Integrated Fuzzy-MCDM and Fault Tree Analysis Framework for Critical Failure Assessment in LNG Cargo Operations

This study integrates the Fuzzy Delphi Method (FDM), fuzzy DEMATEL, Fault Tree Analysis (FTA), and controlled simulator experiments to assess critical failure mechanisms in liquefied natural gas (LNG) cargo operations. FDM identified critical risk factors, fuzzy DEMATEL characterised their causal interactions and an expert-validated fault tree represented Boolean failure pathways using normalised model weights specified by the authors and approved by the expert panel. Ten participants completed one nominal and five abnormal scenarios using the TRANSAS LCHS 2000 LPG/LNG simulator. Instrumentation and Sensor Errors (C2.4), Overfill and Sloshing Risk (C3.4), and Cargo Pump and Compressor Failure (C2.3) were the strongest net driving factors, with R-C values of +0.9617, +0.9254, and +0.9100, respectively. C2.4 and C2.3 each formed a first order minimal cut set and had the highest Fussell–Vesely importance values, at 0.476833 and 0.390260, respectively. The top event index of 0.178359 is a semi-quantitative measure of functional failure in LNG cargo operations, rather than an empirical accident probability. Escalation to harmful consequences depends on subsequent conditions and the performance of protective barriers. Operational performance deteriorated under abnormal conditions, with the poorest observed performance in the combined scenario (S5). However, these findings do not isolate interaction effects, as scenario severity and workload were neither independently matched nor quantified. Causal influence, structural importance, and operational performance thus provide complementary perspectives on risk. The proposed Barrier-Based Dynamic LNG Cargo Safety Management Model (BB-DLC-SMM) translates these findings into prevention, monitoring, detection, emergency response and recovery measures.