Data analytics applied to the electricity consumption of office buildings to reveal building operational characteristics

ABSTRACT Rigorous statistical analysis of whole building, 15-minute interval, time series electricity data enables remote insights into buildings’ operational characteristics. We developed select building markers and applied them to six commercial office buildings located in three different climate zones for comparison. The building markers reveal information about daily operational patterns, scheduling, and the ratio of base to peak load. Time series analysis, clustering, anomaly detection, diffusion index-based forecasting, first-order energy differential, data visualization and data mining techniques were used for marker development. The daily operational pattern marker identifies weekday and weekend energy consumption patterns and was used here to quantify opportunities for alternative weekend energy scheduling to reduce energy consumption. The scheduling marker recognizes the turn-on and turn-off times for HVAC and other scheduled equipment. Here, we quantified an alternative HVAC schedule can reduce on average 2.7% energy consumption in the office buildings. The base to peak load ratio marker identified that the selected office buildings could reduce their baseload by using more aggressive night and weekend temperature setbacks. Ultimately, these building marker functions may be employed on any whole building electricity datasets to gain insights to building operation and characteristics, enabling improved identification of potential energy savings measures.