Pilot disorientation

November 1, 2000

Many case studies have a clear origin and outcome, but the events are puzzling and disturbing in this accident. There was significant ground damage and injury, which created media interest and local political pressure. On average, there are only four serious or fatal injuries to off-airport bystanders per year, but the nature of this incident leaves us with questions and some political problems.

On November 26, 1999, at 10:49 a.m. Eastern Standard Time, the 56-year-old physician pilot, his wife, and daughter departed from Linden, New Jersey, on an IFR flight destined for Washington Dulles International. The weather observation for nearby Newark International Airport was 2.5 miles' visibility, with light rain and mist. Scattered clouds were reported at 600 feet, with a 1,300-foot broken ceiling. The temperature was 61 degrees Fahrenheit with a dew point of 59 degrees. It looked like a routine instrument flight in the V-tail Beech S35 Bonanza.

After departing Linden, the pilot contacted New York Departure Control and was instructed to turn left to a heading of 010 degrees and climb to 5,000 feet. A few seconds later, the controller revised the clearance and instructed him to maintain 2,000 feet. Thirty-four seconds after that the controller asked for a left turn to a heading of 270 degrees, to which the pilot did not reply. The controller reissued the heading, but there was still no response.

The controller made two more attempts to reestablish communications. After the second attempt the pilot responded, "I have a problem." The controller inquired about the problem and the pilot responded, "I had a gyro problem momentarily. It looks straightening now. I must have had water in the system." Twenty seconds after that the controller radioed, "Continue the right turn all the way around...correction, you're in a left turn now." The pilot responded, "Yes, sir...left turn climbing to niner thousand."

The controller corrected, "Stop your climb at two thousand, turn left, left turn heading two-seven-zero," and then asked if the pilot was OK to navigate. The pilot responded, "I think I have a problem," and then requested a climb. The controller instructed the flight to maintain 2,000 feet and requested the current heading. The pilot responded, "Looks like zero-three-zero." The controller repeated the instruction to turn left to 270 degrees, but there was no response. The controller repeated the altitude and heading. Still there was no response. The controller then radioed, "Niner-Two-Mike, I need to be acknowledged please." The pilot replied, "I have a problem."

The Bonanza slammed into the southwest corner of the roof of an abandoned three-story brick building at about 10:53 a.m. Approximately 50 percent of the roof and 50 percent of the third floor were consumed in the post-crash fire. The debris path continued north along the left side of a residential street, crossing a road, a parking lot, and then another road before ending some 760 feet from the initial point of impact. There were no survivors on the aircraft. Two people on the ground received serious injuries (one later died), and 25 received minor injuries. Eighteen build-ings received varying degrees of damage rang-ing from broken windows to structural damage. Three of the buildings were condemn-ed and demolished. The city of Newark estimated the property damage at $1.15 million. At least eight automobiles were damaged or destroyed.

Radar data showed the Bonanza heading east at approximately 900 feet when the pilot first reported a problem. Over the next two minutes, the ground track changed from east to north, to north-east, to northwest, and then back to north. In the last 30 seconds, the target attained a maximum altitude of 2,800 feet and airspeed of 161 kt before beginning a descent that reached 10,000 feet per minute.

The pilot held an airline transport pilot certificate with an airplane single-engine land rating and a commercial certificate for airplane multiengine land, single-engine sea, and glider. An active instrument flight instructor, he was known to participate in type-club training programs to instruct other pilots. His second class medical certificate had been renewed on November 1, 1999. On the medical application, the pilot reported 5,800 hours of total flight experience, with 120 hours in the past six months. His logbook showed 1,308 hours of actual instrument flight. He passed a flight review in a Piper PA-28 and a flight check for his single-engine ATP certificate eight months earlier. The pilot was, by reputation and experience, well-qualified and conscientious.

Examination of the flight instruments and vacuum pumps yielded some interesting findings. The engine was operating at the time of impact, and while both the primary and standby vacuum pumps had separated from their engine mounts, neither unit revealed preimpact failure. Continuity of the vacuum system plumbing could not be verified because of impact damage. It appears that the primary vacuum pump was operating. The Bonanza was equipped with a backup vacuum system, and when the standby clutch assembly was examined, the engine side of the clutch displayed rotational scoring consistent with a momentary impact, but the vacuum pump side did not, indicating that it probably was not activated. Electrical and mech-anical continuity for the clutch assembly was verified.

Three gyros and two gyro cases were also inspected. If there is scoring on the case or on the gyro itself, that usually indicates the gyro was spinning at the time of impact and then contacted the case during the crash sequence, leaving a mark. If there are no score marks, that generally indicates the gyro was not operational at impact. The vacuum-driven gyro from the attitude indicator displayed rotational scoring. No rotational scoring was observed on the horizontal situation indicator's (HSI) vacuum-driven gyro or on the turn coordinator's electric-driven gyro. Static marks consistent with no or little rotation were observed on the HSI gyro housing. The turn coordinator gyro housing was not recovered.

A plausible interpretation is that the attitude indicator was working, but the heading indicator and the turn coordinator were not. This is not the typical IFR training scenario. Usually, it is assumed that the vacuum pump has failed, leading attitude and heading indicators to gradually fail. The electrically driven turn coordinator, or turn needle, is considered the most reliable, and the pilot is expected to transition to that as soon as a fault is recognized to maintain lateral control. This is a typical instrument power configuration, but some aircraft may have electrically driven heading or attitude gyros and, although rare, may have a vacuum turn needle. The gyro flight instruments are required to have a split source of power to assure some redundancy. In the real world, the failure is subtle as the gyros spin down slowly, and it can be quite confusing to determine which instrument or instruments are inoperative.

It is possible but extremely unlikely for two independently powered systems to fail simultaneously. It is pure speculation on my part that possibly the turn coordinator may have been inoperative before the flight. Regardless, the pilot was presented with a very difficult task early in the flight as he was transitioning to the IMC environment. The wandering heading and nonresponse to ATC indicate that he was mentally saturated with sorting out the problem.

At this point, the probable accident cause would have been evident — spatial disorientation because of instrument failure. However, the pilot's autopsy found that he had consumed large doses of a prescription barbiturate drug. According to private medical records during a medical examination in 1976, it was noted that the pilot suffered from migraine headaches. He was using Fiorinal, which contains the barbiturate Butalbital, aspirin, and caffeine, to control the pain.

In 1989, the pilot called a nurse complaining of "severe migraine" headaches and stated that Tylenol or aspirin provided no relief. He "refused to come in for an evaluation and just wanted a refill that would last until morning when he could call a physician." At that time, he was taking one Fiorinal every four to six hours. Pharmacy records show that the pilot was provided with more than 6,000 tablets of Fiorinal or the generic equivalent from 1992 to October 1999. According to The 1999 Physician's Desk Reference, "Fiorinal...is indicated for the relief of the symptom complex of tension (or muscle contraction) headache. Caution is required because Butalbital is habit-forming and potentially abusable.... The most frequent adverse reactions are drowsiness and dizziness."

On the pilot's last FAA medical application, he stated he was not taking any prescription or nonprescription medication, and that he had never suffered from severe or frequent headaches. All prior FAA medical applications contained the same statements. According to the FAA Guide for Aviation Medical Examiners, a history or presence of any of the following conditions would preclude the issuance of a medical certificate: migraine headaches, migraine equivalent, cluster headaches, chronic tension headaches, or conversion headaches. In addition, the publication stated that "pain, in some conditions, may be acutely incapacitating. Chronic recurring headaches or pain syndromes often require medications for relief or prophylaxis, and in most instances, the use of such medications is disqualifying because they may interfere with a pilot's alertness and functioning. The examiner may issue a medical certificate to an applicant with a long-standing history of headaches if they are mild, seldom requiring more than simple analgesics, occur infrequently, and are not incapacitating, and are not associated with neurological stigmata."

Would the pilot have been capable of handling the emergency if he had not taken the drug? That is unknown. As a medical doctor he would have easier access to prescription medicines than the average pilot. It could be debated that the drugs had little or no effect on his abilities since he had been taking them for so long that his system had adapted. After all, he had flown for 23 years without incident.

It is undeniable that lying on the medical application and continuing to fly with a potentially debilitating condition — using prohibited medication — are wrong. It puts a blot on an otherwise exemplary personal and flying career. The accident might well have happened without any medical transgression, and that must be kept in mind. But that negligence adds to the sensationalism. Media-attracting incidents like this provide strong ammunition for those who would further restrict general aviation.

See also the index of "Safety Pilot" articles, organized by subject. Bruce Landsberg is executive director of the AOPA Air Safety Foundation.