Risk Modeling in Process Industries by Stochastic Petri Nets

Abstract

Modeling the risk to safety of personnel in offshore industry is often realized by the application of Event Trees. The risk is then defined as a product of event frequency and its consequences. However, steady-state methods like Event Trees are not suitable for modeling time-dependent probability of fatality based on time-dependent events. This article presents a new possible approach to modeling the risk to safety of personnel by Stochastic Petri nets (SPN). Stochastic Petri nets models for small leak occurrence on an offshore platform is shown, based on a realistic example from the offshore industry. The probabilities of fatalities were obtained from the simulation by using the Moca-RP software and compared to probabilities obtained by Event Trees and direct Monte Carlo simulation methods.