A data-driven analysis of building energy use with emphasis on operation and maintenance: A case study from the UAE

A pre-requisite to reducing the energy intensity of the building sector is to assess and understand current drivers of energy consumption. Drivers include building design characteristics, outdoor environmental conditions, as well as the operation patterns of building systems by facility managers and occupants. In the United Arab Emirates (UAE), where buildings consume more than 70% of the total electricity generated, a comprehensive analysis of the drivers of building energy performance has not yet been conducted. This paper proposes an empirical top-down statistical framework to comprehensively evaluate drivers of energy use for a stock of buildings. The approach is unique in its ability to capture the combined effects of technical, as well as, operation-related building parameters. A combination of data analysis techniques, including CHAID decision trees and regression models, is proposed and applied to data gathered from 713 mixed-use buildings in downtown Abu Dhabi. Results indicate that in addition to physical building parameters such as window tinting, operational parameters, including AC cleanliness, chiller condition, and thermostat temperature setting have shown significant impacts on energy consumption levels. The findings shed light on the crucial - yet often overlooked - role of facility managers and occupants in ensuring efficient energy performance in the building sector.