On 2 July 2003, four people were seriously injured while escaping from a Qantas Boeing 747-400 after the aircraft's captain called for an emergency evacuation using inflatable slides.
The flight from Singapore had been uneventful. It was not until the aircraft was parked at its designated gate at Sydney airport that the alarm was sounded after fire broke out in the right wing and body landing gear.
© PA/AP Photos
BA's Flight 38 accident in January was a prime example of how every evacuation scenario is unique
The worst injury occurred when a passenger fell from an escape slide that had initially inflated, only to be punctured by a blunt object, possibly the heel of a shoe or luggage carried by an escaping passenger, something contrary to the rules and which contributed later to a review of the airline's emergency evacuation procedures.
Contrast that with the British Airways flight BA38 that in early 2008 hit the ground just short of Runway 27L at London Heathrow airport before skidding across the grass. With its undercarriage ripped off, the 136 passengers on board were safely evacuated from the Boeing 777-200ER, with miraculously only 18 people needing treatment for minor injuries.
© PA AP photos 2008
Passenger behaviour during evacuation is key
One a stationary aircraft in the relatively safe precincts of the airport apron the other an aircraft that had taken out the approach lights struggling to clear the perimeter fence: all serving to show that real-life evacuation scenarios are, to borrow from the words of Forrest Gump, like a box of chocolates in that "you never know what you're gonna get".
Not so in the world of aircraft certification. Demonstrating that a new type is deemed "safe" comes down to a simple pass or fail as part of a one-off full-scale live evacuation test that centres on an orchestrated 90s dash by several hundred primed and able-bodied volunteers.
Many even in the industry quite openly acknowledge that those bubblegum sprints represent nothing more than the fact that the aircraft has met an artificial benchmark, one however that gives no indicator of aircraft performance in a real accident.
The accepted nonchalance certainly irritates Ed Galea, UK Civil Aviation Authority professor of mathematical modelling and director of fire safety engineering at the University of Greenwich in London.
He insists computer simulation technology could make evacuation certification of aircraft an infinitely safer and more efficient process. "It could be argued that the certification trial may provide a false sense of security to the travelling public and parts of the aviation industry, who may assume that if the aircraft is certified, it must be 'safe'," says Galea.
© PA AP Photo 2008
Passing an evacuation test does not necessarily mean procedures will work in real life situations
One thing is certain, evacuations whether real, computer-generated or live full-scale experiments are always a fusion of three major components combining aircraft, crew and passengers such as exit availability, configuration, and cabin behaviour. And if you change the nature of the one of those components, you intrinsically change the outcome of the evacuation: something that lends itself perfectly to simulation through the possibility of performing an unlimited number of repeat simulations for any particular scenario.
So what exactly is the problem with how evacuation certification is conducted?
For critics, one of the chief weaknesses of the current system is the lack of realism inherent in the 90s evacuation scenario with volunteers subject neither to trauma nor to the physical effects of smoke, fire and debris.
"So the certification trial provides little useful information regarding the suitability of the cabin layout and design or the cabin crew procedures in the event of a real emergency," says Galea. He cites the 1985 Manchester disaster, in which 55 people died, as an example where the last passenger to escape from the burning Boeing 737 emerged 5.5min after the aircraft had ceased moving, while 15 years earlier in a UK certification trial, the entire load of passengers and crew evacuated the aircraft in 75s.
© PA AP Photos 2008
In the case of this Kalitta Air cargo Boeing 747-200 overrun in Brussels, it was only the crow that needed the evacution slide.
Also, while passengers may be encouraged to exit as quickly as possible during a certification trial, their behaviour is essentially co-operative, whereas in real accident situations people may become competitive and distinctly less accommodating to fellow passengers.
Other criticisms relate to the use of half the available exits, usually on one side of the aircraft, a scenario that bears little resemblance to real accident events. Again, the trial assumes that passengers have no social bond on board and further ignores the presence of passengers with disabilities, both crucial factors that can easily influence passenger behaviour and speed of egress during evacuation.
The one-off nature of the evacuation trial is also at issue, something that its detractors say must surely limit confidence in any test that claims to demonstrate the evacuation capability of the aircraft. From a design point of view, a single test further does not provide sufficient information to arrange cabin layout for optimal evacuation efficiency, and can fail to match the configurations flown by potential carriers. To adopt the same "one size fits all" test standard to all aircraft types is also questioned as to its scientific and engineering validity.
Essentially, as yet there is no framework or procedure for the use of computer models. Until such a framework is in place, Galea, who places the onus on the aviation regulators around the world, believes it unlikely that the industry will voluntarily adopt the use of simulation for evacuation certification analysis.
Aircraft manufacturers might be expected to support its inception, as full-scale evacuations remain a costly headache at around $2 million a time, not to mention the risk of litigation when things go terribly wrong.
While most common injuries to volunteers range from cuts and bruises to broken bones, the industry is haunted by the McDonnell Douglas evacuation certification trial for the MD-11 in October 1991 when one volunteer was left permanently paralysed.
Over and above liability issues, however, is any aircraft manufacturer really going to want revised acceptance criteria, inviting potentially a much more rigorous certification process through the use of computer modelling?
Under the current "make or break" single-test regime, the aircraft will pass as long as the result is below the 90s threshold and it is argued that the current procedures are designed with the sole purpose of passing the test - even if that does not necessarily mean they will work in real situations.
"It is impossible to know whether or not the outcome is a fair reflection of the aircraft's evacuation capability. In contrast, the multiple tests enabled by computer simulation generate a distribution of times, reflecting what would happen if the full-scale evacuation scenario could be repeated. This provides a better indication of the performance capability of the aircraft," says Galea.
Predictably, both Airbus and Boeing have declined to share their views on evacuation certification, other than to recommend a fireside chat with the US Federal Aviation Administration and the European Aviation Safety Agency, which set the rules under which they operate. Boeing, which has now submitted to the FAA its plan on how to demonstrate evacuation from the 787, says it would be "premature" to discuss this until its receives feedback from the US regulator.
Regulators on both sides of the Atlantic, meanwhile, have an interesting take on what is preventing progress in this area and imply that it is inertia on the part of the airframers to collaborate and share the reams of data they have amassed during certification trials.
Norbert Lohl, director of certification at EASA, defends the continued use of live aircraft evacuations as part of the certification process because they have proven to be a valuable certification requirement in light of service experience. "The full-scale evacuation demonstrates the evacuation capability of the aircraft and the effectiveness of crew training and emergency procedures. On several occasions the demonstration has also served to highlight design weaknesses overlooked at other stages in design and certification in the overall evacuation system. Computer models do not have the same ability to do this."
He acknowledges, however, that safety and practical considerations prevent a full-scale evacuation trial from being wholly realistic because they do not reflect passengers' physical disabilities, the effects of panic or the deterioration of the cabin environment.
© PA AP Photos & Kyoto news 2008
All 157 passengers and eight crew escaped from this China airlines Boeing 737 in Okinawa
"However, aircraft designs which include new features, for example new exit types, escape slides etc, will always require some testing, albeit perhaps at less than full scale. This is because computer models must always be validated against test data," says Lohl.
Professor Helen Muir, professor of aerospace psychology at Cranfield University, has led many collaborative programmes with manufacturers and regulatory authorities, on both sides of the Atlantic. She believes there is an undoubted role in the future for simulation, although argues that it is not the sole answer as modelling cannot claim to simulate all aspects of emergency behaviour.
"I see the way forward as being some sort of combination of simulation and testing because new configurations and new sorts of behaviour have to be tested," Muir says. "One of the challenges will remain getting a realistic test without putting an individual at any physical or mental danger. I mean, at the end of the day, how could anyone really test for the real panic experienced by real passengers?"
She points out that there is widespread support for full-scale trials by cabin crew unions, the critical personnel in any evacuation. "A lot of what really happens not only depends on how frightened the passengers are, but, critically, how cabin crew perform. It is important that they are assertive, take command, both vocally and by even sometimes giving the passenger a push out of the aircraft."
Ali Bahrami, FAA manager within the US aviation regulator's air transport directorate, agrees on the critical role of crew members, reinforcing the importance of crew training through to the service environment.
"The big issue with modelling an emergency evacuation is modelling for human behaviour. We are trying to get a handle on that through developing the ability of models to simulate behaviour by reviewing 37 full-scale tests by Lockheed Martin, Airbus, Boeing and McDonnell Douglas. However, an unpredictable behaviour by the very first person that jumps on to the slide may affect the success of the whole evacuation."
- How evacuation certification works
- Regulation: Prospects for change
- Evacuation: Modelling points the way forward
- Experience: Out of an A380 in 70 seconds, but so what?
- Experience: Life and death played out for real
Source: Flight International