By combining data from satellite observations, models, and ground elevation, the REGREEN pilot project aims to develop AI-map to identify multiple renewable energy sources, and enhance the stability and reliability of power generation and transmission.
Deep Blue started a pilot project called REGREEN (Renewable EnerGy souRce gEneration sitE fiNder), co-funded by the EOSC FUTURE (under the European Union Horizon Programme INFRAEOSC-03-2020 – Project ID 101017536). The pilot project aims at developing innovative software solutions to support the transition away from fossil fuels and towards carbon neutrality by developing an AI-based tool to identify the best energy-generation sites of three RES: solar, wind and hydropower. The purpose of the pilot is to advance the idea from its conceptual stage (TRL2) to an initial prototype of the core functionalities (TRL3).
Aiming to achieve “Net Zero” carbon emissions by 2050, in recent years EU countries have been intensifying efforts to promote the transition to renewable energy sources (RES). Despite the generalised awareness of their crucial importance and efforts to foster their production, there are still limitations related to their availability, intermittency, and unpredictability that pose challenges to the widespread adoption of more sustainable solutions. To address these issues, Artificial Intelligence (AI) is emerging as a promising opportunity by optimising renewable energy generation and enhancing its reliability.
AI-Driven and Co-Generation: a promising solution for sustainable energy generation
The REGREEN pilot project aims to create software to support the transition away from carbon neutrality by developing an AI-based solution to identify the best energy generation sites of three RES: solar, wind and hydropower. “The project objective is to produce a map of the sites which are suitable for an efficient production of at least two or three RES” explains Carlo Abate, Head of Environment and Energy at Deep Blue. The decision to prioritise three different renewable energy sources instead of just one is easily said: enabling multiple-energy-source production.
“In principle, energy co-generation is a flexible process that can potentially be optimised to compensate for fluctuation locally, and ensure stable energy transmission to the grid – explain Abate – for instance, by combining solar and hydropower in a production site, the excess energy produced by solar panels in a sunny day could be used to pump water into a reservoir used as an energy storage, which would be emptied during the night to exploit the water power.”
The co-generation presents a significant opportunity to address challenges related to availability, reliability, and power-generation stability without relying on batteries and encountering associated issues. This is the underlying concept, and to make it feasible it is necessary to identify sites with high potential for generating energy from multiple renewable sources, and that’s where AI plays a crucial role.
AI-Powered tools to predict optimal sites for co-generation
AI-powered tools hold the capacity to compute and solve complex problems in a really effective way, that is why this innovative technology has emerged as an effective alternative in several sectors, including environmental and energy ones, showing immense promise in the optimization of renewable energy sources. In fact, from optimising energy generation and distribution to enhancing efficiency and driving innovation, AI is revolutionising the way we harness and utilise renewable energy sources. For instance, predictive algorithms based on neural networks and other machine-learning techniques are being used to forecast weather patterns to optimise the use of wind turbines and solar panels, and AI is also being adopted to optimise the operation of smart grids, which can facilitate the integration of RES into the national grid.
The REGREEN project operates in the identification of potential sites for renewable energy generation, traditionally requiring extensive and expensive measurement campaigns to monitor factors such as solar radiance, wind patterns, and water streams. “By combining data from Copernicus satellite observations, models, ground elevation, and existing renewable energy facilities, we believe AI algorithms can be trained to associate geographical and weather characteristics with the total power generated from different renewable sources” highlights Carlo Abate. The model will use these insights to predict energy generation from each renewable source, thereby identifying the most viable sites for co-generation plants.
The ability to accurately predict energy generation from renewable sources based on AI algorithms holds the potential to reshape the renewable energy sector. By eliminating the reliance on batteries for energy storage, the REGREEN project aims to enhance the stability and reliability of power generation and transmission, and deliver innovative methods to tackle Climate Change. “The final step is to feed the model with the local geography and weather characteristics of any site where there currently are no RES plants – explains Abate – It might be easy to say it in words, but will the AI really be able to predict how much energy can be generated by each renewable source, based on the ‘lessons learnt’ during the training?”.
The objective of the REGREEN pilot project is to answer this question.