Technical Report
Effects of Idle Warm-up and Warm Idle Soak
Periods on CO Emissions and Fuel Consumption at 20°F
Tom Darlington
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.
Inspection and Maintenance Staff
Emission Control Technology Division
Office of Mobile Source Air Pollution Control
Office of Air, Noise, and Radiation
U.S. Environmental Protection Agency

-------
Effects of Idle Warm-Up and Warm Idle Soak
Periods on CO Emissions and Fuel Consumption at 20°F
1.0 Summary
The effects of idle warm-up and warm idle periods on CO emissions and fuel
consumption were studied on two vehicles at 20°F. Both vehicles were Ford
vehicles which had air pump systems that routed pump air to the atmosphere
after 1-2 minutes of engine idling, and as such,these vehicles are not
representative of any particular fleet. The testing revealed that (l) there
is a CO benefit and a fuel economy penalty associated with a short (less than
10 minutes) idle warm-up period which preceeds actual driving; (2) there is a
fuel economy penalty associated with leaving a vehicle running for 5-15
minutes as compared to turning the engine off for the same period of time and
restarting it, and (3) a different and more comprehensive test program would
be needed to test the effects of warm idle periods on CO emissions.
2.0 Background
The 1975 Federal Test Procedure (FTP) requires vehicle exhaust emissions to be
measured in three separate phases of a 31 minute typical urban driving cycle
at 75°F. The distinction between phases is made to characterize the emissions
produced from different modes of engine operation such as cold starting and
engine warm-up, stabilized operation, and hot starting.

-------
2
As one would expect, the first phase, which includes engine starting and
warm-up, is the phase that produces the most CO emissions during an FTP test.
This effect becomes more pronounced as test temperature is decreased. There
are several reasons for this phenomenon. First, at low temperatures, engine
starting requires carburetor choking because of the low volatility of gasoline
at low temperatures. Consequently, much of the gasoline entering the engine
is not completely burned. Second, low temperatures also result in longer
engine cranking times than would be experienced at s&warmer temperature~ Like
choking, long cranking times can add more low-volatility fuel to the engine,
resulting in more incompleted burned combustion products than there would have
been with a shorter cranking time. Third, the choke does not completely open
up at the instant the vehicle is started. Depending on what other choke
controls are present on a vehicle (electric assist, thermostatic coil, etc.)
it^may take several minutes for the choke to completely open up. This further
\
exacerabates /the incomplete combustion products problem. Finally, internal
fricfeiofi in the drive train and the power required to drive the accessories
(heater, wipers) are higher, requiring greater power output from the engine
during warm-up.
When these effects are coupled with the fact that during engine starting and
warm-up the catalyst is cold and unable to operate effectively, high CO
emissions at the tailpipe are typically the cold start and warm-up result.
The aforementioned reasons are usually given as an explanation of the causes
of excessive low temperature CO emission from vehicles. However, driver
behavior may also have some impact on the amount of CO emissions produced by

-------
3
vehicles at cold temperatures. In extreme cold weather (i.e., 20°F or less)
many people warm up their vehicles at idle prior to taking a trip (to work,
shopping, etc.). Presumably people feel this increases driving comfort, but
may also feel it increases fuel consumption. The 1975 FTP has only a 10
second cold idle warm-up included in the driving cycle. The question raised
by this first aspect of cold weather driving is: What are the overall CO
emission and fuel economy effects of this cold weather anomaly, i.e., are the
total emissions produced and fuel used during an i Ms warm-up and a typical
trip greater or less than the total emissions and frael used from a typical
trip without an idle warm-up?
The other aspect of cold weather vehicle operation is the extended warm idle..
In extreme cold weather operation, many people leave their cars running
instead of turning them off when performing a short errand that takes them out
of their car (such as running into a grocery store for 10 minutes). People
feel this increases driving comfort, and may also feel it saves a small amount
of battery power for times when the vehicle really needs it — during an
extreme cold start. The 1975 FTP contains 252 seconds (out of 1371 total
seconds or 18.4%) of idle time in the driving cycle. This accounts for idle
time which is incurred at stop lights and stop signs. But it does not take
into account the emissions produced during extended (5-15 minutes) warm idles
which may be an integral part of cold weather driving habits. The question
raised by this second aspect of cold weather driving is: Are total CO
emissions produced during a warm idle plus a typical trip greater or less than
the emissions produced from a typical trip without a prior warm idle?

-------
4
3.0 Program Description
The test program used to determine the effects of idle warm-up and warm idle
periods on CO emissions and fuel consumption is presented in Figure 1.
Figure 1. Idle Effects Test Sequence
Figure A. Test Sequence
1.	Check tune-up specifications on vehicle
2.	FTP at 75°F
3.	Cool to 20°F
4.	Idle warm-up (2, 5, 10 minutes)
5.	LA-4
6.	Engine-off soak (5, 10, 15 minutes)
7.	LA-4
8.	Cool to 20°F
9.	Cold start LA-4
10.	Idle Soak (5, 10, 15 minutes)
11.	LA-4
12.	Repeat steps 3-11 for next idle warm-up and
warm idle period.
Steps 1 and 2 were performed for the purpose of assuring that each vehicle was
operating properly. Steps 3 through 11 are the core of the test program;
emissions produced from the tests in Steps 4 and 5 were compared to emissions

-------
5
produced from the test in Step 9 to determine the overall effect of an idle
warm-up period on a typical first trip of the day. Similarly, emissions
produced from the tests in Steps 6 and 7 were compared to emissions produced
in Steps 10 and 11 to determine the effect of warm idle period on a typical
subsequent trip.
Idle warm-up periods were 2, 5, and 10 minutes in duration; warn idle periods
were 5, 10, and 15 minutes in duration. These times were selected on the
basis of a 1980 Alaska Department of Environmental Conservation survey which
studied idle warm-up and warm idle times among people who participated in the
1979-80 Anchorage Free Emission Control Test. The results of this survey are
presented in Attachment 1 of the Appendix.
Vehicles - The two vehicles used in this study were a 1978 Ford LDT (8
cylinder) and a 1978 Ford Pinto (4 cylinder - certified for California). Both
vehicles had air pumps, and both vehicles had an air pump control system which
routed pump air to the atmosphere (instead of to the exhaust manifold) after
1-2 minutes of engine operation at idle. Specifications for both vehicles are
listed in Attachment 2 of the Appendix. Both vehicles were tuned to
manufacturer specifications prior to testing.
Controlled Environment Test Cell - The vehicles were tested in the Controlled
Environment Test Cell (CETC) located in EPA's Motor Vehicle Emission
Laboratory in Ann Arbor, Michigan. TThe cell contains an electric dynamometer
for simulating vehicle loads and inertia weights and a constant speed fan for

-------
6
engine cooling. A diagram of the cell is presented in Attachment 3 in the
Appendix. The cell is capable of maintaining any temperature between 20°F and
100°F throughout the duration of a test.
The starting procedures used during testing were the manufacturers recommended
cold start and warm start procedures. The cold start procedures for both
vehicles required the driver to set the choke with the accelerator pedal prior
to turning the key. The warm start procedures requi'^d the driver to turn the
key without depressing the accelerator pedal.
Since a proportional fan control (where fan speed is coupled to dynamometer
roll speed to simulate the actual velocity of cooling air passing over the
vehicle) was not available at the time of testing, the fan speed was set to
simulate a velocity of about 20 mph, which is the average vehicle speed of the
driving cycle of the 1975 FTP. A divider was placed in front of the vehicle
to prevent 20 mph cooling air from the fan from passing over the vehicle when
idle emissions were sampled.
In most instances a vehicle was allowed to "soak" overnight at the test
temperature (20°F) so that there was reasonable assurance that all components
(tires, oil, etc.) of a vehicle were at the test temperature. However, in
order to expedite testing, a "forced cool down" technique was sometimes
employed. In this cooling technique, fan speed was increased to approximately
55 mph while cell air was maintained carefully at 20°F. This reduced the
"soak time" necessary to cool all components (including engine crankcase oil,

-------
7
which is the slowest component to cool) to 20° from about 12 hours (with no
fan cooling) to about 4 hours. Vehicles were determined ready for testing
when their engine crankcase oil temperature reached a value of 20_+ 2°F.
4. Results
Equipment Modifications - When the first 2 minute idle warm-up bag was
analyzed, it was discovered that the CO analyzer Mn its highest operating
range was partially saturated at approximately 135 meter deflections (an
adequate analysis can be obtained only up to about 125 meter deflections). A
higher CO range (range 23, 0-25,000 ppm CO) was installed to alleviate this
problem.
Emissions From Idle Periods - CO emissions from the test vehicles during the
idle periods are summarized in Table 1. It should be noted that although CO
emissions increased with longer idle warm-up periods, the same was not true
with respect to the warm idle periods. Possible reasons for these results are
discussed in the conclusions.

-------
8
Table 1. CO Emissions From Idle Periods
Test
Pinto
LTD
2 min. Idle Warm-up
5 min. Idle Warm-up
10 min. Idle Warm-up
79.A grams
240.2
286.3
86.9 grams
622.6
734.9
5 min. Idle Warm-up
10 min. Idle Warm-up
15 min. Idle Warm-up
1.3 grams
.24
14.8
.43 grams
.12
.53
Idle Warm-up Effects on Emissions and Fuel Consumption - The average (of both
vehicles) emissions produced and the average fuel consumed for the idle
warm-up test as opposed to the cold start test are presented in Table 2. It
is evident from this table that there is a CO benefit and a fuel economy
penalty associated with a short (10 minutes or less) idle warm-up period which
preceeds an LA-4. There also appears to be a small HC benefit to an idle
warm-up of 2 minutes or less. The NOx results are inconclusive.

-------
9
Table 2. Average Emissions (gms.) and
Fuel Consumed (gals.) During Idle Warm-Up
Sequence of Testing
Total Emissions (gms.)
Test Sequence	HC	NOx	CO	Fuel (gals.)
2 min. idle warm-up + LA-4
45.3
9.4
485.6
.55
5 min. idle warm-up + LA-4
71.1
11.9
567.3
.63
10 min. idle warm-up + LA-4
122.4
6.9
591.3
.74
Average of 6 cold start LA-4's 53.7	9.7 618.2 .54
(3 tests on 2 vehicles)

-------
10
There is another way to interpret the data obtained from this phase of the
test program. Although a CO benefit may accompany an idle warm-up coupled
with an LA-4, how great is the benefit compared to the total FTP CO emissions
(i.e., idle + FTP CO emissions)? Figure 1 provides the answer to this
question. This figure was made by adding an actual Bag 3 CO result at 20°F to
the tests which generated Table 2, and applying the appropriate weighting
factors-^ to each bag result to come up with an FTP CO gm/mile value for
each idle warm-up test and cold start test. These -values are listed in Table
3. The ratio of idle warm-up FTP CO to FTP CO (without idle warm-up) is
plotted in Figure 1.
If The equation used was FTP CO (gm/mi) = [.57 (idle + CT) + CS + .43 HT]/7.5 miles,
(40 CFR 86.144-78, July 1, 1977, page 467)

-------
Figure 1 • Idle Warm-up Factors vs. Idle Warm-up Time
Factor = (Idle + FTP) CO gm/mi divided by-
cold start FTP CO gm/mi.
Idle
Warm
Up.
Factor
1 2
1 0
G 8
0 4
0 2
e 0
4	6
HARM-UP TIME CHIN)

-------
u
Table 3. FTP CO gm/miU
Calculated for Idle Warm-ups
end ColJ Starts.
Idle Warn-up Tine
2 Jain.
5 miiu
10 rain.
Ro idle warm-up
FTP CO gm/aile
33»7 gto/ffiile
38. ? gra/raile
40.8 gm/mile
42.2 gtn/mile
Warm Idle Effects - The average (of both vehicles) emissions produce-d (in
gras,) and the average fuel consumed for the warm idle test as compared to the
eogine-off teat is presented in Table 4.

-------
12
Table 4. Emissions and Fuel Consumption
Comparisons Between Idle and Engine-Off
Periods of the Same Duration
Test Sequences
5 min. Eng-off + LA-4 -
10 min. Eng-off + LA-4
15 min. Eng-off + LA-4
5 min. Idle + LA-4
10 min. Idle + LA-4
15 min. Idle + LA-4
Total Emissions (gms.)
HC	NOx
Fuel (gals.)




6.2
10.6
6Jc4
.43
9.0
11.1
78.4
.43
4.3
10.6
77.8
.43
7.2
11.0
53.0
.47
7.8
12.1
48.0
.52
10.0
11.7
50.1
.58
2/ The CO emissions from the second bag of the LA-4 (Bag 4) were averaged with
those from the other tests to reduce variation for comparison purposes.
Theoretically, Bag 4 results should be very similar because the vehicles are
well warmed-up at the time Bag 4 sampling begins.

-------
13
Although it appears from Table 4 that there may be a CO benefit and a fuel
economy penalty associated with allowing a warm engine to idle in cold weather
for a short period of time instead of turning the engine off, several other
considerations yield this result inconclusive. First, the data on the
individual vehicles did not display the same trend as the average of the two.
Table 5 lists the bag-by-bag CO emission results of each vehicle. One should
notice that there is a CO reduction which accompanies a 10 minute warm idle on
the 1978 Pinto, but there is a CO increase which accompanies a 10 minute warm
idle on the 1978 LTD. Second, the idle percent of total emissions in five out
of six cases is less than 1.7%. Third, the variation in Bag 4 CO emissions
from each vehicle, although the vehicles are fully warmed up, is in most cases
higher than the total CO emissions produced from each idle period. (For this
reason, Bag 4 CO emissions were averaged, and the average was used in the
calculation of Total Emissions, Idle % of Total Emissions, and the Total CO
Emissions presented in Table 4.) In summary, this test procedure cannot be
viewed as very effective in testing a phenomenon whose contribution to total
emissions is so small as to be disguised by normal variations in the test
procedure.

-------
Table 5. CO Emissions and Fuel Consumption Comparisons
between Idle and Engine-Off Periods of the
Same Duration


1978 For LDT



1978 Calif
. Pinto




CO
Emissions (gms.
)
Idle %
Fuel
CO
Emissions (gins.
)
Idle %
Fuel
Test Sequence
Idle
Bag 3
Bag 4
Total
of Total
(gals.)
Idle
Bag 3
Bag 4
Total
of Total
(gals.)
5 min Eng-off + LA-4

33.4
7.3
38.3

.51

53.9
22.4
88.5

.35
10 min Eng-off + LA-4

23.1
4.0
28.0

.52

94.3
32.2
128.9

.34
15 min Eng-off + LA-4

29.4
4.2
34.3

.51

86.8
43.4
121.4

.34
5 min Idle + LA-4
.43
23.1
2.0
28.4
1.5%
.56
1.3
43.0
42.0
77.6
1.7%
.38
10 min Idle + LA-4
.12
27.0
7.1
32.0
.4%
.62
.24
29.5
25.9
64.1
.4%
.42
15 min Idle + LA-4
.53
26.0
5.0
31.4
1.7%
.69
14.8
27.0
41.9
68.9
21.5%
.47


Avg
= 4.9




Avg
= 34.6



Notes:
(1)	Both vehicles have air pumps which route air to the atmosphere after 1-2 minutes of engine operation at idle.
Normally the air is routed to the exhaust manifold.
(2)	Bag 3 is the cold transient driving cycle portion of the 1975 FTP. It covers a distance of 3.59 miles and lasts
505 seconds. Bag 4 is the cold stabilized portion of the 1975 FTP. It covers a distance of 3.91 miles and lasts for
867 seconds.
(3)	Bag 4 results should be relatively insensitive to the preceeding test procedures, as the vehicles were fully
warmed up before the Idle and Bag 3 tests were conducted. Consequently, Bag 4 results were averaged and used in the
calculation of Total and Idle % of Total emissions in an attempt to reduce this source of variation.

-------
15
5. Conclusions
Applicability of Results - Since the data obtained from this test program was
from tuned-up vehicles, it cannot be used to predict the behavior of actual-
in-use vehicles. In-use vehicles, on the average, exhibit significantly
higher idle HC and CO concentrations than tuned-up vehicles, and therefore
might exhibit very different idle warm-up and warm idle effects.
CO Emissions Produced From Idle Periods - It was mentioned that CO emissions
increased as idle warm-up time increased, but that this was not the case for
the. warm idle period. One would expect that as the length of the warm idle
period increases, CO emissions would increase also; all other things being
equal. An explanation of this disparity could be in differing idle speeds of
the vehicles. Although the vehicles were fully warmed up and the carburetors
were thoroughly checked prior to testing for warm idle effects, it is possible
that the fast idle cam did not consistently return to the same position each
time a vehicle returned to idle. A higher idle speed would increase CO
emissions from the warm idle period.
Idle Warm-up Effects - There are two competing factors which probably affect
cold start and total emissions when an idle warm-up preceeds actual driving.
Since the vehicle is warming up at idle, the warm-up time is longer than if
the vehicle were immediately driven after it was started. This would seem to
have the effect of increasing cold start and total CO emissions as compared to
a situation where there was no idle warm-up. However, the volumetric flow
rate of air and fuel through the engine is lower during idle than when the
vehicle is driven. This would seem to have the effect of lowering cold start

-------
16
and total CO emissions, because the vehicle is warming-up in a low CO
production mode (as compared to a vehicle which is warmed-up while it is
driven). The results indicate that the second factor out-weighs the first.
Warm-Idle Effects - There are several factors which would affect the amount of
CO emissions produced during either a warm idle period or a period in which
the engine is turned off and restarted. First, increasing time would increase
the amount of CO produced during a warm idle period. However, increasing idle
speed would also increase CO emissions produced at idle. It is conceivable
therefore that a vehicle which idles for five minutes at a high idle speed
could produce the same amount of CO emissions as the same vehicle idling at a
lower idle speed for ten minutes.
For a period in which an engine is turned off and restarted, there are two
other factors which probably affect the quantity of CO produced. First,
increasing the time which the engine is turned off has the effect of decreas-
ing the temperature of the engine, thereby yielding higher restart emissions.
Second, cranking time also probably has an effect on CO emissions. If a
vehicle takes a long time to start, its emissions will be higher than if the
engine immediately fires upon turning the key.
The variation in Bag 4 CO results, the insignificance of warm idle emissions
when compared to total emissions, and the lack of correlation between warm
idle time and CO emissions produced leads to some doubt on the effectiveness
of the test procedure that was used in studying the warm idle phenomenon. It

-------
17
is possible that a test procedure that compared cranking (starting) emissions
to warm idle emissions might be more effective. However, problems that would
surface in designing and implementing this type of testing would center around
(1) defining the cranking period, and (2) accurately analyzing cranking
emissions.

-------
18
Appendix

-------
ATTAQ1NENT 1
: following questions, among others, were as"ked of 500 participants in
! 1979-SO Anchorage Free Emission Control Test (AFECT), eliciting
se responses:
How long do you let the vehicle warn up in the morning on cold
mornings after starting the vehicle for the -first tini&?
How long do you let the vehicle wann up after v.'ork during cold
days?
When you are shopping and leave the store, how long do you warm up
the vehicle before driving it?

-------
WARM-IP	WARM-UP	WARM-UP
TIME - AM	TIME - PM	TIME - poST
ilNUTES %	CUMULATIVE %	%	CUMULATIVE %	SHOPPINu	CUMULATIVE
0
0.5
1.0
1.5
2.0
2.5
5.0
5.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
10. +
8.9
10.0
8.9
3.9
9.0
3.5
6.1
3.8
3.9
0.2
17.3
0.0
0.2
0.7
0.0
5.1
0.0
«
0.0
0.4
18.0
8.9
41.1
75.5
14.3
11.6
15.8
5.5
10.7
3.0
5.8
2.6
3.2
0.2
15.0
0.0
0.2
0.2
0.2
2.4
0.0
0.0
0.4
8.8
14.3
58.2
88.0
36.0
22.2
17.3
4.5
8.1
0.6
2.9
0.4
0.1
6.2
4.5
36.0
88.1
9S.0
2.0

-------
19
Attachment 2 Vehicle Specifications
1978 Pinto (California)
1978 Ford LTD
I.D.
Mileage
EGR
Air Pump
Catalyst
Engine
Displacement
A.H.P.
I.H.P.
I.W.
8R10Y131366
20,392
Yes
Yes
Yes
In line - 4
2.3L
10.7
9.7
2750
F8863F182034F
35,7 39
Yes
Yes
I
Yes
V-8
302 C.I.D.
10.9
9.9
4500

-------
Attachment 3 Controlled Environment Tp.sfc Cell
—	'Environmental Protection Agency

-------