Escalation probabilistic model of atmospheric tank under coupling effect of thermal radiation and blast wave in domino accidents

Abstract

Statistical analyses of accidents have suggested that fires are frequently accompanied by explosions in domino accidents at atmospheric storage tank farms. Owing to the complexity of the escalation probability under the coupling effect of multiple accidents, determining an escalation criterion is of considerable significance for high-accuracy risk assessment of an accident chain. In this study, by establishing the propagation mode of domino accidents and calculating their physical effects, an escalation probability model under the coupling effect of thermal radiation and blast waves is constructed based on a dynamic Bayesian network. To verify the rationality of the model, it is compared to Cozzani's model under a single thermal radiation by conducting case studies. The escalation probability at each stage of a domino accident was calculated using the two models; the results indicated that the former reached 303 times that of the latter in 30 min. Finally, the influence of the layout on the domino accident was studied. The results indicated that the maximum deviation of different layouts on the total escalation probability of the domino accident reached 181.4%. Therefore, the coupling effect between different accidents should be considered for domino accidents. Tank layout should be considered during construction, alteration, and expansion.