Provider intervenes in the household
How smart power grids of the future will work
May 21, 2023, 6:25 p.m
Nuclear power in Germany has been passé since mid-April. In the future, most of the electricity will be generated from renewable energies – for example with photovoltaic systems on private houses. At the same time, millions of households will switch to electric cars and heat pumps in the coming years. The local power grids are not yet designed for this. In order to prevent bottlenecks, they have to become more flexible – with smart technologies. An example from the south of Baden-Württemberg shows how this can work.
What is the problem?
E-car, heat pump, photovoltaic system, own energy storage – this is what the household of the future will probably look like. By 2030 there will be up to six million of these systems in the lower power range. This is what Joachim Seifert, professor for the digital networking of energy systems, at the Technical University of Berlin expects. “This will be a huge challenge for network operators.”
This is mainly due to the low-voltage grids – i.e. the grids that bring electricity to households. “That’s where the bottlenecks are: in the transformers, in the last switch boxes and in the lines that are laid in the streets,” says Seifert. This is because the grids have so far mainly been designed to transport energy from central power plants to consumers. In the future, there will be numerous high-performance consumers such as charging points at home (wall boxes) for electric cars and heat pumps. At the same time, many households are becoming electricity producers who also want to feed this energy into the grid.
And that can push the networks to their limits?
Yes. The keyword is simultaneity. Two examples: when all the electric cars in a residential area are plugged in at the same time on a winter evening and the heat pumps are running. Or when many photovoltaic systems generate electricity at the same time on a sunny afternoon – and want to feed it into the grid. “Then the networks can reach a capacity limit,” says Seifert. In the first case, the grid could not provide enough energy, in the second, it could not absorb the electricity produced. In both scenarios, the result would be an overload. The transport capacity of the network would be exhausted. In the worst case, this could lead to a power failure.
Why do networks have to become “smart”? Isn’t an expansion enough?
The expansion is urgently needed – but there is a lack of time, money and skilled workers. “It won’t work without expanding the grid. We don’t want to have to constantly intervene,” says Carmen Exner from the network operator Netze BW. However, there is a lack of resources to be able to expand the distribution network at the required pace. Since the risk of overloading is still very rare at the moment, this is also inefficient. Load peaks only occur at certain times. It makes no sense to oversize the entire network for this. This is where smart networks come into play: They should be able to control the connected systems – to relieve the network in an emergency. According to Exner, they should also be used to bridge the time until the grid is expanded: “We want to connect all systems to the grid as quickly as possible. But if that’s only possible in a few years, we need intelligent solutions.”
How exactly does this work?
To find out, Netze BW undertook a 17-month field test in Freiamt in southern Baden. For this purpose, 23 households were selected that were equipped with photovoltaic systems and heat pumps. Storage, wall boxes and technology such as intelligent electricity meters and energy management systems were partly provided. This is how the household of the future was simulated, says project manager Exner. In addition, 24 larger PV systems were included in the test. The tests took place in real network operation.
Exner and her team managed to prevent simulated bottlenecks in two ways: In the event of acute overloads, automated commands were sent to households to purchase or feed in less electricity. With the help of measurement and weather data, it was also possible to predict bottlenecks, which were then prevented. In the example, the e-cars would charge at night instead of in the evening. And if the weather forecast announces a sunny day, the PV systems could avoid peak loads at midday by filling the in-house storage system instead of feeding it into the grid.
What does that mean for consumers?
In the best case, the consequences of this so-called grid-friendly control are not or hardly noticeable. Experience from the trial shows: “Most customers didn’t notice whether their heat pump was running a little earlier or a little later – as long as the water and the apartments were warm,” says Exner. Acceptance was high. This also applies to electric cars. The participants were able to indicate preferences when charging – for example when they need the car and with which charge level.
“Smart power grids therefore enable grid operators to prevent overloads by giving them more room to maneuver – without our customers suffering any disadvantages,” says Exner. In the future, consumers could also benefit from digital networks through dynamic electricity prices, for example – and charge their e-cars when there is a particularly large amount of renewable electricity available in the network and is correspondingly cheap. The Federal Government is also concerned with the issue: before the end of this year, the traffic light coalition intends to present a new set of rules on when and how network operators can intervene in the network.
When will this be available across the board?
According to the Federal Ministry of Economics, numerous network operators are running projects on the feasibility of smart power grids. However, the Netze BW project is innovative in demonstrating the interaction. Netze BW wants to use the results as a blueprint for its own network expansion – and encourage the standardization of system interfaces, for example.
This is also urgently needed, since there are still a wide variety of options for switching the systems, says Seifert from the TU Berlin. At best, this should be standardized across Europe. It is difficult to estimate how long it will take for intelligent power grids to be implemented: “The energy industry is very conservative when it comes to digitization. It will certainly be another ten years before all areas from buildings to low-voltage grids are comprehensively digitally developed”.