P/EPA
              United States
              Environmental Protection
              Agency
               Office of Air Quality
               Planning and Standards
               Research Triangle Park NC 27711
September 1983
              Air
Coke Oven Emissions
from Wet-Coal Charged
By-Product Coke Oven
Batteries—Background
information for
Proposed  Standards
     Draft
     EIS

-------
                             DISCLAIMER
This draft document has not been formally released by the Environmental
Protection Agency and should not be construed to represent Agency policy.
The document is being circulated for comment on its technical accuracy
and policy implications.

-------
Coke Oven Emissions from By-Product Coke
  Oven Charging, Door Leaks, and Topside
   Leaks on Wet-Coal  Charged Batteries—
          Background Information for
             Proposed  Standards
                       DRAFT
               Emission Standards and Engineering Division
              U.S. ENVIRONMENTAL PROTECTION AGENCY
                 Office of Air, Noise, and Radiation
               Office of Air Quality Planning and Standards
               Research Triangle Park, North Carolina 27711
                      September 1983

-------
                             TABLE OF CONTENTS
Tables	   ix

Figures	xiii

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	3-31
          3.3.3  Coke Oven Doors and Their Emissions	3-33
               3.3.3.1  Facility Description	3-33
               3.3.3.2  Emission Sources and Pollutants 	   3-39
               3.3.3.3  Emissions from Poorly Controlled Door
                        Leaks	3-40
               3.3.3.4  Emissions from Wei1-Controlled Door
                        Leaks	3-44
               3.3.3.5  Factors Affecting Emissions  from Coke
                        Oven Doors	3-46
                    3.3.3.5.1   Oven pressure	   3-46
                    3.3.3.5.2   Temperature effects	3-49
                    3.3.3.5.3   Miscellaneous factors affecting
                               emissions	   3-51
                                     11

-------
                           CONTENTS (continued)

                                                                       Page
          3.3.4  Topside Leaks and Their Emissions	3-52
     3.4  BASELINE REGULATIONS	   3-54
     3.5  REFERENCES	3-63

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	4-36
               4.1.2.1  Description 	   4-36
               4.1.2.2  Performance of Sequential Charging	4-40
          4.1.3  Scrubber Systems Mounted on Larry Cars 	   4-41
               4.1.3.1  Description 	   4-41
               4.1.3.2  Performance of Scrubber Systems Mounted on
                        Larry Cars	4-45
     4.2  TECHNOLOGY FOR THE CONTROL OF DOOR LEAKS	4-47
          4.2.1  Traditional Oven Door Seal  Technology	4-49
               4.2.1.1  Koppers Door	4-49
               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  	   4-88
     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  AIR POLLUTION IMPACT	   7-1
          7.1.1  Primary Air Impact for Model Batteries  	   7-1
          7.1.2  Nationwide Primary Air Impacts 	   7-6
          7.1.3  Ambient Concentrations of BSD from Dispersion
                 Modeling	   7-7
     7.2  OTHER IMPACTS	   7-9
     7.3  IMPACT OF DELAYED STANDARDS 	   7-12
     7.4  REFERENCES	7-12

8.   COSTS	8-1
     8.1  COST ANALYSIS OF REGULATORY ALTERNATIVES  	   8-1
          8.1.1  Basis for Model Battery Costs  	   8-2
          8.1.2  Cost of Regulatory Alternatives for Charging ....   8-9
               8.1.2.1  Existing Batteries	8-9
               8.1.2.2  New Batteries 	   8-13
          8.1.3  Cost of Regulatory Alternatives for Topside Leaks.  .   8-16
               8.1.3.1  Existing Batteries	8-16
               8.1.3.2  New Batteries 	   8-21
          8.1.4  Cost of Regulatory Alternatives for Door Leaks  .  .  .   8-21
               8.1.4.1  Existing Batteries 	    8-21
               8.1.4.2  New Batteries 	   8-27
          8.1.5  Cost Effectiveness	8-27
          8.1.6  Comparison of Estimated Costs with Actual  Plant
                 Costs	8-30
          8.1.7  Summary of Nationwide Costs  	   8-30

-------
                           CONTENTS (continued)
     8.2  OTHER COST CONSIDERATIONS	8-31
          8.2.1  Desulfurization	8-31
          8.2.2  OSHA	8-33
          8.2.3  Water Treatment	8-33
          8.2.4  Pushing,  Quenching,  and Stack Emission Controls.  .  .   8-38
     8.3  REFERENCES	8-38

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	9-3
               9.1.1.2  Brief History of the Coke Industry in the
                        Overall  Economy 	   9-3
               9.1.1.3  Size of  the Iron and Steel  Industry	9-4
          9.1.2  Production	9-9
               9.1.2.1  Production Description	9-9
               9.1.2.2  Production Technology 	   9-9
               9.1.2.3  Factors  of Production 	  ....   9-11
          9.1.3  Demand and Supply Conditions 	   9-15
          9.1.4  Market Structure 	   9-18
               9.1.4.1  Concentration Characteristics  and Number
                        of Firms.	9-19
               9.1.4.2  Integration Characteristics  	   9-28
               9.1.4.3  Substitutes 	   9-28
               9.1.4.4  Pricing  History 	   9-29
               9.1.4.5  Market Structure Summary	   9-31
          9.1.5  Financial Performance	   9-31
          9.1.6  Projections	   9-35
          9.1.7  Market Behavior:   Conclusions	   9-40
     9.2  ECONOMIC IMPACT  OF REGULATORY ALTERNATIVES	9-41
          9.2.1  Summary	   9-41
          9.2.2  Methodology	9-42
               9.2.2.1  Supply Side	   9-43
                    9.2.2.1.1 Data base	   9-43
                    9.2.2.1.2 Output relationships  	   9-45
                    9.2.2.1.3 Operating costs	'  .   9-45
                    9.2.2.1.4 Capital costs	9-47
                    9.2.2.1.5 Environmental costs	9-48
                    9.2.2.1.6 Coke supply function—existing
                              facilities	   9-50
                    9.2.2.1.7 Coke supply function—new
                              facilities	   9-53
               9.2.2.2  Demand Side	9-53
               9.2.2.3  Synthesis 	   9-58
               9.2.2.4  Economic Impact Variables 	   9-58
          9.2.3  Furnace Coke Impacts	9-62
               9.2.3.1  Price Effects 	   9-65
               9.2.3.2  Production and Consumption  Effects	9-67

-------
                           CONTENTS (continued)
               9.2.3.3  Coal Consumption and Employment Effects .  .  .   9-69
               9.2.3.4  Financial Effects 	   9-69
               9.2.3.5  Battery and Plant Closures	9-69
          9.2.4  Foundary Coke Impacts	9-73
               9.2.4.1  Price and Production Effects  	   9-74
               9.2.4.2  Coal Consumption and Employment Effects .  .  .   9-74
               9.2.4.3  Financial Effects 	   9-79
               9.2.4.4  Battery and Plant Closures	9-79
     9.3  POTENTIAL SOCIOECONOMIC, INFLATIONARY, SMALL BUSINESS,
          AND ENERGY IMPACTS	9-81
          9.3.1  Potential Socioeconomic and Inflationary
                 Impacts	9-81
                 9.3.1.1  Compliance Costs	9-81
                 9.3.1.2  Prices and Consumer Costs 	   9-83
                 9.3.1.3  Balance of Trade	9-83
                 9.3.1.4  Community Impacts 	   9-85
                 9.3.1.5  Conclusions 	   9-85
          9.3.2  Small Business Impacts 	   9-85
          9.3.3  Energy Impacts	9-87
     9.4  IMPACTS OF VARIED ECONOMIC CONDITIONS:  A SENSITIVITY
          ANALYSIS	9-87
          9.4.1  Effects of Change in Coking Time on Furnace
                 Coke Price and Production	9-88
          9.4.2  Effects of Change in Coking Time on Foundry
                 Coke Price and Production	9-93
     9.5  REFERENCES	9-93

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  SUMMARY OF CANCER-RISK ASSESSMENT 	   E-l
                                    vn

-------
                                  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
          1978	  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	3-30
3-7       Types of Ovens in Current Use	  3-38
3-8       Major Components of Coke Oven Gas	  3-40
3-9       Minor Components of Coke Oven Gas	  3-40
3-10      POM Pollutants from Coke Oven Door Leaks	3-41
3-11      Cokeside Shed Test Results	  3-43
3-12      Exponential Model for Door Leak Emissions	3-45
3-13      The Effect of Temperature on the Relative Coking
          Time	  3-50
3-14      Topside Leak Emission Test	3-53
3-15      State Regulations for Coke Oven Emissions	3-55
3-16      Consent Decrees for Charging	  3-56
3-17      Consent Decrees for Door Leaks	  3-58
3-18      Consent Decrees for Topside Leaks 	  3-59
3-19      Summary of Baseline Regulations for Charging	3-61
3-20      Summary of Baseline Regulations for Door Leaks	3-61
3-21      Summary of Baseline Regulations for Lid Leaks 	  3-61
3-22      Summary of Baseline Regulations for Offtake Leaks 	  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
                                   vm

-------
                            TABLES (continued)

Number                                                                 Page

4-7       Particulate Emissions from Scrubbers Mounted on
          Larry Cars	4-46
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	4-90
4-13      Visible Emission Data on Topside Leaks from Charging
          Port Lids, U.S. Steel, Clairton	4-91
4-14      Visible Emission Data on Offtake Leaks,  Kaiser Steel,
          Fontana	4-92
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
6-3       Regulatory Alternatives for Dook Leaks for Batteries with
          Cokeside Sheds of Varying BSO Removal Efficiencies .  .  .      6-10

7-1       Air Emissions from Charging	    7-2
7-2       Air Emissions from Door Leaks	    7-4
7-3       Air Emissions from Lid Leaks	    7-5
7-4       Air Emissions from Offtake Leaks	    7-5
7-5       Nationwide Emissions of BSO	    7-8
7-6       Maximum Average Annual BSO Concentrations
          for Model  Battery 2	    7-10

8-1       Model Batteries	    8-3
8-2       Summary of Requirements for Each Regulatory Alternative.  .    8-4
8-3       Capital Cost Items	    8-6
8-4       Annualized Costs 	    8-8
8-5       Cost Estimates for Charging Regulatory Alternative
          I — Existing Sources	    8-10
8-6       Cost Estimate for Engineering Study of Charging Emission
          Control	    8-12
8-7       Incremental Cost Estimates for Charging  Regulatory
          Alternative II--New and Existing Sources 	    8-14
8-8       Cost Estimate for Charging Regulatory Alternative
          I--New Sources 	    8-15
8-9       Cost Estimates for Topside Leaks Regulatory Alternative
          I--New and Existing Sources	    8-17
8-10      Incremental Labor Required for Topside Leaks
          Regulatory Alternative II	    8-19
8-11      Incremental Cost Estimates for Topside Leaks
          Regulatory Alternative II--Existing Sources	    8-20
                                     IX

-------
                            TABLES (continued)

Number                                                                 Page

8-12      Incremental Cost Estimate for Topside Leaks Regulatory
          Alternative II--New Sources	    8-22
8-13      Labor Requirements for a Door Leak Control  Program ....    8-23
8-14      Cost Estimates for Door Leaks Regulatory Alternative
          I—Existing Sources	    8-25
8-15      Incremental Cost Estimates for Dook Leaks Regulatory
          Alternative II--Existing Sources	8-26
8-16      Cost Estimate for Door Leaks Regulatory Alternative
          I--New Sources	8-28
8-17      Cost Effectiveness of the Regulatory Alternatives 	   8-29
8-18      Nationwide Incremental Costs to Attain Regulatory
          Alternative II	8-32
8-19      Estimated Capital Cost of Desulfurization 	   8-34
8-20      Estimated Capital and Annual Operating Costs of OSHA
          Compliance	8-34
8-21      Cost Estimates for a Biological System Wastewater
          Treatment Facility  	   8-35
8-22      Cost Estimates for a Wastewater Treatment Facility for
          the Model Batteries	8-36
8-23      Cost Estimates for Pushing, Quenching, and  Stack
          Emissions—New Sources	8-37

9-1       Coke Industry Foreign Trade 	   9-5
9-2       Coke Production in the World	9-6
9-3       Value of Shipments, SIC 3312	9-7
9-4       Value Added, SIC 3312	9-8
9-5       Potential Maximum Annual Capacity of Oven Coke
          Plants in the United States on July 31, 1979	9-12
9-6       Typical Cost Breakdowns:  Furnace Coke Production
          and Hot Metal (Blast Furnace) Production	9-13
9-7       Employment in the By-Product Coke Industry	9-14
9-8       Coke Rate	9-17
9-9       Coke Plants in the United States, January 1980	9-20
9-10      Interregional Coke Shipments in 1980	9-26
9-11      Percent of Coke Capacity Owned by Top Firms
          (January 1980)	9-27
9-12      Comparison of Coal Prices and Domestic and  Imported
          Coke Prices	9-30
9-13      Financial Information on Coke-Producing Firms, 1980 ....   9-32
9-14      Financial Ratios for Coke-Producing Firms 	   9-33
9-15      Summary of Steel Industry Projections 	   9-36
9-16      Summary of Coke Industry Projections	9-37
9-17      Projections of Coke Capacity 1985, 1990, and 1995 	   9-39
9-18   .   Estimated Capital Costs of New Batteries	9-49
9-19      Estimate of Elasticities of Steel and Coke  Markets	9-57
9-20      Economic Impact Variables and Affected Sectors	9-60
9-21      1983 Baseline Values for Economic Impact
          Analysis—Furnace Coke	9-63

-------
                            TABLES (continued)

Number                                                                 Page

9-22      Baseline Control Costs—Furnace Coke	9-64
9-23      Price Effects of Regulatory Alternatives—Furnace
          Coke, 1983	   9-66
9-24      Production and Consumption Effects of Regulatory
          Alternatives—Furnace Coke, 1983	9-68
9-25      Coal Consumption and Employment Effects of Regulatory
          Alternatives—Furnace Coke, 1983	9-70
9-26      Industry Capital Requirements of Regulatory
          Alternatives—Furnace Coke, 1983	9-71
9-27      1983 Baseline Values for Economic Impact
          Analysis—Foundry Coke	9-75
9-28      Baseline Control Costs—Foundry Coke	9-76
9-29      Price and Quantity Effects of Regulatory
          Alternatives—Foundry Coke, 1983	9-77
9-30      Coal Consumption and Employment Effects of Regulatory
          Alternatives—Foundry Coke, 1983	9-78
9-31      Industry Capital Requirements of Regulatory
          Alternatives—Foundry Coke, 1983	9-80
9-32      Compliance Costs of Regulatory Alternatives Under
          Scenario A,  1983	   9-82
9-33      Coke, Steel,  Ferrous Foundry, and Consumer Products
          Price Effects of Regulatory Alternatives,  1983	9-84
9-34      Effects of Increased Coking Time on Furnace
          Coke Price	9-89
9-35      Effects of Increased Coking Time on Furnace
          Coke Production	9-90
9-36      Effects of Increased Coking Time on Foundry
          Coke Price	9-91
9-37      Effects of Increased Coking time on Foundry
          Coke Production	   9-92
                                    XI

-------
                                    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       Uncontrolled charge 	   3-32
3-10      Vertical cross section of coke oven door on pusher
          side of oven	 .  .  .  .   3-34
3-11      Outside elevation of a pusherside coke  oven door	3-35
3-12      Vertical cross section of a pusherside  coke oven door .  .  .   3-36
3-13      Horizontal cross section of a coke oven door	3-37
3-14      The effect of time on the internal oven pressure	3-47

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 	   4-35
4-10      AISI/EPA larry car coke oven charging system	4-37
4-11      Flow of gases in the AISI/EPA sequential charging
          system	4-38
4-12      Representative scrubber system on a larry car 	   4-42
4-13      Details of a shrouded drop sleeve	4-44
                                    xi

-------
                            FIGURES (continued)

Number                                                                 Page

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 .  .   9-10
9-2       United States apparent consumption of coke	9-16
9-3       Coke plants in the United States,  1980	9-24
9-4       Economic impact model 	   9-44
9-5       Coke plant cost centers	9-46
9-6       Estimated average cost of furnace  coke  production as a
          function of plant production, 1980	9-51
9-7       Estimated average cost of foundry  coke  production as a
          function of plant production, 1980	9-52
9-8       Marginal and average cost functions for furnace
          coke, 1980	9-54
9-9       Marginal and average cost functions for foundry
          coke, 1980	9-55
9-10      Coke supply and demand	9-59
9-11      Coke demand and supply with and without regulatory
          alternatives	9-61

-------
                                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)Xconsent decree limits).

    II -  For all new and existing batteries, a maximum of 16 seconds
          of visible emissions per charge (log average* of 10 charges).

   Ill -  For all new and existing short batteries, a maximum of
          8 seconds of visible emissions per charge (log average of
          10 charges).   For all  new and existing tall batteries, a
          maximum of 15 seconds  of visible emissions per charge (log
          average of 10 charges).

     Regulatory Alternative I requires full stage charging,  and Regulatory

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).

    II -  Twelve percent leaking doors (PLD) for all new and existing
          batteries.

   Ill -  Seven PLD for all new  and existing short batteries and
          10 PLD for all new and existing tall 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  +1n (*2 ^ D + •  •  •  In (Xn + 1)

X = seconds of visible emissions,  and n = number of observations.
                                 1-1

-------
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).
    II -  Three percent leaking lids (PLL) and 6 percent leaking offtakes
          (PLO).
   Ill -  One PLL and 4 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 Alterna-
tive 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 1,100 Mg/yr (190 to 2,100 Mg/yr).  Regulatory
Alternative II for the combined sources reduces nationwide BSO emissions to
820 Mg/yr (140 to 1,500 Mg/yr).  Regulatory Alternative III for the combined
sources would reduce BSO emissions to 350 Mg/yr (61 to 630 Mg/yr).
     The only alternative that may affect water pollution is Regulatory
Alternative I for charging.  The charging alternative would increase the
total plant wastewater volume by less than 6 percent.  None of the regula-
tory alternatives would have an impact on the generation of solid waste.
Regulatory Alternative I for charging could increase the plant's steam
consumption by 10 percent.  No other energy impacts are known.
1.3  ECONOMIC IMPACT
     The derivation of the following costs was based on a battery-by-
battery cost analysis.   Available data were used to estimate each battery's
status in 1977-1979 and its control requirements for each regulatory alterna-
tive.  A discussion of regulatory baseline costs is also provided because
                                  1-2

-------
I
CO
                      TABLE  1-1.  ASSESSMENT OF  ENVIRONMENTAL AND  ECONOMIC  IMPACTS  FOR  EACH
                                        REGULATORY ALTERNATIVE CONSIDERED
Source
Wet-coal charging


Door leaks


Topside leaks


Administrative action
Regulatory Alternative I
Regulatory Alternative II
Regulatory Alternative III
Regulatory Alternative I
Regulatory Alternative II
Regulatory Alternative III
Regulatory Alternative I
Regulatory Alternative II
Regulatory Alternative III
Air
impact
+4
+2
+2
+4
+2
+2
+4
+2
+2
Water
impact
-2
0
0
0
0
0
0
0
0
Solid waste
impact
0
0
0
0
0
0
0
0
0
Energy
impact
-1
0
0
0
0
0
0
0
0
Noise
impact
0
0
0
0
0
0
0
0
0
       Long-term  impact.
       KEY:   +  Beneficial  impact
             -  Adverse  impact
             0  No  impact
             1  Negligible  impact
2 Small impact
3 Moderate impact
4 Large impact

-------
many batteries have not yet fully incurred the cost necessary to meet the
baseline.
     The total annualized cost to the industry to change from its 1978
status to the regulatory baseline for charging is estimated as $23 million
per year.   Regulatory Alternative II for charging may require a nationwide
capital investment of $6.5 million with a total annualized cost of $5.6 mil-
lion per year above the regulatory baseline costs.   No costs could be
estimated for charging Regulatory Alternative III.
     The total annualized cost to the industry to change from its 1977-1979
status to the regulatory baseline for door leaks is estimated as $77 million
per year.   Regulatory Alternative II for door leaks may require a nationwide
capital investment of $33 million with a total annualized cost of $12 mil-
lion per year above the regulatory baseline.   No costs could be estimated
for Regulatory Alternative III for door leaks.
     The total annualized cost for the industry to change from its 1977-1979
status to the regulatory baseline for topside leaks is estimated as $1 mil-
lion per year.  Regulatory Alternative II for topside leaks may require a
nationwide capital investment of $5.3 million with a total annualized cost
of $2 million per year above the regulatory baseline.   Costs could not be
estimated for Regulatory Alternative III for topside leaks.
                                    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 	.
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 (93 percent in 1978)
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 assure 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 pro-
duced by the slot oven process.  This conversion of coal to coke is per-
formed in long, narrow slot ovens  which are designed to permit separation
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 1978 only 60 by-product coke plants re-
mained in operation.2  The industry has two sectors, and plants are classi-
fied generally as "furnace" and "merchant."  Furnace plants are owned by or
affiliated with iron- and steel-producing companies which produce coke
primarily for consumption in their own blast furnaces,  although they also
engage in some intercompany sales  among steel firms with excesses or de-
ficits in coke capacity.   In 1978  there were 47 furnace plants which
accounted for 93 percent of the total coke production.
     Independent merchant plants which produce  coke for sale on the open
market are typically owned by chemical or coal  firms.   The 13 merchant
                                  3-1

-------
plants operating in 1978 accounted for 7 percent of the total coke pro-
duced.  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 hydrocar-
bon gases that are released when coal is converted into coke; coal firms
have entered the coke industry as a form of downstream vertical  integra-
tion.
     In 1978, 43.8 million megagrams of coke were produced in slot ovens in
the United States.   This production was less than the 1977 production level
of 48.1 million megagrams and was 16 percent less than the 1976 production
level of 52.4 million megagrams.3  In 1978 furnace plants produced 40.7
million megagrams of coke and merchant plants produced 3.1 million mega-
grams 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.
The production of coke from the beehive process accounted for only 0.3
million megagrams,  which was 0.7 percent of the total coke production
during 1978.3  A production and consumption history of coke in the United
States since 1970 is presented in Table 3-1.
     Although coke was produced in 18 States in 1976, 57 percent of the
production occurred in three eastern States:   Pennsylvania,  Ohio, and
Indiana.2  Pennsylvania, with 14.7 million megagrams of output,  was the
leading coke producing State and accounted for 28 percent of U.S. coke
production.   Ohio produced 7.6 million megagrams of coke while Indiana
produced 7.5 million megagrams of coke.   The relative amounts of coke
produced in the various States have changed little in the past decade.  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 1971 through
1976.2 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 which results from the crushing and screen-
ing of the coke after it is removed from the oven.   Although not completely
standardized, the term breeze is generally applied to coke that will pass
                                  3-2

-------
                   TABLE 3-1.   PRODUCTION  AND CONSUMPTION HISTORY OF COKE IN THE  UNITED STATES'









CO
1
CO





Year
1970
1971
1972
1973
1974
1975
1976

1977
1978

Number
of
merchant
plants
10
16
11
14
14
14
13

12
12

Number
of
furnace
plants
48
48
48
40
48
48
48

48
48

Total
number
of
coke
oven
plants
63
64
62
62
62
62
61

60
60

U.S. oven coke produced

Merchant
plant
5.37
5.05
5.11
4.78
4.64
4.28
3.95

3.33
3.09


Furnace
plant
54.19
46.36
49.20
52.82
50.47
46.97
48.42

44.80
40.66



Total
59.56
51.41
54.39
57.60
55.11
51.25
52.37

48.13
43.75

Percent
change
In
production
from
previous
year
+2.5
-13.7
+5.6
+ 6.1
-4.3
-7.0
+2.2

-8.0
-9.1

Coke
imports
0.14
0.15
0.16
0.997
3.17
1.63
1.18

1.90
5.17

U. S. oven coke uses
By
producing
companies
51.99
46.14
48.46
52.59
49.12
42.31
47.04

44.69
47.93

For

commercial
sales
7.76
6.51
6.35
6.88
6.20
5.26
3.93

3.52
3.12

Total
59.25
52.65
54.81
59.47
55.32
47.57
50.97

48.21
51.05

Coke

In iron
furnaces
52.75
46.72
49.54
55.08
53.02
44.29
46.77

43.99
47.32

consumption
All
other
purposes
4.59
4.71
4.93
4.58
5.13
4.42
4.78

5.08
3.73



Total
57.34
51.43
54.47
59.66
50.15
48.71
51.55

49.07
51.05

Dlasl
furnace
iron
produced
83.20
74.12
81.10
91.81
86.99
72.48
78.83

73.75
79.56

Percent
chaiufc
in
production
f rom
prev ions
year
-3.9
-11.0
• 9.4
U3.2
-5.2
-16.7
• 8.8

-6.4
• 7.9

Coke consumption figures include beehive coke.

-------
u>
 I
                     TABLE  3-2.   GEOGRAPHICAL  LOCATION  AND  PRODUCTION  OF COKE  IN  THE  UNITED  STATES2  4
                                   ,  .                     Kentucky,

                                   California.                Missouri.
                                   Colorado.
Year
1971

1972

19/3
1974

19/5

19/6

1977 1
1978 /
aNumber
Alabama
7a
4.865b
7
4.858
4.656
7
4.646
7
4.255
7
4.229
Data not
ol establis
lu 1 uidUQ ,
Utah |
3
2.704
3
2.681
3
3.072
3
3.012
3
2.718
3
2.940
yet aval lable.
ihnenls.
1 linois
*
1.945
A
1.891
4
1.761
*
1. 734
A
1. 745
*
i.548


Indiana
f
7.105

8.338
6
8.468

D
8.231

b
8.388

6
7.49?


Tennessee,
Texas

1.773

5
1.904
5
1. 756

5
1.742

5
1.586

5
1.212


Maryland,
New York
.c
4
5.429

4
4.930
4
6.414

4
5.858

4
4.806

4
5. 752


Michigan

3
3.429

3
3.336
3
3.512

3
2.956

3
2.942

3
3.078

— —
Minnesota.
Wiscons in

3
0.711

3
0.740
3
0.766

3
0.865

3
0. 708

3
0.649

=^— 	 = 	 = 	 i
Ohio

12
6.872

12
8.038
12
8.562

12
8.022

12
7.681

11
7.647

— • -'-• —
Pennsylvania

12
13.844

12
14.396
12
15.159

12
14.808

12
13.966

12
14.674

-j--_- - _-.".._.
West Virginia

3
2.727

3
3.184
3
3.522

3
3.225

3
2.457

3
3 147

•-•=-= — *•-—=—- _-_--
Total

64
51.404

62
54.297
62
57.602

62
55.099

62
51.250

61
52.369

	 .--...
                     Production quantity (millions of metric tons).

                    CI971  includes New Jersey.

-------
through a half-inch screen.  In 1976, coke plants produced 3.9 million
megagrams of breeze, which is equivalent to 51 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 evolved during the
coking process are used within the facility or are marketed.   Typical
products and by-products from the production of coke during 1978 are pre-
sented in Table 3-3.3
     Approximately 93 percent (47.3  million megagrams) of the coke distrib-
uted by U.S.  producers in 1978 was shipped to blast furnace plants, 5 per-
cent went to foundries, and the remainder went to other industrial plants.3
Apparent consumption (or total consumption) of coke in the United States
equaled 51 million megagrams in 1978 and 48.2 million megagrams in 1977.
The apparent consumption is a quantity which 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 1977, apparent consumption
decreased from the 1976 level.  Both apparent consumption and blast furnace
iron production increased in 1978.
     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 19783
              Product
    Total production
     Coke
     Breeze
     Crude tar
     Crude light oil
     Ammonia (sulfate equivalent)
     Coke oven gasa
43.8 million megagrams
 2.2 million megagrams
  2,050 million liters
  623.6 million liters
 0.4 million megagrams
22 billion cubic meters
      Estimated from the 1976 ratio of gas to coke.
                                  3-5

-------
increased significantly during the past 2 decades, the blast furnace coke
rate has declined.   In 1976, an average of only 593.5 kg of coke was re-
quired 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 quan-
tities 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

-------
                                                                             Standpipe Caps
                          Coal Ports
CO
                                     Waste Gas
                                     Stack
Coke Side
                     Coke Guide
                                           Figure 3-1. Schematic diagram of by-product coke battery.

-------
        I COAL HOIST
                                  CONVEYOR BELTS
L__J
R.R. CAf
DUMPER
TAR
STORAGE
W :;:I(M», :s.,
T
FOR FURTHER
PROCESSING
OR FOR FUEL
WEAK AMMONIA
LIQUOR
SETTLING
TANKS
r
1 -«- llli
r^l rH f-Li r*~i
L $2 J ™-nnnn
/ I TTT
COAL BRIDGE , '- . • f
FLUSHING LIQUOR
11
OE^A^R «="»»«"
1 	 «-O ft 	 ,, <-»''«Hft
COKE OVENS COKE - .;;,;,. . 	 . fH
DEHYORATOR f |
GAS STATION WHARF STA
Ms! 1
PRIMARY S»S FOR UNDERFIRING COKE OVENS

3 COKE SCREENINGS
NINO
TION
| TAR p-Ui "]
PRIM!^ li_:J--^-L^J »c»
COOLER
DECANTER AMMONIA LIGHT OIL , ...,.
| PRECIPITATOR 1 ABSORBER SCRUBBER HOt OFR B°OSTER
RHENOL C±T ~ ^ ^^--^^Tl ^^^l^-^T
EXTRACTOR EXHAUSTER T"^ FINAL HYDROGEN
~H COOLER SULFIDE
. 	 . „„, t SCRUBBER
/^\ "™s

Vi_i/ DEBENZOLIZEO WASH OIL
AMMONIA
* BENZOLIZEO WASH OIL

FOR FURTHER
PROCESSING NAPHTHALENE

Figure 3-2.  Flow sheet showing the major steps in the by-product coking process.
             (copyright 1971 by United States Steel Corporation)

-------
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

-------
                                      TOPSIDE PORT
   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
                                                         _s
   a  THE COAL FROM THE LARRY HOPPERS HAS DROPPED INTO THE OVEN CHAMBER  FORMING
      PEAKED PILES.
                                       LEVELING 8AR
                    COAL
   C.  THE  LEVELING OOOR 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 cool down 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

-------
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 GUIDE
          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
                                             r
          THE RAM OF THE PUSHER ADVANCES TO PUSH THE INCANDESCENT COKE OUT OF THE OVEN,
          THROUGH THE COKE GUIDE  AND INTO THE OUENCHING 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 (BSO) 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 BSO 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             BSO (kg/Mg of coal)            BSD emission rate
Charging3         0.059 to 0.5510 n           0.2 to 1.5 kg/mi n of emissions
Doorsb            0.13 to 0.2510 n 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    oors x 0.3 (fraction leaking)
                             = 37

     kg of BSO per year = 37 doors x   JLS. x 8^P_hr  = 6
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    « "a1  x *t™S. x           =  492_000
     Emission factor = 64,800 T 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 BSO 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 BSO 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:
                                    T
                              PLD = ^   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; nowever, ovens are charged sequentially and the collective effect
is a major source of emissions.
                                  3-22

-------
                                                                            LARRY CAR CAB
r\>
CO
          FEED HOPPER
          WITH SHUT-OFF
          VALVE
                 DROP SLEEVE
            CHARGING
            PORTAL
          COKE OVEN

STANDPIPE CAP

 ASCENSION PIPE
     - COLLECTION MAIN
  TOPSIDE
                                           Figure 3.7. Schematic of coke oven topside.

-------
     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 BSO.   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 |jm.   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 BSO are available.
However, a compilation of available data on particulate emissions and the
results of one test that shows the percent BSO 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 charge
34
29
17
0.11






Maximum
concentration
(ppm)




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

     With the assistance of the American Iron and Steel  Institute,  EPA has
compiled and analyzed data on particulate emissions from iron and steel
mills to assist in the definition of particulate 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 BSO 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 pg/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
per kilogram of coal charged (5.6 x lo"4 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

-------
oo
    CO
    ~^

    o
    a:
    UJ
    c/i
    CO
    o
LIJ

CJ
ca


OL_


LU
                                                  >20 PERCENT OPACITY
                                                  PERCENT DIFFERENCE
                                                    30          40           50


                                                              TIME, seconds
                          Figure 3-8.  Comparison of results between recording only visible emissions greater than 20 percent
                                         opacity and recording all visible emissions (U.S. Steel, Clairton).

-------
          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.   Applying the exponential model to the two
estimates of charging emissions yielded the results listed in Table 3-6.
     The results for Model A in Table 3-6 are based on measured emissions
that occurred throughout an average charge time of 5 minutes with 0.5 kilo-
gram of BSD per megagram of coal and 15 megagrams of coal per oven.  The
results for Model B are based on the test at J&L Steel; it is assumed that
visible emissions were occurring during the entire 3.5-minute sampling
period.  The two models predict that 8 to 30 grams of BSD escape when
20 seconds of visible emissions are observed.
     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.
     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.  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.
                                  3-29

-------
          TABLE 3-6.   ESTIMATES OF CONTROLLED CHARGING EMISSIONS

                                                      ,         Minimum
   Seconds of         Exponential  Aa     Exponential  B        estimatec
visible emissions     (g BSD/charge)     (g BSO/charge)      (g BSO/charge)
300
210
40
30
20
15
7,500
3,700
130
75
30
19
--
900
33
18
8
5
--
--
0.24
0.18
0.12
0.09
 Based on 0.5 kilogram per megagram of coal,  15 megagrams  of coal  per oven,
 and 5 minutes of emissions throughout the charge.
 Based on an estimated 900 grams of BSD from  3.5 minutes  of sampling.9
cBased on the measured rate of 0.021 kilogram of BSD per  hour (0.006 g/s)
 for a 1- to 2-meter plume.13
                                  3-30

-------
     The charging emission estimates listed in the third column of Table 3-6
are based on a topside leak mass emission rate of 0.006 gram of BSO per
second of visible emissions (0.021 kg/hr).   This emission rate is for a 1-
to 2-meter plume coming from a single hole in the charging port lid.13  The
minimum estimate should be applied only when all charging emissions origi-
nate from a single small gap.   Any specific application of the results in
Table 3-6 should consider whether light or heavy visible emissions are
observed.  The results in Table 3-6 are judged to be a reasonable range of
estimates for determining nationwide impacts from controlled charging
emissions.
     3.3.2.4  Factors Affecting Emissions from Charging.  Uncontrolled
emissions from charging cause large visible clouds of smoke and fumes.
Figure 3-9 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
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.

                                  3-31

-------
Figure 3-9. Uncontrolled charge.
              3-32

-------
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
lossl (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 3-10 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
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  3-11 shows an
outside view of a coke oven door.  More detail is provided in  Figure 3-12
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 3-13) around the periphery of the door.
It is termed "self-sealing" 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
                                  3-33

-------
              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
  c
  I
BOTTOM LATCH
                 HEARTH PLATE

      Figure 3-10. Vertical cross section of coke oven door on pusher side of oven.22
                                3-34

-------
       LATCH HOOK
SEAL ADJUSTING SCREW
                                          LEVELER OR "CHUCK" DOOR
                                          UPPER LATCH
                                           OVEN NUMBER
                                           PLATE BRACKET
                                           LOWER LATCH
    Figure 3-11. Outside elevation of a pusherside coke oven door.23
                          3-35

-------
LEVELER OR "CHUCK" DOOR
      SPRINGLOADED
       LOCKING BAR
              DOOR JAMB
                                                   REFRACTORY PLUG
       Figure 3-12.  Vertical cross section of a pusherside coke oven door.23
                                 3-36

-------
                 KNIFE EDGE SEAL
oo
 i
00
                                                                                              OVEN REFRACTORY
                   DOOR JAMB
                    BUCKSTAY
                                                                                   LATCH BAR
                                           Figure 3-13.  Horizontal cross section of a coke oven door.'
                                                                                                23

-------
                 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
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.
     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 3-13)  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
                                  3-38

-------
     The pusherside door has a small door at the top called the leveler or
chuck door (Figures 3-11 and 3-12).   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.
     3.3.3.2  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 3-9.2S  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
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 ROMs which have been identified in door leak
emissions are listed in Table 3-10.27

                                  3-39

-------
               TABLE 3-8.  MAJOR COMPONENTS OF COKE OVEN GAS25
                    Compound            Percent volume
H2
CH4
CO
C02
N2
02
C H
m n
H2S
NH3
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
               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-
                                  3-40

-------
           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
                                              ++
       1
     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
                                  3-41

-------
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.   BSO emissions during the test averaged 6 kg/hr from
an average of 31 leaking doors (70 PLD).  Emission factors for BSO 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
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
                                  3-42

-------
                  TABLE 3-11.   COKESIDE SHED TEST RESULTS
Kilograms
of BSO
Test per hour
Wisconsin Steel Shed30



Average
ARMCO, Inc. Shed12


Average
Bethlehem Steel, Burns Harbor10
Bethlehem Steel, hoods10
7.0
5.9
5.4
6.0
6.1
6.8
11
13
10.3
3.9
3.9
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 per
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.
GThis value is for toluene solubles from a pusherside hood over the door.
                                  3-43

-------
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-Controlled Door Leaks.   The emission rate
from two cokeside shed tests averaged 0.2 to 0.7 kilogram of BSD 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.
     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 BSD per hour at 70  PLD and
10.3 kilograms of BSD per hour at 29 PLD, respectively.   The results,
summarized in Table 3-12, are used to estimate emissions from well-controlled
doors.
                                  3-44

-------
          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
aFor comparison, the measured BSD rate for a 1- to 2-meter plume from one
 gap was 0.021 kg/hr.13
 Average results from the cokeside shed tests.
                                  3-45

-------
     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
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 well-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 3-14 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.

                                  3-46

-------
   60
   50
O  40
 CM
X
LLJ
tc  30

C/J
CO
UJ
cc
O_

   20
   10
                                        10
40
                            20         30

                      TIME (MINUTES)

             (scale is linear with square root of time)


Figure 3-14. The effect of time on the internal oven pressure.34
50
70

-------
     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
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.
                                  3-48

-------
     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:

     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.
                                  3-49

-------
     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.
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

     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
                                  3-50

-------
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
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
                                  3-51

-------
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
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 I by
EPA in August 1978.l3  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
                                  3-52

-------
                 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. 00022°
0.021
Small                               0.0017
                                   0.0029
                                   0.0053                        0.000072°
  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.
 These values are 1.4 to 1.8 percent of the BSO.
                                  3-53

-------
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.
3.4  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, but do not set a performance
level.  This section summarizes existing regulations for wet-coal  charging,
door leaks, and topside leaks.
     Existing regulations vary from stringent emission limits for new and
some existing batteries to no emission limits for other batteries.   There-
fore, a variable baseline is defined and described in this section.   The
variable baseline  is expressed as the number of batteries and the percent
of the total U.S.  coke capacity that must meet specific existing emission
limits.  By-product coke ovens are currently operated in the 17 States
shown in Table 3-15.   Three major steel producing States (Pennsylvania,
Ohio, and Indiana) account for over 55 percent of the coke capacity.
All coke-producing States have coke ovens that are currently covered  by
existing State regulations or consent decrees with the exception of charging
emissions in Alabama.   State regulations range from 15 to 34 s/charge, 10 to
16 PLD, 2 to 5 PLL, and 5 to 10 PLO.
     Table 3-16 lists charging consent decrees that were negotiated on a
plant-by-plant basis.   The most stringent limit, 55 seconds for 5 charges
has been applied generally to new or rebuilt batteries.   Consent decrees
for door leaks, listed in Table 3-17, range from 4 to 15 PLD.   The 4  to
5 PLD level has been applied only to new or rebuilt batteries.   Consent
decrees for topside leaks are listed in Table 3-18.   The levels of 1  PLL
and 4 PLO have been applied only to new or rebuilt batteries.   Consent
decrees for existing batteries range from 2 to 5 PLL and 5 to 10 PLO.

                                  3-54

-------
                                 TABLE 3-15.  STATE REGULATIONS FOR COKE  OVEN  EMISSIONS
co
i
en
en

Charging
State (seconds/number of charges)
Alabama
Cal i form' a
Colorado
11 1 i no is
Indiana
Kentucky
Maryland
Michigan
Missouri
New York
Ohio
Pennsylvania
Tennessee
Texas
Utah
Wisconsin
West Virginia
--
60/4
b
125/5C
125/5C
b
160/5
80/4
120/6
150/5
170/5
75/4
b
b
b
125/5
b


Doors
15a
10
b
10
10d
b
10
10-12e'f
159
10f
16
10h
b
b
b
10 1
b
Percent maximum

Lids
5
3
b
5
3
b
3
4
2
2
5
2
b
b
b
51
b
leaking

Offtakes
10
10
b
10
10
b
10
4
10
5
10
5
b
b
b
101
b
.Maintenance and inventory of spare doors required.
cConsent decree in effect for batteries in this State.
.Excludes one observation in
Excludes four doors.
.Jen PLD for doors less than
a total of 20.

or equal to 5 m in height,


12 PLD for doors


greater than 5
Chuck doors are counted as separate doors.
^Pusherside doors only (cokeside shed).
-Excludes two doors, counts
Percent on operating ovens
al 1 door area leaks.
only.






m in height.




-------
                                TABLE 3-16.  CONSENT DECREES FOR CHARGING
      Limit
    Company/location
      Battery
   Date
55 s/5 charges CF&I , Pueblo, CO
U.S. Steel, Fairfield, AL

Republic Steel, Chicago, IL
National Steel, Granite City, IL

U.S. Steel, Clairton, PA



Allied Chemical, Ashland, KY
n _L i_ "i L f j_ T f* r\ • j_ nrv
— 	 bet.nienem bt.ee I, oparrows roint, riu
^~~~- 	 	 	
B, C, D
2
9
New
A
B
13, 14
15
20
B, C
4
A
	 	
1960-1974
New (1978)
Rebuild (1979)
Under construction
New (1979)
New (1980)
Rebuild (1980)
Rebuild (1979)
Rebuild (1978)
Proposed new
Rebuild (1978)
	 Proposed_new
 60  s/5  charges

~84  s/7  charges


 75  s/4  charges


 137 s/7 charges

 100 s/5 charges

 200 s/10 charges

 125 s/5 charges
Lone Star Steel,  Lone Star,  TX

Republic Steel,  Cleveland,  OH
Republic Steel,  Warren,  OH

U.S. Steel, Clairton, PA
Chattanooga Coke,  TN

U.S. Steel, Geneva,  UT

Republic Steel,  Youngstown,  OH

Republic Steel,  Cleveland,  OH

National Steel,  Granite  City, IL
Interlake, South Chicago,  IL
Bethlehem Steel,  Burns Harbor, IN
J&L Steel, Indiana Harbor,  IN
U.S. Steel, Gary,  IN
Bethlehem Steel,  Lackawanna, NY
                          New (1979)
         1
         4
1-3, 7-12, 16-19, 21
       1, 2

    1, 2, 3, 4

       B, C

       6, 7

         C
       1, 2
       1, 2
       4, 9
1, 5, 7, 13, 15, 16
      7, 8, 9
1976
New (1979)

1946-1964
1918-1941

Rebuild (1980)

1950-1962

1952

1961
1956-1957
1969-1972
1956-1961
1950-1970
1952-1970
                                                                                         (continued)

-------
                                                 TABLE 3-16 (continued)
Limit
100 s/4 charges
120 s/hour
200 s/5 charges
Company/location
Wheel ing- Pittsburgh, WV
National Steel, Weirton, WV
Allied Chemical, Ashland, KY
Republic Steel, Cleveland, OH
Battery Date
1, 2, 3, 8 1953-1977
1, 8 1974-1979
3 1953
2, 3, 4 1957-1958
en
—i

-------
                                       TABLE  3-17.   CONSENT  DECREES FOR DOOR LEAKS
CO
I
en
00
        Percent leaking doors
                  10
                  15
          Company/location

-Lane_S£ar_Steel ,_IX_
 U.S.  Steel,  Clairton,  PA
 U.S.  Steel,  Fairfield,  AL   -
 Bethlehem  Steel,  Sparrows  Point,  MD

 CF&I,  Pueblo,  CO

 Allied Chemical,  Ashland,  KY
 U.S.  Steel,  Fairfield,  AL
 Wheeling-Pittsburgh,  WV
 National  Steel, Weirton,  WV

 Republic  Steel, Cleveland, OH
     Battery
                                                                                    13b, 14b, 15b  20b, Bc, Cc

                                                                                                9c
                                                                                                AC

                                                                                              B, C, D
s- ' ~ Q
,~f
Chattanooga Coke, TN
Republic Steel, Youngstown, OH
Republic Steel, Cleveland, OH
Republic Steel, Warren, OH
U.S. Steel , Geneva, UT
ARMCO, Inc. , Houston, TX
1, 2
B, C
J"
1, 2, 3,
1, 2
4
-~->
-3, 7-12, 16-19, 21
    1, 2, 3, 8
       1, 8

       6, 7
         Texas construction permit.

        DNew or rebuilt battery.

        "Proposed new battery.

         Consent decree states  <  7 PLD except for three doors.

        'Pusherside doors only  (cokeside shed).

         Consent decree states  <  10 PLD except for two doors.

-------
                                    TABLE  3-18.   CONSENT  DECREES  FOR TOPSIDE LEAKS
(Jl
10
Percent leaking
lids
1
£-7
2
2
2
,c
V 3
5
5
Percent leaking
offtakes
4
4C
5
6
10
8C
5
10
Company/location
U.S. Steel , Clairton, PA
U.S. Steel, Fairfield, AL
Bethlehem Steel , Sparrows. Point, MD-
Lone Star Steel , TX
U.S. Steel, Clairton, PA
Republic Steel, Cleveland, OH
U.S. Steel, Fairfield, AL
Republic Steel, Warren, OH
U.S. Steel, Geneva, UT
CF&I, Pueblo, CO
Wheeling-Pittsburgh, WV
Chattanooga Coke, TN
Republic Steel, Youngstown, OH
National Steel, Weirton, WV
ARMCO,. I nc.._,_ Houston, TX
Republic Steel, Cleveland, OH
Allied Chemical, Ashland, KY
Battery
13a, 14a, 15a. 20a, Bb, Cb
9h
Ab
\
Ca J
1-3, 7-12, 16-19, 21
;•
1, 2, 3, 4
B, C, D
1, 2, 3, 8
1, 2
B, C
1, 8
__~LJL^
6, 7
3, 4
         New  or  rebuilt  battery.
         Proposed  new  battery.
        cMaximum of  5  leaking  lids  and  5  leaking  offtakes.

-------
     Tables 3-19 through 3-22 summarize the existing baseline conditions
from State regulations and consent decrees.  For charging, 159 batteries
with 94 percent of the total coke capacity have a visible emission limit
that averages up to 34 s/charge.   A total  of 18 batteries with 6 percent of
the coke capacity currently have no applicable seconds per charge limit.   A
total of 105 batteries with 64 percent of the coke capacity must meet a
door leak limit of 12 PLD or less.   For lid leaks, 111 batteries with
66 percent of the capacity have an applicable regulation of 3 PLL or less.
For offtake leaks, 71 batteries with 44 percent of the capacity have an
applicable regulation of 6 PLO or less.
     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.

                                3-60

-------
         TABLE 3-19.   SUMMARY OF  BASELINE  REGULATIONS  FOR  CHARGING
Charging limit
(average s/charge)
No limit
30-34
25
19-20
15
11-12
Number of
batteries
18
40
37
60
7
15
Percent of
capacity
6.1
21.2
26.0
33.4
2.9
10.4
        TABLE 3-20.   SUMMARY OF BASELINE  REGULATIONS  FOR  DOOR  LEAKS
Percent leaking doors       Number of batteries        Percent  of  capacity
        14-16                       72                      36.1
         -12                   '     54                      34.1
         10                         42                      24.2
          8                          3                      1.7
         4-5                         6                      3.8
        TABLE 3-21.   SUMMARY OF BASELINE  REGULATIONS  FOR  LID  LEAKS
Percent leaking lids       Number of batteries        Percent  of  capacity
          5                         58                     27.2
          4                          8                       6.5
          3                         36                     21.6
          2                         70                     41.8
          1                          5                       2.8
                                 3-61

-------
      TABLE 3-22.   SUMMARY OF BASELINE REGULATIONS FOR OFFTAKE LEAKS



Percent leaking offtakes      Number of batteries       Percent of capacity





         10                        106                      55.7





          6                          3                       1.7





          5                         54                      32.3





          4                         14                      10.3
                                 3-62

-------
     When the OSHA requirements are implemented by all batteries, a substan-

tial degree of control will be established for all of the coke oven sources.

However, OSHA does not enforce a performance level or visible emission

limit.   Therefore, the batteries which are listed in Table 3-19 and have no

visible emission limits for charging must have the OSHA equipment and work

practice controls, but the resulting level of emission control is not

specified.

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, and 1978.

 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.


                                 3-63

-------
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.

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.
                                   3-64

-------
26.


27.



28.


29.



30.



31.



32.



33.


34.



35.


36.


37.


38.
Barrett, R. E., et al.   Sampling and Analysis of Coke-Oven Door  Emis-
sions.  Publication No. EPA-600/2-77-213.  October 1977.

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

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.

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

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.

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.

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).

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.
Graham, J. P. and B. P. Kirk.
Control.  The Metals Society.
Problems of Coke-Oven Air Pollution
London SWIY 5DB.   p.  82-100.
Lowry, H. H. (ed.).  Chemistry of Coal Utilization.  Supplementary
Volume.  John Wiley and Sons, Inc.  NY.  1963.

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.
                                  3-65

-------
                     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
tunnelhead 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

-------
                                 MCM
                                 rmw
              i«  r ~ "w'iii'i    I ^
              S?LJ  ST*« CHARGING  I l^gl
               rOSITIONING CAR PRIOR TO CHARGING
                (Mil .jrpng inwunu of cojl in hcppin
                                  SMte&HJa
                                   ~>r—YT7Ji=}'L-"".'". '««««,•."" •• ,:.~] Jk=i;


                                                STEP No.2
                                              STAGE CHARGING
                                           DISCHARGE Hi 11 « HOPPERS
^jj-^Jx^UU^ ^Ip
1TN iL^dLTSas^i^iUyi
                ;; I    STEP No 3
                 1 |   STAGE CHARGING

                  DISCHARGE No 1 HOPPER
                                                 rrn
                                          ^-.  -•..  "_^aj
                                            DISCHARGE Ni.2 HOPPCR
                                             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

-------
             — .-t^>-  —  -^Piping for steam and/or
             „— ^j-. —xT  weak liquor.
           Nozzles
     Return bend -
    Collector
      main
              	til
                   Door
	 V-





^
tj
»** ,~.
ra c
u o
Q -rH
•a «»
0 -•<
3 a.
•o in
c <
h-l ^"
••Ascension pipe




Charging
_ hole
Oven
Figure 4-2.  A typical aspiration system operating in the return bend of the
                            gas collection duct.

-------
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




 10QO
1 900
^.800
>
  700
c;
o


1  600
   400




   300




   200
                               NOZZLE PRESSURE, kg/cm2, gauge

                                 3456
                   20
40            BO           80

    NOZZLE PRESSURE, Ib/in2. gauge
100
                                                           30
                                                           25  I
                                                               E
                                                               ^
                                                               t;

                                                               >

                                                           20  G
                                                               u
                                                               z
                                                               o
                                                                                           15
                                                           10
                                                                                               o.
                                                                                               V)
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 RiN<
               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 aspiration systems
                                  4-9

-------
-pi

h-1
o
                              1001—
                               90
                            =  80
                            =  70
                            c=
                            •s
                               60
C-
o

x  50}--
o    '
                               40
                               30
                                20
                                         10       20        30       40       50       60


                                                                MAXIMUM AIRFLOW, percent
                                                                70
80
90
100
                                             Figure 4-5.  Air flow gooseneck area open in test ovens.
                                                                                                   12

-------
                         OVEN (IN SECTION)
collector
main
 X^.
loading oven from coa
 high pressure
steam aspirator
                         steam suction puffing
                           coke oven gases
                      \    from baking coal
                                    -,.,. tf^y.
        Figure 4-6.  Aspiration of gases from a coke oven into the gas collecting mains.

-------
(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
w
|3.5

"a

23.0
60
C
•* 9
oo «• •
§2.0
  1.5
           Stage Charging
                   AISI/EPA Car(Sequential  Charging
                       o Conventional Charging
                 .  \        . 1 ,  ... I
     10           15       20      25    30   35  40

            Weight of Coal per  Charge, tonnes

     •   .   .  .   I  .  .  .. I  .... i .... I ,,,,!,.. 1,,,.
     11         15       20      25    30   35  40
          Weight of Coal per Charge, net tons
    Figure 4-7.  Variation in the time required to charge coal.
                        4-14

-------
     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 w.ill 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 0.
     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, excjept 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
1
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
11.6
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 & Laughlin (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 lx

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 -
o -
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:
                                             y
                              Log average = e -1
where:
              v _ In (X, + 1) + In (X, + 1) +.  .  .  In (Xn + 1)
              Y *~ 	1.	c. 	               n       j
                                       n
              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
Arithmetic
average
Company Battery (s/charge)
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
Bethlehem Steel, Burns Harbor
U.S. Steel, Clairton3

U.S. Steel , Gary
National Steel, Weirton
U.S. Steel, Clairton3
3These batteries used the opti
16
17
9
B
I
22
3
P2
PA
10
19
4
6
3
7
21
9
cb
2b
2
8
20
lb
lb
11
mization
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 cl
Log 95-percent level
average (log average,
(s/charge) s/charge)
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
harginq 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.
DTall  (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 representa-
tive 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 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.
     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 charg-
ing 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 significantly increas-
ing 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 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
                                 4-29

-------
TABLE 4-5.   COMPARISON OF ARITHMETIC AVERAGE AND LOG AVERAGE
Example (s/charge)











Arithmetic
average =
Log
average =
1
16
16
16
16
16
16
16
16
16
16
16.0
16.0
2
6
16
16
16
16
16
16
16
16
40
17.4
16.0
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

-------
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 pulveriza-
tion.  (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 pulverization has an
effect, it is overshadowed by other process or operating variables which
more significantly affect the performance of stage charging.
     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
                                 4-31

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

Rosedale0
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.

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

-------
     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 pulverization
would be the charging of run-of-the-mine (nonpulverized) coal.   Results
from U.S. Steel Batteries 16 and 17 at Clairton show that nonpulverized
coal has no detectable adverse effect on charging emissions.  These batter-
ies 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 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 mechan-
ical feed requires about 6 min.   The factors mentioned 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
                                 4-33

-------
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 signif-
icant 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 observa-
tion 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 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 general-
ly 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 propor-
tionately.   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
                                 4-34

-------
Figure 4-9. A "puff" of emissions during stage charging.
                     4-35

-------
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 developed.   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 sequen-
tial 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.
     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
                                 4-36

-------
                Figure 4-10. AISI/EPA larry car coke oven charging system.29
B.
C.
D.
E.
F.
G.
H.
Leveler bar
UHF alignment device
Ascension-pipe damper  actuator
Gooseneck cleaner
Ascension-pipe actuator
Lid lifters
Feed hopper drop sleeves
Leveler-door operator
J.  Leveler-door  smoke  seal
K.  Butterfly valves
L.  Controlled environment cab
M.  Coal hopper
X.  Oven to be charged
Y.  Oven to be pushed
Z.  Oven to be dampered  off
                                      4-37

-------
                                                     ASCENSION
                                                     PIPC • CL60W
                                                                IAMMAMV
                                                                WTERlOCK

Figure 4-11.  Flow of gases in the AISI/EPA sequential charging system
                                                                      29
                                  4-38

-------
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 follow-
ing 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.
     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.
                                 4-39

-------
     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 the
AISI/EPA larry car described in Section 4.1.2.  The average result of the
tests is 120 grams of particulate 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 measure-
ments, "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 conditions, 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
                                 4-40

-------
     However, these results do not represent the true potential of sequen-
tial 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 sequen-
tial 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 increas-
ing 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 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.
                                 4-41

-------
                                                     Exhoust
Figure 4-12. Representative scrubber system on a larry car.
                          4-42

-------
     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 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,"
                                 4-43

-------
 Combustion •*•—
   Gases
Igniter CO
     T][~y-.Air Ports
          Gas Port
                        jk
        Coke- Oven'
            Gases
Coal
ill
           \
^Coke Oven
   Gases
 Figure 4-13.  Details of a shrouded drop sleeve.
                                       34
                  4-44

-------
          "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 effective-
ness 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.
                                 4-45

-------
    TABLE 4-7.  PARTICULATE 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.
                                  4-46

-------
     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.
     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
                                 4-47

-------
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
air stream.   A suction hood above each door with a wet electrostatic precipi-
tator 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, replace-
ment, 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
                                 4-48

-------
is expressed in terms of visible emissions as percent leaking doors (PLD).
A discussion of this test procedure is included in Section 3.4.1, and the
test method 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 con-
tain steam, volatile oils, and tars pass through small  defects in the seal-
ing 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
     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
                                 4-49

-------
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.
     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.
                                 4-50

-------
r
en
                                                            Figure 4-14. Koopers door.42
                            Copyright 1971 by United States Steel Corporation.

-------
              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,
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
                                 4-52

-------
I
en
Go
                                                                                                JAMB BRICK
                                                                                                REPLACEABLE
                                                                                                FUSED SILICA
                                                           Figure 4-15.  Wilputte door.

                              Copyright 1971 by United States Steel Corporation.
                                                                                       42
                                                                                                                  RENEWABLE
                                                                                                                  STEEL SEAL
                                                                                                                  DIAPHRAGM

-------
     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
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-
                                 4-54

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

-------
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.
     A few coke oven doors have been observed to be completely free of
visible emissions during the entire process.39  One hundred percent control
                                 4-56

-------
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.
Uneven stress distributions may be caused by the forces from adjusting
                                 4-57

-------
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 PLO.   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 Laughlin 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
              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
Figure 4-18. Ikio door.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  Battelle 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:
1
0
, 1 ,
1
Inches
|
2
                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 1
Sealing System 2
                                      Y/////Y7
                                      i /
Watt 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 3^
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

-------
Figure 4-21. Coke oven door hood that assists in controlling
                  pushing emissions.

                         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

-------
                                      CLEANER
i
^j
en
                                Figure 4-23. Channel for the passage of gas formed by a door hinged to a buckstay.73

-------
     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 ftVmin).   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 BSO emission factor was about 0.25 kg of BSD per megagram
of coal charged.77
     4.2.3.6  Vented Plug.S4 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
Latch Bar
                                   Gasses Vented to
                                   Top of Oven
                                      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

-------
I
oo
                 Bucksta
                 Mount
                                            Slots
  Seal
Surface

Tack
Weld
                                                                        Closed
                                                                        Latch Mooks
                                Figure 4-25. Slotted jamb design.

-------
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 DATA1
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

lb

21
C

22
19

9C
B

20
16
10
D

8
17

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 U.S. Steel, Clairton have implemented the
modified seal technology.

A 6-meter battery without the modified seal technology.

A rebuilt battery without 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 modi-
fied 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 clean-
ing 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.
     Ten of the U.S.  Steel  and CF&I batteries averaged 4  PLD or less with a
95-percent confidence level  of 7 PLD.   In the data base,  the worst of the

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

-------
batteries with modified seal technology is Battery 9 at U.S.  Steel, Clairton
with an average of 9 PLD and a 95-percent confidence level  of 12 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 proce-
dures 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 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.
                                 4-86

-------
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 re-
quired 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
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
                                 4-87

-------
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 fTow 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 7 lidsmen 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, meas-
urement of visible emissions is possible and provides a good indicator of

                                 4-88

-------
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 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
                                 4-89

-------
      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
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).
                                4-90

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





October, 1977






November, 1977






Oecember. 1977






January, 1978






February, 1978






3a:tery
12A
15
15
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
L5
16
17
19
21
22
12A
IS
16
17
19
21
22
Number ofd
Traverses
26
25
25
25
26
26
26
26
29
29
28
29
28
28
21
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.
Observed Per
Rawe
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
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
- 0.6
- 1.2
- 1.2
- 0.3
- 1.1
- 1.5
- 1.1
- 0.6
- 0.2
- 0.3
- 0.8
- 1.1
- 0.9
- 1.1
- 0.5
- 1.2
- 1.2
- 0.8
- 0.5
- 0.6
- 0.5
- 0.9
- 0.4
- 0.8
- 0.4
- 0.9
- 1.2
- 0.9
- 3.0
- 1.2
- 1.2
- 2.0
- 2.6
- 3.5
- 2.3
- 1.8
- 2.9
- :.2
- 3.3
- 1.7
- 2.2
- 0.9
lesks
Travorseb
Average
0.1
0.2
0.2
0.2
0.3
0.3
0.5 .
0.1
0.2
0.1
•o.i
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
O.S
0.3
0.5
0.3
0.6
0.3
    d A traverse is-one recording of leaks frox an entire battery.

    b Percentage of total  potential leaks.
                                  4-91

-------
                TABLE  4-14.   VISIBLE EMISSION  DATA ON OFFTAKE LEAKS,  KAISER  STEEL,  FONTANA86
Percent leaks for each battery
L-jtimj*
manpower Date
1 April 1977
May 197V
June 1977
April - June 1977
2 July 1977
f" August 1977
(JO
ro September 1977
October 1977
November 1977
December 1977
July - December 1977

Aver-
age
6.1
4.3
4.4
4.9
2.7
2.2
2.0
2.9
3.4
4.1
2.9
A
Percent of
readings
at 5 or
less
36
69
57
54
78
91
91
82
;71
57
78

Aver-
age
5.1
4.0
3.8
4.3
2.9
2.1
1.2
2.1
4.2
4.4
2.8
B
Percent of
readings
at 5 or
less
53
69
74
65
76
86
100
96
57
57
79

Aver-
age
12. 0
11.4
8.9
10.8
4.7
3.8
4.1
5.3
5.8
4.3
4.7
C
Percent of
readings
at 5 or
less
16
14
29
20
56
72
68
54
43
57
' 58

Aver-
age
4.7
4.4
3.0
4.0
1.8
2.2
1.9
1.8
1.8
2.3
2.0
0
Percent of
readings
at 5 or
less
58
62
88
69
96
90
97
93
100
66
94

Aver-
age
5.3
5.6
2.7
4.5
1.9
2.0
1.6
2.0
2.6
2.2
2.1
E
Percent of
readings
at 5 or
less
40
42
86
56
95
95
97
89
86
100
94

Aver-
age
4.9
5.3
7.0
5.7
4.3
3.4
2.8
2.7
3.8
4.8
3.6
f
Percent of
readings
at 5 or
less
56
56.
28
47
58
76
87
86
79
43
72

Aver-
age
7.1
7.6
8.2
7.6
3.9
3.4
2.4
5.1
4.3'
3.4
3.8
G
Percent ol
readings
at 5 or
less
22
23
21
22
73
81
87
46
64
86
73
Approximate employees per shift per battery responsible for luting and tending lids.

-------
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 con-
trolled by frequent scrutiny of potential  sources of leaks and prompt
sealing of the leaks when they appear.
     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 aver-
aged 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
                                 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
Range
PLL
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0.4 -
1.2 -
0 -
o -
0 -
0 -
0 -
0.4 -
0 -
0 -
0.4 -
-
o -
-
o -
0 -
-
o -
o -
0 -
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
Number
of
obser-
vations
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
Range
PLL
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0 -
0 -
o -
0 -
0 -
0 -
0 -
o -
-
-
-
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
Number
of
obser-
vations
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 ob-
Range serva-
PLO tions
0 -
0 -
0 -
0.8 -
0 -
0.8 -
0 -
0 -
o -
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

-------
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 confi-

dence level of 6 PLO when averaged over three runs.

     Leaks in battery mains on a well maintained battery will occur infre-

quently.  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.

2.   Barnes, T. M., et al.  Control of Coke Oven Emissions, Ironmaking, and
     Steelmaking.  1975.   p.  165.

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.
                                 4-97

-------
12.   Munson, 0. 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.

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.
                                 4-98

-------
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.

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.
                                 4-99

-------
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.

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.
                                 4-100

-------
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.

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.
                                4-101

-------
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-1iquour
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-sealing 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-seal ing 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 battery
2
Metallurgical
18
15-118 (62)
3-5 (4.0)
0.44

3
Metal lurgical
18
56-87 (71)
(6.0)
0.47
Battery dimensions:
  Length  (m)
  Width (m)
  Height6 (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 exemplary
batteries.  A summary of the regulatory alternatives is presented 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 plant-by-plant basis.
     The visible emission limits associated with this regulatory
alternative correspond to the baseline visible emission limits that
were presented in Tables 3-19 through 3-22 (Chapter 3).   For wet-coal
charging, approximately 159 batteries (94 percent of the total  U.S.
coke capacity) have average visible emission limits of 11 to 34 s/charge.
Approximately 18 batteries have no visible emission limits (seconds
per charge) that are currently applicable to the wet-coal charging
operation.  All coke oven batteries in the United States currently
have applicable visible emission limits for door, lid, and offtake
leaks.   These limits range from 4 to 16 PLD, 1 to 5 PLL, and 4 to
10 PLO.
     Although some batteries have no current visible emission limits
for wet-coal charging, OSHA regulations apply to all batteries.1  The
OSHA regulations 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
                                  6-3

-------
                TABLE 6-2.   SUMMARY OF REGULATORY ALTERNATIVES
Emission
point
Charging
Regulatory
alternative Technology
I Stage charging
II Optimized stage charging
III Exemplary batteries (<5 m)
Exemplary batteries (>5 m)
Emission
limit
11 to 34 s/charge
16 s/charge
8 s/charge
15 s/charge
Door leaks
  I      Leak control  program
 II      Modified seals (or manpower
           for luted doors)
III      Exemplary batteries (<5 m)
         Exemplary batteries (>5 m)
PLD = percent leaking doors.
PLL = percent leaking lids.
PLO = percent leaking offtakes.

aLog average of 10 charges.
4 to 16 PLD

 12 PLD
  7 PLD
 10 PLD
Topside leaks
I
II
III
Leak control program
Luting manpower, modification
of some offtakes
Exemplary batteries
1 to 5 PLL,
4 to 10 PLO
3 PLL, 6 PLO
I PLL, 4 PLO
                                    6-4

-------
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 s/charge at the 95-percent confidence level when averaged over
10 charges.   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 probability 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 s/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; 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
                                  6-5

-------
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 s/charge.   It is
estimated that 82 batteries (47 percent of the total  U.S.  coke capacity)
have current regulations that are as  stringent as the log average
limit, and 95 batteries have existing regulations that are less strin-
gent.
     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.  Table 4-10 (Chapter 4) shows that
at these plants, the battery with the highest average of door leaks
(Battery 9 at Clairton) averaged 9 PLD with a 95-percent confidence
level of 12 PLD 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 and National Steel, Weirton).
     A total  of 105 batteries have current regulations that are at
least as stringent as 12 PLD, and approximately 72 batteries have
regulations that are less stringent.   Exemption of the doors on the
last oven charged in the Pennsylvania standard and other similar
regulations increased the effective PLD limits from 10 PLD to approxi-
mately 12 PLD.
     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
                                  6-6

-------
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.
     An estimated 111 batteries (about 66 percent of the total U.S.
coke capacity) have an existing lid leak limit of 3 PLL or less.   For
offtake leaks, 71 batteries (about 44 percent of the total U.S.  coke
capacity) have existing regulations of 6 PLO or less.   The limits of
3 PLL and 6 PLO would affect lid leaks for about 66 batteries and
offtake leaks for about 106 batteries.
     6.2.1.3  Regulatory Alternative III.  Regulatory Alternative III
was developed as a more stringent alternative than Regulatory Alterna-
tive II.   This alternative would possibly require technology development
because control technologies that are more effective than those previ-
ously discussed have not been demonstrated.   The emission limits  for
Regulatory Alternative III were derived from the best control performance
observed at at least two batteries for each emission point.
     Regulatory Alternative III for charging would impose log average
limits of 8 s/charge for short batteries (oven height of less than
5 meters) and 15 s/charge for tall batteries (oven height of 5 meters
or greater) from 10 charging observations.   The short battery limit is
derived from data in Table 4-4 and the performance of Battery 9 at
U.S. Steel, Fairfield (rebuilt in 1979,  single main),  Battery B at
CF&I (built in 1972, double main), and Battery 1 at U.S. Steel,  Clairton
(built in 1955, double main).   The tall  battery limit is based on the
performance of Battery C at Lone Star Steel  (built in 1979,  double
main) and Battery 1 at U.S.  Steel, Gary (built in 1970, double main).
     Regulatory Alternative III for door leaks would establish a  limit
of 7 PLD for short batteries and 10 PLD for tall batteries.   The  short
battery limit is based on a total of ten batteries in Table  4-10
(Chapter 4) that averaged 4 PLD or less with a 95-percent confidence
level of 7 PLD from a minimum of three runs.  Tall battery performance
is characterized by Battery 1 at National Steel, Weirton with an
average of 2.9 PLD and by Battery 2 at U.S.  Steel, Fairfield which
averaged 7 PLD with a 9 5 percent confidence level of 10 PLD.

                                  6-7

-------
     Regulatory Alternative III for topside leaks would establish
limits of 1 PLL and 4 PLO.   The lid leak limit was derived from Table
4-15 which shows that 14 batteries averaged 0.2 PLL or less during at
least one inspection with a 95-percent confidence level of 1 PLL
(averaged over three runs).   The offtake leak limit was derived from
Table 4-16 which shows that six batteries averaged 0.9 PLO or less
with a 95-percent confidence level of 4 PLO (averaged over three
runs).
6.2.2  Alternative Control  Technologies
     Some technologies which 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 preclude 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
                                 6-8

-------
expensive than modified seals coupled with the door leak control
program, and sheds have not been shown to be significantly more effec-
tive.   However, when sheds are used as a pushing control device, door
leak emissions from the coke side of the battery can be collected and
removed.  Sheds with good capture and control efficiency could provide
a level of door leak control comparable to modified seals and a door
leak control program.
     Separate PLD limits were developed for batteries equipped with
cokeside sheds because cokeside door leaks can be controlled by sheds.
The intent was to limit the estimated mass emissions (based on the BSO
removal efficiency of the shed's control device) from the shed's
control device to the estimated mass emissions from a battery without
a shed that is controlled to the PLD levels required by the regulatory
alternatives.   The mathematical model discussed in Chapter 3 was used
to estimate the new PLD levels for doors controlled by cokeside sheds
of varying BSO removal efficiencies.  Table 6-3 lists the results of
this model.
6.3  REFERENCES
1.   41 Federal Register 46785, 46786.  October 22, 1976.
                                  6-9

-------
     TABLE 6-3.   REGULATORY ALTERNATIVES FOR DOOR LEAKS FOR BATTERIES
          WITH COKESIDE SHEDS OF VARYING BSO REMOVAL EFFICIENCIES

 BSD removal       Regulatory Alternative II     Regulatory Alternative III
efficiency (%)           (PLD limit)3                  (PLD limit)0
20
30
40
50
60
70
80
90
95
96
97
98
99
99.5
13
14
15
16
17
20
23
30
40
44
49
57
76
100
7.7
8.1
8.6
9.2
10
11
13
18
23
25
28
33
44
58
PLD = percent leaking doors.

aE
br
aBased on 12 PLD without a shed.
 Based on 7 PLD without a shed.
                                  6-10

-------
                           7.   ENVIRONMENTAL IMPACT

     This chapter presents environmental impacts that will result from each
regulatory alternative applied to the model  batteries presented in Chapter 6.
Total nationwide impacts are also presented.   Primary impacts (those directly
attributed to operation of an air pollution  control system) and secondary
impacts (indirect occurrences such as emissions that result from generation
of power to operate a control  system) are included.  Average benzene soluble
organics (BSO) concentrations that were determined by atmospheric disper-
sion modeling are given for a model battery.   Other impacts that are dis-
cussed include water pollution, solid waste  disposal, and energy consumption.
7.1  AIR POLLUTION IMPACT
7.1.1  Primary Air Impact for Model Batteries
     The emission levels associated with Regulatory Alternative I correspond
to the baseline emission levels required by  existing SIPs and consent
decrees that were discussed in Chapter 3.  A variable baseline was defined
in Chapter 3 because emission limits of varying stringency are in effect at
different coke oven plants.   However, for the purposes of evaluating the
relative impact of each regulatory alternative, the least stringent of the
State regulations was used to determine baseline emissions for the model
batteries.   Emissions for Regulatory Alternatives II and III were estimated
from the emission limits given in Chapter 6.   Because the least stringent
existing regulation was used to determine the baseline emission levels,  the
emission reduction impacts presented are the maximum impacts that would  be
expected at a battery (except for reductions in charging emissions at the
18 batteries in Alabama that are not currently covered by standards for
charging emissions—see Chapter 3).  Many batteries have existing regu-
lations that are at least as stringent as Regulatory Alternative II; there-
fore, Regulatory Alternative II would not result in emission reductions  for
                                  7-1

-------
those batteries.   Total nationwide emissions, which are presented in
Section 7.1.2, include emission reductions for only those batteries that
will be required to improve emission control.
     Visible emission limits for the regulatory alternatives were used to
estimate mass emissions.   To comply consistently with a limit, a battery
must maintain an average control performance that is lower than the limit.
For example, an average performance of 9 PLD yielded a 95-percent confidence
level of 12 PLD,  and a log average of 11 s/charge resulted in a 95-percent
confidence level  of 16 s/charge.  A battery with an average control perfor-
mance at the limit level  would frequently be out of compliance with the
limit.   Therefore, a factor of 75 percent was applied to each emission
limit to estimate an average performance level.
     A range of mass emission estimates from controlled charging was provided
in Section 3.3.2.3 for various levels of visible emissions.   The lower
bound estimate was based on the extreme case in which the charging emissions
escape from a single opening.  The upper bound estimate was derived by
applying an exponential model to uncontrolled emission estimates.   This
range of estimates was applied to the model batteries to obtain the BSO
emission estimates presented in Table 7-1.

                  TABLE 7-1.  AIR EMISSIONS FROM CHARGING
Regulatory
alternative
I
IIb
IIIb
Model
battery
1
2
3
1
2
3
1
2
3
Seconds
per charge
34
34
34
16
16
16
8
8
15
BSO emissions3
(Mg/yr)
0.002 to 0.6
0.005 to 1.7
0.006 to 1.9
0.001 to 0.22
0.003 to 0.63
0.003 to 0.73
0.0006 to 0.08
0.002 to 0.24
0.003 to 0.66
BSO emission reduction
over baseline (Mg/yr)
0
0
0
0.001 to 0.38
0.002 to 1.1
0.003 to 1.2
0.0014 to 0.52
0.003 to 1.5
0.003 to 1.2
aPer battery.
 Log average limits; for estimating emissions, 5 seconds were added to
 approximate an arithmetic average.
                                  7-2

-------
     The BSO emissions for the model batteries were calculated from the
number of ovens, the cycle time, the visible emission limit, and the esti-
mated range of grams of BSO per charge for a particular visible emission
level.   For example, a visible emission limit of 34 s/charge would require
an estimated average performance of 26 s/charge.   The emission estimate
from Section 3.3.2.3 is 0.16 to 56 grams of BSO in 26 seconds of visible
emissions.   This range is applied in the following sample calculation for
Model Battery 2 which has 62 ovens, an 18-hour cycle time, and an existing
regulation of 34 seconds of visible emissions per charge.
     Sample calculation:
                                                 = <>•«» •«.
An upper bound on the charging emissions is obtained by substituting the 56
g/charge estimate for the 0.16 g/charge estimate.   The resulting range of
emissions for Model  Battery 2 for a limit of 34 seconds of emissions per
charge is 0.005 to 1.7 megagrams of BSO per year.
     Regulatory Alternatives II and III provide emission limits in the form
of log averages for wet-coal charging.   Arithmetic averages cannot be
precisely determined from log averages, but the arithmetic average is
always greater than or equal to the log average.   The discussion in Section
4.1.1.4 and the data in Table 4-4 (Chapter 4) reveal that the arithmetic
average may be 5 or 6 seconds higher than the log average.   To estimate
mass emissions, arithmetic averages were estimated as 5 seconds greater
than the log averages.   This procedure avoids overstating the emission
reduction from the log average emission limits for Regulatory Alternatives II
and III.
     Emission estimates for door leaks are summarized in Table 7-2 for the
model  batteries.   A range of mass emission rates  was estimated in Sec-
tion 3.3.3.4 as 0.01 to 0.33 kilogram of BSO per  hour per leaking door for
well-controlled door leaks.   Total  BSO emissions  were calculated from the
total  number of doors,  the percent of doors allowed to leak,  and the range
                                  7-3

-------
                 TABLE 7-2.   AIR EMISSIONS FROM DOOR LEAKS
Regulatory
alternative
I

II


III


Model
battery
1
2
3
1
2
3
1
2
3
Percent
leaking
doors
16
16
16
12
12
12
7
7
10
BSO emissions
(Mg/yr)
1.
1.
2.
0.
0.
0.
0.
0.
0.
1
8
1
50
86
99
13
23
63
to
to
to
to
to
to
to
to
to
11.
20.
22.
5.
9.
11.
1.
2.
7.
0
0
0
7
8
0
5
6
0
BSO emission reduction
over baseline (Mg/yr)


0.
0.
1.
0.
1.
1.


60
94
1
97
6
5
0
0
0
to 5.
to 10.
to 11.
to 9.
to 17.
to 15.


3
0
0
5
0
0
 Per battery.
of emission rates.   A sample calculation follows for Model  Battery 2 which
has 62 ovens,  124 doors, and a baseline regulation of 16 PLD.   The
estimated average performance for a 16 PLD limit is 12 PLD  with a range of
0.014 to 0.15 kilogram of BSO per hour per leaking door.
     Sample calculation:
(124 x 0.16 x 0.75) leaking doors x "•>,»,.'  "^   x ">'uu "'  x ±"—LJa_
v                 '       y         hour/leaking door     year     kilogram
                                    = 1.8 Mg of BSO/yr
The highest estimate of door leak emissions is obtained by substituting
0.15 kilogram of BSO per hour for the 0.014 kilogram of BSO per hour emis-
sion rate.   The resulting range of emissions for Model Battery 2 with a
limit of 16 PLD is 1.8 to 20 megagrams of BSO per year.
     Emissions for lid and offtake leaks from the model  batteries are
summarized in Tables 7-3 and 7-4, respectively.   The mass emission rate was
estimated in Section 3.3.4 as 0.0033 to 0.021 kilogram of BSO per hour  per
leaking lid or leaking offtake.   This range includes narrow wisps and large
leaks with 1- to 2-meter plumes.   For lid leaks, the emissions are estimated
from the total number of lids,  the percent of lids allowed to leak, and the
range of emission rates.  A sample calculation follows for Model  Battery 2,
                                  7-4

-------
                 TABLE 7-3.  AIR EMISSIONS FROM LID LEAKS
Regulatory
alternative
I


II


III


Model
battery
1
2
3
1
2
3
1
2
3
Percent
leaking
lids
5
5
5
3
3
3
1
1
1
BSO emissions3
(Mg/yr)
0.12 to 0.75
0.20 to 1.3
0.31 to 2.0
0.07 to 0.5
0.1 to 0.8
0.2 to 1.2
0.02 to 0.2
0.04 to 0.3
0.06 to 0.4
BSO emission reduction
over baseline (Mg/yr)
0
0
0
0.05 to 0.25
0.10 to 0.50
0.11 to 0.80
0.10 to 0.55
0.16 to 1.0
0.25 to 1.6
 Per battery.
               TABLE 7-4.   AIR EMISSIONS FROM OFFTAKE LEAKS
Regulatory
alternative
I



II



III



Model
battery
1 .
2A*
2BC
3
1
2A
2B
3
1
2A
2B
3
Percent
leaking
offtakes
10
10
10
10
6
6
6
6
4
4
4
4
BSO emissions3
(Mg/yr)
0.08 to 0.50
0.13 to 0.86
0.27 to 1.7
0.31 to 2.0
0.05 to 0.3
0.08 to 0.5
0.16 to 1.0
0.18 to 1.2
0.03 to 0.2
0.05 to 0.3
0.11 to 0.7
0.12 to 0.8
BSO emission reduction
over baseline (Mg/yr)
0
0
0
0
0.03 to 0.20
0.05 to 0.36
0.11 to 0.70
0.13 to 0.80
0.05 to 0.30
0.08 to 0.56
0.16 to 1.0
0.19 to 1.2
 Per battery.
 Single-main battery.
cDouble-main battery.
                                  7-5

-------
which has 62 ovens, three lids per oven (186 total lids), a baseline limit
of 5 PLL, and a range of 0.0033 to 0.021 kilogram of BSD per hour.  The
average performance is estimated as 75 percent of the limit, or 3.8 PLL.
     Sample calculation:

                                                              "3
     (186 x 0.05 x 0.75) leaking lids x  -      kg. of BSO x  IP
                               b        hour/leaking lid    kilogram

                                                 = 0.20Mg of BSO/yr

The highest estimate is obtained by substituting 0.021 kilogram of BSD per
hour for the 0.0033 kilogram of BSD per hour emission rate.   The resulting
range of emissions for Model Battery 2 at 5 PLL is 0.20 to 1.3 megagrams of
BSO per year.
     For offtake leaks, Model 2A was defined as a single-main battery with
62 offtakes, and Model 2B was defined as a double-main battery with 124
offtakes.  A sample calculation follows for Model Battery 2B for a level of
10 PLO and a range of 0.0033 to 0.021 kilogram of BSO per hour.   The
average performance is estimated as 75 percent of the limit, or 7.5 PLO.
     Sample calculation:

     (124 x 0.10 x 0.75) leaking offtakes x   0.0033 kg of BSO   x lO^g,
                       '       3            hour/leaking offtake   kilogram

                                                     =0.27Mgof BSO/yr

Repeating this calculation for an emission rate of 0.021 kilogram of BSO
per hour yields a range for Model 2B of 0.27 to 1.7 megagrams of BSO per
year for a limit of 10 PLO.
7.1.2  Nationwide Primary Air Impacts
     Nationwide emissions for the regulatory alternatives were estimated on
a battery-by-battery basis by following the same procedure that was out
lined for the model batteries.   The battery-sperif ic approach considered
the number of ovens, doors, lids, and offtakes on each battery,  the coking
cycle time, and the range of emission rates for each emission point.   The
emissions were summed across all the batteries to estimate nationwide
emissions.
                                  7-6

-------
     An oven utilization of 90 percent was used to estimate nationwide
emissions because 10 percent of the nation's ovens may be out of service at
any given time for repair or rebuilding.1  An average control performance
was estimated as 75 percent of the applicable visible emissions limit.
     For Regulatory Alternative I, the existing baseline regulation from
SIPs and consent decrees that are applicable to each battery was used to
estimate mass emissions.  Nationwide emissions for Regulatory Alternatives II
and III were estimated by applying the appropriate emission limit from the
regulatory alternative.   Regulatory Alternative II for many batteries is
less stringent than the existing regulation; therefore, these batteries
will not experience an emission reduction from Regulatory Alternative II.
For example, baseline regulations are more stringent than or equal  to the
Regulatory Alternative II limits for about 82 batteries for charging, 105
batteries for door leaks, 111 batteries for lid leaks, and 71 batteries for
offtake leaks.   In terms of coke capacity, Regulatory Alternative II will
affect batteries with about 53 percent of the industry's coke capacity for
charging, 36 percent for door leaks, 34 percent for lid leaks, and  56 percent
for offtake leaks.
     The nationwide emission estimates are summarized in Table 7-5  for each
regulatory alternative and emission point.  The nationwide estimates were
calculated by summing the plant-specific estimates that are presented in
Tables E-6 through E-8 in Appendix E.   The baseline emissions from  door
leaks appear to contribute the most BSD emissions, followed by topside
leaks and charging emissions.   Regulatory Alternative II for the combined
emission points reduces total  nationwide emissions of BSD by approximately
30 percent from the baseline while Regulatory Alternative III reduces
emissions by approximately 70 percent from the baseline.
7.1.3  Ambient Concentrations of BSO from Dispersion Modeling
     Dispersion modeling techniques were used to estimate the long-term air
quality impacts of fugitive BSO emissions from wet-coal charging,  topside
leaks, and door leaks.   The Human Exposure Model  (HEM)2 is used at  distances
up to 20 km from an origin centered on the coke ovens to calculate  the
omnidirectional average concentration.
                                  7-7

-------
                  TABLE 7-5.   NATIONWIDE EMISSIONS OF BSO
Regu-
latory
Emission alter-
point native
Charging


Door leaks


Lid leaks


Offtake leaks


I
II
III
I
II
III
I
II
III
I
II
III
BSO emissions
(Mg/yr)
0.46
0.36
0.26
150
110
43
22
18
6.9
21
15
11
to
to
to
to
to
to
to
to
to
to
to
to
130
76
37
1,700
1,200
480
140
110
44
130
95
70
BSO emission
reduction over Percent
baseline (Mg/yr) reduction
0
0.10 to 54
0.20 to 93
0
40 to 500
110 to 1,200
0
4 to 30
15 to 96
0
6 to 35
10 to 60
0
40
70
0
30
70
0
20
70
0
30
50
Dav»r-cin+ vaHii^+nrtn -f v*r\m + ho Kacal-Jna fDormla + Ar'w Al'hainna't'iwa T A r* a 1 r* 1 1 —
 lated from the midrange values.   Rounded to one significant figure.

     The HEM estimates the annual average concentrations resulting from
emissions from point sources.   The dispersion model within the HEM is a
Gaussian model that uses the same basic dispersion algorithm as the
climatological form of EPA's Climatological Dispersion Model.3  Gaussian  .
concentration files are used in conjunction with STAR data and emissions
data to estimate annual average concentrations.   Details on this aspect of
the HEM can be found in Reference 2.
     Seasonal or an.iual stability array (STAR) summaries are principal
meteorological input to the HEM dispersion model.   STAR data are standard
climatological frequency-of-occurrence summaries formulated for use in EPA
models and available for major U.S.  sites from the National Climatic Center,
                                  7-8

-------
Asheville, N.C.  A STAR summary is a joint frequency of occurrence of wind
speed and wind direction categories, classified according to the Pasquill
stability categories.  For this modeling analysis, annual STAR summaries
were used.
     The model receptor grid consists of 10 downwind distances located
along 16 radials.  The radials are 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, 0.3, 0.5, 0.7, 1.0, 2.0,
5,0, 10.0, 15.0, and 20.0 kilometers.  The center of the receptor grid for
each plant was assumed to be the plant center as determined by review of
maps.  All concentration calculations were performed assuming the meteoro-
logic conditions of an urban area.
     Dispersion modeling was performed for each of the 53 by-product plants
that use wet coal for each emission point and each regulatory alternative.
These results are used in Appendix E to estimate health impacts from the
BSD emissions for each emission point and each regulatory alternative at
each battery.  Dispersion modeling results from the HEM are presented in
Table 7-6 for Model Battery 2.   For this model battery analysis,  meteoro-
logical data for Pittsburgh were used.   For the plant-specific analysis,
the meteorological data for each plant's location were used in the
dispersion model.
     The results for Model Battery 2 (Table 7-6) correspond to the concen-
trations predicted north of the battery where the maximum concentration
occurs.  The range given in the table corresponds to the minimum and maximum
emission estimates for a given emission point and regulatory alternative.
The lowest concentrations (not shown in the table) occurred southeast of
the battery and were 24 to 31 percent of the maximum concentrations north
of the battery.
7.2  OTHER IMPACTS
     No impacts on secondary air pollution, water pollution, solid waste
disposal, or energy usage are expected for any of the regulatory alter-
natives beyond the current baseline.
                                 7-9

-------
TABLE 7-6.   MAXIMUM AVERAGE ANNUAL  BSO  CONCENTRATIONS  FOR  MODEL  BATTERY 2C
Downwi nd
distance (km)

0.2
0.5
1.0
2.0
5.0
10.0
20.0
Rnwnwi nrl
distance (km)
0.2
0.5
1.0
2.0
5.0
10.0
20.0
Concentration of BSO from charging regulatory
I

1.8 x
6.1 x
2.3 x
7.9 x
1.9 x
6.7 x
2.4 x
(basel i
-3
10 J -
io-4-
io-4-
10" 5 -
io-5-
io-6-
io-6-
ne)

6.3 x
2.1 x
7.7 x
2.7 x
6.6 x
2.3 x
1.2 x
Concentration
I
6.6 x
2.2 x
8.1 x
2.8 x
7.0 x
2.4 x
8.5 x
(basel
lo-1-
lo-1-
io-2-
io"2 -
lO'3-
io-3-
ib

x 10"1
x 10"1
x 10"1
x 10"2
x 10"2
x 10"3
x 10"3
                                                                         (continued)

-------
                                     TABLE 7-6.   (continued)
Downwi nd
distance (km)
0.2
0.5
1.0
2.0
5.0
10.0
20.0
Concentration of BSO
]
1.7
5.8
2.1
7.4
1.8
6.3
2.2
from topside leak regulatory alternatives ((jg/m3)
[ (baseline)
x 10"1 -
x 10"2 -
x 10'2 -
x 10"3 -
x 10"3 -
x 10~4 -
x 10~4 -
1.1
3.7
1.4
4.7
1.2
4.0
1.4
xlO°
x 10"1
x 10
x 10"2
x 10"2
x 10
x 10
9.6 x
3.2 x
1.2 x
4.1 x
1.0 x
3.5 x
1.2 x
II
Kf2-
io"2 -
io-2-
lO'3-
10~3 -
io"4 -
lO'4-
III
6.6 x
2.2 x
8.1 x
2.8 x
7.0 x
2.4 x
8.5 x
io-1
10"1
ID'2
ID'2
io"3
10"3
ID'4
5.5 x
1.8 x
6.8 x
2.4 x
5.8 x
2.0 x
7.1 x
io-2-
io-2-
io-3-
io-3-
io-4-
io-4-
io-5-
3.7 x 10'1
1.2 x 10"1
4.5 x 10"2
1.6 x 10"2
3.9 x 10"3
1.3 x 10"3
4.7 x 10~4
Maximum concentrations occurred north of the battery using Pittsburgh meteorological  data.   The
lowest concentrations were predicted southeast of the battery and were 24 to 31 percent of the
maximum values shown in this table.

The ranges are for concentrations north of the battery and correspond to the minimum and
maximum emission estimates.

-------
7.3  IMPACT OF DELAYED STANDARDS

     Delay of the standard will not have a significant impact on water

pollution, solid waste disposal, or energy.  A delay will result in contin-

ued air pollution at or above the baseline level (Table 7-5).  The health

impact from control at the regulatory baseline level (described in Appen-

dix E) would continue for the duration of the delay.

7.4  REFERENCES

1.    Hogan, W.  T. ,  and F.  T.  Koelble.   Analysis of the U.S.  Metallurgical
     Coke Industry.  Industrial Economics Research Institute.  Fordham
     University.  October 1979, p.  41.

2.    Systems Applications, 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-2/250-2.
     May 1980.

3.    Busse, A.  D.,  and J.  R.  Zimmerman.  User's Guide for the Climatological
     Dispersion Model.   (Prepared for the U.S.  Environmental Protection
     Agency, Research Triangle Park, North Carolina.)  Publication Number
     EPA-R4-73-024.  December 1973.
                                  7-12

-------
                               8.   COSTS

8.1  COST ANALYSIS OF REGULATORY ALTERNATIVES
     This chapter analyzes the costs associated with the regulatory
alternatives discussed in Chapter 6.  The cost impacts of these alterna-
tives include the cost to move from the uncontrolled level  to the
SIP/OSHA/consent decree baseline level and the incremental  cost of the
EPA alternatives as they increase in stringency.   For each  source, the
regulatory baseline is listed as Regulatory Alternative I and is
defined as the level of control that should be achieved in  the absence
of national emission standards.  The cost estimates are derived on a
model battery basis for each source and each alternative.
     The model battery approach is a summary of a more detailed cost
analysis that was used to develop total industry costs on a battery-by-
battery basis.  The results of the more comprehensive analysis of
total industry costs are presented in Chapter 9.   The model battery
cost analysis does not include estimates of the costs incurred as a
result of lost production because of increased cycle time or retrofit
down time.  Production penalties are considered in the economic assess-
ment in Chapter 9.
     The cost estimates for the model batteries are provided for total
installed capital costs and total  annualized costs for existing, modi-
fied or reconstructed, and new batteries.   Because of similar work
practice and retrofit requirements, the costs associated with meeting
the regulatory alternatives on modified or reconstructed batteries are
approximately the same as those estimated for existing batteries.  New
battery costs are presented separately because equipment modifications
and improved design can often be incorporated into the construction
plans at a cost lower than that of retrofitting similar equipment to
                                 8-1

-------
an existing battery.   New batteries are generally the tall (5- or
6-meter) batteries instead of the 3- to 4-meter size of most existing
batteries.
     Section 8.1.5 discusses the cost effectiveness of the regulatory
alternatives on a model battery basis in terms of dollars per kilogram
of benzene soluble organics (BSO) removed.   The total annualized costs
and annual  BSO reductions are used to determine the incremental  cost
effectiveness between EPA alternatives with increasing stringency and
the average cost effectiveness of each EPA alternative calculated from
the regulatory baseline.
8.1.1  Basis for Model Battery Costs
     The model batteries that were developed in Chapter 6 are described
in Table 8-1.  For this cost analysis, a subset of Model  2 has been
defined to account for some of the differences in age and construction
features that will affect retrofit costs.   Model  3 is a 6-meter  battery
that is typical of new sources.
     A summary of the estimated requirements of each alternative is
presented in Table 8-2.  Full stage charging is required for all
models to reach the regulatory baseline; optimization of stage charging
is required for all models to attain the limits of Regulatory Alterna-
tive II.   For door leaks, a work practice and maintenance program is
required to meet the baseline.   Additional  improvement in control is
obtained by the items that are listed in Table 8-2 and are described
in more detail in the following sections.   For topside leaks, luting
manpower and offtake modifications on some batteries are required to
attain the baseline and Regulatory Alternative II limits.
     Table 8-3 lists the installed capital  cost estimates of the major
items used in the cost analysis.   An estimate of  the battery cost is
provided for comparison.   Annual  cost items are listed in Table  8-4.
The plant overhead, estimated from 50 percent of  labor and materials,
is probably overstated.  The incremental overhead is not this large
for small additions to large plants, particularly when the coke
operation is part of a steelmaking complex.   However, this cost  is
counterbalanced by supervisory costs that are not included for the
                                8-2

-------
                                          TABLE  8-1.   MODEL BATTERIES
      Model  1:   Foundry coke battery producing 106,000 megagrams  of coke per year;  36 ovens;  single
                collecting main;  3.5-meter hand-luted doors.

      Model  2:   Metallurgical  coke battery producing 344,000  megagrams  of coke per  year;  4-meter self-
                sealing doors.

            2A:  Double collecting main;  over 10 years since a rehabilitation; requires battery top
                rehabilitation,  door and jamb replacement,  modification of offtakes.

            2B:  Double collecting main;  less than 10 years  since  last major rebuild.

            2C:  Single collecting main;  does not  require rehabilitation.

      Model  3:   Metallurgical  coke battery producing 770,000  megagrams  of coke per  year;  71 ovens;
                double collecting main;  6-meter self-seal ing  doors; representative  of type most likely
 •               to be constructed (new source).
CO                                                                              	

-------
                     TABLE 8-2.  SUMMARY OF  REQUIREMENTS FOR  EACH  REGULATORY  ALTERNATIVE
00
Regulatory
Source alternative
Charging I


II
III

Topside leaks I
II
III
Model battery
1
2A
2B.2C.3
1,2A,2B,2C,3
1,2A,2B,2C
3
All
All
All
Emission limit
11-34 s/charge
11-34 s/charge
11-34 s/charge
16 s/charge
8 s/charge
15 s/charge
1-5 PLL,
4-10 PLO
3 PLL, 6 PLO
1 PLL, 4 PLO
Estimated requirements for cost
analysis
FULL STAGE CHARGING FOR ALL MODELS:
Bunker modifications; battery top
rehabilitation; new larry car; ade-
quate aspiration
Battery top rehabilitation; new
larry car; adequate aspiration
Modified larry car; adequate
aspiration
FOR ALL MODELS:
Optimized stage charging; worker
training; minor equipment
modifications
Unknown
Unknown
Luting manpower
Additional luting manpower; offtake
modifications
Unknown
      See  footnotes  at  end  of  table.
(continued)

-------
                                             TABLE 8-2 (continued)
        Source
Regulatory
alternative
Model battery     Emission limit
Estimated requirements for cost
          analysis
00
i
en
        Door leaks
                1


                2A




                28,2C
                      4-16 PLD       Maintenance and repair;  two luter-
                                     men;  door repair shop

                      4-16 PLD       Door  and jamb replacement;  improved
                                     door  shop;  primary maintenance;
                                     inspection, cleaning, adjustment;
                                     seal  replacement;  refractory repair

                      4-16 PLD       Improved door shop; primary mainte-
                                     nance;  inspection, cleaning, adjust-
                                     ment;  seal  replacement;  refractory
                                     repair

                      4-16 PLD       Primary maintenance; inspection,
                                     cleaning, adjustment; seal  replace-
                                     ment;  refractory repair
II

III

1
2A,2B,2C,3
1.2A.2B.2C
3
12 PLD
12 PLD
7 PLD
10 PLD
Two lutermen per shift
Modified doors and seal
Unknown
Unknown

s


        PLL =  percent  leaking  lids.
        PLO =  percent  leaking  offtakes.
        PLD =  percent  leaking  doors.

         The requirements  are  cumulative as the emission limit stringency increases.

-------
                                 TABLE 8-3.  CAPITAL COST ITEMS (1979 dollars)
      Item
                                Cost estimate
                                  Derived from reference(s)c
00
 i
CD
Coke battery
  Model 2
  Model 3
Decarbonizing air
Decarbonizing scraper
Doors
  3-m, luted
  4-m, self-sealing
  6-m, self-sealing
Door seals
  4-m
  6-m
Gooseneck cleaner
Improved door repair shop
Improved topside luting
Jumper pipe
Leveler bar seal
Modify coal bunker
Modify doors
Modify goosenecks
Modify larry car
34,200,000
48,900,000
    80,000
     3,000

     2,000
    10,000
    22,500

       850
     1,700
   137,000/collecting main
   150,000
       800/oven
     2,500/oven
    62,000
 1,000,000
    13,800/oven
     2,750 each
   500,000
                                                                                 1
                                                                                 1
                                                                                 2
                                                                                 2,  3

                                                                                 Q
                                                                                 Q
Q
Q
2
2
4
2, 5
2, 5, 6
5
Q
7
2, 5, 6
      See  footnotes  at end of  table.
                                                                                        (continued)

-------
                                              TABLE 8-3  (continued)
       Item                            Cost  estimate                      Derived  from  reference(s)

       New larry  car
         3-m                           1,400,000                                   2, 3,  6
         4-m                           1,800,000                                   2, 3,  6
         6-m                           2,500,000                                   2, 3,  8
       Repave  top,  realign  tracks          2,400/oven                              2
       Replace doors  and  jambs                  7/Mg of coke                        4
       Reset or replace lid rings          2,000/oven                              5
       Second  collecting  main             47,000/oven                              2, 6,  Q
oo      Steam lines  and nozzles             2,200/oven/collecting main              2, 5,  6
_^                                                                                                     =
       Q  = derived  from industry questionnaires.
       aEscalated to  current dollars  at a rate  of  10 percent per year.
       This cost is  for  the battery  proper  and does not  include operating  equipment,  the quench  tower,
       coal handling, or the by-product plant.
       Includes  steam lines and nozzles.

-------
                 TABLE 8-4.   ANNUALIZED COSTS (1979 dollars)1
Item
      Cost estimate
Direct operating costs
     Steam
     Electricity
     Labor (operating and maintenance)
     Supervision
     Maintenance materials and supplies
 $ 4.75 per 1,000 pounds
 $ 0.027 per kilowatt-hour
 $15.70 per hour
 $17.04 per hour
 50% of maintenance labor
Indirect operating costs
     Payroll overhead
     Plant overhead
       10% of labor
50% of labor and materials
Fixed and general  expense factors
     Taxes,  insurance
     Capital  recovery factor
     Administration overhead
       2% of capital

       2% of capital
 Calculated with capital  recovery factor formula using assumed  equipment
 lifetimes and an estimated interest rate (minimum attractive  return)  of
 10 percent.
                                    8-8

-------
small increase in labor.   The cost estimates in this section are
expressed in 1979 dollars.   The referenced cost estimates were escalated
to 1979 dollars at a rate of 10 percent per year.
8.1.2  Cost of Regulatory Alternatives for Charging
     8.1.2.1  Existing Batteries.   The cost to attain the regulatory
baseline for charging has been estimated from the requirements for
full implementation of stage charging.  These requirements can vary
widely from battery to battery depending on battery design, age,
condition, and other retrofit considerations.   Costs for the model
batteries were developed by assuming a specific type of battery,
approximate age, and condition.  This approach should bound the probable
range of cost impacts.
     Stage charging has been implemented to varying degrees at most
existing batteries.   Some batteries have fully implemented the practice
and others will incur substantial  expense to complete all of the
aspects described in Section 4.1.1.  The estimated control level that
can be obtained by stage charging  is an average of 5 to 30 seconds of
visible emissions per charge (s/charge).
     The capital cost estimates for stage charging (Table 8-5) range
from $1 million to $3 million.   The basic items include a modified
larry car, adequate aspiration, and battery modifications.  Approxi-
mately 5 percent of the batteries  may require a modified coal  bunker
(Model 1) when there is inadequate clearance for changing the  hopper
size.  Older batteries may require repaying and realignment of the
larry car tracks or resetting the  lid rings to rehabilitate the battery
top. The cost of lost production during the coal bunker modification
or rehabilitation of the battery top is not included, but this cost
may be significant because of the  major disruption of normal  operation.
     The total annualized costs (Table 8-5) range from $624,000 to
$906,000, or $1.81 to $8.46 per megagram of coke.   The labor costs
include a man per shift to ensure  timely removal and replacement of
lids.  The costs for Model  1 will  be incurred by 5 percent of  the
batteries with 2 percent of the industry's capacity.7  The majority of
batteries have already implemented most of the basic requirements of
                                 8-9

-------
             TABLE 8-5.  COST ESTIMATES FOR CHARGING REGULATORY ALTERNATIVE I — EXISTING  SOURCES
                                          (thousands of 1979 dollars)
00
Model
Regulatory Alternative I requirements
Investment requirements
Modify coal bunker
Reset or replace lid rings
Repave top, realign tracks
New larry car
Modified larry car
Steam lines and nozzles
Jumper pipes
Leveler bar seal
Straighten standpipes
Gooseneck cleaner
Decarbonizing air
Decarbonizing scraper
Total installed capital cost
Operating requirements
Utilities - steam
(a) Operating labor (8,760 man-hours)
(b) Maintenance labor (6,200 man-hours)
(c) Materials (1% of capital)
Payroll overhead (10% of a + b)
Plant overhead (50% of a + b + c)
Administration overhead, taxes, insurance (4% of capital)
Total operating costs
Capital recovery (25 yr at 10%)
Total annual ized cost
Dollars per megagram of coke
1

1,000
72
86
1,400
-
79
90
62
16
137
80
3
3,025

21
138
97
30
24
133
121
564
333
897
8.46
2A

-
124
149
1,800
-
272
-
62
32
274
80
3
2,796

67
138
97
28
24
132
112
598
308
906
2.63
2B

-
-
-
-
500
272
-
62
32
274
80
3
1,223

67
138
97
12
24
124
49
511
135
646
1.88
2C

-
-
-
-
500
136
155
62
16
137
80
3
1,089

67
138
97
11
24
123
44
504
120
624
1.81

-------
stage charging and are expected to have incurred costs in the range
given for Model 2 ($1.81 to $2.63 per megagram of coke).
     Regulatory Alternative II imposes a visible emission time of a
log average of 16 s/charge for all batteries.   The incremental cost of
this alternative is estimated based on the requirements described in
Section 4.1.1.3.  An engineering study may be required to determine
what particular aspects of the charging operation prevent consistent
control of emissions.   EPA has conducted engineering studies of charging
control techniques at several batteries to collect visible emissions
data, document work practices and written procedures, analyze control
equipment, and determine if stage charging has been fully implemented.
The analysis of control equipment includes evaluation of drop sleeve
alignment; charge port blockage; charge time sequence; coal  peak
heights, bulk density, and charge volume; roof carbon buildup, adequacy
of aspiration; gooseneck and standpipe openings; chuck door and lid
operation; and collecting main pressure control.  Actual  work practices
are compared to written procedures, deviations are noted, and the
causes of excess visible emissions are investigated.   Findings will
vary from battery to battery because of differences in worker attitude,
level and quality of supervision, training, or minor equipment problems.
     A contractor with experience in conducting several engineering
studies of emission control at coke batteries provided the following
estimate of labor hour requirements for a three-member engineering
team to investigate charging emissions problems.  The estimate is for
a coke plant with three batteries and for a 5-day observation period.9
                                             Technical
               Item                            hours
          Preparation                            16
          Field study                           120
          Analyze data, prepare report          160
          Management                             20
               Total                            316
The total technical hours include 316 hours for various levels of
professional labor.  Total expenses would also include secretarial
time, per diem allowance, and travel expenses.9  The cost estimate for
the engineering study is given in Table 8-6 as $12,000.
                                 8-11

-------
            TABLE 8-6.   COST ESTIMATE FOR ENGINEERING STUDY OF
                 CHARGING EMISSION CONTROL (1979 dollars)
Item
EPA Level P-l (junior professional)
EPA Level P-3 (professional)
EPA Level P-4 (senior professional)
Secretarial
_,. b
Per diem
Travel0
Total
Hours
140
140
40
20
-
-

Rate3
23
33
40
15
-
-

Cost
3,200
4,600
1,600
300
1,200
600
12,000d
 1979 dollars  per hour.
 Three people  for 7 days  at $55 per day.
"Three people  at $200 each.
 Rounded.
                                  8-12

-------
     The major capital expenses will  have been incurred in implementing
stage charging, which is required by existing OSHA regulations,  SIPs,
and consent decrees.   However, a capital  cost of $100,000 has been
used in Table 8-7 as an order of magnitude estimate for "fine-tuning"
equipment items.   For example, either a gooseneck inspection platform,
automatic lid lifters, a larry car positioning system,  improved aspira-
tion through increased steam pressure or altered nozzle design,  or
altered lid design, may be necessary for a specific battery.   The
$100,000 order of magnitude estimate was derived from costs for typical
stage charging improvements such as automatic lid lifters ($188,000),
steam aspiration ($81,000), or a larry car positioning system ($81,000).8
     The requirements for monitoring costs are estimated by assuming
the addition of a smokereader to record visible emissions during
daylight hours.  This smokereader will also observe and record topside
and door leaks.  The total annualized cost for Regulatory Alternative
II is given in Table 8-7 as $99,000 per year per battery unit, or
$0.13 to $0.47 per megagram of coke for the various models.  A battery
unit is defined as a group of ovens serviced by one larry car and may
include one to three batteries that are interconnected.   A battery
unit is typically composed of 50 to 120 ovens.   The number of batteries
operated as a unit is assumed to be two to three for Model Battery 1
(36 ovens), one to two for Model Battery 2 (62 ovens),  and one for
Model Battery 3 (71 tall ovens).
     Regulatory Alternative III applies an 8 s/charge limit to short
batteries and a 15 s/charge limit to tall batteries, based on exemplary
battery performance.   A cost analysis is not presented for this  alterna-
tive because no control technology has been widely demonstrated  that
would permit all  types of batteries to meet this limit.
     8.1.2.2  New Batteries.   The capital cost of implementing stage
charging (Regulatory Alternative I) for a new 6-meter battery is given
in Table 8-8 as $1.2 million, with a total annualized cost of $731,000,
or $0.95 per megagram of coke.  The cost of Regulatory Alternative II
is the same as that given in Table 8-7 to optimize the stage charging
procedure.
                                   8-13

-------
                TABLE 8-7.   INCREMENTAL COST ESTIMATES  FOR  CHARGING  REGULATORY  ALTERNATIVE II—NEW
                                               AND  EXISTING SOURCES
                                  (applicable to  each battery  unit,  1979  dollars)

       Item                                                               Capital  cost ($)

       Engineering study (from Table 8-6)                                     12,000
       Minor equipment modifications                                         100,000

                                TOTAL                                        112,000

       Operating costs                                                 Cost/battery unit ($)

       Monitoring - One man/battery, 2,080 man-hours x  $17.04/hr              35,000

       Training program - One instructor using one-third  of his time
                          per battery,  670 man-hours at $17.04/hr             11,400
00
£      Classroom time - Five men/shift/battery at $15.70/hr, 4 shifts,
                        with 8 hr of emission control training
                        every 6 mo                                             5,000

                                Total  operating cost                          51,400
       Payroll  overhead (10%)                                                  5,100
       Plant overhead (50%)                                                  25,700

                                Total  direct operating  cost                  82,200
       Capital  recovery (25 yr at 10%)                                        12,300
       Administration overhead (2% of capital)                                 2,200
       Taxes,  insurance (2% of capital)                                        2,200

                                Total  annualized  cost                         98,900

                                                                       Dollars  per megagram of coke

                 Model 1 (2 per battery unit)                                      0.47
                 Model 2 (1.5 per battery  unit)                                    0.19
                 Model 3 (1 per battery unit)                                      0.13

-------
    TABLE 8-8.  COST ESTIMATE FOR CHARGING REGULATORY ALTERNATIVE I~
                                NEW SOURCES
                        (thousands of 1979 dollars)

Regulatory Alternative I requirements                         Model  3

Investment requirements

Modifications to larry car                                      500
Steam lines and nozzles                                         312
Leveler bar seal                                                 62
Gooseneck cleaner                                               274
Decarbonizing air                                                80
Decarbonizing scraper                                         	3

                         Total installed capital  cost         1,231

Operating requirements

Utilities - steam                                               151
  (a) Operating labor (8,760 man-hours)                         138
  (b) Maintenance labor (6,200 man-hours)                        97
  (c) Materials (1% of capital)                                  12
Payroll overhead (10% of a + b)                                  24
Plant overhead (50% of a + b + c)                               124
Administration overhead, taxes, insurance (4% of  capital)        49

                         Total operating costs                  595
Capital recovery (25 yr at 10%)                                 136

                        . Total annualized cost                  731

Dollars per megagram of coke                                      0.95
                                  8-15

-------
8.1.3  Cost of Regulatory Alternatives for Topside Leaks
     8.1.3.1  Existing Batteries.   The regulatory baseline for topside
leaks is 1 to 5 PLL and 4 to 10 PLO.   Adequate manpower and conscien-
tious luting of leaks are required to attain the regulatory baseline.
In addition, some older batteries may have incurred costs for repair,
modification, or replacement of standpipes.   In general, four lidsmen
per battery are required if automatic lid lifters are used, and eight
lidsmen per battery are required if lid lifting is performed manually.
At some batteries, the larry car operator or helper seals standpipe
caps; at other batteries the lidsmen perform this function.  Foundry
coke batteries require fewer topside workers than metallurgical  coke
batteries require because of the longer cycle time (30 hours versus
18 hours) and fewer ovens charged.   One source estimates that a foundry
coke plant requires 4.9 lidsmen per battery unit (group of batteries)
and that a metallurgical coke plant requires 9.8 lidsmen per battery
unit.i
     An estimate of the cost for the model batteries to meet the
regulatory baseline is given in Table 8-9.  The estimate may be somewhat
overstated because the topside worker performs functions not directly
related to emission control, such as lid removal and replacement and
assisting in the charging operation.   Some older batteries may have
incurred a capital cost for standpipe modifications or replacement
that may be attributed to emission control.   For the worst case of
replacement of standpipes (at $2,750 each) on a double main battery, a
capital cost of $341,000 would have been incurred for Model Battery 2.
This expense would have increased the total  annualized cost for Model
Battery 2 from $648,000 per year to $700,000 per year.
     Regulatory Alternative II establishes a limit of 3 PLL and 6 PLO,
based on control demonstrated at several  plants.  This control  is
accomplished by adding manpower to aid in luting the leaking lids and
offtakes.  A capital cost of $800 per oven is also included for minor
equipment replacement or modification on batteries that have damaged
offtakes. This estimate is based on the cost given below for a plant
with 315 ovens.4
                                 8-16

-------
         TABLE 8-9.  COST ESTIMATES FOR TOPSIDE LEAKS REGULATORY
                 ALTERNATIVE I—NEW AND EXISTING SOURCES
                  (thousands of 1979 dollars per year)
Operating requirements
Labor ($15.7/hour)a
Materials and supplies (50% of labor)
Payroll overhead (10% of labor)
Plant overhead (50% of labor & material)
Total annual ized cost
Dollars per megagram of coke
Model 1
138
69
14
104
325
3.07
Model
275
138
28
207
648
1.88
2 Model 3
138 to 275
69 to 138
14 to 28
104 to 207
325 to 648
0.42 to 0.84
Model Battery 1 is a small foundry battery with a 30-hour cycle and four
lidsmen.

Model Batteries 2 and 3 are metallurgical batteries with an 18-hour cycle.
Model 2 has manual lid lifting and eight lidsmen.   Model 3 has either
manual or automatic lid lifters with four to eight lidsmen.
                                 8-17

-------
          Supply material and equipment to seal all top ports and
          experiment with luting materials to determine the most
          effective seal ($215,000).
          Replace standpipes and goosenecks where necessary to
          permit proper seals and eliminate leaks from offtake
          piping ($42,000).
     The additional labor required to improve topside leak control
from the most lenient baseline regulation (5 PLL and 10 PLO) to 3 PLL
and 6 PLO is based on existing topside labor requirements.   One plant
reports that 48 lidsmen are used on seven batteries (45 ovens each)
with an 18-hour cycle time.4  This plant has double collecting mains,
manual lid lifting, and three lids per oven.   The job responsibilities,
in addition to lid removal  and replacement, include:4
          Inspect and clean goosenecks and steam jets.
          Coordinate with larry car operator.
          Lute and relute as necessary to maintain smokeless oven
          tops.
A total of 100,000 man-hours (48 lidsmen) per year are required to
perform these job duties.  If all of the effort were devoted to luting
lids and offtakes, 0.13 man-hour per lid or offtake would be required
for the 315 ovens (750,000 lids and offtakes luted per year).
     The incremental labor required for Regulatory Alternative II is
given in Table 8-10 for the model batteries.   The estimate is for
improving the level of control from the most lenient baseline regula-
tions (5 PLL and 10 PLO) to 3 PLL and 6 PLO.   At 0.13 man-hour per
leak, a total of 140 to 720 man-hours would be required for the model
batteries.   The incremental  annualized cost for Regulatory Alternative II
is given in Table 8-11 as $8,000 to $32,000 per year or $0.07 to
$0.09/Mg coke.
     Regulatory Alternative III, with limits of 1 PLL and 4 PLO,
represents the best observed control  of topside leaks at exemplary
batteries.   It is not known how all existing batteries could attain
this level; therefore, no cost estimates are given.
                                 8-18

-------
         TABLE 8-10.   INCREMENTAL LABOR REQUIRED FOR TOPSIDE LEAKS
                         REGULATORY ALTERNATIVE II
Parameter
Number of ovens
Cycle time (hours)
Charges/year
Lids/oven
Lids luted/year3
Offtakes/oven
Offtakes 1uted/yeara
Additional lids to lute/year
Additional offtakes to lute/yearc
Total additional leaks
Man-hours/year
Model 1
36
30
11,000
3
33,000
1
11,000
660
440
1,100
140
Model 2
62
18
30,000
4
120,000
2
60,000
2,400
2,400
4,800
620
Model 3
71
18
35,000
4
140,000
2
70,000
2,800
2.800
5,600
730
 Each lid and offtake on an oven is luted each time the oven is charged.
 The additional lids are equal  to 2 percent of the total (5 PLL - 3 PLL).
 The additional offtakes are equal  to 4 percent of the total (10 PLO -
 6 PLO).
dAt 0.13 man-hours/leak.
                                  8-19

-------
         TABLE 8-11.   INCREMENTAL COST ESTIMATES FOR TOPSIDE LEAKS
                REGULATORY ALTERNATIVE II—EXISTING SOURCES
                        (thousands of 1979 dollars)


         Regulatory Alternative II                          Model
               requirements                        1         2A         2B

Investment requirements

Improved topside luting ($800/offtake)           29.0       99.0      50.0

Operating requirements

Labor (from Table 8-10 at $15.70/hr)              2.2        9.7       9.7
Materials and supplies (1% of capital)            0.3        1.0       1.0
Payroll  overhead (10% of labor)                   0.2        1.0       1.0
Plant overhead (50% of labor)                     1.1        4.9       4.9
Administration overhead, taxes, and insurance     1.2        4.0       2.0
  (4% of capital)

       Total  operating cost                       5.0       21.0      19.0
Capital  recovery (25 yr at 10%)                   3.0       11.0       6.0
       Total  annualized cost                      8.0       32.0      25.0

Dollars  per megagram of coke                      0.08       0.09      0.07
                                  8-20

-------
     8.1.3.2  New Batteries.  The estimate of total cost to meet the
regulatory baseline for a new 6-meter battery was given in Table 8-9.
The cost for meeting the regulatory baseline is estimated as $325,000
to $648,000 per year, based on the assumption that one to two men per
shift perform the topside luting duties.   The cost for Regulatory
Alternative II is given in Table 8-12 and is estimated as $25,000 per
year ($0.03 per megagram of coke), based on additional labor to improve
the level of control from 5 PLL and 10 PLO to 3 PLL and 6 PLO.   Capital
costs for offtake modifications and improved luting mixtures are not
included because the best design should be incorporated into the
construction plans for a new battery.
8.1.4  Cost of Regulatory Alternatives for Door Leaks
     8.1.4.1  Existing Batteries.   The regulatory baseline for doors
ranges from 4 to 16 PLD.  The operating and capital costs for the
regulatory baseline (Regulatory Alternative I) are estimated for the
model batteries by assuming that a door leak control  program must be
implemented as dictated by OSHA regulations.   Requirements will vary
widely from battery to battery and will depend upon many factors such
as battery age, type of door seals, and battery condition.
     The labor estimates for a control program are given in Table
8-13.  Primary maintenance for self-sealing doors includes repairing
and maintaining seals within the original specifications, repairing
nicks, and maintaining the chuck door and end-door components in good
condition.   Luted doors require less maintenance, but they require two
men per shift to clean and lute doors on both sides of the battery.
Self-sealing doors require additional  labor for cleaning at the end of
the coking cycle, adjusting the seal after the door is replaced,
inspecting door and jambs for damage,  replacing damaged seals,  and
repairing refractory.  The control costs for door leaks also include
an annual expenditure for the materials cost of replacing doors and
seals.  Based on responses to industry questionnaires, the installed
cost for doors is estimated as $2,000 for a hand-luted door, $10,000
for a 3.5-meter self-sealing door, and $22,500 for a  6-meter self-
sealing door.   Seal costs are estimated as $850 for Model 2 and $1,700
                                 8-21

-------
         TABLE 8-12.   INCREMENTAL COST ESTIMATE FOR TOPSIDE LEAKS
                  REGULATORY ALTERNATIVE II--NEW SOURCES
                   (thousands of 1979 dollars per year)

Operating requirements                                       Model  3

Labor (from Table 8-10 at $15.70/hr)                           12

Materials and supplies (50% of labor)                           6

Payroll  overhead (10% of labor)                                 1

Plant overhead (50% of labor)                                  _6

                         Total operating cost                  25

Dollars  per megagram of coke                                    0.03
                                 8-22

-------
                         TABLE 8-13.  LABOR REQUIREMENTS FOR A DOOR LEAK CONTROL PROGRAM
                                                                                         Model
      Operation                                 Estimate of man-hoursa            1      2A,  2B,  2C


      Primary maintenance                                        .
        Self-sealing doors                      16 man-hours/door                --         7,936
        Luted doors                             16 man-hours/week              1,000

      Luter                                     2 men/shift                   17,520

      Cleaning                                  1 man/shift                      --C        8,760

      Adjustment                                1 man-hour/door/day              --         9,052

      Door inspection                           8 man-hours/day6                 --C        2,920
CO                                                               ,:
ro    Seal replacement                          5 man-hours/seal                 --            248

      Refractory repair                         4 man-hours/door                 288           496

                               Total manhours                                 18,808        29,412

       Derived from responses to Section 114 questionnaires and engineering estimates.

       Performed every 3 months.

       Performed by luter.

       Assumes 20 percent of the doors need adjustment.
      Q
       Includes replacement as required.

       Seals replaced every 30 months.

-------
for Model 3 from the same sources.   These costs are estimated using
lifetimes of 20 years and 30 months for a door and a seal, respectively.
     The investment requirements to meet the regulatory baseline
depend upon the present condition of the doors and jambs.   Improvements
to the door repair shop are considered a baseline expense and have
been estimated for the model batteries in Table 8-14.   Model  2A has
been defined as a battery that has warped jambs and damaged doors and
will require extensive repairs to reach the regulatory baseline.   The
$2.6 million estimate includes replacement of all doors and jambs at
$7 per megagram of coke.   This estimate does not include the  cost of
extensive trimwork, flue repair, or other repairs that are not directly
related to door leak control but which may be made when the jambs are
replaced.  The total annualized cost to meet the regulatory baseline
(Table 8-14) ranges from $481,000 to $1,263,000 for the models, or
$2.52 to $4.54 per megagram of coke.
     Regulatory Alternative II sets a limit of 12 PLD for all batteries,
based upon control demonstrated at several batteries.   The control
requirement to meet this level is the addition of two men to  provide
more luting (and reluting) manpower for hand-luting doors.  Short
batteries with self-sealing doors will require the door and seal
modifications discussed in Sections 4.2.2.1 and 4.2.2.2.
     The total annualized cost for a battery with luted doors is given
in Table 8-15 as $469,000, or $4.42 per megagram of coke.   The estimate
for batteries with luted doors (approximately 13 percent of total U.S.
coke capacity) represents an upper bound in costs for additional
manpower.  The manpower requirements may be lower for those batteries
operating on a longer cycle time (e.g., 30 hours instead of 18 to 20
hours).   The capital cost to modify self-sealing doors is $856,000 for
Models 2A, 2B, and 2C.   This estimate is based on industry question-
naire responses that indicated a pusherside door and a cokeside door
could be modified at a cost of approximately $13,800.   The total
annualized cost for modified doors and seals for the model is $144,000,
or $0.42 per megagram of coke.
                                 8-24

-------
             TABLE 8-14.   COST ESTIMATES FOR DOOR LEAKS REGULATORY ALTERNATIVE I—EXISTING SOURCES

                                           (thousands of 1979 dollars)
00
 i
tn
Regulatory Alternative I requirements
Investment requirements
Improvements to door repair shop
Replace all doors and jambs
TOTAL
Operating requirements
Labor (from Table 8-13 at $15.70/hr)
Materials
Doors (5% of doors per year)
Seal, chuck doors (40% of seals replaced each year)
Payroll overhead (10% of labor)
Plant overhead (50% of labor)
Administration overhead, taxes, insurance (4% of capital)
Total operating cost
Capital recovery (25 yr at 10%)
Total annual i zed cost
Dollars per megagram of coke

1

75
--
75

295

7
--
28
140
3
473
8
481
4.54
Model
2A

150
2,646
2,796

462

62
42
46
231
112
955
308
1,263
3.67

2B, 2C

150
—
150

462

62
42
46
231
6
849
17
866
2.52

-------
         TABLE 8-15.   INCREMENTAL COST ESTIMATES FOR DOOR LEAKS REGULATORY ALTERNATIVE II-
                                          EXISTING SOURCES
                                     (thousands of 1979 dollars)

                                                                       Model
Regulatory Alternative II requirements                           1         2A, 2B, 2C

Investment requirement

Modify doors ($13,800/oven)                                      --            856

Operating requirements

Labor (17,520 man-hours)                                        275             —a
Materials and supplies
  Luted doors (10% of labor)                                     28
  Self-sealing doors (1% of capital)                             --              9
Payroll overhead (10% of labor)                                  28             --a
Plant overhead (50% of labor)                                   138             —a
Administration overhead, taxes, and insurance (4% of capital)     --             34

                         Total  operating cost                   469             43
Capital recovery (20 yr at 10%)                                	—            101

                         Total  annualized cost                  469            144

Dollars per megagram of coke                                      4.42           0.42

 The operating labor required for modified doors  and seals is  not significantly different from
 the operating labor required for the  basic door  leak control  program.   This labor cost has been
 included in the cost of Regulatory Alternative I (regulatory  baseline).

-------
     Regulatory Alternative III applies limits of 7 PLD to short
batteries and 10 PLD to tall  batteries, based on exemplary battery
performance.   It has not been demonstrated that all types of batteries
could meet this limit; therefore, control  costs could not be estimated.
     8.1.4.2  New Batteries.   The cost of a door leak control program
for a new battery is estimated in Table 8-16 as $902,000 per year, or
$1.17 per megagram of coke.  Although a new battery would have new
doors and seals, the labor requirements for maintenance, cleaning,
adjustment, and replacement of damaged components are needed to maintain
effective long-term control of leaks.  No significant differences in
control equipment or work practices could be defined to account for
the incremental improvement in control performance associated with
Regulatory Alternative II.  Therefore, the total cost listed in Table
8-16 is attributed to the regulatory baseline, and no additional cost
is estimated for Regulatory Alternative II for new tall batteries.
8.1.5  Cost Effectiveness
     For the analyzed model batteries, the cost effectiveness (the
incremental total annualized cost divided by the incremental annual
emission reduction) of the regulatory alternatives is summarized in
Table 8-17.  The emission reductions were obtained from the estimates
for the model batteries in Chapter 7.  The mid-range of the emission
estimates in Tables 7-1 through 7-5 was used to calculate the total
annual BSO abatement in Table 8-17.
     The following discussion of the cost effectiveness of the alterna-
tives is based on the incremental change from the assumed baseline.
The cost effectiveness of Regulatory Alternative II for charging
ranges from $130 to $250 per kilogram of BSO removed.  The most cost
effective alternative appears to be Regulatory Alternative II for door
leaks, which has a cost of $28 to $150 per kilogram of BSO removed.
The higher cost effectiveness for small hand-luted batteries ($150 per
kilogram of BSO) results from the need for two additional lutermen on
a low capacity battery.  The zero cost effectiveness for Model 3
results from identifying no additional controls required for the new
sources to meet an emission limit of 12 PLD.  The cost effectiveness
                                  8-27

-------
           TABLE 8-16.   COST ESTIMATE FOR DOOR LEAKS REGULATORY
                        ALTERNATIVE I — NEW SOURCES
                        (thousands of 1979 dollars)

        Operating requirements                                Model  3

Labor
  Primary maintenance (16 man-hours/door or 9,088 man-hours)    143
  Cleaning (1 man/shift or 8,760 man-hours)                     138
  Adjustment (1 man-hour/door/day or 10,366 man-hours)          163
  Inspection (8 man-hours/day or 2,920 man-hours)                46
  Seal replacement (5 man-hours/seal or 248 man-hours)            4
  Refractory repair (4 man-hours/door or 568 man-hours)           9
Seals, chuck door (40% seals replaced each year at $1,700 each)  97
Payroll overhead (10% of labor)                                  50
Plant overhead (50% of labor)                                   252

                         Total annual cost                      902

Dollars per megagram of coke                                      1.17

 Assumes no incremental capital requirements for new plants.
                                  8-28

-------
                           TABLE 8-17.   COST EFFECTIVENESS OF THE REGULATORY ALTERNATIVES
00
ro


Source
Charging

Door leaks

Topside leaks

NA = not appl
PLO = percent


Regulatory Model
alternative battery
I 1
2A
2B
2C
3
II 1
2
3
I 1
2A
2B.2C
3
II 1
2
3
I 1
2
3
II 1
2A
2B
3
i cable.
leaking offtakes.

Incremental
Emission total annual i zed
limit cost ($/yr)
34 s/charge 897,000
34 s/charge 906,000
34 s/charge 646,000
34 s/charge 624,000
34 s/charge 731,000
16 s/charge' 49,000
16 s/charge 66,000
16 s/charge 99,000
16 PLD 481,000
16 PLD 1,263,000
16 PLD 866,000
16 PLD 902,000
12 PLD 469,000
12 PLD 144,000
12 PLD
5 PLL, 10 PLO 325,000
5 PLL, 10 PLO 648,000
5 PLL, 10 PLO 490,000
3 PLL, 6 PLO 8,000
3 PLL, 6 PLO 32,000
3 PLL, 6 PLO 25,000
3 PLL, 6 PLO 25,000
PLD = percent leaking
PLL = percent leaking
Total annual
BSD abatement
from baseline
(Mg/yr)
0
0
0
0
0
0.2
0.5
0.6
0
0
0
0
3.1
5.2
6.3
0
0
0
0.25
0.70
0.50
0.88
doors.
lids.

Cost
effectiveness
($/kg of BSO)
NA
NA
NA
NA
NA
250
130
170
NA
NA
NA
NA
150
28
0
NA
NA
NA
30
46
50
28


-------
for the EPA alternative for topside leaks ranges from $28 to $50 per
kilogram of BSD removed.
8.1.6  Comparison of Estimated Costs with Actual Plant Costs
     The costs for retrofitting larry cars for stage charging were
estimated as $500,000.   Individual company estimates, which included
the cost of filtered air cabs and gooseneck cleaners, were $400,000,
$600,000, and $750,000 (1975 dollars).2  Other companies responding to
industry questionnaires reported estimates of $494,000 and $600,000
(1978 dollars).5  The cost of steam lines and nozzles was estimated as
$2,200 per oven.  Individual company estimates were $943, $1,000,
$1,500, and $1,345 (1975 dollars) per oven.2  Individual companies
reported that the addition of jumper pipes cost $1,500,  $2,940,  and
$3,330 (1975 dollars) per oven.2  A current estimate of $2,500 per
oven was used.
     The capital cost for new doors and jambs was reported by one
plant as $1.7 million for a 45-oven battery.  The capital cost estimate
for Model 2A, a 62-oven battery with 38 percent more doors, is $2.6
million.   Cost estimates of $2,000 for a luted door and $10,000  for a
3- to 4-meter self-sealing door were used.   One plant reported a cost
of $1,000 (1971 dollars) for hand-luted doors; other plants reported
that costs for individual self-sealing doors ranged between $7,000 and
$10,000 (1979 dollars).
     Many of the capital  cost estimates were derived from plant  esti-
mates.   These estimates include the estimate for new doors and jambs
discussed above and the cost of $13,800 per oven for modified doors
and seals.  Comparisons of actual cost with estimated costs are  limited
because the cost of many of the regulatory alternatives  is operating
cost and procedure cost intensive instead of capital cost intensive.
8.1.7  Summary of Nationwide Costs
     Nationwide costs were estimated on a battery-by-battery basis for
all existing coke plants by following the methodology outlined for the
model plants.  This battery-specific approach considers  the number of
doors,  cycle time, number of ovens, baseline regulations, and other
specific data for a particular battery to estimate the cost to attain
                                8-30

-------
the limits of Regulatory Alternative II.  Costs for each existing
battery are summed to estimate nationwide costs.
     The nationwide incremental capital and annual ized costs to attain
Regulatory Alternative II are summarized in Table 8-18.   These nationwide
costs represent the costs which may be incurred by those existing
batteries that have a baseline regulation that is less stringent than
Regulatory Alternative II.   For those batteries with existing SIP's or
consent decrees that are equal to or more stringent than Regulatory
Alternative II, the costs are attributed to the baseline and are not
included in the incremental costs given in Table 8-18.  For example,
approximately 82 batteries (out of a total of 177 batteries in 1982)
have existing charging regulations that are equal to or more stringent
than Regulatory Alternative II.  For lid, offtake, and door leaks,
approximately 111, 71, and 105 batteries, respectively,  would not
incur additional costs to attain the limits of Regulatory Alternative II.
     Regulatory Alternative II would require an incremental capital
investment of $45 million and an estimated total annualized cost of
$19 million per year.   Further information on baseline costs, Regulatory
Alternative II costs,  and economic impacts is given in Chapter 9.
8.2  OTHER COST CONSIDERATIONS
     A number of costs are currently imposed on coke-making facilities
in response to the Clean Air Act, the Water Pollution Control Act, the
Occupational Safety and Health Administration, State Implementation
Plans, and other environmental regulatory requirements.   Additional
information on these costs is presented in the following sections.
Any future costs attributed to the Resource Conservation and Recovery
Act are not included because these costs could not be identified at
this time.
8.2.1  Desulfurization
     Coke oven gas desulfurization, already required in some States,
limits the hydrogen sulfide (H2S) content to values ranging from 1.1
to 2.3 g/m3 (0.5 to 1.0 grains/ft3).   This limit corresponds to a
removal efficiency of about 75 to 90 percent for the typical H2S
content of coke oven gas.10  Although several removal processes are
                                 8-31

-------
 TABLE 8-18.   NATIONWIDE INCREMENTAL COSTS TO ATTAIN REGULATORY ALTERNATIVE II
                                (1979 dollars)

Emission point          Capital  cost ($)         Total  annualized cost ($/yr)

Charging                    6,500,000                     5,600,000

Topside leaks               5,300,000                     2,000,000

Door leaks                 33,300,000                    11,700,000

     Totals                45,100,000                    19,300,000
                                    8-32

-------
available, the vacuum carbonate process was used as an example to
estimate cost.  Capital cost and operating cost estimates for this
process, as applied to the model plants, are given in Table 8-19 as
$1.54 to $4.98 per megagram of coke.1  However, these estimates may be
greater than the actual cost because a plant would probably install
desulfurization capacity to handle coke oven gas from more than one
battery rather than from a single battery as indicated in the estim-
ates in Table 8-19.
8.2.2  OSHA
     Some OSHA costs which overlap air pollution control  requirements
were included in the regulatory baseline costs.  However, other costs
that are directed strictly to employee protection are associated with
OSHA regulations.   Estimates of these other costs (e.g.,  exposure
monitoring and protective equipment) indicate that they would require
capital expenditures of $3.12 per megagram of coke capacity and would
increase the annual operating cost by $1.92 per megagram of coke
produced.1  These figures were used in computing the OSHA costs for
each of the model  plants, as shown in Table 8-20.
8.2.3  Water Treatment
     By-product coke facilities include a by-product plant that produces
water effluents.   According to the Clean Water Act as amended in 1977,
these effluents must be treated using best practicable technology
(BPT) by July 1,  1977 and best available technology (BAT) economically
achievable by July 1, 1983.  EPA has investigated these water treatment
costs for coke plants.11  However, it is difficult to estimate costs
for the model batteries developed in Chapter 6.  For example, if Model
Battery 1 or 2 is a reconstruction, the water treatment facility may
already exist.  Also, if Model 3 is a new battery, it may only require
an increase in the capacity of an existing treatment facility.
     Considering the above uncertainties, it was decided  that the
conservative approach, which would not underestimate the  costs, was to
assume that a new treatment facility at BAT would be required for each
model.   In Reference 11, cost estimates for a BPT and BAT wastewater
treatment facility are given for a coke plant producing 4,270 Mg coke
                                   8-33

-------
                              TABLE  8-19.   ESTIMATED  CAPITAL  COST  OF  DESULFURIZATION3
                                            Model  1              Model  2              Model  3
                                        (106,000  Mg  of       (344,000 Mg  of       (770,000 Mg of
                                            coke/yr)            coke/yr)             coke/yr)
Total capital cost ($)
Total operating cost ($/yr)
Dollars per megagram of coke
2,573,000
528,000
4.98
5,025,000
807,000
2.35
7,967,000
1,183,000
1.54
       Estimates are based on information  in  Reference  1,  assumed  337  cubic  meters of coke oven gas
        produced per megagram of coal  (Reference  5).   Capital  recovery  is  not included in total  capital
        cost.
        Total  operating cost is exclusive of depreciation  or  capital  recovery.
00
i
"                  TABLE 8-20.   ESTIMATED CAPITAL AND  ANNUAL  OPERATING  COSTS  OF  OSHA COMPLIANCE3

Capital costb ($)
Annual operating costc ($/yr)
Model 1
(106,000 Mg of
coke/yr)
330,000
200,000
Model 2
(344,000 Mg of
coke/yr)
1,100,000
660,000
Model 3
(770,000 Mg of
coke/yr)
2,400,000
1,500,000
       aOSHA requirements that do not overlap  with  pollution  control  requirements.
        Capital  recovery costs are not included.
        Indirect operating costs are not included.

-------
oo
CO
en
                 TABLE 8-21.  COST ESTIMATES FOR A BIOLOGICAL  SYSTEM  WASTEWATER TREATMENT FACILITY
                                            (4,270 Mg/day  coke plant)
Element
Installed capital cost ($)
Direct operating cost ($/yr)
Operation and maintenance
Sludge disposal
Hazardous waste disposal
Energy and power
Steam
Chemicals
Subtotal
Annual ized capital cost ($/yr)
Land ($/yr)
Subtotal
Total annual ized cost ($/yr)
BPTa
5,230,000

183,000
2,600
22,300
41,700
523,000
127,000
900,000
470,000
8,700
479,000
1,380,000
BAT-lb
1,410,000

49,500
	
	
35,400
97,300
12,700
195,000
127,000
5,000
132,000
327,000
Total (July
1978 dollars)
6,640,000

233,000
2,600
22,300
77,100
620,000
140,000
1,095,000
597,000
13,700
611,000
1,710,000
Total C
(December 1979
dollars)
7,500,000

263,000
2,900
25,200
87,100
701,000
158,000
1,240,000
675,000
15^500
691,000
1,930,000
       BPT = Best practical control technology  currently  available.

       BAT = Best available technology economically  achievable.
      cBased on a factor of 1.13  from Chemical  Engineering  Plant  Cost  Index (247.6 -=- 219.2).
       are rounded.
All totals

-------
                         TABLE 8-22.  COST ESTIMATES FOR A WASTEWATER TREATMENT  FACILITY
                                FOR THE MODEL BATTERIES (December 1979 dollars)
00
I
GO
CTi
Element
Installed capital cost ($)
Direct operating cost ($/yr)
Operation and maintenance
Sludge disposal .
Hazardous waste disposal
Energy and power
Steam .
Chemicals
Subtotal
Annual ized capital cost ($/yr)a
Land ($/yr)D
Subtotal
Total annual ized cost ($/yr)
$/Mg coke

1
106,000 Mg
coke/yr
1,490,000

52,300
200
1,700
5,900
47,700
10,700
119,000
134,000
1,100
135,000
254,000
2.40
Model battery
2
344,000 Mg
coke/yr
3,030,000

106,000
640
5,600
19,200
155,000
34,900
321,000
273,000
3,400
276,000
597,000
1.74

3
770,000 Mg
coke/yr
4,910,000

172,000
1,400
12,400
43,000
346,000
78,100
653,000
442,000
7,700
450,000
1,100,000
1.43
       Based on 0.6 power scaling factor applied to Table 8-21.

       Based on a  linear scaling factor applied to Table 8-21.
       "All totals  are rounded.

-------
oo
 i
CO
              TABLE 8-23.   COST ESTIMATES FOR PUSHING, QUENCHING, AND STACK EMISSIONS—NEW SOURCES

                                          (Model 3 = 770,000 Mg/yr)6


Capital
Annual i
Dollars


cost ($)
zed cost ($/yr)
per megagram of coke

Pushing
5,550,000
580,000
0.75
Emissions
Quenching
170,000
13,000
0.02

Stack
1,750,000
277,000
0.36

-------
per day.   This estimate is given in Table 8-21 with a conversion to

1979 dollars.   The costs for the model batteries are estimated by

scaling the costs based on the model battery capacity.   For items such

as installed capital costs and operation and maintenance, the scaling

factor is based on the 0.6 power of the capacity ratio.   Other items,

such as steam and chemicals, are expected to be linearly related to

capacity; consequently, a linear scaling factor was used for these

items.   The cost estimates for the model batteries are given in

Table 8-22 and show a range of $1.43 to $2.40/Mg of coke.

8.2.4  Pushing, Quenching, and Stack Emission Controls

     Emission standards for pushing, quenching, and stack emissions

may generate additional costs for new sources.  Preliminary estimates

of these costs are given in Table 8-23.   These estimates were obtained

from cost functions in Reference 1 for reasonably achievable control

technology for new installations.

8.3  REFERENCES

1.   Technical Approach for a Coke Production Cost Model.  PEDCo
     Environmental, Inc.  Draft report prepared for Research Triangle
     Institute.  Research Triangle Park, NC.  December 1979.   101 p.

2.   Inflationary Impact Statement:   Coke Oven Emissions.  0.  B.
     Associates, Inc., for U.S.  Department of Labor, Occupational
     Safety and Health Administration.  Contract No. J-9-F-6-0015.
     November 5, 1975.

3.   Memorandum from Faber, P. V., Technical Director,  Wilputte Corpora-
     tion, to D. W., Coy, Research Triangle Institute.   May 4, 1978.

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

5.   Draft of Standards Support and Environmental Impact Statement
     Volume I:  Proposed National Emission Standard By-Product Coke
     Oven Wet-Coal Charging and Topside Leaks.  Emission Standards and
     Engineering Division.  U.S. Environmental Protection Agency.
     June 1978.

6.   Development of Air Pollution Control Cost Functions for the
     Integrated Iron and Steel Industry.  PEDCo Environmental, Inc.
     EPA Contract No.  68-01-4600.  July 1979.
                                 8-38

-------
7.    Baldwin, V., 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 No.  68-02-2612.  Task 39.  March
     1978.   112 p.

8.    Bee, R. W., et al.   Coke Oven Charging Emission Control Test
     Program—Volume I.   EPA-650/2-74-062.  July 1974.   163 p.

9.    Letter from Spawn,  Peter D.,  GCA Corporation,-to M. Branscome,
     Research Triangle Institute.   June 15, 1982.

10.   Gorman, P.  G., C. Mumma, and J.  Shum.  Study of Coke Oven Battery
     Stack Emission Control Technology, Final Report.  Volume I--Collec-
     tion and Analysis of Existing Emission Data.  MRI.  Kansas City,
     MO.  EPA Contract No.  68-02-2609 ( Task 5).  March 1979.

11.   Development Document for Effluent Limitation.   Guidelines and
     Standards for Iron and Steel  Manufacturing, Point Source Category.
     Volume II--By-product Cokemaking Subcategory.   EPA 440/1-82/024.
     May 1982.  p.  126 and 130.
                                  8-39

-------
                            9.   ECONOMIC IMPACT

     This chapter discusses the economic impacts of the regulatory
alternatives for topside, charging, and door leaks.  These alter-
natives are described in Chapter 6, and they would apply to new and
existing coke oven batteries.   The Regulatory Alternative II com-
bination of topside, charging,  and door leaks would increase the price
of furnace coke by 0.14 percent and reduce production by 0.16 percent.
Under this combination, the price of foundry coke would increase by
1.6 percent and production would decrease by 1.6 percent.   No furnace
or foundry coke battery is expected to close as a result of Regulatory
Alternative II.  The analysis  also addresses the potential impacts of
compliance with a comprehensive list of environmental and other
controls (see Section 9.3).  These impacts are measured against the
baseline state of control for  all sources.
     Section 9.1 presents a profile of the coke industry.   Section 9.2
contains the impact analysis,  and Section 9.3 presents potential
socioeconomic and inflationary  impacts.   Section 9.4 discusses the
impacts of varied economic conditions, and Section 9.5 presents
the references.
9.1  INDUSTRY PROFILE
9.1.1  Introduction
     Coke production is a part  of 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
STC 3312.  Furthermore, many furnace coke producers are fully integrated
                                  9-1

-------
iron- and steel-producing companies.   Regulation of coke production may
affect the entire blast furnace and steel mill industry, with special
emphasis on coke producers.
     The industry profile has two main purposes.   The first is to
provide the reader with a broad overview of the industry; in this
sense, the profile should be meaningful when read alone.  The second
purpose is to support an economic analysis through an assessment of
the appropriateness of various economic models to analyze the
industry.   Furthermore, the profile provides some of the data neces-
sary to the analysis.
     The industry profile comprises seven major subsections.  The re-
mainder of this introduction, which constitutes the first subsection,
provides a brief, descriptive, and largely qualitative examination of
the industry.   The remaining six subsections of the profile conform to
a particular model of industrial organizational analysis.  This model
maintains that an industry can be characterized by its basic condi-
tions, market structure, market conduct, and market performance.
     The basic conditions in the industry, discussed in the second and
third subsections of this profile, are believed to be major deter-
minants of the prevailing market structure.   Most important of these
basic conditions are supply and demand conditions.   Supply conditions
are largely technological, while demand conditions are determined by
product attributes.
     The market structure and market conduct of the blast furnace and
steel mill industry are examined in the fourth subsection.   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 concerns price and nonprice behavior of sellers.  Of
particular interest is the degree to which industry pricing behavior
can be approximated by the competitive pricing model, the monopoly
pricing model, or some model of imperfect competition.
     The fifth subsection of the industry profile addresses market
performance.  The historical record of the industry's financial
                                  9-2

-------
performance is examined, with some emphasis on its comparison with
other industries.   The sixth subsection of the industry profile presents
projections of key variables such as coke production and steel produc-
tion.  The seventh and final subsection describes market behavior.
     9.1.1.1  Definition of the Coke Industry.   Coke production is a
part of SIC 3312--Blast Furnaces and Steel Mills.   SIC 3312 includes
establishments that produce coke and 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 also are included in SIC 3312.x  In 1980, the total  value of
coke production was less than 10 percent of the total  value of shipments
from SIC 3312.  Total  coke production in 1980 was valued at about
$4,648,413,000.2  Total value of shipments from SIC 3312 in 1980 was
$50,303,900,000.3
     Coke is produced in two types of establishments:   merchant plants
and captive plants.  Many merchant plants, which produce coke to be
sold on the open market, are owned by chemical  companies or companies
other than iron and steel companies.  Captive plants,  which are
vertically integrated with iron and steel companies, use coke in the
production of pig iron.  The majority of coke plants in the United
States are captive plants; at the end of 1979,  20 plants (17 loca-
tions) were merchant plants and 45 plants (43 locations) were captive
plants.   Merchant plants accounted for only 9 percent  of total coke
production for that year.4
     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
comprised less than 1 percent of the gross national product (GNP).5 6
During most of the 1950's, coke production was  about 0.30 percent of
the GNP, and during the 1960's and until  the mid-1970's coke produc-
tion was only about 0.20 percent or less of the GNP.   However, in 1974
coke production rose to above 0.30 percent of the GNP,  a trend that
                                  9-3

-------
continued for the next 2 years.  By 1979, coke production was about
0.2 percent of the GNP.7 8
     Historically, more coke has been exported than imported by the
United States, but that trend may be changing.  The values of all U.S.
imports and exports and U.S. coke imports-and exports are shown in
Table 9-1.  From 1950 to 1972, coke exports were much greater than
coke imports, but after 1973, this trend reversed.   The same pattern
applies to 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 1979, this trend reversed, and coke
imports were a larger proportion of total U.S. imports than coke
exports were of total U.S.  exports.
     U.S.  coke production always has been a substantial portion of
world coke production.  This share has decreased during the past 30
years, as indicated in Table 9-2.   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.   This growth has
fluctuated little; e.g., as shown in Table 9-3,  the 1965 value of
shipments of SIC 3312 was the highest 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.
     Table 9-4 shows the value added by manufacture, the total  number
of employees, and the value added per employee for SIC 3312.   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 shipment value for this  industry was
highest.   The increasing value added per employee might indicate that
this industry is changing to a more capital-intensive production
process.
                                  9-4

-------
                                    TABLE 9-1.   COKE  INDUSTRY FOREIGN  TRADE6  9  10  37  38  39
        Total  U.S.  imports
Year      ($ billions)3
 Coke imports
for consumption
 ($ millions)
Coke imports
as a share of
total imports
U.S.  exports
($ billions)1
Coke exports.
($ millions)'
NA = not available.

 Current dollars.
Coke exports
as a share of
total exports
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
^963

-------
               TABLE 9-2.   COKE PRODUCTION IN THE WORLD2 4 6 11
Year
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
World production
(106 Mg)
182.3
204.1
208.9
225.6
211.5
242.3
256.8
266.1
255.0
260.4
279.7
272.0
272.9
281.7
298.5
310.3
310.4
303.9
315.8
335.8
350.5
342.7
340.5
365.8
367.4
363.3
367.2
373.5
364.7
341.0
NA
U.S. production
(106 Mg)
65.9
71.9
62.0
71.5
54.4
68.3
67.6
69.0
48.6
50.7
51.9
46.9
47.1
49.3
56.4
60.7
61.2
58.6
57.8
58.8
60.3
52.1
54.9
58.4
55.9
51.9
52.9
48.5
44.5
48.0
41.8
U.S. production
as a share of
world production
(%)
36.1
35.2
29.7
31.7
25.7
28.2
26.3
25.9
19.1
19.5
18.6
17.2
17.3
17.5
18.9
19.6
19.7
19.3
18.3
17.5
17.2
15.2
16.1
16.0
15.2
14.3
14.4
13.0
12.2
14.1
NA
NA = not available.
aOven and beehive coke combined.
                                  9-6

-------
TABLE 9-3.   VALUE OF SHIPMENTS, SIC 33123 12 13 40
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
In current dollars
(io6)
15,783.8
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
46,025.3
52,433.7
50,303.9
In 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
30,673.3
32,213.4
28,362.6
                 9-7

-------
TABLE 9-4.   VALUE ADDED, SIC 331212 13 41 42
Value added by manufacture
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
In current dollars
(106)
6,844.4
6,546.3
6,620.9
7,506.4
8,479.6
9,379.8
9,643.6
8,910.1
9,275.8
9,853.2
9,350.5
9,563.1
10,304.7
12,769.4
17,425.8
13,356.2
14,755.5
15,021.4
19,085.7
21,039.0
18,632.2
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
12,719.6
12,925.6
10,505.3
Employees
do3)
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
Value added
per employee —
1972 dollars
(103)
18.1
18.8
18.7
20.9
21.9
22.3
22.5
21.2
21.1
21.1
19.4
20.7
22.0
24.0
29.0
23.3
24.4
24.1
29.6
27.7
22.7
            9-8

-------
9.1.2  Production
     9.1.2.1  Product Description.   Two types of coke are produced:
furnace coke and foundry coke.   Furnace coke is used as a fuel in
blast furnaces, and foundry coke is used as a fuel in the cupolas of
foundries.  Coke also is used for other miscellaneous processes such
as residential and commercial heating.   In 1978, only 2 percent of all
coke used in the United States was used for these miscellaneous
purposes, while 93 percent was used in blast furnaces, and the
remaining 5 percent was used in foundries.11  Time-series data for the
percentage of total U.S. consumption attributable to each use are
shown in Figure 9-1.
     9.1.2.2  Production Technology.   Coke typically is 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 coke's end use.  Both furnace
and foundry coke usually are produced from the carbonization of a
mixture of high- and low-volatile coals.   Generally, furnace coke is
produced from a coal  mix of 10 to 30 percent low-volatile coal, and
foundry coke is produced from a mix of 50 percent or more low-volatile
coal.   Furnace coke is coked an average of 18 hours; foundry coke 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 soda ash produc-
tion.   In such ovens, the by-products of carbonization (e.g., ammonia,
tar, and gas) are collected instead of being emitted into the atmos-
phere 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 operating 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
resulting oven deterioration.  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.14  In a report written for
                                  9-9

-------









Q
01
CO
D
01
O
O

o.
0
vO f~~*
1 ^
s §
DC
01
Q.









3M

92

90

00

86

84
r>
^ /'\
^.'' \ / FURNACE
^•v ."'' \ /
•\ ^..x ."- \ y 	 ' \ /
/N\ / ^x/ \./ v

" /\ /'"
_ / "«/
/
~ ^-*
- /'
•
L /
1 FURNACE


/
12


10


8

6


4

2



- \ OTHER USES
\
\
A
- ^
\
\
\
^\
W 	 	 	 ^x\ FOUNDRY
\ ^ ^—-^^- 	 — ^^^^\X^\

^^x7^ C*--'' ^\ /"~^\ V
v*— . 	 / \ /» / •
"*-^ ^x Nx/ OTHER
USES
i , i , i , i , i . t i i . l . l . l i 1 i I i I i I i — L_J 	 1
50   52   54   56   58   60   62   64   66    68   70   72   74   76   78   80
                                   YEAR
       Figure 9-1.  Uses of oven coke as pcrcents of total coke consumption.2-6'11

-------
the Department of Commerce, Father William T.  Hogan estimated the
potential annual maximum capacity of U.S.  oven coke plants as of July
31, 1979.1S  His estimates are shown in Table 9-5.   Hogan assumes that
almost 10 percent of his estimated total capacity will be out of
service at any given time; therefore, he subtracts out-of-service
capacity from total capacity to obtain maximum annual capacity.
     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 when the
coal mix is changed and the coking time increased.   Furthermore, some
ovens that produce foundry coke could be changed to produce furnace
coke when the coal mix is changed and the coking time decreased.
Also, some variation in the combination of flue temperature and coking
time is possible for either type of coke.   A shorter coking time
results in greater potential annual production.
     9.1.2.3  Factors of Production.   Table 9-6 provides a typical
labor and materials cost breakdown for furnace coke production.   Coal
is the major material input in coke production.  In 1979, greater than
61.0 percent of the coal received by coke plants was from mines that
were company owned or affiliated.17  In this same year, 14 States
shipped some coal to coke plants outside the State.18  Of the coal
received by domestic coke plants, over 81 percent came from West
Virginia, Kentucky, Pennsylvania, and Virginia.19  Any potentially
adverse impact on the coke industry probably will impact these States.
A total of 69.9 million megagrams of bituminous coal was carbonized in
1979.20
     Table 9-7 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
                                  9-11

-------
        TABLE 9-5.  POTENTIAL MAXIMUM ANNUAL CAPACITY OF OVEN COKE
               PLANTS IN THE UNITED STATES ON JULY 31, 197915

In existence
Furnace plants
Merchant plants
Total
Out of service
Furnace plants
Merchant plants
Total
In operation
Furnace plants
Merchant plants
Total
Number of
batteries
169
30
199
(18)
Jl)
(20)
151
28
179
Number of
ovens
10,076
1,337
11,413
(1,026)
(117)
(1,143)
9,050
1.220
10,270
Capacity
(Mg)
53,095,381
4,400,691
57,496,072
(5,255,001)
(460,599)
(5,715,600)
47,840,380
3,940,092
51,780,472
Batteries and ovens down for rebuilding and repair.
                                 9-12

-------
     TABLE 9-6.   TYPICAL COST BREAKDOWNS:   FURNACE COKE PRODUCTION AND
                    HOT METAL (BLAST FURNACE) PRODUCTION16

Furnace coke production

     Labor and materials                                      Percent of cost

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

     Total labor  and material costs                                100.0

Hot metal production                                          Percent of cost

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

     Total labor  and material costs                                100.0
                                  9-13

-------
            TABLE 9-7.  EMPLOYMENT IN THE BY-PRODUCT COKE INDUSTRY21 48
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
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
9,772
9,348
8,924
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
2.18
2.06
2.24
NA = not applicable.
 Figures for 1971-1980 are estimates.  See text for more detail
                                  9-14

-------
ratio of coke used in steel production.   These estimates are also
shown in Table 9-7.
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, coke consumption was fairly high; 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 con-
sumed annually.  Domestic coke consumption increased during the
mid-19601s to mid-1970's to an annual  figure of 57 million megagrams
but did not reach the 1950 to 1957 level.   The late 1970's show
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, which are sensitive to the
performance of the overall economy.  Cycles in coke demand are linked
to cycles in aggregate demand or cycles  in demand for particular
products like 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 coke supplements to heat the blast furnaces.   The
increased use of these supplemental fuels over the past 20 years has
caused the tons of coke used per ton of  pig iron produced (the coke
rate) to decrease.  Other factors that have caused the coke rate to
decline are the increased use of oxygen  in the blast furnaces and the
use of ores of higher metallic content.   Table 9-8 shows U.S.  pig iron
production,  coke consumed in the production of pig iron, and the coke
rate for 1950 to 1980.  (Data limitations make it difficult to cal-
culate the foundry coke rate in  cupola production.)
     Recently, some concern has  been raised about the ability of the
United States' cokemaking capacity to support domestic steel pro-
duction—the major source of coke demand.   The study conducted by
                                  9-15

-------
    70 r-
    65
o
o
u_

O   60
C/5
UJ
    50
    45
    35
                   J_
J_
                               _L
                                    _L
_L
                      JL
J-
                                                                _L
                                             _L
                                       J_
                                                                                      J_
                                                                   J
        50    52   54   56    58    60   62
                      64   66

                      YEAR
                68    70    72   74    76   78    80
                        Figure 9-2. United States apparent consumption of coke.6-42

-------
TABLE 9-8.   COKE RATE2 4 6 11 22 43
Year
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
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
78,904
62,324
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
53,575
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
45,862
37,583
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
0.58
0.60
          9-17

-------
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.23  Imports increased
dramatically in that same year.  Hogan and Koelble attribute this
decline in coke production to the abandonment of coke ovens for
environmental reasons and predict a severe coke shortage by 1982.24
This prediction is disputed in the Merrill Lynch Institutional Report
by Charles Bradford.   The Bradford report attributes the lack of
adequate U.S. coke production in 1978 to two factors:  (1) a coal
miner's strike that caused the drawing down of stocks of coke when
they should have been increasing, and (2) the premature closing of
some coke ovens that came under EPA regulation and that normally would
have been replaced before they were closed.25  The Bradford report
states that a survey of U.S. steel producers revealed that all of the
major steel producers are or soon will be self-sufficient with regard
to cokemaking capacity.26  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.  Furthermore,
there are no signs of a major shortfall in capacity occurring in 1982.
9.1.4  Market Structure
     Market power, the degree to which an individual producer or group
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 potential impacts of new regulations.  It is
important to determine whether the competitive pricing model (price
equal to marginal cost) adequately describes pricing behavior for coke
producers.
     Any analysis of market structure must consider industry charac-
teristics.  This analysis addresses the number of firms producing
coke; the production concentration in specific firms; the degree of
integration in coke production; the availability of coke substitutes;
and the availability of substitutes for commodities for which coke is
a production input.  Also, some information on past pricing in the
                                  9-18

-------
coke industry is presented.   These topics are considered together with
financial performance (see Subsection 9.1.5) and growth (see Subsec-
tion 9.1.6) in an assessment of market behavior (see Subsection
9.1.7).
     9.1.4.1  Concentration Characteristics and Number of Firms.   This
subsection describes various concentration measures that can be
computed for the furnace and foundry coke industries.   Normally,
concentration ratios are used to indicate the existence of market
power.   Although concentration ratios are a useful  tool for describing
industry structure, concentration should not be used as an exclusive
measure of market power.  Many other factors (e.g., availability  of
substitutes, product homogeneity, and ease of market entry) determine
a firm's ability to control  market price.
     As of December 1979, 33 companies operated by-product coke
ovens.27  Fifteen companies are integrated iron and steel producers
and eighteen are merchant firms.   These companies owned and operated
65 coke plants (60 locations); 45 of these plants (43 locations)  were
captive and 20 (17 locations) were merchant.   These companies,  their
plant locations, the major uses of coke at each plant, and plant  coke
capacities are listed in Table 9-9.
     Reported capacities in Table 9-9 are maximum,  nominal figures
that do not allow for outage like that estimated for the overall
industry in Table 9-5.   All  of the largest plants are captive,  and
most of the merchant plants have small capacities.   Furnace coke
production is concentrated in captive plants.   Virtually all  of the
coke that was used in foundries and for other industrial uses was
produced by merchant plants.   If coke plant sites are ranked according
to capacity, the top 5 plant sites and top 10 plant sites have
30.0 percent and 44.6 percent of total coke capacity,  respectively.
(A plant site or location may include more than one complete plant.)
     By-product coke plants are concentrated in States bordering  the
Ohio River; this concentration may be caused by the location of coal
in that area.  Figure 9-3 shows the number of coke  plants in each
State.   Pennsylvania, Ohio,  and Indiana have 12, 14, and 6 plants,
respectively.

                                  9-19

-------
                          TABLE  9-9.   COKE  PLANTS  IN THE  UNITED  STATES, JANUARY  198028  30  4S
Company
Armco , Inc.
Bethlehem Steel Corp.
CF&I Steel Corp.
Crucible Steel , Inc.
^P Cyclops Corp.
o (Empire-Detroit)
Ford Motor Co.
Inland Steel Co.
Interlake, Inc.
J&L Steel Corp.
Kaiser Steel Corp.
Plant location3
Hamilton, OH
Houston, TX
Middletown, OH (2)
Bethlehem, PA
Burns Harbor, IN
Johnstown, PA
Lackawanna, NY
Sparrows Point, MD
Pueblo, CO
Midland, PA
Portsmouth, OH
Dearborn, MI
E. Chicago, IN (3)
Chicago, IL
Aliquippa, PA
Campbell , OH
E. Chicago, IN
Pittsburgh, PA
Fontana, CA
Classification
of plant
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Major uses of coke
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Coke capacity
(103 Mg/yr)
542
306
1,772
2,067
1,921
381
1,623
3,082
888
414
397
1,409
3,711
581
1,215
1,089
1,377
1,788
1,540
Footnotes on last page of table.
(continued)

-------
TABLE 9-9.   COKE PLANTS
IN THE UNITED STATES,  JANUARY 198028  30 45
    (continued)
Company
Lone Star Steel Co.
National Steel Corp.
Republic Steel Corp.
U.S. Steel Corp.
Wheel ing-Pittsburgh
Alabama By-Products
Corp.
Allied Chemical Corp.
Footnotes on last page
Plant location3
Lone Star, TX
Granite City, IL
Detroit, MI
Weir ton, WV
Brown's Island, WV
Cleveland, OH (2)
Gadsden, AL
Massillon, OH
S. Chicago, IL
Thomas, AL
Warren, OH
Youngstown, OH
Clairton, PA (3)
Fairfield, AL
Fairless Hills, PA
Gary, IN
Lorain, OH
Provo, UT
E. Steubenvil le, WV
Monessen, PA
Tarrant, AL
Ashland, KY
of table.
Classification
of plant
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Captive
Merchant
Merchant

Major uses of coke
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Blast furnace
Foundry,
other industrial
Blast furnace

Coke capacity
(103 Mg/yr)
506
712
1,928
1,111
1,095
1,755
757
166
439
315
943
875
6,017
1,818
914
4,220
1,491
1,158
1,507
489
708
961
(continued)

-------
                      TABLE  9-9.   COKE  PLANTS
IN THE UNITED STATES,  JANUARY 198028 30 45
    (continued)
Company
Carondelet Coke
Company
Chattanooga Coke and
Chemical Co.
Citizens Gas and Coke
Utility
Detroit Coke
Donner-Hanna Coke Corp.
vo
ro Empire Coke Co.
Erie Coke and Chemicals
Indiana Gas and
Chemical
Ironton Coke Corp.
(McLouth Steel)
Keystone Coke Co.
Jim Walter
Koppers Co. , Inc.

Plant location3
St. Louis, MO
Chattanooga, TN
Indianapolis, IN
Detroit, MI
Buffalo, NY
Holt, AL
Painesville, OH
Terre Haute, IN
Ironton, OH
Swedeland, PA
Birmingham, AL
Erie, PA
Toledo, OH
Woodward, AL
Classification
of plant
Merchant
Merchant
Merchant
Merchant
Merchant
Merchant
Merchant
Merchant
Merchant
Merchant
Merchant
Merchant
Merchant
Merchant
Major uses of coke
Foundry,
other industrial
Foundry,
other industrial
Foundry
Blast furnace, foundry
Blast furnace
Foundry
Foundry
Foundry,
other industrial
Blast furnace, foundry
Foundry
Blast furnace, foundry
Foundry, other
industrial
Foundry
Blast furnace, foundry
Coke capacity
(103 Mg/yr)
232
130
476
616
860
120
149
132
858
401
994
207
157
562
Footnotes on last page of table.
                                                        (continued)

-------
                         TABLE 9-9.   COKE PLANTS IN THE UNITED STATES,  JANUARY 198028 30  45
                                                     (continued)
CO
   Company
                          Plant location'
Classification
    of plant
Major uses of coke
Coke capacity
 (103 Mg/yr)
Milwaukee
Philadelpl
Sol vay
iia Coke
Milwaukee,
Philadelph
WI
ia, PA
Merchant
Merchant
Foundry, other
industrial
Foundry, other
195
291
     (Eastern Assoc.  Coal
      Corp.)

   Shenango, Inc.

   Tonawanda Coke Co.
                          Neville  Island,  PA

                          Buffalo,  NY
    Merchant

    Merchant
  industrial


  Blast furnace, foundry

  Foundry
     397

     298
aNumbers in parentheses indicate the  number of plants  at  that  location.   If  no  number  is  indicated,  only one
 plant exists at that location.
 An end use is considered a major use if it is at  least 20  percent  of  the plant's  total distribution of coke.

 Residential  and commercial  heating included in other  industrial  category.

-------
to
ro
                                                 Figure 9-3.  Coke plants in the United States, 1980.1*5

-------
     Table 9-10 divides the United States into 11 coke-consuming and
producing regions and shows the amount of coke produced in each region
and the locations of coke consumption.  Most of the regions produce
the bulk of the coke they consume; only three regions produce less
than 80 percent of their own consumption.  Likewise, only one region
produces more coke to be used in other regions than for its own
consumption.   Transporting coke long distances is avoided whenever
possible to reduce breakage of the product into smaller, less valuable
pieces and to minimize freight charges.
     The concentration of production or capacity in specific firms may
have economic importance.   Table 9-11 presents the percent of total
capacity owned by the largest four (of thirty-three) firms.   The
four-firm concentration ratio for the coke industry is 54.4 (the top
four firms own 54.4 percent of total capacity).   The concentration of
coke production has changed little since 1959, when the four-firm
concentration ratio was 53.5.29
     In the preceding discussion, furnace and foundry coke production
are considered jointly.   However, each existing coke battery may be
considered a furnace or foundry coke producer, based on the battery's
primary use.   Separate capacity-based concentration ratios for the two
types of coke are calculated based on this allocation.   The 1980
four-firm concentration ratio is 60.0 for furnace coke and 57.8 for
foundry coke.
     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 pro-
duction attributable to the top four firms declined from 62 to 53
percent.31
     In summary,  concentration exists in the production of both types
of coke and in steel  production.   However, concentration is not
sufficient to guarantee market power, and many companies are involved
in the production of both coke and steel products.   Other factors must
be considered in  any final  assessment of market power.
                                  9-25

-------
                            TABLE  9-10.   INTERREGIONAL  COKE  SHIPMENTS  IN 198046
                                                   (103 Mg's)
==————
Producing
region Ala.
Alabama 1.948
California. 0
Colorado,
Utah
Maryland, 0
New York
Illinois 0
Indiana 0
*P Kentucky. 0.9
ro Missouri.
°^ Tennessee,
Texas
Michigan. 0
Wisconsin
Ohio 0
Pennsylvania 0
Virginia. 0
West Virginia
TOTAL3 1.949

Calif.,
Utah, Md.
Colo. N. V. ' 111.
39 <0.5 155
2.224 0 0


<0.5 3,152 0

0 0 1,534
5.4 <0.5 33
12 0.9 6



6 0 216

0 0 80
0 24 383
0 3.6 4.5

2.287 3.181 2.411
Consuming region 	
Ky. , Mo. ,
Ind. Tenn. , Tex. Hlch.
5.4 225 51
0 1.8 0


189 0 4.5

000
7.987 34 61
211 1,135 42



9 6 2.286

2.7 407 305
181 89 30
5.4 1.8 • 327

8,590 1.900 3,107

Minn..
Wise.
63
0


0

0
20
5



89

15
1.8
0

194

Ohio
127
0


0.9

0
36
17



31

4.863
1.093
6.3

6,174

Pa.
18
0


5.4

0
1.8
<0.5



<0.5

118
7.796
247

8.187

Va. ,
V. Va.
16
0


0

0
0.9
10



0

<0.5
19
1,922

1,968

TOTAL
2.648
2.226


3.352

1.534
8.179
1.440



2.643

5.791
9.617
2.518

39.948
aNot in* Stales  are listed.   Minor shipments lo other States bring the  lota)  interregional  coke shipment to 41.053,000 Hg

-------
TABLE 9-11.   PERCENT OF COKE CAPACITY OWNED BY TOP FIRMS
                   (JANUARY 1980)27
Company
U.S. Steel, Inc.
Bethlehem Steel Corp.
J&L Steel Corp.
Republic Steel Corp.
Sum of four larges
Capacity
(103 Mg)
14,002
7,651
5,469
5,250
t firms 32,372
Percent of
total capacity
23.52
12.85
9.19
8.82
54.38
                        9-27

-------
     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 ver-
tical integration, the firm substitutes intrafirm transfers for pur-
chases from suppliers and sales to distributors.  A firm may seek to
supply its own materials inputs to ensure a stable supply schedule or
to protect itself from monopolistic suppliers.   The firm may seek to
fabricate or distribute further its own products to have greater
control over the consuming markets or to lessen the chance of being
shut out of the market by large buyers or middlemen.   Therefore, the
presence of vertical integration may constitute a firm's attempt to
control costs or ensure input supplies.   Vertical integration does not
guarantee market power (control over market price).
     Many coke-producing firms, especially furnace coke producers, are
vertically integrated enterprises.   As previously mentioned,  45 of the
existing coke plants are captive;  i.e.,  they are connected with blast
furnaces and steel mills.   In addition,  many coke firms own coal
mines, and greater than 61.0 percent of the coal used in ovens was
from captive mines in 1979.17  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.
This furnace coke is consumed by the producing  company.   Accordingly,
the impact analysis for furnace coke (see Subsection  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
substitution 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.   Furthermore, the recent trend toward
                                  9-28

-------
electric arc furnaces and minimi 11s has increased ease of entry into
the iron and steel industry.  Ease of entry reduces market power.
     Imported coke also can be substituted for domestically produced
coke.   In fact, coke imports.have increased recently; however, U.S.
iron and steel producers prefer to rely on domestic sources of coke.
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.
                                        \
     Furthermore,  substitutes  exist for the final products (iron and
steel) to which coke is an input.   Increases in the price of coke and
resulting increases in  the price of iron and steel products can lead
to some substitution of other  materials for iron and steel.   This
situation also reduces  market  power in coke production.   In regard to
foundry coke, analagous substitutions are possible.   Cupola production
of ferrous products, which uses foundry coke,  has competition from
electric arc furnaces that do  not use coke.  Hence,  there is a tech-
nological substitute for foundry coke in the manufacture of ferrous
products.  Furthermore, imported foundry coke  can be substituted for
domestic foundry production.  In conclusion, some substitution for
coke is possible in the manufacture of both steel and ferrous
products.
     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 collects annual data on coke production
and consumption and gives the  quantity and total  value of coke con-
sumed by producing industries,  sold on the open market,  and imported.
Dividing total value by quantity yields an average price for each of
these three categories  of coke.   Time-series data on these three
average values are given in Table 9-12.   (Furnace and foundry coke are
combined in these  figures.)
     Also shown in Table 9-12  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
                                  9-29

-------
     TABLE 9-12.   COMPARISON OF COAL PRICES AND DOMESTIC AND IMPORTED
                             COKE PRICES2 4 6 11 38
Year
Average value of
coal carbonized.
in coke ovens '
     ($/Mg)
Average value of
 oven coke used
  by producers3
      ($/Mg)
Average value of
 oven coke sold   Average value of
  commercially     imported coke )C
     ($/Mg)            ($/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
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
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
106.05
112.42
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
117.53
117.54
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
87.13
 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).
                                  9-30

-------
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, the trend
indicates that the industry can pass through some cost increases.
     9.1.4.5  Market Structure Summary.   The preceding sections
addressed the issue of market power in the coke industry through
concentration, integration, substitution, and historical price trends.
While no perfect method exists for measuring the extent of market
power,, these various characteristics can be used to study the po-
tential for market power.   Concentration statistics indicated that
some potential for market power exists in the coke industry.  However,
these statistics are not conclusive proof of such power.  Similarly,
vertical integration in the steel industry does not identify con-
clusively 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
cokemaking industry.
9.1.5  Financial Performance
     Financial data on many of the coke-producing firms or their
parent firms, including captive and merchant furnace and foundry
producers,  are shown in Table 9-13.   (Data for other firms were not
available.)  From the financial data in Table 9-13, three ratios have
been calculated for each company (see Table 9-14).   The first, a
liquidity ratio, is a measure of a firm's ability to meet its current
obligations as they are due.  A liquidity ratio above 1.0 indicates
that the firm is able to pay its current debts with its current
assets; the higher the ratio, the bigger the difference between
current obligations and the firm's ability to meet them.  All of the
coke-producing firms have liquidity ratios between 1.0 and 3.0.  These
figures are consistent with liquidity ratios for firms in a wide
variety of manufacturing industries.
     The second ratio, a coverage ratio, indicates the firm's ability
to meet its interest payments.  A high ratio indicates that the firm
                                  9-31

-------
                 TABLE  9-13.   FINANCIAL INFORMATION ON COKE-PRODUCING FIRMS I96032  45
                                               (103 1980 $)a
Company
Armco, Inc.
Bethlehem Steel Corp.
CF&I Steel Corp.
Crucible, Inc.
(Colt Industries, Inc.)
Cyclops Corp.
Ford Motor Co.
Inland Steel Co.
Interlake, Inc.
Jones & Laughlin
Steel , Inc.
Kaiser Steel Corp.
National Steel Corp.
Northwest
Industries, Inc.
to Republic Steel Corp.
^ U.S. Steel Corp.
ro Wheeling-Pittsburgh
Steel Corp.
Alabama By-Products0
Allied Corp.
Diamond Shamrock Corp.
Jim Walter Corp.
Koppers Co. , Inc.
McLouth Steel Corp.
Philadelphia Coke
Co. , Inc.
(Eastern Gas &
Fuel Assoc. )
Net sales
5,678,000
6,743,000
614,003
2,165,602

978,976
37,085,500
3,255,898
1,055,883
3,443.157

884,942
3,706,658
2,876.400

3.760,042
12,492,100
1,100,061

201,072
5,519.000
3,143,040
1,900,977
1,929,190
614,012
1,002,727



Net
working
capi tal
851,400
756.600
43,919
540.072

130,775
487,000
346,963
181,145.
765,411

400,637
454,419
491,500

394,488
1,329,700
157,432

33,559
467,000
488,530
704,152
329,006
52,927
72,075



Current
assets
1,685,800
1,800,200
138,581
924,917

276,357
11,559,000
956,672
369,532.
702,359°

716,512
1,235,048
936,300

1,002,283
3,898,800
394,940

61,319
1,649.000
911,553
1,365,029
641,257
188,469
241,251



Current
liabilities
834,400
1,043,600
94,662
384,845

145,582
11,072,000
609,709
188,387
536,948°

315.875
780,629
444,800

607,795
2,569,100
237,508

27,760
1,182,000
423,023
660,877
312,251
135,542
169,176



Annual
interest
expense
59,800
67.800
5,646
24,345

13,579
432,500
46,972
15,747
38.187

26,317
58.388
88,200

17,775
208,900
12,206

9,857
81,000
64,374
83,297
31,190
17,535




Total
assets
3,805,400
5,206,800
390,199
1,438,153

457,583
24,347.600
2,958,242
703,618
2,110,171

1,360.434
3,446,674
2,271,700

3,016,832
11,747,600
983,372

243,977
4,538,000
2,793,048
2,226,210
1,385,888
446,085
1,020,462



Long-
term debt
440,400
209,700
44,143
284,114


2.058,800
805.406
133,020
354,697

267,511

564.900

643,538
2,401,300
279,752

89,549
650,000
809,847
668,153
296.151
166,092
226,804



Tangible
net worth

2,622,000
206,228
677,636


8,567,500
1,299,362
336,707


693,524
1,444,092


1,507,816

409,938

105,916
1,664,000
1,241,714

634,475

434,782



 Blanks in the table indicate values that were unavailable or unclear in the reference.
bParent firms of captive coke plants are listed first, followed by  parent firms of merchant producers.

 Figures are for year 1978.

-------
            TABLE 9-14.   FINANCIAL  RATIOS FOR COKE-PRODUCING FIRMS3
Company Liquidity ratio
Armco, Inc.
Bethlehem Steel Corp.
CF&I Steel Corp.
Crucible Steel , Inc.
Cyclops Corp.
Ford Motor Company
Inland Steel Co.
Interlake, Inc.
J&L Steel Corp.
Kaiser Steel Corp.
National Steel Corp.
Northwest Industries, Inc.
Republic Steel Corp.
U.S. Steel Corp.
Wheeling-Pittsburgh
Alabama By-Products
Allied Chemical Corp.
Diamond Shamrock Corp.
Jim Walter
Koppers Co. , Inc.
McLouth Steel
Philadelphia Coke
1.99
1.60
1.57
2.51
1.75
1.33
1.67
1.92
1.27
1.43
1.71
2.30
2.03
1.67
1.63
2.21
1.43
1.96
1.98
2.24
1.54
1.54
Coverage ratio
6.95
4.59
3.02
6.46
9.82
15.26
5.75
2.17
2.63
2.70
4.83
5.73
8.15
2.31
2.51
2.95
4.90
4.36
3.46
10.76
1.88
1.73
Leverage ratio
1.97
2.09
1.92
2.24
2.22
2.28
2.07
2.15
2.48
2.17
2.33
2.38
1.83
2.00
2.22
2.30
2.54
2.34
3.02
1.99
2.51
2.48
aFigures are for year 1978.
                    Earnings  before  interest

'Coverage ratio  .  Annual  ^teresfexpense


d,           ..       Total  liabilities
 Leverage ratio  =  Tangible  net worth   '
                                 9-33

-------
is more likely to be able to meet interest payments on its loans.
This ratio can also be used to determine a firm's ability to obtain
more loans.  The coverage ratio for 10 of the predominantly coke-
producing firms ranged from 1.7 to 3.5.  Such ratios are close to the
coverage evidenced in most manufacturing industries.
     The last of the ratios, a leverage ratio, indicates the rela-
tionship between the capital contributed by creditors and that con-
tributed by the owners.  Leverage magnifies returns to owners.
Aggressive use of debt increases the chance of default and bankruptcy.
The chance of larger returns must be balanced with 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., additions to debt in the future are less
likely.  The firms with cokemaking capacity had leverage ratios that
ranged from 1.8 to 3.0.  These figures are relatively high among
leverage ratios for firms in many manufacturing industries.   Firms
with cokemaking capacity are engaged in substantial amounts of debt
financing.
     Another measure of financial  performance is the rate of return on
equity, which studies of the iron and steel industry show to be low.
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 projects a rate of return on equity of 1.0 percent for 1980
to 1990.33  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 financing future investments,
including environmental control expenditures.   For the steel  industry,
issuing new stock to raise investment capital  is unlikely under
current circumstances.   If environmental and other control investments
                                  9-34

-------
cannot be financed through new equity, another source of funds must be
found.  Increased debt is one potential source.  However, firms with
cokemaking capacity already have incurred substantial debt.   The TBS
analysis concludes that, to avoid deterioration in its financial
condition, the steel industry is likely to reduce expenditures to
modernize production facilities rather than increase its external
financing.34
9.1.6  Projections
     The demand for furnace coke (which is the largest share of coke)
is derived from the demand for steel produced by processes that use
coke.  Hence, projections of the future production of steel  by process
type are a necessary precursor to development of projections of furnace-
coke production and furnace-coke capacity requirements.
     Initially, projections of steel production developed by Data
Resources Incorporated (DRI) in 1979 were used to estimate coke projec-
tions.  However, the DRI projections were for years up to 1984, whereas
projections of the economic impact of Regulatory Alternative II are
desired for the period beyond 1984.   A revised projection of steel
production by process type (basic oxygen furnace, open hearth, and
electric arc) for 1985, 1990, and 1995 has been developed and is
presented in Table 9-15.   This projection is based on two sources:
     1.   Environmental Policy for the 1980's:   Impact on the American
          Steel Industry, Arthur D.  Little,  Inc., 1981.
     2.   Memorandum from Don Anderson, Economics Department, Research
          Triangle Institute, to Dave McLamb,  U.S.  Environmental
          Protection Agency, November 20, 1981.
When projections for the year 1995 were developed,  it was assumed that
the growth of the projected variables between  1990 and 1995  would be
the same as the growth pattern between 1980  and 1990 presented in the
Arthur D.  Little study.
     Table 9-16 presents the projection of furnace-coke  consumption,
furnace-coke production,  and furnace-coke imports for 1985,  1990, and
                                  9-35

-------
              TABLE 9-15.   SUMMARY OF STEEL INDUSTRY PROJECTIONS
Projections

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Variable
U.S. steel production3
Proportion, basic oxygen furnace3
Proportion, electric arc furnace
Proportion, open hearth furnace
U.S. steel consumption
Steel imports
Steel exports
Labor productivity
Producer price index of steel
nri.ll products (1967 = 100)
Producer price index of ferrous
scrap (1967 = 100)
1985
119.4
63
30
7
141.1
23.9
2.2
347
288
325
1990
125.5
63
33
4
150.3
27.3
2.5
411
324
383
1995
131.6
62
35
3
159.5
30.7
2.8
476
360
442
Note:   Figures for variables 1 and 5 through 7 are given in million megagrams.
       Figures for variables 2 through 4 are in percent.
       Figures for variable 8 are in thousand megagrams per employee.

 Based on estimates given in Environmental Policy for the 1980's:   Impact on
 the American Steel Industry, Arthur 0.  Little, Inc. , 1981.

 Steel consumption = steel production + steel imports - steel exports.
                                     9-36

-------
                 TABLE 9-16.  SUMMARY OF COKE INDUSTRY PROJECTIONS3

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Variable
World coke production
Scenario 1 furnace coke production
Scenario 2 furnace coke production
Foundry coke production
Furnace coke consumption
Foundry coke consumption
Scenario 1 furnace coke imports
Scenario 2 furnace coke imports6
Foundry coke exports
VMP of coke9

1985
427.0
39.2
39.2
2.9
42.8
2.5
3.5
3.5
0.4
208
Projections
1990
460.3
38.4
39.6
2.9
43.1
2.5
4.7
3.5
0.4
263

1995
493.5
37.6
40.0
2.9
43.5
2.5
5.9
3.5
0.4
317
Note:   Figures for variables 1 through 9 are given in million megagrams.
       Figures for variable 10 are in current dollars per megagram.

 The projection methodology includes no explicit assumption of additional
 controls like those assessed in this report.   Projections are based on normal
 growth and intended to represent long-run trends in the industry.

 Scenario 1 assumes imports to grow at the long-term trend and furnace coke
 production = furnace coke consumption - Scenario 1 furnace coke imports.

cScenario 2 assumes imports for 1990 and 1995 at the 1985 trend level  and
 furnace coke production = furnace coke consumption - Scenario 1 furnace
 coke imports.

 Scenario 1 assumes imports to grow at the long-term trend and all  coke
 imports are assumed furnace coke.

eScenario 2 assumes imports for 1990 and 1995 at the 1985 trend level  and
 all coke imports are assumed furnace coke.

 All coke exports are assumed foundry coke.
     stands for value of marginal  product.   This is a measure of the implied
 price of furnace coke based on its value in the production of steel products.
 Historical estimates of VMP were. based on econometric analysis of production
 functions for steel.
                                       9-37

-------
1995.  The projected furnace-coke consumption is based on a contin-
uation of historical trends of furnace-coke consumption in hot-steel
production (steel produced in basic oxygen and open-hearth furnace
processes) and the projected steel production presented in Table 9-15.
Furnace-coke capacity requirements are projected assuming a capacity
utilization rate of 85 percent by the coke producers during the
period.
     The projected coke capacity requirement is sensitive to the level
of coke imports.   Hogan and Koelble (Analysis of the U.S.  Metallurgical
Coke Industry, Industrial Economics Research Institute, 1979) assert
that coke suppliers in western Europe and Japan, which are the major
foreign coke suppliers to the U.S. steel industry, are not likely to
export substantial additional quantities to the United States.   This
is in spite of the fact that U.S.  coke imports have been growing
steadily.   If so, coke imports for 1990 and 1995 are more likely to
remain at about the 1985 level of imports of 3.5 million megagrams.
In Table 9-17, the coke capacity requirements (furnace coke plus
foundry coke) are projected under two scenarios.  Scenario 1 is the
long-run capacity requirement projection, which assumes that imports
will increase as shown in Table 9-16; and Scenario 2 is the capacity
requirement projection, which assumes coke imports at the projected
1985 level through 1995.
     Forecasts of U.S.  coke consumption are sensitive to forecast
steel production and technology.   Other projections have been made of
domestic coke consumption in 1985.  In the Merrill Lynch Institutional
Report, Charles Bradford forecasts furnace-coke consumption for 1985
at between 38.1 and 43.5 million megagrams.47  If furnace coke con-
sumption is assumed to be 91 to 93 percent of total coke consumption
(figures for recent years), this corresponds to a forecasted total
coke consumption of 41 to 46 million megagrams.   The projection
presented in this chapter appears reasonable:   First the projected
furnace coke consumption of 43.5 million megagrams falls in the middle
of the above projected range; and the projection takes into account
the more realistic estimate of steel production from the conventional
                                  9-38

-------
                  TABLE 9-17.   PROJECTIONS OF COKE CAPACITY'
                             1985, 1990, and 1995
Capacity requirements
(scenario)
Scenario 1C

Scenario 2

Projections (106 Mg/yr)
1985
49.6
46.1
49.6
46.1
1990
48.6
45.1
50.0
46.6
1995
47.6
44.2
50.3
47.0
 Coke capacity includes furnace and foundry coke.   Figures in brackets
 represent furnace coke capacity.
 Coke capacity requirements are computed from the estimate of coke production
 in Table 9-16, assuming 85 percent capacity utilization.

cScenario 1 assumes imports to grow at the long-term trend.

 Scenario 2 assumes the imports are held constant at the 1985 level of 3,547
 thousand megagrams per year from 1985 through 1995.
                                     9-39

-------
open hearth and basic oxygen furnaces (70 percent of steel production
from open hearth and basic oxygen furnaces combined as shown in
Table 9-15).  Hogan and Koelble and Lawrence R. Smith (Koppers)
forecast a furnace coke consumption of 51.7 to 53.5 million megagrams
for 1985.48  However, this projection of furnace coke consumption does
not appear to be consistent with the projected share of steel pro-
duction using basic oxygen and open hearth furnaces--70 percent.
Historically, foundry coke production, consumption, and exports have
been quite stable.  The projected annual production of approximately
3 million megagrams of foundry coke during the 1985-1995 period
presented in Table 9-16 assumes that the past trend will continue in
the future.
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 potential for market power, other characteristics
belie it.
     Some concentration exists in coke-making capacity and steel
production;  however, many firms produce coke and iron and steel
products.  Substantial vertical integration appears to result pri-
marily from a desire for increased certainty in the supply of critical
inputs.   Furthermore, substitution through alternative technologies
and coke imports is feasible, and some substitutes for the industry's
final products (iron and steel) are available.   In any industry, the
potential for substitution is a major factor leading to competitive
pricing.   Certainly, the financial profile of coke-making firms is not
indicative of monolopy profits.   Prospects for industry growth are
limited.   An individual  firm must actively compete with other firms in
the industry to improve its profit position.
     No industry matches the textbook definition of perfect compe-
tition.   The important issue is whether or not the competitive model
satisfactorily captures major behavioral  responses of firms in the
industry.
                                  9-40

-------
9.2  ECONOMIC IMPACT OF REGULATORY ALTERNATIVES
9.2.1  Summary
     Economic impacts have been projected for the baseline and for
each regulatory alternative.   Furnace and foundry coke impacts are
examined separately because their production costs and markets differ.
All cost and price impacts are in third-quarter 1979 dollars.  When
measured on a per-unit basis, the costs of meeting baseline regu-
lations for foundry coke plants tend to be greater than those for
furnace coke plants.   There are two reasons for the difference in
per-unit costs.   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.   The cost of compliance with Regulatory
Alternative II (combination of topside, charging, and door leaks)
would be about $19.2 million annually for furnace and foundry coke
producers combined.
     Regulatory Alternative II would generate capital costs of $44.9
million.  Full compliance with baseline regulations measured against
the 1979 state of control would result in annualized costs of $404
million and capital  costs of $1,122 million.
     Price impacts are estimated under the empirically supported
assumption that furnace coke demand is responsive to higher coke
prices.  Foundry coke demand is also assumed to respond to price.
Under these assumptions, the price increases for Regulatory Alter-
native II would be $0.18/Mg (0.14 percent) for furnace coke and $2.9/Mg
(1.6 percent) for foundry coke.   Compliance with baseline regulations
measured against the 1979 state of control increases the furnace coke
price by 5.6 percent and the foundry coke price by 12.3 percent.
     Under Regulatory Alternative II, furnace coke production would
decrease 0.16 percent from the baseline level.   Under Regulatory
Alternative II,  foundry coke production would decrease 1.6 percent
                                  9-41

-------
from the baseline level.  Complete compliance with baseline regu-
lations measured against 1979 compliance would decrease furnace
production by 5.7 percent and foundry production by 10.4 percent.
Complete compliance with baseline regulations produces three potential
furnace battery closures and four potential foundry battery closures.
Regulatory Alternative II is not projected to result in any furnace
foundry battery closures.  As a consequence, Regulatory Alternative II
is not projected to result in any plant or company closures.
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 baseline and the regulatory alternatives will
create.  These long-run adjustments involve investment and shutdown
decisions.   Short-run operating decisions are not considered.   For
example, short-run adjustments to fluctuations in the demand for coke
are frequently made by altering coking times.   Short-run adjustments
are not considered in 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 intra-
firm shipments of coke are not uncommon, as can be inferred from
Table 9-10.   A plant-by-plant review of the coke industry by Hogan and
Koelble also confirms the existence of such exchanges.49
     A set of programmed models has been developed to produce intra-
industry and interindustry estimates of the economic impacts of the
alternative regulations.  The models are applied to both furnace and
foundry coke.   The sectors included are coke,  steel, and ferrous
foundries.   The rest of the economy is incorporated in the inter-
industry portion of the analysis.
     The analytical approach incorporates a production cost model of
the coke industry based on engineering data, an econometric model of
                                  9-42

-------
the steel industry, and an input-output model of the rest of the
economy and final demand.   The interrelationships of these models for
furnace coke are shown in Figure 9-4.   The upper portion of Figure 9-4
encompasses the supply side impacts of the regulatory alternatives,
and the lower portion contains 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 further described 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 func-
tions, with and without additional controls.50  Supply functions are
estimated on a year-by-year basis for furnace and foundry coke plants
projected to be in existence between 1980 and 1990.  Both coke produc-
tion costs and the costs that plants already 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 further compliance with the baseline regulations
and the regulatory alternatives are used to compute the projected
upward shifts in that supply function.   All costs are in 1979 dollars.
     This approach has been used because it provides a method of
estimating the industry supply curve for coke.   The supply curve shows
the alternative coke quantities that will be placed on the market at
alternative prices.  When the supply curve is considered in con-
junction 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 control  cost estimates
developed by the engineering contractor.  These new supply functions,
along with the demand curve, can then be used to compute the equili-
brium price and output rate under each regulatory alternative.
     9.2.2.1.1  Data base.  Plant-by-plant data on over 60 variables
for furnace and foundry coke plants in existence in 1979 were compiled
from government publications, industry contacts,  and previous studies
                                  9-43

-------
 COSTS OF COKE
  PRODUCTION
FOREIGN COKE
   SUPPLY
COSTS OF EXISTING
  ENVIRONMENTAL—\
   REGULATIONS
  COSTS OF
REGULATORY
ALTERNATIVES
                            EXISTING PLANT
                               INVENTORY
                              NEW PLANT   .
                            CONFIGURATIONS
DOMESTIC
  COKE
 SUPPLY
                            FOREIGN DEMAND
                              FOR US COKE
FOREIGN DEMAND
 FOR US STEEL '
DOMESTIC DEMAND
  FOR US STEEL -
                              DEMAND FOR
                               US STEEL
                    SYNTHESIS
                  " BEHAVIORAL"
                   ASSUMPTIONS
 IMPACTS
"ON  COKE
 AND STEEL
 ^IMPACTS
   ON FINAL
DEMAND PRICES
  DEMAND FOR
    US COKE
                                        Figure 9-4.  Economic impact model.

-------
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,51 and to the American Coke and Coal  Chemi-
cals Institute.
     9.2.2.1.2  Output relationships.   For a given battery, the  full
capacity 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 decarbonization time).  The following values for effective
gross coking time were used except where plant-specific values were
available.50 51
Furnace
coke
18 hours
13 hours
Foundry
coke
30 hours
24 hours
          Wet coal
          Preheated coal
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 from the coking process  are esti-
mated from engineering relationships.   These quantities are dependent
on the amount of coal carbonized, the percentage of coal  volatile
matter, coking time, and the 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 environmental control cost centers (Figure 9-5) to  facilitate
the development 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
                                  9-45

-------
en


_,..,.. .-..._-.-._ (0>t Cl»ll» 1
(0>l IMIS1

I.I. CM (CMtiroi Kilt




m
tio«»cf —
1
rot IU«IH(»
'•OCMIIAI;
M 101 full
kit! JU
IIDI
mill*


&*»H — ' • 1 I
	 • — '-* (OM 1I01HI I 1
. yr~ J .„ _
/\J humc
ion meet — ^ "~|




ruiiiifHG HO""* * ......
1 ^=>>, 	 L.I 	
- 	 fiuinns .\J( i ^
'" D £?!• V ! (O.lon.l ^l »!««•« __ CO.
1

• ,!? i 1 <0" "«"« »

i :
»« v«,v — i',r,v L «!•..«««
ettuid | 'win *-• -
j
| ACIO
1
IllUUItll --^ MtCIFIIAlO* -*- I(H(AI(I . ^ '«Jo°»V*l •-
1
AMKSOIH
IU.»lUIC
«?"* -^ rHll>0i ,^^«*^ AHWWl-H,!
MA-U iiioii<" v»'ui;
IOOIIM NKHCXAII COM ((11(1 1

1. •
UMt. ...... ' UAH(UAIH 1
... .. ™?Ml¥*l'* ^.| IHMM1I J
j I91L11S11U j
JL
J_
1 '1 >U)N|liC I«I11I011 1
| OL^OM.OU | | ' (0-.I.W |
| lULilill.U | | u\UL«ii«_L 1
^^^L- 	 1




1 1 tOU HMI*I*6""~" 	 *"
t ~"i — *^ 
-------
identifying the mine that supplies each plant and estimating transporta-
tion 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 costs were
estimated based on the distances traveled, the transport mode (barge
or rail) employed, and appropriate rates for the chosen mode.
     Labor and supervision requirements for maintenance were estimated
for 69 jobs within the coke plant.  Primary variables that determine
the number of maintenance labor and supervision man-hours needed
include type of plant (merchant or captive), number of battery units,
number of plants at a site, size of by-product plant, type of coal
charge (wet or preheated), and coke production.   The labor rates used
for captive plants were $17.04 per hour for supervisory positions and
$15.70 per hour for production labor.   For merchant plants, rates of
$15.80 per hour and $14.40 per hour were assumed.
     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 $5.44 per 103 Ib
of steam, $0.027/kWh electricity, $0.16 per 103 gal.  of cooling water,
and $2.76 per 106 Btu underfire gas.
     9.2.2.1.4  Capital  costs.  Although no net additions to industry
coke-making capacity are anticipated during the 1980 to 1995 period,
some producers will rebuild or replace  existing batteries.   Such
actions will  alter the long-run industry supply curve because the new
batteries will 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.  Hogan and Koelble47
have identified plant-by-plant rebuild/replacement intentions.   These
plans are included in the data base.   The cost of building a new coke
battery with model characteristics and  the cost of major rehabili-
tation of an existing battery have been estimated for the affected
                                  9-47

-------
facilities.  It has been assumed that new furnace construction will be
6-meter batteries and new foundry construction will be 4-meter
batteries.  Pad-up rebuilds are assumed to leave the battery size
unchanged.  Pad-up rebuild costs were estimated as a function of
battery size.   A zero salvage value is assumed for existing batteries.
     The capital cost breakdown for new plants is shown in Table 9-18.
For such plants, the major capital cost items are the battery, quench
tower, quench car, pusher machine, larry car, door machine and coke
guide, by-product plant, coal handling system, and coke handling
system.  Every new (model) battery is assumed to have 60 ovens.
Pipeline charging can increase the coke-making capacity of a given
oven by about 25 percent by reducing gross coking time.   Consequently,
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 the performance deterioration that occurs over
time.   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.34  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 regulations and the
                                  9-48

-------
          TABLE 9-18.   ESTIMATED CAPITAL COSTS OF NEW BATTERIES52
                                   Conventionally
                                  charged battery
Capital costs by element
  (106 1979 dollars)
   Pipeline
charged battery

Capacity (103 Mg/yr)
4-metera 6-metera
450 720
4-metera 6-metera
560 900
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.
                                  9-49

-------
regulatory alternatives under consideration in this study were
incorporated in the model.  The current regulations include workplace
standards (OSHA), water quality regulations (BPT and-BAT), and SIP
requirements.  Estimates of compliance costs already incurred for each
plant in the data base for current regulations were included in the
existing control costs.  Therefore, it was possible to estimate the
remaining environmental costs that plants will incur to meet current
regulations.   The additional costs were included in baseline control
costs.   It has been assumed that costs to comply with OSHA and BPT
water requirements under the Federal Water Pollution Control Act were
incurred by 1981.  Costs for all other existing environmental regula-
tions are assumed to be incurred by 1983.
     The scatter diagrams in Figures 9-6 and 9-7 show estimates from
the coke supply model  of average total cost of production in 1980,
including environmental costs,  for all furnace and foundry coke plants.
A number of factors such as the delivered price of coal,  the age
of the plant, and the by-products recovered create variability in the
average cost of production across coke plants.
     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 percentage, which was estimated from data on the
capital structure for publicly  owned steel companies,  has been used in
this analysis as the minimum acceptable rate of return on new facil-
ities. 53
     The capital costs of controls that are affixed to coke  oven
batteries are annualized with the assumption that when a  battery
reaches the end of its useful life, it is  rebuilt or replaced by a
                                  9-50

-------
    0 J
    w-

2  9^
•y CO"
O ~
h-
V.
g
M a
kw* f"\J — •
^^— V>f "^
o —
,

4
4* 4
4
i
4
i ' \
0
9F
a.

*
^ 4
.0 o
™_ —
* 4
i
CO 4 *
o i

            A  4
    O  .
    O —
    8-
                1
                      1    1
    o_
    03
I    I
                                I    I
                300        600        900       1,200     1,500  .    1,300

                      PRODUCTION  CAPACITY CTONNES*100C/YS)


             Figure 9-6.  Estimated average cost of furnace coke production as a function

                                 of plant production, 1980.
                                                 2. I /"!.<"»
                                                 , 1 t»U
                                          9-51

-------
    a
    co-
    o
    uv
UJ



o
O
    C

n   CO1
CO

O

O
    o
    o-
                          t

                         I CO                200                3CO


                       PRODUCTION  CAPACITY (TONNES*VOCO/YR}



             Figure 9-7. Estimated average cost of foundry coke production as a function of

                                    plant production, 1980.
4CC
                                           9-52

-------
battery of the same height.  If this situation occurs, most of the
control equipment is salvageable.54  Accordingly, each annualization
is performed over the life of the control equipment.
     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 production and the output for each battery is accumulated to
produce a stepped marginal cost function for the plant; plant overhead
costs are averaged for all relevant plant output rates; and average
total costs are computed for each output rate by summing the average
costs for plant overhead and the battery.  Each plant's supply
function is the portion of the marginal cost function above the
average total cost function.   For existing plants where the average
total cost exceeds marginal cost over the entire range of output, the
supply function is the point on the plant's average total  cost
function represented by capacity output (after allowing for outages).
The aggregate long-run supply function for all currently existing coke
plants and batteries is obtained by horizontally summing the supply
function for each plant.   The 1980 industry marginal cost (supply)
curves for existing furnace and foundry coke plants are presented in
Figures 9-8 and 9-9, 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 des-
cribed previously.   These costs include the normal  return  on invest-
ment 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 foundries products.   A demand function for furnace
coke was derived by.econometrically modeling the impacts of changes in
furnace coke production costs on the steel  industry.55
                                  9-53

-------
    o
    
-------
    a
    r—r
    a
    o
-  5-1
LJ  I. !
2

O
t^5  O
8-  ^1
                     AVERAGE COST
                     MARGINAL COST
    o I
    CM-f
  7
   i

0_[_
•M  I
  0
               500
,CCO      I.5CC      2,COO      2,500

 PRCDUCTICN  CTONNES*!OGO/YS3
3.500
                 Figure 9-9.  Marginal and average cost functions for'foundry coke, 1980.
                                         9-55

-------
     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 were used to estimate
the different components of the steel sector.  Visual inspections of
the correlation matrix and a plot of the dependent variable versus the
residuals indicated no multicollinearity or heteroscedasticity
problems.  The Durbin-Watson statistic showed no evidence of auto-
correlation.
     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 analysis.56 57  Estimates of elasticities
for coke and steel functions are presented in Table 9-19.
     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 precise data necessary to estimate
properly 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
                                  9-56

-------
     TABLE 9-19.  ESTIMATES OF ELASTICITIES OF STEEL AND COKE MARKETS

                                                 Point         Interval
                                                estimate       estimate3

1.   Percent change in furnace coke demand       -1.17 'c    (-1.06, -1.29)
     for 1 percent change in the price of
     furnace coke

2.   Percent change in foundry coke demand       -1.03
     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         0.14°      (0.139, 0.141)
     1 percent change in the price of
     furnace coke

5.   Percent change steel demand for 1           -1.86C      (-0.54, -3.18)
     percent change in the price of steel

6.   Percent change in steel imports for          1.51C      (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.

alnterval estimates are based on 95 percent confidence level.

 Derived from the production function  for steel.

 Significantly different from zero at  1 percent level  of statistical
 significance.

 Calculation based on the theoretical  relationship between input demand
 elasticity and input cost share in the production of  foundry products.
 Accordingly, no interval is provided.
                                  9-57

-------
inputs to foundry production.  This elasticity calculation is based on
a 3-year average of the cost share of foundry coke in foundry produc-
tion.  This estimate is presented in Table 9-19.
     9.2.2.3  Synthesis.  Separate linear functions were fit to the
furnace and foundry coke marginal cost values depicted in Figures 9-8
and 9-9.  As illustrated in Figure 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 (PL and Qx in Figure 9-10),
The supply function is reestimated for each regulatory alternative (S1
in Figure 9-10), and the new equilibrium price-quantity values are
predicted.
     9.2.2.4  Economic Impact Variables.  Table 9-20 shows the
specific economic variables for which impacts are estimated.   The
methodology presented 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.   In Figure 9-11, DO'  represents the derived
demand for coke.  The line cd represents the supply of coke.   The
equilibrium price and quantity are P! and QL respectively.   The area
OcdQi is the total cost of coke production, OP^O,! is the total
revenue, and cPxd represents before-tax profits.   The total  cost of
coke production (OcdQi) can be divided into costs incurred to produce
coke per se (OabQi) and the costs being incurred to meet existing
environmental  and other regulations (acdb).
     The regulatory alternatives will increase the cost of coke
production by shifting the supply function to ef.   Given the demand
and supply functions as drawn in Figure 9-11, higher costs of pro-
duction will lead to higher prices.   If there were no substitutes for
coke and no decrease in the production of coke-using products, the
rate of coke consumption would remain at Qj annually and the price of
coke would increase to P2.   The cost of the regulatory alternative
would be cefd.   However, a production decrease is more likely.   As
                                  9-58

-------
  S/Q
P2
                                             • Q/time
                        b)
        Figure 9-10.  Coke supply and demand.
                        9-59

-------
        TABLE 9-20.   ECONOMIC IMPACT VARIABLES AND AFFECTED SECTORS
Variable
                                                   Sector
Furnace
 coke
Foundry
 coke
Steel
 Final
demand
Price

Output

Profits

Costs

Plant closures/openings

Capital requirements

Factor employment
     Labor
     Metallurgical coal

Imports
   X

   X

   X

   X

   X

   X
   X
   X
   X

   X

   X

   X
   X
   X
  X

  X
                                  9-60

-------
Figure 9-11.  Coke demand and supply with and without
                regulatory alternatives.
                         9-61

-------
shown in Figure 9-11, the price would rise to P3 and the quantity
demanded would fall to Q2.  The actual costs of the regulatory
alternative are ceik, and profits before income taxes are eP3i.
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 inter-
firm 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.
     Trend projections through 1995 in coke and steel production and
consumption were presented in Section 9.1.   The baseline values for
1983, presented in Table 9-21, are based on these projections.   The
projected values for 1983 were assumed to reflect full  compliance with
applicable SIP, OSHA, and water.   The coke supply model  was used to
compute the price of furnace coke, costs, revenues, and  profits, given
these trend projections.   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
1979 for all emission sources.  This estimation was used to determine
the impacts of moving from actual industry control  levels to baseline
control for all sources.   Actual  industry control practices were most
recently accessed in 1979.  These impacts are also presented in
Table 9-21.
     Table 9-22 presents the capital and operating costs that have yet
to be incurred, but that must be incurred in meeting the baseline for
all sources.  Table 9-22 also provides an estimate of the costs that
have already been incurred (based on 1979 control levels) to meet
existing and proposed regulations (the baseline) for all sources.
                                  9-62

-------
U>
                    TABLE 9-21.   1983 BASELINE VALUES FOR ECONOMIC IMPACT ANALYSIS—FURNACE COKE

                                                                                  Impact of moving from   .
                                                  Baseline values3            existing to baseline control
Coke market
Price (1979 $/Mg)
Production (103 Mg/yr)
Consumption (103 Mg/yr)
Imports (103 Mg/yr)
Employment (jobs)
Coal consumption (103 Mg/yr)

129
39,500
42,600
3,100
7,100
47,800

6.8
-2,400
-2,100
300
400
2,900
      Steel market
        Price ($/Mg)                                      409                               3.1
        Production (103 Mg/yr)                        117,000                          -2,200
        Consumption (103 Mg/yr)                       137,400                          -1,900
OT       Imports (103 Mg/yr)                            22,600                            +300
        Employment (103 Mg/yr)                        401,100                           6,700

      aBaseline includes:   OSHA (coke oven emissions); desulfurization,  pushing,  coal  handling, coke handling,
       quench tower, and battery stack controls;  and BPT and BAT water regulations.

       The effects of moving from existing (December 31, 1979)  levels of control  to  the regulatory baseline
       are included in the baseline values.

-------
      TABLE 9-22.   BASELINE CONTROL COSTS—FURNACE COKE (106 1979 $)a

                                             Annualized      Incremental
                                               costs        capital costs

Costs already .incurred to meet 1979             499             1,454
  regulations

Costs yet to be incurred to meet 1979           364               947
  regulations

Total cost to meet baseline                     863             2,401

 Calculated for all batteries projected to be in existence in 1983.42 44

 Not including capital charges (sunk costs).

 The capital costs of controls already in place are  estimated as if the
 controls were put in place in the third quarter of  1979.

 Includes:  OSHA;  desulfurization, pushing, coal handling, coke handling,
 quench tower, and battery stack emission controls;  and BPT and BAT water
 regulations.
                                  9-64

-------
     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 just 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 estimated econometrically and
found to be highly responsive to price changes.  The estimated
elasticity of demand for furnace coke is -1.2.  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 (14.3 percent) times the base
price of steel, assuming that there is only one coke price and no
inventories.  This assumption is used to develop maximum steel price
effects.
     Table 9-23 presents the furnace coke and steel price impacts of
the regulatory alternatives.   This table also shows the impacts on
prices caused by the change from 1979 control levels to baseline
levels of control for all sources.  Complete compliance with the
                                  9-65

-------
          TABLE 9-23.   PRICE EFFECTS OF REGULATORY ALTERNATIVES--
                            FURNACE COKE,  1983
                              (1979 dollars)
                                             Coke
                                            ($/Mg)
                 Steel
                 ($/Mg)
Price assuming 1979 controls

Price increase caused by moving
  from 1979 controls to baseline
  controls

Price increase caused by moving
  from baseline to Regulatory
  Alternative II
    Topside
    Charging
    Doors
    Doors
    Topside and doors
    Topside, charging,  and doors
122.2

  6.8
  0.06
  0.09
  0.05
405.9

  3.1
  0.03
  0.04
  0.02
                                             Cost-effective combinations
                                             of regulatory alternatives
  0.05
  0.09
  0.18
  0.02
  0.04
  0.08
                                  9-66

-------
baseline measured against the 1979 control increases the coke price by
6.4 percent.   The regulatory alternatives are not likely to affect
coke prices.
     9.2.3.2  Production and Consumption Effects.   The estimated
demand and supply relationships for coke are used to project the
production and consumption effects of the regulatory alternatives.  As
shown in Table 9-24, the changes in coke production and consumption
are fairly small for Regulatory Alternative II.   However, moving from
1979 control  levels to baseline levels of control  reduces coke produc-
tion by 6.6 percent under Scenario A.
     Imported coke is a close substitute for domestically produced
coke.   Imported coke is not a perfect substitute because coke quality
deteriorates during transit and contractual arrangements between
buyers and sellers are not costless.   However, increases in the costs
of production for domestic plants will increase the incentive to
import coke.
     The projected increases in coke imports are reported in Table
9-24.   As illustrated below, coke imports have increased significantly
since 1972.
                         Year       Imports (103 Mg)
1972
1973
1974
1975
1976
1977
1978
1979
168
978
3,211
1,650
1,189
1,659
5,191
3,605
     The recent increase in imports is believed to be the result of a
coal strike in the United States during 1978 combined with depressed
conditions in the market for steel  in the countries exporting coke to
the United States.  Accordingly, continued 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 regu-
latory alternatives are small.
                                  9-67

-------
TABLE 9-24.   PRODUCTION AND CONSUMPTION EFFECTS OF REGULATORY ALTERNATIVES--
                             FURNACE COKE,  1983



Coke market
Steel market
Production Consumption Imports
(103 Mg/yr) (103 Mg/yr) (103 Mg/yr)
Value assuming
Change in value
from 1979 to
control s
Change in value
from basel ine
Alternative I
^ Topside
oo Charging
Doors
Doors
1979 controls
caused by moving
basel ine

caused by moving
to Regulatory
I


Topside and doors
Topside, charging, and doors
41,900
2,400


-20
-31
-17

-17
-31
-65
44,700
2,100


-18
-27
-15
Cost-effective
-15
-27
-57
2,800
-300


2
4
2
combinations
2
4
8
Production Consumption Imports
(103 Mg/yr) (103 Mg/yr) (103 Mg/yr)
119,200
2,200


-18
-28
-15
of regulatory
-15
-28
-59
139,300
1,900


-16
-24
-13
alternatives
-13
-25
-52
22,300
-300


2
4
2

2
3
7

-------
     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 produc-
tion 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 9-25.
These values were developed assuming constant coal- and labor-output
ratios.  The employment impacts shown do not include the estimated
increases in employment caused by the regulatory alternatives.   There-
fore, the employment impacts represent maximum values.
     9.2.3.4  Financial Effects.   The aggregate capital costs of the
regulatory alternatives are summarized in Table 9-26.  The capital
requirement to meet Regulatory Alternative II for the furnace coke
industry is $42.4 million.   The estimated value of furnace coke produc-
tion (baseline production multiplied by baseline price) is $5.1 billion.
The capital requirements to meet Regulatory Alternatives II or III
represent less than 1 percent of the expected gross furnace coke
sales.   It is reasonable to conclude that the aggregate industry
capital requirements are likely to be small in comparison to the total
cash flow represented by the value of the output.
     9.2.3.5  Battery and Plant Closures.   Battery closure candidates
are batteries that have marginal  costs of operation greater than the
projected price of coke.   Some batteries are closure candidates under
existing regulatory controls while others become closure candidates
when total compliance with baseline regulations is posited.  No other
batteries become closure candidates when the regulatory alternatives
are added.
     Under the 1979 level of control,  nine furnace batteries are
closure candidates.  When compliance with all baseline  regulations is
posited, three additional batteries become closure candidates.   Regula-
tory Alternative II does not increase the closure list.  Hence, the
total impacts under the most stringent combination of regulatory
alternatives include 12 batteries as closure candidates.
                                  9-69

-------
          TABLE 9-25.   COAL CONSUMPTION AND EMPLOYMENT EFFECTS OF
               REGULATORY ALTERNATIVES—FURNACE COKE, 1983
Coal
consumption
for coke
(103 Mg/yr)
Employment (jobs)3
Coal Coke Steel-
mining plant making
Change caused by moving from
  1979 controls to baseline
  controls

Change caused by moving from
  baseline to Regulatory Alter-
  native II

    Topside
    Charging
    Doors
-2,950
-940
-420    -6,750
   •24
   -37
   -20
  -8
 •12
  -6
  -3
  -5
  -3
-55
-85
-46
                                            Cost-effective combinations
                                            of regulatory alternatives
Doors
Topside and doors
Topside, charging, and doors
-20
-37
-79
-6
-12
-25
-3
-5
-11
-46
-86
-181
 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 3,125 megagrams
 annually per job.
                                  9-70

-------
         TABLE 9-26.   INDUSTRY CAPITAL REQUIREMENTS OF REGULATORY
                     ALTERNATIVES--FURNACE COKE, 1983

                                                Capital costs
                                                of regulations3
                                                 (106 1979 $)
Capital costs caused by moving from .
  1979 controls to baseline controls

Capital costs caused by moving from
  baseline to Regulatory Alternative II

    Topside
    Charging
    Doors
    Doors
    Topside and doors
    Topside, charging, and doors
          947
            4.6
            5.3
           32.5

Cost-effective combinations
of regulatory alternatives
           32.5
           37.8
           42.4
 Calculated for all  plants projected to be in existence in 1983.42 44

5The control  sources included in the baseline are described on Table 9-21.
                                  9-71

-------
     The 12 batteries identified as closure candidates under baseline
control are owned by six companies and are located in seven plants.
Three entire plants are projected to close under baseline conditions;
no entire companies are projected to close.
     The number of closure candidates under any regulatory scenario is
extremely sensitive to the projections of coke demand and coke imports
for 1983.   A 10-percent increase in projected 1983 coke production
would reduce the list of closure candidates under the baseline to
seven batteries in five plants, owned by four companies.
     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 inves-
tigated include other plants within the same company, other companies,
and foreign suppliers.   As noted in Section 9.1, some interregional
and international movement of coke occurs.
     Obtaining coke from offsite sources introduces two potential
complications:   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 profitable.   Three of the battery closure
candidates under the baseline are some distance from most other
coke-producing facilities.
     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 12 closure candidates under the
baseline produce at marginal  costs that are less than 3 percent above
the projected market price.   Another two of the 12 closure candidates
                                  9-72

-------
under baseline compliance produce coke at marginal costs that are less
than 5 percent above the market price.  If 5 percent is not an
excessive premium to pay for certainty of supply, these batteries
would not close.
     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,  and required control and other expenditures.
     The developed demand model uses a single coke price, which repre-
sents 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.
     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.
     Under the baseline, over half of the battery closure candidates
have incurred over 80 percent of projected baseline control expendi-
tures.   Three of the battery closure candidates have incurred less
than 30 percent of projected baseline expenditures.   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.
     As can be concluded from the preceeding discussion, 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.
9.2.4  Foundry Coke Impacts
     Oven coke other than furnace coke represents less than 10 percent
of U.S. coke production.  The majority of it is used as a fuel in the
                                  9-73

-------
cupolas of foundries.  The remainder is used for a variety of pur-
poses, especially for heating.
     Projections of various foundry coke variables in the absence of
the regulatory alternatives are presented in Table 9-27.  These
projections are based on historical trends in foundry coke production
and consumption extrapolated to 1983 and on the coke supply model.
Table 9-27 also provides estimates of the impacts of meeting the
baseline for all regulations measured against the 1979 level of
control.  Table 9-28 presents the cost already incurred and the costs
yet to.be incurred to meet current regulations.
     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.   The projected price and
quantity effects of these shifts are presented in Table 9-29.
     Complete compliance with the baseline,  measured against the 1979
level of control, increases the foundry coke price by 15.4 percent.
This price increase causes a 12.2-percent reduction in foundry pro-
duction.  Regulatory Alternative II does not cause any significant
change in coke price or coke output.
     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 9-30.   These  values
were developed assuming constant coal- and labor-output ratios.   The
employment impacts shown do not include any  employment increases
caused by the regulatory alternatives.   Consequently, the employment
impacts represent maximum values.
                                  9-74

-------
             TABLE 9-27.   1983 BASELINE  VALUES FOR ECONOMIC  IMPACT ANALYSIS—FOUNDRY COKE

                                                                         Impact of moving from 1979 .
                                          Baseline values3               control  to baseline control


Coke market
  Price (1979 $/Mg)                              187                                   21
  Production (103 Mg/yr)                        3,000                                 -350
  Consumption (103 Mg/yr)                      2,600                                 -350
  Employment (jobs)                              500                                  -60
  Coal  Consumption (103 Mg/yr)                 3,500                                 -500

aBaseline includes:  OSHA (coke oven emissions);  desulfurization,  pushing,  coal handling, coke hand-
 ling,  quench tower,  and  battery stack controls;  and BPT and BAT water regulations.
 The effects of moving from existing (December 31, 1979) levels  of control  to  the regulatory baseline
 are included in the  baseline values.

-------
      TABLE 9-28.   BASELINE CONTROL COSTS—FOUNDRY COKE3 (106 1979 $)
Total
annual ized
costs
Costs already .incurred to meet 1979 36
regulations
Costs yet to be incurred to meet 1979 44
regulations
Total cost to meet baseline 80
Incremental
capital costs
118
175
293
Calculated for all  batteries projected to be in  existence  in  1983.41  42
 Not including capital  charges (sunk costs).
 The capital  costs of controls already in place are  estimated  as  if the
 controls were put in place in the third quarter  of  1979.
 Includes:   OSHA and BTP and BAT water regulations.
                                  9-76

-------
    TABLE 9-29.   PRICE AND QUANTITY EFFECTS OF REGULATORY ALTERNATIVES
                            FOUNDRY COKE,  1983

                              Coke price impact        Coke quantity impact
                                 (1979 $/Mg)               (103 Mg/yr)

Value assuming 1979                 166                       3,350
  controls

Change in value caused               21                        -350
  by moving from 1979
  controls to baseline
  controls

Change in value caused
  by moving from baseline
  to Regulatory Alterna-
  tive II
    Topside                           1.2                        -20
    Charging                          1.1                        -17
    Doors                             1.4                        -23

                                       Cost-effective combinations of
                                         regulatory alternatives	
    Doors                             1.4                       -23
    Topside and doors                 1.6                       -26
    Topside, charging, and doors      2.9                       -48
                                  9-77

-------
    TABLE 9-30.   COAL CONSUMPTION AND EMPLOYMENT EFFECTS OF REGULATORY
                     ALTERNATIVES—FOUNDRY COKE, 1983
Coal
consumption
for coke
(103 Mg/yr)
Employment (jobs)
Coalb Coke
mining plant
Change caused by moving from           -480
  1979 controls to baseline
  controls

Change caused by moving from base-
  line to Regulatory Alternative II

    Topside                             -28
    Charging                            -23
    Doors                               -32
-155
-60
  -9
  -7
 -10
 -4
 -3
 -4
                                           Cost-effective combinations  of
                                              regulatory alternatives
    Doors                               -32
    Topside and doors                   -36
    Topside, charging, and doors        -67
 -10
 -12
 -21
 -4
 -4
 -8
 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  3,125 megagrams
 per job.
                                  9-78

-------
     9.2.4.3  Financial Effects.  The aggregate capital costs of the
regulatory alternatives are summarized in Table 9-31.  The capital
requirement to meet Regulatory Alternative II for the foundry coke
industry is $2.5 million.   The estimated value of foundry coke produc-
tion (baseline production multiplied by baseline price) is $561 million.
The capital requirement to meet Regulatory Alternative II represents
less than 0.5 percent of the expected gross foundry sales.  It is
reasonable to conclude that the aggregate industry capital requirements
are likely to be small in comparison to the total cash flow represented
by the value of the output.
     9.2.4.4  Battery and Plant Closures.   The decision rule used to
indicate 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 projected price of foundry coke is a
candidate for closure.  According to this criterion and assuming
baseline control, eight batteries that are projected to be in exist-
ence in 1983 are closure candidates.   These batteries are located in
five plants owned by five companies.   Three entire plants are closed
under baseline conditions.   All eight batteries are smaller than the
average foundry battery.
     Three of the eight batteries and one of the three plants are
closure candidates  under the 1979 level of control.   Complete com-
pliance with current regulations is responsible for five potential
battery closures, which result in two potential plant closures.
Compliance with Regulatory Alternative II  should not add to the list
of closure candidates.
     The number of  closure candidates under any regulatory scenario is
extremely sensitive to projected foundry coke production in 1983.   A
12-percent increase in projected production would reduce the list of
closure candidates  under the baseline to four batteries and one plant.
     If the battery closure candidates provide coke to foundries near
the coke plant and  if there are no other sources of coke in the
immediate vicinity, these batteries may continue to operate.   The cost
                                  9-79

-------
         TABLE 9-31.   INDUSTRY CAPITAL REQUIREMENTS OF REGULATORY
                     ALTERNATIVES—FOUNDRY COKE, 1983

                                             Capital costs of regulations
                                                    (106 $, 1979 $)

Capital costs caused by moving from .                    175.0
  1979 controls to baseline controls

Capital costs caused by moving from
  baseline to Regulatory Alternative II

    Topside                                               0.6
    Charging                                              1.1
    Doors                                                 0.8

                                             Cost-effective combinations
                                             of regulatory alternatives

    Doors                                                 0.8
    Topside and doors                                     1.4
    Topside, charging, and doors                          2.5

Calculated for all plants projected to be in existence in 1983.41 42

 The control sources included in the baseline are described on Table 9-27.
                                  9-80

-------
of transporting coke from distant sources may be sufficiently high to
outweigh potential cost and price differences among foundry coke
producers.
9.3  POTENTIAL SOCIOECONOMIC, INFLATIONARY, SMALL BUSINESS, AND
     ENERGY IMPACTS
9.3.1  Potential Socioeconomic and Inflationary Impacts
     Under Executive Order 12291, Federal agencies are required to
prepare a "regulatory impact analysis" for a "major" rule.   A major
rule is any regulation likely to result in:
     1.   An annual effect on the economy of $100 million or more;
     2.   A major increase in costs or prices for consumers; indi-
          vidual industries; Federal, State, or local  government
          agencies; or geographic regions; or
     3.   Significant adverse effects on competition,  employment,
          productivity, innovation, or on the ability  of U.S-based
          enterprises to compete with foreign-based enterprises in
          domestic or export markets.
     The results of the economic analysis are evaluated according to
the criteria below.
     9.3.1.1  Compliance Costs.   The estimated total annualized costs
to coke producers for compliance with the regulatory alternatives are
shown in Table 9-32.  Furnace and foundry coke costs are shown
separately and in total.  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 projected to be in
existence but not necessarily operating in 1983.
     In 1983, Regulatory Alternative II for topside, charging, and
door controls may result in incremental compliance costs (in 1979
dollars) of about $19.2 million per year.  It is expected that 1983
will be the year of maximum impact.
     The ratios of the average costs of compliance per unit of coke
production to the prices of furnace coke and foundry coke under the
regulatory baseline are shown in the right-hand portion of Table 9-32.
These ratios do not exceed 2 percent under any combination of regu-
latory alternatives.  However, meeting the baseline with regard to all
                                  9-81

-------
         TABLE 9-32.   COMPLIANCE COSTS OF REGULATORY ALTERNATIVES
                          UNDER SCENARIO A, 1983
                  Compliance cost9 (106 $/yr. 1979 $)
                  Furnace coke   Foundry coke   Total
                       Ratio of
                    average cost of
                    compliance to. the
                    price of coke  (%)
                    Furnace   Foundry
                     coke      coke
Costs caused by      364.0
  moving from
  1979 controls
  to baseline
  controls

Costs caused by
  moving from
  baseline to
  Regulatory
  Alternative II
    Topside            1.8
    Charging           4.6
    Doors              5.0
44.0
408.0
7.14
7.84
 0.2
 0.9
 6.7
  2.0
  5.6
 11.7
0.04
0.09
0.10
0.04
0.16
1.19
                    Cost-effective combinations of regulatory alternatives
    Doors              5.0
    Topside and        6.8
      doors
    Topside, charg-   11.4
      ing,  and
      doors
 6.7
 6.9

 7.8
 11.7
 13.7

 19.2
0.10
0.13

0.22
1.19
1.22

1.39
 Calculated for all  plants projected to be in existence in 1983.41  42
 Assuming baseline price and production levels.
                                  9-82

-------
regulations produces ratios of 7.1 percent for furnace coke and 7.8
percent for foundry coke.
     9.3.1.2  Prices and Consumer Costs.  The price changes projected
in Section 9.2 are reproduced in Table 9-33.   Coke price changes are
based on the assumption that quantity adjustments indicated by esti-
mated demand and supply functions will occur.  Some part of the
expected changes in the price of steel and ferrous foundry products
will be passed forward by producers that use steel and ferrous
products and some will be absorbed by these producers.  The degree to
which those changes are passed forward as opposed to being absorbed
will depend on the demand and supply conditions of the affected markets.
     Projections of the impacts on consumer prices of changes in the
price of steel and ferrous foundry products were made using input-
output analysis.  These projections are developed under the assumption
that purchasers of these products will pass forward the entire pro-
jected price increases to final consumers.   Hence, this projection
represents a worst-case outcome for consumer prices.   As shown in
Table 9-33, the effects on consumer prices are nominal for Regulatory
Alternative II.
     9.3.1.3  Balance of Trade.  Projecting recent trends in coke
imports implies continued increases in coke imports.   Imposition of
Regulatory Alternative II is expected to slightly reinforce this
trend.   Some increase in steel imports is possible also.   However,
since steel price increases caused by coke price increases are pro-
jected to be quite small, any increase in imports caused by Regulatory
Alternative II should be minor.  Moreover, trade regulations covering
steel imports may mitigate such increases.
     In the aggregate, it appears unlikely that Regulatory Alterna-
tive II would significantly affect the U.S. balance of trade position,
given the small share of international trade represented by coke
imports.  However, compliance with the baseline for all control
sources in the coke plant is estimated to increase coke imports by as
much as 10 percent.
                                  9-83

-------
     TABLE 9-33.   COKE,  STEEL,  FERROUS FOUNDRY,  AND CONSUMER PRODUCTS
              PRICE EFFECTS OF  REGULATORY ALTERNATIVES,  1983
                   Furnace
                     coke
                    ($/Mg ,
                    1979 $)
Foundry
 coke     Steel
($/Mg ,   ($/Mg ,
1979 $)   1979 $)
Foundry
products
 Increase
in consumer
price level
Increase caused      6.8       21.0      3.1        0.33        0.00070
  by moving from
  1979 controls
  to baseline
  controls

Increase caused by
  moving from base-
  line to Regulatory
  Alternative II

    Topside          0.06       1.2      0.03        0.01        0.00001
    Charging         0.09       1.1      0.04        0.02        0.00001
    Doors            0.05       1.4      0.02        0.02        0.00001

                    Cost-effective  combinations  of  regulatory  alternatives

    Doors            0.05       1.4      0.02        0.02        0.00001
    Topside and      0.09       1.6      0.04        0.03        0.00001
      doors
    Topside,         0.18       2.9      0.08        0.05        0.00003
      charging,
      and doors
                                  9-84

-------
     9.3.1.4  Community Impacts.  Furnace and foundry coke and steel
production facilities are located in Pennsylvania, Indiana, Ohio,
Maryland, New York, Colorado, California, Michigan, Illinois, Alabama,
Utah, Kentucky, Tennessee, Missouri, Wisconsin, and West Virginia.
Closure of coke facilities could have impacts on communities in these
States.  Regulatory Alternative II is not projected to result in any
plant closures.  Potential production decreases should not be suffi-
cient to generate significant community impacts.  However, compliance
with 1979 regulations could result in additional battery and plant
closures and the resulting community impacts.
     9.3.1.5  Conclusions.  Regulatory Alternative II would not be
considered a major action under Executive Order 12291 for a number of
reasons.  First, the annual  compliance costs of Regulatory Alter-
native II do not exceed the critical level, $100 million, indicated in
the order, and second, Regulatory Alternative II is not anticipated to
increase consumers' prices significantly.   Furthermore, the alter-
native is not expected to have significant adverse effects on com-
petition, employment, or investment.
9.3.2  Small Business Impacts
     The Regulatory Flexibility Act (RFA) requires consideration of
the potential impacts of proposed regulations on small "entities."
The Small Business Administration (SBA) defines small  firms in terms
of employment.   Firms owning coke ovens are included in SIC 3312, which
includes blast furnaces, steel works, and rolling mills.   The SBA has
determined that any firm that is in SIC 3312 and employs less than
1,000 workers will be considered small  with regard to the RFA.   The
EPA guideline for implementing the RFA specifies that a Regulatory
Flexibility Analysis must be performed if it is expected that the
regulation will have a significant adverse economic impact on a
substantial number of small  entities.  The EPA guidelines stipulate
that 20 percent or greater of the entities is a substantial number.
     Employment data on existing coke-producing firms  were examined
to determine the number of small firms  being affected by Regulatory
Alternative II.  Five out of a total of 33 existing coke-producing
                                  9-85

-------
firms were found to be small (employed less than 1,000 workers).
Employment information could not be found for 2 out of the 33 existing
coke-producing firms so they were treated as small firms.  Because the
standard being proposed is a NESHAP and all existing and new plants
will be legally required to comply, all plants owned by the small
firms not currently in compliance potentially could be impacted adversely.
     The RFA guidelines indicate that an economic impact should be
considered significantly adverse if it meets one of the following
criteria:
          Annual compliance costs (annualized capital, operating,
          reporting, etc.) of the regulation increase total costs of
          production for small  entities, for the pertinent process or
          product being regulated, by more than 5 percent.
          Compliance costs as a percentage of sales for small entities
          are at least 10 percentage points higher than compliance
          costs as a percentage of sales for large entities.
          Capital costs of compliance represent a "significant" portion
          of capital available  to small entities, considering internal
          cash flow plus external financial capabilities.
          The requirements of the regulation are likely to result in
          closures of small entities.
     The first estimation method recommended by the guidelines is to
obtain the percent increase in  the average total cost of producing
coke as a result of the proposed standard.   Present RFA guidelines
state that cost increases greater than 5 percent are considered to be
significant.   None of the small firms identified was found to have an
average cost increase that was  greater than the 5-percent value.
     The second method that the RFA draft guidelines recommended for
estimating economic impacts requires information on annual  sales for
small firms.   However, small, privately owned firms do not report
their annual  sales, and sales data are available from published sources.
     The third method that the  RFA draft guidelines recommended for
estimating economic impacts requires information on average annual
capital spending of firms.   The goal is to determine if the capital
requirements  of regulations will cause capital  availability problems
for small  firms.   Ironically, small, privately  owned firms  frequently
do not report their annual  investment.   Therefore,  it is usually
                                  9-86

-------
impossible to assess small firm capital availability.  No financial
data could be found for the small coke-producing firms previously
identified.
     The economic impact analysis in Section 9.2 included a closure
analysis that identified potential plant closure candidates that
resulted from meeting existing regulations, i.e., OSHA coke oven
emission standards, SIPs, water standards, and other air standards.
These plants are called baseline closure candidates.  The results of
the closure analysis of implementing Regulatory Alternative II revealed
that there will be no additional closure impacts.  As a consequence,
Regulatory Flexibility Analysis is unnecessary.
9.3.3  Energy Impacts
     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, 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  IMPACTS OF VARIED ECONOMIC CONDITIONS:   A SENSITIVITY ANALYSIS
     Steel production in the United States during 1982 fell  below
50 percent of the potential production level.   A number of integrated
steel plants were reported to  have stopped (or planned to stop) produc-
tion temporarily or permanently.58  Plants that continued production
frequently produced below their normal  production levels (85 percent
capacity utilization).   As a consequence of reduced steel production,
steel producers reduced coke production by shutting down coke bat-
teries and by increasing coking time.

                                  9-87

-------
     Historically, domestic steel and coke production activities have
been sensitive to the general level of economic activity.   If the
level of economic activity revives during the coming years, it is
likely the steel industry will respond by increased production activity.
The economic effects estimated in Sections 9.2 and 9.3.do not take
into account the temporary slack in the economic activity during 1982.
However, it is useful to analyze the changes, if any,  in the estimates
of economic effects of the NESHAP regulatory alternatives with varied
economic conditions.  One method coke producers use to reduce output
without shutting down ovens or batteries is to increase coking time.
However, increases in coking time do not cause proportional decreases
in output because increased coking time changes the mix of output
among coke, breeze, gas, and chemical products.   With increases in
coking time, the quality of coke is improved (lower amount of carbon)
and the ratio of coke to breeze is higher than before.   As a consequence
of increased coking time, the unit cost of production of coke will be
higher.  These effects are incorporated in the RTI coke model.
9.4.1  Effects of Change in Coking Time on Furnace Coke Price and
       Production
     The coke supply functions and the estimated economic effects
presented in Section 9.2 represent "average" economic conditions in a
normal year.  The coke supply functions for all  the batteries included
in the previous analysis (Section 9.2) were recomputed by increasing
the reported coking time of each battery by 10 and 20 percent.   The
price and quantity effects of increased coking time are estimated
under 1979 controls, baseline controls, Regulatory Alternative II, and
movement from baseline controls to Regulatory Alternative II.
     As shown in Tables 9-34 and 9-35, a 10-percent increase in coking
time did not cause the projected price and output effects of Regulatory
Alternative II.   However, a 20-percent increase in coking time caused
an increase of 12
-------
             TABLE 9-34.   EFFECTS OF INCREASED COKING TIME ON
                            FURNACE COKE PRICE
Furnace coke price- impact (1979 $/Mg)

Value at 1979 controls
Value at baseline controls
Value at Regulatory
Alternative II
Change in value at Regula-
tory Alternative II
from baseline
Normal
coking time
122.20
129.00
129.18
0.18
10-Percent
increase in
coking time
124.40
131.15
131.33
0.18
20-Percent
increase in
coking time
126.30
133.32
133.62
0.30




aNormal coking time represents the coking time used in estimating the
 economic effects reported in Section 9.2.
                                  9-89

-------
            TABLE 9-35.  EFFECTS OF INCREASED COKING TIME ON
                         FURNACE COKE PRODUCTION
Furnace coke production impact (103 Mg/year)
10-Percent
Normal increase in
coking time coking time
Quantity at 1979 controls 41,962 41,179
Quantity at baseline 39,537 38,760
controls
Quantity at Regulatory 39,462 38,695
Alternative II
Change in quantity at Regula- -65 -65
tory Alternative II
from baseline
20-Percent
increase in
coking time
40,496
37,985
37,880
-105
Normal coking time represents the coking time used in estimating the
economic effects reported in Section 9.2.
                                 9-90

-------
             TABLE 9-36.   EFFECTS OF INCREASED COKING TIME ON
                            FOUNDRY COKE PRICE
Foundry coke price impact (1979 $/Mg)


Value at 1979 controls
Value at baseline controls
Value at Regulatory
Alternative II
Change in value at Regula-
tory Alternative II
from baseline

Normal
coking time
165.8
186.9
189.9
3.0
10-Percent
increase in
coking time
174.8
196.4
201.5
5.1
20-Percent

increase in
coking time
183.6
207.6
212.9
5.3




aNormal  coking time represents the coking time used in estimating the
 economic effects reported in Section 9.2.
                                  9-91

-------
            TABLE 9-37.   EFFECTS OF INCREASED COKING TIME ON
                         FOUNDRY COKE PRODUCTION
Foundry coke production impact



Normal
coking time
Quantity at 1979 controls 3,346
Quantity at baseline
controls
Quantity at Regulatory
Alternative II
Change in quantity at
tory Alternative II
from basel ine
2,997
2,949
Regula- -48
10-Percent
increase in
coking time
3,198
2,840
2,756
-84
(103 Mg/year)
20-Percent

increase in
coking time
3,052
2,655
2,568
-87




Normal coking time represents the coking time used in estimating the
economic effects reported in Section 9.2.
                                 9-92

-------
9.4.2  Effects of Change in Coking Time on Foundry Coke Price and
       Production

     The coke supply functions for all foundry batteries were recomputed

by increasing the coking time in each battery by 10 and 20 percent.

The price and quantity effects are estimated under 1979 controls,

baseline controls, and Regulatory Alternative II and movement from

baseline controls to Regulatory Alternative II.   As shown in Tables 9-36

and 9-37, both the 10- and 20-percent increases in coking time increase

the projected price and production effects of Regulatory Alternative II.

Both impacts are almost doubled over the normal  coking time case.

9.5  REFERENCES

 1.  Executive Office of the President, Office of Management and
     Budget.  Standard Industrial Classification Manual, 1972.   U.S.
     Government Printing Office.   Washington, DC.   1972.  p.  145.

 2.  Office of Coal, Nuclear, Electric and Alternate Fuels, Energy
     Information Administration,  U.S. Department of Energy.  Energy
     and Data Reports:   Coke and Coal Chemicals in 1980.  Washington,
     D.C.  November 1981.   p. 4.

 3.  Bureau of the Census, U.S.  Department of Commerce.  1980 Annual
     Survey of Manufactures.   Statistics for Industry Groups and
     Industries (Including Supplemental Labor Costs).   Washington,
     D.C.  February 1982.

 4.  Office of Coal and Electric Power Statistics, Energy Information
     Administration, U.S.  Department of Energy.   Energy Data Reports:
     Coke and Coal Chemicals in 1979.  Washington, DC.   October 31,
     1980.   p.  33-37.

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

 6.  Bureau of Mines,  U.S.  Department of the Interior.   Minerals
     Yearbook,  1950-1978.   Volume I--Metals and Minerals.   U.S.  Govern-
     ment Printing Office.   Washington, DC.  (Table and page numbers
     vary because of reclassifications).

 7.  Reference 2, p. 4.

 8.  Bureau of the Census, U.S.  Department of Commerce.  Statistical*
     Abstracts of the United States:   1980.  101st Edition.   Washington,
     D.C. 1980.   p. 437.

 9.  Reference 5, p. 874.
                                  9-93

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

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

12.   Bureau of the Census, U.S.  Department of Commerce.   1972 Census
     of Manufactures.  U.S. Government Printing Office.   Washington,
     DC.   1976.

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

14.   Industrial Economics Research Institute.  Analysis of the U.S.
     Metallurgical Coke Industry.   Fordham University.  Bronx, New
     York.  October 1979.   p.  41.

15.   Reference 14, p. 40.

16.   Kerrigan, Thomas J.   Influences upon the Future  International
     Demand and Supply for Coke.  Ph.D.  dissertation.  Fordham Univer-
     sity.  Bronx, New York.   1977.  p.  14, 114.

17.   Reference 4, p.  1.

18.   Reference 4, p.  13.

19.   Reference 4, p.  3-5.

20.   Reference 4, p.  3-5.

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

22.   Reference 4, p.  20.

23.   Reference 14, p. i.

24.   Reference 14, p. ii.

25.   The Outlook for Metallurgical Coal  and Coke.  Institutional
     Report.  Merrill Lynch,  Pierce, Fenner, and Smith, Inc.  1980.
     p. 1.

26.   Reference 25, p. 5.
                                  9-94

-------
27.  PEDCo Environmental, Inc.  Technical Approach for a Coke Production
     Cost Model.  1979.  p. 39-50.

28.  Office of Energy Data and Interpretation, Energy Information
     Administration, U.S. Department of Energy.  Energy Data Reports:
     Coke Producers in the United States in 1978.  Washington, DC.
     September 4, 1979.

29.  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.

30.  Office of Energy Data and Interpretation, Energy Information
     Administration, U.S. Department of Energy.  Energy Data Reports:
     Distribution of Oven and Beehive Coke and Breeze.  Washington,
     DC.  April 10, 1979.  p.  7-8.

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

32.  Moody's Investors Service, Inc.   Moody1s  Industrial Manual,
     Volumes 1 and 2.   New York.  1979.

33.  Temple, Barker, and Sloane,  Inc.  An Economic Analysis of Proposed
     Effluent Limitations Guidelines, New Source Performance Standards,
     and Pretreatment Standards for the Iron and Steel Manufacturing
     Point Source Category.   Exhibit 6.   Lexington, MA.   December 1980.

34.  Reference 33, p.  VI-4.

35.  Reference 8, p. 873.

36.  Reference 2, p. 8-9.

37.  Reference 4, p. 8.

38.  Bureau of the Census, U.S. Department of Commerce.   1979 Annual
     Survey of Manufactures.  Value of Product Shipments.   Washington,
     D.C. October 1981.  p. 19.

39.  Reference 3, p. 16, 19.

40.  Bureau of the Census, U.S. Department of Commerce.   1978 Annual
     Survey of Manufacturers.   Statistics for Industry Groups and
     Industries (including supplemental  labor costs).   Washington,
     D.C.  January 1981.  p.  16.
                                  9-95

-------
41.  Bureau of Mines, U.S. Department of the  Interior.  Minerals
     Yearbook, 1980.  Volume I--Metals and Minerals.   U.S. Government
     Printing Office.  Washington, D.C.  1981.  p. 437.

42.  Reference 2, p. 3.

43.  Reference 2, p. 33-35.

44.  Reference 2, p. 28-32.

45.  Moody1s Investors Service, Inc.  Moody1s Industrial Manual,
     Volumes 1 and 2.  New York.  1981.

46.  American Iron and Steel Institute.  Annual Statistical  Report.
     Washington, D.C.  1981.  p. 21.

47.  Reference 25, p. 3.

48.  The Politics of Coke.  33 Metal Producing.  March 1980.  p. 49-51.

49.  Reference 14, p. 63-111.

50.  Reference 27, pp.  1-69.

51.  Letter from Young, E. F.,  American Iron and Steel Institute,  to
     Pratopos, J., U.S. Environmental Protection Agency.  May 16,  1980.

52.  Reference 27, p. 85.

53.  Economic Impact of NSPS Regulations on Coke Oven Battery Stacks.
     Research Triangle Institute.   Research Triangle Park, NC.  May 1980.
     pp. 8-45, 8-47.

54.  Letter from Kemner,  William F. , PEDCo Environmental, Inc., to
     Bingham, Tayler, Research Triangle Institute.  March 31, 1980.

55.  An Econometric Model  of the U.S. Steel Industry.  Research Triangle
     Institute.   Research Triangle Park, NC.  March 1981.

56.  Ramachandran, V.  The Economics of Farm Tractorization  in India.
     Ph.D.  dissertation.   North Carolina State University.   Raleigh,
     NC.  1979.

57.  Heckman, James J.   Shadow Price, Market Wages, and Labor Supply.
     Econometrica.   42.  pp. 679-694.

58.  "U.S.  Steelmakers Slim Down for Survival."  Business Week.
     May 31, 1982.   pp. 88-89.
                                  9-96

-------
             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 EPA1s_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 Laughlin 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


August 3, 1982
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 1isting notice.
Science Advisory Board review of the listing notice.
                                 A-8

-------
                  APPENDIX B



INDEX TO ENVIRONMENTAL IMPACT CONSIDERATIONS

-------
B-2

-------
                                                    APPENDIX B

                                   INDEX TO ENVIRONMENTAL IMPACT CONSIDERATIONS
oo
i
oo
       Agency guidelines for preparing regulatory
         action environmental impact statements
       	(39 FR 37419)	

       1.  Background and description

         a.  Summary of regulatory alternatives
         b.  Statutory basis for proposing standards
c.   Relationship to other regulatory agency
    actions
         d.   Industry affected by the regulatory
              alternatives
         e.  Specific processes affected by the
             regulatory alternatives
                                              Location within the  Background Information Document
The regulatory alternatives from which standards
will be chosen for proposal are discussed in
Chapter 6, Section 6.2.

The statutory basis for proposing standards is
summarized in Chapter 2.

The various relationships between the regulatory
agency actions are discussed in Chapters 3, 7,
and 8.

A discussion of the industry affected by the regula-
tory alternatives is presented in Chapter 3, section
3.1, pages 3-1 through 3-6.  Further details cover-
ing the "business/economic" nature of the industry
is presented in Chapter 9, Section 9.1.

A detailed technical discussion of the sources
and processes affected by the regulatory alterna-
tives is presented in Chapter 3, Section 3.2.
                                                                                            (continued)

-------
                      INDEX TO ENVIRONMENTAL  IMPACT  CONSIDERATIONS  (continued)
Agency guidelines for preparing regulatory
  action environmental  impact statements
	(39 FR 37419)	       Location within  the  Background  Information  Document

2.   Impacts of the alternatives

  a.   Air pollution                             The  air pollution  impacts  of  the  alternatives  are
                                                discussed  in  Chapter 7,  Section 7.1.

  b.   Water pollution                           The  water  pollution  impacts of  the  alternatives  are
                                                discussed  in  Chapter 7,  Section 7.2.

  c.   Solid waste disposal                       The  impact on solid  waste  disposal  is  discussed  in
                                                Chapter 7,  Section 7.2.

  d.   Energy impact                             The  energy impacts are discussed  in Chapter 7,
                                                Section 7.2.

  e.   Economic impact                           The  economic  impacts are discussed  in  Chapter  9.

-------
       APPENDIX C



EMISSION SOURCE TEST DATA

-------
C-2

-------
                                   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.   Sattery 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, Clairton, 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)--$ixty-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)--$ixty-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)--Eightyfive 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-nine 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 (1979)—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
CS
A8
B3
C8
A10
BIO
CIO
A12
B12
A22
B22
Al
Bl
Cl
= 5.2
C12
B12
B14
C14a
A16
816
C16
A18
B18
CIS
A20
820
A22a
A9
89
C9
All
Bll
Cll
A13
= 5.6
B3b
C3
A5
B5
C5
A7C
B7
C7
A9
B9
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
3
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
R
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
All*
Blld
A19
819
C19
A21
B21
A23
B23
= 6.7
A15H
B15d
CIS
A17
B17
C17
A19
B19
C19,
A21C
B21,
A23d
8»
B8
C8
A10
BIO
CIO
A12
= 4.2
B2b
C2
A4
84,.
C4C
A6
B6
C6
AS
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
6

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

aSteam aspiration turned off or  plugged steam nozzle.
bLong charge time (exceeding the mean charge time by  30  percent or
 more).
Misalignment of the larry car,  shrouds, or drop sleeves.
dlnterference with the observers'  ability to read emissions, not
 used.
                              C-9

-------
          TABLE C-2.   DATA ON VISIBLE EMISSIONS FROM CHARGING,
                          J&L STEEL, PITTSBURGH

                                             Duration of visible
 Battery      ..Date                          emissions (seconds)

   P2       12-12-78                                5
                                                    7

            12-13-78                               28.8a
                                                   12.2
                                                    6.0
                                                    6.0
                                                    7.4
                                                    3.8

            8-15-79                                 3
                                                    0
                                                    1.5
                                                    9
                                                    5

            8-16-79                                 3.6
                                                    7.0
                                                   13.6
                                                    6.2
     Mean =7.4

   P4       8-15-79                                 3.6
                                                    2
                                                   11
                                                    8
                                                    5
                                                    7
                                                    3.4
                                                   11
                                                    6
                                                    5

     Mean =6.2

aProblems with door placement.
                                  C-10

-------
TABLE C-3.  DATA ON VISIBLE  EMISSIONS FROM CHARGING,
          U.S. STEEL,  FAIRFIELD,  JUNE 1979
Battery Date
2 6-25










6-26












6-27














6-28















Mean = 25.3
Oven
no.
47
52
57
4
14
19a
24
29S
9b
39
44
19
9,
24'
34C
29
39
44
49^
54cf
. lf
6
119
16
44e
49
54
1
6
11
16
21
26
36
41
46
51
56
3
56d
3
8
13Q
18e
23
28
33
38
43
48
53
5
10
15
20

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
120
32
2
51
285
42
3
8
2
1
2
6
6
24
1
2

                                               (continued)
                        C-ll

-------
                    TABLE C-3.  (continued)
Battery Date
S 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
4.
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
7-9-74

















Mean

7-10-74




























Mean
9-30-75




Oven no.
38
48
58
68
78
1
Ha
21a
31s
41b
51
61
71
81
3
13
23
33
= 30.4

63
73
83
5
15
25
35
45
55
65
75K
85b
7
17
27
37
47
57
67
77
9
19
29
39
49
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
--
5
11
9
4
6
4
7
6
95
8
13
7
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

V
7
8
30
4
9
6
3
6
7
8
8
45
--
8
12
10
4
7
7
7
6
63
10
22
10
7
6
9
14

4
8
4
5
2
                                            (continued)
                       C-13

-------
TABLE  C-4 (continued)
Date
9-30-75









Mean
11-30-76














Mean
12-1-76









Mean
12-6-76














Oven no.
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
l^K
47b
29
39
49
59
79
2
= 22.9
4
14
24
34
44
54
&4K
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
6
15
2
4
42
10
14
6
7
3
6
3
visible emissions (seconds)
Observer 2
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
5
11
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
7
15
2
4
47
8
16
6
3
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.
71r
81C
3
13
23
33
43
53
63
= 6
65
75h
85b
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
?.o
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)
                         Duration  of visible emissions  (seconds)
Date      Oven no.      Observer  1      Observer 2      Observer 3
12-9-76 24
34
54
54
74.
84b
6
16
26
36
46
56
66
76
8
18
28
38
48
58
68
78
1
11
31
21
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
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
35
4
7
3
8
158
11
4
5
3
8
3
8
7
4
2
3
24
3
3
3
3
3
3
3
4
     Mean = 13.6
 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
Oven
Date no.
6-25 57a
67
9.
19b
29
39
49
59r
69C
2
12
22
32
42
52

6-26 27
47

57
67
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
62H
4d
14
24

6-27 52
62
4
14
24
34
44-
T
54T
46
56
66
8
18
28g
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.
 Last oven, jumper pipe used.
 Emissions from No.  2 boot during leveling.
eLiquor 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
814
C14
A16
816
C16
A18
B18
CIS
A20
= 11
A26
B26
C26
A28
B28
Al
81
Cl
A3
S3
= 14
C19
A21
B21
C21
A23
B23
C23
A25
625
C25
= 8
B27
C27
A29
829
A2
B2
C2
A4
84
C4
A6
= 9
A14.
B14b
C14
A16.
B16b
C16
A18
= 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
11
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
11
6
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
3
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
8
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.
11-2-77











Mean
11-3-77
























Mean
11-4-77














Allb
811
Cll
C13
A15b
815,.
C15C
A17.
B17b
C17
A19
819
= 13.5
A29
829
85
A2
C2
825
C25
A27
827
C27
C21
A23
823
C23
A25fa
A19b
B19b
C19D
A2L
821b
A8
88
C8
A10
810
= 8
828
Al
SI*
Cld
A3
C24
A26
826
C26
A28
822
85
C22
A24
824
Observer 1
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
3
38
14
0
8
16
1

21
4
1
4
2
5
8
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
11-4-77














Mean
Oven no.
814
C14
A16
B16
C16
CIO
A12
B12
C12a
A14e
A8b
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

 Suspected  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)
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
CIS
A18
B18
CIS
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
A10
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,
fi
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
C18
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
24.0C
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
812
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
B24
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'°n
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.
cDelay 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.
^Rapid charge rate, end of shift,  heavy emissions out of all  lids.
 Steam 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
84
A6
B6
= 8.6
A2
B2
C2
A4
84
A6
= 7.8
B9
All
Bll
A13
A15
BIS
= 7.6
Cl
B3
A3
85
AS
87
= 8.9
Bll
A13
B13
A15
815
A17
= 8.4
A25
B25
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
B14
A16
816
A18
818
A20

86
AS
A14
814
A16
816

A17
817
A23
823
A25
825

Bll
All
B13
A13
BIS
A15

817
A19
819
A23
823
A25

A31
831
B6
A3
B8
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
3
6

3
5
4
4
11
3

Oven
no.
820
A22
B5
A7
B7
A9


A18
818
A20
B20



A27
827
A29
B29



817
A17
819
A19



B25 .
A27
B27
A29



BIO
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
829
= 11.6
A16
316
CIS
A18
812
C12
= 6.8
A21
821
C21
A23
823
C23
= 13.1
Visible
emissions
(seconds)
13h
18h
14h
13h
5h
10h

8,
14k
7
8
8
5

15
4
2
7
4
9

Oven
no.
A31
831
A3
88
A10
810

A14
814
A16
C22
A24
B24

Cll
A13
813
C13
C23
A25

Visible
emissions
(seconds)
8h
19b
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
827
C27
A29
829

Visible
emissions
(seconds)
8
7h i
29- -1
11^



7
8
2
2



1
32m
8
1
6
4

aPoor 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:   lidraan failed  to remove all  lids before charging.
IT
 Poor drop sleeve  alignment.
 Most emissions from  lids No. 3 and No. 4 during No. 2 lid discharge and
 leveling.
""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.
83
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
°b
13D
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

 "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






Mean =
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
10.2
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


                        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

Visible
emissions
(seconds)
26
2
1.4
4
2a
43
35
9


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
3a
4a
Mean = 10.8
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





22






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
8-7
8-8

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
5

5
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
2.9
3.0
2.9
2.3
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

                               Start        Finish          Percent
Battery          Date          time          time         leaking lids

   2             6-25          1210          1214              0

                 6-26          0817          0820              0.5

                 6-27          1100          1102              0
                               1557          1600              0

                 6-28          0907          0909              0
                               1529          1532              0

   9             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

Battery
0.0
0.6
0.0
0.0
0.6
0.6
0.6
Percent leaking lids .
PI 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, FAIRFIELO, 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
358
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












aMultiple 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 Date
5 4-21



4-22

4-23


Mean =0.1
6 4-21


4-22

4-23



Mean =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 Percent
time leaking offtakes
1223
0854
1118
1600
0915
1544
1606
1054
1546
0841
1120
1434
10.0
15.9
2.7
1.8
0.9
14.3
0
0
0
1.7
0
0
DATA ON VISIBLE EMISSIONS FROM OFFTAKE LEAKS,
J&L, PITTSBURGH, BATTERY P4





Percent leaking
offtakes
0.0
5.1
1.3
0.0
2.5
0.0
0.0





                        C-48

-------
                                    TABLE  C-35.  VISIBLE  EMISSION  DATA ON OFFTAKE  LEAKS  -  KAISER  STEEL.  FONTANA, CALIFORNIA
Percent leaks
Battery A


O
I
*>
UD





Luting3
manpower Date
1 April 1977
May 1977
June 1977
April - June 1977
2 July 1977
August 1977
September 1977
October 1977
November 1977
December 1977
July - December 1977
Battery B
Battery C
Battery 0
Battery E
Battery F
Battery G
Percent of Percent of Percent of Percent of Percent of Percent of Percent of
readings at readings at readings at readings at readings at readings at readings at
Avg. 5 or less Avg. 5 or less Avg. 5 or less Avg. 5 or less Avg. 5 or less Avg. 5 or less Avg. 5 or less
6. 1
4.3
4.4
4.9
2.7
2.2
2.0
2.9
3.4
4.1
2.9
36
69
57
54
78
91
91
82
71
57
78
5.1
4.0
3.8
4.3
2.9
2.1
1.2
2.1
4.2
4.4
2.8
53
69
74
65
76
86
100
96
57
57
79
12.0
11.4
8.9
10.8
4.7
3.8
4.1
5.3
5.8
4.3
4.7
16
14
29
20
56
72
68
54
43
57
58
4.7
4.4
3.0
4.0
1.8
2.2
1.9
1.8
1.8
2.3
2.0
58
62
88
69
96
90
97
93
100
86
94
5.3
5.6
2.7
4.5
1.9
2.0
1.6
2.0
2.6
22
2.1
40
42
86
56
95
95
97
89
86
100
94
4.9
5.3
7.0
5.7
4.3
3.4
2.8
2.7
3.8
4.8
3.6
56
56
28
47
58
76
87
86
79
43
72
7.1
7.6
8.2
7.6
3.9
3.4
2.4
5.1
4.3
3.4
3.8
22
23
21
22
73
81
87
46
64
86
73
Approximate employees per shift per battery responsible for luting and tending lids.

-------
TABLE C-36.   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
2.6
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
(continued)
       C-51

-------
                          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

-------
o
on
                      TABLE C-38.   BENZENE  SOLUBLE ORGANIC  CONCENTRATIONS  AND  EMISSION  RATES FROM
                                     TESTS  AT  ARMCO,  INCORPORATED,  HOUSTON,  TEXAS1
Stack gas
parameters
Sample Sample
location number
Inlet 1
2
3
Average
Outlet 1
2
3
Average
Date
(1979)
10-4
10-5
10-5

10-4
10-5
10-5

Flow rate
(ft3/min)a
176,000
174,000
172,000
174,000
153,000
149,000
148,000
150,000
Temperature
(° F)
95.0
105.0
109.0
103.0
79.6
77.3
66.9
74.6
Concentration
g/ft3
0.010
0.017
0.020
0.016
0.007
0.011
0.011
0.010
mg/m3
22.8
37.8
45.0
35.2
16.0
25.4
24.9
22.1
Emission rate
Ib/hr
15.0
24.6
28.9
22.8
9.18
14.1
13.8
12.4
kg/hr
6.83
11.2
13.1
10.4
4.16
6.41
6.27
5.61
      Dry standard measurement.

-------
              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
Ib/ft3*
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/m3
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

-------
                    TABLE C-41.   SUMMARY OF DOOR LEAK OBSERVATIONS AT ARMCO, INCORPORATED, HOUSTON, TEXAS1
en
cr>
Sample
number Observer

BSO test
no. 1
Average

BSO test
no. 2
Average

BSO test
no. 3
Average
1
2
3

1
2
3

1
2
3

Total3
number of
leaking
doors
46
28
28
34
68
81
67
72
51
35
51
46
Total
emission
rate for
entire
battery
12
9.0
9.0
10
22
22
22
22
21
14
21
19
Coke oven door emission
doors) by
1
PS
9.7
8.1
6.5
8.1
18
16
18
17
15
9.7
15
13
2
CS PS
16 11
9.7
6.5
NA 9.1
-- 25
29 19
24
NA 22
13
13
42 13
NA 13
CS
--
15
--
NA
--
--
32
NA
45
--
—
NA
3
PS
9.7
9.7
11
10
16
19
18
18
9.7
6.5
13
9.7
rate
run

CS
--
--
11
NA
31
--
--
NA
--
27
—
NA
(percent of
number
4
PS CS
8.1 19
3.2
9.7
7.0 NA
23
16 31
16
18 NA
-b
_-b
__b
--b
total

Average
PS CS
13
7.7
8.4
9.7
20
18
19
19
13
9.7
14
12
18
NA
NA
NA
NA
30
NA
NA
NA
NA
NA
NA
       PS = pusherside.
       CS = cokeside.
       NA = not applicable.

        These numbers  are the combined number of leaking doors from all the runs per observer.
        Because of darkness,  Run 4 was not conducted.

-------
                   APPENDIX D



EMISSION MEASUREMENT AND CONTINUOUS MONITORING

-------
D-2

-------
        APPENDIX D - EMISSION MEASUREMENT AND CONTINUOUS MONITORING

D.1   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 parts:   Part A—Determination of Visible Emissions
During the Oven Charging Period, Part B—Determination of Visible Emissions
from Coke Oven Topside Leaks, and Part C--Determination of Visible Emission
from Coke Oven Doors.  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.I.I  Charging Operation
     Reference Method—Part A was developed to measure emissions  from larry
car systems charging wet coal.   A discussion of the rationale behind the

                                  D-3

-------
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 assign-
ing 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 quantitative 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 encompasses 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 repre-
sentative 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  considered.
The potential advantage of these techniques is that a distinction could be
made between small and large quantitites of emissions that are  emitted in
the same length of time.   Each of these techniques proved unsatisfactory or
                                  D-4

-------
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.
0.1.2  Topside Leaks (Offtakes and stationary  jumper pipes, topside ports,
and collecting mains)
                                 D-5

-------
     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
not appear to be any correlation between plume color and the quantity of
emissions.   The second criterion examined for determining  whether  or not a
                                 D-6

-------
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
this variation is also attributed to the observer location during the
                                  D-7

-------
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
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.
0.2  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 109C.
     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 109C



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 109C 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.
                                      - 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

METHODOLOGY FOR ESTIMATING INCIDENCE OF LUNG CANCER AND
        MAXIMUM LIFETIME'RISK FROM EXPOSURE TO
                  COKE OVEN EMISSIONS

-------
                              APPENDIX E
                   SUMMARY OF CANCER RISK ASSESSMENT

E.I  INTRODUCTION
     The purpose of this appendix is to describe the methodology and
to provide the information used to estimate the incidence of lung
cancer and maximum lifetime risk from population exposure to coke oven
emissions from certain emission sources at wet-coal-charged coke oven
batteries.  The methodology consists of four major components:   esti-
mation of annual average concentration patterns of coke oven emissions
(expressed as benzene soluble organics [BSO]) in the region surround-
ing each plant, estimation of the population exposed to each computed
concentration, calculation of exposure by summing the products  of the
concentrations and associated populations, and calculation of esti-
mated annual lung cancer incidence and estimated maximum lifetime risk
from the concentration and exposure estimates and a health effects
estimate represented by a unit risk factor.   Due to the assumptions
made in each of these four steps of the methodology, there is con-
siderable uncertainty associated with the lifetime individual risk and
lung cancer incidence numbers calculated in this appendix.   The num-
bers presented could be either underestimates or overestimates.   These
uncertainties are explained in Section C.6 of the appendix.   A  descrip-
tion of the health effects and derivation of the unit risk factor for
coke oven emissions is not included in this appendix.   EPA's Carcinogen
Assessment Group (CAG) presented the background literature on the
carcinogenicity of coke oven emissions in a separate EPA document.1
The CAG document includes an assessment of the carcinogenic potency
                                  E-2

-------
(unit risk) of coke oven emissions as indicated by occupational exposure
to coal-tar pitch volatiles, a measure equivalent to the BSO fraction
of coke oven emissions.1
E.2  ATMOSPHERIC DISPERSION MODELING AND PLANT EMISSION RATES
     The Human Exposure Model (HEM) was used to estimate concentra-
tions of BSO around 53 by-product coke plants that operate wet-coal
charged batteries (Table E-l).2  The HEM estimates the annual average
concentrations resulting from emissions from point sources.   The
dispersion model within the HEM is a Gaussian model  that uses the same
basic dispersion algorithm as the climatological form of EPA's Clima-
tological Dispersion Model.3  Gaussian concentration files are used in
conjunction with STAR data and emission data to estimate annual average
concentrations.   Details on this aspect of the HEM can be found in
Reference 2.
     Seasonal or annual stability array (STAR) summaries are principal
meteorological input to the HEM dispersion model.   STAR data are
standard climatological frequency-of-occurrence summaries formulated
for use in EPA models and available for major U.S.  sites from the
National Climatic Center, Asheville, North Carolina.   A STAR summary
is a joint frequency-of-occurrence of wind speed and wind direction
categories, classified according to the Pasquill stability categories.
For this modeling analysis, annual STAR summaries were used.
     The model receptor grid consists of 10 downwind distances located
along 16 radials.  The radials are 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,  0.3, 0.5, 0.7, 1.0,
2.0, 5.0, 10.0,  15.0, and 20.0 kilometers.   The center of the receptor
grid for each plant was assumed to be the plant center as determined
by review of maps.  Industry provided confirmation where ambiguities
existed.  All concentration calculations were performed assuming the
meteorologic conditions of an urban area.
     Plant emission rates for each emission point (charging, door
leaks, and topside leaks) used as inputs to the dispersion model are
                                  E-3

-------
TABLE E-l. BY-PRODUCT COKE OVEN PLANT LOCATIONS
State, city
Alabama
1. Tarrant
2. Holt
3. Woodward
4. Gadsden
5. Thomas
6. Birmingham
7. Fairfield
California
Company
Alabama By-Products Company
Empire Coke Company
Koppers Company, Inc.
Republic Steel Corporation
Republic Steel Corporation
Jim Walter
U.S. Steel Corporation

  8.    Fontana

Colorado

  9.    Pueblo

II linois

  10.  Granite City
  11.  Chicago
  12.  South Chicago

Indiana

  13.  Burns Harbor
  14.  Indianapolis
  15.  Terre Haute
  16.  East Chicago
  17.  Gary
  18.  East Chicago

Kentucky

  19.  Ashland


Maryland

  20.  Sparrows Point
Kaiser Steel Corporation
CF&I Steel Corporation
National Steel Corporation
Interlake, Incorporated
Republic Steel Corporation
Bethlehem Steel Corporation
Citizens Gas & Coke Utility
Indiana Gas and Chemical Corporation
Inland Steel Company
U.S.  Steel Corporation
Jones and Laughlin Steel Corporation
Solvay Division Allied Chemical
 Corporation
Bethlehem Steel Corporation
                                                          (continued)
                                  E-4

-------
TABLE E--1 (continued)
State, city
Michigan
21. Detroit
22. Dearborn
23. Detroit

Missouri
24. St. Louis
New York
25. Buffalo
26. Lackawana
27. Buffalo
Ohio
28. Ironton
29. Hamilton
30. Middletown
31. Portsmouth
32. Toledo
33. Cleveland
34. Massilon
35. Warren
36. Youngstown
37. Lorain
38. Campbell
Pennsylvania
39. Bethlehem
40. Johnstown
41. Aliquippa
42. Pittsburgh
43. Erie
44. Philadelphia
45. Pittsburgh
46. Clairton
47. Fairless Hills
48. Monessen
Tennessee
49. Chattanooga
Company

Detroit Coke
Ford Motor Company
Great Lakes Steel Division
National Steel Corporation

Carondolet Coke

Tonawanda Coke Company
Bethlehem Steel Corporation
Donner-Hanna Coke Corporation

Ironton Coke
Armco Steel Corporation
Armco Steel Corporation
New Boston Coke
Koppers Company, Incorporated
Republic Steel Corporation
Republic Steel Corporation
Republic Steel Corporation
Republic Steel Corporation
U.S. Steel Corporation
Jones and Laughlin Steel Corporation

Bethlehem Steel Corporation
Bethlehem Steel Corporation
Jones and Laughlin Steel Corporation
Jones and Laughlin Steel Corporation
Koppers Company Incorporated
Philadelphia Coke Division
Shenango Incorporated
U.S. Steel
U.S. Steel
Wheeling-Pittsburgh Steel Corporation

Chattanooga Coke and Chemical
                                  (continued)
         E-5

-------
                         TABLE E-l (continued)
   State, city
             Company
Texas

  50.  Lone Star

Utah

  51.  Provo

West Virginia

  52.  Weirton

  53.  Brown's Island

  54.  E. Steubenvi1le

Wisconsin

  55.  Milwaukee
Lone Star Steel Company
U.S. Steel Corporation
Weirton Steel Division, National Steel
 Corporation
Weirton Steel Division, National Steel
 Corporation
Wheeling-Pittsburgh Steel Corporation
Milwaukee Solvay Coke
                                  E-6

-------
given in Tables E-2, E-3, and E-4.  Both maximum and minimum emission
estimates were computed for each plant based on the range of emissions
estimates presented in Chapter 3.
E.3  POPULATION AROUND PLANTS CONTAINING-FUGITIVE EMISSION
     SOURCES OF BSO
     The HEM was used to estimate the population that resides in the
vicinity of each receptor coordinate surrounding each coke plant.  A
slightly modified version of the "Master Enumeration District List--
Extended" (MED-X) data base is contained in the HEM and used for
population pattern estimation.  This data base is broken down into
enumeration district/block group (ED/BG) values.   MED-X contains the
population centroid coordinates (latitude and longitude) and the 1970
population of each ED/BG in the United States (50 States plus the
District of Columbia).   For human exposure estimations, MED-X has been
reduced from its complete form (including descriptive and summary
data) to produce a randomly accessible computer file of the data
necessary for the estimation.   A separate file of county-level  growth
factors, based on the 1970 to 1980 growth factor at the county level,
has also been created for use in estimating 1980 population figures
for each ED/BG.   The population "at risk"  to BSO exposure was considered
to be persons residing within 20 km of coke plants.   The population
around each plant was identified by specifying the geographical  coordi-
nates of the plant.
E.4  POPULATION EXPOSURE METHODOLOGY
E.4.1  Exposure Methodology
     The HEM uses BSO atmospheric concentration patterns (see Sec-
tion E.2) together with population information (see Section E.3) to
calculate population exposure.  For each receptor coordinate, the
concentration of BSO and the population estimated by the HEM to  be
exposed to that particular concentration are identified.   The HEM
multiplies these two numbers to produce population exposure estimates
and sums these products for each plant.   A two-level scheme has  been
adopted to pair concentrations and populations prior to the computation
of exposure.   The two-level approach is used because the concentrations
                                  E-7

-------
                                         TABLE  E-2.   TOTAL BSO EMISSIONS  FROM CHARGING  (Mg/Yr)
CD
                                                      nan EMISSIIIHS IHC/TIII muM CHARUINI:
                                                MINIMUM
                      HlAIII     SHUNI UAIII.MIM            l«Ll flMICHIES
                              «Ll I   'II II  Ail III     Alt I   «LI II   All III
 9HIIHI BAIIEHU3
At I  I   All II  ALI III
 IALI IIAMtNHS
All  I   All  ||  AI.I  III
1
5
4
•i
A
/
ft
if
10
It
| 5
M
in
IS
1 A
1 n
If
I"
19
i?0
*2I
22
/I
2«
2s
2>-
2;

1)
11
oS
u . uoin
o.ulln
o.olon
•o.gost
u.oo'.r
o.vl3o
o.ulon
O.IMIW
u.vosl
u.o
o.uo/s
o.uoii
o.ulnn
o.ol?"
0.002J
v.o
u.unni
v.oo.,2
0.1)0??
O.U"2I
II. 0011 #
o.oosi
O.OOhS
u.un/a
U . u 0 *j 1
U , OOS1
0.0d
u. til no
V . 1 n i \
o.u
I' . u f U 0
O.UI •>»
u.uoiis
U.UH 10
U.UIi>0
II . U II 1 1
V . U II 1 1
(J.VIIJt
0.0020
O.fluSh
0.0054
0.002'
0. no ii)
0 . on M
(i.nion
o.ov4»
n.oixt
0.0
0.00??
o.nols
0.1111,11
II . 0 1 0 II
Q.OOIV
o o?(*o
0.0
O.OOM
o.ni.t?
o.no??
o.oul^
O.OOht.
O.OU1U
0. ooao
0.0016
0.0l»<>
o. no jq
0.0014
o.oi?o
0 . 00 1 3
o.o
0.01 10
0 . oa'l'l
a. onus
II . 0 n 5 U
O.OI?U
0 . 0 II 1 J
n . ou I /
0.01124
o.ontn
o.oo IS
a.aoii
g.uni i
U.OOIH
o.ooso
O.onvl
o.omr
n.aoik
u.o
O.IIOI I
d.onov
o.oov;
0.11061
U.DOI?
0.0
0.00-iJ
O.D04I
o.ooi q
o.uouv
u . n o 1 0
O.OO^J
o.oo«r%
o.ur^n
U.OOI9
O . OO/ 1
o.ooo1'
U.OOMH
O . u*l u 6
o.u
tf. 01 ID
o.onsn
ii.miiin
o.oo,!a
v.uoni
u.ono"
u . v n i I
U . U II 1 K
0.0
0.0
0.0
0.0
0.0
o.ooio
0.0
b . 0
0.0
0.0
o.ongl
o.onjt
0.0
0 • OOrf?
0.0041
0.0
0.0009
0.0
0.0
O.U035
0.0
0.0
O.OOJB
0.0
o.u
0.0
(1.00 If,
o.u
0.0
0.0
o.u
0.0
o.u
0.0
• o.u
o.onjo
0.0
0.0
u.u
V.I)
u.u
0.0
0.0
u.o
0.0
0.0
o.o
0.0014
o.e
o.o
o.o
o.o
0.0
0.00t>8
o.colo
o.o
0 . DO 1 1
o.oois
o.o
o.ouoy
0.0
0.0
O.OUl't
0.0
0.0
0.0012
o.o
0.0
o.o
0.004/
A . O
o.o
0.0
0.0
o.o
o.o
o.o
0.0
o.ooio
o.a
n. o
0.0
o.o
n. if
0.0
0.0
0.0
0.0
0.0
0.0
0.0014
0.0
0.0
0.0
o.a
u.o
U.VOfc}
o.ooi r
o.u
0 .00 lA
0.6011
o.u
0.0009
o.o
• .0
o.u
U.OOJ5
u.u
o.u
O.UO Jl
0.0
o.u
0.0
o.ooos
o.u
v.o
o.u
0.0
u.u
u.u
0.0
*>. a
o.uoio
u.u
tf • u
0.0
o.u
u.o
0.11
1 . MIOO
1 . 2000
q. 5noo
0. JlMIO
Z.luoo
2.HU00
4.7000
I.MIOO
0.4|00
1 .1000
1 .3ung
O.I VOO
0.0
O.hfOO
0.40U
1 .201)0
0.4100
0.41,00
0.3oou
O.I VOV
0.0
o.lsoo
O.I21U
1 .0000
1 .Jono
O.H20U
O.looo
0.0
o. nou
O.r>600
0. 140V
0. lido
0.5600
0. 1100
0. 1)00
0. Jnou
O.?l.0o
0. 1/00
1 .1 IIIIU
o.o
o.v,oo
1 .10110
0 . t V O O
o.i'ino
o.itoa
0 . .1 / Ptl
1 . lono
o. l/oa
0. I >PV
O.f.f"u
0.0
o.o
o.o
0.0
0.0
o.o
o. ti.no
o.u
0.0
o.n
0.0
a. o
;. moo
O.S600
o.n
W.5VOO
1 . 1 V U i>
0.0
0.1000
0.9
0.0
o. rsov
U . 0
o.n
1 . ouov
0.0
0.0
o.u
2. 'ooo
o.o
0.0
0 . fl
II .11
u.u
0.0
n . n
n.ivoo
o.n
0.0
O . 0
o.o
'I. U
n .u
0.0
0.0
o.u
0.0
u.o
u.o
0. IhUO
0.0 .
u.o
u.u
0.0
o.v
1 .Sooo
0.39UO
0.0
U.«i>00
v . F f> u u
o.o
u. louo
0 . 0
0.0
u.u
0. /Son
o.u
u.u
0. J 100
u.o
o.u
u.v
1 . onoo
u.u
v.u
u.o
u.u
u .u
o.u
v.u
u.v
u.r>"*uo
u.o
V . U
u.v
u.v
V . U
u . u
n.o
0.0
o.v
o.o
n.o
II. 0
0. 16110
O.D
0.0
o.o
n.o
0.0
1 . 3000
0. 1-jOO
o.o
o. thmi
n . f.'JW

-------
                                      TABLE  E-2   (continued)
                               IISII CMUJIIINS  (MG/TIM (HUM CHARUINi;
                        MINIMUM
      SIMIIII iKiiiuita             1*1.1. luficiittj
     • I. I  I   Ml II   «LI  III     *LI I   •!. I  II  »LI III
                    M«I|HUM
 SMIIlM  l)*l If 11113             I«I.L 11*11111115
.M.I  I    Al.l II  «LI III    *l. I  I   Ml II  »LI
                                                                                                            III

ni
in
ni
SO
SI
S2
S 1
S4
5%

0.00(1
o.on JS
0.0010
0.0
0.0100
O.OOS2
0.0
o.oorn
0.004)

O.Ooll
O.OUJS
o.ooio
0.0
0.0100
O.OU12
0.0
o.oos>
o.oois

0.0044
O.OD/q
O.UOO 1
0.0
o.ooio
o.uosl
0.0
0.00)6
O.OOIS

0*0 . 0
0.0 . 0
o.o .0
0.0 .0
0.0017 .0012
0.0 .0
0.0 .0
0.0044 .00)0
0.0040 .00)4
0.0 .0

0*0
0.0
0.0
0.0
0.001}
0.0
o.o
0.00)6
0.00)7
0.0

? . s o oo
1 .4000
O.hVOO
0.7000
0.0
7.0000
1.4000
o.o
i.aooo
7.0000

f • j o o o
l.qooo
0.6900
0.2000
0.0
7.0000
l.qooo
0.0
1 .3000
O.S400

) . no oo
O.dliOO
0. SlOO
0 .0** 2U •
0.0
0.9SOO
0.1400
0.0
O.U SOU
6.21 oo'

o.u
0.0
o.o
0.0
0. ISOO
0.0
0.0
1 .2000
1 .0000
0.0

u . o
0.0
0.0
o.o
0. ISOO
0.0
u.o
O.t>)00
o. /quo
u.o

0 .
0.
0.
0.
0. SOU
0.
0.
0. 1600
O.hfOO
0.0
Emissions  for  dry-coal charged  batteries are not  estimated and are  recorded as 0.0.

-------
TABLE E-3.   TOTAL BSD EMISSIONS FROM DOOR LEAKS (Mg/Yr)
PIANI
,
2
3
•
5
6
r
•
n
10
II
I y
\ 3
14
15
Id
i r
IN
2n
*2I
22
J\
24
*"*
26
27
2n
29
10
) 1
J 7
3J
34
*3S
37
)H
39
40
41
42
4 3
44
45
3M(|H 1
>ll 1
1 .8000
1 .4000
5.0000
2. 9000
1 . flOvO
2. (000
3.6000
2.5000
O.b/UO
1 .2000
0./400
0.5'00
0.0
1 .bOOO
1 . 1 000
5.9000
4. IVUO
0.6500
5.0000
o.u
1.6000
1 .2000
2. 1000

1 .20OO
l.20on
3 . *uoo
2.9000
2.00UII

;..touo
0.d2UO
0.0
2 . UOUO
1 1 .UOllO
b .UOUO
3.5000
1 . II Duo
1 .1000
4.0000
U c / 2u 0
U .9IIUO
1 .1000
HllllHIt
All II
l.oonu
0.7500
2. nooo
1 .6000
O.PUOU
t.soou
1 ."ooo
2.5000
0.6700
1 .2000
0. >90U
0.5900
0.0
1 . 1 onrt
0.7500
4.1000
3.0000
0.6500
5.0000
0.0
1 .6000
1 .2000
1 .2000

1 .2000
1 .2000
1 . 6 0 0 II
1 .Qono
0.9000

3.4000
o. J'too
0.0
5. luuo
2.9000
1.5000
1 .noon
1 .1001,
4.001)0
0 . f 2 no
0 .9UOO
1 . a o n ii
HSU EMISSIONS 
P 0 123.2000 D7.2UOO S'i.2000 1 1 . 7oou 1 5 . 6 II II U «4.nunu 1 0 . IIHOU IS.hKIIV 11.2000 a. 4000 11.3600 17.9200 U. 9600 16.0000 21.2600 20.0000 1.5040 13.4400 a.oino 6.6060 0.0 12.3200 4.4000 45.9200 53.1600 33.6000 ;.2 n u n Z.9I20 2.2400 H.SI20 4.6160 2.4b40 4.5V20 1 . 2HOU 12.3200 5.376U 3.6UOO 3.6V60 2.9120 0.0 3.74HO 2.2400 5H.2400 2.9I2U 21. MOO 0.0 1 . M (, 0 5.9360 3 .5U4U 5.6000 « . 720O 4. I44U 3.3600 7 5 M, 0 ;| *n ,. . i < nu 12. 1200 1.7320 0.0 IS. b iino Il.b240 9. (inu 2.9I20 I.V4QV 12.1200 ^ . b tt r u 3.920V o.n n.o o.o . o.o o.o o.o S.0400 0.0 n.o 0.0 o.o o.o 33.6000 14.5600 o.o : IO.«I60 It,. nooo 6.1600 o.o o.u o.n 12. 3200 0.0 0 0 o.o o . n o.o 33.6000 0 U 0 0 0.0 v.n . n.v 0.0 O.n 1 1 ,2oou n.o O.ll O.n n.v n.n UAIUMIEM Al.l II All HI o.u o.o o.v u.o o.o 0.0 5.0400 0.0 o.n o.u v.u ' o.o 22.40UO 10.0(100 0.0 I2.32VO o.v 6. IbUO 0.0 o.u v.u 12. 32UO u.o U 0 o.o u.o o . u I5.bnv0 ' v.v v.o v.u v.u u.u 1 1 .20UO u.v v.v v.v u.o u.v 0.0 o.o 0.0 0.0 o.o 0.0 5.0400 o.v o.o 0.0 o.n 0.0 |1.',OOU 6.31140 n.n 4.411110 u.o 6. IllOO n.o O.I) n.o 7.5040 0.0 0 0 n.o 0 . 0 V . V iq.uliov ' u.o o.u v.o 0.0 o.n 1 .05611 O.I) o.v n.o 0.0 U . II

-------
                                 TABLE E-3  (continued)
IIJII FHI33IIIHS IHf./tKI (HUM
MINIMUM
I'l.ANI SIH.IHI
«LI 1
at, ll.sooo
47 ? . 2pOO
•IX I.2HVO
40 O.ASWO
so o.o
SI 2.0000
•,? 1.3000
s * o.o
54 1 .11000
s1; C.HOOO
II A 1 1 1. H I > »
AI. i it ALI 111
| 1 .SHOO
2 7n4p
1. 2HOO
fl.JSOO
0.0
7.0000
1. tooo
0.0
l.nono
o.nooo
3.3000
o . s**on
0.3200
0. 1400
0.0
o.nsoo
o. /loo
0.0
0. 4 7UO
0. 33uo
IALI. HAIKMU*
ALI i ALI ii
0.0
0.0
0.0
0.0
0.0570
0.0
0.0
1. looo
0.11V9
0.0
0.0
o.o
0.0
0.0
0.0570
o.o
0.0
1 .1000
0.9900
0.0
ALI ill
0.0
o.u
0.0
0.11
o.o5'o
0.0
0.0
0.6900
0.6)00
0.0
IIIIIIH HflUS
MA II MUM
SllllHI HUM HI If
ALI 1 ALI II
I^R.flooo
c! 4 . r. it n i>
1 3.4400
3.9200
0.0
22.400U
o.o
20.lt>00
n.ibOO
I2B.POOO
2 4 . f, 4 0 0
1 (.4400
• 3.9200
0.0
22.4000
14.5600
0.0
2U.lt.00
B.9600
3
ALI 111
.111. 9*00
1. . MIO V
3.SII40
I.S6KU
0.0
9.5^00
7.9'J2t»
0.0
5.Z»i40
3.696V
IAI.L 1
AI. 1 1
0.0
0 . o
0.0
. 0.0
O.MBQ
n.o
o.o
12.3200
II .OflflO
0.0
IIAI ILI'II^U
• II II
0.0
ft . O
v.v
v.o
<>
fl.O
Emissions for dry-coal charged batteries are not estimated and are recorded as 0.0.

-------
               TABLE  E-4.   TOTAL  BSO  EMISSIONS FROM TOPSIDE  LEAKS (Mg/Yr)
PIANI
                        Bsn EMISSIONS  (MC/TH) tuon tUPSIDE UAKS
                  MINIMUM
 aiiuui HI MIMIC)            IALL ruiltmts
•II I   All II  All III     «Lt I  »ll II  Alt  III
                  HO I MUM

 .1IIOHI MAlltHltS            IALI (IAIIIHIE3

all I   All II All III     At. I  I   »ll II  «LI III
1
2
3
4
6
1
II
V
in
1 1
12
1 3
la
IS
16
1 7
III
1 *>
2n
*2I
i?
23
2«
2S
i. t*
2»
2rt
2?
3D
II
»2
33
14
*3S
3h
37
3n
3'l
• 0
1 1
42
4.1
44
15
o. jnuo
t. inuo
1 . 3004
0.7600
0 . JPUO
l.ooon
I .0000
0.5I>00
0.0900
11. ill UK
0.4400
o.o
0. 3?0ll
0. 2200
1.4000
1 . UOUI)
0.3700
3.200H
0.0
U . (j4UO
V.46UO
0.2100
II. 1300
1) . IS0 1)
0.32UO
V.b4UD
0.11UO
U. 5*110
u. SSuo
o.^noo
1 .11100
II. 1500
O.u
0.1,1.01
2. IOUO
l.bnuo
0. 7IOO
0. I'-OO
O.ttllOO
0. I 7UO
0. I6UO
0.2100
0. IHOO
0. 7(100
0.450U
0 . 2 3 Ou
0.3SOU
0.6400
1.3000
O.SbOII
0.520U
0. 3000
0.2700
0.0
. 2bno
.1000
.looo
.0300
. 22no
.5000
.0
.6UOD
.looo
.2100
. 1 3 'I II
. 3*1 00
o. 320U
0. 3'ino
0. 320U
0. 3300
0.2IOU
0. 1 701)
0. 9UOO
0.01SO
o.o
0.470U
1 . 3ono
0.1411 II
0. 71 no
n. ibou
0.6000
0. 1 IOU
0. IbOV
0.1100
O.OKOO
0.3100
0.2000-
0 . 1000
U. Id 00
o.34on
O.bOUO
0.33UO
0.2400
0.1)00
0. 1 100
0.0
0. 1200
0.0120
0.4-700
0.3700
o. looo
1 .2000
0.0
u.2niio
0.2)00
o. 1 Ion
o.u«20
0.2IUD
o . I n u o
0. I5VO
0. 1 700
o . 1 o n o
0.07..0
0 . 1 '» U 0
O.OI'IO
o.u
U.25UO
o.i.tun
U . "1 1 U 0
0.41 UO
0. 1 100
I). 1100
U. 1000
a. inoo
0.0
0.0
0.0
0.0
0*0
0.0
0.2000
0.0
0.0
o.o
0.0
0.0
0.4600
0.2700
0.0
0.2200
0 . 5 30O
o.u
0.3500
0.0
0.0
0.0
0.2700
o.o
o.u
0 • 1 A 00
0.0
o.o
0.0
0.1)400
o.o
0.0
u.u -
0.0
u.o
u.u
u.o
o.u
0.2000
0.0
o.u
o.u
o.u
o.u
0.0
0.0
0.0
0.0
0.0
o.o
0.2000
0.0
0.0
0.0
o.o
0.0
0.61100
0.2100
0.0
0.2000
o » 450 o
0.0
0.2100
0.0
o.o
0.0
0.2500
o.o
0.0
0 1 1100
0.0
0.0
0.0
0.5000
0.0
0.0
n.o
o.o
0.0
n.o
o.o
o.o
0.2HOO
o.o
o.o
O.n
o.o
o.a
0.0
0.0
0.0
0.0
0.0
0.0
0.1300
0.0
0.0
o.a
0.0
o.a
0.3200
o.looo
0.0
O.OHBO
0.0
0.1000
0.0
0.0
0.0
O.I 100
o.o
o.u
0 1100
0.0
o.u
0.0
0.2600
0.0
u.o
u.u
0.0
o.u
u.a
o.u
0.0
0.1400
u.u
u.o
•1.0
U.O
u.o
2 4J20
1 .9200
A. 3200
«.«640
? • *•* 4 i? 0
3.7760
6.4000
11.5200
3.51140
5.6160
3.1360
2.ni60
0.0
2 . n >i n u
1 .4000
A. 9600
II 5/00
6.4000
2. JliflO
20.4(100
o.o
4.0V60
2.V440
1 .11560
O.C320
2.04(10
4.0V6U
3.4560
1. 51140
2.24UO
1 . 7120
C . **OOU
0.9600
0.0
4.2210
I 3 . 440U
I0.240U
4.S44U
1 . U24U
4.3520
1 .nnno
2.1 120
1.4720
1.1520
4.V920
2.HAOO
In I 3 a
• 1 1 < O
2.2400
I.OVbO
n. 3200
3.5040
3.32BO
1.9200
1 . 7200
0.0
1.6640
1. 1520
7.041)0
5.3120
1 .4000
la. oooo
0.0
3.1400
2.4320
1.3440
n.0320
2.0480
2.416U
2.0480
2. M20
1 . J440
1 .UNDO
S. 1600 •
0.6000
o.u
3.00(10
0.3200
.oluo
.5440
.0240
fl 7 20
.1520
.OCBO
.viva
2. 1120
0.7040
n.MPo
2.416U
l.2»no
1.0240
2.1/60
o. 3520
2.1 I2U
1.5360
0.0320
0. 7040
0.0
0. 7hl'U
0.5240
3:i36o :
.36(10
,64no
.6000
.0
. 712U
. 340U
.0320
.521*
4 U P U
.3100
. I52U
0.1I.OU
I . onnu
0.6400
0.1HM
2.n»iiu
0.2HI6
o.n
1 . f> II 'I U
3.1000
3. 1 36U
2.6240
U . 7UQU
0 IbOU

0.6MOU
1 .472U
a.o
0.0
0.0
o.n
0.0
1.2(100
0.0
0.0
0.0
o.o
O.u
5.SU40
1 .12(10
o.a
i.tono
0.0
2.2400
0.0
0.0
o.o
1.7200
n.o
0.0
1 . 1520
o.n
0.0
0.0
5 . \ 1 1, U
o.o
0.0
o.u
0.0
o.o
0.0
u . u
0.0
1 . 7V2U
n . n
o.o
0.0
o.o
0.0
0.0
u.o
0.0
0.0
0.0
I.2HOO
I'.O
0.0
0.0
v.o
0.0
4.352V
1.3400
0.0
I.2DUO
y ^400
0.0
1.3440
0.0
o.u
0.0
1 .1,000
0.0
u. u
1 . 1520
0.0
o.u
o.v
3 . 20UO
u.u
u.u
u.u
0.0
o.u
u . u
O.u
u.u
1 ,/V20
o.u
o.u
U.I.
U . V
u.u
0.0
0.0
o.o
0.0
0.0
O.H320
o.u
o.o
o.n
O.D
o.u
2.04(IB
0.640U
0.0
0.56)2
0.0
0.64011
u.ll
u.u
0.0
0. 7010
o.a
u.u
0 / 04 II
0.0
o.u
0.0
I . 1,1.40
n.o
0.0
n.o
u.o
o.u
O.u
u.o
II . II
1 .2164
o.u
O II
n.o
o.o
0.0
o.o

-------
                                                         TABLE  E-4   (continued)




nan {MISSIONS
(HG/IHt rilllH
luraini: LIAKH
MINIMUM
IM.AHI
3IIIIOI UAIItnlCS
ALl 1 At. 1 II
• * 1
91 V
1M 0
q<9 0
5" 0
il 0
S2 a
SI 0
bi 0
SS 0
.iiouo l.nooo
.•jiai o.sioo
.2000 0.2000
. lion
.0
.1100
.1600
.0
.SIOO
.1000
.0
.7900
.21100
.0
.4000
.1100 0.1000
•It III
2.7000
0. 1/00
0.1)00
*.OS/0
u.o
O.S100
IALL RAIKHIta
ALl 1 All II
.0 0.0
.0 o.o
.0 o.o
.0 0.0
.2100 0.2100
.0 0.0
0.1100 0.0 0.0
0.0 0.1800 O.JO.OV
0.2100 O.I'OO O.JIOO
o.lloa o.o o.o
Al. 1 III
u.o
0.0
0.0
0.0
0. 1600
0.0
0.0
0,2100
0.1900
0.0
SlIIIHI
ALl 1
».»..
1 . 4S60
1 .2000
o.ev«>o
0.0
S.»5ho
2.1010
0.0
3 .2bQO
1.1160
DAI ItHltJ
ALl II
21.SZOO
J.1S60
1.2800
0.6000
0.0
5.0Sh>
1 . 1920
o.g
2.S600
1 .4200
MAXIMUM

All III
1?
2. li.no
O.K12U
o . 1 1* 4 a •
o.a
.IbbU
.1 120
.0
.ino
.»J20
• ALL
Al. 1 1
n.o
D. n
o.o
O.a
1. J«io
0.0
O.o
1.0720
2.0160
o.a
HAIUHKS
«LI II Al
0.0
(1.0
0.11
o.a
i .mo'
0.0

I III
.0
. 0
. a
. a
. 0210
.0
0.0 0.0
2.1)60
I.VfllO
.1010
.2160
o.a o.a
                         Emissions for dry-coal charged  batteries are  not estimated'and are recorded as 0.0.
 i
!—«
CO

-------
are defined on a radius-azimuth (polar) grid pattern with nonuniform
spacing.  At small radii, the grid cells are much smaller than ED/BG's;
at large radii, the grid cells are much larger than ED/BG's.  The area
surrounding the source is divided into two regions, and each ED/8G 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 0.1 km and 2.8 km from the
emission source, populations are divided between neighboring concen-
tration grid points.   There are 96 (6 x 16) polar grid points within
this range.  Each grid point has a polar sector defined by two con-
centric arcs and two wind direction radials.   Each of these grid
points is assigned to the nearest ED/BG centroid identified from
MED-X.   The population associated with the ED/BG centroid is then
divided among al1 concentration grid points assigned to it.   The exact
land area within each polar sector is considered in the apportionment.
     For population centroids between 2.8 km and 20 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 (usually 1 km in diameter).   Since there is a linear relation-
ship 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 geometrically interpolated radially and arithmetically inter-
polated azimuthally from the four receptors bounding the grid cell.
Concentration estimates for 80 (5 x 16) grid cell receptors  at 2.0,
5.0, 10.0, 15.0, and 20.0 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 96 concentration points
within 2.8 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 portions of the grid,  pairing occurs at the ED/BG cen-
troids through the use of log-log and linear interpolation.   (For a
more detailed discussion of the methodology used to estimate exposures,
see Reference 2.)
                                  E-14

-------
E.4.2  Total Exposure
     Total exposure (persons-(jg/m3) is the sum of the products of
concentration and population, computed as illustrated by the following
equation:
                                        N
                      Total exposure =  I  (P-C.)                  (1)
                                       i=l   ] n
where
     P.  =  population associated with point i,
     C.  =  annual average BSD concentration at point i, and
     N   =  total number of polar grid points between 0 and 2.8 km and
            ED/BG centroids between 2.8 and 20 km.
     The computed total exposure is then used with the unit risk
factor to estimate lung cancer incidence.   This methodology and the
derivation of maximum lifetime risk are described in the following
sections.  (Note:  "Exposure" as used in this appendix is the same as
"dosage" in the computer printout, Docket Number 	.
Table E-5 gives the total  exposure around each plant.
E.5  LUNG CANCER INCIDENCE AND MAXIMUM LIFETIME RISK
E.5.1  Unit Risk Factor
     The unit risk is the  lifetime probability of contracting lung
cancer as a result of continuous exposure to 1 ug/m3 of coke oven
emissions (expressed in units of BSO).   This risk factor was derived
by EPA's CAG.1  The derivation was accomplished by extrapolating the
dose-response relationship of occupational  exposure4 and lung cancer
incidence to the general population using a no-threshold, linear
model.5  The resulting estimate of the lifetime risk of contracting
                       -4
lung cancer is 9.6 x 10   if a person is exposed continuously to 1 |jg
of coke oven emissions (expressed as BSO)  per cubic meter of air.
This represents a plausible upper-limit risk estimate in the sense
that the risk is probably  not higher but could be much lower.
E.5.2  Annual Lung Cancer  Incidence
     Estimates of annual lung cancer incidence (the number of cases
per year) associated with  a given plant is  the product of the total
                                  E-15

-------
        TABLE E-5.  BASELINE DISPERSION MODELING AND EXPOSURE DATA
Maximum annual
average BSD
concentration (ug/m3)
Plant
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Minimum
1.
1.
4.
2.
1.
9.
1.
4.
2.
1.
2.
3.
1.
1.
4.
2.
2.
1.
7.
1.

48 x
09 x
28 x
49 x
21 x
85 x
00 x
99 x
50 x
37 x
50 x
32 x
00 x
18 x
76 x
67 x
50 x
71 x
79 x
00 x
b
10°
10°
10°
10°
10°
ID'1
10°
10°
10'1
10°
ID'1
10'1
10°
10°
10'1
10°
10°
10°
ID'1
10°

Maximum
1.64 x
1.20 x
4.67 x
2.83 x
1.37 x
1.09 x
1.00 x
4.76 x
2.50 x
1.33 x
2.50 x
3.09 x
1.00 x
1.26 x
5.11 x
2.88 x
2.50 x
1.85 x
7.61 x
1.00 x
b
101
101
10 1
101
101
101
10 !
10 1
10°
101
10°
10°
10 !
101
10°
101
101
101
10°
101

Total
population
exposed
519,
115,
433,
91,
580,
552,
52-2,
542,
117,
1,173,
2,186,
2,124,
266,
795,
1,177,
1,104,
594,
1,585,
211,
113,
b
230
367
709
844
947
Oil
359
105
975
737
372
404
581
226
061
291
369
139
515
311

Total exposure
(persons-ug/m3)
• Minimum
2.22
6.59
4.72
1.77
2.73
3.81
5.01
3.01
4.45
2.57
3.09
2.14
4.84
5.78
9.04
7.88
7.24
6.35
9.00
5.43

x 103
x 102
x 103
x 103
x 103
x 103
x 103
x 103
x 101
x 103
x 103
x 103
x 102
x 103
x 102
x 10
x 103
x 103
x 102
x 103
b
Maximum
2.46
7.28
5.14
2.00
3.09
4.21
5.51
2.87
4.15
2.49
3.20
2.00
5.15
6.20
9.70
8.49
7.73
6.88
8.77
5.91

x 104
x 103
x 104
x 104
x 104
x 104
x 104
x 104
x 102
x 104
x 104
x 104
x 103
x 104
x 103
x 104
x 104
x 104
x 103
x 104
b
See notes at end of table.
(continued)
                                  E-16

-------
                           TABLE E-5  (continued)
Maximum annual
average BSO
concentration (|jg/m3)a
Plant
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Minimum
5.
2.
1.
3.
5.
8.
1.
1.
1.
9.
7.
3.
1.

8.
2.
2.
1.
4.
6.
1.
00 x
50 x
57 x
36 x
00 x
38 x
80 x
00 x
96 x
77 x
33 x
81 x
00 x
b
51 x
50 x
43 x
90 x
67 x
03 x
73 x
10" l
10'1
10°
10'1
10'1
10'1
10°
10°
10°
10'1
10'1
10°
10"1

ID'1
10°
10°
10°
10'1
ID'1
10°
Maximum
5.00 x
2.50 x
1.70 x
3.78 x
5.00 x
9.45 x
1.98 x
1.00 x
2.12 x
1.10 x
7.98 x
4.24 x
1.00 x
b
9.10 x
2.50 x
2.64 x
2.06 x
5.16 x
6.62 x
1.90 x
10°
10°
101
10°
10°
10°
101
10 *
10 1
101
10°
101
10°

10°
10 !
101
101
10°
10°
101
Total
population
exposed
2,232
1,702
1,209
909
881
941
153
310
278
87
519
1,409
316

419
1,405
403
478
157
270
1,333
,023
,752
,132
,911
,589
,882
,518
,272
,754
,118
,597
,651
,582
b
,447
,180
,879
,354
,101
,803
,396
Total exposure
(persons-ng/m3)
Minimum
3.92
3.28
5.82
3.95
2.89
2.94
2.29
1.56
4.83
5.07
1.22
1.53
4.54

2.42
1.99
5.93
5.94
1.06
7.32
1.29
x 103
x 103
x 103
x 102
x 103
x 103
x 103
x 103
x 103
x 102
x 103
x 104
x 102
b
x 103
x 104
x 103
x 103
x 103
x 102
x 104
Maximum
4.14
3.51
6.28
4.43
3.28
3.31
2.25
1.74
5.23
5.71
1.33
1.70
5.08

2.58
2.21
6.47
6.42
1.17
1.17
1.41
x 104
x 104
x 104
x 103
x 104
x 104
x 104
x 104
x 104
x 103
x 104
x 10s
x 103
b
x 104
x 10s
x 104
x 104
x 104
x 104
x 10s
See notes at end of table.
(continued)
                                  E-17

-------
                           TABLE E-5  (continued)
Maximum annual
average BSO
concentration (ug/m3)
Plant
43
44
45
46
47
48
49
50
51
52
53
54
55
Minimum
4.
3.
5.
5.
1.
5.
3.
5.
5.
6.
5.
2.
4.
73 x
25 x
00 x
66 x
45 x
18 x
38 x
00 x
00 x
14 x
02 x
50 x
16 x
10'1
10'1
10'1
10°
10°
10'1
10'1
10-2
ID'1
10'1
10'1
10"1
10'1
Maximum
4.94 x
3.50 x
5.00 x
5.92 x
1.56 x
5.69 x
3.45 x
5.00 x
5.00 x
6.74 x
5.26 x
2.50 x
4.53 x
10°
10°
10°
101
101
10°
10°
ID'1
10°
10°
10°
10°
10°
Total
population
exposed
216
2,813
1,066
999
815
243
305
19
152
151
151
150
1,076
,118
,430
,334
,120
,270
,974
,804
,574
,264
,637
,637
,580
,337
Total exposure
(persons-|jg/m3)a
Minimum
1.01
2.74
1.62
1.28
4.06
6.64
5.00
1.71
3.60
4.89
4.47
1.21
3.09
x 103
x 103
x 103
x 104
x 103
x 102
x 102
x 101
x 102
x 102
x 102
x 103
x 103
Maximum
1.06
2.95
1.71
1.34
4.35
7.29
5.11
9.31
3.79
5.37
4.68
1.30
3.36
x 104
x 104
x 104
x 10s
x 104
x 103
x 103
x 101
x 103
x 103
x 103
x 104
x 104
 Maximum and  minimum figures  correspond  to  maximum and minimum estimates of
 emissions  of BSO  from the  three  emission  sources  (see Tables  E-2 through
 E-4).

^Emissions  from  dry-coal  charged  batteries  are  not estimated  and  are
 recorded as  0.0.
                                  E-18

-------
exposure around that plant (persons-|jg/m3) and the unit risk factor
               -4
(URF), 9.6 x 10  , divided by 70 years (average life expectancy).
Thus,
   Lung cancer cases per year = (total exposure)
                                x (unit risk factor)/70
(2)
where total exposure is calculated according to Equation 1.   Table E-6
gives the estimated number of lung cancer cases per year at the base-
line level of emissions (Regulatory Alternative 1) from all  plants.
E.5.3  Lifetime Risk
     The populations in areas surrounding coke plants have various
risk levels of lung cancer incidence from exposure to BSD emissions.
Using the annual average concentration of BSO to which any persons are
exposed, it is possible to calculate the lifetime risk of lung cancer
attributable to BSO emissions using the following equation:
                      Lifetime risk = Ci x (URF)                   (3)

where
     C.  =  the concentration at any receptor location where exposed
            persons reside, and
     URF =  the unit risk factor, 9.6 x i(f4 .
Table E-7 gives the lifetime risk estimates for the baseline level of
emissions from all plants.   The maximum lifetime risks are estimated
to occur at the U.S. Steel  plant in Clairton, Pennsylvania.   The
maximum lifetime risk estimates at this plant range from 5.4 x 10   to
        -2
5.7 x 10  , based on the low and high estimates of BSO emissions from
this plant.
E.6  UNCERTAINTIES
     Estimates of both lung cancer incidence and maximum lifetime risk
are functions of the estimated BSO concentrations, estimated popula-
tions,  and the unit risk factor.   The calculations of these  variables
are subject to a number of uncertainties of various degrees.   Some of
the major uncertainties are identified below.
                                  E-19

-------
      TABLE E-6.   ESTIMATED NUMBER OF LUNG CANCER CASES
                          PER YEAR9
Emission
source
Charging
Door leaks
Topside leaks
Total
Regulatory Alternative I
Minimum
6.2 x 10"3
2.0
0.6
2.6
Maximum
1.7
22.9
3.5
28.1
aFor the baseline level of emissions from all plants.

 Maximum and minimum figures correspond to maximum and
 minimum estimates of emissions of BSO from the three
 emission sources (see Tables E-2 through E-4).
                             E-20

-------
    TABLE E-7.   POPULATION DISTRIBUTION OF LIFETIME RISK ESTIMATES
              FROM ALL SOURCES,  REGULATORY ALTERNATIVE I
Lifetime risk level
Greater than 10"2
-2 -3
10-10
io"3 - io"4
io"4 - io"5
io"5 - io"6
Less than 10
Minimum
Population
at risk
0
1,900
165,000
2,975,000
21,000,000
14,209,000
b
Percent
0
c
0.43
7.76
54.76
37.05
Maximum
Population
at risk
2,300
187,000
3,273,000
21,370,000
13,446,000
73,000
b
Percent
c
0.49
8.53
55.73
35.06
0.19
aRisks were calculated on a plant-by-plant basis and summed.   Thus,
 persons exposed to emissions from more than one coke plant were
 counted for each plant's impact.   Therefore, the total  number of
 people actually exposed is overestimated, while the levels of risk
 are underestimated (because the additive risk impact of several
 plants could not be analyzed).

 Maximum and minimum figures correspond to maximum and minimum
 estimates of emissions of BSO from the three emission sources
 (see Tables E-2 through E-4).
cThe percentage is less than 0.01 percent.
                                  E-21

-------
E.6.1  BSO Concentrations
     Modeled ambient BSO concentrations depend upon (1) plant config-
uration, which is difficult to determine for more than a few plants;
(2) emission point characteristics, which can be different from plant
to plant and are difficult to obtain for more than a few plants;
(3) emission rates, which may vary over time and from plant to plant;
and (4) meteorology, which is seldom available for a specific plant.
The particular dispersion modeling used can also influence the num-
bers.   The dispersion coefficients used in modeling are based on
empirical measurements made within 10 km of sources.   These coefficients
become less applicable at long distances from the source, and the
modeling results become more uncertain.   The best model to use (ISC-LT)
is usually too resource intensive for modeling a large number of
sources.  Less complex models introduce further uncertainty through a
greater number of generalizing assumptions.   The HEM assumes that the
terrain in the vicinity of the source is flat.   For sources located in
complex terrain, the maximum annual concentration could be under-
estimated by several fold due to this assumption.   Assuming the inputs
to the dispersion model are accurate, the predicted BSO concentrations
are considered to be accurate to within a factor of 2.6  This uncer-
tainty factor was not included in the calculations in this analysis.
E.6.2  Exposed Populations
     Several simplifying assumptions were made with respect to the
assumed exposed population.   The number of people was assumed to
remain constant over time.  In addition, those that are exposed are
assumed to remain at the same location 24 hours per day, 365 days per
year,  for a lifetime (70 years) and are assumed to be exposed to a
constant BSO concentration over time.  The assumption that exposed
populations remain at the same location is counterbalanced to some
extent (at least in the calculation of incidence) by the assumption
that no one moves into the exposure area either permanently as a
resident or temporarily as a transient.   The population "at risk" was
assumed to reside within 20 km of each plant.   The selection of 20 km
is considered to be a practical modeling stop-point considering the
                                  E-22

-------
uncertainty of dispersion estimates beyond 10 km.   The results of
dispersion modeling are felt to be reasonably accurate within that
distance (see above).   The uncertainty of these assumptions has not
been quantified.
E.6.3  Unit Risk Factor
     Discussions of the limitations of the unit risk factor are con-
tained in the CAG coke oven carcinogen assessment document.1
E.6.4  Other Uncertainties
     There are several uncertainties associated with estimating health
impacts.   Maximum lifetime risk and annual lung cancer incidence were
calculated using the unit risk factor, which is based on a no-threshold
linear extrapolation of lung cancer risk and applies to a presumably
healthy, predominantly nonwhite male cohort of workers exposed to BSO
concentrations much higher than community levels.   It is uncertain
whether the unit risk factor can be accurately applied to the general
population, which includes men, women, children, whites, the aged, and
the unhealthy, who are exposed to concentrations much lower than the
occupational exposure levels.   It is uncertain whether these widely
diverse segments of the population may have susceptibilities to lung
cancer that differ from those of workers in the studies.  Furthermore,
the use of lung cancer incidence as the disease of interest in this
appendix does not address all  of the health impacts from coke oven
emissions.  These emissions may be related to other malignant and
nonmalignant causes of death and disability.   Redmond (1975) reported
a statistically significant risk for nonmalignant respiratory disease
mortality among white and nonwhite coke oven workers.7  Furthermore,
mortality from other sites showed statistically significant (P<0.05)
excesses.   These sites include prostate, kidney, and all malignant
neoplasms.7 8  Excess mortality from pancreatic cancer and cancers of
the large intestine and the pharyngeal and buccal  cavities were reported
among nonoven workers employed in the coke plant.8  In addition,
pollution from coke oven emissions in the presence of an air inversion
may be related to episodes of asthma, bronchitis,  emphysema, or other
acute cardiorespiratory conditions in both children and adults.9
Compared to the incidence of all other conditions discussed above,

                                  E-23

-------
               EPA-RTP LIBRARY


deaths from lung cancer are relatively uncommon events.   Clearly,  lung

cancer mostly affects individuals who are more than  50 years  old.   In

contrast, the other conditions may affect infants, children,  women  of
child-bearing age, and patients with existing chronic  illnesses.   In

spite of these limitations, the estimates of lung cancer  incidence

appear to be as rigorous as possible based on the data presented.

E.8  REFERENCES

1.    Letter from McGaughy,  R.  E., Carcinogen Assessment Group,  to
     Griffin, H.,  Science Advisory Board.   May 27, 1983.

2.    Systems Applications,  Inc.   Human Exposure to Atmospheric  Con-
     centrations of Selected Chemicals.   (Prepared for the  U.S.
     Environmental Protection Agency, Research Triangle Park, North
     Carolina).  Volume I,  Publication Number EPA-2/250-1.

3.    Busse, A.D.,  and J.  R. Zimmerman.   User's Guide for  the  Climato-
     logical Dispersion Model.   (Prepared for the U.S. Environmental
     Protection Agency, Research Triangle Park,  North Carolina.)
     Publication Number EPA-R4-73-024.   December 1973.

4.    Mazumdar, S., C.  Redmond,  W.  Sollecito,  and N.  Nussman.  An
     Epidemiologic Study of Exposure to Coal  Tar Pitch Volatiles Among
     Coke Oven Workers. J.  Air Pol.  Cont.  Assn.   25(4):382-389.  1975.

5.    Telecon, Goldsmith,  David,  Research Triangle Institute,  to Thors-
     land, Todd,  Carcinogen Assessment Group,  EPA.   May 6,  1980.

6.    Albert, R. E.  Carcinogen Assessment Group's Final Report  on
     Population Risk to Ambient BSD  Exposures.   U.S. Environmental
     Protection Agency.  Publication No.  EPA-450/5-80-004.  January
     1979.

7.    Redmond, C.   Comparative Cause-Specific Mortality Patterns by
     Work Area Within the Steel  Industry.   U.S.  Department  of Health,
     Education, and Welfare, PHS, CDC,  NIOSH.   HEW Publ.  No.  (NIOSH)
     75-157.  April 1975.   p.  35-54.   p.  144-150.

8.    Redmond, C.  K.,  H. S.  Wieand,  H.  E.  Rockette, et al.   Long-Term
     Mortality Experience of Steelworkers.   U.S.  Department of  Health,
     Education, and Welfare, PHS, CDC,  NIOSH.   Cincinnati,  Ohio.  June
     1979.

9.    Holland, W.  W.,  A. E.  Bennett,  I.  R.  Cameron, et al.   Health
     Effects of Particulate Pollution:   Reappraising the  Evidence.
     Am.  J. Epidemiol.   110(5):527-659.   1979.
                                  E-24

-------