What happened
During a straight-in approach for landing, the pilot experienced a loss of all power on the left engine. After attempting to switch to the left auxiliary tank, the engine failed to restart. Approximately one minute later, the right engine also lost all power. Because the aircraft was at a low altitude, the pilot focused on performing a forced landing on a city street directly ahead rather than attempting to switch to the right auxiliary tank.
Upon touchdown, the main fuel tanks located at the wing tips collided with the front ends of two approaching school buses. The collision ruptured both main fuel tanks, which were found empty. No injuries were reported among the occupants of the aircraft or the buses.
The investigation
Investigators examined the fuel levels and engine performance. While the exact initial fuel load could not be determined, refueling records indicated the aircraft had at least 1s22 gallons. During the first leg of the flight, a high cylinder head temperature prompted a precautionary landing after 1.3 hours of flight time. The pilot was then instructed to return to the departure point without adding any additional fuel.
The total elapsed flight time for both legs was 2.5 hours. The pilot reported setting a fuel flow of 18 gallons per hour per engine, which exceeds the red line gauge limit; performance charts indicated the correct setting should have been 13.35 gallons per hour per engine. Even at the higher consumption rate, investigators calculated that at least 12 gallons of fuel should have remained at the time of the accident.
The pilot stated he had switched to auxiliary tanks until gauges read 2 and 5 gallons before switching back to the main tanks, though he could not recall exactly when this occurred. To prevent fuel venting, a specific tank use sequence is required: auxiliary tanks should only be used after the main tanks have been in use for at least 60 minutes to allow room for excess fuel returned by the engine-driven pumps.
No evidence of leakage was found around the wing filler caps. While staining was present on the pavement near the buses, it could not be identified as fuel or fire suppressive agent. Post-accident test runs of both engines showed no anomalies.