Ship-Bridge Allision Risk Analysis - AIS-based Failure Event Modelling and Consequence Assessment

Doktorsavhandling, 2025

Increases in ship sizes and in maritime traffic frequency pose a growing risk of ship-bridge allisions. Although relatively rare, such accidents can have catastrophic consequences, including significant economic disruption and loss of life. However, constructing robust bridges capable of withstanding allisions with large ships travelling at high speeds is costly. Therefore, to balance risk and expense, it is crucial to conduct a thorough risk assessment. Traditional risk assessment methodologies often separate the probability analysis from the consequence analysis, but this separation leads to generalisation, less accuracy, and the introduction of overly conservative assumptions, and it may not reflect local traffic patterns or ship manoeuvring failure behaviours. This thesis aims to develop a novel, unified analysis for assessing the risk of ship-bridge allisions by integrating local ship manoeuvring failure statistics, probabilistic modelling, and consequence analysis into a single methodology.
The thesis presents the four stages of the development of a new methodology, STAPS-cons (Ship Traffic Allision Probability using Simulations – consequences). First, five distinct types of ship manoeuvring failure were identified based on Automatic Identification System (AIS) data: (1) loss of propulsion; (2) loss of steering; (3) miss of turning point; (4) wrong course at a turning point; and (5) turn at the wrong location. These failure types were also quantified with respect to frequency and duration, using a combination of trajectory analysis and manual classification. In the second stage, millions of input datasets were generated in a Monte Carlo simulation based on a combination of these failures, environmental parameters, and ship statistics. In the third stage, these inputs were simulated in a time-domain hydrodynamic simulator to model the ship behaviour under respective failure conditions. The number of simulations was determined based on route-specific traffic volumes and failure probabilities. In the final stage, the consequences of failures leading to allisions were analysed using a set of precalculated nonlinear finite element analyses, enabling the estimation of structural damage and collapse probability. The thesis focuses on energy transfer and bridge collapse probability; socioeconomic consequences (e.g., loss of life, or traffic disruptions) fall outside its scope.
The results presented in the thesis demonstrate that the STAPS-cons methodology provides a more detailed and realistic assessment of ship-bridge allision risk than existing models. The methodology also includes real-time replays for identification of high-risk scenarios and evaluation of risk mitigation strategies. The methodology was applied in multiple case studies and compared with established tools (e.g., IWRAP Mk II), wherein it showed comparable or better performance. Future work includes broadening the scope to include analysing offshore wind farms and the probability of ship groundings, automating failure classification and future route paths using AI, and developing fast-time consequence approximations to reduce the need for time-consuming finite element modelling. Overall, STAPS-cons represents a significant advancement in maritime risk assessment by combining probability and consequence modelling in a flexible, simulation-based methodology.