The Ted Stevens Accident

While the NTSB stated that its investigation couldn't determine precisely what happened in the final few minutes of the flight, it did determine that there was a lack of responsiveness on the part of the pilot before the airplane hit the terrain. It came to that conclusion even though it was clear the airplane had been put into a climbing left turn before impact. Although there were survivors who were interviewed by investigators, none were able to speak to the alertness---or incapacitation---of the pilot in the final moments.

The accident took place on August 9, 2010, at about 2:42 p.m. The single-engine turboprop de Havilland DHC-3T, which had floats, struck mountainous tree-covered terrain about 10 nm northeast of Aleknagik, Alaska. The airline transport pilot and four passengers (including Senator Stevens) received fatal injuries, and four passengers received serious injuries.

Although the airplane's ELT activated, it came out of its mount during the crash, and the connection to the antenna broke. The signal couldn't be received by satellites, which the NTSB said delayed evacuation of the survivors.

The airplane had two pilot seats and nine passenger seats. A bulkhead with an open center entryway separated the cockpit from the cabin. The airplane was operated by GCI Communication Corp. under Part 91.

About the time of the accident, marginal VFR was reported at Dillingham Airport about 18 nm south of the accident site. The passengers had been at a lodge owned by GCI, and they were being taken to a GCI fishing camp about 52 nm southeast for an afternoon of fishing.

The pilot had flown the accident airplane to the lodge earlier in the day. During the return flight to Dillingham, the pilot filed a PIREP with the Dillingham Flight Service Station at about 11:05 a.m. He reported ceilings at 500 feet, visibility of two to three miles in light rain, and "extremely irritating!continuous light chop" turbulence that he described as "kind of that shove-around type stuff rather than just bumps." According to GCI lodge personnel, when the pilot returned to the lodge, he stated that the weather wasn't conducive for a flight to the fishing camp because of turbulence and low ceilings.

Passengers told investigators that by the time that they had lunch about 1 p.m., the weather had improved, and the group discussed with the pilot the option of going to the fishing camp. One passenger characterized the conversation as casual, and stated that no pressure was placed on the pilot to make the flight or to depart by a certain time. The GCI lodge manager and some passengers stated that they thought that the pilot checked the weather on the computer during lunch. By 2 p.m., the pilot was saying he was comfortable taking the group to the fishing camp if the group wanted to go.

The passenger who was in the right cockpit seat told investigators that after takeoff the visibility was "fine." He said he had gone on these trips before, and this time the pilot went in a different direction after takeoff, which the pilot said was to avoid "wind and weather." The passenger described the weather as cloudy above with light turbulence. He fell asleep after about 10 minutes. Another passenger, who was seated in the second seat behind the pilot on the left side of the airplane, said he didn't think that the airplane flew into any clouds. He estimated that he fell asleep about three to four minutes after departure. The passengers who fell asleep had taken an over-the-counter drug to prevent airsickness.

The passenger who was in the third seat behind the pilot on the left side of the airplane stated that the pilot kept the airplane below the cloud ceiling and flew along the tree line, followed streams, and maneuvered to avoid terrain. The passenger stated that the airplane banked into a left turn (he said that the bank angle wasn't unusual) and then immediately impacted terrain. Neither he nor the other passenger who was awake at the time of impact recalled noticing any unusual maneuvering, unusual bank or pitch angles, or change in engine noises that would indicate any problem before the airplane impacted terrain. The wreckage was found at an elevation of about 950 feet MSL in steep, wooded terrain in the Muklung Hills, about 16 nm southeast of the GCI lodge.

The pilot, age 62, was ATP rated, and had been a Boeing 737 pilot with Alaska Airlines before retiring. His FAA first-class airman medical certificate required him to wear "...corrective lenses [and] possess glasses for near [and intermediate] vision." Records indicated he had about 27,868 total flight hours, including about 35 hours in de Havilland DHC-3 airplanes.

The pilot's wife said that he had been very health conscious, and that he hadn't been a coffee drinker, smoker or took any medication.

The pilot had reported to the FAA that he experienced an ICH in the right basal ganglia on March 22, 2006. The records indicated that the pilot was hospitalized, improved quickly, and was discharged six days later.

When the pilot applied for an FAA medical certificate in 2007, he was denied. A May 7, 2007, internal FAA memorandum from the Alaska Regional Flight Surgeon noted: "Airman with history of basal ganglia cerebrovascular accident in March 2006. Not hypertensive and no etiology for the cerebrovascular accident. Has family history of [cerebrovascular accidents] at relatively young ages. Has made good recovery but needs two-year recovery period. Deny for history of cerebrovascular accident."

The Alaska Regional Flight Surgeon's office sent two letters to the pilot dated May 10, 2007. One letter denied the pilot's request for an airman medical certificate and stated that "!our policy is to require an adequate recovery and rehabilitation period. The recommended recovery period is [two] years." The other letter, which also denied the pilot's request for an airman medical certificate, stated that the pilot did not meet the medical standards as prescribed in 14 CFR 67.109(b), 67.209(b) and 67.309(b).

In 2008, after having tests done to submit to the FAA, the pilot applied again. An April 8, 2008, internal FAA memorandum from the Alaska Regional Flight Surgeon noted the following about the application: "Sixty-year-old airman status post-cerebrovascular accident as noted, now two years out from the incident. No recurrence and neurology report entirely normal. OK to issue with warning."

The Alaska Regional Flight Surgeon's office sent the pilot a letter dated April 9, 2008, stating the following: "[Y]ou are eligible for a first-class medical certificate.! Because of your cerebrovascular accident, operation of aircraft is prohibited at any time new symptoms or adverse changes occur."

All of the surviving passengers had flown with the accident pilot on previous flights, and none reported anything different about his behavior on the day of the accident. The passenger who was in the third seat behind the pilot on the left side of the airplane stated that the weather conditions during the accident flight weren't remarkable and didn't seem "risky" based on his previous experience with general aviation flights in Alaska. He described the conditions as "characteristic" Alaska flying. He stated that all of the flights were made below the cloud ceilings and that the airplane entered the clouds for a few seconds on rare occasions. He described the pilot as prudent, cautious, very quiet and businesslike but not humorless, and he stated that he got the impression that the pilot had "flown the routes so many times that he had memorized them."

The passenger who was in the right cockpit seat during the accident flight stated that the pilot seemed alert, and that he didn't see the pilot do any paperwork during the flight before the passenger fell asleep.

The airplane had GPS equipment including a terrain warning system, which would have generated audio and on-screen text alerts. Investigators talked with a friend of the pilot, who said when they flew they generally turned off the audio warnings. The airplane also was equipped with a radar altimeter system that would have issued aural and visual altitude alerts about four to six seconds before impact.

The NTSB asked for help from the Armed Forces Institute of Pathology and the Mayo Clinic in reviewing medical records and autopsy reports for evidence that the pilot suffered another stroke or some other incapacitating event. No such evidence could be found.

The NTSB determined that the probable cause of this accident was the pilot's temporary unresponsiveness for reasons that couldn't be established from the available information. Contributing to the investigation's inability to determine exactly what occurred in the final minutes of the flight was the lack of a cockpit recorder system with the ability to capture audio, images and parametric data.

Peter Katz is editor and publisher of NTSB Reporter, an independent monthly update on aircraft accident investigations and other news concerning the National Transportation Safety Board. To subscribe, write to: NTSB Reporter, Subscription Dept., P.O. Box 831, White Plains, NY 10602-0831.