Britain’s Grid Has a Queue Full of Phantom Data Centers

Britain’s energy regulator, Ofgem, is preparing to test whether the data centers crowding the country’s electricity connection queue are likely to be built. Its July proposal, due to be finalized after industry feedback ends in September, would require developers to provide a substantial nonrefundable deposit, identify customers in advance and prove that financing is available. For the largest projects, the deposit could reach hundreds of millions of dollars.

The intervention follows an extraordinary expansion of claimed demand. Ofgem says projects in the connection queue grew from 41 gigawatts in November 2024 to 125 gigawatts in June 2025. New data centers accounted for 73 gigawatts—about one and a half times Britain’s peak electricity demand last year. The government believes much of that demand may never become a working facility.

Scarcity turns a reservation into an asset. When grid capacity is limited and connection dates stretch into the future, a developer can control a valuable place in line before buying servers, securing a tenant or completing the capital stack. The queue then stops behaving like an orderly schedule of construction and starts resembling a thinly supervised property market whose plots happen to be measured in megawatts.

The cost of proving intent

Ofgem’s strongest case is straightforward: a project asking the grid to plan around enormous demand should demonstrate more than appetite. A nonrefundable deposit makes duplicate or casually inflated applications expensive. Evidence of customers and funding tests whether a proposed campus has someone prepared to use it and someone capable of paying for it. Used together, those rules could move viable projects forward and improve the forecasts behind substations, transmission lines and generation.

The industry’s objection also deserves its full weight. Data centers face grid delays in the United States and Europe, while Britain already has costly electricity and limited land. A deposit large enough to frighten off speculators may also favor the largest incumbents, which can immobilize capital more easily than a smaller but serious developer. Britain could clear its queue by making entry prohibitively expensive and then mistake the resulting quiet for efficiency.

The better test is progressive rather than theatrical. Deposits should rise with the capacity requested and become refundable when developers meet published milestones. Customer commitments should be credible without requiring every tenant to be fixed years too early. Financing evidence should be refreshed as costs and ownership change, while projects that repeatedly miss land, planning or procurement deadlines should surrender capacity instead of indefinitely revising the brochure.

These rules would alter the wager. Developers could still pursue uncertain AI demand, but they would pay more for oversized claims and recover that money by doing measurable work. If Ofgem calibrates the milestones well, the queue will become a forecast of construction rather than a registry of ambition. If it relies mainly on a giant entrance fee, Britain may discover only which companies can afford to speculate politely.