Optimal location and sizing of photovoltaic and battery storage in distribution system considering time varying load data

Abstract

The integration of distributed renewable energy generation, especially in Photovoltaic (PV) planning studies, has increased over the past years due to its great benefit in distribution systems. However, the energy generated from PV is highly unpredictable and inconsistent as the generation depends on weather conditions. The rising issues associated with the variability and intermittency of PV sources could be overcome by integrating PV with battery storage systems (PV-BSS). Many studies on PV integration with batteries have been proposed; nevertheless, previous work only considers a constant load model and dispatchable generation unit. In addition, the battery is normally charged when PV output exceeds load demand. The impact of the battery on the technical benefits of the distribution system and the method to size the battery based on PV availability and load demand requirements are disregarded. Therefore, proper PV and battery sizing methods are necessary to supply the loads and avoid mismatches between PV and battery generation. This research proposed the optimization model of the PV-BSS integration to supply load during peak times to reduce power losses in the distribution system. In addition, the proposed model also aimed to reduce high-cost maximum demand charges from utility providers considering the latest Time of Use (ToU) tariffs, weather-dependent PV generation, and time-varying loads. To maximize the technical and economic benefits from PV-BSS integration over the total maximum demand reduction, Particle Swarm Optimization (PSO) is employed to find the optimal location and size of PV and battery by utilizing 13 years of historical solar irradiance data with different time-varying load models. The findings are evaluated based on the comparative analysis of total active and reactive power losses, bus voltage improvement, PV penetration level, and different impact indices, namely, active power loss index (PLI), reactive power loss index (QLI), and voltage deviation index (VDI). The research has been performed on IEEE 33-bus and IEEE 69-bus test distribution systems. Findings revealed that the proposed optimization model effectively determines the optimal location and size of PV and battery with a significant reduction of power losses and improvement in bus voltages. The power loss reduction with PV integration varies between 13.84% to 32.71%. The combination of batteries in PV-BSS helped further reduce power loss by between 34.72% to 62.22% and achieved high maximum demand reduction, which yielded to 27.85% and 24.18% annual energy savings for commercial and industrial consumers, respectively. In addition, the improvement in other performance indices is also significant. The optimization results using PSO have been validated using a Genetic Algorithm (GA), with negligible differences between both techniques. Overall, this thesis contributed to an optimization method to size and locate PV and battery considering time-varying load models to maximize the technical performances of the distribution network and electricity bill savings to the consumer for efficient power use during peak hours.