The DTU10MW model scale wind turbine is adopted in the present work, which was developed as part of the UNAFLOW experimental campaign. In previous efforts for its numerical modelling, the fixed turbine was simulated using ReFRESCO CFD solver using both k−ω SST 2003 RANS and DDES turbulence models. Now, the PANS approach is also used, a novelty in the simulation of wind turbines using a fully-resolved geometry. A Verification and Validation study is presented for the three turbulence approaches, with meshes up to 130 million cells. The direct use of a PANS filter that resolves 75% of the turbulent scales originates a wake with more content in terms of vortical structures in comparison with RANS and DDES. The wake tip vortices are destabilized and start to break due to these new vortices, something that was not observed with RANS and DDES. This phenomenon was already noticed by other authors when using different Scale-Resolving Simulation (SRS) models. Nonetheless, the integral forces with PANS present large discrepancies in relation to both the experiments and numerical results with the other turbulence models. This is due to very premature separation in the whole blade span, since the boundary layer is laminar. To force transition to turbulent flow in the boundary layer of PANS simulations, and avoid premature separation, turbulence generation is included in the inlet. On the other hand, RANS does not require turbulent content in the inflow, due to its statistical nature, fully modelling the turbulence effects. A similar conclusion is applicable to DDES, since it switches to RANS in the boundary layer region. Overall, the ultimate goal is to find a balance between the turbulence closure filter and the inlet conditions that originate an adequate prediction of integral forces on the blades and vortical content in the wake.