Piper PA32RT accident near Pflugerville, Texas, January 27, 2019

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

The pilot departed with about 40 gallons of fuel on board and completed several flights to local area airports. After about 90 minutes of total flight time, the pilot departed the last airport and had flown a few miles when the engine began running roughly. He stated he switched the fuel selector back and forth between both fuel tank positions, turned the fuel pump on, and advanced the mixture to full rich, but the engine roughness continued. He performed an emergency landing to a road, during which the airplane hit a parked car. The right wing separated from the fuselage, and the airplane flipped upside down. Nine gallons of fuel were recovered from the left wing tanks, and the inboard right wing tank was found empty, with the right fuel cap secure on the wing. The outboard right wing tank was ruptured and the right wing fuel lines were compromised during impact, but no evidence of fuel from the right wing tanks was found at the scene. The on-scene examination noted that the fuel selector lever was found in an intermediate position between the left and right tank detents, about 1 to 2 inches past center, more toward the right tank position. Recovered engine data monitor (EDM) data associated with the accident flight showed a consistent decrease in cylinder head temperatures (CHTs) during the final 3 minutes of the recording, and exhaust gas temperature (EGT) data showed a series of temperature fluctuations during the same time period. The decreasing CHTs and varying EGTs were consistent with a rough-running engine with fluctuating engine speeds. Postaccident functional engine test runs were performed, and no anomalies were noted with the engine or fuel system. When the fuel system was configured to mimic fuel available in the left tank and no fuel in the right tank, engine roughness and a partial loss of engine power was duplicated through placement of the fuel selector lever to the approximate position as found at the accident scene. The engine roughness can likely be attributed to a mixture of fuel and air in the fuel system. It was not possible to compute exact fuel usage associated with all the engine operations on the day of the accident due to a lack of information regarding power settings and altitudes flown. However, computations using best economy and best power fuel flow rates for the 94 minutes recorded by the EDM resulted in fuel used of 21.62 gallons for best economy and 37.6 gallons for best power. Although the EDM did not record fuel usage, it is likely the airplane consumed sufficient fuel to empty the right tanks during operations on the day of the accident. The pilot's operating handbook states the electric fuel pump should be turned on before switching tanks, and, to provide continuity of fuel flow, the selector should be changed to another tank before fuel is exhausted from the tank in use. If signs of fuel starvation occur at any time during flight, fuel exhaustion should be suspected, and the fuel selector should be immediately positioned to a full tank and the electric fuel pump switched to the on position. It is likely that the pilot exhausted the fuel in the right tanks before switching tanks, contrary to the recommended operating procedures, which introduced air into the fuel system and resulted in a partial loss of engine power due to fuel starvation. Additionally, the pilot likely continued to switch between the left and right tanks, and ultimately placed the fuel selector handle between the left and right tank positions, instead of selecting and leaving the fuel selector on the left tank with available fuel. This allowed more air into the system and further reduced the fuel available to the engine.

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.