Degradation-aware multi-objective optimization of wind-battery energy systems using NSGA-II

Abstract

Battery energy storage systems (BESS) are increasingly integrated with wind power plants to mitigate production variability and enhance economic performance in electricity markets. However, optimal battery sizing requires simultaneous consideration of economic profitability, operational smoothing capability, and lifecycle degradation, which inherently represent conflicting objectives. This study proposes a degradation-aware multi-objective optimization framework for wind–battery energy systems using the Non-Dominated Sorting Genetic Algorithm II (NSGA-II). The framework integrates dynamic state-of-charge modeling, equivalent full cycle (EFC) based degradation estimation, and capital recovery factor (CRF) based economic evaluation within a continuous optimization structure. Real operational data consisting of 744 h of wind generation and electricity price information from a 70 MW wind power plant are employed to evaluate system performance. The optimization results reveal a well-defined three-dimensional Pareto surface representing nonlinear trade-offs among net profit, grid injection variance, and battery degradation. The deterministic benchmark identifies a profit-maximizing configuration with a storage capacity of approximately 5.0 MWh and a net profit of approximately $632,687 over the analyzed period. In contrast, the NSGA-II optimization generates a diverse Pareto solution set enabling flexible trade-off analysis. The TOPSIS-based compromise solution selected using weights of 0.50 for profit, 0.25 for variance, and 0.25 for degradation, corresponds to a battery capacity of 10.18 MWh and a power rating of 2.00 MW, achieving a net profit of approximately $266,189 over the simulation period. Pareto quality evaluation yields a hypervolume value of 66,038,808 and a spread indicator of 0.455. The results demonstrate that degradation-aware multi-objective optimization provides a complementary framework for wind–battery system planning, offering diverse Pareto trade-offs unavailable to deterministic sizing approaches.