Modelling the accident probability in large-scale, maritime transportation system

The probability of an accident in transportation systems can serve as a measure of these systems safety or risk, depending on the objective. Therefore numerous methods and models for risk evaluation, with respect to maritime, have been developed. However, these models are either too simplified, allowing relatively fast analysis but very often missing the substantial links among the model variables, or they are too slow for effective analysis, due to computational complexity, not necessarily being backed-up with the complexity of the model itself. Thereby, this paper introduces a novel method evaluating the probability of ship-ship collision in the maritime transportation system focusing on the open sea collisions, applying the queuing theory in the simulation model. The model allows relatively fast prediction as it focuses on the specific events (e.g. accidents), instead of simulating the whole traffic. To support this hypothesis a case study is presented focusing on a selected element of transportation system in operation.