Data centres are classified as Critical National Infrastructure. But critical for whom?

In September 2024, data centres joined a select list. Alongside water, energy and the emergency services, they were classified as Critical National Infrastructure: the systems the country cannot function without. Since then the commitments have only grown. National planning status for the biggest campuses. AI Growth Zones from Oxfordshire to Lanarkshire. Tens of billions of pounds of private investment heading into British ground.

Critical, then, is settled. The question that isn't settled is the one hiding inside the word.

Critical for whom?

A data centre can be critical to the nation and invisible to its neighbours. It can hold up the digital economy while the town beside it gets a fence, a gatehouse and some business rates. That is how most of this build-out is currently shaped, and nobody planned it that way. It is simply what happens when every conversation is about what the infrastructure needs, electricity, land, water, grid connections, and almost none is about what it gives.

But these buildings have far more to give than almost anyone at the planning table realises. The most important question isn't what data centres compute. It's what they make possible.

Critical for a town: what that would look like

Picture a January morning, somewhere in England, ten years from now.

The school is warm before the children arrive. The leisure centre pool has been at temperature since six. On the estate behind the high street, radiators tick on in a thousand homes, and nobody thinks about the gas price, because none of this warmth comes from gas. It comes from the campus on the edge of town: a data centre, humming through the night, its servers thinking for half the world and warming this corner of it.

The bills on that estate are lower than they were a decade ago. The GP sees fewer chest infections each winter. Money that used to leave town through the meter gets spent in town instead. And when a rumour went round that the operator might scale back, the town's reaction said everything: people wanted it defended. Their data centre. Their heat.

Nothing in that picture is invented. Every piece of it is already running somewhere. Stockholm feeds surplus heat from data centres into the network that warms its apartment blocks, and has for a decade. Meta's campus in Odense warms around eleven thousand Danish homes. A data centre heated the training pool for the Paris Olympics. In Britain, modular data centres already warm public swimming pools, backed by hundreds of millions in investment to do more.

The only missing piece is the decision to put it all together here.

What a data centre actually makes

To see the opportunity, you need one plain fact about how these buildings work.

Every data centre makes two things. The first is computation: the AI models, the cloud services, the answers. That product is sold, and it is why the money is coming.

The second is heat. Nearly every watt a server draws comes back out as warmth. A large campus is, in effect, a town-sized heater that happens to think. Today that warmth is mostly lifted into the sky by cooling plant. Not because it has no value, but because nothing was ever built to receive it.

Now scale that up. Research published in January estimated that heat from the UK's expanding data centre fleet could warm between 3.5 and 6.3 million homes by 2035, if we build the connections. The low end of that range is more homes than there are in Scotland.

No new fuel. No drilling. No exposure to a global gas market. The warmth arrives with the servers, every hour of every day, for as long as the world keeps asking questions. And the engineering is well understood: heat pumps lift the temperature, pipes carry it, storage smooths it across the seasons. This is plumbing, at civic scale. The countries already doing it are not cleverer than us. They just decided the second product mattered.

Which sharpens our question. If every one of these buildings arrives carrying enough warmth for a neighbourhood, then critical for whom is not philosophy. It is a design decision, taken project by project, and it depends entirely on who gets connected.

Seeing who it could be critical for

Here is why that decision keeps going unmade: because of how we look at places.

When a major project is planned in Britain, each expert examines one slice. Energy in one report, transport in another, health, jobs and carbon in others still. Every report is competent. But a town is not seven slices, it is one living system, and the biggest opportunities sit on the joins between the slices, exactly where nobody is looking.

At Generation 7 we look through a different instrument. We call it HAUK: a place-based intelligence resource that views a place the way a hawk does, from above, where the joins disappear. It reads every place through three lenses, and each lens asks questions anyone can answer about their own street.

People live well. What is it like to live here, day to day? Are homes warm in winter and cool in summer? Can people pay their bills without fear, or does the envelope on the doormat set the mood for the week? Are they healthy? Secure in their homes, or one rent rise from leaving? Do they carry the constant mental load of a house that is working against them? And do they have any real say over any of it? These are the first things a big piece of infrastructure next door should change, and the first things today's planning process never asks about.

Communities thrive. How does this place work as a system of people? Does money stay local, or leak out through energy bills to distant suppliers? Are the NHS and social care carrying pressure that warmer, drier buildings would simply remove? Are children in school, in classrooms fit to learn in? Do people feel they belong somewhere stable, or is the place churning? And do the institutions that serve it, the council, the housing provider, the health system, actually work together, or only meet at ribbon cuttings? A data centre cannot fix all of that. But connected well, it touches every one of those questions from the same set of pipes.

Planet heals. How does this place sit inside larger living systems? Are emissions falling? Is the air clean, indoors and out? Is energy being used twice, or wasted once? Can the place cope with the heatwaves and floods already on their way? Is there room for nature? Do resources circulate, or pass through once and vanish? A data centre that vents its warmth is a straight line: energy in, heat to the sky. Connected, it becomes a circle, and circles are what healing systems look like.

Beneath the three lenses sit eighteen dimensions, six for each. That sounds like a framework, and frameworks sound like paperwork, so hold on to what the dimensions really are: real life, named. The back bedroom above 18 degrees in January. The cough that doesn't come back. The gas bill money spent at the butcher instead.

The reason to look this way is that value compounds along the joins. Connect a campus's warmth to an estate and you haven't done one good thing, you've started a chain: warmth improves health, health lifts school attendance and eases the NHS, steadier bills free household budgets, freed budgets feed the high street, and the displaced gas cleans the air and cuts the carbon. One decision, eighteen dials moving at once. That multiplication is invisible to slice-by-slice appraisal, and it is the real return on infrastructure done well.

The method is three questions, asked in order. SEARCH: what is true about this place, street by street? SEE: how does it connect? FORESEE: what could it become under different choices?

The day the map reorganised itself

We have watched those three questions change a real place. Recently, and close to home.

We were working with an estate in the north of England, several thousand people, with something increasingly precious under its streets: a district heat network, decades old, already connecting nearly two thousand homes. The strategic question was how to take its heat supply into the low-carbon future, and the direction of travel was a familiar one, a water-source solution, well studied and entirely sensible.

Then we asked the data centre question, almost as due diligence. What if compute came to the estate?

Everything changed. Not incrementally. Structurally.

A modular data centre, the kind British companies now deliver as units not much bigger than a plant room, could sit at the energy centre and plug more or less straight into the existing network. No new pipes under every street; the pipes were already there, bought and paid for by a previous generation. The servers would earn their keep computing, and their warmth, instead of being thrown away, would flow into homes that were already connected and waiting.

And then the second-order possibilities started arriving faster than we could write them down. Heat costs for nearly two thousand households, decoupled from the gas market. A revenue stream for the network's operator that could fund its renewal, instead of renewal competing with rent for every pound. Headroom to extend the network to the school, the pool, the streets that had always been just outside the boundary. Resilience, because compute demand doesn't take winters off. Even the beginnings of a different story the estate could tell about itself: not a place that heavy industry left behind, but a place the digital economy moved into, warmth first.

Here is the part worth underlining. The engineering had been available all along. What changed the map was the way of looking. One question, asked early, from above, with the whole place in view, reorganised the entire strategy around a better centre of gravity. A good heat source doesn't join a plan. It becomes one.

That is what SEARCH, SEE, FORESEE is for. And the earlier it happens, the more it is worth. Ask before the masterplan is drawn and the answers cost nearly nothing to act on: the energy centre lands on the edge nearest the homes, pipe corridors are safeguarded while they are only lines on paper, heat becomes a siting criterion alongside grid capacity. The cheapest moment to create a century of value is before anyone draws a red line.

Now scale the same thinking up. On the site of an old brick pit at Stewartby, near Bedford, a campus of national significance is proposed: 143 acres, up to a gigawatt of power. Within a few kilometres sit thousands of homes, schools and public buildings that will need clean heat within twenty years under any credible plan. A campus at that scale would produce more warmth than the whole area could use. Two facts, a few kilometres apart, that nobody has yet been asked to put in the same document. Put them together early enough and Bedford doesn't just host national infrastructure. It inherits a heat source that could serve it for generations.

There are grown-up questions inside that sentence, and each has a known answer. Who owns the heat? Settle it in the consenting deal, so the warmth has a beneficiary before the first rack arrives. What if the operator one day leaves? Build networks fed by many sources, as Stockholm does, so no single facility is load-bearing. What about campuses planning to run early phases on gas? Put the fuel pathway and the heat delivery in one accountable frame, so the transition is a commitment rather than an appendix. None of these is a reason to wait. They are the design brief.

Critical for the seventh generation

There is an old principle, and our company is named for it, that a decision should be weighed by its effect on the seventh generation to come. Not the next quarter, not the next election. The children of the children of the children not yet born.

Most infrastructure debate can't hold that horizon; its clock runs in consenting timelines and refresh cycles. But heat networks are hundred-year assets. The estate in our story is proof in both directions: pipes laid decades ago by one generation are now the reason a twenty-first-century opportunity can simply plug in. Decide well now and the decision outlives the technology, the operator and all of us.

It has happened before, on exactly the ground where some of Britain's newest infrastructure is planned. In 1926 the London Brick Company dug its pit at Stewartby and built a village around it: houses, a school, a hall, playing fields, a pool. Not charity; a theory of industry, that a company which takes something from a place should build something into it. The brickworks closed in 2008. The village is still there. A century on, people still live inside the decision.

Now a new industry arrives at the same pit, and the question in this article's title arrives with it. Critical for whom? There is a simple way to know when we've answered it well. Not whether a town tolerates its data centre, but whether, given the choice, it would fight to keep it. The way a town fights for its hospital. Warmth in ten thousand homes clears that bar easily. And clearing it is the best deal on the table for everyone: operators gain revenue and a welcome no PR budget can buy, towns gain energy security no global market can take away, and the country gains civic infrastructure it was going to build the shell of anyway.

The servers are coming. The heat is coming with them, every hour, for decades. Critical national infrastructure will have earned the first two words of its name when it is also critical to the street next door, and still warming it in a hundred years.

A century ago, a company on that Bedfordshire clay built a village without being asked.

Oliver Riley is the founder and Chief Engineer at Generation 7, where infrastructure, energy and regeneration projects are shaped using HAUK, a place-based intelligence resource that sees every place through three lenses (people live well, communities thrive, planet heals) and eighteen connected dimensions, asking three questions: what is true here, how does it connect and what could this place become?