EPA/AA/CTAB/TA/81-7
TECHNICAL REPORT
Carbon Monoxide and Non-FTP
Ambient Temperature
by
Robert I. Bruetsch
February, 1981
NOTICE
Technical reports do not necessarily represent final EPA decisions or
positions. They are intended to present technical analyses 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.
Control Technology Assessment and Characterization Branch
Emission Control Technology Division
Office of Mobile Source Air Pollution Control
Office of Air, Noise and Radiation
U.S. Environmental Protection Agency
2565 Plymouth Road
Ann Arbor, Michigan 48105
-------
TABLE OF CONTENTS
Page
I. Introduction ....................... 3
II. Conclusions ....................... 5
\
III. Summary of Significant Background Studies ........ 8
IV. Investigation of Data From the 1975 Through 1977
Vehicle Fleet ........................ 23
V. Discussion of Current Data
VI. References ........................ 69
VII. Appendix ......................... 73
-------
Carbon Monoxide and Non-FTP Ambient Temperature
I. INTRODUCTION
In recent years, increasing emphasis has been placed on the importance
of motor vehicle gaseous emissions under non-FTP conditions. Of
special interest, particularly in this report, are the effects of
ambient temperature conditions on carbon monoxide (CO) emissions from
motor vehicles. Under the current Federal Test Procedure (FTP),
carbon monoxide vehicle exhaust emissions are required to be below a
certain level in the temperature range of 68 °F to 86°F (20°C to
30°C). At ambient temperatures outside this range, particularly
colder temperatures, carbon monoxide emissions from mobile sources
increase dramatically. The pollutants given off during the first few
minutes of cold start operation, when the vehicles control system is
wanning up, are known as "cold start emissions" and can account for
the vast majority of all carbon monoxide emitted throughout the whole
vehicle trip. These emissions may be a major factor causing National
Ambient Air Quality Standard (NAAQS) violations for CO in cold weather
conditions.
Ambient carbon monoxide emission levels have shown marked improvement
in the counties monitored, with the greatest change evident in the
East. Nevertheless, violation frequencies are still substantial,
averaging from 40 to 90 days per year in the western areas studied,
and from 20 to 60 days per year for the central and eastern U.S.
counties [41]*.
*Bracketed numbers indicate reference numbers as listed in Section VT
of this document.
-------
-4-
The current MAAQS levels for CO are 9 ppm averaged over 8 hours and 35
ppm for 1 hour, neither to be exceeded more than once per year. The
draft revised EPA criteria document on the health effects of CO [43]
suggests that persons with stable angina (the most common form of
cardiovascular disease) are probably the most sensitive population.
Such persons may show adverse effects from exposure to 15 to 18 ppm CO
for 8 hours. Other studies suggest that there may be no threshold
concentration where stable angina patients would not be affected.
Fetuses, pregnant women, and anemics may also be sensitive groups to
CO exposure. However, the CO levels and frequency of exposure at
which such effects would be considered significant have not been
determined. The statutory CO limit for automobiles is 3.4 grams per
mile (gpm) for model year 1981 and subsequent model years.
The objectives of this report are to 1) summarize the available
information regarding motor vehicle emissions under non-FTP temper-
ature conditions, 2) investigate ambient air quality data to determine
the extent to which CO emissions must be reduced to achieve ambient
air quality standards, 3) evaluate vehicle test data under non-FTP
temperature conditions, 4) estimate the difference between in-use
vehicle emission levels and certification laboratory emission levels,
and 5) make conclusions and recommendations based on these objectives
to provide a baseline from which to consider the need for development
of regulations or guidelines for motor vehicle operation under non-FTP
temperature conditions and future testing procedure revisions.
This report treats the subject of CO emissions and temperature. The
work reported on in this report is only part of the overall evaluation
of CO emissions from motor vehicles that is being undertaken at EPA.
Factors other than temperature may also influence the CO emissions
from cars. For example, the CO emissions during speed/load conditions
outside those experienced on the FTP may also be important.
In addition this study proceeds under an assumption of what the CO
values from the air quality monitors represent. In this report, any
-------
-5-
difference between the CO levels recorded by the monitor and the CO
levels to which people are exposed are ignored. If monitors
overpredict or underpredict actual population exposure then the levels
of control for mobile sources may need to be altered accordingly.
Therefore, temperature should not be considered as the only important
variable.
In September 1978, a draft Advisory Circular on non-FTP Conditions was
made available for public comment [38]. This draft document provided
guidance to manufacturers on the policy that the EPA staff were
considering for the 1980 and later model years with respect to certain
aspects of Sections 202(a)(4) and 206(a)(3) of the Clean Air Act, as
amended. These sections of the Clean Air Act place the burden on the
manufacturers to establish that emission control systems or elements
of design used in new motor vehicles or motor vehicle engines do not
cause or contribute to an unreasonable risk to the public health,
welfare, or safety
before 1979 or later model year certificates of
conformity may be issued.
Even though a vehicle meets the required standards over the Federal
Test Procedure, a potential risk to public health and welfare may
occur at ambient temperatures outside of the 68°F to 86°F (20°C to
30°C) range of the Federal Test Procedure and/or over different
driving schedules than the FTP due to high emission levels of HC, CO
and/or NOx under non-FTP conditions. The draft temperature guidelines
which were made publicly available in 1978 are listed in Table A-l of
the appendix and are shown graphically in Figures 6 to 13.
II. CONCLUSIONS
The following conclusions and recommendations were made based on the
analysis of the new 1980 Mobile Source Emission Factors (Mobile 2),
ambient emission levels, and the in-house vehicle emission data
generated in the Controlled Environment Testing Facility (CETF) at the
EPA Motor Vehicle Emissions Laboratory.
-------
-6-
1. Based on the air quality data, only about 17% of all the exceed-
ences of the National Ambient Air Quality Standard for Carbon
Monoxide occur in the temperature range around 75°F (between 67.5°F
and 87.5°F - see Figure A-9) [29]. This is used here to
approximate the 68° to 86°F temperature range of the Federal Test
Procedure used to determine compliance with all emission standards,
including the CO standard, for motor vehicles.
2. The CO emission performance of motor vehicles at ambient temper-
atures outside those of the FTP is of relatively greater importance
than the CO emission performance of vehicles within the FTP
temperature range.
Specifically, the CO emission performance is of greatest importance
in the temperature range below that of the FTP. From a National
perspective, it appears that the lower bound for concern for CO
temperature performance is in the range from 0°F to 20°F. There
are exceedences of the NAAQS for CO at lower temperatures, but the
lower bound of the temperature range indicated above does tend to
recognize that special climatic conditions exist in some parts of
the U.S.A. that may be approached in a more effective way by those
closer to the problem, as opposed to adjusting the federal
emissions standards for vehicles marketed everywhere.
Restricting the temperature range of concern to between 0°F and
85 °F, for example, captures 96% of all the NAAQS exceedences for
CO, with about 2% of the exceedences below and above that range
[29].
3. This report includes information on the degree of control needed to
eliminate various percentages of the exceedences of the NAAQS for
CO based on actual emissions data collected over a three year
period from 1975 to 1977. These values of the degree of control
needed are shown as a function of temperature in the report. The
results presented are based on the assumption that any growth in
-------
-7-
Vehicle-Miles-Traveled (VMT) will not affect the result. If VMT
growth occurs and does affect the results, then the emission levels
required will be lower than those in this report.
The CO control needed is very sensitive to the percent exceedences
eliminated, especially near 100% exceedences eliminated. For
example, at about 35°F, to eliminate 95% of the exceedences results
in an implied emission level of 15 grams per mile at that
temperature, to eliminate 99% of the exceedences implies 12 grams
per mile, but to achieve the last one percent, i.e. eliminate 100%
of the exceedences, implies a level of 7 grams per mile.
4. From the data that exist to date, it is clear that substantial
increases in CO emissions result when vehicles are soaked at
temperatures lower than those specified for the FTP.
Most of the increased CO emissions come from the first few minutes
of vehicle operation. After vehicles are warmed up, temperature
does not impact their CO emissions as much. On a grams per mile
basis, bag one of the FTP (the first 8.4 irinutes of the 31.3 minute
long test) contributes 81% of the total CO emissions at 75°F, but
contributes 92% of the total CO emissions at 20°F. Cold start
emissions become a larger portion of a larger value, in essence.
For example, for the vehicles EPA has tested at its Ann Arbor
Laboratory, the total CO emissions at 20°F average more than four
times higher than at 75°F.
5. There are wide variations in the temperature sensitivity between
vehicles. Engine type, emission control system design, and
calibration strategy are all factors that influence the temperature
sensitivity.
-------
-8-
For example, two vehicles which are separated by less than 2 grams
per mile at 75°F (both below the 3.4 grams per mile standard) can
be almost 30 grams per mile apart at 20°F.
6. Based on test results from some prototype vehicles it can be said
that suitable emission control technology appears to be available
to achieve any desired level for elimination of NAAQS exceedences
between the FTP temperature range and 20°F.
III. SIMMABY OF SIGNIFICANT BACKGROUND STUDIES
Several studies have been conducted in the past few years to determine
the relationships between motor vehicle emissions and ambient
temperature. The major reports which were analyzed as background
material for this study are, in chronological order, CO Hot Spot
Preliminary Investigation by the U.S. Fnvironmental Protection Agency,
Effect of Cold Weather on Motor Vehicle Emissions and Fuel Consumption
II, by Nicolas Ostrouchov, A Review of Carbon Monoxide Emissions from
Motor Vehicles During Cold Temperature Operation, by the Alaska
Department of Fnvironmental Conservation, Effect of Ambient Temper-
ature on Vehicle Emissions and Performance Factors, by Gulf Research
and Development Company, Temperature Correction Formulae for Adjusting
Estimates of_ Emissions from Automobiles, by Robert L. Farrell of
Vector Research, Incorporated, The Exhaust Emission and Fuel Consump-
tion Characteristics of an Engine During Warmup - _A Vehicle Study, by
Donald J. Pozniak of CM, Emissions at Off-Ambient Temperatures, by
W.F. Marshall and B.H. Eccleston of the Dept. of Energy, Bartlesville,
Oklahoma, and Effect of Ambient Temperature, and Driving Cycle on
Exhaus t Emissions, W. F. Marshall of the Dept. of Energy,
Bartlesville, Oklahoma.
CO Hot Spot Preliminary Investigation
The purpose of the CO Hot Spot study was to generate preliminary data
on the effect of alternate soak times and temperatures prior to the
-------
-9-
1975 Federal Test Procedure and alternate driving cycles on CO
emission test results. The driving cycles run were the 1975 Federal
Test Procedure (FTP) and the New York City Cycle (NYCC). Soak
temperatures were either in the 10 to 25°F or 68 to 80°F ranges and
soak periods ranged from 2 hours to 8 hours or overnight soaks. The
directionality of these effects was known from previous studies. What
was in question was the relative magnitude of each test variable on
emission levels of recent production cars as well as "pre-emission
control" cars.
It was found that with soak temperatures of 10 to 25°F CO exhaust
emissions from bag one of the 1975 FTP increased by factors ranging
from 3 to 7 over identical tests following 68 to 80°F soaks for the
1976 model cars tested. During the first 125 seconds of the 1975 FTP,
CO levels as high as 350 grams/mile were emitted by these vehicles.
The extreme CO emission levels observed at start up following an
overnight soak decayed rapidly to fully warmed emission levels within
approximately 5 minutes after start up. The decay was more rapid for
the 1975 FTP than for the NYCC because the FTP is a higher speed, more
strenuous cycle.
The type of emission control system on the vehicle seemed to have a
strong impact on both the level of CO emissions and their distribution
over the driving cycle. For example, a 1976 Granada and a 1976 CVCC
Honda Civic were the two lowest CO emitters. However, bag one CO
emissions represented 90% of the total 1975 FTP CO value for the
Granada versus 28% for the Civic. Also, when the emissions perform-
ance of 1976 models tested is compared to the "pre-emission control
cars," the results show that emission controls have not reduced CO
emissions for low temperature cold starts as much as for the normal
FTP temperature cold start conditions. It was found that since all of
the cold start effects on emissions occur within the bag one sampling
period, bags two and three are virtually independent of soak time or
temperature. Thus, bag one CO emissions as a percent of 1975 FTP
composite value increase as the soak temperature is lowered and/or as
the soak time is increased.
-------
-10-
On a grams/mile basis, CO emissions over the NYCC were two to three
times those of bag one of the 1975 FTP for a given vehicle and soak
temperature. On a grams/minute basis, the CO emission levels of bag
one of the 1975 FTP were within a factor of two of the NYCC levels for
any given vehicle and soak temperature. On the same grams/minute
basis, decreasing the soak temperature from 77 °F to 10 °F to 25 °F
increased bag one and NYCC CO emissions by factors ranging from 1.5 to
75. This suggests that for ambient air modeling of cold climate
regions, the ambient temperature can be more influential than the
driving cycle.
CO emissions following a two-hour soak at 10 to 25°F are about the
same as those following an overnight soak at approximately 77°F
suggesting that the effect of soak time is strongly soak temperature
dependent. In other words, a vehicle in cold weather can have the
equivalent of multiple FTP "cold starts" in an eight hour period if
soak times are two hours or more.
The use of winter grade fuels instead of the Indolene test fuel for
1975 FTPs following soak periods at 10 to 25 °F had no appreciable
effect on CO emissions, but 50% reductions in HC emissions were
observed during the first 125 seconds of the cycle.
Finally, it was concluded that a relationship between CO emissions and
engine coolant temperature may exist for a given control technology
and driving cycle. Since engine coolant temperature could be
adequately modelled knowing ambient temperature, soak time, possibly
wind speed, and some other basic vehicle information, the engine
coolant temperature may be a useful parameter for predicting vehicle
emissions over a wide range of soak times and temperatures [15].
-------
-11-
Effeet of Cold Weather on Motor Vehicle Emissions and Fuel Consumption
- II
This report is a continuation of the evaluation of the effect of low
ambient temperature and new emission control technology on automobile
emissions and fuel consumption.
Previous studies were conducted during the winters of 1972/73 and
1975/76, and indicated that exhaust emissions and fuel consumption
from light duty vehicles are substantially affected by soaking
temperatures.
To assess the effect of various advanced emission control systems,
including Diesel powered vehicles, further cold weather testing was
done on seven vehicles (more than 160 tests) over a temperature range
of -20°C to 20°C (^°F to 78.8°F) from December 1977 to October 1978.
The cars in the test fleet included two identical 1978 Dodge Colts
(MCA Jet engines with EGR and an oxidation catalyst), two identical
1978 Volvos (three-way catalyst with feedback K-Jetronic fuel
injection), a 1978 Plymouth Fury (EGR, aspirator, and an oxidation
catalyst), a 1978 Oldsmobile Delta 88 (Diesel) and a 1978 Volkswagen
Rabbit (Diesel).
All test procedures, fuels, equipment, and emission and fuel
consumption measurements were identical to those in the previous
studies and were the same as those used in emission testing of motor
vehicles for compliance with emission standards (i.e. 1975 CVS-CH),
except that the off-baseline tests were conducted at various ambient
temperatures instead of the standard test temperature of 20 to 30°C
(68°F to 86°F). It should be noted also, that a heated sample line,
which is normally used for Diesels to prevent condensation of heavy
hydrocarbons, was not available for this program.
-------
-12-
Th e results of the testing indicated, that for late model cars, a
reduction of soaking temperature results in a considerable increase in
HC and CO emissions, but the newer emission control technology is less
sensitive to the soak temperature. For all vehicles, the emissions
during the first phase of the 1975 CVS cycle after a cold start (bag
one) account for almost all the increase in HC and CO emissions
observed at reduced temperatures.
A lowering of soaking temperature resulted in an increase in fuel
consumption and there appears to be a relatively greater loss with
decreasing vehicle weight. The temperature sensitivity of fuel
consumption in vehicles equipped with Diesel engines appears to be
significantly lower than the temperature sensitivity of gasoline
engines.
Emission and fuel consumption variations were correlated mathematic-
ally in terms of vehicle soaking and operating temperatures so that
the resulting equations could be used to estimate temperature related
changes in vehicle emissions and fuel consumption.
This study took place in Canada where the ambient temperatures are
much lower than the standard emissions and fuel consumption measure-
ment temperature range of 20 to 30°C (68 to 86°F). The average
monthly minimum temperature in the most populous Canadian cities for
six months of the year is -15 to 0°C (5 to 32°F). Thus, in modeling
fuel consumption and emissions, Ostrouchov concludes that data from
the Federal Test Procedure must be adjusted to reflect effects
attributable to seasonal and geographical variations in ambient
temperature [13].
A Review of Carbon Monoxide Emissions from Motor Vehicles During Cold
Temperature Operation
The main objectives of this report are to evaluate the extent to which
emissions affect achievement of ambient air quality standards for
carbon monoxide, particularly emphasizing Fairbanks and Anchorage,
-------
-13-
Alaska and to present methods which are currently available to reduce
cold start emissions. No vehicle testing was done, but results of
other studies done by EPA and assorted Canadian test programs were
analyzed.
The conclusions are that at 0°F, 81% of all carbon monoxide emitted
from automobiles tested in these various programs comes from the cold
start phase of operation, whereas at 75 °F only 36% of the carbon
monoxide comes from the cold start phase, as measured by the Federal
Test Procedure. Vehicles range from 1970 to 1980 prototype models,
but the number of vehicles tested and their control systems are not
provided in the report. The authors of the study concluded that the
amount of carbon monoxide emitted during engine cold start conditions
increases by 7 1/2 times as the ambient air temperature decreases from
75 to -20°F. Also, a vehicle's cold start carbon monoxide emissions
are substantially increased with longer cold soak time.
It was deduced in this report that the major contributors to high
carbon monoxide emissions in cold weather are 1) malfunctioning
automatic chokes, 2) ignition timing and carburetor settings which
are not geared to minimizing emissions but rather are set solely to
improve car driveability, and 3) carburetor and intake manifold
systems which do not efficiently atomize fuel before it is burned.
Four methods of CO reduction are proposed in this report including 1)
use of exhaust manifold air injection, 2) use of engine preheat
devices, 3) alternative engine technology and 4) engine tuneups and
regular vehicle maintenance.
According to research in Canada, 1975 California vehicles with exhaust
manifold air injection can reduce low temperature cold start carbon
monoxide emissions by more than 60% compared to 1975 Federal vehicles.
-------
-14-
The use of vehicle engine preheat devices reportedly has the potential
to reduce carbon monoxide cold start emissions between 44% and 85%,
depending on the preheat device used and the amount of preheating
employed. However, it is advised in the report that energy
consumption should be evaluated before selecting any preheat device as
a means to achieve ambient air quality standards. Preheat devices
which store excess engine heat during warm driving conditions to be
later used to start a cold-soaked engine, could be further developed
and considered for installation on new vehicles sold in cold weather
climates.
The authors of the Alaskan report assert that alternative engine
technology which has inherently low cold start emission character-
istics, such as the Honda CVCC engine may offer the greatest potential
for reducing cold start emissions of all possible methods reported in
the current research literature.
The Alaskan study also concludes that engine tuneups and maintenance
should substantially reduce overall carbon monoxide emissions through
more proper timing and choke settings. The report submits, however,
that research needs to be conducted to quantitatively determine the
effect of maintenance on cold climate emissions [1].
Effect of Ambient Temperature on Vehicle Emissions and Performance
Factors
The effect of ambient temperature on exhaust emissions, fuel economy,
catalyst light off time, exhaust temperatures and driveability were
studied using 14 automobiles at eight ambient temperatures which
ranged from 0°F to 110°F (-18°C to 43°C). The vehicles consisted of
three noncatalyst cars: a 1972 Chevrolet, a 1974 Chevrolet, and a 1977
Honda Civic. Three were 49 state vehicles: a 1977 Ford, a 1977
Plymouth, and a 1978 Buick. Five were California vehicles: a 1977
Plymouth, a 1977 Chevrolet, a 1978 Ford, a 1978 VW Rabbit
(non-Diesel), and a 1979 Dodge. Three were prototype vehicles: a
-------
-15-
1980 Mercury, a 1980 Buick, and a Datsun prototype. The studies were
conducted using the 1975 Federal Test Procedure, the New York City
Cycle and the Federal Short Cycle.
The cold transient phase of the FTP displayed high CO concentrations
at 0°F and 20°F (-18 °C and -7°C) for all of the cars. The 1980
prototype Mercury gave over 100 grams/kilometer (161 grams/mile) of CO
bag one FTP emissions at 20°F (-7°C) despite having an advanced
catalyst system. CO was the highest for two Chevrolet noncatalyst
cars. The Honda, which was lower at 0°F and 20°F (-18°C and -7°C)
than several of the catalyst cars, was also quite low at 110 °F
(43°C) . The use of air conditioning usually increased CO, possibly as
the result of increased load. The VW Rabbit showed the least change
in CO emissions with ambient temperature. Overall, the Datsun
prototype gave the lowest CO emissions for the cold transient phase of
the FTP. The hydrocarbon and carbon monoxide variations observed at
the lowest temperatures using the cold transient phase of the FTP
appeared to be a function of the emission control technology used on
the vehicle.
The stabilized phase of the FTP gave some very low CO emissions. The
1978 (49 state) turbocharged Buick, the 1977 (California) Plymouth,
the 1978 (California) Chevrolet, the 1979 (California) Dodge, the 1980
Mercury (prototype), the 1980 Buick (prototype) and the Datsun
(prototype) all had CO emissions of less than 2 grams/kilometer (3.2
grams/mile) from 0 to 80°F (-18 to 27°C). However, rich operation as
indicated by high CO levels at the highest temperatures was evident in
all cars, some more than others. At 110°F (43°C), the use of air
conditioning often resulted in CO emissions that more than doubled.
Data from the hot transient phase of the FTP show that the 10-minute
soak and the startup still had an impact at reduced temperatures.
Again, higher temperatures and particularly the use of air condition-
ers increased CO emissions. The 1978 Chevrolet had the lowest CO
emissions at all temperatures other than 110°F (43°C) where several
-------
-16-
cars appeared to perform better. The 1980 Mercury (prototype) had the
lowest CO emissions at 110°F (43°C), particularly with the air
conditioners in operation.
The Sulfate Emission Test (SET), which is performed using the
Congested Freeway Driving Schedule (CFDS), was run following the FTP
and, after a three minute engine idle period. Results showed higher CO
emissions than those measured using the Highway Fuel Economy Test
(HFET) but generally lower CO emissions than those measured during the
hot transient phase of the FTP. CO emissions were the greatest at
110°F (43°C) and higher at the same temperature with the vehicles' air
conditioning in operation.
The New York City Cycle (NYCC) is made up largely of idle conditions
plus several sharp accelerations and decelerations. Fven the
California cars and the prototype cars have significant CO emissions
when operated on this cycle. The Honda was the only noncatalyst car
tested using this cycle. The NYCC, run with a warmed-up engine, gave
several times the CO emissions obtained with the HFET cycle. The
Honda, with its modified combustion system and no catalyst, was equal
to many of the catalyst cars and much better than some at 110°F (43°C).
CO emissions were also analyzed using the Federal Short Cycle (FSC), a
cycle designed for fast analysis of vehicle emissions. Of most
interest is its correlation with the FTP. In general, the CO
emissions obtained with the FSC were much lower than FTP composite
values at temperatures below 110°C (43°C). At 110°F (43°C) CO
emissions on the FSC were very high.
The light off times and the corresponding catalyst-out exhaust gas
temperatures during the FTP varied widely for the different catalyst
systems. Light off time was defined for this study as the time in
seconds from the start of the cold transient of the FTP until the
temperature of the gases leaving the catalyst exceeded those entering.
The light off times for all eight ambient temperatures varied from 67
-------
-17-
to 419 seconds and the light off temperatures ranged from about 400 to
1100°F (204 to 593°C). There was no consistent relationship between
light off time and ambient temperature. The maximum exhaust gas
temperature measured was 1462°F (794°C). There were larger differ-
ences in exhaust gas temperatures between the various catalyst systems
than there were between test procedures or ambient temperatures.
A considerable number of driveability problems occurred at 0 and 20°C
(-18 and -7°C) using the cold transient phase of the Federal Test
Procedure. However, some driveability problems were encountered at
higher ambient temperatures and with other test cycles. Driveability
problems usually increased most of the exhaust gas emissions [4].
Temperature Correction Formulae for Adjusting Estimates of Emissions
from Automobiles
This analysis, conducted by Vector Research, Inc., was intended to
provide formulae, referred to as correction formulae, which could be
used to estimate the emissions of regulated pollutants at temperatures
other than those used in the standard Federal Test Procedure. Data
were analyzed on a bag by bag basis from more than 100 U.S. and
Canadian vehicles over a 0 to 110°F temperature range to determine the
effects on CO emissions of 1) model year (1967 to 1980), 2) standard
(California or 49-state Federal), 3) emission control system (engine
modifications, air pumps, three-way and oxidation catalysts,
carburetors and fuel injection) and 4) inertia weight. Emission
control systems were analyzed in the following configurations: no
controls, EM, EM/AIR, OC, OC/AIR, TWC, TWC/FI, TWC/OC, and TWC/OC/FI.
The temperature effects observed were evaluated by comparing the CO
emission levels at low temperatures (0 to 67.5°F) and high tempera-
tures (86.5 to 110°F) to those observed at the standard FTP
temperature.
Temperature effect equations were fit to each of the three bags. From
the data, the equations showed increases in emissions at low
-------
-18-
temperatures for bags one and two, with little or no effect for bag
three at low temperatures, and decreases in emissions at high
temperatures for bag one, with increases at high temperatures for bags
two and three.
No trend was shown toward increasing or decreasing bag one CO
emissions since the 1967 model year at temperatures higher or lower
than the FTP temperature (analyzed in the report as the natural
logarithm of emissions of CO in grams per mile at the non-standard
temperature divided by the natural logarithm of CO emissions in grams
per mile at the standard temperature). However, bag one CO emissions
at low temperatures have been consistently higher than bag one CO
emissions at the FTP temperature. Conversely, bag one CO emissions at
high temperatures have been consistently lower (with the exception of
1977 California vehicles) than bag one CO emissions at the FTP
temperature. The maximum cold temperature bag one CO emissions came
from 1?76 Federal vehicles wV th 4 1/2 to 5 times as much bag one
*\
emissions in the 20°F range i\\s in the FTP temperature range. Among
[i
vehicles tested at OF, 1977 Federal cars exhibited bag one CO
emissions 7.6 times greater than they did at FTP temperatures. The
minimum cold temperature bag one CO emissions occurred in 1970 Federal
vehicles with CO levels at 20°F 1.64 times higher than those at FTP
temperatures. Bag one CO emissions at 110°F ranged from 0.40 (for
1970 Federal vehicles) to 1.63 (for 1977 California vehicles) times
the CO emission levels at FTP temperatures. These ratios are again
based on the natural logarithm of grams per mile CO emissions at
non-standard to standard FTP temperatures.
Most bag two CO emissions were not greatly increased as the
temperature got farther away from either end of the FTP range with the
exception of 1978 Federal vehicles. At ]10°F, these vehicles
exhibited a tenfold increase in CO emissions over tests at FTP
temperatures.
-------
-19-
Bag three CO emissions at low temperatures did not vary significantly
from FTP temperature levels for all model years in the data base. The
effect was slightly more pronounced at higher temperatures. Again,
1978 Federal vehicles were the worst case emitting 3.25 times as many
CO emissions at 110°F than at FTP temperatures.
In very few cases did the effect of standards (California or Federal)
make a difference on temperature response of CO emissions. In those
that did, it was found that California vehicles showed less effects of
temperature, with the ratio of California vehicle emissions at any
temperature to those at FTP temperatures being smaller than those
exhibited by Federal vehicles. Perhaps this is due to the fact that
prior to 1980, the California standard (9.0 CO) was stricter than the
Federal standard (15 CO).
Data were examined for eight different categories of emission control
technology to determine their respective effects on CO emissions at
different temperatures. Surprisingly enough, the worst bag one CO
emitter at both low and high temperatures relative to FTP CO emissions
were systems equipped with three-way plus oxidation catalysts with
fuel injection. Very few data were available for this system compared
to the other seven systems analyzed. The next worst systems, in terms
of bag one CO emissions, were systems equipped with three-way
catalysts and a carburetor at temperatures below the FTP, and systems
equipped with oxidation catalyst without an air pump at temperatures
above those of the FTP. The lowest CO emitting systems from bag one
at low temperatures were vehicles with three-way catalysts and fuel
injection and vehicles with engine modifications and air pumps. No
data were available for these two systems at high temperatures. The
best systems of those with data available at high temperatures were
vehicles with oxidation catalysts and air pumps and vehicles with
three-way catalysts and a feedback carburetor. Vehicles with no
emission control system at all exhibited high CO emissions at all
temperatures compared to their emissions at FTP temperatures.
-------
-20-
Cases were examined in which the temperature response was assumed to
vary linearly with vehicle inertia weight. Statistically significant
effects were found to exist for below FTP temperature CO emissions in
bags one and two. The effects were measured in terms of the change in
a temperature response coefficient per thousand pounds of increased
inertia weight from a. mean value of 3760 pounds. CO emissions were
found to increase slightly with increased inertia weight in bag one,
and decrease slightly with increased inertia weight in bag two. No
effect of inertia weight on CO emissions at temperatures above the FTP
range could be found [2].
The Exhaust Emission and Fuel Consumption Characteristics of an Engine
During Warmup - A Vehicle Study
The purpose of this study was 1) to determine the effects of air-fuel
ratio and spark advance on the emissions measured during warmup of a
vehicle when operated in accordance with the FTP, and 2) to identify
reasons for differences in levels of emissions measured during warmup
compared with those measured when the vehicle was fully warm.
Testing was conducted using a vehicle equipped with a 5.0 liter
engine, three speed automatic transmission, EGR, and an oxidation
catalyst. A modified carburetor was used on the test vehicle for this
project. The principal observations made are 1) during warmup,
minimum HC and CO tailpipe emissions occurred near air-fuel ratios of
14.5 and 16, respectively, 2) after warmup, minimum HC was at about
16.5 air/fuel ratio and minimum CO was at an air/fuel ratio of 16 or
leaner and 3) compared with fully warm operation, operation during
warmup increased fuel consumption at all air-fuel ratios and increased
HC and CO emissions at air-fuel ratios only leaner than about 14 and
16, respectively. Spark retard during warmup did not significantly
affect HC emissions or fuel consumption, but decreased NOx emissions
and driveability, and increased CO emissions at lean air fuel ratios
only.
-------
-21-
The primary conclusions reached from the study are 1) during warmup on
the FTP, the engine air-fuel ratio should be maintained in the 14.5 to
16 range as soon as practicable within the constraints of acceptable
vehicle driveability. Disregarding driveability, an air-fuel ratio of
16 offers the best compromise of emissions and fuel economy. 2)
Within the range of variables used in the study, the time required to
attain catalyst reaction temperature is not measureably affected by
engine air-fuel ratio. 3) Gas phase oxidation in the exhaust system
upstream of a catalytic converter can be a major reason for lower HC
and CO emissions from a vehicle when fully warm than when cold [7].
Emissions at Off-Ambient Temperatures
The Department of Energy's (DOE) Bartlesville Energy Center (BETC) was
selected in an interagency agreement between DOE and EPA to conduct
tests on a fleet of vehicles to determine a baseline for automobile
emissions at a variety of operating temperatures. All tests were
conducted in a climate-controlled chassis dynamometer facility at BETC.
A fleet of twenty-five 1970 model-year vehicles had been previously
acquired and tested by Automotive Testing Laboratories, Inc., under
contract with the Motor Vehicle Manufacturers Association. The
vehicles were selected to represent the 1970 model-year nationwide
sales mix. Each vehicle was tested on the 1975 FTP, the HFET and the
NYCC sequences. No additional load was applied for vehicles equipped
with an air conditioner.
The fleet average FTP CO emission rate was 21 grams/mile at 75°F. The
current 3.4 grams/mile standard is a 90% reduction from fleet average
34 grams/mile CO as measured by the revised 1975 FTP [40]. HC and CO
emissions increased significantly as the test temperature was
decreased from the standard ambient of 75°F to 25°F. The increases
were 0.7% and 1.5% per 1°F decrease for HC and CO, respectively, over
the EPA urban FTP driving cycle.
-------
-22-
In the 1975 FTP test cycle, variation from a standard ambient
temperature of 75°F had little effect on emissions over the stabilized
and hot transient phases. Essentially all of the increases in
emissions of HC and CO associated with lower ambient temperatures were
incurred in the cold transient phase.
FTP nitrogen oxide emissions increased slightly with increasing
temperature, approximately 6% greater at 100°F than at 75°F.
Emissions of all three regulated pollutants (on a grams per mile
basis) were greater for the NYCC than for either the FTP or highway
driving cycles.
Fourteen of the 25 cars in the fleet had operational air-conditioning
systems and were tested on the FTP, HFET and NYC cycles at 75 °F and
100°F, both with and without air-conditioning. The use of
air-conditioning caused moderate increases in HC and CO emissions in
the FTP and highway cycles. HC and CO increased by 20% and 60%,
respectively, during the NYCC at 100°F. This is due to the additional
load placed on the engine during idle, a mode which comprises more
than one third of the entire NYCC. The effect on NOx was much greater
and fairly consistent for all three driving cycles. The increased
power requirement could be the source of this increase [5].
Effect of Ambient Temperature and Driving Cycle on Exhaust Emissions
This report is a follow-up study on the baseline examination described
in the previous paper, Emissions at Off-Ambient Temperatures. In that
paper, a fleet of 25 1970 model year vehicles were tested to form a
baseline with which to compare later model vehicles. In this report,
five more 1970 MY vehicles were tested plus seven late model
production vehicles for a total of 37 vehicles in the test fleet. The
late model production vehicles consist of three three-way catalyst
vehicles, two Diesel vehicles, one stratified charge vehicle and one
turbocharged vehicle.
-------
-23-
Emissions were determined at a variety of ambient temperatures to
evaluate the emission trends of new technology vehicles relative to
those of the 1970 MY vehicles.
When emissions at different temperatures are referenced to the 75°F
FTP case, the results appear to show that temperature extremes have a
greater deleterious effect on emissions performance of the current
production fleet than the baseline fleet. CO emissions in grams/mile
were 3.3 times higher at 25°F than at 75°F for the current fleet
compared to 1.8 times higher for the baseline fleet. However, if the
results are referenced to the emissions "standards" applicable to the
fleet, the current production fleet's performance appears to be
markedly better than that of the baseline fleet. (The standard for
1970 baseline vehicles was assumed to be 34 grams/mile). Absolute and
percentage differences in emission rates between the two fleets
indicate that overall, the current fleet's emission levels are about
70% less than the baseline fleet at 25°F [39].
IV. INVESTIGATION OF DATA FROM THE 1975 THROUGH 1977 VEHICLE FLEET
The EPA Mobile Source Emission Factors data of March 1978, hereinafter
referred to as Mobile 1, were analyzed in combination with data from
the Storage and Retrieval of Aerometric Data (SAROAD) and the National
Weather Service (NWS) data bases to determine the certification
emission levels at specific temperatures that would have to be
attained by vehicles to eliminate 90 to 100% of the exceedences of the
3
8 hour lOmg/m National Ambient Air Quality Standard (NAAQS). The
same analysis was also done using the new 1981 Mobile Source Emission
Factors data, hereinafter referred to as Mobile 2.
Ambient Data from the SAROAD and NWS Data Bases
Data from the SAROAD data base provided observations of ambient CO
concentrations at various locations as functions of date and time of
day. The NWS data provided concurrent ambient temperatures that could
-------
FIGURE: i
BflSZ.
1S73-1377 BJ-»J2 33-002 *
100
SO
so
o 70
3. SO
o
50
40
3D
10
DO! EE
EE
932 EE
1002 EE
1 \ 1 1
_L
JL
1 T ?
T
T
3 U 13 15 17 13 21 Z3 2H 27 23 31 33 35 37 39 >ll
m£-P.U3« TEHJ- C0.5J - ISi C70-75J - 23 x*
SS = 00% ineans no stationary source contribution to total CO emissions.
S3 = 05% neans stationary sources amount to 5% of total CO emissions.
Interval 15 corresponds to the 0-5 F temperature range.
Interval 29 corresponds to the 70-75 F temperature range.
The specified tesrperature range is an average of the three hours
immediately preceeding the rush hour that preceeds the ending hour
of the associated 8-nour average concentration.
-24-
-------
HOIULI.
RB9K CH3£ 1 975-1977
SS=05Z
JDO
70
o
tU
eo
SO
/c
1*0
30
20
VI
J_J__I__L
T i » t —I L
_L
S 11 13
PflE-RU3H
»7 13 21 23 25 27 23 31 3
? CD.S) - iSt C70-7S) - 23
33
7 33
-25-
-------
FIGURE 3
MOBILE 2
Bfl3E Cfl3E 1375-1377 HI/H' 33-002
100
70
UJ
ui
_ SO
o 50
ui
e:
t—
•x.
«_>
30
20
10
* :
vV*V*
SOZ EE
=a=» 352 EE
"uvu 332 EE
lOOt EE
./A
-<^
o
t
1
t_
_L
3 11 13 IS 17 13 21 23 2S 27 29 31 32 35 37 33 Ul
PRE-RU2H TEKf* CO.S) - IS; C70-7S) - 23
* Non I/iI. Inspection and nnzntenancc benefits not included.
-26-
-------
FIGURE 4
MOBILE 2
BH3E CB3E 1375-1377 HJ/H 33-052
100
30
to
o
Ul
ac
ED
SO
30
20
• 0
J L
J L
J L
J L
3 11 13 15 17 13 21 23 25 27 23 31 33 35 37 33
PBE-RU3M TEKP CO.5) - iSt C70-75) - 23
-27-
-------
-28-
be related to these ambient CO monitoring locations. Figures 1 to 4
show the percent by which each eight hour average ambient CO level in
over 50 cities must be reduced to eliminate exceedences of the
standard based on Mobile 1 and 2 with stationary source CO
contributions of zero and five percent. Cities were chosen on the
basis of having a large number of 8 hour average concentration values
3
in one year that exceeded 10 mg/m . In general, the 50 cities with
the greatest number of exceedences have been included. The dotted
line indicates the per cent reduction in ambient CO needed to
eliminate 100% of the NAAQS exceedences. The solid line indicates the
level at which 90% of exceedences will be eliminated. Temperatures
range from -50°F to 110°F in five degree intervals. Interval number
15 includes all eight hour averages that were associated with a zero
to five degree (°F) temperature range. Interval number 29 covers the
70-75°F range. Data from SAROAD for the three year period from
January 1, 1975 to December 31, 1977 were used in the analysis. At
the temperature extremes, comparatively few data points determine the
position of these curves. Fairbanks, Alaska, accounts for most of the
very low temperatures and Phoenix, Arizona, accounts for most of the
very high temperatures. The highest eight hour average CO
3
concentration confirmed for the 1975-1978 period is 38.5 mg/m . (A
3
74 percent reduction in this level is required to meet the 10 mg/m
3
standard.) Concentrations reported which were above the 38.5 mg/m
level were excluded from the SAROAD data used in this analysis.
Ambient CO concentrations were recorded on an hourly basis from which
running eight hour CO averages were determined. Table 1 shows an
example of how running eight hour averages were taken from hourly CO
data to determine eight hour average CO exceedences of the applicable
NAAQS. In this example 9 ppm is the assumed standard and eight hour
average concentrations above a 9 ppm are said to be exceedences of the
standard. For the Mobile 2 (new emission factors) analysis, a 9 ppm
3
(10.35 mg/m ) standard was also assumed.
-------
-29-
TABLE 1
Example of How Running Eight Hour Averages
Are Determined from Hourly CO Data
CONSECUTIVE
1-HR AMBIENT
CO OBSERVATIONS
(ppm)
12
11
10
8
7
6
5
5
6
6
6
8
10
13
15
15
16
17
18
17
16
16
16
16
RUNNING
8 - HR AVERAGE
(ppm)
9ppm 8 - HR
AVERAGE
EXCEEDENCE?
8.0
7.3
6.6
6.1
6.1
6.5
7.4
8.6
9.9
11.1
12.5
14.0
15.1
15.9
16.3
16.4
16.5
no
no
no
no
no
no
no
no
yes
yes
yes
yes
yes
yes
yes
yes
yes
If you take the set of exceedences and reduce each one to the
standard, the amount or fraction by which each exceeds the standard is
the "percent reduction needed" to achieve that standard. Table 2
-------
-30-
shows an example of how percent reduction needed to achieve a 9 ppm CO
level was calculated from a given set of exceedences. To eliminate
100% of the exceedences a percent reduction needed associated with the
maximum exceedence, or 45.5%, must be achieved. Figure 5 shows the
relationship between percent reduction needed and percent exceedences
eliminated for this example. A line of equality is shown to
demonstrate that the percent exceedences eliminated is greater than
the percent reduction needed.
TABLE 2
Example of How Percent Reduction Needed and Percent Exceedences Eliminated
Are Calculated From a Given Set of Exceedences
To Meet a 9ppm Ambient CO Level
SET OF (%) REDUCTION (%) EXCEEDENCES
EXCEEDENCES (ppm) NEEDED ELIMINATED
16.5 45.5 100
16.4 45.1 89
16.3 44.8 78
15.9 43.4 67
15.1 40.4 56
14.0 35.7 44
12.5 28.0 33
11.1 18.9 22
9.9 9.1 11
All of the ambient temperatures in this analysis are called "pre-rush
hour" temperatures. Pre-rush hour temperature means the temperature
at which the vehicle is soaked prior to the morning or evening rush
hour. So all exceedences in a given day are assigned a corresponding
temperature which is the average ambient temperature for the 3 hours
which immediately precede the rush hour. This, we believe,
appropriately assigns CO exceedences to the temperature during the
-------
FIGURES.
PERCENT REDUCTION NEEDED VS
EXCEEDENGES ELIMINATED
FOk EXAMPLE IN TA&LE 2
too
2O 4-O 60 80
% REDUCTION NEEDED
100
-31-
-------
-32-
period of time immediately before the cold starts and vehicle
operation during the rush hour. The morning rush hour is taken from
0601-0900 Local Time (LT); the evening rush hour extends from
1601-1900 LT. Daylight Savings Time (DST) begins on the last Sunday
of April and ends the last Sunday of October. A complication which
arose was that not all locations in the U.S. adhered at that time to
DST uniformly. Nevertheless, for computation ease DST was assumed to
apply to all areas from May through October.
Vehicle Emissions Data
Tailpipe CO emissions data were collected from the data bases used in
the low altitude Federal versions of Mobile 1 and Mobile 2. The
collected data were from vehicles and engines which correspond to the
light duty vehicles (automobiles) and the mobile source fleet which
were in-use during the period from January 1, 1975 through December
31, 1977. The in-use CO emissions were calculated for the temperature
range of concern by the computer programs in Mobile 1 and 2. The
calculated emissions were then reduced by the percent reductions
needed to eliminate exceedences of the NAAQS CO standard at the same
temperature (as shown in Figures 1 through 4). This analysis was done
first assuming only improvements in CO emissions from light duty
vehicles (LDV) and then assuming proportional improvements in fleet
emissions. The graphs labeled LDV assume that LDV emissions
completely represent the fleet. The graphs labeled FLEET assume that
"composite" emissions represent the fleet. The CO levels which must
be attained by in-use vehicles to eliminate the NAAQS exceedences are
plotted for pre-rush hour temperatures ranging from 0°F to 100°F in
five degree intervals (see Figures A1-A8). It is important to note
that no high altitude data were included, and that VMT growth was not
considered.
The analyses of Mobile 1 and 2 were done for two separate cases, one
which assumed that 5% of the CO was attributable to stationary source
(ss) contributions and one which assumed no stationary source
contribution. The National Emission Inventory Report [31] indicates
-------
-33-
that the average stationary source CO contribution of the 146 worst
counties is approximately 15% [31]. This stationary source
contribution takes into account, however, "uncontrolled" mobile source
contributions such as off-highway gasoline vehicles, off-highway
diesel vehicles, rail, vessels and aircraft in addition to the true
stationary sources. In order to meet the NAAQS, the regulated mobile
sources, which consist of light duty vehicles, light duty trucks,
motorcycles, heavy duty gasoline vehicles and heavy duty diesel
vehicles, must be controlled to lower CO levels to account for the
uncontrolled mobile sources and the true stationary source
contributions.
In modeling carbon monoxide concentrations, stationary sources are
neavily discounted to account for their normal dispersion [35].
Discount factors of 80% for stationary area sources and 100% for
stationary point sources such as power plants are used. These factors
were selected after considering the results of dispersion models for
point sources and reviewing the relationship between traffic density
and CO levels in several situations. Using these discount factors,
the stationary source contribution is assumed to be about 2% or 3%.
Thus, EPA used 0% and 5% in the analysis as the lower and upper bounds
for the stationary source contribution.
As a final treatment of the data, a certification to in-use factor
(CKF) was applied to the in-use CO emissions levels which must be
attained to eliminate NAAQS exceedences in order to compare these
emissions levels with data from EPA certification type testing and
from the EPA's Controlled Environment Testing Facility (CETF). The
CETF is an in-house testing facility at the Motor Vehicle Emission
Laboratory (MVEL), Ann Arbor, Michigan. The latest EPA estimates for
certification to in-use factors, based on Mobile 2 for low altitude,
49 state fleet emissions (and assuming a 3.4 CO light duty vehicle
standard) are 3.04 (with inspection and maintenance programs) and 5.53
(without inspection and maintenance) [36]. The predicted Mobile 2 50K
CO emission level for 1983 and future model year vehicles without
-------
-34-
inspection and maintenance programs is 18.80 grams/mile. Dividing
this "in-use" number by the 3.4 grams/mile CO "certification" standard
yields a CRF of 5.53. Mobile 2 estimates that inspection and
maintenance programs are effective in reducing CO emissions levels by
45%. The CRF assuming inspection and maintenance programs is
therefore 55% of the CRF without inspection and maintenance, or 3.04.
Plots of CO emission levels versus pre-rush hour temperature for
Mobile 2 with these CRF's applied are shown in Figures 6 to 13 for 90,
95, 99 and 100% of the NAAQS exceedences eliminated.
-------
6.
20
>
r
o
o 10
MOBILE 2 LIGHT DUTY VEHICLES STATIOM/1RY SOURCES - 0% CRF«3.04
- I9"77
EXCEEDENCCS
WITH INSPECTION AND
MAINTENANCE
I9SO DMFT CO
19814
O
20
3O
4O 50
?o
8O
PRE-^USH
9O /OO
(°F)
^Assumes LUV is the entire fleet.
-------
7.
20
V)
£
2
o
u
MOBILE Z LIGHT DUTY VEHICLES STATIONARY SOURCES-STo CRF=3.04
19814 DRAFT CO
6UIDELINE.
IO
ZO
40
1915-1977
%EB =
WITH INSPECTION AND
MAINTENANCE
5O 60 "7O 6O
9O IOO
-------
FIGURE 8.
30
20
iu
I
t*
o
u
10
MOBILE Z FLEET EMISSIONS STATIONARY SOURCES-0%
I97S-I977
- PERCENT
I98O PRAFT CO
GUIDELINE
19 81 +DRAFT CO
6U!D£L|N£
WITM INSPECTION/ ANp
MAINTEMANCE
o
10 ZO 3o 40 5"o 6O 7O
PP.e-R.USH
80
9O
IOO
*Assumes identical CO reduction for all segments of the vehicle fleet.
-------
o
10
MOBILE 2 FLEET EMISSIONS STATIONARY SOURCES«5% CRF-304
90%
1980 DRAFT CO
GUIDELINE
1981+ DRAFT CO
GUIDELINE.
1975-1977
ELIMINATED
WITH INSPECTION AND
MAINTENANCE
o
IO 2O
30 40
GO 70
PRE-PU5H HOOP.
80 9O
IOQ
-------
FIG UP. e IO.
MOBILE 2 LIGHT DUTY VEHICLE STATIONARY SOURCE9 = 0%> CRF-5.53
1975" -1977
%EE = PERCENT
20
i
•£
O
u
I960
CO
IO
PRAFT CO
GUIDELINE.
W/0 INSPECTION! AND
MAINTENANCE
o
10 zo 30 4o 5O GO
PR.E -
"7O 80 9o 100
HOOK, l^M PERTURB (°F)
-------
FIGURE n.
UJ
j
IS
O
,o
MOBILE2 LIGHT DUTY VEHICLES STATIONARY SOURCES =
19814 DRAFT CO
J I L
1975-1917
a PERCENT
ELIMINATED
W/O IN5PECT10M AND
MAINTENANCE
. I960 DRAFT CO
J I I I L
o \o
60 70 8O 90
PKE-ROSH
-------
12.
20
Ul
J
VI
I
O
10
MOBILE 2 FLEET EMISSIONS STATIONARY SOURCES -0% CRF-S.S3
I960 DRAFT
1981+ PRAFT CO
-1977
- PERCENT
W/0 INSPECTIOM AND
MAINTENANCE
o
30
70
PRE-P.(;SM HOUR
9o too
-------
i3.
2o
V)
I
u
O
O
u
MOBILE 2 FLEET EMISSIONS STATIONARY SOURCES-5% CRF-553
-19-77
= PERCENT
ELIMINATED
w/o iNSPEcnorJ AMD
MAINTENANCE
u 1980 DRAFT CO
1981+ DRAFT CO
GUIDELINE
10
30
40 SO
6O
70 8O 9O
T£MP6RATUP,E
IOO
-------
-43-
EPA Controlled Environment Test Facility Data
The previous analyses both relate CO emission levels needed to
eliminate exceedences of the NAAQS to pre-rush hour temperature. The
reason for this is that these temperatures more closely represent the
ambient temperature that in-use vehicles are soaked at for a number of
hours before they are cold started and begin warming up. Rush hour
temperature values more commonly represent vehicles operating under
warmed up and hot start conditions.
Vehicles tested in the EPA's Controlled Environment Test Facility
(CETF) are soaked for 12 to 24 hours at the test temperature before
being run on the Federal Test Procedure (FTP), Highway Fuel Economy
(HFET) and New York City Cycles (NYCC) (see test sequence, Table 3).
The EPA's emission testing program at different temperatures began in
February 1980. To date ten vehicles have completed testing which
includes duplicate tests at 20, 60, 75 and 100°F over the FTP, HFET
and NYCC cycles. Frequent CETF compressor breakdowns, instrumentation
failures, dynamometer problems and competition with other testing
programs have delayed the program since then. It was difficult to
obtain vehicles which represent the industry sales mix and current
emission control technologies. Seven of the ten vehicles were either
certification emission data (5) or fuel economy (2) vehicles and the
other three were production vehicles, two of which had data from other
testing programs with which to correlate. Tests at 75°F in the CETF
compared reasonably well to official certification tests for all
vehicles run on a Clayton Chassis Dynamometer in a light duty
certification test cell.
This program was designed to provide emission test data from vehicles
at ambient conditions other than FTP conditions. Temperature
tolerances in the test cell were held to +5°F during soaks and +2°F
during vehicle tests. Humidity was controlled at 75 +5 gr/lb during
vehicle tests at 75 and 100°F and was not controlled during vehicle
-------
Table 3
Test Sequence
1. Refuel to 40% tank capacity.
2. Set-up vehicle on dynamometer for testing.
3. Stabilize ambient conditions in the CETC to within tolerances of
desired conditions.
4. Precondition with 1 LA-4 cycle. The blower velocity will be
equivalent to a vehicle speed of 20 mph. and the hood will be open.
5. Engine off, 30 minute soak. Air stream will be deflected from the
vehicle, hood closed.
6. Rapid cool-down. Blower at 50 mph, undeflected. Hood open.
7. Cool down until oil tempe~ature in oil pan is + 2°F of test
temperature, then 30 minutes Additional cool-down.
8. Emission Test Sequence
There will be 10 minute soaks between tests. The blower will be at
20 mpg during the sequence a.id deflected during the 10 minute soaks.
The hood will be open during i ests and closed during\ soaks.
!\
a. Cold Start FTP ,
b. HFET i
c. NYC Cycle
9. Refuel to 40% tank capacity.
10. Precondition with 1 LA-4 cycle.
11. 12-24 hour soak. Hood will be closed and airstream deflected. Blower
output within +_ 5°F of desired test condition.
12. Repeat step 8 once.
13. Change temperature and humidity to next settings.
14. Refuel to 40% tank capacity.
15. Repeat steps 4-14 for each temperature. Sequence will be 75° - 20° -
60° - 100°F.
-44-
-------
-45-
soaks. A 10% air conditioning horsepower adjustment was subtracted
from 20°F and 60°F tests. No evaporative emissions tests were run.
For a complete description of the test sequence refer to Table 3. A
complete description of the CETF test vehicles is listed in Table 4.
This test program is continuing and the CETF has recently been
modified for testing of front wheel drive vehicles.
V. DISCUSSION OF CURRENT DATA
As mentioned, Figures 6 through 13 display the emission levels from
Mobile 2 reduced to levels needed to eliminate 90, 95, 99 and 100
percent of the exceedences of the NAAQS with CRFs of 3.04 and 5.53.
There is little difference between the values for Mobile 1 and those
of Mobile 2. Mobile 2 values are slightly higher, but the differences
are mainly at the temperature extremes where there are not much data
of high confidence. For simplicity, therefore, CRFs were only applied
to Mobile 2 plots. The September 1978 EPA draft CO guidelines at
temperatures of 75, 50 and 20°F are shown on each graph as reference
lines (see Table A-l in appendix).
For a CRF of 3.04, the 100% exceedences eliminated curve is above the
1981+ CO guideline in all cases. Assuming there is a correlation
between CO emissions at any given temperature to CO emissions at 75°F,
this would indicate that adherence to this guideline would be
sufficient for eliminating all NAAQS exceedences.
With a CRF of 5.53, the 90% exceedences eliminated curve is below the
1980 CO guideline, and the 1981+ CO guideline (based on a
3.4 grams/mile CO standard at 75°F), approximates a level close to 99%
NAAQS exceedences eliminated.
For 90% NAAQS exceedences eliminated, the necessary CO emissions are
1.7 times higher at 20 °F than they are at 75°F. This ratio increases
up to 2.6 for 100% NAAQS exceedences eliminated. The 1978 CO
guideline assumed a ratio of 1.76 in this same temperature interval.
-------
Table 4
EPA CETF TEST FLEET
MY
1979
1980
1980
1981
1980
1978
1980
1981 1/2
1981
1980
MFR
Ford
Nl ssan
(FB
Nissan
(FB
CM
CM
Ford
GM
Ni ssan
(FB
MODEL
T-Blrd
280ZX
223)
280ZX
178)
Grand
Prix
Cutlass
Supreme
Pinto
Regal
280ZX
289)
Plymouth Reliant
Toyota
Celica
EMISSION
EGR, AIR,
FI, EGR,
FI, THC,
EGR, AIR,
EGR, AIR,
EGR, AIR,
RAIR, TWC
EFI, EGR,
EGR, AIR,
EGR, TOG,
CONTROL SYSTEM
OC
PAIR, OC
CL
OC, TWC, CL
TWC, CL
TWC, OC, FBC
, CL, FBC
TWC, CL, ECCS
CL, TWC, OC,
CLEFI
SALES CLASS
Calif.
49S
Calif.
49S
49S
Calif.
50S
, TC 49S
FBC 49S
50S
AC
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Eng. Size No. of
(Liters) Carb. bbls
5
2
2
4
4
2
3
2
2
2
.8 2
.8
.8
.3 2
.3 2
.3 2
.8 2
.8
.2 2
.6
No. of
Cylinders
8
6
6
8
8
4
6
6
4
6
Trans
A-3
A-3
M-5
L-3
A-3
A-3
A-3
A-3
M-4
A-40D
Inertia
WGT
4500
3000
3000
4000
3500
2750
3500
3000
2750
3000
-------
-47-
The mean FTP results of vehicles tested in EPA's CETF for temperatures
of 100, 75, 60 and 20°F are listed in Table 5. The average of the ten
vehicles tested so far exhibits CO emissions 4.6 times higher at 20°F
than at 75°F on a grams/mile basis. When compared to the certifica-
tion standard at which these vehicles were calibrated, the average is
2.7 times higher at 20°F than the standard. The average CETF FTP CO
vehicle emissions for 75, 60 and 20°F are 4.1, 7.6 and 17.9
grams/mile, respectively. These averages indicate larger CO 20°F/CO
75°F ratios than the levels needed to eliminate 90 to 100% of NAAQS
exceedences indicated in the Mobile 2 analysis (Figures 6 to 13).
-------
Table 5
CETF FTP Emissions and Fuel Economy Results (grams/mile)
CO (grams/mile)
Test Vehicle
Ford Thunderblrd
Datsun 280ZX FB 223
Da t sun 280ZX FB 178
Pontlac Grand Prix
Oldsmoblle Cutlass
Ford Pinto
Bulck Regal
Datsun 280ZX Turbo
Plymouth Reliant
Toyota Cellca Supra
CETF Fleet Average
Cert. Veh?
Yes
Yes
Yes
Yes
Yes*
No
No
Yes
Yes*
No
Cert. CO Standard 100°F
15 8.72
7 2.45
7 1.89
3.4 1.29
7 2.18
(15) 2.46
(7) 8.65
3.4 1.46
3.4 2.45
(7) 0.95
3.25
75°F
12.63
1.86
2.28
2.99
3.17
4.56
7.92
1.11
2.84
1.91
4.13
60QF
24.84
1.57
2.96
15.46
8.41
5.95
8.44
1.33
4.09
2.85
7.59
20°F
47.34
4.36
7.04
31.32
21.70
27.81
16.53
3.81
11.46
7.87
17.92
20°F/75QF
3.75
2.34
3.09
10.47
6.85
6.10
2.09
3.43
4.04
4.12
4.63
20 °F/ Cert. Standard
3.16
0.62
1.01
9.21
3.10
1.85
2.36
1.12
3.37
1.12
2.69
* Fuel economy vehicle. Not an emission data or durability vehicle.
-------
Table 5 (ConC.)
CETF FTP Emissions and Fuel Economy Reaults (grams/mile)
HC (grams/mile)
Test Vehicle
Ford Thunderblrd
Datsun 280ZX FB 223
Datsun 280ZX FB 178
Pontiao Grand Prix
Oldsmoblle Cutlass
Ford Pinto
Bulck Regal
Datsun 280ZX Turbo
Plymouth Reliant
Toyota Celica Supra
100°F
0.47
0.29
0.27
0.30
0.27
0.32
0.45
0.26
0.22
0.12
75°F
0.76
0.36
0.29
0.30
0.32
0.31
0.43
0.30
0.26
0.21
60°F
1.30
0.31
0.34
1.38
0.61
0.35
0.38
0.36
0.50
0.28
208F
3.06
0.58
1.00
3.69
1.40
1.30
0.56
0.73
1.68
1.25
100'F
0.73
1.52
0.37
0.90
0.48
0.64
0.29
0.53
0.52
0.37
NOx (grams/mile)
75°F
0.83
1.52
0.38
0.78
0.54
0.93
0.42
0.56
0.51
0.40
60°F
1.01
2.05
0.36
0.96
0.71
0.67
0.67
0.68
0.66
0.54
20°F
1.10
2.46
0.46
1.06
1.44
0.82
2.24
0.83
0.80
0.70
Fuel Economy (mpg)
100'P
13.30
19.57
20.70
18.30
19.65
24.15
21.05
19.45
25.33
22.35
75'F
13.10
19.35
20.30
17.75
19.10
21.05
20.25
18.85
25.25
21.73
60"F
12.55
19.45
20.30
16.65
18.25
21.80
20.50
18.95
25.30
21.15
20°F
11.45
18.50
18.90
15.15
16.85
20.00
18.25
16.75
21.90
19.15
-------
-50-
The highest CO emissions at all temperatures were exhibited by the
1979 Ford Thunderbird equipped with EGR, air injection and an
oxidation catalyst. The lowest CO emissions at FTP temperatures and
below were displayed by the 1981 1/2 Datsun 280ZX equipped with EGR,
three-way catalyst, closed loop electronic fuel injection, turbo-
charger and an electronic engine control system called ECCS. These
results seem to agree with the findings of Robert L. Farrell's
Temperature Correction Formulae for Adjusting Estimates ^f_ Emissions
from Automobiles [2 and section II. Summary of Significant Background
Studies]. In that report, the system that achieved the lowest bag one
FTP CO emissions was a fuel injection, three-way catalyst system.
The vehicle with the highest ratio of CO emissions at 20°F to those at
75°F is the 1981 Pontiac Grand Prix (CO 20°F/CO 75°F = 10.5) with EGP,
closed loop, three-way plus oxidation catalyst and air injection. The
vehicle with the lowest CO 20°F/CO 75°F ratio (2.1) is the 1980 Buick
Regal with a radial aperture tube air injection, closed loop
carburetor, and three-way catalyst system.
A number of the background studies used for this report also
determined values for cold temperature versus FTP temperature ratios.
In Emissions £t^ Off-Ambient Temperatures, the average CO 20°F/CO 75°F
FTP ratio for a fleet of 25 1970 model year vehicles is 1.82 [5]. In
the follow-up study, Effect of Ambient Temperature and Driving Cycle
on Exhaust Emissions, the new fleet performed worse than the 1970
baseline fleet at extreme temperatures when referenced to the 75°F FTP
CO levels (CO 25°F/CO 75°F = 3.31) [39].
The mean CETF CO emission results on the HFET and NYCC test cycles are
listed in Tables 6 and 7. Two tests were run on each vehicle at all
temperatures. All vehicles displayed HFET CO emission levels of less
than 1 gram/mile at 75°F and 20°F with the exception of the 1980 Buick
Regal and the 1978 Ford Pinto. Eight vehicles were tested on the
NYCC. This is a hot start NYCC, however, so the emissions levels are
much lower than might be expected if each vehicle were soaked
-------
Table 6
CETF HFET Emissions and Fuel Economy Results
CO (grams/mile)
Test Vehicle
Ford Thunderbird
Datsun 280ZX FB 223
Datsun 280ZX FB 178
Pontiac Grand Prix
Oldsmobile Cutlass
Ford Pinto
Bulck Regal
Datsun 280ZX FB 289
Plymouth Reliant
Toyota Celica Supra
100°F
2
4
0
0
0
0
3
0
0
0
.48
.02
.86
.05
.43
.08
.74
.44
.72
.19
75"F
0.84
0.10
0.94
0.05
0.23
0.04
4.96
0.48
0.24
0.15
60°F
0.09
0.02
1.00
0.10
0.19
0.04
1.90
0.46
0.34
0.09
20"F
0.38
0.13
0.99
0.07
0.15
1.89
1.19
0.50
0.39
0.11
100°F
20.2
25.5
33.2
26.6
25.8
31.2
30.4
23.8
35.2
29.7
FE
75BF
20.6
27.3
32.0
26.4
25.8
28.0
29.6
23.5
41.5
29.4
(mpg)
608F
21.1
27.2
32.4
26.1
25.8
31.4
30.9
24.3
41.4
30.1
20"F
20.0
27.2
32.2
25.3
25.8
29.9
30.2
23.6
40.8
29.1
Cert . CO
0.04
0.00
0.99
0.00
0.12
0.05*
-**
_***
0.13
-**
Cert. FE
20.2
26.5
31.0
26.0
24.8
-*
-**
-***
40.1
-**
* Southwest Research Institute data
** Production vehicle, no correlative data
*** 1^81 1/2 certification vehicle, not certification tested as yet
-------
Table 6 (Cont.)
CETF HFET Emissions and Fuel Economy Results
HC (grams/mile)
Test Vehicle
Ford Thunderblrd
Da t sun 280ZX
Da t sun 280ZX
Pontlac Grand
FB 223
FB 178
Prix
Oldsmoblle Cutlass
Ford Pinto
BuJck Regal
Da t sun 280ZX
FB 289
Plymouth Reliant
Toyota Cellca
Supra
100°F
0.11
0.05
0.07
0.04
0.05
0.06
0.14
0.05
0.04
0.01
75°F
0.14
0.03
0.08
0.04
0.05
0.03
0.14
0.05
0.02
0.01
60°F
0.14
0.03
0.08
0.11
0.05
0.03
0.06
0.05
0.02
0.01
20°F
0.16
0.04
0.07
0.07
0.06
0.04
0.04
0.05
0.03
0.01
100*F
0.53
1.25
0.20
0.67
0.35
0.79
0.14
0.77
0.54
0.11
NOx (grams/mile)
75°F
0.96
1.78
0.19
0.64
0.32
0.88
0.18
0.36
0.30
0.13
60BF
1.16
2.07
0.20
0.79
0.37
0.73
0.19
0.28
0.34
0.17
20°F
1.43
2.55
0.20
1.06
0.51
0.68
0.29
0.32
0.43
0.19
Cert. HC
0.14
0.03
0.08
0.03
0.06
0.02
_**
-***
0.02
-**
Cert. NOx
1.30
1.88
0.21
0.67
0.33
0.68
_**
_***
0.29
-**
* Southwest Research Institute data
** Production vehicle, no correlative data
*** 1981 1/2 certification vehicle, not certification tested as yet
-------
Tabla 7
HtCC gmiaaiona and EtaaJL Eeonoay Basalts*
CO (goM/oll*)
100*7
Datson 280ZX ?B 178
Ponciac Grand Prix
Oldsaobila Cudac*
Ford Pinto
Boick Regal
Dacsun 28023C Turbo
Plymouth Reliant
Toyota Callca Supra
5.35
3
3
20
4
11
5
.27
-
.65
.18
.97
.01
.30
73*7 60*7
1.04
3.14
-
1.28
-
0.99
3.32
2.43
2.68
0
3
1
18
0
6
2
.74
.78
.54
.20
.35
.25
.29
20*7
1.93
0.
.30
100*7
0.45
0.37
BC (grau/nila)
75*7
0.31
0.36
2.60
9.56
16.38
1.14
4.76
2.55
0.38
1.25
0.43
0.68
0.28
MOr (graaa/mila)
100 T
Dacsun 23023 FB 178
Ponclac Grand Prix
Oldsmobile Cuclaaa
Ford Pinto
Buick Regal
Dacsun 280ZC Turbo
Plymouth Reliant
Toyoca Calica Supra
0
1
0
0
1
0
0
.79
.07
-
.31
.48
.05
.41
.61
75"F
0.34
1.24
-
0.93
-
1.05
0.46
0.35
60
0
1
0
1
0
1
0
"7
.79
.46
.96
.12
.66
.49
.52
0.41
20*
T
1.16
1.
1.
1.
1.
1.
1.
0.
28
14
26
10
37
35
55
100 °7
11.50
10.50
-
13.60
11.90
12.20
13.30
13.05
0.28
-
0.30
0.74
0.10
60*7
0.48
0.98
0.40
0.32
0.92
0.29
0.60
0.11
20*7
0.28
0.24
0.39
0.63
0.71
0.27
' 0.50
0.09
FE (nog)
75'?
11.65
10.55
-
12.30
-
12.25
13.25
13.17
60 T
11.45
10.40
10.45
12.40
11.90
12.40
13.20
13.30
20 "7
11.45
10.30
10.60
11.25
11.60
11.67
3.90
12.65
*This NYCC is a. hot start cast. Vehicles were soaked at the test cenoerature only 10 minutes
after being run on a HFET ac the same soak temperature. (See Table 3 - Test Sequence.)
-------
-54-
several hours at the test temperature instead of 10 minute soaks
following a HFET.
The CETF FTP CO bag results are listed on a grams/mile basis in Table
8 and on a grams/bag basis in Table 9. Table 8 shows that bag one
cold start emissions are the major CO contributors with 81% of total
FTP CO emissions at 75°F and 92% at 20°F.
Emissions from bag one of the FTP at 20°F increased an average of 5.3
times over the 75°F grams/mile levels. In the CO Hot Spot Study, it
was found that with soak temperatures of 10°F to 25°F, CO exhaust
emissions from bag one of the FTP increased by factors ranging from 3
to 7 over identical tests following 68°F to 80°F soaks for 1976 model
year vehicles [15]. On a grams/bag basis (Tcble 9), the bag one
portion again rose from 80% to 92% of the total CO emissions from 75°F
to 20°F as the average bag one contribution climoed from about 50 to
275 grams.
The relationships between FTP HC, CO and NOx emissions and fuel
economy versus temperature for all vehicles tested in the CETF are
shown in Figures 14-21. Emissions data from a data base complied by
Vector Research, Inc. in a report entitled Draft Data Base for the
Development of Improved Temperature Correction Factors for Emissions,
and CETF CO emissions data were examined to see if a linear
relationship exists between CO emissions at various temperatures and
CO emissions at 75°F. Vehicle CO emissions were plotted for 0, 20,
40, 60, 80 and 100°F versus CO emissions at 75°F. Additionally, CETF
vehicle CO emissions were plotted for 20, 60 and 100°F versus CO
emissions at 75°F (see Figures 22-24). Table 10 lists the statistical
analysis of these plots. A correlation coefficient (R) of 1 indicates
a linear relationship exists, whereas values close to zero indicate no
relationship exists between the two variables.
The results are for the most part, inconclusive. However, it can be
said that in model years preceding 1972 there is very little
-------
Table
CETF FTP CO Bag Results (grams/mile)
1979 Ford T-Bird
Bag 1
Bag 2
Bag 3
1980 Datsun 280ZX (FB 223)
Bag 1
Bag 2
Bag 3
1980 Datsun 280ZX (FB 178)
Bag 1
Bag 2
3
1981 Pontiac Grand Prix
Bag 1
Bag 2
Bag 3
1980 Oldsmobile Cutlass
Bag 1
Bag 2
Bag 3
1978 Ford Pinto
Bag 1
Bag 2
Bag 3
1980 Buick Regal
Bag 1
Bag 2
Bag 3
75 °F
49.39
0.75
7.61
7.63
0.04
1.08
7.34
0.94
1.03
9.96
0.25
2.93
13.69
0.30
0.75
20.36
0.15
1.10
16.31
5.12
6.93
PCX
86%
1%
13%
87%
1%
12%
79%
10%
11%
76%
2%
22%
93%
2%
5%
94%
1%
5%
58%
18%
24%
20°F
203.44
6.27
7.96
20.76
0.00
0.31
29.11
1.29
1.48
150.19
0.16
1.00
102.85
0.55
0.65
93.09
5.73
20.53
60.17
5.27
4.98
PCT
93%
3%
4%
99%
0%
1%
91%
4%
5%
99%
0%
1%
99%
0%
1%
78%
5%
17%
85%
8%
7%
-------
Table 8 (Cont'd)
CETF FTP CO Bag Results (grams/mile)
1981 1/2 Datsun 280ZX (FB 289)
Bag 1
Bag 2
Bag 3
1981 Plymouth Reliant
Bag 1
Bag 2
Bag 3
1980 Toyota Celica Supra
Bag 1
Bag 2
Bag 3
CETF Fleet Average
Bag 1
Bag 2
Bag 3
75 °F
3.16
0.60
0.55
8.59
0.64
2.70
8.11
0.55
0.53
75°F
14.45
0.93
2.52
PCT
73%
14%
13%
72%
5%
23%
88%
6%
6%
PCT
81%
5%
14%
20°F
15.65
0.59
0.98
51.16
0.49
2.57
36.29
0.48
0.55
20°F
76.27
2.08
4.10
PCT
91%
3%
6%
94%
1%
5%
97%
1%
2%
PCT
92%
3%
5%
-------
Table 9
CETF FTP CO Bag Results (grams/bag)
1979 Ford T-Bird
Bag 1
Bag 2
Bag 3
1980 Datsun 280ZX (FB 223)
Bag 1
Bag 2
Bag 3
1980 Datsun 280ZX (FB 178)
Bag 1
Bag 2
Bag 3
75°F
177.29
2.95
27.33
26.94
0.05
3.89
PCT
85%
2%
13%
87%
0%
13%
20°F
730.36
24.71
28.57
74.90
0.00
3.10
PCT
93%
3%
99%
0%
1%
26.39
3.67
3.69
78%
11%
11%
104.69
5.10
5.32
91%
4%
5%
1981 Pontiac Gran Prix
Bag 1
Bag 2
Bag 3
1980 Oldsmobile Cutlass
Bag 1
Bag 2
Bag 3
1978 Ford Pinto
Bag 1
Bag 2
Bag 3
1980 Buick Regal
Bag 1
Bag 2
Bag 3
35.68
0.98
10.51
49.23
1.18
2.72
73.08
0.59
3.98
58.29
19.89
24.79
76%
2%
22%
93%
2%
5%
94%
1%
5%
57%
19%
24%
538.59
0.62
3.60
367.19
2.14
2.32
333.93
22.29
73.81
217.13
20.59
17.89
99%
0%
1%
99%
0%
1%
78%
5%
17%
85%
8%
7%
-------
Table 9 (Cont'd)
CETF FTP CO Bag Results (grams/bag)
1981 1/2 Datsun 280ZX (FB 289)
Bag 1
Bag 2
Bag 3
1981 Plymouth Reliant
Bag 1
Bag 2
Bag 3
1980 Toyota Celica Supra
Bag 1
Bag 2
Bag 3
CETF Fleet Average
Bag 1
Bag 2
Bag 3
75°F
11.34
2.33
1.97
30.54
2.46
9.70
26.63
1.97
1.56
75°F
51.54
3.61
9.01
PCT
72%
15%
13%
71%
6%
23%
88%
7%
5%
PCT
80%
6%
14%
20°F
56.26
2.30
3.53
181.88
1.89
9.00
130.16
1.87
1.98
20°F
273.51
8.15
14.71
PCT
90%
4%
6%
94%
1%
5%
97%
1%
2%
PCT
92%
3%
5%
-------
CETF DRTR 5UMMRRY
FTP H
-------
I.
CETF DRTR 5UMMRRY
FTP HC V5 TEMP
Ld
J
\
in
£
ID
1.20
B.BH
B.CB
Q.HB
-f DATSUN TURBO FB289
X DATSUN 280ZX FB178
A DATSUN 280ZX FB223
D FORD PINTO 8R10Y131366
O TOYOTA CELICA SUPRA MA461001S3
M
n
n
IQ 23 3H HSJ 53 E0 70
TEMPERRTURE CF>
EC! S3 1EO
3/lB/GEl
-------
u
J
z
\
Ul
z
in
a
CETF DRTR GUMMRRY
FTP
-------
U
J
\
in
a
23. Q
2H.E3
20. H
IE.13
12.Q
B.D
H.H
IB
CETF DRTR EUMMRRY
FTP
aa IBQ
3/IB/BEJ*
Tl
M
O
I
-------
2.50
CETF DRTR 5UMMRRY
FTP NDX V5 TEMP
U
J
Z
\
in
z
m
x
D
Z
2.00
l.EH
I .03
0.S0
0
-f- BUICK REGAL 4M47AAH202725
X PONTIAC GRAND PRIX P0943
A FORD THUNDERBIRD 9S1-5.8M-H-400
D OLDSMOBILE CUTLASS 03H2-94472C
O PLYMOUTH RELIANT DIG2
X
o
e:
n.an
10 23 33 H3 SB B2 70
TEMPERRTURE CF>
ea
90
IBB
-------
2.SB .
CETF DRTR 5UMHRRY
FTP NDX V5 TEMP
U
J
Z
\
Ul
£
•LD
X
D
Z
2.0£J
I .ED
I.BB
a.sa
O TOYOTA CELICA SLPKA MA4610018J
+ DATSUN TURBO FB289
X DATSUN 280ZX FB178
A DATSUN 280ZX FB223
D FORD PINTO 8R10Y131366
-A
H.BB
0
10 20 30 MB SB BB 70 BB 50
TEMPERRTURE CF>
IB 13
-------
ID
0,
Z
E
P
Z
D
V
LJ
J
U
U
L
CETF DRTR 5UMMRRY
25.0 ».
FTP FE V5 TEMP
-0
2B.B
1S.0
o
I
C5
ID.B
S.B
+ BUICK REGAL AM47AAH202725
X PONTIAC GRAND PRIX P09A&
A FORD THUNDERBIRD 9S1-5.8M-H-4CO
D OLDSMOBILE CUTLASS 03H2-94472C
O PLYMOUTH RELIANT D162
B.B
IB 20 33 H3 SH B3 ' 70
TEMPERRTURE CF>
Q3
SB
IBB
-------
LD
H
z
a
~z.
.a
u
u
j
u
n
L
25.0
20.0
is.0
10..0
5.0
B.B
CETF DRTR 5UMMRRY
FTP FE VS TEMP
-f- DATSUN TURBO FB289
X DATSUN 280ZX FB178
A DATSUN 280ZX FB223
D FORD PINTO 8R10Y131366
O TOYOTA CELICA SUPRA MA461001b3
10 20 30 M0 S0 E0 70
TEMPERRTURE CF
00 H0 100
3/1 B/EJd
-------
FlGURe 22.
Ul
1L
o
0
-------
FIGURE 23.
0
O
tO
COMPARISON OF CETF CO EMISSIONS
AT 60°f= KELATIVE TO CO
EMISSIONS AT 75°F O^4 THE FTP
SO
20
2. 4 e 8 to
co AT TS°F
12 14. 1
/a 20
-68-
-------
24:
U.
O
O
O
COMPARISON OP CETF CO EMISSIONS
AT IOO°F RELATIVE TO CO
EM155IOMS AT ~75*F ON THE FTP
3?
lO
5"
A
2 4- € 8 10 IZ i4 t
-------
Table 10
CORRELATION OF VRI DATA VEHICLE* CO EMISSIONS
AT VARIOUS TEMPERATURES RELATIVE TO CO
EMISSIONS AT 75°F
Model Year
Correlation
Group
PRE 1968
1968-1969
1970-1971
1972-1974
»
»
1975-1979
••
••
••
••
»
1980-1981
••
Temperature(°F)
vs. 75°F
20
20
20
20
40
100
0
20
40
60
80
100
20
60
100
No of
Vehicles
3
4
7
11
3
2
9
27
18
8
5
8
10
10
10
Coefficient(R)
0.39
0.81
0.02
0.83
0.99
-
0.11
0.89
0.96
0.48
0.13
0.87
0.78
0.83
0.92
R2
0.15
0.65
0.00
0.69
0.99
-
0.01
0.79
0.91
0.23
0.02
0.75
0.61
0.68
0.85
Significance
0.74
0.19
0.06
0.00
0.07
-
0.77
0.00
0.00
0.23
0.83
0.02
* Vehicles from several test programs were merged in a
data base by Vector Research, Inc. (VRI). Reference
42 contains descriptions of the test programs and vehicles.
-------
-71-
relationship between CO emissions at non-FTP temperatures and CO
emissions at 75°F. From 1972 on, a linear relationship exists in most
cases, especially the cases with the most data. Intuitively, one
might expect a better relationship at temperatures closer to 75°F,
with a decreasing relationship as one compares temperatures closer to
0°F. Such is not the case with 1975-1979 vehicles by examination of
the correlation coefficients at 80°F and 0°F. CETF CO emissions data
(1980-81 model year group) show fairly high correlations at all
temperatures relative to 75°F.
This analysis is useful in evaluating whether controlling CO emissions
at different levels at FTP temperatures affects CO emissions levels at
non-FTP temperatures. Unfortunately, the amount of data available to
date is insufficient to make definative conclusions.
-------
-72-
VI. REFERENCES
1) A Review of Carbon Monoxide Emissions from Motor Vehicles During
Cold Temperature Operation, Alaska Department of Environmental
Conservation, (November, 1978), revised March, 1979.
2) Temperature Correction Formulae for Adjusting Estimates of
Emissions from Automobiles, by Vector Research, Inc., EPA Contract
No. A-2098-NASX, Robert L. Farrell, Volumes I & II, September,
1979.
3) EPA memo from Karl Hellman to Robert Garbe entitled Analysis for
Improved Emissions vs. Temperature Influence, April 3, 1979.
4) Effect of Ambient Temperature on Vehicle Emissions and Performance
Factors, by Gulf Research and Development Co., R. S. Spindt, R. E.
Dizak, R. M. Stewart, W. A. P. Meyer, (Appendix A, B, C), Report
No. EPA-A60/3-79-006A, September, 1979.
5) Emissions at Off-Ambient Temperatures, W. F. Marshall, B. H.
Eccleston, DOE, Bartlesville, Energy Technology Center, OK, SAE
Paper No. 800512, presented at the Congress and Exposition,
Detroit, MI, February 25-29, 1980.
6) Methodologies for Projecting the Relative Air Quality Impacts of
Emission Control Strategies, James H. Wilson, Jr., U.S. EPA,
Durham, N.C., presented at the 71st Annual Meeting of the Air
Pollution Control Association (APCA), Houston, Texas, June 25-30,
1978.
7) The Exhaust Emission and Fuel Consumption Characteristics of an
Engine During Warmup-A Vehicle Study, Donald J. Pozniak, General
Motors Research Labs, SAE Paper No. 800396, presented at the
Congress and Exposition, Detroit, MI, February 25-29, 1980.
-------
-73-
8) Microprocessor Control Brings About Better Fuel Economy With Good
Driveability, Kenji Ikeura, Akio Hosaka, and Founeomi Yano, Nissan
Motor Co., SAE Paper No. 800056, presented at the Congress and
Exposition, Detroit, MI, February 25-29, 1980.
9) Study to Determine Accuracy of the 10% Load Factor for Air
Conditioning, Engineering Evaluation Section, Engineering Branch,
Vehicle Emission Control Division, California Air Resources Board,
June, 1978.
10) An Upper Limit Analysis of Carbon Monoxide Monitoring in the
United States, R. L. Ferrari & K. H. Jones, Ph.D., Council on
Environmental Quality (CEQ), APCA Journal, June 15, 1979.
11) Projections of Carbon Monoxide Exposure Risk to Stable Angina
Pectoris Patients in Four Cities, K. Jones, T. Chapman, R.
Ferrari, & N. Buskwick, CEQ, January, 1979.
12) A Continuous Stress Function Relating Ambient Carbon Monoxide
Exposure and Excess Angina Pectoris Attacks, K. H. Jones, CEQ,
John Knelson, EPA, Research Triangle Park (RTP), N.C., 1979.
13) Effect of Cold Weather on Motor Vehicle Emissions and Fuel
Consumption II, Nicolas Ostrouchov, Mobile Sources Division,
Environment Canada, SAE Paper No. 790229, presented at the
Congress and Exposition, Detroit, MI, February 26 - March 2, 1979.
14) The Development of Improved Temperature Correction Factors for
Emissions, Purchase Order by Jay Wallace, EPA, Ann Arbor, MI,
April 13, 1979.
15) CO Hot Spot Preliminary Investigation, Gregg Service, Testing and
Evaluations Branch (TAEB), Emission Control Technology Division
(ECTD), Office of Mobile Source Air Pollution Control (OMSAPC),
EPA, December, 1977.
-------
-74-
16) Ambient Temperature Effects on Exhaust Emissions of Closed-loop
Emission Control System Vehicles, T.M. Fisher, Director,
Automotive Emission Control, GM, July 19, 1978.
17) Emissions Under Non-FTP Temperature and Speed Conditions,
Characterization and Applications Branch (CAB), ECTD, EPA, Lois
Platte, July, 1978.
18) Effects of Low Ambient Temperature on the Exhaust Emissions and
Fuel Economy of 84 Automobiles in Chicago, TAEB, ECTD, EPA, Wayne
Heinmiller, October, 1978.
19) Control Techniques for Carbon Monoxide Emissions, Emission
Standards and Engineering Division (ESED), Office of Air Quality
PI inning and Standards (OAQPS), RTP, EPA-450/3-79-006, June, 1979.
20) Mooile Source Emission Factors (For Low Altitude Areas Only) EPA,
v\
Office of Transportation and Land Use Policy (OT&LUP),
EPA-400/9-78-006, March, 1978.
21) Sensitivity of Exhaust Emissions from Late Model Year Passenger
Cars Under Non-FTP Conditions, Gary T. Jones, CAB, ECTD, EPA,
CAB-9/Non-FTP-2, September, 1978.
22) Rationale and Discussion of Potential Highway Cycle Emissions
Guidelines, Gary T. Jones, CAB, ECTD, EPA, CAB-9/Non-FTP-2,
September, 1978.
23) Ambient Pollutant Violations Versus Average Temperature, Lois
Platte, CAB, ECTD, EPA, August 4, 1978.
24) Ambient Temperature Effects on Exhaust Emissions of Closed Loop
Emission Control System Vehicles (Additional Data), Gerald J.
Barnes, Director, Automotive Emission Control, GM, December 20,
1978.
-------
-75-
25) Transmittal of Draft 202(a)(4) Advisory Circular, Charles L. Gray,
Director, ECTD, EPA, September 12, 1978.
26) Transmittal of GM 1980-X Emission Summary, Craig Harvey, ECTD,
EPA, September 6, 1978.
27) GM Micro-Computer Engine Control System, R.A. Grimm, R.J. Bremer,
S.P. Stonestreet, GM Emission Control Project Center, SAE Paper
No. 800053, presented at the Congress and Exposition, Detroit, MI,
February 25-29, 1980.
28) Letter from M.E. Rivers of Environment Canada to Pon Bright of
Ford regarding HC, CO Step Increase at Low Temperature, December
7, 1979.
29) Memo from Mark Wolcott, TEB, EPA to Charles Gray, Director, ECTD,
EPA entitled The Relationship Between Ambient Carbon Monoxide and
Temperature, April 17, 1980.
30) Memo from F.P. Hutchins, TAEB, EPA to Charles Gray, Director,
ECTD, EPA entitled Preliminary Results of the Temperature Effects
Investigation at Gulf Research, EPA, October 3, 1978.
31) National Emission Inventory Interim Report Number 1, Mark Wolcott,
TEB, EPA, February 26, 1980.
32) Test Results of 4 Vehicles Regulated and Unregulated Emissions
Under Non-FTP Conditions, Jim Braddock, Mobile Source Emissions
Research Branch (MSERB), Office of Research and Development (ORD),
EPA, May 2, 1980.
33) Memo from F.P. Hutchins, TAEB, EPA to Charles Gray, Director,
ECTD, EPA, entitled Michael P. Walsh's Request for Comments on
Environment Canada Letter to Ford (HC & CO Step Increase at Low
Temperature), January 18, 1980.
-------
-/ 0-
34) Vehicle Emissions and Fuel Consumption in Canadian Winter
Temperatures, N. Ostrouchov, Environment Canada, presented at the
73rd Annual Meeting of the Air Pollution Control Association,
June 22-27, 1980.
35) Methodologies for Projecting the Relative Air Quality Impacts of
Emission Control Strategies, James H. Wilson, Jr., U.S. EPA,
presented at the 71st Annual Meeting of the Air Pollution Control
Association, June 25-30, 1978.
36) Compilation of Air Pollution Emission Factors; Highway Mobile
Sources, OMSAPC, EPA, March 1981.
37) Urban-Center CO Air Quality Projections, T.V. Chang, J.M. Norbeck
and B. Weinstock, Ford Motor Co., Journal of the Air Pollution
Control Association, Vol. 30, No. 9, September, 1980.
38) Draft Advisory Circular on non-FTP Conditions, Compliance with the
Requirements of Section 202(a)(4) and 206(a)(3) of the Clean Air
Act for 1980 and Later Model Years, September 19, 1978.
39) Effect of Ambient Temperature and Driving Cycle on Exhaust
Emissions, W.F. Marshall, U.S. DOE, Bartlesville, Oklahoma, EPA
460/3-80-012, June, 1980.
40) Evolution of Federal Light-Duty Mass Emission Regulations, Thomas
A. Huls, Environmental Protection Agency, SAE Paper No. 730554,
May 14-18, 1973.
41) Environmental Quality 1979; The Tenth Annual Report to the
President of the Council on Environmental Quality, December 1979.
-------
-77-
42) Draft Data Base for the Development of Improved Temperature
Correction Factors for Emissions, Vector Research, Inc., VIR-EPA-5
FR79-1, July 27, 1979.
43) Draft Air Quality Criteria for Carbon Monoxide, Environmental
Criteria and Assessment Office, U.S. Environmental Protection
Agency, Research Triangle Park, NC, April 1979.
-------
VII. APPENDIX
-------
Table A-I
1973 CO Gi'idalinas frotn T)raft Advisory Circuli:
on non-FTP Coadj.to.oos [38]
Ratio of Value Over FTP at Specified Temperature to Value
of Emission Standard at 75°
Cycle
Temperature (°F)~
20 50
75
NYCC
FTP
HFETT
2.30
1.76 1.38 1.00
0.42
Cycle
NYCC
FTP _
ETE'T
1980 Model Year LDVs CO Guidelines (gpm)
Temperature (°F)
20 50 75
12.3
9.7
16.1
7.0 (7.4 )
2.9
Cycle
NYCC
FTP .
1981 and Later Model Year LDVs CO Guidelines (gpm)
Temperature (°F)
20 50 75
6.0
4.7
7.8
3.4(3.6^)
1.4
Temperature tolerance for a valid test should be -f 5°F.
NYCC is started frota a hot idle.
The HFET is the EPA highway fuel economy test.
-------
This denotes guideline with air-conditionirg. These guidelines for
air-copcli^ioriaj for HC, CO, tnJ I'Ox should o? vred for any engine
family where air-conditioning is expected to be sold on over 33/1 of tha
•number of vehicles produced. The 75° air-conditioning factor refers tov
the ratio of a:b where "a" is the emission level on a hot start 1972 FTP
at 75° with vehicle air-conditioning on (as described below) -without the
10% road load air-conditioning simulation, and "b" is the emission level
on the 1975 FTP at 75° without vehicle air-conditioning, but with the
10% road load air-conditioning simulation. The 110° air-conditioning
factor refers to the ratio of c:d where "c" is the emission level on the
1975 FTP at 110" with vehicle air-conditioning on (as described below)
without the 10% road load air-conditioning simulation, and "d" is the
emission level on the 1975 FTP at 75° without vehicle air-conditioning
and without the 10% road load air-conditioning simulation. The air-
conditioning tests should be conducted with the air-conditioning control
set at the coolest setting, the interior air recirculated, the fan set
to the highest speed, and the refrigerant charged as recommended by "he
•manufacturer.
References: Kutchins, F.P., 3 internal EPA memos: "Effects of Air
Conditioning Operation on Older Cars' Emissions (67-72-1IY)", 5/23/7S,
"Modification of Air Conditioning Factors", 6/1/78, "Air Conditioning
Effects Factors in the Draft Advisory Circular", 6/16/78. These are
based on two studies: "A Study of Emissions from 1967-1974 Light T)u<-y
Vehicles in Denver, Houston, and Detroit", (EPA-460/3-74-015) , and
"Ambient Temperature and Vehicle Emissions", (EPA-460/3-74-028).
-------
rIGURE A-l
MOBILE 1
BB3E CB3E 1975-1877 BPU2 33-002
100
SO
202 EE
SSZ EE
•««u 632 EE
1002 EE
N.
o
CO
o
F so
10
o 30
20
10
t
1
_L
1 t
\
1
_L
1
t
f
3 tl 13 IS 17 13 21 23 2S 27 20 Si S3 3S 37 39
PKE-RUSII TEH? CO.51 - 15, C70-75) ~ 22
-------
FIGURE A-2
MOBILE 1
BR3E C33S 1375-1377
33-05X
100
SO
,00
302 EE
E52 EE
83X EE
1002 EE
- TO
•x.
CO
SO
e-»
r>
ac
ui
SO
10
1 i
l 1 T t t
J
1 1
T Til 1 t 1
9 \\ 13 IS t7 10 21 23 2S 27 23 31 S3
PRE-nU3H !£»?• CO. 5) - 15; C70-7SJ - 23
35 37 S3 Ul
-------
FIGURE A-3
MOBILE 1
Baas.
1375-1377 p?^ 33-oox
too
20
so
CO
o
r-
o
c:
50
US
ui 30
802 EE
aai352 £E
-uiru 332 EE
I COX EE
to
1 I » »
J 1
J T t T
J L
T T.
3 11 13 IS 17 J2 "21 2S 23 27 23 31 S3 3S 37 SO
PRE-HU3H 1£r? CO.SI •> ISt C70-7S1 - 2a
-------
FIGURL A-4
HOBILE 1
SB3E CR3E 1375-1977
33-052
100
so
302 EE
oa»3S2 EE
332 EE
1002 EE
70
BO
o
»-
o
50
o
•"•»
trt
SO
20
10
w *>"»- V-.1'
* - -- ^>
-------
FIGURE A-5
MOBILE 2
Bfl3£ CB3E 1375-1377 HI/M 33-002
100
3D
SO
50
20
aor EE
3S2 EE
33X EE
IDOZ EE
_L
J_
J L
J_
J_
J_
3 II 13 IS 17 13 21 23 25 27 23 31 33 3S 37 33
PBE-RU3H TEMP CO.5) - I5t C70-7S) - 23
-------
FIGURE A- 6
MOBILE 2
BB3E CB3E 137S~1377
33-05Z
100
30
80
~ 70
60
SO
"30
20
10
J.
r
t t
t
_L
J L
3 11 13 IS 17 13 21 23 25 27 23 31 33 3S 37 33
PRE-HU3H 7E«f* CO.5) - 1S» C70-7S1 - 23
-------
A-7
MOBILE 2
BB3E CH3E 1375-1377 NJt/H 33-002
100
30
SO
70
*» SO
so
2 no
UJ
•J 3D
20
C3C3I
10
1 1
1 > 1 1
J L
J L
J L
3 11 13 15 17 13 21 23 2S 27 29 31 33 35 37 33 «4 3
f»RE-fTJ3H 7EHP CO.SI - ISv C70-7S) - 23
-------
FICURx. A-8
MOBILE 2
BB3S. CB3E 1375-1377 NI/M
100
30
30
S 70
•v.
O
rr, 50
50
tn
Z uo
ui
u> 30
sor EE
3Sr EE
332 EE
1002 EE
*•—-.1
I »
J L
J ! I I t t T T t t
> T
3 II 13 IS 17 IS 21 23 25 27 23 31 33 35 37 33
PRE-HU3.1 TEMP CO.SI - ISt C70-7S) - 23
-------
Figure A-9
Pre-Rush Hour Temperature Distribution of Observed L:.ce?dences
of the National Anbient Air Duality Standard for Carbon x'onn-:ide
10
H 8 H3 CO
2300
auoo
.2000
1600
>•
cs
1.1
1200
800
y
------- |