Abstract
The need for sustainable and reliable power supply in universities in the global south to enhance teaching, research and administrative activities has increased the integration of Solar photovoltaic (PV) into the institutional energy mix of most universities. Unfortunately, the high initial cost of investment limits the installed capacity for these systems relative to demand, thus raising the need to ration power generated among different users, which is a pathway less explored. This study develops an autonomous fuzzy-based user prioritization model for rationing power generated by the proposed PV system among identified energy user categories using the Federal University of Technology, Akure (FUTA) as a case study. The study applied k-nearest neighbour (k-NN) for energy user classification using silhouette values to identify the optimal k value. The ground-mounted solar PV minigrid system was developed and simulated using Helioscope, identifying the system’s hourly yield. A Mamdani fuzzy inference system was developed based on user demand and solar radiation for power rationing among users. Fuzzy rules were developed from expert opinions, and three scenarios of solar irradiance (low, medium, and high) were simulated for prioritization sequence and power distribution over the period of active power generation of the mini-grid system. An optimal k value of 3 was obtained, classifying the university community into Low (LPUs), Medium (MPUs), and High (HPUs) power users. Campus A comprises five (5) HPUs, thirteen (13) MPUs, and forty (40) LPUs. However, for Campus B, only one facility was categorized as an HPU, while three (3) and twenty-six (26) buildings were categorized as MPUs and LPUs, respectively. Based on the hourly sequence run, the system assigns more priority to LPUs at low irradiance, with a minimum of 35% priority observed to be provided to this user category. MPUs receive the highest priority during the peak period of power generation (2 pm) when the solar irradiation is low. Demands from LPUs are given more priority in low solar irradiance, and MPUs are provided with more supply at medium and high levels of solar irradiance. The system is capable of exporting up to 30% excess power to the grid in low demand scenario. However, 10-20% of the peak demand is unmet at high demand scenario, thus necessitating a rationing approach to power supply among user classes. This technique has the prospect of serving as an effective framework for mini-grid systems fostering efficient use of energy and advancing environmental sustainability.