An Aeromedical Mystery Solved
By: Clint Johnson, Chief, Alaska Region, Office of Aviation Safety
This is the fifth blog in a new series of posts about the NTSB’s general aviation investigative process. This series, written by NTSB staff, explores how medical, mechanical, and general safety issues are examined in our investigations.
After nearly 20 years of investigating hundreds of aviation accidents, I recently encountered an invisible killer.
I was enjoying a late summer Saturday afternoon with my wife in Anchorage, Alaska, when my phone rang. My wife – a 20-year-veteran NTSB spouse – knew from the look on my face that our quiet weekend at home had just ended.
An Anchorage Fire Department dispatcher was calling. She reported that rescue crews were on the scene of a fatal airplane crash in a residential neighborhood only 20 minutes away.
When I arrived, I was briefed by a small army of Anchorage Police and Fire Department
crews. Behind the wall of fire trucks, police cars, stunned residents, and TV cameras, I caught a glimpse of the inverted and burned remains of what looked like a float-equipped Piper 11 in the middle of the residential roadway.
We continued to talk as we walked toward the wreckage site. The pungent smell of burned aircraft wreckage filled the air as we proceeded past the yellow police tape. Finally, I was close enough to see that only the welded steel-tube structure and engine remained, with the fuselage and wings barely recognizable. The postcrash fire had incinerated much of the wreckage.
Witnesses had told the police that just before the accident they watched in amazement as the airplane completed two, low-level, high-speed, 360° right turns over the neighborhood – the first 150-200 feet above ground level, and the second much lower. One homeowner stated that the airplane passed over his home about 50 feet above his roof.
Witnesses also reported that the airplane’s bank angle increased significantly on the second 360° right turn; one pilot-rated witness estimated the bank at more than 60°. Witnesses also reported hearing the airplane’s engine operating in a manner consistent with high power settings throughout both 360° turns.
One man was mowing his lawn as the airplane completed the second, steep, 360° right turn. He said that the airplane flew directly over his yard, then the nose of the airplane pitched down and it began to descend rapidly. The engine rpm then increased significantly, and the wings rolled level just before the airplane impacted a stand of tall trees adjacent to his home, severing its floats.
It crashed on a neighborhood road, coming to rest inverted. About 30 seconds after impact, a fire ensued, which engulfed the entire airplane before any of the witnesses made it to the wreckage.
Sadly, after the fire department crews extinguished the fire, they found the remains of the 75‑year-old pilot and his dog still inside the incinerated wreckage.
While we all waited for the medical examiner to arrive, I began interviewing witnesses. Most concluded, or were well on their way to concluding, that the pilot was “just showing off” to someone on the ground. But the NTSB sets a high bar for conclusions. It was way too early for me to go there.
At the scene, I met a family member, along with a close friend of the pilot. Understandably upset, both reported that it was highly unusual and uncharacteristic behavior for the pilot to be flying as the witnesses consistently described to me. They went on to say that to their knowledge, the pilot didn’t know anyone in the area, but that, given the pilot’s anticipated flight route, he would have been flying over the neighborhood while on the return flight home.
Then, as the pair was preparing to leave the scene, the pilot’s friend said something in passing – something about his longtime buddy’s history of cardiac problems, which, in his opinion, caused the pilot’s erratic flight maneuvers.
I pressed him for more information, but it became clear that he wasn’t prepared to provide any additional information on the subject then and there, and I decided that this was neither the time or place to discuss it. As the pair got back into their car and slowly drove away, I knew that the following Monday morning I’d likely be attending the pilot’s autopsy.
For now, I needed to document and examine the wreckage before it was removed. This included determining control cable continuity to the flight control system, engine control continuity, and more.
The engine sustained significate impact damage, but only minimal fire damage. There were no mechanical problems that I could find on-scene that would explain what the witnesses reported. However, a much more detailed wreckage exam would be accomplished later, once the wreckage was moved to a more secure and suitable site.
On Monday morning, I found myself at the State medical examiner’s facility, meeting with the pathologist who would be working my case. I explained to her what I was looking for, and she started the exam.
The entire autopsy took over two hours to complete, and the pathologist found no conclusive evidence for medical incapacitation from an acute cardiac event. However, per standard protocol, the autopsy team took blood and tissue samples to send to the FAA’s Bioaeronautical Sciences Research Laboratory in Oklahoma City for a toxicological exam.
I knew I would not have the tox report for two to three months, but the autopsy yielded at least one more piece of valuable information: the pilot died from trauma, not the postcrash fire. Unbeknownst to me at the time, this would be an extremely important data point that would help solve the case in the end.
Over the next two weeks, I visited the wreckage two separate times at a local aircraft salvage yard. I looked for evidence that would support various theories, but nothing ever panned out. It was one dead end after another.
Then on a cold and snowy autumn afternoon, the FAA’s tox report on the pilot appeared in my e-mail. I opened it and scanned the results, and only then realized just what I had been missing all this time: Carbon Monoxide, an odorless, colorless and tasteless gas – and a silent killer of general aviation pilots.
The pilot’s carboxyhemoglobin (carbon monoxide) level was an extremely high 48%. To put these results in context, nonsmokers may normally have up to 3% carboxyhemoglobin in their blood, and heavy smokers may have levels of 10% to 15%. And according to family members, this pilot did not even smoke.
Since the pilot died of blunt-force trauma prior to the ensuing fire, it was not possible that this CO level was an effect of the fire. But it was possible that it was a cause of the crash.
I realized that over the last few months I had missed an important and somewhat elementary piece of evidence, the airplane’s exhaust system. I quickly reviewed my on-scene photos, and I could clearly see that the entire exhaust system sustained relatively minor damage in the accident.
Within 15 minutes of receiving the toxicology results, I was on my way back to the stored wreckage. I ended up bringing the entire exhaust system back to the office, muffler, heat exchanger/muff and all. Like the autopsy examiners I had met months earlier, I went to work on this simpler machinery, peeling back the heater shroud.
Inside I found a severely degraded muffler with portions missing, which allowed raw exhaust gases to enter the main cabin through the airplane’s heater system.
Unfortunately, neither the family or any of the pilot’s friends could find any maintenance logbooks for the accident airplane, so I was unable to determine just when the last muffler inspection was done (if ever). However, after talking with several friends of the deceased pilot, many said that he did his own maintenance, and he was not an aviation mechanic.
They went on to say that the pilot, with the help of a few other friends, installed the more powerful Lycoming O-320 engine about 5 years earlier, but none could provide any additional information about how the pilot maintained his airplane.
However, I could report directly to the family what circumstances led up to the death of their loved one, and I was able to show them the physical evidence that I found.
The NTSB’s probable cause summed it all up: “The pilot’s severe impairment from carbon monoxide poisoning in flight, which resulted in a loss of control, and a subsequent inflight collision with trees and terrain.”
Often, it takes time, patience, and knowledge of the human operator, the machine, and the environment to solve an accident mystery to provide answers.
NTSB Safety Compass Blog 
Great article, Clint! Excellent depiction of the process where one gradually and methodically peels the layers of the accident puzzle until you get to the probable cause. Bravo!