Using data from connected thermostats to track large power outages in the United States. (15th December 2019)

• Record Type: Journal Article
• Authors: Meier, Alan; Ueno, Tsuyoshi; Pritoni, Marco
• Is Part Of: Applied energy. Volume 256(2019)
• Publication Date: 2019-12-15
• Subjects: Electricity -- Smart grid -- Hurricane -- Metrics -- Resilience -- SAIDI -- Sensor; Power (Mechanics) -- Periodicals; Energy conservation -- Periodicals; Energy conversion -- Periodicals; 621.042
• DOI: 10.1016/j.apenergy.2019.113940
• Languages: English
• ISSNs: 0306-2619

Abstract

Highlights: When Internet-connected thermostats go "dark", then a power outage probably occurred. Power outages caused by nine major weather events were compared to disconnected thermostats. The network of thermostats has a geospatial resolution similar to that of utility SCADA devices. Cable TV nodes could provide sub-distribution level voltage sensing at low cost. Abstract: The detection of power outages is an essential activity for electric utilities. A large, national dataset of Internet-connected thermostats was used to explore and illustrate the ability of Internet-connected devices to geospatially track outages caused by hurricanes and other major weather events. The method was applied to nine major outage events, including hurricanes and windstorms. In one event, Hurricane Irma, a network of about 1000 thermostats provided quantitatively similar results to detailed utility data with respect to the number of homes without power and identification of the most severely affected regions. The method generated regionally uniform outage data that would give emergency authorities additional visibility into the scope and magnitude of outages. The network of thermostat-sensors also made it possible to calculate a higher resolution version of outage duration (or SAIDI) at a level of customer-level visibility that was not previously available.