This report has been produced to highlight the dangers of car fires. It discusses various aspects of such fires, in an attempt to provide information which will enable the public to exercise choice in selecting products which affect their personal safety.


This report uses the best available data to define the frequency of car fires. It examines causes, and considers what measures can be taken to reduce the frequency and severity of such fires. It concludes that speed of exit from a burning car is vital, and that seatbelts slow down speed of exit.

Prepared by Mr. G. Read, August 1990.


The number of car fires reported by fire brigades rose by 6% from 1988 to 1989, to 39,800. This 6% rise is slightly less than the increase in vehicle-kilometres travelled from 1988 to 1989.

A study by Lincolnshire Fire Brigade concluded that some 70% of car fires are not reported to the fire services, and that therefore the total number annually could be as many as 150 thousand.


Most car fires are not the results of crashes or collisions. Only some 3% are from this cause. Many occur while the car is in normal use.


Currently, car fires are the cause of 10 to 12% of all fire deaths in U.K. Deaths and serious injuries from car fires are commonly caused by inability to leave the vehicle instantly. Once a car has caught fire and is well ablaze, the emergency services have little chance of rescuing the occupants. Deaths in road accident fires have risen from 20 (in 1983) to 81 (in 1989). Most car fire deaths are in accidents.


4.1. General.

The latest Home Office Statistical Bulletin on vehicle fires shows causes as follows:-

Deliberate, or possibly deliberate 34%
Wiring on vehicle 19%
Batteries 4%
Ignition-related 5%
Exhaust 13%
Sparks 6%
Overheating under 1%
Crash/collision 3%
Non-vehicle defects 2%
Not vehicle-related 4%
Unspecified 4%
Other 4%

This tabulation shows that at least 28% of fires are electrical and that only a small minority are caused by road traffic accidents. The 34% "deliberate or possibly deliberate" can be ignored in the context of this report, as fire prevention measures will not reduce numbers.

4.2. Electrical Causes.

The type of circuits used in cars can be classified as "high current". In such circuits, especially (as is usual) where the vehicle shell is used as the return path, a wire shorting to the shell can produce a spark or, where the short is not momentary, it can produce a heat source equal to or greater than that of one bar of a standard electric fire (1kw) Such heat will cause the plastic insulation of the wire to burst into flame.

Even in fused circuits, a significant amount of heat can be generated without blowing the fuse. Often, several circuits share one fuse; a typical value might be 35 amps. Then, a heat of 400 watts can be generated without blowing the fuse.

4.3. Fuel System

Although few statistics are available, a significant number of car fires are probably related to fuel leaks. At atmospheric pressure, and given the right mixture of air and fuel vapour, unleaded petrol will self-ignite at a temperature of approximately 400 deg. C.; for 4-Star petrol, the temperature is approximately 490 deg. C. As exhaust manifold temperatures can be above 400 deg. C., there is therefore a possibility that fuel leaking from a carburettor above an exhaust manifold could cause a fire.

These figures can only be approximate, as many variable factors are involved in the ignition of hydrocarbon vapour. It should be noted that liquid petrol is not what ignites; the petrol must vaporise before ignition can occur. Petrol vaporises below 0 deg. C.

In crash situations, fuel tanks or fuel pipes may rupture. Sparks may occur, as may electrical wiring shorts, giving a source of ignition to any petrol vapour present.


5.1. Plastic Materials.

Much plastic material is used in modern cars. Almost all, especially in upholstery and interior trim, is inflammable to a degree which would make its use illegal in domestic furniture. Given a source of fire, this material rapidly burns and extends the size and the ferocity of the fire. This situation exists in most countries: Finland and U.S.A. are notable exceptions.

5.2. Electrical Components.

Unlike components used in domestic and industrial service, the car components do not have to conform to any mandatory standards of safety. This does not mean that all are dangerous, but it does mean that components from different sources can vary greatly in their standards of safety.


There are three basic methods of improving the fire safety of cars. They fall into groups as follows:-
(a) The use of non-inflammable plastic materials.
(b) Re-design of fuel systems, coupled with the fitting of exhaust manifold heat-shields.
(c) Re-design of electrical systems on a ring-main basis and not using the vehicle shell as a return, employing low-current switching and control circuits.

While these improvements could be made, they would inevitably increase car prices. What is more significant is the fact that such changes would not affect existing cars, of which there are some 20 million in U.K. If an average car life of 10 years is taken, it is apparent that such changes in new production would take at least 5 years to have significant effect. It is therefore logical to examine possibilities for improvement which could have immediate effect.


7.1. Fire Extinguishers

German law requires all cars to carry fire extinguishers. On the face of things, this is an obvious step to take. However, such a requirement is likely to have much less effect than its proposers might hope. It is common knowledge amongst the fire services that a large number of members of the public are at best ineffective when attempting to fight a fire, even when suitable equipment is available. Therefore, common advice from Fire Brigades to the public is "get out and leave the fire to us.

A point relevant to fire extinguishers is, surprisingly, their purpose in car fires. The obvious answer is "to put out the fire". This in turn raises the further question of why this is required.

If a car catches fire it will be a "write-off". This being so, the only purpose served by attempting to put out the fire is to gain time for the occupants to leave the vehicle. As most car fires require the fire-fighter to be outside the vehicle in any case, it is apparent that the best course from the personal safety point of view is for the occupants to leave the vehicle as quickly as possible and call the fire brigade (or, in an isolated situation, let the car burn out).

The average fire extinguisher has an effective operation period of about 30 seconds before becoming exhausted. This means that the fire must be extinguished within this time period. An untrained person Is unlikely to succeed, especially as most car fires occur inside compartments which are limited in access (e.g. under dashboards, inside engine compartments, etc.).

7.2. Speed of Exit.

The aviation world has long been aware that speed of exit from the burning vehicle is vital in fire situations. All aviation crash procedures are based upon this philosophy, but it is notable that this philosophy has never been applied to road vehicles. Most car fire deaths are in crashes, 81 out of 94 in 1989 and 20 out of 27 in 1983.

The prime example of disregard of the importance of immediate exit is the four-seated two-door car. The rear-seat occupants cannot leave the vehicle until the front-seat occupants are outside the vehicle. The minimum time in which the rear-seat occupants can leave the vehicle is therefore largely determined by the time taken by the front-seat occupants to leave the vehicle. If for any reason the front-seat occupants cannot leave the vehicle (e.g. they could be unconscious after a crash) the rear-seat occupants are completely trapped.

Professional fire service opinion is that the occupants of a car which has caught fire have at best "a few seconds" to get out. This "few seconds" will be taken as 4 seconds for the purposes of this report.

Following text shows the exit times required by a number of test subjects. In each test the car was stationary: the onset of fire was simulated by a pre-arranged signal from the timing observer to the occupant of the driver's seat and to no other occupant. This simulated the situation in which the driver spotted a fire and the other occupants (if any) of the car were unaware of the fire until alerted by the driver.

No facilities were available for carrying out a large series of tests of exit times. It was therefore decided to carry out a few "best-case" and a few "worst-case" tests. The "best-case" tests involved young and agile males who were familiar with the car used (in most cases, their own car). The "worst-case" tests involved older people who were not familiar with the car used. It is reasonable to conclude that the results of these tests define the lower and upper bounds of probable exit times.

Two aspects of the matter, both unexpected, emerged. These were:-

(a) Modern 2-door cars have latch handles on their front seats. These must be released before the seat will tip forward to permit exit by rear-seat occupants. In normal use, they are usually released by the front-seat occupant as he gets out, so as to permit exit of the rear-seat occupant. No front-seat occupant in the tests thought to release these catches. The occupants of rear seats usually did not know where these catches were.

(b) Older women, especially in the lower socio-economic groups, will rarely obey a sudden order without question. Their normal response is to become agitated and to question the order and the reasons for its being given. This, apparently well-known to psychologists, means that such persons will rarely if ever succeed in rapidly quitting a burning vehicle.

For an unbelted person in a front seat we found that a "best case" time of exit was 3 to 4 seconds. A "worst case" time was at least 8 seconds, and could be as high as 11 seconds. The times of exit for passengers in the rear seats of a 4-door car were similar.

We found that these exit times were increased when a seatbelt was used. The "best case" times were increased by about 1.5 seconds. The "worst case" times were increased by up to 5 or more seconds depending on the subject's familiarity with the car and the seatbelt fastening.

For the occupants of the rear seats of 4-seat 2-door cars, the exit times were at least 1.5 times that of the front seat occupants because the rear seat occupants cannot quit the vehicle until the front seat occupants have left. The exit time from such a rear seat for a stout and elderly person can be very great.

NOTE. These results should not be taken as definitive. They are the results of a small number of pilot tests only. It is expected, however, that a larger series of tests will produce results which fall within the bounds set by these pilot tests.


This initial report leads to the inescapable conclusion that a number of long term measures related to car design would be advisable, but that such measures would have little or no immediate effect on casualties. The only visible way, in our opinion, to have immediate effect on car fire casualties would be to abolish seatbelt compulsion because it is apparent that it is a significant factor in increasing time of exit from a burning car. Another page at this site gives a listing of cases known to CIPS where persons have been harmed by belts. In this listing it is notable that a significant proportion involve fire.

Return to C.I.P.S. home page