United States Office of Air and Radiation EPA-450/3-87-001 b
Environmental Protection (ANR-443) July 1987
Agency , Washington, DC 20460
Air
Draft Regulatory
Impact Analysis:
Proposed Refueling
Emission Regulations
for Gasoline-Fueled
Motor Vehicles —
Volume II
Additional Analysis of
Onboard Controls
-------
EPA-450/3-87-001b
Draft Regulatory Impact Analysis: Proposed
Refueling Emission Regulations for Gasoline-
Fueled Motor Vehicles —
Volume II
Additional Analysis of
Onboard Controls
OFFICE OF AIR QUALITY PLANNING AND STANDARDS
AND
OFFICE OF MOBILE SOURCES
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Radiation
Washington, DC 20460
July 1987
-------
This report has been reviewed by the Office of Air Quality Planning and Standards and the Office of Mobile
Sources, EPA, and approved for publication Mention of trade names or commercial products is not intended
to constitute endorsement or recommendation for use. Copies of this report are available through the Library
Services Office (MD-35), U.S. Environmental Protection Agency, Research Triangle Park, N.C 27711, or
from the National Technical Information Services, 5285 Port Royal Road, Springfield, Virginia 221 61
-------
1.0 INTRODUCTION
This second volume (Volume II) supplements the first
volume of the draft Regulatory Impact Analysis in fulfilling
the objectives of Title 3 - Executive Order 12291 as they apply
to the proposed onboard rulemaking. It provides a summary of
the more extensive support analyses of onboard costs and air
quality benefits that were prepared in the course of developing
the proposed regulations. Alternatives to the proposed
regulations are considered in Volume I.[l]
More specifically, following this introduction (Chapter
1), Chapter 2 of this volume summarizes the economic impact of
the proposed rulemaking. It identifies the fixed and variable
costs to manufacturers for systems development, certification,
facility modifications, and emission control hardware. It also
addresses costs to consumers in terms of both first price
increase for hardware and operating costs or savings. Finally,
the chapter summarizes aggregate costs of the proposed
regulations to the nation by year incurred, and the
socioeconomic impact of the proposed rulemaking. The latter
category includes a brief discussion of the impacts on the
financial status of manufacturers and component vendors,
effects on sales and employment in the automotive and petroleum
industries, effects on energy usage, balance of trade, and on
particular segments of the economy.
Chapter 3 summarizes the air quality benefits of the
proposed rulemaking. Beginning with a brief characterization
of refueling emissions potentially controllable by an onboard
control strategy, the chapter goes on to analyze the effect of
refueling emissions control on ambient ozone levels in current
non-attainment areas, i.e. those areas now in violation of the
National Ambient Air Quality Standards for ozone. The direct
health effects of ozone have been fully documented in other EPA
publications and so they are not included in this analysis.
-------
REFERENCES FOR CHAPTER 1
Draft Regulatory Impact Analysis: Proposed Refueling
Emission Regulations for Gasoline-Fueled Motor Vehicles —
Volume I - Analysis of Gasoline Marketing Regulatory
Strategies, U.S. Environmental Protection Agency, Office
of Air and Radiation, Office of Air Quality Planning and
Standards and Office of Mobile Sources, EPA-450/3-87-001a,
July 1987.
1-2
-------
2.0 ECONOMIC IMPACT
I. Introduction
This chapter assesses the economic impact of onboard
control of refueling emissions in terms of the costs to
manufacturers, costs to the consumer, total cost to the» nation
and the overall socioeconomic impact of the proposed
regulations. The purpose of this chapter is to summarize and
assemble into one comprehensive analysis those cost elements
that have been developed in greater detail than in "Evaluation
of Air Pollution Regulatory Strategies for Gasoline Marketing
Industry - Response to Public Comments" (hereafter referred to
as the "response to public comments document ), and other
support documents.[1]
In addition to controlling refueling emissions, properly
designed and tested onboard control systems also have the
potential to control excess evaporative emissions. However,
other approaches to controlling excess evaporative emissions
are being evaluated separately, so this analysis will focus
primarily on the costs of refueling emissions control alone.
Specifically, the costs and credits involved in onboard control
of refueling emissions described here are incremental to
vehicle-based excess evaporative control costs.
This chapter is organized into four sections (1) costs to
manufacturers, i.e., systems engineering costs, certification
costs, facility modifications and hardware costs, (2) costs to
the consumer, including first price increase and operating
costs or savings, (3) aggregate costs to the nation, by year
incurred, and (4) socioeconomic impacts of the regulations.
The latter would include any effects the proposed regulations
might have on manufacturers' or vendors financial status,
effects on sales and employment in the automotive and petroleum
industries, and effects on energy usage, balance of trade, or
on particular segments of the economy.
Data from a wide variety of sources were utilized to
develop the cost estimates summarized in this chapter. The
primary sources for the hardware cost estimates at the vendor
level are studies done in 1978 and 1983 by Leroy H. Lindgren.
[2,3] Retail Price Equivalent (RPE) markups, used to calculate
costs to the consumer from vendor costs, were developed by an
EPA contractor.[4] These works have been supplemented where
appropriate by other EPA contract studies, comments on the
draft gasoline marketing study from the automotive and
petroleum industries, supplier quotations, trade publications
and previous estimates from EPA regulatory support analyses.
Most of the estimates of vehicle miles traveled (VMT) and fuel
consumption data used in assessing operating costs were
obtained from EPA's MOBILE3 emission factors and fuel
consumption models. Additional vehicle usage data came from
the Department of Commerce "Truck Inventory and Use Survey"
(TIUS) and an SAE paper on fuel economy done by EPA. [5,6]
-------
Vehicle sales estimates are based on the Data Resources
Incorporated 25-year "Trendlong" projections.[7] The
dieselization rates applied to these projections to obtain
gasoline-fueled vehicle sales estimates are from the MOBILES
mode1.
In determining costs to the industry, consumers, and
aggregate costs to the nation, various cost elements must be
amortized over different periods of time, and in portions of
the analysis costs must be discounted to reflect the time value
of money. Systems engineering and certification costs, for
example, are relatively short term and are likely to be
incurred and recovered early in the regulation implementation
period. These costs will therefore be amortized over a five
year period 1990-1994, assuming a 1990 implementation date for
the proposed regulations. The other fixed costs, those for
modification and construction of testing facilities, are
longer-term and will be amortized over a ten year period
1990-1999. Variable costs will be aggregated over the initial
five-year period to show the highest annual costs and will be
discounted to reflect the time value of money, where
necessary. The standard ten percent discount rate will be
used, and all fixed costs will be amortized at a 10 percent
interest rate. All costs are expressed in 1986 dollars.
II. Costs to Manufacturers
Costs to manufacturers can be divided into two general
categories, fixed and variable costs. Fixed costs represent
capital expenditures that must be made before production of
emission control components can begin. As such, they are
relatively independent of production volumes. These costs
include systems engineering costs, certification costs and
facilities modifications. Variable costs represent the vendor
cost of the necessary emission control hardware. They are
directly dependent on production volume and are expressed on a
per vehicle basis. The fixed costs will be amortized and added
to variable costs to provide a total unit cost to the
manufacturer.
The basic source for this section is the response to
public comments document.[1] For additional detail on both
fixed and variable costs, the reader is referred to this
document.
A. Fixed Costs
As stated above, the fixed costs involved in onboard
control of refueling emissions include systems engineering and
certification costs, and the cost of test facilities
modifications. These costs can be expected to vary somewhat
between vehicle classes, however light duty vehicles (LDVs) and
light duty trucks (LDTs) have the same certification procedures
2-2
-------
and quite similar fuel systems and can therefore be grouped
together. Heavy-duty gasoline vehicles (HDGVs) have similar,
but larger, systems and different certification procedures.
Estimating costs for HDGVs requires some different analysis
than that needed for LDVs and LDTs.
1. Light Duty Vehicles and Light Duty Trucks
Systems engineering costs are those incurred in developing
an onboard system that is integrated with the other related
vehicle/engine systems. This includes incorporation of the
onboard system into other vehicle/engine systems (e.g. fuel
system, packaging) as well as consideration of safety and other
emission control requirements. In some cases this is a
straightforward engineering design problem, in others sucn as
vehicle safety or emissions control, it involves not only
design, but also follow-on testing and evaluation.
At this point EPA has little data on which to base a firm
estimate of the systems engineering costs. Costs involved
would generally include engineering design and development,
procurement and modification of prototype hardware and test
vehicles, and the actual testing and evaluation of the
systems. The above mentioned safety testing and evaluation
would also be included. EPA briefly examined these latter
costs and concluded that the cost for ensuring compliance with
the applicable fuel system safety provisions would be about
$34 000 per body configuration, or approximately $6.9 million
fleetwide.[8] Since the number of body configurations is
roughly equal to the projected number of refueling families,
this cost can easily be included in the total systems
engineering cost per family. Accordingly, in the response to
public comments document, EPA estimates a total systems
engineering cost for LDVs and LDTs (including safety costs) at
about $146,000 per refueling family (assumed to be similar in
number to the current evaporative emissions family).
Multiplying by the projected number of families (approximately
140 LDV and 65 LDT) results in an estimated cost of $20.4
million for LDVs and $9.5 million for LDTs. Amortizing these
costs for five years at 10 percent per year and dividing by
projected sales during the period provides an amortized cost
per vehicle of $.45 per LDV and $.69 per LDT, including $.12
per vehicle for fuel system safety compliance.
Certification costs include the costs of vehicle buildup,
mileage accumulation for LDVs and LDTs on durability and
emission data vehicles, and emissions testing for durability
and emission data vehicles. The EPA estimated cost per family
from the response to public comments document is $181,000 for
durability vehicles and $28,000 for emission data vehicles.
Although a total recertification would likely be necessary
because of the refueling regulations, the estimated 10 percent
of the fleet that normally undergoes certification each year
2-3
-------
regardless of new standards must be deducted to determine the
net cost of the refueling regulations. Multiplying the cost
per family by 90 percent of the projected LDV and LOT families
yields a certification cost of approximately $26 million for
LDVs and $11 million for LDTs. The amortized cost per vehicle
is $.61 for LDVs and $.77 for LDTs.
Systems engineering and certification efforts will require
some expansion of and/or improvement to manufacturers' testing
facilities in order to accommodate the additional testing
required. Comments submitted by the Motor Vehicle
Manufacturers Association (MVMA) contained an estimate of
$734,000 per manufacturer for additional equipment and testing
space.[9] Absent any other reasonable estimates of the
requirements, the response to public comments document uses
this figure to project the facilities modifications that would
be required. Although it is unlikely that all 35 LDV/LDT
manufacturers would incur this entire expenditure, it is
possible that some larger manufacturers would have to spend
even more than this amount. To be conservative, then, EPA will
assume that all manufacturers will make the above investment in
additional facilities, for a total industry cost of
approximately $26 million. This results in an amortized cost
per vehicle of approximately $.30.
2. Heavy-Duty Gasoline Vehicles
Due to the wide range of HDGV sizes and applications, it
is necessary to first categorize the HDGV fleet in terms of key
fuel/vapor system parameters (vehicle length, chassis design,
number/size of fuel tank(s), etc.) before manufacturer costs
can be determined. The most convenient breakdown of the HDGV
fleet is into weight classes based on gross vehicle weight
(GVW) rating. While the HDGV fleet is traditionally broken
down into seven GVW classes (Ilb-VIII) insufficient data was
available at the time of the response to public comments
documentU] to accurately describe the HDGV fleet in terms of
seven different groups. It was known, however, that about
ninety percent of all HDGVs fall into one of three weight
classes: Class lib (8501-10,000 Ibs GVW), Class VI
(19,501-26,000 Ibs GVW), and Class VII (26,001-33,000 Ibs
GVW) It was also known that about 75 percent of these
vehicles fall into Class lib. Therefore, a relatively
simplified analysis was performed which assumed that 75 percent
of all HDGVs can be classified as Class lib trucks, and all
remaining HDGVs can be classified as Class VI vehicles since
Class VII vehicles are quite similar to Class VI vehicles.
Finally, it should be noted that Class lib trucks are
essentially heavy LDTs and can benefit directly from the
transfer of LOT technology.
Systems engineering costs for the Class lib vehicles
should be the same as for LDTs, i.e., $.69/truck, since they
are all made by LOT manufacturers and can thus take advantage
2-4
-------
of the LDT work. In fact, some manufacturers may choose to
certify part or all of their Class lib trucks as LDTs.
Class VI HDGV refueling emissions are likely to present
additional control problems. The emission loads are likely to
be greater, due to the larger fuel tanks that are typically
found on these vehicles. Also, most of the heavier HDGVs
currently have open loop carbureted fuel systems, which would
have more difficulty with purge control than the closed-loop
fuel systems with electronic engine controls that are more
common on LDVs and LDTs and are expected on many Class lib
vehicles. Some additional purge control systems engineering is
thus likely to be required for these vehicles. Conversely, it
should be noted that the current HC and CO exhaust emission
standards for these vehicles are not as stringent as those that
apply to the Class lib HDGVs. In the response to public
comments document, EPA estimates a cost of $1.50 per vehicle to
cover development of these control systems. On a fleet
weighted basis, the amortized costs of all HDGV classes is
$0 89 per vehicle. Equivalent total systems engineering costs
can be obtained by summing the net present value of the $.69
Class lib cost and the $1.50 Class VI cost multiplied by _ the
respective projected sales for the two classes for the first
five years of the regulation. This would equate to a total
HDGV systems engineering cost to manufacturers of about $1.4
million.
Control of HDGV refueling emissions would entail
recertification to the exhaust and evaporative emission
standards in addition to refueling. Exhaust emission
recertification would be necessary due to the new requirement
that evaporative/refueling emission control canister(s) be
connected to the engine during testing. Evaporative emission
recertification would be necessary due to the test procedure
changes and the potential interdependence between the
evaporative and refueling control systems.
Exhaust emissions recertification is estimated to cost
$200,000 per family. This entails both durability assessment
and three emission data engines. For 1986, HDGV manufacturers
certified 8 families using 23 emission data engines. Total
exhaust emission recertification costs including durability
assessment are estimated at $1.6 million dollars.
Evaporative emission recertification costs are a bit more
difficult to estimate since abbreviated certification
procedures apply to Classes lib through VI HDGVs and
certification is by engineering evaluation for the heavier
HDGVs. Using the development costs presented in the original
HDGV RIA, certification costs are estimated at $31,000 per
evaporative family/system combination. In 1986, 23
family/system combinations were certified. Using these
figures, total evaporative emission certification costs are
estimated at $713,000.
2-5
-------
Even though abbreviated certification procedures have been
proposed for HDGV refueling, some costs would still be incurred
for durability assessment and demonstration of system
performance on emission data vehicles. However, these costs
would be incremental to those incurred during system
development and exhaust and evaporative emission
recertification. Durability assessment is estimated to cost a
nominal $50,000 per manufacturer over those costs incurred for
exhaust and evaporative recertification, and emission data
vehicle testing is estimated to cost $9000 per family/system
combination for testing and mileage accumulation. Assuming
three primary HDGV manufacturers and the same number of
refueling and evaporative family/system combinations (i.e. 23)
the total certification for refueling is estimated to be
$357,000.
Summing the estimated costs for exhaust, evaporative, and
refueling certification, the total industry cost is estimated
at $2.67 million dollars. Assuming these costs are incurred in
the year prior to the new standard and amortized over vehicles
sold in the 5 year period 1990-1994, the per vehicle
certification cost is $2.00.
The final fixed cost to manufacturers of HDGVs is for
modification of and improvements to test facilities. It is
anticipated that the requirements for Class lib vehicles will
be essentially the same as those required for LDTs; the same
facility may in fact serve both vehicle classes. Class VI
vehicles will also require new or modified facilities for
certification. Thus, while the additional facility
requirements may be minimal, the $.30 per vehicle estimated for
LDVs and LDTs is also extended to HDGVs. This provides an
additional $710,000 to cover any incidental modifications that
may be required for HDGV facilities.
3. Summary
In summary, the projected manufacturers' fixed costs for
LDVs, LDTs, and HDGVs are just under $98 million. These costs
are divided between the various vehicle classes for systems
engineering, certification and facilities modifications as
shown in Table 1. Table 1 also shows these fixed costs on an
amortized per-vehicle basis.
B. Variable Costs
Emission control hardware is the primary variable cost to
the manufacturers. Refueling control hardware costs for
individual components are relatively the same for any vehicle
class/subgroup, but total system costs vary somewhat depending
on the type of control system, the number of fuel tanks, and
the vehicle fuel tank capacity. A brief description of the
onboard control system and a summary of the individual
2-6
-------
Table 1
Fixed Costs to Manufacturers
(Millions of Dollars)
LDV LPT HDGV Total
Systems Engineering 20.4 9.5 1.4 31.3
Certification 26.0 11.0 2.7 39.7
Facility Mods 17.9 8.1 0.7 _?JL1
Total 64.3 28.6 4.8 97.7
Amortized Costs per Vehicle
(Dollars)
LDV LPT HDGV
Systems Engineering 0.45 0.69 0.89
Certification 0.61 0.77 2.00
Facility Mods 0.30 0.30 Q-30
Total 1-36 1.76 3.19
2-7
-------
component costs is given below. This is followed by a
discussion of the key factors which can cause the costs to vary
and a summary of the control system costs used in the analysis.
As described in the Technological Feasibility portion of
the gasoline marketing study, an onboard system is comprised of
a number of new components and modification of several existing
components. These are described briefly below:
1. Fillpipe Seal - Used to prevent gasoline vapor from
escaping to the atmosphere. Liquid or mechanical seal
approaches are possible. This analysis assumes all vehicles
will use liquid seals, except the large HDGVs (Class VI) which
may use mechanical seals. A pressure relief device may be
required for mechanical seals.
2. Fill Limiter - A device inside the tank used to close the
refueling vapor line when the tank is full, eventually causing
actuation of of automatic shut-off on the fuel nozzle.
3. Vapor Line Closure Valve - A valve used to close the vapor
line during normal vehicle operation. This is a key component
in rollover protection. Electronic or mechanical approaches
are possible.
4. Liquid/Vapor Separator - Required to decrease the emission
load to the canister and improve canister durability by
returning entrained liquid fuel droplets to the fuel tank.
5. Vapor and Purge Lines - Hoses to route refueling vapor to
the charcoal canister and purged vapor to the vehicle fuel
system. Net cost depends on canister location and whether the
system is integrated or separate from the evaporative control
systems.
6. Charcoal Canister - Serves as vapor storage device. Cost
and size vary with fuel tank size and whether the control
system is integrated or separate.
7. Packaging - Hardware or vehicle modifications to
accommodate canister and other hardware on the vehicle. Costs
vary depending on vehicle size and type. These costs are
amortized over five years.
8. Modifications - Minor modifications will be necessary to
the fuel tank and vehicle purge system. Fuel tank
modifications will be needed to accommodate the liquid seal and
fill limiter. Purge system modifications will be needed to
efficiently purge the canister while controlling exhaust
emissions. These costs are amortized over five years of
production.
2-8
-------
On a per vehicle basis, three other factors may affect the
total cost of the refueling emission control hardware. These
include the basic system approach, the number of fuel tanks,
and the capacity of the fuel tanks.
Three basic control systems approaches may be used;
integrated, partially integrated, and separate. The integrated
system uses one canister to control both evaporative and
refueling emissions. The partially integrated system collects
diurnal evaporative and refueling emissions in one canister and
uses a separate canister for capturing hot soak emissions. The
third approach, the completely separate system, captures all
evaporative emissions in one canister and all refueling
emissions in another.
EPA expects that all fuel injected vehicles will utilize
fully integrated systems due to the cost advantage and the low
level of non-fuel tank hot soak emissions. Carbureted vehicles
may use partially integrated or separate systems due to the
need to deal with hot soak evaporative emissions. However, EPA
expects most manufacturers will opt for partially integrated
systems to take advantage of the available cost savings.
By 1990, when these regulations are assumed to take
effect, current projections are that approximately 88 percent
of all LDVs and LDTs are expected to be fuel injected and 12
percent carbureted. Class lib HDGV fuel systems are likely to
follow the same split as LDTs. The majority of the heavier
HDGVs (Class VI) are assumed to remain predominantly
carbureted. Thus 88 percent of LDVs, LDTs and Class lib HDGVs
are expected to use integrated systems and the remaining 12
percent of these vehicles classes are expected to use partially
integrated systems. Essentially all of the heavier HDGVs are
assumed to use either partially integrated or separate systems.
On a per vehicle basis, onboard hardware costs also depend
on whether a vehicle has single or dual fuel tanks. Hardware
costs for dual fuel tank vehicles are essentially twice those
of single tanks because each fuel system requires the necessary
control hardware. Dual tank vehicles could use any of the
three control system approaches discussed above, and EPA
assumes that the mix of these control approaches will be the
same as described above for both single and dual tank
vehicles. The fractions of dual tank vehicles used in this
analysis are shown below. These essentially reflect current
conditions.
LDV 0%
LOT 20%
HDGV:
Class lib 20%
Class VI 15%
2-9
-------
While most individual onboard hardware component costs
will not vary among the vehicle models, the cost of the
charcoal canister for any given vehicle model will vary
depending on the capacity of the fuel tank. The cost of a
refueling emissions control canister is a function of the
required control system capacity which in turn depends upon the
size of the fuel tank. EPA estimated fleet average fuel tank
sizes for the three vehicle classes/subgroups based on a
minimum driving range and the fuel economy improvements that
are expected to occur between 1990 and 2000. The projected
fuel tank capacities used in calculating canister costs for the
three vehicle classes/subgroups are shown in Table 2.
Based on this information regarding refueling control
hardware and those factors which influence total per vehicle
cost. Tables 3 and 4 present onboard costs for LDVs, LDTs and
HDGVs. These are manufacturer (vendor level) costs and do not
reflect manufacturer or dealer markups for overhead and
profit. For sake of completeness, the amortized fixed costs
discussed previously are also listed. It should be noted that
the costs presented in Tables 3 and 4 are incremental to the
cost of current evaporative emissions systems.
As can be seen from Tables 3 and 4, costs vary with time
to reflect both the different amortization periods for fixed
costs and changing canister sizes due to projected fuel economy
improvements. Three distinct cost periods can be identified
for LDVs, and LDTs, and HDGVs: 1990-94, 1995-99, and 2000 and
beyond. For the sake of brevity, only the total costs are
shown for the years after 1994 for all vehicle classes. The
cost differences are based on the following assumptions. For
all vehicles, systems engineering and certification costs are
assumed to be fully amortized during the first 5 years, i.e. by
1994, and are not reflected thereafter. Vehicle modifications
(i.e. tank and purge system modifications and packaging) costs
would also disappear after 5 years for LDVs, LDTs and Class lib
HDGVs, as car and truck designs begin to incorporate onboard
controls from the initial design inception. Class VI HDGVs are
assumed to have no packaging costs and so these are not
reflected in the 1990-94 initial cost estimates. Tank and
purge system modification costs would also drop out after five
years for HDGVs, just as they do for LDVs and LDTs. For LDVs,
LDTs, and HDGVs, facility modifications represent a longer-term
investment and are assumed to be fully amortized after 10
years. Therefore they are not included after 1999.
Projected fuel economy improvements have been averaged for
each of the three periods to determine canister size
requirements for LDVs and LDTs. These are also reflected in
the decreasing costs for each period. Fuel economy differences
are not reflected in long-term HDGV costs, since the projected
improvements are slight and have a minimal effect on system
costs.
2-10
-------
Table 2
Projected Fuel Tank Capacity By Vehicle Class
(Gallons)
Vehicle Class
LDV
LOT
Single Tank
Dual Tank
1990-94
12
16
33
5 Year Period
1995-1999
11
16
32
2000 +
10
15
29
HDGV
lib-Single
lib-Dual
Vl-Single
VI-Dual
20
40
30
80
20
40
30
80
20
40
30
80
2-11
-------
Table 3
Manufacturer Costs of Onboard Emission Control Components
Light Duty Vehicles and Light Duty Trucks
LDV
LPT - Single Tank
LOT - Dual Tank
to
I
M
NJ
Category
1990-94
Hardware
Veh. Mods
Subtotal
Amortized
Fixed Costs
Total
Class
Weighted Avg.
1990-94
1995-99
2000 +
Integrated
13.24
1.07
14.31
1.36
15.67
LDV
16.01
14.00
12.87
Part Int/Sep
15.17
1.97
17.14
1.36
18.50
LDT-ST
17.75
15.51
14.29
Integrated
14.58
1.07
15.65
1.76
17.41
LDT-DT
37.23
33.82
32.08
Part Int/Sep
16.54
1.97
18.51
1.76
20.27
All LOT
21.65
19.17
17.85
Integrated Part Int>
32.39 34.21
2.39 3.44
34.78 37.65
2.11 2.11
36.89 39.76
-------
Table 4
Manufacturer Costs of Onboard Emission Control Components
Heavy Duty Gasoline Vehicles
CL. lib HDGV
CL. VI HDGV
Category Single Tank
1990-94
Hardware
Veh. Mods
Subtotal
Amortized
Fixed Costs
15
1
16
2
.66
.25
.91
.99
Dual
35
2
37
2
Tank
.43
.00
.43
.99
Single Tank
28
0
29
3
.91
.75
.66
.80
Dual
62.
1.
63.
3.
Tank
18
50
68
80
Total
19.90
40.42
33.46
67.48
Class
Weighted Avg:
1990-94
1995-99
2000 +
lib
24.00
19.91
19.51
VI
38.56
34.16
33.86
All HDGV
27.64
23.47
23.17
2-13
-------
As was mentioned previously, properly designed and tested
onboard control systems also have the capacity to control the
excess evaporative emissions caused by the higher volatility of
commercial fuels. However, these excess evaporative emissions
could be controlled by increasing the RVP of the fuel used in
exhaust and evaporative emissions certification testing to the
in-use level of 11.5 psi, in which case manufacturers would be
forced to improve their current emission control systems. Thus
it is appropriate to evaluate onboard refueling control costs
incremental to excess evaporative emission control costs.
Calculating the costs of onboard controls incremental to
improved evaporative systems requires the subtraction of the
incremental costs of the improved evaporative systems from the
onboard costs presented in Tables 3 and 4. This is done in
Table 5 for LDVS, LDTs, HDGVs. The cost of these improved
systems is taken from EPA's recent study of gasoline volatility
and vehicle HC emissions.[ 10] For the sake of brevity only a
single cost is shown for each vehicle class. This cost is a
weighted average based on the distribution of configurations
discussed earlier.
Ill. Costs to the Consumer
The cost of the proposed regulations to the consumer can
be divided into two general categories: first price increase
and operating costs. The first price increase consists of
hardware costs, amortized manufacturer fixed costs and the
retail price equivalent (RPE) markups. The RPE reflects the
various overhead and profit markups that are added to the
vendor cost of a the onboard system at the manufacturer and
dealer levels. The second category, operating costs, consists
of maintenance and in-use inspection costs, if any, plus any
change in total lifetime fuel costs due to increased or
decreased fuel consumption rates. Each of these is discussed
below, followed by an estimate of the net lifetime consumer
cost.
A. First Price Increase
The total first price increase to the consumer consists of
the manufacturer's cost per vehicle developed in the previous
section, including amortized fixed costs, multiplied by an RPE
markup factor of 1.26 for LDVs and LDTs and 1.27 for HDGVs.
These markup factors represent industry averages and were
developed by an EPA contractor using available financial data
for the domestic auto and truck manufacturers and their
dealers. [4] Data covering a ten year time span were used in
the determination to dampen the effects of the business cycle.
The markup factor includes corporate overhead and profit plus
dealer profit and other expenses for interest and sales
commissions. No dealer overhead is included, since the
addition of emission control components should not increase the
dealer's cost for storing or selling his vehicles. Hardware
2-14
-------
Table 5
Onboard Hardware Costs to Manufacturers
(Incremental to Costs for Control of
Excess Evaporative Emissions)
LDV LPT HDGV
1990-94 Hardware cost' $16.01 $21.65 $27.64
Less: excess evap. cost -2.25 -2.98 -3.28
Incremental cost2 13.76 18.67 24.36
1995-99 Hardware cost1 14.00 19.17 23.47
Less: excess evap cost -1.52 -1.99 -2 . 70
Incremental cost2 12.48 17.18 20.77
2000 + Hardware cost' 12.87 17.85 23.17
Less: excess evap cost -1.52 -1. 99 -2 . 70
Incremental cost2 11.35 15.36 20.47
Incremental to cost of current evaporative control systems.
Incremental to cost of improved evaporative control systems
capable of controlling excess evaporative emissions,
2-15
-------
costs and amortized costs for certification and facilities
modifications are marked up using these factors. However,
using the markup methodology developed by Lindgren for EPA,
systems engineering costs are essentially R&D and are not
subject to RPE markups, but rather are added to the total cost
after the RPE markup has been applied. Table 6 shows the first
price increase to the consumer when onboard costs are evaluated
incremental to excess evaporative emission control costs.
B. Operating Costs
As discussed in the response to public comments document,
an onboard system should not require any additional maintenance
over the current evaporative emission control system. Also, no
in-use inspections of onboard systems are likely, since
evaporative systems are generally not inspected in current
programs and tampering rates are relatively low. Therefore, no
increases in operating costs related to onboard system
maintenance or in-use inspections are expected.
However, some change in lifetime operating costs are
expected as a result of the positive fuel consumption benefit
realized from the fuel recovery credits. Fuel recovery credits
accompany both the recovery of refueling and excess evaporative
emissions, but this analysis will address only the refueling
emissions which are captured by the onboard system. As was
mentioned previously, a properly designed and tested onboard
system would also have the capacity to control excess
evaporative emissions, but alternative strategies for
controlling excess evaporative emissions may reduce or totally
eliminate the excess evaporative credit.
The lifetime change in fuel consumption resulting from the
recovery of refueling emissions actually is the difference of
two effects: the gross recovery credit less the effect of the
weight penalty associated with the onboard system. The gross
recovery credit is determined by calculating the gasoline
equivalent of the total mass of refueling vapors captured, then
multiplying this figure by the presumed vapor combustion
efficiency. This recovery credit must then be reduced by the
fuel economy penalty resulting from the slight weight increase
associated with the onboard system. The weight penalty was
determined using the estimated weight of an onboard system
together with a weight/fuel consumption sensitivity factor.
Using this approach, the net change in fuel consumption in
each year of the vehicle life was determined and then
multiplied by the value of a gallon of gasoline ($0.98) to get
the monetary value of the recovery credit. These values were
then discounted at 10 percent to the first year of the
vehicle's life to get a monetary value in the same terms as the
RPE.
2-16
-------
Table 6
i
Costs to the Consumer
Onboard Refueling Emissions Control
Costs: 1990-
First Price
94:
Increase:
Refueling FPI
Less :
Refueling
Net cost to
Costs: 1995-
Credits
consumer
1999:
Refueling FPI
Less:
Refueling
Net cost to
Costs: 2000
Credits
consumer
& Beyond:
Total Refueling FPI
Less :
Refueling
Net cost to
Credits
consumer
LDV
17
(4
$12
15
(3
$11
14
(3
$10
.17
.24)
.93
.71
.85)
.86
.29
.53)
.76
LOT
18
(6
$11
17
(6
$10
15
(5
$9
.39
.50)
.89
.01
.31)
.70
.47
.93)
.54
Dual Tank
LOT
42
(3
$39
40
(3
$36
37
(3
$34
.82
.35)
.47
.08
.22)
.86
.89
.09)
.80
Class lib
HDGV
26
(11
$14
21
(11
$10
21
(11
$10
.02
.70)
.32
.86
.20)
.66
.48
.00)
.48
Class
VI HDGV
44
(23
$21
39
(22
$17
39
(22
$16
.29
.10)
. 19
.96
.70)
.26
.58
.60)
.98
Minor differences between the costs presented in this
table and those presented in the response to public
comments document are due to rounding. Costs are
incremental to excess evaporative emissions control costs.
2-17
-------
This analysis was conducted for the five vehicle
classes/subgroups being evaluated here for the three time
periods mentioned earlier for the hardware, ie. 1990-1994,
1995-1999, and 2000 and beyond. The results of the analysis
are shown in Table 6.
C. Net Consumer Costs
The net lifetime consumer costs for onboard control are
shown in Table 6. In this table the costs associated with the
first price increase are partially offset by the net present
value of the recovery credits. Net costs are shown for the
five vehicle class/subgroups and the three time periods
mentioned previously.
IV. Aggregate Costs to the Nation
Aggregate costs to the nation include the fixed costs to
manufacturers for systems engineering, certification and
facility modifications, plus the cost to the consumer for
hardware, less the fuel recovery credits. In calculating these
costs, the net amounts from Table 6, adjusted to remove the
amortized fixed costs, were multiplied by the projected sales
shown in Table 7. The 1988-94 aggregate costs are shown in
Table 8 (undiscounted) and Table 9 (discounted) by the year
they are expected to be incurred. A total discounted cost is
also shown for each vehicle class, representing a lump sum
payment as of the year the standards are effective. Costs are
shown for LDVs, LDTs, HDGVs and a total for all three classes.
All costs before the year the standard is assumed to take
effect (1990) are for systems engineering, certification, and
facility modifications; all costs for 1990 and subsequent years
are for hardware costs, less fuel recovery credits. It is
assumed that all systems engineering and facility modifications
costs will be incurred in 1988 while all certification costs
will be incurred in 1989. All of the costs shown in Table 9
are discounted at 10 percent per year to 1990, the first model
year of the standard. As shown in the table, the total
discounted cost to the nation for the proposed refueling
regulations (in 1986 dollars) is about $892 million.
IV. Socioeconomic Impact
The socioeconomic effects of the proposed regulations
include any impact they may have on manufacturers* cash flow,
sales and employment, energy usage, balance of trade, and on
particular groups of individuals or segments of the economy.
The overall impact of the proposed regulations on
manufacturers' cash flow is expected to be minor. Hardware
costs are recouped by the manufacturers during the year they
are incurred. With current computerized inventory control
systems, manufacturers are not expected to be required to
maintain large, expensive parts inventories during the course
2-18
-------
Table 7
Projected Gasoline-fueled Vehicle Sales
(Millions of Vehicles)
1990-2000
HDGVZ
0.384
0.381
0.379
0.381
0.383
0.384
0.385
0.388
0.392
0.396
0.400
Year
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000 +
1 Based
Fall,
LDV1
11.2
11.2
11.2
11.1
11.0
10.8
10.9
11.0
11.0
11.0
11.0
on DRI "Trendlong" Proji
1984 Long Term Review,
LDT1
3.71
3.68
3.64
3.56
3.47
3.35
3 .40
3.43
3.45
3.46
3.48
ectio
Data
See Reference 11.
2-19
-------
Table 8
Aggregate Incremental Costs
of Onboard Refueling Controls
(Millions of Dollars1)
Undiscounted
Calendar Year LDGVs
1988
1989
1990
1991
1992
1993
1994
38
26
126
126
126
125
124
.30
.00
.90
.90
.90
.75
.65
LDGTs
17
11
57
56
55
54
53
.60
.00
.00
.55
.95
.70
.35
HDGVs
2.
2.
4 .
4 .
4.
4.
4 .
10
70
70
65
60
65
70
Total
58
39
188
188
187
185
182
Costs
.00
.70
.60
. 10
.45
.10
.70
Rounded to the nearest 0.05 million.
2-20
-------
Table 9
Aggregate Incremental Costs
of Onboard Refueling Controls
(Millions of Dollars1)
Discounted 1990
Calendar Year LDGVs
1988
1989
1990
1991
1992
1993
1994
46
28
126
115
104
94
85
.35
.60
.90
.35
.85
.50
.10
LDGTS
21
12
57
51
46
41
36
.30
. 10
.00
.40
.25
.10
.45
HDGVs
2
3
4
4
3
3
3
. 55
.00
.70
.20
.80
.50
.20
Total
70
43
188
170
154
139
124
Costs
.20
.70
.60
.95
.90
.10
.75
Total
601.65
265.60
24.95
892.20
1 Rounded to the nearest 0.05 million.
2-21
-------
of the manufacturing year, so the only significant negative
impact on manufacturers' cash flow would come from the fixed
costs, i.e. systems engineering, certification and facilities
modifications. Although these costs would eventually be
recovered, they would likely be incurred during the two years
prior to the effective model year of the regulations, when
there would be no directly offsetting revenues received.
Assuming that all of the fixed costs would be incurred in the
two years prior to the effective date of the standards, the
average cost per year would be less than $50 million. This
represents less than one-half of one percent of the cash outlay
made by the domestic industry during 1980 and 1981 for
modernization and downsizing.[11] .
The effect of these regulations on sales and employment is
also expected to be minimal. The first price increase to
consumers is only 0.1 to 0.2 percent of the average cost of a
new vehicle for LDVs, LDTs and HDGVs. By contrast, the annual
price increases for new vehicles have been many times these
percentages in recent years, even in times of very low
inflation. The impact of such increases on consumer demand for
vehicles is commonly expressed in terms of the price elasticity
of demand, ie. the reduction in sales corresponding to an
increase in price. EPA has determined a price elasticity of
demand of -1.0 for LDVs & LDTs and -0.9 to -0.5 (average of
-0.7) for HDGVs.[1,12] The resulting decrease in demand would
thus be -.1 to -.2 percent for LDVs, LDTs and HDGVs. However,
a real question arises as to whether it is appropriate to apply
this price/demand model to such small price increases. Thus,
although it is conceivable that consumer demand, and
consequently sales, could be affected slightly, it is highly
unlikely and not measurable. Accordingly, the impact on
employment in the auto industry should also be minimal. In
fact, some additional employment in the manufacture of onboard
control hardware is likely. This employment would occur in the
traditional auto industry but would also extend to suppliers of
activated carbon, vapor lines, and other components.
Onboard controls could result in a slight decrease in the
demand for gasoline, due to the fuel recovery credits.
However, the savings involved are small in comparison to the
projected gains in fuel economy during the period in question
without the credits, or to the normal fluctuations in consumer
demand due to other factors. A slight reduction in demand for
gasoline could in turn result in a slight decrease in the
demand for imported oil and perhaps an improvement in the
nation's balance of trade, but this would likely be offset by
the price increase on imported vehicles. Any effect on
employment in the petroleum industry would also be negligible.
In conclusion, the overall socioeconomic impact of these
regulations is expected to be minimal. Barring a significant
recession in the late 1980'S auto and truck manufacturers are
2-22
-------
expected to have no trouble underwriting the capital investment
required. Vehicles sales should not be impacted measurably,
and employment in the automotive and related industries may
increase slightly. These regulations are not expected to have
any net effect on the balance of trade.
2-23
-------
REFERENCES FOR CHAPTER 2.0
1. "Evaluation of Air Pollution Regulatory Strategies
for Gasoline Marketing Industry - Response to Public Comments,"
U.S. EPA, OAR, July, 1987.
2. "Cost Estimations for Emission Control-Related
Components/Systems and Cost Methodology Description," Leroy H.
Lindgren, Rath and Strong for U.S. EPA, 1978.
3. Draft "Manufacturing Cost and Retail price
Equivalent of Onboard Vapor Recovery System for Gasoline
Filling Vapors," Leroy H. Lindgren for API, 1983.
4. "Update of EPA's Motor Vehicle Emission Control
Equipment Retail Price Equivalent (RPE) Calculation Formula",
Jack Faucett Associates for U. S. EPA, 1985.
5. "1977 Census of Transportation, Truck Inventory and
Use Survey", U.S. Department of Commerce, Bureau of the Census,
1980.
6. "Light Duty Automotive Fuel Economy...Trends Thru
1985", Murrell et. al., SAE Technical Paper Series #850550,
1985.
7. "U.S. Long Term Review, Fall 1984", Data Resources,
Inc., 1984.
8. Memorandum "Cost of Crash Testing to Assure Fuel
System Integrity for Onboard Systems," Johnson to the Record,
U.S. EPA, OAR, QMS, August, 1986.
9. Comments of the Motor Vehicles Manufacturers
Association of the U.S., Inc. on EPA Report "Evaluation of Air
Pollution Regulatory Strategies for Gasoline Marketing
Industry," MVMA, Public Docket A-84-07, I-H-127, 1984.
10. "Study of Gasoline Volatility and Hydrocarbon
Emissions from Motor Vehicles," U. S. EPA, OAR, QMS, 1935.
11. "Regulatory Support Document, Revised Gaseous
Emission Regulations for 1985 and Later Model Year HD Engines,"
U. S. EPA, OAR, QMS, 1983.
12. "Draft Regulatory Impact Analysis and Oxides of
Nitrogen Pollutant-Specific Study," U. S. EPA, OAR, QMS, 1984.
2-24
-------
3.0 ENVIRONMENTAL IMPACT: Ozone Air Quality
I. Introduction
This chapter evaluates the effects on ambient air quality
associated with implementing a 0.10 gram per gallon (9/9*1)
refueling emission standard for light-duty vehicles, light-duty
trucks, and heavy-duty vehicles that use gasoline fuel. The
proposed refueling standard is assumed to take effect with the
1990 model year. The following section serves as the
background for the remainder of the chapter. It includes a
brief characterization of the emissions associated with the
refueling process, a discussion of those emissions which are
potentially controllable with onboard control technology, and
the interaction between refueling emission controls and
concurrent reductions in excess evaporative emissions from
motor vehicles. The last section evaluates the relative effect
of controlling refueling and excess evaporative emissions on
ambient ozone levels in selected urban areas currently in
violation of the National Ambient Air Quality Standard (NAAQS)
for this pollutant. The actual adverse effects of ozone on_man
and the environment will not be described in this section,
since these effects have been fully documented elsewhere.11,2J
In general, this chapter has a relatively narrow focus.
The ambient air guality analyses contained herein have a much
more limited range than the emission inventory analyses
presented in Volume 1 of the draft Regulatory Impact Analysis
(draft RIA). The analyses of this chapter examine only those
non-California urban areas that were in non-attainment status
for the ozone standard on the basis of 1983 design values. A
detailed evaluation of various national emission reduction
scenarios is contained in the first volume of the draft RIA,
and is not repeated here. Also, this chapter does not address
the carcinogenic potential of gasoline vapors. Again, this
topic is discussed in the first volume of the draft RIA.
Readers wishing to obtain further information on the
aforementioned subjects should see the appropriate sections of
this document.
It is important to note that the following analysis was
done to estimate trends in ozone attainment for the group of
urban areas in the sample. It uses VOC emission inventory data
from EPA's National Emission Data System and not city-specific
inventory data. In addition, an up-to date assessment of the
implementation of stationary source regulations for each area
was not made. Therefore, the analysis should be viewed as
indicating probable changes in the magnitude of the ozone
nonattainment problem and not as a prediction of
attainment/nonattainment for specific urban areas.
-------
II. General Description of the Emissions Associated with
Refueling and the Use of Onboard Controls
A. Refueling Emissions
Gasoline vapors are composed of various hydrocarbon (HC)
compounds. During the vehicle refueling process, these vapors
are emitted from a variety of sources. The majority of the
refueling emissions (i.e., about 90 percent) are HC vapors
displaced from the vehicle's fuel tank by the incoming
gasoline. The mass of vapor which escapes per unit of gasoline
dispensed, or the emission factor, is dependent on several
variables, including: (1) the temperature of the dispensed
fuel, (2) the difference between the temperature of the
dispensed fuel and the fuel tank, and (3) the volatility of the
fuel.[3] These highly concentrated vapors are emitted from the
vehicle's fillneck directly into the breathing zone of the
person performing the refueling operation.
In addition to displacement losses from the vehicle tank,
there are two secondary sources of emissions associated with
refueling. The first is spillage, due to "splash back" from
the fill pipe or the escape of some liquid from the dispensing
nozzle when withdrawn from the fill pipe. The second is the
escape of vapor from the vent of the service station's
underground tank following refueling. As fuel is pumped into
the vehicle's fuel tank, ambient air is drawn into the service
station's tank through its tank vent. This "fresh" air causes
fuel in the tank to evaporate until an equilibrium
concentration is reestablished between the vapor and the
liquid. As this process takes place, the total volume of the
ingested air in the underground tank increases somewhat, and
the excess volume is emitted from the vent in the form of HC
emissions. These latter two sources each account for about 5
percent of the total emissions associated with the refueling
process, as described in the first volume of the draft RIA.
The total refueling emission factor is expressed as the
mass of HC emitted per gallon of dispensed fuel (g/gal), and
varies from region to region due to differences in temperatures
and fuel volatility. As an example, using national average
temperature values and national average estimates of future
volatility levels based on historic trends, the total emission
factor for all three sources is 6.6 g/gal.[3]
B. Emissions Potentially Controllable by Onboard
Technology
Onboard controls function during a refueling operation by
sealing the vehicle's fillneck and then routing the displaced
vapors to a storage canister, where the HC molecules are
adsorbed onto the surface of activated charcoal. When the
vehicle's engine is started, fresh air is drawn through the
3-2
-------
canister to purge the HC molecules from the charcoal. The
resulting vapor is transferred to the fuel metering system and
subsequently burned in the engine. Tests have shown that
properly designed and operating onboard systems are capable of
controlling about 99 percent of the displacement vapors.[4]
(The control effectiveness value was revised upward from the 97
percent used in the gas marketing study based on more recent
test data.) Of those emissions not captured, nearly all escape
from the fillneck (i.e., 95 percent). The remainder (i.e., 5
percent appear to be emitted from the canister, which typically
is located in the engine compartment.
Onboard control technology will also have the effect of
eliminating one of the two secondary sources of emissions
associated with refueling mentioned previously, that of splash
back." These emissions, which result from the spillage and
subsequent evaporation of liquid gasoline, should be
significantly reduced through the design of the proposed test
procedure. The primary cause of fuel spillage is dispensing of
the fuel at too rapid a rate, relative to what the vehicle
fillneck is able to accept. The proposed test procedure
specifies a maximum fuel dispensing rate of 10 gallons per
minute. The draft regulations accompanying today's proposal
specify this same maximum fuel flow rate as an in-use standard,
aS well as being part of the test procedure. Since, to be able
to meet the proposed standard, no fuel spillage can be
tolerated during certification testing of the onboard system,
EPA believes that spillage due to high fueling rate will &e
eliminated through fuel flow rate restrictions and standardized
design of fuel delivery nozzles. Of course, spillage from such
causes as nozzle malfunctions will not be affected by the new
procedures. The other secondary source of refueling-related
emissions, emptying losses from service stations' underground
storage tanks, will not be affected by onboard control
technology.
The overall in-use efficiency of capturing displacement
vapors will be somewhat less than 99 percent, due to the
effects of tampering on some vehicles. Tampering is primarily
expected to take the form of removal or disconnection of the
canister. In such instances, the displacement losses revert to
the uncontrolled level, but are re-oriented spatially. with
tampered systems, the majority of the vapors would be emitted
at the location of the missing or disconnected canister, rather
than from the fillneck into the breathing zone of the person
dispensing the fuel.
An issue that is intimately associated with the
implementation of an onboard refueling standard is the
interaction between the requisite control technology and the
elimination of excess evaporative emissions from motor
vehicles. Evaporative emissions are primarily a combination of
breathing losses from the vehicle's fuel tank, due to diurnal
3-3
-------
temperature changes, and losses from the carburetor bowl, fuel
lines, and fuel tank that occur as the result of residual heat
from the engine and exhaust system after the engine is turned
off. Presently, gasoline-fueled vehicles must be certified to
certain evaporative emission standards prior to mass production
and sale. The control technology used to reduce evaporative
emissions to the required levels is similar in design and
function to that described above for onboard refueling hardware
(i.e., carbon canisters). However, data from EPA's emission
factor program indicate that many "in-use vehicles fail to
comply with the applicable standards. The principle cause of
these failures is that in-use fuel typically has a higher
volatility than the fuel specified for certification testing.
The increased amounts of evaporative HC caused by higher in-use
volatilities cannot be adsorbed by current charcoal canister
systems. These additional HC vapors escape into the atmosphere
as excess evaporative emissions.
The Agency is presently considering a variety of
strategies aimed at assuring in-use compliance with evaporative
emission standards. While no decision has been made, one
alternative is to make relatively simple changes to the current
evaporative emission test procedures and to increase the
volatility of certification fuel to be more representative of
in-use levels (i.e., certification fuel volatility equals
in-use fuel volatility). These revisions basically would
result in larger canisters and different purge rates. Since
onboard technology includes a larger charcoal canister than is
necessary to control evaporative emissions, even assuming the
revision of test fuel volatility, and because refueling
normally would not coincide with the occurrence of evaporative
emissions, it is possible to control excess evaporative losses
by integrating the requisite controls with the refueling
control system at little additional cost. In order to
illustrate the effects of such a program, the analyses
presented in the remainder of this chapter will include a
scenario that combines the control of refueling emissions plus
excess evaporative emissions, in addition to a scenario that
evaluates only the control of refueling emissions.
Ill. Air Quality Analysis
A. Selection of Areas for Modeling
The NAAQS for ozone requires that the fourth highest daily
maximum one-hour measured concentration not exceed 0.12 parts
per million (ppm) in any three-year period. On the basis of
ozone air quality monitoring data collected in 1982, 1983, and
1984 (or 1983 design values), which were the most recent
complete data available, there are 73 urban areas with measured
air quality above this standard.[5] Twelve of these 73
non-attainment areas were located in California. As already
discussed, Stage II vapor recovery systems are currently being
3-4
-------
used in nearly all of California, and it is unclear whether
these systems will continue to be used or if they will be
discontinued in favor of onboard controls. Because of this
uncertainty, only the 61 non-California urban areas listed in
Table 1 are included in the air quality analysis. This
simplifies the analysis without affecting _the resulting
conclusions. (The specific design values contained in Table 1
are further discussed in Section III.B.6.)
B. Air Quality Methodology
Future ambient ozone concentrations in specific urban
areas were estimated using the Empirical Kinetic Modeling
Approach (EKMA). This model utilizes a series of ozone
isopleths depicting downward maximum ozone concentrations as an
explicit function of initial non-methane hydrocarbon (NMHC) and
oxides of nitrogen (NOx) concentrations and as implicit
functions of a number of emissions and meteorological
characteristics. Differences between initial and subsequent
emissions as well as changes in concentrations of pollutants
transported into the modeled area are then simulated to
estimate changes in maximum ozone. It should be noted that
EKMA, as used by EPA in this analysis, is a nationwide-average
model. In other words, the only city-specific information used
as input data for the model are the base-year ambient ozone
concentrations (design values) and the emission inventories
from which future concentrations are projected, and the
NMHC:NOx ratios. Meteorological conditions and other input
data are held constant for all the urban areas modeled.
References 6 and 7 contain additional information on the use of
EKMA.
The initial, or base-year, emission inventories used in
the model are based on the information contained in EPA's
National Emissions Data System (NEDS), which is compiled by the
Agency's Office of Air Quality Planning and Standards (OAQPS).
The NEDS inventory includes emissions from both stationary and
mobile sources. The most recent emissions compilation at the
time of this analysis was for calendar year 1983, and is
referred to as the 1983 base-year inventory. Stationary source
inventories for any future year are computed by the model using
anticipated growth and retirement rates, along with estimates
of emission control efficiency. Similarly, future inventories
are constructed for mobile sources using the emission factor
ratios (base-to-future years) and annual compound VMT growth
rates for each vehicle class, derived from EPA's MOBILES
Emission Factor Model and MOBILE3 Fuel Consumption Model,
respectively.
3-5
-------
Table l
61 Non-California Urban Ozone Non-Attainment
Areas and Associated Design Values (ppm)
82-84 EKMA
Area pvs pys
EPA Region l
Boston Metropolitan Area 0.19 0.18
Greater Metropolitan Connecticut 0.23 0.18
New Bedford, MA 0.19 o'. 19
Portland, ME 0.15 o!io
Portsmouth-Dover-Rochester, NH-ME 0.13 0.09
Providence, RI 0.16 0.15
Springfield, MA 0.19 0.17
Worcester, MA 0.14 o!l2
EPA Region 2
Atlantic City, NJ 0.19 0.15
New York Metropolitan Area 0.23 0.24
Vineland-Millville-Bridgeton, NJ 0.14 0.14
EPA Region 3
Allentown-Bethlehem, PA 0.15 0.14
Baltimore, MD 0.17 0.17
Erie, PA 0.13 o!13
Harrisburg-Lebanon-Carlisle, PA 0.13 0.13
Lancaster, PA 0.14 0.09
Philadelphia Metropolitan Area 0.18 0.20
Pittsburgh, PA 0.14 0.14
Reading, PA 0.13 0'. 13
Richmond-Petersburg, VA 0.14 0.14
Scranton-Wilkes Barre, PA 0.13 0.13
Washington, DC-MD-VA 0.16 0.17
York, PA 0.13 o! 13
EPA Region 4
Atlanta, GA 0.17 0.17
Birmingham, AL 0.15 0.15
Charlotte-Gastonia-Rock Hill, NC-SC 0.13 0.13
Chattanooga, TN-GA 0.13 0.13
Huntington-Ashland, WV-KY-OH 0.14 0.14
Louisville, KY-IN 0.15 0.15
Memphis, TN-AR-MS 0.13 0.13
Miami-Hialeah, FL 0.13 0.13
Nashville, TN 0.13 0.13
Tampa-St. Petersburg-Clearwater, FL 0.13 0.13
3-6
-------
Table 1 (cont'd)
82-84 EKMA
Area DVS pys
EPA Region 5
Akron, OH 0.13 0.13
Canton, OH 0.13 0.13
Chicago Metropolitan Area 0.20 0.25
Cincinnati Metropolitan Area 0.15 o!l7
Cleveland, OH 0.14 0.14
Dayton-Springfield, OH 0.13 0.13
Detroit, MI 0.14 0.14
Grand Rapids, MI 0.13 o'. 13
Indianapolis, IN 0.13 o'. 13
Milwaukee Metropolitan Area 0.17 0.17
Muskegon, MI 0.14 0.14
EPA Region 6
Baton Rouge, LA 0.17 0.17
Beaumont-Port Arthur, TX 0.21 0.21
Brazoria, TX 0.14 0!14
Dallas-Fort Worth Metropolitan Area 0.16 0 16
El Paso, TX 0.17 o!17
Galveston-Texas City, TX 0.17 o'l7
Houston, TX 0.25 0^25
Lake Charles, LA 0.15 0.15
Longview-Marshall, TX 0.15 o!l5
New Orleans, LA 0.15 0.15
San Antonio, TX 0.14 0 14
Tulsa, OK o!l3 o!l3
EPA Region 7
Kansas City, MO-KS 0.14 0.14
St. Louis, MO-IL 0.17 0.17
EPA Region 8
Denver Metropolitan Area 0.14 0 14
Salt Lake City-Ogden, UT O.'lS 0.15
EPA Region 9
Phoenix, AZ 0.15 0.15
3-7
-------
The specific stationary source estimates and assumptions
used in the model for this analysis are the same as those that
were used in EPA's report on in-use fuel volatility and
evaporative emissions (hereafter referred to as the volatility
study).[8] Because these model inputs are fully described in
that document, they are not repeated here. The volatility
study also evaluated baseline and excess evaporative control
scenarios that are identical to those considered in this
chapter. Rather than repeat the detailed discussion of the
mobile source inputs for these scenarios, only a few of the key
features of that analysis, along with the basic inputs used to
model the onboard control scenario evaluated here, are
described below.
1. Fuel Volatility
Under the baseline scenario, the Reid vapor pressure (RVP)
of in-use fuel is the maximum summertime value (11.5 pounds per
square inch) recommended by the American Society for Testing
and Materials (ASTM) in "Class C" areas, which include most
areas of the country with significant ozone air quality
problems.[8] This value was chosen because nearly all ozone
NAAQS violations in these locations occur during the summertime
and, based on current fuel volatility trends, it is expected
that this RVP level will be reached by the time onboard
controls enter the marketplace. (In fact, this level is
already being exceeded in some areas, at least at times.) The
RVP of the certification fuel for the baseline scenario is the
currently specified value of 9.0 psi. Under the control
scenarios, the RVP of in-use fuel is again 11.5 psi. However,
the RVP of certification fuel is assumed to be revised to equal
that of in-use fuel.
2. Fuel Temperatures
The mass of gasoline vapors displaced from the vehicle's
fuel tank is highly dependent on the temperature of the
dispensed fuel, as well as the temperature of the fuel already
in the tank. As mentioned with regard to fuel volatility, most
ozone violations occur during the summer months. Therefore,
national average summertime temperatures for the two
fuel-related temperature parameters are appropriate when
modeling ozone air quality. These values are 88.2°F and
78.8°F for fuel tank and dispensed fuel, respectively.[3]
3. Refueling Emission Factors
The refueling emission factor is composed of three parts:
vehicle fuel tank displacement losses, spillage, and service
station tank emptying losses. The amount of displacement
vapors which escape from an uncontrolled vehicle during the
refueling event are found by using the RVP and temperatures
just described, along with the equation developed in an EPA
3-8
-------
technical report that relates these parameters to displacement
losses:[3]
Displacement losses =
-5 909 - 0.0949(AT) + 0.884(TD) + 0.485(RVP)
-5.909 - 0.0949(4-9.4) + 0.884(78.8) + 0.485(11.5)
Where:
To = temperature of the dispensed fuel (°F),
AT = TT - TD = difference between the temperature of
fuel already in the vehicle tank (TT) and dispensed
fuel, and .
RVP = Reid Vapor Pressure of the dispensed fuel (psi).
The resulting value for the displacement losses component of
the refueling emission factor is 5.7 g/gal.
As discussed in greater detail in reference [3], data on
the frequency and magnitude of spillage losses are both scarce
and widely variable. EPA has chosen to use 0.3 g/gal to
represent the spillage portion of the refueling emission factor
and 0.4 g/gal to represent the emptying losses. These values
are taken from Volume 1 of the draft RIA. Summing these three
components yields the uncontrolled refueling emission factor of
6.4 g/gal.
For vehicles equipped with onboard controls, two emission
factors are used in the analysis, depending on whether the
system is properly operating or has been tampered with. Tests
of properly operating (non-tampered) systems have shown the
capability to control 99 percent of the displacement losses.[4]
As noted earlier EPA expects that implementing onboard controls
will also serve to reduce spillage losses. However, since the
amount of spillage which will be controlled is unclear at this
time, EPA has taken the more conservative position of assuming
no reduction in spillage. Therefore, adding the spillage and
emptying losses, the resulting non-tampered emission factor is
0.8 g/gal.
Since there are no externally visible changes to a vehicle
equipped with an onboard control system, the rate of tampering
with such systems is assumed to be no more frequent than that
occurring with current evaporative controls, and will take only
the forms of removal or disconnection of the storage canister.
When this occurs, emissions would revert to their uncontrolled
level (i.e., 6.4 g/gal). MOBILES accounts for the effects of
tampering by utilizing this uncontrolled emission factor with
appropriate tampering rates (see below).
3-9
-------
It should be noted that the uncontrolled emission factor
used to examine ambient ozone air quality differs slightly from
that used to evaluate other health effects (i.e., 6.4 versus
6.6 g/gal). This is the result of using different volatility
and temperature data to estimate displacement losses for the
two analyses. As mentioned above, ozone violations are
primarily a summer problem, so the emission factor for the air
quality analysis is based on national summertime average
values. In contrast, the emission factor for the health
effects analysis was derived using national annual average
values since, unlike ozone violations, these phenomena are not
restricted to certain areas of the country or to a specific
season. Since fuel volatility and fuel temperatures vary over
the course of the year, the two resulting emission factors are
not equal. In any event, it should be noted that the
difference in the emission factors is only about 3 percent.
4. Tampering Rates
In the onboard control scenario, the tampering rates used
in MOBILES for removal and disconnection of the evaporative
emissions storage canister are also used to reflect refueling
canister tampering. It should be noted that the tampering
rates contained in the original MOBILES were revised in the
volatility study to reflect later information from EPA's in-use
surveillance programs.[8] The inclusion of these 1985 survey
data had the effect of slightly increasing the tampering rates,
relative to the original MOBILES tampering rates. These new
rates, which vary as a function of vehicle'mileage, have been
used in this analysis. The tampering rates used in this
analysis for LDVs and LDTs, respectively, are about 2.5 and 6
percent at 50,000 miles, and about 6 and 9 percent at 100,000
miles.[8]
It should also be noted here that with the use of a liquid
seal, which EPA anticipates will be the dominant approach, the
vehicle fillneck is no different with than without an onboard
control system. In fact, the incorporation of onboard
technology with the liquid seal should be virtually transparent
to the vehicle owner/operator, thus presenting no real
incentive to tamper. Thus, it is assumed that whatever
tampering will occur as a result of the implementation of
onboard controls will be limited only to the removal or
disconnection of the canister, as noted above. This would not
be the case with Stage II control systems, or with onboard
control systems using mechanical fillneck seals.
5. Inspection and Maintenance Program
The exhaust emission factors used in the air quality
analysis assume the existence of Inspection and Maintenance
Programs (I/M) in each of the non-attainment areas throughout
the projection period. This assumption has the effect of
3-10
-------
slightly lowering the average emission factors assigned to
those vehicles subject to I/M. This assumption was made in
consideration of the fact that most of the high-population
areas among the 61 non-California urban areas modeled either
have I/M programs in effect already, or have such programs
scheduled to go into effect in the near future. In addition,
I/M programs are a fundamental component of the Agency s
recently announced four-phase ozone compliance strategy (see
the introduction to this section).
6. Design Values and Base Year Inventories
As noted earlier, the design values and inventory
projections in this chapter use 1983 as the base year. More
specifically with regard to design values, two different types
have been used as part of this analysis. The design values
briefly discussed in Section III.A., and used in the
determination of non-attainment status, are based on the 1982,
1983, and 1984 air quality monitoring data without
modification. These values are shown in Table 1 under the
heading "82-84 DVs" and represent the fourth highest daily
maximum hourly ambient ozone concentrations measured in the
three years of monitoring data examined.
Previous ozone air quality analyses have used similarity
defined values as a necessary input for EKMA. In this
application, the value is interpreted as the ambient ozone
concentration in the base year. However, the air quality
analyses for this chapter include a refinement which leads to
the development of a second set of "design values" being used
to represent the base year ambient concentrations. These
values are also shown in Table 1, under the heading "EKMA DVs."
This second set of design values reflect modifications
that attempt to take into account the transport of ozone. This
is accomplished by a computer program known as MASH (Multiple
Airshed) . MASH uses wind directions and the times that ozone
exceedances (i.e., one-hour concentrations in excess of the
0.12 ppm NAAQS) occur to determine whether an exceedance is
predominantly the result of emissions originating in the local
area, or of emissions originating in a nearby major
metropolitan area upwind. It then assigns the measured ozone
levels to either the upwind metropolitan area or the local area
as a function of selected criteria.[5] In regions where ozone
transport is significant (e.g., the northeastern part of the
country), this process tends to lower the design values in
smaller cities or SMSAs and to raise the design values in the
larger metropolitan areas. These MASH-processed design values
are intended for use in ozone air quality modeling, and are not
appropriately used in determining attainment/non-attainment
status.
3-11
-------
It should also be noted that there is relatively little
difference in the two sets of design values, with the "82-84 DV"
and "EKMA DV" being equal for 46 of the 61 areas modeled. Of
the 15 areas where the application of the MASH program led to a
change in the design value, the value used for input to EKMA is
lower than the unmodified value in 10 of the areas and higher
in five of them. The areas with differences between the two
values are concentrated in the Northeast and along the central
East Coast (EPA Regions 1, 2, and 3), where many of the SMSAs
and metropolitan areas are contiguous. The greatest increase
in design value resulting from the MASH program occurs for the
Chicago Metropolitan Area. The increase of 0.05 ppm represents
a 25 percent increase over the unmodified value.
In addition to the design values, EKMA also requires NMHC
inventories and NMHC:NOx ratios. As noted above, the
inventories are developed as part of the National Emission Data
System (NEDS).
To model air quality in a given future projection year, a
number of assumptions must be made about control technology,
control efficiency, and growth rates. For example, modeling
the air quality impact of the proposed refueling emission
standard involves the calculation of emission factors for
refueling emissions from all gasoline-fueled mobile sources
(light-duty vehicles and trucks and heavy-duty vehicles), both
with and without onboard refueling controls. By holding other
parameters constant in both cases (i.e., growth rates, exhaust
emission standards, stationary source control efficiences and
growth rates), the effect of the proposed refueling emission
standard on ozone air quality can be projected.
7. Stationary Source RACT and the FMVCP
The modeling of future NMHC inventories and ambient ozone
concentrations involves looking at emissions from both mobile
and stationary sources. For mobile sources, the key inputs
include the emission factors (as calculated by MOBILE3) and
vehicle miles travelled (VMT) growth rates. The emission
factors for each vehicle/engine class vary by model year,
reflecting the impact of the Federal Motor Vehicle Control
Program (FMVCP), under which increasingly stringent exhaust and
evaporative emission standards have been established.
EPA has developed an extensive data base on the in-use
performance of mobile source emission control technology.
Based on this information, the Agency has a high degree of
confidence in the emissions reductions projected to result from
standards established under the FMVCP. Even when coupled with
projected VMT growth, the FMVCP gains net reductions in the
mobile source portion of future NMHC inventories into the 1990s.
3-12
-------
In the case of stationary sources, the two main inputs
required for modeling future emissions are growth rates and
control efficiencies. Stationary source NMHC emissions are
divided into six major categories in EPA's models.[8,9] For
some stationary sources, EPA has defined "reasonably available
control technologies" (RACT) and issued control technology
guidelines (CTGs). RACT-based emission controls, for those
categories covered by CTGs, are required to be included as part
of each state's plan for attaining the NAAQS for ozone.
Previous EPA modeling results have always indicated
decreasing emission inventories and improving air quality at
least through the mid and late 1980s. Nonetheless, the most
recent ambient ozone monitoring data indicate that the
projected improvements indeed may not be occurring. It is
widely accepted that the mobile source portion of the inventory
will continue to decline as a result of fleet turnover, with an
increasing fraction of the national fleet being comprised of
vehicles subject to the stringent emission standards taking
effect in the early and mid 1980s. Since the FMVCP continues
to yield emission reductions, but monitoring data suggest the
expected improvement in ozone air quality is not occur ing, the
assumptions used in previous EPA analyses concerning the
efficiency and timing of stationary source RACT control appear
to have been overly optimistic.
The Agency knows that some RACT-based emission reductions
had already occurred as of the base year; what is not known is
what fraction of" sources subject to RACT actually were
controlled. Nor is it known, because of various practical
constraints on the application of RACT controls, what fraction
of the sources subject to RACT can actually be expected to be
controlled in the near future. Without this information, the
amount of additional RACT-based control still available is also
uncertain. EPA is attempting to resolve these questions as
part of the abovementioned review of RACT controls. Until EPA
has better data available on the actual rate of implementation
and control efficiency of stationary source RACT, the
"baseline" scenario for this analysis assumes no RACT-based
stationary source emission reductions are assumed to occur
after the base year.
C. Ambient Air Quality Results
Tables 2 and 3 present the results of the EKMA-based air
quality projections for the 61 urban areas currently in non-
attainment of the ozone NAAQS. The information in Table 2 is
also presented graphically in Figure 1. Two measures of air
quality in various calendar years are shown: total NMHC
emissions and the number of non-attainment areas. Each table
displays the model's output for the scenarios analyzed:
baseline (FMCVP only), control of excess evaporative emissions
only, and control of both refueling and excess evaporative
3-13
-------
emissions. A comparison of the results for the three scenarios
indicates the incremental effects of implementing control of
excess evaporative emissions and then refueling controls.
Before discussing the results for the control scenarios,
it is interesting to briefly examine the projections for the
baseline scenario in each table. Generally, the results for
this scenario show significant improvements occur primarily in
response to the stationary and motor vehicle emission standards
which already have been promulgated. However, T.able 3 shows
that a significant number of violations and non-attainment
areas continue to exist throughout the projection period. It
also appears that by about the mid to late 1990s, the trend
toward decreasing ambient ozone levels will be reversed, and
air quality will worsen steadily thereafter. The continued
number of non-attainment areas, in addition to the projected
deterioration in air quality, indicates the need for additional
HC reductions to help attain and maintain the NAAQS for ozone
across the nation.
Regarding the effects of the control scenarios, Table 2
shows that as onboard-equipped vehicles comprise an increasing
percentage of the motor vehicle fleet over time, reductions in
the non-methane hydrocarbon inventory due solely to refueling
controls change from 105,000 tons (about 1.5 percent) in 1995,
to 179,000 tons (about 2.1 percent) in 2010. If reductions in
excess evaporative emissions are also included, the reductions
are approximately double those amounts.
Based on the air quality modeling results presented above,
and in Tables 2 and 3 and Figure 1, it appears that the
implementation of onboard refueling standards for gasoline-
fueled LDVs, LDTs, and HDVs can provide significant long-term
benefits in helping to achieve and maintain the ozone NAAQS.
3-14
-------
Table 2
Total Non-Methane Hydrocarbon Emissions
for 61 Urban Areas (xlOOO Tons)
Scenario
1990
1995
2000
2010
Baseline (FMVCP only)
Excess Evap Control'
Refueling & Excess
Evap Control
6961 6926 7256 8420
(14.2)1 (14.6) (10.5) (+3.8)
6958
(14.2)
6823
(15.9)
7074
(12.8)
8160
(+0.1)
6931 6718 6928 7981
(14.6) (17.2) (14.6) (1.6)
Figures in parentheses represent the percent change
(decrease, unless otherwise noted), relative to the 1983
base-year inventory of 8,111.6 tons for the 61 areas.
No control of refueling.
3-15
-------
9I-E
TOTAL NMHC INVENTORY
(Thousands)
D
oo
CD
>
in
ir?
ID
o
m
x:
m
70 -*
to
m
O
O
I I < I I I I l
-------
Table 3
Estimated Number of Non-California
Urban Ozone Non-Attainment Areas
Scenario 1990 1995 2000 2Q1Q
Baseline (FMVCP only) 47 46 48 55'
Excess Evap Control2 " 47 44 47 53
Refueling & Excess
Evap Control 46 40 44 51
Only the 61 non-California areas listed in Table 3 were
modeled. Additional areas are likely to be in
non-attainment in 2010 under the "Baseline (FMVCP only)"
scenario.
No control of refueling.
3-17
-------
REFERENCES FOR CHAPTER 3
1. "Air Quality Criteria for Ozone and Other
Photochemical Oxidants," U.S. EPA, ORD, April 1978,
EPA-600/8-78-004.
2. "Air Quality Criteria for Ozone and Other
Photochemical Oxidants," U.S. EPA, ORD, August 1986,
EPA-600/8-84-020LaF.
3. "Refueling Emissions from Uncontrolled Vehicles,"
Dale S. Rothman and Robert Johnson, U.S. EPA, QMS, ECTD,
EPA-AA-SDSB-85-6, July 1985.
4. "Vehicle Onboard Refueling Control," American
Petroleum Institute, Report No. 4424, March 1986.
5. "1982-84 Ozone Design Values for Regulatory Impact
Analyses," Memorandum from Richard G. Rhoads, Director of
Monitoring and Data Analysis Division, to Charles Gray,
Director of Emission Control Technology Division, June 16, 1986.
6. "Uses, Limitations and Technical Basis of Procedures
for Quantifying Relationships Between Photochemical Oxidants
and Precursors," U.S. EPA, OAQPS, EPA-450/2-77-021a, November
1977.
7. "Guidelines for Use of City-Specific EKMA in
Preparing Ozone SIPs," U.S. EPA, OAQPS, EPA-450/4-80-027, March
1981.
8. "Study of Gasoline Volatility and Hydrocarbon
Emissions from Motor Vehicles", U.S. EPA, QMS, November 1985,
EPA-AA-SDSB-85-5.
3-18
------- |