Raytheon Aircraft Company G36 accident near Horseshoe Bay, Texas, May 28, 2015

The NTSB's narrative final · quoted from the NTSB record

About six weeks before the accident, the pilot aborted a takeoff when the airplane’s engine did not develop the expected manifold pressure required for takeoff. After the aborted takeoff, the pilot completed several engine runups to troubleshoot the engine issue. The fuel flow to the engine would suddenly decrease when the engine power was increased above 24 inches of manifold pressure. The pilot’s aviation mechanic was unable to replicate the engine issue on two subsequent occasions. On the day of the accident, the pilot completed an uneventful before-takeoff engine runup. The airplane then had a loss of engine power during the takeoff run when the pilot increased the engine power to about 25 inches of manifold pressure. The pilot immediately reduced engine throttle, and the engine resumed normal operation at a lower manifold pressure. The pilot aborted the takeoff, taxied off the runway, and attempted to troubleshoot the engine issue. However, he was unable to replicate the engine issue during a subsequent engine run-up, during which the engine ran normally at 28 inches of manifold pressure. He also verified proper operation of the fuel selector, magnetos, battery and alternator, engine-driven fuel pump, auxiliary fuel pump, and mixture control. The pilot decided that he would fly the airplane to another location where it could be further examined by his aviation mechanic. The pilot selected 30 inches of manifold pressure and 33 gallons per hour (gph) for the takeoff, but the engine power decreased to 25 inches of manifold pressure and 12 gph at some point during the takeoff roll. The pilot stated that he believed the airplane would have departed the end of the runway if he had aborted the takeoff and that, despite the partial loss of engine power, he thought he could save the airplane by continuing the takeoff. The pilot was unable to restore full engine power and a forced landing was completed to a nearby highway. The airplane was subsequently destroyed by a postimpact fire. A mobile phone video showed the airplane began its takeoff roll about 175 ft from the start of the runway. The airport manager reported that the airplane did not appear to accelerate normally during the takeoff roll. Video of the takeoff showed the airplane became airborne about 3,560 ft down the runway from where it started its takeoff roll. Takeoff performance calculations determined that about 1,250 ft of the 5,977 ft-long asphalt runway was required for a normal takeoff at maximum gross weight and with the wing flaps retracted. Based on the available information, the pilot was aware of the partial loss of engine power during the takeoff run and should have aborted the takeoff before liftoff, which required about 3 times the expected runway distance. After liftoff, the airplane was observed in a shallow climb before it rolled to the left and descended into trees and terrain. The airplane was subsequently destroyed during a postimpact fire. Postaccident examination of the engine and its accessories did not reveal any mechanical malfunction that would have resulted in a loss of engine power. However, the engine, fuel system, and turbo-normalizing components sustained extensive damage that precluded the additional testing required to determine why the engine lost power.

The complete narrative as the NTSB published it. The NTSB's docket holds the report as a PDF and any photographs, statements and other documents from the investigation.