Degradation-aware sizing of off-grid PV-Battery systems using a Constructal Theory metaheuristic algorithm

• CTOA enables degradation-aware sizing of an off-grid PV-battery system. • PV decay and battery capacity fade are included in the lifetime sizing model. • CTOA is benchmarked against PSO, DE, and FA on eight test functions. • The optimal system uses 5 PV modules and 2 battery units at low annual cost. • Battery SOH falls to 72%, while PV output drops over the 20-year horizon. Reliable sizing of off-grid PV-battery systems remains challenging because long-term component degradation can significantly affect both cost and adequacy. This study proposes a Constructal Theory Optimization Algorithm (CTOA) for the degradation-aware sizing of a standalone PV-battery system supplying the FabLab of the University Institute of Technology Fotso Victor in Cameroon. The optimization minimizes the Total Annualized Cost while explicitly accounting for photovoltaic efficiency decay and battery capacity fade over the project lifetime. CTOA is first evaluated on eight benchmark functions against Particle swarm optimization (PSO), Differential evolution (DE), and Firefly algorithm (FA), and is then applied to the real case study under site-specific climatic conditions. The optimized configuration consists of five 300 W PV modules and two 2.4 kWh battery units, with a Total Annualized Cost of about 1635.84 US$/year and an LPSP of 0.00001 (dimensionless fraction, i.e., 0.001%). Compared with PSO, DE, and FA, CTOA reaches the best solution in fewer iterations and shows faster stabilization of the cost function. The degradation analysis indicates that battery SOH decreases to about 72% by year 20, while PV performance also declines progressively over the same period. In addition, the annual PV energy output decreases from 2362.16 kWh/year to 1808.76 kWh/year over the 20-year horizon. These results show that CTOA provides an effective framework for degradation-aware sizing of off-grid PV-battery systems under long-term operating conditions.