Dynamic analysis of fire induced domino effects to optimize emergency response policies in the chemical and process industry

Abstract

A fire accident is one the most dangerous accidents that may lead to knock-on effects, especially in the plants dealing with large quantities of hazardous substances. Different from other accidents, e.g., explosions, fires can last certain period of time and the thermal radiations emitted by them have a synergistic effect. At the same time, the heat-up of target installations under the effect of thermal radiation is also a process that takes a certain amount of time. These characteristics make fire induced domino effects a dynamic process. During an emergency response due to a fire accident, emergency personnel may arrive at the fire scene at different times, so they may face different accident situations. In this work, an adaptive timed Petri-net (ATPN) based approach is proposed to model the propagation of fires and perform a dynamic analysis of potential domino effects. The definition of ATPN as well as the enabling and execution rules is provided. Through simulations, not only the probabilities of fires in different installations, but also the probabilities of the fire extension propagated over time can be obtained. An example of a tank farm fire illustrates the approach. Our developed approach for carrying out a dynamic analysis of domino effects is helpful for emergency preparation.