Abstract-Demand Side Management (DSM) using Thermal Electric Storage (TES) presents a promising opportunity for enhancing the system flexibility, resulting in reliable and economic operation of future low-carbon power systems. Systemwide analysis of the flexibility potential of TES necessitates representation of dynamic thermal models in large-scale power systems models. Therefore, this study presents a novel Buildingto-Grid (B2G) model integrating buildings' thermal dynamics and end-use constraints with a security-constrained unit commitment model for energy and reserve scheduling. The behaviour of residential thermal demand is represented through linear state space (RC-equivalent) models for different residential archetypes. The B2G model is subsequently used to evaluate the energy arbitrage and reserve provision potential of TES for a test system and various sensitivity analyses for wind penetration levels and presence of other flexibility options have been conducted. The optimisation results highlight the significant value of TES in terms of annual generation cost savings, reserve provision, peak load reduction and utilization of wind energy. The findings also emphasize the importance of co-optimising energy arbitrage and reserve provision from TES devices vis-a-vis system performance and household energy consumption scheduling.