With surging demand for power grid peak shaving, the traditional independent configuration mode of user-side energy storage systems suffers from low utilization efficiency, long investment payback periods, and severe resource waste. To address these issues, this study proposes a two-layer programming optimization model based on capacity sharing. By coordinating the upper-layer capacity configuration optimization with the lower-level charge-discharge scheduling, the model aims to minimize the total life-cycle cost. Case analysis shows that after adopting the capacity-sharing mechanism, the joint configuration scheme for the user side and the grid side increases the user-side energy storage systems utilization rate by 40%. The total system revenue covers the cost of independent configuration, and the user-side revenue per unit capacity rises by 56.7%. In addition, under the joint configuration scheme, the overall system revenue per unit capacity increases by approximately 20.7%, which proves the model’s effectiveness in balancing the grid’s peak-shaving demand and user-side economy. This study indicates that the capacity-sharing mechanism significantly enhances the operational flexibility of the new power system and has great potential for promotion in regional-level energy storage systems planning.