Abstract: In order to improve the capability of emergency response and risk prevention and control in the storage tank area of the station yard, the domino effect propagation path under different accident types was determined by comparing the extension vector and failure threshold. The probability of domino effect at different levels was extracted by using the Bayesian network and introducing auxiliary nodes, and the case analysis was completed by taking the storage tank area of a station yard as the research object. The results show that the propagation path of domino effect is the widest when the fire occurs in the storage tank located at the corner of the tank farm and in the upwind direction. When the storage tank located in the middle of the tank farm explodes, the tanks are more dangerous in the event of an explosion, as the overpressure hazard is more widespread. The results can provide a theoretical basis for the formulation of disaster prevention and reduction measures.

Key words: domino effect, Bayesian network, storage tank area, accident probability, propagation path