The spread of renewable resources, such as wind and solar, is one of the main drivers to move from a fossil-based to a renewable-based power generation system. However, wind and solar production are difficult to predict; hence, to avoid a mismatch between electricity supply and demand, there is a need for energy storage units. To this end, new storage concepts have been proposed, and one of the most promising is to store electricity in the form of heat in a Thermal Energy Storage reservoir. However, in Thermal Energy Storage based systems, the critical component is the storage tank and, in particular, its mathematical model as this plays a crucial role in the storage unit performance estimation. Although the literature presents three modelling approaches, each of them differs in the considered parameters and in the method of modelling the fluid and the solid properties. Therefore, there is a need to clarify the model differences and the parameter influences on plant performance as well as to develop a more complete model. For this purpose, the present work first aim is to compare the models available in the literature to identify their strengths and weaknesses. Then, considering that the models’ comparison showed the importance of adopting temperature-dependent fluid and storage material properties to better predict the system performance, the authors developed a new and more detailed model, named TES-PD, which works with time and space variable fluid and solid properties. In addition, the authors included the tank heat losses and the solid effective thermal conductivity to improve the model accuracy. Based on the comparisons between the TES-PD model and the ones available in the literature, the proposal can better predict the first cycle charging time, as it avoids a 4% underestimation. This model also avoids overestimation of the delivery time, delivered energy, mean generated power and plant round-trip efficiency. Therefore, the results underline that a differential and time-accurate model, like the TES-PD, even if one-dimensional, allows a fast and effective prediction of the performance of both the tank and the storage plant. This is essential information for the preliminary design of innovative large-scale storage units operating with thermal storage.