End-use disaggregation in commercial buildings with the building automation system trend data

Abstract

End-use submetering is essential for energy management in large commercial and institutional buildings. However, most existing buildings lack adequate submetering even for major end-uses. End-use disaggregation techniques offer an untapped opportunity to supplement deficiencies in a metering network. This study presents an end-use disaggregation method for commercial buildings by using building automation system (BAS) data. The BAS trend data provide contextual information about the operational state of major energy-consuming systems and equipment such as fans, pumps, air handling unit (AHU) heating and cooling coils, and chillers. The method applies a series of multiple linear regression models disaggregating bulk metered heating, cooling, and electricity use data into different end-uses by using BAS data as predictors. The results demonstrate that the method can accurately disaggregate hourly buildinglevel electricity, heating, and cooling use into their end-use categories.