Dynamic Bayesian network-based resilience assessment for a maritime LNG transport system

Abstract

The attack on Qatar's LNG facility, which incapacitated 17% of its production capacity for an extended period and triggered a precipitous escalation in European gas prices, highlights the critical importance of the LNG supply chain to global energy security. However, these complex systems are susceptible to severe disruptions that can trigger catastrophic operational degradation. While previous studies have predominantly relied on static risk assessment, quantifying dynamic recovery trajectories under extreme shocks remains imperative. This paper proposes a quantitative, time-varying resilience assessment framework for an integrated maritime LNG carrier system, explicitly addressing the interplay between technical reliability and external disruptions. The methodology integrates a Dynamic Bayesian Network (DBN) with Markov chains to shift the analytical focus from static failure probabilities to dynamic resilience evolution. By tailoring the 4Rs resilience concept (robustness, redundancy, rapidity, and resourcefulness) to the unique operational constraints of LNG, the model accurately captures the rapid hazard escalation that generic shipping models overlook. The evaluation encompasses diverse disruptions, centering on Natech events, while concurrently accounting for human factors, geopolitical risks, and intentional attacks. A case study of the maritime shipping route between Ras Laffan (Qatar) and Yangkou (China) is conducted, demonstrating the model's efficacy in predicting dynamic recovery trajectories. Ultimately, by quantifying the specific impact of individual operational nodes, the proposed model translates sensitivity results into practical operational measures, providing a reliable predictive tool to enhance long-term maritime LNG resilience under complex disruption scenarios.