An Integrated Dynamic Bayesian Network--Based Framework for Risk-Informed Hazard Ranking: A Case Study of a Thermal Power Plant

Hazard prioritisation in thermal power plants remains a complex challenge due to dynamic risk interdependencies, probabilistic uncertainty, and ranking inconsistencies arising from conflicting multi-criteria decision-making (MCDM) approaches. This paper proposes an integrated Dynamic Bayesian Network (DBN)-based framework for risk-informed hazard ranking that combines temporal hazard modelling, entropy-guided multi-criteria evaluation, and consensusbased prioritisation. The DBN captures time-varying hazard interactions under cascading failure conditions, while Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), VIKOR, and entropy-based ranking methods provide complementary perspectives for hazard evaluation. Inter-method ranking disagreement is systematically quantified and resolved through a consensus prioritisation mechanism that minimizes total pairwise disagreement across input rankings. The framework is validated through a case study involving ten critical hazards in a thermal power plant, achieving a pairwise majority agreement ratio of 0.978, which outperforms individual MCDM-based rankings. The proposed framework offers a robust decision-support tool for early-stage hazard prioritisation in industrial environments characterized by limited and uncertain data.