Project FINE4CAST
Excerpt
New generation of tools for forecasting energy demand and renewable energy production on fine spatial and temporal scales
Georges Kariniotakis, Research Director, Mines de Paris
The main objective of the Fine4Cast project is to improve short-term forecasting (minutes to days) of renewable energy production and consumption on a fine geographic scale (production plants, consumers, territories). Improved forecasts are essential if more renewable energies are to be integrated into power systems. The emergence of new players and new use cases means a need for new products. Forecasters face new challenges due to the proliferation of data sources, and the associated threats and confidentiality constraints.
Fine4Cast offers a holistic approach that covers the entire value and modeling chain of energy forecasting, from data to weather and energy forecasts, and including the optimal use of forecasts for decision-making in power systems and the energy market.
Keywords: Forecasting, Meteorology, High resolution, Data science, Renewable energies, Smart grids
Consortium
The Fine4Cast consortium brings together 6 laboratories with highly complementary expertise. Synergies with other PEPR TASE projects will enable the sharing of use cases and data sets. A reference group of 10 industrial players will provide insight into end-user requirements and the most relevant use cases. These players will also be able to provide datasets for the evaluation and validation of proposed approaches.
- Increase the spatio-temporal resolution of meteorological models, based on the AROME-EPS ensemble forecasting system, and on very high-resolution LES simulations.
- Develop weather forecasting products dedicated to the energy sector.
- Propose forecast evaluation strategies adapted to high-resolution forecasts and end-user needs.
- Extract relevant meteorological information from new data sources such as 3rd generation Meteosat satellite images, Sky Imagers, etc.
- Enable better interpretation of energy forecasts using physical approaches.
- Develop a continuous approach to energy forecasting capable of predicting renewable energy production and local demand at horizons ranging from a few minutes to several days, taking into account all available data sources.
- Predict consumption on a local scale, taking into account new uses and integrating contextual and socio-economic data (textual news, social networks, traffic, news searches, etc.).
- Adapt forecasting models when RE production or consumption is affected by significant flexibility management.
- Link energy forecasting models to the value they bring to the applications for which they are used (RE market participation or grid management).
Project news
Thèse – High-dimensional optimization of distributed assets in smart grids
Thèse – Optimization of flexibility services under multiple local uncertainties in the context of smart grids
No news
Publications
More projects
