Smart4RES: Data science for renewable energy prediction | edp.com

Goals

EDP NEW is participating in a new R&D project named Smart4RES whose overarching objective is proposing a number of forecasting and decision-aid tools to cover gaps and to prepare the energy system of 2030 and beyond, when a massive penetration of weather-dependent Renewable Energy Sources (RES) generation is foreseen. Indeed, the short-term forecasting of RES generation is a key prerequisite for the secure and economic operation of power systems but the moderate accuracy of operational tools lead to non-negligible security margins to ensure secure operation. In addition, for RES power plants without feed-in tariff that are participating in electricity markets, forecasts are not accurate enough to prevent discrepancies for which they are financially penalised by grid operators.

Therefore, EDP NEW, together with Armines, INESCTEC, DNV-GL, DTU and HEDNO, among other partners, have defined six main goals for the Smart4RES project: 

• To define requirements for forecasting technologies to enable near 100% RES penetration.
• To develop a RES-dedicated view of weather forecasting, leading to improvements in forecasting of the relevant weather variables in the order of 10-15%, using various sources of data (satellite, all-sky imagers, NWPs), and the development of very high-resolution forecasting approaches.
• To develop a new generation of RES forecasting tools that are able to improve RES power production forecasting by at least 15%.
• To streamline the process of getting optimal value from data and forecasts, through new forecasting products and data marketplaces, and novel business models.
• To develop a new data-driven optimization and decision-aid tools for enabling the large-scale penetration of renewable energy, combined with storage, into the electricity market as well as to provide system services towards TSOs and DSOs.
• Validation of new models in living labs and assessment of forecasting value.
• The granularity of the RES forecasts, the use cases to be addressed and the types of data required to run the models are summarized in the figure bellow.

The granularity of the RES forecasts, the use cases to be addressed and the types of data required to run the models are summarized in the figure bellow. 

In this project, EDP NEW is responsible for three of the ten validation sites of the project:

• RES plants at different areas in Europe
• Stocare Project – Cobadin Wind Farm in Romania, where EDP Renewables uses batteries to flatten forecasting errors
• Wind Float Atlantic, taking advantage of five years of collected data to validate the Smart4RES models. 

In these demonstrations, EDP Renewables will have a third-party role where it will be responsible for defining which RES plants should be considered for the first validation case and to facilitate the access to data and algorithms which are used for RES power prediction. The demonstrators will have a main role in verifying the prediction tools created during the project and how they compare, in terms of prediction accuracy, to the already available tools, enhancing their actual demonstrated accuracy.

The project had its official kick-off in November of 2019 with the presence of representatives of the thirteen members of the consortium. It will be developed during three and a half years, with a budget of four million euros.