Gas-solid packed beds have been widely studied as a cost-effective means of thermal energy storage in concentrating solar power (CSP) plants. Typically, the operation of packed beds in such systems is modelled by accounting for a finite rate of heat transfer between the fluid and solid media. This approach requires the coupled solution of the fluid- and solid-phase energy equations, which is computationally-costly, especially for year-long performance simulations. The local thermal equilibrium assumption, which assumes an infinite inter-phase heat transfer rate, can be applied to reduce the complexity and thus computational cost of packed bed models. However, the implications of making such an assumption in the context of CSP thermal energy storage system performance modelling is poorly understood. In fact, the application of the approach in long-term simulations has not been investigated before. This work addresses the topic by comparatively evaluating the performance of local thermal equilibrium and local thermal non-equilibrium models in the annual simulation of an air-rock packed bed, hypothetically operating in an open volumetric receiver CSP plant. The level of inter-model agreement is assessed in terms of annual bed exergy yield, bed blowing work, and plant power generation time. In addition, solution times are compared to establish the extent of computational cost savings. A parametric study examining the effect of variations in key bed design parameters on inter-model agreement is also conducted. The results obtained provide a clear indication of the strengths and weaknesses of either modelling approach, as well as of the suitability of the local thermal equilibrium assumption in general.