En poursuivant votre navigation, vous acceptez l'utilisation de cookies destinés à des fins de mesure d'audience, à améliorer la performance de ce site et à vous proposer des services et contenus personnalisés. En savoir plus


EDF group sponsors a research program on solar energy at l'X

Thanks to the Previnergy program, École Polytechnique Dynamic Meteorology Laboratory will conduct research to improve software used to forecast the production of electricity from solar energy. This research will be funded by EDF customers via the offer that has just been launched by the group.

École Polytechnique and EDF signed in November 2015 a research agreement for the "Previnergy" program which aims to improve the prediction of photovoltaic electricity for the industry. The goal on the long-term is to develop and optimize the production of renewable energies from solar farms.

To finance this research, led by the Dynamic Meteorology Laboratory (CNRS, École Polytechnique, ENS and UPMC), EDF is turning towards its customers. The group has just launched a new offer entirely dedicated to renewable electricity: it ensures 25 million individuals that, for all of their consumption, the equivalent in renewable electricity will be produced and injected into the network provider. Furthermore, for every MWh (megawatt hour) consumed, EDF will donate two euros to Previnergy research program.

With this funding, the Dynamic Meteorology Laboratory will recruit a PhD student in 2016 in order to conduct the research. To help manufacturers make the most of their production of solar electricity, the PhD candidate will focus on understanding the processes involved in the life cycle of clouds and mists in particular. The idea is to better understand how they work to improve their forecasting. "The link between solar radiation incident on the surface and clouds is simple: solar energy can be halved during a cloudy episode," says Martial Haeffelin, a specialist in atmospheric observation who works at the Dynamic Meteorology Laboratory.

The solar power generated depends on weather conditions, so it is important to be able to predict the evolution of its production to manage the supply-demand balance. Cloudy situations can create production shortfalls and strong irregularities. Fog and low stratus can affect the production of solar energy unexpectedly as the ability to predict these phenomena is limited.

Improve and expand the use of the Pvscope forecasting software

The work carried out through this research program will help to improve the Pvscope forecasting software. This tool, which was developed through a project carried out by the the Dynamic Meteorology Laboratory with EDF's collaboration since 2011, is used to anticipate the production of solar energy for the hours and days ahead. This forecasting tool incorporates three types of information: production data, satellite imagery data and a Numerical Weather Prediction (NWP) model. "Thanks to mathematical tools, this prediction model allows for simulations of changes in atmospheric parameters. The model incorporates information from regular measures on the state of the atmosphere such as temperature, humidity and atmospheric pressure," says Martial Haeffelin. Equations are used to calculate the evolution over time of these parameters.

If the model autonomously calculates the changes in temperature, humidity and pressure, it is however not able to do so yet for the clouds. "The model does not have a fine enough resolution to visualize accurately the clouds whose evolution scale can vary from a meter to a micrometer, continues Martial Haeffelin. And this is all the more difficult with the fog and low clouds because they are affected by very small scale heterogeneities which impact the complex ways in which they form and dissipate." The idea of the research project is therefore to exploit the information contained in the satellite imagery data to assess the evolution of fog and stratus in real time. By combining information from the NWP with those of satellite observations, it will be possible to improve the prediction of solar energy production for the next 1 to 6 coming hours.

The Previnergy program will thus contribute to the development of new features of the Pvscope software, but il will also help to expand the use of the tool. So far, Pvscope has been used at two solar farms located on the island of Reunion, selected for its sunshine and the complexity of its landscape generating various clouds, but within three years, the software could be also used to forecast the photovoltaic energy produced by medium and large solar farms in France and even in Europe.

In addition to the recruitment of a PhD student, the funding allocated to this research is likely to be reassessed: "As more individuals will subscribe to EDF's offer, more support will be brought to the Previnergy program," says Martial Haeffelin, who hopes to extend the research to other École Polytechnique laboratories working on renewable energies in a second phase.