What’s the best way to evacuate a skyscraper?

Although these buildings have been around for well over a century now, no clear answer has emerged, largely because there is no easy way to calculate an answer.

The simplest models can determine the number of people that can pass down a stairway in a certain time, like an incompressible liquid flowing through a pipe. In theory, at least.

But in real emergencies, all kinds of other factors come into play. In particular, people behave in strange ways that are nothing like liquids. When panic occurs, the flow can drop to zero as people block doorways or become so compressed that they stick fast. The result is a kind of phase change when the flow changes dramatically from one regime to another.

One way to avoid this is to ensure that the density of individuals never increases to the point that triggers any kind of phase change in the flow behaviour.

In recent years, computer models have emerged that can simulate the kind of pressure people are under. Today, Wieslawa Sikora and chums from AGH University of Science and Technology in Krakow, Poland, use such a system to simulate an interesting new evacuation strategy.

One problem is that when people are evacuated from all floors of a building at the same time, the stairs can become clogged, raising the likelihood of jamming. A possible solution is to evacuate one floor at a time but this leads to delays that could be fatal.

So Sikora and co have simulated a staggered evacuation to see how it compares to a simultaneous one. They say that although the staggered evacuation takes longer, it dramatically reduces the “social forces” that individuals experience.

That makes it worth considering the trade off in more detail. Clearly, panic or stampeding behaviour is in nobody’s best interest during an evacuation. So anything that can prevent this is likely to make any evacuation better.

One unanswered question is how well such models simulate people’s real behaviour in an emergency. It’s well known that people behave entirely differently in test evacuations than they do in real emergencies where their lives are at risk.

So whether a definitive answer is possible with the kind of computer models available now is a point worth discussing.

Ref: arxiv.org/abs/1108.3702: Model Of Skyscraper Evacuation With The Use Of Space Symmetry And Fluid Dynamic Approximation