The Case of the Absent Avgas

A doctor, 67 years old, and a friend flew in a club Cessna 182E Skylane from Clemson, South Carolina, to Orlando, Florida, to pick up a crankshaft. The weather was fine and the flight apparently uneventful. After completing their errand at Orlando Executive, they took off for a 35-mile hop to Massey Ranch in Edgewater, possibly to get fuel there for the return trip. The price of fuel was substantially lower at Massey than at Executive, and they could save a hundred bucks or so.

According to the National Transportation Safety Board’s report, on the day before the flight the doctor, as the pilot, had fueled the 182E with 21.1 gallons (or 21.3—both figures are in the report) from a self-service pump at Oconee County Regional in Clemson. Because the pilot was making a longish trip the next day, he would probably have topped the tanks; the nonround number of gallons suggests that he did.

The two wing tanks of the 182E together hold 65 gallons, of which 5 gallons are unusable. This airplane was equipped with a supplemental 18-gallon tank behind the rear seats, from which fuel could be pumped to the right-wing tank, but there is no reason to believe that it was in use.

The flight from Clemson to Orlando, a straight-line distance of 375 nm, took 3.1 hours, for an average groundspeed of 121 knots. The airplane, which had a 285 hp Continental O-520 in place of the stock 230 hp O-470, would be expected to achieve something more like 150 ktas at 75 percent power with a fuel flow of 16 gph. It was equipped with a JPI engine-monitoring system, however, according to which the fuel flow had been 15 gph during almost the entire cruising segment of the flight—corresponding to an expected true airspeed of 147 knots—and 50 gallons had been used. The discrepancy in speed could have been caused by a headwind or a circuitous route.

A little before 11 a.m., after a 14-minute taxi and hold, the 182E took off from Runway 25 at Orlando Executive and turned northward toward Massey Ranch. The local controller issued a series of vectors and altitudes to keep it clear of Class B airspace. About 5 miles out, the pilot declared an emergency, reporting to the local controller that fuel was not flowing from the right tank.

A witness in a boat on a small lake reported seeing the airplane pass overhead northbound. The engine seemed to be sputtering, and the airplane turned back. It appeared to the witness that the pilot intended to ditch in the lake, but the airplane snagged a tree at the shore, nosed over, dived violently into the water and sank. Both occupants died on impact.

Retrieved from 20 feet of water, the 182E was found to have suffered remarkably little damage. Its fuel tanks were not breached; each contained 2 gallons of fuel. Accident investigators secured the airplane to a trailer and returned the drained fuel to a tank. The engine started without difficulty and ran for several minutes.

At the time of the crash, the right tank was selected, the electric boost pump was off, and the transfer pump for sending fuel from the aft tank to the right-wing tank was also off. The engine monitor’s record of fuel flow, CHT and EGT clearly showed the supply of fuel to the engine beginning to fluctuate about two minutes before the pilot’s emergency call.

The NTSB blamed the pilot’s “inadequate preflight planning” for the accident, which had plainly been caused by fuel starvation. It did not attempt to answer, nor even to ask, the obvious question: If the airplane had 60 gallons of usable fuel at the start of the trip, why did it run out of fuel after only 50, or at most 52, gallons had been used?

First, we need to understand the meaning of “unusable fuel.” Unusable fuel can be of two types: fuel that cannot flow from a tank even in level flight and fuel that might become unavailable during a maneuver such as a steep climb, slip or turning takeoff. The relevant FAR, 23.959, does not distinguish between the two categories, so it is not always possible to know how much of the officially unusable fuel is actually usable provided the airplane is more or less level. It’s noteworthy that the 182E handbook warns that the fuel selector should be on both for takeoff because, in a slip or skid, tanks may unport even when one-quarter full.

Let’s assume that the pilot of the 182E—who, with 1,200 hours and an instrument rating, was not a beginner—was aware that the airplane carried 60 gallons of usable fuel in its wing tanks.

Let’s further assume that he did not use the auxiliary tank to any degree because, if he had, the accident would probably not have occurred.

From the totalizing function of the engine monitor, the pilot would have known that he had used 50 gallons for the flight from Oconee to Orlando. Ten presumably remained. The flight to Massey Ranch would take 15 minutes and require around 5 gallons. (He could not anticipate the long hold before takeoff at Orlando, which consumed another gallon.) So, he would expect to arrive with 5 usable gallons, which, at reduced power, would just about represent the regulation 30-minute VFR reserve.

It appears from this back-of-a-napkin calculation that the pilot’s preflight planning, while not exactly conservative, was not strictly “inadequate” either. There was no reason to think the 182E could not make Massey Ranch with a few gallons to spare. (The engine monitor’s fuel-remaining function could have provided the same answer as the simple calculation, but users are cautioned not to rely on that function because it assumes, possibly without evidence, an accurate setting at the start of the flight.)

The numbers can’t all be correct. The totalizer was probably accurate—they usually are—and so the suspicion has to be that the 182E actually left Oconee with tanks less than full. Had the pilot failed to top them completely? Was a bladder faulty? Had someone drained some fuel from the airplane during the night? What did the gauges say at the start of the flight? There’s no way to answer any of these questions.

I think the NTSB’s finding of “inadequate planning” must remain only a charge, not a conviction. What the report fails to tell us—possibly because the information was not, or could not be, obtained—is whether the pilot took the trouble to visually inspect the fuel tanks in Orlando. Most likely, he did not; it’s inconvenient, and he had good reason to believe that he knew how much fuel was aboard. Maybe his reasons were too good, in fact. Precision fuel-flow measurement and digital totalizing may lull us into a false sense of security, so that we leave our dipstick behind.

In 1997, the singer John Denver took off from Monterey, California, for a couple touch-and-goes and a sightseeing jaunt around Monterey Bay. He had just bought his Rutan-designed Long-EZ and had earlier flown it up from Santa Maria after a checkout flight with the previous owner. The nose-down parking position peculiar to the Long-EZ makes it difficult to judge fuel quantity from the cabin sight gauges, but Denver told a mechanic who assisted him that he was going up for only an hour. While he was flying low over the bay, one tank ran dry, and Denver lost control of the airplane while trying to switch to the other. His accident, like the Orlando one, illustrates the potential hazard of taking off with low fuel even for a short flight. Unless you can measure the fuel aboard precisely, you never know how low your fuel really is.

This story appeared in the October 2020 issue of Flying Magazine.