Abstract: Under the "dual carbon" goal, building a new type of power system with high proportion of renewable energy requires a large amount of flexible resources to ensure its reliable operation. Large scale grid side energy storage plays an important role in providing diverse services to the system, such as peak shaving, frequency modulation, and inertia supporting. However, it is hard to recover the storage costs only relying on the electricity market revenue from arbitrage, which has become the main cause to restrain the investment and development of the grid side storage. Therefore, this paper investigates the cost sharing mechanism of the grid side storage based on the externality theory. On the analysis of the externality value characteristics of the grid side storage versus multi-agents under the market environment, a bi-level decision model of the grid side storage including the externality value quantification and the storage cost sharing is established. The upper level gives the optimized planning capacity of its energy storage, and the lower level simulates the impact of storage on the market prices and quantifies the externality value to different market entities. Simultaneously, the lower level feedbacks the market price and cost compensation signals to the upper model to guide the storage investment. Furthermore, a cost sharing method for the grid side energy storage is designed and the cost sharing mechanism is applied relying on the market clearing. The calculation results show that the proposed method is able to guide the scale of the grid side energy storage investment and shorten the cost recovery period of storage projects. The designed cost sharing mechanism meets the compatibility of the multi-agent incentives and is well compatible with the existing electricity market trading rules.