G-OBME was a British Midlands 737-400, flying between Heathrow and Belfast.

During that flight, the crew experienced a sudden and severe vibration, accompanied by smoke in the cabin. There was no fire warning, and although they suspected a fire in the right engine, the crew were unable to determine the cause.

The crew responded by closing the right throttle, then shutting the engine down. Unknown to the crew, the right engine was functioning normally at the time, but cabin crew and passengers had in fact seen flames coming from the LEFT engine.

Unfortunately, this fact was not communicated to the cockpit. The crew diverted to East Midlands airport, where on approach they experienced a major power loss from the left engine. With no time to re-light the right engine, the aircraft struck the M1, approximately one kilometre from the runway. The aircraft was totally destroyed. 47 passengers died and major injuries were sustained by a further 67 passengers and seven crew.

At 1952 hrs. on 8 January 1989, G-OBME left Heathrow for Belfast with the F/O handling the aircraft. It was completing the second leg of a double shuttle between the two airports, and was cleared to FL350 on a direct track to the VOR at Trent.

At 2005 hrs., the aircraft was passing through FL283 about 20nm. SSE of East Midlands Airport when moderate to severe vibration of ’rattling’ and the smell of fire occurred. There was no fire warning. The F/O commented that they had ‘got a fire.’ The Captain took over, and immediately disconnected the autopilot. The F/O continued ‘It’s a fire coming through.’ The Captain said later that although he looked at the engine instruments, he found no clear indication of the cause of the problem. He said he thought the smoke and fumes were coming forward from the cabin, and his understanding of the air conditioning system led him to suspect the right engine.

The F/O also monitored the engine instruments, and when asked which engine was causing trouble replied: ‘It’s the le… it’s the right one.’ The captain responded: ‘OK, throttle it back.’ The auto throttle was then disengaged and the right throttle was closed. The F/O later had no memory of what it was on the engine instruments that led him to make his assessment.

The Captain’s instruction to throttle back was given 19 seconds after the onset of vibration when, according to the Flight Data Recorder, the right engine was operating normally.

The checklist gave separate drills for high vibration and smoke, but no drill for a combination of both. The combination of symptons was outside the crew’s training and experience but they had responded immediately to the emergency.

The F/O advised London ATC of an emergency that looked like an engine fire. The Captain then instructed ‘Shut it down’ 43 seconds after the onset of the vibration, but this action was deferred as the Captain said ‘Seems to be running alright now. Let’s just see if it comes in.’ The F/O was responding to ATC’s advice of the aircraft’s position and request for nomination of alternate landing field. He told ATC they would probably go to EMA but to stand by.

In the cabin, an F/A used the PA to advise passengers to fasten their seatbelts. The three aft cabin crew had seen the flames from the left engine, as had a number of passengers, but this was not communicated to the cockpit.

The F/O told the Captain he was about to start the ‘Engine Failure and Shutdown’ checklist, saying also: ‘Seems we have stabilised. We’ve still got the smoke.’ Action on the checklist was suspended as the Captain called his company, British Midlands Airways, at EMA.

At 2 min. 7 sec. From the onset of the vibration and during a pause in communications with BMA Ops, the fuel cock of the right engine was closed and the APU started. BMA Ops suggested diversion to EMA. The F/O had begun again on the checklist, but was interupted by transmissions by the Captain and the company. Throughout this period there was no fire warning.

The Captain later said that after shutdown, the smoke and smell cleared, reassuring him that action had been correct. Power was further reduced on the left engine, which showed only a higher than normal level of indicated vibration and increased fuel flow. The high vibration level continued for a further three minutes and then fell progressively to a little than higher than normal. The Captain later said indications for the rest of the flight were that the emergency was over.

Soon after shutdown the Captain had called the FSM to the cockpit and asked ‘Did you get smoke in the cabin back there?’ His answer was ‘We did, yes.’ The Captain instructed him to clear up the cabin and pack everything away, but a minute later he returned to the cockpit saying ‘Sorry to trouble you… the passengers are very, very panicky.’ The Captain broadcast to the passengers that there had been trouble in the right engine which had produced some smoke, but that the engine was now shutdown and they would be landing at EMA in about ten minutes.

Afterwards, F/A’s said they had not heard his reference to the right engine, but many passengers had been puzzled by it. None had told the F/A’s of the discrepancy. The smell of smoke had dissipated here too, though some passengers had seen fire in the left engine and many later described the smoke and smell of oil, burning and hot metal.

At engine shutdown, the aircraft was approximately 5nm. South of EMA. London ATC had transferred control to Manchester ATC, who passed headings for the aircraft to steer to descend to the north of EMA and fly to the localiser of the ILS for runway 27.

During descent the Captain continued to manually fly while the F/O carried out the radio work. The workload was high as the F/O gethered weather details and for two minutes tried unsuccessfully to program the flight management system to display the EMA landing pattern.

At 2012 hrs, the Captain attempted to review the situation – ‘Now what indications did we actually get? Just rapid vibrations in the aeroplane…. Smoke….’ Discussion was interupted by ATC passing a new radar heading, descent clearance to FL40, and instructions to change radio frequency.

The F/O began to read the one engine inop approach checklist. ATC asked for a test call to the aerodrome fire service, which the Captain made without response.

The approach checklist was completed vy 2017 hrs. at 15nm. From touchdown and at 6500ft. At 2018 hrs. the Captain accepted a new radar vector to take the aircraft south to increase the distance from touchdown, and shortly called for 1 degree flap. Throughout the descent there were distractions from a small number of other aircraft making radio calls on the same frequency.

At 13nm. Out on the new heading and descending through 3000ft., ATC advised a right turn to bring the aircraft to the centreline. At 2020hrs. during this turn, power was increased on the left engine to level the aircraft at 3000ft., and maximum indicated vibration was then recorded on the FDR. The aircraft began a slow descent to 2000ft. with 2 then 5 degrees of flap.

At 2023 hrs., 2,4nm. From touchdown at 900ft. agl, there was an abrupt drop in power from the left engine.

The captain called immediately to the F/O ‘Try lighting the other one up – there’s nothing else you can do…’ The captain raised the nose of the aircraft in an attempt to reach the runway. Seventeen seconds after the power loss, the fire warning system operated on the left engine, and seven seconds later, the GPWS warning sounded and continued. The captain ordered the F/O not to carry out the fire drill.

No power had become available from the right engine. At 2024hrs. the captain broadcast to the cabin ‘Prepare for crash landing – prepare for crash landing.’ Two seconds later the stick shaker activated as airspeed dropped through 125kts. To 115kts., and the aircraft struck the ground at 2024.43.

The initial ground impact was in a nose – high attitude on level ground just to the east of the M1 motorway. The aircraft then passed through trees and suffered its second and major impact on the northbound carriageway of the M1. The fuselage was extensively disrupted, and the aircraft came to rest entirely on the wooded western embankment approximately 900m. from the threshold of runway 27 and displaced 50m. to the north of the extended runway centreline.

Several of the passengers described heavy vibration immediately prior to the impact, and one passenger, in the rear of the aircraft, described the vibration as being severe enough to open the overhead lockers and cause them to spill contents. Passengers in the rear of the aircraft described two distinct impacts; those in the front appeared only to have been aware of the final impact.

Ground witnesses who saw the final approach saw clear evidence of fire associated with the left engine. The intake area of the engine was filled with yellow/orange fire, and flames were observed streaming from aft the nacelle, pulsating in unison with ‘thumping noises’. Metallic rattling was also heard, and flaming debris was seen falling from the aircraft.

Of the 118 passengers, 47 dies and 67 suffered serious injuries. Seven of the eight crew had serious injuries. The aircraft, manufactured only in 1988, was destroyed.

The 43-year-old Captain(13180hrs. including 765 on the B737-300/400 types) and the F/O were later dismissed.

The official report includes the following paragraphs:

Flight Crew Coordination

Among the important factors that affect the ways in which individual crew members relate to one another are their personalities, relative ranks, roles (ie handling, non-handling) and relative levels of competence.

The commander, although he had no management or training responsibility, had been a captain with the operating company for 14 years, whereas the first officer had flown jet transport aircraft for only six months.

Nevertheless, this wide difference in rank and their previous limited association appear not to have influenced coordination adversely. The CVR did not suggest any undue deference from the first officer to the commander, and the atmosphere appeared relaxed in the early part of the flight with both pilots addressing each other by their first names.

Although the first officer was the handling pilot when the emergency occurred, the commander then disengaged the autopilot and, although no words were said, it was apparent to the first officer that the commanderhad taken control of the aircraft. This change in handling may have had an effect on the first officer’s ability to interpret engine instrumentation.

Since he was likely to have been more concerned with handling the aircraft than with monitoring these instruments during the early part of the flight, he was not, perhaps, as acutely associated with interpreting them as he would have been as the non-handling pilot, and this rapid change of perceptual ‘set’ could have contributed to his identification of the wrong engine.

The relative and absolute levels of competence of the crew members are difficult to guage. Both pilots had met company requirements during conversion training and subsequent base and line checks, and their training records reflect no difficulty in comprehension, or lack of competence.

There is, therefore, no suggestion that any large ability mismatch on the flight deck affected coordination. Indeed, the CVR suggests that the pilots worked together as a team throughout the flight, and that the decisions made on the flightdeck were all accepted jointly.

Flight Deck – Cabin Coordination

It was extremely unfortunate that the information evident to many of the passengers of the fire associated to the left engine did not find its way yo the flight deck even though, when the commander made his cabin address broadcast, he stated that he had shut down the ‘right’ engine. The factor of the role commonly adopted by passengers probably influenced this lack of communication.

Lay passengers generally accept that the pilot is provided with full information on the state of the aircraft and they will regard it as unlikely that they have much to contribute to his knowledge. Even those passengers who noticed the commanders’s reference to the right engine may well have assumed that the commander had made a slip of the tongue, or that the problems they had seen with the left engine were in some way consequential to an important problem with the right engine that the commander had dealt with. It cannot therefore be regarded as surprising that information from the passengers was not made available to the pilots.

The same information was available to the 3 cabin crew in the rear of the aircraft but they, like the passengers, would have had no reason to suppose that the evidence of malfunction they saw on the left engine was not equally apparent to the flight crew from the engine instruments. In addition, it would appear that there was not the same awareness of possible error, since these cabin crew members stated that they had not heard the commander'’ reference to the right engine. This may have been because the cabin crew, engaged on their own duties, were not aware of any more than the general sense of the broadcast.

In addition, cabin crew are generally aware that any intrusion into the flight deck during busy phases of flight may be distracting, and this is particularly so if the flight crew are known to be dealing with an emergency. There can thus be at times a firm division between flight deck and cabin, and it is notable in this context that in this accident the flight service manager made no initial attempt to approach the flight deck until he was called.

However, it must be stated that had some initiative been taken by one or more of the cabin crew who had seen the distress of the left engine, this accident could have been prevented. It must be emphasised, nonetheless, that patterns of airline training at the time did not provide specifically for the exercise of coordination between cabin and flight crew in such circumstances.

The Influence of Stress

One aspect of flying that is extremely difficult to address in training is the stress presented by an emergency. Although all pilots are aware of the general requirement to avoid making hasty decisions in the air, it is much easier to advocate such a policy on the ground than it is to execute it in the air when presented with an unusual emergency. The response of any individual to a given emergency will be affected by three factors – the perceived severity of the problem, the personalities of the individuals concerned and the training they have received.

The noise and smell which suddenly alerted the crew to the emergency quickly led them to believe that they were experiencing a severe problem, and it may reasonably be assumed that this would have had a marked effect on their affective states.

No formal assessment of the personalities of the pilots has been undertaken, but there was nothing in their records to suggest that they were likely to differ significantly from most other pilots in their response to stressful stimuli.

It is notable from many accidents that crews are more likely to remain calm during hazardous events if they understand the situation and have an appropriate drill to implement. In this accident the combination of initial symptoms was outside their experience and training, and there was no specific drill for such a combination.

Thus the combination of severity and novelty must have acted to increase their arousal. Under such circumstances it is understandable that their first desire was to identify the problem. Although this is obviuosly a first requirement in order that action may be taken, uncertainty reduction also has considerable psychological importance in that it is much more comfortable and reassuring to be able to impose structure on a situation and deal with a known rather than an unknown problem.

Two effects of increased arousal on the desire to reduce uncertainty may be contemplated. The first is unneccesary haste in making a decision about the nature of the problem and the second is a failure to question that decision, once it has been made.

Although there is evidence in this accident that both these factors may have prevailed to some extent, there is no evidence that this crew was abnormally affected by stress, or that they responded to the situation in a uniquely unexpected manner. In particular it should be noted that the second of these effects – the reduction in likelihood that they might question the accuracy of their decision – would have been heavily influenced by the fact that the reduction in the right engine power had apparently stopped the shuddering.

There is considerable research on the topic of ‘casualty’ from which it is clear that in this situation it would have required an exceptional crew to question the the association between their action and its apparently obvious, and highly desirable, consequence. The commander attempted, during a period of slightly lower workload, to review the events that had passed. It was unfortunate that further events intervened to curtail this review since it is possible that he may have realised, given more time, that there was a risk that they might have shut down the wrong engine.

A last factor which may have influenced this crew’s behaviour, given the stressful nature of events, is the flight simulator training which they would have experienced. In the simulator virtually all engine problems result in as engine shutdown. Since this crew would have been under both practical and psychological pressure to come up with a program of action, it cannot be regarded as surprising that the actions they embarked upon were those that they had practised in the flight simulator.

Flight Crew Training

The performance of flight crew in emergency situations may be regarded as a product of their natural ability and their training. It is possible to identify three aspects of the circumstances of this accident where a different pattern of training could have favourably influenced the outcome. The ability of the pilots to extract information from the EIS must be questioned, and so must the apparent lack of coordination between the flight deck and the cabin crew. The most important issue,however, concerns the preparation of the pilots generally to cope with unforseen situations which are not covered in their emergency checklist.

Training for Flight Crew/Cabin Crew Coordination

It could be argued that the pilots of the aircraft did not make effective use of the cabin crew as an additional source of information. Such cooperation could be encouraged by joint training exercises between flight and cabin crews. In addition, it should be possible to provide simulator exercises in which it would be appropriate for pilots to ask for cabin crew to give a briefing on events in the cabin and for the role of the cabin crew to be taken, in such exercises, by the simulator instructor.

Such training would serve to provide pilots with the knowledge that cabin crew are a source of information that should be considered in certain emergencies. Equally, cabin crew could be trained to appreciate that one factor which they should consider during any emergency is the provision to the pilots, in a timely way, of a summary of the sights and sounds witnessed in the cabin.

It was therefore recommended that training exercises for pilots and cabin crew should be introduced to improve coordination between technical and cabin crews in response to an emergency. (Made 30 March 1990.)

Technical Training

With the increased technical complexity of modern aircraft systems it has become generally accepted that pilots cannot be expected to have an in-depth technical knowledge of all the systems of their aircraft. In addition, technical development has produced systems with much improved fail-safe characteristics which can give the flight crew continued system performance following anticipated discrete failures within such systems. Because of this technical progress, associated pilot training has become largely based on the ‘nee to know’ principle. This approach has its limitations and largely rests on the assumption that all technical failures with which a flight crew may be confronted can be anticipated. This is an unsafe assumption.

In this accident, the pilots were suddenly presented with an unforeseen combination of symptons that was outside their training or experience. It may be contended that fan/compressor blade contact with the surrounding abradable seals during conditions of severe out-of-balance running could not have been anticipated technically. Even if this is accepted, this effect has now been demonstrated.

It is also apparent that this flight crew did not assimilate the readings on both engine vibration indicators. The reaction of pilots to indications on the flight deck is modified by their general experience and many pilots of earlier gas turbine aircraft have become dismissive of engine vibration indicators due to the inferior performance of such systems. Such views are liable to prevail on modern aircraft unless the technical knowledge of pilots is effectively revised.

The prime factor which appeared to confirm to this crew that the right engine was at fault was the sudden reduction in noise and shuddering which occurred when the right engine was throttled back. It is considered that this effect was due to the autothrottle disconnection which preceeded throttling of the right engine. This is yet another technical systems finding which should be covered in pilot training since it may have implications for engine failure discrimination where the affected engine is experiencing compressor stalls, with autothrottle engaged.

Such findings raise questions concerning the level of technical training available to pilots of modern aircraft. In addition they illustrate the point that such training should also include an appreciation of systems response under abnormal conditions, where the associated symptoms have the potential to mislead flight crews in an emergency situation.

Decision Training

Training of flight crew can be classified under three broad headings. There is training designed to provide the pilot with specific handling or psychomotor skills; that designed to ensure that pilots are familiar with procedures (eg the pattern of behaviour required to to deal with an engine fire); and that designed to provide the pilot with general techniques for dealing with unexpected and possibly poorly defined problems.

Flying training and checking has traditionally concentrated on the first two types of training, and this accident provides evidence of the efficiency of that training. The aircraft was controlled satisfactorily by the captain, and the engine shutdown procedure was carried out with accuracy.Any errors made on the flight deck in this accident were made at the highest decision making level.

The crew was not presented with a clear cut fire warning, with which they would have undoubtedly dealt satisfactorily, but with a noise, shuddering and the smell of burning. Nowhere in their previous experience would they have been presented with this particular situation, and it was therefore up to them, or at least up to the commander, to formulate a plan for dealing with it.

Because it is not possible to anticipate every emergency or combination of circumstances and reduce such situations to a level at which pilots may be trained to deal with the procedurally, it may be argued that it is the essence of the pilo’s task to bring his flexibility and decision making potential to bear on those situations that cannot be anticipated. Such considerations have led to the development of ‘Line Orientated Flying Training’, ‘Cockpit Resource Management’. ‘Flight Deck Management’ and othe concepts designed to provide pilots with experience of evaluating unusual situations, albeit in the flight simulator, in the belief that such practice in making high – level decisions will transfer positively to the real flight deck.

Such courses were not required to be presented to British airline transport pilots, and hence their training tended to contain a heavy procedural bias. This accident emphasised the fact that occasions arise when it is important for pilots to have the ability to evaluate novel situations correctly, and that consideration should thus be given to requiring that Evidence Evaluation and Decision Training, as well as Procedural Training, should be included in company training and checking procedures.

It was therefore recommended that the CAA review their current airline transport pilot training requirements with a view towards considering the need to restore the balance in flight crew technical appreciation of aircraft systems, including systems response under abnormal conditions, and to evaluate the potential of additional simulator training in flight deck decision making.

This recommendation has since been adopted with CRM training (as we now know it) being made a compulsory component of Airline Training, and together with LOFT, is an integral part of Human Factors Training.