AI for Optimizing Efficiency and Cost for Grid Integrated Buildings

Building owners and controllers face numerous challenges when incorporating grid-interactive intelligent elements into their operations. As the electric power distribution grid continues to transform into a complex and active network, the focus for many facility operators is the economics of enhancing efficiency, resiliency, and comfort.

Large buildings (greater than 200,000 square feet) make up less than 1% of total buildings, but account for a staggering 26% of total commercial building energy consumption. Similarly, 11% of buildings range from 25,000 to 200,000 square foot range, but they accounted for 42% of energy consumption in commercial buildings. To address this high energy utilization and potential excess usage of resources, EPRI leveraged its expansive collaborative R&D network and expertise to develop an AI solution.

EPRI researched the control-oriented digitization of commercial buildings, specifically to assess the time, engineering, and amount of data needed to build predictive models that are cost-effective, accurate, and scalable. Central to these efforts are high resolution data sets of operational conditions and machine learning algorithms that incorporate rule-based and model predictive controls. The research is being conducted at two buildings on the east and west coast: a small independently-owned office building in San Leandro, California; and large EPRI-operated office building in Charlotte, North Carolina, which houses variety of large event conference rooms, small offices, open office floor space, various labs and bays, halls and several cafeterias and break rooms of various sizes.

The results of these efforts are expected to improve predictive modeling leading to enhanced efficiency, affordability, resiliency, environmental performance, and reliability as well as a scalable solution for buildings of all sizes. Using this work, utilities and building owners will also be able to manage loads dynamically and co-optimize with energy resources for improving occupant comfort. It’s critical to keep this work in mind while the research develops, as these improvements have the potential to revolutionize the commercial energy landscape.

Learn more about game-changing solutions generated by the convergence of artificial intelligence and the electric power industry during the upcoming Grid-Interactive Buildings meeting on June 22 and AI and Electric Power Summit on September 28-29, 2021.