I attended a conference the other day that dealt with fire behaviour research in buildings put on by the National Research Council of Canada. That is, how does fire actually behave, how do people react to fires, etc. These are my notes from just one of the sessions and even at that, it is only a fraction of the information presented. It is also condensed and reorganized. Overall the conference was very informative and interesting. Unfortunately, the formatting and the lists are a bit off because this forum is limited in its abilities.

The speaker (Guylene’ Proulx) provided the example of the nightclub fire (Great White concert) in which a video camera recorded the participant’s behaviour in the building. Note the capacity of the building slightly exceeded the limit. Almost all the people tried to exit from the doors they used to come into the building even though other exits were available. A closer exit had an illegal door swing and part of the building had a dead-end corridor, which did not come into play during the exit. Due to a changeover from a restaurant to a nightclub, they were required but did not have sprinklers in place. Pyrotechnics were used next to flammable sound foam that was installed to counter complaints from neighbours about the noise.
http://en.wikipedia.org/wiki/The_Station_nightclub_fire
Elapsed Time Event
0 Pyrotechnics activated
9 s Flame visible on wall
19 s Flame progress on wall
20 s Crowd started to react
25 s Flame at ceiling
30 s Band stopped playing
41 s Fire alarm and strobes activated
60 s Smoke at ceiling level throughout room
71 s Cameraman left by front door (One of the first out. Strangely, that cameraman from a local TV station was also a part owner of the club so he was cited in a conflict of interest charge.)
102 s Occupants piled up at the front door and could not exit.

The crowds reacted quicker than in the typical office-building scenario because they could actually see the fire and did not need to get confirmation from others that it was an emergency.

Human behaviour in a fire: Contrary to popular belief, people do not exhibit panic while responding to a fire (only 2%). In fact, people can and do exhibit altruistic behaviour. If a person trips and falls, someone will try to help them up. If a handicapped or very slow person is blocking the stairwell, people will bunch up and proceed at the speed of the slowest person. People will attempt to calm each other down.

The most common behaviour is the absence of any response before committing to an action. Once a warning has been made, people tend to seek affirmation that an event is actually happening – they talk to their neighbours, poke their head up over their desk or around the corner. People tend to want to finish phone conversations or finish typing a thought on their computers before realizing that something is happening. Then they take their time to shut down computers, gather their Blackberries, put away important files, grab their purse, put on their coats, etc. before they actually get up to leave. In a high-rise office evacuation drills, they observed that the pre-movement time to be approximately 1 minute before the occupants began to evacuate. This is in a building where people are generally prepared and practise their fire drills.

Other observations: people who live/work in the building might attempt to put out a fire but visitors to the building (especially theatres or movie house) never attempt to do so probably because of their non-familiarity with the building.

You would think that in a fire people would just go to the nearest exit. This is not necessarily true because people who are separated from their loved ones will often try to go back into a fire area to reconnect with their loved ones, especially if they are separately from their children. This creates a counter flow to the exit pattern, which bunches people up and slows the exiting process.

After the 1993 bombing in the lower levels of the World Trade Centre, the speaker was involved in analysing the exiting from the buildings and after 9/11 was also involved in interviewing the survivors. It took approx 1 minute of pre-movement time before people began to react and a minimum of 10 more minutes before people made the actual decision to leave the building whereupon it took an average of 2.5 hours to evacuate. The pre-movement times were estimated:
WTC Tower 1 (1993) 11 min. 2 sec.
WTC Tower 2 (1993) 25 min. 24 sec.
WTC Tower 1 (2001) 6 min. 0 sec.
Cook County Fire (2003) 5 min. 0 sec. (even after voice communication over speakers)
In the post-collapse of the 2001 buildings, they interviewed over 1000 people to get feedback on what the problems were in exiting the building. One of the prevalent activities was that people could feel the building shake but then took time to look around the corner, check stairwells, look out the window, smell for smoke, etc.

In an apartment building the reactions times are considerably slower. They conducted unannounced fire drills (Thursday at 1930hrs to maximize the number of people in the buildings). They found that the pre-movement time varied from 2.5 minutes to over 9 minutes but averaged 3 minutes, which is considerably longer than the office building experience. The elderly and very young took the longest to mobilize but not drastically different from able adults. It was disturbing that many people did not hear the fire alarm due to the distance from the alarm, location of the alarm was in the hallway not the suite, TV/radio noises, people talking, etc. or just plain did not recognize it as a fire alarm. The testers also had the emergency first responders show up with full gear and flashing lights to make the experience as real as they could. Part of the difference between an office building and a residential tower is that most offices have regular drills and practise leaving the building whereas they do not in residence.

Some factors that affected the response of people to an emergency:
a) Familiarity with the building especially with assembly buildings.
b) Training or lack thereof.
c) Looking at others and the need for a leader to actually start to do something.
d) Following instructions, especially from those who “look” official.
e) People tend to want to use cell phones and Blackberries while in the stairwell to contact others.
f) Evacuees who may be out of shape or overweight, wheelchair users, asthmatic people, all of which slows the ability of the masses to use the stairwell.
g) Women with high heels tend to clog up stairwells.
h) Stairwells that have incurred damage, have water or debris in them, lack signage or proper lighting will slow down the exiting process.

Recommendations:
a) Do not have exits lead to an underground garage even if the garage is a safe area that has an exit to the exterior.
b) Do not build double loaded stairwells. (Architects like to design the “building cores” to maximize rental space but by doing so, damage or cross air contamination can shut down both stairwells!)
c) People will exit from the way they came in unless they are directed to another exit by instructions or by a leader.
d) Practise fire drills in all buildings and building operators should have a fire plan in place with the necessary people to enforce it.
e) Review stairwell size and design and match with crowd densities (2.3 people/m2).
f) Put refuge areas in places other than the stairwell because stairwells (especially upper levels) gather and funnel smoke. Use protect-in-place strategies with better and more fire/smoke resistant compartments.
g) Use a refuge floor in very tall buildings.
h) Consider having a designated “safe” elevator, which has extra barriers to prevent migration of smoke and fire.
i) Civilise and smooth the merging of people from a floor into the traffic in the stairwell.
j) Change the codes to demand more use of photoluminescent marking in exit routes.
k) Use phased evacuation in tall buildings.
l) Have good clear voice communication by loudspeakers and be specific with the instructions.
m) Design for efficient movement in stairs (see above).
n) Inter-floor communication systems for fire fighter use only because often their radios cannot penetrate concrete floors and structures. (Lessons from a firefighter).
o) Leave fire extinguisher in but take out the fire hose out the building fire cabinets because occupants are taught not to use them and even if they did use them, they would not know how to fight a fire with a 1.5” hose pushing water at 60-80 psi. The fire fighters don’t trust the building hoses because they may be never be checked or stored properly so they usually bring their own. Instead, install as many standpipe attachments in the building as possible based on distances to areas. (Lessons from a firefighter).

p) Building Emergency Action Plan should have:
i. Description of the occupants.
ii. Description of the building.
iii. Description of emergencies.
iv. Actions to be taken by building management; emergency team, floor warden and occupants.

q) Train, practise the Plan, assess the plan to see if it works.

Overall, people are rational and can think for themselves. They are able to make decisions based on information available. They will attempt to preserve themselves and their loved ones. They are helpful, courageous and co-operative. They are not fire experts. They lack experience with emergencies. They will use a familiar plan of action. They will follow instructions. They are relatively cool and composed; the trauma comes after the event.