Technique: Running on empty

Fuel management basics

March 1, 2010

We’ve all heard the maxim—“The only time there’s too much fuel in an airplane is when it’s on fire.”

That’s not entirely true, of course, but carrying an insufficient amount of fuel, or not being able to get it to the engine(s), has been a frustratingly persistent cause of aircraft accidents for generations. About 200 GA accidents in the past five years were attributed to pilots running out of fuel, according to the AOPA Air Safety Foundation.

In recent years, however, general aviation pilots have been handed some valuable new tools, and adopted new practices, to avoid such problems—and the results are encouraging. No technologically advanced GA aircraft, for example, has ever had a fuel-exhaustion accident.

Graphical cues such as the range rings on Garmin G1000s, or the color-coded time- and distance-remaining numbers on Avidyne multifunction flight displays (MFDs), have proven tremendously beneficial, even though in most installations they depend on fallible pilots to tell them how much fuel is in the tanks at the start of each trip. The vast majority of the GA fleet doesn’t have such brainy avionics, but items ranging from a simple wooden dipstick to sophisticated electronic fuel computers can do wonders.

The most important first step for pilots is to know the fuel systems in the airplanes they fly. Does gravity send fuel to the engine, or a fuel pump, or a series of fuel pumps? Are the fuel pumps engine-driven, electric, or manual? How much usable fuel is in each tank, and can the pilot transfer fuel between them? Is there a fuel return line from the engine to any particular fuel tank? Can the pilot use fuel from more than one fuel tank at a time? When should the fuel stored in tip tanks be transferred to the mains? At what rate does it flow? And can a pilot pump fuel overboard without realizing it? Such questions are much more than rhetorical, and they carry direct flight safety implications.

Every student learns the FAA fuel requirements for VFR and IFR flights, during the day and at night, in training (30 minutes reserve for day VFR; 45 minutes reserve for night; to the destination, then alternate, at normal cruise power with 45 minutes reserve for IFR). But most experienced GA pilots prefer to land with more than that.

“My own hard limit is 10 gallons,” said John Kounis, a California pilot and founder of Pilot Getaways magazine, who makes several transcontinental trips each year in a Cessna 185. “I’ve got a very accurate fuel computer in my airplane, but I’d never consider flying with less than 10 gallons—about 45 minutes’ endurance. And I try to always land with at least an hour of reserve fuel.”

The AOPA Air Safety Foundation strongly recommends that pilots land with at least one hour of fuel in the tanks. Don’t flight plan for one hour remaining—land with that much.

We all like to fly direct GPS routes to our destinations, but pessimism in flight planning can be a great help in the air. Carry enough fuel so that circuitous routings and unanticipated headwinds aren’t a concern. And when flight planning, assume your airplane will get less than book performance. The numbers in the POH were determined under ideal conditions with a new, perfectly operating aircraft. Your airplane may use more fuel, or travel a bit slower, or both—even if you’ve leaned the engine aggressively in cruise. (And make sure to lean the engine aggressively in cruise.)

High fuel prices have got all of us using online tools to find the least costly avgas, especially on long cross-country trips. But consider calling ahead to the FBO at your planned fuel stop to make sure fuel is available. It would be a shame to find the fuel was sold out, or the fuel truck was out of service, after you landed.

In the cockpit, remember that aircraft fuel gauges are notoriously unreliable. They depend on floats and sensors and other failure-prone components, and the needles tend to jump around in turbulence or when fuel pumps are turned on or off. Also, keep in mind that FAA regulations only require that the gauges display accurately when the tanks are empty.

Last year I made a long cross-country trip in a 1976 Piper with a pair of disco-era float gauges whose indications displayed a distant and tenuous relationship to the amount of fuel within. The airplane also had a brand-new electronic fuel computer that came with its own lawyerly placard warning pilots not to depend on it. Throughout the trip, the fuel computer was astonishingly accurate in judging fuel consumption and range, while the fuel gauges themselves were next to useless.

The Cirrus SR22 has accurate mechanical fuel gauges as well as a state-of-the-art fuel computer. But the fuel computer is precise to the drop while the mechanical gauges stay within a few gallons. On cross-country trips, the usual practice is to take off and climb to altitude on the left tank, then switch to the right tank as soon as 10 gallons of fuel has been consumed. Then stay on the right tank until an additional 20 gallons disappears.

The airplane’s two Garmin GNS430Ws are programmed to remind the pilot to check the fuel quantity every 30 minutes, but that doesn’t mean switching tanks that often. It usually takes about a half-hour to consume the first 10 gallons because the airplane is climbing at full power for most of that time—then it takes almost two hours to burn the next 20 gallons. Cirrus recommends that pilots limit fuel imbalances between the two wing tanks to 10 gallons or less, and that’s far simpler to do with precision and confidence using the fuel computer on the MFD (and the mechanical fuel gauges as a backup) than simply watching the clock.

GPS provides constant groundspeed and estimated time remaining to the destination, invaluable numbers for showing real-time winds aloft. But the GPS doesn’t know the winds ahead, and pilots should know from preflight weather briefings, in-flight weather services, and perhaps onboard satellite weather reports whether the winds aloft are likely to get more or less favorable as each leg progresses.

Some pilots zealously keep the fuel balance between multiple tanks as close as possible during flight. But when approaching the end of a long trip, when the fuel remaining is getting light, equality isn’t necessarily a virtue.

Think about it this way: If you’re on approach in an airplane with two fuel tanks and 10 gallons remaining, would you rather have five gallons in each tank, or all of it in one? Having all of it in one tank, and the fuel selector on that tank, is the better option because it avoids the need to change fuel tanks during a high-workload phase of flight at low altitude. Ideally, pilots should keep a sufficient quantity of fuel stashed away in one tank to fly the entire approach and landing on that tank without a last-minute, low-altitude change.

Some pilots run fuel tanks completely dry in flight because it gives them precise information about actual fuel consumption. The sudden silence when a tank empties lets them know exactly how long it took to drain a known quantity of fuel at a particular power setting, and gives them a strong hint about how much endurance the remaining fuel will give them. It’s a strategy that makes sense intellectually, but passengers generally don’t like the sound of an aircraft engine shutting down—even if the pilot tells them in advance it’s going to happen.

I used to fly a particularly thirsty warbird with four fuel tanks and, when trying to stretch its range, would be reluctant to switch away from any tank that still contained fuel. I’d count down the minutes and stare at the fuel pressure gauge for the flicker that would let me know a tank was about to run dry. My hand would be on the fuel selector knob and poised to make a quick switch to avoid losing engine power. Invariably, that’s the moment ATC would call with a frequency change or an amended clearance, and I’d be tuning the radio or writing when the engine died, the airplane decelerated, and my backseater jumped out of his skin. The engine always revived a few seconds later once the fuel flow was restored—but those were some very long seconds.

E-mail the author at