PrioritEvac: an Agent-Based Model (ABM) for Examining Social Factors of Building Fire Evacuation

Fire evacuation modeling benefits from the application of social science both in terms of accuracy and external validation. This paper describes PrioritEvac, a novel agent-based model which incorporates the social dimension of group loyalty into fire evacuation and responses to fire and smoke. It uses individual priorities, making for a dynamic approach that allows greater agency and nuance. PrioritEvac is programmed in NetLogo and validated using extensive data collected from the Station nightclub fire. The statistical analysis of the results of the model indicate that, compared to historical patterns, it reproduces along multiple metrics including a mean of 114 deaths (std. dev. = 38) over 50 runs, which puts the actual result of the fire within one standard deviation of the mean results of the simulation. Overall, the mean differential along all the metrics is 79, significantly outperforming all published ABM models of the Station nightclub fire that did not incorporate social relationships.