Distribution and transmission operators could trade for flexible capacity at the same time
The secret of systems that are user-friendly and simple is they are often the result of complex mathematical calculations. This is particularly true of the trials for a Local Energy Market (LEM) in southwest England, which sought to prove that decentralised renewables generation assets within the region could provide the grid with sufficient flexibility to self-balance the bigger swings in demand and supply that accompany a shift to localised clean power.
Cornwall is one of the sunniest and windiest regions in the UK with surging levels of renewable energy generation. But it also faces severe grid congestion. The three-year £16.7 million pilot scheme tested the role of flexible demand, generation and storage through a digital trading platform. It allowed network operators to pay energy producers and consumers to either ramp up generation or reduce load in a competitive but coordinated process.
Participants included households and businesses, who volunteered to take part in the trials, receiving free equipment in return. While businesses participated in the auctions themselves, residential players were managed by an aggregator, Kiwi Power, a UK-based energy technology company, which formed a virtual power plant from more than 600 kilowatt-hours (kWh) of battery storage capacity.
The aggregated capacity could then trade autonomously with the National Grid Electricity System Operator (ESO) and Western Power Grid, the distribution network operator (DNO). The platform ran closed-gate auctions, where both the buyers and sellers submitted the quantity and availability of their flexibility capacity. Once the market closed, an algorithm matched the orders in terms of price compatibility, while also considering the capabilities of the technologies involved and the grid.
“You need to ensure that the actions taken on the market will not blow up the transformer. We can ensure that through market-clearing because we input these types of constraints into the system,” explains Adrien Rosen, from N-SIDE, a Belgian software company with expertise in advanced analytics, which provided the basis of the Cornwall LEM digital platform.
The algorithm takes into account the limits of the renewable or storage technologies, for instance, a battery cannot be asked to discharge twice without sufficient time in between to recharge, Rosen says. Techniques such as these enabled the LEM platform to exist in a way that would not have been possible previously, Rosen adds.
Advanced analytics in the decision-making engine enabled close-to-real-time trading on a large scale. Though the type of problems the algorithm was solving could theoretically have been solved manually, the complexity would have led to slow answers. “If it takes me two weeks to solve a problem, it’s completely useless in the scope of energy,” he says.
N-SIDE applied a market-clearing algorithm used elsewhere in wholesale markets but reconfigured it to fit the smaller, local market of the Cornwall project, factoring in the different types of grid constraint, timing and participants. “We needed to offer something that could work for a battery owner, but also for an owner of a diesel generator, or a wind turbine, or a PV panel. Those kinds of constraints are not the same as you would see on international wholesale markets,” he says.
The innovative aspect of the project—facilitated by the new digital products available—was to allow both the ESO as the Transmission System Operator (TSO) and the DNO (Western Power Distribution) to buy flexibility simultaneously. “With two different buyers, we had to really coordinate the actions of both the TSO and the DNO to avoid one requesting flexibility in one direction and the other asking for it in the other direction at the same location. It was quite a challenge but we found a way to take it forward,” Rosen says.
Conflict avoidance was also built into the platform for the website, according to James Atkinson, Centrica’s local flexibility market product owner. This allowed flexibility buyers and sellers to see information on the headroom available at a particular node on the grid, all the flexibility procured at each location, a list of all the offers and sites that are registered on the platform.
“Buyers can see what the other buyer has either requested or contracted and therefore if there are any potential conflicts that might arise during procurement,” he says. The trials were based on simulated needs of the TSO and DNO that mimicked real conditions, rather than real congestion on the grid, but there was no reason that the platform could not be used for flexibility procurement on the national network, Atkinson says.
It was great to prove that the system was viable. “Being able to design and run a market to activate local flexibility for both the distribution and the transmission system operator was definitely a key learning for us. There was a lot of research, but this was the first real-life pilot,” Rosen says.
A key feature of the LEM is in taking a system developed for the wider wholesale markets down to the local level, says Centrica’s Atkinson. “The concepts themselves are relatively standard for energy markets, but using them to coordinate TSOs and DNOs, and using it to model their headroom and the capacity constraints of the local network, that’s the main innovation,” he says.
Community spirit The Cornwall LEM saw businesses and households trade flexibility capacity to both the DNO and TSO. Digital products ensured only those with enough capacity could enter the auctions
The importance of digitalisation in this process will grow as the need for flexibility scales in size and geography, Atkinson believes. Currently, DNOs are procuring flexibility through a relatively manual process, he says. “This is okay at the moment because they’ve really just started to use flexibility as business as usual, and they’re running tenders once or twice a year. Whereas what we think will be the case in a few years’ time, is that they’ll need flexibility at a much, much larger scale across many more locations. That’s where the requirement to automate the process will be—it means you can effectively scale it across as wide an area as you’d like, without it being restrained by the human resources available to make the decisions,” he says.
The system is fully developed in terms of managing the whole process of flexibility procurement for a DNO, but more functionality could be added to improve usability, or even to integrate it with other markets, Atkinson says. “At the moment, flexibility procurement is quite isolated. Integration with wider wholesale markets is still being explored in research projects,” he says.
Theoretically, there is no limit to the amount of flexibility the system could handle, Atkinson says. “The algorithm that clears the market doesn’t see the identity of the buyers and seller, so in theory, it could have an infinite number of buyers and sellers and an infinite number of buy and sell orders. You’d need more and more computer processing power in order to handle it, but no there’s no limit in theory,” he explains.
In future, it should be possible for network operators to automate the creation of bids, so that when they run their load forecast and sensitivities on their own systems, flexibility requests will automatically be created on the platform, he adds. What is missing for the future of flexibility markets, according to Rosen, is international harmonisation so that what was achieved in Cornwall could be recreated in another country. “Today I’m not sure this is the case, it requires massive work to convince the TSO and the DNO to share information about their grids.”
N-SIDE is involved in similar pilots in Belgium, Portugal and Germany, and it is working with grid operators to demonstrate the benefits of market-clearing systems, he says. “Usually, they’re not ready to share information needed about the topology of the grid. We usually succeed, but it takes quite a lot of time, for regulatory reasons,” he says. The UK in contrast has quite advanced innovation programmes from a regulatory perspective, while other countries would be more limited in terms of what type of research they will allow, he adds.
Although the platform was designed to be simple to use for non-energy market players, both Atkinson and Rosen see the near-future local flexibility markets being handled by service aggregators, rather than individual businesses or households. The financial value of each market transaction is relatively small and probably not worth the time it would take participants to familiarise themselves with new markets, Atkinson explains. “Aggregators will develop their own algorithms for determining how much flexibility is available from their pool of assets, what is required, and what they can bid into the platform,” he adds. •
TEXT
Catherine Early
PHOTO
Hannah Woolley