Abstract
Large university campuses are micro-societies and mini-cities with significant opportunities to lead by example in the clean energy transition. They often used Building Energy Management Systems (BEMS) to monitor and control energy supply and demand. However, this data could be further exploited to provide a better understanding of the performane of renewables in specific geographical regions. In this paper we leverage empirical data from a public university BEMS for research purposes, aiming to provide insights for citizen groups such as energy communities who need support to make long term planning decisions. We evaluate statistical and machine learning models of photovoltaic (PV) power generation from PV arrays on university buildings, and provide a tool to estimate PV generation to support energy community planning decisions.