New data centres in London may wait 7 to 13 years for a grid connection, against 3 to 7 years in other European markets, according to the study the Greater London Authority commissioned on the sector. At the midpoints that is 10 years against 5, about twice as long, by our arithmetic.
The London Assembly cited that study on 11 November, in a notice setting a 12 November Environment Committee hearing on the environmental cost of the city’s data centres. London has 99 data centres built and about 26 more planned or in the planning system, the Assembly said.
Business Pill · Grid Cost Allocation Explained
A short explainer on who pays when the grid has to be extended for a new user. It teaches the general idea only and does not discuss London.
The key insight: As we read the study, the binding limit on London’s data centres is time to power: new grid connections may take 7 to 13 years, about twice the 3 to 7 years in other European markets by our arithmetic, while the study projects capacity of 8 GW to 11 GW by 2050 from about 1 GW today.
The London Assembly notice says London “accounts for more than 80 per cent of the UK’s data centre capacity”, citing CBRE. The committee will look at energy demand, water use, air quality, noise, biodiversity and the impact on local communities, it says.
The witnesses include campaigners from Global Action Plan, Park Royal and the “No to East Havering Data Centre” campaign, and speakers from the Environment Agency, Water UK, the Carbon Trust and the University of Exeter. The meeting is at City Hall at 2pm on Thursday 12 November.
Grid connection timelines for new data centres in London, other European markets and the Hillingdon and Ealing hubs, as given in the Buro Happold study for the Greater London Authority (December 2025). Midpoints are our arithmetic.
What the GLA’s Own Study Says
The Buro Happold study for the GLA, dated December 2025, puts London’s operational capacity at about 1 GW in 2025. On grid connection data and future energy scenarios, it says capacity “could grow to between 8 GW and 11 GW by 2050”, which is 8 to 11 times today’s, by our arithmetic.
It calls energy “the single most critical constraint on London’s data centre growth” and says transmission limits have emerged particularly in West London. Elsewhere it reports delays of 5 to 10 years for new grid connections in hubs such as Hillingdon and Ealing, citing Energy UK and CBRE.
On water, the study says a 100MW hyperscale data centre in the UK “can consume up to 2.5 billion litres of water annually”, about the needs of 80,000 people. That is up to 25 million litres per megawatt a year, by our arithmetic.
The Pipeline and the Grid
The 26 planned facilities would add about 26% to the 99 already built, by our arithmetic. The study says the sites are predominantly north of the Thames and that land scarcity is pushing new projects to outer boroughs, “often requiring long-distance grid connections”.
The study also notes that water-related considerations such as cooling, abstraction and discharge “are not comprehensively addressed” in current planning, and that energy planning often relies on aggregated demand data.
The Structural Read
As we read the Assembly notice, the hearing puts water, noise, air quality and local communities next to energy, and two of the three community witnesses come from campaigns at named sites, Park Royal and East Havering.
As we read the study, its own authors flag that water-related considerations are not comprehensively addressed in planning, so the 2.5 billion litres a year it gives for a 100MW site is a ceiling, not a London total.
Buro Happold for the Greater London Authority, December 2025
“Energy is the single most critical constraint on London’s data centre growth”
Three Implications
FOR DEVELOPERS At the study’s 7 to 13 years, a connection requested now would arrive between 2033 and 2039, by our arithmetic.
FOR PLANNERS The 26 planned sites would add about 26% to the 99 built, by our arithmetic.
FOR WATER Up to 25 million litres per megawatt a year for a 100MW hyperscale site, by our arithmetic on the study.
The framework records: “But infrastructure reality is catching up with AI ambitions”.
As we read it, the GLA’s study gives that sentence a local number: connection waits of 7 to 13 years in the city that holds more than 80 per cent of UK data centre capacity, as the Assembly puts it. The framework predates the study and does not discuss London.
What Is Not Established
The Assembly notice previews a hearing; it does not publish findings, and what witnesses will say is not yet on record. The 8 GW to 11 GW range is a projection the study says carries significant uncertainty. The 2.5 billion litres is a ceiling for one 100MW site type, not a measure of London’s current water use, and the study does not give a figure for the share of London’s electricity that data centres use.
This Week In The AI Supercycle
The Great Convergence of Capital, Power, and Geopolitics
This week’s issue: for two years the AI build-out was priced as a supply story, and last week as a credit story. This week it was priced as a permission story, and the answer arrived as a cost, not a ban.
London has 99 data centres, about 1 GW of capacity and 26 more sites in the pipeline. The GLA’s own study says new grid connections may take 7 to 13 years, about twice the 3 to 7 years elsewhere in Europe by our arithmetic, and the Assembly’s Environment Committee has scheduled a hearing on the environmental cost for 12 November.
A note on sourcing. We read the London Assembly’s 11 November 2026 notice in full and the Buro Happold study for the Greater London Authority (December 2025), the evidence the notice cites, from the GLA’s site. The 80 per cent UK share is CBRE’s, as cited by the Assembly. Midpoints, multiples and per-megawatt water figures are our arithmetic. The hearing takes place on 12 November; we report its agenda, not its findings. The Assembly and the GLA were not contacted.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
