Big Data Applications for Improving the Reliability of the French Electricity Distribution Grid

The French electricity network has been developed since 1946. Enedis is in charge of power distribution on 95% of continental France and supplies 35 millions of customers. Access to the electricity network is necessary for a major part of every day life, as well as for governmental services such as health and transport, especially in times of crisis. French territory is subject to the risk of storms, heat waves, thunderstorms, and floods. With the average age of the network increasing, natural wear and tear, climate change, and new uses influencing the load flow, it is necessary to improve the ways of operating and maintaining the distribution network. The massive deployment of smart meters at different network levels, as well as connected sensors providing information on the state of the network, is creating large amounts of new data. Artificial intelligence techniques make it possible to automate data collection on a larger scale. The IT systems required to manage this data are also impacted. This transformation adds challenges to the existing IT systems which have had to adapt significantly since the opening up to competition of the electricity market in the early 2000s, the massive development of embedded renewable generation, and the large-scale deployment of smart meters. These new data sources allow of course improving existing business processes. Also, their use in mass, with big data tools and a centralized data pattern via a data lake, allows multiplying the uses and a better valorization of the new deployed material. Anticipation is enhanced with a more robust network, and reaction in case of incident or climatic crisis is done in better conditions and less time.