VALIDATION OF EVACUATION MODEL USING REAL DATA – WHAT IS THE INFLUENCE OF THE COMPOSITION OF THE POPULATION?

摘要

The influence of the composition and heterogeneity of the population of the evacuation times in real fire drills and the prediction of the total egress times using STEPS are investigated in the current work. The evacuated building was a train in a tunnel in Kors?r, Denmark. The experimental study was carried out in 2012 and engaged 100 participants in the age from 6 to 75 years. Besides able-bodied people, the participating population consisted of subpopulations covering hearing, visual, mobility and cognitive impairments. The mean free walking speed was detected for all subpopulations as a mean of the individual speed taken prior the experiments. They served as input to the software STEPS. The total egress time was measured and calculated with the numerical tool. Different setups with varying compositions, blending in and out different subgroups of the population, were run. It was found that accounting for heterogeneity of the population doubled the total evacuation time compared to neglecting it. The predictions of STEPS could be improved by adjusting the speed distribution. The human behavior occurring in the experiment affected the egress times. The simulations did not account for these phenomena.