The 500-hour pilot held a private license for aircraft single-engine land. He did not have an instrument rating, but had been receiving instruction toward one. The airplane was a J35 Bonanza, manufactured in 1958 and equipped with tip tanks that increased its fuel capacity to 100 gallons.
The pilot was solo when he left New Smyrna Beach, Florida, on a dark February evening for New Orleans, about 480 nm distant. This was a flight that he had practiced on his personal computer, according to his flight instructor, using Microsoft Flight Simulator 2000. The instructor also reported a quirk of the airplane: On one flight, when the right tip tank was selected, the engine began to cough and lose power. Returning to the main tank restored smooth running. The instructor said that the pilot had had a mechanic “lubricate the fuel selector valve assembly,” after which “the fuel system operated normally.” No maintenance log entry was found to document this somewhat peculiar episode, and it is unclear what effect lubricating a selector valve would have on what sounds more like a vent obstruction or a clogged or leaking feed line.
Once airborne, the pilot contacted Daytona Beach TRACON to request an IFR clearance to New Orleans. Having obtained it, he proceeded uneventfully en route at an altitude of 6,000 feet and with a groundspeed of about 150 knots.
At 8:58, when the flight had been airborne for about an hour and 40 minutes, the Center Controller working the Bonanza noticed that the airplane had reversed course and was now eastbound. He inquired whether there was some difficulty, and the pilot replied that he had a fuel problem and was working on it. When the controller told the pilot that if he needed to land the Panama City airport was less than 10 miles away, he replied that he believed he could “make it” to Gulfport-Biloxi, 160 nm to the west and just 60 nm short of New Orleans.
A couple of minutes later, the pilot reported to the controller that he had fixed the problem and wished to continue to New Orleans, his original destination.
At 9:28, after being handed off to Pensacola Tracon, the pilot checked in and added, “Can I amend it to you for fuel?” Upon being asked what he wanted to amend, the pilot stammered, “I wanna … I would like to amend my flight paa … my flight plan to come to you for fuel … you have ah, avgas there?”
“Yes sir, you wanna land at Pensacola Regional Airport, is that correct?”
“I’m, uh, I’m having a fuel situation up here,” the pilot said. “I need to get this thing taken care of.”
Garbled, hesitant speech and irrational questions (does an airport have avgas?) are pretty clear signs of a pilot who is rattled. The controller did not wait for the pilot to declare an emergency, but gave the Bonanza priority handling for an ILS approach into Pensacola where, unfortunately, there was a 200-foot overcast, a challenging condition only slightly mitigated, given the darkness, by two-mile visibility. The Bonanza, however, was then above the clouds and in visual conditions.
The controller was vectoring the Bonanza to the Runway 17 ILS when he asked the pilot, “Were you just burning more fuel than you’d estimated?”
“I have fuel in a tip tank that I cannot get to,” the pilot explained.
The controller cleared the Bonanza to descend to 1,700 feet. When it was about eight miles north of the airport he said, “Assigned altitude is 1,700, you’re on a base leg, eight miles north of the airport,” to which the pilot responded, “Roger, 89D 1,700, base leg is north.” The Bonanza, however, was at 2,700 feet. “I’m showing you at 2,700,” the controller said. “You want a vector across the final for descent, or can you get down from there?”
“I’m, uh, still working ‘er down,” the pilot said. “I think uh … yes, go ahead and give me a vector for descent.” A little later, the pilot asked the controller for the ILS frequency, which he provided.
At 9:51, the controller said, “Bonanza 89D, you appear to be off course, sir. You’re right at Brent [intersection, the final approach fix, 4.4 miles from the runway]. You want me to vector you back around and try to resequence you again?”
“That’s affirmative,” the pilot said. “I’m outta gas.”
“You say you’re out of gas?”
“Just about.”
“All right, you want to continue the approach?”
“Affirmative.”
The controller told the pilot to turn left in order to intercept the localizer. When the pilot reported, “I’ve got the needle,” however, the controller said, “89D, you look like you’re northbound sir, you’re going away from the airport, you need to turn around, you need to go back to the south, sir.” The pilot said: “Roger that, 89D turning left 130.”
The controller vectored the pilot back around to a southerly heading. “I’m descending quickly, I’m having a real hard time up here with fuel,” the pilot said. Then there was a low altitude alert: The Bonanza had descended below the minimum vectoring altitude. The controller instructed the pilot to climb, only to see the radar target curve back to a northeasterly heading. “Can you turn right to 130 degrees?” the controller asked. “Okay,” the pilot said, “I’m trying to turn right to 130, I’m just having a helluva time up here.”
The pilot succeeded in establishing the Bonanza on the localizer, and the controller cleared him to land. He was two miles from the runway.
But then – “[It] looks like you’re drifting off … off course, to the west, over.”
“… Affirmative, I trying to …. “
It was too late. “Bonanza 89D, climb immediately and maintain 1,700, climb and maintain 1,700 immediately, over … Bonanza 89D, how do you hear me?”
“Affirmative, I’m climbing.”
“Roger, turn right heading 330, turn to 330, let’s try it again.”
A second attempt was also unsuccessful and ended in another missed approach.
Around 10:10 the controller tried for a third time to set the pilot up for the ILS.
“Bonanza 89D, turn right heading 040.”
“040, 89D … I’m outta gas … we’re on fumes here.”
“Bonanza 89D roger, turn right heading 090.”
“090, 89D.”
“Bonanza 89D say altitude.”
“… two … “
“Bonanza 89D, maintain 1,700, over.”
“I can’t, I can’t … I’ve lost it … “
Soon after, the airplane disappeared from the radar screen.
Local residents heard a crash, and at three in the morning searchers located the wreckage in swampy, wooded terrain about 14 miles northwest of the airport. NTSB and FAA investigators visited it the next day, but the airplane was too badly fragmented to permit inspection of the fuel system, other than to note that all six of its fuel tanks were breached.
It is often said that accidents occur as a result of unforeseen linkages of events and decisions that were innocuous in themselves. This was not that kind of accident. It was the result of a series of bad decisions, beginning, well before the final flight, with an incurious response to the problem of lack of fuel flow from the right tip tank.
Only partway through instrument instruction, the pilot then chose to file an IFR flight plan for a night flight. Uncharacteristically, the National Transportation Safety Board’s report on the flight omits mention of a preflight weather briefing, if the pilot even obtained one, and of the forecast weather en route and at the destination. There is also no discussion of fueling records, and so the amount of fuel aboard at takeoff is unknown.
The NTSB report makes no mention of fuel odor or of the staining of foliage that occurs when fuel spills from breached tanks; and so it is impossible to know whether there was still fuel in the airplane when it crashed. At the very least, one would expect from the apparent chain of events that the right tip tank would have been full. The pilot’s difficulty holding a heading is consistent with the very considerable imbalance that would be created by fuel trapped in a tip tank while all other tanks were empty.
In airplanes with auxiliary tanks, it’s good practice to use fuel from the auxiliaries before the main fuel. In fuel-injected airplanes, however, including the Bonanza, more fuel is pumped from the supplying tank than the engine needs, and unused fuel returns to one of the main tanks. It’s necessary, therefore, to burn some fuel out of that tank before starting to empty the auxiliaries.
It seems, however, that the pilot did not try to feed fuel from the right tip tank until an hour and 40 minutes into the flight. If, at that point, he felt that he could make Gulfport-Biloxi but not New Orleans, he must have believed that he had sufficient fuel available for about two hours of flight-probably around 30 gallons, with the required reserve. But apparently he had less than that, because he ran out only 75 minutes later.
At first glance it appears that the half-trained pilot was wildly overconfident to undertake a three-hour night IFR flight alone. But it’s possible that he actually expected to make the entire flight in visual conditions, and that he filed an IFR flight plan just to gain experience with the system. The flight would, so to speak, have been a simulation, like the Flight Simulator trips he had made along the same route, but with real live controllers in the role of the desktop computer. To the extent that vectors might have threatened to send him into IMC, he could always have cancelled IFR and continued VFR.
At some point he must have become aware of the overcast blanketing the coast beneath him. The weather was better inland; 40 miles northeast of Pensacola, there were breaks in the overcast. It may have been better still farther north; the accident report unfortunately does not discuss the larger weather picture. If the forecast at New Orleans was good, however, he would not have been the first pilot who flew VFR over a stretch of low-level IMC in order to reach a VMC destination.
The accident report does not say whether the pilot inquired about weather ahead when he decided to pass up Panama City and make for Gulfport-Biloxi; nor is there any mention of his having inquired of the Pensacola controller whether there was a non-IMC airport anywhere within, say, 100 miles, a distance that he could have traveled, at economy cruise, with the fuel that he burned in 45 minutes of unsuccessful ILS approaches.
The sad fact is that the pilot had gotten in over his head some time earlier. His confusion is evident throughout his communications with the Pensacola controller. Furthermore, an ILS approach to 200-foot minimums is a difficult task even for a well-practiced instrument pilot. The controller must have known that the pilot was very unlikely to land safely; but what could he do but keep trying to bring him in?
Flight simulation software has probably convinced quite a few pilots that they were ready for the big leagues long before they really were. What makes the real thing more difficult than the simulation is fear. When you’re alone in an airplane in the dark and running out of fuel and the fog below you is hovering just at airline minimums, you know that if you crash you won’t be getting up and strolling to the refrigerator. Flight Simulator is said to be marvelously realistic; what would make it more realistic still would be to play it while a slightly crazed stranger held a gun to your head.
This article is based solely on the National Transportation Safety Board’s reports of the accidents and is intended to bring the issues raised to the attention of our readers. It is not intended to judge or to reach any definitive conclusions about the ability or capacity of any person, living or dead, or any aircraft or accessory.
