In the first part of this article I wrote that some contractors were reducing integrity/insulation advanced glazing from 60 to 30 minutes.
Some, indeed, are specifying for 30 minutes of integrity only.
This means that the fire will be contained, but not the radiant heat.
In a serious fire, that heat transfer could set off secondary fires. In that worst-case scenario, the whole purpose of containment would be nullified.
The justification for this reduction in fire safety standard is that the building can easily be evacuated within 30 minutes.
Well, maybe. But the real reason, of course, is cost. A glazing system providing 30 minutes of protection costs less than one giving 60 minutes of protection. And an integrity only system is cheaper than one for integrity and insulation. For safe evacuation there is a real need for integrity/insulation advanced glazing.
But while I understand the cost argument, I challenge the justification itself.
The reason being that we are human beings and don’t always behave in line with computer simulations.
The contractor or developer may assume that a building’s inhabitants will swiftly move to an exit as soon as the fire alarm sounds.
But assumptions are never a good foundation for fire safety.
For example, a major UK retailer recently had a series of unannounced fire test evacuation drills.
But contrary to their training, staff didn’t immediately move customers towards the exits.
The majority of staff paused to confirm the fire alarm was real, and therefore delayed evacuation.
In other words, human behaviour can sometimes work against the fast evacuation of a building.
This is backed up by research from the US National Institute of Standards and Technology (NIST).
They found that computer models don’t always take human reaction into consideration.
Because, when a fire alarm sounds, our first instinct is to look for evidence of a fire.
If we can’t see any evidence, we assume it’s a fire test or a false alarm. Until we know for certain, we do nothing.
That period of inactivity is called “pre-movement time.” Our brains are telling us that there is no discernible threat. Therefore, the chances of it being a real fire are remote.
In most cases, a fire alarm is simply a source of confusion. And even when we believe the alarm might be real, we don’t know where the threat is coming from.
Nobody wants to evacuate until we know that our escape route is safe. In other words: more delay.
It’s estimated that as much as two-thirds of the time it takes people to exit a building after an alarm is start-up time. That’s time wasted looking for more information.
The most extreme example was 9/11. Fewer than 9% of the occupants of the World Trade Center immediately evacuated after the alarms sounded.
The average start-up time before people began to move to the exits was between five and eight minutes. Others didn’t start to evacuate for up to 40 minutes.
But even people do evacuate, there’s another complication – exit choice behaviour.
Again, this is something that computational models struggle with because people don’t always exit a building by designated routes.
They’ll exit a building by a familiar route, even it’s a longer journey. For example, the route by which they arrive at their desks in the morning.
But behaviour isn’t only about individuals. Groups of people can influence one another. People might want to evacuate with their colleagues, slowing the entire process down.
In a shopping centre, evacuations might involve families with elderly relatives or small children.
All of those factors, and many others, influence the level of protection that should be applied within a particular building.
It simply shouldn’t be about using a computer model to estimate how long a well-drilled evacuation will take. It must also take into account the vagaries of human behaviour.
And that’s why I question the wisdom of reducing fire safety specifications of integrity/insulation advanced glazing from 60 to 30 minutes, or even removing the need for insulation.
In a complex building, that could spell tragedy.
This is the second of his two-part article, Denis Wright, our chairman, looks at why evacuation times can be underestimated.