In this study a novel approach is presented to optimize the layout of petrochemical units, considering the most likely scenario. To illustrate the method a real case study, exporting petrochemical tank farm (Esfahan petrochemical, Iran), was selected. Three objective functions like, piping cost, occupied land cost and human losses were combined and the total cost function was minimized. The required data for lethality function were obtained using simulation of different scenario by Phast software. A sigmoid function was fitted to the data which can estimate the lethality based on the distance from incident point. Genetic algorithm (GA) was used to minimize the total cost function. Due to vastness of searching area, more than one optimum layout was obtained. Finally, the optimum layouts were simulated with fire ball scenario and a good agreement was obtained between optimum points and simulated data.
