Packed rock-bed thermal energy storage (TES) systems represent a robust and economically attractive solution for large-scale solar energy storage, enabling enhanced efficiency, dispatchability, and operational flexibility of solar thermal power plants. This study presents an in-depth computational fluid dynamics (CFD) investigation of an air–rock packed-bed TES system, in which air is used as the heat transfer fluid (HTF). A two-phase local thermal non-equilibrium (LTNE) model is implemented within a porous media framework, with distinct energy conservation equations solved for the fluid and solid phases and interphase convective heat transfer explicitly modeled through appropriate closure correlations. The governing mass, momentum, and energy equations are numerically solved using a finite-volume approach and rigorously validated against experimental measurements reported in the literature. A systematic parametric study is performed to evaluate the influence of rock particle diameter on flow distribution, interphase heat transfer coefficients, pressure drop, thermocline evolution, and global storage performance during charging and discharging processes. The results demonstrate that particle size exerts a first-order control on the coupled thermal–hydraulic behavior of the packed bed. Specifically, smaller particle diameters significantly increase the specific surface area and enhance interphase convective heat transfer, resulting in steeper axial temperature gradients, thinner and more stable thermoclines, faster thermal front propagation, and improved utilization of the storage volume. These effects lead to higher energy storage capacity, increased charging effectiveness, and improved thermal efficiency; however, they are accompanied by a marked increase in pressure losses and pumping power requirements due to reduced bed permeability and higher flow resistance. In contrast, larger particle diameters promote higher bed permeability and more uniform velocity fields, thereby substantially reducing pressure drop and pumping energy consumption, but they exhibit lower heat transfer coefficients, delayed solid-phase thermal response, increased thermal dispersion, and broader temperature fronts, which degrade thermal stratification and limit effective energy recovery. The analysis further reveals that the optimal particle size represents a compromise between maximizing thermal performance and minimizing hydraulic losses, and that this trade-off becomes increasingly critical at higher operating temperatures and mass flow rates. Overall, the results provide detailed physical insight into the role of particle size in thermal and hydraulic performance and offer valuable CFD-based guidelines for the optimal design and scaling of packed-bed thermal energy storage systems for solar thermal applications.