COST AND ECONOMIC IMPACT ASSESSMENT FOR ALTERNATIVE
LEVELS OF THE NATIONAL AMBIENT AIR QUALITY STANDARD
FOR OZONE
Economic Analysis Branch
Strategies and Air Standards Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
DRAFT
June 1978
-------
COST AND ECONOMIC IMPACT ASSESSMENT FOR ALTERNATIVE
LEVELS OF THE NATIONAL AMBIENT AIR QUALITY STANDARD
FOR OZONE
Economic Analysis Branch
Strategies and Air Standards Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
DRAFT
June 1978
-------
PREFACE
In accordance with the provisions of Sections 108 and 109 of the
Clean Air Act as amended, the Environmental Protection Agency has
conducted a review of the criteria upon which the existing primary and
secondary photochemical oxidant standards are based. The Act specifically
requires that National Ambient Air Quality Standards be based solely
on scientific criteria relating to the level that should be attained
to adequately protect public health and welfare. Based on the wording
of the Act and its legislative history, EPA interprets the Act as excluding
any consideration of the costs of achieving those standards or the existence
of technology to bring about the needed reductions in emissions. However,
in compliance with the requirements of Executive Orders 11821 and 11949
and OMB circular A-107 and with the provisions of the recently issued
Executive Order 12044 for rulemaking proceedings which are currently
pending, EPA has prepared an assessment of the potential cost and economic
impacts associated with efforts to attain the proposed standard as well as
alternative levels of the standard. This document presents the results
of this assessment.
The purpose of the analysis contained herein is to estimate the
relative ranges of national control costs for alternative levels of the
ozone standard. In addition, in order to compare the relative implications
of alternative standards, the range in the number of Air Quality Control
Regions (AQCRs) which might be expected to attain the alternative standards
given various assumptions is also indicated. Because of the many uncer-
:e nties in projecting emission levels and air quality levels and in
-------
determining effective control strategies throughout the nation, it is
important to fully recognize that the results of the analysis should
be viewed only as general guidance which provides an indication of
relative differences in the attainment picture and the associated costs
between alternative levels of the standard. The analysis cannot be
used to precisely determine how many or which specific AQCRs will attain
a given ozone standard through particular control strategies. Rather,
attainment status and control requirements for attainment will have tc
be determined for each geographical area basod on the unique conditions
that sre inherent for that area. Likewise, this analysis cannot ascertain
with a great degree of precision the costs of control strategies that
will be required for all areas of the country to attain alternative
standards. Since thc> actual control costs will be extremely variable,
this analysis is useful in only presenting the relative implications
for costs between alternative levels of the standard.
ii
-------
TABLE OF CONTENTS
Page
Chapter 1.0 Introduction and Summary 1-1
1.1 Approach and Methodology 1-1
1.2 Limitations of the Analysis 1-4
1.3 Major Findings 1-10
Chapter 2.0 Baseline Air Quality and Allowable Emission
Levels 2-1
2.1 Existing Ambient Air Quality Levels 2-1
2.2 Allowable Emission Levels 2-2
2.3 References 2-8
Chapter 3.0 Projection of Emissions in 1987 and Reasonable
Levels of Control Achievable 3-1
3.1 Motor Vehicle Emissions in 1987 3-2
3.2 Uncontrolled Emissions from Stationary Sources . . . 3-3
3.3 Emission Projections with Controls on New
Stationary Sources 3-7
3.4 Reasonably Available Control Measures for
Existing Sources 3-8
3.5 Attainment Status and Additional Reduction
Required 3-14
3.6 References 3-19
Chapter 4.0 National Costs of Control for Alternative
Levels of the Standard 4-1
4.1 Costs of the Federal Motor Vehicle Control
Program 4-1
4.2 Costs for New and Modified Stationary Sources . . . 4-2
4.3 Costs of Applying RACT for Alternative Standards . . 4-3
-------
4.4 Summary of Costs of FMVCP, New Source Control,
and RACT on Existing Sources 4-8
4.5 Estimated Cost of Attainment 4-11
4.6 References t . 4-16
Chapter 5.0 Economic Impact of Reasonably Available
Control Measures on Selected Industries 5-1
5.1 Introduction 5-1
5.2 Petroleum Refining 5-2
5.3 Retail Gasoline Service Stations 5-5
5.4 Gasoline Bulk Plants 5-11
5.5 Automobile Assembly Plants ..... 5-16
5.6 Metal Furniture Industry 5-20
5.7 References 5-25
Appendix A Ozone Design Values for 90 Air Quality Control
Regions A-l
Appendix B Mobile Source Emission Factors B-l
Appendix C Analysis of Costs for Hydrocarbon Control
Measures * .... C-l
Appendix D Estimated Reductions in VOC Emissions Due to
RACT for Stationary Sources D-1
-------
1.0 INTRODUCTION AND SUMMARY
The Clean Air Act, as amended in 1977, requires the Administrato-
of the Environmental Protection Agency to periodically review the basis
of the ambient air quality standards. The National Ambient Air Quality
Standard for photochemical oxidants has been reviewed and a new form of
the standard has been proposed. As part of this review procedure, this
report presents the results of an analysis of the potential impact of
feasible changes in the standard on national costs of control and the
attainment status for various areas of the country. An analysis of the
potential socio-economic impact of the alternative standards has not
been conducted for every affected industry and region of the country,
although an assessment of the impact of control costs on selected industries
is presented to give an indication of the magnitude of the impacts.
1.1 APPROACH AND METHODOLOGY
This report includes an analysis of 90 Air Quality Control Regions
(AQCRs) which currently exhibit ambient ozone concentrations in excess of
the current photochemical oxidant standard (.08 ppm hourly average not to
be exceeded more than once per year). For each AQCR this analysis estimates
potential emissions in 1987 and potential emission reductions achievable
with the Federal Motor Vehicle Control Program (FMVCP), new and modified
source control, application of reasonably available control technology
(RACT) on existing stationary sources and further motor vehicle controls
through transportation control plans. Based on the projected emission
reductions, control costs are estimated for applying technology in an
attempt to attain alternative standard levels. While the analysis considers
each AQCR separately, the results are presented in aggregate form for ell
90 AQCRs instead of each individual AQCR.
1-1
-------
This analysis uses the 1975 emissions inventory from NEDS as the
baseVine emissions for each of the AQCRs. From this base 1987 emissions
are projected based on statutory automotive emissions standards and
reasonable ranges of assumptions for growth in vehicle miles trpveled
(VMT), growth in production for new sources, and retirement rates for
existlr.g sources. In addition, assumptions relating to the control of
new and modified sources are made in order to determine emissions with
i
only new source and statutory motor vehicle controls.
In order to calculate the emission reduction required, if anv, for
existing sources to ensure that alternative standard levels are met, the
following approach is employed. For each standard level, the maximum
allowrble emission levels are calculated for each AQCR based on baseline
ambient concentrations, 1975 emissions of non-methane hydrocarbons (NMHC).
the level of the alternative standard, background ozone concentrations in
urbanized areas of .02 ppm, and assumed relationships between hydrocarbon
emissions and ambient ozone concentrations according to alternative
modeling approaches. These two approaches are the linear rollback model
and the Empirical Kinetic Modeling Approach (EKMA).
Once the allowable emissions have been calculated for each AQCR,
they are compared with projected 1987 emissions levels taking motor vehicle
tailpipe controls and new stationary source controls as the baseline. If
additional reductions are required, available control measures are then placed
! . -.
on existing stationary sources that have not been replaced as well as in-use
vehicle inspection/maintenance proarams or VMT reduction plans for control
of mobile source emissions. RACT is assumed only for those specific
stationary source categories which have been identified and for which
1-2
-------
estimates of control technology and efficiencies have been made. While
these sources represent the primary sources of volatile organic compounds
(VOC), other sources exist, primarily small solvent evaporation sources,
which have not been identified by EPA. As an extension of the analysis,
reasonable levels of control for these miscellaneous solvent sources are
assumed in order to determine their effect on the results of the analysis.
In many instances, control of identified sources and reasonable transpor-
tation controls will not result in the required emission reductions,
meaning that additional control measures will have to be applied in
order to meet the alternative standards. This study does not consider
additional control measures, rather only the additional emission reductions
needed are indicated.
Once the control measures have been identified, the costs of the measures
for the specific sources are then analyzed and estimated. The average cost
per ton of emissions controlled is established for general categories
of sources and combined with estimates of emission reductions achievable
for the 90 AQCRs in order to obtain total annual costs of control in 1987.
Costs are included for the FMVCP*. new source control, and application of
RACT for identifiable sources to the extent required for each of the alter-
native levels of the standard. For areas that need additional control
beyond the identified control measures, a rough estimate of the additional
cost to attain the alternative standard is made which will reflect the
magnitude of differences in total costs of the alternative levels of the
standard. The costs presented in this analysis represent direct annualized
costs of control which will be incurred in 1987, when it is assumed that
*lncljdes costs for both HC and CO control sirce it is not possible to
oV.cc'.te :csts between pollutarrs.
1-3
-------
the control measures will be applied fully to the applicable sources.
No estimate of total capital costs is made. In addition, there is no
consideration of the secondary costs (i.e., costs beyond those encountered
by the source being controlled) or benefits associated with the control measures.
1.2 LIMITATIONS OF THE ANALYSIS
In any analysis of this type many simplifying assumptions have to
be made because of the uncertainties that surround the problem. The choice
of assumptions, analytical tools, data bases, and approach will profoundly
affect the conclusions of the analysis. This analysis attempts to place
reasonable ranges on the assumptions in order to estimate the range of
national costs for controls necessary to attain a given ozone level.
Since the assumptions are so crucial, this section outlines the basic
assumptions employed in the analysis and the limitations the choice of
these assumptions place upon the results of the analysis. Because of
the uncertainties, the results of this analysis are presented as ranges
which serve as realistic bounds for the conclusions.
First of all, this analysis considers 90 AQCRs for which validated
data on ambient concentrations exist which indicate violations of the
current standard or which contain urbanized areas with populations
greater than 200,000 which have been designated non-attainment by EPA.
Undoubtedly, there are many other smaller urban areas as well as rural areas
which currently also experience violations of the standard. However, these
latter areas are not considered in the analysis since validated data have not
been compiled for these areas. While this exclusion will understate the
costs somewhat, it is not expected to be significant since these areas
probably do not significantly exceed the standard and since the AQCRs
which are considered represent well over 60 percent of the nation's
population.
1-4
-------
In order to calculate the emission reductions required in each ,.f
the AQCRs in 1987, assumptions must be made concerning the emissions
inventory, baseline air quality values, growth rates, and the modeling
technique to be used.
This analysis uses the preliminary 1975 emissions inventory from
the NEDS Emission Summary Report, which is the only source of aggregate
information on emission data for all 90 AQCRs. No attempt was made to
verify the emissions inventory with data from state and local agencies,
whose inventories are in some cases at variance with NEDS. Although
there are some inconsistencies between NEDS and local inventories, NEDS none-
theless provides a consistent procedure for conducting the inventories
throughout each AQCR and is believed to be the best available source of
emissions data for national assessments.
Several reasons exist for the differences between NEDS and local
inventories. First, the NEDS inventories have not yet been updated
with the latest mobile source emission factors which have recently
been developed by EPA's Office of Transportation and Land Use Policy.
A review of these latest emission factors reveals that NEDS under-
estimates the mobile source emissions by 10-15 percent. Another appre-
ciable difference between NEDS inventories and local inventories is
that NEDS includes significant emission contributions from miscellaneous
sources of organic solvent applications. Local inventories do not usually
1-5
-------
include the small miscellaneous sources whose emissions are relatively
easy to calculate nationally based on nations! production statistics but
are extremely difficult to calculate on a local basis. Much uncertainty
exists as to the exact composition of sources in this ill-defined
miscellaneous category. EPA currently has underway an extensive study
to better define, classify and locate the sources of solvent evaporation.
The design ozone values used in this analysis represent only an
approximation of the ambient ozone concentration that on the average
will be exceeded once per year in each AQCR. These values were obtained
from an analysis of air quality data for the years 1975 to 1977 using
methods described in Guideline for Interpretation of Ozone Air Quality
Standards (OAQPS No. 1.2-108). Analysis of air quality data indicates
that this value should fall between the third and fourth highest hourly
averages.
Growth rates are crucial parameters in determining the overall
emission reduction which will be required in 1987. This analysis uses
ranges of national average rates of growth in industrial production for
various classes of industrial sources as well as ranges of national average
rates of growth in vehicle miles travelled. This approach of using national
growth rates simplifies the growth projections but introduces some
distortions for some local areas. First of all, there is the implicit
assumption that every AQCR has the same mix of sources as the nationwide
1-6
-------
mix on which the average growth rates are based. Secondly, all areas of
the country will not experience the same rate of growth. For example,
many highly industrialized areas with ozone problems may experience
lower rates of growth than less industrialized areas which are attracting
new industries. Ideally, for each AQCR, the analysis should use regional
growth rates which reflect local economic conditions and the mix of
sources in an area. While such an approach has not been used in this
analysis, EPA is considering modifying the analysis to include regional
growth rates.
Another variable in the analysis is the modeling techniques employed
to estimate the allowable emissions for areas to attain alternative
ozone standards. Depending on the technique used, wide variations can
result in allowable emission levels and the concomitant reduction in
projected emissions that is required. This analysis includes results from
application of the linear rollback model as well as the Empirical Kinetic
Modeling Approach (EKMA). A judgement as to which modeling approach is
more valid or accurate is not made.
The emission reductions that can be achieved with application of
various controls measures is still another source of variability in the
analysis. The emissions inventory is segmented according to eight broad
categories of industrial sources and three categories of mobile sources.
Each of the stationary source categories is composed of many diverse
industrial sources which are subject to varying levels of control. Thus,
a weighted average control level has to be estimated for each broad
category based on the assumed control levels and overall emissions for
-.; i-d-vidual source category. When this weighted average emission
1-7
-------
potential is applied to each broad category in each AQCR, it is again
assumed that the mix of sources in each AQCR is the same as the nation-
wide mix. Thus, this assumption could well affect the results of the
analysis for any AQCR.
Furthermore, RACT has not been identified or estimated with any
certainty for all sources of hydrocarbons. These sources are primarily
in the solvent evaporation category, where the specific sources have not
been identified, control technology and efficiency have not been assessed,
and whether or not these sources are amenable to control in the first place
has not been determined. Without this information currently available
on the remaining sources, further controls are subject to speculation.
As an extension of the analysis, this analysis assumes control of these
sources in order to determine their impact on control costs and attainment
projections.
The approach used in defining control costs is also likely to
introduce another area of variability in the analysis. For each of
the many sources of VOC, the cost per ton of pollutant removed was
determined based on estimated capital and operating costs for "typical"
model facilities. Costs for actual facilities are likely to vary
considerably, both on low and high sides, due to individual circumstances
of the plants. However, the typical plant costs are considered to be
representative of the industry as a whole. Once again, to aggregate
the control costs for the broad categories of sources, the costs were
weighted based on the nationwide mix of sources, which is not represen-
tative for all AQCRs.
1-8
-------
The range of assumptions made in this analysis in regard to these
points appear to be reasonable in terms of estimating costs of control
measures for hydrocarbon control. While the intended purpose of this
analysis is an estimation of the range of national costs, the approach
allows the estimation of the number of AQCRs which will attain alternative
levels of the standard, although this is only a secondary result which
was not considered in the original design of the study. Since the
approach of this study can possibly introduce more errors for an individual
AQCR, the estimation of the number of AQCRs which will attain alternative
standard levels is particularly sensitive to the choice of assumptions.
The estimates of attainment status included in this analysis are only
approximate and are valid only for the assumptions and approach used;
thus, the estimates should be viewed with full recognition of the limitations
of this analysis. Any choice of different assumptions or refinement of
the assumptions used will naturally lead to different results. Refinement
of the assumptions to a fine-tuned basis was beyond the scope of this
cost study and the attainment numbers are presented only to indicate the
relative differences between alternative levels of the ozone standard.
Finally, this analysis does not include a rigorous economic impact
assessment on the affected industries or the impact of growth restric-
tions on the economies of affected urbanized areas. An analysis of the
economic impact of the revised standard on the numerous industries
1-9
-------
affeset€di ts" riot possible to' eomp-'lete iin- tne tfine frame av.arin-.ab:Te. for
analyses.- Howe'ver,- EPA Has conducted econom-ie impact s>tudtes: for th&
major emission sources, though these constitute on-Ty a- portion of the
total number of sources affected. Economic analyses of all affected
industries are being initiated and significant results are expected
within two to three years. Furthermore, the Clean Air Act does not
permit the consideration of costs and economic impact in the setting of
national ambient air quality standards. The incomplete character of the
economic analysis for all affected sources and areas does not mitigate
EPA's legal burden to propose the standard.
1.3 MAJOR FINDINGS
While the results of this analysis are sensitive to the assumptions
used, several conclusions are evident which indicate the relative
differences between alternative levels of the standard and the implications
for control. This section summarizes the major conclusions of the analysis,
Table 1-1 summarizes the results.
1.3.1 Attainment Status for Major AQCRs
Many areas of the country will not be able to attain a .08 ppir.
level based on the statistical form of the standard by 1987. Based
on the results from the linear rollback model, anywhere from one-third
to two-thirds of the major AQCRs in the country will require stationary
control measures beyond NSPS and RACT as well as additional transpor-
tation GbhtrSl measures in order to approach the standard. For a .10
1-10
-------
Table 1-1. SUMMARY OF ESTIMATED CONTROL COSTS AND ATTAINMENT STATUS
IN 1987 FOR VARIOUS LEVELS OF THE OZONE STANDARD
Level of
Standard
.08 ppm
.10 ppm
.!
-------
standard level, all but 10 to 30 AQCRs may achieve the standard by 1987
with the control measures identified in this analysis. The lower estimate
reflects a low growth scenario with extensive RACT measures placed on
all significant stationary source categories, including the solvent
evaporation source categories which have yet to be assessed in detail
to determine their amenability to control. On the other hand, the higher
estimate is based on a higher growth scenario without control of those
solvent evaporation categories which have not been assessed. Finally,
at a .12 standard level, all but five AQCRs could attain the standard
under the low growth/high control scenario, while all but 13 could
attain under the less optimistic case.
The application of EKMA in the analysis does not lead to as optimistic
results since EKMA, with the assumptions used in this analysis, tends to
require significantly more control than rollback. The results from EKMA
could vary appreciably for any urbanized area depending primarily upon
the ratio of ambient non-methane hydrocarbons to nitrogen oxides. Thus,
the results from EKMA would be more accurate if NMHC/NOx ratios were
available for every AQCR. Based on a typical nationwide ratio of 9.5:1, EKMA
predicts that only three or four AQCRs could attain a .08 ppm standard
by 1987. Even at .10 ppm, less than half of the AQCRs could attain the
standard even assuming low growth and high control, while at .12 ppm
more than 15 AQCRS would still be in violation in 1987 without applying
further controls.
1-12
-------
1.3.2 Implications for Further Control
This analysis confirms that, most areas of the country will have to
apply extensive control measures in order to attain any of the reasonable
levels of the standard. Areas will have to place controls on new and
modified sources, aggresively apply RACT to existing sources, institute
effective inspection/maintenance programs for in-use vehicles, and
implement various degrees of measures to reduce vehicle miles traveled
in the urbanized areas. While higher levels of the standard will mean
that some areas will come into attainment automatically or will require
less emission reduction, most major urbanized areas will not be able to
relax control efforts. Because ozone is such a pervasive and intense air
pollution problem, any modest relaxation of the standard will not result
in any significant changes to existing or planned control strategies
for these areas. The major urban areas will need to continue applying
all available control measures.
The results of this analysis also indicate the need for the identi-
fication of miscellaneous solvent uses as well as miscellaneous industrial
processes and the assessment of the applicability of control measures
for these sources. Current data indicate that these sources represent a
significant source of emissions which will need to be controlled if many
areas are to attain the alternative levels of the standard. For instance,
at a .08 ppm level, control of these sources would result in an additional
25 percent of the 90 AQCRs attaining the standard under the low growth
scenario using linear rollback.
1-13
-------
1.3.3 Costs of Control
As Table 1-1 indicates, the annual cost in 1987 of applying reasonable,
identified control measures will range from $4.0 to $6.0 billion for all
levels of the standard under consideration. The cost difference between
alternative standards is not as great as might be expected for two basic
reasons. First, the vast majority of the total annual costs ($3.7 to
$4.3 billion) result from the FMVCP and new source controls, which are
assumed to be the same regardless of the level of the standard. Secondly,
the RACT control costs do not differ greatly from alternative standards
since many areas will have to apply full RACT regardless of the level of
the standard.
Since many AQCRs will not attain alternative levels of the standard
by applying the reasonably available control measures considered in this
analysis, additional control measures will be needed in some areas in
order to reduce emissions even further so that the standards can be
attained. These measures could include restrictive transportation
control measures that significantly reduce vehicle miles traveled in
urban areas, control of stationary sources for which RACT has not been
identified or defined, tighter controls on new and existing sources
other than those achievable with RACT and restrictions on growth. While
the costs of such measures have not been estimated, on the whole they
arc believed to be more costly than current measures. Even though the
costs are not precisely known, it is still useful to estimate the cost
of attainment in some manner in order to better indicate the cost differences
between alternative standards. To do this, a cost estimate of $1,000 to
1-14
-------
$1,500 per ton of emissions controlled is assumed for the additional
emissions reductions required for each AQCR to attain the alternative
standards. Table 1-1 summarizes the estimated costs of total attainment
for the alternative standards. The cost differential between standard
levels is greater, with the 1987 annual costs for the alternative standards
under consideration ranging from $4.0 to $8.0 billion using the results
from the linear rollback model. Based on the results from EKMA, total
attainment for a .08 ppm standard could result in annual costs ranging
from $9.0 to $12.5 billion.
1-15
-------
2.0 BASELINE AIR QUALITY AND ALLOWABLE EMISSION LEVELS
This section examines the current ambient levels of air quality in
non-attainment areas in order to establish a bdbis for determining the
levels of emission control required to meet alternative levels of the
ozone standard. Based on these air quality levels, techniques for
establishing allowable emission levels in individual areas are discussed.
2.1 EXISTING AMBIENT AIR QUALITY LEVELS
Many areas of the country are experiencing levels of ozone and
oxidant concentrations well above the present standard. In this analysis
90 Air Quality Control Regions (AQCRs) are considered which have validated
data that indicate ambient concentrations at or above the present
standard of .08 ppm. While other areas of the country may currently
violate the standard, validated data were not available for additional
AQCRs. Nonetheless, the AQCRs considered contain all the major urbanized
areas with populations greater than 200,000.
The baseline ozone design values for the 90 AQCRs are based upon a
listing contained in Appendix A. These values reflect an approximation
of the ambient concentration that on the average will be exceeded once
per year in each AQCR. Using analytical methods for the statistical form
of the ozone NAAQS, these values were obtained from an analysis of data
2
for the years 1975 to 1977. Using three years of data the design value
would be expected to fall between the third and fourth highest hourly
averages. In selecting the design values, the fourth highest hourly
average over a three year period was used, unless the difference between
the third and fourth highest values exceeded .01 ppm, in which case
the third and fourth highest values were ave-aged.
2-1
-------
Figure 2-1 shows the distribution of the design air quality
values among the 90 AQCRs for which there are data. About 46 percent of
the areas are above twice the level of the present standard and about
nine percent are three times above the standard. The adoption of alternative
levels of the standard in the range of .08 ppm to .14 ppm would have an
immediate effect of bringing six to 32 of the AQCRs into compliance,
depending on the level of the standard.
2.2 ALLOWABLE EMISSION LEVELS
For each alternative level of the standard, the maximum allowable
emission levels are calculated for each AQCR based on baseline air quality
levels discussed earlier, 1975 total emissions of non-methane hydrocarbons
(NMHC), the level of the alternative standard, background ozone levels of
.02 ppm and assumed relationships between hydrocarbon emissions and
ambient ozone concentrations. The maximum allowable emission levels
indicate, based on the parameters listed, the emissions which will permit
attainment and maintenance of the standard. The 1975 emission levels for
each AQCR were obtained from the NEDS Emission Summary Report (NE204),
corrected to exclude methane.
Two modeling approaches were used to estimate the allowable emission
levels. The first is the linear rollback model which assumes a pro-
portional relationship between hydrocarbon emissions and ambient ozone
concentrations, with appropriate adjustment for natural background levels
of ozone. The relationship is given by the following equation:
Allowable emissions _ (Standard level minus background)
Current emissions ~ (Current ambient concentration minus background
2-2
-------
Figure 2-1. Frequency Distribution of Design Values for 90
Non-Attainment AQCRsa
I\J
I
00
ZQ.
10
/
n
PI5
4
<%
17
13
'/?,'
i
n
a
<.08
..16
1.22 1.24 <.26 £.28 <.30 >.30
Based on Appendix A.
-------
The alternative approach is the Empirical Kinetic Modeling Approach
(EKMA), which employs isopleths based on the results of smog chamber
experiments to relate various coiicenlrations of NMHC and nitrogen oxides
(NOx) to resulting concentrations of ozone. In employing the model
it is necessary to make assumptions as to:
1. the prevailing 6-9 a.m. ratio of ambient NMHC to NOx.
2. the relative degree to which NMHC and NOx emissions will be
controlled, and
3. ozone background levels.
While the NMHC/NOx ratios for various cities will vary widely, it is
beyond the scope of this analysis to determine the appropriate ratios
for every city. However, best estimates of typical 6-9 a.m. ratios, based
on an examination of data from a number of monitoring sites, indicate
a median ratio of 9.5:1, which is used in EKMA for purposes of this
analysis. In addition, it is assumed that NOx emissions remain constant
between 1975 and 1987.
A background level of .02 ppm ozone in urban areas is added in both
modeling approaches. While measurements conducted in remote locations
suggest that natural background ozone is about .04 ppm, simulation
results indicate that the impact of natural ozone on peak hourly ozone
concentrations in urban areas ranges from .01 to .03 ppm, with .02 ppm being
3
most likely.
This analysis presents the results from both modeling techniques
since both are considered viable approaches for relating ambient ozone
to organic compounds and oxides of nitrogen. No judgement is made as
to the relative effectiveness or accuracy of the alternative approaches.
For a full discussion of the technical basis, uses and limitations of the
approaches, consult reference 2.
2-4
-------
In essence, EKMA is a rather complex model that has been primarily
validated against data from smog chambers, which represent a simplification
of the urban atmosphere. In addition, the absolute positions of the
standard isopleths represent an approximation since their position
depends upon a number of underlying assumptions concerning meteorological
conditions and emission patterns.
Rollback, on the other hand, is a rather simple approach that contains
the assumptions that ozone concentrations are proportional to NMHC emissions
and that the amount of organic emission controls needed to attain ambient
ozone levels is independent of the prevailing NMHC/NOx ratio. However,
smog chamber experiments suggest that the lower the ratio, the more
effective the hydrocarbon reduction is in reducing the maximum ozone formed.
Thus, at very low NMHC/NOx ratios, linear rollback may underestimate
the effectiveness of organic controls, while at high ratios estimates may
be overly optimistic. Nonetheless, under the range of NMHC/NOx ratios
believed to prevail in most U.S. cities, rollback appears to be useful
in serving as a lower bound for estimates of hydrocarbon controls needed
LO attain the ozone standard.
Application of the two modeling approaches gives somewhat different
results. Table 2-1 illustrates the comparative control levels required
A
for alternative standard levels utilizing both approaches. With the
exception of cases having very low NMHC/NOx ratios, needed NMHC reductions
estimated with rollback are almost always less than those obtained with
the standard isopleth version of EKMA.
2-5
-------
Table 2-1. PERCENT HC REDUCTIONS REQUIRED TO MEET VARIOUS LEVELS OF
THE STANDARD GIVEN SECOND MAXIMUM OZONE CONCENTRATIONS2
Ozone Design Value (Second Maximum Hourly Ozone Conct^'ration)
spzone
[sicln
STDValue
.10
.14
.06
50
7*. OR
31
44
75
25
77
79
75
79
10
67
70
73
77
60
67
71
30
57
.66
69
71
72
72
74
17
38
50
5R
67
69
71
72 '
35
55
64
68
68
71
0
35
46
57
59
EKMA
Assumptions:
(1) Impact of natural background ozone concentrations on maximum
afternoon ozone levels = .02 ppm ozone.
(2) No transport from upwind cities
(3) Default NMHC/NOx Ratio = 9.5:1.
(4) No NOx Control
Rollback:
EKMA
Rollback
- STD
- BKOi
1 100,
-------
Using both modeling approaches, allowable emission levels were
calculated for each AQCR. These levels were then compared to projected
1987 baseline emission levels in order to determine the emission reduction
that needs to be effected so that all areas will attain the standard
level by 1987. These emission projections and control requirements
are discussed in the next section.
2-7
-------
2.3 REFERENCES FOR CHAPTER 2
1. Neligan, Robert E., Monitoring and Data Analysis Division (MDAD),
OAQPS, EPA, memorandum to Bruce Jordan entitled "Ozone Design Values
for 90 Air Quality Control Regions," May 24, 1978.
2. Environmental Protection Agency, Guideline for Interpretation of
Ozone Air Quality Standards, OAQPS No. 1.2-108, Draft.
3. Environmental Protection Agency, Uses, Limitations and Technical Basis
of Procedures for Quantifying Relationships Between Photochemical
Oxidants and Precursors, EPA-450/2-77-021a, November 1977.
4. Freas, Warren P., MDAD, memorandum to Edward J. Lillis entitled
"Estimates of HC Reductions Required to Meet Selected Ozone Levels
Given Second Maximum Ozone Concentrations." December 4, 1977.
2-8
-------
3.0 PROJECTION OF EMISSIONS IN 1987 AND REASONABLE LEVELS
OF CONTROL ACHIEVABLE
This chapter summarizes the emission projections and emission reductions
achievable by 1987 for the affected AQCRs. Reasonable ranges of growth rates
are applied to 1975 emissions in order to determine 1987 baseline emissions
without additional stationary source and mobile source controls other
than mobile source standards mandated by the Clean Air Act. Stationary
source controls in the form of new source standards and reasonably
available control technology (RACT) for existing sources as well as
additional mobile source controls in the form of inspection/maintenance
programs and transportation control measures are assessed and potential
emission reductions quantified. These projected emission reductions are
then compared with the baseline 1987 emissions and the calculated maximum
allowable emission levels to determine if the available controls result
in attainment of the alternative levels of the standards. If the standards
are not attained, the additional emission reduction required to reach the
allowable emission levels is indicated.
For purposes of this study, RACT is broadly defined as technology
readily available for application in categories of sources which will lead
to adequate levels of control based solely on tecnnical considerations.
Because of unique circumstances with individual sources, the reasonableness
of technology for particular sources will be determined by state and local
agencies based on technical guidance issued by EPA and depending upon
economic and energy feasibility.
3-1
-------
Emission reductions due to new and modified source controls are applied
to the growth in new stationary source emissions over the time period in
addition to the emissions from the replacement of existing sources with new
or modified sources. This analysis uses a reasonable range of national
growth rates which will bound the extreme variability in stationary source
growth throughout regions of the country.
Projections are made for each individual AQCR based on its emission
inventory and national estimates of growth rates and control efficiency.
National totals of baseline emissions and emission reductions are obtained
by summing the totals for the individual AQCRs.
The assumptions and methodology used in this analysis were chosen
specifically to arrive at national emission and cost estimates. For planning
in any specific AQCR, more detailed analyses will be required. The procedures
to be followed for preparing detailed plans are presented in the document,
"Control Strategy Preparation Manual for Photochemical Oxidants."
3.1 MOTOR VEHICLE EMISSIONS IN 1987
Mobile sources include light-duty vehicles, other highway vehicles,
and non-highway vehicles (such as aircraft, barges and vessels, railroads
and earth moving equipment). The Federal Motor Vehicle Control Program (FMVCP)
for highway vehicles will significantly reduce emissions from these sources.
This reduction is counterbalanced by an estimated two to three percent growth
2
per annum in total miles traveled for most areas of the country. For
light-duty vehicles, a total reduction of emissions of 60 to 66 percent is
projected by 1987 taking into account both growth and tailpipe controls.
3-2
-------
(See Appendix B). Similar reductions for other highway vehicles are
expected to be 30 to 40 percent through the FMVCP. These factors are
included in the estimated baseline emission projections for 1987 in
Table 3-1.
Even though the FMVCP will reduce mobile source emissions by 50
to 60 percent, this will not be adequate to offset the growth of stationary
source emissions for most areas. Only in the unique circumstance where
an area with a high proportion of mobile source emissions is close to the
standard will the FMVCP alone permit attainment of the standard. For a
.08 ppm standard level, this will pertain to only one or two AQCRs,
while up to eight AQCRs could be affected with a .10 ppm standard level.3
3.2 UNCONTROLLED EMISSIONS FROM STATIONARY SOURCES
In order to determine the total reduction in emissions required in
1987 for each AQCR to achieve the allowable emission level, growth in the
emissions inventory has to be taken into account. In Table 3-1, uncontrolled
emissions in 1987 are projected for eight stationary source categories in
addition to tne three mobile source categories. These projections
include the existing emission inventory as well as the growth of new
sources. For each of the eight stationary source categories, as segmented
in the NEDS Emission Summary Report, a range of representative growth
rates has been determined and applied to the existing emission inventory
in order to determine the uncontrolled emission levels in 1987. Table
3-2 summarizes the range of growth rates used in this analysis for each
4
source category.
3-3
-------
Table 3-1.
PROJECTED NON-METHANE HYDROCARBON (NMHC) EMISSIONS
IN 1987 FOR 90 HQN-ATTAIIWENT AQCRs.1
(Millions of tons per year)
1987 Emissions
With FMVCP &
OJ
•e.
1975
Source Category Emissions3
Light-duty vehicles
Other highway vehicles
Non-highway vehicles
Oil and gasoline
marketing"
Fuel Combustion
Chemical manufacturing
Petroleum Industries
5.30
0.96
1.04
0.79
0.13
0.43
0.43
No Stationary
Source Control
1.80-2.06
0.58-0.65
0.69-0.78
1.00-1.13
0.15-0.17
0.69-0.87
0.53-0.61
Other Industrial
Processes 0.53
Solvent and Petroleum
Evaporation0 8.fi7
Solid Waste 0.24
Miscellaneous 0.07
TOTAL 18.59
0.75-0.84
12.38-15.60
0.19
0.07
18.84-22.97
1987 Emissions
With FHVCP &
New Source
Control
1.80-2.06
0.58-0.65
0.69-0.78
1.00-1.13
0.15-0.17
0.40-0.43
0.27-0.28
0.75-0.84
7.19-7.80
0.19
0.07
13.09-14.40
aThis emissions inventory represents the sum of emissions for each AQCR obtained from the
NEDS Emission Summary Report (NE204)
Includes only service stations.
clncludes emissions from gasoline bulk terminals, bulk plants, and gasoline and crude oil storage
in addition to solvent application sources. This categorization is from the NEDS Emissions
Summary Report, which is currently being revised to more appropriately segment sources.
-------
Table 3-2. RANGE OF GROWTH RATES FOR MOBILE AND STATIONARY SOURCES4
(Compounded Percent per Year)
LOU HIGH1
Mobile Sources(VMT 2 3
Oil and Gas Marketing 2 3
Fuel Combustion 1 2
Chemical Manufacturing 4 6
Petroleum Industries 2 3
Other Industrial Processes 3 4
Solvent and Petroleum
Evaporation 3 5
Solid Waste -2 -2
Miscellaneous C 0
3-5
-------
Baseline emissions from oil and gas marketing, which in NEDS includes
only service stations, are a function of the growth in vehicle miles
traveled, which is estimated to grow at a rate of two to three percent
per year. This growth is tempered somewhat by a projected 10 percent
improvement in fuel economy. While new vehicles in the mid-1980's are
expected to be 20 to 25 percent more efficient than 1975 vehicles, this
10 percent efficiency factor represents an average of the entire vehicle
fleet over time.
Emissions from fuel combustion are expected to grow at a rate of
approximately one to two percent per year. The major industrial process
sources of HC, chemical manufacturing and petroleum refining, are estimated
to grow at annual rates of four to six and two to three percent, respectively.
Other industrial processes are expected to grow at three to four percent
per year.
Solvent and petroleum evaporation is by far the largest source of
emissions, constituting 47 percent of the total hydrocarbon emissions and
77 percent of the stationary source emissions in the 90 non-attainment
AQCRs. In NEDS, this category includes a variety of sources such as
industrial surface coatings, adhesive applications, dry cleaning, asphalt
application, graphic arts, metal cleaning and degrees ing, gasoline bulk
terminals and oulk plants, crude oil storage, and many other small
individual sources, such as architectural coatings.* Growth in these
many sources is expected to vary widely and thus growth of emissions
*The NEDS Emissions Summary Report is currently undergoing revision so as
to more appropriately segment common sources. Nonetheless, this report
utilizes the existing NEDS categorization since NEDS is the only source
AQCR emission inventories.
3-6
-------
could be extremely variable in individual AQCRs depending on the mix of
sources. Nonetheless, an overall growth rate range of three to five
percent is predicted for all the sources if recent solvent use levels
continue. However, this growth could be less than projected because of
increased costs of solvents and the use of high solids coatings or
water-based coatings. Assuming a growth rate of five percent in
emissions, by 1987 the solvent and petroleum evaporation category will
constitute 68 percent of the total hydrocarbon emissions and 81 percent
of the stationary source emissions in the 90 AQCRs.
3.3 EMISSION PROJECTIONS WITH CONTROLS ON NEW STATIONARY SOURCES
In non-attainment areas of the country, the recent Clean Air Act
Amendments require new and modified stationary sources to achieve the
lowest achievable emission reduction (LAER) before the source can be
located in an area or modified. Hence, this requirement will result in
a significant reduction in emissions from new and modified stationary
sources. In order to estimate the reduction in uncontrolled emissions
that can be achieved, assumptions as to the efficiency of new source
controls have to be made. Since EPA has not yet determined what consti-
tutes LAER for stationary sources, this analysis assumes that new and
modified sources will have to achieve a minimum level of control equal to
RACT. The underlying assumptions in the determination of RACT control
levels will be discussed in the next chapter. The assumed new source
control level for applicable sources are summarized below:
3-7
-------
• Chemical Manufacturing 82%
• Petroleum Industries 96%
* Solvent and Petroleum
Evaporation 81%
No new source controls are assumed for the other stationary source
categories since achievable control levels with RACT have not been
identified. In the case of service stations, no new source controls are
included since relatively few new service stations are anticipated to be
built in the near future due to the significant attrition of stations
currently underway.
Table 3-1 summarizes the baseline emissions in 1987 assuming
statutory motor vehicle controls and minimum controls on new and modified
stationary sources. These stationary source estimates include the normal
replacement of existing sources (due to obsolescence) with new, modified,
or reconstructed sources that would be subject to control under LAER. As
shown in the table, even with growth in emission sources, new source
controls on new and modified sources in conjunction with motor vehicle
controls will result in an aggregate 20 to 30 percent reduction in
hydrocarbon emission levels in the 90 AQCRs by 1987.
3.4 REASONABLY AVAILABLE CONTROL MEASURES FOR EXISTING SOURCES
3.4.1 Estimates of RACT for Stationary Source Categories
For areas that fail to attain the standard with mobile source and
new and modified source control, existing sources will be required to
install reasonably available control technology (RACT) in order to attempt
3-8
-------
to attain the standard. RACT will vary among industries and may we!1
vary among sources within an industry. RACT is defined as the lowest
emission limit that a particular source is capable of meeting by the
application of control technology that is reasonably available considering
technological and economic feasibility. Since economic feasibility
is basically source specific, RACT may vary among sources in an industry.
Since this study considers broad categories of sources and does not
consider the feasibility of RACT on individual sources, RACT is broadly
defined for purposes of this study as technology available for application
in categories of sources which will lead to adequate levels of control
based solely on technical considerations. In actual practice, the
reasonableness of technology for particular sources will be determined by
state and local agencies based on technical guidance issued by EPA and
depending upon economic and energy feasibility.
EPA is currently developing guideline documents for sources of VOC
which will assess the technology available for these sources. Based on
studies underway in support of these documents, estimates have been made
as to the efficiency of RACT for many industries which constitute major
sources of VOC. The industries and individual emission sources are
summarized in Table 3-3, which also presents the estimated efficiency
of RACT for these sources.
While these sources represent the primary sources of VOC, other
sources exist which have not been studied by EPA. These sources are
primarily in the solvent evaporation category. In many cases, the
3-9
-------
Table 3-3. ASSUMED EFFlClEdCY ,0£ RACT FOR STATIONARY.
SOURCE CATEGORIES 0>/
Source Category and Sources Included3 Efficiency of RACT (7.)
1. Chemical Manufacturing
t Organic Chemical Manufacturing Industry
- Process Emissions 90X
- Fugitive Emissions 802
- Storage and Loading Emissions 90X
- Secondary Emissions 75X
• Pharmaceutical Industry 95%
• Paint Manufacture 90%
• Rubber Industry 75X
Weighted Average of RACTb 82*
Current Emissions Affected by RACT c 792
Emission Reduction in Source
Category Achievable with RACT 65*
2. Petroleum Industry
• Gas and Crude Oil Production 90X
• Petroleum Refining
- Vacuum Jets
- Waste Water Separators
- Miscellaneous Sources
- Process Unit Turnaround
• Natural Gas and Gasoline Plants
Weighted Average of RACTb
Current Emissons Affected by RACT c
Emission Reduction in Source Category
Achievable with RACT 92%
3- Solvent and Petroleum Evaporation
Auto and Light Duty Truck Manufacturing SOX
Graphic Arts SOX
Flatwood Products 80X
Paper Coating 81X
Fabric Coating SIX
Shoe Adhesive 81%
Wire Coating 90X
Packing Laminates SIX
Can Coating SOX
Metal Furniture 85X
Industrial Machinery 80%
Commercial Machinery SOX
Coil Coating 85%
Fabricated Metal Products BOSS
3-10
-------
Table 3-3. (continued)
Source Category and Sources Included3 Efficiency of
Large Appliances 85%
Small Appliances 80%
Dry Cleaning 65%
Cutback Asphalt Paving 100%
Cold Cleaning 50%
Vapor Degreasing 55%
Gasoline Bulk Plants 90%
Gasoline Bulk Terminals 90%
Gasoline and Crude Oil Storage 75%
Weighted Average of RACTb
Current Emissions Affected by RACTC
Emission Reduction in Source Category
Achievable with RACT 39%
4. Oil and Gasoline Marketing
Service Stations-Storage 90%
Service Stations-Refueling 90%
Weighted Average of RACTb 90%
Current Emissions Affected by RACTC 99%
Emission Reduction in Source Category
Achievable with RACT 89%
5. Fuel Combust iond 0%
6. Other Industrial Processes 0%
7. Solid Wasted 0%
8. Miscellaneous 0%
Sources included are those for which screening studies or guideline
documents are being prepared. Many solvent evaporation sources and other
industrial processes are not included because the sources have not been
identified or control levels have not been defined.
Represents weighted average of RACT for sources listed based on current
national emissions for each category.
Represents the proportion of current national emissions for which RACT can
be applied. Excludes emissions from source categories for which RACT control
levels have not been identified (based on reference 6) and the residual
emissions from sources which have already controlled to the RACT level
RACT has not been identified for these source categories.
3-11
-------
specffi'e sources have not b'eeri' i'dervfifle'd! a'frd' control teefintf'fd'gy aritf
efficiency have" not been a'ssessed, though' these sources wil't Be analyzed
in detail in the future.
Since the emission inventory ftfr each AQ'CR from NEDS is segmented
according to the eight broad source categories,- the average emission
reduction that could be achieved with the application o'f identified RACT
has to be determined for the categories, to do this, it is first necessary
to derive the weighted average of the efficiency of RACT fo'r the source's
in each category for which tfACT has been estimated. This weighted
average .takes into account the relative contribution to current national
emissions and the efficiency of RACT for each source.6 After this
calculation, the total emission reduction" in ea'ch general source category
is determined by multiplying the weighted efficiency of RACT for applicable-
sources by the percentage tif emissions in the genera! source category
which would be affected by RACf. The emissions' affected by RACT include
only those sources for which RACT has been identified and which have not
already controlled emissions to the RACT l§vei. Thus, this excludes
sources not covered by RACT arid those already controlled since these latter
sources already achieve RACT control. 5
As Table 3-3 indicatesj emissions from existing sources cart be reduced
by 65 pe'rcent from the chemical manufacturing industry, 9? percent from the
petroleum industry, and 89 percent from oil arid gas marketing-. However,
controls on ideritified solvent and petroleum evaporation sources will reduce
emissions from the total category by only 39 percent, since oVe'f half of
the emission inventory results from sources whi£h haV6 not been identified
•Ye
-------
and for which control levels have not been estimated. While the control
of identified sources serves as the basis for this analysis, a sensitivity
analysis is included which assumes a moderate level of control or, the
emissions from the unidentified solvent sources. This assumed level of
control is 65 percent of all solvent evaporation emissions. Through the
remainder of the document, the control of identified sources only is termed
"identified" RACT while the assumed control of the entire solvent
evaporation category is termed "advanced" RACT.
3.4.2 Estimates of RACT for Mobile Source Categories
There are measures in addition to tailpipe standards which are
available to reduce emissions from mobile sources. Reasonably available
measures for mobile vehicles include inspection/maintenance as well as
traffic reduction measures such as transit improvements, parking management,
and traffic management. The emission reduction potential for each of these
measures will vary among areas dependent on the applicability of control
measures, the relative contribution of each mobile source category, existing
traffic patterns, and other factors.
In tnis study, it is anticipated that an inspection/maintenance
program for light-duty vehicles could achieve a reduction of up to 30
percent of 1987 emissions from the LDV category, with the FMVCP as
8 9
baseline, though the effectiveness could range from 20 to 50 percent. '
In addition, for purposes of this study, traffic reduction plans are
estimated to reduce vehicle miles traveled by 10 percent compared to the
projected baseline in 1987. Hence, total RACT for light duty vehicles
is estimated to represent a 37 percent reduction. Similarly, the emissions
from other highway vehicles can be reduced an additional five percent
through traffic reduction.
3-13
-------
3.4.3 Total Emission Reductions with RACT
Table 3-4 summarizes the total emission reduction in the 90 non-
attainment AQCRs that can be achieved by applying both identified and
advanced RACT to existing sources that will not be replaced before 1987.
-Baseline emissions in 1987, taking into account FMVCP and controls on
new and modified sources, can be reduced by an additional 30 to 35
percent with the application of RACT. Of course, the reduction in
individual AQCRs will vary depending on the mix of sources in the area.
This reduction with RACT does not take into account the control of
replaced sources that would also be controlled to the LAER level (see
Section 3.3). Considering the control of replaced sources, existing
stationary source emissions can be reduced from 1975 to 1987 by an
average of 52 percent with identified RACT and 66 percent with advanced
RACT.
3.5 ATTAINMENT STATUS AND ADDITIONAL REDUCTION REQUIRED
Even with full application of RACT to identified source categories,
many AQCRs will not be able to attain alternative levels of the standard.
Tables 3-5 and 3-6 summarize the attainment status for the AQCRs for the
alternative standard levels. Using the linear rollback model (Table 3-5),
from 30 to 60 AQCRs will attain a standard level of .08 ppm depending on
the assumptions as to growth rates and the control levels achievable. As
the table indicates, as the standard level increases more AQCRs will
naturally come into attainment, to the point that all but two to nine
3-14
-------
Table 3-4. '"TAL EMISSION REDUCTION ACHIEVABLE IN 1987 WITH FULL
Al'PLJCATlON OF RACT TO IDENTIFIED SOURCES FOR 90
NUN-ATTAIIIMENT AQCRs (millions of tons per year)
OJ
i
Light Duty Vehicles
utlicr Highway Vehicles
in-highway vehicles
I end Gasoline Marketing
Fuel Combustion
Chemical Manufacturing
Petroleum Industries
Other Industrial Processes
lent a-«1 Petroleum Evaporation
Solid Waste
Miscellaneous
TOTAL
!«!_..*.£• ~.*l
1987 Emissions with
FMVCP and Control of
New and Modified Sources
1.80-3.06
0.58-0.65
0.69-0.78
1.00-1.13
0.15-0.17
0.40-0.43
0. 27-0. 28
0.75-0. 84
7.19-7.80
0.19
0.07
13.U9-14.40
Emission Reduction
with RACT
0.66-0.76
0.03
0
0.88-1.00
0
0.22
0.24
0
2.32
0
0
4.35-4.57
1987 Emissions with
Total Control Achievable
1.14-1.30
0.55-0.62
0.69-0.78
0.12-0.13
0.15-0.17
0.18-0.21
0.03-0.04
0.7S-0.84
4.87-5.48
0.19
0.07
8.74-9.83
Advanced
Emission Reduction
with RACT
0.66-0.76
0.03
0
0.88-1.00
0
0.22
0.24
0
3.91
0
0
5.94-6.16
RACT
1987 Emissions with
Total Control Achievable
1.14-1.30
0.55-0 62
0.69-0.78
0.1?-0.13
0.15-0.17
0.18-0.21
0.03-0.04
0.75-0.84
3.?8-3.89
0.19
0.07
7.15-8.24
-------
Table 3-5. ATTAINMENT STATUS AND ADDITIONAL EMISSION REDUCTION REQUIRED
FOR 90 AQCRs UNDER ALTERNATIVE OZONE STANDARDS
(ROLLBACK)
Level of Standard
(ppm)
.08
.10
.12
.14
Assumptions:
Background
Growth Rates
RACT
Allowable Emissions 1987
(106 Tons)
6.7
8.7
10.2
11.3
.02
Approximate No. of AQCRs
Attaining Standard3
60
80
85
88
.02
Low
Advanced
38
68
81'
84
.02
Low
Identified
25
60
75
81
.02
High
Identi-
fied
Additional Emission Reductions
Neededb {106 Tons)
1.2
0.5
0.2
0.1
.02
Low
Advanced
2.3
i.2
0.6
0.2
.02
Low
Identified
3.2
1.8
1.0
0.5
.02
High
Identifie
CO
aThis is only an approximate number based on assumptions outlined in this report.
bThe total reduction in emissions estimated to be needed for all 90 AQCRs to attain the alternative
levels of the standard.
-------
Table 3-6. ATTAINMENT STATUS AND ADDITIONAL EMISSION REDUCTION REQUIRED
FOR 90 AQCRs UNDER ALTERNATIVE OZONE STANDARDS
(EKMA)
.evel of Standard
(ppmj
.08
.10
.12
.14
,'Licns'
acl- ground
rowlh Rates
ALT
Allowable Emissions 1987
(106 Tons)
3.1
5.8
8.1
9.1
.02
Approximate Number of AQCRs
Attaining Standard3
4
43
74
80
.02
Low
Advanced
3
26
62
71
.02
Low
Identified
3
17
49
63
.02
High
Identifie
Additional Emission Reductions
Neededb (10° Tons)
4.1
1.8
0.8
0.5
.02
Low
Advanced
5.6
3.1
1.6
1.1
.02
Low
Identified
6.7
4.1
2.3
1.7
.02
High
Identitie
llii i is only an approximate number based on assumptions outlined in this report.
'l1' total reduction in emissions estimated to be needed for all 90 AQCRs to attain the alternative
iwols of the standard.
-------
AQCRs could attain a .14 standard level by 1987. The use of EKMA, on the
other hand, leads to more pessimistic results (Table 3-6). According to
this modeling approach, only three to four AQCRs could attain a .08
standard by 1987, while at most 80 AQCRs could attain a .14 standard level.
These attainment numbers are only approximate and could vary significantly
depending on assumptions for each AQCR pertaining to design air quality
values, growth rates, control levels, emissions inventory, and the
modeling technique employed.
Tables 3-5 and 3-6 also indicate the total additional emission
reductions estimated to be needed for all 90 AQCRs to attain the alterna-
tive levels of the standard. For comparison purposes, the allowable
emission levels are also presented.
3-18
-------
3.6 REFERENCES FOR CHAPTER 3.0
1. Environmental Protection Agency, Control Strategy Preparation Manual
for Photochemical Oxidants, OAQPS Guideline Series No. 1.2-047,
January 1977.
2. McCurdy, Thomas, Strategies and Air Standards Division (SASD),
OAQPS, EPA, memorandum, "FHWA Projected Increases in Highway Mileage
and VMT." April 6, 1978.
3. Jordan, Bruce, Policy Analysis Staff, OAQPS, EPA, Personal communication
to Kenneth Lloyd, SASD, OAQPS, EPA, March 20, 1978.
4. Data Resources, Inc., (Lexington, Mass.) U.S. Long-Term Review,
Winter 1978.
5. Arthur D. Little, Inc., The Economic Impact of Vapor Recovery Regulations
on the Service Station Industry, Report prepared for EPA and OSHA,
March 1978.
6. Walsh, Robert T., Emission Standards and Engineering Division, OAQPS,
EPA, Memorandum to Bruce Jordan, "Estimated Reductions in Volatile
Organic Compound Emissions Which Could be Effected at Stationary
Sources," August 12, 1977.
7. Walsh, Robert T., ESED, OAQPS, EPA, Memorandum, "Current VOC Emission
Inventory," November 30, 1977.
8. Environmental Protection Agency, Mobile Source Emission Factors,
EPA-400/9-78-005, March 1978.
Environmental Protection Agency, Appendix N to 40 CFR Part 51: Emission
Reductions Achievable Through Inspection and Maintenance of light Duty
Vehicles. Motorcycles, and Light and Heavy Duty Trucks, May 19/7.
3-'
-------
4.0 NATIONAL COSTS OF CONTROL FOR ALTERNATIVE LEVELS OF THE STANDARD
In the previous sections of this report, estimates were made of the
amount of non-methane hydrocarbon control needed to meet alternative
levels of the ozone standard and control levels achievable with available
control measures. A comparison will now be made between alternative
levels in terms of the cost of applying available controls for various
mobile and stationary sources. The average cost per ton of emissions
controlled will be established for general categories of sources and
combined with estimates of emission reductions achievable in order to
obtain total costs of control. The costs presented represent the costs
incurred in the 90 non-attainment AQCRs under consideration in the previous
section.
The costs presented in this section represent annual costs of control
which will be incurred in 1987, when it is assumed that the control measures
will be applied fully to the applicable sources. The annual costs include
operating, maintenance and administrative costs as well as annualized
capital charges that take into account depreciation and interest costs.
No estimates are made of the total capital costs that will result from
application of the control measures outlined. In addition, secondary
costs and benefits associated with control measures are not estimated.
4.1 COSTS FOR THE FEDERAL MOTOR VEHICLE CONTROL PROGRAM
The Federal Motor Vehicle Control Program (FMVCP) includes tailpipe
emission standards for hydrocarbons (HC), carbon monoxide (CO), and
nitrogen oxides (NOx) from motor vehicles. Since control of HC and CO
-------
occurs simultaneously with catalyst control, it is not possible to
allocate specific control for either pollutant. Thus, it also is not
possible to allocate the costs of control between the pollutants. The
costs estimated below pertain to both HC and CO control.
Control of HC and CO to the statutory limits for light duty vehicles
will involve an installed cost of about $125 per vehicle for model years
1975-81 and an installed cost of approximately $218 per vehicle for model
years 1982-87.1 These unit costs are combined with estimates of the
composition of the vehicle fleet for the 90 AQCRs in 1987 with respect
to model year in order to estimate national costs of the program. The
annual cost in 1987 for the 90 AQCRs is estimated to be $2.8 to $3.0
2
billion, which will be the cost irrespective of the level of the standard.
4.2 COSTS FOR NEW AND MODIFIED STATIONARY SOURCES
Appendix C presents a discussion of the many sources of hydrocarbons
and the control techniques that are available. An analysis of the costs
of these techniques for each source of emissions has been made and a summary
is also presented in Appendix C. These costs are current as of July 1977.
Since all of the sources considered in Appendix C fall into the broad
categories of chemical manufacturing, petroleum industries or solvent
evaporation sources, it is assumed that only industries in these categories
will be subject to controls on new and modified sources. In order to
estimate the average cost per ton of hydrocarbons controlled for the general
source categories, the individual stationary sources contained in Appendix C
were segmented into the three source categories. The average cost per ton
4-2
-------
controlled for each category represents a weighted average based on the
cost per ton and the net reduction in emissions {in tons) achievable with
RACT (from Appendix D) for each source. The average costs are summarized
as follows:
0 Chemical Manufacturing $100/ton
0 Petroleum Industries 0 (costs are offset by product recovery)
0 Solvent Evaporation $150/ton
To obtain total new source costs for each AQCR, the average cost per
ton is multiplied by the emission reductions for the source categories
due to the control of new and modified sources. The total national costs
of new source control, which represents the sum of the costs for the 90 AQCRs,
amount to an annual cost of approximately $900 million to $1.3 billion, of
which more than 95 percent comes from solvent evaporation sources while the
rest comes from chemical manufacturing sources.
4.3 COSTS OF APPLYING RACT FOR ALTERNATIVE STANDARDS
4.3.1 Cost-effectiveness of Reasonably Available Control Measures
The approach for determining the average cost per ton controlled for
the sources for which RACT is applicable is the same as the approach outlined
in the previous section for new sources. The hasps for the costs are
discussed in detail in Appendix C. The weighted average cost of RACT for
the source categories is summarized in Table 4-1.
The unit costs for transportation sources were derived somewhat
differently than for stationary sources. For light-duty vehicles, the
cost-effectiveness of an inspection/maintenance is estimated to be about
-------
$340 per ton of hydrocarbons removed at current automotive emission
levels.3 As the emissions of new cars are reduced with statutory require-
ments, though, the cost per ton will increase as there are less baseline
emissions to reduce. It is estimated that by 1987 the average cost
effectiveness of inspection/maintenance will be $420 per ton. Since the
cost of traffic reduction measures are extremely variable, for purposes
of this study it is assumed that reasonable aspects of these measures can
be instituted for around $1000 per ton. Traffic reduction measures for
other highway vehicles are also expected to be costly since it will be
difficult to affect reductions in vehicle miles traveled for these
sources. Once again, the cost of these measures will be extremely
variable, but a cost of $1000 per ton is assumed for analytical purposes.
4.3.2 Analytical Methodology
For each alternative standard, maximum allowable 1987 emission levels
are calculated for each AQCR based on 1975-1977 expected air quality
levels, 1975 total emissions of non-methane hydrocarbons, the level of
the alternative standard and assumed relationships between NMHC emissions
and ambient ozone concentrations using both the linear rollback model and
the Empirical Kinetic Modeling Approach (EKMA). The allowable emissions
for each AQCR are then compared to the projected 1987 emission levels,
based on the national estimates of growth and retirement rates and new
sourre emission levels for each source category over the 12-year interval.
If projected emissions are greater than allowable emissions, existing
source emissions (i.e., those that have not been replaced) are reduced
appropriately due to RACT and RACT costs are computed. The costs are
4-4
-------
obtained by multiplying the average cost per ton for RACT for each
general source category (see Table 4-1) by the emission reduction
due to RACT for each category.
If controlled 1987 emissions are less than allowable emissions for
alternative standards, the application of total RACT is unnecessary.
In this case, the source categories are partially decontrolled in a
manner which minimizes total control costs. This involves applying
controls to the source categories in order of the most cost-effective
first.
After the costs for each individual AQCR have been determined, the
AQCR costs are then summed in order to arrive at national costs. The
costs presented are only for those existing sources which have net been
replaced by 1987 and does not include the costs of controls on replaced
sources, which is included in the new source costs presented earlier.
4.3.3 Total Costs
The total annualized costs in 1987 of applying RACT in the 90 non-
attainment AQCRs are presented in Table 4-2. The costs differ for
alternative levels of the standard since as the level becomes less
stringent more areas can attain the standard without applying full RACT.
In addition, the costs differ depending upon the modeling technique used.
Rollback, using the assumptions in this report, generally requires less
control than EKMA and thus rcore AQCRs -are able to attain the alternative
levels of the standard.
4-5
-------
Table 4-1. AVERAGE RACT CONTROL COSTS3
Control Costs
Source ($ per ton)
Transportation Sources
Light Duty Vehiclesb $530
Other Highway Vehicles 1000C
Non-Highway Vehicles
Oil and Gas Marketing 275
Stationary Sources
Fuel Combustion
Chemical Manufacturing 100
Petroleum Industries 0
Other Industrial Processes 0
Solvent Evaporation
- Identified RACT 150
- Advanced RACT 300e
Solid Waste
Miscellaneous
aData on control costs for reasonably available control technology are
taken from Appendix C.
bDoes not include cost of tailpipe controls, which are assumed as given
and are included earlier in the costs of FMVCP.
°The RACT control for this category is traffic reduction. The estimate of
average cost is a very rough approximation. Only sparse and usually site-
specific information available for verification.
dCosts are offset by credit due to product recovery.
eBased on cost of $150/ton for identified sources and $500/ton for
additional sources.
4-6
-------
Table 4-2. ANNUAL CONTROL COSTS IN 1987 OF APPLYING RACT IN
90 AQCRs FOR ALTERNATIVE STANDARD LEVELS
($ Billions)
Rollback Model
EKMA
Level of
Standard
.08 ppm
.10 ppm
.12 ppm
.14 ppm
Identified
RACT
0.8-1.0
0.6-0.8
0.4-0.6
0.3-0.4
Advanced
RACT
1.3-1.7
0.9-1.3
0.6-0.9
0.4-0.6
Identified
RACT
0.9-1.1
0.8-1.1
0.6-0.9
0.5-0.7
Advanced
RACT
1.5-1.9
1.4-1.8
1.0-1.4
0.9-1.2
-------
Utilizing the rollback model, costs of applying RACT at existing
identified sources range from up to $1.0 billion per year for a .08 ppm
down to about $300 million per year for .14 ppm. For EKMA, the estimated
1987 annual costs for identified sources range from $1.1 billion for .08
ppm down to $500 million for a .14 ppm. The difference in costs between
alternative standards is not as great as might be expected, especially
under EKMA, because reductions in costs occur only in those areas which
meet the particular standard without having to apply full RACT. For
instance, under EKMA, from 20 to 40 percent of the AQCRs under consideration
are not predicted to achieve either a .08 ppm standard or a .12 ppm
standard. Since these areas have to apply full RACT for both standards,
this results in the same cost of control for the areas. This concept also
explains why the cost differences between alternative standards are not
as great under EKMA since so many areas have to apply full RACT regardless
of the level of the standard.
For "advanced" RACT, the costs are naturally greater since more
sources are being controlled at a higher cost. Based on the rollback
results, the costs range from up to $1.7 billion for a .08 ppm level
to a low of $400 million for a .14 ppm level. The costs are greater with
EKMA and there is not as much difference in costs due to alternative
standard levels.
4.4 SUMMARY OF COSTS FOR FMVCP, NEW SOURCE CONTROL, AND RACT ON
EXISTING SOURCES
Tables 4-3 and 4-4 summarize the costs discussed thus far in this
chapter, depending upon the modeling technique used. As can be seen, the
4-8
-------
Table 4-3. SUMMARY 01- COSTS FOR FMVCP, NEW SOURCE CONTROL, AND RACT
ON EXISTING SOURCES ($ BILLIONS)
(Rollback)
Level of
Standard
.08 ppm
.10 ppm
.12 ppm
.14 ppm
Costs of
FMVCP
$2.8-3.0
2.8-3.0
2.8-3.0
2.8-3.0
Costs o1
New Source
Control
$0.9-1.3
0.9-1.3
0.9-1.3
0.9-1.3
Identified
Cost of
RACT
$0.8-1.0
0.6-0.8
0.4-0.6
0.3-0.4
RACT
Total
Costs3
$4.5-5.3
4.3-5.1
4.1-4.9
4.0-4.7
Advanced
Cost of
RACT
$1.3-1.7
0.9-1.3
0.6-0.9
0.4-0.6
RACT
Total
Costs3
$5.0-6.0
4.6-5.6
4.3-5.2
4.1-4.9
i
•o
Total costs represent sum of costs for FMVCP, new source control, and appropriate RACT.
-------
Table 4-4. SUMMARY OF COSTS FOR FMVCP, NEW SOURCE CONTROL, AND RACT
ON EXISTING SOURCES ($ BILLIONS)
(EKMA)
Level of
Standard
.08 ppm
.10 ppm
.12 ppm
.14 ppm
Costs of
FMVCP
$2.8-3.0
2.8-3.0
2.8-3.0
2.8-3.0
Costs of
New Source
Control
$0.9-1.3
0.9-1.3
0.9-1.3
0.9-1.3
Identified
Cost of
RACT
$0.9-1.1
0.8-1.1
0.6-0.9
0.5-0.7
RACT
Total
Costs3
$4.6-5.4
4.5-5.4
4.3-5.2
4.2-5.0
Advanced
Cost of
RACT
$1.5-1.9
1.4-1.8
1.0-1.4
0.9-1.2
RACT
Total
Costs9
$5.2-6.2
5.1-6.1
4.7-5.7
4.6-5.5
aTotal costs represent sum of costs for FMVCP, new source control, and appropriate RACT.
-------
total costs of the measures considered vary relatively little with
alternative standards. For identified RACT, total costs ranging from i
low of $4.08 billion with rollback at .14 ppm to a high of $5.4 billion
with EKMA at .08 ppm. This difference is not great when one considers
that the total costs for any standard level vary up to $.08 billion.
The differences are somewhat greater for "advanced" RACT. While the
costs range from $4.1 billion with rollback at .14 ppm to a high of $6.2
billion with EKMA at .08 ppm, the total costs at any standard level can
vary by as much as a billion dollars.
Regardless of modeling technique the reason for these relatively
small differences for alternative standard levels are twofold. First,
the vast majority of the total costs ($3.7 to $4.3 billion) result from
the FMVCP and new source controls, which are assumed to be the same
regardless of the level of the standard. Secondly, the RACT control
costs do not differ greatly from alternative standards since many areas
have to apply full RACT regardless of the level of the standard. One
point to remember though is that these costs do not represent the cost
for all 90 AQCRs to attain the standard. AQCRs which fail to attain the
alternative standards will have to apply additional control measures,
which will result in a larger difference between the costs associated
with the alternative levels of the standard.
4.5 ESTIMATED COST OF ATTAINMENT
As discussed in Chapter 3.0, many AQCRs will not attain alternative
levels of the standard by applying the control measures outlined in this
-------
report. Thus, additional control measures will be needed in some areas in
order to reduce emissions even further so that the standards can be attained.
These measures could include restrictive transportation control measures that
significantly reduce vehicle miles traveled in urban areas, control of
stationary sources for which RACT has not been identified or defined, and
tighter controls on new and existing sources than those achievable with RACT.
The cost-effectiveness of these measures has not been estimated, though
on the whole they are believed to be more costly than current measures
that have been defined. While some measures can no doubt be implemented
at costs approaching the level of RACT costs identified earlier, many
measures are likely to be much more costly. For example, traffic
reduction plans are extremely area specific and it is likely that.
application of such measures could cost up to several thousand dollars
per ton of emissions controlled, though by the same token they could cost
much less. Likewise, the control of the many solvent evaporation sources
that have yet to be identified is likely to be more costly since these
sources tend to be relatively small and difficult to control from a
technical standpoint. For instance, the cost to control small coin-op
4
drycleaners is estimated to be about $5000 per ton of solvent controlled.
Finally, the cost of applying more stringent controls in place of RACT
on new and existing stationary sources is likely to be high due to
higher marginal costs of control. Moving to more stringent levels of
control for particular sources will involve exponentially higher costs
for a relatively small reduction in emissions. For example, the marginal
cost of achieving 95 percent control at service stations versus 90
percent control is likely to be greater than $5000 per additional ton
removed.
4-12
-------
Even though the costs of additional control measures are not Known,
it is still useful to estimate the cost of attainment is some manner ir
order to better indicate the cost differences between alternative standards.
To do this, a cost-effectiveness estimate of 51,000 to $1,500 per ton
controlled is assumed for the additional emissions reductions past
"advanced" RACT required for each AQCR to attain the alternative standards.
While there are likely to be many measures that are more costly than
this, it is also quite possible that many measures may be applied at
cost levels approaching projected RACT costs discussed earlier in this
study. The $1000-$!500 per ton figure represents a compromise between
the extreme estimates.
To determine the toial cost of attainment in all 90 AQCRs, the
additional emission reduction required (from Table 3-4) is multiplied
by $1,000 to $1,500 per ton, with the resulting costs added to the costs
summarized previously in Tables 4-3 and 4-4. The estimated costs of attain-
ment are presented in Table 4-5. Employing the rollback model, the total
cost of attainment for a .08 ppm standard is estimated to range between
$6.0 and $8.0 billion, falling to $5.0 to $6.5 billion, $4.5 to $5.5 billion,
and $4.0 to $5.0 billion for a .10 ppm, .12 ppm, and .14 ppm standards,
respectively. The costs using EKMA are higher, ranging from $5.0 to $6.5
billion for a .14 ppm to between $9.0 and $12.5 billion for .08 ppm. In
addition to the assumed cost-effectiveness of additional controls making
the cost differences conservative, the differences may also be somewhat
understated due to the fact that more stringent levels of the standard are
4-13
-------
Table 4-5. ESTIMATED COST FOR ALL 90 AQCRs TO ATTAIN ALTERNATIVE STANDARDS
($ Billions)
Level of
Standard
(ppm)
.08
.10
.12
.14
ROLLBACK
Cost of
FMVCP
New Source
Control, and
Advanced RACTa
$5.0-6.0
4.6-5.6
4.3-5.2
4.1-4.9
Cost of
Additional
Reduction
Required0
$1.2-1.8
0.5-0.8
0.2-0.3
0.1-0.2
Total
Cost of
Attainment
$6.2-7.8
5.1-6.4
4.5-5.5
4.2-5.1
EKMA (CONSTANT NOx EMISSIONS)
Cost of
FMVCP
New Source
Control , and,
i Advanced RACTa
$5.2-6.2
5.1-6.1
4.7-5.7
4.6-5.5
Cost of
Additional
Reduction
Required6
$4.1-6.2
1.8-3.0
0.8-1.2
0.5-0.8
Tota.l
Cost of
Attainment
$9.3-12.4
6.9-9.1
5.5-6.9
5.1-6.3
aFrom Tables 4-3 and 4-4.
Emission reduction from Tables 3-5 and 3-6 multiplied by $1000-1500/ton, which is a lower-bounds estimate.
-------
likely to require more sophisticated and extensive, and thus more costiy,
controls. The approach used in this section does not take into consideration
such differences.
4-15
-------
4.6 REFERENCES FOR CHAPTER 4.0
1. Environmental Protection Agency, Auto Emission Control: Current Status
and Development Trends as of March~T976.
2. Personal Communication from Dale L. Keyes, Energy and Environmental
Analysis, Inc., Arlington, Virginia, to Kenneth H. Lloyd, Strategies
and Air Standards Division, OAQPS, EPA, May 30, 1978.
3. Walsh, M.P., Mobile Source Enforcement Division, EPA, "The Need for and
Benefits of Inspection and Maintenance of In-Use Motor Vehicles",
November 9, 1976.
4. Environmental Protection Agency, Control of Volatile Organic Emissions
from Dry Cleaning Operations. Draft, April 1977.
5. Lloyd, Kenneth H., Economic Analysis Branch, OAQPS, EPA, "Cost of
Alternative Vapor Recovery Systems at Service Stations," October 20, 1977.
4-16
-------
5.0 ECONOMIC IMPACT OF REASONABLY AVAILABLE CONTROL MEASURES
ON SELECTED INDUSTRIES
5.1 INTRODUCTION
Control measures to reduce emissions of hydrocarbons and other volatile
organic compounds will affect many industries in the U.S. economy. In
addition to major industries such as chemical manufacturing, petroleum
refining and automobile production, control measures will affect many
small industries which use petroleum products or organic solvents in a
wide range of applications. Countless industries which produce consumer
products use solvents in coating operations, while many industries use
solvents in metal and product cleaning. For examples of the industries
affected by control measures for hydrocarbons, consult Appendix D.
An analysis of the economic impact of the revised standard on the
numerous industries affected is not possible to complete in the time frame
available for analysis. However, EPA has conducted economic impact studies
for the major emission sources, though these constitute only a portion of
the total number of sources affected. Economic analyses of all affected
industries are being initiated and significant results are expected within
two to three years. The Clean Air Act does not permit the consideration of
costs and economic impact in the setting of health related national
ambient air quality standards. The incomplete character of the economic
impact analysis for all affected sources does not mitigate EPA's legal
burden to propose the standard now.
5-1
-------
Nonetheless, an attempt has been made to give an indication of the
economic impact of reasonably available control measures on selected
industries. These are industries for which EPA has conducted previous
cost or economic studies and for which quantitative or qualitative
economic judgements can readily be made.
5.2 PETROLEUM REFINING
5.2.1 Industry Profile
Crude petroluem is refined by 150 companies at 266 refineries located
in 40 different states. Production of refined products in the U.S. totalled
over 15 million barrels per day in 1976, or 93 percent of nameplate capacity.
The industry employs 100,000 workers and is heavily concentrated in the
West South Central region of Arkansas, Oklahoma, Texas, and Louisiana.
These four states employ 44 percent of all industry workers and supply 43
percent of all refined products. Refineries tend to be concentrated in
areas of the country that have oxidant levels that currently exceed the
standard.
The petroleum refining industry is somewhat concentrated. The five
leading producers own 36.5 percent of all industry capacity; the top ten,
58.5 percent. These leading producers are integrated, major oil companies
that engage in exploration, production, refining, distribution, and marketing
on the retail level. Other refiners are independent companies that are
typically not integrated into more than one other segment of the industry.
Prices vary little among companies, although there are occasional examples
of price cutting when there is weak demand and an excess of supply.
5-2
-------
5.2.2 Costs of Reasonably Available Control Measures
Petroleum refineries are significant sources of volatile organic
compound (VOC) emissions. The major point sources of VOC emissions from
petroleum refineries are vacuum producing systems (VPS), wastewater
separators (WWS), process unit turnarounds (PUT), and leaks from
miscellaneous sources such as pumps, compressors, and valves. EPA has
analyzed control techniques for these sources, which reduce emissions
by 95 to 100 percent. For vacuum producing systems, control can be
achieved by venting the emissions to a firebox. Large reductions in
emissions can be accomplished for the wastewater separators through
covering the separators and forebays. Emissions can be controlled from
process unit turnarounds by piping the VOC to a flare or to the fuel gas
systems. Finally, emissions from miscellaneous equipment can be substan-
tially controlled by reducing equipment leaks with a monitoring and
maintenance program.
Costs of these control measures have also been estimated by EPA based
on analysis of a model sized refinery with a throughput of 100,000 barrels
o
per day. Costs for individual refineries will vary considerably due to
differences in size, configuration and age of facilities, product mix,
and degree of control.
The capital cost for piping for controlling emissions from vacuum
producing systems will range from $23,700 to $51,600, depending on whether
the system uses surface condensers or contact condensers. However, this
control measure should result in an annual savings of $89,000 to $96,700
5-3
-------
due to significant credits for the value of the recovered petroleum. For
wastewater separators, the capital cost of covers for the facilities will
be approximately $62,800, but again due to the value of the recovered
petroleum products, a net annual savings of $309,700 should result. The
control method for process unit turnarounds has an estimated capital cost
of $97,600 for piping and valves. Petroleum product recovery is not as
readily recoverable from this control measure, though some refineries
currently have facilities for recovering the hydrocarbons. Nonetheless,
assuming no recovery of petroleum products, an annualized cost of approxi-
mately $25,900 will result. The total capital cost for these three sources
will range from $184,000 to $212,000. However, this capital outlay is more
than offset by the annual recovered petroleum credits valued at $432,000,
which does not include the value of recovered PUT emissions.
Finally, the cost of a monitoring and maintenance program to detect
o
and control equipment leaks has also been estimated. For a model 100,000
barrel per day refinery, the capital cost for monitors will approximate
$8,600. The annualized cost of a program, taking into account monitoring
and maintenance labor, materials, and capital charges, will be about
$103,000. This annualized cost, though, should be offset to some degree
by recovery credits of reduced emissions. However, emission reduction
factors are not presently quantifiable and thus recovery credits will be
extremely variable.
5.2.3 Economic Impact of Control Measures
The economic impact of the control measures outlined in the previous
section is expected to be small for several reasons. First, the capital
5-4
-------
costs are not large in comparison with the capital cost of a refinery as
a whole. The capital cost for a model refinery with a throughput of
100,000 barrels per day is estimated to range from $300 million to S500
million.4 A capital outlay of $220,000 for the VOC control measures
represents an insignificant increase (less than 0.1 percent) in the
capital cost of a refinery.
Secondly, the value of recovered petroleum products offsets the
capital costs entirely in the first year. Thus, the controls can be
justified solely on economic grounds. Lastly, the economic impact should
not be large since a significant portion of the industry has already
instituted the controls. Twenty-five percent of the industry already
controls vacuum producing systems, 80 percent controls wastewate^
separators, and 40 percent controls process unit turnarounds. Finally,
a comprehensive study of the petroleum refining industry indicates that
EPA's total air and water pollution regulations will result in only a
small impact on this sector.
L 5'3
* 5.3.1
5.3 RETAIL GASOLINE SERVICE STATIONS
Industry Profile
In 1977, there were approximately 178,000 gasoline stations in the
U.S. Over 48,000 service stations have closed in the U.S. since the
population peak of 226,000 in 1972. This attrition is expected to continue
at least through the early 1980's to a leveling off point of anywhere from
125,000 to 150,000 outlets. The economies of scale of high volume stations
and the shift to self-service operations are a prime factor in shrinking
retail margins. Consequently, the closure of outlets due to market
5-5
-------
rationalization processes will be most severe for those outlets which
have relatively low sales volume coupled with high unit expenses.
Major oil companies and regional re finer/marketers supplied over
half of the retail service stations in the country with the remaining
43 percent supplied by independent marketers. The traditional retail
marketing strategy of major oil companies has been to operate through
lessee dealers. These lessee outlets still represent approximately
two-thirds of the major oil company stations and almost 50 percent of
all stations in the country. The second largest group of outlets are
known as open dealers. In these operations, the onsite dealer actually
owns or controls the investment in his station where he is physically
employed. Open dealers represent over one-third of the retail outlets in
the U.S. They are generally branded (i.e., station operating under the
brand identification of a major oil company) and supplied either directly
by a major oil company or a branded jobber. The other types of retail
operations are direct salary outlets and convenience stores, which are
low expense, low margin operations which account for less than 25 percent
of the total population of gasoline retailers. A summary of the service
station market segments is presented in Table 5-1.
Retail service stations dispense an average of about 40,000 gallons
per month. In recent years, marketing economics have resulted in a trend
toward stations with larger volumes, with small volume operations being
marginal operations that have to rely on other parts of the retail trade,
-such as mechanical work and sales of accessories, in order to remain in
business. The high volume stations tend to be mostly direct operations
5-6
-------
Table 5-1. 1977 U.S. SERVICE STATION POPULATION
Number of Outlets (%)
en
i
Supplier
Major
Regional Refiner
Independent Marketer/
Wholesaler -
Super Jokker"
jmall Jobber
Direct
6,400 (3.6%)
4,100 (2.3%)
16,600 (9.3%)
5,000 (2.8%)
32,100 (18.0%)
Lessee
50.200 (28.2%)
9,400 (5.3%)
4,400 (2.5%)
19,400 (10.9%)
83,400 (46.9%)
Type of Operation
Open Dealer
27,800 (15.6%)
2,000 (1.1%)
1,100 (0.6%)
21,900 (12.3%)
52,800 (29.6%)
Convenience
Store
700 (0.4%)
200 (OJ%)
7,700 (4.3%)
1,100 (0.6%)
9,700 (5.4%)
Total
85,100 (47.8%)
15,700 (8.8%)
29,800 (16.7%)
47,400 (26.6%)
178,000 (100. OX)
-------
which are controlled and operated by the supplier and operate on relatively
low margins. Low volume stations, those dispensing less than 25,000 gallons
per month, are mostly lessee dealers and open dealers supplied by all classes
of suppliers. These low volume stations, which comprise close to 50 percent
of the total number of stations, are the segment of the retail industry
that is most vulnerable to changes in marketing economics as well as external
costs such as vapor recovery costs.
v/5.3.2 Cost of Reasonably Available Control Measures
Emissions occur from two major sources at service stations - the loading
of underground storage tanks (Stage I) and the refueling of motor vehicles
(Stage II). For Stage I emissions, vapors can be controlled through the
use of a vapor balance system, where vapors are vented by displacement to
an intermediate holding area (usually the tank truck) for ultimate disposal
or recovery at the bulk terminal or bulk plant. Stage II emissions can be
controlled through a variety of systems, the most basic of which is the
I/balance system where vapors from the refueling operation are displaced by
means of a tight fitting nozzle and vapor return lines to the underground
tank. More elaborate recovery systems create a vacuum whre vapors are
drawn from the refueling operation, alleviating the need for a tight nozzle
fit. Vapors are again displaced to the underground storage tanks, with the
excess vapors being incinerated in most cases. There are several variations
of this vacuum assist system which are too numerous and involved to
discuss here.
Since vapor balance systems are less costly than vacuum assist systems
and control anywhere from 80 to 90 percent of the emissions, only the costs
5-8
-------
of the balance systems are included in this analysis. The capital
of the system varies with the number of dispensers at the station, the
number of underground tanks, and the physical layout of the station.
For a typical nine dispenser, three island station, the capital costs
will approximate $8,800. These costs can range from $4,500 for a two-
o
dispenser station to over $11,000 for a 15-dispenser facility.
Essentially the only operating and maintenance cost associated with
the.balance system is that for nozzle maintenance since the system does not
require any power to operate and there are no moving parts associated with
the remainder of the system. Nozzle maintenance requirements will be
extremely variable depending on a number of factors. However, it is
expected that the nozzle will have to be replaced only once a year or
the faceplate and/or boot repaired or replaced no more than twice a year.
This would result in an annual maintenance cost of about $60 per nozzle,
or about $540 for a nine-nozzle station.
Since the vapor balance system is characterized by a high fixed cost
component, the annualized cost per gallon of throughput is naturally highly
dependent on the volume of the station and the cost of investment capital for
the station. Costs range from about 0.1 cent per gallon for high-volume
direct operations to over one cent per gallon for low-volume open dealer.
Costs for other low volume outlets range from 0.5 to 0.6 cent per gallon
while other medium to high volume outlets have costs ranging from 0.2 to
0.4 cent per gallon.
5-9
-------
5.3.3 Economic Impact of Control Measures
The direct economic effect of vapor recovery at service stations
is to reinforce the existing economies of scale in gasoline marketing.
The competitive position of high volume outlets may be strengthened
since their economics will not be significantly affected. On the other
hand, low volume outlets which are already marginal operations will have
their position eroded even further, even though it is expected that most
of these marginal stations will close in the next five years regardless
of the requirement of vapor recovery due to unfavorable station economics.
While the costs for the balance system are insignificant to the
consumer, the costs are still of appreciable magnitude to the dealer,
who typically has a profit margin of one cent or less per gallon on
gasoline. Thus, in some instances, vapor recovery costs could entirely
wipe out profit margins and in other cases severely reduce the margin by
over 50 percent. In addition, some owners of stations may have difficulty
obtaining the capital necessary to finance the vapor recovery equipment.
Highly leveraged firms may not have the capacity to absorb additional
debt and thus could not obtain loans. This aspect is particularly crucial
to large independent marketers who typically own anywhere from 20 to 100
stations and hence would have to come up with a sizable sum of investment
capital for vapor recovery systems.
It is difficult to segregate the marginal stations which will
eventually survive in the marketplace but would have to close with the
requirement of vapor recovery. In a study conducted for EPA and OSHA by
Arthur D. Little, Inc., an attempt was made to estimate the number of
5-10
-------
closures nationwide which would result due to market forces, due to
capital availability constraints for vapor recovery investment, and
finally due to the impact of the vapor recovery costs on the profitability
of stations.9 The analysis indicated that over 20 percent of the current
population could close by 1981 due solely to market forces, with over
75 percent of closures resulting in the lessee dealer segment of the
market. Vapor recovery requirements, on the other hand, would result in
additional closures representing about six percent of the current population,
or just over 10,000 stations nationwide. Around 12 percent of these vapor
recovery-induced closures would be in the large independent segment of
the market where companies with large numbers of stations would be unable
to obtain the required investment to finance vapor recovery systems at all
their stations. The remainder of the closures would be open dealers for
whom the increased costs would severely reduce or eliminate profit margins
and make staying in business unattractive. The closures in this segment of
the market would represent about 17 percent of the total open dealer
stations.
5.4 GASOLINE BULK PLANTS
5.4.1 Industry Profile
Bulk gasoline loading plants are typically secondary distribution
facilities which receive gasoline from bulk terminals by trailer transports,
store it in above-ground tanks, and subsequently dispense it via account
trucks to local farms, businesses, and service stations. Bulk plants may
be owned by a major or independent petroleum refiner, an independent jobber,
or an individual operator. Although operation and ownership of bulk plants
5-11
-------
include cooperatives and salaried employees, the predominant types are
the independent jobber and the commission agent who operates the plant
for a larger refiner but owns his own delivery trucks.
Currently there are less than 20,000 bulk plants in operation in
the U.S. with almost half having daily throughputs less than 4,000 gallons.
This represents a decline of nearly 4,000 stations facilities from the
population in 1972. This trend is expected to continue as major oil
companies dispose of their many bulk plants as they decline in importance
in gasoline distribution and become less profitable. While bulk plants
served useful pruposes in years past, their role in the distribuiton chain
have declined since more stations are receiving deliveries directly from
bulk terminals. Bulk plants are being bypassed since economies of labor
and capital can be realized if the transport truck can deliver directly
from the terminal to an account, thus reducing the cost of gasoline to
the station by an appreciable amount. Another important factor in the
decline of bulk plants has been and will continue to be a decline in the
customer population served by bulk plants. Small stations and commercial
accounts which once depended upon bulk plants are also undergoing a
significant attrition in the retail market. Even commercial accounts
once served by bulk plants are receiving deliveries directly from terminals,
No estimates are available which indicate what the bulk plant population
will be in the next five years or once the market rationalization process
is completed, though it is expected to be significantly less than the
current population.
5-12
-------
5.4.2 Cost of Reasonably Available Control Measures
Control of breathing, working and miscellaneous losses resulting from
storage and handling of gasoline at bulk plants can be accomplished through
submerged fill, balance systems, vapor processing systems and control of
truck loading leaks. Vapor processing systems have not been applied to
bulk plants, but have been used to recover hydrocarbon vapors at bulk
terminals during truck loading.
By changing from top splash loading to submerged fill, vapors generated
by loading tank trucks can be reduced by about 58 percent. Submerged fill
decreases turbulence, evaporation, and eliminates liquid entrainment. The
cost to install submerged fill at the typical plant with three loading arms
is less than $1,000. This cost is more than offset by the cost savings that
result from the elimination of the generation of vapors during loading.
The vapor balance system operates by transferring vapors displaced from
the receiving tank to the tank being loaded. A vapor line between the truck
and storage tanks essentially creates a closed system permitting the vapor-
spaces of the two tanks to balance with each other. In addition, vapor
balancing of incoming transport trucks displaces vapors from the storage
tanks to truck compartments, with the emissions ultimately being treated at
the terminal with a secondary recovery/control system. The vapor balance
system can reduce emissions from the bulk plant by around 90 percent.
The capital costs for vapor balance systems at bulk plants will vary
depending upon a number of factors, such as tht configuration of the plant,
age and condition of tanks, and requirements for additional equipment due
to local regulations such as fire laws. In areas where these regulations
are less restrictive, vapor baUnce costs are substantially lower. Based
5-13
-------
on an analysis of costs from various sources. EPA estimates the capital costs
for converting to submerged fill and installing the cheaper vapor balance
systems to be around $4,000 for a 4.000 gallon per day (gpd) facility
and $5,000 for a 20,000 gpd plant. Except to a minor extent, these
costs are not a function of throughput of the bulk plant since the
number of tanks is relatively independent of throughput and the number
of delivery trucks serviced is small. The annualized costs for these
model facilities are offset by a credit for gasoline recovery. The
credit, which includes only the savings for the emissions which are not
generated in the first place as opposed to the vapors which are returned
to the bulk terminal for processing, is naturally a function of throughput.
The large bulk plant has a sizeable gasoline recovery credit.
In areas where local regulations prohibit use of the cheaper vapor
balance systems, a more complete and expensive balance system will have
to be installed. Capital costs for converting to submerged fill and
installing the complete balance system would range from about $23,000
for a 4,000 gpd facility to close to $26,000 for a 20,000 gpd plant.
The annualized costs for these model facilities amount to around 53,500 for the
small plant (or about 0.3 cents per gallon) and about $750 for the large
plant (less than 0.1 cent per gallon).
5.4.3 Economic Impact of Control Measures
The economic impact on bulk plant operators due to vapor recovery
requirements depends on the system which can be installed at the plant.
If the cheaper balance systems can be used, the capital costs are not
of such magnitude as to cause a significant impact in the industry.
5- It
-------
An economic impact analysis is being conducted by EPA which will quantify
the potential impacts which could result from the range of vapor recovery
system costs.
The magnitude of the costs for the more expensive balance systems
will likely have a more-significant impact on bulk plants. A capital
outlay of around $23,000 represents a significant investment for small
plants. Some sources have reported sales prices for bulk plants that
have been sold by major oil companies which range between $45,000 and
$65,000. Thus, investment for the expensive vapor balance systems
represents one-third to one-half of these transaction prices.10 A
preliminary economic analysis prepared for EPA concludes that bulk plants
having throughputs less than 4,000 pgd are either unprofitable or only
marginally profitable and could possibly be unable to cope with an
expenditure of this magnitude. EPA's current analysis will better
determine the extent of the potential impacts of these costs on small
bulk plants.
When assessing impacts on bulk plants or potential closures of
plants, it is important to consider several points. First, until SIP
revisions are submitted, enforcement discretion is being utilized to
prevent closures from occurring due to pollution control requirements.
Guidance has been developed and provided to the states in developing
the SIP revisions. Secondly, not all bulk plants in the country will
be affected by vapor recovery regulations. Only bulk plants in
5-1-5
-------
non-attainment areas will be required to install controls. Since bulk
plants are concentrated in rural areas, it is likely that a large number
of bulk plants would not be required to install vapor recovery systems.
Finally, most of the bulk plants that could be closed by vapor recovery
requirements are likely to go out of business or dispose of their gasoline
operations in the near future due solely to market forces. Hence, vapor
recovery requirements could only accelerate closures that will take place
even in the absence of such requirements.
5.5 AUTOMOBILE ASSEMBLY PLANTS
5.5.1 Industry Profile
In the model year 1976, over 8.5 million automobiles were sold by
U.S. automakers. They represented a significant rebound from 1974 and
1975 sales levels when, respectively, only 7.3 and 6.7 million cars were
sold. General Motors and Ford Motor Company dominate the industry as the
two companies accounted for 58 and 24 percent of the autos produced in
1976, respectively. The other two major automakers, Chrysler and American
12
Motors, accounted for 16 and three percent, respectively.
There are currently 46 auto assembly plants in the U.S., though this
number can vary due to temporary shutdowns and switchovers to light-duty
truck assembly. GM has 22 of the plants while Ford has 14. The remainder
of the plants are owned by Chrysler and AMC as well as Checker Motors and
Volkswagen, which is opening a new plant in Pennsylvania. The locations of
these plants are indicated in Table 5-2. Essentially all of the plants are
located in non-attainment areas for oxidants.
5-15
-------
Table 5r2.
Manufacturer
American Motors
Chrysler Corp.
Ford Motor Co.
General Motors
Checker Motors
Volkswagen
U.S. AUTOMOBILE ASSEMBLY PLANTS
Location
Kenosha, Wisconsin
Toledo, Ohio
Belvidere, Illinois
Hamtramck, Michigan
Detroit, Michigan
Newark, Delaware
St. Louis, Missouri
Atlanta, Georgia
Chicago, Illinois
Dearborn, Michigan
Kansas City, Missouri
Lorain, Ohio
Los Angeles, Calif.
Mahwah, New Jersey
Metuchen, New Jersey
St. Louis, Missouri
San Jose, California
Twin Cities, Minnesota
Wayne, Michigan
Wixom, Michigan
Arlington, Texas
Baltimore, maryland
Detroit, Michigan
Doraville, Georgia
Fairfax, Kansas
Flint, Michigan
Framingham, Mass.
Fremont, California
Janesville, Wisconsin
Lakewood, Georgia
Lansing, Michigan
Leeds, Missouri
Linden, New Jersey
Lordstown, Ohio
Norwood, Ohio
Pontiac, Michigan
St. Louis, Missouri
South Gate, Calif.
Tarrytown, New York
Van Nuys, California
Willow Run, Michigan
Wilmington, Delaware
Kalamazoo, Michigan
Pennsylvania
5-17
-------
The earnings of the automakers sagged in 1974 and 1975 due to reduced
sales levels. However, in 1976 earnings expressed as return on equity or
return on assets returned to historical levels, though American Motors is
still experiencing financial difficulties.
5.5.2 Cost of Reasonably Available Control Measures
For the paint coating of auto bodies at assembly plants, numerous
options exist for the control of VOC emissions, with control ranging
from 70 to 95 percent. Options potentially consist of process changes,
such as electrodeposition (EDP) of the primecoat and water-borne topcoats,
and add-on control devices such as carbon adsorption, thermal incineration,
and catalytic incineration. For purposes of this study, the most cost-effective
control options for prime and top coating were chosen that resulted in at
least 80 percent control. Other control options could possibly be chosen
in actual existing plants, but this analysis considers only the least costly
option, based on costs furnished to EPA by Springborn Laboratories. The
following control option was chosen:
• Prime coating: EDP with water-borne dip and solvent guide coat
* Prime and top coat spray booths: Catalytic incineration with
primary heat exchange
* Prime and top coat ovens: Catalytic incineration with primary
heat exchange.
Add-on controls in addition to EDP are needed for the prime coating
operation in order to control emissions from the application of the solvent
guide coat.
5-18
-------
Costs for these options have been estimated for a "model" auto
assembly plant producing 211,200 bodies per year. The capital cost of
converting the plant to EDP and adding the ccitrol devices is estimated to
be about $20.2 million, with $15.7 million resulting from the conversion
to EDP, $0.8 million from the prime coat add-on devices, and $3.7 million
from the top coat add-on controls.
Annual!zed costs have also been estimated taking into account operating
and maintenance costs of the processes and devices as well as the depreciation
and interest charges. Only the incremental O&M costs incurred over the
existing base case (solvent-borne prime and otp coats with no control) are
included in the estimates. However, the entire capital charges of the new
processes and devices are included since it is assumed that the existing
equipment has no salvage value. Salvage values will vary significantly
from plant to plant and thus it is difficult to generalize on an appropriate
value. Based on these assumptions, the increased annualized cost of control
is estimated to be almost $34 per car.
5.5.3 Economic Impact of Control Measures
EPA is conducting but has not completed a formal study of the economic
impact of these controls on the automobile industry. However, tentative
conclusions can be drawn.
Currently, about 60 percent of the assembly plants employ EDP to
apply prime coats. Since this process change contributes almost half of
the annualized cost per body, many of the existing plants will be able to
5-19
-------
achieve the required additional control for about $18 per car. In
addition, due to the fact that such a large portion of the industry has
already moved to EDP for economic and technical reasons the economic impact
of such a switch for the remainder of the industry should not be unduly
burdensome.
The impact on sales of automobiles is not expected to be significant.
Rough estimates based on the Ford Econometric Sales Forecasting Model indicate
that a $34 increase in the cost of the average automobile could result in a
reduction in sales of 0.2 percent in 1983.15 This is not a reduction in
sales from current levels, but rather a reduction in levels that would
otherwise occur in 1983. Such a reduction in foregone sales will have a
negligible effect on the return on investment in the industry.
It is important to remember that these general conclusions are based
on model plants and average conditions in the industry. Though none of
the major firms are expected to experience serious impacts, some individual
plants will experience more costly conversions to alternative processes
which could affect their viability. A determination of the individual
plants which have the potential to be severely impacted has not been
determined and is beyond the scope of this analysis.
5.6 METAL FURNITURE INDUSTRY
5.6.1 Industry Profile
The metal furniture industry consists of about 1600 firms employing
nearly 100,000 people and producing around $3.4 billion in metal furniture
5-20
-------
shipments in 1975.16 The industry is highly fragmented, including the
following categories of products: household metal furniture, office metal
furniture, public building furniture, and metal partitions and fixtures.
Around 500 firms manufacture household furniture, another 500 manufacture
partitions and fixtures, over 400 produce public building furniture, and
200 firms engage in office furniture manufacture.
The industry is characterized by relatively small manufacturers.
Whereas single-unit firms, those with one establishment for both manu-
facturing and administration, account for only 19 percent of the value of
shipments for all manufacturing establishments listed by the Census of
Manufacturers, such firms account for over double the average for household
furniture, public building furniture, and partitions and fixtures. Only
the office furniture segment of the market is consistent with the overall
industry average. In addition, over 85 percent of all metal furniture
establishments employ less than 100 people. In fact, more than 50 percent
of the establishments in segments other than office furniture have less
than 20 employees, according to the 1972 Census of Manufacturers. The
significant number of small firms indicates that no economies of scale
are evident which prohibit small manufacturers from competing, especially
in regional markets where low labor productivity may be overcome by lower
distribution costs.
The manufacturing markets of metal furniture facilities vary. Some
plants manufacture furniture to be sold directly to consumers through
5-21
-------
retail stores. In contrast, job shops, which produce furniture on a
contract basis, apply coatings on many different furniture pieces according
to the customer's specifications.
Metal furniture plants are located throughout the U.S. However, the
states of Illinois, California, Michigan, New York, and Pennsylvania
contain over 50 percent of the establishments in the industry.
5.6.2 Costs of Reasonably Available Control Measures
Measures to reduce volatile organic emissions from metal furniture
coating operations consist of process changes as well as exhaust gas
treatment with add-on control devices. Applicable process changes include
conversion to waterborne coatings, high solids coatings, powder coatings,
and electrodeposition (EDP) of waterborne coatings. Add-on control devices
include carbon adsorption and incineration. While each of these measures
achieves reasonable levels of control, the option chosen by an individual
plant will depend upon circumstances specific to the plant.
EPA has estimated costs of the alternative measures on the basis of
18
model plants in order to indicate the relative costs of the alternatives.
These models are one-color lines and are sized based on the annual product
coverage rates for the coating lines.
For electrostatic spray lines, the most feasible control option appears
to be conversion to high solids coatings in order to reduce solvent emissions.
The reduction can range from 50 to 90 percent depending upon the type of
coating used previously. For a three million square feet per year coating
line, the capital cost to convert the line is approximately $15,000, while
5-22
-------
for a large plant (48 million square ft/yr) the capital cost will approximate
$62,000. These costs represent a five to six percent increase in the invest-
ment in the existing line. However, in both cases, the increased capital
costs appear to be justified on economic grounds due mainly to the savings
in lower applied film cost when compared to conventional solvent coatings.
For the larger plant, the cost savings in the first year offset the capital
cost entirely, while for the smaller plant the savings represent a return
on investment approaching 20 percent1.
Conversion to waterborne coatings, appears to be the most feasible option
for dip coating lines. Switching to waterborne coatings would entail a
capital investment of $3,000 for a smaller facility (seven million square
feet/year) and $5,000 for a larger plant (22.5 million square feet/year).
These costs represent an increase in investment of two to three percent.
There is an increase in operating and maintenance costs due to higher
materials costs, resulting in an increase in coating costs of seven percent
for the smaller plant and four percent for the larger facility.
5.6.3 Economic Impact of Control Measures
An analysis of the economic impact of these costs on the segments of
the metal furniture industry has not been conducted by EPA, thus no
definitive conclusions can be drawn. However, it appears from the model
plant analysis that conversion to high solids coatings for electrostatic
spray lines is justified from an economic standpoint and would be of
benefit to that portion of the industry utilizing this coating application
method.
5-23
-------
On the other hand, conversion to waterborne coating for dip coating
lines is more difficult to assess since one has to consider both the capital
investment requirements as well as the increased cost of coating on an
annual basis. With regard to this latter point, the 1972 Census of
Manufacturers indicates that the cost of coating materials comprise
19
0.8 to 1.4 percent of the value of shipments for metal furniture. An
increase of four to seven percent in coating costs resulting from the
conversion to EDP will affect the final selling price for the metal furniture
by an insignificant amount (less than 0.1 percent). In addition, a capital
investment of $3,000 to $5,000 does not appear to be burdensome for most
establishments. Some small, marginal facilities may find such an investment
level unjustified, but the extent of this is not known.
5-24
-------
5.7 REFERENCES FOR CHAPTER 5
1. Energy and Environmental Analysis, Inc. Estimated Cost of Benzene
Control for Selected Stationary Sources,' February 27, 1978, p. 3
2. Environmental Protection Agency, Control of Refinery Vacuum Producing
Systems. Wastewater Separators, and Process Unit Turnarounds,
EPA-450/77-025, October 1977, p. 4-10.
3. Environmental Protection Agency, Control of VOC from Petroleum
Refinery Equipment, Draft, April 1978, p. 4-7.
4. Energy and Environmental Analysis, p. 10.
5. Environmental Protection Agency, Control of Refinery Vacuum Producing
Systems. Vlastewater Separators, and Process Unit Turnarounds,
EPA-450/77-025, October 1977, p. 5-3.
6. Sobotka and Company, Inc., Economic Impact of EPA's Regulations on
the Petroleum Refining Industry, EPA-230/3-76-004, April 1976.
7. Arthur D. Little, Inc., The Economic Impact of Vapor Recovery Regulations
On the Service Station Industry, Report prepared for EPA and OSHA,
March 1978, p. 6.
8. Lloyd, Kenneth H., "Cost of Alternative Vapor Recovery Systems at
Service Stations," Economic Analysis Branch, OAQPS, EPA, October 20, 1977.
9. Arthur D. Little, p. 131.
10. Sobotka and Company, Inc., Bulk Plant Vapor Controls Economic Impact,
August 15, 1977, p. 3.
11. Pacific Environmental Services, Inc., Economic Analysis of Vapor
Recovery Systems on Small Bulk Plants, September 1976, p. 2-2.
12. Motor Vehicle Manufacturers Association, Motor Vehicle Facts & Figures '77.
pp. 8 and 9.
13. Springborn Laboratories, Inc., Study to Support New Source Performance
Standards for Automobile and Light-duty Truck Coating, June 1977,
EPA-450/3-77-020, p. 35.
14. Ibid, pp. 8-42 and 8-43.
15. Personal communication from Tom Alexander, OPE/EPA to Ken Lloyd
OAQPS/EPA, March 13, 1978.
5-25
-------
16. Springborn Laboratories, Inc., Study to Support New Source Performance
Standards for Surface Coating of Metal Furniture, EPA-450/3-78-006,
April 1978, p. 8-1.
17. Springborn, p. 8-15.
18. Environmental Protection Agency, Control of Volatile Organic Emissions
from Existing Stationary Sources - Vol. Ill: Surface Coating of Metal
Furniture. December 1977, pp. 3-8 and 3-9.
19. Springborn, p. 8-22.
5-26
-------
APPENDICES
-------
APPENDIX A
Ozone Design Values for 90 Air Quality Control Regions
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
DATE. MAY 2 4 19/8
SUBJECT Ozone Design Values for 90 Air Quality Control Regions (AQCR's)
FROM Robert E. Neligan, Director <
Monitoring and Data Analysis bivision
TO Bruce Jordan
Environmental Protection Specialist, OAQPS
ije have assembled ozone design values for 90 AQCR's based or
the three year period 1975 - 1977. (See Table 1) This list updates
previous lists which used 1974 and earlier data. Using three years
of data the design value should fall between the third and fourth
highest hourly averages based on the guidance for determining com-
pliance with the statistical form of the ozone NAAQS. In selecting
the enclosed design values we used the fourth highest hourly averages
over the three year period, unless the difference between the third
and fourth highest values exceeded .01 ppm (20 jjg/m3) in which case
we took the average of the 3rd and 4th highest values.
Table 1 lists the following:
(1) the AQCR;
(2) the site which produced the old design value;
(3} the old design value (the j?nd maximum hour) in ppm
based on the 1971 - 1976 trime period;
(4) the year in which the value occurred;
(5) the site which measured the new design value;
(6) the new design value in ppm based on the 1975 - 1977;
(7) the year ir, v;hich the new design value occurred.
The new design value is lower in 53 AQCR's than the old design
value; higher in 19 AQCR's and 18 show no change. In the 19 AQCR's
showing an increase in design values, 12 increased by .01 ppm, 6 by
.02, and 1 by .03 ppm. In the 53 AQCR's showing a decrease in design
values, 18 decreased by .01 ppm, 10 by .02 ppm, 10 by .03 ppm, 3 by
.04 ppm, and 10 by .05 ppm or more. The largest decrease occurred in
AQCR 105. Southeastern LA. - S.E. Texas, which had a design value of
.32 ppm based on 1974 data. The design value has been revised to .18
pp-n based on the 1975 - 1977 period.
below:
Finally, compliance with several possible standards is summarized
No. of AQCR's
in
Non-attainment
OLD STD.
.08 ppm
2nd max
89
EXPECTED VALUE STD
.08 .10 .12
ppm ppin ppm
87 84 74
•iv. r.
i t. IH- v 1 «.l
A-l
-------
If you have any questions regarding this list, feel free to contact
Robert B. Faoro, of my staff, at 541-5351.
Enclosure
cc: H. Jones, SASD
E. Lillis, HDAD
0. O'Connor, SASD
K. Lloyd, MDAD
W. Barber, OAQPS
A-2
-------
TABLE 1. OZONE DESIGN VALUES FOR 90 NON-ATTAINMENT AIR QUALITY CONTROL REGIONS (AQCRs) BASED ON THE 1975-1977 TIME PERIOD
ACCR NUMBER
AtjiK NAME
OLD DESIGN VALUE BASED ON
2D MAX., 1971-1976
SITE VALUE YEAR
PPM
NEW DESIGN VALUE BASED ON THE 4TH
HIGHEST VALUE, 1975-19773
SITE VALUE YEAR
PPM
002
004
005
013
015
016
018
022
024
025
028
029
030
031
033
036
038
041
042
043
045
047
048
049
050
052
055
Columbus-Phoenix, GA
Metropolitan Birmingham
Mobile-Pensacola, AL-FL
Clark-Mohave, AZ-NV
Phoenix-Tucson, AZ
Central Arkansas
Metropolitan Memphis,
AR-MS-TN
Shreveport, LA
Metropolitan Los Angeles
North Central Coast, CA
Sacramento Valley, CA
San Diego, CA
San Francisco, CA
San Joaquin Valley, CA
Southeast Desert, CA
Metropolitan Denver, CO
San Isabel, CO
Eastern Connecticut
Hartford-New Haven-;
Springfield, CT-MA
New Jersey-New York-
Connectiuct (NJ-NY-CT
Metropolitan Philadelphia,
N.J.-PA
National Capital
(DC-MD-VA)
Central Florida
Jacksonville-Brunswick,
(FL-GA)
Southeast Florida
West Central Florida
Chattanooga, TN-GA
no data
011300003
012380011
290320009
030600006
G01
GDI
G01
G01
Summer Study
442340021
GO!
Summer Study
058440003
054860001
056580003
055320003
053140001
052820001
050560001
062210001
060380004
070200001
070570003
070060123
*
391080012
211560001
104900002
101960048
104760001
103980012
440380024
101
'101
F01
101
101
101
101
F01
F01
F03
F01
F01
F01
F01
F01
HOI
G01
HOI
G01
.15 est
.14
.14
.17
.18
.15
.13
.14
.44
.12
.18
.27
.25
.27
.26
.25
.10
.23
•32.
.26
.32
.23
.10
.19
.14
.18
.12
.
1975
1974
1973
1974
1977
1975
1977
1974
1974
1975
1974
1974
1974
1972
1974-
1975
1975
1974
1975
1975
1975
1976
1974
1976
1974
1975
no data
same
103540004
290320001
030600002
041880002
same
452180001
same
050275001
056600001
052460002
same
052820001
050560002
060120002
same
070350123
070700123
same
same
480080009
same
101960055
same
104360035
same
F01
G01
GO!
P05
F03
101
101
101
101
101
F01
F01
F01
HOI
HOI
G02
,15 est
.15
.16b
.15°
.14
.13
.14
.15
.38
.12
.19
.24
.19
.19
.23
.17
.09
.23
.
.32b
/»-.k
.27°
.30
_ _ h
.21°
t f\
.10
.12
• ^k
1 ^D
• 1 */
.11
_
1977
1977
1975
1977
1977
1976
1977
1976
1976
1975
1976
1975
1975
1976
1976
1975
1977
1977
1975
1 n Tf f*
1975
1976
Irt^ *T
977
1975
1f\^ f
976
1975
Irt T /"
976
-------
TABLE 1, OZONE DESIGN VALUES FOR 90 NON-ATTAINMENT AIR QUALITY CONTROL REGIONS (AQCRs) BASED ON THE 1975-1977 TIME PERIOD
AQCR KUK3ER
AQCR NAME
OLD DESIGN VALUE BASED ON
2D MAX., 1971-1976
SITE VALUE YEAR
PPM
NEW DESIGN VALUE BASED ON THE 4TH
HIGHEST VALUE, 1975-197?
SITE VALUE YEAR
PPM
056
060
052
065
057
059
070
073
078
079
080
081
082
085
092
094
099
106
113
115
118
Metropolitan Atlanta, GA
State of Hawaii
Eastern Washington-
Northern Idaho
Burlington-Keokut, IA
i-'.etropolitan Chicago, IL
Metropolitan Quad Cities,
IL-IA
Metropolitan St. Louis,
IL-MO
Rockf ord- Janes vi lie-
Bel oit, IL-WI
Metropolitan Louisville, KY
Metropolitan Cincinnati,
KY-OH
Metropolitan Indianapolis,
IN
Northeast Indiana
South Bend-El khart-
Benton Harbor, IN-MI
Metropolitan Omaha-
Council Bluff, IA-NE
South Central Iowa
Metropolitan Kansas City,
KS-MO
South Central Kansas
Southern Louisiana-
Southeast Texas
Cumber! and-Keyser,MD-WV
Metropolitan Baltimore, MD
Central Massachusetts
111600002 F01
120120001 F01
492040012 F01
146080024 F01
"HI 220025 HOI
163280010 F05
260200002 G01
146680005 F01
182380021 G01
362720006 HOI
152040022 HOI
no data
no data
281880026 G01
161180037 G02
262380022 HOI
173740011 F01
453830003 F01
210800004 A05
210680001 G01
222640012 F01
.16
.08
.09
.10
.23
.11
.23
.18
.23
.21
.15
.17 est
.16 est
.11
.10
.15
.29
.32
.17
.26
.19
1975
1974
1975
1976
1974
1975
1975
1975
1975
1975
1976
-
-
1975
1976
1975
1975
1973
1974
1975
1976
same
same
same
same
148020002 F01
146700002 F01
264200061 HOI
361260001 G01
same
same
152040033 HOI
-
-
same
same
same
173740010 F01
same
same
same
same
.15
<.08
<.08
.12
.26
.13
L
.23b
•17K
.22b
.20
.17
.17 est
.16 est
.10
.11
.12
.17
.19
.12
f\ H
.25
.16
1975
—
1977
1977
1975
1975
1977
1975
1975
1977
-
-
1976
1977
1975
1975
1976
1976
1976
1976
-------
TABLE 1. OZONE DESIGN VALUES FOR 90 NON-ATTAINMENT AIR QUALITY CONTROL REGIONS (AQCRs) BASED ON THE 1975-1977 TIME PERIOD
AQCR NUMBER
AQCR NAME
OLD DESIGN VALUE BASED ON
20 MAX., 1971-1976
SITE VALUE YEAR
PPM
NEW DESIGN VALUE
HIGHEST VALUE,
SITE
BASED ON THE 4TH
1975-1977 a
VALUE YEAR
PPM
119
120
121
122
123
124
125
128
131
151
152
153
158
160
161
162
167
173
174
176
178
184
186
193
Metropolitan Boston, MA
Metropolitan Providence,
MA-RI
Merrimack Valley-Southern
New Hampshire, MA-NH
Central Michigan
Metropolitan Detroit-
Port Huron, MI
Metropolitan Toldeo,MI-OH
South Central Michigan
Southeast Minnesota-La
Crosse, MN-WI
Minneapolis-St. Paul , MN
222340003 F01
220580004 F01
222467001 F01
231580011 HOI
231180020 G01
336000006 HOI
no data
243120019 G05
243300030 HOI
.20
.20
.20
.19
.26
.15
.17 est
.17
.12
1976
1974
1976
1975
1975
1976
-
1975
1975
same
410140002 F03
same
same
same
same
232840007 F01
same
same
.17b
.
.19°
•17b
.17b
h
.23°
.14
.08
.16
.12
Northeast Pennsylvania-Upper
Delaware Valley, PA-NJ-DE
Albuquerque-Mid Riod
Grande, NM
El Paso-Las Cruces-
Alamogordo, NM-TX
Central New York
Genesse-Finger Lake, NY
Hudson Valley, NY
Niagara Frontier, NY
Charlotte, NC
Dayton, OH
Greater Metropolitan
Cleveland, OH
Metropolitan Columbus, OH
Northwest Pennsylvania-
Youngs town, OH-PA
Central Oklahoma
Northeastern Oklahoma
Portland, OR-WA
397620009 F01
320040017 H02
451700028 F01
336620005 F01
2701-08
(State Code)
336020003 F01
330130002 F01
340200011 G01
361660019 G01
361300034 HOI
361460004 F01
367760007 101
372200033 F01
373000127 F02
381200001 F01
.25
.13
.16
.11
.13'
.18
.21
.16
.18
.17
.16
.21
.14
.20
.15
1974
1976
1975
1975
1976
1973
1975
1975
1976
1975
1976
1975
1976
1976
1975
390780017 F01
320040015 H02
same
same
335760004 F01
333500002 F01
same
340700028 GDI
saint:
365320002 G02
same
same
same
same
381580011 F01
.23
.14
.16
.12
.12
.14
.18&
.17"
.18
U
.19b
.16
ft m
.21
.12
.18
.16
1976
1976
1976
1975
1976
1976
1977
1975
1975
1975
1977
1976
1976
1976
1976
1975
1976
1977
1977
1976
1 AT C
1975
• AT C.
1976
1f\ T /•
976
1977
-------
TABLE }.• OZONE DESIGN VALUES FOR 90 NON-ATTAINMENT AIR QUALITY CONTROL REGIONS (AQCRs) BASED ON THE 1975-1977 TIME PERIOD
AQCR NUMBER
AQCR NAME
OLD DESIGN VALUE BASED ON
2D MAX., 1971-1976
SITE VALUE YEAR
PPM
t>
I
NEW DESIGN VALUE BASED ON THE 4TH
HIGHEST VALUE, 1975-1977 a
SITE VALUE YEAR
PPM
195
196
197
199
200
208
212
214
215
216
217
220
223
225
229
230
239
240
Central Pennsylvania
South Central Pennsylvania
Southwest Pennsylvania
Charleston, SC
Columbia, SC
Middle Tennessee
Austin-Waco, TX
Corpus Christi-Victoria.TX
Metropolitan Dallas-
Fort Worth, TX
Metropolitan Houston-
Galveston, TX
Metropolitan San Antonio.TX
Wasatch Front, UT
Hampton Roads, VA
State Capital, VA
Puget Sound, WA
South Central Washington
Southeastern Wisconsin
Southern Wisconsin
394460011 F01
399560008 F01
391560005 F01
no data
421900003 F01
442540011 G0>
450220012 F01
451150001 P0*l
451 31 0045. F01
452330024 F01
454570036 F01
460060001 F01
481440004 F02
481500010 F01
490960001 101
492190003 F01 •
512200041 F01
510600001 F03
.17
.19
.21
.14 est
.14
.20
.14
.15
.19
.30 ;
.18
.17
.18
.18
.13
.15
.26
.13
1975
1975
1975
_
1974
1975
1976
1974
1974
1975
1976
1976
1974
1975
1974
1974
1976
1976
same
same
same
no data
same
443320007 F01
same
455340002 P05
same
same
same
same
482060002 F01
same
490980010 F05
no data
same
510600005 F02
.15
.19
.20
.14 est
1C
• 1 «J
.17
.13
.14
.19
.27
.16
.16
.14
.20
.14
.15
.25
.13
1975
1975
1975
1077
i y/ /
1975
1977
1977
1977
1976
1975
1977
1977
1977
1974
1976
1977
aThe fourth highest hourly average over the 3-year period, 1975-1977, was used unless the difference between the. third and
fourth highest values exceeded .01 ppm (20'yg/m3) in which case the average of the 3rd and 4th highest values was used.
^These values represent the average of the 3rd and 4th highest values.
-------
APPENDIX B
Mobile Source Emission Factors
-------
APPENDIX B
MOBILE SOURCE EMISSION FACTORS
This appendix summarizes the mobile source emission factors which
serve as the basis for emission projections resulting from the Federal
Motor Vehicle Control Program as well as an inspection/maintenance
program. Table B-l presents the non-methane hydrocarbon (NMHC) emission
factors for various classes of vehicles for base years 1975 and 1987.
The 1987 emission factors reflect the -emission standards mandated by the
Clean Air Act Amendments of 1977.
In order to estimate the change .in mobile source emissions for an
area between 1975 and 1987, the following equation is used:
1987 Emissions = 1975 Emissions X {975 Emission Factor X (1 + Annual Growth
Estimates have also been made'regarding the effectiveness of an
inspection/maintenance program in reducing mobile source emissions through
overall improvement in fleet maintenance. Emission factors for light duty
vehicles for various I/M scenarios are summarized in Table B-2. As can be
seen, the effectiveness of an I/M program depends on the stringency level
and the extent of mechanic training. For a 30 to 40 percent stringency
level, an I/M program can reduce emissions in 1987 by 25 to 44 percent
depending on whether there is mechanic training. For purposes of this
study, an emission reduction of 30 percent was assumed to reflect a mid-range
of the estimates.
B-l
-------
Table B-l. NON-METHANE HYDROCARBON EMISSION FACTORS FOR
MOBILE SOURCES3 (grams/mile)
1975 1987
Light Duty Vehicles
(without I/M) 8.27 2,24
Light Duty Trucks
0-6000 Ibs 8.98 3.71
6000-8500 Ibs 12.22 4.99
Heavy Duty Gasoline Trucks 29.99 14.28
Heavy Duty Diesel Trucks 4.42 3.41
Motorcycles 12.07 1.40
Includes evaporative emissions
Source: Environmental Protection Agency, Office of Transportation and
Land Use Policy, Mobile Source Emission Factors. EPA-400/9-78-005
March 1978.
B-2
-------
Table B-2. EFFECTIVENESS OF INSPECTION/MAINTENANCE PROGRAMS
FOR LIGHT-DUTY VEHICLES (1987 NMHC Emission factors)
Grams /Mile
% Reduction
from Base
Base
(w/o I/MJ
2.24
—
30% Stringency Level*
No Mechanic
Training
1.68
25%
Mechanic
Training
1.33
41%
40% Stringent
No Mechanise _
Training ""*
1.60
29%
:y Level*
Mechanic
Training
1.26
44%
*Stringency level is a measure of the rigor of a program based on the estimated
fraction of the vehicle population whose emissions would exceed cutpoints for
NMHC were no improvements in maintenance habits or quality of maintenance to
take place as a result of the program.
Source: Based on Appendix N to 40 CFR Part 51: Emission Reductions Achievable
Through Inspection and Maintenance of Light Duty Vehicles, Mororcyjrje?,
and Light and Heavy Duty Trucks. May 1977.
-------
APPtNUlA t
Analysis of Costs for Hydrocarbon Cnntrnl
-------
Appendix C
ANALYSIS OF COSTS FOR HYDROCARBON CONTROL MEASURES
C.I. Introduction
This analysis presents the costs for selected hydrocarbon control
measures for stationary and mobile sources. The costs are derived f»- on-
numerous EPA reports and documents and represents the Agency's best
estimates of costs as of July, 1977. The sources covered by this analysis
are not the only sources of volatile organic emissions, rather they are
tne sources for which cost -information is readily available . SCIT.C arc
described more completely than others, with the extent of coverage
depending on the availability of information for each source.
-2. Csts •fOy' Stationar c'mtvrp r.nnt.rnl
Tables C-l through C-7 summarize the control costs for selected
stationary sources. The methodology for estimating costs for most sources
involved selecting model facilities of a size or sizes considered typiral
in the industry. For these model facilities, capital costs for the
control equipment were developed which included equipment costs as well
as installation costs.
The annuali zed costs for each model facility include direct
operating costs such as labor and materials, maintenance costs, and
annuali zed capital charges. This latter component accounts for depre-
ciation, interest, administrative overhead, property taxes, and insurance.
The depreciation and interest are computed by use of a capital recovery
C-l
-------
factor, the value of which depends on the operating life of the device
and the interest rate (in most cases, an annual interest rate of 10
percent has been assumed).
In many instances, the annualized costs also include a credit for
product, heat or steam recovery. These credits are substracted from the
costs of control so that the annualized costs included in the. tables
are net costs.
The cost effectiveness of control represents the net annualized
costs divided by the annual tons of hydrocarbons removed. While cost effective-
ness serves a useful purpose as one factor in comparing control measures,
it cannot serve as the only decision-making tool. Cost effectiveness
in itself does not give anyindication of the economic feasibility of
alternatives since it does not take into account the baseline economic
or financial conditions of the source or industry. However, such an
evaluation was beyond the scope of this study, so that cost effectiveness
is the only means available by which to compare control measures. None-
theless, the limitations should be recognized.
Finally, there are in many cases several control measures available
to achieve a certain level of control at the various sources. However,
this study considers costs for only one control measure at each control
level. In choosing the measure, the assumption is made that a prudent
plant manager will choose the lowest cost option on an annualized cost.
basis. In addition, If a higher level of control can be achieved at a
lower annualized cost than a lower level of control, then the latter is
not considered to be a viable option. Thus some of the control measures
C-2
-------
considered may not be as energy efficient as others or recover the
end product, but they are still the least cost options and have there-
fore been chosen for inclusion in the tables.
C.2.1. Oil and Gas Production. Refining, and Storage
Table C-l presents costs of controlling hydrocarbon emissions from
selected sources associated with oil and gas production, petroleum
refining, petroleum storage tanks, and natural gas and gasoline processing
plants. Exact cost figures are not yet available for these sources,
but preliminary consideration of costs indicates that the net costs will
be minimal since the costs of control will be,for the most part, offset
by appreciable savings from product recovery Tn fart, much of the control
equipment is already in operation in many plants or fields, indicating
that the controls must be justifiable from a cost standpoint in many
cases.
C.2.2 Gasoline Handling and Distribution Operations
The costs for controlling hydrocarbon vapor emissions from selected
gasoline handling and distribution operations are shown in Table C-2.
These operations trace the flow of gasoline and resultant hydrocarbon
vapors from the bulk terminal to the bulk plant to the service stations
and finally to the refueling of vehicles. As can be seen from the table,
relative control costs are much higher at smaller facilities as evidenced
by the significantly greater cost effectiveness numbers.
At service stations there are two sources of vapor loss—the under-
ground tanks (Stage I) and vehicle refueling (Stage II). Stage I can
br implemented alone at service stations but Stage II cannot since
C-3
-------
vapors captured during refueling would be lost through the underground
tanks that had no control. Thus, costs are presented for Stage I
alone and Stages I and II In conjunction. The cost effectiveness
numbers for Stages I and II take Into account the total amount of vapors
collected during the two stages Instead of just the Incremental cost of
controlling Stage II emissions.
C.2.3. Surface Coating Operations
Table C-3 summarizes the costs for controlling volatile organic
emissions from selected surface coating operations. Except for the
primer application area and curing oven in automobile and light duty
truck assembly plants, which utilize a process change in applying the
primer, the control measures- considered for these sources
are add-on technology to destroy or recover organic compounds
front exhaust gases. Thermal incineration, which destroys organic emis-
sions, is considered for all of these processes instead of catalytic
incineration because of the variability in the use of the latter.
Catalytic incineration is limited to a more restricted range of applica-
tions as a result of problems with catalyst deactivation, coating with
particulates, poisoning, and type of fuel used (ref. 12, pp. 54-55).
Where applicable, though, catalytic units offer the potential of
significantly lower fuel consumption and smaller, lighter-weight units.
However, it was beyond the scope of this study to judge the effectiveness
of the option in all of the processes considered, so thermal incineration
was used since Us use is not as limited, even though the costs may be
higher.
C-4
-------
Costs for incineration include savings from the use of primary
heat recovery. Primary heat recovery involves the use of incinerator
exhaust to preheat incinerator inlet air instead of using expensive
auxiliary fuel. Further heat recovery, called secondary heat recovery,
is possible in some applications where incinerator exhaust from the
primary heat recovery stage (or from the incinerator directly if there
is no primary heat recovery) replaces energy usage elsewhere in the
plant. This energy can be used for process heat requirements or for
plant heating. However, credits for secondary heat recovery have not
been included in this analysis since the amount of energy that a plant can
recover and use depends on the individual circumstances of the plant
(ref. 12, p. 45).
Carbon adsorption, the other add-on technology considered, separates
and recovers organic vapors from the exhaust stream but is not as efficient
as incineration in general. The annual!zed costs for this control measure
includes a credit for the recovered solvent at its fuel value, which is
lower than the solvent's market value. Nonetheless, the market value of
the recovered solvent has not been used in general because reuse of the
solvent may not be feasible due to mixture of solvents or breakdown
of single solvents. Distillation is possible, but the complexity and
cost are so variable that it is difficult to generalize (ref. 12, p. 32).
Thus, the fuel value of the solvent has been used as a conservative
figure.
C-5
-------
C.2.4. Graphic Arts Processes
Costs of control measures for selected graphic arts processes are
tabulated in Table C-4. Once again, the techniques considered are
incineration with primary heat recovery and carbon adsorption with
solvent recovery credited at the fuel value of the solvent.
C.2.5. Pegreasing, Dry Cleaning, and Cutback Asphalt Paving
Table C-5 presents control costs for selected sources of evaporative
emissions of organic vapors. One such category of sources is organic
solvent metal cleaning, which includes cold cleaners, open top vapor
degreasers, and conveyorized degreasers. Dry cleaning operations include
neighborhood and industrial petroleum solvent plants as well as coin-op,
commercial.^and industrial perchloroethylene solvent plants. Finally,
the control measure for cutback asphalt paving entails the substitution
of emulsified asphalt in order to eliminate evaporative emissions.
C.2.6. Rubber Products Manufacture
In Table C-6 costs are summarized for selected processes in the
manufacture of various rubber products. The manufacture of tires and
inner tubes represents the largest source of emissions in this category
with over 50 percent of the total. Thermal incineration and carbon
adsorption are the primary control measures for most sources. However,
for the sole attachment operation in shoe manufacturing, substitution of
hot melt adhesives accomplishes the control of emissions at a lower cost
than the add-on technologies.
C-6
-------
C.2.7. Chemical Manufacturing Processes
Table C-7 indicates the costs for controlling hydrocarbon emissions
from selected chemical manufacturing processes. For all but one process,
two sets of costs and control measures are given for each level of control
efficiency. The difference depends on whether or not the heat from the
control device is recovered and utilized. If heat is recovered, equipment
has to be added to utilize the heat for process requirements or to
generate steam for process requirements. For this option to be feasible,
the higher capital costs have to be offset by the savings for the heat or
steam recovery. This means that the plant has to have uses for the
recovered heat or steam either for in-plant uses or for other plants
nearby. Thus, the special circumstances of individual plants dictate
the viability of heat or steam recovery. Existing plants may not have
the flexibility necessary to incorporate this option. As a result,
some plants may have to employ incineration without any heat or steam
recovery, making the annualized costs of control higher than would other-
wise be the case.
C.3. Mobile Source Control Measures
While the Federal Motor Vehicle Control Program is intended to
reduce tailpipe hydrocarbon emissions from new vehicles, transportation
control measures are designed to reduce emissions from in-use vehicles.
Such measures can be divided into classes of measures that reduce in-use
automobile emission rates (emissions per mile) and classes of measures
that reduce vehicle miles traveled (VMT). The former class includes
essentially inspection/maintenance (I/M) as the only reasonably available
C-7
-------
measure, while the latter class includes transit improvements, carpooling,
and restrictions on the use of automobiles. In most transportation con-
trol plans, emphasis has been placed on I/M with VMT reduction measures
used where in-use controls are not sufficient.
C.3.1. Inspection/Maintenance
Motor vehicle inspection and maintenance (I/M) is a program of
periodic inspection of vehicles to determine the levels of emissions from
the vehicles. Those vehicles found to emit excessive amounts of pollutants
are failed, and must then be repaired and reinspected.
Reasonable estimates can be made of the costs and cost effectiveness
of an I/M program since data can be developed for each serviced vehicle
on inspection costs, maintenance costs, fuel savings, and emission reduc-
tion. The cost of building an inspection station varies depending on the
size of the station, the cost of land,and the type of test run. For a
two-lane centralized public facility utilizing the idle test, the
capital cost of the station ranges from $117,000 to $237,000. For a
loaded test, where the vehicle is run under simulated driving conditions
on a dynomometer, the capital cost is estimated to be $140,000 to $260,000
(ref. 13, p. III-C-8).
The inspection costs can vary depending upon who operates the pro-
gram, the complexity of required equipment,and whether safety testing is
included in the inspection. Costs can be as little as $0.30 per vehicle
if an idle test is combined with an established inspection program. On the
other extreme, one state which has built a loaded test system from scratch
requires an inspection fee of $5.00 to cover costs.
C-8
-------
The maintenance costs per serviced vehicle depend on the percentage
of vehicles which fail the test. A lower failure rate means the
maintenance cost per serviced vehicle will be higher since only the
worst emitters with more severe problems will be rejected. Naturally,
the cost of servicing these vehicles is higher than the cost of servicing
vehicles with less severe problems. For a loaded test, the maintenance
cost per serviced vehicle is estimated to range from $36 for a 10 percent
failure rate to $26 for a 50 percent failure rate. Of course, more
vehicles will be affected by a high failure rate, making the total cost
of maintenance for all vehicles higher (ref. 13, p. III-C-10).
By serving as a check to insure that vehicles are maintained pro-
perly throughout their lifetime, an I/M program can result in significant
fuel savings by motor vehicles. Once again, the dollar savings per
serviced vehicle depend on the failure rate, with the savings varying
inversely with the failure rate. Estimated annual fuel savings per
serviced vehicle range from $21 at a 50 percent failure rate to $49 at
a 10 percent failure rate (ref. 13, p. III-C-11).
At failure rates of 30 percent or less, maintenance costs are
offset by fuel savings. Hence, the average out-of-pocket costs of an
I/M program will be limited to the inspection fee, which will be about
$5 per vehicle (ref. 14, p. 21). Based on this, result and estimates of
current I/M effectiveness in reducing hydrocarbon emissions, the cost
effectiveness of I/M is estimated to be $340 per ton of hydrocarbons
removed (ref. 14, p. 22). As the emissions of new cars are reduced
through statutory requirements, though, the cost per ton will increase,
possibly to about $550 per ton by 1985.
C-9
-------
C.3.2. WtT Reduction Measures
Most VMT reduction measures are interrelated and require a coordinated
program in order to be an effective means for reducing hydrocarbon emis-
sions. The maximum emissions reductions from transportations measures
will result from coordinated measures designed to discourage low occupancy
auto use and to encourage transit and carpool.use. The following dis-
cussion outlines each group of measures and indicates the range of costs
associated with some of the measures.
C.3.2.1. Ridesharing
Carpools are an effective means of reducing commuter VMT. With
four members in a carpool, VMT can be reduced 75 percent over the VMT if
each rider .drove separately. .
The riders can also experience sizeable savings in travel costs.
For a four-person carpool with a 20 mile round-trip, one report estimates
that each rider will experience annual savings of $475. This savings
increases to $925 for a 40 mile round trip (ref. 15, p. 81).
The primary costs associated with carpooling are promotional costs.
However, an areawide computerized carpool matching service can be operated
for about $4 per participant (ref. 15, p. 80).
C.3.2.2 Preferential Treatment of High-Occupancy Vechiles
Dedicating lanes on freeways and city streets for the exclusive use
of buses and carpools during peak travel periods permits these vehicles
to bypass congested sections of roadways. This increases the attractive-
ness of high-occupancy travel modes since the passenger's travel time is
substantially reduced.
C-10
-------
Techniques used to give preferential treatment to these vehicles
vary from exclusive lanes on freeways to curb bus lanes on city streets
that require only remarking the lines. Costs vary widely depending
on the complexity of the improvements. For example, an exclusive busway
in California cost $4.9 million per mile to construct, while curb lanes
on city streets can cost $3,000 per mile to make the necessary changes
(ref. 15, pp. 31 and 38).
C.3.2.3. Transit Service Improvements
A number of aspects of transit operations can be improved to
enhance the level of service. These include transit marketing, security
measures, transit shelters, transit terminals, and transit fare policies
and fare collection techniques (ref. 15, p. 107). Since these measures
are solely dependent on local conditions, no attempt is made to assess
costs.
C.3.2.4. Parking Management
EPA considers parking restrictions, when coupled with transit and
carpool incentives, to be an effective and necessary means for standards
attainment and maintenance (ref. 16, p. 15). Parking management policies
relating to (1) the location of parking, (2) the amount of on- and off-
street space allocated to parking, (3) the parking charges applied to
the allocated space, and (4) the length of time parking is permitted
all can have a dramatic effect on traffic flow and VMT.
One mesns to discourage parking is through the use of taxes or
surcharges to increase parking cost. Studies have predicted that an
increase in daily parking cost up to one dollar can result in a reduction
C-ll
-------
i;o'f-'-three'*::t6;-Jl 5; perb¥nt:'::ih; V^f •Yrv'ttib' central 'business'" "di'strict whi 1 e
increasing 'transit1 us'e1 ff'^f: 1:5,'p;:l'54).
iA'nfrtheV :paHfhgsma;nage^fient rtreasure" is 'the* use of 'park-arid-ride lots.
'CbupTing1 fr'i'hg'e-or corridor1 parking facilities with' express transit
servic'e to; activity ceh'ters'can c'oritHbute sighificantly to the success
'of parking policies designed to reduce1the number of"CBD-directed automo-
biles. The estimated cost of a surface level,1 "self-park 'fringe parking
lot ranges from $0.50 to $2.00 per^vehic-ie:-per'day' (re'f". 15, p. 70).
-------
Table i. COSTS OFCONTROL MEASURES FOR OIL AND GAS PRODUCTION. REFINING, AND STORAGE
o
i
_j
CO
Somce/Anr-ctc'l Operation
1. Oil and gas production
2. Petroleum Refineries
Miscellaneous Sources
-Process drain and
wastewater separators
-Vacuum producing systems
-Process unit turnaround
3. Storage of rrude oil
and gasoline
4. Natural gas and natural
gasoline processing plants
r.iciiity
SI,-P
62,000 bhl/day
200,000 bbl/day
100,000 bbl/day
250,000 bbl.
Control
Efficiency
"•,
NA2
91
90
90
99
97
97
91
Capitol
C'ist
_j$ooo;
NA
W
NA
NA
NA
NA
9.8
NA
Annual i?Pd
Cost
(Win)
NA
NA
NA
NA
NA
NA
C
2.06
NA
Cost.
Effrrlivm
IS/ Ion]
Min3
Mm3
(64)4
(73)
X * ** /
(?8)5
Min3
200
Mm3
]
^GS Control
Mctisiirp
Detection and [
maintenance
- Detection and
maintenance
- Mechanical Seals
- Flare Header System
- Floating roof covers
- Floating roof covers
• - Detection and
maintenance
- Incineration of
non-condensables
- Combust non-con-
densable vapors
- Secondary seal on
floating roof tank
- Covers for oil -water
separators
- Mechanical seals
- Detection and main-
tenance
Rpf pri»nr o
- "ELsmiEc
1, P. 30
1, p. 17-18
1, p. 20
1, P. 22«*
1, p. 25
1. p. 2
1, p. 27
include credit for product recovery, parentheses indicate net savings.
o
N.A. = not available.
3Exact costs for these control measures are not known. However, it is believed that costs are minimal since- much of the control equipment is
already in operation in many plants or fields. In addition, it is believed that control costs will be offset by savings from product recovery.
These savings pertain only to floating roof covers.
iSThls estimate does not include cost of a condensate receiver for a surface condenser or the cost of covering the barometric hot well
6Em ssion reduction and hence product recovery ancLcost-effectiveness of secondary seals over and above primary seals will vary with the wind
velocity and the true vapor pressure of the stored product.
-------
Table 2. COSTS OF CONTROL MEASURES FOR SELECTED GASOLINE HANDLING AND DISTRIBUTION OPERATIONS
Source/ Affected Operation
1. Marine Terminals
2. Gasoline Bulk Terminals
- Top Splash Fill
- Top Submerqed
Fill and Bottom
Fill
3. Gasoline Bulk Plants
- Top Splash Fill
- Top-submerged
Fill and Bottom
Fill
4. Service Stations
- Underground tanks
(Stage I)
- Underground tanks
and vehicle refuel-
ing (Stages I and
ID
5. Gasoline tank trucks
Facility
Size
107 bbl/day
250.000 gpd
500,000 qpd
250,000 gpa
500,000 gpd
5,000 gpd
20,000 gpd
5,000 gpd
20,000 gpd
3 tanks
20,000 gpn
3 pumps
60,000 gpm
9 pumps
120,000 gpm
12 pumps
Control
Efficiency
%
95
94
94
87
87
58
93
58
93
93
93
93
90
90
90
99
Capital
Cost
(SOOO)
4000
195
231
176
264
0.3
24 3
0 6
57.0
24 8
57.0
0.6
4 4
8 8
10 9
N.A5
Annual ized
Co-tl
($0(0)
925''
(8 .)
(388)
19.9
22.0
(C.7)
I 8
(5.9)
i: 3
'..8
1 . 3
(0.2)
).7
1.5
1.3
N.A.
Cost .
Effectiveness
($/ton)
4000
(24)
(55)
146
i
80
(izo)
350
(120)
200
350
200
(110)
375
260
120
Mm.1'
Control
Measure
Refrigeration/
absorption
Refrigeration
Refrigeration
Refri(ierat1(in
Refrigeration
Conversion to top-
submevged fill
Vapor Balan:e
Conversion to top-
subme'-ged fill
i
Vapor balance
Vapor balance
Vapor balance
Vapor balance
Vapor
Collection
System "* 4
Reference
2. p. 142 '.
3. p. 4-4
.
4. p. 4-3
•
5
5
1. p. 15
'Costs include credit for product recovery. Parentheses Indicate net savings.
2Based on 15 year life and 10X Interest for capital recovery factor.
Three technologies (no latch/no flow balance, hybrid aspirator assist, and vacuum issist) appear capable of achieving Greater than 90
percent control While the control capabilities of the three systems are not significantly different under most conditions, costs have
been indicated for the hybrid system since these costs range between the insts for the other two systems.
Underground tanks controlled with vapor balance system
5N.A. - not available
cost of control will be limited to the cost of installing and maintnininq effr:Hve seals, connections, and pressm p-vari un valves.
This cost is considered minimal when compared lo the value of the product recovered
-------
Table 4. COSTS OP CONTKOL MEASURES TOR SELECTED GRAPHIC ARTS PROCESSES
O
Control Capital Annual ized Cost
Facility Efficiency Cost Cost' effectiveness Control
<-~, .-™/.-f fr.ri™] jjppration Sizp % ($000) CSOOO) ($/ton) Measure Reference
1 '-Jebt> Offset Printing
2 Fle*oora!)hic Printing
1 Rotogravure printing
1 Mebb Letterpress
printing
5000 scfm
20,000 srfm
5000 scfm
20,000 scfm
bOOO scfm
20,000 scfm
5000 scfm
20,000 scfm
95
90
95
90
95
95
95
90
95
90
9S
90
95
90
120 0
170 0
160.0
340 0
120.0
160.0
120.0
170 0
160 0
340 0
120 0
170 0
160.0
340 0
70 0Z
35. 03
35. 0Z
110.03
70. 02
35.03
70 O2
35 O3
35. 02
110 O3
70 O2
35. 03
35 0?
110 O3
Costs include credit for product or hent recovery. Parentheses indicate net savings.
210
240
110
100
210
110
210
240
no
100
210
240
110
100
Direct Flame incinerator
Carbon adsorption
Direct flame incinerator
Carbon adsorption
Direct flame incinerator
Direct flame incinerator
D.F. incinerator
Carbon adsorption
D.F. incinerator
Carbon adsorption
D F. incinerator
Carbon adsorption
D.F. incinerator
Carbon adsorption
1. P. 45
1, p. 52
1, P. 43
1, p. 48
Includes credit for primary heat recovery.
Includes solvent recovery credited as fuel value of solvent.
-------
Table 5. COSTS OF CONTROL MEflSURCS FOR SELECTED SOURCES OF EVAPORATIVE HYDROCARBON EMISSIONS
Control Capital Annual 1zed Cost
Facility Efficiency Cost Cost Effectiveness Control
Source/ Affected Operation Size Z ($000) ($000) ($/ton) Measure Reference
1. Organic solvent metal
cleaning operations
- Cold Cleaners
Low volatility solvent
High volatility solvent
- Open top vapor degreasers
- Conveyorized degreasers
Monorail
Cross-Rod
2. Dry cleaning Operations
- Petroleum plants
•Neighborhood cleaners
dryer, still, misc.
Filter Muck
TOTAL
•Industrial plant
' Dryer. Still, Misc.
Filler Muck
TOTAL
- Perchloroethylene
plants
'Coin-op facil ity
•Commercial plant
•Industrial plant
3. Cutback asphalt paving
Typical
Typical
152
302
30
45
97
50
SO
80"
90
SO
90
66
55
51
99
25
65
0.3
6.5
16.0
8.5
7.b
16. (I
5.0
21.8
71 ?.
_L.?_
76.4
7.J
Z.'J
7.5
N.A.
O.J.
(26)
(800)
84
36
(3735)
(650)
3.5
LA
4.5
4.4
i-LU
1.3
1.8
0.1
(9.4)
N.A.
20
(245)
(365)
25
1
(260)
(no)
735
TOQ-
730
55
(127)
15
5450
55
(345)
Min.
drainage facility
Drainage facility with
mechanically assisted
cover.
Manual cover
Refrigerated chiller
Enclosed design
Refrigerated chiller
Kefrigeratod chiller
Carbon adsorption
Centrifugal separator
Carbon adsorption
Centrifugal separator
Carbon adsorption
Carbon adsorption
Carbon adsorption
"Substitution of emul-
sified asphalt
7, p. 4-7
7, p. 4-1,1*
,
7, p. 4-17
8, p. 48
8, p. 4-10
8, p. 4-18
1, p. 32
1 . . .
2The price difference between the two types of liquified asphalt concrete 1s Insignificant.
-------
Table G. COSTS Of CONTROL' MCA'WCS FOR RUIJRER PRODUCT MANUFACTURE
LriiiUul Cd|illiil Annuitized ' Co'.t
Facility Efficiency Cost Cost' > Lffectlveness Control
Souree/Affrctcd Operation Slz<> X ($000) (JOOO) ($/
9d
K-
3!.
4'.
2900
7!
.221 3
2901 4
JU.
6!
290(
2900
22C
2900
220
2900
220
2900
22)
235
275
Floating covers
Adsorption
Thermal Incineration
Adsorption
Incineration'
Adsorption
Thermal Incineration
Thermal Incineration
Incineration
Adsorption
Thermal Incineration
Absorption: condenser
and scrubber
Thermal Incineration
Thermal Incineration
Thermal Incineration
Adsorption
Thermal Incineration
Adsorption
Thermal Incineration
Adsorption
Thermal Incineration
Hot melt adhesives
Hot melt adhesives
Hot melt adhesives
9. p. 4.1.3
• 9. p. 4.4.3
9. p. 4.6.4
9, p. 4.14.3
9. D. 4.13.3
(IdblL- 1very. Parentheses Indicate net savings.
'includes credit for primary heat rocovery
I
CO
-------
Table 7. COSTS OF CONTROL MEASURES FOR SELECTED CHEMICAL MANUFACTURING PROCESSES
o
VD
Facility
Source/Affected Operation Size
1.
2.
3.
4.
Ethylene Dichlorlde 350 x 103
(oxychlorination) TRY
Acrylomtrlle 100 x 103
TPY
Ethylene Oxide
- Air oxidation plant 100 x 103
TPY
- Oxygen oxidation plant 100 x 103
TPY
Formaldehyde
- Silver catalyst 50 x 103
TPY
- Mixed oxide catalyst7 50 x 103
TPY
'Con^ol
Efficiency
X
97
97
99
99
85
85
85
80
80
93
93
Capital
Cost
($000)
2050
1500
1515
700
485
15
30
88
92
135
10S
Annual i zed
Cost
($000)
Z073
10754
1903
4704
385
(")J
244
(44)6
224
1053
150"
Cost
Effectiveness
($/ton)
20
115
20
45
20
(5)
15
(275)
140
155
220
Control
Measure
Thermal incinerator,
waste boiler And
caustic scrubbing
Thermal incinerator
and caustic scrubbing
Thermal incinerator
ard waste heat boiler
Thermal Incinerator
Catalytic incineration
Steam generator
Thermal incineration
Boiler house vent gas
burner
Thermal incinerator
Thermal Incinerator
Thermal Incinerator
Reference
11,
pp.
11,
pp.
H,
pp.
11,
P-
11,
P-
Vol. 3
EO-39-4
Vol. 2,
AN-32
<*.
Vol. 6,
EO-31-3
Vol. 4,
FS-33
Vol. 5,
FM-Z5
Costs include credit for product or heat recovery. Parentheses indicate n-'t savings.
ZCosts are updated to 1976 values from reference 11.
Includes credit for steam or heat recovery.
Does not include credit for steam or heat.
Includes heating value credit.
Includes credit for heat recovery as process utilizes waste gas as fuel supplement.
Does not include recycling.
-------
REFERENCES FOR APPENDIX C
1. "Air Pollution Control Technology Applicable to 26 Sources of Volatile
Organic Compounds", Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency, May 27, 1977.
2. Radian Corporation, "Background Information on Hydrocarbon Emissions
from Marine Terminal Operations", Vol. I, EPA-450/3-76-038,
November, 1976.
3. "Control of Hydrocarbons from Tank Truck Gasoline Loading Terminals",
Office of Air Quality Planning and Standards, U.S. Environmental
Protection Agency, May, 1977, Draft.
4. "Control of Volatile Organic Emissions from Gasoline Bulk Plants",
guideline document in preparation by Office of Air Quality Planning
and Standards, U.S. Environmental Protection Agency.
5. Based OP internal E"A spBlys^s o^ rnmmontc cyhmitterj in
.^November 1, 1976,- Federal Register notice concerning the proposal
of Stage II vapor recovery regulations.
6. "Control of Volatile Organic Emissions from Existing Stationary Source^--
Volume II: Surface Coating of Cans, Coils, Paper, Fabrics,
Automobiles, and Ligh-Duty Trucks", Office of Air Quality Planning
and Standards, U.S. Environmental Protection Agency, EPA-450/2-77-008,
May, 1977.
7. "Control of Volatile Organic Emissions from Organic Solvent Metal
Cleaning Operations", Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency, April, 1977, Draft.
8. "Control of Volatile Organic Emissions from Dry Cleaning Operations",
Office of Air Quality Planning and Standards, U.S. Environmental
Proection Agency, April, 1977, Draft.
9. Monsanto Research Corporation, "Identification and Control of Hydrocarbon
Emissions from Rubber Processing Operations", EPA Contract No.
68-02-1411, Task 17, March, 1977.
10. "Control of Organic Solvent Emissions from Adhesive Usage During Shoe
Manufacture", guideline document in preparation by Office of
Air Quality Planning and Standards, U.S. Environmental Protection
Agency,
11. Air Products and Chemicals, Inc., "Engineering anrt Cost Study of Air
Pollution Control for the Petrochemical Industry", Volumes 2-6,
EPA-450/3-73-006.
C-20
-------
12. "Control of Volatile Organic Emissions from Existing Stationary
Sources —Volume I: Control Methods for Surface Coating Operations",
Office of Air Quality Planning and Standards, U.S. Environmental
Protection Agency, EPA-450/2-76-028, November, 1976.
13. "Control Strategy Preparation Manual for Photochemical Oxidant",
Office of Air Quality Planning and Standards, U.S. Environmental
Protection Agency, Guidelines Series No. 1.2-047, January, 1977.
14. Walsh, M.P., "The Need for and Benefits of Inspection and Maintenance
of In-Use Motor Vehicles", Mobile Source Enforcement Division,
U.S. Environmental Protection Agency, November 9, 1976.
15. Interplan Corp., "Transportation System Management: State of the
Art", report for Urban Mass Transportation Administration, U.S.
Department of Transportation, February, 1977.
16. "Policies for the Inclusion of Carbon Monoxide and Oxidant Controls
in State Implementation Plans", Office of Transportation and
1975.
U.S ETlVirCryO*"*"3! Py">^or-+-'?nri Anon/Ms
' ' " ..... " ..... •-•••--—
C-21
------- |