Uncertainty-aware electrification of university campuses: Matching demand with renewable generation for zero emissions

Abstract

Urban decarbonization requires robust methodologies to match electrified demand with on-site renewable generation under operational variability. University campuses are particularly well-suited testing environments, as their predominantly daytime demand profiles favor solar photovoltaic (PV) self-consumption. However, most campus-scale studies adopt deterministic approaches and do not explicitly quantify the impact of uncertainty in solar generation and demand, which can significantly affect system performance and design. This study proposes a structured and transferable methodology for campus decarbonization, integrating high-resolution hourly demand analysis, PV generation modeling, electrification measures, and uncertainty quantification within a Best Estimate Plus Uncertainty (BEPU) framework using Wilks’ non-parametric tolerance intervals. The approach is applied to the ETSII buildings at the Universitat Politècnica de València (UPV), within Valencia’s climate-neutrality strategy. For a case study with ∼3 GWh annual demand and 1.3 MWp PV capacity, results indicate that solar PV alone can supply about 70% of annual demand, with self-consumption levels of 67–69% and overall self-sufficiency near 50%. Adding a 1 MW / 4 MWh battery increases self-consumption above 95% and self-sufficiency to ∼65%, despite storage losses. The BEPU/Wilks framework provides statistically robust performance bounds, supporting reliable PV–storage sizing. The methodology provides a robust and scalable framework for planning the decarbonization of urban areas and university campuses.