The French TSO RTE builds grid development plans with power flow and generation adequacy modelling. Both observed and simulated weather data are used as inputs. A Meteo France simulation with Arpege-Climate provides 200 years of gridded data under the current climate. Because power flow simulations for 200 years are computationally costly, our objective is to find a small number of representative years.
Statistical models based on historical data convert weather variables to energy-related variables : consumption, and wind and photovoltaic capacity factors for 25 areas in France. Assumptions on the evolution of electricity generation and consumption (population growth, new wind power plants, technology evolution, electric mobility...) come from our Long-Term Adequacy Report [1].
A clustering procedure is applied to the 200 years of energy-related variables. The same procedure provides day and year clusters. We use a distance between days and we consider that a period of time is a cluster of days. The clustering procedure is validated on power flow data.