Degradation-aware optimization of second-life battery applications in multi-energy charging hubs: Identifying profit-optimal operating conditions

Abstract

The growing number of retired electric vehicle batteries presents both an environmental challenge and an economic opportunity. Repurposing them as second-life batteries (SLBs) can reduce lifecycle costs and material waste, yet their economic value strongly depends on how they are operated. Although the effects of charge–discharge rates (C-rate) and depth-of-discharge (DoD) on battery degradation are established, their long-term economic implications under different operating conditions remain insufficiently examined. This study develops a degradation-aware optimization framework for integrating SLBs into a photovoltaic (PV)-coupled charging hub serving both electric and hydrogen vehicles. The proposed framework consists of a pre-optimization stage and a full system optimization. The cost-minimizing system optimization is first solved as a nonlinear MINLP to capture battery behavior and generate degradation data. The same problem is then reformulated as a mixed-integer model using this information. This two-stage approach replaces long-horizon MINLP simulations with a more efficient MIP formulation, achieving an 87% reduction in solution time while maintaining high accuracy, with a relative error of 0.33% per simulation interval in loading and degradation profiles. Six operating regimes are analyzed by varying C-rate (0.3–0.8) and DoD (0.7–0.8), and long-term performance is evaluated over a 25-year horizon using net present value (NPV). The results indicate that economically optimal operation does not necessarily coincide with minimal degradation; notably, the most aggressive strategy (0.8C, DoD = 0.8) yields a 2.27% higher NPV. Overall, the proposed framework provides a scalable decision tool for enabling profit-oriented SLB operation while supporting reliable and sustainable battery reuse. • Assessed SLB techno-economics in PV-integrated multi-energy charging hubs. • Screened SLB operating conditions comparatively to identify the optimum strategy. • Identified candidate operating conditions via a dynamic-degradation pre-optimization step. • Linearized model reduced computation time by 87% with 0.33% error. • Quantified how operating strategies affect SLB lifetime, replacements, and NPV.