United States Office of Air duality EPA-450 /5-85-006
Environmental Frotjct,on planning and Standards Julv1335
Agency Research Triangle Park NC 277' 1
Air
©EPA Cost and Economic
Assessment of
Alternative Nationa
Ambient Air Quality
Standards for
Carbon Monoxide
(Revised)
-------
EPA-450/5-85-006
Cost and Economic Assessment of
Alternative National Ambient Air Quality
Standards for Carbon Monoxide
(Revised)
.ire.-: ., - ->&
I;. .04
U.S. ENVIRONMENTAL PROTECTION AGENCY
Strategies and Air Standards Division
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
July 1985
-------
ACKNOWLEDGEMENTS AND DISCLAIMER
Early versions of this report were prepared by Energy and Environmental
Analysis, Inc. The firm's staff who were responsible for the report were
Susan E. Schechter and Bruce Henning. EPA project managers were Allen Basala,
Economic Analysis Branch, and Thomas McCurdy, Ambient Standards Branch.
That version was rewritten in part by EPA in 1982 based on new analyses
made under Mr. McCurdy's direction. The current report is the result of
extensive revisions by Mr. McCurdy and GCA/Technology Division in 1985.
The work was performed under EPA Contract No. 68-02-3804. Key GCA staff
members were Mark G. Smith, Mark W. Deese, and Nancy K. Browne. Questions
regarding this report should be addressed to Tom McCurdy at MD-12, U.S. EPA,
Research Triangle Park, N.C., 27711; (919) 541-5655 (FTS 629-5655).
This report has been reviewed by the Office of Air Quality Planning
and Standards, U.S. EPA, and approved for publication. Mention of trade
names or commercial products is not intended to constitute endorsement
or recommendation for use.
ii
-------
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 carbon monoxide
standard 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 various
Executive Orders and OMB Circular A-107, EPA has prepared an assessment of the
potential cost and economic impacts associated with efforts to attain
alternative levels of the standard. This document presents the results of
this assessment.
The assessment also provides the cost information needed for a complete
benefits/cost analysis, as required by Executive Order 12291. This report
does not meet the requirements of that Executive Order, nor is it intended
to. A Regulatory Impact Analysis will be published under separate cover that
meets the 12291 requirements.
The purpose of the analysis contained herein is to estimate the relative
ranges of national control costs for alternative levels of the carbon monoxide
standard. In addition, in order to compare the relative implications of
alternative standards, the estimated number of urbanized areas which might be
expected to attain the alternative standards given various assumptions is also
indicated. Because of many uncertainties involved in projecting carbon
monoxide emission levels and air quality and in determining effective control
strategies throughout the nation, results of the analysis should be viewed
only as being an indication of relative differences in attainment status and
control costs for alternative standard levels. The analysis cannot be used to
determine how many or which specific urbanized areas will attain a given
iii
-------
carbon monoxide standard through particular control strategies. Rather,
attainment status and control requirements for attainment will have to be
determined by the States for each geographical area, considering the unique
conditions inherent to that area. Likewise, this analysis cannot estimate
with certainty the national costs of control strategies needed for all areas
of the country to attain alternative standards. Actual control costs will be
extremely variable, and this analysis is only useful in presenting relative
cost implications for alternative levels of the standard.
iv
-------
TABLE OF CONTENTS
Page
1. SCOPE AND PURPOSE 1-1
1.1 Regulatory Context 1-1
1.2 Study Objectives 1-3
1.2.1 Cost Estimates 1-3
1.2.2 Economic Impacts 1-4
1.2.3 Sensitivity Analysis 1-4
1.3 Organization of the Report 1-5
2. STUDY METHODOLOGY AND TOTAL COST 2-1
2.1 Control Programs 2-1
2.1.1 Federal Motor Vehicle Control Program 2-1
2.1.2 In-Use Mobile Source Control 2-3
2.1.3 Stationary Source Control 2-4
2.2 Total Cost 2-4
3. MOBILE SOURCE COST 3-1
3.1 Cost Estimation Process 3-1
3.1.1. Baseline Emissions 3-1
3.1.2 Air Quality Design Values 3-3
3.1.3 Assumptions Made in Calculating Needed
Reductions in Mobile Source Emissions 3-6
3.1.4 Calculation Procedure 3-6
3.2 Control Strategies: Cost Development and Emission
Reduction 3-8
3.2.1 Inspection/Maintenance Program 3-9
3.2.2 Transportation Control Measures 3-13
3.2.3 Total Cost of Mobile Source Control 3-14
v
-------
Tables of Contents (Continued)
Page
3.3 Cost of Attaining Current and Alternative Standards 3-15
3.3.1 Assumptions. 3-15
3.3.2 Federal Motor Vehicle Control Program 3-17
3.3.3 Mobile Source Cost of Attaining Current and
Alternative NAAQS Level 3-18
3.4 Sensitivity Analyses 3-18
3.4.1 Case #2: 1984 Base Year 3-32
3.4.2 Case #3: 1988 Analysis Year 3-32
3.4.3 Case #4: 1 ppm CO Background 3-33
3.4.4 Case #5: 40 Percent Maximum I/M Stringency 3-33
3.4.5 Case #6 and 7: 50 and 100 Percent I/M
Effectiveness 3-34
3.5 Incremental Costs of CO Control 3-34
3.5.1 Incremental Costs of FMVCP. 3-35
3.5.2 Incremental Costs of I/M and TCM 3-35
4. STATIONARY POINT SOURCE CONTROL 4-1
4.1 Stationary Point Source Costs 4-1
4.1.1 Attainment of a 9 ppm Eight-Hour Standard 4-4
4.1.2 Attainment of Alternative Eight-Hour Standards 4-4
5. MOBILE SOURCE ECONOMIC ANALYSIS 5-1
5. 1 Introduction 5-1
5.2 Analytical Procedure and Assumptions 5-2
5.2.1 Federal Motor Vehicle Control Program 5-2
5.2.2 Inspection/Maintenance Program 5-3
5.2.3 Transportation Control Measures 5-7
5.3 Economic Impacts 5-8
5.3.1 Federal Motor Vehicle Control Program 5-8
5.3.2 Inspection/Maintenance Program 5-10
5.3.3 Transportation Control Measures 5-15
5.4 Conclusion 5-21
VI
-------
Table of Contents (Continued)
Page
6. ECONOMIC IMPACT ON INDUSTRIAL POINT SOURCES 6-1
6.1 Economic Analysis Procedure 6-1
6.1.1. Data and Data Adjustments 6-1
6.1.2 Capital Availability Analysis 6-3
6.1.3 Annualized Cost Impact Analysis 6-4
6.2 Economic Impact to Industry 6-6
6.2.1 Determination of Industries with Significant Impact. 6-6
6.2.2 Profile of Industries Facing REgulatory Impacts 6-7
6.2.3 Capital Cost Impacts 6-9
6.2.4 Annualized Cost Impacts 6-21
6.3 Implication of CO NAAQS Control Costs for Small Business.... 6-29
APPENDIX A FEDERAL MOTOR VEHICLE EMISSION CONTROL PROGRAM (FMVCP).. A-l
A. 1. Light Duty Vehicles A-2
A. 1.1 Initial Cost of Emission Control Systems A-2
A.1.2 Maintenance Costs A-3
A.1.3 Use of Unleaded Fuel A-7
A. 1.4 Emission Control at Altitude A-7
A. 2 Light Duty Trucks A-10
A.2.1 Initial and Annualized Cost of Emission Systems A-10
A.2.2 Maintenance Costs A-10
A.2.3 Use of Unleaded Fuel A-10
A.2.4 Emission Control at Altitude A-14
A. 3 Heavy Duty Vehicles A-14
A. 3.1 Initial Cost of Emission Control Systems A-14
A. 3. 2 Maintenance A-16
A.3.3 Use of Unleaded Fuel A-16
A. 4 Motorcycles A-16
A. 5 Aircraft A-19
A. 6 Summary of FMVCP Costs A-19
vii
-------
Table of Contents (Continued)
Page
APPENDIX B STATIONARY SOURCES: METHODOLOGY AND ASSUMPTIONS B-l
B. 1 Industry Coverage B-l
B.I.I Selection Criteria B-l
B.I. 2 Air Quality Modeling of "Model Sources" B-l
B.I.3 NEDS Screen B-3
B.I.4 Source Category Control Cost Development B-7
B.2 Total Cost Methodology B-7
B.2.1 Evaluation of Ambient Impact B-7
B.2.2 Calculation of Required Emission Reductions B-9
B.2.3 Least-Cost Strategy Determination B-9
B.3 Benefits from Steam Credits B-10
B.4 CO Source Growth B-14
viii
-------
LIST OF TABLES
Number
TABLE 2-1. ANNUALIZED VARIABLE CONTROL COSTS FOR ALTERNATIVE
8-HOUR CO NAAQS ASSUMING FUEL SAVINGS CREDIT
TABLE 2-2. THE NUMBER OF SMSAs AND INDEPENDENT COUNTIES PROJECTED
TO EXCEED ALTERNATIV 8-HOUR CO NAAQS
TABLE 3-1. PERCENTAGE REDUCTION IN MOBILE SOURCE EMISSIONS FROM
I/M PROGRAMS FOR LDV'S
TABLE 3-2. 1987 MOBILE SOURCE VARIABLE CONTROL COSTS SENSITIVITY
ANALYSIS: DATA BASE ASSUMPTIONS
TABLE 3-3. 1990 MOBILE SOURCE VARIABLE CONTROL COSTS SENSITIVITY
ANALYSIS: DATA BASE ASSUMPTIONS
TABLE 3-4. 1995 MOBILE SOURCE VARIABLE CONTROL COSTS SENSITVIITY
ANALYSIS: DATA BASE ASSUMPTIONS
TABLE 3-5. RELATIVE (TO THE PRIMARY CASE) CHANGE IN MOBILE SOURCE
VARIABLE CONTROL COSTS FOR 1987: EMISSIONS DATA BASE
ASSUMPTIONS
TABLE 3-6. RELATIVE (TO THE PRIMARY CASE) CHANGE IN MOBILE SOURCE
VARIABLE CONTROL COSTS FOR 1990: DATA BASE
ADJUSTMENTS
TABLE 3-7. RELATIVE (TO THE PRIMARY CASE) CHANGE IN MOBILE SOURCE
VARIABLE CONTROL COSTS FOR 1995: EMISSIONS DATA BASE
ADJUSTMENTS
TABLE 3-8. 1987 MOBILE SOURCE VARIABLE CONTROL COSTS SENSITIVITY
ANALYSIS: I/M ASSUMPTIONS
TABLE 3-9. 1990 MOBILE SOURCE VARIABLE CONTROL COSTS SENSITIVITY
ANALYSIS: I/M ASSUMPTIONS
TABLE 3-10. 1995 MOBILE SOURCE VARIABLE CONTROL COSTS SENSITIVITY
ANALYSIS: I/M ASSUMPTIONS
TABLE 3-11. RELATIVE (TO THE PRIMARY CASE) CHANGE IN MOBILE SOURCE
VARIABLE CONTROL COSTS FOR 1987: I/M ASSUMPTIONS
TABLE 3-12. RELATIVE (TO THE PRIMARY CASE) CHANGE IN MOBILE SOURCE
VARIABLE CONTROL COSTS FOR 1990: I/M ASSUMPTIONS
2-5
2-7
3-10
3-19
3-20
3-21
3-22
3-23
3-24
3-25
3-26
3-27
3-28
3-29
ix
-------
Tables (continued)
Number Page
TABLE 3-13. RELATIVE (TO THE PRIMARY CASE) CHANGE IN MOBILE SOURCE
VARIABLE CONTROL COSTS FOR 1995: I/M ASSUMPTIONS 3-34
TABLE 4-1. CAPITAL COSTS FOR STATIONARY POINT SOURCES, BY INDUSTRY:
EIGHT-HOUR ALTERNATIVE CO STANDARDS 4-2
TABLE 4-2. ANNUALIZED COSTS FOR STATIONARY POINT SOURCES, BY
INDUSTRY: EIGHT-HOUR ALTERNATIVE CO STANDARDS 4-3
TABLE 4-3. TONS OF CO REDUCTION, BY INDUSTRY: EIGHT-HOUR
ALTERNATIVE CO STANDARDS 4-5
TABLE 5-1. AGE-FAILURE INDEX 5-5
TABLE 5-2. DISTRIBUTION OF MOBILE SOURCE COST FACTORS 5-5
TABLE 5-3. DISTRIBUTION OF FMVCP COSTS BY INCOME 5-9
TABLE 5-4. INITIAL I/M INVESTMENT COST 5-11
TABLE 5-5. DISTRIBUTION OF I/M INSPECTION COST FOR ALTERNATIVE
NAAQS 5-12
TABLE 5-6. DISTRIBUTION OF I/M REPAIR COST FOR ALTERANTIVES NAAQS. 5-14
TABLE 5-7. DISTRIBUTION OF FUEL SAVINGS FROM I/M FOR ALTERANTIVE
NAAQS 5-16
TABLE 5-8. DISTRIBUTION OF NET I/M REPAIR COSTS AND FUEL SAVINGS
FOR ALTERNATIVE NAAQS 5-17
TABLE 5-9. ANNUAL TCM GROSS COST (?) FOR ALTERNATIVE NAAQS 5-18
TABLE 5-10. DISTRIBUTION OF TCM FUEL SAVINGS FOR ALTERNATIVE
NAAQS 5-20
TABLE 6-1. MALE 1C ANHYDRIDE: FINANCIAL PROFILE 6-10
TABLE 6-2. CARBON BLACK: FINANCIAL PROFILE 6-11
TABLE 6-3. STEELMAKING; FINANCIAL PROFILE 6-12
TABLE 6-4. GRAY IRON FOUNDRIES: FINANCIAL PROFILE 6-13
x
-------
Tables (continued)
Number Page
TABLE 6-5. PRIMARY ALUMINUM: FINANCIAL PROFILE 6-14
TABLE 6-6. CAPITAL REQUIREMENTS FOR THE AVERAGE MALEIC
ANHYDRIDE FIRM 6-16
TABLE 6-7. CAPITAL REQUIREMENTS FOR THE AVERAGE IRON FIRM 6-18
TABLE 6-8. CAPITAL COST TO THE AVERAGE ALUMINUM FIRM 6-20
TABLE 6-9. CAPITAL COST TO THE AVERAGE CARBON BLACK FIRM 6-22
TABLE 6-10. ANNUALIZED COST FOR AN AVERAGE ALUMINUM FIRM 6-23
TABLE 6-11. ANNUALIZED COST FOR AN AVERAGE GRAY IRON FIRM 6-24
TABLE 6-12. AVERAGE AFTER-TAX RETURN FROM CONTROL OF CO FOR AN
AVERAGE MALEIC ANHYDRIDE FIRM 6-26
TABLE 6-13. AVERAGE ANNUAL RETURN PER FIRM IN THE STEELMAKING
INDUSTRY 6-27
TABLE 6-14. AVERAGE ANNUALIZED RETURN PER FIRM IN THE CARBON
BLACK INDUSTRY 6-29
TABLE 6-15. CONTROL COSTS AS A PERCENTAGE OF CAPITAL EXPENDITURES
AND VALUE ADDED BY SIZE OF ESTABLISHMENT FOR GRAY IRON. 6-32
TABLE A-l. ANNUALIZED CAPITAL COST INCREASE FOR LDV's DUE TO
FMVCP CO CONTROLS A-4
TABLE A-2. MAINTENANCE CHANGES OVER 100,000 MILES A-5
TABLE A-3. ANNUAL OPERATING AND MAINTENANCE SAVINGS FOR LDV's DUE
TO FMVCP CO CONTROLS A-6
TABLE A-4. COSTS OF UNLEADED FUEL USE DUE TO FMVCP FOR LOT A-8
TABLE A-5. ANNUALIZED CAPITAL COST INCREASE FOR LDV's DUE
TO FMVCP CO CONTROLS AT ALTITUDE A-9
TABLE A-6. ANNUALIZED CAPITAL COST INCREASE FOR LOT's DUE TO
FMVCP CO CONTROLS A-11
xi
-------
Tables (continued)
Number
TABLE A-7.
TABLE A-8.
TABLE A-9.
TABLE A-10.
TABLE A-11.
TABLE B-l.
TABLE B-2.
ANNUAL OPERATING AND MAINTENANCE SAVINGS FOR LDT's DUE
TO FMVCP CO CONTROLS
COSTS OF UNLEADED FUEL USE DUE TO FMVCP FOR LDT.
ANNUALIZED CAPITAL COST INCREASE FOR LDT's DUE TO
FMVCP CO CONTROLS AT HIGH ALTITUDE
COSTS OF UNLEADED FUEL USE DUE TO FMVCP FOR HDV.
TOTAL ANNUALIZED COST OF THE FMVCP FOR CO
POTENTIALLY SIGNIFICANT PROCESS SOURCES OF CO...
PTMAX MODELING RESULTS
A-12
A-13
A-15
A-17
A-19
B-2
B-4
xii
-------
LIST OF FIGURES
Number Page
2-1 Study Methodology for Cost of Attaining CO Standard 2-2
6-1 Annualized Cost Methodology 6-5
B-l Schematic of Incineration System With Primary Heat Recovery. B-12
B-2 Schematic of Incineration System With Primary and
Secondary Heat Recovery B-13
xiii
-------
1. SCOPE AND PURPOSE
The Clean Air Act, as amended in 1977, requires periodic review of ambient
air quality standards. This study, which reviews the National Ambient Air
Quality Standard (NAAQS) for carbon monoxide, analyzes 1) the hypothetical
relative impacts that alternative changes in the standard would have on
national costs of control and on attainment status for various areas of
the country, and 2) the potential socioeconomic impacts of these costs.
This section presents the study's background, scope, and organization.
1.1 REGULATORY CONTEXT
Executive Order 12291, "Federal Regulation," requires among other things that
all major rules be accompanied by a comprehensive analysis of cost and economic
impacts of the federal rule or regulation. This cost and economic assessment
is intended to meet the Executive Order requirement on the cost side. The
Order also requires that a benefits analysis be undertaken for major rules,
but CO NAAQS benefits are not discussed in this report. They are analyzed
in a formal Regulatory Impact Analysis (EPA 1985) sent to the President's
Office of Management and Budget, which includes exerpts from this report
as well.
It should be noted that the Clean Air Act specifically requires that a NAAQS
be established based on scientific criteria relating to the level that should
be attained to protect public health with an adequate margin of safety. The
Act precludes consideration of the cost or feasibility of achieving primary
standards in determining the level of ambient standards. This policy has
been affirmed by the U.S. District of Columbia Court of Appeals in Lead
Industries Association v. Environmental Protection Agency (647 F.2d 1130
[1980J; cert, denied 449 U.S. 1042 [1980]).
1-1
-------
At the present time there are two NAAQS's for carbon monoxide: an eight-hour
standard of 9 parts per million (ppm), and a one-hour standard of 35 ppm.
Both standards are not to be exceeded more than once each year. Only the
costs and impacts of the eight-hour standards are addressed in this study
because extensive examination of the one-hour issue indicates that, in all
cases, both the current and alternative one-hour standards impose less
stringent control requirements on both stationary and mobile sources than do
any of the eight-hour standards analyzed. Control techniques used to attain
the eight-hour standard are probably adequate to guarantee attainment of the
one-hour standard. Therefore, there would be no significant incremental cost
due to compliance with the one-hour standard.
The cost and economic analyses presented here examine the current eight-hour
standard of 9 ppm, one observed exceedance and three alternative standard
formulations.
The one observed exceedance form is a "deterministic" one in that it relates
to a specific value that is fixed by its place in a frequency distribution of
air quality data. In the case of one observed exceedance, it is the
second-highest value in the ranked order set of measured 8-hour averaged CO
values for a year. In this report, this value will be known as the 2nd-high
value.
The three alternative formulations all involve (1) a "statistical"
specification for the allowed violation rate dimension of a NAAQS and (2) a
daily maximum one-hour specification for the periodic aggregation dimension of
a NAAQS.* In summary form, the four alternative CO NAAQS investigated in this
analyses are:
1. 9ppm/2nd-high non-overlappng 8-hour average in a year. (Hereafter
cited as the 9/2nd-high standard.)
2. 9 ppm/1 expected exceedance daily maximum 8-hour average in a year.
(Hereafter cited as 9/1 standard.)
3. 12 ppm/1 expected exceedance daily maximum 8-hour average in a year.
(Hereafter cited as a 12/1 standard.)
1-2
-------
4. 15 ppm/1 expected exceedance daily maximum 8-hour average in a year.
(Hereafter cited as a 15/1 standard.)
1.2 STUDY OBJECTIVES
The primary objectives of this study are to:
Estimate the relative hypothetical costs of attaining alternative
National Ambient Air Quality Standards for carbon monoxide.
Study the economic impacts that the hypothetical have controls on
consumers and on industrial production.
Determine the major sensitivity of results to certain study variables.
Each of these objectives is discussed below.
1.2.1 Cost Estimates
The hypothetical cost estimates consist of the direct costs of control
measures which require carbon monoxide control, including the Federal Motor
Vehicle Control Program. In cases where control of more than one pollutant is
achieved by a particular control measure, costs are apportioned among the
various pollutants, as described in the Cost Methodology sections.
*There are at least six attributes, or dimensions, of a NAAQS (McCurdy,
1984). The dimensions, and the corresponding item for carbon monoxide, are:
Dimension
1. Concentration Level
2. Averaging Time
3. Periodic Aggregation
4. Frequency of Repeated Peaks
5. Allowed Violation Rate
6. Data Handling and Analysis
Conventions
Example
9 ppm
8 hours
Maximum 8-hour average in a day
(Not used)
One expected exceedance per year
Defined by 40 CFR and EPA Guidance
1-3
-------
The cost estimates reflect total annualized cost of attaining alternative CO
standards. This includes annualized capital charges on equipment as well as
annual operating and maintenance/repair costs. Also included in the
annualized cost estimates are potential savings due to decreased energy
requirements resulting from certain control strategies.
It should be noted that no consideration is given here to SIP development
costs, or to other costs incurred by States or local governments in
implementing a CO NAAQS.
1.2.2 Economic Impacts
The second primary objective of this study is to assess economic impacts
associated with the costs incurred by implementing the required set of control
strategies. These economic impacts are divided into three areas, depending on
where the direct effect is felt. The particular cost components differ among
the set of control techniques hypothetically needed, but include estimates of
costs imposed on industry and individuals.
Impacts on individual consumers arise primarily from mobile source control
costs and are examined in two ways: per capita impacts and macro-economic
impacts on the distribution of income. Industrial financial impacts result
from the costs of stationary source control and are examined both in terms of
(1) capital requirements imposed on industry and the ability of companies to
bear the capital expense, and (2) inflationary and competitive impacts of
annualized control costs of total production costs and product prices.
1.2.3 Sensitivity Analysis
The final objective of the study is to determine how sensitive the results are
to certain assumptions. Exogenous variables such as growth rates and control
effectiveness are examined using different assumptions to determine
sensitivity of the resulting costs. Certain of the sensitivity analysis
scenarios are deemed to be significant and therefore are subject to a complete
analysis.
1-4
-------
1.3 ORGANIZATION OF THE REPORT
Section 2 presents the study methodology. While it provides an overview of
the study, it does not present complete methodologies for each analytical
component; rather, it indicates what the components of the study are and shows
how the various analytical pieces interact to produce estimated costs of
attainment. Section 2 also presents the total annual costs of attaining the
current as well as alternative eight-hour standards. These costs summarize
all of the component costs calculated in later sections.
Section 3 presents the methodology for and results of the cost estimation for
mobile source control measures. Included is a discussion of the procedures
used to determine the amount of emissions reduction needed, the control
strategy selected to attain the standard, and the degree of reduction and
associated cost produced by the control measures. Estimates are made of the
number of standard metropolitan statistical areas (SMSAs) that will comply
with each standard using the control strategies under consideration. Results
for various sensitivity analysis scenarios also are presented. Appendix A
provides the detailed analysis of Federal Motor Vehicle Control Program costs
for CO.
Section 4 presents results of the stationary source cost analysis by
industry. The methodology and assumptions used to calculate these costs are
presented in Appendix B.
Section 5 contains an analysis of the economic impacts of mobile source
control requirements. While these impacts are not directly associated with a
NAAQS, they are presented in order to provide a synoptic picture of mobile
source impacts for environmental protection.
Section 6 analyzes stationary source economic impacts. The results indicate
those industries for which control costs may present potential problems.
Control costs for any stationary source category constitute less than 1% of
total CO control costs.
1-5
-------
REFERENCES FOR SECTION 1
1. McCurdys Thomas. March 1984. "Miscellaneous Analyses of Alternative 63
NAAQS Formulations and Their Impact on 63 Non-Attainment and Cost Analysis
Procedures." Ambient Standards Branch, U.S. EPA.
2. U.S. Environmental Protection Agency. 1985. Regulatory Impact Analysis
for Carbon Monoxide (Revised). Ambient Standards Branch, Office of Air
Quality Planning and Standards.
1-6
-------
2. STUDY METHODOLOGY AND TOTAL COST
This section presents the total annualized cost of attaining alternative
National Ambient Air Quality Standard (NAAQS) and various alternative
standards for carbon monoxide. Figure 2-1 depicts the general procedure
followed to determine the total cost of control. The program through which
attainment is achieved consists of three areas of emission control: the
Federal Motor Vehicle Control Program (FMVCP), in-use mobile source control,
and stationary source control.
2.1 CONTROL PROGRAMS
2.1.1. Federal Motor Vehicle Control Program
The Federal Motor Vehicle Control Program provides substantial emission
reductions across the country and their resulting costs comprise a basic
component of the program to attain the NAAQS for carbon monoxide. While FMVCP
control requirements are not directly related to the NAAQS program, they
provide a baseline emission reductions guideline upon which the current and
alternative standards can be based.
FMVCP includes a series of increasingly stringent tailpipe emission standards
with which new motor vehicles must comply. The program includes control of
light-duty vehicles (LDVs), light-duty trucks, heavy duty trucks, and
motorcycles. Emission standards established by the program become
increasingly stringent for later model years; thus, as more early model
vehicles are retired, total fleetwide emissions are reduced. All of the
analyses undertaken in this report assume that the current 3.4 grams per mile
FMVCP standard for LDVs will be in effect throughout the period of analysis.
2-1
-------
cn
<:
o
00
_JO
ill
U_
O
O
ae
o
2-2
-------
Since emission standards require different control technologies, their costs
are calculated separately for each model year. A complete discussion of the
methodology and results of the costs incurred due to FMVCP appears in
Appendix A.
2-1.2 In-Use Mobile Source Control
Control programs used to achieve additional reductions in mobile source
emissions necessary to attain NAAQS are aimed at reducing emissions from
in-use vehicles. For purposes of this analysis, such measures include
Inspection and Maintenance (I/M) programs for light-duty vehicles and
Transportation Control Measures (TCMs) which reduce emissions by reducing
travel and improving traffic flow. In designated nonattainment counties
(those needing emission reductions), the amount of reduction is determined by
current emissions, projected future emissions, and air quality levels that are
present. The estimated emission reductions are expanded to the Standard
Metropolitan Statistical Area (SMSA) level of analysis, which is the areal
basis used in this report.
While the actual cost of these measures will vary considerably among local
areas, this analysis uses average costs based on studies of programs already
implemented in local areas. The magnitude of the cost in individual SMSAs is
dependent upon the estimated vehicle population and the degree of emission
reduction required. The particulars of the cost algorithm are presented in
Section 3.
Because of the different nature of mobile source and stationary source
emission problems and the location of the existing monitoring network, it is
believed that recorded violations in nonattainment areas are a result of
mobile sources and localized area sources. As part of this study, an analysis
of the stationary source problem was conducted which indicates that stationary
source emissions had negligible effects on monitor readings in most counties.
Therefore, all of the needed reductions in each nonattainment county are
provided by control of mobile source emissions.
2-3
-------
2.1.3 Stationary Source Control
Stationary source control requirements were determined by examining the
maximum air quality impact from individual sources in isolation.
Source-specific data for emission and stack parameters were used to determine
the maximum air quality impact of the plant's emissions.
Model plants were used to estimate the unit cost of alternative control
techniques for each source category. The total cost of controlling point
source emissions was calculated by applying the average unit cost data to
actual levels of operation at each source. The costs were aggregated to
obtain a national total. A more complete discussion of the calculation
methodology is presented in Appendix B; the results are discussed in Section 4.
2.2 TOTAL COST
As discussed in-Section 1.1, while the NAAQS for CO is expressed as both a
one-hour and eight-hour standard, only the eight-hour standard alternatives
are examined here, since, in all cases analyzed, compliance with the
eight-hour standard assures compliance with the one-hour standard. Control
requirements estimated by this study for the current eight-hour standard of
9 ppm 2nd-high eight-hour average and three alternative NAAQS formulations
provide emission reductions necessary to meet almost all one-hour standards as
well. Total control expenditures for any of these eight-hour standards exceed
estimated expenditures for any feasible one-hour CO NAAQS.
Table 2-1 presents national variable costs of complying with alternative
NAAQS's, ignoring net FMVCP costs. Variable costs are those that vary with
the NAAQS being investigated. It should be noted that these costs do not
include estimation of any cost of administering the standard. All costs
represent net annualized costs incurred in the years indicated, measured in
1984 dollars. Any benefits arising from reduced fuel consumption for mobile
and stationary sources are included. The costs estimated for compliance with
each standard depend on assumptions made for each scenario. These assumptions
include emission source growth rate, I/M and TCM control program
effectiveness, and the initial air quality levels, as well as policy decisions
2-4
-------
TABLE 2-1. ANNUALIZED VARIABLE CONTROL COSTS FOR ALTERNATIVE
8-HOUR CO NAAQS ASSUMING FUEL SAVINGS CREDIT3
(in constant 1984 dollars x 106)
Alternative NAAQS(in ppm)
All Hours Daily Maximum
Item 9/2nd-High 9/1 ExEx 12/1 ExEx 15/1 ExEx
1987
I/Mb
TCMC
Stationary Sources^
Total Incremental Costs
232.7
75.0
..2.7.3
335.0
184.3
37.5
.27.3
249.1
51.1
6.0
30.5
87.6
34.5
-2.5
16.0
48.0
1990
I/M
TCM
Stationary Sources^
Total Incremental Costs
1995
I/M
TCM
Stationary Sources^
Total Incremental Costs
232.3
53.1
27.3
312.7
197.6
27.1
27.3
252.0
47.2
4.3
30.5
82.0
31.8
-1.8
16.0
46.0
Note: aExcludes Federal Motor Vehicle Control Program costs, which do not
vary by NAAQS analyzed. Annualized FMVCP costs are shown in
Appendix A, Table A-ll. The cost estimates are for the "primary
case".
"I/M = Inspection and Maintenance Programs.
CTCM = Transportation Control Measure.
°Annualized stationary source control costs are non-linearly related
to stringency of NAAQS because of fuel savings associated with CO
controls in certain industry types.
2-5
-------
regarding the level and form of the standard. A complete discussion of these
assumptions appears in Section 3.
Stationary source controls do not vary with the number of allowed expected
exceedances for any particular standard level. The modeling methodology used
to estimate stationary costs is not sensitive to the allowed exceedances
variable, so the number of sources not attaining a standard -- and hence their
control costs -- do not change with number of allowed exceedances. The
relatively small magnitude of stationary source controls vis-a-vis total costs
minimizes the impact that this methodological problem has on the cost
assessment results for most of the standards analyzed. This issue is
discussed in more detail in Section 4 and Appendix B.
The national cost estimates are only one part of the picture with respect to
impacts on the nation of alternative 8-hour CO NAAQS. Associated with each
alternative is an estimate of the number of SMSAs that hypothetically cannot
attain the CO NAAQS using reasonably available control measures (RACM). RACM
control techniques are those that are currently operational and proven to be
effective. For CO, these are inspection and maintenance programs and
transportation control measures for mobile sources; and boilers, flares,
incinerators, and operating changes for stationary sources. All of these
techniques are discussed in more depth below.
If an area cannot achieve an alternative NAAQS using RACM technologies, no
attempt is made in this analysis to "force" attainment by using more
stringent, but unproved control technology. Thus, if the hypothetical control
program devised for each SMSA needing control cannot achieve the standard,
that fact is simply noted. The number of SMSAs projected to exceed the
various CO NAAQS after implementing a RACM program is shown in Table 2-2. The
same set of assumptions that were used to develop the data in Table 2-1 is
used to develop Table 2-2.
2-6
-------
TABLE 2-2. THE NUMBER OF SMSAs AND INDEPENDENT COUNTIES PROJECTED TO
EXCEED ALTERNATIVE 8-HOUR CO NAAQS AFTER IMPLEMENTATION OF
REASONABLY AVAILABLE CONTROL MEASURES3
All Hours
Year 9/2nd-High
-1987 28
1990 10
1995 3
Alternative NAAQSCin ppm)
Daily Maximum
9/1 ExEx 12/1 ExEx
14 2
6 0
2 0
15/1 ExEx
0
0
0
aThese estimates are for the "primary case."
2-7
-------
3. MOBILE SOURCE COST
Mobile sources account for over 70 percent of the mass of carbon monoxide
emitted in the United States (EPA, 1984). Since the CO problem in most urban
areas is caused by mobile sources, the majority of CO control costs will be
borne by the mobile source sector.
This section provides information about 1) the process by which the cost
estimates for mobile source controls were made, including the method by which
needed reductions are calculated, 2) control strategies, and 3) costs of
control. In addition, the critical assumptions made are presented along with
their sensitivity analyses.
Complete details of the procedure for determining mobile control requirements
and resulting costs can be found in a report produced by Stanford Research
Institute, entitled "Methodologies to Conduct Regulatory Impact Analysis of
Ambient Air Quality Standard for Carbon Monoxide" (SRI, 1979). Most of the
inputs to this procedure have been updated to 1984, as noted in this Section.
3.1 COST ESTIMATION PROCESS
The process by which cost estimates were made includes determining baseline
emissions, air quality design values, and needed reductions.
3.1.1 Baseline Emissions
Baseline emissions data were developed to provide information on the magnitude
of CO emissions in each of 280 counties. These counties are considered either
to be in nonattainment under the current standard or to be potentially in
nonattainment under one or more of the alternative standards. The base year
3-1
-------
emission inventory comes from EPA's existing emission data base. Projections
of emissions for three critical future years are made taking into account
growth as well as the mandated tailpipe emission standards for motor vehicles.
Emissions projections are made for the years 1984, 1987, 1990 and 1995. The
baseline emissions data include contributions from both mobile and area
sources, with mobile source emissions as the primary contributors. The
current mobile source data are obtained from the EPA National Emission Data
System (NEDS) file, and were assumed to represent 1982 emissions; growth and
adjustment factors therefore were used to transform 1982 emissions to the 1984
baseline (Platte, 1985).
Projection of future mobile source emissions in 1984, 1987, 1990, and 1995
were calculated using MOBILE3, EPA's mobile source emission model, which
incorporates the latest emission factors (EPA, 1984b). The emission
projections take into account estimates of growth in vehicle miles traveled .
(VMT), as well as reductions in tailpipe emissions expected from the Federal
Motor Vehicle Control Program, as projected by MOBILES. The projections are
based on the currently mandated 3.4 grams per mile (gpm) light duty vehicle
standard. Since MOBILE3 provides no emission factors for California
technology vehicles, separate California mobile source emission projections
were made using data from the California Air Resources Board's equivalent of
MOBILE3, EMFAC6 (GARB, 1985).
The original CO cost model (SRI, 1979) was designed to correct NEDS CO
emissions from an assumed uniform 75°F temperature to each county's average
January temperature, using ratios of mobile emission factors for the
respective temperatures. Emission estimates in the current NEDS file are
based on county-specific annual average temperatures. As described later in
this Section, the current analysis has been performed with and without the
original temperature correction factors to assess the sensitivity of the
methodology to this variable.
3-2
-------
Area source emissions also contribute to the baseline emissions inventory.
Area sources include residential heating and some processes too small to be
individually identified, including woodstoves (to the extent they are included
in NEDS). Current area source emissions (for 1982) were also obtained from
the NEDS data file. Future area source emissions were calculated using
current emissions and an annualized growth rate that can be input to the
program.
There has been some concern recently that high CO emissions from woodstoves
would cause by themselves violations of the current CO standard in urban areas
with a high usage of woodstoves. To date, OAQPS has no data to confirm or not
confirm this statement. Even if woodstoves are a problem in some areas, they
would not contribute much to the monitors that are used in this analysis (see
the next Section). These monitors are traffic-oriented for the most part, and
are not located in residential areas that generally are the source of most
woodstove emissions. In addition, more than one-half of the areas
investigated have relatively warm winters or are major metropolitan areas,
neither of which is conducive to the use of woodstoves for primary wintertime
heating. Thus, an under-representation of woodstove emissions from the NEDS
inventory would not have much impact on results contained in this report.
Perhaps a few areas would need more controls if woodstove emissions were
included, but it is impossible to quantitatively address the issue because of
lack of data.
Total baseline emissions are equal to the sum of mobile source and area source
emissions. This calculation is performed for all counties in the analysis.
3.1.2 Air Quality Design Values
The design value for an area represents the estimated ambient CO concentration
from which emission reductions are calculated in the strategy planning
process. Design values which are used in this analysis were obtained from a
review of 1981-1983 ambient air quality data in EPA's SAROAD data base, which
comes from an extensive monitoring network. The data base represents the only
3-3
-------
readily available source of consistent data for all of the areas
investigated. Sufficient data are contained in the data base to estimate the
approximate values based on alternative forms of the standard.
The design values included in this analysis are approximate and suitable only
for analytical purposes in this assessment. In SIP revisions submitted to
EPA, States calculate the actual design values used for attainment
determinations and for planning purposes. The values are calculated for the
appropriate form of the standard based on guidance provided by EPA (see CFR
Part 51.)
The current CO standards specify that the eight-hour average of 9 ppm must not
be exceeded more than once per year. In addition to assessing alternative
standard levels in this analysis, alternative procedures for calculating
exceedances of the standard also are considered. These procedures affect the
form of the standard. Not only does the form influence determination of the
number of exceedances of the standard, but it also impacts the calculation of
an area's design values.
In its current form, the standard is based on the second-highest monitored
value during a year in an area. However, this deterministic (observed
once-per-year) approach has limitations in that it does not account for the
stochastic nature of maximum CO concentrations. To maintain such a standard
year after year necessitates a zero probability that the second-high value
will ever exceed the standard. On a practical basis, permitting only a single
absolute exceedance in a year means that there is some finite possibility of
occasionally having two or more exceedances in a particular year.
To remedy this conflict and to adjust for the effect of missing data, EPA is
considering defining the standard on a statistical basis whereby the expected
number of exceedances per calendar year is determined. Statistical forms of
the standard vary depending on whether all possible values or just daily
maximum values are used, and how running averages are handled for the
eight-hour standard.
3-4
-------
For purposes of the analyses as summarized in this document, two
interpretations of the statistical standard are used. For the one-hour
standard, the hourly interpretation bases the design value on the ambient
hourly concentration that on average will be exceeded once per year in each
area. The daily interpretation, on the other hand, bases the standard on the
number of days with maximum hourly CO averages above the level of the
standard. This means that a day with two or more hourly values over the
standard level counts as one exceedance of the standard level rather than two
or more.
Statistical forms of the eight-hour standard follow the same basic approach,
but the interpretation is complicated by running averages, as discussed by EPA
in "Guidelines for the Interpretation of Air Quality Data with Respect to the
National Ambient Air Quality Standards" (EPA, 1977). The current CO standard
is chosen so that the second exceedance will not come from an eight-hour
period which contained at least one hour in common with the first exceedance.
In calculating design values, the daily interpretation uses overlapping
eight-hour averages in computing the expected number of exceedances. For each
day, the highest of the 24- possible eight-hour averages is the daily maximum
eight-hour average. With this method, the possibility arises that two daily
exceedances could have common values. The other statistical approach (the
running eight-hour average interpretation) uses all possible eight-hour
averages for the year so that more than one exceedance per day could be
counted. This is more stringent than the current form of the standard in that
exceedances could be overlapping.
In the analysis of the three years of data from 1981 to 1983, design values
based on the eight-hour statistical form of the standard are expected to fall
between the third and fourth highest maximum value (whether hourly or daily).
In selecting design values, the fourth highest value over the three year
period was used. If only two years of data were available, the third highest
value was chosen.
3-5
-------
While design values are available for all major urban areas which comprise 60
to 70 percent of the nation's population, many non-urbanized areas of the
country do not have valid air quality monitoring data from which to judge
whether the area has a CO problem. For these areas, emission densities in the
county were obtained in order to gain an indication of the potential problem
in the county. These values then were compared to equivalent emission
densities which serve as compatible surrogates for the various standard
levels. A basic air pollution model was used to calculate equivalent emission
densities that would lead to a concentration equal to the levels of the
standard under a certain set of conservative conditions (SRI, 1979). Thus,
for those counties not having design value data, emission densities are used
as design value surrogates.
3.1.3 Assumptions Made in Calculating Needed Reductions in Mobile Source
Emissions
One of the most controversial areas of air pollution analysis is the
relationship between emissions and monitored air quality values. Changes in
relative location of emission sources and air quality monitors can radically
alter air quality readings. Therefore, all emissions within a county do not
contribute equally to air quality readings.
The amount of emission reduction required in each county to meet each of the
standards under consideration is calculated through the use of a modified
rollback procedure (SRI, 1979) which assumes a linear relationship between
emissions and air quality. In other words, a 10 percent improvement in the
air quality design value is calculated to require a 10 percent reduction in
emissions.
3.1.4 Calculation. Procedure
The percent rollback was calculated for each county with a design value in the
following manner:
Percent rollback = 100 x (Design value), -(Standard)
Design value - Background
3-6
-------
This provides a percentage indicating the amount of improvement in air quality
necessary to meet the standard. For counties where design values do not
exist, the rollback is calculated by the following equation based on emission
density data:
2
Percent Rollback = 100 x (Total emission density, tons/mi /yr)-(Standard Density)
Total emission density
where the standard density is an estimate, calculated by EPA, of emissions per
square mile that result in air quality readings equal to the standards.
Both of these calculation procedures provide an estimate of the percent
improvement in air quality required to meet the standard. Once this is
accomplished, the allowable emissions are calculated for each county under
each standard for each analysis year. Allowable emissions are defined as the
maximum level of effective emissions within a county that produces air quality
readings that comply with a particular standard, as calculated by the rollback
procedure:
Allowable emissions = (1 ' % rollback) x (Current effective emissions)
1007.
The concept of effective emissions is used to reflect the relative impact of
various source categories. Effective emissions are obtained by considering
the current or future baseline emissions and weighting the components by
factors chosen to estimate the relative contributions of different sources to
air quality readings. Based on results of modeling studies, the assumption is
made that only a portion of the total area source emissions contributes to an
area's maximum CO values, which comes from sites that reflect more localized
sources. For purposes of this study, the effective emissions which are
assumed to impact monitored values are equal to 100 percent of mobile source
emissions and 20 percent of area source emissions (SRI, 1979). These
contributions may be varied in the algorithms used. See Section 3.4.
3-7
-------
Since point source emissions do not affect air quality readings and no area
source control measures are considered, any reduction required to meet the
allowable emissions levels must be supplied through control of mobile source
emissions. Thus, the needed mobile source reduction equation is:
Needed mobile source reduction = Total effective emissions in any year) -
(Allowable effective emissions)
Needed reductions were calculated for 1987, 1990, and 1995, for the current
and relaxed standards, and for 1990 and 1995 for more stringent alternative
standards.
Point source emissions (i.e., those from major industrial processes) are not
considered to contribute to monitored CO values. The stationary sources
analysis (presented in Section 4 and Appendix B) indicates that point source
emissions have negligible effects on monitored readings in most counties.
Since it is the design value and not emissions which determines whether or not
a county requires additional control, no counties are exempted from the need
for additional control by the assumption of the reduced effect of area source
emissions. These assumptions cause mobile source emissions to be weighted
more heavily in attainment strategies, to simulate observed contributions. A
further discussion of this is presented as part of the discussion on emissions
growth sensitivity in Section 3.4.
3.2 CONTROL STRATEGIES; COST DEVELOPMENT AND EMISSION REDUCTION
Mobile source control strategies considered in this study are divided into
two types: Inspection and Maintenance (I/M), and Transportation Control
Measures (TCM). TCM programs include both area-wide measures, which are
designed to reduce total travel in an area, and localized measures, which
respond to "hot spot" problems by improving traffic flow.
While the cost and effectiveness of both I/M and TCM programs vary depending
upon the characteristics of the individual areas and programs, this study uses
3-8
-------
average cost and reduction estimates for the programs to calculate total
cost. These average costs are based on studies of programs that have been
implemented in selected local areas across the country.
3.2.1 Inspection/Maintenance Program
3.2.1.1 Description of_Programs Used
Fleet-wide emission reductions achieved by I/M programs will be dependent on a
number of factors, one of which is the percentage of vehicles that fail the
test and thus have to undergo repairs. Other influencing factors include
ambient temperature, location, year of implementation, and extent of mechanic
training achieved by an I/M program.
Table 3-1 presents example estimated emissions reductions for 50°F for I/M
programs using 20 and 30 percent failure rates for pre-1981 vehicles, with 3.0
and 1.2 percent CO cutpoints for 1981 and later vehicles. The table also
indicates that the amount of reduction obtained is dependent upon the number
of years the program has been in operation, while the year in which the
program starts is dependent upon a status schedule generated by EPA (EPA,
1985). Presently, many areas in the United States are required to establish
I/M programs for the control of CO/HC emissions in order to meet current
ambient air quality standards for CO and/or ozone. It is assumed that any
area with this designation has initiated its I/M program in 1984. Programs
have been operating in some areas for a number of years. Assuming 1984
commencement for all existing and required programs will result in some
underestimation of I/M effectiveness for programs in place earlier, especially
for the 1987 analysis year. Since only six programs began before 1982, the
overall effect of this assumption will be small. Since MOBILE3 does not
provide I/M program effectiveness for California, these factors were estimated
by adjusting MOBILE2 California effectiveness factors for changes in non-I/M
emission factors between MOBILE2 and MOBILE3.
3-9
-------
i-H
1
CO
w
-C
<
s
TON IN
/M PROGRAMS FOR LDVs3
H i i
U
=> s
c o
w oe
OS EL,
H tO
O 2
< O
H HI
2 to
w to
O HI
OS S
U W
ft.
U
U
OS
3
o
CO
b4
_1
H- 1
M
O
in
ON
ON
i 1
1^
cfl
01
5n 0
4-> ON
CU i 1
00
V.
to
H
f""*»
00
ON
i l
cu
CO
4-1 OS
c
cu cu
u t-
M 3
0) i I
ft. 'H
<0
fn
(-
to
01
>
c
o
H
4J
CO
-l
41
H
C
M
C
O
H
4-1
CO
U
o
.
1 1
m I-H r» fo ON m
00 CN r^ fH vO O
CM CO CN CO CM CO
O O 00 CN CO CO
f- i l -d" ON CO r-»
CN CO CM CM CM CM
r~~
oo oo oo
ON ON ON
I 1 I 1 I 1
cu
o
3
4J
H
4-1
I 1
<
?
0
hJ
00 i i ON in «* CM I-H r~-
0)
H
.c
u
CO
CO
c
o
H
4J
u
3
T3
CU
OS
1
CO
inspected
1-1
0)
,0
o
4-1
o
classes are not require
CU
ii
0
I-l
jr
u
>
i-i
CU
.c
4J
O
co
cu
i-H
u
H
"cu
>
>-.
4J
3
a
4J
X
60
H
i 1
1-
o
<4H
;rature .
^v
a
cu
4J
cu
60
CO
^
CU
CO
4-1
I-*
CU
s:
c
CU
^
X
CO
fa
CU
01
i-l
oc
0
"O
0
in
ca
o
4J
>^
i-H
a
c.
CO
Cfl
0)
to
CO
4J
C
CU
u
t.
o
c.
01
CO
4)
x:
H
3-10
-------
Some counties are identified as needing emission reduction for attainment of
one of the more stringent alternative CO standards, without currently needing
an I/M program. If such a county is located in a State which has a current or
planned I/M program, it is assumed that the county will implement I/M in
1986. In other States, a county needing reductions is assumed to begin the
program in 1987. This is to account for the lead time needed to obtain
legislative authority at the State level.
In urban areas, auto emissions affect the air quality of the entire area, not
just the county in which the specific problem occurs. Emission reductions can
be guaranteed only if every county in the urban area implements an I/M program
similar to that required in the problem county. It therefore is assumed that
Bureau of Census Standard Metropolitan Statistical Areas (SMSAs) which require
control in at least one of their counties implement an I/M program at the
tightest stringency level required, and then implement the same program in all
of their other counties as well.
3.2.1.2 I/M Costs
The annual expenditure required for an I/M program is divided into inspection
cost, repair cost, and fuel savings (a negative cost). Total costs for each
area implementing I/M are calculated by estimating the vehicle population in
each analysis year and multiplying the number of cars affected by the unit
costs.
The inspection cost, levied on each car, is designed to include costs of
operating and maintaining the program and an annual capital charge to pay for
the inspection facility. While these costs can vary from $2.50 to $10.00
depending on the particular program (EPA, 1985), the average cost is assumed
to be $7.00 for each car inspected. Cars which fail the inspection will be
required to undergo repairs in order to correct the problem. Preliminary data
provided by EPA's Office of Mobile Sources (OMS) indicates that the average
cost of repairs is a function of (1) the proportion of pre-1981 to post-1981
cars in the fleet inspected, (2) stringency level, and (3) year of analysis.
The average repair cost data takes the form of a matrix (Lorang, 1985):
3-11
-------
Stringency
.207. 307. 40%
1987 $ 20.88 $ 21.67 $ 22.28
1990 17.58 18.79 19.79
1995 14.82 16.60 18.20
There is uncertainty regarding the true repair cost estimate and per-car costs
certainly will vary around these mean estimates.
Improved operation of repaired vehicles is expected to result in some small
fuel savings for post-1981 cars. The fuel savings vary by (1) the proportion
of post-1981 cars in the fleet inspected, (2) I/M stringency level, and
(3) year of analysis. The average fuel savings estimate takes the form of a
matrix (Lorang, 1985):
Stringency
20% 30% 40%
1987 0.4% 0.4% 0.5%
1990 0.6 0.7 0.8
1995 0.9 1.1 1.2
Total fuel savings are calculated based on assumed values of 550, 490, and
420 gallons per year per car for 1987, 1990 and 1995, respectively, and
$1.20 per gallon of gasoline (EEA, 1984; EIA, 1984).
Unit costs presented here do not include allowances for the cost of peoples'
time to have their cars inspected and repaired. While people admittedly will
suffer some inconvenience from I/M programs, the costs of such losses in time
are not large since the actual inspection lasts only a few minutes. In one
study of an operational I/M program specifically concerned with developing
detailed cost information on all aspects of the program, the average value of
time spent waiting for the inspection and needed repairs, was $1.68 per car
inspected (Palmini and Rossi, 1980). Since two pollutants are involved in
this I/M program, CO and HC, the time costs attributable to CO would be
$0.84 per car.
3-12
-------
3-2.1.3 Allocation of I/M Costs for CO Control
Costs mentioned in the previous Section represent the total cost of
implementing I/M in any county. Counties specified by EPA as implementing I/M
by 1985 were assumed to do so because of the need for both CO and HC emission
control. For these programs, total I/M costs should not be attributed totally
to CO control, rather they should be equally divided between the
two pollutants. Therefore, annual I/M costs are divided by two for all
counties instituting I/M by 1985, and for all counties currently exceeding the
ozone NAAQS by 0.03 ppm (i.e., an ozone design value _>0.15 ppm). For other
areas which are identified in this analysis as needing I/M for CO control, all
of the costs are attributed to CO.
3.2.2 Transportation Control Measures
3.2.2.1 Program Description
Transportation Control Measures reduce emissions by reducing total vehicle
miles traveled (VMT), by reducing average daily travel (ADT) for specific
roadways, or by improving traffic flow. Control measures include transit
system improvements, ridesharing programs, road user charges and congestion
pricing, traffic signal system improvement, and exclusive bus or car pool
lanes. These methods are categorized as either local or area TCM measures.
Local TCM include signal timing, computerized control of street flow, freeway
surveillance and control, and truck restrictions on certain streets. These
measures are intended to improve traffic flow and alleviate CO "hot spot"
problems. Areawide TCM, which include ridesharing, express buses, local bus
improvements, and work rescheduling are intended to reduce total emissions in
an area by reducing VMT and the number of trips.
TCM options are used where the needed reduction is 5 percent or less,
(assuming no other county in the urbanized area requires I/M), and in addition
to I/M in counties where I/M is not sufficient to provide the projected needed
reductions in mobile source emissions.
3-13
-------
3.2.2.2 TCM Cost
Emission reductions and costs per unit of reduction of TCM controls can vary
significantly depending on local conditions. However, EPA has undertaken an
assessment of TCM in order to determine generalized emission reductions and
costs for use in this analysis (SRI, 1979). As a result, an estimate of up to
a 3 percent reduction in mobile source emissions is assumed to be obtained
through local TCM. Although the measures will be somewhat different depending
on the locality, the cost of such measures is estimated at $230/ton of CO
reduction (SRI, 1979).
Area-wide TCM may reduce emissions by an additional 2 percent. The marginal
cost of these reductions, however, is not constant. The first one percent
reduction in emissions from area TCM is estimated to cost $550/ton of CO
reduced, with the last one percent costing $12,600/ton reduced (SRI, 1979).
The last one percent includes such costly measures as transit improvements.
A byproduct of reducing VMT is a reduction in fuel consumption. By reducing
total travel, a factor of 1088 gallons/ton of CO reduced is used to estimate
these fuel savings (SRI, 1979). This value then is subtracted from the annual
cost to estimate the net annual cost of the program.
As in the case of I/M, only a portion of these costs are attributable to CO
control in many areas. Since some TCM also can be used to reduce HC,
emissions, reduce traffic congestion and travel time, and save energy, this
analysis allocates a maximum of 50 percent of the cost and savings to CO
control in areas with both a CO and ozone problem. For CO-only non-attainment
areas, 100% of the costs and savings attributable to TCM is allocated to CO..
3.2.3 TptaJLCost. of. Mobile Source Control
The total annual net cost of mobile source control is the sum of three CO-only
components for each area: 1) the cost of FMVCP, 2) the I/M cost, and 3) the
TCM cost. These are discussed in detail in subsequent sections.
3-14
-------
3.3 COST OF ATTAINING CURRENT AND ALTERNATIVE 8-HOUR CO NAAQS
3.3.1 Assumptions
Total nationwide CO NAAQS costs reflect variable expenditures required to
attain an alternative NAAQS for CO plus the costs of complementary mobile
source emission programs (e.g., FMVCP) which do not change under alternative
NAAQS1s. The variable expenditures are associated with 3 major control
programs:
Inspection and Maintenance (I/M)
Transportation Control Measures (TCM)
Stationary Source Emission Controls
The stationary source control estimates are derived via a methodology that is
independent of mobile and area sources impacts. The only factors that
influence stationary source control sources are the source-specific emission
rates and modeling methodology. (See Appendix B.) Because stationary source
control costs are relatively small, sensitivity analyses were not undertaken
on the various inputs that went into the predictive equations. Thus, we have
only a "best estimate" for stationary source control costs.
A number of sensitivity analyses were undertaken on mobile source oriented
control costs inputs, however. These costs include implementating I/M and TCM
programs and FMVCP controls. FMVCP controls do not vary by NAAQS level, so
they are the same for all alternative ambient standards investigated here. A
FMVCP of 3.4 grams per mile instituted for 1981 model year cars and beyond is
assumed. Variable costs associated with alternative NAAQS are those related
to I/M and TCM. We focus on these two programs in this Section.
There are a number of variables that affect total I/M and TCM costs in the
country. They are listed below, along with the primary, or usual, assumption
that the variables take. The usual assumptions are considered to be
conditions associated with reasonably available control measures, or RACM.
These variables are:
3-15
-------
NAAQS Standard, definition; Four alternatives were investigated, as
noted in Section 1.1.
Growth in emissions; Mobile source emissions grow approximately at a
one percent per annum compound rate. The growth rate used depends
upon the time interval; for the 1982-1987 time period the growth rate
is 1.2% per year, for 1987-1990 it is 1.1% per year, and for
1990-1995 it is 1.05% per year. Area source emissions grow at a rate
of 2.5% per year.
Source interaction; Stationary source emissions do not affect the
analysis; only 20% of area source emissions affect the analysis.
CO background concentration; 0 ppm. A one ppm background level is
also considered.
I/M stringency (or re4ectipn rate); 20% of all LDVS, or 30% if
needed. A 40% stringency level is considered to be an alternative.
Once I/M goes into an area, it stays in for the entire period of
analysis.
I/M effectiveness, in cold-weather area: I/M effectiveness is assumed
to vary by temperature. QMS research indicates that I/M
effectiveness can be approximated by a straight-line relationship
between the following end points: (1) 50% effective at 20°F, and
(2) 100% effective at 70°F (Armstrong, 1985). The temperature of
interest is the mean daily average temperature for the three month
period of November-January.
TCM effectiveness: TCM can reduce the critical county mobile source
emissions by a maximum of 5%. (See Section 3.2.2.2.) Once a TCM
program is implemented, it remains in during the period of analysis.
Per unit cost figures:
I/M inspection fee: $ 7.00 per LDV
I/M repair charge: cost per repaired LDV is in 3.2.1.2.
I/M capital charge: $8.00 per LDV
price of gasoline: $ 1.20 per gallon
In the following subsections, the RACM costs of attaining or coming as close
as possible to attainingthe various CO NAAQS are presented and discussed.
In addition, the impact on cost of changing one or several of the above
assumptions will be examined and compared to the primary case as a test of
sensitivity of the cost analysis to analytical assumptions.
3-16
-------
3.3.2 Federal Motor Vehicle Control Program (FMVCP)
The FMVCP controls tailpipe emissions of CO, hydrocarbons, and nitrogen oxides
from cars, trucks, and motorcycles. Appendix A presents an analysis of the
program's cost for these vehicle categories as well as a discussion of the
rationale for allocating certain FMVCP costs to CO. Table A-ll summarizes
total 1987, 1990, and 1995 FMVCP costs attributable to CO for a 3.4 gpm
FMVCP. In general, the costs were divided equally among the pollutants
controlled by each phase and/or component of the program (i.e., one, two, or
all three).
Costs for light-duty vehicles assume that 1981 and later model year cars meet
the 3.4 grams/mile emission standard as mandated by the Clean Air Act. While
there are several alternative technologies which might be used to achieve the
standand, most auto manufacturers plan to use a three-way catalytic converter
in combination with an oxidation catalyst, the latter primarily for CO
control. The costs presented in this analysis reflect the allocation of
estimated costs for such a system among CO, HC, and NOX.
In addition to the hardware costs, FMVCP annualized costs also include
estimates of the costs of savings resulting from changes in operation and
maintenance, changes in fuel economy, and fuel price differentials. The
underlying rationale for these estimates is presented in Appendix A.
As mentioned, FMVCP costs are incurred regardless of the level or form of the
standard. Thus, they remain a constant cost in all scenarios analyzed.
Because FMVCP costs are a constant, they will not be presented in the
sensitivity analysis output tables.
It should be mentioned that FMVCP costs would vary greatly if a waiver of the
3.4 gpm CO standard would be granted. (At one time, a 7.0 gpm FMVCP starting
with 1981 model year cars, and remaining in effect for two years, was
discussed.) Such a waiver is not seriously being contemplated, however, so
its possibility was ignored in this analysis.
3-17
-------
3.3.3 Annualized Mobile Source Cost of Attaining Current and Alternative
NAAQS Levels
Annualized mobile source costs of attaining the various CO standards analyzed
vary widely depending upon the assumptions made. The "primary case"
assumptions, result in an estimated national net control cost for the current
9 ppm 2nd-high standard of about $343 million in 1987, $332 million in 1990,
and $313 million in 1995. However, not all areas can attain this standard by
the years analyzed. The number of residual non-attaining areas is estimated
to be 28 in 1987, 10 in 1990, and 2 in 1995 for the primary case.
Both the estimated annualized national control costs and residual
non-attainment areas decrease for the alternative NAAQS investigated. This is
shown in Tables 3-2 to 3-4 for the primary case and other scenarios.
3.4 SENSITIVITY ANALYSES
Several different sets of alternative analytical assumptions were formulated
in order to determine how sensitive the cost and residual non-attainment
estimates are to the assumptions made. These sets are known as Cases #2
through #7. (Case #1 is the primary case.) The cases are aggregated into two
major case categories: data base assumptions, and I/M assumptions. The
differences in each case are described below, and results are shown in
Tables 3-2 through 3-13.
As mentioned previously, the assumptions used make a large difference in the
cost and residual non-attainment estimates. For instance, annualized 1987
control cost estimates for the current 9 ppm 2nd-high standard vary from a low
of $227.5 million to a high of $335.5 million. This range of $108 million is
almost 35% of the primary case estimate of $307.7 million annualized costs.
The estimated number of residual non-attainment areas in 1987 varies from 14
(for Case #4) to 45 (for Case #5). This is a very large difference. The
primary case estimate is 28 residual non-attaining areas in 1987 for the
9/2nd-high standard.
3-18
-------
CN
t
f")
w
-T
ffl
<;
H
Y ANALYS
H
tH
r- 1
H
CO
J2
w
CO
CO
H
CO
o
CJ
^
O
06
H
Z
O
^
.
^J
a
^
OS
<^
>
iO
2
0
;H
&4
Si
S
CO
CO
^
w
CO
«
s
^
Q
xc""1
o
rH
X
CO
t-l
CO
rH
rH
O
o-
00
ON
r t
C
CO
4>J
CO
c
o
o
X"N
g
0.
a
c
H
s-^
CO
0*
Z
0)
^
f-4
4J
CO
C
01
4-1
< .
CO
T3
u
CO
"O
C
CO
CO
g
3
g
H
^
CO
2
>,
^
CO
Q
CO
I*
3
O
32
rH
<
w
X
W
rH
-*x^.
in
rH
O
CN O
CO
X
W
X
W
rH
**^
CN
rH
i-H
r-« CM
LO
X
w
X
w
1 t
ON
.C
oo
rH
55
1
a
c
CN
ON
00
.
rH <)
CN rH
CN
[^
.
r- 00
O CN
fl
0£
O
CO
a
c
i
oo
ON
41
CO
CO
01
CO
(0
CJ
o
SI-
CO
^sD vD
m CN
CN
sf O
C*1
00 O
00 rH
CN
co o
CN
vO CN
Sf
st <-<
CN
0)
CO
CO
O
^,
^
CO
e
H
r.
r
o
ro
CO
4-1
CO
O
CJ
rH
O
^1
^J
c
o
0
00
CN
co
CO
01
(j
<
C
H
CO
4-1
4J
<
C
o
z
rH
CO
3
T3
H
W3
CJ
B£
}^
CO
0)
cy
ca
co
K
^f
00
ON
fl
CO
1 1
ca
o
CJ
rH
O
IM
4J
C
o
CJ
lA
sf
co
CO
01
r4
<
C
rH
CO
4-1
U
<
1
C
0
rH
CO
3
a
H
CO
01
CO
01
CO
I-l
CO
^
1 1
CO
c
CO
00
ON
r~-
CN
CN
ca
4J
CO
o
CJ
rH
O
^
4-1
C
o
CJ
rH
CN
CO
CO
OJ
Id
<
c
H
CO
4-J
4-1
<
1
C
0
"Z.
rH
CO
3
a
rH
CO
0)
OS
round
oo
CJ
CO
«
O
g
CL
a.
o
ii
CO
4-1
CO
o
CJ
rH
O
I-l
4-J
C
o
CJ
-Attain. Areas 31
c
o
rH
CO
3
-o
H
co
01
3-19
-------
CO
I
CO
TY ANALYS
CO
z
CO
CO CO
6- Z
CO O
O r-l
a
a.
CO
> I- T3
CO
C co
O CU
CJ OS
oo
CO O
CM O
CO
CN
VO
H U
II
CU r-4 CO
CO O 3
CO I- T3
PC 4-1
ON O
CO rH
CM
CO
CO
CU
CO
U 4J
CQ 4-1
at <
>- CO I
U C
CO CO O
H O Z
co O
CM
CO
00
ON
CO
i-l r-l CO
CO O 3
C r< -U
< iJ -i-l
C CO
OO O
-------
st
I
<*>
PQ
CO
a
<
z
<
>-
H
h-1
>
CO
CO
CO
H co
co z
o o
o s:
06 3
H CO
Z CO
O <
K CO
5 SQ
ii <;
(X H
> Q
W
U
o
CO
w
w
o
CO
t-l
CO
o
a
st
00
c
CO
ca
O
a
/ N
E
a.
a
c
H
s^
CO
a-
z
CU
^
H
4-t
CQ
c
>-i
CU
4-1
P-^
CO
o
}-l
CO
a
c
CO
4J
CO
E
E
H
X
>,
H
to
O
CO
l-l
3
O
HM
1 1
i4
X
W
X
U
I 1
^^.
LO
i I
X
w
X
w
1 1
**^,
CN
I 1
X
w
X
u
1 1
*^s^
ON
f.
60
H
"I
1
o
c
CN
0 -.
O O CM O
ro *3~
m co
«
i I O fOO
u-i i--
f*^ r 1
st CN CO \O
CN >£>
CN CN
st m
*
m co r-- o
00 CO rH
CSI C")
CU
ca
co
U
o
CU
ca
to
co
CN
Oi
vo
st en
CN
CN
vO
CN O
st
CO
CN
as
CN
CU
ca
CO
U
^
^4
CO
E
i-t
CO
4J
CO
o
CJ
1 1
o
t^
4J
c
O
CJ
CO
CU
l-l
C
H
CO
4-1
4J
1
C
o
2J
^H
CO
3
T3
H
co
CU
Cfi
^1
CO
CU
^-i
CU
ca
CO
PQ
st
00
CO
4-1
CO
o
CJ
1 1
o
^
4J
c
o
U
ca
to
l~
c
l-l
to
4-1
4J
1
C
o
z
1 1
CO
^
o
H
ca
cu
(X
CO
cu
ca
H
ca
^,
i-i
to
C
00
00
CO
4-1
CO
o
o
1 1
o
i-l
4-1
C
o
CJ
-Attain. Areas
c
o
52
i i
CO
3
o
i-l
CO
CU
round
60
^
U
CTJ
CQ
0
E
a.
a.
co
4J
CO
o
o
1 1
o
^_!
4-1
c
0
CO
co
0)
c
H
CO
4-1
4J
1
C
o
z
1 1
CO
3
*o
H
co
cu
3-21
-------
OS
<
M
O O
0) M
H
w a.
« co
O 00
s: <
z w
03
J
<
O
w r-.
to oo
< (^
C6
O
i-i CO
OS H
CL. CO
O
W 0
H J
O
O OS
H H
N' 12
o
>
tI
H
OS
^-\
E
o.
ex
c
H
N '
CO
o-
z
01
J>
i-l
4-1
to
c
01
<
to
"O
^4
to
o
G
tO
4-1
CO
£
3
£
H
X
to
£
H
to
Q
to
3
Q
"T^
1 1
^
X
w
X
w
i-H
~*^
in
r 1
O
CM O
CO
X
W
X
M
r 1
^^
CN
i 1
i i
r- CM
LO
X
w
X
W
r I
oT
.c
00
^1
i
*o
c
CM
ON
00
r t <&
CM r-4
CN
r-»
r- oo
O CM
ri
to
tfl
0
o
z
01
to
to
00 CM
r-.
ITl CM
CN
m
t
CM I
I
CO
o o
i-H
I
CM CM
00
O
CO
to
u
>l
CQ
E
H
U
0
X
CO
o
o
r- 4
O
4J
C
o
o
to
m
01
c
to
4-1
4-1
1
C
o
z
1 1
to
3
u
H
co
01
OS
to
01
o>
CO
to
03
^f
oo
ON
Change in Costs
4-1
C
01
0
'H
4"
to
tO
0)
CO
3
o
H
to
01
G
i-l
Ol
00
c
CO
f.
o
to
01
CO
H
CO
* 1
to
C
00
oo
ON
to
4-1
cn
o
o
c
1-1
01
00
G
to
o
4-1
c
0)
o
}H
01
a-
CO
to
Ol
i i
CO
3
u
H
CO
0)
c
01
00
G
j~;
Q^
Background
0
E
P.
a
Change in Costs
4-1
C
Ol
o
t-l
0>
cu
in Residual Areas
01
60
C
to
_r"
u
3-22
-------
w
5
a
oa
<;
M
OS
>
w co
CJ Z
OS O
O H
co (X
S
W 3
J CO
I-. CO
PQ <
0
2; a
CO
z <
1-1
a <
0 H
Z <
< O
G
o
W O\
CO f-l
^ \
E
a
a
c
H
^
to
>
H
CO
O
CO
^
2
Q
^
i I
i 1
**
~-V4
X
W
1 1
^^
to
1 1
^
*
O O
trj
X
X
a
,_4
CN
i 1
00
en o
X
W
X
W
r-l
-^
C7\
i i
^\ \o
CN
CN
60
i-l
|^
1
T3
c
.
ON
r*"* ^2
Cw '"^
m
U as
O
>-l fc
OS
< CO
S H
hi CO
OS O
(X O
a p-J
o
H OS
H
o z
H o
a
OS
r-.
rH O
CN
CN O
CN
CN
CN O
iI
I
CN O
CN
CU
CO
CO
CJ
01
co
CO
O
O
00 O
t-H
I
CN
0)
co
CO
O
^
(-
CO
E
l-i
2-
0
X
CO
4J
CO
O
U
^H
o
4-1
C
0
o
-Attain. Areas
c
o
z
,1
CO
3
"O
H
co
0)
OS
IM
CO
cu
>H
01
CO
CO
PQ
^
00
cy\
r t
CO
CO
o
o
H
01
60
c
to
r~.
C_^
4J
C
01
U
I-l
0)
CO
CO
CO
o
CO
01
OS
c
1-1
cu
60
C
tcj
r"
U
CO
CU
CO
H
CO
5^
p 1
CO
C
00
00
ON
1 1
CO
co
0
0
C
H
CU
00
c
CO
.C
o
4_l
c
01
o
^4
01
0,
sidual Areas
CU
OS
c
01
60
c
CO
o
round
60
*s
U
CO
«
O
CJ
E
a
o.
! <
CO
4J
CO
o
0
C
1-1
CU
60
c
to
4=
0
4-1
C
CU
u
I-l
0)
Cu
sidual Areas
cu
OS
c
1-1
0)
60
c
to
.c
0
3-23
-------
I
CO
eo
OS
W co
CJ Z
OS O
3 M
O H
CO OH
2
W 3
^ CO
HH CO
EC <
O
s u
co
m
K ON
O OS
O
OS ""
< CO
Z H
i-i CO
OS O
K
O
OS
H
Z
O
u
u
H-1
H
U
y^.
E
a
a.
c
H
>^
CO
CX
<5
Z
0)
^
1-1
4J
CO
C
i-l
0)
<
X
X
U! U
o
1- 1
CO *-»
u m
C --1
O CO
o o o o
CO ,
H X
CO U
Q X
W
i i
~*^
ON
si
CO 60
3 =:
O 1
3 "d
C
i-l CM
i 1
m to
-I O CM 0
m
CO
CO
o
z
01
co
CO
O
ca
C3
0)
CO I
±J pj
CO O
01 O Z
CO U
CO .-H
CJ ii CO
O 3
>-, 1- -O
|j 4-1 -H
co C CO
E O o>
H CJ OS
CM
CO
CO O
in o
r-l
I
CO
CN
I
00 O
I
CO
f^ O
rI
I
m
\D CN
J^
CO
O)
^H
0)
0}
CO
P3
sfr
oo
cc
u
o
u
C
H
01
60
C
f
CJ
^J
c
0>
u
CU
OH
Areas
CO
3
c
H
W
01
OS
c
H
01
60
c
CO
r*
O
(^
CO
V
^-1
CO
H
CO
>SJ
f^
co
C
<^
00
00
en
CO
o
u
C
r-t
01
60
C
CO
4=
CJ
u
C
01
u
o>
Areas
CO
3
o
H
CO
O)
OS
C
01
60
C
CO
r*
CJ
o
c
3
O
^
60
V
U
CO
»
O
o
E
a
o.
CO
o
O
C
ft
01
60
c
CO
^^
CJ
4J
B
01
u
^4
0)
Areas
CO
3
o
H
co
Ol
OS
C
H
01
60
c
CO
f.
3-24
-------
CO
r-l
CO
$
<
<
><
H
00
1
on
w
r 3
«
^
H
CO
Z
W
CO
H
CO
o
o
r4
o
EH
^
O
0
.
,_;
^
ex
<3^
>
CO
z
o
1 i
H
S
CO
CO
^
1-1
vO
o
1 1
X
CO
i-l
CO
rl
r 1
O
V
^
oo
ON
r t
C
CO
c
o
o
y^x
E
a
a
c
*^/
C/J
Q*
,
H
CO
O
CO
w
3
_O
r 1
<
X
X
W
14
^.
in
i i
O
.
CN O
on
X
w
X
w
1 1
^^
CN
1 1
ii
.
r~ CM
in
X '.
w
X
w
r4 i
^^.
ON
oo
r-l St
CN i 1
CN
60
H
£2
|
o
e
CN
ON
r^
r-- co
O CN
on
in CN
in o f» o
on CM
oo oo
.
OO r-l 1 1 CN
in m
si- in
.
ON on in m
O ' 1 ' ! ' '
CN CM
.
m o
on
O
st 0
st
f^
p** r*^
o
CN
O
i I CO
r*>- i (
CN
w
K
O
0)
CO
CO
r-
co
o
z
4)
CO
CO
u
0)
CO
CO
o
^Sj
^4
CO
£
H
^
a.
CO
44
CO
o
CJ
II
o
i-i
4-1
C
o
o
ttain. Areas
<£
i
c
o
z
1 1
CO
3
o
H
co
0)
ngency
H
r-l
4-1
CO
E
i
H
^
CO
*sr|
j^2
O
st
CO
4-1
CO
o
o
pH
O
i.
4-1
C
o
o
ttain. Areas
<^
t
c
o
z
r«<
CO
3
-o
H
CO
CU
OS
eness
^
I-l
4-1
O
01
14-1
IM
W
z.
^^.
r-l
g^£
o
in
CO
4-1
CO
o
o
1 1
O
^4
4-1
c
o
o
ttain. Areas
<^
i
c
o
z
1 I
CO
3
*O
H
CO
01
veness
r-l
4-1
O
/U
VM
14-1
W
S
r-l
g~5
O
O
i i
CO
4-1
CO
o
u
r-4
O
r*
4J
C
o
o
ttain. Areas
<5
t
c
o
z
r t
CO
3
o
rH
CO
01
K
V43
3-25
-------
c^
1
CO
W
,_3
X
<
H
S
0.
o.
CO C
1 1 i-l
co vx
^f
H-; co
< o-
H ca
M >
> -i-l
M 4-1
H CO
1 1 / N C
CO --C I*
Z O CD
a i i 4j
CO i-l
X <
CO
H co
CO >-
O co
o co ^
*S il
iJ O 0
O " -0
OS H
H CM -xt
CO
u
}_t
CO
c
CO
4J
CO
£
3
E
1-1
X
CO
14
CO
Q
CO
l-l
2
_o
t 1
14
«^
Z JS 00
O 13 CT>
CJ CO i-4
CO
K < 4-1
>' C
£C S co
o
w c
O ' I
^x ^ _-*
p-\
o
CO
w
1 (
JQ
o
o
O"*
1 1
w :
u
1 i '
>^.
tn
C^
o o
F*)
X
w
X
u
1 1
*^
CN
^H
X
w
X
w
r-H
-*-^.
&\
00
CO O
tn
r 1
ON ^
CN
CN
JC
to
H
P" '
1
T3
CN
C^
O\
r*^ o
O i*
rO
CJ
CO
CO
U
w
c^
o;
JM
<^
c
H
CO
4-1
<
CO 1
4-1 C
CO O
CU O Z
en O
CO r*^
U i-l CO
O 3
>J 4-1 TH
co C co
E O ID
i-l O OS
l-l
u"*i C^ «^
.
CN O vO O -3* O
O CN CO
*xj U^i -"st
...
10 O cyx O no
U") «s^" 4.)
H < > < -H <
I-i CO 1 -i-4 (0 1 4J CO 1
4-1 4-1 C 4J4-IC O 4J C
cocao ucao cucao
OZ 4JOZ iMOZ
E O u-i O u-i cj
3 ,-H in ,_| [jj ,_|
gf-^tfl WI-HCO r-4co
rHO3 ^O3 SO3
CO 4-1 -1-1 »». 4-1 -H I-H U -H
SCCO h-ICCO CW
O CD O CU S-S O CU
5^OOS &^£CJOS OOOS
0 O O
O
01
cc
CO
3-26
-------
CO
rI
CO
>"
pj
<
<
CO
w
cc
H
u
CO
CO
H
CO
o
CJ CO
J O
O r-l
as H
H &,
O 3
cj 10
en
w <
X
CO
CO
o
T3
00
c
CO
CO
c
o
u
^"N
e :
c.
O.
c
H
"-'
CO
o-
1
cu
^
F-i
J-l
CO
c
M
d)
4-1
r-l
**
CO
o
CO
*O
C
CO
jj
CO
s
*^
X
CO
2
H
CO
Q
CO
t-l
3
_O
r-l
<
X
'x
w
1 1
"v*
m
1 1
o ^ ^o
O O CM O *£>
CO CO CM
^
U
X
w
,_4
CM
T 1
X
U
X
til
1 1
^s^.
ON
ib
H
"^
1
T3
C
CM
ON
LO i-l 00
-I O CO O l*>
P** ^O ^
...
«^- CM oo CM in
CN r-l CM
CM CM CM
^J- r 1 vO
m co i i co r-^
OO r-l OO
CM CO CM
CJ
o
co
w
S3
O
m
01
CO
CO
u
CO
CO
CJ
CO O
CO
co o
14 O
CM
CM
CM
cu
CO
tfl
CJ
r>1
^J
CO
E
H
J^
PH
CO
J-l
CO
o
CJ
14
O
u
C
o
CJ
CO
CO
U-4
W
S
"^,
1 i
g-<
0
m
CO
4J
CO
o
CJ
1 (
o
1.4
4J
c
o
CJ
en
CO
OS
3-27
-------
u
S
H
W
cj
3 CO
O Z
co O
W H
J C-
CG 3
O co
2 co
U3
oo
ON
co o;
< O
CJ fe
>> CO
OS H
< CO
s o
os
-\ ^
"" O
W OS
"" Z
o
o
E-i
W
1-^
P3
OS
^ v
E
a
a.
c
H
\^
CO
o-
,
I-l
CO
Q
CO
^4
3
O
"T";
r 1
i 1
X
W
^
w
I 1
Ul
1 1
O ON
. .
CM O O
CO t 4
"*"
X
W
X
W
1 1
^^
«N
i 1
^H O
r** CN f")
u""i -{-
X
w
X
w
.4
ON
"so
H
11^*
1
o
p
CN
CO vO
I-H < tn
CM i 1 1
CM
r- O
/*" CO O^
O CN +
c-O
C
1 1
1
f-^
1
C"">
i
0)
CO
n
CJ
o
z
0)
CO
cfl
CJ
ON
I
00
CN
I
vO
i-l O
CN CM
CN I
I
cu
co
CO
CJ
^
w
CO
e
H
^
D^
r*
o
X
v>
^
CO
U
CO
0
u
14
o
4J
c
0
CJ
n. Areas
H
CO
4J
4_)
<^
1
C
o
Z
1 1
CO
3
T3
H
W
0)
OS
CJ
c
D
60
C
H
}_,
^J
CO
E
3
E
4
X
CO
s
s-s
o
%^-
Costs
c
H
01
60
c
CO
r*
o
4U
e
01
o
t-i
0)
Areas
14
CO
3
T3
H
CO
o>
OS
c
H
cu
oo
e
CO
u
CO
CO
0>
c
01
£>
H
4J
o
01
<*.
14-
w
2^
^^
i i
^5
O
in
Costs
C
H
01
00
c
CO
f.
^
JJ
c
0)
CJ
}_4
0)
3-,
Areas
14
CO
3
o
IH
CO
01
K
C
H
0>
eo
c
CO
CJ
CO
CO
Ol
c
01
^
I-l
jj
O
V
u^
UH
W
2
^~^
r 1
j^
O
o
r 1
4-J
CA
O
CJ
c
H
01
60
C
CO
r*
C^J
4U
C
QJ
CJ
t-l
Oi
Areas
i i
CO
3
-a
i_i
CO
01
05
C
H
o>
00
c
CO
.e
C.J
3-28
-------
CN
rI
I
f">
W
3
<
H
W
O
OS
3 CO
o z
CO O
II
W H
.-3 A
H-i S
P5 3
O CO
S co
Cd
O
z
W
CO OS
< O
>H 00
OS H
< 00
s o
I U
O
W OS
S H
H Z
O
p o
W 35
> <
H-i M
H OS
CO
13
l-i
CO
TJ
/-^ C
E co
a. 4-1
a. co
c s
t-i 3
^ E
H
CO X
O* co
< s
<
Z >^
1 1
at 1-1
> co
H O
4J
CO
c
1-
0)
U
f-4
<
03
1-
3
_o
1 I
<
X
W
X
U
1 i
-»»
m
i *
X
w
w
t i
**^
CN
i 1
w
X
EzJ
i i
"*-s.
^
_C
00
!-t
!><-l
1
"O
c
CM
ON
(U
CO
CO
O
O
0>
CO
CO
U
ON CN
o o m o
m +
oo o
en o mo
m +
i i i i
ON \o -* O
CM 1
CN
ON .3-
r^ o ON o
0 ^H +
m
CO
CO
0) CO
^ l-l CO CO
"C < 4J 0)
O CO t-i
" >. O <
C CJ O
X -r^ C ^
CO CU C <0
4-1 60 T-4 3
^^ 4J c T3
< -H CU -H
CO 1 >- 00 CO
4-1 C 4-1 C CU
CO O CO CO OS
0) O Z j=
w 0 E 0 C
cs > i 3 T-I
O rH CO £ 4-1
O 3 ft C 0)
>-, U T3 X 0) 00
(-1 4J -H CO CJ C
CO C W S M co
E O 0) 0) £
H cj os s~s &< u
J-i O
O-i -*
i i m
ON
CN O
^
I
O
00 O
1
fNl
o o
+
^H
,
t-i m
+ +
CO
CO CO
4-1 0)
CO )*
o <
CO CJ
CO < 1
01 C <0
C -H 3
cu -o
> CU -H
H 00 CO
4-1 C CU
O CO OS
01 £
«*H CJ C
U-l -rt
H 4J
C 01
Z oi oo
^ o c
i-l ti co
CU £
s-s a, u
o
m
,
vD
CN
0 0
i-H
+
o
ON O
1 i
1
00
m m
1 1
CN
o oo
t-( 1
1
CO
co co
4-1 CU
CO I*
CO O <
CO U
CU i-i
C B co
CU -i-i 3
> T3
H 0) -H
4J 00 CO
U C O>
01 CO OS
VM ,e
<4-i u e
W -H
4J
S C 01
-~^ Ol 60
M CJ C
I-I CO
f-S 0) J=
O PH U
o
r- 1
«
r-»
3-29
-------
CO
1 1
1
CO
tu
t>
<
t-
w
«
O
co
w
iJ
M
cc
0
z
M
U
z
JjS
o_2
/-^
kj
CO
<
u
OL
<
f t
o:
P_
W
Zbi
H
O
H
W
^
ti
H
,j
w
o;
Is
a
a
c
H
s^
cy
z <
o s
H U
CH >
2 -H
3 4J
CO co
CO C
< tl
IS 4J
->. I <
t 1 -j i i
CO -^
O m
C <J
4-1
CO X
w
E x
3 W
E
H i 1
CO CN
--1
>^
H X
CO W
G X
W
r 1
*
OS
03 00
t4 '*"4
_o T
"~ c
i 1 CN
r^
< Os
0)
CO
O
f"^ r^-
*d" Cvl CN O
CN 1
CN
^t 0
m co os o
oo +
CN
03
CO
4) en
s~** i-J 0) CO
C < 4J 4i
O 03 U
r-> >> O <
C 0 U
X -H C i-l
CO 0) C CO
4-1 00 -^ 3
* ' 4-* C T3
< .i-l ft) -H
03 1 (-1 00 CO
4J C 4J C 3>
co o co co os
v o z .c
03 O E U C
CO »-i 3 'H
U f-4 CO E 4J
o 3 -H n 01
>i (-1 13 X rl CO O C
co C 03 S l-i co
E O CU 0) 4=
H O OS s-S Oi O
U O
a, m
CO
' O
i-H
|
CN
r*^ co
i
ro
O I-H
+ 1
00
O CN
-^- -i.
03
03 CO
01 -rt
T-( 60 03
4J C Oi
O CO OS
0) »C
14H CJ C
UH -H
W 4-J
C CU
S 0) 00
\ o c
i-l l-i CO
CU JM
5-S SH O
O
m
^D
CO
.-I O
F*
_L
r-»
in o
i t
^o
*
i-4 (N
1 1
U-1
00 CN
1 1
M
UJ CO
4J ^3
03 )-i
03 O <
03 U
Ol 1-4
C C co
Ol -^ 3
> -a
-H CiJ *H
4-1 00 03
0 C 4>
cu co os
U-j J^
iw a c
U4 -1-1
4J
SCO)
\ 4) 00
M CJ C
^ co
5~5 01 £
O P-i U
O
r 4
r-!
3-30
-------
The sensitivity results are organized into two sets of 6 tables. Each set
contains the primary case and 3 alternative cases. The first set involves
cases focused on alternative assumptions regarding the data base, including
its base year, analysis year, and CO background concentration. The second set
poses alternative cases for I/M program assumptions. These include stringency
level or rejection rate of the program, and effectiveness of the I/M program
in cold-weather areas.
The first three tables within a set contain estimates of the absolute costs
and numbers of residual nonattainment areas for the three years of analysis.
The remaining three tables within a set contain relative change estimates for
the alternative cases as compared to the primary case. There is one table for
each year of analysis. The relative change indices are computed thusly:
Estimated costs _ Estimated cost
% change in = ^ for the alternative for, the primary case
control costs Estimated costs for the primary case
Change in residual = Estimated residual Estimated residual
non-attainment area areas for the ~ areas for the
primary case alternative
Thus, if an entry in any of the relative change tables is negative, it means
that costs for the alternative are lower than the primary case or there are
fewer residual non-attainment areas for the alternative than for the primary
case.
As can be seen, these can be a positive change in costs associated with an
alternative and a negative change in residual non-attainment areas -- or vice
versa. There also can be no change in one or both of these items with either
a positive or negative change in the other. In other words, there can be
numerous possible combinations of directional changes in the costs and
residual non-attainment areas indices. In addition, the directional changes
can vary over time (i.e., 1987, 1990, or 1995) or by the standard level.
3-31
-------
The reason for the fairly complex pattern of change in the relative change
tables is the logic used for developing each area's control strategy and the
fuel savings attached to the strategies. For example, if by changing a
particular assumption an area goes from a situation requiring I/M at 20% to
one where a TCM program using local strategies only is needed, control costs
would drop dramatically, but the non-attainment situation would not change.
In some areas, relaxing the standard from 12/1 to 15/1 actually causes control
costs to increase because TCMs are no longer required. (At low levels of use,
TCMs save money because of fuel savings.)
We turn to a brief description of the alternative cases investigated in this
sensitivity analysis.
3'4-1 Case #2; 1984 Base Year
The primary case uses 1982 as the base year of analysis because it is the
mid-point of the 1981-1983 time period used for air quality data. Some
reviewers of drafts of this report object to this choice of base year as the
1981-1983 air quality data are used to define the design value in an area as
of January 1, 1984. To them, 1984 should be the base year.
Having a different base year obviously affects the results. A 1982 base year,
in essence, gives MOBILES two more years in which to reduce mobile source
emissions due to FMVCP.
Compared to the primary case, with its 1982 base year, Case #2 results in
about a 9 percent increase in costs for the 9 ppm, 2nd-high standard in 1987.
The number of residual non-attainment areas increases greatly (+17) for the
same standard and time period. This pattern holds time for almost all of the
alternative NAAQS and time periods investigated.
3.4.2 Case #3; 1988 Analysis Year
The EPA MOBILE3 mobile source emission projection model was used to project
composite fleet emissions and I/M effectiveness. Since MOBILE3 projects for
3-32
-------
January 1 of each year, it could be argued that 1988 rather than 1987 would be
the proper MOBILES projection year for assessment of compliance with
December 31, 1987 Clean Air Act for attainment of the CO standard. As shown
in Table 3-2 through 3-7, this change results in somewhat fewer nonattainment
areas and lower costs in 1988 than in the base case of 1987. Costs in 1990
and 1995 are also generally lower, since the first analysis year is the
controlling factor in designation of required control programs.
3.4.3 Case #4; 1 ppm CO Background
The primary case assumes a 0 ppm CO background level. This will not always be
the case, so a case with 1 ppm background was also analyzed. For both 9 ppm
standards analyzed, the number of residual non-attainment areas and control
costs increased slightly. This is simply because a slightly greater emission
reduction would be required to attain the standards. For the higher
standards, there is no change in non-attainment areas, and costs decrease due
to increased use of TCM with high fuel savings.
3.4.4 Case #5_;40 percent Maximum I/M Stringency
The primary case puts a 30 percent limit on the number of cars that are
assumed to fail an I/M program. It is theoretically possible to fail more
cars; one such possibility is to fail 40 percent of them in those areas that
do not attain with a 30 percent stringency program. The 40 percent failure
rate, selectively applied, comprises the essence of Case #5.
This change reduces residual non-attainment areas somewhat. For the 9 ppm,
2nd-high standard, the change results in 3 fewer non-attaining areas in 1987;
there is no difference in 1990 and 1995 non-attaining areas. For most of the
standards analyzed, the change results in substantially increased control
costs. For the 9 ppm standards, this increase is about 9 percent; it is
11 percent for the 15 ppm standard.
3-33
-------
3.4.5 Cases #6 and...7;.,.. 50 and 100 percent I/M Effectiveness
As discussed in 3.3.1, the primary case used a linear relationship between
50 percent effectiveness at 50°F and below (applied as a step-function), and
100 percent effectiveness at all temperatures. For Case 6, slight increases
in non-attainment and decreases in costs are observed. This minimal
difference would be expected since the step function closely approximates the
primary case. The increased I/M effectiveness in Case 7 results in
significantly reduced non-attainment and control costs, as shown in Tables 3-8
through 3-13.
3.5 Incremental Costs of CO Control*
Costs presented in previous sections represent the portion of mobile source
control costs allocated to CO emission control as opposed to total program
cost. Tailpipe emission control systems as well as in-use programs such as
I/M provide for simultaneous control of more than one pollutant. Thus, the
most accurate indication of CO control program costs in obtained from an
allocation of costs among multi-pollutant programs.
Nonetheless, presentation of the incremental costs of CO control is also
informative from the standpoint of those costs resulting directly from setting
an ambient air quality CO standard. In this case, costs of programs or
portion of programs where CO control is the sole emphasis are isolated.
Programs which are implemented for control of other pollutants and which are
required regardless of the need for CO control are not considered. The
following discussion presents incremental costs for controlling CO through
FMVCP as well as through I/M and TCM programs.
*This section deals in a partial sense with reversibility of control costs
associated with systems designed to reduce HC and NOX emissions, as well as CO
emissions. It is not meant to describe total incremental costs for the
alternative standards.
3-34
-------
3.5.1 Incremental Costs of FMVCP
Estimating costs of controlling CO only through the FMVCP is exceedingly
difficult. Since control of CO, HC, and NOX has been mandated simultaneously,
all FMVCP system costs represent control of all three pollutants. Little
attention has been given to the situation of assessing the configuration and
costs of control systems given the premise that CO control is not required.
Because of the lack of pertinent data, the assumption is made that the
three-way catalyst would be used in future technology cars (1981 and later)
even without CO control requirements.
However, inclusion of the oxidation catalyst in combination with the three-way
catalyst can be attributed solely to the requirement of controlling CO to the
statutory limit of 3.4 grams/mile. Hence, the incremental costs of CO control
resulting from FMVCP can be viewed as the cost of requiring oxidation
catalysts on future technology cars. Based on information from auto
manufacturers as well as independent sources, EPA estimates that adding an
oxidation clean-up catalyst to a three-way catalytic converter ranges from $9
to $21 per auto (see Appendix A). Using the approach outlined in Appendix A
for determining annualized hardware costs, the fleet-wide annualized costs for
a 3.4 FMVCP will be about from $2,000 million in 1987 and $2,600 million in
1995. This takes into account hardware costs only and does not assume any
maintenance or fuel benefits or penalties associated with the oxidation
catalyst.
3.5.2 Incremental Costs of I/M and TCM
While the majority of areas that will be implementing I/M programs experience
both ozone and CO problems and thus need control of both HC and CO, several
areas have just a CO problem. While these areas also receive benefits of HC
control from I/M programs, the programs are instituted primarily because of
the needs for CO control. As a result, the incremental costs of I/M for CO
control represent total program costs in these areas. These areas include
those with current plans to institute programs for I/M that are designated as
3-35
-------
non-attainment for CO only, as well as areas identified in this analysis as
possibly needing I/M programs for CO in the future but which are not currently
designated non-attainment.
Using this approach for determining incremental CO control requirements, 13
areas in 1987 could need an I/M program for CO only under a 9 ppm, 2nd-high
standard. In these areas, gross I/M costs are estimated to be $19 million (in
constant 1984 dollars); net I/M costs are estimated to be around $18 million,
assuming fuel benefits are associated with the program.
The incremental cost of TCM for CO cannot be estimated readily because of the
variability of the effectiveness of similar strategies of HC and CO. An
area-wide program implemented to combat an ozone problem may not necessarily
alleviate an area's CO problem, particularly if the problem area is a hot
spot. However, if it is assumed that (1) all TCM measures are equally
effective for HC and CO in areas with both a CO and ozone problem, and (2) TCM
measures do not provide any ozone benefits in areas that do not attain the CO
standard only, then certain cost calculations can be made. Given these
assumptions, the costs of TCM programs are assumed to be split equally for
areas not attaining the CO and ozone NAAQS, and allocated solely to CO for TCM
costs (and savings due to fuel benefits) in CO-only non-attaining areas. This
results in 1987 CO incremental gross costs of $65 million (in constant 1984
dollars) and net costs of $17 million.
3-36
-------
REFERENCES FOR SECTION 3
1. Armstrong, Jane. 1985 Personal Communication. Office of Mobile Sources,
EPA. June 14, 1985.
2. California Air Resources Board (GARB). 1985. EMFAC6 Outputs dated
November 2, 1983 ("Predicted California Vehicle Emissions") and
March 13, 1985 ("EMFAC6D Emission Factors").
3. Energy and Environmental Analysis, Inc. (EEA). November 1984. "The
Motor Fuel Consumption Report--Eleveth Periodical Report." Motor Vehicle
Emission Laboratory, EPA, Ann Arbor, MI.
4. Energy Information Administration (EIA), U.S. Department of Energy.
1984. "Petroleum Marketing Monthly, November 1984."
5. Jordan, Bruce C., Environmental Protection Specialist. 1979. Memorandum
to Ken Lloyd, Economic Analysis Branch, EPA. "Comparison of CO Data from
DRI 'Hotspot' Study to Data in SAROAD."
6. Lorang, Philip A., et al. 1982. "In-Use Emissions of 1980 and 1981
Passenger Cars: Results of EPA Testing." SAE Technical Paper 820975.
7. Lorang, Philip A. 1985. Personal communication. Office of Mobile
Sources, EPA. July 5, 1985.
8. Palmini, D. and Rossi, D. 1980. "What Price Air Quality? The Cost of
New Jersey's Inspection/Maintenance Program." Journal of the Air
Pollution Control Association, 30(10): 1081-1088 (October).
9. Platte, Lois. 1985. Memorandum to Bruce Jordan, Ambient Standards
Branch, EPA. "QMS Review of CO NAAQS Analysis." June 26, 1985.
10. Stanford Research International (SRI). December 1979. "Methodologies to
Conduct Regulatory Impact Analysis of Ambient Air Quality Standards for
Carbon Monoxide." Prepared for U.S. EPA.
11. U.S. Congress. August 7, 1977. Clean Air Act. 42 U.S.C. 1857 as
amended 1977, PL 95-95.
12. U.S. Environmental Protection Agency (U.S. EPA). February 1977.
"Guidelines for the Interpretation of Air Quality Data with Respect to
the National Ambient Air Quality Standards." Guidelines Series OAQPS
1.2-008.
3-37
-------
13. U.S. EPA. March 1978. "Mobile Source Emission Factors."
EPA 400/9-78-006.
14. U.S. EPA, Inspection and Maintenance Staff (OMSAPC). April 1979a.
"Questions and Answers Concerning the Technical Details of Inspection and
Maintenance."
15. U.S. EPA. October 1979b. "Proposed National Ambient Air Quality
Standards for Carbon Monoxide Draft Environmental Impact Statement."
16. U.S. EPA. 1981. "Emission Effects of Inspection and Maintenance at Cold
Temperatures." EPA-AA-IMS-81-24.
17. U.S. EPA. 1984a. "National Air Pollutant Emission Estimates,
1940-1982." EPA-450/4-84-028.
18. U.S. EPA. 1984b. "User's Guide to MOBILE2 (Mobile Source Emissions
Model)." EPA-460/3-84-002.
19. U.S. EPA, Inspection and Maintenance Staff, OMSAPC. 1985. "Inspection
/Maintenance Program Implementation Summary."
3-38
-------
4. STATIONARY POINT SOURCE CONTROL
This section presents the costs of stationary point source control
requirements to meet alternative CO ambient air quality standards. The
sources were examined in isolation, using an assumed background CO level of
2 ppm. The control requirements and costs were selected to meet the ambient
standard at the lowest annualized cost of control including any steam credit
obtained from burning CO. A complete description of the methodology used and
assumptions made in this simulation is presented in Appendix B. Appendix B
has been left unchanged from previous analyses, and thus contains some
references which do not apply to the standards currently under consideration
(e.g. 5 expected exceedance forms). The contents of this section also have
been reviewed and updated from previous analyses.
4.1 STATIONARY POINT SOURCE COSTS
Tables 4-1 and 4-2 present the results of modeling and costing procedures.
Each table indicates the number of facilities which require control under
alternative CO NAAQS. Costs are presented in 1984 dollars and reflect the
total initial capital cost and the annualized cost for a 1987 attainment
date. It should be noted that the results shown here are also applicable to
attainment dates in 1990 and 1995 due to the exclusion of growth (see
Appendix B.4) and retirement. The latter would tend to reduce costs for
existing sources as it would deplete the existing source inventory to some
degree. Lacking information on equipment age, the assumption was made that
the equipment population would remain static; that is, retirement of obsolete
equipment would not occur during the period of concern.
4-1
-------
f <
1
sr
W
hJ
<
H
£
H
CO
p"i
Q
M
PQ CO
CO <
D H
O CO
CO
o
Ij2 u /_v
Mam
o > o
PH M r-<
H -to-
Pi S >3-
< £6 00
OH -i
H iJ ^
H <
H pd
CO P
IS
sc
CO O
H M
CO W
o
u
J
H
M
flj
C^
O
s
_o
p
0
B
O
_o
a
*j
a
OOOOOO r^ C
E CO O CM CM C
a i~ n cr c
a. -
CM 4 IT
in ^^ CM r"
*-*
tf o r-» co f^
CM CM CO *O f^l
«. «
S1OOOOOO I- C
QJ O O ^ P^ ^ r*
sJ m »^ o cs co c
^1 ^4 (N \O f*l <
C~t *f \ fiC
«. «
ElOOOOOO p^ C
aJoooco^fn p^.
Cu o *>n P") ^ CD
n . o
r^l CM & &. m
n CM -^ "-
VI
m
u
u
i
3
w
i
a
«^
o
^
o
u
a
Ct
^
=
3
Z
E
ai-^-aO'^«^o -^ co
U~l
'^
g
ScMO-^-ffCMO 1 j;
CM
"*
olcM-<-H4 15EOC
vjoxsoa-< <
r^uwtg-^uu^u r
aQubu3 c C
SU«O3-< P-H f-
sc
c
*3
v 3
» o
Ui
o
u
V
3
3
E
3
o
4J
o
s
i >s
o
u|
III
to I
CO «"i
CM CO
U.
a
4-2
-------
W
£
H
1-^
Jj~
Q
5s
M
^
PQ
CO
* £j
C/D S
w H H
ei <;
p^ p^ f^x
O W ON
J. ""' t ""* "
r~j hJ ^^
H
CO p3
*~"^
Pi O
P~l 1
H
to ffi
H O
CO M
O W
o
o
w
M
3
***^
Cj
^5
^5
^ 4
0
o
o
a
o
D
O
o
o
V
N
^1
q
3
C
i) =
& <
10 i
*-4
_. _,
«J O^
Q] CO
fSJ " ^
-HJ 1
O 0 0
ao
«
1A
1
0 0 O
r- y-i
00 -* -*
00 f)
1 1
o o r~ c:
K
1 V.
CM ] -
O O r~ C
-3- i/
CO , C
m c
s^ r
J
3
3
»
3
3
1
3
3
1
« \
O 0 O
iTi e*J -H
«^ P^ »^
1
*£>
^H
I
O 0 r* c
a- i-
co r
«>
1-1 ; r-
f
v> «
si o
£j in
CJ tT)
-1 '
03
U
U
bi
3
O
CQ
"3
O
«4
O
Ui
AJ
o
u
1-1
o
w
o
-2
£
3
Z
c
a
in
^
s
a. N
CM
^H
£. (M
^
0\
gn
-1
14
>J "O
u >l
u JI
a c
_3 0
£ u
r^ <*4
O
^
S
0 O O
r^ m o
P4 \o in
CO f^"
i i
< o -^
o -* %
W V Q
BO AJ 1*
U W O
O O l>- S
sf CM r
CO C
n" r
«
O -« I 0
1
CM O x.
3
1
5
1
>
3
5
H
CM O 1 1 Cf>
i r
CM -^ ~H C
r
E l
C C
~c n
Eg c
3 M O
^* 01 C *
CB ^^ -*^ 4J
T* U Q
>. jO 3 1*
U O u CJ
-------
4.1.1 Attainment of a 9 ppm Eight-Hour Standard
4.1.1.1 Costs
Attainment of an eight-hour 9 ppm standard results in a capital expenditure of
almost $84 million (see Table 4-1) and an annualized cost of $27.3 million
(see Table 4-2). Three industries incur positive annualized costs: primary
aluminum amounts to $44 million, followed by gray iron with $410,000 and
incineration with $7,000. The bulk of the negative annualized cost stems from
carbon black, which accounts for $16.5 million in annualized savings. Steel
incurs an annualized savings of $320,000 and maleic anhydride would save
$220,000.
4.1.1.2 Tonnage Reduction
Table 4-3 shows the degree of CO emission reduction under alternative
eight-hour standards. The eight-hour 9 ppm standard results in 964,100 tons
of reduction, over 75 percent of which is produced by the carbon black
industry. Steel accounts for 9 percent of the reduction; gray iron represents
8.5 percent of the reduction (82,000 tons); and primary aluminum accounts for
4.4 percent (42,000 tons).
4.1.2 Attainment of Alternative Eight-Hour Standards
4.1.2.1. Costs
The capital cost required by industry under the alternative eight-hour
standards (Table 4-1) indicates that a 34 percent increase is caused by the
7 ppm alternative. The most significant increases under 7 ppm occur for
maleic anhydride (67 percent), steelmaking (386 percent), and gray iron
(65 percent). The 7 ppm standard shows a positive annualized cost (Table 4-2)
for four industries: gray iron, primary aluminum, automobile manufacturing,
and incinerators. The largest increase in cost occurs in the gray iron
industry. This happens since 50 percent more sources must control than under
either 9 ppm standard (2nd-high or one expected exceedance).
4-4
-------
H
CO
O
U
W
H
o o
n"i O
O 0
O ^^
o o
o --
CO
00
fM
03
o o
o -<
ro
1
st
ia
«
H
1
H
EC
O
H
W
>H*
P-.
'ra'
C
o
4-1
m
o
"~^
0^
u
gcoo-ioo-vooc;
A.**>>>>*
: CM m in o ^- ^-*
l--| 00 -< -1
00
C
H
o
1
V
1I
a
c
O
J3
B
H C
j: o
ca i-i
_B -
ai x
o> a
fn
o
CO
2
O
H
O
H
4-5
-------
Three industriesmaleic anhydride, carbon black, and steelmaking--provide
negative annualized costs sufficient to reduce the annual cost to
$13.0 million compared to the $27.3 million under the 9 ppm standard.
Under a relaxed alternative eight-hour standard of 12 ppm, the total capital
expenditure decreases by 14.7 percent. Relaxation of the standard for both
maleic anhydride and steelmaking under the 12 ppm standard increases capital
expenditure but decreases annualized costs. This happens because the control
strategies which may be used to comply with a 12 ppm standard provide a
greater benefit in terms of fuel savings. This control strategy removes less
CO, however (see Table 4-3), and therefore could not be used for the 7 or
9 ppm standards. Three industries show positive annualized cost for the
12 ppm standard: primary aluminum, gray iron, and incinerators. Under the
12 ppm standard, however, the total annualized cost increases by 10 percent.
Again, this happens because the reduced stringency eliminates some control on
carbon black which provided a net savings of $7.6 million. Under a 15 ppm
standard, capital costs and annualized costs drop to about 42 and 59 percent
of those for 9 ppm, principally due to elimination of controls at one primary
aluminum plant.
4.1.2.2 Tonnage Reduction
The reduction in CO emissions attributable to the alternative standards is
shown in Table 4-3. The 7 ppm alternative results in a 25 percent incremental
reduction over the current standard. This alternative represents the most
restrictive concentration level of any standard as indicated by the level of
CO reduction achieved in response. The 12 ppm alternative reduces emissions
by 538,400 tons from baseline levels and represents a 44.2 percent decrease
from the reduction required to attain 9 ppm standards. A 15 ppm standard
would require a total decrease of 407,100 tons, about 58 percent less than the
reduction for 9 ppm.
4-6
-------
REFERENCES FOR SECTION 4
PEDCo Environmental, Inc. (PEDCo). February 1979. "Control Techniques and
Costs for Carbon Monoxide Emissions -- Interim Report #1."
4-7
-------
5. MOBILE SOURCE ECONOMIC ANALYSIS
5.1 INTRODUCTION
The mobile source control costs presented in Section 3 consist of obligations
to be paid by local governments or by individual car owners. The impact of
these costs is analyzed in terms of their effect on these two sectors.
Four scenarios were chosen for economic analysis. They are the current 9 ppm
2nd high NAAQS, and the 9, 12 and 15 ppm alternative NAAQS for a 1 ex ex form
of the standard. This analysis is performed for the primary case assumptions
described in Section 3. Most of the numbers come from the previous CO cost
economic analysis (EPA, 1982), updated to 1984.
It was assumed that the payments made by local government, primarily for
initial investment of I/M and gross annual TCM costs, are paid directly from
local government revenues. The relative ease or difficulty in paying the cost
depends on the revenue received. The costs therefore are compared to typical
government revenues for the localities involved. I/M programs, once
developed, are typically designed to be self-supported by the user charge, or
fee.
The costs paid by individual car owners, which include FMVCP, inspection, and
I/M repairs, were calculated from average cost estimates. The impact of these
costs on average individuals would be difficult to specify. In addition, the
variance of costs around the average estimates is not known and could be more
significant than the averages themselves.
Examination of costs borne by motorists may be more useful if performed across
income levels, that is, by determining the relative burden on different
economic classes of motorists. The fuel savings provided by I/M programs and
TCM are examined in the same manner. Fuel savings can help reduce the net
cost of the control programs to groups of motorists, but they are received by
this portion of the population only and cannot be transferred to cover other
costs. The lack of liquidity of costs and savings makes the economic analysis
5-1
-------
not only a question of direct and indirect impacts, but also one of the
progressive or regressive nature of costs. In other words, is the cost being
paid by those individuals best able to pay, or is a disproportionate amount of
the cost paid by lower income groups? It should be stressed that this
analysis ignores benefits of control accruing to all income groups. Thus, it
is unfair to just focus on cost burdens to different income groups. Even if
the control requirements are regressive on net, they are consistent with the
"polluter pays" principle.
5.2 ANALYTICAL PROCEDURE AND ASSUMPTIONS
5.2.1 Federal Motor Vehicle Control Program
Regardless of the ambient air quality standard level, the cost of FMVCP must
be paid. FMVCP, while separate from the National Ambient Air Quality
Standards (NAAQS) program, provides emission reductions necessary to attain a
standard. In the absence of FMVCP, emission reductions provided by the
program in a nonattainment area would have to be obtained through additional
control measures.
Cost of the program (see Appendix A) is derived through control of
four sectors: heavy-duty vehicles, light-duty vehicles, light-duty trucks,
and motorcycles. There are no CO standards for aircraft. Most of the total
annual cost of the program results from control of light-duty vehicles,
light-duty trucks, and motorcycles. It is assumed that costs from these
sectors are borne by owners of these vehicles. The costs then, can be divided
by income group using automobile ownership information obtained from the
Department of Transportation (DOT, 1974). it is assumed that ownership
patterns for light-duty trucks and motorcycles are similar to those of
light-duty vehicles.
FMVCP costs for heavy-duty vehicles comprise the remainder of the total cost
of the program. It is assumed that most heavy-duty vehicles are used in
industry or for commercial purposes. Because these costs are imposed on all
heavy-duty vehicles, the cost of control can be passed through to the final
5-2
-------
consumer. The relatively small magnitude of these costs, $129 to $207 million
for all heavy-duty vehicles, depending on the analysis year, indicates that
pass-through of costs would have virtually no measurable effect on the prices
of the goods and services involved.
5.2.2 Inspection/Maintenance Program
5.2.2.1 Initial Capital,.Investment
The estimate of initial capital required to start an I/M program is dependent
solely upon the number of cars inspected under the program. An average figure
of $18.1 per car generates the cost.
It is assumed that initial expenditure is made by the county government.
Problems in budgeting and allocating the county funds are not considered in
this analysis; instead, the initial investment is compared to total county
revenues including transfers from Federal and State Sources. This provides an
indicator showing any difficulties in meeting the capital requirements of the
program at the extremes of county income levels.
5.2.2.2 Inspection Costs
Under an I/M program, all registered passenger vehicles would be inspected
each year. For the main analyses, a fee of $7.00 per vehicle is estimated to
be the annualized cost of the inspection. This fee is assumed to cover the
capital costs and operating expenses of the inspection and would be paid by
the owner of the car.
The total inspection cost, therefore, can be divided into the income groups of
car owners who pay for the inspection. The Department of Transportation
statistics again are used to accomplish this. The resulting distribution of
cost by income group is examined to determine where the burden of cost is
heaviest. Although the average car owner pays $7.00, this distribution does
not tell what, if any, redistributive effect results from the program.
5-3
-------
5.2.2.3. Repair Costs
Any automobile failing the inspection test is required to undergo maintenance,
with the cost of repairs estimated in 3.2.1.2. Each State, in its State
Implementation Plan (SIP) for attainment, will determine the test criteria for
passing and failing the test. A given stringency factor (i.e., number of
failures) can result in several emissions criteria for failure. For each
State, test thresholds could be different.
Historically, even with this variation, there is a correlation between the age
of the vehicle and failure rate. Data regarding this correlation were
published by the Air Pollution Control Association (APCA, 1979). Using these
data, a statistic indicating likelihood of failure by age was created. The
probability of failure was divided by the probability of failure of a car of
mean age (DOT, 1974). This age-failure index is presented in Table 5-1. The
statistic then can be applied to any stringency factor to determine the
increased or decreased probability of failure as a function of age.
A distribution of age of car by income of the owner was found in the National
Personal Transportation Study (DOT, 1974). These data are presented in
Table 5-2. For each income level and corresponding mean age of car, the table
also shows the age-failure index, which is calculated by linear interpolation
between years.
A distribution of repair costs is determined from these statistics by
multiplying income group ownership by the corresponding age-failure index.
The resulting distribution is normalized by dividing by the summation of the
age-failure index times the percent of cars by income in order to indicate a
distribution of cost. In functional form, the statistic is calculated as
follows:
5-4
-------
TABLE 5-1
AGE-FAILURE INDEX
Year
1
2
3
4
5
6 and above
Mean
Percent
Observed Failures
55
49
60
69
70
71
65.8
Index
Observed -f Mean
0.836
0.745
0.912
1.049
1.064
1.079
1.000
TABLE 5-2
DISTRIBUTION OF MOBILE SOURCE COST FACTORS
Income Group
< 6,600
6,600-11,000
11,001-16,600
16,601-22,000
22,001-33,000
>33,000
Unreported
Age
7.0
6.1
5.8
4.8
4.6
4.0
5.5
Age-Failure
Index
1.079
1.079
1.076
1.060
1.057
1.049
1.069
Percent
of Cars (%)
5.91
9.42
20.00
18.02
24.53
12.50
9.61
Age X I
Percent
0.0638
0.1016
0.2150
0.1910
0.2593
0.1311
0.1027
)istribution
by Income
0.0599
0.0955
0.2020
0.1794
0.2435
0.1232
0.0965
5-5
-------
.._.IG. x AF
*(IG x AF)
where
IG = Percentage of cars by income group
AF = Age-failure index
* = Summation over all income groups.
Each factor is presented in Table 5-2.
It should be noted that implicit within this calculation is the assumption
that the mean failure rate of a car whose owner is part of a particular income
group can be estimated by the failure rate of a car of mean age within that
group. Strictly, this would be true only if the age distribution within each
income group were normal. The resulting distribution therefore provides only
an indicator of the burden of cost.
The normal distribution of repair cost is multiplied by the total repair cost
for each scenario to indicate the relative burden of repair cost by income
group.
5-2.2.4 I/M Fuel Savings
The fuel savings accrued by I/M programs are a direct result of the
maintenance performed on vehicles that fail the emission test. The fuel
savings benefit is received only by car owners that pay repair costs. The
value of fuel savings can be divided using the same statistic created for the
distribution of repair costs presented in the preceding section. The money
saved by reduced fuel consumption is subtracted directly from the repair costs
since the costs and savings are not independent.
5-6
-------
5.2.3 Transpprtation Contrp1 Measures
5.2.3.1 Gross Annual Cost
The annual cost of TCM is assumed to be paid by the local county government
(there is no direct charge to the motorist) and is treated as if it were paid
by general county revenue. Since this is an annual cost, the revenue is
assumed to meet the cost each year. It is possible that some portion of the
cost could be paid by state or federal funds. Placement of the burden on the
counties estimates the maximum possible impact for each county requiring TCM.
The average cost of control per county is compared to a range of county
revenues obtained from the 1972 Census of Governments (Bureau of Census). The
revenues were inflated to 1984 dollars using Federal Reserve Bank GNP
deflators. This provides an indicator of the ability of a county to meet the
obligation imposed by TCM annual costs.
5.2.3.2 TCM..Fuel. Sayings
Fuel savings result from TCM by reducing the number of miles or hours traveled
(see Section 3). The reduction in gasoline consumption is assumed to be
constant per ton of CO reduced by TCM. This corresponds directly to the
reduction in VMT. The fuel savings therefore are divided equally among all
drivers on the basis of current VMT.
The National Personal, Transportation Study (DOT, 1974) also provides data on
the percent of. total VMT travelled by each income group. Multiplying these
percentages by the total value of TCM fuel savings under each scenario gives
an estimate of the benefits of fuel savings to each income group.
5-7
-------
5.3 ECONOMIC IMPACTS
5.3.1 Federal Motor Vehicle, Control Program
Table A-ll shows the annual cost of FMVCP, with the total cost for passenger
cars, light-duty trucks, and motorcycles equaling 3.1 billion in 1987.
Table 5-3 shows the estimated distribution of these costs by income group.
Since this distribution is based on the percent of cars owned by each income
group, the first column of the Table 5-3 also reflects the portion of FMVCP
costs allocated to each group.
A previous analysis has shown that the marginal number of cars per increased
dollar in income decreases with higher incomes. Thus, in proportion to
household income, a low income household will face a larger burden from FMVCP
than will a high income household. This reflects the necessity for lower
income households to spend a greater proportion of their incomes on
transportation. This does not imply that the total burden on the low income
group is disproportionate to its segment of the population. For example,
approximately 12.8 percent of all households earned less than $6400 per year
in 1979. By comparison, the same group paid only 10.6 percent of the total
FMVCP. The result, therefore, indicates that there is no real additional
burden to low income households as a group, although individual households
within the low income group pay a substantially higher portion of their income
for FMVCP than do individuals from higher income households. (U.S. EPA, 1982)
This impact is present with all other scenarios discussed. The cost
distribution by income group presented will serve as a basis for comparison
with the costs of the other control techniques.
5-8
-------
TABLE 5-3
DISTRIBUTION OF FMVCP COSTS'1 BY INCOME
Income Group
< 6,600
6,600-11,000
11,001-16,600
16,601-22,000
22,001-33,000
>33,000
Unreported
Total
Percent
of Cars (%)
5.9
9.4
20.0
18.0
24.5
12.5
9.6
Average Number of
Autos per Household
0.4
0.8
1.1
1.3
1.6
1.9
1.2b
1987 FMVCP
Costs ($106)
185
294
622
551
750
379
297
3,078
a - Costs for LDV's, LDT's, and motorcycles.
b - Mean number of automobiles per household.
5-9
-------
5.3.2 Inspection/Maintenance Program
5.3.2.1 Initia 1_ Investment. Cost
Table 5-4 presents estimated initial investment costs for I/M programs. The
initial investment was estimated using a per-vehicle investment of $18.1 and
projected I/M area car counts in 1995. Costs for ozone NAAQS nonattainment
areas were divided by two to allocate half of the investment to CO. Since the
initial investment was determined by projected car counts, average county
investment indicates that areas requiring I/M under the 9 ppm standards have
smaller car counts than those still needing I/M under the 12 and 15 ppm
alternatives. For counties with population of 100,000 or more, the average
revenue is $485 million per year (U.S. DOC, 1972). The range of initial
investments per county in Table 5-4 represents between 0.5 and 2.0 percent of
this average county revenue. Since revenues of areas requiring I/M for the 12
and 15 ppm standards will be higher than this average, it is expected that all
counties will actually fall at the lower end of the range above.
5-3.2.2 Inspection Cost
Table 5-5 presents the distribution of inspection costs by income group for
the scenarios under analysis. Since the distribution of cost is dependent
upon the number of cars in each income group, the portion of the total
inspection cost paid by each income group remains constant. Any fluctuation
in cost of each inspection results from a change in the number of cars tested
under the program.
As in the case of FMVCP, the inspection cost is regressive in nature. As
income rises, there is a diminishing marginal increase in the number of cars
owned by a household. Since the cost is directly proportional to the number
of cars, the relative cost of inspection decreases for higher income levels.
In simple terms, this indicates that a greater portion of the income of low
income families is spent on inspection than that spent by higher income
families. As pointed out for the FMVCP, this does not indicate that the
5-10
-------
TABLE 5-4
INITIAL I/M INVESTMENT COST
(In constant 1984 dollars x 1Q6)
Alternative NAAQS (ppm)
9/2nd High 9/1 ExEx 12/1 ExEx 15/1 ExEx
Initial Investment
Number of Counties
Average County
Investment
316.2
129
2.5
278.4
112
2.5
69.9
7
10.0
55.2
2
27.6
5-11
-------
ITS CO ««"
Cvl
CM
c-i
CNI
tn
f
en
_
ID t/3
-------
proportional burden on the low income group as a whole is greater than their
portion of the population. The additional burden is placed on the individual
low income household with a car.
5.3.2.3 Repair Costs
Repair costs of the I/M programs are divided by income group using statistics
created from the age-failure relationship (see Table 5-2). Results of these
distributions of cost are presented in Table 5-6.
It should be remembered that the distribution of cost by income group
represents a distribution based on historical failure rates. The actual costs
distribution could differ from the estimates presented depending on the
individual threshold values of emissions used to determine failure in each
program.
The effect of increased failure in older cars tends to exacerbate problems of
the low income group. As Table 5-2 shows, inclusion of the age-failure index
causes the distribution of repair costs to be skewed towards low income groups
compared to the original distribution of car ownership. The shift in
distribution, while never more than 0.1 percent for any income group, adds to
the adverse income distribution impact of mobile source costs. Again, the
total burden on the low income group is not disproportionate; it is individual
low income households that may be adversely affected.
Since the percentages calculated to distribute the repair costs are
independent of the scenario analyzed, relative impacts among income groups is
the same for all scenarios. The difference in absolute impacts is brought
about by increases or decreases in program stringency and the number of
counties in which programs are required.
5-13
-------
in
o-
ft)
o m
t_> o.
uu
UJ -* > *»
_l CO
on u_ » o-
«I c=> «x <
^ ce jj
* i-o o as
~ CM CM CM
CM I
CD l^» ITS ^
« CM
u-j
^ «M
CM
CM I CM
eo *M -o r^>
c» -" CM CM ro
»- » 10
^ K> CM CM CO
CM CO
CM
CM CM 10
m * ir> - «r i
_I CM u-> «r ^ ro CM I
in o- o* 10
CO r*- irt *
to o~ ^ w oo
«o ^ i^> ^ ^-
CM CM CM
O J
Z^ i
i
o-
o-
-» I CM
<=> O <=> CJ O
* *O CM IO
^ * CM IO
I I I I
5-14
-------
5.3.2.4 Fuel Savings for I/M
The fuel savings for I/M are related directly to the repairs performed on
vehicles that fail inspection. Fuel savings therefore can be distributed
among income groups using the same percentages used to distribute repair
costs. Table 5-7 presents results of this distribution of fuel savings.
Since fuel savings occur directly from maintenance, the value of these savings
can be subtracted directly from the repair cost to estimate a total cost of
repair. In all cases, the amount of fuel savings for each income group
exceeds the corresponding repair cost. The total annual impact of repairs
including fuel savings is a net savings. Table 5-8 presents these net savings
for each income group.
The statistical nature of the repair cost and fuel savings (i.e., using
average costs for each without a distribution of costs) makes it impossible to
guarantee that every car owner who requires repair will receive a net
savings. There is no assurance of a balance between repair costs and fuel
savings for each motorist. A motorist might be stuck with a repair bill and
no fuel savings. However, in terms of the macroeconomic impact, the combined
fuel savings and repair costs provide a greater savings in proportion to
income for low income groups than for high income groups.
5.3.3 Transportation. Control Measures
5.3.3.1 Gross Annual Cost
TCM's are necessary in a relatively few number of counties compared to I/M
programs. The number of counties requiring this kind of control ranges from
36 for the 9 ppm/second-high standard to 1 for the 15 ppm/1 expected
exceedance alternative, using the primary case assumptions. The average gross
annual cost per county ranges from $11.4 million for the 9 ppm/1 expected
exceedance alternative in 1987 to $0.4 million under the 15 ppm/1 expected
exceedance option in 1995. Results for all scenarios are presented in
Table 5-9.
5-15
-------
in **
in
o*
o-
«r r- ^- ro r* a*
c> -o ^ « ^ <^
r-* r*» r^. o- in
tNJ Ul IO «O bO
in
ro
ex rx \n in ~o
(0
« O- CO O 1O -O-
en at
to i-
as to n
M« C3I ««
CS| Or* O Irt OD «" O^
UJ => 00
=1 cr-
00 0
in » cvi -a
CD *^> co in
c> CM CM
<=> o> <=> o
o o o> o
-JS O O O
ca
01
a
i t i >
5-16
-------
en
to
(O
en
>
en
-^
CM » ro in CM CM
^- r in ro OB
-o CM o- ^- i^>
o* ^ i^* ^ CM
CM CM
in
r
in
en a
5-17
-------
cr-
o-
c/i
C3
cj cn
i en
a
k.
ai
u a
«*- ~t-t
a m
5-18
-------
Gross costs are assumed to be paid by county governments. The average gross
cost per county is calculated as a percentage of the total revenue for a range
of county sizes. This provides an indication of the difficulty involved in
paying the TCM cost. Table 5-9 displays these values. According to the
Census of Governments (DOC, 1972), average county revenue is estimated at
$150 million for counties of 100,000 to 250,000 people, and $70 million for
all counties.
Some small counties could have considerable problems in paying for TCM without
Federal assistance. If a small county with revenue on the low end of the
range is forced to implement TCM, the cost could consume so much revenue that
it would make continuation of county services practically impossible without
additional revenue. A cost of $11.4 million corresponds to a tax increase of
$174 per person using an average county population of 65,181 people (DOC,
1972). Examination of county population indicates that none of the counties
thought to need TCM have populations of less than 50,000. Therefore, while
there is some potential for problems in affording TCM, it does not appear to
be overly burdensome to any county, especially since TCM projects are funded
largely by the U.S. Department of Transportation's transportation system
management program.
5.3.3.2 TCM Fuel Sayings
Fuel savings accrued under TCM cannot be used to offset the cost of control
measures since no tax revenue can be gained. The savings are received
directly by the motorist. It is assumed that savings are proportional to VMT
traveled. Table 5-10 shows the distribution of TCM fuel savings (at a
$1.20 per gallon fuel cost) using the percent of VMT traveled by income group.
Because of a lack of sufficient data points indicating the VMT per household
by income class (only four points), it is impossible to make a meaningful
estimate of the marginal increase in VMT by income. Without these data, the
total effect of distribution of income as a result of TCM fuel savings cannot
be estimated.
5-19
-------
S
<=>
g-s-
=> in c-4
in m i
ir> o
i r-J
r<< i c-i
i 01
O- CM CM -O C4 I
I
I >->
in in -
-. c»< 10 » ~0 K>
i -o oo
«r o- o «
r^. ^ or* 10 r
in in «o ro CM
« -o CM to i->
PM rr> t^>
i i i i ^
5-20
-------
5.4 CONCLUSION
The results of the economic analysis indicate that no particular segment of
the population is forced to pay a disproportionate amount of the total cost of
mobile source control. Therefore, there are no significant adverse income
distribution impacts.
The incremental nature of ownership of automobiles does indicate, however,
that some individual households may face a disproportionate cost. This is due
to the relatively small increase in numbers of automobiles owned by higher
income households. (In other words, a constant portion of total income is not
used to purchase more cars for all households.) As a result, a cost
proportional to the number of cars may place additional burden on individual
low income households.
5-21
-------
REFERENCES FOR SECTION 5
1. Air Pollution Control Association. June 25, 1979. "Reducing Air
Pollution From Motor Vehicles -- Developments in In-Use Strategy." APCA
Paper No. 79-7.1.
2. U.S. Department of Commerce, Bureau of the Census (U.S. DOC). 1972.
Census of Government:. Vol. 3,; Government Finance.
3. U.S. DOC. 1977. County and City Data .Book.
4. U.S. Department of Transportation. December 1974. National Personal
Transportation^Survey. U.S. DOT Report 11.
5. U.S. Environmental Protection Agency (EPA), Office of Air Quality Planning
and Standards. October 1982. Cost and Economic Assessment of Alternative
National Ambient Air Quality Standards for Carbon Monoxide.
6. The Urban Institute. 1978. "Distressed City Indicator." Cited in the
President's National Urban Policy Report, 1978. Prepared by the
Department of Housing and Urban Development.
5-22
-------
6. ECONOMIC IMPACT ON INDUSTRIAL POINT SOURCES
6.1 ECONOMIC ANALYSIS PROCEDURE
Section 4 presented aggregate capital and annualized costs for each industry
along with the number of plants controlled under each standard. In this
section the average cost per controlled plant to a company within the industry
is estimated by dividing the total control cost by the number of plants
controlling.
The control data are obtained from the previous 1979 analysis of economic
impacts, updated to 1984 dollars. The industry financial statistics,
production and other data with which these costs are compared have been left
as presented in the 1979 analysis. Thus, for example, estimated 1984 capital
costs are compared with capital expenditures in 1979 dollars which are based
on several earlier years. Production statistics are also from non-uniform
years. Thus the estimates of percentage of capital investment needed for CO
control and other indices in this Section are believed to be somewhat
inflated. This results in a conservative assessment of industry impacts.
6.1.1 Data and Data Adjustments
6.1.1.1 Cost Data
The average annualized cost figures for plants requiring control must be
altered to account properly for the effect of taxes. In their present form
the figures represent the sum of operating costs, capital charges, and credits
obtained from the use of CO as a fuel. Operating costs are the day-to-day
expenditures required to operate the system. Capital charges are those items
associated with owning the equipment, including depreciation, property taxes,
insurance, and interest on borrowed capital. A uniform capital recovery factor
6-1
-------
of 13.6 percent was chosen to estimate the effect of depreciation and interest
(PEDCo, 1979). This portion of the capital charge is assumed to represent the
only after-tax expense of owning the equipment. All other portions of the
annualized cost represent before-tax expenses.* The cost of insurance,
property taxes, and operating cost are deductible for tax purposes; thus, the
magnitude of these costs is reduced by 48 percent, the corporate income tax
rate.
Steam credits produced by heat recovery represent an income stream into the
firm and are therefore taxable. This is true regardless of whether the steam
is sold or used to reduce energy expenditures. The credits, therefore, must
also be reduced by 48 percent. The tax-adjusted components of annualized cost
then, were added again to represent the average after-tax cost of regulation
to firms within each industry.
6.1.1.2 Financial Data
In addition to the cost data, financial data collected to construct a profile
of the industries were used to determine the impacts. The data form a set of
"typical" firms in each industry.
Financial characteristics of firms in four major industries -- primary
aluminum, carbon black, gray iron, and maleic anhydride -- were compiled in
order to present a composite picture of the capital structure of each
industry. The primary sources of this information were Moody's Industrial
Reports (1979) and the Value Line Investments Survey (Bernhard, 1979), with
background material on the industry provided by the U.S. Department of
Commerce, 1979 U.S. Industrial Outlook.
Production figures for the companies were used to determine typical size,
while long-term debt, current liabilities, total stockholders' equity, and
1977 or 1978 capital investments were used to measure the financial strength
*Some financial analysts may disagree. For example, the uniform capital
recovery factor could yield a stream of rental payments that is a before-tax
expense item. Moreover, there is no room for investment tax credits and rapid
write-off considerations using this approach.
6-2
-------
of an average company in each industry. Also collected were debt/equity
ratios and beta values; the latter are estimators of risk as measured by the
covariance between stock market activity and company stock prices. These
figures are used by investors to determine whether or not an investment in a
company is wise. As such, they serve to indicate whether it would be
difficult to generate capital.
Where data were available for most of the firms within an industry, the mean
of the values collected was used to represent a typical firm. In other
instances, the data collected were used to establish a range within which a
typical firm would fall. For example, the gray iron industry comprises
1300 separate companies, but data are available for only a few of each size;
therefore, the range of data collected established a range within which any
firm is likely to fall.
6.1.2 Capital Availability Analysis
The following steps were used to determine the ability of a firm to shoulder
the burden of capital expenses incurred by pollution control requirements:
(1) screening for industries significantly affected;
(2) comparing capital required to historical capital expenditures of a
typical company;
(3) comparing capital cost to total long-term debt of a typical company;
(4) examining debt/equity ratios and beta values to determine the ability
of a typical company to raise additional capital.
Screening consisted of examining the capital requirements for 1) the number of
sources requiring control, and 2) the percentage of recent capital expenditure
represented by controls under the most stringent standards. If the number of
sources was limited (e.g., one plant) and capital costs were less than
one percent of recent expenditures, the industry was removed from the study.
(Fluctuations of one percent and less were assumed to be normal.) However, an
industry comprised mostly of small firms was retained in the study since
smaller changes in cost can affect small companies more significantly.
6-3
-------
The percentage of recent capital expenditure represented by the control cost
was then used to determine whether large amounts of additional capital need be
raised or whether other investments would be delayed.
Capital costs were also compared to total long-term debt to indicate the
ability of a company to absorb the expense. Small changes in debt would not
affect the posture of a firm; large changes were examined for the effect onthe
debt/equity ratio. An increase in the debt/equity ratio or an already high
ratio together with a high beta value indicates that raising capital may prove
difficult.
6.1.3 Annualized Cost Impact Analysis
Figure 6-1 presents a diagram of the components of the annualized cost
analysis. A screen for significant impact within the industries was used to
eliminate industries whose price impact was less than 5 percent. Five percent
was chosen for consistency with EPA "Criteria for Conducting Regulatory
Analysis" (44 Federal Register 30988. (May 29, 1979)).
The average cost impact was calculated by dividing the tax-adjusted aggregate
annualized cost for each industry by the number of sources controlled under
each standard. It is assumed that this cost is representative of the cost
faced by a typical firm.
By further assuming full pass-through of cost, the increase in the average
cost of production cost would be offset by increases in price. Accounting for
the effect of taxes (since 48 percent of the marginal revenue goes to
corporate income tax), the price impact is approximately twice the increase in
the average cost of production. The magnitude of the cost impact, viewed in
terms of a percentage price increase, can be identified by comparing the price
increase to a weighted average product price.
The price impact was examined to determine if there would be interference with
the firm's competitive posture within its own industry. Foreign competition,
transportation cost, and availability of sources of substitute products were
considered within this effort.
6-4
-------
03
O
_J
O
O
O
=3 O
C3 O
ul o
LU
M
ScS
01
~
TO
6-5
-------
Because credits may be obtained by incineration and heat recovery, the
annualized cost can be negative; that is, the value of the heat recovered (in
steam) exceeds the annual cost of equipment operation. In these cases, the
pollution control investment has been viewed as an exogenous investment; any
benefits obtained are distributed to the shareholders and not reflected in a
product price decrease. This conservative estimate is consistent with the
economic theory that "the dollar price of goods and services resists being
pushed down" (Solow, 1979). It is assumed, therefore, that corporations
simply would not pass the cost savings on to the consumer.
6.2 ECONOMIC IMPACT TO INDUSTRY
6.2.1 Determination of Industries with Significant Impact
The regulatory impact analysis indicated that seven industries would require
control under any alternative 1 ex ex NAAQS investigated in this report. They
are:
Maleic anhydride
Carbon black
Steelmaking
Gray iron
Primary aluminum
Automobile manufacturing
Incineration.
The latter two industries, at maximum, show only one source controlling even
under the most stringent 1 ex ex standard (i.e., 7 ppm).
The maximum capital expenditure incurred by the automobile manufacturing
industry is $26,000 under any alternative. This amount is insignificant
compared to investments made within the industry. General Motors, for
example, invested $4.6 billion in 1978. Even for the much smaller American
6-6
-------
Motors Corporation, the $26,000 would represent less than 0.5 percent of the
$41,233,000 invested in 1978.* The annualized cost of control to automobile
manufacturing is $90,000. While the figures are not readily available, it
seems safe to assume that the operating budget of a company is not less than
its capital investments. Therefore, even for the small automotive
corporation, the cost increase would be less than 0.02 percent.
The most stringent standard requires control of CO emissions from one
municipal incinerator. Both the capital and the annualized cost of control
are approximately $7,000. Since the incinerator is operated by a State or
local government, the increased expenditure would come from governmental
revenues. The aggregate of local government revenues across the United States
for 1977 was $105 billion. Wyoming, with $170 million, ranked last in
revenues received. Individual counties had revenues ranging from $1.7 million
to $29 million (EEA, 1979). A $7,000 expense represents only 0.41 percent of
the smallest county revenue. Therefore, because of control requirements,
incineration is considered not to have a significant impact.
6.2.2 Profile of Industries Facing Regulatory Impacts
6.2.2.1 Maleic Anhydride
Seven companies produce 117 gigagrams (10^ grams) of maleic anhydride annually
at eight separate plants. The corporations have a mean debt/equity ratio of
31.9 percent and mean beta value of 0.96. For five of the companit 78
capital expenditures data were available. Three of the companies are large
integrated corporations showing capital expenditures in excess of $500 million
per company. Two smaller corporations' expenditures were investments for
chemical production alone, not for the entire corporation, and these were
around $33 million each.
*Telephone interview with staff of Automotive Information Council.
6-7
-------
6.2.2.2 Carbon Black
Carbon black, an organic compound often used as reinforcing compound in rubber
and plastic products, is produced by nine companies; financial data were
available for six companies whose aggregaate production is approximately
4.3 billion pounds per year. The corporations producing carbon black have a
mean debt/equity ratio of 30.5 percent and a mean beta value of 0.97. In
1977, four of these companies invested more than $1.7 billion in capital
expenditures, about $440 million each. The industry is associated with the
petroleum refining industry by ownership and is concentrated in Louisiana,
Texas, and Oklahoma.
6.2.2.3 Steel
The steelmaking industry was profiled using 10 of its largest firms, which
account for almost all of the steel produced in the United States
(97.1 million tons in 1978). The 10 companies have a mean debt/equity ratio
of 31.9 percent, with the median at 34.5 percent. The mean beta value for
steel companies is 0.99, indicating that, historically, return on steel
investments has been almost identical to that of the market as a whole. The
1977 capital expenditures of these companies ranged from a low of
$30.2 million to a high of $865 million, with a mean of $236.5 million.
6.2.2.4 Gray Iron
It is difficult to describe a typical gray iron company because of the vast
numbers and varieties of plants. Gray iron is produced at over 1300 U.S.
foundries, ranging from giant casting foundries owned by auto manufacturers to
small, family owned and operated foundries.
A representative cross section was drawn from the Iron Casting Society's
Source Book. A wide variation was noted in the data collected for
five companies ranging in size from 100 to 9000 tons per month. The
debt/equity ratio ranged from 15 to 80 percent with no correlation to size; in
fact, the two largest companies set the limits to the range.
6-8
-------
Four of the companies were able to provide annualized cost data on pollution
control equipment presently installed to control other nonspecified
pollutants. They range from $7500 to $800,000 per year with the larger values
occurring in the larger firms. This range will be compared with required
expenditure to indicate increased burden.
Total 1977 capital expenditures for all investments ranged from $60,000 to
$24,000,000. The latter figure could be a one-year anomaly since it is far
greater than any of the other firms; $300,000 is a more reasonable upper limit.
6.2.2.5 Aluminum--Primary SmeIt ing
According to the Aluminum Association, primary aluminum is produced by
12 companies in 32 locations across the country. Approximately 5.19 million
short tons of aluminum were produced in 1978. Financial information was
available for 9 companies which account for 75 percent of the total or an
average production rate of 432,800 tons per year per company. The mean
debt/equity ratio is 38 percent, with a range of 23.17 to 54.7 percent and a
median of 36.1 percent. Capital expenditures (1977) could be found for only
three companies with values ranging from $0.9 billion to $1.7 billion.
Financial and production data gathered for the industries discussed in
Section 6.2.2 are shown in Tables 6-1 through 6-5. All of the ratios,
percentages, and price impacts shown in later sections of this chapter are
derived from the data presented in these tables.
6.2.3 Capital Cost Impacts
6.2.3.1 Maleic Anhydride
Two sources in the maleic anhydride industry require control to meet
either 9 ppm eight-hour standard. Compliance with these standards requires a
capital expenditures of $1.59 million, an average of $780,000 for each plant.
This represents less than 0.20 percent of the average 1977 capital investment
6-9
-------
n o
4J -J
m
oc
o
o
m in
cc o
o -i
2
5-
i-i N a
a ..
o a
H .u
C.
a
U
n
r*-
e^
CO
oo
IN
3*
O
O
00
O
o
O
o
o
c>
CM
3 0
cr -H /
A
c
o
CM
<
2
<
2
w
«
H
4) £
a >s^
1 '- °
ji zrv>
u a ^.
o
eg ^ ^
^ *J C!
"l O
00 01
c c «
o »
IS
o o
O CO
CM O
G O
O O
m CM
c
00
o
o
CM
o
<
2
O SS
O C-.
a
c
c
c
o
o
o
< o
u
a.
-H f>
CM PH
SI
O
u
=
c r
CJ
u)
i*. a
o «J
«j <
«j
E =
ij
u 5
r n
o s
a;
- a
u
a
6-10
-------
w
hJ
M
TT .
H-l
§
Pu
hJ
<;
CN M
1 U
vd g
U 2
iJ M
« EH
c_; > ' '
e
U
c
o
p4
-J « rt
to N o
o **
t- « ^
->
*j ^*
--, u
3 0
O> « X-N
Ul -u J<
^ eo ^^
4J QC
J=
QJ
Q
00
u ,
HI U
c >« >
^-1 r^ 4J PI
(0 o -^ O
o 2 £<«
t- 0 td >-^
o
^
co
s >.
u ij /^
aj ) n
f- J= =
1 CJ .
OC C
c >-
Q
-J
^_
u« u
O
c
a
3C i.
J2
>v^
4J
*T^ ~
U
CO C5
CQ
U
r*
*j 1^1
c s
ea
z ^
s
1
1
<
(_
in o in o o o
CO O OO C^
o , i o i-i o
0 0 >? ec
< . <
O 5E O 2 CO r-
< u"> m
v>
^
& en o -d- -^ o c
N tn -* Is- r* o to
*^ \O «^ f*^ >T -J1 M*
**"»"* S
r^ c^ <**) CTN vo oo o
*3- O «N O *T m U
\o ** r*» tn a\ c
* » . * - ai
i * CM >j n i i
4-1
<*4
O
^J
?e S
vC u"> CM CT\ CNJ m U
O "O C\ r«* r*» 4j
xj <») fs) f«i CN m £.
O
^O
{JJ
-K
+j
m in co r*» r- o
en r^ m -- c* c 4J
r^- *o m » n -* *-
CM n
c\ (N a^ o « ^o w
W p-i n CM 3
u
| so co cs ao -^- o «
1 O r^ tn m r- O -S
so os m ^o c o a
C1 .t-^rnf^cNc**) u
\O a\ O -T
00
51
9 3«
iJf-^cMcn^jinvo n
=1 a
3 «>.
jl C5
c
tz
_
0
«
u
t>4
e
J
C
VI
~>
u
Jf
u
re
1*4
Q
C
C
w
ee
i
w
c
co
v. r
« 35
C
u -
w
<
C
W _l
^ ^ ,
c = o\
US r-
.-. cr\
<*. !B -H
C -1
.J
CJ CO
= n
(^ 0 H
e z v.
0
4J S C.
18 -" U
U (X
= to
^ ^a
Ci n
VI *f*
r*- ~J
Cv C W
x -3
-i C
u ^
it r: w
C" >v
< ^ "~
^: o
si < z:
-: u
N
eul
>
w
tH
^
c
CJ
=
^
n
u
>
u
-
i
^
r^.
C>
t
-------
a
£
U
1-3
§
U
Z.
fn
o
z
u
H
OO
83
d
O
E-
5
.H
c.
CJ
*H
(C
4J
O
t-
U 1
= J
cr
.u
0
a
09
01 -
S
o
u
o
w
g
J^
01
i
oc
c
o
<0 0
JJ -J
CQ a
Ch
JJ /-s
cio
o> o
03
01 r-
C ON
e
o
*J /*^
(C«
N O
H ^H
**
« '
-^t
ex
CJ
*
j
a
o
* 2
03 **^
ce
«*
3
^ C
S -^
O" VM
U ^^
«.
3 />-s
-u«
2 C
m
Q v>
>._^
IN O
O> CM
0 1
-3- f-i
NO CM
-31 ro
^^ uo
v>
CO O
-^ o
r-. ON
in ON
ON CO
-H CM
^e
CM CM
r» NO
m ^
CM CO
a- o
r- O
-H 00
m <
>I CM
CM
ON O
~sr c
NO 00
m o
o -3-
in i
o
00
o
o
CM
oo
CM
ON
in
in
ON
in
i"^
i i
CM
ON
^y
CO
0
00
m
UO
^
«>
**
ON
r*^
ON
CO
r-t
NO
o
ON
0
CM
O
m
ON
in
^^
o
«^
**
^H
o
M
CO
NO
rj
ON
-3-
CM
ro
ro
ro
CM
m
oo
in
ON
O
-31
CM
ro
CM
in
-3"
vO
ON
CM
CM
NO
^
CO
^J>
ro
CO
^J.
NO
m
-T
CO
,».
r-.
m
<
^4
CO
r*.
.^
^^
ON
NO
CM
-3-
m
^
o
CM
0
NO
m
o
NO
H
CM
ON
CO
m
ro
i~^
uo
r~-
O
f~-
r*
vO
O
CO
CM
CM
ON
NO
i
O
NO
ro
>y
NO
CO
in
CO
CM
i
NO
*3"
NO
-3
UO
^
in
uo
UO
ff^
CO
CM
NO
00
r*.
r-l
CM
CO
UO
CM
ON
rM
O
>*
CO
ro
CM
UO
cs
CM
UO
^J
uO
0
^
0
UO
00
o
o
CM
r~
m
<_
^
»^
o
o
CM
O
0
Csl
r^
UO
H
>
UO
O
^3
O
O
O
ro
O
CM
^
"*
UO
CM
O
ro
CM
ON
,4
O
UO
ro
(M
CO
o
ro
oo
CM
O
CM
IN
ON
O
CM
B C
= O
O 4-1
(Q U 2 CJ O
u fa c 2 t
o o. s -s
H<3
a o
z
01
-j = col
n *
co o
u «> a.
a u u
ON
ON
o
u
< -s
U a
6-12
-------
nj o
i. u
tL4
.
\
^o
w
1 "]
ff\
<3
H
W
E
§
^f
^J
M
U
^j
2
i i
C/5
w
M
Pi
6
^
o
EH
§
Pi
i i
r^
^^
fyj
O
c
o
i* an
« N O
o »
H * ^
H
a
» ^^
i-^ r^ w t"3
ie o -jo
o 2 c- c
w S
f*^ ~ >C.
(3 ~ w
j-> o r
o i- o
t- s- -»
CO -1
c> =
r-" s«/
2-
m
a
£
0
u
o o a o o
o o ~r z c
CM *G CS O C
k K > K V.
-^ IN n <
s
0
ca
e
o o m o o =
CC O ri -T CO <
U3
i SI
u >
a ^.
S 5
co
o ^*
C O O f~ O =
c a o c-" o = s
o ^ c ^ c o i*
* ** ^ -H *>
« ^^ rn rM ^j v;
« t >
= C
.-
" . Si
COOOC T5 " -^
S 01 ^
O O O O C « 15
o a j C aj 4)
^ < ^-i
f! M H
?M C ^J >
o 5 o<
*J CQ Ch
O -J r-
14- o '^ r
O z T
u c 01
a u u
S J3 2 .5
'? " .si""
^ J2 ^ *
r <*- ^ o
O R3 sJ
O O O O C
* « r*^ 'vrf ys (?
f yi ffN -^ S
* tc -r «
5- c
u
c i-
-------
c
CN
a <->
4 N en
IB O
o a *
~* ON
C
c
tM
e
m
m
ce
c
\C
CM
o
cc
CN
vO
E a
u c
W
iJ
M
P-
O
in
I
w
P3
EH
U
s
H
En
>n m
un
m
03
a
pi
c
10 u
O^wl
o
u <
a
<^l
a I
>> » j| O O C O 3 r~
u i. M|
j'joiin m * CM s M
5 ^ «{ r* r*- r^ ,n c ^-<
i 0
3! '
6-14
-------
for the companies. The cost represents only 24 percent of the smallest 1977
capital expenditure for any of the companies for which financial data were
available.
The mean beta value for the firms producing maleic anhydride is 0.96,
indicating both a slightly less-than-average risk associated with
shareholders' investment and, combined with a debt/equity ratio of 31.9, a
strong financial position. Even if the expenditures were all placed in
long-term debt (i.e., 100 percent debt financed), it would represent only a
0.05 percent increase in debt.
Capital requirements for the maleic anhydride industry are shown in
Table 6-6. The impact against 1977 expenditures (i.e., less than one percent
in all cases), indicates little real problem with capital availability; the
difference between 0.19 and 0.27 percent should not hinder the formation of
capital.
6.2.3.2. Steelmaking
To meet either 9-ppm eight-hour standard, only one Steelmaking plant must
control CO emissions. The capital cost for the single plant is $6.0 million.
The mean value of capital expenditures in 1977 for the steel industry was
$236.5 million.
The control equipment capital cost, therefore, represents an average of
2.5 percent of the capital budget for typical corporations within the
industry. In 1977, capital investments ranged as low as $30.2 million for a
single company. A $6.0 million cost would represent 20.0 percent of the
capital expenditure for that firm. However, assessing capital requirements in
isolation can give misleading results. Chapter 4 shows that, on an annualized
cost basis, CO control requirements result in a net cost savings, or credit.
Despite problems with competition from foreign steel products, the industry
maintains a mean beta value of 0.99 and a relatively low debt/equity ratio of
31.9. This indicates that while expansion of the industry may not occur, its
financial structure is reasonably sound.
6-15
-------
TABLE 6-6
CAPITAL REQUIREMENTS FOR THE AVERAGE MALEIC ANHYDRIDE FIRM
Alternative
8-Hour CO
Standards
7 ppm
9 ppm
12 ppm
15 ppm
Capital
Per Plant
($103)
$ 870
780
1210
780
# of Plants
Controlling
3
2
2
1
% of 1977
Capital
Expenditure
0.20
0.19
0.30
0.19
% Increase in
Long-Term Debt
0.050
0.060
0.080
0.060
6-16
-------
Long-term debt for a typical steel producer is $494 million. Total debt
financing of the capital investment would be only a 1.2 percent increase in
debt. Complete debt financing would raise the debt/equity ratio to 32.3, an
increase of 0.4 percent.
6.2.3.3 Gray Iron
The 9 ppm eight-hour standards require 12 gray iron plants to control CO
emissions at a capital cost of $233,000, an average of $19,400 per plant.
This represents from 0.08 to 32.3 percent of the capital expenditures of
foundries in 1977 (Company 2, Table 6-4). The expenditure represents
7.1 percent of the capital expenditures of our benchmark firm, and is
1.9 percent of the company's long-term debt.
Table 6-7 presents average capital costs for the gray iron industry. The
percentage of 1977 capital expenditures is calculated from the expenditure of
the benchmark firm. In the case of the gray iron industry, the effect on the
individual firm of changing the standard is not very significant. The maximum
difference between the capital requirements under any of the standards is
$3,500. The primary effect of a change is to alter the number of sources
subject to the control requirements. Even so, the maximum number of plants
involved (18) is less than one percent of the foundries in the United States.
6.2.3.4 Primary Aluminum
To meet either 9 ppm eight-hour standard, a capital expenditure of
$43.8 million is required for the control of two plants, an average cost of
almost $22 million per plant. Capital expenditure data for 1977 were
available for only three companies which are of small size compared to the
mean production of the industry. The figures for capital expenditure ranged
from $90 million to $1.6 billion. This means that the capital required to
6-17
-------
TABLE 6-7
CAPITAL REQUIREMENTS FOR THE AVERAGE GRAY IRON FIRM
Alternative
8-Hour CO
Standards
7 ppm
9 ppm
12 ppm
15 ppm
Capital
Per Plant
($103)
$ 21.4
19.4
18.5
22.00
# of Plants
Controlling
18
12
4
1
% of 1977
Capital
Expenditure
7.8
7.1
6.7
8.0
% Increase
Long-Term
2.1
1.9
1.9
2.2
in
Debt
6-18
-------
meet the eight-hour standard should represent between 1.4 and 24.4 percent of
the capital investment in any year for these aluminum-producing companies.
The cost is 3.3 percent of the average capital expenditure.
The mean value for debt is $746.7 million, therefore, the control expenditure
would be 2.9 percent of the total debt for the company. This might delay some
other non-mandatory investments by requiring generation of equity to cover the
expenditure. Capital expenditures are being considered at this stage of
analysis in isolation of cost pass-through possibilities. Both capital
expenditures and cost pass-through, or price impacts, must be considered in
evaluating net impact on the primary aluminum industry.
Table 6-8 shows that average capital requirements for firms controlling for CO
do not vary among standards.
6.2.3.5 Carbon Black
Capital investments for the corporations producing carbon black totaled
between $174.8 million and $584.4 million in 1977, with a mean of
$440 million. The 9 ppm eight-hour standards requires control of 11 sources,
which make up almost 35 percent of all the sources (31) that produce carbon
black. The total capital cost is $32.4 million or $2.95 million per plant.
This is 0.7 percent of the mean expenditure or 1.6 percent of the smallest
1977 capital expenditure of companies within the industry. The industry
should have little difficulty raising the capital for such investments,
because: (1) the debt/equity ratio of the firms is lower than for most
industries (between 19.90 and 41.6); (2) the amount of credits obtained from
investments is substantial, and (3) profits are up among many oil companies
with which carbon black production is connected.
6-19
-------
TABLE 6-8
CAPITAL COST TO THE AVERAGE ALUMINUM FIRM
Alternative
8-Hour CO
Standards
7 ppm
9 ppm
12 ppm
15 ppm
Capital Cost
Per Firm # of Firms
($103) Controlling
$ 22,000
22,000
22,000
22,000
2
2
2
1
7. of 1978
Capital
Expenditure
2.9
2.9
2.9
2.9
7o of
Long-Term Debt
2.1
2.1
2.1
2.1
6-20
-------
Table 6-9 compares the average capital cost for firms needing control under
each alternative CO NAAQS investigated. The low average capital cost
differences result in an insignificant impact on the average firm. The
primary effect would be a change in the number of sources subject to control
requirements.
6.2.4 Annualized Cost Impacts
6.2.4.1 Industries with Positive Annualized Cost of Control
6.2.4.1.1 Primary Aluminum
Table 6-10 presents the after-tax annualized cost for the NAAQS options for
the typical firm within the primary aluminum industry. A price increase of
this magnitude may not have a significant effect on the industry for
two reasons. First, transportation of aluminum costs approximately $0.014 per
pound per ton-mile (EEA, 1979). Therefore, an increase in transportation
distance of an alternative source of aluminum of 150 miles would completely
offset the price increase. Sacond, the aluminum industry is currently
experiencing increases in demand. Operating at 90 percent of full capacity
(Bureau of Mines, 1978), sources within the industry are in a strong position
to fully pass through costs without causing a significant shift in competitive
posture.
6.2.4.1.2 Gray Iron
Table 6-11 compares the average after-tax annualized cost for gray iron
foundries. As can be seen, none of the options under consideration cause an
increase of greater than one percent in product price for a foundry requiring
control.
The after-tax annualized cost of the 9 ppm standard is $18,900 per plant,
corresponding to an increase in the average cost of production of $0.001 to
$0.008 per pound, depending on the size of the foundry. The increase in cost
to our benchmark firm (Company 2, Table 6-4) would be $0.001, reflecting a
0.7 percent increase in price.
6-21
-------
TABLE 6-9
CAPITAL COST TO THE AVERAGE CARBON BLACK FIRM
Alternative
8-Hour CO
Standards
7 ppm
9 ppm
12 ppm
15 ppm
Capital Cost
Per Plant # of Plants
($103) Controlling
$ 3,040
2,950
3,100
3,100
12
11
6
4
% of 1978
Capital
Expenditure
0.69
0.67
0.70
0.70
7o of
Lpng-Te rm Deb t
0.48
0.46
0.49
0.49
6-22
-------
TABLE 6-10
ANNUALIZED COST FOR AN AVERAGE ALUMINUM FIRM
Alternative
8-Hour CO
Standards
7 ppm
9 ppm
12 ppm
15 ppm
After-Tax
Annualized Cost
($103)
$ 12,680
12,680
12,680
12,680
# of Sources
Controlling
2
2
2
1
7o Increase in Price*
(
-------
TABLE 6-11
ANNUALIZED COST FOR AN AVERAGE GRAY IRON FIRM
Alternative
8-Hour CO
Standards
7 ppm
9 ppm
12 ppm
15 ppm
After-Tax
Annualized Cost
($103)
$ 17.700
18.900
15.633
20.700
# of Sources
Controlling
18
12
4
1
% Increase in Price*
(@ $0.62/lb)
0.78
0.82
0.69
0.91
*Calculated by dividing before-tax annualized cost by annual revenue.
6-24
-------
6.2.4.2 Industries with Negative Annualized Cost of Control
6.2.4.2.1 Maleic Anhydride
Two sources in the maleic anhydride industry must control CO to meet a 9 ppm
eight-hour NAAQS. The after-tax annualized cost is $24,000 per plant, which
is assumed to be distributed among the stockholders. However, since the mean
stockholders' equity for a company producing maleic anhydride is approximately
$3.11 billion, this is only a 0.0008 percent return to the stockholders. The
negative annualized cost does indicate that this is an investment which would
not change the ability of a company requiring control to operate within the
market.
All of the NAAQS options result in either no cost or negative annualized cost
to the maleic anhydride industry. The magnitude of these after-tax costs are
presented in Table 6-12.
The annualized credit associated with the investment for the 7 ppm standard is
$39,600 per firm, adjusted for taxes. This negative cost is an insignificant
additional dividend to stockholders (approximately 0.0012 percent), but a
savings to the average firm of $415,600 per year over the current standard.
6.2.4.2.2 Steelmaking
The 9 ppm eight-hour standards require controls for one Steelmaking facility.
The after-tax annualized cost was calculated to be $55,000 per year per plant,
representing a $0.004 payment for each dollar of stockholders' equity.
Table 6-13 compares the negative annualized costs incurred under the various
alternatives. It should be noted that the steel industry has the largest
fluctuation of average impact of any of the industries affected. The level of
control required changes the profitability of the pollution control investment.
6-25
-------
TABLE 6-12
AVERAGE AFTER-TAX RETURN FROM CONTROL OF CO FOR AN
AVERAGE MALEIC ANHYDRIDE FIRM
Alternative
8-Hour CO
Standards
7 ppm
9 ppm
12 ppm
15 ppm
Annualized Return
($103)
$ 39.6
24.0
180.2
110.0
# of Sources
Controlling
3
2
2
1
Return on
Stockholders' Equity
0.0012
0.0008
0.0058
0.0035
6-26
-------
TABLE 6-13
AVERAGE ANNUALIZED RETURN PER FIRM IN THE STEELMAKING INDUSTRY
Alternative Annual Return
8-Hour CO per plant # of Sources «f/$ Return on
Standards ($10^) Controlling Stockholders' Equity
7 ppm $ 1100 2 0.080
9 ppm 55 1 0.004
12 ppm 550 1 0.040
15 ppm 0
6-27
-------
6.2.4.2.3 Carbon Black
The after-tax annualized credit under the current 9/1 eight-hour standard is
$655,000 per plant. The savings translates into a $0.040 return on each
dollar of equity. Table 6-14 presents the annualized return for alternative
1 ex ex CO NAAQS options under consideration.
The after-tax annualized savings due to the most stringent standard (7 ppm
eight-hour) is $690,000 per plant. This is only $0.002 per dollar of equity
increase over the 9/1 standard, while the 12 ppm standard yields an after-tax
annualized credit for each plant of $647,000, or a reduction in benefits of
$8,000 per year. Again, this change does not affect any operations of these
large corporations.
Since the impact on the operation of the firm is not affected by an increase
in dividends of less than one percent, the differential impact of the
regulatory options is not significant.
6.3 IMPLICATIONS OF CO NAAQS CONTROL COSTS FOR SMALL BUSINESSES
The preceding material was prepared prior to Congress's enactment of the
Regulatory Flexility Act, requiring that regulatory agencies consider the
economic implications of major federal actions affecting small business
enterprises. This section briefly discusses both the likelihood and degree of
small business impacts.
The probable sources of impacts to small business include two of the
four control programs that appear in Table 1 and elsewhere. These are
Inspection/Maintenance Programs (see Chapter 5); and
Stationary Source Control Programs.
6-28
-------
TABLE 6-14
AVERAGE ANNUALIZED RETURN PER FIRM IN THE CARBON BLACK INDUSTRY
Alternative Annual Return
8-Hour CO per plant # of Sources ^/$ Return on
Standards ($103) Controlling Stockholders' Equity
7 ppm $ 690 12 0.042
9 ppm 655 11 0.040
12 ppm 647 6 0.040
15 ppm 645 4 0.039
6-29
-------
Neither the FMVCP nor the TCM programs should result in small business
effects. The TCM should not affect small businesses because the costs of the
program fall entirely to government. Impacts associated with FMVCP
requirements fall to the automobile manufacturers, who must retool to
accommodate the performance standards. None of the U.S. automobile
manufacturers are classifiable as small businesses by SBA employment size
standards (1000 employees for SIC 3711 Motor Vehicle and Passenger Car
Bodies; 500 employees for SIC 3714 -- Motor Vehicle Parts and Accessories).
In 1979, these two industries together employed 807,500 persons, of which over
80 percent are production workers. Considering the number of domestic
manufacturers (i.e, CMC, Ford, AMC, Chrysler, several specialty manufacturers,
and domestic operations of foreign manufacturers), no small businesses operate
in this industry. Therefore, no impacts fall to this class of business
enterprise.
Additionally, the analysis assumes that FMVCP costs are passed on to the
purchasers of cars, trucks, motorcycles, etc. It is conceivable that the
incremental cost of CO control might affect small enterprises that purchase
motor vehicles for their businesses. However, because the CO part of the
control program adds only about $180 to the price of a. vehicle that may cost
over $12,000, direct impact to the purchaser should be negligible.
The I/M program could affect small businesses in that a decentralized, gas
station-operated inspection program would impact gasoline retailers. However,
I/M administration costs are borne by local governments and costs are largely
recovered through an inspection fee. Moreover, because inspection failures
will produce increased demand for repair services, and higher receipts at
repair centers, any negative impact is likely to be mitigated, if not
overtaken, by increased earnings.
Stationary source CO controls can affect small businesses if the industries
that must control are comprised predominantly of small concerns. Tables 4-1
6-30
-------
and 4-2 present capital and annualized costs of stationary source CO control,
respectively. The tables show that, nationwide, only 29 sources in
six industries will impose control under a 9/1 CO standard. Three of these
industries, which account for 13 sources, experience negative costs, i.e.,
control of CO generates savings because of steam credits, resulting in
positive economic impacts. The industries showing positive costs and the
number of sources controlled in each include: steelmaking (1), primary
aluminum (2), municipal incineration (1), and gray iron foundries (12).
Of this group, iron represents the only candidate for small business effects.
Gray iron companies are classified as "small" by SBA if they have less than
500 employees. Ownership of gray iron foundries ranges from automobile
manufacturing firms to small family owned and operated firms. National total
employment in this industry (SIC 3321) is 152,100, and average employment per
foundry (although not necessarily per firm) is slightly less than 100 persons,
indicating a potential small business impact. The CO NAAQS results in control
of only 12 foundries, however, which constitute less than one percent of the
total population of gray iron foundaries. Those needing control in this
hypothetical analysis are the 12 highest CO emitting domestic foundries. They
represent hourly throughputs in the range of 25-96 tons of charge, which
equals approximately 20-79 tons of gray iron produced. The average facility
in the U.S. produces about 5 tons per hour. The facilities modeled for this
cost and economic analysis used 20 tons per hour as the average size for
larger production facilities and 5 tons per hour as the average size for
smaller production facilities. Hence, the sources needing controls to attain
a 9/1 NAAQS represent large facilities and are probably not small businesses.
Additionally, as shown in Table 6-15, even for smaller firms, CO control costs
do not impose a significant operating cost burden, although very small firms
might face capital availability constraints. Given the size of controlled
facilities, it is unlikely that smaller firms are those controlled by the
standard. It is therefore possible to conclude that the stationary source
control program does not constitute an economic burden for small business
concerns.
6-31
-------
IT)
r-l
1
\^
£s3
rJ
«
<
BE
o o
w cc
C1^ C- * *
O G
C*3 ^C
rn ry-i fv
^ ^ C3
E- J
2 < CSS
W > O
cj r^t
Q£ Q
W 2 6-
0- < 2 /-N
U r-l
< cn s CM
cn eg cn cn
H *J CJ
cn i i es
E- C < cn
cn z r- '
O w cn
CJ &4 U
X
nJ rsl [j ^
o c
P£ wJ
H < W
Z H Ni
O M ^-,
cj a. cn
<
CJ >«
cc
01 4J
4J 00 C
W (3 "O 4)
0 4-1 CO E
CJ C T3 JS
X Oi "C CO
is TJ o <: -H
E- 01 U r-l
1 N 01 01 ^2
V4 -H CU 3 13
01 rH i 1 4-1
4J (3 13 13 CO
< 5 CO
C C3 M-i l-i
*>N^
00 Cn r-l i 1 ,£) J3
d d d d
13
4-1
C
0)
0 E
01 -C
o w
T3 *H /N
< i cn
Xi O
01 CO r-l
3 4-1
t4
Ol
cu
O OO CM \D O ON
»* r~ o cn m in
ON CN *«J CM oo cn
r~ cn CM cn r-i vo
M ** r #. *» «s
r-l sd- r-l O CN O
r-l CM ID VO
1 1
M
OJ
^j 1«4 ^J
ca o 3 e
13 4J 01
O) *r4 £
4-> oo tj x:
CO 13 3 CO
O 4-> 01 -r-l
cj a a. r-i
0) X A
r-l U W «0
13 U 4-1
4J 01 r-l CO
n CU 13 Cd
a. 4->
!S C3 -i-l U
CJ Cu 01
13 CU
CJ
00 O ^ \O CM i 1
C5 l^ rH r-I 0 0
r 1
^^,
13
CO
01
>* J-J
3 3
4.) 01
n E
T3 J=
C ca
01 -H ^^
D-r
r- CO - '
13 W
4J
1-1 1-1
a. oi
13 CU
CJ
P-- CM cn in o r4
14 in vo ^^ r*^ CM
cn as vc o m in
i4 CM OO ON »^ ON
*»
ON *^
( i
01
N f-\
H CO
cn 01
0)
4-1 >v
C O
1 "9-
~w oi
H
r-l U-4
X) O
13
4_j ^*5
CO ^^
tu
ON
CN 0)
ON ON ON ^^ ^4
CN ^T CN ON ^ O
ON CM "»^ ON CM E
O O O O O U
4-» *J 4-J 4-1 4-1 O
0 O C 0 O 0
m o in o o o
ti CM in o in
M A
r-4 CM
.
p2
cn
cn /-s
^^N
CM CO
CM 01
v« ' 't-1
CM 4-1
1 "O
CJ 0
s: E
CO O
01 U
rl
tJ i 1
01 IS
cn «
V-
>> 4-1
l-i IO
4-1 3
CO TJ
3 C
T3 4
C
r-i l-l
O
U
01 X
U 01
3 ">!J
4J e
U -H
13
U- OJ
3 U
3 1
13 )-l
2^ p ,
IM ^«
O 0)
U
tn 3
3 ! *^
CO C
C| U
01 C.
CJ
01
CM .3
r- ^
ON
r- 1 OO
3
^ *r*
CO CO
j J
ca
C ON
0) r~
CJ ON
4-J i4
C 01
OI .3 O
U 4-1 4-1
OI I-
00 ^ ^^ ^3
03 C- O 01
l-i u
01 r* 3 M
> 13 3
ts Ol '-n
3 kJ -3
r-l 13 3 <
13 JS 03 ^
C -^
O
H 10
4-1 tfl U
13 01 CJ
2! rJ P^
^^
^. "^ O
ts xi cn
6-32
-------
APPENDIX A: FEDERAL MOTOR VEHICLE EMISSION CONTROL
PROGRAM (FMVCP)
The costs of the FMVCP are estimated for the following vehicle categories:
Light-duty vehicles
Light-duty trucks
Heavy-duty vehicles
Motorcycles
The primary sources of cost estimates for the vehicle emission standards are
an analysis of FMVCP costs for hydrocarbon control by the EPA Office of Mobile
Sources (Gray, 1983), and the previous regulatory analysis for ozone (EPA,
1979). All costs are updated to 1984 dollars.
The FMVCP controls gaseous emissions from combustion and evaporative emissions
from fuel storage. Evaporative emissions consist of hydrocarbons only, and
therefore are not included in this discussion. FMVCP costs are presented by
initial cost of emission controls, annualized capital charges, maintenance
costs, increased costs of unleaded fuel, and costs for control at altitude.
In previous analyses, fuel economy improvements in later model year vehicles
have been credited to the FMVCP, resulting in substantial credits. These fuel
economy improvements are the result of technical changes brought about by
Federal Fuel Economy Standards, fuel prices, consumer demand for more
efficient vehicles and other factors as well as the FMVCP. Thus it is not
possible to isolate what portion the benefits of fuel economy improvements, if
any, should be allocated to the FMVCP, and no credit is taken in this analysis.
A-l
-------
A.I LIGHT-DUTY VEHICLES
A.1.1 Initial Cost of Emission Control Systems
Effective standards of control for automobile carbon monoxide emissions began
with the 1972 model year; the CO standard, effective through 1974, was
39.0 grams/mile. The standard was achieved using an air pump, which added an
initial cost per vehicle of $41 (EPA, 1979; Automotive. News, 1979). The
control system also reduced hydrocarbon emissions; the initial cost was
apportioned equally between these pollutants, making the cost attributable to
CO approximately $21 per vehicle.
In the 1975-79 model years, an oxidation catalyst (also called a dual
catalyst) was used to achieve a standard of 15 grams/mile. The initial cost
of the oxidation catalyst is $240 per vehicle (EPA, 1978a). Of this cost,
approximately $16 was attributable to an exhaust gas recirculation (EGR)
system used for NOX reduction only. Therefore, $224 per vehicle was spent to
reduce CO and hydrocarbons; again, the CO share was half, or $112 per vehicle.
The 1980 model year required emissions of not more than 7.0 grams/mile; such a
reduction was achieved using a three-way catalyst at an initial cost of $338
per vehicle (EPA, I978a). Removing the EGR cost ($16) and allocating a share
to hydrocarbons and N02, the CO portion was approximately $124.
The emission standard again was lowered, beginning with the 1981 model year,
to 3.4 grams/mile. The typical control technology is a three-way catalyst
augmented with an oxidation cleanup catalyst. An EPA study of control
technology in place in 1981 through 1983 estimated that the portion of total
control system cost attributable to CO control was about $133 in 1981-1982 and
about $145 for 1983 (Gray, 1983). The cost for 1983 was assumed to be static
through the remaining projection years.
A-2
-------
Initial capital charges for the FMVCP for CO in 1987, 1990 and 1995 are
estimated by multiplying the model year cost per vehicle by the projected or
actual new car sales. These products are summed to obtain the cumulative
cost. Table A-l. presents these estimates for a FMVCP scenario that keeps the
3.4 gpm equipment beyond model year 1981 automobiles and light duty trucks.
Table A-l also presents the annualized nationwide capital costs of FMVCP for
the projection years. These costs were calculated by multiplying the total
capital expenditure for each model year by (1) a capital recovery factor of
16.3 percent (assuming the discount rate for household is 10.0 percent and
that cars have an average life of 10 years)5 and (2) the fraction of each
model year fleet in operation at the end of the calendar year.
A.1.2 Maintenance Costs
The use of certain emission control systems leads to different costs and/or
savings on typical maintenance relative to uncontrolled vehicles. As shown in
Table A-2, cars equipped with oxidation or three-way catalysts are assumed to
realize a net annual savings of $88 in ten years, or $8.80 per year. The
items in Table A-2 indicate expected changes in the needed maintenance between
controlled vehicles and 1975 model year vehicles. The costs presented are
considered to be a case of minimum savings since the prices assumed for
sparkplug and point/condenser changes are probably low. The increase in oil
change intervals, which at least in part is due to the use of unleaded fuel,
is not considered. Also, 02 sensors are required only for vehicles with
feedback fuel systems, and such vehicles comprised less than 75 percent of
gas-fueled LDV's through 1983. Thus the entire cost for 02 sensor replacement
is conservative. The $65 miscellaneous cost for repair of emission control
systems is in addition to any repairs attributable to inspection and
maintenance programs.
No maintenance charges are experienced in pre-1975 vehicles. Table A-3
presents the estimated future maintenance savings calculated by multiplying
the savings per vehicle by the future year population for each model year
A-3
-------
a;
00
a-c ai
u o
T5 «>
CO
ej\
3 CO
e i^
C *H
to
^
CO
o
a
oo
«l C
> 1-1
>u 00 O
o c o*
K ON
« c -*
C -4
5 I
u C) P*
u ac co
« Ov
Ul -H
b.
OOOOGOOOOOOOOO = C
coc
cs
-3-
CO
w
pa
^
H
fn
W
t/0
<^
W
OS
u
ss
1 j
H
o
u
W
c
CO
4J
W
c
o
o
c
M
v^x
*- Q
C. '
O h
u a
c u
l-l O
c
O 10 ^
U-D
3 u O
O 01 I
Z -^ ^
00
111
Cfi
CO oo co CO CO CO co co CO
A-4
-------
TABLE A-2
MAINTENANCE CHANGES OVER 100,000 MILES
(In constant 1984 dollars)
Total Cost of Cost Related to
Maintenance Maintenance CO Control
Change 02 sensor twice 2 x $55 = $110 1/3 x 110 = $37
Miscellaneous emissions system $65 1/3 x 65 = $22
repairs
Save five plug changes 5 x $13 - $(-65) 1/2 x (-65) - $(-32)
Save 10 point/condenser
changes 10 x $13 = $(-130) 1/2 x (-130) = $(-65)
Save one muffler change $(-150) 1/3 x (-150) = $(-50)
TOTAL $ (-88) per 10 yrs
SOURCE: Gray, 1983.
A-5
-------
TABLE A-3
ANNUAL OPERATING AND MAINTENANCE SAVINGS FOR LDV's
DUE TO FMVCP CO CONTROLS
(In constant 1984 dollars)
Annual 0 & M
Year Savings ($)
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
198A
1983
1982
1981
1980
1979
1978
1977
1976
1975
1974
1973
1972
1971
8.8
8.8
8.8
8.8
8.8
8.8
8.8
8.8
8.8
8.8
8.8
8.8
8.8
8.8
8.8
8.8
8.8
8.8
8.8
8.8
8.8
0.0
0.0
0.0
0.0
Vehicle
Population ($10°)
1987 1990 1995
Total Annual
Savings ($10°)
1987 1990 1995
9.02
10.27
9.88
9.49
8.19
6.86
7.06
6.78
7.07
6.21
4.91
3.33
2.16
1.65
1.60
1.26
0.98
9.68
11.04
10.64
10.34
9.84
9.15
8.48
6.85
5.30
4.79
3.96
3.54
2.83
2.07
1.37
1.00
0.87
9.92
11.52
11.40
11.19
10.95
10.38
9.66
8.62
7.48
6.10
4.68
3.43
2.31
1.48
1.18
1.06
1.04
79.38
90.38
86.94
83.51
72.07
60.37
62.13
59.66
62.22
54.65
43.21
29.30
19.01
0.00
0.00
0.00
0.00
85.18
97.15
93.63
90.99
86.59
80.52
74.62
60.28
46.64
42.15
34.85
31.15
24.90
18.22
12.06
8.30
0.00
87.30
101.38
100.32
98.47
96.36
91.34
85.01
75.86
65.82
53.68
41.18
30.18
20.33
13.02
10.38
9.33
9.15
96.72 101.75 112.40
802.82 887.74
989.12
A-6
-------
(i.e., new car sales for each model year times the proportion of remaining
vehicles).
A-1.3 Use of Unleaded Fuel
Catalyst control systems require the use of unleaded fuel, which is more
expensive than regular gasoline. This difference is partially due to a
slightly higher refining cost of unleaded fuel and partially because of
pricing strategies of gasoline retailers. If the cost differential between
regular and unleaded gasolines is heavily influenced by production volume,
leaded gasoline may become more expensive than unleaded in the future due to
projected reduction in demand for leaded gasoline. However, in the interest
of keeping the calculations in this document conservative, a constant future
differential of $0.05 per gallon was assumed. (Gray, 1983)
Table A-4 presents estimates of the increased cost of unleaded fuel for LDVs
that can be allocated to the FMVCP for CO. Available estimates of future
total LDV unleaded fuel use and the $0.05 per gallon differential allow
estimation of total added cost of unleaded fuel use for LDV. Added fuel costs
must be divided between HC and CO control for pre-1981 vehicles, and between
HC, CO and NOX for 1981 and later. The fractions due to the CO FMVCP cited in
Table A-4 are based on available EPA estimates of this distribution.
A.1.4 Emission Control at Altitude
The high altitude requirements prior to 1984 will not be addressed here, since
they resulted in a cost of consumer of much less than $1 per car (U.S. EPA,
1983). Almost every 1984 and later model year LDV must be equipped with
self-actuating emission control hardware to achieve standards at both low and
high altitude. This program is expected to add between $13 and $16 to the
purchase price of the average vehicle (FR, 1983). The analysis in Table A-5,
based on $16 per vehicle, provides a conservative estimate of the annualized
cost of this additional hardware.
A-7
-------
TABLE A-4.
COSTS OF UNLEADED FUEL USE DUE TO FMVCP FOR LDT
(In constant 1984 dollars)
Year
1987
1990
1995
LDT Unleaded
Fuel Use
(109 gallons)3
49.9
50.2
51.1
Unleaded
Fuel Price
Differential
($/gallons)b
$ 0.05
$ 0.05
$ 0.05
Total Added
Cost of HDV
Unleaded Fuel
for LDT ($109)
2.50
2.51
2.56
Fraction
due to CO
FMVCPb
0.43
0.40
0.36
CO Portion
of Added
Unleaded Fuel
Cost ($109)
1.03
1.00
0.92
aEEA, 1984
bGray, 1983
A-8
-------
1
s
fa
o
H
Ed
>-~l
P
CO
£>
p
fy*
C
w
<
§
0
z
M
H
0
O
. "j
^
§
U
p
Ed
N]
M
O
0
H ^x O
00
**^ C
O -rt
C
C -H
o «s
4^>r:
w O
&. > <
u S
eg co
u 0
H
^
CB
O U
U eg
CJ
eg *
-3- CTv-s-eNLnwroco
coosO'iriO'-^^-^
mot^e^tor^uieNOv
CN r»» r^ o\
in co r^ ^c
(N CN CM CN
0
eN
eot^et-ovoveocor-vo
oooooocoo
eN kO m -^
co cr> o^ o»
o o o o
CM
CO
o
^DOO-^^COOM©
mcNCNOOCOO-CfvvD
C^e>ocy\c^eiocor^'r^
XSvOMS^^sOMS^vO
^OOej>ONOOmtNO
CNeNCN-.-1-l -H ^
ON ON ON ON ON ON CO CO CO
222222222
ON ON ^ O CT*
^ CO *" ^~<
m en CM o
CM CM fM CO
MD *^ rt f*^
r-. r^ r*. r*
CM CM CM O
m < «n
O O O
CM CO ^~
0^ co co
o o o
* in ^j-
ON ON ON
C3 O O
CM ^ \£> \O
^H so -^ m
ps. \o *^ CO
CM
\o ^ ^o
r- sj- ^H
000
CO CO OO
ON ON ON
A-9
-------
A.2 LIGHT-DUTY TRUCKS
A.2.1 Initial and Annualized Cost of Emission Control Systems
Table A-6 presents estimated initial and annualized capital costs for FMVCP
controls for light-duty trucks (LDT). The initial individual vehicle costs of
emission control systems for (LDT) were assumed to be the same as for
automobiles from 1972 to 1980. The standard schedule is slightly different
from that for LDV's in the 1972-82 period, however, since the most stringent
standards are implemented in 1983 instead of 1981 and the interim 7.0 gpm
FMVCP standard applies for four years while it is only one year for LDV's.
The FMVCP applied only to trucks of 6,000 pounds or less until 1979, when
coverage was extended include trucks up to 8,500 pounds. This change is
reflected in the truck populations in Table A-6.
A.2.2 Maintenance Costs
For LDT operating for 12 years, it has been estimated that the FMVCP will
reduce the number of spark plug replacements from eight to four at a savings
of $30, and reduce exhaust system replacements from two to one, saving $167.
Allocating half of the spark plug saving and a third of the exhaust system
savings to CO, the average annual maintenance savings per LDT is about $5.90.
Table A-7 applies this savings to projected LDT fleets in 1987, 1990 and 1995.
A.2.3 Use of Unleaded Fuel
Table A-8 presents estimates of the increased costs of unleaded fuel for LDT's
that can be allocated to the FMVCP for CO, using the methodology described for
LDV in A.1.4, above. Since only a small fraction of recent LDT's use
three-way catalyst systems, it is assumed that half of the total added cost of
unleaded fuel for LDT should be allocated to CO.
A-10
-------
ta
hJ
ta
en
ox
cue
a o
u -* o
« ox
o ^ ox
«
M 01
H 00
ec a
3 j:
cor-
C co
< OX
09
O
IN
CO
H
s
PX
O
Fb
CO
Ed H-3
W O
< K
W H
cd &
u o
2 U
M
O
H 0
Crt
O PU
o u
^J s
j
<
g
^
/^
w
(-4
CO
tl
r-l
O
TD
OO
Cjs
r-4
4-1
CO
] i
0)
C
o
o
c
1 1
s^^
in
ox
OS
Cfi
Iju
^ c
ON
0 C ^
c c
O -4
H Q
u E
OJ OS CO
) 4
CO'-s
- o
O.1
u
01 tB
u O
o u
H
""^
AJ^2
111
u
ca V
fH Cfi
U CO
-J V
e t-
M o
c
1 (
HO
^^
CO
S a
V V
Ui
fl)
4)
»
SSSSSS^SSKSS^SSSSS
ooooocoooooooooooo
^*r**nQDPncof^cori*>^int «p**f^O'«Oc*400
oo^avcocor^r*.vOvciAu^iA'j'^»a'nfnes
oooooooooooocooooo
<*r«»f">CQr'>cOf"!OOP1^^Ln'-«r"i»r">O
coCTsa^aoeor^r-vOsOini.nm-^'^^t
coooocoooocoooo
S3§S§iS§g§§?SS§»?SSSSSS 5
CMCM « ^osooxvor^insfi^ocopsiesr^n^in^ao c*
eoS§rSlCr'2f^S!a23irtevje^»-)eNj(n"-22 Irt
.
O
OOOOOOOOOOOOO^or^.eMCNrMeNPsicsj^^»-H
co fs^m^eM^o>QOsOu^ocof^oooii'^-tfj«r-.--r*jm
NOu^^-fOcsj^avaDr^vOiAON^-^r^^r-.- ^-cNvOOO
^^^^osoN^ON^^^^^^^^^cnosoN^^^
-3-
CO
i
e
o
"O
CO
CO
/v
in
ox
(4
i
CO
OX
r-<
^^
CO
c
o
l_»
u
01
< >
0
a.
00
OX
f^
h3
bU
E
O
^v
o
CO
OX
1
p*.
OX
»-4
^
CO
4-1
CO
13
CO
u
u
0
4J
a.
u
£
o
4J
o
o
^
^
3
a
4J
O
C
41
41
3
_g
^4
3
>
CO
c.
o
o
o
A
SO
U
41
0
5
CJ
3
U
4J
01
Q
3
*H
U
C
4J
o
c
o
a
OS
p^
CT\
^*
41
W
O
a
"°
ce
0)
u
3
00
c-1
CO
er>
>
a
b
O
E
O
w
1.1
4J
a
o
u
^
V
CO
^
c
-1
~
5
tc
c;
01
£
a
CO
01
o
o
E
3
CO
CQ
a
0
CO
OX
7"
r»
^^
E
O
^
UJ
4J
CN
00
OX
1-4
»
a
C
o
*4
4^
CO
U
c
3
E
0
U
CD
^
W
3
O
o
a
to
^
s
E
CO
3
1
a
CO
L4
CO
01
4->
*
U.
d
CO
^
*
A-ll
-------
TABLE A-7
ANNUAL OPERATING AND MAINTENANCE SAVINGS FOR LOT'S
DUE TO FMVCP CO CONTROLS
(In constant 1984 dollars)
/
Year £
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
1979
1978
1977
1976
1975
1974
1973
innual O&JI
iavings ($)
5.90
5.90
5.90
5.90
5.90
5.90
5.90
5.90
5.90
5.90
5.90
5.90
5.90
5.90
5.90
5.90
5.90
5.90
5.90
5.90
5.90
0.00
0.00
Vehicle Population (106) Total Annual Savings ($106)
1987 1990 1995 1987 1990 1995
3.16
3.54
3.28
2.62
2.02
1.55
1.24
1.35
1.76
0.67
0.65
0.75
0.57
0.70
0.74
3.45
3.87
3.61
3.31
3.03
2.75
2.18
1.65
1.25
1.00
1.09
1.42
0.54
0.51
0.59
0.44
0.52
0.52
3.93
4.43
4.14
3.82
3.50
3.21
2.91
2.64
2.37
2.15
1.94
1.52
1.14
0.86
0.68
0.72
0.39
0.32
18.64
20.89
19.35
15.46
11.92
9.15
7.32
7.97
10.38
3.95
3.84
4.43
3.36
0.00
0.00
20.36
22.83
21.30
19.53
17.88
16.23
12.86
9.74
7.38
5.90
6.43
8.38
3.19
3.01
3.48
2.60
0.00
0.00
23.19
26.14
24.43
22.54
20.65
18.94
17.17
15.58
13.98
12.69
11.45
8.97
6.73
5.07
4.01
4.25
5.25
1.89
24.6
31.73
41.17
136.64 181.07 242.90
A-12
-------
TABLE A-8.
COSTS OF UNLEADED FUEL USE DUE TO FMVCP FOR LDT
(In constant 1984 dollars)
Year
1987
1990
1995
LDT Unleaded
Fuel Use
(109 gallons)3
13.5
15.8
18.9
Unleaded
Fuel Price
Differential
($/gallons)b
0.05
0.05
0.05
Total Added
Cost of HDV
Unleaded Fuel
for ($109)
0.68
0.70
0.95
Fraction
due to CO
FMVCPb
0.50
0.50
0.50
CO Portion
of Added
Unleaded Fuel
Cost ($109)
0.34
0.40
0.48
aEEA
bGray, 1983
A-13
-------
A.2.4 Emission Control at Altitude
Average capital cost for additional gasoline-fueled LDT emission controls for
high altitude is about $20, and annual production of such trucks is about
80,000 (Gray, 1983). This is divided equally between CO, HC, and NOX, with
the results shown in Table A-9.
A.3 HEAVY-DUTY VEHICLES
A.3.1 Initial Cost of Emission Control Systems
The cost to the consumer, beginning with the 1985 model year, of the HC and CO
emission standards for Heavy-Duty Diesel Engines (HDDE) is estimated to be
$230 for each engine. Assuming annual sales of 300,000 engines, the cost
attributable to CO control is 69 million dollars annually. (Gray, 1983)
For Heavy-Duty Gasoline Engines (HDGE), the cost to the consumer of the HC and
CO emission standards for the 1985 and 1986 model years will average about
$125 per engine, in addition to the estimated $185 cost of meeting the 1979
model year standards for controlling HC, CO, and NOX. Assuming an annual
production of 340,000 HDGEs, the annual cost attributable to CO control is
about $21 million through 1984, and about $42 million in 1985 and 1986.
(Gray, 1983)
The HC and CO emission standards for the 1987 and later model years for HDGE
with a gross vehicle weight (GVW) between 8,500 and 14,000 pounds will become
more stringent, requiring catalysts on most of those trucks. This is expected
to result in an increase (above the 1986 model year) of $155 per truck. Sales
of HDGE with GVW between 8,500 and 14,000 pounds are expected to average about
264,000 units annually and sales of those with GVW over 14,000 pounds is
expected to average about 96,000 units annually. Based on these estimates,
the total cost for model years 1987 and later is anticipated to be about
$67 million annually (Gray, 1983). Combining this with the above costs for
previously-required controls, total annual costs of CO control for HDDE and
A-14
-------
O"\
1
pt3
j 1
PQ
H
O
H
W
Q
03
-
O
s
PS S /-s
OH «
ft, i i M
H cfl
W ,-J rH
CD < ii
< O
w sa T3
PS O
U M
u ^-^
-4 JJ
03 tfl
w 0
o u
H
*"*
%>^
u (9
CO J£
O U
CJ 3
rt H
cef ^-^
rt 0)
H (0
S 01
h- U
U
B
r-cor~r~r~NONOinin
ooooooooo
oocoooooo
f^ CO f^ r^
O O O O
0000
0
o
oSSN.££S££3S
ooooooooo
«tf r* n oo
00 ON ON 00
o o o o
CO
»
o
OfSf-S'-9-*-*-*-Sf-«T
in m in in in in m m in
OOOOOOOOO
NONONONO>ONOvONONO
m ^ ^3
o o o
000
r~ NO NO
O O O
O C O
00 r^ r^
O 0 O
o o o
ON in ^H
in in in
0 O 0
f*l 00 O
o o o
r«» n 00
^ 9> 00
O O O
^ "^ *3"
m i/^ in
O O 0
tC *O ^O
r^
o
CO
o
ON
CN
0
CN
ON
00
NO
ON CN
00
m ON
CO
C
IN o
NO -*
* u
O cj
^
C
3
i
o
u
CO
o
C-l k
»
*
NO
C
o
T3
01
CO
CO
.u
C
i
CO
a
n*
CO
LJ
CO
cooooooooocoaocococoooco
oooooooooooo
o
33
O O O O O O
o o o o o
»^ {*! CN i-^ O ON 00 r** sD tO ^1
ON ON ON ON ON CO CO CQ CO 00 00
<
u
u
A-15
-------
HDGE are estimated at $21 million through 1984, $111 million in 1985 and 1986,
and $136 million in 1987 and thereafter.
A.3.2 Maintenance Costs
No changes in maintenance costs are expected for HDDES. For HDGEs for model
years 1985 and 1986, a small increase in annual maintenance costs ($3.90) is
due to minor servicing (operational checks and lubrication) of heated air
intake and automatic chokes. The cost attributable to CO control is
0.6 million dollars annually. For the 1987 and later model years, the
catalyst-equipped HDGE will realize an annual savings (relative to 1986) of
$49 per truck due to improved spark plug and exhaust system longevity
resulting from the use of unleaded gasoline. The related savings attributable
to HC control for the 1987 and later model years is $6.3 million dollars.
A.3.3 Use of Unleaded Fuel
Table A-10 presents estimated HDV unleaded fuel use for 1987, 1990 and 1995,
and the derivation of the portion of added unleaded fuel costs due to the
FMVCP for CO.
A.4 MOTORCYCLES
Based on an estimated average price increase of $102 due to FMVCP-required
motorcycle controls for CO and HC, and an assumed annual production of
725,000 units, the annual cost for motorcycle controls attributable to the CO
FMVCP is about $37 million. While some FMVCP-required design changes may
result in reduced maintenance, data which would allow quantitative estimates
of the potential savings are not available. (Gray, 1983)
It has been estimated that leaner air/fuel mixtures required to meet
motorcycle FMVCP limits will improve fuel economy by about 20 percent. Based
on an estimated motorcycle fuel consumption of 360 million gallons in 1981 and
a gasoline price of $1.20, the annual fuel savings could amount to
$54 million. (Gray, 1983; EIA, 1984)
A-16
-------
TABLE A-10.
COSTS OF UNLEADED FUEL USE DUE TO FMVCP FOR KDV
(In constant 1984 dollars)
Year
1987
1990
1995
Unleaded
Fuel Use
(109 gallons)3
0.95
2.47
4.06
Unleaded
Fuel Price
Differential
($/gallons)b
0.05
0.05
0.05
Total Added
Cost of HDV
Unleaded Fuel
for ($109)
0.048
0.124
0.203
Fraction
due to CO
FMVCPb
0.5
0.5
0.5
CO Portion
of Added
Unleaded Fuel
Cost ($109)
0.024
0.062
0.102
aEEA
bGray, 1983
A-17
-------
A.5 AIRCRAFT
EPA has withdrawn aircraft emission standards for CO.
A.6 SUMMARY OF FMVCP COSTS
Annualized costs of the FMVCP for CO as estimated in this Appendix are
summarized in Table A-ll.
A-18
-------
TABLE A-11
TOTAL ANNUALIZED COSTS OF THE FMVCP FOR CO ($1Q6)
(In constant 1984 dollars)
Passenger Cars (LDV)
Hardware
Unleaded Gasoline
Operating and Maintenance
Altitude Controls
Total (LDV)
1987
1,994.8
1,030.0
-802.8
100.7
2,322.7
1990
2,293.3
1,000.0
-887.7
180.2
2,585.8
1995
2,639.0
920.0
-989.1
274.4
2,844.3
Light-Duty Trucks (LDT)
Hardware
Unleaded Gasoline
Operating and Maintenance
Altitude Controls
Total (LDT)
571.6
340.0
-136.6
0.3
775.3
812.8
400.0
-181.1
0.5
1,032.2
1,158.1
480.0
-242.9
0.7
1,395.9
Heavy-Duty Vehicles (HDV)
Hardware
Unleaded Gasoline
Operating and Maintenance
Total (HDV)
111.0
24.0
-6.3
128.7
111.0
62.0
-6.3
166.7
111.0
102.0
-6.3
206.7
Motorcycles
Operating and Maintenance
-17.0
-17.0
-17.0
Grand Total
3,209.7
3,767.7
4,429.9
A-19
-------
REFERENCES FOR APPENDIX A
1. Automptiye__New8. April 25, 1979. "Increase in Retail Price of
Automobiles Due to Federal Requirements, 1968-1978." 1979 Market Data
Book Issue #4573, p. 95.
2. Automotive News. "1985 Market Data Book." Grain Automotive Group, Inc.,
Detroit, MI. April 24, 1985. Energy and Environmental Analysis, Inc.
(EEA). 1980. "Techniques for Estimating MOBILE2 Variables" Arlington,
VA. (Prepared for U.S. EPA, Ann Arbor, MI.)
3. EEA, Inc. 1982. "The Highway Fuel Consumption Model: Sixth Quarterly
Report." Arlington, VA. (Prepared for the U.S. Department of Energy
(DOE), Washington, DC).
4. EEA, Inc., November, 1984. "The Motor Fuel Consumption Report: Eleventh
Periodical Report. "Motor Vehicle Emission Laboratory, EPA, Ann Arbor,
MI.
5. Energy Information Administration (EIA), U.S. Department of Energy.
1984. "Petroleum Marketing Monthly, November 1984." Washington, DC.
6. Federal Register, February 18, 1983. "Control of Air Pollution From New
Motor Vehicles and New Motor Vehicle Engines; Emission Standards for 1984
Model Year Light-Duty Vehicles," (48 FR 7392-7397).
7. Gray, Charles L., Emission Control Technology Division, EPA, Ann Arbor,
MI. Memorandum to Bruce Jordan, Ambient Standards Branch, Office of Air
Quality Planning and Standards, EPA, Research Triangle Park, NC. "Updated
Cost Estimates of Controlling HC Emissions from Mobile Sources."
November 28, 1983.
8. McCurdy, Tom, Ambient Standards Branch, Office of Air Quality Planning and
Standards, EPA, Research Triangle Park, NC. 1984 (undated). Memo
entitled "Various Mobile Source Inputs to NAAQA Cost and Economic
Analyses."
9. U.S. EPA. Emission Control Technology Division. March 1978. "Analysis
of Technical Issues Related to California's Request for Waiver of Federal
Preemption With Respect to Exhaust Emission Standards and Test Procedures
for 1981 and Subsequent Model Years LDV."
10. U.S. EPA. 1979. "Cost and Economic Impact Assessment for Alternative
Levels of the National Ambient Air Quality Standards for Ozone."
EPA-450/5-79-002.
A-20
-------
11. U.S. EPA, Office of Mobile Source Air Pollution Control.
February, 18, 1983. "Controlling Emissions from Light-Duty Motor Vehicles
at Higher Elevations - A Report to Congress", EPA-460/3-83-001.
12. Ward's Communications, Inc. 1982. "1982 Ward's Automotive Yearbook,"
Detroit, MI.
A-21
-------
APPENDIX B
STATIONARY SOURCES: METHODOLOGY AND ASSUMPTIONS
B.I INDUSTRY COVERAGE
B.I.I Selection Criteria
The industries included for evaluation were selected from a preliminary
list of potentially significant CO sources (PEDCo, 1979b). Average process
and stack parameters were gathered from recent EPA regulatory and engineering
studies for approximately 25 industrial processes. Table B-l lists those
source categories identified. The resulting data (PEDCo, 1979b) indi-
cated, for each process, two production sizes typical of the range of
operation and the uncontrolled emission rate, stack height and exhaust
gas velocity, temperature, and volume appropriate to each size. Probable
control equipment and efficiencies also were identified.
In order to focus the derivation of process-specific control costs on
those processes likely to experience some need for emission reduction
under any of the alternative standards, it was decided to screen the
source list by evaluating the air quality impact of each "model" source
B.I.2 Air Quality Modeling of "Model Sources"
A single source dispersion model, PTMAX, was used to calculate a predicted
maximum concentration for each source (and size of source). PTMAX is an
interactive program which calculates the maximum short-term (one-hour)
ambient concentration from any point as a function of exhaust and stack
characteristics, wind speed, and stability. The program is part of the
User's Network for Applied Modeling of Air Pollution (UNAMAP) maintained
B-l
-------
TABLE B-l
POTENTIALLY SIGNIFICANT PROCESS SOURCES OF CO
Petroleum refineries
Carbon black
furnace/oil
furnace/gas
Kraft pulp and paper
Iron foundries
Steelmaking
sintering
basic oxygen furnace (EOF)
Primary aluminum
Gas pipelines
Electric utilities
Acrylonitrile
Maleic anhydride
Industrial combustion
Charcoal
Formaldehyde
Dimethylterephthalate
Phthalic anhydride
Coke production
Cyclohexanol
Gas-fired process heaters
Ethylene dichloride
Reciprocating internal combustion engines
B-2
-------
by the Division of Meteorology, EPA. To determine their maximum impact,
sources were modeled using their uncontrolled emission rate; no back-
ground level was assumed.
PTMAX predicts one-hour levels only. Because the ambient standards con-
sider both one- and eight-hour averaging periods, a method for converting
the predicted one-hour into expected eight-hour concentrations was necessary.
Persistence factors of 0.5 and 0.7, which represent ratios of eight-hour
to one-hour concentrations exhibited by monitored data, were used (Schewe,
1979). The 0.7 factor represents a ratio of maxima and, therefore, produces
a worst case estimate. The results of the PTMAX air quality modeling
are presented in Table B-2.
Eight of the model processes experienced potentially significant air
quality impacts. For the screen, "significant" was defined as 50 percent
or more of the most stringent ambient standard to be considered. These
processes were:
Basic oxygen furnace
Aluminum anode prebake
Aluminum prebake cells
Maleic anhydride
Cyclohexanol
Formaldehyde
Ethylene dichloride
Coke oven charging.
B.I.3 NEDS Screen
The PTMAX modeling indicated potential impacts for source categories
with a specific set of process and stack characteristics. The probabil-
ity that actual sources may deviate from those "model" parameter sets
created the need to screen further process categories which might cause
ambient violations.
B-3
-------
TABLE B-2
PTMAX MODELING RESULTS
Industry
Petroleum refining
Carbon black
Kraft pulp
Kraft pulp
Gray iron
Iron and steel
Steelmaking
Steelmaking
Primary aluminum
Primary aluminum
Primary aluminum
Gas pipeline
Electric utility*
Electric utility*
Electric utility*
8-hour Bounds
Process Source
Catalytic
cracking
Furnace process
Recovery boiler
Kiln
Cupola
Sinter plant
Electric arc
BOF
Prebake cells
Anode prebake
HSS cell
Engine exhaust
Coal
Residual oil
Gas
Size
25,000 BPD
90,000 BPD
176,000 TPY
44,000 TPY
1,500 TPD
1,000 TPD
1,500 TPD
1,000 TPD
25 TPH
1 TPH
3.65 MTPY
1.19 MTPY
0.10 MTPY
1.13 MTPY
1.61 MTPY
3.78 MTPY
225,000 TPY
140,000 TPY
225,000 TPY
140,000 TPY
225,000 TPY
140,000 TPY
1,500 HP
1,000 MWe
100 MWe
1,000 MWe
100 Mwe
1,000 MWe
100 MWe
Lower
(ppm)
1.49
1.58
1.77
2.54
0.003
0.002
0.19
0.17
1.71
0.79
1.00
0.91
0.02
0.03
29.92
62.20
0.87
1.12
7.32
6.64
0.45
0.55
0.004
0.001
0.000
0.001
0.000
0.001
0.000
Upper
(ppm)
2.09
2.21
2.48
3.55
0.004
0.003
0.26
0.23
2.40
1.09
1.40
1.27
0.03
0.04
41.89
87.02
1.21
1.56
10.24
9.30
0.63
0.77
0.005
0.001
0.001
0.001
0.001
0.001
0.000
Distance
(km)
0.80
1.56
17.50
0.62
2.07
1.82
0.51
0.40
0.37
0.21
1.13
0.58
0.63
1.58
0.62
0.60
1.01
0.36
0.36
0.38
1.44
0.99
0.22
2.50
1.15
2.24
1.15
2.29
41.93
* 75 m. stack height.
B-4
-------
TABLE B-2 (cont'd)
PTMAX MODELING RESULTS
Industry
Acrylonitrile
Maleic anhydride
Industrial boilers
Industrial boilers
Industrial boilers
Charcoal
Charcoal
Formaldehyde
Dimethyltereph-
thalate
Phthalic anhydride
Iron and steel
Cyclohexanol
Process heater
Ethylene dichloride
Internal combustion
8-hour Bounds
Process Source
Absorber vent
Condenser
Coal
Distillate
Gas
Kiln
Continuous
furnace
wood-drying
Absorber vent
Scrubber
Condensors
Coke oven
charging
Product recovery
vent
Gas-fired
Fractionation
vent
Diesel
Dual fuel
Natural gas
Size
0.154 MTPY
0.121 MTPY
0.027 MTPY
0.006 MTPY
50 lO^Btuh
K
250 10 Btuh
1 lO^Btuh
10 10 Btuh
1 lO^Btuh
10 10 Btuh
215 Lbh
7.5 TPH
7.5 TPH
0.100 MTPY
0.040 MTPY
0.120 MTPY
0.030 MTPY
0.065 MTPY
0.047 MTPY
0.35 MTPY
0.72 MTPY
0.110 MTPY
0.025 MTPY
150 lO^Btuh
50 10 Btuh
0.440 MTPY
1,200 HP
4,300 HP
1,500 HP
4,400 HP
Lower
(ppm)
0.60
0.57
5.65
4.33
0.001
0.001
0.000
0.001
0.000
0.001
0.10
0.04
1.21
1.48
2.67
0.52
0.44
0.63
0.56
0.82
1.24
7.13
3.99
0.000
0.000
3.31
0.03
0.08
0.004
0.01
Upper
(ppm)
0.83
0.80
7.90
6.06
0.002
0.002
0.000
0.002
0.000
0.001
0.15
0.06
1.70
2.08
3.73
0.73
0.62
0.88
0.78
1.14
1.73
9.98
5.58
0.000
0.000
4.63
0.04
0.12
0.01
0.01
Distance
(km)
0.61
0.61
0.31
0.29
0.41
0.58
0.21
0.25
0.21
0.25
0.17
1.06
0.40
0.18
0.20
0.44
0.27
0.48
0.44
0.24
0.22
0.27
0.17
0.62
0.41
0.37
0.20
0.31
0.22
0.48
B-5
-------
To conduct a second screen, the PTMAX program was modified to calculate
each source's effective plume rise (a measure of the dispersion character-
istics of emissions) and to determine, for that plume rise, the CO emis-
sion rate that would yield a carbon monoxide ambient concentration of
0.5 ppm (under worst case meteorological conditions, i.e., wind speed of
2.0 m/s and stability class A). The modified PTMAX program was applied
to the source data in the National Emission Data System (NEDS) point
source subfile. Therefore, based on actual rather than "model" stack
and emission characteristics, this PTMAX screen produced a subfile com-
prised of those point sources in NEDS which would produce ambient con-
centrations above 0.5 ppm; the 0.5 ppm concentration was selected as a
conservative threshold of potential to cause ambient violations even
under the most stringent alternative.
The resulting subfile included over 700 point sources. Additional source
categories beyond those eight previously listed then were identified for
control cost development based on 1) their expected magnitude of air
quality impact, or 2) the number of sources within that category con-
tained in the subfile.
Data validity checks were performed on a number of sources in the subfile.
Emission rates (AP-42 uncontrolled vs. NEDS) and their consistency with
operating and stack data were verified. After correction for apparent
data errors, six additional process categories were identified for cost
development; they were:
Carbon black
Iron sintering
Electric arc furnace
Gray iron cupola
Conical wood incinerator
Municipal incinerator.
B-6
-------
B.I.4 Source Category Control Cost Development
For the 14 potentially significant processes screened in the preceding
two steps, capital and annualized costs were obtained for two typical
operating rates within each process category.* The technology selection
was based on currently demonstrated control equipment; no experimental
technology was considered (PEDCo, 1979a).
As indicated, costs (capital and annualized) were developed for two pro-
cess sizes which are deemed reflective of the range of typical opera-
tion. The two cost estimates were used to construct an exponential
equation for each process category which, upon inputting a source's
specific annual capacity or production level, would produce the costs of
a unit of appropriate size. The log-log curve represented by the equa-
tion is used to capture economies or diseconomies of scale inherent in
the two size-specific point estimates.
In three cases, a single process size was indicated, hence the control
costs for that size were assumed to be linear per unit of production (a
uniform $/ton of product). For one process, the conical wood incinerator,
the costs of control do not vary by size; therefore, the capital and
annualized costs are level.
B.2 TOTAL COST METHODOLOGY
B.2.1 Evaluation of Ambient Impact
In order to determine the costs incurred by each alternative standard,
the evaluation of the extent to which existing sources violate mandated
concentration levels constitutes the first step. PTMAX was applied to
the 700 point sources in the NEDS subfile. For the purposes of costing,
There are several exceptions where only one size was costed. See
PEDCo, 1979b.
B-7
-------
the meteorological assumptions were stability class B and a wind speed
of 2.5 m/sec, as it was felt that these represent more realistic atmos-
pheric conditions and would produce consistently valid results from PTMAX
(i.e., plume rises of less than 500 m).*
A background level of 2 ppm CO was included to reflect the probable con-
centration of area sources at points of maximum point source impacts.
The 2 ppm background was assumed sufficient since point sources generally
are removed from highways and other major transportation routes where
high CO concentrations from area sources occur. No natural background
was considered.
The background level, assumed equal for all point sources, can be consi-
dered either to increase the point's impacts on air quality (since the
source's concentration is no longer considered truly in isolation) or to
reduce effectively, by 2 ppm, each alternative ambient standard. For
ease of analysis, the background was incorporated in the latter manner.
Each point source's maximum concentration was calculated using PTMAX and
revised atmospheric conditions. The concentrations of multiple points
at any facility were aggregated to the plant level, thereby assuming
that the plant impact equals the sum of the process maximum impacts.
This procedure will overestimate the actual plant impact because it
assumes that these maxima occur at the same location. No interaction
among plants was assumed; the source file produced under the revised
meteorological assumptions contained 275 sources which did not exhibit a
significant degree of co-location. It was believed that the level of
source interaction would not be of a magnitude that would jeopardize the
conclusions reached using PTMAX.
* PTMAX will calculate ambient concentrations for plume rises above 500 m,
but will identify the prediction as invalid.
B-8
-------
The calculated plant concentration level then was compared to each alter-
native CO standard adjusted for background. (The plant total was cor-
rected by the 0.7 persistence factor for comparison with eight-hour stan-
dards.) It should be noted that the form of the standard, whether 1 or 5
expected exceedances, was not considered. The calculated impact was used
directly. The adjustment to a single or multiple exceedances form of a
standard would require varying meteorological conditions to produce the
variability exhibited by monitored data. Such a data manipulation was
beyond the scope of this analysis. The result is, as with the sura of
process maxima, a bias upward in the forecasted i- -?:t.a3i_cu
B.2.2 Calculation of Required Emission Reductions
Emission reductions required under each alternative standard were calcu-
lated assuming proportionality between reduction in emissions and air
quality levels; a linear rollback equation was used:
o/ r. . _ , Plant Concentration - (Ambient Standard - Background)
% Emission Reduction = - - - 6 - -
Plant Concentration
If the plant impact does not violate any standard, the equation produces
a negative result and no reductions are calculated.
B.2.3 Least-Cost Strategy Determination
The required reduction in emissions refers to plant totals. Since many
plants constitute the combined impact of multiple emission points, emis-
sion reductions can be produced by a variety of control option combinations.
For this analysis, the required emission reductions were obtained through
an option or combination of options which resulted in the lowest total
annualized cost for the plant, thus accounting for initial capital outlays.
Each plant's control options were calculated and ranked by cost effective-
ness, that is, lowest to highest cost per ton of CO removed. All options
needed to produce the tonnage emission reduction are applied to each
B-9
-------
source in order of decreasing cost effectiveness; that is, the lowest
$/ton removed controls and sources are selected first.
Because of the discrete nature of control technology, it is possible
that a plant might employ a less cost-effective means of control simply
by choosing control options which have the lowest average annualized
cost per unit of reduction. For example, a facility consists of two
points: point A emits 100 tons per year and point B emits 200 tons per
year. Assume that the required emission reduction is 50 percent for the
plant, or 150 tons; point A can be controlled at 90 percent efficiency
for $l/ton; point B may be controlled at 75 percent efficiency for $2/ton.
By a cost-effective ranking, the choice would be to control point A first;
however, since point A only provides 90 tons of reduction, point B also
must be controlled. As a result, the plant expends $390 in annualized
cost for 240 tons of reduction. However, if only point B were controlled,
it could achieve the entire 150 ton reduction at a total annualized cost
of $300. A verification procedure was used, therefore, to ensure lowest
total annual cost, in this example, to check whether the controls on
point A were necessary;"" since they were not, the least-cost algorithm
would control point B only. The results of the costing procedure are
presented in Section 3.5.
B.3 BENEFITS FROM STEAM CREDITS
The burning of carbon monoxide is an exothermic reaction. If the exhaust
stream of CO from a process is sufficiently dense, the energy released
from burning the gas may be used in one of two manners. The energy from
combusted CO can be used to heat the continuing stream of process gas
* The need for controls on point B are not reverified, however, since
point B's controls obviously were necessary to achieve the total reduction.
B-10
-------
before combustion. This procedure, called primary heat recovery and
illustrated in Figure B-l, can reduce the amount of supplemental fuel
needed to burn the process gas. Secondary heat recovery, as illustrated
in Figure B-2, generates steam from the heat released from the combusted
CO. This steam may be used in certain processes within the plants thereby
reducing costs of burning fossil fuels to produce needed steam.
The use of secondary heat recovery increases capital cost, however. For
example, using incineration with primary heat recovery to control CO
from an iron ore sinter plant producing 1.19 x 10 tons per year incurs
a capital cost of $3.4 million. With the addition of secondary heat
recovery the capital cost becomes $3.86 million, an increase of over 13
percent.
However, the secondary heat recovery provides enough annual steam credit"'"
to more than offset the capital cost increase; using the same sinter
plant, control with primary heat recovery operates at an annual cost of
$3.5 million. The addition of secondary heat recovery provides enough
steam to provide a net annual savings of $3.3 million (PEDCo, 1979b).
It should be noted that not all sources can use secondary heat recovery.
Its application will depend on the quantity/concentration of CO in the
off-stream and the degree of steam-heating requirement (i.e., if supple-
mental fuel for combustion is necessary). Moreover, it is possible that
certain processes would not be amenable to secondary heat recovery because
no steam is required. In this event, credits for steam are not valuable
since steam normally is not purchased or produced. As a result, firms,
in choosing to minimize expenditures, likely would select primary recovery
since, without credits, secondary recovery is more expensive.
The steam credit is a product of calculating the cost of steam (based
on fossil fuel combustion) times the quantity of steam produced by a
CO boiler.
B-ll
-------
FIGURE B-l
SCHEMATIC OF INCINERATION SYSTEM
WITH PRIMARY HEAT RECOVERY
TO STACK
PRIMARY
HEAT
EXCHANGER
SUPPLEMENTAL
FUEL
1
COMBUSTION
CHAMBER
Source: PEDCO Environmental Inc. June 1979.
Processes." Prepared for U.S. EPA.
"Carbon Monoxide Control Costs for Selected
B-12
-------
FIGURE B-2
SCHEMATIC OF INCINERATION SYSTEM WITH PRIMARY
AND SECONDARY HEAT RECOVERY
TO STACK
WATER
SECONDARY
HEAT
EXCHANGER
k
>STEAM
PRIMARY
HEAT
EXCHANGER
SUPPLEMENTAL
FUEL
COMBUSTION
CHAMBER
Source: PEDCO Environmental, Inc. June 1979. "Carbon Monoxide Control Costs
for Selected Processes." Prepared for U.S. EPA.
B-13
-------
The application of secondary heat recovery by any process was based on
engineering judgment concerning the off-stream characteristics and the
utility of the steam output. Since secondary recovery produces a nega-
tive cost, it always is selected by the costing algorithm for those
processes where it is an option. The possibility that actual units may
not be appropriate for secondary recovery, as opposed to the "model"
processes costed, implies that the credits calculated may be optimistic.
In the event that, say, half the credits were not technologically or
economically feasible, the annualized costs presented in Section 4 are
understated while the capital costs are overstated. Because the capital
cost picture changes in the absence of steam utility, it is not necessarily
the case that annual costs merely are halved. Rather, on a case-specific
basis, a determination would be needed regarding the advantage of secondary
versus primary recovery relative to the degree of steam needed. For the
purpose of this analysis, 100 percent credit was assumed.
B.4 CO SOURCE GROWTH
Inclusion of growth in CO emission sources would serve two purposes:
To add costs automatically incurred due to New Source Perfor-
mance Standards (NSPS) to the baseline costs
To identify the level and cost of new source control motivated
by the ambient air quality standard for CO.
Towards these ends, an evaluation was made of the extent to which new
source controls are mandated by air quality limitations for CO.
At present, two process categories have NSPS promulgated for CO: the
catalytic cracker (petroleum refining) and the electric submerged arc
furnace (ferroalloy production). The control technology for the cata-
lytic cracker is a CO boiler. For a new source, this type of control
would be used in the absence of the standard because its operation
B-14
-------
provides a net credit due to steam generation and, thus, an economic
benefit. Since economics, rather than emissions control, constitutes
the prime motive for CO recovery, the costs are not attributed to NSPS.
Moreover, even uncontrolled FCCU have negligible ambient impact (refer
to Table B-2); it is improbable that any of the alternative standards
would motivate more stringent control, or significant costs.
Ferroalloy furnaces do not employ end-of-pipe controls specifically for
CO reduction. In open furnaces, the CO combusts as part of the process
while, in semi-enclosed furnaces, flaring the exhaust gas to reduce visible
emissions also reduces CO. For this reason, the NSPS for this process
was not considered to incur cost. Further, since the NEDS screen did
not capture any ferroalloy furnace, it appears logical to conclude that
this process does not constitute a significant source of ambient impact.
Accordingly, additional control or costs under any alternative s
are not likely.
Beyond NSPS, other processes and industries were examined. Since incor-
poration of growth would be made, analytically, on a model source basis,
only five processes would exhibit potential impacts (see Table B-2) at
uncontrolled levels, thus motivating controls and incurring costs.
However, an examination of the typical control measures for these processes
revealed either of two practices:
Use of secondary heat recovery (e.g., CO boilers) to capture
CO as a fuel, or
Use of incineration of hydrocarbons to reduce adequately CO
emissions.
In either case, it would be inappropriate to assign the total costs or
credits associated with these types of measures to the carbon monoxide
standard. Moreover, because the total costs incurred by existing point
B-15
-------
sources are relatively small,"" the significance of a growth increment
likely would be small as well. The analysis indicated, therefore, that
the incorporation of growth would not have noticeable effects on the
magnitude of the costs estimated for attaining CO standards. For this
reason, especially, the cost analysis is confined to the existing base
of stationary point sources.
Vis-a-vis mobile source total costs,
B-16
-------
REFERENCES FOR APPENDIX B
PEDCo Environmental, Inc, (PEDCo). May 14, 1979a. Letter to EEA,
Report PN3264-BB.
PEDCo. June 1979b. "Carbon Monoxide Control Costs for Selected
Processes."
Schewe, G. (EPA). April 12, 1979. Memo to J. Sonmers (OMSAPC). "Reply
to Request for Concentration Estimates Near Roadways Due to Mobile Source
Emissions of Sulfuric Acid and Diesel Particulates (TSP and BaP)."
B-17
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO. 2.
EPA-450/5-85-006
4. TITLE AND SUBTITLE
Cost and Economic Assessment of Alternative National
Ambient Air Quality Standards for Carbon Monoxide
(Revised)
7. AUTHOR(S)
Mark Smith, Susan Schechter, and Thomas McCurdy
9. PERFORMING ORGANIZATION NAME AND ADDRESS
GCA/Technology
Chapel Hill, NC
Energy and Environmental Analysis, Inc.
Aleyandria VA
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Air Quality Planning and Standards
Office of Air and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
3. RECIPIEN ~ 3 ACCESSION NO.
5. REPOR ." DATE
July 1985
6. PERT-ORMING ORGANIZATION CODE
GCA/Technology
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68^02-3804
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The report presents cost and economic impact information needed to undertake
a complete benefits/cost analysis of setting alternative carbon monoxide NAAQS.
Four alternative 8-hour NAAQS were analyzed: 9 ppm, one observed exceedance,
and 9,12,15 ppm one expected exceedance of daily maxima. The cost estimates
are for both total societal and industrial sector control technology installation
and operation. These include capital changes, annualized capital expenditures,
and annual operating and maintenance/repair costs. Mobile source expenditures are
analyzed separately from stationary source control costs. The economic impact
analysis includes data on industry's ability to bear the capital expense and the
inflationary and competitive impacts of incurring the control costs. Three time
periods of analysis are used: 1987, 1990, and 1995.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Air Pol lution
Carbon Monoxide
18. DISTRIBUTION STATEMENT
Release to Public
b. IDENTIFIERS/OPEN ENDED TERMS
Air Quality Standards
Cost Analysis
Economic Analysis
Control Techniques
19 SECURITY CLASS (This Report,
Unclassifipd
20 SECURITY CLASS /Thispage:
Unclassified
c. COS AT I F-'ield/Group
21 NO. OF PAGES
162
22. PRICE
Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
-------
INSTRUCTIONS
1. REPORT NUMBER
Insert the EPA report number as it appears on the cover of the publication.
2. LEAVE BLANK
3. RECIPIENTS ACCESSION NUMBER
Reserved for use by each report recipient.
TITLE AND SUBTITLE
'"itle should indicate clearly and briefly the subject coverage of the report, and be displayed prominently. Set subtitle, if used, in smallei
oe or otherwise subordinate it to mam title. When a report is prepared in more than one volume, repeat the primary title, add volume
mber and include subtitle for the specific title.
5. REPORT DATE
Each report shall carry a date indicating at least month and year. Indicate the basis on which it was selected (e.g., date of issue, date of
Pi. oval, date of preparation, etc.).
6. PERFORMING ORGANIZATION CODE
Leave blank.
7. AUTHOR(S)
Give name(s) in conventional order (John R. Doe, J. Robert Doe, etc.). List author's affiliation if it differs from the performing organi
zation.
8. PERFORMING ORGANIZATION REPORT NUMBER
Insert if performing organization wishes to assign this number.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Give name, street, city, state, and ZIP code. List no more than two levels of an organizational hirearchy.
10. PROGRAM ELEMENT NUMBER
Use the program element number under which the report was prepared. Subordinate numbers may be included in parentheses.
11. CONTRACT/GRANT NUMBER
Insert contract or grant number under which report was prepared.
12. SPONSORING AGENCY NAME AND ADDRESS
Include ZIP code.
13. TYPE OF REPORT AND PERIOD COVERED
Indicate interim final, etc., and if applicable, dates covered.
14. SPONSORING AGENCY CODE
Insert appropriate code.
15. SUPPLEMENTARY NOTES
Enter information not included elsewhere but useful, such as: Prepared in cooperation with, Translation of. Presented'at conference of,
To be published in, Supersedes, Supplements, etc.
16. ABSTRACT
Include a brief (200 words or less) factual summary of the most significant information contained in the report. If the report Contains a
significant bibliography or literature survey, mention it here.
17. KEY WORDS AND DOCUMENT ANALYSIS
(a) DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper authorized terms that identify the major
concept of the research and are sufficiently specific and precise to be used as index entries for cataloging.
(b) IDENTIFIERS AND OPEN-ENDED TERMS - Use identifiers for project names, code names, equipment designators, etc. Use open-
ended terms written in descriptor form for those subjects for which no descriptor exists.
(c) COSATI FIELD GROUP - Field and group assignments are to be taken from the 1965 COS ATI Subject Category List. Since the ma-
jority of documents are multidisciplinary in nature, the Primary Field/Group assignment(s) will be specific discipline, area of human
endeavor, or type of physical object. The application(s) will be cross-referenced with secondary Field/Group assignments that will follo\
the primary postmg(s).
18. DISTRIBUTION STATEMENT
Denote releasability to the public or limitation for reasons other than security for example "Release Unlimited." Cite any availability to
the public, with address and price.
19. &20. SECURITY CLASSIFICATION
DO NOT submit classified reports to the National Technical Information service.
21. NUMBER OF PAGES
Insert the total number of pages, including this one and unnumbered pages, but exclude distribution list, if any
22. PRICE
Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
EPA Form 2220-1 (Rev. 4-77) (Reverse)
------- |