Self-propulsion predictions are important to understand propeller–hull interactions and the communication between the main engine and the propulsion system. Experimental work encapsulates the core of self-propulsion estimations, but due to their high cost, initial calculations are made using empirical relations suggested by IMO. However, these empirical relations may, at times, fail predicting some of the self-propulsion parameters. In this study, a specific example was chosen in which IMO recommendations fail to predict the wake fraction. Numerical results utilizing RANSE are used to solve for the hull-propeller system to assess their validity in contrast with the experiments. RANSE is implemented in two different methods. The first one is by simulating directly the hull-propeller system, while the second is the hull-virtual disk system, in which the propeller is only modelled as a circular disk. It was found out that numerical results are more robust with respect to empirical relations advised by IMO.