| Résumé | While understanding end-uses and their flow within a building is essential for energy management, many existing buildings still do not have adequate submetering for important end-uses such as heating, cooling, air distribution, lighting and plug loads. To this end, this paper presents a new end-use disaggregation method for commercial buildings. Unlike previous disaggregation methods which mainly rely on standalone devices to sample power draw at high frequencies, we used common building automation system (BAS) data types which indicate the operational state of fans and pumps to disaggregate low-frequency electricity data into three major end-uses: lighting and plug loads, distribution, and chillers. The method employs regression models that associate BAS data with related end-uses – e.g., variable speed air handling unit (AHU) fan state and fan electricity use. The model parameters are estimated by the genetic algorithm minimizing the misfit between the predicted and metered electricity use subject to several practical constraints. The accuracy of the method is evaluated by using a dataset from an office building with high-resolution submetering. Six different disaggregation scenarios representing various model forms are used in this investigation. The results indicate that the method can accurately disaggregate hourly building-level electricity use into three end-use categories for lighting and plug loads, distribution, and chillers with the contextual information provided by a few common BAS data types. |
|---|
