The aim of this paper is to present a probabilistic framework useful for the risk assessment of refinery plant components located in areas prone to seismic hazard. The classical quantitative risk analysis (QRA) in fact allows to evaluate a mean frequency value of a potential release as a function of the fault tree diagram without taking into account uncertainties related in the estimation of temporal occurrence of each singular initiator event. In addition, natural hazards are not frequently considered in QRAs due to challenges related to the evaluation of their consequences in the chemical process industry. For these reasons, this work presents a probabilistic framework to be applied in the fault tree analysis (FTA) to estimate the occurrence probabilities of small, medium and large release from an atmospheric steel tank belonging to a chemical plant located northeast Italy, a national area particularly prone to seismic risk.
