Resilience-based design of barrier system to mitigate fire-driven escalation in process plants

Abstract

Fire-driven escalation is a typical type of domino effect, which may lead to severe consequences in process plants. A barrier system comprising by add-on barriers and procedural barriers could effectively mitigate fire-driven escalation, but its optimization poses great challenges to safety managers. In this paper, a resilience-based methodology is proposed to identify the optimal design scheme of barrier system considering the cost-effectiveness. Possible schemes of barrier system under a limited budget and their effects are explored. Dynamic Bayesian network is used to model the evolution patterns of fire escalation, while the dynamic uncertainties associated with accident evolution are quantified. Resilience, a metric that could capture the dynamic change of plant performance in the accident and recovery period, is employed and quantified to determine the barrier system with highest profit on fire-driven escalation mitigation. A case study of a tank farm is applied to illustrate the applicability of the proposed methodology, in which different design schemes of barrier systems for increasing resilience are discussed and some recommendations are given. The developed methodology directly associates safety budget with plant resilience, which could help safety managers and decision-makers of new or existing chemical plants identify the optimal scheme of barrier allocation addressing fire-driven escalations.