FBN-based analysis of operational risks in marine compressed air systems

Compressed air systems are vital to maritime vessels, supporting key functions like engine starting, control systems, and emergency protocols. Failures in these systems can lead to serious operational, safety, and environmental issues. This study proposes a Bow-Tie-based Fuzzy Bayesian Network (BT-FBN) method to conduct a dynamic, equipment-specific risk assessment for marine compressed air systems. By combining the visual structure of Bow-Tie analysis with the probabilistic reasoning of Bayesian Networks–enhanced by fuzzy logic to handle uncertainty–the model focuses on compressor failure as the top event. Using real operational data and expert input, the study examines root causes, potential consequences, and the effectiveness of control measures. Results show that operational errors, limited technical knowledge, and poor maintenance are the primary contributors to failures. Sensitivity analysis reveals their significant impact on system reliability. The model's accuracy and reliability are validated through a three-axiom approach. Overall, the study introduces a comprehensive, data-informed framework for assessing and managing risks in marine compressed air systems. The BT-FBN approach offers practical insights for enhancing maintenance practices and improving the safety and efficiency of maritime operations, with potential for broader application to other critical shipboard systems.