Switzerland's first local electricity market has successfully completed field testing in January 2020. For one year, 37 households in Walenstadt traded locally produced solar power within their own neighbourhood in a project called Quartierstrom ('District Power'). Participants could buy and sell solar power directly within their neighbourhood via a portal on which the participants could set their own purchase and sales price limits. The resulting transactions were processed automatically via a blockchain. The local electricity supplier, Water and Electricity Works Walenstadt (WEW), not only provided access to its distribution grid but also purchased surplus solar power and supplied the community with 'normal' power when the supply of solar power was insufficient. This innovative project, which is supported by the Swiss Federal Office of Energy (SFOE) as a flagship project, aimed not only to verify technical feasibility in the field but also to study user behaviour.
Significantly higher local consumption
After a year of field trials, project participants from research and industry give a positive appraisal. Thanks to the local electricity market, the purchase of the locally produced solar power almost doubled. The 37 households covered 33 per cent of their electricity demand with solar power produced in the neighbourhood – twice as much as before. These figures might have been anticipated, but it was surprising how well the project was received. The participating households played a very active role and perceived the electricity market as green, local, and fair. "After initial scepticism, even the energy industry has shown a great deal of interest and sees a lot of potential in the development. We've managed to really provoke discussion," says Christian Dürr, Managing Director of the Water and Electricity Works Walenstadt. Verena Tiefenbeck, leader of the Bits to Energy Lab at ETH Zurich, also gives a positive appraisal: "Quartierstrom was the first project of its kind in the world, and we did pioneering work on many different fronts. We're especially delighted that the technology operated effectively – apart from the usual teething troubles." For the Swiss Federal Office of Energy, the application of these new technologies was a particular area of interest. "The project allowed us to study the extent to which blockchain and artificial intelligence might be suitable for the direct marketing of electricity from decentralised energy sources and what role the energy supplier plays in a bottom-up approach of this kind. These findings should be helpful for the future development of the electricity market," says Benoît Revaz, Director of the SFOE.
Almost no one wants to pay more
One new feature of the system was that participants could use a portal to set a minimum sales price for their solar power and a maximum purchase price to buy solar power from their neighbours. "The participants frequently adjusted the price limits, especially at the beginning. But the price limit they set for buying local solar power was rarely higher than for normal power from the grid," says Tiefenbeck. On average, the participants were willing to pay just under 18 euro cents per kilowatt hour – less than the cost of mains power, which stands at 19.4 euro cents. Fewer than 10 per cent of offers were above this rate, despite the fact that many people had declared their willingness to pay more for local solar electricity in the surveys conducted beforehand. "This gap between attitude and action is seen in behavioural research time and time again," says Tiefenbeck. The researchers also attribute this to the fact that the participating households knew that local solar power was subject to lower grid fees and that, accordingly, the power suppliers were getting more for their power, even at lower prices. For their part, the households with solar power systems also sought to make a profit, asking for around 6 euro cents per kilowatt hour. When selling to the power plant they made as little as just under 4 euro cents.
Automatic pricing more effective
In order to compare different market models, the researchers deactivated the feature allowing individual pricing for a period of one month and replaced it with an automatic pricing system. When demand and production coincided, the solar power was distributed locally. The price varied automatically depending on whether the solar power was in relatively scarce supply or in surplus. With individual pricing, on the other hand, a small proportion of the solar power could not be sold because the price requirements of suppliers and consumers did not match. In surveys, a little over half of households stated that they preferred automatic pricing. "What was surprising was that participants who used the portal frequently tended toward automatic pricing," says Tiefenbeck. "Based on our experience, we don't consider individual pricing to be decisive for a local electricity market in the future."
Effectively raising awareness
What does seem to be important, on the other hand, is that participants are able to monitor production and consumption, as well as their purchases and sales, in real time. This function was very popular with users and contributed to raising awareness. Indeed, many participants said that they now use electrical appliances more when the sun is shining. They saw the peak and off-peak tariff system that still applies today as outdated when applied to renewable energies. Christian Dürr: "The participants developed an understanding of the energy market, thereby helping to balance supply and demand. This reduces the burden on the infrastructure and puts surplus power to sensible use."
Reliable software but a need for hardware optimisation
While the software proved highly reliable, the project team repeatedly struggled with hardware failures. As there was no smart metre with an application processor available on the market, the project team had to use devices with self-developed modules (Raspberry Pi) instead. "These devices have an error-prone SD-card memory system," says Arne Meeuw, who developed the blockchain system. For larger applications, projects like Quartierstrom would require certified and stable smart metres with an integrated application processor that could run different software tools.
The blockchain system, on the other hand, proved highly robust – although its capacity was limited. Twenty-seven prosumers acted as validator nodes to approve the transactions in the blockchain. These nodes represent the critical variable when it comes to scaling. "The system could still handle about five more solar installations," says Meeuw. The number of consumers, on the other hand, could be increased further. The system would remain stable with up to 600 pure consumers or other clients, such as batteries or flexible loads. "It would be possible to scale the system up by building multiple blockchains for different neighbourhoods," says Meeuw. In turn, these could then exchange electricity among themselves.
Low power consumption
Unlike public blockchains, such as those used for Bitcoin, the Quartierstrom blockchain is private. Moreover, the approval of transactions does not rely on elaborate computational processes. "The nodes reach an agreement on a proposed energy trade," says Meeuw. This mechanism does not require a great deal of computing power. The small computers that are used as smart metres and blockchain nodes consumed around 3,300 kilowatt hours of energy over the duration of the project. In terms of the volume of power traded in the local market, this consumption amounted to around 4 per cent.
Follow-up project in the pipeline
The pilot phase of the local electricity market as part of the SFOE flagship project has now come to an end. However, a follow-up project has been launched to ensure a seamless transition, albeit in modified form. The user portal has been redesigned and streamlined slightly, and prices are now set automatically. In the coming months, the hardware will gradually be replaced with series-produced equipment, and there are also plans to develop the trading platform into a marketable product. This goal is being pursued by the spin-off company Exnaton, which was founded by members of the development team at ETH Zurich. One option under development would allow participants to determine their preferred suppliers of local solar power, rather than setting prices. In other words, they could choose to buy electricity from their aunt's roof or from the farmer who sells them their eggs. After all, Quartierstrom has also demonstrated that in a local market emotions play an even larger role than price. (mfo)
