Millions of electric vehicles, heat pumps, and smart appliances will soon need to coordinate their energy use in near-real-time to prevent local power grids from buckling under the strain of net-zero electrification. The problem is computational. Current systems cannot process the flood of data from millions of grid-edge devices quickly enough to balance local and national electricity flows. Without solving this, the flexibility those devices could offer—shifting charging times or reducing demand during peaks—remains largely untapped. FleXEdge aims to remove that bottleneck by designing new machine-learning methods and a cloud-to-edge computing architecture that can coordinate devices across seconds to months, and from a single feeder to the national transmission network. If successful, the project could make local flexibility markets work at scale. That would let distribution network operators reward device owners for adjusting their energy use, increase the grid’s capacity for renewables, and reduce the cost of the net-zero transition. National case studies in Great Britain and Luxembourg will quantify the potential value and produce policy recommendations for market reforms. The work is applied, not fundamental science—it targets a specific engineering barrier to a functioning clean-energy grid.
View original technical description
The FleXEdge project is focused on accelerating and reducing the cost of the net-zero transition by unlocking demand flexibility at scale from grid-edge devices including electric vehicles, electric heating, and smart appliances. Specifically, the project aims to address the computational bottlenecks which currently prevent electricity system operators from fully utilising grid-edge flexibility as a potentially major resource for local and national power system balancing. The electrification of transport and heating are key components of the UK and EU net-zero strategies. Over the next decade, millions of electric vehicles and heat-pumps will be deployed within local distribution networks. To manage this transition, all six UK distribution network operators are trialling local flexibility markets which can incentivise coordination between grid-edge devices to manage distribution network constraints and increase the hosting capacity for renewables and electric transport/heating. Ofgem's ambitious future vision is for local flexibility markets to support the widespread coordination of grid-edge flexibility, not just for managing local power flows, but also to provide national flexibility services. Research by the investigators and others has shown that grid-edge flexibility could be highly valuable if integrated into electricity system operation across spatial scales (local feeders to national transmission) and timescales (months- to seconds-ahead flexibility services). However, realising this value requires fundamental research gaps to be overcome related to the computational challenges of large-scale near real-time coordination and local-national flexibility market integration. FleXEdge will address these challenges through a research programme into the design and national integration of local flexibility markets, crossing disciplinary boundaries between distributed AI-based coordination, market mechanism design, and cloud-to-edge computing. The project will deliver scalable machine learning methods for coordinating grid-edge flexibility, linked with incentive mechanisms which reward grid-edge device owners for the power system value of their flexibility and data. These will be underpinned by a new computing architecture targeted at the domain-specific challenges which have previously limited the use of distributed AI for power system control applications. The methods and tools that are delivered will be valuable for electricity system operators and flexibility aggregators, as well as the wider energy research community. National case studies for Great Britain and Luxembourg will be completed to quantify the overall potential value of the new designs and to develop policy recommendations for market reforms which could unlock additional value. The project is a collaboration between Dr Thomas Morstyn at the University of Oxford, Dr Pierre Pinson at Imperial College London, Mr Davy Virdee at EPCC (the University of Edinburgh's centre for high performance computing and data science) and Dr Jun Cao and Dr Pedro Rodriguez at the Luxembourg Institute of Science and Technology (LIST). EPCC will provide state-of-the-art computing facilities and engineering expertise in optimising, scaling, and deploying data-intensive scientific software. LIST brings aligned capabilities in AI-based edge computing and hardware-in-the-loop testing. The project is supported by industry partners in the UK and Luxembourg, including power system operators (National Grid ESO and SP Energy Networks), smart grid technology providers (Piclo, Siemens and Typhoon HIL), and the Energy Systems Catapult, which supports energy sector collaboration, innovation, and policymaking.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know