A risk analysis model for Arctic escort operations based on STPA and Bayesian Networks

Abstract

This study proposes a risk analysis model integrating Systems-Theoretic Process Analysis (STPA) and Bayesian Networks. The model comprehensively incorporates environmental, vessel, human, and organisational/management factors to assess the risks of collision between the escorted vessel and the icebreaker, collision between the escorted vessel and ice, and besetting of the escorted vessel in ice across the five sea areas along the Northeast Passage. The results demonstrate that the proposed model effectively captures the distribution characteristics of risks associated with icebreaker escort operations across the various sea areas. Consequently, it can serve as a valuable tool for guiding icebreaker service operations and navigational decision-making. Furthermore, this study confirms that August and September represent the optimal navigation period along the Northeast Passage. Compared to the other sea areas, the East Siberian Sea exhibits substantially higher risks during icebreaker escort operations. Ice thickness, ice concentration, navigation experience, emergency response capability, vessel speed, propulsion and manoeuvring system performance, ice condition information services, and emergency response and rescue plans are identified as the primary risk factors influencing icebreaker escort operations in this region.