Government targets speculative demand in grid queue

Britain’s electricity connections problem has moved beyond queue management and into a more political argument about allocation. After transmission demand applications rose by 460% in the six months to June 2025, the government has opened a consultation aimed at stopping speculative demand projects from clogging the system and moving viable, strategically important loads forward faster. By the end of June, the transmission demand queue stood at 96GW, with a further 29GW sitting at distribution level.

The headline targets are AI data centres, EV charging hubs, and electrified industrial sites, but the shift is broader than a clampdown on speculative server capacity. Ministers want tougher conditions for joining and remaining in the queue, while Ofgem is considering stronger financial requirements, including deposits or milestone-linked fees, to prove projects are real and progressing.

The consultation also proposes powers for government to publish a list of strategically important demand projects that can be prioritised as capacity is released or created.

From queue clean-up to control

That matters because the numbers have started to look detached from the system they are supposed to describe. Ofgem says contracted demand offers rose from 41GW in November 2024 to 125GW in June 2025, against peak electricity demand in Great Britain of 45GW on 11 February 2026. NESO’s evidence also points to around 140 data centres accounting for roughly 50GW in the queue, with only part of that capacity tied to schemes that have reached final investment decision.

The underlying problem is familiar enough to anyone watching connection reform over the past two years. If the private cost of securing a queue position is too low, speculative projects can sit on capacity that has value far beyond the seriousness of the application itself. The result is a queue that looks full on paper, a planning system forced to react to phantom demand, and viable projects in industry, transport, housing, and digital infrastructure pushed further back than ministers, network operators, or developers can now tolerate.

Under the government’s proposed approach, NESO and network companies would gain wider scope to reserve future capacity, reallocate capacity released when projects exit the queue, and prioritise strategic demand in future batched exercises. Alongside that, the related connections work now under way at Ofgem points towards more flexible, non-firm, or ramped connections, as well as wider use of self-build or developer-led high-voltage assets where that can accelerate delivery. The issue is no longer simply who joins the queue, but who can credibly be connected, when, and under which commercial and technical model.

Industry reacts

“It’s no surprise that the government is reviewing the demand connection process given the surge in recent months as the interest in data centres has boomed,” says Jamie Baxter, Associate Partner at Carter Jonas. “The challenge is determining which connections are ‘real’ and backed by investment, but we are certain that the UK will need more data centres to keep up with needs around AI, cloud computing and other uses for advanced GPU capacity.

“The approach of increasing financial commitments through project milestones seems like the right ‘stick’ to ensure that projects do progress with developers and operators, mirroring what we are seeing for generation projects. We will be watching for the results of Ofgem’s consultation which closes this week to understand this in more detail. Clarity on the AI Growth Zones and what are deemed priority projects is to be expected from a government which has made growth a priority.

“There is considerable potential for data centre developers to deliver high voltage networks and a great opportunity to provide competitive pricing and speed. We already know that Independent Connection Providers (ICPs) and Independent Distribution Network Operators (IDNOs) are speeding up delivery of connections and providing valuable commercial models at the distribution level so it would be positive to see that reflected in the transmission space.

“We hope that lessons have been learned from the delays experienced through the current Connections Reform process. Alongside this strategic view and policy reform, the government needs to ensure that there is sufficient resourcing to seize this growth opportunity for both data centres and the additional renewable energy generation sources required to support these developments.”

The consultation is clearly designed to impose more discipline on the front end of the process, but it also hints at a more practical change in how large demand customers might get connected, especially where developers are prepared to take on more of the asset build-out themselves. That is not a minor technical tweak; it points towards a system in which the ability to organise capital, land, planning, and connection works in parallel could matter as much as getting into the queue early.

Lawrence Turner, Director of Boyer, said: “The Government is right to clamp down on speculative grid connection requests if the queue has genuinely become clogged with projects that were never likely to move forward. Tightening the conditions for entering and remaining in the queue should help redirect grid capacity toward schemes that are genuinely deliverable. If applied robustly, that could materially shorten connection times for viable projects.

“However, the reforms go much further than tidying up the system. By signalling that sectors such as AI data centres and major industrial projects may be prioritised, ministers are effectively introducing politics into planning, by deciding which types of development get electricity first.

“That has real implications for developers. For many of our commercial and industrial clients, particularly those involved in data centres, logistics and advanced manufacturing, these changes could unlock schemes that have been stuck behind speculative projects in the queue.

“But there is a flip side. Housing delivery increasingly depends on access to grid capacity, and in some areas, power constraints are already slowing development. If strategic infrastructure begins to receive preferential access to electricity connections, there is a real possibility that new homes could find themselves competing with server farms for power.

“The question becomes, when electricity is scarce, do we power servers first, or homes? Both are strategic development.”

That warning is uncomfortable, but it is also close to the policy choice now being made in public. Government has been explicit that speculative demand can hold up housing, transport, manufacturing, and AI Growth Zones, which means future reform will not simply be about removing bad applications. It will also be about ranking good ones, and once that happens the queue stops behaving like a neutral waiting list and starts functioning as a live instrument of industrial policy.

The timetable is short enough to keep the pressure on. Ofgem’s current call for input closes on 13 March 2026, a fuller consultation on the detailed demand-side reforms is expected in spring, and the government’s strategic-demand consultation runs to 15 April 2026. Condemning speculative queueing is the easy part; the harder task, now heading into codes, licences, and project sequencing, is deciding whose demand the electricity system is prepared to treat as urgent.