Burning Power

When the fire in an engine escapes the cylinders, the problem isn't staying aloft, but staying alive.

July 1, 1983

The Beechcraft Bonanza was being flown from San Francisco to Santa Monica, California. While the Model A35 was at cruise altitude, an engine fire developed as the result of a fuel leak. Recognizing the need to land, the pilot headed for nearby Paso Robles Airport, overflying a number of suitable emergency-landing sites. By the time he was on final approach, the fire had developed into an inferno that was clearly visible to horrified observers on the ground. Panic apparently dictated the tragic events that followed.

As the aircraft shot across the runway threshold with an extraordinary abundance of airspeed, the right-seat passenger opened the door, jumped out in an attempt to escape the flames and was killed in the process. The Bonanza then landed heavily on its nosewheel and seconds later was seen cart-wheeling into a smoldering aluminum ball, claiming the lives of its pilot and the two rear-seat passengers.

Accident investigators concluded later that everyone aboard could have survived the consequences of the engine fire if the pilot had abided the recommended procedures for such an emergency. They concluded also, however, that most pilots have only a vague idea of how to react to an in-flight engine fire (as determined by an informal survey conducted several months after the accident).

During the 31 months from April 1979 to October 1981, 47 general aviation accidents were attributed to engine fires. That is 18 per year. Although the chances of experiencing an engine fire are relatively poor, the nature of the emergency — demands immediate action and abiding by a series of guidelines. There is precious little time available to consult an aircraft operating handbook or check list, both of which usually offer inadequate guidance anyway.

The seriousness of an engine fire must not be underestimated. Consider, for example, this advice offered by Britain's Royal Air Force to its pilots: "Landing with an aircraft on fire is seldom justified and very dangerous; there is little chance of saving the aircraft after landing, and there is great risk that the fire may increase or cause structural failure at an altitude where it is unsafe to abandon the aircraft."

Unfortunately, the alternative — bailing out — is an option most general aviation pilots do not have, so they must cope with whatever Fate hands them.

Engine fires usually are caused by the mechanical failure of the engine or an engine-driven accessory, defects in the induction or exhaust system, a flooded carburetor (caused by a sticking float valve) or a leak in the fuel, oil or hydraulic system.

Although the sight of flames or smoke from under the cowling is convincing evidence of a fire, these are not always the first indications. Pilots of single-engine airplanes who have survived engine fires report that the first sign often was the heat generated in the vicinity of the rudder pedals. They literally got a hot foot (or two). Others stated that the first indication was the smell of smoke. One pilot mentioned that he was aroused by the indication of reduced fuel pressure. He thought at first that the fuel pump was failing, but the flames that developed soon thereafter made him realize that the pressure loss was caused by a serious leak.

When a pilot first notices an engine fire, it already may be well developed, leaving no choice other than to shut down the engine as soon as possible. An engine fire must not be allowed to continue unabated.

In theory, the firewall separating the engine compartment from the cabin of a single-engine airplane is supposed to keep a fire at bay. But it does not always work that way. Although these bulkheads are ground tested by exposure to flame, this does not mean that they can contain a fire while airborne. The variables of airspeed and flame intensity make it impossible to determine how long the integrity of the firewall can be maintained. In other words, a firewall is fire-resistant, not fireproof. Sufficiently intense and persistent flames can penetrate a firewall and cook the pilot's goose.

In a twin, the danger of allowing an engine fire to continue is equally serious. Once a firewall begins to fail, the very structure of the wing is in jeopardy. And should the flames attack a nearby fuel tank, the results are likely to be explosive.

Once the decision has been made to shut down the offending engine, it must be done without hesitation, even when flying over inhospitable terrain in a single-engine aircraft. A controlled crash landing is preferable to aerial cremation.

The obvious purpose of a shutdown is to deprive the fire of additional fuel and is. accomplished by closing the throttle and retarding the mixture control to the idle cutoff position. Additionally, turn off the electric fuel pump (if applicable), and turn the fuel selector valve to the Off position. Do not, however, be in a rush to turn off the magnetos. Leaving on the ignition for a moment or two may allow residual downstream fuel to be burned in the safest possible place, the cylinders.

With some luck, starving the engine compartment of fuel may cause the fire to die naturally. But even if the flames, smoke and heat seem to disappear, do not assume that the problem has been resolved. The. fire may only have diminished temporarily, or the smoke and flames may be coming out the bottom of the cowling.

Under no circumstances should a pilot entertain the notion of a re-start. The fire may rekindle with even greater intensity.

Several aircraft operating handbooks also suggest that the master switch be turned off during the shutdown procedure. Unless the fire is electrical in nature or a crash landing is imminent, this is of dubious value. Deactivating the electrical system precludes the possibility of transmitting a Mayday message (assuming a pilot has the time, desire and presence of mind to make such a broadcast). More importantly, this prevents the use of electrically operated landing and flap systems during the subsequent descent and landing.

The possibility of asphyxiation also is of immediate concern. To prevent smoke from entering the cockpit, dose all fresh-air and cabin-heat vents located either in the firewall or along the sides of the fuselage. (Some aircraft have a firewall shutoff that should be closed as well.) Air vents located in the wing roots of high-wing aircraft or in other locations free of streaming smoke or flame should be opened to increase the flow of fresh air through the cabin. (Those not thoroughly familiar with the sources of air for ventilation and cabin-heat ducts should review the plumbing of the environmental control systems for their aircraft. Vents and the like often are taken, for granted and given only cursory attention but can be vitally important.)

If the cabin does fill with smoke, the air often can be cleared expeditiously by opening a door or a window on the right side of the cockpit (away from the pilot) as long as the opening is clear of trailing smoke and flame. Do not open the pilot's door or window because this may cause the smoke to pass in front of him on the way out and compound his breathing problem. (The smoke-removal check list of a Boeing 727 calls for depressurizing the aircraft and opening the first officer's sliding-glass window when airspeed is less than 250 knots indicated, a procedure that seems to have the unanimous approval of captains.)

Howard Hughes was said to be almost paranoid about the possibility of asphyxiation. This probably stemmed from his fiery crash landing of the XF11 photoreconnaissance-plane in 1946. Prior to flying his HK-1 "Spruce Goose" a year later, he had a large-diameter tube installed near his seat that would allow ambient air from outside the aircraft to blow in his face in the event smoke filled the cockpit for any reason. (None of the other crewmembers was provided such emergency equipment.)

Occasionally pilots are told that it is possible to dive an airplane to blow out an engine fire. This technique may be successful. On the other hand, the effect of increasing the mass airflow across the engine might cause fire intensity to increase. It all depends on the combustible mixture. For example, increasing airspeed encourages the burning of a rich mixture but may extinguish a lean mixture. Since a pilot has no way of assessing the combustible mixture, he cannot know whether to combat the fire by increasing or decreasing airspeed. (If some sixth sense suggests that a dive is in order, do not forget to open the cowl flaps — if available — to maximize airflow through the cowling.) One thing is certain, however: An oil-fed fire is much more difficult to blow out than one fed by fuel.

It is not particularly difficult to distinguish between fuel and oil fires. Burning avgas produces bright orange flames and leaves a smoke trail that often can be seen only by looking behind the airplane. An oil fire, however, is characterized by a profusion of dense, black smoke and may be accompanied by indicati6ns of low oil pressure and high oil temperature, depending on the size of the oil leak.

Although oil dripping on a hot exhaust manifold is sufficient to cause ignition, an oil fire also may be the result of a fuel fire that bums through an oil line. So it is quite possible that an engine fire that begins as one type may develop into another.

This indicates further the urgency of engine shutdown. If a fuel-fed fire is denied a supply of avgas soon enough, it may die of starvation before igniting an oil fire. This latter type can be much more persistent and difficult to extinguish. This is because an engine-driven oil (or hydraulic) pump continues to pump flammable liquid and feed the fire as long as the propeller windmills. In this respect, the pilot of a multiengine aircraft has a decided advantage. He can feather the propeller, resulting in a stilled engine and an impotent oil pump. Although this does not guarantee that the fire will go out, it limits its destructive potential.

If the pilot of a single-engine airplane is sitting behind a raging oil fire, he too can combat the inferno by halting propeller rotation. However, he must raise the nose and reduce airspeed until the resistance of engine compression overcomes the waning aerodynamic forces that cause a propeller to windmill. (Stopping a constant-speed propeller can be assisted by fully retarding the pitch-control lever to the minimum-rpm position.)

This process of arresting engine rotation has the disadvantage of temporarily reducing the sink rate and increasing the time required to lose altitude. If a pilot does only one thing correctly when faced with an engine fire, it should be an expeditious descent and landing. This is not the time to stretch the glide in an attempt to reach some optimum landing site unless the fire is known to be out. Instead, select a nearby piece of terrain and plant the aircraft on it as quickly as is possible and survivable. One must adopt the attitude that the airplane has been lost, that it can be sacrificed if necessary to preserve the lives of those inside. When confronted with a persistent engine fire, the element of time is extremely critical. Give a fire enough time and it will consume you.

This message is particularly important to the pilot of a twin, who may be tempted to use the remaining engine to hobble to some distant haven. When fire rages, the only value of an operative engine is the assistance it can provide in accurately placing the airplane on a selected touchdown point.

The method of descent is at the discretion of the pilot, but most experts agree that the airplane be configured so as to achieve the maximum rate of descent consistent with keeping the wings and tailfeathers firmly attached. With some aircraft, this may necessitate extending the landing gear or the flaps or both and then diving at the maximum-applicable airspeed.

Although pilots correctly are taught never to exceed various limiting airspeeds, this may not be the time to be so conscientious. If the aircraft is on fire, who cares whether the maximum-allowable, landing-gear-extended airspeed is violated? Besides, airplanes are built with enough of a safety margin so as to be able to withstand a one-time emergency violation of structural limits. (But do not get too carried away.)

When descending at high speed in a single-engine airplane after propeller rotation has been stopped, consider that the increased aerodynamic forces may overcome engine compression and cause propeller windmilling to resume.

When discussing engine fires in single-engine airplanes, the subject of slipping usually arises. Presumably this is recommended when smoke from an oil fire obscures forward visibility or when flames lick the cockpit. Slipping may divert the smoke and flames toward one side of the airplane, but it is essentially a low-speed maneuver that results in less of a sink rate than diving, factors that must be weighed. A pilot might decide that diving initially would be more beneficial and that slipping would be more appropriate on final approach when forward visibility is more critical.

There comes a point during the descent when a pilot must shift mental gears and transition into the landing approach. This requires slowing the airplane and planning for the safest possible landing. It is of little benefit to survive an engine fire if the flight terminates in a lethal crash landing.

If a pilot is so fortunate as to be landing at a controlled airport, he should not hesitate to request that emergency equipment stand by, even if the fire appears to be out. The reduction of airspeed during an approach and landing could cause a smoldering fire to rekindle with a vengeance. This also explains why an orderly evacuation should be conducted as soon as the aircraft comes to rest and why everyone should get away from it as quickly as possible.

Many engine fires can be avoided by methodical preflight inspections. Pilots should be on the alert for puddles, streaks and stains of fuel, oil and hydraulic fluid, both under and inside the cowling. They also should be able to locate the various fuel and oil lines and know how to check their security and condition. Sometimes a good sniff under the cowling can detect a slow fuel leak before it develops into one that can be seen. Be sure also to check exhaust manifolds for security by tugging gently on the ends of the stacks. Exhaust leaking from a cracked stack can be an ideal source of ignition.

Of course, many pilots do not take the time to inspect the engine properly, trusting the statistical infrequency of in-flight engine fires and modem powerplant reliability to insulate them from a horrifying airborne emergency. But in so doing they give up one of the only two weapons against fires that we have; to lessen the odds, we have only preflight inspections and procedural proficiency. The rest is luck. And a sudden gush of flame is harsh notice that your luck has run out.


Barry Schiff, AOPA 110803, is a captain for a major airline, an FAA-designated examiner and has more than 16,000 hours in 200 types of aircraft.