Doing the maths on Britain’s data centre ambitions
The data centre boom looming over Britain is nothing short of staggering. The scale and speed of projected investment, already encompassing about 40% of peak electricity demand in final investment decisions, is unprecedented, and promises Britain a head start in exploiting innovation in AI-powered technologies – unlocking a new era of economic growth.
But the maths of how all this would happen doesn’t quite add up. At least, not without serious trade-offs for communities and other national objectives, including energy affordability.
There are problems with powering ahead. Rising electricity prices are the first port of call for opposition to data centres. In the United States, reports that residential electricity prices have increased by 267% in areas near data centre clusters have led to large-scale pushback, including the consideration of outright bans on new data centre developments. The British electricity market is configured differently, so it is unlikely to face the same scale of upward pressures. But it is hard to imagine a world where data centres don’t contribute to electricity price increases.
This is because of the scale of energy demand data centres introduce. In the UK, the queue of data centre projects is around 100 gigawatts (GW), or double peak electricity demand. Building even part of this queue, e.g., the 20 GW that have secured financial commitments, would be significant.
The issue is whether new energy supply can keep up – but it is likely to be outpaced.
The US is, yet again, an illustrative example. The Independent Energy Monitor for the PJM grid (the largest electricity grid in the US, and home to the nation’s highest concentration of data centres) notes the role of supply crunches in raising prices, and that these constraints are almost entirely due to large load additions from data centres.
Do data centres work on a (carbon) budget?
So, the energy needed to power the projected fleet of UK data centres is immense. It is also unlikely to be met with renewable energy, despite claims to the contrary – which have unravelled under even superficial inspection. This is because of the stable, round-the-clock power required for data centres, which entrenches reliance on fossil-fuel generation. In the UK, this role will fall to gas – which will have meaningful implications for carbon emissions today and in the future.
Again, it’s the scale of deployment is what makes these issues significant. By 2035, back-of-the-envelope calculations from the Centre for Economic Transition Expertise suggest that just 10 GW of new data centres could add about 12 million tonnes of CO2-equivalent in emissions. For context, by 2030, the UK is targeting emissions of ~10 million tonnes from total energy supply.
We could certainly try to meet all new data centre demand with renewables. But the amount of land this would require is almost impractical. Meeting 20 GW of data centre-led energy demand entirely with onshore wind capacity could require about 600,000 hectares of land, which is roughly 2.5 times the size of Luxembourg. The resource demands imposed by data centres reflect technologies built for a world of abundance, not scarcity. This is already forcing policymakers into difficult decisions.
For example, driven by the confluence of constraints in gas supply and transmission constraints from recent heatwaves, parts of the US have had to restrict access to power for data centres.
In the UK, it is also becoming increasingly apparent that there may not be enough water for the nation’s data centre ambitions. This is particularly concerning as UK data centres use drinking water for cooling, in the context of a shortfall exceeding 220 million litres per day.
A single-minded strategy of aggressively pushing data centre expansion at all costs could lead to outcomes that are unacceptable. The interaction of data centre deployment with other central policy objectives should form the cornerstone of how, where, and how many of these facilities are built.
Key to all this, is transparency. Being upfront about emissions, energy requirements, how data centre projects are being prioritised, and their water usage, would go a long way to providing policymakers with the information necessary to make informed decisions on this highly disruptive, controversial, and yet, potentially hugely beneficial development.
Britain’s data centre and AI ambitions are not inherently flawed, nor mutually exclusive to the achievement of sustainability objectives. But we do need to stop pretending we can have both without a plan.
Pausing to reflect on how to best make the astronomical levels of investment in data centres work for and alongside the nation’s energy, water, and climate objectives is perhaps the only way to ensure data centres deliver on what they promise.
Isabel Jackson is a recent graduate working as a politics and media consultant, as well as a writer and editor at Comment Central. Having just completed her studies, she brings a fresh, energetic perspective to political commentary, quickly establishing herself as an emerging voice on current affairs, policy, and the evolving media landscape.
Most Popular
More from Technology
