About 120 nautical miles separate Winchester and Lynchburg, Virginia. But even a nearby destination may prove strangely difficult to reach.
The pilot-owner of a 1964 Cessna 210D, a 50-year-old physician with about 300 hours, left Winchester for Lynchburg, alone, a few minutes before 8 o’clock on an April evening. He had made the same trip, in both directions, many times. Several of those had been in the past few weeks, and, with a newly acquired instrument rating, he had filed IFR on each of them. One of those trips was at night, but the eight-tenths of an hour of solo actual instruments that he had logged had been in daylight. He had flown in actual instrument conditions, with an instructor, during 4.6 hours of his training; the rest was simulated. This was his first solo night IFR flight in weather. Its every detail would be recorded and preserved, second by second, by an electronic engine and systems monitor and data logger.
Lynchburg was reporting 600 overcast and three miles in mist, with calm winds. The flight began uneventfully, but after 15 minutes the Potomac Tracon controller noticed that the 210’s heading was 270, whereas it should have been 210. He called the discrepancy to the pilot’s attention. The pilot, then level at 8,000, said that he was “just a little disoriented here” and proceeded to correct his heading.
Tracon then offered vectors around weather, which the pilot accepted, and for the next quarter-hour he weaved, with the controller’s guidance, among areas of moderate rain.
He appears to have been hand-flying the airplane, whether because its autopilot was not working or because he thought it desirable to gain experience hand-flying in IMC. His altitude control was generally accurate, but his heading wandered continually. About 27 minutes into the flight, the 210 turned around northward in the space of a minute and lost 1,000 feet of altitude. The doctor, a specialist in diabetes, was himself an insulin-dependent diabetic, and one of the conditions of his medical certification was that he check his glucose level at one-hour intervals while flying and within 30 minutes before landing. Perhaps the brief excursion from controlled flight occurred during such a check.
About an hour after the start of the flight, he was cleared down to 3,000 with vectors for the Runway 4 ILS at Lynchburg. He missed the approach, reporting to Lynchburg Tower that he had problems with his “engine monitor.” Indeed, the manifold pressure and percent-of-power traces on the engine log drop to near zero before the start of the approach and remain there, implausibly, even during the climb portion of the miss while several other indicators show the expected increase in power. But it is unclear why those obviously spurious indications would affect an ILS approach.
Back with approach control and making right turns to regain the ILS, the pilot neglected his altitude, which decayed from 3,000 to 2,200 feet. The controller issued a low-altitude warning, and the pilot replied, “That’s affirmative. I’m just trying to get the VORs in here.” This was an uncharacteristically muddled transmission; his communications and clearance read-backs had been, to this point, generally precise and well worded.
The controller, who by now was well aware that he had a pilot who was, at least potentially, in trouble, offered a GPS approach, but the pilot replied that he would prefer another ILS. Again, there was a problem: The pilot reported himself to be on the localizer when he was in fact well to the right of it. The controller instructed him to climb immediately to 3,000. On both ILS approaches, the logged GPS altitude trace drops to near 1,000 feet, just a couple of hundred feet above the sparsely populated terrain south of the airport.
Now the pilot accepted the suggested GPS approach, and the controller vectored him to Kilbe Intersection, 14 miles out on the ILS, to give him plenty of time to get it set up. The controller gave him a block altitude above 2,800 and vectored an arriving Piedmont 737 into a holding pattern. “I’m going to need to make a 360 right here,” the pilot said. “I’m still having some trouble with my engine monitors.”
In the course of this turn his altitude again decayed, and the controller again issued an alert. Shortly after this, the pilot said, “I’ve got a complete gyro failure; it looks like I need some help with both heading and altitude monitoring.” The controller began to provide no-gyro vectors. Asked if he was tracking the Lynchburg VOR, the pilot replied, “I’m a little dizzy, thanks.” Continuing to climb past 3,000 feet, he reported his altitude a few minutes later as “5,200 elevation,” which agreed with the controller’s Mode C readout.
Beginning with the 360-degree turn, which took place in the vicinity of Kilbe, the 210 made several random circles, both right and left. It climbed to above 6,000 feet. Was the pilot hoping to get into the clear on top? The controller, in the meantime, was coordinating with Washington Center, trying to find an airport with VFR weather. There was none within range. “The pilot is very disoriented,” he said. “He reported some equipment problems … but I really do feel like this pilot is disoriented more than anything else. … ”
Responding to an offer of a no-gyro surveillance approach, the pilot said, “Roger that. I’m going to turn right to 180, start tracking outbound, see if I can’t get back in line with the approach and see if you can give me a no-gyro approach as well.”
That was his last transmission. Less than a minute later, after the 210 appeared to have stabilized on an easterly heading, radar suddenly showed it in a tight, rapidly descending right turn. A few witnesses on the ground saw it emerge from the fog and mist, very low, moving fast, in a nearly vertical bank, before disappearing among trees.
The NTSB attributed the accident to “the pilot’s loss of airplane control due to spatial disorientation [and his] lack of experience in actual night instrument conditions.”
This accident must have presented investigators with a confusing mix of clues. It is natural to wonder whether the cascade of instrument malfunctions reported by the pilot was real. Some power-related items from the systems monitor are obviously incorrect, but in every other respect the recorded engine parameters appear normal. Vacuum is steady at five inches. The gyros were heavily damaged in the crash but on disassembly showed no evidence of previous problems, and the fact that the pilot had kept the airplane upright through a long series of meandering turns suggests that the attitude indicator was working. “Physical evidence of an in-flight instrument failure,” the board said, “was inconclusive.”
Low blood sugar can produce confusion, but it was impossible to know whether or not the pilot’s glucose level had affected his performance. Nondiabetic pilots too have been known to experience this kind of disorientation. Investigators found an empty “energy bar” wrapper in the wreckage but not a blood glucose measuring device. The pilot’s mention of disorientation and dizziness might suggest vertigo just as easily as hypoglycemia; it is not clear whether it was because of the pilot’s statements or for other reasons that the Roanoke approach controller — whose handling of the situation was exemplary — felt that the problem was disorientation rather than actual hardware failures.
New instrument pilots are advised to advance gradually into weather flying. Real flight, with its noise, vibration, turbulence, visceral sensations and, for a novice at least, nervous anxiety about actual rather than make-believe dangers, belies the simulator’s implication that instrument flying is purely a matter of following procedures with reasonable accuracy. This low-time pilot was perhaps taking on a lot in a maiden night-IFR flight that would end with an approach in below-VFR conditions. If his problems had begun only after his first missed approach, one would suspect that some combination of vertigo and nerves might have been the cause. But his first use of the word disoriented — we don’t know if he meant it literally — came early in the uneventful cruise segment of the flight; and his first approach, which should have been easy, ended in a miss attributed, bafflingly, to the “engine monitor.”
Something went wrong. What was it? We, and the NTSB, can only guess.
This article is based on the NTSB’s report of the accident and is intended to bring the issues raised to our readers’ attention. 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.
