Energy meets AI: flexibility instead of idle wind turbines
Machine learning can help us to manage green energy – so that we can push ahead with the energy transition. The research project DESIGNETZ provides evidence of this.
Germany is on its way to having a sustainable energy supply: by the year 2030, 80 percent of the country's electricity requirements will be covered by renewables, according to the coalition agreement of the federal government. In 2020, the figure was around just 46 percent.
But despite the fact that we are still not producing enough electricity from renewable sources, it is not unusual for wind turbines to be standing idle. One reason: wind power and photovoltaic installations are dependent on the weather; it is therefore much more difficult to plan how much they can feed into the grid than it is with conventional large power plants. If there is a surplus, for example because of stormy weather in the north of the country, the plants are brought to a halt.
This lack of stability can be balanced out by integrating flexibility into the power grid, for example by coupling volatile electricity generation plants with heat storage units. The storage units can take up the excess energy, providing relief for the power grid and thus preventing the generating plants from being switched off. The decisive currency for the success of the energy transition is thus not just the megawatt – but also flexibility.
How can flexibility be predicted?
The provision of flexibility was therefore the focal point of the research project DESIGNETZ, a sub-project of SINTEG, a project funded by the Federal Ministry of Economic Affairs (BMWi). As well as looking at other ways of using new technologies, the focus here was on the use of Power-to-Heat (P2H) and combined heat and power (CHP) plants.
A key question: How can flexibility be predicted in a better way? This is important in order that available storage capacity can also be made use of. A team from umlaut examined this question as part of the DESIGNETZ project, set up the demonstration units for managing the flexibility of the P2H and CHP plants, and integrated them into the overall system.
Diagram showing how P2H can be integrated into the power grid (source: DESIGNETZ Dokumentation Band 2, p. 121)
Using Artificial Intelligence to forecast flexibility potentials
The core of the project: a machine learning model that helps to predict flexibility potentials. With this goal in mind, sensor data was used to first of all determine the current values for the available flexibility of the individual plants across a defined testing period. This allowed training data for the prediction of flexibility to be collected over a period of almost a year. In the case of the P2H plants, an average flexibility of 40kW was determined for the 29 plants which were connected up to the demonstration unit. For the three CHP plants that were connected up to the demonstration unit, evidence of the availability of flexibility was provided in up to 9 percent of the times tested during operation.
The availability of flexibility is, however, dependent on the weather to a large extent because the intention is to balance out weather-dependent power generation and consumption through the systems provided – their utilisation will thus also be influenced by current and expected weather conditions. As well as the operating data collected, we therefore also recorded the corresponding weather data so that we could train a so-called support vector regression model to allow us to predict the flexibility of the individual systems. The method serves as a way of describing and quantifying statistical dependencies. The result: on the basis of weather forecasts for the coming three days, the trained model is able to make predictions with regard to flexibility potentials.
Integrating prognoses and measurement data
A crucial factor if the operation of the P2H and CHP plants is to be beneficial to the grid is the integration of the flexibility potentials into an overall system into which current electricity and gas network measurement data and plant data are also fed. In the DESIGNETZ project, this was provided by the so-called system cockpit, which serves as a central element for the exchange of data and the initiation of requests for flexibility. In addition to setting up the demonstration units, the project team from umlaut was also responsible for the realisation of the data connections to the system cockpit, as well as the implementation of requests for flexibility (see Figure 2).
Such aspects of data connection, data preparation and data management regularly come up as hurdles in data analytics and AI projects, aside from the actual development of the model, and have to be resolved by the project team. As well as having a good understanding of data and modelling, it is therefore crucial that those working on such projects also possess interdisciplinary knowledge and skills in the area of energy and systems technology if project implementation is to be successful.
Diagram showing the connection of the individual plants to the central system cockpit (source: DESIGNETZ Dokumentation Band 2., p. 71)
Flexibility as the key element of sector coupling
umlaut was able to show that it would be possible from a technical point of view to provide flexibility through the use of P2H and CHP plants. In view of the high availability of flexibility which could be demonstrated during the testing period, it has been shown that the coupling of the electricity and heating sectors has a great deal of potential in terms of future energy provision.
The exploitation of these synergies and the realisation of the associated potential for efficiency is essential in creating a sustainable energy supply from renewable sources, without at the same time endangering the stability of the network and security of supply. In addition to the power grid, further infrastructures could also benefit from targeted requests for flexibility. In the case of P2H, for example, switching on electricity-based heat generators could reduce the requirement for gas, thereby taking pressure off of the gas grid.
In the future, the rapidly expanding decentralised storage capacity from electric-powered vehicles can also be used to feed electricity back into the grid. In this case, a key prerequisite for the use of such decentralised storage units is knowledge of the available flexibility in terms of when and where. As was shown in the example of the P2H and CHP plants, AI-based forecasting models present a suitable means for supplying this knowledge. If network operators and energy suppliers are able to set up the necessary information base for their training, these models could be an important tool on the path to achieving a sustainable energy supply.