737 “Kegworth” crash in 1989 (UK)

This crash where a new version the 737 came down on a motorway just short of the diversion airport after the captain had shut down the good engine in the belief–based on his knowledge of the previous version–that the smoke must be coming from that one.

Furthermore, the pilots had felt the vibration and aircraft shudder, but did not refer to the engine vibration indicators clearly showing which engine had the problem because they had got out of the habit of doing so because the ones in the previous version were unreliable. This was not true for those in the new version of the 737.

When the problem engine failed completely they were too low on their approach to the airport and going too slowly to restart the good engine.

Extract from Air Crashes and Miracle Landings (85 Cases):

In total, forty-seven passengers perished, sixty-seven passengers and seven crew members were seriously injured, and four passengers and one crew member had slight or no injuries. Though considered the classic case of what not to do, there were a number of contributory factors to what came to be called the Kegworth Air Disaster, in view of its proximity to the village of that name. These include the following:

1.   False positive
Safety expert Professor Peter Ladkin says this is the only case he is aware of where a “false positive” features in an air accident. That is to say the mistaken corrective action (shutting down the good engine) seemed to be solving the problem, thus making the pilots think they had done the right thing. This is unlike in medicine, where the long period of time over which recovery or improvement of the patient for any unrelated reason can be attributed to action by the doctor or surgeon means such false positives are well known.

      The cessation of the vibrations was one thing, but to confirm things by saying that the smoke disappeared when the pilots shut down the number two engine was, with hindsight, rather dubious thinking, since smoke would not normally disappear immediately.

2.   Engine instrument system (EIS) difficult to read
Before the introduction of two-man flight crews, the primary instruments, showing the performance of the engines were in front of the pilots, and the secondary instruments, indicating the condition of the engines, such as oil temperature and pressure and vibration, were in front of the flight engineer. However, with the sidelining of the flight engineer, these secondary instruments had to be in front of the pilots.

       In the earlier version of the aircraft, the B737-300, this was done by having traditional cockpit dials with mechanical hands, as in traditional clocks, there being two panels side by side, one with the main flying instruments, and the other showing the condition of the engines. These earlier ones with needles were easy to read at a glance.

       However, as anything mechanical is liable to go wrong and anyway requires costly maintenance, LEDs were used instead of mechanical hands. However, rather than redesigning the panels to take full advantage of the virtues of an electronic display, the designers wanted to maintain the same general layout so pilots could switch from one model of the aircraft to another without expensive recertification.

       In reality, LEDs could not simulate the previous clocklike hands, because those available at that time could not be bunched up at the center of the dials to look like a continuous line. Instead, the designers placed three rather pathetic-looking LEDs at intervals around the perimeter of the dials.

       These could still be read by pilots with good eyesight when looking for a particular reading but made comparison and noticing anything unusual more difficult. In addition, Boeing had reduced the size of the secondary engine display relative to that for the primary display instruments.

      The captain and first officer had very little experience (twenty-three and fifty-three hours respectively) on the 737-400 version, and the airline did not yet have a simulator where they could have practiced using the new engine information system (EIS), with its diodes. In addition, the captain said his considerable experience with other aircraft had led him to distrust vibration readings in general, and he did not include them in his usual scan of the instruments. His conversion training had not included instruction that technical improvements meant that spurious vibration readings were very unlikely.

3.   Training and checklists
In the training of the BMA 737 pilots, the need to think or check things out before taking precipitous action was stressed, but as already mentioned there had not been training on a flight simulator with the new hybrid EIS display. There was a checklist for what to do in case of vibration from the engines and one for what to do when smoke occurred, but not one for when they happened simultaneously.

       At the time pilots at BMA had not been made fully aware that there was no need to shut down engines completely because of vibration, nor that engine fans which are vibrating or not properly aligned could have their fan tips touching the rubber seals on the periphery and that this could produce smoke and a smell of burning but did not mean the engine was on fire. Thus, as the investigators said, the situation was outside the pilots’ experience and training.

4.   Workload and stress: Fear of fire
In many emergencies, airlines usually insist that captains take control. Captains also tend to take control in difficult situations when it is not quite an emergency. Doing something physical makes the captain feel he is coping and relieves stress. The trouble with this is that the captain is concentrating on the physical task of flying the aircraft, or, as in the case of SQ006 at Taipei, maneuvering it over the slippery taxiway in bad visibility and heavy rain, and misses the larger picture.

      The flight data recorder (FDR) revealed that when the captain disengaged the autopilot, the aircraft yawed sixteen degrees to the left, a sign that the left engine was producing less power than the one on the right, but he did not seem to notice, as he did nothing to correct it. The fact that the first officer reported to ATC early on that they had an “emergency situation like an engine fire” shows they were concerned about fire, even though up to then none of the engine fire alarms had triggered.

      It is an interesting psychological point that a smell can instantly transport one mentally to a certain place, and the shaking of the aircraft followed by the smell of burning may have caused the pilots to react more instinctively and precipitously than they would have done in the event of a fire-warning light coming on. Anyway, a fire warning would have immediately indicated which engine had the problem.

      The official report made the additional point that having another pilot take over the handling of the aircraft—as PF (pilot flying)—meant monitoring of the instruments was less consistent than it might have been.

     Up until the onset of the vibration, the first officer had been flying the aircraft and would have been concentrating on the main instruments, not the engine vibration indicator, it being the role of the PNF (pilot not flying; in this case the captain, who did not believe in scanning vibration readings) to do the general monitoring. The captain must have thought the first officer had good reason to say it was the right engine that was giving trouble.

5.   Unfortunate timing.

     The pilots did not have the height or speed to restart the good engine, and not enough height to choose a flat place to land. Had the airport been farther away, they would have had found the problems with the number one engine when still high enough to restart the other one.

6.   Passengers and three cabin crew knew
Passengers at the rear who had seen the “sparks” from the left engine when the initial trouble occurred were somewhat perplexed when the captain said he had shut down the right engine but did not inform the cabin crew because the captain sounded supremely confident.

      The three members of the cabin crew who had also seen the sparks apparently did not notice the captain saying the right engine had been shut down. They knew the purpose of the announcement was to reassure the passengers and were no doubt extremely busy with their own duties as they got ready for the unexpected landing.

A retired British Airways flight attendant has suggested to the author that the failure to pick up on the captain’s mistake might have come about because cabin staff themselves often get confused about left and right, as they face backwards when addressing the passengers.

Just after shutdown of the number two engine, the captain called the flight service manager (FSM) to the flight deck to tell him to clear things for landing, and at the same time asked him, “Did you get smoke in the cabin back there?” He got the reply “We did. Yes.”

This perhaps only confirmed the captain’s mistaken view that the right-hand engine must have been at fault. The FSM departed but returned a minute later to say the passengers were panicky, and it was only then that the captain announced to the passengers that a little trouble with the right engine had produced some smoke, but it would be okay, as they had shut it down, and would be landing about ten minutes thereafter.

Leave a Comment