The used battery packs with inhomogeneous cells suffer from energy loss and poor state of charge (SOC) estimation accuracy. In order to reduce the cell-to-cell variations by minimizing the energy loss and SOC estimation error, they need to be reconfigured. In this paper, a framework for cell selection method is developed. To describe the cell characteristics, the first-order RC model is selected. Then, the cell testing profile with static capacity and dynamic parameter test is proposed. By means of the linear squares estimation, the cell parameters such as nominal capacity and dynamic parameters are obtained. Lastly, a screening algorithm minimizes the cell-to-cell variations with respect to both nominal capacity and dynamic parameters. Thereby, the optimal combination of selection parameters (sigma-range for nominal capacity and weighting factors for dynamic parameters) is studied customized to the design requirements (number of packs and number of cells within pack). The results of case study imply that the developed framework reduces the maximum cell-to-cell variation by 60-70%. The reconfigured second-life packs can be offered with lower price compared to new battery packs by guaranteeing reliable energy supply. Therefore, they are the cost-effective and reliable energy storage solution for stationary power applications.