October 1974
PREFERRED
STANDARDS PATH REPORT
FOR POUYCYLIC
ORGANIC MATTER
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air Quality Planning and Standards
Strategies and Air .Standards Division
Durham, North Carolina
-------
PREFERRED
STANDARDS PATH REPORT
FOR POLYCYLIC
ORGANIC MATTER
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF AIR QUALITY PLANNING AND STANDARDS
STRATEGIES AND AIR STANDARDS DIVISION
DURHAM, NORTH CAROLINA
October 1974
-------
CONTENTS
Page
LIST OF FIGURES v
LIST OF TABLES vi
EXECUTIVE SUMMARY vii
1. INTRODUCTION 1
1.1 Background 2
1.2 Scope of Problem 4
2. RATIONALE AND RECOMMENDATIONS FOR EPA STRATEGY FOR BaP . . . . 7 '•'
2.1 Criteria Under the Clean Air Amendments of 1970 7
2.1.1 National Emission Standards for Hazardous
Air Pollutants 9
2.1.2 National Ambient Air Quality Standards 10
2.1.3 Standards of Performance for New Stationary
Sources 11
2.1.4 Emission Standards for Moving Sources 12
a. Motor Vehicle Emission Standards 13
b. Regulation of Fuels 13
c. Aircraft Emission Standards 14
2.1.5 Combination of Options 15
2.1.6 Total Ban 16
2.2 The Preferred Standards Path 17
2.2.1 NESHAP 17
2.2.2 NAAQS 18
2.2.3 ESFMS 19
2.2.4 NSPS 19
2.2.4.1 Alternatives to Control Under Section 111 21
2.2.5 Alternative Control Strategy 21
2.2.6 Advisability of Federal Control for BaP 24
2.3 Recommendations ' 25
3. SOURCES OF BaP 27
3.1 Stationary 27
3.1.1 Heat and Power Generation 28
3.1.2 Refuse Burning 29
3.1.3 Industrial Activity 31
3.2 Mobile 33
iii
-------
Page
3.2.1 Gasoline-powered Vehicles 33
3.2.2 Diesel-fuel-powered Vehicles 34
3.3 Natural 36
4. CONTROL TECHNOLOGY 37
4.1 Stationary 37
4.2 Mobile 41
4.3 Natural 43
5. AMBIENT CONCENTRATIONS 45
5.1 Measurement Technique 45
5.2 Ambient BaP Concentrations 47
5.3 Trends Analysis for BaP 49
5.4 National Urban and Nonurban Trends for BSO 51
6. EFFECTS OF BaP 55
6.1 Epidemiological Studies 55
6.1.1 Classification of Studies 56
6.1.2 Consideration of Other Variables 57
6.1.3 Conclusions from Epidemiological Studies .... 58
6.2 Other Effects 59
7. PROGRAMS TO STUDY BaP (AND OTHER POM) 61
7.1 Current Research Projects 62
7.1.1 Analysis Methodology . . . 62
7.1.2 Effects Investigation 62
7.1.3 Stationary Sources 64
7.1.4 Ambient Analysis 64
7.2 Recommended Research Projects 64
8. REFERENCES 67
APPENDICES 71
Appendix A - Overview of a Preferred Standards Path A-l
Appendix B - Lung Cancer Mortality in Selected SMSAs .... B-l
Appendix C - National Air Surveillance Network Ambient
Air Measurements for Benzo(a)pyrene C-l
Appendix D - Description of By-product Coke Production ... D-l
Appendix E - Costs of Control and Growth Patterns for
Coke Ovens E-l
Appendix F - Emission Estimates for Coke Ovens F-l
iv
-------
LIST OF fIGURES
Figure Page
1 Typical Structural Formula of POM, Benzo(a)pyrene .... 3
2 Typical Structural Formula of an Aza-arene,
Dibe.nzo(c,g)carbazole 4
3 Annual POM Concentrations (10 Compounds) Versus Annual
BaP Concentrations in Birmingham, Ala., 1964 - 1965 ... 6
4 Preferred Standards Path: Guide for Determination
of Regulatory Action 8
5 Trends in Annual BaP Concentrations in Cities with and
without Coke Ovens 48
6 Trends in BSO and in BSO Percentage of TSP at 32 Urban
and 19 Nonurban Stations 52
A-l Preferred Standards Path: Guide for Determination
of Regulatory Action A-4
D-l Koppers-Becker Underjet Low-differential Combination
Coke Oven with Waste-gas Recirculation D-2
-------
LIST OF TABLES
Table Page
I Major Sources of Estimated BaP Emissions in the United
States by Descending Quantity of Contribution (1971-73) . 20
II Estimated Benzo(a)pyrene Emissions from Heat and Power
Generation Sources in the U.S. (1972) 29
III Estimated Benzo(a)pyrene Emissions from Refuse Burning
in the United States (1968) 30
IV Summary of Estimated Industrial Benzo(a)pyrene
Emissions in the United States (1972) 32
V Estimated Vehicular Benzo(a)pyrene Emissions
in the United States (1970) . . 35
VI Estimated Benzo(a)pyrene Emissions from Catalytic
Cracking Sources in the United States (1968) 39
VII Contributions to National Totals of Benzo(a)pyrene
by Source and State (1972) 42
VIII Automotive Benzo(a)pyrene Emission Factors (1972) .... A3
IX Three Year Summary of TSP, BSD, and BaP for 121 NASN
Sites in the United States - 1968, 1969, and 1970 .... 47
X Applicable Current POM Research Projects 63
B-l White Male Deaths and Death Rates per 100,000 Population
per Year by SMSA for Malignant Neoplasm of Trachea,
Bronchus, and Lung (1959-1961) B-2
C-l National Air Surveillance Network Ambient Air
Measurements for Benzo(a)pyrene: 121 Selected Sites
(1968-1970) C-2
C-2 Listing of 40 NASN Sites Selected for 1971-72
BaP Analysis C-7
C-3 Annual BaP Averages for Selected Cities (1966-1972) . . . C-7
E-l Capital Costs (in $1,000) E-3
E-2 Annual Costs (in $1,000) E-4
E-3 Summary of Projected New Oven Construction (U. S.). . . . E-6
vi
-------
EXECUTIVE SUMMARY
Polycyclic Organic Matter (POM) is an aromatic hydrocarbon group
that includes certain compounds which are proven carcinogens at ele-
vated levels in laboratory animals. Some of these compounds have been
linked with the occurrence of cancer in humans, but none have been
directly related to cancer from exposures to ambient air. Nevertheless,
it seemed prudent to consider POM a potential air pollution problem and
to evaluate the need for additional formal regulatory action. The
objective of this report is to assess the POM problem, determine the
requirement for additional regulatory action under the Clean Air Act,
and identify the most appropriate regulatory tool if such action
is necessary.
POM includes thousands of compounds which vary widely in physical
and chemical characteristics and in their capacity to act as carcinogens.
Data on total POM are extremely limited. Most source emission estimates
are for a single POM compound, benzo(a)pyrene (BaP), and not for total
POM. BaP is a proven carcinogen commonly used as a surrogate for total
POM. No analytical techniques exist for ambient air measurements of
total POM. The majority of these measurements are for BaP only.
Limited dose-response data that have been assembled also are for BaP
and not total POM.
Sources of BaP normally are associated with the incomplete combustion
of organic compounds, especially coal. The nationwide emissions of BaP
.are estimated at about 900 tons per year based on 1971-1973 data.
vii
-------
Major source categories and their estimated contribution to this total
are listed below.
BaP Emissions % of
Source Type (tons/year) Total
(1971-1973)
1. Coal Refuse fires 310 34.7
2. Residential furnaces, coal (hand-stoked) 300 33.6
3. Coke production 170* 19.0
4. Vehicle disposal (open burning) 25 2.8
5. Wood burning (fireplaces, etc.) 25 2.8
6. Mobile sources, gasoline 11 1.2
7. Forest and Agricultural burning 11 1.2
8. Tire degradation 11 1.2
9. Open burning (domestic—municipal) 10 1.1
10 Intermediate coal furnaces 7 0.8
11. Petroleum refineries 7 0.8
12. Enclosed incineration (apartment—municipal) 3 0.3
13. Other 4_ 0.5
TOTALS 894 100.0
*A range of emissions has been given for this value. Appendix F
contains the details.
viii
-------
As illustrated by this list, stationary sources account for about
97% of the nationwide emissions of BaP. The inefficient combustion of
coal is the largest single source of BaP. Thus the three largest
emission sources are coal refuse fires, residential coal furnaces, and
coke ovens. Together, they comprise more than 85% of national emissions
of BaP.
Annual concentrations of BaP range from 0.07 to abour 17 nanograms
3
per cubic meter of air sampled (ng/m ). Urban levels are higher than non-
urban levels by as much as 10 times under some conditions. Annual
ambient air concentrations of BaP have declined significantly over the
past few years. In order to assess these declines, BaP trends data
were compiled for selected cities from 1966 through 1972. Cities
containing large coke oven facilities were singled out as a separate
category since coke ovens are a major source of POM and were considered
for special regulatory action. In addition, trends in the benzene
soluble organic fraction (BSO) of total suspended particulates were
analyzed for urban and non-urban sites since data for this fraction
are more complete than data for BaP. Both BaP and POM are found in
the benzene soluble portion of total particulates.
Trends data for BaP (shown in the following figure) indicate a
significant decline in ambient concentrations. From 1966 to 1972,
concentrations have decreased about 55% in coke oven cities and 77%
in non-coke oven cities. For all cities sampled, the urban composite
average BSO concentrations (shown in the second figure) have decreased
ix
-------
5
I 3
o
o
O-
ra
CD
1966
1967
1970
1971
1968 1969
TIME, year
Trends in annual BaP concentrations in cities with and without coke ovens.
1972
12
10
cc
o
o
o
o
KQ
CO
Till
32 URBAN STATIONS
(ONLY 18 NONURBAN STATIONS
INCLUDED IN THIS COMPOSITE
AVERAGE)
19 NONURBAN STATIONS
0
1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970
TIME, year
Trends in BSD and in BSO percentage of TSP at 32 urban and 19 nonurban stations.
-------
3 3
from 10.6 ug/m in 1960 to 4.8 ug/m in 1970, or a 55% decrease.
These decreasing trends are significant and can be attributed to the
success of existing control programs as well as the move awav from
the use of coal in small, inefficient furnaces.
Coke ovens are a special problem. Based on available data,
they are suspected of being a major source of BaP and are difficult
to control. Demonstration projects sponsored by the Environmental
Protection Agency and industry groups indicate that particulate
emissions from coke ovens can be controlled by a combination of
process alteration and equipment design modifications. Preliminary
results show that application of good particulate emissions control
will also reduce BaP emissions by 85-90%. Federal emission standards
for particulates being developed for new coke ovens, and enforcement
actions underway against existing ovens, ensure adequate regulation
of this remaining significant source of BaP.
Existing Regulatory Control Program
Coal refuse fires, residential coal-fired furnaces, and coke ovens
(the three major source categories of POM emissions) are being con-
trolled, or are scheduled for control, through the following actions:
1. Coal refuse fires are regulated in the four States which
include the majority of coal refuse fires. The nine other States
with coal refuse fires can control emissions from this source by open
burning regulations or other regulations. In addition, the Bureau of
Mines, Department of Interior, has proposed Federal regulatory action
for this source. They presently require control of spontaneously
started fires under existing Federal regulations.
xi
-------
2. Emissions from coal-fired residential sources decreased 28%
between 1968 and 1972 and this trend is expected to continue for the
following reasons:
a) Socio-economic conditions have established a trend
away from coal-burning in the home because of its dirtiness
and inconvenience.
b) All states except one are imposing particulate emission
limits on coal-fired residential furnaces. In many
cases, these limits, by requiring particulate control
systems, discourage the continued use of coal for res-
idential heating.
c) Residential coal-fired furnaces are banned directly in
some urban areas such as Chicago, St. Louis, and Milwaukee.
Others have indirect bans through stringent sulfur regu-
lations which preclude the use of most, if not all,
sources of coal.
3. New coke ovens will be controlled bv national particulate
emission standards being developed for this source. Control equipment
has already been installed on several existing coke ovens as a result
of local regulations and Federal enforcement actions. EPA has fourteen
enforcement actions against coke ovens for particulate controls. Pre-
liminary data from EPA demonstration projects indicate that about 90%
control can be achieved for particulate emissions during charging
operations, which historically has been responsible for most emissions
from this source. These same data show an equivalent degree of BaP
emission control can be expected.
xii
-------
Most of the remaining sources for BaP are small, numerous and widely
dispersed. Several of these sources are not possible to control effectively
(e.g., home fireplaces, forest and agricultural burning). Others
(petroleum refineries, vehicle disposal, and open burning) are being
regulated through New Source Performance Standards and State Implementation
Plans. Although some concern has been expressed regarding increased
use of coal for electrical power generation under our current national
energy policy, such increased use is not anticipated to cause increased
emissions of POM because power plants utilize very efficient combustion
systems and emit essentially no POM.
Conclusions and Recommendations
The health effects of ambient air concentrations of POM or BaP
are not well documented and information is not available for selecting
harmful air concentrations. Although the presence or magnitude of the
threat from POM cannot be determined at this time, selected polvnuclear
organic materials have been shown to be carcinogenic in some situations.
Therefore, it seems prudent to minimize emissions of POM and to use
all practical means to reduce levels of POM in the air.
. i
Ambient concentrations of BaP and benzene soluble organics have
declined significantly over the period for which trends data are
available. From 1966 to 1972, BaP concentrations have decreased 55%
in coke oven cities and 77% in non-coke oven cities. This decline in
ambient levels is attributed to current particulate control programs
and the phase-out of coal-fired home furnaces. Although these actions
have not been oriented specifically to reducing BaP emissions, they do
have a direct effect on those sources responsible for most BaP emissions
xiii
-------
and are an effective national control strategy. An analysis of additional
regulatory steps does not suggest other effective measures.
Control regulations specifically for BaP or POM are not warranted
nor practical at this time. Ambient monitoring for BaP should be under-
taken in selected cities to ensure the continued adequacy of the control
program. Also, special emphasis should be given to the prompt enforce-
ment of existing regulations for incinerators, open burning, coal
combustion, burning refuse piles, and coking operations.
xiv
-------
PREFERRED
STANDARDS PATH REPORT
FOR POLYCYLIC
ORGANIC MATTER
1. INTRODUCTION
Polycyclic organic matter (POM) and benzo(a)pyrene (BaP) were two
of several pollutant groups for which control strategies were to be
considered as listed under National Objectives in the Administrator's
"Guidelines for Development of Fiscal Years 1973-1978 Program Plans."
Since BaP is a specific POM compound and is generally used as a
surrogate for POM the two tasks were combined.
The procedure used in preparing this report and arriving at
recommendations included:
(1) studying control options available under the 1970 Clean Air
Act, an internal EPA procedure known as a preferred standards
path analysis.
(2) reviewing and evaluating comprehensive studies available on
POM, including:
1
-------
3*
(a) "Particulate Polycyclic Organic matter" by the National
Academy o,f Sciences, 1972.
(b) "Draft NERC/RTP Position Paper on Particulate Polycyclic
Organic Material (PPOM)" by PPOM Task Force Panel, U.S.
EPA, NERC/RTP, 1972.
(c) "Preliminary Air Pollution Survey of Organic
Carcinogens" by Litton Systems, Inc., for HEW, 1969.
8
(d) "Sources of Polynuclear Hydrocarbons in the Atmosphere"
by R.P. Hagebrauck, e_t ^1., AP-33, 1967.
(3) evaluating available ambient air data.
After thorough evaluation of this information, the following
abridgement is presented.
1.1 Background
Polycyclic organic matter (POM) is organic matter with a
multiple ring (two or more) structure, as shown in Figure 1. A
typical structure of five rings is shown in Figure 1 with
the dashed circle representing the carbon-carbon (C-C) bond and the
numbers around the outside representing points where, in this case,
hydrogen atoms are attached. Other atoms (or groups of atoms) may be
substituted for carbon at the numbered points. When such a
substitution occurs, polycyclic azaheterocyclic compounds (aza-arenes)
are produced. Other names which are given to POM include (1)
* Numbers shown as superscripts refer to references at the end of
the Report.
-------
5
Figure 1. Typical structural formula of POM,
benzo(a)pyrene .
polynuclear aromatic hydrocarbons (PNAH, PAH); (2) polynuclear
aromatics (PNA); (3) polycyclic aromatic hydrocarbons (PAH); or
simply, (4) aromatic ring compounds. Generally, when any of these
names, or variations thereof, are used the broad class of POM is being
referenced.
The primary reason for interest in POM is that some of these
compounds have been proved to be carcinogens in experimental animals.
They may also act as cocarcinogens, or be active in the presence of
other cocarcinogens. (A carcinogen is a substance or agent which
produces or incites cancer. Agents which induce altered physiological
states that may increase the risk to carcinogenesis are
.cocarcinogens). Furthermore, epidemiological evidence connects lung
cancer etiology with occupational exposure to certain particulate POM
(PPOM) compounds in specific industries. A few POM compounds which
have been identified as definite carcinogens are :
-------
7,12-Dimethylbenz(a)anthracene
Dibenz(a,h)anthracene
3-Methylcholanthrene
Benzo(c)phenanthrene
Benzo(a)pyrene
Dibenzo(a,h)pyrene
Dibenzo(c,g)carbazole
The last compound in the list is an example of a carcinogenic aza-
arene and is illustrated in Figure 2.
Figure 2. Typical structural formula of an
aza-arene, dfbenzo(c,g)carbazole.
1.2 Scope of Problem
Under the Clean Air Amendments of 1970 the EPA was provided with
alternative approaches to control strategies for candidate pollutants.
The process of looking at these alternatives is known as a preferred
standards path analysis (PSP), which is explained in greater detail in
Appendix A. The PSP analysis in this report uses BaP as an Indicator
for PPOM. Reasons for limiting this analysis to BaP as an index for
PPOM are:
-------
a. Existing data support using BaP as a rough indicator for PPOM
(Figure 3 and Section 5). If effective control is achieved for
BaP other POM compounds should also be controlled to a large
degree because their characteristics are similar to BaP.
b. .Lack of available data for POM:_ _
1. Emissions estimates are for BaP and not for POM; some
scientists have limited data on total POM.
2. Majority of ambient air measurements are for BaP only;
selected research studies have produced limited ambient
data for POM.
3. Limited dose-response information does exist for BaP but
none is available for total POM.
c. BaP has received primary emphasis in measurement techniques
development for specific POM compounds; none are available
for total POM.
Thus, this report presents the preferred standards path for BaP as
a surrogate preferred standards path for POM. The NAS report, the
ORM/NERC/RTP position paper on PPOM, extensive personal contacts, and
other literature have been used for support^ d?~the^ scenario contained
in this report.
-------
e
oi
O
I-
ce
o
o
o
a.
180
160
140
120
100
80
60
40
20
10
20
30
40
BaP CONCENTRATION, ng/m3
Figure 3. Annual POM concentrations (10 compounds) versus annual BaP concentrations
in Birmingham, Ala., 1964-1965.
-------
2. RATIONALE AND RECOMMENDATIONS
FOR EPA STRATEGY FOR BaP
2.1 Criteria under the Clean Air Amendments of 1970
Criteria available to the Environmental Protection Agency (EPA) for
recommending a preferred standards path for any pollutant from either a
stationary or moving source, as described in the Clean Air Amendments of
1970 (Act), include the following factors: (1) Presence and magnitude
of health and/or welfare effects of a pollutant; (2) Nature and distri-
bution of pollutant sources; and (3) Supporting data (implied). Those
sections of the Act in which these factors are included, along with the
six basic standard setting options, are described and then discussed
below. The characteristics of a candidate pollutant are compared with
criteria for each option given (Figure A). This chart should not be con-
sidered a strict decision-making tool but merely a guide that provides one
of many inputs into the selection process. Analysis must begin with an
estimation of health effects since this is the common thread among all
six options. (See block 1 of Fig. 4.) Once the determination is made
that a candidate pollutant may contribute to adverse health and/or
welfare, one or more of the options for control should be chosen. The
approach to a control strategy requires different procedures depending
upon the chosen option.
-------
oo
1
Adveise Health and
Welfare Effects: or
Possibility of En-
dangerment of Health
and Welfare
3 Y"
Magnitude of Health and Welfare
Effects: Contribute to Increase
in Mortality, or Serious Irrever-
sible or Incapacitating Reversi-
ble Illness
No
7
Nature and C
of Sources:
ness - Wheth
or Diverse, '
Moving, or B
No
istribution
Ubiquitous-
er Numerous
tationary or
oth
No
4 5
< UnTCi Adequate Data to Nature and Distri- No
See NOTE 1 Show Hazardous El- IiS ». butjon ., sources. "° »
Yes (ects at Current Am' Ubiquitous?
bient Concentrations
No Optional,
8
Is the Pollutant
Moving Sources? lft 12
u Can Tontrnl Rp Adequate Data
_ „ N° SiSsr INn . £»** YK .
Yes * No Ac ion SS" "* ." * .Y" = =
*" ° Under ESFMS "'• Q3n Control Be and Welfare
Yes Regulation^ T7
9 1 ^ 1 [No
Is the Pollutant r *
* Prominent From No L.
Stationary Sources? 1& f ^ St-c NntP 7 F
Yes to Establish and Support yes
for Health and Welfare
Effects?
No N°, Consider
ACTION
2
No Action
Under Clean
Air Act
6
Section 1 12, NESHAP"
Most Likely Appropriate1
13
Section 211, ROF,"
Most Likely Appropri-
ate*
15
Sections 202 and 231,
ESFMS,** Most Likely
Appropriate*
7
Section 109, NAAQS,**
Most Likely Appropri-
ate*
18
Section 111, NSPS,~
Most Likely Appro-
priate*
NOTE 1: A yes answer here simply means the pollutant is a candidate for regulatory action; however, it does not denote mandatory action.
NOTE 2: The dashed line indicates that significant pollution from stationary sources may remain even after utilizing ESFMS. Land use and transportation
controls may need to be implemented in addition to Section 202.
*MOST LIKELY APPROPRIATE simply means that the indicated option may be the most logical without considering external factors. However, other ramifications
may preclude its use. Also, more than one option (e.g., Sections 109 and 202; Sections 109 and 111; etc.) may be used in combination for any given pollutant in
order to achieve the full intent of the Act. (Also see NOTES.) Section lll(d) is applicable to non-criteria pollutants.
"Acronyms «iplamed in text.
Figure 4. Preferred standards path: Guide for determination of regulatory action.
-------
2.1.1 National Emission Standards for Hazardous Air Pollutants
(NESHAP - Section 112)
Under the Act the Administrator is allowed to use his judgment to
determine whether a pollutant is hazardous, i.e. "may cause, or contri-
bute to, an increase in mortality or an increase in serious irreversible,
or incapacitating reversible, illness" [Section 112(a)(l)]. Within 180
days after publishing a list of suspected hazardous pollutants the
Administrator must publish proposed regulations setting emission standards
together with a notice of public hearings. "Not later than 180 days
after such publication, the Administrator shall prescribe an emission
standard for such pollutant, unless he finds, on the basis of informa-
tion presented at such hearings, that such pollutant clearlv is not a
hazardous air pollutant, [then] the Administrator shall establish any
such standard at the level which in his judgment provides an ample
margin of safety to protect the public health from such hazardous air
pollutant" [Section 112(b)(1)(B)].
A major consideration in making a preferred standards path choice
is to assess the largest magnitude of adverse health effects, i.e.,
possible mortality or serious illness. This requires evaluating the
candidate pollutant for the hazardous option (Section 112) initially.
Here, the effect of atmospheric emissions on health must be analyzed in
relation to (a) current and expected ambient levels (concentrations) and
(b) what an ample margin of safety is. In practice, this relationship
is affected by several factors, including terrain, number of sources,
extent of buildup or persistence in the environment, etc. Setting a
national emission standard despite these variables may logically require
-------
a level of emissions to be either zero or based upon the worst signifi-
cant situation observed or anticipated. If in applying these criteria,
specific measured or expected ambient concentrations of the candidate
pollutant are shown to be hazardous, the hazardous pollutant option mav
be the choice. If so, a national emission standard which provides an
ample margin of safety must be proposed and promulgated. The standard
will apply to stationary sources, new and existing. The Act does not
specifically reciuire that the Administrator have information on control
technology to initiate action under this option. However, from a practical
point of view, since the Administrator must oversee implementation and
enforcement of emission standards knowledge of control technologv is
required.
2.1.2 National Ambient Air Quality Standards (NAAQS -
Sections 108-110)
The Act requires the promulgation of primarv "ambient air quality
standards the attainment and maintenance of which in the judgment of the
administrator, . . . allowing an adequate margin of safety, are req-
uisite to protect the public health" [Section 109(b)(l)]. Similarly,
secondary NAAQS are required to "protect the public welfare from any
known or anticipated adverse effects" [Section 109(b)(2)]. Ambient air
quality standards are based upon criteria which delineate "all identi-
fiable effects on public health or welfare" from a pollutant whose
"presence ... in the ambient air results from numerous or diverse
mobile or stationary sources" [Section 108(a)]. The Act further requires
each State to "adopt and submit to the Administrator ... a plan which
provides for implementation, maintenance, and enforcement of such . . .
10
-------
standard in each air quality control region . . . within such State" [Sec-
tion 110(a)(l)].
In controlling pollutants under the ambient option the effect of
existing ambient concentrations on health and welfare must first be
analyzed. Such data must be published in a criteria document simultane-
ously with a proposed national standard for a specific ambient concentra-
tion which can be supported. Then the States are left to establish the
relationship between ambient concentrations and emission levels from
sources. This relationship is affected by such factors as terrain,
number of sources, and effect of buildup or persistence of the candidate
pollutant in the environment. States are responsible for prescribing
and enforcing emission standards, procedures for control of number or
location of sources, etc. However, the Administrator must issue control
techniques information simultaneously with criteria documents [Section
108(b)(l)]. Also from the practical point of view, development of im-
t
plementation plans by States and the Administrator's review and approval
of such plans require some knowledge of control technology.
2.1.3 Standards of Performance for New Stationary Sources
(NSPS - Section 111)
The Act specifies that the Administrator include a category of sources
on a proposed list for standards of performance "if he determines it may
contribute significantlv to air pollution which causes or contributes
to the endangerment of public health and welfare" [Section lll(b)(1)(A)].
Further, within 120 days "the Administrator shall propose regulations,
establishing Federal standards of performance for new sources within
such category" and ". . . promulgate within 90 days" [Section lll(b)(1)(B)].
11
-------
The Act further requires the Administrator to "prescribe regulations . . .
under which each State shall submit ... a plan which (A) establishes
emission standards for any existing source for any air pollutant (i) for
which air quality criteria have not been issued or which is not included
on a list published under Section 108(a) or 112(b)(l)(A) but (ii) to
which a standard of performance under subsection (b) would apply if such
existing source were a new source, and (B) provides for the implementation
and enforcement of such emission standards" [Section lll(d)(l)].
When a limited number of source categories or limited number of
predominant categories exist, the section on standards of performance
for new sources may be the most practical approach to controlling a
candidate pollutant. If Section 111 is used, the effect of atmospheric
emissions of the candidate pollutant on health and welfare must be
analyzed first. However, the Act does not require calculating a re-
lationship between ambient concentrations and emissions since the
standard will reflect 'the best demonstrated system of emission reduction
for the affected new source (taking cost into account). Thus, the level
of control is not related directly to health and welfare effects of the
candidate pollutant. Analysis for health and welfare effects do or
could exist. For unmodified existing sources, States must enact emission
standards using available practical approaches, subject to EPA review
and approval [Section lll(d)].
2.1.4 Emission Standards for Moving Sources (ESFMS - Title II)
Title II of the Act, Emission Standards for Moving Sources (ESFMS),
provides for control of air pollutants from moving sources. This title
presents criteria for motor vehicles, fuel regulations, and aircraft.
12
-------
a. Motor Vehicle Emission Standards (MVES - Section 202)
Under this section of the Act "the Administrator shall by
regulation prescribe . . . standards applicable to the emission
of any air pollutant from any class or classes of new motor
vehicles or new motor vehicle engines, which in his judgment
causes or contributes to, or is likely to cause or contribute
to, air pollution which endangers the public health or welfare.
Such standards shall be applicable to such vehicles and engines
for their useful life" [Section 202(a)(l)]. The Act further
requires the Administrator to allow time for "development and
application of the requisite technology, giving appropriate
consideration to the cost of compliance within such period"
[Section 202(a)(2)]. Subsection (b) of Section 202 presents
exceptions which apply to standards and time requirements for
hydrocarbons (HC), carbon monoxide (CO), and nitrogen oxides
(NO ).
A.
b. Regulation of Fuels (ROF - Section 211)
"The Administrator may, ... by regulation, control or
prohibit the manufacture, ... or sale of any fuel or fuel
additive for use in a motor vehicle or motor vehicle engine
(A) if any emission products of such fuel or fuel additive
will endanger the public health or welfare" [Section 211(c)
(1)(A)]. The Act further provides that control under the
subsection just quoted is prohibited "except after considera-
tion of all relevant medical and scientific evidence available
to him, including consideration of other technologically or
13
-------
economically feasible means of achieving emission standards
under section 202" [Section 211(c)(2)(A)]. Furthermore, in
connection with registration of fuels and additives the Act
provides means whereby "the Administrator may also require the
manufacturer of any fuel or fuel additive - (A) to conduct
tests to determine potential health effects of such fuel or
additive (including, but not limited to, carcinogenic, ...
effects)," [Section 211(b)(2)(A)].
c. Aircraft Emission Standards (AES - Section 231)
The Administrator shall issue, based on studies
conducted by him, "proposed emission standards applicable
to emissions of any air pollutant from anv class or
classes of aircraft or aircraft engines which in his
judgment cause or contribute to or are likely to cause or
contribute to air pollution which endangers public health
or welfare" [Section 231(a)(2)]. Provision is made for
final regulations to be issued within 90 days after the
proposal following public hearings within the most af-
fected air quality control regions (AQCR) [Section 231
(a)(3)]. However, the regulation shall not take effect
until necessary time has elapsed to develop and apply
requisite technology and consultation with the Secretary
of Transportation [Section 231(b) and (c)].
In considering the nature and distribution of sources of the
candidate pollutant, one of the first questions is whether emissions are
from moving sources. If so, then additional questions must be asked
14
-------
(block 8 of Figure 4) because at least two options become available.
However, regulation of fuels (Section 211) can be employed only after
considering all possibilities of control by emission standards under
Section 202, as well as relevant medical and scientific evidence. If
emission standards are not, or can not be, effective without fuel re-
gulation, then this option may be used also.
Emissions from aircraft must also be considered in connection with
moving sources. If such emissions have a significant impact on health
and welfare effects the Administrator, in consultation with the Secre-
tary of Transportation, is authorized to set standards. Enforcement of
these standards is the responsibility of the Secretary of Transportation,
after consultation with the Administrator [Section 232(a)]. Standards
can be set and enforced only after allowing what is, in the Administra-
tor's [Section 232(a)]. Standards can be set and enforced only after
allowing what is, in the Administrator's judgment, sufficient time for
development and application of requisite technology. Cost of compliance
must be considered also [Section 231(b)].
2.1.5 Combination of Options
Pollutants may be regulated under more than one section of the Act
simultaneously (Figure 4). However, this approach cannot be used in at
least two specific situations. These situations occur when pollutants
are potential candidates for control under Section 112 (which cannot be
applied to a pollutant for which an ambient air quality standard exists)
or Section 111, subsection (d) (which cannot be applied to any pollutant
for which ambient or hazardous standards exist). However, Section 111
may supplement Sections 109 and 112 to control the same pollutant if
15
-------
specified criteria (as discussed in Subsections 2.1.1-2.1.4 above and
Appendix A) are fulfilled. Once a pollutant is regulated under Section
111, Section lll(d) becomes effective for existing sources of the pol-
lutant, unless ambient or hazardous pollutant standards apply.
Standards for moving sources, Sections 202, 211, and 231, are more
likely to be used in conjunction with Section 109, NAAQS, than other
possible combinations (Figure 4). For example, Section 202 of the Act
specifies emission limitations for carbon monoxide, hydrocarbons, and
nitrogen oxides from automobiles. Concurrently, ambient air quality
standards have been set under authority of Section 109 for these same
pollutants. However, standards for moving sources should not be ex-
pected to be used solely with Section 109. They could also be used with
Sections 111 and 112 to control the same pollutant if specified criteria
of the Act are satisfied.
2.1.6 Total Ban
If health data warrants, a total ban on emissions could be achieved
directly under the hazardous standard option, indirectly under the
ambient standard option through stringent levels, and directly under the
new and/or moving source options if an adequately demonstrated system
exists to achieve zero emissions (taking cost into account).
Using the preceding mechanisms, alternatives and criteria needed in
selecting the preferred standards path (selecting a particular option)
are illustrated in Figure 4. Decisions listed in Figure 4 are qualified
as "most likely appropriate" because the Act is subject to interpre-
tation (Appendix A). Since the Act allows flexibility, the choice of a
particular option may create controversy because another alternative
16
-------
could be just as valid. However, the first requirement for taking any
action under the Act is the establishment of adverse health and/or wel-
fare effects. Once this is confirmed, alternatives can be evaluated,
using Figure 4 as a guide.
2.2 The Preferred Standards Path
Now, available information for BaP will be examined in relation to
the first block in Figure 4, i.e., establishment of the possibility of
adverse health or welfare effects. Specifically, the following reasons
are sufficient justification to consider Federal control of BaP.
a) BaP is a proven carcinogen at elevated levels in animals under
experimental conditions.
b) Epidemiological evidence connects lung cancer etiology with
occupational exposure to BaP, and other coal tar volatiles, in specific
industries. Data relating to each of the six options will now be pre-
sented in the order shown in Figure 4.
2.2.1 NESHAP
Although an apparent urban factor exists causing a 2.0 to 2.5 times
increase in lung cancer incidence, no evidence exists to link this in-
crease solely to BaP or other POM in the ambient air. Many areas with
high lung cancer incidence (Appendix B) do not coincide with high ambient
concentrations of BaP as measured by the NASN prior to 1970 (Appendix C).
Unfortunately, many cities with high lung cancer incidence have not been
sampled for BaP. Nevertheless, experimental animal studies and limited
epidemiological studies do suggest that BaP may cause or contribute to
lung cancer incidence. Naturally, lung cancer results in an increase in
mortality and/or an increase in serious irreversible, or incapacitating
reversible, illness.
17
-------
However, data do not support that these hazardous effects (e.g. lung
cancer incidence) are due to inhalation of (or other exposure to) ex-
isting or anticipated atmospheric concentrations of BaP. Neither do data
define levels at which effects occur as no threshold of effects can be
conjectured. In short, with current data and extrapolation techniques
BaP has not been shown with certainty to be hazardous at ambient levels.
Of course, the Act does not require that ambient levels be the deciding
factor, but from a practical point of view, effects should be related
to levels that people are likely to receive in routine activities. If
these are occupational levels then EPA would not be the agency to establish
standards. Also, the nature and distribution of sources (block 5) should
have a bearing on the decision reached. If sources other than stationary
contribute a significant amount of BaP to the total then Section 112
of the Act would not be adequate remedy. Thus, the conclusion is that the
use of Section 112, NESHAP, for BaP control cannot be justified at this
time.
2.2.2 NAAQS
For BaP the NAAQS option (block 7) must be examined because both
source classifications, moving (block 8) and stationary (block 9),
contribute to ambient concentrations. The moving source classification
also satisfies the criteria of numerous or diverse. However, available
data do not support a specific concentration for an ambient standard.
Establishment of an ambient air quality standard requires the identifi-
cation of a defensible air quality number (including an adequate margin
of safety) where adverse health effects are observed. Arbitrarily
establishing a number without sound medical support for it would be
18
-------
inconsistent with the Clean Air Act. Thus, the conclusion is that
Section 109, NAAQS, is not an appropriate regulatory tool for control-
ling BaP because evidence to support an ambient air quality standard for
BaP is not available.
2.2.3 ESFMS
If the decision block (no. 8) for ESFMS in Figure 4 is chosen
because moving sources are a significant source of the candidate pol-
lutant, two options become available. The first option, emission
standards, does not appear feasible because emissions from mobile sources
are insignificant as shown in Table I. This estimate does not include
emissions from aircraft or diesel-burning moving sources. No estimate
has been attempted for aircraft but emissions are generally believed to
be slight. Emissions from diesel engines have been estimated to be
minimal.
If the emission standards option were not selected, it would cause
a shift to the regulation of fuels option shown in the diagram (block
11). Based on negligible emissions of BaP from gasoline burning
engines the ROF option can not be justified. Thus, based on available
data, emission standards and fuel regulation do not appear feasible.
2.2.4 NSPS
While it is concluded that data do not support action by any of the
previous options, the question remains whether Federal action should be
considered under Section 111 of the Act. Data presented thus far verifies
that Section 111 would be the only feasible regulatory option at this
time if additional regulatory action is needed.
19
-------
Section 111 of the Act authorizes the Administrator to establish
standards of performance which reflect best systems of emission reduction
for categories of new sources. The Act further requires States to
establish emission standards for existing sources of the same categories
for pollutants not listed under Sections 108 or 112. Section 111 can be
used if the source category "may contribute significantly to air pol-
lution which causes or contributes to the endangerment of the public
health or welfare." To be defensible, performance standards under
Section 111 must be applicable to significant categories of BaP. Sources
in these categories are identified in Table I based on available emission
estimates (Section 3.0).
Quantitatively, emissions of BaP to the ambient air are minimal
(Table I). However, since BaP is a proven carcinogen (and/or cocar-
Table I. MAJOR SOURCES OF ESTIMATED BaP EMISSIONS IN THE
UNITED STATES BY DESCENDING QUANTITY OF CONTRIBUTION (1971-73)
Source type
1 . Coal refuse fires
2. Residential furnaces, coal
(hand-stoked)
3. Coke production
4. Vehicle disposal (open burning)
5. Wood burning (fireplaces, etc.)
6. Mobile sources, gasoline
7. Forest and agricultural burning
8. Tire degradation
9. Open burning
(domestic - municipal )
10. Intermediate coal furnaces
11. Petroleum refineries
12. Enclosed incineration
(apartment - municipal)
13. Other
TOTALS
BaP emissions
(tons/year)
310
300
170*
25
25
11
11
11
10
7
7
3
4
894
% of total
34.7
33.6
19.0
2.8
2.8
1.2
1.2
1.2
1.1
0.8
0.8
0.3
0.5
100.0
*A range of emissions has been given for this value. Appendix F contains
the details.
20
-------
cinogen) under given conditions, it may be considered to contribute
significantly to air pollution. This contribution of BaP is from about
three major categories of sources that are amenable to control, using
either Section 111 or other means (e.g. local ordinance). Emissions of
BaP should be reduced because it is a carcinogen.
2.2.4.1 Alternatives to Control Under Section 111
Two major alternatives under this option are:
a) Control all major sources of BaP by §111, assuming adequately
demonstrated control techniques, or
b) control only selected sources of BaP.
Either of these alternatives could be used but would result in a com-
pletely new national strategy for BaP only, thus invoking §lll(d) since
BaP is not regulated directly under §§109 or 112. Strong evidence indi-
cates that such a specific strategy is not warranted due to other con-
trol actions in progress, as discussed in the next section.
2.2.5 Alternative Control Strategy
In addition to options already discussed and dismissed, another
alternative exists: no direct Federal action, based either on the lack
of demonstrated control technology or that existing Federal, State, or
local controls are adequate and effective. Control technology does
exist for, and has been applied to, many of the sources listed in Table
I; for others it is non-existent (e.g., forestry burning) or very difficult
to apply (e.g., burning coal refuse.piles). Federal, State, and local
agencies are also making progress in controlling BaP from all sources.
Analysis of ambient BaP concentrations shows a definite declining
trend in all cities (except one) for which data are available. Ambient
21
-------
concentrations of BaP analyzed in twenty-three cities with coke ovens
show a decrease of 55% over a six-year period (1966-1972). Furthermore,
i
trends analysis for the benzene soluble organics (BSO) fraction of particulate
samples (that portion which includes several POM compounds) shows a 55%
decrease for a ten-year period (1960-1970). (See Section 5.0.) Such
definite declines as these indicate that the following positive actions
by control agencies are having concurrent effects on ambient concen-
trations as measured at receptor sites.
1. Source No. 1, coal refuse fires, is regulated in four
States which include the majority of coal refuse fires.
Other States can control emissions from this source by
open burning, or other, regulations. In addition, the
Bureau of Mines, Department of Interior, has proposed
Federal regulatory action for these sources and are
requiring spontaneously started fires to be controlled
under existing Federal regulations.
2. Emissions from coal-fired residential sources decreased
3 30
28% between 1968 and 1972 ' and this trend is expected
to continue for the following reasons:
a) Socio-economic conditions have established a trend
away from coal-burning in the home. Urban renewal
is underway in practically every U.S. city. All of
these projects avoid the use of coal for heating
because it is dirty, relatively inconvenient to use,
and in many places cannot meet air quality emission
standards. The manufacturer of units which burn
22
-------
solid fuel for home heating has declined to the
point that an industry directory for 1973 did not
list coal-fired units as being manufactured.
b) All States except Arkansas have promulgated general
particulate emission regulations for small fuel
burning sources. These regulations, in effect, re-
quire particulate collectors which are not cost-
effective for residential sources. Five States
limit maximum emission from these sources to 0.8 Ibs
of particulates per million BTU; this rate is
slightly less than estimated emission rates for
coal-fired home furnaces. Of the remaining forty-
four States, twenty-two limit emissions to 0.6 Ibs
per million BTU, and the rest require less than
0.6 lb/10 BTU. For example, Oregon limits particu-
late emissions to 0.33 lb/10 BTU and Connecticut
has a limit of 0.2 lb/106 BTU while Alaska limits
emissions to 0.1 lb/10 BTU. These regulations have
been a contributing factor in the phase-out of coal-
fired residential furnaces.
c) Residential coal-fired furnaces are banned directly
in some urban areas, such as Chicago, St. Louis and
Milwaukee. Others have such stringent sulfur
content regulations that coal is banned indirectly.
For example, Detroit, Philadelphia, and New York,
allow a maximum of 0.3% sulfur content.
23
-------
3. Emissions from all other sources listed in Table I are
in the process of being reduced, or are scheduled to
be in the near future. The remainder of the sources
contribute less than one-third of the total quantity
of BaP nationwide and are dispersed fairly widely
across the U.S. Moreover, the majority of these re-
maining sources are amenable to control and will likely
be controlled by various means in meeting primary
ambient air quality standards for particulates or other
pollutants. For example, by-product coke ovens
(Appendix D) are being controlled through Federal en-
forcement actions as well as State and local actions.
At least fourteen enforcement actions are in progress
and more are expected. In addition, EPA is gathering
data through its demonstration projects in preparation
for proposing performance standards for new sources
(Section 111). These NSPS's for particulate and hydro-
carbon controls are expected within the next twelve
months.
These actions that are in progress and downward trends of ambient
concentrations indicate that major sources of BaP are being con-
trolled.
2.2.6 Advisability of Federal Control for BaP
The advisability of Federal control should be based on an
evaluation of the magnitude of effects of ambient BaP as well as
effectiveness of existing and planned actions. Available data have
24
-------
already been presented which identify those factors of importance in
deciding whe&her Federal action should be specific for BaP. These data
are restated here for emphasis. First, evidence to show that ambient
POM is a health problem does not exist. Secondly, a significant nega-
tive trend in ambient BaP concentrations indicates that such a relation-
ship to adverse health effects will be even more difficult to establish.
Controls by State/local agencies and those resulting from Federal en-
forcement actions should assist in continuing this downward trend.
Finally, NSPS for particulates and hydrocarbons expected by late 1975
should also result in further control of new sources.
2.3 Recommendations
As a result of appraising available options and supporting data
and considering the advisability of Federal action, the following
recommendations are made.
1. In order to minimize POM emissions EPA should continue to
support those national and local actions that are responsi-
ble for the downward trends in BaP levels. However, a
Federal regulatory program designed specifically for con-
trol of BaP is not warranted at this time.
2. Monitoring of ambient concentrations of BaP in selected
cities should be continued on a routine basis. Specifically,
quarterly composited NASN hi-vol samples for BaP concen-
trations at the same 30-40 sites studied for this report
(Appendix C, Table C-2) should be analyzed routinely. Section
7.0 gives additional information on current research
under way by NERC/RTP.
25
-------
3. Special emphasis should be given to prompt enforcement of
existing regulations for coal combustion, incinerators,
open burning (including coal refuse piles), and coking
operations. The same emphasis should be placed on new
regulations as they are promulgated for these and other
sources of POM.
26
-------
3. SOURCES OF BaP
The literature indicates that BaP usually is associated with partic-
ulate matter. Formation of BaP generally accompanies any incomplete
comhustion of fossil fuels or organic compounds, especially coal. The
amount of BaP produced and emitted is postulated to be directly pro-
portional to the efficiency of combustion and control techniques applied.
Of course, controlled combustion is not the only source of BaP; some
occurs from natural causes, e.g. forest fires. Also, natural synthesis
of BaP has been postulated by several scientists. The more common
sources of BaP are listed in Table I and will be mentioned in more
detail below. Values given in this section are estimates based on the
best available information. Methods used to arrive at these estimates
may be obtained from the references.
3.1 Stationary
The category of stationary sources of BaP encompasses a wide variety
of processes that contribute to local BaP concentrations and accounts
for 98% of the nationwide estimate. This category is subdivided into
heat and power generation, refuse burning (which includes forest and
agricultural burning), and industrial activities. The most comprehen-
sive examination of stationary sources of BaP available is an EPA pub-
Q
lication (AP-33) by Hangebrauck et al. Other POM compounds were
measured also, but BaP received primary emphasis; this, results of
emissions are reported as units of BaP.
27
-------
3.1.1 Heat and Power Generation
The majority of heat and power produced and used in the United
States comes from fossil fuels (coal, oil, gas, or wood) and accounts
lor nbonL '17% ol nationwide iiaP emissions. Of these the primary source
of BaP appears to result from coal combustion in hand-fired residential
furnaces. Data from all four fuels, in terms of quantities of BaP
produced, may be seen in Table II. Emissions are dependent on efficiency
of combustion rather than the type of fuel used. This is shown by the
low emission factor for intermediate size coal units and coal-fired steam
power plants, in relation to the gross heat input. Oil and gas have
relatively low emission factors, while the emission factor for wood is
low when compared with hand-stoked residential furnaces. These data are
consistent with knowledge of POM formation processes, i.e., reducing
conditions caused by insufficient oxygen. As may be expected, local
variations in quantities of emissions will exist in given areas of the
United States depending on the type fuel, number of industries, and
other local conditions.
Total BaP emissions given in Table II must be regarded as subject
to change because of several developments. The two main developments
relate to two residential heating practices, woodburning fireplaces and
residential heating with coal. These type units have the highest emission
rates and are affected by trends of an affluent society. Wood-burning
fireplaces are on the increase in pooularity in new urban areas; local
air pollution control agencies are banning, or attempting to ban, resi-
i.
dential coal burning. Based on available data the total BaP emissions
from heat and power generation in the United States are estimated at
30
approximately 340 tons annually.
28
-------
Table II. ESTIMATED BENZO(a)PYRENE EMISSIONS
FROM HEAT AND POWER GENERATION SOURCES IN THE U.S., 1972*
Type of Unit
Coal
Hand-stoked residential
furnaces
Intermediate units (chain-
grate and spreader
stokers)
Coal -fired steam power
plants
Oil
Low-pressure air-atomized
Other
Gas
Premix burners
Wood
Gross Heat
(BTU/hr)
0.1 x 106
60-250 x 106
1,000-2,000 x 106
0.7 x 106 ,
0.02-21 x 10b
0.01-9 x 106
-
Benzo(a)pyrene
Emission Factor
(ug/106 BTU)
1.7-3.3 x 106
15-40
20-400
900
100
20-200
5 x 104
Benzo(a)
pyrene
Emissions
(tons /year)
300
7
<1
2
2
25
*Adapted from references 3 and 30.
3.1.2 Refuse Burning
In refuse burning, as in coal burning, efficiency of combustion
governs BaP emissions. Inefficient combustion occurs in all open
burning and also in most small incinerators; thus, formation of BaP
results. On the other hand, enclosed incinerators designed and operated
for specific tasks should result in efficient combustion and less BaP
emissions. Presently, this category contributes about 40% to the
national total.
Q
Hangebrauck et al give BaP emission factors for: (1) municipal
and commercial incinerators for certain wastes; (2) municipal and
agricultural open burning; and (3) junked vehicle disposal. As ex-
29
-------
Table III. ESTIMATED BENZO(a)PYRENE EMISSIONS
FROM REFUSE BURNING IN THE UNITED STATES
(1968)*
Source of Benzo(a)pyrene
Benzo(a)pyrene
Emissions (tons/year)
Enclosed incineration
Municipal
Commercial and industrial
Institutional
Apartment
Open burning
Municipal
Commercial and industrial
Domestic
Forest and agricultural
Vehicle disposal
Coal refuse fires
23
2
8
4
10
10
11
50
340
*Adapted from reference 3: an estimate based-on 1972 data indicates a total
of approximately 340 tons being emitted, ' accounted for as follows:
a) coal refuse fires - 310 tons
b) forest and agricultural.- lltons
c) vehicle disposal - S^tfths
d) other open burning - 10 tons
e) enclosed incineration - 3 tons
The method of estimating emissions may have varied for the two years.
pected, these emission factors vary widely and again show the importance
of efficient combustion. Large municipal incinerators (50-250 tons of
refuse/day) have a BaP emission rate of 0.1 - 6 mg/lb of charged refuse.
Commercial incinerator (3-5 tons/day) emission factors ranged from 50-
260 mg/lb. Emission factors for municipal open burning are given an
average value of 150 mg/lb while vehicle disposal results in about
mg/lb of refuse. Higher emissions of BaP from refuse burning than
originally estimated (20 tons/year) by Hangebrauck e£ ai^ have been
30
-------
projected recently. These later emissions, shown in Table III, are
based on higher estimates of nationwide refuse burning, and the in-
clusion of coal refuse fires, rather than significantly different
emission factors. The single largest contributor identified is the coal
refuse pile under "Open Burning." Such waste piles are common in coal
mining areas and are frequently referred to as culm (or gob) piles (or
banks).
The given estimates should be regarded as an order—of-magnitude
approximation because of the uncertainty of factors used to estimate
emissions. Therefore, approximately 340 tons/year of BaP from refuse
burning is the best available value.
3.1.3 Industrial Activity
Some industrial processes are conducive to direct source measure-
ment while others must use indirect means, often imprecise, for esti-
mating emissions. The former category includes catalytic cracking of
petroleum and air-blowing of asphalt, probably the most obvious sources
of BaP in the petroleum refining industry. The latter category of
emissions estimating, indirect sampling, may be illustrated by coke
production in the steel industry; carbon black, coal-tar pitch, and
asphalt hot-road-mix processes; and general chemical processes. The
procedure generally used consists of sampling the ambient air as near
the expected source, or complex of sources, as possible and back-cal-
culating emissions. This technique necessarily yields much less accurate
results because of background ambient concentrations and other factors.
31
-------
Industrial emissions of BaP are summarized in Table IV and account
for about 20% of the national total. Experience has indicated that
those industrial processes that must be sampled indirectly contribute
insignificant quantities of BaP to the environment, with the exception
of coke production. Estimates of BaP emissions from coke ovens range
from about 0.06 tons per year to approximately 170 tons/year (Appendix
F). The higher value is used here because it is based on data from the
U.S. The high BaP discharge from coke production appears to be associated
with the effluent from the charging and coking processes. A crude emission
factor of 2.5 g of BaP/ton of coke produced was applied to estimate BaP
emissions from coke production. A more accurate assessment of coke oven
emissions is being developed.
Table IV. SUMMARY OF ESTIMATED INDUSTRIAL
BENZO(a)PYRENE EMISSIONS IN THE UNITED STATES
(1972)*
Source of
Benzo(a)pyrene
Petroleum
Asphalt air-blowing
Coke production
Benzo(a)pyrene
Emissions
(tons/year)
(0.06-) 170
*References 31; 8; and Appendix F.
Obviously, other industrial processes are sources of BaP (and POM)
but are more localized and less significant as a nationwide problem.
Thus, the current best estimate for BaP emissions from industrial sources
is about 175 tons/year in the United States.
32
-------
3.2 Mobile
Generally, when one thinks of mobile sources, thoughts of automo-
biles immediately enter one's mind. Of course, sources other than the
automobile contribute to this group; for this reason, many people
prefer to call this category "transportation" sources rather than
"mobile." Because of difficulties associated with the type testing
O / "7 O O
required, literature ' ' ' on vehicular sources of BaP emissions is
sparse.
3.2.1 Gasoline-powered Vehicles
Of the various transportation sources contributing to emissions of
BaP, the automobile has been studied the most. Contributions of gasoline-
powered vehicles can be separated into vehicular characteristics and
fuel composition effects. The first category includes the effects of
the air:fuel ratio; emission control devices; driver operating modes;
engine deterioration; and combustion chamber deposits. The second
category is composed of fuel aromaticity, BaP levels, additives, and
lubricants.
As with other sources of BaP, efficient combustion apparently
results in less BaP emissions. Air:fuel ratios greater than stoichio-
metric, i.e. use of excess air, promote efficient combustion in the
cylinders of internal combustion engines. Recent data indicate that BaP
production decreases about 30-fold with an increase of the air:fuel
2
ratio from 10:1 to 14:1.
33
-------
4 7 32
Various researchers ' ' have shown that other factors mentioned
above also increase emission quantities of BaP. Engine age causes
deterioration of operating parts, thus permitting more oil consumption
and concomitant BaP emissions. Also, excessive wear creates more space
for deposits to accumulate in the combustion area. These deposits,
through a complex process, cause increased BaP emissions.
Since BaP is composed of aromatic compounds, it stands to reason
that the more basic aromatics in the fuel the easier for BaP to be
formed and emitted. However, based on recent research, higher fuel
aromaticity may be offset by using unleaded gasoline which changes the
32
character of combustion chamber deposits. Other research in progress
tends to dispute such a conclusion. Nevertheless, one thing appears to
be certain, with the use of unleaded gasoline the distribution of BaP
emissions from mobile sources will be different. Development of emission
control devices introduces further complicating variables in the overall
32
effect but studies thus far indicate these devices are beneficial.
3.2.2 Diesel-fuel-powered Vehicles
Studies on diesel-fuel-powered vehicles have been limited mainly to
trucks and buses under laboratory conditions. Actual on-the-road
operation of diesel-powered vehicles can result in high BaP emissions
due to engine overloading, poor maintenance, and other factors. If
diesel engines are not overloaded, laboratory tests have indicated that
higher BaP emissions occur at idle than at any load condition. Further-
more, these same tests indicated that half-load produced highest BaP
emissions during operation. Emissions dropped sharply at full load,
presumably because of increased combustion chamber temperatures re-
3
suiting in more efficient burning.
34
-------
lier mohi/e. Sources, likely e.prrlribu'fe, *£ J7 e/sj/ss/
-------
3.3 Natural
Data presented thus far provide good substantiation that BaP is
primarily a product of incomplete combustion. However, a few sources of
BaP may be defined as naturally occurring. The first of these has
already been alluded to—it is bituminous coal. In addition to being
analyzed for BaP content, coal has also yielded benzo(a)anthracene and
other unspecified POM. Two of three types of asbestos used industrially
contain appreciable quantities of natural oils. These oils have been
found to contain BaP.
Two additional sources may be more difficult to verify but no
question of their natural occurrence should arise. One researcher has
identified various molds In the environment as a source and BaP. Several
scientists have indicated that BaP is synthesized naturally in the
environment.
36
-------
4. CONTROL TECHNOLOGY
The nature of emissions referenced above provides a broad hint to
control procedures. Since inefficient combustion of fossil fuels is a
primary source of BaP, by far the most important control technique is
to burn fuels efficiently. Fuel or process substitution may be
directly applicable to many processes of BaP emissions. For example,
gas or oil are inherently more efficient than coal; incineration is
much more desirable than open burning. These two methods have been
adopted widely recently.
A desirable end result may be possible only by adding control
equipment in line with the regular process. In particular, reduction
of BaP from a source may be achieved by treating the effluent gas
stream. Bag filters, scrubbers, or electrostatic precipitators are
standard devices for removing particulates upon which BaP is likely
absorbed. Afterburners, catalytic mufflers, or condensation may be
applicable techniques to control vapor phase POM.
4.1 Stationary
Since the most important contributors of BaP emissions in
heat and power generation are hand-fired coal furnaces and wood
burning, the preferred choice of control would be by alternative fuel
sources. One might argue that substitution is the only means of
37
-------
control for these two sources as neither is amenable to better
engineering controls.
In regard to refuse burning, efficient incinerators are being
installed for municipal, commercial, industrial, and apartment
building sources. Relative contributions from these sources are
diminishing in comparison to coal refuse banks. Existing burning
refuse banks are required to be extinguished if burning began
spontaneously. Federal regulations have been proposed for all burning
coal refuse piles which will eliminate most emissions as these
regulations are implemented. Future culm piles are being eliminated
by more specific attention to refuse accumulation practices in mining
areas.
Benzo(a)pyrene emissions from catalytic cracking in the petroleum
industry apparently are receiving necessary action through utilizing
carbon monoxide (CO) waste boilers for effective control (see Table
VI). From Table VI it is clear that CO waste-heat boilers have a
signficant beneficial effect on BaP emissions, and thus on other POM
emissions.
Much of the BaP produced at by-product coke ovens (Appendix D)
quite likely could be removed in the by-product stream, except for
that escaping from a leaky system. However, BaP emissions from most
coke processing still appears to be quite high, whether it originates
from charging, leaks, or directly from the gas stream. Some possible
approaches to controlling BaP from coke ovens are under investigation
by EPA and industry groups; two demonstration projects for particulate
38
-------
Table VI. ESTIMATED BENZO(a)PYRENE EMISSIONS
FROM CATALYTIC CRACKING SOURCES IN THE UNITED STATES
(1968)*
Type of Cracking Unit
FCC
no boiler
CO boiler
Subtotal
HCC
no boiler
CO boiler
Subtotal
TCC (air-lift)
no boiler
CO boiler
Subtotal
TCC (bucket-lift)
no boiler
CO boiler
Subtotal
Total
Petroleum Consumption
(million barrels/year)
424
1,230
1,654
14
55
69
27
118
145
17
75
92
1,960
Benzo(a)pyrene
Emission (tons/year)
0.08
0.02
0.10
'3.4
0.0
3.4
2.4
0.0
2.4
0.0
0.0
0.0
5.9
*Adapted from reference 3; reference 31 gives 1973 petroleum consumption
as 2,372 million barrels/year which yields about 7 tons of BaP annually.
control from coke ovens are in progress and BaP sampling is being done
simultaneously.
The Control Systems Laboratory, NERC/RTP, began coke oven
particulate control demonstration projects in 1972. These projects
were designed 'to demonstrate the feasibility of controlling
particulate emissions from coke ovens, on both the charging and
pushing sides of the oven. Particulate control demonstration projects
are, for the most part, being accomplished on existing coke ovens.
The purpose of demonstrating the availability of control technology
was to gather background information for standards of performance for
f\ f\
new coke ovens. However, since the demonstration projects are on
39
-------
existing sources, preliminary results indicate that particulate
controls are technically feasible for existing as well as new sources.
Since coke oven controls are complex and many questions must be
answered, the various projects underway by EPA are divided into
several phases. Only initial phases of both charging and pushing
control demonstration have been completed. However, from observations
during these phases, particulate control appears to be feasible.
In the charging-phase demonstration project, source sampling for
BaP was performed in conjunction with particulate sampling.
Preliminary analysis of sampling results indicate that about 90
23
percent control of particulates is achieved. With this level of
particulate control, a similar degree of reduction in BaP emissions
was expected. Preliminary evaluation of the BaP fraction of the
particulate from the controlled versus the non-controlled larry car
(vehicle that drops coal into the oven) indicates that approximately
85-90 percent control is achieved.^3 if these results are obtained for
other types of particulate control (e.g., pipeline charging*), control
processes should also reduce BaP emissions. Such results confirm
previous expectations of control engineers that particulate control
will significantly reduce BaP emissions.
Presently, no data are available on the amount of BaP control
achievable on the pushing side of the coke oven. The second phase of
a demonstration project is underway to obtain control estimates on the
pushing side. Based on observations of installed systems, engineers
* Pipeline charging is a method for pre-heating coal and charging
it into coke ovens via pipelines in a totally enclosed system;
licensed by Coaltek Associates.
40
-------
estimate that control on the pushing side should work as well as
on
controls on the charging side of the oven.
Thus, we see from Tables I-IV that BaP emissions originate
primarily from stationary sources, accounting for about 97 percent of
total nationwide BaP emissions. However, due to inadequate sampling
techniques these emissions should be considered estimates of relative
ranking. Furthermore, since stationary sources account for the
majority of BaP, these emissions are highly localized. Table VII
depicts some degree of localization for the three major sources. Such
localization indicates those areas where control efforts should be
emphasized.
4.2 Mobile
Before the current concern for reducing vehicle emissions was
instigated by the Clean Air Act, most vehicles operated with fuel-rich
carburetion to promote smooth performance and quick power response.
Preliminary modifications prior to the 1970 Clean Air Amendments
resulted in leaner fuel-air mixtures and lower BaP emissions. This
trend is indicated in Table VIII by the emission factors estimated in
1972 for the periods indicated. "Advanced systems" in Table VIII
refer to prototype emission control devices spurred into development
by the 1970 Act. Generally, advanced systems utilize catalytic
converters and thermal reactors.
Emission control devices will effect control of BaP concomitantly.
Data to support this view are based on current exhaust sampling •
32
techniques. If such a high degree of control is achieved in actual
practice, mobile source pollution caused by BaP will assume a much
41
-------
Table VII. CONTRIBUTIONS TO NATIONAL TOTALS OF
BENZO(a)PYRENE BY SOURCE AND STATE
(1972)*
Source of
benzo(a)pyrene
emission
Coal refuse burning
Total
Residential coal-
fired furnaces
Total
Coke production
Total
State
Pennsylvania
West Virginia
Ohio
Kentucky
Virginia
Alabama
Illinois
Pennsylvania
Kentucky
Ohio
Tennessee
Pennsylvania
Indiana
Ohio
Alabama }
Michigan /
W. Virginia )
Percent of U.S. total
from given source
contributed by given state
46
37
16
99
23
16
7
7
7
60
27
15
15
21
78
*References 28; 33; and 34.
less significant role. At any rate, one may logically expect a
reduction of BaP emissions from light-duty vehicles as a result of the
EPA motor vehicle control activity. A higher degree of uncertainty
exists for other type transportation sources, e.g., buses, heavy diesel
trucks, aircraft, etc.
42
-------
Table VIII
AUTOMOTIVE BENZO(a)PYRENE EMISSION FACTORS
(1972)*
Source
Benzo(a)pyrene emission factors
Ug/gal of fuel consumed)
Uncontrolled car (1956-1964)
1966 Uncontrolled car
1968 Emission-controlled vehicle
Advanced systems**
170
20-80
6-24
*Adapted from reference 3 & 32
**Prototype emission control systems
4.3 Natural
Control of BaP from natural sources is even more limited than
other types control because natural emissions are less well-defined.
For bituminous coal, product substitution would be the logical
approach to control. Of course, suitable substitutes for specific use
and application must be found. Asbestos control under NESHAPs should
preclude BaP from the environment also.
As for control of BaP from natural synthesis and molds, no
suggestion has been forthcoming. Their contribution to ambient
concentrations is so negligible that it is likely not worth time and
money required for exploration.
43
-------
5. AMBIENT CONCENTRATIONS
Nationwide ambient concentrations for most pollutants of interest,
including BSO and BaP, are available through the National Air
Surveillance Network (NASN). Concentrations available through this
network provide only relative nationwide comparisons because of
limited sampling sites in a given area. An additional problem in
using NASN data for ambient BSO and BaP concentrations is that none of
these sites were selected specifically to sample for organic
pollutants. However, representative data are available from some
suspended particulate sites in the NASN. In addition, summary reports
on POM reference various special studies that have been undertaken
since the late 1950's.
If a detailed breakdown of NASN data by quarter were examined, a
seasonal trend for BSO and BaP would be obvious with few exceptions.
Higher concentrations would be noticeable for the fall and winter
quarters at most sites while the summer generally has the lowest
concentrations. Studies conducted in the late 1950's indicated this
12 11
same pattern. '
5.1 Measurement Technique
To better understand the measurement of ambient BaP data, a
brief description of the process should be helpful. Particulate
samples are collected upon tared fiber glass (flash-fired) filters by
45
-------
drawing air through the filters at a relatively high volume of air
flow (aproximately 50-60 cfm) during a 24-hour day. These filters are
sent to a laboratory for preliminary weight analysis, after which
organics are extracted from a portion of the filter by the Soxhlet
method for about six hours using a suitable solvent [ordinarily
benzene—thus the term benzene soluble organics (BSD)]. Subsequently,
BaP is separated from total BSD by thin layer chromatography (TLC)
followed by ultraviolet (UV) spectral analysis.
A minimum of 50 mg of BSO material is required for a BaP
analysis using TLC and UV. In order to obtain a sufficient quantity
of sample for this analysis several organic fractions from one station
must be pooled. Usually each urban 24-hour hi-vol particulate sample
will contain about 25 mg of benzene extract. Generally, samples for a
three month period, consisting of 5-7 samples, are composited to
obtain one data point. Thus, an annual station mean would consist of
averaging four quarters of composite samples for one station (see
Appendix C). Table IX is an annual summary of BSO, BaP, and TSP for
the last three years of analysis from 121 selected NASN sites where
BaP was analyzed on a routine basis. Table C-2 of Appendix C lists
those sites which were analyzed after a special request for 1971 and
1972 BaP data. TSP is an abbreviation for total suspended
particulates, which is obtained from a weight analysis of hi-vol
filters and relating this to average air flow for the 24-hour sample
period. Note from Table IX that BaP trends follow the average BSO
content and not TSP sampled.
46
-------
Table IX. THREE YEAR SUMMARY OF TSPa, BSO,
AND BaP* FOR 121 NASN SITES IN THE
UNITED STATES - 1968, 1969, AND 1970
Year
1968
1969
1970
Pollutant
TSPa
(yg/m3)b
89 (88)d
84 (81)
96 (88)
.BSOa
(yg/m3)
5.7 (5.5)
5.0 (5.0)
4.3 (4.3)
BaPa
(ng/m3)c
2.4 (1.6)
2.3 (1.6)
2.0 (1.2)
*See Appendix C for the annual average for each
selected site for each year.
aTSP = total suspended particulate; BSO = ben-
zene soluble organics; and BaP = benzo(a)-
pyrene.
yg/m3 = micrograms of sample/cubic meter of
air.
cng/m3 = nanograms of sample/cubic meter of air.
Numbers in parens are median station values
for the 121 sites.
No known routine studies by State or local agencies are being
conducted on ambient levels of BSO and BaP. Several States have
inquired about analytical procedures but have given no indication of
follow-through. At any rate, no BaP data from State or local programs
are recorded in OAQPS's national data bank.
5.2 Ambient BaP Concentrations
Figure 5 is a graphical representation of annual average BaP
concentrations for coke-oven and noncoke-oven cities for 1966 to 1972.
This figure is presented to show the general trend in the two type
cities. The overall trend is decreasing in each case. With the
exception of the dip in 1968 for the coke-oven cities, and 1967 for
47
-------
1967
1968 1969
TIME, year
1970
1971
1972
Figure 5. Trends in annual BaP concentrations in cities with and without coke ovens.
noncoke-oven cities, the trend is quite similar for both type cities.
In each case, the number of cities included in the annual average
varies from year to year because only those cities with a full year's
data (4 valid quarters) were included (Appendix C, Table C-3).
Average BaP concentrations for coke-oven cities have
decreased from 4.74 nanograms per cubic meter (ng/m ) in 1966 to 2.14
ng/m3 in 1972 (Figure 5 and Table C-3). This is a 55% decrease in a
six-year period. The decrease in noncoke-oven cities is even more
conspicuous: a 77% decline occurred over the same six-year period,
o O
with concentrations dropping from 2.76 ng/m to 0.64 ng/m .
Individual trends were decreasing for most of the 34 cities used in
computing BaP averages. However, some cities showed an upward
variation in some years. One city, Chattanooga, Tenn., showed no
decline at all over the six-year period based on a rank correlation
coefficient of zero. Two other cities, Pittsburgh, Pa. and Ashland,
48
-------
Ky., showed a weak decreasing trend pattern with the rank correlation
of BaP versus time of -0.14 and -0.40, respectively, because of
positive fluctuations in the trend line.
Realizing that whatever is happening in the noncoke-oven
cities is also happening in the coke-oven cities for the overall
trend, one can still surmise certain occurrences from information
available. Even though limited data do indicate that coke ovens are a
major contributor to localized BaP concentrations, one must remain
aware that coke ovens are not the sole source of BaP emissions. For
example, the stockpiling of steel in 1967 in anticipation of a steel
strike in 1968 likely contributed to the noticeable decrease of BaP
concentrations in 1968 for coke-oven cities (Figure 5). On the other
hand, an abrupt decline in BaP concentrations occurred in noncoke-oven
cities in 1967, indicating the possibility that less coal was
available for nonsteel industry use, especially small consumer use.
Again, one must remember that no readily evident method
exists for identifying the specific contribution of coke ovens because
the amount of coal delivered to individual cities is not available.
However, the circumstantial evidence presented is quite strong. The
trends analysis presented below establishes that a significant
difference exists in BaP concentrations for cities with and without
coke ovens.
5.3 Trends Analysis for BaP
Two statistical techniques were used to establish whether the
apparently decreasing trend (Figure 5) in BaP concentrations was
significant. These techniques were a linear regression and Spearman
49
-------
rank correlation analyses. The Spearman rank correlation procedure, a
nonparametric approach correlating the rank of BaP concentrations and
time, was used to avoid the assumption of a certain distributional
form for the data, and other requirements inherent in parametric
correlation and regression analysis. This procedure was used on the
data for each of the cities studied while the regression technique was
applied to individual averages separately for the coke-oven and
noncoke-oven cities. An analysis of variance (ANOVA) was used to
establish whether a significant difference in BaP concentrations
existed between coke-oven and noncoke-oven cities.
The general trend for BaP is decreasing. Regression analysis
performed on the cities with and without coke ovens for BaP
concentrations versus time verify that the trends shown in Figure 5
are significant, i.e., the decreasing pattern in BaP concentrations
over time are unlikely to have occurred by chance alone under the
hypothesis that a trend is not present in the data. Therefore, this
suggests that the data support an alternative hypothesis of a trend
component being present in the data. The linear regression for both
the coke and noncoke cities could be described as parallel lines,
i.e., lines with the same negative slope but being displaced by a
significant difference in concentration between the two sets of lines
representing the different cities (see Figure 5).
The Spearman rank correlation analysis was applied to the
individual cities by looking at relative rankings of annual average
BaP concentrations versus time (years). Of the 34 individual sites
analyzed, all have negative correlation coefficients, except
50
-------
Chattanooga which was zero. Of these, 12 sites were statistically
significant. A statistically significant result (a = 0.05) implies
that an observed correlation coefficient was not likely to be derived
from a population of coefficients whose true correlation is zero.
Statistically, we can infer that a significant negative association
exists between BaP and time for about 35% (12 of 34) of the sites.
These results support the conclusion that a decreasing trend does
exist for BaP concentrations over time.
In establishing a significant difference between cities with
and without coke ovens an analysis of variance procedure was used to
evaluate BaP concentration variations. With no way for considering
the many additional uncertainties (other than time) which influence
BaP concentrations in cities with coke ovens, the ANOVA shows that a
statistically significant difference does exist between the two type
cities. In fact, only a very small probability (0.0001) exists that
the difference in the observed BaP concentrations between the two sets
of cities could have occurred by chance alone, under the hypothesis
that no difference exists in BaP concentrations between the two sets
of cities. In other words, based on the seven year's available data,
BaP concentrations are higher in cities having a coke oven, although
it is impossible to evaluate the contribution of other sources of BaP
to explain this difference.
5.4 National Urban and Nonurban Trends for BSD
Composite annual averages of BSD concentrations and BSO
expressed as a percent of TSP are shown in Figure 6 for 32 urban and
19 nonurban NASN stations. The urban composite average concentrations
51
-------
o
o
o
in
CO
12
10
1960
(ONLY 18 NONURBAN STATIONS
INCLUDED IN THIS COMPOSITE
AVERAGE).
1961
1962 1963
1964 1965 1966
TIME, year
1967
1968
1969
1970
Q.
KO
o
OO
CO
1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970
TIME, year
Figure 6. Trends in BSD and in BSD percentage of TSP at 32 urban and 19 nonurban stations.
of BSD have decreased from 10.6 Mg/m3 in 1960 to 4.8 Mg/m3 in 1970.
This represents a 55 percent decrease. The national trend in BSO
concentrations shown closely parallels the trends observed at the vast
52
-------
majority of the stations included in the composite sample.
Statistically downward trends were found for 27 out of 32 stations
represented by this grouping. (The same statistical techniques
described under BaP trends were used for BSD.) Only in Helena,
Montana, was there some indication found of an increasing trend
pattern. Portland, Oregon, did not exhibit a clear-cut trend pattern
and was described as having "no trend," whereas Providence, Rhode
Island, Seattle, Washington, and Milwaukee, Wisconsin, exhibited
weaker decreasing trend patterns than was the general case. The
overwhelming consistency in trend patterns points to some general
phenomenon exclusive of geographical location within the country, city
size, or industrial sources present in a particular city.
The decrease in the composite average concentrations of BSO
in 1963 — approximately 1 vg/m — together with similar decreases
in 1964 and 1970, represent the years with the largest concentration
changes. The nonurban average BSO concentration, which is a factor of
approximately 4 or 5 less than the urban averages, exhibits a more
uncertain pattern. However, since 1966, nonurban BSO averages have
decreased from 2.2 Mg/m3 to 1.2 Mg/m3 in 1970.
The BSO percent of TSP has also decreased for the urban
stations over the period considered. The pattern is characterized by
rather sudden decreases from 1963 to 1964 (8 to 6 percent) and from
1969 to 1970 (6 to 4.5 percent). In other years, the BSO percent of
TSP remained relatively constant. It appears that significant
decreases in national urban levels of both BSO and BSO percent of TSP
have occurred, notwithstanding sizable, short-term, year-to-year
fluctuations.
53
-------
Nonurban trends in the BSO concentration and the BSD percent
of TSP have also been downward during the 4-year period, 1966-1970.
It is interesting to note that the averages of the BSO concentration
and the BSO percent of TSP for the nonurban stations (with the
exception of Curry County, Oregon) were lower in 1965 and 1966.
However, insufficient pre-1965 data exist to judge whether trends in
i-
nonurban BSO concentrations have paralleled the concentration
decreases at the urban sites.
54
-------
6. EFFECTS OF BaP
Description of lung diseases related to dust or airborne particles
was advanced first in the sixteenth century. Two other researchers
demonstrated more than three hundred years later that lung cancer was a
prominent pulmonary disease among miners.^ Now in the mind of the general
public, lung cancer is associated with cigarette smoking because of
superabundance of publicity, as well as "the warning on the label." The
relation between cigarette smoking and lung cancer has stimulated interest
tn tho role of air pollution in cancer. This interest has arisen because
some pollutants found in urban areas, including BaP, are similar to
those found in cigarette smoke. The necessity of defining all significant
etiologic factors in lung cancer has become evident because of its
rising incidence and poor prognosis. At this stage, primary prevention
is the most effective means of lung cancer control.
6.1 Epidemiological Studies
Some types of comparison strongly support the hypothesis that
increased lung cancer mortality rates are related to urban pollution.
The epidemiolopic method of studying air pollution effects on lung
cancer incidence entails comparing lung cancer death rates in communities
that have markedly different pollution levels, but are similar in other
potentially causal factors. Comparisons between urban and rural areas
within the same country are attractive in that, generally, fewer variables
are involved and pollution level differences are maximized. The disadvantage
55
-------
is that very little data on ambient concentrations of pollution are available
for rural areas. Thus, comparisons between urban areas offer an advantage
over comparisons between urban and rural areas but have other disadvantages
in that different variables may affect the study. Each comparison
requires a suitable adjustment for any known extraneous variables which
may affect it.
6.1.1 Classification of Studies
Four groups of epidemiologic studies are used to arrive at evidence
of environmentally related lung cancer incidence in humans. They are
studies to:
1. Compare urban metropolitan populations with rural
populations in relation to lung cancer mortality rates,
usually without examining etiologic factors.
2. Compare lung cancer death rates in migrants to those
in country of origin and those in country to which they
migrate; and examine change in rate in relation to change
in the environment.
3. Compare demographic units and study the relation between
lung cancer death rates and various indices of pollution.
Multiple regression techniques are generally used to
separate environmental effects from other factors.
4. Sample characteristics of lung cancer decedents through
family interviews and compare these with corresponding
characteristics of the remaining population.
Through such studies the prospect is offered of identifying a relatively
sharp distinction between strongly related factors and those incidentally
associated with lung cancer.
56
-------
Urban-rural differences undoubtedly offer clues to problems of
pulmonary carcinogensis in man. The general association between
urbanization and increased lung cancer is not questioned. However,
characteristics of the urban environment which are primarily
responsible for lung cancer are a subject of controversy. The problem
of identifying causal factors is amplified by difficulties of
obtaining either accurate or extensive measures of exposure for most
factors. Further intensification of the problem occurs from the close
association of urban pollution factors with one another. For example,
an area with high BaP levels will generally tend to have high concen-
trations of other hydrocarbons and sulfur dioxide.
6.1.2 Consideration of Other Variables
Special attention should be paid to imperfections or contradic-
tions in the correlation of air pollution measurements and lung cancer
incidence. General agreement apparently exists on the increment in the
incidence of pulmonary cancer caused by smoking; interpretations differ,
however. This is of immense interest, because urban air and cigarette
13
smoke have carcinogenic substances, e.g. BaP, and some gases in common.
Such similarity is likely the reason frequent reference is made to
studies of cigarette smoking and pulmonary cancer. However, complex-
ities of the problem of pulmonary carcinogensis are compounded, rather
than simplified, by that fact.
Considering the numerous variables that are known and anticipating
that some are likely completely unknown, one should focus attention
on the target tissue and consider the air as the primary point of
contact with carcinogenic agents. Two specific variables which must
57
-------
be considered in epidemiologic studies are irritant or toxic gases
in the atmosphere and diseased tissues at the target site. These
variables are difficult to quantify or to determine their specific
relation to carcinogenicity. The common link between these two
variables is their potentiating action on carcinogenic properties
of BaP. Tissues altered by disease may incur a precancerous state
*.
which appears to contribute to carcinoma susceptibility.
No attention has been given to effects on domestic animals
because no information appears to be available. Of course, except for
epidemiological studies, most data on the carcinogenicity of BaP
have been obtained by using experimental animals. Various techniques
have been used to verify carcinogenicity in these animals. A compre-
3
hensive review of these studies is presented in the NAS report
and the NERC/RTP position paper on PPOft.
6.1.3 Conclusions from Epidemiological Studies
Studies of lung cancer thus far have indicated the following
conclusions, generally resulting from the four broad study classifi-
3
cations listed above.
1. Lung cancer has emerged as the single greatest cause
of cancer death in males and is a significant cause in
females in the U. S. Its incidence has increased in the
last 30 years.
2. A major etiologic factor appears to be cigarette smoking;
however, smoking does not account completely for this
increased incidence.
58
-------
3. Urban dwellers have approximately twice as high an
incidence of lung cancer as those living in rural areas
after correcting for smoking. Within urban areas, in-
cidence hns been greater in areas of industrial pollution.
4. Renzo(n)pyrene, found in cigarette smoke in high concen-
trations, causes cancer of the lung and other organs in
experimental animals. Also, it is present in the air at
relatively high levels in industries whose workers have
been found to have high mortality rates due to lung
diseases, including cancer. Finally, it prevails at
different levels in all urban cummunities investigated.
5. Generally, immigrants have an incidence of lung cancer
between that noted for their countries of origin and that
of the countries to which they migrate. The higher their
ages at migration the nearer their incidence rates to
their cohorts who do not migrate. In studies where the
home country had a greater cancer incidence rate, rates
of migrants decreased significantly, even if their cigarette
smoking increased.
6.2 Other Effects
Other health effects of POM are related mainly to skin disorders
from occupational exposure of man to high levels. Studies of these
effects do not attempt to differentiate exposures by air from exposures
by other means, e.g., contact. This aspect should be considered care-
fully in attempting to extrapolate the information derived for occupa-
tional problems to the possible hazards of ambient air pollution. No
59
-------
documentation exists that participate materials containing POM in ambient
air have resulted in adverse skin effects.
Welfare effects of BaP are even more inconclusive because few
specific studies have been conducted in this area. However, the prob-
ability of altering soil microbial populations appears to exist and
such action may upset the ecosystem. The nature of BaP indicates
that it may be produced in many plants or plant products; however,
these data are speculative.
60
-------
7. PROGRAMS TO STUDY BaP (AND OTHER POM)
Polycyclic organic matter has been studied for two hundred years
among occupational xjorkers. Descriptions of lung disease related to
exposure to dust or airborne pollutants were said to exist as early as
the 16th century. The first recorded description relating the burning
of fossil fuel to occupational disease was written in 1775 by Percivall
Pott. This was related to constant exposure to soot among chimney
sweeps who had developed cancer of the scrotum. Then in 1879 two German
scientists demonstrated that lung cancer was a prominent pulmonary
3
disease among miners. Thus, occupational exposures provided the first
evidence of cause-effect relations from airborne pollutants that now
have been associated with BaP.
Since those early years many studies have been undertaken to
define cause-effect relationships between airborne contaminants and lung,
disease and to identify the causal agent in airborne pollutants. In more
recent years, beginning in the mid- to late 1950's, specific programs
have been initiated to establish such a relationship with the frequency
and complexity of studies increasing each year. Such has been possible
because more and more people and institutions, or sponsors, have become
interested in obtaining as much information as possible concerning BaP
and other POM compounds. However, in all of this interest no forcefu],
common objective has been evident. As a result, much effort has been
duplicated and much money utilized unnecessarily.
61
-------
The purpose of this secton is to give an analysis of those programs
concentrating on BaP and other POM that are underway by NERC/RTP.
7.1 Current Research Projects
Table X is a synoptic listing of those programs which have a
direct relation to emissions of BaP and/or other POM compounds. These
projects were extracted from the NKRC/RTP draft Position Paper on PPOM.
Items of interest pertinent to the listed subjects will be given below.
7.1.1 Analysis Methodology
Several approaches to the improvement of analytical techniques are
under investigation. The primary objective of most of these projects is
to develop more rapid, and hopefully less expensive, analytical method-
ologies for various POM. Analytical techniques that can be easily
applied are needed so that any agency can routinely monitor and measure
BaP. Some analytical methods which are under investigation include
projects 1, 2, 4, and 5 of Table X.
7.1.2 Effects Investigation
Project No. 11 has been delayed for about 1 1/2 years because
exposure-room furnishings have been impeded by equipment delays.
In the meantime, characterization of aerosols, design of containment
systems, sampling techniques, etc. are continuing so that the inhalation
study can be initiated immediately after the room is operable. The
feasibility study listed as project 10 may be one of the more critical
in relating BaP (and other selected POM) exposure to adverse human
health effects. This project was begun in FY 1974.
62
-------
Table X. APPLICABLE CURRENT POM RESEARCH PROJECTS
Project
1.
2.
3.
4.
5.
6.
7.
0.
9.
10.
11.
12.
Development of fluorescent
detector for analysis of
carcinogenic HC
High pressure (up to
1,2000 psi ) liquid chro-
matographic determination
of total PAH
PAH profile in urban air
(ratios of various com-
pounds)
Rapid, direct gas chro-
matographic determination
of BaP
Rapid GC determination of
atmospheric PAH
Synthesis & purification
of carcinogenic air
standards
Characterization & con-
trol of air pollutant
emissions from fuel
combustion
Enclosed coke pushing &
quenching demonstration
Smokeless coke &
charging demonstration
Feasibil ity study:
Determine which tissues
best reflect PPOM expo-
sures. Develop methodo-
logy for tissue analysis.
Physiologic & pathologic
responses of animals to
inhalation of carcinogenic
hydrocarbons
Development of station-
ary source sampling
procedures for PNA
Date
begun
(FY)
1972
1972
1972
1972
1972
1972
1972
1972
1970
-
1973
1973
Scheduled
Completion
1974
1974
1974
1974
1973
1975
1977
1973
1973
1975
1978
1974
Estimate
of cost?
(51000)
47
3
31
54
55
28
200
800
950
200
500
34.4
Time
required
(Yrs.)
2
2
1.5
1.5
1
3
5
1
3
1
5
1
NOTES:
from NERC/RTP Research Objective Accomplishment Plans (ROAP)
and NERC/RTP Position Paper on PPOM (Draft), Dec. 1972.
p
Estimates of costs for multi-year projects are for informational
purposes only; they are not to be construed as accurate since an
order of magnitude could separate estimates & actual costs. Single
year projects scheduled to end in 1973 should be nearly accurate,
discounting extension of completion time.
63
-------
7.1.3 Stationary Sources
Four projects (#7, 8, 9, & 12) involving BaP (and other POM) from
stationary sources were in progress during FY 1973. Three of them were
initiated earlier by the Control Systems Laboratory. The fourth (// 12) ,
which began the latter part of FY 1973, is a project to develop an
improved source sampling procedure for BaP (and other POM). The initial
phase will be for about one year with options to explore any problems in
more detail if funds are made available. Project 8 was delayed approximately
six months because of a job accident. Fuels referred to in project 7
are such things as methanol, low BTU gas, etc.
7.1.4 Ambient Analysis
Data available under this grouping should be of primary interest
because ambient concentrations of POM compounds are ill-defined. Project
3 is of major importance in achieving a better understanding of ambient
levels of BaP and other compounds. The objective of this study is to
define, as accurately as possible in selected urban areas, specific
ratios of BaP to other POM compounds and vice versa. Project 5 could
easily fall in this category also, because standard compounds must be
available for reliable analytical results. A limited number (about 6)
of POM compounds were scheduled to be purified under the present contract.
7.2 Recommended Research Projects
A study which should receive early emphasis is the development of
methodology to simplify and expedite ambient measurements. Analytical
technique appears to be the primary area of need in expediting ambient
measurements. The need for simplified analytical methodology was force-
fully emphasized during preparation of this report when additional
64
-------
ambient BaP concentrations for 1971 and 1972 were required for trends
analysis in selected cities. A significant time delay was entailed
because of time required to analyze for BaP. Simplified analytical
methodology will expedite availability of ambient data for tracking
trends of BaP (and other POM) concentrations. When new methodology is
developed, it should be more easily applicable by State and local
agencies also.
65
-------
8. REFERENCES
1. Andreuzzi, Frank C., A Method for Extinguishing and Removing Burning
Coal Refuse Banks, Bureau of Mines, U.S. Dept. of Interior, 1C
8485, Washington, B.C., 1970.
2. Begemen, Charles R. and Joseph M. Colucci, "Polynuclear Aromatic
Hydrocarbon Emissions from Automotive Engines," Society of
Atomotive Engineers, Mid-year Meeting, Detroit, Mich., May 18-22,
1970.
3. Committee on Biologic Effects of Atmospheric Pollutants, Particulate
Polycyclic Organic Matter, National Academy of Sciences, Washington,
B.C., 1972.
4. Coordinating Research Council, Inc., CRC-APRAC Status Report,
New York, N.Y., Jan. 1973.
5. Dixon, J.R., B.B. Lowe, B.E. Richards, L.J. Cralley, and H.E.
Stokinger, "The Role of Trace Metals in Chemical Carcinogenesis:
Asbestos Cancers," Cancer Research. 30. pp. 1068-1074, April 1970.
6. Duncan, L.J., "Analysis of Final State Implementation Plans -
Rules and Regulations," U.S. EPA Publication //APTB-1334, Research
Triangle Park, N.C., July 1972.
7. Esso Research and Engineering Co., "Gasoline Composition and Vehicle
Exhaust Gas Polynuclear Aromatic Content," 2nd Annual Report, CRC-
APRAC Project Cape-6-68, Linden, N.J., April 1971.
8. Hangebrauck, R.P., B.J. von Lehmden, and J.E. Meeker, Sources of
Polynuclear Hydrocarbons in the Atmosphere, U.S. BHEW, Public
Health Service Publication No. AP-33, Cincinnati, Ohio, 1967.
9. Intersociety Committee, Manual of Methods of Ambient Air Sampling
and Analysis. Vol. 2, January 1970; also published as Health Laboratory
Science, Vol. 7, No. 1 (Supplement), American Public Health Association
Albany, N.Y., Jan. 1970.
10. McKay, Lewis M., Coal Refuse Fires, An Environmental Hazard.
Bureau of Mines, U.S. Dept. of Interior, Information Circular 8515,
Washington, D.C., 1971.
11. Olsen, Douglas and James L. Haynes, Preliminary Air Pollution
Survey of Organic Carcinogens, Litton Systems, Inc., NAPCA Publi-
cation APTD 69-43, Raleigh, N.C., October 1969.
67
-------
12. Rawlcki, Eugene, "Airborne Carcinogens and Allied Compounds,"
Archives of Environmental Health, Vol. 14, pp. 46-53, American
Medical Assoc., Jan. 1967.
13. Sawicki, Eugene, Walter C. Elbert. Thomas R. Hauser, Francis T.
Fox, and Thomas W. Stanley, "Benzo(a)pyrene Content of the Air of
American Communities," AIHA Journal. Vol. 21, No. 6, pp. 443-451,
Dec. 1960.
14. Shabad, L.M., Y.L. Cohan, A.P. Ilnitsky, A. Ya. Khesina, N.P.
Shcherbak, and G.A. Smirnov, "The Carcinogenic Hydrocarbon Benzo
(a)pyrene in the Soil," Journal of the National Cancer Institute,
Vol. 47, No. 6, pp. 1179-1191, Dec. 1971.
15. Stenburg, Robert L., Darryl J. von Lehmden, and Robert P. Hangebrauck,
"Sample Collection Techniques for Combustion Sources - Benzopyrene
Determination," AIHA Journal, Vol. 22, No. 4, pp. 271-275, Aug.
1961.
16. Task Panel, National Environmental Research Center, "Draft NERC/RTP
Position Paper on Particulate Polycyclic Organic Material (PPOM),"
U.S. EPA, Research Triangle Park, N.C., December 1972.
17. U.S. Bureau of Mines, Methods and Costs of Coal Refuse Disposal and
Reclamation, U.S. Dept. of Interior, Information Circular 8576,
Washington, D.C., 1973.
18. U.S. Dept. of Health, Education, and Welfare, Air Quality Criteria
for Particulate Matter, Public Health Service, NAPCA Publication
No. AP-49, pp. 128-144, Washington, D.C., Jan. 1969.
19. U.S. Dept. of Health, Education, & Welfare, Occupational Exposure
to Coke Oven Emissions (Criteria for a Recommended Standard...),
Public Health Service NIOSH, 1973.
20. U.S. Dept. of Health, Education, & Welfare, United States Metro-
politan Mortality: 1959-1961, Public Health Service Publication,
AP-39, pp.v-xi and 59-87, Cincinnati, Ohio, 1967.
21. Unpublished Contractor's report, Preliminary Results by personal
communication with the CSL staff, EPA Contract //68-02-0290, Feb.
1974.
22. Personal communication with Robert McCrillis, CSL, 2/12/74.
23. Personal communication with R. V. Hendricks, CSL, based on preliminary
results from EPA Contract //CPA 70-165, 2/6/74.
68
-------
24. Faber, Paul V., "Pipeline Coke-Oven Charging," Chemical Engineering.
Dec. 24, 1973, pp. 36-37.
25. E. Allegrini, "Cost of Retrofitting Coke Oven Particulate Controls,"
EPA Contract //68-02-0299, Vulcan-Cincinnati, Inc., Cincinnati,
Ohio, 1974.
26. Sebasta, William, "Ferrous Metallurgical Processes," Ch. 36, Air
Pollution, 2nd ed., Vol. 3, A. C. Stern, ed., Academic Press, New
York, 1968, pp. 144-145.
27. Shreve, R. N., Chemical Process Industries. 2nd ed., McGraw-Hill
Book Co., Inc., New York, 1956, p. 90.
28. TI. S. EPA, 1972 National Emissions Report: National Emissions
Data Systems (NEDS) of the Aerometric and Emissions Reporting
System (AEROS). EPA-450/2-74-012, Research Triangle Park, N.C.,
June 1974.
29. Brinkerhoff, Ronald J., "Inventory of Intermediate-Size Incinerators
in the United States - 1972", Pollution Engineering, 5:11, November,
1973, Greenwich, Conn. pp. 33-38.
30. Personal Communication, John R. O'Connor, OAQPS, Cost Analysis Branch,
August 27, 1974.
31. Cottrell, Ailleen, "Annual Refining Survey", Oil & Gas Journal,
72:13, pp.82-84, April 1, 1974, Tulsa, Okla.
32. Gross, George P., "The Effect of Fuel and Vehicle Variables on
Polynuclear Aromatic Hydrocarbon and Phenol Emissions," Society of
Automotive Engineers, presented at Automotive Engineering Congress,
Detroit, Mich., Jan. 10-14, 1972.
33. U. S. Bureau of Mines, Minerals Yearbook 1972; Metals, Minerals,
and Fuels, Vol. I, pp. 427-437, U. S. Dept. of Interior, Washington,
D.C., 1974.
34. ,U. S. Dept. of Interior Task Force to Study Coal Waste Hazards, "List
of Coal Waste Banks," Internal Document, June 15, 1972.
69
-------
APPENDICES
71
-------
APPENDIX A - OVERVIEW OF A PREFERRED STANDARDS PATH
Generally, six major options for control of air pollutants are
available to the Environmental Protection Agency (EPA) under the Clean
Air Amendments of 1970 (Act): (1) National ambient air quality stand-
ards (NAAQS) - Sections 108, 109 and 110; (2) standards of performance for
new sources (NSPS)- Section 111; (3) national emission standards for
hazardous air pollutants (NESHAP)— Section 112; and (4) Title II, Emis-
sion Standards for Moving Sources (ESFMS) - Sections 202 (automobiles);
211 (regulation of fuels); and 231 (aircraft emissions). A seventh
option for consideration, if data evaluation indicates it, is that of no
Federal action. Other options may be identified by subdividing or
combining the above seven options. The process of selecting the most
appropriate option is called a preferred standards path (PSP) analysis.
However, a PSP decision may not be possible if adequate data are not
available for determining the preferred option. Thus, an eighth major
option becomes available: postponement of a decision until more data
become available. Of course, this latter option should carry with it
recommendations of what data are needed and a suggested priority for
obtaining such data.
A. Purpose of a Preferred Standards Path
The purpose of a PSP is to recommend a regulatory course of
action to develop applicable standards resulting in control of
A-l
-------
a given pollutant. Recommendations must be based on a thorough
state-of-the-art assessment of emissions and health effects
for the pollutant as related to objectives of the Act. Basic
objectives of the Clean Air Act are described below.
1. Protection of health and welfare by attainment and main-
tenance of a designated numerical ambient standard:
*
This objective is applicable to those pollutants for
which adverse effects levels (health or welfare) may be
determined and that exist in excessive concentrations
which are relatively easilv measureable in the ambient
air. Usually, this goal is achieveable by setting NAAQS
following procedures set forth in Sections 108-110 of the
Act.
2. Enhancement of air qualitv: This goal is for those
pollutants which may be tolerated in exiguous quantities
but their increase is limited in order to prevent the
creation of new problems. As related to stationary
sources, this goal may be achieved best by applying
Section 111 of the Act. Title IT—Sections 202, 211,
and 231—will be used for moving sources.
3. Protection of life: This objective is the ultimate
purpose of the Act. It applies to those pollutants that
are present in the ambient air and must be reduced to
safe levels. CTenerally, these pollutants cannot be
readily measured and concentrations must be reduced to
a minimum (or possibly to zero if the air pollutant is
A-2
-------
hazardous as defined in the Act) to achieve the goal of
protecting life. Thus, Section 112 of the Act would be
applicable to such pollutants.
B. Necessary Analysis for a Preferred Standards Path
Information needed to develop a standard will depend to a
large degree on the recommended control option, i.e., which
section(s) of the Act is (are) to be used in attaining the
prescribed objective. A thorough evaluation which leads to
such a recommendation is the heart of a PSP analysis. There-
fore, most items which should be considered in this type
analysis are included in the following subsections.
1. Summary of the existing problem - Without an exhaustive
knowledge of the existing situation no effective analysis
can be undertaken. Specific factors to be examined in
summarizing current conditions are described briefly
herein. yFigure A-l is provided as a guide in the logic
of analyzing a candidate pollutant for determining what
control strategy may be applicable.
a. Information on the pollutant's effects: A quanti-
tative criteria document is not required to deter-
mine a pollutant's effects on human health and
welfare. However, health effects must be quantified
for a candidate pollutant prior to implementing
Sections 109 and 112 of the Act. Both these sec-
tions of the Act provide general guidance to the
A-3
-------
•Adverse Health and
Welfare Ellecls: or
Possibility ol En-
dangerment of Health
and Welfare
3
Magnitude of Health and Welfare
Effects: Contribute to Increase
in Mortality, or Serious Irrever-
sible or Incapacitating Reversi-
ble Illness
No
7
Nature and C
of Sources:
ness - Wheth
or Diverse, S
Moving, or 6
No
istribution
Ubiquitous-
er Numerous
tationary or
oth
No
ACTION
2
No Action
Under Clean
Air Act
4 5-6
<• .n-rr, Adequate Data to Nature and Distri- No
See NOTE 1 Show Hazardous El- 1" * bution of Sources- 1 »•
Yes lecls al c""ent Am- Ubiquitous?
bient Concentrations
No Optional
8
Is the Pollutant
Moving Sources? 12
u Can Control BP Adequate Data
n irhi.u=^ RU Available to Yes
— 1' ' Hcmeveo By .N , *• ,.-, D.J , — — ••
Yes * NoAcL " f™«» *» U - "° Yes ^,~
- 3; Under ESFM5 . £_,„ rjont[0| Be and Welfare
Yei Kegulationi1 [7
9 | ^_ 1 I^No
Is the Pollutant I "
*" Prominent From No ^
Stationary Sources? I'D " See Note 7
' ' Adequate Data Available
Yes to Establish and Support yes
for Health and Welfare
Effects?
N° No^ Consider
1 -
1
Section 1 12, NESHAP"
Most Likely Appropriate*
13
Section 211, ROF,"
Most Likely Appropri-
ate*
15
Sections 202 and 231,
ESFMS, ** Most Likely
Appropriate*
,7
Section 109, NAAQS,"
Most Likely Appropri-
ate*
18
Section 111, NSPS,-
Mosl Likely Appro-
priate*
NOTE 1: A yes answer here simply means the pollutant is a candidate lor regulatory action; however, it does not denote mandatory action.
NOTE 2: The dashed line indicates that significant pollution from stationary sources may remain even after utilizing ESFMS. Land use and transportation
controls may need to be implemented in addition to Section 202.
'MOST LIKELY APPROPRIATE simply means that the indicated option may be the most logical without considering external factors. However, other ramifications
may preclude its use. Also, more than one option (e.g., Sections 109 and 202: Sections 109 and 111; etc.) may be used in combination For any given pollutant in
order to achieve the lull intent of the Act. (Also see NOTES.) Section lllidl is applicable to non-criteria pollutants.
**Acronyms explained in text.
Figure A-1 . Preferred standards path: Guide for determination of regulatory action.
-------
nature and severity of effects needed before action
can be taken under either section. Some differen-
tiation between chronic and acute effects is needed.
The objective is to estimate the likelihood of
occurrence and severity of adverse effects.
b. Ambient levels: Without a good estimate of ambient
concentrations of the candidate pollutant little
appreciation for its health and/or welfare effects
is attained. An evaluation of ambient levels
should determine whether the pollutant is ubiquitous
or is related only to specific point sources. If
sources are numerous or diverse and cause adverse
health effects at ambient levels then Section 109
may be applicable. If effects are caused by limited
point sources Section 111 may be recommended. Of
course, Title II must be used for moving sources.
c. Sources of the pollutant: A precise emission
inventory is not required; however, the following
information is needed: (1) an estimate of emissions
from various sources, along with the probability of
resulting ambient levels; (2) the expected growth
and/or decline of the number and type of source
categories that emit the pollutant; and (3) the
number of individual sources within a given category.
This information determines whether observed effects
are caused by localized or numerous and diverse
A-5
-------
sources. Contributions from mobile sources must be
included also.
d. Technology assessment: An evaluation of the state-
of-the-art in monitoring and control technology must
be available. Monitoring technology includes ambient
and source sampling instrumentation. Control tech-
nology includes the feasibility of applying existing
methods to various process configurations within an
industry. Furthermore, this section should evaluate
the effectiveness of reducing emissions beyond those
levels achievable with current control methods.
e. Environmental impact: A PSP should consider the
impact of the suggested option upon the environment.
The completeness of this consideration may be dicta-
ted by the option under consideration. For example,
if a NESHAP option is being studied, cost is pre-
cluded as an item for consideration; health effects
are of primary importance. Generally, this section
should contain such things as the accumulation of
the pollutant in the environment, its effect on the
fate of other pollutants, and other impacts.
2. Summary of existing regulatory efforts - In this sub-
section, the status of existing regulatory efforts should
be examined in relation to how the selected PSP would
impact these regulations. State, local, and Federal
control programs should be considered in order that
A-6
-------
pertinent actions for a new pollutant would not be con-
tradictory, or cause some additional obstacle in en-
forcing existing regulations. These possibilities may be
considered in the following two ways.
a. Existing State, local, or Federal regulations may
achieve partial control for a candidate pollutant.
If so, the degree of control may be determined by
*
summarizing the number and type regulations and by
determining their degree of enforcement.
b. Some control of the candidate pollutant may result
from emission reductions of other pollutants, voluntary
controls by the emitter, or Federal policy. Here,
EPA must decide whether this method yields the
desired control, or if the procedure results in the
best control.
3. Summary of options available - The three options available
for controlling air pollution from stationary sources are
described in greater detail in this subsection. As pre-
viously stated, "no control" or "postponement of a
decision to control" are also valid options that may
result from a PSP analysis. One of the latter options
may be chosen if based on existing evidence, an adverse
effect cannot be shown to result from a candidate pol-
lutant. The Administrator of EPA has such broad, dis-
cretionary powers to reach a decision under the Act that
more attention is given below to the options.
A-7
-------
a. National Ambient Air Quality Standards: This approach
may be used if the candidate pollutant has "an
adverse effect on public health and welfare" and
ambient levels result "... from numerous or diverse
mobile or stationary sources" [Section 108(a)(l)].
A NAAQS includes a primary and secondary standard.
Primary standards must be at levels which, "based on [air
qualitv] criteria and allowing an adequate margin of
safety, are requisite to protect the public health" [Sec-
tion 109(b)(2)].
Since ambient levels are specified, NAAQSs are not
directly applicable to sources. However, the attainment
of an ambient standard may require source control. States
decide how to control source emissions in order to comply
with the ambient standards. Each State prepares an
implementation plan (SIP), subject to approval by the
Administrator [Section 110(a)(2)]. SIP regulations may
utilize emission standards, land use policies, trans-
portation control measures, or other approaches. The
State has wide leeway for controlling new and existing
sources except for moving sources, which are specifically
exempted [Section 116]. If a State submits an unaccept-
able plan, EPA is required to promulgate regulations that
will achieve the standard(s) [Section 110(c)]. Likewise,
if a State does not enforce its regulations this respon-
sibility reverts to the Federal government [Section
113(8)1' A-g
-------
The milestones prescribed by the Act for NAAQS are more complex
than for the other options and thus can result in longer delays before
control is achieved. Significant milestones are listed below.
MILESTONE ACTION
1. Start
2. Within 12 months
3. Within 3 more months
4. Within 9 more months
5. Within 4 more months
6. Within 2 more months
7. Within 3 more years
(4)
Inclusion of candidate
pollutant on Section 108
list.
Issue air quality criteria
and proposed standards
(primary and secondary).
Promulgate standards.
Submit SIP's
Approval/disapproval of
SIP. If-not submitted or
disapproved, prompt proposal
of plan by EPA.O)
Promulgate EPA plan if SIP
remains inadequate.
Attainment of primary(5)
standard.
The Act delineated deadlines for initiating the regulatory process
in each option (including the other two described below) for certain
pollutants and sources (e.g., the six pollutants with existing air
quality criteria documents, Group I NSPS, etc.). It also provided for
subsequent inclusion of pollutants on relevant lists "from time to
time." Once begun, however, the statutory timetable binds EPA to the
described milestones.
2
The Administrator may extend the deadline for submitting SIP's for
secondary standards an additional 18 months.
3
If a State fails to revise an existing SIP within 60 days, or a
longer prescribed period, upon notification by the Administrator, EPA
revisions are to be proposed promptly.
4
If adequate technology is unavailable, an extension of up to two
years may be granted by the Administrator for the attainment of a
primary standard.
Secondary standards must be attained within a reasonable time as
specified within the SIP.
A-9
-------
b. New Source Performance Standards: This option mnv
he used if a category of sources "mav contribute
significantly to air pollution which causes or
contributes to the endangerment of public health or
welfare" [Section lll(b)(A)]. National standards
established under Section 111 are applicable to (])
new stationary sources for specific categories and
(2) modifications resulting in new emissions of air
pollution. The Administrator can distinguish among
classes, types, and sizes within categories. Section
lll(d)(l) requires States to control existing sources
covered by NSPS applicable to the specified category
unless the pollutant is covered under NAAQS or
NESHAP. If the pollutant is covered bv Section
108(a), NAAQS, the States can prevent construction
or modification of stationary sources which may
interfere with attainment or maintenance of a NAAQS
[Section 110(a)(A)]. With proper enforcement, SIP's
control existing sources as necessary to attain and
maintain the NAAQS. Moreover, if a pollutant has
been designated as hazardous the standard will apply
to all designated sources—existing, modified, and
new.
The level of the standard "reflects the degree of
emission limitation achievable through the application of
the best system of emission reduction which (taking into
A-10
-------
account the cost of achieving such reduction) the Adminis-
trator determines has been adequately demonstrated"
[Section lll(a)(l)]. Thus, the degree of control (which
likely will not be more stringent than the standard) is
not related directly to health and/or welfare. States
also may use best demonstrated control methods in
setting standards for existing sources.
Enforcement of NSPS is a Federal responsibility
initially [Section 113(a)(3)] but may be delegated to
States if acceptable enforcement procedures are adopted
[Section lll(c)(l)]. In addition, each state must
adopt plans to regulate unmodified existing sources of
the same category, provided the pollutant is not a
criteria or hazardous pollutant. These plans are
similar to procedures required to achieve NAAQS [Section
lll(d)].
Milestones for NSPS do not allow for time extension
as under NAAQS. Milestones as given in the Act follow.
MILESTONE ACTION
1. Start Include category of
sources on Section
111 list.
2. Within 120 days Issue proposed standards.
3. Within 90 more days Promulgate standards.
Effective upon promulgation.
6
Presently no guidance is available to states in meeting specific
deadlines for controlling emissions from unmodified existing sources for
a given category. However, such guidelines have been proposed.
A-ll
-------
c. National Emission Standards for Hazardous Air Pollutants:
This approach may be used if the pollutant "mav
cause, or contribute to, an Increase In mortality or
an increase in serious irreversible, or incapacitating
reversible, illness" [Section 112(a)(l)]. Standards
under Section 112 are applicable to new and existing
stationary sources, with the provision of a 90 day
delay for existing sources [Section 112(c)(l)].
Authority to make distinctions or classifications,
as in NSPS, is subject to question. These emission
standards are established at a level which "provides
an ample margin of safety to protect the public
health from such hazardous air pollutant" [Section
112(b)(1)(B)]. Technology or cost of control can
not be considered, except perhaps in deciding what
margin of safety is ample. In order to assure an
ample margin of safety the worst case expected in
practice should be considered. This will involve
reviewing the total number of sources, ambient
concentrations, persistence of the pollutant, build-
up in body tissue, etc. If an ample margin of
safety, based on review of the above factors, re-
quired zero emissions a total ban could be accomplished
under Section 112. (A similar result may be possible
under Sections 108-110 if the ambient standards were
A-12
-------
set low enough. Zero emissions can be achieved
under Section 111 if technology to achieve 100 per
cent control is adequately demonstrated.) Enforce-
ment of NESHAP is a Federal responsibility [Section
113(a)(3)], but the Administrator may delegate author-
ity to States under certain conditions [Section 112
(d)]. However, this does not preclude the Federal
enforcement prerogative. Again, provisions for time
extensions are granted in the Act [Section 112(c)(l)
(B); (c)(2)]. Milestones outlined for NESHAP result
in the second quickest means for attainment of standard,
However, with provision for extension and exemption
for two-year periods, no definite time for attainment
of standards can be estimated. Milestones designated
in the Act are as follows:
MILESTONE ACTION
1. Start Include candidate pollutant
on Section 112 list.
2. Within 180 days Issue proposed standards.
3. Within 180 more Promulgate standards. Effee-
days tive for new source upon
promulgation, and 90 days
later for existing sources.
C. Discussion of Various Options
Under this section the ramifications and reasoning for choos-
ing a specific option should be given. Support for the recom-
mended control option as decided by the PSP analysis must be
substantiated by items discussed in this section. In other
A-13
-------
words, a detailed discussion of factors outlined in "B" above
must support the chosen control option. Insight into legal or
other challenges should be given in detail. Except for those
cases where prohibited by statute (e.g., Section 112) the
economic impact for each option should be included.
D. Recommendations
The objective of this Section is to succinctly state the
recommended control option and develop an accomplishment plan
to implement the recommended approach.
A-14
-------
APPENDIX B - LUNG CANtER MORTALITY IN SELECTED SMSAs
Table B-l is a ranking of Standard Metropolitan Statistical Areas
(SMSAs) by descending order of deaths from lung diseases. In comparing
the principal cities of the SMSA (column 2) in this table with sampling
sites of Table C-l in Appendix C, one may see that the majority of those
cities with relatively high lung cancer deaths have not been sampled for
BaP in the NASN.
A definitive program to sample and analyze the ambient air for BaP
in selected cities where high lung cancer incidence is established needs
to be undertaken. This should be considered a first step in defining
the relationship between the urban factor and carcinogens, specifically
BaP or other POM.
B-l
-------
Table B-l. WHITE MALE DEATHS AND DEATH RATES PER 100,000 POPULATION
PER YEAR BY SMSA* FOR MALIGNANT NEOPLASM OF TRACHEA, BRONCHUS, AND LUNG
1959-1961**
Rank SMSA
1 Charleston, S. C.
i? Albany, Ga.
3 Galveston, Texas
4 Lake Charles, La.
5 New Orleans, La.
6 Newport News, Va.
7 Montgomery, Ala.
8 Jersey City, N.J.
9 Shreveport, La.
10 Baton Rouge, La.
11 Norfolk, Va.
12 Jacksonville, Via.
13 Birmingham, Ala.
14 Dubuque, Iowa
15 Baltimore, Md.
16 Jackson, Mich.
17 Tyler, Texas
18 Houston, Texas
19 Monroe, La.
20 Mobile, Ala.
21 Pensacola, Fla.
22 Charleston, W. Va.
23 Honolulu, Hawaii
24 Miami, Fla.
25 Portland, Me.
26 Toledo, Ohio
Death
rate
65
64
63
63
61
59
58
56
56
55
55
54
53
53
52
52
52
51
51
50
50
49
49
49
49
49
Total
deaths
75
23
79
65
468
85
61
556
121
74
214
206
278
55
945
92
49
536
44
111
80
141
55
683
134
313
Rank
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
, 43
1 44
i
! 46
:' 47
48
49
50
51
SMSA
Albany-Schen., N.Y.
Beaumont, Texas
Memphis, Tenn.
Omaha, Neb. -la.
Richmond, Va.
Buffalo, N.Y.
Decatur, 111.
Savannah, Ga.
Dallas, Texas
Des Moines, Iowa
Philadelphia, Pa. -N.J.
St. Louis, Mo. -111.
Atlanta, Ga.
Cincinnati , Ohio
Erie, Pa.
Flint, Mich.
New York, N.Y.
Paterson, N.J.
Syracuse, N.Y.
Tampa, Fla.
Topeka, Kansas
Washington, D.C.
W. Palm Beach, Fla.
Wheeling, W.Va.-Ohio
Wilmington, Del .
Death
:ate
48
48
48
48
48
47
47
47
46
46
46
46
45
45
45
45
45
45
45
45
45
45
45
45
45
Total
Deaths
492
134
214
280
184
831
75
61
495
162
2482
1178
377
610
155
185
7025
752
354
682
79
723
160
141
186
*Standard Metropolitan
**Reference 20.
Statistical Area.
B-2
-------
APPENDIX C - NATIONAL AIR SURVEILLANCE NETWORK
AMBIENT AIR MEASUREMENTS FOR BENZO(a)PYRENE
Data given in this appendix are arithmetic averages of the four
quarterly composites of the listed urban sampling site. The annual
average is computed only if 3 or more quarterly composite values exceed
the minimum detectable concentration of 0.2 nanograms/cubic meter (ng/m^).
Table C-l contains annual averages for 121 selected NASN urban
sampling sites. These sites were selected from the total NASN system
based on valid data for each site for all three years. Each site had to
meet minimum requirements of the NASN for each quarter in order to have
a valid annual average for the given site. In order to be included in
the summary, each sampling station must have a minimum of five scheduled
samples collected and analyzed for any given quarter.
Table C-2 is a listing of 40 sites selected to update BaP con-
centrations in cities with and without coke ovens. Three sites were
selected in National Parks so nonurban background readings would be
available. A limited number (40) was chosen because of time and re-
source restrictions since routine analysis of NASN BaP was discontinued
in 1970.
Table C-3 gives valid annual averages of BaP for the sites listed
in Table C-2 for all years of record (1966-1972). Numbers in paren-
thesis are the number of valid sites included for the year.
C-l
-------
Table C-l. NATIONAL AIR SURVEILLANCE NETWORK AMBIENT AIR MEASUREMENTS
1
Location,
state and cjty
Alabama
Gadsden
Hunts vi lie
Montgomery
Alaska
Anchorage
Arizona
Gra/id Canyon National Park
Maricopa County
Tucson
Arkansas
Little Rock
Montgomery County
West Memphis
California
Glendale
Humboldt County
Long Beach
Los Angeles
Oakland
Riverside
Sacramento
San Bernardino
San Diego
San Francisco
Colorado
Denver
Connecticut
Hartford
New Haven
Florida
Jacksonville
Tampa
Georgia
Atlanta
Hawaii
Honolulu
Idaho
Boise City
Butte County
Benzo(a)pyrene [ng/m3(25°C)]
1968
2.37
2.74
2.93
1.68
0.19
0.47
0.67
0.89
0.23
2.24
1.55
0.31
2.09
1.83
1.62
1.29
1.43
1.00
1.22
1.83
2.25
1.42
1.38
2.94
1.46
1.83
0.55
2.01
0.17
1969
1.76
1.82
2.04
1.28
0.18
0.27
0.53
1.05
0.18
2.41
1.63
0.49
2.27
1.86
1.61
0.82
1.79
0.90
1.38
1.15
2.51
1.95
2.10
2.30
1.00
1.86
0.57
5.96
0.09
1970
2.50
1.56
1.31
0.77
0.10
0.28
0.41
0.65
0.13
0.59
0.97
0.11
1.00
1.23
0.95
0.67
0.72
0.83
0.65
0.63
2.20
1.37
1.21
1.35
0.46
0.92
0.19
1.14
0.07
C-2
-------
Table C-l (continued). NATIONAL AIR SURVEILLANCE NETWORK AMBIENT AIR
MEASUREMENTS FOR BENZO(a)PYRENEa, 121 SELECTED SITES, 1968-1970
Location,
state and city
Illinois
Chicago
Springfield
Indiana
East Chicago
Hammond
Indianapolis
Monroe County
Parke County
Iowa
Des Moines
Kansas
Topeka
Wichita
Kentucky
Ashland
Covington
Louisiana
New Orleans
Maine
Acadia National Park
Maryland
Baltimore
Massachusetts
Worchester
Michigan
Detroit
Flint
Grand Rapids
Trenton
Minnesota
Duluth
Minneapolis
Moorhead
St. Paul
Nebraska
Omaha
Thomas County
Benzo(a)pyrene [ng/m3(25°C)]
1968
3.10
1.06
4.90
2.08
4.07
0.54
0.43
1.12
0.69
1.04
9.30
3.56
1.55
0.33
2.31
1.65
5.13
0.78
3.44
1.44
2.73
1.14
0.85
1.82
1.91
0.20
1969
3.88
1.27
6.75
3.29
5.18
0.25
0.26
0.92
0.44
0.67
10.89
4.11
1.52
1.12
2.76
1.48
3.91
1.69
1.70
1.58
2.08
1.43
1.04
1.75
1.55
0.13
1970
2.00
0.85
5.26
1.67
2.32
0.16
0.39
0.69
0.31
0.50
6.67
4.38
1.14
0.20
2.06
1.63
2.56
1.49
0.87
0.84
1.09
0.62
1.59
1.01
1.01
0.12
C-3
-------
Table C-l (continued). NATIONAL AIR SURVEILLANCE NETWORK AMBIENT AIR
MEASUREMENTS FOR BENZO(a)PYRENEa? 121 SELECTED SITES, 1968-1970
Location,
state and city
Nevada
White Pine County
New Hampshire
Concord
Coos County
New Jersey
Camden
Glassboro
Jersey City
Newark
Paterson
Perth Amboy
Trenton
New Mexico
Albequerque
New York
• Jefferson County
North Carolina
Cape Hatteras National Park
Charlotte
Durham
North Dakota
Bismarck
Ohio
Cincinnati
Cleveland
Columbus
Dayton
Toledo
Youngstown
Oklahoma
Cherokee County
Oklahoma City
Tulsa
Oregon
Curry County
Portland
Benzo(a)pyrene [ng/m3(25°C)J
1968
0.14
0.98
0.20
1.57
1.20
2.35
2.14
1.95
1.17
1.04
1.77
0.19
.22
5.56
7.96
.86
1.77
3.00
2.21
2.36
1.80
5.64
.21
.73
.77
.14
4.13
1969
0.07
0.65
0.11
2.41
1.09
2.68
1.82
1.24
1.20
1.46
1.12
0.25
.11
4.85
3.38
.96
2.90
3.75
2.73
1.88
1.49
9.86
.18
.71
.47
.08
2.60
1970
0.12
0.61
0.14
1.92
1.18
4.65
1.53
1.20
1.01
1.10
1.06
0.24
.21
1.85
3.38
.44
2.58
2.78
1.57
1.48
1.38
7.12
.22
.88
.79
.09
2.31
C-4
-------
Table C-l (continued). NATIONAL AIR SURVEILLANCE NETWORK AMBIENT AIR
MEASUREMENTS FOR BENZO(a)PYRENEa, 121 SELECTED SITES, 1968-1970
Location,
state and city
Pennsylvania
All en town
Altoona
Bethlehen
Clarion County
Harrisburg
Philadelphia
Pittsburgh
Reading
Scranton
Wilkes-Barre
York
Rhode Island
East Providence
Providence
South Carolina
Greenville
Tennessee
Chattanooga
Memphis
Nashville
Texas
Matagorda County
San Antonio
Utah
Ogden
Salt Lake City
Vermont
Burlington
Orange County
Virginia
Danville
Hampton
Lynchburg
Norfolk
Portsmouth
Roanoke
Shenandoah National Park
Benzo(a)pyrene[ng/m3(25°C)]
1968
1.17
17.96
2.05
.97
1.32
2.87
6.31
2.36
6.08
1.58
1.91
1.17
1.95
18.55
7.39
1.34
5.96
.16
.85
.82
.97
.71
.31
2.47
1.53
8.72
4.90
10.17
7.65
.30
1969
1.92
22.28
2.00
1.23
1.45
4.03
13.75
1.75
7.65
1.54
2.00
1.21
2.15
7.00
4.17
.74
2.80
.12
.63
.67 .
.65
.48
.28
1.79
.88
6.28
3.91
3.39
5.30
.31
1970
2.40
19.25
2.71
1.23
1.53
2.44
5.86
1.56
2.88
1.30
1.21
1.22
2.11
3.39
5.54
1.36
3.61
.27
.95
2.49
1.44
.74
.16
2.69
1.08
4.49
1.67
4.94
6.20
.21
C-5
-------
Table C-l (continued). NATIONAL AIR SURVEILLANCE NETWORK AMBIENT AIR
MEASUREMENTS FOR BENZO(a)PYRENEa, 121 SELECTED SITES, 1968-1970
Location,
state and city
Washington
Seattle
West Virginia
Charleston
Wisconsin
Kenosha
Madison
Milwaukee
Superior
Wyoming
Casper
Cheyenne
Benzo(a)pyrene[ng/m3 (25°C)]
1968
1.97
4.57
1.41
1.33
4.65
3.30
.90
.60
1969
1.57
2.63
1.74
9.62
4.03
1.58
.60
.46
1970
1.51
2.11
1.31
1.07
2.46
1.50
.44
.43
Annual average.
'-' /-
C-6
-------
Table C-2. LISTING OF 40 NASN SITES SELECTED FOR 1971-72 BaP ANALYSIS
*1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
*12.
13.
14.
15.
Coke oven
Birmingham
Gadsden
Chicago
Gary
Indianapolis
Terre Haute
Ashland
Baltimore
Dearborn
Detroit
Trenton
Duluth
St. Paul
St. Louis
Buffalo
ci
16
17
18
19
*20
21
22
23
24
25
26
ties
. Cleveland
Toledo
Youngstown
. Bethlehem
Erie
. Philadelphia
. Pittsburgh
Chattanooga
. Houston
. Spokane
. Milwaukee
Noncoke oven cities
1.
2.
3.
4.
5.
6.
*7.
8.
9.
10.
*11.
12.
13.
*14.
Montgomery
Jacksonville
Honolulu
Hammong
Baton Rouge
New Orleans
Acadia National
Park
( a 1 1\
New York
Newport News
Norfolk
Shenandoah Nat'l .
n _ ._ i .
Hark
Seattle
Charleston, W. Va.
Grand Canyon
*Not included in Figure 5. Birmingham
data; Duluth had known extraneous infl
and Erie had insufficient
uences in 1971 and 1972.
Table C-3. ANNUAL BaP AVERAGES FOR SELECTED CITIES
1966-1972
Year
1966
1967
1968
1969
1970
1971
1972
Coke
4.74 (15)*
5.34 (15)
3.75 (18)
4.41 (23)
3.02 (21)
2.18 (11)
2.14 (19)
Non-coke
2.76 ( 7)
2.29 ( 8)
2.64 ( 8)
2.14 (11)
1.41 (11)
1.22 ( 8)
0.64 (11)
*Number of cities included in average (no. with
full year's valid data).
c-7
-------
APPENDIX D - DESCRIPTION OF BY-PRODUCT COKE PRODUCTION
Coke is the carbon residue of certain grades of bituminous coal
after destructive distillation. It is used as a fuel, primarily in
making pig iron which is an essential ingredient in steel. Destructive
distillation (coking) is accomplished principally in by-product ovens
and consists of driving certain volatile matter off coal, leaving a
residue with a high percentage of carbon and relatively small amounts of
impurities.
A brief description of the by-product coking process is presented
here as extracted from reference 26 of the text. By-product ovens are
usuallv constructed in groups called batteries. They consist of a block
of many long, narrow firebrick ovens with heating chambers made of
similar brick located between the ovens, so that a battery of ovens is a
huge block of coking cells separated from an intricate system of combustion
chambers (Figure D-l). The fuel used to heat coke ovens may be blast <-
i
furnace gas, coke oven gas, or natural gas, each fuel requiring a different
set of burner adjustments. None of the coal being coked is burned to
provide heat for the coking operation.
The carbonization of the coal begins soon after crushed and sized
coal is loaded into red-hot ovens. The gaseous products and condensates
are conveyed continuously via the large collecting mains from the coke
oven battery to an adjacent by-product plant. Usually each oven has a
steam jet aspirator which aids in conveying the gaseous carbonization
products into the collecting main.
D-l
-------
V
ro
CROSSOVER FLUE
CONTROL
HOUSE
/^r-S
T=?
SECTION THROUGH
AIR PORTS
7 BLAST
^ FURNACE
GAS MAIN
I ONG SECTION THROUGH LONG SECTION TRANSVERSE SECTION THROUGH
INSPECTION HOLES AND THROUGH CROSSOVER HEATING FLUES AND
WASTE-GAS AND AIR PORTS UNDERJET GAS-DUCTS
«£CIRCULATING DUCTS.
TRANSVERSE SECTION THROUGH
AN OVEN AND REGENERATOR
Figure D-1. Koppers-Becker underjet low-differential combination coke oven with waste-gas recirculation
(Courtesy of Koppers Company, Inc.)2?
-------
The charging holes on the top of the battery are closed almost
immediately after charging to minimize escape of gas or dust into the
atmosphere. Spacing of the charging holes, and sleeves on the charging
car hoppers, have been used to minimize such emissions. Smoothness of
operation and the physical characteristics of the coal being used aid in
the reduction of these emissions. Most operators attempt, through door
maintenance programs and operating procedures, to further reduce the
escape of gas through the charging holes. Various methods of control-
ling particulate and gaseous emissions during charging have been inves-
tigated. More widespread application of these methods has occurred in
Europe and Japan than in the U. S. However, recent attempts at application
have been undertaken by several U. S. companies, with varying degrees of
success.
In the by-product plant, tar, benzene, naphtha, and other com-
mercial products are segregated and separately removed from the coke
oven gas. The by-product recovery plant has processes similar to those
in petroleum refineries and chemical plants. The cleaned gas, having a
medium heating value, is customarily used as fuel in boilers and other
furnaces, including those of the coke oven. Since sulfur has an unde-
sirable influence in steelmaking, the steel industry has always en-
deavored to use materials with low sulfur content. Therefore, low
sulfur coal is used for coke making. Coke oven gas usually contains
some sulfur-bearing compounds derived from the sulfur in the coal.
After the coking period, the incandescent coke is pushed from the
oven into a quenching car where burning in air takes place. This car is
a large special-type railroad car with perforated sides, which is used
D-3
-------
to carry the flaming coke to an area where a huge quantity of water is
sprayed onto the coke to quench it, i.e., extinguish its incandescence.
During this operation, which stops the burning of the coke, large quanti-
ties of steam are generated. Hot coke extinguishing is usually performed
in a quenching tower. The quiescent coke is then conveyed to a cooling
wharf where the coke is spread out so that the excess water may drain
away. After cooling, the coke is sized in a plant similar to that for
screening rock and ores. As with the charging phase of coke production,
various attempts at controlling emissions from the pushing of finished
coke have been attempted with various degrees of success.
D-4
-------
APPENDIX E - COSTS OF CONTROL
AND GROWTH PATTERNS FOR COKE OVENS
Most polycyclic organic matter (POM) is emitted when coal is
charged to coke ovens, during the coking process, or when the finished
coke is pushed out of the oven.
Charging Controls
The charging process is not generally controlled at present. Two
possibilities for control are described. One is to equip the larry car
(vehicle that dumps coal into the ovens) to prevent escape of fumes.
This process is now in the demonstration stage but estimates indicate
that the capital cost for retrofit of an existing 70 oven battery would
approximate $1,500,000. Several things could increase this cost: (1)
the necessity to add additional vents to each oven; (2) adding holes in
the roof for vents; and (3) strengthening the batterv structure if
necessary to bear the weight of the new larry car, which weighs 100 tons
compared to 50 tons for a non-controlled car. Operating costs have not
been compiled as yet. This type of information is expected to be
available by the end of CY 1974.
Another control method already in use is pipeline charging. This
method eliminates the larry car and the dumping cycle. The entire
charging operation is confined within a pipe. For a new 70 oven battery
costing about $24,000,000, the control equipment can be added for
$8,000,000. However, since the coke is preheated in the process, claims
are made that the production capacity of the battery is increased by
50%. To obtain the same capacity without pipeline charging would cost
E-l
-------
$36,000,000. Therefore, the control system results in a 10% capital
i
saving at the greater production figure.
Retrofit costs will run $12 to $14 million for a 70 oven battery.
However, if a market is assumed for the additional coke produced,
installation of control equipment would cost about the same as adding
ovens. Claims are also made that operating savings from the use of a
higher percentage of cheaper, high volatile coal; the increase in yield
of blast furnace coke; and the shorter cycles would more than compensate
for the operating costs involved.
Coking Controls
Gases generated during the coking process are vented to a by-
products plant and thus are not released to the air. However, con-
siderable leakage often occurs around door seals. At present the only
remedy is careful operation and adequate maintenance, including cleaning
the seals at each cycle.
Oven Discharge Controls
Although most of the POM evolves during the coking cycle, it is
possible that some may evolve during the discharge cycle. One company
is testing an indexing hood connected to a scrubber to control parti-
culates. The company is continuing its tests of this control method for
charging and discharging. At this stage of testing, insufficient data
are available to judge the adequacy of performance or to accurately
estimate costs. Company representatives estimated total costs would
approximate $10,000,000 to retrofit their batteries that contain a total
of about 220 slots.
E-2
-------
Cost Summary
Subject to reservations stated above, the tables below detail
25
retrofit control costs for a 70 oven battery. These costs are estimates
only.
Table E-l. CAPITAL COSTS
(in $1,000)
Pipeline charging
Material and field costs $11,858
Modification to ovens and 546
installation of charging
main
Coordination and start-up 388
Subtotal $12,792
Pushing
Material and field costs 1,619
Overhead and profit 26
Coordination and start-up 35
Subtotal $ 1.680
Total $14,472
E-3
-------
Table E-2. ANNUAL COSTS
(in $1,000)
Pipeline charging
Maintenance at 6% Inv. $ 768
Utilities 318
Labor 229
Taxes, Ins. at 2% Inv. 256
Interest (8%) 1,023
Depreciation (15 yrs) 853
Subtotal$3,447
Pushing
Maintenance at 6% Inv. 101
Utilities 92
Labor 0
Taxes and Ins. at 2% Inv. 34
Interest (8%) 134
Depreciation (15 yrs) 112
Subtotal $ 473
Total $3,920
Assuming that a 70 oven coke battery produces 630,000 tons of
coke annually, the average cost per year to achieve about 85-90% BaP
control is approximately $6.25 per ton of coke. This compares with
33
the cost of coke at $40.70 per ton in 1972 (latest published data),
or about 15% of the market value. This annualized cost of retrofit
is significantly greater than that for a new coke battery installing
pipeline charging, where the estimated cost is about $0.10 per ton
of coke.
Capital costs for retrofitting a battery with a modified larry car
21
have been estimated at 1/5 to 1/10 those for pipeline charging. However,
operating costs have been projected to be much greater, even to the extent
E-4
-------
of accounting for the difference in capital costs. These operating costs
are only estimates and are not based on operational data because the first
phase of the demonstration project has just been completed. Additional
time will be required before operating costs can be obtained. Therefore,
no definitive conclusion can be stated for costs at this time. Neverthe-
less, if the above cost estimates are reasonable one can immediately see
the large differences between retrofitting coke ovens and installing
controls on a new battery.
Growth Patterns
Despite the historic growth of the steel industry, the consumption
of coke per ton of steel poured continues to drop. In 1960, 1542 pounds
of coke were used per ton of steel; by 1970 the amount of coke dropped
to 1248 pounds; and by 1980 a further drop to 1100 pounds is projected.
Thus, even though hot metal production is expected to increase at a
normal rate, coke production will not change appreciably.
However, because of air pollution regulations and obsolescence many
existing batteries will have to be replaced. The table below was taken
3?y
from a bv-product coke oven survey of Koppers Company.
E-5
-------
Table E-3. SUMMARY OF PROJECTED NEW OVEN CONSTRUCTION
UNITED STATES
Estimated date
for operation
1974
1975
1976
1977
1978
1979
1980
Number
of
batteries
2
5
7
5
3
2
5
Number
of
ovens
139
386
455
326
196
109
465
Annual
capacity
total coke
(1000 tons)
1,881
4,440
5,260
3,470
2,370
1,465
6,230
New coke
capacity
(1000 tons)
457
1,445
810
700
270
—
—
E-6
-------
APPENDIX F - EMISSION ESTIMATES FOR COKE OVENS
Emission data for BaP from coke ovens are limited. Estimates of
emissions which have been made from available data vary extensively, e.g.,
from 0.06 to 166 tons per year. Presently, crude sample data from
uncontrolled plants in the United States are available. Some data
4
have also been reported from the U.S.S.R and Czechoslovakia.
Demonstration projects are in progress at several plants within the U.S.
Data on BaP emissions from uncontrolled and controlled sources are being
gathered bv contractors; these data should be available soon. Until
these data are available, the following calculations, based on crude
data, are presented to demonstrate the method of arriving at the
estimate contained in the report. Total tons of by-product coke produced
in the U.S. are used in the calculations because total nationwide
emissions are of interest. However, beehive oven production is excluded
because in 1972 it accounted for approximately only 1% of the U.S. total.
All units are in U.S. net tons.
A. Estimates Based on U.S. Data:
1. Given:
a. Ratio of BaP to Total Particulates
= 1.734 x 10~3
(Reference 1, Smith)
b. Through-put loss of particulates in the coking
operation = 1.1 x 10 tons of particulate
F-l
-------
per ton of coal charged (minus unloading and
quenching)
(Reference 2, AP-42)
c. Quantity of coal consumed per ton of coke
produced 3 1.45
(Reference 3, Bureau of Mines)
d. Tons of coke produced in 1972 = 59.9 x 106
Tons produced at furnace plants = 54.2 x 10°
(Reference 3, Bureau of Mines)
2. Find:
Estimated BaP emissions from Coke ovens
(l.734xlO-3 - tonJteP - Vi.lxio-3 ton particulateX
V ton particulate/\ ton of coal /
tons. of coal\ / q Q6 6 tons coke\
M year )
t tons of coke
166 tons of BaP/yr.
B. Estimate Based on Soviet Data:
1. Given:
Ratio of BaP/Total Particulates = 1.05 x 10~3
(This value is the average of the maximum ratios
of two Soviet coal-coke plants.)
(Reference 4, Masek, Table 2)
F-2
-------
2. Find:
Estimated BaP emissions for U.S. coke ovens
assuming B.I is the accurate ratio.
^1.05 x 10~3 \Lff tons of BaP\ in. .. , ., D/
:H[166 J =101 tons of BaP/yr,
V1.734 x 10~V\ year
C. Estimate Based on Czechoslovak Data:
1 .
Ratio of BaP/Total Particulates = 6 x 10
(This estimate is the average value of BaP from
dust on the upper floor of the battery.)
(Reference 4, Table 5)
2. Find:
Estimated BaP emissions from U.S. coke
plants assuming C.I is the accurate ratio.
6 x 10~ \/,,-/• tons of BaP \ n », .. c „ „/
1166 I "0.06 ton of BaP/yr.
1 «• A *•»•»* I J
a.734 x 10
-3/1 year
F-3
-------
REFERENCES FOR APPENDIX F
1. Smith, William M., "Evaluation of Coke Oven Emissions," Presented
to the 78th General Meeting, AISI, New York, N.Y., May 28, 1970.
(Also presented at the 63rd Annual Meeting of APCA, St. Louis, Mo.,
June 14-18, 1970.)
2. OAQPS, Compilation of Air Pollutant Emission Factors, 2nd Edition,
U.S. EPA, pp. 7.2-1 - 7.2-3.
3. U.S. Bureau of Mines, Minerals Yearbook 1972; Metals, Minerals,
and Fuels, Vol. I, pp. 427-437, U.S. Dept. of Interior, Washington,
D.C., 1974.
4. Masek, Vaclav, "The Composition of Dusts from Work Sites of Coke
Ovens," Staub - Reinhaltung Luft, Vol. 30, No. 5, [English edition]
pp. 34-37, May 1970.
------- |