Flexible Funding: New Entrants to Energy Networks Community
A total of two New Entrants to the Energy Networks Community awarded funding for innovative projects that explore artificial intelligence within the context of Energy Networks.
Congratulations to Xue Yong (University of Liverpool) and Victor Gutierrez-Basulto (Cardiff University).
Project date:
Awarded
AI-Guided Infrastructure Planning: Graph Neural Network-Based Modelling of Commuter Patters for EV Charger Allocation in Railway Networks
South Wales faces a critical “charging divide” hindering EV adoption, particularly for those without off-street parking. Our project tackles this by integrating EV charging with railway station car parks, leveraging cutting-edge Graph Neural Networks.
This AI-driven approach will create a comprehensive framework and practical tool to strategically roll out EV charging infrastructure. We’ll forecast EV adoption and charging demand at railway stations, identify optimal locations and capacities for chargers, and simulate future scenarios using Welsh Government data.
By considering socio-demographic factors, commuter patterns, and grid constraints, we aim to ensure equitable access to charging, maximising coverage and supporting wider socio-economic benefits. This project will deliver data-informed recommendations for efficient and fair deployment, ultimately accelerating e-mobility in South Wales
Project date:
Awarded
Smart Electrolysers, Smarter Grids: AIPowered
Catalyst Design for Dual Grid and
Hydrogen Gain
The UK is rapidly shifting to renewable electricity, with wind and solar providing nearly half its power. But these variable sources don’t always match demand, causing wasted energy—like in winter 2022–23, when enough wind power to supply over a million homes was lost due to grid limits.
A promising solution is to use this excess electricity to produce hydrogen via electrolysis. This “green hydrogen” can be stored and used later, and electrolysers can help stabilise the grid. However, current electrolysers degrade under fluctuating power, respond slowly, and rely on costly materials. This also limits optimisation of grid costs, stability, and reliability.
Therefore, this project will focus on optimising the grid’s dynamic performance by developing AI-driven tools that link catalyst atomic properties to real-time grid conditions. These tools will be built by analysing data on catalysts, grid behaviour, and weather patterns to reduce energy waste, cut costs, and improve reliability. This approach will accelerate the UK’s transition to a fully renewable energy future, benefiting industry, researchers, and society.
