Re-identification of Smart Meter data
Erik Buchmann, Klemens Böhm, Thorben Burghardt and Stephan Kessler
Personal and Ubiquitous Computing, 17(4), 653-662.
The Smart Grid approach enhances the power grid with information technology. Smart Meters are an important part of the Smart Grid. They record the energy consumption of households with a high resolution and transfer consumption
records to the energy provider in real-time. In liberalized energy markets, many different parties have access to this
data. This puts the privacy of consumers at risk. In this paper, we analyze to which degree Smart Meter data, as collected by our industry partner, can be linked to its producer, using simple statistical measures. We devise features of the energy consumption, e.g., the first peak of demand in the morning, and we describe an analytical framework that quantifies how well these features can identify households. Finally, we conduct a study with data from 180 households. Our study shows that 68% of the energy-consumption records can be re-identified by simple means already. This insight is important for Smart Grids that are operational, as it emphasizes the need for research and use of anonymization techniques for the Smart Grid.