Device-level plug load disaggregation in a zero energy office building and opportunities for energy savings

Abstract

Along with heating, cooling, ventilation, and lighting, plug loads are one of the principal consumers of energy in commercial buildings. Managing ever-changing plug loads is a significant challenge given the quantity and variety of devices in commercial buildings and the cost of monitoring equipment. To address this issue, we propose a method for developing a disaggregated model of an office building's plug loads that utilizes power data from a small portion of monitored devices and a device inventory. Using data from metering and control devices in the Research Support Facility at the National Renewable Energy Laboratory, we compared our model to the building's plug load submeters. We found that the model was effective in predicting the shape of the building's average plug loads; however, it did not account for the entire magnitude of the load. With the disaggregated breakdown, we identified devices that contributed significantly to the building's unoccupied load, such as the audio visual equipment as well as devices that contributed significantly at specific times of the day, such as the microwaves at noon. This disaggregated information allows building owners to make more informed decisions with respect to plug load controls and energy efficiency upgrades. In addition, we highlight how the plug loads in the Research Support Facility have changed over time and offer recommendations for implementing this disaggregation method in other buildings.