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)
------- |