What airlines should include in their type conversion and recurrent pilot training is a question of balance, but is the mixture right at present?

It is easy to sympathise with an airline crew whose adrenaline levels go through the roof when the stickshaker and audible stall warning alarm both activate at the gear-up point just after take-off. But in the case of a false stall warning under the same circumstances, are the pilots now reading this sure they would have - under stress - the available mental capacity to deduce the fact that the warnings were false, and to carry out a drill that would keep the aircraft flying?

The question as to whether pilots are trained and prepared to cope successfully with either a real or a false stall warning close to the ground is the main issue that has emerged from the report on the January 2000 Kenya Airways Airbus A310 accident at Abidjan, Ivory Coast. As this accident has shown, the trouble with a false stall warning is that the alerts may continue despite efforts by the pilot to carry out the stall recovery drill, so the pilots can believe that a stalled condition continues. A real stall warning is simpler in the sense that applying the correct drill will usually make the warnings stop quite quickly.

Fixation by a crew on a single problem to the extent of ignoring plenty of information warning them of developing but manageable trouble has been the cause of many fatal accidents. In this case, fixation on the stall warnings was the crew's problem. The co-pilot - the pilot flying (PF) - acted as if he assumed that the stalled condition still existed as the aircraft's speed rapidly increased in stable flight. Meanwhile, the cockpit voice recorder (CVR) picked up no voice communication or actions by the captain during the 25s between warning onset and crash, except for an exclamation of surprise when the alert sounded and the altimeter advice "go up" 1s before impact. This total crew preoccupation with an alert system killed the whole crew and almost all the passengers.

If it had been daylight the accident would probably not have happened. The total darkness and the fact that take-off took the aircraft directly over a calm sea meant the crew lost the compelling visual cues that would have warned them of their approach to the surface. Fatal accidents that began with a climb over the sea at night have claimed three aircraft since 2000: the Kenya Airways A310, a Gulf Air A320 at Bahrain later the same year, and a Flash Airlines Boeing 737 at Sharm el Sheikh, Egypt this January. There are differences, however: the Gulf Air event was loss of control through crew disorientation; the Flash Airlines crash - not yet fully investigated - looks as if the crew's awareness of aircraft attitude was a factor, and the Kenya Airways accident was effectively controlled flight into terrain owing to crew lack of awareness of closure with the surface and their failure to realise that the aircraft was, in fact, fully under their control. In the A310, the ground proximity warning system (GPWS) which would normally have given a pull-up alert did not do so because the stall warning audible alert takes precedence over GPWS in the warnings hierarchy. Nevertheless, the radio altimeter voice count-down from 300ft should have alerted the crew to closure with the surface, but their fixation did not allow them to heed it.

The report's main recommendation is that exposure to a false stall warning during crew type conversion and recurrent training should become a requirement. The question is, should it? Every accident could lead to new training requirements until the lengthened training programmes become ineffective because the more probable accident causes become obscured by those - like false stall warnings - which are rare occurrences.

This is a plausible item to train for - especially in conventionally controlled aircraft that are not flight envelope protected - if only because the instant reaction to a false stall warning is identical to that for a real one. Stalling is a potential killer as every pilot knows, but the recovery procedure is relatively simple. Making the reaction to a stall almost instinctive is worthwhile, and one of the essentials of which pilots need to be reminded is that at the onset of a stall warning the aircraft is close to - but not yet in - a stalled condition.

That is a crucial reminder for pilots faced with a stall warning at low altitude, because the need to increase power to maximum and level the wings immediately is paramount, and the angle of attack may not need changing if the other reactions are fast. Loss of height is not necessarily a prerequisite, and practice of the drill would ensure that pilots, having made the correct initial reaction, can then familiarise themselves quickly with the routine for examining whether they are really stalled or not.

Source: Flight International