Cost-Based Optimization of Off-Grid Photovoltaic and Battery Energy Storage Systems Using Genetic Algorithm

Cost-optimal sizing of photovoltaic (PV) and battery energy storage systems (BESS) in off-grid settings is challenging due to nonlinear interactions between solar variability, battery degradation, and system reliability. Existing studies often rely on simulated load profiles, providing limited insight into how genetic algorithm (GA) parameters affect solution quality. This paper proposes a GA-based optimization framework that incorporates high-resolution hourly irradiance and measured load data from a building in Taipei to determine economically optimal PV-BESS capacities. The objective function accounts for capital, operation, degradation, and penalty costs, while enforcing energy-balance and state-of-charge constraints. A sensitivity analysis of GA parameters, population size, crossover rate, mutation rate, and generation count, reveals that balanced operator settings improve convergence stability. The optimized configuration achieves a minimum total cost of 2.882 × 10 NTD, with a reliable match between generation and storage capacity. The results demonstrate that GA is a practical and adaptable approach for techno-economic planning of off-grid PV-BESS systems, utilizing operational data.