To reduce the operational costs of multi-microgrids and address the issues of optimal configuration and fair cost allocation of shared energy storage, a dual-layer optimization configuration for shared energy storage under time-of-use pricing of the main grid and an improved Shapley value method for fair allocation of shared energy storage are proposed. Firstly, a dual-layer optimization configuration model for shared energy storage in multi-microgrids is established. The upper layer aims to minimize the annual investment and configuration costs of shared energy storage, while the lower layer aims to minimize the operational costs of multi-microgrids incorporating shared energy storage. This model determines the charging/discharging strategy of shared energy storage, the optimized configuration capacity, and the operational costs of multi-microgrids. Secondly, based on the emergent benefits derived from minimizing the operational costs of the multi-microgrids system, a contribution evaluation model is constructed. Finally, an improved Shapley value method based on contribution is applied to allocate the cost savings generated by the shared energy storage configuration. Case simulations demonstrate that this method provides theoretical support for fair cooperation among multi-microgrids, effectively resolves the challenge of equitable cost and benefit allocation during interactions between the main grid and microgrids, and holds practical value for fair collaboration among multi-microgrids.