Endogenous Battery Degradation Optimization in PV-ESS Systems: Identifying the Lifecycle-Cost Sweet Spot Using NSGA-II

In grid-connected PV-ESS plants, battery degradation is often treated as an external correction rather than a quantity coupled directly to sizing and dispatch decisions. This study develops a constraint-aware NSGA-II framework in which cycle- and calendar-induced capacity fade are evaluated from the simulated state-of-charge trajectory at every fitness evaluation and optimized jointly with net present cost (NPC), expected energy not served (EENS), and curtailment. The key innovation is in the explicit coupling of four-objective degradation and optimization, along with identifying an intermediate lifecycle-cost operating region, rather than assuming that minimum degradation is always the most economical choice. The framework runs in MATLAB/Simulink R2023b. Degradation performance is assessed against laboratory data and compared with 18 months of battery management system data from Nigeria’s Olorunsogo site; cases in South Africa and India are simulated and serve as transferability tests rather than field validations. In the pooled NMC simulation results, a descriptive cost fit has an interior minimum near 2% per year; the selected regional solutions range from 2.2% to 2.8% per year. Relative to an aggressive 4% per year cycling baseline, the model estimates approximately 11% lower 20-year lifecycle cost in this intermediate operating region and with later battery replacement. While the exact optimal point varies with chemistry and assumptions, sensitivity analysis indicates that an intermediate, moderate degradation rate stays economically advantageous within the tested uncertainty ranges. In practice, this framework helps designers balance initial battery size and cycling frequency against replacement risk, rather than choosing systems based only on immediate cost.