Expert insight: AI, data centres and the UK’s twin crunch – energy and water
By Professor John Patsavellas, member of the IET’s Sustainability and Net Zero Policy Centre
The BBC’s recent article on data centres being expanded crystallises an awkward truth: the UK’s AI boom is colliding with the physical limits of our energy and water systems, especially in the south and east of England [1]. If we want the productivity upside of AI, we must manage siting, energy, and water with engineering discipline, before “workarounds” become the norm.
The Government’s AI Energy Council is a welcome step toward aligning compute growth with the power system [2][3]. But intent now needs standards that are transparent and enforceable: how projects source firm low-carbon electricity, how they shape loads to low-carbon hours, and how cooling choices respect local hydrology. Without that, we risk grid-bypass solutions and public resistance in water-stressed catchments.
Water risk is not hypothetical. Labour’s first AI “growth zone” at Culham is seven miles from the proposed Abingdon reservoir, prompting legitimate concern that we could trade public water resilience for private compute unless water is planned and metered as tightly as power [4].
A recent application for a 314MW hyperscale data centre in Thurrock shows why methodology matters. The Environmental Statement for the proposed hyperscale concludes “Minor Adverse – Not Significant” for greenhouse-gas impacts, in part because annual operational emissions are a small fraction of the overall UK’s national carbon budgets [5]. But IEMA’s 2022 guidance is clear: all project emissions contribute to climate change; “minor adverse (non-significant)” is a high bar, reserved for schemes aligned with a 1.5 °C trajectory, not simply those that are a small percentage of a big pie [6]. Planners and promoters should read those pages together.
Meanwhile, Britain’s electricity got markedly cleaner in 2024. Carbon Brief estimates average carbon intensity around 124 gCO₂/kWh with fossil generation at record lows [8]. That raises the bar on efficiency (PUE) and on when we run energy-hungry training jobs. Every avoided kWh, and every shift to low-carbon hours now delivers more decarbonisation than it did a decade ago.
There’s also a system-planning piece. Energy UK’s recent Powering the Cloud report calls for joined-up delivery of low-carbon power, networks, and data-centre capacity, underscoring that land, water and energy must be planned together—not sequentially [7]. DESNZ’s new evidence review reiterates the uncertainty around demand trajectories and the need to couple growth with clean, firm supply and timely grid reinforcement [9]. These aren’t “nice to haves”; they are the difference between durable public consent and policy whiplash.
This is the golden thread here: AI can be both a productivity engine and a climate ally but only if the sector treats energy and water as first-order design constraints, applies IEMA’s significance logic as written, and reports performance with the same regularity it reports uptime. That is the level of engineering discipline UK communities expect and deserve.
References (APA 7th; numbered)
[1] BBC News. (2025, February 5). Concern the UK’s AI ambitions could lead to water shortages. https://feeds.bbci.co.uk/news/articles/ce85wx9jjndo
[2] Department for Science, Innovation and Technology. (2025, April 8). AI Energy Council to ensure UK’s energy infrastructure ready for AI revolution. https://www.gov.uk/government/news/ai-energy-council-to-ensure-uks-energy-infrastructure-ready-for-ai-revolution
[3] Department for Science, Innovation and Technology. (2025, July 1). Technology and Energy Secretaries chair second AI Energy Council meeting. https://www.gov.uk/government/news/technology-and-energy-secretaries-chair-second-ai-energy-council-meeting
[4] Horton, H. (2025, January 13). Water shortage fears as Labour’s first AI growth zone sited close to new reservoir. The Guardian. https://www.theguardian.com/technology/2025/jan/13/labour-ai-datacentre-growth-zone-water-shortages-abingdon-reservoir
[5] RED Engineering (for Global Infrastructure UK Ltd). (2025, April). Environmental Statement—Chapter 5: Climate Change and Greenhouse Gases (Thurrock Data Centre, Land at the former Arena Essex). https://docs.planning.org.uk/20250625/107/SW5PMMQGMBJ00/16dj7we5p2sp2foh.pdf Planning Portal
[6] Institute of Environmental Management & Assessment. (2022). Assessing greenhouse gas emissions and evaluating their significance (2nd ed.). https://www.iema.net/media/xmgpoopk/2022_iema_greenhouse_gas_guidance_eia.pdfiema.net
[7] Energy UK. (2025, June 17). Powering the Cloud: How data centres can deliver sustainable growth. https://www.energy-uk.org.uk/publications/powering-the-cloud-how-data-centres-can-deliver-sustainable-growth/
[8] Carbon Brief. (2025, January 2). Analysis: UK’s electricity was cleanest ever in 2024. https://www.carbonbrief.org/analysis-uks-electricity-was-cleanest-ever-in-2024/ Carbon Brief
[9] Department for Energy Security and Net Zero; Europe Economics. (2025, August 14). Impact of growth of data centres on energy consumption. https://www.gov.uk/government/publications/impact-of-growth-of-data-centres-on-energy-consumption