The impact of the mushy zone parameter (Amushy) and the chosen numerical model during the solidification of a commercial paraffin-type phase change material (PCM) in a vertical cylinder under T-history conditions was examined through numerical simulations. The cooling process was modeled using three methods implemented in the CFD software ANSYS Fluent 2020 R2: the enthalpy–porosity method, the apparent heat capacity (AHC) method, and a new model proposed by the authors which incorporates heat capacity directly into ANSYS Fluent. To accurately define the boundary conditions, radiative heat transfer between surfaces was taken into account. Furthermore, the influence of the mushy zone parameter on the simulation accuracy and solidification rate was investigated, with the parameter being treated as a function of the liquid fraction. The results indicate that the proposed model aligns closely with experimental data regarding cooling temperature, offering better predictions compared to the other models. It was observed that temperature varies with time but not with position, suggesting that this model more effectively satisfies the lumped system condition—an essential characteristic of the T-history experiment—compared to the other methods. Additionally, the analysis showed that a higher mushy zone parameter enhances the accuracy of simulations and predicts a shorter solidification time; approximately 11% for the E-p and 7% for the AHC model. Using a variable mushy zone parameter based on the liquid fraction also produced similar results, resulting in an increased solidification rate.