# How are evacuation tests made as realistic as possible?

I've seen the video of the A380 evacuation test and similar ones.

While the fast evacuation is impressive, a criticism often heard (and read in the comments on YouTube) is that these people were prepared and (roughly) knew what was about to happen. Also, people in the back of the queue were not in fear of their lives, so they just waited on their turn.
No one tried to go against the stream to look for loved ones seated in different rows, no one tried to get at their personal belongings to try and take those with them.

What measures are taken to make evacuations tests as realistic as possible?

I've heard that incentives are offered to passengers to get out as fast as possible, for instance $100 for the first X passengers to get out. Is that true? If not, anything else? I of course appreciate the conflicting interests that while ideally the test should be as realistic as possible, the test passengers can't be endangered. • Yes, the test is under what some may consider as ideal conditions, however they all are and can be compared to the standard (90 seconds) that way to show that it is possible. An actual evacuation will most likely be slower, but like most things in aviation a safety factor was likely taken in to account when the benchmark was set. If it can't be done in 90 seconds even under ideal conditions, then the problem will be even bigger in an actual evacuation. – Lnafziger Sep 23 '15 at 12:02 • Filling the cabin with non-toxic smoke and adding heaters to simulate fire without telling the participants that was going to happen would certainly help make the testing more realistic. – FreeMan Sep 23 '15 at 19:15 • @MasonWheeler There are plenty of smoke machines in use at parties/concerts/etc that will produce a non-toxic smoke/fog that will make it hard for the test subjects to see where they're going. Something like these – FreeMan Sep 23 '15 at 20:07 • Arguably the passengers on an actual aircraft being evacuated know what's coming to: They do that safety dance...er briefing...before each flight, and if there's time to give the cabin advance warning (e.g. if there's a known landing gear issue) the flight attendants will re-brief everyone. Still it would be interesting to see these tests repeated with "spoilers" added - people who specifically are told to grab luggage, check on someone in another row, etc. – voretaq7 Sep 23 '15 at 20:09 • As the manufacturer has the duty to prove the plane can be evacuated in 90 seconds, you can be sure the passengers have actually been rehearsed to evacuate (but not in the previous 6 months, this is not allowed). That's not a problem however, each aircraft must comply with the 90 seconds limit. An actual evacuation is not expected to be done in 90 seconds. Note that the test is aimed to also test how fast the exits can be opened and the slides can be inflated. All that is done in "night conditions". See Aircraft Evacuation Testing. – mins Sep 23 '15 at 21:05 ## 5 Answers The only ‘realistic’ way of carrying out a full scale evacuation of an aircraft would be to put people in it and set it on fire. That said, FAA takes some efforts to keep the evacuation tests as realistic as possible without seriously endangering the safety of the 'passengers'. The FAA Advisory Circular 25.803-1A - Emergency Evacuation Demonstrations lays out some rules regarding this. • Only half the exits are used for evacuation by a representative mix of passengers, with some debris (luggage, pillows etc) strewn in the path. • The aircraft is to have the maximum number of passengers and flight crew, with a representative mix of age and genders with dolls for children. • Not training any of the participants in the test, i.e. both the flight crew and the participants are not trained to evacuate the (that particular) aircraft. According to the Advisory Circular, The flight attendants should not be trained for specific demonstration conditions, except that specific training should be given which relates to the safety of the participants prior to and during the demonstration. • The participants are clearly told what they are going to do. The prospective participants should be informed of the purpose of the demonstration and the expected duration of their participation. • Though the participants know 'why' they are there, they are not supposed to know 'when' demonstration is going to start. Neither the crew nor passengers should hear or otherwise receive any indication that the demonstration is about to begin. The first indication to participants should be the extinguishing of the normal cabin lighting. • The participants should not know about the evacuation plan beforehand. If safety devices or any other equipment external to the airplane could indicate to the test participants which exits are to be used in the demonstration, passengers and crew should enter the airplane through a tunnel or other means that will prevent them from seeing that indication. • The participants are more than likely paid (employees with previous knowledge of the system are not allowed), though there is no mention of any 'prize' for first getting out. The regulations simply state, Prospective participants should be informed of any direct benefits to them (e.g., pay, meals, etc.) and of benefits to society • The participants are prevented from interacting with outside and to recieve any information from outside the aircraft. ...exterior windows should be obscured from the outside to prevent both viewing of the outside conditions and any ground lighting from shining into the airplane. There is intense debate over how 'realistic' the evacuation tests are. For example, the evacuation of British Airtours Flight 28M, a Boeing 737 took more than 5 minutes (resulting in the deaths of more than 50 people), though the aircraft had passed the test. According to Aircraft Evacuation Testing: Research and Technology Issues, there are multiple problems with this type of testing: • There is only one data point (Full scale evacuation tests are costly; each one costs nearly$1m)
• The conditions do not reflect the actual ones experienced (for example, fire, smoke etc.)
• The participants know why they are there (Employees of the manufacturer are not barred, though they may not sit near the exits).
• The participants do not reflect the actual demographics of the flying public, and if considered, increased the evacuation time significantly (obviously, disabled people and kids are not there in demonstration).
• The participants do not experience any trauma. Of course adding smoke or fire is not an option as already around 4-5 percent people are getting hurt.

Image from Aircraft Evacuation Testing: Research and Technology Issues September 1993 OTA-BP-SET-121 NTIS order #PB94-107620

As a result of the aforementioned BA28M accident, UK's, Cranfield Institute of Technology (CIT) carried out a number of tests which included a case where financial incentives were offered to foster competitive behavior among test participants. The study determined that,

the presence of a competitive element had a significant impact on egress rates for evacuations through the bulkheads, but did not affect the rate of evacuation through the Type III exit

Note: According to FAA, This type is a rectangular opening of not less than 20 inches wide by 36 inches high, with corner radii not greater than one-third the width of the exit, and with a step-up inside the airplane of not more than 20 inches. If the exit is located over the wing, the step-down outside the airplane may not exceed 27 inches.

The US Department of Transportation also carried out studies, which resulted in similar conclusions.

Image from Aircraft Evacuations Onto Escape Slides and Platforms I: Effects of Passenger Motivation, DOT/FAAAM-96/18

The study found that the introduction of smoke did not have much effect, but noted,

Although this research design biased against revealing any deleterious effects of smoke on egress, previous studies using smoke had shown robust reductions in egress speed.

Giving monetary incentives to participants reduced the time required for egress.

Image from Aircraft Evacuations Onto Escape Slides and Platforms I: Effects of Passenger Motivation, DOT/FAAAM-96/18

In this study, the 'competitive' group were offered a financial bonus of \$50 to be among the first 25% passengers to evacuate the arcraft. However, the passengers in the competitve trials

...(became) more aggressive and climbing over seats, outmaneuvering other passengers, etc. to get out quickly.

resulting in broken over seat backs, among others. I suspect participant safety is the main reason FAA doesn't offer any incentives duing demonstration.

Because of these reasons, simulations are being increasingly used in modelling evacuations which include variables like crowd behavior, state (of the person), mobility, etc.

• This is one of those once-a-month-for-the-whole-network answers that really justifies the existence of Stack Exchange by teaching so much more than merely what was asked. Well done! +1 – Todd Wilcox Sep 23 '15 at 18:09
• This answer is also another great example of "Got a question? There's an Advisory Circular for that!" – voretaq7 Sep 23 '15 at 20:11
• I love your opening clause...I was just about to comment on the question "Put a fire under 'er!" – called2voyage Sep 23 '15 at 20:22
• "Put people in it and set it on fire" ... If I see Ashton Kutcher at the bottom of this emergency slide, someone will be hearing from my lawyer – corsiKa Sep 24 '15 at 16:49

IIRC, the passengers that were interviewed in the video you linked to explicitly said that they did not know exactly what was about to happen. They only knew they were going to have to evacuate in the dark with debris on the floor with half of the exits inoperable. They did not know when the evacuation was going to happen; they sat around in a dark plane for an indeterminate amount of time. If your airplane is crashing, you pretty much also know all of that in advance.

Likewise, in a real evacuation, there is an even bigger incentive than a stack of dollar bills to get out fast: your life.

Note that the A380 evacuation certification test actually did lead to one injured test passenger. The most unrealistic part is the vast number of people helping with the evacuation on the ground outside of the airplane. However, those do not affect the speed of the evacuation, only prevent serious injuries (e.g. by getting people off the slides more quickly than in a real evacuation, preventing a possible "pile-up" with passengers slamming into one another at high speed).

Some of them look pretty well disoriented to me in the IR videos.

• Of course, the incentive is bigger in a real evacuation, but that is my point. Passengers in the back will wait in a test, but may be panicking in real life and trying to get out as fast as possible. Also, they may try to look for friends seated in a different row or try to bring personal belongings with them. I'll try editing my question to reflect this. – SQB Sep 23 '15 at 11:49
• Two points to add: There are requirements from the authorities (FAA/EASA) about the distribution of passengers - you must have a certain number of passengers over a certain age, some must be carrying pillows (simulating infants), etc. Also, even though the test participants knew there was going to be an evacuation, they didn't know when it would happen exactly (they were sitting around in the dark plane for a while), and they didn't know which exits would be available. – Lightsider Sep 23 '15 at 12:18
• I'm not quite sure people helping on the ground do not affect the evacuation. You may hesitate a fraction of seconds when seeing the pile of bodies at the bottom of the escape slide may. Multiply the fraction of seconds by the number of passengers for each door and it may slow down the evacuation by few seconds. – Manu H Sep 23 '15 at 13:04
• @ManuH I suspect hesitating in the face of a pileup would only be significant in non-time critical evacuations. If the plane's on fire/sinking the mob pressure'd be too high for anyone to pause at the threshold. – Dan Is Fiddling By Firelight Sep 23 '15 at 14:12
• @Lightsider: thanks, I meant to mention that the "passengers" (nor the flight attendants, I might add) didn't know when exactly the evacuation would happen, but I forgot. It is implicit in the "they only knew", but deserves an explicit mention. – Jörg W Mittag Sep 23 '15 at 15:57

What measures are taken to make evacuations tests as realistic as possible?

• The maximum possible number of "crew" and "passengers" are used.
• There are rules about the mix of ages and sexes.
• The "passengers" are not briefed or trained in advance. They must not have participated ina similar exercise in the last 6 months
• The aisles and exit routes are strewn with luggage I think it is typical carry-on luggage. About half the likely contents of overhead storage in a full plane.
• Half the exits are declared inoperative
• There are financial incentives to get out quickly
• Passengers are prevented from seeing what is happening outside (I think)
• Could you add any references? – Manu H Sep 23 '15 at 16:20

In addition to the other excellent answers, the simulator is equipped with speakers, extra lights, and a fog machine - or at least one airline has such a simulator and I expect others do also. I have done one simply for fun, we were not being timed, and the sound effects, including screaming and weird roaring whooshing noises, plus the orange fire-like light from an unexpected direction and the presence of "smoke" that luckily doesn't smell of smoke all got the adrenaline going and made things more challenging (and therefore realistic) than they otherwise would have been. And we knew exactly what we were doing, we had been offered the treat of experiencing a simulated evacuation.

As for the thing about going backwards - an airline could always instruct someone to misbehave like that, but one of the reasons the target time is set to 90 seconds with only half the exits in use is so that even if such misbehaviours double the exit time, you still get everyone out in 3 minutes. And in real life you have staff who could tell people not to do that.

(Or other passengers: 30 years ago news reports about an evacuation from a soon-to-be-burning plane included tales of the man with the big deep voice who called people over to the exits and urged them to leave during the 90 seconds they got before the smoke turned to flames and those still on board all died, including the folksinger famous for his big deep voice. Two bodies were found far from the exits even though all passengers had moved close to the exits during the descent; it was concluded they had gone past the exits in panic. This is why some airlines instruct you to count rows to the exit from your seat. This flight is also why there are smoke detectors in the lavs.)

May be paid too, in mock drill aircraft manufacturer/airliner wants to show this as added advantage for how fast evacuation can be done in that particular model of aircraft. If they have not paid for this activity, then passengers might not be as quick as you can observe them in video.

Even in software companies, when they do Fire mock drill, you can see employees coming out of the company slowly and record the time with very slowest evacuation.

Coming to point "test should be as realistic as possible" - you may really need to simulate the evacuation (may be accident/emergency) to continue this evacuation

1. In case if accidents
• For passengers, it might not be easy to cross all the possible hurdles created by accident
• Passengers near the evacuation doors are not giving up the way
• Emergency Doors might not be opening as quick as you will able to do it in mock drill
• Accidents will not cause same damage/scenario each time it occurs

With all these reasons you will not be able to evacuate passengers as fast as you do it in mock drills.

I don't know how (on what basis) this particular equipment is certified.

• * passengers grabbing their carry-ons – ratchet freak Sep 23 '15 at 11:43
• This doesn't answer the question, which is how are the drills made realistic. – David Richerby Sep 25 '15 at 7:59