DIRIS : utilisation des diagrammes d’influence dans l’analyse des risques pour les installations industrielles

Abstract : Any industrial system is constrained by regular human interventions conditioned by organizational decisions, which must be taken into account by the risk analysis. The research led in the DIRIS project aims at developing a methodology, a tool and a «risk model» that can be adapted to case studies. Our work considers a probabilistic frame and focuses on the modeling of safety barriers, that are key elements including human and organizational influences. The «bayesian network» was chosen, because it combines the calculation of probabilities with event trees represented in a graphic manner. It integrates all the probabilistic links within the network, and deterministic links can be introduced. We defined a generic «risk model» composed of three levels: a technical level, a level of human actions, and an organizational level. The last one is represented through 7 «pathogenic organizational factors» (POFs), such as «weakness of control» or «production pressure». This model is then applied to a real industrial site, with the study of the explosion of a vessel. The probabilistic links between the variables are modified by quantitative degradation coefficients that take into account upstream influences. This industrial application offered the opportunity to develop bayesian networks as replicable tools, and proved the feasibility of the whole method.
Document type :
Journal articles
Complete list of metadatas

https://hal-ineris.archives-ouvertes.fr/ineris-01869251
Contributor : Gestionnaire Civs <>
Submitted on : Thursday, September 6, 2018 - 1:04:52 PM
Last modification on : Friday, September 7, 2018 - 1:11:29 AM
Long-term archiving on : Friday, December 7, 2018 - 5:33:44 PM

File

2009-451.pdf
Publication funded by an institution

Identifiers

  • HAL Id : ineris-01869251, version 1

Collections

Citation

Régis Farret, Jean-Christophe Le Coze. DIRIS : utilisation des diagrammes d’influence dans l’analyse des risques pour les installations industrielles. Rapport Scientifique INERIS, 2009, 2008-2009, pp.74-76. ⟨ineris-01869251⟩

Share

Metrics

Record views

9

Files downloads

11