How we model demand
Our demand modelling captures both established consumption patterns and the growing influence of behind-the-meter solar and battery installations.
Data sources and forecasting method
Backtest: Our model uses historical demand data sourced from ENTSO-E, EirGrid, and Elexon.
Forecast: We build a 15-minute demand shape based on historical demand seasonality patterns, scaled to align with annual consumption projections:
- Continental Europe: We use demand data from ENTSO-E’s ERAA 2024 (short to medium-term) and TYNDP 2024’s (long-term) Distributed Energy scenario. From the total annual demand numbers, we exclude the corresponding demand from batteries and pumped storage since these are obtained implicitly when modelling storage
- Great Britain: We use NESO’s FES 2025 Holistic Transition pathway
Gross versus net demand
To reflect the impact of behind-the-meter assets, we distinguish between two types of demand:
- Net demand
- The metered demand seen by individual TSOs
- Gross (total) demand
- In addition to net demand, this includes behind-the-meter generation—for example at CHP plants and from residential sources like rooftop solar
Modelling gross demand
We are careful when modelling gross demand patterns to first adjust the shape of the demand. This is done by supplementing net demand shape with behind-the-meter sources such as rooftop solar, before scaling the overall demand values.