Aircraft Accidents and Lessons Unlearned XLVI: British Midlands 92

On January 8, 1989, British Midlands flight 92 (BRM92), a Boeing 737-400, registration number G-OBME, crashed one-half mile east of East Midlands Airport (EMA) near Kegworth, Leicestershire. The flight crew was attempting to land following an ambiguous engine failure with ‘moderate to severe vibration and a smell of fire’ on climb. In addition, the aircraft experienced longitudinal and lateral flight control issues normally associated with aileron and elevator input.

As BRM92 was diverted to EMA, the flight crew misidentified the failed engine as the number two engine (No 2) and shut it down. When BRM92 turned right to approach from the east, the number one (No 1) engine’s power was increased; at this point the mistake was discovered. 2.4 miles from the end of the runway, the flight crew tried, unsuccessfully, to relight the number two engine. The number one engine lost power and the aircraft bellied in short of the end of runway 27.

On page 148 of the Department of Transport Air Accident Investigations Branch’s accident report 4/90, the Cause stated, “The cause of the accident was that the operating crew shut down the No 2 engine after a fan blade had fractured in the No 1 engine. This engine subsequently suffered a major thrust loss due to secondary fan damage after power had been increased during the final approach to land.”

The investigators added the two contributing factors:

1.      “The combination of heavy engine vibration, noise, shuddering and an associated smell of fire were outside their [flight crew’s] training and experience,” and,

2.      “They [flight crew] reacted to the initial engine problem prematurely and in a way that was contrary to their training.”

That the investigator team (IT) described the flight crew’s failed reaction to the problem was curious; it demonstrated the IT’s unfamiliarity with an inflight engine vibration event. It also put into question the IT’s experience from a B737 pilot’s perspective.

The root cause of the BRM92 accident was not that the flight crew shut down the wrong engine; the flight crew’s reaction was contributory. The root cause: The No 1 engine failure was brought about by the fan blade separation; this failure of the No 1 engine translated into the pilots’ confusion. Assuming the Operations investigator understood, firsthand, an air carrier pilot’s training – there have been Operations investigators without this experience – why the IT focused on training as a major contributor was not made clear in the report.

The IT report relied heavily on speculation, not facts. For instance, on page 57, the report section 1.16.1 Engine Tests to Identify the Cause of Fan Blade Fatigue, documented testing the IT accomplished to identify why the suspect blade – number 17 – failed. The report determined that there was no “… material or geometric deficiency in the blade, or to any maintenance related actions.” Here the IT ruled out the blade’s material, geometric integrity and any maintenance performed. The report then stated that the manufacturer checked the fan abradable liner; “No evidence of any such influence [fan abradable liner] was found.” The IT’s engine test results established that a fan imbalance recorded on BRM92’s Flight Data Recorder (FDR), was, “… consistent with that obtained on testing an engine with a single fan blade outer panel missing.” These tests produced probable causes; they were not conclusive. How, then, does an engine’s fan suddenly become out of balance? A fan blade could have become loose in its mount; a metal abnormality subject to temperature could have caused a failure or a migrating crack reached too far across the blade’s span. These and many more reasons could have led to the imbalance, but the report did not say. The suspect blade was recovered; metallurgical tests, even in 1989, could have narrowed the focus on why the suspect blade became damaged.

The IT investigators were as listed: Investigator in Charge; Operations (qualified?); Engineering – Powerplants; Engineering – Systems; Engineering – Structures; Medical – Survivability and Flight Recorders. No one represented Aircraft Maintenance; there was no one on the IT who had balanced engine discs or blended fan blades. This conscious decision by investigative agencies to compromise safety by dismissing Aircraft Maintenance investigators has always been an unfortunate mistake.

Accident investigations, such as Ethiopian Airlines 302, Lion Air 610, National Air Cargo 102, and others, were demonstrative of how investigators, lacking experience in any and all maintenance issues, continue to make investigatory mistakes, leaving aviation less safe. All Federal Aviation Administration (FAA) Airworthiness (Maintenance) inspectors are certificated for aircraft maintenance, yet Aircraft Maintenance is still dismissed. The National Transportation Safety Board (NTSB), by contrast, benefited from experienced aircraft mechanics during only one period, when former Board Member John Goglia (1995-2004) was on the Board. Member Goglia’s guidance and experience exposed the NTSB to the correct diagnosing of maintenance issues, like blade blending and fan balance.

In cases where blades are blended; fans are balanced or vibration troubleshooting is accomplished, engineers are not involved. Engineers, if consulted by the manufacturer, overhaul facility or air operator, could be a mile away from the engine, several states away or separated from the aircraft by twelve time zones. I know this because when I worked engine vibration issues in the 1980s, there were no engineers anywhere nearby; Aircraft Maintenance consulted the maintenance manual, prepared the engine, ran the engine and balanced the engine’s fan with no help from engineering.

There is a lot to be said about identifying vibrations during high-stress situations, like during the landing cycle. BRM92’s FDR showed that while climbing through 28,300 feet, the moment the engine failed, there were “… significant fluctuations in lateral and longitudinal accelerations.” There were no fire alarms – either audio or visual, that normally accompany an engine failure – to call attention away from flight controls to engines. Smoke in the air conditioning system and the vibration could have suggested an air cycle machine failure, the flight control problems to aileron, elevator or hydraulic issues. Could the crew have turned their attention to the engine instruments only after the No 1 vibration settled?

The report stated the number five factor that “… contributed to the incorrect response of the flight crew” was, “They were not informed of the flames which had emanated from the No 1 engine and which had been observed by many on board, including 3 cabin attendants in the aft cabin”. If the flight crew was interpreting information available to them in the cockpit, making decisions during a high-stress diversion and landing, why would they consult the cabin attendants? That statement made no sense.

Airframe vibrations are not as revealing of engine problems as one would believe. When running a turbofan-powered aircraft on the ground, out-of-balance vibrations are more discernable because of the on-site conditions. Vibrations translate to the ground through the landing gear; the rigidity of the solid ground resists, reflecting the vibrations back through the landing gear and amplify through the airframe. An out-of-balance engine is much easier to identify in this situation for two reasons: First, the mechanic is well aware of which engine has the balance problem; the vibrations are expected. Second, the other engine(s), operating with correctly balanced fans, make the unbalanced engine stand out.

The report referred to several conflicting problems; per the FDR readouts, the IT felt the pilots’ attentions were misplaced. However, lateral and longitudinal fluctuations pointed to flight controls; smoke, that smelled like burning, without a fire alarm pointed to pneumatics or air conditioning. What kind of burning did they smell: plastic, metal, rubber, fuel? These separate events could have been why the pilots would not have looked at engine instruments first, which allowed the instruments to settle down. In flight, the airframe absorbs the engines’ vibrations, confusing their source; there is no ground resistance to reflect the vibration back. Engine vibration sensors may have originally isolated the disturbances, but vibration sensor measurements were not infallible. In addition, did the pilots’ confusion stem from the No 1 engine’s vibrations quickly dissipating? Did both pilots suspect another cause, such as the ailerons or elevators? Were their attentions pulled from the engines because there were no alarms? By pulling the No 2 throttle back, did this further mask the out of balance No 1 engine?

The report’s Safety Recommendations were ineffective; of the thirty-one recommendations, there was nothing that would have changed the way aircraft were inspected, pilots were trained, instrumentation reported anomalies or systems were certified, that would have increased safety. It appeared that by repeating what requirements were already in place, only with sterner words, was supposed to improve oversight and inspection, but the effort did not advance either.