OFGEM Data Centre Queue: 16 September Is Not a Consultation Deadline. Its an Investment Decision.

Now that we have all returned from the summer holiday, 16 September is not far away.

For most people, it is simply the closing date for an Ofgem consultation. For UK data-centre developers, investors and landowners, it may mark a more consequential shift: a grid connection can no longer be treated as a low-cost option on future capacity.

Ofgem is consulting on proposals designed to remove speculative data-centre projects from the demand-connections queue. The proposal includes a Data Centre Commitment Fee of £237,500 to £712,500 per MW, alongside hard evidence of a credible end-user, procurement of long-lead electrical equipment and the financial and technical capability to develop the project. Ofgem's consultation closes on 16 September 2026.

Therefore, at 200 MW, that implies a proposed upfront and possibly abandoned commitment of £47.5m to £142.5m. This is not an administrative charge. It is a board-level capital, delivery and risk-allocation decision.

The projects that will win are not those with the best planning drawings or the most optimistic load forecast. They will be the ones able to demonstrate that the data centre, power system, equipment, capital and customer commitments can all move together.

A connection offer is no longer enough

The old development sequence: secure land, obtain a connection offer, advance planning, then solve the detailed engineering and commercial work backed by devex funding that values work completed. That sequence is no longer robust for a large power user.

The proposed Ofgem regime makes the missing work VERY visible. A developer will need a credible compute end-user, quality evidence that long-lead equipment such as switchgear is being procured, and financial and technical capacity to deliver the scheme. Ofgem is, in effect, asking whether the project is real enough to deserve scarce network capacity.

That is where many schemes will fail. Not because the land is unsuitable or the planning case is weak, but because the critical-path decisions have been taken in isolation and therefore the long lead information is not yet available. A grid adviser may optimise the connection application. A lawyer may draft the contracts. An equipment supplier may reserve manufacturing capacity. A funder may assess the capital structure. None of those steps alone creates a deliverable power plan.

The practical test is sequencing. Does the grid route support the energisation date? Does the energisation date support the customer commitment? Does the customer commitment support the capital raise? Does the equipment order match the technical design? Can embedded generation, BESS, flexibility or a private-wire route protect availability if the network timetable moves? And can you document and explain this all coherently?

This is not theoretical. Nscale’s £2bn Loughton AI data-centre project, planned at 50 MW and scalable to 90 MW, has reportedly faced a grid connection delay that threatens its 2027 opening. Microsoft was expected to be an anchor tenant, while Nscale has discussed interim solid-oxide fuel-cell power with Bloom Energy. Data Centre Review reports that the delay could create financial exposure because capacity has already been committed.

The lesson is not that every data centre should install fuel cells. It is that a customer commitment without a credible power-delivery plan can turn a grid delay into a commercial liability.

The numbers are large. The data is still imperfect.

Ofgem’s July announcement describes a 125 GW demand-connections queue, including around 73 GW of data-centre demand. It says data centres accounted for at least 80 GW of the growth in demand applications. Ofgem

Those numbers rightly attract attention. But smart decisions require more than repeating them.

DESNZ has now published its first meter-based estimate of data-centre electricity consumption. It estimates that Great Britain’s in-scope data centres consumed 4.5 TWh in 2024, or 2% of grid electricity. Slough alone consumed 1.3 TWh, equal to 65% of all electricity drawn from the grid in the borough. DESNZ Energy Trends

The same publication explains why published demand estimates differ so widely. DESNZ measures metered consumption. NESO’s higher power estimate is based on installed capacity and utilisation assumptions. Enterprise data centres are excluded from the DESNZ series. The lesson is straightforward: data-centre power is now too important for vague statistics. Any investment case needs a clear definition of demand, location, availability requirement and operating profile.

Slough is the warning. Once a local system reaches that concentration of demand, power is no longer a utility bill or a procurement workstream. It is the asset strategy.

AI Growth Zones are not a substitute for delivery

AI Growth Zones remain important. Government policy has indicated meaningful locational electricity discounts for a 500 MW data centre, including £24/MWh in Scotland, £16/MWh in Cumbria and £14/MWh in the North East. Baker McKenzie’s analysis of the policy paper sets out the design.

However, a discount is not power. Nor is a designation an energisation date.

The right question is not “Can this project qualify in an AI Growth Zone?” It is “What combination of connection route, network reinforcement, interim generation, flexibility, planning and commercial structure makes the promised capacity deliverable?”

That is a much tougher question. It is also the one that investment committees, credit providers and end-users will increasingly ask before they commit.

What should happen before 16 September

The immediate task is to assess whether your live project can survive the proposed test.

Three questions should be answered:

1. Connection credibility: Can the project evidence a genuine end-user, technical standard, long-lead equipment plan and credible delivery capability?

2. Power security: If the grid date moves, what is the least-cost combination of BESS, gas or fuel-cell generation, flexibility, private wire and commercial contracts that protects the business case?

3. Capital discipline: What commitment can the project make without destroying returns or creating an unacceptable stranded-capacity risk?

CM Energy Insight works on this intersection: transmission and distribution connection strategy, embedded and behind-the-meter generation, wholesale power and PPA structures, and the capital plan needed to take a power-intensive asset to FID.

The industry has spent two years discussing the queue. The more important question is now whether a project is credible enough to stay in it.

Working on a UK data centre or large-load project? Contact Chris Moore at CM Energy Insight for a confidential discussion on connection credibility, power security and delivery sequencing.