Conference Paper

Probabilistic Modeling of Failure Domino Effects in Chemical Plants

PDF Online

Authors Diego SierraJuan BriceñoHector BuitragoBrian RozoLeonardo MontecchiIvan Mura
Abstract
Chemical process plants are subject to risk caused by the handling and storage of hazardous substances. Major accidents may occur in this kind of facilities, particularly in those unfortunate circumstances when a triggering event produces a cascading accident that propagates to other units, a failure propagation scenario known as domino effect. An important aspect of designing such industrial plants is to properly arrange hazardous equipment such that, in the event of failures, cascading effects are minimized. In this work, we present a modeling approach to perform a probabilistic analysis of the likelihood of domino effects caused by propagating vapor cloud explosions. The approach combines the modeling of accident propagation based on physical properties of gas clouds, such as released mass and explosion distance, with the probabilistic modeling of cascading effects based on Stochastic Petri Nets. The proposed methodology is subsequently applied to a case study where different layouts of atmospheric gasoline tanks are analyzed, in order to evaluate the effect of this factor on the domino effect.
DOI 10.1109/LADC.2018.00016
Event 8th Latin-American Symposium on Dependable Computing (LADC 2018)
Venue Foz do Iguaçu, Brazil
Date October 8-10, 2018
Pages 57-66
Publisher IEEE
ISBN ELECTRONIC: 978-1-5386-8489-4
ONDEMAND: 978-1-5386-8490-0
 
 

© 2017-2019 Leonardo Montecchi