Time to take a fresh look at aircraft emergency egress testing

It’s understandable if airframers and regulators are reluctant to discuss the emergency egress requirements for commercial airliners. Nobody wants to encourage passengers considering flying, one of the safest modes of transportation, to think of the dread risks — low-probability, high-consequence events like aircraft accidents — since research shows that passengers who choose to drive instead of flying are putting themselves at significantly greater risk of injury or death.

Yet as the passenger experience industry continues to create innovative uses of onboard space, the silence around reviewing emergency egress requirements and modeling is of concern. Increased aircraft survivability during incidents is impressive, but the evacuation patterns of documented incidents and new understandings of passenger behaviour do raise a number of questions.

Trends like the increasing focus on airport profitability via the use of duty free sales come into play. Images of passengers fleeing airliners clutching their sealed plastic bags of Johnnie Walker abound. Similarly, increasing the amount of passenger overhead bin storage is laudable from a #PaxEx point of view, but passengers have a propensity to try to fish out their rollaboards even when escaping after an incident. A complicating factor, of course, is the reduction in at-seat storage resulting from some slimline economy seats, necessitating the stowing of valuables overhead. While egress testing includes items strewn on the floor of the aircraft, there appears to be no requirement for a proportion of test passengers to open overhead bins or attempt to bring items with them.

Also seemingly lacking from tests: passengers outside optimal ranges of fitness. The 2006 Airbus A380 egress testing involved volunteers “drawn 50/50 from Airbus employees and local gymnasium members”. Certification authorities tend to require a certain minority of passengers be over a certain age, with this testing required 35 percent of participants over 50, with 15 percent women over 50. Yet what discussion has been had about whether this is the correct split for regulators to demand, reflective of aging populations?

If cabin crew trainees with restricted body sizes, who have been medically certified as fully fit, can’t slide down from the A380 upper deck without smashing their faces into the tarmac even from a sitting start and trained body positioning (see video below), what about the greater momentum of a larger passenger? Or a passenger who is pushed, slips or falls onto the slide? Or an older passenger or someone with lower bone density? What impact would these diverse groups of passengers have on their own safety or on the safe egress of others?

Moreover, increased aircraft density, particularly among airlines in developing markets — Cebu Pacific’s all-economy A330-300, with its 16.5” wide 3-3-3 seats starting at 30” pitch, seats 436 passengers, four shy of the maximum capacity of 440 — is unlikely to combine well with new populations of passengers unaccustomed to air travel and unfamiliar with the operation of seat belts, evacuation slides, procedures and so on.

Cebu Pacific A330

Cebu Pacific A330

Aircraft evacuation and certification testing also takes place in the language of the volunteer evacuees. The Lufthansa flight attendant shouting “Schneller! Schneller! Schneller!” may have been useful to encourage a largely German-speaking group of test subjects in a Hamburg hangar, but what testing has been done to take account of, say, a Mandarin-only speaker on a Korean Air aircraft flying between Los Angeles and Sao Paulo?

With increasingly spacious premium cabins combined ever-denser economy cabins, the centre of gravity in terms of passenger density is shifting to the rear of the aircraft. At the same time, the Asiana Flight 214 crash in San Francisco that demonstrated that evacuation slides inflating inside the cabin can present an impediment to egress. What would happen if the fifty percent of doors inoperable in a incident involving structural damage to the rear of (for example) a five door 777-200ER were the four doors behind the wing, necessitating the entire economy cabin evacuating through doors 2 ahead of the wing?

Asiana 214 economy cabin --NTSB

Asiana 214 economy cabin post-crash

Furthermore, the configuration of the 50 percent of exit doors simulating inoperability during testing is largely opaque for most aircraft. The A380 live evacuation tests in Hamburg, for example, rendered all doors on the left-hand side of the aircraft inoperable. Requests from RGN to both Boeing and Airbus for information about cabin configurations for other aircraft certification tests were not answered.

Airlines, airframers, regulators and the rest of the industry need to ask these questions — and they need to answer them.


  1. @masterdPm

    That is an interesting piece.
    I do, however have comments.

    The evacuation tests are textbook tests for several reasons.
    A – evacuations are messy, and having non-trained people with physical limitations may cause more injuries. For the A380, Airbus had people on the ground helping the pax leaving the landing area of the slides. This did not prevent one broken leg or arm. During a real evacuation, this is obviously worst due to the fact that the airplane may not be in the usual position. Loss of nosewheel, or ditching are good examples.
    B – evacuations are messy. You cannot predict every possible scenarios because they are almost infinate. So regulators came on with a textbook evacuation certification, and aircraft manufacturers abide by them. The limitations imposed by the regulation are “average really bad luck scenarios”. Worst case scenario ? Doors don’t open, or slides don’t deploy. Can we seriously test something like that ? No. But it can happen. Not sure it did, but it can. That’s the reason begind the half-the-doors rule and the ninety seconds. Do real evacuations take less than 90 seconds ?
    C – evacuations are messy, and part of the reason for that comes from the behavior of the pax. Take one pax, alone, and do the evacuation. No problem. He/she will probably evacuate like in the certification test. But crowds are really different. People react to other people in the wierdest way. People get scared, freeze, get angry, even violent. People won’t even notice they are walking on other people. Some will not care. I saw videos of people walking by fire in stores as if it were normal. Some scientists are trying to identify behaviours, and to recreate them in simulations. And early results show that it’s messy. People in a room with one exit will run to the opposite side of the exit.

    With that in mind, you start to realise that you may not be able to save everyone. That’s a horrible thing to say, but that is the truth. Some people will react in a way that you will end up with people left behind.

    That’s also why F/As are trained to be coercitive. Studies showed that if you give orders loudly with a hard tone, people tend not to question and to follow blindly. That’s what F/As are trained to do. They do not give other options. They even push. People will follow. Regardless of the spoken language. Some will understand the words and/or the gestures, others will follow. Most of them. Some won’t. On the A380, you have one F/A that stands at the top of the stairs to direct people away from them. But if one pax insists, or even goes vilent, the F/A won’t oppose the pax and risk falling down the stairs. The F/A will sidestep and keep on working to evacuate the other pax… and save other lives.

    A lot of work has been done on F/As and evacuation procedures following events (in particular the British Airtour 28M). Even on the doors themselves.

    Keep also in mind that evacuation is the last ressort. A lot of work is done to prevent the situations that would require an evacuation. I’m not denying that there’s room for improvement. There probably is. But I’m not sure there are real unanswered questions here. At least for the aircraft manufacturers.

    Airlines, however, could reinforce the safety messages to the pax, especially new flyers. But they will answer that asking people not to take heels because they could hurt themselves while evacuating is not a good selling point ! But I really think that educating the passengers is essential to decrease the number of injuries or even losses following an evacuation. It’s just like CPR… The more people know how to do CPR, the more people might be saved; the more people know how to evacuate and how to prepare for this unlikely event, the more people might be saved.

    My two cents… @masterdPm

    • Thanks for your extensive and very interesting comment. I don’t think that any of the very valid points you raise are points of disagreement between us:

      A) I’m not actually suggesting that we push grannies down the slides for evac testing, but it does strike me that we are very quiet about survivability for people who aren’t gym-fit. This ties in to a lot of the discussion going on about whether crash test dummies are accurately reflective of the general population. Could we, for example, use mobility limitation suits with trained stunt people to measure the effects? Or attach Fitbit-style accelerometer sensors to evacuation participants to measure the force their bones are subject to?

      B) I agree with you that the scenarios are set up to be less than catastrophic. But it seems to me that there are more of those accidents we’d previously think of as catastrophic (Asiana at SFO, TransAsia at TSA) where people walk away, thanks to the ongoing work of the industry’s learning on safety and survivability. Questions that I’m asking include whether and how we’ve changed testing in recent years to take account of that.

      C) Passenger behaviour is one of the things I’m particularly interested in and concerned about. It is clearly complex to replicate one deeply primal urge — run, jump, survive an entirely unexpected incident — in tests where passengers go into it knowing that they’re going to be evacuating and with the interest of getting the aircraft certified. I also think that there are real questions to be asked about the role of FAs and how passengers react to how they’re presented. Without wanting to put too fine a point on it, there are many airlines that advertise their cabin crew as pliable, come-hither young women at passengers’ beck and call. How have we as an industry examined passengers’ response to that?

      D) I’m popping your points about passengers’ safety knowledge in a separate item, because I think these are really important. In an increasingly globalised world, where air travel is more accessible than ever before, are we using all the tools at our disposal to communicate those messages in multiple languages and cultural contexts?

  2. We are particularly interested in knowing how Airbus plans to get away with raising the certified exit limit for A321 to 240 pax ? This is 20 pax on top of today’s exit limit of 220 pax for the same aircraft. Said score was reached originally in the early 90-ies, with 4 x Type B emergency exit doors on both sides of the fuselage. Present plans are to cancel doors 2R+2L, install twice Type III midcabin (overwing) exits, and to re-position doors 3R+3L some frames further aft …. this is what we shall have.

    The irony in the foregoing is that with an installed THEORETICAL exit limit of 4 x 75 = 300 pax, Airbus achieved a CERTIFIED exit limit of 220 pax. The cabin pitch (with 3+3 seating) was somewhere in the 29″ to 30″ range. So now they intend to achieve 240 pax CERTIFIED exit limit (by 2017 ?) using a cabin fitted with the same (3+3) seatin but superslim, @ 28″ pitch, whereas the THEORETICAL (installed) exit limit with the new doors has been reduced from the previous 300 pax to 290 pax ? We all agree that emergency evacuations are a messy affair but here it could well be that Airbus are presuming too much !

    You have pointed to passenger BEHAVIOUR as random or hectic or unreckonable … I myself offer to see some gregarian PATTERNS in passenger behaviour during an emergency evacuation. One such pattern (for whatever reason ?) is a natural propension of the egress escape flow to pressure onto the mid-cabin (overwing) Emergency Exits. KNOWING THIS, Airbus has actually REDUCED the mid-cabin escapes to 2 x Type III (where the flow will bottleneck) … ?

    To conclude : TwinAisleFeeders wish GOOD LUCK to Airbus with their forthcoming re-certification of the A321 passenger cabin exit limit where we will be taking a close look at the counters after the famous 90 seconds … bookmakers are taking the bets ? I don’t know the odds, but I personally think those counters will stop somewhere around 226 pax … who says more ?

  3. Is there any study regarding speed of evacuation for variable pitch let’s say 30″ vs 32″ vs 34″?

    It’s easy to evacuate first/business class but if people jammed in 28″ economy then it’s becoming bottleneck no matter how many exits airplane has

    • Mary Kirby

      It doesn’t appear so, Nikolay, but we’d certainly like to get confirmation on this point. When one is boxed in so tightly in economy (with seat pitch as low as 28 inches), one has to wonder if egress (and ergo survivability) is impacted. Logic would suggest that it is, but the industry doesn’t seem too eager to discuss. The fact that we humans are growing taller and wider (as our living space on board is shrinking) must also be considered. It seems the time is ripe for fresh egress testing.

  4. I agree with you, Mr Klimchuk : the challenge of a successful emergency evacuation demonstration becomes acuter, the shorter the pitch in economy. In the A321 the aisle is 19″ wide so the aisle pax density at stand-up upon the initiation of the general egress rush to safety is 6 pax in each 19″ x 28″ = 3.7 sq.ft. The row EMF (excuse-me factor) is 6 : both seat extraction jamming AND aisle jamming are unavoidable, as are agoraphobia and crowd panic. On top, those who first rushed towards the mid-cabin (over-wing) location only to discover that the two Type III exits are bottlenecked, will thereafter change direction and move towards 1L/1R or 3L/3R and will fight their way counterwise against the rest of the crowd-flow in the aisle.

    You can install all the exits you can dream of and yet you’ll fail to successfully evacuate the target paxload of 240 pax in 90 s ?!

    I suggest for RGN to set up a poll : in the forthcoming A321 evacuation demo, at what pax-number will the counters stop after 90 seconds ? Reformulated : by what number of pax will Airbus fall short of its target new A321 certified exit limit of 240 pax ?

  5. Its all a revenue game, and safety is the secondary concern. A harsh reality, but that is the real situation. I remember laughing at a B777 evacuation test that only used healthy young and fit Boeing employees.

    More passengers, smaller seats, and tighter pitches, creates more sales of aircraft and eventually more revenue for the airlines. Until and unless regulators step in, both Boeing and Airbus will continue this slide to the bottom. Pun intended.

    My thoughts on Seat pitch vs seat pinch… http://wp.me/p3Ose5-8p

  6. Sean H.

    I know this is a late reply based on when this article was originally published here, but I found it very interesting. I know airlines are now charging extra for the exit seats, but would it not be better to offer those seats to licensed pilots first, aircraft mechanics, and/or other people that have FAA Survival Training. Yes, the FAA does offer survival training, which is free and uses a smoke filled aircraft fusealage for egress and much, much more. Placing people with more knowledge of aircraft are likely to get less of a delay when opening that overwing door and keeping it from being blocked. Lets face it, when the flight attendant asks the person sitting at the exit rows if they are willing and capable of assisting, I would say that most are not, but just don’t want to move or want the extra leg room and if there was an emergency, however unlikely, those people may just be responsible for the deaths of many, just because the airlines wanted to make an extra dime or the person wanted extra space.

    To truly increase safety, put the professionals in the seat. I would imagine that there are plenty of pilots on board every flight whether it be just a Private Pilot or an ATP and they would be much less likely to freak out in an emergency situation. The airline could easily just offer these seats to those that can provide a pilots license, A&P cert, or FAA training card to prove competency.

    Just my thoughts!!!

  7. Jeff Johnston

    Recently while exiting our flight in Atlanta, I noticed at least 10 wheel chairs lined up waiting for passengers…I wondered if there was a limit to the number of passengers who needed assistance that are allowed on a flight? A quick online search revealed the following…”r. Airlines may not limit the number of persons with disabilities on a flight.” Even the discussion above avoids comments about testing evacuation times with some number of individuals with disabilities. Are individuals with disabilities destined to be left behind or slow the evacuation process? JMJ