Updated September 2, 2026
The London Data Centre Grid Connection Queue now averages 7 to 13 years, compared with 3 to 7 years in competing European cities. That gap explains much of what looks strange about the London market right now: a planning pipeline running roughly 20% above its historical norm, 26 monitored projects sitting at the approved stage, and none independently verified as operational.
London is not short of consented sites. It is short of years.
Why the London Data Centre Grid Connection Queue Has Become the Whole Story
For most of the last decade, the constraint on London data centre development was land and consent. Find a well-connected industrial site, satisfy the borough, and build.
That model has broken. Research produced for the Greater London Authority by engineering consultancy Buro Happold, published on 27 July 2026, assessed constraints across energy, water, land use, and digital infrastructure, and identified energy availability as the primary physical limit on further capacity in the capital.
The scale is easy to state. London currently operates around 99 data centre sites drawing approximately 760 MW at peak, roughly the electricity consumption of three-quarters of a million homes. Sitting behind that is a connection queue amounting to about ten times existing operational capacity, with more than 8 GW in total.
Nobody expects that queue to be built. That is the point. A queue filled with positions that will never energize is not a pipeline; it is a traffic jam. Genuinely deliverable schemes are stuck inside it alongside everything else.
Seven to Thirteen Years: What the London Data Centre Grid Connection Queue Does to an Investment Case
The most consequential number in the GLA evidence base is not the gigawatts. It is the wait.
The London Data Centre Grid Connection Queue creates an average delay of 7 to 13 years, compared with 3 to 7 years in comparable European markets. The differential is not marginal. It is long enough to change where capital goes.
Consider what that does to a scheme underwritten today. A developer securing consent in 2026 with a mid-range queue position is modeling energization somewhere in the mid-2030s. Over that horizon, the hardware generation shifts several times, the cost of capital moves, the anchor tenant’s requirements change shape, and the planning conditions attached to the original consent may no longer reflect current policy on water or carbon.
That is why the approval count has become a weak signal. A consent granted in 2026 is not a 2026 asset. It is an option on a 2030s asset, priced as though it were nearer.
The forward demand picture worsens the squeeze rather than easing it. Buro Happold’s analysis projects London data centre demand expanding by 8 to 11 GW by 2050, against roughly 1 GW of active capacity today. Greater London already hosts around 80% of total UK data centre capacity, so there is no easy internal release valve.
What the Pipeline Looks Like When You Separate the Stages
Aggregate approval counts obscure all of this. The more useful view separates each project into the stage it has actually reached and verified.
Authority Inc does exactly that through its London data centre pipeline monitor, which tracks 50 qualifying projects assembled from 566 matching public planning records rather than counting raw applications. The August 2026 stage breakdown looks like this:
- One project at pre-consent stage
- Six still working through planning
- Twenty-six with planning approval secured
- One with verified construction evidence
- Ten recorded as build complete in the planning record
- Six refused or withdrawn
- Zero independently verified as operational
Understanding the London Data Centre Grid Connection Queue alongside planning approvals provides a more realistic picture of which projects can actually reach construction and operation.
Set against the connection queue figures, the problem’s shape is clear. The planning system has already delivered a substantial consented pipeline. The grid has not delivered the counterpart, and the monitored population contains no verified operational site to prove the route through.
Momentum on the planning side is also cooling. The tracked index reads 119.5 for the trailing year to June 2026 against a fixed 2021 to 2024 baseline of 100, down from a peak near 259 in early 2026 as the large 2025 filing wave rolled out of the twelve-month window. The index remains above normal, but it is decelerating.
Ofgem’s Reform Is a Queue Cull, Not a Capacity Release
On 29 July 2026, Ofgem opened a consultation aimed directly at speculative data centre projects, closing on 16 September 2026. Two measures matter.
The first is a data centre commitment fee, which projects must secure from the point they accept a connection offer until they are ready to connect. Projects can forfeit the fee if they fail to progress under the rules. Reporting puts the fee at about 2.5% to 7.5% of average project cost, refundable on energization.
The second is a set of data centre-specific queue management milestones, covering evidence of financial capability and planning progress, with removal from the queue for projects that fall behind.
For developers caught in the London Data Centre Grid Connection Queue, these reforms could significantly change queue positions and project timelines.
The national context explains the urgency. Demand for connection applications across Great Britain rose from 41 GW to 125 GW in under a year, driven largely by data centres.
Here is the part that gets misread. These reforms do not add a megawatt of capacity. They reorder who stands where. For a funded developer with consent, an operator, and a credible program, that reordering is straightforwardly good. For a landowner holding a consent as an option, it converts a free position into a paid one with a deadline attached.
Expect the visible pipeline to shrink and its quality to rise. A smaller queue that actually connects is worth more to the market than a large one that does not.
Where New Capacity Is Physically Being Unlocked
Reform sorts the queue. Network investment is what lengthens it.
The most significant London example arrived on 13 August 2026, when National Grid launched its North West London Upgrade: a new substation, around 60 km of new cable, and more than 200 km of upgraded overhead line. The company has said the project will enable new connections, including five data centres.
This is the right category of evidence to weigh heavily because it is committed capital tied to physical assets, not stated intent. It improves the medium-term outlook for schemes in the west and northwest of the capital, where recent approvals cluster, including Southall in Ealing, the Heathrow Estate in Hounslow, and Premier Park in Park Royal.
It does not fix any individual project’s date. Energization still depends on queue position under whatever rules survive the Ofgem process.
The Geography Is Already Responding
When power sets the timetable, location follows power rather than demand.
As the London Data Centre Grid Connection Queue lengthens, developers are increasingly looking beyond established clusters for locations with better access to available power.
JLL research found that European hyperscale campuses due to open within the next two years sit an average of 175 km from major urban hubs, compared with 46 km for facilities delivered between 2022 and 2025.
London is showing the same drift on a smaller scale. A major scheme has emerged in Charlton at around 180 MW across roughly 100,000 m² of floorspace, and a 600 MW campus proposal sits at pre-consent stage in Havering, both well outside the traditional western cluster.
There is a policy dimension too. Since November 2025, data centres have been formally treated as nationally significant infrastructure, narrowing the role of individual local authorities in the largest cases and pushing decisions toward national frameworks.
Combined with the next London Plan, which draws on the GLA evidence base, the strategic layer is becoming more decisive than the borough layer.
How to Read the London Data Centre Market Without Being Misled
A more defensible method, whether you are underwriting, advising, or reporting, is to separate planning progress from grid availability.
Treat the London Data Centre grid connection queue and planning approvals as independent variables. A scheme with planning permission but no viable connection timeline is not necessarily closer to operation than a project with a secure grid position. It is different along.
Deduplicate before counting. One development routinely generates a dozen planning records across outline applications, reserved matters, conditions, and amendments. Raw counts can substantially overstate the pipeline.
Ask for the energization date, not the approval date. The approval date tells you when a committee voted. The energization date tells you when revenue starts.
Distinguish recorded stages from verified ones. A borough recording “build complete” confirms a building exists. It does not confirm the facility is powered or in service.
Watch the Ofgem outcome as a filtering event. The projects that survive the milestone tests are the real pipeline. Everything else was always notional.
Frequently Asked Questions
Q1. How long is the London Data Centre Grid Connection Queue?
Answer: Research produced for the Greater London Authority puts average wait times at 7 to 13 years for London projects, compared with 3 to 7 years in competing European cities. Energy availability is the primary physical constraint on further capacity in the capital.
Q2. How much capacity is sitting in the London Data Centre Grid Connection Queue?
Answer: More than 8 GW, roughly ten times the capacity of London’s existing operational data centres. The capital currently runs about 99 sites drawing approximately 760 MW at peak demand.
Q3. What is Ofgem doing about speculative data centre connection applications?
Answer: Ofgem opened a consultation on 29 July 2026, closing 16 September 2026, proposing a data center commitment fee and data centre-specific queue-management milestones. The aim is to prioritize viable projects and remove those that cannot progress.
Q4. Does a London planning approval guarantee a data centre gets built?
Answer: No. In the August 2026 pipeline snapshot, 26 monitored projects were approved, one had verified construction evidence, and none had independently verified operational status. Power, finance, construction, and commissioning are separate hurdles.
Q5. How much is London data centre demand expected to grow?
Answer: Analysis for the GLA projects London data centre demand to expand by 8 to 11 GW by 2050, against roughly 1 GW of active capacity today. Greater London hosts around 80% of the UK’s total data centre capacity.
Final Thoughts
London’s data centre market continues to attract strong demand, investment, and development activity, but the London Data Centre Grid Connection Queue remains its biggest constraint. With connection timelines far longer than in competing European markets, planning approval alone does not guarantee a project will become operational. Until grid capacity expands and viable projects move through the queue, London’s large approved pipeline will remain potential rather than immediate capacity.
For investors, developers, occupiers, and advisers, the focus should therefore shift from the number of approved projects to their ability to secure power and reach energization. Ofgem’s proposed reforms and major network upgrades could improve the situation over time, but grid access will ultimately determine which projects move from planning approval to operational data centres.
Recommended Articles
We hope this guide helps you understand how the London Data Centre Grid Connection Queue is shaping data centre development, investment, and infrastructure planning in the capital. Explore these recommended articles for more insights into data centre infrastructure, energy demand, grid connectivity, planning approvals, and the future of the UK data centre market.
