Probable Cause #21: No Pulling Over


This article originally appeared in IFR Refresher, Oct. 2005.

Many years ago a friend of mine who owned a Chevy Nova wanted to know exactly how far he could go on a tank of gas. So he put a five-gallon can of gas in the trunk and drove until the tank ran dry. I happened to be riding with him to a Civil Air Patrol meeting when the engine sputtered and died. We pulled over to the side of the road, poured the five gallons into the gas tank and continued on our way. I don’t remember how many miles my friend said he got from that full tank of gas, but the accident we’re going to discuss reminded me of his experiment.How many times have you driven with the fuel gauge hovering on “empty” and have passed a gas station for one reason or another? I would say that most of us are guilty of this at least once or twice. If we run out of gas, we simply pull off to the side of the road like my friend did. It might be an inconvenience because most of us don’t carry five-gallon cans of gas in the back of the car, but it would be unusual for someone to get hurt or killed just because they ran out of gas.The ramifications of running the tanks dry in an airplane are far more serious and our flight instructors tell us this from the first day we set foot into a cockpit. Most of us heed that lesson but there are always some who don’t. None of us can afford to pass up an airport where fuel is available because we think we can make it to our destination. If we don’t know for sure that we have the fuel on board to get where we want to go then we are bound by good judgment and common sense to make a fuel stop and get some more, even if it means we will be late for Cousin George’s third wedding.The pilot of a Beech B60 Duke apparently thought he had enough fuel to make it to his destination, but he miscalculated by about 13.5 miles. The pilot had departed Houston’s Hobby Airport (KHOU) on an IFR flight to Dekalb-Peachtree Airport (KPDK) in Atlanta but he would never make it. Three hours and 13 minutes later the airplane crashed half a mile short of the runway at Fulton County Airport (KFTY), an impromptu alternate and just 13 miles from KPDK. The pilot, who was the only person on board the aircraft, was killed.It would seem that this is an open and shut case of fuel exhaustion, and it is. Even so, there is more to the story, facts that can’t help but raise questions about the pilot’s attitude towards fuel planning and safety.

A Known Problem

The story begins nearly three weeks earlier, on a day late in July, when the pilot flew his Duke to Houston from Atlanta. Upon landing at Hobby Airport, he experienced a complete loss of engine power. He told the line service supervisor that he had flown approximately 100 miles with one engine feathered and that the second engine quit when the aircraft was on the ground. However, he was able to get both engines restarted and taxied in to the ramp. He told the supervisor to put 40 gallons of avgas in the aircraft’s tanks, 20 gallons per side, and the engines apparently ran normal after that.Twelve days later the aircraft returned to KHOU, but this time the pilot turned the airplane over to a maintenance shop on the field with the complaint that both engines quit at one time or another when the throttles were brought back to the idle position.A mechanic ran the engines the next day and verified the complaint. He suggested to his supervisor that both fuel servos should be sent out to a repair station for evaluation. The units were removed from the aircraft and sent out to a local facility. When they were returned, the facility said that nothing could be found wrong with them, although the right one ran rich at low speeds.The facility spent part of several days making adjustments to the engines’ fuel systems. In the meantime, the shop supervisor contacted the pilot and asked him if the problem had occurred only when the weather was extremely warm. The pilot indicated that that was the case and that it had occurred once with each engine. The shop supervisor then asked the pilot if he had the fuel boost pumps on when this occurred. This time, the pilot wasn’t sure and acknowledged that he may not have turned them on.The Duke is powered by two Lycoming TIO-541-E1C4 engines that produce 380 horsepower each. This is a high-performance, turbocharged, fuel-injected powerplant that requires skill and knowledge to operate. The Duke’s flight manual states that the boost pumps should be turned on for takeoff and landing and that they should also be on in cruise when needed to prevent fuel flow fluctuations.The shop supervisor felt the problem occurred because the boost pumps were not turned on, allowing the heat inside the cowlings to vaporize the fuel in the lines and creating a vapor lock. That would cause the engines to run poorly on the ground and possibly to quit when the throttles were retarded. He did some additional research and was told that, to be safe, he should be certain that the fuel systems were properly adjusted.The following Friday the pilot called the shop and told them to button up the airplane because he had to fly back to Atlanta. The supervisor explained to the pilot that they were not yet finished with the work: The right engine setup was complete, but the left engine needed to be matched to the right. The pilot said he needed the airplane, but that he would return in a week and that they could finish the work at that time.This put the supervisor in a tough spot. While he felt the airplane was safe to fly, he couldn’t return the aircraft to service because the work was not yet complete. That apparently didn’t faze the pilot, who agreed to take the airplane even though there was no sign-off in the logbooks.The aircraft was towed out of the hangar and line service personnel topped off the plane with fuel.

Need Some Help

At 5:50 p.m. Central Time, the pilot picked up his IFR clearance to KPDK and was airborne by 6:31 p.m. A little less than three hours later, the pilot, who was now talking to Columbus (Ga.) Approach, reported that he was getting low on fuel and requested a direct clearance to KPDK. The controller asked if he was declaring an emergency but the pilot said no. The controller then switched the aircraft to Atlanta Approach Control.Upon initiating contact with Atlanta Approach, the controller verified the aircraft’s low fuel state and asked how much fuel the pilot had. He responded that he had about 30 minutes remaining. Although there were no delays going into KPDK, the controller suggested that the pilot land at another airport and offered to help him find a suitable alternate. The pilot agreed and asked for an airport with more than 6,000 feet of runway.The controller suggested Fulton County Airport, which was 45 miles and about 15 minutes from the Duke’s position. He told the pilot that the airport had a control tower and a runway that was 5,796 feet in length. The pilot advised that the runway length was fine and requested a change of destination to the Fulton County airport. The controller also advised the Duke pilot that KFTY was 13 miles from Peachtree and that he could always re-evaluate the situation once he got closer to Fulton County as to whether he would land at Fulton or continue on to Peachtree. But the Duke pilot told the controller that he would land at Fulton County, perhaps an indication that the pilot realized that his fuel state was reaching a critical status.At 10:39 p.m. local time, the pilot told the controller that he had the airport in sight. The controller cleared the aircraft for a visual approach to Runway 8 and instructed him to contact Fulton County Tower. Just before making the switch, though, the pilot reported to the approach controller that he was having “some problem with the engine” and requested a straight-in approach, which the controller approved.When the pilot made his initial call to the Fulton County tower, he advised that, “We are probably going to need some help off the runway.” One minute later he told the controller that he had lost both engines and that he didn’t think he could make the runway. That was the last transmission from the pilot.

Know The Limits

The aircraft came to rest on a grassy island between the parking lot of a commercial business and a public service road in an industrial park. An FAA inspector was at the site of the accident approximately 45 minutes after it occurred. Although both fuel tanks were ruptured, the inspector found no sign or smell of fuel at the accident site. The only fuel retrieved was about three ounces from the plane’s fuel system.A subsequent investigation of the engines and fuel system failed to reveal any contributing mechanical problems. The weather at the time of the accident was VFR with calm winds.The 40-year-old Duke pilot held a private pilot certificate with single-engine, multi-engine and instrument ratings. The NTSB report stated that the pilot’s total time was approximately 1,800 hours, but an FAA inspector’s report put his total time at 1,282 hours, 26 hours in type, and 30 hours of flight in the preceding 30 days. He had a current medical certificate and biennial review. The inspector’s report also indicated that the pilot had taken recurrent training during the last three years, but there was no information about how long he had been flying the Duke.The fact that the pilot ran out of fuel on the westbound leg to Houston on one trip, and then had the problem with the engines not idling on another, seems to me to be unrelated. As the mechanic suggested, the hot July and August weather probably played a part in the idling problem. Coupled with the pilot’s apparent lack of knowledge of the airplane’s systems, his not turning on the boost pumps before landing probably contributed to the incident.Lycoming turbocharged engines run hot. When there is little cooling air inside the cowlings, like there would be after touchdown on a hot day, the boost pumps force cooler fuel from the airplane’s tanks into the engines. If the pumps are not turned on the fuel in the lines under the cowling will vaporize before reaching the engines causing them to sputter and quit.The question is, why would the pilot attempt the same flight, albeit going eastbound, when he – from what we can tell – ran out of fuel the first time? Most pilots who have that experience will not have it a second time, because they will insure that the amount of fuel on board is sufficient at all times.Another question is whether or not the engines were being properly operated. The Duke has an advertised range of 1,010 NM so the flight from KHOU to KPDK – a direct-line distance of 615 NM – is well within the plane’s capabilities. But to achieve that requires careful power/ fuel management and monitoring. Failure to do so can cause excessive fuel consumption resulting in a considerable reduction in range.We’ll probably never learn the answer to those questions, but we can still learn a few lessons from this accident:

  1. If you don’t know how many hours you can fly your airplane to dry tanks, then you need to find out. That doesn’t mean you will do what my friend did with his Chevy Nova. Instead, over the next two or three short trips, figure out what the overall fuel burn per hour is for the aircraft and divide it by the number of gallons you have on board. That will give you a good idea of how long you can stay in the air. On longer flights the fuel burn should improve, but the difference between the short trip figures and the long trip figures will give you a little extra time aloft beyond your planning times in case of an emergency. However, you should never allow the fuel levels to get so low that you need to depend on that difference.
  2. If the airplane is new to you and you don’t have enough experience with it to know exactly what the fuel burn is, be very conservative when flight planning. Each time you fill up the tanks you will have more fuel burn information that will help you with your future trip planning.
  3. Determine how many hours you are willing to stay aloft in IFR conditions and VFR conditions and never exceed those times. If it becomes obvious after takeoff that a flight will take you beyond your maximum allowable flight time, find an airport, land and get some more fuel.

If you plan your flights so tight that you are on the ragged edge of making it to the airport, what are you going to do if you are told to hold at the destination because there is an airplane stuck on the runway or the landing gear doesn’t come down and you need time to troubleshoot the problem? You need fuel on board to give you the time to do these things. Don’t press your luck. An extra fuel stop may take some time, but it is better than winding up half a mile short of the runway being carried away on a stretcher.

More accident analyses are available in AVweb’s Probable Cause Index. And for monthly articles about IFR flying including accident reports like this one, subscribe to AVweb’s sister publication, IFR Refresher.