More solar panels, wind turbines, electric cars and heat pumps are needed to achieve the Dutch climate ambition. However, the grid is already reaching its limits in many places. Which means that new sustainable generation or users of energy must wait a long time until there is room on the grid. To accelerate the next steps of the energy transition, better use must be made of grid capacity.
This can be achieved by matching supply and demand of energy and making use of extra space on the grid when it is available. Additionally, optimizing usage of grid capacity can reduce the cost of grid reinforcements, which are expected to reach up to 102 billion by 2050 in the Netherlands alone [PwC, 2021].
To effectively match supply and demand, it is important to proactively gain insight into the expected generation and consumption. This can be done by having AI algorithms analyse data from various sources, such as weather forecasts, energy production and patterns of energy consumption. This allows the AI algorithm to learn and predict variations in generation and consumption so that it can respond proactively. This allows network operators, together with market parties, to better manage the flows on the grid. This reduces costs and increases the stability and reliability of the grid. AI is crucial to make better use of the electricity grid and enable the energy transition.
Grid operators, knowledge institutions and market parties have joined forces for this and have started the AI for the Future Energy System project (AIFES) with the help of AINed. In the first phase from April to September 2023, a Proof-of-Concept (POC) is realized, in which market parties, regional and national grid operators show for a concrete case how the grid can be optimally balanced and transport capacity can be optimally made available to the market parties using state-of-the-art AI technologies. Experts in energy data exchange help achieve this in a standard way, while experts from knowledge institutions ensure that the POC uses and is in line with the academic state-of-the-art.
Participating parties are Alliander, Tennet, CWI, TUD, HvA, SIA Partners, MFFBAS and Giga Storage. The project is partly funded by AINed.
This visual shows the importance of forecasts within the energy system. On all levels and for all actors’ forecasts either are or will be important for daily operation.
The full project proposal of AIFES describes five core domains that will be worked on to achieve a system breakthrough in the energy transition over a 3-year period starting in 2024, with a size of €10mln.
The first domain, forecasts, aims to reduce the uncertainty within the energy market for grid operators. Forecasts are highly relevant not only for intermittent renewable generation and new sources of demand (e.g., from electrification of transport, heating), but for all types of generation and consumption including the differentiation of generation types to consider probability distributions. By improving forecasts, the safety margins in operating the electricity grids can be reduced, freeing up much-needed capacity to connect or increasing customer connections, which currently can take up to eight years with the status quo solutions.
A key technology to advance is the domain of Artificial Intelligence (AI). For forecasts, but also for the three domains of local behaviour/ flexible energy markets, data markets for information sharing and self-optimizing software for actors, AI provides the possibility to achieve big gains. Applying the latest research from the academic partners in AIFES to the operational work of the other partners to confirm hypotheses and find new state-of-the-art applications is core to the AIFES project. Finally, AIFES aims to enable as many people as possible to take part in and understand the project’s results. Therefore, the AI algorithms and methods are further improved on, such that they are explainable and transparent.
The total project costs are approximately €10mln of which €5mln is contributed by AINed. The subsidies lead to an approx. funding of 30-40%[1] for non-academic partners. Note that the proposal for AIFES will be submitted by the end of September 2023.
[1] The final funding ratio will be determined with the complete consortium in place and their contributions known.
Big interests and opportunities for the Netherlands with AI – AiNed
Announcement incl AIFES: Update development AiNed Chain project – AiNed
Netherlands (sia-partners.com)
Giga Storage | The future of energy storage, today! (giga-storage.com)
TenneT | Lighting the way ahead together
Intelligent Electrical Power Grids (IEPG) Group (tudelft.nl)