Sizing optimization of generator sets for LV grid maintenance interventions

Each year, the French DSO Enedis operates generator sets (gensets) during maintenance work on the low-voltage (LV) distribution grid to ensure continuity of power supply. Historically, genset capacity was determined using simplified models based on transformer ratings and aggregated annual power demand data from Linky smart meters. However, these methods often resulted in overdimensioned gensets, leading to increased costs, carbon emissions, and underutilization of low-emission generators. To address these challenges, Enedis developed Prodige, an AI-driven tool that predicts the required power during interventions by analysing smart metering data, seasonal trends, and grid characteristics. Prodige optimizes genset sizing, reducing inefficiencies and improving the use of low-emission gensets and lithium-ion models. Deployed nationally in 2024 after three years of experimentation, Prodige represents a significant step towards Enedis’ carbon neutrality goals and enhances operational planning across its regional divisions [1].