The scale of power battery decommissioning increases steadily as the rapid development of electric vehicles, but current methods to recycle retired batteries cannot utilize their residual value effectively. The echelon utilization of retired batteries in energy storage systems becomes the focus of research. However, the inability of existing capacity allocation strategies to balance the economy and reliability is an urgent problem. Therefore, a two-stage hybrid energy storage system (HESS) optimal configuration model is proposed in this paper. Firstly, HESS consisting of fresh batteries and echelon batteries retired from electric vehicle is developed, whose life cycle cost (LCC) model is built. Secondly, the proposed strategy takes the total capacity of HESS as the decision variable, and the minimum LCC as the objective function in the first stage. In the second stage, taking the capacity ratio of the echelon battery as the decision variable and the network loss as the objective function, the power flow is optimized, and the HESS income under the optimal flow is calculated, which is a part of LCC in the first stage. Finally, the proposed two-stage optimization model is solved by adaptive genetic algorithm (GA) combined with second-order cone programming (SOCP). According to the results, the differences in investment cost and grid operation situations caused by the capacity proportion of echelon batteries in HESS are analyzed, verifying the ability of proposed scheme to achieve both the economy and reliability of power system.