spreaut$1. oh.
the odie gas (4+ £ We naa #.Ea ; § l ; f 4 £ - . > ka § 3 - l f f & R Caf A R a a » » ~ % s | l : . f u m a i 4 I t ; f l | t - ~ f & s - 3 - d % 2 a " L T a } f f | f f ro ys" R 3 ag f 7 & s . a f 7 e $ i y f > 9 % 5 P s f | i Ch l w j I f Vi f h 5 L f ; | ~ a as | f V | " d A is 2 & a o : 3 f ; . . | f > P A } j C4 bile | . f | h 'o « v a j . f : f c F : C 3 & : ; & | f + i 7 4 + F in § 7 f f f ¥ a § & | i ¥s, f 3 u : 4 + - 8 f | l | S a0% x + R . } . é v , 6 u B a f é | f a a * g x. s f b f ; L 2 « -ol % v 2 a - - ha ie - la 5 . | 8 f & | A 3 k f . | : g F r - -< t : : f f | a a a va U h p, 'F , -a 44 ra f : 7 } | I. s: i F A t% - a# f a i I : 3 - ie £ f > x z x R L z a % § j E s | f 4 u a s | . ; f f | e v 4 A " «ih f § j f ( R a - x £ p v $ } # f s © | * % } f ~ ~ | - F s i @ n o r : J ; l > ald £ f : f . ~ n | : p A é é Pik am - # v f a & -¢ § f ; I ~ ia - + f ¢ ~ f I f f % G * T 5 s j E % | f E f R % j f p A | f y T Re - i . | 3 | = -» a 5< § f r. U ; . L | J : ; f } . Alar f f | ; ; f t f - ' 7 < . hey » s He 2 ' $ . a ; < C P a e: fa # i j 4 3 g . j f * # a k II \ I e miAUK -t #> f 3 s : | ) f 3% f ; . : f & : # f | f f 6 RA a g y R s f : | V . F ' A aC: l ® h p" $ f I é f i s 2 x 2 f 4 f f & « » # s f : | . c | y & * m , & & f \s t* A , z l $ ~PC F ® h a M 0 hess l § f f | .U f | f > a § - & + x « f | E @ f f ; gk < f & 4 f a I ; i ~ s r R ~ ; f '
Statens véig- och trafikinstitut (VTI) - 581 O1 linkoping
National Road & Traffic Research Institute - S-581 01 Linkoping - Sweden\lr 207 A - 1981
SSN 0347-6030
Influence of ambient temperature
and cold start on automobile
fuel consumption
FOREWORD
This study intended for the OECD working group on
"Automobile Fuel Consumption in Realistic Traffic Conditions" has been carried out at the National
Swe-dish Road and Traffic Research Institite (VTI). The
work has been financed by funds of the institute, pro-vided by the government for studies of energy
CONTENTS
ABSTRACT
I INTRODUCTION
2 INFLUENCE OF AMBIENT TEMPERATURE ON
FUEL CONSUMPTION WITH FULLY WARMED-UP CAR
3 INFLUENCE OF AMBIENT TEMPERATURE ON
FUEL CONSUMPTION DURING COLD START
4 CALCULATION OF FUEL CONSUMPTION AS A
FUNCTION OF AMBIENT TEMPERATURE AND TRIP LENGTH AFTER COLD START
5 POSSIBILITIES OF REDUCING COLD START
FUEL CONSUMPTION REFERENCES Page 10 l4 l7
Influence of ambient temperature and cold start on automobile fuel consumption
by Olle Odsell
National Swedish Road and Traffic Research Institute
8-581 01 LINKUPING Sweden
ABSTRACT
Ambient temperature has a significant influence on
automobile fuel consumption, both with warmed-up car and during cold start. As a great percentage of all trips made by car Covers less than 8 10 km, the
cold start fuel consumption is of vital importance for
the car owner. However, car manufacturers have during
recent years concentrated on lowering the fuel con-sumption during standardized driving cycles, like CVS,
HDC and ECE. These cycles do not take into account
cold start or driving at low temperatures, and
there-fore there is still a lot to be done in this field.
The report describes how ambient temperature affects fuel consumption with warmed-up car and during cold
start. A formula is presented, that can be used for
estimates of the fuel consumption as a function of ambient temperature and trip length after cold start.
Finally some possibilities of reducing the cold start
INTRODUCTION
Standardized methods for measuring fuel consumption are today used in many countries. The fuel consumption
figures can be used for regulation purposes or for
information to prospective car buyers about the fuel consumption of different cars. In this respect the
figures are relevant as they make it possible to
com-pare the fuel consumption of different cars when they are driven under exac qrtfmasame driving conditions.
However, the fuel consumption figures from a
standard-ized test cannot tell how much fuel the car will con sume in actual traffic in the hands of a driver. Under
actual driving conditions the fuel consumption is
dependent on many factors, - for instance speed,
driving manners, road-quality, surrounding traffic,
wind and temperature.
One important factor that has influence on the fuel
consumption of a car is ambient temperature. It has the greatest influence during cold start but also when
the engine is fully warmed-up. The standardized fuel consumption tests that are used in Europe (the ECE method and the CVS method) are performed in a nominal temperature of +200C, and only the CVS test takes into
account one cold start from +200C for extended during periods during the year. To be able to estimate the
fuel consumption of cars under realistic driving conditions it is therefore necessary to know the relationship between fuel consumption and ambient temperature, and to know how this relationship is affected by different trip lengths after cold start.
In an American study (l)X it is estimated that about 62% of all trips made by car in USA cover distances
of 5 miles (8 km) or less. These trips account, however,
for only 16% of the total vehicle miles traveled, but require an estimated 31% of the total fuel comsumed,
(see figure 1 and 2). The percentage of short trips
made by car is probably less in EurOpe, because of better public transportation and a greater use of
bicycles. However, a Swedish survey (2) estimates that
trips less than 10 km and made by car account for about 7% of the total amount of kilometers traveled.
In accordance with the American results this indicates that about 14% of the total fuel consumed by cars in
Sweden, would be consumed during trips of 10 km length
or less.
X Numbers in parentheses refer to References at the
um 90-50+- //// 40
30_
/
20 /// 10,PE
RC
EN
UR
MW
OW
10 20 30 40 50 60 70 TRIP LENGTH. MILES
Figure 1 Cumulative vehicle miles traveled vs. trip length, (from SAE paper 750004).
'é
8
\
go
u;
mp
tl
ou
\
l
\
\
M
\
\
PE RC EN T OF 0(Eu
B
6
I 1.. .; O 10 20 30 40 50 60 70 TRIP LENGTH. MILESFigure 2 Cumulative fuel consumption vs. trip
INFLUENCE OF AMBIENT TEMPERATURE ON FUEL CONSUMPTION WITH FULLY WARMED-UP CAR
Even with a fully warmed-up car the fuel consumption increases when the ambient temperature decreases. This
is due to larger heat losses and higher friction in engine and transmission, because of higher viscosity
of the lubricants, The change in fuel consumption with temperature has been studied in several American
investigations (3,4, 5). Based upon these results and upon additional data from automobile manufacturers,
SAE (Society of Automotive Engineers) has established
different correction factors. In SAE standard J 1082, "Fuel Economy Measurement - Road Test Procedure" a
correction factor is used that gives a change in fuel
consumption of 0.25% per OC change in ambient tempera
ture. The correction factor is validated for a tempe-rature range of -l.l to + 32.20C (3O - 9OOF).
The factor 0.25% per 0C is used when the fuel is
measured by weight, which is equivalent to actual
energy consumption. However, it is more common to
measure the fuel by volume, and then it is necessary to take into consideration also the change in specific gravity of the fuel. This correction is 0.11% per 0C, but counteracts the earlier mentioned correction for ambient temperature. If one assumes that the fuel
temperature follows the air temperature, the correction factor thereforevHJJ_be 0.14% per OC, when the fuel
consumption is measured by volume.
The SAE correction factor is a constant, and it shows
how an "average" car changes its i m . consumption with ambient temperature. Tests (6) performed at the National Swedish Road and Traffic Research Institute,
VTI, have shown, however, that the change in fuel con-sumption is progressive, especially at temperatures
below i 00C. In figure 3 is shown how the fuel
con-sumption for a European medium size car (engine dis-placement 2 litres) changes with ambient temperature. The figure shows the average specific fuel consumption
measured over a 34 km long driving cycle on the road
with fully warmed-up car (lower curve) and with cold start (upper curve). The change in fuel consumption with warmed-up car corresponds rather well to the SAE correction factor 0.14% per 0C in the temperature
range of +50C to +250C. However, below +50C the fuel
consumption rises more rapidly and the change is about 0.5% per 0C in the temperature range below -lOOC.
specific fuel consumption (l/lOO kmh 11.0 I 1 \\\\ old start 10.5 \
10.0
\
\
warmed:up\\\\\
9.0
\\
\.
\ 8 . 5 engine \ \\\'\ \ 8.0 7.5 -20 -15 -1O -5 0 +5 +10 +15 +20 +25 ambient temperature (0C)Figure 3 Average specific fuel consumption at
diffe-rent ambient temperatures, measured with a
EurOpean medium size car over a 34 km long driving cycle with mixed city/highway driving
According to these results one can expect about 3% higher fuel consumption at iOOC compared with +200C, with a fully warmed-up car. These changes in fuel
consumption are measured at dry road surfaces and with
radial summer tyres. In cold climate, where there are roads with snow and ice, and with the use of studded
winter tyres, it is quite possible that the fuel con sumption will be about 10% higher in winter than in
summer, even with a fully warmed-up car.
INFLUENCE OF AMBIENT TEMPERATURE ON FUEL CONSUMPTION DURING COLD START
Even though the fuel consumption changes with ambient temperature with a fully warmed-up car, the change is
much greater when also cold start is included. Tests
(6, 7) have shown that in low ambient temperatures it
takes about 0.5 litres of gasoline to warm up the
en-gine of a EurOpean medium size car to the normal ope rating temperature around +800C. However, also in
summer temperatures the engine warm-up takes O.l - 0.2 litres extra fuel after a cold start.
The two curves in figure 3 show the difference in
ave-rage specific fuel consumption after driving a dis-'tance of 34 km with warmed-up car and with cold start. The main part of the extra amount of fuel that is
consumed for the warm-up is, however, consumed during the first 5 - 10 kilometres, and therefore the
influ-ence on the average specific fuel consumption will be greater at short trip lengths after cold start.
As an example figure 4 shows how the average specific
fuel consumption changes with trip length, after cold
start in ambient temperatures -l8OC and ibOOC. The two lower curves show the fuel consumption when driving exactly the same route with fully warmed up car in the
same temperatures. In this case the driving cycle
consisted of mixed city/highway driving on the road, and
started with idling during one minute, which explains the higher specific fuel consumption directly after start, even with warmed-up car. The fuel consumption during the first minute of idling after cold start in temperatures below i_OOC was 100 - 120 cm3, while the
fuel consumption during idling with warmed-up engine was 20 - 25 cm3 per minute.
specific fuel
consumption (l/lOO km)
1
A30 _
\
\
\ 25 qu-180c
\f i 0°C
//____ 20 1 15 \\\\ \x\\
\ Na
old start * \*____*§ \ \&- x _. _H 10 A xV N \/ \ X nqhq- m ~ *- * - w - + »:£:=$==$=a-kr " A _;,_ c + +_n o o o-warmed-up en-5 gine 0 e:o
5
10
15
20
trip length (km)25
30
35
Average specific fuel consumption asfunction of trip length, for a EurOpean
medium size car (engine displacement 2 litres)
Figure 5 shows another example from an investigation
(8) in USA, in which the average specific fuel
con-sumption for four cars with engine displacements be-tween 2.3 - 6.5 l was measured in four different am-bient temperatures. In this case the tests were made
on a chassis dynamometer, using the EPA 1975 Federal Test Procedure. In general the curves in figure 5 show the same tendencies as in figure 4, although the more severe driving cycle and the larger cars give higher
absolute fuel consumption values. Attention should be paid to the effects of air conditioner use on fuel consumption at high temperatures.
average specific fuel consumption (1/100 km) 38* 4?
361
34 32 3O 2 8T 26 24, 22 204 7 C +380C *)K
\a
a 4} U.\' +7Oc
+21OC
*) with air-conditioning Figure 5 2 4 6 8 10 12 14 16 18 20 trip length (km)Average Specific fuel consumption as function
of trip length. Average values for four Ame-rican cars with engine displacements between
lO
CALCULATION OF FUEL CONSUMPTION AS A FUNCTION OF AMBIENT TEMPERATURE AND TRIP LENGTH AFTER COLD START
In a Swedish investigation concerning fuel consumption under varying driving conditions (9) an empirical for-mula has been presented, describing the relation
between specific fuel consumption, ambient temperature and trip length after cold start. The formula is as
follows:
_ . , B
B
BO [1 + a
At (1+ 5)]
where B = actual specific fuel consumption (l/lOO km) B = specific fuel consumption with fully
warmed-up car (l/lOO km)
a = a constant that describes the relation be-tween fuel consumption and ambient tempera-ture with fully warmed up car '
At: 30 t, where t = ambient temperature (QC),
(t s 300C)
8 = a constant that describes the relation be-tween fuel consumption and trip length after cold start
s = trip length after cold start (km)
The purpose of the formula is to provide rough estimates as to how the average fuel consumption for different
classes of cars, (for instance different weight
classes), changes when they are driven under varying
conditions. Therefore the value BO can be calculated from for instance composite values of the ECE urban cycle and the constant speed tests. In Sweden the mandatory fuel consumption declaration, which is a composite value of 55% CVS (1972) cycle and 45% HDC cycle, forms a good basis for the calculation of B0
11
The formula can of course also give an indication as
to how the specific fuel consumption of individual
cars changes with ambient temperature and trip length after cold start. In this case it must be remembered,
however, that different cars react differently to changing ambient temperature, depending on type of choke arrangement etc. Therefore the formula can only give a rough estimate of the fuel consumption for
individual cars.
The constant a represents the correction factor for
the larger friction in the transmission etc, at lower
temperatures. According to earlier discussions in this report the constant a should actually be substituted
with a function of the ambient temperature, but since
the formula should only be used for estimates, a
constant value gives sufficient accuracy and a simple
formula. The constant could, however, be given different
values depending on the temperature range and the weather conditions in which the studied cars are nor mally driven. If cars are used in snow and with studded
tires for a considerable period of the year, this can
be accounted for when choosing the value of the constant.
This is the case in Sweden and a value of d==0.004 has been shown to give relatively good correspondance
with practical measurements. In warmer climates, where
the temperature seldom falls below iOOC, a value of d==0.002 might be more appropriate.
The value At is chosen as the temperature difference from +3OOC. This means that the fuel consumption,
according to the formula, is the same with a cold
start from +3OOC as with fully warmed-up cars. This is not absolutely true, but anyway the formula gives realistic results in the normal temperature range in
12
The constant B is chosen according to type of engine,
as a diesel engine does not consume so much extra fuel as a gasoline engine during cold start. For gasoline engines a value of B = 20 has been shown to give good
correspondance with empirical data. For diesel
engi-nes it is more difficult to find test results, but a
plausible value could be around 8 = 5.
With values of the constants, suitable for the climate in northern EurOpe, the formula takes the form:
B = BO [l + 0.004 (30 - t)(l + gglj(gasoline
en-gines)
B = B [l + 0.004 (30 - t)(l + El] (diesel
engi-O S nes)
In figure 6 is exemplified some results that the use
of the formula could give. The curves show how much the fuel consumption rises at cold start in three different temperatures, compared with the fuel consumption with
fully warmed-up car. If one assumes that a car has an average specific fuel consumption of B = 10 1/100 km in fully warmed-up condition, it can be seen in the
figure that after cold start and driving 5 km it has
an average specific fuel consumption of 12 1/10 km at
+200c, l6 1/10 km at lLooc and 20 1/10 km at ~200c.
These calculated values correspond rather well with
experimental results. 'Bhis indicates that the formula gives realistic values of the fuel consumption under varying driving conditions.
l3 .
-200c
-:0°c
+20°C
> I 2 4 6 8 1O 12 14 16 18 20 s(km)Ratio between cold start fuel consumption and warmed-up fuel consumption as function of trip length, at different ambient
l4
POSSIBILITIES OF REDUCING COLD START FUEL CONSUMPTION
The establishment of standardized fuel consumption test procedures has been of vital importance for car
manufacturers and governmental agencies in the strive
towards reducing car fuel consumption. However, it has
resulted in car manufacturers concentrating their
efforts on obtaining low fuel consumption figures under
the standardized test procedures and this may result in solutions not necessarily Optimal for the actual
driving conditions.
The standardized test procedures CVS, HDC and ECE do not include cold start in low ambient temperatures,
and therefore little work has been done by can
manu-facturers in order to reduce the cold start fuel
con-sumption. As this, however, is an important factor for many car owners, for instance when the car is used daily between work and home, it is worth looking at
how the cold start fuel consumption can be reduced.
The use of the choke at cold start is of great import-ance for the fuel consumption. Many cars are equipped
with automatic choke control, but the function of these systems has often been unreliable, with raised fuel consumption as a result. During recent years there
has therefore been a tendency towards wider use of
manual choke control, but new and more s0phisticated
systems for automatic control are being developed. The use of fuel injection systems offers good possi bilities of optimizing the cold start fuel consump
tion, - especially the recently develOped systems with integrated electronic control of ignition and fuel
mixture.
Most of the excess fuel that is consumed during cold start is due to warming up of engine and transmission.
15
A way to reduce this fuel consumption is to warm up the coolant and lubricants by other means, for instance with electricity. The capacity of the car battery is not sufficient to supply a heater, but during parking
it is possible to use external electricity for heating
purposes. In northern EurOpean countries an electric
engine heater for the coolant is a common accessory, mainly because it makes cold start at low temperatures
easier and it gives better comfort with quicker heating
of the passenger compartment. However, investigations (6) have shown that electric heating of the coolant also gives a reduction of the fuel consumed during
warm up with 0.1-0.3 litres, for a 2 litre engine. If the heater is connected to a timer that restricts the
using time to a maximum of 2 3 hours it is cost effective
as the saved fuel costs more than the supplied
elec-trical energy of 1.0-1.5 kWh. An extension of the heating also to the engine oil and transmission oil
is possible, but it is doubtful whether this can be
cost-effective.
The use of a thermostat controlled cooling fan reduces
the time for engine warm-up and hence the cold start
fuel consumption. Special arrangements regarding the
flow of the coolant around the engine may also have a fuel saving potential. Some attempts have been made by various car manufacturers to modify the engine inlet manifold in order to reduce cold start fuel
consumption. Plastic coating of the inner surfaces of the manifold can reduce condensation of the fuel,
and electric heating devices in the throttle Opening are used for the same purpose. Further work in these areas can be awaited in order to reduce the cold start
fuel consumption.
Finally it should be mentioned that synthetic engine oils and transmission oils have considerably lower
l6
viscosity at low temperatures than conventional oils. This can result in fuel savings of up to 5% at cold
start, and an overall fuel consumption reduction of
2 3% under varying driving conditions. The use of syn-thetic or part-synsyn-thetic lubricants is expected to
17
REFERENCES
1. Austin, T.C., Hellman, K.H.,
economy as influenced by trip length.
paper 750004.
Passenger car fuel
SAE
Swedish travel patterns 1978. The national central
bureau of statistics, Sweden, 1980.
Customer fuel
eco-SAE
Scheffer, C.E., Niepoth, G.W.,
nomy estimated from engineering tests.
paper 650861.
The develOpment of the new SAE motor vehicle fuel
economy measurement procedures. SAE paper 750006.
Claffey, P., Running costs of motor vehicles as affected by road design and traffic. Highway Research Board NCHRB report 111, 1977.
Odsell, 0., The use of electric engine heaters
- influence on cold start fuel consumption.
VTI report No. 190, 1980, (in Swedish).
Everall, P.F., Northrop, J., The excess fuel con-sumed by cars when starting from cold. TRRL
report LR 315, 1970.
Eccleston, B.H., Hurn, R.W., Ambient temperature
and trip length - influence on automotive fuel economy and emissions. SAE paper 780613. Odsell, 0., Estimates of specific fuel consumption
for different vehicle categories under varying driving conditions. VTI bulletin No. 214, 1980,
