United Air Lines Flight 266

Investigation

The National Transportation Safety Board (NTSB) was unable to determine why the #2 generator had failed after it had become the plane's sole power source, nor why the "standby electrical system either was not activated or failed to function."

Several witnesses saw Flight 266 take off and reported seeing sparks emanating from either engine #1 or the rear of the fuselage, while others claimed an engine was on fire. Salvage operations were conducted to recover the wreckage of the aircraft, but not much useful information was gleaned as the cockpit instruments were not recovered. The wreckage was in approximately 930 ft of water and had been severely fragmented, however the relatively small area in which it was spread indicated an extremely steep, nose-down angle at impact. There was little in the way of identifiable human remains at the wreckage site, only two passengers were identified and only one intact body was found. The #2 and #3 engines suffered severe rotational damage from high RPM speeds at impact, but the #1 engine had almost no damage because it had been powered off. No evidence of any fire or heat damage was found on the engines, thus disproving the witnesses' claims. The small portion of the electrical system that was recovered did not provide any relevant information. The Cockpit Voice Recorder took nearly six weeks to locate and recover. NTSB investigators could not explain the sparking seen by witnesses on the ground and theorized that it might have been caused by debris being sucked into the engine, a transient compressor stall or an electrical system problem that led to the eventual power failure. They also were unable to explain the engine #1 fire warning in the absence of a fire, but this may have resulted from electrical system problems or a cracked duct that allowed hot engine air to trigger the temperature sensors. The sensors from the #1 and #2 engines were recovered and exhibited no signs of malfunction. Some tests indicated that it was indeed possible for the #2 generator to fail from an overload condition as a result of the operating load shifting to it following the #1 generator's shutdown, this was maintained as a possible cause of the failure.

N7434U had recently been fitted with a generator control panel that had been passed around several different UAL aircraft because of several malfunctions. After being installed in N7434U the month prior to the ill-fated flight, generator #3 once again caused operating problems and was swapped with a different unit. Since that generator was subsequently tested and found to have no mechanical issues, the control panel was identified as the problem after it caused further malfunctions with the replacement generator. Busy operating schedules and limited aircraft availability meant repair work on N7434U was put on hold, the #3 generator was disabled. The NTSB investigators believed that the inoperative #3 generator probably was not responsible for the #2 generator's in-flight failure since it was assumed to be isolated from the rest of the electrical system.

The flight control system would not have been affected by the loss of electrical power, since it relied on hydraulic and mechanical lines, so it was concluded that the loss of control was the result of the crew's inability to see in the dark cockpit. It was theorized that the non-activation of the backup electrical system might have been for one of several reasons:

The aircraft's battery, which powered the backup electrical system, could have been inadvertently disconnected by the flight engineer following the shutdown of engine 1, as he made sure that the galley power switch (which was similar in shape and adjacent to the battery switch) was turned off (in accordance with procedures for operating with only one functional generator).

The battery, or its charging circuitry, could have malfunctioned, rendering it unable to power the backup electrical system.

The flight engineer could have mistakenly set the aircraft's essential power switch to the APU position, rather than the standby (backup) position; the switch has to pass through a gate when turning from the APU position to the standby position, and the flight engineer, turning the switch until he encountered resistance, may have assumed that this meant that the switch had reached the end of its travel and was now in the standby position, when it had actually hit the detent between the APU and standby positions. The 727's APU is inoperative in flight.

The flight engineer could simply have neglected to switch the aircraft to the backup electrical system; the United Airlines procedures for the loss of all generators did not, at the time, explicitly tell the crew to switch to backup power (instead focusing on regaining at least one generator), and it is possible that the flight engineer repeatedly tried to bring a generator back online instead of immediately switching the aircraft to the backup system.

The CVR and FDR both lost power just after the crew informed ATC of the fire warning on engine #1. At an unknown later point, both resumed operation for a short period of time. The FDR came back online for 15 seconds, the CVR nine seconds, during which time it recorded the crew discussing their inability to see where the plane was. No sounds of the plane impacting the water could be heard when this second portion of the recording ceased.

At the time, a battery-powered backup source for critical flight instruments was not required on commercial aircraft. The accident prompted the Federal Aviation Administration to require all transport-category aircraft to carry backup instrumentation, powered by a source independent of the generators.

Probable cause

The NTSB's "probable cause" stated: