National Grid Confronts AI-Driven Capacity Crunch
Context and Chronology
Large-scale compute deployments are colliding with transmission limits across the UK and in parts of Europe and the U.S. as cloud and AI operators signalled demand that many grid planners did not foresee. Waiting lists for formal connections near the British system now include roughly 30 GW of prospective load; industry trackers and operator statements suggest a parallel surge in planned captive generation and private firming capacity worldwide. The connection backlog accelerated through 2024 and into 2025, with some metrics pointing to a virtual-queue expansion measured in the low hundreds of percent during early 2025.
Network owners are deploying operational fixes to buy time: re-conductoring, rerouting flows, dynamic line ratings and targeted switching to unlock marginal megawatts. Those interventions can blunt immediate constraint but do not remove the hard thermal and stability limits on many high‑voltage corridors. Major transmission projects remain slow: planners and operators expect typical new interregional builds to take on the order of 7–14 years from planning through commissioning in the UK context, leaving a multi-year window in which alternatives will dominate project choices.
Two market responses have emerged in parallel. First, hyperscalers and large colo operators are increasingly vertically integrating power stacks — buying renewables, pre-booking firming plants, and siting batteries and engines onsite to guarantee energisation. Trackers place planned captive capacity at roughly 56 GW across major markets, with industry estimates that about three-quarters of that capacity is likely to be gas-fired in the near term. Second, policymakers and system operators are talking about reshaping connection rules and locational signals: proposals include priority-for-growth schemes that would let some large projects move ahead of earlier applicants, conditional connection agreements that tie energisation to mitigation commitments, and locational pricing to price in system impacts.
That policy debate carries distributional consequences. Ministers are reported to be considering rules that prioritise projects judged to deliver jobs or economic output — a step that would concentrate scarce connection headroom with deep-pocketed energy-intensive investors and risk deferring housing and smaller community loads. At the same time, system operators have recommended siting the very largest inflexible loads where they can act as local sinks for curtailed renewables, an operationally driven reframing of siting choices that tightens the link between connection policy and dispatch outcomes.
Supply-chain dynamics amplify short-term pressure. Turbine and balance-of-plant lead times have lengthened substantially — some supply chains now signal waits of up to seven years — and manufacturers’ announced capacity lifts (for example, selective increases from major OEMs) are unlikely to erase near-term shortages. The result is a market where well-capitalised buyers can pre-book slots and crowd out smaller developers, raising costs and delaying utility-led reinforcement projects.
Operational impacts are already visible. Very large, inflexible data‑centre campuses increase balancing actions, complicate reserve management and raise system-wide costs that can be socialised unless contracts and queue rules internalise them. Developers respond with a mix of mitigations — private wires, on‑site generation, battery pairing, staged energisation, demand‑response commitments and bespoke commercial deals — each of which shifts costs, complexity and execution risk in different directions.
Across jurisdictions the practical effect is a near-term market reallocation: winners include turbine and battery OEMs, engineering and transmission contractors, and hyperscalers with balance-sheet heft; losers are smaller developers, some utilities and projects reliant on public reinforcements. If the current trajectory persists, expect a wave of captive thermal procurement, accelerated battery orders, more private-wire projects, contested queue-priority rules, and a patchwork of conditional approvals that reshape who pays for reinforcement and who gets energised first.
Read Our Expert Analysis
Create an account or login for free to unlock our expert analysis and key takeaways for this development.
By continuing, you agree to receive marketing communications and our weekly newsletter. You can opt-out at any time.
Recommended for you

Global AI datacenter boom risks oversupply and wasted capacity
Rapid expansion of GPU‑heavy datacenter capacity for generative AI is outpacing measurable production demand and colliding with local permitting, financing and grid constraints. Absent tighter demand validation, better utilization mechanisms and coordinated grid planning, the sector faces lower returns, schedule risk and heightened public pushback.

NTT Global Data Centers to Scale Capacity to 4 GW, Targeting AI Demand
NTT Global Data Centers plans to deploy roughly 4 GW of nameplate IT power across 34 projects within about two years, accelerating a shift to GPU‑dense, high‑power facilities. The program sharpens near‑term pressure on interconnection, transformer and cooling supply chains and forces an energy‑strategy choice—embedded generation, contracted renewables, or hybrid solutions—that will determine usable capacity and local political risk.
UK grid operator warns that very large data centres could raise energy bills
The chief executive of the UK’s electricity system operator warned that ultra-high-power, inflexible data‑centre campuses can increase wholesale and balancing costs unless they are sited where the system already has flexibility. He urged concentrating the very largest users — those drawing around 1 GW — near locations where renewables are often curtailed or where operators can absorb variable output, while noting developers and regulators can also blunt impacts through bespoke connection arrangements and demand‑side measures.

AI data centres prioritized for grid access; builders warn housing squeeze
UK proposals would let designated ‘strategic’ projects jump the national electricity-connection queue, favouring AI data centres, EV charging hubs and industrial electrification. Regulators and the system operator stress conditional terms, locational and commercial levers to internalise costs, while builders warn the shift risks prolonged delays to new homes in constrained areas.

AI data centers push U.S. electricity costs higher, Goldman projects
Goldman Sachs warns that rapid expansion of AI-focused data centers is a major contributor to recent and projected electricity demand growth, driving notable wholesale and retail power price increases through 2027 and easing in 2028. The pressure is uneven: concentrated buildouts have spurred local political pushback and roughly $64 billion of delayed projects, raising financing and underutilization risks that will shape who ultimately bears higher bills.

Texas moves to reassess data‑center grid approvals, injecting fresh uncertainty into investments
Texas regulators and grid managers are reviewing recent permissions for large data‑center power connections, a move that could slow project timelines, add technical or financial conditions, and amplify already growing local opposition seen nationally. The reassessment comes as permitting fights and community pushback across multiple states have contributed to roughly $64 billion of delayed or canceled U.S. data‑center projects, raising the stakes for how upgrade costs and mitigation obligations are allocated.

Applied Digital and Partners Commit 1.2GW Gas Plants to Power AI Campuses
Major data‑centre operators are building on‑site gas generation to meet AI power demand, with a reported $2.4B project delivering 1.2 GW . That move signals accelerating direct fuel deals, higher natural‑gas volumes, and greater strain on emissions targets.

Administration pushes tech firms to underwrite $15B in PJM power capacity
The federal government is urging PJM to procure about $15 billion in new generation through 15‑year capacity contracts and wants major technology companies to commit even if they don't immediately need the electricity. The proposal aims to shore up supply amid surging data‑center demand but shifts long‑term financial risk onto corporate buyers and could lock the region into large, slow-to-build fossil assets instead of flexible renewables.