The Water and Electricity Works Walenstadt (WEW) has announced the completion of its year-long blockchain pilot, which created a local energy market between 37 homes. The goal was to see how homes that produced rooftop solar electricity could trade with other participants, and the role that the central grid would play in the process, using a blockchain system to keep track of these transactions.
Backed by the Swiss Federal Office of Energy (SFOE), the Quartierstrom project saw WEW purchase the excess electricity from the homes, but also supply electricity when solar was not available. WEW hasn’t committed to deploying the system in the rest of its utility footprint, but other utilities globally will be interested in the full findings from Quartierstrom.
The ETH Zurich and the University of St. Gallen were academic participants, which saw a private blockchain deployed – to record data pulled from Raspberry Pi computers, which were used because Quartierstrom could not collectively find suitable smart meters. Unsurprisingly, hardware optimization is cited as an area for improvement.
Of the 37 homes, 28 had solar capacity and were described as ‘prosumers.’ These homes ran blockchain nodes on the Raspberry Pis, using a customized module apparently, while the other 9 homes had no solar generation, and were called ‘pure consumers.’
Strangely, the infrastructure needed for the project consumed 4% of the total produced energy, or some 3,300 KWh, which really sounds quite high for a few dozen Raspberry Pis. The third-gen Pi uses a 5W power supply, meaning that it would require 120 watt-hours per day, or 43,800 watt-hours in a year – so 43.8 KWh.
With 37 nodes, that should equate to around 1,620 KWh, which would be around 2% of the total solar produced, but instead we are told that the figure was 3,330 KWh – suggesting that the modules are quite power hungry. This could be solved with dedicated silicon designs, however, inside purpose-built smart meters. This figure would also be blown out of the water once you factor in a year’s worth of server hardware too, squirreled away in a data center, so this is a relatively minor complaint in the grand scheme of things.
The SFOE is hoping that Quartierstrom can be successfully developed into a marketable product, and plans to develop bespoke hardware to this end. It is not clear whether the SFOE would remain in the Swiss market, or if it has global ambitions, but there are a number of other blockchain-based experiments in this field – with LO3, the Energy Web Foundation, and Power Ledger perhaps the most prominent.
To be clear, the homes were still buying the vast majority of their electricity from WEW, however, when you examine just the solar power generated, it is good to see that most of it was used locally – with the winter months of January and December showing around 90% of the solar being used by the generating home itself, or its neighbors. In the image below, you can see that in the summer months, over half the generated solar is sold off to WEW.
But as you might have inferred, in the winter months, the proportion of the total consumed electricity used that is generated by solar is miniscule, with January and December seeing north of 85% of energy provided by WEW from the central grid, as shown below. Notably though, in the summer months, the local solar provides around 45% of the energy mix, which is encouraging from a DERs perspective. However, solving the winter problem still needs some other fuel source from the central grid, and that’s likely going to have to transition to wind and nuclear, if we are focused on cutting carbon emissions.
The blockchain-powered portal would allow those with PV panels to set a minimum selling price, and the prospective buyers could set their own maximum buying price. However, it was found that buyers did not attach much of a premium for locally-generated, environmentally friendly solar power.
“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,” said Verena Tiefenbeck, leader of the Bits to Energy Lab at ETH Zurich.
This was in part due to the flexibility in pricing, where the blockchain-powered portal would allow those with PV panels to set a minimum selling price, and the prospective buyers could set their own maximum buying price. It was found that only 10% of the buyer offers were above the price of electricity sold by WEW from the central grid.
For the solar sellers, however, there was a higher profit margin to be had when selling to the other participants, instead of to WEW, because there were no grid-fees involved in the sale. One of the most striking conclusions was that the project saw home solar consumption double, as part of the energy mix, reaching 33% of the homes’ total usage.
There was an uptick in residents scheduling their appliances to run when the sun was brightest too, and the ability to monitor energy usage was very important to them. To this end, they became very aware of how the energy market functions, which is a promising indicator for utilities that want to deploy such projects at national-scale. It suggests that consumers would understand how such a system would operate, and could change their behavior accordingly, thanks to price sensitivity.
However, there was a significant intervention in the project. Initially, the selling and buying prices were set manually by the participants, but this led to situations where there was local energy was ‘wasted,’ because there were no buyers willing to pay the required price. This led to Quartierstrom changing the system to an automated auction-based one, in order to ensure that the local solar generation was used locally – as the localization of energy generation was a primary focus for the project.
“What was surprising was that participants who used the portal frequently tended toward automatic pricing and vice versa. Based on our experiences, we don’t consider individual pricing to be decisive for a local electricity market in the future,” added Tiefenbeck.
The announcement reads a little like a post-mortem in the IT world, and it is refreshing to get this much detail from any blockchain project. As for follow-ups, Quartierstrom says that a new project has already been launched, with some changes. These include a redesigned user portal, with those aforementioned automated prices, and series-produced hardware will also be installed to replace the Raspberry Pi boards. A spin-off from the ETH Zurich lab, Exnaton, is going to be trying to create the marketable product.