Datacentres, AI Growth Zones, and the UK Power Grid

​

50 GW of Datacentre Demand Just Hit the UK Grid Queue

The AI infrastructure arms race has arrived on British shores — and it is reshaping the UK's power system at a speed that network planners, regulators, and energy investors are still calibrating. Contracted demand offers on the UK grid rose from 41 GW in November 2024 to 125 GW by June 2025, driven overwhelmingly by hyperscale datacentre projects. In February 2026, The Register reported that the total datacentre-related grid demand queued across the UK had reached 50 GW — equivalent to more than twice the UK's average electricity demand.

​

This is not a distant projection. It is happening now. And the implications for energy developers, investors, and network planners are profound.

 

The Scale of the Demand Shift

Oxford Economics forecasts that UK datacentre electricity consumption will grow from 5 TWh in 2023 to 26.2 TWh by 2030 — equivalent to 8.8% of total UK electricity demand. NESO estimates that datacentres could drive up to 71 TWh of additional grid demand over the next 25 years.

​

The UK already has 477 datacentre facilities in operation, making it the world's third-largest market — behind only the US and Germany. That base is set to expand rapidly, with major developments planned across London and the South East, Greater Manchester, Wales, and Scotland.

​

Scotland is emerging as a strategic location of particular significance. Its access to renewable energy from wind and hydro, combined with a lower land cost base and emerging government support structures, is attracting projects at scale. In Scotland alone, 4,749–5,249 MW of datacentre capacity is currently in the planning system — already exceeding Scotland's entire peak winter electricity demand. The Killellan site in Argyll, a renewable-powered sovereign AI cloud planned to scale from 100 MW to more than 2 GW, illustrates the ambition.

 

AI Growth Zones: Opportunity and Complication

The government's response has been to designate five AI Growth Zones (AIGZs) across England, Scotland, and Wales, with planning reform and energy access incentives designed to fast-track development. The zones offer energy price discounts — up to £24/MWh in Scotland and £16/MWh in Cumbria and the North East — as well as access to a new Connections Accelerator Service that aims to bypass the standard grid connection queue.

​

The North Wales AI Growth Zone is particularly notable. It is the only zone explicitly linked to a specific energy supply solution: the first Rolls-Royce SMR fleet at Wylfa. The government's framing positions Wylfa's SMRs and the AI Growth Zone as a co-located clean energy corridor — compute powered by 24/7 nuclear generation, designed for the load profile that AI workloads demand.

​

However, the BBC's March 2026 analysis captured the tension: prioritising AI datacentres for grid capacity and planning permission could displace other uses, including housing and conventional energy development. The Connections Accelerator Service effectively creates a two-tier connection system — and the criteria for "strategic" status are not yet settled.

 

The "Bring Your Own Generation" Model

The most sophisticated hyperscalers are not waiting for the grid. As detailed in CM Energy Insight's earlier commentary on Google's $4.75 billion acquisition of Intersect Power, the dominant strategic model is now co-location: solar, battery storage, and backup generation sited alongside the datacentre, connected to the grid at the substation boundary but keeping most power flows internal.

​

In the UK, this model is being explored through private-wire connections, behind-the-meter generation, and microgrid configurations. National Grid's Data Centre Impact Study has modelled the implications of widespread adoption. For UK energy developers, this creates a direct opportunity: battery storage projects, flexible generation, and renewable assets near planned AI Growth Zones are now being actively evaluated as potential anchor assets for co-located datacentre power supply.

​

The commercial structure — long-term, creditworthy, high-capacity offtake from hyperscalers — could provide the revenue certainty that merchant BESS and renewables projects have historically lacked.

​

The intersection of AI infrastructure and energy supply creates complex investment questions. CM Energy Insight provides the cross-domain advisory to navigate them.