Optimizing Small-Scale Pumped Hydro Storage Operation for Off-Grid PV: Maximizing Stored Energy Under LPSP and Carbon Footprint Constraints

This study explores the utilization of Particle Swarm Optimization (PSO), Grey Wolf Optimization (GWO), and Puzzle Optimization Algorithms (POA) methodologies to improve the performance of a small-scale PHS integrated in an off-grid PV system in a building. The optimization objective function is to minimize the LPSP value, with a carbon footprint of less than 10 kg/kWh. The LPSP value is related to the total energy deficit and total load, and the carbon footprint value is related to the total heat generated by the PHS pump, generator, and carbon intensity value. The optimization setup uses a multi-objective function that has been simplified into a single weighted objective function with normalized and justified weights. The case study is conducted on a 5 kW PV system in a building with a water level of 24 meters and a PHS reservoir of 5 m3. The system is tested under two conditions, namely during the rainy season (January) and the dry season (August). The PSO and GWO algorithms, based on testing results in January and August, demonstrated better performance than POA. This is based on the higher average total PHS energy compared to POA, as well as lower LPSP, LOLE, and EENS values. Meanwhile, for the average stored energy and carbon footprint values, the POA algorithm performs better than PSO and GWO, as indicated by the higher average stored energy and lower carbon footprint values.