Ofgem Grid Fee Would Make Data-Centre Developers Prove Demand

Data Center Knowledge covered an Ofgem proposal that would make large British data-centre grid applicants put refundable money and delivery evidence behind their connection requests, shifting a crowded queue away from simple chronological order.

The regulator is consulting on a Data Centre Commitment Fee for large developments.

Applicants would have to show that funding, commercial preparation and supplier work are moving far enough to justify a place in the queue; money paid under the scheme would come back at energisation and be lost if a project dropped out early.

The queue pressure gives the policy its urgency.

Contracted demand offers rose from 41 GW in November 2024 to 125 GW in June 2025, and data centres made up roughly 73 GW of the pipeline.

Ofgem wants to stop speculative applications from occupying network capacity, slowing credible schemes and distorting future grid planning.

The proposed cost is large enough to change who can stay in line.

A Walker Morris analysis turns the fee structure into a project-level hurdle: a 100 MW development would need to post about £23.8 million to £71.3 million before counting existing transmission security obligations.

DCByte research manager Kristina Lesnjak framed the change as a move from "who applied first" toward a test of whether a developer can show a credible route to delivery.

Clear evidence requirements and firm deadlines become central if the system is to remove weak applications without turning capital strength into the only proof that matters.

That is where the market effect becomes uneven.

Established operators, hyperscalers and developers with a history of delivering large campuses are better positioned to reserve capital for years and satisfy milestone checks.

Smaller developers and new entrants may still have viable sites, but a high deposit can push them behind strategic projects or force them to redesign their power strategy before land plans mature.

The same logic is spreading beyond Britain.

Data Center Knowledge linked the Ofgem move to PJM's US proposal for large loads, where future users may need to bring generating capacity, sign bilateral supply deals or accept curtailment in shortages.

In both cases, grid access is becoming a test of power support rather than a passive utility step after site selection.

For data-centre developers, the practical question now moves upstream.

Sites in constrained clusters may need paired generation, non-firm connection plans, shared financing or a shift toward less congested regions such as northern England and Scotland.

Final rules still need to show how the regulator would separate speculative queue-squatting from credible smaller schemes that lack hyperscaler-scale capital.