EPA-AA-TSS-85- 5
Technical Report
Economic Commission for Europe
Inland Transport Committee
Group of Experts on the Construction of Vehicles
Group of Rapporteurs on Pollution and Energy (GRPE)
Emissions of In-Use Vehicles: Update
on the U.S. Experience
By
Philip A. Lorang
September 1984
NOTICE
Technical Reports do not necessarily represent final EPA
decisions or positions. They are intended to present
technical analysis of issues using data which are
currently available. The purpose in the release of such
reports • is to facilitate the exchange of technical
information and to inform the public of technical
developments which may form the basis \for a final EPA
decision, position or regulatory action. • T\
Technical Support Staff
Emission Control Technology Division
Office of Mobile Sources
Office of Air and Radiation
U. S. Environmental Protection Agency
-------
1. INTRODUCTION
This paper is an update to a paper transmitted for discussion
at the 8th Session of GRPE, August-September 1983, entitled
"New Vehicle Regulation and Emissions of In-Use Vehicles:
The U.S. Experience." That paper emphasized the results
which new vehicle regulations in the U.S. have achieved in
terms of emissions from vehicles in general use by ordinary
motorists.
At the time the previous paper was prepared, the U.S.
Environmental Protection Agency (EPA) was reluctant to draw
conclusions regarding in-use emissions of hydrocarbons (HC)
and carbon monoxide (CO) of 1981 and newer vehicles, because
the available data were limited to low mileage vehicles. In
the interim, additional data have been obtained on higher
mileage vehicles in this category. This paper presents
information on these vehicles and describes conclusions EPA
has reached regarding the likely HC and CO emissions of these
vehicles in the future. EPA has also revised its assessment
of the oxides of nitrogen (NOx) emissions from these
vehicles. The new assessment is described.
EPA has also adopted a new conceptual approach to assessment
of emissions of in-use vehicles. The approach seeks to
isolate and better quantify the effects of tampering to major
emission control subsystems and of the improper introduction
of leaded fuel into catalyst-equipped vehicles". ("Tampering"
refers to the removal, disconnection, or other interference
with the proper operation of a subsystem. Examples are
removal of a catalytic converter or disconnection of a fuel
vapor hose.) The new approach and some of its major results
are described. Further observations are also made on the
phenomena of catalyst tampering and improper use of leaded
gasoline, which may be of interest to countries contemplating
more stringent emission standards and fuel lead content
regulations.
2. TESTING AND ANALYSIS OF 1981 AND NEWER VEHICLES
EPA conducts a continuing In-Use Emission Factor Surveillance
Program which borrows vehicles from their owners and tests
them with the standard U.S. emission test procedure. Because
of the 50% reduction in allowable emissions of NOx in 1981
and the resulting major change in emission control technology
(introduction of three-way catalytic converters,and computer
controlled fuel mangement), 1981 and newer vehicles have been
a subject of intense interest to EPA. The surveillance
program began testing 1981 vehicles as soon as they were in
the hands of owners. After an -initial period of recruiting
vehicles without respect to accumulated service, EPA began to
solicit and test vehicles which had accumulated many more
miles of service for their chronological age than typical for
-------
-2-
passenger cars in the U.S. This was done to learn more about
the durability and emissions of 1981 vehicles at advanced
mileage than would have been possible to learn otherwise
except by the passage of time. In the U.S., the average age
of a passenger car is about 58,000 miles when weighted for
vehicle use, and EPA was concerned that emissions at that age
and above might be much different than observed from younger
vehicles. EPA has also been testing 1982 and 1983 vehicles,
necessarily at lower mileages. Table 1 displays some
information about the numbers and mileages of vehicles tested
to date.
i
EPA has most recently completed a test program phase which
emphasized 1983 vehicles. EPA selected for testing those
1983 models which represent emission control technologies
that EPA and industry sources predict will dominate the U.S.
fleet in future model years, especially new fuel injection
applications.
Table 2 and Figure 1 show average emissions of the 1981,
1982, and 1983 vehicles tested by EPA and their average
accumulated mileage. CO emission data for the 1981 and 1982
vehicles are separated into one group which were certified to
meet the statutory CO emission standard of 3.4 gram per mile
and another group which received a temporary - waiver to a CO
level of 7.0 gram per mile under another provision of the
U.S. Clean Air Act.
When emission levels of individual model years are graphed
with the corresponding accumulated mileages as in Figure 1,
an increasing trend in emissions with mileage is apparent.
There are three reasons why this apparent trend should not be
taken too quickly as the true trend of emissions of all
in-use 1981 and newer vehicles:
1. EPA has confirmed that the sample of test vehicles
used in the surveillance program described here
contains fewer cases of tampering and improper use
of leaded gasoline than does the corresponding
vehicle.population as a whole. This underrepresen-
tation is caused by the reluctance of owners of
such vehicles to lend them to EPA for testing.
V
2. The 1981, 1982, and 1983 vehicles tested by EPA do
not represent exactly the various mixes of
emission control system types that will be sold in
later model years.
3. It is more accurate to examine the relationship of
emissions to mileage within each model year, than
to simply compare model years as in Figure 1.
-------
-3-
For these reasons, EPA has used another method to estimate
in-use emissions of 1981 and newer vehicles. The method
first estimates the emissions of vehicles that have not been
subjected to certain severe forms of tampering or to improper
use of leaded gasoline. For this purpose, EPA believes the
data from the surveillance program are adequate. Second, an
estimate of the additional or excess emissions caused by the
previously excluded severe forms of tampering and by the
improper use of leaded gasoline are estimated and added to
the first value. These two steps will be briefly described
and the results presented in the next two sections.
3. EMISSIONS OF VEHICLES NOT SUBJECT TO TAMPERING OR
IMPROPER USE OF LEADED FUEL
Estimating the emissions from current technology vehicles is
relatively straight forward when adequate in-use test data
are available. Typically a trend in the emission levels of
the vehicles versus age (usually measured in accumulated
mileage) can be determined and used to predict the emission
levels of in-use vehicles. When in-use data are inadequate
or have only a small mileage range a modeling approach
becomes necessary to predict emission levels.
An important consideration in estimating the emissions of
future model years is the type of technology used to control
emission levels. Starting in 1981, for example, most
passenger vehicles produced in the U.S. began using feed back
control or "closed-loop" systems to control the air/fuel
mixture. At the same time EPA parameter control regulations
limited the in-use adjustability of certain engine
parameters. Both of these changes have had significant
effects on the in-use emission performance of the newer model
years.
A major new trend in technology that is likely to have a
significant effect on in-use emission levels is the
increasing use of fuel injection. As can be seen in Figure
2, EPA estimates that use of fuel injection in future model
years will approach 90% of passenger cars by"1990. Figure 3
shows that changes in catalyst type will accompany the-trend
towards fuel injection, with fewer vehicles receiving an
oxidation catalyst in addition to the three-way catalyst.
Fuel injected vehicles can be expected on engineering grounds
to have different emission characteristics'' than the
carbureted vehicles common today. These differences include
lower evaporative emissions due to the elimination of
carburetor float bowls, lower cold start emissions due to
elimination of mechanical and electrical problems of the
choke mechanism, and fewer'instances of catastrophic emission
control failure which still allows the vehicle to be driven.
-------
-4-
EPA has used a simple regression of the emission ' levels
versus mileage for all of the pre-1981 model years. The
emissions data set available from these vehicles is of
sufficient size and mileage range to produce acceptable
estimates. However, for 1981 and later model years, the
significant changes in technology and the lack of emission
data over a wide mileage range have caused EPA to develop a
model to account for all of the factors EPA believes will
affect the emission behavior of these vehicles.
For the analysis, all obviously tampered or misfueled
vehicles in the data base were removed. EPA believes that
tampered and misfueled vehicles are not properly represented
in the vehicle 'test samples since the testing program is
voluntary. The excess emissions in the fleet caused by
tampering and misfueling are calculated separately. This
calculation is described in Section 4.0.
Since technology trends show that the mix of technologies is
changing rapidly, the existing emission data set from 1981
and newer vehicles is divided into four ' separate
technologies. These are:
--Closed-Loop Carbureted with Any 3-way Catalyst
—Closed-Loop Fuel Injected with Any 3-way Catalyst
--All Open-Loop with Any 3-way Catalyst
—All Oxidation Catalyst Only
The results from the analysis of each of these technologies
can then be weighted together to represent any predicted mix
of the technologies. The explanation following concerns the
treatment of a single technology.
For purposes of predicting HC and CO emissions, the vehicles
from one technology are divided into three strata based on
emission levels. The lowest stratum contains the large
majority of the vehicles. It was defined to include all
vehicles emitting less than 1.5 gram/mile of the HC (compared
to. the 0.41 allowable level) and less than 20 gram/mile of CO
(compared to the 3.4 or 7.0 allowable level). Despite the
rather high limits, the vast majority of cars in this stratum
were found to emit very close to their allowable levels. The
average emissions of cars in this stratum as a .function of
mileage was estimated by simple least-squares' regression.
The next, or middle, stratum was defined as including all
remaining vehicles except for four vehicles with extremely
high emission levels. This stratum was kept separate from
the previous stratum for two reasons. First, vehicles in the
lower group generally passed the type of emission short test
used in the U.S. for annual inspections of in-use vehicles.
-------
-5-
Vehicles in this middle group commonly failed these short
tests. Keeping the two strata separate allowed the
subsequent estimation of the benefits of annual inspection
programs. Second, to use just one stratum encompassing both
groups would be to assume that the available test sample
represented the relative frequency of occurrence of these two
vehicle types at all mileages. EPA's hypothesis was that to
the contrary the relative sizes of these two groups would
vary with mileage, with the middle or malfunctioning group
accelerating in size at the expense of the lower, properly
functioning group. In fact, this hypothesis was verified by
analysis of the available data in the range of zero to about
65,000 miles, and EPA extrapolated the pattern in a
piece-wise linear manner to higher mileages. The average
emission level of vehicles in the middle stratum was assumed
to be a constant independent of mileage, equal to the average
of the vehicles tested.
The third, or upper stratum, consisted of only four
vehicles. These vehicles had emission levels much higher
than those of the next highest emitting vehicles, and were
clearly in a different class of malperformance. EPA chose to
assume that the number of such vehicles would increase with
mileage but at a comparatively slower rate. The average
emission level of the tested vehicles was assumed to apply
independently of mileage.
There were no fuel injected vehicles tested which were in the
upper stratum and only a single fuel injected vehicle in the
middle stratum. However, all the fuel injected vehicles were
still at low mileage, and EPA was unwilling to assume that
the absence of higher emitting vehicles would continue
indefinitely. Therefore, EPA estimated the population of the
middle and upper strata for the fuel injection technology by
reasoning backwards from observation of carbureted vehicles.
That is, fuel injected vehicles were assumed to be like
carbureted vehicles except that carbureted vehicles were
eliminated for this purpose if they had malperformances which
engineering analysis showed could not occur with fuel
injection.
Once the sizes and average HC and CO emission levels of all
three strata for one technology were estimated . from the
available data, they were arithmetically combined,A in a simple
fashion to arrive at a prediction for the average emissions
of all in-use vehicles of that technology (excluding the
effects of tampering and improper use of leaded fuel). This
average depends on mileage, of course.
For NOx, there appeared to be no advantage to separation into
strata, and each technology was therefore represented by a
single regression line of NOx emissions versus mileage.'- • •
-------
-6-
Once emission levels of the four technologies have been
estimated, they are weighted together for each model year
using historical and forecasted sales fractions. The result
is the overall estimate of in-use emissions from all vehicles
sold in the U.S. in a given model year, but without the
effects of tampering and misfueling. Figure 4 shows the
results for selected model years. As can be seen, the shifts
in technology mix have a small but discernable impact on
predicted emissions.
4. EXCESS EMISSIONS DUE TO TAMPERING AND IMPROPER USE OF
LEADED FUEL
The estimates of emissions presented thus far were prepared
by first removing vehicles from EPA's data base which had
evidence of certain types of tampering to the emission
control system or of improper use of leaded fuel.
Specifically, vehicles with evidence of catalyst removal,
interference with the positive crankcase ventilation system
(PCV) causing release of crankcase vapors, disconnection of
the cannister used to store evaporative emissions,
disconnection or removal of the air pump system, or habitual
misfueling were removed. (Habitual misfueling means improper
use of leaded fuel over a long enough period to have
permanent and substantial adverse impact on the catalyst.)
While there were some such vehicles in EPA's owner-loaned
test sample, there were fewer than EPA has reason to believe
are present in the U.S. fleet as a whole. Vehicles with only
other forms of tampering such as electrical wire
disconnections were retained, as these problems are more
likely to have been inadvertant and owner reluctance to lend
vehicles to EPA should be much less.
Having removed these vehicles from the data base before
calculating the results displayed in Figure 4, EPA needed to
add back to the emission estimates the additional or excess
emissions caused by the true incidences of tampering and
misfueling in the U.S. fleet.
Figure 5 shows EPA's estimates of the incidence or rate of
each of the six types of tampering/misf ueling in the U.S.
fleet. These estimates are derived from observations of
vehicles in a survey performed by EPA in '-approximately 10
cities per year. In this survey, vehicles';rare ' randomly
selected from the traffic flow and ordered to pull to the
roadside, where technicians examine the vehicles for
tampering and misfueling. Under. U.S. law, owners may decline
this inspection, but in practice virtually all consent to it
so the representativeness of the sample is believed to be
good.
-------
-7-
Figure 5 shows that tampering and misfueling are more
frequent among vehicles which have accumulated more miles,
i.e., those that are older. Since age and year of
manufacture are related, it is .not entirely certain which
factor is the more important. U.S. automobile companies have
stated their belief that progressively better vehicle designs
will be less subject to tampering and misfueling. At
present, EPA assumes that age is responsible for all of the
effect shown in Figure 5. Figure 5 also shows that
light-duty trucks are much more frequently tampered than
passenger cars of the same age. Although used primarily for
personal transportation, the same as passenger cars, the
popular wisdom in the U.S. is that light trucks are commonly '
owned by persons less willing to accept government
restrictions. Because of the small number of light-duty
trucks in the survey sample, a separate regression on mileage
was not considered reliable.
The rates of catalyst removal and habitual misfueling shown
in Figure 5 are unfortunately high. At the average age of
catalyst passenger vehicles, they are about 7 percent and 13
percent, respectively. About one-half of the vehicles from
which catalysts have been removed have also been misfueled,
so the combined rate is 16 to 17 percent. Vehicle owners who
intend to misfuel may remove their catalysts believing that
their vehicles will be further damaged if they do not remove
them. However, these owners are probably misinformed or
overly cautious. EPA has no reason to believe that any
significant number of catalysts are being plugged because of
lead residues, even though most misfueled vehicles have their
catalysts present throughout the misfueling experience.
The rates of air pump disablement and EGR system disablement
are also high. Roughly one-half of U.S. vehicles are not
equipped with air pumps, so the rate in the U.S. fleet as a
whole is about one-half that shown in the figure. Virtually
all U.S vehicles have EGR, but fortunately nearly all U.S.
cities have acceptable nitrogen dioxide air quality levels
despite the high rate of EGR tampering. The rates of PCV and
evaporative cannister disablement are very low, which is
expected since the- nonparasitic nature of these systems is
such that owners have no reason to expect any fuel economy or
performance inprovement from disabling them. v
• *-\
EPA has found that vehicles with one type of tampering often
have multiple types. Therefore, the number of vehicles with
some form of tampering is less than the sum of the separate
rates in Figure 5. This pattern has been accounted for in
EPA's calculations and is reflected in subsequent figures.
Figure 6 shows EPA's estimates of the emissions increases
that occur with each form of tampering/misfueling.'' < -(THC
-------
-8-
denotes total .hydrocarbons including nonreactive methane.
Data not presented show that nearly all of the increase shown
for total hydrocarbons is in species other than methane.)
These estimates are developed from testing vehicles in a
properly operating condition and then again with tampering or
misfueling present. Figure 6 is for closed-loop three-way
catalyst vehicles. The earlier oxidation-catalyst vehicles
tend to have larger effects from catalyst removal and
misfueling, but these are not shown.
Figure 7 shows the net effect of all the forms of tampering
and misfueling combined, with correct accounting for multiple
instances on one vehicle. Since tampering rates for
individual components increase sharply with vehicle mileage,
so does the overall excess for each pollutant. These
excesses must be added to the emissions shown in Figure 4 to
give total in-use emissions.
Given the high rates of catalyst removal and misfueling as
shown in Figure 5 and their large impact on emissions as
shown in Figure 6, one would expect them to be a large part
of the total excess shown in Figure 7 as being due to all
tampering and misfueling combined. This is the case. Figure
8 shows the percentage contributions of individual tampering
types to the excess emissions at 50,000 miles. The
contributions are nearly the same at other mileages.
Because catalyst removal and misfueling have been recognized
by EPA as important concerns for some years, EPA has sought
to determine why they happen. EPA has questioned vehicle
owners, and EPA pays attention to statements made in the
automotive repair and hobby press. The most important reason
is the price difference between leaded and unleaded fuel in
the U.S., which has usually been in the range of U.S.
$0.04-0.06 per U.S. gallon. The immediate savings from use
of leaded fuel are obvious to price-conscious buyers, and
most owners are less aware of the fact that maintenance costs
increase with use of leaded fuel. Other countries would be
well advised to consider ways of avoiding the incentive of a
large price difference in favor of leaded fuel.
Performance is the next reason motivating vehicle owners to
misfuel. Some, vehicles have octane requirements in excess of
the octane provided by regular unleaded fuel'; These are the
only vehicles which should be expected to . give improved
performance on leaded fuel, which has higher octane in the
U.S. than unleaded. (Even owners of these vehicles are
purchasing leaded fuel for price reasons, since a premium
grade of unleaded fuel is widely available in the U.S.)
However, many owners seem to believe they get better
performance from their vehicles when using leaded fuel even
-------
-9-
when knock is not present when operating on unleaded fuel.
Many also believe that leaded gasoline contains more energy
per gallon and that fuel economy is improved with leaded
gasoline.
Misfueling tends to become habitual. About one-half of
leaded fuel purchased for misfueling is purchased by owners
who misfuel over 90% of the time. These owners comprise
about 4 percent of all vehicle owners. A larger number of
owners misfuel less often.
5. NET EMISSIONS FROM IN-USE 1981 AND NEWER VEHICLES
The U.S. paper at the 8th session included illustrations that
showed how the progressively more stringent new car standards
in the U.S. have achieved lower and lower emissions from
passenger vehicles in service. These figures have been
updated and are presented here as Figures 9, 10 and 11 for
HC, CO, and NOx emissions respectively. The lines in these
figures are total emissions including the effects of
tampering and misfueling.
The figures show that succeedingly more stringent emission
standards have achieved progressively better in-use
performance. In addition, the evolution of technology
following stabilization of the HC and NOx standards in 1981
and of the CO standard in 1983 is predicted to lead to still
lower emissions, although only slightly.
-------
-10-
Table 1
1981 and Newer In-Use Vehicles Tested by EPA*
Model
Year
1981
1982
1983
Number of
Vehicles
Tested
852
143
117
Model
Year N
1981 408
444
852
1982 28
115
143
Distribution of
Mean
Mileage
28,506
9,763
17,771
Emissions
CO
Standard
3.4
7.0
Both
3.4
7.0
Both
0-25,000
475
129
92
Table
Vehicles
By Mileage
25,000-50,000 50,000-75,000 75,000
202
10
25
2
144 - 31
3 1
0 0
of 1981 and Newer Vehicles*
Mean
Mileage
30,168
26,978
28,506
5,880
10,709
9,763
Mean Emissions (gram per mile)
HC
0.60
0.63
0.62
0.27
0.31
0.30
CO
7.37
8.67
8.04
2.51
4.03
3.73
NOx
0.84
0.85
0.84
0.60
0.71
0.69
1983
117
3.4
17,771
0.37
3.46
0.73
*As of August 10, 1984
-------
FIGURE lv
Light-Duty Passenger Vehicle Emissions
Exhaust Hydrocarbon Emissions
too
0.79
0.90
0.29
0.00
0.0
10.0
i 1.0-
8
fc
4.0
2.0-
0.0
0.0
ISO
t2S
too
0.79
0.50
0.29
0.00-
1M1
IMS
IM2
10.0
20.0 30.0 40.0
In Thousands Of MD«s
Carbon Monoxide Emissions
IMI
•
IMI
"MY"
•
IM2
IMI
10.0 20.0 JO.O 40.0
Vshldt MIlMg* In Thousands Of Mites
Oxides Of Nitrogen Emissions
IM1
IMS
IM2
0-0 10.0 20.0 30.0
Vthlcte MtkMo* In Thousands Of Mfl*s
0.41 MC STDl
JO.O
^^|».4 CO STOl
90.0
•M.O HOl »TO|
90.0
-------
-12-
PERCENT
100
FIGURE 2,
Fuel System
MULTi-POINT
FUEL INJECTION
THROTTLE-BODY
FUEL INJECTION
Closed Loop Carburetor
1982
1989 199Q
-------
-13-
FIGURE3.
Catalyst System
PERCENT
100'
80-
60-
40-
3-Way Catalyst
20-
1982
1 i
1983 1984 1985 1986
1987 1988 1989 1990
-------
FIGURE 4,
In-Use Emission Levels of 1981 & Newer Vehicles
Not Subject to Tampering or Improper Use of Leaded Fuel
1 (By Model Year)
0.75-
0.50-
0.25
0.00
0.0
Exhaust Hydrocarbon Emissions
MJ.O 20.0 30.0 40.0
Vehicle Mileage In Thousands Of Miles
MODEL YR
1981
t»82
1983
1984-1986 _
1987 «t Lfl*»f
SO.O
Carbon Monoxide Emissions
10.0 20.0 30.0 40.0
Vehicle Mileage In Thousands Of Miles
SO.O
1.25-
1.00-
0.75-
0.50-
0.25-
Oxides Of Nitrogen Emissions
0.0
10.0 20.0 30.0 «0.0
Vehicle Mileage In Thousands Cf Miles
90.0
-------
FIGURE 5,
Rates of Tampering And Improper Use of Leaded Fuel
Rate of Air Pump Disablements
-. Versus Mileage
o" i " i i « » • * • * *
Vihide Mileage In 10.000 Mil* Increments
Rate of Catalyst Removal
Versus Mileage
OI2J4S»7S»10
Vehicle Mileage In K>.000 Mil* Increments
Rate of EGR System Disablement
Versus Mileage
Rate of Habitual MisfueTmg
Versus Mileage
Vehicle Mileage In 10.000 Mile hcrements
OI234»*?8tM
Vehicle Mileage In 10.000 Mile Increments
Rate of Evaporative Canister Disablement
Versus Mileage
21
20
•5 "^
Vehici* Milage In 10,000 Mile hcremenrs
20-
•5 "1
5-
Rate of PCV System Disablement
Versus Mileage
OI2J4S«7»»IO
Vehicle Mileogvln 10.000 Mile kicrements
-------
FIGURE 6,
3.0
JO)
3 2.5- —
E 2.0-
6
c
.0 1.5- —
u>
*E
Ul
x 1.0 H
o
O 0.5
0.0
Average Effects of Tampering And Improper Use of
Leaded Fuel On A 1981 or Later 3WC Vehicle
PCV Evaporative Air
System Canister Pump
Emissions
GZ3 THC
C3 CO
ca NOX
Improper Use Catalyst CGR
Of Leaded Re'moval System
Fuel
- -20.0
30.0
-25.0
0)
O
-15.0
C
o
-10.0 -F
O
o
-5.0
0.0
-------
FIGURE 7,
Excess Emission Levels for 1981 & Newer Vehicles
Due to Tampering And Improper Use of Leaded Fuel
Excess Total Hydrocarbon Emissions
0.5
0.*
0.3
i"
10 20 30 40 90
Vehicle Mileage In Thousands Of Miles
Excess Carbon Monoxide Emissions
J-
8
t '
<0 20 30 40
Vehicle Mileage In Thousands Of Miles
Excess Oxides Of Nitrogen Emissions
0.5
0.4
0.3-
6*
i
0.2-
0.0
W 20 30 40
Vehicle Mileage In Thousands Of Miles
90
-------
FIGURE 8,
Excess Emission Levels for 1981 & Newer Vehicles
Due to Tampering And Improper Use of Leaded Fuel
Total Hydrocarbons
Carbon Monoxide
Oxides of Nitrogen
-------
-19-
15
o>
Q.
to
2
o
to
c
o
'to
fto
LJ
O
10-
5-
Pre-Control
1972-:
1975-1979
10 20 30 40 SO 60 70 80 90
Vehicle Mileage In Thousands Of Miles
K
Figure 9 - Hydrocarbon (HC) Emissions
of In-Use Passenger Vehicles Produced
Under U.S. Emissions Standards
1980
1981
1983
1990
100
-------
-20-
110
100-
90-
5 80 H
E
o
O
O
70-
60-
c
•§ 50 H
40-
30-
20-
10-
.990
Pre-Gontrol
1972-1974
10 20 30 40 50 60 70 80 90
Vehicle Mileage In Thousands Of Miles
1981
1983
1990
1980
100
Figure 10 - Carbon Monoxide (CO) .Emissions of
In-Use Passenger Vehicles Produced Under U.S.
Emissions Standards
-------
-21-
3-
s
o
w
C 2
O
.
UJ
X
o
1-
Pre-Control
1973-197
1975-1976
10 20 30 40 50 60 70 80
Vehicle Mileage In Thousands Of Miles
90 100
Figure 11 - Oxides of Nitrogen (NOx) Emissions
of In-Use Passenger Vehicles Produced,.,JJnder
U.S. Emissions Standards
------- |