EPA-340/1-76-012
May,1977
Stationary  Source
Enforcement Series
              OF  A
              STATIONARY
              COKE-SIDE
              BURNS  HARBOR  PLANT
              BETHLEHEM  STEEL CORPORATION
              CHESTERTON, INDIANA
           U.S. ENVIRONMENTAL  PROTECTION AGENCY
           OFFICE  OF ENFORCEMENT
           OFFICE  OF GENERAL ENFORCEMENT
           WASHINGTON,D.C. 20460

-------
    g-  UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                      WASHINGTON, D.C. 20460

                         6 APR 1977.
                                               OFFICE OF ENFORCEMENT
MEMORANDUM
Subject:     Transmittal of Volume I of DSSE's Report  "Source
             Testing of a Stationary Coke-Side Enclosure"

To:          See Distribution
    Attached is a copy of the subject report.  This volume
contains all of the discussion, conclusions and  summaries of
all of the results of our extensive  testing program at  the
Burns Harbor coke-side shed  (Bethlehem Steel Corporation,
Chesterton, Indiana), March  3-7, 1975. The remaining  eleven
volumes of the report contain specific test methodology,
test and process data.  Because they are  so voluminous, we
have limited their distribution to ESED and DSSE.

    Do not hesitate to contact Louis Paley (202-755-8137)
of my staff for any additional information.
                  Stanley/^?. Legj

Attachment

cc:  State and Local Agencies
     (See Attached List)

-------
                          -2-


                     DISTRIBUTION
EPA Libraries
NTIS
EPA Regional Contacts-
   Howard Bergman
   Dave Brooman
   Ken Eng
   Lew Felleisen
   Donald Goodwin
   Lois Green
   Geoff Grubbs
   Robert Hendricks
   John Hepola
   Mark Hooper
   Reid Iverson
   Larry Jones
   Dave Kee
   Pete Kelly
   Larry Kertcher
   Fred Longenberger
   Lee Marshall
   Gary McCutchen
   Bruce Miller
   Walter Mugden
   Roy Neulicht
   Gary Parrish
   Norman Plaks
   Steve Rothblatt
   Ben Stonelake
   Andrew Trenholm
   David Ullrich
   Lance Vinson
   Thomas Voltaggio
   Richard Watman
   James Wilburn
   Gale Wright
   Gary Young

-------
                       —3 —

                   ATTACHED LIST

Director, New York - Department of Environmental
          Conservation
Director, Pennsylvania - Department of Environmental
          Resources
Director, Pennsylvania - Allegheny County Bureau
          of Air Pollution
Director, Maryland - State Department of Health and
          Mental Hygiene
Director, West Virginia - Air Pollution Control Commission
Director, Alabama - State Department of Public Health
Director, Alabama - Jefferson County Health Department
Director, Kentucky - Air Pollution Control Commission
Director, Tennessee - Chattanooga - Hamilton County
          Air Pollution Control Bureau
Director, Illinois - Environmental Protection Agency
Director, Illinois - Chicago"- Department of Environmental
          Control
Director, Indiana - State Board of Health
Director, Indiana - Department of Air Qualty Control
Director, Michigan - Department of Natural Resources
Director, Michigan - Wayne County Health Department
Director, Minnesota - Pollution Control Agency
Director, Ohio - Environmental Protection Agency
Director, Ohio - Department of Public Health and Welfare
Director, Ohio - Air Pollution Control Division
Director, Ohio - Pollution Control Agency
Director, Ohio - Portsmith City Health District
Director, Ohio - Mahoning-Trumbull Air Pollution Agency
Director, Wisconsin - Department of Natural Resources
Director, Texas - State Department of Health
Director, Texas - Department of Public Health
Director, Missouri - Air Conservation Commission
Director, Missouri - St. Louis - Division of Air Pollution
          Control
Director, Colorado - Department of Health
Director, Colorado - Attorney General's Office
Director, Utah - State Division of Health
Director, California - Air Resources Board
Director, California - San Bernardino County Air Pollution
          Control District

-------
SOURCE TESTING OF A STATIONARY COKE-SIDE ENCLOSURE

                 (Volume 1 of 12)
                Burns Harbor Plant
            Bethlehem Steel  Corporation
                Chesterton,  Indiana

              Contract  No. 68-02-1408
                   Task  No.  10
                   Prepared  for:

    Division of  Stationary  Source  Enforcement
             Technical Support Branch
       U.S.  Environmental Protection Agency
             Washington, D.C.   20460
                 Project Officer:

               Louis R. Paley, P.E.




                   May 20, 1977




                   Prepared by:

      Clayton Environmental Consultants, Inc.
               25711  Southfield Road
           Southfield, Michigan   48075

-------
This report was furnished to the U.S. Environmental Protection
Agency by Clayton Environmental Consultants, Inc., Southfield,
Michigan, in fulfillment of Contract No. 68-02-1408, Task Order
No. 10.  The contents of this report are reproduced herein as
received from the contractor.  The opinions, findings, and con-
clusions expressed are those of the authors and not necessarily
those of the U.S. Environmental Protection Agency.
The Enforcement Technical Guideline series of reports is issued
by the Office of Enforcement, Environmental Protection Agency,
to assist the Regional Offices in activities related to enforcement
of implementation plans, new source emission standards, and haz-
ardous emission standards to be developed under the Clean Air Act.
Copies of Enforcement Technical Guideline reports are available -
as supplies permit - from Air Pollution Technical Information
Center, Environmental Protection Agency, Research Triangle Park,
North Carolina, 27711, or may be obtained,for a nominal cost,
from the National Technical Information Service, 5285 Port Royal
Road, Springfield, Virginia, 22161.

-------
                        ACKNOWLEDGEMENTS
This report was prepared under the direction of Mr. John Mutchler
with the assistance of principal authors Thomas Loch, Richard
Powals, and Janet Vecchio of Clayton Environmental Consultants,
Inc.  The Project Officer for the U.S. Environmental Protection
Agency was Mr. Louis Paley.  The authors are grateful to Mr.
Paley for his recommendations, comments, and review throughout
the execution and report development phases of the study.  The
authors also appreciate the valuable contributions of Mark Antell,
Bernard Bloom, and Kirk Foster (Division of Stationary Source En-
forcement) to this study.  Finally, the assistance of the addi-
tional following people at the field study site is very gratefully
acknowledged.
              U.S. EPA

     Dave Brooman (Region VIII)

     Don Carey (D.S.S.E.)

     Stanley Couer (Audio Visual)

     Paul DePercin (Region V)

     Basim Dihu (Region V)

     Steve Florin (Region V)

     Joseph Kunz (Region III)

     Ron Mitchell (Audio Visual)

     Dave Shulz (Region V)

     R. Edwin Zylstra (Region V)
Bethlehem Steel Corporation

C. A. Trageser

Robert Harvey

Norm Hodgson

Tom Kreichett

John Dunn

Carolyn Mance

Ron Spalding

Dave Fisher

Gerald Marchant

-------
               TABLE     OF    CONTENTS


VOLUME 1                                                    Page

GLOSSARY OF TERMS 	   V

LIST OF FIGURES 	 viii

LIST OF TABLES 	  ix

LIST OF APPENDICES 	  xi


1.0  INTRODUCTION	    1


2.0  SUMMARY AND CONCLUSIONS	  12

     2.1  Coke-Side Particulate Emissions 	  12

          2.1.1  Overall continuous coke-side particulate
                 emissions 	  12
          2.1.2  Continuous particulate emissions from the
                 exhaust duct	  13
          2.1,3  Peak particulate emissions from  the ex-
                 haust duct	  13
          2.1.4  Fugitive particulate emissions from the
                 shed	  14
          2.1.5  Particulate emissions for pushing opera-
                 tions 	  14
          2.1.6  Particulate emissions for non-pushing
                 operations 	  14

     2.2  Shed Particulate Capture Efficiency 	  15

          2.2.1  Evaluation of shed capture efficiency ....  15
          2.2.2  Possible causes of leakage	  15

     2.3  Chemical Composition of Particulate Emissions ...  16

     2.4  Particle Size Distribution 	  16

     2.5  Emissions of Other Materials 	  16

     2.6  Dustfall Measurements	  16

     2.7  Indices of Visible Emissions 	  17

          2.7.1  Degree of greenness 	  17
          2.7.2  Opacity	  17

-------
                               ii
                                                            Pa^ge


           2.7.3  Percent of doors leaking 	  18

      2.8  Process and Emissions Correlations 	  18

      2.9  Representativeness of Process and Shed
           Conditions 	  19


3.0  PROCESS DESCRIPTION AND OPERATIONS 	  20

      3.1  Process Description 	  20

      3.2  Representativeness of Process and Shed Con-
           ditions 	  25

           3.2.1  Criteria for comparison 	  25
           3.2.2  Conditions during sampling periods  	  28

      3.3  Identification of Possible Normalizing Factors..  32


4.0  SAMPLING AND ANALYTICAL METHODS 	  33

      4.1  Test Protocol 	  33

      4.2  Location of Sampling Points 	  34

      4.3  Continuous Particulate Emissions from Shed
           Exhaust Duct 	  35

      4.4  Determination of Peak Particulate Emission Period 39

      4.5  Peak Particulate Emissions from Shed Exhaust Duct 43

      4.6  Particle Size Distribution	  43

      4.7  Emissions of Other Materials 	  44

      4.8  Dustfall Measurements 	  45

      4.9  Subjective and Visual Emission Parameters  	  46

           4.9.1  Degree of greenness 	  46
           4.9.2  Opacity of shed exhaust 	  47
           4.9.3  Percent of doors leaking	  47
           4.9.4  Visual estimates of fugitive emissions...  48

      4.10 Fugitive Particulate Emissions From the Shed ...  48

      4.11 Calibrations, Quality Assurance, and Sampling
           Integrity 	  49

-------
                               iii


                                                            Page

5.0  PRESENTATION AND DISCUSSION OF RESULTS 	  51

     5.1  Coke-Side Particulate Emissions 	  51

          5.1.1  Continuous particulate emissions from the
                 exhaust duct 	  51
          5.1.2  Fugitive particulate emissions 	  53
          5.1.3  Overall continuous coke-side particulate
                 emissions	  54
          5.1.4  Peak particulate emissions from the shed .  56
          5.1.5  Particulate emissions for pushing opera-
                 tions	  59
          5.1.6  Particulate  emissions for non-pushing
                 operations 	  62

     5.2  Particulate Capture Efficiency of the Shed 	  65

          5.2.1  Evaluation of shed capture efficiency ....  65
          5.2.2  Possible causes of leakage 	  65

     5.3  Chemical Composition of Particulate Emissions ...  69

     5.4  Particle Size Distribution	  69

     5.5  Emissions of Other Materials 	  77

     5.6  Indices of Visible Emissions 	  80

          5.6.1  Degree of greenness	  80
          5.6.2  Opacity 	  88

                 5.6.2.1  Emissions from exhaust duct 	  88
                 5.6.2.2  Fugitive emissions 	  89

          5.6.3 , Percent of doors leaking 	  90

     5.7  Emission-Related Correlations 	  92

          5.7.1  Correlations between emission factors and
                 indices of visible emissions 	  92
          5.7.2  Correlations between emission factors and
                 process conditions 	  93
          5.7.3  Correlations involving particle size
                 distributions	  94
          5.7.4  Correlations between indices of visible
                 emissions and process conditions	  94
          5.7.5  Correlations among visible emissions
                 measurements ..	 101

     5.8  Effect of the Shed Upon Dustfall 	 103

-------
                                iv


                                                             Page

      5.9   Impact  of the  Shed Upon Airborne Agents Within
           the  Shed	 115

      5.10  Precision of Test  Results	 115


6.0   REFERENCES  	 117
 7.0   SOME  ANTICIPATED  QUESTIONS AND ANSWERS RELATED TO THIS
      PROJECT	 118
VOLUME  2

     Appendices  A-E  (See  page  xi  for titles)

VOLUME  3

     Appendices  F-H  (See  page  xi  for titles)

VOLUME  4

     Appendices  1-0  (See  page  xi  for titles

VOLUME  5

     Appendices  P-MM  (See  pages xi  and  xii for titles)

VOLUME  6

     Appendices  NN-WW  (See  page xiii for  titles)

VOLUME  7

     Appendices  XX-ZZ  (See  page xiii for  titles)

VOLUME  8

     Appendix AAA  (See page xiii  for ti.tles)

VOLUME  9

     Appendix BBB  (See page xiii  for titles)

VOLUME  10

     Appendix CCC  (See page xiii  for titles)

VOLUME  11

     Appendix DDD  (See page xiii  for titles)

VOLUME  12

     Appendices  EEE-GGG  (See page xiv for titles)

-------
                        GLOSSARY OF TERMS
1.  Abnormal operating conditions

    Net coking time outside the normal range of net coking time
    or any coke pushing stoppage greater than 30 minutes duration.

2.  Atypical operating conditions

    Extremely infrequent major process changes (or upsets).

3.  Coke-pushing emissions

    An intermittent source emission lasting about 15 to 45 seconds,
    occurring on an irregular cycle with an average interval be-
    tween pushes of 13 minutes.

4,  Coke-pushing operations emissions

    The aggregate of two source emissions:  1) coke-pushing emis-
    sions and 2) quench car movement emissions which occur under
    the shed.

5.  Coke side

    That side of a coke-oven battery from which the ovens are
    emptied of coke.

6.  Continuous particulate emissions

    The mass particulate emissions measured on the coke side of
    the coke battery on a continuous basis, spanning periods when
    pushes occurred as well as intervals between pushes (unless
    process upsets or downtime exceeded 30 minutes).

7.  Degree of greenness of a coke-oven push

    A subjective, visual estimate of the quantity of emissions
    released during a single coke-oven push by estimation of the
    plume obscuration immediately above the quench car.

8.  Door leakage

    Any visible emissions observed emanating from coke-side oven
    door^, push-side oven doors,  or push-side chuck doors.

9.  Filterable particulate

    Material captured at a specified temperature,  pressure, and
    chemical activity, on or before the front filter in a partic-
    ulate sampling train.

-------
                               vi
10.  Fugitive particulate emissions
      1 ~ n '        '   —^••—i -i. L r i   i  n 	 _ i -      ^

     Farticulate emissions which escaped capture by the shed and
     passed unrestrained into the atmosphere.  This does not in-
     clude emissions resulting from quench car travel outside the
     shed.

11.  Minimum coking time

     The elapsed time, in minutes, specified by the operator of
     the coke production facility as being the minimum net coking
     time necessary to provide adequate quality coke for produc-
     tion purposes.

12.  Net coking time

     The elapsed time, in minutes, between the charging of a coke
     oven with coal and the pushing of that same oven.

13.  Normal operating conditions

     Any typical operating conditions not abnormal.

14.  Normalization factor

     A variable used to relate a mass emission rate to the rate of
     processing.  An example is "tons of dry coal charged."

15.  Overall continuous coke-side particulate emissions

     The sum of the continuous particulate emissions and the con-
     tinuous fugitive particulate emissions.

16.  Peak particulate emissions

     The mass particulate emissions from the exhaust duct measured
     on the coke side of the coke battery during only the initial
     3-minute periods beginning with the commencement of each coke-
     oven push.

17.  Precision of a test result

     The statistical confidence interval associated with the mean
     value of a series of replicate measurements at a decision-
     risk level of five percent.

18.  Push-only particulate emissions

     The mass particulate emissions measured on the coke side of
     the coke battery and resulting only from the pushing opera-
     tions .

-------
                               vii
19.  Quench car movement emissions

     An intermittent source emission emanating from the coke in
     the quench car and lasting about 15-45 seconds, from the end
     of a coke-oven push until the quench car exits from the shed,

20.  Total particulate

     Material captured at a specified temperature, pressure, and
     chemical activity in the entire particulate sampling train,
     i.e., filterable and condensible fractions.

21.  Typical operating conditions

     Any process operating conditions not atypical.

-------
                              viii
FIGURE 3.1-1



FIGURE 3.1-2

FIGURE 3.1-3


FIGURE 4.2-1


FIGURE 4.2-2


FIGURE 4.2-3


FIGURE 4.4-1


FIGURE 4.4-2


FIGURE 5.1.5

FIGURE 5.4


FIGURE 5.7.4-1


FIGURE 5.7.4-2


FIGURE 5.7.4-3


FIGURE 5.7.4-4


FIGURE 5.7.5-1


FIGURE 5.7.5-2


FIGURE 5.7.5-3
     LIST OF FIGURES
                                            Page

Schematic Diagram of By-Product, Metal-      21
lurgical Coke Manufacturing Facility
with Dustfall Sites

Schematic Diagram of Coke-Side  Shed          23

Schematic Diagram of Coke-Side  Shed End      24
Openings

Schematic Diagram of Sampling Point Loca-    36
tions for "Continuous" Particulate Samples

Schematic Diagram of' Sampling Point Loca-    37
tions for "Peak" Particulate Samples

Schematic Diagram of Sampling Point Loca-    38
tions for  All Samples Except Particulate

Sequential Filter Obscurity Test on          40
February 24, 1975

Sequential Filter Obscurity Test on          41
March 3, 1975

Schematic Diagram of Sampling Schedule       61

Particle Size Distributions in  Exhaust       74
Duct

Net Coking Time Versus Opacity  for Partic-   96
ulate Sampling Days

Degree of Greenness Versus Net  Coking Time   98
for Particulate Sampling Days

Opacity Versus Flue Temperature for Partic-  99
ulate Sampling Days

Degree of Greenness Versus Flue Temperature  100
for Particulate Sampling Days

Opacity Versus Degree of Greenness for Par-  102
ticulate Sampling Days

Composite Graph of Shed Exhaust Duct Opac-   104
ity Versus Time
Shed Exhaust Duct Opacity Versus Time for
Various Net Coking Times
105

-------
                               ix
TABLE 1.0-1
TABLE

TABLE

TABLE

TABLE
1.0-2

1.0-3

3.2.1

3.2.2-1
TABLE 3.2.2-2


TABLE 4.4


TABLE 5.1.1


TABLE 5.1.2


TABLE 5.1.3


TABLE 5.1.4


TABLE 5.1.5


TABLE 5.1.6


TABLE 5.3-1


TABLE 5.3-2


TABLE 5.3-3


TABLE 5.4


TABLE 5.5
       LIST OF TABLES
                                           Page

Generalized Roster of Emissions              3
Investigations

Purpose(s) for Sampling Each Contaminant     5

Project Participants                        10

Pertinent Parameters                        27

Comparison of Key Process Parameters        29
(Battery No. 1)

Comparison of Key Process Parameters        30
(Battery No. 2)

Determination of Peak Particulate Emission  42
Period From the Exhaust Duct
                            .•*''.
Summary of Continuous Particulate Emissions 52
From the Battery No. 1 Exhaust Duct

Measured Fugitive Particulate Emissions     55
Escaping From the Shed

Summary of Overall Continuous Particulate   57
Emissions From the Shed

Summary of Peak Particulate Emissions From  58
the Battery No. 1 Exhaust Duct

Calculation of Filterable Particulate       63
Emission Factor for Pushing Operations

Calculation of Filterable Particulate       64
Emission Factor for Non-Pushing Operations

Summary of Metals and Sulfate Content of    70
Particulate Samples

Summary of Average Rates of Particulate     71
Emissions From Exhaust Duct

Summary of Water Soluble pH and Acidity/    73
Alkalinity of Particulate Samples

Particulate Concentration and Acetone-      76
Soluble Content of Particle Size Samples

Summary of Average Emission Rates of        78
"Other" Emissions

-------
TABLE 5.6.1-1



TABLE 5.6.1-2



TABLE 5.6.1-3



TABLE 5.6.1-4



TABLE 5.6.1-5



TABLE 5.6.1-6



TABLE 5.6.1-7


TABLE 5.6.3

TABLE 5.8-1


TABLE 5.8-2



TABLE 5.8-3

TABLE 5.8-4
Characteristics of Individual Pushes
During Particulate Sampling — Continuous
Particulate Test No. 1
                                           Page

                                            81
Characteristics of Individual Pushes During 82
Particulate Sampling—Peak Particulate
Test No. 1

Characteristics of Individual Pushes During 83
Particulate Sampling—Continuous
Particulate Test No. 2

Characteristics of Individual Pushes During 84
Particulate Sampling—Peak Particulate
Test No. 2

Characteristics of Individual Pushes During 85
Particulate Sampling—Continuous
Particulate Test No. 3

Characteristics of Individual Pushes        86
During Particulate Sampling — Peak
Particulate Test No. 3

Push Characteristics During Particle Size   87
Sampling

Door Leakage on Particulate Sampling Days   91

                                           107
Summary of Dustfall Measurements at
Batteries 1 and 2

Summary of Acetone-Soluble and
Cyclohexane-Soluble Content of Selected
Dustfall Samples
                                           109
Summary of pH of Selected Dustfall Samples 110
Format Used for Analyses of Dustfall
Data (gm/m2/wk)
                                           113

-------
                               xi

                       LIST OF APPENDICES


VOLUME 2

A    BSC Coke-Oven Pushing Schedule

B    BSC Pusher Reports

C    BSC Coke-Oven Daily Data (Coal and Coke Analyses)

D    BSC Coal Charge Rate and Coal Analyses

E    BSC Coke-Oven Fuel-Gas Analyses


VOLUME 3

F    BSC Coal Weights for Each Charge

G    BSC Cross-Wall Temperature Graphs

H    BSC Full-Span Pyrometer Data


VOLUME 4

I    BSC Oven Pressure and Temperature Data

J    BSC Collector-Main Pressure and Underfire Gas Flow  Data

K    BSC Flue Inspection Reports

L    BSC Coke Inspection Reports

M    BSC Fan Curves and Fan Power Data

N    BSC Visible Emissions Ratings

0    BSC Meteorological Data


VOLUME 5

P    EPA Method 1 — Sample and Velocity Traverses for Stationary
     Sources

Q    EPA Method 2 — Determination of Stack Gas Velocity  and Volu-
     metric Flowrate (Type S Pitot Tube)

R    EPA Method 5 — Determination of Particulate Emissions from
     Stationary Sources

S    Determination of Particulate Emissions from Coke-Oven Pushing

-------
                               xli
T    Determination of Peak Particulate Emission  Period  from  ah
     Exhaust Duct

U    Determination of Particle Sizing During  Coke-Oven  Pushing

V    EPA Method 8 — Determination of Sulfuric Acid Mist  and
     Sulfur Dioxide Emissions from Stationary Sources

W    EPA Method 11 — Determination of Hydrogen Sulfide  Emissions
     from Stationary Sources

X    Determination of Ammonia Emissions  from  Coke-Oven  Pushing

Y    Determination of Various Emissions  Absorbed  in  Sodium
     Hydroxide from Coke-Oven Pushing

Z    Determination of Various Emissions  Adsorbed  on  Activated
     Carbon from Coke-Oven Pushing

AA   Determination of Various Emissions  Captured  in  a Glass  Gas
     Burette During Coke-Oven Pushing

BB   Determination of Various Emissions  Absorbed  in  Cyclohexane
     from Coke-Oven Pushing

CC   Determination of Dustfall (Particulate Fallout) Near Coke
     Ovens

DD   Determination of Degree of Greenness of a Coke-Oven Push

EE   EPA Method 9 — Visual Determination of the  Opacity of Emis-
     sions from Stationary Sources

FF   Determination of the Percent of Doors Leaking

GG   Determination of Estimated Fugitive Emissions from Coke-Oven
     Pushing

HH   Assessment of the Fugitive Particulate Emissions Escaping
     from a Coke-Side Shed

II   Calibration Procedures

JJ   Calibration Data

KK   Void Samples

LL   Field Sampling Data

MM   Analytical Data

-------
                              xiii
VOLUME 6

 NN  Chain-of-Custody Procedures for Air Pollution Emission
     Sampling

 00  Example Calculations

 PP  Correspondence from Bethlehem Steel Corporation

 QQ  Results of Previous BSC Sampling

 RR  Results of Continuous Particulate Sampling

 SS  Results of Peak Particulate Sampling

 TT  Results of Particle Size Sampling

 UU  Results of Sampling for "Other" Emissions

 VV  EPA Degree-of-Greenness Data

 WW  EPA Evaluation of Visible Emissions Exhausted from the Shed


VOLUME 7
 XX  EPA Evaluation of Visible Emissions for Shed End Leakage

 YY  EPA Door Leak Observations

 ZZ  Push Emission Charts — Battery No. 1


VOLUME 8

AAA  Record for Coke-Side and Push-Side Doors — Battery No.  1


VOLUME 9

BBB  Shed Capture Performance and Miscellaneous Data—Battery
     No. 1


VOLUME 10

CCC  Push Emission Charts—Battery No. 2


VOLUME 11

ODD  Record for Coke-Side and Push-Side Doors — Battery No.  2

-------
                              xiv
VOLUME 12




EEE  Preliminary Tabulations Forwarded to EPA




FFF  Additional Attempted Correlations




GGG  Results of Dustfall Sampling

-------
                         1.0  INTRODUCTION



     The U.S. Environmental Protection Agency commissioned Clayton




Environmental Consultants, Inc. (Task 10, Contract No. 68-02-1408)




to quantify the nature and extent of particulate and gaseous emis-




sions typically emanating from the coke side of Coke Battery No. 1




at the Burns Harbor plant of Bethlehem Steel Corporation in




Chesterton, Indiana.  This information was obtained to help provide




a basis for:




     1.  Development of EPA policy on coke-side coke battery emis-




         sions and their control;




     2.  Assessment of the adequacy of State Implementation Plans




         to achieve Primary Air Quality Standards in areas contig-




         uous to coke plants; and




     3.  Assessment of the adequacy of control devices being pro-




         posed for abatement of such emissions.




     Measurement of the normally fugitive coke-side emissions was




facilitated at Burns Harbor by the existence of a permanent, 400-




foot long, canopy-type hood, commonly termed "coke-side shed,"




that semi-enclosed the coke side of Battery No. 1.




     The following two major components comprised the coke-side




emissions released into the shed:




     1.  Coke-pushing operation emissions resulting from:




         a.  Coke pushing—an intermittent source emission lasting




             about 15 to 45 seconds  and occurring on an irregular




             basis with an average interval between pushes of 13




             minutes;




         b.  Quench car movement—an intermittent source emission




             emanating from the coke in the quench car and lasting

-------
                              - 2 -


             about 15 to 45 seconds, from the end of

             a coke-oven push until the quench car exits

             from the shed; and

     2.  Leaking coke-side doors emissions; in the aggregate,

         the 82 coke-side doors of Battery No. 1 released emis-

         sions at a fairly cbnstant rate.

These two emission components—especially the pushing operation—

caused the emissions conveyed through the shed exhaust duct to vary

widely with respect to particulate concentration, opacity, chemical

composition, temperature, and particle size as a function of time.

     Since the shed was installed to capture and transport all of

the coke-side emissions to a retrofitted control device (not in-

stalled at the time of this study) , the original testing protocol

specified emission tests only in the (induced draft) duct that

exhausted the shed.  During the tests, however, visibly-significant

quantities of particulate emissions were observed leaking from the

shed, indicating that the shed's capture and transport efficiency

was less than 100 percent.  Therefore, the scope of the project

was expanded to provide an estimate of the magnitude of these leaks,

     Finally, to be fully responsive to the needs and objectives
                                                            (1,2)
of this test program, a large number of additional, expected

air contaminants were measured during this study as shown in

Table 1.0-1.  The rationale and purposes for sampling each of these

materials are given in Table 1.0-2.

     The field sampling portion of the study was performed on March

3-7, 1975, after some initial range-finding determinations were

made on February 24, 1975.  The range-finding determinations

included exhaust gas flowrate, moisture content, gas composition

-------
                                              TABLE 1.0-1
                               GENERALIZED ROSTER OF EMISSIONS  INVESTIGATIONS

                                             Burns Harbor Plant
                                        Bethlehem Steel  Corporation
                                           Chesterton, Indiana
                                             March  3-7,  1975
         Sampling  Method
           Contaminant
      Analytical Method
1. EPA 5 (Final In-House Draft
   7-25-74) Modified as  per
   Appendix S
2. Brink Cascade Impactor outside
   stack at 4 Individual points
3. ASTM D1739-70
*• EPA 11

5. SPA 8
6. Absorption in diluted sulfuric
   acid
Particulate

a.  Acetone-soluble content
b.  Water-soluble content
c.  Water-soluble arsenic
d.  Water-soluble chloride
e.  Water-soluble simple cyanide
f.  Water-soluble mercury
g.  Water-soluble pH
h.  Water-soluble acidity/alkalinity
i.  Metals (Ca, Fe, Mg, Pb, Al, Cd,
    Cu, Be, Se, Ti) content
j.  Total sulfate

Particle Size

    Acetone-soluble content

Dustfall

a.  Weight retained on No. 18 sieve
b.  pH
c.  Acetone-soluble content
d.  Cyclohexane-soluble content

Hydrogen Sulfide

Sulfur Dioxide-Sulfuric Acid
Mist (as S03)

Ammonia
EPA 5 (Final In-House Draft
7-25-74)Modified as per Appendix S
c.  Atomic absorption
d.  Ion-selective electrode
e.  Ion-selective electrode
f.  Atomic absorption
g.  pH electrode
h.  Filtration/titration
i.  Atomic absorption and visi-
    ble spectrophotometry (Se, Ti)
j.  Visible spectrophotometry

Brink Manufacturing specifica-
tions (gravimetry)
ASTM D1739-70

a.  Gravimetry
b.  pH electrode
EPA 11

EPA 8


Ion-selective electrode
                                                     Clayton Environmental  Consultants,  Inc.

-------
                                         TABLE  1.0-1  (continued)
                               GENERALIZED ROSTER OF  EMISSIONS INVESTIGATIONS

                                             Burns Harbor Plant
                                         Bethlehem  Steel  Corporation
                                           Chesterton,  Indiana
                                             March 3-7,  1975
         Sampling Method
           Contaminant
      Analytical Method
7. Absorption in sodium hydroxide
8. Adsorption in activated carbon
9. Grab collection in glass flask
10. Prefilter + absorption in
   cyclohaxane
                                                         -2,
 11.  Filtration
Soluble Chloride
Total Sulfite (as S03"
Total Sulfate
Insoluble Sulfate
Simple Soluble Cyanide
Complex Soluble Cyanide
Insoluble Cyanide
Total Soluble Phenolics (as CgH5OH)
Total Insoluble Phenolics (as
Nitrate + Nitrite (as NOg )

Pyridine
Beta-naphthylamine
Benzene Homologues
Benzene
Total Light Hydrocarbons (as
Methane and Homologues (as
Ethylene and Homologues (as
Acetylene
Carbon Monoxide
Cyclohexane-soluble content
Cyclohexane-insoluble content
Fluoranthene
Pyrene
Chrysene + Triphenylene + 1,2-
  benzanthracene (as Chrysene)
Benzo(a+e)pyrene (as benzo(a)pyrene'

Fugitive Emissions
Ion-selective electrode
Visible spectrophotometry
Visible spectrophotometry
Visible spectrophotometry
Ion-selective electrode
Ion-selective electrode
Ion-selective electrode
Gas chromatography
Gas chromatography
Visible spectrophotometry

Gas chromatography
Gas chromatography
Gas chromatography
Gas chromatography

Gas chromatography
Gas chromatography
Gas chromatography
Gas chromatography
Gas chromatography

Gas chromatography
Gas chromatography
Gas chromatography
Gas chromatography
Gas chromatography

Gas chromatography

Gravimetry
                                                     Clayton Environmental Consultants,  Inc.

-------
                                           TABLE  1.0-2

                            PURPOSE(S) FOR SAMPLING EACH CONTAMINANT

                                       Burns Harbor Plant
                                   Bethlehem  Steel  Corporation
                                       Chesterton,  Indiana
                                         March  3-7,  1975
               Contaminant
                 Purpose  of  Measurement
Particulate
  a.   Acetone-soluble content
  b.   Water-soluble content

  c.   Water-soluble arsenic

  d.   Water-soluble chloride


  e.   Water-soluble simple cyanide



  f.   Water-soluble mercury

  g.   Water-soluble pH


  h.   Water-soluble acidity/alkalinity
1.  Measure particulate emission rate
2,  Measure difference between door leakage and pushing
    operations emissions
3.  Obtain data for emission factor for coke-oven
    battery
4.  Measure difference between front half and total
    particulate

1.  Select control device
2.  Possible hazardous material
3.  Determine organic content
4.  Corrosion resistance

    Standard technique

    Possible hazardous material

1.  Select control device
2.  Corrosion resistance

1,  Select control device
2.  Corrosion resistance
3.  Possible hazardous material

    Possible hazardous material

1.  Select control device
2.  Corrosion resistance

1.  Select control device
2.  Corrosion resistance
                                                Clayton Environmental Consultants, Inc.

-------
                                     TABLE  1.0-2  (continued)

                             PURPOSE(S) FOR SAMPLING EACH  CONTAMINANT

                                       Burns Harbor Plant
                                   Bethlehem Steel  Corporation
                                      Chesterton,  Indiana
                                        March 3-7, 1975
               Contaminant
                 Purpose  of  Measurement
  i.   Metals content (Ca, Fe, Mg, Pb,
        Al, Cd, Cu, Be, Se, Ti)

  j.   Total sulfate
Particle Size
  a. ' Acetone-soluble content
Dustfall
  a.   Weight retained on No. 18 sieve
  b.  pH
  c.   Acetone-soluble content
  d.   Cyclohexane-soluble content
1.  Possible hazardous material
2.  Standard EPA data-gathering procedure

1.  Corrosion resistance
2.  Measure sulfate in catch, possible pseudo-
    particulate

1.  Determine particle size distribution
2.  Select control device
3.  Compare with filterable particulate results

1.  Possible hazardous material
2.  Compare with particulate catches

1.  Measure shed versus non-shed particle fallout
    (dustfall)
2.  Measure dustfall on bench
3.  Measure bench versus ground-level dustfall
4.  Obtain data for emission factor for coke side of
    battery

1.  Exclude big chunks of coke
2.  Measure shed versus non-shed dustfall

1.  Select control device
2.  Corrosion resistance

1.  Possible hazardous material
2.  Compare "organics"

1.  Possible hazardous material
2.  Compare "organics"
                                                 Clayton Environmental Consultants, Inc.

-------
                                     TABLE 1.0-2 (continued)

                            PURPOSE(S) FOR  SAMPLING EACH  CONTAMINANT

                                       Burns Harbor Plant
                                   Bethlehem Steel Corporation
                                       Chesterton, Indiana
                                         March  3-7, 1975
               Contaminant
                 Purpose  of  Measurement
Sulfur Dioxide and Sulfuric Acid Mist
  (as S03)


Ammo n i a

Soluble Chloride
Total Sulfite  (as S03~2), Total Sulfate,
  and Insoluble Sulfate

Simple Soluble Cyanide, Complex Soluble
  Cyanide, and Insoluble Cyanide
Total Soluble Phenolics  (as C^E^OE) and
  Total Insoluble Phenolics (as

Nitrate + Nitrite (as N03~2)

Pyridine

Beta-naphthylamine

Benzene and Homologues of Benzene

Total Light Hydrocarbons  (as CH^)

Methane and Homologues (as

Ethylene and Homologues  (as
1.  Select control device
2.  Corrosion resistance
3.  Environmental impact

    Environmental impact

1,  Select control device
2,  Corrosion resistance

1.  Select control device
2.  Corrosion resistance

1.  Select control device
2.  Corrosion resistance
3.  Possible hazardous material

    Environmental impact
    Select control device

    Possible hazardous material

    Possible hazardous material

    Possible hazardous material

    Measure organic emissions

    Measure organic emissions

    Measure organic emissions
                                                 Clayton Environmental Consultants. Inc,

-------
                                     TABLE  1.0-2  (continued)

                            PURPOSE(S) FOR SAMPLING EACH CONTAMINANT

                                       Burns Harbor Plant
                                   Bethlehem  Steel  Corporation
                                       Chesterton,  Indiana
                                         March  3-7,  1975
               Contaminant
                 Purpose of Measurement
Acetylene

Carbon Monoxide

Cyclohexane-soluble content and cyclo-
  hexane-insoluble content

Fluoranthene

Pyrene

Chrysene + Triphenylene + 1,2-
  Benzanthracene (as Chrysene)

Benzo(a+e)pyrene (as Benzo (a)pyrene)

Fugitive Emissions
Hydrogen Sulfide
1.

2.
Measure organic emissions

Possible hazardous material

Possible hazardous material


Possible hazardous material

Possible hazardous material

Possible hazardous material


Possible hazardous material

Obtain data for emission factor for coke side of
battery
Evaluate capture efficiency of the shed and the
significance of shed leakage

Environmental impact
                                                         i
                                                         oo
                                                Clayton Environmental Consultants,  Inc.

-------
                               - 9 - .





(during pushing only and also as a continuous or integrated




measurement), temperature, and filter obscurity.  A list of




project participants is given in Table 1.0-3.




     Some possible questions and answers that may arise while




reading this report are listed in Section 7.0.

-------
                              -  10  -


                           TABLE 1.0-3

                       PROJECT PARTICIPANTS

                       Burns  Harbor  Plant
                   Bethlehem Steel Corporation
                       Chesterton, Indiana
                         March 3-7, 1975
U.S. Environmental Protection Agency
Division of Stationary  Source Enforcement

     Louis R. Paley, P.E.,  Project  Officer
     Mark Antell
     Bernard Bloom
     Don Carey
U.S. Environmental Protection Agency,  Region  III

     Joseph W. Kunz


U.S. Environmental Protection Agency,  Region  V

     R. Edwin Zylstra
     Dave Shulz
     Paul R. DePercin
     Steve Florin
     Basim Dihu
U.S. Environmental Protection Agency
National Enforcement Investigation Center,  Region  VIII

     Dave Brooman
U.S. Environmental Protection Agency
Audio Visual Branch	

     Ron Mitchell
     Stanley Couer
Bethlehem Steel Corporation

     C.A. Trageser
     Robert M. Harvey
     Norm D. Hodgson
     Tom Kreichett
     John T. Dunn
     Carolyn Mance
     Ron K. Spalding
     Dave Fisher
     Gerald Merchant

-------
                             - 11 -
                     TABLE  1.0-3  (continued)

                      PROJECT PARTICIPANTS

                       Burns Harbor Plant
                   Bethlehem Steel  Corporation
                       Chesterton,  Indiana
                         March  3-7,  1975

Clayton Environmental Consultants, Inc.

     Field Team
     Richard J. Powals, P.E.
     Victor W. Hanson
     Fred I. Cooper
     Richard G. Keller
     Richard C. Marcus
     Richard J. Griffin
     Gerald E. Hawkins
     Kent D, Shoemaker
Project Leader
Senior Environmental Control  Specialist
Group Leader, Source Sampling Studies
              Control Specialist
              Control
              Control
              Control
Environmental
Environmental
Environmental
Environmental
Chemist
Specialist
Specialist
Specialist
     Data Analysts

     Janet L. Vecchio
     Rebecca B. Cooper
Group Leader, Data Processing
Environmental Control Specialist
     Laboratory Analysts

     Aileen G. Hayes
     David J. Holmberg
     John Knowles
     Michael D. Kelly
     Nathan C. Riddle
     Kent D. Shoemaker
     James M. McClain
Assistant Director, Laboratory Services
Laboratory Shift Supervisor
Chemist
Chemist
Chemist
Chemist
Chemist
     Managing Consultant

     John E. Mutchler, P.E.
Vice-President, Engineering Services
                                Clayton Environmental Consultants,  Inc.

-------
                             -  12  -


                  2.0  SUMMARY AND CONCLUSIONS

2.1  Coke-Side Particulate Emissions

     Two types of particulate emissions were observed to emanate

from the coke side of the shed at Battery No. 1 at the Burns

Harbor plant of Bethlehem Steel Corporation.  These emissions

comprised particulate matter discharged through the shed exhaust

duct and fugitive particulate matter which escaped the shed.  The

combination of these two emissions is referred to below as "Overall

Continuous Coke-Side Particulate Emissions."


     2.1.1  Overall Continuous Coke-Side Particulate Emissions

            The overall filterable coke-side particulate emissions

     ranged from 0.89 to 0.93, and averaged 0.91 pound per ton of

     dry coal fed to the ovens (+ 0.06 pound per ton).*  These

     emission measurements inherently include contributions from

     the following sources:  coke-pushing operations  (coke push-

     ing and quench car movement), door leaks, and residual par-

     ticulate concentrations within the shed volume from previous

     pushes, as well as emissions which escaped the shed.  The

     hourly emission rate corresponding to these emissions ranged

     from 143 to 151, and averaged 146 pounds per hour (+ 10

     pounds per hour).
         The notation "+ 0.06 pound per ton" is  an estimate
         of the  statistical precision of the average value
         based upon a 95-percent  level of confidence.  Al-
         though  the precision is  _+ 0.06, the confidence
         interval for a  concentration, emission  rate, or
         emission factor is always bounded by a  minimum
         value of zero.   (See Section 5.10.)

-------
                       - 13 -


2.1.2   Continuous Particulate Emissions from the Exhaust
        Duct

        Filterable particulate emission  measurements made

in the exhaust duct evacuating the shed on a continuous

basis indicated an average emission factor of 0.78 pound

per ton of dry coal fed to the ovens (+ 0.04 pound per ton)

The corresponding average hourly emission rate for continu-

ous particulate emissions was 124 pounds per hour (+ 10

pounds per hour).  These emission measurements inherently

include coke-pushing operations emissions, door leaks, and

residual emissions from previous pushes, but exclude

fugitive emissions.


2.1.3   Peak Particulate Emissions from the Exhaust Duct

        Particulate emission measurements made during the

initial 3-minute period when pushing emissions were being

evacuated from the shed (heaviest visible emission period)

indicated an average emission factor of 0.64 pound of

filterable particulate per ton of dry coal charged to the

ovens (+ 0.34 pound per ton).  The corresponding emission

rate for this period averaged 93.2 pounds of filterable

particulate per  hour (+ 47.9    pounds per hour).  It

should be noted that because of the frequency and overall

duration of sampling, the emission rates for these peak

emissions have been adjusted to reflect typical operations;

i.e., 4.5 pushes per hour.  In addition, these values in-

herently include door leaks and residual emissions from pre-

vious pushes, but exclude fugitive emissions.

-------
                      - 14 -





2.1.4   Fugitive Particulate Emissions from the Shed




        The fugitive particulate emissions from the shed




occurred at four positions:  The north and south ends of




the shed, the Askania valves, and the boundary between the




shed and the coke battery.  On a continuous basis, these




(filterable) fugitive emissions were estimated to average




0.14 pound per ton of dry coal fed to the ovens, or 21.9




pounds of fugitive particulate per hour.  Related to the




continuous filterable particulate emissions, they averaged




15 percent of the overall emissions.






2.1.5   Particulate Emissions for Pushing Operations




        Using the particulate emissions data presented




previously and a straightforward calculational procedure,




it was possible to obtain a rough estimate of the particulate




emissions attributable to pushing operations alone at the




Burns Harbor plant.  These emissions were estimated to




average 0.69 pound of filterable particulate per ton of




dry coal fed to the ovens (+ 0.51 pound per ton).  This




emission factor has been adjusted to include fugitive




emissions from the shed.






2.1.6   Particulate Emissions for Non-Pushing Operations




        The overall coke-side emissions for non-pushing




operations were roughly estimated to average 0.22 pound of




filterable particulate per ton of dry coal fed to the ovens




(+ 0.46 pound per ton).  This factor has been adjusted to




include fugitive emissions.

-------
                           - 15 -





             Using these data, pushing operations were found,




     on an average basis, to account for 76 percent of the over-




     all coke-side particulate emissions, while 24 percent were




     attributable to non-pushing operations.






2.2  Shed Particulate Capture Efficiency




     Because significant visible fugitive emissions were ob-



served escaping from the shed during the study, and in order




for EPA to evaluate the cost-effectiveness of the shed concept,




it was necessary to evaluate the particulate capture efficiency




o f the shed.






     2.2.1   Evaluation of Shed Capture Efficiency




             The efficiency of the shed in capturing and exhaust-'




     ing coke-side emissions from pushing (based upon particulate




     emission measurements) was found to be approximately 85



     percent.  Thus, on a "continuous" basis, an average of 15




     percent of the particulate emissions escaped from the shed.






     2.2.2   Possible Causes of Leakage




             Several potential causes for the existence of fugi-




     tive particulate emissions have been suggested.  These




     include the following:




             1.  The overall magnitude of the shed's holding




                 volume appeared to be too small relative to the




                 magnitude of the emissions, and the effective




                 exhaust rate of the shed may have been too low;




             2.  It is possible that "short circuiting" of the




                 outside air to the exhaust duct occurred; and




             3.  The shape, size, and location of the holding

-------
                          - 16 -





                 chamber and/or the exhaust duct, as well as




                 the shed wall and end openings, may have




                 affected the capture efficiency of the shed.






2.3  Chemical Composition of Particulate Emissions




     The particulate matter samples taken during this study were




subjected to 19 separate analyses to determine particulate com-




position.  The results indicated that the particulate matter was




predominantly carbonaceous with undetectable or trace amounts of




nearly all other constituents for which analyses were performed.






2.4  Particle Size Distribution




     The size distribution of particulate matter varied greatly




as a function of sampling position in the exhaust duct, probably




due to the numerous changes in direction of the exhaust gas flow




within the duct.  On an average basis, however, approximately 32




percent of the particulate was smaller than seven microns and




approximately seven percent was smaller than one micron.






2.5  Emissions of Other Materials




     In addition to particulate, sampling was conducted to deter-




mine the concentration of 29 other potential air contaminants from




coke pushing.  Cyclohexane solubles and insolubles, ethylene and




homologues, and total light hydrocarbons were found to be




discharged at emission rates exceeding 100 pounds per hour.  All




other measured contaminants were detected at levels that averaged




less than 16 pounds per hour.






2.6  Dustfall Measurements




     Dustfall measurements were taken within the shed on Battery



No. 1 and at similar locations on the adjoining unshedded but

-------
                           - 17 -





generally-comparable battery.  The purpose of these measurements




was to assess the shed's effect on dustfall rate at employee




work stations.




     For three of the four locations considered, dustfall




(settleable particulate) rates beneath the shed were  statisti-




cally greater than those at corresponding locations in the




unshedded No. 2 Battery.  The dustfall at the bench location,




the primary work station, did not differ between the  two batteries




As expected, greater dustfall rates were experienced  at the No. 1




Battery near the shed wall than at locations nearer the bench.




Thus, the shed's design effectively causes the increased quantity




of dustfall to drop away from the work stations to a  location




near the wall of the shed.






2 .7  Indices of Visible Emissions




     2.7.1  Degree of Greenness




            The average value for the degree of greenness of



     the pushed coke (product of the sum of the greenness ratings




     and the duration of the push) ranged from 222 to 285 for the




     three particulate sampling periods.  The third particulate




     sampling period was found to contain pushes of higher




     greenness ratings than the other two sampling periods.






     2.7.2  Opacity




            Opacity data were acquired for the two stacks dis-




     charging emissions from the shed exhaust duct during the




     study.  For the 3-minute "peak" periods during particulate




     sampling, the average opacity was found to range from 40




     to 60 percent.  The third particulate sampling period

-------
                          - 18 -





     resulted in an average opacity which exceeded that of the




     other two sampling periods.






     2.7.3   Percent of Doors Leaking




             For particulate sampling days, coke-side oven door




     leakage was found to vary from 27 to 69 percent on Battery




     No. 1 and from 39 to 64 percent on Battery No. 2.  Push-




     side door leakage for both batteries was found to be less




     than that of coke-side door leakage and was less variable.






2.8  Process and Emissions Correlations




     Linear correlation techniques were attempted but revealed




no significant relationship between the continuous filterable




particulate emission factors or the filterable particulate




push-only emission factors and average degree of greenness or




average opacity.  Further, no statistically significant




relationships were found between continuous filterable particu-




late emission factors and net coking time or average flue tem-




perature.  The small sample size, however, limited the sensi-




tivity of the statistical analyses in these cases.




     For the particle sizing samples, no linear correlation was




apparent between variations in size distributions for each of




the samples and the greenness of the push.  In addition, no




significant correlation was apparent between particle size and




net coking time.  Again, however, the small sample size limited




the sensitivity of the statistical technique.




     Both peak opacity and greenness were found to be very highly



correlated with net coking time,  minus a constant of 1000 min-




utes, when the reciprocal of each of the values was used.  In

-------
                          - 19 -





addition, peak opacity and greenness were each very highly




correlated with flue temperature.




     A highly significant statistical relationship was apparent




between the degree of greenness and the opacity of the exhaust




duct emissions.  This relationship characterized opacity as a




function of the logarithm of greenness.






2.9  Representativeness of Process and Shed Conditions




     In order to document that the measured results were repre-




sentative of Battery No. 1's actual emissions during normal pro-




duction, six criteria for comparison were established.  On the




basis of these criteria, all samples obtained during this study




were found to be taken during generally representative process




and shed conditions.  Only three minor deviations from the




criteria were observed during the more than 300 observations




used to establish representativeness.

-------
                              -  20  -
             3.0  PROCESS DESCRIPTION AND OPERATIONS




3 .1  Process Description




     Bethlehem Steel Corporation operates a by-product metallurgical




coke manufacturing facility at its Burns Harbor plant in Chesterton,




Indiana.  This operation includes a destructive distillation pro-




cess, generally  termed "coking," that occurs when coal is heated




in an atmosphere of low oxygen content.  By-product organic com-




pounds, generated during the coking, are recovered from the coke-




oven off-gases.  The main product, de-gasified coa1, commonly known




as  "coke," is a critical raw material used  in the production of




iron.               ">*!*



     The by-product coking process occurs in a "coke battery,"




a series of contiguous, rectangular, refractory-lined ovens.  At




the Burns Harbor plant, two coke batteries, each containing 82




ovens, are positioned end-to-end (See Figure 3.1-1).  Each oven,




20 feet tall, 18 inches wide (average width; the oven is actually




tapered), and 50-feet long, is capable of producing about 24.5




tons of coke per push.  At the beginning of a coking cycle, coal




is charged (dumped) through ports in the top of each oven.  Subse-




quently, each port is sealed and heat is applied to the oven to




maintain a temperature of 2300 to 2450°F.  About 18 hours later, at




the end of the coking cycle, the incandescent coke is pushed from




the oven with a mechanical ram into a specially-designed railroad




car, called a "quench car."  The load of hot coke in the quench car




is subsequently flooded with water at the quenching station.




     The large number of ovens on each battery makes it possible




to average 4.5 pushes per hour, utilizing the same equipment for

-------
                                    Push Side
              Battery No. 2
       (Koppers Design - 82 ovens)
  Spare
  Door
    2
Bench
 223
Bench
 252
           Bench
            278
                                    Coke
Wharf
            Ground
              236
Ground
  251
Ground
  271
                                         Battery No. 1
                                 (Wilputte Design - 82 ovens)
            Bench
             178
          Bench
           152
Bench
 123
Door
 l
                            Side
                                                i  i >      i
                                                r Ttiri
                                                                111
                                                                ~IT
4  GroundV
/   177    \
                                             Ground
                                               151
                                            Vail
                                             121
i

N)
                                                                  Wharf
                                         End 1 & 2
                                         (Ground 180)
                                            Mid 1 & 2
                                            (Ground 152)
                               FIGURE 3.1-1
   SCHEMATIC DIAGRAM OF BY-PRODUCT, METALLURGICAL COKE MANUFACTURING FACILITY
                               WITH DUSTFALL SITES

                               Burns Harbor Plant
                          Bethlehem Steel Corporation
                               Chesterton, Indiana
                                March 3-7,  1975
                                              Clayton Environmental Consultants,  Inc

-------
                           -  22  -


charging the coal to each  oven.  During the  coking  cycle,  the

oven is sealed on both ends  with refractory-lined doors which

are locked into place just prior to oven  charging.   The doors

are then removed just prior  to  pushing the coke  from the oven.

     Emissions from the coke ovens can occur  throughout the

cycle from around the sealed doors ("door  leaks"),  as well as

at the end of the cycle when the coke is  pushed  from the oven

("pushing").  The duration of the coke pushing operations  phase

of the oven cycle is brief,  lasting about  45-90  seconds (approx-

imately 30 seconds for coke-pushing emissions and 15  to 60 sec-

onds for quench car movement emissions).   Nevertheless, emis-
                                     '  '    V.HH
sions during this brief period  can be very copious.

     The emissions generated from the coke side  of  Battery No. 1

at the Burns Harbor plant  from  door leakage  and  coke  pushing are

predominantly captured by  a  semi-enclosed  structure  termed the

"shed."  A schematic of this enclosure is  presented  in Figure

3.1-2.  The shed is designed to capture these emissions, which

exhibit significant thermal  rise, and exhaust them  through the

duct located at the shed's apex.  This mechanically-exhausted,

coke-side shed is a canopy-type hood that  is  about  400 feet

long and encloses a volume of about 225,000  cubic feet.  De-

spite the design, fugitive emissions escape  this enclosure on

both ends, as shown in Figures  3.1-2 and  3.1-3.  The  source

testing performed during this project was  designed  to measure

total coke-side emissions.   Therefore, measurements were made

of the emissions collected and  exhausted  through the  duct as

well as the fugitive emissions  from the shed.  These  fugitive

emissions were documented  as they related  to  coke battery and

shed operating parameters.  Additionally,  the study documented

-------
                                Battery 2  (without  shed)
                                                    Topside Askania Valve Leakage
                                                           Battery 1 (with shed)
                                                                 Holding chamber
                                                                 or "henhouse"
                                                                         Topside  Shed
                                                                      Boundary Leakage
   Shed - .|
   Exhaust
   Duct
Sampling Ports
        Shed
                             FIGURE 3.1-2

                  SCHEMATIC DIAGRAM OF COKE-SIDE  SHED

                          Burns Harbor Plant
                     Bethlehem Steel Corporation
                         Chesterton, Indiana
                           March  3-7, 1975
                                             Clayton Environmental Consultants,  Inc.

-------
             North  End
                                            South End
Holding Chamber
     or
 "Henhouse"
                          Exhaust Duct
                             Sampling
                               arcs
                           Oven
                                 Oven
Holding Chamber
     or
 "Henhouse"
                                                                                   to
                                                                                   -p-
                             FIGURE  3.1-3
         SCHEMATIC  DIAGRAM OF COKE-SIDE SHED END OPENINGS

                        Burns  Harbor Plant
                    Bethlehem Steel Corporation
                       Chesterton, Indiana
                         March 3-7, 1975
                                           Clayton Environmental Consultants, Inc.

-------
                          - 25 -






the "dustfall" rates within the shed and at similar positions




on the unshedded Battery No. 2 (see Figure 3.1-1).






3.2  Representativeness of Process and Shed Conditions




     Because it cannot be assumed (without documentation) that




measured results are representative of the actual emissions of




a source, it was necessary to:  (1) define thoroughly the ob-




jective^) of the test program prior to developing the test




protocol; (2) identify specifically the process, control device,




test,  and analytical conditions required to achieve the objec-




tive^); (3) define, in advance of testing, the acceptable




range  for each parameter; and (4) document that the required




conditions were maintained during the test period.






     3.2.1  Criteria for Comparison




            Before the results presented in this report could




     be considered representative of non-test period operations,




     it was necessary to document that all relevant process and




     operational conditions for test and specific non-test




     periods were acceptably "similar."  The criterion for ac-




     ceptably similar data was arbitrarily defined as + 10 per-




     cent of the average typical operating conditions.




            Further, to more clearly define the terminology re-




     garding coke-pushing operations, the following definitions




    were formulated:




            Atypical operating conditions;   extremely infrequent




            major process changes (or upsets).




            Typical  operating conditions:  any  process operating




            conditions not atypical.

-------
                         - 26 -


             Abnormal operating conditions;  any typical

             operating conditions during which net coking time

             is outside the normal net coking time or during

             which any coke-pushing stoppage greater than 30

             minutes duration occurs.

             Normal operating conditions:  any typical operating

             conditions not abnormal.

Therefore, "abnormal" and "normal" operating conditions are com-

plementary subsets within the category "typical" operating

conditions .

     After a preliminary assessment of the parameters given in
                  ^-•.,.i •
Table 3.2.1, it was determined that some of the process varia-

bles were nearly constant (Askania valve pressure, average daily

oven cross-wall temperature, etc.) while many others (such as

net coking time) were not constant with time.  Additionally,

based upon the effect of a given parameter upon the shed exhaust

duct opacity and the particulate emission rates reported from

earlier source tests at other coke-oven facilities, Mr. Paley

(U.S. EPA) and Mr. Powals (Clayton Environmental Consultants,

Inc.) decided to limit the testing of the battery's coke-side

emissions to periods when the conditions were maintained within

the ranges given below:

     1.  Net coking time of 17-1/4 to 18-1/2 hours

         (1035-1110 minutes),

     2.  Coke-pushing cycle duration up to 30 minutes

         long (within + 10 percent) ,

     3.  Coal feed rate of 35 tons (wet) per charge

         (within _+ 10 percent) ,

-------
            _  27  -          TABLE 3.2.1
                       PERTINENT PARAMETERS
                       Burns Harbor  Plant
                   Bethlehem Steel Corporation
                       Chesterton, Indiana
                         March 3-7, 1975

Coking Time (minutes)
     Net
     Minimum

Average Daily Cross-Wall Temperatures  (°F)

Coal
     Feed rate (wet pounds  charged per  oven)
     Feed rate (dry pounds  charged per  oven)
     Chemical and physical analyses  (average values)

Average Underfire Gas Flow  (103  CFH)

Coke
     Average rate of production  (tons per day)
     Physical analyses (average  values)

Coke Oven Gas
     Average rate of production  (10^ CFH)
     Chemical analyses

Average Askania Valve Pressure  (mm H20)

Battery Operations
     Number and location of empty ovens
     Door maintenance
     Scheduled/unscheduled downtime
     Use of experimental doors
     Occurrence of atypical or abnormal events

Coke-Pushing Operation
     Clock time for each oven pushed
     Duration of each push  (seconds)
     Duration of each push cycle  (minutes)
     Greenness of coke-oven push.

Shed Evaluation
     Average exhaust rate  (actual cubic feet per minute)
     Fan curves
     Amperage and voltage used
     Duration required to clear  the  peak  (push) emissions

Oven Door Leak Observations
     Coke side
     Push-side and chuck doors

Local Surface Wind Conditions
     Average speed (mph) and direction  (degrees) during  test  period
     Speed (mph)  and direction  (degrees) during non-test period
     Persistence at 45° and _+ 6  mph  of winds

Shed Design Parameters (size, shape)

                           Clayton Environmental Consultants,  Inc.

-------
                     -  28  -





4.  Coke production rate of 24 tons per push




    (within + 10 percent) ,




5.  Average pushing rate of 4.5 pushes per hour




    (within + 10 percent) ,




6.  Coal analysis (average percent coal moisture




    and BTU/lb coal) comparable to that of typi-




    cally charged coal (within + 10 percent), and




7.  Shed evacuation rate of 300,000 SCFM (within




    _+ 10 percent) .






3.2.2   Conditions  during Sampling Periods




        On the basis of the criteria presented in Section




3,2.1, all samples  obtained during this study were taken




during generally representative process and shed conditions




by interrupting sampling during abnormal and atypical




periods.  Tables 3.2.2-1 and 3.2.2-2 indicate the average




values and/or the range of values for each of the six




criteria discussed  previously for periods before, during,




and after particulate sampling.  Exceptions to the first




of the six criteria occurred on the two occasions when




Oven 191 was pushed, i.e., during Particulate Tests 1 and




3.  The net coking  time for this oven was approximately




1400 minutes due to its position at the end of the battery;




this is typical for Oven 191.  There was also a single




exception to the second criterion during Particulate Test




No. 2, when a 39-minute interval between pushes occurred.




These three observations were the only deviations, however,




from the six criteria (encompassing over 300 observations)

-------
                             -  29  -

                         TABLE 3.2.2-1

             COMPARISON OF KEY PROCESS  PARAMETERS
                        (BATTERY  NO.  1)

                      Burns  Harbor Plant
                  Bethlehem Steel Corporation
                      Chesterton, Indiana
                        March 3-7, 1975
Parameter
Average Net Coking
Time (minutes)
Range of Time Between
Pushes (minutes)
Average Time Between
Pushes (minutes)
Average Wet Coal Feed
Rate (tons/charge)
Average Coke Produc-
tion Rate (tons/push)
Average Number of
Pushes/hour
Average Coal Moisture
Content (percent)
Average BTU/lb Coal
Average Shed Evacua-
tion Rate (DSCFM)
Time Period
Before
Sampling
All
Data
1096
3-105
13
**
**
4.4
6.7
1162
All
Typical*
Data
1076
3-99
13
**
**
4.5
6.6
1160
295,000
(BSC Data)
During
Continuous
Particulate
Sampling
1071
7-39
12
35.0
25.6
4.9
6.9
1160
268,000
After
Sampling
All
Data
1073
5-53
13
34.8
25.4
4.5
7.3
1166
All
Typical*
Data
1071
5-53
13
34.8
25.4
4.5
7.3
1166
	
 * Typical data is all data other than that for which  five  (4.5
   + 10%) or more consecutive net coking times were outside of
   the range of 17-1/4 to 18-1/2 hours, i.e., 1035 to  1110 min-
   utes.

** Information requested but not received.
                          Clayton Environmental Consultants,  Inc,

-------
- 30 -
               TABLE  3.2.2-2
  COMPARISON OF KEY  PROCESS  PARAMETERS
             (BATTERY  NO.  2)

           Burns Harbor  Plant
       Bethlehem Steel Corporation
           Chesterton, Indiana
             March 3-7,  1976
Parameter
Average Net Coking Time
(Minutes)
Range of Time Between
Pushes (Minutes)
Average Wet Coal Feed Rate
(Tons/Charge)
Average Coke Production
Rate (Tons/Push)
Average Number of
Pushes/hour
Average Coal Moisture
Content (Percent)
Average BTU/lb coal
Time Period
Before
Sampling
1028
4-122
34.8
25.4
4.8
6.7
1281
During
Sampling Days
(March 3-7, 1975)
1003
6-56
35.0
25.6
4.9
6.9
1207
After
Sampling
1003
5-66
35.3
25.8
4.9
7.3
1227
               Clayton Environmental  Consultants,  Inc.

-------
                          -  31  -





established for representative sampling.




     The intent to ensure that these tests were representative




of typical operations at the maximum production rate required




the sampling to be delayed several months during a coal strike




until the plant had been at typical operating conditions for




at least one week.  Two days before the scheduled initiation



of the emission tests, the plant incurred a major upset (coal




feed conveyor breakdown) which caused a few additional days of




atypical operating conditions.  This resulted in a second delay




in the test schedule.  A third delay occurred when, after a pre-




liminary traverse, the stack gas exhaust rate was found to be




somewhat under the criterion mentioned in Item 7 above.  However,




after assurances were given by Bethlehem Steel Corporation



personnel that the exhaust rate was at maximum and representa-




tive conditions, sampling commenced.  During sampling, the



stack gas exhaust rate was nearly within the _+ 10 percent cri-




terion.  A few plant equipment problems did cause abnormal



operations during the test period.  However, since the effects




of such irregularities passed rapidly, the response to such




events was only to interrupt the tests until the process and




shed conditions were again operating normally.




     The procedure used to ensure these representative process



and shed conditions included documenting, by comparison to data




from other operating periods, the fact that the sampling period



was representative of typical operations.  All process opera-




tions and shed performance data acquired for the sampling period



and for periods prior to and following the sampling period are




provided in Appendices A-0 (Volumes 2-4).

-------
                          -  32  -





3.3  Identification of Possible Normalizing Factors




     Source sampling and analytical data must frequently be




normalized using process or performance data to obtain a more




representative characterization of emissions due to varying




operational conditions.  Prior to, during, and subsequent to




this study, potentially significant process, shed, meteoro-




logical and emission data were obtained and recorded by




both Bethlehem Steel Corporation and EPA personnel.  Those




parameters considered as possible normalizing factors were




given in Table 3.2.1.




     Two rationales were considered in the selection of a




normalizing factor:




     a.  Emission data should be normalized with reference to




         some process parameter to reflect the average (or




         "normal") particulate emission rate.




     b.  Emission data should not be normalized with reference




         to some process parameter to reflect the maximum




         emission rate (assuming continuous pushing and not




         just one push about every 13 minutes).




To reflect actual operating conditions, the first of these two




alternatives was chosen.  Further, and traditionally, emission




rates have been compared and normalized to the process input




rate, based in part upon the concept of a material balance.




Finally, in this case the process input rate (coal feed rate)




is a directly measurable quantity.  Thus, particulate emissions




data have been normalized with respect to coal feed rate to




facilitate interpretation of the particulate emissions data.




Such normalized particulate emission rates are presented in




Sections 2.0 and 5.0.

-------
                          -  33  -





           4.0  SAMPLING AND ANALYTICAL METHODS






4 .1  Test Protocol




     In general, the sampling, sample handling, calibration,




and analyses performed in this study incorporated the latest,




most well-established methods available, including those




promulgated by EPA and ASTM.  There were several instances,




however, as described in subsequent paragraphs, when modified




or novel techniques were required in order to ensure representa-




tive results.  In all cases, a method was selected only if it




satisfied the following criteria-questions:




     1.  Would the sampling procedure quantitatively catch the




         analyte of interest?




     2.  Would the analyte  of interest be caught in a medium or




         media in which it  could be separated and quantified?




     3.  Would the analytical procedure characterize all of the




         contaminant species of interest while minimizing the




         necessary number of tests?






     This study was designed to investigate thoroughly the




typical, normal emissions produced on the coke side of Battery




No. 1, including any fugitive emissions that might occur.  Thus,




the approach to meeting this objective was to define both the




process conditions and sampling and analytical methods required




to measure the emissions during typical, normal operation




conditions.  Sampling and analytical procedures were therefore




specifically designed around the process and capture system




characteristics to provide  the maximum amount of information




with a reasonable expenditure of effort.  Additionally, much

-------
                          - 34 -




care was taken to ensure that sampling occurred  only during




"normal" operations  (as agreed mutually by Bethlehem Steel




personnel and Messrs. Bloom and Paley of the DSSE, U.S. EPA;




see Section 3.2).




     After initial discussions with persons from Bethlehem




Steel Corporation and Clayton Environmental Consultants,  Inc.,




it was mutually agreed that the U.S. EPA would be responsible




for acquisition of all process data from Bethlehem Steel




Corporation while Clayton Environmental Consultants would be




responsible for acquisition of the sampling and analytical




information. The EPA was also responsible for visible emissions




data acquisition, including still and motion photography.




Accordingly, the three parties (Bethlehem Steel Corporation,




Clayton Environmental Consultants, and the U.S. EPA) worked




closely together to acquire the information necessary to  docu-




ment the results of this study.  The types of process data




acquired were reviewed in Table 3.2.1 (and are presented  in




Volumes 2 through 4), while the types of sampling and analytical




data acquired were given in Table 1.0-1.






4.2  Location of Sampling Points




     "Continuous" particulate sampling was conducted in the




shed exhaust duct shown in Figure 3.1-2.  The sampling plane




location and sampling point locations met the minimum require-




ments specified in "Method 1 - Sample and Velocity Traverses




for Stationary Sources, U.S. EPA, In-house Draft, 7-18-74"




(Appendix P, Volume 5).  The sampling plane was located 3.2




duct diameters downstream of the nearest potential disturbance

-------
                          -  35  -





and 1.2 duct diameters upstream of the stacks.  The two




sampling ports were located 90 degrees apart in the circular




duct.  The duct was subdivided into 48 equal areas; the 48




sampling points are shown in Figure 4.2-1.




     Peak particulate sampling was conducted through the same




two ports used for continuous particulate sampling.  In order




to avoid probe crossover, however, while obtaining continuous




and peak samples concurrently, the duct was subdivided into




20 equal areas for peak particulate sampling.  The 20 peak




sampling points are shown in Figure 4.2-2.




     Each of the four particle-sizing samples was taken at a




single sampling point accessed through the horizontal port.




These four tests are identified in the summary tables and graphs




by the sampling-point numbers shown in Figure 4.2-3.




     All other samples, such as sulfur oxides and samples




collected in sodium hydroxide, were taken at a single point 30




inches into the duct through the top port.  This sampling point




is indicated in Figure 4.2-3.






4.3  Continuous Particulate Emissions from Shed Exhaust Duct




     "Continuous" particulate sampling was conducted to obtain




an estimate of the "continuous" particulate emissions from the




coke-side shed and included coke-pushing operation emissions,




door leaks, and residual emissions from previous pushes.  These




samples were taken by sampling continuously except when opera-




tional upsets or downtime occurred of greater than 30-minutes




duration.




     Continuous particulate sampling was conducted in accordance




with the then-most-recent drafts of EPA Methods  1, 2, and 5

-------
                                - 36 -
   Ladder
                   Top
                         Platform
                                               Sampling Point Locations
                                               Position
                                      Horizontal
 Platform
TIT
                  Vertical
Exhaust duct diameter:  138"
3.2 exhaust  duct  diameters  downstream  of
   nearest potential  disturbance
1.2 exhaust  duct  diameters  upstream  of
   nearest potential  disturbance
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Distance
(inches)

   1-1/2
   4-3/8
   7-5/8
  10-7/8
  14-1/2
  18-3/16
  22-3/16
  26-3/4
  31-3/4
  37-1/2
  44-9/16
  54-5/16
  83-1/16
  93-7/16
 100-1/2
 106-1/4
 111-1/4
 115-13/16
 118-13/16
 123-1/2
 127-1/8
 130-3/8
 133-5/8
 136-1/2
                              FIGURE  4.2-1

           SCHEMATIC DIAGRAM OF SAMPLING POINT LOCATIONS
                 FOR "CONTINUOUS"  PARTICULATE  SAMPLES

                          Burns  Harbor  Plant
                     Bethlehem Steel  Corporation
                         Chesterton,  Indiana
                          March 3-7  1975
                             Clayton Environmental  Consultants,  Inc,

-------
                                 -  37 -
     Ladder

                 Top
                       Platform
            ,--""     Ho
         /'"        la       \.                    Sampling Point Locations
\
    *&                                            Distance
                                  Position         (inches)
    .7             v.
                   \                  1              3-7/16
    •6     •         \                 2             11-5/16
                     »                 3             20-1/8
      - £   ?.  ?.-?'?. Horizontal       *             31-3/16
                     i                 5             47-3/16
    I                 /                 6             90-13/16
    ?•$"              /                 7            106-13/16
                   /                  8            117-7/8
    •4            /                   9            126-11/16
                                     10            134-9/16
    •I
     !
Vertical
                    !
   \
      _, Platform
 Exhaust duct diameter:  138"
 3.2 exhaust  duct  diameters  downstream of nearest
    potential disturbance
 1.2 exhaust  duct.diameters  upstream of nearest
    potential disturbance
                              FIGURE 4.2-2

            SCHEMATIC DIAGRAM OF SAMPLING POINT  LOCATIONS
                     FOR "PEAK" PARTICULATE SAMPLES

                           Burns Harbor Plant
                       Bethlehem Steel Corporation
                          Chesterton, Indiana
                            March 3-7, 1975
                               Clayton Environmental Consultants,  Inc,

-------
                           - 38  -
                                     Platform
        Ladder
   Horizontal
    Platform
                                                   Particle Sizing
                                                    (one  sample  at
                                                    each  position)
Exhaust duct diameter:  138"
3.2 exhaust duct diameters downstream
   nearest potential disturbance
1.2 exhaust duct diameters upstream of
   nearest potential disturbance
of
Horizontal
 Position

     1
     2
     3
     4
Distance
(inches)

  9-1/4
 34-1/2
103-1/2
128-3/4
             Other Samples
                                               Vertical
                                               Position
                       Distance
                       (inches)

                          30
                       FIGURE 4.2-3

       SCHEMATIC  DIAGRAM OF SAMPLING POINT LOCATIONS
            FOR ALL  SAMPLES EXCEPT  PARTICULATE

                    Burns Harbor Plant
               Bethlehem Steel Corporation
                   Chesterton, Indiana
                     March 3-7, 1975
                             Clayton Environmental Consultants, Inc.

-------
                          -  39  -                                    ;




 (Final  In-house Drafts  7-18-74,  7-21-74,  and  7-25-74,  respec-




 tively).  Copies  of  these methods  are  presented  in Appendices




 P-R, respectively (Volume 5).  It  should  be noted  that this




 draft of EPA Method  5 allows the  filter  in  the  sampling train




 to be maintained  at  a temperature  other  than  "about 250°F"  and




 also allows measurement  of  the impinger  catch.   Because no




 change  in particulate concentration was  anticipated by sampling




at any temperature up to that of the stack gas,  the  temperature




of the filter for the continuous particulate  tests  was  maintained




at the average stack gas temperature, which ranged  from 84 to  98°F



 Detailed descriptions of the sampling  and analytical methods




 are given in Appendix S  (Volume  5).  The  particulate samples




 were analyzed for the materials  indicated in  Table 1.0-1.






 4.4  Determination of Peak  Particulate Emission  Period




     This study was  designed to  independently measure  both  the




 continuous particulate  emission  rate and  the  intermittent




 particulate emission rate during  the coke-pushing  operation,




 i.e., peak particulate  emission  rate.  To ascertain the average




 duration of the period  of peak particulate  emissions from  the




 exhaust duct, two sets  of filter  obscurity, opacity, and




 temperature measurements were acquired.   These  data are pre-




 sented  in Figures 4.4-1  and 4.4-2  and  Table 4.4.  The  times




 indicated on Figures 4.4-1  and 4.4-2 represent  the time interval




 since the beginning  of  the  push  during which  the sample was




 taken.  Evaluation of these data  indicates  that  the peak partic-




ulate  emission period was  approximately  three minutes. The




method of data analysis  is  described in  detail  in  Appendix  T




 (Volume 5).

-------
                               - 40 -

                            FIGURE 4.4-1

       SEQUENTIAL FILTER OBSCURITY  TEST  ON  FEBRUARY  24,  1975

                         Burns Harbor Plant
                    Bethlehem  Steel  Corporation
                        Chesterton,  Indiana
                          March 3-7,  1975
120-135 sec.
150-165 aec.
                        10
                     270-285 sec.
180-195 sec.
                        11
                   540-555 sec.
210-225 sec,
                       12
                  570-585 sec,
 650-665  sec,
                             Clayton  Environmental Consultants, Inc.

-------
                              -  41  -

                           FIGURE 4.4-2

        SEQUENTIAL FILTER OBSCURITY TEST ON MARCH 3, 1975

                        Burns  Harbor  Plant
                   Bethlehem Steel Corporation
                       Chesterton,  Indiana
                         March 3-7, 1975
                                        75-90 sec.
                               105-120  sec,
                              \
                                       195-210 sec.
                                            11
                             225-240 sec,
                                   12
                   285-300  sec.
           360-375 sec.
420-435 sec,
480-495 sec.
Clayton Environmental Consultants, Inc.

-------
                          -  42  -
                        TABLE 4,4

        DETERMINATION OF PEAK PARTIGULATE EMISSION
               PERIOD FROM THE EXHAUST DUCT

                    Burns Harbor Plant
               Bethlehem Steel Corporation
                    Chesterton,  Indiana
                      March  3-7,  1975
Test Number
1 (2-24-75 at
1345)











2 (3-3-75 at
1158)











Filter
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
1**
2
3
4
5
6
7
8
9
10
11
12
13
Start
Time
(sec)
0
30
60
90
120
150
180
210
240
270
540
570
650
0
45
75
105
135
165
195
225
255
285
360
420
480
Stack
Temp.
(°F)
*
*
*
*
*
*
*
*
*
*
*
*
*
75
120
145
125
105
100
95
83
80
78
78
78
78
Average
Plume
Opacity
(%)
15
50
50
50
50
40
25
20
20
20
15
15
15
20
80
60
40
25
20
20
20
20
20
20
20
20
 * No data acquired
** 30-second duration filter obscurity sample
                            Clayton  Environmental  Consultants, Inc

-------
                          -  43  -





4.5  Peak Particulate Emissions from Shed Exhaust Duct




     "Peak" particulate sampling was performed to determine




the individual contribution of pushing operation emissions to




the average, continuous particulate emission rate from the




uncontrolled coke-side shed.  Peak particulate samples were




acquired using EPA Method 2 (Final In-House Draft, 7-21-74) and




variations of EPA Methods 1 and 5  (Final In-House Draft, 7-18-74




and 7-25-74, respectively).  These methods are included in




Appendices P-R (Volume 5).




     The major modification to EPA Method 5 was that the peak




particulate emissions were measured only during the evacuation




of the shed during and immediately after coke pushing, i.e.,




for three minutes out of approximately 13 minutes.  Thus, each




peak particulate sample consisted of intermittently sampling



20 individual coke-oven pushes with the probe stationary during




each push. It should be noted that, in general, the peak and




continuous samples were acquired concurrently.




     Because the peak particulate emission tests were conducted




on an interruptible basis (i.e., sampling for three minutes and




then stopping for about 10 minutes, then sampling again), the




filter temperature was maintained just above the dewpoint of




the stack gases.




     The peak particulate samples were analyzed for the materials




indicated in Table 1.0-1.  All sampling and analytical methods




are presented in Appendix S (Volume 5).






4.6  Particle Size Distribution




     Particle size tests were performed to define the distri-




bution of particles entering a potential control device.  Since

-------
                          - 44 -





previous coke pushing studies have used the cascade  impact ion




method, the sizing determinations were performed using an




outside-the-stack Brink cascade impactor.  Each of the four




particle size samples were taken at a different single point,




each representing an equal area, (at an isokinetic rate)




because accurate calculation  of the aerodynamic diameter of




each impaction stage is impossible if the sampling rate varies




during sampling  (such as would  result from maintaining an




isokinetic rate while traversing a number of points).  In




addition, the particulate  concentration for each of  the four




particle size samples was  calculated for comparison  to the




filterable particulate concentrations resulting from 3-minute




peak particulate samples.  The  sampling and analytical methods




for particle sizing are presented in Appendix U (Volume 5).






4.7  Emissions of Other Materials




     A large number of additional emission measurements were




made during the course of  the study.  A complete roster of the




emissions measured is presented in Table 1.0-1, and  the purpose




for obtaining each of these measurements is specified in




Table 1.0-2.




     Sulfur dioxide and sulfuric acid mist (sulfur trioxide)




samples were acquired and analyzed according to U.S. EPA




Reference Method 8.  Hydrogen sulfide samples were acquired and




analyzed according to U.S. EPA Reference Method 11.  These




methods are included in Appendices V and W (Volume 5), respectively.




     Sampling and analytical methods for all other contaminants




measured during the study are not covered by EPA-standardized

-------
                          -  45 -





procedures.  Other standard methods, such as ASTM methods, were




used as much as possible.  All sampling procedures were reviewed




and accepted by Mr. Louis Paley, P.E., DSSE, U.S. EPA, prior to




their use.  Similarly, all analytical procedures were reviewed




and accepted by Mr. Mark Antell, DSSE, U.S. EPA, prior to their




use.  A complete description of each of these sampling and




analytical methods is given in Appendices X-BB (Volume 5).






4.8  Dustfall Measurements




     In order to assess the impact of the coke-side shed upon




particulate deposition, dustfall measurements were made using




the general principles outlined in ASTM Method D-1739-70.




Measurements were taken within and near the coke-side shed and




at similar locations on the unshedded (No. 2) coke battery, as




shown in Figure 3.1-1.  Dustfall sample buckets were located




at approximately equivalent positions on both benches about 13




feet above grade.  Near the shed wall (on the far side of the




quench car tracks), the dustfall buckets were elevated about 10




feet above grade on the far side of the railroad tracks.  Each




dustfall bucket was carefully positioned to avoid, as much as




possible, incineration from falling coke, deformation, or other




means of destruction.  In addition, eight pairs of dustfall




samples were acquired about three feet above grade to assess the




precision of the technique.




     A dustfall measurement made in the ambient air usually




requires a period of 30 days.  Because dustfall levels are




orders of magnitude higher in the immediate vicinity of a coke




battery than in ambient air, it was only necessary to expose

-------
                          -  46  -





the dustfall buckets for periods of hours in these tests.




     It should be noted that the settled dust in each can was




filtered during the laboratory analysis through a No. 18 mesh




screen (1-mm square) to remove large chunks of fallen coke.




In addition, selected dustfall samples were analyzed for pH,




acetone-soluble content, and cyclohexane-soluble content.  The




sampling and analytical techniques are summarized in Appendix




CC (Volume 5) .






4.9  Subjective and Visual Emission Parameters




     4.9.1   Degree of Greenness




             The semi-quantitative measurement scale used to




     estimate visually the relative quantity of particulate




     matter released during a coke-oven push is termed "degree




     of greenness."  In applying this technique, the duration




     of the coke-pushing operation was estimated and divided



     into thirds.  The amount of visible particulate generated




     during each third was estimated by mentally integrating




     the quantity of particulate generated and recording the




     value on a scale of one to four ("very little" emissions to




     "copious" amounts of emissions, respectively).  Each obser-



     vation represented the total obscurity caused by the emis-




     sions from both the falling coke and the coke in the quench




     car.  The resulting numbers for each third of each coke-oven




     push were then summed to give a semi-quantitative measure




     of emissions generated from each coke-oven push on a scale




     of three to 12.  In addition, the same observer recorded




     the actual duration of the coke-pushing operation with a

-------
                     -  47  -






stopwatch.  A second estimate of the amount of emissions




released, the product of the sum of the ratings and the




duration of the push, was also determined.  The detailed




method for determination of the degree of greenness of a




coke-oven push is given in Appendix DD (Volume 5).






4.9.2   Opacity of Shed Exhaust




        Although Figure 3.1-2 shows the shed exhaust duct




with one exit stack, in fact, two stacks were used to dis-




charge the emissions from the shed exhaust duct during the




study.  Further, a third exhaust stack existed but was




sealed completely during this study,,*' Opacity data were




acquired by U.S. EPA personnel for the two functional stacks




using EPA Method 9.  These values were averaged and there-




after treated mathematically as if there were only a single




stack.  A copy of the method is contained in Appendix EE




(Volume 5).






4.9.3   Percent of Doors Leaking




        Since door emissions appeared to be predominantly




independent of either coke-pushing or quench car  particu-




late emissions, an observation technique that recorded the




quantity of oven doors leaking visibly during a short-term




observation period was developed on-site.  These  observa-




tions yielded an estimate of the percent of coke-side doors




leaking.  Similar observations were made and results were




calculated for the push side of the coke battery  to document




general process conditions.  A detailed description of the




basic method is provided in Appendix FF (Volume 5).

-------
                          - 48 -





     4.9.4   Visual Estimates of Fugitive Emissions




             The sampling  program developed  to evaluate  the




     shed leakage rate  (fugitive particulate emissions)  included




     on-site evaluation of opacity and the subsequent use  of




     photographs of these  emissions.  Opacity observations




     were made by EPA personnel under several different  process,




     wind, and shed leak-rate conditions using EPA Method  9




     (see Appendix EE, Volume 5).  Twenty-seven observations




     of the shed end  leakage were made during randomly-selected,




     complete pushing cycles on four days.  The observations




     were made while  looking diagonally through the plume, as




     close as possible to  the point of emission (shed end).




             Both 16-mm black-and-white and color movies,  and




     35-mm color stills were taken randomly during the fugitive




     particulate sampling  periods and during other instances of




     end and topside  shed  leakage.  These photographs were used




     primarily to estimate the  cross-sectiona1 area (i.e.,




     height and width) of  the fugitive plume emanating from the




     ends and side of the  shed.






4.10 Fugitive Particulate Emissions from the Shed




     In addition to the opacity data and photographs discussed




in Section 4.9.4, the sampling  program for fugitive emissions




included the measurement of fugitive particulate concentrations




at three points of leakage from the shed.  Each sample was taken




during (visually determined) "peak" emission periods.  These




samples were collected using a  47-mm diameter glass-fiber  filter,




a calibrated limiting orifice,  and a leakless diaphragm  pump.

-------
                          - 49 -





A varie-axial anemometer and a stopwatch were used  to  estimate



the exhaust gas velocity.  The  sampling rate for  each  sample



was held constant using the critical  orifice in  the  sampling



train.  The critical orifice  was  sized  initially  to  the



average anticipated velocity  based upon preliminary  vane-axial



anemometer measurements made  in the various areas  of  fugitive



particulate emissions.  Due to  spatial and  temporal  variations



in velocity, however, these samples were acquired  anisokinetically.



A detailed description of  the sampling and  analytical  methods is



presented in Appendix GG  (Volume  5).



     These short-term (approximately  1-1/4  minutes)  fugitive



particulate measurements were then extrapolated  to a  continuous,



fugitive particulate emission rate estimate.  This was accomplished



by using the shed leak opacity  data (converted to  "attenuation


             (3")
coefficients1^ ') as the basis  for extrapolation  from  the  short-



term basis to a continuous basis.  This was possible  because  it



has been shown that mass emissions can be correlated with  the


                                       (2)
attenuation coefficient for coal  dust.  '   A full  description



of the technique is provided  in Appendix HH (Volume  5).





4.11 Calibration, Quality Assurance,  and Sampling  Integrity



     Chain-of-custody procedures  utilized during  this  study were



followed conscientiously.  Each sample was  uniquely  identified,



and at all times either one member of  the Clayton  test team was



with the samples or the samples were  locked securely  in  storage.



     Calibrations of all instruments  were performed  both prior



to and after the sampling  period.  The critical  orifices used



in the fugitive particulate sampling  were calibrated,  even




though no promulgated air  pollution regulations  in the United

-------
                          - 50 -






States required calibration of  these devices at  the  time  of




the study.  Finally, the sample handling and analysis  tech-




niques were approved by Mr. Mark Antell  (DSSE, U.S.  EPA)  after




consultation with Clayton  personnel.




     The methods used for  instrument calibration  in  this  study




are presented in Appendix  II  (Volume 5).  The calibration data




are given in Appendix JJ (Volume 5).  A  list of  the  samples




voided during the conduct  of  the study and the reasons  for void-




ing them are listed in Appendix KK  (Volume 5).   Field  sampling




data sheets are provided in Appendix LL  (Volume  5),  while




analytical data are presented in Appendix MM (Volume 5).   "Chain




of Custody" procedures are given in Appendix NN  (Volume 6),




and example calculations are  shown  in Appendix 00  (Volume 6).




Copies of all correspondence  with Bethlehem Steel  Corporation




are included in Appendix PP (Volume 6).  Results  of  previous




sampling by Bethlehem Steel Corporation at the Burns Harbor




plant are presented in Appendix QQ  (Volume 6).

-------
                     - 51 -


        5.0  PRESENTATION AND DISCUSSION OF RESULTS


5.1  Coke-Side Particulate Emissions

     Two types of particulate samples were collected in the

exhaust duct from the coke-side shed at the Burns Harbor plant.

The first, termed "continuous" particulate emissions, spanned

the entire period when pushing occurred, as well as the inter-

vals between pushes.  Sampling continued during these intervals

unless process upsets or downtime exceeded 30 minutes.

     The second type of sampling estimated "peak" particulate

emissions from the shed.  These samples were acquired by

sampling intermittently during the periods found to have maxi-

mum visible emissions, i.e., the 3-minute interval which

immediately followed the beginning of a push.  Since both

continuous and peak emissions included essentially the same

pushes, the two types of samples were, in a sense, simultaneous

     It should be noted that both types of samples necessarily

included quench car emissions while it was under the shed and

door leaks, occurring constantly, as well as residual emissions

from "old pushes."  Neither type, however, included the

emissions that were fugitive from the shed.  It was therefore

necessary to estimate these fugitive emissions by another

technique.


     5.1.1   Continuous Particulate Emissions from the Exhaust
             Duct

             The continuous particulate emissions from the shed

     exhaust duct are summarized in Table 5.1.1.  These values

     represent emissions from pushing, door leaks, quench car

-------
                                            TABLE 5.1.1

          SUMMARY OF CONTINUOUS PARTICULATE EMISSIONS  FROM THE BATTERY  NO.  1  EXHAUST DUCT

                                        Burns Harbor Plant
                                    Bethlehem Steel Corporation
                                        Chesterton, Indiana
                                          March  3-7,  1975
N)

I
Test
No

1
2
3
Average
Stack Gas
Conditions

Temp
(°F)
84
94
98
92
Flowrate
(DSCFM)
269,000
268,000
266,000
268,000
Particulate
Concentration
(gr/DSCF)*

Filter-
able
0.056
0.054
0.053
0.054
Total
0.058
0.055
0.056
0.056
Particulate
Emission Rate
(lbs/hr)*

Filter-
able
129
123
121
124
Total
134
127
127
129
Process Weight
Rate

tons wet
coal/hr
183
167
168
173
tons dry
coal/hr
170
156
156
161
Particulate Emission Factor*
Filterable

Ibs/ton
dry coal
0.76
0.79
0.78
0.78
Ibs/ton
coke +
0.97
1.0
0.99
0.99
Total

Ibs/ton
dry coal
0.79
0.81
0.81
0.80
Ibs/ton
coke +
1.0
1.0
1.0
1.0
* These values do not include fugitive particulate emissions  (see Table 5.1.3  for overall  emissions)

+ Bethlehem Steel Corporation has indicated that 0.73 ton of  coke is  produced  per  ton of
    wet coal charged.
                                                 Clayton Environmental  Consultants,  Inc.

-------
                - 53 -






movement, and the residual emissions from previous pushes,




but inherently do not include fugitive emissions from the




shed.  The emission factors for both filterable and total




particulate are relatively consistent among themselves




and average 0.78 and 0.80 pound of particulate per ton of




dry coal fed, or 0.99 and 1.0 pound of particulate per ton




of coke produced, respectively.  A more complete summary




of sampling times, sampled volumes, concentrations, and




emission rates can be found in Appendix RR (Volume 6).






5.1.2   Fugitive Particulate Emissions




        Just prior to commencement,,fpf the study, fugitive




particulate emissions were observed to be leaking from




the shed at the four positions shown in Figure 3.1-2:  1)




the north end of the shed,  2) the south end of the shed,




3) the Askania valves, and  4) the boundary between the




shed and the coke battery.  The Askania valves and shed-




battery boundary leakage were observed to be essentially




constant, while the ends leakage went through an increas-




ing-decreasing opacity cycle similar to that of the




exhaust duct emissions.  Because the primary goal of this




project was to measure the emissions from the coke side




of the battery and not just from the shed exhaust duct,




the fugitive particulate emissions escaping the shed were




estimated.  To accomplish this, a measurement technique




was developed in the field that included short-term




anisokinetic, fugitive particulate emission measurements.

-------
                - 54 -



        Table 5.1.2 summarizes the results  of  these


measurements.  The fugitive particulate  concentrations


ranged from 0.002 to 0.124 grain per dry  standard  cubic


foot  (gr/DSCF).  These concentrations were  extrapolated


to estimate the continuous fugitive particulate  emission

    (3)
rate    by the methodology presented in Appendix HH


(Volume 5).  The resulting estimate was  21.9 pounds  of


fugitive particulate per hour.  Based upon  an  average


feed rate during continuous particulate  sampling of  161


tons of dry coal per hour  (see Table 5.1.1), this  emission


rate corresponds to an emission factor of 0.14 pound of


filterable particulate per ton of dry coal  fed to  the


ovens, or 0.17 pound of filterable particulate  per  ton of


coke produced.


        Since these values estimate the fugitive  filterable


particulate emissions on a continuous basis, they  may be


related to the continuous filterable particulate emissions


from the exhaust duct, which averaged 0.78  pound per ton


of dry coal fed (see Section 5.1.1).  The percentage of


the emissions that were fugitive can then be calculated


as follows:
           °-14      -  100 = 157.
        0.14 + 0.78


Thus, approximately 15 percent of the filterable particu-


late emissions escaped from the shed on a continuous basis.



5.1.3   Overall Continuous Coke-Side Particulate Emissions


        Using the particulate emissions data presented in


the previous two sections, it is possible to estimate the

-------
             - 55 -

           TABLE  5.1.2

 MEASURED  FUGITIVE  PARTICULATE
EMISSIONS ESCAPING FROM THE SHED

       Burns Harbor Plant
   Bethlehem Steel  Corporation
       Chesterton,  Indiana
         March  3-7,  1975
1975
Date
3-5
3-5
3-5
3-5
3-6
3-6
3-5
3-6
3-5
3-6
3-6
3-6
3-5
3-5
Sample Description
Topside at crossover butterfly for ovens
112 & 113; Pushing oven 114
Topside at oven 112; Charging oven 114
Topside at oven 134; Charging oven 134
Topside; Off main at oven 134
North end; Pushing oven 105
North end; Pushing oven 123
North end; Pushing oven 124
North end; Pushing oven 125
North end; Pushing oven 126
North end; Pushing oven 183
South end; Pushing oven 113
South end; Pushing oven 181
South end; Pushing oven 184
South end; Pushing oven 191
Fugitive
Particulate
Concentration
(gr/DSCF)
0.105
0.002
0.029
0.077
0.124
0.006
0.117
0.022
0.053
0.024
0.017
0.051
0.002
<0.006
Velocity
(ft/min)
_
—
160
151
276
230
—
201
353
221
337
409
171
312
           Clayton Environmental Consultants, Inc.

-------
                 -  56 -






overall particulate  emissions emanating  from  the  coke




side of the shed on  a continuous basis.   Table  5.1.3  pre-




sents both the continuous filterable particulate  emissions,




which do not include  fugitive emissions,  and  the  overall




filterable particulate emissions.  These  overall  emissions




include emissions  from pushing, door leaks, quench  car




movement under the shed, the residual  from previous pushes,




and fugitive emissions from the shed,  but do  not  include




emissions from the quench car during transit  outside  the




shed .




        The overall  emission rates presented  in Table




5.1.3 have been  estimated  by adding the  average  continu-




ous fugitive emission rate developed in Section 5.1.2 to




the continuous emission rate for each  particulate sample.




Emission factors for  estimated fugitive emissions were




determined by dividing this average emission  rate by  the




coal feed rates for  the individual samples presented  pre-



viously in Table 5.1.1.  The overall emission factors were



then estimated to range from 0.89 to 0.93, and average 0.91




pound  of filterable particulate per ton of dry coal fed, or




1.2 pound of filterable particulate per ton of coke produced,






5.1.4   Peak Particulate Emissions from the Shed




        Table 5.1.4  presents the peak  particulate emissions




measured in the shed  exhaust duct.  These peak  emissions




were quantified by sampling during only the period  of




greatest visible emissions, i.e., the  first three minutes




of the approximately  13-minute interval when  pushing

-------
                           TABLE  5.1.3

SUMMARY OF OVERALL CONTINUOUS PARTICULATE EMISSIONS FROM  THE  SHED

                       Burns Harbor Plant
                   Bethlehem Steel  Corporation
                       Chesterton,  Indiana
                         March  3-7,  1975

Continuous
Particulate
Test
No.



1
2
3
Average
Continuous Filterable
Particulate Emissions


Emission
Ra t- 1*
(lbs/hr)


129
123
121
124
Emission
Factor

Ibs/ton
dry coal

0.76
0.79
0.78
0.78
Ibs/ton
coke

0.97
1.0
0.99
0.99
Estimated Fugitive
Particulate Emissions


Emission
T?a t-*>
(lbs/hr)


21.9
21.9
21.9
21.9
Emission
Factor

Ibs/ton
dry coal

0.13
0.14
0.14
0.14
Ibs/ton
coke

0.17
0.18
0.17
0.17
Overall Continuous
Filterable Particulate
Emissions Estimate


Emission
Rate
(lbs/hr)


151
145
143
146
Emission
Factor

Ibs/ton
dry coal

0.89
0.93
0.92
0.91
Ibs/ton
coke
c
1.1
1.2
1.2
1.2
                                Clayton Environmental  Consultants,  Inc.

-------
                                            TABLE 5.1.4

             SUMMARY OF PEAK PARTICULATE EMISSIONS FROM THE BATTERY  NO.  1  EXHAUST DUCT

                                        Burns Harbor Plant
                                    Bethlehem Steel  Corporation
                                        Chesterton,  Indiana
                                          March  3-7,  1975
Ul
00
Test
No.
1
2
3
Average
Stack Gas
Conditions
Temp
(°F)
113
128
131
124
Flowrate
(DSCFM)
257,000
262,000
251,000
257,000
Particulate
Concentration
(gr/DSCF)*
Filter-
able
0.148
0.221
0.195
0.188
Total
0.162
0.230
0.219
0.204
Particulate
Emission Rate*
(lbs/hr)+
Filter-
able
73.4
112
94.2
93.2
Total
80.5
116
106
101
Process Weight
Rate +
tons wet
coal/hr
158
157
157
157
tons dry
coal/hr
147
146
146
146
Particulate Emission Factor*
Filterable
Ibs/ton
dry coal
0.50
0.77
0.65
0.64
Ibs/ton
coke;4
0.64
0.98
0.82
0.81
Total
Ibs/ton
dry coal
0.55
0.79
0.73
0.69
Ibs/ton
coke^
0.70
1.0
0.92
0.87
* These values do not include fugitive particulate emissions.

+ Emission rates and process weight rates assume typical operations; i.e., 4.5  pushes/hour.


£ Bethlehem Steel Corporation has indicated that 0.73 ton of coke is produced per  ton  of  wet
    coal charged.
                                                 Clayton Environmental  Consultants,  Inc.

-------
               - 59 -






emissions were being evacuated from the shed.  This




3-minute sampling period was determined empirically by




the methodology discussed in Section 4.4.  It should be




noted that because of the sampling technique, the




emission rates and feed rates presented in Table 5.1.4




have been adjusted to assume typical operations; i.e.,




4.5 pushes  per  hour.   The resulting emission factors




for filterable and total particulate average 0.64 and




0.69 pound of particulate per ton of dry coal fed, or




0.81 and 0.87 pound of particulate per ton of coke pro-




duced, respectively.




        In addition to pushing emissions, emissions from




door leaks, quench car movement, and residual concentra-




tions from previous pushes were also observed to be




exhausted during the 3-minute "peak" periods.  Thus, the




peak particulate emissions data presented in Table 5.1.4




should not be considered an estimate of the particulate




emissions from coke pushing, per se.  These additional,




variable sources of particulate emissions (doors and




quench car) likely account for the wider range of data




reported for the peak particulate emissions from the




exhaust duct as compared to the results of continuous




sampling.  A more complete summary of sampling times,




sampled volumes, concentrations, and emission rates can




be found in Appendix SS (Volume 6).







5.1.5   Particulate Emissions for Pushing Operations




        Using the particulate emissions data presented




previously, together with a straightforward calculational

-------
                        - 60 -





procedure, it is possible to estimate the particulate emis-



sions due to pushing operations alone at Battery No. 1,  By



assuming that the aggregate door leak emission rate within



the shed is essentially equivalent during pushing periods and



non-pushing periods, the emission rate due to door leaks can



be estimated algebraically by consideration of emission rate



measurements from the two particulate sampling modes.



       Referring to Figure 5.1.5, the total shaded area B^



(diagonal lines upward and to the right) includes a single



push and the time following that push and preceding the next



push, and represents the mass of emissions occurring during



the continuous particulate sampling period tc., consisting of



door leaks, residuals of old pushes, and push emissions occur-



ring during tc..  Area A^ (diagonal lines upward and to the



left) represents the mass of push emissions, residual emis-



sions, and door leaks measured during the peak particulate



sampling period, tp..  The time period tp-  represents the



time required to evacuate the shed of most of the "current



push emissions."  The difference of the two areas, B^ - A^, is



the rectangular area that represents the mass of emissions due



only to door leaks and residual concentrations  from previous



pushes.  A time-weighted fraction of this rectangle, (B^ - A^) *



tPt/ (CC4 ~ fcp')^ represents the "baseline" rectangular area's



contribution to the total peak emissions, A^.  Subtraction



then gives an estimate of the mass emissions ascribed only to



pushing:



       Push-Only Mass Emissions = Ai - (Bj -
T    'PI
 L 'ci-'pi

-------
Test Type
Peak
Continuous
Duration
Sampling Time = P^ = PI+I
Sampling Time =2 ci
Emissions
Collected
Ai * Ai+l
SBi
Symbol
^
w//>
                                       •i
Emission
  Rate
                        Residual Current Push
                                        -^Residual Old, Pushes/
Constant Door -I
                                                       Leaks
                                                                                                Residual
                                                                                                 Current
   Push
                                        Time
                                                                            Push
                                               FIGURE 5.1.5
                                  SCHEMATIC DIAGRAM OF SAMPLING SCHEDULE
                                            Burns Harbor Plant
                                        Bethlehem Steel Corporation
                                            Chesterton, Indiana
                                              March 3-7, 1975
                                                          Clayton Environmental Consultants,  Inc.

-------
                         - 62 -






        The  calculational procedure described above is shown




 in  Table  5.1.5.   Since all samples were taken on a multi-push




 basis,  measured  emissions were normalized to a "per-push"




 basis  by  dividing by the number of pushes included in each




 sample.   Then, using the sampling time per push and the equa-




 tion above,  the  pushing operations emissions captured by the




 shed were estimated.




        Fugitive  emissions which escaped capture by the shed




 were estimated,  as discussed in Section 5.1.2, to be 15 per-




 cent  (on  a  continuous  basis) of the emissions captured by the




 shed.   Thus,  in  order  to determine total push emissions, the




 emissions captured by  the shed must be adjusted upward by di-




 rect ratio  using this  factor.   Using the process weight rates,




 the overall  filterable particulate emission factors for the




 pushing operations can also  be calculated.   These values were




 found  to  range from 0.48 to  0.89,  and  average 0.69 pound per




 ton of dry  coal  fed to the ovens,  or 0.87 pound per ton of




 coke produced.






 5.1.6  Particulate Emissions for  Non-Pushing Operations




       Using  the  evaluations presented in Sections 5.1.3 and




 5.1.5, the  emissions from non-pushing  operations are calcu-




 lated by  difference, as  shown  in  Table 5.1.6.   These emission




 factors include  door leaks,  residuals  from past pushes,  and




 emissions from quench  car movement.   They have been corrected




 for fugitive  emissions  and thus provide  an  estimate of the




 overall coke-side  emissions  for non-pushing operations.   The




 values range  from  0.04  to 0.41, and  average 0.22 pound of




filterable particulate per ton of  dry  coal,  or 0.3 pound per

-------
                                            TABLE 5.1.5

                          CALCULATION OF FILTERABLE  PARTICULATE  EMISSION
                                   FACTOR FOR PUSHING OPERATIONS

                                        Burns Harbor  Plant
                                    Bethlehem Steel Corporation
                                        Chesterton, Indiana
                                          March  3-7,  1975
Type
of
Test

Continuous


Peak

Test
No.
1
2
3
1
2
3
Filterable
Particulate
Emission
Rate
(Ibs/hr)
129
123
121
73.4*
112 *
94.2*
Sampling
Time
(Minutes)
288
288
288
60
60
60
Number
of
Ovens
Pushed
25
23
23
(4.5)
(4.5)
(4.5)
Filterable
Particulate
Emissions
(Ibs/push)
24.8
25.7
25.3
' 16.3
24.9
20.9
Sampling
Time Per
Push
(Minutes)
11.5
12.5
12.5
3
3
3
                                                                                                   Co

                                                                                                   I
Peak
Particulate
Test No.
1
2
3
Average
Push
Emissions
Captured
by the Shed
(Ibs/push)
13.3
24.6
19.5
19.1
Total
Push
Emissions
(Ibs/push)
15.6
28.9
22.9
22.5
Total
Push
Emissions
(Ibs/hr)*
70.2
130
103
101
Process Weight
Rate*
tons wet
coal/hr
158
157
157
157
tons dry
coal/hr
147
146
146
146
Filterable Particulate
Emission Factor for
Pushing Operations
Ibs/ton
dry coal
0.48
0.89
0.71
0.69
Ibs/ton
coke +
0.61
1.1
0.90
0.87
* These emission rates and process weight rates assume typical operations;  i.e.,  4.5  pushes/hour.
+ Bethlehem Steel Corporation has indicated that 0.73 ton of  coke  is produced  per ton of  wet  coal

                                                    Clayton  Environmental  Consultants, Inc.
charged.

-------
                     TABLE 5.1.6
CALCULATION OF FILTERABLE PARTICULATE EMISSION FACTOR
              FOR  NON-PUSHING  OPERATIONS

                  Burns  Harbor Plant
             Bethlehem Steel Corporation
                 Chesterton, Indiana
                   March 3-7,  1975
Continuous
Particulate
Test
No.
1
2
3
Average
Overall Continuous
Filterable Particulate
Emission Factor
Ibs/ton
dry coal
0.89
0.93
0.92
0.91
Ibs/ton
coke
1.1
1.2
1.2
1.2
Filterable Particulate
Emission Factor for
Pushing Operations
Ibs/ton
dry coal
0.48
0.89
0.71
0.69
Ibs/ton
coke
0.61
1.1
0.90
0.87
Filterable Particulate
Emission Factor for
Non-Pushing Operations
Ibs/ton
dry coal
0.41
0.04
0.21
0.22
Ibs/ton
coke
0.5
0.1
0.3
0.3
                            Clayton  Environmental Consultants, Inc.

-------
                        - 65 -









     ton of coke produced.  This table thus indicates that  the




     pushing operations account for 76 percent of the overall




     emissions, while the non-pushing operations account  for




     24 percent.






5.2  Particulate Capture Efficiency of the Shed




     5.2.1  Evaluation of Shed Capture Efficiency




            The average emission factor for continuous  filterable




     particulate emissions from the exhaust duct was 0.78 pound




     per ton of dry coal fed to the ovens, as shown in  Table




     5.1.1.  Using the average fugitive emission factor developed




     in Section 5.1.2, 0.14 pound per ton of dry coal,  the  par-




     ticulate capture efficiency of the shed may be calculated




     as follows:






            [j.  -  0.147(0.14 + 0.78)] X 100 = 85%.






     Thus, on a continuous basis, an average of 85 percent  of  the




     filterable particulate emissions are captured by the shed.






     5.2.2  Possible Causes of Leakage




            The following are possible reasons for the  shed's




     fugitive particulate emissions:




            1.  The overall size of the shed's holding  chamber




                (see Figures 3.1-2 and 3.1-3) appeared  to be




                too small relative to the magnitude of  the  emis-




                sions and the effective evacuation rate of  the




                shed.  This was substantiated by the exhaust




                duct opacity observations, which documented  that




                the shed was not completely cleared of  push

-------
         - 66 -







emissions in 2 to 3 minutes (as designed, per




Mr. Robert Harvey of Bethlehem Steel Corporation)




Instead it appeared that clearing of the push




(peak) emissions sometimes took longer than 10




minutes (perhaps 14 to 15 minutes).  This is




important for two reasons:




a.  The shed's holding and evacuation capacities




    may have been exceeded in the many (32 per-




    cent) instances when the time between pushes




    was only 8 to 9 minutes; i.e., below the




    "average" cycle duration of 13 minutes.




    This meant an 8- to 9-minute push-to-push




    interval was slightly below the "observed"




    period required for push emissions clearing.




    Shed leakage was likely also increased by the




    (not infrequent) occurrence of highly emis-




    sive ("green") coke-oven pushes.  When the



    shed's  capacities  were exceeded,  the  particu-




    late  emissions "overflowed"  from any  openings




    below the shed's holding chamber.   In this




    event,some  of the  particulate  material  sus-




    pended  in the shed's  holding chamber  likely




    moved  beyond  the "capture"  range  of  the  ex-




    haust  duct  and into the  region where  wind




    effects  were  more  pronounced.   The probable




    result  of this "undersizing" was  leakage




    from  any  shed openings  such  as  those  located




    at both  ends  of  the shed and topside.

-------
             -  67  -






    b.  The hot, particulate-laden gases may have




        also dropped beyond the reach of the exhaust




        duct because of cooling caused by the attempt




        to hold the  emissions in the holding chamber




        beyond  the design duration.






2.  As implied  in the first item, perhaps the shed




    exhaust rate was too low.  During the test period}




    exhaust gas flowrate measurements indicated that




    airflow was about 10 percent below the rate identi-




    fied by the facility to be its optimum rate on the




    basis of emission clearing time.  However, assur-




    ances were  given by Bethlehem Steel Corporation




    that the exhaust rate was at maximum conditions




    and,since the stack gas exhaust rate was within  the




    jf10 percent criterion  discussed  in  Section 3.2.1,




    sampling commenced.



3.  It is possible that "short circuiting" of the




    outside air (which enters the shed through its




    ends and open side) to the exhaust duct occurred




    due to the:  (1) magnitude of the openings,




    particularly at  the ends of the shed; and (2)




    the varying cross-sectional area of the openings.




    If this happened, the "actual" emission exhaust




    rate would  be reduced.  This phenomenon could




    have further reduced the shed's performance




    because the resulting inlet airflow pattern would




    disturb, rather  than enhance, the desired pattern




    of airflow  in the shed.

-------
             - 68 -









4.  The capture problem may have been caused, at




    least in part, by the holding chamber, its  inlet,




    and/or the exhaust duct.




    a.  The shape of the shed's holding chamber, in




        conjunction with the size of its inlet




        ("throat") might have affected the shed's




        initial emission capture efficiency and sub-




        sequent holding capacity.




    b.  The shape, size, and location of the exhaust




        duct (located at the top of the holding




        chamber) may have significantly affected the




        rate and efficiency of its emission exhaust.






5.  The shed wall and end openings may have further




    affected the performance in three ways:




    a.  Such openings provided potential escape routes




        for fugitive emissions.




    b.  At least some of the openings appeared to




        allow the wind to interact with the emissions



        within the shed.  Several instances were ob-




        served when the wind blew coke-side door leaks




        directly out the end of the shed, before they




        were captured.




    c.  The relatively large end openings may not




        have permitted optimum use of the inlet air.




        Ideally, the air should have entered the




        shed uniformly, and preferably only along




        its open side.  This would have enhanced the

-------
                         - 69 -






                    entrainment of  the  particulate  emissions  by




                    reinforcing the  spiral air  pattern  being




                    established in  the  shed  by  the  combined  effect




                    of the rising hot emission  gases  and  the  shape




                    of the shed holding chamber.





5.3  Chemical Composition of Particulate Emissions




     Nineteen separate analyses of  the  particulate  samples  (both con-




tinuous and peak) were performed.   Table 5.3-1  presents  the  results




of the analyses  for sulfate and 10  metal ions in  terms  of percent of




both filterable  and total particulate weight.   The  contribution  of




these substances to the particulate  emissions was quite  small.   The




only substance found  in an amount greater than  one  percent was  sul-




fate (2.3 to 4.5 percent).  Thus, carbonaceous  material  (coke)  con-




stituted the majority of the particulate matter captured.




     Table 5.3-2 presents the average emission  rates  for  these  11




substances as well as the average emissions  of  acetone-soluble  par-




ticulate, water-soluble particulate, and other  water-soluble




substances.  Note that in this  table emission rates  for  peak  samples




have been adjusted to an average pushing rate of  4.5  pushes  per  hour




The acetone-soluble content of  the  filterable continuous  and  peak




emissions averaged 13 and 12 percent, respectively.   The  water-




soluble content  averaged two percent for both types  of  samples.




     Table 5.3-3 presents the pH and acidity of the  front and back-



half catches of  each  of the particulate samples.  Although  the




values vary, all fractions were found to be  acidic.





5.4  Particle Size Distribution




     The results of the four particle size samples are  plotted  in




Figure  5.4.   The lines are labeled according to the point within

-------
                             TABLE 5.3-1

SUMMARY OF METALS AND SULFATE CONTENT OF PARTICULATE SAMPLES  (PERCENT)

                         Burns Harbor  Plant
                    Bethlehem Steel Corporation
                        Chesterton, Indiana
                          March 3-7,  1975
Sampling
Conditions

Continuous


Peak

Test
No.
1
2
3
1
2
3
Portion
of
Sampling
Train
Front
Total
Front
Total
Front
Total
Front
Total
Front
Total
Front
Total
Percent of Particulate Weight
Ca
0.01
0.02
0.03
0.04
0.02
0.03
0.03
0.06
0.01
0.02
0.005
0.02
Fe
0.9
0.9
0.9
0.9
0.5
0.5
0.8
0.7
0.8
0.8
0.8
0.7
Mg
0.01
0.01
0.007
0.008
0.01
0.01
0.02
0.02
0.01
0.01
0.001
0.003
Pb
0.01
0.01
0.004
0.004
0.006
0.006
0.005
0.005
0.003
0.003
0.004
0.004
Al
0.6
0.6
0.2
0.2
0.6
0.5
0.4
0.4
0.2
0.1
0.3
0.3
Cd
0.0004
0.0004
0.0002
0.0002
0.0002
0.0004
0.0008
0.0007
0.0008
0.0009
0.0003
0.0004
Cu
0.02
0.02
0.002
0.002
0.002
0.002
0.0006
0.0005
0.003
0.003
<0.0003
0.001
Be
<0.0006
<0.0005
<0.0006
<0.0007
<0.0006
<0.0007
<0.001
<0.001
<0.0007
<0.0008
<0.0008
<0.0008
Se
0.002
0.002
<0.0008
<0.002
<0.002
<0.002
<0.003
<0.004
<0.002
<0.003
<0.002
0.004
Ti
0.1
0.09
0.04
0.04
0.04
0.03
0.04
0.04
0.06
0.06
0.05
0.04
SO^
2.8
4.5
3.3
4.2
4.1
4.5
3.1
4.4
2.3
2.5
2.5
2.7
                                            Clayton Environmental Consultants, Inc.

-------
                                               TABLE 5.3-2

                   SUMMARY  OF  AVERAGE RA-TJES- OF PAR-TICULATE" EMISS IONS FROM" EXHAUST DUCT

                                          Burns Harbor Plant
                                      Bethlehem Steel Corporation
                                          Chesterton, Indiana
                                            March 3-7, 1975
Material
Filterable Particulate
Total Particulate
Front-half Acetone Solubles
Total Acetone Solubles
Front-half Water Solubles
Total Water Solubles
Front-half Water-Soluble Arsenic
Total Water-Soluble Arsenic
Front-half Wa ter-Soluble Chloride
Total Water-Soluble Chloride
Front-half Water-Soluble Simple Cyanide
Total Water-Soluble Simple Cyanide
Front-half Water-Soluble Mercury
Total Water-Soluble Mercury
Front-half Calcium
Total Calcium
Front-half Iron
Total Iron
Average Emission Rate
Continuous Emissions
Ibs/hr
124
129
16.6
18.5
1.9
5.2
<0.004
0.003-0.006
2.8
3.0
<0.003
<0.005
<0.001
<0.002
0.03
0.04
1.0
1.0
Rgs/hr
56.3
58.5
7.5
8.4
0.91
2.4
<0.002
0.001-0.003
1.3
1.4
<0.001
<0.002
<0.0007
<0.001
0.01
0.02.
0.5
0.5
Peak Emissions*
Ibs/hr
93.2
101
10.8 .
17.6
1.4
2.7
0.001-0.002
0.001-0.003
1.5
2.4
0.0004-0.001
0.0004-0.003
<0.0008
<0.001
0.02
0.03
0.7
0.7
kgs/hr
42.2
45.8
4.9
8.0
0.62
1.2
0.0004-0.0009
0.0004-0.001
0.68
1.1
0.0002-0.0008
0.0002-0.001
<0.0003
<0.0006
0.007
0.01
0.3
0.3
* All data converted to typical  operations;
  i.e., 4.5 pushes/hour
Clayton Environmental  Consultants,  Inc;

-------
                                         TABLE 5.3-2 (continued)

                   SUMMARY OF  AVERAGE RATES  OF PARTICULATE EMISSIONS FROM "EXHAUST DUCT

                                          Burns Harbor Plant
                                       Bethlehem Steel Corporation
                                          Chesterton, Indiana
                                            March 3-7, 1975
Material
Front-half Magnesium
Total Magnesium
Front-half Lead
Total Lead
Front-half Aluminum
Total Aluminum
Front-half Cadmium
Total Cadmium
Front-half Copper
Total Copper
Front-half Beryllium
Total Beryllium
Front-half Selenium
Total Selenium
Front-half Titanium
Total Titanium
Front-half Sulfate
Total Sulfate
Average Emission Rate
Continuous Emissions
Ibs/hr
0.01
0.01
0.007
0.007
0.6
0.6
0.0004
0.0005
0.01
0.01
<0.0007
<0.0008
0.001-0.002
0.001-0.003
0.06
0.06
4.2
5.7
kgs/hr
0.006
0.006
0.003
0.003
0.3
0.3
0.0002
0.0002
0.005
0.005
<0.0003
<0.0004
0.0005-0.0009
0.0005-0.001
0.03
0.03
1.9
2.6
Peak Emissions *
Ibs/hr
0.01
0.01
0.004
0.004
0.3
0.3
0.0006
0.0008
0.0009-0.001
0.001
<0.0007
<0.0009
<0.002
0.001-0.003
0.04
0.04
2.4
3.1
kgs/hr
0.005
0.005
0.002
0.002
0.1
0.1
0.0003
0.0003
0.0004
0.0006
<0.0003
<0.0004
<0.0009
0.0007-0.001
0.02
0.02
1.1
1.4
* All data converted to typical operations;
  i.e., 4.5 pushes/hour
Clayton Environmental  Consultants,  Inc.s

-------
                         - 73 -
                       TABLE 5.3-3

              SUMMARY OF WATER SOLUBLE  pH
     AND ACIDITY/ALKALINITY OF PARTICULATE  SAMPLES

                   Burns Harbor Plant
               Bethlehem Steel Corporation
                  Chesterton, Indiana
                    March 3-7, 1975
Sampl ing
Cond itions

Continuous


Pushes Only

Test
No.
.; 1
2
2
3
3
1
1
2
2
3
3
Portion
of
Sampling
Train
Front
Back
Front
Back
Front
Back
Front
Back
Front
Back
Front
Back
PH
3.0
4.9
6.6
4.2
3.1
3.8
5.8
4.5
4.4
4.3
2.9
4.3
Acidity*
2.2
0.6
<0.3
0.2
1.3
0.3
<0.3
0.4
0.2
0.5
0.6
0.4
* Total milliequivalents of NaOH added to attain a pH  of  7.0
                            Clayton Environmental Consultants, Inc.

-------
Effective
Particle
Diameter
(microns)

                     Point 1 •
                     Point 2 A
                     Point 3 X
                     Point 4
         FIGURE 5.4
PARTICLE SIZE DISTRIBUTIONS
      IN EXHAUST  DUCT
     Burns  Harbor Plant
Bethlehem  Steel  Corporation
    Chesterton,  Indiana
      March 3-7.  1975
See Figure 4.2-3
 for  location of
 points in duct
                 
-------
                         - 75 -






the exhaust duct at which the  sample was obtained, shown previously




in Figure 4.2-3.  The weights  for each stage and the  individual




distributions are presented  in Appendix TT  (Volume 6).  These




data show that the particle  size distribution varies  somewhat




across the duct, probably due  to the changes in direction of the




exhaust gas flow within the  duct.  However, a statistical compari-




son (chi-square test for  independence) of the percentage of parti-




culate less than one micron  and the percentage less than five




microns shows no statistically significant  difference among the




four particle size distributions.




     The data in Figure 5.4  indicate that between 12  and 62 percent




(by weight) of the filterable  particulate was smaller than seven




microns (aerodynamic diameter).  Further, between four and 17




percent of the filterable particulate was smaller than one micron




(aerodynamic diameter).   The average stack-gas particle-size




distribution can be estimated  by averaging  the distributions in




Figure 5.4.  Based on this procedure, about 32 percent of the




filterable particulate was smaller than seven microns, and about




seven percent was smaller than one micron.




     The concentration of filterable particulate matter was also




calculated for each of the particle size samples.  The results,




displayed in Table 5.4, indicate a range from 0.105 to 0.260 gr/DSCF




Since the sampling period for  each of these tests included a single




push and ranged from eight to  12 minutes, these concentrations




should fall (on an average basis) between those obtained during




continuous and peak particulate sampling, which they  do.  Thus,




these results compare favorably with those  obtained during the




particulate tests.

-------
                   -  76  -
                  TABLE 5.4

PARTICULATE CONCENTRATION AND ACETONE-SOLUBLE
      CONTENT OF  PARTICLE  SIZE  SAMPLES

             Burns Harbor  Plant
         Bethlehem Steel Corporation
             Chesterton, Indiana
               March 3-7, 1975
Sampling
Point
Number
P-l
P-2
P-3
P-4
1975
Date
3-6
3-6
3-7
3-6
Sampling
Period
Start
09:05
18:26
08:27
12:59
Stop
09:13
18:38
08:35
13:07
Particulate
Concentration
(gr/DSCF)
0.260
0.142
0.105
0.147
Percent
Acetone
Solubles
48
63
13
34
                      Clayton  Environmental  Consultants, Inc

-------
                         - 77 -






     Table 5.4 also indicates the acetone-soluble content for each




of the samples.  These values vary greatly, ranging from 13 to 63




percent.  In addition, the acetone-soluble content was determined




for three fractions of each test:  the particulate matter collected




in the cyclone, that collected on the five stages of the impactor,




and that collected on the back-up filter.  This breakdown of the




acetone-soluble content by size ranges is given in the tables




included in Appendix TT (Volume 6).  A statistical test  (one-way




analysis of variance) indicates that there is no statistically




significant difference between the mean acetone-soluble  content




of these various siz,e ranges.






5.5  Emissions of Other Materials




     Table 5.5 presents the average emissions for all contaminants




other than particulate matter.  The following substances were




measured at levels exceeding 100 pounds per hour:  cyclohexane




insolubles (203 Ibs/hr), cyclohexane solubles (291 Ibs/hr) , ethylene




and homologues (147 Ibs/hr), and total light hydrocarbons (131




Ibs/hr).  Other contaminants were detected at levels that averaged




less than 16 pounds per hour:  acetylene, ammonia, benzene, ben-




zene homologues , benzo (a+e)pyrene , p-naphtylamine, carbon monoxide,




soluble chloride, chrysene plus triphenylene plus 1,2-benzanthracene ,




complex soluble cyanide, insoluble cyanide, simple soluble cyanide,




fluorafcthene, hydrogen sulfide, methane and homologues, nitrate




plus nitrite, total insoluble phenolics,  total  soluble  phenolics,




pyrene,  pyridine,  insoluble  sulfate,  total  sulfate,  total  sulfite,




sulfurdioxide, and  sulfuric  acid mist.   The  sampled  volumes  and




sampling  conditions,  as well  as  concentrations  and  emission  rates




for individual  tests,  are  presented  in Appendix  UU  (Volume  6).

-------
                            - 78 -

                          TABLE  5,5

   SUMMARY OF AVERAGE EMISSION RATES OF  "OTHER" EMISSIONS

                       Burns Harbor Plant
                  Bethlehem  Steel Corporation
                      Chesterton, Indiana
                        March 3-7,  1975

Acetylene
Ammonia
Benzene
Benzene Homologues (as C6Hg)
Benzo(a+e)Pyrene
Be ta- Naphthy lamine
Carbon Monoxide
Soluble Chloride
Chrysene + Triphenylene + 1,2-
Benzanthracene (as Chrysene)
Complex Soluble Cyanide
Insoluble Cyanide
Simple Soluble Cyanide
Cyclohexane Insolubles
Cyclohexane Solubles
Ethylene & Homologues (as C2H4)
Fluoranthene
Total Light Hydrocarbons (as CH^)
Hydrogen Sulfide
Methane & Homologues (as CH4)
Nitrate + Nitrite (as NO^)
Total Insoluble Phenolics (as
C6H50H)
Total Soluble Phenolics (as
C6H5OH)
Pyrene
Pyridine
Insoluble Sulfate
Total Sulfate
Total Sulfite
Sulfur Dioxide
Sulfuric Acid Mist (as 803)
Average Emission Rate
Ibs/hr
0.4-0.5*
0.34-0.44
4.1
<1.7
0.9-1.2
<0.35
6.9*
4.6

0.8-1.3
0.03
0.01
0.03
203
291
147 *
0.7-1.2
131 *
0.93
5.8*
0.33-0.40

<0.06

0.89
; <0.86
<0.15
<0.13
15.7
6.2
12.6
2.2
kgs/hr
0.2*
0.16-0.20
1.9
<0.77
0.4-0.5
<0.16
3.2*
2.1

0.4-0.6
0.01
0.004-0.005
0.01
92
132
67*
0.3-0.6
60*
0.42
2.7*
0.15-0.18

<0.03

0.40
<0.39
<0.07
<0.05
7.2
2.8
5.7
1.0
* Emissions measured during peak periods. These data have been
  converted to typical operations;  i.e., 4.5 pushes/hour.  All
  other samples were taken on a
continuous basis.
                         Clayton Environmental Consultants,  Inc.

-------
                          - 79 -
     The sum of the average cyclohexane-soluble and  cyclohexane-




insoluble emissions (termed category  "1")  should be  comparable  to




the sum of the average emission rates  for  total particulate  and




organic mists and gases  (termed category "2").  This  is  to be




expected because the sampling and analytical procedures  tend to




indicate that the materials captured,  measured, and  classified




by the two procedures should have approximately the  same aggregate




value.  A comparison of  these two categories does indicate approxi-




mately the same emission rate; category "2" is about  86  percent




of category "1."  Thus,  the two individual contaminant categories




 "cyclohexane  solubles"  and "cyclohexane  insolubles"  should  be




 considered  a  single  category  to  be  compared  with  the sum of  the




 total  particulate  and  organic materials  emissions,  and not  as




 separate  emissions.




      It  should  be  noted  that, in the  field,  low ambient  temperatures




 caused freeze-up  of  the  (standard)  impinger  solution containing




 cyclohexane;  midget  impingers (where  temperature  could be  controlled)




 were  substituted,  but  sampling  rates  were  below isokinetic  con-




 ditions.   Because  the  ambient temperature  precluded  isokinetic




 sampling  for  cyclohexane solubles,  cyclohexane insolubles,  fluoran-




 thene, pyrene,  chrysene  plus  triphenylene  plus 1,2-benzanthracene,




 and  benzo(a+e)pyrene,  the  reported  results may be  somewhat  high




 for  these  contaminants.   In retrospect,  this  alteration in  field




 sampling  was  favorable  with respect to the measurement of  fluoran-




 thene, pyrene,  chrysene  plus  triphenylene  plus 1,2-benzanthracene,




 and  benzo(a+e)pyrene  because  these  contaminants may otherwise  have




 been found  to  be  below limits of analytical  detection.

-------
                         -  80  -
5.6  Indices of Visible Emissions




     5.6.1  Degree of Greenness




            During the design of the study, it became necessary to




     develop a semi-quantitative measurement scale to document the




     relative degree of visible particulate emissions generated by




     the coke ovens during pushing.  The method and measurement




     scale formulated to characterize an observer's estimate of




     the emissions (visible obscuration) generated from a single




     coke-oven push (coke fall and quench car movement) incorporates




     the term "degree of greenness," a term used widely in the steel




     industry as a subjective assessment of the appearance of visi-




     ble emissions generated from coke-oven pushing.




            Specific applications of this index are demonstrated




     in Tables 5.6.1-1 through 5.6.1-6, which present the degree-




     of-greenness ratings, the sum of the ratings, and the product




     of the sum of the ratings and the duration of the push for




     each individual push during both the continuous and peak




     particulate samples.  The average product of the sum of the




     ratings and the duration of the push for the six tests ranged




     from 222 for Peak Particulate Test No. 1 to 285 for Peak




     Particulate Test No. 3.  Thus, the third sampling period con-




     tained pushes of higher greenness ratings than did the other




     two sampling periods.




            Table 5.6.1-7 presents the degree-of-greenness data




     for the particle size samples, each of which was conducted




     during a single push.  The three samples for which greenness




     data were obtained had relatively high values, ranging from

-------
                           TABLE 5.6.1-1
CHARACTERISTICS OF INDIVIDUAL  PUSHES  DURING PARTICULATE SAMPLING
                CONTINUOUS PARTICULATE TEST  NO.  1

                       Burns Harbor  Plant
                   Bethlehem Steel Corporation
                       Chesterton, Indiana
                          March  4^ 1975
Push
Time
09:42
09:50
10:03
10: 17
10:27
10:39
10:54
11:04
11: 14
11:31
11:44
11:53
13:28
13:48
13:55
15:47
15:56
16:10
16: 18
16: 28
16:36
16:46
16:53
17:04
17:13
Oven
Pushed
133
143
153
163
173
183
105
115
135
145
155
165
117
127
137
191
129
139
149
159
169
179
189
102
112
Average
Net
Coking
Time
(Minutes)
1044
1042
1047
1051
1052
1038
1044
1044
1044
1046
1035
1035
1092
1098
1096
1444
1112
1102
1101
1101
1101
1108
1048
1048
1047
1081
Degree of Greenness
Ratings
3,3,3
3,2,3
3,2,3
3,2-,2
3,2,3
3,3,3
4,4,4
3,4,3
3,2,4
3,2,3
223
^ , £ , j
323
_> , z , _>
3,3,3
3,2,3
2,2,3
2,2,3
3,2,3
3,4,3
3,2,3
3,2,3
4,4,3
3,2,3
4,3,3
3,2,3
3,4,3
—
Sum
9
8
8
7
8
9
12
10
9
8
7
8
9
8
7
7
8
10
8
8
11
8
10
8
10
9
Dura tion
(Seconds)
25
23
27
28
26
23
23
28
24
23
—
27
26
23
24
24
23
28
28
28
32
28
30
26
28
26
Sum X
Duration
225
184
216
196
208
207
276
280
216
184
—
216
234
184
168
168
184
280
224
224
352
224
300
208
280
227
Average Opacity for
Two Exhaust Stacks, %
0-3 Minutes
__
45
45
45
50
45
55
45
40
40
65
—
35
25
25
—
—
—
—
—
—
—
—
—

45
Rema inder
__
15
15
20
15
10
15
15
15
20
20
—
5
5
5
—
—
—
—
—
—
—
—
—

15
Average Flue
Tempera ture
(°F)
2420
—
—
2410
2400
—
—
2370
—
—
—
—
2370
2410
—
—
— '
—
—
—
—
—
• —
—

2400
                                                                                    I
                                                                                    oo
                                  Clayton  Environmental  Consultants,  Inc.

-------
                           TABLE 5.6.1-2

CHARACTERISTICS OF  INDIVIDUAL  PUSHES  DURING PARTICULATE SAMPLING
                  PEAK PARTICULATE TEST  NO.  1

                       Burns Harbor Plant
                    Bethlehem Steel Corporation
                       Chesterton, Indiana
                          March 4, 1975
Push
Time
09:42
09:50
10:03
10: 17
10: 27
10:39
10:54
11:04
11: 14
11:31
13:28
13:48
13:55
15:47
15:56
16: 10
16: 18
16:28
16:36
16:46
Oven
Pushed
133
143
153
163
173
183
105
115
135
145
117
127
137
191
129
139
149
159
169
179
Average
Net
Coking
Time
(Minutes)
1044
1042
1047
1051
1052
1038
1044
1044
1044
1046
1092
1098
1096
1444
1112
1102
1101
1101
1101
1108
1090
Degree of Greenness
Ratings
3,3,3
3,2,3
3,2,3
3,2,2
3,2,3
3,3,3
4,4,4
3,4,3
3,2,4
3,2,3
3,3,3
3,2,3
2,2,3
2,2,3
3,2,3
3,4,3
3,2,3
3,2,3
4,4,3
3,2,3
—
Sum
9
8
8
7
8
9
12
10
9
8
9
8
7
7
8
10
8
8
11
8
9
Dura tion
(Seconds)
25
23
27
28
26
23
23
28
24
23
26
23
24
24
23
28
28
28
32
28
26
Sum X
Duration
225
184
216
196
208
207
276
280
216
184
234
184
168
168
184
280
224
224
352
224
222
Average Opacity for
Two Exhaust Stacks, %
0-3 Minutes

45
45
45
50
45
55
45
40
40
35
25
25
—
—
—
—
—
—

40
Remainder

15
15
20
15
10
15
15
15
20
5
5
5
—
—
—
—
—
—

15
Average Flue
Tempera ture
(°F)
2420
—
—
2410
2400
—
—
2370
—
—
2370
2410
—
—
—
—
—
—
—

2400
                                                                                   oo
                                                                                   ro
                                  Clayton Environmental Consultants, Inc.

-------
                           TABLE 5.6.1-3
CHARACTERISTICS OF INDIVIDUAL  PUSHES  DURING PARTICULATE SAMPLING
               CONTINUOUS PARTICULATE  TEST NO. 2

                       Burns Harbor  Plant
                   Bethlehem Steel Corporation
                       Chesterton, Indiana
                          March  5, 1975
Push
Time
09:07
09:15
09:54
10:38
10:47
10: 55
11: 07
11: 18
11:40
14:23
14:43
14:50
15:04
15: 13
15: 22
15:33
15:45
15: 54
16:02
16:17
16:28
16:36
16:44
Oven
Pushed
119
129
139
169
179
189
102
112
122
134
154
164
174
184
191
116
126
136
146
156
166
176
186
Average
Net
Coking
Time
(Minutes)
1035
1035
1060
1068
1069
1065
1069
1069
1080
1045
1048
1045
1050
1048
1397
1045
1049
1050
1050
1058
1062
1063
1035
1069
: Degree of Greenness
Ratings
3,3,3
4,3,3
3,2,3
3,3,3
3,2,3
4,3,3
4,4,3
3,2,2
3,3,2
3,3,4
3,2,3
3,2,3
3,2,3
3,2,2
—
32 3
_>,.£,_>
3,2,3
2,3,3
3,2,3
223
^ > *• , J
3,2,3
4,3,4
4,4,3
—
Sum
9
10
8
9
8
10
11
7
8
10
8
8
8
7
—
8
8
8
8
7
8
11
11
9
Dura tion
(Seconds)
23
25
25
25
28
25
24
25
25
32
31
31
30
27
—
30
33
30
32
26
28
28
32
28
Sum x
Duration
207
250
200
225
224
250
264
175
200
320
248
248
240
189
—
240
264
240
256
182
224
308
352
241
Average Opacity for
Two Exhaust Stacks, %
0-3 Minutes

—
—
65
55
65
—
35
50
40
—
40
60
30
25
40
30
30
40
— ' -
— •
—

40
Remainder

—
—
15
10
10
—
10
10
10
—
10
10
5
10
10
10
15
10
— -
—
—

10
Average Flue
Temperature
(°F)
2420
2420
—
—
—
. —
2400
2480
—
2430
—
—
—
—
—
—
—
— —
-*—
—
—
—

2430
                                                                                   oo
                                                                                   w
                                 Clayton Environmental  Consultants,  Inc.

-------
                           TABLE 5.6.1-4
CHARACTERISTICS OF INDIVIDUAL PUSHES  DURING PARTICULATE SAMPLING
                   PEAK  PARTICULATE  TEST NO. 2

                        Burns Harbor Plant
                   Bethlehem Steel Corporation
                        Chesterton, Indiana
                          March 5, 1975
Push
Time
09:54
10:38
10:47
10:55
11:07
11:18
11:40
11:49
12:00
12:09
14:32
14:43
14:50
15:04
15: 13
15:33
15:45
15:54
16:02
16: 17
Oven
Pushed
139
169
179
189
102
112
122
132
142
152
144
154
164
174
184
116
126
136
146
156
Average
Net
Coking
Time
(Minutes)
1060
1068
1069
1065
1069
1069
1080
1080
1083
1081
1045
1048
1045
1050
1048
1045
1049
1050
1050
1058
1061
Degree of Greenness
Ratings
3,2,3
3,3,3
3,2,3
4,3,3
4,4,3
322
J > *• > *•
3,3,2
2,3,2
4,2,3
3 2 ^
J } *• i J
3 2 ?
3 > *• » *
3 2 "3
J , £. , J
3 ? "\
_> , z , j
3,2,3
3,2,2
3,2,3
3,2,3
2,3,3
3,2,3
2,2,3
—
Sum
8
9
8
10
11
7
8
7
9
8
7
8
8
8
7
8
8
8
8
7
8
Dura tion
(Seconds)
25
25
28
25
24
25
25
26
26
25
32
31
31
30
27
30
33
30
32
26
28
Sum X
Duration
200
225
224
250
264
175
200
182
234
200
224
248
248
240
189
240
264
240
256
182
224
Average Opacity for
Two Exhaust Stacks^, 7,
0-3 Minutes

65
55
65
—
35
50
30
45
—
45
—
40
60
30
40
30
30
40

45
Remainder

15
10
10
—
10
10
10
10
—
10
—
10
10
5
10
10
15
10

10
Average Flue
Temperature
<°F)

—
—
—
2400
2480
—
—
—
2540
2340
—
—
—
—
—
—
—
—

2440
                                                                                    I

                                                                                    00
                                  Clayton Environmental Consultants, Inc.

-------
                           TABLE 5.6.1-5
CHARACTERISTICS OF  INDIVIDUAL  PUSHES  DURING PARTICULATE SAMPLING
                CONTINUOUS PARTICULATE TEST NO. 3

                       Burns Harbor Plant
                    Bethlehem Steel  Corporation
                       Chesterton,  Indiana
                           March 6,  1975
Push
Time
09:50
10:05
10:17
10:36
10:58
11:07
11: 15
12:02
12:15
12:25
13:33
13:42
13:50
14:13
14:28
14:36
14:47
15:04
15:13
15:24
15:40
15:51
15:59
Oven
Pushed
156
166
176
186
108
118
128
168
178
188
151
161
171
181
103
113
123
191
133
143
153
163
173
Average
Net
Coking
Time
(Minutes)
1038
1040
1045
1052
1035
1036
1037
1050
1035
1035
1046
1047
1035
1035
1035
1035
1035
(1534)
1051
1040
1049
1038
1035
1062
- Degree of Greenness
Ratings
2,3,4
3,2,3
4,3,4
3,3,4
4,4,4
4,3,4
3,2,3
4,3,3
3,3,4
4,4,4
3,2,3
2,1,3
3,3,4
3,3,2
4,3,4
4,4,3
2,2,4
1,1,2
3,2,3
3,2,3
3,2,3
3,2,3
3,4,4
—
Sum
9
8
11
10
12
11
8
10
10
12
8
6
10
8
11
11
8
4
8
8
8
8
11
9
Duration
(Seconds)
28
27
30
32
30
26
26
30
28
32
28
28
29
30
57
27
28
—
28
25
28
29
30
30
Sum x
Duration
252
216
330
320
360
286
208
300
280
384
224
168
290
240
627
297
224
—
224
200
224
232
330
283
Average Opacity for
Two Exhaust Stacks, %
0-3 Minutes
45
45
80
70
70
75
65
—
60
70
40
35
60
35
80
50
50
30
—
—
— • ..
—

55
Remainder
20
25
25
25
25
25
20
—
30
30
25
25
25
25
25
25
25
25
—
— -
— •
'' —

25
Average Flue
Temperature
(°F)

—
2340
2290
2340
2380
—
—
2360
2320
2410
—
—
—
2290
2380
—
'— -
—
• •' • —
- —
—

2350
                                                                                    oo
                                                                                    Ul
                                  Clayton Environmental Consultants, Inc.

-------
                           TABLE  5.6.1-6
CHARACTERISTICS OF INDIVIDUAL PUSHES  DURING PARTICULATE SAMPLING
                   PEAK  PARTICULATE   TEST NO.  3

                        Burns Harbor  Plant
                   Bethlehem Steel Corporation
                        Chesterton, Indiana
                          March  6, 1975
Push
Time
09:50
10: 17
10:36
10:58
11:07
11:15
11:52
12:02
12:15
12: 25
13:33
13:42
14: 28
14:36
14:47
15:04
15: 13
15:24
15:40
15:51
Oven
Pushed
156
176
186
108
118
128
158
168
178
188
151
161
103
113
123
191
133
143
153
163
Average
Net
Coking
Time
(Minutes)
1038
1045
1052
1035
1036
1037
1040
1050
1035
1035
1046
1047
1035
1035
1035
(1534)
1051
1040
1049
1038
1066
: Degree of Greenness
Ratings
2,3,4
4,3,4
3,3,4
4,4,4
4,3,4
3,2,3
3,2,4
4,3,3
3,3,4
4,4,4
3,2,3
2,1,3
4,3,4
4,4,3
2,2,4
1,1,2
3,2,3
3,2,3
3,2,3
3,2,3
—
Sum
9
11
10
12
11
8
9
10
10
12
8
6
11
11
8
4
8
8
8
8
9
Duration
(Seconds)
28
30
32
30
26
26
30
30
28
32
28
28
57
27
28
—
28
25 .
28
29
30
Sum X
Duration
252
330
320
360
286
208
270
300
280
384
224
168
627
297
224
—
224
200
224
232
285
Average Opacity for
Two Exhaust Stacks, %
0-3 Minutes
45
80
70
70
75
65
—
—
60
70
40
35
80
50
50
30
—
—
—
-
60
Rema inder
20
25
25
25
25
20
—
—
30
30
25
25
25
25
25
25
—
• —
—

25
Average Flue
Temperature
(°F)

2340
2290
2340
2380
—
—
—
2360
2320
2410
—
2290
2380
—
—
—
—
—

2350
                                                                                   oo
                                                                                   ON
                                  Clayton  Environmental Consultants, Inc.

-------
                   TABLE  5.6.1-7

PUSH CHARACTERISTICS DURING PARTICLE  SIZE  SAMPLING

                Burns Harbor Plant
            Bethlehem Steel  Corporation
                Chesterton,  Indiana
                  March 6-7,  1975
I
oo
Sampling
Point
Number
P-l
P-2
P-3
P-4
Oven
Number
116
107
171
131
Net
Coking
Time
(Minutes)
1035
1035
1100
1040
Degree of Greenness
Ratings
3,4,3
4,3,3
—
4,4,3
Sum
10
10
—
11
Duration
(Second s )
30
28
—
28
Sum X
Duration
300
280
—
308
Average Opacity for
Two Exhaust Stacks, %
0-3 Minutes
—
—
—
55
Remainder
—
—
—
30
Average
Flue
Temperature
(°F)
—
—
2480
—
                          Clayton Environmental Consultants,  Inc.

-------
                    - 88 -







280 to 308.  The field data sheets for these greenness




ratings, as well as those from the particulate samples, are




presented in Appendix VV (Volume6).






5.6.2  Opacity




       5.6.2.1  Emissions from Exhaust Duct




               Although Figure 3.1-2 schematically shows




       the shed exhaust duct with  one exit stack, in  fact,




       two stacks were used to discharge the emissions from




       the shed exhaust duct during the study.  Further, a




       third exhaust stack was sealed completely during this




       study.  The opacity data acquired for the two  stacks




       were averaged and thereafter treated mathematically as




       if there were only a single stack.  These average opa-




       city data for the 3-minute peak periods during the




       particulate sampling, as well as the entire period




       following each push, are presented in Tables 5.6.1-1




       through 5.6.1-6.  The average 3-minute opacities for




       pushes during the sampling  periods ranged from 40 for




       Peak Particulate Test No. 1 and Continuous Particu-




       late Test No. 2, to 60 for  Peak Particulate Test No. 3




       The field data sheets from which these values were




       summarized are presented in Appendix WW (Volume 6).




               The exhaust duct opacity data were used for




       the following purposes:



               1.   To assess the  length of time required to




                    evacuate the coke-pushing emissions from




                    the shed (see Section 4.4);

-------
             - 89 -






        2.  To develop correlations with:




            a.  Greenness ,




            b.  Net coking time, and




            c.  Average crosswall temperature; and




        3.  To determine the representativeness of




            process conditions.






5.6.2.2  Fugitive Emissions




         Observations of the coke-side shed made




during the course of the study indicated that fugi-




tive particulate emissions escaped topside, from




Askania Valve positions, and from both ends of the




shed, as shown in Figure 3.1-2.  The sampling program




developed to evaluate this shed leakage  (fugitive par-




ticulate emissions) included both the evaluation of




the opacity of these fugitive emissions  and the use




of photographs (stills and motion picture) in addition




to particulate sampling.  Both types of  data were




acquired by U.S. EPA personnel.




        The shed leak opacity data are included in




Appendix XX (Volume 7).  These data were used to con-




vert the peak mass emission rates of the leaks, which




averaged 1-1/4 minutes in duration, to a continuous




(about 13 minutes) fugitive emission rate.  The photo-




graphs were used to estimate the average cross-sectional




area of the fugitive plumes.  Together with the fugitive




particulate concentration data discussed in Section 5.1.2,




these  visual  estimations were  used  to  determine  the con-



tinuous emission rate of fugitive emissions.  The method-



ology is discussed in detail in Appendix HH (Volume 5).

-------
                        - 90 -






5.6.3  Percent of Doors Leaking




       An additional component of the coke-side particulate




emissions, door emissions, appeared to be predominantly inde-




pendent of either coke-pushing or quench car particulate emis-




sions.  Therefore, an observation technique which recorded




the quantity of oven doors leaking during a short-term obser-




vation period was developed.  Any visible leak from a door




was considered a door leak.  A number of coke-side oven doors




were usually obscured by the coke guide during the short ob-




servation period, and therefore, the quantity of these ob-




scured doors was also recorded.  These observations yielded




an estimate of the percent of coke-side doors leaking.  Simi-




lar observations were obtained and results calculated for the




push side of the coke battery and for both sides of Battery




No. 2 for process documentation purposes.  All field data




sheets documenting door-leak observations are presented in




Appendix YY (Volume 7).




       Table 5.6.3 presents the door-leakage data for Batteries




1 and 2 on the days of particulate sampling.  These results




indicate that coke-side oven door leakage varied, ranging from




27 to 69 percent on Battery No. 1 and from 39 to 64 percent on




Battery No. 2.  Push-side oven door leakage was somewhat less




variable than coke-side oven door leakage; it ranged from 26




to 37 percent on Battery No. 1 and from 8 to 19 percent on Bat-




tery No. 2.  Finally, push-side chuck door leakage was the




least variable type of door leakage, ranging from 18 to 22




percent on Battery No. 1 and from 36 to 47 percent on Battery




No. 2.

-------
               TABLE 5.6.3

DOOR LEAKAGE ON PARTICULATE SAMPLING  DAYS

           Burns Harbor Plant
      Bethlehem Steel Corporation
           Chesterton,  Indiana
             March  3-7,  1975
vo
!-•
i
Coke Oven
Battery
No. 1
(Shedded)
No. 2
(Unshedded)
1975
Date
3/4
3/5
3/6
Average
(Total)
3/4
3/5
3/6
Average
(Total)
Coke-Side Oven Doors
Number
of Doors
Observed
58
55
63
(176)
69
62
72
(203)
Number
of Doors
Leaking
40
15
31
(86)
44
27
28
(99)
Percent
of Doors
Leaking
69
27
49
49
64
44
39
49
Push-Side Oven Doors
Number
of Doors
Observed
57
54
63
(174)
68
61
72
(201)
Number
of Doors
Leaking
21
14
23
(58)
13
9
6
(28)
Percent
of Doors
Leaking
37
26
37
33
19
15
8
14
Push-Side Chuck Doors
Number
of Doors
Observed
57
54
63
(174)
68
61
72
(201)
Number
of Doors
Leaking
10
12
13
(35)
32
22
28
(82)
Percent
of Doors
Leaking
18
22
21
20
47
36
39
41
                     Clayton Environmental Consultants,  Inc.

-------
                          -  92  -


5.7  Emission-Related Correlations

     Since this project was essentially investigative in nature,

several process parameters and indices of visible emissions were

examined to determine whether they were directly related to the

emissions measured.  Many potential correlations were examined,

using both the emission factor obtained from pushing operation

samples and the emission factors obtained from continuous particu-

late samples.  Supportive information for this section and addi-

tional attempted correlations are provided in Appendices ZZ to FFF

(Volumes 7 to 12) .


     5.7.1  Correlations Between Emission Factors and Indices
            of Visible Emissions
            Emission factors for continuous particulate samples

     were presented in Table 5.1.1.  Indices of visible emissions

     for these samples — degree of greenness and opacity for the

     peak emission period — were presented in Tables 5.6.1-1,

     5.6.1-3, and 5.6.1-5.  Linear correlation techniques reveal

     no statistically significant relationship between the three

     continuous filterable particulate emission factors and average

     degree of greenness or average opacity.  The linear correlation

     coefficients for the three pairs of data involved in these two

     potential relationships were 0.419 and -0.143, respectively.

            Correlations were also attempted using the three

     filterable particulate push-only emission factors given in

     Table 5.1.5 and the indices of visible emissions for peak

     particulate samples shown in Tables 5.6.1-2, 5.6.1-4, and

     5.6.1-6.  Again, linear correlation techniques yielded no

     statistically significant relationships between these

-------
                        - 93 -


parameters.  The linear correlation coefficients for these

potential relationships with degree of greenness and opacity

were 0.070 and 0.281, respectively.  In each of these cases,

the small quantity of data available may have partially

caused the poor correlations.


5.7.2  Correlations Between Emission Factors and Process
       Conditions

       Net coking time and average flue temperatures for the

continuous particulate sampling were summarized in Tables

5.6.1-1, 5.6.1-3, and 5.6.1-5.  Correlation techniques using

these parameters and the three continuous filterable particu-

late emission factors yielded no statistically significant

relationships.  The linear correlation coefficients for po-

tential relationships with net coking time and average flue

temperature were -0.761 and 0.189, respectively.

       Correlations were also attempted using these same two

process conditions for peak particulate samples and the three

push-only filterable particulate emission factors.  Although

the linear correlation coefficient for the potential relation-

ship with net coking time was quite high, -0.949, no statis-

tically significant relationship was found.  When the loga-

rithms of these values were used, an even higher linear cor-

relation coefficient resulted, -0.976.   Nevertheless, this

value, as well, was not statistically significant, likely be-

cause only three particulate samples were obtained.  The cor-

relation coefficient for the emission factor as a function of

average flue temperature, 0.405, was also not significant.

-------
                     -  94  -



5.7.3  Correlations  Involving Particle  Size  Distributions

       The particle  size  distributions  were  presented  graph-

ically in Figure 5.4,  and characteristics of the distribu-

tions were given in  Table 5.6.1-7.  Using these data,  linear

correlations were attempted between the distributions  and

characteristics of the pushes.  Although the correlation

coefficients were quite high, no statistically significant

correlation was apparent  between variations  in size distribu-

tion  (weight fractions less than one micron  and weight  frac-

tions less than five microns) determined for each of the sam-

ples and the greenness of the push.  This was likely due to

the few data pairs available  (one of the four samples was

lacking greenness data).

       No statistically significant correlation was apparent,

either, between particle  size and net coking time.  It  was

not possible to correlate particle size with flue tempera-

tures or opacity because  of the lack of data in both cate-

gories.


5.7.4  Correlations  Between Indices of  Visible Emissions and
       Process Conditions

       Two parameters  in  this study can be considered as

indices of visible emissions:  opacity  and greenness.   The

values for these two indices during particulate sampling

were indicated in Tables  5.6.1-1 through 5.6.1-6.  Two

parameters which can be considered indicative of process

conditions were also shown in these tables:  net coking time

and flue temperature.  In order to determine whether these

-------
                    - 95 -






indices of visible emissions could be considered a function




of process conditions, several correlations were attempted.




To use the most complete data base possible, the data in




Tables 5.6.1-1 through 5.6.1-6 were combined with the other




data obtained during particulate sampling days.  All data




sets can be found in the tables in Appendix ZZ, Volume 7,




       Peak opacity was found to be highly correlated with




net coking time during particulate sampling days.  Several




relationships were evaluated using these data, including the




linear form, logarithms, and inverses.  A relationship in-




volving inverses, however, was found to be statistically




superior.  A modification to the net coking time variable,




subtraction of a constant of 1000 minutes, improved the cor-




relation further.  The constant factor of 1000 was selected




because none of the net coking times witnessed in a review




of two years of data from the Burns Harbor Plant was less




than this value (Appendix ZZ, Volume 7).




       The.final regression function, plotted in Figure




5.7.4-1, was found to be:






        Peak Opacity   =  0'0386 ' 0'730 ( NCT -' 1000 ) *




The correlation coefficient for this relationship, which com-




prised 60 pairs of data, was -0.626.  This value is statis-




tically significant at a level exceeding the 99-percent




level.




       A very highly statistically-significant correlation




was also obtained for greenness as a function of net coking




time for the particulate sampling days.  Again several forms

-------
                                      FIGURE 5.7.4-1
                            NET  COKING TIME VERSUS OPACITY FOR
                                PARTICULATE  SAMPLING DAYS
                                    Burns Harbor Plant
                               Bethlehem  Steel  Corporation
                                   Chesterton,  Indiana
                                     March 3-7,  1975
 Opacity
(Percent)
                                                  iV-i .-I-: l\: i :•; -.X : I Mi
                                               Confidence
Interval
                                               Clayton Environmental  Consultants, Inc.

-------
                    - 97 -






of the relationship were attempted, but the one relating the




inverse of greenness and the inverse of net coking time minus




a constant of 1000 was found to be superior:




       The regression equation, which is plotted in Figure




5.7.4-2, was:






        Greenness   =  °-00510 ' °'0431 ( NCT -' 1000 ) '




The linear correlation coefficient for this relationship,




which comprised 104 sets of data, was -0.335.  This coeffi-




cient is statistically significant at a level exceeding the




99-percent level.




       Correlations were also attempted using a second process'




conditions parameter — average flue temperature.  When peak




opacity was considered as a function of flue temperature, a




function involving the logarithm of temperature was found to




be superior.  The regression equation, which is plotted in




Figure 5.7.4-3, was:




       Peak Opacity = 4122 - 523   [ln(TempM  .




The correlation coefficient for this relationship, which in-




volved 24 sets of data, was -0.655.  This value is signifi-




cant at a level exceeding the 99-percent level.




       A potential relationship between greenness and flue




temperature was also considered.  The linear relationship




between the logarithms of both values was found to be supe-




rior.   The equation, which is plotted in \Figure 5.7.4-4, was:




       In (Greenness) = 75.8 - 9.04  lln(Temp)]  .

-------
Degree of Greenness
  (Sum *  Duration)
FIGURE 5.7.4-2
                             DEGREE OF GREENNESS  VERSUS NET COKING TIME
                                    FOR PARTICULATE SAMPLING DAYS
                                          Burns Harbor  Plant
                                     Bethlehem Steel Corporation
                                         Chesterton, Indiana
                                           March  3-7, 1975
                                        \ •• ii : J i t-\ -; .• \. i i . . \i •; i , • ,\-i-
                                        n   95%  Confidence
                Interval
                                                  Clayton  Environmental Consultants,  Inc.

-------
                IH
 Opacity
(Percent)
           FIGURE 5.7.4-3

OPACITY VERSUS FLUE  TEMPERATURE FOR
     PARTICULATE  SAMPLING DAYS

         Burns Harbor Plant
    Bethlehem Steel  Corporation
        Chesterton,  Indiana
          March  3-7,  1975
Itl:
14-
     rt-s
                                                                                     j-ii-
                                                                                     m
                             !&•
                                                Clayton Environmental Consultants,  Inc.

-------
                                               FIGURE 5.7.4-4
Degree
  (Sum
 of  Greenness
* Duration)
                        DEGREE  OF  GREENNESS VERSUS FLUE TEMPERATURE
                               FOR PARTICULATE SAMPLING DAYS
                                             Burns Harbor Plant
                                        Bethlehem  Steel  Corporation
                                            Chesterton,  Indiana
                                              March 3-7,  1975
rrrrffi+r
 95%
             Confidence
                  Interval
                                                   Clayton  Environmental CoRsultaoits, Inc.

-------
                    - 101 -







The correlation coefficient for this relationship, which




involved 31 pairs of data, was -0.709.  This value is sta-




tistically significant at a level exceeding the 99-percent




level.






5.7.5  Correlations Among Visible Emissions Measurements




       Attempted correlations involving opacity as a func-




tion of greenness for the particulate sampling days also




resulted in a highly statistically-significant correlation.




The final equation, which is plotted in Figure 5.7.5-1, was:




       Peak Opacity = -166 + 39.0  jln(GreennessM  .




The correlation coefficient for this relationship, covering




54 sets of data, was 0.646; the statistical significance of




this value exceeds the 99-percent level.




       In order to further evaluate the opacity of the emis-




sions from the shed exhaust stack, the opacity data obtained




during particulate sampling days were combined with all




available opacity data taken by certified visible emission




observers during a one-year period prior to the start of




the field testing.  The results were then grouped in two




ways.  The first method clustered the data into the four




categories listed below:




       1.  Particulate test days  (typical, normal conditions




           only);




       2,  Non-test days  (typical, normal conditions only);




       3.  Typical but abnormal conditions (i.e., coke-




           pushing cycle duration exceeding 30 minutes); and




       4.  Typical but abnormal conditions (i.e., net coking




           time greater than 18-1/2 hours).

-------
                                  FIGURE 5.7.5-1
                      OPACITY VERSUS DEGREE OF GREENNESS FOR
                             PARTICULATE  SAMPLING  DAYS
                                Burns Harbor Plant
                            Bethlehem Steel  Corporation
                               Chesterton,  Indiana
                                 March  3-7,  1975
 Opacity
(Percent)
                                                  95% Confidence  Interval
                                              Clayton Environmental Consultapts,  Inc.

-------
                          -  103  -






     The opacity data, taken at intervals of approximately  15




     seconds, were then averaged for each of these  four categories,




     The results are plotted in Figure 5.7.5-2.  The numbers in




     parentheses above each line on the graph indicate the  number




     of sets of data averaged to obtain the curve.




            To investigate  this relationship further, the data




     were regrouped based upon  net coking time.  Six 15-minute




     net coking time intervals  were established, using the  1,035-



     minute minimum net coking  time specified by Bethlehem  Steel




     Corporation as a baseline.  The results are plotted in Figure




     5.7.5-3.  Again, the numbers in parentheses above each line




     on the graph indicate  the  number of data sets  averaged to




     obtain the curve.  This figure indicates that  increasing net




     coking time yields predictably decreasing shed exhaust opac-




     ities.






5.8  Effect of the Shed Upon Dustfall




     In addition to collecting  door, quench car movement, and push-




ing emissions, the shed on  Battery No. 1 also acts  as a large set-




tling chamber for coarse dust,  especially along the shed wall.




During a push, and for a period of about two or three minutes




thereafter, a worker or observer under the shed may experience




a "fallout" of settleable particulate matter along  the length of




the shed, especially along  the  shed wall.




     To determine the magnitude of this effect, and to determine




how it varies with location, dustfall (settleable particulate)




measurements were made on and about both Batteries  1 and 2.  Dust-




fall jars were exposed at fixed locations on both batteries:

-------
                                  - 104  -
Average
Opacity
 80,
 75
                      FIGURE  5.7.5-2

.COMPOSITE GRAPH OF SHED EXHAUST DUCT OPACITY VERSUS TIME

                   Burns  Harbor  Plant
                Bethlehem  Steel .Corporation
                   Chesterton, Indiana
                      March 3-7,  1975


                       •• Partlculate  test  days  (typical,  normal)
                       — Non-test days  (typical,  normal)
                       —•_- Abnormal condition-after downtime (>30 mln.)
                          Abnormal condition-long  coking (>18-l/2 hr.)
                  45    60    75    90   105   120  135  150  165

                  Time After Commencement of Push  (seconds)

                                 Clayton Environmental  Consultants, Inc.
                                                     180  195  210  225

-------
- 105 -
                 FIGURE 5.7.5-3
           SHED EXHAUST DUCT OPACITY VERSUS TIME  FOR
                   VARIOUS NET COKING TIMES

                      Burns Harbor Plant
                  Bethlehem Steel Corporation
                      Chesterton, Indiana
                        March 3-7, 1975
   15
30   45    60   75   90   105   120  135 150   165
    Time After Commencement of Push  (seconds)
180  195   210
                          Clayton Environmental Consultants, Inc.

-------
                          -  106  -



     1.  Along the bench  of Battery 1   (under the  shed);


     2.  Along the bench  of Battery 2   (unshedded);


     3.  Along the shed wall of Battery 1, about 30 feet  from


         the side of the  bench; and


     4.  Along a line geometrically equivalent to  that described


         in  (3) above, near Battery 2.


It should be kept in mind that this technique provides only rela-


tive values of dustfall intensity.


     Table 5.8-1 shows dustfall data for comparable locations on


Batteries 1 and 2.  The units in each case are grams/m^/week (con-

                                                f\
vertible to the usual "ambient" units of tons/mi^/month by multi-


plying by 11.4).  The geometric mean value for each location is


also reported.  All dustfall sampling periods, dustfall weights,


and the percentage of particulate matter collected on the sieve


are presented in Appendix GGG (Volume 12).


     In comparing the data in Table 5.8-1, several general comments


should be kept in mind:


     1.  Pushing emission rates for these two coke batteries are


         not necessarily  identical.  The batteries are of differ-


         ent design with  respect to their heating  system, Battery


         2 being newer.


     2.  Battery 1 produces about 10 percent more  coke per hour.


     3.  Variations in wind speed and direction may affect the


         significance of  the data.


     A shed visitor's perception that dustfall is  severe under a


shed should be interpreted in the context of the physical position


of the observer.  It is common for such observations to be made at


ground level near or under the shed wall on the far side of the

-------
                                -  107 -

                              TABLE  5.8-1

         SUMMARY OF DUSTFALL MEASUREMENTS AT BATTERIES 1 AND 2
                              (gm/m2/wk)
                          Burns Harbor  Plant
                     Bethlehem Steel  Corporation
                          Chesterton, Indiana
                            March  3-7, 1975
Sampling
Location
ii.. -
Shedded ^
Spare Door
Unshedded +
Spare Door
Bench 123
Bench 152 A
Bench 178
Shedded Bench
Geometric Mean
Bench 223
Bench 252 +
Bench 278
Unshedded Bench
Geometric Mean
Ground 151 ^
Ground 177
Shedded Ground
Geometric Mean
Ground 236
Ground 251 +
Ground 277
Unshedded Ground
Geometric Mean
Shedded Wall
Wall 121 A
Mid 1
Mid 2** A
Shedded Mid
Geometric Mean
End l**
End 2 A
Shedded End
Geometric Mean
1975 Sampling Period
3/3-4
^_
198
^^™
—
1400
2930
2010
2020
26,600
12,000
17,900
357
450
101
253
11,600
3180
1380
2090
1530
1280
1400
3/4
1660
56
2690
2410
3790
2910
6010
1500
26,900*
3000
21,900
15,300
18,300
1130
492
746
11,400
4110
5850
4900
1530
2560
1980
3/4-5
3700
275
3300
3650
3050
3320
3800
3380
4640
3910
11,200
11,200
645
576
610
12,500
—
—
—
—
3/5
3260
2320
484*
—L.
1710
4010
2320
2520
27,800
7340
14,300
1010
1010
18,400
3550
3370
3460
1630
1780
1700
3/5-6
3080
352
4810
3590
4160
2000
2540
2080
2190
22,200
11,100
15,700
5*
136
136
16,300
,^_
—
—
—
3/6
332
222
2840
6270
5000
4470
2300
4500
3220
25,200
30,400
27,700
120
739
137
230
10,600
2380
2410
2390
3640
1670
2470
Geometric
Mean
1830
286
3320
3810
3800
3620
2510
2730
2890
2690
24,600
13,100
17,500
364
494
250
365
13,200
3240
2850
3040
1930
1770
1850
* Statistical tests indicate that these values are suspect.
  not used in the statistical analyses.
                                                              They were
** Duplicate Samples

A  Battery No. 1
+  Battery No. 2
                              Clayton Environmental Consultants,  Inc,

-------
                             -  108  -






quench car tracks.  For reasons of accessibility  and  safety,  it  is




easier to observe there than on the coke-side bench,  which  offers




no clearance between a battery and its door machine.   Dustfall




under the wall is not necessarily  representative  of a  work  sta-




tion, i.e., normal worker exposure.




     In addition, the small numerical difference  in Table 5.8-1




between dustfall rates at bench locations on the  shedded and  un-




shedded batteries means that the shed had no apparent  measurable




effect on dustfall at this key work station.




     Table 5.8-2 presents the acetone-soluble and cyclohexane-




soluble content of five selected dustfall samples.  Neither ex-




traction resulted in a concentration which exceeded 0.1 percent.




Table 5.8-3 presents the pH of five other samples; the values




ranged from 6.0 to 7.1.




     In order to identify how the coke-side shed  affects measur-




able dustfall rates, other potential influential  factors were




first evaluated.   These other variables were:  a) greenness of




the pushes, b) pushing rate, and c) location of the dustfall




bucket.   The data shown in Table 5.8-1 were used  for the analyses.




All statistical analyses were performed u'sing the logarithms  of




the dustfall rates since dustfall rates are known to be log-nor-




mally distributed.^




     Because dustfall measurements inherently are relatively  crude,




the precision (reproducibility) of the method was estimated by




exposing pairs of dustfall jars at the same site.  The eight




pairs of samples  identified in Table 5.8-1 as "mid" and "end"

-------
             -  109  -
          TABLE 5.8-2

SUMMARY OF ACETONE SOLUBLE AND
  CYCLOHEXANE  SOLUBLE  CONTENT
 OF SELECTED DUSTFALL SAMPLES

      Burns Harbor Plant
  Bethlehem  Steel  Corporation
      Chesterton,  Indiana
       March  3-7,  1975
Site
Ground 177
Wall 121
Bench 278
Bench 223
Bench 152
Sampling Period
Start
Date
3/3
3/3
3/4
3/4
3/6
Time
16:47
16:31
08:26
15:44
09:52
Stop
Date
3/4
3/4
3/4
3/5
3/6
Time
09:17
09:05
15:49
10: 19
17:46
Percent
Ace tone
Solubles
0.008
0.002
0.005
0.087
0.042
Percent
Cyclohexane
Solubles
0.002
0.004
0.006
0.01
0.02
          Clayton Environmental Consultants,  Inc.

-------
          -  110 -
         TABLE 5.8-3

     SUMMARY  OF  pH  OF
 SELECTED DUSTFALL SAMPLES
    Burns  Harbor  Plant
Bethlehem Steel Corporation
    Chesterton, Indiana
      March 3-7, 1975
Site
Bench 152
Ground 151
Mid 1
End 1
End 2
Sampling Period
Start
Date
3/4
3/3
3/6
3/3
3/4
Time
08:17
16:40
09:42
17: 12
10:35
Stop
Date
3/4
3/4
3/6
3/4
3/4
i
Time
15:30
09: 10
16:01
09: 18
16:30
pH of
Sample
6.6
6.2
7.1
6.0
7.1
        Clayton Environmental  Consultants,  Inc,

-------
                         - Ill -







samples were simultaneous pairs.  To determine the precision of



each of these pairs of samples, the difference in the  logarithms



of the paired values was divided by the geometric mean of the




pair.  These precision values ranged from 0.2 to 11 percent.




These eight precision values were then evaluated to determine if



any pair could be considered an "outlier."  No pair of samples




could be classified as an outlier by this method.  In all addi-




tional evaluations, the geometric mean dustfall rate was then



used for the paired samples.  The "mid" and "end" samples are




indicated in further analyses as "ground" samples taken at the




4-foot level.



     An average greenness for ovens pushed during each dustfall




sample was determined by averaging the "sum times duration" values



for the pushes that occurred during the sampling period.  These



average greenness values were then arranged in ascending order




and a median value of 230 found.  All greenness values below 230



were labeled "low" and all above 230 were labeled "high."  It is



important to note that 77 percent of the pushes under the shed




had high average greenness values while only 14 percent of the




unshedded pushes had average greenness values that were con-



sidered high.




     Pushing rate for a dustfall sample was determined by count-




ing the number of either shedded or unshedded ovens,  as applicable,



that were pushed during a sample and dividing by the  time duration




of the dustfall sample.  Again the values were arranged in ascend-




ing order and the median was found to be 4.7 pushes per hour.




All pushing rates below this value were considered "low" and all




rates equal to or abovp this value were considered "high."

-------
                         - 112 -







     The dustfall data were then arranged into several  cells  in




order to best eliminate any confounding effect of the multiple




variables.  These cells, shown in Table 5.8-4, were determined by




first dividing the data into that applicable to shedded and un-




shedded areas.  Each area was subdivided into one of four common




locations:  "spare door," "bench," "ground" at the 4-foot level,




or "ground" at the 10-foot level.  Other locations sampled in




this study were not used in the analyses because samples taken




at these locations were taken only within the shed.  The next two




subdivisions were those of "low" and "high" pushing rates and




"low" and "high" greennesses.  Tests for outliers were  then con-




ducted within each of these cells; a single cell now contained




the most homogeneous subset of data available.  Only the three




values indicated in Tables 5.8-1 and 5.8-4 were found to be out-




liers.




     In order to determine whether greenness and dustfall rate




were correlated, the number of subdivisions was reduced by one so




that greenness was no longer used as a basis of subdivision.  In




each of the remaining 16 cells, the logarithm of dustfall rate




for each sample was paired with its average greenness value.  The




linear correlation coefficient for the pairs in each cell was then




determined.  Only the value for the nine shedded bench  samples




with low pushing rates was found to be significant at the 95-percent




level.  On the basis of the fact that only one of the correlation




coefficients was found to be significant,  it was concluded that




the relationship between greenness and dustfall rate was not sig-




nificant for the overall data set.

-------
                           - 113 -

                          TABLE 5.8-4

     FORMAT USED FOR ANALYSES  OF DUSTFALL DATA  (gm/m2/wk)

                      Burns Harbor Plant
                  Bethlehem Steel  Corporation
                      Chesterton,  Indiana
                        March  3-7,  1975

Greenness
Pushing
Rate

Spare Door






Bench








Ground
(41 level)





Ground
(10* level)



S h e d d e d
Low
Low
1,660


2,690
2,410
3,790








4,900
1,980




21,900
15,300
11,400





High




























High
Low
3,700
3,260
3,080
3,300
484*
4,810
3,650
6,270
3,050
3,590




3,460
1,700




27,800
22,200
11,200
7,340
11,100
12,500
18,400
16,300
High
332


2,840
5,000









2,390
2,470




25,200
30,400
10,600





Unshedded
Low
Low
275


3,800










136
137












High
56
352
222
6,010
2,000
2,300
1,500
3,380
2,540
26,900*
4,640
2,320
2,080
4,500
1,130
5*
120
J_ fm \J
645
739
492
576








High
Low




























High
2,320


1,710
4,010









1,010













* Statistical tests indicate that these values are suspect.
  They were not used in statistical analyses completed after
  the test for outliers.
                          Clayton Environmental Consultants,  Inc

-------
                          -  114  -






     Since greenness and  dustfall rate were not  found  to  be  cor-




related, those dustfall rates which did not have a greenness




rating associated with them could now be included in further




analyses.  Thus, these values were added to their respective




cells determined in the previous analysis, and the tests  for out-




liers were repeated.  No  additional suspect values were found.




     The correlation between pushing rate and dustfall rate was




evaluated next in a similar manner.  The number of subdivisions




was reduced by one by eliminating pushing rate as a basis of sub-




division.  In each of the remaining eight cells, the logarithm of




dustfall rate was paired with its pushing rate.  The linear corre-




lation coefficient was determined for each cell.  None of the




values was found to be significant at the 95-percent level.  It




was thus concluded that pushing rate and dustfall rate were not




correlated for this set of data.



     Two factors remained to be considered — the location of the




dustfall bucket and the shed effect, i.e., shedded versus unshed-




ded areas.  To determine whether or not location was a significant




factor, two separate one-way analyses of variance were performed.




The door-bench-ground location samples were compared to each other




for the shedded and unshedded areas.  Under the shed, the geo-




metric mean of the 10-foot-level ground samples was significantly




higher than that of the other three locations.  The geometric mean




of the bench samples, in turn, was significantly higher than the




geometric mean of the spare door samples.  The geometric mean of




the 4-foot-level ground samples did not differ significantly from




that of the bench samples or the spare door samples.

-------
                         - 115 -






     For the unshedded area, the geometric mean of the bench sam-




ples was found to be significantly higher than the geometric means




of the 4-foot-level ground samples and the spare door samples.   In




this area, the geometric mean dustfall rates for the ground sam-




ples and the spare door samples were essentially the same.




     Since location of the dustfall bucket appeared to be a sig-




nificant factor, a one-way analysis of variance was done for each




of the locations to determine whether or not the shed was a sig-




nificant factor.  At three of the four locations — the spare doors




and both ground levels — the geometric mean dustfall rates under




the shed were found to be significantly higher than those for sam-




ples not taken under the shed.  However, for the bench location




the geometric mean dustfall rates under the shed were not statis-




tically different from those found at the unshedded location.  It




can thus be concluded that both the presence of the shed and the




location of the dustfall bucket have a significant influence upon




measured dustfall rates in this study.






5.9  Impact of the Shed Upon Airborne Agents Within the Shed




     The question of whether a semienclosed shed adjacent to a




coke-oven battery has a significant effect upon the quality of




the work environment within the shed was not addressed in this




study.  Two studies by the National Institute of Occupational



Safety and Health (NIOSH), however, did address this issue. 5»6)






5.10 Precis ion of Test Re suits




     Although the terms "precision" and "accuracy" are often re-




garded as synonymous, they do have different technical meanings.




The accuracy of a measurement signifies the closeness with which

-------
                         -  116 -


the measurement approaches  the true value.   Precision,  on  the

other hand, characterizes the repeatability  of the measurements.

Thus, the precision of a measurement denotes the  closeness with

which a given measurement approaches the average  of  a  series of

measurements taken under similar conditions.  Clearly,  if  the

bias is large, a measurement may be very precise  but very  inac-

curate.

     Many techniques exist  to evaluate the precision of a  result.

Ideally, simultaneous replicate samples are  taken and  the  coeffi-

cient of variation, the standard deviation expressed as a  percent-

age of the mean, is used as a measure of precision.  This  technique

was used for eight pairs of dustfall samples taken in  this study

and reported in Section 5.8.

     When the sample at hand is the only measure  of the variability

of data at given conditions, a confidence interval may be used to

bracket the true mean of the population.  This interval may be re-

garded as a first estimate  of the precision  of the results.  In

this study, such confidence intervals were constructed at the 95-

percent level, implying a 5-percent risk of  not bracketing the true

mean of a series of test measurements.   This confidence interval is

expressed in the Summary and Conclusions (Section 2.0) as m (Hh r),

where m is the arithmetic mean and 2r is the confidence interval.

This technique was used in  the evaluation of particulate emission

rates and emission factors.
This report prepared by:   Thomas A. Loch, Ph.D., P.E,
                           John E. Mutchler, P.E.
                           Richard J. Powals, P.E.
                           Janet L. Vecchio

-------
                             - 117 -






                         6.0  REFERENCES






1.   United Nations Report, Economic Commission for Europe,




     "Air Pollution by Coking Plants," ST/ECE/Coal/26, 1968.






2.   U.S. EPA, Division of Stationary Source Enforcement, "Study



     of Coke-Side Coke-Oven Emissions, Great Lakes Carbon Corpora-




     tion, St. Louis, Missouri," in print.






3.   Conner, W.D. and J.R. Hodkinson, Optical Properties and




     Visual Effects of Smoke Stack Plumes, U.S. EPA, Office of




     Programs, Publication No. AP-30, May, 1972.






4.   TR-2 Air Pollution Measurement Committee, "Recommended




     Standard Method for Continuing Dustfall Survey (APM-1,




     Revision 1)," Journal of Air Pollution Control Association,




     16:7, pp. 372-377, 1966.






5.   National Institute for Occupational Safety and Health,



     Division of Technical Services, Industrial Hygiene Services




     Branch, "An Industrial Hygiene Survey of the Bethlehem Steel




     Corporation (Burns Harbor Facility) Coke Side Emission



     Collecting Shed," Project No. 75-32, April 28, 1975.






6.   National Institute for Occupational Safety and Health,




     Division of Technical Services, Industrial Hygiene Services




     Branch, "An Industrial Hygiene Survey of the Great Lakes




     Carbon Corporation (Missouri Coke & Chemical Division) Coke




     Side Emission Collecting Shed," Project No. 75-31,




     April 28, 1975.

-------
                          -  118 -



        7.0  SOME ANTICIPATED  QUESTIONS AND ANSWERS
                    RELATIVE TO THIS PROJECT


1.  Q.  Were these emission tests  truly representative  of  the  typi-

        cal conditions occurring at Coke Battery No.  1  at  the  Burns

        Harbor plant?

    A.  Yes.  A great deal of  care was taken and much documentation

        was obtained to ensure that both the process  operations  and

        the sampling and analytical procedures would accurately

        represent typical conditions existing at the  subject battery

        (see Section 3.2).


2.  Q.  Just how reasonable is the choice "+ 1070" for defining

        typical conditions?

    A.  Quite reasonable.   Basically, two off-setting conditions

        are at work.   One  is the inherent variability of process

        parameters and the other is the need for maintaining, as

        close as possible, maximum operating conditions during the

        test period which  are representative of "normal" condi-

        tions.  We believe that _+ 5% is probably too strict a  cri-

        terion for process variables which would not materially

        affect the outcome of testing.  However, anything more than

        jf 1070 could very likely cause significant changes in emis-

        sion concentrations, rates, and characteristics.  Therefore,

        the criterion of _+ 10% was chosen to represent  "typical"

        conditions.


3.  Q.  Were the frequency and extent of our observations sufficient

        to characterize the particulate emissions as a  function of

        process input rate?

-------
                         - 119 -





3.  A.  While emission factors and emission  rates  did  not  correlate




        significantly with process input rate,  some  related  correla-




        tions were found to be significant statistically.   Those




        correlations relating the indices of visible emissions  to




        various process parameters, such as  net  coking time,  proved




        significant.  The fact that emission factors or emission




        rates could not be correlated to pertinent process  param-




        eters is due predominantly to the small  number of  data




        points available (three) for emission measurement  tests .




        Because so much of the data acquired during  these  source




        tests appears to be relatively precise,  several correla-




        tions were examined nonetheless.






4.  Q.  Why was the term "tons of dry coal fed"  used as a  normali-




        zation factor?




    A.  Dry coal feed rate rather than wet coal  feed rate  was used




        because it was an accurate measurement and compatible with




        a mass balance concept which historically  has  been  the




        "process weight rate" method of normalizing  emission  data




        to production rate.






5.  Q.  Does the Burns Harbor study provide  sufficient basis  for




        expressing an emission factor for coke-side  emissions?




    A.  The Burns Harbor tests provide emission  factors for  coke-




        side emissions for a host of contaminants  for  a single




        coke-oven battery at a s ingle production rate  and  a  relatively




        narrow range of operating conditions.  Nonetheless,  the




        degree to which data analysis has revealed statistically




        significant correlations between emission  factors  and

-------
                         - 120 -




        process parameters indicates that extrapolation of these




        results to other batteries may be appropriate and meaning-




        ful only if similar ranges of process parameters exist at




        the untested battery.  Such emission factors should, however,




        be refined as more data are acquired.






6.  Q..  Why were so many correlations attempted?




    A.  This study was, in some ways, a prototype for subsequent




        studies.  Therefore, it was important to learn the rela-




        tionships, if any, between process and emission variables




        that could describe variations in emission rates.






7.  Q.  Were any especially good correlations developed as a result




        of this study?




    A.  Yes.  Net coking time appears to be one of the most signifi-




        cant variables affecting at least the opacity and degree




        of greenness (see Figure  5.7.4-1 and 5.7.4-2) and probably




        the mass rate of particulate emissions, although data




        analysis did not reveal any significant correlations be-




        tween emission factor and net coking time.  This may also be




        true for other contaminants.  However, no data were acquired




        to substantiate the latter postulation.  Obviously, other




        process conditions must be maintained relatively constant




        for any of these correlations to be developed and, in fact,




        they were relatively constant during the Burns Harbor sampling




        Unfortunately,  the  small number of particulate samples




        precluded  the possible  development of a  statistical




        relationship between particulate  emissions and other



        parameters.

-------
                         - 121 -




8-  Q.  Who obtained the process information for this study?




    A.  Bethlehem Steel Corporation personnel gathered the  infor-




        mation which was then provided either directly to U.S.




        EPA personnel or to Clayton personnel.






9.  Q.  Who obtained the sampling and analytical data?




    A.  Clayton personnel (see Appendix RR, Volume 6).






10. Q.  Who obtained the visible emissions data?




    A.  U.S. EPA personnel (see Appendix RR, Volume 6).






11. Q.  Were there any experimental (atypical) oven doors,  unusual




        maintenance, or other peculiar operating conditions during




        the tests?




    A.  Yes.  Experimental doors were observed in use on a  few of




        the coke ovens.  No unusual or abnormal maintenance or




        operating conditions were noted on any of the data  pro-




        vided to Clayton personnel.






12. Q.  Did an analysis of Bethlehem Steel Corporation pollution




        and/or inspection reports document that the plant,  or at




        least the oven doors, were being handled and/or maintained




        in an unusual fashion immediately before or during  the




        test period as compared  to other periods?




    A.  No.  However, inspection reports are too infrequent to define




        these conditions  very effectively (see Appendix YY, Volume 7)






13. Q.  What is the optimum net coking time for reduction of emis-




        sions?




    A.  Without an analysis of the minimum net coking time  neces-




        sary for preventing coke-oven "stickers," this answer can-




        not be provided.

-------
                         - 122 -




14. Q.  What is the best method of describing "greenness" in a




        quantified sense?




    A.  The  product of the  sum of the greenness ratings and the




        duration of the push  was  judged to be the best.






15. Q.  Is there any specific reason for very green pushes?



    A.  This could not be determined with any degree of certainty




        simply because there  weren't enough "very green pushes" to




        acquire statistical information.  Only a few very green



        pushes were observed  during the source testing.  However,




        as mentioned previously,  the shorter the net coking time,




        apparently the greater the degree of greenness and the




        higher the opacity of exhaust duct emissions.  An especially




        "cold" oven, operating significantly below the average cross-




        wall temperature of the other ovens, could very likely cause




        a very green push.






16. Q.  Could the causes of leaks from the coke-oven shed be quanti-




        fied sufficiently to  indicate which variable was most sig-




        nificant regarding fugitive particulate emissions?




    A.  No.   Too many independent variables were working together




        to affect leakage from the shed.  These included wind,




        exhaust gas flowrate   through the shed system, location of




        the  oven pushed, greenness of a specific push, etc.






17. Q.  Why was a three-minute peak sampling period chosen instead




        of a 2-1/2-minute peak period?




    A.  The  three types of data acquired to obtain an accurate esti-




        mate of the maximum period of pushing emissions evacuation




        from the shed (temperature, filter obscurity, and opacity)

-------
                         -  123 -




        all indicated that  the best measure  of  peak  pushing  emis-




        sions was one taken over a three-minute  interval.   In  fact,




        the preliminary procedures to indicate  and quantify  this




        peak pushing emissions evacuation period were  performed




        not only prior to beginning emission testing,  but  even




        during the preliminary tests on February 24, 1975  so that




        the best indicators possible of that subject :period  would




        be obtained.  It turned out that the choice  of  three minutes




        was very propitious, especially when the data  were evaluated




        after the source testing was performed.  The opacity data




        indicated strongly  that the choice of three minutes  was




        not only fortunate but very accurate as well (see  Figures




        5.7.5-2 and 5.7.5-3).






18. Q.  Why was the probe rotated in the stack  rather  than being




        pulled out between sampling points (for  the  peak emission




        tests)?




    A.  The insertion and withdrawal of the probe  would more  likely




        cause sampling error and possible sample losses.  Additionally,




        it was important to rotate the face of  the nozzle away from




        the entering stream lines of the exhaust gas flow during




        non-sampling periods.  Therefore, the face of  the  nozzle




        was rotated with the probe at least 90°  off  of  facing  directly




        into the exhaust gas flow whenever the  sampling train  was




        shut down.






19. Q.  Why wasn't the attenuation coefficient  a good  indicator of




        the particulate concentration in the duct?

-------
                         - 124 -





    A.  The closeness of the measured particulate concentrations




        among the three sample runs precluded the ability to




        distinguish between average attenuation coefficients for




        the continuous particulate emission tests.  Thus, the




        attenuation coefficient may have been a good indicator of




        particulate concentration but could not be correlated with




        such a narrow range of average particulate emission con-




        centrations.






20. Q.  Was the proper path length utilized for attenuation coeffi-




        cient calculations?




    A.  Yes.  Mr. Kirk Foster of the U.S. EPA helped in this en-




        deavor.






21. Q.  Why were the probe and filter maintained at the temperature




        of the stack?




    A.  Because it was part of the objective to neither create nor




        diminish filterable particulate matter.






22. Q.  How frequently was the A_ checked for isokinetic sampling?



    A.  It was monitored continuously.  For peak sampling, it was



        recorded every 30 seconds (see Appendix LL, Volume 5).






23. Q.  Why was acetone used rather than some other organic solvent




        for cleaning up the sampling train?




    A.  Not only is it the recommended procedure in U.S. EPA Method




        5, but acetone effectively removes the deposition of particu-




        late matter and allows for good clean-up of the components




        of the sampling train when sampling coke-oven emissions.

-------
                            - 125 -






24. Q.  Should cyclohexane be used as the solvent for sampling




        train clean-up for coke-oven particulate tests in the




        future?




    A.  The results of sampling for cyclohexane solubles and insolu-




        bles presented in Section 5.5 indicate that cyclohexane may




        be a preferable solvent.  Acetone was used in this study,




        however, because it is required by the standard EPA Method




        5 procedure.  In actuality acetone may be a better solvent




        because it is a better wetting agent than cyclohexane, i.e.,




        removes particulate matter by wetting rather than solubility.






25. Q.  How did we attempt to quantify the organic fraction of the




        particulate material?




    A.  Through the use of acetone solubility, the cyclohexane cap-




        ture technique, the activated carbon adsorption technique,




        and grab flask samples (see Appendix LL, Volume 5).






26. Q.  How did we account for or avoid potential sampling and




        analytical problems with sulfate, nitrite, nitrate,  hydro-




        gen chlorides, and pseudo-particulate?




    A.  By maintaining the sampling conditions very close to the




        stack conditions extant during the type of tests (whether




        continuous or peak),  and with pre-planning regarding the




        analytical techniques used.   These techniques minimized




        potential problems with pseudo-particulate generation in




        the impingers.  The extent to which the aforementioned




        species affected results is  minimal since particulate




        analyses reveal small amounts of these materials (see




        Tables  5.3-2 and 5.5).

-------
                         - 126 -






27. Q.  Do we have a true representation of the test period com-




        pared to the typical process at Battery 1 of the Burns




        Harbor plant?




    A.  Yes (see Section 3.2).







28. Q.  How well do our particle size results represent the EPA



        Method 5 filterable particulate catch?




    A.  Very well (see Section 5.4 for complete details).






29. Q.  Why were the peak and continuous particulate emission rates




        different for each test?




    A.  Because of variability of individual coke-oven pushes




        and door leaks (see Tables 5.6.1-1 through 5.6.1-6).






30. Q.  Why was the average "pushes per hour" figure for the con-




        tinuous particulate samples slightly different than the




        typical condition?




    A.  Because sampling had to be interrupted at numerous times




        for process malfunctions and/or for changing the probe




        from port to port.  Additionally, the procedure (stopping




        the sampling whenever the push-to-push time exceeded 30




        minutes) slightly increased the pushes per hour figure.




        Nevertheless, it was still within the + 10% criteria  (see




        Table 3.2.2-1).  Further, even though this was the case,




        no apparent change in the emissions data occurred between




        the first continuous particulate sample run which operated




        at a rate of about 5.2 pushes per hour and the second and




        third tests which each operated at a rate of about 4.8




        pushes per hour.  Apparently, then, the push-per-hour

-------
                          -  127  -





        figure, which obviously related to the tons of coke charged




        per hour for each sampling period, is not a significant cri-




        terion for establishing an average continuous particulate




        emission rate from the coke battery.







31. Q.  What was the stability of percentage of door leakage,




        whether push-side oven door, coke-side oven door, or push-




        side chuck door?




    A.  Widely variable (see Table 5.6.3).






32. Q.  Can we estimate from door leakage inspections how long it




        takes for a door to stop leaking?




    A.  There are insufficient data to make this estimate from




        the data acquired at Burns Harbor (see Appendix YY, Volume




        7).






33. Q.  What was the stability of net coking time for the individual




        coke-oven pushes?




    A.  Widely variable (see Appendix ZZ, Volume 7).






34. Q.  What was the stability of greenness by oven day-to-day?




    A.  Quite unstable except in the case of a few sets of ovens




        which apparently were responsible for slightly greener




        pushes (see Appendix ZZ, Volume 7).






35. Q.  What was the stability of the percentage of moisture of




        the coal mix?




    A.  Quite stable (see Appendix D, Volume 2).






36. Q.  Who built Battery No. 1?




    A.  Wilputte.

-------
                          - 128 -





37. Q.  Who built Battery No.  2?



    A.  Koppers .






38. Q.  How extensive were chain-of-custody procedures?




    A.  Very extensive.  Someone from the Clayton crew was always




        present with the samples or the samples were locked securely




        in storage (see Appendix NN, Volume 6).






39. Q.  Do you have  other major recommendations  for subsequent




        coke-oven test work?




    A.  Yes.  Include a complete industrial hygiene/occupational




        health survey at the same  time the emission testing is




        performed on a subject battery; acquire  visible emissions




        data continuously during each particulate sample run; and




        continuously refine the required roster  of materials to be




        mea sured.






40. Q.  What reservations do you have regarding  this study?




    A.  1.  We believe the information acquired  here, although of




            excellent quality, should not be extrapolated  indis-




            criminately to all coke ovens.




        2.  It would have been helpful to substantiate  (with




            "official" process information) that the pre-study




            average  coal feed rate and the pre-study coke  production




            rate were equivalent to those recorded during  and after




            the  study  (see Table 3.2.2-1).

-------
                              TECHNICAL REPORT DATA
                        (Please read Instructions on the reverse before completing)
1. REPORT NO.
   340/1-76-012
                         2.
                                                   3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
   Source Testing of a Stationary  Coke-Side
   Enclosure
                                                   5. REPORT DATE
                                                     5-20-77
                            6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
                                                   8. PERFORMING ORGANIZATION REPORT NO.
   Thomas A.  Loch,  John E. Mutchler,
   Richard J.  Powals, Janet L.  Vecchio
9. PERFORMING ORGANIZATION NAME AND ADDRESS
                                                   10. PROGRAM ELEMENT NO.
   Clayton Environmental Consultants,  Inc.
   25711 Southfield Road
   Southfield,  Michigan   48075
                            11. CONTRACT/GRANT NO.

                             68-02-1408; Task  10
12.SPONSORING AGENCY NAME AND ADDRESS
   U.S. Environmental Protection  Agency
  -Division  of Stationary Source  Enforcement
                            13. TYPE OF REPORT AND PERIOD COVERED
                            14. SPONSORING AGENCY CODE
   401 M Street,  S.W.
   Washington,  D.C
20460
15. SUPPLEMENTARY NOTES
                 Volumes 2-12 of  this  report are appendices that supple-
   ment Volume  1 and  are available  from the sponsoring  agency ^above)*  *'
16. ABSTRACT
          This report  summarizes an emission study that  documents the na-
ture  and extent of  particulate and gaseous  emissions typically emanating
from  the coke side  of Coke Battery No.  1  at the Burns Harbor Plant of
Bethlehem Steel Corporation, Chesterton,  Indiana.  The  information was ob-
tained  to help provide a  basis for:

      1.   Development  of EPA policy on  coke-side coke battery emissions
          and their  control.         ,

      2.   Assessment of the adequacy of  State Implementation Plans (SIPs)
          to achieve Nat tonal Air Quality Standards in areas  proximate- to
          coke plants.

      3.   Assessment of the adequacy of  control devices  being proposed for
          abatement  of coke-side emissions.

          The source testing included measurement of 48  different contamin-
ants, and the project  resulted in several process-emission  correlations.
17.
                           KEY WORDS AND DOCUMENT ANALYSIS
               DESCRIPTORS
                                       b.lDENTIFIERS/OPEN ENDED TERMS
                                       c.  COSATI Field/Group
  Coking
  Air  Pollution
  Opacity
  Visual Inspection
  Particles
  Particle Size Distribution
                 New  Source Perform-
                 ance  Standards

                 Emission Testing

                 Performance Tests
 13B
 14D
13. DISTRIBUTION STATEMENT

  Unlimited
                 19. SECURITY CLASS (This Report)
                  Unclassified
21. NO. OF PAGES
 128 (Vol.  1)
                                       20. SECURITY CLASS (This page)
                                         Unclassified
                                                               22. PRICE
EPA Form 2220-1 (9-73)

-------