Assessment of human contribution to cargo ship accidents using Fault Tree Analysis and Bayesian Network Analysis

Abstract

Maritime accidents are significant contributors to human casualties, environmental damage, and economic losses. This study examines the role of human factors in very serious maritime casualties involving cargo ships, employing Fault Tree Analysis (FTA) and Bayesian Network (BN) methodologies. Using data from the European Maritime Casualty Information Platform (EMCIP) for incidents between 2010 and 2020, the analysis focuses on 60 maritime accidents linked to human actions. FTA displays causes related to human error that can lead to accident, while BN models the relationships and dependencies among Risk Influencing Factors (RIFs). The combined approach enables a comprehensive evaluation of system risks, highlighting key contributors such as shipboard operations and shore management practices. The study also explores minimal cut sets and mutual information to assess the influence of environmental and operational factors on accident probabilities. Results indicate that factors like crew resource management and workplace conditions significantly affect the likelihood of casualties. Scenario analyses further demonstrate the dynamic interactions between RIFs and their impact on maritime safety. This dual methodology provides actionable insights for improving risk management strategies and reducing human error in maritime operations, offering a robust framework for enhancing safety in the shipping industry.