Airworthiness certification standards for larger aeroplanes require that all occupants can evacuate in 90 seconds yet there have been several evacuations when this time has been significantly exceeded. NICHOLAS BUTCHER FRAeS and Dr JOHN BARNETT FIET, FCILT, FIRSE of the RAeS Flight Operations Group (FOG) review the requirements of EASA/FAA 25.803 and question if these are still appropriate.
One of the most important criteria of EASA/FAA 25.803 is that only 50% of emergency exits are available and there are many examples of where, in spite of a catastrophic accident having occurred and with some pairs of exits being unavailable, a good rate of egress was achieved with evacuations taking 90 seconds or less.
However, on 19 October 2012 a Boeing 737-800 at Glasgow, UK, experienced a significant amount of smoke in the cabin prior to take-off and the commander ordered an evacuation. With 100% of the emergency exits having been operated by cabin crew and passengers, the evacuation took almost four minutes – an evacuation which potentially could have been achieved in some 45 seconds.
This document reviews the differences between airworthiness testing and the realism of an actual operational emergency evacuation, and considers what might be the variable factors such as the location and availability of emergency exits, crew procedures and training and the co-operation of passengers in following crew instructions.
There is little doubt that passenger attitude to flight safety has changed over the years and that compliance with crew instructions in normal and emergency circumstances is less than it used to be.
The 90-second certification criteria
Airbus A380 undergoing evacuation certification testing in 2006. (Airbus)
For aircraft having a seating capacity of more than 44 passengers, 25.803 requires that all occupants must be able to evacuate under the following conditions:
● Only 50% of emergency exits are active.
● The test participants (acting as passengers), as well as the aircrew do not know which exits will be active.
● The demonstration is conducted in conditions of reduced lighting.
● Cabin baggage and blankets are positioned in the cabin aisles.
● All test participants must evacuate and be on the ground within 90 seconds without assistance from external safety personnel.
This requirement is usually demonstrated in testing by the aircraft manufacturer, although analysis and partial testing has been accepted by National Aviation Authorities (NAAs), and in some cases credit has been given for earlier tests on aircraft or variants with similar types of exits and/or exit configurations.
For the A380 25.803 evacuation test, Airbus assessed the agility level of test participants (passengers). This was the first time that such an assessment had been made prior to such a test. While there is no doubt that this may be a reasonable approach to protect test participants from unnecessary potential injury, it cannot be said to reflect normal line operations.
Future aircraft types and variants might be assessed by analysis and partial testing rather than full-scale testing. It is obviously desirable that test participants are not unnecessarily placed in ‘harm’s way’, but if there are shortfalls in the previous requirements of 25.803 then these shortfalls will be perpetuated in the certification of future aircraft.
Perhaps it might be questioned why the 25.803 requirements are only an airworthiness standard when actual evacuations are the responsibility of flight operations.
Possible shortcomings in the validation process
Cranfield University opened a large cabin evacuation simulation facility in 2001 - the only one in academia. (Cranfield University)
The 25.803 test criteria might by default result in a somewhat false emergency scenario, for example:
● The test participants know that an evacuation will happen and relatively soon after they have boarded the aircraft.
● Some test participants are employees of the manufacturer and will no doubt be well motivated to achieve a positive test result.
● The aircraft test crew may be from an operator (often the ‘launch customer’) who has ordered the aircraft and is likely to be similarly well motivated. They will have been specifically trained in evacuation techniques by the aircraft manufacturer shortly before the test.
● Although cabin baggage and blankets may be scattered throughout the cabin, the baggage is filled with low-density material and for which the test participants have no affinity – unlike their own cabin baggage.
● Historically, deactivation of the emergency exits has been such that there is always one of a pair of exits available. No 25.803 test has been identified where a pair of emergency exits has been deactivated.
● 25.803 tests are usually conducted in an orderly manner with passengers being co-operative rather than competitive.
● Given that such evacuation tests are conducted inside the controlled environment of a building, there are no weather conditions, such as wind, rain or snow that might have an adverse affect on evacuation slides.
The 25.803 test scenario also differs considerably from an actual emergency evacuation in which some or all of the following constraints may apply:
● The passengers will not be expecting or prepared for such an event.
● The aircrew may not have received their evacuation training shortly before an evacuation and it could be up to one year since their previous training check.
● The experience of a line cabin crew having to conduct an emergency evacuation might be less than a cabin crew specifically trained for a 25.803 test.
● In an actual evacuation where there is a perceived danger and threat to life, passengers are more likely to compete with each other to reach an emergency exit and therefore possibly to disrupt the flow of the evacuation.
● Passengers taking cabin baggage with them.
● Passengers taking photographic images.
● The number of usable emergency exits might be more or less than the 50% criteria and some pairs of exits will not always be available.
● The demographics of passengers in an actual evacuation will differ from those in a 25.803 test, such as children and infants, the elderly and those with reduced mobility.
Although some evacuations are subject to serious difficulties, such as unusable exits or evacuation slides, there are evacuations where most or all exits were available – but in spite of this the 90 second criteria was exceeded. There might be a number of reasons for this including:
● Delays in the flight crew, or, if necessary, the cabin crew, in commanding an emergency evacuation.
● Lack of communication between flight crew, cabin crew and passengers.
● Unserviceable communication equipment, such as public address and interphone systems.
● Crew not providing effective evacuation commands.
● Passengers not understanding the safety information they have been provided with.
● Passengers ignoring crew instructions, notably taking cabin baggage with them.
● Passengers experiencing panic and not being able to take appropriate actions.
Accident statistics demonstrate that, in a number of instances, either emergency exits were not usable or passengers took baggage with them or took photographic images during the evacuation or where the crew had no opportunity to specifically prepare the passengers for evacuation.
Only the most recent accidents have been listed above indicating that the problems are still prevalent. Indeed, especially in the case of passengers taking baggage with them in an emergency, they are in fact becoming more frequent – undoubtedly because of the increasing perceived value placed by passengers on the contents of their cabin baggage.
Injuries sustained in emergency evacuations
In any evacuation there is a potential for aircraft occupants to sustain injuries. This is true both of actual evacuations and 25.803 evacuation tests. Most injuries sustained by occupants in an evacuation are of a minor nature, such as friction burns and sprains caused by using evacuation slides. However, more serious injuries have also been recorded, such as fractured bones and in one 25.803 test a participant (passenger) suffered a life-changing injury. Placing passengers and crew in a potentially dangerous situation simply for the purposes of a 25.803 evacuation test is surely questionable, especially when other options might be available. NAAs should now consider alternative methods which might be already be available, such as mathematical modelling.
Mathematical modelling
Could simulation and modelling, such as the Fire Safety Engineering Group (FSEG)'s Exodus tool be used to update evacuation standards? (FSEG)
Mathematical modelling has the potential to replace or partially replace the actual testing requirements of 25.803. This could have the following advantages:
● Reduce or eliminate injury to test participants.
● Study variable factors seen in actual evacuations, such as pairs of emergency exits not being available, passengers taking baggage with them, actions of elderly persons or those with reduced mobility, etc.
● Enhance cabin crew evacuation procedures, taking into account variable factors, such as exit availability.
While mathematical modelling has several potential advantages, NAAs would need to be satisfied that for each model there is empirical evidence for satisfactory correlation to each factor being tested.
Availability of emergency exits
Sukhoi Superjet crash in Moscow in 2019. (Russian State Investigative Committee)
Some operators might base their cabin crew emergency evacuation procedures on recommendations made by aircraft manufacturers for specific aircraft. If such procedures are based on one of a pair of emergency exits always being available then in actual evacuations where this is not the case, such procedures might be compromised.
Potential future options
NAAs should carefully consider possible shortfalls in the 25.803 criteria. How many previous 25.803 tests would have successfully met the 90 seconds criteria if actual scenarios had been included, i.e. unavailability of pairs of emergency exits and passengers evacuating with their cabin baggage? If such testing is inherently flawed what might be possible options for the future? One way might be for NAAs to review previous 25.803 tests, now using mathematical modelling to see where the problems actually are and to see how crew members might best be trained to deal with the situations presented to them in an actual accident rather than a controlled test environment.
Summary
The British Airtours disaster in 1985 was a catalyst for improvements in evacuation safety. (AAIB)
In 2020 the RAeS Specialist Paper entitled “Emergency Evacuation of Commercial Passenger Aircraft” recommended that NAAs consider the feasibility of introducing a certification requirement for means of remotely locking from the flight deck overhead bins for taxi, take-off and landing, as well as other critical phases of flight. Given the number of recent accidents where passengers have taken baggage with them in evacuations, perhaps it is time for NAAs to now consider such a proposal. Fortunately, many recent evacuations were conducted safely but in more catastrophic events requiring more urgent actions, different outcomes might well occur.
The US Department of Transportation’s Office of Inspector General (OIG) has stated concerns “……..about the validity of the assumptions that drive FAA’s evacuation standards and industry tests and simulations for certifying new aircraft”. The OIG urges the FAA to review whether passengers really can evacuate a packed aircraft in the required 90 seconds. The OIG is of the opinion that in emergency evacuations, issues, such as passenger behaviour and demographics, seat dimension, cabin baggage, and smoke in the passenger cabin, are important factors, some of which have changed in recent years and are not necessarily reflected in the FAA requirements for certification in 25.803.
Some accidents might suggest potential shortfalls that exist in the 25.803 criteria and the need for NAAs to revisit the current requirements which were established many years ago with little amendment since. Of particular importance is the combination of the factors identified in this paper, i.e. pairs of emergency exits not being available, passengers taking cabin baggage, the lack of crew and passenger readiness for an evacuation, compounded by passengers not understanding or ignoring crew emergency instructions.
NAAs need to consider how ‘fit for purpose’ the requirements of 25.803 actually are and in particular how passenger behaviour has changed in the intervening years since 25.803 was last amended.
Recommendations
Do evacuation standards need updating for the 21st century? (CBS)
In its Specialist Paper titled ‘Emergency Evacuation of Commercial Passenger Aircraft’, the FOG made the following recommendation.
“NAAs should evaluate the use of computer-based mathematical modelling to facilitate different evacuation scenarios for initial aircraft type certification. Manufacturers, operators and cabin designers should use such mathematical modelling in the development of cabin crew evacuation procedures and whenever the number of required cabin crew is to be reduced from the number involved in initial aircraft type certification or where there is a significant change in passenger numbers.”
The FOG also makes further recommendations:
1. The FAA, EASA and other NAAs should review the requirements of 25.803, taking into account the issues addressed in this document and the relevance of such criteria with the way that passengers behave in actual emergency evacuations, including the taking of cabin baggage and non-compliance with crew instructions.
2. NAAs should review evacuations occurring on aircraft on their aircraft registers where the 90 second criteria has been significantly exceeded in order to determine the factors which might have delayed the evacuation.
3. NAAs should further investigate if, how and when mathematical modelling might be used to look at various evacuation scenarios, such as the deactivation of pairs of exits and passenger behaviour issues.
4. Additionally, NAAs should consider introducing a requirement for passenger safety briefings to include an instruction for passengers not to take photographic images during an emergency evacuation.
Further reading
Information on many operational and airworthiness evacuation issues can be found in the RAeS Specialist Paper titled ‘Emergency Evacuation of Commercial Passenger Aeroplanes’. https://www.aerosociety.com/media/14129/raes-emergency-evacuation-of-commercial-passenger-aeroplanes-paper-second-edition-june-2020.pdf
In August 2016 the RAeS issued an Insight Blog which considered ideas for preventing passengers taking cabin baggage with them in an evacuation and how it might be time for a rethink of the regulations governing emergency evacuation.
Further information can be found in a paper issued by the RAeS in June 2018 titled ‘Emergency evacuation – time for a rethink’.
In May 2020 the RAeS issued an Insight Blog regarding the accident that occurred to an SSJ-100 aeroplane at Moscow in May 2019.
In November 2020 the RAeS issued an Insight Blog regarding an evacuation of a DHC-8 following a ground collision at Toronto in May 2019 highlighting recurring issues which affect regional turbo-prop aeroplanes in similar ways experienced by larger aeroplanes.
In July 2021 the RAeS launched a video to remind passengers of the key safety factors they need to know if they need to evacuate an aeroplane.
In October 2021 the RAeS issued an Insight Blog (also published in the December 2021 AEROSPACE) regarding the safety briefings of passengers.