EPA-450/3-77-029
October 1977
ATMOSPHERIC
BENZENE EMISSIONS
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711

-------
TECHNICAL REPORT DATA
(l'l< n\i rnhl htUJucHons *>n the reverse bi jore eomplctmy}
; •< i ft in N( •
1-:PA -4 30/ 3 - 7 7-029
! I I ; 1.1. A N I) S u II 7 I 7 L L
Data Base of Atmospheric Benzene Emissions
Auinom;))
T.M• Briggs, J.T. Bertke, and D.W. Augenstein
3. RE
f$fTSicris8w-
5 RE PORT OATE
Date of Issue:August 1977
6. PERFORMING ORGANIZATION CODE
8 PERrOHMING ORGANIZATION REPORT NO.
3264-A
9 PERFORMING ORG M\ll Z A TION NAME AND ADDRESS
PEDCo Environmental, Inc.
11499 Chester Road
Cincinnati, Ohio 45246
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-2515, T.0. 1
12 SPONSORING AGF NCY NAM I AND ADDRESS
Pollutant Strategies Branch
Strategies & Air Standards Division
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA 200/04
27711
15 SUPPL(:MF NTAMY NOT i S
SASD Project Officer: Richard Johnson
16 AHSTHACI
All major potential sources of atmospheric benzene emissions were
reviewed and evaluated to develop a comprehensive national inventory.
Emission sources are categorized as mobile, stationary and natural.
Gasoline-powered motor vehicles represent by far the largest overall
source. The mere diverse stationary sources are categorized by produc-
tion, consumption and storage and transport. Maleic anhydride productior
represents the largest known source in this category. Available process
control techniques are also described and, whenever possible, the best
control technology for specific sources is identified. A projected
emission inventory for the base year 1985 has been developed from avail-
able emission factor data and market projections for all of the major
source categories.
Di scniP t oh:;
KL v wonus AND OOCUMf NT ANALYSIS
klOr.N m IERS/OPI N I NOl D TERMS
Air Pollution, Air Quality Data,
Emission Inventory, Emission Factors
Control Technology
Benzene
Atmospheric Emissions
,Air Pollution Control
Air Quality Projec-
tions
Organic Compounds
. COSATI I lclil/<
-------
EPA-450/3-77-029
ATMOSPHERIC
BENZENE EMISSIONS
by
PEDCo Environmental, Inc.
(Chester Towers
II4')9 C.hester Road
Cincinnati. Ohio 452 Ki
Contract No. f>ii-02-2!>I>r>
Task No. I
EPA Project Officer: Justice A. Manning
EPA Task Officer: Hichard Johnson
Prepared for
l .S. ENYTHONYIENTAI. PROTECTION AGENCY
Office of Air and Waste Vlunagement
Office of Air Quality Planning and Standards
Research Triangle Park. INorlli Carolina 27711
()i lobi'f IV77
-------
This report was furnished to the U.S. Environmental Protection
Agency by PEDCo Environmental, Inc., Cincinnati, Ohio, in
fulfillment of Contract No. 68-02-2515, Task No. 1. The contents
are reproduced herein as received from the contractor. The
opinions, findings, and conclusions expressed are those of the
author and not necessarily those of the U.S. Environmental
Protection Agency. Mention of company or product names does
not constitute endorsement by the Environmental Protection Agency.
Copies of this report are available free; of charge as supplies
permit - to Federal employees, current contractors and grantees,
and non-profit organizations from the Library Services Office
(MD~35), U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina 27711, or may be purchased from the National
Technical Information Service, 5285 Port Royal Road, Springfield,
Virginia 22161.
Publication No. EPA-450/3-77-029
-------
ACKNOWLEDGMENT
This report was furnished to the U.S. Environmental
Protection Agency by PEDCo Environmental, Inc., Cincinnati,
Ohio. Terrence Briggs was the PEDCo Project Manager and
Richard Gerstlo functioned as Service Director. Principal
authors of the report were J. Thomas Bertke, Terrence
Briggs, Leslie Ungars, and David Augenstein.
Richard Johnson was the Task Officer for the U.S.
Environmental Protection Agency. The authors appreciate the
contributions made to this study by Mr. Johnson and other
EPA personnel.
-------
TABLE OF CONTENTS
Page
1.0 SUMMARY	1-1
2.0 INTRODUCTION	2-1
3.0 CHEMICAL AND PHYSICAL FACTORS AFFECTING	3-1
ATMOSPHERIC LOADING
3.1	Natural Sources	3-1
3.2	Atmospheric Reactions	3-1
3.3	Atmospheric Factors	3-3
4.0 PRIMARY MAN-MADE BENZENE EMISSION SOURCES &	4-1
LEVELS
4.1	Mobile Sources	4-1
4.2	Stationary Sources	4-19
4.3	Benzene Emission Source Characteristics	4-69
5.0 CONTROL METHODS	5-1
5.1	Mobile Sources	5-1
5.2	Stationary Sources	5-11
5.3	Costs of Control	5-32
6.0 EMISSION TRENDS	6-1
6.1	Mobile Sources	6-1
6.2	Stationary Sources	6-6
REFERENCES
APPENDIX A	A-1
APPENDIX B	B-l
APPENDIX C	C-l
APPENDIX D	D-l
APPENDIX V,	E-l
i.v
-------
LIST OF FIGURES
No.	Page
4-1 Location of United States Oil Refineries	4-20
4-2 Geographical Location of Petroleum-Based	4-21
Benzene Production Plants
4-3 Processing Plan for Typical Intermediate	4-22
Refinery
4-4 Processing Plan for Typical Complete Refinery 4-22
4-5 Geographical Location of Coal-Derived Benzene 4-26
Production Plants
4-6 Schematic Diagram of By-product Coke Oven	4-29
Showing Possible Atmospheric Emission Sources
for Benzene
4-7 Flow Chart for Benzene Derivatives and Their	4-34
Uses
4-8 Geographical Location of Benzene-Consumption	4-35
Plants
4-9 The Gasoline Marketing Distribution System	4-57
in the United States
4-10	Vcipor and Liquid Flow in a Typical Bulk	4-58
Terminal
5-1	Operation of a Dual-Bed Catalyst	5-3
5-2 Closed Positive Crankcase Ventilation System	5-8
5-3 Adsorption-Regeneration Evaporative Emissions 5-10
Control System
5-4 Vapor Pressure of Liquid Benzene at Various	5-16
T€;mperatures
-------
LIST OF FIGURES (continued).
No.	Page
5-5 Two-Unit Fixed-Bed Adsorber	5-22
5-6 Vertical Adsorber with Two Cones	5-22
5-7 Adsorption of Benzene on Activated Coconut	5-24
Charcoal at Atmospheric Pressure, Out-Gassed
at 550°C
5-8 Small-Capacity Vaporsavor Gasoline Absorption 5-28
Unit
5-9 Carbon Adsorption Capital Costs	5-34
5-10 Carbon Adsorption Capital Costs - Small Units 5-35
5-11 Carbon Adsorption Unit Operating and Main-	5-37
tenance Cost vs. acfm and Hydrocarbon
Concentration
5-12 Catalytic Incinerator Prices	5-38
5-13 Catalytic Incinerator Operating and Maintenance 5-39
Cost vs. acfm and Hydrocarbon Concentration
5-14 Prices for Thermal Incinerators with Heat	5-40
Exchangers
5-15 Prices for Thermal Incinerators without Heat	5-41
Exchangers
5-16	Thermal Incinerator Operating and Maintenance 5-4 3
Cost vs. acfm and Hydrocarbon Concentration
6-1	Historical and Projected Domestic Aniline	6-8
Production and Corresponding Benzene Consump-
tion
6-2 Historical and Projected Domestic Ethlylbenzene 6-10
Production and Corresponding Benzene Consump-
tion
6-3 Historical and Projected Domestic Maleic	6-11
Anhydride Production and Corresponding Benzene
Consumption
vi
-------
LIST OF FIGURES (continued).
No.	Page
6-4 Historical and Projected Domestic Cyclohexane 6-13
Production and Corresponding Benzene Consump-
tion
6-5 Historical and Projected Domestic Cumene	6-14
Production and Corresponding Benzene Consump-
tion
vi i
-------
LIST OF TABLES
No.	Page
1-1 Estimate of 1976 Benzene Emissions	1-2
1-2 Estimated Benzene Emissions from Production	1-4
of Chemicals
1-3 Projections of 1985 Benzene Emissions	1-5
3-1 Significant Properties of Benzene	3-2
3-2	Atmospheric Reactivities of Benzene and Other 3-4
Hydrocarbon Compounds
4-1	Estimated National Benzene Emission	4-2
Inventory - 1976
4-2 Summary of Benzene Emissions for Two Mobile	4-3
Source Categories, 1976
4-3 Seasonal Benzene Concentrations in Different	4-5
Brands and Grades of Gasoline
4-4 Benzene Content of Gasoline Blending Components 4-6
4-5 Benzene Levels in U.S. Gulf Oil Gasolines,	4-8
October 1976
4-6 Automobile Emission Factors	4-14
4-7 Benzene and IIC Emissions from Diesel Engine	4-17
Under Varying Operating Conditions
4-0 Exposures of Workers to Benzene in a Coal-	4-31
Derived Benzene Production Plant
4-9 Annual (.1976) Production Capacity for Some	4-36
Major Chemical Compounds in Which Benzene
is Used as an Intermediate
vi i i
-------
LIST OF TABLES (CONTINUED).
No.	Page
4-10 Estimated Benzene Emissions from Gasoline	4-60
Transportation Operations
4-11 Inventory of Service Station Benzene Emissions 4-62
4-12 Summary of Benene Emission Factors and Annual 4-65
Emissions from Storage/Distribution of
Gcisoline and Benzene
4-13 Stationary Source Emission Characteristics -	4-67
19 76 Basis
4-14 Other Air Pollutants Emitted from Benzene	4-68
Sources
4-15	Estimated Emissions from an Average Facility	4-70
Using Benzene as an Intermediate
5-1	Estimated Usage of Control Devices on Cars	5-9
iri Operation in 1976
5-2	Control Methods for Major Point Sources of	5-12
Atmospheric Benzene
6-1	Summary of Projected Benzene Emissions from	6-2
Mobile Sources in 1985
6-2 Projected Benzene Emissions from Gasoline-	6-3
Powered Motor Vehicles in 1985
ix
-------
1. 0 EXECUTIVE SUMMARY
Documented data on atmospheric benzene emissions were
reviewed and evaluated. In general, these data were rela-
tively sparse for most facets of this study.
Because benzene is highly toxic and widespread in the
atmosphere, it has been listed as a hazardous pollutant
under Section 112 of the Clean Air Act. Available data
indicate that benzene is minimally reactive photochemically;
thus it must be assumed that benzene is essentially unchanged
chemically in the atmosphere before reaching receptor sites.
Benzene is emitted to the atmosphere in large quanti-
ties, principally from automotive and industrial sources.
Benzene is not only a significant component of motor gasoline,
but also is used as an intermediate in numerous chemical
processes and is a by-product of others. In 1976, it ranked
thirteenth among chemicals produced in the United States,
with an output of 10.6 billion pounds.
It is estimated that benzene emissions from known
sources exceeded 540 million pounds in 1976. Table 1-1
lists by category the primary sources and their estimated
emissions. Although these figures clearly point to gaso-
line-powered motor vehicles as the largest total emitter,
dispersion analysis of emissions data indicates that some
stationary sources may be causing the highest levels of
population exposure.
1-1
-------
Tabic L - 1. ESTIMATE OF .19 76 BENZENE EMISSIONS
Emission estimates,
Source	million lb/yr
Gasoline-powered motor vehicles	443.6
Petroleum refineries	4.1
Coke-oven operations	7.8
Known processes using benzene	60. 0a
as an intermediate
Storage and distribution of	24.8
gasoline and benzene
Solvent operations	b
Other miscellaneous	4.0
Total known losses	544.3
Does not include aniline or cyclohexane production for
which no emissions estimates were obtainable.
Unknown.
Mobile sources emit benzene from engine exhausts and by
evaporative losses. The emission rate is a direct function
of the controls applied and of the benzene content of
gasoline. Older automobiles (1975 and earlier models) emit
much larger quantities of benzene than newer model catalyst-
equipped cars. Estimates of the extreme upper and lower
limits of auto emissions (518.5 million lb/yr and 166 million
lb/yr) are crude because benzene emissions data on cars
without exhaust controls are very limited. Further, it is
difficult to estimate the benzene level in gasoline (a
critical factor in estimating auto emissions) because that
level varies greatly by geographical region, sources of
crude oil, and season of the year.
1-2
-------
Stationary sources of benzene emissions are diverse.
Thus, these sources have been grouped into three categories:
benzene production, other chemical manufacturing, and
storage and distribution of benzene and gasoline. Petroleum
refineries and coke-oven operations are the major benzene
production sources; however, estimates of the magnitude of
emissions from these sources are unreliable because of lack
of documented emissions data. The generally poor vapor
containment in coke-oven operations indicates that benzene
emissions from this source may have a significant impact on
local ambient concentrations. Although petroleum refineries
provide far better containment, they may be a significant
source of benzene emissions because of their size and
number.
Maleic anhydride production, which uses benzene as the
raw material, not only is probably the largest source of
benzene emissions in the source category of other chemical
production, but also is probably the largest stationary
source. Table 1-2 lists the major chemical products in
which benzene is used, with estimates of emissions from
these sources during 1976.
In the subcategory of storage and distribution emis-
sions (primarily gasoline evaporation), service stations
appear to be the most important source, particularly
because of the large number of stations in operation and
their widespread distribution nationwide. The importance of
this source is increasing with the trend toward self-service
operations, which could lead to greater exposure of the
general population. Again, the benzene content of the
gasoline is an important variable affecting emissions
from this source.
1-3
-------
Table 1-2. ESTIMATED BENZENE EMISSIONS FROM
PRODUCTION OF CHEMICALS
(million lb/yr)
Product
1976
Nitrobenzene
7.5
Aniline
a
Ethylbenzene
2.7
Styrene
6.5
Maleic Anhydride
34. 8
Cyclohexane
a
Cumene
0.5
Phenol
1.7
Chlorobenzenes
5.3
Detergent alkylates
CO
•
o
Other nonfuel uses
0.7
Total
O
•
o
a Unknown.
Total atmospheric benzene emissions in 1976 are esti-
mated to have been 544.3 million pounds. They are projected
to decrease to 232.1 million pounds in 1985 (see Table
1-3). Essentially all of this decrease results from the
phasing out of vehicles without exhaust controls (from 443.6
million pounds in 1976 to 107. 1 in 1985) . Maleic anhydride
production is expected to remain the major stationary
source. It is likely that the use of benzene in solvent
operations will be eliminated by 1985 as a result of com-
pliance by industry with the Occupational Safety and Health
Administration's proposed regulations for benzene use; thus,
any benzene emissions from solvent operations will arise
primarily from benzene impurities in other solvents. The
1-4
-------
emission factors used to project the 1985 estimates of
emissions from stationary sources (Table 1-3) are based on
1976 market trend data in the respective industries.
In oenzene production facilities and facilities that
use benzene as a chemical intermediate, the recovery of
vapors containing organic compounds is frequently practiced
for economic reasons and therefore is considered a part of
the process in addition to being a means of controlling air
pollution. When required, more stringent control of benzene
vapors can be achieved by incineration, adsorption, or other
techniques.
Table 1-3. PROJECTIONS OF 1985 BENZENE EMISSIONS
Source
Emission estimate,
million lb/yr
Increase over
1976 estimate,
o
*6
Gasoline-powered
vehicles
107.1
-76
Petroleum refineries
6.5
+4 2
Coke-oven operations
9.3
+ 19
Benzene consumption
processes
79. 9
+ 33
Storage and distri-
bution of gasoline
and benzene
25.3
+ 2
Other miscellaneous
4.0
+ 3
Total
232.1

1-5
-------
2.0 INTRODUCTION
The purpose of this report is to establish an informa-
tion base concerning sources of atmospheric benzene. Ben-
zene is used widely throughout the United States in large
volumes and in diverse applications. It is a clear, color-
less, extremely flammable liquid that is highly toxic. For
these reasons benzene has been classified as a hazardous
pollutant under Section 112 of the Clean Air Act.
Benzene is a by-product of petroleum and coal. It is
used as a component of gasoline, as a solvent, and as an
intermediate in the production of numerous chemicals,
including ethylbenzene, cumene, cyclohexane, nitrobenzene,
maleic anhydride, chlorobenzene, and detergent alkylate.
Benzene is released to the atmosphere from a variety of
mobile and stationary sources; principal emitters are
gasoline-powered combustion engines, gasoline storage and
distribution facilities, petroleum refineries, by-product
coke ovens, and chemical and industrial operations that
consume or handle benzene. No information is available
concerning naturally occurring benzene emissions.
Information for this study was obtained from government
publications (largely those of the U.S. EPA and NIOSH),
technical journals, industry publications and documents,
data from state agencies, PEDCo technical files, and per-
sonal contacts, all of which are cited as references. No
data were generated by emissions tests or by tour of in-
dustrial facilities.
2-1
-------
The scope of the study and the structure of this report
are summarized as follows:
Evaluation of the chemical and physical factors affec-
ting atmospheric benzene concentrations (Section 3).
Identification of the known primary sources of atmos-
pheric benzene; with estimates of the benzene emissions
from these sources (Section 4).
Review of the currently available techniques for
control of benzene emissions from the major mobile and
stationary sources (Section 5).
Projection of benzene production and consumption to the
year 1985, with estimates of the 1985 emissions from
the major sources (Section 6).
2-2
-------
3.0 CHEMICAL AND PHYSICAL FACTORS AFFECTING
ATMOSPHERIC LOADING
The atmospheric chemistry of benzene was reviewed to
evaluate the possible impact of benzene emissions on ambient
levels of photochemical oxidants and to identify possible
secondary pollutants formed as a direct or indirect result
of benzene oxidation. Benzene is not a normal atmospheric
reaction product, therefore, the review was limited to
consideration of the degradation of primary benzene emis-
sions .
Factors affecting the atmospheric dispersion of benzene
from major source categories were evaluated to identify
atmospheric loading characteristics.
3.1	NATURAL SOURCES
No literature was found to indicate the likely major
natural sources of benzene or their contribution to the
total environmental benzene inventory. Benzene appears to
be a relatively minor constituent of the total environmental
burden of hydrocarbons from natural sources.
3.2	PHYSICAL AND CHEMICAL PROPERTIES OF BENZENE
Benzene is a clear, colorless, extremely flammable
liquid with a distinct odor. It is sometimes called benzol,
phenyl hydride, coal naphtha, phene, benzole, or cyclohexa-
triene. Table 3-1 presents selected chemical and physical
properties of this compound.
3-1
-------
i n-5 1 9 *"}
Table 3-1. SIGNIFICANT PROPERTIES OF BENZENE ''
Formula
Chemical: Structural: (^7);
Molecular weight
78.11
Melting point
5.51°C
Boiler point
80. 093° to 80. 094°C
Flash point
-11°C (closed cup)
Specific gravity
0.9894 at 20°C
Autoignition
temperature
5 38 °C
Vapor pressure
100 mm Hg at 26.1°C
Vapor density
27 7 g/ml
Explosive limits
1.4 to 6.8% by volume
Solubility
Very soluble in most organic
solvents; slightly soluble in
water (700 ppm at 20°C)
3.3 ATMOSPHERIC REACTIONS
Lack of experimental data prohibits a comprehensive
evaluation of benzene chemistry. Because of the low rela-
tive reactivity of benzene, gaseous reactions at ambient
conditions have not received the attention given to benzene
solution-phase chemistry. Very limited quantitative data
are available concerning benzene reaction rates and degra-
^ 4. ¦	, , 24,156 '
dation products.
The relative reactivity of benzene and other hydro-
carbons can be evaluated in several ways. Analysis of the
influence of benzene on the rate of oxidation of nitric
oxide (NO) to nitrogen dioxide	most common.
Since the generation of ozone (0..) is proportional to the
24
NO^/NO ratio, this index provides an approximate measure
3-2
-------
of the ozone formation potential of hydrocarbons. Chemical
reactivity also is measured by the relative rates of reac-
tion with important oxidizing species such as hydroxyl
3
radicals (OH), atomic oxygen [0( P) ] , and ozone. This index
provides an approximate rate of formation of secondary
degradation products. Data on reactivity of benzene and
other hydrocarbon compounds (presented in Table 3-2) indi-
cate that benzene is substantially less reactive than other
common citmospheric hydrocarbons. For example, reactivity of
benzene is one-tenth that of propylene and one-third that of
n-hexane.
The reactivity relationships of benzene with alkyl-
benzene compounds are consistent with electrophilic attack
on the aromatic ring. The xylenes, trimethyl-benzenes, and
tetramethyl-benzenes have reactivities one order of magni-
tude larger than that of unsubstituted benzene. The low
apparent reactivity of benzene compared with that of all
olefins, most aromatics, and most paraffins suggests that
benzene does not participate significantly in photochemical
oxidant formation in the immediate vicinity of the source.
It is conceivable, however, that benzene could react during
long-range transport of urban air masses. Data are in-
adequate: for evaluation of the role of benzene relative to
naturally generated unsaturated hydrocarbons and paraffins
within the urban air mass.
The products of benzene reactions depend primarily on
the type's of oxidation agents available, since benzene does
not absorb ultraviolet light in the ambient spectrum (Refer-
ence 23 gives benzene absorption-band data). The most
important oxidizing species could consist of hydroxy!
radicals, atomic oxygen, excited molecular oxygen, and
ozone. Although the benzene decay paths and ultimate reac-
3-3
-------
Table 3-2. ATMOSPHERIC REACTIVITIES OF BENZENE
AND OTHER HYDROCARBON COMPOUNDS9
Compound
Based on NO

oxidation
OH
01062,124,156
<0.05
0.2562'156
N.D.
0.5062'156
N.D.
0.5062'156
N.D.
1.0062'156
o
o
r-i
0. 34 62
N.D.
1.5062,156
N.D.
1.OO62
N.D.
Based on

O(^P)
Aromatics
Benzene
Toluene
p-xylene
o-xylene
m-xylene
Ethylbenzene
1,3,5 Trimethylbenzene
1,2,3,4 Tetramethyl-
benzene
Paraffins
Methane
Propane
n-Butane
n-Hexane
Olefins
Ethylene
Propylene
Isobutene
Trans-2-butene
0.10
0. 40
0.75
0.93
62
62
62
62
0. 50
62,156
, (basis)
1. 00
1. 0 0 6 2, 124 , 156
35062,124,156
.124
>
b
,156
156
<0.001
,156
0.08
0. 20
N.D.
156
<0.01
0.03
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
124,156
156
156
0.001
,156
0.04
0.05
N.D.
156
0.2124'156 i 0.2124'156
i.oo(basls)l i.oo(basis)
3. 00
4.10
124,156!
124,156
4. 50
| 5.50
124,156
124,156
Values shown represent an approximate averaqe of data in the
references cited. The references are summary documents with
citations to the original experimental work.
N.D. - No data.
3-4
-------
tion products have not been determined, the products prob-
ably include formaldehyde, peroxides, and epoxides.
3.4 ATMOSPHERIC TRANSPORT
In analysis of the dispersion of benzene in the atmos-
phere after being released from a source, it is assumed that
benzene disperses in much the same manner as an inert gas;
i.e., it neither has any appreciable weight nor adds any
buoyancy to the dispersing medium. Thus, the primary
consideration should be the type of source from which the
155
benzene is released and how it affects the plume rise.
Gasoline-powered automobiles account for about 82
percent of the total nationwide emissions. However, this is
not to say that approximately 8 2 percent of the measured,
ambient benzene in any one area results from automobile
emissions. Varying source characteristics and local meteoro-
logical conditions are the determining factors governing the
benzene source-receptor relationships. The source-receptor
relationships for three types of benzene sources [auto-
mobiles, point sources (e.g. maleic anhydride plants), and
gasoline: distribution J are discussed in the following
paragraphs.
When considered collectively on a nationwide basis,
automobiles constitute the largest source of benzene emis-
sions; however, individually they are probably the smallest
of the benzene emitters. It would be difficult to relate
benzene emissions from individual automobiles to ambient air
quality; therefore, the analysis is approached by consider-
ing total emissions from a length of highway. The emission
rate then depends on the number of vehicles traveling along
the designated length at any one time and the number of
lanes in the highway.
-------
Ground-level concentrations of a gaseous pollutant
(benzene) downwind of a highway depend on the orientation of
the highway with respect to the wind direction. For example,
downwind concentrations of benzene will be considerably less
when the wind is moving perpendicular to the highway than
when it is moving parallel to the highway. Because of their
low emission height and horizontal exit velocities, benzene
emissions from automobiles have greater impact close to the
roadway than would emissions from an elevated stack.
Benzene emissions from maleic anhydride plants exhibit
a different dispersion pattern than automobile emissions
because the characteristics of the two sources differ, e.g.
the height at which the emissions are released. Typical
stack heights at maleic anhydride process plants are 60 to
100 feet, whereas automobile exhausts are essentially at
ground level. Vertical exit velocities are generally much
higher at a process plant, and the greater plume rise causes
the pollutants to disperse to a greater degree. Because the
benzene is emitted at a greater height, its maximum ground-
level impact occurs further downwind from the source than
that of emissions from an automobile.
Leaks in pump seals and valves are another source of
benzene emissions within a maleic anhydride plant (and most
other benzene-consuming processes). Since they have no
appreciable vertical exit velocities or elevated tempera-
tures, these sources are termed fugitive sources. They
exert their greatest impact close to the source, probably
within the plant boundaries. Collectively, the magnitude of
fugitive source emissions probably approaches that of the
process stack emissions and should be seriously considered
in an analysis of the benzene problem.
3-6
-------
Characteristics of the gasoline distribution sources,
e.g. service stations, are very similar to those of auto-
mobiles when considered individually. Emissions occur
essentially at ground level, and their exit velocities are
very low. Unlike those from automobiles, however, emissions
from gasoline distribution exit near-ambient temperatures
and they occur at a stationary point. Benzene emissions
from gasoline distribution have their greatest impact very
close to the source and thus affect most strongly the
persons working at these facilities.
Diurnal patterns also influence dispersion of emissions
from the various sources. The process sources operate at a
more or less constant level throughout the day, whereas
automobile emissions undergo a definite diurnal cycle,
118
peaking at the two rush-hour periods.	The diurnal
pattern of emissions from gasoline distribution sources may
closely resemble that of automobile emissions, at least the
portion of the distribution emissions that occurs during
filling of automobiles.
3-7
-------
4.0 PRIMARY BENZENE EMISSION SOURCES AND LEVELS
In this section the known atmospheric benzene emissions
are identified and, where possible, emission estimates and
emission factors are presented. Sources are categorized as
mobile and stationary. Table 4-1 presents estimated emis-
sions from each source, identifies the report section de-
scribing the source and the estimated emission levels, and
indicates; the degree of accuracy of the emissions estimates.
The accuracy estimates are subjective and are based upon
PEDCo's evaluation of the available data.
4.1 mob::le sources
Gasoline-powered motor vehicles constitute the major
source of benzene emissions in the United States. Aircraft,
boats, off-road vehicles, construction equipment, motor-
cycles, and diesel-powered motor vehicles also emit benzene,
but to a lesser degree. Gasoline-powered vehicles pre-
dominate because of their great numbers nationwide and the
6 7
large quantities of gasoline they consume. Table 4-2
summarizes estimates of benzene emissions from both gaso-
line- and diesel-fueled motor vehicles in 1976. Because of
the lack of quantitative emission data, a range of emissions
and a "best estimate" are presented for gasoline-powered
motor vehicles.
4-1
-------
Table 4-1. ESTIMATED NATIONAL BENZENE EMISSION
INVENTORY - 1976
Source
Reference
sect i on
in report
Fmi ss on
est imatc,
106 lb/yr
I
1
Accuracy ^
of estimate'
Mob i 1 <• .'lou r ces
4 . 1


(latto) i n t" — 1 'ow<' red Motor Vehicles
4.1.1
44 1. f,n
I;
f> ]<•«;<• j llf-o vy - I'm t y Vr• f11 c ] <•?.
4.1.2
i . H'1
r
.vi«• r Mob. I' >nr'
4. \ . :>
b

t . j ~ i or..i r y .';oi, r c" s
A .


!'• n/< ri<* f'rod •. k:Lion
4 . 1' . J


»#¦ r ro refineries
4.2.1
4 . 1
c
be:; z(•:.«. fj j r i f j ca t j on
4.2.1
0 . 2
lj
Coke oven operations
A. 2. 2
7 . 8
e
H'.'n'/ciif Conrur.pt ion f'roccfl'-.cj:
4.2.1


N 11 roben zerte

7.5
c
An iline

b

Ethylbenzene

2.7
c
Styrene

6.5
R
Maleic anhydrjde-

34.8
A
Cyclohexano

h

Cumene

0 . '¦>
C
Pheno]

1 . 7
b
Ch lorobenzenes

5 . 3
r
Detergent alkylate

0. i
C.
Solvent operations

c

Other nonfuol usc«;

0.7
c
Storage and Transport
4.2.4


Gasoline storage

1 . 7
h
benzene storage

0.6
B
Gasol ine dir.tr ibut ion

3.8
B
Benzene d i s tri but. ion

2 . 0
h
Crude oil operations

2.2
r
Sorvicc: ttt.it ions

14.6
p
- Known loMMe*; -

'<44. J

Source tc'il d.itn utilized in r.Orul.i
* Unknown, r«-Uit ive)y nunoi source-.
U n k now i:, po f en t i n 1 1 y ma j o r muirri-.
j
ACciu ncy of e'lt iin.it'- rat i r: ¦ i :
i ng benzene <
•rri i *i.«; l oriK .

A - (iouci em i hb j on data available.
H - Fair •stimate, *100'.
C - Poor iIai a base, ordor-of-maqnitudo estimate.
rRepr^duceT^o^^^
4-2
-------
Table 4-2. SUMMARY OF BENZENE': EMISSIONS FOR TWO
MOBILE SOURCE CATEGORIES, 1976
(Gg/year)

Exhaust
emissions
Evaporative
losses
Total
emissions
Gasoline-powered motor
vehicles



Best estimate
169. 4
31.8
201.2
Upper limit
198.7
36. 5
235.2
Lower limit
43.5
31. 8
75.3
Diesel-powered heavy-
duty vehicles
1.7
N. A.

N.A. Not available.
Emissions of benzene from gasoline combustion by motor
vehicles are regulated indirectly by Federal standards that
limit total hydrocarbon (THC) emissions from tailpipes and
evaporative losses. The 1976 national (excluding California)
hydrocarbon exhaust emission standard (0.93 g/km) for light-
duty* passenger vehicles was lowered to 0.25 g/km for
1977/19713 . 113 > 26 The national (excluding California)
standard for light-duty trucks for 1976/1977 is 1.25 g/km.^"^'^
The 1977 evaporative emission regulation for light-duty
vehicles is 6 g/test by the SHED+ procedure.^
4.1.1 Emissions from Gasoline-Powered Motor Vehicles
This subsection presents a review of recent literature
relevant to benzene emissions from gasoline-powered motor
*	31
Vehicles rated at weighing 2.7 Mq (6000 lb) or less.
Sealed Housing Method for Evaporative Determinations.
4-3
-------
vehicles. These data form the basis for the benzene emission
estimates presented in Section 4.1.2.
In 1974, total U.S. consumption of gasoline by highway
3	8
vehicles was 365 Mm (964 x 10 gal). This figure includes
private and commercial, Federal, civilian, state, county,
61
and municipal usage. Highway vehicles account for greater
than 96 percent of the total gasoline consumption in the
United States. Automobiles are the major consumer, but the
6 7
category also includes light- and heavy-duty trucks.
Benzene Concentration in Gasoline
The benzene level in gasoline depends on numerous
factors including the source of the crude oil, the geographic
location of the source and the refiner, the grade of gaso-
line, the refinery operations, and the seasonal blends
produced by each refinery. A report by the National Insti-
tutes of Occupational Safety and Health (NIOSH) on analysis
of several different gasolines, based on 1976 data, shows an
average benzene concentration of 1.24 liquid volume (lv)
64
percent. This value was determined by weighting averages
of concentrations in the three grades of gasoline (premium,
regular, and unleaded) from four commercial-brand service
stations. Table 4-3 summarizes these data.
With the exception of alkylates and butanes, essen-
tially all products manufactured for blending into finished
motor gasolines contain benzene, in concentrations ranging
22
from 0.5 to 8 volume percent. Table 4-4 presents the
benzene contents of gasoline blending components produced at
two refineries. Both cautioned that their data on the
benzene content of these individual streams are limited and
22
should be considered as order-of-magnitude estimates.
4-4
-------
Table 4-3. SEASONAL BENZENE CONCENTRATIONS IN
DIFFERENT BRANDS AND GRADES OF GASOLINE39
Company
(typical
service
station)
Gasoline
grade
Volume of benzene in
bulk sample, %
Average
volume of
benzene, %
Summer
Winter
Tresler-Comet
Premium
1.11
1.10
1.11

Regular
1.21
1. 00
1.11

Unleaded
1.41
1. 60
1.51
Bonded
Regular
0. 88
0. 88
0.88

Unleaded
1. 19
1. 60
1. 40
Bonded
Regular
0. 88
0. 88
0. 88

Unleaded
1 . 20
1.60
1.40
Clark
Regular
0. 97
2.00
1. 49

Unleaded
1.09
1. 10
1.10
4-r;
-------
22
Table 4-4. BENZENE CONTENT OF GASOLINE BLENDING COMPONENTS

Ref inery
A
Refinery B
Components3
in complex
refinery gasoline
pool, volume %
Stream
Average volume
% in pool
1 Volume %
! benzene
|
Vo 1 urae %
benzene
Heavy reformate
30
; 1-7
0.5-1.5
29.0
Raff inate^

N. A.
0.9
2.1
Straight-run naphtha
10
i 1-3
0.5-2.0
4.6
(nat. gaso . )
c
Pyolysis in gasoline
5
1 2-8d
0. 5-1.0e
None
Lt. fluid cat-cracked
qasoline
Hvy. fluid cat-cracked
gasoline
40
2-3
i <1-2
;
0.7-2.0
0.1-0.4
30 . 6
f
Hydrocracked gasoline*

I N . A.
0.5-2.0
3.4
Hydrocracked reformate"

N .A.
0.5-1 . 5

Butane

: 0
0
7.2
Alkylate

I 0
0
12.5
Coker gasoline



1 . 1
Isomerized naphtha

¦


a Arthur D. Little computer run for Texas Gulf Complex Refinery, 1985 (100 percent
unleaded gasoline production).
k From benzene-toluene-xylene extraction.
c From steam cracking of gas oils.
^ Assumes no benzene extraction.
e C_ and C_+ gasoline; assumes benzene extracted from C, qasoline.
f 5	7	6
Light hydrocrackate is sent directly to qasoline.
^ Reformed heavy hydrocrackate to qasoline.
-------
Benzene content of commerical U.S. gasoline prior to
1974 was recorded as less than 1.0 Iv percent. The maximum
121
level in a single test sample was 1.3 lv percent.	A
1972 report on benzene analyses of 37 unleaded and low-lead
gasolines from 15 companies in the United States showed a
range from 0.3 to 2.0 percent benzene by volume, with an
19
average of 0.8 percent. Table 4-5 presents data on a
series of gasoline samples obtained from domestic Gulf Oil
refineries in October 1976. These data show an overall
average of 1.25 lv percent in three grades of leaded gaso-
lines. The highest value reported is 2.39 lv percent and
122
the lowest, 0.54 lv percent.
A nationwide survey during February/March 1977 showed
benzene concentrations in gasoline ranging from 1.25 to 5
144
percent, with a national average of about 2.5 percent.
The survey indicates that benzene levels tend to be higher
during winter than in other seasons. The investigators
estimate that the benzene concentration in gasoline averages
approximately 2 percent through the year. An American
Petroleum Institute (API) study by duPont, will present a
144
further survey of benzene levels m U.S. gasolines.
For this report, the average value of 2 percent benzene
in gasoline (from Reference 144) is used in emission esti-
mates. At the present time no regulations are known to
restrict benzene concentrations in gasoline. ^
Sources of Emissions from Gasoline-Powered Vehicles
Hydrocarbon emissions from gasoline-powered vehicles
without emission controls originate from four sources:
the carburetor, the fuel tank, the crankcase, and the engine
exhaust. Hydrocarbons are generated from the carburetor by
evaporation of fuel after a hot engine is shut off. Evapora-
tion from the fuel tank occurs through vents in the tank and
4-7
-------
Table 4-5. BENZENE LEVELS IN U.S. GULF OIL GASOLINES,
122
OCTOBER 1976
Refinery source
Liquid volume of benzene, %
Good Gulf
Gulf Crest
No-No
X
A
0. 54
0. 88
1. 16
B
1. 99
1. 45
0.85
C
1. 19
1. 21
0. 81
D
1. 59
1.18
1 . 49
E
1.2 5
1. 98
2. 39
F
0.85
0. 82
0. 88
Arithmetic mean
1.24
1. 25
1.26
Standard deviation
0. 52
0.43
0.61
4-8
-------
increases as the temperature .in the tank increases. Crank-
case emissions result f ro™ 'blowby" pa ft. the piston rings.
'Blowby* denotes leakage of the air/fuel charge into the
crankcase during the compression power strokes in an inter-
nal combustion engine: these emi anions, unless properly
controlled, escape to the atmosphere tblough a vent tube or
the crankcase ventilat.io.' cap. Ccankcase losses were a
significant source o r ; iyd r ocar he.! e« ;. <••?> « or,--* irs motor
vehicles mcinufactu'eri dori nr. the I960 ? s. Such emissions
have? been reduced in riew-sr automobilfs by venting the
V]
crankcase into the engine j. r.tii're.. Jl/drorvrbon emissions
in the engine exhaust gasas	from incomplete combus-
cion.^"' Available data indicate' that rtstr-rch effort has
been directed mosc.lv tovarus cont rol of hydr.ocarbon emis-
sions from exhau-t syshems.
For our emission :s >'^ ma t e:->, ove.porati^?> emission data
for 1972 and older automobile a axe t:a ken from EPA Publi-
cation AP-42, Supplement: 5. 'Pable 1). .1-27. 1 Data for newer
vehicles are taken from a -U) c£..t study by Exxon Research and
Engineering Corp . j. v. v'n.i c'i -jv/fer.-:..!.? c'iroc¦ J..L: v evaporative
3 n
loss was measured ar 8. ' cj/ SHED test,' " The sample in-
cluded a cross section of iM 7 .<¦ -'7cat h with /representative
evaporative controls. Most of the evaporative loss occurs
during the "hot soak" chase, which is estimated to contri-
30
bute 85 percent of this total, or 7.'i g/tes-t: , In our
emission estimates, gasoline is assumed to contain 2 percent
by volume of benzene a.od therefore benzene is assumed to
constitute 2 percent of the? hydrocarbon entiss.i on.+ With
¦k
Sealer. Housing Method for Evaporative Determinations.
+ This assumption dees not consider the relative volatility
of benzene in the gasoline?, which may result in some over-
estimation, Because- of the .,rariety of temperature and
pressure conditions thnit must be considered in determining
evaporative losses and because most, emissions occur during
"hot soak," which is at an ele /a ted engine temperature,
this cissumption should not cause a significant overestimation.
4-9
-------
•k
approximately 96 million cars in current operation, benzene
losses by evaporation would be 31.8 Gg/yr (70.1 million
lb/yr).+
Our estimates of total exhaust emissions are based on
HC emission test data given in AP-42."^ Benzene emission
rates are presented below primarily to determine the ben-
zene-to-THC ratio. These data are restricted almost en-
tirely to catalyst-controlled vehicles, which have been in
widespread use in the U.S. only since 1975. With few excep-
tions, the published data have not reported the benzene
level in the gasoline under test conditions. Since this is
a critical factor in estimates of benzene emissions, the
usefulness of published studies is limited. Also, because
certain components, such as ethylbenzene and other aromatics
within the blend, can break down to form benzene, benzene
can be emitted in engine exhaust even though it is non-
existent in the gasoline. Thus, benzene emissions depend
not only on benzene levels in gasoline but also on parti-
cular characteristics of the gasoline blend.
Black and Bradow reported the benzene content of hydro-
17
carbon emissions from engine exhausts of 10 automobiles.
These low-mileage vehicles included one 1972 production car
with no catalytic emission control, three 1975 prototypes,
and six 1975 production cars equipped with catalytic control
devices. The cars were fueled from a single batch of
unleaded test gasoline containing only 0.03 weight percent
benzene. The vehicles were driven on a chassis dynamometer
simulating a route typical of urban and highway driving
conditions. Detailed hydrocarbon composition analysis (to
~
Nationwide data from the Department of Transportation and
the Motor Vehicle Manufacturers Association.
+ An average estimate determined from data on hydrocarbon
emissions from controlled/uncontrolled vehicles.95
4-10
-------
determine benzene content) of exhaust, samples was made by
54
gas chromotography. Results given in Table A-l show
hydrocarbon levels ranging from a high of 0.63 g/km (1.0
g/mile) to a low of 0.15 g/km (0.242 g/mile); these values
conform to the expected effectiveness of exhaust controls.
The highest benzene concentration among the catalyst-equipped
cars was 9 mg/km (0.015 g/mile), whereas the concentration
for the older uncontrolled 1972 vehicle was almost double
that amount. The average benzene concentration was 2.3
percent of the total hydrocarbons.
Olson Laboratories conducted benzene analyses of the
52
exhaust from forty-five 1975 vehicles. These data are
presented in Table A-2. The mean benzene concentration was
9.94 mg/km (0.016 g/mile), with an average benzene-to-THC
ratio of approximately 2.5 percent. The Olson study indi-
cates a somewhat higher level of benzene exhaust emissions
than the other studies mentioned and a wide variability in
the data, with a equalling 0.018. Since these data are
based on California vehicles and the benzene level in the
test gasoline was not reported, they may not be representa-
tive on a nationwide basis.
Hydrocarbon emissions from a 1976 production vehicle
using a lean burn system have been analyzed for benzene.
Emissions were not significantly different (-0.0134 g/mile)
from those from cars with catalyst systems. Benzene emis-
sions from a 1974 production vehicle equipped with a thermal
reactor system were similar to those of an uncontrolled
vehicle. It is anticipated that to meet the more stringent
hydrocarbon emission standards of the future, all highway
qasolins-powered vehicles will be equipped with some type of
catalyst system by 1985.
Our estimate of the benzene exhaust emissions in
1975-1976 (years in which catalyst control systems were
4-11
-------
installed on cars), utilizes a value of 1.0 g/mi of THC
31
(from AP-42) and a benzene-to-THC ratio of 0.025 (from the
Olson studies) to arrive at an average level of 0.025 g of
benzene/ mile. This benzene-to-THC ratio is based on data
from the Olson studies for which no benzene level in the
gasoline was reported. This level, although higher than the
mean benzene emission rate of 0.016 g/mile from Olson (Table
A-2), was chosen to reflect the impact of other control
systems used on these vehicles and the effects of different
engine conditions and gasoline benzene contents.
4
In 1973 the Bureau of Mines conducted a study of
vehicles with five different emission-control systems to
explore the effects of fuel compositions and ambient tem-
perature on exhaust HC emissions and benzene content. Five
of six vehicles studied showed a decrease in benzene emis-
sion with decrease in fuel aromaticity. Ambient temperature
tests indicated that the total hydrocarbon emissions and
benzene emissions were lowest at 24°C; emissions increased
at higher or lower temperatures, the most pronounced rate of
increase occurring at lower temperatures. A more thorough
review of this research is presented in Appendix A.
EPA is funding a number of catalyst characterization
projects. Results of this work, to be available in 1978,
should provide a more comprehensive data base concerning
52
benzene emissions from auto exhaust.
In summary, benzene content of gasoline is approxi-
mately 2.0 liquid volume percent, varying with the source of
crude oil, refinery operations, grade of fuel, season, and
geographic area. Whereas research concerning benzene emis-
sions from vehicle evaporative sources has been limited,
much effort has been directed towards analysis of tailpipe
emissions. Studies confirm that benzene emissions from
controlled vehicles are lower than those from uncontrolled
4-12
-------
vehicles. Also, tests show that an increase in fuel aromati-
city is associated with an increase in benzene emissions.
This emission increase, however, may not result from altered
chemical composition but rather from changes in physical
properties of the fuels and failure to adjust the vehicles
for use of fuels of higher aromatic content. Studies also
reveal that altering ambient temperatures in either direc-
tion from 24°C (simulating seasonal variations) increases
benzene content of exhaust emissions. Tests at temperatures
below 24°C generally produced higher benzene emissions than
tests at temperatures above 24°C.
4.1.2 Benzene Emissions Calculations
Table 4-6 shows the information used to calculate the
total 1976 benzene emissions from autos.
4.1.2.1	Best Estimate - Best estimate of 1976 auto emis-
sions - exhaust plus evaporative losses.
G *	+
Exhaust emissions = (96 x 10 ) (9494) Z a x b x c
- (96 x 106)(9494)(0.186)
= 16 9.4 Gg/yr (373.5 million lb/yr).
Evaporative losses - (96 x 10^)(9494)(0.02)3 Z a x d
= (96 x 106) (9494) (0. 02) (1 . 744)
= 31.8 Gg/yr (70.1 million lb/yr).
Total emissions = 201.2 Gg/yr (443.6 million lb/yr).
4.1.2.2	I-Iigh Estimate - The high estimate of benzene emis-
sions assumes that (1) no catalyst exhaust controls are used
(exhaust emissions remain at the 1974 level, column b2 Table
4-6) and (2) evaporative losses remain at the 1972 rate, as
*	9 5
Estimated number of autos in operation in 1976.
f	9 5
Average mileage traveled per car in 1974.
("i	144
Average fraction of benzene in gasoline.
4-13
-------
Table 4-6. AUTOMOBILE EMISSION FACTORS



*
Exhaust emissions

Evaporative
losses


a
b
c
bl
b2
d
Vehicle
model
year
Vehicle
age,
years
Fraction
of miles+
traveled
HC
emissions,
g/mi
Benzene-
to-HC 5
ratio"
Low HC
emissions
~ *
case
g/mi
High HC
emissions
case
g/mi
HC
g/mi
1976
1
0.116
1.0
0. 025
1.0
3.8
1.24
1975
2
0.135
1.2
0 . 025
1.2
3.8
1.24
1974
3
0. 125
3.8
0 . 041
1.4
3.8
1.24
1973
4
0. 122
4.1
0. 041
1.6
4.1
1.24
1972
5
0.106
4.4
0.041
1.8
4.4
1.76
1971
6
0. 086
5.5
0. 041
2.0
5.5
1.78
1970
7
0.083
6.8
0. 041
2.2
6.8
2.53
1969
8
0. 072
6.8
0. 041
2.4
6.8
2 .53
1968
9
0.051
8.6
0. 041
2.6
8.6
2.53
1967
10 plus
0.105
1.	000
9.1
0. 041
2.8
9.1
2.53
All emission data are based on low-altitude, nonCalifornia cases.
AP-42, Supplement 5, Table 3.1.2-5
AP-42, Supplement 5, Table D.l-1, reference year 1976.
§
Ratios for 1975 and 1976 are based on experimental data; ratios for other years are
based on noncatalyst levels."'"
* ~
Catalyst-controlled emissions are based on a 1976 base year rate of 1.0 g/mi and
deterioration increases from AP-42, Supplement 5, Tables D.l-5 to D.1-1-17.
-------
31
presented in AP-42 (designated as d^). The benzene-to-HC
ratio is assumed to be constant (0.041) for all model years.
Exhaust emissions =	(96 x 10^) (9494) (0.041) I a x
=	(96 x 106) (9494) (0. 041) (5. 32)
=	198.7 Gg/yr (438.1 million lb/yr).
Evaporative losses = (96 x 10^) (9494) (0.02) T, a x d^
=	(96 x 106) (9494) (0. 02) (2. 00)
=	36.5 Gg/yr (80.5 million lb/yr).
Total emissions = 235.2 Gg/yr (518.6 million lb/yr).
4.1.2.3	Low Estimate - The low estimate is based on all
auto exhaust emissions at the 1976 catalyst-controlled level
(column b^ Table 4-6). Increases in exhaust rate with age
of vehicle are because of catalyst deterioration and are
based on AP-42 data."^
C.
Exhaust emissions - (96 x 10 ) (9494) (0.025) T, a x b^
= (96 x 106) (9494) (0.025) (1.909)
= 43.5 Gg/yr (95.9 million lb/yr).
Evaporative emissions (from subsection 4.1.2.1)
= 31.8 Gg/yr (70.1 million lb/yr)
Total emissions = 75.3 Gg/yr (166.0 million lb/yr).
4.1.2.4	Benzene Emission Variations - The best estimate of
total benzene emissions from automobiles (see Table 4-2) is
201.2 Gg/yr (443.6 million lb/yr). The high and low esti-
mates are 235.2 Gg/yr (518.6 million lb/yr) and 75.3 Gg/yr
(166 million lb/yr). These high and low estimates are
intended to indicate an overall emissions range and were
derived by using extreme values for the emission factors.
The effect of benzene in the gasoline on the benzene emis-
sion rate is not well-defined for either catalyst-controlled
or precontrol vehicles. Also, the emission changes attribut-
able to control method and vehicle age (deterioration rate)
4-] 5
-------
have not boon tested adequately. Benzene emissions also may
vary widely as a result of fractions of benzene and other
aromatics in gasolines from different geographic regions in
the United States (possibly an important factor in deter-
mining the levels of benzene to which specific populations
are exposed).
4.1.3 Calculation of Emissions from Other Mobile Sources
Diesel-fueled engines and aircraft are not considered
potentially large benzene emission sources. Diesel-fueled
engines include power buses, light- and heavy-duty trucks,
and some automobiles. Also included are locomotives;
military, pleasure, and commercial vessels; and agricultural
and heavy-duty-construction equipment. Total hydrocarbon
emissions from diesels are relatively low because compres-
sion-ignition engines generally provide fairly complete
33
combustion. Crankcase and evaporative emissions are low,
and the exhaust system is the only major hydrocarbon emis-
3 3
sion source. Studies conducted by Southwest Research
Institute on emissions of benzene and other volatile hydro-
carbons from a heavy-duty diesel truck engine indicate a
benzene-to-TIIC mean ratio of 0. 0073 under varied engine
52
stress conditions. Table 4-7 presents data from this
study. By use of this factor, a value for national benzene
emissions from heavy-duty diesel trucks (assumed to be the
greatest contributors of benzene in the category of "other
mobile sources") is estimated as follows:
(9 , 630, 754 , 000 gal.)"(5.1 miles/gal. )+(4. 6 g/mile) ( 0 . 00 7 3 )
= 1.7 Gg benzene.
~
The total 1975 diesel fuel consumption in the United
States.141
Reference G3.
Reference 31.
4-1 f,
-------
Table 4-7. BENZENE AND HC EMISSIONS FROM DIESEL ENGINE UNDER
VARYING OPERATING CONDITIONS52
Engine condition
Hydrocarbons, j
ppm C '
Benzene,
ppm C
Benzene/hydrocarbon,
o,
o
Intermediate speed, 2% load

101
!
1.0
0. 99
Intermediate speed, 50% load

89 :
0. 5
0.56
Intermediate speed, 100% load

100 1
1
0.7
0.70
High speed, 2% load

92 j
0.7
0.76
High speed, 50% load

111 i
0.9
o
CO
High speed, 100% load

85
0.9
1.06
Idling speed, 2% load

141
l
0.3
0.21

i
Mean =102.7 |
0.7
0.73
-------
It i. s anticipated that additional data on benzene
emissions from diesel trucks (resulting from recently
52
initiated EPA contracts) will be available in early 1978.
During a study of occupational exposure to benzene at ser-
vice stations, NIOSII was unable to detect benzene in diesel
64
fuel samples taken from survey sites. Measured concentra-
tions were reported to be below the detectable limit range
of 0.5 liquid volume percent. Shell Oil reports that the
benzene level in diesel fuel should be in the parts-per-
4 3
million range because of flash point specifications.
Although no data are available, the greatest potential
for benzene emissions from aircraft probably occurs during
ground operations (taxiing and idling after landing and
before takeoff) when hydrocarbon emissions are most signif-
. 31
Leant.
4.2 STATIONARY SOURCKS
The primary U.S. stationary sources of atmospheric
benzene emissions are petroleum refineries, petrochemical
plants, and coal coking operations. The projected breakdown
of 1977 benzene production is as follows: 50 percent from
refinery reformate, 16 percent from pyrolysis gas olefin
streams, 5 to 8 percent from coking operations, 27 percent
from hydrodealkylation of toluene, and 4 percent from
imports.^"'" The estimated volume of benzene production for
1977 is 1.56 billion gallons (5.6 MmV^
Variations in benzene production from year to year
reflect not only changes in petrochemical, demands for
benzene but also increases in gasoline demand. One of the
effects of l h i demand is thai refineries divert more aroma-
ties to the gasoline pool to meet octane requirements,
especially with the phasing down of lead antiknock additives.
4-1 8
-------
Major benzene emission sources, addressed individually
in the following sections, include petroleum refineries,
coke ovens, major benzene consumption processes, and storage
and distribution evaporative sources. Some minor sources
are also discussed.
4.2.1 Petroleum Refining and Benzene Production
Petroleum refineries appear to represent a significant
source of benzene emissions, although the emissions have
never been quantified. Figure 4-1 presents the geographical
distribution of U.S. refineries, and Appendix B-l lists the
refineries, with crude capacities and gasoline output.
Major refinery products of interest here are commercial
grades of benzene and gasoline. Appendix B-2 presents a
list of U.S. producers of petroleum-based benzene, their
locations, and their capacities; Figure 4-2 presents the
geographical distribution of these facilities in the United
States. These facilities are mostly along the Texas-Louisiana
coastline of the Gulf of Mexico.
Benzene emission sources in refineries include (1)
process emissions from light and heavy naphtha streams from
the crude unit (Figures 4-3 and 4-4); fluid catalytic crack-
ing units; hydrocracking units; gasoline treating units; and
pumps, flanges, and other sources of fugitive emissions; and
(2) nonprocess emissions from wastewater treatment facilities,
heaters and boilers, and facilities for storage and handling
of benzene and gasoline.
Benzene is recovered from the reformate of catalytic
reforming units. For benzene production, toluene dealkyla-
tion processes are used to convert higher aromatics in
reforma-e-mixed aromatics to benzene. The average distri-
bution of aromatics in reformate is 2 to 4 percent benzene,
4-19
-------

• ••
• •
• •
/•
::

•• ••
••••
^ •
• is:
••
Figure 4-1. Location of United States oil refineries (not including
Alaska, Hawaii, Puerto Rico, and asphalt or very small plants).
-------
I V J
h-	


f- _
L.._.

• •.SOU*'	>s»
\ ^
Figure 4-2. Geographical location of petroleum-based
benzene production plants.
-------
?» r i -j
\ ""J
C\

¦. gr/. r-e
-CEEE
:p-	
! -i»-.-r.
¦['.".•j'U
/~A
C' J-1-
b.:IZ3

Wesulo
f
,y 904
T r
i " i i i
'0\ ; •' ir i!
',iV
"i" "" 1
>r-'"

4Z
' .'fOi'jM f Jf- '/IV >
: '|T;
i
I •••' }l"
•'J'
i
\Z7
» ' ,'' oil
W
€3
V - I-" . j --1
t
' F uei gas
. PG
fA t(j' ') I'.oHnf
. (Jti: •• jciSOl-ne
f »5' osi^e
.	f C' i
jipsei *je'
r..j|fur
*- '.V.'iWj
H(-nvy 'oC- Oil
Aspholt
F i (jure
Source
4-4. Process iruj pi ri n for typical complete refinery.
Ref. 130.
4-2 2
-------
15 to 19 percent toluene, 28 to 30 percent xylene, and the
107
remainder other aromatics.	In the reforming-separation
process, a petroleum fraction is catalytically reformed;
benzene and other aromatic hydrocarbons removed from the
reformate by solvent extraction then are distilled frac-
tionally to separate the solvent. Benzene originally
present in the petroleum fraction goes through the process
essentially unchanged. The total benzene yield includes
both this unchanged benzene and benzene from various pre-
43
cursors, primarily cyclohexane and. methylcyclopentane.
Benzene is also produced by dealkylation of toluene or
toluene-containing stocks such as heavy catalytic refor-
mates, catalytic cycle oils, and the like. Several pro-
cesses, both catalytic and thermal, are utilized. Dealkyla-
tion is performed in the presence of hydrogen (often catalytic-
reformer hydrogen), and no solvent extraction of the product
59 81
is necessary. ' This process is commonly practiced in a
petrochemical complex rather than in a petroleum refinery.
Whether benzene is produced as a refinery product or
mixed in the gasoline pool may not significantly affect
atmospheric concentrations, since the same quantity of
benzene is being handled (unless toluene dealkylation is
performed). Although separation of benzene probably creates
more fugitive emissions, all refineries emit benzene but
in ill-defined quantities.
The only available data on benzene exposures concern
occupational exposures and indicate a common 8-hour tiine-
41
weighted-average (TWA) exposure of 1 to 5 ppm. Young et
16 3
al. also report benzene concentrations at various refineries.
The highest concentrations were measured near pumps (1 to 16
ppm) and near a benzene tower reboiler (<1 to 6 ppm).
4-2 3
-------
Refineries contain numerous tanks for blending gasoline
and storing components; evaporation from these tanks prob-
ably is a major source of benzene emissions. Emission
estimates for storage tanks are presented in storage and
transport emissions {Section 4.2.4).
No data are available on benzene emissions from re-
fineries, nor have estimates been published. Since process
emissions appear to occur from leaks, only an order-of-
magnitude estimate can be made. Refinery emission factors
for hydrocarbons from revised EPA AP-42 data are used in
117
these estimates.	In emissions from cooling towers,
valves and flanges, relief valves, pump seals, sampling and
other miscellaneous losses, we estimate 1 percent of the
hydrocarbons is benzene. In hydrocarbon emissions from
blowdowns, process drains, and vacuum jets, we estimate 0.1
percent is benzene. Benzene emission factors used for
cooling towers are 10 percent of revised AP-42 levels, based
44
upon data provided by Shell Oil. Benzene emissions from
catalytic cracker regenerators are assumed to be very low
because a combustion temperature of approximately 1200°F and
an atmosphere of excess oxygen, as used in these operations,
are not conducive to benzene formation.
The emission factors used from AP-42 estimate emissions
117
from controlled and uncontrolled sources.	Although all
refineries have some control, it assumed that refineries
with less than 35,000 barrels per stream day (BPSD) are
uncontrolled. In 1973, this segment of the industry repre-
7 6
sented 19.6 percent of total U.S. crude capacity. Because
most subsequent capacity increases have been in larger
refineries, the portion of uncontrolled capacity is reduced
to 15 percent for these estimates. The overall emission
factors, expressed as pounds of benzene emitted per 1000 bbl
4-24
-------
of crude feed, are 0.415 for controlled refineries, 2.27 for
uncontrolled refineries, and a weighted estimate of 0.759
for the industry average.
The estimated total U.S. annual benzene emissions from
refineries = (15.4 x 10^') (350)+ (0. 759 x 10
= 2.1 Gg/yr (4.1 million lb/yr).
Estimated benzene emissions from a 150,000 BPSD "con-
trolled" refinery are 21,800 lb/yr; estimated emissions from
a 10,000 BPSD "uncontrolled" refinery are 7900 lb/yr.
Estimates of benzene emissions from benzene production
plants are based on a Union Carbide emission factor of 0.035
130
lb per ton of benzene produced.	Because this factor is
based on an aromatics plant utilizing a crude olefins by-
product stream, it is not necessarily representative of all
benzene production operations.
The total U.S. annual benzene production emissions
based on this emission factor
= (1.45 x 10S)§ (7.365**/2000) (0.035)
= 0.08 Gg/yr (0.19 million lb/yr).
4.2.2 Coke Ovens and Coke Oven By-Production
Although coke ovens represent a relatively minor
benzene production source (about 5 to 8 percent of U.S.
total), they are a potentially significant local environ-
mental benzene source because of their poor vapor contain-
ment. Appendix C-1 lists the coal-derived benzene pro-
ducers in the United 3tates, their locations, and their
2 8
production capacities„ Figure 4-5 depicts the geographic
distribution of these facilities.
*	6	115
U.S. refinery crude feed rate, 15.4 x 10 barrels/day.
+ Days operation/year.
ri	9	5
U.S. benzene production, 1..49 x 10 gallons/year.
**
Conversion for benzene lb/gallon.
4-2 5
-------
Figure 4-5. Geographical location of coal-derived benzene production plants.
-------
Major aromatic hydrocarbon constituents of coke-oven
gas qenerated during coking operations are about 0.66
percent by volume benzene, 0.13 percent toluene, 0.05 per-
59
cent xylene, and less than 0.10 percent other aromatics.
Higher temperatures produce greater amounts of aromatic
hydrocarbons, particularly benzene, and lesser quantities of
paraffinic, napthenic (saturated alicyclic), and unsaturated
59 153
hydrocarbons. '	The yield of light oil from coke ovens
3
producing blast-furnace coke ranges from 13 to 17 Mm /Mg (3
81
to 4 gallons per ton) of coal carbonized. The principal
constituents of this oil are benzene (60 to 85 percent),
toluene (6 to 17 percent), xylene (1 to 7 percent), and
153
solvent naphtha (0.5 to 3 percent).	Light oil is re-
covered by the following methods: '153
° Compression and cooling at pressures of 1.0 MPa
(10 atmospheres) and temperatures below -70°C.
° Adsorption on solids such as activated carbon (or
silica gel) and recovery from the carbon by direct
or indirect steam heat.
° Continuous countercurrent absorption by solvents,
e.g., petroleum wash oil, a coal-tar fraction, or
other absorbent, followed by steam distillation
stripping to recover the light oil.
This crude light oil is then fractionated to produce
benzene, toluene, and xylene. A typical amount of benzene
3
recovered from coke-oven gas is 7.7 Mm /Mg (1.85 gallons per
34
ton) of coal input.
Distillation of coal tar is another source of benzene.
Tho composition of coal tar, although chiefly determined by
the coking and recovery processes, also varies with the
cjradc of raw coal. Coke-oven by-production generates 33 to
3
50 Mm per Mg (8 to 12 gallons per ton) of tar according to
4-27
-------
t.ho grade of coal coked. This tar is fractionally distilled
to yield light., middle, heavy, and anthracene oils. The
light oil fraction (approximately 5 percent of the tar) is
generally added to the major portion of light oil recovered
from coal gas to be refined for its benzene content.
Figure 4-6 illustrates the basic coke-oven operation
8 3
and major points of air pollutant emissions. Emissions
occur from various points during charging, pushing, and
quenching operations; from doors during the coking cycle;
and from the waste-gas stack.
An uncontrolled charging operation is believed to have
the greatest potential for emissions of hydrocarbons, and
thus of benzene. When the moist coal comes into contact
with the hot oven floor and chamber walls, evaporation and
coking (distillation and cracking of volatile components)
start immediately. If not controlled, the liberated vapors
and gases escape to the open air."^^ These emissions may
vary widely depending on the characteristics of the coal and
operating practices. Hydrocarbon emissions from charging
have been reported to range from 57 to 800 grams per ton of
14
coal charged at some European coke plants. Total emis-
sions of about 5.4 Mg (6 tons) per day [including 0.15 Mg
(0.4 ton) of benzene-containing hydrocarbons] are generated
25
from a battery of 65 high-capacity 6 m (19.7 ft) ovens.
Topside emissions of fugitive benzene occur from a
variety of sources including leakage from weakened refrac-
tory materials, ascension-pipe-elbow covers, leveling
apertures, badly fitted charging-hole covers, and collecting
main pipe valves.
Coke pushing involves the discharge of incandescent coke
from an oven on the coke side by operation of a ram on a
pusher machine on the push side. This operation is believed
4-28
-------
figure 4-6.
showing PosS
o£ by-ProduCt fof benzene.83
5ible atmospheric
-------
to be a major source of benzene emissions during the coking
13 6
cycle.	Under optimum conditions, however, coal is
completely carbonized, and coke pushing results in minimal
benzene emissions because most of the aromatic compounds
already have oeon drown off !.o be recovered from the oven
gas. Three tests performed to measure emissions during the
pushing operation shov/ed benzene levels in the range of
14 5
0.003 to 0.04 lb benzene per ton of coke produced.
Door leaks represent one of the most significant emis-
sions sources. A 1975 national survey assessing the com-
pliance status of coke-oven facilities revealed that 57
percent did not comply with state air pollution regulations
during the coking cycle because of door leaks on both push
8 3
and coke sides of the ovens. " Leaks can be minimized by
good maintenance of door seals or by using a special luting
mixture to seal the oven before charging.
The combustion chamber gas (natural or cleaned coke-
oven gas) used for firming leaves the battery after combus-
tion and is emitted to the atmosphere via the waste-gas
stack. Waste-gas stacks normally are in compliance with
state regulations, even though they rarely are equipped with
control devices. Thus, they are not likely to be important
sources of atmospheric benzene in well-maintained operations.
Emissions can be significant, however, if coal or coke
escapes through leaks into areas between the oven and the
heating flue.
Quenching involves the flooding of pushed hot coke
84
(1.100+°C) with a huge quantity of water to cool it rapidly.
The large quantities of steam generated by this operation
vary during the quench cycle. [-'missions from quenching
depend on the completeness of coking and the sources of
quenching water. No data are available on benzene emissions
from this operation.
4-3 0
-------
The only available data on ambient benzene concentra-
tions attributable to coking operations concern occupational
exposures. Typical exposures to ambient benzene (in 8-hour,
time-weighted averages) by occupation were measured within a
coal-derived benzene recovery plant; these are listed in
Table 4-8.
Table 4-8. EXPOSURES OF WORKERS TO BENZENE IN
49
A COAL-DERIVED BENZENE PRODUCTION PLANT
Occupation
Benzene
concentration
(8-hour
time-weighted
average), ppm
Range
ppm
r
Agitator operator
6.0
0.5 -
20
Benzene loader and
loader helper
4.0
0.5 -
15
Benzene still operator
4.0
1 -
15
Light oil still
operator
2.5
1 -
15
Naphthalene operator
10.0
2 -
30
Analyst
10.0
4 -
30
Chemical observer
10. 0
4 -
50
Foreman
1.5
1 -
10
In Czechoslovakia, atmospheric benzene emissions
generated in the coking and by-product recovery phases of
benzene processing were assessed. Within coke-oven battery
working areas, ambient benzene concentrations ranged from 50
3	3
|jg/m to 13.0 mg/m ; in the recovery plant, the concentra-
3	3
tions ranged from 50 yg/m to an exceptional 145 mg/m ,
almost triple the limit imposed by Czechoslovakian standards
3 93
(50 mg/m ). Within working areas of tar-processing
3	93
plants, the analysis detected 0.3 mg/m of emitted benzene.
4-31
-------
Our order-of-magnitude estimate of total annual benzene
emissions from coking operations in the United States is
based on a hydrocarbon emission factor of 6.9 lb emitted per
31
ton of coal coked. The benzene fraction of total hydro-
carbon emission is based on coke-oven gas containing 34.8
percent hydrocarbons and 0.776 percent benzene. The value
for hydrocarbon content is an averaye of two reported
values, 36.1 and 33.6 percent.	The benzene concen-
tration in coke-oven gas (0.776%) is the average of three
plant measurements: 1.028 percent at a USSR plant, 0.686
percent at another USSR plant, and 0.615 percent at a West
1 . 77,82,85
German plant.
Annual U.S. benzene emission from coking operations
= (50.4 x 106)* (6.9) +(0.0223)§
= 3.52 Gg/yr (7.76 million lb/yr).
The above calculation assumes that all hydrocarbons
contain the same benzene fraction as those in coke-oven gas.
This fraction probably depends on the time in the coking
cycle and the source of the emission, and therefore is not a
good index of benzene emissions from the by-product recovery
process. It is applied because no specific data are avail-
able on emissions from by-product recovery.
Benzene emissions from a Czechoslovakian coke plant
producing 4 million tons of coke per year are reported to be
94
1660 tons. If a 75 percent by weight coal conversion to
-4
coke is assumed, this benzene emission rate of 4.15 x 10
kg/kg coal charged leads to a U.S. benzene emission of 19.0
Gg/yr (41.8 million lb/yr). Thus the coking operations are
*	'	~	79
Total annual U.S. coke-oven feed rate, tons of coal.
Hydrocarbon omission factor for coke ovens, lb/ton of
coal.
Fraction of benzene in total hydrocarbons, utilizing level
of 0.776 percent benzene in coke-oven gas and 34.8 percent
hydrocarbons in coke-oven gas.59
4-32
-------
potentially a major industrial source of benzene emissions.
In summary, data on benzene emissions from coke-oven
and benzene-recovery plant operations are very meager, but
pertinent occupational exposure levels have been measured.
A study on by-product plants being conducted by Research
Triangle Institute may improve the accuracy of the esti-
mates. Data on by-product recovery plants indicate that
exposures vary with occupation and over extended periods of
time. Charging of coal is regarded as the greatest poten-
tial source of benzene emission in coke-oven operations;
however, leakage from doors may be an equally or more
serious source in U.S. plants. Lesser quantities are attri-
buted to topside leakage, pushing, quenching, and waste-
stack emissions.
4.2.3 Benzene Consumption
This subsection addresses benzene emissions from the
major commercial uses of benzene. Total 1976 U.S. consump-
3
tion of benzene in such uses was 5.3 to 5.6 Mm (1.4 to 1.5
5
billion gallons). Primary usage is for the manufacture of
such chemicals as nitrobenzene, ethylbenzene, maleic an-
hydride, cumene, phenol, chlorobenzenes, cyclohexane, and
detergent alkylate, as well as other nonfuel products and
solvents, as shown in Figure 4-7. The production capacities
for the' major compounds are presented in Table 4.9. Pro-
ducers of various compounds (excluding solvents) are listed
in Appendix D-l, with their locations and 1976 capacities.
Figure 4-8 shows the geographic distribution of these
facilities. Primary locations are along in the Gulf coast,
in the northeastern section of the United States, and in
Puerto Rico.
4-33
-------
AilUTE
DOT
*01*- V
C^?OSAU$ JU^CX? k»5!S<
:*TS€C TIC lots
MOLDINGS
COATINGS
COATINGS
M3ERS
Figure 4-7. Flow chart for benzene derivatives and their uses.
Source: Ref. 66.
-------

/•
lt
•x««rr «.< o
Figure 4-8. Geographical location of benzene-consumption plants.
-------
Table 4-9. ANNUAL (1976) PRODUCTION CAPACITY FOR
SOME MAJOR CHEMICAL COMPOUNDS IN WHICH BENZENE
IS USED AS AN INTERMEDIATE

o p
1976 Production ^
Chemical
Gg/yr
10" lb/yr
Nitrobenzene
480
1065
Ethylbenzene
3890
8566
Maleic anhydride
188
415
Cyclohexane
1230
2706
Cumene
1720
3785
Chlorobenzene (mono)
313
690
Detergent alkylate
392
865
Nitrobenzene
Nitrobenzene (CgH,-N02) is prepared by direct nitration
of benzene, using mixed nitric and sulfuric acids (see
Appendix Figure D-l) in both batch and continuous opera-
tions. A typical batch reaction vessel (or nitrator) is
sized for 450 to 700 kg (1000 to 1500 lb) of benzene and
operates on a time cycle of 2 to 4 hours. The nitrobenzene
produced in this reaction is principally used directly in
the manufacture of aniline. Newer plants use the continuous
process, which is fundamentally the same as the batch pro-
cess except for smaller reaction units, lower nitric acid
concentrations, and higher reaction rates. Overall, it is
estimated that some 85 percent of the chemical is used in
59
the manufacture of aniline.
Reactors and acid concentration systems are the main
sources of any atmospheric emissions during processing.
These emissions are vented to absorbers, where any NO^ is
4-16
-------
oxidized with air and absorbed in water to produce recycled
nitric acid.
Monsanto Research Corporation reports that ambient
benzene emissions from nitrobenzene production in 1975
97
totaled 3.39 Gg (7.47 million lb). On the basis of a
total 1975 domestic nitrobenzene production capacity of 480
Gg (532,500 ton), the resulting benzene emission factor is
81
0.007 kg per kg of product.
Aniline
Aniline (C^HcNH~) is produced either by chemical
d b 2.
reduction or catalytic hydrogenation of nitrobenzene. The
traditional batch method (now seldom used) involves reduc-
tion with iron filings and hydrochloric acid. Most modern
plants practice continuous catalytic vapor-phase hydrogena-
tion because it is more economical.59'66
No data are available on atmospheric benzene emissions
from the manufacture of aniline. Because benzene is present
principally as an impurity in nitrobenzene, this does not
appear to be an important benzene emission source.
Ethylbenzene/Styrene
Ethylbenzene (CgH^C^H^) is prepared primarily by the
liquid-phase Friedel-Crafts alkylation of benzene with
ethylene. Styrene (CgH^CHC^) is produced solely by the
catalytic dehydrogenation of ethylbenzene.
In the production of ethylbenzene, aluminum chloride is
the catalyst and ethyl chloride or hydrogen chloride is the
promoter. The process takes place within a brick-lined
steel tower or a glass-lined reactor operating at 90 to
100°C and at atmospheric pressure. Liquid products from the
alkylator are cooled, then passed through a settler, where
4-37
-------
the aluminum chloride complex is removed and returned to the
alkylator (Appendix Figure D-2). The water-washed alkylate
is scrubbed with 20 percent caustic for neutralization, then
separated into components in a series of distillation
columns, the first of which removes the benzene for recycle
to the reactor. The gaseous products leaving the top of the
alkylator, (unreacted ethylene, some benzene, hydrogen
chloride, and inerts) enter a condenser, from which the
recovered benzene flows back to the alkylator. The off-gas
is scrubbed for final benzene recovery, with recycled
polyethyl benzenes as absorbents; the gas is then water-
washed for HC1 removal before being vented or compressed for
, , 81,59,106
use as fuel.
Other major ethylbenzene production processes using a
benzene feed are the Alkar process and the Mobil/Badger
ethylbenzene process. The Alkar process utilizes a fixed-
bed catalyst and can handle a mixed olefins feed to produce
106
high-purity ethylbenzene.	The reactor effluent is cooled
and flashed to remove noncondensibles when required, i.e.
when mixed olefins contain saturated hydrocarbons. The
Mobil/Badger ethylbenzene process flow is similar. In both
processes the uncondensed reactor effluent vapor can be
10 6
vented to the process fuel gas system.
Recovery of ethylbenzene from mixed xylenes is the only
other significant source of ethylbenzene capacity; it
47
accounts for approximately 6 percent of domestic production.
This process produces no benzene emissions.
Direct dehydrogenation of ethylbenzene is the only
commercial process currently employed for the manufacture of
styrone. Preheated ethylbenzene and superheated steam are
fed into a reactor containing a selective dehydrogenation
catalyst. Available production routes involve either an
4-3B
-------
adjabatic or an isothermal reactor. Steam serves as a heat
source and a diluent to improve conversion. The reactor
effluent is cooled and flashed, the noncondensibles going to
fuel and liquid going to a distillation train. '
A case study (with unspecified monitoring methodology)
of an ethylbenzene plant of 101 Gg/yr (223 million lb/yr)
capacity reported a scrubber vapor effluent containing 11 kg
6 6
of benzene per Mg of product (22 lb per ton of product).
If uncontrolled effluent were vented into the atmosphere,
annual emissions would be 1.1 Gg (1226.5 tons) of benzene.
Ethylbenzene production using a benzene feed process can be
calculated as the total domestic production of 2.14 Tg (2.36
5
million tons) minus the ethylbenzene production from mixed
aromatics (0.190 Tg, 0.209 million tons), or 1.95 Tg (2.15
million tons). Application of this emission factor yields a
hypothetical uncontrolled annual emission of 21.46 Gg (47.3
million lb) of benzene. This emission rate is reported to
be representative of older technology, resulting in an
emissions value that is much higher than the present in-
37
dustry wide total. The controlled benzene emission from a
combined ethylbenzene/styrene facility reported by Union
37	-5
Carbide' is 0.065 x 10 kg/kg of product. Application of
this emission factor gives total benzene emissions of 0.14
Gg/yr (0.308 million lb/yr). It not known what percentage
of domestic producers apply emission controls comparable to
those a- the facility described.
The national emission inventory for styrene production
6 8
plants reported by Houdry indicates that benzene emissions
range between 2 x 10 ^ and 2.2 x 10 ^ kg/kg of styrene,
based on data from 1.40 million tons of the total 1975
5
production of 2.2 million ton styrene. Using an average
_ 3
emission factor of 1.5 x 10 kg/kg, PEDCo's estimates of
4-39
-------
total benzene emissions is 2.9 Gg/yr (6.48 million lb/yr).
Major emission sources are the distillation column hotwell
4- 68
vents.
The Houdry data for styrene plants show higher emission
rates than Union Carbide data for ethylbenzene plus styrene
6 8
trains. To estimate the ethylbenzene process emissions,
it is assumed that reactor effluent vents are controlled and
that fractionation vacuum jet vents represent the major
emission sources. Houdry reports that the total hydrocarbon
losses from distillation vents at styrene plants are 0.00062
6 8
ton/ton of styrene. This emission factor is used in
estimating benzene emissions from ethylbenzene processes
because benzene is the major volatile hydrocarbon present in
the fractionation train of these processes. In ethylbenzene
production a benzene feed of 1.95 Tg (2.15 x 10 tons)
yields an estimated annual emission rate of 1.2 Gg (2.67
million lb).
In summary, benzene emission estimates are
1.2 Gg/yr (2.67 million lb/yr) - for ethylbenzene production
2.9 Gg/yr (6.48 million lb/yr) - for styrene production.
Maleic anhydride
Maleic anhydride CHCO is produced from the oxidation
II P
CHCO
of benzene in the presence of a suitable catalyst (see
Appendix Figure D-3). Standard commercial practice is to
conduct the reaction in the vapor phase utilizing a ^O^.-
based catalyst. Benzene is mixed with an excess of pre-
heated air before being admitted to a multitubular catalytic
reactor. Effluent vapors, consisting of maleic anhydride
(MAN), maleic acid, carbon oxides, water, and benzene, are
cooled and passed through a partial condenser and separator,
where the bulk of the maleic anhydride is separated from the
4-40
-------
noncondensibles. The overhead material from the separator
still contains some maleic anhydride; this material is
recovered as maleic acid through absorption in aqueous (or
nonaqueous) solvents. The crude maleic acid is converted to
the anhydride and distilled to meet purity specifications.
Maleic anhydride is also obtained through vapor-phase
oxidation of butenes and as a by-product of phthalic anyhy-
105
dride.	However, these techniques are used less frequently
than the method described above.
Two percent of the U.S. MAN capacity of 485 million
lb/yr is derived from phthalic anhydride by-production and
55
16.7 percent from butenes. Thus, for purposes of cal-
culating benzene emissions, MAN output from these processes
is deducted and only the benzene-derived MAN production
(81.3 percent) is considered.
The most significant process source of benzene emis-
sions is the product-recovery condenser vent. Scrubbers are
usually equipped to recover uncondensed maleic anhydride
before venting effluent gases to the atmosphere. Data from
stack tests at one MAN plant indicate benzene emissions from
the off-gas (with an inoperative incinerator): the total
hydrocarbon level was measured at 1150 ppm (as methane
45
equivalents), with 180 ppm of benzene. A 1972 emissions
inventory of seven other MAN plants, some of which are known
to be equipped with large scrubber units, shows controlled
benzene emissions ranging from 0.06 to 0.20 kg of benzene
per kg of maleic anhydride produced. Monsanto Research
reported total benzene emissions from all plants to be 15.8
8 8
Gg/year (34.8 million lb/year). The reported emission
factor for a plant with a product scrubber is 0.0967 kg per
8 8
kg of product.
4-41
-------
Because of these high rates of emission of benzene and
other pollutants, maleic anhydride production is the largest
stationary source of benzene emissions.
Cyclohexane
Cyclohexane	is manufactured commercially by
hydrogenation of benzene and by distillation from petroleum.
Of 803 Gg (1.7 billion lb) produced domestically in 1975,
approximately 75 percent was produced by catalytic hydro-
5 66
genation of benzene. ' This process is carried out in the
liquid phase by contacting benzene, hydrogen, and recycled
cyclohexane over a platinum catalyst. The yield of cyclo-
hexane is approximately 95 percent.^Although off-gases
may contain benzene, they are rich enough to be sent to the
fuel gas system.*^' ^ Thus, benzene emissions probably are
mainly fugitive. Because of its relative simplicity, this
process appears to be a minor source of benzene emissions.
Cumene
Cumene [CgH^CH(CH )^], a naturally occurring compound
present in many petroleum crudes, is manufactured syntheti-
cally by liquid- or vapor-phase alkylation of benzene with
propylene. It is used predominantly for synthesis of
phenol and acetone. Solid phosphoric acid (SPA) and alumi-
num chloride are used in the most common processes for
commercial production of cumene. The SPA process requires
a reactor and three distillation columns, one each for
rejection of propane, recycle of unreacted benzene, and
rerunning the cumene product to reject a minor quantity of
polyalkylated materials (see Appendix Figure D-4). The
aluminum chloride process is similar, but requires additional
equipment to dry recycled streams and neutralize reaction
products.
-------
Domestic production of cumene in 1975 totalled 893 Gg
5
(1.97 billion lb). Union carbide reports a benzene emis-
-4
sion factor of 2.45 x 10 kg/kg of cumene produced.
Factoring this value with 1975 cumene production data yields
a total estimated annual benzene emission rate of 0.22 Gg/yr
(0.48 million lb/yr).
Phenol
Phenol (CrH_OH) can be manufactured synthetically by
O D
benzene sulfonation, chlorobenzene and toluene oxidation,
cumene peroxidation, and the regenerative Raschig process.
Domestic phenol production in 1975 was 780 Gg (1.72 billion
lb).5 Synthetic processes have long accounted for nearly
all U.S. production. Because the cumene method accounts for
81
over 90 percent of total U.S. phenol capacity, this is the
only mechanism described. This process essentially involves
only two steps (see Appendix Figure D-5). First, air is
introduced into a vigorously stirred emulsion of slightly
alkaline,, aqueous sodium carbonate with purified cumene to
105
produce cumene hydroperoxide.	Then diluted sulfuric acid
is added to effect cleavage of the compound directly into
phenol and acetone.
Atmospheric emission of "spent air" containing benzene
exhausted from the oxidizer is the largest single source of
105
emission from cumene-process phenol plants.	Normally,
effective multiple-stage refrigeration condensation systems
and other equipment (such as scrubbers, activated carbon
absorbers, or incinerators) are used for recovery and
recycle of unconverted cumene (and benzene). One respondent
to a phenol plant survey reported stack test data showing a
3
controlled emission of 1.5 x 10 kg benzene per kg of
product phenol from an oxidation unit at a plant with an
105
average production of 113 Gg (0.25 billion lb) per year.
4-4 3
-------
This plant relies on refrigeration condensation equipment
primarily to remove cumene from the oxidizer off-gas and
only secondarily to control emissions. Another plant in
this survey reported a benzene emission rate of 2 x 10
kg/kg of phenol from the postoxidizer concentration conden-
105
sor vent.	A new phenol plant under construction shows no
44
benzene point-source emissions. Also, Union Carbide
reports that the reported benzene emission rate of 1.5 x
-3
10 kg/kg is one or two orders of magnitude higher than the
37
controlled emission levels it achieves.
Uncontrolled benzene emissions from plants utilizing
toluene oxidation and benzene sulfonation are estimated to
-3	-4
be 1.2 x 10 kg/kg and 1.2 x 10 kg/kg of phenol, respec-
4. • ^ 66
tively.
In estimation of annual benzene emissions from phenol
plants, it is assumed that half of the production capacity
is from controlled operations and that control efficiency is
95 percent. Based on phenol production of 2.2 billion
pounds in 1976 and the following production breakdown by
process, the estimated benzene emissions are as follows:
Process
Portion of
total
capacity %
Phenol
production,
106 lb/yr
Emission
factor
uncontrolled,
ton/ton
Emission
rate,
106 lb/yr
Cumene
peroxidation
89. 2
1962
0.0015
1.62
Toluene
oxidation
5.2
114
0.0012
0.08
Benzene
sulfonation
5.6
123
0.00012
0.01
Total 0.77 Gg/yr (1.71 million lb/yr)
4-44
-------
Chlorobenzenes
Chlorobenzenes (C-C CI ) are manufactured by batch or
b b X x
continuous processes (Appendix Figure. D-6) . In the batch
method, chlorine is bubbled through dry benzene contained in
a cast-iron or steel tank (chlorinator), with iron turnings
as catalyst. The solution is then neutralized with aqueous
(10%) caustic soda and separated into several fractions by
distillation, the bulk of which (75%) is pure chlorobenzene.
Residue from the distillation is the principal source of
para- and orthodichlorobenzenes. Hydrogen chloride formed
in the chlorinator is scrubbed with benzene or chlorobenzene
to remove organic spray and is absorbed in water to give
hydrochloric acid. Neutralization causes the settling out
of a sludge rich in dichlorobenzenes. This sludge is
further processed to recover the para- and orthodichloro-
benzene isomers.
Continuous processing involves a combination of chlori-
nation and fractionation, wherein the monochlorobenzene is
isolated, as quickly as it is formed. Monochlorobenzene
withdrawn from the fractionating column is neutralized and
59
distilleid in the same manner as in the batch process. The
manufacture of paradichlorobenzene, the primary product, is
accomplished by chlorinating benzene in the presence of
aluminum chloride. An alternative process for production of
monochlorobenzene is a catalytic vapor-phase oxychlorination
of benzene with hydrogen chloride and air (Raschig process).
Presumably, the chlorination step is the principal
source of atmospheric benzene emissions in processing of
these chlorinated compounds. No quantitative emission data
are given for specific production plants; however, Monsanto
attributes total annual (1975) ambient benzene levels of
1065, 692, and 529 Mg (1170, 750, and 600 tons) to the
4-4 5
-------
manufacture of chlorobenzene, paradichlorobenzene, and
97
orthodichlorobenzene, respectively.
Based on the assumption that the average facility
operates at 80 percent of capacity, emission factors are
estimated as follows;
Chemical
Annual
capaci ty,
10 6 Kg
Emission,
Gg
Benzene
emission
factor,
kg/kg
Chlorobenzene
303. 9
1. 065
0.0035
Orthodichlorobenzene
69. 4
0.692
0.0100
Paradichlorobenzene
73.0
0. 529
0.0072
Total
446. 3
2. 286
!
0.0051
Detergent alkylate
Detergent alkylate (dodecylbenzene) is the name common-
ly given the product formed from the alkylation of benzene
with dodecene. Any of three catalysts can be used: alumi-
num chloride, hydrogen fluoride, or sulfuric acid. The
product, either branched or linear alkylbenzene (dodecyl-
benzene) , depending on the olefin reactant, is separated
from the catalyst, washed, and purified by distillation.
Linear alkylbenzene is commonly made by alkylating benzene
with a chlorinated n-paraffin at 38 to 66°C with aluminum
chloride catalyst. Dodecylbenzene can be sulfonated into a
surface-active agent.
Purifying the end product by distillation can cause
emissions of airborne benzene from incineration of the light
and heavy organic ends. No data are available on benzene
emissions from production of branched alkylbenzene. The
reported annual (1975) generation of benzene from linear
97
alkylbenzene manufacture was 13.7 Mg (15 ton). The
4-46
-------
omission rate for benzene from a controlled linear alkylate
-4	37
process was reported as 5 x 10 kg/kg of product. Based
on an annual linear alkylbenzene production of 286 Gg (630
million lb), the annual benzene emission is estimated to be
55
144 Mg (0.32 million lb). The latter emission rate is
used in this inventory because it is based on the most
recent emission data.
Solvent operations
Although benzene has been used widely as an industrial
50
solvent," increasing awareness of its toxicity has caused
users to replace it with other solvents for most applica-
40
tions. Because of the relatively poor vapor containment
in most solvent operations, solvent emissions for nonhalo-
92
genated solvents are essentially equal to solvent usage.
In 1966 the Council of Europe recommended that such materials
as solvents, paints, and adhesives should not contain more
than 5 percent benzene (unless enclosed and labeled), with
eventual reduction to 1 percent. In 1971 an International
Labor Office (ILO) conference adopted a recommendation
specifying a ceiling of 25 ppm for workplace exposure to
benzene or products containing benzene at more than 1 per-
cent by volume. It is reported that 0.5 percent would
represent an extremely high estimate of the average benzene
3 8
level in solvents. The continual lowering of allowable
levels of benzene to which workers may be exposed has caused
benzene solvent users to reduce their consumption and/or
institute emission controls.
Degreasing operations using benzene as a solvent are
reported to have consumed as much as 73 Gg/yr (16.10 million
lb/yr), with this amount ultimately being released to the
92
atmosphere. The degreasing operations covered in the
reported figure are as follows:
4-47
-------


92

L
Senzene usage,
Operation
Gg/yi:
rrillion lb/yr
Industrial degreasing


Metal furniture
1 . r
2.2
Primary metals
f ;!
2.2
Fabricated products
1 . 0
2.2
Other miscellaneous
15.0
77.2
Textile plants
3 5.0
77.2
Total
7 3.0
161.0
Textile plants, the largest consumers of benzene as a solvent,
use benzene for fabric scouring or cleaning. It is reported
that the textile industry still uses benzene for this opera-
92 1
tion, even thoaen it now uses Mai>y othar solvents as well. '
Chevron reports bar decjrea&ers generally use either
chlorinated hydrocarbons or higher-boiling solvents in which
benzene would be present only as an impurity. Thus, they
believe that the degreaser contributions to benzene emis-
sions from solvent operations should be very much less than
38
10 percent.
Finally, the re-refining of benzene-containing solvents
is a potential source of benzene emissions. The typical
solvent recovery system (Appendix Figure D-7) consists of a
tank in which solids settle out, a pot sr.ill topped by a
fractionating section, and a surface condenser to distill
and liquify the vapor prior to collection in a holding tank.
A salt such as sodium carbonate is then added to aid in
separating water from the pure solvent. Although no quanti-
tative data are reported, apparent sources of emissions
include the settling tanks during filling and sludge drawoff,
the still during drawoff of bottoms, the product receivers,
and the water jet reservoir (.if vacuum is produced by a
4-43
-------
barometric water jet). By creating a vacuum, the water jet
entraps solvent vapors from the still.^
Other nonfuel uses
Other nonfuel uses of benzene that are regarded as
potentially significant emission sources are listed below,
with annual production rates and estimated atmospheric
emissions:

Annual
production,
97
Benzene emission
Chemical
106 lb/yr
10J kg/yr
10J lb/yr
Fumaric acid
75
300.1
662
Cumene sulfonate
ammonium salt
N.A.
1.6
3.5
Cumene sulfonic acid
N.A.
1.1
2.4
Benzonitrile
N.A.
0.3
0.7
Anthraquinone
N.A.
N.A.
N.A.
Total emission

~ 303
= 669
N.A. = Not available.
In summary, maleic anhydride appears to be the largest
source of benzene emissions among the processes using ben-
zene as an intermediate. The poor quality of the available
data on benzene emissions from many of the processes, par-
ticularly solvent operations, makes it difficult to evaluate
the environmental impacts of these processes.
Hydroscience Inc. is making detailed engineering studies
of the following benzene consumption processes: ethyl
benzene, maleic anhydride, nitrobenzene, cyclohexane,
chlorobenzenes, and alkyl benzenes. Because these studies
will include source testing and evaluation of best control
4-49
-------
technology, their results should greatly improve the reli-
ability of the characterization of benzene emissions from
these processes.
4.2.4 Storage and Transport Emissions
Storage and distribution of gasoline and benzene (in-
cluding 'feedstock' benzene stored by noncaptive consumer
industries) are potentially more significant sources of
benzene emissions than benzene manufacturing processes. For
the purposes of this report, it is assumed that little or no
benzene is emitted in the storage and transport of consumer
chemicals (ethylbenzene, curaene, cyclohexane, etc.) for
which benzene is used as a feedstock in production. This
assumption is justified by the normally low benzene concen-
trations in these chemicals.
4.2.4.1 Storage - Petroleum refinery bulk terminals and
petrochemical plant storage facilities consist of closed-
storage vessels, which include pressure, fixed-roof,
floating-roof, and conservation tanks.
Pressure tanks are designed to withstand relatively
high pressures and normally are used to store volatile
fluids. They are never used for gasoline storage and seldom
for benzene storage.
Ordinary fixed-roof tanks (vertical cylinders) normally
are used to store low-volatility liquids. Many of the flat,
conical, or domed roofs of these tanks have free vents open
to the atmosphere; others have conservation vents that open
at very slight positive pressures. (Appendix Figure E-l
shows a diagram of a typical fixed-roof storage tank.)
Fixed-roof tanks are still used for gasoline storage, but
state regulations require that these tanks be small.
Floating-roof tanks minimize vapor space and normally
are used to store more volatile liquid fractions, such as
4-50
-------
gasoline and benzene, that emit vapors presenting potential
fire or explosion hazards. These tanks are differentiated
by type of floating-roof design: pan, pontoon, or double-
deck. Pan-type tanks have a flat, metal plate roof. Extreme
tilting or holes in the plate may cause it to buckle or
sink, resulting in high losses of vapor. Other losses may
occur from solar irradiation if the metal roof is in direct
contact with the liquid surface. High liquid surface tem-
peratures can cause hydrocarbons to boil and escape through
the opening around the perimeter of the roof. The pan-type
tank, placed in service more than 40 years ago, is seldom
used in new construction. Pontoon-type tanks were designed
to overcome the pan-roof hydrocarbon losses. Pontoon sec-
tions were added to the top of the exposed deck, with a
drain placed in the center to solve any drainage problem.
Ordinarily this roof is commonly used on tanks with very
large diameters. In some pontoon-roof tanks a vapor trap is
installed on the underside to retain vapors formed by the
surface liquid boiling induced by solar heat. Double-deck
floating roofs (Appendix Figure E-2) were introduced to
reduce the effects of solar boiling and to provide rigidity.
In this design, compartmented, dead-air space covers the
entire liquid surface. The bottom deck of the roof is
normally coned upward to trap any vapors entrained with
incoming liquid or any vapors formed earlier in storage.
Conservation storage vessels include lifter-roof tanks
and tanks with internal, flexible diaphragms or internal,
plastic, floating blankets. Appendix Figure E-3 shows a
diagram of a typical lifter-roof tank with variable vapor
space. This type of vessel is similar to the fixed roof
tank and xs used for low-volatility liquids or low-pressure
gaseous products. Fixed-roof tanks with plastic floating
4-51
-------
blankets are effective for controlling evaporation from
stored oil but do not reduce significantly the emissions
from stored gasoline.
Ordinary fixed-roof tanks normally are used to store
less volatile petroleum products. FIcatiny-roof vessels
usually are used to store gasoline. All tanks may lose
vapors at gauging hatches, sample hatches, and relief vents
unless these points are designed and maintained for proper
35
closure. Fmissions of hydrocarbons (including benzene)
from gasoline stored in a floating-roof tank occur primarily
because of improper fit of the seal and shoe to the vessel
shell, which exposes some liquid surface to the atmosphere. ^
Small losses also occur when vapor escapes between the
flexible membrane seal and the roof and during withdrawal.
Withdrawal losses result from evaporation of liquid from the
117
tank wall as the roof descends during emptying operations.
'Breathing' losses caused by the expansion and contraction
of vapors promoted by diurnal changes in atmospheric tempera-
ture are primarily associated with fixed-roof tanks. Because
of the low volume of gasoline stored in fixed-roof vessels,
'breathing' losses are not considered a significant source
of ambient benzene.
Limited data have been reported on benzene exposures
adjacent to storage facilities. A survey of industry by
OSHA (April 19 75) reported an average 375 ppm of benzene
measured next to the sampling port atop a benzene storage
4- 1, 48
tank.
Total benzene emissions from gasoline storage are
estimated ori the assumption that, al l gasoline is stored in
3
floating-roof tanks wj th typical capacities of 8700 m (2.3
21
million gallons). Applying emission relationships pre-
4-52
-------
117
sented by Radian yields estimated hydrocarbon losses
of 132 lb/day per tank during storage (standing) and 0.025
3
lb/10 gallons during withdrawal. Emission factors given
by Radian are based on American Petroleum Institute (API)
data for floating-roof tanks. A subsequent study for the
Western Oil & Gas Association (WOGA) of standing losses from
floating--roof tanks found the hydrocarbon losses to be 58
140
percent of the API levels.	This lesser emission rate may
not be representative of industry-wide practice; thus the
API value will be utilized in calculations.
Our calculations are based on an annual domestic gasoline
3	6 7
consumption rate of 378 Mm (100 billion gallons). It
is assumed that the average tank retention (turnover) time
is 30 days and the average tank is 75 percent full. The
resulting annual hydrocarbon emissions are 25.46 Gg (55.4
million lb) for standing losses and L.l Gg (2.5 million lb)
121
for working losses. Data presented by Runion were applied
to convert to benzene losses. The essentially linear
relationship between benzene in gasoline and the vapor-phase
benzene level is 0.4 x volume percent benzene in the vapor
for each volume percent of benzene in the liquid. Utilizing
*
the national average of 2 percent benzene level in gasoline
yields total annual benzene standing losses of 0.20 Gg (0.44
million Lb) or working losses of 0.01 Gg (20,000 lb). The
calculations are as follows:
*
As developed in Section 4.1.
4-51
-------
Gasoline storaqe standing losses =
(100 x 10 ) (365)	x (132)# x (0. 02) (0. 45)**
(0. 75) (2. 3 x 106) 5 (12) +
= 0.7 7 Gg/yr (1.70 million lb/yr).
-4	3
and the benzene emission factor = 1.66 x 10 lb/10 gal.
Gasoline storage withdrawal losses =
(100 x 109)** x (0.025)++ x (0.02)(0.45)**
(1000)
= 0.01 Gg/yr (20,000 lb/yr).
Benzene normally is stored in floating-roof storage
tanks. Based on a typical storage tank having a capacity
3
of 8700 rrT (2.3 million gallon), the estimated corrected
benzene emission rate would be 4.6 kg/day per tank (10.3
3
lb/day) for standing loss and 0.89 g/day per m (0.0074
lb/day per 10~* gallon) for withdrawal loss.	On the
3
basis of a 5.67 million m (1.5 billion gallon) annual
benzene consumption and a 30-day tank retention time, the
annual benzene losses are 0.12 Gg/yr (0.26 million lb/yr)
for standing and 5.0 Ilg/yr (11,000 lb/yr) for withdrawal.
The calculations follow:
*	67
Annual gasoline usage, gallons/yr.
+ The tank emission factor is converted from lb/day to
lb/year by assuming 12 fillings per year and a 30-day
retention time.
J Average tank size determined from the 2701 terminals
listed and total national gasoline storage capacity of
23.5 Mm^ (6.2 billion gallons) for these terminals in
1967; 21- each tank is assumed to bo 75 percent full.
jj	H7
Gasoline omission, lb/day for this average tank.
~	*
Benzene conversion, 2 percent of gasoline and vapor-phase
concentration at 40 percent of liquid concentration (con-
verted to vapor weight basis, this is 45 percent).
++Withdrawal loss for gasoline, lb/103 gallon.
4-54
-------
Benzene' storage standing losses =
q * + u
(1.5x10) (30)	(10.3)
(0.75) (2.3 x 106)5
- 0.12 Gg/yr (0.26 million lb/yr).
3
and the benzene emission factor 0.01 lb/day per 10
ga1Ion.
Benzene storage withdrawal losses =
(1.5 x lo9)* (0. 0074)** 5.0 Ma/yr = (11, 000 lb/year).
(1000)""
4.2.4.2 Crude Oil Operations - Benzene emissions from crude
oil operations are based upon 0.15 Iv percent of benzene in
117
crude oil.	Radian reports a total evaporative hydro-
carbon emission rate of 0.80 Gg/yr (1.77 billion lb/yr) from
crude oil production and transportation."'""''^ If it is
assumed that the benzene concentration in the hydrocarbons
evaporated is 0.40 of the liquid, the benzene emission rate
is 1.1 million lb/yr. Hydrocarbon emissions also result
from storage of crude oil at the refinery. The estimated
hydrocarbon emissions caused by standing in a fixed-roof,
crude-oil storage tank with a diameter of 150 feet would be
798 lb/day per tank. Based on 0.15 lv percent benzene, 0.40
vapor/liquid concentration, and 4000 crude-oil storage tanks
in use nationally, the total benzene emission rate is 0.3
Gg/yr (0.7 million lb/yr). Crude-oil working losses are
117
estimated to be 3.53 lb hydrocarbons/1000 gallons oil.
Based upon a crude-oil consumption of 4.8 billion bbl/yr and
the 0.15 lv percent benzene concentration in the crude, the
~
1976 benzene consumption, lb/year.
1 30-day retention time.
ti
Arithmetic average calculated from data on old and new tank
operations, assuming tanks are typically 75 percent full.
#	117
Tank emission factor, lb/tank-day.
*	*	3
Benzene withdrawal loss rate, lb/10 gal/day.
4-55
-------
estimated benzene omission loss is 0.2 Gg/yr (0.4 million
18
lb/'yi • Total benzene emissions from crude-oil operations
are estimated to be 1.0 Gg/yr (2.2 million lb/yr). This is
117
not considered an accurate estimate.
4.2.4.3 Distribution - Distribution of gasoline from manu-
facturing facility to the consumer (service station, com-
mercial user, etc.) is also a significant source of atmos-
pheric benzene. An integrated network of pipelines, tank
trucks, rail tank cars, barges, and ocean tankers connects
intermediary storage and loading stations, as illustrated
in Figure 4-9. Intermediate stations, termed "bulk ter-
minals," are distinguished from stations supplied by tank
vehicles, which are called "bulk plants." Service stations
that fuel public motor vehicles are supplied by tank trucks
from either bulk terminals or bulk plants. Privately owned
commercial operations, such as those providing fuel for
vehicles of a company fleet, are generally supplied by tank
trucks from an intermediate bulk installation. Gasoline is
loaded into the transport vehicles at bulk stations and
refineries by means of loading racks.
Working losses, or vapors emitted during filling and
emptying of tanks and tank trucks, occur at a rate directly
proportional to the amount of gasoline charged to either
vessel and to the rate of charging. Hydrocarbon emissions
from tank truck loading racks are potentially much greater
21
than emissions from bulk terminals. As empty tank trucks
are filled, hydrocarbons in the vapor space are displaced to
the atmosphere unless vapor collection facilities have been
provided. The quantity of hydrocarbons contained in the
displaced vapors depends on the vapor pressure, temperature,
method of tank filling, and conditions under which the truck
was previously loaded. Figure 4-10 is a schematic drawing
4-56
-------
SERVICE STATION
BULK PLANTS
COMMERCIAL,
RURAL USERS
SHIP, RAIL, BARGE
PIPELINE
TANK TRUCK
BULK TERMINALS
TRUCKS
AUTOMOBILES
TRUCKS
REFINERY STORAGE
I''i.(jure 4-9. The gasoline marketing distribution
21
system m the United States.
4 - 7
-------
PIPELINE GASOLINE
TO STORAGE-
VENT GAS
LOADING VAPORS
TO RECOVERY UNIT
L RECOVERED
GASOLINE
GASOLINE TO
LOADING
RACK
TERMINAL
TRANSPORT
LOADING
RACK
VAPOR
RECOVERY
UNIT
STORAGE TANK (S)
Figure 4-10.
Source: Ref.
Vapor and liquid flow in a typical bulk terminal
(floating-roof tank).
21.
-------
of liquid and vapor flow through a typical bulk terminal.
Exposures of workers to benzene were measured at
several bulk loading installations in Britain (1969).
Atmospheric benzene was measured by personal activated-
charcoal samplers in 70 total determinations. Normal pro-
cedures for handling of gasolines with benzene contents
ranging from 0.4 to 6.8 percent by volume resulted in
ambient benzene concent rations rang ing from 0.1 to 7.7 ppm;
one sample showed 19,5 ppm. A 1972 study concerned benzene
exposures during loading and weighing of rail tankers with
gasoline from storage tanks. The mean concentration of
benzene vapors escaping open tanker ports was 1.2 ppm,
representing a time-weighted average over an 8-hour workday.
The mean concentration of vapors drifting into the weighing
room was 20 ppm on a time-weighted average basis. This
49
exposure is equivalent to 14 ppm over an 8-hour workday.
Gasoline is loaded from storage tanks to transport
tri-.cks (tank cars) by two basic methods: top loading and
bottom loading. Top loading can be done by splash fill or
submerged fill. In spash filling, the gasoline droplets
undergo a 'free fall' and thus may evaporate or become
entrained in the expelled vapors. In subsurface or submerged
filling, the gasoline is introduced below the liquid surface
in the tank. Bottom loading of gasoline is comparable to
submerged top loading. Hydrocarbon emission factors have
been developed to approximate the amount of uncontrolled
21
gasoline vapor emitted by each of these loading operations.
Our calculations assume a 2.0 percent level of benzene in
~
gasoline and a 0.45 factor for the vapor benzene concentra-
tion based on the liquid benzene concentration.
*
Shell Oil reports a benzene concentration in the hydro-
carbon vapor of 0.1 weight percent for barge loading of
gasoline containing 1,5 weighted percent benzene.^
4-59
-------
Vapor containment efficiency of bottom-loading equip-
ment may approach 100 percent if the truck or rail car has
21
no leaks. Although difficult to quantify, vapor collec-
tion efficiency for top loading is generally lower than for
21
bottom loading. An overall 95 percent efficiency of vapor
recovery/containment is assumed for both loading and unload-
ing .
It is assumed that half of the gasoline is transported
by tank cars or trucks and half by marine vessels and that
the loading of tank cars and trucks is 25 percent by the
splash method and 7 5 percent by the submerged method. Table
4-10 presents estimated benzene emissions from gasoline
transportation operations.
Table 4-10. ESTIMATED BENZENE EMISSIONS FROM
GASOLINE TRANSPORTATION OPERATIONS
Emission
Uncontrolled
emission factor
Benzene emission rate,
source
Gg/yr (million Ib/yr)
Loading-tank
car/truck



Submerged
0.6 kg/103£
(5 lb/103 gal)


Splash
1.4 kg/103£
(12 lb/103 gal)


Marine
0.22 kg/10
(1.8 lb/103 gal)


Loading total

0.43
0. 96
Transit
0.4 kg/wk-103?.
(3 ib/wk- 1 0 3 tjal )
1. 22
2. 70
Unloading
0.26 kg/1031
(2.1 lb/10 3 gal)
0. 04
0. 09

Total
1. 69
3.75
4-60
-------
Service Stations
Filling of underground gasoline storage tanks at
service stations and service station operations together
represent another source of evaporative benzene emissions.
Emissions occur from displacement of hydrocarbon vapors to
the atmosphere while the gasoline is being loaded into the
storage tanks, underground tank breathing, and vehicle
refueling operations. Estimates of benzene emissions from
service stations are based on the rate of gasoline consump-
tion and the benzene levels in gasoline presented earlier.
Table 4-11 itemizes service station emissions. Storage tank
filling and vehicle refueling are the largest sources. A
percentage breakdown of the controls utilized in these
operations is assumed, although no nationwide data are
available to substantiate this assumption.
Within the past 2 years, Gulf Oil Corporation and
others have monitored benzene exposure levels at service
stations. In 1976 NIOSH conducted a brief study to deter-
mine potential exposure of attendants. The five service
stations sampled were selected because of their large
throughputs of gasoline and their location in a warm climate.
Personal charcoal tubes were placed on the attendants. Most
of the monitors indicated that benzene exposures were below
detectable limits; some indicated concentrations ranging
64
from 0.0 3 to 0.12 ppm. A weighted average benzene content
of the gasoline at the five stations was 1.24 lv percent.
A similar study of Gulf Oil service stations and
marketing terminals, conducted in September 1976, showed an
average benzene concentration of 0.32 ppm. The benzene
content of the gasoline in this study was approximately 1.25
lv percent.
Service stations are now an important emission source
category. As popularity of self-service operations increases,
4-61
-------
Table 4-11. INVENTORY OF SERVICE STATION BENZENE EMISSIONS
Emission
source
Hydrocarbon
emission rate,
kg/103x, (lb/103 gal)
throughput^17
Usage,
% of total3
Benzene
emission rate,
Gg/yr (million lb/yr)
Filling tank




Submerged
0. 88
7 . 3
50

Splash
1.38
11. 5
25

Balanced
0. 04
0.3
25
2.70 (5.95)
Underground tank




breathing
0.12
1.0
100
0.41 (0.90)
Vehicle refueling




displacement losses




Uncontrolled
1. 08
9
75

Controlled
0. 11
0.9
25
2.85 ( 6.28)
Spillage*3
0. 084
0.7
100
0.63 ( 1.40)
Total benzene emissions
6.59 (14.53)
These data represent estimated frequency of usage,
k All of the benzene in spillage is assumed to evaporate.
-------
these operations could constitute a major source at which
the population is exposed to benzene,
Transport of Benzene
Benzene is transported principal]y by rail tank cars,
tank trucks, and bargeu on inland waterways. As with gaso-
line, major emissions are assumed to occur during loading
and unloading operations These, emissions may be generated
by top (splash or submerged) and bottom loading of benzene
into transit vehicles or by evaporative loss from spillage
due to overfilling of. car.k cars or accidental discharge of
9 0
benzene retained in transfer lines,
In estimating emissions we assume that half the benzene
is transported by tank car or truck and half by marine
operations. Test data indicate a tank car/truck loading
emission factor of 0.92 lb/103 gallon uncontrolled (we
assume 95 percent control) and a marine loading emission
1	117 * A
factor of 1,5 lb/10 gallon. ' ** These factors yield a
total benzene loading loss of 0.34 Gg/yr (0.76 million
3
lb/year). An assumed transit loss of 0.78 lb/wk-10 gallon
117
leads to an annual emission of 0.53 Gg (1.17 million lb).
Emission data are limited, but i.r. one instance disconnection
of the hose at a loading point resulted in an exposure
48
measurement of 75 ppm benzene. During an industry survey,
OSHA measured exposures at distances? of 7, 15, 30, and 60
meters (25, 50, 100, and 200 feet) from loading facilities
48
for tank cars, tank trucks, and barges. In all cases
exposure was reported to be less than 10 ppm over an 8-hour
period. Depending on the transport means and distance from
the loading point, short-term exposures ranged from zero to
60 ppm. Samples collected directly over hatch openings at
barge and tank-truck loading ports showed average concentra-
90
tions of 100 and 400 opm, respectively.
4-6 3
-------
In summary, floating-roof tanks nm normally used to
store the more volatile petroleum fractions such as gasoline
and benzene. Fixed-roof tanks are still used occasionally
for gasoline storage, but they are being phased out. More
atmospheric benzene is generated annually from gasoline
storage than from benzene storage because a much higher
volume of gasoline is manufactured. Crude oil operations
may represent a significant source of benzene emissions,
but reliable data are not available for estimating. Both
benzene and gasoline are transported by tank truck, rail
tank car, and marine vessel. This study indicates that in
the handling and distribution of gasoline the greatest
quantity of atmospheric benzene is emitted from underground
storage tank filling and refueling of motor vehicle tanks at
service stations. Lesser amounts are emitted during highly
efficient, vapor-contained loading and unloading of bulk
gasoline, as 'breathing' losses from service station under-
ground tanks, and by evaporation of gasoline spillage.
Table 4-12 presents benzene emission data and overall ben-
zene emission factors for each source category.
4.2.5 Other Miscellaneous Sources
Hydrocarbons emitted because of incomplete combustion
in fuel-burning equipment present a potential source of
atmospheric benzene. Combustion efficiency governs the
amount of hydrocarbon (and therefore benzene) emissions, and
this efficiency depends on the operation and design of the
combustion equipment. No emission data are available for
this source.
Although still in the pilot staqe, coal gasification
plants present a definite potential for benzene emissions.
The light-oil product from a Hy-Gas plant, a potential coal
8 6
gasification process, contains about 8 percent benzene.
4-64
-------
Table 4-12. SUMMARY OF BENZENE EMISSION FACTORS AND ANNUAL
EMISSIONS FROM STORAGE/DISTRIBUTION OF
GASOLINE AND BENZENE

Benzene
emission factor,
kg/m' (lb/1000 gal.)
Annual
benzene emission,
Gg (106 lb)
Storage




1. Gasoline: standing
3.3 x 10~5
(2.9 x 10~4)
0.B
(1.7)
Withdrawal
2.6 x 10"5
(2.2 x 10"4)
0.0
(0.0)
Total

0.8
(1.7)
2. Benzene: standing
1.3 x 10"5
(1.1 x 10"4)
0.2
(0.5)
Withdrawal
8.6 x 10~4
(7.4 x 10~3)

(0.0)
Total

0.2
(0.5)
Crude oil operations




1. Production and
Transport


0.5
(1.1)
2. Refinery storage -
Standing


0.3
(0.7)
Withdrawal


0.2
(0.4)
Total


1.0
(2.2)
Distribution systems




1. Gasoline losses




Loading
1.1 X 10~4
(9.6 X 10~4)
0.4
(1.0)
Unloading
1.1 x 10"5
(0.9 x 10~4)
0.0
(0.1)
Transit
0.4/week
(3.0/week)
1.2
(2.7)
Total

1.7
(3.8)
2. Benzene losses




Loading
6.0 x 10'5
(5.1 x 10-4)
0.3
(0.8)
Unloading
low -
unknown
low
Transit
0.1/week
(O.B/week)
0.5
(1.2)
Total


0.9
(2.0)
Service Station Gasoline




Loading
6.9 x 10"3
(5.9 x 10"2)
2.7
(6.0)
Refueling vehicles
7.3 x 10"3
(6.3 x 10"2)
2.9
(6.3)
Underground tank




breathing losses
1.1 x 10"3
(0.9 x 10"2)
0.4
(0.9)
Gssoline spillage
1.6 x 10"3
(1.4 x 10"2)
0.6
(1.4)
Total

6.6
U4.6)
TOTAL

(24.*)
4-65
-------
Other potential sources of benzene emission include
solid waste incineration, forest fires, structural fires,
coal refuse fires, and agricultural burning. No data are
available on benzene emissions from these sources.
Investigation into emission of atmospheric benzene from
sanitary landfills appears warranted. In a March 1976 study
conducted at a site where large quantities of chemical
wastes were being dumped, measurements of ambient benzene
3
ranged from trace amounts to 1.5 5 mg/m , depending upon
72
proximity to the landfill. * Samples were taken both
upwind and downwind of the site within an approximate 1-mile
radius. A single measurement was made directly atop the
dump mound adjacent to the liquid waste pits. Samples were
analyzed by gas chromatography with mass spectrometry.
Evaluation of the data showed that average benzene concen-
trations were higher downwind. Analysis of the single
sample taken atop the mound gave a benzene concentration of
9 00 ug/m^.
4.3 BENZENE EMISSION SOURCE CHARACTERISTICS
This subsection provides a brief summary of available
data from major benzene emission sources. Table 4-13 pre-
sents benzene emission factors for major stationary sources
of various capacities. Maleic anhydride plants and coke
oven batteries have by far the highest emission factors.
Table 4-14 summarizes emission factors for other air
pollutants from major benzene emission sources. This infor-
mation is meager for many of the sources.
Table 4-15 presents estimated emissions of benzene and
other pollutants from a typical facility using benzene as an
intermediate. These data were developed by incorporating
plant size and emission factors (where listed) from previous
tables. The data show that maleic anhydride is the most
significant individual source of emitted benzene and other
pollutants.
4-66
-------
Table 4-13. STATIONARY SOURCE EMISSION CHARACTERISTICS - 1976 BASIS

Benzene
i




emission
!



factor,
1 Plant capacity
106 lb/yr
Total no. of


kg/kg
• -
. . 	1
faci1ities
Emission

10-3
t Largest ;
Average
nationally
j characterization
Nitrobenzene
7.0
310
100
9
,Point-absorber
Aniline
c
' 200
100
7
;Fugitive
Ethylbenzene





- Monsanto
0 . 62
1965
55C
19
!Scrubber-vent
- Alkar





- Mobi1/Badger

;


i
Styrene
1.50
j 1300
460
13
Column ve.'it, reactor et^ergoncy





vort
Maleic anhydride
96 . 7
! 105
50
9
1 Product recovery ?crubber
Cyc iohexane
c
85
45
1 0
• Fi.i g i t \ ve
C'jjnene
0.25
650
270
14
'Fugitive
F h e p. o 1 —C ut*. e n e a
: . 5
| 5 00
200
14
' Oxi J . \ , wc = t•.'xidat *on condenser
Toluene
1 . 2
i 135


j c
Ben 2ene
0.1
; 48


; c
Chlorobenzene
3.5
> 300
100

' c
o-dichl
10.0
25
12

1 c
p-dichl
7 . 2
! 60
i
25
13
'Chlorinator, PDCD rec. sysrem
Solvent
c


c
;Evaporat ion
Other nonfuel
d
'

d

Ref iner:es
Low
I

248
Fugitive
Coke ovens
-2*
7.6 x 10
1 1

66
i
[Primarily fugitive
t
a
Fmissinn factors are for uncontrolled operations. Fmis?ion calculated for each process,
assuming half of throughput was controlled at 95%.
Based on coal charged to coke-oven batteries.
C Data unavailable.
No overall value available.
-------
Table 4-14. OTHER AIR POLLUTANTS EMITTED FROM BENZENE SOURCES
Benzene source
Pollutant
Emission factor,
kg/kg product
Reference
I
T\
cu
Gasoline-powered motor vehicles
Diesel heavy-duty vehicles
Benzene production sources
Petroleum refineries
Coke-ovens
Benzene consumption sources
Nitrobenzene
Aniline
Ethyllienzene
Styrene
HC/CO/NO^
HC/CO/NO
Mixed HC
SO 2
Particulates
S02
CO
nh3
h2s
HCN
N0x
Phenols
PNA particulates
N0X
Nitrobenzene
d
d
Toluene
Ethylbenzene
Styrene
Nonaromatic HC
kg/kg of coal feed,
k No emission estimates available.
C Low level.
No data on pollutant identification.
(Continued)
4.0
9.5
1.8
1.9
1.9
2.7
2.5
2
2
5
9
,-3a
x 10l4a
x 10 4
x 10"4a
x 10-4a
x 10"53
x 10"5a
x 10_4a
9 x 10
c
-5
x 10
x 10
x 10
x 10
-4
-5
-5
-5
31
31
5 5,94
!5 , 94
! 5 , 9 4
85
94
66
66
105
105
105
105
-------
Table 4-14.(continued).


Emission factor,

Benzene source
Pollutant
kg/kg product
Reference
Maleic anhydride
Formaldehyde
6.3 x 10"3
88

Formic acid
2.1 x 10"3
88

Maleic acid
3.1 x 1Q-3
o o

Maleic anhydride
3.2 x 10~3
88

Xylene
11.7 x 10"3
88

Total HC
106.6 x 10"3
88

CO
633.3 x 10"3
88

NO 2
0.3 x 10 3
88
Cyclohexane
CO, sulphur
c
66
Cumene
Cumene
7.3 x 10~4
44

Formaldehyde
1.3 x 10"4
44

Acetaldehyde
1.6 x 10-5
44

Methanol
1.6 x 10"5
44
Phenol
Acetaldehyde
9.0 x 10~4
105

Formaldehyde
3.2 x 10~J
105

Acetone
1.2 x 10-3
105

EthyIbenzene
2.0 x 10-4
105

Cumene
2.0 x 10-3
105

Methyl Styrene
4.3 x 10
44

Phenol
1.2 x 10-6
44

Mesityloxide
5 x 10"8
—
105

Toluene
4 x 10"5
105
Chlorobenzene
d


a kg/kg of coal feed.
No emission estimates available.
c Low level.
^ No data on pollutant identification.
(Continued)
-------
Table 4-15. ESTIMATED EMISSIONS FROM AN AVERAGE FACILITY
USING BENZENE AS AN INTERMEDIATE
Benzene
source
Emissions
Benzene,
Gg/yr (10 lb/yr)
Other pollutants,
i Gg/yr (10 lb/yr)
Nitrobenzene
Ani1ine
Ethy lbcMizene
Styrene
Maleic anhydride
Cyclohexane
Cumene
Phenol
Chlorobenzene
p-dichlorobenzene
Detergent alkylate
Solvent operations
0.15
0. 31
2.20
0.14
(0.34)
(0.69)
(4.84)
a'
0.03 I (0.07)
(0.30)
0.16 (0.35)
0.08 (0.18)
NOx
Nitrobenzene
0.00 | (0.01)
b i
To]uene
0.04
(0.09)
Ethylbenzene
0.00
(0.01)
Styrcnc-
0.01
(0.02)
Nonaromjtic lie
0. 02
(0.04)
Formaldehydc
0.14
(0.32)
Formic acid
0.05
(0.11)
Maleic acid
0 . 07
(0.1C)
Maleic anhydride
0 . 07
(0.16)
Xylene
0.27
(0.59)
Total HC
2 . 20
(5.33)
CO
14.37
(31.67)
no2
0.01
(0.02)
CO, sulfur
b

Cumene
0.09
(0.20)
Formaldehyde
0 .02
(0.04)
Acetaldehyde
b

Methanol
b

Acetaldehyde
0.08
(0.18)
Formaldehyde
0.29
(0.64)
Acetone
0.11
(0.24)
Ethylbenzene
0.02
(0.04)
Cumene
0.18
(0.40)
Methyl Rtyrene
0. 04
(0.09)
Phenol
b

Mesityloxide
b

Toluene
0.00
(0.01)
p-dichlorobenzene
Chlorobenzene
Dichlorobenzene
HCj, Cl2
Toluene
Xylene
| Other HC
b
b
b
b
No emission estimates available.
Low level.
No data on pollutant identification.
Phenol production from cumene assuming uncontrolled operation.
4-70
-------
5.0 CONTROL METHODS
Specific information on the effectiveness of the con-
trol methods for benzene emissions is very limited. Thus,
attention in this section is directed primarily to hydro-
carbon control. Descriptions of methods are somewhat
general. Hydrocarbon control on automobiles is achieved by
positive crankcase ventilation, evaporative emission control,
catalytic oxidation, and engine modification. Benzene
vapors are recovered at stationary sources of benzene pri-
marily by absorption or condensation.
5.1 EMISSION CONTROLS FOR MOBILE SOURCES
Although some data are available on the effectiveness
of systems for direct control of benzene, extensive research
has been aimed at control of hydrocarbons (HC). Thus, this
section focuses on the effectiveness of HC control by
systems now under development. Attention is given also to
emission levels of carbon monoxide (CO) and nitrogen oxides
(NO^), because regulation of these emissions is mandatory
and their control influences the level of HC and benzene
control.
Emission control systems in use on current-model motor
vehicles include positive crankcase ventilation (PCV), ex-
haust gas recirculation (EGR), evaporative controls, and
catalytic oxidation systems. PCV and evaporative controls
minimize hydrocarbon emissions at the source, whereas
catalytic oxidation reduces hydrocarbon emissions in the
exhaust by oxidizing unburned hydrocarbons. In addition,
engine modifications are made to minimize formation of
contaminants in the cylinders. These modifications include
5-]
-------
improved air/fuel mixing and distribution systems and al-
teration of ignition characteristics.
113
5.1.1	Catalyst Technology
Several types of catalytic converters are in use or
under development. The oxidizing catalytic converter, a
device in the exhaust stream that stimulates oxidation of
unburned fuel, changes HC and CO emissions into harmless
carbon dioxide (CC^) and water. The reducing type of cata-
lytic system stimulates conversion of NO into nitrogen gas
(N2). These two converters can be combined in a dual-bed
system, as shown in Figure 5-1.
The third basic type of catalytic system is the three-
way catalyst, which appears to be in an advanced state of
development in California. In this system HC, CO, and NO
are converted in one catalyst bed.
Catalyst-engine matching has also been improved signif-
icantly by feedback control systems that closely control
air/fuel ratios. Oxygen sensors respond to excess oxygen in
the exhaust gases and signal the carburetor or the fuel
injection system to control the air/fuel mixture.
Table 4-6 (subsection 4.1.1) shows how effective
catalytic converters are in reducing hydrocarbon and benzene
emissions. Total HC emissions range from 0.133 to 0.489
g/km (0.214 to 0.788 g/mile), with benzene emissions gen-
2
erally below 6 mg/kg (0.01 g/mile).
113
5.1.2	"Lean Burn" Technology
Engines calibrated for very lean air/fuel mixtures
achieve low emissions of CO, HC, and sulfuric acid even
without a catalytic system. An advanced form of "lean burn"
technology would utilize various sensors, including an
oxygen sensor, a mini-computer, feedback circuits to opti-
mize spark timing, exhaust gas recirculation rate, air/fuel
5-2
-------
UNTREATED
EXHAUST
TREATED
EXHAUST
FINAL
EXHAUST
INTAKE
AIR I FUEL
/CO + HC *\
CO + HC + NO
WITHIN LEGAL
LIMITS
IN EXCESS
OF LEGAL
LIMITS
COMBUSTION
OXIDIZING
CATALYST
REDUCING
CATALYST
Figure 5-1. Operation of a dual-bed catalyst.
Source: Ref. 113.
-------
ratio, and transmission shift points. These improvements
might cost no more than current catalyst technology and
would control emissions to the 1975 California levels
(HC/CO/NO - 0.9/9.0/2.0 g/mile), with little or no loss of
fuel economy as compared with uncontrolled cars. Reductions
of emissions much below these levels will most likely re-
quire a catalyst in conjunction with some or all of the
"lean burn" system components. One domestic manufacturer is
actively considering full-scale production of lean-burn-type
vehicles.
11 3
5.1.3 Exhaust Gas Recirculation
Recirculating exhaust gas through the intake system
controls NO^ emissions by reducing the peak combustion
temperature. The exhaust gas, an inert substance, does not
contribute to the combustion process. Refinements of EGR
systems have been made in the laboratory, the most signif-
113
icant recent work being that of Gumbelton.	He reports an
EGR system yielding NO levels of 0.62 g/km (1.0 g/mile)
X
from full-size cars without fuel economy penalties. The
conditions that promote low NO levels, however, also
promote high HC emissions. Achieving low NO^ emissions can
increase HC emissions to the point that 0.41 HC standard is
more difficult to meet than the 0.4 NO standard.
x
It is clear that optimum EGR operation will require
more advanced systems than are currently available. Labo-
ratory work on electronically modulated EGR and spark-timing
systems has been reported as encouraging, but no commitments
have been made to produce such systems.
11 3
'j. 1.4 Uuestor System
The Questor system combines high-temperature oxidation
of HC and CO with catalytic reduction of NO^. This system
5-4
-------
has shown considerable potential for achieving the 0.41/3.4/
0.4 standards. In tests of standard-size cars by several
manufacturers, emissions were reduced below the standards.
Problems remain, however, with the degree of mixture enrich-
ment and the high temperatures required to achieve adequate
control of HC, CO, and NO . The rich mixtures used with the
x
Questor system increase fuel consumption and exhaust tempera-
tures, causing degradation of the system with use. Over 3
years, however, the fuel economy associated with the Questor
system has been improved to a position of parity with 1974
models (13% lower than in 1975 models).
5.1.5 Alternative Engines ^
Several alternatives to the conventional internal
combustion engine now under development may have some
bearing on emissions in the next few years.
The stratified-charge engine uses a layered fuel
mixture in the combustion chamber. The idea is to provide a
fuel-rich mixture near the point of ignition in the cylinder
while keeping the rest of the mixture lean.
Honda's CVCC stratified-charge engine has been shown to
be capable of bringing even full-size cars into compliance
with 0.41/3.4/2.0 standards with essentially no increase in
fuel consumption over the total of average 1975 cars.
Ford's Proco stratified-charge engine uses an open
combustion chamber different from the Honda prechamber
system. The Proco can achieve NO^ levels of 0.4 in 2.0-Mg
(4500-lb) cars without catalytic control of NOx because of
the combination of stratified-charge combustion and high
EGR. HC emissions have presented the greatest problem.
Although the Proco vehicle can simultaneously achieve
0.41/3.4/0.4 with oxidation catalysts, its HC emissions
levels have exceeded 0.41 before 40,000 km (25,000 miles) at
-------
the 0.4 NO^ calibration. Without controls, the Proco engine
uses substantially less fuel than conventional engines, but
the HC emissions are high. So far, reducing HC levels has
required throttling, which makes the engine's fuel consump-
tion comparable to that of conventional engines. Once
throttling is used to control HC emissions, further measures
to reduce NO^ have little effect on fuel consumption. At
present, there are no production versions of the Proco
engine.
The rotary engine uses rotating drive elements to
replace the reciprocating pistons of current internal
combustion engines. In the past 5 years Toyo Kogyo has made
major improvements in control technology for engines. Cur-
rent prototypes of stratified-charge rotary engines show
potential for meeting the 0.41/3.4/0.4 standards. Pro-
totypes have achieved 0.33/1.7/0.38, while consuming less
fuel than 1974 and 1975 production versions of the engine.
Diesel-engine passenger cars currently produced by
several manufacturers show potential for meeting the 1978
standards for HC and CO. The automotive industry is studying
greater use of diesel engines in passenger cars. It appears
that diesel-powered passenger cars can be designed to
achieve emission levels of 0.41 HC and 3.4 CO, with NO
x
levels in the range of 1.0 to 1.5. The prospects for
meeting 0.4 NO with a diesel automobile are not known but
x
are considered unlikely.
Projects for development of Rankine cycle, gas turbine,
and Stirling cycle engines have been initiated. Although
these systems offer long-term potential for meeting very low
emission standards, their large-scale manufacture will be
later than that of the other systems discussed here, cer-
tainly after 1980.
5-6
-------
33
5.1.6	Positive Crankcase Ventilation (PCV)
PCV systems circulate air through the crankcase to pick
up crankcase blowby gases and carry them to the intake
manifold,, where they enter the combustion chamber. The PCV
system, as shown in Figure 5-2, is equipped with a control
valve that restricts air flow when intake manifold vacuum is
high but allows flow at idle conditions. In a closed sys-
tem, the blowby gases, rich in hydrocarbons, are conducted
to the air cleaner through the intake air hose. In an open
system, the gases are vented to the engine compartment
through the crankcase breather cap. When operating prop-
erly, these systems virtually eliminate hydrocarbon emis-
sions, including benzene.
5.1.7	Fuel Evaporative Controls
Fuel evaporative control systems were installed on all
1971 model vehicles. Gasoline tanks and carburetors are the
major sources of evaporative emissions. The most common
control technique is the adsorption-regeneration system,
which virtually eliminates evaporative hydrocarbon emis-
sions, including benzene.
The vapor-recovery system, in which the crankcase is
used as a. storage volume for vapors from the fuel tank and
carburetor, is not currently used.
In the adsorption-regeneration system, a canister of
activated carbon traps the vapors and holds them until such
time as they can be fed back into the induction system for
burning in the combustion chamber. During a hot-soak
period, vapor from the fuel tank is routed to a condenser
and separator, and liquid fuel is returned to the tank. The
remaining vapor, along with fuel vapor from the carburetor,
is vented through a canister filled with activated carbon,
which traps the fuel vapor. When the engine is started,
5-7
-------
AIR CLEANER HOSE	AIR CLEANER HOSE
CLOSED OIL
^FILLER CAP
CLOSED OIL
FILLER CAP
VENTILATOR VALVE HOSE
VENTILATOR VALVE
HOSE
CYLINDER
--HEAD
COVER
CYLINDER HEAD
CO i/ER
VENTILATOR VALVE
ASSEMBLY
V-8 ENGINES
SIX-CYLINDER ENGINES
Figure 5-2. Closed positive crankcase ventilation system.
Source: Ref. 33.
-------
fresh purge air drawn through the canister strips the acti-
vated carbon of the trapped fuel vp.por and carries it to the
33
combustion chambers. Figure 5-3 shows a typical adsorp-
tion-regeneration system employing a three-way check valve.
The following excerp': summer!jses results of a 1976
assessment of evaporative ront.ro 1. .systems by Exxon Research
30
and Engineering Company.
A new test method, known at	SHED method, has shown
that the evaporative emission controls in the field are
less effective than originally estimated and in the
case of some critical late model vehicles, evaporative
emissions are of the sam-i magnitude as the exhaust hy-
drocarbon emission.,.- The average evaporative loss for
a 20 car group was 8.7 grams/SHED test. The lowest
loss was 0.5 gram and the highest was 30.6 grams....
Hardware was developed to improve ECS performance. Six
vehicles were modified to demonstrate the feasibility
of improving current systems. The modifications
involved: (1) the venting of the carburetor bowl to
the canister to alleviate air cleaner overflow, (2)
utilization of larger carbor; beds, (3) adaptation of
increased purge rates, and (4) sealing and capping of
leak sources. These modifications were successful in
lowering the evaporative emissions to 2.0 grams/SHED
test or lower for each of the six modified vehicles.
We have concluded from this work that it is feasible to
markedly improve the performance of current evaporative
control systems,
5.1.8 Current Status of Mobile Erniesion Control
Table 5-1 presents a percentage breakdown of controls
used on cars in operation in 1976.
Table 5-1. ESTIMATED USAGE OF CONTROL DEVICES
ON CARS IN OPERATION IN 197695
Device
No. of cars
Percentage
Crankcase controls
89,234,000
94
Exhaust controls
69,076,000
73
Full evaporative
controls
45,213,000
47
No controls
3,998,000
4
5-9
-------
VAPOR separator
CARBON CANISTER STORAGE
Figure 5-3. Adsorption-regeneration evaporative emissions control system.
Source: Ref. 33.
-------
5.2 EMISSION CONTROLS FOR STATIONARY SOURCES
In recent years, regulations on control of organic
emissions from stationary sources have been limited to those
compounds that are classified as photochemically reactive.
Benzene is classified as a nonphotochemically reactive
compound.
Benzene producers and consumers whose operations
constitute distinct point sources of benzene (usually chemi-
cal reactors) usually recover vapors by absorption or con-
densation. However, recovery is for economic reasons and
can be considered as part of the process rather than as an
air-pollution control. When emissions involve occupational
health hazards or local nuisances (odors), more stringent
control of organic vapors (benzene) can be achieved by
incineration or adsorption.
This section describes control of benzene emissions
from gasoline distribution systems; coke ovens; and produc-
tion of cumene, styrene, maleic anhydride, and paradichloro-
benzene.
5.2.1 Status of Source Control
Little information is available on the usage of air
pollution control devices in benzene-producing and -consuming
industries. Table 5-2 indicates the typical emission
controls for the major point sources of benzene in these
industries.
The most common control systems in benzene production
and the major processes using benzenes are absorption and/or
condensation, which are practiced primarily to recover the
valuable organic compounds.
Fugitive gaseous emissions from leaky valves and pumps
and evaporative emissions from storage tanks and water
5-11
-------
Table 5-2. CONTROL METHODS FOR MAJOR POINT
SOURCES OF ATMOSPHERIC BENZENE
Product
Process
Control method
Nitrobenzene
Nitrator
Absorption with water
Aniline
Reactor
N/A
Kthylbenzene
Reactor
Condensor/absorption
Maleic anhydride
Hydrogenation
reactor
Condensor/scrubber
Cyclohexane
Hydrogenation
reactor
N/A
Phenol
Oxidizer
Multistage condensers
with scrubbers, adsorp-
tion, or incineration
Chlorobenzene
Chlorinator
absorption tower
Absorption with water
Paradichlorobenzene

Wet scrubbers or
absorption columns
Cumene
Oxidizer vent
Condensor
Styrene
Benzene-toluene
recovery
Primary distilla-
tion
Stripper vent
Condensor or absorbers
Condensor
N/A
N/A - Not available.
5-12
-------
treatment and collection systems can be significant and
generally are difficult to control.
5.2.2 Control Methods
Methods applied commercially to control organic emis-
sions (benzene) and/or to recover valuable products are
described in this section. Control measures include ab-
sorption,, adsorption, condensation, and incineration.
Valves, pumps, and piping can release fugitive organic
emissions in petroleum refineries and chemical processing
plants as a result of leakage caused by vibration, heat,
pressure, corrosion, and improper maintenance. Inspection
and preventive maintenance can help to reduce these emis-
sions. Also, some pump designs include special seals for
minimizing leakage of volatile organic vapors.
By far the most important control technique is to
design basic equipment so as to make the most efficient use
of the materials being processed.
34
5.2.2.1 Absorption
Introduction - In the absorption process, a soluble
component of a gas mixture is dissolved into a relatively
nonvolatile liquid. In addition to being dissolved, the gas
mixture :nay react chemically with the liquid. This tech-
nique is common and profitable in petroleum and petro-
chemical operations in which a gas includes a relatively
high concentration of solvent vapor.
Absorption of hydrocarbons is an important manufac-
turing step in the petroleum and petrochemical industries.
This type of absorption must be considered as a preliminary
step in air pollution control, because it does not usually
remove enough of the pollutant material to provide final
control.
5-13
-------
Common absorbents for organic vapors are water, mineral
oil, nonvolatile hydrocarbon oils, and aqueous solutions
(e.g., solutions of oxidizing agents, sodium carbonate, or
sodium hydroxide). Trace benzene emissions are absorbed
primarily with water; however, they sometimes are absorbed
with toluene or ethanol.
Types of Absorbers - Gas absorption equipment is de-
signed to provide thorough contact of the gas and the liquid
solvent, enough to permit interphase diffusion of the ma-
terials. Contact is provided by several types of equipment:
bubble-plate columns, jet scrubbers, packed towers, spray
towers, and venturi scrubbers.
34
5.2.2.2 Condensation
Introduction - Condensation and subsequent removal of
organic compounds is a proven method of reducing organic
emissions. At a given temperature, if the partial pressure
of a compound is increased until it is equal to or greater
than its vapor pressure at that temperature, the compound
will condense. Alternatively, condensation will occur if
the temperature of a gaseous mixture is reduced to a satura-
tion point at which the vapor pressure equals the partial
pressure of one of the constituents. In most air pollution
control applications, condensation is effected by reducing
the temperature.
Because of their relatively high boiling points, many
organic compounds condense readily even though they are not
highly concentrated. Benzene, however, has a relatively low
boiling point of 80°C.
Condensers are applied widely in the organic chemicals
industry to condense concentrated vapors in the primary
process rather than to reduce contaminant emissions. Ap-
plied in the primary process, they allow recovery of valu-
able products.
5-1 4
-------
Control of an organic emission by condensation is
limited by the equilibrium partial pressure of the organic.
Figure 5-4 shows a vapor pressure curve for benzene. As
condensation occurs, the partial pressure of the material
remaining in the gas phase decreases rapidly, making com-
plete condensation impossible. For this reason, condensers
usually must be followed by a secondary control system such
as an afterburner, which combusts the noncondensible gases
and provides a high degree of overall efficiency.
Condensers are of two types, surface and contact. In a
surface condenser, the vapor to be condensed and the cooling
medium are separated by a metal wall; in a contact conden-
ser, the vapor and the cooling medium are brought into
direct contact. Surface condensers include the common
shell-and-tube heat exchangers. In these devices, the
cooling medium, usually water, flows through the tubes, and
vapor condenses on the outside surface. The condensed vapor
forms a film on the cool tubes and drains to a vessel for
storage or disposal. Air-cooled condensers are usually
constructed with finned tubes, and the vapor condenses
inside the tubes.
Contact condensers cool the vapor by spraying a cold
liquid, usually water, directly into the gas stream. The
condensed vapor and water mixture are then treated and
disposed of or recovered. Contact condensers usually are
less expensive, more flexible, and more efficient in re-
moving organic compounds than surface condensers. Examples
of contact condensers are the steam or water ejector and the
barometric condenser, in which the condensing vapors create
a negative pressure, inducing the flow of additional vapor
from the process.
5-15
-------
TtK?tMTU»f. f
32	W_W 70..60. 90 IfO _r. 1£Q._ . UQ_. . 1 op	]^0. ?00 ..
10-
J 50r
100
TlMPfMluRr, C
Figure 5-4. Vapor pressure of liquid benzene
at various temperatures.
Source: Ref. 160.
5-16
-------
The main design factors to be considered for conden-
sation of organic compounds are the type(s) of compounds and
their temperature, volume, concentration, vapor pressure,
and specific heat. In streams containing a variety of
hydrocarbon species, condensor design is based on the
component with the highest boiling point.
34
5.2.2.3 Incineration - Incineration controls organic
emissions by combustion. The objective is to oxidize com-
pletely the organic vapors and gases emitted from a process
or operation. Emissions, of course, include particulate as
well as gaseous matter. Combustible particulates may be
controlled by incineration. Incinerators are used widely
and successfully for reducing organic emissions. They also
offer potential for heat recovery.
Devices in which dilute concentrations of organic
vapors are burned by addition of fuel are known as after-
burners. Devices for burning of waste gases having enough
her.ting value to burn without added fuel are known either as
flares (if gases are not premixed with air) or incinerators
(if gase3 are premixed with air). Combustion in after-
burners is accomplished either by direct-flame incineration
or by catalytic oxidation. Under the proper conditions, the
firebox of a process heater or boiler may also be used as an
afterburner.
Attempting to control organic vapors that contain
halogens or sulfur by means of combustion alone is not
considered wise because the combustion products of such
materials are usually undesirable and often corrosive.
Removal of these contaminants may require a secondary con-
trol system, such as a scrubber.
5-17
-------
Properly designed and operated direct-flame after-
burners usually achieve more than 95 percent efficiency in
removal of organic vapors. Burner utilization must be
considered in conjunction with the overall combustor design
to achieve the optimum temperature and turbulence.
Catalytic afterburners are much like direct-flame
types, but they provide a solid active surface on which the
combustion reaction takes place, usually at a significantly
lower temperature than is required for combustion by direct
flame. Because they can be operated at much lower tempera-
tures and are self-sustaining, catalytic afterburners offer
the advantage of lower fuel costs in some applications.
Fuel requirements for a direct-flame afterburner can be
reduced by means of a heat exchanger, however, and the total
cost of operating a direct-flame afterburner then may be
comparable to the cost of operating the catalytic type. The
primary problems with catalytic afterburners are relatively
high maintenance costs and catalyst deactivation by feed
contaminants. These devices will not function if the
catalyst becomes fouled by particulate matter or coated by
polymers. Catalytic incineration typically reduces emis-
sions 95 percent or more.
An afterburner that does not achieve substantially
complete combustion of the organic emissions can produce CO,
aldehydes, and other partially oxidized substances, which
may be more objectionable than uncontrolled emissions.
34
5.2.2.4 Adsorption - The condensation and retention of a
gaseous or liquid substance on a solid surface is known as
adsorption. Gas-puri f icration processes involving this
principle arc? based upon the physical properties of certain
granular solids, known as adsorbents. These adsorbents
attract selected components of a fluid to their surfaces and
5-18
-------
retain them there. Certain materials, such as activated
alumina, silica gel, and activated carbon, have been de-
veloped 1-0 adsorb substantial quantities of gases and
liquids. These materials are highly porous, with very high
ratios of surface to volume. A liquid can penetrate the
material and contact the large surface area available for
adsorption. Complete package adsorption systems are avail-
able from a number of manufacturers.
In vapor-phase adsorption (essentially an exothermic
gas-solid equilibrium process), the approach to equilibrium
is governed by the rate of adsorption. Conditions that
shift the equilibrium toward saturation usually improve the
process. Consequently, the system is more efficient near
the dew point of the adsorbate (substance being absorbed),
and a vapor-phase adsorption system generally should operate
at the highest pressure and the lowest temperature within
the process limitations.
After some period of usage, the adsorbent becomes
saturated with the contaminant and will no longer function.
When this occurs, it must be regenerated or replaced.
Regeneration may be done in a number of ways. The most
common method is to withdraw the adsorbed gas in a stream of
easily condensible gas such as steam. The stripped gas is
then recovered by condensing the mixture. Thus, the adsor-
bent is restored to activity, and the adsorbed material is
removed for disposal or recovery.
If the gases or vapors to be adsorbed consist of sev-
eral compounds, adsorption of the various components is not
uniform. Generally, these components are adsorbed in an
approximately inverse relationship to their volatilities.
5-19
-------
Although fixed-bed adsorbers generally are not in-
stalled to recover organic solvent vapors when the vapor-
laden stream contains less than 0.2 pound of solvent per
1000 scf of gas (2700 ppm), they can recover much lower
concentrations very efficiently. The reason is that adsorp-
tion is virtually complete and independent of concentration;
however, the maximum bed loading (saturation point) is
dependent on initial concentration.
In a certain range of vapor concentration (from a few
ppm to about 1000 ppm), organics cannot be recovered prof-
itably with either regenerative or nonregenerative adsorp-
tion. In this range, adsorption is often unsuitable for
nonregenerative systems because of the high cost of adsorb-
ent replacement. It is also uneconomical for regenerative
systems because costs of adsorbate recovery generally exceed
the value of the material recovered. For materials within
such a range, newly developed oxidative destruction of the
adsorbate can be applied instead of desorption. The carbon
is impregnated with a small amount of catalyst, which is
inactive during adsorption. It may be activated by heating
the air stream and bringing about catalytic oxidation of the
adsorbate.
For satisfactory adsorption, the molecular weight of a
substance should be greater than 45. Benzene easily meets
this requirement with a molecular weight of 78. Despite its
disadvantages, adsorption appears to be the most economical
method for control of organic vapors in the concentration
range of 100 to 200 ppm, virtually eliminating the emissions.
Adsorbers may be classified as regenerative or non-
reqenerative. Regenerative systems are used when the
adsorbent is to be reactivated by desorption and the de-
sorbod vapors recovered for reuse or disposal. Nonregen-
5-20
-------
erative systems are used when the adsorbent is to be re-
placed with fresh materials; the displaced material is
returned to the vendor for regeneration or is discarded.
The use of silica gel and aluminum oxide is limited to
dry streams because they preferentially adsorb water from a
gas mixture containing water vapor and organic solvents.
Thus their use in emission control is restricted because the
water content of a gas stream is often greater than the
organic vapor content- Silica gel and activated alumina
disintegrate in the presence of liquid water; hence a wet
steam cannot be used for desorption.
A typical adsorption process is depicted in Figure 5-5.
One adsorber handles the vapor-laden stream while the other
is undergoing regeneration. When the first, or onstream,
adsorbent bed approaches the breakpoint, the second adsorber,
which had undergone regeneration, is placed on stream. This
procedure ensures that the vapor is removed from the air
continuously.
Adsorbers can be used in fixed, moving, or fluidized
beds and can be set vertically or horizontally. A single
fixed-bed unit is satisfactory if the adsorbent can be
regenerated during process downtime. For example, an
adsorber for a spray-paint booth that is in use only 6 hours
a day can be designed to extract the total emission during
the spray operation for desorption lateir.
The simplest fixed-bed adsorber is a vertical cylin-
drical vessel fitted with perforated screens that support
the carbon. In another type the beds are arranged in the
shape of a cone, as shown in Figure 5-6. The cone shape
allows more surface area for gas flow and accommodates
higher air rates at lower pressure drops than does a flat
5-21
-------
EXHAUST AIR
TO ATMOSPHE RE
'j T t A M I > L U
',01  * '
• 1 '
1 /
: i
I i \ ¦ \ \
; / vCARBON ¦¦ \
	'
\
CYLINDRICAL
SHELL
HOUSING
VAPOR-
FREE
	>Aift OUT
Figure 5-6. Vertical adsorber with two cones
(permitting studies on different depths of carbon beds).
Source: Ref. 34.
5-22
-------
bed of the same diameter using the same total weight of
adsorbent. Where the cone shape will provide adequate
adsorption, its use tends to reduce costs of moving the
gases through the system.
Moving-bed adsorbers move the adsorbent into and out of
the adsorption zone. The fluidized-bed adsorber contains a
number of shallow beds of activated adsorbent. As the air
flows upward through these beds and fluidizes them, the
solvent is progressively adsorbed onto the carbon. An
adsorption isotherm for benzene adsorption on activated
coconut charcoal is given in Figure 5-7.
5.2.3 Control of Emissions from Distribution Systems
35
5.2.3.1 Bulk Terminal Storage - Emission control tech-
nology for bulk terminals is the most highly developed in
the gasoline marketing industry. For many years, certain
regions have required emission controls at bulk terminals;
additionally, the petroleum industry has viewed control of
losses at bulk terminals as an economical means of con-
serving valuable fuel products.
Uncontrolled storage tanks can account for half of the
gasoline emissions from a bulk terminal. Tankage losses
occur from breathing, evaporation, filling, emptying, and
wetted walls. The best technology for control of tankage
losses consists of (1) floating a rigid cover on the surface
of the stored gasoline to eliminate the vapor space and
(2) containing the tankage vapors within a sealed system
that incorporates a variable vapor space tank to provide
surge capacity. The efficiency of these two approaches to
controlling tankage emissions is greater than 90 percent
(over uncontrolled tankage losses). The value of recovered
gasoline has justified the cost of these tankage control
systems.
5-23
-------
LIQUID
 cc
-10 1 2 3 4 5 6
BENZENE CONCENTRATION, ppm (LOG-jQ)
Figure 5-7. Adsorption of benzene on activated coconut
charcoal at atmospheric pressure, out-gassed at 550°C.
Source: Ref. 69.
5-24
-------
5. 2. J. 2 Loading Faci .1J ties - Gasoline and petrochemicals
are distributed to the consumers by pipeline, trucks and
trailers, railroad tank cars, and marine tankers, which can
be deployed in various combinations to form a distribution
system. Significant emissions of hydrocarbons (benzene) in
these operations occur when gasoline or other hydrocarbons
are loaded into tank trucks or stationary tanks. These
operations and their emissions are described in Section
4.2.4. This section describes the control methods currently
34
used.
34
Overhead Loading - Four types of vapor collectors
have been developed for use during overhead loading opera-
tions. All are essentially plug-shaped devices that are
inserted into a fitting for the hatch opening. Gasoline
flows through a central channel in the device into the tank
compartment. This central channel is surrounded by an
annular space into which vapors enter through openings on
the bottom of the hatch fitting. The annular space is in
turn connected to a hose or pipe leading to a vapor disposal
system.
A device developed by the Mobil Oil Corporation is
connected to a vapor chamber with a transparent section to
allow the operator to see the calibrated capacity markers in
the tank compartment. Height of this closure is adjustable.
Constant downward force is required to keep it firmly in
place during filling. This device is built to fit only
hatches 20 to 25 cm (8 to 10 inches) in diameter.
The Chiksan device incorporates the hatch closure, the
vapor return line, and the fill line into an assembled unit.
This unit incorporates features that prevent overfills,
tipping off, or filling when the assembly is not properly
seated in the tank hatch.
5-25
-------
The Greenwood vapor closure, developed by the Vernon
Tool Company, also requires downward force during filling
and ordinarily does not include a transparent vapor chamber.
This closure can be fitted with an adapter for hatches
larger than 25 cm (10 inches).
The SOCO device, developed by Standard Oil Company of
California, has a positive clamp for the hatch opening,
which automatically actuates the vapor chamber when closed.
The device also includes a safety shutoff float that senses
the gas level and prevents overfilling. This device can be
used with an adapter for hatches larger than 20 cm (8
inches) in diameter.
The slide positioner of the Mobil Oil Corporation
device can allow leakage of vapors and requires close
attention by the operator during adjustments for fitting and
submerged loading. The inner valves of the SOCO device make
it considerably heavier than the other types. This device
increases pressure drop and reduces the loading rates.
Mobil and Greenwood devices both require check valves in the
vapor-gathering lines to prevent the vapor from discharging
back to the atmosphere when the assembly is withdrawn. They
also require nearly vertical entry of the loading tube into
the hatch opening to ensure a tight seal against vapor
leaks. An assembly is available that maintains the Green-
wood device in the vertical position.
34
Bottom Loading - Bottom loading facilitates collec-
tion of displaced vapors because the filling line and the
vapor collection line are independent of each other. The
vapor collection lino consists of a flexible hose or swing-
type arm connected to a quick-acting valve fitting on the
dome of the vehicle. The vapor collection line must be
5-2 6
-------
equipped with a check valve to prevent backflow of vapors to
the atmosphere when the connection to the tank is broken.
Vapor Disposal - Vapors collected during loading
operations may be used as fuel or may be recovered as a
product. When fired heaters or boilers are available, the
displaced vapors may be directed through a drip pot to a
small vapor holder, which is gas-blanketed to prevent
formation of explosive mixtures. The vapors are drawn from
this holder by a compressor and discharged to the fuel gas
34
system.
If the loading facility is near a refinery, the vapor
line can be connected from the loading facility to an
existing vapor recovery system through a regulator valve.
If not, packaged units may be used to recover vapors.
Figure 5-8 shows an absorption unit developed by the Supe-
rior Tank and Construction Company. It includes a tank
equipped with a flexible membrane diaphragm, a saturator, an
absorber, compressors, pumps, and instruments. These units
use the gasoline product as the absorbent. Accumulation of
explosive mixtures is prevented by passing the vapors dis-
placed at the loading rack through a saturator counter-
currently to gasoline pumped from storage. The saturated
vapors then flow to the vaporsphere, where a diaphragm
controls a compressor, which draws the vapors from the
2	2
sphere and injects them at about 14 kg/cm (200 lb/in. )
into the absorber. Gasoline used to absorb the hydrocarbon
vapors is returned to storage, whereas the remaining gases,
mostly air, are released to the atmosphere through a back-
pressure regulator. Some difficulty has been experienced
with air entrained in the gasoline returning to storage.
Any such air released in the storage tank is saturated with
hydrocarbon vapors as it is discharged. This air can be
5-27
-------
igure 5 8. Small capacity vaporsaver gasoline absorption
Source: Ref. 35.
-------
properly removed by flashing the liquid gasoline from the
absorber in one or more additional vessels operating at
-i ~i	34
successively lower pressures.
In 197 2, a comparison was made of alternative gasoline
vapor control systems installed on gasoline-loading ter-
164
minals.	Since none of the systems, hydrocarbon oxida-
tion, absorption, or adsorption, could be justified economi-
cally, selection of the best system, was based on recovery
efficiency, demonstrated performance, and ability to meet
air quality standards. Hydrocarbon oxidation is preferred
for terminals pumping less than 3 million barrels per year.
For larger terminals, a recovery system using an activated
carbon adsorption unit in series with a conventional absorp-
tion column may be preferred. Tests performed on these
systems show that hydrocarbon oxidation will effectively
reduce emissions by 90 percent or more.
5.2.4 Control of Emissions from Coke Ovens
Major effort is currently being expended to control
coke-oven emissions, primarily because of concern over their
polynuclear aromatic constituents. It is believed that use
of advanced technology could significantly reduce benzene
emissions from the coke ovens.
A procedure commonly used to reduce emissions during
coal charging is termed "charging on the main." A steam
ejector in the ascension pipe produces a draft in the oven
during the charging period, which causes the gases to be
pulled into the gas collector main. Although it has under-
gone some improvements, this technique is generally con-
sidered to be ineffective. The two most feasible improve-
ments are charging on the main with closed coal ports and
pipeline charging. Charging on the main with closed coal
r>~29
-------
ports can be readily incorporated into an existing plant.
Features of this system include sequencing of coal charging
from the larry car with charging port lids off only during
charging, maintenance of a steady negative pressure in the
oven, and continual drawoff of evolved gases. Pipeline
charging involves preheating coal to drive off moisture and
transporting it by pipeline. Although a 40 to 50 percent
increase in oven capacity can be realized with this method,
it is considerably more expensive, requiring far more ex-
tensive equipment changes than charging on the main with
closed coal ports. Furthermore, the capacity of the by-
product recovery system must be increased to utilize the
increase in oven capacity.
Several systems have been considered for control of
emissions resulting from the discharge of newly produced
coke. One system recently put in operation consists of an
enclosed quench car, which is attached to a second car that
carries a wet scrubber unit. During the pushing operation,
emissions are collected in the hood and withdrawn through
the coupled duct work between the cars to the wet scrubber
.. 103
uni t.
5.2.5 Control of Emissions from Sources Using Benzene
5.2.5.1 Styrene - The major source of benzene emissions in
the production of styrene is from the benzene-toluene
_ 3
recovery columns. One column emits 4.2 x 10 kg HC/kg
styrene, consisting mainly of benzene and toluene. Columns
6 8
are controlled by condensors or absorbers.
Stripper vents, another major source, have been re-
- 3
ported to emit 3.2 x 10 kg ethylbenzene/kg styrene in one
6 8
plant. Stripper vents can be controlled by conventional
methods discussed previously.
5-30
-------
Benzene emissions from distillation columns are con-
trolled well by vapor recovery systems. At one plant,
primary distillation is controlled by a condensor and emits
-3
0.4 x 10 kg benzene/kg styrene. At another plant, primary
distillation emits an uncontrolled 1.85 kg benzene/kg
68
styrene.
5.2.5.2	Cumene - The oxidizer vent is a major source of
benzene emissions in the manufacture of cumene. At one
_3
plant, about 1.5 x 10 kg benzene/kg of cumene is emitted
6 8
from a condensor operating at 73 psig. At another plant,
emissions from the condensor are reduced to only 0.004 lb
/u	68
benzene/ton cumene.
5.2.5.3	Maleic Anhydride (MAN) - The product recovery
condensor vent is the most significant source of atmospheric
88
benzene emissions in maleic anhydride manufacturing. The
uncontrolled emissions from a condensor are in the range of
8 8
0.06 to 0.20 kg benzene/kg MAN. Condensors are vented to
a scrubber to recover the MAN vapors. The uncondensed MAN
vapors can also be controlled effectively by incineration in
series with the wet scrubber. This system has been shown to
practically eliminate hydrocarbons (including benzene) in
the off gas.®®
5.2.5.4	Paradichlorobenzene - Wet scrubbers and absorption
columns are the most common benzene control devices used in
paradichlorobenzene production. A single wet scrubber
controls 90 percent of the benzene emissions and a single
97
absorption column, 95 percent.
5.3 CONTROL COSTS
This section discusses approximate capital and annual
costs for controlling gaseous emissions, including benzene.
The costs presented are limited to control devices and major
auxiliary equipment; they do not include duct work, hoods,
5-31
-------
and stacks because these costs can be determined only on a
case-by-case basis.
Cost figures cover absorption columns, carbon adsorbers,
catalytic incinerators, and thermal incinerators. Because
condensors are generally considered process equipment, they
are not discussed here.
5.3.1 Absorption Columns
Purchase costs for absorption columns were obtained
] 4 6
from a vendor. " They correspond with flow rates, as
shown:
Flow rate, scfm	Cost, $ (mid-1976)
22,600	14,200
28,400	17,500
41,000	27,300
These costs include the basic column with internal piping
and packing. Construction is of fabricated reinforced
polyethylene or carbon steel. Bed depth is 5 feet. The
costs do not include pumps, fans, or instrumentation. The
equation relating these costs to the flow rate is as follows:
Purchase cost, $ = 0.231 Q^'"^
Q = volume, scfm
This equation applies to packing bed depths of 5 feet and
flow rates between 10,000 and 80,000 acfm.
Costs of deeper beds (giving higher efficiency) in-
146
creases according to the following equation:
Purchase cost, $ = 0.231 Q"*" (1 + 0.05 [Z — 5] )
Q = volume, scfm
Z = bed depth, ft.
One other major system cost is the fan. The pressure
drop through the bed (tellerette rings) is 0.4 to 0.6 in. WG
146
per foot of packing.	Assuming J in. WG pressure drop
through the rest of the system, the following equation
relates bed depth and flow rate to fan size in horsepower:
5-32
-------
2.6 x 10 4 (1 + 0.5 z) Q = fan size, hp
Fans in this size range can be purchased for $75 to 100 per
horsepower, including base, motor, and drive.
Feed pump costs are insignificant in relation to those
of the column and fan.
Total installed costs are usually 3.4 to 4.7 times the
8 0
equipment purchase costs. This factor includes not only
direct installation costs but also engineering, tax, freight,
and contingencies.
Annual operating costs include power, maintenance,
labor, and fixed costs. Power costs are insignificant in
relation to total annual costs.
Maintenance costs are assumed at 5 percent of the total
installed costs (TIC). Labor is assumed at 0.1 man-hour/hour
operation at $10/man-hour. Capital charges are 12 percent
of TIC; taxes, insurance, etc. are about 4 percent of TIC.
Annualized costs can be roughly estimated by the
following equation:
Annualized costs, $ = 0.24 (TIC) + 1.0 H
TIC - total installed cost, $
H = hours of operation per yr
5.3.2 Carbon Adsorption
Carbon adsorption units for benzene removal are de-
8 0
signed for 0.06 kg benzene per kg carbon. For example, a
gas stream discharging 45.4 kg/hr (.100 lb/hr) benzene would
require 757.5 kg (1670 lb) carbon (density ~ 0.48 g/cc or 30
lb/ft3) regenerating every hour. Beds are designed at 60 to
100 fpm superficial bed velocity and are 15 to 25 inches
deep.
Costs of adsorption units are based on the weight of
carbon. Figures 5-9 and 5-10 show total purchase prices for
carbon adsorption systems, including fan, controls, and
5-3 3
-------
DATA VALID FOR DECEMBER 1975
Prices

it-! i j.'-I i i
iii! Mi!
1
Stem
U
! i I M ! ' : 1
80	100
C. POUNDS OF CARBON, 1000 LB
Figure 5-9. Carbon adsorption capital costs.
-------
UA1A VALiU rUK UtLCfWtK 49/9
LH
I
OJ
LH
3
S

CM
0CMC1

• i
M*nJal!
"looo
3000
4000	5000	6000
C, rouNOS Of CMS0M, IB
7000
8000
9000 10,000
Figure 5-10. Carbon adsorption capital costs - small units.
80
-------
steam regenerators. Total installed costs (not including
duct work, stack, and hooding) are 1.8 to 2.6 times the
3 0
equipment purchase price. This factor includes engi-
neering, tax, freight, and contingencies, as well as in-
stallation costs.
Operating and maintenance costs are shown as a function
of acfm and ppm benzene in Figure 5-11. The annualized
costs, including fixed costs, may be obtained by adding
about 15 percent of the total installed cost.
5.3.3	Catalytic Incinerators
Costs of catalytic incinerators are based on volumetric
throughput. Heat exchangers may be provided to save on
energy costs, but they increase capital costs. Figure 5-12
shows purchase prices for catalytic incinerators with and
without heat exchangers.
Total installed costs are usually 1.8 to 2.6 times the
8 0
purchase price. They include not only direct installation
costs but also tax, freight, engineering, and contingency
costs; they do not include duct work, fan, stack, or hooding
costs.
Annual operating and maintenance costs are given in
Figure 5-13 as a function of flow rate (acfm) and benzene
concentration. The annualized cost, including fixed costs,
may be obtained by adding about 15 percent of the total
installed costs.
5.3.4	Thermal Incinerators
Thermal incinerators are designed to allow 0.3 to 0.6
second retention time for combustion of low concentrations
of benzene. Figures 5-14 and 5-15 show purchase prices for
thermal incinerators, with and without heat exchangers, as a
function of acfm and retention time. Total installed costs
(not including fan, duct work, stack, hooding, etc.) are
3
about 1.8 to 2.6 times the purchase price. These costs
5-36
-------
DATA VALID FOR DECEMBER 1975
erf*
lOO-PPM
I ,
Curve
off to
Costs
ut-iili
s; based orr concen
tratlor
ne 1n| PPM.
cludelalll labQ
r ana
lectHon
costsl-from col
stack- exhaus
poi nt^'tc
Cpsts' Includej adsorber
cement:
10"	10J
ADSORBER INLET VOLUME, ACFM
Figure 5-11. Carbon adsorption unit operating and main-
tenance cost vs. acfm and hydrocarbon concentration.^
5-37
-------
DATA VALID FOB DECEMBER 1975
CO
II
4*CWU*
Tcfi! Uf ITS
TOW
£
t:
S 50

40	50	60
INCINERATOR CAPACITY, 1000 AC FN
Figure 5-12. Catalytic incinerator prices.
80
-------
100 DATA VALID FOR DECEMBER 1975
100-ppm
WO/ HE
nnS NOTE: 1. Curves based on ppm concentration
of hydrocarbons such as toluene,
ketones, and napthas.
2.	Costs include all labor and
utility costs from collection
point to stack exhaust.
3.	Cost include catalyst replacement.
4.	Costs with and without heat
exchangers included.
WO/HE
—-A 100" ppm
l/>
O
o
Ui
3x10
INCINERATOR INLET VOLUME, ACFM
Ficjure 5-13. Catalytic incinerator operating and maintenance
cost vs. acfm and hydrocarbon concentration.80
r>- 39
-------
280
260
240
220
200
180
V*
1160
•««
g 140
gl20
i
*100
u
Z
80
60
40
20
0
OATA VALID FOR DECEMBER 197S
ed bp
of!
i M
andj
w
II ©CO
s eyrvq is
lneintttiBt
or| bl( Wer;i
hdnqti -
! 11
lie tent
Hattr
peci
Heat! it
luttm

INCINERATOR CAPACITY. 1000 ACFN
-14.
8 0
Prices for thermal incinerators with heat exchangers.
-------
70
60
i
CEHBEK 1S75
FfTffF
• H-
50 }..».,
I ¦
U-U
a
tliii
«1f
ri:
l:
j-V-

1 'p'iM'ilijU

«"¦ '¦
'j-i •~•I s ¦r-r-i-t
M
41 'iliir1
r I"!"t";! it
Jrifl-liil.
¦I! -
I!
ij-r mi ;m pij -4
11 ' j i ri 11 i i ¦ -* • • —
U'l il ji i'! i Ml r rif ij^-ir Vt:
Hn f \/<< I i i ; ! I HJH «,5-ff ft i M i ! k-: ?t - r 11 • t H r !-r 1 i ! 1 i	i ! 11 U"r:RTivi-? ,!»M J ilLi": I I'liT !
•HiUis'fH	• rf 1-f-ri-i H • M H i t tlU-rr r1 i M ¦ itif - n'' ¦ ' <1'': • : ' i ! 1¦ ; T~: I nnft]
30 I"Ht fcj4l t-Trl
f-i 1 1 / i-l i t :/i •
10
Y !-U4 Hit H rU-~-H 111 ¦ f f ftr i ; <1 • M- 1-t i ; i . 	
Tfit; i-j-j f IjSo-v* * :A t i 1-4 t 4-i-p-i-f 111	i~H ¦ fi •; ? i! It';! -i •! ~ i;
tr-i i	i]1 i ji h U ji iiiilii -1! iirN ri]i i-i! '_i ni j IH-; i^fl*
t'tTrrLi":. Hvj i i I" "i fr-rTr"i i7 ayti :WR:f;-frtrr wr-	/a.T LJ¦ JI
	H"1 F[! i] Kfctiiwijridfl i'tNH IoirM ;is; }±*sojEf |MM "'l"f
'i1 b&inF itW-ihiv*;1 • "ir'fhr?
i^pRI^ lUtfWPES'
| fc.OSER;,, c$i}t«;01
J PRICES N!Lm V'AR'i
¦ ¦ ''11 kttc
, jjcihj!
015 I AND; :
Tt
20
I.
40	50	60
INCINERATOR CAPACITY, 1000 ACFH
I
-iTCRilPANi Ql
KSiRufePfU
FUNCmi pf
ON
.] Materia)} of* coi
i«1' Cantj-O
Head con(f at:
! H-tI

s.tr(jcqi.«
r
poll
hr r
1 a.nt

70
80
90
Figure 5-15. Prices for thermal incinerators without heat exchangers.
-------
include tax, freight, engineering, and contingencies, as
well as direct installation costs.
Annual costs for operating and maintenance of incinera-
tors with and without heat exchangers are shown in Figure
5-16 as a function of benzene concentration and flow rate.
Annualized costs may be estimated by adding about 15 percent
of the total installed costs to include fixed costs.
5-42
-------
10*
DATA VALID FOR DECEMBER 1975
S
UJ
a_
o
00L3PM
i W/HEr
ed_on
roaar
ketone
lab
., pm • cjol ljec
tjacklejihaLStii/!
and- vnthautrlhe
ricTude
-------
6.0 EMISSION TRENDS
However they are calculated, emission predictions are
always inexact. This is particularly true of benzene emis-
sion projections because of future uncertainties regarding
the entire chemical market (especially benzene production
industries) and gasoline consumption. Demand for benzene as
a petrochemical has a direct bearing on its availability for
blending into the gasoline pool. All these factors affect
the type and quantity of projected benzene emissions. Total
national benzene emissions are predicted to decrease by
approximately 57 percent between 1976 and 1985, mostly as a
result of reduced automobile emissions.
6.1 MOBILE SOURCE TRENDS
Available data indicate that gasoline-powered motor
vehicles will remain the major source of benzene emissions
nationally. They will be a less dominant source than at
present, however, because 1985 auto emissions are projected
to fall to about 24 percent of the 1976 level. This sharp
decrease is predicated primarily on the assumption that
nearly all cars will have catalyst control systems to meet
the anticipated 0.41 g/mile HC regulation. It is assumed
that such a standard will not be retroactive for older
vehicles; therefore, AP-42 data, which incorporate the
effect of catalyst deterioration, are utilized in the trend
31
projections. Evaporative losses presented by AP-42 for
the post-1979 period are greatly reduced to reflect the
assumption that better control will be available to meet the
6-1
-------
more stringent overall standards.^ Table 6-1 summarizes
the estimates of 1985 benzene emissions from gasoline- and
diesel-fueled motor vehicles.
Table 6-1. SUMMARY OF PROJECTED BENZENE EMISSIONS
FROM MOBILE SOURCES IN 19 8 5
Source
Benzene
emissions, Gg
Gasoline-powered motor vehicles

Evaporative losses
20. 8
Exhaust emission (catalyst)
controlled
27.8
Total
48.6
Diesel-powered heavy-duty vehicles

Evaporative losses
a
Exhaust emission
1.6
a Not available.
6.1.1 Projected Emissions from Gasoline-Powered Motor
Vehicles
Tighter emission controls are expected to reduce
benzene emissions from both exhaust and evaporative sources.
It is assumed that the benzene level of gasoline will remain
at 2 percent by volume, the value utilized in the 1976
estimates, and that the benzene-to-HC ratio in exhaust will
52
be 0.025, the level determined in recent exhaust tests.
Projections of benzene emissions from gasoline-powered
motor vehicles in 1985 are detailed in Table 6-2. All
emission data are based on low-altitude, non-California
cases. This projection assumes that both exhaust and
6-2
-------
Table 6-2. PROJECTED BENZENE EMISSIONS FROM GASOLINE-
POWERED MOTOR VEHICLES IN 1985
Vehicle
model
year
Age,
years
a
Fraction
of miles
traveled*
b
Exhaust HC
emissions,
g/mi+
c
Evap. HC
losses,
g/mi §
1985
1
0.116
0.27
0.5
1984
2
0.135
0. 32
0.5
1983
3
0.125
0. 38
0.5
1982
4
0.122
0.43
0.5
1931
5
0.106
0.49
0.5
1980
6
0. 086
0. 54
0.5
1979
7
0. 083
0. 59
1.76
1978
8
0.072
0.65
1.76
1977
9
0.051
2.6
1. 76
1976
10
0.037
2.8
1.76
1975
11
0. 023
3.0
1.76
1974
12 plus
0. 045
6.2
1.76

1. 000

/• 11	it it
Exhaust emission estimate = (123 x 10 ) (9494) (0.025) Z a x b
= (123 x 106) (0494) (0.025) (0.95)
= 27.8 Gg/yr (61.3 million lb/yr)
Evaporative losses	= (123 x 10^)(9494)(0.020) Z a x c
= (123 x 106) (9494) (0.020) (0. 89)
= 20.8 Gg/yr (45.9 million lb/yr)
Total 1985 benzene losses = 48.6 Gg/yr (107.2 million lb/yr)
*	31
AP-42, Supplement 5, Table 3.1.2-5.
+ AP-42, Supplement 5,"*1 Table D.l-17, reference year 1985.
5	31
Evaporative loss estimates from AP-42, Supplement 5,
Table D.l-27.
ft
National vehicle population projection data from DOT and the
Motor Vehicle Manufacturers Association.
* *	95
Average annual miles traveled per vehicle in 1974.
6-3
-------
evaporative emission control technology available by about
1980 will control benzene emissions more effectively than
the best available control technology in 1976. A more
conservative projection of 1985 emissions can be developed
by assuming that the best control technology as of 1976 will
be used in 1985. Thus, the 1985 projection for exhaust
emissions for the years 1978 to 1985 vehicles is altered to
include the Olson data of 0.016 g/mile for new catalytically
52
controlled vehicles and a 1976-level evaporative loss rate
of 1.24 g/mile for 1980 to 1985 vehicles.^ A catalyst
deterioration rate of 10 percent per year is also incorpo-
rated. The resulting exhaust emissions are 38.7 Gg/yr (85.3
million lb/yr) and evaporative losses are 32.7 Gg/yr (72.1
million lb/yr), giving a total emission rate of 71.4 Gg/yr
(157.4 million lb/yr).
Over 95 percent of all 1975 and later model automobiles
now have catalyst exhaust controls; a limited few have
'lean-burn' engines, and even fewer have thermal-reactor
(stratified-charge) systems for hydrocarbon control.To
meet the more stingent hydrocarbon emission standards anti-
cipated by 1985, all gasoline vehicles must employ some type
of catalytic system. Since benzene emission is a function
of its constituent level in the gasoline, any increase or
decrease in blending affects exhaust-pipe benzene levels.
Concentrations in gasoline are dependent on all of the
following factors, which vary among refineries (i.e., the
major benzene source) and with seasonal blends:
° Benzene content of the crude oil
° Operating characteristics of catalytic reformers
° The extent to which catalytic reformate is used as
a blending component
6-4
-------
° Use of benzene for petrochemicals and the re-
sulting volume available for the gasoline pool.
Published projections indicate that even though the volume
of crude oil processed in 1985 probably will be about 9
percent greater than in 1976, total gasoline production will
154
decrease some 5 percent (to 100 billion gallons).
Conversely, the quantity of refinery benzene withdrawn for
use as a chemical intermediate is expected to increase
substantially. Energy and Environmental Analysis, Inc.,
(EEA) applies a value of roughly 1 percent benzene in gaso-
line blends (markedly lower than Short's 2 percent) and
projects that 1985 gasoline blends will not exceed 2 per-
154
cent.	EEA stipulates that this estimate involves a
degree of uncertainty and that although the aromatic content
of gasoline is likely to increase, the toluene and xylene
levels will probably increase more than the benzene level.
Both of these available constituents are higher volume
gasoline constituents and have higher octane numbers (al-
though benzene is a desirable blending component because of
its volatility and other performance characteristics).
Also, neither toluene nor xylene is as much in demand as
benzene as a feedstock for petrochemical production.
6.1.2 Projected Emissions From Other Mobile Sources
Benzene emissions from diesel-powered motor vehicles
are predicted to decrease 8 percent between 1976 and 1985,
primarily because of an estimated decline in diesel fuel
usage consistent with that of gasoline. This projection
assumes no variance from the emission calculation given in
Section 4.1.3 other than a decrease in diesel consumption
from 9.6 billion to 9.2 billion gallons over the 9-year
period. The fuel decline is based on the projected decline
in 1985 gasoline production and on the assumption that the
6-5
-------
154
diesel fraction of the market remains constant.	This
decline may not be wholly realistic in view of the control
methods available to meet the more stringent hydrocarbon
standards of the future, once promulgated. If the use of
light-duty diesel vehicles offers a convenient means of
meeting the lower emission limitations, their use may
increase.
Because of the scarcity of information regarding ben-
zene emissions from aircraft and other mobile sources, it is
impossible to make a realistic projection of 1985 emissions.
6.2 STATIONARY SOURCE TRENDS
As discussed in Section 4.2, the primary stationary
sources of atmospheric benzene emissions include the pro-
duction of the compound, its use as a chemical intermediate,
and the storage and distribution of benzene and gasoline.
The portions of stationary-source benzene emissions contri-
buted by these three major sources in 1976 are approximately
14, 64, and 22 percent, respectively. Projected 1985 dis-
tribution figures (14, 68, and 18 percent) do not differ
significantly despite the anticipated 30 percent increase in
stationary sources of benzene emissions between 1976 and
1985.
6.2.1 Benzene Production
Although less atmospheric benzene is generated and lost
during benzene production than from other stationary sources,
the volume is still significant. Manufacture of benzene as
a by-product of coke-oven operations represents a relatively
minor portion of total benzene production, but it is a
significant source of localized atmospheric benzene because
of the poor vapor containment within the typical oven
battery. Although emissions from coking operations are
6-6
-------
projected to increase approximately 20 percent by 1985,
improved coke-oven battery emission controls could reduce
these emissions to well below 1976 levels. Benzene emis-
sions from refineries are projected to increase approxi-
mately 42 percent by 1985, primarily because of the expected
increase in benzene production volume.
6.2.2 Benzene Consumption Processes
As the result of a predicted growth rate of 6 percent
per year between 1976 and 1985, demand for benzene as a
chemical intermediate is expected to be about 2.4 billion
gallons per year by 1985.^ The 1976 benzene demand for
154
consumption processes was only about 1.5 billion gallons.
This increase in volume of consumption is expected to cause
a 39 percent increase in benzene emissions by 1985. Pro-
jected benzene emissions are based on 1976 emission factors
and estimated 1985 production rates.
Nitrobenzene (Aniline)
Nitrobenzene is prepared by direct nitration of benzene
using mixed nitric and sulfuric acids. Generally, about 85
percent of the chemical is used in the manufacture of
aniline. The demand for aniline corresponds directly to the
demand for a wide variety of plastics, synthetic fibers,
rubber chemicals, drugs, and dyes in which it is used.
Domestic nitrobenzene production capacity in 1975 was
55
approximately 1.07 billion pounds. Benzene emissions from
nitrobenzene manufacture are projected for 1985. Figure
6-1 correlates the future demand upon benzene supply with
increased aniline production. Because of the lack of
documented data on atmospheric benzene emission from aniline
manufacture, future levels are not projected.
6-7
-------
900
800
700
600
500
o
2 A 00
700
100
©•HISTORICAL DATA
DCHEMICAL AND ENGINEERING NEWS 5/2/77
¦CHEMICAL AND ENGINEERING NEWS 4/4/77
1985
1975
1980
1965
1970
YEAR
Figure 6-1. Historical and projected domestic aniline
production and corresponding benzene consumption.
a Linear regression for the consumption of benzene was
developed from 10 year historical data (1965-1975).
^ Linear regression for chemical production develooed
from 10 year historical data (1965-1975).
6-8
-------
Ethylbenzene (S c.yrem;)
Production of ethylbenzene, primarily from the alkyla-
tion of benzene with ethylene, has been the principal con-
sumer of domestic benzene. As early as 1965, ethylbenzene
manufacture consumed 41 percent of the entire domestic
benzene production. Over 90 percent of the ethylbenzene
produced in the United States today is used for the manu-
facture of styrene and its derivatives. Present domestic
styrene capacity greatly outweighs demand, and no change in
this situation is predicted through 1980. Ethylbenzene/
styrene demand should continue to increase at the present
rate of 6 to 7 percent annually until 1980, after which a
decrease to 5 percent annually is predicted.	It is
anticipated that ethylbenzene and its styrene derivatives
will use approximately 50 percent of the domestic benzene
produced in 1985. Figure 6-2 presents projected demands on
benzene supply based on the increase in ethylbenzene/styrene
production. The 198 5 benzene emissions from processing both
of these compounds are expected to increase some 90 percent
over 1976 estimates.
Maleic Anhydride
Of the present domestic maleic anhydride supply, 81.3
percent is produced from benzene oxidation, 16.7 percent is
derived from butanes, and the remaining 2 percent is a
phthalic: anhydride by-product. Figure 6-3 relates future
demands on benzene supply to the projected increase in
maleic anhydride production. Nearly half of all the atmos-
pheric benzene emitted from the petrochemical category
results from the manufacture of maleic anhydride. The
projected 1985 increase in benzene emissions from this
source is approximately 14 percent.
6-9
-------
9000
8000
7000
fe 6000
LO
E
—i
5c
5000
4000
3000
2000
~l	1	1	1	1	«~
vs-#"" t -

J	L
J	I	I	I	I	1	I	L
O»HIST0RICAL DATA
~CHEMICAL AND ENGINEERING NEWS 5/2/77
¦CHEMICAL AND ENGINEERING NEWS 4/4/77
__l	I	I	I	I	I	1	I	
1965
1970
1975
YEAR
1980
1985
Figure 6-2. Historical and projected domestic ethlylbenzene
production and corresponding benzene consumption.
Linear regression for the consumption of benzene and the
production of ethylbenzene were developed from 10 year
historical data (1965-1975).
6-10
-------
500
450
400-
350 -
300
V •
250
o
200-
x
O
150
100
O#HISTORICAL DATA
~ CHEMICAL AND ENGINEERING NEWS 5/2/77
¦ CHEMICAL AMD ENGINEERING NEWS 4/4/77
1965
YEAR
Figure 6--3. Historical and projected domestic maleic anhydride
production and corresponding benzene consumption.
a Linear regression for the consumption of benzene was
developed from 10 year historical data (1965-1975).
k Linear regression for chemical production developed from
10 year historical data (1965-1975).
6-11
-------
Cyclohexane
In the United States approximately 25 percent of the
cyclohexane manufactured commercially is produced by frac-
tionation of petroleum and 7 5 percent by catalytic hydro-
generation of benzene. Figure 6-4 relates future demands on
benzene supply to the projected increase in cyclohexane
manufacture. Although the literature search produced no
documented data on atmospheric benzene emissions from cyclo-
hexane production, it is anticipated this will remain a
relatively minor source.
Phenol (Cumene)
Among the several methods available for manufacture of
phenol, cumene peroxidation accounts for over 90 percent of
the total U.S. phenol capacity. Phenol production is the
second largest consumer of benzene. Approximately 20 percent
of total domestic benzene was used for cumene/phenol produc-
tion in 1965; the present figure is the same and is pre-
dicted to apply through 198 5. Since future U.S. benzene
supply is not expected to meet phenol demand, more phenol
will be imported from foreign producers. Figure 6-5 depicts
future demands on benzene supply in correlation with pro-
jected increases in cumene/phenol production. Based on
these projections, benzene emissions from phenol production
in 1985 will be about 50 percent higher than the 1976 volume
unless abatement procedures are improved. Similarly, pro-
jected 1985 benzene emissions from the manufacture of cumene
are proportionately higher.
Chlorobenzenes
The annual benzene emission level attributed to the
manufacture of chlorobenzenes, as determined by Monsanto, is
approximately 2.29 Gg.^ Emissions are projected to increase
nearly 40 percent by 1985.
6-12
-------
1000 ¦
J	L
I
O • HISTORICAL DATA
8 CHEMICAL AND ENGINEERING NEWS 5/2/77
D CHEMICAL AND ENGINEERING NEWS 4/4/77
J	1	I	I	I	1
_L
1965
1970
1975
YEAR
1980
1985
Figure 6--4. Historical and projected domestic cyclohexane
production and corresponding benzene consumption.
Linear regression for the consumption of benzene was
developed from 10 year historical data (1965-1975).
Linear regression for chemical production developed from
10 year historical data (1965-1975) ,
6-13
-------
T ¦' 1" 1 1 1 T ! 1 1 ! 1 1
i i i ¦ t r ¦ i -T ¦ i - r ¦

		 	
	






.•»* o
9 c\

• w
6 °
		Ill
O* HISTORICAL data
~ CHEMICAL AND ENGINEERING NEWS 5/2/77
¦ CHEMICAL AND ENGINEERING NEWS 4/4/77
i i 1 I I I i I I
1965
1970
1975
YEAR
I960
1985
Figure 6-5. Historical and projected domestic cumene (phenol)
production and corresponding benzene consumption.
a Linear regression for the consumption of benzene was
developed from 10 year historical data (1965-1975).
b Linear regression for chemical production developed from
10 year historical data (1965-1975).
6-14
-------
Detergent alkylate
As stated in Section 4.2.3, detergent alkylate, is the
name commonly given to a product formed from the catalytic
alkylation of benzene with dodecene. The product made may
be either branched or linear alkylbenzene (dodecylbenzene).
Benzene emission data are obtainable only for the latter
production, and these were used to project a 1985 emission
increase of approximatly 36 percent, or 0.19 Gg total. This
estimate could be low.
Solvent operations
Benzene has a wide range of applications as an indus-
trial solvent but is being replaced by less toxic products.
Although no benzene emission data are available on which to
base projected future atmospheric levels, solvents could
remain an important emission source. The benzene contami-
nant level in other solvents may be a major determinant in
estimating the total solvent emissions.
Other nonfuel uses
Additional nonfuel uses of benzene that are regarded as
potential emission sources include the production of fumaric
acid, cunene sulfonate ammonium salt, cumene sulfonic acid,
benzonitrile, and anthraquinone. The increase in benzene
emissions from the manufacture of these chemicals is pro-
jected to be approximately 40 percent by 1985. This figure
depends on the volume of benzene production and consumption.
6.2.3 Storage and Transport Emissions
Although storage and distribution of gasoline and
benzene I including "feedstock" benzene stored by noncaptive
consumer industries as described in Section 4.2.3) are
significant sources of benzene emissions, they are expected
to increase only slightly in overall volume (approximately
6-15
-------
2.5 percent) by 1985. It is assumed that little or no
benzene is emitted in the storage and transport of consumer
chemicals (ethylbenzene, cumene, cyclohexane, etc.) for
which benzene is a raw material; therefore, this source is
not considered in the projection. Any projected increase in
benzene emissions from storage and transport systems is
offset considerably by the decline in the total gasoline
supply predicted for 1985. More effective vapor control
methodology is also anticipated. Combined benzene emissions
from gasoline storage and distribution and those from ser-
vice station operations are predicted to be lower by some 5
percent than 1976 levels. General population exposure from
service station operations will remain significant. The
increasing number of self-service operations make this
potentially the largest exposure source by 1985. Emissions
from the storage and transport of product benzene are pre-
dicted to increase significantly (approximately 65 percent)
because of the projected increase in production volume by
1985. As stated above, this increase is offset by the
estimated decrease in benzene emissions from gasoline
storage/distribution.
6-16
-------
REFERENCES
1.	AICE Survey of USSR Air Pollution Literature, Vol. XV.
A Third Compilation of Technical Reports on the Biologi-
cal Effects and the Public Aspects of Atmospheric Pol-
lutants. American Institute of Crop Ecology. Research
Triangle Park. U.S. Environmental Protection Agency.
1972.
2.	Air Pollution from Nitration Processes. Processes Re-
search, Inc., Industrial Planning and Research, Cincin-
nati, Ohio. Contract No. CPA 701. EPA Office of Air
Programs. March 1972. pp. 14-22.
3.	Air Quality Criteria for Hydrocarbons. Washington,
D.C. AP-64. U.S. Department of Health, Education, and
Welfare. March 1970.
4.	Aldehyde and Reactive Organic Emissions from Motor
Vehicles, Parts I and II. U.S. Bureau of Mines. Ann
Arbor, Michigan. Report No. APTD-1568 a/b. U.S.
Environmental Protection Agency. March 1973.
5.	Anderson, E.V. Output of the Top 50 Chemicals Drops
Sharply. In: Chemical and Engineering News. May
1976. pp. 34-37.
6.	Anderson, E.V. Top 50 Chemical Regain Output Lost in
1975. Chemical and Engineering News, May 2, 1977.
pp. 32-39.
7.	Annual Refining Survey. A. Cantrell (ed.). In: The
Oil and Gas Journal. March 1976. pp. 124-156.
8.	Annual Report for Calendar Year 1972. Environmental
Toxicology Research Laboratory, National Environmental
Research Center. Cincinnati. U.S. Environmental Pro-
tection Agency. January 1973.
9.	API Questionnaries on OSHA's Proposed Benzene Standard.
10. Aronatics in 1980 and Beyond. Hydrocarbon Processing.
September 1971. pp 147-150.
6-17
-------
11.	Aromatics Outlook: Abundant Supply, Chemical & Engi-
neering News. April 4, 1977. 55(14):1011.
12.	Atmosphere Emission from Petroleum Refineries - A
Guide for Measurement and Control. Washington, D.C.
PB-198-096. U.S. Department of Health, Education, and
Welfare. 1960. 56 pp.
13.	Bachman, J. Documentation of Benzene Automative Exhaust
Emission Estimates. Internal Memorandum, U.S. Environ-
mental Protection Agency, Pollutant Strategies Branch.
June 2, 1977.
14.	Barnes, T.M., A.O. Hoffman, and W. Bownie. Evaluation
of Process Alternatives to Improve Control of Air
Pollution from Production of Coke. Battelle Memorial
Institute. Columbus, Ohio. January 31, 1970.
15.	Bavley, H., and L.D. Paguotto. How to Handle Ventila-
tion for Benzene Control. In: Air Engineering,
Detroit, Business News Publishing Company. December
1962. pp. 18-22.
16.	Benzene in Motor Gasoline. International Study Group
for Conservation of Clean Air and Water, Western Europe
(Stichting Concawe). The Hague. 1973.
17.	Black, F.M., and R.L. Bradow. Patterns of Hydrocarbon
Emissions from 197 5 Production Cars. U.S. Environ-
mental Protection Agency. Research Triangle Park, N.C.
1976.
18.	Both Oil Imports & Refining Capacity up Sharply in U.S.
in '76, Say API. Hydrocarbon Processing. March 1977.
p. 13.
19.	Bowden, J.H. Status of Unleaded and Low-Lead Gasoline
Composition. Contract No. DAAD05-72-C-0053. U.S.
Army Coating and Chemical Laboratory. Aberdeen Proving
Ground, Maryland. Interim Report FLRL 16:48. 1972.
20.	Brownstein, A.M. U.S. Petrochemicals. The Petroleum
Publishing Company. Tulsa, Oklahoma. 1972. pp.
154-191.
21.	Burklin, C.E., E.C. Cavanaugh, J.C. Dickerman, and S.R.
Fernandes. A Study of Vapor Control Methods for Gaso-
line Marketing Operations: Vol. I - Industry Survey
and Control Techniques, Vol. II - Appendix. Radian
Corporation. Austin, Texas. EPA-450/3-75-046-a/b.
U.S. Environmental Protection Agency. April 1975.
6-18
-------
22.	Burr, R.K. Benzene Extraction from Motor Gasolines.
Internal Memorandum to J.F. Durham. U.S. Environmental
Protection Agency. June 21, 1977.
23.	Calvert, J.G., and J.N. Pitts, Jr. Photochemistry.
John Wiley and Sons, Inc. 1966.
24.	Calvert, J.G., and McQuigg. The Computer Simulation
of the Rates and Mechanisms of Photochemical Smog
Formation. In: Symposium on Chemical Kinetics of
Tropospheric Reactions. Warrentown, Virginia. April
1974. pp. 113-153.
25.	Carbone, W.E. Coke-Oven Emission Control. In: Iron
and Steel Engineer. December 1971. pp. 56-60.
26.	40 CFR 86 Emission Standards for 1977 Light-Duty
Vehicles, Trucks, and Heavy-Duty Diesel Engines.
27.	Chemical and Engineering News. American Chemical
Society. Washington, D.C. November 1976.
28.	Chemical Information Services. Stanford Research In-
stitute, 1976 Directory of Chemical Producers. U.S.A.
1976.
29.	Chemical Marketing Reporter (Formerly Oil, Paint & Drug
Reporter). December 20, 1976.
30.	Clark, P.J. Investigation and Assessment of LightDuty-
Vehicle Evaporative Emission Sources and Control. Exxon
Research and Engineering Company. Linden, N.J. Pre-
pared for U.S. Environmental Protection Agency. Ann
Arbor, Michigan. Contract No. 68-03-2172. June 1976.
40 pp.
31.	Compilation of Air Pollutant Emission Factors, 2nd Edi-
tion. U.S. Environmental Protection Agency. Research
Triangle Park, North Carolina. Publication No. AP-42.
April 1976.
32.	Considine, D.M. Chemical and Process Technology Ency-
clopedia. McGraw-Hill Book Company. New York. 1974.
pp. 161-163.
33.	Control Techniques for Carbon Monoxide, Nitrogen Oxide,
and Hydrocarbon Emissions from Mobile Sources. AP-66.
U.S. Department of Health, Education, and Welfare.
National Air Pollution Control Association. Washington,
D.C. March 1970.
6-19
-------
34.	Control Techniques for Hydrocarbon and Organic Solvent
Emissions for Stationary Sources. AP-68. U.S. De-
partment of Health Education and Welfare. Public
Health Service. Washington, D.C. March 1970.
35.	Control Techniques for Hydrocarbon and Organic Solvent
Emissions from Stationary Sources. North Atlantic
Treaty Organization. Research Triangle Park, N.C.
U.S. Environmental Protection Agency. October 1973.
36.	Control Techniques for Lead Air Emissions. PEDCo
Environmental, Inc. Cincinnati, Ohio. Contract No.
68-02-1375. U.S. Environmental Protection Agency.
October 1976.
37.	Correspondence from F.D. Bess, Union Carbide Corp.,
Chemicals & Plastics Div., to R.J. Johnson, U.S. Environ-
mental Protection Agency. March 3, 1977.
38.	Correspondence from I.H. Gilman, Chevron U.S.A., Inc.,
to R. Johnson, U.S. Environmental Protection Agency.
March 18, 1977.
39.	Correspondence from R. Hartle, National Institute for
Occupation Safety and Health. December 1976.
40.	Correspondence from K.J. Kaminski, B.F. Goodrich
Chemical Div., to Richard Johnson, U.S. Environmental
Protection Agency. Marcy 4, 1977.
41.	Correspondence from I.M. Lackazczyk, Ryckman/Edgerley/
Tomlinson & Assoc., to R. Young, National Institute for
Occupational Safety and Health. February 197 6.
42.	Correspondence from L. Platte, Office of Air and Water
Programs. United States Environmental Protection
Agency. Ann Arbor, Michigan. December 1976.
43.	Correspondence from A.G. Smith, Shell Oil Co., to
Richard Johnson, U.S. Environmental Protection Agency.
March 4, 1977.
44.	Correspondence from A.G. Smith, Shell Oil Company,
Houston, Texas, to D.R. Patrick, Emission Standards &
Engineering Div., U.S. Environmental Protection Agency.
March 16, 1977.
45.	Correspondence from J. Tarr, Texas State Air Pollution
Control.
6-20
-------
46.	Correspondence from Dr. N.K. Weaver, American Petroleum
Institute, to R. Young, National Institute for Occu-
pational Safety and Health. March 1976.
47.	Correspondence from R.C. Widgery, Gulf Oil Chemicals
Co., to Richard Johnson, U.S. Environmental Protection
Agency. February 25, 1977.
48.	Correspondence from R. Young, National Institute for
Occupational Safety and Health.. Cincinnati, Ohio.
December 1976.
49.	Criteria for a Recommended Standard--Occupational
Exposure to Benzene. U.S. Department of Health, Educa-
tion, and Welfare. NIOSH 74-137, 1974.
50.. Danielson, J.A. Air Pollution Engineering Manual. 2nd
Edition. Air Pollution Control District County of Los
Angeles. Research Triangle Park, N.C. AP-40. U.S.
Environmental Protection Agency. May 1973.
51.	Debreczeni, E.J. Future Supply and Demand for Basic
Petrochemicals. Chemical Engineering, 84(12), June 6,
1977. pp. 135-141.
52.	Dekany, J.P. Evaluation of OAQP1s Draft Report on
Benzene. Internal Memorandum, U.S. Environmental Pro-
tection Agency, Emission Control Technology Division.
January 21, 1977.
53.	Dickerman, J.C., T.D. Raye, and J.D. Colley. The
Petroleum Industry. Radian Corporation. Austin,
Texas. U.S. Environmental Protection Agency. May
1975.
54.	Dimitriades, B., and D.E. Seizinger. A Procedure for
Routine Use in Chromatographic Analysis of Automotive
Hydrocarbon Emissions. In: Environmental Science and
Technology, Vol. 5. Easton. American Chemical Society.
March 1971. pp. 223-229.
55.	1977 Directory of Chemical Producers - U.S.A. Menlo
Park. Chemical Information Services. Stanford Research
Institute, 1977. 1060 pp.
56.	Dwyer, F.G., P.J. Lewis, and F.J. Schneider. Efficient,
Nonpolluting Ethylbenzene Process. In: Chemical
Engineering, C.S. Cronan (ed„). McGraw-Hill Book
Company. New York. January 1976. pp. 90-91.
6-21
-------
57.	Encyclopedia of Occupational Health and Safety, Vol.
1/11. McGraw-Hill Book Company. New York. 1974.
58.	Ess, T.J. The Modern Coke Plant. Iron and Steel
Engineer. C3-C36. January 1948.
59.	Faith, W.L., D.B. Keyes, and R.L. Clark. Industrial
Chemicals. 3rd Edition. John Wiley and Sons, Inc.
New York. 1966.
60.	Federal Register, Vol. 41, No. 164. August 23, 1976.
61.	Finkelstein, H. Air Pollution Aspects of Pesticides.
Litton Systems, Inc. Prepared for U.S. Department of
Health, Education, and Welfare. September 1969.
62.	Glasson, W.A., and C.S. Tuesday. Hydrocarbon Reactivi-
ties in the Atmospheric Photoxidation of Nitric Oxide.
Environmental Science and Technology. Vol. 4, No. 11.
November 1970. pp. 916-924.
63.	Guide for Compiling a Comprehensive Emission Inventory.
2nd Edition. Office of Air Quality Planning and
Standards. Research Triangle Park, North Carolina.
December 1974.
64.	Hartle, R., and R. Young. Occupational Exposure to
Benzene at Service Stations. National Institute for
Occupational Safety and Health. Cincinnati, Ohio.
June 197 6. 7 pp.
65.	Hay, E.B. Exposure to Aromatic Hydrocarbons in a Coke-
Oven By-Product Plant. In: Industrial Hygiene Journal.
July-August 1964. pp. 386-391.
66.	Hedley, W.H. Potential Pollutants from Petrochemical
Processes. Monsanto Research Corporation. Westport,
Conn. Contract No. 68-02-0226. U.S. Environmental
Protection Agency. 1975.
67.	Highway Statistics. U.S. Department of Transportation,
Federal Highway Administration. Washington, D.C.
Stock No. 050-001-00107-1. 1974.
68.	Houdry Div./Air Products and Chemicals. Survey Reports
on Atmospheric Emissions from the Petrochemical Industry,
Volume III, EPA-450/3-73/005-d.
6-22
-------
69.	Hougen, O.A., K.M. Watson, and R.A. Ragatz. Chemical
Process Principles, Part I. Bombay, P.S. Jayasinghe,
Asia Publishing House. 1962.
70.	Hum, R.W., J.R. Allsup, and F. Cox. Effect of Gasoline
Additives on Gaseous Emissions. U.S. Bureau of Mines.
Washington, D.C. EPA-650/2-75-014. U.S. Environmental
Protection Agency. December 1974. 64 pp.
71.	Iammartino, N.R. Petrochemicals Sing to Ethylene's
Tune. In: Chemical Engineering, C.S. Cronan (ed.).
McGraw-Hill Book Company. New York. April 1975. pp.
52-54.
72.	Identification and Analysis of Ambient Air Pollutants
Using the Combined Techniques of Gas Chromatography and
Mass Spectrometry. Research Triangle Institute.
Research Triangle Park, N.C. Contract No. 68-02-2262.
U.S. Environmental Protection Agency. June 24, 1976.
56 pp.
73.	Ifeadi, C.N. Screening Study to Develop Background
Information and Determine the Significance of Air
Contaminant Emissions from Pesticide Plants. Battelle
Institute. Washington, D.C. Contract No. 68-02-0611,
Task 12. U.S. Environmental Protection Agency. March
1975.
74.	Industrial and Engineering Chemistry. October 1965.
p. 163.
75.	Jones, A.R., and R.S. Brief. Evaluating Benzene
Exposures. In: American Industrial Hygiene Association
Journal. Esso Research and Engineering Company.
Linden, New Jersey. September 1971. pp. 610-613.
76.	Jones H.R. Pollution Control in the Petroleum Industry.
Noyes Data Corporation. Park Ridge. 1973.
77.	Kagasov, V.M., et al. Evaluation of the Process of
Obtaining Aromatic Hydrocarbons from Coke-Oven Gas.
Koks Khim 6_s38. 1970.
78.	Kenyon, R.L., and N. Boehmer. Phenol by Sulfonation.
In: Modern Chemical Processes, Vol. II. Reinhold
Publishing Corporation. New York. January 1960. pp.
33-42.
6-23
-------
79.	Keystone Coal Industry Manual, By-Product Coke-Oven
Plants in the United States and Canada. New York
Mining Information Services. McGraw-Hill, Inc. 1974
and 197 5.
80.	Kinkley, M.L., and R.B. Neveril. Capital and Operating
Costs of Selected Air Pollution Control Systems. Gard,
Inc. Niles, Illinois. For U.S. Environmental Protec-
tion Agency. Research Triangle Park, N.C. EPA Con-
tract No. 68-02-2072. March 1976. (Draft).
81.	Kirk-Othmer Encyclopedia of Chemical Technology. A
Stanton (ed.)., 2nd Edition. John Wiley and Sons, Inc.
New York.
82.	Kogan, B.A., et al. Experimental Verfication of a
Method for Calculating Benzene Hydrocarbon from Coke-
Oven Gas. Kaks Khim 1_0;49. 1972. As Reported in:
Atmospheric Emissions of Coking Operations - A Review
(an unpublished paper), Lowell White, Margaret Stasikowski.
83.	Kulujian, N.J. By-Product Coke Battery Compliance
Evaluation, Vol. I. PEDCo Environmental, Inc. Cincinnati.
Contract No. 68-02-1321. U.S. Environmental Protection
Agency. June 1975. pp. 3-25.
84.	Kulujian, N.J. Environmental Impact Due to Installa-
tion of Pushing Controls on By-Product Coke Batteries.
PEDCo Environmental, Inc. Cincinnati. Contract No.
68-00-3150. U.S. Environmental Protection Agency.
September 1975. pp. 7-36.
85.	Kunc, J., et al. Changes of the Gas Composition at
Various Distances from the Chamber Wall of a Coking
Furnace in the Course of the Coking Process. Breunt-
Chem. 4£:41. 1967. As reported in: Atmospheric
Emissions of Coking Operations - A Review (an unpublished
paper), Lowell White and Margaret Stasikowski.
86.	Lee, B.S. Status of the Hygas Program. Institute of
Gas Technology. In: Seventh Synthetic Pipeline Gas
Symposium. Chicago. October 27-29, 1975. 47 pp.
87.	Leonard, M.J., E.L. Fisher, M.F. Brunelle, and J.E.
Dickinson. Effects of the Motor Vehicle Control Program
on Hydrocarbon Concentrations in the Central Los Angeles
Atmosphere. In: Journal of the Air Pollution Control
Association, Vol. 26, No. 4. April 1976.
6-24
-------
88.	Lewis, W.A., and T.W. Hughes. Source Assessment;
Maleic Anhydride. Monsanto Research Corp. EPA-600/2-
76-032. U.S. Environmental Protection Agency - Office
of Research and Development. March 1977.
89.	Light Oil. Chapter 11. In: Coal, Coke, and Coal
Chemicals. Chemical Engineering Series, P.J. Wilson,
Jr., and J.H. Wells (eds.). McGrawHill Book Co. New
Yor^. 1950. pp. 336-337.
90.	Lippian, J.M. The Transportation of Hazardous Material:
Transport of Benzene by Tank Car. U.S. National Tech-
nical Information Service. A.D. Rept. No. 771105. May
1973.
91.	Maier, A., R. Ruhe, R. Rosensteel, and J.B. Lucas.
Health Hazard Evaluation/Toxicity Determination Report
72-90-107. National Institute for Occupational Safety
and Health. Rockville. NIOSH-TR-107A-74. U.S.
Environmental Protection Agency. February 1974.
92.	Marn, P.J., D.A. Horn, and T.W. Hughes. Solvent
Evaporation - Degreasing. Monsanto Research Corporation.
Dayton. Contract No. 68-02-1874. U.S. Environmental
Protection Agency. January 1976.
93.	Mashek, V. Contents of Gases and Liquids in Coke-Oven
Plant Atmospheres. In: Erdol and Kohle, Russia.
1972. As reported in: Atmospheric Emissions of Coking
Operations - Review; Lowell White and Margaret Stasikowski
(an unpublished paper).
94.	Masek and Sadlak. Emissions from Coke Plants, Hutnilisty
25:149. 1970. As reported in: Atmospheric Emissions
of Coking Operations - A Review; Lowel White and Margaret
Stasikowski (an unpublished paper).
95.	Motor Vehicle Facts and Figures '76. Motor Vehicle
Manufacturers' Association of the United States, Inc.
Detroit, Michigan. 1976. 160 pp.
96.	Neligan, R.E. Hydrocarbons in the Los Angeles Atmos-
phere. In: Archives of Environmental Health. Los
Angeles. 1962. pp. 67-77.
97.	Noncriteria Pollutant Emissions Report for Selected
Organic Emissions. Monsanto Research Corporation.
Dayton, Ohio. Contract No. 68-02-1874. U.S. Environ-
mental Protection Agency. December 1975.
6-25
-------
98.	Octane Crunch Threatens Petrochemicals. Chemical Week.
September 15, 1976.
99.	Otterson, E.J., T.A. Richard, M.R. Lischefski, and R.E.
Zimmerman. Furniture Stripping with Benzene-Based Sol-
vents. A Small Plant Health Problem. Wisconsin
Division of Health, Section of Occupational Health.
Thirty-Third Annual Meeting of the American Conference
of Governmental Industrial Hygienists. Toronto. May
1971. 93 pp.
100.	Outlook for Benzene, Toluene Told in Study. Hydro-
carbon Processing. April 1977.
101.	Overcapacity, Slower Demand Hit Styrene. Chemical and
Engineering News. April 4, 1977. p. 11.
102.	Patterson, R.M., M.I. Bornstein, and E. Garshick.
Assessment of Benzene as a Potential Air Pollution
Problem. G.C.A. Corporation. Bedford, Massachusetts.
Contract No. 68-02-1337, Task Order No. 8. U.S.
Environmental Protection Agency. January 1976. 32 pp.
10 3. PEDCo Environmental, Inc. Company Files.
104.	Perch, M., and R.E. Muder. Coal Carbonization and
Recovery of Coal Chemicals. In: Riegel's Handbook of
Industrial Chemistry, Seventh Edition. Van Nostrand
Reinhold. 1974. New York. pp. 193-206.
105.	Pervier, J.W., R.C. Barley, D.E. Field, B.M. Friedman,
R.B. Morris, and W.A. Schwartz. Survey Reports on
Atmospheric Emissions from the Petrochemical Industry,
Vol. III. Houdry Division/Air Products and Chemicals.
Marcus Hook. Contract No. 68-02-0255. U.S. Environ
mental Protection Agency. April 1974.
106.	1975 Petrochemical Handbook Issue. Hydrocarbon Process-
ing. November 1975.
107.	Petroleum Refining Handbook, Hydrocarbon Processing.
September 1976. p 177.
108.	Pilar, S., and W.F. Graydon. Benzene and Toluene
Distribution in Toronto Atmosphere. In: Environmental
Science and Technology, Vol. 7, No. 7. Easton, American
Chemical Society. July 1973. pp. 628-631.
6-26
-------
109.	Polyanskii, V.A., et al. Air Pollution, The Grounds of
a Modern Oil Refinery. In: Hygiene and Sanitation,
Vol. 33, No. 10-12. October-December 1968.
110.	Ponder, T.D. Benzene: Outlook through 1980. In:
Hydrocarbon Processing. Gulf Publishing Company.
Houston, Texas. November 1976.
111.	Powell, A.R. Gas from Coal Carbonization, Preparation,
and Properties, Chapter 25. In: Chemistry of Coal
Utilization. H. Lowry (ed.). John Wiley and Sons.
New York City. 1945.
112.	Preferred Standards Path Report for Polycylic Organic
Matter. U.S. Environmental Protection Agency. October
1974.
113.	Progress in the Implementation of Motor Vehicle Emis-
sion Standards through June 197 5. U.S. Environmental
Protection Agency. Washington, D.C. EPA 230/1-76-001.
June: 1975.
114.	Ragland, K.W., and J.J. Peirce. Boundary Layer Model
for Air Pollutant Concentrations Due to Highway Traffic.
Journal of the Air Pollution Control Association.
25(1):. January 1975.
115.	1976 Refining Process Handbook, Hydrocarbon Processing.
September 1976.
116.	Restricting Emission of Dust, Tar Mist, and Gas When
Charging Coke Ovens. Anthracite-Mining Association,
Essen. Washington, D.C. U.S. Department of Health,
Education, and Welfare. June 1962. 26 pp.
117.	Revision of Evaporative Hydrocarbon Emission Factors.
Radian Corporation. Research Triangle Park, N.C.
EPA-450/3-76-039. U.S. Environmental Protection
Agency. August 197 6.
118.	Reynolds, S.D., et al. Mathematical Modeling of Photo-
chemical Air Pollution-Ill Evaluation of the Model.
Atmospheric Environment. 8:563-596. 1974.
119.	Roberts, B.M., and R.L. Chamberlain. The Detarring of
Coke;-Oven Gas by Electrostatic Precipitator in the
United States. Dechema Monographier 48:307. 1973.
6-27
-------
120
121
122
123
124
125
126
127
128
129
130
131
Ross, J.F. Which Raw Material for Petrochemicals? In:
Hydrocarbon Processing, Vol. 54, No. 11. Gulf Publishing
Company. Houston, Texas. November 1975. pp. 760-76VV.
Runion, H.E. Benzene in Gasoline. In: American
Industrial Hygiene Association Journal. 36: 338-350.
May 1975.
Runion, H.E. Benzene in Gasoline II. Gulf Science and
Technology Company. In: International Workshop on
Benzene. Paris. November 9-11, 1976. 9 pp.
Sax, I.N. Dangerous Properties of Industrial Materials,
Fourth Edition. Van Nostrand Reinhold Company. New
York. 1975.
Seinfold, J.H., T.A. Hecht, and P.M. Roth. Existing
Needs in the Experimental and Observation Study of
Atmospheric Chemical Reactions. Report No. EPA-R4-73-
031. June 1973.
Settig, M. Organic Chemical Process Encyclopedia,
Second Edition. Noyes Development Corporation. Park
Ridge, Illinois. 1969.
Sherwood, R.J. Benzene: The Interpretation of Monitor-
ing Results. American Occupational Hygiene Association
Journal 15:409-21. 1972.
Solvent Emission Control Laws and the Coatings and
Solvents Industry. Environmental Science Services
Corporation. Stamford, Connecticut. 1967.
Stack Sampling at Petro-Tex Chemical Corporation.
Texas Air Control Board. Houston, Texas. Account No.
110-348-2. February 1976. 4 pp.
Steel and the Environment: A Cost Impact Analysis. A
Report to the American Iron & Steel Institute. Arthur
D. Little, Inc. May 1975.
Stern, A.C. Air Pollution, Second Edition, Volume III.
Sources of Air Pollution and Their Control. Academic
Press, Inc. New York City. 1968.
Stuart, J.A. Metropolitan Zone Air Quality and Meteorol-
ogy 1975 Annual Report. Southern California Air
Pollution Control District. El Monte, California.
1975.
6-28
-------
132.	Synthetic Organic Chemicals. U.S. Production and
Sales. U.S. Tariff Commission. 1940-1969.
133.	Synthetic Organic Chemicals. U.S. Production and
Sales. U.S. Tariff Commission. 1965-1970, 1972.
134.	Synthetic Organic Chemicals. U.S. Production and Sales
of Cyclic Intermediates. U.S. International Trade
Commission.
135.	Technical Guide for Review and Evaluation of Compliance
Schedules for Air Pollution Sources. PEDCo Environ-
mental, Inc. Washington, D.C. EPA-340/l-73-001-a.
U.S. Environmental Protection Agency. July 1973.
136.	Telephone Conversation between Mark Antel, EPA Station-
ary Source Enforcement, and Terry Briggs, PEDCo.
April 11, 1977.
137.	Telephone Conversation between R. Garbe, Emission Control
Technology Division (Ann Arbor) and J. Bertke, PEDCo
Environmental, Inc. July 1977.
138.	Telephone Conversation with D. Hardin, Certification
Policy and Support Branch of Certification Division.
Ann Arbor, Michigan. December 1976.
139.	Telephone Conversation with B. Johnson, Directory
Staff Manager, Stanford Research Institute. December
1976.
14 0. Telephone Conversation between Pete Jonker, Union Oil
Co., and Terry Briggs, PEDCo. April 18, 1977.
141.	Telephone Conversation with E. Kreher, Statistics
Department of Motor Vehicle Manufacturers Association.
April 1977.
142.	Telephone Conversation with R. Kruze, Standards Develop-
ment and Support Branch of Emission Control Technology
Division for Mobile Sources. Ann Arbor, Michigan.
December 1976.
14 3. Telephone Conversation between Jack Murray, American
Textile Manufacturers Inst., and Terry Briggs, PEDCo.
April 15, 1977.
6-29
-------
144.	Telephone Conversation with Dr. Leigh Short, Department
of Chemical Engineering, University of Massachusetts.
April 1977.
145.	Telephone Conversation between A. Trendholm, Environmental
Protection Agency, and Terry Briggs, PEDCo. April 11,
1977.
146.	Telephone Conversation with William Vatavukt, Environ-
mental Protection Agency, Cost Analysis Branch, and J.W.
McDonald of Ceilcote Corp. July 30, 1976.
147.	Telephone Conversation with Wendel Ward. American
Petroleum Institute. Washington, D.C. December 1976.
148.	Telephone Conversation with Marcia Williams, Office of
Air and Water Programs. October 197 6.
149.	Telephone Conversation with Ronald Young, National In-
stitute for Occupational Safety and Health.
150.	Telephone Conversation with R. Young, Representative,
American Petroleum Institute. New York, N.Y. December
1976.
151.	The Economics of Clean Air. Annual Report to the
Congress of the United States. U.S. Environmental
Protection Agency. February 1972.
152.	The Iron and Steel Industry. PEDCo Environmental, Inc.
Cincinnati. Contract No. 68-02-1321. U.S. Environ-
mental Protection Agency. December 1976.
153.	The Making, Shaping, and Treating of Steel, Ninth
Edition. H.E. McGannon (ed.). United States Steel.
Pittsburgh, Pa. December 1970. pp. 104-177.
154.	The Present and Projected Relationship Between Petro-
chemical Markets and Benzene Levels in Gasoline.
Energy and Environmental Analysis, Inc. Arlington,
Virginia. April 11, 1977.
155.	Turner, D.B. Workbook of Atmospheric Dispersion
Estimates. U.S. Environmental Protection Agency.
Research Triangle Park, N.C. AP-26. 1970.
156.	Vapor-Phase Organic Pollutants, Volatile Hydrocarbons
and Oxidation Products. National Academy of Sciences.
U.S. Environmental Protection Agency. Contract No.
68-02-0542. 1975.
6-30
-------
157.	Walker, P. Air Pollution Assessment of Benzene. Mitre
Corporation. U.S. Environmental Protection Agency.
April 1976. 105 pp.
158.	Walkley, J.E., L.D. Panotto, and H.B. Elkens. The
Measurement of Phenol in Urine as an Index of Benzene
Exposure. In: Industrial Hygiene Journal. October
1961. pp. 362-367.
159.	Watkins Cyclopedia of the Steel Industry, 12th Edition.
Steel Publications, Inc. Pittsburgh, Pa. 1969. pp.
31-44.
160.	Weast, R.C. Handbook of Chemistry and Physics, 48th
Edition. The Chemical Rubber Company. Cleveland.
1967.
161.	Whitma, N.E., and A.E. Johnston. Sampling and Analysis
of Aromatic Hydrocarbon Vapors in Air: A Gas-Liquid
Chromatographic Method. In: American Industrial
Hygiene Association Journal. Vol. 25. September-
October 1964. pp. 464-469.
162.	Wiley, S.K. Hydrocarbon Emissions from Chemical Load-
ing Operations. Resource Management Consultants. In:
68th Annual Meeting of the Air Pollution Control
Association. June 15-20, 1975. 15 pp.
163.	Young, R., et al. Walk-Through Survey of Gulf Oil
Refinery. Petro-Chemical Div. Philadelphia, Pennsyl-
vania. November 20, 1975.
16 4. Walker, D.C. et al. Demonstration of Reduced Hydro-
carbon Emissions from Gasoline Loading Terminals.
Amoco Oil Company. U.S. Environmental Protection
Agency. June 1975. 50 pp.
165. Telephone Conversation with John Sigsby, Emissions
Testing and Characterization Section, Environmental
Protection Agency. May 1977.
6-31
-------
APPENDIX A
A-1
-------
APPENDIX A
Data from three automobile exhaust emission studies are
presented. Table A-l gives emissions data from the Black
17
and Bradow study; Table A-2 gives emission test data from
52
Olson Laboratories. A Bureau of Mines auto emissions
4
study is described, and relevant test data are presented.
BUREAU OF MINES STUDY4 OF THE EFFECT OF AUTOMOBILE OPERATING
VARIABLES ON EXHAUST EMISSIONS
Six 1970-1972 vehicles were tested in a 1973 study.
4
Results of this test are listed in Table A-3. These
vehicles were equipped with the following emission-control
systems:
° A base-metal oxidation catalyst system with
exhaust gas recirculation (EGR)
° An Esso rapid-action manifold (RAM) thermal
reactor system with EGR
° Two systems with platinum oxidation catalyst and
EGR
° An Ethyl lean-reactor system with EGR
° An Esso dual-catalyst system (Monel reduction
catalyst and platinum oxidation catalyst).
The experimental program revealed two areas of interest:
first, the influence of fuel composition on exhaust hydro-
carbon emission and constituent benzene content; second, the
effect of temperature variation (simulating seasonal condi-
tions) on benzene emissions. In the fuel composition phase
of the study, all six vehicles were operated on high-aromatic
A-2
-------
Tab] r< A-l. TOTAT., HYDROCARBON AND BENZENE EMISSIONS
FROM TEN 197 5 AUTOMOBILES17
Vehicle
FIAa
Q
hydrocarbons,
g/mile
GCb total
hydrocarbons,
g/mile^
GCb
benzene,
g/mile^
197 2 Chevrolet
1.00+0.19 (3.00)
1.150+0.11
0.029
1975 Ford Proto-
0.307+0.082 (0.41)
0.214+0.034
0.008
type



197 5 Olds. Proto-
0.436+0.051 (1.50)
0.485+0.073
e
type



1975 Ply. Proto-
0.242+0.017 (0.41)
0.257+0.044
0.008
type



197 5 Chev. Impala
0.279+0.060 (1.50)
0.249+0.044
0 .004
197 5 Chev. Nova
0.291+0.030 (1.50)
0.264+0.048
0.003
1975 Ply. Fury
0.560+0.120 (1.50)
0.488+0.074
0.010
1975 Ply. Scamp
0.731+0.130 (1.50)
0.658+0.026
0.015
19 7 5 Ford Granada
0.581+0.115 (1.50)
0.550+0.041
0.004
197 5 Ford Torino
0.950+0.091 (1.50)
0.788+0.034
0. 015




Measurements determined by flame ionization analyzer (FIA).
k Measurements determined by gas chromatograph (GC).
c
Numbers in parentheses indicate applicable emission standard.
Multiply by 0.621 for conversion to g/km.
0
Data not reported.
A-3
-------
Table A-2. HYDROCARBON AND BENZENE EMISSION DATA PROM OLSON
LABORATORIES TESTING PROGRAM (1975 CALIFORNIA VEHICLES)52
Vehicle No.
Make
Model
CID
Hydro-
carbon
g/niile
Benzene,
g/mile
Benzene
hydrocarbon,
*
1
Chev
G10V
350
0.48
0 .007 '
1.4
2
Chev
Scout
350
0.66
0 .016
2.4
3
Chev
G10V
350
CO
o
**4
0 .020
1.8
4
Chev
G10V
350
0.74
0 .010
1.4
5
Dodge
TRAD
318
0.67
0 .022
3.3
6
Ford
F100
300
0.68
0 .021
3.1
7
Ford
F100
302
2.43
0 .108
4.4
8
Ford
Econ
3 51
0.56
0 . 016
2.8
9
Ford
F100
351
0.72
0 .010
1.4
10
CMC
Si cr
350
1.38
0 .028
2.0
U
AMC
Pacer
2 58
0. 35
0.012
3.4
12
Buir:k
Cnty
350
0.89
0 . 054
6.1
13
Duick
E3ta
455
0.54
0 . 018
3.3
14
Bu iek.
Regal
350
0.38
0.010
2.6
15
Cadi
Devi
500
0.77
0 .015
1.9
16
Chev
Monz
140
0. 30
0 .007
2.3
17
Chev
Came
250
0.49
0 .010
2.0
18
Chev
Mont
350
0. 51
0 .007
1.4
19
Chev
Caroe
350
0.86
0 .023
2.7
20
Chev
Mont
350
0.62
0 .010
1.6
21
Chev
Impa
400
0.44
0.007
1.6
22
Chry
Mewp
400
0.44
0 .011
2.5
23
Dodge
Dart
318
0. 22
0 .007
3.2
24
Dodqe
Dart
225
0.38
0 .005
1.3
25
Ford
Must
140
0.12
0.003
2.5
26
Ford
Gran
302
0.78
0.012
1.5
27
Ford
Gran
250
0. 33
0.008
2.4
28
Ford
Tori
351
0.57
0.011
1.9
29
Ford
LTD
460
0.64
0.012
1.9
30
Merc
Ma r
-------
Table A-3. INFLUENCE OF FUEL COMPOSITION ON TOTAL HYDROCARBON
EMISSION AND CONSTITUENT BENZENE IN EXHAUST FROM
PROTOTYPE LOW EMISSION VEHICLES4
Fuel
Totala
HC emissions,
£
Benzene emission
Total HC
wt. pet
g/mile
mg/I:m
g/mile
g/km






1972 Oldsmobile Delta 88(Car 403) With A 455-CID Engine And
Equipped With A Base-Metal Oxidation Catalyst And EGR
Typical clear I,
35 percent
aromatic (7202)
Indolene clear,
22 percent
aromatic (7203)
High alkylate,
10 percent
aromatic (7212)
0.33
0.43
0. 50
. 21
0. 27
0. 31
7.7
6.0
2.9
0.025
0.026
0. 015
1971 Ford LTD (Car 810) With A 351-CID Engine And Equipped
With ESSO Ram Reactors And EGR
Typical clear I,
3 5 percent
aromatic (7202)
Indolene clear,
22 percent
aromatic (72 03)
High alkylate,
10 percent
aromatic (7212)
0.17
0.11
0.13
0.11
0. 07
0.08
4.5
4.3
1.8
0.008
0.005
0.002
1971 Plymouth Fury III (Car 333) With A 360-CID Engine And
Equipped With Platinum Oxidation Catalysts And EGR
'.'ypical clear II,
40 percent
aromatic: (7221)
Indolene clear,
22 percent
aromatic (7203)
(Continued)
0.40
0.39
0.25
0.24
A-5
5.8
3.9
Is
16
0.023 i 14
0.015
-------
Table A-3 (continued).
Fuel
High alkylate,
10 percent
aromatic (7212)
Total
HC emissions,
g/ro4-le g/km
Benzene emission
Total Tic"
wt. pc L g/mile
0. 50
31
1.5
0.008
mg/km
1972 Ford Torino (Car 724) With A 351-CID Engine And Equipped
With Platinum Oxidation Catalysts And EGR
Typical clear II,
40 percent
aromatic (7221)
tndolene clear,
22 percent
aromatic (7203)
High alkylate,
10 percent
aromatic (7212)
0.72
0.79
0.45
0.49
0.86 I 0.53
4 . 2
3.2
1.3
0.030
0. 025
0.011
19
15
1971 Plymouth Fury III (Car 775) With A 360-CID Engine And
Equipped With The Ethyl Lean Reactors And EGR
Typical clear II,
40 percent
aromatic (722],)
Indolene clear,
22 percent
aromatic (7203)
High Alkylate,
10 percent
aromatic (7212)
(Cent iiracu)
0.43
0.44
0. 36
0.27
0.27
0 . 22
5.0
3.5
1.7
0.022
0. 015
0.006
14
A-r,
-------
Table A-3 (continued).
Total3
HC emissions,
Benzene emission3
Fuel
g/mile" g/km
Total HC
wt. pet g/mile mg/km
1970 Chevrolet Impala (Car 58) With A 350-CID Engine And
Equipped With GEM Monel NOx Reduction Catalysts And
Platinum Oxidation Catalysts
Typical clear II,
40 percent
aromatic (7221)
0. 30
0.19
3.7
0. 011
7
Indolene clear,
22 percent
aromatic (7203)
0. 34
0.21
2.5
0.009
6
High alkylate,
10 percent
aromatic (7212)
0.40
0.25
1.1
0.004
*
a All tests were conducted at 24°C ambient temperature.
Data are weighted in accordance with the 1975 Federal
test procedure and each value represents the average
of three replicate tests.
A-7
-------
(up to 40%), midrange-aromatic (21 to 22%), and low-aromatic
(8 to 10%) fuels. Table A-4 lists the fuel characteristics.
Five of the vehicles showed a decrease in benzene emission
with the decrease in fuel aromaticity. Three showed an
increase in hydrocarbon emission with decreasing fuel
*
aromaticity. The remaining vehicles showed few conclusive
effects attributable to fuel composition.
In the temperature-variation phase of the study only
three of the six vehicles were used in tests at ambient
temperatures of -4, 7, 24, and 35°C. Test procedures and
fuels were the same as in the earlier tests. Table A-5
lists the test vehicles and the data collected. The data
show that total hydrocarbon emissions were lowest at 24°C
and increased when the ambient temperature was either
higher or lower than 24°C. Generally, the same trend was
apparent in benzene emissions in most of the tests. The
increase in THC was much more pronounced when the tempera-
ture was decreased than when it was increased.
For comparison with tests of these prototype low-
emission vehicles, an additional study on the influence of
fuel composition was conducted with 10 production vehicles
4
(1970-1973) that had no exhaust control devices. Table A-6
lists these vehicles and the results of the study, which was
conducted with the same fuels and testing procedures used
earlier. Results show that total hydrocarbon emissions from
these uncontrolled vehicles were far higher than those from
the controlled vehicles.
~
It should be noted that increases in fuel aromaticity
result in increased fuel density, which can affect
carburetor metering, which in turn can affect air/fuel
ratio. Therefore, THC emissions may be influenced by
changes in air/fuel ratio rather than changes in fuel
aromaticity.
A-8
-------
Table A-4.
FUEL DATA'1

Typical
clear I
(7202)
Typical
clear II
(7221)
Indolene
clear
(7203)
High
alkylate
(7212)
Reid vapor pressure	
8.8
9.1
9.0
9.1
Specific gravity 	
.745
.755
. 720
.704
API gravity	
58.4
55.7
65.0
69.5
Octane number, research




method	
91
93
91
94
Distillation, °C




IBP	
34
3 r>

34
10 pet evaporated . . .
5 6
58
53
58
50 pet evaporated . . .
105
.1 07
96
101
90 pet evaporated . . .
163
16]
150
147
End point 	
202
199
189
197
Composition, vol pet (FIA)




Aromatics 	
33
39
21
8
Olefins 	
9
8
10
8
Paraffins 	
58
53
69
84
Composition, vol pet (GLC)




Aromatics 	
35
40
22
10
Olefins 	
9
7
11
9
Paraffins 	
56
53
67
81
A-9
-------
Table A-5. INFLUENCE OF AMBIENT TEMPERATURE ON TOTAL
HYDROCARBON EMISSION AND CONSTITUENT BENZENE IN
4
EXHAUST FROM PROTOTYPE LOW EMISSION VEHICLES
Fuel
Ambient
temperature
Total HC
emissions,'
g/mile
Benzene emissions
Total HC
wt. pet
g/mile
1972 Oldsmobile Delta 88 (Car 403) With a 455-CID Engine
And Equipped With A Base-Metal Oxidation Catalyst And EGR
Typical clear I,
35 percent
aromatic (7202)
Indolene clear,
22 percent
aromatic (7203)
High alkylate,
10 percent
aromatic (7212)
25
45
75
95
25
45
75
95
25
45
75
95
-4
7
24
35
-4
7
24
35
-4
7
24
35
0.82
0.45
0.33
0. 63
0.62
0.44
0.43
0. 52
0.98
0. 59
0. 50
0. 58
5.2
7 . 2
7.7
4.7
6.2
6.3
6.0
5.2
2,
2
2,
2.
0. 043
0. 032
0. 025
0. 030
0.038
0. 028
0.026
0. 027
0.018
0.014
0. 015
0. 012
1972 Ford Torino (Car 724) With A 351-CID Engine and Equipped
With Platinum Oxidation Catalysts And EGR
Typical clear II,
40 percent
aromatic (72 21)
High alkylate,
10 percent
aromatic (7 212)
24
-4
4.23
4.3
0.182
113
45
7
2.36
4 . 4
0.104
65
75
24
0. 72
4 . 2
0. 030
19
95
35
0. 82
4.4
0.036
33
25
-4
4 . 97
1.8
0. 090
56
45
7
3. 31
1.7
0. 056
35
75
24
0. 86
1.3
0.011
7
95
35
1. 06
1.5
0.016
1
(oonti nued)
A-10
-------
Table A-5 (continued).
Fuel
Ambient
temperature
Total HC
emissions,a
(V™ ile
Benzene emissions
Total HC
wt. pet
g/mile
mg/km
°F
°C
1971 Plymouth Fu.
Equipped
Typical clear II,
4 0 percent
aromatic (7221)
High alkylate,
10 percent
aromatic (7212)
ry III
ith T
25
45
75
95
25
45
75
95
(Car 7'
le Ethy!
-3.9
7.2
24
35
-3.9
7.2
24
35
'5) With A 360
L Lean Reactor
1.85
0.74
0.43
0. 47
2.15
0.77
0.36
0. 63
-CID Engine
s and EGR
4.7
5.8
5.0
3.7
1.6
2.0
1.7
1. 3
And
0.087
0. 043
0.022
0. 017
0.034
0.015
0.006
0. 008
54
27
14
11
21
9
4
5
Data are weighted in accordance with the 1975 Federal test
procedure and each value represents the average of three
replicate tests.
A-11
-------
Table A-6. INFLUENCE OF FUEL COMPOSITION ON TOTAL HYDROCARBONS
EMISSION AND CONSTITUENT BENZENE IN EXHAUST FROM 1970-1973
TEST VEHICLES4
Fuel
Total
HC emissions
a/mile
g/kra
Benzene emission
Total HC
wt. pet
g/mile
mg/km
1972 Oldsmobile 98 (Car 151) With A 455-CID Engine
Typical clear II,
40 percent
aromatic (7221)
Indolene clear,
22 percent
aromatic (7203)
1.79
1.72
l.li
1. 07
4.5
4. 2
0.081
0.072
1971 Ford Galaxie (Car 707) With A 351-CID Engine
Typical clear II,
40 percent
aromatic (7221)
Indolene clear,
22 percent
aromatic (7203)
4. 88
5.56
3.03
3.45
4.8
3.8
0. 234
0.211
1971 Plymouth Fury III (Car 76) With A 360-CID Engine
Typical clear II,
40 percent
aromatic (7221)
Indolene clear,
22 percent
aromatic (7203)
3.32
3. 54
2. 00
2.20
1972 Ford Torino (Car
Typical clear II,
40 percent
aromatic (7221)
Indolene clear,
22 percent
aromatic (7203)
(Con t j.nub;)
2.16
2.29
5.1
4.2
0. 164
0.149
769) With A 351-CID Engine
1. 34
1.42
6.0
4.5
0.130
0. 103
50
45
145
131
102
93
81
64
A-1 2
-------
Table A-6 (continued)
Fuel
Total
HC emissions
g/mile | g/km
Benzene emission
"Total HC 1
wt. pet g/mile
1970 Chevrolet Impala (Car 595) With A 350-CID Engine
Typical clear II,
40 percent
aromatic (7221)
Indolene clear,
22 percent
aroi latic (7-¦ 0 3)
3. 58
3.^
2.22
2.4.
5.5
4 . 3
0.197
0.169
1970 Pontiac (Car 400) With A 400-CID Engine
Typical clear II,
40 percent
aromatic (7221)
Indolene clear,
22 percent
aromatic (7203)
5.86
5.92
3. 64
3. 68
3.9
3.5
0.229
0. 207
1970 Volkswagon (Car 365) With A 1,600-CC Engine
Typical clear II,
40 percent
aromatic (7221)
Indolene clear,
2 2 percent
aromatic (7203)
2.44
1. 91
1. 52
1.19
5.2
4.3
0.127
0.082
1971 Chevrolet Vega (Car 68) With A 2,300-CC Engine
Typical clear II,
40 percent
aromatic (7221)
Indolene clear,
22 percent
aromatic [7203)
('".or tinucf)
4.92
5.54
3.06
3.38
3.8
3.4
0. 187
0.185
A-13
-------
Table A-4 (continued).
'uel
Total
I1C emissions
g/mi le
g/km
Benzene erniss ion_
Total 11 cr i
v.'t. - pet. j g/mile
g/km
1973 Ford Torino (Car 146) With A 351-CID Engine
Typical clear II,
40 percent
aromatic (7221)
Indolene clear,
22 percent
aromatic (7203)
2.79
2. 68
1.73
1. 67
5. 3
4.2
0. 148
0. 113
1973 Chevrolet Impala (Car 110) With A 351-CID Engine
92
70
Typical clear II,
40 percent
aromatic (7221)
2. 16
1. 34
5.2
0.112
69
Indolene clear,
22 percent
aromatic (7203)
2 .12
1. 32
4.3
0.091 :
i
56
All tests were conducted at 24°c ambient temperature.
Data are weighted in accordance with 1975 Federal test
procedure.
A-14
-------
Because of the minimal effect of variation in fuel
aromaticity, it was concluded that HC emissions are affected
mainly by changes in physical properties of the fuels (which
affect carburetor metering and thus the air/fuel ratio)
rather than by changes in chemical composition. Benzene in
the total hydrocarbons emitted by all 10 vehicles using the
higher aromatic fuel averaged 111 mg/km (0.178 g/mile).
This is approximately 10 percent higher than the volume of
benzene emitted from the vehicles using the lower aromatic
fuel. Had the carburetor of each vehicle been adjusted for
each fuel, the significant difference in the total hydro-
carbon (and therefore benzene) emissions attributable to
variation of fuel aromaticity might not have occurred.
A-15
-------
APPENDIX
B-l
-------
53
Table B-l. U.S. REFINERIES AND THEIR PRODUCTION CAPACITIES
LQKPNfT
LOCATION
REFINERY*
CAPACITY

PKOJCTlOi CAPACITY

LUBES*
(¦tyd**)
ASPHALT4
(•3/d*y)
COKE*
(¦etrtc tpd)
•OPERATING*
GASOlIKE
OUTPUT
(shtyo
-------
Table B-l (continued)
CCWAlTf
LOCATION
R£fIKEBY*
CAPACITY
(«3 /ity)
PRODUCTION CAPACITY
LUBES*
ASPHALT*
(atyd*y)
CCK£*
(a#trie tptf)
b
VERATitt0
tttOLIXE
OUTPUT
(¦3/d*y)
Cenex
Laurel, Hon.
6630



3070
QuMpllo fetrolM Co.
Corpus Christl, Ten.
9360
	
	
		
5100

Wilmington, Calif.
4610
	
—
522
398

Enid, Okla.
7870
175
223
150
5380
Charter International Oil Co.
Houston, Texas
11100
...
636
...
4450
Chevron Asph«1t Co.
Baltimore Kd.
2150

1750
...

Chevron Oil Co.
Perth Aoboy, N.J.
14000
...
3980
...
3420

El Paso, Tex.
11300
...
795
	
5450
Cities Service OU Co.
Salt Lake City, Utah
7160
—
—
...
2850
Lake Charles, La.
42600
1110
	
908
17300
Clark OU & Refining Corp.
81ue Island, 111.
10800
...
716
	
7600

Hartford, 111.
5720


409
6483
Coastal States Petrochemical Co
Corpus Christi, Tex.
21500
...
80
454
7030
Continental 011 Co.
Billings, Hon.
8350
	
557
...
3380

Commerce City, Colo.
4770
...
525
—
17C0

Ponca City, Okla.
186C0
366
477
454
9<:o

Westlake, La.
11200
—
—
454
5600

Wrensfvati, Minn.
3740
	
—

1CS0
Cosden 011 & Chemical Co.
Big Spring, Tex.
10300
	
1270
III
8740
Cotton Valley Solvents Co.
Cotton Valley, La.
1230
...

....
	
CRA Inc.
CoffeyviUe, Kan.
S570
239
...
272
2860

PhUlipsburg, Kan.
3180
...
318
. ..
1510

Scottsbluff, Neb.
795

	

350
Cross OU & Refining Co.
Smackover, Ark.
771
207
223
...
---
Crown Centra) Petroleifli Corp.
Houston, Texas
15900
...

272
8045
Crystal Xefining Co.
Carson City, Hich.
9S6
...

...

Delta Refining Co.
Memphis, Tenn.
4610

477
	
2230
Derby Refining Co.
Wichita, Kan.
4110
...
	
145
2020
Oiaaond Shac*ock 011 1 6aS Co.
Sunray, Texas
7470
...
398
	
4130
Douglas 011 Co. of Calif.
Paramount, Calif.
5570
—
2230
	
966

Santa Monica, Calif.
1300
...
922
...
...
Eddy Refining Co.
Houston, Texas
517
—
—

...
Edgington OU Co.
Long Beach. Calif.
4610
...
1190

—
Edglngtcn Oxnard Refinery
Canard, Cai i f.
398
—
—
—

Evangeline Refining Co., Inc.
Jenings, La.
684
...
—
—
95
Euon OU Co.
Benica, Calif.
13800
—
—
817
12000

Baytown, Tex.
63600
3980
1910
...
26500

Baton Rouge, Li.
70800
2390
4600
1960
33000

B111ings, Hon.
7160
...
2070
227
3700

Lindon, N.J,
43700
...
7310
	
18400
Faatrlss OU Corp.
Honument, tt. Hex.
795
...
...
	
95
Fletcher Oil 4 Refining Co.
Carson, Calif.
2510
—
...
—
557
Hint Che*. Co.
San Antonio, Tex.
191
—-

...
	
Gary Western Co.
Grand Junction, Colo.
1320


363
	
6ttV Oil Co., Ik.
Delaware City, Del.
a wo


1630
11700
-------
Table B-l (continued)
COHPAKT
LOCATIOM
REFINERY*
CAPACITY
(»J/
-------
Table B-l (continued).
amucr
LOCATION
RTFUERT*
CAPACITY
MMrfTM CVACIT?
LUBES*
(¦3/d»y)
ASPHALT*
(»3/ Calif.
USO

477
—
...
¦orth American Petroled Corp
Sha)low Water, Kan.
795
...
...
—
—
iorthMestem Refining Co.
St. Paul Park, Hinn.
9540
...
3500
...
3330
OKC Refining Inc.
Okmulgee, Okla.
3420
—-
223
...
1990
Osceola Refining Co.
West Branch, Mich.
1510
...
...
	
—
Ptsco, Inc.
Sinclair, Uy.
6360
...
366

2630
Pennzoll Co.
Rouseville, Penn.
1590
$61
...
...
448

Falling Rock, W. Va.
795
207
...
...
318
Phillips Petroleum Co.
Borger, Tex.
15100
—
—
—
10600

Kansas City, Kan.
13500
398
477

6121

Great Falls. Hon.
906
...
127

262

Woods Cross, Utah
3660
...
3S0
...
1750

Sweeney, Texas
13500
...

—
9060

Avon, Calif.
17500
266
—
1090
13600
Plateau, Inc.
Bloomfield, N.Mex.
811
...
...

350
Powerline Oil Co.
Santa fe Springs, Calif.
4530
...
795
...
2130
Pride Refining Inc.
Abilene, Tex.
2240
...
...
...

Quaker State 011 Ref. Corp,
Emlenton, Penn.
528
270
---
...
169
Farmers Valley, Penn.
1030
398
—
...
366

Newell. W. Va.
1540
572
—
...
398

St. Mary's, w. Va.
771
270
...

183
Quintana-Hcwel1
Corpus Christi, Tex.
1590
...
...
	
—
lock Island Refining Corp.
Indianapolis, Ind.
4690
—
477
...
2500
Sage Creek Refining
Cowley, Wy.
159

...
...
...
Son Joaquin Refining
Oildale, Calif.
2700
...
S34
...
...
Sequoia Refining Corp.
Hercules, Calif.
4290
...
—

2660
Seolnole Asphalt Ref.
St. Harks, F1 a.
795
...
393
	
—
Sfcell Oil Co.
Martinez, Calif.
15900
716
16S0

7680

Wilmington, Calif.
15300
...

1630
7360

Wood River, 111.
41300
890
3580
—
24EOO

Norco, La.
38200
...
9S4
817
20200

Cin1*a. N.hex.
3160
...
134
—
1830

Deer Park, Tex.
45800
954
604
—
19100

Odessa, Tex.
5090
...

	
2420

Anacortes, Wash.
14500
...
...
—
7710
Skelly Oil Co.
£1 Oorado, tefl.
11700
...
—
454
61S0
Somerset Refinery
Somerset, Ky.
445
•
...
	
—
Sound Refining Inc.
TacoM, Wash.
716
302
413
...
—
Southland Oil Co.
Crupp. Miss.
668

229


-------
Table B-l (continued).











flUfiUCTlON CAPACITY
k








REFINERY*



OPERATIHS*


LUELS*
ASPHALT*
(a^/day)
COKE*
(setiic tpd)
GASOLINE
COMPANY
LOCATION
CAPACITY
(¦Vdiy)
OUTPUT
(¦Vday)
Southland Oil Co.
Lumfcerton, Kiss.
954

372



Sandersville. Hiss.
1750
...
554
..j
	
Southwestern OH ft Ref. Co.
Corpus Christi » T*x.
15900
	
	
—
3720
Southwestern fief. Co.
LaBarge. Wy.
52
	
	
—
	
StomUrd Oil of Calif.
Kena i. Alaska
3500

48
—
	

BakersHeld, Calif.
4130
	
7S
—
1160

£1 Scgundo, Calif.
36600
	
1320
2000
21400
*
Richmond, Calif.
30200
1590
1 750
...
20300

Barber's Point, Hawaii
6360
...
207
...
1860

Portland, Oregon
2230
—
1370
...
	

Richmond Beach, Wash.
716

636
...
—
Standard 01] of Ky.
Pascaqoula, Hiss.
38200
—
...
...
22800
Standard 011 of Ohio
Lima. Ohio
26200
334
—
5S4
8700

Toledo, Ohio
19100
...
1110
S90
8700
Son 011 Co.
Toledo, Ohio
19900

—
—
13800

Ouncan, Okla
7710

...
363
4 350

Tulsa, Ok la.
14100
—
668
272
6420

Marcus Hook, Penn.
26200
2700
1910
—
13500
Stfiland Ref. Co.
Bakersfield, Calif.
1400
...


151
Suntide Refining
Corpus Christi• Ten.
9060
...
—
213
4710
Tenneco Oil Co.
Bakersfield, Calif.
191
	
—
—
...

Chalmette, La.
15400
—
—
318
9810
Ttsoro-Alaskan Petr.
Kenai, Alaska
6040
...
...
...
...
Tcsoro Petrolewi
Wolf Point, Hon.
Carrizo Springs, TtJU
398
2070
...
:::
...
—

Newcastle, Wy.
1670
...
...
—
47?
Tcjuco, Inc.
Wilmington, Calif.
11900
—
—
1500
11800

Lawrencevilie. 111.
13400
—
429
—
7080

Lockport, III.
11400
—
—
272
5990

Convent, Is.
223CO
—
...

1230

Westville, N.J.
14000
—
—
—
5250

Anacortes, Wash.
10000
	


3550

Casper, Wy.
3340
—
239
"in
1500

Port Arthur, Tex.
64600
3180


26209

Amarillo, Tex.
3180
...
—
91
1620

El Paso, Tex.
2700
...
...
91
1510

West Tuisa, Okla.
7950
—


3580

Port Neches, Tex.
74 70
—
1430
...
	
Texas Asphalt & Ref. Co.
Fort Worth,. Tex.
477
—

...

Texas City Refining
Texas City, Tex.
9540
—
—
...
3550
THe Refinery Corp.
Covwnerce City, Colo.
2780
...

	
2590
Three Rivers Refinery
Three Rivers, Tex.
238
127
19
	
	
Tfcriftway Co.
Bloonfield, N. Hex.
338
...
—
	
	
Tonkawa Refining
Tonkawa, Okla.
954
...
—
	
	
Toscopetro Corp.
Bakersfield, Calif.
4220
...
...
191
4300
Total Leonard. Inc.
Alna, Hich.
6S00
...
477
...
3530
U.S. OU ft Refining
Tacoaa, Uasfc.
2540

477

398
-------
Table B-l (continued)
COMPANY
LOCATION
REFINERY*
CAPACITY
(•3/dajr)
rwcucnoi capacity
LUBES*
(¦3/nt, Ml.
24200
...
318
908
11300

Kederland, Tex.
18400
$57
8S9

8060
Union Texas Petroleui
Winnie, Tex..
1510
...
—

1256
United Refining Co.
Warren, Penn.
6040
—
636

2290
ValvoHne OU Co.
Freedom, Penn.
986
207
—
...
...
tickers Petrolei*
Ardwore, Okla.
5000
—
79S
	
2610
Vulcan Asphalt Ref.
Cordova, Ala.
477
—
—
	
...
Warrior Asphalt Co.
Hott. Ala.
385
---
274
—
...
West Coast 011 Co.
Bakersfield. Calif.
2540
...
636

318
Westeo Refining Co.
Cut fiar.k. Hon.
741
...


242
Mestland Oil Co.
Williston, N.D.
741
...
—

318
Winston Refining
Fort Worth. Te*.
2390

---
	
405
Wlrtback Oil Co.
Plymouth, 111.
238
—
...
_ _ _
...
Wolfs head 011 Ref.
Reno, Penn.
334
80

— — «¦

Tetter OU Co.
Collar, 11).
159
...
-—

...
Young Refining Corp.
Douglasvllle, St.
398
i 207
...

TOTAL	UCMM	32S31	107073	3W28	10SM5*
Source: (i) "Annual	Svrvvy.* Oil t tn J. I April 1974.
(b) Nationil Petrol®** Newt. fit Boot. md-Kty 1974, H.Y., Hc6r«w-H111, 197*.
-------
Table B-2. PETROLEUM-BASED BENZENE PRODUCERS (1976)


Benzene a


Capacity,
Company
Location
106 gal -
Allied Chemical
Winnie, Texas
3
Amerada Hess Corp.
St. Croix, Virgin Islands
25
American Petrofina, Inc.
Big Spring, Texas
45

Port Arthur, Texas
15
Ashland Oil, Inc.
Ashland, Kentucky
j 50

North Tonawanda, New York
15
Atlantic Richfield Co.
Houston, Texas
44

Wilmington, California
12
Charter International
Houston, Texas
5
Oil Co.


Cities Service Co., Inc.
Lake Charles, Louisiana
! 25
Coastal States Gas
Corpus Christi, Texas
1 70
Prod. Co.


Commonwealth Oil Refining
Penuelas, Puerto Rico
185
Co.


Crown Central Petroleum
Pasadena, Texas
23
Corp.


Dow Chemical Co.
Bay City, Michigan
30

Freeport, Texas
50
Exxon Corp.
Baton Rouge, Louisiana
65

Baytown, Texas
62
Gulf Oil Corporation
Alliance, Louisiana
70

Philadelphia, Pa.
33

Port Arthur, Texas
38
Kerr-McGec Corp.
Corpus Christi, Texas
16
Marathon Oil Co.
Texas City, Texas
6
Mobil Oil Corp.
Beaumont, Texas
60
Monsanto Co.
Chocolate Bayou, Texas
75
Pennzoil United, Inc.
Shreveport, Louisiana
15
Phillips Petrol. Co.
Sweeny, Texas
22

Guayama, Puerto Rico
110
Shell Oil Co.
Deer Park, Texas
75

Odessa, Texas
6

Wood River, Illinois
40
Skelly Oil Company
El Dorado, Kansas
13
Southwestern Oil £.


Ref. Co.


Standard Oil Co. of
El Scgundo, California
23
California


Standard Oil Co. (Ind.)
Texas City, Texas
85
Standard Oil Co. (Ohio)
Marcus Hook, Pa.
8
Sun Oil Co.
Marcus Hook, Pa.
15

Corpus Christi, Texas
35

Tulsa, Oklahoma
24
Tenneco, Inc.
Chalmette, Louisiana
10
Texaco, Inc.
Port Arthur, Texas
45

Westville, N.J.
35
Union Carbide Corp.
Taft, Louisiana
70
Union Oil Co. of
Lemont, Illinois
19
California


Union Oil-American
Beaumont, Texas
19
Petrof ina


Union Pacific Corp.
Corpus Christi, Texas
10

Total
1,701
a To convert to cubic meters multiply by 3.785 x 10~3,
B-8
-------
APPENDIX C
C-l
-------
Table C-l. COAL-DERIVED BENZENE
2 8
PRODUCTION FACILITIES (1976)
Company
Location
Benzene capacity
January 1. 1976
10° gal km3
Armco Steel Corp.
Bethlehem Steel Corp.
C.F. & I. Steel Corp.
Interlake, Inc.
Jones & Laughlin
Steel Corp.
Middletown, Ohio
Bethlehem, Pa.
Lackawanna, N.Y.
Sparrows Point,
Maryland
Pueblo, Colorado
Toledo, Ohio
Aliquippa, Pa.
4
8
15
3
1
10
Northwest Industries,
Inc.
Lone Star, Texas
1
4
U.S. Steel Corp.
Clairton, Pa.
Geneva, Utah
45
4
170
15

Total
94
355
C-2
-------
APPENDIX D
D-l
-------
Table D-l. LOCATIONS AND CAPACITIES OF ORGANIC
2 ft
CHEMICAL PLANTS UTILIZING BENZENE FEEDSTOCKS
Company
Capacity Production
January 1, 1976. "Mllior.a of
Nitro- j	Ethyl-	Maioic |
benzene 'Aniline . ix.nzeno [Styrcne anhydride j Cua-^nc ,?r.c lol
Kor.o-
C^iioro-
b€.n*er»e
DicMcto-
ben2enc
(0- and ?-
» • IL.;
Cyclohcxar»e 6".
•» r «r.;
.Ldr
ir.cr.e65
A. .c- «.	i v-orp.
Ar< ; iv .v. Cyanaitiid Co.
A -• « i ;v r <»r ln».
A: i • •/»•
¦j , inc.
oil. ll.'.
C 1 .i: ' Ci Arnl bc-
r, . , I n<; .
S' .ir <"S CJfc,
1 r.t; v'o .
Cor.t .r-.-'itAl tul Co.
C^ > • v.i t . Inr .
t.:. Torn, d«.
S'a i 1 a , H. y.
H. i (if.. Tux .
wa»i«.
«»r j.U«jeville» Pa.
Cicero, til.
Follanabee. W. Va.
I Djy town, i«x.
| 5.uitt, r<> Sprin^a, Cal.
[ I*.i jr"a
-------
TABLE D-l (continued).
Company
Location



Capacity Production
January 1, 1176 'millions of
•
lb».)




Wltro-
bansana
Anlllna
Ithyl-
bansana
Styran*
Malalc
anhydrlda
Cua«r.a
t'hanol
Mono-
Chloro-
bansana
Oichlcro*
bans ana
(0- and PS
1
fr
Cyclohaxana
Oa'.tr^er.t
al*yl«;.e
(Lir.aar i-.d
Braachad)
Marathon Oil Co..
City, Tax.





190





VericSt'^ Co.
llo'Jiton, To*.






N.A.




MoMy fhcrr.ii'« 1
:;»w M«rt 1 n*v 11 la, W. v«
135
100









Kctvirt-iti, Ci>.
Chora late Dn/uu, Tax.
!'¦ ¦> in Jt* t., lil.
' '<•*«¦ CI ty , T"<.
.«* . 1 ou : » , :*.t>.
10

•
•
1300
105
(SO
500
315
28

2^5
Mom roru Ch'f lcmlcaic.
Ire.
Tuscaloosa, Ala.
L 11, N.J.
> o r r i•, III.




w
o ©

ISO




Rut-. c CV.c-.iral s ,
1p.- •
feis-.'Jr, La.
?5
55









S*e : i-y Oi * v-'o.
11 Doi.v'.o, Kan*.
1



135
95




Sol'v-xt Ch(J.">iCAi Co.,
Inc -
Mass.
Nitxj.tro ValxS, N.Y.
i





N.A.
N.A.
3«
20


$pcc<*lty Cry.>:»Lr»,
Ire.
Trvi iuI.i !<.'. Cf*l.
¦
,






3


StJi
i:r 1 >w.n »• City, Dai .
k.-.» r r>y , s..







75
•0*
16«


M.w «. » ci l Co. of
c.u .t,...
rirli ..«•>,1. V «l .





10n
V.




ftJCl.1t .1 , • , 1 ( lei. )
'>l.n <" I t y . 1 >* •



*40

v.





nt 1- - 1 I" f{ r,,.







N.A.




Sou Oi 1 v i'-
			 f'!i - int I , lax.


>5
• 0

250





T»r,i>r..., liu-.
N..1.


21








I.K .
W • ' vl I If, N.vl.
I" V t Ai r hur , 1 .tX.





200



40

Union Carbide, Corp.
r.-nuplm, »\R.
•J t-ijr « 't, Tax .
1* ;>mi(i hrook. N.J.
Ci.n lot ton . M. Va.


140
300

MO
2P0
ISO



l^C-
Union Ci I Co. of
Ca1i'ornIA
Li?aur.ont, 7gx j








34

linin - J-aci'ic Corp.
C-);pi..« Ciuiatl, Vax. j








27

Unitrd Steol
Nrvjllc laland, Pa.
C * j rton, Va . .
Hivarhlil, Ohio
I



40

N.A.
2C0




Wiico Chw^lcal Corp.
Cicaon, Cal.










55
TOTALS

10«J
••5
• 56«
7075
415
J7I5
275$
690
2(J
416
L«a
4 To ooAwri 1° MlNlil^ly toy 0.414.
6 Production OtfMUy 1a *111 tan• •» oallafta-
0 Coabinad athylbansan<» production at Chocolata Bayou* Taxaa and «t Taxaa City, Taxaa OQuala 1450. taparata valuaa
not avtilabla.
< Production capacity for o~dichlorobansana only.
* Production capacity for p-dichloroban»ana only.
D-3
-------
AMBIENT
EMISSIONS
TO ANILINE
PRODUCTION
CRUDE
NITROBENZENE
AIR
WATER, DILUTE
SODIUM CARBONATE
BENZENE
NITROBENZENE
(REFINED)
MIXED
ACID
WASH-WATER
WASTE
WASTE
SPEND ACID
WASHER
NITRATOR
co
ct
TO RECOVERY
CgHg + HN03H2S04t C^HrNOn + H20
Figure D-l. Flow chart for nitrobenzene manufacture
from benzene and nitric acid.
Source: Ref. 59.
-------
BENZENE RECYCLE
TO REACTOR
ETHYLBENZENE
ETHYLENE
ALUMINUM
CHLORIDE
COMPLEX
HEAVY
(POLYETHYL) BENZENES
AND TAR
BENZENE
RECYCLED
AND FRESH
RECYCLE (POLYETHYL)BENZENES
CONDENSER
OFF-GAS
SCRUBBING
SYSTEM
WATER WASH
AND SETTLER
uj x:
f-sl ZD
UJ o
CO <_>
>- u
CAUSTIC WASH
AND SETTLER
c_>
cc
A1C1,
C6H6 + C2H4 *"C6H5C2H!
Figure D-2. Flow chart for ethylbenzene manufacture
from benzene and ethylene.
Source: Ref. 81
-------
TO ABSORBERS
BENZENE
MALEIC
ANHYDRIDE
i-CHCO + 2H,0 + 2CO
CHCO
WASTE
MIXER
VAPOR
COOLER
< o
> <_>
FUSED
SALT
COOLER
CONVERTER
CONDENSER
Figure D-3. Flow chart for the manufacture ot maleic anhydride
by catalytic vapor-phase oxidation of benzene.
Source: Ref. 59.
-------
FRESH
BENZENE


OC


LU
CxL
O

rsi
M
z
o
<


CL
o
OC


Q£
O.



LU



Q
CONDENSER
COMBINED
FEED DRUM
k-l
o
<_>
Ui
CONDENSER
o
o
a:
LU
OC
CONDENSER
CUMENE
BOTTOMS
CUMENE
FRESH PROPYLENE
PROPANE
PROPANE
C6H6 + CH3CH
CH,
PHOSPHORIC
ACID
SOLID
C6H5CH(CH3>2
Figure D-4. Process for the manufacture of cumene.
Source: Ref. 81.
-------
IMPURE CUMENE RECYCLE
CUMENE
HYDROGEN
EMULSIFIERS
ACETONE
>• PHENOL
AIR
SULFURIC ACID
RECYCLE ACID
o
I— o

OXIDIZER
SEPARATOR
HYDROGENATOR
ACIDIFIER
ACETOPHENOHE
CH6H5(CH3)2 + °2	^C6H5C^CH3^2 °°H
c6h5c(ch3)2 00H—K6H50H + (CH3)2 CO
Figure D-5. Flow diagram for the manufacture of phenol by the
cumene peroxidation process.
Source:
Ref.
59.
-------
BENZENE OR —
CHLOROBENZENE
HYDROCHLORIC ACID,
BENZENE,
CHLORINE.
CHLORINATOR
D
I
WATER
1
-~VENT
cc

on
00

CO


CO



ZD
oc


oc
o


o
V)


vo



SODIUM
HYDROXIDE
I
CHLOROBENZENE
t
HYDROCHLORIC
ACID .
NEUTRALIZING
TANK
o
<_>
SETTLING
TANK
—*-DICHLORO- AND
POLYCHLOROBENZENES
TO DISTILLATION
CgHg + Clg—~"CgHgCl + HC1
CgHgCl + Clj-^CgH4Cl2 + HC1
DICHLOROBENZENE
SLUDGE TO RECOVERY
Figure D-6. Flow diagram for the manufacture of
chlorobenzene and by-product dichlorobenzenes.
Source: Ref. 59.
-------
WATER OUT
WATER IN
RECOVERABLE
SOLVENT
MIXTURE
SODIUM CARBONATE
STEAM
SLUDGE
CONDENSATE
PRODUCT
WATER
SEWER
U. LU
OL O
CO O
o
PREHEATER
SETTLING
TANK
DEHYDRATING
TANK
3
O
an
Ll.
BOTTOMS
Figure D-7. Typical solvent re-refining installation.
Source: Ref. 50.
-------
APPENDIX E
E-l
-------
PRESSURE-VACUUM
VENT
GAUGE HATCH
MANHOLE
LIQUID LEVEL
MANHOLE
NOZZLE
Figure E-l. Fixed-roof storage tank.
Source: Ref. 29.
WEATHER SHIELD
LIQUID LEVEL
DRAIN
VENT
ROOF SEAL
(NOKMETALLIC
OR
METALLIC)
f4r
GUIDE RODS
HINGED CENTER SUPPORT
= NOZZLE
MANHOLE
Figure E-2. Double-deck floating-roof storage tank
(nonmetallic seal).
Source: Ref. 31.
E-2
-------
ROOF CENTER SUPPORT
FLEXIBLE DiAPriKAGM RClGh
GAUGE HATCH
LIQUID LEVEL
MANHOLE
NOZZLE
Figure E 3. Variable vapor space storage tank (wet-seal lifter type).
Source: Ref. 31.
-------