NTSB Factual Report

On May 10, 2015, about 0911 central daylight time, an Air Tractor AT-502 airplane, N879JA, collided with terrain during a forced landing following a partial loss of engine power shortly after takeoff from Dighton Airport, Dighton, Kansas. The commercial pilot was not injured, and the airplane was substantially damaged. The airplane was registered to and operated by Jordan Air Inc. under the provisions of Title 14 Code of Federal Regulations Part 91 as a positioning flight without a flight plan. Day visual meteorological conditions prevailed for the cross-country flight that had an intended destination of Dodge City Regional Airport, Dodge City, Kansas.

The pilot reported that he observed no anomalies during his preflight inspection of the airplane or during his before-takeoff engine runup. The airplane was last fueled (topped-off) a couple of days earlier at another airport and was not refueled before the accident flight. The pilot stated that the engine produced full torque and speed during the takeoff from runway 35, and that he made a left turn shortly after liftoff to avoid a powerline located about 400 yards north of the runway. The pilot stated that the airplane experienced a partial loss of engine power shortly after he made an initial power reduction during initial climb. His corrective actions did not restore full engine power, so he made a forced landing to a nearby muddy field. The pilot reported that the engine continued to run after the accident with a gas generator speed of about 35%. The pilot stopped the engine by turning off the electric fuel pump, repositioning the fuel valve to OFF, and moving the condition lever to the fuel cutoff position.

On May 11, 2015, a Federal Aviation Administration (FAA) maintenance inspector with the Wichita Flight Standards District Office examined the wreckage. The wreckage had been recovered to the departure airport before the examination. The FAA inspector reported that the right wing, aft fuselage, and tail were substantially damaged during the forced landing. A fuel sample obtained from the fuel control unit inlet filter revealed evidence of water contamination. Additional fuel samples were obtained from both wing fuel tanks and the firewall-mounted primary fuel filter. These additional samples initially appeared to be free of any water contamination on the day of the examination; however, over the next two days the fuel samples became opaque with a yellow tint. The FAA Inspector noted that the airplane was in an abnormal attitude when the wing tanks were drained and, as such, the samples were not likely representative of the fuel found at the lowest point of each wing fuel tank in a normal ramp attitude. On May 13, 2015, the wreckage salvage company reported that after moving the airplane they observed contaminated fuel with an orange/red color in both wing tanks and the firewall-mounted primary fuel filter. An engine examination revealed no evidence of a mechanical malfunction that would have precluded normal operation.

At the request of the National Transportation Safety Board (NTSB) investigator, the wreckage salvage company obtained fuel samples from the header tank and the fuselage-mounted primary fuel filter. These fuel samples were submitted to the National Institute of Standards and Technology's Applied Chemicals and Materials Division, Boulder, Colorado, for laboratory testing. A visual examination of the fuel samples revealed two immiscible liquid phases in their respective containers; the top phase was consistent with a hydrocarbon while the bottom phase appeared to be aqueous. The top and bottom layers of the header tank sample were clear and colorless. The top layer of the primary fuel filter sample was clear and colorless, and the bottom layer was a pale brown or orange color with similarly colored fine precipitate on the bottom of the container. The initial boiling temperature for the bottom layer of each sample was consistent with water. The initial boiling temperatures for the top layer of each sample was consistent with aviation turbine fuel. Volatility measurements using the advanced distillation curve (ADC) method confirmed that the top hydrocarbon layer of each sample was consistent with aviation turbine fuel. The water concentration was about 162 parts per million (ppm) and 160 ppm for the hydrocarbon layer of the primary fuel filter sample and header tank sample, respectively. The specified limit for water concentration in aviation turbine fuel is 30 ppm. Volatility measurements for the lower aqueous layer of each sample was consistent with water, and the water concentration was about 93% and 100% for the aqueous layer of the primary fuel filter sample and header tank sample, respectively.

The Air Tractor AT-502B Airplane Flight Manual specifies that during a preflight inspection a pilot should obtain a fuel sample from both wing fuel tank sumps, header fuel tank sump, and the fuel valve sump.

NTSB Probable Cause

The pilot's failure to follow the Preflight checklist and his inadequate preflight inspection, during which he failed to detect the water contamination of the fuel, which resulted in a partial loss of engine power during initial climb.