From Expert Judgment at the Early Design Stage to Quantitative Resilience Curves Using Fuzzy AHP and Dynamic Bayesian Networks

The early design stage is the most effective time to introduce cost-effective measures to increase the resilience of the engineering systems against accidents and disruptive events, yet, since the current resilience assessment methodologies require sufficient knowledge on system characteristics and accidents scenarios, the resilient design is usually overlooked at this stage. This paper proposes a practical methodology for resilience assessment at the early design stage and links the qualitative assessment of system characteristics and expert judgment to a dynamic quantitative resilience assessment. In particular, the resilient characteristics of a system is identified and evaluated by the experts. Then, Fuzzy Analytic Hierarchy Process (AHP) is used to evaluate the contributing factors of the resilient design. Finally, Dynamic Bayesian Network (DBN) is used to make a dynamic mathematical model that represents the system response to disruptive events and integrates the identified system characteristics and expert judgment into a model that quantifies the dynamic resilience curve. The application of the methodology is demonstrated in a carbon capture and storage (CCS) system against the loss of containment accident. This paper presents a feasible methodology for the industries to introduce the system resilience at the early design stage and helps them to design safer, more reliable and available systems.

Keywords: Resilience, Design stage, Dynamic Bayesian Network (DBN), Fuzzy AHP, Expert judgment, Carbon capture and storage.

Download PDF