May 2010 Archives

When the Cessna Citation 500 crashed into houses near Biggin Hill aerodrome on 30 March 2008, it mystified all the aviators I spoke to who tried to understand what had happened. Now the Air Accident Investigation Branch has released its final report and we are not much the wiser despite their painstaking work. There were no recorders on board - there were not required to be.

The AAIB's verdict is that, two minutes after take-off while the aircraft was flying across the airfield on its cleared departure routeing, the crew misinterpreted vibration in the air conditioning unit as an engine fault, The Board says there was no evidence of a fault in either engine, ruled out fuel contamination and a host of other possibilities, but concluded that 70sec before impact neither engine was delivering any power. At impact both engines were powering up at different stages in their re-start procedure.

The Board believe that if the crew had restarted one engine only it would have supplied sufficient power in time for the crew to have recovered the aircraft, but that starting two together slowed the process.

What a tragedy. Imagine being in the left hand seat of this Citation on the kind of day when everything you do turns to dross. You misinterpret an "expensive" noise produced by the failing bearings of an ancillary component as impending or actual engine failure. What happens next is not clear and never will be, but maybe something tells you that your right engine is the problem and you shut it down. Then, with supreme irony (if the AAIB's theory is correct), reducing power on the left engine causes fuel cut-off because of a mechanical snag. You go for engine restart and it's too late - just. You tell Biggin tower "we're going in".

If there's a moral to this story for all pilots, it's don't shut down an engine that's producing power while you are still very low, and you're overhead an airfield you can land at quickly. Easy to say. But we never will know what the crew saw and heard that made them act as they did. We don't even know for sure exactly what they did.

It's time to demand cockpit voice and flight data recorders in turbine aircraft, however small, which is what the AAIB is calling for. 

The tendency for national judicial systems almost automatically to bring criminal prosecutions in the event of commercial aircraft accidents is a growing global phenomenon, but it is not clear what is causing it.

This issue was addressed at a Royal Aeronautical Society conference in London on 28 April, and the delegates - including some very experienced lawyers who regularly act for airlines - came to some surprising conclusions. They also, with an air of mild surprise, generally voted it the most successful conference in the long history of the RAeS Air Law group.

Its ambitious agenda was entirely summed up in the event name: "The Proposed EU Regulation on Air Accident Investigation; the Criminalisation of Air Accidents and the 'Just Culture'".

I would never have guessed before the event that the European Commission's presentation of the principles behind its new draft regulation on improving accident investigation in Europe would have been generally well received, but it was. More of that later.

Back to criminalisation. Most people present, including the aviation lawyers who dominated the delegate list, think the growth of automatic criminal prosecution following aircraft accidents is undesirable and illogical. At the same time there were several robust challenges to the aviation industry's almost evangelical contention that a "just culture" system encouraging internal voluntary safety reporting is intrinsically good, and it should therefore be embraced unconditionally by the judiciary.

The benefits of a just culture were fully recognised, but the way in which the argument in favour of a just culture is being pursued was criticised, basically on the grounds of naivety.

It is naive, the conference heard, to believe the aviation industry can appeal for a privilege not accorded to any other business sector - that of conditional immunity from judicial investigation.

So has a "just culture" been killed off by the law?

No, not entirely, but it hasn't been achieved yet anyway, so we haven't lost anything.

But what of the future? Well, the lawyers present believed a better balance can be struck between the seemingly incompatible demands of running a good safety reporting system and a judicial system, and it also seems that the European Commission's draft rules for a new European air accident investigation system may offer some steps toward providing that balance.

I won't go over it all again here, because you can read about it in this week's Flight International magazine and make up your own mind about whether a new European law might show the world the way to a Just Culture.   

This Afriqiyah Airlines crash at Tripoli was not an ordinary approach accident.

Ordinary approach accidents that involve an impact with flat terrain in the last kilometre before the runway threshhold do not usually smash the aeroplane into tiny pieces. They normally leave it crumpled but more or less complete, or otherwise the structure fractures into large but recognisable components.

Apart from the fin, the aircraft was smashed into small pieces 

The tail fin's completeness in this case doesn't count: the fin is a monolithic composite structure of considerable strength that tends to snap off in impacts. So what kind of impact was this to break the rest of the aircraft up so totally? The debris field does not look as if the aircraft cartwheeled, and if it had done there would normally still have been some recognisable large components.

The hull of the Turkish Airlines Boeing 737-800 accident on final approach to Amsterdam Schiphol in February 2009 was fractured but completely identifiable. That aircraft impacted flat, clear terrain about the same distance from the runway. The Turkish 737 stalled into the ground with a high rate of descent and very low forward speed.

The Turkish 737 hit flat ground on short final approach when it stalled 

The undeniable fact is that the Afriqiyah aircraft impact was particularly high-energy.

The Airbus A330 has a terrain awareness warning system that would have alerted the crew to a high terrain closure rate, and its flight envelope protection system would have prevented the aircraft from stalling unless the crew had de-selected the "normal" flight control law, but there is no information to suggest that they would have had a reason to do that.

The weather was not challenging, in the sense that the visibility was adequate, there was little or no cloud and the wind was gentle. There is no suggestion of a technical snag because the crew did not report an emergency. The pilots were on the last leg of a long, uneventful night flight, arriving at their home airport. 

The general visibility was 6km, although there is a suggestion that about the time of the accident the visibility dipped to 2km in haze. The pilots were flying an approach to runway 09 (due east) a few minutes before dawn, but being above sea level they would have had the sun shining into their eyes through the brightening haze, which can provide the crew with an impression of hanging in the sky with no sense of speed, no external visual references,  with the ground beneath them almost invisible in the pre-dawn shadow.

Although technically the visibility would have allowed the crew to see the runway from at least 2km, it was probably invisible to them because the haze was illuminated, the pilots were dazzled by the sun's bright red disc, and the runway was in shadow.

It is not clear what approach aid the crew were using, but a precision aid like ILS was not available for 09, so the pilots had no glideslope guidance. They would have been tired after an all-night flight of more than eight hours, and looking forward to going home. 

Having the dawning sun in your eyes when you are very tired after flying all night makes your eyelids feel like lead.

Flying into the low, rising sun near your home base at the end of a night flight is perhaps the most overpoweringly soporific combination of circumstances known to man.

The aircraft, we know, flew into the ground.

If it flew into the ground without any control inputs whatever to arrest the rate of descent, that could explain the degree of damage.

I've just been in Geneva attending the European Business Aviation Conference and Exhibition, and Bombardier's amazing learning experience for aviators - the Safety Standdown - is now an embedded part of what's on offer there: a learning experience par excellence among the gleaming, glamorous hardware that EBACE presents.

When the study of human factors began in earnest during the 1970s, the focus was finding out why humans make mistakes, and the aim was to stop errors happening.

This was followed by the age of Reason (Prof Jim Reason of 'Swiss cheese model' fame).

Managers in aviation absorbed the considerable wisdom of  Reason and other industrial psychology gurus, and now the currently accepted practice is is to design your safety systems on the assumption that human error will occur, so as to mitigate the outcome that results from mistakes.

Reason's model is a great tool for understanding how humans perform within systems, and how systems can be designed, but many people have made a couple of highly misleading extrapolations beyond it: that humans do not really fail, only systems do; and that human error is inevitable.

Has "The System" (maybe your system?) stopped trying to prevent, or at least reduce, human error? It seems that many have lost sight of the individual in the organisation. Good training is traditionally the best way of keeping individuals up to scratch because it reduces the kind of human error produced by ignorance or incompetence.

Bombardier's amazing, free-at-the-point-of-delivery Safety Standdown training seminars challenge this loss of human factors focus. The Standdown draws in pilots, engineers/mechanics, and cabin crew and trains them in the stuff they didn't learn during even if they were well trained. They learn about themselves, about their weaknesses, their strengths, and they also increase the depth of their knowledge about stuff they first encountered during their training but may have forgotten since.

And no, you don't have to fly a Bombardier aeroplane to qualify to attend the Standdown. 

Now I'll let the video of my interview with Dr Tony Kern do the talking. His presentation at the Safety Standdown reminds individuals of the power they have over their own performance standards.