMENT OF REGULATORY ALTERNATIVES
6.2.1  Regulatory Alternatives
     Regulatory alternatives are alternate courses of action that EPA
could take to regulate emission sources.   The regulatory alternatives
for limiting emissions from wet-coal charging, doors, lids, and offtakes
of coke ovens were formulated on the basis of the demonstrated perform-
ance of the control systems presented in Chapter 4.   The stringency of
these regulatory alternatives ranges from the control levels required
by existing regulations to the best control levels observed at new or
rebuilt batteries.  A summary of the regulatory alternatives is pre-
sented in Table 6-2, and details are provided in the following sections.
     6.2.1.1  Regulatory Alternative I.   Regulatory Alternative I
permits the coke oven batteries to operate without National Emission
Standards for Hazardous Air Pollutants (NESHAP).  Under this regulatory
option, the coke oven batteries would be regulated by State and local
regulations, OSHA coke oven regulations, and consent decrees negotiated
on a pi ant-by-plant basis.
     The visible emission limits associated with this regulatory
alternative correspond to the baseline visible emission limits that
were presented in Tables 3-15 through 3-18 (Chapter 3).
     Currently, the least stringent emission limit for charging is
32 seconds per charge.  The most common baseline regulation is an
average of 19 seconds per charge and is in effect at 53 batteries
(40 percent of total).  The most stringent limit is 11 seconds per
charge and is currently in effect at 12 batteries.   The least stringent
limit for door leaks is 16 PLD, which is in effect for two tall batteries
in Ohio.   Approximately 97 batteries (72 percent of total) have a
limit of about 10 PLD.  Exclusions of 2 to 4 leaking doors are granted
to 70 of these batteries.   The most stringent baseline limits for
doors are 4 to 5 PLD and are required at 7 new or rebuilt batteries.
The baseline for lid leaks ranges from 1 to 5 PLL with the majority of
the batteries (63 percent) controlled at levels of 3 PLL or less.   The
limits for offtakes range from 4 to 10 PLO with 84 batteries at 10 PLO
and 50 batteries at 5 PLO or less.
                                 6-3

-------
           TABLE 6-2.   SUMMARY OF REGULATORY ALTERNATIVES
Emission Regulatory
point alternative Technology
Charging
Doors
Lids
Offtakes
I
II
III
I
II
III
I
II
III
I
II
III
Stage charging
Optimized stage charging
Rebuild batteries
Leak control program
Modified seals, inspect,
clean, repair
Rebuild batteries
Leak control program
Luting manpower
Rebuild batteries
Leak control program
Luting manpower, repair
slip joints
Rebuild batteries
Emission limit3
11 to 32 s/chgrge
16 s/charge
11 s/charge
4 to 16 PLD
10 PLD
5 PLD
1 to 5 PLL
3 PLL
1 PLL
4 to 10 PLO
6 PLO
4 PLO
 PLD =  Percent  leaking  doors.
 PLL =  Percent  leaking  lids.
 PLO =  Percent  leaking  offtakes.

5Log average  of 10  charges.
                                 6-4

-------
     The OSHA regulations1 require implementation of stage charging
with modified larry cars, adequate aspiration, and written operating
procedures.  In addition, the cost analysis of the OSHA regulation
included estimates for replacing or resetting lid rings, repairing the
battery top, realigning tracks, adding steam lines, nozzles, jumper
pipes, leveler bar seals, and decarbonizing equipment, and straightening
and cleaning standpipes.   For door leaks, OSHA requires a corrective
action program of inspection, cleaning, and repair or replacement of
damaged doors and seals.   Door repair facilities, an adequate number
of spare doors, and chuck door gaskets are also required.   Topside
leak regulation requirements include regular inspection and maintenance,
luting repairs, and prevention of miscellaneous topside leaks.1
     6.2.1.2  Regulatory Alternative II.   Regulatory Alternative II
represents application of the control technology discussed in Chapter 4.
The control technology for wet-coal charging is optimized stage charging,
which includes an engineering study, optimization of the operating
procedure, extensive worker training, visible emission observers, and
minor equipment changes.   This technology has been implemented by CF&I
and U.S.  Steel, Clairton; the results are listed in Table 4-4 (Chapter 4).
At these plants, the battery with the highest log average of charging
emissions (Battery 11 at Clairton) would meet a limit of a log average
of 16 seconds per charge at the 95-percent confidence level when
averaged over 10 charges.  In addition, no more than 1 charge in 10
can exceed 45 seconds in duration.  An emission limit based on the
95-percent confidence level from the worst case means that all of the
batteries in the data base could meet the limit at least 95 percent of
the time.   The limit also means that there is only a 5-percent probabil-
ity of citing a battery in violation when it is actually controlling
emissions as effectively as the batteries in the data base.
     Data in Table 4-4 also show that other battery types can meet a
log average of 16 seconds per charge.  These 25 batteries have the
following characteristics:   3- to 6-meter ovens; single and double
mains; old, new, and rebuilt status; three or four charging holes; a
variety of oven manufacturers; different geographical locations;
                                  6-5

-------
gravity or mechanical  feed;  manual  or automatic lid lifters;  inspections
during different times of the year;  different coking coal  sources;  and
a range of coal sizes.
     A comparison of log and arithmetic averages is necessary because
existing regulations use arithmetic averages.   A precise arithmetic
average cannot be determined from a log average; however,  the log
average is always less than or equal to the arithmetic average.
Therefore, a log average of 16 seconds indicates an arithmetic average
of at least 16 seconds.   The charging data base shows that some data
sets have arithmetic averages that are 5 or 6 seconds higher than the
log average.  Analysis of charging data indicates that the current
Pennsylvania standard (75 seconds for four charges) is probably equal
to or more stringent than a log average limit of 16 seconds per charge.
It is estimated that 80 batteries (60 percent of the total) have
current regulations that are as stringent as the log average limit,
and 54 batteries have existing regulations that are less stringent.
     Regulatory Alternative II for door leaks is based on implementation
of the basic door leak control program required by OSHA and installation
of temperature-resistent door seals that are more durable than those
used currently on most batteries.  This technology has been demonstrated
by CF&I and U.S. Steel, Clairton.  The discussion in Chapter 4 showed
that batteries at these plants could achieve average performance
levels of 7 PLD or less with a 95-percent confidence level of 10 PLO
when averaged over three runs.  The 20 batteries in the door leak data
base also include a rebuilt battery (Battery 9 at U.S. Steel, Fairfield)
and two 6-meter batteries (Battery 2 at U.S. Steel, Fairfield, and
Battery 1 at National  Steel, Weirton).
     A total of 30 batteries have existing limits greater than 10 PLD,
20 batteries have a limit of 10 PLD excluding 4 doors, and 50 batteries
have a limit of 10 PLD excluding 2 doors.  A total of 34 batteries
have current limits of 10 PLD or less.  This alternative would have
the most significant  impact on the 30 batteries with limits greater
than 10 PLD.  The impacts on batteries with 10 PLD limits and 2 to 4
doors excluded would  be minor compared to the impacts for batteries
with higher limits.
                                  6-6

-------
     Regulatory Alternative II for topside leaks represents sealing
with luting material, additional topside labor, and modification or
replacement of offtakes on some batteries.  Table 4-15 (Chapter 4)
shows 22 batteries that average 1.8 PLL or less with a 95-percent
confidence level of 3 PLL or less (averaged over three runs).   Table
4-16 (Chapter 4) lists 15 batteries that averaged 3.5 PLO or less with
a 95-percent confidence level of 6 PLO or less (averaged over three
runs).   These topside leak limits require conscientious luting of lids
and standpipe caps after charging and reluting whenever topside leaks
are observed.
     Approximately 84 batteries (63 percent of total) have existing
limits of 3 PLL or less.  Consequently, this alternative would affect
only 50 batteries with limits of 4 to 5 PLL.  A total of 84 batteries
have limits of 10 PLO and 50 batteries have limits of 5 PLO or less.
Regulatory Alternative II for offtake leaks would affect only those
batteries with a current limit of 10 PLO.
     6.2.1.3  Regulatory Alternative III.  Regulatory Alternative III
was developed as a more stringent alternative than Regulatory Alterna-
tive II.  This alternative is based on Lowest Achievable Emission Rate
(LAER) limits that have been applied to new and rebuilt batteries.
This alternative could require the extensive rebuilding of the industry's
existing batteries.   In addition, the limits are difficult for some
rebuilt batteries to meet and no technology, short of continual rebuild-
ing, could be identified that would allow new batteries to continue to
meet the limits as they age.
     Regulatory Alternative III would impose a limit of 11 seconds per
charge for 10 consecutive charges.  Currently 12 new or rebuilt batteries
have existing regulations as stringent as 11 seconds per charge.   For
door leaks, this alternative would impose a limit of 5 PLD.  Currently,
only 7 new or rebuilt batteries have existing limits of 5 PLD or less.
The limit for lid leaks would be 1 PLL, which is currently in effect
at 6 batteries.   Offtake leaks would be controlled to 4 PLO, which is
currently required at 13 batteries.
                                  6-7

-------
6.2.2  Alternative Control  Technologies
     Some technologies that are capable of controlling charging and
door leak emissions were not chosen as regulatory alternatives.   The
alternative control technologies for wet-coal  charging included sequen-
tial charging and larry-car-mounted scrubber systems.   These technol-
ogies were not chosen as bases for regulatory alternatives because
they were judged to be less capable than stage charging and optimized
stage charging for controlling emissions.   None of the regulatory
alternatives precludes use of an alternate technology, provided that
the coke oven batteries attain the visible emission limits specified
in the alternatives.
     The door leak control  technologies, which were discussed in
Chapter 4, include Ikio seals, the Battelle (EPA-AISI) seal, Nippon
gas seal, prechamber doors, and door hoods.  These technologies were
not considered as bases for regulatory alternatives because it has not
been demonstrated that they are superior to the control technology of
modified seals coupled with a door leak control program.   Cokeside
sheds with control devices, such as scrubbers or electrostatic precipi-
tators, also were discussed as possible controls for door leak emissions.
However, sheds with pollution control devices have been designed
primarily for control of particulate emissions from the pushing opera-
tion.  Most batteries do not use sheds to control pushing emissions,
and other types of pushing controls do not collect and remove door
leak emissions.  Cokeside sheds were not considered as a basis for a
regulatory alternative for door leaks because they are much more
expensive than modified seals coupled with the door leak control
program, and sheds have not been shown to be significantly more effective.
For example, the shed's control device may not remove gaseous pollutants
(such as benzene and sulfur oxides) that are contained within the
by-product recovery system by a door leak control program.  In addition,
the removal efficiency for the different compounds in the BSD fraction
may vary with the compound.  The shed's capture efficiency may also be
variable and will depend upon the design, construction, and operational
characteristics; the extent and location of emissions; and meteorological
conditions, such as wind speed and direction.
                                  6-8

-------
6.3  REFERENCES
1.    41 Federal Register 46/85, 46786.   October 22, 19/6.
                                 6-9

-------
                             7.   ENVIRONMENTAL IMPACT
     This chapter presents the environmental impacts that will result from
additional control of coke oven emissions for the Regulatory Alternatives
described in Chapter 6.   The primary impact directly attributable to
emission control is a reduction in atmospheric emissions of benzene soluble
organics (BSO) from coke ovens.  The estimated emission rates generated in
this chapter are used in the Human Exposure Model (HEM) to predict ambient
BSO concentrations around coke plants.  The HEM couples the predicted
concentrations with the population exposed to the BSO emissions and the
unit risk factor for BSO to predict annual incidence and maximum lifetime
risk.   The dispersion modeling and health impact analysis are discussed in
Appendix E.
     Other impacts, such as water pollution, solid waste disposal, and
energy consumption, were also examined.   The additional control of emissions
is accomplished by preventing leaks and containing potential emissions
within the by-product collection system.   Consequently, no significant
impacts on water pollution, solid waste disposal, or energy usage are
expected for additional  control of coke oven emissions from charging or
leaking doors, lids, and offtakes.
7.1  EMISSION RATE ESTIMATES
     The emission estimates for batteries at their current level of control
are based on the current visible emission limits in effect from existing
State regulations and consent decrees.   The existing emission limits for
each source at each battery were summarized in Chapter 3.  For a battery to
meet the limit a high percentage of the time, the battery's average perform-
ance must be less than the limit.
                                 7-1

-------
     The charging data base was examined to develop an empirical
correlation between the charging emission limits and an overall
logarithmic average performance.  The data base included 528 observa-
tions at 15 batteries.  The results are listed in Table 7-1 as the
estimated log performance that would be required to meet the emission
limit approximately 95 percent of the time.  For example, a log average
performance of about 17 seconds per charge would be required to meet
arithmetic limits of 32 seconds per charge or 25 seconds per charge
with the highest observation in 20 excluded.   The table indicates that
requiring a log average of 16 seconds per charge maximum is roughly
equivalent to a limit of 20 seconds per charge as an arithmetic average
because both have an estimated average performance of 10 seconds per
charge.
     The emission rates in grams of BSD per charge are based on the
estimates given in Chapter 3 and the estimated average performance.
The minimum emission estimate from Chapter 3 is 0.006 g BSO/s and
represents the unusually well-controlled case with only small puffs of
emissions from a single opening for every charge.  Although theoretically
achievable, this level of emissions probably underestimates actual
charging emissions significantly; however, this approach should provide
a reasonable lower bound for the charging emission estimate.  The
grams of BSO per charge (minimum) is estimated by multiplying the
average performance by 0.006 g/s (e.g., 17 s/charge x 0.006 g/s =
0.102 g/charge).
     The upper bound estimate is based on an assumed exponential
reduction in emissions as visible emissions decrease (exponent = 2).
The emission estimates are given in Table 3-6.  These estimates from
the exponential model should provide a reasonable upper bound for
well-controlled emissions.  However, for relatively dense emissions or
emissions that are long in duration, the linear model discussed in
Chapter 3 may be more appropriate and emissions would be higher than
those estimated from the exponential model.  Consequently, the maximum
estimate for charging does not  represent a worst-case estimate and  is
more typical of a consistently  controlled battery that does not
                                 7-2

-------
     TABLE 7-1.  CHARGING BASELINE LIMITS AND ESTIMATED EMISSION RATES
Charging emission
limit (s/charge)
Estimated average
    (s/charge)
                                                   BSO per charge (g)
Minimum'
Maximum
32
30 (-l/20)c
25 (-1/20)
25
24
20
19
12
11 .
16 (log)d
17
24
17
13
12
10
9.3
5.6
5.0
10
0.102
0.144
0.102
0.078
0.072
0.060
0.056
0.034
0.030
0.060
24
48
24
14
12
8.3
7.2
2.6
2.1
8.3
 Minimum = (average s/charge) * 0.006 g/s.
DMaximum = (average s/charge/300)2 x 7,500.

"(-1/20) = Excludes highest observation in 20.

 Logarithmic average of 16.  All others are arithmetic averages.
                                  7-3

-------
experience the dense emissions often seen with charges  that exceed 25
seconds of emissions.
     The emission estimates are given in grams of BSD per charge.   To
estimate annual emissions,  the grams/charge is multiplied by the
number of charges per year.
     The baseline regulations for percent leaking doors (PLO) are
summarized in Table 7-2.   The average performance was determined from
the statistical analysis  of the Poisson distribution and represents
the average required to meet the limit roughly 95 percent of the time.
The leak rates are based  on the exponential models described in
Chapter 3 and the results listed in Table 3-12.   For comparison,
consider that a simulated plume that was 1 to 2 meters  in length
yielded a measured emission rate of 0.021 kg/hr.   Door leaks may have
multiple plumes from a single door, and the plume may exceed 1 to 2
meters in length.  Because the minimum and maximum estimates include
the measured plume rate of 0.021 kg/hr, the estimates in Table 7-2 may
be slightly low.  Another factor that indicates the estimates may be
low is that the data from measurement of captured door leaks on poorly
controlled doors ranged from 0.1 to 0.7 kg/hr.  The lower estimates in
Table 7-2 result from the exponential door leak model that predicts
much lower emission rates because of smaller gap sizes.  For doors
that are leaking heavily, such as doors under a cokeside shed where
the measured data were collected, an estimated rate of 0.1 to 0.7
kg/hr per leak may be more appropriate.  For doors that are consistently
controlled around 10 PLD without any heavy plumes, the leak rates in
Table 7-2 should provide reasonable bounds on the emission estimates.
     The emission limits  and estimated average performance for percent
leaking lids (PLL) and percent leaking offtakes (PLO) are summarized
in Table 7-3.  The minimum and maximum emission rates per leak were
derived from the capture and measurement of simulated leaks on the
topside of a battery.   The minimum estimate of 0.0033 kg/hr was the
average rate measured for a plume ~0.3 meters in length and should be
representative of the small wisps seen from lid leaks.   The maximum
rate of 0.021 kg/hr was for a plume that was 1 to 2 meters in length
                                 7-4

-------
TABLE 7-2.
DOOR LEAK BASELINE
LIMITS AND ESTIMATED
EMISSION RATES
BSO/hrj>er leak (kg)
Door limit
(PLD)
16
15
12 (-2)C
12 .
10 (-4)d
10 (-2)
10
5
4
Estimated average
(PLD)
12.1
11.3
10.2
8.8
10.0
8.6
7.2
3.3
2.6

Minimum
0.0137
0.0123
0.0106
0.0085
0.0103
0.0082
0.0063
0.0019
0.0014
K
Maximum
0.153
0.138
0.118
0.095
0.115
0.0916
0.0702
0.0218
0.0152
aMinimum = (average PLD/70)1'5
 Maximum = 11.2 x minimum.
°(-2) = Excludes 2 doors.
 (-4) = Excludes 4 doors.
0.19
7-5

-------
           TABLE 7-3.   LID AND OFFTAKE LEAK BASELINE LIMITS AND
                         ESTIMATED EMISSION RATES
Lid limit
(PLL)
5
4
3
2
1
Estimated average
(PLL)
3.5
2.6
1.9
1.1
0.6
Offtake limit
(PLO)
10
6
5
4

Estimated average
(PLO)
6.5
3.5
2.9
2.2

aThe minimum estimated emission rate is 0.0033 kg/hr per leak.
 The maximum estimated emission rate is 0.021  kg/hr per leak.
                                 7-6

-------
and may be more representative of a large leak from an offtake.  The
rates assume one leak per source; consequently, these emission rates
would underestimate the emissions from multiple leaks on a single lid
or offtake.
     The average performance levels and emission rates in Tables 7-1,
7-2, and 7-3 are used to estimate the current (baseline) emissions.
The emission rates for Regulatory Alternative II are estimated from
the performance levels achieved by the improved control technology
implemented at CF&I and U.S. Steel's Clairton Works.  The description
of control technology and the associated performance levels are provided
in Chapter 4.  The upper limits, average performance, and estimated
emission rates are summarized in Table 7-4 for each source.  Emission
estimates for Regulatory Alternative II are based on the average
performance and mass emission rates listed in Table 7-4.   The upper
limits, average performance, and estimated emission rates for Regulatory
Alternative III are summarized in Table 7-5.
7.2  EMISSION ESTIMATES FOR MODEL BATTERIES
     Model batteries were defined in Chapter 6 and will be used in
this section to estimate emissions for the types of batteries commonly
found in the industry.  The characteristics of the model  batteries are
summarized in Table 7-6.   Model Battery 1 is a small battery that
produces foundry coke on a 30-hour cycle.   This model has a single
collecting main (one offtake per oven) and three lids per oven.  Model
Batteries 2 and 3 produce metallurgical coke on an 18-hour cycle.
Both have double collecting mains and four lids per oven.  Model
Battery 2 is 4 meters in height and Model  Battery 3 is 6 meters in
height, which is characteristic of new batteries constructed in the
past 10 to 15 years.
     The current baseline limits shown in Table 7-6 were chosen to
represent the least stringent baseline limits currently in effect.   In
this sense the model  batteries represent those batteries that will  be
most affected by requiring improved emission control from a national
standard.   The average performance levels shown were obtained from
Tables 7-1, 7-2, and 7-3  for the corresponding visible emission limits.
                                 7-7

-------
        TABLE 7-4.   DEMONSTRATED PERFORMANCE LEVELS3 AND ESTIMATED
               EMISSION RATES FOR REGULATORY ALTERNATIVE II

Upper limit          Average performance            BSD emission rate
16 s/charge 10 s/charge
10 PLD 7.2 PLD

3 PLL 1.9 PLL

6 PLO 3.5 PLO

0.06-8.3
0.0063-0.
leak
0.0033-0.
leak
0.0033-0.
leak
g/charge
0702 kg/hr per

021 kg/hr per

021 kg/hr per

 Performance levels demonstrated by CF&I and U.S. Steel's Clairton Works.
 Log average of 10 charges.
                                 7-8

-------
            TABLE 7-5.   EMISSION LIMITS AND ESTIMATED EMISSION
                   RATES FOR REGULATORY ALTERNATIVE III
Upper limit
Average performance
  BSO emission rate
 11 s/charge

  5 PLD


  1 PLL


  4 PLO
    5.0 s/charge

    3.3 PLD


    0.6 PLL


    2.2 PLO
0.03-2.1 g/charge

0.0019-0.0218 kg/hr per
 leak

0.0033-0.021 kg/hr per
 leak

0.0033-0.021 kg/hr per
 leak
                                7-9

-------
                 TABLE 7-6.   MODEL BATTERY CHARACTERISTICS
Model batteries
Item
Capacity (1,000 Mg/yr)
No. ovens
Cycle time (hr)
Hours/yr
Charges/yr
No. doors
No. lids
No. offtakes
Current limits
s/charge
PLD
PLL
PLO
Average performance1
s/charge
PLD
PLL
PLO
No. of leaks1-'
Doors
Lids
Offtakes
la
106
36
30
7,880
9,460
72^
108^
36T
25(-l/20)h
15
5
10

17
11.3
3.5
6.5
8.1
3.8
2.3
2
344
62
18
7,880
27,100
124
248e
1249
25(-l/20)
15
5
10

17
11.3
3.5
6.5
14
8.7
8.1
3
770
71
18
7,880
31,100
142
284*
1429
25(-l/20)
16
5
10

17
12.1
3.5
6.5
17
9.9
9.2
aModel  Battery 1 is a foundry coke battery while Models 2 and 3 are metal-
 lurgical coke batteries.
 Assumes in operation 90 percent of the time (365 x 24 x 0.9).
cCharges/yr = No.  ovens 4- cycle time x 7,880.
 Three lids per oven.
eFour lids per oven.
 Single main.
^Double main.
 (-1/20) = Excludes highest observation in 20.
Estimated average performance required to meet the emission limit a high
 percentage of the time.
••'Average leaking times number on battery (e.g., 11.3 percent of 72 doors =
 8.1 leaking doors).

                                  7-10

-------
     The number of charges per year were estimated from the number of
ovens, cycle time, and average annual hours of operation.   The battery
is assumed to operate 90 percent of the time and the balance is lost
time or down time.  This is equivalent to assuming that at any given
time 10 percent of the ovens are out of service for repairs.  The
average number of leaks was estimated by multiplying the average
percent leaking by the total number on the battery.
     The emissions are estimated for the model batteries for each
regulatory alternative in Tables 7-7 and 7-8.  For charging, the
estimates for Model Battery 1 are calculated from the number of charges
times the g BSD/charge (from Table 7-1):
     9,460 charges/yr x 0.102 g/charge x 10~3 kg/g =1.0 kg/yr
     9,460 charges/yr x 24 g/charge x lo"3 kg/g = 230 kg/yr.
For door leaks, the estimates are based on the number of leaks and the
estimated leak rate.  For Model Battery 1 with 8.1 door leaks, the
estimated leak rate is 0.0123 to 0.138 kg/hr.  The calculation for
Model Battery 1 is:
     8.1 leaks x 0.0123 kg/hr x 7,880 hr/yr = 790 kg/yr
     8.1 leaks x 0.138 kg/hr x 7,880 hr/yr = 8,800 kg/yr.
The calculations for emissions from lid and offtake leaks are based on
the number of leaks and an emission rate of 0.0033 to 0.021 kg/hr per
leak.  A sample calculation for lid leaks on Model Battery I is:
     3.8 leaks x 0.0033 kg/hr x 7,880 hr/yr = 100 kg/hr
     3.8 leaks x 0.021 kg/hr x 7,880 hr/yr = 630 kg/yr.
The Regulatory Alternative II emission estimates are based upon control
performance improved to the levels summarized in Table 7-4 with average
performance levels of 10 seconds per charge, 7.2 PLD, 1.9 PLL, and 3.5
PLO.   These average performance levels correspond to the levels achieved
by the control technology implemented at CF&I and U.S.  Steel's Clairton
Works.  Emissions were estimated in the same manner described previously
                                 7-11

-------



















1_1_1
:E
h-
Qr
O
LI-

CO
LU
1 — I
0£
LU 1— 1
1— 1— 1


I-H
_l h-
LU -
a; ce:
o o
i i i
U. |
<
CO —1
LU Z3
1— O

 s^^
0 0
0 0

-0 0 -
S~>: CD O O OO
O rH CO LO CM
in i - - i
r-. O rH rH O
1 O 1 1 0
CM r-^ o o CM
. -co «t -
CO rH CM CM CM
s~/v"-" "~-'v-/
o CD o o o co CM
CO O CO r- 0 rH CO
CO CO CD 00 O O O
*t *1 • *
CD CM 0 O
rH rH





<"> /—v
O 0
0 0

-o o -
-— NUO o o oo
O rH ^f CO rH
in i - - i
CO O rH rH O
1 O 1 1 O
00 «t 0 0 00
. -CO T-H -
CM rH CM CSJ rH
**~S **~S ^-S ***S ^^

OOOOO CD 00
CO O CM CD O CM in
co CM co r»» o o o
f. +> . ,
00 0 0 O
rH


/— N
»^s O
O O
^>O '-N 0
O CO O '•> -
CO -CO O O
CM 00 CD 00 rH
1 1 1 co 1
O O O 1 O
• cr> o o in
rH r~- rH CD CD


un o o o o CM
rH O UD CM O in O
rH 00 CO CM in O rH
— — • •
«d- in o o

re
S- I-H
>1
\ CU
Di >
v/ , . *
jt, «»-a j
v-' -P
re<)
C C f
O i- !
•i- cu :
10 -P ">
in r—
•r- -
0) CO in a> -^ >,
c ai -^ o s-
o -r- ^: ouo
COM re i — u -P
•r- i. s.  +j re c re-r
r- reo-a4--p coo r—
cu.co'r-<4-o s:-p 3
to OQ_IOI— vv.\ en-
re oxi cu
OO -^ r- Q£


^^S /-"N
O O
/ — - O s^s~ > O
0 CO 0 0 UD
CO -CD CM
CM m oo oo r^
1 1 1 1 1
CD O O O CD
• o «3- co r^
rH LO rH rH |-«
v_/v^/^^s_xs^
ooooo in
COOCMOOO mo r^.o«^-inm
rHrHin«^-CM OrH CDf^-«S-<-CO
0 O
CO ^- . .
0 0








s~\ /— >
o o
/•~N O S*^ *^*s ^
O CD O O IO
CO -00 CM -
CM ^- r^ r-- CD
1 1 1 1 1
CD O 0 O 0
• ^- CM rH r^
rH ^ rH rH VD
\^.S ^S *^*S ^^/ **^*

ooooo I-HCM r~»( — minco
rHOinCMO rHCM CDCO«^-<-CD
rH l>- •* ^- r-^ 00
*- *• • »
CM co o o




/••S <^N
O O
O 0
/ — > CD .*->,,—••, in
CD -O O -
1^ CM ^ rH CO
1 1 CO CM 1
co o i i o
• CD co co in
o CM in co co
*^™^ VSl— / K^^/ \*^ l^™^

csoooo CDOO inr-^^-in'*
<*OOCMO rHCO COCD'^-^-CD
CD CM rH O O O
— — • .
rH CM O O








-v C
0
>> •'~
•v -P
Jl U
*: =s
^ cu ~a
O m o J^ ai O in
y>c a) -^ o S- c cu
= •!- ^ O U -r- -i^
DCOIO re> — u -potn rei —
t-t-m^jnj c c s-s-in-pre
ji reoT3**--p coo cu reo"Of--P
n -co-r-'i-o S:-P u ^0-1-14-0
- OQ_IOI— \\ i- oc3_ioi—
= D) -O. CU
1) ^ ^- OL.




















.
co
(U
in
cu
JZ
-p
c
cu
i.
re
O-

c
•r- O
_i
in cu
cu
•P CD
re
E T3
•r- C
-P re
in
i \
f~) y^
-a co
CU rH
^
0 «-
r- O
^~
O t/>
ll , 1
H— +^
CU E
-P -r-
ro •—
E
•r- t-
•P 0
to f-
CU
in
cu c
O O
C •!-
re in
i- to
T3 ••-
•r- E
s: LU
re J2
7-12

-------














oo
LU
1 	
— 1
OO
oo
1— 1
•s.
LUro
oo
I— 1 LU
I—I 1— (
1— 1 Q£
LU
LU 1—
> I—
i— ' «=C
r- 03
^C
Ty '
Q£ LU
LU O
h- O
— 1 2!
- re
fy |—
o
I— OL
< 0
	 1 1 1
"*"*>
C3
LU
c*:


*
00
i-L

LU
	 1
CD
1
a
O)
c
0
*r—
t/)
- •* oo CM
•^^•^s^s^.
COOOOO COID rHtDOOlDCM
oo-=j-vooo I-HCM cncnooiDcn
«3- rH CO O O O
0 O
0 CD












S^\
0
'""^ O
P** /^^ t"^ **"** UD
LO O O O -
1 O LO LO rH
rH r»- CM <^- |
00 1 1 1 0
. rH  rH CT>
CD ID 00 r*. rH
N«*S *^*S ^*-S ***** *****

O1OOOO UDCM i— 1 LO CO U> rH
CMOO'!l-l£>O CMLO CTlOlOOlDCTl
OO rH CM CT» O O
O O

O CZ)






/****
CD *""1N /**^N /**N ^~S
CM O O O O
1 rH rH 00 rH
00 •* rH rH r»-
CM 1 1 1 I
. ID r*. rH O
O ro rH CM 00

CD CD ^^ UD CO r^ *«i" rH LO CM LO OO
rH CM tf*i p*^ CTi CO r^- CTi CD CO ID CT>
CM CO O O
0 O

O O









C
o
•£
u
3
•o
O) in cu cu D5 w
c ai ... cu s- c cu
•I- ^ O ^ •!- -ilS
O) i/) ro r— cj o -P O^ in ro r—
S-S-in-pro O C £_S-(/>-PrO
fO O "C3 ^~~ ^^ CJ1 Qi) ^ O "^J *^* rt ^
C™ Q *r~* *+— O *S»" | 1 (J ^ Q (|_ tXj^ Q
C_>Q— IOr- ^v^ S- O Q — 1 O I—
C^ Q Q)
... r— Q_


















,
in
CU
in
cu
• f
O -P
— J C
Q_ CU
s_
^t ro
a.
T3
c c
ro -r-

r> (/)
—1 CO

Q- ro
g
rH *r-
^j
„ (/)
O CU

Q- E

LO E

» X
CU ro
0) E
S-
rO "O
J=. C
O ro

i- E
CU 3
a. E
•r-
in c
•a T-
C E
O

CU ^1
in
•a
rH 
-------
for the baseline case by substituting the reduced average performance
level, number of leaks, and the leak rate for those used in the baseline
estimate.
     The estimates for Regulatory Alternative III in Table 7-8 were
derived in a similar manner.   The average performance levels and
emission rates in Table 7-5 were used with the model battery
characteristics (from Table 7-6) to estimate emissions from the model
batteries.
7.3  ESTIMATES OF NATIONWIDE EMISSIONS
     Emission estimates were generated for each battery and are based
on the battery's specific characteristics and the visible emission
limit currently in effect.   The data for the number of ovens, doors,
lids, and offtakes are summarized in Table 7-9.  A cycle time of 30
hours was used for foundry coke producers and 18 hours was used for
metallurgical coke producers.   The estimated average performance for
each battery is given in Table 7-10 and is based on the estimated
performance required to meet the baseline limit a high percentage of
the time  (approximately 95 percent).
     For charging, the baseline emission estimate for each battery is
based on the number of charges per year (number of ovens/cycle time x
7,880 hr/yr) multiplied by the grams of BSO per charge.  An example
calculation is provided below for Battery A at Plant 1 in Tables 7-9
and 7-10:
     Charges/yr = 78/30 x 7,880 = 20,500
     Average =9.3 seconds per charge
     g BSD/charge = 0.056-7.2 (from Table 7-1)
     Midrange =3.6 grams per charge
     Midrange emissions = 20,500 x 3.6 x 10 3 = 74 kg/yr
     The emission estimates for door leaks are based on the estimated
number of doors leaking on each battery and the range of leak rates
given in Table 7-2.  For Battery A at Plant 1, an average of 17.6
doors leaks result from an average control level of 11.3 PLD for 156
                                 7-14

-------
TABLE 7-9. 8ATTERY CHARACTERISTICS
No.
1


2

3




4

5
6


7



8


9

10
11

12


13

14




15





16

Plant
AL Byproducts, Tarrant, AL


Empire Coke, Holt, AL

toppers, Woodward, AL




LTV Steel, Gadsden.AL

LTV Steel, Thomas.AL
Jim Walters, Birmingham, AL


U.S. Steel, Fairfield, AL



National Steel, Granite City, IL


Interlake, Chicago, IL

LTV Steel ,So.Chicago, IL
Bethlehem Steel, Bums Harbor, IN

Citizens Gas, Indianapolis, IN


IN Gas, Terre Haute, IN

Inland Steel, E. Chicago, IN




U.S. Steel, Gary, IN





LTV Steel, E.Chicago, IN

Battery
No.
A
5
6
1
2
1
2A
26
4
5
2
3
1
3
4
5
2
5
6
9
A
6
C
1
2
2
1
2
E
H
I
1
2
6
7
8
9
10
1
5
7
13
15
16
4
9
Status No. of Cycle No. of No. of No. of
Ovens Time Doors Lids Offtakes
(hr)
0
1
0
0
0
0
0
0
0
0
0
0
2
0
0
0
2
2
2
2
0
0
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
1
1
1
0
1
2
2
78
25
29
20
40
60
38
40
58
30
65
65
65
30
30
60
57
77
77
63
45
45
47
50
50
60
82
82
47
41
72
30
30
65
87
87
87
51
85
77
77
77
77
77
75
87
30
30
30
30
30
24
24
24
24
24
18
18
18
30
30
30
18
18
18
18
18
18
18
18
18
18
18
18
30
30
30
30
30
18
18
18
18
18
18
18
18
18
18
18
18
18
156
50
58
40
80
120
76
80
116
60
130
130
130
60
60
120
114
154
154
126
90
90
94
100
100
120
164
164
94
82
144
60
60
130
174
174
174
102
170
154
154
154
154
150
150
174
234
100
116
60
120
180
114
120
174
90
260
260
260
120
120
240
228
308
308
252
180
180
188
150
150
240
328
328
188
164
216
90
90
260
348
348
348
153
340
231
231
231
231
231
300
348
78
25
29
20
40
120
78
40
58
30
65
65
65
30
30
60
114
72
72
63
90
90
94
50
50
60
82
164
47
41
144
60
60
65
87
87
87
51
170
154
154
154
77
77
75
87
                                                   (continued)
             7-15

-------
TABLE 7-9. (continued)
No.
17

18








19



20

21


22
23


24
25


26
27
28





29






30

Plant:
Armco Inc., Ashland, KV

Bethlehem Steel .Sparrows Pt., MD








Rouge Steel, Dearbom,MI



National Steel, Detroit, MI

Carondolet, St. Louis, MO


Tonawanda, Buffalo, NY
Bethlehem Steel, Lackawama, NY


LTV Steel .Warren.GH
Armco Inc., Middletowi,CH


New Boston, Portsmouth, OH
toppers, Toledo, OH
LTV Steel, Cleveland.*





U.S. Steel, Lorain, OH






AL Byproducts, Keystone, PA

Battery
No.
3
4
1
2
3
4
5
6
11
12
A
A
Ax
8
Dx
4
5
1
2
3
1
7
8
9
4
1
2
4
1
C
1
2
3
4
6
7
D
G
H
I
J
K
L
3
4
Status No. of Cycle No. of No. of No. of
Ovens Time Doors Lids Offtakes
(hr)
0
0
1
1
2
2
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
2
2
2
2
2
2
2
2
2
76
70
63
60
63
63
63
63
65
65
80
45
10
55
27
78
85
40
18
35
60
76
76
73
85
57
57
76
70
57
51
51
51
51
63
63
59
59
59
59
59
59
59
55
55
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
30
30
30
30
18
18
18
18
18
18
18
30
30
18
18
18
18
18
18
18
18
18
18
18
18
18
30
30
152
140
126
120
126
126
126
126
130
130
160
90
20
110
54
156
170
80
36
70
120
152
152
146
170
114
114
152
140
114
102
102
102
102
126
126
118
118
118
118
118
118
118
110
110
228
280
189
180
189
189
189
189
260
260
320
135
30
165
81
234
340
160
72
140
180
228
228
292
255
228
228
228
210
171
153
153
153
153
189
189
177
177
177
177
177
177
177
220
220
76
140
63
63
63
63
63
63
130
65
160
90
20
110
54
156
85
40
18
35
60
76
76
73
170
114
114
152
70
57
51
51
51
51
63
63
118
59
59
59
59
59
59
55
55
                                         (continued)
     7-16

-------
TABLE  7-9. (continued)
No.
Plant
Battery
No.
Status No. of Cycle No. of No. of No. of
Ovens Time Doors Lids Offtakes
(hr)
31



32
33




34

35

36











37

38


39

40
41



42
43






Bethlehem Steel, Bethlehem, PA



LTV Steel, Aliquippa.PA
LTV Steel, Pittsburgh, PA




Koppers, Erie, PA

Shenango, Pittsburgh, PA

U.S. Steel, Clairton, PA











U.S. Steel, Fairless Hills, PA

Wheeling-Pitt, Honessen, PA


Chattanooga Coke, Chattanooga, TN

Lone Star Steel, Lone Star, TX
U.S. Steel, Provo, UT



Weirton Steel, Brown's Is.,WV
Wheeling-Pitt, E.Steubenville,WV






A
2
3
5
A1
P1
P2
P3N
P3S
P4
A
8
1
4
1
2
3
7
8
9
15
19
20
21
22
B
1
2
1A
18
2
1
2
C
1
2
3
4
1
1
2
3
8



0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
0
0
0
0
0
2
2
2
0
0
0
0
0
0
0
0
0
2
0
0
0
0
Status
0=online
1=hot idle
2=cold idle
80
102
102
80
106
59
59
59
59
59
23
35
56
35
64
64
64
64
64
64
61
87
87
86
87
75
82
82
37
37
19
24
20
70
63
63
63
63
87
47
47
51
79



18
18
18
18
18
18
18
18
18
18
30
30
30
30
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
30
30
18
18
18
18
18
18
18
18
18
18



160
204
204
160
212
118
118
118
118
118
46
70
112
70
128
128
128
128
128
128
122
174
174
172
174
150
164
164
74
74
38
48
40
140
126
126
126
126
174
94
94
102
158



320
510
510
320
318
177
177
177
177
177
92
140
168
105
256
256
256
256
256
256
244
348
348
344
348
300
246
246
111
148
76
72
60
280
189
189
189
189
348
188
188
204
316



160
102
102
80
106
59
59
59
59
59
23
35
112
70
128
128
128
128
128
128
122
174
174
172
174
150
164
164
37
74
38
24
20
140
126
126
126
126
174
47
47
51
158



3=under construction
  7-17

-------
TABLE 7-10. ESTIMATED AVERAGE VISIBLE EMISSIONS
No. Plant
1 AL Byproducts, Tairant, AL


2 Empire Coke, Holt, AL

3 tappers, Woodward, AL




4 LTV Steel, Gadsden.AL

5 LTV Steel, Thonas.AL
6 Jim Walters, Birmingham, AL


7 U.S. Steel, Fair-field, AL



8 National Steel, Granite City, IL


9 Interlake, Chicago, IL

10 LTV Steel ,So.Chicago,IL
11 Bethlehem Steel, Burns Harbor, IN

12 Citizens Gas, Indianapolis, IN


13 IN Gas, Terre Haute, IN

U Inland Steel, E. Chicago, IN




15 U.S. Steel, Gary, IN





16 LTV Steel, E.Chicago, IN

Battery
No.
A
5
6
1
2
1
2A
28
4
5
2
3
1
3
4
5
2
5
6
9
A
8
C
1
2
2
1
2
E
H
I
1
2
6
7
8
9
10
1
5
7
13
15
16
4
9
Status
0
1
0
0
0
0
0
0
0
0
0
0
2
0
0
0
2
2
2
2
0
0
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
1
1
1
0
1
2
2
Average Baseline Emissions Level
s/chg
9.3
9.3
9.3
9.3
9.3
9.3
9.3
9.3
9.3
9.3
9.3
9.3
9.3
9.3
9.3
9.3
5.0
9.3
9.3
5.0
5.0
5.0
5.0
17.0
17.0
5.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
PLO
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
7.2
11.3
11.3
3.3
7.2
7.2
7.2
7.2
7.2
7.2
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
PLL
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
1.1
3.5
3.5
0.6
3.5
3.5
3.5
3.5
3.5
3.5
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
PLO
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
2.9
6.5
6.5
2.2
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
                                                        (continued)
                             7-18

-------
TABLE 7-10. (continued)
No.
17

18








19



20

21


22
23


24
25


26
27
28





29






30

Plait
Armco Inc., Ashland, KY

Bethlehem Steel, Sparrows Pt., MO








Rouge Steel ,Oearbom,MI



National Steel, Detroit, HI

Carondolet, St. Louis, MO


Tonawanda, Buffalo, NY
Bethlehem Steel, Lackawanna, NY


LTV Steel ,Warren,OH
Armco Inc., Middletom,OH


New Boston, Portsmouth, OH
Koppers, Toledo, OH
LTV Steel, C1eveland,OH





U.S. Steel, Lorain, OH






AL 8yproducts,Keystone,PA

Battery
No.
3
4
1
2
3
4
5
6
11
12
A
A
Ax
8
Dx
4
5
1
2
3
1
7
8
9
4
1
2
4
1
C
1
2
3
4
6
7
D
G
H
I
J
K
L
3
4
Status
0
0
1
1
2
2
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
2
2
2
2
2
2
2
2
2
Average Baseline Emissions
s/chg
13.0
5.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
5.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
24.0
13.0
13.0
13.0
17.0
17.0
17.0
17.0
17.0
17.0
5.6
17.0
17.0
17.0
10.0
10.0
17.0
17.0
17.0
17.0
17.0
17.0
17.0
9.3
9.3
PLD
7.2
7.2
7.2
7.2
7.2
7.2
7.2
7.2
7.2
7.2
3.3
7.2
7.2
7.2
7.2
7.2
8.8
11.3
11.3
11.3
7.2
7.2
7.2
7.2
7.2
12.1
12.1
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
PLL
3.5
3.5
1.9
1.9
1.9
1.9
1.9
1.9
1.9
1.9
0.6
2.6
2.6
2.6
2.6
2.6
2.6
1.1
1.1
1.1
1.1
1.1
1.1
1.1
3.5
3.5
3.5
3.5
3.5
3.5
1.1
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
1.1
1.1
Level
PLO
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
2.2
3.0
3.0
3.0
3.0
3.0
3.0
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
2.9
6.5
6.5
6.5
2.9
2.9
6.5
6.5
6.5
6.5
6.5
6.5
6.5
2.9
2.9
                                 (continued)
     7-19

-------
                             TABLE 7-10.  (continued)
No.
31



32
33




34

35

36











37

38


39

40
41



42
43



Plant
Bethlehem Steel, Bethlehem, PA



LTV Steel, Aliquippa.PA
LTV Steel, Pittsburgh, PA




Koppers, Erie, PA

Shenango, Pittsburgh, PA

U.S. Steel, Clairton, PA











U.S. Steel, Fairless Hills, PA

Wheeling-Pitt, Monessen, PA


Chattanooga Coke, Chattanooga, TN

Lone Star Steel, Lone Star, TX
U.S. Steel, Prove, UT



Weirton Steel, Brown's Is.,WV
Wheeling-Pitt, E.Steubenville.WV



Battery
No.
A
2
3
5
A1
PI
P2
P3N
P3S
P4
A
6
1
4
1
2
3
7
8
9
15
19
20
21
22
B
1
2
1A
18
2
1
2
C
1
2
3
4
1
1
2
3
8
Status
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
I
1
1
1
0
0
0
0
0
2
2
2
0
0
0
0
0
0
0
0
0
2
0
0
0
0
Average Baseline Emissions
s/chg
9.3
9.3
9.3
9.3
9.3
9.3
9.3
9.3
9.3
9.3
9.3
9.3
5.0
9.3
9.3
9.3
9.3
9.3
9.3
9.3
5.0
9.3
5.0
9.3
9.3
5.0
9.3
9.3
9.3
9.3
9.3
9.3
9.3
5.6
9.3
9.3
9.3
9.3
13.0
13.0
13.0
13.0
13.0
PLD
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
8.6
3.3
8.6
8.6
8.6
8.6
8.6
8.6
8.6
3.3
8.6
3.3
8.6
8.6
3.3
8.6
8.6
8.6
8.6
8.6
7.2
7.2
2.6
7.2
7.2
7.2
7.2
8.6
8.6
8.6
8.6
8.6
PLL
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
0.6
1.1
1.1
1.1
1.1
1.1
1.1
1.1
0.6
1.1
0.6
1.1
1.1
0.6
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
Level
PLO
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.2
2.9
2.9
2.9
2.9
2.9
2.9
2.9
2.2
2.9
2.2
2.9
2.9
2.2
2.9
2.9
2.9
2.9
2.9
6.5
6.5
2.2
2.9
2.9
2.9
2.9
6.5
6.5
6.5
6.5
6.5
Status
0=online
1=hot idle
2=cold idle
3=under construction
s/chg= seconds per charge.
PLD= percent leaking doors.
PLL= percent leaking lids.
PLO percent leaking offtakes.
                               7-20

-------
total doors.  The emission rate for a level of 11.3 PLD was estimated
as 0.0123 to 0.138 kg/hr per leak (Table 7-2).   The midrange of the
leak rate is 0.075 kg/hr.  Annual emissions for Battery A are estimated
as:
     17.6 leaks x 0.075 kg/hr x 7,880 hr/yr = 10,400 kg/yr.
     The procedure for estimating emissions from lids and offtakes was
the same as the procedure for doors.  The emission rates for topside
leaks range from 0.0033 to 0.021 kg/hr per leak with a midrange estimate
of 0.012 kg/hr.  For Battery A, average levels of 3.5 PLL and 6.5 PLO
yield 8.2 lid leaks and 5.1 offtake leaks.  The midrange emission
estimates for lids and offtakes are then calculated:
     8.2 lid leaks x 0.012 kg/hr x 7,880 hr/yr = 780 kg/yr
     5.1 offtake leaks x 0.012 kg/hr x 7,880 hr/yr = 480 kg/yr.
     The results of the nationwide emission estimates for each battery
are summarized in Table 7-11 with the midrange values given.  Emission
estimates were also generated for emission limits of 16 seconds per
charge, 10 PLD, 3 PLL, and 6 PLO.  The estimates of reduced emissions
are based on an emission reduction at those batteries with current
visible emission limits that are less stringent than 16 seconds per
charge, 10 PLD, 3 PLL, and 6 PLO.  No reduction in emissions from the
baseline level was attributed to those batteries with current visible
emission limits more stringent than those listed above.
     The results of the battery-by-battery analysis are summarized in
Table 7-12.   Door leaks contribute about 82 percent of the baseline
emissions of BSO followed by lids (8 percent),  offtakes (7 percent),
and charging (3 percent).  The estimates for emission control improved
to at least 16 seconds per charge, 10 PLD, 3 PLL, and 6 PLO (Regulatory
Alternative II) on a nationwide basis are also given in Table 7-12.
The midrange estimate of emissions is reduced from 719 Mg/yr at the
baseline to 420 Mg/yr for improved control.  Most of the 42-percent
reduction in total  emissions results from the 44-percent reduction in
door leaks,  which are the major emission source.
                                7-21

-------
TABLE 7-11. BASELINE EMISSION ESTIMATES
No. Plant
1 AL Byproducts, Tarrant, AL


2 Empire Coke. Holt, AL

3 toppers, Woodward, AL.




4 LTV Steel, Gadsden.AL

5 LTV Steel, Thomas.AL
6 Jim Walters, Birmingham, AL


7 U.S. Steel, Fairfield, AL



8 National Steel, Granite City, IL


9 Interlake, Chicago, IL

10 LTV Steel .So.Chicago, IL
11 Bethlehem Steel, Bums Harbor, IN

12 Citizens Gas, Indianapolis, IN


13 IN Gas, Terre Haute, IN

14 Inland Steel, E. Chicago, IN




15 U.S. Steel, Gary, IN





16 LTV Steel, E.Chicago, IN

Battery
No.
A
5 '
6
1
2
1
2A
28
4
5
2
3
1
3
4
5
2
5
6
9
A
8
C
1
2
2
1
2
E
H
I
1
2
6
7
8
9
10
1
5
7
13
15
16
4
9
Status
0
1
0
0
0
0
0
0
0
0
0
0
2
0
0
0
2
2
2
2
0
0
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
1
1
1
0
1
2
2
Midrange Baseline Emissions (Hg/yr)
Charges
0.0744
0.0239
0.0277
0.0191
0.0382
0.0716
0.0453
0.0477
0.0692
0.0358
0.1034
0.1034
0.1034
0.0286
0.0286
0.0573
0.0264
0.1225
0.1225
0.0292
0.0208
0.0208
0.0218
0.2648
0.2648
0.0278
0.4343
0.4343
0.1494
0.1303
0.2288
0.0953
0.0953
0.3443
0.4608
0.4608
0.4608
0.2701
0.4502
0.4078
0.4078
0.4078
0.4078
0.4078
0.3972
0.4608
Doors
10.44
3.35
3.88
2.68
5.35
8.03
5.09
5.35
7.76
4.02
8.70
8.70
8.70
4.02
4.02
8.03
2.47
10.31
10.31
0.39
1.95
1.95
2.04
2.17
2.17
2.60
8.09
8.09
4.64
4.04
7.10
2.96
2.96
6.41
8.58
8.58
8.58
5.03
8.38
7.59
7.59
7.59
7.59
7.40
7.40
8.58
Lids Offtakes
0.784
0.335
0.389
0.201
0.402
0.603
0.382
0.402
0.583
0.302
0.871
0.871
0.871
0.402
0.402
0.804
0.240
1.032
1.032
0.145
0.603
0.603
0.630
0.503
0.503
0.804
0.597
0.597
0.342
0.298
0.393
0.164
0.164
0.473
0.633
0.633
0.633
0.278
0.619
0.420
0.420
0.420
0.420
0.420
0.546
0.633
0.485
0.156
0.180
0.124
0.249
0.747
0.485
0.249
0.361
0.187
0.405
0.405
0.405
0.187
0.187
0.373
0.317
0.448
0.448
0.133
0.560
0.560
0.585
0.311
0.311
0.373
0.510
1.021
0.293
0.255
0.896
0.373
0.373
0.405
0.541
0.541
0.541
0.317
1.058
0.958
0.958
0.958
0.479
0.479
0.467
0.541
                                                (continued)
                   7-22

-------
TABLE 7-11. (continued)
No.
17

18








19



20

21


22
23


24
25


26
27
28





29






30

Plant
Anrco Inc., Ashland, KY

Bethlehem Steel .Sparrows Pt., HD








Rouge Steel, Dearborn, MI



National Steel, Detroit, HI

Carondolet, St. Louis, HO


Tonawanda, Buffalo, NY
Bethlehem Steel, Lackawanna, NY


LTV Steel ,Warren,CH
Armco Inc., Hiddletown,OH


New Boston, Portsmouth, OH
toppers, Toledo, OH
LTV Steel, Cleveland,OH





U.S. Steel, Lorain, OH






AL Byproducts, Keystone.PA

Battery
No.
3
4
1
2
3
4
5
6
11
12
A
A
Ax
B
Dx
4
5
1
2
3
1
7
8
9
4
1
2
4
1
C
1
2
3
4
6
7
D
G
H
I
J
K
L
3
4
Status
0
0
1
1
2
2
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
2
2
2
2
2
2
2
2
2
Midrange Baseline Emissions
Charges
0.2357
0.0324
0.3337
0.3178
0.3337
0.3337
0.3337
0.3337
0.3443
0.3443
0.0370
0.0827
0.0184
0.1011
0.0496
0.1434
0.1562
0.0441
0.0199
0.0386
0.3796
0.2357
0.2357
0.2264
0.4502
0.3019
0.3019
0.4025
0.2225
0.1811
0.0296
0.2701
0.2701
0.2701
0.1158
0.1158
0.3125
0.3125
0.3125
0.3125
0.3125
0.3125
0.3125
0.0525
0.0525
Doors
3.30
3.04
2.73
2.60
2.73
2.73
2.73
2.73
2.82
2.82
0.49
1.95
0.43
2.39
1.17
3.38
6.09
5.35
2.41
4.69
2.60
3.30
3.30
3.17
3.69
9.05
9.05
5.14
4.74
3.86
3.45
3.45
3.45
3.45
4.26
4.26
3.99
3.99
3.99
3.99
3.99
3.99
3.99
3.72
3.72
(Hg/yr)
Lids Offtakes
0.764
0.938
0.344
0.327
0.344
0.344
0.344
0.344
0.473
0.473
0.184
0.336
0.075
0.411
0.202
0.583
0.846
0.169
0.076
0.147
0.190
0.240
0.240
0.308
0.855
0.764
0.764
0.764
0.704
0.573
0.161
0.513
0.513
0.513
0.633
0.633
0.593
0.593
0.593
0.593
0.593
0.593
0.593
0.232
0.232
0.473
0.871
0.392
0.392
0.392
0.392
0.392
0.392
0.809
0.405
0.337
0.259
0.057
0.316
0.155
0.448
0.244
0.249
0.112
0.218
0.373
0.473
0.473
0.454
1.058
0.710
0.710
0.946
0.436
0.355
0.142
0.317
0.317
0.317
0.175
0.175
0.734
0.367
0.367
0.367
0.367
0.367
0.367
0.153
0.153
                                 (continued)
      7-23

-------
                            TABLE 7-11. (continued)
No.
31



32
33




34

35

36











37

38


39

40
41



42
43



Plant
Bethlehem Steel, Bethlehem, PA



LTV Steel, Aliquippa.PA
LTV Steel, Pittsburgh, PA




Koppers, Erie, PA

Shenango, Pittsburgh, PA

U.S. Steel, Clairton, PA











U.S. Steel, Fairless Hills, PA

Wheeling-Pitt, Monessen, PA


Chattanooga Coke, Chattanooga, TN

Lone Star Steel, Lone Star, TX
U.S. Steel, Provo, UT



Weirton Steel, Brown's Is.,WV
Wheeling-Pitt, E.Steubenville,WV



Battery
No.
A
2
3
5
A1
PI
P2
P3N
P3S
P4
A
B
1
4
1
2
3
7
8
9
15
19
20
21
22
8
1
2
1A
18
2
1
2
C
1
2
3
4
1
1
2
3
8
Status
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
0
0
0
0
0
2
2
2
0
0
0
0
0
0
0
0
0
2
0
0
0
0
Midrange Baseline Emissions
Charges
0.1273
0.1622
0.1622
0.1273
0.1686
0.0938
0.0938
0.0938
0.0938
0.0938
0.0220
0.0334
0.0156
0.0334
0.1018
0.1018
0.1018
0.1018
0.1018
0.1018
0.0282
0.1384
0.0403
0.1368
0.1384
0.0347
0.1304
0.1304
0.0589
0.0589
0.0302
0.0229
0.0191
0.0406
0.1002
0.1002
0.1002
0.1002
0.2698
0.1458
0.1458
0.1582
0.2450
Doors
5.41
6.90
6.90
5.41
7.17
3.99
3.9S
3.99
3.99
3.99
1.56
2.37
0.35
2.37
4.33
4.33
4.33
4.33
4.33
4.33
0.38
5.89
0.54
5.82
5.89
0.46
5.55
5.55
2.50
2.50
1.29
1.04
0.87
0.24
2.73
2.73
2.73
2.73
5.89
3.18
3.18
3.45
5.34
Lids
0.337
0.537
0.537
0.337
0.335
0.186
0.186
0.186
0.186
0.186
0.097
0.147
0.097
0.111
0.270
0.270
0.270
0.270
0.270
0.270
0.140
0.367
0.200
0.362
0.367
0.172
0.259
0.259
0.117
0.156
0.080
0.076
0.063
0.295
0.199
0.199
0.199
0.199
0.367
0.198
0.198
0.215
0.333
(Mg/yr)
Offtakes
0.444
0.283
0.283
0.222
0.294
0.164
0.164
0.164
0.164
0.164
0.064
0.097
0.236
0.194
0.355
0.355
0.355
0.355
0.355
0.355
0.257
0.483
0.367
0.478
0.483
0.316
0.455
0.455
0.103
0.205
0.106
0.149
0.124
0.295
0.350
0.350
0.350
0.350
1.083
0.293
0.293
0.317
0.983
Status
0=online
1=hot idle
2-cold idle
3=under construction
                              7-24

-------











re
CO
LU
h-
-
rv*
»^
^^
y
~^
CO


.
CM
1
r**>
LU
— i
CO
^c
h-
























-Q
^
-P
C
Ol
o
S-
O)
Q.

I-H
h— I
^J
t- CU
o >
-P •!-
re -P
i— re
3 C
cn S-
cu cu
ry* 4^
r—
^£




J

4_)
C
CU
u
S-
cu
Q_






t-H
1— 1
>*
I- CU
o >
-P •!-
03 -P
•— re
3 c:
en t-
O) CU
r^ ,|_>
r—

c
o

Ifl
in

i




3
c
0
•r—
-P
O
3
•a
cu
S-




/^>
S-
^.
''x^
O)
s:


in
c
o

in
in
• r-
E
CU



, — >
S-
^>^
*x^
cn
^*
^^

in
c
0
•r-
in
in
ft—
E
CU






CU
u

3
O
CO




LT>
00




o
,
(^^
1
o
rH
•
O


IT)

ro









00











s*~\
^-
CM

00
rH
.
0
\^/

CM
rH








X*^S
^O
^^
1
*^t"
CM
,
o
^-s

ro
CM





cn
c
• r-
cn
s_
re
.c
0




CM
cn




CM
t
rH

1
0
•
00


CD

cn
^J"








<^*
^^









S***\
O"^
o
CD
1

.
^j-
IT)
\^^/

CM
ro
ro




s*^
a
CO
o
*»
rH
1
LT>
•
CD
cn


00
00
LO







V)
t_
O
o
0




0





^
•
00
CM

^f
•
^^
\^s

^"
«
CD
rH








^^
CM








s~^.
CO

r-i

i
ro
.
rH
rH
**^s

m

p-i
*d-




/*x
^^
*
^~
CT1*
1
00

^^
rH
•^^

CD

^^
Lf)







W
•a
•^
_j




t-H
^




^
•
oo
LO
I
^t"
•
00


o

r-i
CO








^«
CO









^•^
^3
.
O
CD

^f
.
cn
^*S

r^.
.
^j-
oo




s~**
*^
.
rH
O^
1
«^f
.
"sj~
rH
*^s

a>
,
OsJ
LO



C/l
^
-Q

T3
CU
2c
O

r—
o
M-

CU
3
P—
ro
>

cu
cn
c.
re

"O

*y
ro



































.
in
CU
3
i~-
re
cu
cn
c
re
5-
T3

'i

C
O

•a
cu

re
-Q
cu
c
,n
cu
(/)
re
.a

E
O
J_
M-

C
0
•r—

U
3
•^
cu
S-

-p
c
cu
u

cu
a.
JD
7-25

-------
     Emission estimates for Regulatory Alternative  III  with  limits  of
11 seconds per charge,  5 PLD,  1 PLL,  and 4 PLO are  also given  in  Table
7-12.   This alternative would  reduce  the estimated  nationwide  emissions
of 8SO from 719 Mg/yr to 101 Mg/yr,  an 86-percent reduction.
                                7-26

-------
                              8.   COSTS

8.1  INTRODUCTION
     This chapter analyzes the costs associated with control of emissions
from wet-coal charging and leaking doors, lids, and offtake systems.
The control of these fugitive emissions involves the coordination of
equipment modification, work practices, emission monitoring, and
corrective action.  Emission control at coke plants has improved
significantly over the past 10 to 15 years as the industry has developed
and implemented new control equipment, improved operating procedures
and worker training, and included more diligent supervision of the
emission control program.  The improved controls were implemented in
response to regulations promulgated by the Occupational Safety and
Health Administration (OSHA), State agencies, and technology-forcing
consent decrees negotiated on a pi ant-by-plant basis by the U.S.
Environmental Protection Agency (EPA).  The emission control levels
required by these existing regulations vary in stringency and have
resulted in different performance levels at different batteries.
     Cost data for the control of coke oven emissions were received
for 16 batteries at four coke plants in response to detailed question-
naires and plant visits.  The coke batteries at these four plants were
chosen to include the general types of batteries found in the industry
and include metallurgical and foundry coke producers; old, new, and
rebuilt batteries; single and double collecting mains; hand-luted and
self-sealing doors; short (<5 m)  and tall (>5 m) batteries; and existing
emission limits that range from the most stringent to the least stringent
regulations currently in effect.
     The cost data supplied by the industry represent the costs currently
being incurred by the various batteries to meet their existing regula-
tions.   The analysis in this chapter compares differences in the costs

                                  8-1

-------
incurred at these batteries to differences in the required control
performance.   This comparison is used in an attempt to estimate reason-
able incremental increases in control costs that might be associated
with incremental improvements in emission control.   The following
sections will first describe the batteries, control techniques, and
the data they supplied.   Next, these data are used to estimate control
costs for each emission point.  Incremental costs for improved control
are estimated from the comparison of differences in total costs to the
differences in required control performance.
8.2  COST COMPONENTS
     The cost data from Empire Coke are summarized in Table 8-1.1  The
two batteries are operated as one with one larry car that serves all
60 ovens.  This plant produces foundry coke with a normal coking time
of 30 hours.   The batteries are the Semet-Solvay type and were originally
constructed in 1904 to 1912.   The collecting main on Battery 1 was replaced
in 1960, and both the collecting main and offtake system were replaced
on Battery 2 in 1979.  The new collecting main is larger than the one
it replaced to accommodate the larger gas volumes from stage charging.2
     The company reported that extensive modifications were required
on these old batteries to implement stage charging.  For example, an
expenditure of $715,000 (1979 dollars) was reported for replacing the
collecting main, modifying the coal belt, installing a weigh scale,
and extending the larry car tracks.  Permanent jumper pipes were added
to improve the aspiration system, the battery top was repaved, and the
larry car was modified.   The control of leaks from doors, lids, and
offtakes is accomplished by the manual application of a luting mixture.
Estimates are provided in Table 8-1 for inspecting, cleaning, materials,
and luting labor to control these leaks.
     The cost data from Bethlehem Steel's plant at Sparrows Point are
provided in Table 8-2.3  Battery 1 is a short (3-m) battery with a
single collecting main and hand-luted doors.   Additional aspiration
for stage charging is provided by a jumper pipe.  Battery 11 is a
double main battery with self-sealing doors.   At the time the informa-
tion was gathered (September 1984), the company was in the process of
                                  8-2

-------
                  TABLE 8-1.   COST DATA FROM EMPIRE COKE1
Plant:   Empire Coke, Holt, AL
Battery:  1, 2
No ovens:   60 (single main)
Capacity (1,000 Mg/yr):   160.6

    Item description
Cost
Year
Index   1984 $
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.

13.
14.
15.
16.
17.
18.

19.

20.

21.

22.
23.
24.
25.
Basic stage charging system
Basic written procedures
Aspiration system
Inspect, clean goosenecks
Inspect, clean standpipes
Inspect, clean roof carbon
Inspect, clean steam nozzles
Inspect, clean liquor spray
Larry car modifications
Level er boot
Engineering study
Detailed procedures, all
sources
Training program, all sources
Battery modifications*
Monitoring, all sources
Repave oven top
Inspect, clean, replace lids
Inspect, clean, replace
standpipes
Inspect, clean, replace
goosenecks
Inspect, clean, replace
slipjoint seal
Labor for luting, reluting
lids, caps (6 hr/day)
Inspect and clean doors
Labor to lute doors and caps
Labor to mix luting material
Cost of luting material
397,000
500
10,000
3,100/yr
1,600/yr
1,700/yr
10,800/yr
10,800/yr
14,000
5,000
74,000
1,000

1,000/yr
715,000
10,000/yr
6,000
3,100/yr
1,600/yr

3,100/yr

1,300/yr

46,000/yr

7,300/yr
61,000/yr
26,000/yr
20,000/yr
1974
1974
1974
1984
1984
1984
1984
1984
1979
1979
1974
1984

1984
1979
1984
1974
1984
1984

1984

1984

1984

1984
1984
1984
1984
2.0a 794,000
2.4° 1,200
2.0a 20,000
3,100/yr
1,600/yr
1,700/yr
10,800/yr
10,800/yr
1.3C 18,000
1.3r 6,500
2.4° 178,000
1,000

1,000
1.3C 930,000
- 10,000/yr
2.0a 12,000
3,100/yr
1,600/yr

3,100/yr

1,300/yr

46,000/yr

7,300/yr
61,000/yr
26,000/yr
20,000/yr
Note:   For the footnotes to the indices see Table 8-5.
^Includes replacing collecting main, modifying coal belt, installing a
 weigh scale, and extending larry car tracks.
                                  8-3

-------
















CO
_J
UJ
LU
1 —
OO

21
LU
31
UJ
— 1
n:
i—
LU
03

S
0
or
LL.

<

05

•"-
''0
cu
10
1
4-
r—
CU
t/>
II

*-O
oo

*,
•Q





Q O ID CM OO 00
S  LO CM OO LO
10 rH tO • OO IO
3 00 OO
O
i-
i.
ro OO OO rH — I OO
O_ i — I to f*"^
OO CM

r.
t— • '
CU f — »
cu s-
-p >>
00 \
o>
E S
Ol
f— ^-^|
J— V_J
0) O
r- O
-C "
-P .. rH
CU • • /•— s • • " — '
CQ 10 E r-
.. C v^ >
>> CU CU -P
.. S- > -P • • U) •!-
4-> CU O -C U1 U
c P 0) c j_ ro
mj > ._ «,— /-J C"\
.f . -1— 1 \J u.
i — ro o cu ro o ro
CL. 03 Z IE S Q C_>

•f?"

00
CD
rH

X

S_
QJ
-P
^_>
ro
J3


C
o
• f—
+->
O-
5-
U
CU
~rj

E
CU
+J
1— 1




O OOOOO OOOOOOOO OO O OO OOLO
O OOOOO OLOOOOOOO OO UD UDO OOI^-
OlOOOO«d-i— llrHOOOOOOOtO OILn rHl<^-OlOCOCM
LD OOOLn«^-CM CM OOOOCMOLT) r^OO rH«J- O/^
LO1^- LOrHLnrHLn • tOrH
OO ^- CO CM 1 — rH
LO
CU CU 4— 4 — O) CD
to to r^ to CM CM
II - . 1 1 1 1 1 1 .... I 1 II 1 1 1 1 1 1 1 1
rH rH rH rH rH rH
^^.^ ^ior,oo«-^ ^ ^ ^^ ^^^
ooif^i~~-coooooioooor^.r^ooooooooiooooo ooiooooioooooo
CD CD CD CD CD CD CD CD CD CD CD CD CD CD E CD CD CD CD CD CD CD CD
rH rH rH rH rH i— 1 rHrHrHrHi-HrHrHrHCUrHrH rH rHrH rHrHrH
-P
LT>
i —
10 ro E
o> =} cu
OOOOO 'r->,-OOO-P O
>>EEE EE>i>,>, S- i- ••- E E I-H OEO
-P i- 1 1 1 i i S- S- S_ Ol O) > 1 1 ElE
ui ro-pi-s-s-i s-t_4J-p+Jroroccucu roi oro ccucu
1 1 1 Q_ ro -p -p 4-^ +J -P ro ro ro co .0 f~* *r~ -P 4-) ~o * * ' f ^ uu ro -p >
CQ +J P -P P +-* CO CQ CQ 1 -P -P CU i — -P O
ro ro ro ro ro CD j- c ro ro "o ci_ ro
CO CQ CQ CQ CQ 1 >- 'r- CQ CO 13 m
S- i—
>- ~o L)
CU C
T3 i — I
3
i —
ocuooooocuoooooooouoo ocuoocuoom
or-ooooor— OLDOOOOOOCOO Ln^Loor- oor^
OJQOOOO^"rH^2rHOOOOOtDOlOi— ICDUD rHjQ^-OJaOCOCM
LO O)Ln O LO "^- CM D5CM OOOLOCDOLn 1^00 COrH «*• D)O r^.
•r- •=*• OO •!- O OO CM LO -r- -r- tO rH
i — OOCM i — LOrHtO i — i — rH
O5 - CJ) OT O5
cu oo cu cu cu
z z z z

rH *~^
rH rH in in
^~> rH ,-> S- 5-
v — ' *~s i — I  <~~^ "U
-'-N ^-/v^^rH O O rH rH C C
i-H — ' cu cu >> >> - s- ' — •* ro
i — . — rororH +->P -^ — >O -P l/>
IO IO INI N S- S- ^— ' CC rHOCOinc: C
^ai J^ NI N O- O- ''""^ ro n3 "^-^ "^.^ "O ' — •! — CU
uoooioiotorH /— >, /"> i — i — <•"> ro ro >
OICUCC Ci— IrHrHO-Q. rH r— r— S->
miororo33+JrH _^ ^ -/ — s in in S- +-> i — -P
m oocuo~o""ro^— 'o_a.oo r-t<: mo inro-r-
CU E E O O 4^ 4-^ "i — *f CJ O O CJ CJ >j"x^ ^^ 4-3 4-3 CU O 4-3 4-3 r—
S_CUCUO5O)W)ini— r--r-r~4->4J "D CCS-T3U1CUCUT-
. — -34->4-3 M-ro •— •— 3D) D)-i- -r- 3 O -^ U U
rH "O t/)lOCCCZCCC *' — CU 5^ ^> < — ' — -P C C O O "O •• — O to *r- ro
v-^cu>>>>rorororororo"oi/)j-s-roroio •!-•!- -i-j-r-j cu i — ro > 4-
OU1U)CUCUCUCUCUCUO CUCU r- i — O-O-O3 -O1S-
o-s-ccoutjuuu ro4-34-3cno)c:cucui — i — s-'r-s-s-m-i-
•r-o-.oo s-jarorocc-r-tni/jioioo-zoo ro
c-p-P-p-p+j-P-P-Pcjs- s-t-cuooaj-^-^L)uaji — -PS-
0^ -P £-t-Q^> r~~- > > -p -p C 03 (Tj fl3 fU -P fO QJ -^ TO
Q» ^-> •!-•!- Q-Q-O-Q-Q-Q-S- 0) 03 «J «i- «r- «r- -P -P i — . — -P-P S- (J S- S-
£T-CLO.(/)(/>(/)(O(/>I/)&->O.Q.CC om- 4- o. o_*r- w n3 n cu o
^i-l/)COCCCCCCIfOQJQJQJOOC M — ^- CU QJ t- C Q_ _C Q_ O
•"5 3* ^C *^ *~H I-H h—l l-H h— I HH 	 1 	 1 Q£ Q£ 2! S! Lj-J ^3 C* *"**"" Oi 3S '""' t/) CJ C3 "O



rHCsJ(T)^iX)^r^COCT^i^rHrHrHrHrHrHrHrHrHCSjC\JC^JC\ICMC\JC\l
8-4

-------
equipping 10 doors on Battery 11 with vented door plugs and modified
seals made of NiCuTi alloy.  Jambs are repaired at a rate of about
2 per week by welding cracks and grinding to a smooth finish.4
     Battery A is a new 6-meter battery equipped with a double collecting
main and self-sealing doors with a knife edge seal made of NiCuTi
alloy.   The cleaning of door seals and jambs is performed by an automatic
door cleaning machine and is supplemented by manual cleaning at a rate
of once per month.  Two persons per shift are assigned to make door
seal adjustments and to perform manual cleaning as needed to reduce
leaks.   Lids are luted manually, and standpipe lids have a water seal
that requires cleaning about once per month.4
     The company data given in Table 8-2 include the major capital and
labor cost items for emission control.  Some of the data were reported
as a plant total, which was converted to a per-battery basis by dividing
the total cost by the number of batteries at the plant.
     Armco's coke plant at Middletown, Ohio, has one short battery and
two recently reconstructed Carl-Still 6-meter batteries.  All three
batteries are equipped with double collecting mains and self-sealing
doors.   The short battery was installed in 1953 and has not required a
major rebuild or rehabilitation, probably because the battery has not
been cooled down.  The 6-meter batteries were built in 1976 and extensive
modifications were made recently to improve structural integrity and
door leak control.5
     Armco is finishing a door leak control program that included jamb
straightening or replacement, pressure-reducing door plugs, modified
door seals and adjusting springs, and buckstay reinforcement.  Extensive
repairs were required on the 6-meter batteries because the original
door seals leaked badly and caused door fires when the batteries were
new.  A total of about $7.5 million was spent to modify the entire end
closure system on the tall batteries and includes expenditures to
correct a structural design problem with the batteries as originally
constructed.   Door repair is performed onsite and is supplemented by
the services of an outside contractor.  Armco personnel believed that
tighter tolerances on seal specifications plus a higher frequency of
door repair or replacement would be needed to reduce door leaks.5
                                  8-5

-------
     The cost data from Armco are summarized in Tables 8-3 and 8-4.6
The extensive door modifications for the tall  batteries in Table 8-3
total $7.5 million.   In addition, other modifications were made on the
tall batteries and are summarized in Table 8-4.   Most of the items
listed in the tables would tend to improve emission control; however,
it is not clear what portion of the cost, if any, should be attributed
to emission control  and what portion of the costs should be attributed
to maintenance and prolonging the life of the battery.  Table 8-5
presents the footnotes to indices in Tables 8-1 through 8-4.
     Cost data were also received from U.S.  Steel's Clairton Works
where extensive research, development, and implementation of improved
control techniques have occurred over the past 10 to 15 years.7 8 9
The cost and labor hour estimates provided by the company were claimed
to be confidential and are not presented here.   The control costs
incurred by U.S.  Steel included the costs for monitoring, cleaning,
repairing, and luting various equipment items.   In addition, costs
were incurred for maintaining or replacing slip joint seals, jamb
repairs, door seals, and the door leak control  program.
     The batteries at U.S. Steel's Clairton Works include old and
rebuilt batteries and a new 6-meter battery.  All batteries have
double collecting mains and self-sealing doors that were modified to
improve emission control.  The company optimized its stage charging
procedure, modified doors, seals, and procedures to obtain consistent
emission control, and installed new slip joint seals to alleviate
problems with offtake leaks.10 X1
8.3  COST ESTIMATES
     In this section the individual cost items supplied by the companies
are compiled for each emission point.  The major capital and annual
cost components are identified and total annualized costs are estimated.
Capital cost items are annualized over an average 20-year lifetime at
a rate of 6.2 percent (after adjustment for inflation).
8.3.1  Charging Control Costs
     The charging control costs for Empire Coke's battery are given in
Table 8-6.  The two major capital items include extensive battery

                                  8-6

-------
to

cj
O
O

a:
o
a:
<

o
co
o
co

oo
03
, CU  O
                           01
                          o
                          o
                          o
                    cu
                    >
                    o  in  u
                           ro
                       -^  Q.
                       ro  ro
                       2: «_>


•w-

oo
en
rH


X
cu
T3
C
l— l
S-
ro
CU



-P
C





-p
in
U

-P
c




/*^
.
o

>J
S-
cu
-p
-p
ro
o
Sw>

C
o
•f—
-p
Q.

i-
U
in
cu
T3

£
CU
•P
i— t


O
o
o
o
CM
rH




1

00
en
rH

i.
1
g
*-
CU
-p
+J
(0
CQ
O
O
0
O
CM
i-H

s=
0
in
i-
Ol
Q.
v— '
in
S-
o
0
~rj

S-
o
4—

s-
o
ja /•>
ro rH
r— v^^

i — /-••,
ro -P
C <4-
0 ••-
• r— J"
+-> CO
•r—
"0 S-
"O cu

CM
^
in
J-
0
O
•o

S-
o
4—

S-
o
J3
rO


r—
ro
c
O
• r—
-P
»^-
•o
•a
<
CM

O
CD
O
LT>

CM




|

00
en
rH

J-
1
c
>—
cu
-P
-P
ro
OO
O
0
0
LD
en
CM













^•^
i— i
^^/

in
-P
in
0
u

S-
•f—
rO
Q.
CU
S-

S-
o
0
o
CO

0
o
o
co
en
CM




i

«3-
00
en
rH

S-
1
r
9—
cu
-P
-P
ro
oo
o
o
0
co
en
CM










(**-••»
CO

r*
OJ
^s

t/>
-p
in
O
u

s-
•i—
fO
Q.
CU
S-

S-
o
0
o
«3-

o
o
0
ID
rH





1

00
cn
rH



.C
ro
UJ



0
O
0
ID
rH

















/• — s
rH
\^^

00
1
1
in
i.
o
o
•a

cu
S-
ro
Q.
CO
LO

0
o
o
LT)
CM





1

00
cn
rH



JZ
ro
UJ



0
o
o
LT)
CM













s***>
ro

«s
CSJ
^fc^/

ID
rH
1

in
i.
o
0


cu
i.
rO
Q.
CO
ID

O 0
0 0
«3- o
O
0
CO




1 1

CO 00
cn cn
rH rH
in
cu
5-
j= cu
ro -P
UJ rO
o

CM
O O
0 0
«tf- o
O
O
co










s^*
CO

r>
CM
in •^^
S-
o e.
O (0
*^3 t»
cn
\s O
U J-
3 0-
JZ
O -P
in
CU CU
U -P
rO
i— S-
a. o
cu o
Q£ 0
r^ oo

oooooooooooo
oooooooooooo
oooooooooooo
cnr^coiDr^-^t-coooiDocn
OOCMCOCMOOOO^-r^^-lDOCO
cn CM r^»* ^* rH ID CM
** f
CM rH


1 1 1 1 1 1 1 1 1 1 1 1

000000000000000000000000
cncncncncncncncncnoScncn
rHrHrHrHrHrHr-HrHrHrHrHrH
i/)ininininininininininin
cucucucucucucucucucucucu
S-S-S-t-J-S-S-S-S-J-S-S-
cucucucucucucucucucucucu

rO T3 rO ro ro ro ro rO ro ro ro ro
o ^2 f^ o f> ^3 f^ o o f*> o o

CMCMCMCMCMCMCMCMCMCMCMCM
OOOOOOOOOOOO
OOOOOOOOOOOO
OOOOOOOOOOOO
cnr^-coiDp^^-coooiDoen
00 CM CO CM 00 00 ^" f^*- S_OS- D.ororOU
in *r— o o o ^> T— cu cu cu
o in-Po~aoroi— -PI— i— s-
i— r— UT3 T3 -P r— U U U
Uincucu CU inrocu -P
O ^_) ^o • t- C *^ 4-> Q. ?- J3 d
T3ECOj3rocuuininoECU
crOT-5-rOQ.>3CCorO>
UJ'-J — IQ.U_cOOQQi— 'l— ica^UJ
cn

0 O 0
o o o
o o o
*3- O «^-
0 CM CO
f""1*- p-1 t^
"
ID


1 1 1

00 00 00
cn cn cn
rH rH rH
in in in
cu cu CU
S- s- s-
ai cu cu
i % i % i s
•i—* -i— * +-*
ro ro ro
O o o

CM CM CM
O O O
O O 0
O O O
*s- o «*
c^ c\j oo
r^* I~H («D
ID




















cn >,
C (J
•r- C
r— S- CU
(0 Q) cn

O C •!-
-P -r- -P
ja cn c

CO UJ O




o
o
o
00
IT)
^~
r>
^


1

00
cn
rH
in
cu
S-
cu
i ^
•+-*
-P
ro


CM
O
0
O
00
LD

^


























r..
ro
•P
O
I—



•a
at
^
c
-P
f~
o
u
****
























































                                                             8-7

-------























,^^
73
CU
=3
c
-P
c
o
o
^ 	 ^

OO
1
CO

l-l-J
	 1
CQ
«=c
I—








































tfl-

*3~
00
en
r-l
X
-





-p
•1 —
C
m









-p
V)
o
u

-p
• f—
c;
r?








r^-%

o
•z.

>>
J-
O)
-p
-p
(C
jQ
•%^>

C
O

-p
O-
• t —
5-
>
S- 1
>, ui
1 QJ
>>•>-
i- i-
CU CU
-P -P
-P -P
ai<3-^i-^j-'3-'3-<=i-
ooooaooor--r^.cocooooooooo
CJ1O1CT)CTlCTIC7>CT>C71CT>Cncr)
r-l r-H ?— 1 rH i— 1 i— 1 r- 1 i— 1 i— 1 i— ) i— \ r— 1
i. i.
01 tn >, en >,
ai cu s- s- i ai i
•r- >-, >, >>•!- >, >, 1^1 •!- 10
1- S- S_S-t-liO)S-O)
cu-c:cu^:a)aia)S->,-i—  « c
CM CM -P a) ro CM ro
^ — ' **.*> ro S- ^- — ' i —
s~^s~*, U O -^~\ p
en in i— ( i— i -r- - — >~ac\j roin
C!~v~^^-'c-l-rHa) C - O -V
•t- •'- (SI (S) ~<3 O i— 1 rH CM -i- O
-P-P Q.Q.O >, S- V- ' v— ' •-— ' +J +- CU 0) C in O •>- •!-•!- 4- S-S-i--i-^oo:z^'-OUjQS33:<:s

CMCO^lDliJr^COCDOl-HCMrO
1-HrHi-Hi— Ii-Hi— li— trHCMCMCMCM





























.
in
i
CO

£U
S
ro
t_
r-~
OJ
01
in

m
O)
U

•5
c
•1 —

a>
^z
4->

O
-p

in
CU
+J
O
c
+J
o
o
4-

(U
JZ
-p

5-
O
U_


, .
CU
1 1
^^
o
z:
8-8

-------













00
LU
1 — i
LU
1—
r—
f^
CO

I i I
l-LJ
1—
LU
s:

CD

jg
LU
z

o
jt
h*~

ry
o
U_
00
h-
oo
o
C_J

0
o
^T"
o;

CO
o
o

p
•r—
C
~*^






















c
o
•r™
P
Q.
• p-
£-
o
cu
"73

E
OJ
P
i— i



O O
o o
0 0
0 O
oo p***
<3- CO
rH rH

x: _c J

rH i-H


00 00
r-- r--
cn cn


co co
cu cu
i- J-
cu cu
P P
-p -P
ro ro
XJ .0.

CM CM






OO CM
rH 1^*-
CM 00
CO CO
CM CO
O CO
». *s
rH rH



*

s/
U
CU
c *
CU in
CO CU
O >
O p—
oj ro
>
-o
c «•
ro co
S-
eO CU
CU >
Q. 0
•r- CO
Q. CO
•a o
C i-
ro u
•P
cn »*
c
ro •!- *
t- ro m
-P E S-
x cu
 4-j +j 4-)
fO (0 co fi3
^^ _^2 O o

CM CM CM CM






r^ o CM o
r-» o o o
r^ oo o LO
co CM cn oo
rH •!*• CD f--
LO rH CO rH
•>
rH



E eo
au cn i
•PC ro
CO -r- U
>> s-
LO a. 4-
CO •!-
t— "O
cu - o
P co E
ro -a
3: o i-
i. ro
>> co U
U CU C
c -p- o >>
CU -P -P- S-
O) -P S-
s- » ro ro
CU CO O i —
E >>T-
cu ro M- s.
•P T- 0
•« co "O 4—
CT!-^ O
CUED)
•r- 3 C
p* r"> i_ >r—
•r- ro s-
O- S- U CU
cu cu
E T- >^ C
ro > s- T-
cu ro s- cn
•p cu ro c
00 31 — 1 LU

• ...
CO •* UO IXD
O O O O O
O O O O O
CO O O O O
ID rH "3" ID CO
CO rH O CD CO
rH rH CO 00


x: x: x: x: x:
^ -sj- ^J- •=!• «^-
rH rH rH rH rH


00 00 00 00 00
r***« r^> r^> r^* r^»
cn cn cn cn cn


CO CO CO CO CO
cu cu cu cu cu
i- i. i. s. i.
cu cu cu cu cu
+J 4J +J +J +J
+j .p +j p +j
ro ro ro ro ro
o o XJ f^ f™t

CM CM CM CM CM





•f*-
LO O LO i-H rH
CM 0 r-N O O
ID 00 00 O O
rH CD «* to LO
tx* r** r^ r^» CD
^* LO








S-
in in o
CU "O O
•— T- T3
O i—
x: cu
T3 -O
O5 C -i-
c ro co
•r- cu
cn co .*:
1- CU CO O
ro •— co
x: o o
ox: co T- CM
\s .p
4- cn o ro 4-
o c cu o o
•r- C •!-
3: cn cu 4- P
O S- CO T- CO
s- ro o -o o
x: o o o
r- o cn E
ro ~a
c -a -a c cu
o cu cu T- to
•r- -r- T- xi ro
CO +J 4- q_ O)
C •!- •!-•!- P- J-
o T3 -o T3 ro o
•r- -a o o o c
+j <: s: s: o t-n

• • • 0 rH
r^% oo cn rH i— i
o
CD
o
00
CO
CM


•1- J.
o
CM


00
r**
cn


CO
cu
J_
cu
-p
-p
ro
.a

CM






cn
cn
cn
CO
rH
rH








-o
C
rO

CO
CU
r-
•r~
x:
o
ro
£•

s-
o
o
-o

i.
o
4-

cn co
c cu
•r- TJ
CO 1- •!—
CU CU 3
c cu cn
•r- C
x: -i- cu
o cn-i*:
ro c o
E LU 0

CM
rH
o
o
o
rH
^-
LO


- J
<*
rH


00
[**^
cn


CO
cu
j_
cu

1 ^
ro
XJ

CM






rH
LO
r^
V£)
00
CO






•P
c
cu
E
a.
»p-
3
CT
CU

S-
cu
• r-
>
ro
cu


i.
o
4-

P
C
cu
E
CU 10
o s-
5- 0
O O
4- T3
C
•r- S-
cu o
o: 4-

ro
rH
OOO OOOOOOOO
000 OOOOOOOO
OOO OOOOOOOO
CD f^> CD ^f 00 rH CD 00 CD CM 00
CD LO CD LO CD Cn ^" LO CO rH CD
CD CO IX1 r*^- rH CM rH ^J- r^^ OO CD
CO 00 CM CO

= -r-x: x:x:x;x:x:x:x:x: J

l-HCMrH i-HrHrHi— IrHrHrHrH


000000 OOOOOOOOOOOOOOOO
r^^ r*^ r^*> r*^ r^ r^ r^* r*^« t**« r**^ i^^*
cncncn cncncncDcncncncn


cococo coinincococococo
cucucu cucucucucucucucu
S-S-l- S_S_£-S_S-£-l_l-
cucucu cucucucucucucucu
-P -P +-^ 4^ 4^ P -P -P -P -P )-%
+JPP pp.p-ppp.pp
rororo rorororororororo
o o X} o o XI X) X) r"i xj X)

CMCMCM CMCMCMCMCMCMCMCM






OO ^~ !^ OO LO ^t* ^D O*) C^ LO ^*""* LO
OOO CJ^r^OLOLnrHCXJOO
OLOCNJC\JCDLOO
CNJ oo io oo r**» r***- o** r^ rH c\j CT^
^j- px. PN. (Y) r*1* CD CNJ CNJ CNI CNJ *sj"
rHrHLO LO CNjLOCOLOCNJpH
^ +. •* r.
CM CM rH CM





i-
O
4-
co
CU E
C CU
•r- C -P
x: cn co
O T- >,
ro XJ co c c co
E E CU O o
ro T3 -r- -r- S-
£- T-J -P +J 0
cu Q. ro ro 3
x: "O 3 co "o -P cr
co in C X) C C T-
CU3ro T3 E 3 CUr—
co_ cu ro o co E
•r- 4_ -p-3 4- ro 3 en
x; s- i. cu a) i- c
ooo cu s-cncus-4J-f-
ro4-ocoxicocuc>roinx:
s -a c s--o -r-'i- ceo
cn o »OT-i-cos_T-3C
S-C -•!- in O 3t 3 C -i- r— -r-
cu *r- ^ .p r— 13 o cu ro "o 4— ro
XT S- s- ro ro " E o. CL cu E
eoCUOOCUCUi-l-XcUeOT3
3 cu 3- *r- co(-curocus*roc -o
OL c -s^ 4— ro *r- cu ro c
•>-OT-J-D->CCUS_£- o
o cn-r- -aocorocus-oos_o
3:cs-oo cu>oocrocu
h-LUQQ EQ003:OS:ai-<|— eo

^-LOl£> 1 — OOCnOrHCMCO^t"
rHrHi-H rHrHrHCMCMCMCMCM
O
O
O
i->r
o
«=t




r-i


00
p^-
cn
•
rH

en
CU
J-
cu

p
ro


CM






o
o

o
CD
CM










p—
o
S-
p
c
o
o

>,
-p
•p-
co
c
cu


"T^j
c
ro

O)
c

X
•r—
C

r-
ro
o
o

LO
CM
































.
LO

00

cu
r^
ro
i—

cu
cu
eo

CO
CU
U

-a
c
•P_

cu

p

o


en
cu
P
o
c
p
o
o
4-

cu

p
s_
o
u_

. ,
cu
. ^
•4-"
o
z
eo
a
s-
4-
^J
ro
'CL
eo
S-
•p-
ro
P
co
CO
ro
2p
•^
'ro
in
^.
O
ro

3
cn
cu


T3
c
ro

cu
CO
3
O


CD
C
•r-
P
ro

3
cn
cu
S-

„
cn
c
•p-
O-

O-

S-
o
3
cr

i—

cn •
c eo
•p- O-
x: ro
eo s-
3 -P

4— r*
O
CO 4->
CU -P-
•0 Q.
3

O CO
C -P
•r- S-
o
0 Q.
CO Q.

-------















J —
00
O
O
CD
i — i
h-

Q
Q.
-3
o;
o
u_

OO
o
O
2
1 — 1


LD
CO

LU
	 1
GQ
I—





















>><*
i— CO
rs cn
O rH

O
CO
cn
i-H

cn
cn
rH

CO
cn
rH



r-~
cn
rH



U3
r--
cn
rH



rt
cn































1X3

un
^-
00
in
,
CM
cn
CM
LD
^
LD
CM
UD
CD

CM


cn
o
CM
CM


00
.
LO
0
CM


CM
r-i
p^.
f— ^


in
-P
s-
o
o.
§"
in
„
^>.
S-
Ol
•^
u
ro
£
C
cu
E
o.

^
cr
LU



l-i
CM
.
OO
ip
CM
OO

<^J-
0
CM
cn
<*•
cn
i— i
CO
i-n
oo
rH


CM
00
r^x
I—I


CM
.
^t"
r^
rH


00
00
<*D
rH












S-
ro


C
•i —

in ro
•i— LU
s_
Ol
Q.
^
-a o»
c u
ro T-
-o
C. 1— I
»l —
S-
OJ
O!
c
•i- OJ
O5 -P
c o
LU C
.p
O
o
OO Lu

































<-^-cncnr^.ix>ocococn
r^^ r^-- r^** r — r*^- r^*> oo i — * r^- r^^
cncncncr>cncncncncncn
1-HrHrHrHrHrHrHi— li— It—t

rHOOrHOOrHrHrHr-HOOCM




o*d-oor-^<£>r^.cM*3-o*3-
CMCMrHrHrHrHrHrHCMrH

II II II II II II II II II II


C\J rH LO LQ O^ 00 I-O CO CSJ O*i
r*H CT) *^^ O^ c*™) i_f^ c\l ^3 00 ^^"
rHrHCvJrHCMCMCMCMrHi-H
^^ ^\ *\ *~\ ""\ "X^ "\ ^s^ \^ ^-^
inr^ti3r^.ix)i-D<-O(£ir^cM
OOOOrOOOOOOOOOOOOOCM





II II II II II II II II II II

(T3 _Q (J "O CD M— O^ f~ 'r— 'i —



8-10

-------
                 TABLE 8-6.   CHARGING COSTS AT EMPIRE COKE
Plant:   Empire Coke, Holt, AL
Battery:   1, 2
Description:  60 ovens, foundry coke, single main,  160,600 Mg/yr

A.   Charging costs—major capital items

     Item                                             $1,000
Basic stage charging system
Develop procedures
Aspiration system, jumper pipes
Larry car modifications
Level er smoke boot
Engineering study
Battery modifications3
Repave oven top
Total
794
2.2
20
18
6.5
178
930
12
1,960
B.  Annual costs

     Item                                            $l,000/yr

Inspect, clean goosenecks                                3.1
Inspect, clean standpipes                                1.6
Inspect, clean steam nozzles                            10.8
Inspect, clean roof carbon  .                           0-1.7
Inspect, clean liquor sprays                          0-10.8
Training program                                         1.0
Monitoring      .                                         5.0
Increased steam                                         34
Capital recovery6                                      174
     Total                                             230-242
     $/oven-yr                                       3,800-4,000
     $/Mg                                             1.43-1.51

 Includes replacing collecting main, modifying coal belt, installing a
 weigh scale, and extending larry car tracks.
 Assumes 0 to 100 percent attributable to baseline regulations.
 Assumes 50 percent of plant total is for charging.
 Estimated as $0.21/Mg coke.
eCapital recovery factor = 0.0886 (20 years at 6.2 percent).
                                  8-11

-------
modifications ($930,000) and the installation of the basic stage
charging equipment ($794,000).   Recurring expenses such as periodic
inspection, cleaning, training, and monitoring are included as annual
costs.   A range of 0 to 100 percent of the cost to clean roof carbon
and liquor sprays is attributed to emission control  because these work
practices may be needed for normal operation independent of emission
control.
     The increase in steam usage was estimated as 17 kg steam/Mg
coke13 and the cost of steam at a coke plant is estimated as $10.50/Mg
in 1979 dollars ($12.50/Mg in 1984 dollars).14  These two figures
yield an estimated cost of $0.21/Mg coke for increased steam for stage
charging.
     The costs for Armco's Wilputte battery are summarized in Table 8-7.
The major capital components are modifications for stage charging and
a detailed engineering study to optimize the equipment changes and
operating procedures.  Additional labor added to the topside of the
battery  is the major annual cost item.
     The costs for Armco's two tall batteries are summarized in Table 8-8.
Extensive modification of the larry car, an important component of
basic stage charging, is the major capital cost item.  The company
also attributed the costs of a second collecting main and the associated
offtake  system to charging emission control.  The range used for these
two items assumes that their cost may be attributed either entirely to
emission control or entirely to the cost of constructing (or recon-
structing) a battery.  Most tall batteries have been constructed with
two collecting mains.
     The costs for Bethlehem Steel's batteries are summarized in
Table 8-9 and again the major capital costs are for the basic stage
charging procedure (larry car modifications and improved aspiration).
In addition, the top of Battery 11 was repaved and leveled at a cost
of $756,000.  Labor is required for inspecting and cleaning equipment
and for  visible emission monitoring.
     The costs associated with control of charging emissions were
estimated for three U.S. Steel batteries and the totals are given in

-------
         TABLE 8-7.  CHARGING COSTS FOR ARMCO'S WILPUTTE BATTERIES

Plant:   Armco, Inc., Middletown, OH
Battery:   1
Description:   76 ovens, double main, 448,000 Mg/yr

A.   Charging costs—major capital items

     Item                                                $1,000

Stage charging modifications                              485
Engineering study                                         304
Development of procedures3                                  8.5
     Total                                                797.5

B.   Annual costs

     Item                                               $l,000/yr

Additional laborb                                         120
Worker training                                             5.4
Monitoring3                                                 1.3
Steam            .                                          94
Capital recovery                                           70.7
     Total                                                291
     $/oven-yr                                          3,800
     $/Mg                                                   0.65

 Assumes 50 percent of plant total is for charging.

 Assumes 1 of 2 additional people per shift on topside added for charging
 control  (lid removal, replacement, gooseneck inspection and cleaning).
cAt $0.21/Mg coke.

 Capital  recovery factor = 0.0886 (20 years at 6.2 percent).
                                  8-13

-------
           TABLE 8-8.   CHARGING COSTS FOR ARMCO'S TALL BATTERIES
Plant:   Armco, Inc., Middletown, OH
Battery:   2, 3
Description:  57 6-m ovens each, 664,000 Mg/yr each
A.   Charging costs—major capital items

     Item
Larry car modifications
Engineering
Modified charging holes and lids
Modified goosenecks and standpipes
Coal mixing and density control
Leveler bar seal                      .
Second collecting main and accessories
Extra standpipes and goosenecks
Askania modifications
Development of procedures0
     Total

B.  Annual costs

     Item
  $1,000
   1,830
     357
     111
     104
     407
     213a
   0-1,870
   0-1,430
      15
      17
 3,054-6,354
$l,000/yr
Worker training
Monitoring
Steam
Labor ,
Capital recovery
Total
Per battery-yr
$/oven-yr
$/Mg
8.1
2.7
279
60
271-563
621-913
311-457
5,400-8,000
0.47-0.69
 Derived from Bethleherr, Steel questionnaire.

 ""Assumes 0 to 100 percent attributable to baseline regulations.

 'Assumes 50 percent of the plant total is for charging.

 JAt $0.21/Mg coke.

 "Assumes one-half of additional topside labor is for charging  (lid removal,
 replacement, gooseneck inspection and cleaning).

 Capital recovery factor = 0.0886 (20 years at 6.2 percent).
                                  8-14

-------
              TABLE 8-9.  CHARGING COSTS FOR BETHLEHEM STEEL
Plant:   Bethlehem Steel, Sparrows Point, MD
Battery:   1, 11
Description:  Battery 1 has 63 ovens, single main, 273,000 Mg/yr.
              Battery 11 has 65 ovens, double main, 363,000 Mg/yr.
A.   Charging costs—major capital items
     Item
                                                        $1.000
                   11
Larry car modifications
Improved aspiration system
Leveler bar seal
Repave battery top
Jumper pipe
Engineering study
     Total

B.   Annual costs
   850
   448
   213
  850
  448
  213
  750
Included in above items
 1,516
2,261
                                                       $1.000/yr
Item
Inspect, clean goosenecks
Inspect, clean steam nozzles
Inspect, cjean liquor sprays3
Monitoring
Steam ,
Capital recovery
Total
$/oven-yr
$/Mg
1
70
25
0-25
2.8
57
134
289-314
4,600-5,000
1.06-1.15
11
70
25
0-25
2.8
77
200
375-400
5,800-6,200
1.03-1.10
 Assumes 0 to 100 percent is attributable to baseline regulations.

3Assumes 50 percent of plant total is for charging.

:At $0.21/Mg coke.

 Capital recovery factor = 0.0886 (20 years at 6.2 percent).
                                  8-15

-------
Table 8-10.   The derivation of total  costs for these batteries  is
consistent with the previous cost estimates and the cost components
are similar to those identified in the estimates for the other  three
plants.   One difference in the estimates derived for U.S.  Steel  compared
to the other plants is a more labor-intensive effort to clean,  repair,
and maintain goosenecks, steam nozzles, collecting mains,  and charging
ports.  A higher cost for visible emission monitoring was  also  estimated
for U.S.  Steel's Clairton Works.
     The costs for each battery used in the analysis are summarized in
Table 8-11.   Charging emission control costs range from $230,000 to
$485,000 per year or $0.47 to $1.51/Mg coke.  Also given in Table  8-11
is the current charging emission limit (baseline) applicable to each
battery.   The average performance given in the right-hand  column of
the table represents the estimated average control level that will
enable the battery to meet the limit a high percentage (approximately
95 percent) of the time.  The costs summarized in Table 8-11 are costs
that have already been incurred because of essentially complete conversion
by the industry to some form of stage charging.
     The costs in Table 8-11 appear to be inversely related to  the
baseline regulations, i.e., lower costs are seen for higher baseline
limits.   Battery 20 at the Clairton Works is an exception  in that  the
costs for all of the Clairton batteries are similar; however, Battery 20
is not achieving the stringent limits (11 seconds per charge) 100  percent
of the time.  The average midrange costs for each plant were determined
on a per battery basis and are listed below with the estimated average
performance required to meet the emission limit:
         Plant           $/battery-year      Average s/charge
     Empire Coke            236,000                24
     Bethlehem Steel        345,000                17
     Armco, Inc.            338,000                17
     U.S. Steel             431,000                 9.3
A linear regression of cost versus average seconds per charge yielded
the equation:
                                  8-16

-------
          TABLE 8-10.   CHARGING COSTS FOR U.S.  STEEL AT CLAIRTON

Plant:   U.S.  Steel, Clairton, PA
Battery:   2,  20, 21
Description:   Battery 2 has 64 ovens, double main,  300,000 Mg/yr.
              Batteries 20, 21 each have 87 ovens,  double main,
                535,000 Mg/yr
Cost estimate
   Annual cost ($l,000/yr)
   $/oven-yr
   $/Mg
 Battery 2

  377-436
5,900-6,800
 1.26-1.45
Battery 20. 21

   425-485
 4,900-5,600
  0.79-0.91
                                   8-17

-------
























>-
a;
 o>
ro O>
i-
"O ro
CU -C
-P U
ro \
£ in
-P
in
UJ
ro
CD O)
c cn
•r- 5-
•— ro
CU -C
in u
ro \
CO in





CD
s:

•w-




5-
>i
C
CU
o
\
o
0
o
A
rH
te-


s_
>,

<*


rH
LO

rH

OO
«d-

r-i





O
^t-
1
00

oo






CM
«*
CM
1
O
OO
CM




CM

r*
rH




-P
'o
n:

»•
0)
S-
•r-
O.
E
LU







r-- r^
rH rH













CM CNJ
OO OO


in o
rH rH
* t
rH rH
1 1
 UO
1 1
LO CO

«* in






«*• o
rH 0
00 •sd-
1 1
CD LD
oo r-.
CM OO






rH
rH rH






E

r-*
S^
+J
•» r~~
rH rH






	 ti 	 s
O O
CM CM
^^ *\
rH rH
^*^ *^w^

in LO
CM CM


CTI
UD
m •
LD O
• 1
o r^-
<3-

o





o
00 00
• 1
OO «*•
,
LO






r~-~
LO
rH •*
cn i
CM rH
rH
OO




OO

~
rH CM










O
U
£
£_
- CM CM
00 •* Sf






rH 0
CNJ CM CM





i —
CU
QJ
-P
CO


CO
.
^










































.
o
CM

C
• 1—

C
o
-P
ro
>
i.
Ol
in
JD
O
•P
in

-------
          Cost per battery ($l,000/yr) = 560 - 13 (s/charge).
The slope of the above equation implies that a battery would incur a
cost of $13,000 per year to decrease charging emissions by 1 second
per charge.   Presumably, this additional cost would reflect additional
labor hours for inspecting, monitoring, training, supervision, repairs,
and other additional components of U.S. Steel's optimized charging
procedure.   The comparison of costs and average performance levels
suggest that the Bethlehem Steel and Armco batteries would incur costs
of approximately $100,000 per year for each battery to improve emission
control to the levels required at the Clairton Works.   This cost is
roughly one-third of the total costs already incurred at the two plants
to implement stage charging.   For the battery at Empire Coke,  annual
costs may increase by $195,000 per year (82 percent of the costs
already incurred) to achieve the levels at Clairton.  This estimate is
generated by multiplying the difference in average performance (24-9)
by $13,000 per year.
     The costs for any specific battery to improve charging emission
control will be determined by a myriad of interrelated factors and
battery-specific characteristics and could be quite different from the
incremental  cost estimated by the above correlation of cost and
performance.  However, the data supplied by industry do offer a reason-
able alternative for estimating costs on a nationwide basis if the
costs for the batteries in the data base are generally representative
of the industry as a whole.
8.3.2  Door Leak Control Costs
     The door leak control costs for Empire Coke are summarized in
Table 8-12.   Door leaks are controlled by the manual application of
luting material and, consequently, the major cost component is the
annual cost of labor.  A range of 50 to 100 percent of the total labor
is attributed to emission control for OSHA and environmental regulations
because the doors would be luted routinely as part of normal operation
even in the absence of regulations.
     The door control costs for Armco, Inc., are given in Table 8-13
and Table 8-14 and also reflect a range of 50 to 100 percent attributed
to OSHA and environmental regulations.  The end closure modifications
                                  8-19

-------
              TABLE 8-12.   DOOR CONTROL COSTS FOR EMPIRE COKE
Plant:   Empire Coke, Holt,  AL
Battery:   1, 2
Description:  60 ovens;  hand-luted doors;  160,600 Mg coke/yr
A.   Door costs

     Item                                          Range ($l,000/yr)


Inspect and clean doors                                 3.7-7.3
Labor to lute doors3'         .                         27.5-55.0
Labor to mix luting maternal '                          11.7-23.4
Cost of luting material '                                9.0-18.0
Monitoring                                              2.5
     Total                                             54.4-106.2
     $/oven-yr                                          910-1,800
     $/Mg                                              0.34-0.66

 Assumes 50 to 100 percent of plant total is attributable to baseline
 regulations.

 Assumes 90 percent of total luting costs are for doors and 10 percent
 for topside.

cAssumes 25 percent of total monitoring is for door leaks.
                                  8-20

-------
        TABLE 8-13.  DOOR CONTROL COSTS FOR ARMCO'S WILPUTTE BATTERY

Plant:  Armco, Inc., Middletown, OH
Battery:   1
Description:  76 4-m ovens, self-sealing doors, 448,000 Mg/yr
A.   Door costs - major capital items

     Eight additional spare doors at $16,000 each or $128,000 total

B.   Annual door costs

     Item                                               $l,000/yr

Door repairs3                                            150-295
Additional labor                                           120
Worker training                                            1.1
Monitoring       ,                                          0.7
Capital recovery                                            11
     Total                                               283-428
     $/oven-yr                                         3,700-5,600
     $/Mg                                               0.63-0.96

 Assumes 50 to 100 percent attributable to baseline regulations.
 Assumes 10 percent of plant total is for doors.
 Assumes 25 percent of plant total is for doors.
 Capital recovery factor = 0.0886 (20 years at 6.2 percent).
                                  8-21

-------
         TABLE 8-14.   DOOR CONTROL COSTS FOR ARMCO'S 6-m BATTERIES
Plant:   Armco, Inc., Middletown,  OH
Battery:   2, 3
Description:  6-m Carl-Still, 57  ovens each,  664,000 Mg/yr each.
A.    Door costs—major capital  items

     Item                                                $1,000

Door test program                                          300
Spare doors              .                                  400
End-closure modifications                                0-7,500
                                                       700-8,200

B.    Annual costs

     Item                                               $l,000/yr

Door repairsc                                            293-585
Additional labor                                          None
Worker training                                            2.2
Monitoring      f                                          1.4
Capital recovery                                         62-727
     Total                                              359-1,320
     $/oven-yr                                        3,100-11,600
     $/Mg                                               0.27-0.99

a!6 additional spare doors at $25,000 each.

 Assumes 0 to 100 percent attributable to baseline regulations.

 Assumes 50 to 100 percent attributable to baseline regulations.

 Assumes 10 percent of plant total is for doors.

eAssumes 25 percent of plant total is for doors.
 Capital recovery factor = 0.0886 (20 years at 6.2 percent).
                                  8-22

-------
for Armco's tall batteries ($7.5 million) may enable these batteries
to meet their existing door leak standard.  However, the modifications
and repairs would probably have been needed at some point in the near
future to maintain the operational status of the batteries.   The labor
for door repairs is a major component of the annual costs for all
three Armco batteries.  In addition, the costs for recovering the
capital investment to rebuild the end closure systems on the tall
batteries contributes significantly to their total annual cost.
     The door control costs for three batteries at Bethlehem's Sparrows
Point Plant are summarized in Table 8-15.  The major cost component is
the labor costs for repairs and a range of 50 to 100 percent is used
because some portion of this total cost is likely to be incurred even
in the absence of regulations.  The new Battery A apparently offers
economy of scale because the estimated control costs are lower for
this high capacity battery on the basis of coke production ($0.12 to
0.25/Mg coke).
     The total for the estimated control costs at the Clairton Works
are summarized in Table 8-16 for eight batteries.   The major cost
components used to derive the estimates are increased operating expense
from more frequent cycling through the door repair shop, more extensive
repairs and attention to detail in the door repair shop, jamb repairs,
and capital recovery for the installation of modified doors and seals.
U.S.  Steel's new Battery B also offers some economy of scale because
this high capacity battery has the lowest costs on the basis of coke
production ($0.33 to $0.36/Mg coke).
     The estimated control costs for door leaks are summarized in
Table 8-17 along with the current door leak standard applicable at
each battery.   The door leak control  costs do not appear to be related
to the regulations and vary widely by battery type and choice of
assumptions.   For example, Batteries 2 and 3 at Armco have undergone
an extensive rebuild of the end closure system, and, depending upon
whether this is completely attributed to emission control, the costs
                                  8-23

-------
             TABLE 8-15.   DOOR CONTROL COSTS FOR BETHLEHEM STEEL

Plant:   Bethlehem Steel,  Sparrows Point,  MD
Battery:   1, 11, A
Description:  Battery 1 has 63 3-m ovens, hand-luted doors, 273,000 Mg/yr.
              Battery 11  has 65 3.5-m ovens, self-sealing doors,
                365,000 Mg/yr.
              Battery A is a new 6-m, 80  ovens, self-sealing doors,
                1,148,000 Mg/yr

A.    Door costs—major capital  items
Item
Install NiCuTi door, seals
Original door seals
Spare doors
B. Annual door costs
Item
Operate, service all ovens and
door repair facility
Monitoring f
Capital recovery
Total
$/oven-yr
$/Mg

1
24
24
1
105-210
1.4
2.1
109-214
1,700-3,400
0.40-0.78
$1,000
11
190
26
216
$l,000/yr
11
108-215
1.4
19.0
128-235
2,000-3,600
0.35-0.64

A
0-220
32
32-252
A
132-264
1.4
2.8-22
136-287
1,700-3,600
0.12-0.25
a$l,450 each.
DAssumes 0 to 100 percent attributable to baseline regulations at $1,400
 each.
"Assumes 5 percent extra for spares at $4,000 each (from Table 8-2) to
 maintain spares.

 Assumes 50 to 100 percent is attributable to baseline regulations at
 $3,300/oven-yr.

^Assumes 25 percent of plant total is for doors.

Capital recovery factor = 0.0886 (20 years at 6.2 percent).
                                  8-24

-------
         TABLE 8-16.  DOOR CONTROL COSTS FOR U.S.  STEEL AT CLAIRTON

Plant:   U.S.  Steel, Clairton, PA
Description:   Batteries 1, 2, 7 have 64 ovens,  296,000 Mg/yr;  Battery 17
              has 61 ovens, 302,000 Mg/yr; Batteries 19, 20,  21 have 87
              ovens, 530,000 Mg/yr; Battery B is a new 6-m with 75 ovens
              and 804,000 Mg/yr.


                                              Battery number
Cost estimate
Annual cost ($l,000/yr)
$/oven-yr
$/Mg
1,2,7
230
3,600
0.78
17
283
4,600
0.94
19,20,21
311
3,600
0.59
B
269-287
3,600-3,800
0.33-0.36
                                  8-25

-------
                      TABLE 8-17.   DOOR COSTS SUMMARY
  Plant
Battery   $l,000/yr   $l,000/oven-yr
                              $/Mg
                         Baseline
                           PLD
Armco,
  Middletown
Empire, Holt
U.S.  Steel,
  Clairton
U.S.  Steel,
  Clairton
U.S.  Steel,
  Clairton
Armco,
  Middletown
Bethlehem,
  Sparrows Pt.
Bethlehem,
  Sparrows Pt.
Bethlehem,
  Sparrows Pt.
U.S.  Steel,
  Clairton
U.S.  Steel,
  Clairton
 2, 3

 1, 2
 1, 2, 7

 17

 19, 21

 1

 1

 11

 A

 20

 B
180-660

54.4-106
  230

  283

  311

283-428

109-214

128-235

136-287

  311

269-287
3.1-11.6

0.9-1.8
  3.6

  4.6

  3.6

3.7-5.6

1.7-3.4

2.0-3.6

1.7-3.6

  3.6

3.6-3.8
0.27-0.99

0.34-0.66
   0.78

   0.94

   0.59

0.63-0.96

0.40-0.78

0.35-0.64

0.12-0.25

   0.59

0.33-0.36
16

15
11.6

11.6

11.1

11.3

10

10

 5

 5

 5
                                  8-26

-------
vary from $3,100 to $ll,600/yr per oven.  The hand-luted doors at
Empire Coke and at Bethlehem Steel (Battery 1) appear to show higher
costs for a tighter baseline (15 to 10 percent leaking door [PLD])
with an increased cost of $800 to $l,600/yr per oven.   The remaining
batteries, with self-sealing doors, are grouped in the range of 5 to
11.6 PLD with costs that are similar ($1,700 to $5,600/yr per oven).
     Because the cost data in Table 8-17 show no obvious correlation
between cost and performance level, an attempt was made to place
bounds on door control costs by generating a lower bound and an upper
bound estimate.  A procedure for generating the lower bound estimate
is to examine the improvement in emission control at each battery that
was obtained by incurring the total cost.  For example, the total cost
for each battery in Table 8-17 represents the cost to improve control
from some "uncontrolled" level in the absence of regulations to the
current control level.  This "uncontrolled" level is difficult to
define but observations of door leakage on poorly controlled batteries
have been observed to range from 40 to 70 PLD.  Data from U.S. Steel's
Clairton Works showed a range of 38 to 45 percent leaking ovens11
before the implementation of their door leak control program in response
to their negotiated consent decree.  An average starting point of
55 PLD is assumed for all of the batteries (except U.S. Steel where
the specific 45 percent leaking ovens was observed), and the endpoint
is the estimated level of current performance.
     In Table 8-18, the difference between previous and current control
in PLD is divided into the annual cost to provide an incremental cost
for an improvement of 1 PLD.   For example, the estimated improvement
in control at Armco's Battery 1 is 44 PLD (55-11) at a total cost per
oven of $3,700 to $5,600 per year.  The annual cost per oven ranges
from $84 to $127 for an improvement of 1 PLD.   This approach assumes
that the cost is linearly related to the PLD reduction and yields
estimates for a 1-PLD improvement that range (per oven) from $34 to
$62/yr for hand-luted doors and $61 to $210/yr for self-sealing doors.
     The approach described above is likely to bound costs on the
lower end because control costs are more likely to be exponentially
                                  8-27

-------
                  TABLE 8-18.  DOOR COSTS AND PERFORMANCE
Company
Armco, Inc.
Armco, Inc.
Empire
Bethlehem
U.S. Steel
Battery
1
2,3
l,2d
ld
11
A
All
Annual Cost
($l,000/oven-yr)
3.7-5.6
3.1-11.6
0.9-1.8
1.7-3.4
2.0-3.6
1.7-3.6
3.6-4.6
PLD
From
55a
55C
55a
55^
55a
55a
459
To
llb
20C
15b
14ef
14f
11T
59
$/oven-yr PLD
Midrange
106
210
34
62
69
61
103
Range
84-127
89-331
23-45
41-83
49-88
39-82
90-115
Assumed PLD level prior to control program.
Assumed to be meeting existing State regulation.
From Armco trip report, Reference 5.
Hand-luted doors.
Assumed to be the same as Battery 11.
From Bethlehem Steel trip report, Reference 4.
     provided by U.S. Steel in Reference 11.
                                   8-28

-------
related to PLD.  The initial control of emissions generally provides
the greatest and easiest reduction.  Subsequent improvements may
require additional equipment or redoubling of labor efforts to obtain
a small increment of improvement.  For example, if the costs are the
same to improve control from 0 to 90 percent, 90 to 99 percent, and 99
to 99.9 percent, then a linear assumption between cost and control
performance from the first  increment (0 to 90 percent) would tend to
understate the costs for subsequent incremental improvements in control.
     The cost data provided by U.S. Steel's Clairton Works are used to
generate an upper bound estimate of door leak control costs.  The
technology and achievable performance levels at Clairton have been
demonstrated and well documented.  From Table 8-17, the overall average
cost at Clairton is estimated as $3,700/oven-yr for their current door
leak control program.  The  improvement in control associated with this
cost was reported as roughly 42 percent leaking ovens to 5 PLD.
Currently, the most lenient regulations for door leaks are 15 and
16 PLD.  For the worst case, assume that if these batteries must
improve control from 15 to 10 PLD, then they will incur costs equal to
the full costs incurred at Clairton.  The assumption in effect is that
there may be existing batteries that do not have the modified door
seals and the extensive monitoring, inspection, and repair program
implemented at Clairton, and that the entire program must be implemented
to achieve the reduction from 15 to 10 PLD.   The cost for this incremental
improvement is about $740/oven-yr-PLD ($3,700/oven-yr divided by
5 PLD).
     Many batteries may merely supplement their existing door leak
control programs.   For these batteries, the range from Table 8-18
($34-210/oven-yr-PLD) may be more appropriate, particularly if improve-
ment is obtained by increasing labor with no significant capital
investment.   An overall range for door leak control costs of $61/oven-
yr-PLD (average from Table 8-18) to $740/oven-yr-PLD is recommended
for self-sealing doors.
     The estimate of $740/oven-yr for an improvement of 1 PLD is
believed to be a reasonable upper bound estimate because all existing
batteries have already improved control well beyond the 45 percent
                                  8-29

-------
leaking ovens at which the Clairton Works  started.   Many aspects  of
Clairton's program were likely implemented by the other batteries to
improve control to their current levels (all  with regulations  of
16 PLD or less).  However, the upper bound estimate assumes that  the
other batteries must implement the entire  program of U.S.  Steel  to
achieve the same performance levels obtained at the Clairton Works.
8.3.3  Topside Leak Control Costs
     The control of leaking lids and offtakes on the topside of  the
battery is accomplished primarily by the manual application of a
luting mixture.  In addition,  damaged offtakes and slip joints may
require repair or replacement.  The topside leak control costs for
Empire's coke plant are summarized in Table 8-19 and indicate that
luting labor is the primary component.   A range of 50 to 100 percent
of the luting labor is attributed to emission control regulations
because the lids and offtakes  would be luted to some extent as a part
of routine battery operation to prevent fires and loss of valuable
by-products.
     The estimated topside leak control costs for Armco's batteries
are given in Table 8-20.  Topside labor is a major component of the
costs for both lids and offtakes.  In addition, a capital expenditure
of $1.1 million is included for the two tall batteries for replacement
of the offtake systems.  The tall batteries appear to offer an economy
of scale even with the major capital expenditure to replace offtakes
with costs of $0.03 to $0.09/Mg coke for lids and offtakes.
     The control costs for topside leaks at Bethlehem's batteries are
given in Table 8-21.  Labor again is a primary component of the annual
cost.  Although offtakes were  not replaced, the cost to maintain the
slip joint seals on the offtake system is another major cost component.
Bethlehem's new Battery A also provides an economy of scale with
topside control costs that range from $0.03 to $0.06/Mg coke.
     The total costs estimated for topside leak control at U.S.  Steel's
Clairton Works are provided in Table 8-22.  The major annual cost
items included in the estimate are cleaning and luting labor and
maintenance of slip joint seals.  Capital  costs were also estimated
for replacing the offtake systems.  As observed with the other tall
                                  8-30

-------
          TABLE 8-19.  TOPSIDE LEAK CONTROL COSTS FOR EMPIRE COKE

Plant:   Empire Coke, Holt, AL
Battery:   1, 2
Description:  60 ovens; 3 lids and 1 offtake per oven; 160,600 Mg/yr
A.    Topside leak control costs'
                                                       $1.000/yr
Item
Luting labor
Maintain slip joint seal .
Inspect, clean, repair lids .
Labor to mix luting material '
Cost of luting material '
Monitoring
Total
$/oven-yr
$/Mg
Lids
17.3-34.5
-
1.6-3.1
1.0-2.0
.8-1.5
1.3
22-42.4
370-710
0.14-0.26
Offtakes
5.8-11.5
1.3
-
0.3-0.7
0.3-0.5
1.3
9.0-15.3
150-260
0.06-0.10

 (3 lids and 1 offtake).

 Assumes 50 to 100 percent is attributable to baseline regulations.
"Assumes 10 percent of plant total is for topside leak control  (90 percent
 for luted doors).

 Assumes 25 percent of plant total is for topside leaks and divided
 evenly between lids and offtakes (one traverse for each).
                                  8-31

-------
          TABLE 8-20.   TOPSIDE LEAK CONTROL COSTS FOR ARMCO,  INC.

Plant:   Armco, Inc., Middletown,  OH
Battery:   1, 2, 3
Description:  (1) 76 ovens, 3 lids and 2 offtakes per oven, 448,000 Mg/yr;
              (2, 3) 57 ovens each, 6-m height, 4 lids and 2 offtakes per
                oven,  664,000 Mg/yr each.
A.    Topside capital costs'
     Item
                                                   $1,000
Lids   Offtakes
                                                               2. 3
Lids   Offtakes
Replace offtake system
Replace standpipe caps
Develop procedures
Total
B. Annual costs3


Item
Additional topside laborc
Monitoring ,'
Training '
Capital recovery
Total
$/oven-yr
$/Mg
-
-
2.5
2.5



Lids
72
0.4
1.4
0.2
74
970
0.17
-
11.7
1.7
13.4

$1
1
Offtakes
48
0.4
1.4
1.2
51
670
0.11
-
-
5.7
5.7

,000/yr

Lids
40
0.7
2.1
0.5
43.3
380
0.03
1,100
-
2.8
1,103


2, 3
Offtakes
20
0.7
2.1
98
120.8
1,100
0.09
 Items applicable to both lids and offtakes are split 60/40 percent for
 Battery 1 and 67/33 percent for Batteries 2 and 3.

DAssumes 25 percent of plant total is for topside leaks.

"Assumes one-half of additional labor is for topside leaks and one-half
 for charging.
•j
 Training and monitoring costs are divided evenly between lids and
 offtakes.
"Capital recovery factor = 0.0886 (20 years at 6.2 percent).
                                  8-32

-------



















_J
LU
LLJ
1 —
CO
s:
LU
in
LU
_1
31

LU
CQ

C£
o
u.

co
(—
co
o
o

	 1
o
Q£
h-

o
C_3

^
,
>o
i- » coo
\ 2: «
CO 00
2: o «t
O rH £-
O O - >,
O »rH \
o m o
-ID - O
co co c o
r^ co
CM - > rH
CO W
* CO
C > S-
CO O CO
> Q.
O S-
cu in
I- Q. CO
CO -^
O- in 03
CO -P
CU ^ 4-
^ 03 4-
03 -P O
-P 4-
4- 4- CM
O 4- O
2! 0 T3
CM C
rH 03
•P -a
c ~a c in
•r- C 03 T3
O «3 •!-
Q. in i —
in -a
in "O •!- ^-
5 •<- r- 03
o r- in
S- «3- in +J











T3

_l















rH
rH











in
CO
.^
03
-P
4-
4-
O






in
T3
•r-
_i















rH











in
CU
^
03
-P
4-
4-
O





in
•a
•r"
_J



s- *t c in
tO " CU O
a. - m > u
co in c o
C CU f—
CO > E O
r— > O 1 J-
CO O CD -P
co in c
-P CO CD O O
CO CD 00 U
«C X-N
e /-M-H /— v .V
CO -rH rH  CO
0) rH i—
i —
.C - .. QJ
•P rH C T3
CO O •!-
CQ •<- in E
.. -P Q. CO
>> Q. 0 -P
• • t- T- (— i— I
•P CO S-
C -P CJ
03 -P in
r— 03 CO
Q- CO Q  O
1 1 1
r^ o cn
. O 0
CM in •
CO O





cn LT>
rH CO r-.
1 1 •
CD 00 0
• rH
cn

r~« m
. O rH
<*• r-^ .
in oo o
i i i
CO O 00
• m o
oo *i- •
CM 0





CD
Is-.
i r--
^ i .
o
00
CO
r^. o oo
• O CM
CD CM •
r-> -o
1 rH 1
rH 1 <3-
. 0 rH
cn CM •
CO CD CD




u

03
CU
in

-P
C
• r-
o
JD T-,
cn s-
C Q. >,
•f- •!- |
•P r-=D C
3 10 cn r— cu
r— c 03 > cn
C -r- +J 0 2:
CU -r- S- 0 \ \
TS 03 O h- *9-<**
•r- 4-> 4->
in c -i-
O-'i- c
O 03 O
(— 2: 2:
-P
C.
cu
u
S-
CU T3
a. c
03
CO
CO C
^N. co cu
r^ c cu
CD 5
C -P
TJ 03 CO
c cu .a
03 r—
u ~a
rH O>
•a
>> CO -r-
i. C >
CO ••- -r-
P -P T3
P • 0
03 WO) >,
CO C Q. r—
O in c
J- T- C CU
O -P •!- >
4-03 CU
r— S-
-P 3 o in
c cn 4- T-
cu cu
(J S- r— +J
i- 03 in
CU CU -P 0
O- CO U
•r- +J
Or- T3
00 CU CU C
~-v in ^= 03
0 03 P
CM .a in
4- ^
P O O 03
•r- P CU
•— -p I—
Q- CU C
in i — cu co
jQ U T3
CU 03 S- -r-
S- -P CU in
03 3 o- a.
.a o
in •!- o -P
cu s- in
-^ P S-
03 -P O O
-p 03 -P 4-
4-
4- in in in
O •<- CM -i-

-a -P in r-
c in -i- (O
03 O -P
u cu o
in u -P
•a 4- c
•i- O O3 -P
r— C C
•P CU 03
JZ C +J i—
-P • CU C O.
O —
03 jO 0 -O
8-33

-------






















"21.
o
r-
or
_j
o

r-


— i
r~
^


ce
0


en
h-

O


— I
o

1—
z:
0




UJ
_J

UJ
o
1 — 1
t/1

Q


































s- .
>> t-
*c >,

•-E cn
S- E

^ 0 0
O)O O
s: -o
tn -
Q ro *g-
Q if) O
O GO

O C -
O QJ C
ro > a>
o >
o
C t-
a»  Q. a;
0 Q.
in
j £- O> W

Q- TO -^
-P TO
t/i V— -P
at "+- s—
j^ O **-
t— i (TJ O
1 +-> CM
«4- esj

CSJ
C\j
1
CO

LU
	 1
CO
 ID
1 C * O >
O in O
*j c r-
s- 

| ro cn o ^^ vo r— rH C_) - rt CM UT> r-4 LO f^ j -CM a; -CM CM co .P CM •• 1 • t/) CM C 0 • • 4^ >> o. +J QJ S- C +J U i — m QJ ! Q. cn a CD I- a» ro CD rH CM j-' •S +j ro CD QJ 03 ij o TJ ^ ^ m V- o in •o — F CNJ >s a; TO CO tn OJ 03 <4— l^_ O T3 _J 0) fO E -t-5 in OJ •»-* 0 CO in m i ro cn CM CM ro 00 1— 1 CO CO rH I cn CM CM I 00 rH CO CO I_Q cn 0 CO rH CO rH CO f~*. rH 1 in CO CO >, o o o r-\ ^ ul O U ro 3 C C <£ rH 0 O rH 1 in o ID CM rH CM CM CSJ i rH ID CM 1 rH CM CM CM 1 rH rH CO CM 1 CM CM CM 1 rH ^ 1 C


-------
batteries, U.S. Steel's Battery B requires the lowest control cost on
a coke production basis at a level of $0.01 to $0.09/Mg coke.  The
U.S. Steel costs appear to be higher than the costs at other batteries
and the major difference is in the estimated amount of topside labor.
The replacement of offtakes also increases U.S.  Steel's cost, but the
offtakes were also replaced at Armco's tall batteries.
     The control costs for lid leaks are summarized in Table 8-23 for
all batteries.  Current baseline regulations for percent leaking lids
(PLL) are also given.  The control costs on a per oven basis appear to
increase as the stringency of the baseline regulation increases.   The
midrange costs for each plant on a per oven basis from Table 8-23 are
listed below:
                                                           Average
     Plant               $/Oven-yr      Baseline PLL         PLL
     Empire Coke             540               5             3.5
     Armco, Inc.             680               5             3.5
     Bethlehem Steel         825               3             1.9
     U.S. Steel            1,800              1-2          0.6-1.1
The average PLL given above is the estimated average performance that
would be required to meet the current regulation (baseline PLL) a high
percentage of the time.  A linear regression of cost versus average
PLL yielded the equation:
           Cost per oven ($/yr) = 1,930 - 400 (Average PLL).
The slope of the above equation suggests that an annual expenditure of
$400 per oven is required to obtain a decrease of 1 PLL.
     The cost estimate derived above can be applied to a typical
battery to determine if the $400/oven-yr estimate is reasonable in
terms of luting labor.   Consider a 60-oven battery with 4 lids per
oven and an 18-hour cycle time.   For a 1-percent improvement in PLL,
an additional 2.4 lid leaks (4 x 60 x 0.01) per cycle must be located
and luted.   This is equal to about 3.2 additional lid leaks per day
(2.4 x 24 -r 18) or 1,170 additional leaks per year to be luted.  The
incremental cost is estimated as $400/oven-yr or $24,000/yr for a
                                  8-35

-------










OO
1—
00
o
(_ }
i
o
or
h-
o

<£
LU
— 1
O
l — i
-*
OO
CM
00
LU
	 I
GO
*^
1—

















CU
c
•r-
i — 	 1
CU —1
l/> Q.
03
CQ


D)
s:

^
i
•o
.r—
r—
\

o
o
0
o
	 1
p»l

s_
\
o
o
o
i-H
>,
cu
-p
-P
03
CQ




-P
C
O3
f—
CL.




LO LO



CM
• r^
0 rH
rH
O
O
^~
CM LO
1 CM
O 00
CM
|



rH
. l~^
O CD
1
oo
.
O


<*
CM
^«
1 **J"
CM r^>
CM

CM

«*
i-H rH

CU •
-^ (J
o c
C_) 1— 1

oi -
S- O
•r- U
Q. E
S S-
LU  00 OO O

OrHO OOOOOOO
          I    I   I   I   I   I   I

         rH O O OO rH rH O

         OOOOOOO
          OOOOOOO
          1   I   I   I   I   I    I
         OOOOOOO
         l£> OO rH Lf> 00 00 OO
         rH rH rH 00 CM CM i-H
             oo in
          CM CD 00 f~- CM CM O

OCDOOrHOOCMCMCMrH
           I   I   I   I   I   I   I

          ID LT) -3-
           .   .   . t-H r-H rH O
          CO O O
           .   .   . 00 00 00
          ID ^ oo r~~ oo oo oo

        •   I   I   I   I   I    I   I

       CM   .   .   .   • CD CD oo
          CD CM -3- 00
          OO OO 00 00
       ro
             i-H        O rH
rHrHCMrHi—I
-------
60-oven battery.   The cost of additional  luting of lid leaks is about
$20/1 id leak, or about 48 minutes per leak at $25/hour.   The estimate
of 48 minutes appears reasonable if it includes a distribution of
total plant labor (e.g., monitoring, maintenance, and supervision).
The estimate of $400/oven-yr for a 1-PLL improvement in control does
not appear to understate costs in terms of the labor required to lute
a leaking lid.
     Control costs for offtake leaks are summarized in Table 8-24 and
show that the highest estimated costs have been incurred at U.S.
Steel's Clairton Works.   The average midrange costs per oven are given
below for each plant:
         Plant           $/Oven-yr    Baseline PLO    Average PLO
     Empire Coke             205           10             6.5
     Armco, Inc.              885           10             6.5
     Bethlehem Steel         665           10             6.5
     U.S.  Steel            2,400           4-5          2.2-2.9
The higher estimated costs for U.S. Steel reflect an estimate of more
manpower for topside labor and the replacement of offtake systems.
The average PLO given above represents the estimated average performance
that would be required to meet the current standard (baseline) a high
percentage of the time.   The incremental  cost difference between
plants ranges from $1,500 to $2,200/oven-yr for an improvement of
about 3.6 units in average PLO (6.5 to 2.9).   This estimate yields  a
range of $420 to $610/oven-yr for each percent decrease in PLO.  This
difference in cost between batteries implies  that Armco1s and Bethlehem's
costs could increase by a factor of 3 to 4 to attain the control level
in effect at U.S.  Steel's plant.   Because of  the marginal improvement
required (6.5 to 2.9 PLO), the range of $420  to $610/oven-yr is likely
an upper bound estimate of cost for control of offtake leaks.
     The application of the cost function to  a 60-oven battery with
double collecting mains yields an annualized  cost of $25,000 to $37,000/yr
for a 1-percent improvement in PLO.  The 1-percent improvement represents
about 1.6 additional offtake leaks controlled per day or 584 offtake
leaks per year.   The cost corresponds to $43  to $63/offtake leak or
                                  8-37

-------
       CD
       C
       d)
       fO
       co
           01
                o       LD  oo un o co  co
                i—I       rH  rH O CO *d"  ^J"
                  . rH CTl	rH
                OrHOOOOOOOO
                 I     ••  I   I   I   I   I    I

                O       O  i—f O <3" 00  CO

                O       O  O CD O CD  O
                                    O  CD O
                                    CD  O O
CO         -<7>ooo
Qi         O     rH       -^CMrHi-Hi-lrH
O
C_J
LU



LU
           I
           C
          O
          o
          o
                 CM
r^.     CD
oo CD r^ oo CD
                 OCDrHOi—(OCSJCMCMrH
                 I    ••  I   I   I   I   I    I
                 LnOr-tLncOOOCSJrHr-(O
                 i—i       ^" LO co
                            .   .   . CM  CM CM
                 o       o o o
CM

00
00
«=c
          o
          o
          o
                 co       r*- <* r-»
                            .  .   . O  CT> CTi
                 LO       ^ i^- o oo  CM CM
                 i—I O ^~ LO CD CD rH  Csl CM 00
                 I    ••  I    I   I   I   I   I

                   . LT> CD   •  •   . CO  00 00
                        00 <* O rH  i-H rH
                          CM CO CO
          i-
          0)

          -P
          (0
          co
                 CM
                       CO
                         -i-H        O i-H
                 rHi-HCMrHrH
 
                                                       8-38

-------
1.7 to 2.5 person-hours per leak at a labor rate of $25/hr.   The
estimated cost appears to be high in terms of labor hours because
minutes instead of hours should be required to lute a leak.   However,
capital costs for replacing offtakes and repairing slip joint seals
were significant elements in U.S. Steel's total costs.   Consequently,
the costs for offtake leak control include the annualized cost of
capital in addition to the simple labor cost of luting and reluting.
The increased cost of control is likely to be lower than stated above
for most batteries because few of the batteries are likely to replace
entire offtake systems to improve average performance from 6.5 to
2.9 PLO.
8.4  MODEL BATTERY AND NATIONWIDE COSTS
8.4.1  Model Battery Costs
     The cost functions developed in Section 8.3 are applied in this
section to model batteries to illustrate the procedure for estimating
nationwide costs.  The procedure is based on battery-specific character-
istics, such as the number of doors and ovens, and on the required
improvement in control.   For the example case of model  batteries, the
costs are based on improving control to the levels demonstrated by CF&I
and U.S. Steel's Clairton Works as discussed in Chapter 4.   The demon-
strated control levels are a log average limit of 16 seconds per
charge, which requires an average performance of 10 seconds  per charge,
a limit of 10 PLD with an average performance of 7 PLD, a limit of
3 PLL with an average performance of 1.8 PLL, and a limit of 6 PLO
with an average of 3.5 PLO.   These control  levels were chosen as
Regulatory Alternative II.
     Costs for the model batteries that were developed in Chapter 6
are presented in Table 8-25.   The baseline limits that were  chosen for
the illustration represent the most lenient regulations currently in
effect and,  consequently, the costs represent estimates for  the most
affected population of batteries.  Many batteries are currently subject
to emission  limits that are  roughly equal to or more stringent than
the levels demonstrated by CF&I and U.S.  Steel.   Batteries with current
limits equal  to or less  than those demonstrated by the  recommended
                                  8-39

-------












t— H
I — 1
UJ
h^
•=c

oi
LU
h-

«=c

2>-
oi
O

<£
	 1
ID
LU
ce
oi
0
u_
i/i
1
t
00
0
o

*2r-
C£.
LU
h-
I—

^
OJ
-P
-p
ro
ca

, —
O)
-a
o
2:


CM

S-
0)
^_>
^_)
m
CQ

i —
01
-a
o



rH
^
t-
O)
4_>
-P
ro
co

f—
0>
-a
o
s:






























o
.
O <£> i-H
co r^-
O

o
r**^
p**-







0
,
O ^- C\J
O UD
O
r>
^~
^d~
CO







LO
.
O ro *£)
O CO
O

tj3
0
rH














/ 	 s
s- in
>> c
\ 0>
O5 >
2: o

"~"EV
>>s— ' O
-p
•r— -P £-
U -C O)
ro cnja
Q."- e
ro cu -3
0 31 Z

^
rH s-^s~^
i 	 • • LO LO
r». CM • •
rH rH OO UD
°*~s "^— ^ **~s ^— ^
LO IO LO O
CM rH i— 1








,— ^
OO ' — >' — *
, — . • LO LO
f**- rH • •
rH rH 00 l£>
**^s ^~~s **^t *v/
LO LO LO ^~~>
CM rH rH








/^x
OO **"^ ^^
, 	 	 • LO LO
^. | — | . .
rH rH OO VD
^^ ^^ ^** ^**
LO LO LO CD
CM rH i-H


s**^
i-
^^
^v^
O
o
o
rH
" — '
.a -P
a> m
O) O
i- U
ro
-C T3
U Q)
N
t- •!-
CU i—
Q. ro

ro m c
CJ T3 C
c c ro
•i- O
r- (J CD — 1 O r—
ai cu — i — i — i re
in oo Q_ o_ Q- -P
ro o
CQ 1—




rH LO 00 O O
CTl «3- **• rH OO
rH rH CM













i-H 1 — CM ID O
O"> O «^ CT) O
rH CM













rH CM •=!• UD LO
CT> ID CM LO rH
rH

















/*~»,
E
^3
E
•r—
X
ro
O5 V) S
C CD ^_x
•r- ^
CT in ro in
t- s- in -P s_
ro o -a t- o
-C 0 ••- M- 0
0 Q — 1 0 Q


CM
CM cn i£>
rH rH CD
000











CD rH CM
CM 00 rH
O CD O















in oo oo
CO LO CM

CD O O

























O)
cr
O) m
s- S- in
ro o TJ
-C 0 -r-
C_3 Q 	 1



<3" O
I-H 00
O O











00 00
CM LO
O O















00 00
LO O

O rH



















E
^3
E
•r-
X
ro
i/) E
_aj ^
ro m
•p i-
<4- O
M- 0
o a


















•
CD
m
-C
-p
c
01
&_
ro
o.

c
•r-
a>
u
d
ro
E
J_
O
<^ 	


•P
ro

-p
in
a>

>»
~°
TI
ai
U
0

r—
0

-p
•1 —
E
•r—
	 1
ro



in
-P
r—
in
ai
O
in
ro
a>
0)
s_
ro
u
^\
in
r^
i— I
i,_
O

0)
O)
ro
S-
0)

ro
a>
^ in
L f—
r^ •—
o
•i —
. in
O 3
CM i —
U
c x
•r- CD
C O
0 C

-P .C
ro -P
s- 5
O)
in 01
-a cn
0 S-
ro
u
Ol^x
c in
• i —
TJ CM
3 00
'(J M-
X O
01
-P
a> T-
01 E
J- -r-
ro r-
_f—
u ro

in E
o
LO i-
C\J M-
-Q
8-40

-------
technology would not incur any significant costs from a national
standard based on the technology.
     The cost estimates in Table 8-25 were derived by applying the
cost functions developed in Section 8.2 to the specific characteristics
of the model batteries.  For charging, an estimate of $13,000/yr was
derived for each battery as the cost to obtain an emission control
improvement of 1 second per charge.  The baseline limit for the example
is 25 seconds per charge (excluding one observation in 20) with an
estimated average of 17 seconds per charge.   The technology demonstrated
a level of 16 seconds per charge (log basis) with an average performance
of 10 seconds per charge.   The cost estimate is based on an average
improvement of 7 seconds per charge (17 to 10) times $13,000/yr,  or
$91,000/yr for each battery.
     The costs for doors were estimated as $61 to $740/oven-yr for a
decrease of 1 PLD.  A midrange value of $400/oven-yr was used in the
example and was multiplied by 4.3 PLD (11.3 to 7) and the number of
ovens for Model Batteries 1 and 2.   The most lenient regulation for
tall batteries is a limit of 16 PLD with an estimated average of
12.1 PLD.   The door control costs for Model  Battery 3 are estimated as
5.1 (12.1 to 7) times $400/oven-yr times the number of ovens.
     The estimated cost to improve lid leak control by 1 PLL is
$400/oven-yr.  The average improvement used in the example is 1.7 PLL
(3.5 to 1.8), which is multiplied by $400/oven-yr and the appropriate
number of ovens to estimate the total annual cost.
     A midrange value of $515/oven-yr ($420 to $610/oven-yr) was used
for offtakes for an improvement of 1 PLO.   The average improvement in
the example is 3 PLO (6.5 to 3.5),  which is multiplied by $515/oven-yr
and the number of ovens to estimate the total annual cost.
     A worst-case or maximum cost was derived for the case of door
leak controls and was based on the upper end of the estimated range
($740/oven-yr).  The maximum case assumes that each of the model
batteries must implement the entire door leak control program that is
used at U.S. Steel's Clairton Works.   No credit for costs incurred to
reach their current baseline are given to the batteries under this
assumption.
                                  8-41

-------
     The costs expressed in $/Mg of coke indicate that the taller and
higher capacity batteries are less affected in terms of production
costs.  Most new batteries are the 6-meter type represented by Model
Battery 3.   Consequently, the lower costs (in $/Mg coke) shown for
Model Battery 3 are probably representative of costs that might be
incurred for new sources.
     Reasonable estimates of capital costs for improved emission
control generally require identification of specific equipment and its
installed costs.  The industry-supplied data indicated that the control
of coke oven emissions was a labor-intensive effort.  Major equipment
modifications have already been made by most batteries as a result of
OSHA regulations, State regulations, and consent decrees.   Dramatic
improvements in control performance have been observed in recent
years.  Consequently, the effect of a national standard based on the
demonstrated emission limits (Regulatory Alternative II) will require
only a nominal improvement in control for most batteries.   This improve-
ment in control is small compared to the improvement made in recent
years from a poorly controlled status to the current baseline status.
The required improvement may be accomplished by most batteries through
additional labor and changes in work practices.
     No capital costs are estimated for charging.  The basic technology
of some form of stage charging requires a significant capital invest-
ment; however, this cost is entirely attributable to the baseline.
The improvement in charging emission control is expected to result
from optimization of procedures, additional inspection and cleaning of
equipment, more supervision, and visible emission monitoring to trouble-
shoot emission control difficulties.  These items are treated as
recurring annual expenses.  A few batteries may choose to invest
capital in some minor items, such as new steam nozzles or some modified
design.  However, these capital expenses will be very specific to a
battery's characteristics and are expected to be minor when compared
to the major expenditures already incurred for stage charging.
     Some batteries may  incur capital costs for door leak control if
they choose to modify their doors and seals as suggested by U.S. Steel.
                                  8-42

-------
The data supplied by industry indicated that a major expense for door
leak control was labor.  These labor costs resulted from more frequent
cycling through the door shop, more extensive repairs to doors,
maintaining seals within tight tolerances, inspection and cleaning of
seals and jambs, and visible emission monitoring.  For the purpose of
estimating nationwide capital costs, the annualized cost of capital
was estimated as 10 percent of the total annual cost, or $40/oven-yr
for the midrange value.  The capital was annualized over 20 years at
6.2 percent with a capital recovery factor of 0.0886.  The capital
investment is then estimated as $450/oven (40^-0.0886) for an
improvement of 1 PLD.  Applying this cost function to the model
batteries yields estimated capital costs of $70,000, $120,000, and
$160,000, respectively, for the three models.  The worst case or
maximum approach for door leaks yields an estimated capital requirement
of $835/oven for a 1-PLD improvement.  The capital requirements for
the models on this worst-case basis are $130,000, $220,000, and $300,000,
respectively.
     Lid leak control is accomplished almost entirely by the labor
effort to lute, locate leaks, and relute as necessary.   No capital
costs are estimated for this emission point.
     The cost data for offtake leaks indicated that some batteries
incurred a cost to replace offtakes or slip joint seals to improve
emission control.   However, the major expenses were primarily recurring
labor items for inspecting, cleaning, repairing, luting, reluting, and
monitoring.   To estimate nationwide capital costs, the annualized cost
of capital was again assumed to be 10 percent of the annual costs.
This assumption yields an estimate of $350/oven as the capital cost to
improve offtake leaks by 1 PLO.   For the models, the capital costs are
then estimated as $38,000, $65,000, and $75,000, respectively.
8.4.2  Nationwide Costs
     Nationwide costs were estimated on a battery-specific basis using
the procedure illustrated for the model batteries.  Data were available
for each battery for the number of ovens and current baseline require-
ments in terms of visible emission limits.   The baseline limits were
                                 8-43

-------
converted to an average performance level.   The average performance
levels of the demonstrated technology (10 s/charge,  7 PLD,  1.8 PLL,
3.5 PLO) were used with the baseline average to estimate the average
required improvement in emission control.   The average improvement,
the cost function, and the number of ovens at each battery  were then
used to estimate total cost.   (A listing of batteries with  their
current baseline is given in Chapter 3.)
     The nationwide costs of Regulatory Alternative II for  improved
charging emission control from the analysis described above is estimated
as $4.7 million per year.  For door leaks, the nationwide cost is
estimated as $5.4 million/yr with a potential capital requirement of
$6 million.   The annualized cost for lid leaks was estimated as $1.6
million/yr.   For offtakes, the nationwide cost was estimated as $7.6
million/yr with a potential capital requirement of $5.2 million.
     A maximum cost was estimated for door leaks from the upper bound
of the cost functions.  This approach assumes that the affected batteries
may be required to implement essentially the entire door leak control
program of U.S. Steel.  On a nationwide basis, these maximum door
control costs were estimated as $10 million/yr with a capital require-
ment of $11 million.
     The cost analysis approach used in this chapter contains a signifi-
cant element of uncertainty, especially when the cost functions are
applied to any specific battery.  An improved approach would be to
examine each individual battery, estimate the specific items that
might be required to improve control, and then generate an estimate
for each individual battery in the industry.  Such an approach would
require extensive resources, and because assumptions at each battery
would still  be required, the approach would still yield a significant
amount of uncertainty in the final estimate.
     The data used in this analysis were supplied by industry and
should represent actual costs that have been incurred.  If the batteries
in the data base and their cost information are generally representative
of the industry, then the nationwide cost estimates are likely to be
representative on an aggregated basis.   However, the costs for any
                                  8-44

-------
specific battery may be higher or lower than the cost estimated from
the cost functions.
     Several factors indicate that the cost estimates for door leak
and offtake leak controls may be high.  The estimated requirements for
these sources could be less than estimated because only marginal
improvements in control performance are required.  These improvements
may be accomplished at many batteries without the capital investment
and intensive labor effort on which this cost analysis is based.  In
addition, the nationwide cost estimates include totals for batteries
that are currently shut down and may not be restarted.  Consequently,
the estimated nationwide costs probably do not understate costs and
may be somewhat higher than the costs that will actually be incurred.
Each plant would likely decide on the most cost-effective method to
improve control at each specific battery.   The plant's approach may be
accomplished by techniques and redistribution of labor efforts that
are less costly than those estimated by the cost functions.
8.5  OTHER COST CONSIDERATIONS
     Coke plants are subject to a number of regulations that relate to
emissions and emission control.  OSHA has developed occupational
health rules that restrict personal exposure of workers to 10 ppm of
benzene and to 150 (jg/m3 of benzene soluble organics, both on a time-
weighted basis.  The coke plants also have existing regulations for
control of air emissions from charging, doors, lids, offtakes, pushing,
quenching, and battery stacks.  In addition, water regulations are in
effect for the effluent discharged from wastewater treatment plants.
The costs for OSHA regulations and other air and water regulations
have been included in the baseline costs,  which will be analyzed in
Chapter 9, Economic Impact.
8.6  REFERENCES
1.    Section 114 Questionnaire Response from Empire Coke, Holt, Alabama.
     December 3, 1984.
2.    Coy, D. W.  Trip Report for Empire Coke, Holt, Alabama.
     September 19,  1984.
                                  8-45

-------
3.    Section 114 Questionnaire Response from Bethlehem Steel Corpora-
     tion, Sparrows Point, Maryland.  November 15, 1984.

4.    Coy, D. W.  Trip Report for Bethlehem Steel Corporation, Sparrows
     Point, Maryland.  September 17, 1984.

5.    Branscome, M.  R.  Trip Report for Armco, Inc., Middletown, Ohio.
     September 26,  1984.

6.    Section 114 Questionnaire Response from Armco, Inc., Middletown,
     Ohio.  October 23, 1984.

7.    Section 114 Questionnaire Response from U.S. Steel Corporation,
     Pittsburgh, Pennsylvania.  September 6, 1974.

8.    Section 114 Questionnaire Response from U.S. Steel Corporation,
     Clairton, Pennsylvania.  October 24, 1984.

9.    Section 114 Questionnaire Response from U.S. Steel Corporation,
     Clairton, Pennsylvania.  December 10, 1984.

10.   Coy, D. W.  Trip Report for U.S. Steel Corporation, Clairton
     Works, Pennsylvania.   September 13, 1984.

11.   Control of Emissions from Coke Oven Doors:   Design and Operational
     Guidelines.  Published by U.S. Steel Corporation, Pittsburgh,
     Pennsylvania.   1983.

12.   Cost Indices from Chemical Engineering.  Published by McGraw
     Hill, New York.   1974-1984.

13.   Development of Air Pollution Cost Functions for the Integrated
     Iron and Steel Industry.   PEDCo Environmental, Inc.  EPA Contract
     No.  68-02-3071.   July 1979.

14.   Technical Approach for a Coke Production Cost Model.  PEDCo
     Environmental, Inc.   EPA Contract No. 68-02-3071.  December 1979.
                                  8-46

-------
                           9.0  ECONOMIC IMPACT

     This chapter addresses the economic impacts of Regulatory Alternative
II for coke oven topside, charging, and door leaks emissions sources.  This
alternative applies to both new and existing coke ovens and is more fully
described in Chapter 6.   Because of uncertainties in the unit compliance
cost estimates, impacts are presented for two cost scenarios:   mid-range
cost estimate and upper bound cost estimate.  Regulatory Alternative II is
projected to increase the price of furnace coke by 0.2 percent and reduce
production by 0.3 percent.   The price of foundry coke would increase by
less than 1 percent and production would decrease, at most, by 2.8 percent.
No furnace or foundry coke battery is expected to close as a result of
Regulatory Alternative II.
     Section 9.1 presents a profile of the coke industry.   Section 9.2
contains the analysis of the impacts of the regulatory alternatives.   These
alternatives are outlined in Table 9-1.   The impacts are measured against
the baseline state of control for all sources.   Section 9.3 presents poten-
tial socioeconomic and inflationary impacts.
9.1  INDUSTRY PROFILE
9.1.1  Introduction
     Coke production is a part of Standard Industrial Code (SIC) 3312--
Blast Furnaces and Steel Mills.   Coke is principally used in the production
of steel and ferrous foundry products, which are also part of the output of
SIC 3312.  Thus coke is both produced and principally consumed within
SIC 3312.  Furthermore,  many producers of furnace coke are fully integrated
iron- and steel-producing companies.   Any regulation on coke production is
expected to have some impact on the entire blast furnaces and steel mills
industry with special emphasis on coke producers.
                                  9-1

-------
     TABLE 9-1.   COKE OVEN CONTROL  OPTION:   REGULATORY  ALTERNATIVE  II
Emission source
Charging
Door leaks
Lid leaksb
Offtake leaks'3
Emission limits3
16 seconds/charge
10 PLD
3 PLL
6 PLO
Control technology
Optimized stage charging
Door control program, modified
doors and seals
Adequate luting manpower
Adequate luting manpower,
repair or replacement
aPLD = Percent leaking doors
 PLL = percent leaking lids
 PLO = percent leaking offtakes.

 Lid and offtake leaks are collectively termed "topside"  leaks  in this
 chapter.
                                  9-2

-------
     This profile has two purposes:  (1) to provide the reader with a broad
overview of the industry and (2) to lend support to an economic analysis by
assessing the appropriateness of various economic models to analyze the
industry.  Further, the profile provides some of the data necessary to the
analysis itself.
     The industry profile comprises six major sections.  The remainder of
this introduction, which constitutes the first section, provides a brief,
descriptive, and largely qualitative look at the industry.   The remaining
five sections of the profile conform with a particular model of industrial
organizational analysis.  This model maintains that an industry can be
characterized by its basic conditions, market structure, market conduct,
and market performance.
     The basic conditions in the industry, discussed in the second and
third sections of this profile, are believed to be major determinants of
the prevailing market structure.  Most important of these basic conditions
are supply conditions, which  are largely technological in nature, and
demand conditions, which are determined by the attributes of the products
themselves.
     The market structure and market conduct of the blast furnaces and
steel mills industry are examined in the fourth section.  Issues addressed
include geographic concentration, firm concentration, integration, and
barriers to entry.  Market structure is believed to have a major influence
on the conduct of market participants.  Market conduct is the price and
nonprice behavior of sellers.  Of particular interest is the degree to
which the industry pricing behavior can be approximated by the competitive
pricing model, the monopoly pricing model, or some model of imperfect
competition.
     The fifth section of the industry profile addresses market perform-
ance.  The historical record of the industry's financial performance is
examined, with some emphasis on its comparison with other industries.   The
sixth section of the industry profile presents a discussion of industry
trends for the coke and steel sectors.  The seventh section discusses
market behavior.
                                  9-3

-------
     9.1.1.1  Definition of the Coke Industry.   Coke production is  a part
of SIC 3312--Blast Furnaces and Steel  Mills,  which includes  establishments
that produce coke and those that primarily manufacture hot metal, pig iron,
silvery pig iron, and ferroalloys from iron ore and iron and steel  scrap.
Establishments that produce steel from pig iron, iron scrap, and steel
scrap and establishments that produce basic shapes such as plates,  sheets,
and bars by hot rolling the iron and steel are also included in SIC 3312.l
The total value of shipments from SIC 3312 in 1982 was $36,931,900,0002 and
an approximate value for total coke production in 1982 was $3,220,011,000,3
or less than 10 percent of the total value of shipments.
     Coke is produced in two types of plants:  merchant and  captive.
Merchant plants produce coke to be sold on the open market,  and many are
owned by chemical or other companies.  The majority of coke plants in the
United States are captive plants which are vertically integrated with iron
and steel companies and use coke in the production of pig iron.  At the end
of 1984, 15 plants were merchant and 36 were captive, and merchant  plants
accounted for only 12 percent of total coke production.4 5
     9.1.1.2  Brief History of the Coke Industry in the Overall Economy.
Traditionally, the value of coke produced in the United States has  con-
stituted less than 1 percent of the gross national product (GNP).6  7
During most of the 1950's, coke production was about 0.3 percent of GNP,
and during the 1960's and until the mid-1970's, coke production was only
about 0.2 percent or less of GNP.  However, in 1974, coke production as a
percent of GNP rose to above 0.3 percent.  This trend continued for the
next 2 years.  By 1982, coke production was about 0.1 percent of GNP.3 8
     Previously, U.S. coke exports had been greater than imports, but that
trend has fluctuated.  The values of all U.S. imports and exports and U.S.
coke imports and exports are shown in Table 9-2.  From 1950  to 1972, coke
exports were much greater than coke imports, but after 1973, this trend was
reversed.  In 1982 and 1983, exports again exceeded imports.  Data for the
second quarter of 1984 indicate that coke imports are again  on the rise.
Imports for the first two quarters of 1984 totaled 247,604 megagrams (Mg)
compared to 6,874 Mg for the same period in 1983, and to 32,000 Mg for all
                                  9-4

-------




1
1



in


v
i-r
"
CM

2

a
re
LU
a
az
*~
Zi
13

LU
as
o
U-
>-.
1 —

~^
a



LU

0



•
<\j
i
en

LU

CD
 *- "
a> TO TO
o */» o
o TO +•>


U1)
O '•"N
cxv*
X
OJO
0
0) rH
0 ^"*
o

Ul

O -~v
CLV9-
OJ Ji
O

r>

>4— I/)
U» O •*->
*J t.
S- 0) O
o s- a.
j2--g
P.
o> re re

O i/> O
O (0 4->



c
o
Ifl •*-

t. Q.^-.
0 S *»
O. 3
E in to
•—CO
0 i-t
o> u ^^
N^
O (-
(_J Q
<*-





U)

I.
o
0.
SfQ
•F— <~ \

uioi
• o
^'~'
4J
O
1 —




&-
0) -
Or^OOOOOOOOOOOOOOOOOr-ir-i— 'OOOOOOOOOOOO
OOOOOOOOOOO'OOOOOOOOOOOOOOOOOOOOOOC





u ~
csir^r-rotvjCMLO«-Lr1u)Lncnoor-rHvar- r^LOLDOr^fnLOLOrnLT>vi>OLOLO«'^-i^r^f-»oOt-tLr>csiO(»iLO
OLOLr>Lr)LOiX)cn'or^i^drt^roLor^C)rt^odcsjrocnrtcd'-~^-C5t\ja-LOOLOmor~cncM«i-LOOCOcMOcnu3eOLnr«oocrioor-rHono
ad-HdddrHcvjmrJLn^TrLOr^aJ^LriLOroLOCnLOLrJdroaj^^LdcsJcnLO^cj
r-t r- 1(— trH

01 o
ce.

3
1

Ol
>

. «j
•o re
ai —
C 3

U 3
r— U
10
U i —
re
o c
c c


c;
re .
f-H
•a co
Oi en
— rH

U '-
' — OI
re a
u E
01

ai o
.* z
o
U -C
CT
re p
= 4-J
« 0
re >

•a 
•a 01
9-5

-------
of 1983.16  Exports for the first two quarters of 1984 and 1983 were
307,540 Mg, and 300,283 Mg, respectively,  and were 603,288 Mg for all  of
1983.16
     The same pattern applies to the percentages of coke imports and exports
                                                   &
within total U.S.  imports and exports.   From 1950 to' 1972, coke exports
were a larger percentage of total U.S.  exports than coke imports were of
total  U.S. imports.  Again, from 1973 to 1981, this trend reversed, and
coke imports were a larger proportion of total U.S. imports than coke
exports were of total U.S. exports.   Percentage shares of exports were
greater than imports in 1982 and 1983.
     U.S.  coke production has always been a substantial portion of world
coke production.   This share has decreased during the past 30 years, as
indicated in Table 9-3.  From 1950 to 1977, world coke production generally
increased while U.S. coke production decreased.  This trend explains the
decline in the U.S. percentage of world coke production.
     9.1.1.3  Size of the Iron and Steel Industry.  The value of shipments
of SIC 3312 has increased since 1960.  There have been a few fluctuations
in this growth; for example, as shown in Table 9-4, the 1965 value of
shipments of SIC 3312 was the highest value between 1960 and 1972.  Since
1972,  the value of shipments has remained around $30 billion, with the
highest value being $35 billion (1972 dollars) in 1974.  After reaching
another peak of $34 billion (1972 dollars), the value of shipments declined
to a 23-year low of about $18 billion (1972 dollars).  This result reflected
conditions in the steel industry.  In 1982, the steel industry sustained
record financial losses close to $3.2 billion  (1982 dollars).23  In 1983,
an additional $3.6 billion was lost.24
     For SIC 3312, Table 9-5 shows the value added by manufacture, the
total  number of employees, and the value added per employee.  Current and
constant (1972) dollar figures are included.   Both the total value added by
manufacture and the value added per employee peaked in 1974, the same year
in which the value of shipments for this industry was the highest.  The
increasing value added per employee might indicate that this industry is
changing to a more capital-intensive production process.  This aspect is
discussed in Section 9.1.6.
                                  9-6

-------
              TABLE 9-3.   COKE PRODUCTION IN THE WORLD6
                                                        17 18
Year
           World production
               (106 Mg)
U.S.  production
   (106 Mg)
                                                             U.S.  production
                                                              as a share of
                                                             world production
Oven and beehive coke combined.

Information on world coke production not available after 1979.
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978.
1979°
182.3
204.1
208.9
225.6
211.5
242.3
256.8
266.1
255.0
260.4
279.7
272.0
272.9
281.7
298.5
310.3
310.4
303.9
315.8
335.8
350.5
342.7
340.5
365.8
367.4
363.3
367.2
373.5
364.7
341.0
65.9
71.9
62.0
71.5
54.4
68.3
67.6
69.0
48.6
50.7
51.9
46.9
47.1
49.3
56.4
60.7
61.2
58.6
57.8
58.8
60.3
52.1
54.9
58.4
55.9
51.9
52.9
48.5
44.5
48.0
36.1
35.2
29.7
31.7
25.7
28.2
26.3
25.9
19.1
19.5
18.6
17.2 .
17.3
17.5
18.9
19.6
19.7
•19.3
18.3
17.5
17.2
15.2
16.1
16.0
15.2
14.3
14.4
13.0
12.2
14.1
                                 9-7

-------
TABLE 9-4.   VALUE OF SHIPMENTS, SIC 33122 19 20 21 22
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
Current dollars
(106)
15,738.8
14,873.3
15,571.6
16,418.0
18,840.1
20,841.7
21,193.9
19,620.6
21,161.1
22,299.0
21,501.6
21,971.3
23,946.7
30,365.5
41,671.7
35,659.8
39,684.1
41,897.8
49,055.4
55,695.8
50,303.9
57,472.9
36,931.9
1972 Dollars
(io6)
22,981.7
21,468.4
22,071.7
22,933.4
25,914.9
28,043.2
27,610.6
24,829.9
25,628.1
25,713.8
23,535.0
22,882.0
23,946.7
28,700.9
35,917.7
28,038.8
29,643.8
29,645.4
32,879.0
34,358.9
28,244.8
29,473.3
17,884.7
                     9-8

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

-------
9.1.2  Production
     9.1.2.1  Product Description.   There are two types of coke:   furnace
coke and foundry coke.   Furnace coke is used as a fuel  in blast furnaces;
foundry coke is used as a fuel in the cupolas of foundries.   Coke is also
used for other miscellaneous processes such as residential and commercial
heating.  In 1983, only 3 percent of all  coke used in the United States was
used for these miscellaneous purposes, 92 percent was used in blast furnaces,
and the remaining 5 percent was used in foundries.25  Time-series data for
the percent of total U.S. consumption attributable to each use from 1950 to
1980 are shown in Figure 9-1.
     9.1.2.2  Production Technology.  Coke is typically produced from coal
in a regenerative type of oven called the by-product oven.  The type of
coal used in coke production and the length of time the coal is heated
(coking time) determine the end use of the coke.  Both furnace and foundry
coke are usually obtained from the carbonization of a mixture of high- and
low-volatile coals.  Generally, furnace coke is obtained from a coal mix of
10 to 30 percent low-volatile coal  and is coked an average of 18 hours, and
foundry coke is obtained from a mix of 50 percent or more low-volatile coal
and is coked an average of 30 hours.
     The first by-product oven in the United States was built in 1892 to
produce coke and to obtain ammonia to be used in the production of soda
ash.  In such ovens, the by-products of carbonization (such as ammonia,
tar, and gas) are collected instead of being emitted into the atmosphere as
they were in the older, beehive ovens.
     The total amount of coke that can be produced each year is restricted
by the  number of ovens in operation for that year, and not all ovens are in
operation all of the time.  Oven operators try to avoid closing down a
group of ovens for any reason because of the time and energy lost while  the
ovens cool and reheat and because of the oven deterioration that results
from cooling and reheating.  However, it is estimated that at any time,
approximately 5 to 10 percent of existing coke oven capacity is out of
service for rebuilding or repair.28  In a report written for the Department
of Commerce, Father William T. Hogan estimated the potential annual maximum
capacity of U.S. oven coke plants as of July 31, 1979.29  Hogan assumed
                                  9-10

-------
    LU
    z
    cr
    D
       \
                                                                 o
                                                                 ca
                                                                             to
                                                                             CM
                                                                             rx
                                                                             O
                                                                             r-
                                                                            O

                                                                           'i


                                                                            in

                                                                            O
                                                                            O
                                                                            0)
                                                                           .*

                                                                            8

                                                                            re
                                                                            +^
                                                                            O
                                                                                 LU
                                  01
                                  O
                                  <
                                  Z
                                  c:
                              LU
                              CO
                              D

                              C
                              LU
                                                                              (N
                                                                              CS
                                                                              O
                                                                              o
                                                                              cs
                                                                              1C
                                                                              o
                                                      LD
                                                                              CM
                                                                              ta
                                                                                        I
                                                                                        o
                                                                                       o
                                                                                        en
c:
      tN
o
Cl
                  ca
cs
CO
                                                     J   i  l
                                                     J_.i
                                         CN
                                                           O
                                                                              o
                                                                              to
               0350
                                            lN30H3d
                                          9-11

-------
that almost 10 percent of his estimate of total  capacity would be out of
service at any given time; therefore,  he subtracted the out-of-service
capacity from total  capacity to obtain maximum annual  capacity.   The actual
number of ovens which are out of service in a given year varies greatly.
In December, 1983, 112-of 6,978 ovens, or 1.6 percent were being rebuilt or
repaired, and annual capacity totaled 35,575,000 Mg.30  In November 1984,
1,756 of 8,204 ovens or 21.4 percent were out of service, and annual capac-
ity totaled 51,180,000 Mg.5  Table 9-6 presents the data for November 1984.
     In actuality, ovens which are removed from service and placed on "hot
idle" status ar*e those likely to be returned to production in the short
term.  Ovens which are placed on "cold idle" status are less likely to be
returned to service and, historically, have not been returned to service.
The capacity of these ovens is included in a plant's total capacity for
bookkeeping purposes even though the ovens may be scheduled for demolition.31
     Within the limits of the number of ovens available for coking, both
furnace and foundry coke production levels vary.  Some ovens that produce
furnace coke can be switched to produce foundry coke by changing the coal
mix and increasing the coking time.  Furthermore, some ovens that produce
foundry coke could be changed to produce furnace coke by changing the coal
mix and decreasing the coking time.  Al'so, some variation in the combina-
tion of flue temperature and coking time is possible for either type of
coke.  A shorter coking time results in greater potential annual produc-
tion.
     9.1.2.3  Factors of Production.  Table 9-7 provides a typical labor
and materials cost breakdown for furnace coke production.  Coal is the
major material input in the production of coke.   In 1979, greater than 61
percent of the coal  received by coke plants was from mines that were company
owned or affiliated.33  In this same year, 14 States shipped some coal to
coke plants outside their borders.34  Of the coal received by domestic coke
plants, over 81 percent came from West Virginia, Kentucky, Pennsylvania,
and Virginia.34  Any potential adverse impact on the coke industry probably
will have some impact in these States.  A total of 33.6 million megagrams
of bituminous coal was carbonized in 1983.35
                                  9-12

-------
          TABLE 9-6.   MAXIMUM ANNUAL CAPACITY OF OVEN COKE  PLANTS
                  IN THE UNITED STATES IN NOVEMBER,  19845

In existence
Furnace plants
Foundry plants
Total
Out of service
Furnace plants
Foundry plants
Total
T *• b
In operation
Furnace plants
Foundry plants
Total
Number of
batteries
105
35
140
(25)
J2)
(27)
80
33
113
Number of
ovens
6,638
1,566
8,204
(1,646)
(110)
(1,756)
4,992
1,456
6,448
Capacity
(Mg)
44,810,000
6,370,000
51,180,000
(9,828,000)
(402,000)
(10,230,000)
34,982,000
5,968,000
40,950,000
 Batteries and ovens down for rebuilding and repair,  or  on  cold  idle  prior
 to permanent closure.

bDefined as "online" or "on hot idle."
                                 9-13

-------
     TABLE 9-7.   TYPICAL COST BREAKDOWNS:   FURNACE COKE PRODUCTION AND
                    HOT METAL (BLAST FURNACE) PRODUCTION32

Furnace coke production

     Labor and materials                                      Percent of cost

     Coking coal
     Coal transportation
     Labor (operation and maintenance)
     Maintenance materials

     Total labor and material costs                                100.0

Hot metal production                                          Percent of cost
                                               •
     Charge metal!ics
          Iron ore
          Agglomerates
          Scrap
     Fuel inputs
          Coke
          Fuel oil
     Limestone fluxes
     Direct labor
     Maintenance
     General expenses

     Total labor and material costs                                100.0
                                  9-14

-------
     Table 9-8 shows employment in the by-product coke industry from 1950
to 1970 and the percentage of total SIC 3312 employees in the by-product
coke industry.  This table shows decreasing employment in the by-product
coke industry.  A similar decline in employment has occurred in SIC 3312.
Unfortunately, employment data for the by-product coke industry are not
available after 1970; however, these figures can be estimated by regressing
employment in the by-product coke industry on total iron and steel industry
employment and on the ratio of coke used in steel production.*  These
estimates are also shown in Table 9-8.
9.1.3  Demand and Supply Conditions
     Domestic consumption of coke from 1950 to 1980 is graphed in Figure 9-2.
In the early 1950's, the amount of coke consumption was fairly large; an
average of 65 million megagrams was consumed annually between 1950 and
1958.  The late 1950's and early 1960's showed a sharp decrease in coke
consumption, with an average of only 48 million megagrams consumed annually.
Domestic consumption of coke increased during the mid-1960's to mid-1970's
to an annual figure of 57 million megagrams but it did not reach the 1950
to 1957 level.  The late 1970's showed another slump in coke consumption.
     The variation in coke consumption shown in Figure 9-2 has both cyclic
and trend components.  The demand for coke is derived from demands for iron
and steel products, and these demands are sensitive to the performance of
the overall economy.  Cycles in coke demand are linked to cycles in aggre-
gate demand or cycles in demand for particular products such as automobiles.
     The trend component in coke consumption results from changes in blast
furnace production techniques.  Coke is used as a fuel in blast furnaces,
but it is not the only fuel that can be used.   Coke oven gas, fuel oil, tar
and pitch, natural gas, and blast furnace gas have all been used as supple-
ments to coke in heating the blast furnaces.   The increased use of these
supplemental fuels over the past 20 years has caused the amount of coke
used per ton of pig iron produced (the coke rate) to decrease.   Other
causes of the decline in coke rate are increased use of oxygen in the blast
furnaces and use of higher metallic content ores.  Table 9-9 shows U.S. pig
     ^Regressions performed by Research Triangle Institute in 1980 and 1985.
                                  9-15

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



aFigures for 1971-1982 are estimates.  See text for more detail
                                  9-16

-------
                                         o
                                         CO
                                         03
                                         to
                                         cs
                                         o
                                         r-
                                         co
                                         CD
                                         CO
                                         to
                                            cr
                                            <
                                         •a-  uj
                                         co  >
                                         CM
                                         CO
                                         O
                                         (O
                                         03
                                         CO
                                  (O

                                   OJ
                                  ^
                                   O
                                   u
                                  «^
                                   O


                                   O


                                  1



                                   tf)

                                   O
                                   O
                                  4~»


                                   I
                                   CO
                                   Q.
                                   a
                                   re

                                  CO
                                  psl

                                  CT>

                                   £
                                   3
                                  .2*
                                  iZ
                                         «M
                                         in
                                         o
                                         in
CD
        o
        CD
in
in
o
in
                             in
                                    in
3x00 do swvyDvoaiAj NOITIIIAI
                         9-17

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

-------
                            TABLE 9-9 (continued)
                                               Coke used in
             Pig iron production              blast furnaces
Year              (103 Mg)                       (103 Mg)          Coke rate
1979
1980
1981
1982
1983
78-, 926
62,325
66,951
39,282
46,267
45,862
37,583
37,832
21,918
25,009
0.58
0.60
0.56
0.56
0.54
                                  9-19

-------
iron production,  coke consumed in the production  of pig iron,  and the  coke
rate for 1950 to  1983.   (Data limitations  make it difficult to calculate
the foundry coke  rate in cupola production.)
     Recently, there has been some concern about  the ability of the United
States'  coke-making capacity to support domestic  steel  production—the
major source of coke demand.   The study conducted by Hogan and Koelble of
the Industrial Economics Research Institute at Fordham University indicates
that in 1978, U.S.  production of coke was  14.1 percent below domestic
consumption.39  Imports increased dramatically in that same year.  Hogan
and Koelble attributed this decline in coke production to the abandonment
of coke ovens for environmental reasons and predicted a severe coke shortage
by 1982.40  This  prediction was disputed in a Merrill Lynch Institutional
Report by Charles Bradford.  The Bradford  report  attributed the lack of
adequate U.S. coke production in 1978 to two factors:  (1) a coal miner's
strike, which caused the drawing down of stocks of coke when they should
have been increasing, and (2) the premature closing because of EPA regula-
tion of some coke ovens that normally would have  been replaced before  they
were closed.41  The Bradford report stated that a survey of U.S. steel
producers revealed that all of the major steel producers were or soon  would
be self-sufficient with regard to coke-making capacity.42  The Bradford
explanation of 1978 coke imports seems more reasonable because 1979 coke
imports decreased about 1.6 million megagrams compared to the 1978 level.
     The following values describe the situation in the 1980s with respect
to production, imports, and apparent consumption of coke (thousand mega-
grams). 16
     Year     Production     Imports     Consumption     Distributor Stock
                                                               7,009
                                                               5,556
                                                               7,141
                                                               4,024
                                                               2,776
         *Two quarters of 1984

     Production is less than apparent consumption  in 1981, 1983, and 1984.
For each of these years, stocks and  imports more than accommodate the short-
1980
1981
1982
1983
1984*
41,851
38,815
25,506
23,413
14,446
598
478
109
32
248
37,447
39,975
23,384
27,080
14,886
                                  9-20

-------
fall.   Coke producers were operating at 80 percent of total  capacity in
November 1984.5  Thus, it is unlikely that major shortages will  develop in
the near future; in fact, unless there is a turnaround in the steel  industry,
surpluses appear more likely.
9.1.4  Market Structure
     Market power,  the degree to which an individual  producer or groups of
producers can control market price, is of particular economic importance.
Market structure is an important determinant of market power.  Pricing
behavior is relevant to the choice of the methodology used in assessing the
potential impacts of new regulations.   It is important to determine  if the
competitive pricing model (price equal to marginal cost) adequately  des-
cribes pricing behavior for coke producers.
     Any analysis of market structure must consider the characteristics of
the industry.  This analysis addresses the number of firms producing coke;
the concentration of production in specific firms; the degree of inte-
gration in coke production; the availability of substitutes for coke; and
the availability of substitutes for the commodities for which coke is an
input to production.  Also, some information on past pricing in the  coke
industry is presented.  These topics will be considered together with
financial performance (Section 9.1.5) and trends (Section 9.1.6) in  asses-
sing market behavior (Section 9.1.7).
     9.1.4.1  Concentration Characteristics and Number of Firms.  This
section describes various concentration measures that can be computed for
the furnace and foundry coke industries.   Normally, concentration ratios
are used as an indication of the existence of market power.   While concen-
tration ratios are a useful tool for describing industry structure,  concen-
tration should not be used as an exclusive measure of market power.   Many
other factors (e.g., availability of substitutes, product homogeneity, ease
of market entry) determine a firm's ability to control market price.
     As of November, 1984, 23 companies operated by-product coke ovens.5 43
Twelve companies are integrated iron and steel producers; 11 companies are
merchant firms.   These companies owned and operated a total  of 51 coke
plants; 36 of these plants were captive and 15 of them were merchant.  A
list of these companies,  their plant locations, the major uses of coke at
                                  9-21

-------
each plant, and plant coke capacities is given in Table 9-10.   A plant site
may include more than one complete plant.
     Reported capacities in Table 9-10 are maximum,  nominal  figures,  which
do not include any allowance for outage like that determined for the  overall
industry in Table 9-6.   All but one of the largest plants are captive, and
most of the merchant plants have very small capacities.   Furnace coke
production is concentrated in captive plants.   Virtually all of the coke
used in foundries and in other industries was  produced by merchant plants.
If coke plant sites were ranked according to capacity, the top five plant
sites and top ten plant sites would have 37.1  percent and 54.6 percent of
total coke capacity, respectively.
     By-product coke plants are concentrated in the States bordering  on the
Ohio River, probably because of the coal in that area.  Pennsylvania  contains
12 plants, and Ohio and Indiana each have 8 plants.5
     Table 9-11 divides the United States into 11 coke-consuming and coke-
producing regions and shows the amount of coke produced in each region and
the locations of coke consumption in 1977.  Most of the regions produce the
bulk of the coke they consume; only three regions produced less than  80
percent of their own consumption and only one  produced more than it needed
for its own consumption.  Transportation of coke across long distances is
avoided whenever possible to reduce breakage of the product into smaller,
less valuable pieces and to minimize freight charges.46
     The concentration of production or capacity in specific firms may have
economic importance.  Table 9-12 presents the  percent of total capacity
owned by the largest four (of 23) firms.  The  four-firm concentration ratio
for the coke industry has increased over the years.   In 1959, the four-firm
concentration ratio was 53.5 (the top four firms owned 53.5 percent of
total capacity)47; in 1984 it was 69.9 percent.  Consolidation of the
industry through mergers, acquisitions, and closures has encouraged this
trend.
     In the preceding discussion, furnace and foundry coke production are
considered jointly.  However, each existing coke battery may be considered
a furnace or foundry coke producer, based on the battery's primary use.
Separate capacity-based concentration ratios for the two types of coke are
                                  9-22

-------














r?


in
»^
00
cn
t— i

s-
QJ
.a
QJ
o
z
UJ
1 	 -,-
f—
t—
CO

Q
111
1 I i
1—
(-H
Z
""•%

UJ
1—
z
t—
_1
Q-

LLJ
*~s
O
CJ


•
o
tH
1
cn

UJ
	 i
oo
<^
t—























-p /-->
•r- S.
<-> >>
re ">.
Q. O5
re s
u

QJ O
-i<: rH
o •— '
re
QJ
•w*
o
(J

4-
O

in
QJ
in'
3
j_
O
re
z

c
o
•r-
•P -P
re c
UfH
 P**
M
I— 1




a> QJ
u u
re re
c c
i. J-
3 3
4- 4-

•P P
in in
re re
00 CQ






QJ QJ
> >
•P -P
Q. Q.
re re
o o



JD
/*^
CSJ
s^/

z:
o

2»— #»
""•^ C
3
- O
•a -P
c QJ
re r—
r- TJ
-C T3
in •»-
 <7> O"> O
CM 1 — CM un
M *t •» *»
CM tH i— 1 ro




QJ QJ QJ QJ
U U U (J
re re re re
c c c c
S- i. S- i-
3333
4— 4— 4— 4~*

+J +J +J ^_>
in in in in
re re re re
QQ OO CQ CQ





^j
QJ QJ C QJ
> > re >
+j +j "5 -p
a. o. i. a.
re re QJ re
o c_> s: o




o
^"
Z
h— t •*
>- -P
 re
a» 3 re a.
CQ CQ — 11/5

Q.
s_
0
C_J

^—
QJ
QJ
•P
I/)

ps
QJ
-C
QJ
r^
JZ
P
QJ
CQ







00

px.






QJ
U
re
c
S-
3
4-

-P
in
re
CQ






QJ
>
-P
a.
re
o










i — i
^^

^
c
^
o
o
i-
re
QJ
0










r—
QJ
QJ
+J
1/5

QJ
0)
3
0
a:







LO
I— 1
f^^
ft
co




QJ
(J
re
c
S-
3
4-

4J
in
re
03






QJ

-P
a.
re
o




s^
CO
\^^

z


ft
o
05
ro
u

f~
r i

.
UJ





.
o
o

r~
QJ
QJ
I '•
<-O

•o
c
re
i —
c
HH







CM
00
Lf)






QJ
U
re
c

3
4—

P
in
re
ca






QJ

p
a.
re
0











— i
i — i

ft
0
O5
re
u

f™
o






*
u
c
t— 1

M
QJ

re
r^
s-
QJ
4^
C
1— <
9-23






CO *""*^ 00 00 ^^ f^O ^5 LO
^^ CD *^ LO tji ^^ i~H ^f
c\i r^- ^ r***- r^ to c^o cn
*. *, ^
r- i i— 1 rH




CJOJOCDOJ>>>>>>>
4J+J4J+J^J+J+J4J
O.Q.Q.Q.Q.Q.Q.Q.
flJfQrtJ'OfOfOfOnj
oooooooo





/^-H,
CSJ

U Z < — 1
< a: HH a. 1-1
a. o

- - o «c ^= o — J a:
re *o 05 05 05 «^ o
a. c re - s- re

•r- r— T- QJ JD -r— in C
3 QJ r" T^J (/i r* f^ QJ
cr>o mpcj s :.
•r- QJ TJ -P O S-
i— r— . re T- . -c re

h-
_J

QJ

(—







[•"•».
0
LT)






QJ
0
re
c:
S-
3
4-

4->
in
re
CQ






ai

+->
a.
re
o









X
H-

n
i-.
re
•p
CO

QJ
C
o
	 1

•a
.
o
CJ

r™~
QJ
QJ
-P
OO

S-
re

CO

QJ
C
o
—I







oo r^
ID cn
00 00
^
rH




QJ QJ
U U
re re
c c
S- i-
3 3
4- 4-

•P -P
in in
re re
CQ CQ






QJ QJ

•P -P
a. a.
re re
O C_3






^1
H— 1

*.
^*j
•4-3 t-H
• r— y
C_)

QJ -P
•P •(-
•i- O
C t-
re -P
!- QJ
cj a
QJ

a.

o
o

I— •
QJ
QJ
4->


^^
re
c
o
•f—
1 »
re
z







r***
o^
o
•*
rH




QJ
(J
re
c
s.
3
4-

-P
in
re
CQ






QJ

+J
Q.
re
^_5




^^
"^

«,
TJ
C
re
P—
in
l-H

in

C

0

CQ


,
a.
5-
o
o

r—
0)
QJ
+J


C
O
4.)
i-

QJ


T3
QJ
3
C
•r"
•p
«^- c
in o
ro u
s»>





QJ
U
re
c
S-
3
4-

4J
in
re
00






QJ

!|3
a.
re
o








~^~
0

«.
j=

3
O

in
j S
^
o
a.
4-
Q.

O
O

QJ

O
CJ

c
o
4-1
in
O
CQ

3:
QJ
Z


-------































"O
cu
3
C
4J
C
o
u
N^ /

o
1
CD
1 1 1
_J
CQ
1 —






































>^
•P /*"N
•1- S_
Q. O)
re 2:
o
CO
cu cz*
^ r-l
O x—x
CJ





re
cu
0
u

t|
o

in
CU
3

^0
re
2:
c
o

'-p -p
re c
.)£ plH
M- Q.

in 4-
in o
re
CJ






c
o
•r-
-P
re
u
o


-P
c
re
r—
a.









CJ
£
re
c

>^
c
re
Q.

o






«d- CM CD oo co o
cn CM i— i CM cn co
CM CO Cn CM «d- rH
in rH «3- rH iH









CU CU CU CU CU CU
U CJ U CJ U CJ
re re re re re re
c c c c c c:
S. i- i. i. i- i.
333333
<+~ <+_ df« (-^— <+— *+—

co u) tnt/) 1/5 t*o
TO TO TO TO TO tJ
CO CO GO CQ CO CO




GJ >>>>>
4J 4J +J +J .fj +J
a. CL Q. o_ a. ex
TO TO TO ''O TO TO
CJ O O O O O






, re >
re •!- -r- s- i_ o
r— re re re o &-
O U_ U_ C3 — 1 O-





.
a.
s_
0
CJ

r—
CU
cu
•p
CO

.
CO
.
Z5





cn o
o cn
ID *J-
rH









CU CU
U (J
TO fl3
c c
1- S-
3 3
1— *+•»

CO U)
re re
CO CO




cu cu
> >
-P -P
o. a.
rt3 n3
o o






^

a*
i— —
r™-
•r-  Q-
C
CU ~
JO C
3 CU
cu in
•P 10
CO 0)
e
. o
LU S

CT
^*
OJ
S-
3
o
in
-P
4^
«r™ •
a. a.
1 S-
O5 O
C CJ
• r*
r— • r—
CU CU
cu cu
JZ -P
3: co





cn
cn
•>f



S-
•rj
C
3
0
il-

CU
CJ
re
c
s_
3
<4«

CO
re
CO



4J
C
re
/—
>
CJ
S-
O)
2:









____!
^c

«s
^s
TO
x:
o^
c

i
S-
• r-
ca





.
a.
t.
o
CJ

^
cu
-p
r—
re
3:

E
•r~
"D





r*x r^x ro
O LD CD
CM rH Lf)



>>
•o
C
3
O
<*-

CJ CU
e~ rj
4— ' re
Or- C
re i-
-•i- 3
^> ^ ^> *^~
5- j-> S_
c 3 c cn
3 -a 3 re
Or- o ^~
ki. thj r^
U_ -r- U. 00



+^ 4^ 4^
c c c
re re re
U CJ U
i- J- S-
cu cu cu
2: s 2:











1
>
•^
C
3
0
14—

CU
u
re
c
t.
3
L^»

in
re
CO



-P
c
re
o
i-
cu
2:






^
a.
f.
T3
C
re

CO
t— »

CJ

r—
•r"
>
CJ









CJ
c
>— H

•*
0
01
c
re
c
CU
JZ
CO





ro CM
00 O
Lf) *f







JZ
r«-
re
s_
4->
CO
3
•a
- c
^} "r™ ^*>
1- S-
c cu c
3 JZ 3
01 \ *^
+-1 CJ
U_ O Lu



•P 4J
C C
re re
0 O
S- S-
CJ CU
2: 2:











 C
c o
re -P
t- 10
s- >>
re o



in
-P
u
3
Tj
O
S-
a.
i
>^
CO

re .
E 0.
re i-
Q O
re o
r—

-------











































/^—\
•o
CU
3
C

-P
C
o
u
•~+^


o
1— t
1
en

uu
_i
CO
>
-P '•^
•i- S-
U >,
ro \.
O. O)
ro 2:
u
«
0) O
24 rH
o ••— '
CJ








ro
cu
j*
o
u

4—
O

CO
CU
CO
3

S.
O
•r—>
ro
s





c
o

•p -p
ro c
u ro
•f— r—
"+- 0-
•r—
CO 4-
CO O
ro

C_3














c
o
• r—
-P
ro
u
o
P—

•p
c
ro
»—
Q_













0)
e
ro
c

>>
c
ro
o,
E
o
C_3




1^.
l—f
CO





















>>
1?
T3
C
3
O
u_










•p
c
ro
.c
u
j_
CU






















»— 1
s

M
-p
• l~
o
J_
-p
CU
Q







,
o.
S-
o
o

CU
^
o
o

.p

o
J_
•p
0)
Q




rH
CO
rH





















>,
S?
•o
c
3
O
u_










•P
c
ro

t3
S-
cu
s:
























— 1
<

M
-»J

o











,
o
o

0)
.*:
o
C_}

CU

•f— •
O-

LU




CsJ
n
rH












r—
ro
•i*—
S-
-P
CO
3
T3
- C
>>•--
S_
-O S-
C CU
3 -C
0 -P
Ll_ O










•P
C
ro
^:
u
s_
cu
s:

















•z.
i— i

**
O)
4->
3
ro
oc

cu
f
c
cu
r—









.
T3 Q.
C S-
ro o
o
CO
ro r—
cu ro
o
fO -r-
C E
ro cu
•r- £1
-a cj
c
»— *




en
en
CSJ





















>>
s_
T3
C
3
O
u.










-P
c
ro
j=
u
S-
cu
s:





















>•
z.

f
0
r—
ro
4-
4-
3
CQ





.
a.
5_
o
o

cu
-*
o
CJ

ro
T3
c
ro
3
ro
c
o
I—


.
cu
j*:
o
o

4-
o

c
o
• r—
+J
3
jQ

S-
-P
CO
'^
-a

r~
ro
-P
0
-P

ui
_
4->
C
ro

a.

HI
_c
4->

4-
O

+J
C
CU
u
i.
O)
a.

o
CSJ

4->
CO
ro
cu
i—

-p
ro

co
•r—

-P
• r™

4-
• t—

CU
CO
3

S-
O
•1-5
ro
E

ro

•o
cu
i-
/I*
cu
•o
• r-
CO
C
0
(J

CO
•i—

a>
vO
3

•a
c
cu

c
<
ro .

rH
00
cu . 01 en
C +J -C T- rH
0 C -P 3 -P
ro •«- co ro C
>> ^- 3 c x: -r-
/— Q. 3 ^
c i_ s- -a
o ro cu CQ i— ai
•i- O5 CU r—
c s- c co -r-
•a ro cu cu M- 4-
a> > s en
*J i— i- CO rH
ro >, CO CU -P rH
(J CO -P CQ C
•r- c -r- ro s-
•o c "o -P a>
c cu en c i — -P <3-
•r- a. co ro 3 Q. co
en co ro en
CO rH - C -C rH
•r- ro .00
a. i— a. u -c
s- a. «r- s- s- u
CU T- S- O -P CU S-
jQ 3 Q. O CU T3 ro
E CT < ^ C 2
3 -r- r— S- 3
c i— c cu ro c
 cu T- -a
JZ CU S • -P CU
4— +j u i — ocoro i —
i— i c ro co s- s_ . •!-
-P 3 C CU CU >, t-
ro o o ~o a. s-
• C •*— . 'r- O O rH
C r— C -PCUro OlrH
o cu ro ro-ps-co cu
•i- cu z. ro •!- • -P s-
+->+j " TDCU cnrocu
ro CO r— C|_+J.COU-P
u cu o T: ro • en a.
o 4- cu ais-o-rHi— ro
r— o -P >>i— o$- ro j=
co c 3 a. o r— -i-o
•P C roTDS-CJ-r-l-
ro o s- o-cuo s-+->s-
jz -r- ro E-CUCUQ.COCU
•P 4-> -P O(J ^)Cu ^->co4- -i-'r-
O i— C C -P C O5
CO S-r-O  CU
T3>>, -CCU 3Q.CQ.
4— cucuc -P^II — cu ro •!— o
o t. ro ro z .c
COQ. -CO'r— CJT3CO
S- O3E .^^J4- cu-r-
cu +J+JO ucucos--a
ja rou ccucu 03-
E un+J i— i 3 +j >, 4- r—
3 OOlO-P OC (J,-oro o O-I-OT-
j-> ro-r-a. S--304-C
s i a>cn o T- o
cu -acu S-OOC-P .-PO
•P . c i— ^r cucn co.ro
roc-i— OP -PrHco CU S- CUi —
uo o c-i-s-o-cro
•r-'i-C - >-Hi — -r--zjz3 Q.CTO ro
^->4JcoCS_co O1OS-S-U
C -PO-PCUC3S-O3 — i 3 cu ccu-i-r— ro
CCOC T3J2r— 1— 1 -P O. (Q O)
•r- T- 3 +J -P C r— CO 1 •.- O
X OC CO 1— 1 S_CUr— O5-P O
coaicrocu a)4-ro^:ccc
s- ci — 3>>oi c+J-r-curo
 roOi-CCU S-.o 3 r— 73 -P
jsc -PI— aio_-.-ocu-'--p
£ro>cu s-j- SS-+J — i cu co ro
3 r— i — .e oro oro ujz cu-c
ZQ._lr— ZU. 
-------
















«
en
rH
Z
t— i
LO
1—
Z
LU
Q.
H— *
T"
OO

LU (f>
^ E
o re
o t-
O)
_j re
< cn
Z CJ
0 E
1 — 1
C3 co
LU O
a: I-H
LU
H-

HH

r-i
rH
1
en
LU
GO

I—


























, —
re
•P
0
h-




















C
0
•r*
D5
0)
i.

cn
c
•r-»
E
13
V)
c
o
0















> 3:


Z
0_



"JET
o

HH
3:

_

h— 1
^g*



«t
o x
2! h-

M M
>- Z
•^ }—





z
h— 1


_,
I-H


Q >-
•SL Z


ro

M
< o
0 0


_J
<£




c
o

cn
QJ


CO
C

tj
3
•a
0
s_
Q_
rH
LO
CO
•*
CO
rH
in


rH



^
rH
rH

r--



m
cn
rH




LO
LO
^J»






CM
r-H
rH


rH
00


o
rH



p^
CSJ



00
CM
CM
CM













— )

•> *.
re o
*^~ ^3
C ^
s_ s_
0 0
4— r—
•r- O
<— 0
03
-

^
CD
Z

**
•a
c
re
r—
^j
i-
re
2:
•=r uo
CM r-
•5}" 00
«t «t
rH r-
O 0



o co



O CM
LD


0 rH
rH


o r^
cn





0 LO
ro







0 <•
O^
LO
M
^
«* cn
CM LO
^J-
rH
0 0




o in




0 0














i/>
•r- re
o c.
c. re
• r— *r-
r- T3
f— C
1— 1 I-H
CO
CO
i—t
•<
rH
0
CSJ


o



co
rH


O



m
CM
rH




00
CM
^^






LX^




LO
I— t


O




00
rH



<•
rH



•i- > ^
^: a;
u c
3 C
-p a>
C h-
HI
^£
CO rH
LO f-~
LO LO
•<
CM
0 0



0 rH



LO in



0 00
LO
_J
^^

Cn rH
CO LO
LO
CM



rH rH








r^v co
r^



o en
LD
CM

o o




0 LO




O 0



c
U1
c
o
u

.f_
3:

M
re
C 0
re -P
cn m
•i- CD
-C C
(J C

SI HE!
r-- o
rH CM
LO rH
•N M
r^- co
rH
CM CO
rH


CM P^
LO
CM
O
rH
LO O
in r^
CO CO
LD rH
O O



O CM
LO LO
CM




cn co
p^
CO






LO *5j"
LO CO
CO rH


00 rH
co *r
rH CM
rH
0 rH
m



•3- 0




o cn










re
• r—
C.
re
>
r—
>^
tn
O c
•r- C
j= a>
O Q-
cn
cn
o
*.
CO
m
LO

CM
^
rH
CM

O



O



CM
rH
*fr




CO








O




o



o




o




o







re
• r—
C
• r-
O)
i.
• r-
*• ^>
re
C -P
•r— UJ
CO CU
i- 3:
•r-
>
cn
LO
0
f>
00
rH
m
LO
CM
LD
LO
LO
o
rH
CM
CO
cn
^
LO
f^X
_J
^^

cn
CM
cn
CO



CM
^^
00

rH




00
00
CSJ
*.
00
o
LO
CO
co
CO
m
^
«*"

rH
CO
p*^
*•
CM
rH
in
CM
CM














— 1
<

O
1—
9-26

-------
TABLE 9-12.   PERCENT OF COKE CAPACITY OWNED BY TOP FIRMS
                  (NOVEMBER, 1984)5
Firm
U.S. Steel, Inc.
Bethlehem Steel Corp.
The LTV Steel Corp.
Inland 'Steel Co.
Sum of largest four firms
Capacity
(103 Mg)
14,916
8,841
8,299
3,715
35,771
Percent of
total capacity
29.14
17.27
16.22
7.26
69.89
                      9-27

-------
calculated based on this allocation.   The 1984 four-firm concentration
ratio for furnace coke is 75.4;  the 1984 four-firm ratio for foundry coke
is 65.2.5
     Concentration in the steel  industry has economic relevance because a
large fraction of all furnace coke is produced by integrated iron and steel
companies.  Historically, the eight largest steel producers have been
responsible for approximately 75 percent of industry production.  However,
from 1950 to 1976, the share of  production attributable to the top four
firms declined from 62 percent to 53 percent.48  In 1981, the seven largest
steel companies produced about 70 percent of steel made in the United
States.49
     In summary, concentration exists in the production of both types of
coke and in steel production.  However, the concentration is probably not
sufficient to guarantee market power, and many companies are involved in
the production of both coke and  steel products.  Other factors must be
considered in any final assessment of market power.
     9.1.4.2  Integration Characteristics.  When one firm carries out
activities that are at separate  stages of the same productive process,
especially activities that might otherwise be performed by separate firms,
that firm is said to be vertically integrated.  Through vertical integra-
tion, the firm substitutes intrafirm transfers for purchases from suppliers
and/or sales to distributors.  A firm may seek to supply its own materials
inputs to ensure a stable supply schedule or to protect itself from monopo-
listic suppliers.  The firm may seek to fabricate further or distribute its
own products to maintain greater control over the consuming markets or to
lessen the chance of being shut  out of the market by large buyers or middle-
men.  Therefore, the presence of vertical integration may constitute a
firm's attempt to control costs  or ensure input supplies.  Vertical integra-
tion does not guarantee market power (control over market price).
     Many coke-producing firms,  especially furnace coke producers, are
vertically integrated enterprises.  As previously mentioned, 36 of the
existing coke plants are captive; i.e., they are connected with blast
furnaces and/or steel mills.  In addition, many coke firms own coal mines,
and greater than 61.0 percent of the coal used in ovens was from captive
                                  9-28

-------
mines in 1979.33  Assurance of coal supply to coke production and coke
supply to pig iron production appears to be the motivation behind such
integration.
     One implication of vertical integration is that much of the furnace
coke used in the United States never enters the open market—it is consumed
by the producing company.   Accordingly, the impact analysis for furnace
coke (Section 9.2.2) uses an implied price for furnace coke based on its
value in producing steel products, which are transferred on the open market.
     9.1.4.3  Substitutes.  Substitutes for a given commodity reduce the
potential-for market power in production of the commodity.  The substitu-
tion of other inputs for coke in blast furnaces is somewhat limited, but
not totally unfeasible.  In addition, electric arc furnaces, which do not
require coke, are becoming increasingly important in steel production.  The
trend toward electric arc furnaces and mini-mi 11s has eased entry into the
iron and steel industry, which in turn reduces market power.
     Imported coke can also be substituted for domestically produced coke.
In fact, although U.S. iron and steel producers prefer to rely on domestic
sources of coke, coke imports have increased most recently.  If the cost of
domestic coke increased substantially compared to the cost of imported
coke, U.S.  iron and steel  producers might attempt to increase imports even
more.  Correspondingly, if costs of imported coke are reduced due to improved
foreign technology and productivity, reductions in foreign labor cost, or
other reasons, imports might become more desirable.
     Furthermore, substitutes exist for the final products (iron and steel)
to which coke is an input.  Increases in the price of coke and the result-
ing increases in the price of iron and steel products can lead to some
substitution of other materials for iron and steel, which also reduces
market power in the production of coke.  Analagous substitutions for foundry
coke are possible, and cupola production of ferrous products, which uses
foundry coke, has competition from electric arc furnaces that do not use
coke.  Hence, there is a technological substitute for foundry coke in the
manufacture of ferrous products.   Furthermore, imported foundry coke can be
substituted for domestic foundry production.  In conclusion, some substitu-
tion for coke is possible in the manufacture of both steel and ferrous
products.
                                  9-29

-------
     9.1.4.4  Pricing History.   As previously indicated,  a significant
portion of all U.S.  coke production is not traded on the  market.   However,
the Bureau of Mines and the Energy Information Administration collect
annual data on coke production and consumption and give the quantity and
the total value of coke consumed by producing industries,  sold on the open
market, and imported.   Dividing total  value by quantity yields an average
price for each of these categories.   Time-series data on  these three average
values are given in Table 9-13.   (Furnace and foundry coke are combined in
these figures.)
     Also shown in Table 9-13 are data on the average value of coal  that is
carbonized in coke ovens.  An examination of coke and coal prices reveals
that increases in coal prices generally coincide with increases in coke
prices.  In fact, only 3 years show an increase in the price of coal that
was not accompanied by an increase in the price of the two categories of
coke.  Although it is impossible to conclude from this trend that individual
firms have market power, it indicates that the industry can pass through
some increases in costs.
     9.1.4.5  Market Structure Summary.  Although there is no perfect
method for measuring the extent of market power, the preceding sections
addressed four characteristics used to measure the potential for market
powei—concentration, integration, substitution, and historical price
trends.  Concentration statistics indicated that some potential for market
power exists in the coke industry, yet, these statistics  are not conclusive
proof.  Similarly, vertical integration in the steel industry is not conclu-
sive in identifying the presence of market power because  vertical integration
is a method of controlling the cost and ensuring the quality and supply of
inputs.  Finally, the possibility of substitution represents a strong
argument against the existence of extensive market power  in the coke-making
industry.
9.1.5  Financial Performance
     Financial data on the coke-producing firms or their  parent firms,
including captive and merchant furnace and foundry producers, are shown in
Table 9-14.   Firms for which data are not available are noted.
                                  9-30

-------
     TABLE 9-13.  COMPARISON OF COAL PRICES AND DOMESTIC AND IMPORTED
                         COKE PRICES6 50 51 52 53

      Average value of  Average value of  Average value of
      coal carbonized.    oven coke used    oven coke solg   Average value
      in coke ovens '     by producers      commercially     imported coke
           ($/Mg)    •         ($/Mg)           ($/Mg)            ($/Mg)
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
9.56
9.85
10.17
10.19
9.92
9.74
10.31
10.92
10.90
10.89
10.90
10.79
10.86
10.46
10.23
10.48
10.78
11.05
11.03
11.49
13.46
15.43
17.34
20.19
40.22
48.73
48.68
50.99
57.37
55.88
62.09
69.29
71.62
65.36
14.26
14.50
15.11
15.36
17.33
17.90
19.39
19.98
19.82
19.16
19.92
19.12
19.53
18.88
19.17
17.89
18.40
18.58
19.57
21.54
30.30
32.86
35.76
41.34
82.32
92.84
93.83
90.57
105.79
117.39
123.42
125.52
131.24
124.57
14.54
15.72
17.63
17.96
18.95
18.52
20.27
21.51
21.90
23.03
22.32
23.30
23.36
23.24
22.85
23.90
24.49
24.99
24.25
27.01
33.04
41.29
44.87
47.31
72.47
96.61
104.01
111.95
118.03
107.54
113.24
124.34
126.24
124.67
13.34
13.17
15.96
12.02
11.98
12.26
12.38
14.43
14.25
12.89
13.06
13.44
14.42
14.78
16.10
16.95
20.60
20.41
22.31
21.36
25.46
31.93
27.70
40.16
60.14
94.84
. 93.35
--
--
94.32
87.12
89.59
84.61
61.45
Both furnace and foundry coke and the coals used to produce furnace and
foundry coke are included in these figures.
Market value at the oven (current dollars).

General customs value as reported by the Department of Commerce (current
dollars).
                                 9-31

-------
TABLE 9-14. FINANCIAL INFORMATION ON COKE-PRODUCING FIRMS, 1983
• (million 1983 dollars)3 S4 ss 56 57
Company name
Armco, Inc.
Bethlehem Steel Corp.
Ford Motor Co.
(Rouge Steel)
Inland Steel Co.
Interlake, Inc.
The LTV Steel Corp.
McLouth Steel Corp.1"3
(New Boston Coke Corp.)
National Intergroup, Inc.
(National Steel Corp. )
Northwest Industries'11
(Lone Star Steel Co. )
Shenango Furnace Co., Inc.n'°
(Shenango, Inc. )
U.S. Steel Corp.
Weirton Steel Corp.'3
Wheeling-Pittsburgh Steel Corp.q
Jim Walter Corp.n
Koppers Co. , Inc.
Alabama Byproducts Corp.
Carondelet Coke Corp.1"
Chattanooga Coke and Chemicals
Co. , Inc.
Citizens Gas and Coke Utility
Detroit Coke Corp.1"'1
Indiana Gas and Chemical Corp.
McWane, Inc. (Empire Coke Co.)
Tonawanda Coke Corp.
Net sales
4,165
4,898
44,455
3,046
835
4,578
11
2,993
1,608
145
16,869
1,000
772
2,025
1,566
229
k
17
316
k
64
k
k
EBITC
(526)
(239)
2,166
(177)
38
(252)
(0.09)
(177)
(104)
k
(1,208)
k
(72)
113
42
.<
k
k
k
k
(0.2)
k
'<
Cash flowd
70
1299
5,542
106g
67
(164)9
k
161
154g
k
1.5639
k
(70)9
159
1749
192
k
k
91
k
k
k
k
                            (continued)
9-32

-------
TABLE  9-14 (continued)
Company name
Armco, Inc.
Bethlehem Steel Corp.
Ford Motor Co.
(Rouge Steel )
Inland Steel Co.
Interlake, Inc.
The LTV Steel Corp.
McLouth Steel Corp.1>J
(New Boston Coke Corp. )
National Intergroup, Inc.
(National Steel Corp.)
Northwest Industries
(Lone Star Steel Co.)
Shenango Furnace Co., Inc.n'°
(Shenango, Inc.)
U.S. Steel Corp.
Weirton Steel Corp.p
Wheeling-Pittsburgh Steel Corp.q
Jim Walter Corp."
Koppers Co. , Inc.
Alabama Byproducts Corp.
Carondelet Coke Corp.1"
Chattanooga Coke and Chemicals
Co., Inc.
Citizens Gas and Coke Utility
Detroit Coke Corp.r>
Indiana Gas and Chemical Corp.
McWane, Inc. (Empire Coke Co.)
Tonawanda Coke Corp.r
Net working
capital
563
271
503
233
203
538
(10)
252
338
16
789
147
102
136
282
50
k
k
20
1
2
48
k
Current
assets
1,576
1,259
10,819
789
378
1,848
2
875
762
43
4,298
332
343
1,594
527
73
k
k
82
13
13
58
k
Current
liabil ities
1,013
988
10,316
556
175
1,310
12
623
424
27
3,509
185
241
1,458
245
23
k
k
62
12
11
10
k
(continued)
      9-33

-------
TABLE 9-14  (continued)
Company name
Armco, Inc.
Bethlehem Steel Corp.
Ford Motor Co.
Inland Steel Co.
Interlake, Inc.
The LTV Steel Corp.
McLouth Steel Corp.1 lj'
(New Boston Coke Corp.)
National Intergroup, Inc.
(National Steel Corp. )
Northwest Industries
(Lone Star Steel Co.)
Shenango Furnace Co., Inc.n'°
(Shenango, Inc.)
U.S. Steel Corp.
Weirton Steel Corp.p
Wheeling-Pittsburgh Steel Corp.q
Jim Walter Corp.
Koppers Co. , Inc.
Alabama Byproducts Corp.
Carondelet Coke Corp.
Chattanooga Coke and Chemicals
Co. , Inc.
Citizens Gas and Coke Utility
Detroit Coke Corp. r>
Indiana Gas and Chemical Corp.
McWane, Inc. (Empire Coke Co.)
Tonawanda Coke Corp.
Annual
interest
expense
154
104
567
68
12
171
k

62

63

k

1,074
k
58
140
26
5
*
k

5
k
k
0.5
k
Total
assets
3,609
4,457
23,369
2,626
674
4,406
11

2,649

1,811

k

19,314
357
1,241
2,609
1,175
243
k
*

343
28
36
112
k
Long
term
debt
832
1,134
2,713
788
116
1,560
3

606

451

10

7,164
149
514
1,151
233
52
k
k

145
19
0
21
k
Tangible-
net worth
1,213
1,088
7,545
1,118
314
985
(4)

875

530

73

4,570
k
247
717
554
243
k
k

136
(3)
22
74
!<
                                    (continued)
       9-34

-------
                          TABLE 9-14 (continued)
aValues in parentheses represent negative numbers.
 Parent firms of furnace coke producers are listed first, followed by
 parent firms of foundry coke producers.  Subsidiaries are listed in
 parentheses below parent companies.
 EBIT = earnings be'fore interest and taxes.
 Cash flow = operating income + depreciation - interest expenses - taxes.
 Net working capital = current assets - current liabilities.
 Tangble net worth = equity - intangible assets.
9Received income tax credit in 1983.   Income tax represented as zero in
 cash flow calculation.
 McLouth Steel Corp. has debtor-in-possession status.  The parent company
 filed for bankruptcy in 1981, and filed a petition for reorganization in
 December 1984.   Financial information listed is for the subsidiary.
^Figures are interim values repoKed for first eleven months of 1984.
 Converted to 1983 dollars using GNP implicit price deflator.
k
 Information not available.
 A merger between National Intergroup, Inc., the parent company of National
 Steel Corp., and Bergen Brunswig Corp. fell through in April 1985, two
 weeks before its scheduled date.   Some market consultants feel that
 National Intergroup, Inc. is now a potential target for corporate raiders.
mNorthwest Industries, Inc., the parent company of Lone Star Steel,
 announced in April 1985 its merger with Farley Industries.
 Producer of both furnace and foundry coke.
 Financial information listed applies to subsidiary rather than parent
 company.
"Employees formally took control in January, 1984.   All figures are interim
 values reported for first three months of 1984.   Conversion to 1983
 dollars using GNP implicit price deflator.
qWheeling-Pittsburgh Steel Corp. filed for Chapter 11 in April 1985.
rOwned by James D.  Crane.   Financial  information denied.
sChattanooga Coke and Chemicals Co.,  Inc. has debtor-in-possession status.
 The company filed for arrangement under Chapter 11 in March 1984.
 Latest information available is for 1982.   Conversion to 1983 dollars
 using GNP inplicit price deflator.
                                  9-35

-------
     Ten companies show negative earnings before interest and taxes.   Of
these, nine are furnace coke producers,  whose earnings reflect the disas-
trous years for the steel industry.   As  mentioned,  in 1983,  steel  firms had
financial losses totaling $3.6 billion.   The balance of steel trade favored
imports by a 14 to 1 import-export ratio.  Imports  totaled 20.5 percent of
apparent supply in 1983.58
     Two integrated steel producers exhibit negative cash flows, while a
third has negative calculated working capital, as financial  resources have
dwindled with the recession.  Two companies, one furnace coke producer and
one foundry coke producer, are operating under bankruptcy status.
     From the financial data in Table 9-14, three ratios have been calculated
(Table 9-15).   The first, a liquidity ratio, is a measure of a firm's
ability to meet its current obligations  as they are due.  A liquidity ratio
above 1.0 indicates that the firm is able to pay its current debts with its
current assets; the hfgher the ratio, the bigger the difference between
current obligations and the firm's ability to meet them.  All of the coke-
producing firms have liquidity ratios between 1.0 and 4.0, with the excep-
tions of McLouth Steel (0.17) and McWane, Inc. (5.80).  These figures are
consistent with liquidity ratios for firms in a wide variety of manufactur-
ing industries.
     The second ratio, a coverage ratio, gives an indication of the firm's
ability to meet its interest payments.   A high ratio indicates that the
firm is more likely to be able to meet interest payments on its loans.
This ratio can also be used to determine the ability of a firm to obtain
more loans.  The coverage ratio of the coke-producing firms ranged from 0.8
to 3.9.  Seven firms for which information was available had negative
coverage ratios due to negative EBIT values.  The positive ratios are
comparable to the coverage evidenced in  most manufacturing industries.  The
poor performance of those firms with negative ratios may be due to problems
in the steel industry.  However, many firms continue to make investments
funded through mergers, joint ventures,  and other means.
     The last of the ratios, a leverage  ratio, indicates the relationship
between the capital contributed by creditors and that contributed by the
owners.  Leverage magnifies returns to owners.  Aggressive use of debt
                                  9-36

-------
TABLE 9-15.  FINANCIAL RATIOS FOR COKE-PRODUCING FIRMS
Company name Liquidity ratio
Armco, Inc.
Bethlehem Steel Corp. •
Ford Motor Co.
(Rouge Steel)
Inland Steel Co.
Interlake, Inc.
The LTV Corp.
McLouth Steel Corp.
(New Boston Coke
Corp. )
National Intergroup, Inc.
(National Steel Corp.)
Northwest Industries
(Lone Star Steel Co.)
Shenango Furnace Co. ,
Inc. (Shenango, Inc.)^
U.S. Steel Corp.
Weirton Steel Corp.
Wheel ing-Pittsburgh
Steel Corp.
Jim Walter Corp.9
Koppers Co. , Inc.9
Alabama Byproducts Corp.
Citizens Gas and Coke
Utility
Detroit Coke Corp.
Indiana Gas and
Chemical Corp.
1.56
1.27
1.05
1.42
2.16
1.41
0.17
1.40
1.80
1.59
1.22
1.79
1.42
1.09
2.15
3.17
1.32
1.08
1.18
Coverage ratioc Leverage ratio
-3.42e 1.52
-2.30s 1.95
3.82 1.73
-0.25s 1.20
3.08 0.93
-1.47s 2.91
f -3.75s
-2.85s 1.40
-1.66s 1.65
f 0.51
1.12 2.34
f f
-1.25s 3.06
0.81 3.64
1.59 0.86
f 0.35
f 1.52
f -10.3s
f 0.50
                                                  (continued)
                     9-37

-------
                           TABLE 9-15 (continued)
Company name3         Liquidity ratio     Coverage ratio0   Leverage ratio

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

aParent firms of furnace coke producers are listed first, followed by
 parent firms of foundry coke producers.   Subsidiaries are listed in
 parentheses below parent companies.  No ratios were calculated for
 Carondelet Coke Co^p.,  Chattanooga Coke and Chemicals Co., Inc. and
 Tonawanda Coke Corp. due to lack of information.

b, .   ....    ..      Current assets
 LiqmdHy ratio = current liabilities


Coverage ratio = Alnnua1 interest expense

d,           ...    Total  liabilities
 Leverage rat10 = Tang1b1e net wortn _

6Negative values are not meaningful.

 Information not available.

^Produces both furnace and foundry coke.
                                  9-38

-------
increases the chance of default and bankruptcy.   The chance of larger
returns must be balanced with the increased risk of such actions.   The
leverage ratio indicates the vulnerability of the firm to downward business
cycles.  Also, a high ratio reveals a low future debt capacity, i.e.  addi-
tions to debt in the future are less likely.   The firms with coke-making
capacity had leverage ratios that ranged from 0.3 to 3.7.  Six companies
had ratios below one, while one firm experienced a negative ratio.  These
figures highlight the poor financial condition of many firms in the coke
industry.  Currently, firms with coke-making capacity are engaged in substan-
tial amounts of debt financing, while continuing to make investments.
     Another measure of financial performance is the rate of return on
equity.  Studies of the iron and steel industry show low rates of return on
equity.  In an analysis performed by Temple,  Barker, and Sloane, Inc.
(TBS), the real (net of inflation) rate of return in the steel industry was
 •
estimated to be 0.2 percent for the period 1970 to 1980.   The TBS analysis
projected a rate of return on equity of 1.0 percent for 1980 to 1990.59
These estimates of historical and projected return on equity compare very
poorly with estimates of the required return on investment in the steel
industry.  A difference between realized and required returns implies that
equity financing of capital expenditures may be difficult.
     As noted, low rates of return on equity affect common stock prices and
have implications for future investment financing, including environmental
control expenditures.  The following data represent total pollution abatement
capital expenditures (PACE) as a percentage of new capital  expenditures
(NCE) for SIC 3312.2 60 61 62
               Year           Percentage PACE of NCE
               1975                   20.25
               1976                   20.92
               1977                   22.95
               1978                   22.20
               1979                   25.57
               1980                   20.11
               1981                   15.75
               1982                   12.32
PACE as a percentage of NCE peaked at 25.57 percent in 1979, after having
been fairly steady throughout the latter part of the 1970s.   The trend is
                                  9-39

-------
the 1980s shows a PACE as a declining percentage of NCE.   This decrease may
reflect the capital availability restrictions experienced by the steel
industry during this period.
     For the steel industry,  issuing new stock to raise investment capital
is unlikely under current circumstances.   If environmental and other control
investments cannot be financed through new equity, another source of funds
must be found.   Increased debt is one potential source.  However, firms
with coke-making capacity already have incurred substantial amounts of
debt.   The TBS analysis concluded that to avoid deterioration in its finan-
cial condition, the steel industry is likely to reduce expenditures to
                              •
modernize productive facilities rather than increase its external financ-
ing.63
     The steel  industry has had to resort to more creative forms of financ-
ing to provide funds for modernization of facilities.   This upgrading is
key to gaining and maintaining a competitive position with respect to
imports.  Cash flow for the industry has been below capital requirements
for the past two decades.  Mergers, joint ventures, shared production
arrangements, abandonment of uneconomic facilities, and the sales of assets
are likely to continue being sources of capital.64  Funds advanced by
customers, with repayment geared to earnings have been used for equipment
modernization.65
9.1.6  Industry Trends
     The demand for coke is derived from the demand for steel produced by
processes that utilize coke.   Hence, description of steel industry trends
in technological development and production are useful indicators of future
coke production and coke capacity requirements.
     As mentioned, there has been a technological shift toward labor-saving
technology which is expected to continue.  Trends in modernization are away
from open-hearth furnace production and toward electric arc furnaces and
basic oxygen furnaces.  In 1960, these processes accounted for 88.2 percent,
9.5 percent, and 3.3 percent of U.S. production, respectively, while in
1982 these values were 8.2 percent, 31.1 percent, and 60.7 percent.66  In
1985, basic oxygen furnaces are expected to account for 61.5 percent of
steel production, with electric furnaces contributing 34.0 percent or more,
                                  9-40

-------
and open hearth furnaces declining to 6.1 percent or less.67  Electric arc
and basic oxygen furnaces represent reductions' in production time, as well
as shifts to less expensive inputs.67  The increased use of these types of
furnaces will result in some decrease in demand for coke.
     Other changes have improved industry productivity, quality of yields,
and energy efficiency.'  In-ladle processes (performed after the melting
furnace stage) include inert gas stirring and vacuum treatments.68  These
techniques yield higher quality steel.
     The use of continuous casters, which convert molten steel directly
into shapes ready for rolling, has increased from 18 percent of production
in the late 1970s to 35 percent in 1984.69  Yield of finished product per
ton of raw steel may be boosted to 95 percent from the current 76 percent
by use of this process.   For each ton of finished steel produced using this
technology, 15 percent to 20 percent less raw steel is required, while
40 percent to 50 percent less energy is needed.67  The impact of these
technological developments on the coke industry is unknown.  Any effects
will be through productivity improvements in the steel industry.
     Technological trends have reduced steel use per unit of output of
durable goods.  Since the 1970s, the decline in consumption of steel per
dollar of gross national product has averaged 4 percent annually, with
continued decline expected.70  Increases in economic growth are predicted
to offset this effect, resulting in an increase in steel use to 95 million
megagrams by 1988, with domestic shipments representing a 5 percent average
annual rate increase over the 1983-1988 period.65  Projections by the
Bureau of Mines predict U.S. raw steel demand will be 138 million megagrams
in 1990, and 164 million megagrams in 2000.71
     Steelmaking capacity utilization has recently been low, averaging
47.3 percent in 1982 and 55.4 percent in 1983.72  In 1984, this rate rose
to 82 percent in April before dropping to 57 percent in September.73
Capacity utilization is an important measure of industry performance due to
high fixed costs for the industry.   The larger the volume of production,
the smaller the cost per unit of steel produced.   For the steel industry,
the breakeven point for operations is at approximately 65 percent of capa-
bility, though this figure is highly dependent on prices.73  This means
that steel companies have been operating at losses for several years.
                                  9-41

-------
     The steel industry has responded to this financial  difficulty by
permanently reducing capacity.  Since 1983, this reduction has been more
than 10 percent, with perhaps another 5 percent cut necessary.73  From
122.5 million megagrams in 1984, capacity is likely to be trimmed to
109 million megagrams by the late 1980s.73  However, the Bureau of Mines
predicts U.S. production of raw steel will rise to 113 million megagrams in
1990, and to 132 million megagrams in 2000, under assumptions of slow
growth in the rate of production, and increases in demand.71  Changes in
capacity utilization affect coke production only to the extent that coke is
an input to the steel production process.  Reductions in steel production,
coupled with shifts to non-coke energy input's could greatly reduce demand
for coke.
     The emergence of mini mills to supply regional demand for steel has
"had an impact on the operation of the larger integrated steel mills.  Mini
mills now account for approximately 20 percent of U.S. steel production, at
a cost per ton of installed capacity about 75 percent less than for inte-
grated plants.  The use of electric arc furnaces in mini mills may result
in dramatic reductions in coke demand if the minis claim 40 percent of the
steel market by 2000, as some predict.74
     The combination of the factors described in this section indicate that
coke consumption is destined to continue declining.  Technological improve-
ments are likely to result in an input shift away from coke, while reduced
capacity in the integrated steel industry signals a decrease in amounts of
coke needed for blast furnace steel production.
9.1.7  Market Behavior:  Conclusions
     Market structure, financial performance, and potential growth influence
the choice of a methodology to describe supply responses in the coke-making
industry.  Although some characteristics of this industry indicate a poten-
tial for market power, other characteristics belie it.
     Some concentration exists in coke-making capacity and steel produc-
tion; however, many firms produce coke and iron and steel products.  Vertical
integration is substantial; however, integration appears to result primarily
from a desire for increased certainty in the supply of critical inputs.
Furthermore, substitution through alternative technologies and coke imports
                                  9-42

-------
is feasible, and some substitutes for the industry's final products (iron
and steel) are available.  In any industry, the potential for substitution
is a major factor leading to competitive pricing.  Certainly, the financial
profile of coke-making firms is not indicative of monopoly profits.  Pros-
pects for industry growth are limited.  An individual firm must actively
compete with other firms in the industry to improve its profit position, or
even to remain viable.
     No industry matches the textbook definition of perfect competition.
The important issue is whether or not the competitive model satisfactorily
captures major behavioral responses of firms in the industry.  Based on the
factors outlined in this section,'the competitive pricing model adequately
describes supply responses for coke-making firms.
9.2.  ECONOMIC IMPACT OF REGULATORY ALTERNATIVE
9.2.1  Summary
     Economic impacts are projected for Regulatory Alternative II for two
unit compliance cost scenarios:  mid-range cost estimate and upper-bound
cost estimate.  Furnace and foundry coke impacts are examined separately
because their production costs and markets differ.  All cost and price
impacts are in second-quarter 1984 dollars.  Where necessary, conversion to
1984 values were made by multiplying 1979 values by 1.362, the ratio of
1984 second-quarter GNP implicit price deflator to the 1979 GNP implicit
price deflator.14 75  The 1983 values were converted by multiplying by
1.032, the ratio of the producer price index for second-quarter 1984 to the
same index for 1983.76
     When measured on a per-unit of output basis, the costs of meeting
baseline regulations for foundry coke plants tend to be greater than those
for furnace coke plants for two reasons.   First, some economies of scale
are present for some of the controls.   Foundry plants tend to be smaller
than furnace plants, thus, they have higher per-unit control costs.  Second,
for a given battery, foundry coke output will be less than furnace coke
output because foundry coke coking time is about two-thirds longer than
furnace coke coking time.
     Regulatory Alternative II has annualized compliance costs of $14 to
$17 million above baseline for furnace and foundry coke producers combined.
                                  9-43

-------
Regulatory Alternative II requires capital  expenditures of $9 to $13 million
above baseline for furnace and foundry coke producers combined.   These
costs differ from engineering estimates due to the calculation of costs
based on batteries with marginal cost of production below price, rather
than all batteries.
     Price impacts are estimated under the empirically supported hypothesis
that furnace coke demand is responsive to higher coke prices.  Foundry coke
demand is assumed to respond to price.  Regulatory Alternative II would
have maximum impacts of $0.25/Mg (0.2 percent change) on the price of
furnace coke, and $1.34/Mg (0.8 percent change) on the price of foundry
coke under Import Scenario A (1984 dollars).  Under Scenario B there are no
foundry coke price effects.
     Regulatory Alternative II would have less than a I percent negative
impact on the production of either furnace or foundry coke under Scenario A.
Under Import Scenario B, Regulatory Alternative II would decrease foundry
coke production by 2.8 percent.  There are 14 furnace coke batteries that
currently appear uneconomic.  There are no uneconomic foundry coke batteries.
Regulatory Alternative II would place one more furnace coke battery into
the uneconomic production region.
9.2.2  Methodology
     The following approach focuses on the long-run adjustment process of
furnace and foundry coke producers to the higher costs of coke production
that the regulatory alternatives will create.  These long-run adjustments
involve investment and shutdown decisions.   Short-run adjustments, such as
altering coking times, to meet the fluctuations in the demand for coke are
not the subject of this analysis.
     Because of differences in production costs and markets, furnace and
foundry coke producers are modeled separately.  Both are assumed to behave
as if they were competitive industries selling coke in a market.  This
assumption is somewhat more realistic for foundry than for furnace coke
producers because most furnace coke is produced in plants captive to the
steel industry.  However, interfirm and intrafirm shipments of coke are not
uncommon, as can be inferred from Table 9-11.  A plant-by-plant review of
the coke industry by Hogan and Koelble also confirmed the existence of such
exchanges.77
                                  9-44

-------
     A set of programmed models has been developed to produce intraindustry
and interindustry estimates of the economic impacts of the alternative
regulations.  The models are applied to both furnace and foundry coke, and
the sectors included are coke, steel, and ferrous foundries.  The rest of
the economy is incorporated into the interindustry portion of the analysis.
     The analytical approach incorporates a production cost model of the
coke industry based on engineering data, and an econometric model of the
steel industry.   The interrelationships of these models for furnace coke
are shown in Figure 9-3.  The upper portion of Figure 9-3 encompasses the
supply side impacts of the regulatory alternatives; the lower portion con-
tains the demand side impacts.  In the synthesis step, the two sides are
brought together and the equilibrium price and quantity relationships are
determined.   An analogous diagram for foundry coke would substitute ferrous
foundry products for steel.  The methodology is described further in the
following subsections.
     9.2.2.1  Supply Side.   A production cost model that incorporates
technical relationships and engineering cost estimates is used with plant-
specific information to compute separate industry supply functions, with
and without additional  controls.78  Supply functions are estimated on a
year-by-year basis for furnace and foundry coke plants projected to be in
existence between 1984 and 1995.  Both coke production costs and the costs
that plants incur to meet existing environmental regulations are computed
to estimate the industry supply curve before any additional controls are
applied.  Estimates of the costs of control for compliance with the regula-
tory alternatives are used to compute the projected upward shifts in that
supply function.  All costs are in 1984 dollars.
     This approach provides a method of estimating the industry supply
curve for coke,  which shows the alternative coke quantities that will be
placed on the market at alternative prices.  When the supply curve is
considered in conjunction with the demand curve, an equilibrium price and
coke output rate can be projected.   Supply curve shifts caused by the
regulatory alternatives can be developed from the compliance cost estimates
made by the engineering contractor.   These new supply functions, along with
the demand curve, can then be used to compute the equilibrium price and
output rate under each  regulatory alternative.
                                  9-45

-------
                                           — c_
                                        — o <:

                                              g
                                              U.J
                                        OO UJ LU
                                        I— Si H-
                                        o o oo
                                        < u
                                        c_    o
                                        00


                                        OO
                                        OO
                          1 00
                        <: z
                        cr O
                        C —


                        il
                        — 2
                        UJ OO
                        S3 oO
o
I—' >•

z c.
^ c_

l_l WO
O
»•••«
H-
oo
                   O
                   o
                                            c:
                                            O uj
                                            l_ ^
                                               O
                                            O <_5
                                                              0)
                                                              •a
                                                              o


                                                              I
                                                              o.
                                                              E
                                                                                                        E
                                                                                                        o

                                                                                                        o
                                                                                                        u
                                                                                                       111

                                                                                                       CO
                                                                                                       O)
                                                                                                        0)

                                                                                                        O)
             cr
             O
          CJ
                oo
                z
                O
             I— —
             Z t—
             <: <
             _/ cr
             e. =>
               00
            Z =5
            CJ
            — C£
            UJ O
            c: u_
            o
            O —I
            U_ UJ
               UJ
            C3 >—
            Z 00

            S oo
 >£.
 C  Z
 t-;  o
•—  O
oo  cr
O  —
 H—  < OO
 00  H- Z
 —  z o
 >C  UJ —
 uj  Z v—
    z <
 u.  O —'
 O  ee 3
    — t;
 00  > UJ
 H-  Z C=
 00  UJ
 C
      oO
      UJ
u_ >- >
O cs —•
   O l—
oo t— <:
i— < z
00 _J SL
O 3 uj
00^—
   UJ —I
O
Z —I
< UJ
S UJ
UJ h—
O oO

Z 00
uj ce
e: O
O u_
                            o  '
                            Z —J
                            < UJ
                            X uj
                            UJ I—
                            O OO

                            O oo'
                                                                                           on sr
                                                                                           uj O
                                                     9-46

-------
     9.2.2.1.1  Data base.  Plant-by-plant data on over 60 variables for fur-
nace and foundry coke plants in existence in 1979 were compiled from govern-
ment publications, industry contacts, and previous studies of the coke
industry.  The data base was sent to the American Iron and Steel Institute,
which submitted it to their members for verification, corrections, and
additions,79 and to the American Coke and Coal Chemicals Institute.   The
data were adjusted to account for the 1984 plant inventory in the reanalysis.
Capacity, number of ovens, and status (hot idle, cold idle, under construc-
tion, or online) were updated for each battery.5
     9.2.2.1.2  Output relationships.  For a given battery, the full capac-
ity output of coke, measured in megagrams per year, is dependent on the
nominal coal charge (megagrams of coal per charge) per oven, the number of
ovens, and the effective gross coking time (net coking time plus decarboni-
zation time).  The following values for effective gross coking time were
used except where plant-specific values were available.78
                                      Furnace          Foundry
                                       coke             coke
             Wet coal                 18 hours        30 hours
             Preheated coal           13 hours        24 hours
An age-specific outage rate that varies from 4 to 10 percent is assumed to
allow for normal maintenance and repair.  Thus, the model assumes some
increase in such costs as plants age.
     The quantities of by-products produced are estimated from engineering
relationships.   These quantities depend on the amount of coal carbonized,
percentage of coal volatile matter, coking time, and configuration of the
by-product facility at a plant.  The by-products included in the model are
coke breeze, coke oven gas, tar, crude light oil, BTX, ammonium sulfate,
anhydrous ammonia, elemental sulfur, sodium phenolate, benzene, toluene,
xylene, naphthalene, and solvent naphtha.   All plants are assumed to produce
breeze and coke oven gas.
     9.2.2.1.3  Operating costs.  The major costs of operation for a coke
plant are expenditures for coal, labor, utilities,  and chemicals.   The
activities within the coke plant were allocated to five production and ten
                                  9-47

-------
environmental control cost centers (Figure 9-4) to facilitate the develop-
ment of the operating cost estimates.
     Coal is the major operating cost item in coke production.   Plant-
specific estimates of the delivered price of coal were developed by identi-
fying the mine that supplies each plant and estimating transportation costs
from the mine to the plant.  When it was not known which coal mine supplied
a particular plant, it was assumed that the coal came from the nearest
mines supplying coal of the same volatile matter and ash content as that
used by the plant.  Transportation cost estimates were based on the dis-
tances traveled and the transport mode (barge or rail) employed.
     Maintenance labor and supervision requirements were estimated for 69
jobs within the coke plant.  Primary variables that determine the number of
maintenance labor and supervision man-hours needed include type of plant
(merchant or captive), number of battery units, number of plants at a site,
size of by-product plant, type of coal charge (wet or preheated), and coke
production.  The 1979 labor rates used for captive plants are $23.21/h for
supervisory positions and $21.38/h for production labor.  For merchant
plants, rates of $21.52/h and $19.61/h are assumed.  These values are
scaled by the GNP  implicit price deflator to 1984 dollars.
     The major utilities at a coke plant are steam, electricity, water, and
other fuels.  Utility requirements were estimated from the data on the
plant configuration and output rates for coke and the by-products.  The
prices used for the utilities are $7.41/103 Ib steam; $0.037/kWh electric-
ity; $0.22/103 gal cooling water; and $3.76/106 Btu underfire gas.  These
values are 1979 figures scaled to 1984 dollars by the GNP implicit price
deflator.
     9.2.2.1.4  Capital costs.  Although no net additions to industry
coke-making capacity are anticipated during the 1984 to 1995 period, a
number of producers had plans to rebuild or replace existing batteries in
1979.  In 1984, three new  batteries had been constructed and one was under
construction.5  Such actions alter the long-run industry supply curve
because the new batteries  typically have lower operating costs per unit of
output than the batteries  they replace and, most  importantly, their capital
costs will be reflected in the new supply curve.
                                  9-48

-------













r

























































*
R
^ 1








—












































W































i












-


•r
.
•r
•r


























•^
—4














m
i


























J
•M
!••




i
*







































t
s «
c
"1


^


1


































—
—







S

•i











riji
i?i !_
Ǥrii
? SI I
| S! •
| c J

"• — J
-« 1
fi 1
IS |—
w w |

\ — J














.
i
w
*^H — '

1





























*


















/
1 	 \


|
x
5









i 1
s a
^ E
« E
H
* s
r "
is
a5



H
«_*

i





o5

1
|s
I5
-• I'

£

\^^St
f
T
1 	
y
i
ill
9 •" w»
•r

1











•r
w















m\
5
• Si

^-






s *«



£
«
»
X
t


;


















e
§M

5
—
•e
6
8
a



	






— II
r~s
1
•r *
M 7 V'
H
m
~
;


s g
S S




























f






R
M
i
,»
•M

•r
§""
•




^«
5,i
^|



































—








,

a-
r "••
j!

r
L


















e










	 —










5,
1
it

•c
•s
1
"5=
6*a
1
i
1
w-r
°|
** w
w **
1
!
«
£g

^


II
|
«

t
E

5
*

i
!










I!
-. y

I S | fa
1111
i i i J
uJ.i 	 L

5|
_rin L.D
"^iii* s'-
HE 1 i *
IS S

1






_



§5


"
?" §
5
S






-1
f w V fc
£5 —•• » «
£= $s
-^ ll
T j
1 *
. fSo

5 S

















£
(!)
4-1
C
8
*•<
0
u
CO
a
r—~ s
Jc
Sgf 0)
L __^ 3
" .2*
iZ


«
5;
=
«l










•
9-49

-------
    _The capital cost breakdown for new plants is shown in Table 9-16.   For
such plants, the major capital  cost items are the battery, quench tower,
quench car, pusher machine, Tarry car,  door machine and coke guide,  by-
product plant, coal handling system, and coke handling system.   A 60-oven
battery is assumed.  Pipeline charging  can increase the coke-making  capacity
of a given oven by about 25 percent by  reducing gross coking time.   Conse-
quently, the per-unit operating cost is reduced.   The capital costs  show
economies of scale, i.e., larger plants have smaller per-unit-of-capacity
capital costs.  The capital cost per unit of capacity is higher for  pipeline-
charged batteries than for conventionally charged batteries.
     Periodically, batteries must undergo major rehabilitation or rebuilding
because of performance deterioration.   The costs  of pad-up rebuilds  will
vary from site to site depending on battery maintenance, past operating
practices, and other factors.  Average  estimates  of the cost of rebuilding
were developed for this study and are shown in a report by PEDCo.81   The
economic life of coke-making facilities is subject to considerable variation
depending upon past maintenance and operating practices, which also  affect
current operating costs.  For this study, 25 years was used as the average
preferred life of a new coke-making facility; however, many batteries are
operated for 35 to 40 years.  If 35 to  40 years is a more reasonable battery
lifetime, use of a 25-year lifetime will result in some overestimation of
the annual costs of new or rebuilt facilities.  However, firms will  probably
not plan or expect to wait 35 to 40 years to recoup an investment in coke-
making capacity.
     9.2.2.1.5  Environmental costs.  Plant-specific estimates of the
installed capital and operating costs for current environmental regulations
are incorporated in the model.   The current regulations include workplace
standards (Occupational Safety and Health Administration [OSHA]), water
quality regulations best practicable technology [BPT] and best available
technology [BAT], and State implementation plan (SIP) requirements.   Compli-
ance expenses incurred for all  plants in the data base for each of the
current regulations assumed baseline control costs were estimated.   Costs
to comply with OSHA and BPT water requirements under the Federal Water
Pollution Control Act were assumed incurred by 1981.  Costs for all  other
baseline environmental regulations were assumed to be incurred by 1983.
                                  9-50

-------
          TABLE 9-16.  ESTIMATED CAPITAL COSTS OF NEW PLANTS80
                                   Conventionally
                                   charged battery
   Pipeline
charged battery

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

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

7.92
3.20
2.28
2.10
39.75
23.60
8.80
1.80
$141.20
64.60
2.45

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

7.92
3.20
0
2.10
43.74
26.70
10.00
1.91
$182.12
 In the production cost model, new foundry batteries were assumed to be
 4-meter batteries and new furnace batteries were assumed to be 6-meter
 batteries.
                                  9-51

-------
     The scatter diagrams in Figures 9-5 and 9-6 show estimates from the
coke supply model of average total  cost of production in 1984,  including
environmental costs, for all furnace and foundry coke plants.   A weak,
inverse relationship between the average cost of production and the size of
the plant is evident in Figures 9-5 and 9-6.   However, a number of other
factors create variability in the average cost of production across coke
plants.  The most important of these factors are the delivered price of
coal, the age of the plant, and the by-products recovered.
     9.2.2.1.6  Coke supply function—existing facilities.   The operating
and capital cost functions were used to estimate the cost of production,
including relevant environmental costs, for all plants in the data base.
This cost does not include a return on investment for existing facilities.
The capital costs for these facilities have already been incurred and do
not affect operating decisions.
     Capital costs that have not yet been incurred are annualized at 6.2
percent, which is estimated to be the real (net of inflation) cost of
capital for the coke industry.  (This percentage is an after-tax estimate.)
This figure, which was estimated from data on the capital structure for
publicly owned steel companies, has been used in this study as the minimum
acceptable rate of return on new facilities.82
     The capital costs of the regulatory alternative are annualized over
the life of the control equipment (20 years).  This action is tantamount to
assuming either that, all ovens have a remaining life of at least 20 years
or that the control equipment is salvageable.
     The supply function for each plant is estimated as follows:  the
average cost of production is computed for each battery in the plant; these
batteries are arranged in increasing order of their average costs of produc-
tion and the output for each battery is accumulated to produce a stepped
marginal cost function for the plant; plant overhead costs are averaged for
all relevant plant output rates; and average total costs are computed for
each output rate by summing the average costs for plant overhead and the
battery.  Each plant's supply function is the portion of the marginal cost
function above the average total cost function.  For existing plants where
the average total cost exceeds marginal cost over the entire range of
                                  9-52

-------
n











D
D

D





n
n


D
n '

PV-i
1^^*
~ "\-J
D n
D
u
cP

ff a


D 0
n D
n

D n

s ^
—

O
«• .
CM


^

CO
T-

r ' >
L_
(^> ^~
~ ' X\
*"" C^
•«r-
_ -*Z
J. °
-TO
"" .
o
o
CSJ °.
,_ T—
*^~^
^B
— L_
O 13
~ — 0
3
O
CO Q_
" o"
^


o
•^

	 1_


S
O)
*~
c
o
ts
3
•o
2
a
c
J2
Q.
>*-
o
c
o
'^
c
3
to
13
o
3
•a
2
a
8
8
CO
c
3
"o
8
{O
aS
(0
1
ie
E
tt
HI
in
0)
w
3,
       a
f-f ij
j^
™»(
i
O
I
O
1
o
i
o
i
o
o
•
c
uoipnpojd
      9-53

-------
1
a
i

a









D
















D


a

D


CS
a
a
a



o
- 6







o
"
- c
UL>





^



q
- o
m
- 1
r^ rc
V^1 
-------
output, the supply function is the point on the plant's average total cost
function represented by capacity output (after allowing for outages).  The
aggregate long-run supply function for all currently existing coke plants
and batteries is obtained by horizontally summing the supply function for
each plant.   The 1984 industry marginal cost (supply) curves for existing
furnace and foundry coke plants are presented in Figures 9-7 and 9-8,
respectively.
     9.2.2.1.7  Coke supply function—new facilities.  The cost of coke
production for new furnace and foundry batteries was estimated from the
engineering cost model, assuming the new model plants described previously.
These costs include the normal return on investment and allowances for
depreciation and corporate income taxes.  When expressed on a per-unit
basis, these costs are the minimum price at which it is attractive to build
new facilities.
     9.2.2.2  Demand Side.  The demand for coke is derived from the demand
for products that use coke as an input to production—primarily steel and
ferrous foundry products.  A demand function for furnace coke was derived
by econometrically modeling the impacts of changes in furnace coke produc-
tion costs on the steel industry.83
     The econometric model of the steel industry has two sectors:  steel
and coke.  The steel sector includes domestic steel supply, steel imports
and exports, and steel consumption (steel supply plus imports minus exports).
Similarly, the model of the coke sector consists of domestic coke supply,
domestic coke demand, and coke imports and exports.  The two sectors are
linked by a derived coke demand function, which includes as variables steel
production, steel price, and quantities and prices of other inputs to steel
production.   The domestic supply of steel is assumed equal to domestic
demand for U.S. steel plus world demand for U.S. steel minus U.S. import
demand.
     Both linear and nonlinear specifications were used to estimate the
steel-sector model.   Two-stage least squares was used to estimate the
different components of the steel sector.  Visual inspections of the corre-
lation matrix and a plot of the dependent variable versus the residuals
indicated no multicollinearity or heteroscedasticity problems.   The Ourbin-
Watson statistic showed no evidence of autocorrelation.
                                  9-55

-------

I \
V. \
•^ 1
1 1
! 1
( \
--
1
\
1
\



\
\
\\
\
i 1
' \
\
I I
I I


\
1 \
i \
) \
V \
1 \
\
I
\
\ \
\ \
U
\\
t N\\
\

i i i i i i i i i i
3OOOOOOOOOOC
"> " /TS CD r*^ <<;~v 10 "i~ r^ ^i — <"^ /*
Mif «p»* ^** » «^ ^» « * • • * « S.^ ta
N| ^— T— T— T— ^— ^— T— T— •»— *—

O
•
- co

o
*
CM


••*
o
•
- o



-5


™*


q
- 
- CO
>v> 05
O *"
oT
£& 8
y1^. fl}
^V i 	 fO
	 *^ £•
^r^ fli ^
2 > -2
< 0
O *"
0 i i
q e
o 1 §
o t
0 . S
o
v— 0)
O)
**— •* CO
03
c s
2 1
-!_• CO
^ "to
3 .=
T? 2*
2 i
Q."S rf
c> ®
CN 99
^^ ka
__ .5"
Q "•
^C

L.
V
ZE

i
i
1
i





9-56

-------
                                                                    CN!
                                                                    
                                                                          O
                                                                          i_
                                                                          Q.
                                                                                 S
                                                                                 t   «
                                                                                 C
                                                              _o
                                                              +3
                                                              O

                                                              3
                                                              O
                                                              O
                                                              C3

                                                              s
                                                              (O
                                                              T5
                                                              re
                                                              15
                                                              c
                                                              '§>
                                                              re
                                                          -   2
                                                          (O   co
                                                          Q   dj
                                                          Q   2
                                                          _   a
o
O
O

O
O
O
                                9-57

-------
     The econometric estimation of the coke sector was complicated by the
small share of total domestic production that is traded in the market.   The
fact that very little coke is actually sold creates concern over the reported
price of coke.  Therefore, estimates of the implied price of coke were
developed, based on the value of coke in steelmaking,  and used in the
estimation of elasticities.84 85  Estimates of elasticities for coke and
steel functions are presented in Table 9-17.   Actual prices for coke pro-
duced and used internally by the producing companies were used in the
reanalysis.
     An attempt was made to derive a demand function for foundry coke in an
                                                                            •
analogous manner.   However, the relevant coefficient estimates were not
statistically significant at a reasonable level.  A direct estimation of
the demand function, based on the prices of foundry coke, foundry coke
substitutes, and complementary inputs, was also attempted.  Unfortunately,
the data necessary to properly estimate the demand function was not readily
available from published sources.  Accordingly, the elasticity of demand
for foundry coke was estimated based on the theoretical relationship between
the production function for foundry products and the derived demand function
for inputs to foundry production.  This elasticity calculation is based on
a 3-year average of the cost share of foundry coke in foundry production.
This estimate is presented in Table 9-17.  This elasticity assumes U.S.
demand for foundry coke is supplied entirely from domestic production
(Scenario A).
     Another scenario is that imported foundry coke competes with the
domestic product in the open market (Scenario B).  A simplifying assumption
is that they are perfect substitutes in the production processes which
utilize foundry coke.  In this case, a reduction in U.S. supply is compen-
sated by imports,  so that price need not rise if the quantity of imports
purchased is increased.  Both scenarios are examined in the reanalysis.
     9.2.2.3  Synthesis.  Separate linear functions were fit to the furnace
and foundry coke marginal cost values depicted in Figures 9-7 and 9-8.  As
illustrated in Figures 9-9 and 9-10, each supply function is used with the
demand function for the appropriate type of coke to compute the initial
equilibrium price-quantity values (Px and Qx in Figures 9-9 and 9-10).   In
                                  9-58

-------
      TABLE 9-17.   ESTIMATES OF ELASTICITIES OF STEEL AND COKE MARKETS

                                                 Point         Interval
                                                estimate       estimate

1.    Percent change in furnace coke demand       -1.29
     for 1 percent change in the price of
     furnace coke

2:   Percent change in foundry coke demand       -1.03C            --c
     for 1 percent change in the price of
     foundry coke

3.    Percent change in import demand for 1        1.88       (-1.68, 5.44)
     percent change in the price of furnace
     coke

4.    Percent change in price of steel for         0.11
     1 percent change in the price of
     furnace coke

5.    Percent change steel demand for 1           -1.86d      (-0.54, -3.18)
     percent change in the price of steel

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

Note:   Estimates are based on the empirical analysis using annual  data for
       the years 1950-1977 with a structural econometric model of steel  and
       coke markets.

 Interval estimates are based on 95 percent confidence level.
b
 Derived from the production function for steel, and input cost shares.

 Calculation based on the theoretical relationship between input demand
 elasticity and input cost share in the production of outputs.  Accord-
 ingly, no interval is provided.
 Significantly different from zero at 1 percent level of statistical
 significance.
                                  9-59

-------
S/Mg
     P2




     PI
                                       Q2      Ql
                                                              10   M   / Yr
                   Figure 9-9. Coke supply and demand without import competition.
                                      9-60

-------
$/Mg
     PI
                                                                       S2
                                                                        SI
                            Q2
Ql
                                                              10   Mg /Yr
                    Figure 9-10. Coke supply and demand with import competition.
                                     9-61

-------
*the  case  where  imports  are  not  perfect  substitutes  for  domestically  produced
 coke,  the supply  function  is  reestimated  for  each regulatory  alternative
 (S2  in Figure 9-9),  and the new equilibrium price-quantity  values  (P2  and
 Q2  in  Figure 9-9)  are predicted.
     The  case where  imports compete  with  domestically produced  foundry coke
 is  shown  in Figure 9-10.   As  in Figure  9-9, the  supply  curve  shifts  upward
 to  S2.  However,  because imports are available,  no  change  in  price and
 quantity  need be  experienced  by the  consumer.  Though domestic  production
 is  reduced by Qi-Q2, the share  of the market  supplied by  imports  increases
 by  this same amount.  The  new equilibrium price  and quantity  for  domestic
 coke are  Px and Q2 in Figure  9-10.
     9.2.2.4  Economic  Impact Variables.   Table  9-18 shows  the  specific
 economic  variables for  which  impacts are  estimated.   The  methodology pre-
 sented previously was designed  to provide industry-level  estimates of these
 impacts.   The conventional  demand and supply  partial equilibrium  model of  a
 competitive market was  chosen for this  analysis  because it  was  believed to
 represent the key characteristics of the  coke market and  many of  the impacts
 of  interest can be readily estimated from this model.
      Figure 9-11  represents the markets for  furnace coke  and  for  foundry
 coke which is free of  import competition  (Scenario  A).   In  Figure  9-11,
 area abQIO represents the  costs of coke production, are aOPlbQl the  total
 revenue of producers, and  area  Plba producer  surplus—the difference between
 total  revenues  and the  opportunity costs  of  production.   Area Plcb represents
 the total benefits of consumption,  PlbQlO total  expenditures, and area Plcb
 consumer  surplus--the difference between  total benefits and total  expendi-
 tures. With  the  regulatory alternatives  total cost change  by adeh - Q2ebQl.
 Producer  surplus  changes by PlP2ef - deha -  gbh.  Consumer  surplus changes
 by  -PlP2eb.   Cancelling gains to producers that  are offset  by losses to
 consumers leaves  the cost of the regulation  as adebh.   Compliance costs are
 adeh.
     When imported coke is a perfect substitute  for domestic  coke, and
 since  imported  coke  is  assumed  to be available at price P.,, domestic consum-
 ers purchase  more imports  and less domestic  coke (Figure  9-12).  The results
 are that  domestic production decreases  to Q2, imports increase by Qi-Q2,
                                   9-62

-------
        TABLE 9-18.   ECONOMIC IMPACT VARIABLES AND AFFECTED SECTORS
Variable
Price
Output
Profits
Costs
Plant closures/openings
Capital requirements
Factor employment
Labor
Metallurgical coal
Imports

Furnace
coke
X
X
X
X
X
X
X
X
X
Sector
Foundry
coke
X
X
X
X

X
X
X
xa

Steel
X
X





X
Impacted under Scenario B.
                                 9-63

-------
$/MG
                                                                  MG/YR
             Figure 9-11. Economic impact variables without import competition.
                                         9-64

-------
$/MG
                      Q2
Ql
MG/YR
              Figure 9-12. Economic impact variables with import competition.
                                        9-65

-------
and price remains at P1.   The costs of the regulation to producers are
represented by area adij.   Compliance costs are adik.   There are no costs
to consumers because they purchase quantity Qt at price Px as they were
before the regulation.
9.2.3  Furnace Coke Impacts
     As described in Section 9.2.2 of this analysis,  the furnace coke
industry has been modeled as a competitive industry supplying coke to the
steel industry.   This definition implies the existence of interfirm and
intrafirm shipments of coke.  However, no allowance has been made for coke
transportation costs, although coal transportation costs are included in
the cost of coke production estimates.  Coke plants and their associated
steel mills are typically clustered together.   As noted in Section 9.1.4.1,
most coke is consumed within the region where it is produced.  Hence,
transportation across great distances is uncommon.   Therefore, the omission
of coke transport costs should not greatly influence the calculations.
     The baseline values for 1983, presented in Table 9-19, are actual data
for 1983, except for coke prices, which are calculated by the model.   The
values for 1983 are assumed to reflect full compliance with applicable SIP
and OSHA air quality regulations and water quality regulations.  The coke
supply model was used to compute the price of furnace coke, costs, revenues,
and profits, given these actual values.  Coal consumption and employment
projections were made using current coal- and labor-output ratios.  The
supply function was reestimated assuming control levels being practiced in
1984 for all emission sources.  This estimation was used to determine the
impacts of moving from baseline industry control levels to alternative
regulations control for all sources.
     Table 9-20 presents total costs incurred by companies in SIC 3312 in
meeting environmental regulations up to 1983.  These costs represent indus-
try efforts to achieve baseline compliance.  Expenditures are segmented by
type of pollutant treated.   Total cost for abatement increased throughout
the late 1970s and peaked at $956.5 million (1972 dollars) in 1979.  Expend-
itures declined slightly in 1980 and 1981, and dropped to an eight-year low
of $549.8 million (1972 dollars) in 1982.
                                  9-66

-------
          TABLE 9-19.   BASELINE VALUES FOR ECONOMIC
         IMPACT ANALYSIS—FURNACE COKE, 198337 86 87

                                            Baseline values3

Coke market

  Price (1983 $/Mg)                                106.25b
  Production (103 Mg)                            20,462
  Consumption (103 Mg)        '                  24,380
  Imports (103 Mg)c.                             3,918
  Employment (jobs)                              6,236
  Coal consumption (103 Mg)                     29,787

Steel market

  Price (1983 $/Mg)                                319.97
  Production (103 Mg)                            76,763
  Consumption (103 Mg)                          75,710
  Imports (103 Mg)                              15,486
  Employment (jobs)6                           295,000

 Baseline assumes companies meet existing regulations
 including OSHA (coke  oven emissions); State regulations on
 desulfurization, pushing, coal  handling, coke handling,
 quench tower, and battery stack controls; and BPT and BAT
 water regulations.

 Calculated.  Market price for furnace coke was $123.51 in
 the fourth quarter of 1983.

 Calculated.  Imports  = Consumption - Production.

 Calculated.  Furnace  coke employment = Employment in
 byproduct coke industry x Proportion of coke production
 represented by furnace coke sector.

Represented"by employment in SIC 3312 (Blast furnaces and
 steel mills).
                             9-67

-------













s
CM
rH
CO
CO

O
00

0£.
O
U_

oo
LU
f*S
*""3
h-
HH
O
Ty
LU
Q_
LU

t—

LU
?>"
LU
H-

CQ


rzr_
o
i— i
h-
~*i
—i
— i
o
Q.


,
0
CM
1
cn

LU
— 1
CO

•P o to
C U 0
O) Ol rH
s- o; • — '
^
3
u

CO
0 T3
i— 1 T3 Ol
v_/ -1- -P
r— CO
in o 03
Ol GO 3
S-
^
•4_>
• r—
•a u
c c
O) O
a. s- T-
X <3J t "^
Ol -P 3
03 .—
•P 15 "—
c o
01 a.
E
Ol
+j
03
f} C
< 0

5— 

- co r^ ^« i-D rH . cn o CO rH CO r- 1 rH , CO <5" 00 UD 0 CO CO , p*^ f^s. ^^ LO f^ cn rH 0 CM CM p^. O t O LO cn CO r-». rH «^" r-H LO LO cn 0 CO cn • i PI o t QJ t o f*^. cn r-l ^" CM rH CO CO CO CM rH rH rH CO rH LO . LO CO CO LO CO CO rH . LO r*>. I O pv. p*^ cn rH CM ^j- CM 'CO rH . O CO C\J rH rH LO rH CM T-H' rH rH cn ,3- CM ^J- rH . cn o r- co P*» cn rH uO tp uO cn p>* . CM «^- LO rH O rH LO , cn cn i-O rH rH LO **D , CM CO cn cn p*^ cn rH LO 1^3 LO CO ^~ . cn CM LO rH CO CO rH cn p^ CM rH cn ^j. cn ^^ ^" . LO CM cn 0 00 cn rH cn P>, o CO CM . •=!- CO LO rH CO CM CM 0 CO LO rH U3 CM rH LO cn o «^ cn rH 00 cn rH CO cn ^* LO LO LO ^^ I— 1 rH O CM CM ^^ . CM cn CO CO 0 ^f CO . **Q i-O UD CM CO cn rH . in 4-) !3 r— P— 0 a. r- 03 S_ 0 M- T3 Ol £ 3 in -P in o U •a Ol . i. it ^ o u Ol 1 in -P in O U O5 C •r" •P 03 S- ai Q. O -|. Ol S- 3 4J • i— -rj C Ol a. X 0> r— 03 4_> • r— CL C_J II +J in 0 U 'ra 4J O H- O3 O 4_> m i- 03 •— O c Ol S- i. 3 ^_) •P i. Ol ^ c o u o -P •a Ol in 3 S- -P U Ol in • in 3 J3 E 03 C O c Ol .c 4_> 0 4- S- o 03 ,_ 4- O> •a o> u 'r- « S- in a. cu 3 •P i— •r- 03 U > r^ CM S cn HH t— 1 Xi Ol in 3 O) in O CL •a c m c o u O (J 03 0) S -P Ol in in TO 3 O •r- TJ 3 a. s_ o in -P • r— c 3 -P C a> E c s_ O) o +J c CD tT3 a. in 01 U c u o in S- o •P C ai e c: s- 01 > o O! o -P c O) 03 a. in O) •a 3 O C 9-68


-------
     9.2.3.1  Price Effects.  The price of furnace coke is assumed to be
established in a competitive market.  In the basic model of a competitive
market, the interaction of supply and demand determine the equilibrium
price.   This price is dependent on the costs of production of the marginal
producer and the value of the product to the marginal buyer.   The marginal
producer is the producer who is willing to supply the commodity at the
market price because he is just covering all his costs at that price.  The
marginal buyer is just willing to pay the market price.  Other buyers who
value the product more still pay only the market price.
     Estimates of the demand and supply functions for furnace coke are
necessary to develop projections of the equilibrium price for furnace coke
with and without increased control.  The supply of furnace coke as shown
previously would be shifted by the regulatory alternatives.  The demand for
furnace coke has been econometrically estimated and found to be responsive
to price changes.  The estimated elasticity of demand for furnace coke is
-1.3.  -This responsiveness reflects the substitution of other fuels for
coke in blast furnaces; the substitution of other inputs, primarily scrap,
for pig iron in steelmaking; and the substitution of other commodities for
steel throughout the economy.
     Higher prices for coke will increase the cost of steel production
unless there is a perfect substitution between coke and other inputs to
steelmaking.  In that case, the consumption of coke would decrease to zero.
If substitutions for coke in steelmaking were not possible (i.e., input
proportions were fixed), the steel price increase would be the percentage
change in coke price times the share that coke represents in the cost of
steelmaking (10.7 percent) times the base price of steel.
     Table 9-21 presents the furnace coke and steel price impacts of the
regulatory alternatives.  The proposed regulatory alternatives raise coke
prices only slightly; about 0.2 percent regardless of the cost scenario.
     9.2.3.2  Production and Consumption Effects.   The estimated demand and
supply relationships for coke are used to project the production and con-
sumption effects of the regulatory alternatives.   As shown in Table 9-22,
the changes in coke production and consumption are fairly small for the
                                  9-69

-------
 TABLE 9-21.  PRICE EFFECTS OF REGULATORY ALTERNATIVE II—
                    FURNACE COKE, 1984

  Compliance                       Coke            Steel
cost scenario                     ($/Mg)           ($/Mg)

Midrange estimates                 0.22             0.07
                                  (0.2)            (0.0)

Upper bound estimates              0.25             0.08
	(0.2)	(O._0)	

aValues in parentheses are percentage changes from baseline.
                             9-70

-------









1
1
1— 1
1— (

LU
h— 1
 r-l
LU
LL. •*
U_ LU
LU ^
O
Z 0
o
»— t LU
1— 0
a. <
2: z
— i Q£
CO ZD
Z U_
0
O

O
z
«^

z
o
1—4
H-
U
a
0

a.

•
OJ
CNJ
cn
LU
	 1
ca

to

















^-,
s-
>>
>SNN
CJ
s:

CO
o
rH
^^

+J
ai
v^
5-
03
s

ID

O
O








co
i.
o
Q.
HH

C
o
•£
Q.
E

CO
C
o
o

c
o
•r™
4J
u
3
•a
o
i.
O.

CO
^>
i-
O
O.
E
1—1



C
O
• r-
+J
Q.
E
3
CO
C
O
O



c
o

+J
u
3
•a
o
i^.
a.

















O
•r-
0> i-
O 03
C C
O3 0)
•r- U
r— v/j
a.
E •*->
O o>
cj o
u

o
«
ID O
•— ^




O
•
i-H O
co i
i ^"-^





^*\
o
.
^ o
CO 1
I ^"-^




f*"S
^>
•
tn o
r-t ^>






x"s
C\J
*
CT> O
CO 1
1 s*-^*









/*"%
CO
*
^ o

1 "*^-^














c/>
0)
•+-*
(T5
•r-
+J
t^
0)

01
O)
c
03
5-
T3
• r~
s:

o
*

4^
03
gs
• r"
•t-1
01

•a
c
3
O
-Q

1-
01
a.
a.
^2





































aJ
c
»^-
^—
01
CO
(0
O

E
O
C—
<4_

CO
01
D)
C
03
.C
U

01
O)
•2
c
01
u
0>
a.

ai
03
co
0>
CO
01

-p
c
0>
03
a.

c


CO
O)
3

03

03
9-71

-------
regulatory alternative.   Changes in production and consumption are less
than 0.3 percent.
     Imported coke is a close substitute for domestically produced furnace
coke.   Imported coke is not a perfect substitute because coke quality
deteriorates during transit and contractual  arrangements between buyers and
sellers are not costless.   However, increases in the costs of production
for domestic plants will increase the incentive to import coke.
     The projected increases in coke imports are reported in Table 9-22.
Imports increase by a maximum of 0.5 percent under the alternative.   As
illustrated below, coke imports increased significantly since 1972,  but
peaked in 1978 and began a marked decline.
Year
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
Imports (103


3
1
1
1
5
3




168
978
,211
,650
,189
,659
,191
,605
598
478
109
32
Mg)












     The increase in imports in the 1970s is believed to be the result of a
coal strike in the United States during 1978 combined with depressed condi-
tions in the market for steel in the countries exporting coke to the United
States.   Accordingly, future importation at a high level may depend upon
future market conditions for steel in other countries.  In any case, the
change in coke imports projected for all the regulatory alternatives is
smal1.
     9.2.3.3  Coal Consumption and Employment Effects.  Any reductions in
coke and steel production are expected to cause reductions in the use of
the factors that produce them.  The major inputs to coke production are
coal and labor.  Labor is also an important input in coal mining.
     The coal consumption and employment implications of the projected
reductions in coal, coke, and steel production are shown in Table 9-23.
                                   9-72

-------
          TABLE 9-23.   COAL CONSUMPTION AND EMPLOYMENT EFFECTS OF
              REGULATORY ALTERNATIVE II—FURNACE COKE, 1984a
Compliance
cost scenario
Midrange estimates
Upper bound estimates
Coal
consumption
for coke
(103 Mg/yr)
-79
(-0.3)
-92
(-0.3)
Employment (jobs)
Coalc
mining
-22
(-0.3)
-26
(-0.3)
Coke
plant
-17
(-0.3)
-19
(-0.3)
Steel-
making
-139
(-0.0)
-161
(-0.1)
 Values in parentheses are percentage changes from baseline.

 Employment impacts are based on input-output relationships  and production
 impacts.   Impacts on coke plant employment do not include jobs created by
 the relevant controls.

cAnnual labor productivity in coal  mining is estimated as  3,515 megagrams
 per year per job.
                                  9-73

-------
For Regulatory Alternative II, coal  consumption and employment impacts are
less than 0.3 percent.   These values assume constant coal- and labor-output
ratios.  The employment impacts shown do not include the estimated increases
in employment caused by the regulatory alternatives.  Therefore,  the employ-
ment impacts represent-maximum values.
     9.2.3.4  Financial Effects.   The aggregate capital costs of the regula-
tory alternatives are summarized in Table 9-24.  Capital costs have also
been summed across member plants to determine the cost to each coke-producing
company of meeting alternative regulations.  The total capital costs by
company may be used to produce percentages that express the relation between
total capital cost and the annual average net capital investment of the
company and the annual  cash flow of the company.  This analysis is pre-
sented to give some insight into the distribution of the financial effects
across coke-producing firms.
     Total capital cost as a percentage of average annual net investment is
an indicator of whether the usual sources of investment capital available
to the firm will be sufficient to finance the additional capital  costs
caused by the regulatory alternatives.  The larger this percentage, the
greater the probability that investment needed to comply with the regulatory
alternatives would significantly reduce investment in other areas.  This
percentage provides some insights regarding the degree to which firms will
be able to finance the controls required to meet the regulatory alternatives
without a serious impact on their financial position.
     Total capital cost as a percentage of cash flow provides similar
information.  Cash flow data accounts for revenues, operating costs, depre-
ciation, expenditures on dividends, interest expenses, and taxes.  Thus, it
is a more realistic measure of the funds available to the firm.  However,
this index may not be consistent across firms, because depreciation account-
ing varies across firms.  As with the net investment ratio, the larger the
ratio, the greater the probability that cash flow will be diverted from
other sources to finance compliance expenditures.
     Financial analysis is necessarily restricted to companies for which
financial data are accessible.  Therefore, financial analysis cannot be
conducted for some privately owned companies for which reporting has been
                                  9-74

-------
         TABLE 9-24.  INDUSTRY CAPITAL REQUIREMENTS OF REGULATORY
                    ALTERNATIVE I I — FURNACE COKE, 1984

  Compliance                                    Capital costs
cost scenario                                    (106 1984 $)


Midrange estimates    .                                  7

Upper bound estimates                                  10


Calculated for all plants in operation in 1984.44 46
                                  9-75

-------
restricted.   These companies are usually the smallest in a given industry,
and they probably experience higher per unit costs of regulation and higher
costs for securing financing than do larger companies.
     A further complication of financial analysis is that many coke-produc-
ing companies are wholly owned subsidiaries of larger,  highly diversified
corporations.  Financial data are available for the parent corporations
only.  Analysis of these data will probably lead to the conclusion that the
parent companies have ample resources to finance additional capital costs.
However, the extent to which these corporations will make such investments,
or will cease some coke operations in favor of other investment opportuni-
ties evidencing higher rates of return, cannot be determined without knowl-
edge of the required return on investment specific to the firm and the
other investment opportunities that exist for the firm.
     Table 9-25 provides the capital costs as a percentage of average
annual net investment by company for Regulatory Alternative I.I.  The average
annual net investment was calculated from financial records for each company
by averaging net investment data (in constant 1983 dollars) for 1979 to
1983.  These values are converted to 1984 dollars using an implicit GNP
price deflator.  In some instances less than 5 years of data were available.
In the case of subsidiaries, net investment of parent companies was used.
The  regulatory alternative imposes capital costs as percentages of average
annual net investments of less than 1 percent.
     Table 9-26 shows capital cost as a percentage of cash flow for firms
for which information is available.  Cash flow data were derived from
Table 9-14.   The values for 1983 were converted to 1984 dollars using an
implicit GNP price deflator.  As for net investment, in the case of subsi-
diaries, cash flow of parent companies  is used.  Capital costs as percent-
ages of cash flows range from 0 to 1.4  percent for the regulatory alterna-
tive.
     The leverage ratios presented in Table 9-15 indicate that coke-producing
firms are engaged in a substantial amount of external financing.  These
firms may be reticent (or unable) to borrow more heavily, especially in the
current economic climate for steel.  Furthermore, financing capital expendi-
tures by issuing additional common stock would tend to dilute existing
stockholder equity.  Considering the low historical return on  investment  in
                                  9-76

-------











I—

LU
2:
i—
oo
LU
^> i
-ytrt
1— ' •=*•
00
t— 01
LU r-t
Z.
U_ 00
O C£
LU
LU 0

< Q
1— O
^Z, ****
LU Q-

Q£ LU
LU i^
0. O
O
^C
LU
00 0

Z.
LU C£
O Z3
z u_
< i
i— i i
_! KH
Q- I-H
s:
O LU
0 >
t-H
U- 1—
0 <

00 C£.
t— LU
00 1—
O -J
0 <

—1 >
<^ ^v
1— 0

SI 

C3 0



iH r- 1
, ,
O O


in
0)
4_>
re
E

^>
in "O in
Ol CO)
3 -P
oi o re
Ol J2 E
C •!-
re s» ^^
i. oi in
T3 Q. Ol
• i- Q.
2: r>

in
N^^
TJ
O
O
s:

«t
r—
re
3
c
re
*jT

r~"
tJ
•n-
J_
+J
V)
Z3
T3
c

in
«
^^
^3
o
o
^

c
•r-

in
01
r—
• r~
14-
o
i-
Q.

t-
0 0
0 -P
x^
U_
•a
•a oi
C +J
re i-
Ol
T3 >
t. C
re o
•a u
c
re -a
•P C
oo re

" in
in
-P
4-
O
a
Ol
en
0)
en
c
re
j=
u
X
LU

^
U
o
-P
oo

^
i-
o
^™

3
>, 0)
C
re
^^

Q.-O
E
O
U

E
0
i-
4-

T3
01
•P
re
^~
3
U
r—
re
u

Ol
E
0
U
c
• r—

^Jt
O)
c

^~
re
3
C
C
re

Ol
01
re
s-
Ol
>
<^
re
J-
re
T3
c
re
-P
oo
• r—
in
re
JD

J-
re
^!
o
•a

00
00
Ol
i-H
.
-P in
c oi
re T-
•P C
in re
C Q.
0 E
(J O
o
re
o
C 3
O -P
.
Ol S- i-
•a o o
re -P 4-
E re
•— Ol
01 4- r-
i- oi ja
oi TJ re
3 i—
01 •!-
"^3 u\ CJ rO
C C T- >
re o s- re
•r- 0.
"-P -P
«* re -p o
00 •— -i- C
Ol 3 U
rH U •<- Ol
i — i — S-
- re a. re
-*: «_> e
t. 	 /.r- 4->
o c
>- Q- O)
• Z S
3 «* C3 -P
O) 00 m
z 01 re oi
iH >
01 c
Ol » C •!-
U ^ T-
•r- i. in r—
> o 3 re
5- >- 3
oi in c
oo 3 ai c
oi 3 re
S- Z r-
0 03 C
-P - > O
in
oi f"* ^* re
> S- 00 -P
c o 01 re
I— i O i— 1 Q
jQ
9-77

-------
       TABLE 9-26.   CAPITAL COST AS A PERCENTAGE OF ANNUAL CASH FLOW
           REGULATORY ALTERNATIVE—FURNACE COKE PRODUCERS, 19843

  Compliance    	Furnace Coke Producers6	
cost scenario   ABCCDEF     GCHIJ     K    I


Midrange       0.1   -   0.1  0.4  1.0  0.9    -   0.0  0.9  0.2   0.1  0
  estimates
Upper-bound     0.1   -   0.2  0.6  1.4  1.3    -   0.0  1.3  0.2   0.2  0
  estimates

aCash flow data is from Table C-14 in Appendix C.   Values are converted to
 1984 dollars using a GNF implicit price deflator.

 There are three furnace coke producers for which cash flow data was not
 available.
cThis company had negative cash flow.
                                  9-78

-------
the industry, this dilution would probably be-unacceptable.  An analysis of
the iron and steel industry undertaken by Temple, Barker, and Sloane, Inc. ,63
addressed the question of external financing with regard to water pollution
control expenditures.  This analysis concludes that to avoid deterioration
in its financial condition, the industry is likely to reduce expenditures
to modernize production facilities rather than increase its external financ-
ing.   In fact, as discussed in Section 9.1.5, capital expenditures for
pollution abatement as a percentage of new capital expenditure are decreas-
ing.
     In summary, the capital costs of the regulatory alternatives are in
the seven-to-ten-mi 11 ion dollar range.  However, these amounts do not
appear unduly burdensome when compared with normal investment expenditures
or cash flow for companies for which data are available.
     9.2.3.5  Battery and Plant Closures.  Uneconomic batteries are those
that have marginal costs of operation greater than the price of coke.
Theoretically, these batteries are candidates for closure.  There are 14
batteries operating in the uneconomic portion of the supply curve (above
the point where priceline intersects the supply curve) under baseline
conditions.   They are owned by 10 companies and are located in 11 plants.
Regulatory Alternative II adds one battery to this list.  It is important
to note, however, that this does not necessarily imply that the regulation
would cause the closure of this battery.
     The decision to close a battery is more complicated than the basic
closure rule would indicate; this is particularly true for integrated iron
and steel producers.   Continued access to profits from continued steel
production is a key factor in the closure decision for a captive battery.
Before closing or idling a coke battery,  managers would want to know where
they would get coke on a reliable basis in order to continue making steel.
The obvious sources to be investigated include other plants within the same
company, other companies, and foreign suppliers.   As noted in Section 9.1.4.1,
some interregional and international movement of coke occurs.
                                   9-79

-------
     Obtaining coke from offsite sources introduces two potential  complica-
tions:   the cost of transporting coke and the certainty of the coke supply.
Obtaining coke from a nearby source might be the most profitable alternative
to transporting coke.   If coke must be shipped over long distances, onsite
production at a cost above the projected market price might be more profit-
able.
     If coke must be purchased, certainty of supply is a complication.
Steel  producers prefer to have captive sources of coke to safeguard against
supply interruptions, and they may be willing to pay a premium for this
security.  Producing at a cost above market price would involve such a
premium.   Five of the fourteen uneconomic batteries under baseline compli-
ance produce coke at marginal costs that are less than 5 percent above the
market price.  Five percent does not appear to be an excessive premium to
pay for certainty of supply.
     Several other factors could affect a particular plant's decision to
close a battery.  These factors relate to the relationship of coke quality
to the type of steel commodities produced, the existence of captive coal
mines, the costs of closing a battery and potential costs of restarting it
in the future, and required control and other expenditures.
     An alternative to closure-for a financially troubled company is to
file for Chapter 11 bankruptcy.  This option allows firms to continue
operating coke plants under a restructured debt payment schedule.   Of firms
owning the fourteen uneconomic batteries under baseline, one has filed for
Chapter 11, and another is expected to file in the future if the steel
industry continues to sustain large losses.88  This action may improve a
firm's competitive situation.  McLouth Steel Corporation, which filed for
bankruptcy in 1982, has modernized equipment and reduced overhead to enable
it to capture market share from larger companies.
     The demand model uses a single coke price, which represents an average
quality of coke used to produce a weighted average mix of steel products.
If high production costs for a particular battery are associated with a
higher than average quality of coke, continued production might be justified.
Production would also be justified if the firm produces only the most
highly valued steel products.
                                  9-80

-------
     Some coke-producing firms also own coal mines and may wish to secure
continued access to profits from coal mining.   Because profits in the coal
sector may be subject to less effective taxation because of depletion
allowances, these profits may be extremely attractive.
     Furthermore, an integrated iron and steel producer must consider the
question of necessary expenditures for its entire steel plant.  If the
steel facility is old or if substantial additional expenditures will be
necessary to comply with regulations on other parts of the facility, then
closure is more likely.
     Closure decisions are so specific to individual situations and managers'
perceptions regarding their future costs and revenues that exact projections
of closures should be viewed with caution.  In the current market for
steel, it is difficult to say whether or not uneconomic batteries will be
closed.  Of companies owning uneconomic batteries, three have cut back
capacity by closing batteries, though it is unknown whether they are those
projected as candidates for closure.88
9.2.4  Foundry Coke Impacts
     Oven coke other than furnace coke represents less than 15 percent of
U.S.  coke production.   The majority of it is used as a fuel in the cupolas
of foundries.  The remainder is used for a variety of purposes, especially
for heating.
     Values of various foundry coke variables in the absence of the regu-
latory alternative are presented in Table 9-27.   These values assume base-
line compliance with the regulations listed in the footnote to the table.
     9.2.4.1  Price and Production Effects.  In developing the estimates of
price and quantity impacts, a vertical shift caused by each regulatory
alternative has been projected in the linear estimate of the foundry coke
supply function generated under the regulatory baseline.  This shift is
used in conjunction with an estimated elasticity demand for foundry coke of
-1.03, and is designated Import Scenario A.  Under this scenario, domesti-
cally produced foundry coke does not compete with imported coke.   This
analysis also provides estimates of the effects of alternative regulations
assuming that foundry coke producers must compete with imports in open
market sales.  In this case,  consumers of foundry coke may purchase imported
                                  9-81

-------
       TABLE 9-27.   1983 BASELINE VALUES FOR ECONOMIC
            IMPACT ANALYSIS--FOUNDRY COKE,  198395

                                            Baseline values3

Coke market

  Price (1983 $/Mg)                               169.58b
  Production (103 Mg)            •               2,951
  Consumption (103 Mg)                          2,938
  Employment (jobs)                               542
  Coal Consumption (103 Mg)                     3,809

aBase1ine assumes companies meet existing regulations
 including OSHA (coke oven emissions); State regulations on
 desulfurization, pushing, coal handling, coke handling,
 quench tower, and battery stack controls;  and BPT and BAT
 water regulations.

 Calculated.  Market price for foundry coke was $149.66 in
 the fourth quarter of 1983.

cCa1culated.  Foundry coke employment = Employment in
 byproduct coke industry x proportion of coke production
 represented by foundry coke sector.
                             9-82

-------
coke as a perfect substitute for domestically produced coke.  The price of
imports is assumed to be constant.   As regulations cause less domestic coke
to be produced, its price relative to imported coke rises.   Consumers are
able to purchase as much coke as before at the same price,  but a larger
proportion of sales is.made up of imports.  Thus, there are quantity effects
for domestic producers,  but no price effects.  This shift is designated
Import Scenario B.   Impacts are assessed for both scenarios.  Impacts for
Import Scenario B represent the maximum effect of import substitution in
the foundry coke market.
     The projected price and-quantity effects are shown on Table 9-28.
Under Import Scenario A, both price and quantity impacts are less than
1 percent of baseline for Alternative II.   Under Import Scenario B, there
are no price impacts.  Quantity impacts are a maximum of 2.8 percent of
baseline.
     9.2.4.2  Coal  Consumption and Employment Effects.  Any reductions in
foundry coke production are expected to cause reductions in the use of the
factors that produce the foundry coke.  The major inputs to foundry coke
production are coal and labor.  Labor is also an important input in coal
mining.
     The coal consumption and employment implications of the projected
reductions in coke production are shown in Table 9-29 for both scenarios.
These values were developed assuming constant coal- and labor-output ratios.
The employment impacts shown do not include any employment increases caused
by the regulatory alternative.  Consequently, the employment impacts repre-
sent maximum values.
     Under both scenarios, effects on employment in the coal mining industry
are negligible.  Under Import Scenario A,  coal consumption impacts or
employment impacts in coke plants less than a 1 percent change from baseline.
Under Import Scenario B, coal consumption and employment in coke plants are
reduced by less than 3 percent.
     9.2.4.3  Financial  Effects.   The aggregate capital costs of the regula-
tory alternatives are summarized in Table 9-30.   The capital requirements
to meet Regulatory Alternative II for the foundry coke industry are $2 to
$3 million.   Capital  costs have also been summed across member plants to
determine the cost to each company of meeting alternative regulations.

                                   9-83

-------
  TABLE 9-28.   PRICE AND QUANTITY EFFECTS OF REGULATORY ALTERNATIVE II—
                            FOUNDRY COKE,  1984
                                              a
  Compliance                  Coke price impact        Coke quantity impact
cost scenario                   (1984 $/Mg)                (103 Mg/yr)
Midrange estimates
Import Scenario A
Import Scenario B
Upper bound estimates
Import Scenario A
Import Scenario B

1.13
(0.7)
0.00
(0.0)

1.34
(0.8)
0.00
(0.0)

-20
(-0.77)
-69
(-2.3)

-24
(-0.8)
-82
(-2.8)
aVa1ues in parentheses are percentage changes from baseline.

 Coke consumption is not affected due to imports.   Imports under Scenario B
 are equal in magnitude and of opposite sign to quantity impacts.   Under
 Scenario A, imports are zero.
                                  9-84

-------
   TABLE 9-29.   COAL CONSUMPTION AND EMPLOYMENT EFFECTS OF REGULATORY
                   ALTERNATIVE II—FOUNDRY COKE, 1984a
Compliance
cost scenario
Midrange estimates
Import Scenario A
Import Scenario B
Upper bound estimates
Import Scenario A
Import Scenario B
Coal
consumption
for coke
(103 Mg/yr)
.
-26
(-0.7)
-89
(-2.3)

-31
(-0.8)
-106
(-2.8)
Employment
Coal
mining

-7
(-0.7)
-25
(-2.3)

-9
(-0.8)
-30
(-2.8) •
(jobs)b
Coke
plant

-3
(-0.7)
-13
(-2.3)

-4
(-0.8)
-15
(-2.8)
Values in parentheses are percentage changes from baseline.

Employment impacts are based on input-output relationships and production
impacts.   Impacts on coke plant employment do not include jobs created
by the relevant controls.

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

Zero due to rounding.
                                 9-85

-------
         TABLE 9-30.   INDUSTRY CAPITAL REQUIREMENTS OF REGULATORY
                      ALTERNATIVE II—FOUNDRY COKE, 1984
                                                   Capital  costs
  Compliance                                      of regulations
cost scenario                                       (10* 1984 $)


Midrange estimates                                       2

Upper bound estimates                                    3


Calculated for all plants in operation in 1984.48 50
a
                                  9-86

-------
These company capital costs, along with firm-specific financial data, are
used to produce the same financial percentages as described above for
furnace coke, total capital cost as a percentage of net capital investment
and of annual cash flow.  Financial data are not available for many of the
foundry coke producers-that are privately held companies.   Therefore,
percentages for these companies are not included in the analysis.
     Capital costs as percentages of average annual net investment for the
foundry coke producers are provided in Table 9-31.   The costs of moving
from baseline to Regulatory Alternative II are never more than 3.3 percent
of the average annual net investment.  Foundry coke production plants
operate at a significantly lower production rate for the same level of
investment as compared with furnace coke production rates.  This is due to
the longer coking time for foundry coke.  Furthermore, in looking at the
available data on the age of the batteries used in the production processes
within each plant, there appears to be a correlation between the age of the
battery used and the level of compliance costs facing the firm.  The data
suggest that the foundry coke producing plants which are facing the highest
pending compliance costs are operating with batteries which were installed
between 1919 and 1946.  Conversely, the foundry coke producers which are
facing the lowest pending compliance costs are operating, for the most
part, with batteries installed between 1950 and 1979.
     Table 9-32 provides capital cost as a percentage of annual cash flow.
The Regulatory Alternative results in capital costs no greater than 0.6
percent of cash flow for foundry coke producers for which information is
available.
     Firms would use internal financing, additional equity financing, debt
financing, or perhaps some of the methods mentioned in Section 9.1.5, to
make these capital expenditures.  Since many of the foundry plants are
owned by private corporations, data are insufficient to assess the eventual
sources of capital that these firms will use.  Therefore, only qualitative
statements can be made concerning the impacts of financing regulatory
investments.  Any internal financing would reduce return on equity by
directly reducing dividends or by reducing productive capital expenditures.
Debt financing may reduce the return on equity by increasing the cost of
                                  9-87

-------
      TABLE 9-31.   CAPITAL COSTS OF COMPLIANCE AS A PERCENTAGE OF NET
    INVESTMENT REGULATORY ALTERNATIVE II—FOUNDRY COKE PRODUCERS, 1984a

  r   ,.                              Foundry coke producers
  Compliance                  	•*	c	
cost scenario                 AA                 BB                CC
Midrange estimates
Upper bound estimates
0.1
0.2
0.1
0.1
2.3
3.3
aAverage annual net investment calculated from company profiles in Moody's
 Industrial Manual, Moody's Investor Service, New York, 1984, Standard
 New York Stock Exchange Reports, Standard and Poor Corp.,  New York, 1984,
 and Dun and Bradstreet Financial Profiles, 1985.  (Calculations were made
 on a constant 1983 dollar basis and converted to 1984 dollars using the
 GNP implicit price deflator.)

 There are eight foundry coke producers for which annual  investment data
 are not available.
                                  9-88

-------
   TABLE 9-32.   CAPITAL COST AS A PERCENTAGE OF ANNUAL CASH
         FLOW REGULATORY ALTERNATIVE II—FOUNDRY COKE
                       PRODUCERS, 1984a

  Compliance     .              	Foundry coke producers*3	
cost scenario                  AA       BB       CC       DD


Midrange estimates             0.1      0.1      0.4      0.2

Upper bound estimates          0.1      0.1      0.6      0.3


aCash flow data is from Table C-14 in Appendix C.   Values
 are converted to 1984 dollars using the GNP implicit price
 deflator.

 There are six foundry coke producers for which cash flow
 data was not available.
                             9-89

-------
debt financing.   Financing regulatory capital  requirements using new common .
stock issues will have a tendency to dilute present owner's equity.   This
dilution could be substantial in cases where regulatory capital requirements
are high.   As an alternative, one foundry coke firm entered bankruptcy
status.   No information on the competitive status of this firm is available.
For the firms for which data are available, the capital requirements of the
regulatory alternative do not appear unduly burdensome.
     9.2.4.4  Battery and Plant Closures.   The decision rule used to indi-
cate closure candidates among furnace batteries is also used for foundry
batteries.   Any foundry battery for which the marginal cost of operation is
greater than the price of foundry coke is an uneconomic battery.  According
to this criterion and assuming baseline control, no batteries that were in
operation in 1984 are uneconomic to operate.  The regulatory alternative
does not cause any batteries to move into the uneconomic region under
either scenario.
9.3. POTENTIAL SOCIOECONOMIC AND INFLATIONARY IMPACTS
9.3.1  Compliance Costs
     The estimated total annualized costs to coke producers for compliance
with Regulatory Alternative II are shown in Tables 9-33 and 9-34.  Costs
for furnace and foundry producers are differentiated because of differences
in coke prices and control costs per unit of output.  The costs are for all
plants in operation in 1984 are calculated.  As shown in Table 9-33, in
1984, Regulatory Alternative II would result in compliance costs of $11 to
$14 million per year for furnace coke producers.  Compliance costs for
foundry coke producers is shown in Table 9-34.  For Regulatory Alternative  II,
this cost is $3 to $4 million per year.  Combined compliance cost for
furnace and foundry coke producers is $14 to $17 million per year.
9.3.2  Balance of Trade
     Recent trends indicate that imports are decreasing.  Imposition of the
regulatory alternatives is expected to slow this trend.  Some increase in
steel imports is possible also.  However, since steel price increases
caused by coke price increases are projected to be quite small, any increase
in imports caused by the regulatory alternatives should be minor.  Moreover,
trade regulations covering steel imports may mitigate such increases.
                                   9-90

-------
TABLE 9-33.  COMPLIANCE COSTS OF REGULATORY ALTERNATIVE II-
                FURNACE COKE PRODUCERS, 1984

       Compliance                        Compliance
     cost scenario                  cost (106 1984 $/yr)


Midrange estimates                          11.2

Upper bound estimates                       13.9
                             9-91

-------
TABLE 9-34.   COMPLIANCE COSTS OF REGULATORY ALTERNATIVE II-
                FOUNDRY COKE PRODUCERS,  1984

     Compliance                          Compliance
   cost scenario                    cost (106 1984 $/yr)


Midrange estimates                           3.0

Upper bound estimates                        3.8
                             9-92

-------
     In the aggregate It appears unlikely that the regulatory alternative
would significantly affect the U.S. balance of trade position, given the
small share of international trade represented by coke imports.
9.3.3  Community Impacts
     Furnace and foundry coke and steel production facilities are in Pennsyl-
vania, Indiana, Ohio, Maryland, New York, Michigan, Illinois, Alabama,
Utah, Kentucky, Tennessee, Texas, Missouri, Wisconsin, and West Virginia.
Closure of coke facilities, if they occur, could have impacts on commun-
ities in these States.  The regulatory alternatives are not necessarily
projected to result in closures.  Potential production decreases should not
be sufficient to generate significant community impacts.   However, compli-
ance with other proposed regulations could result in additional  battery and
plant closures and the resulting community impacts.
9.3.4  Small Business Impacts
     The Regulatory Flexibility Act (RFA) requires consideration of the
potential impacts of proposed regulations on small "entities."  A regulatory
flexibility analysis is required for regulations which have a "significant
economic impact on a substantial number of small entities."  For the NESHAP
for coke oven by-product plants, small entities can be defined as small
furnace and foundry coke firms.  The Environmental Protection Agency Office
of Standards and Regulations recently drafted a memo on RFA compliance.89
This section addresses the requirements that relate to the economic aspects
of the RFA.  Steps necessary for determination of applicability of the RFA
are:
          Identification of small firms impacted by the NESHAP,  and
          Estimation of the economic impact of the NESHAP on these
          small firms.
     The guidelines for conducting a regulatory flexibility analysis define
a small business as "any business concern which is independently owned and
operated and not dominant in its field as defined by the Small Business
Administration Regulations under Section 3 of the Small Business Act."  The
Small Business Administration (SBA) defines small firms in terms of employ-
ment.  Firms owning coke ovens are included in SIC 3312,  which also includes
blast furnaces, steel works, and rolling mills.  The SBA has determined
                                   9-93

-------
that any firm that is in SIC 3312 and employs less than 1,000 workers will
be considered small under the Small  Business Act.
     Table 9-35 shows employment data for all U.S. firms that operate
by-product coke ovens.   Six firms in the list--9,  14, 16, 19, 20, and
23—can be designated as small  based on SBA definitions.  Because the
standard being proposed is a NESHAP and all existing and new plants will be
expected by law to comply, all  plants of the small firms not currently in
compliance could be adversely impacted.  A "substantial number" of small
business is defined as  "more than 20 percent of these entities."89  This
rule implies that at least two firms be impacted to qualify as a "substan-
tial number."
     After the affected small firms are identified, the guidelines for the
RFA require an estimate of the degree of economic impact.  Four criteria
are applied in assessing whether significant economic impact occurs from
the regulation.  The first criterion determines whether annual compliance
costs increase average total production costs of small entities by more
than 5 percent.  None of the small firms identified was found to have an
average cost increase greater than 5 percent.
     A second criterion compares compliance costs as a percentage of sales
for small entities with- the same percentage for large entities.  If the
result for small entities is at least 10 percentage points higher than for
large firms, the impact is considered significant.  Based on net sales data
in Table 9-14 and compliance cost data in Tables 9-31 and 9-32, no small
company is significantly impacted.  It should be noted that sales data are
not available for all small entities.
     The third criterion for assessing significant impact is whether capital
costs of compliance represent a "significant" portion of capital available
to small entities.  The criterion recommends examining  internal cash flow
in addition to external sources of financing.  Small, privately owned firms
often do not report their annual investment expenditures, so that determina-
tion of capital availability is impossible.  No financial data could be
located for the small coke-producing firms previously identified.  However,
capital costs as percentages of net investment and cash flow for other
firms in the industry are not substantial (see Table 9-25, 9-26, 9-31, and
9-32).   Based on this information, the capital costs of compliance are not

                                   9-94

-------
     TABLE 9-35.   EMPLOYMENT DATA FOR
   U.S.  FIRMS OPERATING COKE OVENS, 1983

Company                        Employment
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
48,071
52,800
28,700
9,107
163,356
37,300
16,000
32,000
210
1,300
2,400
19,200
7,300
102
7,512
164
1,245
98,722
150
150
14,518
2,200
200
                   9-95

-------
expected to represent a significant portion of capital  available to small
firms.
     The final criterion is whether the requirements of the regulation are
likely to result in closures of small  entities.   None of the small firms
identified is projected to close as a result of the regulatory alternatives.
     The regulatory alternatives are unlikely to result in adverse economic
impact on a "substantial number" of small  entities (as defined by SBA).
Based on the four criteria used by the U.S. Environmental Protection Agency
for which assessment may be made, no firm is expected to be "significantly
impacted" under the second criterion.   No significantly impacted firms were
identified under the other rules.
9.3.5  Energy
     The regulatory alternatives do not have any significant direct energy
impacts.  Although some indirect impacts are possible, they are likely to
be minor in nature.
     Indirect impacts could include the substitution of fossil fuels for
coke in blast furnaces or an increase in use of electric furnaces, further
reducing the coke rate.  Some reduction is projected to occur in any case,
but technological limits govern the degree to which the coke rate can be
reduced.  Furthermore, projected coke price increases are minor when compared
to recent and projected fossil fuel price increases.  Of course, if imports
increase, fuel will be needed to transport them.  Furthermore, if imports
replace domestic coke production, excess coke oven gas, some of which is
currently used in other parts of the steel plant, may be replaced by other
fuels.   But if steel production decreases, there will be some reduction  in
fuel consumption.
9.4  REFERENCES
 1.  Standard Industrial Classification Manual, 1972.  Executive Office  of
     the President, Office of Management and Budget.  Washington, DC.  U.S.
     Government Printing Office.  1972.  p. 145.
 2.  1982 Census of Manufactures.  Preliminary Report, Industry Series.
     Blast Furnaces and Steel Mills (Industry 3312).  Bureau of the Census,
     U.S. Department of Commerce.  Washington, DC.  MC82-I-33A-1(P).
     July 1984.   p.. 3.
                                 9-96

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

 4.   Energy Data Reports:  Coke and Coal Chemicals  in 1979.  Office of Coal
     and Electric Power Statistics, Energy Information Administration, U.S.
     Department of Energy.  Washington, DC.  DOE/EIA 0121(79).  October 31,
     1980.   p. 33-37.

 5.   Research Triangle Institute.  Coke Battery Survey and information from
     the American Coke and Coal Chemicals Institute and the American  Iron
     and Steel Institute.  Unpublished data.   1984.

 6.   Minerals Yearbook.  Bureau of Mines, U.S. Department of the Interior.
     Washington, DC.  1950-1978.  (Table and page numbers vary because of
     reclassifications.)

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

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

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

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

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

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

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

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

15.   Reference 8.  Table 1471.  p. 832.
                                  9-97

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

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

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

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

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

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

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

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

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

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

26.   Reference 4.   p. 3-5.

27.   Reference 18.   p.  3-5.

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

29.   Reference 28.   p.  40.
                                   9-98

-------
30.  Quarterly Coal Report, January - March 1984.  Appendix A.  Office  of
     Coal, Nuclear, Electric, and Alternate Fuels, Energy  Information
     Administration, U.S. Department of Energy.  Washington,  DC.   DOE/EIA-
     012K84/1Q).  July 1984.  p. 62.

31.  Telecon.  Peterson, J. , U.S. Steel Corporation, with  Lohr, L. ,  Research
     Triangle Institute.  December 10, 1984.  Operational  status  of  cold
     idle coke batteries.

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

33.  Reference 4.  p. 1.

34.  Reference 4.  p. 13.

35.  Reference 25.  p. 54.

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

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

38.  The Outlook for Metallurgical Coal and Coke.  Institutional  Report.
     Merrill Lynch, Pierce, Fenner, and Smith, Inc.  1980.  p.  3.

39.  Reference 28.  p. i.

40.  Reference 28.  p. ii-iii.

41.  Reference 38.  p. 1.

42.  Reference 38.  p. 5.

43.  PEDCo Environmental, Inc.  Technical Approach for a Coke Production
     Cost Model.   1979.  p. 39-50.

44.  Reference 18.  p. 33-35.

45.  Energy Data Reports:   Distribution of Oven  and Beehive Coke  and Breeze.
     Office of Energy Data  and Interpretation, Energy Information Adminis-
     tration, U.S. Department of Energy.  Washington, DC.  April  10, 1979.
     p. 7-8.

46.  The Politics of Coke.  33 Metal Producing.  March 1980.  p.  49-51.

47.  DeCarlo, J.  A., and M. M. Otero.  Coke Plants in the  United  States on
     December 31, 1959.  Bureau of Mines, U.S. Department  of  the  Interior.
     Washington,  DC.  1960.  p. 4-10.


                                    9-99

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

49.   Schottman, Frederick J.  Iron and Steel.  Mineral Commodity Profiles,
     1983.  Bureau of Mines, U.S. Department of the Interior.  Washington,
     DC.   1983.  p. 4-5.

50.   Reference 18.  p. 3-4,8.

51.   Reference 3.  p. 53-55,58.

52.   Reference 25.  p. 53-55,58.

53.   Reference 4.  p. 3-4,8.

54.   Moody's Investors'  Service.  Moody's Industrial Manual.  Vol. 1  and  2.
     New York.  1984.

55.   Moody1s Investors'  Service.  Moody1s OTC Manual.  New York.   1984.

56.   Standard and Poor's Corp.  Standard and Poor's New York Stock Exchange
     Stock Reports.  New York.  1984.

57.   Dun and Bradstreet, Inc.  File reports printed for use of Research
     Triangle Institute.  January 31, 1985.

58.   Reference 24.  p. S16.

59.   Temple, Barker, and Sloane, Inc.  An Economic Analysis of Proposed
     Effluent Limitations Guidelines, New Source Performance Standards, and
     Pretreatment Standards  for the Iron and Steel Manufacturing Point
     Source Category.  Exhibit 6.  Lexington, Massachusetts.  December 1980.

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

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

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

63.   Reference 59.  p. VI-4.

64.   Reference 24.  p. S28-S29.

65.   Reference 23.  p. 18-5.
                                 9-100

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

67.   Reference 24.  p. S22-S25.

68.   Reference 49.  p. 3.

69.   Reference 24.  p..521.

70.   Reference 23.  p. 18-4.

71.   Reference 49.  p. 14.

72.   Reference 24.  p. S3.

73.   Reference 24.  p. S19-S20.

74.   Reference 25.  p. SI.

75.   Economic Report  of the President,  February  1984.   U.S.  Government
     Printing Office.  Washington, DC.   1984.  p.  224.

76.   Reference 14.  p. S-6.

77.   Reference 28.  p. 63-111.

78.   Reference 43.  p. 1-69.

79.   Letter from Young, E. F., American  Iron  and Steel  Institute,  to
     Pratopos, J. , U.S. Environmental  Protection Agency.   May 16,  1980.

80.   Reference 43.  p. 5.

81.   Reference 43.  p. 85.

82.   Research Triangle Institute.  Economic  Impact of NSPS Regulations on
     Coke Oven Battery Stacks.  Research Triangle Park,  NC.   May 1980.
     p. 8-46, 8-47.

83.   Research Triangle Institute.  An  Econometric Model  of the U.S.  Steel
     Industry.  Research Triangle  Park,  NC.   March 1981.

84.   Ramachandran, V.  The Economics of  Farm  Tractorization in India.
     Ph.D. dissertation.  North Carolina State University.   Raleigh, NC.
     1979.

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

86.   Reference 24.  p. S16, S27.

87.   Reference 25.  p. 55-57,61.
                                 9-101

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

89.   Memorandum from Smith,  C.  Ronald, Office of Standards and Regulations,
     U.S.  Environmental Protection Agency, to Steering Committee Representa-
     tives, EPA.  April 12,  1983.  pp. 1-2 and Attachment I.  Guidelines
     for compliance with Regulatory Flexibility Act.
                                  9-102

-------
             APPENDIX A



EVOLUTION OF THE PROPOSED STANDARDS

-------
A-2

-------
               APPENDIX A - EVOLUTION OF THE PROPOSED STANDARDS

     Work on development of standards for coke oven emissions was officially
begun in March 1975.  The initial effort was directed at limiting participate
emission discharges from coke oven charging, and topside leaks at charging
lids, offtakes, and collecting mains.  A study contracted by the Environmental
Protection Agency in March 1975 provided information on the history and trends
of the byproduct coking industry, industry statistics, processes and emis-
sions, emission control technology and economics.   In this initial effort six
United States plants that were reported to have the best emission control were
visited.  Those plants were as follows:
                                Plants Visited
     U.S. Steel, Clairton, Pennsylvania
     U.S. Steel, Fairfield, Alabama
     Allied Chemical, Detroit, Michigan
     Bethlehem Steel, Franklin & Rosedale Plants,  Johnstown, Pennsylvania
     Inland Steel, East Chicago, Indiana
     Jones and Laughlin, Aliquippa, Pennsylvania
     Prior to and during the data gathering process, discussions were held
with individual steel companies.  Meetings were also held with an Ad Hoc
Committee of the American Iron and Steel Institute (AISI) on March 25, 1974,
August 16, 1974, and May 8, 1975.
     Other meetings were held to review the project with the National Air
Pollution Control Techniques Advisory Committee (NAPCTAC) on February 26,
1975, and April 9, 1975, and an EPA Working Group on February 11, 1975.  The
NAPCTAC is composed of 16 people from industry, state and local air pollution
control agencies, and environmental groups and others with expertise in air
pollution control.  The NAPCTAC meetings were open to the public and were at-
tended by representatives of the steel industry and other interested persons.

                                     A-3

-------
The EPA Working Group is composed of representatives of various offices within
the Agency.
     At the May 8, 1975 meeting with AISI, the NAPCTAC meetings and the first
two meetings with the Working Group, the following recommended standards were
discussed:
     a.   No owner or operator shall cause to be discharged into the atmos-
          phere during charging of a coke oven any emissions which are visible
          in excess of 20 seconds per charge when averaged over 18 consecutive
          charges.
     b.   No owner or operator shall cause to be discharged into the atmosphere
          from:  (1)  the standpipes of a coke oven battery any visible emis-
          sions from more than 10 percent of the total number of standpipes,
          and (2) the charging holes or ports of a coke oven battery when in
          place and sealed any visible emissions from more than 2 percent of
          the total number of charging holes.
     At both the August 1975 Working Group and the September 1975 NAPCTAC
meetings concern was expressed by environmental groups and EPA enforcement
personnel that the recommended emission limits for charging and topside leaks
were too lenient.  It was suggested that the data gathered at U.S. Steel,
Clairton Works and U.S. Steel, Fairfield Works which formed the data base for
the recommended standards, did not reflect present levels of control at either
plant.  Because of the efforts of local control agencies, control at these
plants had reportedly improved significantly since tested by EPA in 1974.  To
verify this reported improved control, additional emission tests were conduct-
ed by  EPA at both plants in the fall of 1975.
     The results of these tests were presented at the December 18, 1975,
NAPCTAC meeting.  Based on these tests, new recommendations were drafted.
These  recommended emission limits were discussed at the February 1976 Working
Group  meeting and were as follows:
     No owner or operator shall cause to be discharged into the atmosphere:
     (a)  Any visible emissions from any charging system and charging ports
          for more than 90 seconds for any 10 or less consecutive charging
          periods.
     (b)  Any visible emissions from more than 12 percent of the offtake
          systems.
                                  A-4

-------
     (c)  Any visible emissions from any collection main.
     (d)  Any visible emissions from more than two percent of the charging
          ports except for oven lids which are removed during coke oven battery
          charging periods and during decarbonizaton periods.
     In May 1976 a draft report entitled "Standards Support and Environmental
Impact Statement:  Standards of Performance for Coke Oven Batteries" was
published and circulated to obtain comments on technical accuracy and policy
implications.
     Insufficient emission data on tall batteries delayed further considera-
tion of the recommended emission limits.  Plant visits to gather additional
data on charging emissions and topside leaks were made to Bethlehem Steel's
Bethlehem plant and U.S. Steel's Fairfield plant late in 1976 and early 1977.
     Growing attention to the presence of carcinogenic substances in coke oven
emissions led to a decision on April 8, 1977 to develop National Emission
Standards for Hazardous Pollutants (NESHAP) contained in coke oven emissions.
In the second week of May 1977 tests were performed at the Wisconsin Steel
coke plant.   Measurements were made of particulate and POM concentrations at
the inlet and outlet of a wet electrostatic precipitator applied to a coke-side
shed.   Sampling was performed during non-pushing periods when door leaks could
be characterized.
     At a steering committee meeting on August 26, 1977 a development plan for
NESHAPs was reviewed.  Approval was withheld pending resolution of certain
issues.
     During the summer and fall of 1977 several plant visits were made to the
U.S.  Steel Gary coke plant to perform measurements of tall oven stage charging
emissions.
     In December 1977 EPA contracted for a study of retrofit information and
other data essential to determining problems in fitting existing batteries
with controls for charging and topside leak emissions.   A report entitled,
"Study to Develop Retrofit Information and Other Data for Use in Setting
Standards for Coke Oven Emissions" was issued in March 1978.
     A Working Group Meeting was held April 26, 1978 to review the issues and
standards development of the coke oven NESHAPs.
     On May 5, 1978 a Section 114 letter was sent to U.S.  Steel requesting
self-inspection data on topside leaks from Batteries 16 and 17 at Clairton
Works covering the period of January 1977 through February 1978.
                                    A-5

-------
     The health risk assessment for coke ovens was reviewed by EPA's Science
Advisory Board on May 30-31, 1978.
     In June 1978, a draft document entitled,  "Standards Support and Environ-
mental Impact Statement:   Volume I:  Proposed  National  Emission Standards
By-Product Coke Oven Wet-Coal Charging and Topside Leaks" was published and
circulated for review prior to a scheduled July NAPCTAC meeting.
     At the NAPCTAC meeting held on July 11-12, 1978, a presentation entitled
"National Emission Standards for By-Product Coke Oven Wet-Coal Charging and
Topside Leaks" was given.   The proposed standard presented for charging emis-
sions was that accumulative visible emissions  from charging could not exceed
88 seconds for 8 consecutive charges.   Leakage topside was proposed to be
limited to 2 percent of the charging lids, 5 percent of the offtakes, and zero
leaks from the collecting main.  Following this meeting the work on standard
development for charging and topside leaks was delayed to examine OSHA con-
siderations and to permit standard development for door leaks to proceed to a
similar point for simultaneous consideration.   The principal concern with OSHA
regulations was potential overlap between the OSHA regulations and any that
might be developed for coke oven sources.  It was necessary to determine
whether OSHA regulations were sufficient to limit emissions to protect public
health.
     On August 29, 1978 a Section 114 letter was sent to Jones and Laughlin
Steel to obtain additional cost data for charging emission control.
     To initiate development of a standard for leaking coke oven doors, EPA
contracted for a study of coke oven door emission control technology in April
1978.  Further important events that have occurred in the development of back-
ground information are presented below in chronological order.
     Date                                       Activity
July 6, 1978             Visit by EPA contractor to Battelle Memorial
                         Institute's coke oven door study laboratory.
August 8, 1978           Plant visit to  Koppers coke plant in Erie,
                         Pennsylvania to examine hand luted technology.
August 22, 1978          Plant visit to  Jones  and Laughlin coke plant at
                         Pittsburgh, Pennsylvania to discuss door  and jamb
                         modifications.

                                    A-6

-------
September 8, 1978

October 4, 1978

October 5, 1978

October 6, 1978

October 26, 1978

November 29, 1978

January 15, 1979
February 14, 1979

February 14, 1979


June 1, 1979

July 12, 1979


July 23, 1979
September 7, 1979
Plant visit to U.S. Steel coke plant at Clairton,
Pennsylvania to discuss door and jamb modifications.
Plant visit to Jones and Laugh!in coke plant at
Pittsburgh, Pennsylvania to obtain a tar sample.
Plant visit to Koppers coke plant in Erie,
Pennsylvania to obtain a tar sample.
Plant visit to Jim Walters Resources coke plant at
Birmingham, Alabama to obtain tar samples.
Plant visit to Bethlehem Steel coke plant at
Baltimore, Maryland to obtain a tar sample.
Door project review meeting to discuss potential
formats for the standard and essential test data.
Meeting between Standards Support Sections and EPA
contractor to discuss the relationship of OSHA regula-
tions to potential EPA standards for doors, charging,
and topside leaks.
Plant visit to Armco Steel coke plant at Houston,
Texas to discuss door hood system and cokeside shed.
Official letter redirecting efforts of contractor
toward a study of the comparability of OSHA regula-
tions with potential EPA regulations.
Work resumed on coke oven door leak standard develop-
ment.
Section 114 letters with questionnaire mailed to
Jones and Laughlin, Bethlehem Steel, Kaiser Steel,
Koppers, United States Steel, and Allied Chemical.
Initial test request was submitted to EMB.
Meeting between Emission Standards and Engineering
Division and EPA contractor to discuss resumption of
work on standards development for wet-coal charging
and topside leaks to be combined with door leaks
standard development.
                                  A-7

-------
September 11, 1979

September 12, 1979

September 18, 1979

September 20, 1979

October 1-5, 1979


November 21, 1979
January 22, 1980

April 22, 1980

June 24, 1980
July 15, 1980

August 26-27, 1980

January 13, 1981

April 21, 1981
June 26, 1981
Plant visit to Bethlehem Steel,  Bethlehem,
Pennsylvania to read door emissions.
Plant visit to Koppers coke plant in Erie,
Pennsylvania to read door emissions.
Plant visit to National  Steel  coke plant at Detroit,
Michigan to read door emissions.
Plant visit to Kaiser Steel coke plant in Fontana,
California to read door emissions.
Plant visit to Armco Steel coke plant in Houston,
Texas to test door emissions captured by coke-side
shed.
Meeting between Emission Standards and Engineering
Division to review coke oven standards development
for wet-coal charging, topside leaks, and door
leaks.
Input data for the economics study transmitted to
EPA.
Meeting with American Iron and Steel Institute
coke oven committee to discuss standard development.
EPA Working Group meeting conducted.
Meeting with American Iron and Steel Institute
coke oven committee.
National Air Pollution Control Techniques Advisory
Committee (NAPCTAC) meeting conducted.
Meeting with American Iron and Steel Institute
coke oven committee.
EPA Steering Committee meeting conducted.
Assistant Administrator concurrence package prepared.
Package returned unreviewed pending completion of
the listing notice.
                                     A-8

-------
August 3, 1982



March 12, 1984
September 13, 1984


September 17, 1984



Sepbember 18, 1984

September 19, 1984

September 21, 1984

November 29, 1984

February 2, 1985



May 1, 1985
Science Advisory Board (SAB) review of the listing
notice.  Listing notice postponed pending revisions
and responses to SAB's comments.
EPA Working Group Meeting conducted.
Plant visit to U.S. Steel's coke plant, Clairton,
Pennsylvania, to gather data on emission controls
and costs.
Plant visit to Bethlehem Steel's coke plant,
Sparrows Point, Maryland, to gather data on emission
controls and costs.
Coke oven emissions listed as a hazardous air
pollutant.
Plant visit to Empire Coke's plant, Holt, Alabama,
to gather data on emission controls and costs.
Plant visit to Armco's coke plant, Middletown,
Ohio, to gather data on emission controls and costs.
Meeting with American Iron and Steel Institute
coke oven committee.
Briefed representatives of State air pollution
agencies on the status of regulatory development
for coke oven emissions.
National Air Pollution Control Techniques Advisory
Committee (NAPCTAC) meeting conducted.
                                      A-9

-------
                  APPENDIX B



INDEX TO ENVIRONMENTAL IMPACT CONSIDERATIONS

-------













































QQ

X
1— 1
Q
Z
UJ
Q,
Q-
^£








































































OO
z
o
1— 1
p_

Z
UJ

o
i—

5<
UJ
Q
Z
I— I





















































































>^
t.
O 10
-p -p
fO d
i— 0)
3 E
cn cu
O) P
s- ro
-P
cn to
c
•r- .p
t. u
ro ro

i ^
c
cu
a

u
o
Q

r~
O
•r-
•P
ro
e
s_
0
4-

I — t

73
C
3
0
S-
cn
v/
U
ro
oa

cu
^
-P

c
• r—
.c
+3
•r—
3:

c
o
•I—
•p

u
o
—1








































































^-^
cn
Q. CLrH
cu e
J- T-
Q.
1 —
s- ro
O 4->
4- C
cu
in S
o> c
c o
•1- $-
r— •!— •
0) >
•O C
•r- CU
3
cn c
0
>\ •!•"
U -P
c u
01 ro
cn
^^
p^.
oo

rv
U_

cn
00
^s


























c
o
•p-
•P
Q.
• r—
S-
u
in
CU
-a

73
c:
ra

73
c
^3
O
i.
cn
\/
u
ro
GO


rH




IO
-a

ro
-a c
C •!-
ro
•P 73
in o)
in
JZ 10
U 3
•r- U
f~ i/}
3-  10
•r- O
-p a. •
ro o CM
C S- .
J_ Q_ ^Q
O>
-P i- C
i— O O
ro 4- -r-
-P
>> c u
S. O) O)
o in oo
•P O
ro jz -
i — U (D
Zf
cn cu s-
fii r> at
S- +->
r— 0.
cu r- ro
r~ *r— r"
l~>0








m
QJ
>
•;—
-P
«
S-
ro
s
s

oo


.
ro







in
(r_

in

J-
ro
-o
c

+j
in

cn
c
•r-"
in
0
a.
o
s-
cx

s- c\l
o
4- S-
cu
in -P
•r- Q.
in ro
CO JZ
-O C_)

>> c
J- -^
o
-P "O
3 0)
-P N
ro -r-
-P S-
(f> CO
E
CD E
x: 3
1— in



in
73
5-
ro
T3
c

-p
in

cn
c
.,—
10
O
a.
0
S-
Q.

S-
o
4-

in
• r—
in
ro
-Q

>}
S-
0
-p

.p
ro
-P
OO


.
-Q





^w
j_
O -
-P r^
ro

3 OO
cn
CD 10
i- j-
0>
O) -P
-C Q.
-P ro
r"
C O
0)
o> c
5^ "I""
-P
0) T3
o CU
in
in in
a. 3
•r- U
-C 10
10 •!-
c ~a
o
•r- CU
-P 5-
ro ro
P—
01 in
J- C
o
in •!-
3 -P
0 0
•r- ro
i.
ro >» •
> U CO

O> CO T3
^= cn c
H- m ro



^>.
u
c
0>
cn
ro

>»
S-
o
-p
CO

3
cn
cu
S-

1-
O)
f~
-p
o

o
-p

a.
•i —
f—
in
a
o in
•r- C
-P 0
ro •!-
— -P
O) U
o: ro


,
u

1 C
ro o i
r- -r- J- >,
3 -P 01 S-
a> o > -P
oi o> o m
S- 00 U =J
T3
O) - in c
JZ OO r— -r-
-P •!-
i- ro cu
>, OI -P -C
J3 -P CU -P •
a. -a T-H
•o ro 4- .
cu -c s- o cn
-POO)
o ^.  -P OO
S- C . =
-p O) U5 O -
in in i -i— cn
3 ai oo e
•at- o t-
C Q-JZ C O)
•r- cn o -P
in 3 o a.
oi -i- o 0) ro
jz ^ \ JT
+-> in jz in o
01 -p in
4- > CU C
O 'I- I—I C •!-
-P 1 -r-
c ro oo m "O
0 C 30)
•p- s- in .Q -P
in o> Q) = C
in -P cn o»
3 i — ro O» in
u ro Q..C 01
LO -P S-
•r- >, - Q-
T3 S- rH Cn
o . c in
.
£_
o
^>
ro

3
cn
cu
S-

a>

-P

>.
J3

T3
CD
-P
U
CU
4- in
4- CU
ro >
•r-
5"} "P
s- ro
•P C
in i-
3 CU
•a -P
C r—
(— i ro


,
-a







in
CU
U
S-
3
O
in

CD
f~
-P

*+—
0

c
o
• r—
in
in
3
U
in
•r-
-a

f-~
ro
u
•i —
c
f~
o
cu
-p

-o
cu

•r-
ro
-P
cu
73

^
J2

T3
cu

u
cu

4-
<0

in
cu
in
in
O)
U
O
1-
D.

^j
>r-
'I
1r.
U
O)
a.
00



CD




1
ro
c
S-
o> •
•P CM
t — .
ro oo

^>^ c
S- 0
0 •!-
-p -p
ro u
r— 0)
3 00
Cn
cu -
s- oo

0) 1-
JZ OI
-P -P
Q.
>, ro
-Q -C
o
•^3
cu c
•P •>-
u
o* *o
4— OI
4- -P
ro c:
OI
in in
CU O)
in s-
in Q.
CD
U in
o •<-
S-
Q. w
0)

C T-
ro •>->



















in
O)

•r—
-p
ro
c
i.
OI
•p

ro

^^
£_
o
-P
ro

3
cn
cu
S-









^ 	 V
T3
O)
3
C
•i —
•P
f
O
u
\^s




















































































B-3

-------





















/•""•s
-o
Ol
^
c
•f—
-p
c
o
u
s^^

(/>
z
o
1 — 1
t—

z
LU
O
l~
x:
LU
O
Z
1— t





















-p
c
Ol
e

u
o
a

c
0

-p
re
E

O
4-
c
1 — I

•a
c
3
o
i-
01

u
re
ca

Ol

-P

c
•r-
.c
-P
•r™
j£

C
o
•r™
-P
re
u
o



Ol
s_
re

IO
O)
>
•i —
•p
re
c
5-
Ol
•p
^~
re

Ol
-C
-p

4—
O


-P
u .
re r**^
Q.
E J-

4_>
C Q.
o re
•r- JZ
-P 0
3
r— C
^— »f—
O
Q.-0
O)
5- 10
•r- IO
re 3
u
01 in
f~ tr—
t- -a


£?
O m
-P -P
re c
r— Ol
3 E
0) -P
S— fD
en in
c
•r— -p
S- U i — >
re re cr>
Q- QLI-H
Oi E "t1
5— *r— f^-
Q. CO
r— -
s- re Q£
0 -P U-
4— C
Ol CTl
en E ro
01 C ^
c o
•r- S.
1 — •(—
Ol >
TD C
•r- Ol

en c
o
~^t '*"*
O -P
c: u
o> re
«C








in
ai
.^
•P
re
c
j-
01
-p

're
c:
QJ 0
_f »r~
-P -P
3
4— i —
O i—
O
in a.
+J
U 1-
re ••-
Q. <
E
i— i
*
re
CM
0)
S-
re

10
o>
>
•i —
+j
re
(~
S-
Ol
-p
f»
re

01
-C
+J

4-
O

10
+J
U
re
a. •
E f^^
•f—
S-
C Ol
0 -P
••- a.
-p re
3 .C
i— 0

o c
Q.-I-

s_ ~^j
0) Ol
•P 10
re 10
jg 3
U
O) 10
-C -r-
(— -a



















c
o
•r—
-P
3
r-»
r—
O
a.

i.
o>
-P
re
3:


.
^


C
•r-

-o
Ol
in
in
3
U
in
•r—
•a

in
• r—

r^
re
in
o
ex
in
•r-
-a

O)
-p
10
re
3

•a
• r-
(^
O
IO

f~
O

^^ .
O 1^
re
Q. S-
E Ol
•r- .p
Q.
01 re
.c x:
(— C_5














i —
re
10
o
Q.
10
•r—
T3

ai
^_>
10
re
3s

-a
•r—
r—
0



.
u







f**»

i.
Ol
+->
Q.
re
-C
t_)

c
*l—

•a
O)
10
10
3
U
l/l
• i —
•a


c
0>

01
.c
l—





















4_)
U
re
Cl.
E
•1 —

>.
D)
5-
01
C
LU


,
-a





en

S-
ai
-P
a.
re
-C
CJ

c
ir—

•a
01
in
in
^
U
10
• i —
-a

Ol
S-
re

in
4J
(J
re
a.
E
•r™

U
•p-
E
0
C
o
u
Ol

Ol
.c
1—



















.p
(J
re
Q.
E
• i—

u
•r—
E
o
C
o
o
LU


,
Ol

B-4

-------
       APPENDIX C



EMISSION SOURCE TEST DATA

-------
                                   APPENDIX C
                            EMISSION SOURCE TEST DATA

     Tables C-l through C-14 present the visible emission data for charging.
The footnotes on Tables C-l through C-14 are notations made by observers
for individual readings of visible emissions.  These notations include
deviations from best emission control practices; such deviations were caused
by observed and suspected lapses in good work practices or equipment failure.
Cases of interference with observer readings caused by fugitive emissions
from other sources such as door leaks or burning coal are also noted.  The
data points associated with observer interference were not judged to be
accurate measurements of charging emissions and were deleted from the data
base.  However, all the remaining data are included in the data base and in
the statistical analysis.
     Tables C-15 through C-22 present the visible emission data gathered by
the Environmental Protection Agency (EPA) for door leaks.  The U.S.  Steel,
Clairton batteries and Batteries B, C, and D at CF&I have modified doors
and a door program with repair and maintenance procedures.   Battery 2 at
U.S. Steel, Fairfield is a 6-meter battery, and Battery 9 is a rebuilt
4-meter battery.  Kaiser Steel uses operating and maintenance procedures
with its Koppers doors, and the Koppers batteries in Erie represent hand-
luted door control.  The data on National Steels' Battery 5 are for 6-meter
ovens equipped with Ikio doors.
     Data for percent leaking lids are presented in Tables C-23 through
C-32. The data on visible emissions from offtakes are compiled in Tables C-33
through C-37.
     The following are brief descriptions of the batteries that are included
in the data base.  The dates in parentheses are the dates of initial opera-
tion or the most recent rebuild.
                                    C-3

-------
U.S.  Steel, Clalrton,  Pennsylvania
     All batteries—Metallurgical  coke,  double mains,  gravity feed,  manual
lid lifters, four (14 in) charging holes,  visible emissions observer,  and
extensive worker training program with monitoring of worker performance.
The company has conducted engineering studies to optimize its stage  charg-
ing procedures and to implement improved door leak controls.
     Batteries 1, 2, and 3 (1955)--Sixty-four Wilputte ovens each, 3.6 meters
in height, and charges 15.6 tonnes of coal  per oven.   To be pad up rebuilt
or replaced in 1986.
     Batteries 7, 8 (1954), and 9 (1964)--Sixty-four Koppers ovens each,
3.6 meters in height,  and charges 15.6 tonnes of coal  per oven.   To  be
abandoned in 1983 and replaced by new Battery C in 1984.
     Batteries 10, 11 (1957), and 12 (1958)—Sixty-four Wilputte ovens
each, 3.6 meters in height, and charges 15.6 tonnes of coal per oven.
Battery 12 was abandoned in 1979,  and Batteries 10 and 11 are to be  replaced
by new Battery B in 1983.
     Battery 12A (1948)--Eighty-five Koppers ovens, 3 meters in height, and
charges 12.5 tonnes of coal per oven.  Abandoned in June 1979.
     Batteries 13, 14, (1953), and 15 (1953/1979)--Sixty-one Koppers ovens
each, 3.6 meters in height, and charges 16.3 tonnes of coal per oven.
Batteries 13 and 14 are being pad up rebuilt to operate in 1980.  Battery 15
had a pad up rebuild in 1979.
     Batteries 16 and 17 (1951)—Sixty-one Wilputte ovens each, 3.6  meters
in height, and charges 16.8 tonnes of coal  per oven.   Battery 17 has a
cokeside shed.  To be abandoned in 1982.   End flue repairs in 1975.
     Batteries 19 (1951/1976) and 20 (1951/1978)—Eighty-seven Koppers-Becker
underjet ovens each, 4.3 meters in height,  and charges 20.8 tonnes of coal
per oven.
     Battery 21 (1947)—Eighty-seven Koppers-Becker underjet ovens,  4.3 me-
ters in height, and charges 20.8 tonnes of coal per oven.  Rebuilt from the
floor up in 1972.  Pad up rebuild planned for 1986-1989.
     Battery 22 (1946)—Eighty-seven Koppers-Becker underjet ovens,  4.3 me-
ters in height, and charges 20.8 tonnes of coal per oven.  Plan pad  up
rebuild in 1986-1989.
                                  C-4

-------
U.S. Steel, Fairfield, Alabama
     Battery 2 (1978)—Fifty-seven Carl Still  ovens,  6.2 meters in height,
four (17 in) charging holes, magnetic lid lifters, mechanical  gooseneck
cleaners, screw feed, wet or dry coal, and double main.   Replaced Batteries 3
and 4.
     Battery 6 (1957)--Seventy-seven Koppers ovens, 3.4 meters in height,
gravity feed, single main with jumper pipe, and charges 14.7 tonnes of coal
per oven.  May be replaced or rebuilt in 1981-1982.  Mechanical repairs in
1973.
     Battery 9 (1954/1979)--Sixty-three Koppers-Becker ovens,  4.3 meters in
height, four (14 in) charging holes, single main with jumper pipe, gravity
feed, mechanical gooseneck cleaner, and charges 18.1 tonnes of coal per
oven.  Major repairs in 1972, mechanical repairs in 1974, pad up rebuild in
1978.
U.S. Steel, Gary, Indiana
     Battery 1 (1970)--Eighty-five Koppers-Becker underfired ovens,
6.2 meters in height, four (17 in) charging holes, disc cutter elbow clean-
ers, magnetic lid lifters, turntable feed, and charges 34.3 tonnes of coal
per oven.  Retrofitted a second main, down for flue repairs and back in
operation in late 1979.
CF&I, Pueblo, Colorado
     All batteries--CF&I conducted an extensive study to optimize stage
charging and to improve door leak controls.  The company uses visible
emission monitoring and a training program.
     Battery B (1972)--Sixty-five Koppers ovens, 4 meters in height, four
(16 in) charging holes, double main, magnetic lid lifters, screw feed,
hydraulic gooseneck cleaner, and charges 18.7 tonnes of coal per oven.  No
major repairs anticipated.
     Battery C (1974)--Forty-seven Koppers ovens, 4 meters in height, four
(16 in) charging holes, double main, magnetic lid lifters, screw feed,
hydraulic gooseneck cleaner, and charges 18.7 tonnes of coal per oven.  No
major repairs anticipated.
     Battery D (1930/1960)—Thirty-one Koppers ovens, 4 meters in height,
four (16 in) charging holes, double main, magnetic lid lifters, screw feed,
hydraulic gooseneck cleaner, and charges 18.7 tonnes of coal per oven. Flue
repairs in 1974.

                                  C-5

-------
Kaiser, Fontana, California
     Batteries F and G (1959)--Forty-five Koppers-Becker ovens each,  4 meters
in height, three (14 in) charging holes,  double main,  table feed,  and
charges 15 tonnes of coal per oven.   Flue rebuild planned for 1982-1983.
Koppers, Erie, Pennsylvania
     Batteries A (1952) and B (1943)--Fifty-eight Wilputte ovens each,
3.6 meters in height, foundry coke,  luted doors, and charges 16.6 tonnes  of
coal per oven.  Cold shut down in 1967,  restarted by Koppers in 1971.   Good
condition, no major repairs anticipated.
J&L Steel, Pittsburgh, Pennsylvania
     Battery P2 (1961)--Fifty-nine Wilputte ovens, 4 meters in height,
three (17 in) charging holes, single main with jumper pipes, rotating table
feed, and charges 17.3 tonnes of coal per oven.  Flue repairs in 1977, at
least 10 years of operation remaining.
     Battery P4 (1953)--Seventy-m'ne Koppers-Becker underjet ovens,  4 meters
in height, three (17 in) charging holes,  single main with jumper pipes,
gravity feed, magnetic lid lifters,  and charges 17.2 tonnes of coal  per
oven.  Flue repairs in 1977.
Lone Star Steel, Lone Star, Texas
     Battery C (I979)--Seventy Koppers ovens,  5 meters in height,  and
double main.  New battery.
Shenango Inc., Neville Island, Pennsylvania
     Battery 3 (1950) and 4 (1952)--Thirty-five Koppers-Becker underjet
ovens each, 4 meters in height, three (16 in)  charging holes, double main,
gravity feed, and merchant, foundry, and metallurgical coke.  A 30-hour
cycle time was in effect during the  October 1980 inspection.  Furnace walls
rebuilt in 1974-1975.  Battery 3 may require pad up rebuild.
Bethlehem Steel, Burns Harbor, Indiana
     Battery 2 (1972)--Eighty-two Koppers ovens, 6 meters in height,  double
main, screw feed, magnetic lid lifters.   New larry cars in 1979.
National Steel, Weirton, West Virginia
     Battery 1 (1973)--Eighty-seven  Koppers ovens, 6 meters in height,
double main, screw feed, automatic lid lifters.
     Sampling was conducted at Armco, Inc. in Houston, Texas, to measure
door leak emissions which were collected by a cokeside shed.  These emissions

                                     C-6

-------
were sampled at the inlet and outlet of the wet electrostatic precipitator
(WESP) during nonpush periods in an attempt to measure emissions from door
leaks alone.  The purpose of this sampling was to attempt a correlation
between mass emissions at the WESP inlet and low values of percent leaking
doors (PLD).  Unfortunately, the door leakage rate ranged from 27 to 45 PLD;
therefore, a correlation of mass emissions with very low percent leaking
doors values could not be obtained.
     Samples were analyzed for benzene soluble organics (BSD) and benso(a)-
pyrene (BaP) using EPA draft methods dated July 3, 1978, and March 20,
1978, respectively.   The level of benzene was determined by Draft Method 111,
and door leak observations were recorded according to Part C of Draft
Method 109.
     Three test runs were conducted.  The first run was not considered
typical of normal operation because the plant was operating on an extended
coking cycle because of equipment failures.  The runs for Tests 2 and 3
were conducted under normal operating conditions—a coke production rate of
820 tonnes per day.   An average of 34 PLD was observed under the cokeside
shed during Tests 2 and 3.
     The BSD results in Table C-38 show a removal efficiency of 40 to 60
percent.   This low efficiency occurred because one of the two parallel
WESPs was inoperative during the tests.   BaP results are shown in Table C-39.
The BaP rates for Tests 2 and 3 were two to three times higher at the stack
than at the WESP inlet.  There was no obvious explanation for these anomalous
results.
     The benzene data (Table C-40) ranged from 1 to 3 ppm at both the inlet
and outlet of the WESP.  This measurement translates into a benzene emission
rate of 0.6 to 2.4 kilograms per hour.
     Table C-41 lists the results of leaking door observations for the
three tests.  Although relatively low PLD numbers were obtained for Test 1,
these low values result from an extended coking time.  Tests 2 and 3 are
considered representative of normal production.
                                    C-7

-------
TABLE C-l.  DATA ON  VISIBLE EMISSIONS FROM CHARGING,
                  CF&I,  BATTERY B
Date
7-26-77



















Mean
7-27-77



















Mean
7-28-77









Oven no.
C2
A4
84
C4
A6
B6
C6
A8
88
C8
A10
BIO
CIO
A12
812
A22
822
Al
81
Cl
= 5.2
C12
812
814,
C14a
A16
816
C16
A18
818
CIS
A20
820
A22a
A9
89
C9
All
811
Cll
A13
= 5.6
B3b
C3
A5
85
C5
A7C
87
C7
A9
89
Duration of
Observer 1
3
3
4
2
4
7
3
6
3
7
3
7
2
7
5
7
6
3
4
4

3
4
8
15
3
2
5
2
3
3
0
3
19
7
8
5
6
9
5
3

10
2
10
7
4
15
9
5
14
6
visible emissions (seconds)
Observer 2
8
5
9
4
7
6
4
9
6
9
5
2
2
10
9
6
11
3
6
3

6
5
11
8
4
2
3
2
2
4
1
0
20
5
5
4
6
6
8
4

10
2
7
4
5
12
8
5
10
4
Observer 3
4
6
7
2
4
7
3
5
5
5
2
2
1
6
6
9
7
1
11
2

1
5
9
10
1
3
2
1
1
1
1
2
25
6
12
5
6
11
5
7

15
5
8
4
4
20
8
5
17
4
                                             (continued)
                        C-8

-------
                   TABLE  C-l.   (continued)
Duration of visible emissions (seconds)
Date
7-28-77









Mean
7-29-77



















Mean
7-30-77











Mean
Oven no.
C9d
Alld
Blld
A19
B19
C19
A21
B21
A23
B23
= 6.7
A15.
B15d
CIS
A17
B17
C17
A19
819
C19r
A21C
B21,
A23d
B23
A8
B8
C8
A10
BIO
CIO
A12
= 4.2
B2b
C2
A4
B4
C4C
A6
B6
C6
A8
B8
C8
A10
= 4.5
Observer 1
5
4
4
4
6
8
4
3
7
4

4
2
2
1
1
3
3
2
1
5
8
6
5
5
6
5
2
5
9
5

4
4
6
5
7
3
5
5
2
11
8
3

Observer 2
4
6
4
3
3
6
4
4
2
4

6
2
3
1
1
4
3
3
2
5
5
5
2
5
4
4
1
3
7
10

1
6
9
4
8
5
6
3
2
2
8
1

Observer 3
3
5
8
8
3
6
6
10
5
3

5
2
3
1
0
9
4
4
4
4
9
6
6
5
6
3
1
6
6
13

0
6
6
3
6
3
5
2
1
2
7
2

Steam aspiration turned off or plugged steam nozzle.
Long charge  time (exceeding the mean  charge time by 30 percent or
more).
Misalignment of the larry car, shrouds, or drop sleeves.
Interference with the observers'  ability to read emissions,  not
used.
                             C-9

-------
         TABLE C-2.   DATA ON VISIBLE EMISSIONS FROM CHARGING,
                         J&L STEEL,  PITTSBURGH
Battery
   Date
Duration of visible
emissions (seconds)
  P2
12-12-78
           12-13-78
           8-15-79
           8-16-79



    Mean =7.4

  P4       8-15-79
    Mean =6.2
       5
       7

      28.8£
      12.2
       6.0
       6.0
       7.4
       3.8

       3
       0
       1.5
       9
       5

       3.6
       7.0
      13.6
       6.2
                                        3.6
                                        2
                                       11
                                        8
                                        5
                                        7
                                        3.4
                                       11
                                        6
                                        5
Problems with door placement.
                                 C-10

-------
TABLE C-3.   DATA ON VISIBLE EMISSIONS  FROM CHARGING,
           U.S.  STEEL,  FAIRFIELD, JUNE  1979
Oven
Battery Date no.
2 6-25 47
52
57
4
14
19a
24
29b
9b
39
44
6-26 19
9
24
34C
29
39
44
49r
54f
lf
6
11^
16
6-27 44e
49
54
1
6
11
16
21
26
36
41
46
51
56
3
Duration of
visible
emissions
(seconds)
5
6
13
8
10
9
6
7
21
5
5
5
11
36
49
2
6
4
5
50
246
22
59
1
21
3
3
5
1
12
2
15
2
9
1
70
3
7
59
                                  23             42
                                  28             3
                                  33             S
                                  38             2
                                  43             1
                                  48             2
                                  53             6
                                   5             6
                                  10             24
                                  15             1
                                  20             2

   Mean =25.3

                                                (continued)
                        C-ll

-------
                    TABLE C-3.  (continued)
Battery Date
6 6-28



















Mean =10.2
Oven
no.
64
74
84
6
16
36
46
56
66
76
78
1
11
21
31
41
51
61
71
81

Duration of
visible
emissions
(seconds)
11
4h
30h
18
9
10
8
2
3
4
6
6
2
4
34
5
3
6
6
33

Emissions during lid replacement.
 Nozzle may be plugged.
C8oot not seating properly.
 Suspect  low steam pressure.
Insufficient aspiration.
 No.  4 sleeve leaking.
^Suspect  steam problems.
 Jumper not connected.
                              C-12

-------
TABLE C-4.  DATA ON  VISIBLE EMISSIONS FROM CHARGING,
        U.S. STEEL,  FAIRFIELD,  BATTERY 6


Date Oven no.
7-9-74

















Mean =
7-10-74



























Mean =
9-30-75




38
48
58
68
78
1
11
21a
31b
41b
51
61
71
81
3
13
23
33
30.4
63
73
83
5
15
25
35
45
55
65
75b
85°
7
17
27
37
47
57
67
77
9
19
29
•5
59
69
79
2
14.2
14
34
44
54
64
Duration of
Observer 1
18
16
14
5
4
11
3
99
4
150
15
11
7
66
20
16
13
16

7
7
50
6
9
3
5
7
18
8
5
115
5
8
8
10
5
—
—
—
—
—
--

	
--
—
—

4
14
6
7
3
visible emissions (seconds)
Observer 2
38
15
11
4
6
12
4
87
7
147
33
20
12
69
31
36
11
7

7
16
37
3
9
6
5
5
9
7
7
98
—
6
11
9
4
6
4
7
6
95
8
13
6
3
8
14

6
18
7
7
4
Observer 3
15
14
9
10
—
9
4
98
—
160
23
11
9
105
8
40
15
12

7
8
30
4
9
6
3
6
7
8
8
45
—
8
12
10
4
7
7
7
6
63
10
11
1
6
9
14

4
8
4
5
2
                                           (continued)
                       C-13

-------
TABLE  C-4 (continued)


Date Oven no.
9-30-75









Mean =
11-30-76














Mean =
12-1-76









Mean =
12-6-76













74
6
16
26
36
46
56
66
76
8
8
71
81
23
33
43
53
63
73
83
5
15
25
35
45
55
27
77
9
19b
47b
29
39
49
59
79
2
22.9
4
14
24
44
54
64,
84b
6
16
26
36
56
66
48C
Duration of
Observer 1
6
13
10
19
11
5
4
7
6
13

37
22
13
32
33
36
19
15
28
45
47
43
100
48
18

2
19
9
73
—
10
32
33
14
8

12
7
4
ib
2
4
42
10
14
6
7
3
6
3
visible emissions (seconds)
Observer ?
7
24
12
7
13
5
6
5
5
13

5
12
5
20
9
51
11
14
17
39
--
34
68
28
23

2
16
10
42
7
10
33
26
16
8

8
5
5
-i
2
5
42
5
16
4
4
2
5
2
Observer 3
4
11
6
6
4
3
5
6
3
10

7
12
5
12
8
33
18
7
22
34
48
5
71
39
12

3
23
18
112
11
16
37
42
21
10

10
6
4
u
"2
4
47
8
16
6
8
2
6
4
                              (continued)
       C-14

-------
TABLE  C-4 (continued)
Date
12-6-76









Mean
12-7-76














Mean
12-8-76
























Mean
Oven no.
58C
68
78
1
11
21
31
41
51
61
= 7.4
11
21
31
41
51
61
71=
81C
3
13
23
33
43
53
63
= 6
65
75b
85°
7
27
37
47
57
67
77
9
29
39
49
59
69
79
2a
12
52
72
82
4
14
24
= 12.1
Duration of
Observer 1
4
3
12
6
2
4
2
5
3
7

12
5
1
16
4
4
5
9
6
7
6
9
3
4
7

6
5
9
6
28
6
8
5
8
6
31
5
15
11
8
4
5
119
14
25
10
27
44
20
7

visible emissions (seconds)
Observer 2
2
3
12
4
2
3
2
5
2
11

13
7
2
16
3
4
7
6
5
5
3
8
2
3
7

3
4
8
3
25
9
8
3
4
7
28
6
14
9
8
6
5
22
15
22
10
16
40
18
10

Observer 3
4
2
14
6
3
3
2
7
4
10

9
6
1
13
5
4
4
7
5
21
4
6
2
2
7

4
4
7
4
20
7
7
3
4
5
19
3
6
5
6
4
4
17
12
14
10
9
34
12
8

                             (continued)
     C-15

-------
                    TABLE  C-4 (continued)
Date Oven no.
12-9-76 24
34
54
64
74b
84°
6
16
26
36
46
56
66
76
8
18
28
38
48
58
68
78
1
11
31
21
Mean =13.6
Duration of
Observer 1
31
5
6
3
9
185
15
4
6
2
13
4
13
3
2
3
5
25
4
4
4
4
5
5
3
4

visible emissions (seconds)
Observer 2
26
7
4
4
20
167
10
4
5
3
14
3
20
5
3
3
4
26
4
7
5
6
5
4
7
8

Observer 3
35
4
7
3
8
158
11
4
5
3
3
3
8
7
4
2
3
24
3
3
3
3
3
3
3
4

Misalignment of the  larry car, shrouds, or drop sleeves.

 Failure to use a jumper pipe or improper seating of jumper pipe.

cLong charge time (exceeding the mean  charge time by 30  percent
 or more).
                            C-16

-------
          TABLE C-5.  DATA ON VISIBLE EMISSIONS FROM CHARGING,
              U.S. STEEL, FAIRFIELD, BATTERY 9, JUNE 1979
Date
6-25















6-26










Oven
no.
57a
67
9h
19b
29
39
49
59
69C
2
12
22
32
42
52

27
47
57
67o
37e
9
19
29
39
49
Mean =5.4
Duration of
visible
emissions
(seconds)
38
12
4
12
4
4
2
6
20
3
4
5
3
4
3

4
2
8
5
4
4
6
2
4
2

Oven
Date no.
6-26 2
12
22
32
42
52
62 .
4d
14
24

6-27 52
62
4
14
24
34
44,
54f
46
56
66
8
18
289
38

Duration of
visible
emissions
(seconds)
4
2
2
4
3
5
2
20
5
4

1
2
3
2
4
2
1
10
2
2
2
2
1
8
13

 Emissions from smoke box.
 Collector pipe emission.
cLast oven, jumper pipe used.
 Emissions from No. 2 boot during leveling.
p
 Liquor spray nozzle plugged.
 Emissions from No. 1 boot.
^Small leak on smoke box.
                                  C-17

-------
         TABLE C-6.   DATA ON VISIBLE EMISSIONS FROM CHARGING,
                  LONE STAR STEEL,  BATTERY C,  OCTOBER 1979



Date
10-23









Duration of
visible
emissions
(seconds)
3
4
10
6
3
6
11
10
4
3
Duration of
visible
emissions
Date (seconds)
10-25 10
13
11
6
9
5
2
8
9
3
10-24
14
 8
10
 5
 6
 6
 6
15
12
10
    Mean =7.4
                                 C-18

-------
TABLE C-7.  DATA ON  VISIBLE EMISSIONS FROM CHARGING,
             U.S.  STEEL,  GARY,  BATTERY 1
Duration of visible emissions (seconds)
Date Oven no.
6-17-76









Mean
6-18-76









Mean
6-28-76









Mean
6-29-76










Mean
11-1-77






Mean
A14
B14
C14
A16
816
C16
A18
818
CIS
A20
= 11
A26
B26
C26
A28
B28
Al
Bl
Cl
A3
83
= 14
C19
A21
B21
C21
A23
B23
C23
A25
B25
C25
= 8
827
C27
A29
B29
A2
I'..'
C2
A4
B4
C4
A6
= 9
A14.
B14b
C14
A16.
B16b
C16
A 18
= 14
Observer 1
-.
5
0
3
2
3
44
33
26
22

15
29
19
15
12
18
15
15
6
2

5
11
8
5
6
10
1
19
8
11

9
8
2
5
4
U
3
8
21
20
~-

20
109
19
5
146
0
8

Observer 2
5
8
0
2
1
3
36
21
20
6

12
25
9
14
9
26
11
9
4
1

4
10
7
3
5
7
2
11
4
7

25
3
—
6
13
U
b
12
21
10
11

46
60
37
12
30
2
16

Observer 3
10
2
2
1
2
1
21
16
11
7

._
—
14
20
15
38
18
11
4
0

9
16
14
3
12
6
15
40
14
11

13
3
3
5
15
7
j
3
20
6
11

15
11
19
3
1
0
3

Observer 4
17
7
1
—
--
--
--
—
--
--

__
—
—
--
—
—
—
—
--
--

6
13
5
6
2
5
9
6
10
8

105a
2
--
4
21
3
6
13
9
16
—

_.
--
--
--
--
--
--

Observer 5
—
—
--
—
--
—
--
—
--
—

..
—
—
--
—
—
--
—
—
--

1
36
5
1
3
12
4
1
5
9

21
0
4
3
7
11
4
10
11
12
7

__
--
--
--
--
--
--

                                             (continued)


                     C-19

-------
TABLE C-7.  (continued)
Duration of visible emissions (seconds)
Date Oven no. Observer 1
11-2-77











Mean
11-3-77
























Mean
11-4-77














All"
811
Cll
C13b
A15b
B15r
C15C
A17.
B17b
C17
A19
B19
= 13.5
A29
829
85
A2
C2
825
C25
A27
827
C27
C21
A23
823
C23
A25.
A19b
819.
C19b
A2L
B21b
A8
88
C8
A10
810
= 8
828
Al
Bin
Cld
A3
C24
A26
826
C26
A28
822
85
C22
A24
B24
12
18
15
17
8
7
32
14
0
15
13
0

25
4
3
3
2
12
7
13
6
6
3
8
3
18
15
14
2
9
13
21
18
17
4
19
5

20
4
5
150
17
9
11
18
31
18
14
6
3
9
12
Observer 2
29
15
6
14
5
6
23
8
0
5
16
0

17
3
2
4
2
6
11
5
6
8
0
6
7
7
15
10
0
8
3
14
11
12
2
18
5

25
7
13
139
15
10
11
16
38
15
8
6
3
13
11
Observer 3 Observer 4
23
26
17
18
3
Q -, ™
38
14
0
8
16
1 — .»

21
4
1
4
2
5
Q __
5
7
11
0
6
10
10
8
12
1
6
7
13
15
12
2
17
2

24
6
7
145
18
11
6
22
33
-_
9
4
1
10
16
Observer 5
—
--
—
--
--
--
--
__
«
--
—
—

--
--
--
—
--
—
--
—
--
--
--
--
--
—
«
--
—
--
--
--
--
--
—
--
--

--
--
--
--
--
—
--
--
--
--
--
—
--
--
--
                                 (continued)
      C-20

-------
                       TABLE C-7.  (continued)
Duration of visible emissions (seconds)
Date Oven no.
11-4-77














Mean
814
C14
A16
816
C16
CIO
A12
812
C12
A14e
A8.
B8b
C8
A10
BIO
= 20.2
Observer 1
15
12
3
14
9
13
15
9
11
105
27
22
11
10
6

Observer 2
15
9
5
18
7
12
14
11
13
123
42
19
10
14
2

Observer 3 Observer 4 Observer 5
—
9
3
15
4
8
10
10
9
94
45
18
10
11
5

aSuspected  recording error, not used.
 Interference with the observers'  ability to read emissions,  not used.
cMisalignment of the larry car, shrouds, or drop sleeves.
 Closed damper between the offtake and main.
 Long charge time (exceeding the mean charge time by 30 percent
 or more).
                               C-21

-------
   TABLE C-8.   DATA ON VISIBLE EMISSIONS  FROM CHARGING,  U.S.  STEEL,
      CLAIRTON, ALL BATTERIES (double collector main,  3  to  4.2 m)
Battery
1
2
3
7
8
9
10
11
16
17
19
20
21
22
Number of
charges
124
124
124
124
112
124
124
124
124
124
124
150
120
52
Average duration of visible
emissions per charge (seconds)3
8.4
4.6
5.7
6.4
6.9
8.3
4.2
4.5
2.6
3.6
11.6
6.4
9.6
3.5
aThe data were collected by U.S.  Steel,  Clairton personnel  in
 December 1979.
                                  C-22

-------
TABLE C-9.   DATA ON VISIBLE EMISSIONS FROM CHARGING, U.S.  STEEL,
                   CLAIRTON, AUGUST 7-10, 1979
Battery
2


















Mean =
20





















Mean =
Oven
no.
A27
B29
A29
B31
B6
A6
B8
A8
BIO
A10
B12
A12
B14
A8
BIO
B12
A12
B16
A16
9.0
A12
B12
C12
A14
B14
C16
A18
B18
C18
A20
B20
C20
A22
B22
C22
C8
A10
BIO
CIO
A12
B12
C12
12.0
Visible
emissions
(seconds)
2.2
9.1
3.5
3.2
5.7
17.6
11.8
13.0
9.2
4.2
4.0
9.5
5.0
2.5
3.0
11.0
6.5
8.0
2.5

6.5
8.0
7.0
13.0
8.5
6.0
7.5
3.0
6.0
7.5
8.5
9.5
6.0
4.5
7.0
5.8
5.0
4.8
4.8
6.5
3.8
4.2

Oven
no.
A24
B26
A26
B28
A28
B30
A30
Al
A24
B26
A26
B28
A28
B30
A30
Cl
Al
Bl
B9

A14
B14
C14
B18
C18
A20
B20
C20
A22
B22
C22
A24
B24
C4
A6
B6
C6
A8
B8
C8
AID
BIO

Visible
emissions
(seconds)
8.0
5.0
6.0
13.0
2.0
6.0
4.0
13.0
21.0
6.0
5.0
5.0
6.0
5.0
4.0
5.0
11.0
7.0
10.8

3.5
2.0
3.5
5.0
8.5
4.0
5.0
10.5
8.0
7.0
6.0
7.5
4.5,
127.0°
13.0
4.5
7.0
5.5
2.0
6.0
4.0
4.0

Oven
no.
A9
Bll
All
A13
B13
B9
A9
Bll
All
B13
A13
B15
A15
B17
A17
B19
A19



A12
C14
A16
B16
C16
A18
B18
CIS
A20
B20
C20
A16
B16
C16
A18
B18
C18
A20
B20
C20
A22


Visible
emissions
(seconds)
14.0
13.7
10.0
7.2
5.2
22.8
9.6
19.5
21.5
19.8
15.8
10.8
8.4
10.0
11.0
11.7
10.2



24. Oa
15.5
11.5
7.0
10.0
10.0
8.5
25.0,
120.0°
6.0
17.0
24. Oc
9.5
12.0
14.0
17.0
15.0
8.5
10.5
15.5
9.5


                                                           (continued)
                            C-23

-------
                         TABLE C-9.  (continued)
Battery
22




















Mean
Oven
no.
A19
B19
C19
A21
B21
C21
A23
B23
C27
A29
B29
C29
A2
B2
C2
A4
B4
C4
A6
B6
B27
= 6.6
Visible
emissions
(seconds)
4.5
9.2
2.6
5.5
9.3
9.5
7.0
8.5
2.5
0.5
4.0
2.5
5.5
10.0
2.0.
41. 01
3.0.
0.5J
4.0
2.0
1.0

Oven
no.
C27
A29
B29
C29
A2
B2
C2
A4
B4
B6
C6
A8
B8
C8
A10
BIO
CIO
A12
B12
C12
A14

Visible
emissions
(seconds)
3.0
8.5
13.5
9.0
3.5
3.0
1.0
7.0
4.0
6.0
5.0
6.5
3.0
3.5,
10. On
5.5
3.0
1.5
6.5
15.0
13.0

Oven
no.
A18
B18
C18
A20
B20
C20
A22
B22
C22
A24
824
C24
A20
C20
A22
B22
C22
A24
B24
C24


Visible
emissions
(seconds)
18. Oe
4.0f
13. 0T
4.0
4.0
12.0
9.5
4.5
2.0^
19. 5g
4.5
3.0
9.5
7.0
2.0
3.0
11.0
4.0
6.5
2.0


 Coal stuck in No. 3 hopper.
 Larry car No. 14 problem with volumetric metering of coal into hopper No. 3.
P
 Delay waiting for leveling machine.
 All emissions from lid No. 3 after drop sleeve was raised.
Emissions from lids 1 and 4 while charging.
 Emissions from lid 2 while charging.
^Riv-iil , harg*; r^te, end >>f shift, heavy emissions out of all lids.
hSteam turned off too soon.
Plugged liquor spray.
••'Pushing emissions interfered for 1 to 2 minutes.
                                  C-24

-------
TABLE C-10.
U.S.
Battery
1






Mean
3





Mean
7





Mean
8





Mean
9





Mean
10





Mean
Oven
no.
A2
B2
C2
A4
B4
A6
B6
= 8.6
A2
82
C2
A4
B4
A6
= 7.8
B9
All
Bll
A13
A15
B15
= 7.6
Cl
B3
A3
B5
A5
B7
= 8.9
811
A13
813
A15
B15
A17
= 8.4
A25
825
A27
827
A29
829
= 8.8
DATA ON VISIBLE
STEEL, CLAIRTON,
Visible
emissions
(seconds)
2
2a
82a
2
7
3
7

3
3
5
8
9
8

8
7
11
8
12
7

15
7
9
10
8
10

20d
8
5
7
3
10

22f
9
9
46
4
6

Oven
no.
A8
814
A16
B16
A18
818
A20

B6
AS
A14
814
A16
816

A17
817
A23
823
A25
825

Bll
All
813
A13
815
A15

817
A19
819
A23
823
A25

A31
831
86
A8
88
A10

EMISSIONS FROM CHARGING,
FEBRUARY 4-9, 1980
Visible
emissions
(seconds)
4
2
8
2
9b
12°
6

10
7
5
3
9
10

3
9
5
6
10.
6C

8
3
13
7
13
7

8
3
10
6
8
6

3
5
4
4
11
3

Oven
no.
820
A22
85
A7
87
A9


A18
818
A20
820



A27
827
A29
829



817
A17
819
A19



825
A27
B27
A29



810
A12
812
814



Visible
emissions
(seconds)
3
4
4
7
3
2


15
15
4
10



7
6
7
9



8
5
16
3



18e
5
12
6



2
3
8
2



                           (continued)
C-25

-------
                       TABLE C-10.  (continued)
Battery
11





Mean
19





Mean
21





Mean
Oven
no.
A25
B25
A27
827
A29
B29
= 11.6
A16
B16
C16
A18
812
C12
= 6.8
A21
821
C21
A23
323
C23
-' 13.1
Visible
emissions
(seconds)
13h
isj
14h
13n
5h
10h

8k
14*
7
8
8
5

15
4
2
7
4
9

Oven
no.
A31
B31
A8
88
A10
810

A14
814
A16
C22
A24
824

Cll
A13
813
C13
C23
A25

Visible
emissions
(seconds)
8b
19°
3
14
4
10

7
5
2
8
7
11

8
6
8
14,
96 '
10

Oven
no.
A12
812
A14
814



C24
A26
826
C26



825
A27
B27
C27
A29
B29

Visible
emissions
(seconds)
8
7. .
29 .'
11^



7
8
2
2



l
32m
8
1
6
4

 Poor drop sleeve  alignment on No. 1 lid; heavy emissions from No.  2 and
 No.  3 lids during charging; bursts of emissions when chuck door opened and
 leveling commenced.
 Emissions at the  start of and during leveling.
cSmoke from cokeside  door obstructed view during first minute.
 Larry car drifted off holes.
eDamper arm fell on cokeside, lids No. 3 and No. 4 emitting.
 Carbon buildup blockage possible.
9Puffs from No.  1  lid.
 Emissions during  leveling.
^uch puffing during  charging.
•'Abnormal charge:   lidman failed  to remove all  lids before charging.
I,
 Poor drop sleeve  alignment.
 Most emissions from lids  No.  3 and No  4 during No. 2 lid discharge and
"^Replacing collar on pusherside causing  steam  disruption.
                                 C-26

-------
     TABLE C-ll.  DATA ON VISIBLE EMISSIONS FROM CHARGING,  U.S.  STEEL,
                       CLAIRTON, AUGUST 13-15,  1979

Oven
no.
B3
A9
B9
All
A23
A25
Bll
A13
All
Bll
A13
B13
B19
A21
B21
B23
A19
B19
A21
B21
A23
B23
A25
B25
Al
B2
A4
B4




Mean =
Battery 16a
Visible emissions
(seconds)
1.5
0
1
1
0
0
0
0
0
0
0
0
1
0
1
0
0
0
0
2
0.
13b
0
0.5
0.5
0
0.5
5.5




1.0

Oven
no.
B6
A8
B8
A10
A14
A16
B14
B16
A18
B18
A20
A22
A18
A20
A22
B22
A28
B28
A30
B30
A30
B30
Al
Bl
A3
B3
A5
B5
Bll
A13
A15
B13
Mean =1.2
Battery 17a
Visible emissions
(seconds)
4
1
2
1
0
10
4
2
1
0
0
2
1
1
1
0
0
0.5
0.5
0.5
1
1
0
0.5
1
0
0
0
0.5
0.5
1
0

a,,
  Run of the mine" (nonpulverized) coal was charged.
5Drop sleeves No.  1 and No. 3 not sealed.
                                  C-27

-------
TABLE C-12.   DATA ON VISIBLE EMISSIONS FROM CHARGING,
         SHENANGO, INC., OCTOBER 20-22, 1980

Oven
no.
94
104
84
114
82
92
89
109
79
3
107
87
81
101
111
91
83
103
113
93
Mean =
Battery 3
Visible emissions
(seconds)
4
3
3
6
4
10
6
4
1
23
9
6
6
3
7
4
12
3
4
4
6.1

Oven
no.
124
135
144
154
129
139
149
119
117
127
137
147
121
131
141
151
123
133
143
153
Mean =6.6
Battery 4
Visible emissions
(seconds)
3
8
4
4
9
8
7
4
4
11
7
10
4
4
9
8
6
7
5
10

                       C-28

-------
        TABLE C-13.   DATA ON VISIBLE EMISSIONS FROM CHARGING,
               NATIONAL STEEL, WEIRTON, WEST VIRGINIA,
                              BATTERY 1
Date
8-81






Mean =
6-82






Oven
no.
A- 7
A- 9
B-3
C-3
B-5
C-5
C-7
13.5
B-10
C-10
B-12
C-12
A- 14
B-14
C-14
Visible
emissions
(seconds)
5.0
8.0
3.0
30.0
7.0
12.0
12.0

14.0
7.0
15.0
3.0
6.0
11.0
28.0
Oven
no.
B-7
B-9
B-ll
C-ll
B-13
B-ll
A- 7

A- 16
B-16
A- 18
B-18
B-10
C-10
B-12
Visible
emissions
(seconds)
6.0
8.0
4.0
12.0
26.0
8.0
32.0

18.0
2.5
3.5
24.0
5.0
16.0
22.0
Oven
no.
A- 9
A- 11
A- 15
A- 17
C-ll
B-13
C-13

C-12
C-14
B-14
C-14
A- 16
B-16

Visible
emissions
(seconds)
4.5
7.0
3.0
43.0
17.0
15.0
22.0

2.0
4.0
3.5
11.0
3.0
5.0

Mean =10.2
                                C-29

-------
         TABLE C-14.   DATA ON  VISIBLE EMISSIONS  FROM CHARGING,
          BETHLEHEM STEEL, BURNS HARBOR,  INDIANA,  BATTERY 2,
                            MAY 19-23,  1980

Oven
no.
224
244
254
264
274
236
246
256

Mean =
Visible
emissions
(seconds)
26
2
1.4
4a
2a
43
35
9

10.8

Oven
no.
266
276
286
208
218
234
238
286


Visible
emissions
(seconds)
8.6
11.9
19.2
9.7
12.5
3.2
4
10a



Oven
no.
222
208
218
224
238
248
258
268
278

Visible
emissions
(seconds)
10.5
1.5
1.0
13.5
2.0
2.0
1.5
3
4a

Interference with the observer's ability to read emissions, not used.
                                 C-30

-------
         TABLE C-15.   DATA ON VISIBLE EMISSIONS FROM DOOR LEAKS,
                      U.S.  STEEL, FAIRFIELD,  JUNE 1979

                               Start        Finish          Percent
Battery          Date          time          time        leaking  doors

   2             6-25          1152          1159             12.7

                 6-26          1225          1241              2.6
                               1603          1611              6.1

                 6-27          1118          1127             10.2
                               1559          1604              8.1

                 6-28          1235          1247              5.3
                               1553          1607              5.6

   9             6-25          1727          1733              3.5

                 6-26          1128          1134              3.4
                               1551          1559              2.7

                 6-27          0852          0912              4.8
                               1122          1129              4.8
                               1436          1448              4.0
                                  C-31

-------
        TABLE  C-16.   DATA  ON  VISIBLE  EMISSIONS  FROM  DOOR  LEAKS,
              KAISER  STEEL, FONTANA,  CALIFORNIA,  MAY 1979
Date
5-8
5-9
5-10
5-11

Battery F
7.8
5.6
3.3
4.4
0.0
2.2
1.1
5.6
Percent leaking doors
Battery G
3.3
2.2
7.8
5.6
0.0
0.0
2.2
7.8
         TABLE C-17.   DATA ON VISIBLE EMISSIONS FROM DOOR LEAKS,
            KOPPERS,  ERIE, PENNSYLVANIA,  SEPTEMBER 11,  1979

                                           Percent leaking
            Battery                             doors

               A                                  2
                                                  1
                                                  2
               B                                  7
                                                  0
                                                  5
TABLE C-18.   DATA ON VISIBLE EMISSIONS FROM DOOR LEAKS, NATIONAL STEEL,
          ZUG ISLAND, MICHIGAN, BATTERY 5, SEPTEMBER 18, 1979
Percent leaking doors
Run
no.
1
2
3
Observer
CS
61
64
39
1
PS
39
55
65
Observer
CS
38
44
33
2
PS
25
52
58
Observer
CS
34
46
47
3
PS
27
62
62
CS = cokeside.
PS = pusherside.
                                  C-32

-------
TABLE C-19.   DATA ON VISIBLE  EMISSIONS FROM DOOR LEAKS,
            U.S.  STEEL,  CLAIRTON, AUGUST 1979
Battery
1







2







19





20









21











Date
8-7

8-8

8-9

8-10
Average
8-7

8-8

8-9

8-10
Average
8-7
8-8
8-9

8-10
Average
8-7
8-8

8-9

8-10
8-13
8-14
8-15
Average
8-7
8-8
8-9

8-10
Average
s-s

8-9

8-10
Average
Number of
doors leaking
1
3
1
0
2
1
1

2
7
0
1
5
0
5

6
6
8
6
8

5
8
9
13
11
8
6
4
6

5
4
2
10
b

4
4
9
6
8

Percent
leaking doors
0.8
2.3
0.8
0
1.6
0.8
0.8
1.0
1.6
5.5
0
0.8
3.9
0
3.9
2.2
3.4
3.4
4.6
3.4
4.6
3.9
2.9
4.6
5.2
7.5
6.3
4.6
3.4
2.3
3.4
4.5
2.9
2.3
1.1
5 7
2. ')
3 0
^ 3
2.3
5.2
3.4
4.6
3.5
                          C-33

-------
TABLE C-20.  DATA ON VISIBLE EMISSIONS FROM DOOR LEAKS,
          U.S. STEEL, CLAIRTON, FEBRUARY 1980
Battery Date
3 2-4
2-5
2-6
2-7
Average
7 2-4
2-5
2-6
2-7
Average
8 2-4
2-5
2-6
2-7
Average
9 2-4
2-5
2-6
2-7
Average
10 2-6

2-7
Average
11 2-6

2-7
Average
16 2-4
2-5
2-6
2-7
Average
17 2-4
2-5
2-6
2-7
Average
Number of
doors leaking
0
1
1
2

3
5
7
0

7
8
8
8

10
12
5
21

7
6
9

7
9
18

5
6
8
5

4
3
3
5

Percent
leaking doors
0
0.8
0.8
1.6
0.6
2.3
3.9
5.5
0
2.9
5.5
6.3
6.3
6.3
6.1
7.8
9.4
3.9
16.4
9.4
5.6
4.8
7.1
5.8
5.5
7.0
14.1
8.9
4.2
5.0
6.7
4.2
5.0
6.6
4.9
4.9
8.2
6.2
                          C-34

-------
TABLE C-21.   DATA ON VISIBLE EMISSIONS FROM DOOR LEAKS,
             CF&I, FEBRUARY AND MARCH 1978
Date
2-15
2-16
2-17
2-22
2-23
2-24
3-2
3-7
3-8
3-9
3-13
3-15
3-16
3-20
3-21
Average

Battery B
2.3
2.3
1.5
3.8
4.6
6.9
7.7
7.7
2.3
4.6
6.9
9.2
2.3
4.6
2.3
6.2
3.8
2.3
1.6
5.4
0.8
2.3
2.3
5.4
2.3
6.2
6.9
4.6
4.3
Percent leaking doors
Battery C
3.2
1.1
5.3
2.1
5.3
2.1
1.1
3.2
4.3
5.3
3.2
3.2
3.2
1.1
4.3
2.1
1.1
2.1
3.2
2.1
6.4
2.1
4.3
6.4
3.2
3.2
3.2
3.2

Battery D
8.1
3.2
6.5
8.1
8.1
4.8
6.5
8.1
6.5
9.7
16.1
6.5
4.8
1.6
6.5
9.7
3.2
1.6
6.5
4.9
1.6
3.2
6.5
6.5
1.6
3.2
4.8
5.9
                         C-35

-------
     TABLE C-22.   DATA ON VISIBLE EMISSIONS FROM DOOR
      LEAKS,  NATIONAL STEEL,  WEIRTON,  WEST VIRGINIA,
                   BATTERY 1, JUNE 1982
Date
6-7-82
6-8-82
6-8-82
6-9-82
6-10-82
Number of door
area leaks
7
4
8
3
2
Percent leaking
doors
4.2
2.4
4.8
1.8
1.2
aFor a total of 166 doors on operating ovens.
                           C-36

-------
TABLE C-23.  DATA ON VISIBLE EMISSIONS FROM LID LEAKS,
         U.S. STEEL, FAIRFIELD, JUNE 1979
Battery Date
2 6-25
6-26
6-27
6-28
Start
time
1210
0817
1100
1557
0907
1529
Finish
time
1214
0820
1102
1600
0909
1532
Percent
leaking lids
0
0.5
0
0
0
0
       6-25          1557          1559              0

       6-26          1047          1049              0
                     1541          1543              0

       6-27          0835          0839              0
                     1113          1116              0
                     1427          1430              0
                       C-37

-------
TABLE C-24.   DATA ON VISIBLE EMISSIONS FROM LID LEAKS,
      KAISER STEEL, FONTANA, CALIFORNIA, MAY 1979
Date
5-8
5-9
5-10
5-11

Battery F
0
0
0
0
0
0
0
0
Percent leaking lids
Battery G
1.7
0
0
0
0
0
0
0
TABLE C-25.  DATA ON VISIBLE EMISSIONS FROM LID LEAKS,
                 J&L STEEL, PITTSBURGH
Date
12-12-78
12-13-78
8-15-79
8-16-79
Percent
Battery PI
0.0
0.6
0.0
0.0
0.6
0.6
0.6
leaking lids
Battery P4
0.4
0.0
0.0
0.4
0.0
0.0
0.0
                        C-38

-------
  TABLE C-26.   DATA  ON VISIBLE EMISSIONS FROM CHARGING PORT LID LEAKS,
                  U.S. STEEL, FAIRFIELD, BATTERY 5
Date
7-8-74





7-9-74




7-10-74

7-11-74









7-12-74





9-30-75







10-1-75














Clock
time
1200
1245
1315
1405
1457

1215
1300
1400
1500

708

715
1005
1015
1115
1254
1335
1430
1440
1550

705
835
1030
1245
1458

930
1010
1011
1013
1023
1125
1131

852
853
855
858
859
900
1025
1027
1028
1030
1032
1033
1041
1334
1343
Percent
leaking
lids
0
0.3
0.3
0
0

0
0
0
0

0

0-1. 9a
0
0-0. 3a
0
0-1. Oa
0
0
0
0

0.3-2.93
0
0-1. 3a
0
0

0
0
0
0
0
0
0

0
0
0
0
0
0
0
0
0
0
0
0
2.3
0
0.3
Clock
Date time
11-30-76 852
952
1059
1311
1409
1511
1611

12-2-76 804
910
1009
1104
1331
1431
1517
1622

12-3-76 833
934
1037
1143
1331
1436
1528

12-4-76 747
844
948
1047
1224
1327
1420

12-9-76 728
827
921
1019
1214
1321
1423












Percent
leaking
lids
0
0
0
0.3
0.6
2.9
3.2

0
0.3
0
0
0.3
1.0
2.6
1.3

0.3
0
0
0
0.3
0
0.3

0
0
0
0
0.6
0.3
0

0
0
1.0
1.0
0
0
0.6












Multiple observers.
                             C-39

-------
          TABLE C-27.   DATA ON VISIBLE EMISSIONS FROM CHARGING
                       PORT LID LEAKS, U.S.  STEEL,
                           FAIRFIELD, BATTERY 6
Date
7-9-74






7-10-74





9-30-75










10-1-75













Clock
time
707
810
820
906
915
1030
1117
705
717
855
1000
1030
1115
1303
1338
1343
1346
1502
1509
1717
1719
1720
1726

936
937
941
943
948
950
1100
1102
1103
1108
1110
1112
1333
1341
Percent
leaking
lids
0
2.9
0
0.3
1.6
0
0
0
0
0
0-0. 3a
0
0
0
0
0
0
0
0
0
0
0.3
0

0
0
0
0.3
0
0
0.3
0.3
0
0
0
0
0
0
Clock
Date time
12-1-76 858
1109
1334
1433
1528
1620
1625
12-6-76 756
847
937
1251
1324
1432
12-7-76 738
827
908
1105
1219
1318
1416

12-8-76 757
853
935
1037
1228
1321
1426

12-9-76 713
813
908
1009
1208
1309
1411


Percent
leaking
lids
0.3
0.3
0.3
0.3
0.3
0.3
0
0
0
0
0
0
0
0
0
0
0
0
0
0

0
0.6
0
0.3
0
0
0.6

0.3
0
0
0.6
0
0.6
0.6


Multiple observers.
                                C-40

-------
TABLE C-28.  DATA ON VISIBLE EMISSIONS FROM CHARGING PORT LID LEAKS,
                  U.S. STEEL, FAIRFIELD, APRIL 1976
Battery
5








Mean
6








Mean
Date
4-21



4-22

4-23


= 0.1
4-21


4-22

4-23



= 0.1
Clock
time
1228
1312
1355
1439
0728
1116
0753
1025
1430

1340
1417
1456
0800
0937
0816
1047
1400
1442

Percent lid leaks
Observer 1
0
0
0.3
0
0.3
0
0
0
1.0

0.3
0.3
0.3
0
0
0
0
0.7
--

Observer 2
0
0
0.3
0
0
0
0
0
0

0
0.3
0
0
0
0
0
0
—

Observer 3
0
0
0.3
0
1.0
0
0
0
0.7

0
0.3
0
0
0
0
0
0
1.0

                              C-41

-------
       TABLE  C-29.  VISIBLE EMISSIONS DATA ON TOPSIDE LEAKS FROM
           CHARGING PORT LIDS, U.S. STEEL, CLAIRTON
Date
September 1977






October 1977






November 1977






December 1977






January 1978






February 1978






Battery
12A
15
16
17
19
21
22
12A
15
16
17
19
21
22
12A
15
16
17
19
21
22
12A
15
16
17
19
21
22
12A
15
16
17
19
21
22
12A
15
16
17
19
21
22
Number of
traverses
26
26
25
25
26
26
26
26
29
29
28
29
28
28
21
1 26
26
26
23
24
25
14
25
26
25
22
26
27
25
14
20
20
20
22
21
16
27
27
27
27
27
27
Percent leaks .
observed per traverse
Range
0-0.6
0-1.2
0-1.2
0-0.8
0-1.1
0-1.5
0-1.1
0-0.6
0-0.8
0-0.8
0-0.8
0-1.1
0-0.9
0-1.1
0-0.6
0-1.2
0-1.2
0-0.8
0-0.6
0-0.6
0-0.6
0-0.9
0-0.4
0-0.8
0-0.4
0-0.9
0-1.2
0-0.9
0-3.0
0-1.2
0-1.2
0-2.0
0-2.6
0-3.5
0-2.3
0-1.8
0-2.9
0-1.2
0-3.3
0-1.7
0-3.2
0-0.9
Average
0.1
0.2
0.2
0.2
0.3
0.3
0.5
0.1
0.2
0.1
0.1
0.2
0.2
0.3
0.1
0.1
0.1
0.1
0.1
0.2
0.1
0.1
0.0
0.1
0.1
0.1
0.3
0.2
0.3
0.2
0.3
0.4
0.4
0.9
0.7
0.7
0.8
0.3
0.5
0.3
0.6
0.3
A traverse is one recording  of  leaks  from  an  entire battery.
Percentage of total  potential  leaks.
                               C-42

-------
TABLE C-30.   DATA ON  VISIBLE  EMISSIONS FROM LID LEAKS,
           U.S.  STEEL,  CLAIRTON, AUGUST 1979
Battery
1





2





3





7



8



9



10



11



12



16








Date
8-7

8-8

8-9
Average
8-7

8-8

8-9
Average
8-7

8-8

8-9
Average
8-7
8-8
8-9
Average
8-7
8-8
8-9
Average
8-7
8-8
8-9
Average
8-7
8-8
8-9
Average
8-7
8-8
8-9
Average
8-7
8-8
8-9
Average
8-7
8-8
8-9
8-13

8-14

8-15
Average
Number of
leaking lids
0
0
1
1
0

0
5
0
0
0

0
5
1
0
0

3
6
3

0
0
1

2
3
1

1
3
1

0
0
0

0
0
0

0
0
0
0
1
0
0
0

Percent
leaking lids
0
0
0.4
0.4
0
0.2
0
2.0
0
0
0
0.4
0
2.0
0.4
0
0
0.5
1.2
2.3
1.2
1.6
0
0
0.4
0.1
0.8
1.2
0.4
0.8
0.4
1.2
0.4
0.7
0
0
0
0
0
0
0
0
0
0
0
0
0.4
0
0
0
0.1
(continued)
                     C-43

-------
TABLE C-30 (continued)
Battery Date
17 8-7
8-8
8-9
8-13

8-14
8-15
Average
19 8-7
8-8
8-9
8-13
8-14
8-15
Average
20 8-7
8-8
8-9

8-13
8-14
8-15
Average
21 8-7
8-8
8-9
8-13
8-14
8-15
Average
22 8-7
8-8
8-9

8-13
8-14
8-15
Average
Number of
leaking lids
0
0
0
1
0
0
2

2
7
12
0
4
0

3
6
5
6
1
0
1

1
0
3
3
2
0

2
6
0
1
4
6
2

Percent
leaking lids
0
0
0
0.4
0
0
0.8
0.2
0.6
2.0
3.4
0
1.1
0
1.2
0.9
1.7
1.4
1.7
0.3
0
0.3
0.9
0.3
0
0.9
0.9
0.6
0
0.5
0.6
1.7
0
0.3
1.1
1.7
0.6
0.9
       C-44

-------
TABLE C-31.  DATA ON VISIBLE EMISSIONS FROM LID LEAKS,
           U.S. STEEL, CLAIRTON, FEBRUARY 1980
Battery Date
1 2-4
2-5
2-6
2-7
Average
2 2-4
2-5
2-6
2-7
Average
3 2-4
2-5
2-6
2-7
Average
7 2-4
2-5
2-6
2-7
Average
8 2-4
2-5
2-6
2-7
Average
9 2-4
2-5
2-6
2-7
Average
10 2-6

2-7
Average
11 2-6

2-7
Average
Number of
leaking lids
1
1
0
0

3
0
0
0

8
3
1
1

0
0
2
2

0
0
2
1

0
1
1
2

0
0
0

1
0
1

Percent
leaking lids
0.4
0.4
0
0
0.2
1.2
0
0
0
0.3
3.1
1.2
0.4
0.4
1.3
0
0
0.8
0.8
0.4
0
0
0.8
0.4
0.3
0
0.4
0.4
0.8
0.4
0
0
0
0
0.4
0
0.4
0.3
                                                (continued)
                        C-45

-------
TABLE C-31 (continued)
Battery Date
16 2-5
2-6
2-7
Average
17 2-5
2-6
2-7
Average
19 2-4
2-5
2-6
2-7

Average
20 2-4
2-5
2-6
2-7

Average
21 2-4
2-5
2-6
2-7
Average
22 2-4
2-5
2-6

2-7
Average
Number of
leaking lids
1
0
0

0
1
0

2
0
3
0
2

3
5
0
1
2

2
1
0
2

3
1
0
0
0

Percent
leaking lids
0.4
0
0
0.1
0
0.4
0
0.1
0.6
0
0.9
0
0.6
0.4
0.9
1.4
0
0.3
0.6
0.6
0.6
0.3
0
0.6
0.4
0.9
0.3
0
0
0
0.2
          C-46

-------
          TABLE C-32.   DATA ON VISIBLE EMISSIONS FROM LID LEAKS,
                       SHENANGO,  INC., OCTOBER 1980

                                 Number of             Percent
Battery          Date          leaking lids          leaking lids
3 10-20
10-21

10-22

Average
4 10-20
10-21

10-22

Average
0
0
0
0
0

0
0
0
0
0

0
0
0
0
0
0
0
0
0
0
0
0
                                 C-47

-------
TABLE C-33.  DATA ON VISIBLE EMISSIONS FROM OFFTAKE LEAKS,
            U.S.  STEEL, FAIRFIELD, JUNE 1979
Battery Date
2 6-25
6-26
6-27
6-28
9 6-25
6-26
6-27
TABLE C-34.
Date
12-12-78
12-13-78
8-15-79
8-16-79
Start
time
1216
0850
1116
1557
0912
1539
1601
1052
1544
0838
1117
1431
Finish
time
1223
0854
1118
1600
0915
1544
1606
1054
1546
0841
1120
1434
DATA ON VISIBLE EMISSIONS FROM OFFTAKE
J&L, PITTSBURGH, BATTERY P4





Percent leaking
offtakes
0.0
5.1
1.3
0.0
2.5
0.0
0.0
Percent
leaking offtakes
10.0
15.9
2.7
1.8
0.9
14.3
0
0
0
1.7
0
0
LEAKS,





                        C-48

-------
     e
     5
 O
i- re
QJ 
in 01
*> en-—
c c

u xi o
I- re
QJ O) in
Q. t-

?
4J
O t/>
c c *"

 .1- 1-
O XI O
u m
a; QJ in
Q- 1-

CT



•*— re in
O 10
in OJ
c c
(U -i- t-
U XI O
I- TO
a» a> m
Cv S-


a>

•*-*
O m
*J O)-—
C C
Qt -r- i.
O X) O
u re
a* aj m
Q- i-


>






ai

re
a





TO U
03 *
C O
"- CX
4-J C
—1 E





CM CM CM



f~^ r*- co




in in CM





CT, m o


O CM v£
« « 00




n us r>
m in CM




00 CM CO
in ID oo



r*- «j- o




ID ^j- cn
rH rH CM






o «r cn
CM rH CO





m cn ^>
m 10 i*^






rH O CO




$ £ In













cn cn
• — r- 1 r- 1

a. TO c













CM
CM



^










^


^
in




m
«•




cn




o




o
CM






CO
o





in







m




S







^
cn


(U
c


(
j 	

[-
a.















-> m CM in «• m




in r»- oo co r** ^





m ^t co r* oo oo


in in r** cn vc o
cn en cn co oo o




cn o u? o ic CM
rH CM rH CM CM CM




cn cn cn cn o 00



co c, a, oo oo ^




In £ S S 5 In






r^ ao «-H m GO ro
^ ro ^ in in ^~





^D ^D o i^ r^ r*.
r"* oo o cT* in in
rH





Cn i— 1 CM rH CM ^*




CO rH rH CM rH P*.
r*- cn cr> oo r^ m








r*-


r*- en rH r»> cn cn
Cn rH Cn rH rH
i-H t- r-*
+-> u s- t-
>, in J3 t. QJ a>
•— 3 E OJ J3 .O
3 C71 Ol ^3 E E
^ 3 +j o a»  u
4> 

E
u
Of
a



^
r~>












































in
XI


O!
c

X!
C
0)
+J
XI
(0
C7-
C
'^
3

j_
O
if-



JD
in
C
o
OJ
L.
>
OJ
*J
TO


t-
ai
a.
•*-
in
O)
a.


a*
(U

o
i
0)

a>
TO
x
o
i-
d
Q.
TO*
                      C-49

-------
TABLE 036.   DATA ON VISIBLE EMISSIONS FROM OFFTAKE LEAKS,
                   U.S.  STEEL, CLAIRTON
Battery Date
1 8-7-79

8-8-79

8-9-79


Average
2 8-7-79

8-8-79

8-9-79

8-10-79
Average
3 8-7-79

8-8-79

8-9-79

8-10-79
Average
16 8-7-79
8-13-79

8-14-79

8-15-79

Average
17 8-7-79
8-13-79

8-14-79

8-15-79

Average
Number of
leaking offtakes
4
1
1
0
0
0
0

0
3
0
0
2
2
0

1
2
0
1
0
3
1

0
0
2
3
0
0
1

1
1
1
2
4
4
1

Percent
leaking offtakes
3.1
0.8
0.8
0
0
0
0
0.7
0
2.3
0
0
1.6
1.6
0
0.8
0.8
1.6
0
0.8
0
2.3
0.8
0.9
0
0
1.6
2.5
0
0
0.8
0.7
0.8
0.8
0.8
1.6
3.3
3.3
0.8
1.6
                                                       (continued)
                         C-50

-------
TABLE C-36 (continued)
Battery
7




8




9




10



11



19





20





21





Date
2-4-80
2-5-80
2-6-80
2-7-80
Average
2-4-80
2-5-80
2-6-80
2-7-80
Average
2-4-80
2-5-80
2-6-80
2-7-80
Average
2-6-80

2-7-80
Average
2-6-80

2-7-80
Average
9-16-81
9-16-81
9-17-81
9-17-81
9-18-81
Average
9-16-81
9-16-81
9-17-81
9-17-81
9-18-81
Average
9-16-81
9-16-81
9-17-81
9-17-81
9-18-81
Average
Number of
leaking offtakes
1
1
3
2

2
4
0
2

2
5
1
3

0
1
1

8
4
0

3
2
0
5
3

1
3
3
1
3

7
2
1
0
14

Percent
leaking offtakes
0.8
0.8
2.3
1.6
1.4
1.6
3.1
0
1.6
1.6
1.6
3.9
0.8
2.3
2.2
0
0.8
0.8
0.5
6.3
3.1
0
3.1
1.8
0.6
0.0
3.1
1.8
1.5
0.6
1.9
1.8
0.6
1.9
1.4
4.4
1.3
0.6
0.0
8.5
3.0
       C-51
                                     (continued)

-------
                          TABLE  C-36  (continued)
                                  Number  of                  Percent
Battery          Date          leaking offtakes           leaking offtakes

   22            9-16-81            6                         4.3
                 9-16-81            2                         1.2
                 9-17-81            3                         2.5
                 9-17-81            1                         0.6
                 9-18-81           14                         9.0
                 Average                                     3.5
                                  C-52

-------
        TABLE C-37.   DATA ON VISIBLE EMISSIONS FROM OFFTAKE LEAKS,
                       SHENANGO, INC.,  OCTOBER 1980
Battery
Date
   Number of
leaking offtakes
    Percent
leaking offtakes
                 10-20

                 10-21


                 10-22


                 Average

                 10-20

                 10-21


                 10-22


                 Average
                    2
                    2

                    0
                    0
                    2
                    0

                    1
                    2
                                4.3

                                2.9
                                2.9

                                0
                                0

                                2.0

                                0

                                2.9
                                0

                                1.4
                                2.9

                                1.4
                                 C-53

-------




s
o

lo-
co
LU
^
O2
•z.
o
H- 1
COrH
»-H CO
s: 
c
O)
u
C
0
<_>





<
S-
<"
^^
D)
^^





t-
-C
.a
•~




CO
E

Ol






CO
4^
*^-
^"^
O)



i)
S-
•3
•4
(
^_
j
i i
d)
Q-O
in E ^>
U) S- (
tO 0) r
O) 4-*
 cr>
•p r^^
10 CT>
Q i-l



N^y


 ^o r^»» rH rH ^3
iHrHCMrHOrHrHrH
OOOOOOOO
• •••••••
OOOOOOOO







ooooioooaicD
inir>cT>cocT>r~-cD«^-
o^ co co co r^» r^> co f^-
rH rH rH







OOOOOOOO
OOOOOOOO
OOOOOOOO
«t«s«t«s«t«k «tn
^•D *s^ ^\l l^" CO ^^ 00 f^
^^^^^^^^


^t IT) if) *$• if) if)
III III
0 O O 0 0 O
iH rH rH rH rH rH


0) CD
D) cn
id co
rHCMCO S-rHCMCO i-
0) CU

^C ^C



^J
•P 0)

C 3
>-• O




























































^>
C
a>
E
a)
^
in
10
a>
E

-o
S-
10
C
10
•p
in

>*)
S-
o
ro
C-54

-------
              TABLE C-39.   BaP EMISSION RATES FROM TESTS AT
                   ARMCO,  INCORPORATED, HOUSTON, TEXAS2
Sample
location
Inlet


Outlet


Sample
number
1
2
3
1
2
3
Date
(1979)
10-4
10-5
10-5
10-4
10-5
10-5
Concentration
lb/ftaa
5.7 (10~9)
6.5 (10~9)
5.2 (10~9)
3.6 (10~9)
1.2 (10~8)
1.8 (10~8)
mg/maa
0.092
0.103
0.084
0.057
0.191
0.294
Emission rate
Ib/hr
0.056
0.070
0.056
0.036
0.117
0.174
kg/hr
0.026
0.032
0.025
0.016
0.053
0.079
Dry standard measurement.
         TABLE C-40.  BENZENE CONCENTRATIONS AND EMISSION RATES
            FROM TESTS AT ARMCO, INCORPORATED, HOUSTON, TEXAS1
Sample
location
Inlet



Outlet



Sample
number
1
2
3
Average
1
2
3
Average
Concentration
(ppm)
0.7
2.1
2.2
1.7
0.7
2.9a
2.1
2.0
Emission
Ib/hr
1.6
4.4
4.6
3.5
1.4
5.3
3.7
3.5
rate
kg/hr
0.71
2.0
2.1
1.6
0.63
2.4
1.7
1.6
    aThis result was determined from a small air sample, possible
     leaky bag.
                                 C-55

-------






T-t
OO

0

gr
fV
«=t

I—

to
Z
O
1— 1
r-

-^>
QL
LU
00
CO
O
^
LU


cv*
O
O
1=1

U_
o

^~
a:

2:

"^
LO


.
rH

CJ

LU
	 1
CO
^f_
\—


















re
-P
0
4->

O

•p
c
CO
u s-
i- CO
<0 jQ
a. E
\^s 3
C
co
•p e
re 3
s- s-

c >
O X3
•r-
in /"•>
in in
•t- l-
E 0
O) O
-a
5-
o
o
•O

C
CO

0
CO

o
cj







at
C31
re
5-
a>
^

00
CJ



to
a.






^j-




to
o




to
CL.








CO



to
o




to
a.







CM





OO
f ^




OO
a.







r^





to
0




to
a.



c: s-
0 0 CO >,
r- -r- <4- S- S?
to in •!- a>
•P in co -p -P
O -i- -P C -P
t— E re co re
CO S- X5




H-
re o o>
i— cm
re t- -t- s-
•P CO ^ O
o xi re o
r- E CO -0
3 i—
C



t-
CU
>
S-
at
v)
^3
O





<1) S-
r*— OJ
ft ^*)
E E
re rj
oo c

00 ^t ^£ ^"C ^t CD *3I ^Z *^ ^j^ ^r - «tf- r^ r~-
co f*^ oo cn o oo cn cn co CT> «^ CM
rH CMrHrHrHrH rHrH



CD 1 1  1 l<£ 1 CTl | -CM rHini-HlO
KI-CMCMCO loootor^ income









rHCMCO rHCMCO rHCMCO




at co at
O) O) O)
re re re
t- $- t-
-P at Pat P co
in i— I > WCM> inco>
CO 
s-
co
in
o
O

s-
ai
Q.

in
c
3
S_


u
<4- 3
O T3
C
J- O
a> u
XI
e -P
3 0
C C

~a i/>
at re

•r*
ft ^^-
E
O C
(J 3
Q£
CO
. -P in
at in
•— CO CO
XJ 5- C
re re ^
co u s-
"O . -r- in re
•r- co •— J- -a
m "o Q. at
S_ -r- O- XJ 1-
ai m re E o
x: co 3
in ^ P c at
3 o o m
a. u c at 3
m re
II II II at u
x: at
to to 
-------
                  APPENDIX D



EMISSION MEASUREMENT AND CONTINUOUS MONITORING

-------
        APPENDIX D - EMISSION MEASUREMENT AND CONTINUOUS MONITORING

D.I  EMISSION MEASUREMENT METHODS
     The measurement of emissions resulting from coke oven charging opera-
tions, topside leaks, and door leaks presents a unique problem, due to the
unconfined nature of the emissions.   Neither mass measurement techniques
nor conventional techniques that measure only the maximum or average opacity
are considered suitable as a means of quantifying the emissions for the
purpose of establishing or enforcing a numerical emission standard.  There-
fore, beginning in 1973, a new measurement technique based upon determina-
tion of the total duration of visual emissions (for charging) or the percent-
age of potential sources with visual emissions (for topside leaks and door
leaks) was developed and tried by the Environmental Protection Agency (EPA)
and State and local agencies.
     Initial tests indicated that the method was repeatable and could be
used as a means for quantifying coke oven emissions.  Therefore, the method--
"Determination of Visible Emissions from Coke Oven Batteries"--was used to
obtain data for the development of an emission standard.  The method is
divided into three main parts:   Procedure for Determining Visible Emissions
from the Charging System During Charging, Procedure for Determining Visible
Emissions from Coke Oven Doors, and Procedure for Determining Visible
Emissions from Topside Ports and Offtake Systems.  All tests made by EPA to
obtain data adhered to the basic procedures of this method.   The method and
emission tests are discussed in more detail in the following paragraphs.
                                  D-3

-------
D.I.I  Charging Operation
     Part 3 of the Reference Method was developed to measure emissions from
larry car systems charging wet coal.   A discussion of the rationale behind
the development of the Reference Method follows.   Visible emissions from
larry car charging operations are not emitted from a single source or point
but from the drop sleeve area around each topside port,  the top of the
hoppers on the larry car, and the offtakes.   The emissions can be steady
but are usually intermittent or pulsating during good charges.  These
intermittent emissions result from momentary buildups of pressure which
occur as coal is charged to and leveled in the coke oven.  Except for these
pressure surges, the oven is under a slight vacuum during the charge.  Due
to the nature of the operation, procedures for directly measuring the mass
of particulate and gaseous emissions were not considered feasible.  The
approach of assigning an actual opacity value to the emissions according to
the procedures of Method 9 is potentially applicable and was considered;
however, due to the intermittent emissions which characterize coke oven
charging operations, this technique is not necessary to provide the quanti-
tative measure of control system effectiveness desired for a standard of
performance.  The technique selected involves recording the total time that
emissions are visible during a complete charge.   The charging period encom-
passes the period of time commencing when coal begins flowing into the oven
and ending when the last topside port lid is replaced.  Analysis of data
gathered by EPA demonstrates that this parameter, total  emissions duration,
is a representative measure of control effectiveness.  Other techniques,
such as recording the duration of emissions over a specified opacity value
or only recording emissions greater than 1 meter in plume length, were
                                  D-4

-------
considered.   The potential advantage of these techniques is that a distinc-
tion could be made between small and large quantitites of emissions that
are emitted in the same length of time,   bach ot these techniques proved
unsatisfactory or unnecessary for this source.   The relationship between
the opacity of the emission and the quantity of the emission is poor.
Determining the size or length of the emissions proved difficult due to the
quickness with which emissions momentarily explode from the ovens.  Since
it is not necessary to make these distinctions in order to measure control
effectiveness, and since not having to determine opacities or plume length
simplifies the measurement technique, total emission duration was chosen as
the parameter to measure effectiveness of control.
     In regard to the procedures of the Reference Method, two areas are
worth noting.   First, during the larry car charging operation it is some-
times standard operating procedure to reopen topside ports immediately
after charging (but prior to wet sealing of topside port lids) in order to
sweep spilled coal into the oven.  During the EPA study, the emissions
during this period were not monitored.  Therefore,  as defined in the Refer-
ence Method, the charging period for larry car systems does not include the
period of time during which lids are reopened for the purpose of sweeping
spilled coal into the oven.  The charging period ends when the last topside
port lid is put into place immediately after the coal charging is completed.
Secondly, during the EPA study, the observers also recorded the elapsed
time required for each oven to be charged; this information was desired for
the EPA study.  However, this information is not necessary to determine
compliance during a performance test; therefore, it is deleted in order to
simplify the test procedure.
                                  D-5

-------
D.I.2  Topside Leaks (Offtakes and stationary jumper pipes,  topside ports,



and collecting mains)



     Because of the numerous potential  topside leaks on a battery,  it would



be nearly impossible to assign opacities to or record the duration  of every



emission.  Therefore, the technique selected is to count the sources of



visible emissions.   This technique is relatively simple; the primary diffi-



culty is in defining a "leak."  Leaks from offtakes, topside ports, or



collecting mains can emit emissions of a large quantity or can emit only a



"wisp" of visible emissions (much like a smoking cigarette).  Furthermore,



since pressure in any specific portion of an oven or battery will  vary, the



leaks can be intermittent.   With these considerations in mind, the  technique



used involves a quick inspection (approximately 10 minutes per run) of all



the ovens on a single battery at one time.  During this inspection, the



observer notes and records the number of offtake systems or stationary



jumper pipes and the number of topside ports with any visible emission



(except for steam); the observer also notes and records all  visible emissions



(except for steam) from the collecting main(s).  As already stated, the



magnitude of the topside leaks can vary tremendously.  During the EPA



study, the question arose as to whether or not very small leaks ("wisps")



were "significant" emission sources.  Obviously there was a problem with



defining "significant."  Nonetheless, for several tests during the EPA



study the observers were instructed to differentiate between "total" and



"significant" leaks.  The first criterion used for determining whether or



not a leak was "significant" was plume color.  Brown/yellow leaks were



considered "significant", while white/gray leaks were not considered "signi-



ficant."  This procedure was deemed totally unacceptable because there did
                                  D-6

-------
not appear to be any correlation between plume color and the quantity of
emissions.   The second criterion examined for determining whether or not a
leak was "significant" was plume length.   Any visible emission (other than
steam) was recorded in order to determine the "total" leaks; any visible
emission (other than steam) with a visible plume length of more than 1
meter was recorded in order to determine "significant" leaks.   Although
this criterion and test procedure appeared to be feasible, the test proce-
dure ultimately chosen for the Reference Method makes no distinction between
"significant" and "total" leaks.  It requires only that all visible emissions
(except steam) be recorded.  This test procedure was chosen for two reasons.
First, a test method based on measuring total leaks is preferred because
use of this procedure is less complicated.  Secondly, measuring all visible
leaks provides the best indicator of the degree of emission control.  Note
that for the most part, the emission data collected during the tests con-
ducted for "total" and "significant" leaks are still valid since the proce-
dures used for determining "total" leaks during these tests were the same
as the procedures required in the Reference Method.
     During the EPA study, the question was raised as to what effect sun
angle has on observers' readings.  At one facility, it was noted that in
the early morning when the sun was low in the sky and behind the offtakes
(i.e., the offtakes were between the observer and the sun) the readings
were higher than in the afternoon hours with the sun at the observer's
back.  However, this difference cannot be solely attributed to the effect
of sun position.  Some of this variation is attributed to the difference in
luting procedures between shifts, including the difference attributable to
reduced visibility during the night shift (reduced visibility at night
would make it more difficult to determine leaks requiring luting).   Some of
                                  D-7

-------
this variation is also attributed to the observer location during the
inspection (catwalk rather than topside).   Although some of the difference
in emission levels is undoubtedly due to sunlight effects, the Reference
Method does not restrict the observer's position relative to the sun.   The
rationale supporting this approach is that data supporting the proposed
standard were collected under all conditions of observer position relative
to the sun.
D.I.3  Door Leaks (Oven and chuck doors)
     During the coking cycle, leaks in the seal between the oven door and
oven frame (door jamb), as well as leaks in the seal between the chuck door
and chuck door frame, can cause emissions to occur.  In addition, emissions
may occur from cracks in the refractory around the door jamb or buckstay.
Due to varying pressure within the coke oven, a leak from any one oven door
can intermittently occur, stopping and starting at any time during the full
coking cycle.  Each oven door on the battery can be a potential emission
                                                     t
source.  Due to the large number and variability of potential leaks, it
would be nearly impossible to identify the occurrence and duration of all
door area leaks on a battery at any given time and this is not the intent
of this method.  Instead, the method involves a successive inspection of
all the oven door areas on a single battery.  To perform the inspection,
the observer chooses to first inspect either the pusher side or coke side
of the battery.  The observer takes a position on the ground at one end of
the chosen side of the battery and then walks along the ground to the other
end of the coke battery looking at all the oven door areas and noting the
door areas that have visible emissions.  The door area for a given oven is
defined as the vertical face of the coke oven between the bench and top of
the battery and between two adjacent buckstays.  The inspection is completed
                                  D-8

-------
by observing the door areas on the other (coke or pusher) side of the
battery.  During the inspection, the number of each oven with a leaking
door area is recorded; an emission rate, percent door areas leaking, is
calculated based on the number of potential sources.
     During EPA studies of door emissions, opacity readings were also taken
in addition to the counting of door area leaks.  The approach of assigning
an actual opacity value to door emissions according to procedures similar
to Method 9 were considered for performance tests.  However, opacity readings
would unnecessarily complicate the test method and require certified ob-
servers, while the recommended technique of determining the percentage of
leaks provides an adequate quantitative measure for the demonstration of an
emission standard.
D.2  MONITORING SYSTEMS AND DEVICES
     Due to their unconfined nature, the emissions from oven charging,
topside leaks, and door area leaks are not suited to measurement by continu-
ous emission monitoring systems.
D.3  PERFORMANCE TEST METHOD
     Available data show Reference Method 109, "Determination of Visible
Emissions from Coke Oven Batteries," used by EPA during the emission mea-
surement program to be applicable to the performance testing.   Since the
majority of the tests conducted by EPA involved multiple observers record-
ing emissions simultaneously, these data provide a source of information on
the repeatability of this method.   Examination of the data indicates that
the variance caused by observer imprecision or bias between different
observers is small especially when compared to the variance caused by the
process, i.e., the variation between charges or number of leaks.   Therefore,
                                  D-9

-------
observer error (differences between observers  as  well  as  imprecision of a
given observer) is not a significant problem inherent  in  the application of
the Reference Test Method.
     While conducting a performance test for charging, 15-20 charges could
be observed in a 1-day period; this includes the  time  required on the first
day of testing for the observer to become familiar with the particular
facility being tested.  No special test equipment is necessary, and data
reduction is extremely simple.  However, special  safety precautions are
required because coke oven emissions are suspected carcinogens.  The pre-
cautions required can be found in the OSHA regulations pertaining to exposure
to coke oven workers (Federal Register, Vol. 41,  No. 206, Part III, Friday,
October 22, 1976).  The cost of conducting an emission test is estimated at
$500.00 to $1,000.00 per day (15-20 charges per day).   This cost estimate
is based on participation of two persons per day, one  person to monitor and
record process information and one person to determine emissions.
     The cost of a single emission test for topside leaks or door leaks is
estimated to be $500.00 to $1,000.00.  This cost estimate is based on a
single observer performing one run each for offtakes,  topside ports, and
collecting main emissions or a single observer performing one run each on
oven door areas.  A second person would monitor and record process operating
information.  In reality, the emission tests for topside leaks, door leaks,
and charging could be performed by the same person(s)  in a single trip.
Hence, all emission tests could be performed for a cost of $500 to $1,000
per day.  The same safety and exposure precautions briefly mentioned for
observing charging emissions also apply to performing emission tests for
topside and door area leaks.
                                  D-10

-------
     A new approach by the iron and steel industry for control of pushing
emissions directly affects the determination of by-product battery door
area leaks.   Large enclosures, usually called sheds, have been installed on
the coke side of several batteries which extend for the entire length of
the battery and beyond the outside track for the pusher car.  The design of
these sheds requires that they be as fully enclosed as possible in order to
obtain maximum capture efficiency for pushing emissions.   Typically, this
results in minimal clearances for operation of the pusher car and access
for maintenance.  It is extremely difficult for an observer to traverse for
door area leaks under shed primarily because of severe safety hazards due
to movement of the pusher car.  In addition, low light levels and visual
obstructions from ground level under the shed do not permit suitable observa-
tions for door leaks as specified in Method 109.
     The problem presented by shed installations was investigated by survey
door leak observation tests and two method development tests at three
different by-product coke plants.1 2 3 4  The survey trips were made to try
conducting door leak observations from the bench under the coke side sheds.
Light level  readings were made from the bench which indicated that sufficient
illumination is provided to permit bench level door leak inspections.
There are more stringent safety considerations involved with making observa-
tions from the bench so that persons assigned this duty must familiarize
themselves with any special safety rules in effect as required by company
or Occupational Safety and Health Administration regulations.   A principle
finding of the survey tests was a definite positive bias introduced by
observing for door area leaks from the bench.   Two method development tests
were conducted at two separate by-product plants to characterize and quantify
                                  D-ll

-------
this positive bias.   The plants selected for method development tests were
chosen by process and operational  performance similarity.   Controlled
experimental observation door area leak traverses were made on both push
and coke sides at five different batteries at the two plants.   These Method 109
tests were done with two pairs of experienced observers making simultaneous
traverses from the bench and from the yard.   Observation rotations were
used to minimize any statistical influence of individual observer bias or
observer location.  The method development tests confirmed a significant
positive bias due to application of Method 109 from bench level.  Evalua-
tion of the data from these tests provided a means of adjusting bench
observed leaks to a level reasonably equivalent to leak frequencies expected
from the standard ground level (yard) traverse specified in the method.
     It is recommended that observations for door area leaks be made from
the bench at by-product coke batteries equipped with sheds for pushing
emissions control.  The number of leaks observed during the bench traverse
can be corrected to a "yard traverse equivalence" using the following
equation.
                              Ey= Lb - 0.07N
where:
     E  = Equivalent yard leaks.
     L  = Number of leaks observed from bench.
      N = Number of ovens in battery.
     The number of ovens per battery is included to provide a proportional-
ity related to battery capacity.  Calculations of battery door area leak
performance for coke side shed installations is acheived by combining the
number of "equivalent yard leaks" on the coke side with the observed number
of push side leaks to determine the total number of leaks for the battery.

                                  D-12

-------
D.4  REFERENCES

1.    The EPA Report.  U.S. Steel, Clairton, PA.  October 1980.  Prepared by
     PEDCo Environmental under EPA Contract No. 68-02-3546, Task No. 5.

2.    The EPA Report.  Carondelet Coke, St. Louis, MO.  November 1980.
     Prepared by PEDCo Environmental under EPA Contract No. 68-02-3546.
     Task No. 5.

3.    The EPA Report.  Method Development 1.  U.S. Steel, Clairton,  PA.
     August 1981.  Prepared by PEDCo Environmental under EPA Contract No.
     68-02-3546, Task No. 5.

4.    The EPA Report.  Method Development 2.  U.S. Steel, Provo, UT.
     September 1981.  Prepared by PEDCo Environmental under EPA Contract
     No. 68-02-3546, Task No. 5.
                                  D-13

-------
              APPENDIX E

COKE OVEN EMISSIONS RISK ASSESSMENT FOR
  WET-COAL CHARGED COKE OVEN BATTERIES

-------
                  COKE OVEN EMISSIONS RISK ASSESSMENT FOR
                    WET-COAL CHARGED COKE OVEN BATTERIES

E.I  INTRODUCTION
E.I.I  Overview
     The quantitative expressions of public cancer risks presented in  this
appendix are based on (1) a dose-response model that numerically  relates
the degree of exposure to airborne coke oven emissions (expressed as benzene
soluble organics (BSO)) to the risk of getting respiratory cancer, and (2)  numerica
expressions of public exposure to ambient air concentrations of BSO estimated
to be caused by emissions from stationary sources.  Each of these factors
is discussed briefly below and details are provided in the following sections
of this appendix.
E.I.2  Relationship of Exposure to Cancer Risk
     The relationship of exposure to the risk of contracting respiratory cancer is
derived from epidemiological studies in occupational settings rather than
from studies of excess cancer incidence among the public.   The epidemiological
methods that have successfully revealed associations between occupational
exposure and cancer for substances such as asbestos, benzene, vinyl chloride,
and ionizing radiation, as well as for BSO, are not readily applied to the
public sector, with its increased number of confounding variables, much
more diverse and mobile exposed population, lack of consolidated  medical
records, and almost total absence of historical exposure data. Given  such
uncertainties, EPA considers it improbable that any association,  short of
very large increases in cancer, can be verified in the general population
with any reasonable certainty by an epidemiological study.  Furthermore, as
noted by the National Academy of Sciences (NAS)l, "...when there  is exposure
to a material, we are not starting at an origin of zero cancers.   Nor  are
we starting at an origin of zero carcinogenic agents in our environment.
Thus, it is likely that any carcinogenic agent added to the environment
will act by a particular mechanism on a particular cell population that is
already being acted on by the same mechanism to induce cancers."   In discuss-
ing experimental dose-response curves, the NAS observed that most information
                                    E-3

-------
on carcinogenesis is derived from studies  of  ionizing  radiation with experi-
mental animals and with humans  which  indicate  a  linear  non-threshold dose-
response relationship at low doses.  They  added  that although  some evidence
exists for thresholds in some animal  tissues,  by and large, thresholds  have
not been established for most tissues.   NAS concluded  that establishing
such low-dose thresholds "...would require massive, expensive, and  impractical
experiments..." and recognized that the U.S. population "...is a  large,
diverse, and genetically heterogeneous  group exposed to a  large variety of
toxic agents."  This fact, coupled with the known genetic  variability to
carcinogenesis and the predisposition of some  individuals  to some form  of
cancer, makes it extremely difficult, if not  impossible, to  identify a
threshold.
     For these reasons, EPA has taken the  position, shared by  other Federal
regulatory agencies, that in the absence of sound scientific evidence to
the contrary, carcinogens should be considered to pose some  cancer  risk at
any exposure level.  This non-threshold presumption is based on the view
that as little as one molecule of a carcinogenic substance may be sufficient
to transform a normal cell into a cancer cell.  Evidence is  available from
both the human and animal health literature that cancers may arise  from a
single transformed cell.  Mutation research with ionizing  radiation  in  cell
cultures indicates that such a transformation  can occur as the result of
interaction with as little as a single  cluster of ion  pairs.   In  reviewing
the available data regarding carcinogenicity,  EPA found no compelling
scientific reason to abandon the no-threshold  presumption  for  BSO.
      In developing the exposure-risk  relationship for  BSO, EPA has  assumed
that a linear no-threshold relationship exists at and  below  the  levels  of
exposure reported in the epidemiological studies of occupational  exposure.
This means that given the non-threshold presumption, the estimates  of cancer
risk to the general population can be derived from the occupational studies
even though exposures in the latter were much  higher.   The EPA believes
that this assumption is reasonable for  public  health protection  in  light
of presently available information.  Assuming that exposure  has been  accurately
quantified, it is the Agency's belief that the exposure-risk relationship
used by EPA at low concentrations represents  a plausible upper-limit  risk
                                    E-4

-------
estimate in the sense that the risk is probably not higher  than the  calculated
level and could be much lower.
     The numerical constant that defines the exposure-risk  relationship
used by EPA in its analysis of carcinogens is called the unit  risk estimate.
The unit risk estimate for an air pollutant is defined as the  lifetime
cancer risk occurring in a hypothetical population in which all individuals
are exposed throughout their lifetimes (about 70 years) to  an  average con-
centration of 1 ug/m3 Of the agent in the air which they breathe. Unit
risk estimates are used for two purposes:  (1) to compare the  carcinogenic
potency of several agents with each other, and (2) to give  a crude indication
of the public health risk which might be associated with estimated air
exposure to these agents.
     The unit risk estimate for BSD that is used in this appendix was based
on the risk of respiratory cancer only.  Although other tumor  sites  are
associated with BSD emissions, data are insufficient to derive unit  risk
factors.  The methodology used to derive the respiratory cancer risk factor
is described in E.2 below.
E.I .3  Public Exposure
     The unit risk estimate is one of several factors necessary to derive
quantitative expressions of public health risks.  Another factor needed
is a measure of the extent of human exposure, i.e., the numbers of
people exposed to the various concentrations of BSO.  The difficulty of
defining public exposure was noted by the National Task Force  on Environ-
mental Cancer and Health and Lung Disease in their 5th Annual  Report to
Congress, in 1982,2  They reported that "...a large proportion of the
American population works some distance away from their homes  and experiences
different types of pollution in their homes, on the way to  and from  work,
and  in the workplace.  Also, the American population is quite  mobile, and
many people move every few years."  They also noted the necessity and
difficulty of dealing with long-term exposures because of "...the long
latent period required for the development and expression of neoplasia
[cancer]..."
     The EPA's numerical expression of public exposure is based on two
estimates.  The first is an estimate of the magnitude and location of
                                    E-5

-------
long-term average ambient air concentrations  of BSO  in the vicinity of
emitting sources based on air dispersion  modeling using  long-term estimates
of source emissions and meteorological  conditions.   The  second  is an estimate
of the number and distribution of people  living in the vicinity of emitting
sources based on 1980 Bureau of Census  data which  "locates" people by
population centroids in census tract areas.   The people  and concentrations
are combined to produce numerical expressions of public  exposure by an
approximating technique contained in a  computerized  model.  The methodology
is described in E.3 below.
E.I.4  Public Cancer Risks
       By combining numerical expressions of  public  exposure with the unit
risk estimate, two types of numerical expressions of public cancer risks  are
produced.  The first, maximum individual  risk, relates to the person or persons
estimated to live in the area of the highest, long-term  concentration as  esti-
mated by the computer model. Maximum individual risk is  expressed as  "maximum
lifetime risk."  As used here, the work "maximum"  does not mean the greatest
possible risk of cancer to the public.   It is based  only on the maximum
annual average exposure estimated by the  procedure used. The second measure,
aggregate risk, is a summation of individual  risks across the exposed
popualtion  (generally limited to the population within 50 kilometers of
the emmitting sources).  The aggregate  risk  is expressed as the  incidence of
cancer (number of cases) expected in the  exposed population as  result of  70
years; for convenience, it is often divided  by 70  and expressed as expected
annual cancer incidences.  These calculations are  described in  more detail
in E.4 below.
     There are also risks of nonfatal cancer and other potential health
effects, depending on which organs receive the exposure. No  numerical
expressions of such risks have been developed; however,  EPA considers all
of these risks when making regulatory decisions on  limiting emissions of
BSO.
E.2  UNIT RISK ESTIMATE 3
     The following discussion is summarized  from  a more  detailed description
of the Agency's derivation of the inorganic  arsenic  unit risk estimate  as
found in EPA's "Carcinogen Assessment of  Coke Oven Emissions"  (EPA-600/
6-82-003F).

                                    E-6

-------
     Several models have been used to relate cumulative lag-time-adjusted
exposure from coke oven emissions to observed increases in the human
respiratory cancer death rate.  The parameters in the models were estimated
using approximate exposure variables and mortality data generated by Lloyd,
Redmond, and Mazumdar, and reworked by Land.3  The resulting explicit models
were in turn used to estimate the lifetime respiratory cancer death risk
from a continuous exposure to 1 ug/m3 of coal tar pitch volatiles from coke
oven emissions.
     Using the Weibull model, these estimates ranged from 1.30 x 10~8 for
the 95 percent lower-bound, zero lag-time, and the nonwhite male background
death rate assumption, to 1.05 x 10~3 for the 95 percent upper-bound, 15-year
lag-time, and total background death rate assumption.  However, these limits
do not allow for the possibility that the Weibull model does not describe
the true dose-response relationship at low exposure levels.  As a result,
even this range of 5 orders of magnitude does not fully describe the un-
certainty associated with the point estimate.
     To allow for possible model differences, the multistage model  is also
employed to estimate risk.  This model has the advantage of simultaneously
taking into account both parameter estimates and model difference variability,
The 95 percent upper limit of this model is based upon finding a polynomial
with the largest linear term that is still consistent with the observed
data.  Using this model, point estimates were obtained for the lifetime
risk due to a constant 1 ug/m3 exposure ranging from 1.76 x 10'6 for the zero
lagtime case to 6.29 x 10~* for the 15-year lag-time case.  The 95 percent
upper bounds corresponding to these extremes range from 2.57 x 10~^ to
1.26 x 10~3.  The geometric mean of the 95 percent upper bounds for the four
lag times is calculated to be 6.17 x 10"*, which is taken to be the composite
unit risk estimate.
     Numerous additional uncertainties exist concerning these estimates.
The effects of age, sex, race, general health, and cigarette smoking on the
sensitivity of responses to coke oven emissions are unknown.  The way in
which the data were collected and summarized by the researchers could also
introduce additional biases and uncertainties.  However, because of the
unavailability of sufficient data to correct for these factors, the impact
of these factors cannot be addressed in this assessment.
                                    E-7

-------
E.3  QUANTITATIVE EXPRESSIONS OF  PUBLIC  EXPOSURE
E.3.1  EPA's Human Exposure Model  (HEM)  (General)
     The EPA's Human Exposure Model  is a general model  capable of producing
quantitative expressions of public exposure to  ambient  air  concentrations  of
pollutants emitted from stationary sources.  HEM contains  (1) an atmospheric
dispersion model, with included meteorological  data,  and  (2) a population
distribution estimate based on Bureau  of Census data.  The  input data  needed
to operate this model are source data, e.g., plant location, height  of the
emission release point, volumetric rate  of release, and temperature  of the
off-gases.  Based on the source data,  the model estimates  the magnitude and
distribution of ambient air concentrations of the  pollutant in  the  vicinity
of the source.  The model is programmed  to estimate these  concentrations
for a specific set of points within a  radial distance of  50 kilometers from
the source.  If the user wishes to use a dispersion model  other than the
one contained in HEM to estimate ambient air concentrations in  the  vicinity
of a source, HEM can accept the concentrations  if  they  are put  into  an
appropriate format.
     Based on the radial distance specified, HEM numerically combines  the
distributions of pollutant concentrations and people to produce quantitative
expressions of public exposure to the  pollutant.
E.3.1.1  Pollutant Concentrations Near a Source
     The HEM dispersion model is a climatological  model which  is a  sector-
averaged gaussian dispersion algorithm that has been simplified to  improve
computational efficiency.^  Stability  array (STAR) summaries are the
principal meteorological input to the  HEM dispersion model. STAR  data are
standard climatological frequency-of-occurence summaries  formulated  for use
in EPA models and available for major  U.S. meteorological  monitoring sites
from the National Climatic Center, Asheville, N.C.  A STAR summary  is  a
joint frequency-of-occurence of wind speed, atmospheric stability,  and wind
direction, classified according to Pasquill's categories.   The  STAR  summaries
in HEM usually reflect five years of meteorological data  for each  of 314
sites nationwide.  The model produces  polar coordinate  receptor  grid points
consisting of 10 downwind distances located along  each  of  16 radials which
represent wind directions.  Concentrations are estimated  by the  dispersion
model for each of the 160 receptors located on this grid.  The  radials are

                                    E-8

-------
separated by 22.5-degree intervals beginning with 0.0 degrees  and proceeding
clockwise to 337.5 degrees.  The 10 downwind distances for  each  radial  are
0.2, 0.5, 1.0, 2.0, 5.0, 10.0, 20.0, 30.0, 40.0, and 50.0 kilometers.   The
center of the receptor grid for each plant is assumed to  be the  plant center.
Concentrations at other points were calculated by using a log-linear scheme
as illustrated in Figure E-l.
E.3.1.2  Methodology for Reviewing Pollutant Concentrations
     Before making HEM computer runs, EPA reviewed small-scale U.S. Geological
Survey topographical maps (scale 1:24,000) to verify location  data for  each
BSO source.  Plants were given accurate latitude and longitude values which
were then incorporated into the HEM program.
     The HEM results were compared to the limited data available for ambient
BSO concentrations from monitoring sites near coke ovens.5  The ambient
concentrations are difficult to relate to predicted concentrations because
of the many other sources of BSO and its nonspecific nature.  However,  the
available BSO data and the HEM results were examined to determine if the
predicted increases in BSO concentrations were reasonable when compared to
the differences in ambient measurements upwind and downwind of coke plants.
     The data from Wayne County (Detroit), Michigan are summarized  in
Table E-l.  Meteorological data for Detroit were examined and  indicated
that the prevailing wind is from the west.  (In February  and March the  wind
shifts to the WNW, and during July to September it is generally  WSW.)   Con-
sequently, the results for monitoring sites east of the plants are  listed
in Table E.I as the downwind concentrations.  The BSO data  in  Table E-l
show clearly that as the distance from the coke plants increases BSO con-
centrations decrease.  Average downwind concentrations (east of  the plants)
range from 1.5 to 4 ug/m3 greater than the upwind concentrations (west  of
the plants).  Assuming no other significant BSO sources in  the area of  the
coke plants, the data indicate that the coke oven emissions could add over
1 ug/m3 to ambient BSO concentrations.
     Dispersion model predictions for Rouge Steel (subsidiary  of Ford Motor
Company) and National Steel, which represent 83 percent of  the coke capacity
in Wayne County, are shown in Table E-2.  (Detroit coke was not  modeled
because it is a dry coal battery not covered by the recommended  NESHAP.)
                                    E-9

-------
Concentrations are presented for  the areas  east of  the plants where the
model predicted that the maximum  concentrations would occur.  The modeling
predicts that the combined effects  of the two  major plants, if they were
located on the same site, would increase ambient concentrations by 0.16 to
1.7 ug/m3 at 1 km, 0.01 to 0.13 ug/m3 at 5  km, and  0.004 to 0.043 ug/m3 at  10
km.  The Wayne County data showed increases of 1.5  to  4 ug/m3 when comparing
downwind to upwind ambient measurements.  The  comparison of measured  differences
to the model's predicted differences indicates that the dispersion modeling
results are probably not overstated, and may be slightly low for the  lower
end of the range.
     During 1978-1980, EPA Region 5 conducted  ambient  monitoring around 4
steel plants.  Although the primary emphasis was on total particulate
matter, BaP data were collected for all four sites  and BSO data were  obtained
for one plant.*>  The BaP data for the four  plants were collected at distances
of 0.8 km (0.5 miles) to 3.2 km (2 miles) from the  plants.  Samples collected
downwind from the plant all showed BaP levels  from  12  to 20 times higher than
background BaP levels.  At the U.S. Steel plant  (Lorain, Ohio), 16 out of
68 samples were  identified as distinctly upwind or  downwind.  These samples
were collected 1.5 to 3 km (1-2 miles) from the Lorain plant3:

                                 BaP (ug/m3)          BSO  (ug/m3)

              Downwind mean          12.6                  7.1
              Upwind Mean             0.6                  3.2
              Increase               12.0                  3.9

     The NESHAP  dispersion model  predictions for  U.S.  Steel  at  Lorain
(existing control level) are compared to the measured  increases below:

                 Model Prediction                        Measured  Increase
Distance (km)      BSO  (ug/m3)        Distance (km)         BSO  (ug/m3)

     1              0.44 - 4.6
     2              0.14 - 1.5           1.5 - 3                3.9
     5             0.033 to 0.34
                                    E-10

-------
Figure E.I   Group 2 BG/ED Interpolation
                                                      — A
                               R           R2
Given:
A   -  The angle in radians subtended clockwise about  the  source  from due
       south to the BG/ED centroid;
Al  -  The angle from due south to the radial  line immediately  counter-
       clockwise of A, or passing through A if there is  an exact  match;
A2  -  The angle from due south to the radial  line immediately  clockwise of
       Al (A2 is 0 if it is due south);
R   -  The distance in km from the source to the BG/ED centroid;
Rl  -  The distance from the source  to the largest circular arc of  radius
       less than R;
R2  -  The distance from the source  to the smallest circular arc  of
       radius greater than or equal  to R;
Cl  -  The natural logarithm of the  concentration value  at (Al, Rl);
C2  -  The natural logarithm of the  concentration value  at (Al, R2);
                                    E-ll

-------
C3  -  The natural  logarithm of  the  concentration value at (A2, Rl);



C4  -  The natural  logarithm of  the  concentration value at (A2, R2);



then:



RTEMP - ln(R/Rl)/ln(R2/Rl);



ATEMP - (A-A1)/(A2-A1);



CA1   - exp(Cl + (C2-Cl)xRTEMP);



CA2   - exp(C3 + (C4-C3)xRTEMP); and



CX    - CA1 + (CA2-CAl)xATEMP,



where CX is the interpolated concentration  at  the BG/EO centroid.
                                    E-12

-------







co
UJ
>
O
UJ
x^
O
o
>-
r-
Z
O
O

UJ
z
>-
<
3E
u.
o
>-
r-
»— 4
Z
»-^
0
t— t
>
LU
-^
_i_
h-

r-H
•a:

CD C
0)
o
c
o
o

















o
4->
(U
U
c

>— i





o
CT>
i— 4



CT>
31
r-4





00
^
r- 1






^i
^— "
in
4->
c
fl3
CL.
0)
^^
O
LJ





O
4->
C
ie
a.



OCOCTilOCMOOVOCO
rHCMr^CTlCMCOCMLOi-H
ocOLO^J-r^'a-«a-ro«a-



ocMr^i-ioo'v-LOo^'"
^•LOi-HOO*T'*r-4LOO
OOLO«*COLOCOCOCMCO





CT> VO LO t»^ CM CM
00 1 LO CM LO 1 1 CTi CTi
• • • • • •
VO ^ LO LO CO CO




CMLOOCTt'frCOCTiVOCTi
t— iCTi^CMCTiOr-.CT>LO
CTir-»vo*S-voio'*ro^-




oor>-cocooocv)vocMCO
LOvocor>.LOr~-.r^LOCo
OOr»«tOcO**'
i— 1 .-4





r>.VOODCTiLOCTiOOvO':'"
vor~-*TvOi— 4CTiooovo
LOOOVOLOCTiLO^CO^t
1—4








CO Qj d)
• VO OJ O VO VO CTi
LO CO i— 1 (O r-l t— 1 CM IO
00 t_ t_
• 1 1 1  >
«tt^-«*•>•> III 111 111
CM CM CM Z Z Z
• • • o o o
«— • <— i •— < i—i z: z z
c

JT
4->
•r-
2
XJ
OJ
4->
IO
•r-
0
o
trt
(/>
(O

CO
f
W-
fO

4->
c
*o

Q.
vl)
JZ
4->
V|-
o
4_>
(/)
(O
0)
trt
0
4->
•r~
c
g

*
O)
o
l[
QJ
t_
0)
JZ
4->

• *«
4->
(/)
vU

O)
JZ
4->
Q
c_
M-

•r—
•o
c

$

CTt •
c in
•r- C
•— o
•r- -r™
IO I/)
> l/>
d) •!-
(- E
CL O)
(O




























0>
\s
o
o
4->
•1—
0
t_
4->
0)
Q
II

co
• *1
D
O)
•«->
co

,_
IO
c
0
•r"
4^
IO
II
CM
» M
'a;
fll
Vi/
4J
co

a>
o>
3
o
en
II
1— 1
JD




























•
J_)
*•*
c
IO
Q.
a>
s.
4->
l|_
O
40
CO
0)
2

•f—
c
i
ai
^
i <
4->
IO
JZ
4-)


-------
           TABLE E.2.  DISPERSION  MODEL  PREDICTIONS OF BSO FOR
                            WAYNE  COUNTY COKE BATTERIES
                   0.5 km        1  km          5  km            10  km


A.  Rouge Steel

BSO (ug/m3)     0.21-2.2     0.066-0.69    0.0049-0.051     0.0017-0.017




B.  National Steel

BSO (ug/m3)     0.32-2.7     0.098-1.0    0.0073-0.076     0.0025-0.026



C.  Combined13

BSO (ug/m3)     0.53-4.9     0.16-1.7     0.012-0.13     0.0042-0.043
a Results for existing baseline controls.   Ranges  are  derived  from the
  ranges in the emission estimates.

b Assumes worst case with both plants at the same  location.
                                 E-14

-------
These results again indicate that the range of estimates  from the  dispersion
model does not overstate the increase in ambient concentrations  when  compared
to actual measured increases from ambient data.
     The comparisons made in this brief review are not refined and should
be reviewed with caution because of the numerous assumptions  required.
However, the ambient data and emission test data show that coke  ovens are
an identifiable source of BSD emissions and that coke ovens contribute  to
ambient 8SO concentrations.  The ambient data also appear to  indicate that
the range of baseline dispersion modeling predictions for the increase  in
BSO levels is reasonable when compared with actual measured increases in
BSO concentrations from coke ovens.5
E.3.3  Site-Specific Modeling
       In its original risk assessment, EPA did not consider  terrain  effects
or the effects of buoyancy of the fugitive emissions escaping from the  coke
oven batteries.  At the same time, since the emissions are warmer  than
ambient air, they tend to rise.  Consequently, after experiencing  downwash
initially, the plume may lift-off, thus lowering ground-level concentrations
downwind.  However, this effect can be offset in the presence of rising
terrain.
     Since the combined effect of terrain, downwash, and  buoyancy  on  air-
borne BSO concentrations was unclear, additional dispersion analyses
were carried out for two facilities.  The coke ovens examined were U.S. Steel
at Clairton, Pennsylvania, and LTV Steel at Chicago, Illinois.  U.S.  Steel-
Clairton is intended to be representative of a facility located  in a  river
valley while LTV Steel-Chicago is intended to be representative  of one
located in flat terrain.7
     The analysis at Clairton was highly site-specific.  On-site meteoro-
logical data were used.  Each coke oven battery was represented  individually
based on a facility site plan.  Plume rise from the coke  oven batteries was
considered explicitly.  The calculations employed a buoyant line source
treatment to account for natural convection about the ovens.   The  effect of
downwash was also included.  Actual ground elevations in  the  Clairton area
were input to the LON6Z model to account for the effects  of terrain.
     At Chicago, detailed source information was not available.  For  this
reason, all emissions were assumed to be collocated.  However, the effects

                                   E-15

-------
of plume rise and downwash were both  considered  in  the analysis using the
treatment of coke oven batteries at Clairton  as  a prototype.  National
Weather Service meteorological  observations from Chicago Midway Airport
were processed and input to the Industrial Source Complex Long Term-Urban
(ISCLTU) model.  Review of the  HEM inputs  for these two facilities  indicates
that no plume rise should have  been calculated by the program.  Also, HEM
selected meteorological data from Greater  Pittsburgh Airport  for  U.S. Steel-
Clairton and from Chicago O'Hare airport for  LTV Steel-Chicago.
     A comparison of the LON6Z  and ISCLT estimates  with those from  HEM  is
shown graphically in Figures E.2 and  E.3.  These figures show normalized
concentration estimates as a function of distance along a radial  line to
the north.  The north radial was selected  because at both facilities maximum
concentrations beyond a kilometer occur in the north sector.  Normalized
concentrations are plotted to avoid confounding  the comparison where
differenct emission rates had been used in the various analyses.  Normalized
here means that the concentrations (ug/m3) were  divided by  the emission rate
(Kg) to yield a normalized concentration (nr3) as to emission rate.
     The LONGZ results show the substantial effect  that terrain can have on
ambient concentrations.  Between a half and one  kilometer,  concentrations
remain  level as rising terrain  intercepts  the coke  oven plumes.   Between
three and five kilometers the rate of decrease in concentration  increases
as dispersion changes from a predominately rural regime to  a  predominately
urban one.  Beyond about 20 kilometers concentrations  decrease  less rapidly
as dispersion becomes limited by the  depth of the mixing  layer.   At all
distances, the LONGZ estimates  are lower than HEM.
     The ISCLTU results show a  monotonic decrease with distance as  expected
in the  absence of terrain.  Again, beyond  about  20  kilometers concentrations
decrease less rapidly due to the limitation  in vertical mixing.   At all
distances, the ISCLTU estimates are lower  than HEM.
     Also of interest is a comparison of the  maximum estimated concentrations
At U.S. Steel-Clairton, LONGZ predicts a maximum normalized concentration
of 2.5  x lO'^m"3.  By comparison, the maximum normalized  concentration  from
HEM is  12.5 x 10-6nr3, a factor of 5  higher.   At LTV Steel-Chicago, ISCLTU
predicts 3.75 x lQ-6nr3.  This  compares to an HEM  value of  10.8  x 10~6m-3,
a factor of three higher.

                                   E-16

-------
                    Figure E.2  Comparison between Normalized
                                Concentrations Predicted by
                                HEM  and LONGZ
       10.0 -_.!;^;
UD

 O
CO
 I
 O
 
-------
                Figure E.3  Comparison  between  Normalized

                            Concentrations  Predicted by
                            HEM  and  ISCLT
(JO
 I
 o
n
 i
 CO

 i.
 Ol
 u

 o
 o

 •o
 O)
 Nl
       10.0--.
        1.0-:
0.1-;:
       0.01-
      0.001
                             10.0
                                       1.0
0.1
                                  Distance  (km)




                                    E-18

-------
     In summary, this evaluation shows  that the  HEM dispersion  algorithm
provides conservative concentration estimates  for  coke  oven  batteries when
applied in this matter.  Maximum concentrations  from HEM  are as much as a
factor of five higher than more detailed analyses  indicate.   As previously
discussed, however, measured ambient concentrations of  BSD are  higher than
the HEM predictions.  Although BSD may  originate from sources other than
coke ovens, these measured data indicate that  the  HEM predictions  are not
unreasonable.
     At the remaining coke ovens where  site-specific air  dispersion
analysis was not performed, the standard  analysis (HEM)  as  described in
section E.3.1 was used.  Comparison of  concentration profiles that were
predicted by HEM, LONGZ and the ISCLT models and the comparison of modeling
results to measured ambient concentrations indicate that  the standard HEM
analysis produces similar (although somewhat higher) results to the
sophisticated air dispersion models.  Since site-specific analysis is
resource intensive and was not producing significantly  different results
from the standard analysis, acceptable  risk estimates for the remaining
smelters were produced by the HEM analysis.
E.3.4  Population Near A Source
       To estimate the number and distribution of  people  residing within 50
kilometers of the coke oven, the HEM model uses  the 1980  Master Area Reference
File (MARF) from the U.S. Bureau of Census. This  data  base  consists of
enumeration district/block group (ED/BG) values.  MARF  contains the population
centroid coordinates (latitude and longitude)  and  the 1980 population of
each ED/BG (approximately 300,000 EO/BG) in the  United  States (50 States plus
the District of Columbia).  The HEM identifies the population around each
plant, by using the geographical coordinates of  the plant, and  identifies,
selects, and stores for later use those ED/BGs with coordinates falling
within 50 kilometers of plant center.
E. 3.5  Exposure^
       The Human Exposure Model (HEM) uses the estimated  ground level
concentrations of a pollutant together  with population  data  to  calculate
public exposure.  For each of 160 receptors located around a plant, the
concentration of the pollutant and the  number  of people estimated by the
HEM to be exposed to that particular concentration are  identified. The HEM

                                    E-19

-------
multiplies these two numbers to  produce  exposure estimates and sums these
products for each plant.
      A two-level scheme has been adopted in  order  to pair concentrations
and populations prior to the computation of exposure.  The two-level approach
is used because the concentrations are defined  on a radius-azimuth  (polar)
grid pattern with non-uniform spacing.  At small radii, the  grid cells are
usually smaller than ED/BG's; at large radii, the grid cells are usually
larger than ED/BG's.  The area surrounding the  source  is  divided into two
regions, and each ED/BG is classified by the  region in which its centroid
lies.  Population exposure is calculated differently for  the ED/BG's located
within each region.  For ED/BG centroids located between  0.2 km and 3.5 km
from the emission source, populations are divided between neighboring con-
centration grid points.  There are 64 (4 x 16)  polar griti points within
this range.  Each grid point has a polar sector defined by two concentric
arcs and two wind direction radials.  Each of these grid  points and respec-
tive concentrations are assigned to the  nearest ED/BG centroid identified
from 1980 U.S. Census Bureau data.  Each ED/BG  can  be paired with one or
many concentration points.  The population associated with the ED/BG cen-
troid is then divided among all  concentration grid  points assigned  to it.
The land area within each polar sector is considered in the  apportionment.
     For population centroids between 3.5 km  and  50 km from  the source,
a concentration grid cell, the area approximating a rectangular shape
bounded by four receptors, is much larger than  the  area of a typical ED/BG.
Since there is an approximate linear relationship between the  logarithm of
concentration and the logarithm of distance for receptors more than 2 km
from the source, the entire population of the ED/BG is assumed to be exposed
to the concentration that is logarithmically  interpolated radially  and
arithmetically interpolated azimuthally  from  the  four  receptors bounding the
grid cell.  Concentration estimates for  96 (6 x 16) grid  cell  receptors at
5.0, 10.0, 20.0, 30.0, 40.0, and 50.0 km from the source  along each of 16
wind directions are used as reference points  for  this  interpolation.
     In summary, two approaches are used to arrive  at  coincident
concentration/population data points.  For the  64 concentration points
within 3.5 km of the source, the pairing occurs at  the polar grid points
using an apportionment of ED/BG population by land  area.  For  the remaining

                                    E-20

-------
portions of the grid, pairing occurs  at the ED/BG  centroids themselves
through the use of log-log and linear interpolation.   (For a more detailed
discussion of the model used to estimate exposure, see Reference  4.)
E.3.6  Public Exposure to Coke Oven Emissions  (BSO)
E.3.6.1  Source Data
     Forty-three coke ovens are included in the analysis.  Table  E-3  lists
the model parameters used as input to the Human Exposure Model  (HEM)  for
all of the plants.  Table E.4 lists the specific plant names and  locations.
Table E.5 lists the emission rates used for baseline  and BAT scenarios.
E.3.6.2  Exposure Data
     Table E.6 lists, on a plant-by-plant basis, the  total number of  people
encompassed by the exposure analysis  and the total exposure.  Total exposure
is the sum of the products of number of people times  the ambient  air  concen-
tration to which they are exposed, as calculated by HEM.  Table E.7 sums,
for the entire source category (43 plants), the numbers of people exposed
to various ambient concentrations, as calculated by HEM.   (Source-by-source
exposure results are provided in the EPA docket numbered A-83-33.)
E.4  QUANTITATIVE ESTIMATES OF PUBLIC CANCER RISKS FROM BSO
E.4.1  Methodology (General)
E.4.1.1  Two Basic Types of Risk
     Two basic types of risk are dealt with in the analysis.  "Aggregate
risk" applies to all of the people encompassed by  the particular  analysis.
Aggregate risk can be related to a single source,  to  all of the sources in
a source category, or to all of the source categories analyzed.  Aggregate
risk is expressed as incidences of cancer among all of the people included
in the analysis, after 70 years of exposure.  For  statistical convenience,
it is often divided by 70 and expressed as cancer  incidences per year.
"Individual risk" applies to the person or persons estimated to live  in the
area of the highest ambient air concentrations and it applies to  the  single
source associated with this estimate as estimated  by  the dispersion model.
Individual risk is expressed as "maximum lifetime  risk" and is  defined as
the probability of contracting cancer if continuously exposed to  the  estimated
maximum ambient air concentration for 70 years.
                                    E-21

-------
TABLE E.3.  MODEL PLANT PARAMETERS INPUT DATA FOR EACH COKE OVEN*
           Stack Height                   5 meters
           Area of Release               10 square meters
           Stack Diameter                 0 meters
           Gas Exit Velocity            0.1 meters/second
           Gas Exit Temperature         298 degrees Kelvin
  a These were model parameters which were used for all forty-three
    coke ovens.
                               E-22

-------
           TABLE E.4.  WET-COAL  CHARGED COKE PLANTS AND LOCATIONS

PLANT AND LOCATION                          LATITUDE             LONGITUDE
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11
12

13.
Alabama By-Products, Tarrant,  AL
Enpire Coke, Holt, AL
Jim Walter, Birmingham, AL
Koppers Company, Woodward, AL
LTV Steel, Gadsden, AL
LTV Steel, Thomas, AL
U.S. Steel, Fairfield, AL
Interlake, St. Chicago, IL
National Steel, Granite City,  IL
LTV Steel, S. Chicago, IL
Bethlehem Steel, Burns Harbor, IN
Citizens Gas & Coke Util.
 Indianapolis, IN
Indiana Gas 4 Chemical
 Terre Haute, IN
Inland Steel, E. Chicago,
                              IN
                           IN
14
15. LTV Steel, E. Chicago,
16. U.S. Steel, Gary, IN
17  Armco Inc., Ashland, KY
18. Bethlehem Steel, Sparrows  Point,  MD
19. Rouge Steel, Dearborn, MI
20. National Steel, Detroit,  MI
21. Carondolet Corporation,  St.  Louis, MO
22. Tonawanda Coke Co., Buffalo
23. Bethlehem Steel, Lackawanna
24. Armco Steel, Middletown,  OH
25. Koppers Company, Toledo,  OH
26. New Boston Coke, Portsmouth
27. LTV Steel, Cleveland, OH
28. LTV Steel, Warren, OH
29. U.S. Steel, Lorain, OH
30. Bethlehem Steel, Bethlehem,
31. LTV Steel, Aliquippa, PA
32. LTV Steel, Pittsburgn, PH
33. Koppers Company, Erie, PA
34. Shenango, Inc., Pittsburgh,
35. U.S. Steel, Clairton, PA
36. U.S. Steel, Fairless Hills,  PA
37. Wheeling-Pittsburgh, Monessen,
38. Chattanooga Coke & Chem.
     Chattanooga, TN
39. Lone Star Steel, Lone Star,  TX
40. U.S. Steel, Provo, UT
41. Weirton Steel, Browns Island,  WV
42. Wheeling-Pittsburgh
     E. Steubenville, WV
43. Alabama By-Products
     Keystone, PA
                                 NY
                                 NY
                                 OH
                                PA
                                PA
                                   PA
33°34'57"
33°14'25"
33°33'40"
33°26'13'
34°00'46"
33°32'47"
33°29'22"
41°39'22"
38°41'40"
41°41'29"
41"37'41"
39°45'16"

39°26'48"

41 "37'53"
41°39'48"
41°36'55"
38°30'07"
39°13'10"
42°18'19"
42°15'16"
38e>32'08"
42°58'56"
42°49'20"
39°29'45"
41°40'10"
38°44'57"
41°28'26"
41°13'13"
41°26'56"
40°36'51"
40°37'16"
40°25I34"
42°08'43"
40°28'49"
40°18'04"
40°09'28"
40°09'46"
35°02'16"

32°54'59"
40°18'43"
40°24'58"
40°20'36"

40°05'12"
                                                                 86° 46 '47"
                                                                 87°30'11"
                                                                 86° 48 '38"
                                                                 86°57'50"
                                                                 86°02'38"
                                                                 86°50'13"
                                                                 86°55'32"
                                                                 87°37'32"
                                                                 90°07'42"
                                                                 87°32'50"
                                                                 87°10'20"
                                                                 86°06'49"

                                                                 87°23'47"
87°27
87°26
87°20
82°40
76°29
83°09
83°07
90°16
78°56
'15"
'42"
'03"
'08"
'19"
'40"
'43"
'05"
'19"
                                                              84°23'15"
                                                              83°29'31"
                                                              82°56'01"
                                                              81°39'55"
                                                              80°48'40"
                                                              82°07'50"
                                                              75021'13"
                                                              80°14'24"
                                                              79°57'47"
                                                              80°01'32"
                                                              80°03'34"
                                                              79°52'21"
                                                              74°44'32"
                                                              79°53'47"
                                                                  94°42'57
                                                                  80°35'16"
                                                                  80°36'25"

                                                                  75°18'59"
                                    E-23

-------
TABLE E.5.  COKE LOCATIONS AND EMISSION RATE FOR
             BASELINE AND BAT SCENARIOS

PLANT
BASELINE BAT
EMISSIONS EMISSIONS
(KG/YR) (KC/YR)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
16
19
20
21
22
23
24
25
26
27
26
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
AL Byproducts, Tarrant, AL
Empire Coke, Holt, AL
Koppers, Woodward, AL
LTV Steel, Gadsden.AL
LTV Steel, Thomas, AL
Jim Walters, Birmingham. AL
U.S. Steel. Fairfield, AL
National Steel, Granite City, IL
Interlake, Chicago, IL
LTV Steel. So. Chicago, IL
Bethlehem Stieel, Burns Harbor. IN
Citisens Gas, Indianapolis, IN
IN Gas, Terre Haute. IN
Inland Steel, E. Chicago, IN
U.S. Steel. Gary. IN
LTV Steel. E.Chicago. IN
Armco Inc. .Ashl and.KY
Bethlehem Steel .Spar rows Pt . , MD
Rouge Steel .Dearborn. MI
National Steel, Detroit, MI
Carondolet, St. Louis. MO
Tonawanda. Buffalo. NY
Bethlehem Steel, Lackawanna, NY
LTV Steel. Warren, OH
Armco Inc.. Middletown.OH
New Boston. Portsmouth, OH
Koppers, Toledo. OH
LTV Steel, Cleveland. OH
U.S. Steel. Lorain, OH
AL Byproduct s .Keystone, PA
Bethlehem Steel, Bethlehem. PA
LTV Steel, AIiquippa.PA
LTV Steel. Pi t tsburgh.PA
Koppers, Erie. PA
Shenango, Pittsburgh, PA
U.S. Steel, Clairton, PA
U.S. Steel, Fairless Hills. PA
Wheel ing-Pi 1 1 . Monessen, PA
Chattanooga Coke, Chattanooga. TN
Lone Star Steel, Lone Star, TX
U.S. Steel, Provo, UT
Weirton Steel. Brown's Is.,WV
Wheeling-Pitt, E.Steubenvi 1 le.WV
20100
9070
34600
20200
10100
18500
27600
9550
6500
3810
19800
18800
7180
44200
56300
19000
9650
32200
8000
11900
13500
3550
12600
6050
28900
6100
4970
27800
37200
8310
28200
7970
22200
4390
3400
53800
12800
7210
2360
868
13500
7600
18700
7110
3190
12400
7230
3610
6600
12280
7920
5400
3270
9430
6930
3400
20960
26520
9050
8150
28800
7730
9110
4840
3060
11720
4880
11110
3730
3040
17550
22510
5650
19350
5400
15020
2980
2550
38180
8600
4950
2230
870
13530
4880
12090
                         E-24

-------
            TABLE E.6.
TOTAL EXPOSURE AND NUMBER OF PEOPLE EXPOSED
  (COKE OVEN BATTERIES BASELINE EMISSIONS)
Plant
                    Total
                  Number of
                People Exposed
     Total
    Exposure
(People - ug/rn-*)
Alabama By-Products - Terrant, AL
Empire Coke - Holt, AL
Koppers Company - Woodward, AL
LTV Steel - Gadsden, AL
LTV Steel - Thomas, AL
Jim Walters - Birmingham, AL
U.S. Steel - Fairfield, AL
National Steel - Granite City, IL
Interlake - S. Chicago, IL
LTV Steel - S. Chicago, IL
Bethlehem Steel - Burns Harbor, IN
Citizens Gas & Coke - Indianapolis, IN
Indiana Gas 4 Chemical - Terre Haute, IN
Inland Steel - E. Chicago, IN
U.S. Steel - Gary IN
LTV Steel - E. Chicago, IN
ARMCO Inc. - Ashland, KY
Bethlehem Steel - Sparrows Point, MD
Rouge Steel - Dearborn, MI
National Steel - Detroit, MI
Corondolet Corporation - St. Louis, MD
Tonawanda Coke Co. - Buffalo, NY
Bethlehem Steel - Lackawanna, NY
ARMCO Steel - Middletown, OH
New Boston Coke - Portsmouth, OH
Koppers Company - Toledo, OH
LTV Steel - Cleveland, OH
LTV Steel - Warren, OH
U.S. Steel - Lorain, OH
LTV Steel - Aliquippa, PA
Bethlehem Steel - Bethlehem, PA
Koppers Company - Erie, PA
U.S. Steel - Clairton, PA
U.S. Steel - Fairless Hills, PA
Shenango, Inc. - Pittsburgh, PA
LTV Steel - Pittsburgh, PA
Alabama By-Products - Keystone, PA
Wheeling-Pittsburgh - Monessen, PA
Chattanooga Coke & Chem.- Chattanooga, TN
Lone Star Steel - Lone Star, TX
U.S. Steel - Provo, UT
Weirton Steel - Browns Island, WV
Wheeling-Pittsburgh -
E. Steubensville, WV
810,000
189,000
790,000
335,000
803,000
806,000
802,000
2,180,000
6,210,000
6,090,000
2,810,000
1,190,000
242,000
5,340,000
4,400,000
5,470,000
436,000
2,530,000
4,170,000
4,040,000
2,210,000
1,240,000
1,260,000
2,170,000
298,000
831,000
2,570,000
999,000
1,720,000
2,230,000
1,170,000
333,000
2,280,000
4,350,000
2,410,000
2,340,000
4,910,000
2,110,000
575,000
184,000
750,000
1,230,000
1,030,000

2.29E+04
3.80E+03
3.39E+04
1.01E+04
1.39E+04
3.26E+04
2.64E+04
1.42E+04
2.00E+04
7.12E+03
5.63E+03
2.98E+04
6.64E+03
6.51E+04
4.01E+04
1.51E+04
4.36E+03
3.32E+04
1.65E+04
1.64E+04
3.30E+04
2.67E+03
2.01E+04
2.70E+04
3.28E+03
4.93E+03
4.72E+04
4.27E+03
2.39E+04
6.49E+03
2.97E+04
2.05E+03
5.72E+04
1.50E+04
6.51E+03
6.38E+04
2.40E+04
5.98E+03
2.88E+03
4.61E+01
3.31E+03
3.45E+03
7.75E+03

                                       E-25

-------
         TABLE E.7.   PUBLIC  EXPOSURE  TO BASELINE EMISSIONS AS
                      PREDICTED  BY THE HUMAN EXPOSURE MODEL
     Concentration
     Level
  Population
   Exposed
  (Persons)3
    Exposure
(Persons  -
      5.47E+01
      5.00E+01
      2.50E+01
      l.OOE+01
      5.00E+00
      2.50E+00
      l.OOE+00
      5.00E-01
      2.50E-01
      l.OOE-01
      5.00E-02
      2.50E-02
      l.OOE-02
      5.00E-03
      2.50E-03
      l.OOE-03
      5.00E-04
      2.50E-04
      l.OOE-04
      5.00E-05
      2.50E-05
      2.41E-05
       105
     1,580
     3,800
    13,900
    47,300
   125,000
   297,000
 1,000,000
 2,100,000
 4,770,000
12,300,000
21,900,00
38,400,000
64,600,000
80,200,000
86,000,000
88,700,000
88,800,000
88,800,000
88,800,000
    1.92E+01
    1.92E+01
    3.38E+03
    2.47E+04
    4.03E+04
    7.50E+04
    1.24E+05
    1.77E+05
    2.37E+05
    3.43E+05
    4.19E+05
    5.12E+05
    6.29E+05
    6.96E+05
    7.54E+05
    7.98E+05
    8.09E+05
    8.12E+05
    8.12E+05
    8.12E+05
    8.12E+05
    8.12E+05
a Column 2 displays the conputed value,  rounded  to  the  nearest  whole  number,
  of the cumulative number of people exposed  to  the matching  and  higher
  concentration levels found in column  1.   For example, 0.5 people would  be
  rounded to 0 and 0.51 people would be  rounded  to  1.

b Column 3 displays the computed value of  the cumulative exposure to  the
  matching and higher concentration levels found in column  1.
                                   E-26

-------
C.4.1.2  Aggregate Risk
     Aggregate risk is calculated by multiplying the  total  exposure produced
by HEM (for a single source, a category of  sources, or  all  categories of
sources) by the unit risk estimate.  The product is cancer  incidences among
the included population after 70 years of exposure.   The  total  exposure,
as calculated by HEM, is illustrated by the following equation:
                                      N
                     Total Exposure * I
where

     I  - summation over all grid points where exposure is  calculated

     P} - population associated with grid point i,

     C-j - long-term average inorganic arsenic concentration at  grid point  i,

     N  * number of grid points to 2.8 kilometers  and number of ED/BG
          centroids between 2.8 and 50 kilometers  of each source.

To more clearly represent the concept of calculating aggregate  risk, a
simplified example illustrating the concept follows:

                                  EXAMPLE

     This example uses assumptions rather than actual data  and  uses
only three levels of exposure rather than the large number  produced by
HEM.  The assumed unit risk estimate is 6.2 x 10~4 at 1 pg/m3 and  the
assumed exposures are:
                                    E-27

-------
          ambient air
        concentrations

          2    ug/m3
          1    ug/m3
          0.5  ug/m3
               number  of people  exposed
                to  given concentration

                         1,000
                        10,000
                       100,000
The probability of getting cancer  if  continuously exposed to the assumed
concentrations for 70 years is  given  by:
      concentration

       2    ug/m3    x
       1    ug/m3    x
       0.5  ug/m3    x
 unit risk

6.2  x 10-4
6.2  x IO-4
6.2  x 10-4
probability of cancer

      1 x lO'3
      6 x 10-4
      3 x 10-4
The 70 year cancer incidence among the people  exposed  to  these concentrations
is given by:
       probability of cancer
       at each exposure level
       number of people at
       each exposure level
1
6
3
X
x
x
io-3
io-4
io-4
x
x
x
1
10
100
,000
,000
,000
         after 70 years
           of exposure
                                                                    1
                                                                    6
                                                                   30
                                                            TOTAL  = 37
The aggregate risk, or total cancer incidence,  is  37  and,  expressed
as cancer incidence per year, is 37 * 70,  or  .5 cancers per  year.
The total cancer incidence and cancers per year apply to the total  of
111,000 people assumed to be exposed to the given  concentrations.
E.4.1.3  The Calculation of Individual  Risk
     Individual risk, expressed as "maximum  lifetime  risk,"  is  calculated
by multiplying the highest concentration to  which  the public  is exposed, as
reported by HEM, by the unit risk estimate.   The product,  a probability of
                                     E-28

-------
getting cancer, applies to the number of people which HEM reports as  being
exposed to the highest listed concentration.  The concept involved is a
sinple proportioning from the 1 ug/rn^ on which the unit risk estimate is
based to the highest listed concentration.  In other words:

       maximum lifetime risk          the unit risk estimate
     highest concentration to    a             1
     which people are exposed

E.4.2  Risks Calculated for BSO Emissions
     The explained methodologies for calculating maximum lifetime risk and
cancer incidences were applied to each coke oven, assuming a baseline level
of emissions.  The baseline level of emissions is the level of emissions
assumed for each battery at the performance levels defined by current state
regulations or consent  decrees.
     Table E.7 summarizes the calculated risks.  The analytical uncertainties
of these numbers are discussed in section E.5 below.
E.4.2.1  Control Scenarios
     The EPA completed HEM estimates of risk and incidence for the baseline
case at each of 43 coke ovens (see discussion, Section E.3.1.2).  These
estimates are outlined in Table E.8.  To ascertain the effect on maximum
individual lifetime risk and on annual incidence, the BAT control scenario
was also examined.  Using BAT-based emissions estimates as inputs to the HEM
model (Table E.5), EPA calculated risk and incidence values for the control
option.  Identical procedures were followed in risk and incidence calculations
for the baseline and control sceneries.
E.5  ANALYTICAL UNCERTAINTIES APPLICABLE TO THE CALCULATIONS OF PUBLIC
     HEALTH RISKS CONTAINED IN THIS APPENDIX
E.5.1  The Unit Risk Estimate
     The procedure used to develop the unit risk estimate is described in
Reference 3.  The model used and its application to epidemiological data
have been the subjects of substantial comment by health scientists.  The
uncertainties are too complex to be summarized sensibly in this appendix.
Readers who wish to go beyond the information presented in the reference
should see the following Federal Register notices:  (1) OSHA's "Supplemental
                                    E-29

-------
            TABLF E.8.   ANNUAL  INCIDENCE AND MAXIMUM LIFETIME
                        RISK FROM COKE  OVEN EMISSIONS
               PLANT
      BASELINE                  BAT
Incidence    Max Risk    Incidence    Max Risk

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
39
36
37
38
39
40
41
42
43

toppers. Woodward, AL
U.S. Steel. Gary. IN
U.S. Steel. Lorain. OH
Inland Steel, E. Chicago. IN
LTV Steel. Gadsden.AL
AL Byproducts, Tarrant, AL
LTV Steel, Cleveland, OH
Bethlehem Steel. Bethlehea. PA
Armco Inc.. Middletown.OH
Carondolet, St. Louis. MO
Jim Walters, Birmingham. AL
LTV Steel. Pi t tsburgh.PA
Bethlehem Steel. Lackawanna, NT
LTV Steel. Thomas. AL
U.S. Steel, Fairless Hills. PA
Bethlehem Steel. Purns Harbor. IN
National Steel, Granite City. IL
LTV Steel. E.Chicago. IN
Wheeling-Pitt. E.Steubenvi lle.WV
U.S. Steel. Fairfield. AL
Citixens Gas, Indianapolis, IN
U.S. Steel, Prove. UT
AL Byproducts, Keystone. PA
Empire Coke. Holt. AL
Wheeling-Pitt. Monessen, PA
New Boston, Portsmouth. OH
Rouge Steel .Dearborn, MI
LTV Steel, Aliquippa.PA
Weirton Steel. Brown's Is.,WV
IN Gas. Terre Haute. IN
Interlace. Chicago. IL
Tonawanda, Buffalo. NT
LTV Steel, Warren. OH
U.S. Steel. Clairton. PA a
Chattanooga Coke, Chattanooga, TN
Koppers, Toledo, OH
Shenango, Pittsburgh, PA
Armco Inc. .Ashland, KT
Lone Star Steel. Lone Star. TX
Koppert. Erie. PA
National Steel. Detroit. MI
Bethlehem Steel .Sparrows Pt., MD
LTV Steel. So. Chicago, IL b

0.300 2
0.350
0.210
0.580
0.090
0.200
0.420
0.260
0.240
0.290
0.290
0.570
0.180
0.120
0.130
0.050
0.130
0.130
0.069
0.230
0.260
0.029
0.210
0.034
0.053 :
0.029 I
o.iso :
0.057 i
0.031 :
0.059
0.160
0.024 i
0.038 1
0.220 J
0.026 -
0.044 !
0.058
0.039
0.0004 1
0.018 !
0.140 ,
0.290 1
0.031 J

I.39E-02
. 60E-02
.37E-02
-04E-02
.97E-02
. 86E-02
.77E-02
.45E-02
.34E-02
.26E-02
. 24E-02
.17E-02
.02E-02
.84E-03
. 83E-03
. 16E-03
.91E-03
.79E-03
.63E-03
. 06E-03
.04E-03
. 54E-03
.07E-03
.09E-03
J.80E-03
J.75E-03
J.54E-03
J.44E-03
J.28E-03
J.07E-03
J.01E-03
5.67E-03
6.79E-03
E.60E-03
8.33E-03
E.32E-03
I.79E-03
I.07E-03
I.61E-04
E.91E-04
E.86E-04
5.62E-04
S.81E-04

0.107
0.165
0.127
0.275
0.032
0.071
0.265
0.179
0.092
0.104
0.103
0.386
0.167
0.043
0.090
0.024
0.108
0.062
0.045
0.102
0. 124
0.029
0.143
0.012
0.036
0.018
0.145
0.039
0.020
0.028
0.150
0.021
0.031
0.156
0.025
0.027
0.044
0.033
.000
0.012
0.107
0.259
0.027

1.21E-02
1.23E-02
1.43E-02
9.68E-03
7.06E-03
6.93E-03
1. 12E-02
9.96E-03
5. 15E-03
4.51E-03
4.41E-03
7.92E-03
9.45E-03
3.53E-03
6.77E-03
4.37E-03
7.39E-03
4.18E-03
5.59E-03
3.59E-03
3.83E-03
7.54E-03
4.80E-03
1.44E-03
2.61E-03
2.29E-03
3.42E-03
2.33E-03
2.11E-03
1.45E-03
2.50E-03
2.48E-03
2.25E-03
1.85E-03
2.20E-03
1.42E-03
1.34E-03
9.04E-04
6.61E-04
1.97E-04
2.19E-04
2.52E-04
2.41E-04
a The value in the table was estimated with LONGZ.  The HEM  results  for annual
  incidence and maximum  lifetime risk are as follows:
  and BAT - 0.362, 2.01xlO"z.
                   Baseline - 0.51, 2.83x10
                                                                                -2
  The value reported in the table was estimated with  ISCLTU.   The  HEM results
  for annual incidence and maximum  lifetime risk  are  as  follows; Baseline  -
  0.063, 1.76xlO"J and BAT - 0.054, 1.51x10   .

                                      E-30

-------
Statement of Reasons for the Final  Rule",  48 FR  1864 (January  14,  1983);
and (2) EPA's "Water Quality Documents  Availability" 45  FR  79318  (November
28, 1980).
     The unit risk estimate used in this  analysis  applies only to  respiratory
cancer.  Other health effects are possible;  such as  neoplasms  of the kidney
and prostrate gland.  No numerical  expressions of  risks  relevant to these
health effects are included in this analysis.
     There are a number of uncertainties  associated  with the unit  risk
estimate.  The estimate Mas made from an  occupational  study that  Involved
exposures only after employment age was reached.  In estimating risks from
environmental exposures throughout  life,  it  was  assumed  through the absolute-
risk model that the Increase 1n the age-specific mortality  rates  of lung
cancer was a function only of cumulative  exposures,  irrespective  of how the
exposure was accumulated.  Although this  assumption  provides an adequate
description of all of the data, 1t  may be 1n error when  applied to exposures
that begin very early in life if coke oven emissions are early stage car-
cinogens.  Similarly, the linear models possibly are inaccurate at low
exposures, even though they provide reasonable descriptions of the
experimental data.
     The risk assessment methods employed were severely  constrained by the
fact that they are based only upon  the analyses  performed and  reported by
the original authors—analyses that had been performed for  purposes other
than quantitative risk assessment.   For example, although other measures of
exposure might be more appropriate, the analyses were necessarily  based
upon cumulative dose, since that was the  only usable measure reported.
Given greater access to the data from these  studies, other  dose measures,
as well as models other than the simple absolute-risk model, could be
studied.  It is possible that such  wide analyses would Indicate that other
approaches are more appropriate than the  ones applied here.
E.5.2  Public Exposure
E.5.2.1  General
     The basic assumptions implicit 1n  the methodology are  that all exposure
occurs at people's residences, that people stay  at the same location for 70
                                    E-31

-------
years, that the ambient air concentrations  and  the  emissions which cause
these concentrations persist for 70 years,  and  that the concentrations are
the same inside and outside the residences.  Hence, public exposure  is
based on a hypothetical premise.  It is  not known whether this results in
an over-estimation or an underestimation of public  exposure.
E.5.2.2  The Public
     The following may affect the public health impacts as estimated in
this analysis, however, the amount and type of  effect  is not known:
     1.  Studies show that all people are not equally  susceptible to cancer.
There is no numerical recognition of the "most  susceptible" subset of the
population exposed.
     2.  Studies indicate that whether or not exposure to a particular
carcinogen results in cancer may be affected by the person's exposure to
other substances.  The public's exposure to other substances is  not
numerically considered.
     3.  Some members of the public included in this analysis  are  likely to
be exposed to BSD in the air in the workplace,  and  workplace air concentrations
of a pollutant are customarily much higher  than the concentrations found in
the ambient, or public air.  Workplace exposures are not numerically
approximated.
     4.  Studies show that there is normally a  long latency period between
exposure and the onset of lung cancer.  This has not been numerically
recognized.
     5.  The people dealt with in the analysis  are  not located by actual
residences.  As explained previously, people are grouped by census districts
and these groups are located at single points called the population  centroids.
The effect is that the actual locations  of  residences  with respect to the
estimated ambient air concentrations are not known  and that the  relative
locations used in the exposure model may have changed  since the  1980 census.
     6.  Many people exposed to coke oven emissions are subject  to exposure
to ambient air concentrations of BSO where  they travel and shop  (as  in
downtown areas and suburban shopping centers),  where they congregate (as in
public parks, sports stadiums, and schoolyards), and where they  work outside
(as mailmen, milkmen, and construction workers).  These types  of exposures
are not numerically dealt with.
                                    E-32

-------
E.5.2.3.  The Ambient Air Concentrations
     The following are relevant to the estimated ambient air  concentrations
of BSO used in this analysis:
     1.  Flat terrain was assumed in the dispersion model. Concentrations
much higher than those estimated would result if emissions impact on  elevated
terrain or tall buildings near a plant.
     2.  The estimated concentrations do not account for the  additive impact
of emissions from plants located close to one another.
     3.  The increase in concentrations that could result from re-entrainment
of BSD-bearing dust from, e.g., city streets, dirt roads, and vacant  lots,
is not considered.
     4.  Meteorological data specific to plant sites are not  used in  the
dispersion model.  As explained, HEM uses the meteorological  data from the
STAR station nearest the plant site.  Site-specific meteorological data
could result in significantly different estimates, e.g., the  estimated
location of the highest concentrations.
     5.  Fugitive emissions from coke ovens are very difficult to measure
accurately and not only from battery to battery but also vary at the  same
battery overtime.  For door leaks, emission estimates range from 0.021
kg/hr for a single small leak to 0.7 kg/hr for a heavily leaking door.
Assumptions and theoretical models are used to relate visible emissions
(e.g., percent leaking doors) to mass emissions.  Consequently, the emission
estimates are subject to uncertainties that may be significant for a  specific
battery depending upon the site-specific characteristics of the battery and
its emissions.
                                    E-33

-------
E.6 REFERENCES
1.  National  Academy of  Sciences,  "Arsenic," Committee on Medical and
    Biological Effects of  Environmental Pollutants, Washington, D.C., 1977.
    Docket Number (OAQPS 79-8)  II-A-3.
2.  U.S. EPA, et.al., "Environmental Cancer and Heart and Lung Disease,"
    Fifth Annual  Report  to Congress by the Task Force on Environmental Cancer
    and Health and Lung  Disease, August,  1982.
3.  Carcinogen Assessment  of  Coke  Oven Emissions - Final Report.  EPA-600/
    6-82-003F, February  1984.
4.  Systems Application, Inc.,  "Human Exposure to Atmospheric Concentrations
    of Selected Chemicals."  (Prepared for the U.S. Environmental Protection
    Agency, Research Triangle Park, North Carolina).  Volume I, Publication
    Number EPA-2/250-1,  and Volume II, Publication Number EPA-1/250-2.
5.  Branscome, Marvin.   Research Triangle Institute.  Ambient Data for Indicators
    of Coke Oven Emissions.  Docket Number A-79-15.  August 1985.
6.  Regan, G. F.  EPA Region  5. Benzo-a-pyrene as a Tracer for Coke Oven
    Emissions.  Paper provided  by  the author.
7.  Pearson,  Johnnie L.  (1985). Evaluation of Coke Oven Emission Dispersion
    Estimates, Memorandum to  Robert G. Kellam, EPA, Office of Air
    Quality Planning and Standards, Research Triangle Park, North Carolina
    27711, August 1985.
                                    E-34

-------
                                    TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
   EPA-450/3-85-028a
                              2.
                                                            3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
   Coke Oven  Emissions from  Wet-Coal Charged By-Product
   Coke Oven  Batteries—Background Information  for
   Proposed Standards	
                                                            5. REPORT DATE
                                                              April  1987
             6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
             8. PERFORMING ORGANIZATION REPORT NO. J

                                              1
9. PERFORMING ORGANIZATION NAME AND ADDRESS
                                                            10. PROGRAM ELEMENT NO.
   Office of Air Quality Planning and Standards
   U.S. Environmental Protection Agency
   Research  Triangle Park, North Carolina  27711
              11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
   DAA for Air Quality Planning and Standards
   Office of Air and Radiation
   U.S. Environmental Protection  Agency
   Research Triangle Park, North  Carolina  27711
              13. TYPE OF REPORT AND PERIOD COVERED
                Draft
              14. SPONSORING AGENCY CODE

                 EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
        National  emission standards to control coke  oven emissions  from new and
   existing wet-coal  charged by-product coke oven  batteries are being  proposed under
   Section 112  of the Clean Air Act.   The standards  apply to charging,  topside, and
   door leaks.   This  document contains information on  the background and authority,
   emission control  techniques, regulatory alternatives considered, and the economic
   and health  impacts associated  with the proposed standards.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS  C.  COSATI Field/Group
   Air Pollution
   Wet-Coal Charged Coke Oven Batteries
   Pollution Control
   Coke Oven Emissions
   Emission Standards
   Control Costs
 Air  Pollution Control
 Hazardous Air Pollutants
18. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report)
  Unclassified
21. NO. OF PAGES
   500
                                               20. SECURITY CLASS (Thispage)

                                                 Unclassified
                                                                          22. PRICE
EPA Form 2220-1 (R«v. 4-77)    PREVIOUS EDITION is OBSOLETE

-------