Uncertainties in simulation model for ice loads prediction in level ice with respect to structural reliability

Arctic shipping is increasingly in focus due to the diminishing ice cover due to the effects of climate change. With new shipping routes opening up, navigation in ice-covered waters receives increasing attention in industry and academia. One of the key issues in the design of ice-going vessels is the required plate thickness to withstand the ice loads. Several regulations have been developed, containing formulae for determining the place thickness for different ice classes. On the other hand, various engineering tools have been developed for simulating the ship performance in ice, with several of these tools explicitly aiming to determine the ice loads on the hull. Such tools in principle provide alternative means to determine the safety level of the hull elements based on first-principle modeling. In this report, a risk-based design approach is taken as a starting point for contextualizing the determination of the required plate thickness based on a structural reliability analysis. The focus of the work is to analyze the importance of the uncertainties in the ice loads resulting from a ship performance in ice simulation model, in relation to the structural reliability as calculated based on the simulated ice load time series. The study centered on the influence of the applied empirical parameters in the model design on the predicted ice loads with respect to the long-term safety and reliability analysis of the ship. The various assumptions in the simulation model for ship performance in ice are systematically varied to quantify the uncertainty about the safety index and failure probability of a plate under ice loading. The results showed that the predicted loads from the simulation model lead to some about of uncertainties across the investigated parameters, indicating the limited usefulness of the simulation model in a risk-based design context. The estimated uncertainties were however based on the relative variations of the simulated ice loads and the estimated safety index corresponding to each case studied. With these results, the suspicion for inherent uncertainties in the ice prediction model’s performance associated with the applied parameters was justified. Hence, to obtain better or improved performance from the ice prediction model, the sources of uncertainties should be minimized or eliminated. Also, with the ultimate aim of developing models and tools for risk-based design of ships in Arctic and other ice-covered waters, areas of further research and development in context of the ship performance in ice simulation model are highlighted to reduce the uncertainties