Britain’s data centre pipeline has run into a problem that has little to do with computers. The country has far more demand for electricity connections than the network can accommodate. Ofgem says applications jumped from 41 GW to 125 GW in less than a year, with at least 80 GW of those applications concerning data centres. The regulator now wants projects to back up connection requests with evidence of progress. That could change more than the queue. It could change how Britain designs, funds and builds its next data centres.

Britain is putting a price on holding grid capacity

Ofgem’s proposal starts with a practical concern: a connection request can occupy capacity long before a building is ready to use it.

Why the 125 GW queue has become a problem

The 125 GW figure needs context. It does not mean Britain has committed to building 125 GW of data centres. Many applications will change, slow down or disappear.

That creates a planning problem. NESO and network companies need to decide where to strengthen the system, but they cannot easily tell which projects will actually arrive. Meanwhile, a viable project can be deferred behind an earlier application that has little chance of meeting its connection date.

Ofgem’s wider demand-connections work now puts queue curation at the centre of reform. For data centres, the regulator is proposing a separate set of financial and progression requirements. The consultation remains open until 16 September 2026.

What the proposed commitment fee changes

Ofgem’s July 2026 proposal would require relevant data centre projects to secure a commitment fee from acceptance of their connection offer until the project is ready to connect. The fee would not return if the project failed to progress under the proposed rules.

That changes the economics of keeping a connection alive. A developer can no longer treat a grid offer as something with value but little carrying cost. However, the proposal is still under consultation. The final design could change before Ofgem makes its decision.

Ofgem wants to make speculative projects less attractive and free capacity for projects that can actually move forward. The fee puts a price on uncertainty, rather than on electricity itself in practice.

The fee could change when developers spend their money

The more interesting effect may come before construction, when developers decide how much capital to risk while several parts of the project remain unresolved.

When does a data centre become financially real?

A developer can control a site without a final customer. It can have financing discussions while planning remains unresolved. Grid access sits across all three issues.

The proposed rules add another layer. Developers must demonstrate financial and technical feasibility and show proof at specified intervals. That might give investors a clearer way to separate a working development from a long-term option.

For instance, think of a 200 MW campus with a good site and some early customer interest. The developer may still have substantial costs ahead. A commitment fee, equipment deposits, and engineering work could compel a financial decision earlier than anticipated.

Could the reforms favour developers with deeper pockets?

This is where the policy becomes less clear-cut.

A big hyperscaler can digest an early order for equipment more easily than a smaller developer. An infrastructure fund or one with a strong balance sheet may be able to take on risk that a newer operator cannot. Meanwhile, a developer who has signed a customer contract is exposed to a different risk than one still in the business of marketing capacity.

Ofgem says it is aiming to weed out non-viable projects and increase access for projects that are able to connect. Even so, the market impact is worth watching. Developers may require stronger customer commitments, earlier financing and more skin in the game before they can carry a big connection.

Britain could end up with fewer speculative projects & a more concentrated development market.

Grid milestones could change the construction programme

This is where the reform becomes especially relevant to design & engineering teams. Proposed evidence requirements can reach into decisions that normally sit inside the project programme.

Why switchgear could become proof that a project is moving

Ofgem cites the purchasing of long-lead electrical equipment, such as switchgear, as possible evidence of progress. That’s significant.

Switchgear is a function of the electrical design, fault levels, protection philosophy and equipment layout. Such decisions need to be made; a project can not simply order it on a whim before those decisions mature. However, waiting on every last design detail could jeopardize the grid slot if procurement is tied to a milestone.

That does create sincere tension. Teams might have to freeze portions of the electrical design sooner. They might also have to give up a little less flexibility when it comes to choice of suppliers or equipment. A switchgear purchase order might become more than just a purchase. It might be proof that a project is really progressing.

What happens when the grid timetable and building timetable diverge?

A data centre does not develop on one clock.

The grid operator has a connection programme. The planning team has a separate approval process. The water company might have separate infra work. The developer may yet bargain with buyers. Civil construction has a different order.

For example, electrical goods might be delivered prior to planning permission being granted, resulting in the developer having costly equipment sitting in a warehouse. It’s the other way around too. A building can be in late construction stage while a network upgrade is on hold on a different programme.

So the practical change is not an extra dashboard. It is a single joined-up programme comprising grid milestones, planning, finance, procurement, civil works and commissioning. This is particularly important for large London and Hertfordshire campuses where multiple infrastructure projects may impact on the same delivery timetable.

Britain may start building data centres differently

The biggest consequence may sit beyond the fee itself. Developers could change the size, phasing & power strategy of their campuses to reduce exposure to a difficult connection process.

Could phased campuses become the safer bet?

A massive campus can book a huge future workload years before the full site is in place. That model works when demand is predictable, and grid access is relatively easy to obtain. It gets tougher when a builder has to keep demonstrating progress.

Phasing provides an alternative route. A developer could build the initial 50 or 100 MW, demonstrate the demand, then build the rest. That can keep capital closer to the point of revenue. It can also reduce how much capacity goes unused.

That comes with a catch. Subsequent stages will also require space for substations, a cooling plant, roads, water systems and utility corridors. Good masterplanning therefore matters more, not less.

Will more developers build around the grid instead of waiting for it?

Britain’s AI Growth Zone scheme already permits applicants to use credible behind-the-meter solutions to show they are able to secure at least 500 MW of power by 2030. Applicants must substantiate that path with technical information, supplier contracts and permit applications. They must also demonstrate sufficient water, wastewater provisions and a minimum of 100 acres of land.

That opens the door to dedicated generation, private-wire arrangements and hybrid supply models. These solutions introduce new costs and risks, but they can change the grid conversation.

Above all, they shift responsibility. A big developer might have to think like an infrastructure provider as well as a datacenter operator. It may also have to deal with generation, electrical grids, water and building interfaces that a traditional network connection would have isolated somewhere else.

To sum up

Britain’s suggested data-centre colocation reforms could disrupt the market well before the first building contractor shows up. A fee can make capacity holding more costly. Milestones for progress may draw equipment orders and design decisions ahead. Developers may respond with smaller phases or more private power infrastructure.

The 4th Data Centre Design, Engineering & Construction Summit UK is scheduled for 6–7 October 2026 in London, UK. It unites developers, engineers, contractors, and infrastructure experts to discuss the delivery challenges that will define Britain’s future data centres. The focus on cross-disciplines makes the timing particularly appropriate.