NOVEMBER 1977
HUMAN POPULATION EXPOSURES TO COKE-OVENS
ATMOSPHERIC EMISSIONS
US. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
WASHINGTON, D.C. 2O460
Project Offtcw: ALAN f. CARLIN
Technical Monitor: JUSTICE A. MANNING
CONTRACT 6*01-4314
.SRf Project EGU-5794.
CENTER FOR RESOURCE AND ENVIRONMENTAL SYSTEMS STUDIES
R«port No. 27
-------
NOVEMBER 1977
HUMAN POPULATION EXPOSURES TO COKE-OVENS
ATMOSPHERIC EMISSIONS
By: BENJAMIN E. SUTA
SRI INTERNATIONAL
MENLO PARK, CA 94025
Prepared for:
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
WASHINGTON, D.C. 20460
Project Officer: ALAN P. CAR LIN
Technical Monitor: JUSTICE A. MANNING
CONTRACT 68-01-4314
SRI Project EGU-5794
CENTER FOR RESOURCE AND ENVIRONMENTAL SYSTEMS STUDIES
Report No. 27
-------
NOTICE
This is a preliminary draft. It has been released by the U. S.
Environmental Protection Agency (EPA) for public review and comment
and does not necessarily reflect Agency policy. This report was
provided to EPA by SRI International, Menlo Park, California, in partial
fulfillment of contract No. 68-01-4314. The contents of this report
are reproduced herein as received by SRI after comments by EPA. The
opinions, findings, and conclusions expressed are those of the authors
and not necessarily those of EPA. Mention of company or product
names is not to be considered as an endorsement by the EPA.
ii
-------
ACKNOWLEDGMENT
It is a pleasure to acknowledge the cooperation and guidance given
by Alan Carlin and Justice Manning of the U.S. Environmental Protection
Agency in the preparation of this report. In addition, a number of other
people generously supplied input data. They include:
Mark Antell, U.S. Environmental Protection Agency
Robert Armbrust, New York State Department of Environmental
Conservation
Thomas Au, Pennsylvania Department of Environmental Resources
Bernard Bloom, U.S. Environmental Protection Agency
Walter Cooney, Maryland State Environmental Health Administration
Ron Dubin, Pennsylvania Department of Environmental Resources
Arvid Ek, Allegheny County Health Department
Clemens Lazenka, Philadelphia Air Quality Division
Jim Payne, Texas Air Control Board
C. B. Robison, Jefferson County, Alabama Board of Health
Terry Sweitzer, Illinois Environmental Protection Agency
Peter Warner, Wayne County, Michigan Health Department
Robert Yuhnke, Pennsylvania Department of Environmental Resources.
iii
-------
PREFACE
There is a substantial body of evidence, both direct and indirect,
that the mixture that coke oven emissions represent is carcinogenic
and toxic. Current U. S. Environmental Protection Agency (EPA) policy
states that there is no zero risk level for carcinogens. To determine
what regulatory action should be taken by EPA on atmospheric emissions
of coke ovens, three reports have been prepared: (1) a health effects
assessment, (2) a population exposure assessment, and (3) a risk
assessment document based on the data in the first two assessments.
This document is the human population exposure assessment and presents
estimates of the numbers of people in the general population of the
United States exposed to atmospheric concentrations of coke oven
emissions. Estimates are provided of population exposures to ambient
concentrations of benzo(a)pyrene Bap and benzene soluble organics (BSO)
material caused by coke oven emissions.
iv
-------
CONTENTS
ACKNOWLEDGMENT 11
LIST OF ILLUSTRATIONS v
LIST OF TABLES vi
I INTRODUCTION 1
II SUMMARY AND CONCLUSIONS 2
A. Overview 2
B. At-Risk-Populations 5
C. Population Estimation 5
D. Population Exposures to BaP Emitted by Coke-Ovens . . 7
E. Population Exposures to BSO Emitted by Coke-Ovens . . 8
F. Considerations in the Use of the Annual Average as a
Measure of Exposure to Coke-Oven Emissions 12
G. Accuracy of Estimated Exposures 15
H. Other Potential Human Exposure Routes 16
III SOURCES OF COKE OVEN EMISSIONS 19
A. The Coking Process 19
B. Environmental Emissions During Coking 20
C. Coke Processing Plants 21
IV A METHOD OF ASSESSING BaP AND BSO CONCENTRATIONS IN THE
VICINITY OF COKE-OVENS 30
A. General 30
B. Categorization of Coke Plants by Emission Control . . 31
C. Background Concentrations 32
D. Evaluation of Ambient Concentration Data for Coke Plant
Locations 35
E. Relationship Between BaP and BSO Atmospheric Concentra-
tions 40
F. Population Exposure Estimates 43
-------
Appendices
A AMBIENT ATMOSPHERIC BaP AND BSO CONCENTRATIONS 45
rt
A. General 45
B. Atmospheric BaP and BSO Concentration Data Recorded
Near Coke Manufacturers 45
C. Ambient Background BaP and BSO Concentration Data . . 62
B STATISTICAL EVALUATION OF BaP ATMOSPHERIC CONCENTRATION
DATA RECORDED IN THE VICINITY OF COKE PLANTS 80
A. General 80
B. Statistical Distribution of 24-Hour BaP Atmospheric
Concentrations 80
C. Precision of Estimates Based Upon Small Sample Sizes 82
D. Evaluation of Ambient Concentration Data as a Function
of Distance from Coke Plant Locations 85
C DETAILED ESTIMATES OF POPULATIONS AND BaP CONCENTRATIONS
FOR INDIVIDUAL COKE FACILITIES 101
BIBLIOGRAPHY 105
vi
-------
ILLUSTRATIONS
II-l
II-2
IV-1
IV-2
B-l
B-2
B-3
B-4
B-5
B-6
B-7
B-8
B-9
B-10
B-ll
B-12
B-13
B-14
Estimated Population Exposures to BaP
Estimated Population Exposures to BSD
Relationship Between BSD and BaP Atmospheric Concen-
trations for all Locations
Relationship Between BSD and BaP Atmospheric Concen-
trations for Coke Oven Locations ....
Statistical Distribution for Atmospheric BaP Concen-
trations
Atmospheric Concentrations of BaP for Johnstown,
Pennsylvania
Atmospheric Concentrations of BaP for Geneva, Utah
Atmospheric Concentrations of BaP for Wayne County,
Michigan
Atmospheric Concentrations for BaP for Allegheny
County, Pennsylvania
Atmospheric Concentrations of BaP for Buffalo, New
York
Atmospheric Concentrations of BaP for Birmingham,
Alabama
Atmospheric Concentrations of BaP for Granite City,
Illinois
Atmospheric Concentrations of BaP for Sparrows Point,
Atmospheric Concentrations of BaP for Cleveland, Ohio
Atmospheric Concentrations of BaP for Monessen, Penn-
sylvania
Atmospheric Concentrations of BaP for Gadsden, Alabama
Atmospheric Concentrations of BaP for Duluth, Minnesota
Atmospheric Concentrations of BaP for Philadelphia,
10
13
41
42
81
86
87
88
89
90
91
92
93
94
95
96
97
Pennsylvania 93
vii
-------
TABLES
II-l Estimated BaP Emissions in the United States (1972) 4
II-2 Summarization of Ambient BaP and BSO DATA 6
II-3 Annual Average Exposure Concentrations for BaP
Emitted by Coke Ovens 9
II-4 Summarization of Population Exposures to BaP from
Coke Oven Emissions 11
II-5 Annual Average Daily BSO Inhalation for Persons Re-
siding Near Coke Plants 14
II-6 Benzo(a)pyrene Concentrations in Foods 17
III-l Correlations Among PAH Compounds in the Air over
Greater Birmingham, Alabama, 1964 and 1965 .... 22
III-2 Correlation Coefficients Among Log Concentrations
of 13 PNA and BSO Samples Taken Within Five Coke
Plants 23
III-3 By-Product Coke Plant Locations and Capacities . . 25
III-4 Estimated Size and Productive Capacity of By-Product
Coke Plants in the United States on December 31, 1975 29
III-5 Directory of U.S. Beehive-Coke Plants 24
IV-1 Assumed Emission Weighting Factors for Plant Com-
pliance Status 33
IV-2 Classification of Coke Plants into Emission Cate-
gories (1974-1975) 34
IV-3 Estimated Annual Background Concentrations of BaP
for Coke Plant Locations 36
A-l Monessen Air Study, 24-Hour Sample Characteristics 47
A-2 Ambient BaP Concentrations for Allegheny County,
Pennsylvania 48
A-3 BaP Data Obtained During First Stage Alerts at
Liberty BoroughSite 8790 49
A-4 Atmospheric BaP Concentrations Near the Geneva Works
in Utah 49
A-5 Ambient BaP Concentrations for Wayne County, Michigan 51
A-6 Ambient BaP Concentrations for Buffalo, New York 52
A-7 Ambient BaP Concentrations for Duluth, Minnesota 53
viii
-------
A-8 Ambient BaP Concentrations for Gadsden, Alabama . . 53
A-9 Ambient BaP Concentrations for Birmingham, Alabama 55
A-10 CHAMP Site Ambient Atmospheric BaP Data for the
Birmingham Area (1975 Data) 55
A-ll Ambient BaP, BSO, and TSP Concentrations for Johns-
town, Pennsylvania 56
A-12 Ambient BaP, BSO, and TSP Concentrations for Philadel-
phia, Pennsylvania 57
A-13 Ambient BaP Concentrations for Granite City, Illinois 59
A-14 Additional Atmospheric Ambient Data for Granite City,
Illinois 59
A-15 Ambient BaP Concentrations for Houston, Texas ... 60
A-16 Ambient BaP Concentrations for Cleveland, Ohio . . 61
A-17 Ambient Atmospheric BaP and BSO Concentrations for
Sparrows Point, Maryland 63
A-18 Ambient BaP and BSO Concentrations for Chattanooga,
Tennessee 63
A-19 Annual Average Ambient BaP Concentrations at NASN
Urban Stations (ng/m3) 64
A-20 Seasonal Variations of Benzene Soluble Organic Sub-
stances 66
A-21 Summarization of Ambient BaP and BSO Data 68
A-22 Annual BaP Averages for Selected Cities 68
A-23 Ambient BaP Concentrations for Pennsylvania, 1976 70
A-24 Distribution of BaP Concentrations in Ambient Air
at Charleston, South Carolina 75
A-25 Ambient Atmospheric BaP and BSO Concentrations for
Maryland Locations 76
A-26 Atmospheric BaP and BSO Concentrations for CHESS
and CHAMP Sites (1975 Data) 79
B-l Statistical Summary for Sampling Data Taken from a
Number of Locations 83
B-2 Estimated Parameter Values for Regression Approxima-
tions to Ambient Data 100
C-l Detailed BaP Population Exposures (ng/m3) 102
C-2 BaP Exposures for Persons in Locations Having More
than One Coke Facility 104
ix
-------
LIST OF CHEMICAL ABBREVIATIONS
A anthracene
Ant anthanthrene or anthanthrene
BaA benz(a)anthracene
BaP benzo(a)pyrene
BbF benzo(b)fluoranthene
BcA benzo(c)acridene
BeP benzo(e)pyrene
BghiP benzo(g,h,i)perylene
BjF benzo(j)fluoranthene
BkF benzo(k)fluoranthene
BSD benzene soluble organics
Chr chrysene
Cor coronene
DBahA dibenz(a,h)anthracene
Flu fluoranthene
Per perylene
Pyr pyrene
TSP total suspended particulates
-------
I INTRODUCTION
The primary objective of this study has been to quantify the environ-
mental atmospheric exposure of the general human population to coke-oven
emissions of benzo(a)pyrene (BaP) and benzene soluble organics (BSO).
To do so, we have located and characterized coke production plants,
estimated atmospheric environmental concentrations of pollutants resulting
from coke production, and estimated human populations exposed to various
levels of these pollutant concentrations.
In this report, we indicate human exposure to coke-oven emissions in
terms of the average amount inhaled per day for each population subgroup.
Note that this study reports exposures that took place before biological
sorption occurred and that the degree of sorption is not considered. In
addition, because the results of this study are intended to serve as input
to another study in which health effects are to be assessed, health effects
are not addressed. Another study is also being conducted to describe the
chemical and physical properties of coke-oven emissions; therefore, these
results are not included in this study.
The main findings of this report are provided in tables and
figures. The text describes the methodologies, assumptions, and data
sources used. All estimates given in this report depend in large part on
data reliability and availability, both of which varied widely. Some
discussion of this variability is provided in Appendices A and B.
-------
II SUMMARY AND CONCLUSIONS
A. Overview
There are 65 by-product coke plants in the United States. (Some
authors list 62, omitting separate operations for three of the locations.)
These plants consist of an estimated 231 coke oven batteries, containing
13,324 ovens that have a theoretical maximum annual productive capacity
of 74.3 million tons of coke. The industry generally operates at about
of the theoretical capacity.
Environmental emissions occur in the coking operation during charging,
from leaks in the oven doors and the tops of ovens, from the waste gas
stack, during pushing and quenching, and from by-product processing.
The various batteries are characterized by different types of control and
operational procedures which affect the amount of their emissions. In
general, the measurement of environmental emissions from coke-ovens has
been limited to some atmospheric sampling of BaP for about one-third of
the locations. Atmospheric concentrations of TSP have also been measured
for many of the locations, and the BSO fraction of the TSP has been
measured for a few locations. Atmospheric concentrations of other sub-
stances that may be emitted by coke-ovens have generally not been recorded.
In addition, very little work has been done to characterize detailed
emission factors for coke-ovens. Because of these limitations, this
report's estimates of nonoccupational exposures to coke-oven emissions
are based on the two substances for which some atmospheric concentration
data are availableBaP and BSO. These two substances might be considered
as substitute or surrogate measures of total exposure. However, much
more monitoring data will be required before we can conclude that concen-
trations of these two substances always correlate well with other emitted
substances that are important from a health viewpoint.
-------
Atmospheric concentration data recorded during 1964 and 1965 for
Birmingham, Alabama, with several coke plants located in the vicinity,
showed that the correlation coefficient for BaP with 11 other polynuclear
aromatic compounds ranged from 0.65 to more than 0.99. For BSO with 11
other substances, it ranged from 0.58 to 0.88 (U.S. EPA, 1975). In addi-
tion, occupational exposure data recorded by NIOSH (1974) show correlation
coefficients between BSO and 13 other polynuclear aromatic compounds to
range from 0.71 to 0.94. The same study also showed correlation coeffi-
cients for BaP with 12 other polynuclear aromatic compounds ranging from
0.57 to 0.95. The substances used in these correlation studies are given
in Section III of this report.
It is difficult to use ambient data "to assess exposures to coke-oven
emissions; most communities have other sources of the same substances,
generally associated with coal and other fossil fuel combustion. Hence,
any evaluation of population exposures to coke-oven emissions must separate
the background concentration from the coke-oven contribution. Of course,
for health risk assessment, the summation of the two is important. Table
II-l reports a BaP emission inventory made by the Environmental Protection
Agency (EPA) for 1972. Stationary sources account for 98% of the nation-
wide estimate. Estimates of BaP emissions from coke ovens range from
about 0.06 ton per year to approximately 170 tons per year, depending on
assumptions used. EPA (1974) used the higher value because it is based
on data from the United States (using a crude emission factor of 2.5 g
of BaP per ton of coal processed). The coke production is estimated to
account for approximately 19% of the nationwide BaP emissions. EPA is
currently working on better factors to characterize coke-oven emissions
(Manning, August 1977).
BaP may also have natural sources, including bituminous coal which
also contains benzo(a)anthracene and other polycyclic organic matter.
Two of three types of asbestos used industrially were found to contain
oils with BaP. Mold may constitute another source (U.S. EPA, 1974).
The National Air Surveillance Network (NASN) routinely monitors
suspended particulate levels in urban and nonurban areas. This program
is described in more detail in Appendix A. BaP and BSO are monitored for
-------
Table II-l
ESTIMATED BaP EMISSIONS IN
THE UNITED STATES (1972)
Emissions
Source Type tonne/yr
Stationary Sources
Coal, hand-stoked residential furnaces 300
Coal, intermediate-size units 7
Coal, steam power plants <1
Oil, residential through steam power type 2
Gas, residential through steam power type 2
Wood, home fireplace 25
Enclosed incineration-apartment through
municipal 3
Vehicle disposal 25
Forest and agriculture 11
Other open burning 10
Open burning, coal refuse 310
Petroleum, catalytic cracking 7
Asphalt air blowing <1
Coke production (0.06)-170
Mobile Sources
Gasoline-powered automobiles and trucks 11
Diesel-powered trucks and buses <1
Tire degradation 11
Source: US EPA (1974).
-------
40 locations Chat include cities with and without coke-ovens and rural
areas.* The BaP and BSD concentrations recorded for this program are sum-
marized in Table II-2. The BaP concentrations are generally 0.1 ng/m
for rural locations. Most urban locations without coke-ovens have aver-
age concentrations of less than 1 ng/m (the average is 0.38 ng/m );
however, areas with coke-ovens generally have average concentrations in
3 3
excess of 1 ng/m (the average is 1.21 ng/m ).
B. At-Risk-Populations
The at-risk populations to coke-oven emissions are defined as the
resident populations exposed to coke-oven atmospheric emissions. Exposure
is based on the estimated average annual concentrations occurring at the
place of residence of at-risk population subgroups. Average daily human
exposure is calculated as the product of the average annual concentration
and human daily ventilation rate.
C. Population Estimation
jAn evaluation of the concentration data shown in Appendix A indicates
that coking operations may possibly affect atmospheric concentrations out
to a radius of 15 km from the operations.I For most cases, the affected
radius is considerably less than 15 km; however, for conservative analysis,
population residing within a 15-km radius from each coke plant is considered
as the maximum potential exposure population. For the estimation of
populations at-risk to selected concentrations resulting from coke ovens,
the resident populations were calculated in a series of five concentric
rings about each coke plant. The spacing of the rings was based on the
shape of the concentration versus distance functions illustrated in
Appendix B. The distances are 0-0.5, 0.5-1.0, 1.0-3.0, 3.0-7.0, and
7.0-15 km.
Geographic coordinates of most of the coke plants were obtained from
the U.S. EPA NEDS data system. The remainder were obtained from consult-
ing maps or by telephone conversations. The population residing in each
concentric ring about each coke plant was obtained from the Urban Decision
Systems, Inc., Area Scan Report, a computer data system that contains the
BSO monitoring was discontinued in 1972.
-------
Table II-2
SUMMARIZATION OF AMBIENT BaF AND BSO DATA
Pollutant
BaP (ng/m3)
1975 Data
BSO (pg/m3)
1971-72 data
Statistic
Average
Sample size
Range
Average
Sample size
Range
Cities
with
Coke
Ovens
1.2]
21
0.3-4.7
4.21
25
2.1-7.3
Cities
without
Coke
Ovens
0.38
13
0.03-0.9
3.75
12
1.9-5.6
Rural
3
<0.
0.
2
0.8-1
Areas
1
95
.1
-------
1970 census data in the smallest geographic area available (city blocks
and census enumeration districts). The total population residing in each
of these rings for all the coke plants is as follows:
Distance from Resident
Coke Plant (km) Population
0-0.5 32,700
0.5-1 116,000
1-3 1,644,000
3-7 7,226,000
7-15 22,200,000
|The total population residing within 15 km of the coke plants is approx-
imately 31,220,000 people.I In the exposure calculations it was found
-i *
that only about 17,100,000 of these people were affected by coke oven
emissions.
D. Population Exposures to BaP Emitted by Coke-Ovens
The annual average BaP atmospheric concentrations were estimated
for each of the five concentric rings around each of the coke plants.
Recorded ambient data were used for those locations having a sufficient
number of samples and monitoring sites; otherwise, the extrapolative
modeling technique described in Section IV was used. For locations with
several coke plants, a procedure was devised to assess the combined
atmospheric concentrations by summing the contribution for individual
plants for areas in overlapping geographic rings. The population within
the rings was assigned to the overlapping sections by using uniform
distribution assumptions.
Two calculations were made for the atmospheric concentration of each
ring: the concentration resulting from only coke oven emissions and the
total concentration, which includes background plus coke oven emissions.
The background concentrations used are given in Section IV and range
*
Coke oven emissions resulted in an increase in the average annual
atmospheric BaP concentrations of 0.1 ng/m^ or more.
-------
2
from 0.04 to 1.6 ng/m . The wide range In background concentrations
indicates the variations of other BaP-emitting activities in the cities.
Table II-3 summarizes the number of people exposed to various BaP concen-
tration levels. Detailed exposure estimates are given in Appendix C. The
cumulative distribution for these exposure concentrations is given on
Figure II-l. Total average BaP concentrations range from a high of
3 3
100 ng/m to a low of below 1 ng/m . The median population exposure
concentration is around 7 ng/m .
Potential human exposures from inhalation are given in Table 11-4.
3
For these exposures a human ventilation rate of 15 m /day was assumed.
This is the amount of air inhaled per 24-hr day by a standard man as
defined in the Radiological Health Handbook (1960). The approximately
17,100,000 exposed people inhale between 3 to 1,500 ng BaP per day on
an annual average basis. About one-half of these people inhale more
than 100 ng BaP per day.
E. Population Exposures to BSO Emitted by Coke-Ovens
Sufficient data have not been collected near coke plants nor have
emission factors been developed for adequately assessing the atmospheric
BSO concentrations resulting from the plants' emissions. The approach
taken here is to estimate the BSO concentrations, based on the estimated
BaP concentrations. A number of problems are associated with this approach,
however, and the results can, at best, be described as "ballpark estimates."
Further work on assessing plant emission factors or measuring environmental
concentrations should help to improve the quality of future estimates.
The approach taken has been described in Section IV of this report.
Three methods of estimation were tried. All three methods estimate BSO
contributions attributable to coke-ovens and add this to estimated back-
ground concentrations. One method involved using empirical formulas
derived from data taken from coke oven areas where both BaP and BSO
* 3
In a kepone assessment report, U.S. EPA (1976b) used a rate of 8.6 m /day
for an average adult. The exposures given here can be converted to the
8.6 m-Vday rate by multiplying by 0.57.
-------
Table II-3
ANNUAL AVERAGE EXPOSURE CONCENTRATIONS FOR BaP
EMITTED BY COKE OVENS
Subgroup
Cumulative Number of People Exposed
Concentration
Range (ng/m3)
95-100
50-55
45-50
40-45
35-40
30-35
25-30
20-25
15-20
10-15
8-10
6-8
5-6
4-5
3-4
2-3
1-2
0.5-1
0.2-0.5
Background, plus
Coke Oven Emissions
1,800
2,670
2,720
4,220
5,920
9,320
14,120
19,120
82,820
630,220
705,320
981,020
1,097,720
1,345,920
3,069,020
7,335,620
15,148,620
16,754,020
17,106,620
Coke Oven
Emissions Only
1,800
2,670
2,720
4,220
5,920
8,320
9,920
18,920
82,620
219,920
662,620
798,920
995,220
1,182,320
1,971,620
3,216,820
8,243,520
12,923,120
17,106,620
Number exposed to indicated concentration or more.
-------
100
I
i
o
111
u
o
o
111
o
tc.
Ill
3
Z
Q
UJ
i
II Illlll
1I Illlll
I I I I I III
BACKGROUND CONCENTRATIONS ARE
ASSUMED TO BE 0.04 to 1.6 ng/m3 FOR
INDIVIDUAL LOCATIONS
I I I I Mil
I ' I I I Ml
104 106 106
NUMBER OF PEOPLE EXPOSED TO INDICATED CONCENTRATION OR MORE
107 2x107
FIGURE 11-1. ESTIMATED POPULATION EXPOSURES TO BaP (Background, plus Emissions)
-------
Table II-4
SUMMARIZATION OF POPULATION EXPOSURES TO
BaP FROM COKE OVEN EMISSIONS
(Background Plus Coke Oven Emissions)
Subgroup
Concentration^
Range (ng/m3)
95-100
50-55
45-50
40-45
35-40
30-35
25-30
20-25
15-20
10-15
8-10
6-8
5-6
4-5
3-4
2-3
1-2
0.5-1
0.2-0.5
Subgroup
Population
Exposure (ng/day)*
1,425-1,500
750-825
675-750
600-675
525-600
450-525
375-450
300-375
225-300
150-225
120-150
90-120
75-90
60-75
45-60
30-45
15-30
7.5-15
3.0-7.5
Number
of People
in Subgroup
1,800
870
50
1,500
1,700
3,400
4,800
5,000
63,700
547,400
75,100
275,700
116,700
248,200
1,723,100
4,266,600
7,813,000
1,605,400
352,600
Cumulative
Number
of Exposed
People**
1,800
2,670
2,720
4,220
5,920
9,320
14,120
19,120
82,820
630,220
705,320
981,020
1,097,720
1,345,920
3,069,020
7,335,620
15,148,620
16,754,020
17,106,620
**
Based on the annual average.
Number exposed to indicated concentration or more.
11
-------
atmospheric concentrations have been measured. The other two methods
involved multiplying the BaP concentration attributable to coke ovens by
a constant that relates BSO to BaP and adding this to an assumed background
BSO concentration. The two factors used were 1 ng BaP = 0.1 yg BSO and
1 ng BaP = 0.5 yg BSO. The 0.1 factor appears to be reasonable, based on
occupational exposure data whereas the 0.5 factor should give an upper
limit. An average background BSO concentration of 3.75 yg/m3 was assumed
for all locations.
The exposures estimated by these methods are given on Figure II-2.
The results for the empirical formula and the 0.1 factor were very similar,
only one of which is plotted on Figure II-2. The estimated human popula-
O
tion exposures based on a standard human inhalation of 15 m of air per
day are given in Table II-5.
Average annual BSO concentrations are estimated to range from
3 1
3.75 yg/m (assumed background) to 11 yg/m . The median population expo-
3
sure concentration is 5.4 yg/m . Based on these exposure concentrations,
the exposed population would inhale an average of between 45 to 165 yg/day.
If we use the upper limit estimates for BSO concentrations, the human
inhalation exposure could rise to almost five times these values.
F. Considerations in the use of the Annual Average
as a Measure of Exposure to Coke-Oven Emissions
Exposure estimates in this report are given in terms of the daily
exposure averaged over a year. Statistically, this measure represents
the expected daily exposure; multiplied by 365, it gives the total
expected annual exposure. However, the statistical distribution of con-
centrations for a specific location is not symmetrical; rather, it takes
the form of many relatively small observations and a few relatively
larger observations. Examples of these distributions are given in Appendix
B. The averages for these types of distributions are much larger than the
median and, generally, only 20 to/0% of the observations might be expected
to exceed the mean in value. The geometric average rather than the arith-
metic average is a better measure to characterize the central location of
these distributions; however, exposure estimates based on the geometric
average are difficult to interpret. The overall arithmetic average was
found to be 1.8 times as large as the geometric average (Appendix B).
12
-------
35
30
o
.E
o>
a.
I
g
cc
UJ
o
§
o
o
01
OC
UJ
_
z
Q
IU
1
25
20
15
10
II TT
i
I I I I I I II
I I I I I II I
1 I I I I Ml
HIGH ESTIMATE, ASSUMES 0.5 fig BSO FOR
1 ng BaP EMITTED BY COKE OVENS
> ESTIMATE BASED ON FUNCTIONS DERIVED
FROM EMPIRICAL DATA
BACKGROUND CONCENTRATION
10e
IO7 2 K IO7
NUMBER OF PEOPLE EXPOSED TO INDICATED CONCENTRATION OR MORE
FIGURE 11-2. ESTIMATED POPULATION EXPOSURES TO BSO (Background, plus Emissions)
-------
Subgroup 1
Concentration
Range Cpg/m3)
10.8-11
8-9
7-8
6-7
5-6
4-5
3-4
Table II-5
ANNUAL AVERAGE DAILY BSD INHALATION FOR
PERSONS RESIDING NEAR COKE PLANTS
Subgroup Exposure
Range (pg/day)*
162-165
120-134
105-134
90-105
75-90 *
60-75
45-60
Number
of People
Exposed
1,800
2,420
8,400
54,000
1,034,000
13,900,000
2,106,000
Cumulative
Number
of People
Exposed
1,800
4,220
12,620
66,620
1,100,620
15,000,620
17,106,620
Estimated values are based on functions derived
from empirical data and include background and
emissions. Exposures are based on the annual averages.
14
-------
Calculations of averages and standard deviations are given in
Appendix B for BaP concentration data recorded over a number of different
days at a specific location. For most of these locations, the average
was found to equal the standard deviation. Thus, concentrations for an
individual worst case day could easily be three times the annual average.
Conversion methods given by Thuillier (1977) show that the 24-hour worst
case can be expected to be four times the average. This large difference
between the annual average and the worst case is quite logically explained
by the variations in meteorological conditions over a year.
G. Accuracy of Estimated Exposures
The accuracy of the exposure estimates are difficult to assess because
many relevant factors associated with the various monitoring programs are
unknown (e.g., the accuracy of the monitoring data and if the monitoring
days were selected at random). Another important source of error arises
from using a general model based on a sampling of coke plants to represent
all coke plants. The general model is not expected to give highly accurate
estimates for the concentrations at any location because only limited
plant-specific data went into the model. The model is, however, expected
to give fairly accurate estimates of overall national exposures because
it was formulated by using averages of parameters that represent a range
of meteorological, geographical, and emission control conditions.
The potential errors in estimated exposure concentrations were
addressed by using the model to predict average annual concentrations
for 1- and 3-km distances for a number of coke plants for which environ-
mental BaP monitoring data are available. The differences between the
observed and predicted concentrations then provide an estimate of the
accuracy of the procedure. The one-standard deviation value for these
differences was about 100%. This indicates that the predicted annual
average concentration for any location for a specific coke plant could
differ from the actual average by 100% or more. The standard deviation
for the overall national estimated exposures should be considerably less
than for individual locations because many sources of error are, in effect,
being averaged. The one-standard deviation value of the overall national
15
-------
estimated exposure concentrations is on the order of 10 to 100% of the
actual concentration. The size of the error depends primarily on the
number of coke plants having exposed populations in an exposure concen-
tration subgrouping. These accuracy estimates exclude potential errors
associated with the monitoring data. Note that these percent errors are
given relative to the actual exposure concentrations. Thus, a 100% error
indicates that the actual concentration may range from one-half to twice
the estimated value.
H. Other Potential Human Exposure Routes
There are potential human exposure routes for coke-oven emissions
other than inhalation. These include ingestion of contaminated food and
water and dermal contact. In addition, family members of occupational
workers might be exposed through particulates brought home on clothing
and other equipment such as lunch pails and automobiles. An assessment
of potential human exposures via these routes was excluded from the scope
of this study because they either appear to be much less significant
than the inhalation route or because of the lack of available data.
The dermal exposure would result from contamination of clothing or
the skin directly from atmospheric concentrations. Hence, the atmospheric
concentration estimates given in the summary tables of this report can be
used to provide estimates of dermal exposures.
Foods can become contaminated because of atmospheric fallout of parti-
culates or by way of contaminated water released by the coke plant. The
contamination may be on the surface of plants from fallout or included by
root-uptake. Animals can become contaminated by drinking contaminated
water, eating contaminated foods, or breathing contaminated air. Con-
tamination may also result from other man-made or natural sources.
Processed foods may contain additional contaminations from the combustion
of fuels used in smoking, roasting, or broiling. Foods in general have
been found to contain concentrations of polynuclear aromatic hydrocarbons
such as benz(a)anthracene, chrysene, and benzo(a)pyrene (Radding et al.,
1976). Table II-6 lists concentration levels of BaP in some foods. As
16
-------
Table II-6
BENZO(A)PYRENE CONCENTRATIONS IN FOODS
Concentration
Food
Cereals
Potato peelings
Potato tubers
Barley, wheat, rye
Cabbage
Spinach
Lettuce
Tomatoes
Fruits
Refined fats and oils
Fresh fish
Broiled meat and fish
Smoked fish
Smoked meat/sausage
Roasted coffee
Roasted coffee
Teas
Whiskey
(ug/ke)
0.3-0.8
0.36
0.09
0.2-4.1
24.5
7.4
2.8-12.8
0.22
2.0-8.0
0.9-15
<0.1
0.2-0.6
1.0-78.0
0.02-107.0
0.3-0.5
*
0.1-4.0
*
3.7-3.9
*
0.04
Reference
A
A
A
B
B
C
B
B
C
C
D
C
E
C
B
C
B
B
A - Shabad (1972)
B - Grummer (1968)
C - IRAC (1973)
D - Gorelova (1971)
E - Andelman (1970)
17
-------
expected, the BaP concentration of certain prepared foods is higher than
for other foods. At present, insufficient information is available to
assess the potential contamination of foods by coke-oven emissions.
Polynuclear aromatic hydrocarbons find their way into waterways
already absorbed onto aerosols or bacteria. Although their solubility
in pure water is essentially zero, they may exist in water in association
with organic matter or colloids (Radding et al., 1975). The IRAC (1973)
report lists BaP concentrations in drinking water of 0.0001 to 0.023 yg/1.
18
-------
Ill SOURCES OF COKE OVEN EMISSIONS
A
A. The Coking Process
Coke is a porous cellular residue from the destructive distillation
or carbonization of coal. It is used as a fuel and reducing agent in
blast furnace operations, and in foundries as a cupola fuel. Of the
approximately 60 million tons of coke produced annually in the United
States, 92% is used in blast furnaces, 5% in foundry operations, and 3%
in other types of industrial plants. Of the total coke production, approx-
imately 90% is produced by steel industry plants, 8% by foundry plants,
and 1% by beehive ovens.
Two basic processes are used in the production of coke: One recovers
vapors and other by-products from the coking process (by-product ovens),
and one does not (beehive ovens). The beehive oven, an older design,
that has been steadily replaced by the newer by-product design is excluded
from this analysis.
A by-product coke battery consists of 10 to 100 ovens made up of
chambers for heating, coking, and regeneration. Heating and coking flues
alternate with each other so that there Is a heating flue on either side
of a coking flue; the regenerative flues are located underneath.
The coking cycle begins with the introduction of coal into the coke
oven. This operation, called "charging," is carried out with a mechanical
"larry car" on rails on the top of the battery. The larry car receives
a load of coal from the coal bunker at the end of the battery. The car
moves down the battery to the pven to be charged. The lids on the oven
charging holes are removed, the larry car ±s positioned over the holes,
and the hoppers are emptied. During the charge, the oven is aspirated
The material contained in this section is summarized from the Federal
Register (October 22, 1976).
19
-------
by steam jets in the standpipes connecting the by-product gas collector
main with the oven. This operation, called "charging the main" is designed
to limit the escape of gas from the oven during the charging process.
After charging is completed, the lids are replaced and the aspiration
system is shut off.
The "coking time," the time required to produce coke from coal, is
governed by numerous factors, including the condition and design of the
oven heating system, width of the coking chamber, coal moisture, and
the nature of the coals being coked. The coking time for blast furnace
coke varies from 16 to 20 hours. Coking times for foundry coke are longer
than for blast furnace coke because coke of different physical character-
istics is required for foundry operations.
When the coal is coked, the doors on each side of the oven are re-
moved and the coke is pushed out. A large mechanically operated ram
attached to a pusher machine moves the coke out the opposite side of the
oven called the "coke side," through the "coke-guide" attached to the
door machine and into a railroad car called the "hot car" or "quench car."
The quench car moves down the battery to a "quench tower" where the hot
coke is cooled with water. The quenched coke is then dumped onto the
coke wharf, from which it is conveyed to the screening station for sizing,
then to the blast furnace, or removed for other purposes. When the doors
on the oven are replaced, the oven is ready to be charged again.
B. Environmental Emissions During Coking
Environmental emissions can occur during charging; during coking
from leaks in the doors and on the top of the oven; from the waste gas
stack; and during pushing and quenching, and from by-product processing.
Coke-oven emissions are described as a complex mixture of particulates,
vapors, and gases (Federal Register, October 22, 1976). (A detailed
assessment of the chemical and physical properties of these emissions is
being prepared as a separate document and, therefore, is not included
here.)
20
-------
Because of the effort and complexity that would be required in
characterizing all of the constituents of coke-oven emissions, various
surrogate measures have been used in the past. These usually are of
three types: TSP, BSD, and BaP. TSP is generally considered not to be
a specific enough measure for assessing total occupational health effects
(Federal Register, October 22, 1976). The concept of a surrogate measure
would be valid if it could be shown that that measure correlates well
with the presence of other emitted substances known to have adverse
health effects. Atmospheric concentration data recorded during 1964
and 1965 for Birmingham, which has several coke plants in the surrounding
area, showed that the correlation coefficient for BaP with 11 other
substances ranged form 0.65 to more than 0.99. For BSO with 11 other
substances the coefficient ranged from 0.58 to 0.88 (U.S. EPA, 1975),
indicating a fairly good association. These are given in Table III-l.
In an occupational exposure study, the atmospheric concentrations of 13
polynuclear aromatics (PNAs) and the total benzene soluble organics were
recorded. A correlation study was made of these data using logarithmic
transformations because the data followed a log-normal distribution
(NIOSH, 1974). The correlation of the PNAs with BaP and BSO are given
in Table III-2. Except for one case, all the correlation coefficients
exceeded 0.7, thus indicating a fairly good correlation. The correlation
of BSO with the 13 PNAs was generally better than the similar correlations
for BaP.
The occupational and the Birmingham correlation studies provide some
justification for using a surrogate measure rather than trying to identify
and control each of the PNA compounds emitted by coke-ovens.
C. Coke Processing Plants
In 1975, 57.2 million tons of coke were produced in the United States.
By-product ovens produced 98.7% of the total production, with beehive
ovens accounting for the remaining 1.3%. Approximately 90% of the coke
is used in blast furnace plants, whereas 2% is exported. The remainder
is primarily used in foundries. The yield of coke from coa], which
averaged 68.4% in 1975, has remained fairly constant during the past
decade (Sheridan, 1976).
*
TSP - total suspended particulates.
21
-------
Table III-l
CORRELATIONS AMONG PAH COMPOUNDS
IN THE AIR OVER GREATER BIRMINGHAM,
ALABAMA, 1964 AND 1965
Compound
Flu
Pyr
BaA
Chr
BeP
BaP
Per
BghiP
A
Cor
TSP
BSO
BaP
0.916
0.935
0.988
0.980
0.998
1.000
0.985
0.966
0.971
0.815
0.789
0.651
Compound
BSO
0.582
0.684
0.597
0.746
0.677
0.651
0.689
0.804
0.672
0.867
0.880
1.000
TSP
0.668
0.730
0.742
0.842
0.823
0.789
0.830
0.839
0.716
0.856
1.000
0.880
Source: U.S. EPA (1975).
22
-------
Table III-2
CORRELATION COEFFICIENTS AMONG LOG CONCENTRATIONS
OF 13 PNA AND BSO SAMPLES TAKEN WITHIN
FIVE COKE PLANTS
Compound
Flu
Pyr
BcA
Chr
BaA
BbF
BjF
BkF
BeP
BaP
DBahA
BghiP
Ant
BSO
BaP
0.797
0.740
0.569
0.857
0.824
0.776
0.768
0.813
0.950
1.000
0.694
0.855
0.892
0.914
BSO
0.914
0.862
0.713
0.936
0.909
0.884
0.894
0.915
0.922
0.914
0.725
0.875
0.905
1.000
Source: NIOSH (1974).
23
-------
In the United States, 65 plants produce coke. (Some authors list
only 62 by combining three pairs of closely co-located plants, where each
pair of plants are owned by the same corporation.) The 65 plants are
listed in Table III-3 which also lists the coal capacity and the 1974
coal consumption on a plant-by-plant basis. The plants consist of an
estimated 231 coke-oven batteries containing 13,324 ovens that have a
theoretical maximum annual productive capacity of 74.3 million tons of
coke. Because of depressed economic activity in 1975, the industry
operated at only 76% of this capacity. Coke production on a state-by-
state basis is given in Table III-4.
The Keystone Coal Industries Manual (1975) lists six beehive-coke
plants. These operate in two states (Pennsylvania and Virginia). Although
excluded from this analysis, they are listed in Table III-5.
Table III-5
DIRECTORY OF U.S. BEEHIVE-COKE PLANTS
Name or Location
of Plant
Pennsylvania
1. Mahoning
2. Daugherty
3. Laughead
Virginia
5. Vansant
6. Esserville
County
Company
Armstrong Caipentown Coal & Coke Co.
Fayette Bortz Coal Company
Fayette Ruane Coal & Coke Company
Buchanan
Wise
Jewell Smokeless Coal Corp.
Christie Coal & Coke
Source: Keystone Coal Industries Manual (1975).
24
-------
Table III-3
BY-PRODUCT COKE PLANT LOCATIONS AND CAPACITIES
State. City
N)
Alabama
1. Tarrant
2. Holt
3. Woodward
4. Gadsden
5. Thomas
6. Birmingham
7. Fairfield
California
8. Fontana
Colorado
9. Pueblo
Illinois
10. Granite City
11. Chicago
12. Chicago
13. South Chicago
Indiana
14. Chesterton
15. Indianapolis
16. Terre Haute
17. East Chicago
18. East Chicago
19. Gary
20. Indiana Harbor
Plant Name
Company
Tarrant Plant
Holt Plant
Woodward Plant
Gadsden Plant-
Thomas Plant
Birmingham Plant
Fairfield Plant
Fontana Plant
Pueblo Plant
Granite City Steel Div.
Chicago Plant
Wisconsin Steel Works
South Chicago Plant
Burns Harbor Plant
Prospect Street Plant
Terre Haute Plant
Plant No. 2
Plant No. 3
Gary Plant
Indiana Harbor Plant
Alabama By-Products Co.
Empire Coke Co.
Koppers Company, Inc.
Republic Steel Corp.
Republic Steel Corp.
U.S. Pipe and Foundry Co.
U.S. Steel Corp.
Kaiser Steel Corp.
CF&I Steel Corp.
National Steel Corp.
Interlake, Inc.
International Harvester Co.,
Wisconsin Steel Div.
Republic Steel Corp.
Bethlehem Steel Corp.
Citizens Gas & Coke Utility
Indiana Gas and Chemical Corp.
Inland Steel Co.
Inland Steel Co.
U.S. Steel Corp.
Youngstown Sheet and Tube Co.
Annual Coal
Capacity
(tons)
1,200.000
150,000
800,000
820,000
185,000
1,175,000
2,500,000
2,336,000
1,332,000
1,132,000
949,000
991,000
590,000
1974
Coal Consumption
(tons)
1,760,000
900,000
643,000
2,630,000
675,000
204,000
3,102,000
1,642,000
3,700,000
2,100,000
2,525,000
584,838
193,000
3,096,000
1,258,000
1,750,000
Sources: Keystone Coal Industries Manual (1975) and Varga (1974).
-------
Table III-3 (Continued)
State. City
Kentucky
21. Ashland
Maryland
22. Sparrows Point
Michigan
23. Detroit
24. Dearborn
25. Zug Island
(Detroit)
Minnesota
26. St. Paul
27. Duluth
Missouri'
28. St. Louis
flew York
29. Buffalo
30. Lackawana
31. Buffalo
Ohio
32. Iron ton
33. Hew Miami
34. Mlddletown
35. Painesville
Plant Name
Company
Annual Coal 1974
Capacity Coal Consumption
(tons) (tons)
Seraet
Sparrows Point Plant
Semet
Steel Plant
Zug Island Plant
St. Paul Plant
Duluth Plant
St. Louis Plant
Harriet Plant
Lackawana Plant
Donner-Hanna Plant
Ironton Plant
Hamilton Plant
Hiddletown Plant
Painesville Plant
Solvay Dlv., Allied Chemical Corp. 1,600,000
Bethlehem Steel Corp. 4,820,000
Solvay Div., Allied Chemical Corp. 900,000
Ford Motor Co. 1,800,000
Great Lakes Steel Dlv., National
Steel Corp- 2,850,000
Koppers Company, Inc. 250,000
U.S. Steel Corp. 850,000
Great Lakes Carbon Corp., Missouri 450,000
Coke & Chem Div.
Seraet-Solvay Div., Allied Chemical
Corp. 400,000
Bethlehem Steel Corp. 4,250,000
Donner-Hanna Coke Corp. 1,387,000
Semet-Solvay Div., Allied Chemical
Corp. 1,230,000
Annco Steel Corp. 934,000
Arnco Steel Corp. 748,000
Diamond Shamrock Corp. 215,000
4,100,000
3,385,000
210,000
-------
State. City
36. Portsmouth
37. Toledo
38. Cleveland
39. Massilon
40. Warren
41. Youngstown
42. Lorain
43. Campbell
Pennsylvania
44. Swedeland
45. Bethlehem
46. Johnstown
47. Johnstown
48. Midland
49. Aliquippa
50. Pittsburgh
51. Erie
52. Philadelphia
53. Pittsburgh
54. Clairton
55. Fairleas Hills
56. Moneseen
Tennessee
57. Chattanooga
Texas
58. Houston
59. Lone Star
Utah
60. Provo
Plant Name
Empire
Toledo Plant
Cleveland Plant
Massilon Plant
Warren Plant
Youngstown Plant
Lorain Cuyahoga Works
Campbell Plant
Alan Wood Plant
Bethlehem Plant
Rosedale Div.
Franklin Div.
Alloy & Stainless Steel
Div.
Aliquippa Plant
Pittsburgh Plant
Erie Plant
Philadelphia Plant
Neville Island Plant
Clairton Plant
Fairless Hills Plant
Wheeling
Chattanooga Plant
Houston Plant
E. B. Germany Plant
Geneva Works
Table III-3 (Continued)
Annual Coal
Capacity
Company (tons)
Detroit Steel Div. of Cyclops
Corp- 600,000
Interlake Inc. 438,000
Republic Steel Corp. 2,220,000
Republic Steel Corp. 250,000
Republic Steel Corp. 650,000
Republic Steel Corp. 1,500,000
U.S. Steel Corp. 2,700,000
Youngstown Sheet and Tube Co. 2,300,000
Alan Wood Steel Co. 803,000
Bethlehem Steel Corp. 2,210,000
Bethlehem Steel Corp. 550,000
Bethlehem Steel Corp. 1,680,000
Crucible Inc., Div. Colt
Industries 657,000
Jones and Laughlln Steel Corp. 2,250,633
Jones and Laughlin Steel Corp. 2,587,404
Hoppers Company, Inc. 290,000
Philadelphia Coke Division 715,400
Shenango Inc. 1,022,000
U.S. Steel Corp. 9,670,000*
U.S. Steel Corp. 1,800,000
Pittsburgh Steel Corp. 750,000
Chattanooga Coke and Chemicals Co. 204,400
Armco Steel Corp. 584,000
Lone Star Steel Co. 498,000
U.S. Steel Corp. 2.000,000
1974
Coal Consumption
(tons)
1,895,116
2,105,000
545,000
1,645,000
630,000
385,000
823,900
492,000
Based on a 1973 emission inventory.
-------
Table III-3 (Concluded)
West
61.
62.
63.
64.
State, City
Virginia
Weirton
Weirton
Fairmont
Follonsbee
Plant Name
Weirton Mainland Plant
Weir tor's Brown's Island
Plant
Fairmont Plant
East Steubenville Plant
Company
Weirton Steel Dlv., National Steel
Corp.
Weirton Steel Dlv., National Steel
Corp.
Sharon Steel Corp.
Wheeling-Pittsburgh Steel Corp.
Annual Coal
Capacity
(tons)
2,500,000
1,825,000
300,000
2,500,000
1974
Coal Consumption
(tons)
284,000
Wisconsin
65.
Milwaukee
Milwaukee Solvay Coke Co.
A Division of Picklands Mather and
Co.
347,000
INJ
00
-------
Table III-4
ESTIMATED SIZE AND PRODUCTIVE CAPACITY OF BY-PRODUCT COKE PLANTS
IN THE UNITED STATES ON DECEMBER 31, 1975
ro
vo
State
Alabama
California
Colorado
Illinois
Indiana
Kentucky
Maryland
Michigan
Minnesota
Missouri
New York
Ohio
Pennsylvania
Tennessee
Texas
Utah
West Virginia
Wisconsin
Undistributed
Total
Number of
Plants
7
1
1
4
6 (7)
1
1
3
2
1
3
12
12 (13)
1
2
1
3 (4)
1
-
62 (65)
Number of
Batteries
28
7
4
9
31
2
12
10
5
3
10
35
51
2
3
4
13
2
-
231
Number of
Ovens
1,401
315
206
424
2,108
146
758
561
200
93
648
1,795
3,391
44
140
252
742
100
-
13,324
Maximum Annual
Theoretical
Productive
Capacity (tons)
6,961,000
1,547,000
1,261,000
2,523,000
11,925,000
1,050,000
3,857,000
3,774,000
784,000
257,000
4,053,000
9,960,000
18,836,000
216,000
839,000
1,300,000
4,878,000
245,000
74,266,000
Coke
Production
in 1974
(tons)
5,122,000
C1)
1,912,000
9,073,000
C1
C1)
3,259,000
C1)
C1)
C1)
8,842,000
16,318,000
C1)
J1)
3,555,000
12,656,000
60,737,000
Included in Undistributed.
Source: Sheridan (1976).
-------
IV A METHOD OF ASSESSING BaP AND BSO
CONCENTRATIONS IN THE
VICINITY OF COKE-OVENS
A. General
All available ambient concentration data recorded for BaP and BSO
in the vicinity of coke-ovens are presented in Appendix A and analyzed
in Appendix B. These data (mostly for BaP) have been recorded in 15
locations, some of which contain several coke plants; as a result, approx-
imately one-third of the coke plants are represented. However, in many
cases, the data were recorded for only a few days and for only a few
sampling stations, thus making exposure estimates based solely upon them
unreliable. Moreover, it was necessary to devise some method of predicting
ambient concentrations for coke plant areas in which no atmospheric data
have been recorded. A procedure for doing this is given here. One approach
considered was to model the concentrations mathematically, basing it in
part on emission factors, amount of coal processed, and local meteorology.
When this approach was tried by the EPA (Youngblood, 1977), it was con-
cluded that, because of the uncertainties in characterizing the sources
themselves, definitive estimation of air quality impact of coke-ovens by
means of dispersion calculations is impossible at this time. The EPA
is currently working on developing better emission factors for coke-ovens.
Because these will not be available for some time, however, it was decided
to develop a procedure to extrapolate the available ambient data that have
been recorded in the vicinity of coke plants to other locations for which
no data has been recorded. When possible and when they seem reliable,
the actual recorded ambient concentration data have been used to estimate
population exposures.
The procedure that was devised required the following steps, which
are described in more detail in subsequent sections of this report:
30
-------
(1) Information on the type of environmental controls at coke plants
is evaluated to determine if facilities can be grouped by their
degree of control.
(2) The background concentrations estimated for each coke plant
location are those that would exist if the batteries were not
in operation.
(3) Existing ambient concentration data are evaluated to determine
if atmospheric concentrations can be expressed as a function
of distance from the coke plants.
(4) These concentration functions are evaluated to determine if
relationships can be derived from them, based on the amount
of coal processed and the degree of environmental controls.
(5) The functions are then used to estimate atmospheric concen-
trations in the vicinity of coke plants, with subsequent
estimation of human population exposures.
B. Categorization of Coke Plants by Emission Control
Emission factors are not well-developed for coking operations. Among
other factors, they are thought to be a function of process equipment,
environmental controls, and operating procedures. In theory, a different
set of emission factors exists for each battery. These battery emission
factors would be composed of emission factors for such sources as charging,
door leaks, pushing, topside leaks, by-product processing, quenching,
and the waste gas combustion stack.
The most detailed source of information on coke battery pollution
control compliance is based on a survey conducted by PEDCo during
September 1974 to April 1975 (Kuliyian, 1976). Among other items reported
in this survey was the compliance status of each plant or battery with
regard to charging, doors, waste gas combustion stacks, pushing, and
quenching. Compliance or noncompliance provide only a general indication
of environmental emissions. In addition, some of the batteries have
reduced their emissions since 1975. However, this time frame is consistent
with the dates when much of the environmental concentration data were
recorded.
31
-------
Weighting factors were assigned to each compliance status listed
in the PEDCo survey (in, out, at least one battery out, under a legal
plan, undetermined). These weighting factors are based on work performed
by EPA personnel, who were familiar with coke operations, to roughly
estimate BaP emission factors (Manning, March 18, 1977). This assignment
of weights assumes that an in-compliance status indicates low emissions
and that an out-compliance status indicates high emissions. Because the
EPA work gives emission factors for clean and dirty operations, the clean
factor was assigned to the in-compliance status and the dirty factor was
assigned to the out-compliance status. Plants having at least one battery
out of compliance and at least one battery in compliance were assigned
a weighting factor half-way between the out and in factors. These weight-
ing factors are given in Table IV-1. Note that the quenching weighting
factors dominate those for all other sources. Individual weights were
assigned to each compliance status within plants and summed to give a
total for each plant. These sums formed the basis for classifying plants
into two groupings. Plants for which no compliance data are available are
assigned to a separate group. Plant assignments are shown in Table IV-2.
This method of assignment can, and obviously has, led to some misclassi-
fications. At best, it should be regarded as a technique to be used to
form strata for statistical sampling. In theory, stratified samples
usually have increased precision over simple samples. As will be later
shown, atmospheric concentration versus distance from the coke plant
relationships for the two strata, when scaled for plant production, were
different. This indicates that the stratification method did, in this
case, provide increased precision.
C. Background Concentrations
Because substances emitted to the atmosphere by coke ovens can
also be emitted by other sources, it is necessary to consider atmospheric
concentrations as a sum of background plus coke-oven emissions. The coke
plants should only be assigned responsibility for their contribution to
the total.
32
-------
Table IV-1
ASSUMED EMISSION WEIGHTING FACTORS FOR
PLANT COMPLIANCE STATUS
Compliance Status
Emission Source
Charging
Doors
Pushing
Topside
Quenching
Waste gas stacks
In
1.5
16
*
N
1.6
175
N
Out
80
130
3
65
350
0.7
Undetermined
80
130
3
65
350
0.7
At Least One
Battery Out
40
73
1.5
33
260
0.4
Under a
Legal Plan
40
73
1.5
33
260
0.4
N - Negligible.
Topside compliance was assumed to be the same as door compliance.
33
-------
Table IV-2
CLASSIFICATION OF COKE PLANTS INTO EMISSION
CATEGORIES (1974-1975)
Plant
Number
*
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Classification
**
K
F
K
K
F
F
F
F
F
F
F
K
F
F
F
F
K
X
K
K
K
F
Plant
Number
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Classification
F
F
F
F
K
F
K
K
F
K
K
K
F
K
K
K
F
X
F
F
F
F
Plant
Number
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
Classification
F
F
F
F
F
F
F
X
F
K
X
K
F
K
F
F
X
X
X
F
F
**
Plant numbers correspond to plant names given in Table II1-3.
F indicates clean and K indicates dirty. The X indicates that
insufficient data were available to classify the plant.
34
-------
Background concentrations are difficult to assess because ambient
concentrations are seldom measured in an area when the coke-ovens are not
in operation. Moreover, upwind ambient concentrations, recorded near coke
plants, appear to have been influenced by the coking operations. In fact,
ambient atmospheric concentrations of BaP or BSD have not been measured
at all for many of the coke-oven locations. It is therefore necessary
to estimate background concentration by using data recorded at a sufficient
distance from the coke plant or by using data recorded at "similar"
locations. Either of these methods has inherent error. In addition,
background concentrations have been shown to vary with location within a
city and with season.
The available BaP atmospheric concentration data for cities without
coke plants are given in Appendix A. They were reviewed to identify a
"similar" noncoke plant location for each coke plant location. For
example, the average BaP concentration over Montgomery, Jacksonville, and
Charleston was used to represent Birmingham. The assumed annual average
BaP backgrounds are given in Table IV-3. They vary from 0.04 ng/m3 for
Houston to 1.6 ng/m3 for Pittsburgh.
D. Evaluation of Ambient Concentration Data for Coke Plant Locations
Available ambient data that were recorded in the vicinity of coke
plants have been evaluated to determine if it is possible to represent
the relationship of concentration mathematically as a function of distance
from a coke plant. An analysis of the results of the dispersion calcula-
tions performed by EPA (Youngblood, 1977) indicate that such a procedure
should be possible. An analysis of data given in Appendix B shows that
the BaP atmospheric concentration versus distance relationship about coke
plants can be represented by a double logarithmic function (power curve).
The procedure taken here is to modify the power curve formulation to
include allowances for background concentrations and for coke plant
capacities. The function selected is as follows:
35
-------
Table IV-3
ESTIMATED ANNUAL BACKGROUND CONCENTRATIONS
OF BaP FOR COKE PLANT LOCATIONS
Plant
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
BaP
(ng/m3)
0.4
0.4
0.4
0.4
0.4
0.4
0.4
1.2
0.6
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.4
0.8
1.1
1.1
1.1
0.4
0.3
**
Remarks
Montgomery, Jacksonville,
Montgomery, Jacksonville,
Montgomery, Jacksonville,
Montgomery, Jacksonville,
Montgomery, Jacksonville,
Montgomery, Jacksonville,
Montgomery, Jacksonville,
Average of 5 sites in the
area
Spokane
Hammond
Hammond
Hammond
Hammond
Hammond
Hammond
Hammond
Hammond
Hammond
Hammond
Hammond
Norfolk, Charleston
Riviera Beach, MD
Site 30 km away
Site about 30 km away
Site about 30 km away
NASN site
NASN site
Charleston
Charleston
Charleston
Charleston
Charleston
Charleston
Charleston
Los Angeles
*
Plant numbers correspond to plant names given in Table III-3.
Cities on locations used for reference concentrations.
36
-------
Table IV---3 (continued)
Plant A
Number
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
BaP
(ng/m3)
0.3
0.8
0.8
0.8
0.6
0.6
0.6
0.6
0.6
0.4
0.5
0.6
0.6
0.8
0.6
0.6
0.8
0.8
0.8
0.8
0.8
0.8
1.6
0.8
0.8
1.6
1.6
0.8
0.8
0.4
**
Remarks
NASN site
Site about 30 km away
Site about 30 km away
Site about 30 km away
Average of Pennsylvania and Ohio sites
Average of Pennsylvania and Ohio sites
Average of Pennsylvania and Ohio sites
Average of Pennsylvania and Ohio sites
Average of Pennsylvania and Ohio sites
NASN site
Site about 12 km away
Average of Pennsylvania and Ohio sites
Average of Pennsylvania and Ohio sites
Average of several Pennsylvania basins
Average of Pennsylvania and Ohio sites
Average of Pennsylvania and Ohio sites
Average of several Pennsylvania basins
Average of several Pennsylvania basins
Average of several Pennsylvania basins
Average of several Pennsylvania basins
Average of several Pennsylvania basins
Average of several Pennsylvania basins
Sites about 10 km away
Average of several Pennsylvania basins
Average of several Pennsylvania basins
Sites about 10 km away
Sites about 10 km away
Average of several Pennsylvania basins
Average of several Pennsylvania basins
Montgomery, Jacksonville, Charleston
37
-------
Table IV-3 (concluded)
Plant t BaP
Number (ng/m3) Remarks
58 0.04 Austin and Brownwood
59 0.04 Austin and Brownwood
60 0.5 Sites 20 to 30 km away
61 0.5 Charleston
62 0.5 Charleston
63 0.5 Charleston
64 0.5 Charleston
65 0.7 Hammond
38
-------
B + V A D* (1)
where, C is the atmospheric BaP at some distance (D) from
the coke plant.
B is the location's nominal background concentration.
V is the amount of coal processed annually by the
coke plant.
A and x are constants determined by regression.
D is the distance from the plant.
Least squares techniques were used to fit the available data to this
function to estimate values for A and x.
To extrapolate these functional representations from areas where data
are available to areas where data are not available, it is first necessary
to determine if the functional parameters (A and x) are consistent within
the emission control grouping given in Table IV-2. If they are found to
be consistent within groupings, average values can be used to represent
a group. The parameter designated as A in Equation (1) relates to the
atmospheric concentration resulting from coke-oven emissions at a distance
of 1 km from the plant. It could be estimated for more plants than the
slope parameter (x) because of the type of available data. For five
plants representing the better control classification group, the A
parameter had an average value of 2.8 x 10~6, whereas for eight plants
representing the poorer control group, the average was 7.3 x 10~6. There
were not enough data to show a difference in the slope parameter (x) for
the two control groupings. The average value for five locations was found
to be approximately -1.0. This is consistent with the dispersion modeling
data, which gave values of about -0.9 to -1.0. The average based on the
data will be used. Hence, this analysis suggests that Equation (1) be
used with a value of -1.0 for the parameter x. The value of the parameter
A will depend upon the grouping in which the plant is placed. For the
F grouping a value of 2.8 x 10~6 will be used, and for the K grouping a
value of 7.3 x 10~6 will be used.
39
-------
E. Relationship Between BaP and BSO Atmospheric Concentrations
Because so few data are available for BSO atmospheric concentrations
taken in the vicinity of coke production plants, an analysis has been
made to determine if the BaP data can be used to predict BSO atmospheric
concentrations, that is, to determine if some mathematical relationship
exists between BaP and BSO concentrations. Some of the potential hindrances
to establishing this type of relationship are that BaP and BSO are emitted
from other sources besides coke ovens and that the precise relationship
of BSO to BaP for coke battery emissions is unknown.
The available BSO concentration data (Appendix A) have been plotted
against the BaP data on Figure IV-1 for sampling sites that collected
both types of data. Average values were used. Data sources included
the 1972 NASN urban data, data recorded at sampling sites near coke
plants, and Maryland data. The data from the various sources appear to
form an increasing function with the cities without coke-ovens representing
the lower end of the scale and the data recorded near coke plants repre-
senting the upper scale. Figure IV-2 is a plot of only data found near
the coke plants.
Statistical regression techniques were used to fit mathematical
functions to various selected combinations of data given in Figures IV-1
and 2. The functional equation used was of the type:
BSO = A BaPX (2)
where, BSO is the atmospheric BSO concentration (ug/m3)
BaP is the atmospheric BaP concentration (ng/m3)
A, x are constants.
The values of the constants were found to be as follows:
Parameter
Data Set
All data
Data for noncoke locations
Data for coke locations
A
3.80
3.82
3.93
X
0.19
0.15
0.15
Data for coke locations with
BaP greater than 5 ng/m3 2.20 0.35
40
-------
40
"5
=L
I
§
u
8
o
E
i
I I I I I I I
1 I I I 1 I I I
I I I I I 1 II
IIIIIT
0.1
ALL DATA
DATA FOR NON-COKE OVEN
CITIES ONLY
I I 1 I I I I I
1 I I I 1 1 I I
I I I I I I 1 1
I 1 I I I < I 1
10
-1
10" 10'
ATMOSPHERIC B*P CONCENTRATION -
10'
10"
FIGURE IV-1. RELATIONSHIP BETWEEN BSO AND BaP ATMOSPHERIC CONCENTRATIONS
FOR ALL LOCATIONS
-------
100
I
o
H
in
u
§
o
in
at
o
UJ
a.
I
I I 1 I I I I I
1 1 I I I I I
1 I I I | I I I
1I I I I I L
Q NASN DATA FOR COKE
OVEN CITIES
0 DATA TAKEN NEAR
COKE PLANT LOCATIONS
DATA WITH
BaP> 5
DATA FOR NON-COKE
OVEN CITIES
II
I 1
ATMOSPHERIC BaP CONCENTRATION - ng/m3
FIGURE IV-2. RELATIONSHIP BETWEEN BSO AND BaP ATMOSPHERIC CONCENTRATIONS
FOR COKE OVEN LOCATIONS
-------
The regression coefficients (R2) were found to be around 0.4, indicating
a less than good fit to the data. The equation fit to all of the data
appears to underestimate the BSD concentrations for the higher BaP
concentrations. Consequently, an equation was fit to the data for coke
oven locations having BaP concentrations in excess of 5 ng/m3; this
equation had a higher slope. Based on occupational exposure data taken
within coke plants, there is evidence that the slope would continue to
increase as the BaP concentration increases (Antell, 1977). The occupa-
tional data indicate that 1 ng of BaP might correspond to 0.1 pg of BSO
with an upper bound of around 0.5 yg of BSO. However, due to a number
of possible chemical and physical processes and to dilution due to back-
ground, the relationship in the outside environment may not be the same
as in the occupational environment.
Three procedures are suggested for estimating atmospheric BSO
concentrations based on BaP concentrations. The first, which should
give an upper limit, follows: The estimated background BSO concentration
is added to 500 the BaP concentration that is due solely to coke-
oven emissions. The second procedure is similar to the first, except the
BaP concentrations due solely to coke-ovens are multiplied by 100 and
added to background concentrations. The third procedure is based on the
empirical functions fit to the data. All three of these procedures were
tried. The empirical procedure and the procedure in which 1 ng of BaP
corresponds to 0.1 pg of BSO were found to give almost identical results.
The results of the empirical procedure were used in the exposure estimates.
F. Population Exposure Estimates
The estimated population exposures to coke-oven emissions are given
in the summary section of this report and are not repeated here. However,
a general discussion of the approach is included.
Resident populations were estimated for five concentric geographic
rings about each plant. The radii of the rings were taken as 0 to 0.5,
0.5 to 1, 1 to 3, 3 to 7, and 7 to 15 km. These spacings were selected
to correspond to the shape of the concentration versus distance curves
43
-------
shown in Appendix B. Resident population for each of the geographic rings
was obtained from the Urban Decision Systems, Inc., Area Scan Report, a
computer data system that contains the 1970 census data in the smallest
geographic area available (city blocks and census enumeration districts).
Average annual BaP concentrations for each geographic ring were
estimated by using the empirical model for those coke plants for which
questionable or no monitoring data were available. This model was used
for 45 of the 65 coke plants. The plant specific best fit equations given
in Appendix B were used for locations for which sufficient monitoring
data were available. On a few locations, the monitoring data were used
to fix the concentration at a distance of 1 km from the plant and the
empirical model slope of -1.0 was used to estimate concentrations at
other distances. In all, some monitoring data were used in making
exposure estimates for 20 of the coke plants. For locations with more
than one coke plant, the population residing in overlaps of the geographic
rings was estimated by assuming a uniform population distribution. BaP
concentrations for the overlapping rings were estimated as the sum of
the applicable concentrations for individual coke plants.
Concentration subgroups were then developed, based on the range of
concentrations for the estimated exposures, and the total number of
residents for each exposure subgroup were calculated. The population
residing within a subgroup was excluded if its average annual BaP con-
centration due only to coke-oven emissions was less than 0.1 ng/m3.
These subgroupings were made for exposures to coke-oven emissions only
and to coke-oven emissions plus background concentrations.
Population exposures to BSO were calculated using che procedures
given in Section IV-E. These procedures estimate BSO exposures based
on estimated BaP exposures.
44
-------
Appendix A
AMBIENT ATMOSPHERIC BaP AND BSD CONCENTRATIONS
A. General
This appendix presents BaP and BSO atmospheric concentration data
recorded in the vicinities of coke manufacturing plants. Data are also
presented that give background concentrations for locations that contain
and do not contain coke ovens. All data used in this report are based
on high-volume filter samples. In addition, many of the sampling pro-
grams were conducted over a relatively few days within 1 or 2 consecutive
months; thus, they may not be entirely representative of an area's average
annual concentration. The implications of this sampling approach in es-
timating population exposures is described in further detail in Appendix B.
B. Atmospheric BaP and BSO Concentration Data Recorded Near Coke
Manufacturers
Atmospheric data that have been recorded near coke manufacturers are
described in the following paragraphs.
1. Monessen Area Air Quality Study, Pennsylvania
The Pennsylvania Department of Environmental Resources conducted
an air quality study to determine the distribution and magnitude of total
suspended particulates (TSP), benzene soluble organics (BSO), and benzo
(a)pyrene (BaP) concentrations in the Monessen area. The impact and
extent of air pollution due to sources at the Wheeling-Pittsburgh Steel
Corporation, Monessen, were evaluated, with sampling conducted from
April 6 to June 21, 1976, at three sites near the steel plant. Meteoro-
logical and selective sector actuator techniques were included in the
sampling program (DER, 1977A).
45
-------
A statistical summary of the data for the three sites is given
3
in Table A-l. The average TSP concentrations ranged from 79 to 166 yg/m ,
3
average BaP concentrations from 2.7 to 40.8 ng/m , and the average BSD
from 2.6 to 9.2 ng/m . Selective sector actuator sampling and a concen-
tration-wind direction frequency weighting technique all confirmed that
the steel plant is the major source of TSP and BaP. The average concen-
trations found in areas in the direction of winds coming from the plant
are between 1.5 and 3 times the average concentrations for winds from
all other directions (DER, 1977A).
2. Allegheny County, Pennsylvania
Three coke batteries are located in Allegheny County: U.S.
Steel Corporation in Clairton, Jones and Laughlin in Hazelwood, and
Shenango, Inc. on Nevell- Island. From April to September 1976, high-
volume particulate samples taken from 11 sites were analyzed for BaP.
The sampling schedule included two 10-week periods of four and two
samples per week, respectively (Ek, 1977).
Table A-2 shows the results obtained during the sampling.
The average BaP concentrations for the 11 locations varied between 1.64
3
and 51.95 ng/m . Eight additional samples were collected during first-
stage alerts at Liberty Borough in April and June 1976. Four were
collected over 24 hours and four over 8 to 12 hours. These data which
are given in Table A-3, show average BaP concentrations about six times
higher than for the regular sampling given in Table A-2.
3. Geneva Works, Utah
The data collected for BaP concentrations near the U.S. Steel
Geneva Works located near Prpvo, Utah, are summarized in Table A-4.
Eight stations within 4 km of the coke batteries showed average BaP
concentrations of 1.47 to 3.81 ng/m . Two background stations 20 to
3
30 km away showed average BaP concentrations of 0.12 and 0.83 ng/m .
46
-------
Table A-l
MONESSEN AIR STUDY, 24-HOUR SAMPLE CHARACTERISTICS
TSP
Station 2
Station 6
Station 7
BaP (ng/m3)
Station 2
Station 6
Station 7
BSD (Mg/m3)
Station 2
Station 6
Station 7
Sample
Size
29
28
31
29
28
31
29
28
31
Average
166.0
79.0
113.0
40.8
2.7
22.8
9.2
3.3
2.6
Range
27.0-360.0
22.0-165.0
26.0-300.0
0.3-206.4
0.2- 10.8
0.4-100.3
1.5- 25.4
0.6- 9.1
0.9- 19.3
Geometric
Mean
145.0
71.0
93.0
10.0
1.6
10.1
6.5
2.6
3.8
Standard
Deviation
1.76
1.64
1.91
7.60
2.78
4.57
2.34
2.01
2.36
Station 2 is 1 km ESE of the coke ovens.
Station 6 is 2.1 km NW of the coke ovens.
Station 7 is 1.8 km ENE of the coke ovens.
Source: DER (1977A).
-------
Table A-2
AMBIENT BaP CONCENTRATIONS FOR ALLEGHENY COUNTY, PENNSYLVANIA
(ng/m3)
CO
Site
Number
7102
5702
8601
8704
5802
7570
8602
6903
8790
5602
7004
Site Location*
10.5 km N of USS, 8 km E of J&L, 21.5 km SE of S
18 km NW of USS, 4.5 km N of J&L, 12 km ESE of S
0.5 km SE of USS, 14 km SE of J&L, 28 km SE of S
2.0 km NE of USS, 12 km SE of J&L, 27 km SE of S
18 km NW of USS, 6 km NW of J&L, 10 km SW of S
8.5 km NNE of USS, 9.5 km ESE of J&L, 24 km SE of S
1.5 km NNW of USS, 12 km SE of J&L, 26 km SE of S
12 km NW of USS, 1 km SSE of J&L, 16 km SE of S
2 km NE of USS, 13 km SE of J&L, 27.5 km SE of S
16 km NNW of USS, 5 km NNE of J&L, 15 km ESE of S
12.5 km N of USS, 6 km E of J&L, 18.5 km SE of S
Sample
Size
2
2
6
5
2
5
4
10
20
2
2
Average
1.64
2.62
13.63
15.00
2.29
6.12
28.17
3.95
51.95
3.78
1.66
Range
0.2- 3.1
0.3- 4.9
0.9- 67.7
0.3- 40.4
1.8- 2.8
0.9- 20.1
2.8- 83.5
0.5- 19.0
0.4-310.0
3.1- 4.5
1.4- 1.9
USS is U.S. Steel, J&L is Jones and Laughlin, and S is Shenango.
Source: Ek (1977).
-------
Table A-3
BaP DATA OBTAINED DURING FIRST STAGE ALERTS AT
LIBERTY BOROUGHSITE 8790
(ng/m3)
Sample
Number
1
2
3
4
24-Hour Data
427.9
277.8
320.4
171.0
Average 299.3
8-12 Hour
405.8
458.8
189.8
155.6
302.3
Data
Table A-4
ATMOSPHERIC BaP CONCENTRATIONS
Station
Number
1
2
3
4
5
6
7
8
9
10
GENEVA WORKS
(ng/rn^)
Location in Relation
to Battery
2.0 km NW
2.7 km NW
2.4 km NW
1.8 km N
1.3 km NE
2.4 km SE
4.0 km NW
2.6 km S
30.0 km S
20.0 km N
IN UTAH
Sample
Size '
9
6
9
11
11
11
3
11
11
9
NEAR THE
Average
2.08
3.81
3.15
2.41
3.13
1.63
2.10
1.47
0.12
0.83
Range
0.40-4.42
2.52-5.27
0.97-6.30
0.44-5.85
0.54-6.29
0.46-3.44
0.87-3.53
0.38-3.35
0.01-0.32
0.05-2.77
49
-------
4. Wayne County, Michigan
Three companies operating coke batteries are located in Wayne
County, Michigan: Solvay, Ford, and Great Lakes Steel. Ambient atmo-
spheric BaP concentration data were reported annually for seven sites
in the general area for 1971 to 1975 and are given in Table A-5. Annual
BaP concentrations for the various sites varied between 0.34 to 14.72
ng/m .
5. Buffalo. New York
Three companies operate coke batteries near Buffalo, New York:
Semet-Solvay, Bethlehem Steel, and Donner-Hanna. Atmospheric BaP con-
centration data were recorded from 1973 to 1974 on 13 sites, in addition
to data recorded at the National Air Surveillance Network (NASN) site.
These data, which are given in Table A-6, indicate the average BaP con-
3
centrations ranged from 0.45 to 27.10 ng/m .
6. Duluth, Minnesota
Thirty-eight samples for ambient BaP concentrations were ob-
tained from two sites within 3 km of the U.S. Steel coke batteries in
Duluth, Minnesota. These data are summarized in Table A-7. Average
3
BaP concentrations of 0.22 and 1.45 ng/m were found for the two sites.
When most of the samples were collected, the wind was blowing in the gen-
eral direction of the collection sites from the plant.
7. Gadsden, Alabama
The Republic Steel Corporation operates coke ovens in Gadsden,
Alabama. Atmospheric BaP concentrations were sampled at two sites within
1.6 km of the coke ovens during 1974, 1975, and 1976. The data from this
sampling which are summarized in Table A-8, indicate the annual atmo-
3
spheric BaP concentrations varied from 0.44 to 5.06 ng/m .
50
-------
Table A-5
AMBIENT BaP CONCENTRATIONS FOR WAYNE COUNTY,
MICHIGAN
(ng/m3)
Site
Number
02
04
05
06
08
11
NASN
1971
3.00
2.97
9.32
3.62
2.39
1.30
1.40
1972
2.44
3.14
5.95
2.62
2.56
1.32
1.90
1973
3.02
4.16
11.78
3.12
2.70
2.00
1.00
1974
1.46
1.70
10.83
0.52
0.44
0.34
1975
3.43
4.85
14.72
1.47
2.54
0.73
1.00
Average
2.67
3.36
10.52
2.27
2.13
1.14
1.33
No. 2 is 14 km NE of Solvay, 14.5 km NE of G.L.*, and 18 km ENE of Ford.
No. 4 is 7.2 km NNE of Solvay, 9.3 km NE of G.L., and 9 km NE of Ford.
No. 5 is 1.6 km N of Solvay, 4 km NNE of G.L., and 4.4 km E of Ford.
No. 6 is 15.3 km NNW of Solvay, 16.5 km NNW of G.L., and 11.7 km NNW of Ford.
No. 8 is 10.5 km SW of Solvay, 8.5 km SW of G.L. , and 9.3 km SSW of Ford.
No. 11 is 30 km SW of Solvay, 29 km SW of G.L., and 30 km SW of Ford.
G.L. - Great Lakes Steel.
51
-------
Table A-6
AMBIENT BaP CONCENTRATIONS FOR BUFFALO, NEW YORK
Ul
10
Sice
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
NASN
Site Location
3.1 km E of Beth, and 3.4 km SE of D-H*
1.9 km ESE of Beth, and 3.8 km S of D-H
3.8 km NE of Beth, and 1.5 km SE of D-H
1 km'ENE of Beth, and 2.8 km S of D-H
3 km N of Beth, and 1.4 km WNW of D-H
4.3 km NE of Beth, and 1.1 km ENE of D-H
5.6 km NE of Beth, and 2.1 km NNE of D-H
0.6 km W of Allied
1.2 km ESE of Allied
2.4 km NE of Allied
30 km SW of Beth, and 34 km SW of D-H
3.2 km NW of Allied
4 km SW of Beth, and 6.2 km S of D-H
8.8 km NNW of Beth, and 5.6 km N of D-H
Sample
Size
37
81
48
7
78
65
41
73
76
44
44
28
7
Average
5.99
8.99
11.38
27.10
2.78
9.10
7.29
1.29
3.80
3.74
0.82
0.45
1.29
0.70**
Range
0.27-30.5
0.26-48.7
0.06-68.4
2.76-48.8
0.05-23.8
0.20-65.6
0.07-46.2
0.05-21.2
0.13-81.4
0.01- 9.3
0.04-24.5
0.06- 3.1
0.40- 3.7
**
Beth, is Bethlehem Steel; D-H is Donner-Hanna.
Two-year composite.
-------
Site
Number
1
2
Table A-7
AMBIENT BaP CONCENTRATIONS'FOR DULUTH, MINNESOTA
(ng/ra3)
Distance from Coke Ovens
2.1 km SW .
2.7 km N
Sample
Size Average
18
20
Range
1.45
0.22
BDM-7.02*
BDM-1.25
BDM - below detectable minimum.
Source: Jungers (1977A).
Site
Number
1
2
NASN
Table A-8
AMBIENT BaP CONCENTRATIONS FOR GADSDEN, ALABAMA
(ng/m3)
Average Concentrations*
Distance from Coke Ovens
1974
1975 1976 3-Year
1.6 km E
1.1 km SW
Same as Station 1
5.06(0.44) 0.75 0.58 2.13(0.60)**
0.97 0.44 1.89 1.10
0.50 0.60 0.55
Sample size for each year for Sites 1 and 2 was 5.
**
i O
Excludes one high observation of 23.55 ng/m .
Source: Jungers (1977B).
53
-------
8. Birmingham, Alabama, Area
Five coke battery facilities, which are within about 20 km of
Birmingham, are located at Tarrant, Woodward, Thomas, Birmingham, and
Fairfield. Atmospheric BaP concentrations were sampled at Tarrant and
Fairfield during 1976, and NASN data are available for Birmingham. These
data are given in Table A-9. The average BaP concentrations ranged from
2.5 to 4.5 ng/m . BaP data were also recorded for five CHAMP sites in
the Birmingham area. These data are given in Table A-10.
9. Johnstown Air Basin, Pennsylvania
Two coke plants are located near Johnstown, Pennsylvania
(Bethlehem Steel's Franklin and Rosedale Divisions). An air quality
study was conducted from August through November 1975 to determine the
distribution and magnitude of TSP, BSO, and BaP concentrations in the
Johnstown area. Concentration data were obtained for eight sampling
sites 0.6 to 7.8 km from the Franklin Works (Table A-ll). BaP concen-
trations ranged from 85.3 ng/m for the site nearest the Franklin plant
to 3.6 ng/m for the site farthest from the plant.
Wind-actuated sampling was also conducted for TSP, BSO, and
BaP. For all three, in-sector sample concentrations were almost double
the out-sector concentrations.
10. Philadelphia, Pennsylvania
One coke facility is situated in Philadelphia (Philadelphia
Coke Diviison), and another two coke facilities are within 12 km of the
city at Alan Wood and Fairless Hills. Air quality data were collected
at four different times from November 1976 to January 1977 to determine
the distribution and magnitude of TSP, BSO, and BaP in Philadelphia.
Concentration data were obtained for 13 sampling stations about 2 to
14 km from Philadelphia Coke Division. These data are summarized in
Table A-12. The average BaP concentrations for the 13 sampling sites
j
ranged from 0.97 to 4.70 ng/m . BSO average concentrations ranged from
3
3.05 to 8.56 ug/m .
54
-------
Table A-9
AMBIENT BaP CONCENTRATIONS FOR BIRMINGHAM, ALABAMA
(ng/ra3)
Site
Number
1
2
NASN*
Distance from Coke Plants
Tarrant (0.5 km NW)
Fairfield (0.5 km ESE)
Sample
Size
2
3
Average
4.46
2.79
2.50
Range
0.06-8.86
1.10-5.31
__
1974 sample composite.
Source: Jungers (1977B).
Table A-10
CHAMP SITE AMBIENT ATMOSPHERIC BaP DATA FOR THE
BIRMINGHAM AREA (1975 Data)
Site
Number
304
305
306
307
323
331
Distance from Batteries (km)
Fairfield
11.4
25.8
10.3
4.9
16.4
13.4
Birmingham
3.8
5.0
10.8
18.4
2.4
3.2
Tarrant
2.4
13.0
13.4
21.4
1.9
5.4
Sample
Size*
6
12
12
6
12
12
BaP (ng/m3)
Average
4.2
1.8
1.5
2.4
2.9
3.5
Range
0.7-9.2
0.6-3.7
0.4.-4.0
0.9-4.3
1.2-6.6
1.4-5.5
Number of months for which data are available; for
individual months data were generally collected for
25 to 31 days.
55
-------
Ln
O>
Table A-ll
AMBIENT BaP, BSO, AND TSP CONCENTRATIONS FOR JOHNSTOWN, PENNSYLVANIA
Site Discance from Franklin Number of
Number Coke Ovens Samples
1 7.8
2 3.8
3 2.9
4 1.0
5 4.6
6 3.4
7 0.6
8 1.9
km
km
km
km
km
km
km
km
WSW
W
SW
NNE
SSW
SSW
ESE
SE
30
32
33
32
28
31
34
31
BaP
Average
3.6
13.
7.
23.
6.
6.
85.
19.
8
7
4
0
8
3
9
(ng/mj)
Range
0.
2.
0.
3.
1.
1.
1.
1.
5- 15.
0-110.
9- 41.
6-246.
5- 11.
4- 24.
5-575.
2-102.
4
9
8
6
0
5
9
9
Average
TSP*
32
70
71
142
55
58
179
70
(pg/mj)
BSO
2.
5.
5.
9.
3.
4.
14.
5.
2
5
4
7
9
1
1
6
*
Geometric mean.
Source: DER (1977B),
-------
U1
Table A-12
AMBIENT BaP, BSO, AND TSP CONCENTRATIONS FOR PHILADELPHIA, PENNSYLVANIA
Site Location from Philadelphia Number of
Number Coke Ovens Samples
1 14.1 km SW 4
2 3.5 km SW 3
3 19 km SW 4
4 12.5 km WSW 4
5 13.7 km WNW 4
6 9.3 km NNE 4
7 5.8 km W 4
8 10 km SW 4
9 2 km WNW 4
10 8.8 km SW 4
11 13 km SSW 3
12 10 km SW 3
13 5.2 km SW 4
NASN
BaP
Average
3.82
1.61
2.27
0.97
1.34
2.54
4.24
1.96
2.78
4.42
3.68
4.70
4.10
2.10
(ng/m3)
Range
2.09-8.81
0.82-2.26
1.09-5.35
0.21-1.81
0.35-2.31
1.38-4.53
1.90-6.29
1.38-2.31
1.26-4.46
2.28-7.15
1.55-6.70
2.44-8.06
1.57-9.90
_
Average
TSP
76.5
130.5
102.5
44.8
36.3
48.0
85.5
60.3
60.0
109.7
133.3
95.0
102.3
_
(pg/m3)
BSO
5.44
4.00
5.75
3.05
3.11
4.15
7.22
4.77
6.42
7.76
8.05
8.56
6'.59
4.66
Source: Lazenka (1977).
-------
11. Granite City. Illinois
BaP was measured at eight sampling sites between 0.5 to 3.5 km
from the National Steel coke ovens in Granite City, Illinois. The data
obtained during this sampling, which are summarized in Table A-13, indi-
cate that average atmospheric BaP concentrations for the stations ranged
from 2.6 to 12.2 ng/m .
More recently TSP, BSO, and BaP data have been obtained for 3
days on two sites within 0.8 km of the coke batteries (Table A-14). BaP
measurements from individual observations ranged from 1.6 to 278 ng/m .
12. Houston, Texas
Atmospheric BaP samples were obtained from seven sites located
up to 5.5 km from the Armco Steel coke ovens situated in Houston, Texas.
Samples were recorded at various times from 1973 to 1976. The data
summarized in Table A-15 show that average concentrations by site varied
3
from 0.03 to 0.28 ng/m . These concentrations are much lower than those
recorded at similar distances from other coke-oven locations, perhaps
indicating any of the following factors:
Good emission control.
Faulty measurement techniques.
All samples recorded upwind.
Ovens not in operation when measurements
were recorded.
13. Cleveland. Ohio
King et al. (1976) have reported on atmospheric BaP concen-
trations for a number of sites in Cleveland, Ohio. These data are
summarized in Table A-16. The geometric means are given rather than
the arithmetic means.
14. Sparrows Point, Maryland
The Maryland State Division of Air Quality Control measures
ambient BaP and BSO concentrations for many sites within the state,
58
-------
Table A-13
AMBIENT BaP CONCENTRATIONS FOR
GRANITE CITY, ILLINOIS
(ng/m3)
Site
Number
LN1
NW2
008
006
007
009
010
Oil
Distance from Sample
Coke Ovens Size
0.7 km N 3
0.6 km SSW 3
1.1 km NE 2
2.4 km WNW 2
1.8 km NW 2
1.5 km W 1
3.5 km WNW 2
2.9 km WSW 2
Table A- 14
Average
8.60
4.83
2.65
8.15
5.20
3.50
12.15
7.15
Range
1.1-16.5
0.6- 9.5
1.8- 3.5
8.0- 8.3
3.9- 6.5
-
1.8-22.5
3.9-10.4
ADDITIONAL ATMOSPHERIC AMBIENT DATA FOR
GRANITE CITY, ILLINOIS
Station
1
2
Distance from
Coke Ovens Pollutant 1
0.8 km N TSP (gg/m3) 113
BSD (pg/m3) 4.9
BaP (ng/m3) 2.1
0.5 km S TSP (yg/m3) 193
.BSO (pg/m3) 17
BaP (ng/m3) 124
Day
2 3
344 268
18 14
278 202
113 83
4.2 0.5
2.6 1.6
Average
238
12
161
130
7
43
59
-------
Table A-15
AMBIENT BaP CONCENTRATIONS FOR
HOUSTON, TEXAS (ng/m )
Site
Number
256034
256015
233006
256017
256019
256028
256005
Distance from
Coke Ovens
1.0 km NW
0.9 km NNW
2.2 km NE
0.8 km W
2.2 km WSW
0.8 km SSE
5.4 km SW
Sample
Size Average
6
5
4
6
4
7
1
0.17
0.15
0.05
0.03
0.33
0.28
0.16
Range
0.05-0.62
0.07-0.35
0.02-0.11
0.02-0.05
0.04-1.00
0.05-0.34
60
-------
Table A-16
AMBIENT BaP CONCENTRATIONS FOR
CLEVELAND, OHIO
Site Location from
Number Coke Battery*
1 0.8 km N
3 4.8 km SW
4 4.4 km NE
5 4.4 km SE
6 12.0 km NE
7 7.2 km W
8 6.8 km SW
9 1.0 km SE
10 6.0 km NNE
12 13.2 km ESE
13 4.4 km S
14 9.6 km SE
15 3.2 km W
17 6.4 km NE
20 16.8 km NE
21 4.0 km NNW
Sample
Size
21
37
23
28
22
38
28
30
33
32
23
22
21
32
19
22
BaP
Geometric
Mean**
1.40
0.62
0.64
0.58
0.71
0.46
0.44
3.60
0.74
0.43
0.85
0.47
0.51
0.91
0.50
1.10
(ng/m3)
Maximum
41.0
3.1
15.0
3.3
3.0
2.1
2.3
130.0
7.2
2.0
14.0
3.7
3.5
49.0
6.9
17.0
Locations are only approximate.
**
Arithmetic mean not reported.
Source: King et al. (1976).
61
-------
four of which are located within approximately 12 km of the coke batteries.
Data are given in Table A-17.
15. Chattanooga, Tennessee
As part of the CRESS and CHAMP programs, BaP samples were
collected for nine sites in the Chattanooga area. These data are sum-
marized in Table A-18.
C. Ambient Background BaP and BSO Concentration Data
Because coke ovens are not the only sources of BaP and BSO concen-
trations in the atmosphere, the coke oven contributions must be placed
in perspective with each area's nominal background concentrations. Data
are presented here for ambient background concentrations measured in
cities in which coke ovens are located, cities without coke ovens, and
remote rural areas.
1. NASN Air Quality System Data
NASN routinely monitors suspended particulate concentration
levels in urban and nonurban areas, generally reporting them as quarterly
composites for stations in the network. The composite, which pools all
samples collected during the quarter, assists in generating sufficient
material for laboratory analysis.
Before 1971, BaP analysis was made for more than 120 sites
per year. For 1971 and subsequent years, the sites were limited to 40
because of time and resource restrictions. These 40 sites were selected
to update BaP concentrations in cities with and without coke ovens.
Three sites were selected in National Parks to provide nunurban back-
ground readings (U.S. EPA, 1974).
Annual average BaP concentrations for 1967 to 1975 are given
in Table A-19 for the 40 NASN sites. Table A-20 gives BSO data recorded
at these sites for 1971 and 1972. The BaP and BSO concentrations are
summarized in Table A-21. The BaP concentrations are generally less
than 0.1 ng/m for rural locations. Most urban locations without coke
3
ovens have average concentrations of less than 1 ng/m (the average
62
-------
Table A-17
AMBIENT ATMOSPHERIC BaP AND BSD CONCENTRATIONS
FOR SPARROWS POINT, MARYLAND
Distance from
Coke Batteries*
12 km N
7 km NNW
3 km W
4 km SSW
Sample
Size**
2
12
10
10
BaP (ng/m3)
Average
1.4
1.4
1.9
2.4
Range
1.1-1.7
0.2-4.4
0.1-2.6
0.4-5.4
**
Locations are only approximate.
k
Number of months for which data are available.
BSO
Average
6.5
5.4
4.8
Range
6.3-6.8
3.0-8.0
2.6-8.7
Table A-18
AMBIENT BaP AND BSO CONCENTRATIONS
FOR CHATTANOOGA, TENNESSEE
Site
Number
621
622
631
632
633
634
635
641
642
Distance from
Coke Ovens
7.6 km
8.9 km
20.2 km
15.2 km
16.4 km
23.8 km
14.3 km
13.0 km
15.1 km
Number of
Samples
12
12
12
12
12
12
12
10
12
BaP (ng/m3)
Mean
3.83
3.49
1.63
1.85
1.55
0.82
1.23
2.35
2.66
Range
1.0-8.4
0.4-8.5
0.2-3.6
0.2-5.9
0.1-4.2
0.0-2.7
0.1-3.0
0.2-8.6
0.2-5.6
BSO
Mean
3.69
4.51
3.04
2.93
2.33
1.73
2.66
3.26
3.60
(pg/m3)
Range
2.2-6.5
2.6-9.1
1.4-5.2
1.1-6.4
0.6-4.6
0.3-2.7
1.5-4.1
1.8-5.3
1.5-7.0
Number of months for which data are available; for individual
months data were generally collected for 20 to 31 days.
63
-------
Table A-19
ANNUAL AVERAGE AMBIENT BaP CONCENTRATIONS AT NASN URBAN STATIONS (ng/m3)
Location
Montgomery AL
Chicago , IL
Detroit, MN
New York, NY
Toledo, OH
Philadelphia, PA
Pittsburgh, PA
Shenandoah Park, VA
Charleston, WV
Grand Canyon, AZ
Gads den, NM
Gary, IN
Indianapolis, IN
Baltimore, MD
Trenton, NJ
St. Louis, MO
Youngs town, OH
Chattanooga, TN
Spokane , WA
1967
2.3
3.0
5.4
3.9
1.9
5.9
7.0
0.3
-
0.2
-
-
5.7
3.8
-
2.3
8.2
22.9
1968
2.9
3.1
5.1
-
1.8
2.9
6.3
0.3
4.6
0.2
2.4
-
4.1
2.3
1.4
-
5.6
7.4
-
1969
2.0
3.9
3.9
3.6
1.5
4.0
13.8
0.3
2.6
0.2
1.8
-
5.2
2.8
1.6
3.3
9.9
4.2
-
1970
1.3
2.0
2.6
3.0
1.4
2.4
5.9
0.2
2.1
0.1
2.5
-
2.3
2.1
0.8
-
7.1
5.5
-
1971
0.5
2.5
1.4
2.3
0.8
2.3
6.1
-
0.9
-
1.2
1.6
0.9
2.8
0.7
0.8
3.7
-
1.7
1972
0.5
1.3
1.9
1.8
0.4
0.9
10.6
0.1
0.7
-
1.2
1.2
4.9
1.3
0.5
0.6
3.2
9.9
1.5
1973
0.3
0.4
1.0
0.7
0.4
0.6
-
0.1
0.2
>0.1
0.8
0.3
0.4
0.4
0.1
0.2
1.1
-
0.4
1974
0.4
-
-
0.9
0.2
0.8
1.3
-
0.5
>0.1
0.5
0.5
-
0.5
-
0.3
1.9
-
-
1975
0.3
1.0
1.0
0.9
0.4
0.6
2.1
-
0.5
-
0.6
2.2
-
0.6
-
0.3
2.1
0.8
0.6
-------
Table A-19 (Concluded)
Ul
Location 1967
Milwaukee, WI
Birmingham, AL
Jacksonville, FL
Honolulu, HI 0.5
Terre Haute, IN 3.7
Ashland, KY
Baton Rouge, LA
New Orleans, LA 1.8
Dearborn, MI
Duluth, MN
Buffalo, NY
Cleveland, OH 2.9
Bethlehem, PA 2.9
Erie, PA
Houston, TX
Newport News , RI
Norfolk, VA 3.5
Seattle, WA 1.8
St. Paul, MN 2.3
Arcadia National PK, ME
Hammond , IN 2.5
1968
it. 7
-
2.9
0.6
-
9.3
-
1.6
-
2.7
-
3.0
2.1
-
-
-
4.9
2.0
1.8
0.3
2.1
1969
4.0
-
2.3
0.6
4.0
10.9
-
1.5
-
2.1
-
3.8
2.0
-
-
-
3.9
1.6
1.8
0.1
3.3
1970
2.5
-
1.4
0.2
2.8
6.7
-
1.1
-
1.1
-
2.8
2.7
-
1.2
-
1.8
1.5
1.0
0.2
1.7
1971
1.8
4.0
2.2
0.2
9.0
0.4
0.9
4.8
0.9
1.5
0.5
0.4
1.2
0.5
0.5
3.8
1972
3.6
2.3
0.4
0.1
1.1
8.5
0.2
0.4
0.6
19.1
1.5
1.3
0.8
2.4
0.4
0.3
0.6
0.5
0.5
0.3
1.4
1973
0.6
1.5
0.2
0.2
2.9
0.1
0.3
1.0
0.3
0.6
0.5
0.7
0.4
0.2
0.4
0.3
0.1
0.2
1974
2.5
0.4
0.3
0.1
0.3
1.7
0.2
0.8
0.1
0.6
0.2
0.2
0.5
0.1
0.4
1975
1.1
0.4
0.03
0.6
4.7
0.1
0.2
3.1
0.3
0.5
0.4
0.2
0.2
0.4
0.4
0.1
0.7
-------
Table A-20
Birmingham, AL
Gadsden, AL
Montgomery, AL
Grand Canyon, AZ
Jacksonville, PL
Honolulu, HI
Chicago, IL
Gary, IN
Hammond, IN
Indianopolis, IN
Terre Haute, IN
Ashland, KY
Batan Rouge, LA
New Orleans, LA
Baltimore, MD
Dearborn, MI
Detroit, MI
Trenton, NJ
Duluth, MN
KL VARIATIONS OF BENZENE SOLUBLE ORGANIC SUBSTANCES (u
1971
1
3.1
2.9
3.4
1.2
4.3
2.3
4.3
4.7
3.8
2.6
4.1
6.8
2 JE>
4.0
7.3
-
2.6
1.7
1.8
2
6.7
3.6
4.2
0.9
3.0
0.1
5.7
2.7
4.7
3.1
7.4
1.9
3.5
4.5
3.2
3.0
3.0
2.5
3
-
2.1
2.4
-
2.1
1.2
-
-
6.0
3.2
3.6
4.0
-
3.1
-
3.1
2.4
2.6
2.1
4
4.8
4.5
3.4
-
2.2
1.4
4.5
5.7
7.0
3.7
8.3
3.4
3.5
4.3
-
3.9
2.8
3.6
1
3.6
2.7
3.3
-
2.3
1.4
3.4
2.7
2.1
2.9
2.5
7.8
3.2
3.7
5.0
3.6
3.2
1.8
2.0
g/m )
1972
2
7.5
4.2
2.9
-
5.4
2.3
2.5
4.1
9.4
4.9
5.7
7.2
4.1
4.9
3.6
7.3
3.3
1.7
5.9
3
4.0
2.4
2.2
-
4.4
3.3
2.7
3.0
6.3
3.0
4.0
7.9
3.5
5.5
-
4.6
3.4
2.0
4.5
4
5.2
2.3
2.6
-
6.0
3.0
3.9
2.5
5.0
-
6.3
9.2
5.3
4.2
4.5
4.5
3.0
1.5
12.5
Average
4.99
3.09
3.05
1.05
3.71
1.88
3.86
3.63
5.54
3.34
4.37
7.33
3.43
4.05
4.87
4.38
3.10
2.14
4.36
-------
Table A-20 (Concluded)
1971
1972
St. Paul, UN
St. Louis, MO
Buffalo, NY
New York, NY
Cleveland, OH
Toledo, OH
Youngstown, OH
Bethlehem, PA
Erie, PA
Philadelphia, PA
Chattanooga, TN
Houston, TX
Newport News, VA
Norfolk, VA
Shennandoah, VA
Pittsburgh, PA
Seattle, WA
Spokane, WA
Charleston, WV
Milwaukee, WI
1
-
5.5
-
5.5
3.6
2.1
2.9
3.8
2.6
6.0
4.8
3.5
2.7
4.9
-
3.8
5.6
3.5
-
2
2.8
3.1
-
6.2
3.6
2.4
4.9
3.8
2.5
4.0
5.1
3.2
2.9
4.0
0.7
4.4
4.1
4.4
5.0
3.8
3
2.2
2.0
2.9
-
-
-
3.5
3.6
-
3.8
-
3.8
3.1
3.5
0.9
-
5.3
3.5
2.0
3.8
4
3.0
-
3.6
4.5
-
3.1
6.4
3.8
4.2
7.4
-
5.7
4.5
4.2
-
6.9
5.4
3.6
3.6
4.8
1
2.9
2.9
3.0
4.5
3.1
1.8
4.1
2.9
1.3
4.7
4.1
4.5
1.3
2.5
1.0
6.1
1.6
3.8
2.8
3.0
2
7.9
3.7
9.3
5.3
6.5
-
3.9
5.2
6.8
3.8
11.0
5.9
3.3
3.8
0.9
4.9
4.2
3.1
2.7
6.5
3
5.6
2.3
3.4
4.7
-
2.9
4.2
4.2
5.8
4.3
3.7
5.9
4.4
3.8
0.8
4.7
3.9
4.6
3.6
3.2
4
4.6
2.6
7.8
3.9
4.6
2.7
4.6
4.6
-
3.3
-
4.9
3.7
3.2
0.6
6.6
5.4
4.0
2.8
3.4
Average
4.14
3.16
5.00
4.94
4.28
2.50
4.31
3.99
3.87
4.66
5.74
4.68
3.24
3.74
0.81
5.34
4.44
3.81
3.21
4.07
-------
Table A-21
SUMMARIZATION OF AMBIENT BaP AND BSO DATA
Cities
Cities With Without Rural
Pollutant
BaP (ng/m3) 1975
data
BSO (yg/m3)
1971-72 data
ANNUAL
Year
1966
1967
1968
1969
1970
1971
1972
1975
Statistic Coke Ovens Coke Ovens Areas
Average
Sample size
Range 0
Average
Sample size
Range 2
Table A-22
BaP AVERAGES FOR
(ng/m3)
Cities With
Coke Ovens
4.74 (15)*
5.34 (15)
3.75 (18)
4.41 (23)
3.02 (21)
2.18 (11)
2.14 (19)
1.21 (21)
1.21 0.38 <0.10
21.00 13.00 3.00
.3-4.7 0.03-0.9 <0.10
4.21 3.75 0.95
25.00 12.00 2.00
.1-7.3 1.9-5.6 0.8-1.1
SELECTED CITIES
Cities Without
Coke Ovens
2.76 ( 7)
2.29 ( 8)
2.64 ( 8)
2.14 (11)
1.41 (11)
1.22 ( 8)
0.64 (11)
0.38 (13)
Number of cities included in average.
Source of 1966-1972 data: U.S. EPA (1974).
68
-------
3
is 0.38 ng/m ); however, areas with coke ovens generally have average
3 3
concentrations in excess of 1 ng/m with Ashland's 4.7 ng/m the highest
and Dearborn's 3.1 ng/m the next highest. Coke ovens are located in
both Ashland and Dearborn. The overall average for cities with coke
3
ovens is 1.21 ng/m .
The BSO concentrations were generally less than 5 pg/m . The
3
average concentrations of most urban locations range from 1 to 4 vg/m .
Ashland, Chattanooga, Pittsburgh, Buffalo, and Hammond have concentra-
3
tions exceeding 5 yg/m .
Table A-22 shows the change from 1966 to 1975 in BaP concen-
trations in the atmospheres for cities with and without coke ovens.
Both classes of cities have shown a reduction; however, the atmospheric
difference between the two types of cities has been fairly constant
since 1968.
2. Pennsylvania Air Quality System
The Pennsylvania Division of Technical Services and Monitoring,
Bureau of Air Quality and Noise Control has systematically surveyed air
quality since 1970. As part of this program, the division monitors sus-
pended and settleable particulates at 91 locations. Suspended particu-
lates are collected on a glass fiber filter with a high-volume air sampler.
Each sample represents the particulate matter filtered from approximately
3
2000 m of air over 24 hours. Samples are taken from midnight to mid-
night every 6 days (DER, 1977).
During 1976, samples taken by this surveillance system were
also analyzed for BaP concentrations. The yearly average for these data,
based on one day sampled per month, are given in Table A-23 by sampling
location within the air basin. The highest average annual concentration
3 3
was 56.38 ng/m for Montessen and the next highest was 17.10 ng/m for
Johnstown. Both locations have coking operations. The lowest average
3
concentration was 0.40 ng/m for Hanover Green.
69
-------
Table A-23
AMBIENT BaP CONCENTRATIONS FOR PENNSYLVANIA, 1976
(ng/m^)
Allentown-Eastern Air Basin
Allentown
Tatamy
Bethlehem
Easton
Bethlehem East
Emmaus
Allen Twp.
Northampton
(Basin average)
Beaver Valley Air Basin
New Castle
Bessemer
Koppel
Beaver Falls
Vanport
Rochester
Ambridge
Baden
Midland
Brighton
(Basin average)
Erie Air Basin
Millcreek Twp.
Erie Central
Erie South
Erie East
'early
Lverage*
0.71
0.80
1.11
1.86
1.46
1.29
0.55
0.76
1.08
3.06
1.41
9. -43
5.03
2.27
4.19
6.18
9.00
3.13
2.42
4.73
0.45
2.04
1.16
1.62
Monthly
Range
0.09- 2.30
0.11- 2.74
0.22- 4.15
0.39- 9.28
0.24- 6.34
0.10- 7.62
0.06- 2.24
0.10- 3.43
0.13-11.36
0.46- 2.21
0.30-78.08
0.42-12.65
0.16- 5.44
0.35-13.96
0.75-31.96
0.40-43.48
0.31- 8.60
0.34- 9.74
0.12- 0.87
0.26- 7.13
0.21- 3.77
0.23- 6.33
Based on one sample per month for 12 months.
Source: Dubin (1977).
70
-------
Table A-23 (Continued)
Yearly Monthly
Average Range
Harborcreek Twp. 0.60 0.13- 3.40
(Basin average) 1.20
Harrisburg Air Basin
Middletown 0.83 0.12- 2.10
Swatara Twp. 0.65 0.18- 1.42
Steelton 1.03 0.32- 2.98
Lemoyne 0.92 0.28-2.38
Susquehanna Twp. 0.90 0.13- 2.55
Harrisburg 0.81 0.15- 2.00
Summerdale 0.61 0.14- 1.60
(Basin average) 0.82
Johnstown Air Basin
Westmont 1.00 0.15-5.05
Johnstown North ' 17.14 0.31-75.54
Johnstown Central 4.41 0.24-10.69
E. Conemaugh 16.30 1.21-50.74
Johnstown South 4.78 0.32-23.01
Hornerstown 3.17 0.13- 8.16
(Basin average) 7.51
Lancaster Air Basin
Lancaster Twp. 0.54 0.15- 1.77
Lancaster General 1.01 0.21- 2.74
Lancaster East 10.82 0.19-122.7
Lancaster North 0.72 0.27- 2.53
Lancaster West 0.91 0.25- 3.10
Neffsville 0.68 0.11- 1.81
Manheim Twp. 0.73 0.12- 2.75
(Basin average) 2.28
71
-------
Table A-23 (Continued)
Yearly Monthly
Average Range
Monessen Valley Air Basin
New Eagle 2.78 0.31-7.51
Monessen 56.38 1.05-206.3
Lover 2.61 0.61- 9.66
Elco 0.96 0.12- 3.94
Brownsville 9.05 0.59-57.00
Charleroi 2.47 0.13- 6.99
(Basin average) 12.69
Reading Air Basin
Leesport 0.56 0.07- 1.-60
Reading South 0.94 0.18- 3.20
Shillington 1.02 0.10- 4.09
Sinking Spring 0.73 0.05- 2.26
Reading Central 0.83 0.17- 2.67
Temple 0.90 0.15- 3.73
Laureldale 0.94 0.20- 3.32
(Basin average) 0.85
Scranton-Wilkes-Barre Air Basin
Hanover Green 0.40 0.09- 1.04
Dickson City 1.35 0.18- 3.32
Jessup 2.00 0.15-13.70
Pittston 1.49 0.14- 3.60
Swoyersville 1.67 0.42- 3.67
Nanficoke 0.94 0.11- 3.26
Wilkes-Barre 1.82 0.19-9.00
Scranton 2.06 0.28- 4.25
Dupont 1.30 0.27- 2.41
Avoca 0.44 0.11- 0.97
West Nanticoke 0.79 0.14- 2.05
(Basin average) 1.32
72
-------
Table A-23 (Continued)
Yearly Monthly
Average Range
Southeast Pennsylvania Air Basin
Pottstown 1.06 0.36- 3.09
Bristol 0.91 0.20- 3.03
Willow Grove 1.05 0.32- 3.93
Dowingtown 0.69 0.19- 2.48
Doylestown 0.76 0.12- 3.21
Media 1.00 0.24- 3.11
Chester 0.56 0.14- 1.78
Perkasie 0.73 0.16- 2.53
Quakertown 0.48 0.08-1.84
West Chester 0.81 0.11- 2.62
Lansdale 1.36 0.18- 4.57
Conshohocken 2.06 0.40- 3.24
Phoenixville 0.80 0.12- 2.47
Morrisville 0.79 0.07- 2.55
Coatsville 0.64 0.07- 1.42
(Basin average) 0.92
York Air Basin
York East 0.98 0.17- 2.59
York Central 0.96 0.17- 3.13
West Manchester Twp. 0.78 0.12- 3.38
Manchester Twp. 0.41 0.07- 1.12
West York 0.77 0.19- 2.16
Springettsbury 1.15 0.11- 7.49
(Basin average) 0.84
Altoona Area
Altoona Central 3.49 0.29-17.10
Altoona East 5.80 0.31-22.20
(Area average) 4.64
73
-------
Table A-23 (Concluded)
Yearly Monthly
Average Range
Farrell-Sharon Area
Farrell 2.46 0.44- 8.54
Sharon 2.45 0.24- 9.22
(Area average) 2.46
Williamsport Area
Williamsport Central 1.02 0.23-4.14
Williamsport East 1.28 0.15- 8.56
(Area average) 1.15
74
-------
3. Charleston, South Carolina
BaP was analyzed for three collection sites in Charleston,
South Carolina, which has no coke ovens. The data are summarized in
3
Table A-24. The average concentration tor the city was 0.69 ng/m .
Table A-24
DISTRIBUTION OF BaP CONCENTRATIONS IN AMBIENT AIR
AT CHARLESTON, SOUTH CAROLINA*
(ng/m3)
Site
Number
1
2
3
Location
Radio Station WTMA
Queen St. Fire Station
Mt. Pleasant, Post Office
Sample
Size
22
22
22
Average
0.5711
0.7441
0.7448
Range
0.0028-1.2409
0.1693-1.6787
0.1995-1.9767
Total 66 0.6866 0.0028-1.9767
There are no coke ovens in Charleston.
Source: Spangler and de Nevers (1975).
'*. Maryland Atmospheric Data
The Maryland State Division of Air Quality Control reports
monthly composite BaP and BSO concentrations for many sites throughout
the state. Data, primarily for 1976, are summarized in Table A-25. The
Average annual BaP concentrations ranged from 0.43 ng/m for Harwood to
6 ng/m for Catonsville.
5. CHESS and CHAMP Site Data
Atmospheric BaP and BSO data have been recorded for a number
of CHESS and CHAMP sites through out the country. These data are sum-
marized in Table A-26. Average annual concentrations ranged from 0.63
ng/m for Thousand Oaks, California to 4.2 ng/m for one site in
Birmingham, Alabama.
-------
Table A-25
AMBIENT ATMOSPHERIC BaP AND BSD CONCENTRATIONS
FOR MARYLAND LOCATIONS
Location
Cumberland
Hagerstown
Adams town
Frederick
Myersville
Buckeystown
Glen Burnie
Harmons
Harwood
Linthicum
Odenton
Riviera Beach
Annapolis
Baltimore
Lexington and Gay
Sun Avenue
1900 Argonne
5700 Reisterstown
5700 Eastern
Samp le
Size*
12
12
7
12
7
8
11
12
12
12
12
12
11
12
12
12
12
12
BaP
Average
4.48
1.40
0.83
1.29
0.17
0.80
1.03
0.54
0.43
0.96
0.71
0.80
0.75
1.95
2.03
1.21
1.37
1.92
(ng/mj)
Range
0.40-20.22
0.20- 3.87
0.31- 2.12
0.18- 3.55
0.05- 0.55
0.16- 2.52
0.15- 2.82
0.09- 2.10
0.04- 1.13
0.15- 2.18
0.08- 1.75
0.10- 2.79
0.10- 1.83
0.41- 4.46
0.50- 4.43
0.27- 3.51
0.42- 4.76
0.38- 5.36
Sample
Size*
12
12
-
12
-
-
12
-
12
12
12
12
12
12
12
12
12
12
BSO 1
Average
8.44
4.74
-
4.80
-
-
4.84
-
2.93
3.99
3.69
4.25
3.76
6.27
6.88
4.21
5.26
6.38
'Mg/m )
Range
6.03-18.14
3.34- 7.40
-
3.48- 6.77
-
-
2.67- 7.76
-
1.52- 4.51
2.21- 7.49
1.93- 5.74
2.50- 6.14
2.29- 4.93
4.34- 8.64
3.35- 9.50
2.51- 7.24
3.25- 8.11
4.25-10.06
-------
Table A-25 (Continued)
Sample
Location Size*
Baltimore (Continued)
Fonthill St
Cockeysville Ind Pk
Cockeysville Police
Station
Police Barracks
3001 Eastern Blvd
Catonsville
Dundalk (8801 Wise Ave)
Edge me re
Essex
Fort Howard
Towson
Middle River
Dundalk (Kavanaugh Rd)
Dundalk (Reg. Voc. Training)
Westminster
Gaithersburg
Silver Spring
(1901 Randolph)
Kensington
12
11
8
12
11
12
10
10
12
10
10
2
11
10
12
10
10
11
BaP
Average
1.66
0.80
0.50
0.78
1.29
5.98
1.10
1.85
1.37
2.39
0.75
1.43
1.61
3.30
0.49
0.62
1.14
0.59
(ng/m3)
Range
0.37- 4.80
0.09- 4.83
0.12- 1.76
0.10- 2.06
0.18- 3.97
0.09- 2.08
0.56- 2.42
0.07- 2.60
0.19- 4.36
0.38- 5.41
0.09- 2.95
1.14- 1.71
0.53- 4.60
0.31-10.35
0.09- 1.49
0.09- 2.13
0.09- 5.80
0.09- 2.07
Sample
Size*
12
-
8
12
11
12
-
-
12
10
3
2
-
-
12
10
11
11
BSD
Average
5.24
-
4.11
4.79
4.89
4.02
-
-
5.35
4.79
5.48
6.52
-
-
3.07
3.60
4.83
4.26
(ug/mj)
Range
3.01- 8.68
-
2.15- 9.99
2.35- 7.40
3.27- 6.60
2.53- 6.22
-
-
3.00- 7.98
2.64- 8.66
4.16- 6.41
6.25- 6.79
-
-
1.95- 4.78
2.00- 5.17
2.51-11.69
3.05- 7.56
-------
00
Sample
Table A-25 (Concluded)
BaP (ng/m3)
Sample
BSO (yg/m3)
Location
Poolesville
Silver Spring (Rock
Creek Forest)
Rockville
Bethesda
Accokeek
Cheverly
Largo
Laurel
Orme
Oxon Hill
Laplata
Elkton
Cambridge
Salisbury
Size*
11
4
12
4
2
11
1
10
10
8
10
12
11
9
Average
0.46
1.83
0.96
1.25
1.06
0.65
1.30
0.52
0.49
0.70
0.34
1.02
0.62
0.58
Range
0.06- 1.40
0.58- 3.93
0.07- 4.57
0.71- 1.68
0.73- 1.38
0.17- 1.50
-
0.09- 1.37
0.08- 1.90
0.17- 1.50
0.17- 1.68
0.18- 2.73
0.18- 1.58
0.10- 1.59
Size*
12
-
12
-
-
11
-
10
10
8
10
12
11
10
Average
3.36
-
4.51
-
-
4.57
-
3.56
3.56
4.03
3.27
4.60
4.05
4.46
Range
1.89- 5.18
-
2.56- 8.89
-
-
3.45- 6.24
-
1.90- 5.31
2.51- 5.33
3.08- 5.80
2.38- 4.95
3.55- 6.76
2.55- 5.80
3.03- 5.50
Number of months of data used in calculating the average and range.
-------
Table A-26
ATMOSPHERIC BaP AND BSO CONCENTRATIONS FOR
CHESS AND CHAMP SITES (1975 DATA)
Location *
Charlotte, NC (1)
Charlotte, NC (2)
Riverhead, NY
Queens, NY
Brooklyn, NY
Bronx, NY
Ogden, Ut
Salt Lake City, Ut
Kearns, Ut
Magna, Ut
Vista, CA
Santa Monica, CA
Thousand Oaks, CA
Garden Grove, CA
Glendora, CA
West Covina, CA
Anaheim, CA
Sample
Size**
6
6
12
12
12
12
12
12
12
12
12
6
12
12
12
12
12
BaP
Mean
1.44
2.36
0.66
1.07
1.57
2.11
2.05
2.37
1.20
1.09
1.03
1.46
0.63
2.42
0.91
1.98
2.36
(ng/m )
Range
0.3-2.7
0.5-4.4
0.0-3.6
0.1-3.1
0.3-4.0
0.2-4.3
0.0-7.2
0.2-5.0
0.1-3.6
0.1-2.9
0.1-4.9
0.2-3.5
0.1-1.4
0.3-7.5
0.1-2.2
0.2-5.0
0.4-7.1
BSO
Mean
1.79
2.91
1.29
1.99
3.70
3.24
2.41
3.26
1.43
1.48
2.07
3.91
2.31
3.86
4.13
5.85
4.77
(ug/m3)
Range
1.1-2.3
2.2-3.8
0.6-2.0
0.9-2.8
1.1-7.6
2.0-3.6
0.7-8.8
1.6-7.7
0.7-3.2
0.5-3.4
0.8-6.7
1.1-6.1
1.1-4.8
0.8-11.9
0.5-6.5
2.6-9.5
1.6-11.2
*
Data for Birmingham and Chattanooga are given with the city
coke oven data in Tables A-8 and A-15, respectively.
**
Number of months for which data are available; sample size for
individual months generally ranged from 20 to 31 days.
79
-------
Appendix B
STATISTICAL EVALUATION OF BaP ATMOSPHERIC CONCENTRATION
DATA RECORDED IN THE VICINITY OF COKE PLANTS
A. General
This appendix presents a statistical evaluation of the BaP atmospheric
concentration data recorded in the vicinity of coke plants. Factors
addressed here include the following:
What is the statistical distribution for atmospheric BaP
concentrations over time at a given location?
What errors are introduced by using estimated annual
atmospheric concentrations, based on a small sample
size?
Can the average BaP concentrations around a coke plant
be described as a mathematical function relating average
concentration to distance from the plant?
B. Statistical Distribution of 24-Hour BaP Atmospheric Concentrations
Because of changes in meteorological conditions and other factors,
the atmospheric BaP concentration at a specified location in the vicinity
of a coke plant will vary from day to day. The day-to-day variations
in the recorded 24-hour concentrations form a statistical distribution.
The long-term concentration for a specified location is generally
characterized by some central parameter for the distribution like the
arithmetic or geometric mean or the median. Obviously, the atmospheric
concentration data have been found to follow a distribution having rela-
tively many small values, with a few observations ranging to fairly high
values. These are called skewed distributions, as contrasted with symmetricaJ
distributions. They are sometimes found to follow what is known as two-
or three-parameter lognormal distributions.
Figure B-l illustrates the cumulative statistical distribution for
BaP atmospheric concentrations from some sampling locations. Because the
80
-------
1000
i r
\ I r
n i i i i r
n
E
O
DC
z
UJ
O
§
O
£
U
£
UJ
i
100
10
O JOHNSTOWN STATION No. 7
H MONESSEN STATION No. 2
©
CD
j i
2%
10
20
40 60
PERCENTAGE
80
90
95
98%
FIGURE B-1. STATISTICAL DISTRIBUTION FOR ATMOSPHERIC
BaP CONCENTRATIONS
81
-------
plotted points approximate a straight line, the statistical distributions
might be approximated by a lognormal distribution. The central measure
that best characterizes this type of distribution is the geometric rather
than the arithmetic average. The geometric average for these types of
distributions is smaller than the arithmetic average.
The properties of the lognormal distribution should be used when
describing the probability that a particular BaP atmospheric concentra-
tion will occur at a specified location. However, the arithmetic average
should be used when estimating expected population exposures. That is,
the arithmetic average concentration when used with daily human ventilation
rates gives the expected daily inhalation exposure. This expected daily
inhalation exposure multiplied by 365 gives the estimated total annual
exposure. The point here is that the arithmetic average should be used
in estimating expected population exposures, and the properties of the
lognormal distribution should be used in estimating the probability of
a specified exposure.
Table B-l summarizes the arithmetic and geometric average and standard
deviations for samples recorded at a number of stations. It is of interest
and potentially useful that the coefficient of variation (standard deviation
divided by the average) has a value near 1 (i.e., the standard deviation
generally equals the average).
C. Precision of Estimates Based Upon Small Sample Sizes
Most of the ambient sampling data available for this study are based
on 24-hour samples collected during limited sampling days. The ambient
concentrations recorded for these dates, for a given location, are averaged
and used as an estimate of the annual average concentration for that loca-
tion. It is, therefore, desirable to know how well an estimated average
annual concentration approximates the actual annual concentration.
From a statistical viewpoint, it is first necessary to know if the
sampling dates or period of dates were selected at random. In fact,
sampling was probably conducted when people get around to it or are
forced to do it and not because of any particular meteorological or
82
-------
Table B-l
STATISTICAL SUMMARY FOR SAMPLING DATA TAKEN FROM A NUMBER OF LOCATIONS
Arithmetic
Sampling Location
Monessen, 2
Monessen, 6
Monessen, 7
Johnstown, 1
Johnstown, 2
Johnstown, 3
Johnstown, 4
Johnstown, 5
Johnstown, 6
Johnstown, 7
g Johnstown, 8
Utah, 1
Utah, 2
Utah, 3
Utah, 4
Utah, 5
Utah, 6
Utah, 7
Utah, 8
Utah, 9
Utah, 10
Gadsden, 1
Gadsden, 2
Duluth, 1
Duluth, 2
Sample
Size
29
28
31
30
32
33
32
28
31
34
31
9
6
9
11
11
11
3
11
11
9
5
5
18
20
Average
40.8
2.7
22.8
3.6
13.8
7.7
23.4
6.0
6.8
85.3
19.7
2.1
3.8
3.2
2.4
3.1
1.6
2.1
1.5
0.1
0.8
0.6
1.9
1.5
0.2
Standard
Deviation
58.9
2.8
26.3
3.3
19.8
7.5
43.2
3.0
5.0
112.0
28.0
1.3
0.9
2.0
1.6
1.9
1.0
1.3
1.0
0.1
0.8
0.6
1.4
2.0
0.3
Coefficient
of
Variation*
1.4
1.0
1.2
0.9
1.4
1.0
1.8
0.5
0.7
1.3
1.4
0.6
0.2
0.6
0.7
0.6
0.6
0.6
0.7
1.0
1.0
1.0
0.7
1.3
1.5
Geometric
Average
10.0
1.0
10.1
2.6
8.6
5.7
13.2
5.2
5.6
44.5
8.3
1.7
3.7
2.6
2.0
2.5
1.4
1.8
1.2
0.1
0.5
0.3
1.6
0.3
0.1
Standard
Deviation
7.6
2.8
4.6
2.3
2.5
2.2
2.5
1.8
1.9
3.6
3.9
2.2
1.3
1.9
2.0
2.2
1.9
2.0
2.2
2.9
3.4
1.8
1.9
13.2
17.7
*The standard deviation divided by the average.
-------
seasonal reasons. If this is the case, it might be assumed that the
sampling period was selected in a quasirandom manner.
The next point has to do with weighting the samples for individual
dates by the fraction of time the meteorological condition on that date
occurs during the year. This generally is not done because in some cases
the meteorological conditions at the time of sampling are not reported or
because representative sampling is not available for a range of probable
meteorological conditions. If it can be assumed that the sampling period
is taken at random and that no weighting of the samples is to be made,
the estimation reduces to a simple statistical random sampling problem.
In this case, the average of the available data becomes an unbiased estimate
of the average concentration over the year. However, the number of dates
for which data are available greatly affects the precision of the estimated
annual -average.
The precision of an estimated value is measured by its standard
deviation. For a simple random sampling problem, the standard deviation
for an estimated annual average is given by:
P = Sf,
where
P = the standard deviation of the estimated annual average.
S = the calculated standard deviation for the sampling data.
f = /365-n
V 365n
n = the sample size.
The factor labeled as f is called the finite sample correction factor,
some values of which follow:
Finite Sample
Sample Size Correction Factor
1 0.999
5 0.444
10 0.319
20 0.217
30 0.175
50 0.131
100 0.085
200 0.048
365 0.000
84
-------
Note that the finite sample correction factor reduces in size rapidly with
additional sampling when the sample size is small. Depending on the
standard deviation for the sampling data, one might reasonably want sample
sizes in excess of 30. Estimates based on sample sizes of less than 10
might be suspected of being quite imprecise.
D. Evaluation of Ambient Concentration Data as a Function of
Distance from Coke Plant Locations
Available ambiend data that have been recorded in the vicinity of
coke plants (Appendix A) are evaluated to determine if it is mathematically
possible to represent the relationship of BaP concentration as a function
of distance from the coke plant. To investigate the feasibility of an
approach using recorded ambient concentrations, the average atmospheric
concentrations have been plotted as a function of distance from the
coke plants (Figures B-2 through B-14). As might be expected, the
atmospheric concentrations decrease with increasing distance from the
coke plants, thus indicating that the coke plants are a possible source
of BaP. The moderate amount of scatter in these relationships is probably
due to such factors as the location of the sampling site with respect to
the coke plant, local meteorological conditions, and local geography.
In addition, because many of the areas have several coke plants, it is
difficult to characterize the contribution to the environment for an
individual plant. If ambient data are to be used to characterize human
exposure, it would be desirable to have data from many monitoring sites
located at different directions from the coke plant and to have data
recorded for each monitoring site over a large number of days. Much of
the recorded data do not meet these requirements; the number of sampling
stations by plant ranged from 1 to 16.
The relationship of average atmospheric concentration to distance
does appear to follow a mathematical function of the type:
C = ADB>
where
85
-------
DISTANCE PROM COKE PLANT - km
FIGURE 8-2. ATMOSPHERIC CONCENTRATIONS OF BaP FOR JOHNSTOWN. PENNSYLVANIA
86
-------
10
w&
'?, U.
s
B
±ttt
n-
-rf
ii
I
ffl
H-i H
H-rrt
m
^
m^
-Hn
4-U
H-T
-
1
§
8
w
DC
ui
X
m
rn r
±t
i
I:
-t
r ,
1
;
fit
T
rr
m
~
;"
I
_L
t
ir^S^^E
:r
EH
--r
-2
ti
i
Uj:
mf
ai
ttlttt
^ j I i
rr
1
0.1
DISTANCE FROM COKE PLANT - km
FIGURE B-3. ATMO»»HERIC CONCENTRATIONS OF BaP FOR GENEVA. UTAH
-------
too
RELATION TO SOLVAY
RELATION TO GREAT LAKES
A RELATION TO FORD
DISTANCE FROM COKE PLANT - km
FIGURE B-4. ATMOSPHERIC CONCENTRATIONS OF BaP FOR WAYNE COUNTY, MICHIGAN
-------
100
oo
i ^.L^L^I ITS * . \ n&t ; i H
RELATION IB UX STEEL
RELATION TO J&L
DISTANCE PROM COKE PLANT - km
FIGORE B-5. ATMOSPHERIC CONCENTRATIONS FOR BaP FOR ALLEGHANY COUNTY. PENNSYLVANIA
-------
I | ;
3T
ft-r
rm
H-
lit
Traq.':a.i z^iJi-' M .., . ^ .. .
LJ..J..I.I * - -. «-* i.. ,.,,,,!*,..,1 ,.:...., ...in _
O RELATSO TO MTHf LEM
D RELATED TO BONNER-HANNA
A RELATED TO ALLIED
XI1
_ij
,i
-,d
i i : I Ml!T
TT- ^-t-14' "
_
w
F
I
It
rr -
t
"ft
1
r
i
Lffl
Pffl
\
ALLIED
T^
t
f p:;
I
if
4
f
-t
"
±-
*
0.1 u
0.1
:
jfTtft:
i '- m,
I
1.0
DISTANCE FROM COKE PLANT -km
-L^ ...
10
FIGURE B-fl. ATMOSPHERIC CONCENTRATIONS OF BaP FOR BUFFALO, NEW YORK
90
-------
RELATIVE TO TARRENT
D MCIATIVC TO
A RELATIVE TO BIRMINGHAM
0.1
DISTANCE FROM COKE PLANT - km
FIGURE B-7. ATMOSPHERIC CONCENTRATIONS OF BaP FOR BIRMINGHAM. ALABAMA
-------
1«
; :
Pip
0.1
I
I
M
I
1L
g
-
I
;
r
-i
i
H-
±
Tft *L
c::
fi
O SMMKftM PflNOB 2
[J SAMPLING PERIOD 1
4
if
t
r
= = =Ff
T
I
t
JX
+1 jj^.
a H
-HI
tttt
!
4
fr
tm
B
i
1
!
FT!
t
A
I
6.1
1.0
DISTANCE FROM COKE PLANT - km
FIGURE B-8. ATMOSPHERIC CONCENTRATIONS OF BaP FOR GRANITE CITY, ILLINOIS
92
-------
» «
o
IT
DISTANCE FROM COKE PLANT - km
FIGURE B-9. ATMOSPHERIC CONCENTRATIONS OF BaP FOR SPARROWS POINT, MARYLAND
-------
C4 -f~U-
DISTANCE FROM COKE PLANT - km
FIGURE B-10. ATMOSPHERIC CONCENTRATIONS OF BaP FOR CLEVELEND, OHIO
94
-------
tT.[_-
:JR '"';'
1
iff
i«
r.'j
i
1 =
.
t
&
I
i-it
f
I!
T
ttt
tt
1
H--
T
I
_
T
±
I
i
u-
ffi
i i
5-t"
T
I
i
t-
! , H
-fl
r
\
-
I
s
I
^
P
r
i
±:
.«
i
m
0.1
0.1
f-
R
;
!:
1.0
10
DISTANCE FROM COKE PLANT - km
FIGURE B-11. ATMOSPHERIC CONCENTRATIONS OF BaP FOR MONESSEN. PENNSYLVANIA
95
-------
DISTANCE FROM COKE PLANT - km
f IGURE i-12. ATMOSPHERIC CONCENTRATIONS OF BaP FOR GADSDEN, ALABAMA
96
-------
m
l;;!!l
S:
a
i
Jii
.1
i
i
i
r
n-
"
:
j :.
f
ftt
ur
IT
l-t-
a
!
tr
i
:
ai
'!tf
lit
DISTANCE FROM COKE PLANT - km
FIGURE B-13. ATMOSPHERIC CONCENTRATIONS OF BaP FOR DULUTH, MINNESOTA
97
-------
00
oc
UJ
(J
8
y
ALL UPWIND STATIONS
0.1 t
DISTANCE FROM COKE PLANT - km
FIGURE B-14. ATMOSPHERIC CONCENTRATIONS OF BaP FOR PHILADELPHIA. PENNSYLVANIA
-------
C = the average concentration at some distance from the coke
plant.
A,B = model parameters fit by regression techniques.
D = the distance from the plant.
Least squares regression techniques have been used to fit the data to
this mathematical function for each coke plant for which data are avail-
able. The results of this evaluation are given in Table B-2. The regres-
sion coefficients given in Table B-2 indicate how well the data fit the
function. For most cases, the regression coefficients ranged from 0.5 to
1.0, suggesting fairly good approximations. The coke plants with only
two monitoring stations showed, as expected, regression coefficients
with a value of 1. The model parameters based on actual ambient data can
be compared to similar fits to the atmospheric modeling data conducted by
Youngblood (1977). Two conditions are given in Table B-2 for comparison
(one for a dirty plant and one for a clean plant).
The magnitude of the model parameter A relates to the amount of BaP
emitted from the source, and the model parameter B relates to decay in
the concentration versus distance function. Note that for all coke loca-
tions with more than sampling stations on Table B-2 the B parameter varies
between -0.32 to -1.42 with an average of -0.9. When locations that have
more than one coke plant are a]so excluded, the parameter has an average
value of -1.0. This compares favorably with the modeling data, which
give a value of B of about -0.95.
99
-------
Table B-2
ESTIMATED PARAMETER VALUES FOR REGRESSION
APPROXIMATIONS TO AMBIENT DATA
Location
Johnstown
Gadsden
Duluth
Monessen
Utah
Wayne County
Buffalo - Beth.
Buffalo - D.H.
Buffalo - Allied
Cleveland
Pittsburgh - USS
Pittsburgh - J&L
Tarrant
Granite City
Sparrows Point
Fairfield
Dirty Plant Model
Clean Plant Model*
Number of
Stations
8
2
2
3
10
6
5
7
4
16
5
6
4
10
4
2
5
4
Model Parameters Regress:
A
42.86
1.28
379.39
49.99
4.70
13.09
15.96
8.40
1.96
12.42
22.00
4.32
3.98
11.50
3.07
2.67
135.84
60.66
B
-1.22
-1.62
-7.50
-2.92
-0.84
-0.69
-0.99
-0.75
-0.33
-1.42
-0.32
-0.37
-0.33
-0.57
-0.33
-0.07
-0.96
-0.95
Coefficii
0.96
1.00
1.00
0.64
0.76
0.92
0.79
0.60
0.06
0.72
0.16
0.52
0.75
0.10
0.61
1.00
0.99
0.98
Uses only data for distances of 1 km and greater.
100
-------
Appendix C
DETAILED ESTIMATES OF POPULATIONS AND BaP CONCENTRATIONS
FOR INDIVIDUAL COKE FACILITIES
This appendix includes the detailed population and BaP concentration
estimates for each defined geographic population ring (i.e., 0 to 0.5,
0.5 to 1, 1 to 3, 3 to 7, and 7 to 15 km) about each coke facility.
These estimates are given in Tables C-l and C-2. The concentrations
include the summation of atmospheric concentrations from both the coke-
ovens and background. The population within a geographic ring was con-
sidered not to be excessivly exposed to coke-oven emissions if its
i
estimated average annual BaP concentration was less than 0.1 ng/m . For
some locations, several separate coke facilities are located within 15 km
of each other. In these cases, it was necessary to estimate geographic
population ring overlaps and total ring BaP concentrations.
101
-------
Table Ol
DETAILED BaP POPULATION EXPOSURES (ng/m )
Coke Emissions Plus Background
Distance from Coke Facility (km)
Site No.'
1
2
3
4
5
6
7
B
9
10
11
12
13
14
IS
17-18**
19
20
21
22
23
24
23
26
27
28
29
30
31
32
33
34
0 -
Population
388
0
0
478
1,656
975
0
0
0
0
827
0
2,618
0
57
512
0
0
99
552
11
0
0
0
991
0
3,368
P
0
1,184
0
6
0
0.5
Concentration
F
1.5
15.0
6.0
1.7
F
5.0
19.0
10.0
2.0
7.3
19.0
4.8
19.0
5.4
2.1
87.0
68.0
39.0
30.0
4.3
F
F
F
2.2
4.2
3.9
2.7
F
F
23.0
18.0
14.0
0.5
Population
1.756
0
0
916
532
5,279
0
1,416
0
7,244
7,307
0
2.494
0
8,176
3.059
53
33
482
0
3,416
3,008
2,197
51
3.901
0
2.450
202
3.113
2.537
2,373
2,608
0
- 1
Concentrat ion
F
0.9
8.2
2.0
1.1
F
4.2
10.0
5.6
i.4
4.2
10.0
2.9
11.0
3.2
1.5
47.0
37.0
21.0
16.0
3.5
F
F
F
1.3
2.0
2.0
2.1
F
F
13.0
9.7
7.9
1
Population
13.880
0
21,495
19.693
36,497
28.195
22.105
9,493
0
30,475
58,244
0
27,666
0
71,661
36,083
21,279
26,829
25,740
15,511
15,948
37,283
46,224
33,878
61,720
4,219
30,734
7,874
17,140
58,527
11,701
8,567
455
- 3
Concentration
F
0.6
3.3
0.7
F
3.1
4.5
2.5
2.0
'4.3
1.4
4.4
1.7
1.0
18.0
14.0
8.4
6.2
2.5
F
F
F
0.8
O.B
0.9
1.6
F
F
5.1
3.7
3.3
3
Population
114,873
5,843
80,440
36,345
158,506
120,414
120,356
51,629
0
46,890
248,247
2,828
187.310
16,253
207.269
42.851
51,533
122,882
70,004
35,072
97,933
289,066
322.403
172,788
255,071
7,036
224,719
111,218
138,521
257,202
55,346
57.955
749
- 7
Concentration
F
1.6
F
2.4
2.5
1.4
1,2
2.2
--
2.2
1.1
7.6
6.1
3.8
2.7
1.9
F
F
F
~
0.6
1.2
F
F
2.4
2.0
1.7
7 -
Population
278,354
7,802
196,316
21,099
283.180
297.002
256.857
161.178
52
704,796
1,153,057
14.649
1.223.577
73.329
444,465
24,392
520.919
265.439
637,625
90,904
481.929
1,190.455
1,274.124
892,126
727.327
76.894
714,782
601.056
531,748
533.320
37,066
83,131
17,189
15
Concentration
F
0.9
--
F
F
1.8
0.9
1.4
0.9
3.9
3.2
2.1
1.5
-
F
F
F
--
-
F
r
1.4
1.2
1.1
-------
Table C-l (Concluded)
Distance from Coke Facility (km)
Site Ho.*
35
36
37
38
39
40
41
42
43
44
45-46**
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61-62**
63
64
65
0 -
Population
0
0
663
1.530
0
0
0
13
0
0
3,960
1,804
0
0
0
2.628
2,433
0
1,024
2,374
0
1,185
6
0
19
0
0
3
1,300
0.5
Concentration
2.1
12.0
8.4
42.0
2.4
13.0
11.0
20.0
18.0
6.4
16.0
100.0
5.4
16.0
F
2.8
14.0
F
F
34.0
50.0
1.8
11.0
3.5
10.0
79.0
6.0
18.0
3.1
0.5
Population
0
0
1,597
4,228
71
0
1,986
2,858
4,074
0
1,593
870
0
0
10,170
10,041
4,526
75
1,365
1,757
1.559
5.483
0
0
0
0
0
0
4,100
- 1
Concentration
1.4
6.4
4.7
25.0
1.5
6.9
6.4
11.0
9.2
3.8
9.1
50.0
3.3
9.2
F
1.9
7.8
F
F
18.0
38.0
1.2
5.7
1.9
5.8
- 43.0
3.4
9.8
2.0
1
Population
10,963
5,352
22.961
73,580
17,663
20,260
45,234
72,578
33,723
26,142
42,961
35.460
573
19.922
115,372
56.243
54.405
31,991
16.819
30,671
22,763
16,595
4.866
1,046
3.044
0
0
20,971
43,000
- 3
Concentration
0.9
2.8
2.0
4.5
1.0
3.0
2.9
4.4
3.8
1>9
3.9
17.0
1.7
4.0
F
1.2
3.4
F
F
7.4
10.0
0.7
2.1
1.0
2.6
16.0
1.6
4.0
1.2
3
Population
30,503
34,067
155,445
399,565
40,897
82,028
147,771
378,615
115,698
110,829
64,472
51.502
12,276
61,371
396,226
87,797
486,896
122,412
83,779
117,606
42,498
81 , 104
99,968
2.219
28,887
3,570
0
39,901
219,000
- 7
Concentration
^
1.5
1.0
2.3
1.6
1.6
2.1
1.9
1.3
2.0
5.2
1.2
2.1
F
1.0
1.8
F
F
3.4
2.0
--
0.9
0.3
1.2
6.9
0.9
1.9
0.9
7 -
Population
50.851
24,213
258,780
859,264
152,877
96,129
133,429
860.216
164,425
734,938
229,293
38,180
91,115
116.901
748,696
40.701
1,685,267
632,088
407,475
354,575
80.359
117.888
350,402
8,902
72,123
8,598
1.410
65.031
HE
15
Concentration
__
1.0
0.7
1.0
1.2
1.3
1.2
1.0
1.4
2.2
1.0
1.4
F
1.3
F
F
2.0
0.4
3.4
0.7
1.1
Site numbers correspond to coke facilities listed in Table III-3.
Indicates that the two facilities were treated as though they were colocated.
F - Indicates that one or more coke facilities are located within 15 km of that facility.
Estimated concentrations are given in Table C-2.
Indicates that the atmospheric BaP concentration due to the coke facility was less
than 0.1 ng/m3.
-------
Table C-2
BaP EXPOSURES FOR PERSONS IN LOCATIONS HAVING
MORE THAN ONE COKE FACILITY
Location
Birmingham, Alabama
Detroit, Michigan
Buffalo, New York
Pittsburg, Pennsylvania
Exposed
Population
975
388
14,025
7,035
28,054
110,000
135,893
51
5,000
2,197
41,000
3,008
19,900
330,000
14,913
1,274,124
3,113
1,184
19,000
2,537
11,300
39,000
533,000
1,024
1,365
83,800
407,500
10,170
147,363
396,226
16,819
Exposure
Concentration
(ng/m3)
8.2
5.8
5.6
4.5
3.0
2.6
1.6
12.0
8.1
7,8
6.7
4.5
4.0
3.6
3.4
2.1
22.0
17.0
12.0
10.0
8.0
5.0
1.6
30.0
24.0
13.0
10.0
4.8
3.0
2.0
1.8
104
-------
BIBLIOGRAPHY
Andelman, J. B. and M. J. Suess, "Polynuclear Aromatic Hydrocarbons in
the Water Environment," World Health Organization, 43:479-508 (1970).
Antel, M., personal communication (Hay 1977).
Department of Environmental Resources, "Monessen Area Air Quality Study,
Monessen, Pennsylvania," Commonwealth of Pennsylvania (1977A).
Department of Environmental Resources, "Johnstown Air Basin Air Quality
Study, Johnstown, Pennsylvania," Commonwealth of Pennsylvania (1977B).
Department of Environmental Resources, "1976 Pennsylvania Air Quality
Surveillance System Air Quality Report," Commonwealth of Pennsylvania
(April 1977C).
Dubin, R., personal communication, Pennsylvania Department of Environmen-
tal Resources (May 1977).
Ek, A., private correspondence with B. E. Suta, Allegheny County Health
Department (May 2, 1977).
Federal Register, "Exposure to Coke Oven Emissions," Occupational Safety
and Health Administration, 4.1(206):46742-46790 (October 22, 1976).
Gorelova, N. D., P. P. Dikun, L. D. Kostenko, 0. P. Gretskaya, and A. V.
Emshanova, "Detection of the Possible Presence of 3,4-Benzopyrene
in Fresh Fish," Novosti Onkol, 8-12 (1971).
Grimmer, G., "Carcinogenic Hydrocarbons in the Human Environment,"
Deut. Apoth.-Ztg. 108(16):529-533 (1968).
Jungers, R. H., "BaP Analysis of Duluth Samples," memorandum to J. Manning,
U.S. Environmental Protection Agency (February 2, 1977A).
Jungers, R. H., "BaP Analysis, Alabama," memorandum to J. Manning, U.S.
Environmental Protection Agency (February 4, 1977B).
King, R. B., J. S. Fordyce, A. C. Antoine, H. F. Leibecki, H. E. Neustadter,
and S. M. Sidik, "Extensive One-Year Survey of Trace Elements and
Compounds in the Airborne Suspended Particulate Matter in Cleveland,
Ohio," Lewis Research Center (1976).
Keystone Coal Industry Manual (McGraw-Hill Publications, New York, New
York, 1975).
105
-------
Kulujian, N. J., "By-Product Coke Battery Compliance Evaluation," PEDCo-
Environmental Specialists, Inc. (1975).
Lazenka, C., private communications to Ben Suta, City of Philadelphia,
Department of Public Health (May 9, 1977).
Manning, J., "Coke Oven Dispersion Analysis for BaP and BSO," memorandum,
U.S. Environmental Protection Agency (March 18, 1977).
Manning, J., personal communication (August 1977).
National Academy of Sciences, "Particulate Polycyclic Organic Matter
Report on Biological Effects of Atmospheric Pollutants," Report No.
PB-212 940/1 (1972).
NIOSH, "Correlations Among Log Concentrations of 13 Polynuclear Aromatic
Compounds (PNAs) and the Log Concentrations of the Benzene Soluble
Fraction," memorandum (December 6, 1974).
Radding, S B., T. Mill, C. W. Gould, D. H. Liu, H. L. Johnson, D. C.
Bomberger, and C. V. Fojo, "The Environmental Fate of Selected Poly-
nuclear Aromatic Hydrocarbons," prepared by Stanford Research Insti-
tute for the Environmental Protection Agency, Report No. EPA 560/2-
75-009 (1976).
Shabad, L. M. and Y. L. Cohan, "Contents of Benzo(a)pyrene in Some Crops,"
Arch. Geschwulstforsch. 40_(3) : 237-243 (1972).
Sheridan, E. T., "Supply and Demand for United States Coking Coals and
Metallurgical Coke," U.S. Department of the Interior (1976).
Spangler, C. and N. deNevers, "Benzo(a)pyrene and Trace Metals in Charles-
ton, South Carolina," U.S. Environmental Protection Agency, Report
No. EPA-450/2-75-004 (1975).
Thuillier, R. H., "Estimating Air Quality Background," Stanford Research
Institute, Menlo Park, California (1977).
U.S. Environmental Protection Agency, "Preferred Standards Path Report
for Polycyclic Organic Matter" (1974).
U.S. Environmental Protection Agency, "Scientific and Technical Assess-
ment Report on Particulate Polycyclic Organic Matter (PPOM)," Report
No. EPA-600/6-75-001 (1975).
U.S. Environmental Protection Agency, "Compilation of Air Pollutant Emis-
sion Factors," Report No. AP-42 (1976A).
U.S. Environmental Protection Agency, "The Carcinogen Assessment Group
Analysis of Kepone," Draft Report (1976B).
106
-------
U.S. Public Health Service, "Radiological Health Handbook," U.S. Depart-
ment of Health, Education, and Welfare (1960).
Varga, J., Jr., "Screening Study on By-Product Coke Oven Plants," Draft
Report, Battelle Research Institute (1974).
Youngblood, P. L., "Rough Estimation of Ambient Impact of Organic Emis-
sions from Coke Ovens," Draft Memorandum, U.S. Environmental Protec-
tion Agency (May 27, 1977).
107
------- |