Description

On 8 January 1989, a Boeing 737-400 being operated by British Midland Airways on a scheduled passenger flight from London Heathrow to Belfast Aldergrove experienced a malfunction in an engine and the engine believed to be at fault was shut down. It was decided to make an en route diversion to East Midlands, but when thrust was increased on the engine believed to be serviceable on final approach, it failed and a fire warning was annunciated. Loss of control was followed soon afterwards by terrain impact in the vicinity of East Midlands which led to the destruction of the aircraft. There was no post crash fire but 39 of the passengers died as a result of the impact and a further 82 of the 126 occupants sustained serious injuries which in 8 cases led to their subsequent death.

Investigation

An Investigation was carried out by the Aircraft Accident Investigation Branch (UK) (AAIB). Data from the 25 hour Flight Data Recorder (FDR) and the 30 minute Cockpit Voice Recorder (CVR) were successfully recovered and the absence of any post crash fire in the fuselage permitted an unusually comprehensive and valuable examination of the wreckage.

A Special Bulletin S2/89 was issued on 20 March1989 to summarise factual information then available and to list the 7 Safety Recommendations which had been made to the UK CAA up to and including 23 February 1989.

It was established that as the aircraft was climbing through FL283 with the First Officer as PF, a series of compressor stalls had occurred in the No 1 engine. These resulted in airframe shuddering, ingress of smoke and fumes to the flight deck and fluctuation of the corresponding engine indications.

The aircraft commander had almost immediately disengaged the AP and “although no words were said, it was apparent to the First Officer that the Commander had taken control of the aircraft”. The crew responded rapidly and instinctively by disconnecting the A/T and reducing thrust on the No. 2 engine without any apparent assessment of the appropriate or otherwise available actions and without noting the indications on flight deck engine instrumentation.

ATC were advised of an emergency involving an engine malfunction and possible fire. The aircraft commander had immediately begun communications with Company Operations on the second radio whilst the First Officer was following a critical Quick Reference Handbook (QRH) procedure during which confirmation was required at various points. Shortly after the No 2 engine had been shut down some 5nm away from East Midlands, where the Company Head Office and engineering base was located, the Commander had agreed to a request from Company Operations to divert there and ATC had been advised accordingly.

Once the A/T was disconnected and the No 2 engine was being shut down, the No 1 engine vibration level slowly began to reduce from the maximum level achieved during the compressor stalls. Since the thrust now required was relatively low in order to accomplish a descent before turning back towards East Midlands, the No.1 engine appeared to be functioning normally, albeit at a slightly lower N1 speed than before. This had the effect of reducing the shuddering apparent on the flight deck which helped to convince the aircraft commander that the No. 2 engine had been correctly identified as the source of the symptoms previously evident. With the APU started and shutdown of the No. 2 engine complete, the smoke and fumes which had been present in the flight deck since the onset of the engine malfunction symptoms also disappeared. The Vibration Indicator for the No 1 engine continued to reduce as thrust was reduced until it reached and remained at 2 units, still a little above normal. Despite an indication of a higher than normal fuel flow, it was concluded by the crew that the No 1 engine was running normally.

Even with the minimum track miles required to achieve descent, a landing at East Midlands was still about 20 minutes flying time from the point where engine shutdown had occurred but crew workload nevertheless remained high. The First Officer attempted to re-programme the FMS for the diversion airport and, some 7.5 minutes after the initial problem, the Commander attempted to review the initial engine symptoms, but this process was interrupted by ATC instructions and not completed thereafter.

With the No. 2 engine shut down, the Commander advised the Flight Service Manager (FSM, the senior cabin crew member) of the intended diversion and instructed them to secure the cabin. In response to a closed question from the Commander, the FSM confirmed that there had been smoke in the cabin. About a minute after this, the FSM called to say the “passengers are very very panicky” and the Commander then made a PA stating that the right (i.e. the No. 2) engine had been shut down due to a malfunction and that the aircraft would be diverting to East Midlands and landing in abut 10 minutes time. Three of the cabin crew who had seen signs of fire on the left engine later stated that they had not heard the Commander's reference to the right engine in this PA but “many of the passengers who saw fire from the No 1 engine heard and were puzzled by the Commander's reference to the right engine, but none brought the discrepancy to the attention of the cabin crew, even though several were aware of continuing vibration”.

Fifteen minutes after the engine problem had occurred and descending through 3000 feet Altimeter Pressure Settings just under five minutes prior to ground impact, the Commander had increased thrust on the engine in use (No. 1) as the aircraft was establishing on the Instrument Landing System (ILS) LOC for runway 27. This action was followed immediately by the vibration indication for the engine again rising to its maximum 5-unit reading, but this was not noticed by either pilot.

Then, just under a minute before ground impact, with the aircraft at 900 feet agl and 2.4 nm from the runway with the landing gear down and 15° flap selected, there was an abrupt loss of thrust from the No. 1 engine and a concurrent engine fire warning. The commander had immediately instructed the First Officer to attempt to restart the No. 2 engine but the attempt was not successful. It became clear that it would not be possible to reach the runway and ten seconds before impact, the Commander broadcast “Prepare for Crash Landing” on the PA.

As the airspeed fell below 125 knots, the stick shaker activated and continued to do so until the aircraft struck level ground in a nose high attitude just to the east of the M1 motorway, which at this point was in a cutting below the general level of the terrain and running perpendicular to the extended runway centreline. The aircraft then passed through some trees before a second and major impact 70 metres to the west and 10 metres below the first one on the opposite (i.e. western/northbound) carriageway of the motorway on the lower part of its western embankment. It then slid up this embankment, through more trees breaking into three main sections and coming to a final stop approximately 900 metres from the threshold of runway 27 some 50 metres to the north of the extended runway centreline (see the illustration below). There was no significant post crash fire and what remained of the No. 1 engine fire was quickly extinguished by the emergency services.

In the final stages of flight the aircraft was found to have demolished fencing and a motorway lighting column and a detached landing gear assembly had struck and deformed a motorway crash barrier.

Witnesses on the ground who had observed the aircraft shortly before the crash reported having seen definite evidence of fire in the left engine with the intake appearing to be on fire and flames being emitted intermittently from the jet pipe in unison with “thumping noises”. Others reported hearing “metallic rattling” and seeing flaming debris falling from the aircraft.

It was found that after the aircraft had crashed, a Company Engineer who had been positioning on the flight had entered the flight deck and switched off the main battery switch and the standby power switch, later returning to switch off the engine start/ignition switches and fuel booster pumps. The engine start levers (fuel valves) were subsequently found in the cut-off position by investigators but it could not be established who had moved them to that position or when this had occurred.

Following the impact the majority of the survivors were unable to evacuate the aircraft because of the extent of their injuries and/or because they were trapped or due seat failure or the effects of debris falling from overhead. “Only 14 of the passengers were able to make a significant contribution to effecting their own escape” and both pilots and one of the two forward-stationed cabin crew were trapped.

The arrival of the emergency services was assisted by the declaration of a full emergency by ATC at East Midlands Airport some 17 minutes prior to the crash occurring so that the Airport Rescue and Fire Fighting Services and supporting civil emergency services were already standing by. RFFS crews in the vehicles waiting near the runway saw the approaching aircraft descend below the motorway embankment and immediately drove to the crash site. They were able to broadcast the exact location of the aircraft six minutes after impact and had begun extinguishing the engine fire soon afterwards. Emergency services personnel, initially assisted by passing motorists, began to recover the survivors from the aircraft to waiting ambulances and military helicopters which had been deployed. The recovery of the survivors was a complex operation and it took almost 8 hours for all of them to be removed from the aircraft.

Airworthiness Issues

The investigation sought to establish the cause of the initial symptoms of engine malfunction which had been observed and confirm which engine(s) had been involved. It was found, after reference to FDR data and the completion of a comprehensive examination of the engine, that the No.2 engine had functioned normally throughout and that all indications available to the crew had indicated this. The No. 1 engine was found to be the only source of the excessive vibration which had followed the detachment due to fatigue of the outer panel of one of its fan blades. This was found to have led to a 22 second period of compressor stalls which had continued until the A/T was disconnected. It was found that “the severe mechanical imbalance which arose because of the outer panel separation led to blade-tip rubbing”, particularly on the abradable seals of the fan and booster sections, and that this had led to smoke and the smell of burning to be passed into the aircraft air conditioning system.

FDR data showed that, although the No. 1 engine began to stabilise almost as soon as the A/T was disconnected, a maximum vibration indication had continued “for at least 3 minutes until this engine was throttled back for the descent”. It was concluded that “the evidence indicated that the timing of the sudden recovery of the No 1 engine from the compressor stalling was related to the autothrottle disengagement at a point when it had demanded a lower throttle lever angle than that required for rated climb, thereby allowing this engine to achieve stabilised running at a slightly lower speed”.

In respect of the loss of thrust and the activation of the fire warning when the final approach to land was attempted using the malfunctioning No.1 engine, FDR data also showed that “fifty three seconds before ground impact, the No.1 engine abruptly lost thrust as a result of extensive secondary fan damage” which was accompanied by compressor stalls, heavy buffeting and the emission of pulsating flames. It was considered that this damage “was probably initiated by fan ingestion of the blade section released by the initial failure, which was considered to have partially penetrated, and temporarily lodged within, the acoustic lining panels of the intake casing before having been shaken free during the period of high vibration following the increase in power on the final approach.” Pieces of fan blade found on the ground at this point included two small fragments which were determined to be remnants of the initially detached blade section.

It was concluded that the attempt to restart No 2 engine at this point probably failed “because there was insufficient bleed air pressure from the No 1 engine, pressure air from the APU was not connected and the bleed air crossfeed valve was closed”. It was in any case also concluded that even if air pressure had been available, it was unlikely that useful thrust could have been obtained before the aircraft hit the ground.

It was found that the No.1 engine fire warning which had begun 36 seconds before ground impact, had been initiated by a secondary fire on the outboard exterior of the engine fan casing and that this had been caused by the effects of prolonged running of the engine with excessive vibration. This would have probably loosened fuel/oil system unions and seals on the exterior of the fan casing and, in conjunction with damage to the engine inlet ducting by fan blade debris, would have been sufficient to lead to the “ignition of atomised fuel/oil sprays by titanium 'sparks' and/or intake flame”.

Finally on the airworthiness front, it was concluded that the root cause of the initial fan blade failure was a fatigue fracture which had been initiated “as a result of exposure of the blade to a vibratory stress level greater than that for which it was designed, due to the existence of a fan system vibratory mode, induced under conditions of high corrected fan speed at altitude, which was not detected by engine certification testing”.

Crew Performance

The Investigation considered the nature of the flight crew response to the initial failure at some length. It was found that both pilots had only very limited experience on the Boeing 737 after gaining the overwhelming amount of it on other aircraft types, in the case of the First Officer, the 737 being his first jet type. At the time of the accident, the Commander had accumulated 763 total hours on the 737, nearly all on the -300 variant fitted with electro-mechanical instruments and only 23 hours on the -400 variant which had an EFIS installation. The First Officer had initially qualified on the 737 type less than six months previously and had accumulated 192 hours on type including 53 on the -400 variant.

The lack of experience of EFIS and the very short differences course required in order to operate on the -400 variant after initial type conversion on the -300 variant was considered to be of potential significance in respect of the crew response to engine vibration. The Investigation considered that EFIS engine indications provide “reduced conspicuity, particularly in respect of the engine vibration indicators” and noted that “no additional vibration alerting system was fitted that could have highlighted to the pilots which of the two engines was vibrating excessively”.

Nevertheless, it was noted that whilst the particular combination of symptoms which followed the initial fan blade failure had not been covered in training and was outside the previous experience of both pilots, that training had been in accordance with CAA requirements. It was note that training had not included FFS training on the recognition of engine failure on the (EFIS) instrumentation fitted “or on decision-making techniques in the event of failures not covered by standard procedures.”

However, despite this context, it was considered that “the speed with which the pilots acted was contrary to both their training and the instructions in the Operations Manual” and “if they had taken more time to study the engine instruments it should have been apparent that the No. 2 engine indications were normal and that the No. 1 engine was behaving erratically”. Having failed to “assimilate the readings on the engine instruments before they decided to throttle-back the No. 2 engine” they then “did not assimilate the (continuing) maximum vibration indication apparent on the No 1 engine before they shut down the No 2. engine.” Their response occurred when there were applicable QRH drills for both high engine vibration and for smoke, although it was observed that there was no drill which acknowledged the possibility of both occurring simultaneously.

It was noted that the Commander had “stated that he had judged the No. 2 engine to be at fault from his knowledge of the aircraft air conditioning system”. He had apparently had “thought the smoke and fumes were coming forward from the passenger cabin” and since he had assumed that the air for the passenger cabin came mostly from the No. 2 engine, that the problem was associated with that engine. The Investigation noted that whilst such thinking might have been applicable to aircraft types he had had more experience on, “it was flawed in this case because some of the conditioning air for the passenger cabin of the Boeing 737-400 comes from the No. 1 engine”. It was in any case noted that his “assessment” was not supported by the evidence, as it was not until much later that he had sought to confirm from the FSM that there had also been smoke in the passenger cabin. It was therefore concluded that it was “unlikely that in the short time before he took action, his thoughts about the air conditioning system could have had much influence on his decision” and that it was much more likely that he had accepted the First Officer’s initial and erroneous identification of a No. 2 engine malfunction without any attempt at verifying its validity.

It was considered that “whilst the commander's decision to divert… with the minimum of delay was correct, he thereby incurred a high cockpit workload which precluded any effective review of the emergency or the actions he had taken”. The fact that the crew had “remained unaware of the blue flashes and flames which had issued from the No. 1 engine during the period of heavy vibration and which had been observed by many passengers, in addition to the three aft cabin crew” was considered significant in respect of any subsequent review of the initial response.

It was found that the six-member Cabin Crew were all relatively recent recruits to the operator with less than three years service and all under the age of 30. Their experience on the 737 ranged from 14 months in the case of the FSM to less than one month in the case of three junior crew members.

The Investigation considered that it was “extremely unfortunate” but not surprising that the observations of many of the passengers of fire associated with the left engine “did not find (their) way to the flight deck” even when the PA made by the commander stated that he had shut down the 'right' engine. On the other hand, with similar awareness of signs of left engine distress to many of the passengers, it was considered that the three cabin crew in the rear of the aircraft might have been expected to have at least the same awareness as some of the passengers of possible error when they heard the commander's reference to the right engine. It was concluded that inaction on their part may have been because they had not assimilated any more than the general sense of the content of the PA. It was also noted that “cabin crew are generally aware that any intrusion into the flight deck during busy phases of flight may be distracting and (that) this is particularly true if the flight crew are known to be dealing with an emergency”. Despite such a firm division between flight deck and cabin in an emergency situation, it was considered notable that the FSM had made no attempt to initiate contact with the flight crew, although the lack of any initiative on the part of one or more of the cabin crew who had seen the distress of the left engine was also a potential factor. It was considered that airline crew training generally did not “provide specifically for the exercise of co-ordination between cabin and flight crew in (emergencies)”.

Survivability

It was concluded that occupant survivability in the accident had been materially increased by the absence of any post crash fire affecting the fuselage. The effects of the fire which began in the No. 1 engine and its nacelle whilst the aircraft was still airborne remained confined to that location after impact and this small fire had been rapidly extinguished by fire fighters. It was considered that, although there had been very little fuel spillage from the aircraft, “had it been a significantly longer time before fire fighting was possible, it is probable that a much greater loss of life would have resulted”.

The severe damage to the fuselage left many occupants who survived the impact with serious injuries and a significant number of them trapped. It was found that “the incidence of passenger fatality was highest where the floor had collapsed in the forward section of the passenger cabin and in the area just aft of the wing” and much less where the cabin floor and passenger seating had remained almost entirely intact in the over wing and tail sections. It was considered that survivability had been significantly enhanced because the passenger seats fitted were designed to withstand an impact force considerably greater than the extant regulatory requirements. In respect of the failure of the cabin floor, it was noted that certification of the -400 variant had occurred in accordance with the same airworthiness requirements as had been applied to the certification of the 737-100 in 1967 (a common process in certification of new aircraft type variants often referred to as “Grandfather Rights”.

It was considered that the main reason why there had been no major post crash fire was because both landing gear assemblies and both engines had separated from the wing without rupturing the wing fuel tanks. It was noted that whilst the separation of the landing gear was in accordance with design, the engine separations had occurred because of pylon fracture and the fuse-pin bolts had remained intact.

Other findings in respect of survivability included:

• Injuries sustained by a mother and child highlighted the advantages of infants being placed in car-type child seats rather than being restrained in an adapted loop-type seat belt.

• Although the overhead stowage bins met the appropriate airworthiness requirements for static loading, all but one of them separated from their attachments, despite the fact that it was concluded that their contents had been well within the permitted maximum weight.

• Some of the overhead stowage bins had also opened during the last seconds of flight which it was considered had demonstrated the need for some form of improved latching of the doors.

It was formally concluded that 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 major thrust loss due to secondary fan damage as power was increased during the final approach to land”.

It was further concluded that five Contributory Factors had been influential in the incorrect response of the flight crew:

1. The combination of heavy engine vibration, noise, shuddering and an associated smell of fire were outside their training and experience.
2. They reacted to the initial engine problem prematurely and in a way that was contrary to their training.
3. They did not assimilate the indications on the engine instrument display before they throttled back the No. 2 engine.
4. As the No 2 engine was throttled back, the noise and shuddering associated with the surging of the No 1 engine ceased, persuading them that they had correctly identified the defective engine.
5. 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.

A total of 31 Safety Recommendations were made during the course of the Investigation as follows:

On 11 January 1989:

• that the CAA consider increasing the frequency of existing engine inspections and engine health monitoring on Boeing 737-300 and Boeing 737-400 aircraft until the causes of the engine failure(s) are established. (Precautionary Recommendation)

• that the CAA call for an examination of the Boeing 737-300 and Boeing 737-400 engine Fire/Overheat and Vibration monitoring circuitry for left/right engine sense.

(Precautionary Recommendation)

On 10 February 1989:

• that the CAA in conjunction with the engine manufacturer, consider instituting inspection procedures for the examination of the fan stage of CFM56 engines to ensure the early detection of damage that could lead to the failure of a blade.

• that the CAA review the advice given in the Boeing 737-400 Maintenance Manual concerning the excessive generation of heat during blending operations with power grinding and blending tools.

On 23 February 1989:

• that the CAA should take action to advise pilots of Boeing 737-300/400 aircraft and of other types with engines which have similar characteristics, that where instances of engine-induced high vibration occur, they may be accompanied by associated smoke and /or smells of burning entering the flight deck and/or cabin through the air-conditioning system, due merely to blade tip contact between fan/compressor rotating assemblies and the associated abradable seals.

• that the CAA should review the current attitude of pilots to the engine vibration indicators on Boeing 737-300/400 aircraft, and other applicable types with turbofan engines, with a view towards providing flight crews with an indication of the pertinence of such vibration instruments when engine malfunctions or failures occur.

• that the CAA should request the Boeing Commercial Airplane Company to produce amendments to the existing aircraft Flight Manuals to indicate what actions should be taken when engine-induced high vibration occurs, accompanied by smoke and/or the smell of burning entering the flight deck and/or cabin.

On 30 March 1990:

• that the CAA should require that pilot training associated with aircraft which are equipped with modern vibration systems, and particularly those aircraft which are fitted with high by-pass turbo-fan engines, should include specific instruction on the potential value of engine vibration indicators in assisting the identification of an engine which has suffered a failure associated with its rotating assemblies.

• that the Regulatory Requirements concerning the Certification of new instrument presentations should be amended to include a standardised method of assessing the effectiveness of such displays in transmitting the associated information to flight crew, under normal and abnormal parameter conditions. In addition, line pilots should be used in such evaluations.

• that the CAA should require that the engine instrument system (EIS) on the Boeing 737-400 aircraft type (is) modified to include an attention-getting facility to draw attention to each vibration indicator when it indicates maximum vibration.

• that the CAA should ensure that flight crew currency training in simulators includes practice reprogramming of flight management systems, or any other such systems which control key approach and landing display format, during unplanned diversions so that they remain practised in the expeditious use of such systems.

• that the CAA should review the current guidance to air traffic controllers on the subject of offering a discrete RT frequency to the commander of a public transport aircraft in an emergency situation, with a view towards the merits of positively offering this important option.

• that the CAA should review current airline transport pilot training requirements to ensure that pilots, who lack experience of electronic flight displays, are provided with familiarisation of such displays in a flight simulator, before flying public transport aircraft that are so equipped.

• that Training Exercises for pilots and cabin crew should be introduced to improve co-ordination between technical and cabin crews in response to an emergency.

• that the CAA should review 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.

• that the Type Certification Requirements for gas turbine engines should be amended so that it is mandatory to perform instrumented flight tests to demonstrate freedom from damaging vibratory stresses at all altitude conditions and powers which an engine will encounter in service.

• that the potential for fuel and oil system leakage within the fan case area of high by-pass turbo fan engines, during conditions of excessive vibration, should be reviewed by the engine manufacturers and the CAA with a view towards modifying such systems to minimise such leakage and the associated fire risk.

• that the CAA should review the existing Joint Airworthiness Requirements concerning fuel tank protection from the effects of main landing gear and engine detachment during ground impact and include specific design requirements to protect the fuel tank integrity of those designs of aircraft with wing-mounted engines.

• that the CAA should expedite current research into methods of providing flight deck crews of public transport aircraft with visual information on the status of their aircraft by means of external and internal closed circuit television monitoring and the recording/recall of such monitoring, including that associated with flight deck presentations, with a view towards producing a requirement for all UK public transport aircraft to be so equipped.

• that the Manufacturers of existing Flight Data Recorders which use buffering techniques should give consideration to making the buffers non-volatile and hence recoverable after loss of power and EUROCAE and the CAA should reconsider the concept of allowing volatile memory buffering in flight data recorders.

• that where engine vibration is an available parameter for flight data recording, the CAA should consider making a requirement for it to be recorded at a sampling rate of once every second.

• that the CAA should actively seek further improvement in the standards of JAR 25.561/.562 and the level of such standards should not be constrained by the current FAA requirements.

• that the CAA should require that, for aircraft passenger seats, the current loading and dynamic testing requirements of JAR 35.561 and .562 be applied to newly manufactured aircraft coming onto the UK register and, with the minimum of delay, to aircraft already on the UK register.

• that, in addition to the dynamic test requirements, the CAA should seek to modify the JARs associated with detailed seat design to ensure that such seats are safety-engineered to minimise occupant injury in an impact.

• that the CAA should initiate and expedite a structured programme of research, in conjunction with the European airworthiness authorities, into passenger seat design, with particular emphasis on:
  • Effective upper torso restraint.
  • Aft-facing passenger seats.

• The Certification Requirements for cabin floors of new aircraft types should be modified to require that dynamic impulse and distortion be taken into account and these criteria should be applied to future production of existing designs.

• that the CAA should initiate research, in conjunction with the European airworthiness authorities, into the feasibility of a significant increase in cabin floor toughness beyond the level of the current JAR/FAR seat requirements.

• that the CAA (should) implement a programme to require that all infants and young children, who would not be safely restrained by supplementary or standard lap belts, be placed in child-seats for take-off, landing and flight in turbulence. (subsequently amended on 8 August 1990)

• that the CAA (should) expedite the publication of a specification for child seat designs.

• that the Certification Requirements for cabin stowage bins, and other cabin items of mass, should be modified to ensure the retention of these items to fuselage structure when subjected to dynamic crash pulses substantially beyond the static load factors currently required.

• that the CAA consider improving the airworthiness requirements for public transport aircraft to require some form of improved latching to be fitted to overhead stowage bins and this should also apply to new stowage bins fitted to existing aircraft.

The Final Report of the Investigation and the Appendices to the Report was submitted to the Secretary of State for Transport on 25 August 1990 and published later the same month.

Further Reading

• Loss of Control
• Compressor Stall
• In-Flight Fire: Guidance for Flight Crews
• Engine/APU on Fire: Guidance for Controllers
• Situational Awareness
• Crew Resource Management
• Grandfather Rights
• Bleed Air Leaks
• Cabin Fumes from Non-Fire Sources
• Communication from the Cabin Crew to the Cockpit Crew, Flight Safety Foundation, January 1990