ASSESSMENT OF HUMAN
EXPOSURES TO ATMOSPHERIC
ETHYIENE DICHLORIDE
Final Report
May 1979
By:
Benjamin E. Suta
Prepared for:
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
Task Officer: Jack K. Greer, Jr.
Project Officer: Joseph D. Cirvello
Contract No. 53-02-2835 Tas,k 17
SRI Project CRU-6780
Center for Resource and Environmental Systems Studies
Report No. 82
-------
NOTICE
This report has been provided to the U.S. Environmental Protection
Agency (EPA) by SRI International, Menlo Park, California, in partial
fulfillment of Contract 68-02-2835. The opinions, findings, and
conclusions expressed herein are those of the authors and are not
necessarily those of EPA. Mention of company or product names is not to
be considered an endorsement by EPA.
-------
CONTENTS
LIST OF TABLES v
ACKNOWLEDGEMENTS vii
I INTRODUCTION 1
II SUMMARY 2
III CHEMICAL AND PHYSICAL PROPERTIES OF EDC
AND ITS ENVIRONMENTAL BEHAVIOR 8
Introduction 8
Physical Properties 8
Chemical Properties 9
Environmental Behavior 11
IV EDC PRODUCTION AND USES • • • 14
Production 14
Uses . . . . 14
EDC Producers and Users . 17
V POPULATION EXPOSURES FROM EDC PRODUCTION 21
General ....... 21
Sources of Emissions 21
Emissions 21
Atmospheric Concentrations 23
Exposure Estimates 28
VI POPULATION EXPOSURES FROM PRODUCERS
THAT USE EDC AS A FEEDSTOCK 34
General 34
Sources of Emissions 34
Emissions . 35
Atmospheric Concentrations 35
Exposure Estimates 35
iii
-------
VII POPULATION EXPOSURES FROM EDO IN AUTOMOBILE GASOLINE .... 40
General . . 40
Exposures from Self-Service Operations 41
Exposures in the Vicinity of Service Stations 46
Urban Exposures Related to Automobile Emissions .... 54
Summary of Urban Exposures from Automobile Gasoline . . 59
VIII OTHER ATMOSPHERIC EXPOSURE ROUTES 60
General • 60
Dispersive Uses 60
Transportation 61
Waste Disposal 64
BIBLIOGRAPHY 66
iv
-------
TABLES
II-l Summary of Estimated Population Exposures to
Atmospheric EDC from Specific Emissions Sources .... 4
II-2 Estimated Atmospheric Emissions of EDC for 1977 .... 5
III-l Physical Properties of EDC 10
IV-1 EDC Consumption 16
IV-2 EDC Producers and Major Consumers 18
IV-3 1977 EDC Production by Direct Chlorination and
Oxychlorination 19
IV-4 1977 Use of EDC Production Capacities 20
V-l EDC Oxychlorination Vent Emissions 22
V-2 Estimated Atmospheric Emissions from EDC Production
Facilities 24
V-3 Atmospheric EDC Monitoring Data for Calvert City,
Kentucky 25
V-4 Atmospheric EDC Monitoring Data for Lake Charles,
Louisiana 26
V-5 Atmospheric EDC Monitoring Data for New Orleans,
Louisiana 27
V-6 Estimated One-Hour Average Downwind Atmospheric
Concentrations of EDC (yg/m3) . . . 29
V-7 Estimated Human Population Exposures to
Atmospheric EDC Emitted by Producers 31
V-8 Comparison of EDC Monitoring and Modeling
Atmospheric Concentrations (ppb) . . 34
VI-1 Estimated EDC Atmospheric Emissions (g/s) for
Plants that Use EDC as a Feedstock 37
VI-2 Estimates of Population Exposures to Atmospheric EDC
Emitted by Plants that Use EDC as a Feedstock in
Various Products 39
VI-3 Estimates of Total Population Exposures to Atmospheric
EDC Emitted by Plants that Use EDC as a Feedstock ... 40
VII-1 Self-Service Operations 43
VII-2 Gasoline Market Share of Self-Service Stations in
Four AQCRs, Spring 1977 44
-------
VII-3
VI1-4
VII-S.
VII-6
VII-7
VII-8
VII-9
VII-10
VIII-1
VIII-2
Gasoline Market Share of Self-Service Stations
in Two Metropolitan Areas, 1976
Sampling Data from Self-Service Gasoline Pumping . . .
Estimates of EDC Exposures from Self-Service
Gasoline Pumping
Service Station Density in Four Metropolitan AQCRs . .
Rough Dispersion Modeling Results for EDC Emissions
for Gasoline Service Stations
Automotive EDC Emission Factors
Distribution of Cities by 1970 Population
Estimated U.S. City Exposures to EDC from the
Evaporation of Automobile Gasoline ......
Summary of Uncontrolled Emission Factor for the
Transfer of Benzene
Estimated 1977 EDC Emissions as Solid Waste and to
Water from EDC Production
46
47
49
51
54
56
58
59
63
65
vi
-------
ACKNOWLEDGEMENTS
It is a pleasure to acknowledge the cooperation and guidance given
by several individuals of the U.S. Environmental Portection Agency,
Office of Air Quality Planning and Standards. Ren Greer, Strategies and
Air Standards Division, and Dr. George H. Wahl, Jr., EPA Consultant,
provided direction throughout the study. David Mascone of the Emission
Standards and Engineering Division gave valuable assistance in regard to
control technology and emission factors and George Schewe of NOAA
provided input on atmospheric dispersion modeling.
Mr. Casey Cogswell, SRI International, Chemical Industries Center,
generously provided information and guidance concerning the
manufacturing and uses of ethylene dichloride.
vii
-------
I INTRODUCTION
This report is one in a series that SRI International is providing
for the U.S. Environmental Protection Agency (EPA) to estimate
populations at risk to selected pollutants. Primarily, this study has
sought to estimate the environmental exposure of the U.S. populaton to
atmospheric ethylene dichloride (EDC) emissions. The principal
atmospheric sources we consider in this report are facilities at which
EOC is produced or used as a chemical intermediate and gasoline that
contains EDC as a lead scavenger. Possible exposures from
transportation of EDC, disposal of EDC wastes, and other minor uses are
also described.
-------
II SUMMARY
EDC is one of Che highest volume chemicals used in the United
States, with approximately 5 million metric tons (mt) synthesized during
1977. More than 80% of the EDC produced is used in the synthesis of
vinyl chloride monomer (VCM). The majority of the remaining production
is used in the synthesis of 1,1,1-trichloroethane (1,1,1-TCE or methyl
chloroform), trichloroethylene (TCE), perchloroethylene (PCE),
vinylidine chloride (VDCM), and ethyleneamines (EA); EDC is also
employed directly as a lead scavenger for gasoline.
EDC is a colorless, oily liquid that has a sweet taste, a
chloroform-like odor, and a volatility similar to that of gasoline. It
boils at 83.5 C, melts at -35.4 C, and has a specific gravity of
1.2. It is relatively stable in water, but evaporates rapidly from
water to the atmosphere where it is destroyed by photooxidation.
Estimates of EDC's half-life in the atmosphere range from weeks to
months, a period sufficiently long for aerial transport to play a major
role in its distribution but relatively short for it to accumulate in
>
the terrestrial and aquatic environment. It has a potential for
bioaccumulation; however, no firm evidence now exists to support
bioaccumulation in the marine environment or in other biota.
EDC does not occur naturally in the environment. Environmental
exposures occur mainly from EDC lost during production, from EDC used as
a chemical intermediate in producing other chemicals, or in its use in
gasoline as a lead scavenger. Minor environmental exposures may occur
through dispersive uses of EDC such as in grain fumigants, paints,
coatings, adhesives, cleaning, and in the preparation of polysulfide
compounds. Additional environmental exposures may occur from spills and
venting during EDC transportation and from evaporation resulting from
waste disposal.
-------
Monitoring data for occupational exposures to EDC have been report-
ed in the literature for more than four decades; however, the environ-
mental monitoring data that have been collected date only from around
1975 and they are limited. A number of environmental monitoring studies
have failed to find atmospheric EDC in the general U.S. environment at
the ppt detection level (Grimsrud and Rasmussen, 1975; Singh et al.,
1977; and Hanst, 1978). PEDCo (1978) found EDC at the ppb level in the
atmosphere surrounding three EDC production facilities. (The maximum
integrated 24-hr value was 180 ppb.) Pellizzari (1978) reported the
detection of EDC concentrations of less than 55 ppb near a chemical dis-
posal site in New Jersey.
The current Occupational Safety and Health Administration (OSHA)
standard for occupational exposure to EDC is 50 ppm (8-hr time-weighted
average). In March 1976, the National institute of Occupational Safety
and Health- recommended an exposure limit of 5 ppm (time-weighted averge
for a 10-hr workday or less, a 40-hr workweek). These levels, however,
were designed to protect against toxic effects other than cancer and may
not provide adequate protection from potential carcinogenic effects
(NIOSH, 1978).
Human population exposures to atmospheric EDC have been estimated
for emissions resulting from its production, its use as a feedstock in
the production of other chemicals, and its use as a lead scavenger in
automobile gasoline. Other potential exposure routes such as emissions
resulting from transportation and emissions from other product uses have
also been described. The population exposure estimates given in Table
II-l are based on the calculated atmospheric emissions given in Table
II-2.
These emission and exposure estimates have necessitated reliance on
very limited data. Because of the paucity of measured atmospheric EDC
data, it was necessary to approximate concentrations through the use of
dispersion modeling. Moreover, the resulting estimates are subject to
considerable uncertainty in regard- to: (1) the quantity of EDC emis-
sions, (2) EDC production and consumption levels, (3) certain source
3
-------
Table II-l
SUMMA&Y OF ESTIMATED POPULATION EXPOSURES TO ATMOSPHERIC EDC FROM SPECIFIC EMISSIONS SOURCES
Annual Average
EDC Concentra-
tion (ppb)a
10
6.00-10.00
3.00- 5.99
1.00- 2.99
0.60- 0.99
0.30- 0.59 1
0.10- 0.29 4
0.000-0.099 1
*" 0.030-0.059 3
0.010-0.029
Total 12
Production Facilities'1 Gasoline
1,1,1- . Lead Service Automobile
EDC VCM TCE TCE PCE EA VDCM Scavenger Stations0 Emissions'1
1,700
3,300
28,000
280,000
400,000 1,300
,500,000 360 70
,300,000f 30,000 1,700 390 80 17,000 270 1,900
,900,000f 42,000 16,000 10,000 500 8,000 3,400 3,400
,SOO,OOOf 260,000 83,000 47,000 17,000 43,000 34,000 25,000
550,000£ 940,000 170,000 140,000 250,000 37,000 90,000 350,000 1,000,000 13,000,000
,500,000 1,300,000 260,000 200,000 270,000 110,000 130,000 380,000 1,000,000 13,000,000
Automobile
Refueling6
( 8 >
30,000,000
• To convert to jtg/m^, multiply each exposure level by 4.1.
b Production facilities that either produce EDC or use EDC as a feedstock in the production of another chemical.
c These are exposures to people who reside near gasoline service stations.
d These are exposures from evaporative emissions from pre-1975 automobiles.
e These are exposures to people while refueling their automobiles at self-service gasoline stations.
These are underestimates because the dispersion modeling results were not extrapolated beyond 30 km from each EDC production facility.
There are additional people who are exposed to EDC concentrations of 0.01-0.1 ppb at distances greater than 30 km from the larger producton
facilities.
8 Estimated aa 30 million people exposed to an EDC concentration of 1.5 ppb for 2.2 hr/yr. The annual average time-weighted exposure is
0.0004 ppb.
-------
Table II-2
ESTIMATED ATMOSPHERIC EMISSONS OF EDC FOR 1977
Emissions (1,000 mt/yr)
EDC production
Fugitive 5.2
Storage 14.5
Direct chlorination 6.3
Oxychlorination 17.9
Subtotal 43.9
Production using EDC as
Feedstock
VCM 1.1
1,1,1-TCE 0.4
TCE 0.2
PCE 0.3
EA 0.3
VDCM 0.2
Lead scavenger 0.2
Subtotal 2.5
Automobile gasoline
Service stations 0.1
Auto emissions 1.2
Subtotal 1.3
Other
Dispersive uses 5.0
Transporation3 . -.-
Waste disposal3 -.—
Total 52.7
aNot included. Rough order estimates place these emissions as much
less than 2,400 mt/yr for transportation and much less than 29,100 mt/yr
for waste disposal.
-------
locations, (4) control technologies employed, (5) deterioration in con-
trol technologies over time, (6) physical characteristics of EDO
sources, (e.g., stack height), (7) details on atmospheric dispersion and
degradation, and (8) living patterns of the exposed population. Given
these complex and variable factors, the accuracy of the estimates could
not be assessed. Nevertheless, the estimates are believed to be a
reasonable approximation of actual conditions. Comparisons of atmo-
spheric monitoring and modeling data for EDO concentrations near 3 sites
used in this report shows agreement within 70% and averaging 25%.
During 1977, 18 facilities produced an estimated 5 million mt of
EDC. It is estimated that approximately 12.5 million people are exposed
to average annual EDC concentrations of 0.01 to more than 10 ppb from
this production. Estimates of exposures to concentrations of less than
0.1 ppb from production facilities are underestimates because the dis-
persion modeling results were not extrapolated beyond 30 km from the
plants. There are additional people who are expected to be exposed to
EDC concentrations of 0.01-0.1 ppb at distances of greater than 30 km
from the larger production-facilities. However, it is generally assumed
that disperions modeling results are unreliable beyond 20 to 30 km from
the source.
Estimates are given for exposures to EDC used as a feedstock in the
production of VCM, 1,1,1-TCE, ICE, PCE, EA, VDCM, and gasoline lead
scavengers. Many of these chemicals are produced at the same facilities
that produce the EDC feedstock. Of the 28 producton plants involved, 18
also produce EDC. In 1977, approximately 5 million mt of EDC was re-
quired, with more than 80% used in producing VCM. More than 2 million
people are exposed to annual average EDC concentratons of 0.01 to 1.0
ppb from these operations.
Leaded gasoline additives contain EDC as a lead scavenger.
Although the EDC is expected to be destroyed during combustion, evapora-
tive emissions occur during refueling operations and from the gas tanks
and carburetors of automobiles. These emissions are expected to de-
crease as newer model automobiles replace the pre-1975 models. It has
6
-------
been estimated that approximately 30 million people are exposed to EDC
concentrations of 1.5 ppb for 2.2 hr/yr while refueling their auto-
mobiles at self-service stations. Similarly, approximately 1 million
people residing near gasoline service stations are exposed to average
annual EDC concentrations of 0.01 to 0.03 ppb from refueling losses.
Another 13 million are exposed to annual average EDC concentrations of
0.01 to 0.03 ppb from automobile evaporative emissions.
-------
Ill CHEMICAL AND PHYSICAL PROPERTIES OF EDC
AND ITS ENVIRONMENTAL BEHAVIOR*
Introduction
The Chemical Abstracts Service registry number of EDC is 000107062;
the NIOSH number is K005250. To minimize confusion between EDC
(C2H4C12) and cis and trans dichloroethylene (C2H2C12),
Drury and Mammons (1978) recommend that EDC be referred to as
1,2-dichlorethane in place of ethylene dichloride. Many synonyms and
trade names are also used: Brocide; Destrucol Borer-Sol;
Di-chlor-mulsion; sym-dichloroethane; alpha, beta-dichloroethane; di-
chloroethylene; Dutch liquid; EDC; ENT 1,656; ethane dichloride;
ethylene chloride; glycol dichloride; and oil of the Dutch chemists
(NOISH, 1977; Mitten et al., 1970).
The composition and structure of 1,2-dichloroethane (EDC) are in-
dicated by the molecular formula, C H,CL9, and the line diagram,
H H
I I
Cl - C - C - Cl
.1 I
H H
Physical Properties
EDC is a colorless, oily liquid that has a sweet taste and a
chloroform-like odor (Hawley, 1977). It is volatile and evaporates at a
rate 0.788 time that of carbon tetrachloride or gasoline (Whitney,
*The discussion given here has been summarized from a draft report by
Drury and Hammons (1978).
-------
1961). Air saturated with EDC contains 350 g/m at 20°C and
537g/m at 30 C. EDC is completely miscible with ethanol, chloro-
form, ethyl ether, and octanol (Windholtz, 1976; Johns, 1976). The par-
tition coefficient, log P, of EDC between octanol and water is 1.48
(Radding et al., 1977), reflecting preferential solubility in organic
media.
Vaporized EDC solvent is readily ignited—the closed cup flash
point being only 13°C. The liquid is also flammable, burning with a
smoky flame, but the ignition temperature is high, 413°C. Under a
pressure of 1 atm, EDC steam distills at 71.9°C. The binary azeotrope
contains 19.5% water; 14 other binary azeotropes are known (Mitten et
al, 1970). A ternary azeotrope containing 78% 1,2-dichloroethane, 17%
ethanol, and 5% water boils at 66.7°C. Other properties are given in
Table III-l. .
Chemical Properties
EDC is stable at ambient temperatures but slowly decomposes in the
presence of air, moisture, and light. During decomposition the liquid
EDC becomes darker in color and progressively acidic. It can corrode
iron or steel containers, but these deleterious reactions are completely
inhibited by small concentrations of alkyl amines (Bardie, 1964).
Both chlorine atoms in EDC are reactive and can be replaced by
other substituents. This bifunctional nature of EDC makes it useful in
the manufacture of condenstion polymers (Rothon, 1972). Hydrolysis of
EDC, with slightly acidic 160°C to 175°C water at 15 atm, or with
140°C to 250°C aqueous alkali at 40 atm, yields ethylene glycol; at
120°C, the addition of ammonia under pressure to EDC yields ethylene-
diamine. 1,1,2-TCE and other higher chloroethanes are formed by chlor-
inating EDC at 50°C in light from a mercury vapor lamp. EDC reacts
with sodium polysulfide to form polyethylene tetrasulfide, and with
oleum Co give 2-chloroethylsulfuryl chloride. With Priedel-Crafts ca-
talysis, both chlorine atoms in EDC can be replaced with aromatic ring
compound (Bardie, 1964).
-------
Table III-l
PHYSICAL PROPERTIES OF EDO
Molecular weight 98.96
Density, g/ml at 20°C 1.2351
Melting point, °C -35.36
Boiling point, °C 83.47
Index of refraction, 20°C 1.4448
Vapor pressure, torr, at °c
-44.5 1
-13.6 10
10.0 40
29.4 100
64.0 400
82.4 760
Solubility in water, ppm w/w at °c
20 8,690
30 9,200
Biochemical oxygen demand (5 days), % 0
Theoretical oxygen demand, mg/mg 0.97
Measured chemical oxygen demand, mg/mg • 1.025
Vapor density (air =1) 3.42
Flash point, open cup, °C 13.0
Ignition temperature, °c 413.0
Explosive limit, % volume in air
Lower 6.2
Upper 15.9
Specific resistivity 9.0 x 106
Viscosity, cP, at 20°C 0.840
Dielectric constant,€ - 10.45
Surface tension, dyne/cm 33.23
Coefficient of cubical expansion, 10°C-30°C 0.0016
Latent heat of fusion, cal/g 21.12
Latent heat of vaporization, cal/g, at boiling point 77.3
Specific heat, cal/g °C
Liquid at 20°C 0.308
Vapor, 1 atm at 97.1°C 0.255
Critical temperature, °C 288
Critical pressure, atm 53
Critical density, g/cm^ 0.44
Thermal conductivity, Btu/hr-ft2 at 20°C 0.825
Heat of combustion, cP, kcal/g-mole 296.36
Dipole moment, ESU 1.57 x 10~18
Conversion factors, 25°C, 760 torr 1 mg/L* = 1. g/nr* = 247 ppm
1 ppm =4.05 gm/m3 =4.05 g/L
Source: Draft report by Drury and-Hammons (1978).
10
-------
Environmental Behavior
Bioaccumulation and Biomagnification
EDO's physical and chemical properties exhibit opposing tendencies
with respect to bioaccumulation, but high vapor pressure and low latent
heat of vaporization argue that the compound is exhaled from the lungs
in the same condition in which it was inhaled. In fact, no firm evi-
dence exists for the bioaccumulation of EDC in food chains under envi-
ronmental conditions (Radding et al., 1977). Pearson and McConnell
(1975) in searching for simple aliphatic chlorocarbons in several tro-
phic levels of the marine environment near the industrialized area of
Liverpool, found no evidence of EDC. In laboratory studies on oysters
and fish using EDC labeled with carbon -14, Pearson and McConnell did
see rapid storage of the chlorinated hydrocarbon up to an'asymptotic
level, but this accumulation was followed by loss of EDC on transfer of-
the organisms to clean sea water. Parallel analyses by chromatographic
techniques showed reduced levels of EDC in the organisms, indicating
that metabolism of the compound occurred in the tissues of both fish and
oysters.
Biological Degradation
The conclusions of the few literature references to microbial de-
gradation of simple chlorinated hydrocarbon compounds conflict. Some
authors report these compounds are not metabolized either by aerobic or
anaerobic microorganisms (Pearson.and McConnell, 1975). Other micro-
biologists believe biodegradation can occur via co-metabolic processes
(Horvath, 1972), but no evidence supporting biodegradation of EDC has
been found. There is general agreement, however, that mammals metabo-
lize these compounds, producing chlorinated acetic acids either directly
or via chloroethanols. All of the resulting chlorinated acetic acids
are susceptible to further degradation by microorganisms in sea water
(McConnell et al., 1975).
11
-------
Chemical Degradation
Photooxidative reactions involving atmospheric EDC probably result
in monochloroacetyl chloride, hydrogen chloride, and monochloroacetic
acid (Spense and Hanst, 1978). Alcohols, ketones, alkyl nitrates, and
cleavage produces arising from intermediate alkoxy radicals are also
possible products. Preliminary data indicate the half-life of EDC in
the atmosphere may be about 3 to 4 months (Pearson and McConnell, 1975;
EPA, 1975). Based on an average HO radical concentration of
0.8 x lO'^M, Radding et al. (1977) estimated a combined
oxidativephotolysis half-life of 234 hr. The recent calculations of
Altshuller and the recent experiments of Snelson et al. (1978)
indicate tropos.phericTifeT1mes"bf ~EDC of approximately 0.75 to 1
year. Although the half-life remains to be determined definitively,
available estimates make it clear that the lifetime of EDC in the
troposphere, although short in an absolute sense, is sufficiently long
for aerial transport to play a major role in its distribution.
EDC is resistant to hydrolysis. Radding et al. (1977) estimated a
hydrolysis half-life of approximately 50,000 yr. This estimte is much
longer than the 6- to 18-month half-lives observed for similar, but not
identical, compounds subjected to a combination of hydrolysis, oxida-
tion, and photolysis (Dilling et al., 1975); nevertheless., it appears
that hydrolysis of EDC is slow compared to other pertinent environmental
processes, such as volatilization or photolysis.
Dilling et al. (1975) and McConnell et al. (1975) studied the re-
moval of compounds similar to EDC from water by adsorption on several
common substrates. Dilling et al. observed little or no adsorption of
chlorinated hydrocarbons on clay, limestone, sand, and peat moss in lab-
oratory experiemtns that'involved aqueous solutions containing 1 ppm
organic contaminant. McConnell et al. reached similar conclusions about
adsorption of chlorinated hydrocarbons from seawater by coarse gravels,
but they found relatively high adsorption by Liverpool Bay sediments
rich in organic detritus. The divergent conclusions of these studies
probably reflect different experimental conditions: The hydrocarbon
12
-------
concentrations in the experiments of Billing et al. were well below the
solubility limits of the various compounds used, and adsorption under
these conditions is less likely than in Liverpool Bay, which receives
large volumes of industrial and domestic effluents.
The relatively high vapor pressure of EDC causes rapid volatil-
ization of the hydrocarbon from aqueous effluents. After 96 min at am-
bient temperature, about 90% of the EDO initially present in water at a
concentration of 1 ppm evaporated (Billing et al., 1975). This rate
corresponds to a vaporization half-life of 29 min. Comparison of these
data with those for other environmental removal processes indicate that
volatilization is the chief process for removal of EDC from water.
Persistence .
EDC has a long hydrolysis half-life, a short vaporization half-life
from water, and a relatively short photooxidative half-life in the atmo-
sphere. It is unlikely to accumulate in the environment. Note, how-
ever, that one of EDC's photooxidative products is chloroacetyl
chloride, which may be sufficiently stable to reach the stratosphere and
interact destructively with the ozone layer.
Environmental Transport
Because the vapor pressure, of EDC is moderately high, most emis-
sions from manufacturing operations occur as vapors that are vented
directly to the atmosphere. Even when initially present in wastewater
or solid waste products, EDC tends to transfer rapidly to the atmo-
sphere. This volatility, coupled with an atmospheric half-life suf-
ficiently long for aerial transport, results in most distribution of EDC
in the environment occurring by aerial transport (HcConnell et al.,
1975; Pearson and McConnell, 1975). Some transfer of EDC from air to
water also occurs, particularly as a result of rainfall. This effect is
assumed to be minor when compared to aerial transport, but quantitative
data comparing these transport routes are lacking.
13
-------
IV EDC PRODUCTION AND USES
Production
The annual EDC production capacity for U.S. plants is approximately
7.3 million mt. From 1973 to 1977 the industry operated at about 60-70%
of capacity, producing 4.6, 4.7, 3.7, 5.0, and 5.2 million mt, respec-
tively. The overall production during these years may have been even
higher than indicated because captive production is not always adequate-
ly recorded in published data. Future growth of the market is expected
to average 4% to 5%/yr through 1981, at which time the demand for EDC is
expected to be 6.6 million mt. Five of the major producing, companies
are currently expanding production facilities or are planning increased
production in the near future (Chemical Marketing Reporter, 1977).
Uaes
EDC is used primarily as a raw material in the synthesis of other
chemicals, in particular for VCM, 1,1,1-TCE, TCE, PCE, VDCM, EA, and as
a lead scavenger for gasoline. Primary uses of these compounds are as
follows:
VCM Its major use is in the production of
PVC and its copolymer resins. Small
' amounts are used in polyvinylidene
chloride and other copolymers.
1,1,1-TCE Its major use is for solvent clean-
ing. Minor uses include aerosol pro-
pellant, solvent in adhesives and
coating formulations, drain cleaner,
and fabric spotting fluid.
TCE It is almost entirely used as a
metal-cleaning solvent.
14
-------
PCE Its major uses are for metal cleaning
and dry cleaning.
EA Its major uses are as a chelating
agent and carhamate fungicide. Other
uses include detergents and softening
agents, specialty resins, epoxy hard-
eners, and corrosion inhibitors.
VDCM It is used mainly in the production
of polyvinylidene copolymers
Lead Scavenger It is used in gasoline antiknock mix-
tures.
The quantities.of EDC consumed for these and other uses are shown in
Table IV-1. More than 80% of the EDC produced is used in the manufac-
ture of VCM. Each of the other compounds listed above requires 2% to 3%
of the total EDC produced. Exports account for about 3.4% of the EDC
produced, and other minor products require less than 0.2% of EDC produc-
tion.
Auerback Associates (1978) estimated EDC consumption for other
minor uses in 1977 at about 5,000 mt. Of this- subtotal, about 28% was .
used in the manufacture of paints, coatings, and adhesives.. Extracting
oil from seeds, treating animal fats, and processing pharmaceutical pro-
ducts required 23% of the subtotal. An additional 19% was consumed in
cleaning textile products and polyvinyl chloride manufacturing equip-
ment. Nearly 11% was used in the preparation of polysulfide compounds.
Grain fumigation required about 10%. The remaining 9% was used as a
carrier for amines in leaching copper ores, in the manufacture of color
film, as a diluent for pesticides and herbicides, and for other
miscellaneous purposes.
EDC Producers and Users
Table IV-2 lists the major EDC producers and consumers along with
their estimated January 1979 installed production capacity. As the
15
-------
Table IV-1
EDC CONSUMPTION
(Thousands of metric* tons per year)
YEAR
Use
VCM
1,1,1-TCE
TCE
PCE
EA
VDCM
Lead scavenger
Other
Net exports
Total
1973
3,645
184
141
104
128
83
106
b
167
4,558
1974
3,871
198
121
98
132
92
97
b
(133)
4,742
1975
3,015
155
92
89
123
83
80
b
(26)
3,663
1976
4,079
213
99
89
132
88
93
b
(199)
4,992
1977
4,300
215
83
87
136
97
89
b
177
5,194
19823
5,635-
6,140
260-280
85-110
87-95
113-119
125-135
39
b
180
6,524-
7,098
Source: SRI estimate.
^Projected consumption.
Bother uses, which are not included in consumption, in 1974 were estimated
at 7,000 rat and at 5,000 mt in 1977.
16
-------
table indicates, most EDO producers have the capacity to use most of the
EDO they produce as feedstock for other products within their own
plants. In fact, in recent years only a small fraction (10% to 15%) of
the total production of EDC has been sold on the open market (US. Inter-
national Trade Commission, 1973-1977).
EDC is produced by the "balanced process." This process involves a
combination of direct chlorination of ethylene and oxychlorination of
ethylene using hydrogen chloride, which in turn is produced in the
cracking of EDC to VCM. The EDC manufactured by oxychlorination of
ethylene is generally used captively as an intermediate in VCM pro-
duction. Table IV-3 shows the percentages of EDC produced by direct
chlorination and by oxychlorination, by producer.
Chemical producers rarely operate at maximum production capacity
for a specific chemical. Table IV-4 shows the percentage of production
capacity employed in 1977 to produce EDC and the major chemicals in
which it is used as a feedstock.
17
-------
oo
Producer
Location
—fable IV-2
EDC PRODUCERS AND MAJOR CONSUMERS
(January 1, 1979, production capacities in thousands of metric tona)
Canacitv VCM 1.1.1-TCE TCE PCE BA VDCM
Borden Chemical
Conoco Chemical
Diamond Shamrock
Diamond Shamrock
Dow Chemical
Dow Chemical
Dow Chemical
Dow Chemical
duPont
duPont
duPont
Ethyl Corp.
Ethyl Corp.
B. F. Goodrich
Houston Chemical
ICI America*
Monochem
Nalco Chemical
PPG Industries
PPG Industries
Shell Chemical
Shell Chemical
Stauffer Chemical
Stauffer Chemical
Union Carbide
Union Carbide
Vulcan Chemical
Vulcan Chemical
Ceismar, LA
Lake Charles, LA
Deer Park, TX
La Porte, TX
Freeport, TX
Oyater Creek, TX
Pittaburg, CA
Plaquemine, LA
Antioch, CA
Corpus Christi, TX
Deepwater, NJ
Baton Rouge, LA
Houston, TX
Calvert City, KY
Beaumont, TX
Baton Rouge, LA
Geismar, LA
Freeport, TX
Lake Charles, TX
Cuayanilla, PR
Deer Park, TX
Nor co, LA
Caraon, CA
Louisville KY
Taft, LA
Texas City, TX
Geismar, LA
Wichita, KS
Total
Source: SRI estimates.
7,316
6,218
a Plant waa purchased from Allied Chemical in September 1978.
b Proceaa doea not uae EDC aa a feedstock.
c Rough order eatimatea.
Scavenger
524
145
719
726
499
953
318
118
454
318
544
379
635
544
154
68
68
150
224
525
17 45
749
150 167 51 Ob 60 45C
525 Ob
ob
936 112 45C
20C
ob
20C
248 15 14 20C
20C
749
15C
224
ob
5C
229 130 68 54 30°
375
629
525
130
Ob
70
60
Ob ' 41
Ob
409
151
154
190 120
100
-------
Table IV-3
1977 EDO PRODUCTION BY DIRECT CHLORINATION AND OXYCHLORINATION
Direct
Chi or i nation
Oxy-
chlorination
Producers
Conoco Chemical
Diamond Shamrock
Dow Chemical
Dow Chemical
Dow Chemical
Ethyl Corporation
Ethyl Corporation
B. F. Goodrich
ICI America3
PPG Industries
PPG Industries
Shell Chemical
Shell. Chemical
Stauffer Chemical
Union Carbide
Union Carbide
Vulcan Chemical
Locations
Lake Charles, LA
Deer Park, TX
Freeport, TX
Oyster Creek, TX
Plaquemine, LA
Baton Rouge, LA
Houston, TX
Calvert City, KY
Baton Rouge, LA
Lake Charles, LA
Guayanilla, PR
Deer Park, TX
Norco, LA
Long Beach, CA
Taft, LA
Texas City, TX
Geismar, LA
49.2
35.8
57.1
(b)
51.7
52.7
100.0
33.3
66.7
77.2
(b)
66.2
(b)
69.2
100.0
100.0
0.0
50.8
64.2
42.9
(b)
48
47,
0.0
66.7
33.3
22.8
(b)
33.8
(b)
30.8
0.0
0.0
100.0
Source: Draft report by Drury and Hammons (1978).
aPlant was purchased from Allied Chemical in September 1978.
bNot available.
19
-------
Table IV-4
1977 USE OF EDC PRODUCTION CAPACITIES
(Thousands of metric tons)
Product
EDC
VCM
1,1,1-TCE
TCE
PCE
EA
VDCM
Lead scavenger
EDC
Capacity5
7,316
6,218
409
151
154
190
120
100
EDC used
in 1977
Production^
5,194
4,300
215
93
87
136
97
89
Percent
Capacity Used
71.0
69.2
52.6
61.6
56,
71
.5
.6
80.8
89.0
aSee Table IV-2. This is the amount that would be used annually if the
product was produced at 100% of capacity.
bSee Table IV-1.
20
-------
V POPULATION EXPOSURES FROM EDC PRODUCTION
General
As was shown in Table IV-2, most of the EDC produced is used as
feedstock in the production of other chemicals, particularly VCM. the
majority of both the EDC produced and the chemicals that use EDC as
feedstock are made at the same facilities. Thus, people residing near
these production facilities can be exposed to atmospheric EDC from
several types of production. Section VI sets forth the exposure from
chemical production facilities that use EDC as a feedstock.
Sources of Emission
EDC producers and their individual capacities are listed in Table
IV-2. The total annual capacity of the 18 plants listed is 7.3 million
mt. Table IV-4 indicates that approximately 71% of the production capa-
city was used during 1977. Because production data for each plant are
unavailable, we have assumed that each operates at 71% of capacity.
Emissions
Four principal sources of emissions have been identified: direct
chlorination vent stack, oxychlorination vent stack, fugitive emissions,
and emissions from tank storage.
As part of the Synthetic Organic Chemical Manufacturing Industry
study under way in EPA's Emission Standards and Engineering Division of
the Office of Air Quality Planning and Standards and as a result of the
detailed study of the VCM industry, a significant amount of engineering
data on the EDC industry are available. Table V-l summarizes the data
collected for oxychlorination vent emissions for 10 .Production plants:
We applied the average vent emission factor of those plants (1.0%) to
plants for which no emission data are available. We estimated an emis-
sion factor of 0.22% for the direct chlorination process vent emissions,
with fugitive emissions estimated at 0.1% of plant
21
-------
Table V-l
EDC OXYCHLORINATION VENT EMISSIONS
Plant and Location
Oxychlorination
Production*
(g/s)
Conoco, Lake Charles, LA 5,987
Diamond, Deer Park, TX 2,871
Ethyl, Baton Rouge, LA 3,392
Goodrich, Calvert
City, KY 6,819
ICI America, Baton
Rouge, LA 2,366
PPG, Lake Charles, LA 2,800
Shell, Deer Park, TX 4,464
Stauffer, Long Beach, CA 1,071
Vulcan, Geismar, LA 3,408
Dow, Oyster Creek, TX 5,609
EDC
Emissions'3
(8/s)
12 3
2.7
43.6
26.4
70.1
0.0
25.2
19.6c
81.6c
0.0
Emission Factor
(g emission/
g production)
0.0021
0.0009
0.0129
0.0039
0.0296
0.0000
0.0056
0.0183
0.0239
0.0000
Average
0.0097
aTotal capacity given in Table IV-2 times 71% use times percent
oxychlorination production given in Table IV-3.
bSource: EPA (1978).
cBased on an EPA engineering estimate.
22
-------
production. Emissions for storage tanks at production facilities were
estimated as 2.8 g/yr/kg of annual capacity (Mascone, 1978). Table V-2
gives the estimated emissions resulting from emission factors. Total
EDO atmospheric emissions are estimated as 43.9 thousand mt/yr (1,312
g/s) or about 0.8% of the amount produced. The storage and direct
chlorination emission rates are for uncontrolled plants. The industry
now has some controls on these two emission points but data are insuf-
ficient to estimate the present degree of control.
Atmospheric Concentrations
Atmospheric monitoring data have been collected from three loca-
tions that have EDC production facilities (PEDCo, 1978). These threie
locations are (1) near the B. F. Goodrich plant in Calvert City, Ken-
tucky, (2) near the Conoco plant in Lake Charles, Louisiana, and (3)
near the Shell plant at Norco, Louisiana and the Union Carbine plant at
Taft, Louisiana. The Goodrich, Conoco, and Shell plants each have an-
nual EDC production capacities of approximately 500,000 mt. The Union
Carbide plant has an annual capacity of approximately 70,000 mt.
Twelve monitoring stations were positioned around each location.
Data were recorded for 10 days in New Orleans, 12 days in Lake Charles,
and 13 days in Calvert City. The preliminary results of the monitoring
data are summarized in Tables V-3 through V-5. Average 12- to 13-day
atmospheric concentrations ranged from 0 to 5 ppb for the Calvert City
stations, 1 to 43 ppb for the Lake Charles stations, and 0.1 to 12.1 ppb
for the New Orleans stations. Individual 24-hr concentrations were much
higher. Generally, the concentrations for locations near the plants in
the Lake Charles area were almost 10 times those for the Calvert City
and New Orleans areas. The differences may be attributable to meteor-
ological conditions, plant production at sampling times, emission con-
trols, positioning of the monitoring stations with respect to plant
location and wind direction, or other EDC sources in the areas.
_The_monitoring_data—show—chat—elevated—EDG—concentrations—ex-i-»t—i-n-
the vicinity of at least 3 EDC production facilities; however, the data
23
-------
Table V-2
ESTIMATED ATMOSPHERIC EMISSIONS'FROM EDC PRODUCTION FACTILITIES
Production3
Emisaiona (g/s)
Plant
Conoco
Diamond
Diamond
Dow
Dow
Dow
Ethyl
Ethyl
Goodrich
lCI America
PPG
PPG
Shell
Shell
Btauffer
Union Carbide
Union Carbide
Vulcan
Location
Lake Charles, LA
Deer Park TX
La Porte, TX
Freeport, TX
Oyster Creek, TX
Plaquemine, La
Baton Rouge, LA
Houston, TX
Calvert City, KY
Baton Rouge, LA
Lake Charles, LA
Guayanilla, PR
Deer Park, TX
Norco, LA
Carson, CA
Taft, LA
Texas City, TX
Geisoar, LA
103 ot/yr
372
103
510
515
354
678
226
84
322
226
386
269
451
386
109
48
48
107
(g/s)
11.800
3,265
16,190
16,345
11,235
21,455
7,160
2,660
10,220
7 , 160
12,250
8,533
14,295
12,250
3,470
1,530
1,530
3,380
Fugitive
11.8
3.3
16.2
16.3
11.2
21.5
7.2
2.7
10.2
7.2,
12.3
8.5
14.3
12.3
3.5
1.5
1.5
3.4
Storage
33.0
9.1
45.3
45.7
31.4
60.1
20.1
7.5
28.6
20.1
34.3
23.9
40.0
34.3
9.7
*-3
4.3
9.5
Direct
12.8
2.6
17.8
20.5
12.4
24.4
8.3
5.9
7.5
10.5
20.8
9.4
20.8
13.5
5.3
3.4
3.4
0.0
Oxychlorination
12.6
1.9
15.8
67.3
0.0
99.5
43.7
0.0
26.6
70.6
0.0
41.4
27.1
59.4
19.6
0.0
0.0
80.8
Total
70.2
16.9
95.1
149.8
55.0
205.5
79.3
16.1
72.9
108.4
67.4
83.2
102.2
119.5
38.1
9.2
9.2
93.7
local
5,194
164.9
461.2
199.3
566.3
1,391.7
Source: SRI estimates.
Assumed to be 71Z of production capacity.
-------
Table V-3
ATMOSPHERIC EDC MONITORING DATA3 FOR GALVERT CITY, KENTUCKY
Site
No.
1
2
3
4
5
6
7
8
9
10
11
12
Relation to
Goodrich Plant
0.8
1.8
1.7
2.0
3.3
2.9
2.5
3.4
2.3
2.8
2.3
3.0
km SE
km SW
km SSW
km SSE
km SE
km E
.km ENE
km NE
km NE
km N
km NNW
km NW
Average
(ppb)
2.0
2.3 .
,1
.7
0.
0.
0.2
0.0
1.2
1.5
5.1
3.6
2.3
0.6
Average
(ue/m3)
8.0
9.3
0.5
2.8
0.6
0.1
4.8
6.2
20.6
14.6
9.4
2.3
Rangeb
(yg/m3)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
-37.7
- 72.2
- 4.1
- 18.0
- 3.2
- 0.5
- 36.3
- 22.4
- 67.8
- 59.9
- 55.0
-28.7
Source: Based on draft data supplied by PEDCo (1978).
Observations are for thirteen 24-hr periods between August 27, 1978, and
September 18, 1978.
''When duplicate quality control samples were taken at one site, the average
of the two samples has been used.
25
-------
Table V-4
ATMOSPHERIC EDC MONITORING DATA* FOR LAKE CHARLES, LOUISIANA
Relation to
Conoco Plant
1.0
km S
km WNW
2 km WNW
7 km W
9 km SW
km WSW
3.0 km NW
2.8 km NNW
0 km NNW
5 km NNW
7 km NE
0.7
1.
0.
0.
1.3
2.
1.
0.
Average
(ppb)
26
61
1.8 km ESE
5.0
35.4
40.2
11.2
1.1
1.0
1.6
1.7
20.1
12.3
Average
(pg/m )
106.
248,
20,
143.4
162.7
45.4
4.5
4.0
6.5
6.7
81.4
49.9
Rang
1.4 -
6.0 -
0.0 -
1.8 -
0.5 -
0.0 -
0.0 -
0.0 -
0.0 -
0.0 -
0.0 -
0.5 -
**
e
)
269.5
651.7
67.2
744.8
383.3
171.6
27.3
32.8
30.2
36.2
581.6
497.8
Source: Based on draft data supplied by PEDCo (1978).
Observations are for twelve 24-hr periods between September 24, 1978, and
October 5, 1978.
bWhen duplicate quality control samples were taken at one site, the average
of the two samples has been used.
26
-------
Table V-5
ATMOSPHERIC EDC MONITORING DATA3 FOR NEW ORLEANS, LOUISIANA
Site
No.
1
2
3
4
5
6
7
8
9
10
11
12
Relation to
Shell
4.0 km NNW
4.0 km NW
3.0 km WNW
0.4 km SW
1.0 km NE
6.0 km WSW
3.0 km SW
2.0 km SW
1.5 km S
2.0 km SE
3.0 km SSE
14.0 km S
Union Carbide
6.0 km NNE
4.0 km NNW
3.0 km NNW
2.0 km NE
4.0 km NE
4.0 km NW
0.8 km NNW
1.5 km NE
2.0 km WNW
3.0 km WNW
3.0 km WSW
1.2.0 km S
Average
(ppb)
0.1
0.4
0.4
12.0
0.5
0.9
1.5
2.3
1.4
0.8
0.5
0.6
Average
(Mg/m3)
0.4
1.7
1.7
48.5
2.0
3.8
5.9
9.4
5.6
3.1
1.9
2.5
Rangeb
0.0- 1.2
0.0- 6.2
0.0- 5.8
0.6-169.0
0.5- 9.1
0.0- 20.7
0.5- 24.3
0.5- 29.1
0.0- 17.6
0.0- 13.1
0.0- 8.7
0.0- 6.4
Source: Based on draft data supplied by PEDCo (1978).
Observations are for ten 24-hr periods between October 10, 1978, and
October 17, 1978.
bwhen duplicate quality control samples were taken at one site, the average
of the two samples has been used.
27
-------
available are insufficient for estimating population exposure for all
•EDC producers. It is, therefore, necessary to use dispersion modeling
to estimate neighboring population exposures. In keeping with the
generalized nature of this study, approximate dispersion estimates were
made using rough-cut Gaussian-plume techniques. Centerline, ground
level, one-hour concentrations were calculated assuming a wind speed of
4 m/s and neutral ("D") stability. Based on engineering data
characteristic of the production facilities, a typical stack height of
25 m was used to assess point source emissions (oxychlorination process
vent and direct chlorination process vent). Fugitive and storage
emissions were treated as area sources. The production area was assumed
2 •
to be 0.01 km , and the area of storage tanks at production facilities
were taken from the following equation supplied by Mascone (1978):
Storage tank area (ft ) = ( v^ncT! tanks - 1)60
, „ , total production capacity in 10 Ib/yr
. where: No. tanks = ^ T~Z^ —o
1. bo x J
Table V-6 lists the results of the dispersion modeling, giving the
average 1-hr downwind concentrations. These 1-hr average concentrations
were adjusted to annual average omnidirectional concentrations by first
dividing them by 20 for conversion to maximum annual values, then
further dividing them by 2.5 to smooth the maximum annual values with
respect to direction. These factors were derived by Youngblood (1978)
and are based on empirical data from studies of industrial sources
similar to those modeled here. The atmospheric concentrations shown in
Table V-6 are in ug/m . These concentrations can be converted to
parts per billion (ppb) by multiplying by 0.244.
Exposure Estimates
We used the emission factors for the four sources of EDC emissions
(fugitive, storage, direct chlorination vent, and oxychlorination vent)
in Table V-2 to scale the generalized dispersion curves in Table V-6.
In this way, we estimated atmospheric EDC concentrations as a function
of distance from each plant, the atmospheric concentrations from the
28
-------
Table V-6
ESTIMATED ONE-HOUR AVERAGE DOWNWIND
ATMOSPHERIC CONCENTRATIONS OF EDC* (pg/m3)
Downwind Point Source Emitter with Emitter with
Distance (km) Emitterb 0.0625-km2 Area0 0.01-km2 Areac
0.30 3,400 4,000 10,000
0.45 4,800 3,400 7,700
0.60 4,400 2,900 5,700
0.75 3,700 2,500 4,300
1.00 2,700 2,000 2,900
1.25 2,100 1,600 2,200
1.60 1,500 1,200 1,600
2.50 800 720 810
4.00 410 380 410
6.00 230 . 220 220
9.00 120 120 120
14.00 66 64 66
20.00 39 39 39
aAssumes an emission rate of 100 g/s for each source, neutral ("D")
stability atmospheric conditions with a wind speed of 4 m/s.
bSingle stack 25 m high.
GEffective emission height of 10 m.
Source: Modeling data provided by P. Youngblood (EPA, 1978).
29
-------
four emission sources were summed at each downwind distance to give
total concentration. The point source emission concentrations of Table
V-6 were used for oxychlorination and direct chlorination sources; the
2
0.01 km emissions area concentrations were used for fugitive emis-
2
sions; and linear interpolation between the 0.01 and 0.0625 km area
concentrations were used for storage emissions, depending on the com-
puted storage area. The total annual average EDC concentration esti-
mates as a function of distance from each plant were used to determine
the radii at which the specified annual average concentrations (i.e.,
0.01, 0.3, 0.6, 1.0, 3.0, 6.0, and 10.0 ppb) are attained in the vicin-
ity of each plant.
The population residing within the radial distances to the concen-
trations specified above was estimated by SRI's computer system, BESTPOP
(Suta, 1978). The population file consists of a grid of 1-km square
sections that span the continental United States. This file was created
by assigning the 1960 and 1970 populations to the grid network and by
assuming uniform distribution of population within each of 256,000 enum-
eration districts. The computer software accesses the population file
and accumulates residential population within radial rings specified
about any given point. In addition, a rectangular map that is printed
out for an area around each specified point shows the population by
square kilometer.
We determined the latitude and longitude for each facility by con-
tacting the company directly, from regional planning groups, or from
other studies completed for EPA.
Table V-7 gives estimated population exposures to EDC from produc-
tion facilities. The number of people exposed to concentrations of less
than 0.1 ppb is underestimated because the dispersion modeling was not
extrapolated beyond 30 km from any plant. The larger EDC producers are
estimated to cause exposures of 0.01 to 0.1 ppb at distances beyond 30
km from their locations.
30
-------
Table V-7
ESTIMATED HUMAN POPULATION EXPOSURES
TO ATMOSPHERIC EDO EMITTED BY PRODUCERS
Annual Average
Atmospheric EDC Number of People
Concentration (ppb) Exposed
;
10.0 1,700
6.00 -10.00 3,300
3.00-5.99 28,000
1.00 - 2.99 280,000
0.60-0.99 400,000
0.30 - 0.59 1,500,000
0.10 - 0.29 4,300,000
0.060- 0.099 l,900,000a
0.030- 0.059 3,500,000a
0.010- 0.029 550,000a
Total 12,500,000
aThese are underestimates because the dispersion modeling results were not
extrapolated beyond 30 km from each EDC production facility.
31
-------
It is estimated that 12.5 million people are exposed to annual EDC
concentrations greater than 0.01 ppb from EDC producers. Approximately
6.5 million of these are exposed to concentrations greater than 0.1 ppb.
Comparison of Monitoring and Modeling Concentrations
Monitoring data were available for three locations having EDC pro-
duction plants (Calvert City, Lake Charles, and New Orleans). The moni-
toring data for each location are given in Tables V-3 through V-5. As
has been previously 'described, the monitoring data were recorded as
24-hr samples taken for 12-13 days at each plant in the period of August
to October, 1978. Table V-8 averages these data for various distances
from each plant and also presents the average concentrations over the
three locations. The corresponding annual average dispersion modeling
concentrations are also given in Table V-8. Therefore, when comparing
the monitoring and modeling data, it is necessary to remember that the
two are not expected to agree precisely since the monitoring data are
site-specific and were recorded over a relatively short period of time
while the modeling data are based on general assumptions and are intend-
ed to represent annual average conditions. Thus, the degree to which
the annual and average monitoring concentrations are comparable depends
in part on the local meteorological conditions during sampling, the
placement of the monitoring stations, plant production during monitor-
ing, averaging times, and assumed source configurations. It is believed
that the monitoring data typify the year's average conditions.
Both monitoring and modeling data indicate that elevated concentra-
tions of EDC occur at distances of at least 14 km from the plants. Com-
parisons of the monitoring and modeling data at various distances from
the plants show that:
o The monitoring concentrations are approximately 20% higher
than the modeling ones for distances of less than 1 km.
o The modeling concentrations are 30-70% higher than those moni-
tored for distances of 1-4 km.
32
-------
o For distances of 4-14 km, both monitoring and modeling results
appear to be of about the same magnitude.
Thus, there is sufficient agreement between the monitoring and mod-
eling concentrations to conclude that the generalized modeling analysis
gives a reasonable first-cut estimate of ambient EDC concentrations near
production facilities allowing estimates of potential population expo-
sures.
33
-------
Table V-8
COMPARISON OF EDC MONITORING AND
MODELING ATMOSPHERIC CONCENTRATIONS (ppb)
3-Location
Modeling
Average8
13.5
9.1
6.5
4.3
2.1
1.6
1.2
0.4
aData are the average 24-hr concentrations over 10 to 13 days for monitoring
and estimated annual averages for modeling.
bThe EDC emissions have been estimated in Table V-2 as 72.9 g/s for B. F.
Goodrich, Calvert City, KY; 70.2 g/s for Conoco,. Lake Charles, LA; and 119.5
g/s for Shell, New Orleans, LA.
cIndicates that no monitoring data were collected.
Monitoring Average Concentrations*
Distance
(km)
0.7-1.0
1.1-1.5
1.6-2.0
2.1-3.0
3.1-4.0
4.1-5.0
5.1-6.0
14.0
Calvert
Cityb
2.0
c
1.0
2.1
0.9
c
c
c
Lake
Charlesb
36.7
6.0
7.0
1.1
c
c
c
c
New
Orleansb
6.3
1.4
1.6
0.7
0.3
c
0.9
0.6
3-Location
Average
15.0
3.7
3.1
1.3
0.6
c
0.9
0.6
34
-------
VI POPULATION EXPOSURES FROM PRODUCERS
THAT USE EDC AS A FEEDSTOCK
General
In estimating human population exposures to atmospheric EDC from
chemical production facilities that used EDC as a feedstock, products
considered include VCM, 1,1,1-TCE (or methyl chloroform), TCE, PCE, EA,
VDCM, and gasoline lead scavenger. Many of these chemicals are produced
at the same facilities that produce the EDC feedstock. Exposure
estimates are given for each product considered separately and also for
all products combined.
Sources of Emissions
Table IV-2 lists producers that use EDG feedstock and their
cpacities. The number of producers of each chemical that uses EDC as a
feedstock is as follows:
Chemical Producers Using EDC
VCM 14
1,1,1-TCE 3
TCE 4
PCE 4
EA 3
VDCM 3
Lead scavenger 6
See Table IV-2.
Table IV-4 indicates that, depending on the chemical, from 52 to 89% of
the production capacity was used in 1977 for the preceding chemicals.
35
-------
Because actual production data for each chemical at each plant are
unavailable, we have assumed that each operates at the percent of capac-
ities shown in Table IV-4.
Emissions
The VCM production process is well controlled to limit emissions.
These controls also reduce emissions from the EDO used as feedstock. It
is estimated that the EDO emission factor for VCM is 0.025% of EDC
input, the EDC emission factor for other processes that use EDC as a
feedstock is estimated as 0.2% of the EDC input (Mascone, 1978). The
estimated EDC emissions are given in Table VI-1. It is assumed that
these emissions are of a low level fugitive type resulting from leaks in
valves and other processing equipment and from storage-tank evaporation.
Atmospheric Concentrations
Because so few atmospheric monitoring data exist for the vicinities
of production plants that use EDC as a feedstock, it has been necessary
to use dispersion modeling to estimate neighborhood population
exposures. (Dispersion modeling is described in Section V.) The
O ' *' ' '. '— .""!*'
dispersion estimates for a 0.01-km area source emitter (Table V-6)
was used for assessing exposure.
Exposure Estimates
The EDC emissions given in Table VI-1 were used to scale the
dispersion curve to estimate atmospheric EDC concentrations as a
function of distance from each plant for each product. Concentrations
were similarly estimated about each plant for emissions from all
products. The tables showing annual average atmospheric EDC
concentrations as a function of distance from each plant were used to
determine the radii at which specified annual average concentrations
(i.e., 1.0, 0.6, 0.3, 0.1, 0.06, 0.03, and 0.01 ppb) are attained. The
population residing within the distances to the concentrations specified
above was estimated by SRI's computer system, BESTPOP (Suta, 1978). We
determined the latitudes and longitudes for each facility "by contacting
the company directly, by using information from regional planning
36
-------
Table VI-1
ESTIMATED EDC ATMOSPHERIC EMISSIONS (g/s) FOR
PLANTS THAT USE EDC AS A FEEDSTOCK
Plant8
Borden
Conoco
Diamond
Diamond
Dow
Dow
Dow
duPooC
duPont
Ethyl
Ethyl
Goodrich
Houston
1C I America
Nalco
PPG
PPG
Shell
Shell
Stauffer
Union Carbide
Union Carbide
Vulcan
Location
Geismar, LA
Lake Charles, LA
Deer Park, TX
La Porte, TX
Freeport, TX
Oyater Creek, TX
Plaquemine, LA
Antioch, CA
Deepwater, NJ
Baton Rouge, LA
Houston, TX
Calvert City, KY
Beaumont, TX
Bacon Rouge, LA
Freeport, TX
Lake Charles, LA
Guayanilla, PR
Deer Park, TX
Norco, LA
Caraon, CA
Taft, LA
Texas City, TX
Geismar, LA
1,1,1
VCM TCE
1.2
2.9
4.1
0.8 5.6
2.9
5.1 3.7
1.4
4.1
1.2
1.3 4.3
2.1
3.4
2.9
0.7
TCE
0.7
2.0
PCE
1.6
EA
VDCM
2.7
2.3
2.3
0.6
0.5
2.7
1.9
1.5
Total
34.1
13.6
6.0
1.5
5.5
3.2
2.7
8.6
Lead
Scavenger
1.1
1.1
1.1
1.1
0.9
0.3
"
Total
1.2
2.9
2.3
13.4
2.9
8.8
3.4
1.1
3.6
1.1
4.1
0.9
1.2
0.3
11.7
2.1
3.4
2.9
0.7
3.2
2.7
1.5
6.1
5.6
79.5
"Blanks indicate the chemical ia not manufactured at the plant in question or that the plant has no EDC emissions.
Source: SRI estimates.
-------
groups, or from other studies completed for EPA. The population
exposures to EDC for individual products that require EDC as a feedstock
are given in Table VI-2.
VCM production results in the largest number of exposures — about
1.3'raillion people — about half of all the exposures for products using
EDC as a feedstock.
Table VI-3 gives the total exposures from all EDC feedstock
producers for all products. As shown in Table IV-2, many facilities
produce several products requiring EDC. Thus, two alternative estimates
of total exposures are given in Table VI-3: Alternative A combines all
the emissions reported in Table VI-1 and then uses these combined
emissions to estimate total exposures about each plant. Alternative B
is a summation of the individual product exposures shown in Table VI-2.
Because Alternative B counts some people twice (or more times), it
results in an overestimate of total exposures; however, the exposure
concentrations are lower than for Alternative A.
38
-------
Table VI-2
ESTIMATES OF POPULATION EXPOSURES TO ATMOSPHERIC EDC EMITTED
BY PLANTS THAT USE EDC AS A FEEDSTOCK IN VARIOUS PRODUCTS
u>
Annual Average
Atmospheric EDC
Concentration (ppb)
0.600-0.999
0.300-0.599
, 0.100-0.299
0.060-0.099
0.030-0.059
0.010-0.029
Product
VCM
1,300
360
30,000
42,000
260,000
940,000
1,1,1-TCE
1,700
16,000
83,000
170,000
TCE
390
10,000
47,000
140,000
PCE
80
SOO
17,000
250.000
EA
70
17,000
8,000
43,000
37.000
VDCM
270
3,400
34,000
90,000
Lead
Scavenger
1,900
3,400
25,000
350,000
Total
1,300,000
260,000
200,000 270,000
110,000
130,000
380,000
-------
Table VI-3
ESTIMATES OF TOTAL POPULATION EXPOSURES TO ATMOSPHERIC EDC
EMITTED BY PLANTS THAT USE EDC AS A FEEDSTOCK
Annual Average Alternative A Alternative B
Atmospheric EDC Sum of Sum of
Concentration (ppb) Emissions3 Exposures^3
0.600-0.999 1,300 1,300
0.300-0.599 2,100 430
0.100-0.299 110,000 51,000
0.060-0.099 210,000 83.000
0.030-0.059 520,000 510,000
0.010-0.029 1,500.000 2,000,000 .
Total 2,300,000 2,600,000
aExposures are based on the total feedstock emissions for each plant given
in Table V-l.
"Exposures'are"based on the sum of the exposures for each product given in
Table V-2.
-------
VII POPULATION EXPOSURES FROM EDO IN AUTOMOBILE GASOLINE
General
Leaded gasoline contains EDC as a lead scavenger, with the amount
of EDC depending on lead content. Catalytic converters were first
required for the 1975 model year. In that year, 100% of Ford and GM
cars and 96% of Chrysler cars required unleaded gasoline. In 1976,
approximately 20% of the gasoline sold in the United Satats was unleaded
(Ethyl Corporation, 1976).
The antiknock "motor mix" added to gasoline is a combination of
ethylene dibromide (EDB), EDC, lead, and other alkyl groups. Quantities
of EDC and EDB are used sufficient to supply two atoms of chlorine and
one atom of bromine for each atom of lead. Major gasoline antiknock
mixes (which are added in small quantities to leaded gasoline) typically
contain 18.8% EDC by weight and 17.9% EDB by weight (SRI estimates).
Whereas the current average lead content in all gasoline is 1.5 g/gal,
leaded gasoline contains approximately 2.5 g/gal. According to EPA's
phase-down schedule for lead, the average lead content for all gasoline
is expected to be 0.5 g/gal by 1 October 1979 (Stolpman, personal
communication, 1977). Because of the low lead content in unleaded
gasoline (approximately 0.01 g/gal), EDC is not required and is not
added. Therefore, our analysis of population exposures related to
gasoline use considers only leaded gasoline. If we assume that gasoline
contains 0.425 units of EDC per unit of lead (by weight), this results
in an estimated 1.1 g of EDC per gallon of leaded gasoline, or 0.02% EDC
by volume.
We have used gallons rather than liters to represent gasoline volume
since these are the units commonly used in the United States for
gasoline sales.
41
-------
We evaluate nonoccupational population exposures to atmospheric
emissions of EDC from leaded gasoline for three sources:
o Exposures to people who refuel their automobiles at
self-service stations
o Exposures to people who reside in the vicinity of service
stations
o General urban population exposures from the evaporation of EDC
from automobiles.
Exposures from Self-Service Operations
Sources of Emissions
Service station types are characterized by the services they offer
and their business operations; they include (1) full-service stations,
(2) split-island stations, (3) self-service stations, and (4)
convenience store operations. In full-service stations (1), attendants
offer all services, including gasoline pumping and other mechanical
check-ups. If fuel is obtained at any class of stations (2) through
(4), the customers themselves may fill their tanks. In split-island
stations (2), both self-service and full-service are offered. At
stations (3) and (4), only self-service is available.
While pumping gasoline, an individual is exposed to EDC released as
vapor from the gasoline tank.. Although occupants in the car at both
self-service and full-service operations are exposed to some EDC, the
highest exposures are to the person pumping the gas. Because it is
difficult to estimate level and length of exposure for car occupants,
only those who pump gasoline from self-service pumps are considered
Vapor recovery systems affixed to the gasoline nozzle can reduce
exposure levels significantly if they are working properly and are
operated correctly. Such systems are required for service stations in
parts of California.
42
-------
here. (Note it is not within the scope of this report to evaluate
occupational exposures.)
Self-service sale of gasoline is a relatively new marketing method
pioneered by independent operators in the West Coast and in the southern
United States. Today, it accounts for 30% of gasoline sold. The
national market share of the major gasoline producers has decreased
recently as independents and others specializing in high-volume,
low-margin sales capture a larger percentage. Of the approximately
184,000 conventional service stations with some self-service operations
account for 39% (Arthur D. Little, 1977). Table VII-1 indicates the
types of service stations offering self-service gasoline.
Table VII-1
SELF-SERVICE OPERATIONS
Outlets Offering Self-Service % of U.S. Total
Total self-service 9
Split island with self-service 26
Convenience stores 4
Total outlets with self-service 39
Source: Arthur D. Little (1977)
An Arthur D. Little report (1977) revealed that 71,300 outlets
offer self-service gasoline. Gasoline sold at U.S. service stations for
a
the year ending May 30, 1977 equalled approximately 87.4 x 10 gal.
9
Of that amount, 27.0 x 10 gal (31%) is estimated to have been
dispensed at self-service pumps. The market share of self-service
stations was surveyed for four metropolitan Air Quality Control Regions
(AQCR): Boston, Dallas, Denver, and Los Angeles. The market share held
by self-service operations varied from 9% in Boston to 45% in Denver
(see Table VII-2). Another study by Applied Urbanetics, Inc. (1976)
surveyed Baltimore and Madison, Wisconsin. The results of that study
43
-------
Table VII-2
GASOLINE MARKET SHARE OF SELF-SERVICE STATIONS
IN FOUR AQCRs SPRING 1977
Type of Operation
Boston AQCR
Full-service
Self-service (total)
Split island
Self-service
Convenience stores
Number of
Outlets
2,253
100
8
92
Sales Volume
(106 gal/yr)
1,045.1
108.6
Market
Sharing
Percent
91.0
9.0
Dallas AQCR
Full-service 2,094
Self-service (total) 1,124
Split island 480*
Self-service 444
Convenience stores 200
924.6
593.8
61.0
39.0
Denver AQCR
Ful1-service
Self-service (total)
Split island
Self-service
Convenience stores
621b
656
310a
226
120
292.1
235.7
55.0
45.0
Los Angeles AQCR
Full-service 2,518
Self-service (total) 4,780
Split island 3,632a
Self-service 1,022
Convenience stores 126
2,472.6
2,154.8
53.0
47.0
Split-island operations offering full service, and self-serve islands.
these, 445 are split-island operations that offer full service and
mini-serve (attnedant-operated) islands.
Source: Arthur D. Little (1977).
44
-------
are shown in Table VII-3. It appears that self-service operations
account for about 40% of the market in urban areas.
Emissions
To estimate the people exposed to EDO from service stations,
several assumptions were necessary. The gasoline pumped.through
Q
self-service outlets is estimated at 27.0 x 10 gal/yr. The annual
average fuel consumption per vehicle is 736 gal (DOT, 1974a). If it is
assumed that on the average, a person who primarily uses self-service
gasoline makes one trip per week to the gasoline station, an average
fill-up amount of 14 gal is determined by dividing 736 gal/vehicle/yr by
52 wk/yr. By dividing the average fill-up into the self-service gallons
pumped, we estimate trips per year to self-service operations at 1.9 x
g
10 . When this number is divided by 52 trips per person per year, the
people exposed to pumping self-service gasoline is estimated at 37 x
10 . We can further assume that only 80% of these people are pumping
leaded gasoline containing EDO. Therefore, the people exposed from this
source is estimated to be 30 x 10 . For this estimate of the
population exposed, we assume that the individuals using self-service
gasoline obtain all"ofthe'ir'gasoline at self-service stations.
Atmospheric Concentrations
A rough estimte of EDC exposures was made by extrapolating the
results of the Battelle (1977) benzene monitoring. In that study, three
samples of ambient air were taken in the breathing zone of persons
filling their tanks at self-service gasoline stations. The results,
shown in Table VI1-4, indicate a wide range in the benzene
concentrations of the emissions. The variations seem to be related to
the subject's position in relation to the tank opening and the wind
direction. Because all measurements were taken on the same day and at
approximately the same time, ambient temperature did not cause the
variation. .Basically, if the subject was downwind of the tank opening,
higher levels were recorded.
45
-------
Table VII-3
GASOLINE MARKET SHARE OF SELF-SERVICE STATIONS
IN TWO METROPOLITAN AREAS, 1976
Type of Operation
Baltimore SMSA '
Full-service
Self-service (total)
Split island
Self-service
Sales Volume
gal/yr)
111.53
90.5
25.5
65.0
Market
Sharing
Percent
55.0
45.0
Madison SMSA
Ful1-service
Self-service (total)
Split island
Self-service
77.0
17.0
60.0
42.0
58.0
alncludes the sales from mini-serve (attendant-operated) stations and 50% of
the sales from split islands.
Source: Applied Urbanetics, Inc. (1976).
46
-------
Table VII-4
SAMPLING DATA FROM SELF-SERVICE GASOLINE PUMPING
Customer
1
2
3
Sampling
Rate
(mL/min)
31
31
31
Nozzle
Time
(min)
2.5
1.1
1.6
Gallons
Pumped
14
8
9
Sample
Volume
(L)
78
34
50
Benzene
Level
g/m3
115
324
1,740
ppb
r, i i
43
121
647
Source: Battelle (1977).
No EDC monitoring data obtained in the vicinity of gasoline
stations are available, therefore, by determining the evaporation rate
of EDC with respect to benzene, benzene monitoring data can be used to
provide a rough estimate of EDC exposures. It is known that the
evaporation rate is proportional to the vapor pressure, solubility, and
the square root of the molecular weight. Thus, the following equation
can be used to estimate the EDC emission factor (or emission rate)
related to evaporation:
E P S./m~
a
e . e e e (7>1)
PbV
where the subscript e refers to EDC and the subscript b refers to
benzene; E is the emission rate (or emission factor); P is the vapor
pressure; S is the solubility; and m is the molecular weight.
For an estimation, (s) (\TBT) may be approximated by Xj the molar
fraction or concentration, thus, Equation (7.1) can be written as
follows:
47
-------
E P
a
6 (7.2)
Eb pb *b
The Battelle benzene monitoring data were taken when the temperatures
was about 20°C. Because the vapor pressures for EDC (70 mm) and
benzene (80 mm) at 20°C are known, and the volume concentrations of
EDC (0.02%) and benzene (2.0%) in gasoline are also available, the
emission factor (or emission rate) of EDC can be estimated by the
following equations:
70 0.02
Ee = 80 x-2To-Eb <7'3>
Eg = 0.009 Eb . (7.4)
This factor can be used to scale benzene atmospheric concentrations
( g/m ) to corresponding EDC concentrations becuse it is assumed that
atmospheric concentrations are proportional, the corresponding EDC
exposures were estimated, based on these data and are given in Table
VII-5. .
Table VII-5
ESTIMATES OF EDC EXPOSURES FROM SELF-SERVICE GASOLINE PUMPING
Nozzle Gallons Estimated EDC Level
Customer Time (min) Pumped g/m^ppb
1 2.5 14 1.04 0.27
2 1.1. 8 2.91 0.71
3 1.6 9 15.66 3.83
Average nozzle time = 1.7 min
Time weighted average exposure = 1.45 ppb
Source: SRI estimates based on Battelle monitoring data (1977). the
conversion is based on Equation (7l4).
48
-------
Exposure Estimates
The estimated exposure levels are based on the information con-
tained in Table VII-5. It is recognized that these data are limited and
highly variable. However, they do allow a reasonable estimate of ex-
pected exposure levels from self-service gasoline pumping. In states
where vapor recovery systems are used, the estimated exposure level may
be much lower. Approximately 30 x 10 persons use self-service sta-
tions. While filling their tanks once a week, they are exposed to an
estimated EDG level of 1.5 ppb for 2.5 min (time required to pump 14
gal). Their annual exposure is estimated at 2.2 hr. This equates to an
annual averge time-weighted exposure of 0.0004 ppb.
Exposures in the Vicinity of Service Stations
Sources of Emissions
People residing in the vicinity of service stations, may be exposed
to EDC from the evaporation of gasoline pumped by attendants and cus-
tomers, and from gasoline loaded by distribution trucks. These expo-
sures are in addition to those assessed previously for persons using
self-service gasoline stations. The amount of EDC emitted depends on
the ambient temperature, vapor recovery controls, the EDC content in
gasoline, and the volume of leaded gasoline pumped. Approximately 80%
of the gasoline currently sold is leaded. With approximately 184,000
service stations in the United States, it is expected that many people
are exposed to EDC from those sources. Because most service stations
are located in urban areas and because their location is expected to be
highly correlated with urban population density, only urban areas are
considered in this analysis.
Emissions
An EDB emission factor of 0.00157 g of EDB per gram of lead per
gallon from refueling losses has been estimated, based on testing at
EPA's Mobile Source Air Pollution Control Laboratory in Ann Arbor,
Michigan (Kittredge, 1977). The factor considers spilling, vapor dis-
placement, entrained liquid gasoline losses, and volume of gasoline
pumped. Assuming an average lead content in gasoline of 2.5 g/gal, the
49
-------
estimated emission factor for EDB is 0.00039 g/gal. The EDB emission
factor can be used to estimate the EDC emission factor through the use
of Equation (7.2) (and by substitution of EDB factors for benzene fac-
tors in the equation). We have assumed that the EDB vapor pressure is
12 mm at 25°C and that it constitutes 0.05% of the gasoline (by vol-
ume). Hence, we estimate that the EDC emission factor for automotive
refueling, losses (E ) is:
e
80 0 02
E = 7^ x ^f x 0.00039 = 0.001 g/gal. (7.5)
e 12 0.05 °
The number of service stations in urban areas can be estimated,
based on urban service station density and total U.S. urban population.
Service station density in urban areas can be extrapolated from the data
presented in Table VII-6. The service station density shown for four
metropolitan AQCRs varies, with no regional pattern evident. Based on
these data, we estimate an average of 0.7 service stations per 1,000
population. This number can be applied generally to urban areas
*
throughout the United States. Urbanized areas provide the best popu-
lation base. The 1970 population residing in urbanized areas was
118,447,000 (Bureau of the Census, 1975). Thus, service stations in
urbanized areas are estimated at 83,000, or 45% of all stations.
An emission rate can be estimated by employing the following
assumptions:
(1) 70.0 x 10? gal of leaded gasoline are sold annually by
service stations.
(2) The average number of gallons pumped per service station is
3.8 x 105 gal. (The United States has approximately 184,000
service stations.)
Defined by the Bureau of Census as the central city or cities and
surrounding closely settled territories. All sparsely settled areas in
large incorporated cities are excluded by this definition. Densely
populated suburban areas, however, are included (U.S. Department of
Commerce, Bureau of the Census, 1972 County and City Data Book).
50
-------
Table VII-6
SERVICE STATION DENSITY IN FOUR METROPOLITAN AQCRs
Number ofa AQCRb Service
Service Stations Population Stations0 per
AQCR (1977) (1975) 1.000 Population
Boston 2,353 4,039,800 0.6
Dallas 3,218 2,970,900 1.1
Denver 1,277 1,389,000 0.9
Los Angeles 7,298 14,072,400 0.5
Sources:
aA. D. Little (1977).
^Bureau of Economic Analysis (1973)
CSRI estimates.
51
-------
(3) All service stations have uniform pumping volumes.
(4) The EDC emission rate for a service station is:
(Vol. of gasoline pumped)(emission factor) = emission rate;
that is, (3.80 x 105 gal/yr)(0.001 g/gal) = 380 g/yr
= 1.2 x 10~5 g/s.
Atmospheric Concentrations
Dispersion modeling of benzene emissions from gasoline service
stations (Youngblood, 1977) employed the Single Source (CRSTER) Model
(DSEPA, 1977). Meteorological data for Denver, Colorado were used to
represent a reasonable worst-case location. EDC emissions from gasoline
service stations are thought to be from sources similar to those for
benzene. The dispersion modeling assumed that the sources of emissions
i
are dispersed over a 50 sq ft area. -The benzene concentrations of the
previous study were modified to reflect EDC emissions by first
multiplying.the benzene concentrations (ppb) by 3.2 to convert them to
3 3
/Ag/m and then by multiplying the /Jg/m by 0.244 to convert to ppb of
EDC.
Table VII-7 presents the results of the benzene modeling modified
for EDC; an EDC emission of 0.01 g/s is assumed. Two conditions are
given: (1) the 8-hr worst-case concentrations for a service "station
that operates only during daytime, and (2) the annual average
concentrations for a service station that operates 24 hr/day.
Exposure Estimates
Population exposures to EDC emissions from gasoline service
stations have been estimated by assuming that the population in
2
urbanized areas is uniformly distributed with a density of 1,318/km
(based on the 1970 census). We have also assumed that no one resides
within 50 m of a gasoline service station. To calculate atmospheric
concentrations as a function of distance, the annual average dispersion
modeling data in Table VII-7 were-sealed by the estimated EDC emissions
52
-------
Table VII-7
ROUGH DISPERSION MODELING RESULTS FOR EDC EMISSIONS
FOR GASOLINE SERVICE STATIONS3
Distance (m) 8^hr Worst Case (ppb)b Annual Average (ppb)c
50 12 1.0
100 6 0.5
150 3 0.3
200 2 0.2
300 1 0.1
aAssumes an EDC emission of 0.01 g/s during operation.
^Assumes continuous operation from 8 a.m. to 4 p.m., 6 days per week.
cAssumes continuous operation 24 hr per day, 7 days per week.
Source: Modified from Youngblood (1977) by adjusting ppb of benzene to ppb of
EDC.
53
-------
(ratio of estimated emissions to emissions on which the dispersion curve
is based); Two cases were considered: (1) the 83,000 urban service
stations were distributed by assuming that no two are closer than 300 m
to each other and that no people reside closer than 50 m, and (2) the
urban service stations were geographically distributed so that three are
always located in close proximity. Thus, for the 27,633 triplets we
have assumed that no people reside closer than 100 m. The EDO emission
rate for the single service station case is taken as 1.2 x 10 g/s
and for triplet service station case as 3.6 x 10 g/s. The actual
geographic distribution of urban service stations is assumed to be
somewhere in between these two cases; therefore, the exposures estimated
for these two cases are expected to bound actual exposures.
Because of the population exclusion radii (50 and 100 m) and
assumed emissions, no exposures are estimated to occur for EDO annual
average concentrations greater than 0.03 ppb. In some cases, people may
reside closer to service stations than these exclusion radii permit. In
these cases, some would be exposed to atmospheric concentrations in
excess of 0.03 ppb. It is estimated that 600,000 people are exposed in
the single service station case and that 1.4 million are exposed in the
triplet service station case. All of these exposures are estimated to
be in the 0.01 to 0.03 ppb annual average concentration range.
These estimates, which are only rough approximations, are based on
assumptions of uniform distribution of service stations in urbanized
areas, uniform pumping volumes, average populaton density, and on
dispersion modeling. In reality, more service stations are located in
commercial areas than in residential areas, and pumping volumes vary
substantially.
Urban Exposures Related to Automobile Emissions
Sources of Emissions
Urban exposures to EDO come from many sources, including gasoline
evaporation, gasoline service stations, losses through transportation
and storage of gasoline, and emissions from production facilities. Most
54
-------
of these sources have been treated as point sources and their emissions
are evaluated elsewhere in this report. This section presents analysis
of exposures due to emissions of EDC from automobiles.
Emissions
. As previously discussed, the EDC content in leaded gasoline
averages 0.02% by volume, and leaded gasoline accounts for 80% of all
gasoline sold. Tests by EPA's Mobile Source Air Pollution Control
Laboratory in Ann Arbor, Michigan, have indicated that EDB is destroyed
in the combustion process (Kittredge, personal communications, 1977),
and it has been calculated that EDC is similarly destroyed (Mabey,
1978). However, evaporation from the carburetor and from the fuel tank
does occur. We have been unable to locate data on EDC emissions from
this type of evaporation; however, data are available for tests
measuring EDB (Table VII-8).
Table VII-8
AUTOMOTIVE EDB EMISSION FACTORS
(G/G OF LEAD PER GALLON OF GASOLINE)
Vehicle Type Low High
Uncontrolled vehicle (pre-1972) 0.00144 0.00362
Pre-1978 controlled vehicle 0.00098 0.00250
Post-1978 controlled vehicle 0.00033 0.00085
Source: Kittredge, 1977
The average EDB emission factor for uncontrolled vehicles, based on
Table VII-8, is 0.00253 g of EDB per gram of lead per gallon. Assuming
2.5 g of lead per gallon of leaded gasoline, the estimated EDB emission
factor is 0.0063 g of EDB per gallon. The EDB emission factor can be
used to estimate the EDC emission factor through the use of Equation
(7.2). By substitution in this equation, the EDC emission factor (E )
for uncontrolled automobile evaporation becomes:
55
-------
on n no
Ee = ll X irof X 0'0063 = °-017 8 of E°C/Sal (7.6)
This factor will provide a slightly high estimate of ambient EDO levels
because it assumes all automobiles using leaded gasoline have emissions
comparable to pre-1972 models. In fact, the 1975 model year was the
first in which automobiles were required to run on unleaded gasoline,
but controls to reduce :oarburetor evaporation were introduced in the
1972 model year.
Dispersion Modeling
The Hanna-Gifford area-wide dispersion model (Gifford and Hanna,
1973) as applied by Schewe (1977) for benzene is used for this analysis
and modified for EDC. Mara and Lee (1978) contains a discussion of this
model and its application to benzene. Because EDC is destroyed during
combustion (Mabey, 1978), only evaporation is considered. The modified
equation to estimate the emission rate for EDC is as follows:
Q = (0.017 g/gal) I atmual travel -"lies /vehicle \
evap ° \ average miles/gal /
(vehicles registered) - (7.7)
If 12,000 mi/yr for each vehicle and 12 mi/gal are assumed (DOT,
1974b), the above equation becomes
Qevap - (5.4 x ID'7 g/s) x ( vehiclesj^istered J (?>8)
To calculate the annual average areawide EDC concentration, the
following equation is used:
225 Q
= _ evaP (7.9)
56
-------
where u is wind speed (m/s) and )r is the atmospheric concentration in
3
g/m . The average annual wind speed, u, in the area.of study was
obtained from AP101 (EPA, 1972). Because wind speed (and thus
dispersion) increases in the afternoon, the morning values were used to
estimate higher concentrations, the number 225 is an empirical factor
derived from several studies that give very good results for long-term
averages for low-level emission sources (Gifford and Harm a, 1973).
Estimates of Exposures
Cities whose vehicular densities are higher should have the higher
EDC concentration from automobile evaporation. Because 916 cities in
the United States have populations greater than 25,000, we used a
statistical sampling approach to evaluate EDC exposures. Table VII-9
shows the distribution of cities by size. Because of the expected
higher EDC concentratons in the larger cities, it was decided to
evaluate the exposures for the 26 cities with populations greater than
500,000 and to do a fractional sample of the cities in the smaller
groups. However, because vehicle registration data were unavailable for
Boston and New Orleans, only 24 of the largest 26 cities were
evaluated. Of the cities in the 250,000-500,000 size range, 25% were
Table VII-9
DISTRIBUTION OF CITIES BY 1970 POPULATION
Number of Combined
Population Size Cities Population
1,000,000 6 18,769,000
500,000-1,000,000 20 12,967,000
250,000- 500,000 30 10,442,000
100,000- 250,000 100 14,286,000
50,000- 100,000 240 16,724,000
25,000- 50,000 520 17,848,000
Source: U.S. Bureau of the Census, Statistical Abstracts of the United
States-1974.
57
-------
sampled, as were 15% of the cities in the 100,000 to 250,000 size
range. No EDC exposures greater than 0.01 ppb were estimated for cities
with population less than 100,000. Exposures were estimated for people
within the sample, and these sampling results were then projected to all
cities, based on the ratio of total population to sample population.
Table VII-10 sets forth the exposure estimates for the cities ,
sampled. No exposures were found to exceed 0.03 ppb. Table VII-11
gives the calculations used to project the sampling data to the total
population. Based on this projection we estimate that approximately 13
million people are exposed to annual average EDC concentrations of 0.01
to 0.03 ppb from the evaporation of gasoline from automobiles.
Summary of Urban Exposures from Automobile Gasoline
Exposures to EDC from leaded gasoline have been estimated for
people refueling their automobiles at self-service stations, for those
residing near service stations, and for those exposed to EDC evaporation
from automobiles. We estimate that approximately 30 million people are
exposed to an EDC concentraton of 1.5 ppb for 2.2 hr/yr while refueling
their automobiles. Approximately 600,000 to 1,400,000 (average of 1
million) people residing near gasoline service stations are exposed to
annual average EDC concentrations of 0.01 to 0.03 ppb from refueling
losses. Another 13 million people are exposed to annual average EDC
concentrations of 0.01 to 0.03 ppb from automobile evaporation.
58
-------
Table VII-10
Rank
Name
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
17
18
20
21
22
23
24
25
26
27
32
37
42
47
52
57
62
72
82
87
92
97
112
117
122
127
132
137
142
151
157
162
New York, NY
Chicago, IL
Los Angeles, CA
Philadelphia, PA
Detroit, MI
Houston, TX
Baltimore, MD
Dallas, TX
Washington, DC
Cleveland, OH
Indianapolis, IN
Milwaukee, WI
San Francisco, CA
San Diego, CA
San Antonio, TX
Memphis, TO
St. Louis, MO
Phoenix, AZ
Columbus, OH
Seattle, WA
Jacksonville, FL
Pittsburgh, PA
Denver, CO
Kansas City, KA
Atlanta, GA
Minneapolis, MN
Oklahoma City, OK
Miami, FL
Norfolk, VA
Akron, OH
Richmond, VA
Corpus Chriati, TX
Ft. Wayne, IN
Fresno, CA
Santa Ana, CA
Lubbock, TX
Riverside, CA
Peoria, IL
Macon, GA
Savannah, GA
Columbia, SC
Alexandria, VA
Al lent own, PA
Hollywood, FL
Duluth, MN
Pueblo, CO
Sunnyvale, CA
ESTIMATED U.S. CITY EXPOSURES TO EDO FROM THE EVAPORATION OF AUTOMOBILE GASOLINE
Qevapc
(10-10 g/8-02)
Population8
(1,000)
Automobiles'5
(1,000)
Wind Speed
(m/s)
7,895
3,363
2,816
1,949
1,511
1,232
906
844
757
751
746
717
716
697
654
624
622
582
539
531
529
520
515
507
497
434
367
335
308
275
250
204
178
166
157
149
140
127
122
118
114
111
110
107
101
97
95
1,707
1,476
1,515
954
796
692
412
741
391
392
237
328
367
388
333
306
295
357
334
273
355
252
331
264
354
237
238
221
144
153
132
102
112
86
98
80
73
66
72
63
63
57
58
82
45
50
55
.11.9
13.8
6.8
15.5
12.0
3.3
11.0
5.8
13.3
10.8
1.9
7.2
16.9
2.6
3.8
2.9
10.1
3.0
5.2
6.8
1.0
9.5
7.3
1.7
5.6
9.0
0.8
13.4
5.7
5.9
4.6
2.1
4.5
4.3
7.6
2.2
2.1
3.7
3.1
4.9
1.2
8.1
6.8
6.8
1.4
4.6
5.4
7
5
3
6
6
6
6
6
5
5
5
5
3
3
6
5
6
4
5
5
6
5
4
6
5
6
6
6
7
6
5'
6
5
3
3
6
3
5
5
5
5
5
3
5
6
4
3
Concentration
(ppb)
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
a 1970 census city population.
b Registered automobiles by SMSA as given in DOT (1974b) have been assigned to cities, based on the ratio of city
population to SMSA population.
c See Equation (7.8).
59
-------
Table VII-11
CALCULATIONS OF NATIONAL EXPOSURES* TO EDC
FROM AUTOMOBILE EVAPORATION
Projected
Fraction Sampled Population
Total U.S. Population Exposed to Exposures.
City Population Population Sampled 0.01 ppb 0.01 ppb
1,000,000 18,769,000 18,769,000 0.43 8,130,000
500,000-1,000,000 12,967,000 11,733,000 0.19 2,460,000
250,000- 500,000 10,442,000 2,670,000 0.13 1,310,000
100,000- 250,000 14,286,000 1,892,000 0.08 1,140,000
Total 13,040,000
*A11 exposures are in the 0.01'to 0.03 ppb range.
60
-------
VIII OTHER ATMOSPHERIC EXPOSURE ROUTES
General
Environmental exposures to EDO may occur through any of its
dispersive uses, including grain fumigants, paints, coating, adhesives,
cleaning, and the prepartion of polysulfide compounds. Additional
environmental exposures may occur from spills and venting during
transportation and from evaporation at waste disposal operations. Few
data are available on the amount of EDO used or lost to the environment
from these potential exposure routes; consequently, only qualitative
descriptions of exposures can be given.
Dispersive Uses
Auerback Associates, Incorporated (1978) estimated consumption of
EDO for minor uses in 1977 at about 5,000 mt. Of this total, about 28%
(1,400 mt) was used in the manufacture of paints, coatings, and
adhesives. Extracting oil from seeds, treating animal fats, and
processing pharmaceutical products required 23% of the total (1,150
mt). An additional 19% (950 mt) was consumed cleaning textile products
and PVC manufacturing equipment. Nearly 11% (550 mt) was used in the
preparation of polysulfide compounds. Grain fumigation required about
10% (500 mt). The remaining 9% (450 mt) was used as a carrier for
amines in leaching copper ores, in the manufacture of color film, as a
diluent for pesticides and herbicides, and for other miscellaneous
purposes. It is generally assumed that all of this material is
eventually released to the atmosphere (Drury and Hammons, 1978).
Atmospheric exposures to EDC from th.ese dispersive uses occur as
point source losses from the industrial sites where these products are
manufactured and from the use of end products.
61
-------
Atmospheric dispersion modeling was used to provide crude
estimates of magnitude of exposures that might be attained from the
manufacture of these other products. Use of the dispersion modeling has
shown that a plant would have to use more than 90 mt/yr of EDC to have
concentrations in excess of 0.01 ppb at the assumed plant boundary 0.5
km from the plant center. A plant that uses 1,000 mt/yr of EDC would
have concentrations of 0.1 ppb at the plant boundary (0.5 km) and 0.01
ppb at 2.5 km from the emissions. Production data are not available for
these dispersive uses. Since a total of 5,000 mt/yr are involved in all
dispersive uses, it is highly unlikely that any one production plant
would use as much as 1,000 mt/yr of EDC. Based on these preliminary
calculations, it appears that the EDC exposures to people residing near
these other manufacturing plants are minimal.
Nonoccupational exposures from end product use would occur
primarily from the use of paints, coatings, adhesives, and solvents, and
to people who inadvertently enter fumigated areas. All of these
exposures would, be intermittent and are extremely difficult to estimate.
Transportation
EDC may be emitted to the atmosphere during transportation from
inadvertent spills and from venting. The amount of EDC transported each
year is not well known because companies transport the chemical between
their plants as well as to other companies or to places for export to
other countries. We estimate that at least 672,000 mt of EDC were
transported during 1977 (approximately 13% of production). This
includes 5,000 mt required for minor dispersive uses, 177,000 rat for
exportation, and an estimated 490,000 mt transported within the United
t
States for use in the major products shown in Table IV-2. The estimated
For the modeling, it has been assumed that EDC emissions are 1% of
input and that the emissions occur over a 0.01 km^ area (see Table
V-6).
62
-------
490,000 me transported for major product use was obtained as the sum of
shortages between EDC production and EDC required at individual plants.
We assumed that each plants' production, for each product, was at the
total capacities given in Table IV-2 times the percent capacities used
during 1977 (Table IV-4). However, a chemical plant is flexible in
regard to the products it actually makes, and this factor could result
in considerably more EDC actually being transported than estimated.
The major transportation emissions would probably result from
venting and spillage during loading and unloading transportation
containers. Thus, these emissions would occur at the processing plants
and would add to the EDC emissions from other plant activities. Other
emissions could result from venting while in transport and from
accidental rupture of a container.
No data are available on EDC emissions during loading and
unloading; however, Mara and Lee (1978) give uncontrolled emission
factors for the transfer of benzene (Table VIII-1) by inland barge, tank
truck, and rail car. The benzene emission factors can be adjusted to
rough order EDC emission factors by adjusting for the differences in
vapor pressure through the use of Equation (7.2). This adjustment gives
an uncontrolled EDC emission factor of approximately 10 g/gal or 0.18%
by weight. This must be regarded as an upper limit because some of the
Table VIII-1
SUMMARY OF UNCONTROLLED EMISSION FACTORS
FOR THE TRANSFER OF BENZENE
Benzene Emission
Operation Factor (g/gal)
Inland barge 0.76
Tank truck 1.8
Rail car . 1.8
Average 1.45
Source: Compiled by Mara and Lee (1978)
63
-------
transfer areas have controls. If we assume that 672,000 mt of EDC are
transported each year, that emissions might occur during loading and
unloading, and that the emission factor is less than 0.18%, an extreme
upper emission estimate of 2,400 mt/yr results.
Waste Disposal
EDC wastes may be generated in any process in which the chemical is
involved. The largest quantities of EDC wastes occur during the
synthesis of the compound and in the production of VCM, processes that
involve all of the EDC production and most of its consumption. The
treatment of wastes is of concern in estimating atmospheric emissions
because of EDC's moderately high volatility (see Section III).
Liquid wastes result from scrubbing vented gases or crude EDC with
water or caustic solutions. The treatment of these wastes varies from
plant to plant, but usually this wastewater is used for pH control in
other processing areas or is sent off site for final processing (EPA,
1974).
Solid wastes are usually disposed of by burial in a landfill or by
incineration (Patterson, 1975). Solid wastes also occur in EDC
manufacturing plants that use fluidized bed, rather than fixed bed,
reactors (EPA, 1974). These solid wastes are periodically removed from
the rejected water settling ponds and transported to landfills.
Monsanto Research Corporation (1975) has estimated EDC emission
factors emitted to solid waste and water for EDC formulation. We have
used Monsanto's conclusions, shown in Table VIII-2, to estimate that
during 1977, 10,600 mt were discharged to solid wastes and that 18,500
mt were discharged to water. Additional solid waste and water
discharges occur as a result of production of chemicals that used EDC as
a feedstock.
64
-------
Table VHI-2
ESTIMATED 1977 EDO EMISSIONS AS SOLID WASTE
AND TO WATER FROM EDC PRODUCTION
Solid Waste' Water
Emission Factor3 (kg/mt)
Direct chlorination 1.5 2.9
Oxychlorination 2.8 4.6
Emissionsb (1,000 mt/yr)
Direct chlorination 4.5 8.5
Oxychlorination 6.1 10.0
Total Emissions 10.6 18.5
aBased on Monsanto (1975).
bAssumes 58% direct chlorination and 42% Oxychlorination (Patterson,
1976) and an EDC production of 5,194,000 mt/yr.
Estimating atmospheric exposures from the solid waste and water EDC
emissions would require:
(1) An estimate of the rate of return to the atmosphere
(2) Identification of the contaminated sites, the amount deposited
at each site, and the site's location in respect to population
(3) A method for transforming emission estimates to estimates of
atmospheric exposure.
Currently, the data available on items (1) and (2) are insufficient for
estimating exposures. Table II-2 indicates that an estimated 52,000 mt
of EDC is emitted directly to the atmosphere annually from sources other
than transportation and waste disposal. If all the EDC solid waste and
water emissons evaporate to the atmosphere, an additional 29,100 mt/yr
of atmospheric emissions would result. This estimate, however, is
uncertain since the rate of evaporation from landfills and water
emissions cannot be adequately assessed.
65
-------
BIBLIOGRAPHY
Altshuller, A. P., "Lifetime of Organic Molecules in the Troposphere and
Lower Stratosphere," Environ. Sci. Tech. (to be published).
Applied Urbanetics, Inc., "Market Share Study," FEA Contract No.
CO-06-60435 (1976).
Arthur D. Little, Inc., "Self-Serve Market Shares in Four Metropolitan
Areas," memo to Richard J. Johnson, EPA, from E. Quakenbush and P. E.
Mawn (June 1977).
Auerbach Associates, "Miscellaneous and Small-Volume Consumption of
Ethylene Dichloride," unpublished report prepared for EPA under
Contract EPA-68-01-3899 and Auerbach Associates, Inc.,
AA1-2431-104-TN-1 (1978).
Battelie-Columbus Laboratories, letter to Richard J. Johnson, EPA, from
C. W. Townley concerning "Results of Self-Service Exposure Samples"
(May 1977).
Bureau .of the Census, Statistical Abstract of the United States (1975).
Bureau of Economic Analysis, "Projections of Economic Activity for Air
Quality Control Regulations," NTIS PB-259-870 (1973).
Chemical Marketing Reporter, "Profile: Ethylene Dichloride," 212(3):9
(1977).
66
-------
Billing, W., N. Tefertiller, and G. Kallos, "Evaporaton Rates and
Reactivities of Methylene Chloride, Chloroform,
1,1,1-Trichloroethane, Trichloroethylene, Tetrachloroethylene, and
Other Chlorinated Compounds in Dilute Aqueous Solutions," Environ.
Sci. Technol. 9:833-837 (1975).
Drury, J. S., and A. S. Hammons, "Investigations of Selected
Environmental Pollutants: 1,2-Dichloroet.hane," Oak Ridge National
Laboratory, Draft Report prepared for the Office of Toxic Substances,
EPA (1978).
Ethyl Corporation, "Yearly Report of Gasoline Sales by States" (1976).
Federal Register, "Consolidation of Hazardous Material Regulations,"
41(74):l5972-15990 (1976).
Gifford, F. A., and S. R. Hanna, "Technical Note: Monitoring Urban Air
Pollution," in Atmospheric Environment, Vol. 7, Pergamon Press (1973)
Grimsrud, E., and R. Rasmussen, "Survey and Analysis of Halocarbons in
the Atmosphere by Gas Chromatography - Mass Spectrometry," Atmos.
Environ. (England), 2:1014-1017 (1975).
Hanst, P., "Noxious Trace Gases in the Air," Chemistry 51(2);6-12 (1978)
Hardie, D., "Chlorocarbons and Chlorohydrocarbons," in Kirk-Othmer
Encyclopedia of Chemical Technology, 2nd ed., Vol. 5, pp. 171-178,
Interscience, New York (1964).
Hawley, G.G. (ed.), The Condensed Chemical Dictionary, 9th ed., Van
Nostrand Reinhold Company, New York (1977).
Horvath, R., "Microbial Co-metabolism and the Degradation of Organic
Compounds in Nature," Bacteriol. Rev. 36(2);146-155 (1972).
67
-------
Johns, R., "Air Pollution Assessment of Ethylene Dichloride," MTR-7164,
The Mitre Corporation (1976).
Kittredge, G. D., memo to files concerning "Up-to-date Estimate of
Automotive Emission Factors," (26 September 1977).
Mabey, W. R., Physical Organic Chemistry Laboratory, SRI International,
personal communications (December 1978).
Mara, S. J., and S. S. Lee, "Assessment of Human Exposures to
Atmospheric Benzene," SRI International (1978).
Mascone, D., EPA, personnel communications (16 November 1978).
i
McConnell, G., D. Ferguson, and C. Pearson, "Chlorinated Hyrdocarbons
and the Environment," Endeavor 34;13-18 (1975).
Mitten, M., K. Dress, W. Krochta, F. Ewald, and D. DeWitt,
"Chlorocarbons," in Encyclopedia of Industrial Chemical Analysis, Vol.
9, F. D. Snell and L. S. Ettre, eds., pp. 437-510, Interscience, New
York (1970).
Monsanto Research Corporaton,, "Potential Pollutants from Petrochemical
Processes," as cited in Drury and Hammons (q.v.).
National Institute for Occupational Safety and Health, "Criteria for a
Recommended Standard Occupational Exposure to Ethylene Dichloride,"
NIOSH-76-139 (1976).
, Registry of Toxic Effects of Chemical Substances, Vol. II,
p.388 (1977).
_, "Current Intelligence Bulletin #25: Ethylene Dichloride
(1,2-dichloroethane)" (19 April 1978).
68
-------
Patterson, R. M., M. I. Bornstein, and E. Garshick, "Assessment of
Ethylene Dichloride as a Potential Air Pollution Problem," GCA
Corporation (1976).
Pearson, C., and G. McConnelL, "Chlorinated C and C~ Hydrocarbons
in the Marine Environment," Proc. R. Soc., London, Ser. B,
189:305-332 (1975).
PEDCo, "Draft of Preliminary Monitoring Data," collected for EPA (1978).
Pellizzari, E. D., "Electron Capture Detection in Gas Chromatography,"
J. Chromat. 99;3-12 (1974).
Radding, S., D. Liu, H. Johnson, and T. Mill, "Review of the
Environmental Fate of Selected Chemicals, SRI International,
EPA-560/5-77-003, Office of Toxic Substances, EPA (1977).
Rothon, R. N., "Petroleum and Organic Chemicals," in Chemical
Technology: An Encyclopedic Treatment, Vol. 4, Barnes and Noble, New
""" York "01972)7 "'
Schewe, G. J., EPA, memos concerning "Estimates of the Impact of Benzene
from Automotive Sources" to R. J. Johnson (20 June, 9 August, 12
August 1977).
Singh, H., L. Salas, and L. Cavanagh, "Distribution, Sources, and Sinks
of Atmospheric Ualogenated Compounds," J. Air Pollut. Control Assoc.
27(4);332-336 (1977).
Snelson, A., R. Butler, and F. Jarke, "Study of Removal Processes for
Halogenated Air Pollutants," U.S. Environmental Protection Agency,
EPA-600/3-78-058 (1978).
Spense, J., and P. Hanst, "Oxidation of Chlorinated Ethanes," J. Air
Pollut. Control Assoc. 28(3):250-253 (1978).
69
-------
Stolpman, P., EPA, personal communication (October 1977).
Storck, W., "Big Chemical Producers Post Moderate Growth," Chem Eng.
News 56(18);31-37 (1978).
Suta, B. E., "BESTPOP: A Fine-Grained Computer System for the Assessment
of Residential Population," SRI International (1978).
Toxic Materials News, "NCI Finds Ethylene Dichloride to be Carcinogenic"
(27 September 1978).
U.S. Environmental Protection Agency, "Mixing Heights, Wind Speeds, and
Potential for Urban Air Pollution Throughout the Contiguous United
States," in Publication No. AP-101 (1972).
, "Engineering and Cost Study of Air Polluton Control for the
Petrochemical Industry," Vol. 3, "Ethylene Dichloride Manufactured by
Oxychlorination," EPA-450/3-73-006-C (1974).
, "Report on the Problem of Halogenated Air Pollutants and
Stratospheric Ozone," EPA-600/9-75-008 (1975).
, "Users' Manual for Single-Source (CRSTER) Model,"
EPA-450/2-77-013 (1977).
, draft material relating to human exposure to atmospheric
ethylene dichloride near production facilities (1978).
U.S. Department of Transportation, Federal Highway Administration,
"Highway Statistics" (1974a).
_, "Motor Vehicle Registrations by Standard Metropolitan
Statistical Areas," Table MV-21 (1974b).
70
-------
U.S. International Trade Commission, "Synthetic Organic Chemicals, U.S.
Production and Sales" (1973-1977).
Whitney, W., "Fumigation Hazards as Related to the Physical, Chemical,
and Biological Properties of Fumigants," Pest Control 29(7);16-21
(1961).
Windholtz, M., The Merck- Index, 9th ed., Merck and Company, Rahway, New
Jersey (1976).
Youngblood, P. L., EPA, memo concerning "Use of Dispersion Calculations
in Determining Population Exposures to Benzene from Chemical Plants"
to R. Johnson (20 September 1977).
Youngblood, P. L., "Dispersion Modeling for Determining Population
Exposure to Benzene," EPA memo to R. Johnson (4 January 1978).
71
------- |