Engineering Information > Detail

Category: Engine
Posted: 12-02-07

< Return to previous page


Carburettor icing is a known killer. It can cause engine stoppage or loss of power, leaving the pilot of a single engine aircraft facing a forced landing on whatever terrain happens to be beneath at the time and the eventual outcome will then depend upon a combination of the pilot’s skill and luck. In many cases skill and luck will be sufficient for the event to conclude in the survival of the aircraft’s occupants and no damage to the aircraft. In other cases the conclusion will be occupants with only light or no injuries and a damaged aircraft. In some, however, skill and luck will be insufficient to prevent serious injuries or fatalities.

The Air Accident Investigation Branch (AAIB) reports a continuing toll of accidents where carburettor ice is suspected of being a factor. During the period 1998 to 2002 the average of such accidents was between six and ten per year and no doubt there are many more carburettor icing events that, thankfully, do not fall to be investigated.

The likelihood of formation of carburettor ice depends principally upon the ambient humidity. A carburettor has the ability to reduce the temperature of the air induced through it by as much as 35 degrees C. If there is dampness in the air (i.e. it is humid) then in the temperatures found in the UK there is a strong possibility of the moisture in the air condensing on the interior of the carburettor and forming ice there, particularly on the butterfly valve. High humidity is a usual feature of the UK’s moist maritime climate and consequently a high incidence of carburettor icing is encountered in this country. Most General Aviation (GA) engines are of American design and manufacture and experience of flying in California and Florida – two states where a great deal of American light aircraft flying takes place – suggests that the usual rule over there is not to use carb heat at all unless flying in cloud. No doubt there are parts of North America, particularly Canada, where high humidity similar to that experienced in the UK can be found. Nonetheless, most light aircraft operation in the USA seems to take place in an environment of lower humidity than is to be found in the UK. Thus American engines are generally more likely to encounter carburettor icing when used in the UK than they will in the USA. The same is likely to be true of engines coming from other sources where humidity is generally less than it is in the UK. Carburettor icing is a particular problem in regions of high humidity and the UK is such a region.

The General Aviation Safety Council (GASCo) believes that the time has arrived for a major initiative throughout GA in Britain to address the problem of carburettor icing and for the general introduction of permanent defences against it. There are twenty eight members of the GASCo Council. The members include the representative bodies of all the significant General Aviation activities in the United Kingdom, relevant learned societies and Government Departments concerned with General Aviation. A list of the member organisations of GASCo is included in the Appendix.

Neither the CAA nor the Air Accident Investigation Branch who are members of the GASCo Council have considered it appropriate for them to be involved in the preparation of this paper.

This paper does not address the relatively rarer induction icing problems associated with impact icing at the induction inlet, icing of ram air devices on some piston engines nor ice formation at very lower temperatures of water contained within fuel.

Aircraft engines can offer a variety of defences against carburettor icing, as follows:

Manually applied carburettor heat
Normally aspirated piston engines are often fitted with a carburettor heat device that warms the air before it reaches the carburettor. With the air heated to a sufficiently high temperature it will still remain above freezing level after subsequent cooling in the carburettor and ice will therefore not form. On Continental, Lycoming and many other aircraft engines the heat is derived from an airbox surrounding part of the exhaust system. Two stroke Rotax engines using the Bing 45 carburettor can be fitted with an electric carburettor heater powered directly from the engine generator. Heating the air reduces the volumetric efficiency of the engine, causing a power loss of anything up to 15%. To get around this loss of efficiency the engines are normally run without carburettor heat being applied. Application of carburettor heat is made by manual operation of a knob or switch and typically pilots of aeroplanes are trained to apply carburettor heat for at least 30 seconds at about five minute intervals in the cruise. Applications must also be made before descending, before final approach and before take off. These devices theoretically provide a more or less complete defence against carburettor icing. In practice, however, pilots forget to apply heat on all occasions that their training has taught them and sometimes they pay a heavy price for their forgetfulness. Pilots used to flying aircraft not fitted with manual carburettor heat sometimes forget altogether to apply heat. Every issue of Flight Safety, GASCo’s quarterly magazine includes a competition and one that was run in 2002 was devoted to carburettor ice. The degree of misinformation that was evidenced in the replies was quite disturbing: the majority of respondents – and they presumably would not have entered the competition if they had entertained many doubts about their knowledge of the subject – believed that air temperature rather than humidity was the major determinant of carb ice vulnerability and that if an engine faltered when carb heat was selected then it should be deselected. It is clear that a great deal could usefully be done to improve GA pilots’ knowledge about carburettor ice and how to deal with it.

Even when pilots comply absolutely with their training, however, they can still come to grief. AAIB Bulletin 4/2003, Ref. EW/G2002/09/12 tells of a Jabiru pilot who flew into clear air downwind of the Drax power station at 1,500 ft. Power was lost and in the subsequent forced landing on a recently seeded flat field the nose wheel dug in at slow speed, the new microlight overturned and was badly damaged. That pilot and the world of GA generally have now learned that even when in clear air form, the exhaust from cooling towers will have an extremely high humidity and this may persist for some miles downwind. The AAIB concludes that it is highly likely that carburettor icing caused the loss of power in this case. Whatever the theoretical reliability of manually applied carburettor heat, the accident records show that reliance upon the pilot applying heat in a sufficiently timely manner is sometimes ineffective.

Carburettor temperature instruments
Some aircraft are fitted with a gauge that reports the temperature within the carburettor. If the carburettor temperature is below freezing point then there is a risk of carburettor icing although the degree of risk will depend upon the humidity of the air. A pilot can adjust the setting of the carburettor heat control so as to maintain a carburettor temperature above freezing and thus eliminate the risk. However, a change of throttle setting, mixture setting or ambient temperature will require re examination of the carburettor heat setting. There will be an inclination on the part of the pilot to aim for a temperature just above freezing so as to reduce the loss of volumetric efficiency to a minimum. However, just a small drop in carburettor temperature subsequently may produce conditions highly conducive to carburettor icing. The device is passive and requires monitoring but it is likely to increase the likelihood of carburettor heat being applied conscientiously. One such device, however, did not prevent a Robinson R22 pilot’s death by way of carburettor icing in December 2002 (AAIB Bulletin 10/2001. EW.C2000/12/3)

Carburettor ice warning devices
A 100% reliable ice formation warning device coupled to a wholly effective warning system (pilots sometimes postpone attention to warning horns until it is too late) could be the near complete answer to the problem. On receiving the warning the pilot applies carburettor heat and continues to do so until the warning ceases. The ‘Iceman’ is probably the best known carburettor ice warning device available at present. A probe within the carburettor reports the actual formation of ice and activates warning devices – a flashing light and/or warning horn which can be piped into the headset if required. A degree of endorsement of the Iceman is implied by the CAA’s willingness to lessen modification approval charges for those fitting the device. We have encountered anecdotal objections that the Iceman carburettor probe is prone to becoming fouled with oil or other matter, causing false alarms although the device includes a zeroing facility that is intended to cope with partial fouling of the probe. Firm evidence from existing users needs to be assembled about the reliability in use of this device and if the Iceman is in fact reliable and trouble free then its use should be actively encouraged. If not, its improvement or other availability of a 100% reliable ice detector would be very welcome.

Permanently applied carburettor heat
The example is often quoted of the DHC Chipmunk which, when in military use, had its carburettor heat wired permanently on. This inevitably caused a permanent loss in engine efficiency but it is instructive to reflect that the military decided that the loss in engine efficiency was a worthwhile trade off against the potential loss of aircraft resulting from carburettor icing. Considering that pilots with much greater experience and skills than the average GA pilot were operating these aircraft it is arguable that the military solution might have considerable relevance to GA operations and aircraft generally. There are engine aspiration issues such as air filtration to be considered when this solution is applied to an installation originally intended for manual heat application. There is also the question of whether the loss of maximum obtainable power might leave some aircraft dangerously under powered. An alternative example of permanently applied carburettor heat is to be found in the 912 versions of the Rotax engine, which can be fitted with a carburettor heater that derives heat constantly from the engine coolant. In the British climate this heat is applied permanently and the suppliers claim that as the amount of heat applied to the carburettor (not directly to the induction air in this case) is small, the loss of volumetric efficiency is negligible.

Normally aspirated installations that are carburettor icing free
There are normally aspirated engine installations not fitted with a dedicated air or carburettor heater that do not suffer from carburettor icing. These usually route the induction air past warm parts of the engine causing sufficient rise in the induced air temperature to avoid carburettor ice. This therefore is a very similar solution to wiring manually applied carburettor heat permanently on. However, air filtration and other issues can be considered ab initio. The Limbach engine (a derivative of which became the engine of the Volkswagen Beetle) makes an interesting case in point. In some installations the engine is permanently proof against carburettor icing, with no actions being required of the pilot, and in others it is fitted with carburettor heat. The latter arrangement is found in many Druine Turbulents where the engine ( an ‘Ardem’ version) in this installation is notorious for suffering carburettor icing at the least provocation whenever heat is not being applied.

Fuel injection engines
Not having a carburettor these engines do not suffer from carburettor icing. They are often fitted with an ‘alternate air’ control, which seems similar to a carburettor heat knob, but this is in fact intended to cope with the case of snow impacting or ice forming against the induction air filter and blocking it. This is not likely to be a situation encountered by most GA aircraft in the UK. Fuel injection seems to be a complete answer to the carburettor icing problem and engines of 150 HP and upwards are frequently fitted with fuel injection. GASCo is encouraged to see the spread of the popularity of fuel injection. The new Cessna 172s and the new Robinson Raven R44s are both fitted as standard with fuel injection engines, as are many other new aircraft. This is by no means universal, however, and we are disappointed to report that the new Piper Warrior III still has a normally aspirated engine, even when offered on the British market. A carburettor ice detector is offered as an option for an extra $1,300. The latest Robinson R22 unfortunately still has normal aspiration and manually applied carburettor heat with a temperature gauge fitted as standard.

Diesel or compression engines seem likely to become a significant part of the European piston engine market for aeroplanes before long. One of their attractions is proof against carburettor icing as they have no carburettor.

Fuel additives and Teflon coating of carburettor interiors
The National Research Council of Canada produced a paper in 1970, Aircraft Carburettor Icing Studies, which dealt with research into the prevention of carburettor icing either by the introduction of fuel additives or by coating the interior of a carburettor with Teflon. An engine was run at what seemed to be the throttle setting that produced the greatest vulnerability to carburettor icing and water was sprayed into the induction system. The conclusion was that the use of ethylene glycol monomethyl ether (EGME) at 0.10 to 0.15% by volume proved to be the most effective additive and that Teflon coating was also relatively effective. A combination of both solutions seemed to be absolutely proof against icing. In spite of the apparent promise of these findings we have not discovered any subsequent commercial application of either of these solutions nor are we aware of any subsequent research along the same lines. We have discussed the fuel additive issue with a technical expert at Shell but that company does not market EGME either as a constituent of avgas or as an additive. It does market isopropyl alcohol (IPA) as an additive for both jet fuel and avgas , but this is intended to deal with the potential for water contained within fuel (there is usually some) from forming ice crystals when aircraft are operated at temperatures a good deal below freezing. IPA is not offered as a solution for carburettor icing.

Further research
We understand that the CAA intends to carry out some research into possible measures against carburettor icing and we welcome this.


1. GASCo believes that the present toll of deaths, injuries and aircraft write offs connected with carburettor icing could be reduced significantly, and this is an eminently desirable aim.

2. Experience shows that reliance on the intermittent application by the pilot of carburettor heat is an inherently unsatisfactory solution as GA pilots cannot all be relied upon to apply heat conscientiously on all occasions. Much needs to be done to improve the level of knowledge about this issue amongst GA pilots and instructors, both when training and subsequently. GASCo would be happy to become involved in such an initiative.

3. So far as new aircraft are concerned a number of proven ways exist to remove the likelihood of carburettor icing occurring. Fuel injection, induction systems that heat the air continually, and heated carburettors are all in regular and satisfactory use. Diesel engines may well join them. For replacement and/or overhauled engines fitted to existing airframes, in many cases carburettor ice proof versions of the existing engines might be offered as substitutes for the existing carburettor ice prone engines. For new aircraft and new or overhauled engines for the British register GASCo believes that insistence on reasonable carburettor ice proofing without the need for manually applied carburettor heat is a reasonable standard to look for. There will be various ways of achieving this standard and the manufacturer might be best able to choose which way is most suitable in the case of each particular engine.

4. For the existing fleet and engines which have not yet reached the need for replacement or overhaul GASCo believes that, so far as those engines that rely on manually applied carburettor heat are concerned, more research is required into ways making these engines safer in the hands of inherently forgetful pilots. Permanently applied carburettor heat may be the best solution for some aircraft and a proven 100% reliable carburettor ice detector may answer best for others. No final conclusion can be made without further authoritative research. GASCo would be willing to become involved in any such research, subject to the availability of resources.


AAIB Bulletins. AAIB website:

Piston Engine Icing. CAA Safety Sense Leaflet No 14A. or LASORS.

Aircraft Carburettor Icing Studies by L Gardner and G Moon, Division of Mechanical Engineering, National Research Council of Canada. Information on obtaining this
document from CISTI's Help Desk/Judy Parisien. Tel 001 613/993-9206, Fax 001 613/998-2399

Induction Icing by Miles McCallum. Flyer magazine, January 2003.

Rotax engine carburettor heaters. Skydrive website:


Contact: General Aviation Safety Council