The Human Factor in Aircraft Incidents

The Human Factor in Aircraft Incidents
By Antoni Andre Kepczynski

This paper sets out to analyse the events leading to the crash of a British Midland Airways Boeing 737 on the western embankment of the M1 on 8 January 1989. The issue is why did the captain shut down the wrong engine leading to an attempted emergency landing at East Midlands Airport. It was to result in 47 passengers losing their lives. Two key factors which contributed to the accident were the first officer identifying the wrong engine (possibly a perceptual 'set'), and this was compounded by the lessening of smoke in the cockpit to suggesting to the captain the malfunctioning engine had indeed been identified correctly by the first officer.

Mechanical failure or pilot error, it may be a combination of both, severe or freak weather conditions (windshear) that can quite literally tear an aircraft apart, an explosion in mid air and the question arises as to whether it is a terrorist bomb or mechanical failure. TWA Flight 800 departing on the 17 July 1996 from JFK last year exploded in mid air and crashed into the Atlantic Ocean off the coast of Long Island immediately after departure. On board the aircraft were 212 passengers and 18 crew members. Subsequent investigation has ruled out a bomb or ground launched missile thus leaving mechanical failure as the likely cause of the disaster. One theory is high temperature air leaking from an air conditioning unit leaked directly onto a fuel tank. The tragedy has left the FAA (Federal Aviation Administration) and the NTSB (National Transportation Safety Board) accusing each other of dragging their feet since the cause of the accident still remains a mystery.

However, far more air disasters are solved than are unsolved and usually very quickly. The FDR (Flight Data Recorder) or black box as it is commonly called is the first piece of equipment sought for from the wreckage by accident investigators. It provides on tape every detail of the flight and will tell the investigators what happened. The CVR ( Cockpit Voice Recorder) records all communication between the flight crew and ground control. This is a vital piece of equipment since it will quite often answer the question "Why did it happen".

Accidents have been attributed to pilot error or human error. This is a somewhat ambiguous term and is inherent in our haste to attribute an accident to somebody - find a scapegoat. There is also the belief that flying an aircraft is a difficult skill to attain. But the evidence points to simple mistakes and errors being the major factors in air disasters. Comprehensive training courses, flight simulators and stringent medical checks are designed to eliminate the human error. At age 60 it is mandatory retirement for a commercial airline pilot. The aircraft have many systems e.g. pneumatic, hydraulic, electrical, and fuel.

Each system will have a backup system to take over if the primary system fails and in some instances there will be a backup to the backup. The turbojet has replaced the piston engine and is far more reliable due to having less moving parts. An engine shutdown is an extremely rare occurrence. Also, metallurgy has improved immensely. Far more is now known about metal fatigue, the cause of several De Havilland Comet disasters in the 50's. Early production Comets had square passenger windows and a fracture was forming at the corner of the window. As the aircraft was pressurising this fracture was unable to withstand the increase in pressure and the result was that the aircraft exploded in mid air. All Comets were immediately grounded pending investigation of the accidents. Subsequent aircraft were built with oval or elliptical windows.

Undoubtedly, a cause of pilot error are the commercial pressures to deliver resulting in stress and its companion fatigue. These symptoms can induce laterality, a mixing of left and right. This was all too evident on the evening of 8 January 1989.

British Midland Airways Boeing 737 Series 400 took off from London Heathrow bound for Belfast. It was the second leg of a double shuttle between the two airports. The co-pilot (first officer) was flying the aircraft. At 2005 hrs the aircraft was climbing through FL283 20 nautical miles south-south-east of East Midlands Airport when the aircraft began to shake and smoke fumes filled the flight deck. The captain took control of the aircraft and immediately disengaged the autopilot. Looking at the instruments he was unable to gain a clear indication of the problem. His knowledge of the air conditioning system immediately lead him to suspect No. 2 (Right) engine. My reaction is that this WAS a premature conclusion and indeed the accident investigators also reached the same conclusion. The first officer when asked by the captain which engine is causing the trouble replied " IT'S THE LE....IT'S THE RIGHT ONE" to which the captain responded by "OK THROTTLE IT BACK". We now have indecisiveness on the part of the first officer and he later stated at the inquiry he had no recollection of what he saw on the engine instruments. This is quite a damning admission and must have played no small part in his subsequent dismissal from the company. The accident report states he may well have experienced perceptual 'set'. Having closed the throttle for No. 2 (Right) engine the smell and visual signs of smoke abated. This convinced the captain he had selected the correct engine. At this point one must ask "how convinced was he that he had shut the correct engine down". I think he was very sure he had initiated the correct procedure. This would be substantiated by a subsequent lengthy communication to East Midlands Airport. The captain now reduced power on the No. 1 (Left) engine due to a slightly higher level of vibration and fuel flow. This action progressively reduced the level of vibration to slightly higher than normal. The captain was even more convinced he had taken correct the correct action.

If the captain and first officer had got it all wrong on the flight deck the three flight attendants and some passengers had seen signs of fire from the NUMBER 1 (LEFT ENGINE). This was described as "fire", "torching", or "sparks". The captain had broadcast over the cabin address system that there was a problem with the NUMBER 2 (Right) Engine and they could expect to land at East Midlands Airport in approximately 10 minutes. The flight attendants were understandably puzzled by the captains reference to the right engine. Some alert passengers had also been puzzled by the captains reference to the right engine having seen the fire and sparks emitting from the left engine.

Shortly after shutting down No. 2 (right engine) the first officer had obtained clearance for an emergency landing at East Midlands Airport. The landing phase is the most critical phase of flight and the workload on the flight deck was intense. Approximately 2-4 nautical miles from touchdown at a height of 900' there was a sharp decrease in power from No. 1 (left engine). Possibly, at this point it may have occurred to the captain his decision to shut down the right engine had been a catastrophic mistake. He immediately called for the right engine to be restarted but it was already too late. On the verge of a stall, the initial impact was on high ground just to the east of the M1 motorway. Passing through trees the aircraft finally impacted on the lower part of the western (northbound) carriageway of the motorway and the lower part of the western embankment.

[Photograph]

Thirty nine passengers died in the accident and a further eight passengers died later from their injuries. Of the remaining 79 passengers 74 suffered serious injury.

The cause of the accident was attributed to the flight crew shutting down No. 2 (Right) engine after a fan blade had fractured in the No 1 (Left) engine. This engine subsequently suffered a major loss of thrust due to secondary fan damage after power had been increased during the final approach to East Midlands Airport.

The Air Accident Investigating Board attributed two factors to the incorrect response of both pilots:

1. The combination of heavy engine vibration, noise, and shuddering was outside their training and experience.

I would concur with the AAIB report in their conclusion, "the combination of heavy engine vibration, noise, and shuddering was outside their training and experience". My suggestion at this point is that if he had read the Operations Manual Bulletin as he should have done it would not have been outside his training and experience.

It is reasonable to say at this point that the simulator training he had received must have been very influential at this point and would have "conditioned" his mind into thinking every engine malfunction is a shutdown procedure.

I would take the view that the responsibility he had as a commercial airline pilot should have allowed for a higher degree of flexibility but then so should of his simulator training. Total flying hours he had accumulated were 13,176. This amounts to considerable experience not only on jets but also in command. He was not familiar with the Operations Manual Bulletin issued by Boeing in March 1988 which introduced the procedure to be followed in the event of high engine vibration. This bulletin implicitly drew attention to the vibration indicators and I find it difficult to understand as to why he did not read this bulletin. It WAS the Operations Manual and as such was mandatory reading. The first officer had also not read the Operations Manual.

2. They reacted to the initial engine problem prematurely and in a way that was contrary to their training.

Indeed there is no evidence that either pilot consulted the engine instruments or attempted any other analysis of the situation before shutting the right engine down. This is a crucial omission because if the right engine had not been shut down the accident would probably never have happened. The confidence shown by the captain that he had shut the correct engine down was evident by the subsequent lengthy communications with British Midland Airways Operations.

The relative and absolute levels of competence of the captain and first officer are not well documented. The captain had been with British Midland Airways for 14 years and had no management or training responsibilities. The first officer had a total on all types of 3,290 flying hours but only 6 months jet experience. Both were current with regard to training requirements. This wide difference of experience and rank do not appear to have affected crew co-ordination. Also, the Cockpit Voice Recorder did not suggest any deference from the first officer to the captain. Indeed, the atmosphere on the flight deck appears to have been informal and relaxed both captain and first officer addressing each other by their first names.

When the emergency occurred the first officer was flying the aircraft. The captain immediately took control of the aircraft and disengaged the autopilot. It is possible this change in handling may have had an effect on the first officer's ability to interpret the engine instrumentation and the confusion between the left and right engines. This rapid change of perceptual 'set' could have influenced his identification of the wrong engine. If he did in fact experience this perceptual 'set' it was as we have seen crucial to the resulting chain of events.

Also of consequence to this tragic chain of events was that although three flight attendants and many passengers had seen the fire in the left engine the news did not make its way to the flight deck. Unfortunately patterns of airline training do not provide specifically for the exercise of co-ordination between the cabin and the flight deck in such circumstances, and one has to assume this is why no visual check was initiated.

In drawing conclusions from this accident regarding the responsibility of the captain and first officer to this tragedy I must admit to apportioning the blame entirely to the crew until having read the accident report. This was easily done since the first question I asked myself was "how can you shut down the wrong engine"? However the report does give a detailed account of the chain of events that contributed to the accident and from this I have identified two crucial points; the perceptual 'set' the first officer may have experienced which resulted in the him identifying the wrong engine and the elimination of the smoke after shutting the right engine down. This undoubtedly gave the captain the assurance he had shut the correct engine down. The report leaves these two issues not clarified. Did the first officer experience perceptual 'set' and was the elimination of the smoke sufficient grounds to give the captain the confidence he had to shut the correct engine down? These questions remain unanswered.

I would contend that the crew had not fully understood the problem and reacted prematurely. Unfortunately they did not even think to question the decision to shut the engine down. One could argue and with some justification it would have taken an exceptional crew to have identified the malfunctioning engine.

I remember the accident, have seen the crash site, and did start work at East Midlands Airport shortly after this accident. The subsequent inquiry also put the blame at the captains feet which caused quite a row with regard to the way he was treated. He did gain support from the Pilots Union and the passengers on the flight did praise his flying skill. Another 100' or so of altitude and he would have cleared the western embankment of the M1and there would have undoubtedly been less casualties.

Professional investigators had this to say about the incident:

"As even experts admit .....disorientation can be simply a matter of misinterpretation of increasingly complex flying systems leading to an inability to distinguish left from right. The most famous consequence of such an incident was the crash of a British Midlands Boeing 737 on a motorway at Kegworth.

Beatty quoted "It is important for each accident investigation department to learn from each other. Because of his training had this captain been the captain of the Herald of Free Enterprise, no airline pilot would have moved away without reassurance from a status light and human reassurance from a crew member that the doors were shut and locked. Conversely, the knowledge that "radar-assisted collisions"are repeatedly a common cause of ship accidents might have been a warning to aircraft manufacturers and airlines of the dangers inherent in trusting, interpreting and relying on machines". (Beatty, D. (1991) p. 265.

Bibliography

Beatty, D. (1991). The Naked Pilot. Methuen London

Gero, D. (1993). Aviation Disasters. Haynes Publishing Group P.L.C.

Faith,N. (1996) Black Box Boxtree,Macmillan Publishers Ltd.

Air Accident Investigations Board Report No. 4/90 (EW/C1095)

This page was coded by SkyNet Research

Disclaimer: While every effort is made to ensure accuracy, no representation is made as to the accuracy of, and no acceptance or any legal responsibility is taken for any errors, ommissions, mis-statements or mistakes within the pages of this web site or on other web sites which may be linked to this site from time to time.

Page layout by: SkyNet Research (c) - Click Here