In order to improve the applicability of probabilistic estimation methods for process safety accidents

the propagation process of deviations should be considered in the basic event stage. A method for estimating the probability of process safety accidents based on scale effects is proposed. A new basic event solving model is established from the point of view of process deviations. Multi-scale ideas are introduced to modify the basic event probability

on the large scale

by considering the effects of human errors on process deviations. The fuzzy Petri net model is used to estimate the accident probability. Finally

the model is analyzed and verified by the case analysis of the fractionating tower accident. The results show that the participation of human has a great influence on the probability of process deviations. The proposed method is more suitable for the practical conditions and process safety. Since the probability estimate is based on the statistical data

the deviation probability is calculated and the subjectivity of the calculation parameters is avoided.