Urban energy informatics: Improving the usability of building energy data for community energy efficiency

Abstract

Of the 150 US cities that have adopted 100% clean energy resolutions, most are considerably far from achieving this goal. Concurrently, the rise of advanced and affordable sensors offering continuous monitoring of city infrastructure has directed research attention towards how data-driven approaches can help cities become 'smart' and achieve sustainability goals. As buildings account for the majority of energy consumption in cities, they have become a key focus for smart city initiatives. The influx of measurements on building energy and infrastructure at urban-scales add substantial complexity in handling of this information; an important area for research is developing approaches to translate this data into metrics that can be helpful for energy management and decision making at the community and city scale. Across the three studies in this dissertation, I take a multidisciplinary approach and draw on areas across data analytics, human-computer interaction, and public policy analysis to transform building energy data for improving community energy decisions. In the first study, I present a new approach for building energy benchmarking using building electricity smart meter data across the Georgia Tech campus. The results aim to support building portfolio owners and municipalities in identifying and prioritizing specific energy efficiency opportunities across a group of buildings. In a second study, I enhance the visibility and awareness of the same data through the development of a community-scale energy feedback system, and evaluate Georgia Tech community member understanding and reactions to having access to campus energy information. In a final study, I explore the impact of built infrastructure on renewable energy deployments across urban areas to inform urban planning design and policy. The results of this work seek to contribute to research efforts within building energy efficiency fields and enhance our understanding of how advances in data science and computing can be connected to energy management and decision making practices. As cities strive to make substantial changes in their energy systems, emerging data sources may provide immense opportunities to make more effective and informed decisions; however, enabling this will require integration of new data science techniques with existing decision making practices. Continued connections between these two areas are likely to foster unique insights and pave the way for cities to attain a low-carbon future.