Bayesian Regression Model for Estimating Economic Loss Resulting from Two-Ship Collisions

Abstract

This study aims to develop a Bayesian regression model to evaluate the economic loss resulting from two-ship collisions using ten years’ ship collision data for occurrances in Fujian waters. The model results show that the involvement of passenger/cruise ships could cause the largest increment on economic loss in ship collisions. Interestingly, it is found the involvement of fishing ships could greatly increase the ship collision consequence in terms of economic loss. Results also reveal that the higher economic loss is associated with the collisions in the straits/sea areas, under the strong wind/wave conditions, during nighttime period, and in poor visibility conditions. The impact analysis results highlight that judgment errors play a decisive role in increasing the economic loss as compared with the other two types of human errors: lookout failure and operation errors. The results of this study are useful for policy-makers in proposing efficient strategies to mitigate the economic loss from two-ship collisions. The developed model is also beneficial for insurance companies in determining the appropriate ship insurance rates.