In 1979, when The Clash recorded London Calling, the city was hardly at the height of its powers. Britain was emerging from a decade of economic turbulence, the three-day week was still a recent memory, and London carried the visible scars of industrial decline. Joe Strummer’s voice, urgent and half-apocalyptic, conjured a city threatened by forces considerably larger than itself: nuclear meltdown, famine, environmental catastrophe and economic collapse. The song’s final moments dissolve into the scratch and crackle of a guitar pickup, spelling out an SOS in Morse code.
It was not, however, quite the straightforward cry of despair that it first appeared to be. London Calling was a warning, certainly, but also a demand that people pay attention. The danger lay not only in the crises themselves, but in the possibility that we might become so accustomed to them that we stopped noticing.
Nearly five decades later, London is calling again.
This time the source of the anxiety is rather less dramatic than a nuclear meltdown, but it is no less fundamental to the way the city will function. London needs electricity, water, land, housing and, increasingly, somewhere to put the extraordinary volume of computing infrastructure on which its economy now depends. The problem is that all of those things are in demand at the same time.
Data centres have consequently found themselves in an unusual position. They are simultaneously among the least visible and most consequential pieces of infrastructure in the modern city. Most people will never see the servers that process their bank transactions, store their photographs, run their businesses or deliver the applications they use every day. Yet when a new data centre is proposed on a piece of land in their neighbourhood, suddenly that invisible digital world acquires a very physical presence: a large building, a substantial electricity connection, cooling equipment, traffic, noise and questions about water.
And increasingly, those questions are being asked rather loudly.
The timing is interesting. As London’s new Draft London Plan comes under scrutiny, the capital is being forced to confront the role that data centres should play in its future. For the first time, the plan contains a dedicated policy for the sector, asking boroughs to consider electricity and water capacity, opportunities for heat recovery, and the potential effects on housing, employment land and local communities.
It is difficult to imagine a more revealing list.
Because London does need data centres. Indeed, it needs rather a lot of them.
The capital became Europe’s leading data-centre market for reasons that have little to do with fashion and everything to do with geography and economics. London is home to one of the world’s great concentrations of financial and professional services, a vast population of businesses and consumers, extensive fibre connectivity and an ecosystem of cloud providers, carriers, technology companies and other digital infrastructure. For an organisation that needs its systems close to customers, markets and networks, London remains an exceptionally useful place to be.
This is why the argument about data centres can so easily become distorted. It is tempting to regard them as enormous sheds full of computers competing with hospitals, homes and businesses for scarce electricity. But that describes the building rather better than it describes its purpose.
The servers inside those buildings are supporting much of the economy outside them.
The question, then, is not whether London should have data centres. It plainly should. The more interesting question is whether building ever more of them is necessarily the best way of satisfying the demand for computing capacity.
That distinction matters because the numbers now attached to London’s data-centre expansion are becoming difficult to ignore.
Current estimates put the peak electricity demand of London’s existing data centres at around 760MW. At the same time, approximately 8GW of additional data-centre capacity is understood to be sitting in the electricity connections queue. In other words, the amount of power being sought by prospective data-centre developments is more than ten times the capacity currently being consumed by the sector.
That is an extraordinary number, particularly in a city which is simultaneously trying to accommodate hundreds of thousands of new homes, decarbonise its energy system and improve resilience to overheating and climate change.
It also raises a deceptively simple question.
What if London’s next data-centre challenge isn’t simply how to build more capacity, but how to make better use of the capacity it already has?
That question has acquired a human dimension in recent months. In west London, planning committees have been considering substantial new facilities while councillors and residents have raised concerns about their demands on electricity and water. In Southall, another major data centre has been approved on the former Honey Monster, or Quaker Oats, food-factory site, close to several other proposed facilities.
Yet the site had previously been earmarked for a very different vision of the future: Galliard Homes’ Quayside Quarter scheme, comprising 1,997 flats and 28,000 square metres of creative employment space. Planning permission was granted in 2020, but the scheme never materialised and Galliard sold the site in 2022.
And in Tower Hamlets, the controversy surrounding a proposed development near Brick Lane has become a much broader argument about land, housing, local democracy and what kind of city London wants to become.
The Brick Lane debate is particularly instructive because it illustrates the problem that statistics cannot quite capture. To a developer, a brownfield site with excellent connectivity may look like an obvious location for digital infrastructure. To a community facing a social-housing waiting list of more than 30,000 households, the same piece of land can look rather different.
Neither perspective is inherently irrational.
The difficulty is that London has become a city in which every major infrastructure decision increasingly has competing claims attached to it. Land that could accommodate a data centre might accommodate housing. Electricity that could support a new digital facility could support other forms of development. Water consumed by cooling systems has to come from the same underlying resource that serves households and businesses.
This is not an argument against data centres. It is an argument against treating their requirements in isolation.
There is, after all, something slightly strange about discussing a future in which London may need thousands of megawatts of additional computing capacity without asking how effectively the infrastructure already operating in the city is being used.
There are existing facilities with power connections, fibre networks, cooling systems and established operational expertise. There are buildings that already form part of London’s digital infrastructure rather than requiring an entirely new physical and electrical footprint. There is colocation capacity, managed infrastructure and the possibility of consolidating or modernising existing environments rather than automatically creating another building every time demand increases.
None of this eliminates the need for new London data centres. Computing demand is growing, and some of that demand will require genuinely new capacity. Artificial intelligence alone is changing the scale and nature of the computing infrastructure required. London cannot simply freeze its digital infrastructure in place and expect the economy to continue expanding around it.
But perhaps the choice is not as binary as the current debate sometimes suggests.
We do not have to choose between unlimited data-centre construction and a moratorium on the industry. There is a more interesting middle ground: use existing infrastructure as efficiently as possible, build new capacity where there is a demonstrable need, and make the environmental and social consequences part of the calculation rather than an afterthought.
That might mean asking rather different questions when a new project arrives on a planner’s desk. How much of London’s existing capacity is actually available? Is new electricity infrastructure genuinely necessary, or can existing connections be used more effectively? What will happen to the heat? How much water will the facility require, and can that requirement be reduced? What does the development contribute to the surrounding economy? And, perhaps most importantly, is this particular site the right place for this particular kind of infrastructure?
These are not questions designed to stop the data-centre industry. They are questions designed to make it better.
There is another reason why location matters. London’s success as a data-centre market is not simply the result of developers having discovered a convenient place to build large buildings. It is the product of decades of investment in connectivity and digital infrastructure. The capital’s dense concentration of fibre, carriers, cloud providers, financial institutions and technology businesses creates an ecosystem that is difficult to reproduce elsewhere.
For many organisations, proximity still matters. Being close to customers, networks, markets and the infrastructure that keeps their operations running can make a genuine difference to performance, resilience and cost.
That is why the answer cannot simply be to push every future data-centre development somewhere beyond the capital’s boundaries.
Nor can it be to pretend that the pressures created by the sector don’t exist.
The smarter answer may be to make better use of what London already possesses, while being considerably more selective about what it builds next.
London’s data-centre story should therefore be about more than planning permissions and megawatts. It should be about utilisation, efficiency, resilience and the contribution that digital infrastructure makes to the city around it. Existing facilities should be part of the conversation, not simply the backdrop against which the next generation of buildings is planned.
In 1979, The Clash ended London Calling with an SOS.
Perhaps London’s digital infrastructure debate deserves a rather different message.
Not ‘stop’. Not ‘build’.
Just: think.
Because the future of London’s data-centre industry should not be measured simply by the number of new buildings that appear on a planning map. It should be measured by how much digital capability the city can deliver while making intelligent use of its land, its electricity, its water and its existing infrastructure.
London still needs to call. We just need to make sure we are listening.
——————————————————————————————————————————————————
Image: Photographed by Pennie Smith, the iconic image of bassist Paul Simonon smashing his guitar was stylised by graphic designer Ray Lowry for The Clash’s 1979 London Calling album cover and repeated on this original release poster.
If you found this article of interest, please don’t forget to sign up for our NEWSLETTER for the latest industry news and insights delivered direct to your mailbox.
Comments are closed.