Degradation-aware operation of heterogeneous second-life electric-vehicle battery modules in photovoltaic–storage–grid systems

The growing retirement of electric-vehicle batteries creates new opportunities for second-life stationary storage, especially in photovoltaic (PV)–battery energy storage system (BESS)–grid applications. Whether these heterogeneous modules can provide reliable operational value, however, depends on how dispatch decisions account for module-level capacity dispersion, state-of-health differences, degradation costs, engineering constraints, and limited computation time. This paper assesses the operation of heterogeneous second-life EV battery modules in PV–storage–grid systems through a module-resolved, degradation-aware scheduling model, a deployment-oriented scenario study, and a certified dispatch layer. The model captures time-coupled battery dynamics, discrete operating actions, PV/load/grid energy balances, optional inter-module balancing, and separable piecewise-linear cycling and calendar aging costs. The certified dispatch layer is used as a decision-support tool: for each declared balancing architecture, it reports lower/upper bounds and an interruption-time optimality gap for the same model instance. In the Sc2 case over 10 seeds, the proposed dispatch achieves a median finite-horizon cost saving of 22.9% relative to no storage under the stated initial SoC and free-terminal-inventory convention, increases PV utilization from 84.0% to 97.5%, and reduces curtailment from 16.1% to 2.5%. Across H0–H4, which jointly vary usable-capacity dispersion and degradation response, median savings decline from 24.2% to 20.6%, while SoH-loss dispersion increases by more than sixfold. These results show the system-level relevance of module heterogeneity without separating the contribution of each attribute. The results show how second-life battery operation should be evaluated as a joint question of system value, aging impact, and verifiable decision quality.