EPA-AA-TEB-EF-85-]
EVAPORATIVE HC EMISSIONS FOR MOBILES
August 1984
Test and Evaluation Branch
Emission Control Technology Division
Office of Air and Radiation
U. S. Environmental Protection Agency
-------
Table of Contents
1.0 INTRODUCTION
2.0 MOBILE2 ASSUMPTIONS
3.0 LDGV DATA ANALYSIS FOR MOBILE3
3.1 Indolene Rates
3.1.1 Low Altitude
3.1.1.1 Tampering
3.].1.2 California Data vs. Federal Data
3.1.1.3 Differences in Fuel-Injected and Carbureted Vehicles
3.1.1.4 Summary of Low Altitude Indolene Rates
3.1.2 High Altitude
3.2 Commercial Fuel Rates
3.2.1 1981+ Vehicles
3.2.2 Pre-1981 Vehicles
3.3 Summary of LDGV Evaporative Emissions
4.0 OTHER VEHICLE TYPES
4.1 LDGT1
4.2 LDGT2
4.3 HDGV
-------
1.0 INTRODUCTION
It has been recognized that in estimating the mobile source emissions, a
sizeable portion of the hydrocarbon (HC) emissions comes from the
evaporation of fuel from both carburetor and fuel tank of a parked
vehicle. The evaporation of fuel from a vehicle at the end of a trip is
called the hot soak loss. The primary source of emissions during hot
soak is the carburetor system. A small amount of the hot soak emissions
may also occur due to fuel tank heating. Evaporation of fuel from a
vehicle due to changes in ambient temperature resulting in an expansion
of air-fuel mixture in a partially filled fuel tank is called the diurnal
breathing loss. The primary source of emissions in this phase is the
gasoline tank. As the fuel temperature rises, evaporation of fuel
increases.
All current automobiles since 1971 use a carbon-filled canister (or
canisters) to collect hot soak or diurnal fuel vapors. These vapors are
then pulled into the engine during engine operation to purge the
canister. The hot soak or diurnal emissions can escape if the canister
is saturated or if there are other.problems in the evaporative emission
control system. Evaporative emissions are generally associated only with
gasoline-powered vehicles and trucks. For diesel-powered vehicles and
trucks, the evaporative emissions are relatively insignificant due to a
very low volatility of the diesel fuel.
Since 1971, EPA's Emission Factor (EF) programs have been collecting-
evaporative emission data from in-use light-duty vehicles and trucks.
Indolene has been the primary fuel used for evaporative tests because
Indolene is the specified certification fuel, and the quality of Indolene
is generally more consistent than other fuels. This consistency in fuel
quality was desired to reduce the variability in exhaust and evaporative
emission data. Data from light-duty vehicles (LDGVs) were analyzed and
the calculated Indolene evaporative emission rates were used in MOBILE2.
Evaporative emissions are estimated for both the hot soak and diurnal
phases of testing. The sum of these two emissions is the value to be
compared with the standards. The evaporative emission standards for all-
vehicle types are summarized in Table 1.
At some test sites both Indolene and commercial fuels were used on
certain vehicles.
-------
Table 1
Evaporative Emission Standards
Veh. Test
Type Procedure*
LDGV None
Canister
Canister
SHED
SHED
SHED
LDGT1 None
Canister
Canister
SHED
SHED
SHED
LDGT2 Same as HDGV
SHED
SHED
SHED
HDGV None
Canister
SHED
SHED
SHED
Standard**
(grams)
None
6.0
2.0
6.0
2.6
2.0
None
6.0
2.0
6.0
2.6
2.0
None
6.0
2.6
2.0
None
2.0
6.0
3.0/4.0
2.0
Low Altitude
Non-California
Pre-1971
1971
1972-77
1978-80
1981+
pre-1971
1971
1972-77
1978-80
1981+
Pre-1979
1979-80
1981+
Pre-1985
1985+***
California
Pre-1970
1970-71
1972-77
1978-79
1980+
Pre-1970
1970-71
1972-77
1978-79
1980+
Pre-1977
1978-79
]980+
Pre-1973
1973-77
1978-79
1980+
High Altitude
Non-California
1977
1982-83
1984+
1977
1982+
1982+
* The two test procedures are: canister (or carbon) trap method and
Sealed Housing for Evaporative Determinations (SHED) method.
** The sum of hot 'soak and diurnal emissions.
*** The 1985+ evaporative standards are split by the gross vehicle
weight, with 3.0 grams/test for vehicles less than 14,000 Jb. GVW,
and 4.0 grams/test for vehicles greater than or equal to 14,000 Ib.
GVW.
The purpose of this report is to analyze the evaporative emission data
collected from EF programs and to provide evaporative emission rates for
MOBILES. Some of the basic assumptions used In MOBILE2 are carried over
to MOBILES. For clarity, assumptions used in MOBILE2 are briefly
presented in Section 2. Section 3 Is a detailed discussion of the LDGV
EF data. Section 4 provides evaporative emission rates for both the
light-duty and heavy-duty trucks.
-------
2.0 MOBILE2 ASSUMPTIONS
The LDGV evaporative emission data from EF programs prior to 1981 were
used to establish evaporative emission rates for MOBILE2. Only tests on
Indolene fuel were used. In calculating emissions, MOBILE2 used the
following assumptions'
a. The available LDGV test data from low altitude California
and Federal 49-state sites (EF71 through EF80 programs) were combined
where their evaporative emission standards were the same to establish one
set of evaporative emission rates to represent both low altitude
non-California (Federal) and California regions.
b. Since there were very few data on 1981 vehicles available
when MOBILE2 was being developed, the evaporative emission levels for
1981 and later model years were based upon 1980 in-use data factored for
changes in 1980 and 1981 certification data.
c. The high altitude evaporative emission rates were the rates
from low altitude multiplied by an altitude correction factor.2 This
altitude factor was derived from a mathematical model which calculated
low and high altitude (5200 ft.) evaporative emissions from uncontrolled
test vehicles. The basic premise of the model is that evaporative
emissions are inversely related to the atmospheric pressure. The derived
ratio of high to low altitude uncontrolled emissions was 1.30. MOBILE2
used this ratio for all model year groups except one, the one exception
being the 1977 model year In which the high altitude evaporative emission
standard was the same as the low altitude standard.
d. Since the evaporative emission standards were the same for
LDGVs and LDGTls3, the MOBILE2 LDGT1 evaporative emission rates were
the same as the LDGVs.
e- The LDGT2^ evaporative emission rates for pre-1979 model
years were the same as those for HDGVs. For 1979 and later years, the
LDGT2 evaporative emission standards were equivalent to LDGTls,
therefore, the MOBILE2 emission rates were the same.
Michael W. Leiferman, "Effect of Altitude on noncontrolled
Evaporative Emissions from Gasoline Fueled Vehicles," January 1979,
an EPA technical report from Standards Development and Support Branch.
Light-duty trucks with 0-6000 Ibs. GVW.
Light-duty trucks with 6001-8500 Ibs. GVW.
-------
f. The HDGV evaporative emission rates for the pre-1968 low
altitude non-California region were derived from nine in-use heavy-duty
vehicles.^ In MOBILE2, the same rates were applied for years 1968-83
since there was no evaporative emission standard for HDGVs. MOBILE2
assumed a proposed standard of 3.0 grams/test for 1984 and later
HDGVs^, and the evaporative emission rates for 3.0 gram HDGVs were
derived by observing the changes in in-use LDGV evaporative emission
rates between 6.0 gram vehicles and 2.0 gram vehicles, and assuming a
portion of those changes would also occur in heavy duty trucks if the
standard were 3.0 grams/test.
See discussions on Section 4.3.
Subsequent to the publication of MOBILE2, the HDGV evaporative
emission standards have been set as 3.0 or 4.0 grams depending on the
gross vehicle weight. See footnote under Table 1, section 1.0.
-------
7
3.0 LDGV DATA ANALYSIS FOR MOBILE3
More evaporative emission test data have become available since MOBILE2,
mostly from FY82 and later EF Programs. In November 1983, the EF
programs started to collect evaporative emission data based upon both
Indolene and commercial fuels. Also available are Indolene fuel
evaporative emission data from the California Air Resources Board
(GARB)7. Only LDGV data from as-received testing are included in this
analysis.
The analysis on LDGV data is divided into two parts: 1) the evaporative
emission rates based on Indolene fuel, and 2) the commercial fuel
emission levels. Since more vehicles have been tested on Indolene fuel
than commercial fuel, the Indolene rates are used to draw conclusions on
issues such as deterioration and fuel injection versus carburetion.
However, the MOBILES evaporative emission rates are based on commercial
fuel data.
3.1 Indolene Rates
As almost all the Indolene test data collected since MOBILE2 are on
vehicles designed to meet the 2.0 gram SHED standard, this section Is
concentrating mainly on results of these vehicles. Attachment 1 is a
summary of the sample size in the complete Indolene fuel evaporative data
base stratified by model year and region.
3.1.1 Low Altitude
For 2.0 grams/test SHED standard, available Indolene fuel test data
representing the Federal region are the EF data collected in Ann Arbor.
The California data, on the other hand, include results from 1980 and
1981 California vehicles collected by EPA and data obtained from CARB
surveillance programs.
Discussions on low altitude evaporative emissions are divided into three
specific areas- tampering, California data vs. Federal data, and
differences in fuel injected and carbureted vehicles.
3.1.1.1 Tampering
One of the major revisions in MOBILE3 computer model is that the effect
of visible tampering on mobile source emissions is to be considered
7 "Test Report of the High Mileage (Three-Way) Catalyst Vehicle
Surveillance Program, Series 2 (HMCVSP2)," California Air Resources
Board, March, 1983.
-------
8
separately from the untampered emission factors. For this reason,
vehicles that are suspected of being tampered are to be analyzed
separately".
Tampering is defined as the disablement of any component of the emission
control system, whether it was done deliberately, inadvertently, or
through neglect. The components of the evaporative emission control
system include the charcoal canister (or canisters), purge valves, vent
lines, vacuum lines, and gas cap. Therefore, the disablements of the
evaporative emission control system include disconnected, misrouted, or
cut lines (bowl, tank vent or vacuum) or hoses (purge control), missing
canisters, missing gas caps, and the removal of the entire evaporative
emission control system.
Under these criteria, one California vehicle and five Federal vehicles
were classified as tampered vehicles. Attachment 2 is a list of tampered
vehicles, their evaporative emission test results, and their tampering
characteristics. Mechanic comments on vehicle conditions indicated that
the California vehicle and three of the Federal vehicles had disconnected
vacuum lines. For the two remaining Federal vehicles, one had misrouted
vacuum lines plus missing gas cap, and one had canister missing.
Tampering of evaporative emission control system has a significant effect
on evaporative emissions, this is illustrated in Attachment 2. For-
example, vehicle No. 4095 was found with canister missing, the Indolene
fuel evaporative emission total was about ten times the emission standard
(2.0 grams/test). Several vehicles listed in Attachment 2 had
disconnected vacuum lines, the Indolene fuel emission totals for these
vehicles were from three to five times the emission standard.
To derive the untampered evaporative emissions for MOBILES, all tampered
vehicles were excluded from the sample. Thus, subsequent analyses and
conclusions are all based upon the data base with no.tampered vehicles.
3.1.1.2 California Data versus Federal Data
The evaporative emission rates for MOBILE2 were derived from combining
Federal data with the California data where the evaporative emission
standards were the same. For example, the emission rates for vehicles
built to meet the 2.0 gram SHED standard were developed from 1980 and
later California vehicles and 1981 and later Federal vehicles. However,
the current analysis has shown that these two samples appear to be
"Anti-Tampering and Anti-Misfueling Programs to Reduce In-Use
Emissions From Motor Vehicles", EPA-AA-TSS-83-10, pages 35-36.
-------
nonhomogeneous. This nonhomogeneity is evident by examining the mean
emissions from these two samples. Table 2 is a summary of the mean
emission rates and mean mileages of the two samples.
Table 2
Comparisons of Indolene Evaporative Data
From the Two Low Altitude Regions
(SHED 2.0 Grams/Test Standard)
Region N
MYR
Low Altitude
Non-California
Low Altitude
California*
242 1981-83
99
157
256
1980
1981
1980-81
Mean Mileage
(Miles)
37,749
16,889
9,109
12,118
Mean Emissions (gms)
Hot Soak Diurnal Total
1.97
1.64
0.84
1.15
2.85
0.88
1.10
1.01
4.82
2.52
1.94
27T5"
* This includes 104 vehicles tested by CARB. The mean emissions of the
CARB data are not significantly different from the mean emissions of
data on California vehicles collected by EPA.
EPA investigated a number of areas to try to explain the emission
differences between these two samples. Areas investigated include:
a. Evaporative test procedures (i.e., prep cycles)
b. Technology
c.
mileage
Rates of observed malfunctions related to differences in age or
d. Canister working capacities related to differences In
age/mileage, or mean Reid Vapor Pressure (RVP) of summer and winter fuels.
The results of this investigation were summarized in an EPA memo." The
conclusions were that the emission differences of the two samples could
be traced to items a and c and possibly item d. It is decided that the
analysis of Indolene fuel evaporative emission rates should be based on
the Ann Arbor data alone.
" "Differences Between California and Ann Arbor Evaporative Data,"
from Tom Darlington to Charles L. Gray, August 24, 1984.
-------
10
3.1.1.3 Differences in Fuel-Injected and Carbureted Vehicles
The type of fuel delivery system is considered to be one of the factors
which may have an effect on the evaporative emissions, especially for the
hot soak emissions. Vehicles with fuel injection (both ported and
throttle body injection) do not have carburetor float bowls which are the
major source of the hot soak emissions in carbureted vehicles.
Table 3 provides mean Indolene evaporative emissions and mean mileages
for the fuel-injected and carbureted vehicles tested in Ann Arbor (sample
distributions by model year are given in Attachment 3). The hot soak and
diurnal emissions of the fuel-injected vehicles are lower than the
emissions of carbureted vehicles. These differences are significant at
the 95% level.
Table 3
Indolene Evaporative Emission Rates
by Fuel Delivery Systems
Mean Mean Emissions (gms)
Sample N Mileage Hot Soak Diurnal
Carbureted 179 44,009 2.27 3.08
Fuel Injected 63 20,241 1.12 2.18
One reason for the difference in hot soak emissions between the
fuel-injected and carbureted vehicles is the absence of float bowls on
the fuel injection systems. However, the difference in hot soak
emissions might be less if the two samples were at equivalent mileages.
There is the potential for leaks around injectors in vehicles with ported
injection, and the probability of leaks increases with the vehicle's
mileage and age. This may be verified by future testing of high mileage
fuel-injected vehicles.
There are a variety of reasons which could explain al] or part of the
differences in diurnal emissions of fuel-injected and carbureted
vehicles. Possible reasons include:
a. Differences in mean fuel tank sizes (assuming that the
canister sizes are the same)
b. Differences in observed rates of malfunctions related to
age or mileage
c. Differences in unobserved malfunctions (such as
deterioration in canister working capacity) related to age or mileage
-------
11
d. Differences in purging technology, and
e. Differences in canister working capacity related to the lower
hot soak emissions of fuel-injected vehicles.
Fuel Tank Sizes - One of the determinants of the amounts of vapor
generated during a diurnal test is the vehicle's fuel tank volume or
size. More vapor is generated from a large tank than a small one when
both are filled to the same level. However, an examination of mean fuel
tank sizes indicated that the fuel-Injected and carbureted samples had
the same mean fuel tank size of 14.5 gallons.
Malfunctions - The fuel-injected vehicles in the EF sample are mostly of
late model year vehicles (1982 and 1983). The carbureted vehicles in the
EF sample are high mileage 1981 model year vehicles. As a result, the
fuel-injected vehicles have a much lower average odometer reading than
carbureted vehicles.
Table 4 shows the different rates of malfunctions between the two samples
and the mean emissions of malfunctioning and well-functioning vehicles.
Table 4
Indolene Evaporative Emission Rates
% of Mean Mean Emissions (gms)
Sample N Sample Mileage Hot Soak'Diurnal
Vehicles with no malfunctions
Carbureted 139 77.7 41,066 1.94 2.40
Fuel-Injected 58 92.0 20,35] 1.01 1.45
Vehicles with malfunctions
Carbureted ' 40 22.3 54,235 3.43 5.47
Fuel-Injected 5 8.0 18,964 2.40 10.65
The system malfunctions of the fuel-injected and carbureted vehicles are
identified in Attachment 4. Note that 22% of the carbureted vehicles
have some kind of system malfunction, while only 8% of the fuel-injected
-------
12
vehicles have malfunctions. The rate of malfunctions may be associated
with vehicle's mileage and age, and it is possible that the fuel-injected
vehicles may have similar numbers of malfunctions when they get to the
age of the carbureted vehicles. However, the number of observable
malfunctions between the two samples does not explain all of their
emission differences, because the emission differences still remain when
the malfunctioning vehicles are removed.
Unobserved Malfunctions - Examples of possible unobserved malfunctions
are age cracks in hoses^ and deterioration in canister's usable
working capacity. Although canister's usable working capacity is known
to stabilize at a certain level after declining somewhat from the "green"
or new condition, it is possible for the canister working capacity to
suffer a sharp drop from a stabilized level if the canister is subjected
to a large quantity of vapor. This is more likely to occur on carbureted
vehicles than the fuel-injected vehicles.
The method.used to characterize the unobserved malfunctions is to perform
regressions on the carbureted and fuel-injected samples that exclude
vehicles with observed malfunctions. If the deterioration rates are the
same and the zero mile levels are not significantly different, the
conclusion would be that the fuel-injected vehicles should emit the same
amount of emissions as the carbureted vehicles at similar mileages.
Least square regression analysis is used first to see the relation
between the emissions and vehicles' mileages for the carbureted and
fuel-injected vehicles. Results indicate that deterioration for the
diurnal emissions of both types of vehicles was nonsignificant (the
deterioration for hot soak emissions was also nonsignificant), as shown
in Table 5. For fuel-injected GM vehicles, which made up the majority of
the fuel-injected sample, regression analysis is performed again with
result indicating that the deterioration is stil] nonsignificant. For
the carbureted vehicles, a least square analysis of covariance is
performed by using the manufacturers as strata to test if vehicles from
various manufacturers have the same deterioration rate. Results,
indicated in Table 5 also, show that carbureted vehicles from domestic
manufacturers have the same and significant deterioration rate on diurnal
emissions.
Recently, vacuum testing of all hoses has been incorporated into
the diagnostic work done on testing EF vehicles to detect age
cracks or other leaks that are not readily visible.
-------
13
Table 5
Regression Results* on Indolene Evaporative Emissions
From the Problem-Free Vehicles
Sample fJ
Carbureted-all 139
GM 49
Ford 43
Other Domestic 20
Imports 27
Fuel-Injected-all 58
GM 47
Ford 8
Imports 3
Hot Soak
ZM
2.11
1.83
2.13
2.21
3.10
1.05
1.02
DR
-0.04
-0.01
-0.01
-0.01
-0.01
-0.02
0.03
Diurnal
ZM
2.27
1.16
2.50
0.67
1.64
1.57
1.14
DR
0.03
0.31**
0.31**
0.31**
-0.09
-0.06
0.09
* ZM=emissions in grams per test at zero miles, DR=deterioration rate in
grams per test per 10K miles.
** Coefficient is significant at a level of 0.10.
Since most of the fuel-injected vehicles were from GM, it was decided to
perform an analysis of covariance on the GM fuel injected versus carbureted
vehicles to test the equality of the regression slopes and zero-mile levels.
The results are shown in Table 6. This test showed that the deterioration
rates, and their zero mile levels, were not statistically different for the GM
fuel-injected and carbureted vehicles.
Table 6
Regression Results* on Indolene Evaporative Emissions
From the Problem-Free GM Vehicles
Sample
Carbureted
Fuel-Injected
49
47
Hot Soak
ZM
1.30
1.02
Diurnal
DR
0.14
0.03
ZM
1.17
0.74
DR
0.30**
0.30**
* ZM=zero mile in grams per test, DR=deterioration rate In grams per
test per 10K miles.
** Coefficient is significant at a level of 0.10.
-------
14
This latter data provides a preliminary indicator that the fuel-injected
vehicles may have diurnal emissions similar to the carbureted vehicles
when they are at equivalent mileages. However, there is not yet enough
data to demonstrate this for the fleet.
Excess Canister Working Capacity of Fuel-Injected Vehicles - The
evaporative test procedure leading up to the diurnal test consisted of
the following sequence:
1. Drain and refuel test fuel to 40% level
2. Precondition (either a 10 minute road route or 10 to 10.5
minute dynamometer cycle)
3. Minimum 12 hour soak
4. Diurnal test
The preconditioning period produces a hot soak, which tends to add more
vapor to the canister for a carbureted vehicle than is added to the
canister of a fuel-injected vehicle.
If the canister sizes were similar, the fuel-injected vehicles would have
more reserved capacity going into the diurnal test. This could also be a
reason for the lower diurnal emissions of fuel-injected vehicles in these
samples. However, the magnitude of this effect is difficult to
quantify. Also, pressures from competitive pricing may lead
manufacturers to downsize the canisters of fuel-injected vehicles if
there is excess capacity.
Purging Technology - When the diurnal emission differences between
fuel-injected and carbureted vehicles were first noted, the automobile
manufacturers suggested that there might be differences in purging
technology which would lead to the differences in diurnal emissions. EPA
therefore investigated these potential differences, and found them to be
mostly insignificant. Both fuel-injected and carbureted vehicles use
engine vacuum (at the Intake manifold or at the carburetor) to pull purge
air through the canister and into the engine. Since many 1981 and later
vehicles use closed-loop fuel control and have on-board computers, both
fuel-injected and carbureted vehicles use various engine sensors
(temperature, manifold vacuum, and/or engine rpm) to control
purging.11 Here there is one small difference - fuel-injected vehicles
may use more precise sensors. The temperature sensor on fuel-injected
vehicles must be a continuous temperature sensing device, so that the
proper mixture is delivered to the engine. A continuous temperature
sensing device is not necessary on most carbureted vehicles, because the
mixture during cold start is controlled by the choke. The temperature
sensing devices on carbureted vehicles are therefore on/off devices which
sense when the engine passes a pre-set temperature point.
11 Purging is avoided during cold start and idle because the additional
HC causes enrichment of the fuel/air ratio, causing an increased
risk of stalling.
-------
Therefore, control of purging in fuel-injected vehicles may be somewhat
more precise than in carbureted vehicles. However, the more precise
control of purging may not necessarily lead to lower emissions. It is
concluded that there may not be any differences in purging technology for
the fuel-injected and carbureted vehicles that would lead to significant
differences in diurnal emissions.
Conclusions on the Diurnal Emission Differences Between Fuel-Injected and
Carbureted Vehicles - The differences in diurnal emissions between
fuel-injected and carbureted vehicles have been attributed to differences
in observed malfunctions, possibly unobserved malfunctions (both of these
two are probably related to age/mileage differences between the two
samples), and possibly also the current excess capacity of canisters used
on fuel-injected vehicles due to smaller hot soak loading. This excess
capacity may not be a permanent difference. The first two differences
would suggest that fuel-injected vehicles will have diurnal emissions
similar to the carbureted vehicles when they are at equivalent mileages.
Therefore, it is concluded that the diurnal emission rates from
carbureted vehicles are the most representative for all vehicles
(fuel-injected and carbureted) at this time.
Conclusions on the Hot Soak Emission Differences Between Fuel-Injected
and Carbureted Vehicles - The differences in hot soak emissions between
carbureted and fuel-injected vehicles may be attributed to their
differences in a fuel delivery system (with and without the carburetor
float bowls). Therefore, it is concluded that the hot soak emissions
should be derived from separate hot soak rates for the two types of fuel
delivery systems and weighted according to their respective market shares.
3.1.1.4 Summary of Low Altitude Indolene Rates
A summary of low altitude Indolene fuel evaporative emission rates is
given in Table 7. The MOBILE2 emission rates, which were based upon
Indolene fuel data, are also listed for comparison purposes. Note that
the MOBILE2 hot soak and diurnal emission rates for 1981 and later years
were updated. Rates for pre-1978 years remain unchanged.
The MOBILE2 low altitude 1978-80 rates, 1.56 grams for hot soak and 2.65
grams for diurnal, seem unrealistically low, in comparison with the new
1981+ rates. There have been no significant design changes-^ made to
the evaporative emission control hardware since 1978 model year
There are no significant hardware differences between the 1978-80
and 1981 and later model year group low altitude vehicles. The only
exception is the 1980 California vehicles that had extra charcoal in
their canisters because of the new SHED 2.0 grams/test standard that
came into effect in 1980. That extra charcoal, however, has been
taken out for 1981 and later California vehicles.
-------
16
vehicles. MOBILE2 derived 1978-80 evaporative emission rates based upon
mostly low mileage non-California vehicles (with mean mileage at 14,100
miles, from Table 8, Section 3.1.2) and a few California vehicles. To
use the MOBILE2 rates of 1.56 grams for hot soak and 2.65 grams for
diurnal would imply that 1978-80 SHED 6.0 grams/test standard vehicles
represented a better emission control than the 1981+ SHED 2.0 grams/test
standard vehicles. It is more likely that the difference is due to
sample differences. Since the sample of 1981+ vehicles is much larger
and at more appropriate mileage, the 1978-80 emission rates were set
equal to the 1981+ rates. No fuel-injection sales are assumed for
1978-80 vehicles. Therefore, the Indolene emission rates of 2.27 grams
for hot soak and 3.08 grams for diurnal from the carbureted vehicles
(from Table 3) are used also for 1978-80 model year vehicles.
The Indolene hot soak emission rates of 1.12 grams from the fuel-injected
vehicles and 2.27 grams from the carbureted vehicles (from Table 3) are
used to derive hot soak rates for 1981 and later model year vehicles-
These two rates are weighted together by sales projections of fuel
injected and carbureted vehicles to derive overall Indolene hot soak
rates. The LDGV fuel injection sales projections are presented in
Attachment 5. The Indolene emission rate of 3.08 grams for the diurnal
loss from the carbureted vehicles (from Table 3) is used for the diurnal
emission rate of all 1981 and later vehicles.
The Indolene evaporative emission rates, although not used directly in
MOBILES, are useful in fuel volatility studies. The conclusions from
Indolene fuel emission analyses, such as deriving hot soak rates from
fuel injected and carbureted vehicles, and using only the carbureted
sample for diurnal emissions, are carried over into the analysis of
commercial fuel rates used for MOBILE3.
-------
17
Table 7
Low Altitude LDGV Indolene Evaporative Emissions (gms)
Equivalent MOBILES
Indolene Rates MOBILE2
MYR Group
Pre-1971
1971
1972-77
1978-80*
1981**
1982
1983
1984
1985-86
1987-89
1990+
Hot Soak
14.67
10.91
8.27
2.27
2.16
2.06
1.95
1.81
1.57
1.36
1.25
Diurnal
26.08
16.28
8.98
3.08
3.08
3.08
3.08
3.08
3.08
3.08
3.08
Hot Soak
14.67
10.91
8.27
1.56
0.63
0.63
0.63
0.63
0.63
0.63
0.63
Diurnal
26.08
16.28
8.98
2.65
1.07
1.07
1.07
1.07
1.07
1.07
1.07
* Emission rates for model year group 1978-80 are set equal to the
emission rates from 1981 and later carbureted vehicles (2.27 grams
for hot soak and 3.08 grams for diurnal). There are no
fuel-Injection sales (see Attachment 5).
** For 1981 and later years, the hot soak rates a sales weighted
combination of 1.12 grams for the fuel-injected vehicles and 2.27
grams for the carbureted vehicles. Sales of fuel-injected vehicles
can be obtained from Attachment 5. The diurnal emission of 3.08
grams for the carbureted vehicles is used for all Indolene diurnal
rate.
-------
18
3.1.2 High Altitude
There are no separate evaporative emission standards for the high
altitude region except for model years 1977 and 1982 and later (see Table
1). For all model years, manufacturers have generally used the same
evaporative emission control system on vehicles without differentiating
whether the vehicles were to be sold at high or low altitude. MOBILE2
used the altitude factor of 1.30 for most of the model years to adjust
the low altitude evaporative emission rates to high altitude (discussed
previously in Section 2.0). The exception was 1977, in which high and
low altitude rates were assumed to be the same because their emission
standards were the same.
There were no high altitude evaporative emission data from in-use
vehicles available for the pre-1972 years. An altitude correction factor
of 1.30 was derived from emissions of uncontrolled test vehicles (see
discussions on high altitude evaporative emission rates under Section
2.0). Since pre-1972 vehicles are basically uncontrolled vehicles (no
evaporative emission standard for pre-1971 vehicles and 6.0 gram canister
for 1971 vehicles), this altitude factor of 1.30 is probably appropriate
for use. Therefore, the pre-1972 high altitude evaporative rates are
derived from low altitude rates adjusted by the altitude factor of 1.30.
EPA's Emission Factor programs also collected evaporative emission data
from in-use light-duty vehicles operated and tested in Denver. These
data have shown that the altitude factor of 1.30 developed from emissions
of uncontrolled vehicles may not be appropriate for the controlled
vehicles. Table 8 is a comparison of high versus low altitude
evaporative emissions. All vehicles with malfunctions have been included
in the mean emission results, while the tampered vehicles have been
excluded.
Note that for the 2.0 grams/test canister standard, the low altitude
rates are represented by combining the non-California and California
data, since there were practically no low altitude non-California data in
this model year group. For the two SHED standards, only the low altitude
non-California data are used for comparison. The 2.0 gram California
vehicles have been excluded in estimating the 2.0 gram low altitude
emission rates, for the reasons discussed in section 3.1.1.4.
-------
Standard
Group
Canister
2.0
SHED 6.0
SHED 2.0
19
Table 8
Comparison of Indolene Evaporative Data
for High vs. Low Altitudes
Altd.
High
Low&Calif
Ratio*
High
Low
Ratio
High
Low
Ratio
1972-76
1972-77
1978-80
1978-80
1981
1981+
N
60
96
169
124
64
242
Mean
Mean Emissions (gms)
Mileage Hot
(miles) Soak Diurnal
10,170
17,500
13,100
14,100
6,800
37,821
14.07
8.27
1.70
4.26
1.62
2.63
1.36
1.97
0.69
17.15
8.98
6.77
2.67
2.54
2.81
2.85
* Ratio of high to low altitude mean emissions.
The ratios of high to low altitude Indolene fuel evaporative emissions
for the SHED standard of 2.0 grams/test may not be valid because of the
mileage difference between the high and low altitude data (6,800 versus
37,821 miles). For the other two standard groups, the mean mileages for
the two regions are similar, and the ratios of high versus low altitude
mean emissions are from 1.70 to 2.63, all considerably higher than the
altitude factor of 1.30 used for MOBILE2.
One reason the high to low altitude emission ratios are higher for
controlled vehicles may be that the canister has a capacity limit in
holding the fuel vapor, whether at low or high altitude. For example, in
an uncontrolled vehicle, 30 grams of fuel vapor may be produced by the
fuel tank during a diurnal at low altitude. At high altitude because of
a lower atmospheric pressure, the fuel tank may produce 39 grams of fuel
vapor — yielding a high to low altitude emission ratio of 1.30 (39 grams
divided by 30 grams). However, if the canister can only hold 25 grams of
fuel vapor before saturation in both low and high altitude regions, 5
grams of fuel vapor will be emitted at low altitude (30 grams minus 25
grams), while 14 grams will be emitted at high altitude (39 grams minus
25 grams). The resulting ratio of high to low altitude emissions is 2.8
(14 grams divided by 5 grams).
-------
20
For the years of 1972-76, high altitude data from testing 60 In-use
vehicles were available through EF programs. The average emission rates
of 14.07 grams for hot soak and 17.15 grams for diurnal should be used as
the Indolene rates to represent the emission rates at high altitude.
Since the evaporative emission standard was the same for both low and
high altitudes in 1977, MOBILE2 combined the high altitude data with the
low altitude data to derive average emission levels and used them for
both regions. No new data are available since MOBILE2, therefore the
MOBILE2 1977 rates should remain unchanged in MOBILES.
For the 1978-80 and 1981 model years, the altitude factor used to adjust
the low altitude rates to high altitude is 2.59, the average of hot soak
and diurnal ratios (from Table 8) from the 1978-80 samples. The average
2.59 of the diurnal and hot soak ratios was used rather than the separate
ratios for the sake of simplicity, since the two ratios were not
significantly different.
The average emission rates from 1978-80 high altitude vehicles could be
used directly for MOBILE3. However, this would result in an
inconsistency of lower evaporative emissions for 1978-80 years, in
comparison with the 1981 rates (4.26 grams for 1978-80 vs. 5.88 grams for
1981 on hot soak, and 6.77 grams for 1978-80 vs. 7.98 grams for 1981 on
diurnal). This would imply that the 1978-80 vehicles represent a better
level of control than the 1981 vehicles. To avoid this inconsistency,
the altitude factor of 2.59 is used for both 1978-80 and 1981 years.
Since there are very limited data available from high altitude for 1982
and 1983, (only four 1982 vehicles were tested, and there was no 1983
vehicles tested) the high altitude rates are those from the low altitude
multiplied by a ratio of the evaporative emission standards of 1.30 (SHED
2.6 grams/test standard for the high altitude vs. SHED 2.0 grams/test
standard for the low altitude). Since both regions have the same SHED
2.0 gram standard for years 1984 and later, low altitude hot soak rates
of 1.12 grams from the fuel-injected vehicles and 2.27 grams from the
carbureted vehicles are used for high altitude Indolene hot soak rates.
The low altitude diurnal rate of 3.08 grams from carbureted vehicles is
used for the high altitude diurnal emission level for all 1984 and later
year vehicles.
A summary of the high altitude Indolene evaporative emission rates is
given in Table 9. MOBILE2 rates are also listed for comparison
purposes. Note that for 1981 and later years, the Indolene hot soak
rates are derived from the high altitude hot soak emissions of fuel
injected and carbureted vehicles weighted by their sales projections.
The high altitude fuel injection sales are assumed to be the same as the
low altitude.
-------
21
Table 9
High Altitude LDGV Indolene Evaporative Emissions (gms)
MOBILE2
MYR
Group
Pre-1971
1971
1972-76
1977
1978-80
1981
1982
1983
1984
1985-86
1987-89
1990+
Ihdolene
Hot Soak
19.07
14.18
14.07
8.27
5.88
5.60
2.68
2.54
1.81
1.57
1.36
1.25
Rates
Diurnal
33.90
21.16
17.15
8.98
7.98
'7.98
4.00
4.00
3.08
3.08
3.08
3.08
Hot Soak
19.07
14.18
10.75
8.27
2.03 .
0.82
0.82
0.82
0.82
0.82
0.82
0.82
Diurnal
33.90
21.16
11.67
8.98
3.45
1.39
1.39
1.39
1.39
1.39
1.39
1.39
Note 1: 1972-76 model year emission rates are the average emissions from
60 in-use vehicles tested at Denver available through EF
programs.
Note 2: The 1978-80 Indolene emission rates are based on low altitude
1981+ emission rates from carbureted vehicles (there are no
fuel-injection sales, see Attachment 5), adjusted by a factor of
2.59, i.e., hot soak of 5.88 grams is from 2.27 grams multiplied
by 2.59, and diurnal of 7.98 grams is from 3.08 grams multiplied
by 2.59.
Note 3: The Indolene hot soak rates for 1981 and later are derived from
weighting the sales of fuel-injected and carbureted vehicles.
The Indolene diurnal emissions are based on the average from the
carbureted vehicles alone.
Note 4: For model years 1981 and 1982, an emission standard ratio of
1.30 is used to adjust the low altitude rates to be the high
altitude rates.
Note 5: For 1984 and later model years, the high altitude rates are the
same as the low altitude because their emission standards are
the same.
-------
22
3.2 Commercial Fuel Rates
Evaporative emissions are sensitive to the volatility of the test fuel.
Using a higher volatility test fuel results in higher evaporative
emissions. The Indolene fuel used for laboratory tests is of low
volatility (approximately 9.0 RVP) in comparison with commercial fuel
used by motor vehicles. It has also been noted that during recent years
the spread between the volatilities of Indolene and commercial fuels has
increased. Consequently, the MOBILE2 evaporative emission rates, which
were based on tests with Indolene fuel, are unrealistically low at this
time.
The EF programs since November of 1983 have been collecting evaporative
emission data on both Indolene and commercial fuels at Ann Arbor. An
unleaded summer grade gasoline with an average RVP of 11.5 has been
selected as the representative commercial fuel used in this EF testing.
All of the testing on this 11.5 RVP commercial fuel has been conducted in
Ann Arbor.
The commercial fuel data collected under FY83 and later EF programs are
from 2.0 gram SHED standard (1981 and later) vehicles. Evaporative
emission rates derived from these commercial fuel data were used in
MOBILES to represent 1981 and later model year vehicles. For pre-1981
years, MOBILE3 commercial fuel rates were estimated. Thus, the
discussions on commercial fuel evaporative emission rates for low
altitude vehicles are divided into two parts: 1981 and later model year
vehicles and pre-1981 vehicles.
3.2.1 1981+ Vehicles
Ampng the 115 untampered vehicles that were tested on both Indolene and
commercial fuels in the EF programs (indicated by parentheses in
Attachment 3), 53 are from high mileage carbureted vehicles and 62 from
relatively low mileage fuel-injected vehicles. A comparison of mean
evaporative emissions from the two fuels is given in Table 10.
Table 10
Evaporative Emissions
Indolene vs. Commercial Fuels
Hot Soak Diurnal
Mean (gm/test) Mean (gm/test)
Sample IJ Indol. Comm. Ratio Indol. Comm. Ratio
Carbureted 53 2.74 3.98 1.45 4.22 9.31 2.21
Fuel-Injected 62 1.32 1.55 1.38 2.21 3.13 1.42
-------
23
Note that there exists a large difference in commercial fuel diurnal
emissions between the carbureted and fuel-injected vehicles (9.31 grams
vs. 3.13 grams from Table 10). The discussions on the reasons for the
Indolene fuel diurnal emission differences between fuel-injected and
carbureted vehicles (section 3.1.1.3) are applicable here also.
Based upon the above data and the conclusions obtained from the Indolene
fuel data (Section 3.1.1.4), MOBILES has used the commercial fuel hot
soak emissions of 1.55 and 3.98 grams (from Table 10) for fuel-injected
and carbureted vehicles, respectively. These two rates are weighted
together by sales projections of fuel injected and carbureted vehicles to
derive overall MOBILE3 hot soak emission rates for 1981 and later years.
The fuel injection sales projections are presented in Attachment 5. The
commercial fuel diurnal emission rate of 9.31 grams from carbureted
sample are used directly to represent MOBILE3 diurnal rates for all 1981
and later year vehicles.
3.2.2 Pre-1981 Vehicles
The Indolene emission rates for the SHED 2.0 gram vehicles are used also
for the 6.0 gram SHED standard vehicles (model years 1978-80), as
discussed in Section 3.1.1.4. To be consistent with the Indolene rates,
the commercial fuel emission rates from the 2.0 gram SHED standard are
also used for 6.0 gram SHED standard vehicles.
To establish LDGV commercial fuel levels for pre-1978 evaporative
emission standards, the Test and Evaluation Branch of EPA initiated a
testing program to test six vehicles on both Indolene and commercial
fuels. The six vehicles represent the three pre-1978 evaporative
emission standard eras as:
Standard MYR Vehicles
None pre-1971 1963 Ford Galaxie
1970 Chrysler Newport
Canister 6.0 1971 1971 Ford Galaxie
1971 Ford LTD Wagon
Canister 2.0 1972-77 1974 Buick Century
1975 Chevrolet Nova
Each vehicle was tested at least twice with each fuel, with a total of 28
tests completed in early February of 1984. Five tests were not used in
calculating the commercial to Indolene fuel emission ratios because of
some technical problems detected after the data had been collected. Test
results are summarized in Table 11.
-------
24
Table 11
Evaporative Test Results
Indolene vs. Commercial Fuels
Evaporative
Emissions Standard
Hot Soak
Diurnal
None
Canister 6.0
Canister 2.0
None
Canister 6.0
Canister 2.0
MYR
pre-1971
1971
1972-77
pre-1971
1971
1972-77
Mean(gm/test)
MOBILE2
14.67
10.91
8.27
26.08
16.28
8.98
Indolene
16.03
16.40
10.38
12.09
5.18
4.04
Cotmn •
24.59
24.24
29.45
22.26
12.26
13.72
Ratio*
,53
,48
,84
1.84
2.37
3.40
* Ratio = Commercial fuel emissions divided by Indolene fuel emissions.
The Indolene emission rates obtained from the new tests are different
from those used in MOBILE2. The MOBILE2 rates for Indolene fuel were
obtained from in-use vehicles tested at Los Angeles, with a much larger
sample size and are, therefore, used in MOBILES. The MOBILES commercial
fuel evaporative rates for pre-1978 cars are estimated from MOBILE2
Indolene rates, multiplied by commercial to Indolene emission ratios
either from Table 11, or from Table 10.
The ratios of 1.53 for hot soak and 1.84 for diurnal (Table 11) developed
from pre-1971 test vehicles are used to adjust the pre-1971 Indolene
emissions to MOBILES commercial rates. The ratios of 1.48 for hot soak
and 2.37 for diurnal developed from 1971 test vehicles are used to adjust
the 1971 Indolene rates to MOBILES commercial fuel evaporative
emissions. These ratios appear to be reasonable in comparison with the
SHED 2.0 gram ratios (1.45 for hot soak and 2.21 for diurnal) from the
carbureted sample in Table 10.
The ratios for the 2.0 gram canister standard vehicles from Table 11
appear too high (2.84 on hot soak and 3.40 on diurnal) when compared with
all other ratios from precontrolled and controlled vehicles^ (for
example, 1.48 on hot soak and 2.37 on diurnal for the 6.0 gram canister
13
Tests on 1978 vehicles in St. Louis on both Indolene and commercial
fuels yielded a hot soak ratio of 1.54 and a diurnal ratio of 2.12-
These were not used, however, because of the uncertainty of the RVP
of the commercial fuel used.
-------
25
standard vehicles). The ratios of 1.49 for hot soak and 2.62 for diurnal
emissions are used to adjust the Indolene levels to commercial rates for
the 1972-77 vehicles. These ratios are derived from 32 in-use high
mileage 1981 carbureted vehicles. They were used in the draft version of
MOBILE3. They are used also in the final version of MOBILE3.
-------
3.3
26
Summary of LDGV Evaporative Emissions
Table 12 is a summary of the developed LDGV evaporative emission rates.
The low altitude Indolene rates are from Table 7. For pre-1978 model
years, the commercial fuel emissions are derived from emission ratios of
different fuels. For 1978 and later years, the commercial fuel hot soak
emissions are 1.55 grams for the fuel-injected vehicles and 3.98 grams
for the carbureted vehicles. The commercial fuel diurnal emissions for
1978 and later years are 9.31 grams.
The high altitude Indolene rates are from Table 9. The low altitude
emission ratios of different fuels are also used for the high altitude
for the pre-1978 vehicles. For 1978 and later years, the high altitude
commercial fuel emissions are derived from the low altitude commercial
rates. For example, an altitude factor of 2.59 is used for 1978-81, an
emission standard ratio of 1.30 is used for years 1982 and 1983, and the
high altitude rates are the same as the low altitude rates for 1984 and
later years, as discussed in Section 3.1.2.
Table 12
LDGV Evaporative Emission Rates (gms)
Hot Soak
Diurnal
Region
Low Altitude
MYR Group Indol. Ratio Comm. Indol. Ratio Comm.
pre-1971
1971
1972-77
1978-80
1981
1982
1983
1984
1985-86
1987-89
1990+
14.67
10.91
8.27
2.27
2.16
2.06
1.95
1.81
1.57
1.36
1.25
1.53
1.48
1.49
22.45
16.15
12.32
3.98
3.75
3.54
3.31
3.01
2.50
2.05
1.82
26.08 1.84
16.28 2.37
8.98 2.62
3.08
3.08
3.08
3.08
3.08
3.08
3.08
3.08
47.99
38.58
23.53
9.31
9.31
9.31
9.31
.9.31
9.31
9.31
9.31
High Altitude pre-1971
1971
1972-76
1977
1978-80
1981
1982
1983
1984
1985-86
1987-89
1990+
19.07
14.18
14.07
8.27
5.88
5.60
2.68
2.54
1.81
1.57
1 . 36
1.25
1.53
1.48
1.49
1.49
29.18
20.99
20.96
12.32
10.31
9.71
4.60
4.30
3.01
2.50
2.05
1.82
33.90
21.16
17.15
8.98
7.98
7.98
4.00
4.00
3.08
3.08
3.08
3.08
1.84
2.37
2.62
2.62
62.38
50.15
44.93
23.53
24.11
24.11
12.10
12.10
9.31
9.31
9.31
9.31
-------
27
Figure 1 (obtained from A DOE survey) shows the trends in fuel
volatility from 1959 to 1982. The RVP of the summer grade leaded
gasolines was around 9.0 Ibs until about 1972, and has been increasing
since 1972 to about 10.0 Ibs in 1980. The commercial fuel used in EF
programs was 11.5 RVP, the average RVP of summer grade unleaded gasolines
used around Detroit which is one of the gasoline volatility Class C
cities^ with summer temperature ranges between 60"F and 84"F. The
majority of non-California cities were classified as either Class B or
Class C in 1983. The average volatility levels of summer grade unleaded
gasolines used among the Class B cities such as Dallas are lower (about
10.0 RVP) yet those cities have summer temperature ranges between 65°F
and 95°F, which would produce similar amount of diurnal emissions as the
11.5 RVP commercial fuel because of their higher temperatures. Based on
this, EPA feels that the commercial fuel evaporative emission rates given
in Table 12 are appropriate for 1980 and later calendar years. Because
of the lower in-use fuel RVP for pre-1980 calendar years, these
commercial fuel evaporative emission rates may not be suitable for use
for those calendar years. But the pre-1980 calendar years are not of
main concern to most MOBILES users.
Figure 1
Trends in RVP, Leaded Gasoline
1959-82
Source: DOE, "Motor Gasolines"
14.0
1959'60'6 I '62 '6J '61 '65'66 '67 '68 '69 '70 '71 '72 '73 '71 '75 '76 '77 '78
80 '81 '62 '83
The American Society for Testing Materials (ASTM) recommends maximum
gasoline volatility levels for different cities based on their
summer temperatures. Warm cities such as Phoenix, Dallas have lower
recommended RVP maximums. The recommended maximum volatility levels
for summer fuel in 1983 are 9.0 in New Mexico (Class A), 10.0 in
Dallas (Class B), 11.5 in Detroit (Class C), and 13.5 in cities of
Alaska (Class D). These recommended fuel maximum volatility levels
are intended to prevent or reduce problems such as vapor lock in
vehicles.
-------
28
4.0 OTHER VEHICLE TYPES
LDGV evaporative emission rates obtained from the previous section are
used as the basis to derive evaporative emission rates for other gasoline
powered vehicles.
4.1 Light-Duty Gasoline Class 1 Trucks
Limited Indolene evaporative emission test data from LDGTls are available
from EF programs. The LDGT1 data are not significantly different from
the LDGV data, and the evaporative emission standards for these two
vehicle classes are exactly the same. Therefore, low altitude LDGV
evaporative emission rates on both the Indolene and commercial fuels are
also used for LDGTls. The only difference between these two vehicle
classes is the fuel injection sales projections. As can be seen from
Attachment 5, LDGT1 fuel injection sales were projected to begin in the
year of 1984 and continue at a different percentage until the year of
1988. Therefore, the LDGT1 hot soak rates between the years of 1981 and
1987 are slightly different from the LDGV rates.
The high altitude LDGV evaporative emission rates on both the Indolene
and commercial fuels for pre-1984 model years are also used for high
altitude LDGT1 emission rates. For 1984 and later years, the high
altitude LDGTls have a SHED 2.6 grams/test emission standard, while the
low altitude LDGTls have the same evaporative emission standard (SHED 2.0
grams/test) as the LDGVs. Therefore, in parallel to the LDGV
relationships the high altitude emission rates for 1984 and later trucks,
Indolene or commercial fuel, are the low altitude rates multiplied by an
emission standard ratio of 1.30.
Table 13 is a summary of the LDGT1 evaporative emission rates.
-------
29
Region
Low Altitude
Table 13
LDGT1 Evaporative Emission Rates (gms)
MYR Group
pre-1971
1971
1972-77
1978-80
1981-83
1984
1985
1986
1987
1988-89
19904-
Hot Soak
Indol.
14.67
10.91
8.27
2.27
2.27
2.09
1.90
1.72
1.56
1.36
1.25
Ratio Comm.
1.53 22.45
1.48 16.15
1.49 12.32
3.98
3.98
3.59
3.20
2.81
2.47
2.05
1.82
Diurnal
Indol. Ratio Comrn.
26.08
16.28
8.98
08
08
08
08
08
08
08
1.84
2.37
2.62
3.08
47.99
38.58
23.53
9.31
9.31
9.31
9.31
9.31
9.31
9.31
9.31
High Altitude pre-1971
1971
1972-76
1977
1978-81
1982-83
1984
1985
1986
1987
1988-89
1990+
19.07
14.18
14.07
8.27
5.88
5.88
2.71
2.47
2.23
2.02
1.76
1.62
1.53
1.48
1.49
1.49
-
29.18
20.99
20.96
12.32
10.31
10.31
4.67
4.16
3.65
3.21
2.67
2.37
33.90
21.16
17.15
8.98
7.98
4.00
4.00
4.00
4.00
4.00
4.00
4.00
1.84
2.37
2.62
2.62
-
62.38
50.15
44.93
23.53
24.11
12.10
12.10
12.10
12.10
12.10
12.10
12.10
4.2
Light-Duty Gasoline Class 2 Trucks
There are very few LDGT2 test data in the EF programs. For pre-1979
model years, LDGTs with gross vehicle weight (GVW) over 6,000 Ibs. were
classified as HDGVs. Therefore, the emission rates and standards for
LDGT2s for pre-1979 model years are the same as those for HDGVs. Since
LDGT2s have the same evaporative emission standards as LDGTls for 1979
and later years, the same emission rates are used.
Table 14 is a summary of the LDGT2 evaporative emission rates. Note that
for the hot soak rates, the LDGT2 fuel injection sales projections are
the same as the LDGTls. For high altitude 1984 and later vehicles, the
emission rates are the low altitude rates multiplied by an emission
standard ratio of 1.30.
-------
30
Table 14
LDGT2 Evaporative Emission Rates (gms)
Hot Soak
Diurnal
Region
Low Altitude
MYR Group Indol. Ratio Comm. Indol. Ratio Comtn.
pre-1979 18.08 1.53
1979-80 2.27
1981-83 2.27
1984 2.09
1985 1.90
1986 1.72
1987 1.56
1988-89 1.36
1990+ 1.25
High Altitude pre-1979 23.50 1.53
1979-81 5.88
1982-83 5.88
1984 2.71
1985 2.47
1986 2.23
1987 2.02
1988-89 1.76
1990+ 1.62
27.66
42.33 1.84
3.98
3.98
3.59
3.20
2.81
2.47
2.05
3.08
3.08
3.08
3.08
3.08
3.08
3.08
1.82
3.08
35.96
10.31
10.31
4.67
4.16
3.65
3.21
2.67
55.03
7.98
4.00
4.00
4.00
4.00
4.00
4.00
1.84
2.37
4.00
77.89
9.31
9.31
9.31
9.31
9.31
9.31
9.31
9.31
101.26
24.11
12.10
12.10
12.10
12.10
12.10
12.10
12.10
4.3
Heavy-Duty Gasoline Vehicles
MOBILE2 used 12.90 grams for hot soak emissions and 31.90 grams for
diurnal emissions for the evaporative emission rates of pre-1985
uncontrolled HDGVs (also pre-1979 LDGT2s) in the low altitude
non-California region. These emission rates were derived from nine
in-use heavy-duty vehicles. -^ There are no new in-use test data
available since MOBILE2. However, in comparison with LDGV Indolene
evaporative emission rates for the no-standard era (pre-1971), the hot
soak of 12.90 grams is lower than that of 14.67 grams from the LDGVs.
This seems to be unreasonable as HDGVs generally have larger carburetors
and experience more severe usage than the LDGVs.
15
Memo on "Quantity of Evaporative HC Emissions from Heavy-Duty
Vehicles" from Michael W. Leiferman to Marcia Williams, May 11, 1977.
-------
31
Test data submitted by GM for the HDGV evaporative emission
rulemakingl^ have shown an average full-SHED hot soak emissions of
18.08 grams and diurnal emissions of 42.33 grams based upon Indolene
fuel. These results were obtained from testing of two uncontrolled
vehicles, one a heavy-duty pick-up, and one C-7 series tractor. These
data were submitted with the intention to show a good correlation between
the GM proposed component test procedure and the EPA proposed "full-SHED"
test procedure for HDGV engine certifications. These emission rates
appear to be reasonable, in comparison with the LDGV rates from the
precontrolled period. The higher rate of hot soak emissions of HDGVs
(18.08 vs. 14.67 grams from LDGVs) can be attributed to their larger
carburetors and more variable operating temperatures. The higher rate of
diurnal emissions (42.33 vs. 26.08 grams from LDGVs) can be attributed to
the fact that HDGVs have larger fuel tanks. Therefore, these rates of
18.08 and 42.33 grams are used as the Indolene rates for low altitude
pre-1985 HDGVs and pre-1979 LDGT2s. The corresponding high altitude
Indolene emission rates are the low altitude Indolene emissions
multiplied by the altitude factor of 1.30.
The emission rates for 1985 and later model year vehicles for low and
high altitude (at SHED standards of 3.0/4.0 grams/test) are estimated
from LDGV emission rates for the SHED standard of 2.0 grams/test. The
HDGV emissions are calculated from the LDGV emissions and the ratio of
the HDGV standards to the LDGV standard. For example, since the HDGV 3.0
grams/test standard is a factor of 1.5 in comparison with the LDGV 2.0
grams/test standard, the assumption is that the ratio of the emissions
will be the same as the ratio of the standards. Therefore, the
evaporative emission rates for HDGV 3.0 gram standard are those for the
LDGVs multiplied by 1.5. Similarly, the rates for HDGVs 4.0 gram
standard are those for the LDGVs multiplied by 2.0. This methodology is
used to derive HDGV evaporative emission rates, for both Indolene and
commercial fuels. The derived two ratios are then weighted by their
respective GVW sales fractions for HDGV in 19871?, which are estimated
to be 81.5 percent for the under 14,000 Ib. weight classes and 18.5
percent for the over 14,000 Ib. weight classes. No fuel injection sales
are assumed for HDGVs. The resultant low altitude emission rates are
3.62 grams for hot soak and 4.90 grams for diurnal on the Indolene fuel,
and 6.34 grams for hot soak and 14.83 grams for diurnal on the commercial
fuel.
"Summary and Analysis of Comments to the Notice of Proposed
Rulemaking: Evaporative Emission Regulation and Test Procedure for
Gasoline-Fueled Heavy-Duty Vehicles," Standards Development and
Support Branch, US EPA Docket #OMSAPC-79-l.
"Historical and Projected Emissions Conversion Factor and Fuel
Economy for Heavy-Duty Trucks, 1962-2002," Energy and Environmental
Analysis, Inc., prepared for Motor Vehicle Manufacturers
Association, December 1983.
-------
32
The derived Indolene fuel evaporative emissions for 1985 and later HDGVs
under different emission standards (SHED 3.0, SHED 4.0, and SHED 3.0/4.0)
are listed in Table 15. Additional test data were available from the
HDGV evaporative emission rulemaking. These data were submitted by Ford
in support of a proposed 3.0 gram standard for HDGVs under 12,000 GVW
classes. With inertia test weight settings between 6,250 to 7,000 Ibs.,
which are equivalent to GVW's of 12,500 to 14,000 Ib. HDGVs, two F-350
vehicles with a total of 51 tests had an average total Indolene
evaporative emissions of 2.70 grams. (Note, from Table 15, that the
estimated total emissions for the 3.0 gram standard are 8.03 grams.) Two
C-700 vehicles, which were the representative vehicles for the over
14,000 Ib. HDGV classes, with 65 tests, had an average total evaporative
emissions of 3.70 grams. (The estimated total emissions for the 4.0 gram
standard from Table 15 are 10.70 grams.) These results show lower
emission levels in comparison with the estimated Indolene rates from
Table 15. However, since these results were submitted to be
representative of vehicles at certified levels, they are expected to be
lower than the in-use evaporative emission rates. Therefore, the derived
Indolene emission rates presented in Table 15 are considered to be
reasonable.
Table 15
Low Altitude Indolene Evaporative Emissions (gms)
for 1985 and Later* HDGVs
Standard
SHED 3.0
SHED 4.0
SHED 3.0/4.0
Hot Soak
3.41
4.54
3.62
Diurnal
4.62
6.16
4.90
Total
8.03
10.70
8.52
The proposed Indolene emission rates are those for the SHED 3.0/4.0
split standards.
-------
33
The MOBILES commercial fuel evaporative emission rates for pre-1985 HDGVs
are derived from the Indolene rates (for example, the low altitude 18.08
grams for hot soak and 42.33 grams for diurnal) and the commercial to
Indolene emission ratios of precontrolled LDGVs (1.53 for hot soak and
1.84 for diurnal). For 1985 and later years, the commercial fuel
evaporative emission rates for the HDGV 3.0/4.0 split standards are
derived the same way as the Indolene rates, that is, scaled from LDGV
emission rates by the ratio of HDGV and LDGV emission standards, then
weighted by the HDGV GVW sales fractions. Table 16 is a summary of the
HDGV evaporative emission rates. High altitude commercial fuel emission
rates are the low altitude rates multiplied by an altitude factor of 1.30.
Table 16
HDGV Evaporative Emission Rates (gms)
Hot Soak Diurnal
Region MYR Group Indol. Ratio Comm. Indol. Ratio Comm.
Low Altitude pre-1985 18.08 1.53 27.66 42.33 1.84 77.89
1985+ 3:62 6.34 4.90 14.83
High Altitude pre-1985 23.50 1.53 35.96 55.03 1.84 101.26
1985+ 4.71 8.24 6.38 19.28
-------
34
Attachment 1
Summary of Indolene Evaporative Data Base*
Number of Vehicles Tested
(Including Tampered Vehicles)
Model Low Altitude High Altitude
Year Non-California California Non-California
Pre-1970 - 102
1970 13
1971 - 21 21
1972 - 20 20
1973 - 20 20
1975 - 20 20
1977 36 - 32
1978 49 50(50) 49
1979 75 49(49) 75
1980 - 100(61) 45
1981 175 157(43) 64
1982 25 - 4
1983 47 -
Total 407 552(203) 350
Data obtained from CARB are indicated in parentheses.
-------
35
Attachment 2
List of Tampered Vehicles
Indolene
Mileage Emissions (gms)
Tampering
Veh ID* MYR
Federal Region
4055 1981
MFC
GM
(Miles)
67,
375
Hot Soak Diurnal
1
.98
0
.70
Characteristic
Misrouted vacuum line
& missing gas cap
4086
4095
4143
4315
1981
1981
1981
1983
Ford
Ford
Toyota
Ford
74,
84,
63,
19,
475
454
747
286
2
10
9
0
.44
.89
.91
.36
4
10
1
7
.97
.73
.88
.82
Disconnected
line
vacuum
Missing canister
Disconnected
line
Disconnected
line
vacuum
vacuum
California Region
204
High
158
188
220
16
1980
Ford
5,
458
0
.94
14
.28
Disconnected
line
vacuum
Altitude Region
1980
1980
1980
1981
GM
GM
Toyota
GM
14,
12,
5,
6,
577
418
920
474
11
3
2
3
.37
.00
.70
.33
3
4
6
1
.75
.77
.29
.19
Disconnected
line
Disconnected
line
Disconnected
line
Disconnected
vacuum
vacuum
vacuum
vacuum
line
* Only one Federal vehicle (#4315) was a fuel-injected vehicle, all
others were carbureted vehicles.
-------
36
Attachment 3
Low Altitude 1981 and Later LDGV Sample Distributions*
(Untampered Vehicles)
MYR Carbureted Fuel-Injected
1981 168(43) 3(2)
1982 0 25(25)
1983 11(10) 35(35)
Total 179(53) 63(62)
* Vehicles tested with both Indolene and commercial fuels are indicated
by parentheses.
-------
37
Attachment 4
Evaporative Emission Control System Diagnosis
Number of Vehicles
Problem Carbureted Fuel-Injected
Canister saturated with fuel* 18 3
Canister broken 1 0
Canister filter dirty 6 0
Purge valve disconnected 1 0
Purge valve leaked vacuum 5 0
Purge valve was sticking 10 0
Vacuum line plugged 1 0
Vacuum line damaged 3 0
Vent line damaged 1 1
Non-OEM gas cap 1 1
Inoperative vacuum control valve 7 0
Overall** 40 5
* Among these vehicles with saturated canisters, 20 were found to
have purge control problems, such as sticking purge valves, etc-
** Vehicles that had one or more problems.
-------
38
Attachment 5
Fuel Injection Projections
To account for different hot soak emission rates due to vehicles' fuel
delivery types, it is necessary to use sales projections on fuel
injection for LDGVs and LDGTs. The following is a summary of the fuel
injection projections with all manufacturers combined.
VEH TYPE
LDGVs
LDGTls & LDGTZs
MYR Group
1981
1982
1983
1984
1985-86
1987-89
1990+
1981-83
1984
1985
1986
1987
1988-89
19901-
Percent of
Fuel Injection
Projection
9.4
18.3
27.6
40.0
61.0
79.5
88.8
0.0
16.0
32.0
48.0
62.0
79.5
88.8
The LDGV projections-^ are obtained from separate estimations on 1982,
1983, 1984, 1985, 1987, and 1990 model years. The sales projections for
LDGTls and LDGT2s combined^ are obtained from years 1983, 1987, and
1995, with the growth between the years of 1984 and 1987 assumed to be
linear, and the same projections as LDGVs assumed for 1988 and later
years.
Letter from Energy and Environmental Analysis, Inc. to Phil Lorang
of Technical Support Staff, November 28, 1983.
Dana Jones and LeRoy H. Lindgren, "Automotive Technological
Projections Based on U.S.A. Energy Conservation Policies," December
17, 1983.
------- |