The growing number of distributed energy resources (DERs) in distribution networks brings new opportunities for local energy sharing. This paper proposes a multi-timescale energy sharing approach among DER aggregators and distribution system operators (DSOs) considering grid-battery energy storage system (BESS) capacity rental and network operations. An energy sharing coordinator is created to manage the energy sharing with price determination. In an hour-ahead stage

the buying/selling energy and required grid-BESS rental capacity are optimally determined by the aggregators while the network operation is robustly considered by the DSO. In addition to renewable generation and loads

the power exchanges of the aggregators are treated as uncertainties. Then during each hour

15-min-ahead energy transaction and controllable DERs are optimized to track uncertainty realization. The uncertainties in the aggregators and the DSO are addressed by stochastic and robust optimization methods

respectively. To efficiently solve the proposed energy sharing problem

a distributed solution algorithm with step length control and step reduction techniques is developed. The simulation results verify the high efficiency of the proposed energy sharing approach.