Hageland Aviation Services

On October 2, 2016, about 1157 Alaska daylight time, Ravn Connect flight 3153, a turbine powered Cessna 208B Grand Caravan airplane, N208SD, collided with steep, mountainous terrain about 10 nautical miles northwest of Togiak Airport (PATG), Togiak, Alaska. The two commercial pilots and the passenger were killed, and the airplane was destroyed. The scheduled commuter flight was operated under visual flight rules by Hageland Aviation Services, Inc., Anchorage, Alaska, under the provisions of Title 14 Code of Federal Regulations Part 135. Visual meteorological conditions prevailed at PATG (which had the closest weather observing station to the accident site), but a second company flight crew (whose flight departed about 2 minutes after the accident airplane and initially followed a similar route) reported that they observed unexpected fog, changing clouds, and the potential for rain along the accident route. Company flight-following procedures were in effect. The flight departed Quinhagak Airport, Quinhagak, Alaska, about 1133 and was en route to PATG.

The Cessna had departed about 3 minutes prior on a scheduled passenger flight and the Piper was en route to a remote hunting camp when the two airplanes collided at an altitude about 1,760 ft mean sea level over a remote area in day, visual meteorological conditions. The airline transport pilot and two passengers onboard the Cessna and the commercial pilot and the passenger onboard the Piper were fatally injured; both airplanes were destroyed. Post accident examination revealed signatures consistent with the Cessna's outboard left wing initially impacting the Piper's right wing forward strut while in level cruise flight. Examination revealed no mechanical malfunctions or anomalies that would have precluded normal operation of either airplane. Neither pilot was in communication with an air traffic control facility and they were not required to be. A performance and visibility study indicated that each airplane would have remained a relatively small, slow-moving object in the other pilot's window (their fuselages spanning less than 0.5° of the field of view, equivalent to the diameter of a penny viewed from about 7 ft away) until about 10 seconds before the collision, at which time it would have appeared to grow in size suddenly (the "blossom" effect). From about 2 minutes before the collision, neither airplane would have been obscured from the other airplane pilot's (nominal) field of view by cockpit structure, although the Cessna would have appeared close to the bottom of the Piper's right wing and near the forward edge of its forward wing strut. The Cessna was Automatic Dependent Surveillance-Broadcast (ADS-B) Out equipped; the Piper was not ADS-B equipped, and neither airplane was equipped with any cockpit display of traffic information (CDTI). CDTI data would have presented visual information regarding the potential conflict to both pilots beginning about 2 minutes 39 seconds and auditory information beginning about 39 seconds before the collision, providing adequate time for the pilots to react. The see-and-avoid concept requires a pilot to look through the cockpit windows, identify other aircraft, decide if any aircraft are collision threats, and, if necessary, take the appropriate action to avert a collision. There are inherent limitations of this concept, including limitations of the human visual and information processing systems, pilot tasks that compete with the requirement to scan for traffic, the limited field of view from the cockpit, and environmental factors that could diminish the visibility of other aircraft. Given the remote area in which the airplanes were operating, it is likely that the pilots had relaxed their vigilance in looking for traffic. The circumstances of this accident underscore the difficultly in seeing airborne traffic by pilots; the foundation of the "see and avoid" concept in VMC, even when the cockpit visibility offers opportunities to do so, and particularly when the pilots have no warning of traffic in the vicinity. Due to the level of trauma sustained to the Cessna pilot, the autopsy was inconclusive for the presence of natural disease. It was undetermined if natural disease could have presented a significant hazard to flight safety.

The pilot departed on a scheduled commuter flight at night from an unlit, rough and uneven snow-covered runway with five passengers and baggage. During the takeoff roll, the airplane bounced twice and became airborne, but it failed to climb. As the airplane neared the departure end of the runway, it began to veer to the left, and the pilot applied full right aileron, but the airplane continued to the left as it passed over the runway threshold. The airplane subsequently settled into an area of snow and tundra-covered terrain about 100 yards south of the runway threshold and nosed over. Official sunset on the day of the accident was 48 minutes before the accident, and the end of civil twilight was one minute before the accident. The Federal Aviation Administration's (FAA) Airport/Facility Directory, Alaska Supplement listing for the airport, includes the following notation: "Airport Remarks - Unattended. Night operations prohibited, except rotary wing aircraft. Runway condition not monitored, recommend visual inspection prior to using. Safety areas eroded and soft. Windsock unreliable." A postaccident examination of the airplane and engine revealed no mechanical anomalies that would have precluded normal operation. Given the lack of mechanical deficiencies with the airplane's engine or flight controls, it is likely the pilot failed to maintain control during the takeoff roll and initial climb after takeoff.

The certificated airline transport pilot, with one non-revenue passenger, departed in the twin engine turboprop airplane from a rural airport on a CFR Part 135, VFR cargo flight over ocean waters. The flight did not reach its destination, and was reported overdue. Search personnel searched along the airplane's anticipated route of flight, over ocean waters. Shortly after initiation of the search, airborne search personnel reported sighting floating debris, consisting of what appeared to be aircraft seats, cardboard boxes, and small portions of aircraft wreckage, about 30.5 miles southwest of the flight's departure airport, and about 10 miles from shore. The airplane is presumed to have sunk in ocean waters estimated to be between 50 and 70 feet deep. Underwater search and recovery efforts were unsuccessful, and the airplane, pilot, and passenger remain missing. A review of archived radar data disclosed that as the accident airplane approached an area about 30.5 miles southwest of the departure airport, it descended to 500 feet msl, and then entered a right turn. As the turn progressed, the airplane continued to descend to 400 feet msl, with a radar-derived ground speed of 180 knots. The last radar return was recorded with the same radar-derived groundspeed, on an approximate heading of 200 degrees. A pilot who is familiar with geographical locations in the area reported that migrating whales are commonly sighted in the area where the radar depicted a descending right turn.

The pilot, who was also the station manager, arrived at the airport earlier than other company employees to prepare for a scheduled commuter flight, transporting seven passengers and cargo to another village during hours of arctic, predawn darkness. Heavy frost was described on vehicles and airplanes the morning of the accident, and the lineman who serviced the airplane described a thin glaze of ice on the upper surface of the left wing. The pilot was not observed deicing the airplane prior to flight, and was described by the other employees as in a hurry to depart on time. The pilot directed the lineman to place fuel in the left wing only, which resulted in a fuel imbalance between 450 and 991 pounds (left wing heavy). The first turn after takeoff was into the heavy left wing. The airplane was observed climbing past the end of the runway, and descending vertically into the water. No preimpact mechanical anomalies were found with the airplane or powerplant. The aileron trim indicator was found in the full right wing down position. Postaccident flight tests with left wing heavy lateral fuel imbalances, disclosed that approximately one-half of right wing down aileron control deflection was used to maintain level flight, thus leaving only one-half right wing down aileron control efficacy. Research has shown that frost on airfoils can result in reduced stall angles of attack (often below that required to activate stall warning devices), increases in stall speeds between 20% and 40%, asymmetric stalls resulting in large rolling moments, and differing stall angles of attack for wings with upward and downward deflected ailerons (as when recovering from turns).

The pilot had contacted the FSS 11 times on the day of the accident to obtain weather briefings. The conditions were below VFR minimums, which were required to conduct the passenger carrying commercial flight in a single-engine airplane. The conditions later improved and the pilot departed under a special VFR clearance. The pilot performed two approaches at the destination airport in IMC that were consistent with the two GPS approaches that were available there. Weather data and witnesses indicate that daylight conditions, low clouds and poor visibility prevailed, with cloud tops at 1,000 feet. After the second approach, the pilot radioed that he was heading back to the departure airport because he could not see the airport. No distress calls or unusual engine noises were heard. The airplane subsequently flew north of the airport and away from the departure airport at an altitude beneath the minimum radar coverage of 2,200 feet. It impacted the frozen Arctic Ocean in a right bank and at a 60-degree nose-down attitude about three miles away from the location of the pilot's last radio transmission. An examination of the airplane (before it sank through cracking ice) revealed no pre-impact mechanical malfunctions. An examination of the propeller revealed that it was under a power setting consistent with a maneuvering airspeed at the time of impact. An examination of the autopilot annunciator filament revealed that the autopilot was not engaged at impact. The airplane was nearly full of fuel and over its published maximum gross weight at impact. Small pieces of clear ice, about 1/4-inch thick, were found on portions of the tail surfaces. Interviews with operator employees and the pilot's wife revealed that the pilot may have felt pressure from himself and passengers to complete the flight.