NATIONAL ASSESSMENT
OF VOC, CO, AND NOx
CONTROLS,
EMISSIONS, AND COSTS
EPA Contract No. 68-W8-0038
Work Assignments 6, 9, and 35
Prepared for:
Oftice of Policy Planning and Evaluation
U.S. Environmental Protection Agency
Washington, DC 20460
Prepared lay:
E.H. Pechan & Associates, Inc.
5537 Hempstead Way
Springfield, VA 22151
September 1988

-------
!>>
U
flj
a
CO
Ifi
o
u
"o
u,
_ —>
p
-------
EXECUTIVE SUMMARY
INTRODUCTION
The Clean Air Act mandated deadline of December 31, 1987,
elapsed with a long list of areas still not attaining national
ambient air quality standards (NAAQS) for ozone and carbon
monoxide (CO), In anticipation of this shortfall, the U.S.
Environmental Protection Agency (EPA) developed a program to
address the 1ikelihood that many areas would not attain the NAAQS
and published this proposed policy in the Federal Register
November 17, 1987. This announcement prompted much interest
among state and local air pollution control agencies and at EPA
to determine what effect this new policy might have' on the
remaining nonatta inment areas. A substantial number of volatile
organic compound (VOC) (precursors of ozone) and CO control
measures have been imposed since the Clean Air Act was passed,
especially in the large metropolitan areas of the United States.
Legislation introduced before Congress in 1987 and 1988 to
address some of the same issues included in the EPA nonatta inment
policy prompted calls for quantitative analyses of each of the
Congressional bills and proposals as well as the EPA policy.
Initial interest among the Congressional alternatives focused on
S. 1894, introduced by Senator George Mitchell and otherwise
known as the Mitchell bill. In the House, a bill introduced by
Rep. Henry Waxman (H.R. 3054) presented some alternative
nonattainment provisions. These were followed by another
proposal formulated by nine Congressmen, which has come to be
known as the Group-of-Nine Proposal. This report presents a
quantitative assessment of the control costs and emission
reductions that might be expected from each of these three
Congressional alternatives and compares them with what would be
expected to happen under the EPA policy.
MODELING METHODS
In reviewing analytic tools available for performing an
analysis of VOC, oxides of nitrogen (N0X), and CO control costs
for dif ferent ozone and CO nonatta inment control approaches, it
was found that no single model was available for any of the three
pollutants that could perform all of the required analyses in a
timely fashion. Therefore, new models were developed to meet the
speci f ic objectives of this project. These models vary in
complexity, with the NOx analysis being the simplest, VOC the
most complex, and CO somewhere in between. Most of the
analyt ical ef fort in this study was spent on model ing est :	-tated
future year VOC emissions and costs, so this summary focuses
primarily on that part of the analysis.
The Emission Reduction and Cost Analysis Model for VOC
(ERCAM-VOC) was developed to analyze alternative measures for
reducing emissions of VOCs, precursor to ozone. The model runs
11,]
-------
on a personal computer and is programmed in dBase III Plus. It
is designed to simulate the process that states and metropolitan
areas might use to move toward attainment of the ambient ozone
standard (the ozone NAAQS) under alternative policies.
The modeling data base is the 1985 NEDS point and area
source emission inventory of VOC sources, which was augmented to
provide the best possible representation of ozone season
emissions and controls in place. This data base was chosen
because it was the product of a multi-year research effort to
develop an accurate and comprehensive inventory of 1985
emissions. The 1985 inventory was also selected because it
matches the time period of the air quality data used in the
study. Emissions and control data were organized by source
classes designed to reflect common emission and control
characteristics of different sources.
The organization of the emissions data input to the model is
by Metropolitan Statistical Area (MSA) and by attainment/
nonattainment area (as well as source category). Nonattainment
areas are categorized according to the level of severity of their
nonattainment problems.
The first modeling step is to compare 1985 NEDS listed
control levels with those mandated by state and local
regulations. Control costs and emission reductions are then
estimated for all sources not in compliance with these
regulations.
New source growth is estimated using Bureau of Economic
Analysis growth rates by industry for stationary sources and
vehicle miles traveled projections for mobile sources. The
applicable New Source Performance Standards (NSPSs) and state and
local regulations for each source category are used to estimate
new source emission rates and control costs. Federal motor
vehicle emission standards affect future motor vehicle emissions
in all areas.
Scenario files allow individual control options beyond those
already being applied for each source category to be selected.
Discretionary control measures included in the analysis include
methanol-fueled cars, more stringent vehicle inspection and
maintenance programs, tighter vehicle emission standards,
consumer solvent controls, and restrictions on hazardous waste
treatment, storage, and disposal facility emissions. The model
has been used to assess the VOC control cost and emission impacts
for projection years 1995 and 2000. Model results are at the
national, state, and Metropolitan Statistical Area (MSA) levels
by industry and source type.
CAVEATS
Any analysis that attempts to estimate how future laws or
regulations will affect the behavior of individuals, firms, and
i v
-------
state arid local regulatory agencies must incorporate simplifying
assumptions. In addition, data bases are employed which may not
be perfectly designed for the analysis being performed. The most
important caveats and assumptions associated with this analysis
are listed below. The most important of these caveats is that,
as a general rule, the model results presented in this study are
more useful for comparing the relative impacts of alternative
policies and bills than they are in estimating absolute values.
. Growth in motor vehicle travel is estimated using national
averages for all areas. These national average growth rates
are different for each of the four vehicle types modeled
(light-duty gasoline vehicles, light-duty gasoline trucks,
heavy-duty gasoline vehicles, and heavy-duty diesel
vehicles). Area specific growth rates are typically
available, but they do not permit separate rates to be
specified for the four vehicle types modeled so they were
not used. In any case, motor vehicle projections in this
analysis will not capture city-by-city differences in
travel.
. New stationary source growth is estimated using Bureau of
Economic Analysis values published in 1985. These rates may
overestimate growth in areas with petroleum-based economies.
. New source costs include all the costs of going from zero to
the indicated level of control. Some controls may be
undertaken for economic, process, or non-ozone related, non-
pollution control reasons. Therefore, total cost estimates
probably overestimate the costs of the policies/bills for
new sources.
. The modeling approach used in this study may also be biased
toward estimating higher costs to existing sources than
might actually occur. Whenever a controlled existing source
is forced to increase its control level, ERCAM-VOC estimates
the cost of the new control equipment without taking into
account the salvage value or reduction in operating cost
associated with the previous control technique. Less costly
upgrades to current control systems are also not considered.
. The 1985 NEDS VOC emission estimates for some area source
categories were adjusted downward to account for likely (but
not recorded) control levels in nonattainment areas. This
change affected emission estimates for the following area
source categories: paper surface coating, decreasing,
rubber and plastics manufacturing, and stage I gasoline
marketing. This change makes 198 5 VOC emission estimates
lower and provides less opportunity for future emission
reductions. No adjustments were made to base year motor
vehicle VOC emission estimates to try to include excess
evaporative and running loss emissions because quantitative
estimates of these values were not available during the
study period.
-------
. Rule effectiveness is almost always less than originally
predicted. The proposed EPA policy states that areas can
take only 80 percent emission reduction credit for various
measures. The effect of this 80 percent rule effectiveness
provision in the policy is not modeled in this analysis.
, Where bills and policies call for control measures which
have not been previously demonstrated or studied, there is
considerable uncertainty in control costs. To avoid
omitting important source types from the analysis, default
cost per ton values have been adopted for a number of
different control options, including controls for
miscellaneous point sources, consumer solvent controls, and
discretionary controls beyond those for which there are some
data.
. Ozone and CO design values from 1983 to 1985 data have been
used in this study. The presence of the generally high
values measured in 1983 is to some extent representative of
the high values that have been measured in the summer of
1988. Nevertheless, estimated control requirements by MSA
would change if more recent data were used. Note also that
these control requirements have been estimated with a
simplified ozone trajectory model with considerable
uncertainty.
. Not all of the policy and bill provisions could be
explicitly included in this analysis. For instance, no
attempt was made to quantify the effects of changing new
source review procedures. Future effects of cold start
certification testing for motor vehicles were also not
included in this analysis.
. The point source data file used in this analysis has
incomplete data for plants that emit less than 100 tons per
year of VOC. Therefore, this study may underestimate
emission reductions associated with regulatory approaches
that make smaller VOC emitters subject to controls.
. Control of emissions in ozone transport regions as defined
in the bills is not assumed to assist in reaching
attainment. Ozone transport regions contain attainment
areas that contribute emissions through atmospheric
transport to other areas not in attainment. Thus, while
costs are estimated for controls in these regions, any
benefits are not included.
. M0X costs have only been estimated for the ex plicit
provisions of the Waxman and Mitchell bills that require NOy
controls. Additional NOx controls may be undertaken in some
areas under the proposed EPA policy or the Group of Nine
proposal, but no attempt has been made to capture these
costs. The effects of N0X control on ozone concentrations
v i
-------
{plus or minus) have been ignored in all cases. These
assumptions could lead to overestimating or underestimating
M0>; control costs and benefits, depending on the area
involved.
RESULTS
Costs to attain the ozone NAAQS and the CO NAAQS were
analyzed using the ERCAM models for VOC and CO and a similar but
simplified analysis method for N0X. This summary briefly
highlights the results. The reader should consult Chapter IX for
attainment costs by MSA and by state.
Ozone Nonattainment
Because attainment/progress requirements affect emission
reductions and costs of the policies/bills, those requirements
are summarized first in Table I. Note that while the Mitchell
bill does not require areas with ozone design values above 0.27
parts per million (ppm) to attain by 2000, the yearly percentage
reduction requirements of that bill effectively force all areas
to attain by then.
Figure 1 shows how the estimated ozone precursor control
costs differ among the EPA policy and the alternative
Congressional bills and proposals. Both 1995 and 2 000 cost
estimates are shown. Expected additional ozone control cost
expenditures under the pre-1988 EPA policy are delineated in the
figure as part of the total EPA policy cost. Although estimates
of the total costs of the EPA policy and the alternative
Congressional bills/proposals are presented, Figure 1 is most
useful for showing the relative costs of the different control
approaches. The total costs should be used with caution because
they do not include the historical costs of VOC control such as
Federal Motor Vehicle Control Program costs and costs of existing
controls for stationary sources.
While Figure 1 shows the Group of Nine costs to be lower in
2000 than the expected EPA policy costs, this lower value depends
in part on high levels of consumer solvent VOC emissions control
being achievable by 2000 at $2,000 per ton. This is a lower cost
per ton than that used to estimate the cost of reducing
"residual" tons for the other alternatives. The consumer solvent
control level is limited in the other simulations. This issue is
discussed more fully in Chapter VIII.
When costs of the different policies/bills are compared, so
should the number of remaining ozone nonattainment areas. Table
II presents ERCAM-VOC estimates of residual nonattainment areas
in '1995 and 2000. Thus, of the three legislative approaches, the
lower costs of the. Group of Nine Proposal must be balanced
against the longer list of expected nonattainment areas. Note
also that the Table II list of residual nonattainment areas
represents what the policies/bills require and is not an
v 1 ]
-------
Table I
Ozone NAAQS Attainment/Progress Requirements of Proposals Analyzed
1995
2000
EPA Policy
Attain Standard or
achieve 3% per year
reductions, whichever
is binding
Attain Standard or
achieve 3% per year
reductions, whichever
is binding
Waxman
Moderate and Serious
must attain
All areas must attain
Severe areas must reduce
emissions by 10% of the
reduction required to
attain the standard each
year
Mitchell Moderate must attain
Serious and Severe must
achieve a 55% reduction
or attain whichever is
less stringent
All except areas with
design values above 0.27
ppro must attain
Group of Moderate I and II must
Nine	attain
Serious must achieve
78% of attainment
target
All except Severe must
attain
Severe must achieve 71%
of attainment target
Severe must achieve
41% of attainment
target
v i i i.
-------
Figure 1
Ozone Nonattainment Control Cost Summary
1995 Costs	2000 Costs
EPA Folic}'
Mitchell Bill
Waxman Bill
, _ |	j	 |	Group of Nine
! 1	11	Proposal
. — |	|					 1	l	,
0 4 8 12 16 20	0 4 8 12 16 20 24 28 32
Estimated New Expenditures (Billion $) Estimated New Expenditures (Billion $)
Ranges reflect costs of controlling residual tons using a range of $2,000 to $10,000 per ton.
Pre-1988 EPA Policy costs are not included here but can be found in Chapter V.
-------
Table II
Residual Ozone Nonattainment Areas by Projection Year*
EPA
Pol icy
Mitchell
Bill
Waxman
Bill
Group of Nine
Proposal
1995
Chicago
Houston
Los Angeles
Milwaukee
New York
San Diego
San Francisco
Chicago
Houston
Los Angeles
New York
San Diego
Chicago
Houston
Los Angeles
New York
Philadelphia
San Diego
Greater Conn.
Massachusetts
Chicago
Cincinnati
Dallas
El Paso
Fresno
Houston
Los Angeles
Milwaukee
Modesto
New York
Philadelphia
Phoenix
Sacramento
San Diego
San Francisco
Santa Barbara
Greater Conn.
2000
Los Angeles
New York
Ch icago
Houston
Los Angeles
New York
San Diego
San Francisco
Greater Conn.
*Emission reduction targets have
area using EKMA. Uncertainties
reduction is needed to bring an
results presented here.
been estimated for each urban
in estimating how much emission
area into attainment affect the
>:
-------
expectation of when specific areas might attain the ozone
standard.
One of the important findings of this study (and other
similar studies) was that there are not enough identifiable
control measures to calculate how much it might cost for all
metropolitan areas to attain the ozone NAAQS. Therefore, the
cost of controlling "residual" tons after all identifiable
controls are applied was estimated using a range of $2,000 to
$10,000 per ton. Thus, ranges of cost estimates are presented
for each of the policies/bills in Figure 1 and Table III.
Carbon Monoxide Nonattainment
Because a number of areas not attaining the CO standard were
also ozone nonattainment areas, costs of measures to help MSAs
(and non-MSAs) attain the CO ambient standard presented in
Chapter IX are those in addition to what is estimated to be spent
to comply with the ozone related provisions of the policy or
bill. This effort to avoid double counting control costs affects
I/M costs. Thus, the bill with the most stringent I/M
requirements for CO may not have the highest costs, because
similarly stringent ozone requirements have probably already-
accounted for most of the cost increase in areas that violate
both standards.
Table IV summarizes estimated CO costs by control measure
for the EPA policy and the three legislative approaches. CO
costs of the EPA policy are much lower than the costs of the
three legislative approaches. The only CO control measure
modeled as if it were mandated by the EPA policy is enhanced I/M.
While the proposed EPA policy mentions 17 ppm as a possible
cutoff for requiring enhanced I/M, a lower cutoff was used in
this analysis because preliminary simulations showed that many
areas with design values below 17 ppm would not be able to
demonstrate short-term attainment without new measures. Thus,
enhanced I/M is modeled as if it would be the preferred
"discretionary control measure" adopted by urban areas to attain
the standard under the EPA policy.
Total CO costs for the Mitchell bill, the Waxman bill, and
the Group of Nine Proposal are similar in magnitude. The cost
burden is distributed differently for each legislative approach,
however. The Mitchell bill places more of the cost burden on
stationary sources. The Group of Nine proposal costs affect only
motor vehicles.
All of the policies/bills have additional I/M costs. These
costs can include improving the effectiveness of existing I/M
programs and establishing new I/M programs in areas where they
currently do not exist. Both the Mitchell bill and the Group of
Nine proposal have alternative; fuel programs in severe CO
nonattainment areas. These programs are estimated to cost $27
mil. lion. The alternative fuels case modeled is a CO season
>: i
-------
Table III
Ozone Nonattainment Control Cost Summary*
Estimated New Expenditures (Billion $)

1995
2000

EPA Policy
4.2- 8.6
8.9 -
18.3
Mitchell Bill
8.3 - 15.5
14.7
26. 5
Waxman Bill
7.8 - 11.5
11.0 -
24 . 0
Group of Nine Proposal
5.8 — 6.3
8.5 -
14.8
* Ranges of costs reflect costs of controlling residual tons
using a range of $2,000 to $10,000 per ton. Costs of pre-1988
policy requirements are not included here but can be found in
Chapter V.
-------
Table IV
Additional Carbon Monoxide Control Costs*
1995 Projection Year
(millions)
Policies/Bills
Control	EPA	Mitchell	Waxman	Group of Nine
Measures	Policy	Bill	Bill	Proposal
Motor Vehicle Measures
Enhanced I/M	$38	$67	$128	$132
Alternative Fuels**	—	27	—	27
Stationary Source	0	40	0	0
Controls
Emission Fee		0	34	13	0
Totals	$38	$168	$141	$159
* Costs are those in addition to what is estimated to be spent to
comply with ozone provisions.
** The alternative fuels case modeled is a CO season (winter)
switch from straight gasoline to an ethanol blend.
Note: Effects of cold start certification testing for motor
vehicles have not been included in this analysis.
-------
(winter) switch from straight gasoline to an ethanol blend. This
program is similar to the one currently being used in the Front
Range of Colorado.
The CO stationary source controls called for by the Mitchell
bill are estimated to cost $40 million. These are the costs of
applying the control techniques listed in Table III.1 to serious
and severe CO nonattainment areas.
Stationary source emission fees of $100 per ton are applied
in both the Mitchell and Waxman bills. Costs are higher for the
Mitchell bill because the fee is applied in both serious and
severe nonattainment areas. The Waxman bill only has an emission
fee for sources in severe nonattainment areas.
Estimates of expected attainment dates depend on which
source types are assumed to be contributing to observed CO
standard exceedances. With the assumption that mobile sources
and a percentage of stationary area sources (20 percent) affect
the design value monitor, there are three residual CO
nonattainment areas in 1995 in the simulations for the proposed
EPA policy and the Waxman bill. The Mitchell bill and Group of
Nine proposal simulations show one remaining CO nonatta inment.
area in 1995. If all sources within an MSA are assumed to
contribute equally to CO standard exceedances, many more areas
are projected to fail to attain the standard by 1995.
Note also that M0BILE3 modeled CO I/M credits are higher
than what has been observed in recent surveys (Sierra Research,
1988). If I/M programs are less successful than indicated by
M0BILE3, the number of remaining CO nonattainment areas in 1995
will increase.
>;j v
-------
CONTENTS
Page
I INTRODUCTION 			1
A.	BACKGROUND AND PURPOSE	1
B.	ORGANIZATION OF THE REPORT	2
II VOC MODEL DEVELOPMENT	4
A.	MODELING OBJECTIVES 			4
B.	MODEL OVERVIEW	5
1.	Definitions			5
2.	Modeling Methods Summary 		6
C.	1985 EMISSION INVENTORY 				11
D.	CONTROL COST EQUATIONS	'	18
1.	Point Sources			20
2.	Area Sources			24
3.	Motor Vehicles	29
E.	GROWTH PROJECTIONS			34
F.	ESTIMATING EMISSION REDUCTIONS AND COSTS	4 3
1.	SIP Regulations. . 			4 3
2.	NSPS File			4 3
3.	Expansion of Nonattainment Areas and Ozone
Transport Region Controls 	 . 		45
4.	Scenario Control Measures			45
5.	National Control Measures. .... 		47
6.	Miscellaneous Measures 		47
G.	RESULTS REPORTING 		4 7
III CARBON MONOXIDE (CO) MODEL DEVELOPMENT 		54
A.	1985 EMISSION INVENTORIES 	 .....	56
B.	CONTROL COST EQUATIONS	56
1.	Stationary Sources			57
2.	Area/Mobile Sources		6 3
C.	GROWTH PROJECTIONS	66
1.	Motor Vehicles			.67
2.	Area Sources	67
IV NOx COST ESTIMATES		 . 70
A.	EMISSION INVENTORY			70
B.	N0X CONTROL COST EQUATIONS.	72
C.	USE OF DEFAULT VALUES	.74
V EPA NONATTAINMENT POLICY ANALYSIS	79
VI MITCHELL BILL ANALYSIS	99
VII WAXMAN BILL ANALYSIS	.106
VIII GROUP OF MINE PROPOSAL AHA I-VP IS	110
IX SUMMARY OF RESULTS	116
A.	OZONE NONATTAINMENT	116
B.	CARBON MONOXIDE NONATTAINMENT	135
-------
CONTENTS (continued)
Page
X	SENSITIVITY ANALYSIS	139
XI	CAVEATS 		14 7
ABBREVIATIONS AND ACRONYMS	150
REFERENCES			152
APPENDIX A	157
A.	RACT LEVEL CONTROL EQUATIONS	157
1.	Industrial and Utility Boilers	157
2.	Internal Combustion Engines	"	168
3.	Gas Turbines			169
4.	Process Heaters		170
B.	BACT LEVEL CONTROL EQUATIONS	170
1.	Industrial and Utility Boilers	170
2.	Internal Combustion Engines .... 	 .172
3.	Gas Turbines			 .17 3
4.	Process Heaters	175
-------
TABLES
Numbers	Page
11.1	Basic Structure of Emission Data Base. ........	12
11.2	Changes in Texas Control Levels by Pod and Control
Device			15
11.3	Texas Miscellaneous Point Source (Pod 90) VOC Control
Efficiency Changes 		16
II. 4 National Summary of 1985 VOC Emissions	19
II. 5 Sample of Cost Equation Input File	21
11.6	NEDS SCCs Added to Cost Pods			2 3
11.7	Consumer Product Subcategories Ranked in Order of
Average Total Emissions (for California) 		26
II. 8 Cost Pods and Control Options. . 		35
11.9	VOC Model Industrial Categories			39
11.10	Earnings Projections by Industry: U.S. Totals ....	40
11.11	Current and Projected Nationwide Vehicle Miles
Traveled by Year and Vehicle Type			41
11.12	Motor Vehicle Emission Factors by Year and Control
Option . . . 				4 2
11.33 Number of Motor Vehicles by Year and Vehicle Type. . .	44
11.14	Ozone Attainment Categories	46
11.15	Attainment Category/Pod Report	.49
11.16	State/Industry Category Report 	 ...	50
11.17	MSA Summary Report			51
11.18	Ozone Design Values and Emission Reduction
Requirements 		52
III. 1. Carbon Monoxide Control Cost Equations for Retrofit
Applications 		58
III.2 Carbon Monoxide Control Cost Equot ions lor New
Applications			59
xv i i
-------
TABLES (continued)
Numbers	Page
III.3 Applicable SCC Codes for Stationary Source CO
Categories			61
III. 4 CO Reductions for I/M Programs			65
111.5	Motor Vehicle CO Emission Factors 	 ..... 68
111.6	Total Population Estimates by Projection Year 	 69
IV.1 N0X Control Cost Equations for Utility and Industrial
Boilers	75
IV.2 N0X Control Cost Equations for IC Engines, Gas
Turbines, and Process Heaters 	 .76
IV. 3 Default Cost per Ton Values for N0X Emitters	77
V.l Key EPA Ozone Policy Provisions	80
V.2 Costs of Pre-1988 EPA Ozone Policy: National Summary. .	81
V.3 EPA Ozone Policy Costs: National Summary 		82
V.4 Discretionary Controls for EPA Policy 		84
V.5 Summary of VOC Emissions by Category for Ozone
Nonattainment Areas 				86
V.6 Transportation Control Measures Analyzed for Kansas
City	88
V. 7 Additional Reductions Needed to Meet Att.ainment/3
Percent Line 1995 EPA Policy: VOC Emissions	9 3
V.8 Additional Reductions Needed to Meet Attainment/3
Percent Line 2000 EPA Policy: VOC Emissions	96
VI.1 Outline of Key Mitchell Bill Provisions ....... .100
VI. 2 Mitchell Bill National Measures and New CTGs	101
VI.3 Incremental Cost of Mitchell Bill: National
Summary	10 3
VII.1 Outline of Key Waxman Bill Provisions . 		107
VII.2 Incremental Cost of Waxman Bill: National
Summary	108
VIII. 1 Outline of Group of Nine Proposal Provisions	Ill
>:v i i i
-------
TABLES (continued)
Numbers	Page
VIII. 2 Group of Nine Proposal Motor Vehicle Emission
Standards	112
VIII.3 Group of Nine Proposal Attainment/Nonattainment
Categories			,113
VIII. 4 Incremental Cost of Group of Nine Proposal: National
Summary	114
IX.1 Attainment/Progress Requirements of Proposals
Analyzed			117
IX.2 Residual Ozone Nonattainment Areas by Projection
Year. . 				119
IX.3 New Versus Existing Stationary Source Costs	121
IX.4 Ozone and CO Nonattainment Control Cost Summary by
CMS A : 1995 	 123
IX.5 Ozone and CO Nonattainment Control Cost Summary by
CMSA: 2000 	 127
IX.6 Ozone and CO Nonattai nment Control Cost Summary by
State: 1995	 131
IX,7 Ozone and CO Nonattainment Control Cost Summary by
State: 2000			13 3
IX.8 Additional Carbon Monoxide Control Costs. ...... .136
X.1 National Average Growth Rates by Industrial Category
Used in Sensitivity Analysis			140
X.2 Annual VMT by Vehicle Class and Year			.141
X - 3 Nonattainment Area VOC Emissions by Alternative
Growth	143
A.1 NOx Control Cost Equations for Utility and Industrial
Boilers	158
A.2 NQX Control Cost Equations for IC Engines, Gas
Turbines, and Process Heaters	159
A.3 SCC Codes Corresponding to NOy Control Cost
Equations	.160
-------
FIGURES
Numbers	Page
II.1 ERCAM-VOC Flowchart. ... 	 7
III.l Carbon Monoxide Model Organization . . 	 55
IV.1 1985 Point Source MOx Emissions by Source Type:
Attainment Area vs. Nonattainment Area	71
IV.2 1985 Point Source N0X Emissions by Ozone Nonattainment
Area Severity			71
IX.1 Ozone Nonattainment Control Cost Summary	118
X.1 VOC Emission Differences for Alternative Growth	144
X.2 National Cost Differences for Growth Analyses. .... .145
xx
-------
I INTRODUCTION
A. BACKGROUND AND PURPOSE
A substantial number of volatile organic compound (VOC) and
carbon monoxide {CO) control measures, especially in the large
metropolitan areas of the United States, have been imposed since
the Clean Air Act was passed. Despite this, the Clean Air Act
mandated deadline of December 31, 1987, elapsed with a long list
of areas still not attaining national ambient air quality
standards (NAAQS) for ozone and CO. In anticipation of this
shortfall, the U.S. Environmental Protection Agency (EPA)
developed a program to address the likelihood that many areas
would not attain the NAAQS, publishing the proposed policy in the
Federal Register November 17, 1987. This announcement prompted
much interest among state and local air pollution control
agencies and at EPA to determine what effect this new policy
might have on the remaining nonattainment areas.
The introduction of bills in Congress in 1987 and 1988 to
address some of the same issues included in the EPA nonattainment
policy prompted calls for quantitative analyses of each of the
Congressional bills and proposals as well as the EPA policy.
Initial interest among the Congressional alternatives focused on
S. 1894, introduced by Senator George Mitchell and otherwise
known as the Mitchell bill. In the House, a bill introduced by
Rep. Henry Waxman (H.R. 3054) presented some alternative
nonattainment provisions. These were followed by another
proposal formulated by nine Democratic Congressmen, which has
come to be known as the Group-of-Nine Proposal. This report
presents a quantitative assessment of the control costs and
emission reductions that might be expected from each of these
three Congressional alternatives and compares them with what
would be expected to happen under the EPA policy.
In reviewing analytic tools available for performing an
analysis of VOC, oxides of nitrogen (N0X), and CO control costs
for different ozone and CO nonattainment control approaches, it
was found that no single model was available for any of the three
1
-------
pollutants that could perform all of the required analyses in a
timely fashion. Therefore, new models and analysis tools were
developed to meet the specific objectives of this project. These
models and tools vary in complexity, with the NO>; analysis being
the simplest, VOC the most complex, and CO somewhere in between.
Effort in the N0X analysis focused on developing current cost
equations for Reasonably Available Control Technology (RACT) (low
N0X burner) and Best Available Control Technology (BACT)
(selective catalytic reduction (SCR)). These cost equations were
then applied to sources in the 1985 National Emissions Data
System (NEDS) emission inventory to estimate costs of various
bill provisions. For VOC, a more complete model was developed
that included current control information, control cost
equations, and new source growth emission projections. Scenario
files were designed to allow different levels of VOC controls in
areas depending on the severity of their nonattainment problem.
Results can be provided by Metropolitan Statistical Area (MSA),
by state and industry, and by source category for the entire
United States.
The CO model is similar in design to the VOC model, but it
was a much simpler model to construct and operate because of the
limited number of important CO source categories and control
options. New source growth and control was incorporated into the
CO model for motor vehicles and other area source emitters; point
source controls and costs were evaluated for only the existing
set of sources.
B. ORGANIZATION OF THE REPORT
The organization of this report is such that modeling
methods are presented first for each of the three pollutants
(VOC, CO, and N0X) followed by results for the EPA policy and the
three legislative alternatives. Sensitivity analyses and a list
of study caveats are provided following the results chapters.
With most of the attention in this study on costs to attain
the ozone standard, much effort was spent developing a VOC
control cost model. VOC model input data and calculation
-------
procedures are described in Chapter II. Discussed in this
chapter are the base year emission inventories, control cost
equations, growth projections, emission constraints, and results
reporting. Chapter III presents similar information for the CO
model. Although a model was not developed for the NQX control
cost analysis, organization of the N0X emission data base and
development of N0X control cost equations are explained in
Chapter IV.
Chapters V through VIII present analysis results for the EPA
policy, Mitchell bill, Waxman bill, and the Group of Nine
Proposal, respectively. Summary results for all policies/bills
are presented in Chapter IX. Because CO control costs are much
lower than those for VOC and N0X, they are only reported in
Chapter IX. Because results are sensitive to the growth rates
used in the VOC model, a sensitivity analysis was performed. The
results of this analysis are presented in Chapter X. Key
analysis caveats are delineated in Chapter XI.
-------
II VOC MODEL DEVELOPMENT
Photochemical oxidants (measured as ozone) are products of
atmospheric reactions involving organic pollutants, nitrogen
oxides, oxygen, and sunlight. All of the evidence presently
available shows that in and around urban centers that have severe
ozone pollution, anthropogenic organics and N0X are the major
contributors. The photochemical formation of ozone is the result
of two coupled processes: (1) a physical process involving
dispersion and transport of precursors to ozone (e.g., organics
and N0X), and (2) the photochemical reaction process. Both
processes are strongly influenced by meteorological factors such
as dispersion, solar radiation, temperature, and humidity.
The Emission Reduction and Cost Analysis Model for VOC
(ERCAM-VOC) described in this chapter was developed to analyze
alternative measures for reducing emissions of VOCs, precursor to
ozone. The model runs on a personal computer and is programmed
in dBase III Plus. It is designed to simulate the process that
states and metropolitan areas might use to move toward attainment
of the ambient ozone standard (the ozone MAAQS) under alternative
policies.
ERCAM-VOC does not include an air quality modeling
component. Instead, it uses as input VOC emission reduction
targets estimated from an ozone trajectory model.
A. MODELING OBJECTIVES
The major objectives in developing the model were that it be
a PC based model that can be used by other parties, that it
provide quick turnaround analyses, that it report results on
various geographical levels (national, state, and MSA), and that
control selection be exogenous to the model.
While; the first objective (use by other parties) has not yet
been realized, once the model is documented and some of the
computer code developed to respond to quick turnaround issues is
reprogrammed, PC users familiar with dBase III should be able to
run ERCAM-VOC. While a normal disadvantage of a PC based model
4
-------
maintenance (I/M) programs (U.S. EPA, 1987d). The regulations
are specified by pod/control strategy combinations. This pair is
then matched to the cost equation file to obtain emission
reduction and cost information. In cases where the specified
reduction is more stringent than the existing level of control
for a source as specified in NEDS, emissions are reduced to the
level specified and the additional control cost is estimated.
Scenario control strategies are next applied to existing
source emissions. A separate file is used for existing and new
source scenario constraints on emissions. For simplicity and
because most of the proposed new VOC control provisions are
stipulated by nonattainment severity, the scenario file is
organized by attainment category. Attainment areas are one
classification while nonattainment areas are divided into
separate classes depending on nonattainment severity. For each
attainment classification and pod, a control strategy (including
zero control) is specified in an input file. In simplified form,
for sources in the data file not already controlled to the level
indicated (after applying regional constraints), scenario level
emissions and costs are calculated.
There are two exceptions to this form. A penetration factor
(the percentage of sources in a cost pod affected by a
regulation) is used for solvent evaporation area sources to
reflect the fact that some sources are small and may be exempt
from control by the chosen technique because of their size. This
factor is used to calculate the resulting control level for the
pod. The second exception is for point source RACT controls.
These include controls on all point sources for which Control
Technique Guidelines (CTGs) have not been published. For these
sources, a size cutoff is specified. Sources under the size
cutoff and sources which are already controlled are not subject
to the scenario file constraint.
Additional scenario control measures outside of the
attainment classification and pod framework may also be analyzed
using the model. Gasoline Reid Vapor Pressure (RVP) reductions,
onboard controls, and new motor vehicle emission standards are
9
-------
modeled on the national level. Emission fees are applied to
sources above a size cutoff by attainment classification.
Expansion of nonattainment areas to the Consolidated Metropolitan
Statistical Area (CMSA)/MSA level involves applying SIP
regulations to the entire CMSA/MSA rather than sped f ic counties.
Ozone transport region controls are specific measures mandated
for regions believed to significantly impact the ozone
concentrations in surrounding areas. Costs are estimated for
controls that might be mandated in these areas, but no credit is
given to areas downwind whose emissions may not have to be
reduced as much because transported ozone is less. -
New source emissions (except motor vehicle) are projected
using Bureau of Economic Analysis (BEA) growth rates by industry
and MSA (BEA, 1985). Growth factors are applied to 1985
uncontrolled VOC emissions for each MSA/state/pod/industry
category combination to estimate future year uncontrolled VOC
emissions. Future year motor vehicle emissions are estimated
using national average vehicle miles traveled (VMT) growth (EEA,
1987) and changes in future year emission factors (Lorang, 1988;
U.S. EPA, 1984 and 1987a).
After calculating new growth emissions, NSPS (Battye et al.,
1987) and SIP constraints are applied. NSPS regulations apply to
all areas and are designated by pod and control strategy. The
methodology in applying the constraints is the same as for
existing source constraints except that average cost per ton
values are used in estimating control costs since source size
specific information is not determined. The average cost per ton
values for each cost pod were estimated by applying the cost
equation to the average sized new source (Battye et al., 1987).
After applying NSPS and SIP constraints, new sources are
then subject to possible further emission reductions via the
scenario constraints. Again, this calculation parallels that of
the existing sources except that the dollar per ton values are
used to estimate control costs.
Results reporting is on three geographical levels:
national, state, and MSA. National level results are provided by
l o
-------
is that runtime is much longer than it would be for a comparable
mainframe model, an ERCAM-VOC national simulation runs in 2 hours
on a PC with an 80286 microprocessor using the dBase compiler
Clipper, which significantly reduces runtime.
The objective of providing quick turnaround analyses was
realized, as the model has been used to provide EPA with analyses
of proposed policy and bills within a few days. Model results
are available by source category and attainment category on the
national level. State level results are reported by industry.
Emission and cost totals may be reported by MSA. ("Sources are
categorized by pollution and control characteristics and several
source categories may exist within an industry.)
ERCAM-VOC was developed to analyze regulatory alternatives
for attaining the ozone standard. It was designed so that
combinations of measures are selected for analysis rather than
having the model select the most cost efficient set of measures.
An endogenous control measure selection is less desirable in a
situation where all available controls (or more) are necessary to
meet control targets. Also, some of the provisions of the
proposed alternatives in this study specify controls that must be
used in areas according to nonattainment severity, thus it is
necessary to design a model in which controls are specified and
analyzed according to various attainment classifications rather
than as a least cost analysis. (Controls mandated may not always
be the most cost effective or even necessary for a specific
area.)
B. MODEL OVERVIEW
1. Definitions
Two terms used frequently in the text need to be well
understood before reading the VOC model discussion. These terms
are "cost pod" and "design value."
Control and cost information for the model is organized by
cost pod. A cost pod is a group of source types, as defined by
NEDS Source Classification Codes (SCCs) or NEDS area source
categories, which have sirtiilar emission characteristics, control
-------
techniques, and control costs. A cost pod may have one or
several control strategies (which consist of control option,
efficiency, and cost information). All sources inventoried are
classified into cost pods. In all, the model comprises 42 point
source and 23 area source cost pods. All of the emission
reduction and control cost calculations are performed at the cost
pod level.
The ozone design value is a measure of the maximum ozone
concentration expected to occur within an area. Any area with a
design value of 0.125 ppm or greater is considered nonattainment.
This means that an area is expected to exceed the 0-. 12 ppm
standard more than once per year on average. In general, the
higher the design value the greater the VOC reduction requirement
to reach attainment (although this varies depending on the amount
of WQX present and other factors). Design values are important
in this analysis both for determining VOC emission reduction
requirements and for classifying areas into attainment and
nonatta inment categories.
2. Modeling Methods Summary
Six primary inputs are used to estimate the effect of
current and future regulations on VOC emissions and costs:
.	1985 VOC emission inventory,
.	existing source regulations,
.	scenario control strategies,
.	set of control cost equations,
.	set of growth factors, and
.	NSPS regulations.
The interrelationship of these inputs is diagrammed in Figure
II.1 and a brief description follows. A more detailed
description is contained in the following sections.
The 1985 VOC emission inventory is taken from the 1985 NEDS
point and area source inventories. From the point source
inventory, source specific information was retained for sources
emitting more than 50 tons per year. Smaller sources were
aggregated by MSA/state region and cost pod into 0 to 25 ton per
year sources and 25 to 50 ton per year sources. Changes were
made to the control efficiencies for combustion sources and
sources within the state of Texas because they were believed to
6
-------
attainment category and cost pod. State level results are
reported by industry. This level of information may be useful
for an economic analysis. Costs and emissions totals can be
provided for each MSA. This output is useful in determining
whether necessary emission reductions toward achievement of the
ozone standard are being met and in estimating what additional
costs might have to be incurred to make adequate progress toward
compliance.
C. 1985 EMISSION INVENTORY
The base year emissions data file was developed from the
1985 NEDS point and area source files. The 1985 NEDS Emissions
Inventory was selected for use in this study because it was the
product of a multi-year research effort to develop an accurate
and comprehensive inventory of 1985 emissions from sources
thought to be important in acid deposition processes. Therefore,
quality control procedures for this inventory were much more
rigorous than those employed in other NEDS data files. The 1985
inventory was also selected because it matches the time period of
the air quality data used in the study.
The data elements in the emissions data base are outlined in
Table II.1. The state, SCO and controlled emission levels are
taken directly from the NEDS emission inventory. The
uncontrolled emissions were calculated based on the controlled
emissions and reported control efficiency. MSA is a four digit
code referring to the metropolitan statistical area and is based
on the state and county. The pod indicates the source types for
emission reduction and costing purposes. The industrial category
code is a two-digit grouping based on the Standard Industrial
Classification (SIC) code for point sources and an assigned two
digit code for area sources. The attainment category is used for
modeling purposes and is determined from the ozone design value
for an area. This element is analysis specific. The number of
vehicles is used for motor vehicle control cost estimates (all
other costs are based on uncontrolled emissions). A set of SIP
regulations are available for each MSA, but the applicability of
11
-------
Table II.l
Basic Structure of Emission Data Base
Element Name
ATTCAT
AEROSSTATE
MSA
POD
SQQ
SCC
NVOC
UVOC
NUMVEH
SPSW
Description
Attainment category (ozone design value
dependent)
AEROS State code
Metropolitan Statistical Area code
Cost Pod Identifier
Industry identifier
NEDS SCC code*
1985 NEDS VOC emissions
1985 uncontrolled VOC emissions
Number of motor vehicles
SIP switch
~includes codes for additional sources not covered by NEDS
** indicates whether sources within an MSA are located in
counties with existing SIP regulations
12
-------
these regulations may not extend to all counties within the MSA.
The SIP switch indicates whether the emissions are from sources
subject to or exempt from the set of SIP regulations for the MSA.
Other elements are added to the data base to hold calculation
results.
Source specific information was retained for sources shown
as emitting 50 tons per year or more in the 1985 NEDS point
source inventory. Smaller sources were aggregated by cost pod,
industrial category, and MSA/state region code into two types of
records. The first type includes sources emitting 25 to 50 tons
per year. The second type includes sources emitting less than 2 5
tons per year. Coverage of smaller sources within the 1985 NEDS
point source inventory is limited since attention was focused on
100 ton-per-year emitters. Emissions from smaller stationary
sources are represented by area source categories.
Two changes were made to point source records in the base
year emissions data file. First, many combustion sources (zero
pod) had reported VOC efficiencies, many over 90 percent. Since
fuel combustion is in itself, a VOC control, it seems unlikely
that additional controls are put on these sources. Since future
emissions are projected using uncontrolled emissions, this
produced high future emissions and no control techniques were
available for reducing these emissions. The uncontrolled
emissions were set equal to the controlled emissions (simulating
no control devices) for ail combustion point sources to keep
growth within a reasonable bound.
In early runs of the model, some areas in Texas showed
higher VOC emissions in the forecast years than in 1985, even
with all available controls being added. This occurred because
current VOC control efficiencies for many point sources were
unrealistically high. Texas Air Control Board point source
emission inventory surveys do not ask about control efficiencies,
only emissions and control equipment data are collected. Then,
Texas assigns default control efficiencies for sources based on
control equipment type and SCC. These control efficiencies were
1 3
-------
higher than the maximum reported in other areas for the same type
of process and control equipment in many cases.
Uncontrolled emissions are based on the reported controlled
emissions and the control device efficiency. If the efficiency
is overestimated, uncontrolled emissions will also be
overestimated. For example, a 100 ton per year source with a 98
percent efficient control device will have uncontrolled emissions
of 5,000 tons per year. If the control equipment efficiency is
overestimated at 99 percent, uncontrolled emissions will be
estimated as 10,000 tons per year. If the efficiency is
overestimated to 99.9 percent, uncontrolled emissions will be
100,000 tons per year. If growth of 3 percent per year is
applied for 10 years, new emissions assuming the 99.9 percent
control efficiency will be 34,400 tons. When using the correct
efficiency of 98 percent, new growth will be only 1,700 tons.
Texas control efficiencies were adjusted to more reasonable
levels as outlined in Table II.2 and Table II.3.
New control efficiencies for Texas were established in two
ways. First, for cost pods specific to industries, i.e., pods 2
through 36, control levels were set equal to either the level of
control achieved by similar pods in other nonattainment areas or
to the level of control specified in the Texas SIP. These
changes are documented in Table II.2. For sources in pod 90
(miscellaneous sources) a different approach was used. Each
source or source/control device combination was evaluated
separately and engineering judgment was used to establish a
likely limit on device effectiveness. The engineering judgment
was based on both the magnitude and characteristics of the VOC
emissions and the probable effectiveness of the control
equipment. The changes made to these sources are documented in
Table II.3.
Changes were also made to the NEDS area source inventory.
First, the solvent evaporation and gasoline marketing categories
were separated into smaller categories for control purposes. The
solvent evaporation emissions are apportioned among eight
14
-------
Pod
2
4
5
7
10
17
20
22
23
24
25
26
Table II.2
Changes in Texas Control Levels by Pod and Control Device
Source Category
Name
Control Device
Number
of
Sources
NEDS
Control
<*>
Revised
Control
<*>
Printing/publishing
19-Catalytic Afterburner
1
99
0
86
Fixed roof crude tanks
47-Vapor Recovery
3
99
5
98
Fixed roof gasoline tanks
47-Vapor Recovery
10
99
0
96
Floating roof gasoline tanks
47-Vapor Recovery
1
99
0
95
Bulk gasoline terminals
51 -Gas Absorption
1
99
0
90
Terephthalic acid mfg.
23-Flaring
1
99
9
98
Refinery fugitives
13-Gas Scrubber
2
99
0
70

23-Flaring
1
99
9
70

47-Vapor Recovery
3
99
0
70

48-Carbon Absorption
1
99
9
70
Styrene-butadiene rubber
52-Spray Tower
1
99
9
98

47-Vapor Recovery
1
99
0
94
Polypropylene mfg.
47-Vapor Recovery
1
99
0
98
Polyethylene mfg.
47-Vapor Recovery
1
99
0
98
Ethylene mfg.
47-Vapor Recovery
6
99
0
98
Refinery W treatment
47-Vapor Recovery
1
99
0
95
-------
Table II.3
Texas Miscellaneous Point Source (Pod 90)
VOC Control Efficiency Changes

Control Device
1985 NEDS

see
Code/Description
X Control
Change
30100104
13-Gas Scrubber
99.0
Delete
30100799
47-Vapor Recovery
99.0

30101199
47-Vapor Recovery
99.9
Delete
30101801
50-Gas Absorption
99.0

30103399
13-Gas Scrubber
99.0

30103399
47-Vapor Recovery
99.9
99
30109101
48-Carbon Adsorption
99.0

30109153
13-Gas Scrubber
99.0

30112001
53-Venturi Scrubber
99.0
98
30112005
47-Vapor Recovery
99.0

30112011
13-Gas Scrubber
99.0
95
30112599
52-Spray Tower
99.0
98
30115380
47-Vapor Recovery
99.0
90
30117613
47-Vapor Recovery
99.9
99
30118105
52-Spray Tower
99.9
99
30125001
60-Proc. Gas Recover
99.9
95
30125099
47-Vapor Recovery
99.0

30125099
23-Flaring
99.9
98
30199999
47-Vapor Recovery
99.9
90
30199999
53-Venturi Scrubber
99.9
90
30300399
23-Flaring
99.9

30699999
13-Gas Scrubber
99.0

30699999
47-Vapor Recovery
99.0

30699999
48-Carbon Adsorption
99.9
90
40400204
46-Process Change
90.0

40400250
47-Vapor Recovery
99.0
90
40688801
47-Vapor Recovery
99.0
95
40688801
23-Flaring
' 9
95
40700816
47-Vapor Recovery
0
95
40782001
23-Flarlog
4 9

40899999
53-Venturi Scrubber
9: 0

50390006
19 -Af terburner
99.0
Delete
-------
categories based on national average factors (Battye et al.,
1987).	The breakdown of this category is as follows:
.	architectural surface coating (12,1 percent),
.	paper surface coating (4.2 percent),
.	degreasing (17.4 percent),
.	dry cleaning (7.6 percent),
.	printing (4.4 percent),
.	rubber and plastics manufacture (3.3 percent),
.	commercial and consumer solvents (25.7 percent), and
.	miscellaneous solvent use (25.3 percent).
Gasoline marketing was divided into underground tank evaporative
losses, or stage I "(38.4 percent), and self service refueling
losses, or stage II, plus spillage (61.6 percent), based on the
emission factors used to estimate gasoline marketing emissions in
the 1985 NEDS area source inventory (Kimbrough, 1988). Emissions
are estimated by multiplying gasoline throughput by the emission
factors for stage I, stage II, and spillage. The ratio of the
emission factors will equal the ratio of emissions for the three
emission sources.
NEDS solvent evaporation emission estimates are based on the
solvent usage for each county. It is assumed that all solvent
used is emitted to the atmosphere. Many areas have SIPs
regulating solvent evaporation sources in such a way that the
solvent is destroyed (i.e., incineration). If this is the case
for an area, solvent evaporation emissions will be overestimated
by NEDS. Therefore, adjustments were made to solvent evaporation
emissions for nonattainment areas with SIP regulations for those
sources to better reflect the actual emissions in 1985 (Johnson,
1988).
Emissions for gasoline marketing, stage I and stage II, are
estimated in NEDS using uncontrolled emission factors. Many
current state regulati. require the use of stage I controls
(usually on sources above 120,000 gallons per year throughput)
which control emissions by 95 percent. Emissions in areas
designated as having stage I controls (Battye, 1987) were
adjusted to reflect these controls. Stage II controls are
already in place in Los Angeles, San Francisco, and the District
17
-------
of Columbia. Base year stage II emission estimates were adjusted
to the estimated control level of 86 percent.
Motor vehicle emissions for 1985 are estimated by county in
NEDS and include the effects of I/M programs. NEDS uses MOBILE3
to estimate 1985 highway vehicle emissions for four vehicle
types: light-duty gasoline-powered vehicles (LDGVs), light-duty
gasoline-powered trucks (LDGTs), heavy-duty gasoline-powered
vehicles (HDGVs), and heavy-duty diesel-powered vehicles (HDDVs).
Reductions in VOC (and other pollutant) emissions that should be
observed in areas with I/M programs were also simulated in NEDS
using M0BILE3, with M0BILE3 program inputs being determined by
summaries of I/M program characteristics by area provided by
EPA's Office of Mobile Sources. It is believed that the M0BILE3
credit calculated for I/M programs is overestimated (Sierra
Research, 1988). Emissions were adjusted for areas having I/M
based on national averages. Emissions were adjusted from a 22
percent credit to a 15 percent credit for basic I/M.
The number of vehicles for each MSA/state/pod combination
were also added to records for motor vehicle pods. National
totals of vehicle registrations by vehicle type for 1985 (EEA,
1987) were apportioned to individual MSA/state combinations by
population (U.S. Bureau of the Census, 1985). The number of
vehicles is used to estimate costs for many motor vehicle control
options.
A summary of the ERCAM VOC emissions data base is shown in
Table II.4. Included are the VOC emissions and the 1985 level of
control by source category.
D. CONTROL COST EQUATIONS
The starting point for the control cost equation data file
is the set of equations developed for the ozone NAAQS Cost Model
by Alliance (Battye et al., 1987). These include equations for
34 point source and 7 area'source categories (cost pods). Cost
equations were developed from model plant data using linear and
exponential least squares curve fitting techniques.
IS
-------
Table II.5
Sample of Cost Equation Input File



¦
Capi tal
Capital


Recovery Recovery



Reduction
Cost
Cos t
0&M
0&M
Credi t Credi t
Pod
CS
Pod Name
CS Name (X) Coefficient
Exponent
Coefficient
Exponent
Intercept Slope Cost*
21
0
Cell. Acetate
No Control 0 0
SESSSSSZSSS
0
:3SZSSS SltS £3 2 SSS ££ SI
0.0
0.0
3SS33SSSZS:Z=;33atSS3SSS = Z= = '
0 0 0
21
1
Cell. Acetate
Carbon Adsorber 54
90,809
0.60
10614.0
0.600
0 320 537.0
21
2
Cell. Acetate
Carbon Adsorber 72
115,789
0.60
12110.0
0.600
0 448 579.8
* COST = $/uncontrolled ton for stationary sources
= $/vehicle for motor vehicles
fcfcTES:
Equations use uncontrolled VOC emissions as independent variable
Capital and O&M cost equations are exponential
Recovery credit equations are linear
-------
Table II.5
Sample of Cost Equation Input File
Pod CS Pod Name
CS Name
0	Cell. Acetate No Control
1	Cell. Acetate Carbon Adsorber
2	Cell. Acetate Carbon Adsorber
Reduction
(X)
0
54
72
Capi tal
Cost
Coefficient
Capi tal
Cost
Exponent
O&M
Coefficient
O&M
Exponent
90,
115,
0
809
789
0
60
60
0,
10614.
12110.
0.0
.600
.600
Recovery
Credi t
Intercept
0
0
0
Recovery
Credi t
Slope Cost*
0
320
448
537
579
0
.0
.8
* COST = $/uncontrolled ton for stationary sources
= $/vehicle for motor vehicles
ru
W)TES:
Equations use uncontrolled V0C emissions as independent variable
Capital and 0&M cost equations are exponential
Recovery credit equations are linear
-------
Table II.6
NEDS SCCs Added to Cost Pods
Pod/Source Type
NEDS
SCCs
1-	Degreasing
2-	Printing and Publishing
4-Fixed	Roof Tank, Crude Oil
5-Fixed	Roof Tank, Gasoline
6-Floating	Roof Tank, Crude Oil
20-Petroleum	Refinery Fugitives
21-Cellulose	Acetate Manufacture
33 —Automobile Surface Coating
35-General Surface Coating
40100306

40500211,
12.
40500311,
12
40500411,
12
40500412,
13
40301011

40301002,
03
40301103,
04
30600811-
20
30688801-
05
30102501

40201601,
06
40201901

2 3
-------
f.	Charcoal Manufacturing
Charcoal manufacturing was identified as a large emitter at
37,200 tons in 1985. EPA1s AP-42 emission factor document (U.S.
EPA, 1985) states that the use of an afterburner can reduce
emissions by an estimated 80 percent. The default cost for
afterburners from the Radian VOCM is $1,685 per ton (Radian,
1985}. While charcoal manufacturing is a large national emitter,
it is not a significant source in nonattainment areas.
g.	Miscellaneous and Combustion Point Sources
No other point source categories were identified as being
large emitters and having readily available control information.
The remaining sources were classified as pod 90, miscellaneous
point sources, except for combustion. Since it is unlikely that
additional VOC controls are placed on combustion sources because
combustion is a VOC control, these were categorized separately as
pod 0. The sources in the miscellaneous pod have an average
level of control around 90 percent. Future growth from this
category is large because of high growth rates and high
uncontrolled VOC emissions in the base year, so future emissions
must be reduced or the growth will offset reductions achieved in
other categories. A default cost of $1,250 per ton reduced was
assigned at a 90 percent control level. This cost was chosen to
represent an average RACT level control cost for sources which
did not involve surface coating.
2. Area Sources
All area sources were also classified into pods. Control
cost equations were developed if information was available. In
addition to the 6 area source pods defined for the NAAQS model
(Battye et al., 1987), 11 new pods were developed including the
pods for the new sources added to the inventory (TSDFs, etc.).
Area fuel combustion sources (except wood stoves) were combined
with point fuel combustion sources.
a. Gasoline Marketing-Stage II
Gasoline marketing-stage II can be controlled using vapor
balance systems, onboard controls, or both. Vapor balance at
24
-------
maximum enforcement can yield reductions of 86 percent at an
estimated cost of $900 per ton reduced (U.S. EPA, 1987a). This
cost estimate is based on applying stage II in 11 nonattainment
areas. The cost per ton varies greatly depending on size cutoffs
(i.e., exempting those emitters with throughputs below a
specified level) and slightly with the number of areas involved.
Costs for nationwide stage II with no size cutoffs can be in
excess of $1,800 per ton reduced. This type of system is already
being used in Los Angeles, San Francisco, and the District of
Columbia. A reduction of 95 percent was assumed when combining
both vapor balance and onboard controls.
b.	Architectural Surface Coating
Emissions from architectural surface coating can be reduced
by reformulating to waterborne coatings. The Federal
Implementation Plan (FIP) study (U.S. EPA, 1987c) estimates an
overall reduction of 52 percent at little or no additional cost.
A draft CTG (U.S. EPA, 1981) estimates an overall reduction of 23
percent at a savings of $1,250 per ton. This means that the cost
for waterborne coatings will be an estimated $1,250 less per ton
of VOC emitted than solvent borne coatings. It is assumed that
many of the coatings yielding large savings have already been
reformulated, so the more recent FIP study information was used
in this analysis. For modeling purposes, a cost of zero dollars
per ton reduced was used.
c.	Commercial and Consumer Solvents
Emissions from commercial and consumer solvent use totaled
1.2 million tons in 1985. These emissions come from a wide
variety of products, each accounting for a small portion of
emissions, forming a large source category when aggregated. A
breakdown of consumer products ranked by average total emissions
in California is shown in Table II.7. Control options for
reducing emissions include product reformulation and banning. A
report on reducing VOC from underarm products (CARB, 1987)
estimates that emissions from these products can be reduced by 54
percent at a cost of $300 per ton reduced. Underarm deodorants
are only a fraction of all consumer solvents, however, so this
25
-------
Table II.7
Consumer Product Subcategories Ranked in Order
of Average Total Emissions (for California)
Total VOC Emissions (tons)
Consumer Product Sub-Category	Per Year in California
Paints, primers, varnishes (aerosols)	11,408
Hair sprays	8,095
All purpose cleaners	6,463
Insect sprays	5,558
Car polishes & waxes	4,625
Room deodorants & disinfectants	4,650
Consumer adhesives	3,830
Caul king & sealing compounds	2,380
Moth control products	2,098
Window & glass cleaners	1,970
Herbicides, fungicides	1,803
Personal deodorants	1,614
Auto antifreezes	1,165
Carburetor & choke cleaners	1,051
Brake cleaners	1,032
Engine degreasers	1,088
Engine starting fluids	949
Rug & upholstery cleaners	930
Lubricants and silicones	913
Metal cleaners & polishes	660
Waxes & polishes	621
Tile & bathroom cleaners	590
Pharmaceuticals	550
Styling mousse	543
Windshield deicer	501
Insect repellents	396
Starch & fabric finish	365
Auto cleaners	354
Floor waxes or polishes	309
Colognes	303
Shaving lathers	271.
Animal insecticides	255
Aftershaves	205
Undercoatings -	188
Oven cleaners	185
Shoe polishes, waxes & colorants	18 3
Paints-other related products	170
Perfumes	135
Spot removers	12 7
Waxes & polishes liquids	9 7
Hair care products - shampoos	89
Carpet deodorizers	69
Suntan lotions	4 1
Depilatories	11
Anti-static sprays		3
68,84 0
Source: U.S. EPA, '1987c
2 6
-------
gives an overall reduction of only 2 percent for all commercial
and consumer products. Since no other control cost information
is available at this time, a default cost of $2,000 per ton
reduced is used for various control levels as specified by the
analyses. It is likely that regulations reducing emissions from
consumer solvents will be in the same form as suggested by
California for underarm products. Emissions may be reduced by
limiting the vapor pressures, relative evaporation rates, or
amount of VOCs in a product. The impact these types of
regulations will have on individual products is difficult to
assess since the formulations may vary widely. Some products may
already meet the standards, some probably can be easily
reformulated, and others would have to be dropped from a
company's product line.
d.	Hazardous Waste Treatment, Storage, and Disposal
Facilities (TSDFs)
Emissions from TSDFs can be controlled using capture and
control techniques such as storage tank covers and carbon
adsorption. It is estimated that emissions can be reduced by 90
percent at a cost of $900 per ton reduced {Bunyard, 1988).
e.	Bakeries
The Bay Area Air Quality Management District has examined
the control of VOC emissions from bakeries. Preliminary results
show that emissions can be reduced by 90 percent via incineration
at an estimated cost of $1,275 per ton reduced (Cutino, 1987).
This cost is based on the control of ethanol from a large bread
baking establishment with five ovens.
f.	Cutback Asphalt
Emissions from cutback (petroleum based) asphalt can be
eliminated by switching to emulsified (water based) asphalts.
The cost difference depends on the price of petroleum and
generally results in a cost savings (U.S. EPA, 1978). A 100
percent reduction at zero cost was used for modeling purposes.
g.	Publicly Owned Treatment Works (POTWs)
It is believed that the most cost effective ways to reduce
emissions from POTWs are those that reduce the VOC content of the
27
-------
industrial wastewater upstream before it reaches the POTW. It is
expected that emissions can be reduced by 90 percent at an
estimated cost of $1,000 per ton reduced (Bunyard, 1988). This
is based on the estimated cost for firebox covers at refinery
wastewater treatment units. The cost will vary depending on the
selected control technique. For example, costs for steam
stripping are expected to be higher than the costs used in this
analysis.
h.	Railroad Engines
Costs for locomotive diesel-engine controls were estimated
using Radian (1988a). This study evaluated both new and existing
engine controls and assumed that the technologies used to control
emissions from other types of diesel engines would be
transferrable. Control techniques for existing engines are
assumed to be applied during a rebuilding process. New engine
controls were evaluated at both intermediate (achievable in 3
years) and advanced (involving further research and development)
levels. Costs and emission reductions applied in this study
represent imposing intermediate technology
emissions standards both on new and existing locomotives.
Emission controls reduce both VOC and NOx. VOC reductions range
from 37.5 percent for new engines to 51.2 percent for existing
engines. The cost effectiveness of these controls depends on
whether VOC and N0X reductions are considered individually or
collectively. N0X emissions are reduced more than VOC emissions,
so if VOC reductions are considered alone in estimating cost
effectiveness, the cost per ton ranges from $19,600 (existing) to
$26,200 (new). Costs used in the model were for VOC plus NOx and
ranged from $1,073 (new) to $1,332 (existing) per ton reduced.
i.	No Available Control Costs
Cost equations have not been developed for the remaining new
area source pods. These pods include (1) off-highway vehicles,
(2) aircraft and vessels, (3) open burning, forest fires, and
prescribed burns, and (4) incineration. It should be noted that
28
-------
emissions from open burning, forest fires, and prescribed burns
are assumed to remain constant with time in ERCAM-voc.
j¦ Miscellaneous Surface Coating
In addition to developing the new area source pods, the cost
data for pod 45, miscellaneous surface coating, was updated based
on information specific to the individual source types in the
category. This category comprises emissions from auto
refinishing, miscellaneous industrial solvent use, and
miscellaneous surface coating. Two control strategies have been
developed for the analysis. The first is the control of
automobile refinishing emissions. Preliminary findings
(Blaszczak, 1988) indicate an overall reduction of 75 percent can
be achieved by using three techniques: an enclosed cabinet to
recycle cleanup solvent, replacement of the application technique
to improve transfer efficiency, and the elimination of lacquers.
All of these options result in a cost savings due to decreased
solvent usage. An overall savings of $3,260 per ton of emissions
reduced can be expected. Based on the percentage breakdown of
the pod into the three emission categories, an overall reduction
of 14 percent of total miscellaneous surface coating emissions
can be achieved by controlling automobile refinishing sources.
The second control strategy modeled combines auto
refinishing control with reductions in industrial solvent use
emissions. These emissions are generally reduced by decreasing
solvent consumption through better working practices. Since no
control cost information was available, a cost of $2,000 per ton
was used. A 2 5 percent reduction of industrial solvent emissions
was used translating to a 10 percent overall reduction for the
pod. Combining this with the automobile refinishing control
opt ion gives an overall reduction of 24 percent at a savings of
$1,070 per ton of emissions reduced.
3. Motor Vehicles
Control strategies and costs were developed for motor
vehicle pods to match the provisions for these sources outlined
in the EPA policy and the bills being studied. The control
strategies modeled include basic I/M, enhanced I/M, a gasoline
7 9
-------
RVP reduction from the current average of 11.5 psi to 9.0 psi,
new motor vehicle emission standards, and alternative fuels.
a.	Inspection and Maintenance
Eiasic I/M is available for reducing emissions from light
duty gasoline vehicles (LDGVs) and light duty gasoline trucks
(LDGTs). The average credit for basic I/M is 15 percent (Sierra
Research, 1988) at a cost of $20.20 per vehicle (U.S. EPA,
1987b) . Enhanced I/M is available for LDGV, LDGT, and heavy-
duty gasoline vehicles (HDGVs). Although cost estimates were
available for heavy duty diesel vehicle (HDDV) I/M, this control
strategy was not used in this analysis as there is no evidence of
achievable emission reductions.
An incremental reduction and cost for enhanced I/M of 7
percent and $6.48 per vehicle over basic I/M is used for LDGV and
LDGT. An emission credit of 13 percent and a cost of $19 per
vehicle is used for HDGV (Lorang, 1985). In a recent APCA paper
(Wright and Klausmeier, 1988), potential emission reductions for
including HDGV in an I/M program were estimated at 8 percent for
1ight-HDGV and 11 percent for medium HDGV for 1988. The exact
reduction depends on the mix of light versus medium HDGV, the mix
of model years, and the VMT of each type.
b.	Reid Vapor Pressure (RVP) Reductions
Emissions from gasoline fueled motor vehicles can be reduced
by reducing the RVP of the gasoline. This option was modeled
assuming a national regulation. Costs per ton of VOC emissions
redacted could be significantly higher if only certain areas
adopted RVP limits. It is expected that decreasing the RVP of
gasoline to 9.0 psi will result in an incremental cost of 0.225
cents per gallon of gasoline (Weiser, 1988). Based on the
average fuel consumption by motor vehicle type derived from
information in the Motor Fuel Consumption Model (EEA, 1987), the
resulting per vehicle annual costs are $1.20 for LDGV, $1.08 for
LDGT, and $2.76 for HDGV. The emission reductions are modeled
through changes in projection year emission factors for motor
vehicles and are discussed in Section II.E, Growth Projections.
3 0
-------
c.	Onboard Controls
Onboard controls can be expected to reduce refueling
emissions by 95 percent but this technique takes time to phase in
due to vehicle fleet turnover. The cost for onboard controls is
attributed to new motor vehicles at an average cost of $14 per
vehicle for gasoline powered motor vehicles (U.S. EPA, 1987b).
The regulatory impact analysis (U.S. EPA, 1987a) of the gasoline
marketing regulation estimated that a little over 50 percent of
consumption would be controlled by onboard controls in 1995
assuming that the controls began in model year 1989. It is
assumed onboard controls will have full impact by the year 2000.
d.	New Motor Vehicle Standards
Motor vehicle emissions can also be reduced by establishing
more stringent new motor vehicle standards. These reductions are
also modeled by adjusting the projection year emission factor.
Expected costs per new motor vehicle are $83.5 for LDGV, $92.4
for LDGT, and $164.7 for HDGV (U.S. EPA, 1987b). The standards
are not expected to reduce VOC emissions from HDGV but are
expected to reduce N0X emissions. Per vehicle costs include the
costs of reducing all applicable pollutants, so they include CO
and NOx control costs, as well as those for VOC. Both CO and N0X
control have been shown to be of benefit in reducing ozone levels
in some areas. These benefits have not been accounted for in
ERCAM-VOC projections of emission reductions needed to reach
attainment.
e. Alternative Fuels
Strategies have been incorporated in the VOC model that
simulate the cost and emission reductions of burning less
polluting fuels in motor vehicles. The costs of these measures
can vary a great deal depending on the assumptions made,
especially for future fuel prices. Two situations are modeled in
the bills that were examined. One is a provision that would
require fleet vehicles (taxis, corporate vehicles) to use less
polluting engines or fuels. There are a number of options
available, but for modeling purposes, it was assumed that fleet
vehicles would meet this requirement by adding a capability to
-------
burn compressed natural gas (CNG). A typical conversion of a
gasoline powered vehicle to natural gas uses two cylinders for
gas storage at a cost of about $2,500 (Flynn, 1985), The payback
period for this conversion depends on the price spread between
natural gas and gasoline. The yearly fuel savings were estimated
using a natura1 gas price of $5.08 per MMBtu and a gasoline price
of $7.70 per MMBtu. The resulting net annual cost per vehicle of
CNG conversion was $55 per year. Fleet conversions to CNG are
assumed to affect LDGVs and LDGTs. Fleet vehicles are assumed to
constitute 5 percent of the total number of vehicles, but 13
percent of the vehicle miles traveled (Lorang, 1988). VOC
emissions from CNG vehicles are estimated to be 60 percent of
those with gasoline engines (U.S. EPA, 1988a).
Bill provisions that call for alternative fuels/engines on a
percentage of all vehicles in the fleet are analyzed using a
different set of assumptions. The most likely situation was
judged to be the production of methanol fueled vehicles that
would begin to be sold in nonattainment areas sometime after
1995 .
A number of studies by government agencies, private
companies, and independent evaluators have pointed out the
significant potential of methanol (compared with other potential
alternative fuels) as the most likely near term replacement for
petroleum. Methanol contains about 50 percent of the energy
content of gasoline. Efficiency improvements are achievable
through the properties of methanol like its higher octane value
and its capability to be burned at very lean air-to-fuel ratios.
Price comparisons between methanol and gasoline presented here
take these factors into account.
In practice, it is expected that a small amount of gasoline
(15 percent) would be added to the pure methanol, making fuel
methanol (M85). Gasoline is added to the pure methanol to
improve engine starting and as a safety measure to reduce in-
vehicle tank explosion hazard and to add luminosity to the flame.
The emission characteristics of vehicles using M85 were modeled
in this study.
3 2
-------
These methanol fueled vehicles would replace both gasoline
and diesel powered vehicles. The key item in estimating costs
for methanol vehicles is the price difference between methanol
and either unleaded gasoline or diesel fuel. Apparently, the
additional cost of methanol can be estimated to be anywhere from
a net savings to a net cost of 7 0 cents per gallon. A more
narrow range for modeling seemed to be 0 to 10 cents additional
per gallon so 10 cents per gallon was used in the simulations to
provide a reasonable cost estimate. Per vehicle capital costs
for methanol versus gasoline (or diesel) can vary depending on
the number of vehicles that are manufactured in a year with
methanol capability. If many cars are being produced, the
capital cost is no different. If there is limited production,
methanol fueled vehicles are estimated to cost $400 per vehicle
more than gasoline vehicles. The cost difference for methanol
versus diesel is $300, but this represents the cost of a
catalyst, not a production cost difference. The $400 and $300
per vehicle costs for gasoline and diesel, respectively, were
used along with the 10 cent per gallon fuel difference in the VOC
model to estimate costs for the methanol option.
f. Application of Motor Vehicle Control Costs
All motor vehicle control options are costed on a per
vehicle basis. Costs for options applying to all registered
vehicles (I/M, enhanced I/M, RVP) are calculated by multiplying
the per vehicle cost by the number of vehicles for each MSA/state
combination. Options which apply on ly to new motor vehicles
(onboard, new motor vehicle standards) are evaluated by applying
the per vehicle cost to the number of new vehicles sold each year
for the MSA/state combination. The number of new vehicles is
calculated by multiplying the fraction of registered vehicles for
the area (number of vehicles for the area divided by the total
number of registrations) by the total number of new vehicles sold
each year {U.S. EPA, 1987b). Costs for options applying to
specific fractions of vehicles (alternative fuels), are
calculated by applying this fraction to the per vehicle cost and
then multiplying by the number of vehicles.
3 3
-------
A complete listing of the control strategies available in
ERCAM-VOC for each source category is given in Table II.8,
Included is the control technique, estimated reduction, and
average cost.
E. GROWTH PROJECTIONS
New growth emissions for stationary sources are estimated
using Bureau of Economic Analysis (BEA) projections of income by
industry and MSA (BEA, 1985). The current growth file contains
factors for projecting emissions to the years 1995, 2000, and
2010. The industrial category breakdown and the corresponding
match to the BEA data is given in Table II.9. The growth factors
are applied to uncontrolled 1985 VOC emissions to estimate future
year uncontrolled emissions. Average annual percentage growth
rates over the time period of the analysis are shown in Table
II.10.
Motor vehicle emission projections are based on national
averages of growth in VMT and changes in VOC emission factors.
The VMT projections for the study years are from the Motor Fuel
Consumption Model and are shown in Table 11.11. The emission
factors used are dependent on the control options being
simulated. Base emission factors simulate the effects of the
Federal Motor Vehicle Control Program. Other options include RVP
and new motor vehicle emission standards. The emission factors
for each option modeled are shown in Table 11.12.
The motor vehicle emission factors in Table 11.12 for 1985,
1995 Base, 2000 Base, and 2010 Base were estimated using MOBILE3.
These are weighted average emission factors estimated using three
different vehicle speeds (20, 45, and 55 mph). The fraction of
travel at each of these three speeds is used to estimate a
composite emission factor for each vehicle type. This method is
used to try to match the calculation procedure used in the NEDS
Area Source File to estimate base year motor vehicle emissions.
Emission reductions that might be achieved in restricting
RVP of gasoline to 9.0 psi are estimated using weighted national
average hydrocarbon (HC) emission factors from the gasoline
34
-------
Table iI.8
Cost Pods and Control Options
VQC Emission
Cost	Reduction
Pod / A * Description	Control Technique	(X)	$/Ton*»*
0
1
p
p
Combustion
Solvent metal cleaning





Solvent metal cleaning
Freeboard cover
23
-483


Solvent metal cleaning
Refrigerated freeboard
42
- 364


Solvent metal cleaning
Carbon adsorber
54
¦ 104
2
p
Printing and publishing





Printing and publishing
Carbon adsorber
75
-139


Printing and publishing
Carbon adsorber
as
-113
J
p
Dry cleaning





Bry cleaning
Recovery dryers
70
65
%
p
Fixed roof crude tanks





Fixed roof crude tanks
Internal floating roof
98
-39
5
p
Fixed roof gasoline tanks





Fixed root gasoline tanks
Internal floating roof
96
-245
6
p
EFR crude tanks





EfR crude tanks
Secondary seal
90
2722
?
p
EFR gasoline tanks





EfR g.iso! ine tanks
Secondary seal
95
¦ 11
d
p
Bulk terminals • Splash





Bulk terminals ¦ Splash
Submerged loading
59
•206


Bulk terminals - Splash
Submerged,balanced,carbon adsorber
78
-175


Bulk terminals • Splash
Submerged,balanced,carbon adsorber,testing
91
• 188
9
c
Bulk Terminals - Balanced




Bulk Terminals • Balanced
Carbon adsorber
67
-198


Bulk Terminals • Balanced
Carbon adsorber/truck testing
87
-212
to
p
Bulk Terminals • Submerged



Bulk Terminals • Submerged
Ba1anced,carbon adsorber
46
- 76


Bulk Terminals • Submerged
Balanced,carbon adsorber,truck testing
79
- 154
1 1
p
Stage I



Stage 1
Vapor balance
95
52
12
p
S t age 1 1




St age 1 1
Vapor balance • minimal enforcement
56
893
IS

Stage 1 1
Vapor balance ¦ maximum enforcement
06
900
p
Ethylene oxide manufacture



Ethylene oxide manufacture
Inc i nerat ion
98
246
16
p
Phenol Manufacture



Phenol Kanufacture
Incineration
98
703
1 ?
p
T rreph thai ic acid manufacture



'rrephthaltc acid manufacture
t ncineration
98
830
i a
p
Ac rylonitri1e manufacture

19
9
Ac ryI onit ri1e manuf sc ture
SOCH I f ug m»(s
1ncineratlon
98
176


SOCWI f ugitivps
Equipment and maintenance
37
- 63
68


SOCH i f ugitives
Equipment and maintenance
56
69


Petroleum refinery fugitives
Petroleum refinery fugitives
Equipment and maintenance


Petroleum refinery fugitives
Equipment and maintenance
80
38
2035


Petroleum refinery fugitives
Equipment and maintenance
93
-------
G"\
Cos t
Pod	P/
	Tf	"T
22	P
23	P
24	P
25	P
26	P
2?	P
28	P
29	p
30	P
31	P
32	P
33	P
34	P
35	P
36	P
36	P
37	f»
40 A
4 1 A
42 A


Table 11.8
-------
Iable 11.8 (coot. )
Cost Pods and Control Options
Cos t
Pod P/A* Description
Control Technique
4 3	A
^4	A
45	A
46	A
4 7	A
'.8	A
49	A
50	P
51	£•
52	P
53	P
54	P
55	P
60	A
61 A
62 A
Pr s nt i rig
Printing
P r i nt' I rig
Rubber and plastics mfg
Rubber and plastics mfg
Rubber and plastics r.f g
Carbon adsorber
Carbon adsorber
Carbon adsorber
Carbon adsorber
M i seel Ianeous
Miscellaneous
HiscelIaneous
Stage I
Stage 1
Stage II
Stage II
Stage
surface
surface
coat
coat
surface coat
ng
ng
ng
Architectural surface coating
Architectural surface coating
Consumer solvents
Consumer solvents
Coke ovens - door and topside
Coke ovens • door and topside
Coke oven by-product plants
Coke oven by-product plants
Aircraft surface coating
Aircraft surface coating
Aircraft surface coating
aging
aging
Whiskey fermentation -
uhiskey fermentation -
Charcoal manufacturing
Charcoal manufacturing
Marine vessel loading
Marine vessel loading
Light duty gasoline vehicles
tight duty gasoline vehicles
Light duty gas vehicles
light duty gas vehicles
Light duty gas vehicles
tight duty ges vehicles
Light duty gasoline trucks
Light duty gasoline trucks
Light duty trucks
Light duty trucks
Light duty trucks
Light duty trucks
gasoline vehicIes
gasoline vehicIes
duty gasoline vehicles
gasol me vehicles
Heavy duty
Heavy duty
Heavy
Heavy
duty
Auto refinishing control
Auto ref and industrial solvent control
Vapor balance
Vapor balance - minimal enforcement
Vapor balance ¦ maximum enforcement
Reformulate to waterborne
Default reduction
Inc i neration
Inspection and maintenance
H i gh solids coat i ng
Incinerat ion
Carbon adsorption
Inc ineration
Vapor balance
UK
Enhanced l&M
Alternate fuels to fleet vehicles
Alternate fuels to regular vehicles
Uaxman alternate fuels
Enhanced i&M
Alternate fuels to fleet vehicles
Alternate fuels to regular vehicles
Uaxman alternate fuels
Enhanced ISM
Alternate fuels to regular vehicles
Uaxman alternate fuels
VOC Emission
Reduc t i on
(%)	t/Ton**
75	-133
85	-104
70	238
83	334
!4	-3260
24	-754
95	745
56	893
86	900
52	0
20**	2000
90	373
63	92
79	4898
88	7020
85	32
80	1688
90	2000
15	7669
7	5725
6	2610
10	15332
5	11259
15	4556
7	3141
6	1545
10	8582
5	6304
13	4534
10	14734
5	10735
-------
Tabte 11.S (coot.)
Cost Pods and Control Options
Lost
Pod
P/A* Description
Control Technique
VOC Emission
Reduc 11on
(S)
$/Ton*
63
A
Heavy duty diesel vehicles
Heavy duty diesel vehicles
Alternate fuels to regular vehicles
10
80853


Heavy duty diesel vehicles
Waxman alternate fuels
5
59343
64
A
Off highway vehicles





Off highway vehicles
Default
90
2000
65
A
Railroads





Ra11 roads
Control of existing engines
51
1150


RaiIroads
Control of new engines
38
1150
66
A
Burning and fires



67
A
Area source incineration



68
A
Aircraft and marine vessels



70
A
TS0F





TSDf
Covers and carbon adsorption
90
900
71
A
Baker i es





Baker ies
Afterburners
90
1278
72
A
Cutback Asphalt





Cutback asphalt
Switch to emulsified asphalts
100
0
73
A
Public treatment works





Public treatment works
Covers and carbon adsorption
90
1111
90
P
Miscellaneous point





Miscellaneous point
Default reduction
90
1250
* P/A: P*pomt source A®area source
** Higher reductions than 20 percent are required (and modeled) as part of the Group of Nine Proposal.
•" Cost per ton for point sources is estimated by applying the cost equation to the average sized new
source (Battye et al., 1987).
Cost per ton for motor vehicle control options are based on a 1995 projection year. Cost per ton
increases in 2000 since motor vehicle emissions per vehicle decrease as a result of the
Federal Motor vehicle Control Program. Cost per ton also increases if RVP or new motor vehicle
control options are applied.
-------
Table I I.9
VOC Model Industrial Categories
VOC Model
!ndustri3! Category
food and Agriculture
M 1 n i rtg Operat 1 oris
wood Products
Printing and Publishing
Chemieals
Petrol cum Ref1ning
Mineral Products
Metals
Machinery 1. Equipment Mfg.
Crude Oil Production,
Storage, and Transfer
Elect r1c Utilities
Other fuel Combustion
Petroleum Product Prod.,
Storage, and Transfer
Other Transportation
Dry Cleaning
Other
Desc r1pt1 on
SIC 1,2,7,8,9,20,21
baker 1es
SIC 10,11,12,14
SIC 24,25,26
SIC 27, printing
SIC 28
rubber & plastics mfg
SIC 29
SIC 32
SIC 53,34
SIC 35,36,37,38,39
SIC 13
SIC 49
Other fuel combustion
SIC 51,55
Off highway vehicles,
raiI, air, £ water trans
SIC 72, dry cleaning
AlI other sources
SEA Industrial Designation for MSA Projections
Agricultural Services, Forestry,
Fisheries, and Other
M t m ng
Manufacturing •	Durable Goods
Manufacturing •	nondurable Goods
Manufacturing -	Nondurable Goods
Manufacturing - Nondurable Goods
Manufacturing - Durable Goods
M.inufactur 1 ng - Durable Goods
Manufacturing • Durable Goods
M1ni ng
Transportation and Public Utilities
Total Earnings
Wholesale Trade
Transportation and Public Utilities
Serv1ces
Total Earnings
-------
Table 11.10
Earnings Projections by Industry
United States Totals


Average


Annual
SIC

Percent;
Code
Industry
Growth*
07
Agricultural Services
4.1%
10-14
Mining
3.4
15-17
Construction
3.4
'20-39
Manufacturing
3.4

Nondurable goods
2.4
20
Food
1.5
26
Paper
2.3
27
Printing
2.8
28
Chemicals
3 . 0
29
Petroleum
2.5

Durable goods
4 . 0
24
Lumber
3 . 6
32
Stone, clay & glass
3 . 3
3 3
Primary metal
3 . 3
34
Fabricated metal
4 . 5
40-49
Transportation
3 . 5
70-84
Services
4 . 0
* The average annual percentage growth was computed over the
period 1983-1995. All industries are not included in this table,
just an illustrative sample.
Source: Bureau of Economic Analysis, 1985
40
-------
Table II.11
Current and Projected Nationwide Vehicle Miles Traveled
by Year and Vehicle Type
VMT (billions)
1985	1995	2000
2010
LDGV
1,354.9 1,748.4 1,898.7 2,199.3
LDGT
286 . 8
484 .9
578.4
965. 4
HDGV
80. 1
93 . 1
104 . 3
126. 7
HDDV
109. 7
158 . 3
188.0
247 . 5
Totals 1,831.5 2,484.7 2,769.4 3,538.9
LDGV
LDGT
HDGV
HDDV
Equivalent Annual Growth Rates
1985-1995	1995-2000
Average
2.6%
5.4
1.5
3 . 7
3 . 1%
1.7%
3.6
2	. 3
3	. 5
2 . 2%
2000-2010
1. 5%
5.2
2 . 0
2 . 8
2 . 5%
Source: EEA, 1987
41
-------
Table 11.12
Motor Vehicle Emission Factors by Year and Control Option
Emission Factor (grams/mile)
LDGV
1985	2.20
1995 Base	1.06
1995 RVP	0.79
1995 RVP and New Standards 0.77
2000 Base	0.97
2000 RVP	0.79
2000 RVP and New Standards 0.75
2010 Base	0.95
2010 RVP	0.79
2 010 RVP and New Standards	0.75
LDGT
4.13
1.91
1. 56
1. 50
1.53
1.30
1.19
1.42
1.28
1.16
HDGV
9.82
3 . 49
2	. 98
3	. 09
2.75
2 .93
2.66
HDDV
2 .01
0.93
0.85
0 .83
Sources: Lorang, 1988
U.S. EPA, 1984
U.S. EPA, 1987a
42
-------
volatility study performed by EPA (U.S. EPA, 1987a). The
relationship between weighted national average emission factors
at 9.0 psi and 11.5 psi was used to estimate the emission
reductions that might be achieved by RVP limits. A separate
calculation was performed for each of the three gasoline-powered
vehicle types.
In addition to projecting future year emissions for motor
vehicles, the number of vehicles must also be projected for
costing purposes. Vehicle numbers are elevated based on national
growth in vehicle registrations. Projections of the number of
vehicles by type and year are shown in Table 11.13.
F. ESTIMATING EMISSION REDUCTIONS AND COSTS
1.	SIP Regulations
The SIP regulations were taken from the file (Battye, 1987)
developed for the ozone NAAQS Cost Model which specifies SIP
applicability by state/county and cost pod and from information
on existing and planned I/M programs (U.S. EPA, 1987d). The file
indicates for each county which source categories are currently
regulated. Each pod is assigned a SIP control level
corresponding to an available cost equation. If an existing
source does not meet the requirements of the corresponding
regulation, the emissions are reduced and a control cost
calculated.
2.	NSPS File
The NSPS file is a file of pod and control strategy
combinations designed to simulate the effects of NSPS
regulations. The pods and control levels specified as NSPS
regulations are those designated for the NAAQS model (Battye et
a 1., 1987). Since some source categories do not have NSPSs, but
are regulated, SIP regulations are also applied to new sources.
It is assumed in the ERCAM simulations that all new sources will
be controlled to at least the same level as existing sources.
-------
Table 11.13
Number of Motor Vehicles by Year and Vehicle Type
LDGV
LDGT
HDGV
HDDV
Millions of	Vehicles
1985 1995	2000	2010
111.983 135.748	147.638	171.418
34.835 55.190	65.836	87.128
5.297 6.270	6.829	7.947
4.922 6.900	8.065	9.423
Source: EEA, 1987
4 4
-------
3.	Expansion of Nonattaininent Areas and Ozone Transport
Region Controls
MSA-specific regulations are modeled in the same way as are
SIP regulations. Two examples which have been modeled are
expansion of nonattainment areas to the MSA/CMSA level and ozone
transport region controls. Expanding the nonattainment area
classification to the MSA/CMSA level is modeled by subjecting all
sources in each county within the MSA or CMSA to SIP regulations.
Ozone transport region controls specify controls for areas which
may contribute to the nonattainment status of neighboring areas.
For example, controls may be specified for the entire Northeast
Corridor in an effort to reduce ozone in areas such as New York
City and Boston. Controls for all MSAs in the northeast region
are added for each category specified by the measure. These
controls are applied to both new and existing sources.
4.	Scenario Control Measures
Scenario constraints are organized to apply future controls
to sources by attainment area classification for simplicity and
because most of the proposed new VOC control provisions are
stipulated by nonattainment severity. Attainment areas are
handled as one class while the other classifications are based on
ozone design values. The exact definition of nonattainment can
differ according to the particular provisions being examined.
The attainment categories which have been used in the analyses
are shown in Table 11.14.
The scenario constraint file is designed so that controls
can be specified for each pod by attainment/nonattainment area
class. Separate scenario files are created for existing and new
sources. As an example, one facet of the proposed EPA policy is
potentially requiring enhanced I/M on LDGV and LDGT in
nonattainment areas with ozone design values above 0.16 ppm.
This would be simulated by indicating enhanced I/M as the motor
vehicle control option for serious and severe nonattainment
areas.
45
-------
Table 11.14
Ozone Attainment Categories
Category	Design Values (ppm)
EPA Proposed Policy, Mitchell Bill,
Waxman Bill
Attainment	_<. 0.12
Moderate Nonattainment	0.13, 0.14
Serious Nonattainment	0.15 to 0.18
Severe Nonattainment	> 0.18
Group of Nine Proposal
Attainment	_<_0.12
Moderate I Nonattainment	0.13
Moderate II Nonattainment	0.14, 0.15
Serious Nonattainment	0.16 - 0.18
Severe Nonattainment	> 0.19
-------
5.	National Control Measures
Several national options for motor vehicles control have
been included in this study. These options include the
following:
. Base case — simulating the effects of the Federal Motor
Vehicle Control Program (FMVCP),
. RVP — simulating the effects of FMVCP combined with
reductions in gasoline RVP,
. RVP plus new motor vehicle standards — simulating the
effects of RVP combined with new emission standards for
motor vehicles.
6.	Miscellaneous Measures
Emission fees can be used as a resource to help maintain
regulatory programs. The revenue generated can be used to fund
the program and help develop new control techniques. Fees are
also considered to be technology forcing measures in that they
encourage emitters to develop cost effective ways to reduce
emissions and thus the emission fee. ERCAM-VOC applies emission
fees to the remaining emissions from existing stationary sources
after all other controls have been applied. Varying dollar per
ton fees are applied based on size cutoffs and selected
attainment categories.
A RACT cutoff can also be simulated by ERCAM, and is
included in some modeling cases. Any non-CTG stationary source
below the size cutoff will not be subject to any controls
specified in the scenario file. This cutoff is only used for
point source emissions. No attempt was made to determine what
fraction of area source emissions are affected by size cutoffs.
G. RESULTS REPORTING
The VOC model currently provides aggregated results at the
national, state, and MSA level of detail. National level results
are reported by attainment category and cost pod. This report is
used to compare national costs for specific provisions of the
bills and policies being examined. It can also be used to
identify source categories where additional reductions might be
4 7
-------
achieved. A sample of an attainment category/pod output is shown
in Table 11.15.
State level results are reported by industry category. An
example of this report is given in Table 11.16. This level of
information may be used as a predecessor to economic analysis.
It will show what industries in each state will be expected to
bear costs under the provisions being examined.
Cost and emission totals are reported in the MSA level
output as shown in Table 11.17. This report is useful in
determining which MSAs will reach attainment or meet the progress
requirements mandated by the policy or bill being examined. A
simplified version of a trajectory ozone model (EKMA) is used to
estimate the required reduction to reach attainment for each area
based on the ozone design value, an assumed amount of transported
ozone, and the ambient nonmethane organic compounds (NMOC) to NOx
ratio. A list of MSAs and corresponding ozone design values and
required VOC emission reductions is shown in Table 11.18. EKMA
calculations are not part of ERCAM-VOC, though. Required VOC
emission reductions from Table 11.11 are an input to the model.
The ozone design values in Table 11.18 were taken from 1983
to 1985 monitoring data. These years were chosen because of the
relatively high concentrations measured in 1983 and because the
ambient data matched the time period of the emission inventory.
Note also that the design value monitors are not always
physically located in the MSAs listed — concentrations are
transport design values which are often downwind of the urban
area.
It should be noted that only the attainment status of
nonattainment areas identified via 1983 to 1985 ambient ozone
data has been investigated in this study. It is likely that some
attainment areas will grow into nonattainment and require
additional controls. This model does not attempt to predict
where this would occur or what the cost would be to bring these
areas back into attainment.
48
-------
Tab!e I I . 1S
Attainment: Category/Pod fiepo
ERCAH V'GC VERSION S2-5/88
SCENARIO:TEST MODEL REF:M85
CASE TEAR:1995 FEE: $ 100
RUN DATE:05/17/88 T1 ME:18:16:49
GRLAB:RN RVP:yes NUHVC:yes OOSIP:yos
BSIM:20 RACTCUT;25 FEECUT:3
••• ATTA INMEM T CA7ECORT/POO REPORT
POD POD
NAME
1985 NEDS
EMISSIONS
{TONS/TEAR)
PROJECTED
BASE ENISS.
AT CASE in.
(TONS/YEAR)
SCENARIO
CONTROLLED
EMISSIONS
(TONS/YEAR)
'.ATTAINMENT AREA
ZERO POD
SLV.MET,CLN
PR T »PUB
DRY CLNING
FXRFTK -CRD
FXRFTK-GASO
EFR-CRD
EFR-GASO
8	BGT-SPL
9	BGT -SU8/BAL
10 BGT SUB/NOBL
15 ETHLOX-MFG
17	TERACID-MFG
18	ACRYLON-MFG
19	SOCMI-FUGS
20	PETREF-FUGS
21	CELACT-MFG
22	STYBUT-MfG
23	POLYPRP-MFG
POLYETH-NFG
ETHYLEN-HFG
26	PETREF-UW
27	PETREF-VACDS
28	VEGOIL-MFG
29	PNUVAR-MFG
30	RUBR T I RE-MFG
31	GRNT I RE -MFG
32	CRBN8LK MFG
33	AUTOSRF-COAT
34	BEVCAW-MFG
35	GENSURF COAT
36	PAPRSR F- COAT
37	MISCSRF COAT
•»0 PAPRSR F - COAT
41 OEGREASIMG
2 4
25
153761
21492
50693
48
J9617
19915
9493
14482
474
577
1539 7
29
3990
1842
8074
23161
23034
11323
2109
19887
6868
17909
13313
5272
3572
7142
3201
31937
83499
19142
24 762
21997
83576
55985
295971
205074
40606
218273
60
63138
27651
12497
42154
716
1220
19596
39
26762
36487
25553
104573
31136
31493
3588
157593
15881
22892
18666
27031
6257
8337
3492
43435
122911
26589
30978
30691
119455
62678
323697
205074
34376
56195
17
30495
15827
5235
9761
302
411
13773
29
4446
2535
15765
42423
26762
17374
3588
157593
15881
13470
12762
27031
6257
4606
2177
43435
40634
16972
30978
10220
119455
60411
301977
SCEN
PCT ,
RED.
SIP ~ COST EFF.
SCENARIO FROM PROJCTD
COST
(1000$)
BASE LEVEL
(S/TON)
1.0
> .3
.3
1.7
1.7
?.8
1.1
.8
r.S
>.3
>.7
i. 6
S.4
!. 1
i. 3
>.4
i. 0
t.8
1.0
.0
1.0
1.2
! .6
.0
).0
..8
',7
1.0
>.9
>.2
1.0
>.7
1.0
1.6
>. 7
0
-1912
• 12745
3
-3086
-4432
27880
2627
-17
-110
-410
2
18514
5972
664
-6952
4351
1455
0
0
0
-1501
130
0
0
1641
3
0
485592
9881
0
-1799
0
9373
- 43
0
-307
-79
64
-95
-375
3839
81
-41
-136
-70
242
830
176
68
-112
995
103
0
0
0
-159
22
0
0
440
2
0
5902
1027
0
•88
0
4134
-2
-------
Table 11.16
State/Industry Category ftepo
ERCAM VOC VERSION S2-5/88
SCENARIO;TEST MODEL REF:M85
CASE YEAR: 1995 FEE: S 100
RUN DATE:05/26/S8 I I ME:13:01:08
GRI AS:RH RVP:yes NUMVC:yes DOS I P.-no
8SIM:?0 RACTCUT:25 FEECUT:3
•" STATE/INDUSTRY CLASS REPORT ***
!NOS FRY	INDUSTRY	1985 HEDS	PROJECTED
CLASS	CLASS	EMISSIONS BASE EHISS.
NUMBER	
-------
Table !!.17
MSA Summary Report
£ R CAM VOC VERSION SI. - 5/88
SCENAH10:TEST MODEL RE f:M85
CASE TEAR: 1995 f€c : S '100
RUN DATE : OS/26/88 T1 ME:13:01:08
>B;RN
RVP:yes NUMVC:yes DQSIP:no






1:20 RACTCUT : 25 f EECUT : 3






MSA SUMMARY REPORT "*






MSA
PMSA Region
1985 NEDS
PROJECTED
SCENARIO
SCENARIO
SIP ~
COST EFF.
HO.

EMISS10NS
BASE EMISS.
CON TROt LED
PERCENT
SCENARIO
fRM PRJCT0


(TONS/YEAR)
AT CASE YR.
EMISSIONS
REDUCTN
COST
BASE LEVEL

*

(TONS/YEAR)
(TONS/YEAR)

(1000$)
< VTON)
60
ABILENE, TX MSA
8810
765 1
6857
10.4
1055
1329
80
AKRON, OH PMSA
47616
48448
30747
36.5
29412
1662
120
ALBANY, GA MSA
9190
8884
7559
14.9
1594
1203
160
ALBANY-SCHENECTADY TROY, NY MSA
49051
124988
112696
9.8
30133
2451
200
ALBUQUERQUE, NM MSA
34628
25646
23175
9.6
3946
1597
220
ALEXANDRIA, LA MSA
8317
7486
6370
14.9
3460
3100
240
ALLENTOUN-BETHLEHEM, PA-NJ MSA
45639
45486
25476
44.0
29985
1498
280
ALTOONA, PA MSA
7338
6494
5250
19.2
4310
3464
320
AMAR!LLC, TX MSA
29704
30413
26079
14.3
1955
451
360
ANAHEIM SANTA ANA, CA PMSA
145929
128331
90405
29.6
64346
1697
380
ANCHORAGE, AK MSA
12873
12038
10469
13.0
5186
3305
400
ANDERSON, IN MSA
13052
13370
7141
46.6
29434
4725
405
ANDERSON, SC MSA
11679
10594
9367
11.6
1506
1227
440
ANN ARBOR, MI PMSA
21252
22527
11888
47.2
18856
1772
450
ANNISTON, AL MSA
7087
6011
5331
11.3
1184
1741
460
APPLETON-OSHKOSH NEENAM, Wl MSA
37975
39692
27694
30.2
12377
1032
480
ASHEVILLE, NC MSA
12606
11249
9814
12.8
1564
1090
500
ATHENS, GA MSA
9354
7490
6632
11.5
1328
1548
520
ATLANTA, GA MSA
181599
179977
94643
47.4
155900
1827
560
ATLANTIC CITY, NJ MSA
16891
16422
10263
37.5
9375
1522
600
AUGUSTA, GA-SC MSA
26702
24291
21787
10.3
3854
1539
620
AURORA-ELGIN, IL PMSA
23687
23263
13840
40.5
19911
2113
640
AUSTIN, TX MSA
46185
38264
34392
10.1
5858
1513
680
BAKERSFIELD, CA MSA
42360
41576
20498
50.7
29730
1410
720
SAL I 1 MORE, MD MSA
136178
139053
83973
39.6
110576
2008
733
BANGOR, ME NECMA
8539
8003
6172
22.9
4556
2488
760
BATON ROUGE, LA MSA
84069
156031
33475
78.5
106548
869
780
BATTLE CREEK, Mi MSA
10318
9605
7383
23.1
4735
2131
840
BEAUMONT-PORT ARTHUR, TX MSA
130875
424140
54972
87.0
157015
425
84 5
BEAVER COUNTY, PA PMSA
12235
12486
5395
56.8
9818
1385
860
BELL INGHAM, UA MSA
12156
11698
10596
9.4
1391
1262
870
BENTON HARBOR, Ml MSA
12136
11073
8723
21.2
5399
2297
875
BERGEN-PASSAIC, NJ PMSA
844S7
77597
49903
35.7
44335
1601
880
BILLINGS, MT MSA
12732
12718
11162
12.2
1780
1144
920
BILOXIGULFPORT, MS MSA
13200
11391
10440
8.3
1627
171 1
960
8INGHAMTOM, NY MSA
18369
16424
13182
19.7
3557
1097
1000
BIRMINGHAM, AL MSA
62086
58095
37817
34.9
50942
2512
-------
Table 11.18
Ozone Design Values and Emission Reduction Requirements
1983-1985	1985 NEDS REQUIRED**
STAR
MSA/CMSA

MSA
OZONE
NMOC/NOX*
VOC
REDUCTION
NUM


CODE
DES.VALUE
RATIO
EMISSIONS
(%)
6022
Massachusetts

22
0.17
7.6
384
796
36
3901
Allentown-Bethlehem,
PA-NJ
240
0.13
10.4
45
639
8
1101
Atlanta, GA

520
0.16
7.7
181
599
32
,3101
Atlantic City, NJ

560
0.16
10.4
16
891
40
502
Bakersfield, CA

680
0.16
10.4
42
360
40
2101
Baltimore, MD

720
0.17
6.1
136
178
27
1901
Baton Rouge, LA

760
0.16
14.9
84
069
48
4502
Beaumont, TX

840
0.17
U.6
130
875
49
101
Birmingham, AL

1000
0.13
9.8
62
086
8
0
Bradenton, FL

1140
0.13
10.4
9
527
8
5001
Charleston, WV

1480
0.13
10.4
23
607
8
3402
Chariot te-Gastonia,
NC-SC
152.0
0.13
10.4
77
904
8
4401
Chattanooga, TN-GA

1560
0.13
16.7
40
689
8
6014
Chicago CMSA

1602
0.25
8.5
529
572
.58'
6036
Cincinnati CMSA

1642
0.17
9.1
121
743
42
3604
Cleveland CMSA

1692
0. 14
7.5
190
252
21
4505
Dallas CMSA

1922
0.16
13.0
266
233
47;
3606
Day ton-Springfield,
OH
2000
0.13
10.4
71
385
8
603
Denver CMSA

2082
0.13
8.2
140
469
8
2301
Detroit CMSA

2162
0.13
10.4
318
674
8
4506
El Paso, TX

2320
0.16
12.0
35
069
44
3904
Erie, PA

2360
0. 13
10.4
17
710
"~8
504
Fresno, CA

2840
0.17
10.4
37
700
46
102
Gadsden, AL

2880
0.13
10.4
9
707
8
2303
Grand Rapids, MI

3000
0.13
10.4
47
777
8
3905
Harrisburg-Lebanon,
PA
3240
0.13
10.4
33
779
8
4509
Houston CMSA

3362
0.25
10.8
370
531
:65,<
5002
Hun t ington-Ashland,
W-KY-0H
3400
0.14
10.4
31
976
20
1504
Indianapolis, IN

3480
0.13
10.9
121
961
8
1002
Jacksonville, FL

3600
0.14
10.4
55
784
20
5103
Janesville-Beloit, WI
3620
0.13
10.4
12
817
8
1701
Kansas City, M0-KS

3760
0.14
9.2
125
335
16
1903
Lake Charles, LA

3960
0.14
24.3
26
294
30
4510
Longviev, TX

4420
0.15
10.4
22
015
33
6005
Los Angeles CMSA

4472
0.36
10.4
824
055
49'
1802
Louisville, KY-IN

4520
0.15
10.4
77
299
33
4404
Memphis, TN

4920
0.15
13.9
65
116
36
6010
Miami CMSA

4992
0.14
13.3
157
426
19
6051
Milwaukee CMSA

5082
0.17
10.4
109
465
<46;
2402
Minneapolis-St. Paul, MN-WI
5120
0. 15
10.4
189
531
33
506
Modesto, CA

5170
0.16
10.4
20
050

2305
Muskegon, MI

5320
0. 14
10.4
15
822
20
4405
Nashville, TN

5360
0. 14
10.4
73
025
20
52
-------
Table II.18
Ozone Design Values and Emission Reduction Requirements
STAR MSA/CMSA
NUM
1983-1985	1985 NEDS REQUIRED**
MSA OZONE NMOC/NOX* VOC REDUCTION
CODE DES.VALUE RATIO EMISSIONS (*)
6033 Nev York CMSA	5602	0.24
4801	Norfolk, VA	5720	0.14
6039 Philadelphia CMSA	6162	0.20
301 Phoenix, AZ	6200	0.16
3909	Pittsburgh CMSA	6282	0.13
3803 Portland CMSA	6442	0.13
4101 Providence, RI	6483	0.15
3910	Reading, PA	6680	0.13
4802	Richmond-Petersburg, VA	6760	0.13
511 Sacramento, CA	6920	0.18
2601 St. Louis, MO-IL	7040	0.16
4602 Salt Lake City, UT	7160	0.15
513 San Diego, CA	7320	0.21
6006	San Francisco CMSA	7362	0.19
516 Santa Barbara, CA	7480	0.16
5109 Sheboygan, UI	7620	0.17
519 Stockton, CA	8120	0.15
1007	Tampa-St. Petersburg, FL	8280	0.13
3702 Tulsa, OK	8560	0.13
521	Visalia-Tulare-Porterville, CA 8780	0.13
901 Washington, DC	8840	0.17
1008	West Palm Beach, FL	8960	0.14
3914 York, PA	9280	0.13
522	Yuba City, CA	9340	0.13
6007	Greater Connecticut CMSA	9999	0.19
11.7
10.4
6.8
10.4
10.4
10.4
10.4
10.4
11.1
10.4
9.6
10.3
10.4
10.4
10.4
10.4
10.4
10.4
14.4
10.4
8.2
13.3
10.4
10.4
6.1
887,534
78,350
344,747
114,562
122,802
95,120
57,881
22,933
66,718
87,463
185,377
72,379
126,559
375,354
28,805
8,586
25,799
107,549
54,034
19,478
168,375
43,946
28,161
9,836
127,499
67
20
41
41
8
8
33
8
8
50
38
32
58
52
40
46
33
8
8
8
38
19
8
8
34
Source: U.S. EPA, 1988b
* A ratio of 10.4 to 1 vas assumed for MSAs vith no available data.
** Estimated using EKMA with carbon bond 3 chemistry. While this is a
relatively sophisticated technique, more complex models such as AIRSHED
may estimate different reduction requirements. Uncertainties associated with
using different chemistries in EKMA (carbon bond 4 instead of carbon bond 3)
have not yet been quantified.
53
-------
-------
^2 one
'MSA
Oes
Vai
T°bl* «.J8
\rk CHSA
«> Va
elPhia CMC,
AZ MSA
'J*h CMSA
d CMSA
iee, RJ
PA
:Petersbu
r°. C/i *'
HO-n
Ccly- W
» wl
!sco CMSA
CA
yj
:a
'e	a
iDe"Portervnie,
"*ch, -t-j
•** ,90t,5*= Se,"'re^"
C0°'DEs^'7>r.,5«»SOs,
5602 " - **^0 Voc
S?20 °-24	^/OWs «**Tr<»
1 ' A •> 0. )/. J J 7 * - • .	f fj
Cicu^ CWS4
M' J9385
5602
5?2o
616 2
fi200
6282
6442
6483
6680
6760
6920
7 040
7160
?320
?362
?480
?620
8120
8280
_ 8560
CA 8?ao
8840
8960
9280
9,340
9999
0-24
0-H
°-20
0-16
0. j3
°-l3
°-!5
0. 1J
°-13
°-l8
0. J6
0. J5
0-2}
0. 19
°- 16
0.l7
O.i 5
13
0.13
0. J3
0,|;
0.| 4
0.13
0. 13
0, 19
' J. }
'0.4
6-8
JO. 4
JO. 4
JO.4
JO. 4
JO.4
H.I
10.4
9.6
10. 3
JO. 4
'0,4
'0.4
JO.4
'0.4
10. 4
/ 4. 4
JO.4
®.2
H. j
JO.4
'0.4
6
I
887 r ' "
, ' • 5 34
121,80?
'?.!*>
:7**i
*'.4«j
'"V l;»
'**, !.?»
! ¦"» -
*>¦ i?f
'•'V 154
§.
• 5- '*•
'0'. %,»
,>4.aj4
I f # 4
'M.I/,
* ' * - -
. I i I
».<*
J"0
*1
*1
*
#
yg
mPW
If
••
m
1J
' 1 vas „
aisumed f0r
f£^ vfth	"h fto —	4.,.
s f i ca fed \ °" b"nd ] -	,
ierpnr ,'echnilue, mnr* *">¦ »«J.
1 **• <^rr*
»
%
¦¦
-------
Ill CARBON MONOXIDE (CO) MODEL DEVELOPMENT
In conjunction with the ozone (VOC and NOx) nonattainment
analyses a CO model was developed, following the same basic
design criteria as the VOC model discussed in Chapter II of this
report. ERCAM-CO is a nonoptimizing, deterministic data base
management system for performing comparative analysis of the
regional and industry costs and CO emissions effects of proposed
bills and regulatory policies. The model's primary data base
consists of approximately 18,000 records, containing CO emissions
data in tons for area and stationary sources at the state,
county, MSA, and SIC level of detail. The model is written in
dBase III Plus for IBM-compatible personal computers and uses the
Clipper dBase compiler for computing efficiency. Figure III.l
shows a flowchart of the salient data and analytic
characteristics of the model.
A major difference between the VOC and CO models is that the
CO model separates the stationary sources from area sources
(mainly transportation) sources both in terms of data file
management and the internal model logic and cost equations. This
segregation was done mainly because the most stringent and
complex CO regulations proposed under pending legislation are for
area sources (transportation and residential fuel use), rather
than stationary sources.
The CO model was designed to be run in tandem with the VOC
model for the same scenario or bill. Certain rules governing the
interrelationships between VOC and CO controls and costs were
followed. Control strategies which affect both CO and VOC are
treated as if the CO-related aspects were costed in the VOC
model, to prevent double counting of cost elements in reporting
results. This applies mainly to the application of vehicle I/M
programs and possible new motor vehicle emission standards which
affect both pollutants. Basic I/M programs are not costed for CO
if they are specified under the VOC provisions of a bill. If
basic I/M is already in effect and the VOC aspects of the bill do
not increase the I/M level but the CO aspects of the bill do,
-------
Figure III. 1
Carbon Monoxide Model Organization
Source	Growth	Future Emission	Control Options
Categories	Factor	Rates	Available
None
Population
Point Sources
None
Other Area Sources
No Change
No Change unless
Controlled
Vehicle Miles
Traveled by
Vehicle Type
No Change Unless
Controlled
Residential Fuel Burning
Wood
LDGV
LDGT
HDGV
HDDV
Control Options by
Source Type
(See Table 111.1)
1995 Emission Factor
— 1985 Emission Factor
by Vehicle Type
1.	Enhanced l/M
2.	Oxygenated Fuels
3.	New Emission
Standards
(affects LDGTs only)
1.	NSPS Affects New
Wood Stove Emissions
2.	Add-on Catalysts for
Existing Wood Stoves
3.	No Controls for
Fireplaces
-------
then an incremental cost of the enhanced-rainus-basic I/M program
is included as a cost.
A,	1985 EMISSION INVENTORIES
The 1985 NEDS point and area source files we^e the source of
the emissions data used in the CO model. The 1985 area source
file data were grouped into six source categories for the
analysis as follows:
.	light-duty gasoline powered vehicles (LDGVs),
,	light-duty gasoline powered trucks (LDGTs),
.	heavy-duty gasoline powered trucks (HDGVs),
.	heavy-duty diesel powered trucks (HDDVs),
.	residential fuel burning-wood, and
.	other area sources.
Point source records for CO emitters were organized into three
types. The first type includes point sources emitting greater
than 100 tons per year. Individual records were retained for
these sources. Sources emitting less than 100 tons per year were
aggregated by six digit SCC code and MSA/state region into the
second type of record. Of these aggregated sources, those
emitting less than 10 tons per year were aggregated by MSA/state
and assigned a separate SCC code. Groupings of SCCs were
according to common emission and control characteristics. These
are described in more detail in the next section.
B.	CONTROL COST EQUATIONS
As discussed earlier, the CO model separates the treatment
of stationary sources and area/mobile sources, since current
control proposals and estimating methods for these sources are
distinct. For stationary sources, for instance, costs are based
on size/operating rates, whereas for transportation sources, they
are based on numbers of vehicles, vehicle miles traveled, and
vehicle fleet growth/replacement rates. The mobile sources have
been subject to many specific policy proposals, such as
requirements/incentives for oxygenated fuel blends or neat
alcohol for targeted fleet use. Motor vehicle CO emissions are
generally expected to continue their downward trend, due to more
stringent new vehicle standards and fleet turnover between 1985
56
-------
and either 1995 or 2000. This section discusses the cost and
emissions control reduction methodology in detail, first for
stationary, then area/mobile sources.
1. Stationary Sources
Point source controls have rarely been considered an
important part of any urban area CO control plan. CO standard
exceedences have usually been associated with motor vehicle
emissions, and in some cases with residential wood combustion.
Thus, there has been little recent work to examine control
efficiencies and costs for CO stationary source controls.
Fortunately, for the most recent CO regulatory analysis, capital
and annualized costs of CO control systems were estimated for a
number of industrial processes (PEDCo, 1979) . This information
was reviewed and judged to be appropriate for application in this
analysis after being updated to reflect current dollars.
Equations were put in exponential form for each combination of
industrial process and control equipment type. This information
is summarized in Tables III.1 and III. 2.
As can be seen in the tables, all of the carbon monoxide
control devices are highly efficient. With the exception of the
o2 analyzers, installing the appropriate control equipment can
potentially reduce CO emissions from 90 percent to 99.5 percent
over the uncontrolled rate. This leads to a relatively low cost
efficiency in comparison with the control of other pollutants,
such as N0X or S02.
The stationary CO sources for which control costs were
estimated fall into four general industry classifications:
. iron and" steel,
. aluminum,
. solid waste disposal, and
. chemicals.
The most common CO control method is to use some form of thermal
incineration, either in conjunction with primary heat recovery,
primary and secondary heat recovery, or no heat recovery.
Incineration achieves from 90 percent to 99.5 percent CO
reduction, depending on the source type.
57
-------
Table III.l
Carbon Monoxide Control Cost Equations
for Retrofit Applications
(1985 Dollars)
Retrofit	Retrofit
Capital	O&M	Control Default
CO Source
Control Device
a
b
a
b
Essssaa
Eff. (X)
Cost/Ton
Carbon Black
Incin. w/PR
35.0
0.98
4.07
0.94
99.5
3
Carbon Black
Incin. w/PR & CO
Boiler 449.0
0.85
-37.20
1.06
99.5
-47
Iron Ore Sinter Plant-Windbox
Incin. w/PR
276.0
0.73
39.60
0.82
90.0
172
Iron Ore Sinter Plant-Windbox
Incin. w/P&SR
206.0
0.76
-0.06
1.32
90.0
-288
Carbon Steel Electric Arc Furnace
Direct Shell Evacuation 534.0
0.65
40.60
0.74
90.0
248
Gray Iron Cupola
Thermal Incin.
4,160.0
0.15
0.99
0.91
90.1
5
Conical Wood Burner
02 Analyzer
8,060.0
0.00
4,310.00
0.00
50.0
9
Municipal Incinerator
02 Analyzer
272.0
0.40
138.00
0.42
50.0
102
Basic Oxygen Furnace
Open Hood System
229.0
0.73
-1.51
0.99
95.0
-21
Prebake Aluminum Cells
Incin. w/PR
65.1
1.06
41.20
1.10
99.0
824
Aluminum Anode Baking
Incin. w/PR
2.3
1.09
0.62
1.10
99.0
83
Maleic Anhydride
Incin. w/PR
3,100.0
0.57
57.10
0.93
98.0
50
Maleic Anhydride
Incin. w/P&SR
1,630.0
0.65
2.93
1.22
98.0
45
Coke Oven Charging
Stage Charging
458,000.0
0.04
8,650.00
0.30
99.0
2,613
Cyclohexanol
No Heat Recovery
10,600.0
0.24
68.00
0.64
98.0
38
Cyclohexanol
Incin. w/PR
110,000.0
0.11
334.00
0.49
98.0
43
Ethlyene Bichloride
Incin. w/PR
254.0
0.60
1.08
1.00
98.0
--
NOTES: Equations are of the form COST = a*(SIZE)*b
Incin. v/PR is a Thermal Incinerator with Primary Heat Recovery
Incin. v/P&SR is a Thermal Incinerator with Primary and Secondary Heat Recovery
-------
Table III.2
Carbon Monoxide Control Cost Equations
for New Applications
(1985 Dollars)
New Capital	New Q&M	Control
CO Source
Control Device
:iC32S3SSa3Sa = = S£X = =S5'= =:=' =
a
b
a
SS===SSB=SS3i
b
ebs=: = sr
Efficiency
Carbon Black
Incin. v/PR
26.9
0.98
4.07
0.94
99.5
Carbon Black
Incin. w/PR & CO Boiler
345.0
0.85
-37.20
1.06
99.5
Iron Ore Sinter Plant-Windbox
Incin. w/PR
230.0
0.73
39.60
0.82
90.0
Iron Ore Sinter Plant-Windbox
Incin. w/P&SR
172.0
0.76
-0.06
1.32
90.0
Carbon Steel Electric Arc Furnace
Direct Shell Evacuation
445.0
0.65
40.60
0.74
90.0
Gray Iron Cupola
Thermal Incin.
3,210.0
0.15
0.99
0.91
90.1
Conical Wood Burner
02 Analyzer
7,330.0
0.00
4,310.00
0.00
50.0
Municipal Incinerator
02 Analyzer
249.0
0.40
138.00
0.42
50.0
Basic Oxygen Furnace
Open Hood System
176.0
0.73
-1.51
0.99
95.0
Prebake Aluminum Cells
Incin. w/PR
59.2
1.06
41.20
1.10
99.0
Aluminum Anode Baking
Incin. w/PR
2.1
1.09
0.62
1.10
99.0
Maleic Anhydride
Incin. w/PR
2,820.0
0.57
57.10
0.93
98.0
Maleic Anhydride
Incin. w/P&SR
1,480.0
0.65
2.93
1.22
98.0
Coke Oven Charging
Stage Charging
352,000.0
0.04
8,650.00
0.30
99.0
Cyclohexanol
No Heat Recovery
9,640.0
0.24
68.00
0.64
98.0
Cyclohexanol
Incin. w/PR
100,000.0
0.11
334.00
0.49
98.0
Ethlyene Dichloride
Incin. w/PR
230.0
0.60
1.08
1.00
98.0
NOTES: Equations are of the form COST ¦ a*(SIZE)Ab
Incin. w/PR is a Thermal Incinerator with Primary Heat Recovery
Incin. w/P&SR is a Thermal Incinerator with Primary and Secondary Heat Recovery
-------
The greatest variety of control methods occurs within the
iron and steel industry. In addition to thermal incineration for
emissions from iron ore sinter plant windboxes and gray iron
cupolas, three other control methods are available. For carbon
steel electric arc furnaces, a direct shell evacuation system is
the control device considered. Although this system primarily
controls particulate emissions with a fabric filter, it also
controls up to 90 percent of the CO emissions by aspirating air
through an air gap and then combusting the CO. Using an open
hood system, 90 percent of the CO emissions from basic oxygen
furnaces can be removed. In such a system, CO emissions are
collected in an open hood and air is added to insure complete
burning of the CO in the hood. After the gas is cooled, it
passes through an electrostatic precipitator to remove
particulates and then the cleaned gas passes out of the stack.
The final control method applied to this industry is stage
charging which can control up to 99 percent of the CO emissions
which occur during coke oven charging. This control method,
whereby coal is charged at a reduced rate and suction on the oven
is maintained during the charging, is a modification of the
typical coke oven charging technique. By making these
modifications, CO emissions should remain within the oven
collection system without leaking to the atmosphere.
In the solid waste disposal category, the two source types
both use 02 analyzers and recorders to optimize combustion. It
is expected that by optimizing combustion, approximately 50
percent of the CO emissions can be reduced.
The remaining source categories, the aluminum and chemical
industries, all utilize forms of thermal incineration. The see
codes to which all the above mentioned control measures are
appl ied in the model are listed in Table III. 3.
Cost components for installing appropriate control systems
on major stationary sources of CO emissions were used (PEDCo,
1979). For each source, costs were reported for installing
controls on two different facility sizes. The facility sizes
60
-------
Table III.3
Applicable SCC Codes for Stationary Source CO Categories
Category
Applicable SCCs
Iron & Steel Industry
Basic Oxygen Furnace
30300913, 14
Carbon Steel Electric Arc Furnace 30300904, 08
30400304
30400701
Coke Oven Charging
Gray Iron Cupola
Iron Ore Sinter Plant Windbox
30300302
30400301
30300813
Solid Waste Disposal
Conical Wood Burner
Municipal Incinerator
50100508
50200105
50300105
50100101, 02
Aluminum Industry
Aluminum Anode Baking
Prebake Aluminum Cell
30300105
30300101-03
Chemical Industry
Carbon Black
Cyclohexanol
Ethylene Dichloride
30100503, 04
30115801.-03
30115821, 22, 80
30112501, 02
30112504-06
30112509
Maleic Anhydride
3 0.110002
61
-------
were chosen to be representative of a large plant and a small
plant. The costs for each facility were broken down as follows:
. Capital Cost
-	Direct
-	Indirect
-	Contingency
-	Total Cost of New Installation
-	Retrofit Factor
-	Total Cost with Retrofit
. Annualized Cost
-	Direct
-	Indirect
-	Credit
-	Total Annual Cost for New Installation
-	Total Annual Cost with Retrofit
The retrofit factor ranged from 10 percent to 3 0 percent of the
total capital cost of a new installation depending on the
difficulty associated with applying a control system to an
existing source. The credit component of the annualized cost
represents any cost savings due to such factors as improved
process efficiency, decreased fuel costs, lowered steam
requirements, or other process improvements. The indirect
portion of the annualized cost was assumed to be the annualized
capital cost. Therefore, the total O&M cost used was the sum of
the direct and credit components of the annual cost.
Before developing the cost equations, the total capital
costs and the total O&M costs were escalated from 1978 dollars to
1985 dollars using the Chemical Engineering economic index (Chem.
Engr., 19 8 8). To develop the cost equations from the available
cost data (PEDCo, 1979), it was assumed that the control cost
varied with the facility size in the form COST = a*(SIZE)b where
a and b are constants, SIZE is the facility size or operating
rate in tons of product per year, and cost is in 1985 dollars.
The cost and size data were converted to logarithmic values and
linear regression was used to find the best line fitting the
equation. Separate equations were developed for capital cost and
O&M costs. The exception to this method was for ethylene
dichloride costs. The data included only one source size, and so
it was assumed that capital costs would vary with size to the 0.6
62
-------
power and that O&M costs would be linear. The annual cost for
each source in the model was calculated by multiplying the
capital cost by the capital recovery factor of 0.09 and adding
the O&M cost. Table III. 1, referred to earlier, listed the
equation constants for retrofit applications and Table III.2
listed the constants for new applications (only retrofit
equations were used in this analysis).
To calculate the control costs for sources falling in the
categories covered by the equations but with no plant size listed
in the emission inventory, default cost effectiveness values were
computed. For each source category, the capital, O&M, and
annualized (net annualized portion of both capital and O&M costs)
costs were calculated for a plant considered to be of the average
size for its type. The emission factor for that source type
(U.S. EPA, 1985) was multiplied by the average source size and
the control efficiency to find the tons of CO emissions reduced.
The total annual cost divided by the tons of CO reduced gave the
cost effectiveness for that source type. This figure was then
used as a default for all plants in that category with missing
source sizes. By multiplying the default cost efficiency by the
uncontrolled emissions for a source, an estimate of the annual
cost could be made. It was necessary to have this default
because a significant number of sources in the data base had
missing values or zero as the source size while the uncontrolled
emission rate was almost always included.
2. Area/Mobile Sources
In the current version of the model, area sources comprise
the four motor vehicle categories (LDGV, LDGT, HDGV, HDDV) plus
residential fuel burning, with wood stoves and fireplaces broken
out for specific control measures, and an "all other area source"
category.
Five motor vehicle control options are available for
selection in the CO model. Other potential CO control options
which have not been analyzed include transportation control
measures (VMT reductions) and increasing passenger vehicle
occupancy.
63
-------
a.	New Motor Vehicle Emission Standards
New motor vehicle emission standards are a control option
linked to the VOC model. New motor vehicle standards affect all
three pollutants (VOC, NOx, and CO) and are costed solely in the
VOC model. These more stringent emission standards are expected
to reduce CO emissions only from light-duty gasoline trucks. The
emission reduction is modeled by changes in future year emission
factors rather than an emission reduction percentage. This is
discussed in more detail in Section III.C, Growth Projections.
b.	Inspection and Maintenance (Basic and Enhanced)
CO reductions for basic and enhanced I/M programs are based
on M0BILE3 values and are listed in Table III.4. The estimated
cost for a basic I/M program is $20.20 per vehicle (LDGV and
LDGT). The incremental cost for enhanced I/M is $6.48 per
vehicle.
One of the more complex aspects of the CO modeling is to
ensure that no double counting of costs occurs for I/M. If basic
or enhanced I/M is already in place in an area according to the
VOC provisions of a bill, an emission credit is given in the CO
model at no cost. An extract of the VOC regional constraint file
is used to determine which areas already have basic or enhanced
I/M programs due to SIPs, expansion of nonattainment areas to the
MSA/CMSA level, and ozone transport region controls. In
addition, the user also specifies which ozone related attainment
categories have enhanced I/M according to the VOC provisions of
the policy or bill being examined. This information is used in
the CO model to give emission and cost credits where appropriate.
c.	Alternative Fuels
Two different alternative fuel options are available for
selection in the CO model. Either a 10 percent ethanol blend or
a Methyl Tertiary Butyl Ether (MTBE) blend may be selected. It
is assumed that these blends will affect all gas-powered vehicles
and be used one-third of the year (winter). A 10 percent ethanol
blend will reduce CO emissions from gasoline vehicles by an
estimated 21.95 percent in 1995 and 19.30 percent in 2000 at an
incremental cost of 0.5 cents per gallon (after the federal
64
-------
LDGV	13%
LDGT
Table III.4
CO Reductions for I/M Programs
iqqc	Projection Years
Basic I/M	Basic I/M	Enhanced I/M
44%
37%
12%	42%
49%
Source: MOBILE3
65
-------
subsidy). Based on the average fuel consumption by motor vehicle
type, the resulting per vehicle costs are $2.69 per LDGV, $2.43
per LDGT, and $6.15 per HDGV applied to one-third of the
vehicles.
MTBE blends are expected to reduce total gasoline vehicle
emissions by an estimated 13.45 percent in 1985 and 12.30 percent
in 2000. The incremental cost for fuel is 3.2 cents per gallon
resulting in per vehicle costs of $5.74 per LDGV, $5.17 per LDGT,
and $13.12 per HDGV assuming the fuel is used one-third of the
year.
d. Residential Wood Burning
Residential wood burning includes bi -ning in both fireplaces
and wood stoves. Controls examined are for reducing CO emissions
from woodstoves only. No controls are analyzed for reducing CO
emissions from fireplaces. It is assumed that one-third of
existing wood stoves can be retrofit with a catalyst at a cost of
$150 per stove. Based on average emissions from a wood stove,
the estimated cost becomes $50 per ton of CO reduced. All new
stoves are controlled also at an annual cost of $50 per ton of CO
reduced. The expected reduction over uncontrolled stoves is 65
percent. It is also assumed for modeling purposes that of total
residential woodburning emissions, 33 percent is from fireplaces
and 67 percent from wood stoves. This split was derived from
NEDS emission estimates for an "average" county.
C. GROWTH PROJECTIONS
Growth factors are used to produce future year estimates of
motor vehicle and area source emissions. For simplicity and
since point sources are not considered a major contributor to CO
nonattainment, point source emissions show zero growth in the CO
model. It is important to produce future year estimates of motor
vehicle and area source emissions since these have a greater
impact on CO nonattainment. and since the most stringent and
complex CO regulations which have been proposed are for these
sources.
6 6
-------
1.	Motor Vehicles
Growth in motor vehicle emissions is based on the joint
effects of estimates of VMT and the CO emission rate in grams per
mile. Since motor vehicle control costs are based on fixed
estimates in dollars per vehicle, a growth rate is also required
for the number of registered vehicles, in addition to VMT. All
of the motor vehicle growth parameters are based on national
averages. Future year emission factors for CO are presented in
Table III.5. Future year VMT and vehicle registrations are the
same as those found in the VOC model.
There is one special case in the growth rate calculation.
If new vehicle standards for CO are required, a special (lower)
regulated new vehicle emissions rate, given in parentheses in the
table, is used. This control option affects only light-duty
gasoline trucks.
2.	Area Sources
All other area source CO emissions are projected to grow in
proportion to (national) population growth. Thus, there are no
MSA or state growth rates used to account for regional growth
differences for these sources. Total population estimates for
the model projection years are shown in Table III.6.
67
-------
Table III.5
Motor Vehicle CO Emission Factors
Emission Factor	(grams CO/mile)
Year	LDGV	LDGT	HDGV	HDDV
1985	12.11	21.64	87.72	5.20
1995	5.46	9.78	(9.41) 23.08	2.52
2000	5.06	7.86	(7.11) 17.28	2.43
2010	4.97	7.58	(6.58) 14.84	2.35
* ( ) indicates CO emission factor used if new motor vehicle
standards are in effect.
-------
Table III.6
Total Population Estimates by Projection Year
Year	Population
1985	232,300,000
1995	252,200,000
2000	259,800,000
2010	274,800,000
69
-------
IV N0X COST ESTIMATES
No model was developed for the N0X portion of the analysis,*
instead, existing (1985) stationary source NOx emitters were
evaluated to determine control costs and emission reductions for
different control levels proposed by the Waxman and Mitchell
bills. Effort in the N0X analysis focused on developing current
cost equations for RACT (low N0X burner) and BACT (SCR) level
controls. This chapter describes how the 1985 NEDS point source
file and the control cost equations were organized to perform the
NQX portion of this analysis.
A. EMISSION INVENTORY
Preparing the 1985 NEDS point source emission inventory for
this analysis involved two major tasks (after all non-NOx
emitters were removed from the file). The first task was to sort
the data by MSA/CMSA and indicate for each source whether it was
in a moderate, serious, or severe nonattainment area or in an
attainment area. Secondly, data file information for boilers had
to be organized differently than that for other sources for
costing purposes because boilers burn more than one type of fuel
in most instances. Thus, it was necessary to identify a primary
fuel for each multiple fueled boiler. The primary fuel was
estimated by establishing a hierarchy of fuel types, and choosing
from this hierarchy the likely primary fuel.
To gain some perspective on ozone nonattainment area N0X
emissions, Figure IV.1 summarizes how much of the 1985 point
source N0X emissions are in ozone attainment versus nonattainment
areas and in different N0X emitting source categories. Of the
9.6 million tons of NOx emitted by point sources in 1985, almost
70 percent is in attainment areas and would not be affected by
any of the bill provisions to control N0X emissions. Figure IV.1
also shows that utility boilers are the predominant point source
of N0X emissions in both areas. Industrial boilers are also a
major NQX source in both attainment and nonattainment areas. Gas
turbines and refinery process heaters are the only source types
7 0
-------
Figure IV. 1
1985 Point Source NOx Emissions by Source Type
Attainment Area vs Nonattainment Area
Utility Boilers
Industrial Boilers


Gas Turbines

IC Engines
Refinery Heaters "ft
All Others
S"
Attainment
Total
NNN Nonattainment
Total
0	1	2	3	4	5	6
NOx Emissions (million tons per year)
Figure IV. 2
1985 Point Source NOx Emissions by
Ozone Nonattainment Area Severity
NOx 4
Emissions
(million tons) ^
Attainment Moderate Serious
Severe
7 1
-------
with greater emissions in nonattainment areas than in attainment
areas.
Control cost equations were developed for five source types,
whose emissions are delineated in Figure IV.1. Almost 90 percent
of the point source NOx emissions in nonattainment areas are from
source types for which cost and control information is available.
Figure IV.2 shows how 1985 point source NOx emissions differ
according to the severity of the ozone nonattainment area. Of
the 30 percent of point source N0X emissions from nonattainment
area sources, the emissions are evenly distributed among
moderate, serious, and severe nonattainment areas.
B. NOx CONTROL COST EQUATIONS
A set of equations was developed to provide an estimate of
the costs associated with implementing the N0X control measures
listed in the proposed bills and policies examined. No attempt
was made to develop control cost equations for all types of
stationary source N0X emitters. The sources of greatest concern
for this model were those contributing significant amounts of N0X
emissions and for which N0X control techniques have been
demonstrated with available cost and control information.
The options for controlling N0X emissions from stationary
sources fall into two general categories: Reasonably Available
Control Technologies (RACT) and Best Available Control
Technologies (BACT). The RACT level of N0X control is typically
some form of combustion modification yielding an intermediate
level of control. One example of this control type is low N0X
burners (LNB) which are designed to reduce NOx emissions by
altering the way the air and fuel mix during combustion so that
NOx formation is inhibited. BACT level controls, on the other
hand, are primarily post-combustion control devices, such as SCR,
and produce a stringent level of control. SCR can achieve
between 80 and 90 percent N0X reduction by selectively reducing
the N0X in the flue gas to nitrogen by reacting it with ammonia
over a metal catalyst.
7 2
-------
For each source category, one N0X control method was chosen
from the RACT level control options and one from the BACT level
control options. SCR was chosen as the BACT control device for
each of the source groupings except for gas turbines for which
SCR was used in conjunction with water injection. At the RACT
level, LNB was the desired control method because of its proven
ability to remove significant amounts of N0X at a relatively low
cost. In cases where no cost information was available on LNB,
the best available option in the intermediate control range was
chosen. Control cost equations were developed for the following
source categories of N0X emitters:
.	utility boilers,
.	industrial boilers,
.	internal combustion engines,
.	gas turbines, and
.	process heaters.
The NEDS source classification codes (SCC) belonging to each of
these source categories are listed in Appendix A.
Within each source category, data were generally available
in the literature for the capital costs and the O&M costs of
controlling two or more source sizes typical for that category.
All costs were converted to 1985 dollars using the Chemical
Engineering economic indices (Chem. Engr., 1988). The desired
object was to put the cost information in the following equation
form:
y = axb
where
y = capital or O&M cost (1985$)
x = boiler design capacity (MMBtu/hr) for boilers and either
design rate (SCC units/hr) or operating rate (SCC
units/yr) for other source types
a,b = equation constants
The exponent of the equation indicates the degree to which
economies of scale exist. For all of the capital cost equations,
b is less than 1. This indicates nonlinearity in the costs due
to a savings for large pieces of equipment. The O&M costs, on
the other hand, are more nearly linear since the operating costs
are typically proportional to the design or operating rate of the
73
-------
unit, with no economies of scale. A variance from 1.0 in the O&M
cost equation exponent indicates the presence of a fixed cost
component.
A description of how the cost equations were derived is
given in Appendix A. The resultant equations for the control of
N0X emissions from stationary sources are listed in Tables IV.1
and IV.2. For boilers, the equation variable is the boiler
design capacity, given in MMBtu/hr. For the other source types,
the equation variables used are the maximum hourly design rate or
the annual operating rate of the unit. These variables are
defined by their SCC units (Stockton and Stelling, 1987). SCC
units are assigned according to the production variable of a
process responsible for its emissions. For a utility boiler, the
SCC units would be the amount of fuel burned (e.g., tons of coal
or barrels of oil}. For an industrial process, emissions would
be primarily determined by amount of raw material or amount of
product produced (e.g., tons of pulp produced by a pulp and
paper mill).
C. USE OF DEFAULT VALUES
The 1985 NEDS point source file has many instances of
missing boiler design capacities, design rates, and operating
rates, as well as many instances where these variables are
incorrectly listed as zero. Nevertheless, the MOx emission rate
is almost always listed. To prevent the omission of a large
number of units from the N0X cost analysis, a default cost-
effectiveness value in $/ton N0X removed was calculated for each
of the equations. By multiplying this default value by the N0X
emission rate and the control efficiency, an estimate of the
annual cost was obtained. To calculate the default values, a
typical or average sized unit was chosen to represent each source
category. These source sizes are listed in Table IV.3. Using
the equations 1isted in Tables IV.1 and IV.2, the net annual cost
for each of these units was calculated, using a capacity
utilization factor of 65' percent to calculate the annual
operating rate from the design rate. The amount of NOx which
74
-------
Table IV.1
NOx Control Cost Equations for Utility and Industrial Boilers
Capital Cost
Equations
Operating & Maintenance
Cost Equations
Source Type
Control
Device
Coefficient Exponent Coefficient Exponent
Control
Eff. X
Ui
Utility Boilers
sz sssssassassissszssssscssiass:
issaasssass'ss
rsiatsssessssssss
assssssaas
¦sassssstm-.
PC - Wall/Opposed
LNB
7,860
0.72
393
0.72
50
PC - Tangential
LNB
232,400
0.40
11,620
0.40
50
Residual Oil
SCA
10,480
0.62
600
0.84
42
Gas
FGR
6,610
0.43
450
1.00
31
Stoker
LEA
3,730
0.44
-67
1.11
21
Coal
SCR
292,400
0.60
4,500
1.00
80
Oil/Gas
SCR
265,800
0.50
2,370
1.00
80
Default
Cost Per Ton
87
232
353
983
-525
2911
3120
Industrial Boilers
Pulverized Coal
SCA
1,910
0.70
186
0
96
36
2198
Stoker
LEA
3,730
0.44
-67
1
11
21
-337
Residual Oil
SCA
10,480
0.62
600
0
84
42
827
Distillate Oil
LEA
3,960
0.36
-690
1
00
36
-4592
Gas
FGR
6,610
0.43
450
1
00
31
1025
Coal
SCR
147,900
0.70
4,600
0
95
80
3278
Oil/Gas
SCR
134,450
0.60
2,425
0
95
80
3667
NOTES: All equations are of the form COST * COEFFICIENT*(BOILER DESIGN CAPACITY)~EXPONENT
Units for BOILER DESIGN CAPACITY are in HMBtu/hr
All costs are in 1985 dollars
-------
Table IV.2
NOx Control Cost Equations for IC Engines, Gas Turbines, and Process Heaters
SOURCE
IC Engines
Gas
Oil
Gas-
Oil
CONTROL
METHOD

Gas Turbines
J! Gas
Oil
Gas
Oil
CAPITAL COST EQUATIONS
O&M COST EQUATIONS
CONTROL
EPF(Z)
Change A/F Ratio
Change A/F Ratio
SCR
SCR
Water Injection
Water Injection
SCR+Water Injection
SCR+Water Injection
0
0
8»802,000*(DESRATE)*0.86
1,556,000*(DESRATE)*0.86
1,393,000*(DESRATE)"0.52
508,000*(DESRATE)"0.52
10,031,000*(DESRATE)A0.74
2,283,000*(DESRATE)'0.74
574*(OPRATE)	30
65.8*(0PRATE)	30
131*(OPRATE)+5,355»000*(DESRATE)	80
18.1*(0PRATE) + 714,000*(DESRATE)	80
174*(OPRATE)	70
22.1*(0PRATE)	70
179*(OPRATE) ~1,700,000*(DESRATE)	94
23.1*(0PRATE) + 227,000*(DESRATE)	94
DEFAULT
COST PER TON
1126
935
964
936
1560
1020
3730
2480
Process
Gas
Oil
Gas
Oil
Heater
SCA
SCA
SCR
SCR
47,260*(DESRATE)~0.67
12,830*(DESRATE)A0.67
5,774,000*(DESRATE)* 0.60
1,780,000*(DESRATE)A0.60
-65,100*(DESRATE)
-9,300*(DESRATE)
221*(OPRATE)
29.8*(OPRATE)
45
45
90
90
-306
-110
7810
2760
NOTES: DESRATE is the maximum design rate in SCC units per hour
OPRATE is the operating rate in SCC units per year
All costs are in 1985 dollars
-------
Table IV.3
Default Cost per Ton Values for NOx Emitters
Source Type
RACT Level Control
Utility Boiler,
Wall/Opposed
Utility Boiler,
Tangential
Utility Boiler,
Stoker
Utility Boiler
Utility Boiler
Industrial Boiler
Industrial Boiler
Stoker
Industrial Boiler
Industrial
Industrial
IC Engine
IC Engine
Gas Turbine
Gas Turbine
Process Heater
Process Heater
Boiler
Boiler
Primary Fuel
Pulverized Coal
Pulverized Coal
Coal
Residual Oil
Natural Gas
Pulverized Coal
Coal
Residual Oil
Distillate Oil
Natural Gas
Natural Gas
Oil
Natural Gas
Oil
Natural Gas
Oil
Average Design
Rate Used*
5, 250
5, 250
5, 250
5, 250
5,250
250
250
250
250
250
>.0214
0. 15
0. 15
1. 125
0. 066
0.463
Default
Cost Per Ton
($/ton NOx)
87
232
-525
353
983
2 , 198
-3 3 7
827
•4 , 592
1, 025
1,126
935
1, 560
1, 020
-306
-110
BACT Level Control
Utility Boiler
Utility Boiler
Industrial Boiler
Industrial Boiler
IC Engine
IC Engine
Gas Turbine
Gas Turbine
Process Heater
Process Heater
Coal
Oil/Natural Gas
Coal
Oil/Natural Gas
Natural Gas
Oil
Natural Gas
Oil
Natural Gas
Oil
5, 250
5, 250
250
250
0.0214
0.	161
0.15
1.	125
0. 066
0.463
2,911
3 , 120
3 , 278
3,667
964
936
3,730
2 , 480
7,810
2 , 760
*Des ign Rate for boilers is in MMBtu/hr
Design Rate for other source types is in SCC units/hr
77
-------
would be reduced each year was calculated using the control
efficiencies listed in Tables IV.1 and IV,2 and published
emission factors (Stockton and Stelling, 1987). The resultant
default costs per ton are also listed in Tables IV.1 and IV.2.
Establishing default values was essential for performing a
NO cost analysis for Texas sources. Design capacities and
operating rates are confidential data items in the Texas emission
inventory system and are, therefore, not subrni 11 ed to EPA•
-------
V EPA NONATTAINMENT POLICY ANALYSIS
Estimates of the costs of the proposed EPA policy were of
special interest for two reasons: it was necessary to know what
additional cost might be incurred to control emissions when
compared with the pre-1988 ozone program, and the costs of the
proposed policy were used as the baseline for estimating costs
for the Congressional bills analyzed. Primary differences
between the current (pre-1988) EPA ozone policy and the proposed
policy are outlined in Table V.l. All the provisions mentioned
were modeled explicitly here except the possibility of N0X
controls. while the costs and benefits of NOx controls as part
of an ozone reduction strategy are not explicitly modeled, it is
recognized that NOx controls may be cost effective in helping
some areas reach attainment for ozone.
The emission reductions and costs associated with the
proposed EPA ozone policy were calculated as those above what
would be achieved via the current policy. Thus, the first step
in the analysis was to simulate the costs and future emission
levels for each MSA under the provisions of the current EPA
policy. Projected 1995 costs of the pre-1988 EPA ozone policy
are presented in Table V.2 for four different policy provisions.
The increase in the cost of planned I/M programs between 1995 and
2000 results from an increase in the number of vehicles being
inspected.
Table V.3 summarizes ERCAM net annual cost estimates for
1995 and 2000 by VOC control measure. Among the national
measures, architectural surface coating is listed as having no
cost. Switching from solvent borne to waterborne coatings is
estimated to be at no cost. The negative numbers for an auto
body ret inishing CTG represent cost savings.
The ERCAM simulation of the cost of the EPA policy is
performed in two parts. First, ail of the explicit (mandated)
provisions of the policy are modeled. Then, each MSA is
evaluated with respect to the 3 percent per year and attainment
requirements to see if these have been met. Additional
79
-------
Table V.1
Key EPA Ozone Policy Provisions
1.	Nonattainment area boundaries expand to equivalent of MSA or
CMS A
2.	National measures include the following:
A.	RVP control
B.	Onboard VOC control
C.	TSDF control
D.	POTW control
E.	Architectural surface coating control
F.	Other possible national measures (CTGs) - autobody
refinishing
3.	Annual 3 percent emission reduction requirement until
attainment
4.	For NMOC/NOx ratios above 10:1, areas are required to
consider NOx control as part of its ozone control strategy
5.	Enhanced I/M in all areas with design values >.0.16 ppm for
ozone
Note that while some of the above provisions have been
assumed by this analysis to be adopted as final for modeling
purposes, they are not proposed as explicitly in the policy.
-------
Table V.2
Costs of Pre-1988 EPA Ozone Policy*
National Summary
Stationary
Sources
199 5
Cost
(million $)
$ 878
449
3, 137
2000
Cost
(million $)
$1,213
449
4 ,407
Provisions
Included
NSPS
Non-CTG RACT
LAER for new
sources
expected to
be > 100 tpy
Motor
Vehicles
Totals
712
$5,176
926
$6,995
Planned I/M
programs not
in effect in
1985
* Costs presented in this table are not historical control costs,
and therefore do not capture costs of meeting motor vehicle
emission standards or costs of stationary source controls in
place in 1985. Current air pollution control expenditures are
approximately $33 billion per year. More than one-third of
this cost is for motor vehicle emission controls.
81
-------
Table V.3
EPA Ozone Policy Costs*
National Summary
Net Annual Costs
(billion$)
National Measures
Architectural Surface Coating
Hazardous Waste Treatment,
Storage, and Disposal
Facilities (TSDF)
Publicly Owned Treatment
Works (POTW)
RVP Control
Onboard Control
New CTGs
Autobody Refinishing
Additional Measures
Bringing all Existing Sources
into compliance with SIPs
CTGs Not Already in Place
Expansion of Nonattainment Area
to MSA or CMSA Level
Enhanced I/M in Serious and Severe
Nonattainment Areas
Discretionary Controls Applied
to Serious and Severe
Nonattainment Areas**
Cost for All Areas to Meet
3 Percent Line or Attain
(at $2,000 to $10,000 per ton)
Total
1995
$0
0.82
0.02
0. 24
0. 19
-0.41
0.61
0.10
0.01
0.68
0.81
1.11 to 5.57
$4.17 to $8.63
2000
0
0.90
0. 02
0.27
0. 19
-0.46
0.61
0.10
0.01
0.76
4 . 20
2.3 4 to 11.72
$8,94 to $18.32
* All costs are incremental to the cost of the pre-1988 ozone
policy.
** Discretionary controls applied in this analysis are listed in
Table V.4
82
-------
discretionary VOC controls are applied in the second ERCAM
simulation to areas not meeting their emission reduction targets.
Because controls in the ERCAM scenario file cannot be put on
individual areas, discretionary controls were put on all serious
and severe nonattainment areas. All severe and most serious
nonattainment areas had difficulty meeting their 1995 and 2000
targets. The cost of discretionary controls in serious and
severe nonattainment areas is estimated to be $0.81 billion in
1995 and $4.20 billion in 2000. The dramatic increase in
discretionary control costs from 1995 to 2000 results from
serious and severe nonattainment areas beginning to see market
penetration of methanol-fueled vehicles into the vehicle fleet
sometime shortly after 1995, such that 30 percent of vehicles in
these areas are methanol-fueled by 2000. The $3.18 billion cost
estimate for alternative fuels to all vehicle types reflects an
assumed 10 cent per gallon price difference between methanol and
gasoline. (Some forecasts show no expected price difference
between the two fuels. If the latter assumption is used, the
cost of methanol-fueled vehicles is negligible.) Estimated costs
for all the discretionary controls included in the analysis of
the EPA policy are shown in Table V.4. The "more stringent
existing source controls" option shown in Table V.4 refers to
applying the most efficient control technique to each cost pod.
Only five source categories were found that were not already
required to control to the highest levels.
Even after all discretionary controls (that can be
identified) are applied, not all areas have met their
attainment/progress requirements.
Because emission reduction targets cannot be achieved in all
areas even with currently available controls applied, it was
necessary to assign a cost to the residual tons of VOC beyond
those for which explicit cost equations exist. An analysis of
this issue proceeded in four steps: (1) determining the source
categories that have opportunities for further control in 1995
and 2000, (2) assessing costs of possible control technologies
that might be applied to these sources, (3) estimating how the
83
-------
Table V.4
Discretionary Controls for EPA Policy
Cost
(billion $)
Option
RACT to 50 tpy
More Stringent Existing
Source Controls*
Industrial Solvents
Consumer Solvents
Enhanced I/M on HDGV
Railroad Engines
Bakeries
Alternative Fuels to Fleet
Vehicles
Alternative Fuels to All
Types
1995
0.04
0.06
0.16
0.24
0.04
0.05
0.03
0. 18
2000
0.05
0. 07
0.21
0.30
0.05
0.07
0.04
0.23
3.18
* Affects Synthetic Organic Chemicals Manufacturing Industry
(SOCMI) fugitives, petroleum refinery fugitives, cellulose
acetate manufacture, paper surface coating, and aircraft surface
coating
Note: Costs are for all serious and severe nonattainment areas
not meeting attainment/progress requirements
84
-------
candidate residual tons to be reduced might be allocated among
the controllable categories, and (4) using the results of steps 1
through 3 to estimate an average cost per ton reduced.
Table V.5 provides information about which major source
categories are candidates for additional emission reductions in
the projection years. The 1995 and 2000 VOC emission totals are
from an EPA policy simulation after discretionary controls have
been applied. Percentage reductions are from 1985 uncontrolled
levels. Table V.5 suggests that controllable VOC emissions will
be concentrated in three categories: solvent use, consumer
solvents, and mobile sources. These three categories constitute
68 percent of the 1985 emissions inventory and 77 percent of the
year 2000 emissions inventory. These categories have relatively
low levels of control as well.
Among the remaining categories in the emission inventory,
point sources, service stations and miscellaneous point sources
are essentially completely controlled. Any further control would
have to come from improved capture and ducting systems,
incineration, and flaring technologies which would be at least as
expensive as those discussed below for solvents. New area source
categories are also subject to very little additional control,
because this consists of several sources (TSDFs) that are or will
be controlled to the maximum extent possible, and several sources
(forest fires) that are essentially uncontrollable.
The analysis concentrates, therefore, on the kinds of
controls that might be applied to (industrial) solvent use,
consumer solvents, and mobile sources. Incineration is the only
method of getting consistently high control of emissions from
solvent use. Options that are less universally applicable
include switching to conforming coatings (at little or no cost)
or switching to water soluble cleaning materials (where controls
in the 90 percent plus range cannot be guaranteed). Costs for
the latter two options are lower than incineration costs, but
control levels may not be high enough to ensure that VOC
reduction targets are met. Therefore, cost estimates were made
for a hypothetical incinerator on a 10 and a 2 5 ton per year VOC
85
-------
Table V.5
Summary of VOC Emissions by Category
for Ozone Honattainment Areas
VOC Emissions 1985 " VOC Emissions 1995 VOC Emissions 2000
Source	Percentage	Percentage	Percentage
Category	(tons) Reduction	(tons) Reduction	(tons) Reduction
Point Sources
792,339
82
406,147
93
473,920
92
Solvent Use
1,743,165
12
1 ,134,202
57
1,248,270
57
Service Stations
333,703
38
35,279
95
38,890
95
Consumer Solvents
682,629
0
735,617
20
811,811
20
Mobile Sources
3,429,299

1,828,113

1,818,079

New Area Source Categories
1,242,182
0
513,038
68
547,674
69
Misc. Point Sources
391,322
87
284,186
93
319,806
92

sssssss==
8,614,639
xs.
44
ssssssias
4,936,582
72
ssssassss
5,258,450
3S3S
73
Proj ection year (1995 and 2000) emissions are those after all EPA policy provisions and discretionary
controls have been applied. Percentage reductions are from 1985 uncontrolled levels.
-------
emitter. The Economic Analysis Branch (EAB) Control Cost Manual
(U.S. EPA, 1986b) was used to estimate the costs for thermal and
catalytic incinerators handling a waste gas flow rate of 5,000
cubic feet per minute (the smallest source size to which the
equations can be applied). The annual cost derived for this
source was used to calculate dollar per ton values assuming a 90
percent reduction. The cost for thermal incineration ranged from
$2,550 per ton for a 25 ton per year source to $6,390 per ton for
a 10 ton per year source. For catalytic incinerators, the
resulting costs were $2,310 per ton and $5,770 per ton for a 25
and 10 ton per year source, respectively. These four estimates
of incineration costs based on two technologies and two source
sizes can be averaged to obtain a value of $4,255 per ton.
There are no good estimates of the cost of controls on
consumer solvents. Industry sources claim costs as high as
$30,000 per ton. The basis for these numbers is being examined
by EPA to see if they are at all reasonable. For purposes of
this analysis, the considerably more cautious assumption has been
made that these controls can be accomplished at an average of
$2,000 per ton.
Additional mobile source reductions must come primarily from
one of two sources: Transportation Control Measures (TCMs), or
additional reductions through greater use of alternative fuels,
such as methanol or natural gas.
Table V.6 is information supplied by EPA's OAQPS on a TCM
analysis of a typical SIP (Kansas City). A simple average of the
measures for which cost data are supplied yield a figure of
$107,000 per "ton. Excluding all measures costing in excess of
$100,000 per ton, this average becomes $17,500 per ton. Figures
from Dallas and Tarrant counties for which tons as well as costs
are available are $21,500 and $6,300 per ton, respectively. The
average of these estimates is $15,000. If it is assumed, to be
conservative, that two thirds of this cost per ton could be
offset by other benefits (such as CO control), the net cost is
about $5,120.
87
-------
Table V.6
Transportation Control Measures Analyzed
for Kansas City
Measures
Cost Effectiveness and
Main Drawbacks	
2	.
3	.
4	.
5.
6.
7	.
8	.
9	.
Short-Range Public Transit Improvements
Pedestrian/Transit Mall
Light Rail Transit
Expand Regional Rideshare Program
Encourage Bicycling to Work Through
Employer-Based Program
Encourage Commuters to Use Transit or
Carpools One Day Each Week
$ 62,060/Ton of
reduction (TOR)
$491,800/TOR
$485,720/TOR
$ 2 , 777/TOR
$ 9,910/TOR
Voluntary Program
Encourage the Use of Variable Work Schedule $ 2,8IS/TOR
Encourage the Use of the 4 Day Work Week Voluntary Program
Improve Traffic Signalization
10.	Improve Highway Surveillance and
Information
11.	Truck Delivery Restrictions in Central
Business Districts
12.	Institute More One-Way Streets, Where
Feasible
$146,OOO/TOR
Increases VOC
Through 1990
$ 7,941/TOR
Only 3.1 TPY
reduction
13.	Switch Traffic Control Devices to Flashing Only 0.6 TPY
Mode During Off-Peak Hours	reduction
14.	Prohibit Left Turns on Congested Streets
15.	Adjust Speed Limits to Reduce Congestion
on Selected Streets
16.	Reduce Amount of On-Street Parking and
Improve Enforcement of On-Street Parking
Controls in Downtown Areas and Along
Congested Streets
Only 3.l TPY
reduction
Safety Problems
$16,3 8 0/TOR
88
-------
Table V.6 (continued)
Transportation Control Measures Analyzed
for Kansas City
17.	Implement Education Program on Vehicle
Idling
18.	Restrict Truck Idling
19.	Encourage the Substitution of
Communications for Transportation
$ 2,151/TOR
$ 24,500/TOR
This will occur
without public
sector
involvement.
20. Encourage Home Delivery of Goods
$ 31,710/TOR
Source: Kansas City State Implementation Plan
89
-------
The range of options on use of alternative fuels is
reasonably well captured by conversion of fleet vehicles to
natural gas, and use of methanol. Natural gas conversion can be
accomplished for $2,000 to $4,400 per ton. The cost of methanol
controls depends heavily on the relative cost of gasoline and
methanol. Assuming a $0.10 per gallon fuel price difference and
a $400 per gasoline powered vehicle capital cost difference,
methanol use would yield reductions at $20,000 per ton.
It seems unlikely that TCMs will be used to achieve the bulk
of the reductions required of motor vehicles. Localities have
been reluctant to rely on them, and when they have, the measures
account for relatively small reductions in the inventory. If it
is assumed that TCMs would account for no more than a third of
all required residual tons gained from mobile sources, the three
mobile source numbers {$5,120 for TCMs, $3,200 average for
natural gas, and $20,000 for methanol) can be averaged to give an
average cost of $9,440 per ton for mobile source reductions.
To allocate residual tons to controllable categories, this
analysis assumes that the residual tons required to attain will
be drawn from the three "controllable" categories in proportion
to the relative share of each category in the total
"controllable" inventory. If, for example, the controllable
inventory (i.e., the sum of the tons in the controllable
categories) consisted of 30 percent solvent use, 20 percent
consumer solvents, and 50 percent mobile sources, these
proportions could be used to allocate required residual tons to
these three categories.
This general allocation principle can be improved slightly
by accounting for the fact that the relative proportion of the
controllable inventory may shift somewhat over time. Table V.5
suggests, for example, that solvent use is increasing as a
percentage of the inventory over time, while mobile sources are
declining. To reflect this fact, this analysis averaged the
proportions of the 1995 and 2000 "controllable" inventory. This
procedure yields an allocation of residual tons to controllable
90
-------
categories as follows: 33 percent to solvent use, 20 percent to
consumer solvents, and 47 percent to mobile sources.
The results in the previous steps can be combined by
multiplying the costs per ton for each controllable category by
the proportions listed above. This yields an average of $6,200
per ton. Given all the uncertainty in the assumptions, the
possibility of new lower cost control technologies being
developed, balanced against the possibility of much higher costs
for consumer solvent controls, a range of $2,000 per ton to
$10,000 per ton was used to estimate costs for controlling
remaining VOC emission after all available controls have been
applied. The $2,000 per ton scenario reflects a situation where
much of the reductions of residual tons can be achieved through
switching to conforming coatings, switching to water soluble
cleaning materials, reducing per motor vehicle VOC emissions even
further through cost effective methods, and the development of
new, less polluting technologies. At the other end of the
spectrum, the $10,000 per ton case reflects a scenario where
there are problems achieving emission reductions from consumer
solvent use, higher levels of control are needed to reach
attainment than the less expensive, moderate efficiency controls
can achieve, and time is too short for new technologies to
penetrate the market in significant quantities.
Model results show that besides the alternative fuel cost
already discussed above, the biggest difference between 1995 and
2000 cost estimates is in the cost to meet the 3 percent per year
reduction requirements or attain. Because all of the controls
except alternative fuels have been imposed by 1995, new source
growth overtakes reductions in emission rates to show a net
increase in VOC emissions between 1995 and 2000 for many areas.
Onboard vehicle evaporative VOC controls will continue to provide
net emission reductions past 1995, but most other controls,
including motor vehicle tailpipe emission standards, may reduce
per source emissions, but not the total emissions for a category.
Thus, many areas which are predicted to meet their .1995 emission
targets exceed their 2000 targets. MSA-level calculations of VOC
91
-------
emissions and 3 percent reduction versus attainment targets are
illustrated in Tables V.7 and V.8. The equation used to
calculate 1995 3 percent line emission targets was as follows;
0.82 (Revised 1985 Base) - Federal Measure Reductions
where the revised 1985 base emissions are those after all current
SIP requirements are complied with. Federal measures for which
areas get no reduction credit {toward the 3 percent reduction
requirement) include Federal motor vehicle emission standards and
existing I/M programs, plus RVP limits and onboard vehicle
evaporative emission controls. Note that Federal measures do not
include TSDF controls or municipal landfill controls in this
simulation,
Year 2000 3 percent line emission targets shown in Table V,8
were estimated using the equation shown above with the 0.82
coefficient changed to 0.67 to reflect five more years of
emission reductions at 3 percent per year. Under these
assumptions, the proposed EPA policy would force all ozone
nonattainment areas except two, Los Angeles and New York, to
attain by 2000.
-------
Table V.7
Additional Reductions Needed to Meet Attainment/3 Percent Line
1995 EPA Policy
VOC Emissions (tons)
Additional



Revised
Federal
Attainment
3 Percent

Redact ion
CMSA Name
1985
Baseline
Measures
Target
Target
1995*
Needed
Massachusetts
384
796
353
163
80
142
246,269
209
452
299,398
53,129
A1lentown, PA-NJ
45
639
41
610
9
510
41,988
2 4
610
32,531
0
Atlanta, GA
181
599
162
577
49
241
123,487
84
072
121,551
0
Atlantic City, NJ
16
891
16
572
5
047
10,135
8
542
11,526
1,391
Bakersfield, CA
42
360
34
212
10
344
25,416
17
710
23,710
0
Baltimore, MD
136
178
129
262
30
633
99,410
75
362
98,076
0
Baton Rouge, LA
84
069
41
819
8
884
43,716
25
408
40,536
0
Beaumont, TX
130
875
58
353
7
911
66,746
39
938
57,184
0
Birmingham, AL
62
086
58
317
14
482
57,119
33
338
46,046
0
Bradenton, FL
9
527
8
883
2
572
8,765
4
712
8,013
0
Charleston, wv
23
607
19
505
3
795
21,718
12
199
18, 198
0
Charlotte, WC-SC
77
904
69
787
18
583
71,672
38
642
64,818
0
Chattanooga, TN-GA
40
689
32
394
7
393
37,434
19
170
36,429
0
Chicago CMSA
529
572
4 58
393
102
165
222,420
273
717
316,346
42 ,629
Cincinnati CMSA
121
743
103
335
26
941
70,611
57
794
70,018
0
Cleveland CMSA
190
252
178
864
41
918
150,299
104
750
135,356
0
Dallas CMSA
266
233
248
808
74
931
141,103
129
092
177,834
36,731
Dayton, OH
71
385
64
343
14
603
65,674
38
158
50,131
0
Denver CMSA
140
469
135
251
42
154
129,231
68
752
102,177
0
Detroit CMSA
318
674
281
305
65
314
293,180
165
3 56
213,016
0
El Paso, TX
35
069
32
447
10
687
19,639
15
920
19,580
0
Erie, PA
17
710
16
386
3
500
16,293
9
937
12,888
0
Fresno, CA
37
700
35
174
9
172
20,358
19
671
26,431
6, 07 3
Gadsden, AL
9
707
6
633
1
754
8,930
3
685
4 ,820
0
-------
Additional Reducti
Revised
CMSA Name	1985 Baseline
Grand Rapids, MI
47-
777
45
195
Harrisburg, PA
33
779
32
969
Houston CMSA
370
531
277
022
Huntington, WV-KY-OH
31
976
26
313
Indianapolis, IN
121
961
79
063
Jacksonville, FL
55
784
53
176
Janesville, WI
12
817
9
119
Kansas City, MO-KS
125
335
104
167
Lake Charles, LA
26
294
14
532
Longview, TX
22
015
16
821
Los Angeles CMSA
824
055
789
525
Louisville, KY-IN
77
299
65
417
Memphis, TN
65
116
55
227
Miami CMSA
157
426
147
539
Milwaukee CMSA
109
465
100
599
Minneapolis, MN-WI
189
531
158
431
Modesto, CA
20
050
18
271
Muskegon, MI
15
822
10
900
Nashville, TN
73
025
59
022
New York CMSA
887
534
852
933
Norfolk, VA
78
350
67
694
Philadelphia CMSA
344
747
303
863
Phoenix, AZ
114
562
108
336
Pittsburgh CMSA
122
802
110
833
Table V.7
Needed to Meet Attainment/3 Percent Line
1995 EPA Policy
VOC Emissions (tons)
Additiona1
Federal Attainment 3 Percent	Reduction
Measures	Target Target	1995*	Needed
8
900
43
955
28
160
41
840
0
8
681
31
077
18
354
24
578
0
67
482
129
686
159
676
255
757
96,081
4
774
25
581
16
803
23
457
0
18
311
112
204
46
521
78
735
0
12
540
44
627
31
064
43
317
0
2
064
11
792
5
414
7
480
0
22
733
105
281
62
684
83
939
0
2
781
18
406
9
135
12
986
0
4
410
14
750
9
383
14
011
0
168
297
255
457
479
114
581
450
102,336
15
153
51
790
38
489
50
417
0
12
681
41
674
32
605
41
982
308
44
080
127
515
76
902
108
024
0
20
860
59
111
61
631
72
096
10,465
40
716
126
986
89
197
121
867
0
5
438
12
030
9
544
13
166
1, 136
2
226
12
658
6
712
8
856
0
14
223
58
420
34
175
48
196
0
192
644
292
886
506
761
586
087
79,326
17
079
62
680
38
430
53
516
0
68
265
203
401
180
903
215
599
12,198
30
242
67
592
58
594
83
762
16,170
27
323
112
978
63
560
77
859
0
i
-------
Table V.7
Additional Reductions Needed to Meet Attainment/3 Percent Line
1995 EPA Policy
VOC Emissions (tons)
Add it ion.a L
Revised Federal Attainment 3 Percent	Reduction
CMSA Name	1985 Baseline Measures	Target	Target	1995*	Needed
Portland CMSA
95
120
89
309
19
776
87,510
53,457
75,626
0
Providence, RI
57
881
53
631
11
156
38,780
32,821
40,026
1,246
Reading, PA
22
933
20
726
4
611
21,098
12,384
16,324
0
Richmond, VA
66
718
53
384
13
467
61,381
30,308
45,906
0
Sacramento, CA
87
463
80
840
23
924
43,732
42,365
55,997
12,265
St. Louis, MO-1L
185
377
153
137
38
216
114,934
87,356
114,257
0
Salt Lake City, UT
72
379
69
377
21
088
49,218
35,801
45,499
0
San Diego, CA
126
559
121
385
34
212
53,155
65,324
89,336
24,012
San Francisco CMSA
375
354
358
562
82
807
180,170
211,214
261,539
50,325
Santa Barbara, CA
28
805
25
884
6
283
17,283
14,942
19,933
2 , 650
Sheboygan, WI
8
586
8
035
1
707
4, 636
4 , 882
6, 090
1, 2 08
Stockton, CA
25
799
24
089
7
633
17,285
12,120
16,643
0
Tampa, FL
107
549
103
609
29
522
98,945
55,437
80,379
0
Tulsa, OK
54
034
50
281
15
801
49,711
25,429
37,499
0
Visalia, CA
19
478
18
430
4
399
17,920
10,714
14,665
0
Washington, DC
168
375
162
274
49
566
104,392
83,499
98,926
0
West Palm Beach, FL
43
946
43
168
10
534
35,596
24,864
38,253
2, 657
York, PA
28
161
27
175
5
855
25,908
16,428
22,208
0
Yuba City, CA
9
836
9
502
1
834
9,049
5,958
7,952
0
Greater Conn, CMSA
127
499
123
427
29
113
84,149
72,097
89,230
5, 081

8,614,
639
7,690,
585
1,865,
083
5,129,002
4,441,198
5,811,878
557,417
* 1995 VOC after discretionary controls are applied where needed
The reader is cautioned that MSA level results are even more uncertain than national level
results
-------
Table V.8
Additional Reductions Needed to Meet Attainment/3 Percent Line ,
2000 EPA Policy	• / f k
¦i pi 't th 7
VGC Emissions (tons)	^"s,3a J
^	1	Additional
Revised Federal Attainment 3 Percent	Reduction
CMSA Name 1985 Baseline Measures Target Target	2000*	Needed
Massachusetts
3 84,
796
353
163
79
688
246
269
156
931
329
542
8 3,271
Al.lent.own, PA-NJ
4 5
639
4.1
610
9
200
41
988
18
679
34
137
0
Atlanta, GA
181
599
162
577
48
335
123
487
60
592
117
834
0
Atlantic City, NJ
16
891
16
572
5
089
10
135
6
014
12
458
2 , 3 2 3
Bakersfield, CA
42
360
34
212
10
260
25
416
12
662
25
178
0
Baltimore, MD
136
178
129
262
29
680
99
410
56
926
94
163
0
Baton Rouge, LA
84
069
41
819
8
903
43
716
19
116
38
106
0
Beaumont, TX
130
875
58
353
7
796
66
746
31
301
52
784
0
Birmingham, AL
62
086
58
317
14
136
57
119
24
936
50
380
0
Bradenton, PL
9
527
8
883
2
589
8
765
3
363
9
312
547
Charleston, WV
23
607
19
505
3
772
21
718
9
296
18
407
0
Charlotte, NC-SC
77
904
69
787
18
245
71
672
28
512
74
980
3, 308
Chattanooga, TH-GA
40
689
32
394
7
295
37
434
14
409
42
845
5,4 11
Chicago CMSA
529
572
4 58
393
99
450
222
420
207
673
330
915
108,495
Cincinnati CMSA
121
743
103
335
2 6
351
70
611
42
883
73
063
2,452
Cleveland CMSA
190
252
178
864
40
610
150
299
79
229
143
458
0
Dallas CMSA
266
233
248
808
73
818
141
103
92
883
193
586
52,4 83
Dayton, OH
71
385
64
343
14
190
65
674
28
920
52
985
0
Denver CMSA
140
469
135
251
41
184
129
231
49
434
113
796
0
Detroit CMSA
318
674
281
305
63
275
293
180
125
199
223
956
0
El Paso, TX
35
069
32
447
10
423
19
639
11
316
20
533
894
Erie, PA
17
710
16
386
3
417
16
293
7
562
13
642
0
Fresno, CA
37
700
35
174
8
884
20
358
14
683
28
389
8,031
Gadsden, AL
9
707
6
633
1
688
8
930
2
756
5
195
0
-------
Table V.8
Additional Reductions Needed to Meet Attainment/1 Percent Line
2 000 EPA Policy
VOC Emissions (tons)
Addit iona1
Revised Federal Attainment 3 Percent	Reduction
CMSA Name	1985 Baseline Measures	Target Target	2000*	Needed
Grand Rapids, MI
47
7 77
45
195
8
645
4 3
955
21
63 6
45,419
1 , 4 64
Harrisburq, PA
33
779
32
969
8
499
31
077
13
590
26,271
0
Houston CMSA
3 70
531
277
022
66
785
129
686
118
820
293,201
163,515
Huntington, WV-KY-QH
31
97 6
26
313
4
676
25
581
12
954
24,388
0
Indianapolis, IN
121
961
79
063
17
771
112
204
35
201
90,007
0
Jacksonville, FL
55
784
53
176
12
304
44
627
23
324
47,069
2,442
Janesville, WI
12
817
9
119
2
058
11
792
4
052
7,988
0
Kansas City, MO-KS
125
335
104
167
22
050
105
281
47
742
89,308
0
Lake Charles, LA
26
294
14
532
2
815
18
406
6
921
12,916
0
Longview, TX
22
015
16
821
4
361
14
750
6
909
13,544
0
Los Angeles CMSA
824
055
789
525
158
489
255
457
370
493
618,240
2 4 7,747
Louisville, KY-IN
77
299
65
417
14
742
51
790
29
087
48,211
0
Memphis, TN
65
116
55
227
12
685
41
674
24
317
45,753
4 , 079
Miami CMSA
157
426
147
539
43
702
127
515
55
149
119,865
0
Milwaukee CMSA
109
465
100
599
20
173
59
111
47
228
75,477
16,366
Minneapolis, MN-WI
189
531
158
431
39
682
126
986
66
467
116,956
0
Modesto, CA
20
050
18
271
5
441
12
030
6
801
14,198
2 , 168
Muskegon, MI
15
822
10
900
2
141
12
658
5
162
9,339
0
Nashville, TN
73
025
59
022
14
127
58
420
25
418
52,056
0
New York CMSA
887
534
852
933
187
054
292
886
384
411
617,939
233,528
Norfolk, VA
78
350
67
694
16
720
62
680
28
635
59,458
0
Philadelphia CMSA
344
747
303
863
66
126
203
401
137
462
227,290
23,889
Phoenix, AZ
114
562
108
336
30
232
67
592
42
353
93,179
25,587
Pittsburgh CMSA
122
802
110
833
26
496
112
978
47
762
82,237
0
-------
Table V.8
Additional Reductions Needed to Meet Attainment/3 Percent Line
2000 EPA Policy
VOC Emissions (tons)
Additiona1



Revised
Federal
Attainment
3 Percent

Reduct ion
CMSA Name
1.98 5
Baseline
Measures
Target
Target
2000*
Needed
Portland CMSA
95
,120
89
309
19,472
87,510
40,365
82,716
0
Providence, RI
57
881
53
631
10,808
38,780
25,125
42,572
3 ,792
Reading, PA
22
933
20
726
4,451
21,098
9,435
17,365
0
Richmond, VA
66
718
53
384
13,185
61,381
22,582
51,577
0
Sacramento, CA
87
463
80
840
23,515
43,732
30,648
58,985
15,253
St. Louis, MO-IL
185
377
153
137
37,072
114,934
65,530
108,186
0
Salt Lake City, UT
72
379
69
377
20,655
49,218
25,828
48,542
0
San Diego, CA
126
559
121
385
33,790
53,155
47,538
96,263
4 3,108
San Francisco CMSA
375
354
358
562
78,938
180,170
161,299
278,749
98,579
Santa Barbara, CA
28
805
25
884
6,221
17,283
11,121
21,398
4,115
Sheboygan, WI
8
586
8
035
1,659
4,636
3,724
6,443
1, 807
Stockton, CA
25
799
24
089
7,521
17,285
8, 619
15,895
0
Tampa, FL
107
549
103
609
28,943
98,945
40,475
89,593
0
Tulsa, OK
54
034
50
281
15,510
49,711
18,178
41,405
0
Visalia, CA
19
478
18
430
4,280
17,920
8,068
15,786
0
Washington, DC
168
375
162
274
48,472
104,392
60,252
104,129
0
West Palm Beach, FL
43
946
43
168
10,585
35,596
18,338
42,860
7 ,2 64
York, PA
28
161
27
175
5,656
25,908
12,551
23,685
0
Yuba City, CA
9
836
9
502
1,802
9,049
4,564
8,373
0
Greater Conn. CMSA
127
499
123
427
28,255
84,149
54,441
95,064
1.0, 915
8,614,639 7,690,585 1,816,842 5,129,002 ,j35,85Q 6,179,539 1,172,835
* 2000 VOC after discretionary control are applied
The reader is cautioned that MSA level results are even more uncertain than national level
results
-------
VI MITCHELL BILL ANALYSIS
S. 1894 is a bill sponsored by Sen. George Mitchell (D-ME)
to amend the Clean Air Act to establish new requirements for
areas that have not yet attained health-protective ambient air
quality standards. The legislation would provide new deadlines
for such attainment, delay the imposition of sanctions, attempt
to better protect against interstate transport of pollutants, to
control existing and new sources of acid deposition, and for
other purposes. This bill is organized into five parts, or
titles: Title I -- Requirements for Nonattainment Areas, Title
II -- Acid Deposition Control, Title III — Mobile Source and
Other Federal Controls, Title IV — Ambient Air Quality
Standards, and Title V — Hazardous Air Pollutants. The analysis
reported on in this chapter covers costs and emission reductions
associated with Title I and III.
An outline of the key provisions of the Mitchell bill is
shown in Table VI.1. As is shown in this table, the
nonattainment area (MSA/CMSA level) definition is the same in the
Mitchell bill as it is in the proposed EPA policy. Significant
parts of the Mitchell bill that differ from the EPA policy
include more stringent motor vehicle emission standards,
alternative fuels and engines to a portion of the vehicle fleet,
emission fees, stationary source N0X emission controls, and ozone
transport regions (where some controls are imposed on sources in
attainment areas directly upwind of nonattainment areas).
An explicit 1ist of all the national measures and new CTGs
proposed by the Mitchel1 bill is shown in Table VI.2. National
measures are applied to sources in all areas of the country,
while new CTGs are only applied in nonattainment areas. The
effect of some of the national measures and new CTGs were not
included in this analysis, because the base year emission
inventory did not include explicit emission estimates for those
source types. Those source categories are indicated on Table
VI . 2 .
99
-------
Table VI.5
Outline of Key Mitchell Bill Provisions
1.	Nonattainraent area deftmtion is the same (MSA/CMSA) as in
EPA policy
2.	National level controls include the following:
A.	RVP controls
B.	Onboard control of VOC
C.	Mobile source emission standard changes
-LDV Hydrocarbon  1 tpy emitters install ma*
practicable control by 12/90
-Emission limits for stationary engines and off-Hwy
vehicles as stringent as those for lOVs
4. Establishes ozone.transport regions consisting of the
following states:
(A) CI, 0C, DE, HE, NO, HA, NH, NJ, NY, OH, PA, RI, VT, VA
(not all of NY and VA are included)
{B) i L, IN, MI, WI
100
-------
Table VI.2
Mitchell Bill National Measures and New CTGs
National Measures
Commercial and Consumer Solvents
Architectural Surface Coating
RVP Controls
Onboard Controls
New Mobile Source Emission Standards
Pesticide Application*
Traffic Marking Coatings*
Metal Parts Coating in Military and Aerospace Applications*
New CTGs
Wood Furniture Coating
Autobody Refinishing
Hazardous Waste Treatment, Storage, and Disposal Facilities
(TSDF)
Bakeries
Publicly Owned Treatment Works (POTW)
Coke Oven By-product Plants
Metal Rolling*
socmi Distillation*
SOCMI Batch Process*
Web Offset Lithography*
Plastic Parts Coating*
* Effects of controls on these source categories were not
included in the modeling analysis.
101
-------
Table VI.3 presents a national summary of the incremental
costs of the Mitchell bill, i.e., those costs above what were
estimated to be incurred to comply with the proposed EPA policy.
New motor vehicle emission standards are estimated to cost $1.2
billion per year more than the current set of emission standards.
(Costs of VOC, NOx, and CO control are all included in this
number.) Costs to reach the attainment/progress requirements for
1995 add another $1.5 billion to the estimated 1995 cost of the
bill. S. 1894 requires all moderate ozone nonattainment areas to
attain by 1995. Serious and severe nonattainment areas must
achieve a 55 percent emission reduction or attain, whichever is
less stringent.
It should be noted that when estimating progress toward
attainment or meeting interim reduction requirements of the
Mitchell bill in this analysis, no emission reduction credit is
given for N0X emission reductions or for VOC emission reductions
in ozone transport regions or any upwind area outside the
nonattainment MSA/CMSA. Costs to attain the ozone standard may
be lower if H0X emission reductions reduce ozone production or if
there is less transported ozone. Costs may be higher, however,
in cases where ratios are low.
N0X costs increase dramatically between 1995 and 2000. To
simulate applying RACT to greater than 25 ton per year emitters
in nonattainment areas, a moderate RACT definition has been used,
so 1995 costs are not high. By the year 2000, though, severe
nonattainment. areas will have been required to reduce major
stationary source emissions fay 65 percent or more. RACT level
controls will not achieve this, so SCR or a similar technology at
80 percent to 90 percent control will have to be applied to these
sources. Costs to achieve NOx reductions above 50 percent are
high.
Note also that NGX control requirements will probably
produce some fuel switching. These effects have not been
captured in this analysis. The cost for all areas to attain by
2000 in Mitchell (all areas are effectively required to attain by
2000 by the 5 percent per year reduction requirement) is
102
-------
Table VI.3
Incremental Cost of Mitchell Bill
National Summary
Annual Cost
(billion $}*
1995
2000
1.	RACT Level N0X Controls
2.	SCR Level N0X Controls in
Severe Nonattainment Areas
3.	New Motor Vehicle Standards
4.	Ozone Transport Region
Controls
5.	Consumer Solvent Controls
6.	Stage II in all NA Areas
7.	Enhanced I/M in all NA
Areas
8.	RACT Cutoff to 25 tpy
Sources
9.	Alternate Fuel to Fleet
Vehicles in NA Areas
10.	Wood Furniture Coating
and Bakery CTGs
11.	Cost for all Areas to Attain
or Meet 5 Percent Line
$0.40
1.20
0.61
0.41
0.29
0.51
0.07
0. 10
0. 02
0.40
2.90
1.20
0. 77
0.41
0. 32
0. 56
0. 06
0.09
0.02
0.44 to 3.30 -0.96 to I
Total
$4.05 to $6.91 $5.77 to $8
* Costs are those above what were estimated to be incurred to
comply with the proposed EPA policy
103
-------
indicated as a small negative number in Table VI.3. This does
not indicate that attainment costs are small, merely that costs
are not much different than those under the EPA policy, which are
already high. The Mitchell bill effectively requires all sources
to attain by 2000 because the bill requires a 65 percent
reduction in 1997, and 5 percent per year every year thereafter.
The maximum VOC emission reduction target is 67 percent.
An overall comparison of the attributes of the Mitchell bill
and the EPA policy shows that the primary cost differences are
for the NOx controls, new more stringent motor vehicle emission
standards, and ozone transport region controls. In the year
2000, these three Mitchell bill provisions account for almost 80
percent of the cost difference between the EPA policy and this
bill. Most of the rest of the cost difference can be attributed
to controls applied in moderate nonattainment areas under S.
1894, which are not shown to be necessary to enable these areas
to attain in the EPA policy analysis, and represent over control.
Thus, much of the additional cost of the Mitchell bill would be
borne by moderate nonattainment areas.
Of the 34 moderate nonattainment areas, in 1995, only one is
estimated to be nonattainment under the proposed EPA policy case
and attainment with the provisions of the Mitchell bill. With
this one exception, the Mitchell bill makes moderate
nonattainment areas control more of their VOC emissions than are
needed to reach attainment of the ozone NAAQS. In 1995, this
overcontrol costs about $950 million.
The year 2000 simulation shows that under the proposed EPA
policy, six moderate nonattainment areas have VOC emissions
increases to the point that they need additional discretionary
controls to continue meeting the standard. (Only two of these
areas have the same problem under the provisions of the Mitchel1
bill.) About 21,000 tons of VOC would need to be reduced in the
six areas for all moderate nonattainment areas to demonstrate
attainment by 2000. If it is assumed that these 21,000 tons can
be reduced at an average cost of $2,000 per ton, the cost to
moderate nonattainment areas of overcontrol1ing VOC is $1
104
-------
billion. Because serious and severe nonatta inment areas need all
available controls, plus new as yet unidentified controls in many
cities, their costs are nearly the same under the Mitchell bill
as they are under the EPA policy.
105
-------
VII WAXMAN BILL ANALYSIS
The Waxman bill (H.R. 3054) offers amendments to the Clean
Air Act that specifically address ozone and carbon monoxide
nonattainment problems. This bill addresses the same areas
covered by Title I and III of the Mitchell bill, i.e., new
attainment deadlines, stationary source control requirements, and
it proposes changes to the current motor vehicle emission
standards.
An outline of the key Waxman bill provisions is shown in
Table VII.1. The Waxman bill provision that differs most from
the EPA policy and the Mitchell bill is the requirement for
catalytic control technology on all greater than 25 ton per year
emitting boilers in severe nonattainment areas by 1991 (natural
gas, methanol, and ethanol fired boilers are exempted). National
measures are the same as those in the Mitchell bill. There are
not as many prescribed measures in Waxman as there are in
Mitchell, but the attainment deadlines are shorter. Assumed new
CTGs for Waxman are the same as those included in the Mitchell
bill simulations listed earlier in Table VI. 2.
A national summary of the incremental cost of the Waxman
bill in 1995 and 2000 is shown in Table VII.2. As expected, the
cost of applying catalytic controls to boilers in severe
nonattainment areas is substantial — about $2 billion. The
incremental cost of new motor vehicle emission standards is $1.2
billion, the same cost estimated for S. 1894. The cost of ozone
transport region controls required by the Waxman bill are $0.53
billion in 1995 and $0.60 billion by 2000. These costs are
slightly lower than those for Mitchell because S. 1894 includes
some Midwestern states in its ozone transport region that are not
covered by Waxman.
Costs of including heavy-duty gasoline-powered vehicles in
enhanced I/M programs are small in 1995 and no different from the
EPA policy in 2000 because enhanced I/M is adopted as a
discretionary control in serious and severe nonattainment areas
under the EPA policy to meet attainment/progress requirements.
106
-------
Table VII.1
Outline of Key Waxman Bill Provisions
Nonattainment area definition is the same (MSA/CMSA) as in
EPA policy
National measures include the following:
A.	New motor vehicle emission standards
B.	RVP restrictions
C.	Onboard control
D.	Commercial and consumer solvent controls
E.	Architectural coatings
F.	Pesticide applications
G.	Traffic coatings
H.	Military specification coating
Nonattainment areas in three categories
Moderate: Attainment deadline is 3 years after enactment
Requirements include the following:
-	Achieve the specified percentage reduction in VOC
and N0X emissions until attainment
Serious: Attainment deadline is 5 years after enactment
Requirements include the following:
-	Moderate nonattainment area requirements plus
-	Enhanced I/M
Severe: Attainment deadline is 10 years after enactment
Requirements include the following:
-	Moderate and serious nonattainment area
requirements plus
-	Alternative fuel capability in 30 percent of newly
registered vehicles by 1997
-	Emission fee for greater than 25 ton per year
emitters
-	Stage II vapor recovery
-	Selective catalytic reduction on all greater
than 25 ton nongas fired boilers
Establishes ozone transport regions (Northeast Corridor)
107
-------
Table
Incremental Cost
National
VII.2
of Waxman Bill
Summary
Annual Cost
(billion $)*
1995	2000
1.	Catalytic Controls on Boilers	$2.00	$2.00
in Severe NA Areas
2.	New Motor Vehicle Standards	1.20	1.20
3.	Ozone Transport Region	0.53	0.60
Controls
4.	Consumer Solvent Controls	0.41	0.41
5.	Stage II in Severe NA Areas	0.09	0.10
6.	Enhanced I/M on HDGV in	0.03
Serious and Severe Nonatta inment
Areas
7.	Wood Furniture Coating	—	0.02
and Bakery CTGs
8.	Cost for Moderate and Serious -0.68 to -1.32
Areas to Attain and Severe
Areas to Reach 60 Percent
of Reduction Toward
Attainment (at $2,000 to
$10,000 per ton)
9.	Cost to Bring All Areas	—	-2.2 3 to 1.3 3
into Attainment
(at $2,000 to $10,000 per ton)
Total	$3.58 to $2.94 $2.10 to $5.66
* Costs are those above what were estimated to be incurred to
comply with the proposed EPA policy
108
-------
The base motor vehicle costs under the Waxman bill (before
residual tons are costed) are lower than those of Mitchell and
the EPA policy because of the way alternative fuel costs are
estimated. The Mitchel1 bill says that by December 31, 1997, not
less than 15 percent, and by December 31, 2002, not less than 40
percent of the total registered motor vehicle fleet shall have
been converted to alternative fuels or power sources in severe
nonattainment areas. The Waxman bill, though, requires that low
emission vehicles constitute at least 30 percent of new motor
vehicles registered by 1997. Therefore, for the year 2000
simulations, it was estimated that 30 percent of the vehicle
fleet would be methanol fueled under the Mitchell bill
provisions, but only 11 percent would be methanol fueled under
the Waxman bill provisions. Thus, the explicit provisions of the
Waxman bill cost less and remove less VOC than the explicit
provisions of the Mitchell bill. EPA policy costs for
alternative fuels are higher than those of the Waxman bill
because the discretionary controls applied to serious and severe
ozone nonattainment areas assume 30 percent methanol fuel
penetration into the vehicle fleet by 2000.
109
-------
VIII GROUP OF NINE PROPOSAL ANALYSIS
Nine Democratic members of the House Committee on Energy and
Commerce developed a proposal that addresses both ozone and CO
nonattainment problems. This approach has come to be known as
the Group of Nine Proposal. This proposal starts by recognizing
that nonattainment, particularly for ozone, is a long-term
problem that will take more than a decade to solve in some areas,
and that emission reductions from many different source types
will be needed to achieve attainment.
Table VIII.1 outlines the key Group of Nine Proposal
provisions from a cost and emissions reduction perspective.
National level consumer solvent controls called for are different
from those in other bills. A 25 percent reduction by 1995 and a
50 percent reduction by 2000 are stipulated. It was assumed in
this analysis that these reductions are from 1985 emission
levels, so with growth included, actual emission reductions in
1995 and 2000 are greater than 25 and 50 percent. Motor vehicle
emission standard changes proposed by the Group of Nine are
somewhat different than those provided for in the Waxman and
Mitchell bills, so they are delineated in Table VIII.2. Note
also that the Group of Nine proposal does not include the heavy-
duty vehicle emission standard changes required by the other
bills.
In the Group of Nine Proposal, nonattainment areas are
categorized according to the degree to which they exceed the air
quality standard. There are four categories for ozone and three
for CO. Group of Nine Proposal ozone nonattainment categories,
design values, and attainment deadlines are shown in Table
VII1.3.
The Group of Nine Proposal calls for new CTGs for 11 source
categories. These new CTGs would be applied in all nonattainment
areas except the Moderate I class. Other more stringent measures
to be applied in moderate, serious, and severe nonattainment
areas are as outlined in Table VIII.l.
National level results of the Group of Nine Proposal
simulations are shown in Table VIII.4. Proposal provisions with
an estimated cost of more than $1 billion include new motor
110
-------
fable VII I . 1
Outline of Group of Win* Proposal Provisions
Nonattainment area definition is the same (MSA/CMSA) as in
EPA poli cv
National level measures include the following:
A, R VP controls
S. Onboard control of VOC
C.	Consumer solvent controls
-25X reduction in VOC by 1995
•	5OX reduction in VOC by 2000
D.	Mobile source emission standard changes
-same as Uaxman and Mitchell Standards except this
proposal does not include HOV Standard changes
Nonattainment area controls
A. Hew CTGs for 7 source categories
-SOCHI disti11 at ion
•Auto body refinishing
¦landfills
•Industrial wastewater
-Clean-up solvents - industrial
-SQCM1 batch process
-Marine vessels - loading and unloading
-Hazardous Waste TSDFs
¦Uood Furniture Coating
•	Baker i es
-Coke Oven By-product Plants
Nonat ta i nment areas in 4 categories
Area 2: Moderate I: Attainment deadline is 3 years after
enactment
Requirements include continuing to apply current regulations
as long as attainment deadline is met, but new CTGs are not
applied in these areas
Area 3: Moderate II: Attainment by 12/31/95
Requirements include the following:
-	RACT applied to all VOC sources
-	CTGs applied to all 50 tpy or larger sources
-	Basic l/M
Area 4: Serious: Attainment by 12/31/97
Requirements include the following:
-	RACT applied to all VOC and NO sources (exemption for
NO allowed)
¦	CTGs applied to all 50 tpy or larger sources
-	Enhanced 1/N
¦	Stage II for large volume gas stations
-	Alternative fuels program {fleet vehicles only) or TCMs
Area 5: Severe: Attainment by 12/31/2005
Requirements include the following:
• same as Area 4 plus
-	each VOC emitter of 25 tpy or more must reduce emissions
by 6X every 3 years or pay S2.000 per ton emitted
111
-------
Table VIII.2
Group of Nine Proposal
Motor Vehicle Emission Standards
Vehicle Type
Start Model
Year
HC
(gm/mile)
NQX
(qm/mile)
Light-Duty Gas
1993*
0.25
0.7
Light-Duty Truck
1993*
0. 50
0.8
* New motor vehicle emission standards are phased in starting
with the .1993 model year. Each manufacturer must have 30 percent
of 1993 model year vehicles, 60 percent of 1994 model year
vehicles, and 90 percent of 1995 or newer model year vehicles
meeting the listed standards.
112
-------
Table VIII.3
Group of Nine Proposal
Atta inment/Nonattainment Categories
Ozone Nonattainment
Categories	Design Value (ppm)	Attainment
Moderate I	.13	1992
Moderate II	0.14, 0.15	1995
Serious	0.16, 0.17, 0.18	1997
Severe	> 0.19	2005
1 1 3
-------
Table VIII.4
Incremental Cost of Group of Nine Proposal
National Summary
Annual Cost
(billion $)*
2	.
3	.
6,
7
8
New Motor Vehicle Standards
Consumer Solvent Controls
Savings from not having TSDF
and POTW Controls in
Moderate I Areas
Lost Savings for not Having
Autobody Refinishing in
Moderate I Nonattainment Areas
Industrial Clean-up Solvent
CTG
Additional I/M Cost
RACT to 50 tpy
Stage II in Serious and
Severe Nonattainment Areas
$2,000/ton Emission Fee for
> 2 5 tpy Sources in Severe
Nonattainment Areas
10.	Alternate Fuel to Fleet
Vehicles in Serious and
Severe Nonattainment Areas
11.	Attainment/Progress
Requirements
Total
1995
$1.04
1.21
-0. 15
0. 07
0. 04
0. 12
0. 03
0. 19
0.15
0.01
•1.13 to -5.04
2000
$1.16
2.09
-0. 15
0.08
0.02
0. 14
0. 03
0.21
0. 15
-0.04
•4.13 to -7.23
$1.59 to $-2.34 $-0.44 to $-3.54
* Costs are those above what were estimated to be incurred to
comply with the proposed EPA policy.
114
-------
vehicle emission standards and consumer solvent controls.
Meeting attainment and progress requirements of the Group of Nine
Proposal is less costly than meeting those of the proposed EPA
policy because the attainment schedule is not as strict and
because the simulation allows a higher level of consumer solvent
controls in the Group of Nine Proposal modeling than it does
under the EPA policy. EPA policy simulations limit consumer
solvent controls to a 20 percent emissions reduction. This
forces MSAs to adopt controls that are more expensive than $2,000
per ton (the cost of consumer solvent reductions) to meet
progress reguirements or attain. Thus, estimates of costs to be
incurred under the Group of Nine Proposal may be biased downward
relative to EPA policy or Mitchell or Waxman bill costs.
The EPA policy case used a 20 percent VOC emission reduction
assumption because it was judged to be realistic and potentially
achievable in the time horizon of the emission projections in
this study.
115
-------
IX SUMMARY OF RESULTS
A. OZONE NONATTAINMENT
Previous chapters have presented results separately for the
proposed EPA policy, the Mitchell bill, the Waxman bill, and the
Group of Nine Proposal. This chapter summarizes the results for
the analyses of all the policies and bills. Because
attainment/progress requirements affect emission reductions and
costs of the policies/bills, those requirements are summarized
first in Table IX.1. Note that while the Mitchell bill does not
require areas with ozone design values above 0.27 ppm to attain
by 2000, the yearly percentage reduction requirements of that
bill effectively force all areas to attain by then.
Figure IX.1 shows how the estimated ozone precursor control
costs differ among the EPA policy and the alternative
Congressional bills and proposals. Both 1995 and 2 000 cost
estimates are shown. Expected additional ozone control cost
expenditures under the pre-1988 EPA policy are delineated in the
figure as part of the total EPA policy cost. While estimates of
the total costs of the EPA policy and the alternative
Congressional bills/proposals are presented, Figure IX.1 is most
useful for showing the relative costs of the different control
approaches. The total costs should be used with caution because
they do not include the historical costs of VOC control such as
Federal Motor Vehicle Control Program costs.
While Figure IX.1 shows the Group of Nine costs to be lower
in 2000 than the expected EPA policy costs, this lower value
depends on high levels of consumer solvent VOC emissions control
in 2000 at $2,000 per ton. The consumer solvent control level is
limited in the other simulations. This issue is discussed more
fully in Chapter VIII.
When costs of the different policies/bills are compared, so
should the number of remaining ozone nonatta inment areas. Table
IX.2 presents ERCAM-VOC estimates of residual nonattainment areas
in 1995 and 2000. Thus, of the three legislative approaches, the
lower costs of the Group of Nine Proposal must be balanced
116
-------
Table IX.1
Attainment/Progress Requirements of Proposals Analyzed
EPA Policy
Waxman
1995
Attain or achieve 3%
per year reductions,
whichever is binding
Moderate and Serious
must attain
2000
Attain or achieve 3% per
year reductions,
whichever is binding
All areas must attain
Severe areas must reduce
emissions by 10% of the
reduction required to
attain the standard each
year
Mitchell Moderate must attain
Serious and Severe must
achieve a 55% reduction
or attain whichever is
less stringent
All except areas with
design values above 0.27
ppm must attain
Group of Moderate I and II must
Nine	attain
Serious must achieve
78% of attainment
target
All except Severe must
attain
Severe must achieve 71%
of attainment target
Severe must achieve
41% of attainment
target
117
-------
Figure IX. 1
Ozone Nonattainment Control Cost Summary
1995 Costs
Zl
0 4 8 12 16 20
Estimated New Expenditures (Billion $)
[7/1 Pre-1988 EPA Policy
2000 Costs
7
A
0 4 8 12 16 20 24 28 32
Estimated New Expenditures (Billion $)
—r
8
i—
12
EPA Policy
—i—
20
Mitchell Bill
Waxman Bill
Group of Nine
Proposal
r
24
Ranges reflect costs of controlling residual tons using a range of $2,000 to $10,000 per ton.
-------
Table IX.2
Residual Ozone Nonattainment Areas by Projection Year*
EPA
Pol icy
Mitchell
Bill
Waxman
Bill
Group of Nine
Proposal
1995
Chicago
Houston
Los Angeles
Milwaukee
New York
San Diego
San Francisco
Chicago
Houston
Los Angeles
New York
San Diego
2000
Los Angeles
New York
Chicago
Houston
Los Angeles
New York
Philadelphia
San Diego
Greater Conn.



/
Massachusetts
^Chicago
—ei-neirmati—
Dallas /
^^^pnrt —
X U U1IU
V	Houston
V	Los Angeles
/Milwaukee
c-Modesto
/New York
/Philadelphia
Phoenix 
-------
against the longer list of expected nonattainment areas. Note
also that the Table IX.2 list of residual nonattainment areas
represents what the policies/bills require and is not an
expectation of when specific areas might attain the ozone
standard.
Table IX.3 shows how VOC control costs are distributed
between new and existing sources for 1995 and 2000 for each of
the four alternatives studied. In all cases, new source costs
are higher than those for existing sources. New source costs
increase in the year 2000 because of growth. Existing source
costs increase between 1995 and 2000 only for the Group of Nine
Proposal. This occurs because ".he Group of Nine Proposal calls
for 50 percent reductions in c isuirter solvent emissions by 2000
while the 1995 emission reduction requirement for this source
category was only 25 percent.
The difference between costs for new and existing sources is
highest for the EPA policy, with new source costs almost three
times higher than existing source costs in 1995 and four times
higher in 2000. For the Congressional alternatives, new source
costs are roughly twice existing source costs in 1995 and two to
three times existing source costs in 2000.
Unless existing source regulations are made more stringent
in the future, new source costs will almost always be higher than
existing source costs because costs of existing source controls
are only estimated for sources which have to install additional
controls to meet regulatory requirements in future years. Thus,
if all existing sources in a category are controlled to 90
percent efficiency, and any regulations expected to affect this
category require no more than 90 percent control, there will be
no control costs estimated for existing sources. Costs are
estimated for all new source emissions affected by a regulation.
While the above may lead to concern that new source costs
are overstated, this is not necessarily so. Because NSPS
Background Information Documents are used to develop cost
equations for many point source categories, and recovery credits
are taken into account in these equations, it is unlikely that
120
-------
Table IX.3
New Versus Existing Stationary Source Costs*
(billion $)
1.995
2000
Existing
Source
Control
Costs
New
Source
Control
Costs
Existing
Source
Control
Costs
New
Source
Control
Costs
EPA Policy	1.48
Mitchell Bill	2.47
Waxman Bill	2.12
Group of Nine	3.03
Proposal
4 . 12
4 . 38
4 .33
4 . 59
1.48
2	. 47
2. 12
3	. 56
5.77
6. 14
6.07
6 . 69
* Includes costs for current policy requirements
121
-------
costs are overstated for these categories. Problems are more
likely to occur for the miscellaneous point source category,
where cost equations have not been developed for specific
combinations of source type and control equipment. For
miscellaneous point sources, a generic cost per ton value is
applied to estimate new source control costs. For industries not
well represented by this generic cost, costs will be in error.
Cost equations were not developed for the many source types
categorized as miscellaneous point sources because in some cases
there are so few plants of that type that it is impossible to
specify a general relationship between controls and costs. Each
individual facility may be of a design different to such a degree
that the control techniques applied differ from plant to plant.
As a general rule, it is important to look closely at
analysis results to see why emission projections and costs
differ. If differences occur largely for categories where
results are very sensitive to analysis assumptions, and not that
much is known about controls and costs for those categories, then
actual differences may not be as great as the analysis shows.
Tables IX.4 and IX.5 show total ozone precursor and CO
control costs for each alternative by CMSA for 1995 and 2000,
respectively. Note that only CO control costs are reported for
CMSAs which are in attainment with the ozone standard but not in
attainment with the CO standard. Many areas have cost ranges
reported. This is due to the costing of residual tons necessary
to meet ozone attainment/progress requirements. A range of
$2,000 to $10,000 per ton reduced was used to estimate these
costs.
Tables IX.6 and IX.7 contain the same information given by
state. Attainment/progress requirement costs for CMSA's crossing
state boundaries were apportioned among the states according to
county population (U.S. Bureau of the Census, 1985). The state
level costs include costs for both ozone precursors and CO for
all areas within the state, both attainment and nonattainment.
122
-------
Table IX.4
Ozone and CO Nonattainment Control Cost Summary by CMSA
CMSA
Albuquerque, NM*
Allentovn-Bethlehem, PA-NJ
Anchorage, AK*
Atlanta, GA
Atlantic City, NJ
Bakers field, CA
Baltimore, MD
Baton Rouge, LA
Beaumont, TX
Birmingham, AL
Boise City, ID*
Bradenton, FL
Charleston, W
Chariot te-Gastonia, NC-SC
Chattanooga, TN-GA
Chicago CMSA
Chico, CA*
Cincinnati CMSA
Cleveland CMSA
Colorado Springs, CO*
Dallas CMSA
Davenport-Rock Island, IA-IL*
Dayton-Springfield, OH
Denver CMSA
Des Moines, IA*
Detroit CMSA
Dubuque, IA*
El Paso, TX
Erie, PA
Fort Collins, CO*
EPA Policy
9.6
15.8
1.1
151.5
8.2-19.3
25.6
89.6
97.1
111.3
22.2
1.0
2.1
35.5
38.8
16.1
403.4-712.Q
0.0
70.9
81.9
0.0
260.4-554.2
0.0
24.6
31.4
0.0
212.8
0.0
16.9
4.0
0.9
	1995 Cost (millionS)	
Mitchell Bill	Waxman Bill
10.0
35.7
1.3
187.1
14.1-27.3
45.8
129.5
109.5
139.1
58.6
1.2
8.2
50.6
69.3
33.4
629.1-1,265.1
0.1
102.5-102.6
155.6
1.9
335.7-677.8
0.0
60.3
63.8
0.0
323.8
0.0
25.5-28.2
15.2
1.1
9.6
27.8
1.1
172.7
12.5-27.0
29.0
119.5
101.3
114.3
29.2
1.0
3.4
37.3
57.6
19.6
982.0
0.0
85.8-86.3
106.9
1.8
313.7-693.6
0.0
33.0
46.9
0.0
251.1
0.0
23.5-33.2
13.9
0.9
Group of Nine
Proposal
10.0
22.5
1.3
198.8
11.2-11.7
33.4
132.2
107.7
122.3
27.3
1.1
4.1-9.7
37.3
51.1-51.7
23.0-34.7
527.1
0.0
103.2
133.8
0.0
290.0-396.5
0.0
34.8
45.8
0.0
250.2
0.0
25.4
6.9
1.0
-------
Table IX.4
Ozone and CO Nonat tainment Control Cost Summary by CMSA
		1995 Cost (millionS)			
Group of Nine
CMSA	EPA Policy	Mitchell Bill	Vaxman Bill	Proposal
Fresno, CA
•
45.2-93.7
54.2-106.0
51.4-107.5
46.9-67.6
Gadsden, AL

3.1
7.5
3.8
3.5
Grand Rapids, MI

28.5
54.4
35.0
42.0-58.6
Greater Connecticut
CMSA
47.8-88.4
90.3-143.4
165.5
80.7
Greeley, CO*

2.8
3.0
2.9
2.9
Greensborough, NC*

0.1
0.4
0.1
0.1
Barrisburg-Lebanon,
PA
4.3
25.3
23.9
8.4
Houston CMSA

763.8-1,532.4
902.5-1,666.0
855.8-1,285.3
693.5
Huntington-Ashland,
W-KY-OH
36.8
51.5
39.2
46.4
Indianapolis, IN

91.1
137.3
101.0
100.4
Jacksonville, FL

10.2
43.4
16.4
32.6
Janesville-Beloi t,
VI
32.7
38.4
33.9
34.2
Kansas City, MO-KS

88.3
146.6
132.5
157.1
Lake Charles, LA

16.2
21.2
17.5
18.3
Las Vegas, NV*

2.8
3.9
3.5
3.2
Lexington, K.Y*

0.0
0.1
0.0
0.0
Longview, TX

25.4
32.5
26.8
27.0
Los Angeles CMSA

471.6-1,290.0
907.6-2,687.2
909.0-1,922.5
508.5
Louisvilie, KY-IN

43.8
62.9
52.2
51.3
Manchester, NH*

1.5
0.3
0.0
1.8
Massachuset ts

378.6-803.7
504.1-980.8
513.2-1,042.8
456.4-663.7
Medford, OR*

0.8
1.1
0.8
0.9
Memphis, TN

23.1-23.3
37.7-38.4
31.5-32.5
34.8-53.2
Miami CMSA

35.2
139.6
57.7
115.8
Milwaukee CMSA

62.6-146.3
98.3-211.1
89.7-221.8
79.1-98.9
Minneapolis-St. Paul, MN-WI
110.5
153.9
137.6
183.1
Modesto, CA

13.1-22.2
17.9-29.4
16.7-30.3
16.3
Muskegon, MI

9.0
17.3
10.4
13.7
Nashville, TN

41.4
69.4
49.0
59.9
New York CMSA

368.8-398.6
674.4-748.4
1,170.6-1,200.6
702.3
-------
Table IX.4
Ozone and CO Nonattainment Control Cost Summary by CMSA


---	1995 Cost
(millionS)	
Group of Ni
CMSA
EPA Policy
Mitchell Bill
Uaxman Bill
Proposal
Norfolk, VA
34.1
79.5
76.2
69.1
Oklahoma City, OK*
0.0
5.3
5.1
0.0
Peoria, IL*
0.0
0.4
7.9
6.0
Philadelphia CMSA
256.6-263.5
354.7-364.1
582.8
377.7
Phoenix, AZ
69.6-198.9
103.0-255.2
94.1-259.2
89.0-I52J
Pittsburgh CMSA
34.0
86.6
78.8
45.8
Portland CMSA
18.7
49.0
36.1
28.6
Providence, RI
24.7-34.6
40.7-55.7
39.6-61.7
36.9-57.6
Provo, UT*
1.2
1.8
1.9
1.4
Raleigh-Durham, NC*
3.2
3.8
3.2
3.6
Reading, PA
7.9
20.3
19.3
11.3
Reno, NV*
1.1
1.3
1.1
1.3
Richmond-Petersburg, VA
83.4
115.3
110.5
88.5
Rockford, IL*
0.0
0.0
0.0
0.0
Sacramento, CA
55.5-153.6
75.7-179.9
70.1-187.9
59.0-85.8
Salem, OR*
0.0
0.1
0.0
0.0
Salt Lake City, UT
22.0
38.0
30.8
29.8-30.0
San Diego, CA
98.5-290.6
148.9-421.2
103.7-188.7
89.6
San Francisco CMSA
262.2-664.8
423.4-1,122.9
386.5-510.7
261.6
Santa Barbara, CA
18.7-39.9
25.4-52.0
23.7-53.7
21.1-26.8
Seattle, UA*
0.2
23.9
20.1
8.5
Sheboygan, WI
5.8-15.4
8.7-20.1
7.9-21.9
7.1-12.0
Spokane, UA*
0.0
33.6
1.9
0.0
Steubenville, OH-VV*
0.0
25.0
3.4
2.6
Stockton, CA
12.4
18.3
15.6
14.9
St. Cloud, MN*
0.0
4.0
3.7
2.8
St. Louis, MO-IL
136.5
181.9
157.6
182.5
Tampa-St. Petersburg, FL
19.6
91.3
34.3
30.8
Toledo, OH*
0.0
0.1
0.0
0.0
Tucson, AZ*
0.0
3.1
3.0
0.0
-------
-------
Table IX.5
Ozone and CO Nonattainment Control Cose Summary by CMSA
CMS A
Albuquerque, NM*
Allen town-Bethlehem, PA-NJ
Anchorage, AK*
Atlanta, GA
Atlantic City, NJ
Bakers field, CA
Baltimore, MD
Baton Rouge, LA
Beaumont, TX
Birmingham, AL
Boise City, ID*
Bradenton, FL
Char 1 eston, UV
Char lotte-Gastonia, NC-SC
Chat tanooga, TN-GA
Chicago CMSA
Chico, CA*
Cincinnati CMSA
Cleveland CMSA
Colorado Springs, CO*
Dallas CMSA
Davenport-Rock Island, IA-IL*
Dayton-Springfield, OH
Denver CMSA
Des Moines, IA*
Detroit CMSA
Dubuque, IA*
El Paso, TX
Erie, PA
Fort Collins, CO*
EPA Policy
9.6
18.8
1.1
178.3
11.4-30.0
29.5
111.2
126.9
140.9
25.5
1.0
3.6-8.0
42.1
51.6-75.1
21.7-32.5
561.4-1,347.0
0.0
79.7-80.1
92.9
0.0
330.7-750.6
0.0
26.7
38.4
0.0
243.1
0.0
21.1-28.2
4.4
0.9
	2000 Cost (million!)	
Mi tchell Bill	Uaxman Bill
Group of Nine
Proposal
10.0
39.7
1.3
221.0-238.2
18.4-41.5
51.3-55.8
152.8
139.8
169.0
65.0
1.2
9.2
58.1
77.8
39.1-43.8
1,602.5-2,456.6
0.1
110.9-111.9
165.4
1.9
418.8-929.2
0.0
65.6
74.4
0.0
347.4
0.0
31.8-48.3
16.5
1.1
9.6
28.6
1.1
209.8-245.7
15.3-39.9
36.7-51.0
135.4-146.6
132.5-138.0
144.0
32.8
1.0
4.2-5.3
44.0
65.0
24.9-34.2
1,360.5-2,296.7
0.0
95.3-96.7
118.6
1.8
396.5-945.3
0.0
35.4
54.8
0.0
282.3
0.0
29.5-53.1
13.9
0.9
10.0
.5
3
.7
7-38,
0
28.
1
240.
18.
40.1-43.6
164.3
139.6
153.4
33.6
1.1
7.6-23.4
44.5
80.8-149.4
40.1-101.6
643.6-782.7
0.0
125.2-143.8
159.8
0.0
434.3-869.9
0.0
41.6
61.3
0.0
299.1
0.0
33.3-45.7
8.6
1.0
-------
Table IX.5
Ozone arid CO Nonat tainmerst Control Cost Summary by CMS A
CMSA
EPA Policy
•	2000 Cost (million*)	
Mitchell Bill	Vaxraan Bill
Group of Nine
Proposal
Fresno, CA
Gadsden, AL
Grand Rapids, MI
Greater Connecticut CMSA
Greeley, CO*
Greensborough, NC*
Harrisburg-Lebanon, PA
Houston CMSA
Huntington-Ashland, UV-KY-GH
Indianapolis, IN
Jacksonville, FL
Janesville-Beloit, WI
Kansas City, MO-KS
Lake Charles, LA
Las Vegas, NV*
Lexington, KY*
Longview, TX
Los Angeles CMSA
Louisville, KY -IN
Manchester, NH*
Massachuset ts
Med ford, OR*
Memphis, TN
Miami CMSA
Milwaukee CMSA
Minneapolis-St. Paul, MN-VI
Modesto, CA
Muskegon, MI
Nashville, TN
New York CMSA
60.9
3.3
38.8
68.2
2.8
0.1
4.9
i,095.1
45.8
105.1
16.7
36.0
96.4
19.4
2.8
0.0
32.4
824.7
50.0
1.5
513.5
0.8
26.8
41.9
82.3
124.1
16.5
9.6
48.8
467.2
-125.1
-50.6
-155.5
-2,403.2
-36.3
-2,806.7
-1,179.7
-28.5
-213.2
-33.8
-555.0
72.0
8.1
64.1
273.4
3.0
0.4
27.9
2,117.5
60.7
155.4
47.8
42.2
157.8
24.7
3.9
0.1
39.7
1,708.1
69.2
0.3
660.3
1.1
42.5
156.5
123.5
169.5
22.5
18.4
78.4
1,931.9
-146.0
371.7
¦3,475.1
-3,374.0
-1,431.5
-45.2
¦275.4
•45.2
2,057.4
69
4
42
224
2
0
23
1,373
48
115
19
37
141
20
3
0
34
1,677
58
0
641
0
35
65
110
152
21
10
56
1,440
0-147.3
1
8
2-347.8
9
1
8
0-2,892.9
2
5
7-25.6
3
0
7
5
0
2-35.3
3-4,871.8
8-59.1
0
6-1,466.8
.8
.6-38.6
6
.5-282.8
.2
0-45.9
.9
.7
.1-1,574.3
72.
4.
58.
102.
2.
0.
11.
1,108.
57.
119.
39.
38.
175.
22.
3.
0.
34.
1,040.
61.
1.
681.
0.
48.
141.
122.
208.
23.
15.
72.
1,161.
9-141.2
I
3-91.4
7
9
1
2
9-1,909.9
0
4
2
2
3
0
2
0
6
3-2,620.3
2
8
2-1,31 1.2
9
2-91.9
4
2-246.9
6
8-41.5
3
1
5-2,305,3
-------
Table IX.5
Ozone and CO Nonat tainment Control Cost Summary by CMSA
CMS A
Norfolk, VA
Oklahoma City, OK*
Peoria, II.*
Philadelphia CMSA
Phoenix, AZ
Pittsburgh CMSA
Portland CMSA
Providence, RI
Provo, UT*
Raleigh-Durham, NC*
Reading, PA
Reno, NV*
Richmond-Petersburg, VA
Rockford, IL*
Sacramento, CA
Salem, OR*
Salt Lake City, UT
San Diego, CA
San Francisco CMSA
Santa Barbara, CA
Seattle, UA*
Sheboygan, WI
Spokane, UA*
Steubenville, QH-WV*
Stockton, CA
St. Cloud, MN*
St. Louis, MO-IL
Tarapa - St. Petersburg, PL
Toledo, OH*
Tucson, AZ*
EPA Policy
38.6
0.0
0.0
318.9-332.3
96.5-301.2
40.7
23.1
34.6-64.9
1.2
3.2
8.6
1.1
115.7
0.0
67.4-189.4
0.0
27.1
147.4-492.3
385.2-1,173.8
24.1-57.0
0.2
7.4-21.9
0.0
0.0
13.9
0.0
152.8
23.9
0.0
0.0
	2000 Cost (millionS)	
Mi tchell Bill	Vaxman Bill
88.3
5.3
0.4
879.7-895.4
138.4-388.9
93.5
55.8
53.3-96.0
1.8
3.8
22.1
1.3
150.3
0.0
92.8-237.2
0.1
49.5-70.8
262.2-618.4
875.9-1,708.8
32.1-74.0
23.9
10.2-25.4
33.6
25.0
20.1
4.0
198.3
102.3
0.1
3.1
79.2
5.1
7,9
700.5-721.3
128.1-391.7
78.2
41.1
48.3-98.3
1.9
3.2
19.6
1.1
141.8
0.0
86.4-244.5
0.0
44.6-77.3
217.2-603.4
684.5-1,598.3
30.2-75.5
20.1
9.5-
1.9
3.4
17.2
3.7
175.4-
39.3
0.0
3.0
27.3
178.2
Group of Nine
Proposal
81. 1
0.0
6.0
465.6
147.7-362.6
59.2
45.3-71.0
49.6-82.9
1.4
3.6
13.7
1.3
123.7
0.0
96.7-229.9
0.0
46.3-73.7
156.7-336.4
404.6-742.0
33.2-70.1
8.5
10.5-23.6
0.0
2.6
17.7
2.8
210.9
48.7-72.6
0.0
0.0
-------
Table IX.5
Ozone and CO Nonat tainment Control Cost Summary by CMSA
			2000 Cost (millionS)						
Group of Nine
CMSA	EPA Policy	Mitchell Bill	Vaxman Bill	Proposal
Tulsa, OK
33.0
44.4
39.4
41.2
Visalia-Tulare-Partervilie, CA
3.1
13.3
5.2
5.4
Washington, DC
76.1
123.9-133.0
106.8-119.3
147.2-163.3
West Palm Beach, PL
21.9-80.0
45.2-73.0
24.3-68.7
37.8-62.1
Yakima, WA*
3.9
4.2
3.9
4.0
York, PA
3.6
22.1
10.7
10.4
Yuba City, CA
1.2
5.3
2.0
1.9-2.1
Note: Costs include both ozone precursor (VOC and NOx) and CO control costs unless otherwise noted.
Control of residual tons necessary to meet attainment/progress requirements at $2,000
to $10,000 per ton produces a cost range for some areas.
* Indi cates CO nonat tainment area which is in attainment of the ozone standard. Costs reported for
these areas include only the CO control costs.
-------
Table IX.6
Ozone and CO Nonattainment Control Cost Summary by State
				1995 Cost (million$)	-
State
EPA Policy
Mitchell Bill
Waxman Bill
Proposal
Alabama
176.6
241.1
207.3
212.8
Alaska
5.5
9.3
8.8
9.6
Arizona
74.6-203.9
120.8-273.0
111.6-276.7
106.3-169.4
Arkansas
27.2-27.3
48.0-48.7
46.6-47.6
53.9-54.9
Callfornia
1,013.3-2,603.7
1,754.2-4,699.7
1,631.8-3,071.7
1,207.3-1,773
Colorado
41.6
86.7
69.0
68.9
Connecticut
70.9-141.4
143.7-270.8
247.8-277.8
118.9
Delaware
52.4-59.2
68.7-78.1
107.8
62.1
Washington, D.C
11.6
15.3
19.2
20.2
Florida
91.0-112.3
364.3
173.1-183.9
290.1-387.8
Georgia
170.0
236.7
218.7
254.8-257.5
Hawaii
1.9
8.9
8.9
10.5
Idaho
3.7
11.4
10.8
12.9
Illinois
523.3-831.9
856.6-1,492.6
899.9
736.2-976.3
Indiana
295.8-323.1
546.4-602.8
506.9-507.4
368.9-390.1
Iowa
9.3
31.2
31.1
37.7
Kansas
43.6
79.8
74.3
87.9
Kentucky
101.1
145.8-146.5
130.1-133.5
138.5
Louisiana
209.4
253.1
240.8
254.2
Maine
5.9
42.7
46.8
17.8
Maryland
123.3-124.3
207.6-209.1
190.1
196.5
Massachuset ts
378.9-803.9
506.2-982.8
514.3-1,043.9
458.6-673.6
Michigan
268.0
527.5
344.3
364.8-380.0
Minnesota
140.2
202.5
185.0
232.7
Mississippi
24.4-24.5
45.3-46.1
44.5-45.6
50.8-51.9
Missouri
198.2
288.9
259.5
301.0
Montana
3.3
9.3
8.8
10.0
Nebraska
4.2
16.0
15.9
19.3
-------
Table IX.6
Ozone and CO Nonattainment Control Cost Summary by State
		---1995 Cost (million$)				
Group of Nine
State	EPA Policy	Mitchell Bill	Vaxman Bill	Proposal
Nevada
8.8
17.0
16.3
18.1
New Hampshire
21.3
47.3
50.5
34.6
New Jersey
341.7-579.6
555.8-1,101.9
610.2-826.0
452.0-452.7
New Mexico
14.8
25.2
24.3
26.6
New York
380.7-786.2
787.7-1,793.4
1,170.8-1,579.1
537.9
North Carolina
149.5
221.3
208.0
219.7
North Dakota
1.5
6.0
6.0
6.8
Ohio
217.9
407.7-411.0
316.6-334.0
377.2-388.3
Oklahoma
56.7
90.9
85.9
84.6
Oregon
76.2
111.8
103.2
99.2
Pennsylvania
220.1-283.0
534.0-620.8
630.6
355.2
Rhode Island
25.6-35.6
43.1-58.1
42.5-64.6
38.8-59.9
South Carolina
38.1
69.0
67.1
76.0
South Dakota
2.7
7.6
7.5
8.6
Tennessee
219.7-221.9
299.4-311.2
262.3-279.1
283.0-308.0
Texas
2,223.1-3,285.6
2,538.3-3,646.6
2,436.5-3,255.7
2,277.1-2,385
Utah
32.2
53.2
46.0
45.5
Vermont
1.1
18.5
20.3
7.2
Virginia
166.1
318.9
365.1
251.4
Washington
87.0
184.2
141.3
137.3
West Virginia
122.3
178.9
137.2
142.8
Wisconsin
123.0-221.5
277.1-411.9
241.8-387.9
177.1-206.6
Wyoming
2.5
5.9
5.8
6.2
Note: Costs include both ozone precursor (VOC and NOx) and CO control costs.
Control of residual tons necessary to meet attainment/progress requirements at $2,000
to $10,000 per ton produces a cost range for some areas.
-------
Table IX.7
Ozone and CO Nonattainment Control Cost Summary by State
State
EPA Policy
	-2000 Cost (million})	
Hi tchell Bill	Vaxman Bill
Group of Nine
Proposal
Alabama
244.3

313
1
276.1
296
5
Alaska
6.0

9
9
9.4
11
6
Ar izona
103.0-307.7

158
2-408.7
147.6-411.2
133
9-197.1
Arkansas
37.9-39.6

59
7-62.4
58.2-61.2
76
2-77.2
Callfornia
1,582.1-4,934.
1
3,214
1-6,356.5
2,867.9-7,683.3
1,475
1-2,040.8
Colorado
50.6

99
6
79.4
91
7
Connecticut
109.9-285.1

418
4-642.3
343.8-601.6
153
1
Delaware
61.2-74.6

136
1-151.8
123.7-144.5
72
3
Washington, D.C
14.0

20
3-29.3
22.3-34.9
24
9
Florida
122.9-204.9

413
2-441.0
204.2-255.6
364
0-461.7
Georgia
204.4-215.2

278
1-300.0
263.9-309.0
317
4-320.0
Havai i
2.0

9
3
9.3
13
9
Idaho
4.2

12
2
11.6
17
1
Illinois
702.0-1,487.
5
1,620
9-2,474.9
1,290.6-2,229.6
851
2-1,091.3
Indiana
380.2-450.1

837
5-913.9
611.6-696.0
471
1-492.3
Iowa
11.1

33
6
33.4
50
3
Kansas
47.8

86
4
79.1
103
4
Kentucky
123.5-126.6

170
9-177.7
154.6-165.6
173
8
Louisiana
262.2

307
4
295.8-301.3
321
5
Maine
7.5

47
9
47.2
24
4
Maryland
151.4-153.5

248
1-272.9
215.1-260.7
242
7
Massachuset ts
513.8-1,180.
0
662
3-1,433.6
642.7-1,467.9
577
7-792.7
Michigan
313.1-324.8

589
3
388.6
448
3-463.5
Minnesota
167.5

232
3
213.9
279
1
Mississippi
33.1-35.0

54
8-57.9
54.0-57.4
69
9-71.0
Missouri
237.4

333
4
302.4-311.1
365
8
Montana
4.1

10
2
9.8
13
2
Nebraska
4.8

17
1
17.0
25
9
-------
Table IX.7
Ozone and CO Nonat tainment Control Cost Summary by State
			2000 Cost (milllon$) —	-
Group of Nine
State	EPA Policy	Mitchell Bill	Waxman Bill	Proposal
Nevada
10.7
19.2
18.5
24.0
New Hampshire
26.5
54.8
54.1
46.5
New Jersey
524.4-1,184.1
1,188.3-2,114.6
948.8-1,954.7
572.4-573.1
New Mexico
16.3
27.0
26.1
32.4
New York
644.4-1,838.2
1,835.2-3,542.5
1,685.3-3,510.7
680.6
North Carolina
204.2-227.7
273.5
259.0
300.5
North Dakota
1.7
6.3
6.3
8.8
Ohio
253.8-269.9
458.6-493.4
359.4-408.6
463.7-474.8
Oklahoma
68.2
104.2
98.3
107.7
Oregon
107.8
145.9
135.8
141.9
Pennsylvania
276.5-399.6
897.5-1,041.6
736.7-927.5
445.2
Rhode Island
35.9-66.3
56.2-99.0
51.3-101.2
49.1-70.2
South Carolina
49.1-52.0
80.7
78.5
102.5
South Dakota
3.5
8.5
8.4
11.5
Tennessee
285.1-346.5
366.2-425.3
329.8-407.6
356.1-381.0
Texas
3,1 17.5-4,852.7
4,366.9-6,251.3
3,533.6-5,627.0
3,078.8-3,186
Utah
39.8
67.5-88.8
62.5-95.2
60.0
Vermont
1.2
20.4
20.1
10.0
Virginia
217.1
383.6-401.8
412.8-438.1
327.8
Washington
107.0
206.1
162.5
174.9
West Virginia
151.8
210.1
167.0
178.0
Wisconsin
153.4-311.8
390.9-572.2
277.7-483.3
217.3-246.7
Wyoming
3.4
6.9
6.7
8.4
Note: Costs include both ozone precursor (VOC and NOx) and CO control costs.
Control of residual tons necessary to meet attainment/progress requirements at $2,000
to $10,000 per ton produces a cost range for some areas.
-------
B. CARBON MONOXIDE NONATTAINMENT
Costs of measures to help MSAs (and non-MSAs) attain the CO
ambient standard that are presented in this chapter are those in
addition to what is estimated to be spent to comply with the
ozone related provisions of the policy or bill. This effort to
avoid double counting control costs affects I/M costs. Thus, the
bill with the most stringent I/M requirements for CO may not have
the highest costs, because similarly stringent ozone requirements
have probably already accounted for most of the cost increase.
Table IX.8 summarizes estimated CO costs by control measure
for the EPA policy and the three legislative approaches. CO
costs of the EPA policy are much lower than the costs of the
three legislative approaches. The only CO control measure
modeled as if it were mandated by the EPA policy is enhanced I/M.
While the proposed EPA policy mentions 17 ppm as a possible
cutoff for requiring enhanced I/M, a lower cutoff was used in
this analysis because preliminary simulations showed that many
areas with design values below 17 ppm would not be able to
demonstrate short-term attainment without new measures. Thus,
enhanced I/M is modeled as if it would be the preferred
"discretionary control measure" adopted by urban areas to attain
the standard under the EPA policy.
Total CO costs for the Mitchell bill, the Waxman bill, and
the Group of Nine Proposal are similar in magnitude. The cost
burden is distributed differently for each legislative approach,
however. The Mitchell bill places more of the cost burden on
stationary sources. Group of Nine Proposal costs affect only
motor vehicles.
All of the policies/bills have additional I/M costs. These
costs can include improving the effectiveness of existing I/M
programs and establishing new I/M programs in areas where they
currently do not exist. Both the Mitchel1 bill and the Group of
Nine Proposal have alternative fuel programs in severe CO
nonattainment areas. These programs are estimated to cost $27
million. The alternative fuels case modeled is a CO season
135
-------
Table IX.8
Additional Carbon Monoxide Control Costs*
1995 Projection Year
(millions)
Policies/Bills
Control
Measures
EPA
Policy
Mitchell
Bill
Waxman
Bill
Group of Nine
Proposal
Motor Vehicle Measures
Enhanced I/M
Alternative Fuels**
Stationary Source
Controls
$38
$67
27
40
$128
$132
27
0
Emission Fee
Totals
$38
34
$168
13
$141
$159
* Costs are those in addition to what is estimated to be spent to
comply with ozone provisions.
** The alternative fuels case modeled is a CO season (winter)
switch from straight gasoline to an ethanol blend.
Note: Effects of cold start certification testing for motor
vehicles have not been included in this analysis.
136
-------
(winter) switch from straight gasoline to an ethanol blend. This
program is similar to the one currently being used in the Front
Range of Colorado.
The CO stationary source controls called for by the Mitchell
bill are estimated to cost $40 million. These are the costs of
applying the control techniques listed in Table III.1 to serious
and severe CO nonattainment areas.
Stationary source emission fees of $100 per ton are applied
in both the Mitchell and Waxman bills. Costs are higher for the
Mitchel1 bill because the fee is applied in both serious and
severe nonattainment areas. The Waxman bill only has an emission
fee for sources in severe nonattainment areas.
Estimates of expected attainment dates depend on which
source types are assumed to be contributing to observed CO
standard exceedances. With the assumption that mobile sources
and a percentage of stationary area sources (20 percent) affect
the design value monitor, there are three residual CO
nonattainment areas in 1995 in the simulations for the proposed
EPA policy and the Waxman bill. The Mitchell bill and Group of
Nine Proposal simulations showed one remaining CO nonattainment
area in 1995. If all sources within an MSA are assumed to
contribute equally to CO standard exceedances, many more areas
are projected to fail to attain the standard by 1995.
Note also that MOBILE3 CO I/M credits are higher than what
has been observed in recent surveys (Sierra Research, 1988). If
I/M programs are less successful than indicated by M0BILE3, the
number of remaining CO nonattainment areas in 1995 will increase.
The weighting procedure employed in this study to estimate
whether areas are expected to attain the CO NAAQS by 1995 is one
that has historically been used by the EPA (U.S. EPA, 1980;
1985). As it says in the "Cost and Economic Assessment of
Alternative NAAQSs for Carbon Monoxide":
Because of the different nature of mobile source and
stationary source emission problems and the 1 ocation of the
existing monitoring network, it is believed that recorded
violat ions 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
137
-------
which indicates that stationary source emissions had
negligible effects on CO monitor readings in most counties.
138
-------
X SENSITIVITY ANALYSIS
The emission projections and cost results for future years
are dependent on the growth rates used in the analysis. As an
alternative to the baseline growth used (MSA-level BEA growth
rates and national average VMT growth from the motor fuel
consumption model), projections were made using a set of SIC
national average growth rates (U.S. EPA, 1980). The MOBILE3 Fuel
Consumption Model was used as an alternate source of national VMT
growth projections. This alternative case is referred to as the
national growth case while the baseline growth is referred to as
the MSA growth case (although national VMT growth rates are used
in both cases). Table X.l provides the national average growth
rates by ERCAM industrial category derived from the SIC annual
growth rates. National average VMT projections used for the
national growth case are shown in Table X.2. VMT by vehicle type
used in the MSA growth case was shown earlier in Table II.7.
Average annual VMT growth for all vehicle types between 1985
and 1995 is 3.1 percent per year for the MSA growth case. In
contrast, the national growth case shows an average VMT growth of
1.9 percent per year for the same period. (When compared with
historic evidence and alternative forecasts, EPA's MOBILE 3 Fuel
Consumption Model projections are on the order of 10 to 30
percent lower than forecasts prepared by other organizations.)
On the stationary source side, average annual growth from 1985 to
1995 for chemical manufacturing pods is 2.5 percent per year for
the MSA growth case compared with 3.1 percent per year for the
national growth case. The largest difference in annual growth is
for sources classified as "other" under ERCAM's industrial
classifications. The national growth case uses population based
growth of 0.8 percent per year. The MSA growth case uses total
earnings as the basis for growth projections in the "other"
classification. The average growth varies by pod since the
growth rates are MSA dependent. Consumer solvents, classified as
"other," show an average growth between 1985 and 1995 of 3.1
139
-------
Table X. 1
National Average Growth Rates by Industrial Category
Used in Sensitivity Analysis
ERCAM
Industrial Category	Average Annual Growth (1977-2000)
Food and Agriculture	0.6
Mining Operations	1.3
Wood Products	2 . 3
Printing and Publishing	2.3
Chemicals	3.1
Petroleum Refining	1.9
Mineral Products	1.3
Metals	2.4
Machinery & Equipment Mfg.	2.6
Crude Oil Production,	2.2
Storage, and Transfer
Electric Utilities	3.5
Other Fuel Combustion	3.5
Petroleum Product Production,	1.9
Storage, and Transfer
Other Transportation	2.9
Dry Cleaning	0.8
Other	0.8
Source: U.S. EPA, 1980
140
-------
Table X.2
Annual VMT by Vehicle Class and Year
VMT (billions)
LDGV
LDGT
HDGV
HDDV
Totals
1985
1,075.5
357 . 2
55. 3
104 . 0
1,592.0
1995
1,298.8
438 . 6
55.9
137 . 8
1,931.1
2000
1, 410.0
479.3
58.7
154 . 3
2,102.3
2010
1,632.4
560. 2
68.2
182 . 8
2,443.6
Equivalent Annual Growth Hates
LDGV
LDGT
HDGV
HDDV
Average
1985-1995
1.9%
2 . 2
0.1
2.8
2 . 0%
1995-2000
1.7%
1. 8
1	. 0
2	. 3
1.7%
2000-2010
1.4%
1. 5
1.5
1.7
1. 5%
Source: U.S. EPA, 1984
,14 1
-------
percent per year while TSDFs show an average of 3.5 percent per
year.
Nonattainment area emission projection results for the two
alternative growth cases are compared in Table X.3. The national
growth case projects lower emissions in all analyses. The 1995
difference (MSA growth-national growth) ranges from 767 thousand
tons for the Mitchell Bill to 837 thousand tons for the EPA
Policy. Approximately 70 percent of the difference can be
accounted for by four or five categories as shown in Figure X.l.
The categories accounting for this difference are the same for
both the EPA Policy and the Mitchell Bill with the exception of
TSDFs. TSDFs are not as important an emissions difference in the
EPA policy analysis because this source is well controlled in all
areas. The Mitchell Bill does not mandate TSDF controls in
attainment areas.
The national total cost differences by alternative and year
are shown in Figure X.2. The MSA growth case total costs are
higher than the national growth case costs in all cases. The
cost difference in 1995 ranges from $711 million for the EPA
policy to $1,058 million for the Group of Nine Proposal. For the
2000 results, the difference ranges from $304 million for the EPA
policy to $961 million for the Group of Nine Proposal. The cost
difference decreases from 1995 to 2000 for two reasons. Many new
source costs are negative, denoting a cost savings (savings on
solvent usage) for the control. Also, many of the organic
chemical manufacturing industry sources show higher growth in the
national growth case than in the MSA growth case.
With grawth rates for key manufacturing industry categories,
such as the chemical industry, not being appreciably different
between the two alternatives used in this sensitivity analysis,
the choice between MSA-level growth rates versus national
averages will only lead to signi f icant differences in results if
nonattainment area growth rates (especially those for serious and
severe nonattainment areas) are much higher than those elsewhere
in the country. Analysis results for chemical industry sources
indicate that this is not the case. Serious nonattainment area
142
-------
Table X.3
Nonattainment Area VOC Emissions by Alternative Growth
(thousand tons)
	1995			2000	
MSA	National	MSA	National
1985 Growth	Growth	Growth	Growth
EPA Policy 8,626 6,173	5,336	6,774	5,748
Mitchell Bill 5,685	4,918	6,147	5,230
Waxman Bill 5,892	5,112	6,396	5,454
Group of Nine
Proposal 5,921	5,120	6,252	5,328
* Projected VOC emissions before discretionary controls and
attainment/progress requirements
143
-------
Figure X.l
VOC Emission Differences for Alternative Growth
1995 EPA Policy
(thousand tons)
Degreasing
(174)
Motor vehicles
(399)

All other
(546)
Consumer solvents
(325)
Miscellaneous
surface coating
area source
(295)
1995 Mitchell Bill
(thousand tons)
TSDF
(203)
All other
(508)
Motor vehicles
(360)
Consumer
solvent
(260)
Degreasing
(169)
Miscellaneous
surface coating
area source
(295)
Notes: Emission difference = base case - national growth.
Emission difference is for all areas, attainment and nonattainment.
All other includes sources with less than 5% absolute of total emission difference
and includes some negative values.
i.	i
J
-------
Figure X.2
National Cost Differences for Growth Analysis
(MSA Growth Cost - National Growth Cost)
U1
m
a
o
0)
o
0
<0
u
Q>
CO
o
o
2000
7


/

EPA Policy
Mitchell Bill
Waxman Bill
Group of Nine
Proposal
-------
VOC emissions as a whole in 1995 are estimated to be higher in
1995 using national average growth rates than they are estimated
to be using the MSA specific rates. The reverse is true for
severe nonattainment areas.
146
-------
XI CAVEATS
Any analysis that attempts to estimate how future laws or
regulations will affect the behavior of individuals, firms, and
state and local regulatory agencies must incorporate simplifying
assumptions. In addition, data bases are employed which may not
be perfectly designed for the analysis being performed. The most
important caveats and assumptions associated with this analysis
are listed below. As a general rule, the model results presented
in this study are more useful for comparing the relative impacts
of alternative policies and bills than they are in estimating
absolute values.
. Growth in motor vehicle travel is estimated using national
averages for all areas. These national average growth rates
are different for each of the four vehicle types modeled.
Area specific growth rates are typically available, but they
do not permit separate rates to be specified for the four
vehicle types modeled, so they were not used. In any case,
motor vehicle projections in this analysis will not capture
city-by-city differences in travel.
. New stationary source growth is estimated using Bureau of
Economic Analysis values published in 1985. These rates may
overestimate growth in areas with petroleum-based economies.
. New source costs include all the costs of going from zero to
the indicated level of control. Some controls may be
undertaken for economic, process, or non-ozone related, non-
pollution control reasons. The costs of control designed at
the outset for newly constructed plants may well be lower
than the simple product of a cost per ton add-on control
times potential uncontrolled emissions based on present day
systems. Therefore, total cost estimates probably
overestimate the costs of the policies/bills for new
sources.
. The modeling approach used in this study may also be biased
toward estimating higher costs to existing sources than
might actually occur. Whenever a controlled existing source
is forced to increase its control level, ERCAM-VOC estimates
the cost of the new control equipment without taking into
account the salvage value or reduction in operating cost
associated with the previous control technique. Less costly
upgrades to current control systems are also not considered.
. The 1985 NEDS voc emission estimates for some area source
categories were adjusted downward to account for likely
control levels in nonattainment areas. This change affected
-------
emission estimates for the following area source
paper surface coating, degreasing, rubber and Di ,C?tegori««
manufacturing, and stage I gasoline marketing J! lcs
makes 1985 VOC emission estimates lower and provf^ chan9*
opportunity for future emission reductions w	lea«
were made to base year motor vehicle VOC emiss? ad^Ust*®nta
to try to include excess evaporative and running 68ti,,at*«
emissions because quantitative estimates nf *h 88
were not available during the study period	valuea
Rule effectiveness is almost always less than t ,
predicted. The proposed EPA policy states th»? "ally
only take 80 percent emission reduction credit- f"®*" c*n
measures. This 80 percent rule effectiveness r,^|V?rloaB
the policy is not modeled in this analysis. P 8lon
Where bills and policies call for control measur—
have not been previously demonstrated or studi«dfk!~ «
considerable uncertainty in control costs To
omitting important source types from the analy«i«Lftall
cost per ton values have been adopted for a nirnb*^
different control options, including controls for
miscellaneous point sources, consumer solvent controls
discretionary controls beyond those for which tiiar* ar« mam
data.
Ozone and CO design values from 1983 to 19§5 data have bmmm
used in this study. Estimated control r*qu 1 r***nt* by
would change if more recent data were used. Hot* also
these control requirements have been astlaatad viu a
simplified ozone trajectory model with consider**!*
uncertainty.
Not all of the policy and bill provision* cow Id tm
explicitly included in this analysis. For ii**t*«e». a®
attempt was made to quantify the *ff*ct* of vfUJa*t«9 «¦<
source review procedures. Futur* effect* of c*Jd at*#*
certification testing for motor vehici** w*r*
included in this analysis.
The point source data file	|*	tan*
incomplete data for pi.nt. th.t -..	^
year of VOC. Therefor.. thl. ««JY !%.!.<«> —
emission reductions	_9 out,te*-f
that make smaller VOC e.itt.r. .uei
Controi	of	emissions^^^		****
in the bills is nu*.	, a
attainment	Thus
^ct'of	'-.IT £• ••—•
effect or ri j	. t
assessing ^	:|	».«•>¦*<•** mm
needed in ar_• - .,c>i«i, •« *
Boston, are
-------
for the particular MSA, with assumed background ozone and
precursors. This may over or under estimate controls needed
to attain the standards, depending on the MSA involved.
. The modeling approach does not incorporate market
adjustments for existing sources as they respond differently
than anticipated to a new policy initiative. If this
occurs, the model probably overstates costs because
efficiencies associated with technological innovations,
economies of scale, process and product substitution, and
geographical migration are ignored.
. N0X costs have only been estimated for the explicit
provisions of the Waxman and Mitchell bills that require NOx
controls. Additional N0X controls may be undertaken in some
areas under the proposed EPA policy or the Group of Nine
proposal, but no attempt has been made to capture these
costs. The effects of N0X control on ozone concentrations
(plus or minus) have been ignored in all cases. These
assumptions could lead to overestimating or underestimating
MGX control costs and benefits, depending on the area
involved.
149
-------
ABBREVIATIONS AND ACRONYMS
ARB	Air Resources Board
BACT	Best Available Control Technology
BEA	Bureau of Economic Analysis
CCWI	Cost Components of Water Injection System
CCWT	Cost Components of Water Treatment
CMSA	Consolidated Metropolitan Statistical Area
CNG	compressed natural gas
CO	carbon monoxide
CTGs	Control Technique Guidelines
EAB	Ecoromic Analysis Branch
ERCAM	Emission Reduction and Cost Analysis Model
FIP	Federal Implementation Plan
FMVCP	Federal Motor Vehicle Control Program
HC	hydrocarbon
HDDVs	heavy-duty diesel-powered vehicles
HDGVs	heavy-duty gasoline-powered vehicles
I/M	inspection and maintenance
LDGTs	light-duty gasoline-powered trucks
LDGVs	light-duty gasoline-powered vehicles
LEA	low excess air
LNB	low N0X burners
MSA	Metropolitan Statistical Area
MTBE	Methyl Tertiary Butyl Ether
NAAQS	National Ambient Air Quality Standards
NEDS	National Emissions Data System
NESHAP	National Emissions Standards for Hazardous
Air Pollutants
NH 3	ammonia
NMOC	nonmethane organic compounds
N0X	oxides of nitrogen
NSPSs	New Source Performance Standards
O&M	operation and maintenance
POTWs	Publicly Owned Treatment Works
150
-------
ABBREVIATIONS AND ACRONYMS (continued)
PPM	parts per million
RACT	Reasonably Available Control Technology
RVP	Reid Vapor Pressure
SCCs	Source Classification Codes
SCR	Selective Catalytic Reduction
SIC	Standard Industrial Classification
SIP	State Implementation Plan
S02	sulfur dioxide
SOCMI	Synthetic Organic Chemicals Manufacturing
Industry
TSDFs	treatment, storage, and disposal facilities
VMT	vehicle miles traveled
VOC	volatile organic compound
VOCM	VOC Model
WCAP	Water Flow Capacity
151
-------
REFERENCES
Battye, 1987; William Battye, Alliance Technologies Corporation,
Chapel Hill, MC, "Ozone Cost Study Files," Memorandum and
computer files to Jim Wilson, E.H. Pechan & Associates,
Inc., April 3, 1987,
Battye et a 1., 1987: William H. Battye, Mark G. Smith, and Mark
Deese, Alliance Technologies Corporation, Chapel Hill, NC,
"Cost Assessment of Alternative National Ambient Air Quality
Standards for Ozone," Draft, prepared for U.S. Environmental
Protection Agency, OAQPS, Ambient Standards Branch, Research
Triangle Park, NC, October 1987.
Battye and Deese, 1988: William Battye and Mark Deese, Alliance
Technologies Corporation Chapel Hill, NC, "Breakdown of
Unassigned Emissions in the Ozone Model," Memorandum to Tom
McCurdy and Frank Bunyard, February 2, 1988.
BEA, 1985: Bureau of Economic Analysis, U.S. Department of
Commerce, Washington, DC, "1985 OBERS BEA Regional
Projections: Volume 2 - Metropolitan Statistical Area
Projections to 2035," 1985.
Blaszczak, 1988: Robert Blaszczak, U.S. Environmental Protection
Agency, Research Triangle Park, NC, personal communication,
March 30, 1988.
Bowen and Jennings, 1982: Michael L. Bowen and Mark S. Jennings,
Radian Corporation, Durham, NC, "Costs of Sulfur Dioxide
Particulate Matter, and Nitrogen Oxide Controls on Fossil
Fuel Fired Industrial Boilers," EPA-450/3-8 2-021, prepared
for U.S. Environmental Protection Agency, Office of Air,
Noise, and Radiation, OAQPS, August 1982.
Bunyard, 1988: Frank Bunyard, U.S. Environmental Protection
Agency, OAQPS, Research Triangle Park, NC, personal
communication, March 1988.
CARB, 1987: California Air Resources Board, "A Technical
Assessment Document for Reducing Photochemica11y Reactive
Organic Compound Emissions from Underarm Products,"
Preliminary Draft, May 1987.
CARB and SCAQMD, 1987: California Air Resources Board and South
Coast Air Quality Management District, "Technical Support
Document for a Suggested Control Measure for the Control of
Emissions of Oxides of Nitrogen from Industrial,
Institutional, and Commercial Boilers, Steam Generators and
Process Heaters," prepared for the Statewide Technical
Review Group, April 1987.
Chem. Engr., 1988: "Chemical Engineering," Economic Indicators,
McGraw-Hill, Inc., January 18, 1988.
152
-------
Cherry and Hunter, 1980: S.S. Cherry and S.C. Hunter, KVB, Inc.,
"Cost and Cost Effectiveness of N0X Control in Petroleum
Industry Operations," API Publication 4331, prepared for the
American Petroleum Institute, October 1980.
Cutino, 1987: Judy Cutino, Bay Area Air Quality Management
District, personal communication, July 7, 1987.
EEA, 198 2: Energy and Environmental Analysis, Inc., "Cost and
Economic Assessment of Regulatory Alternatives for NO2
NAAQS," draft, prepared for U.S. Environmental Protection
Agency, June 1982.
EEA, 1987: Energy and Environmental Analysis, Inc., Arlington,
VA, "The Motor Fuel Consumption Model," 13th Periodical
Report, prepared for Martin Marietta Energy Systems, Inc.,
Oak Ridge, TN, May 26, 1987.
EPRI, 1985: Electric Power Research Institute, "S02 and NOx
Retrofit Control Technologies Handbook," EPRI CS-4277-SR,
October 1985.
EPRI, 1986: Electric Power Research Institute, "TAG - Technical
Assessment Guide Volume 1; Electricity Supply - 1986," EPRI
P-4463-SR, December 1986.
Flynn, 1985: Peter C. Flynn, "Natural Gas as a Vehicle
Transportation Fuel," in Proceedings of the 3 5th Canadian
Chemical Engineering Conference. Vol. 2. Ottawa, Ontario,
Canadian Society for Chemical Engineering, 1985.
Hess, 1987: Peter F. Hess, Bay Area Air Quality Management
District, San Francisco, California, "Ozone Control
Strategies in the San Francisco Bay Area," presented at the
80th Annual Meeting of the Air Pollution Control
Association, New York, NY, June 1987.
Johnson, 1988: David Johnson, OAQPS, U.S. Environmental
Protection Agency, Research Triangle Park, NC, personal
communication, May 10, 1988.
Jones and Johnson, 1979: Gary D. Jones and Kevin L. Johnson,
Radian Corporation, Austin, TX, "Technology Assessment
Report for Industrial Boiler Applications: N0X Flue Gas
Treatment," EPA-600/7-79-178g, prepared for U.S.
Environmental Protection Agency, Office of Research and
Development, December 1979.
Jordan, 1988: Bruce Jordan, Chief, Ambient Standards Branch,
AQMD, OAQPS, Research Triangle Park, NC, "Review of Cost for
Unassigned Categories," memorandum to Gene Durman, OPPE,
February 8, 1988.
153
-------
KVB, 1984: KVB, Inc., "Evaluation of Natural - and Forced -
Draft Staging Air Systems for Nitric Oxide Reduction in
Refinery Process Heaters, Volume 1," EPA-600/7-84-080a,
August 1984.
Kimbrough, 1988: Sue Kimbrough, U.S. Environmental Protection
Agency, OAQPS, Research Triangle Park, NC, personal
communication, January 6, 1988.
Lorang, 1985: Phil Lorang, Office of Mobile Sources, U.S.
Environmental Protection Agency, Ann Arbor, HI, "New York
I/M Credits for Heavy Duty Gasoline Vehicles," Memorandum to
John Walsh, Air Programs Branch, Region II, May 9, 1985.
Lorang, 1988: Phil Lorang, Office of Mobile Sources, U.S.
Environmental Protection Agency, Ann Arbor, MI, "Information
on Motor Vehicle Emissions Reductions for OPPE's
Contractor," Information sheet, February 1988.
Pechan, 1986: E.H. Pechan & Associates, Inc., "Acid Deposition
Control Techniques for the New England States," prepared for
Northeast States for Coordinated Air Use Management, July
1986.
Pechan, 1987: E.H. Pechan & Associates, Inc. "New Technologies
and Utility Emissions," prepared for Office of Policy
Analysis, U.S. Environmental Protection Agency, January
1987.
PEDCo, 1979: Letter from Donald J. Henz to Susan E. Schechter,
PEDCo Environmental, Inc., Cincinnati, OH, May 14, 1979.
Radian, 1985: Radian Corporation, "Control Costs for the
Volatile Organic Compound Model," prepared for U.S.
Environmental Protection Agency, November 1985.
Radian, 1988a: Radian Corporation, Sacramento, CA, "Feasibility
and Cost-Effectiveness of Controlling Emissions from Diesel
Engines in Rail, Marine, Construction, Farm, and Other
Mobile Off-Highway Equipment," February 1988.
Radian, 1988b": Radian Corporation, "Technical Support for
Proposed Prevention of Significant Deterioration Rule for
Nitrogen Oxides, Volume II," prepared for U.S. Environmental
Protection Agency, February 1988.
Sierra Research, 1988: Sierra Research, Inc., Sacramento, CA,
"The Feasibility and Costs of More Stringent Mobile Source
Emission Controls," prepared for U.S. Congress Office of
Technology Assessment, January 20, 1988.
Stockton and Stelling, 1987: Margie B. Stockton and John H.E.
Stelling, Radian Corporation, Research Triangle Park, NC,
"Criteria Pollutant Emission Factors for the 1985 NAPAP
154
-------
Emissions Inventory," EPA-600/7-87-015, prepared for U.S.
Environmental Protection Agency, Air and Energy Engineering
Research Laboratory, May 1987.
U.S. Bureau of the Census, 1985: U.S. Bureau of the Census,
Washington, DC, "Statistical Abstract of the United States
1986," 106th Edition, 1985.
U.S. EPA, 1978; U.S. Environmental Protection Agency, OAQPS,
Research Triangle Park, NC, "Summary of Group I Control
Technique Guideline Documents for Control of Volatile
Organic Emissions from Existing Stationary Sources," EPA-
450/3-78-120, December 1978.
U.S. EPA, 1980: U.S. Environmental Protection Agency, OAQPS,
Research Triangle Park, NC, "Methodology to Conduct Air
Quality Assessments of National Mobile Source Emission
Control Strategies," EPA-4 50/4-80-02 6, October 1980.
U.S. EPA, 1981: U.S. Environmental Protection Agency, OAQPS,
Research Triangle Park, NC, "Control of Volatile Organic
Emissions from Architectural Surface Coatings," Draft, March
1981.
U.S. EPA, 1984: U.S. Environmental Protection Agency, Motor
Vehicle Emission Laboratory, Ann Arbor, MI, "User's Guide to
M0BILE3 (Mobile Source Emissions Model)," EPA 460/3-84-002,
June 1984.
U.S. EPA, 198 5: U.S. Environmental Protection Agency, OAQPS,
Research Triangle Park, NC, "Compilation of Air Pollutant
Emission Factors: Volume 1," AP-42, 4th ed., September
1985.
U.S. EPA, 1986a: U.S. Environmental Protection Agency, OAQPS,
Research Triangle Park, NC, "Control Techniques for Volatile
Organic Compound Emissions from Stationary Sources," 3rd
ed., Draft, March 1986.
U.S. EPA, 1986b: U.S. Environmental Protection Agency, Research
Triangl-e Park, NC, "EAB Control Cost Manual, " 3rd ed. ,
November 1986 .
U.S. EPA, 1987a: U.S. Environmental Protect ion Agency, Office of
Air and Radiation, Washington, DC, "Draft Regulatory
Analysis: Proposed Refueling Emission Regulations for
Gasoline-Fueled Motor Vehicles: Volume 1," March 1987.
U.S. EPA, 1987b: U.S. Environmenta1 Protection Agency, Ann
Arbor, MI, "Further Reflections on the Costs of the Mobile
Source Provisions of the Mitchell Bill," Draft, September
1987 .
155
-------
U.S. EPA, 1937c: U.S. Environmental Protection Agency, OAQPS,
Research Triangle Park, NC, "Implications of Federal
Implementation Plans (FIPs) for Post-1987 Ozone
Nonattainment Areas," Draft, March 1987.
U.S. EPA, 1987d: U.S. Environmental Protection Agency, Office of
Mobile Sources, Ann Arbor, MI, "Inspection/Maintenance
Program Implementation Summary," January 1987.
U.S. EPA, 1988a: U.S. Environmental Protection Agency, Office of
Mobile Sources, Ann Arbor, MI, "Guidance on Estimating Motor
Vehicle Emission Reductions from the Use of Alternative
Fuels and Fuel Blends," EPA-AA-TSS-PA-87-4, January 29,
1988.
U.S. EPA, 1988b: U.S. Environmental Protection Agency, Office of
Air Quality Planning and Standards, Ambient Standards
Branch, table of ozone design values and emission reduction
requ -ements, March 1988.
Weiser, 1 38: Ted Weiser, Office of Mobile Sources, U.S.
Environmental Protection Agency, Ann Arbor, MI, personal
communication, April 4, 1988.
Wright and Klausmeier, 1988: David A. Wright and Robert F.
Klausmeier, Radian Corporation, Austin, TX, "Potential
Emission Reductions From Including Heavy-Duty Gasoline
Powered Vehicles in Inspection and Maintenance Programs,"
presented at the 81st Annual Meeting of APCA, June 198®.
-------
APPENDIX A
N0X CONTROL COST EQUATION DEVELOPMENT
A. RACT LEVEL CONTROL EQUATIONS
1. Industrial and Utilitv Boilers
The cost equations for the RACT level control of NOx from
industrial boilers were derived from cost data given in an
industrial boiler cost report (Bowen and Jennings, 1982). For
each type of fuel and control method, at least three different
boiler sizes were costed. When more than one control method was
listed for a given type of boiler, the control technique yielding
the highest N0X control efficiency was chosen.
The cost equations for stokers and oil and gas-fired
industrial boilers were also applied to the same types of uti lity
boilers for lack of any better data for these utility boilers.
The validity of applying the cost equations developed for
industrial boilers to utility boilers is uncertain. Considering
that the same types of modifications would be made in applying
the same types of control techniques to either utility or
industrial boilers, it is expected that this assumption is
reasonable. The greatest difference between utility and
industrial boilers is size. (Utility boilers are generally
larger than industrial boilers.) In many instances, though, no
real distinction exists between the two types. Therefore, it is
expected that the application of the industrial boiler cost
equations to utility boilers should not cause a large degree of
error. The equations used are all listed in Tables A.l and A.2.
The SCC categories to which the cost equations were applied are
listed in Table A.3.
The low excess air (LEA) control technique, used for
distillate oil boilers and stokers, results in a net savings.
This results from an increase in the boiler efficiency when
implementing LEA. The capital costs for this technology are
relatively low and so the savings in O&M expenses produce an
overall cost savings. The high savings per ton achieved by
distillate oil industrial boilers is somewhat misleading.
157
-------
Table A.1
NOx Control Cost Equations for Utility and Industrial Boilers
Capital Cost
Operating & Maintenance


Equations
Cost Equations



Control




Control
Defaul
Source Type
Device
Coefficient
Exponent
Coefficient
Exponen t
Eff. %
Cost Per
tSS SS SSZ X 9C «En 3£ £S- SS S ££ £ 3 23 £5 Si ££' -£¦» -3. 3
Utility Boilers







PC - Wall/Opposed
LNB
7,860
0.72
393
0.72
50
87
PC - Tangential
LNB
232,400
0.40
11,620
0.40
50
232
Residual Oil
SCA
10,480
0.62
600
0.84
42
353
Gas
FGR
6,610
0.43
450
1.00
31
983
Stoker
LEA
3,730
0.44
-67
1.11
21
-525
Coal
SCR
292,400
0.60
4,500
1.00
80
2911
Gi1/Gas
SCR
265,800
0.50
2,370
1.00
80
3120
Industrial Boilers







Pulverized Coal
SCA
1,910
0.70
186
0.96
36
2198
Stoker
LEA
3,730
0.44
-67
1.11
21
-337
Residual Oil
SCA
10,480
0.62
600
0.84
42
827
Distillate Oil
LEA
3,960
0.36
-690
1.00
36
-4592
Gas
FGR
6,610
0.43
450
1.00
31
1025
Coal
SCR
147,900
0.70
4,600
0.95
80
3278
011/Gas
SCR
134,450
0.60
2,425
0.95
80
3667
NOTES: All equations are of the form COST - COEFFICIENT*(BOILER DESIGN CAPACITY)"EXPONENT
Units for BOILER DESIGN CAPACITY are in MMBtu/hr
All costs are in 1985 dollars
-------
Table A.2
NOx Control Cost Equations for IC Engines, Gas Turbines, and Process Beaters
SOURCE
IC Engines
Gas
Oil
Gas
Oil
Gas Turbines
Gas
m Oil
•sO Gas
Oil
CONTROL
METHOD
: s s a a x :
Change A/P Ratio
Change A/F Ratio
SCR
SCR
water Injection
Water Injection
SCR+Uater Injection
SCR+Water Injection
Process	Beater
Gas	SC A
Oil	SCA
Gas	SCR
Oil	SCR
CAPITAL COST EQUATIONS
tsssssss:
0
0
8,802,000*(DESRATE)"0.86
1,556,000*(DESRATE)"0.86
1,393,000*(DESRATE)"0.52
508,000*(DESRATE)-0.52
10,031,000*(DESRATE)"0.74
2,283,000*(DESRATE)*0.74
A7,260*(DESRATE)"0.67
12,830*(DESRATE)"0.67
5,774,000*(DESRATE)"0.60
1,780,000*(DESRATE)"0.60
O&M COST EQUATIONS
CONTROL DEFAULT
EFF(Z) COST PER TON
issscsas
sassssasssaasac:
: = s: =
574*(0PRATE)
65.8*(OPRATE)
131*(OPRATE)+5,355,000*(DESRATE)
18.1*(0PRATE) + 714,000*(DESRATE)
174*(OPRATE)
22. 1*(0PRATE)
179*(OPRATE)+1,700,000*(DESRATE)
23.1*(0PRATE) ~ 227,000*(DESRATE)
•65,100*(DESRATE)
-9,300*(DESRATE)
221*(OPRATE)
29.8*(0PRATE)
30
30
80
80
70
70
94
94
45
45
90
90
1126
935
964
936
1560
1020
3730
2480
-306
-110
7810
2.760
NOTES: DESRATE is the maximum design rate in SCC units per hour
OPRATE is the operating rate in SCC units per year
All costs are in 1985 dollars
ft
-------
Table A.3
SCC Codes Corresponding to NOx Control Cost Equations
Source Category	Control Type	Applicable SCCs
Utility Boilers
Pulverized Coal
Wall/Opposed	LNB
Pulverized Coal
Tangentially	LNB
Residual Oil	SCA
Gas
FGR
Stoker	LEA
10100101
10100201
10100202
10100221
10100222
10100301
10100212
10100226
10100302
10100401
10100404
10100405
10100406
10100601
10100602
10100604
10100701
10100702
10100102
10100204
10100205
10100224
10100225
10100304
10100306
160
-------
Table A.3
SCC Codes Corresponding to NOx Control Cost Equations
Source Category	Control Type	Applicable SCCs
Coal	SCR	10100101
10100201
10100202
10100221
10100222
10100301
10100212
10100226
10100302
10100102
10100204
10100205
10100224
10100225
10100304
10100306
Oil/Gas	SCR	10100401
10100404
10100405
10100406
10100501
10100504
10100505
10100601
10100602
10100604
10100701
10100702
161
-------
SCC Codes
Source Category
Table A.3
Corresponding to NOx Control Cost Equations
Control Type	Applicable SCCs
Industrial Boilers
Pulverized Coal
SCA
10200201
10200202
10200212
10200221
10200222
10200226
10200301
102.00302
10300101
10300102
1Q30C 35
10301 06
10300216
10300221
10300222
10300226
10300305
10300306
Stoker
LEA
10200204
10200205
10200206
10200224
10200225
10200304
10200306
10300207
10300208
10300209
10300224
10300225
10300307
10300309
162
-------
Table A.3
SCC Codes Corresponding to NOx Control Cost Equations
Source Category	Control Type	Applicable SCCs
Residual Oil	SCA	10200401
10200402
10200403
10200404
10300401
10300404
Distillate Oil	LEA	10200501
10200502
10200504
10300501
10300504
Gas	FGR	10200601
10200602
10200603
10200701
10200704
10200707
10300601
10300602
163
-------
SCC Codes
Source Category
Table A.3
Corresponding to NOx Control Cost Equations
Control Type	Applicable SCCs
Coal	SCR	10200201
10200202
10200212
10200221
10200222
10200226
10200301
10200302
10300101
10300102
10300205
10300206
10300216
10300221
10300222
10300226
10300305
10300306
10200204
10200205
10200206
10200224
10200225
10200304
10200306
10300207
10300208
10300209
10300224
10300225
10300307
10300309
164
-------
Tafa1e A,3
see Codes Corresponding to NOx Control Cost Equations
Source Category	Control Type	Applicable SCCs
Oil/Gas	SCR	10200401
10200402
10200403
10200404
10300401
10300404
10200501
10200502
10200504
10300501
10300504
10200601
10200602
10200603
10200701
10200704
10200707
10300601
10300602
IC Engines
Gas
Change AFR
20100202
20100702
20200202
20200204
20300201
Oil	Change AFR	20100102
20100902
20200102
20200104
20200301
20200401
20200501
20200902
20300101
20300301
165
-------
Table A.3
SCC Codes Corresponding to NOx Control Cost Equations
Source Category	Control Type	Applicable SCCs
Gas
SCR
20100202
20100702
20200202
20200204
20300201
Oil
SCR
20100102
20100902
20200102
20200104
20200301
20200401
20200501
20200902
20300101
20300301
Gas Turbines
Gas
Water Inj
20100201
20200201
20200203
20300202
Oil	Water Inj.	20100101
20200101
20200103
20300102
Gas	Water Inj.
& SCR	20100201
20200201
20200203
20300202
166
-------
Table A.3
SCC Codes Corresponding to NOx Control Cost Equations
Source Category	Control Type	Applicable SCCs
Oil
Water Inj
& SCR
20100101
20200101
20200103
20300102
Process Heaters
Gas
. SCA
30600104
30600105
30600106
Oil
SCA
30600103
Gas
SCR
30600104
30600105
30600106
Oil
SCR
30600103
167
-------
Distillate oil-fired boilers have a much lower emission rate than
coal-fired boilers. Thus, any reduction in NOx emissions will
also be relatively small. Dividing the negative annual cost by a
small number leads to this large savings per ton, making LEA
appear to be very cost effective for distillate boilers. In
actuality, if the emission rate had been greater, leading to a
larger reduction in emissions, while maintaining the same annual
cost savings, the cost effectiveness would actually decrease.
This contradiction is due entirely to the negative cost.
The remaining two RACT level control cost equations, for LNB
applied to pulverized coal-fired utility boilers, were based on
equations given for this control technique in a (Pechan, 1987)
report. These retrof:. equations were based on the size of the
boiler in MW and were simply converted to accept the boiler size
in MMBtu/hr. The use of LNB is expected to decrease NOx
emissions from wall-fired and opposed-fired utility boilers and
tangentia1ly fired utility boilers by 50 percent. Tangentially
fired boilers are much more difficult to retrofit with LNB than
either wall-fired or opposed-fired units, and they emit only
about one half as much N0X in the uncontrolled state as the wall-
fired and opposed-fired boilers. As a result, the cost per ton
to control the tangentially fired units is much higher than that
of the other types of pulverized coal-fired utility boilers.
2. Internal Combustion Engines
Cost equations for reciprocating internal combustion engines
(EEA, 1982} were updated and revised for this analysis. The RACT
level method of control used is a combustion modification of fine
tuning the engine controls and changing the air/fuel ratio of the
engine. This technique is expected to give a 30 percent
reduction in MOx emissions. No capital costs are incurred by
making these adjustments. These process modifications do incur
O&M expenses, however, including a fuel penalty for the
additional fuel consumed. The retrofit equations for O&M costs
and the fuel penalty were combined since the other N0X cost
equations incorporate fuel costs or savings into the O&M
168
-------
equations. The fuel costs used are the expected long-term fuel
prices (Pechan, 1986). The natural gas price used was
$5.08/MMBtu, and $4.54/MMBtu was used as the oil price. The cost
equations are listed in Table A.2.
3. Gas Turbines
The set of cost equations for water injection applied to gas
turbines was derived from data in Radian (1988b). These
equations apply to a N0X removal efficiency of 70 percent, using
a water to fuel ratio of 1:1. Using water injection with gas
turbines leads to approximately a 1 percent reduction in engine
efficiency. Therefore, before calculating any costs, the actual
fuel consumption rate with the water injection system in place
was calculated.
The total capital cost of applying a water injection control
system to a gas turbine is composed of capital cost components
for the water injection system and for water treatment. Both of
these components are based on the water flow capacity (WCAP), in
gallons per minute, of the water injection system. The capital
cost components of water treatment (CCWT) and of the water
injection system (CCWI) are given by the following equations:
CCWT = 59,200 * (WCAP) °-53
CCWI = 45,300 * (WCAP) °*5
The total capital cost of water injection is the sum of these two
cost components multiplied by an assumed retrofit factor of 1.2.
The resulting capital cost equations for applying water injection
to gas turbines are listed in Table A.2.
The annual O&M costs associated with controlling gas
turbines by water injection include the cost of water consumption
as well as the cost of increased fuel consumption due to the
reduction in engine efficiency. The cost of water is a function
of the operating rate since the amount of water used is directly
proportional to the amount of fuel consumed.
Using a unit cost of water in 1985 of $0.60/1,000 gal
(Radian, 1988b) and with the operating rate in SCC units per
169
-------
year, the component for the annual cost of water is given by the
following expressions:
AGWqil = 534 * (0PRATE)
ACWgas = 3. fa 3 * (0 PRATE)
The other cost component which must be included in the final
O&M cost equations is the annual cost for the increase in fuel
use due to the decrease in turbine efficiency. The following
expressions were derived for the increase in the annual cost of
fuel:
ACF0I - 21.6 ~ (0 PRATE)
ACFGAS 171 # (0PRATE)
Th# final OfcM vquatlona for	wmtmr I n f met I an U
NOx	fr'fw -]«• f urtiiniM v«r» <~< • t mmn& if	-t »«' «r	,,»-**»* t«
T * t> ! ¦ K ;

¦	* f#
¦ » iili * !'
-------
reduction in N0X emissions. The actual capital costs involved in
retrofitting a specific boiler with SCR depend on the site
requirements of the unit- Therefore, capital costs could be as
much as two times greater or two times less than those predicted
by the cost equations. The O&M equations are based on an
operating capacity factor of 1.0. Units operating at less than
100 percent capacity will incur O&M costs proportional to their
operating capacity. The cost equations for retrofitting oil-
fired industrial boilers with SCR are based on data reported by
the California Air Resources Board (CARB, 1987).
Because there were no relatively current data available for
SCR costs applied to pulverized coal industrial boilers, the 1979
Technology Assessment Report (Jones and Johnson, 1979) was used
to determine the relationship between SCR costs for coal-fired
and oil-fired industrial boilers. Data for parallel flow SCR for
both types of boilers were compared and the following
relationship for capital costs was obtained, with the size in
MMBtu/hr:
CAP^oal = 1-1 * (SIZE)0"1
CAPoil
This ratio was applied to the capital cost equation for oil-fired
boilers to derive the equation for SCR capital costs for
pulverized coal-fired boilers.
The same procedure was followed in deriving the O&M cost
equation. The O&M cost of SCR for a coal-fired boiler is 1.9
times greater than the O&M cost of applying SCR to an oil-fired
industrial boiler.
The cost equations for applying SCR to coal-fired utility
boilers are based on EPRI (1985) studies. EPRI provides cost
estimates for applying SCR, yielding 80 percent N0X removal, to
four plants of the same size but with different retrofit
difficulties. The average of these four cases, $7 3/kW, was taken
as the base case for capital costs. To derive a capital cost
equation, it was assumed that cost varies with size to the 0.6
.17 1
-------
power, a frequently used assumption when data are unavailable.
The average O&M cost of the four cases was 8 mills/kWh. The
resultant G&M equation assumes that the O&M costs vary linearly
with size.
To determine the costs of SCR retrofitted to oil-fired
utility boilers, it was assumed that the same relationship
existed between coal-fired and oil-fired utility boilers as was
found to exist between coal-fired and oil-fired industrial
boilers. The resultant equations are listed in Table A.l.
2. Internal Combustion Engines
Information on the cost of SCR applied to internal
combustion engines, as well as the cost information on SCR for
all the nonboiler sources, was obtained from Radian (1988).
Items included in the calculation of the capital costs for SCR
are the catalyst, the reaction vessel, the ammonia injection
system, and the ammonia injection control system. The final
capital cost equations for applying SCR to oil-fired and gas-
fired internal combustion engines are listed in Table A.2.
The annual O&M cost consists mainly of the cost of catalyst
replacement and ammonia. It was conservatively assumed that the
catalyst would need to be replaced every 2 years. Thus, the O&M
catalyst replacement cost will be approximately one-half of the
installed SCR equipment cost, excluding the cost of the ammonia
control system. With the design rate in SCC units/hr, the O&M
catalyst replacement cost in 1985 dollars is given by the
following equations:
GAS-FIRED: 0&MCAT = 5,355,000 * (DESRATE)
OIL-FIRED: 0&MCAT = 714,000 * (DESRATE)
The amount of ammonia required is dependent on the inlet
rate of N0X to the SCR reactor. In determining the annual cost
of ammonia, it was assumed that the molar ratio of ammonia (NH3)
to NOx would be 0.93:1 (Radian, 1988). A value of $150/ton NH3
was used as the unit cost of ammonia in accordance with the EPRI
17 2
-------
Technical Assessment Guide (1986). The equations used to
calculate the annual cost of ammonia are provided below:
GAS-FIRED: 0&MNH3 = 131 * (0PRATE)
OIL-FIRED: G&MNH3 = 18.1 * (OPRATE)
The total annual O&M costs for internal combustion engines
were obtained by adding the catalyst replacement cost and the
annual cost of ammonia. Because the amount of catalyst needed is
dependent on the size of the engine, while the amount of ammonia
needed is dependent on the actual system operating rate, both the
design rate and the operating rate are included in the final O&M
cost equations which are listed in Table A.2.
It is assumed that applying SCR to IC engines will result in
an 80 percent reduction in N0X emissions. This reduction will
actually vary somewhat, depending on the catalyst. When the
catalyst is new, it is likely to remove approximately 90 percent
of the N0X emissions, but the ability of the catalyst to reduce
NOx emissions will diminish as it ages. For this reason, it is
important that the catalyst be replaced on a regular basis to
insure high reduction potential.
3• Gas Turbines
At the BACT level of N0X control, a combination of water
injection and SCR can produce an overall reduction in N0X
emissions of 94 percent. The water injection removes the first
70 percent of the N0X emissions and the SCR can remove an
additional 80 percent of the NOx emissions entering the SCR
reactor. The cost, size, and performance of the water injection
system are unaffected by the presence of SCR. The capital cost
of SCR is not affected by the presence of water injection, but
the O&M costs for the SCR will be reduced over those of an SCR
system alone. The amount of ammonia required will be decreased
since the amount of NOx entering the SCR reactor has already been
reduced by 70 percent. The derivation of the water injection
cost equations has already been described in a previous section.
Therefore, only the costs relating to the SCR system and the
173
-------
combined water injection and SCR cost equations are discussed
here.
The capital cost of the SCR system is broken down into two
components -- the capital cost of the catalyst and the capital
cost of the remaining equipment. The remaining equipment
includes the reactor housing, the ammonia injection system, and
the ammonia control system plus the cost of installation of the
SCR system.
The capital cost of the catalyst is expected to be directly
proportional to the size of the turbine with no economies of
scale. This is because the catalyst is sized in direct
proportion with the gas flow rate entering the system to achieve
a given removal efficiency. Since the gas flow rate determines
the amount of catalyst needed and because the catalyst is made
from an expensive metal oxide, any economy of scale which might
exist would be minimal.
As opposed to the capital cost of the catalyst, the capital
cost of the remaining equipment is expected to have an economy of
scale. The cost should vary with size to the 0.6 power. The
total capital cost of the SCR system is given by the following
equations:
GAS-FIRED: CAPSCR = 8,641,000 * (DESRATE)0•78
OIL-FIRED: CAPSCR = 1,801,000 * (DESRATE)0*78
The final capital cost equations for water injection combined
with SCR applied to gas turbines are listed in Table A.2.
The O&M costs for water injection plus SCR are the sum of
the water injection O&M costs plus the O&M cost of SCR using the
reduced gas flow rate entering the SCR reactor. The O&M costs
included for the SCR system are the catalyst replacement cost and
the ammonia cost. Assuming that the catalyst must be replaced
every 2 years to maintain the desired catalyst activity, the O&M
catalyst cost is one-half of the capital catalyst cost. The
following equations give the O&M catalyst costs with the design
rate in SCC units/hr:
174
-------
GAS-FIRED: 0&MCAT = 1,700,000 * (DESRATE)
OIL-FIRED: ()&MCAT = 227,000 * (DESRATE)
The SCR ammonia costs are affected by the use of water
injection since the amount of ammonia required for the SCR system
is dependent on the amount of N0X in the gas stream. The SCR
will remove 80 percent of the N0X emissions remaining after water
injection. The ammonia O&M cost equations for the SCR system are
given below, with OPRATE being the turbine operating rate in SCC
units/year:
GAS-FIRED: ACA = 5.01 * OPRATE
OIL-FIRED: ACA = 1.001 * OPRATE
To obtain the final O&M cost equations for a combined water
injection and SCR control system applied to gas turbines, the O&M
component for the water injection costs, the SCR catalyst
replacement costs, and the ammonia costs were summed. These
final O&M equations, listed in Table A.2, are functions of both
the design rate and the operating rate.
4. Process Heaters
The cost equations for SCR applied to process heaters were
based on Radian {1988} data. It was assumed that the capital
cost would vary to the 0.6 power with size. A retrofit factor of
1.2 was applied to account for the difficulties of applying SCR
to site specific conditions as opposed to applying SCR to a new
unit. The resulting capital cost equations are listed in Table
A. 2 .
The O&M costs were assumed to vary linearly with the
operating rate of the unit. Radian's base case cost estimates
were converted to 1985 dollars in SCC units. As the Radian
estimate was based on a 100 percent capacity utilization, the
equation was divided by 8,760 hours per year to allow the annual
operating rate to be used as the equation variable. The final
0&M cost equations for SCR applied to process heaters are listed
in Table A.2.
1 7 5
-------
I
I
-------