EPA-450/2-89-012b
April 1990
ANALYSIS OF AIR TOXICS EMISSIONS,
EXPOSURES, CANCER RISKS AND
CONTROLLABILITY IN FIVE URBAN AREAS
VOLUME II
Controllability Analysis And Results
U. S. ENVIRONMENTAL PROTECTION AGENCY
Office Of Air and Radiation
Office Of Air Quality Planning And Standards
Research Triangle Park, North Carolina 27711
-------
This report has been reviewed by the Office of Air Quality Planning
and Standards, U. S. Environmental Protection Agency, and has been
approved for publication. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
EPA 450/2-89-012b
-------
CONTENTS
EXECUTIVE SUMMARY vii
I OVERVIEW OF STUDY 1
A. PURPOSE OF STUDY 1
B. NATURE OF URBAN AIR TOXICS PROBLEM 2
C. STUDY FOCUS 3
1. Emphasis on Cancer 3
2. Modeling Scenarios 6
3. Caveats 8
4. Methodology 12
II STUDY METHODS 15
A. DATA BASE 15
B. REGULATORY IMPACT MODEL 20
1. Model Overview 20
2. Growth Rates 25
3. Sample Calculation 25
C. CONSTRAINT FILES 30
1. Overview 30
2. Quality Assurance 38
3. Model Assumption Review 39
III RESULTS 43
A. BASE YEAR SCENARIOS 43
B. FUTURE YEAR SCENARIOS 46
1. Excess Cancer Incidence 46
2. Maximum Individual Risk 55
C. DISCUSSION 58
ABBREVIATIONS AND ACRONYMS 61
REFERENCES 63
APPENDICES
A. CHANGES TO BASE CASE INVENTORY A-l
B. MOTOR VEHICLE EMISSION PROJECTIONS B-l
C. CONSTRAINT FILES C-l
iii
-------
TABLES
Number Page
I.I Summary of Scenarios (1995) 7
II.1 Total VOC Emissions in Each Source Category 18
II.2 Benzene Emissions Added by Source Category 18
II.3 Methylene Chloride Emissions Added by Source Category 19
II.4 Perchloroethylene Emissions Added by Source Category 19
II.5 Formaldehyde Emissions Added by Source Category ... 19
II.6 Population and Economic Growth Factors (1980-1995) —
State Two-digit SIC Allocated to Selected Study
Areas 26
II.7 Stationary Source Sample Calculation —'Regulatory
Impact Model 28
II.8 NESHAP and NSPS Constraints — Expected Controls
Scenario 32
II.9 Miscellaneous Measures Added to the Additional
Controls Scenario 33
11.10 Measures Added to the PM Additional Controls
Constraint File 35
11.11 Candidates for Federal Rules and CTGs — VOC
Additional Controls Case 36
11.12 Candidates for ACT and Air Toxic Regulations —
VOC Additional Controls Scenario 37
11.13 Changes to RIM VOC Constraint File 41
III.l Comparison of All Study Area Base Year VOC Emissions
and Related Incidence (Excluding Formaldehyde) under
NEDS Control Efficiency vs. SIP Control Level
Assumptions 44
III.2 Comparison of All Study Area Base Year VOC Emissions
and Related Incidence under NEDS Control Efficiency
vs. SIP Control Level Assumptions 45
III.3 Annual VOC Related Incidence 48
III.4 Annual PM Related Incidence 49
III.5 Changes in Annual VOC Related Incidence Via
Additional Controls 54
A.I Methylene Chloride Emissions Added by Source Category A-4
A.2 Formaldehyde Emissions Added by Source Category . . . A-4
A. 3 Benzene Emissions Added by Source Category A-5
A.4 Perchloroethylene Emissions Added by Source Category. A-5
B.I Motor Vehicle Air Toxic Emission Factors B-2
C.I VOC Current Rules Constraint File C-2
C.2 VOC Additional Controls Constraint File C-5
C.3 PM Current Rules Constraint File C-10
C.4 PM Additional Controls Constraint File C-12
IV
-------
FIGURES
Number Page
1 Pollutants Contributing to Five City Average
Aggregate Cancer Incidence x
2 Sources Contributing to Five City Average Aggregate
Cancer Incidence xi
3 Five City Controllability of Air Toxics in 1995. . . xiv
4 Incidence Changes by Scenario — Total Five City
Excess Annual Cancer Cases xv
5 Number of Sources with Maximum Individual Risk Greater
than One in a Million (1X10~6) xvi
I.I Pollutants Contributing to Five City Average Aggregate
Cancer Incidence 4
I.2 Sources Contributing to Five City Average Aggregate
Cancer Incidence 5
II.1 Allocation of Surface Coating Emissions to Different
Uses ' 17
II.2 Air Toxics Controllability Study Methodology .... 21
II. 3 Regulatory Impact Model 22
III.l 1995 Estimated Incidence by Scenario Compared with
1980 Base Year 47
.111.2 Five City Controllability of Air Toxics in 1995. . . 51
III.3 Number of Sources with' Maximum Individual Risk Greater
than One in a Million (1X10~6) 57
III.4 Total Five City Incidence Changes by Scenario —
Excess Annual Cancer Cases 60
A.I Allocation of Surface Coating Emissions to Different
Uses A-2
-------
EXECUTIVE SUMMARY
OVERVIEW
The Environmental Protection Agency (EPA), under its
National Air Toxics Strategy (U.S. EPA, 1985a), has been
encouraging State and local air pollution control agencies to
assess and mitigate their urban air toxics problems. These
problems are characterized by complex multi-source interactions
and multi-pollutant exposures. Numerous studies, including
Volume I of this report, suggest that area-wide lifetime excess
cancer risks from urban air toxics may range from about one in
10,000 to one in 1,000, and that cancer incidence may range from
1 to 23 excess cases per year per million population (Lahre,
1988). In addition to the typical area source problems in urban
areas, high risk point sources in the proximity of urban areas
can pose problems for individuals in areas of maximum exposure to
those sources.
Our understanding of the nature of the urban problem is
evolving. The basic tools (e.g., emission factors, potency
numbers) to estimate impacts for the direct emissions of certain
compounds change over time. Also, more needs to be done to
determine the consequences of secondary pollutant formation where
the potency of certain compounds may increase dramatically.
The purpose of this study (Volume II) is to gain some
initial insight into the controllability of the urban air toxics
problem as it is now understood. In particular, the objective is
to investigate the prospects for reductions in aggregate cancer
risk that may result from current national and local regulatory
activities and to estimate the potential for further reductions
that certain additional measures might achieve.
NATURE OF URBAN PROBLEM
The most commonly used measure of cancer risk in urban
assessments is "aggregate cancer incidence," which is a measure
of the excess cancer cases over an entire area associated with
multi-source, multi-pollutant exposures to air toxics. (This is
vii
-------
also called "population risk.11) Incidence is typically expressed
as the number of excess cancer cases expected in a single year,
but is estimated based on an assumed 70 year "lifetime" exposure.
In this report, incidence is considered additive in that the
individual incidence from all the pollutants under study are
added together. Incidence is also population-normalized by
adjusting to a per million persons rate. As such, the measure of
cancer risk most commonly expressed herein is "excess aggregate
additive incidence per year per million population."
Normalization by population allows incidence from one urban area
to be compared with another.
Volume I of this report explores the nature and magnitude of
the urban air toxics problem and presents a base year analysis
involving dispersion modeling of emissions data for five urban
areas in the United States. In this analysis, the available
emissions and source data were compiled and used as input to
EPA's Human Exposure Model (HEM) (U.S. EPA, 1986) to estimate
ambient air concentrations and population exposures to the
following known or suspected air carcinogens:
arsenic ethylene oxide
asbestos formaldehyde
benzene gasoline particulate
benzo(a)pyrene, or B(a)P gasoline vapors
beryllium manganese
1,3-butadiene mercury
cadmium methylene chloride
carbon tetrachloride nickel
chloroform perchloroethylene
chromium (VI and total) trichloroethylene
diesel particulate vinyl chloride
ethylene dichloride
These compounds are suspected contributors to aggregate
cancer incidence in urban areas and are those for which cancer
unit risk numbers have been established. From this modeling
investigation, estimates were made of the sources and pollutants
contributing to additive (i.e., multi-pollutant) cancer risk
throughout each urban area. Emphasis was placed on estimating
area-wide population risk, i.e., aggregate incidence, from multi-
pollutant, multi-source exposures. The year 1980 was nominally
Vlll
-------
defined as the base year in this analysis, although some later
year data were selectively incorporated into the data base.
Figures 1 and 2 are based on average results from the five
cities and indicate the pollutants and sources, respectively,
that may be of the most concern in a typical urban area. Area-
wide individual lifetime cancer risfcs averaged about 4 X 10~4,
ranging from 1.5 X 10"~4 to 7 X 10~4. Maximum individual risks
can range higher and are generally associated with large point
sources. The major contributing source categories tend to be
motor vehicles and small point and area sources. The major
contributing pollutants tend to be chromium, formaldehyde,
products of incomplete combustion (including gas and diesel
vehicle emissions), benzene, and 1,3-butadiene.
NATURE OF THIS STUDY
The object of Volume II is to analyze the base year
emissions data base developed in Volume I (with minor
adjustments) and determine (1) what urban risk reduction is
likely to occur as a result of ongoing regulatory activities and
(2) what further reductions might be possible with the
application of additional measures. Current and expected
programs which will address air toxics within the time frame of
this study (i.e., 1995) include several Federal regulations for
stationary and mobile sources, [e.g., National Emission Standards
for Hazardous Air Pollutants (NESHAP), New Source Performance
Standards (NSPS), and the Federal Motor Vehicle Control Program
(FMVCP)], revisions to existing State implementation plan (SIP)
requirements that will reduce both air toxics and criteria
pollutant emissions, and specific air toxics rules established at
the State and local level. A past study indicated that up to 50
percent of the overall air toxics problem may have been reduced
by such measures (Haemisseger et al., 1985). In addition, there
are potential risk reductions from other requirements which may
be available for future consideration (e.g., adopting for all
sources within a category the maximum control on any facility
within that category). The three control scenarios which are the
IX
-------
If)
w
o
-------
1.8 I
=•8
s
FX 8
§§8 ~
8§gg «
§o^§^
ooo~ ^
UUH O
W
O
xi
-------
focus of this Volume II study are labeled as (1) current rules,
(2) expected controls, and (3) additional controls.
The current rules analysis reflects the effect of existing
rules addressing volatile organic compounds (VOC) or particulate
matter (PM) (or specific toxics) emissions, whether they are
local, State, or Federal in nature. The expected controls
analysis assumes that candidate Federal rules currently under
consideration including NESHAP, NSPS, FMVCP, and Gasoline Reid
Vapor Pressure (RVP) limits will be implemented in each area.
The additional controls scenario added (1) control measures under
consideration by EPA as part of the Federal Implementation Plan
(FIP) for VOC, (2) the most stringent SIP level controls for
particulate matter within the five study areas, and (3) the most
stringent SIP control levels for a source category applied to all
VOC stationary source emitters within the category. For
particulate, if no category-specific regulations existed, a
default control level of 98 percent was used for point sources.
This 98 percent control level was selected by inspecting the
requirements for point source PM control in the five study areas
and choosing a control level representative of the maximum
available. (There are source types achieving greater than 98
percent control of PM, but for categories with no specific
regulations it seemed unrealistic to choose too high a value.)
Area sources emitting PM were controlled further only if category
specific information about controls was available. A vehicle
miles traveled (VMT) reduction of 5 percent was also assumed in
this scenario as a possible additional control measure.
MODELING ANALYSIS
All quantifiable control measures under the three scenarios
were evaluated for VOC and PM emissions reductions by 1995 using
the Regulatory Impact Model (RIM) developed by Radian
Corporation. The RIM starts with a 1980 emissions data base and
simulates how those emissions might be expected to change by
source category in 1995. A more detailed description of how RIM
simulates future emission changes and a summary of the three 1995
xii
-------
modeling scenarios included in this study are provided in Chapter
III. Analyses were also performed to compare the effect of the
control scenarios on the maximum individual risk from point
sources.
RESULTS
Figure 3 displays the projected reduction in excess cancer
cases per year per million population for the three control
scenarios and the source categories contributing the greatest
risk. Figure 4 contains the specific incidence reduction values
by source category. The results in these figures assume 100
percent rule effectiveness (i.e., the applicable regulations are
fully effective).
A review of Figures 3 and 4 indicates that a 27 percent
reduction in excess cancer cases is estimated between the base
year and 1995 under the current rules scenario. This reduction
comes essentially from the FMVCP and is due to projected
reductions in 1,3-butadiene, benzene, formaldehyde, and products
of incomplete combustion. There is an estimated net increase in
incidence from stationary source emissions even though emissions
from some source categories are reduced (e.g., wood stoves,
glass, and brick manufacturing). These emission decreases,
however, are offset by growth in emissions and associated
incidence from other source categories (e.g., cooling towers, use
of miscellaneous solvent cleaners).
Under the expected controls scenario, an additional 20
percent reduction in incidence from the base year is projected
resulting in an overall 47 percent reduction. This reduction
results primarily from chromium source control of industrial and
comfort cooling towers and chrome platers expected under the
NESHAP program and the Toxics Substances Control Act.
The additional rules scenario projects another 13 percent
incidence reduction from the base case to be possible if the most
stringent regulations currently in effect for chromium sources
and hospital sterilizers were applied in each urban area.
Xlll
-------
U
>— 4
X
o
H
O
-H
W «
o
ft!
H
Z
O
U
><
H
hH
u
.2 «
:-.-••!
!*>
X-^XX, - „ ,
^x,l>»jr*V
.JV:\^;^^'
•tl*,;^|W'c'
^L^A:&^
p?t
m
\-f /' ••
*^^
£
i,
•o
<
II
O. =
x o
<5
oe
I
CO
O
oo
1
o
«o
I
I
o
en
I
I
I
S3LLID S NI y VHA HHd S3SVD «HDNVD SSHDX31V1O1
XIV
-------
§§
88
ON
0£
s
UJ
00
CQ
o
CO
•
«—• ^H o :cN 10 CN
o d :O
-------
en
_
SccT
b
a J
r* •—•
I a
cd cd
a a
A -1
8
o
w fe
«4-( -*->
O cd
o
o
CO
C\J
O
to
CN
\
O
CN
I
I
I
I
I
T
oooooo
CNIOCO<0*<*-
-------
Figure 5 shows the results of the analysis to determine what
effect the control scenarios would have on maximum individual
risk from point sources. The measure chosen for demonstration
purposes was the number of point sources that would cause a
maximum individual risk above one in a million. From the base
case to the current rules scenario, there was a 27 percent
reduction in the number of sources above this cutoff level. The
expected controls and the additional controls scenarios resulted
in additional 1 and 32 percent reductions, respectively, for a
total reduction of 60 percent.
In summary, this study estimates that the area-wide cancer
incidence may be reduced by approximately half from the base case
as a result of currently planned controls. It is interesting to
note that this reduction is almost equally divided between
reductions in VOC and PM emissions and that approximately half of
the incidence reduction is due to mobile source control and the
other half to stationary source control. The study also
indicates that the control scenarios studied may significantly
reduce the maximum individual risk contributions from point
sources.
DATA BASE LIMITATIONS
While this study provides an initial insight into the
potential for mitigating urban problems, it is important to
remember the limitations of the available data base in terms of
both data accuracy and source category comprehensiveness. The
methods, data, and assumptions reflected in any study such as
this tend to change, sometimes rapidly, as one's understanding of
the urban air toxics problem evolves and matures. For example,
new sources and pollutants may be uncovered, and new assumptions
may be adopted concerning how exposure and risk characterizations
should be conducted. A fundamental problem with this type of
analysis is the adequacy of emissions inventory data for air
toxics. The past efforts in emissions inventory development work
have focused more on criteria pollutants and less on air toxics,
leaving a gap in the air toxics inventory. In addition, there
xvn
-------
are many pollutants and sources encompassed by air toxics, few
are covered by emission factors, and the emission factors that
are available may not be very accurate. The emissions inventory
data utilized in this study probably did not allow for
identification of all air toxics sources, particularly the
smaller ones. The full extent of the air toxics problem,
therefore, may not be reflected and the potential for control may
well be over- or underestimated.
USE OF RESULTS
Not all of the data and procedures used in this analysis
have been reviewed by the State and local air agencies whose
jurisdictions encompass the study cities. In many cases,
especially with small point and area sources, EPA and its support
contractor made their own emission estimates based on national
data and "top down" procedures. Due to these and other data
limitations, these results should not be associated with any
particular city. Moreover, because of the many assumptions and
limitations inherent in this type of assessment, the composite
results for all five cities probably provide a better
representation of the urban air toxics problem and its overall
controllability than the results for any single city or the
relative results of a particular control measure.
The study, therefore, should be considered as an analysis
whose results give an indication of the potential and direction
for an urban air toxics mitigation strategy. A factor to bear in
mind when reviewing this report is that some source categories
(e.g., small stationary sources) may not show up as significant
contributors to risk due to data base shortfalls. Nevertheless
this study is useful in giving some indication of the potential
for risk reduction in urban areas from anticipated programs.
xviii
-------
I OVERVIEW OF STUDY
A. PURPOSE OF STUDY
This Volume II study was performed to explore the overall
controllability of urban air toxics. It is hoped that this study
will provide some indication of the impact of some mitigation
measures that might be implemented by State or local agencies to
address the cancer risk portion of the urban air toxics problem.
Within current legislative authorities, ongoing EPA efforts will
contribute toward further reductions in air toxics emissions and
associated incidence. These efforts include New Source
Performance Standards (NSPS) for wood stoves and several volatile
organic compound (VOC) sources, the National Emission Standards
for Hazardous Air Pollutants (NESHAP) program, continued progress
under the Federal Motor Vehicle Control Program (FMVCP), Resource
Conservation and Recovery Act (RCRA) air emission limits for
Treatment, Storage and Disposal Facilities (TSDFs), development
of the upcoming national ozone strategy for reducing VOC
emissions, and promulgation of PM^0 ambient air quality
standards. In addition, continued implementation of existing
State and local plus Federal requirements will reduce emissions
which are likely to contribute to the current urban air toxics
problem.
This study provides a quantitative assessment of how air
toxics emissions and associated risk might change between the
base year and 1995 under current rules for a sample of compounds.
Analyses are also provided for possible additional measures
beyond what are currently required. The study was designed to
examine criteria pollutant and compound specific emission
reductions and associate with them potential reductions in
estimated cancer incidence and maximum individual risk.
This study, like most urban assessments to date, should be
considered a screening (or scoping) analysis, performed to yield
an order-of-magnitude estimate of the relative nature of the
urban cancer problem rather than to provide an absolute
prediction of incidence and individual risks. It is especially
-------
critical in this type of study to ensure that the scope and use
of the conclusions be kept compatible with one's confidence in
the underlying data base and analyses. Most studies to date of
this type are acknowledged to be screening or scoping studies
whose results should be used only in a relative sense for
providing broad program direction to suggest where more detailed
and focused follow-up work is needed.
Not all of the data and procedures used in this analysis
have been reviewed and approved by the States or local air
agencies whose jurisdictions encompass the study cities. In many
cases, especially with small point and area sources, EPA and its
support contractor made their own emission estimates based on
national data and "top down" procedures. For these reasons, no
results are associated with any particular city in this report.
Because of the many assumptions and limitations inherent in this
type of assessment, and because of different characteristics of
each of the five cities, the composite results for all five
cities may provide a better overall representation of the urban
air toxics problem in the United States than the results for any
single city.
B. NATURE OF URBAN AIR TOXICS PROBLEM
Volume I of this study showed that for the base year,
aggregate cancer incidence across the five cities in this study
averaged about 6 excess annual cases per million persons, ranging
from about 2 to 10 in individual cities. Area-wide individual
lifetime cancer risks averaged about 4 X 10~4, ranging from about
1.5 X 10~4 to 7 X 10~4. Note that these risks are not maximum
individual risks, which can be as high as 10~3 or even 10"2 at
specific receptor sites around some large point sources
(Haemisegger et al., 1985). Instead, these are individual risks
averaged over entire urban populations. Volume I of this study
did not attempt to estimate maximum individual risks. Results of
the limited maximum individual risk (MIR) modeling performed in
this study are presented in Chapter III of this report.
-------
Figures I.I and 1.2 show the major pollutants and sources
contributing to urban cancer incidence from air toxics in the
base year, based on average results from the five cities. Figure
1.2 shows that the major contributors to total aggregate
incidence tend to be small point and area sources and road
vehicles, the latter figuring importantly in most urban studies.
Not surprisingly, the pollutants of primary importance tend to be
those associated with these same source contributors. Total
cancer incidence associated with POM in Figure I.I is largely due
to diesel particulate (45 percent), gasoline particulate (32
percent), and wood smoke (17 percent), all of which are area
sources. Total cancer incidence associated with chromium is
predominantly due to hexavalent chromium (Cr+6) emitted from
industrial cooling towers (19 percent), comfort cooling towers
(28 percent), and chrome platers (51 percent). Cancer incidence
from benzene and 1,3-butadiene exposure is primarily due to road
vehicles.
Risk from formaldehyde exposures is attributable both to
secondary (or photochemically produced) formaldehyde and to
primary (or directly emitted) formaldehyde. This study suggests
that direct formaldehyde emissions account for about 40 percent
of the total formaldehyde-related cancer risk whereas secondary
formaldehyde accounts for about 60 percent. The primary VOC
sources contributing to secondary formaldehyde production are
road vehicles (35 percent), solvent use (29 percent), gasoline
marketing (8 percent), and refining (6 percent).
C. STUDY FOCUS
1. Emphasis on Cancer
This study estimates cancer risks from long-term (i.e.,
annual average) exposures to multiple pollutants. This type of
analysis has predominated urban air toxics assessments. To date,
limited work has been done to quantify noncancer risks associated
with short-term (i.e., acute and subchronic) exposures to air
toxics.
-------
*H fc
n
w
-------
I
-------
The most commonly used measure of cancer risk in urban
assessments is "aggregate cancer incidence," which is a measure
of the excess cancer cases over an entire urban area associated
with multi-source, multi-pollutant exposures to air toxics.
(This is also called "population risk.") Incidence is typically
expressed as the number of excess cancer cases expected in a
single year, but is estimated based on an assumed 70 year
"lifetime" exposure. In this study, incidence is considered
additive (covering multiple pollutants) and is
population-normalized by adjusting per million persons. As such,
the measure of cancer risk most commonly expressed herein is
"excess aggregate additive incidence per year per million
population." Normalization by population allows incidence from
one urban area to be compared with another.
2. Modeling Scenarios
Three primary modeling scenarios, or groups of control
measures, were developed for examination in this study. The
three scenarios include a "current rules" analysis, "expected
controls" analysis, and "additional controls" analysis. All
quantifiable control measures under the three scenarios were
evaluated for VOC and PM emissions reductions by 1995 using the
Regulatory Impact Model (RIM) developed by Radian Corporation.
The RIM starts with a 1980 emissions data base and simulates how
those emissions might be expected to change by source category in
1995. A more detailed description of how the RIM simulates
future emission changes is provided in Chapter II.
A summary of the three 1995 modeling scenarios included in
this study is provided in Table I.I. The current rules analysis
reflects the effect of existing rules affecting VOC or PM (or
specific toxic) emissions, whether they are local, State, or
Federal in nature. For stationary sources, current rules include
SIP regulations, NSPS, and NESHAP. Mobile source controls
include the effects of the FMVCP as well as existing inspection
and maintenance programs.
The expected controls analysis adds additional emissions
reductions to the current rules scenario. The additional
-------
10
Ox
OS
OT
O
cu cu
rH o
J3 CO
at
H MH
(0
co
OT
CU
rH
3
on
c
0)
I-l
IJ
3
O
OT
rH
O
W
C
O
rH
(0
O
•H
4-1
•H
*C
•o
W
r™t
o
I-l
c
0
"V
0)
^
cu
ex
p»^
n
iH
n
rH4
J*
^
CO
c
cu
M
M
3
0
£
(0
C
o
•H
4->
(0
C/}
OT
§
cu se
•w CU
(0
•0 CU M
•H MO
"O CO MH
4J OT
0 C rH
CU O
•0 bO I-l
C C *-
<0 -H C
h O
O *J 0
O OT
> rH
4-1 CU
I-i OT >
O O CU
««H 3C rH
Cu
£
CO
W
z
•w
c
(0
W
cu
CO
z
OT
CU
o
i-l
3
o
c
o
nH •>
4J CU
CU O rH £>
rH -H C p CO
O rH bOrH
I-l Q, MH CU fH
4-> Q, iH 14 BJ
C (0 >
O « 0 (0
O CU OT i-l
rH OT CU MH W
C (0 CU Cfl 0 E*
CU rH 4J rH CU
bO-H 4-> 3 0. C
e cc 1-1 a OT o
i-( > E fH 1 iH
M (0 CU <-•>»*-•
4-> M ti C0
OT OT rH CO O B
rH rH O. bO I-l
4-> CU (0 CU O
OT > l-t *J MH
O CU O O (0 C
S rH 4-> b CJ i-l
MH
O
rH
CU
I
rH
rH
0
G
O
CJ
4->
rH
3
MH
CU
(0
e
o
o.
i-i
o
*W
T3
CU
OT
3
OT
(0
>
^"?
CO
•
OT
CU
CJ
I-l
3
O
OT
CU (T)
OT
e
"O iH
CU
OS 0 U s-^
O Cu
OT O
OT (0
13 tj j^|»
rH bO y
o. o cu
cu CXMH
£a= «
Sr^5
OT
01 CU
rH O
iH S-l
-O 3
O O
Z CO
-------
reductions are attributable to candidate Federal rules currently
in process. For mobile sources, the expected rules case includes
gasoline Reid Vapor Pressure (RVP) limits.
The third, and most stringent, control scenario modeled is
labeled the additional controls scenario and includes expected
Federal Implementation Plan (FTP) control measures for VOC and
the most stringent SIP level controls for particulates and
control technique guidelines (CTGs) being considered as part of
the EPA post-1987 ozone policy. As shown in Table I.I, this
scenario applies the most stringent control levels available to
all stationary source emitters. For particulates, if no
category-specific regulation information was available, a default
control level of 98 percent was used for point sources. This 98
percent control level was selected by inspecting the requirements
for point source PM control in the five study areas and choosing
a control level representative of the maximum available. (There
are source types achieving greater than 98 percent control of PM,
but for categories with no specific regulations it seemed
unrealistic to choose too high a value.) PM emitting area
sources were controlled further only if category specific
information about controls was available. A vehicle miles
traveled (VMT) reduction of 5 percent was added to the motor
vehicle control measures modeled in the expected controls case to
simulate the maximum expected for these sources.
3. Caveats
The reader is cautioned when drawing conclusions from the
results of this report because they involve considerable
uncertainty. Estimated incidence can change by a large magnitude
and the ranking in importance of various sources and pollutants
can change depending on the assumptions made in the analysis —
and these assumptions will continue to change as new information
becomes available.
As an example, an initial modeling study conducted for the
same five cities produced base year incidence estimates that were
800 percent higher than the current results. This dramatic
decrease in incidence occurred despite adding a number of
8
-------
compounds to the study that were not analyzed previously and
changing the approach for estimating formaldehyde-related
incidence to account for secondary formation. Many of the unit
risk values have changed as new information has become available.
The other major change affecting analysis results has been the
method for estimating polycyclic organic matter (POM) related
incidence. The comparative potency approach used here produces
much lower POM incidence estimates than the previous approach of
using benzo(a)pyrene (B(a)P) as a surrogate for POM.
While this study provides an initial insight into the
potential for mitigating urban problems, it is important to
remember the limitations of the available data base in terms of
both data accuracy and source category comprehensiveness. The
methods, data, and assumptions reflected in any study such as
this tend to change, sometimes rapidly, as one's understanding of
the urban air toxics problem evolves and matures. For example,
new sources and pollutants may be uncovered, and new assumptions
may be adopted concerning how exposure and risk characterizations
should be conducted. A fundamental problem with this type of
analysis is the adequacy of emissions inventory data for air
toxics. The past efforts in emissions inventory development work
have focused more on criteria pollutants and less on air toxics,
leaving a gap in the air toxics inventory. In addition, there
are many pollutants and sources encompassed by air toxics, few
are covered by emission factors, and the emission factors that
are available may not be very accurate. The emissions inventory
data utilized in this study probably did not allow for
identification of all air toxics sources, particularly the
smaller ones. The full extent of the air toxics problem,
therefore, may not be reflected and the potential for control may
well be over- or underestimated.
Other specific caveats and modeling assumptions are listed
below.
(1) Personal exposure to air toxics was estimated using annual-
average concentration estimates and it was assumed that
exposures occur where people reside. In addition, only
outdoor exposures were modeled. Thus, this methodology
ignores people's movements throughout the urban area, travel
-------
outside the urban area, and indoor exposures. Because
exposures were simulated over a 70-year period, it is
unclear how much this restricted modeling methodology
affects the study results.
(2) The study relied solely on quantitative estimates of cancer
risk associated with inhalation of ambient air. Acute and
subchronic effects were not included, and cancer cases
associated with exposure routes other than inhalation of
ambient air were not quantified.
(3) Only a selected number of compounds were included in this
study, although monitoring studies have shown that urban
atmospheres typically contain many more pollutants. The
compounds selected for study were chosen because they were
estimated to be the most important contributors to excess
cancer incidence.
(4) Annual-average emission estimates were used to estimate
concentrations of air toxics. Thus, the study focused on
routine, continuous emissions. Accidental releases were not
modeled.
(5) Unit risk factors employed in this study represent the
chance of contracting cancer from a lifetime (70 years)
exposure to a given concentration of that pollutant. The
carcinogenic potency estimates used in this study were
developed by EPA's Carcinogen Assessment Group.
(6) Cancer incidence estimates are presented for existing
conditions (1980) and a 1995 projection year. These
incidence estimates are based on the assumption that
emission levels for each scenario remain constant for a 70-
year period. In reality, emissions will vary from year to
year.
(7) In assessing cancer risk within an urban area, each of the
compounds under study has been analyzed individually. Any
possible synergistic or antagonistic health effects of these
compounds have been ignored.
(8) Sources included in the exposure modeling data set for each
study area were limited to those in the counties under
study. Therefore, while contributions from these sources to
areas outside the county boundaries were considered,
contributions of sources located outside the county
boundaries to air toxic concentrations within the study
areas were not.
(9) Modeling results presented in this report for exposure to
gasoline vapors do not include self-service exposures at
service stations.
10
-------
(10) Except for secondary formaldehyde formation, atmospheric
transformation of toxic compounds has been ignored. Both
secondary formation and scavenging may occur for the
compounds included in this study. Thus, it is difficult to
quantify how neglecting transformation might affect the
final results.
(11) Incidence from secondary formaldehyde exposure was estimated
using ambient monitoring data and assuming that everyone
within an urban area is exposed to the same concentration.
This is a relatively crude technique especially since it has
been found that ambient ozone causes a negative interference
with the dinitrophenylhydrazine (DNPH) method for measuring
formaldehyde. Because so much of formaldehyde is formed
secondarily, however, this procedure was judged to be
preferable to modeling direct formaldehyde concentrations
and ignoring secondary formation.
(12) It has been suggested that background concentrations of some
toxic compounds, notably carbon tetrachloride, may be
contributing significantly to observed ambient readings. No
attempt has been made in the dispersion modeling performed
for this study to account for background concentration.
(13) While the comparative potency approach used to estimate POM
risks is judged to be an improvement over previous
particulate modeling techniques, which used B(a)P as a
surrogate for POM, the particle unit-risk estimates are
based on few measurements, especially for the important
motor vehicle categories, and the uncertainty in these
values should be recognized.
(14) Analyses of control measures for numerous VOC emitting
source categories were limited by whether categories were
covered in the emission inventories for the five cities
included in the analysis. Thus, if no web-offset
lithography sources appear in the emissions data base, then
regulations to control this source type will show no benefit
in the modeling analysis.
(15) Results of this study may be biased toward showing current
regulations, particularly NESHAPs, to be more effective in
reducing incidence than they may be in practice. This bias
might occur because regulated sources are those for which
the most information is available. If other sources turn
out to be just as important as the well known ones, then
these NESHAP regulations expected to be in place by 1995
will not achieve the percentage reductions in overall
incidence that are estimated in this study. The percentages
are affected because the problem is bigger than is
estimated. The actual benefit is the same, but would show
up as less of a percentage change if the total picture were
known.
11
-------
(16) The handling of some point sources as area sources for
modeling purposes may introduce some upward bias in the
resulting exposure/risk estimates since HEM distributes area
source emissions by population and since area source
emissions are released closer to ground level.
(17) Caution is urged when applying the study results to other
cities since the sources and pollutants may not be
representative of some other areas. For example, no study
area includes the heavy concentration of wood stoves that
characterize some northern cities; hence, the relative
importance of wood smoke may be understated in this report.
(18) Hazardous waste treatment, storage, and disposal facilities
(TSDF) emissions in the five areas selected for study here
are lower on a per capita basis than the national average
and, therefore, TSDF controls may be more effective in other
areas where the share of TSDF emissions is larger.
(19) Any study such as this represents a "snapshot in time" of
one's collective understanding of the urban air toxics
problem. In fact, the emission estimates and dose-response
relationships used in this study are subject to frequent
revision as newer data become available. Hence, care should
be taken when interpreting any results from this study or
comparing these results to those from other studies where
different data have been used.
4. Methodology
This section provides an overview of the study methods used
in this analysis including the areas of data base development,
modeling methodology, and constraint file development. Each of
these areas is discussed in greater detail in Chapter II.
a. Data Base
The data base used in this study represents conditions
generally existing between 1980 and 1985. The starting point of
this inventory was the 1980 National Acid Precipitation
Assessment Program (NAPAP) emission inventory and incorporated
updated information from local and state agencies as well as EPA
and other governmental studies. Several additional emission
factor modifications were made to the base inventory developed
for the Volume I study after examining a number of source
categories and identifying several that were lacking emission
factors in the Volume I inventory.
12
-------
b. Modeling
Once the data base for the base year was established,
inventories for 1995 for each of the three cases were developed.
The future year emission projections were made using RIM. Source
category growth rates and replacement rates were used as inputs
to the model along with a constraint file that simulates how
regulations will change new, replaced, and existing source
emissions. The output of the RIM model, the percentage changes
in VOC and PM emissions between 1980 and 1995 for each source,
was input to the Human Exposure Model (HEM). Other inputs to HEM
include unit risk factors for the pollutants of concern and
population and meteorological data. The HEM model produces
estimates of excess cancer incidences due to the toxic emissions
calculated by RIM. More importantly, the output from HEM was
used to show changes in incidence due to the three different
control cases.
c. Constraint Files
The three different control cases modeled for this study are
current rules case, expected controls, and additional controls.
A separate constraint file was used to model the regulations in
effect for each case. The constraint file for the current rules
case included regulations in existence, being implemented, or
under serious consideration for implementation as of 1985. The
constraint file for the expected controls case includes the
regulations included in the current rules case plus rules that
are expected to be established or come into effect by 1995. The
constraint file for the additional controls case includes the
constraints from the expected controls case with the addition of
constraints applying the most stringent control level available
for all sources. These constraint files were applied in the RIM
model to determine emissions in 1995 for each source category.
13
-------
II STUDY METHODS
A. DATA BASE
A base year inventory was previously compiled for each of
the five metropolitan study areas from a number of different
sources of information. This effort mainly upgraded those areas
of most importance reflecting newer source and emissions data.
The base year nominally represents the year 1980 but, in fact,
the data were not this well resolved temporally. Thus, the base
year should really be considered to represent conditions
generally existing between 1980 and 1985. Motor vehicle
emissions were estimated using 1980 emission factors, but
emission estimates for other categories were updated to current
conditions when better information was available than existed in
the 1980 emission inventory. These revisions included deleting
plants from the data base which were known to have shut down
after 1980. The first reliable1 ambient formaldehyde data, used
to estimate total formaldehyde-related risk, were not available
until 1987.
The base year inventory was compiled from a number of
sources. The starting point was the 1980 NAPAP emission
inventory (Wagner et al., 1986). This inventory was improved by
incorporation of the following:
. information from local surveys;
. comments from State and local air agencies;
. information and methodologies from EPA's Integrated
Environmental Management Projects (IEMP) "geographic"
studies;
. updated emission factors and emission estimates from EPA's
NESHAP program;
. updated emission factors and emission estimates from EPA's
Office of Mobile Sources (QMS);
l-The data are likely to have been underestimated due to
ozone interference with the DNPH method.
15
-------
. updated emission factors from EPA's emission factor
documents; and
. special contractor studies of area source activity levels
and emissions (hospital sterilizers, waste oil combustion,
dry cleaning, residential wood combustion, and wastewater
treatment).
Readers interested in a more detailed treatment of the emission
inventory compilation are referred to Volume I of this report
(Pechan, 1989). What follows is a discussion of how the toxics
data base has been augmented since Volume I was completed.
Using the Crosswalk data retrieval system (U.S. EPA, 1987b)
and the VOC Speciation Data System (Radian, 1989), each of the
source categories currently being considered for control by
Federal rule, CTG, or ACT was examined to determine whether any
of the toxic species of interest to this study could be emitted
by that category. This assessment led to three types of changes
or additions being made to the inventory. The first involved
adding toxic emission factors for point source SCCs for
lithography and marine vessel loading. The second change
consisted of adding a new area source SCC to the inventory to
include VOC and toxic emissions from hazardous waste TSDFs. The
final change consisted of adding five area sources (architectural
surface coating, traffic paints, autobody refinishing, industrial
maintenance coating, and miscellaneous industrial surface
coating) whose VOC emissions were already included in SCC
99999971, Surface Coating. This SCC was eliminated after being
broken down into five new area source SCCs to allow for different
toxic emission factors and control efficiencies for each of the
individual surface coating area sources. This breakdown is
illustrated in Figure II.1. The total VOC emissions in each
source category affected by these changes are shown in Table II.1
with the toxic pollutant emissions that were added to the
inventory listed in Tables II.2 through Table II.5. A more
detailed discussion of the changes made to the base case
inventory is included in Appendix A.
16
-------
CO
CD
JH
Q
o
-P
CO
fl
O
CO
CO
-------
I—*
M
M
fH
w
i-H
-ft
rt
H
^
U
0
bO
0)
rt
O
O
W
9
O
CO
J3
(0
iH
W
B
o
fH
W
•H
B
M
O
O
rH
rt
0
H
0)
C
fH
M
03
3C
uO
(U
E*
rH
A
j^
4->
CO
T3
C
M
•
O
CO
«e
rH
QJ
CO
01
(U
>
J^J
91
tn
UJ
O
01
a
c
a>
c
fH
rt)
X
o
(Q
<4H
g
CO
>>
O
ft
o
3
o
M-4
LM
(0
u
H
rH
«J
U
O
0>
fH
JS
u
u
bO
C
fH
"^
A
O
J
a
(0
bO
0
"5
*?4
j
Q
CO
H
C
4-1
O
o
c
0
o
bo
c
X!
W
•H
B
•H
*4H"
£
C
"rt
cu
B
•H
*t6
O
O
^^
>>
Q.
4J
\^
^^
£
4->
^
0,
^
,^N
>»
0,
^
^rx
g
. .
^v
^^
O.
1^^
>%
o«
4-1
\^ x
^.^
>.
o.
^
>.
4-*
fH
u
-*Q
*^t
<
*z>
in
o
CM
ON
CM
•*
CO
ON
O
vO
pn
—,
o
ON
pn
pn
in
m
pn
ON
o
CM
i
i
i
i <;
^x
•
CO
^>
O
•
CO
m
ON
en*
r^
f— 1
*
pn^
CM
CM
0
pn*
rH
PQ
^*
^Pt
<
cn
pn
ON
VO
O
pn
0
sr
o
rH
^H
pn
00
v— 1
rH
CO
CM
m
CM
0
m
o
^J
*p>^
o
•
£
sf
pn
f*^
VO
CM
CM
ON
•-<
O
m
§
pn
r*
^-
f-H
pn
rH
ON
pn
CM
VO
fs^
cn
ON
Q
^C
^p
o
•
m
o
o
o
pn
CM
CM
CM
m
m
m
pn
pn
VO
pn
m
VO
CM
-*
^j.
00
O
^
u
o
CO
in
o
•
o
CO
o
st"
ON
CO
°1
CO
vO
in
m
f*
en
pn
vO
rH
CO*
m
pn
iv!
en
rH
•*
VO
r^
<— <
r^"
vO
**
CM
Q
CM
pn
CO
en
_j
^^
n)
4-1
O
EH
rH bO
OJ C "^
OT -rl >,
to TD a.
CO
X > o
Q >>
CO Q,
H -M
pn co co pn o
CM en r~ r*» o
pn
pn
CM
CM
M
M
V
rH
(0
O
bO
a>
n>
(U
a
M
o
co
(V
to
o
CO
B
(U
4->
•O B O •
B fH U
00
>> c
O JS •
JO CO
•4- in i-t CM CM
C5 CM C> O O
pn
\o pn vn
in
X
Q.
O ON CM «* in
-H VO
CM
ON
(U
»*H B >»
>4H fH Q,
rt aj 4-1
SH OH >-<
H
rt
3 bo
O IH >»
H) 1-1 p.
4J rt 4--
fH O ^
J3 O
O
in rH rH CM CM
O rn CD O O
CM
•
sj-
ON m m co o
— CM O O r-l
m
vO
pa
w
i-H
R)
4-1
O
H
-------
Table II.3
Methylene Chloride Emissions Added by Source Category
City
A
B
C
D
E
Totals
Architectural
Coating
(tpy)
31.6
206.9
7.6
14.2
16.4
Industrial
Traffic Maintenance
Paint Coating
(tpy) (tpy)
8.1
52.8
1.9
3.6
4.2
6.5
42.4
1.6
2.9
3.4
Web
Offset
Lithography
(tpy)
NA
1,359.9
NA
NA
29.3
276.6
70.5
56.7
1,389.2
Table II.4
Perchloroethylene Emissions Added by Source Category
City
A
B
C
D
E
Totals
TSDF
(tpy)
0.1
7.5
0.7
0.7
0.0
9.0
Table II.5
Formaldehyde Emissions Added by Source Category
City
A
B
C
D
E
Totals
Web
Offset
Lithography
NA
12.6
NA
3.7
4.6
20.9
19
-------
B. REGULATORY IMPACT MODEL
1. Model Overview
All measures analyzed in this study were evaluated for VOC
and PM emissions reductions in 1995 using the RIM developed by
Radian Corporation. The RIM starts with a 1980 emissions data
base and simulates how those emissions are expected to change in
1995 by source category. Important variables in the projections
include new source growth rates, replacement rates, and a
constraint file which simulates how new, replaced, and existing
source emissions are expected to be affected by regulations. The
RIM also contains cost information for each control option, so
that both projected emissions and control costs are an output of
the model.
For estimating future year incidence, the RIM provides
percentage changes in emissions between 1980 and 1995 for each
source category for VOC and PM, and this information is used in
the HEM to estimate incidence changes. As mentioned previously,
in this study three scenarios were simulated using the RIM: a
current rules case, an expected controls case, and an additional
controls case. VOC and PM simulations were performed separately,
with constraint files for each pollutant.
The overall study methodology is summarized in Figure II.2.
This figure shows the relationship among the 1980 emissions data
base, the projections to 1995, and the exposure models. Of most
interest in this study is the application of the RIM, which is
shown schematically in Figure II.3.
The RIM emission data bases (separate data bases are used
for VOC and PM emissions) are organized by source category (SCC),
with aggregations of controlled and uncontrolled emissions for
each region or study area (these are also divided by large and
small sized sources). Uncontrolled emissions are an important
indicator of activity levels for a category since they eliminate
source-by-source variations in control effectiveness. These
uncontrolled emissions are calculated based on controlled
emissions and control efficiency on a source-by-source basis and
then are aggregated at the SCC level for input to RIM.
20
-------
o
o
o
Q
O
t
^
>-
ID
H-
00
-------
CD
cu
nd
o
o
-
O
-t->
cd
tiO
CU
V)
c
0
"55
in
"E
ID
TO
"o
•«,
"o"1
a!
^_4
o
0
ol
11
o '
c
f/1
u
c
o
JC
u
"c
i S.E
2. ^
> a- .52
. OJ « X
|z a: ui
1 c-* oi to
c
g
"u>
(0
E
UI
s
o:
1
1
0)
M I
5
.2 i
o 1
Q
|
-o
l_
•^^
c
o
o
5
-o
J)
~o
w
c
o
o
•o
c
o
>,
V)
c
g
"55
M
E
V
c
0
"01
u
•a
^
o
en
0)
"o
o
j
^
22
-------
The RIM creates year by year projections of uncontrolled
emissions based on growth and replacement rates. These emissions
are separated into emissions from new, replaced, and existing
sources. Constraints and control costs are specified for each of
these three types of emissions. New source emissions added in a
given year are estimated by applying annual growth rates to the
total emissions in the previous year. Growth is compounded
annually, increasing exponentially with time (or decreasing if
the growth rate is negative indicating a decrease in activity for
the source type). Replaced emissions in any year are calculated
by multiplying the replacement rate by the base year (1980)
uncontrolled emissions. Total replaced emissions from the base
year exhibit a linear increase. As an example, a replacement
rate of 3 percent per year results in 45 percent of the 1980
emissions being replaced by 1995. Since existing emissions are
calculated by simply subtracting the replaced emissions, 55
percent of the base year emissions would be labeled as existing
in the above example. Existing source emissions are those
emissions remaining which were in place in the base year. The
RIM keeps track of replaced emissions since they are often
subject to more stringent regulations than are existing source
emissions. New emissions coming on line after 1980 are not
subject to replacement since equipment life generally exceeds the
projection time period.
Uncontrolled emissions are projected on a year by year basis
since regulatory constraints become effective in different years.
For example, an NSPS starting in 1989 would not apply to new
emissions prior to 1989. Along with applying the regulatory
options found in the constraint files, the RIM also applies an
implicit existing level of control (ELOC) constraint. This is
applied to all future emissions since it is believed that within
a given source category, all new and replaced emissions will be
controlled at least to the level of existing sources. The ELOC
is calculated by RIM based on the uncontrolled and controlled
emissions. It is a measure of the average control level in place
in the base year. The ELOC is specific to each source category
23
-------
and study area. Therefore, an ELOC constraint of 90 percent in
Study Area B would not be applied to the same category in Study
Area C.
Application of ELOC as a constraint is an important concept
because of the potential effect overestimating base year control
efficiencies can have on estimating future year emissions and
incidence. First, it is necessary to understand the origin of
the controlled and uncontrolled emission estimates. The
controlled emissions in the National Emissions Data System (NEDS)
may be either (1) estimated based on the operating rate, emission
factor, and control efficiency, hereafter referred to as emission
factor estimates, or (2) estimated based on stack tests, material
balance measurements, or other methods, hereafter referred to as
measured estimates.
Assuming that all other inputs are correct, the effect of
overestimated control efficiencies varies depending on whether
the emissions are measured or emission factor estimates. In the
case of measured estimates, if controlled emissions are correct
and the control efficiency is overestimated, then uncontrolled
emissions will also be overestimated. With growth based on
uncontrolled emissions, projections of new and replaced emissions
will also be overestimated. With both the future year
uncontrolled emissions and future year control level
overestimated, the effect on controlled emissions, and thus
incidence, is uncertain.
Overestimated control efficiencies for cases where emission
factor estimates are used will have a different effect on future
emissions. Uncontrolled emissions will, in effect, be based on
the operating rate and emission factor and will not be affected
by the control efficiency. The controlled emissions in the base
year will be underestimated and ELOC will be overestimated„
Projections of new and replaced uncontrolled emissions will be
correct based on the uncontrolled emissions. Since ELOC is
overestimated, future year control levels will be overestimated
(given that ELOC is the effective constraint), reducing emissions
and thus incidence. If constraints of a higher degree of control
24
-------
than ELOC are applicable to a category, ELOC will have no effect
on the future year emissions and incidence for these cases.
After projections have been completed by the RIM, estimates
of the percentage changes from base year controlled emissions by
region and category are produced. These values are then used
with the HEM results for the base year to project future year
incidence.
2. Growth Rates
Growth in new source emissions was estimated for the 1980 to
1995 period using two-digit Standard Industrial Classification
(SIC) level of detail for each industry. Population was used to
estimate growth for some nonindustrial source types. Industry
growth factors are from the Bureau of Economic Analysis (BEA)
(U.S. Department of Commerce (DOC), 1981) and represent growth in
industry earnings. Population projections are from the same
source. Because Metropolitan Statistical Area (MSA) level
industry earnings projections are only available at the one-digit
SIC level, the method used to provide two-digit SIC growth rates
was to allocate state level two-digit growth estimates to each
MSA using that MSA's share of growth at the one-digit SIC level.
This is believed to be a reasonable compromise between simply
providing one-digit SIC growth rates for each MSA and assuming
state-wide growth rates are representative of any area within
that state. Table II.6 shows the growth factors by industry for
each study area. Note that the services sector growth factors
shown in Table II.6 are used to estimate growth in motor vehicle
travel. Information on the motor vehicle emission factors used
in estimating 1995 emissions is presented in Appendix B.
3. Sample Calculation
While Figure II.3 presented an overall picture of how the
RIM operates, it is also useful to work through a sample
calculation to show how a source category's emissions change
under different scenarios. For this sample calculation, changes
to Study Area B graphic arts VOC emissions (NEDS SCC = 40500599)
are shown (1) for the 1995 base case scenario and (2) after
expected controls are applied.
25
-------
Id
a
•o
3
to <
«—i ~* o •—' •—i I-H t—1«—
4J
co
m «—« 4-4 i—i «a/ o
2 . . . . ^ 2 2 *^ 2 2
I T3
O 3
00 "
CT> CO
OT «-•
U U
O l-i
w •<
TJ (d
3 0)
*-• U
C/) <
•O (TJ
3 l-i
CO <
04
in oo co •-•
2 • • 2 • • • 2
C>1 *-^ **^ •— 4 O *~~* ^^
^^ *~* *~H CN *— < ^
*— ^ CM CM *~H CM ^-4 «— t
—« O
'Z.'Z.
00-4*00
_l _4 r-1 _4,_l_4«CMr-ICM~HCMr-«CM~H<—
o
bO
CJ
C
o
n)
r—I
3
CL
O
O-,
bO
0) C bO
l-l -rH C
3 C -rH
4-> -|H C
3 3C
O rH
•rH (d rH
tl 4-1 (d
bO % rt
l-l CJ
co u
•rH CU
S CO
00
CJ
O
a
co
I— CM
O —
r-
i
3
O
co
26
-------
The 1980 NAPAP Emissions Inventory point source file
indicates that there are 1,585 tons of VOC emitted by sources
with an SCC of 40500599 in Study Area B. These emissions and the
VOC control efficiencies for each source with an SCC of 40500599
are used to calculate a weighted average ELOC for this SCC and
region. For this particular example, uncontrolled VOC emissions
were estimated to be 6,191 tons, which gives an existing level of
control of 74.4 percent.
The growth rate for SIC 27, the printing industry, is used
to estimate new source growth for graphic arts. The growth
factor of 1.52 for Study Area B translates into an annual
compounded rate of 2.83 percent. The estimated retirement rate
for this industry is 4.6 percent.
The RIM constraint file shows two regulations which affect
future year VOC emissions for graphic arts facilities in Study
Area B. One is an NSPS that begins in 1986 and calls for 75
percent control of VOC emissions. A 100 percent penetration
factor means that all new and replaced sources in the category
are assumed to be affected by the NSPS. For this particular
category, though, the SIP constraint is more stringent than the
NSPS because it reguires 85 percent control of existing, new, and
replaced source emissions. Again, the penetration factor is 100
percent (applies to all sources regardless of size). A 1983
target year is listed for this regulation, but for a SIP
regulation, the year of implementation is unimportant as long as
it is prior to 1995, because all sources (existing and new) are
affected.
As Table II.7 shows, for the 1995 base case scenario,
controlled emission calculations are straightforward because all
three categories of sources (existing, new, and replaced) have
the same 85 percent control level. Calculation of 1995
uncontrolled VOC emissions is more complicated.
Shares of existing, new, and replaced emissions are
estimated using the following three equations.
27
-------
kl
re
Q)
^4
c
o
•H
o
cu
"o1
u
a,
o m
1-1 ON
*•• ON
(Q ^H
,4
3 rH
O CU
re o
o x
CU 4-1
f">. rH O
• a. re
M B a.
M « s
CO M
cu
•o
CU rH
4J 0
U U rH
CU •*-* CU
a. c >
X 0 CU
Cd CJ J
•a o
CU rH i-t
•<-> O M
o 1-1 re
a> *-> e
0. C CU
X O 0
w o to
4-> rH
c o
CU CO H rH
tl CU +* CU
U rH C >
3 3 O CU
C_) OS O J
CO
rH
3 O
OS -rl
•^ (0
ss
1 _l y
U C/)
e_>
•o
rH
rH
O
Li U
4-« O
c >
o
o
n?
rH CU >,
^0 0 U rH
rt tw
CO CU
co
re
CQ
o
CO
ON
0
U rH
G >
O CU
0 J
"O
CU
rH
rH
0 O
tj C5
V >
c
o
CJ
•o
CU
rH
O
W 0
*J O
C >
0
U
C
cu
a.
H
CU
o
U
3
O
co
^*9 £,j ^e
m m m
ON ON ON
VO [ ^f
ON VO rH
«— i OJ
)*•?
in m m
CO CO CO
co m rn
co oo sr
CM sr so
ON CO CM
««4 t-H PS.
ON CM CM
rH CO ST
^^
^f
• i i
sr i i
p^.
m
CO
m i i
•• i i
tf«4
•^H
ON
•-« i i
— i t
SO
bO -O
C CU
iH 0
*-* (T)
CO r-l
•rl > Q.
X CU CU
pq Z PS
ON
O
ON
m
00
m
ON
re
4-1
O
re
u
cu
o.
co
o
U
(0
0)
o
iH
CO
CO
•H
E
41
o
0)
4-f
o
z
03
re
cu
U
3
CO
c
•rl
CO
o
-rl
CO
CO
-rl
e
cu
o
I
co
Q.
cd
ON
ON
m
o
o
m
o
o
o
10
CO
o
03
rH
3
a
rH
re
U
cu
i-H
a.
6
tO
CO
CO
••H
28
-------
New 1995 = Existing 1980 * (1 + g)n - Existing 1980
Replaced ^.995 = Existing igso * (n * r)
Existing i995 = Existing igso ~ Replaced i99s
Where: g = new source growth rate
r = retirement rate
n = number of years between the base year and
the projection year
Applying these eguations to the graphic arts SCC for Study
Area B to estimate 1995 uncontrolled VOC emissions is shown
below.
New 1995 = 6,191 * (1.0283)15 - 6,191 = 3,218 tons
Replaced 1995 = 6,191 * (15 * 0.046) = 4,272 tons
Existing i995 = 6,191 - 4,272 = 1,919 tons
Controlled 1995 VOC emissions under the base case scenario
are calculated as l minus the control efficiency (85 percent)
times uncontrolled 1995 emissions. As shown in Table II.7, this
total is 1,412 tons. This 1995 emission total for the base case
scenario is then compared with the 1980 base year value for this
SCC of 1,585 tons and the percentage change from 1980 to 1995 is
estimated to be -10.9 percent. This percentage change is then
used to estimate the change in emissions between 1980 and 1995
for all organic toxic compounds emitted by SCC 40500599 in Study
Area B.
For the expected controls case, all the steps are the same
with the exception that all new, replaced, and existing sources
are assumed to be controlled by 95 percent instead of the 85
percent control assumed in the base case. This lowers the VOC
emission estimate for the category to 471 tons and changes the
percentage difference between 1980 and 1995 organic emissions to
-70.3.
Note that while this sample calculation was performed for a
relatively simple case, the calculations are more complex for
some categories, e.g., where new and existing source regulations
29
-------
differ and where some sources are exempted from a regulation and
these exemptions are simulated via penetration factors. Note
also that for some source categories the existing level of
control is higher than the highest level of control in the
constraint file and thus will be binding on future year emissions
rather than the constraint file controls.
C. CONSTRAINT FILES
1. Overview
This section provides an overview of the VOC and PM
constraint files used to project future year emission changes.
Separate constraint files are used for VOC and PM projections.
The three scenarios examined include current rules, expected
controls, and additional controls. Listings of the constraint
files are in Appendix C.
a. Current Rules Case
The constraint file represents regulations in existence,
being implemented, or under serious consideration for
implementation as of 1985. Existing Federal and State
requirements considered included SIPs, NESHAPs, NSPS, New Source
Review (NSR), Best Available Control Technology (BACT), and
Lowest Achievable Emission Rate (LAER). The principal
information sources included the Federal Register. CTG documents,
SIPs, and the Bureau of National Affairs Environmental Reporter.
In addition, State officials within each study area and EPA staff
having responsibility for pertinent criteria pollutant programs
were contacted by Radian to identify control programs in the
implementation process or under serious consideration for
implementation. The SIP provisions for characteristics of
specified sources to be controlled, NSR program policy,
Prevention of Significant Deterioration (PSD) program policy, and
local regulatory control plans were reviewed to identify any more
stringent regulatory measures.
Review of earlier developed constraint files revealed that
the expected reductions for gasoline marketing/vehicle refueling
were likely overestimated. The expected level of control for
30
-------
Study Area B was changed from 95 percent to 86 percent, which is
the current EPA estimate of in-use efficiency of stage II
controls with an annual enforcement program. The expected level
of control for Study Areas A, D, and E was changed from 95
percent to 40 percent reflecting the likely effects of stage II
controls with source size exemptions and minimal enforcement by
1995 for these regions.
In the current rules scenario, motor vehicle emissions were
assumed to be controlled by both the FMVCP and inspection and
maintenance (I/M) programs since all of the study areas currently
have some form of I/M program in effect.
b. Expected Controls Scenario
The expected controls scenario represents new NESHAP and
NSPS expected to be in effect by 1995 based on the EPA's expected
schedule for currently considered air toxics regulations added to
the current rule constraints. Table II.8 lists these NESHAP and
NSPS constraints. All of the constraints for this scenario were
given beginning and target years of 1992.
c. Additional Controls Scenario
The additional controls scenario is defined as the best
control performance utilized for any source within the category
and therefore includes the expected control scenario constraint
file plus the most stringent control level available for all
sources. Since no FIP study was available for PM emissions,
reasonably achievable rules for PM were defined as the maximum
constraint level for Study Area B from the current rules case.
The constraint levels from Study Area B were applied to the same
source types (new, replacement, or existing) as designated in the
original constraint. Other measures for PM (listed in Table
II.9) include residential oil combustion controls (correct tuning
and operation of the furnace with repair or replacement of poor
performers), fugitive dust controls, wood stove controls, and an
incinerator NSPS. These controls were applied to new,
replacement, and existing source emissions. Additional control
levels were also taken from four sources for the PM constraint
file: the current rule constraint file, the PM cost file, the
31
-------
Table II.8
NESHAP and NSPS Constraints
Expected Controls Scenario
NESHAP Reduction (%) Start Date
VOC: Hazardous Organic NESHAP 50 1993
Butadiene Production
Ethylene Dichloride Production
Chlorinated Hydrocarbon Production
Neoprene Production
Chlorinated HC Use in Chemical Production
Pesticides Production
Pharmaceutical Production
Coke Oven Emissions 50 1992
Charging and Topside Leaks
Coke Oven By-product Plants 63 1989
Municipal Waste Combustion 80 1991
Sewage Sludge Incineration 80 1991
PM: Comfort Cooling Towers 100 1989
Industrial Cooling Towers 67 1993
Chrome Electroplating 95 1993
Coke Oven Emissions 56 1992
Municipal Waste Combustion 99 1991
Sewage Sludge Incineration 99 1991
NSPS
VOC: Hospital Waste Incineration 80 1993
PM: Hospital Waste Incineration 99 1993
32
-------
Table II.9
Miscellaneous Measures Added to the Additional
Controls Scenario
Emission
VOC: Reduction f%)
Wood Stoves 40
Perchloroethylene Dry Cleaning 95
Petroleum Solvent Dry Cleaning 72
PM:
Wood Stoves 40
Incineration (NSPS) 99
Residential Oil Combustion 24
Fugitive Dust 50
33
-------
SIP strategy measures, and potential air toxics regulations. The
SIP strategy and air toxic measures were applied to all three
source types, while BACT levels from the constraint and cost
files were applied as specified in the files. The SIP strategy
and air toxic measures included in the PM additional controls
scenario are listed in Table 11.10. For all sources, the maximum
control level from any of the sources was chosen as BACT with the
exception of SCCs using the default PM control values. Since
these control levels were not derived from information specific
to any of the sources, the lowest control level, 98 percent, was
used for point source SCCs. Area source SCCs were controlled
only if control information specific to that category was
available. The beginning and target years for the constraints
were specified as 1992.
For the VOC constraint file, reasonably achievable rule
control levels were taken from a number of different sources.
One source was EPA's FIP Study (U.S. EPA, 1987a). The control
levels were applied to new, replacement, and existing source
emissions in all study areas. The FIP study identified the
expected SIP level of control for NEDS Source Classification
Codes (SCCs) emitting over 1,000 tons of VOC per year nationally.
Candidate Federal and CTG measures added for this scenario are
listed in Table 11.11. Miscellaneous measures for VOC based on
guidance from EPA are listed in Table II.9. The additional
control levels for VOC were also taken from the VOC current rules
constraint file, the VOC control cost file, candidates for
available control technologies (ACTs), SIP strategy measures, and
additional air toxics regulations being considered. The BACT
constraints taken from the current rule constraint file or the
VOC control cost file were applied to the source types as
designated. Candidates for ACTs and air toxics regulations added
to the VOC additional controls scenario are listed in Table
11.12. All VOC additional control constraints were designated
with beginning and target years of 1992.
34
-------
Table 11.10
Measures Added to the PM Additional Controls
Constraint File
SIP Strategy Measures Reduction (%)
Ceramic Clay Manufacture 99
Feed and Grain Terminals 95
and Country Elevators
Air Toxic Regulations
Chromium Electroplating 99.8
Industrial Cooling Towers 85
35
-------
Table 11.11
Candidates for Federal Rules and CTGs
VOC Additional Controls Case
VOC Emission
Measure Reduction (%)
TSDF 93
Commercial and Consumer Solvents 20
Marine Vessel Loading 90
Architectural Coating 50
Industrial Maintenance Coating 65
Traffic Paint 80
SOCMI Distillation 85
POTW (Industrial Wastewater) 75
Autobody Refinishing 60
Petroleum Wastewater 50
Web Offset Lithography 80
SOCMI Reactor Processes 85
36
-------
Table 11.12
Candidates for ACT and Air Toxic Regulations
VOC Additional Controls Scenario
Assumed Potential
Available Control Technologies Reduction f%)
Cleanup Solvents 50
Adhesives 65
Ink Manufacture 35
Paint Manufacture 35
Pesticide Application 50
NESHAPs
Hospital Sterilizers (Ethylene Oxide) 99
Pulp Manufacturing (Chloroform) 92
Ethylene Dichloride Production 94
(more stringent control)
37
-------
2. Quality Assurance
Quality assurance checks were carried out on both the
emission inventory data taken from the NEDS file to prepare the
base year emission inventory and on the assumptions used in the
development of the modeling assumptions (i.e., the emission
constraint file).
With only a limited number of areas included in this study
(five cities), checks could be carried out on a source by source
basis for a substantial number of sources. Thus, many sources
were reviewed individually to determine the logical and
engineering consistency of the data. The source records were
reviewed to determine the following:
. If the record was internally consistent as to size,
temperature, flow rate, etc. Engineering judgment was used
to determine if the device/source combination was
reasonable.
. If the control device listed was appropriate to the source
and is able to control a given pollutant at the efficiency
listed. It was not possible to determine the actual
efficiencies of the control devices applied, but rather if
the reductions were possible considering the application.
For example, while a fabric filter is capable of achieving a
99.9 percent reduction in total particulates from a coal-
fired boiler, it would be unlikely for a cyclone to reach
that efficiency.
All controlled VOC point sources in Study Area C were
reviewed individually, especially the VOC control efficiencies.
While a number of problems were found with the VOC emission
estimates and operating rates, because VOC control efficiencies
were few in number and appeared to be reasonable, no changes were
deemed to be warranted for this analysis. A similar analysis for
PM emitters in Study Area C showed that source category/control
equipment/control efficiency combinations were reasonable and did
not need to be changed. The NEDS emission inventory also shows
that the estimated total suspended particulates (TSP) emissions
are almost always less than 100 tons per year at the source
level, so changes to the control efficiencies for any individual
source are unlikely to have a significant effect on toxic
emissions for Study Area C.
38
-------
3. Model Assumption Review
The key modeling assumptions that distinguish the different
modeling cases relate to the degree of emission reduction
required. For each strategy or scenario, all of the applicable
regulations possible for each source category SCC were reviewed.
An example is the 1995 current rules case where SIP regulations
are applied to existing sources. For each SCC, all possible SIP
regulations — such as concentration limits, process weight
limits, or opacity limits as well as source category specific
regulations — were reviewed to determine the most restrictive
regulation. Average size sources in each SCC were used to
determine the allowed emission rates under regulations which
varied by size.
Applicable RACT, NSPS, BACT, and LAER regulations were
handled in a slightly different manner. For these requirements,
the average percentage reductions specified in EPA support
documents were applied.
a. Particulates
Resulting changes to the PM constraint file included
increasing the current rule control percentage for a number of
industrial sources in Study Area B that are controlled by a
process weight curve. The RIM constraint file had a number of PM
emitters controlled to 80 percent in 1995, when higher control
percentages are more reflective of current limits. Changes to
the PM constraint file for Study Area D were largely to include
constraints for source types known to exist in Study Area D that
had no constraints listed previously.
Changes were also made to the PM constraint assumptions for
cooling towers. Previously, comfort cooling towers and
industrial cooling towers were combined in a single category.
For this analysis, though, comfort and industrial cooling towers
were separated in order to make control assignments more
straightforward. For comfort cooling towers, a control
percentage of 100 percent was assumed in the expected controls
scenario reflecting the anticipated chromium ban enacted under
the Toxic Substances Control Act (TSCA). For industrial cooling
39
-------
towers, a new NESHAP was assumed to reduce chromium emissions by
67 percent (expected rule). A higher control percentage of 85
percent was applied to industrial cooling towers in the
additional controls case.
For chrome platers, an assumed zero control efficiency in
the base year was changed to 70 percent to reflect the current
mix of controlled and uncontrolled platers. An expected chrome
plating NESHAP is assumed to raise the average control level on
chrome plating shops to 97 percent from uncontrolled levels.
Reductions up to a control level of 99.8 percent are assumed to
be achievable in the additional controls case.
The EPA's promulgated wood stove NSPS will reduce
particulate emissions from new wood stoves by about 75 percent
from uncontrolled levels. An overall 40 percent reduction in
particulate levels is assumed to be achievable in the expected
controls case. This 40 percent reduction reflects a combination
of measures such as burning bans, no burn days, and accelerated
replacement programs.
b. Orqanics
In general, changes were made to the VOC constraint file to
modify CTG/RACT reguirements based on current information and to
estimate the effect of general VOC control regulations. An
example of a general VOC control regulation is the one applied to
Study Area D. It applies to installations not affected by
specific VOC control regulations. For instance, installations
constructed before May 12, 1972, that emit more than 200 pounds
per day of organics must reduce their emissions by 85 percent or
more. Installations constructed on or after May 12, 1972, must
reduce their VOC emissions by 85 percent or more if they emit
more than 20 pounds per day. Changes were made to the RIM
constraint file for VOC to reflect these general regulations.
While many of the changes to the current rule constraint
file for PM were to increase the estimated future control levels,
changes to the VOC constraint files were largely in the opposite
direction (future control levels were adjusted downward). These
changes to the VOC constraint file are summarized in Table 11.13.
40
-------
Table 11.13
Changes to RIM VOC Constraint File
Source Category
Dry Cleaning-
Perchloroethylene
Revised Constraint
File Assumption
RACT = 63%
NSPS = 72%
Expected =95%
Rationale
Rhode Island
State regulations.
Hospital Sterilizers Max Control = 99%
Petroleum Refinery
Wastewater
Surface Coating
NSPS
SIP = 85%
Degreasing
RACT =54%
NSPS = 57%
Ethylene oxide control
(vent and drain
sources).
Begins in 1989.
From city-specific
regulations and
general regulations
assumed to control
coaters in other study
areas.
NSPS for cold
cleaners.
Graphic Arts
RACT =70%
(75% for
rotogravure)
Added and used to
replace previous SIP
listings. NSPS
dropped for
1ithographic.
41
-------
Ill RESULTS
A. BASE YEAR SCENARIOS
Two base year scenarios were analyzed before projections to
1995 were made. These two base year scenarios included (1) base
year emissions estimated using NEDS control efficiencies, and (2)
base year emissions with SIP control levels substituted for
control efficiencies reported in NEDS (in instances when SIP
requirements were less stringent than NEDS control levels).
The second scenario was postulated because of concern that
the VOC control efficiencies in the 1980 NAPAP Emissions
Inventory overestimated the amount of control at VOC sources.
This was of interest because it appeared that some controls in
NEDS were at higher efficiencies than would be expected given the
regulations in force in 1980.
When base year VOC and PM emissions were reestimated by
substituting estimated SIP control levels for NEDS control
efficiencies in instances when SIP control requirements were
lower, there was little change in base year VOC emissions (1,2
percent). The average VOC control efficiency changes very little
across the five study cities when SIP control levels are
substituted, as illustrated in Table III.l. Changes in related
annual incidence from one case to the next are negligible. Table
III.2 shows the incidence estimates with formaldehyde included.
Table III.2 estimated incidence is lower for area sources in the
"with SIP control" level assumptions case because its percentage
of total VOC emissions is lower, and this percentage is the basis
for estimating contributions to formaldehyde-related incidence.
A 10 percent increase in particulate emissions was observed
when PM sources were assumed to be controlled only to SIP levels.
Problems with particulate control levels in NEDS seem less common
than for VOC control levels, however, as PM sources are more well
defined than VOC sources and less likely to be exceeding
prescribed limits.
With the above analysis showing only a negligible change in
VOC emissions with SIP control efficiencies substituted for NEDS
43
-------
O>
rt
^x
•o
^^
r;
Ol
•o
rH
rt
E
i-i
O
Cu
bO
C
•i-i
-o
3 CO
rH C
X iH
Cd *••
S.X Cl*
E
Ol 3
O CO
C W
Ol <
•a
i-s i-H
O Ol
c >
M O)
_J
T3
0) rH
4J O
rt ij
rH *-<
ni r*
U/ k»
OS O
U
•a
C Cu
to M
co
co
C •
o to
1-1 ^
CO
W >,
"E c
Cd I)
•rH
C_> 0
O iH
> M-l
i_i cd
rt
!>H rt
»^^ W
i-l
QJ *-.
co c
rt o
CQ O
Ol O
U Cd
< Z
>, I-i
•o 01
3 -0
>-> C
CO 3
rH
rH
<
U-J
O
C
o
co
•rH
Jj
rt
a.
fr
0
O
to
rH
>
Ol
rH
O
j_|
4_j
c
o
u
a,
M
CO
r^
4-1
•H
>
CO
ai
•i-i
o
c
cv
•rt
o
f-l
VM
VM
Ed
i— I
0
u
4-1
C
o
o
co
a
Cd
z
J=
4->
•r-l
&
r— 1
rt
3 01
c o
c c
•o
(11 r I
w U
•" C
rt rH
ai
Oi
Q) rH
bO O rH
03 U QJ
1-4 ^— * ^
01 C 0>
> 0 J
<: o
CO
C »-X
O CO
•* c
co o
co *-•
•r*
SJ-*l
v^
Cd O
O
O •—
O *•— '
>
rH
rt
3 a*
c o
c c
< o»
•o
T3 T-*
fll r)
W V
rt M
rH
O>
p£
O* rH
bO O rH
rt ^ QJ
i-l *-* ^
0) C OJ
> O J
•< o
to
c ^
o co
i-l G
co o
Wi i
*•*
iH
e o
Cd O
o
C_5 •— «
O *— '
O O r^ co O >o ^T
O O O MO O -* o O
CM CM (^
«* CO CT» in CM vO O
O O O CO •* —i r*
^-* •— 4 CO *^
CM — ' r^
O O m co o vo -^-
O O O CT\ O ^* ^3"
o o -* o o o —
5s?
f^. ^> o^ ""^ "^ ^D ^3
CM CM t^
*^ r*^ o^ vO CM *>o co
O o in o> ^ ^^ m
r~* CO f^ CO
CM «— < r-»
*-N
Ol
rH
•rH
JO
O
E
bo
rH C
C C rt -r<
O O CO CO TD
•K iH -rH CO O 3
(0 *•> *J CU O. rH
01 co co o co u
O 3 3 O -rH C
1-1 -Q JO U O -rH
3 E e a, ^
O O O C 01
CO O O rH O •" CO
rt -rH CO Ol
>M— 1 rH -rH 4J rt O
'~t rt rt jj rt £x >-"
rt c c *•• ij 3
ci-iucoo-ocflo
O Ol O) 3 O.-rH5ccO
•rH 4-1 4-1 -O « rH E-l
*->xcc>oort
rtCdMrHCdCOOi QJ
c
o
•rH
4-1
rt
N
CO C
o rt
Kn
• bo
CO O
E
QJ
to
Ss
CO
rt
i-t
rt
o
CO
c
o
•rH
co
CO
•rH
E
CM Cd
sJ
o rt
— C
—> O
TH
4-f
rt
z
to
•*
^^
CU
Cd
C
•rH
CO
co co
• (11
* W
n o*
n >%
4-1
0)
a
jj
3
0
to
i-l
0
rt
E
Ol
4-1
01
^^ W
- fft
• TO
CO
CO ~O
_ ai
r-* 4-J
to
IH
^^j
r-H
CO
0)
1-1
tj
o
bO
ai
4-1
rt
0
o>
0
Jj
3
O
CO
>^
Lj
rt
c
o
1-1
4-*
rt
4-*
CO
*
co
44
-------
CO
<—I
O>
>
01
3 01
c a
O" (J
*-> c
as M
I— I
a>
02
r^
O
o
Li
4-1
C
91 ,0
CJ t>
g a.
•S rl
•5 CO W
r!J ^
13 -'
•0 Q. &
01 £
*-• 3
co co
rH CO
. **
flj r* M\ CN CJ rO rr* M^ ^«
If C 41 r\jc'sCOC
> O nj CM IT)
< O
CO
C x-v
O CO
•
**"S
EO ^W^lOcMvo
w§ oddco's;^
f
"X3 Ol
C >
CO Ol
, 1
CO
Cr_|
• ^
0 0
••-1 Li „
co 4^ w
•2§ "
1 rH
CO
2 *•
s a
•s5 I *!
CM a, <" ^t 7? co ,.« — _
M^K «
S° G
hS . *M 1
U_ i I
« S2 a
rl Ol 1
v^ •»-( _ . -^ • • u | CD CSI
0;y ^OvDcoQ^g
COM °OcNl>3-dc>cr)
'oi ^
K
-° >« Ss ^ 1
CO o 0
£-. oi C U
« S «
co -i-i 51
ra.^ o
CO m ml
oi IM w 1
L, IS Q
< w a
rH Z
flj J^ «
**/ ^-f
bO 0 rH
™ Li Ol j^
5! g| ^0SJ^^oo
< CJ
>. 0 _ I CO
^ ^ "5 C x-x
3 4-> 1
aj s
•~< to
•<« Q 1
o co
•* c
co o -. _
C04-. ^^^^CMVOCO
Sg 2°'gos--:,n
W ' O '^ •— ' i^'
u-i ^ r , v-^ r^
r*. ^ C_? t— 1
° d -^s
L. > ^
CO)
0 T3
CO C
"*""* 3
Li "^
CO 0)
Q. rH
E •£
0 JQ
O 0
E
rH "0
C C ^ C
o o co S •*
••-1 --H co S "^
co *-> 4^ oi S 3
01 co co cj S" ^
0 = 3 C3 j? w
Li J3 o < ^^ C
3-f-g^ Q -5
sss.-jgs :
^-HrH^i; 2 ?
ggg^g^ g
IS2SS35I
^ x c "c ? 7^ E-
COWMMM£? «
4W ^* "4 CO Q< 0>
CO Li
bo o
(tJ 4-f
4-»
c co
> O 01 c
0 o
Li ^-,
Ol t~>
Q. 3
JO
CO --t
•-* Li
iH 4w
01 0
CO CJ
OS i ^ g g,
NV 1 " fll „_,
z \ z ^2
' ,T « B
CO IH
CO 4->
o co
Ol
CO
rH Ll
0 0
Ll 14-1
4-1
c co
O -rt
O CO
CO
| CLi _O
1 vo rH
« co«
o _ -e
1 «n := «
••-I CO
^ v-t
Ol Ol
JS bo
*-• co
4-»
C C
•M oi
CJ
CO Li
01 Ol
w ex
Li
3 CO
O "rH
CO JS
4-J
CO
1 •* £"2
1 . Li r;
CO CO «
oo
— w Li -
~ 0 L, .
l*-l Ol Ol
S CJ
m ° c
^ rH 01
? "U
O M — i
rH -r-l CJ
c
CO CO •-<
•*« c
„ 0 T3
01 n oi
CJ CO 4-.
C CO to
-------
control efficiencies, it was decided to continue to use NEDS
control efficiencies as reflective of base year conditions for
modeling all the future year scenarios.
B. FUTURE YEAR SCENARIOS
This section presents the results of three 1995 control
scenarios: current rules, expected controls, and additional
controls. The control measures included in each of these cases
were described in Chapter II.
1. Excess Cancer Incidence
Figure III.l summarizes total annual excess cancer incidence
by scenario for the five study areas combined. As this figure
shows, incidence would be expected to drop from 93 cases in the
base year to 67 cases under the current rules case, a 28 percent
decline. Reductions of up to 47 percent from the base case might
be observed if all expected controls were adopted. With the
maximum anticipated controls applied, incidence reductions of as
much as 60 percent might be realized.
a. Base Case
Tables III.3 and III.4 show the source categories that
contribute most to 1980 expected incidence for VOC and PM,
respectively. The motor vehicle categories are prominent in both
VOC and PM incidence contributions. With the exception of wood
stoves, none of the nonmotor vehicle categories appear on both
Tables III.3 and III.4.
One observation that can be made from these tables and
Figure III.l is that VOC-related and PM-related incidence are
about equal in importance, with organic compounds contributing
slightly more than one-half (55 percent) of the base year
incidence. Estimated incidence in the 1995 current rules
scenario is split evenly between organics and particulates, as
current rules prove to be more effective in reducing organic
toxics than they are in reducing particulates. The converse is
true in the expected controls scenario as particulate related
incidence drops dramatically, while VOC related incidence changes
little. Some further reductions in incidence are observed in the
46
-------
JH
cd
O
(U
"o
a
V
T*
•i-t
O
d
a)
i— i
V
(H
O
en
03
•
CO
r-
(O
i
o
i
o
i
o
i
o
i
o
o
i
o
03
0
w
«
"d
o;
CJ
o
OD
03
47
-------
Table III.3
Annual VOC Related Incidence*
Source Categories
LD Gas Vehicles
LD Gas Trucks
Misc. Solvent Use
HD Gas Trucks
Hospital Sterilizers
Diesel Vehicles
Misc. Surface Coating
Wood Stoves
Residential Incineration
Dry Cleaning
Cold Cleaners
POTtfs
Petroleum Waste Water Treatment
Misc. EDC
Off-Highway Vehicles
Other
Base Current
Case Rules
1980
1995
Expected
Controls
1995
Additional
Controls
1995
9.99
6.40
2.80
1.86
1.54
1.25
1.22
0.91
0.67
0.54
0.51
0.48
0.43
0.39
0.37
7.76
2.64
3.16
0.88
1.79
0.72
1.39
0.87
0.72
0.19
0.37
0.55
0.04
0.53
0.42
7.76
2.64
3.16
0.88
1.79
0.72
1.39
0.87
0.72
0.19
0.37
0.55
0.04
0.53
0.42
7.35
2.50
2.18
0.84
0.02
0.68
1.39
0.54
0.72
0.17
0.14
0.14
0.02
0.53
0.42
11.26
8.98
8.39
5.67
Totals
Percentage
50.62 31.00 30.42
77.76 71.03 72.42
23.32
75.69
* VOC related incidence estimates include formaldehyde.
48
-------
Table III.4
Annual PM Related Incidence
Source Categories
Diesel Vehicles
Chrome Plating
LD Gas Vehicles
Comfort Cooling Towers
Industrial Cooling Towers
Wood Stoves
LD Gas Trucks
Fireplaces
Glass Mfg.
Brick Mfg.
HD Gas Vehicles
Waste Oil Burning
Coke Ovens
Residential Dist. Oil
Utility Resid. Boilers
Totals
Totals - All Categories
Percentage of All Categories
Base
Case
1980
Current
Rules
1995
Expected
Controls
1995
Additional
Controls
1995
11.23
7.99
6.17
4.37
2.99
2.88
1.43
1.42
0.66
0.44
0.44
0.35
0.35
0.24
0.20
41.17
42.06
97.88
5.21
11.23
3.41
5.97
3.47
2.66
0.11
1.59
0.15
0.01
0.41
0.41
0.04
0.28
0.22
35.17
36.03
97.61
5.21
1.75
3.41
0.00
1.14
2.66
0.11
1.59
0.15
0.01
0.41
0.41
0.04
0.28
0.22
17.39
18.25
95.29
4.96
0.08
3.24
0.00
0.52
1.66
0.10
1.59
0.15
0.01
0.39
0.41
0.04
0.21
0.01
13.37
14.00
95.50
49
-------
additional controls scenario, but reductions here are not as
significant as those in the other two cases.
Table III.3 shows that in the base case, 15 source
categories contribute 78 percent of the voc-related incidence in
the five cities. The motor vehicle categories are prominent in
VOC-related incidence in all cases modeled. Many of the nonmotor
vehicle categories that appear in Table III.3 are there because
they are large VOC emitters, and are assumed to contribute to
formaldehyde formation, not because they are direct emitters of
specific toxic compounds. An obvious exception to the above is
hospital sterilizers, which are ethylene oxide emitters.
The combined PM and VOC incidence for the base case can be
seen in Figure III.2. As the figure shows, highway vehicles
account for over half of the total incidence in 1980. The
incidence for the next three largest categories—chrome platers,
comfort cooling towers, and industrial cooling towers—comes
entirely from PM. Six categories account for 75 percent of the
total annual incidence with the remaining incidence coming from
numerous emitters of small amounts of toxic substances.
b. Current Rules Scenario
Reductions in VOC related incidence from the base year to
1995 under the current rules scenario comes mostly from motor
vehicles and can be attributed to the current FMVCP with its
associated emission standards and vehicle inspection and
maintenance programs. Almost 90 percent of the incidence
reduction in the current rules scenario is through these motor
vehicle control programs. Other VOC-emitting categories where
current rules are effective in reducing incidence include dry
cleaners, cold cleaners, and petroleum refinery wastewater
treatment.
For particulates, current rules are much less effective in
reducing incidence. As with organics, motor vehicle emission
controls provide most of the expected reductions in PM-related
incidence. Some of the reduction observed from motor vehicle
emission controls is offset by expected growth in emissions from
50
-------
ON
ON
*—<
z
•—<
u
I—<
O
H
CN
O
o
«
H
Z
O
u
^H
H
»—<
u
CO
o
o w
l^H* ^^
CO —
c s
CD S
o 3
CO g
2 ! I
•4—• 2 n
c^™ o •*
o * *
V-> e -S
c
g
"to
to
"E
LLJ
§
"8
-
o
o
UIT)
O
•o
•o
to
,1
I
O
•3
I
•&
-\y ' - *
,%% ^ "To -
oo
cs
C3
o
oo
O
C—
_L
J_
I
I
SHI1ID S NI HVHA
S3SVD HHDMVD SS3DX3 1V1O1
51
-------
important contributors to incidence such as chrome platers,
industrial cooling towers, and comfort cooling towers.
As Figure III.2 illustrated, the overall contribution of
highway vehicles to the total incidence significantly declines
from the base case to the current rules scenario. At the same
time, the combined contribution of chrome platers, comfort
cooling towers, industrial cooling towers, wood stoves, and
hospital sterilizers increases from 22 percent in the base case
to 39 percent in the current rules scenario. The only category
in this group to show a decrease in incidence is wood stoves.
The magnitude of the "other" category decreases by two cases of
annual incidence in the current rules scenario.
c. Expected Controls Scenario
Motor vehicle-related incidence does not change in the
expected controls scenario. Summertime gasoline RVP limits would
not reduce the average VOC emissions from motor vehicles by 1995
for a number of reasons. There is a current RVP rule in
California, so any Federal rule will not provide any benefits in
that State. Second, summertime gasoline RVP in 1980, the base
year for this study, was not as high as in-use RVP values in the
mid-to-late 1980s. A MOBILE4 (U.S. EPA, 1989) sensitivity
analysis performed using regional July RVP values for 1980 showed
negligible differences in 1995 emissions with and without RVP
control. Third, RVP limits are only expected to be in effect
during the five warmest months of the year. Because annual
average exposures are of interest for toxics, only 5/12 of the
time do RVP limits reduce organic emissions.
Reductions in particulate-related toxics incidence in the
expected controls scenario are significant and are largely
attributable to probable NESHAP regulations. Three source
categories are responsible for most of the incidence reduction in
this scenario: chrome plating, comfort cooling towers, and
industrial cooling towers. For chrome platers, reductions
reflect going from a 70 percent average control level in the base
year to a NESHAP that could result in a 95 percent level of
control on all plating shops. Comfort cooling tower related
52
-------
incidence drops from 4.4 cases in the base year to none in the
1995 expected controls case. This reflects the anticipated
chromium ban under the Toxic Substances Control Act which would
affect all comfort cooling towers. Industrial cooling tower
related incidence is also expected to drop substantially in the
expected controls scenario with anticipated NESHAP controls. The
NESHAP requirements are estimated to be achievable by use of a
high efficiency drift eliminator.
The effect of these reductions on the overall incidence was
previously illustrated in Figure III.2. While the incidence
associated with highway vehicle emissions remains unchanged from
the current rules case to the expected controls scenario, the
combination of industrial cooling towers, comfort cooling towers,
and chrome platers makes up only 6 percent of the total incidence
in the expected controls scenario compared with 31 percent in the
current rules case. The "other" category has decreased by about
0.6 cases per year, but the relative importance of this category
has increased, now accounting for 24 percent of the total
incidence.
d. Additional Controls Scenario
Applying the set of controls included in the additional
controls scenario by 1995 provides some additional reductions in
incidence when compared with the expected controls scenario. The
additional reduction in incidence is about 11.5 cases per year.
Seven of these cases are from organic toxic emission reductions
and four are from particulate reductions. The largest reduction
in VOC-related incidence in this case comes from controls being
applied to hospital sterilizers, which emit ethylene oxide.
Reductions shown in Table III.3 for miscellaneous solvent use
result from an assumed 20 percent reduction in consumer solvent
emissions. Motor vehicle emission reductions reflect an assumed
5 percent reduction in estimated 1995 vehicle miles traveled
being achieved via transportation control measures.
Table III.5 itemizes the changes in annual incidence that
might be expected from candidate Federal rules and CTGs being
considered by EPA to reduce tropospheric ozone levels. Rules
53
-------
Table III.5
Changes in Annual VOC Related Incidence Via
Additional Controls
Source Categories
Candidates for Federal Rules
TSDF (Accelerated and Comprehensive)
Commercial/Consumer Solvents
Marine Vessel Loading
Architectural Coating
Industrial Coating
Traffic Paint
Change in Incidence:
Expected to Additional
Controls Scenario
Before After Change
0.27
2.02
0.01
0.30
0.06
0.08
0.02
1.62
0.00
0.15
0.02
0.02
0.25
0.40
0.01
0.15
0.04
0.06
Total Federal Rules
Candidates for CTGs
SOCMI Distillation
Indus. Wastewater (and POTW)
Autobody Refinishing
SOCMI Reactor and Batch Processes
Petroleum Wastewater
Web Offset Lithography
Total CTGs
2.74
<0.01
0.55
0.12
0.02
0.04
0.05
0.78
1.83
<0.01
0.14
0.05
<0.01
0.04
0.05
0.28
Candidates for Available Control Technologies (ACTs)
Cleanup Solvents
Adhesives
Ink Manufacturing
Paint Manufacturing
Pesticide Application
Transportation Control Measures*
Total ACTs
Other Categories
Other Federal Rules**
Other SIP Rules***
Total - All Categories
13.32 12.02
1.98
4.07
0.02
1.63
22.89 15.78
0.91
0.00
0.41
0.07
0.02
0.00
0.00
0.50
1.14
0.004
0.001
0.005
0.17
12.00
0.56
0.003
0.000
0.000
0.09
11.37
0.58
0.001
0.001
0.005
0.08
0.63
1.30
1.96
2.44
7.11
* e.g., measures such as trip reduction ordinances, employer
based transportation management, improved public transit,
and parking management.
** e.g., categories including hospital EtO sterilizers and pulp
and paper production.
*** e.g., categories such as carbon black production, degreasing,
by-product coke manufacture, pertroleum refinery fluid
catalytic cracking units, service stations, and solvent
use.
54
-------
listed in this table are those with nonzero incidence estimated
for the base case. The results for toxics are similar to the
results for ozone control in that no single measure provides
large reductions in VOC emissions or incidence. This results
from the fact that there are many VOC-emitting source categories
and targeting a few of them for control is not likely to be
effective in reducing total VOC emissions or incidence.
Reductions in wood stove-related incidence in the additional
controls scenario reflect an assumption that SIP regulations
could be enacted that could produce a 40 percent reduction in
existing stove particulate emissions.
Table III.5 also itemizes the changes in VOC related
incidence that might be achieved in the additional controls
scenario from control measures considered as ACTs by the ozone
program. Transportation control measures and reductions in
cleanup solvent emissions are estimated to be the most effective
of the ACTs in reducing incidence.
The most significant reduction in annual PM incidence in
this scenario comes from applying chrome plating emission
controls that have been proposed in California to all study
areas. Motor vehicle incidence reductions shown in Table III.4
for particulates reflect the same 5 percent VMT reduction that
was used to estimate organic emission changes.
Figure III.2 showed that the near elimination of emissions
from hospital sterilizers, cooling towers, and chrome platers
accounts for most of the overall reduction in incidence from the
expected controls scenario to the additional controls scenario.
Incidence from the "other" category decreased by about three
cases per year and incidence from highway vehicles decreased by
one case per year.
2. Maximum Individual Risk
While the results presented thus far in this chapter have
focused on estimated annual incidence, maximum individual risk is
another measure that can be used to characterize the hazards
associated with ambient air toxics. The analyses presented here
consist of determining the number of sources with MIR
55
-------
contributions above one in a million (1 X 10~6) for individual
compounds, and how that number changes under the different
control scenarios.
Estimates of MIR (ground-level concentration times unit risk
factor) were made using the HEM. Risks from industrial cooling
tower chromium emissions were estimated from a separate analysis
using the Climatological Dispersion Model (CDM) (Irwin et al.,
1985). Industrial cooling towers were not modeled as point
sources using HEM because they were included in the point source
emission inventory after the HEM runs were completed. CDM does
not estimate concentrations and risk at as many receptors as HEMf
but would be expected to capture most, if not all, of the peak
values when modeling elevated sources.
The source categories with greater than one in a million MIR
total over the five-city study area in the base year are listed
below.
Chemical Manufacturing
Iron/Steel Manufacturing
POTWs
Gasoline Marketing
Nonferrous Metals Production
Glass Manufacturing
Industrial Boilers - Oil
Dry Cleaning
Hospital Sterilizers
Solvent Evaporation
Petroleum Refining
Stage II - Vehicle Refueling
Utility Boilers - Oil
Utility Boilers - Coal
Surface Coating
Printing
Pulp & Paper Manufacturing
Figure III.3 presents the results of the MIR analysis. Note
that each source-pollutant combination contributing to one in a
million MIR is counted every time it occurs. Thus, if the MIR
for a source is above the 1 X 10~6 threshold for both benzene and
methylene chloride, it is counted as two occurrences. This
accounting procedure does not make a big difference in reporting
the results, however, as 215 of the 247 occurrences reported for
1980 are unique sources.
56
-------
en
tn
GO cd cd
s ^ .s
8 S
u
•a
u
o
a
a
o
a
of
a
•»^
V
a
o
a
o
u
03
fl
*^-l
&
B
o
u
fl
aj
o
ex,
I
u
o
o
on
O
-------
When the base year results are compared with the 1995
current rules scenario, there is a 27 percent reduction in 1 X
10~6 MIR occurrences. Only a 1 percent reduction in these
occurrences is estimated with expected controls applied. The
difference between the additional controls scenario and the
expected controls scenario greater than 1 X 10"^ MIR estimates is
44 percent.
Contrasting the results in Figure III.3 for MIR with those
from Figure III.l for incidence shows that the percentage
reductions in both MIR and incidence from the base year to the
1995 current rules case are nearly equal. The percentage
decrease in incidence from the current to the expected rules case
is larger than the associated decrease in MIR threshold
exceedances. This difference occurs because most of the
incidence reduction in the expected rules case is attributable to
particulate control measures, and although area sources are
important contributors to PM related toxics incidence, the MIR
for area sources cannot be estimated. Conversely, the maximum
control case shows a much higher reduction in MIR threshold
exceedances than it does in incidence reduction potential.
C. DISCUSSION
The modeling results presented in the first two sections of
this chapter show that the most effective rules for reducing
future year expected incidence and MIR are motor vehicle emission
controls and NESHAPs. Both particulate and VOC related toxics
are reduced via motor vehicle controls. NESHAPs are expected to
be especially effective in reducing particulate toxics.
The motor vehicle emission controls that are shown to be
particularly effective in reducing expected incidence in future
years include both the Federal Motor Vehicle Emission Control
Program and inspection and maintenance programs.
The NESHAP that the modeling shows to be particularly
effective in reducing PM-related incidence are modeled in this
analysis as "expected controls." Thus, they have not yet been
proposed. This analysis shows the importance of proceeding with
58
-------
the potential NESHAPs for chrome platers and industrial cooling
towers from an incidence reduction standpoint. The anticipated
chromium ban under the Toxic Substances Control Act would affect
all comfort cooling towers and would also provide significant
reductions in incidence if enacted.
Many VOC regulations are targeted as ozone reduction methods
and so some of the measures included here are geared more toward
reducing total VOC emissions rather than the specific toxic
compounds that are responsible for the associated incidence. The
PM regulations, on the other hand, are often targeted to reduce a
specific toxic compound, and are therefore more effective in
reducing stationary source-related incidence. As an example,
sources like chrome platers and cooling towers are not normally
considered to be contributors to ambient particulate problems,
but future regulations for these source types are likely to be
effective in reducing particulate related cancer incidence.
Figure III.4 provides a summary of the emission sources with
the 15 highest related incidences. As this figure shows, current
rules are expected to get as much as an additional 27 percent
reduction in excess cancer incidence in the five study areas.
Adding rules that could be expected to be achieved by 1995 would
almost double the anticipated reduction in cancer incidence
expected via the current rules. Up to 60 percent of the base
case incidence could be reduced by applying additional or best
available controls.
59
-------
W
oo ,0
I'
02
CQ
»— i
o
Ol
1-H
»— 1
CN
i— i
»— t
OO
oo
HIGHWAY VEHICLES
»— t
o
oo
T-H
CN
1-H
r— 1
O
oo
CHROME PLATING
1
*
*
0:°*.
I
o:""}
:en
I
vo :en
•
xf joo
xf- jen
co :
COMFORT COOLING TO
WOODSTOVES
*o
o
1— 1
r-t
en
O
en
2
INDUSTRIAL COOLING r
I
•
• C^
cs :o
CN :O
4
*
CN : oo
en : «-•
*
*
^^J * O^3
en •' «— i
j
oo • «o
•
•
MISC. SOLVENT
HOSPITAL STERILIZERS
VO
,_,
VO
r—!
VO
r— i
FIREPLACES
\
•
*
•
vo
o
VO
0
o
to
o
d.
CN
o
o
s
O
3
0
O
TREAT
PETROL WASTEWATER
^+.
O vo
O oo
Tf
P oo
0 "-
^J-
^^S vj
o ^>i
xf <*}
0 ^
COKE OVENS
OTHER
o
VO
^-^
en
t-~
en
^^^
^^r
oo
oo
o
r-
vo
CN
ON
PQ
O
TOTAL 5 CITY INCIDEN
(Percent reduction from base year)
60
-------
ABBREVIATIONS AND ACRONYMS
ACT available control technology
B(a)P benzo(a)pyrene
BACT Best Available Control Technology
BEA Bureau of Economic Analysis
COM Climatological Dispersion Model
CTG control technique guideline
DNPH dinitrophenylhydrazine
DOC Department of Commerce
ELOC existing level of control
EPA Environmental Protection Agency
FIP Federal Implementation Plan
FMVCP Federal Motor Vehicle Control Program
GC/MS gas chromatograph/mass spectrometer
HC hydrocarbon
HDDV heavy-duty diesel vehicle
HEM Human Exposure Model
IEMP Integrated Environmental Management Projects
I/M inspection and maintenance
LAER Lowest Achievable Emission Rate
MRI maximum individual risk
MSA Metropolitan Statistical Area
NAPAP National Acid Precipitation Assessment Program
NEDS National Emissions Data System
NESHAP National Emission Standards for Hazardous Air
Pollutants
NSPS New Source Performance Standards
NSR New Source Review
OMS Office of Mobile Sources
PM particulate matter
POM polycyclic organic matter
PSD Prevention of Significant Deterioration
RCRA Resource Conservation and Recovery Act
RIM Regulatory Impact Model
RTI Research Triangle Institute
61
-------
RVP Reid Vapor Pressure
SCC Source Classification Code
SIC standard Industrial Classification
SIP State Implementation Plan
TSCA Toxic Substances Control Act
TSDF Treatment, Storage and Disposal Facility
TSP total suspended particulates
VMT vehicle miles traveled
VOC volatile organic compounds
62
-------
REFERENCES
Carey, 1987: Penny M. Carey, U.S. Environmental Protection
Agency, Office of Mobile Sources, "Air Toxic Emissions from
Mobile Sources, Technical Report," EPA-AA-TSS-PA-86-5,
January 1987.
Haemisegger et al., 1985: E. Haemisegger et al., "The Air Toxics
Problem in the United States: An Assessment of Cancer Risks
for Selected Pollutants," U.S. Environmental Protection
Agency, Washington, DC, May 1985.
Irwin et al., 1985: John S. Irwin, Thomas Chico, and Joseph
Catalano, "COM 2.0—Climatological Dispersion Model," EPA-
600/8-85-029, PB86-136546, Atmospheric Sciences Research
Laboratory, Office of Research and Development, U.S.
Environmental Protection Agency, Research Triangle Park, NC,
November 1985.
Lahre, 1988: Thomas Lahre, "Cancer Risks From Air Toxics in
Urban Areas," Paper No. 88-127.6, presented at 81st Annual
Meeting of APCA, Dallas, Texas, June 19-24, 1988.
Pechan, 1989: E.H. Pechan & Associates, Inc., "Analysis of Air
Toxics Emissions, Exposures, Cancer Risks and
Controllability in Five Urban Areas: Volume I—Base Year
Analysis and Results," EPA-450/2-89-012a, prepared for U.S.
Environmental Protection Agency, Office of Air Quality
Planning and Standards, Research Triangle Park, NC, July
1989.
Radian, 1989: Radian Corporation, "VOC/PM Speciation Data System
User's Guide," prepared for U.S. Environmental Protection
Agency, Noncriteria Pollutants Program Branch, Research
Triangle Park, NC, February 1989.
Rogozen et al., 1985: Michael B. Rogozen et al., Science
Applications International Corporation, Hermosa Beach, CA,
"Development and Improvement of Organic Compound Emission
Inventories for California," Final Report, prepared for
State of California Air Resource Board, Sacramento, CA,
January 1985.
Scheff et al., 1989: Peter A. Scheff et al., "Source
Fingerprints for Receptor Modeling of Volatile Organics,"
JAPCA, Volume 39, No. 4, April 1989.
U.S. DOC, 1981: U.S. Department of Commerce, Bureau of Economic
Analysis, "1980 OBERS: BEA Regional Projections - Economic
Activity in the United States, Volume 1, Methodology,
Concepts, and State Data," July 1981.
U.S. EPA, 1985a: U.S. Environmental Protection Agency, "A
Strategy to Reduce Risks to Public Health from Air Toxics,"
Washington, DC, June 1985.
63
-------
U.S. EPA, 1985b: "Compilation of Air Pollutant Emission Factors,
Volume II: Mobile Sources," AP-42, Fourth Edition, Office
of Mobile Sources, Ann Arbor, MI, September 1985.
U.S. EPA, 1986: U.S. Environmental Protection Agency, "User's
Manual for the Human Exposure Model (HEM)," EPA-450/5-86-
001, Office of Air Quality Planning and Standards, Research
Triangle Park, NC, June 1986.
U.S. EPA, 1987a: U.S. Environmental Protection Agency, Office
of Air Quality Planning and Standards, "Implications of
Federal Implementation Plans (FIP's) for Post-1987 Ozone
Nonattainment Areas," Draft, March 1987.
U.S. EPA, 1987b: U.S. Environmental Protection Agency, "Toxic
Air Pollutant/Source Crosswalk - Information storage and
Retrieval System User's Manual," EPA-450/4-87-0236, Office
of Air Quality Planning and Standards, Research Triangle
Park, NC, December 1987.
U.S. EPA, 1989: U.S. Environmental Protection Agency, "User's
Guide to MOBILE4 (Mobile Source Emission Factor Model),"
EPA-AA-TEB-89-01, Office of Air and Radiation, Office of
Mobile Sources, Ann Arbor, MI, February 1989.
Wagner et al., 1986: Janice K. Wagner et al., "Development of
the 1980 NAPAP Emissions Inventory," EPA-600/7-86-057a,
Alliance Technologies Corporation, Bedford, MA, December
1986.
64
-------
APPENDIX A
CHANGES TO BASE CASE INVENTORY
There are several differences between the inventory used to
perform the analyses in this report and the inventory used to
perform the base year analysis documented in the Volume I
companion report (Pechan, 1989). These changes were made to
better and.more fully assess the impact of regulating emission
sources that are candidates for Federal rules, Control Technique
Guidelines (CTG), or Achievable Control Technologies (ACT) rules.
The methodology used to determine what changes and additions to
make to the inventory is discussed below.
Using the Crosswalk data retrieval system (U.S. EPA, 1987b)
and the VOC Speciation Data System (Radian, 1989), each of the
source categories currently being considered for control by
Federal rule, CTG, or ACT was examined to determine whether any
of the toxic species of interest to this study could be emitted
from each category. This assessment led to three types of
changes or additions being made to the inventory. The first
involved adding toxic emission factors for point source SCCs for
lithography and marine vessel loading. The second type of change
consisted of adding a new area source SCC to the inventory to
include VOC and toxic emissions from treatment, storage, and
disposal facilities (TSDFs). The final change consisted of
adding five area sources (architectural surface coating, traffic
paints, autobody refinishing, industrial maintenance coating, and
miscellaneous industrial surface coating) whose VOC emissions
were already included in SCC 99999971, Surface Coating. This SCC
was eliminated after being broken down into the five new area
source SCCs to allow for different toxic emission factors and
control efficiencies for each of the individual surface coating
area sources. This breakdown is illustrated in Figure A.I.
In the first grouping of changes, several point source SCCs
were identified as being toxic emitters which previously had no
toxics listed. The categories of web-offset lithography and
marine vessel loading were found to be lacking toxic emissions in
A-l
-------
o>
CO
-------
the previous version of the inventory. Toxics for both of these
categories were identified using the VOC Data Speciation System
(Radian, 1989). The profile used for SCC 40500101, Lithography -
Inking and Drying - Direct Fired Dryer, is based on composite
survey data of the industry and a gas chromatograph/raass
spectrometer (GC/MS) analysis of a sampling train catch. This
profile lists methylene chloride as making up approximately 35
percent of the VOC emissions from this source. The other
lithography source for which a toxic emission factor was added is
SCC 40500401. The profile for this source, Printing Press -
Lithography Inking and Drying, is also based on composite survey
data and a GC/MS analysis. About 22 percent of the VOC emissions
from this source are formaldehyde emissions. The toxic emissions
added for each city in this category can be seen in Tables A.I
and A.2.
Benzene emission factors were added to four SCCs in the
marine vessel loading category. These emission factors were
determined through the VOC Speciation Data System (Radian, 1989).
The VOC emissions for SCCs 40600243, 40600248, and 40600251
contain 2.4 percent benzene according to the profile Fixed Roof
Tank - Crude Oil Refinery. The speciation for this profile is
based on an engineering evaluation of test and literature data.
The profile for SCC 40600245, Gasoline - Summer Blend, shows 0.77
percent-benzene in its speciation. This profile information was
based on vapor samples composed of four product types combined in
proportion to 1979 sales figures for California. The vapor
samples were analyzed using a dual detection FID/PID GC. The
benzene emissions added for this category are listed in Table
B.3.
A new area source SCC (99999814) was added to the inventory
to include emissions from hazardous waste TSDFs. The VOC
emissions for each county from this source were obtained from a
TSDF data base developed by Research Triangle Institute (RTI).
The toxic emissions from TSDFs were determined using the VOC
Speciation Data System (Radian, 1989) and are shown in Tables A.3
and A.4. The applicable profile, Solid Waste Disposal - Average,
A-3
-------
Table A.I
Methylene Chloride Emissions Added by Source Category
City
A
B
C
D
E
Totals
Architectural
Coating
(tpy)
31.6
206.9
7.6
14.2
16.4
Industrial
Traffic Maintenance
Paint Coating
(tpy) (tpy)
8.1
52.8
1.9
3.6
4.2
6.5
42.4
1.6
2.9
3.4
Web
Offset
Lithography
(tpy)
NA
1,359.9
NA
NA
29.3
276.6
70.5
56.7
1,389.2
Table A.2
Formaldehyde Emissions Added by Source Category
City
A
B
C
D
E
Totals
Web
Offset
Lithography
(tpy)
NA
12.6
NA
3.7
4.6
20.9
A-4
-------
Table A.3
Benzene Emissions Added by Source Category
City
A
B
C
D
E
Totals
Industrial
Architectural Traffic Autobody Maintenance
Coating Paint Refinishlng Coating
(tpy) (tpy) (tpy) (tpy)
1.9
12.3
0.5
0.8
1.0
0.5
3.1
0.1
0.2
0.2
10.6
69.3
2.5
4.7
5.5
0.4
2.5
0.1
0.2
0.2
16.5
4.2
92.7
3.4
TSDF
(tpy)
2.3
193.8
17.8
17.3
0.0
231.3
Table A.4
Perchloroethylene Emissions Added by Source Category
City
A
B
C
D
E
Totals
TSDF
(tpy)
0.1
7.5
0.7
0.7
0.0
9.0
A-5
-------
shows 0.10 percent perchloroethylene and 2.57 percent benzene.
This profile was developed as an average of the original profiles
representing the source category 5XXXXXXX (solid waste disposal).
The final group of changes to the inventory involved the
breakdown of an existing area source and the addition of toxic
emission factors to the newly created SCCs. These changes
affected SCC 99999971, Surface Coating, which has now been
eliminated from the inventory. The new split of this area source
is illustrated in Figure A.I. The allocation of emissions from
the surface coating category was based on data showing the total
1979 national solvent use in paints and coatings by type of
application (Rogozen et al., 1985). Thinner usage was
apportioned to each category in direct proportion to the
percentage of paints and coatings in each category. The toxic
emissions from these surface coating categories are listed by
city in Tables A.I and A.3.
The profile for architectural coatings was based on the
composite of three profiles: solvent based coatings (composite
of profiles for lacquer, primer, and enamel in proportion to
usage in Southern California), thinning and cleanup solvents
(composite based on sales volume from nine solvents used with
architectural coatings), and water based coating (composite of
seven coatings in proportion to 1980 sales figures). These three
profiles were found in the VOC Speciation Data System (Radian,
1989) and were combined in proportion to the percentage of VOC
emissions from each of these categories from a 1984 survey of the
New York major metropolitan area and the entire State of New
Jersey (Scheff et al., 1989). This composite included methylene
chloride as 1.51 percent of VOC emissions and benzene as 0.09
percent of VOC emissions. Traffic paints and maintenance
coatings were assumed to have the same profile as architectural
coatings.
The profile used for autobody refinishing, Autobody Repairs,
was based on a GC/MS analysis of the semivolatile compounds of 12
automotive aftermarket paint and thinner samples of acrylic
lacquer and alkyd enamel from three paint manufacturers (Radian,
A-6
-------
1989). The samples were combined using market statistics. This
profile identified benzene as making up 1.51 percent of VOC
emissions. The profiles for miscellaneous industrial surface
coating did not include any of the toxics being studied in this
report.
A-7
-------
APPENDIX B
MOTOR VEHICLE EMISSION PROJECTIONS
Motor vehicles are an important contributor to estimated
base year incidence, so care was taken to ensure that the most
recent information available was used to estimate future year
changes in motor vehicle emissions. Motor vehicle emission
factors for 1980 and 1995 are shown in Table B.I. The vehicle
categories used match those in the EPA National Emissions Data
System (NEDS). There is only one diesel-powered vehicle category
in NEDS (heavy-duty diesels), but the emission factors for heavy-
duty diesel vehicles (HDDVs) listed in Table B.I represent a
weighted average factor for all diesels including light-duty
diesel vehicle and light-duty diesel truck travel.
Organic toxic emissions were estimated as a percentage of
total hydrocarbons using MOBILES (Federal Test Procedure
conditions) and toxic fractions (Carey, 1988). Particulate
emission estimates were made using the current schedule of motor
vehicle particulate standards and registration and travel
fractions from MOBILE3 (U.S. EPA, 1985b).
Because formaldehyde incidence is estimated based on total
VOC emissions, there was no need to apply the Table B.I
formaldehyde emission factors for this study. In addition,
because the difference between 1980 and 1995 emission factors for
total hydrocarbon (HC), benzene, and 1,3-butadiene emissions was
approximately the same, total HC was used as a surrogate for
estimating 1995 emission levels for all organic toxics. This
allowed area specific I/M program effectiveness values for VOC to
be taken into account in the Regulatory Impact Model (RIM).
B-l
-------
Table B.I
Motor Vehicle Air Toxic Emission Factors
(mg/mile)
Pollutant Vehicle Type 1980
Formaldehyde
Benzene
1,3-Butadiene
Particulates
LDGV
LDGT
HDGV
HDDV
LDGV
LDGT
HDGV
HDDV
LDGV
LDGT
HDGV
HDDV
LDGV
LDGT
HDGV
HDDV
61.7
76.2
290.8
152.5
168.6
229.4
412.4
51.3
12.7
20.5
32.8
15.9
14.9
18.2
31.3
1,973.0
1995
No I/M
11.6
27.4
78.7
80.6
52.8
90.1
115.9
28.3
4.1
8.7
8.9
8.6
6.0
6.9
19.2
566.5
1995
With I/M
7.9
17.2
78.7
80.6
36.9
60.0
115.9
28.3
2.8
5.5
8.9
8.6
5.1
5.1
18.0
566.5
LDGV = light-duty gasoline-powered vehicle
LDGT = light-duty gasoline-powered truck
HDGV = heavy-duty gasoline-powered vehicle
HDDV = heavy-duty diesel-powered vehicle
B-2
-------
APPENDIX C
CONSTRAINT FILES
Presented in this appendix are listings of the VOC and PM
constraint files for the current rules and additional controls
scenarios. Table II.8 in the main body of the report lists the
constraints that were added to the current rules constraint file
to simulate the expected controls scenario. The information
presented below can be used as a guide in interpreting the
information in the Appendix C tables:
No - constraint number
Name - constraint name (first 19 characters)
Region IDs - regional (city) applicability of constraint
Ind Cat IDs - applicable industrial categories (SCCs)
(? = wildcard)
N - applies to new sources?
R - applies to replaced sources?
E - applies to existing sources?
A - applies in attainment areas?
N - applies in nonattainment areas?
Beg - constraint beginning year
Targ - constraint target year for full implementation
(The beginning and target year do not affect the
results for constraints applicable to existing
sources)
R Ctr - relative control; based on process rate
(not applicable to this study)
A Ctr - absolute control; reduction from uncontrolled
emissions
Pentr - constraint penetration; fraction of emissions
affected
C-l
-------
c
a
«- c
• o
u u
O 1)
3 "3
a ae
c
4)
c.
L.
3
U
O
a>
0.
c.
o
L.
ac
Ol
a
Ol
z
40300150'
ro CM in
ooo
in «- •-
ooo
ooo
40300205'
ro
o
CM
O
o
ro
o
O 0
rvi CM
O 0
0 O
0 0
o oooo o ooooooo o
in inooo o ooooooo o
O OOOO O O O O O CO O O O
o oooo o ooooooo o
§o •» rvi in ro o co f1- ro -o in in r«-
o o r- r«- eo O o eo O O r«. N- m
o oooo o ooooooo o
co cocococo co in in m in vt in in co
co cocococo in m in in in in in m in
X XXXX X XXXXXXX X
»- f- o in x> ro o -» »- ro •» «- ro in
CO CO CO CO CO 00 CO CO CO 09 CO O CO CO
o oooo o ooooooo o
«- «- o «» •« ro o •» •- ro «» •- ro in
co eo co co co eo co eo co co eo o co eo
O OOOO O OOOOOOO 0
- "-- _,-.-.-,-
ro ^™ c*» ro ro c^ ^~ CM
O CO C>- O OOOO
eo Oroeo «- «- eo to
o Ooo oooo
o ooo oooo
^ O in *Ov*— rvjcoro
OOO"-OOOOOO«— O'-fMOOOOO-O
ooeooooooooocooooooooo
oooooooooooooooooooo
3 IMO CO Ul Ul UJ »~ _J t— ^
u. z — r o >— ac ac oc is *- in «.
uizzoe 31— uj r-
ui m o — < x z o* u. co -i ui ae ui
Z Z 33 — O > — X
NZJ-UJ— — I—IUI 1— H-
z cj z _i z o ox — — — co oe z
ui o — cj a. ee ae >- x x X o ui ui
ca co > >- «t i— i— _i u u u o co >
ee oc ui uiooooom -i
• • • o cs a. o. o. co co co 3 3 o
oe co
a. a, o. • > >
Z Z Z CO CO CO CO CO CO CO CO CO CO CO
co co to a. a. a. 0.00.0.0.0.0. o
Ul UJ Ul CO CO CO CO CO CO CO t/1 CO CQ OT
zrzzzz z zzzzzzz z
o*- rvi ro ~» in -o N- oo o o
§o
o
o
0 «-
o o
o in
ca eo
O 0
CO CO
CO CO
x x
(M O
CO CO
00
N o
co eo
30500104-
' 30500202'
rvj o
0 O
•- rvj
0 O
o o
in in
o o
CM O r-
o o o
•*a* ^- rvj
rvj o o
ooo
ooo
21 NSPS- SYNTHETIC FIB Al
22 RACT - ASPHALT BLOW Al
o o
0 0
o o
o o
o in
0 0
CO CO
CO CO
X X
o o
CO CO
o o
CO CO
«- r- CM
O C-- f-
•- ro o
*- o o
ooo
ooo
UJ
u
ae
§
in
UJ
ae
i— i—
U CJ
ac oc
ro >»
rvj rvj
O O
0 0
O O
O O
0 0
CO CO
co co
X X
o o
COCO
o o
CO CO
99999981'
>9999973'
«- in
O 0
O 0
o o
25 RACT - DRYCLEANING Al
26 RACT - GRAPHIC ARTS Al
ooo
ooo
ooo
ooo
co in ro
ooo
CO CO CO
CO CO CO
XXX
O N. O
O eo co
So o
CO CO
' 30600803-
1
(0600802*
(0600805-
rvj ro «- -*
oooo
eo in co co
«— o o o
oooo
oooo
ae ui co
z u
>- — 3
H- U. U-
— UJ
co a: u.
Z Ul
Ul • OC
O CJ
CO t-
— _t UJ
z i a.
i i i
H- H- ^
U U U
oe oe ae
P- CO O
o o
o o
O co
o o
o o
r- in
o o
ca co
co co
X X
0 0
o o
o o
•a -o
CO CO
0 0
o rvj
O st
o rvi
CO O
0 0
— CO
U -1
3 O
u. >
X X
U CJ
0 0
CO CO
h- h-
<_> U
ae ac
O "-
O 0
0 0
0 0
0 0
O in
in O
0 0
co in
in in
X x
in o
o o
0 0
o -o
0 CO
0 0
o
ro
o
1*1
c
o
to
o
v— c*« P"
ooo
CM r- O
ooo
ooo
52 RACT - STYRENE/BUTA Al
53 RACT - VOL STORAGE Al
C-2
-------
C
«J
U CJ
V a)
J3 3
(0 a!
01
3
U
CJ
O
i_
c
0
o.
CJ
L.
U
Of
Ol
c.
eg
(—
o»
m
z
Ul
Of
tn
a
u
o*
CVJ^v-^^CQ CM *" ^ *^^»j— *» ^ N. 6O O O» ^- O*
oooooo* o *- o ooo o o o o «— o- o O-
oooooo* oo o ooo o 00000*00
K» ** fo m to o» K» KI >o -0*00 in CM CM CM CM ~*t ^f **T ^f "^" *^ *^ "^ ^f ^f **^ ^f **^ *^
Kl O »-»*•- 0- Kl KICM T~ «- >» «t -O 00 «- «- •- CM «»• O IM CO
O «— Q O r- <*• O O»- ^-OOKI«»O O O O O C-- O in O
o o o o «- o o oo oooooo o 000000*-
ooooooo oo oooooo o ooooooo
KI-»'KIKIKI«-O KIKI -» -O O -O -O ~O in fMCMCMCMCMKIKI
OOOOOOO OO OOOOOO O OOOOOOO
ev.
OOOOO«-OO*-CMOO>-OOOOOOOOOOOOOOOOOOOO
ooooooooooooooooooooooooooooooooo
OOOOOOOO-OOOOOOOOOOOOOOOOOOOOOOOOO
-1-1 -I _J _J _l —1 _l _1 _J
— 1 — J — J CM in — J _J_I_J_I _ 1 CM CM ^ in _l CM
«t< • 3 a
U" —1 COCOCOCOCO —
z o h- QC cu « a: _i
•—< Ul < o oooo— a.
i— i— i—
oo o Q. a. a. 0.0.0.0. 0.0.0.0.0.0. a.
ce oe accococo cocococo couicocococo co
-tin -ON-COO o «— 30601101? 30500199?
•
CM O
0 O
CM CM
0 O
0 O
m m
O 0
o o
o o
00
o o
o m
co co
o o
CO CO
CO CO
X X
in >o
CO OO
Kl O
eo co
0 0
1- H-
> 301015??? 301020??? 1
1
C>. o
r- 0
-* o
«- o
0 0
o o
CM -J-
0 0
z
o u.
z X
CO Ul
u z
z u
>- u
< tn
o —
U X
a. o.
CO CO
Kl -*
in in
0 0
o o
o o
o o
o m
CO CO
o o
CO CO
CO CO
X X
CO CO
0 0
CO CO
0 0
0 O
o o
o o
o o
0 0
O 0
Kl Kl
Kl -t
o o
CO CO
C3 O
Ul Ul
h- t-
CO CO
CJ CJ
CO CO
X X
a. a.
CO CO
in -o
in in
o o
0 0
o o
o
o
CO
o
CO
CO
X
•o
CO
o
CO
o
o
0
0
o
Kl
Kl
o
o
0
Of
a.
CO
h-
CO
_J
a.
a.
0
Kl
m
o
CO
CO
X
o
CO
o
o
40600126? 40600131L 1
0
o
o
•o
o
o
Kl
O
Of
Ul
1—
CO
1—
CJ
Of
a.
40600136? 40600141L 40600144L
99999804?
CO CO
t^ 03
in in
ooo
ooo
in in m
ooo
ooo
N. f<- rw
r-- h. N.
ooo'
tn in in
tn in in
XXX
O CM N.
00 CO 03
OOO
OOO
CO CO 03
000
P RACT - GAS BULK 04 99999806? 1
P RACT - GAS BULK 03 99999806? 1
P RACT - GAS BULK 05 99999806? 1
co tn v>
O 0 «-
in -o -o
C-3
-------
e C.1
e Constraint File
3 "5
a at
0
i.
3
0
0
O
c.
a.
u
u
u
a
a
OJ
o
a
u
c
IA
O
ai
41
ce
z
o
ooo o oooo o o o o o o o o o o o o o o o o
N. N. O O O S> O O O O N. O O O O O O O O O O O O O
OOO O OOOO UtOOOOOOOOOOOOOr-O
ooo o oooo o o o o o o o o o o o o o o o o
moo •o in o o in -»ininooinininininminu"iOom
co co x* eo o o o in m o o o >» eo co eo eo co eo eo eo o o co
ooo o oooo o o o o o o o o o o o o o o «- o
CO CO CO CO CO tO tO CO COCOCOCOCOCOCOCOCOCOCOtOCOtOCOIO
co to t/> co co to to to tococotocotococotocototococococn
XXX X XXXX XXXXXXXXXXXXXXXX
N. IS. r- (M v- m N. N- rs. N. M N. O Is. O O O O O O O O IS. O
co co eo co eo co co co co eo eo co eo co eo eo eo co to eo co eo co eo
«- v- in o «- «- «- o o o o o o o o o o o o o o o o o
co co co co co co eo co eococoeocoeocoeocoeocoeocoeocoeo
K> Kl •* «- »-
o o m o ft
ooo o o
So o o o
*O (^ ^9 ^O
OOO 0 O
CM OO C\J OO IS. -O O r-
OOOOOKI'O'— KIO>Of\J»-OOOOrvJ
OOOOOOOOOOOOOOOO'-OOOOOOOOO
ooooooooooooooooooooooooooo
in
o
in in in 4
o ooo
OOO 0 OOOO OOOOOOOOOOOOOOOO
uj — — rsi a: — — a: _i u.o3zo£(/>t/>ujujui
o >— uj ui uico^^o UJU.U.UIO zi •*»«>.O3^^co
ac i — — —xujco — ac o£ co i- xoz
a:3> > > h- a: _j zzca z- >- o tocotooeoxuiu. z i a: i z i o
O.CLCO co cococo> <: — — i— 3:z_i«i3Z)Z)Oi3 —
i i . _ uj — U> — — 1OOOOOI
OOO O OOOO OOOO^GeOCUJlUH-h-^-h-»— t— Z
oeoecK ce ac ae ae at aeaeae(/>co>
-------
c
a
a>
c
o
u
u o
I_
41 w
Ja o
ID CJ
ca
C
o
•o
•o
u
1
Q.
e.
4*
U
L.
4-*
ee
01
i.
a
o>
4)
z
a
a
u
•a
c.
o
c
o
o>
41
ac
4)
ro
o
z
o o
o o
o o
o o
9999804'
.0600141
(060080?
R- PETROLEUM REF AL
Ul
o o
o o
O 0
o o
So1
o o
CO CO
CO CO
Z X
o o
CO 60
0 O
o o
CO 00
0 0
1
.0300103? 1
(0600503'
.0300102?
0 0
in «-
o o
0 O
0300201
.0300208
o
in
o
0
o
o
0
o
CO
eo
z
CO
00
0
(0600803? 1
(0600802?
(0600801
HAP BENZENE FUGI AL
CO
Ul
(0600805-
(0600804
o o o
in o o
o o o
o o «-
000
O •» PJ
o o r-
o o o
CO CO CO
CO CO CO
XXX
*- 0 in
co co eo
«- o -»
00 CO CO
0 O 0
1
1
>9999981? 1
o ro
in O
PJ O
*- o
«- o
0 0
OOP-
o eo i-
o — o
00 O O
O 0 O
HAP SOCMI BENZEN AL
HAP - VINYL CHLO AL
S - DRYCLEANING AL
co co a.
Ul UJ CO
o
o
o
o
o
in
0
CO
CO
,
00
o
.05003??? 1
.05002011
O P-
«- in
o o
o o
in in
0 O
3 - GRAPHIC ARTS AL
a.
eo
z
o ooooooo o oo ooooooo oooo
o ooooooo o oo ooooooo oooo
O OOOOOOO O OO OOOOOOO OCOOO
o ooooooo o oo ooooooo oooo
ro o eo r— ro in in in p. o in o in ro in co in ro r^ in o in
o ooooooo o oo ooooooo oooo
co m in in in in in in in in in in in in in in in in in in in in
co in in in in in in in in in in in in in in m in in in in in in
Z ZZZZZZZ Z ZZ ZZZZZZZ ZZZZ
ro o«*«-roo«-ro in CMO OOOOOKO oomo
eo coooGoeocoooo oo oooo coooooooooooo oooo
ro o -» «- ro o •- ro in PJO ooooooo o -o o o
oo eooocoeocoooo oo oooo cocooooocococo cooooeo
O OOOOOOO O OO OOOOOOO OOOO
j*. P. P. p. p. p. p. p. p. p*.
ro roo«-pj-»pj r- ro in
O OOOOOO CO 0 •-
eo «r-^ooro«-pj o eo o
O OOOOOO O O «-
O OOOOOO O O O
-
3 t— CO Ul 1— — _1 CJ z C£ >- — • 3 — -3
X z°au.co3 at ui u. cfl ££_<,_ u. u. C3_icaui
33 — Q£ — Z 3Z— Ul 3O^O
ui ui ct •• d ui i— o i— ococeu. u- > uj <
O O Z *- — • ^- O os z ^-< ~J3cui*eic *— ^- ui o
ac ct >- z I i— >- ui ui uiz — i— — i ej u ui eo co > za. ui>-< cot— ucj>-co
ui uiooozaca — i i— eo aecea: — -~ui 001—
a. o.o.coeo^O3 o z^ • o
" ''-'>-3 CO' , 111,1 ,ic>
CO , ,
a. 0.0.0.0.0.0.0. a. o.c-1 ejocjeJcjcjej cjejcjej
co co co co eo — eo co co co
-------
(0
L.
u
at
c
o
CJ
rg to
o -o -o -o m
oooooo oo oooo o
to o ~- ^» •- I*, to ro(M ^ ^ «» -» -o eo «—
o »— o o T- <>• o o •- ^- o o ro »» O o
oooo«*oo oo oooooo o
OOOOOOO OO OOOOOO o
ooooooo oo oooooo o
0.
^~ oo IM ro o I** IM <*• 9~ ^-» co ^* *~ c3 <*• ^^ o ro r** cv« «•• *•• <^« ^~ c^
oo^-oor"Ooooo^-^^-o»»ro-*OtV'OOf»'OO
OOOO--OO«-CMOO«— OOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOOOOOOOO
ooooooooooooooooooooooooo
*CT- VOL STORAGE: ALL '
i
l
i
t
\CT-SOLVENT METAL ALL >•
(
p 02 :
p 05 :
P ALL i
t
t
P ALL !•
P ALL ^
P ALL *
t
P ALL *
<
t
P ALL 4
P - GRAPHIC ARTS 02 <•
P - GRAPHIC ARTS 02 l
P - GRAPHIC ARTS 04 <
P - GRAPHIC ARTS 05 i
t
ac ac cacaca ca c/> ca ca ca ca ca c/> ca
in >o r^coo o »— IM ro ^-m>ON.co
ro ro tororo ^^^ %j ^^•>*>*^t
o
o
0
o
in
00
o'
ca
ca
x
o
00
0
o
00
H-
H-
t-
1—
^.
0 0
0 0
(M IM
O O
.0200301? t
,0200601? t
o o
0 O
0 0
o o
_l
_J
<
3
_J
ce
ta
o
z
a.
C/)
O
o o
o o
IM IM
o o
,0200802? t
o
o
o
o
02009??? i
o
o
o
0
o
o
o
•
o
in
9-
•
O
ca
ca
x
o
00
o
CO
o
H-
h-
t—
t—
»-
o oo o to
o o o «-
oooo
IM ro ro ro
oooo
0200899? t
o
o
o
IM
O
in O O
•- «- o
o «- «-
ooo
ro to ro
ooo
ooo
ooo
ooo
ooo
*
IM
o
o
_l
X
o
UJ
_l
o
ce
t-
UJ
a.
a.
ca
0
in
0 0
o o
0 0
o «-^
0 0
o ro
in o
0* 0*
ca ca
ca ca
X X
«- o
00 CO
o o
T»- 0
N- CO
o o
t- >-
1
0500104? 30500201? 1
(0601101? 30500199?
ooooo
«- tO «— (M IM
•-OOOO
ooooo
ooooo
/
P - SCAQMD RULE 4 02 2
P ASPHALT BLOWING 2 2
•
i
ta ca
«— (M
in in
OOO
ooo
ooo
ooo
o in o
co co eo
ooo
ca ca ca
ca ca ca
ooo
ooo
ooo
ooo
in o ro
co eo in
ooo
ca ca ca
ca ca ca
X X X X X X
in O «- •-
O CM O O
oooo
oooo
o
-* to ro
ooo
ca • uj
•S 0 I-
o o
a: v:
LU a. _i
f— 3
ta ca ca
-------
c
(0
l_
frj
m
z
o
z
o
r-
o
0
o
§
o
CO
x
03
o
s
(0800199-
- PNEUMATI 05
u
a.
in
O
O
§
0
o
NT
O
CO
CO
X
03
o
8
> 40600403? 1
(0600402'
(0600401*
in
0
- SERVICE 04
cj
a.
CO
oo
§
o
0
o
o
(0600499'
o
o
o
o
o
in
in
X
03
O
O
03
' 40600403? 1
i
fM O3
O 0
sfr 03
O O
0 0
«0 0
0 0
t- O
0 O
0 0
o o
•O -O
0 0
CE STATION 02
a:
UJ
a.
CO
in
•o
o o
O r«-
0 O
o o
o o
in 0
0 0
CO CO
CO CO
X X
0303
03 03
(0300307-
O O
to o
O 0
o o
CE STATION 02
- SYNTHESI 03
Of U
UJ 40600101? 1
(0600136*
o o
o «-
•0 0
O 0
o o
in
o
- SYTHESIZ 04
- VOL STOR 02
U CJ
a. a.
CO CO
oo O
o o
> 40600131L
(0600126-
(06001411
O
o
in
o
o
CO
CO
X
03
O
03
>9999806"
• GASOLINE 2
u
a.
eo
o
0 0
00
O 0
in in
0 0
CO CO
CO CO
X X
03 03
o o
03 03
(0600401-
(0600401'
in
o
fM tO
o o
u. u.
UJ UJ
u u
co co
X X
• '
U CJ
a. a.
CO CO
«- (M
N. (v-
OOOOOOOOOOOOOO O OOOO OOOOOOOOO
OOOOOOOOOOOOOO O OOOO OOOOOOOOO
oooooooooooooo o oooo ooooooooo
OOOOOOOOOOOOOO O OOOO OOOOOOOOO
oominininininininooino in eo to o o oinoininooom
O-tfOOaaeOCOOOCOCOOOOOCOO O O O fM O in-OCOOON-OmCOtQ
OOOOOOOOOOO^OO O OOOO OOOOOOOOO
tnminininininininininininin in in in in in ininmatinminatin
ininininininininininmininin in in in in in ininininininminin
XXXXXXXXXXXXXX X XXXZ XZXXXXXXX
COeOeOeO030303CO0303CO03COO O OOOO OOOOOOOOO
OOOOOOOOOOOOOfM (M (M fM fVJ fM fMfMfMfMfMrJrMfMrM
030303030303030303030003030 o oooo ooooooooo
(MinfMOOin-O -*o fM
o o o o o (M --r in o
00 O3 fM <•" to *~ fM fM in
O O O «— O O OO fM
oooooo oo «-
oooooo oo o
o o^^OO"— O«- IOO3 ^- ^ t—
o oooooooo ^»»a- ooo
to coco«-«-oioin~* fMfM stinoo
o oooo«-ooo oo ocMin
O OOOOOOOO OO O «— fM
o oooooooo oo ooo
f. OOOOOOOOO f'f-' OOOOOOOOrOO.-S.KI>±.-<-«-OO — OOO-J
t- O»-tOOOfM«— OOOOfM'-OOO—oOOOOOOOOOO'-OOOOOfM
ooo — oooooooooooooooooooooooooooooo--*-
oooooooooooooooooooooooooooooooooooo
in
o
o
OOOOOOOOOOOOOtf z — i u. — caoo —
co t- zoz u- co — a: u.z3Ooc z
ZZO Z>- >- >- i— cj a: < — i— c/>o.-i— i-za. a. 0.000 ooooaoaoo
O(— -<— • *-* taHUJUIUJ 111 1 1 1 111 111 ... i.i ... ... i . .
0.0.0.0.0.0.0.0.0.0.0.0.0.. t ....
COCOCOCOCOCOCOCOCOCOCOIOtOX X XXXX XXXXXXXXX
o
N-r^N-l^r^-N-N-COOOCOCOOOCOOO 00 COCOOO OOOOOOOO*—
0
o
o
o
o
o
(O
CO
z
o
0
fM
o
o
"
'99999851
CO
UJ
o
CO
o
0
X
o
C-7
-------
ain
U)
c
o
U
U O
4> «-•
3 O
(0 U
10
c
o
U
O
L.
I
Q.
L.
•M
U
U
oc
O)
10
o>
«
tn
o
a
u
c
0)
0
o
o>
oc.
10
z
z
ooo ooooooo o o o o o o
ooo oooo-ooo oooooo
ooo o o o o it o o oooooo
ooo ooooooo oooooo
in CM o tn tn tn tn tn tn tn tn tn tn tn tn
tn tn tn tn m in in in tn tn tn v> tn tn tn tn
XXX X X X X X X X X X X XX X
CMCMKI CM o — K> CNJ CM CM f\i r\j r\j CM fvi JOOOO —
oo«->-ooooooooooooo«-ooooo
ooooooooooooooo—ooooooo
—I _J _J _I_I_J_J_I_I_I _1 _J _1 _J _J _l
<<< <<<<<<< Z
z r> LU «>-
Z O O Z — • _l > U. LU — J <
< a: tu U < — i i— I i— ^ _i
LUO< LUfMMh~O Z C/}(_>^-'
-IZN i^ Z Z — 00 00 — LU
U""1^ OtUOCL LU< O — J O O
>-zz o m z oo a. > a. — < u_ o
a: < 03— o i— z a:
o LU a. a.a.a.zzco°d — — a.
-Jrt «t«r< - OO 00 00 CO Q. _ 1 O Z LU O 13 O —
u
CO
rj
rvi
CO
_
O
o
o
•
o
in
o
o
in
in
X
fM
£
OJ
o
06006??'
.0301197'
C" r-
o »-
0 0
o o
O o- r^-
«— •— o
ooo
ooo
1—
U
<
m
rsi
0
_
o
o
o
in
o
(99999? <
o
0
o
o
ooo
000
ooo
ooo
o N. in
eo o co
ooo
tn tn tn
in tn tn
xxx
CM
tn
X
(M
0
O
CM
O
0
06001031
(0600301-
(06006021
>00102? ]
>00201? I
>99999? 3
ooo
oooo
oooo
oooo
oooo
U
<
CO
M
M
to
—
o
o
o
o
o
0
tn
tn
X
o
o
OJ
o
0
in o
0 0
CO O
0 0
0 0
•0 0
0 0
>00804? !
>00803? 9
0 0
ooo
0 — 0
ooo
ooo
t—
o
<
CO
M
-»
—
0 0
o o
o o
o o
-O 0
•* o
o o
tn tn
tn tn
X X
r\j IM
0 O
o o
(V (M
O O
0 O
f\. f»'
o o
o o
o o
o o
0 0
t— h-
U U
< <
m co
r*J r-j
in
-------
c
i_
4-1
M
C
o
J "o
41 «->
_ C
ID U
t-
a
o
•5
•O
u
§
•-
C
L.
U
u
*->
ee
01
en
0)
o
u
•o
c
in
o
o
en
0)
at
»Or\J-»O KIo~o-oo-oo^o-o-o -o -o o o
oooooo ooooooooo oo o o
Soo|Soo5oS:SS5o8:S5ooo^3?:g:5ooN:
oooooooooooooooooooooooooooo
oooooooooooooooooooooooooooo
oooooooooooooooooooooooooooo
U UUUU CJ UCJ CJUOUUCJ
M MMMM M MM MMMMMM
M MMMM M MM MMMMMM
CM fMtvroro Ki foro KifnMMM^
C-9
-------
a
c.
u O
u
t)
— 10
o
o
CD
-
o o o o
o o o o
O O 0 O
in
0
ro
o
VCT 01
\CT : SEC ZINC/GAL AL
\CT SEC. LEAO/KETT 05
\CT/LAER Al
CO CO CO CO
r\i fo ~* in
O O O O
O o O O
O O O O
in in o o
O o o o
O O O O
0 0 «- 0
CO CO CO CO
CO (0 CO CO
X X X X
CO 00 CO CO
o o ** o
00 CO 00 CO
? 30400209? 30400215? 1
?
? 1
? 30400104? 30400199? 1
o £ P: o
0000
o o o o
o o o o
o
\CT/LAER Al
4CT/LAER (Secondar Al
\ER 02
\ER for Secondary AL
co m _i _i
sfi r«* on o
O O
0 0
0 0
in -0
0 0
CO CO
CO CO
X X
O CO
co r—
0 CO
00 r-
0 0
•- o
o o
o o
o
CM
0
\ER/Primary Lead S AL
5PS 01
-1 Z
o «—
OOOOOOO OOOOOOOOOOOOOOOOOOOOO
o o o o o o o ocoooooooooooooooooooo
OOOOOOO OOOOOOOOOOOOOOOOOOOOO
in in o co o o r- ocoo o >•- o s.oo-ooins.oomKi-o>*orv-in~»oo-o-co
OOO OOO OOOOOOO--OOOOOOOOOOOOO
CO CO CO CO CO CO CO COCOCOCOCOCOCOCOCOCOCOCOCOCOCOCOCOCOCOCOCO
CO CO CO CO CO CO CO COCOCOCOCOCOCOCOCOCOCOtOCOCOCOCOCOCOCOCOCO
xxxxxxx xxxxacxxxxxxxxxxxxxxxx
CO O 00 00 O ~ oooN.K.N.coeocooocoooe9oocON.r>.r>-cooococo
ro o «^'
v» -
0 O O
ro ro ro
f\j o f\i r*- c*. c»
O O o •— f>. O
-» -* KI
~ o ui ui
>• ui — i ui CM ae
Of > _1 C£ Z
< O < O O Z
o i— i— z a: •* o Q O
Z COUI^3COUJUIUI»~
o ozz^—jcococo
CJ O "-CO — — • —
Ul O Z Z CO .>>>
CO 3O3. O
«». o
0 0
o
a
CO
0
o
o
o
o
0
CO
CO
X
o
CO
in
o
o
o
o
o
o
a.
CO
o
o
o
o
ro
CO
o
CO
CO
X
o
CO
o
CO
o
? 10200402? 1
o
0
o
o
o
a.
on
CM
o o
o o
o o
o o
in oo
r>- m
0 O
CO CO
CO CO
X X
0 0
CO 00
0 O
CO CO
o- o
_1 CO
o o
O 0
0 0
0 0
o o
a. a.
o
0
0
o
o
o
CO
CO
X
o
CO
0
§
_J
o
0
o
o
o
a.
00
in
o o
o o
O 0
in rvi
-* KI
>O IM
0 O
CO CO
CO CO
X X
o o
oo co
0 0-
o o
CO 00
o o
CO -i
«— fw
O 0
0 0
0 0
O 0
o o
a. a.
CO CO
•o ^*-
C-10
-------
c
ID
!*•» B)
O O
— ' U)
.a u
§
? 30300922? 30300808? 1
0
o
0
o
ro
o
ro
0
o
CO
0
at
a.
CO
co
o o
0 0
0 ro
«- o
o o
N-0
•o «-
oo
CO CO
CO CO
X X
P.P.
OOP.
0 P.
ao P-
o o
0 0
0 O
-» in
O 0
ro io
1 SEC. LEAD 0
\1K POLLUTIO 0
<
—i -j
o. o.
CO CO
o o
v* in
o o o o
o o o o
o o o o
o o o o
o o O in
09 9- O 09
0000
o o o o
CO CO CO CO
CO CO CO CO
X X X X
p. «- «- O
p. N. p. 00
t>- «- •- o
P. p. P- 00
o
in m
0 O
o o
O 0
in o
t> ooooo«-«-
oop^ooeoooeoN-N.
§£§§§§?:£
? 306011??? 306012??? 1
' 1
? 1
? 1
' 1
? T
? 1
' 1
p.. OOOOOOO
•-0.-OOOOO
oooooooo
oooooooo
rorororororororo
o
oooooooo
vr -!• cj cj o «— «-
UJIAJ (UOOO^*^
>_i— >uiuiuia: (-
cK3 3 at oc ot ^ re
3 at at a. a.
(J at at at
' — — — in in
co ••
-------
C
5
u
w
at
C
O
CJ
~» in
J ~o
J O
ID CJ
10
O
•a
•a
u
c
o.
L.
U
L.
ee
a>
L.
(0
o>
tA
O
•M
CJ
T?
U)
o
c
0
O)
oe
01
z
o
ooooooooooo
ooooooooooo
ooooooooooo
ooininOininoOin>o
OOOOOOOOOOO —
ooooooooooo
r*^OOOOO^OOO
tococococococococococn
cocococncocncncocncnen
XXXXXXXXXXX
coeooooococooooococor-
eoeoeoeoooeoooeocoeorv
ooooooooooo
? 1
? 1
? 1
? 1
? 1
? 30400209? 30400215? 1
? 1
? 1
? 30400104? 30400199? 1
? 1
? 1
£ £ S 5 £ o £ £ S o 5
ooooooooo — o
ooooooooooo
ooooooooooo
in •*
O 0
ro -» ru
o o o
a. _i »— i. co
< H- ID >.
O) o UJ 'O c. 'O
c >-. ^ c mm
••- CJ ^ O "DO)
<-> Z O 0 C _J
— — < .
E -i vx eu c.
CO O CO (D
uj • of ee at E
E GOCJUJUJUJ C_ • —
3 ui < < < o (-
c • • eo _i _j _i •*- a.
ooi>-in>-oouiroo
OO— OKO N- O O O O ut r- O O in K» -O -* O N. in >» O O O CO
oo ooo ooooooo — ooooooooooooo
cocococococo wcoencocococoeococococoGocococncococoenen
xxxxxx xxxxxxxxxxxxxxxxxxxxx
§00 CO O ~O fM —-*v»forO-OO-»OOOOOOOor^S-N.O
r^i^-oocON. Oaoh-r^^ooeoeaaoooeocoeooococooooooooooo
O O CO O OOOOOOO-O-O-OOOOO
co co t^ eo oo K. oeotvK-K.aoaooocococoooeoeoN.rvN.eoeocoeo
oooooo ooooooooooooooooooooo
(\. f». (S.
ro o r-
O — P-
o o <^
o o o
•* <» r^
o o o
ro ro ro
r*° in ^. f*. fv. c^-
IM O IM h<. r- C^.
OOO— l»- O
•* »» ro cu ro o
o o o o o o
o o o o o o
•*•*!/»•* -O -O
o o o o o o
ro ro ro ro ro ro
^f (5 C3 ^* *°° C3 O 00 C3 o* ^** C? C9 ^* ^* ^* £3 ^3 O C3 (^* C3 C3 CO f^> ^* C3 ^<
ooorviooooooooooooooooooooo — or-
oooooooooooooooooooooooooooo
oooooooooooooooooooooooooooo
UJ U UJ
-I t-> Z3 >
— o uj uj
3^ UJ ' iti f\J m*
at > -i ae z
< O > >
CO 3O3
-------
a
L.
*J
(0
O
U
O
01 **
S o
CD U
ra
c
o
u
C
V
o.
u
t.
ee
at
f_
(0
O)
eu
CD
z
*
*
(0
o
(D
U
C
01
0
c
o
0)
oe
eu
z
o
z
oooooooooooooooooo oo ooooooooooo
inooooooooooooooooo oo omooooooooo
COOKIOOOOOOOOOOOOOOO OO OfMOOOOinCOOOO
oooooomooinooos-oooo oo oo«--0ocoooooo
oS.OflOOOooo«-o-*coincoinino-o Klin o«O>OOOooooo«»
h-CnCOCO-«— "-oooooooooooo oo o-ooooo — «-(MOO
coeor<-N.rvr»eocoooa3cooocoooeoeoeoeo eo co eoN-coooeoooi^^oOO
OOf-K»-^-OOOOOOOOOOOO OO O-OOOOOr-^fMOO
cocoK-|vf.l-~cocoo3cococoooeoo3eocoeo 03 co o3S.ooo3cocof-.N.oo-O
o oo o o KI o «— o
O OO OOOOO O
O OO OOOOO O
CM o c*. r». Kl >»• c*« €*• (*• f*- **
(M *~ ^* o* CO O c*> c*- t*. o* C9
O OOt»-OO»-O'— OO
O OOOOOOOO OO
Kl ^-KICM>»-*Klin>>>?>>>> > > > — i Ul UJ Ul Of <- X >—
T- _i_jrMinzzz3=)3Z)3r)=il):) =)l) §5!jCK0'0:0'2iten — uj
• •OO a.o.a.uejuucjuuuu uu <-> r> &.&.&. uj i— Q
oua.a.ujuj— j*j_j os M**M—*inm>— uJUj
as •— •— itaMtM^^^^jujuji^jujuji^j^i^j ujiij uj£3EujujuJUJujen2Ec>s
«tNxo:-OUO.
0<<4<of*- coe>o*— CMrON*in*oh-co
^>*inininininin in minin-o-o-o^o-o-o -0*0 >o>or^s.N.r*-r^.N-^r^^
o
0
o
o
o
in
o
CO
CO
X
o
o
§
99999957? 99999958? 1
in
in
o
O
O
a
UJ
>—
o
z>
u.
O.
en
X
o
o
o
o
o
o
o
CO
CO
X
o
CO
0
eo
03
o
o
o
o
- NESHAP COMFORT AL
X
o
o
o
o
o
in
o
o
CO
CO
X
Kl
0
5
99999811? 1
o
o
a.
oi
X
ej
a.
0 0
0 0
0 0
0 0
4
a.
u
z
Z)
z
a.
<
z
en
•z
X
~»
o
0
o
o
0
o
o
CO
CO
X
0
o
5
o
' 50200505? 1
in
O
o
o
o
r*- in
c». o
•o in
0 0
0 0
0 0
•— ro
C-13
-------
4)
u_
4-«
C
•^
(.
•M
0)
C
o
u
-* in
0 O
u
« w
3 O
a u
•—
PM Additioi
u
«J
C
41
O.
t.
CJ
<
L.
<*
e»
u
t—
O)
CO
(A
o
*V
Q
U
C
U)
a
c
o
en
oaoc>r»CMOO
O OOOOOOOOOOO
o oooooooooo
co cococococococacococo
co cococococococacococo
X XXXXXXXXXX
CM OJ PJ PJ CM CM PJ CM CM PJ PJ
CM (MCMPJCMPJPJCMCMCMCM
P» P» P» P« P" P' P' P' PH P>» P» Pl«
roro-*p-p-p-»-»- ro oo ro
O O O P* P" P' O O O «•" O* O
in'OOP-r-p-^r- \f\ CM^ ro
OOOP-P-P-OO O OO O
OOOP-P-P-OO O OO O
cMCMCMroop-rofM CM •* in m
oooooooo o oo o
^^-^-*— foro*— •«* ro roro ro
CMCMCO«-P'P'P-»- CMIO r--* ^^CM
OOOOP-P-P-O OO ^O OO
in*Or*CMp-p'p*in in*™ CM^- roin
OOO«-P'P-P-O OO OO OO
OOOOP-P-P-O OO OO OO
cMCMCMCMp-p-p.ro CM>O «»in min
o o o o t>. ro P" o oo oo oo
•—«—«— ^-cMroin^- roro roro roro
o o o o p« p- P" o o o P- o v- o p. o o o o P- P"
in"Op^<-p'p«p'**om'OOPjcMCMtoinrorooa«—
OOO«-P-P-P-OOOOOOOOOOOOOO
OOOOt^. r-r-OOOOOOOOOOOOOO
OOOOOIMfv. OOOOOOOOOOOOOO
< OC>OkOOOO'v-'-
C-14
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
1. REPORT NO. 2.
EPA-450/2-89-012b
4. TITLE AND SUBTITLE
Analysis of Air Toxics Emissions, Exposures, Cancer
Risks, and Controllability in Five Urban Areas
Volume II - Controllability Analysis and Results
7. AUTHOR(S)
Jim Wilson, Bob Coleman, and Erica Laich of E.H.Pechan
Roger Powell, EPA, OAQPS
9. PERFORMING ORGANIZATION NAME AND ADDRESS
E.H. Pechan and Associates, Inc.
5537 Hempstead Way
Springfield, VA 22151
12. SPONSORING AGENCY NAME AND ADDRESS
Noncriteria Pollutant Programs Branch
Air Quality Management Division
Office of Air Quality Planning and Standards
3. RECIPIENT'S ACCESSION NO.
S. REPORT DATE
Aoril. 1990
6. PERFORMING ORGANIZATION CODE
OAQPS
8. PERFORMING ORGANIZATION REPORT ">
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVEREC
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report (Volume II) is the second phase of a study to define the urban
air toxics problem and to discern what combination of control measures can best
be employed to mitigate the problem. Volume I of this study documented the base
year analysis (nominally the year 1980), involving dispersion modeling of
emissions data for 25 carcinogenic air toxics in five U.S. urban areas and a
subsequent assessment of estimated aggregate cancer incidence. This Volume II
report applies various control strategies and analyzes the resulting reduction
in aggregate cancer incidence that would occur between 1980 and 1995.
Control scenarios consisted of (1) efforts that were currently underway to
reduce air toxics emissions at the time of this study, (2) efforts that were
expected to occur by 1995, mainly national standards that were under development,
and (3) a series of selected more rigorous controls. Current rules would reduce
cancer incidence by 27 percent, expected rules would gain another 20 percent, and
selected additional controls would add another 13 percent reduction. Reduction
would be almost equally divided between volatile organic compound and particulate
emissions and approximately half of the incidence reduction would come from
mobile source control.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Air Toxics
Cancer incidence, air toxics
Controllability, air toxics
Hazardous air pollutants
Mitigation, air toxics
Risk assessment, air toxics
Urban air toxics (urban soup)
18. DISTRIBUTION STATEMENT
b. IDENTIFIERS/OPEN ENDED TERMS
19. SECURITY CLASS (This Report)
20. SfcCURITY CLASS (This page)
c. COSATl Held/Group
21 NO. OF PAGES
107
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
------- |