Ofgem Moves to Price Speculative Data Centres Off Britain's Grid Queue
Britain's energy regulator proposes commitment fees of up to £712,500 per megawatt to clear phantom demand from an overloaded connections pipeline.
Britain's electricity grid connections queue has become a bottleneck, and Ofgem wants developers to pay to stay in it. On Wednesday (2026-07-29), the regulator launched a consultation proposing an upfront "data centre commitment fee" ranging from £237,500 to £712,500 per megawatt — equivalent to roughly 2.5% to 7.5% of average project costs — aimed at flushing out projects with no realistic path to construction.6,4
The scale of the problem explains the urgency. Applications for electricity demand connections surged from 41 GW in November 2024 to 125 GW by June 2025, driven largely by data centre developments, Montel reported. Ofgem's concern is that a large share of those applications represent speculative positions — land-banking for grid access rather than firm investment intent — leaving genuinely investment-ready projects unable to connect.5
Grid queue congestion is not merely administrative friction. Every gigawatt of phantom demand sitting in the queue delays the real infrastructure Britain needs to hit its clean power targets and displace fossil generation. Ofgem approved a £24 billion grid investment programme last year to expand capacity and allow more renewables onto the system; letting speculative data centre applications absorb that headroom before a shovel is in the ground undermines the entire exercise.6
The commitment fee is designed to change that calculus. A charge worth tens of millions for the largest projects would force developers to decide early whether their plans are genuine. Those who proceed would effectively be pre-paying for access; those who withdraw would free up capacity that could be reallocated to credible projects. Ofgem has framed this explicitly as protecting serious investment from being crowded out.4
The backdrop is a doubling of anticipated demand. Data centre electricity consumption has grown fast enough to trigger a policy response across multiple jurisdictions, and Britain is following a familiar trajectory. The International Energy Agency projected that data centres could account for nearly 10% of US electricity consumption by 2030, up from around 5% now, and the dynamic in the UK market is structurally similar even if the absolute scale differs.2
Not everyone in the British policy establishment treats data centre demand as a problem to be managed. Mission Control chief Chris Stark, speaking at the FT Climate & Impact Summit on Thursday (2026-06-18), argued that data centre expansion would lower energy prices rather than raise them, and said regulators should be "more tolerant" of large consumers rather than steering them toward special utility tariffs. Panellists at the same event suggested data centres could return up to 60% of wasted energy to the grid — a flexibility dividend that the commitment fee debate has largely overlooked.3
That flexibility argument has some quantitative backing. A Montel study published on Monday (2026-05-18) found that flexible operation of data centres could cut their peak power demand contribution by up to 45% by 2035, avoiding the need for 4 GW of fossil fuel back-up capacity. If that potential is real, a policy designed primarily to reduce queue size could inadvertently deter projects that might have operated as grid assets rather than pure load.1
Ofgem's proposal is a consultation, not a final rule. The outcome will depend on how the regulator weighs the short-term queue management objective against longer-term questions about which projects get built and under what terms. Developers with genuine capital behind their plans may welcome the fee as a barrier that filters out the noise competing for the same grid slot. Those still in early-stage fundraising will feel the squeeze hardest.
The political dimension runs in both directions. Restricting speculative data centre connections could draw criticism from technology investors and AI infrastructure advocates who see Britain's grid queue as a competitive disadvantage. Doing nothing risks wasting years of grid expansion spending on paper demand that never materialises, while shovel-ready renewables and industrial electrification projects queue behind it.
The consultation outcome, and how the fee is ultimately calibrated within that £237,500 to £712,500 per megawatt range, will determine which type of data centre operator — speculative developer or committed hyperscaler — finds Britain's grid workable. Whether the upper end of that range is set high enough to actually deter the largest phantom applicants, or whether it becomes a manageable cost of holding a queue position, is the question left open.6,4