The UK did the painful thing with its connection queue. Now comes the harder part.

About 700 GW of projects were sitting in Great Britain's connections queue at the start of 2025. That's roughly four times what the country actually needs. Most of it was speculative. Some of it had been in the queue for years. First-come, first-served had stopped functioning as anything other than a waiting room for ghosts.

So on 29 January 2025 the National Energy System Operator stopped accepting applications. Just paused the whole thing. By December the cull was done — NESO confirmed a retained pipeline of 238 GW, having moved roughly 217 GW of generation to "Gate 1" because the projects couldn't show land, consents, or progress. Battery storage survived best at 83 GW. Solar lost around 36 GW. The TMO4+ reforms went live in June. Ofgem's five-year settlement earmarks up to £90bn for transmission investment.

On paper, a functioning reset.

Except the interesting thing wasn't the cull. It was what was happening to demand at the same time.

The UK demand connections queue — the one data centres sit in — went from 41 GW in November 2024 to 125 GW in June 2025. Up 205% in seven months. NESO's call for input identified around 140 specific data centre facilities, totalling roughly 50 GW, most of which are expected to clear Gate 2 readiness.

Peak UK electricity demand on 11 February 2025 was around 45 GW.

That comparison is the one that should make site selection consultants put their coffee down. The demand queue, if half of it materialised on schedule, would exceed the largest day on which Britain has ever drawn power.

It isn't going to happen on those timelines. The physics won't let it. What happens instead is the political decision the CP30 Action Plan and the strategic-designation machinery are designed to surface: which demand projects get built this decade, and which get quietly deferred into the 2030s. The government has made its preference clear already. AI Opportunities Action Plan. AI Growth Zones with streamlined planning. Connections Accelerator Service. Demand connections reform moving from pure readiness (TMO4+) to strategic alignment, where operators near existing high-capacity parts of the grid get prioritised ahead of operators requiring fresh infrastructure.

I know this is going to annoy anyone working in housing or electrified industry, and I'm partly playing devil's advocate, but Britain has effectively decided data centres are the strategic priority. The policy documents won't phrase it that way. The outcome will look that way.

The question this raises isn't "will my site get built". It's "which version of me will the UK regime actually build".

Readiness evidence is now a live competitive surface. The Gate 2 Criteria Methodology rewards operators who can document land rights, planning consents, and capital commitment in specific formats by specific dates. Operators with good internal project tracking and fast legal turnaround clear faster than operators assembling the paperwork on request.

Strategic designation is zone-based. Northern and southwestern England and parts of South Wales show capacity gaps in the retained generation pipeline — that's where the grid is thin. Conversely, sites near strong existing transmission and established demand clusters will be prioritised for new DC builds. If your site selection deck still defaults to Slough because the fibre is there and the historical DC ecosystem is there, Slough increasingly isn't on the prioritisation list for new connections. That reordering matters commercially in a way the industry hasn't fully internalised yet.

And flexibility is becoming a de facto condition without ever being written into statute. The Balancing Mechanism is open. The demand connections reform weights applications partly on system impact, which is a polite way of saying a facility that can flex its draw looks better to NESO than one that can't. Operators who show up at Gate 2 with a credible flexibility proposition get cleaner decisions than operators who show up with flat 24/7 load curves. The UK didn't mandate dispatchable on-site capacity the way the Irish CRU did. It has something softer that produces similar effects without the legal instrument.

A few practical things worth knowing about the pipeline as of spring 2026.

Offer dates inside the next four years are mostly already allocated to projects that cleared the January cull. New applications going through the second Gate 2 window will land in the second half of the decade at best, with realistic connection energisation for strategically designated sites in the 2029-2032 range. Mid-decade is possible in the right geography. The 15-year wait figure that got quoted at the worst of the 2024 backlog was an artefact of a broken process, not planning reality.

Multiple transmission upgrades are now funded. Eastern Green Link 2 alone is £4.3bn, the single largest investment in GB electricity transmission ever. These projects create real capacity in specific corridors, not everywhere. Reading the project list carefully and matching intended sites against corridor timing is now a core commercial exercise, not a technical footnote.

Britain is further through this reset than any other major European market. Which cuts both ways — the constraints are sharper, but the rules are clearer, and operators who can work the new regime get answers faster than in markets that haven't started their own version of this conversation yet.

The same playbook is going to get exported. Dutch, German, French DNOs are already reading the TMO4+ documents. Worth reading them yourself while it still feels abstract.