The data centre stress test: Power, planning and the limits of legacy advantage

The UK wants and needs data centre expansion, yet the system required to support it is increasingly misaligned. inspiratia’s “Data Centres: Power, Risk & Resilience” event, which took place in London on 18 June, examined a sector constrained by both local opposition and by deeper challenges in power, policy and infrastructure delivery. On regulation, planning and market design, one conclusion stood out: the binding constraint is no longer land or permitting alone but power, and whether the UK can deliver it.

From planning challenge to infrastructure reality

The UK’s data centre debate is a stress test of whether Britain can still build the energy, grid and regulatory systems on which a digital economy depends. The familiar framing centres on the development challenges data centres face amid local resistance. The more consequential story lies beyond that lens, in a sector now exposed to a far broader structural problem: power has become the major binding constraint.

The West London cluster, and Slough in particular, now faces a false trilemma: the narrative of choosing between building new residential areas, expanding Heathrow airport or installing more data centres. The real policy failure is not that one use is crowding out another, but that the UK still lacks a credible system for enabling all of them at once.

The market is already deeply embedded. The borough of Slough hosts more than 30 data centres, making it one of Europe's most significant data centre clusters. In that context, data centres are no longer an unfamiliar land use but part of the local ecosystem. Many Slough sites sit on long-established industrial land or within a simplified planning zone. This focus on brownfield sites allows for more public buy-in than an unstructured approach that pits development against Green Belt protections. Even so, data centres are still not widely understood as infrastructure, despite their role as essential systems that support everyday economic life. Public perception has yet to catch up to that reality.

Putting data centres and AI at the heart of the UK’s growth strategy, without an accompanying public-facing narrative about the infrastructure required to achieve it, has left a vacuum filled with frequent objections to large buildings, heavy energy use, water consumption and too few visible local jobs.

Upgrading existing data centres with more powerful racks, or building new facilities, could triple a site’s consumption overnight, often requiring additional energy and grid assessments. That raises a central question: whether building up baseload generation to power data centres is compatible with the continued expansion of renewables.

Ultimately, the constraint lies in the energy mix and in whether a low-carbon mix is achievable through accelerated integration of renewables or nuclear power. This, in turn, places weight on both the planning process and the broader public consultation process.

Policy is moving fast, but not yet coherent

Current UK power reforms aim to clean up the grid connection queue and accelerate viable projects. Following the overhaul of the connections process under the TM04+ system, which replaced “first-come, first-served” rules with priority-based access, regulator Ofgem has launched a broader demand connections reform programme. Ofgem and the National Energy System Operator (NESO) are consulting on whether to introduce a new Independent Transmission Owner (ITO) licence that would allow data centres to self-build or commission high-voltage connections to the grid.

There is, however, a distinction between regulatory incentives and political coherence. At the political level, the debate appears less mature, still dominated by the rapid surge of data centres and AI without a sufficiently developed plan for how the power system should support that ambition. In this environment, picking which projects qualify as strategically important is as much a political question as a technical one.

A comparison with France is instructive. France holds a far more strategic position, on the assumption that investors ultimately follow a simple equation: cheap power, available land, fast permitting and clear grid access. France is treating data centres as a sovereign necessity, while the UK is still struggling to align its regulations.

London’s standing as one of Europe’s top financial hubs, and the country’s large market for AI inference and low-latency digital services, are real strengths. But demand alone will not keep capital anchored if costs and connectivity remain uncompetitive.

The UK must work to maintain its advantage

The most contentious question is whether on-site generation could offer a workaround. Gas-backed generation for data centres is both technically and legally feasible in the UK, even if politically awkward. There is also a mismatch: the Department for Energy Security and Net Zero (DESNZ) requires a 95% renewable energy connection to qualify for the AI growth zone framework, even though the primary reason developers seek growth zone status is to obtain grid connections.

Infrastructure politics is often less about the asset itself than about siting, consultation and whether a local community sees any upside. Yet gas is not the long-term answer. The more durable strategic angle lies elsewhere: in co-location, private network buildout, and potentially an SMR (small modular reactor)-enabled nuclear solution. Two possible UK advantages stand out: emerging scope for independent transmission buildout, and the country’s strong nuclear licensing regime, particularly around SMRs. That does not indicate a market breakthrough yet, but it does suggest the opportunity may depend less on imitating the hyperscale expansion seen in the US and more on creating investable models that combine firm power, faster delivery and long-term resilience.

The deeper point is that the sector has become a mirror held up to the UK’s wider state of infrastructure. The country still has clear strengths: entrenched demand, a powerful financial centre, established data centre clusters and political recognition that the sector matters. But strengths inherited from history are not the same as a forward-looking strategy. The UK has often traded on legacy advantages for too long; in data centres, that habit may now be reaching its limit.