A data-centered multi-factor seaport disruption risk assessment using Bayesian networks

Seaports are vital for facilitating sheer volumes of maritime transportation and global economies. With their intricate operations, seaports confront a broad spectrum of risks that cause serious disruptions in their functionality. However, very few studies have aimed explicitly at the holistic seaport disruption risk analysis. To that end, this study investigates the impact of wide-ranging interdependent causation factors on the seaport disruption risk. The study draws from literature, 267 accident reports spanning 2010 to 2020, and expert insights to select causation factors and establish their causal relationships. Subsequently, a Bayesian network (BN) model is developed to evaluate seaport disruption risk, with probabilities quantified based on the collected accident reports. The evolution of seaport disruption risk scenarios and the interdependency of the considered risk factors is reflected through forward and inverse inferential analysis in the BN model. The results reveal that without setting evidence at any particular condition, weather extremities, workforce disputes, and port congestion were the highest contributors to the occurrence of seaport disruption risk. A sensitivity analysis was conducted to identify the most critical disruption causation factors. Based on the insights, the study proposes plausible safety procedures to mitigate the incidence of seaport disruption and circumvent the consequence severity.