Packed beds of particles have various applications in different sectors such as mining, chemical, and food industries. Determination of heat exchange rate between solid particles and fluid has been of great interest in various energy-related activities. Therefore, the main objective of this study is to investigate the influence of particle size on the overall performance of the thermal energy storage system. In order to investigate the effect of design and operating parameters on heat transfer characteristics of packed bed, an extensive numerical study followed by multi-objective optimization based on response surface methodology is conducted in the present study. A comparison between the obtained results for volumeaverage-based models (local thermal equilibrium and local thermal non-equilibrium) and experimental data is made. The numerical results show an accurate prediction of thermocline only when local thermal non-equilibrium model is used. After validating the numerical results, the desirability function is used to find the optimum particle diameter for different values of pumping power. The particle diameter that yields the highest energy storage capacity and lowest fan power is reported for the considered thermal energy storage system. The optimization procedure is further extended such that the optimal particle diameter for different values of fan power is also reported.