Segmentation and Classification of Commercial Building Occupants by Energy-Use Efficiency and Predictability

Drawing inspiration from utility-scale customer segmentation research initiatives, a new set of metrics is introduced that serve as quantitative measures of building occupant energy efficiency and energy-use predictability. Building occupant energy-use data is segmented to facilitate the construction of independent energy-use profiles for workdays, nonworkdays, work hours, and nonwork hours, which in turn enable further classification of building occupants according to their energy-use patterns. The three new metrics, building occupant energy-use efficiency, entropy, and intensity enable the design of more targeted energy conservation campaigns. Building occupants with relatively low energy-use efficiency scores can be individually targeted for behavioral interventions aimed at increasing the efficiency of their energy-use. Furthermore, building occupants with relatively low energy-use entropy scores can be sent timely behavior intervention notifications based on predictions of their future energy-use. Finally, building occupants with relatively high energy-use intensity scores can be targeted for equipment upgrades in order to reduce their overall energy consumption. We present the methodology behind the construction of these metrics and demonstrate how they can be applied to classify commercial building occupants based on their energy-use.