Robust multi-objective optimization of safety barriers performance parameters for NaTech scenarios risk assessment and management - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Reliability Engineering and System Safety Année : 2023

Robust multi-objective optimization of safety barriers performance parameters for NaTech scenarios risk assessment and management

Résumé

Safety barriers are to be designed to bring the largest benefit in terms of accidental scenarios consequences mitigation at the most reasonable cost. In this paper, we formulate the problem of the identification of the optimal performance parameters of the barriers that can at the same time allow for the consequences mitigation of Natural Technological (NaTech) accidental scenarios at reasonable cost as a Multi-Objective Optimization (MOO) problem. The MOO is solved for a case study of literature, consisting in a chemical facility composed by three tanks filled with flammable substances and equipped with six safety barriers (active, passive and procedural), exposed to NaTech scenarios triggered by either severe floods or earthquakes. The performance of the barriers is evaluated by a phenomenological dynamic model that mimics the realistic response of the system. The uncertainty of the relevant parameters of the model (i.e., the response time of active and procedural barriers and the effectiveness of the barriers) is accounted for in the optimization, to provide robust solutions. Results for this case study suggest that the NaTech risk is optimally managed by improving the performances of four-out-of-six barriers (three active and one passive). Practical guidelines are provided to retrofit the safety barriers design.

Dates et versions

hal-04103505 , version 1 (23-05-2023)

Identifiants

Citer

Francesco Di Maio, Stefano Marchetti, Enrico Zio. Robust multi-objective optimization of safety barriers performance parameters for NaTech scenarios risk assessment and management. Reliability Engineering and System Safety, 2023, 235, pp.109245. ⟨10.1016/j.ress.2023.109245⟩. ⟨hal-04103505⟩
15 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More