The damaged stability assessment for a passenger ship is a process requiring the simulation of multiple damage scenarios. Nevertheless, the stochastic nature of the damage stability framework requires the analysis of a statistically significant number of cases. On the other hand, the probability density functions used to estimate the possible damage dimensions and locations along the ship generate many scenarios that are not critical for the ship's survivability, especially for large passenger ships. It is standard to apply empirical rules to restrict the number of damage scenarios, such as critical damages is only above two compartments, considering that damage stability regulations currently in force ensure survivability levels beyond this extent of breaches. However, a rigorous approach is lacking. To this end, in the present work, it is proposed to use more scientific-based methods to identify critical damages. This paper presents three original approaches developed in the context of a multi-level damage stability assessment. The first method relies on preliminary static calculations, the second on the energy absorbed by the ship during an impact, and the third on a purely dynamic approach. Here, the methods are critically compared on two sample passenger ships for collision damages, showing their respective advantages and disadvantages.
