Summary
- The UK data centre pipeline is being tested by planning approvals, speculative grid applications, and rising scrutiny of large demand projects.
- Slough’s Manor Farm decision shows how ministers are weighing quantified capacity need against Green Belt, power, and local planning constraints.
- Capacity claims increasingly need evidence of land control, grid status, substation delivery, customer demand, financing, and realistic commissioning.
Britain’s data centre pipeline has become large enough to require a better test than counting announced megawatts. A project may have a site boundary, a capacity figure, and a place in a grid queue, but none of those makes it usable infrastructure. Deliverable capacity depends on land control, planning permission, energisation, substations, cooling, financing, customers, construction capacity, and commissioning.
The government’s recent decision to approve a hyperscale data centre and battery energy storage system at Manor Farm, on land north of Wraysbury Reservoir near Slough, gives that test a practical form. The recovered appeal covered a scheme including a data centre, BESS, ancillary substation, emergency backup generators, fuel storage, drainage, landscaping, and access works. In his decision, the Secretary of State accepted that unmet data centre capacity need in the Slough Availability Zone stood at 1,259MW, after considering expected supply up to 2030 and further potential at Slough Trading Estate.
The decision did not give data centres a blank cheque over land use, nor did it settle the balance between digital infrastructure and Green Belt policy. It showed instead how specific the evidence has become. The appeal turned on a named availability zone, a defined capacity deficit, an assessment of alternative sites, and a conclusion that the scheme could help meet demand in a market where proximity, latency, fibre, and established infrastructure still shape location.
That evidence burden is growing because the UK grid queue now contains demand that cannot all be treated as equally credible. The government’s consultation on accelerating electricity network connections for strategic demand says demand is set to more than double by 2050 and warns that connection delays are holding back projects linked to national priorities. It says the transmission demand queue stood at 96GW at the end of June 2025, with a further 29GW waiting at distribution level, and identifies data centre connection applications as a major driver.
| Indicator | Figure | What it shows |
|---|---|---|
| Transmission demand queue, end-June 2025 | 96GW | Large demand projects are competing for capacity in a queue far beyond near-term network headroom. |
| Distribution demand waiting queue | 29GW | Connection pressure is not confined to transmission-level schemes. |
| Data centre projects responding to NESO demand call | Around 140 projects / about 50GW | Data centre demand is now large enough to shape queue reform. |
| Available supply in the Slough Availability Zone up to 2030 | Up to 1,152MW | Existing and expected capacity did not close the accepted need gap in the Manor Farm decision. |
| Accepted unmet need in the Slough Availability Zone | 1,259MW | The planning case relied on quantified local capacity need. |
| London existing data centre estate | 99 sites / around 760MW peak demand | The capital already has a substantial operating base before new queued demand is considered. |
| London data centre grid requests | Around 10 times current London data centre capacity | Requested demand is far larger than today’s operating footprint. |
Ofgem’s demand connections reform is aimed at the same problem from the network side. Its “Curate, Plan and Connect” work identifies a demand queue containing likely non-viable projects, well-progressed projects unable to connect quickly enough, and no mechanism to prioritise strategically important demand. Financial requirements, readiness tests, post-acceptance progression milestones, and incentives for less mature projects to exit the queue are all being considered.
Around London and the Thames Valley, those reforms will meet one of Europe’s most concentrated data centre markets. City Hall says London has 99 data centre sites delivering around 760MW at peak demand, broadly equivalent to the electricity use of 750,000 homes. It also says around 10 times the capacity currently used by existing London data centres has already been requested in the grid connection queue. A dedicated data centre policy is expected in the draft London Plan.
The first proof point for any project is still land. A project with ownership, option rights, a defined planning boundary, environmental assessments, access studies, and known constraints is in a different position from one that has only a headline IT load and a speculative grid application. Land also decides whether a scheme can make a credible case on traffic, noise, visual impact, backup generation, heat rejection, biodiversity, drainage, and community benefit.
Planning status adds the next layer. Some projects will move through local planning authorities. Others may be called in by ministers, challenged through judicial review, or shaped by national policy routes linked to AI Growth Zones and infrastructure need. Green Belt, industrial land protection, housing targets, transport access, flood risk, and power infrastructure can each decide whether an announced campus has a realistic programme.
Power evidence remains the central test. A grid connection agreement is not energised capacity, and a connection date is not a completed substation. Developers need to show where power will come from, what reinforcement is required, which network assets are shared with other users, and how delays will affect phasing. Where a scheme includes batteries, backup generation, or flexible demand commitments, those systems need to be treated as part of the engineered project rather than optional extras.
Manor Farm shows how much physical infrastructure sits beneath a single capacity figure. The approved scheme includes battery storage, an ancillary substation, emergency backup generation, fuel storage, drainage, access works, and associated landscaping. The decision also gave weight to the BESS element, noting its role in storing and distributing excess energy from renewable sources and adding resilience to a constrained power network.
Commercial evidence is harder to judge from outside, but it is no less important. A hyperscale facility backed by an anchor customer sits in a different risk category from a speculative campus relying on future leasing. Pre-lets, reservation agreements, financing, phasing, and customer credit all affect whether a project will move from consent to construction. In a market where power reservations can carry strategic value, customer demand needs to be distinguished from optionality.
Construction capacity then becomes the filter that planning consent cannot solve. High-voltage equipment, transformers, switchgear, generators, chillers, pumps, controls, commissioning engineers, specialist contractors, and grid interface works all sit on the delivery path. A consent may create land value immediately, but it does not produce long-lead electrical equipment or the labour needed to install and commission it.
Queue reform will make these distinctions harder to avoid. Projects with land, planning progress, credible customers, financing, and network evidence will be able to argue for priority. Projects that mainly hold a queue position will be more exposed. Between those two groups sits a large middle ground of schemes that may be viable but not yet mature enough to justify scarce grid capacity.
The next round of UK data centre growth will be judged through files as much as forecasts. Planning authorities will want quantified need, site evidence, and local impact assessment. Network operators will want readiness and progression. Investors will want customer demand and power certainty. Customers will want energisation dates and resilience commitments. Local communities will want clarity on power, water, backup generation, traffic, heat, and landscape impact.
Announced capacity will still shape market sentiment, but usable capacity will come from projects where the physical and commercial evidence holds together: land that can be developed, power that can be delivered, planning arguments that can be defended, customers that can occupy, and a construction programme tied to real equipment, real labour, and real energisation dates.
