Load profile analysis and short-term building load forecast for a university campus

Nowadays, advanced metering infrastructure allows us to record building energy consumption at much higher resolutions. This paper presents the methods for building load profile analysis and forecast using the 15-minute data collected at a substation feeder at the Centennial Campus of North Carolina State University from May 2012 to April 2014. Building load signatures for benchmarking building load consumptions are first analyzed and extracted from load related environmental factors such as time of the day, temperature and humidity, etc. Then, methods to forecast hours-ahead building energy consumptions are developed. Comparisons are made to derive the benefit of having high resolution environmental data sets.