Summary
- DESNZ estimates that GB data centres consumed 4.5TWh of grid electricity in 2024, around 2% of GB grid consumption.
- London and the South East accounted for 77% of data centre electricity use, with Slough alone consuming 1.3TWh.
- The figures give grid planners and policymakers a firmer baseline for data centre concentration, reporting, and future connections.
The Department for Energy Security and Net Zero has estimated that data centres consumed 4.5TWh of grid electricity in Great Britain in 2024, equal to about 2% of total GB grid electricity consumption.
The figures were published as a special feature article in the June 2026 edition of Energy Trends and cover colocation, managed hosting, and hyperscale data centres from 2020 to 2024. DESNZ describes the work as official statistics in development.
The DESNZ release shows strong regional concentration. London and the South East accounted for 77% of GB data centre electricity use in 2024. Slough alone consumed 1.3TWh, or 29% of GB data centre electricity consumption and 65% of all grid electricity consumed in Slough.
Measured load changes the debate
Slough has long been treated as the UK’s main data centre cluster by reputation and market activity. DESNZ has now put official numbers around that concentration. The South East consumed 1.8TWh of data centre electricity in 2024, with Slough accounting for most of that regional total.
The data gives planning and grid discussions a firmer base. Data centre demand is often described through connection queues, announced megawatts, or developer pipelines. Consumption statistics show what operational sites are actually drawing from the grid. They do not capture all future pipeline demand, but they make the existing load visible.
That distinction matters for grid reinforcement and local planning. A region can have a large future connection queue with modest current consumption, or a large existing load and limited headroom. Slough now has official evidence of both concentration and dependency, making further growth harder to separate from electricity-network planning.
Reporting moves closer to the facility
The publication arrives as data centre electricity use has moved higher up the UK policy agenda. Government, regulators, and network operators are trying to reconcile AI growth, cloud demand, clean-power targets, connection reform, and local opposition to large schemes.
A national 2% share may appear manageable in isolation, but regional concentration changes the operational picture. Grid reinforcement, substations, heat rejection, water use, emergency planning, and construction pressure fall unevenly. Slough, London, and the South East carry a different infrastructure burden from regions with little existing data centre load.
The figures may also increase pressure for more granular reporting from operators. PUE and renewable-energy claims have limited value without location, load, grid mix, demand timing, water strategy, and heat-reuse evidence. As data centres sit inside the UK’s critical national infrastructure framework, the expectation for evidence-based reporting is likely to rise.
DESNZ’s work is still labelled statistics in development, and the methodology may evolve. Even so, the publication creates a reference point for future connection decisions, local plans, energy forecasts, and sustainability reporting. Proposed AI campuses and hyperscale expansions can now be weighed against a measured 2024 baseline rather than a purely speculative demand curve.
