The use of probabilistic plume predictions for the consequence assessment of atmospheric releases of hazardous materials

Abstract

Toxic load-based toxicity models have been increasingly applied to model the effects of atmospheric releases of hazardous materials. Although the majority of atmospheric transport and dispersion models predict only a ‘mean’ plume, real-world personnel are exposed to one of many possible individual realisations of a plume, and never to a ‘mean’ plume. Hazard prediction assessment capability (HPAC) model not only predicts the ensemble-mean dosage, but also the dosage variance, making it possible to construct a fully-probabilistic plume. We compare the ‘mean plume’ and ‘probabilistic plume’ approaches to dosage-based consequence assessment using HPAC simulations of a small-scale chemical artillery attack.