Dry-tank disasters

August 1, 2001

Bruce Landsberg has served as executive director of the AOPA Air Safety Foundation since 1992.

Football coaches all over the country will tell teams, veterans, and rookies alike when returning for summer practice that good fundamentals win games. Some pilots have not mastered a really fundamental concept of safe flight: having fuel on board in sufficient quantity and quality. Additionally, managing the system properly seems self-evident, but there are some traps that catch more of us than you might suspect.

The AOPA Air Safety Foundation has teamed up with the FAA's Aviation Safety Program to make a concerted effort to stop dry-tank disasters. From now until the end of the year, the FAA and ASF will conduct seminars on fuel management. The FAA, working with ASF, sponsored a videotape and Safety Advisor publication to get the word out. Phillips Petroleum's aviation division also assisted with the advisor as a public service.

Have you heard the fuel-management story before? Probably, but then so had the accident pilots. Is the concept simple? Yes, but if it was that simple, why the disturbing record? The number of aircraft damaged or destroyed by this little oversight averages more than one a week. According to ASF's 2000 Joseph T. Nall Report, looking back at 1999's accidents, there were 61 fuel-mismanagement accidents, but fortunately only five involved fatalities. Since 1983, general aviation has racked up more than 2,400 aircraft with no combustible fluid in the combustion chamber. The only bright spot in this picture is that the numbers have dropped from well over 200 accidents per year in the early 1980s to about one-third of that in the past few years. That's still a lot of grief for such a preventable cause!

As with most other pilot-related accident categories, personal flight operations had, by far, the greatest number—and while flight instruction was underrepresented in proportion to the amount of instructional flying, there were still eight accidents.

The airlines occasionally have their problems as well, and the reasons are much the same as in general aviation. For example, on December 28, 1978, a United Airlines DC-8 crashed into a wooded area near Portland, Oregon, during an approach. The flight had loitered at low altitude for about one hour while the crew coped with a landing gear malfunction and prepared the passengers for the possibility of a gear failure upon landing. The airplane crashed about six miles southeast of the airport when all four engines flamed out. The aircraft was destroyed and of course there was no fire. Of the 189 occupants, eight were killed and 23 were seriously injured. The NTSB determined that the probable cause was the captain's failure to monitor the aircraft's fuel state and to respond to crew advisories regarding fuel state. This resulted in fuel exhaustion to all engines.

This was one of the accidents that led to the crew resource management (CRM) movement. The captain had his own idea of how to handle the situation and ignored the crew's concerns. The first officer and the flight engineer also did not state firmly to the captain that "in 20 minutes, sir, we will be landing whether the cabin is ready or not, because there is no more fuel."

In 1989, an Avianca Airlines Boeing 707 crashed on Long Island, New York, while attempting a second approach into John F. Kennedy International Airport after being held three times for a total of one hour and 17 minutes after a flight from South America. The adverse weather in the Northeast with thunderstorms and wind shear compounded the problem, as did ATC procedures. Miscommunication between the pilots and ATC contributed to the accident. The controllers failed to appreciate the criticality of the problem, and the crew, having some difficulty with English language proficiency, were unable to declare an emergency.

The common lesson is that the pilot in command must monitor the fuel condition. Do not think in terms of gallons or pounds—that is an abstract approach to a very real problem. Always consider your fuel state in terms of time. Fuel burn is nearly a constant but the other conditions affecting the flight are not. As a pilot becomes knowledgeable about a particular aircraft he can closely estimate the burn in gallons or pounds per hour.

What changes dramatically from flight to flight, thanks to wind, is groundspeed. A classic GA fuel-exhaustion accident occurred in November 1994 when a Beech B55 Baron crashed just a mile short of the runway of Fulton County Airport-Brown Field in Atlanta (see " Safety Pilot: Too Little, Too Late," March 1996 Pilot). The pilot had made the trip at least a dozen times before, but the Baron was slugging into headwinds beginning at 25 knots and increasing to 40 knots as the flight moved south. An IFR flight plan was filed estimating 3.5 hours en route with 4.5 hours of fuel on board. The alternate was listed as Dekalb-Peachtree Airport, which is about 30 miles from the destination airport Covington Municipal in Georgia. Dekalb-Peachtree Airport was below minimums so the pilot decided on Brown Field. The weather was forecast to improve with low clouds and fog but did not. There was ample opportunity to check the trend and divert to an alternate with better weather, rather than continue into very low conditions with practically no fuel. The aircraft ran out of fuel within minutes of when the pilot's operating handbook said it would; it's just that the range was far less than it had been on earlier trips.

Everybody can recite the rules for fuel requirements. For VFR flight, FAR 91.151 requires enough fuel to get to the destination at normal cruise speed and continue on for another 30 minutes. If you plan to arrive after dark, then 45 minutes' reserve is required. Under IFR, FAR 91.167 makes life slightly more complicated but not much. Carry enough fuel to get to point A, shoot the approach, go to point B (the alternate) and shoot the approach, and then be able to fly another 45 minutes after that at cruise speed. There are some exceptions to this if the weather meets certain criteria. However, these are the minimums and if things get slightly fouled up, as they have a tendency to do when you really need for them to go right, the engine becomes quiet and the aircraft lands—now.

I'd propose a modestly better idea. Plan to be on the ground, not looking for a place to land, with one hour's reserve. There has been one occasion in my personal experience when the pilot I was with pushed the envelope. There are few things more unpleasant than trying to figure out if you've got enough to get there and knowing that the whole fiasco is of your own making. I've heard some creative excuses in the aftermath, but guess where the buck stops? With the pilot in command. Apparently the FAA and the insurance companies seem to agree on this—it's not an audience I would welcome.

How do pilots get into such binds? Let me count the ways. First, is the tank full—really? Everyone knows that aircraft fuel gauges leave something to be desired, but they are required by the FARs to be accurate in only one place—empty. If the gauge says empty, believe it. Some gauges show full when a tank may be down by as much as 30 minutes. The Cessna 210 series, because of the geometry of the tank, is one aircraft that requires careful and patient refueling.

Knowing the system is another part of the deal. There are all manner of boost pump operation schemes—sometimes they're supposed to be on for takeoff and landing, and in other aircraft they are only used for priming and engine-driven pump failure. Get it wrong and the flow may not go.

Starvation is different from exhaustion. Either way, the engine stops but in starvation there is still fuel on board—it's just that the system isn't configured to get it to the engine. Ever forget to switch tanks? Probably everyone has at some point but we usually catch it because the aircraft gets wing heavy; we look at the clock or a gauge and get the message.

A Beech A36 Bonanza made a forced landing with no injuries. According to the pilot, when the airplane was approximately six miles from an airport (not his destination) he experienced a total loss of engine power. He advised ATC of the situation, and the controller provided radar vectors. The pilot kept the airplane in a "clean" configuration to improve glide performance, but the flight contacted trees and slid to a stop in the grass only 300 feet from the runway threshold. The pilot said the loss of engine power was from "fuel starvation" and that he was distracted from monitoring the fuel because of a conversation with a passenger. He added that the autopilot was on, and the weather was "beautiful." With the autopilot on, it will mask the wing heaviness that shows a developing fuel imbalance.

There are some strategies to help with this perfectly human failing. If the aircraft has low-fuel warning lights, as some of the newer ones do, that's a nice backup feature. Barring that, set a timer on takeoff and mount it prominently where it will be in your scan. Switch tanks after one hour of flight, then fly two hours on the other tank, then go back to the original tank. For most aircraft, you'll be about ready to land fairly soon after that and this procedure will give you something to do on an hourly basis. Some new avionics have fuel timers built in so that at whatever time you designate, the box flashes to remind you that it's time to switch.

In the ASF database there are literally thousands of accidents to choose from, and every one has a story—but the basics fall in to three general categories: The pilot knew it was going to be tight and pushed it anyway; the pilot was distracted by something and failed to take action at the appropriate time; or the pilot did not understand how the system worked.

Here's another system gotcha. An aircraft has fuel tanks that prohibit takeoff when the gauge is in the yellow arc, which is typically the bottom third or quarter. The pilot ignores the warning, ironically, because he's just going to a nearby airport to refuel. Shortly after liftoff the engine quits, probably because of unporting. The manufacturer has learned through hard experience that when the attitude of the aircraft changes from level to climb with so little fuel, the standpipe may be uncovered momentarily and the engine is sucking air.

If you ever get into a fuel "situation" advise ATC that you have minimum fuel and cannot take any delays. That usually will get you to the head of the line with no enforcement repercussions. If you don't get satisfaction, then declare an emergency. If you've followed the procedures that we discuss in the seminar, the odds of the problem escalating to this magnitude are slim. If you've done everything you should, there should be nothing to fear, but do let someone know while there is still time to resolve the problem. We encourage you to attend either the FAA or ASF program on fuel management before the end of the year. If nothing else, it may inoculate you against one of aviation's most common mishaps. Don't be left high and dry.

ASF's Fuel Awareness Safety Advisor is available on the Web site or can be ordered by calling ASF (800/638-3101). Links to additional information about fuel awareness may be found on AOPA Online.

See also the index of "Safety Pilot" articles, organized by subject.