ScottishArchitecture.com

The real test of anything we build is not aesthetic, practical nor even economic – but what happens in an emergency.  In extremis, after a serious fire or explosion, the structure must hold together long enough to allow people to escape.  However, whether they get out safely, or not, is down to human nature as much as building design … vehicle design … or indeed aircraft design.

In order to “type certify” a new airliner, trial evacuations are carried out.  The testing of the Airbus A380 – the SuperJumbo – was the most recent, during which an airframe parked inside a hangar at Hamburg was fully loaded with people.  In this case, 853 passengers, plus cabin crew.  When the command to evacuate was given, the aircraft was emptied in an astonishingly fast 78 seconds.  For the purposes of the test, a regular Lufthansa crew was in charge; some smoke and loose objects had been introduced into the cabin; it was dark (although the emergency lights were working); and some exits had been blocked off.  The speed of the passengers’ egress wasn’t down to Teutonic efficiency alone, though – the guinea pigs were well briefed beforehand, and had time to consider the best way to escape.  Tellingly, they co-operated with each other, as they were not in mortal danger.

Most people treat all alarms as false alarms, until proven otherwise – just watch any building site once it’s near to completion, and each time Kidde, Minerva or ADT set off the alarm, workmen come sidling out long after the sirens first began to sound.  There’s an air of unreality about it, especially when it proves to be yet another false signal, or system test.  Yet once people believe they really are in peril, which often takes longer than it should, the alarm instills a sense of panic into their behaviour.  Sounding the tocsin goes back to prehistory, when the great war horns gave a warning of trouble: Fear–Fire–Foes.  In medieval times, the pealing of the cathedral’s bells warned the city when it was about to be laid siege to.  Ring the bells – Wake the town – A call to arms, as the song lyrics have it.  That led in time to the banshee screaming of the air raid sirens during modern wars, then the klaxons alerting RAF crews to scramble in the ‘60’s: the Three Minute Warning had sounded.  Very early in our lives, a connection is made between alarms and danger: self-preservation is our deepest instinct, and will over-ride everything else.

It’s the difference between our responses to a practice run, and the real thing, which is almost impossible to replicate.  That’s where evacuation tests on aircraft, and the fire drills of the sort we all experienced at school and university, fall down.  They can’t represent the terror of a genuine emergency, simply because the mind isn’t adept at self-deception.  It operates in a unified way, so if the higher rationalising part knows this is just a drill, then the response of the more primitive instinctual response will be muted.  Words are really inadequate to describe what happens when you do have to flee a building, though.  The air fills with alarms, instinct kicks in, and you concentrate on only one thing – the brain suspends any functions which aren’t critical to escaping.  Adrenaline takes over and you flee.  The advice about walking calmly towards an exit means nothing when real danger is close at hand.  You move as fast as you can, but afterwards you can’t recall any detail of that 30 seconds, which subjectively felt like a lifetime.  The routines hard-wired into us succeeded – we escaped, we survived to tell the tale.  Yet sometimes things turn out differently.

In the wake of the 1985 accident at Manchester, when a British Airtours Boeing 737 suffered an engine fire on take-off and 55 people died in the resulting crash, Cranfield University made a detailed study of aircraft evacuation.  Critically, it took five-and-a-half minutes for the last passenger to emerge from the burning B737 at Manchester Ringway – the aim of the research was to find out why.  The researchers used a retired Hawker Siddeley Trident, plus a long line of cash-strapped volunteers.  Uniquely, most of the participants were students, who were paid £10 to turn up, with the promise of another fiver each time they succeeded in being among the first few to escape from the plane.  The cash bonus was handed over as soon as they reached terra firma.  The professor conducting the experiment judged that the unholy mixture of the students’ natural competitiveness, and the promise of hard cash, would prove “as compelling an incentive to escape as life itself”.  Or almost so …

You can imagine the Pavolvian reaction when the stewards called on the passengers to evacuate – “the desperation to escape quickly was quite alarming as volunteers battled to be the first through the exits,” wrote Max Kingsley-Jones in the magazine Flight International.  People were carried along with the throng, crushed under seats, wedged in the aisles and caught in the corners of the cabin against bulkheads.  While the Airbus trial achieved a rapid evacuation, thanks to the passengers co-operating with each other and escaping in an orderly manner (“women and children first”), the Trident trial was a closer reflection of reality.  Although it was carried out in the late 1980’s, and has never been repeated, the trial was closely examined by UK authorities such as the Civil Aviation Authority.  The fact that the Germans carried out the A380 trial as they did suggests that they weren’t paying attention, as they didn’t come across either the panic reflex, or the other extreme, abject resignation to your fate.  Sometimes people just give up, and huddle in a corner to await their fate.  Both those reactions are illustrated by Dad’s Army, that popular TV re-enactment of World War 2: when trouble came along, Fraser resignedly exclaimed, “We’re aa doo–oomed!”, whereas Jones cried out, “Don’t panic! Don’t panic!”

One way around the problems is methodical training.  Although it seems that occasional fire drills don’t prepare us to face disaster, over-familiarity with crisis situations does seem to work for firefighters and airline pilots.  In fact, a large proportion of an airline pilot’s training is devoted to preparing him for emergencies, in order to make his responses as automatic as possible.  Several hours are spent on the simulator every month or two, practicing stall recovery, flame-outs, forced landings and other dire situations: the intention is that the pilot “over-learns” the skills needed, because the shock of the emergency may well diminish his ability.  Secondly, over-learned responses and realistic situations give the pilot confidence to stay calm: but however realistic the simulator, the emotional shock of a genuine emergency is still missing.  Psychologists have understood for decades that the brain doesn’t function well when overloaded with stimuli, and the tragic illustration of this is a passenger trapped in the blazing wreckage of an aircraft who continues to struggle with an unyielding emergency exit, yet ignores the gaping hole in the fuselage close by.  The brain is swamped, causing the person to fixate on one thing to the exclusion of all else.

More recently, both aeronautical and architectural fire engineers have begun to use software modelling to replicate evacuations.  For the building project I’m currently running, a computer model representing 12,000 sq.m. of floorplate, and 1150 people, was created by SAFE Fire Engineering in Glasgow.  The actual evacuation sequence is like a scene from a Lowry painting brought to life: but the matchstick people behave differently each time, as computer algorithms try to take account of the randomness of human behaviour – panic, confusion, reactions to other peoples’ irrationality, and the heat, smoke and toxic gases.  The software’s ability to run evacuation scenarios over and over again generates an “envelope” of performance, rather than a single datum, providing a better representation of real life.  Software also has the advantage over full-scale aircraft certification trials that the latter cost £1million a time and volunteers are sometimes injured, or worse.  It does rely on the software being suitably sophisticated and nuanced that it can predict how fickle humans react in an emergency, though.

Fire engineering is a specialist field, and only a small proportion of buildings benefit from it.  For the rest, architects rely on the prescriptions of the Technical Standards to guide them on how the building should assist people to escape from a fire.  Yet are we, or the people who write the Standards, any closer to understanding how people react?  That strange mix of crowd psychology, brain chemistry and self-preservation – how will that turn out, when the VESDA sensors first sniff out smoke, then the sounders are activated, zone by zone, and the alarms grow louder?  The corridor smoke doors swing shut, the power goes off and the emergency lights glow on.  It’s not a drill this time.  It’s for real.  How will *you* react?