Every infrastructure revolution leaves a different kind of place behind

Every infrastructure revolution in Britain's history has also been a place-making revolution, whether it intended to be or not.

The railways did not simply move goods. They made towns.

Crewe was fields until the engine sheds arrived. Swindon already existed as a market town on its hill, so the Great Western Railway built another settlement beside it: New Swindon, with houses, a hospital and a workers' library.

The great industrialists went further. Saltaire, Bournville, Port Sunlight and hundreds of less celebrated company villages were built around the belief that industry and place should prosper together.

We can question the paternalism behind them. We cannot question the conviction that infrastructure owed something to the ground on which it stood.

Then, somewhere in the twentieth century, we began to unlearn that relationship.

The motorway age was enormously productive. It created the distribution geography on which Britain still depends: the warehouse parks of the Midlands' golden triangle, the retail parks and business estates beside motorway junctions, and the logistics systems that supply almost everything we buy.

But it created relatively few places to which anyone feels they belong. It produced efficient, necessary and often placeless edge-lands: landscapes that people use but seldom love, inhabit or defend.

Now another infrastructure revolution has broken ground.

In September 2024, the government designated data centres as Critical National Infrastructure, placing them alongside systems such as water and energy. AI Growth Zones are being established, major investments announced and individual data centre proposals directed into the national infrastructure consenting regime.

The facilities being planned today may still be operating in the 2060s.

They will remake the places around them either way. That is not a prediction. It is what infrastructure does.

The question is which tradition this wave will join: the one that made towns, or the one that made nowhere.

The current model asks communities to contribute more than they receive

The early signs suggest that the relationship is not working, and the places being asked to host these facilities have noticed.

A 2026 review by engineering consultancy Hoare Lea examined 33 disputed UK data centre applications. It found planning delays averaging 490 days, driven by concerns that included inadequate engagement, unclear community benefit, unsuitable locations, infrastructure constraints and energy use.

It may be tempting for the industry to view this as a communications problem — but I think it's deeper than that.

A hyperscale campus asks a place for land, grid capacity, water, infrastructure and consent. In return, it may offer some jobs, some business rates and a secure perimeter.

Communities that question that exchange are not anti-technology. They are looking at the proposed terms and asking what the development will create for the place that hosts it.

That question matters to developers too.

Social licence is no longer the soft material of consultation exercises and corporate reports. Hundreds of days of planning uncertainty make it a material constraint on the construction of Britain's digital infrastructure.

Community support cannot be purchased after the design is fixed. A genuine relationship between infrastructure and place has to be designed in.

Every data centre produces a second product that Britain largely throws away

What could a data centre offer the town around it?

Part of the answer is already inside the building, being thrown away.

A large data centre is, among other things, a town-sized heater that happens to think. Almost all the electricity used by its servers eventually becomes heat, and more energy is then required to remove it. In some cases, this happens within reach of homes that cannot afford to stay warm.

Every data centre therefore produces two things.

The first is computation: the models, cloud services, storage and answers for which the facility exists.

The second is heat.

Today, much of that heat is treated as a waste product. Fans, pumps and cooling systems run continuously to move it out of the building and release it into the surrounding environment. Energy is purchased to operate the servers, and further energy is used to manage the heat they produce.

This does not mean that all data centre heat can readily warm homes. Much of it is low-grade heat, at roughly the temperature of bathwater. Its usefulness depends on the cooling technology, the temperature at which it can be recovered, the distance to potential users, the presence of a heat network and the pattern of demand throughout the year.

But where the conditions are right, the engineering is established.

Heat pumps can raise recovered heat to a useful temperature. Insulated pipes can carry it beneath streets in much the same way that water and gas are distributed. Heat networks can then supply homes, schools, hospitals, leisure centres and businesses from a shared source.

Stockholm has incorporated data centre heat into its district heating system for years. Meta's campus in Odense supplies heat to up to eleven thousand Danish homes. Similar schemes are being developed elsewhere in Europe.

Industry modelling published in January 2026 by EnergiRaven and Viegand Maagøe estimated that recoverable heat from Britain's expanding data centre fleet could theoretically supply between 3.5 and 6.3 million homes by 2035. The range depends on assumptions about future data centre capacity, cooling systems, heat-capture rates, transmission losses and proximity to demand.

The lower estimate is still greater than the number of homes in Scotland.

Nor is this merely a Nordic possibility. The analysis identifies British clusters, including Greater Manchester, where existing and proposed data centres sit close to planned housing, existing communities and areas of fuel poverty.

Honesty matters. A remote campus several miles from significant heat demand may never connect economically. A facility designed without recoverable cooling architecture may offer less useful heat. Summer supply is less valuable than winter supply. Heat generated using fossil fuels cannot automatically be described as clean.

The opportunity is real, but it is conditional.

The important point is that those conditions can be investigated before anything is built.

So why does almost nobody check?

How one northern estate revealed a very different possible future

We recently applied this way of thinking to an estate in the north of England, and the results changed our own view of what was possible.

The estate is home to several thousand people. Beneath its streets sits something increasingly valuable: a district heat network, built decades ago and connecting nearly two thousand homes.

Mapping the place revealed both its assets and the pressures it was carrying.

There was an ageing energy centre. Households remained exposed to the volatility of the gas market. Renewal costs competed with housing investment for every available pound. A school and swimming pool were potentially within reach of the existing network.

Most importantly, the estate possessed a network of pipes that a previous generation had already paid for.

Plans for the network's low-carbon future were moving in a sensible direction. A water-source heat-pump solution had been studied and represented a reasonable route forward.

Modelling the future meant testing that proposal properly. But it also meant testing an additional scenario as a matter of due diligence.

What if a modular data centre—not a hyperscale campus, but a facility delivered in a unit little larger than a substantial plant room—were placed beside the energy centre and connected to the existing network?

The appraisal identified more than an alternative heat source.

The scenario suggested that heating for nearly two thousand homes could be decoupled from the gas market. The data centre could create a revenue stream capable of contributing to network renewal, rather than allowing renewal costs to compete with rent and housing investment. It could also provide headroom to extend the network towards the school, the swimming pool and streets beyond the estate boundary.

None of this has been built. It remains analysis, and it should be described honestly as such.

But the exercise changed the decision available to the estate's leaders. Instead of being presented with one technically reasonable proposal seeking approval, they could compare different, evidenced futures and understand what each might create for households, public services, investment and carbon.

The engineering had been available all along.

The greatest value is missed because nobody is paid to see the whole place

The principal barrier is not engineering alone, and it is not simply a shortage of money.

It is the way Britain decides what gets built.

When a major project is proposed, each specialist examines a particular slice. An energy consultant models the power requirement. A transport consultant models the junctions. Other professionals assess jobs, health, carbon, noise, water, ecology and flood risk.

Seven consultants. Seven methodologies. Seven documents. One place.

Each report may be technically competent. Together, they can remain blind.

A town is not a collection of professional disciplines. It is one connected, living system, and much of its real value sits on the joins—precisely where nobody is commissioned to look.

Follow one winter along one street.

A warm, dry home helps end a child's persistent cough. The child misses fewer days at school. A parent misses fewer shifts at work. The local surgery has a quieter winter. Money that previously left the town through the energy meter is spent in local shops. Gas that supplied the heat inefficiently is no longer burned.

The consequences move through health, education, employment, household finance, public services, local economies and carbon.

No single document in the present system contains that whole chain.

Our response to this blindness has often been to commission more reports: social value assessments, health impact assessments, climate strategies, economic appraisals and local area energy plans.

Each may be useful. Each is trying to answer a version of the same question: what will happen if we do this?

But each tends to answer alone, within its own professional language, budget and boundary.

Britain is not short of potentially useful heat. It is short of decisions that connect infrastructure to place.

Slough shows what can happen when cumulative effects are never examined

The cost of failing to see the whole is not limited to opportunities missed. Sometimes it may be felt on the skin.

Slough, ten miles west of Heathrow, contains one of Europe's largest concentrations of data centres. An estimated 30 to 40 facilities operate close to the centre of the town.

In June 2026, the Guardian reported residents describing summer heat that had become increasingly difficult to bear. A weather station close to the principal data centre campus recorded 36.7°C during a heatwave, while some surrounding stations were cooler.

The report also examined emerging research led by Cambridge academics into what the researchers call the “data heat island effect.”

The research, published as a preprint and not yet peer-reviewed, examined changes in land-surface temperature around data centres internationally. It estimated an average increase of approximately 2°C after facilities began operating, with larger effects close to some sites.

That evidence must be handled carefully.

Land-surface temperature is not the same as the air temperature experienced by residents. Slough has also been shaped by decades of urban growth, roads, buildings, hard surfaces and a warming climate. A weather station recording a high temperature near a campus does not establish that the data centres caused it.

Nobody should claim that data centres have been proven to be responsible for Slough's heat.

But something else can be said with greater confidence: the cumulative question was never properly modelled.

Each facility was consented as an individual application, assessed largely within its own boundary, one at a time, over approximately two decades.

Nobody was responsible for asking what the relationship between an expanding cluster and the surrounding town might eventually become—in heat, water, grid demand, noise, construction, public infrastructure or anything else.

Whatever the science ultimately concludes, the planning process was poorly equipped to anticipate the outcome because the process did not examine the cluster as part of a whole place.

Set Slough beside the opportunity described earlier and the lesson becomes clear.

The same physical process that may contribute to warming a place unintentionally could, under the right conditions, help warm its homes deliberately.

The difference is not whether data centres exist.

It is whether their relationship with the place was mapped, modelled and chosen.

A practical method: map the place today, model tomorrow and choose the future

That relationship can be designed.

The method is not complicated. It consists of three steps, undertaken before the masterplan is fixed.

Map todayUnderstand the place as it exists today

Establish what is true about the place now.

Where are the households experiencing fuel poverty? Where are the existing heat networks, schools, leisure centres, hospitals, surgeries and energy centres? What infrastructure does the place already possess? What pressures are its residents, services and institutions carrying?

The purpose is to build a sufficiently complete picture of the place to understand where the opportunities and constraints connect.

The desktop is where almost every modern project begins, and it is a valid starting point for place-based regenerative work — provided it is recognised as the start of the process, not the whole of it. It should be tested, challenged and enriched on the ground, with the people who know the place because they live, work and depend on it.

Model tomorrowModel the asset twice to reveal genuinely different futures

Treat the place as the connected system it is and test more than one possible future.

Model the proposed asset in isolation, within its own boundary.

Then model it again as part of the surrounding place.

Test its recoverable heat, grid connection, water use, construction period and cumulative effects alongside existing and planned development. Trace the consequences across household bills, housing, public services, local investment, health and carbon—not only the effects inside the red line.

The comparison may show that a heat network is viable. It may show that another source is cleaner or more economical. It may show that the data centre should be located differently, designed differently or not connected at all.

Without predetermining the answer, the objective is to make the alternatives visible.

Choose the futureCompare the evidence and choose deliberately while change remains affordable

Place the scenarios side by side, with the benefits, costs, dependencies and trade-offs clearly stated.

Then decide which future creates the greatest combined value for people, place, money and carbon while the decision is still inexpensive to change.

Create one field of view broad enough to make the existing reports add up to a decision.

The cheapest moment to create lasting value is before the masterplan is fixed

Timing determines how much of this potential survives contact with reality.

Asked before a masterplan is fixed, whole-place questions can be relatively inexpensive to answer and act upon.

An energy centre can be positioned on the edge of a campus closest to potential heat users. Pipe corridors can be protected while they are still lines on a drawing. Heat recovery can influence cooling-system design. Local heat demand can become a siting consideration alongside land, fibre, water and grid capacity.

The commercial model can recognise heat as a potential product rather than treating it as an inconvenient by-product.

Asked after consent, the same questions tend to produce additions around the edge: a community fund, landscaping, a plaque or a promise to investigate heat reuse at some later date.

By then, the location, cooling systems, energy centre and development economics may all have been fixed.

The cheapest moment to create a century of value is before anyone draws the red line.

The policy environment is beginning to recognise this wider field of view.

The 2026 Green Book strengthens the role of distributional and place-based analysis in public appraisal and makes clear that a benefit-cost ratio is one input into a balanced judgement, not a verdict on its own.

Heat network zoning is also developing in England. The intention is to identify locations where heat networks are likely to offer a cost-effective route to decarbonisation and, subject to legislation and appropriate exemptions, enable certain large buildings and useful heat sources to be required to connect.

The government has also begun directing individual data centre proposals into the Nationally Significant Infrastructure Project regime and has indicated that a dedicated National Policy Statement for data centres will be developed.

These decisions will establish defaults that influence a generation of development.

Once defaults are embedded in policy, finance and professional practice, they are what everyone builds.

Infrastructure decisions made today will still shape places a century from now

One of Britain's old company villages still stands on the Bedfordshire clay.

In 1926, the London Brick Company opened its works at Stewartby. It built not only an industrial site but houses, a school, a hall, playing fields and a swimming pool around it.

The brickworks closed in 2008.

The village remains. People are still living inside that industrial decision.

Now a data centre campus of national significance is proposed for the former clay pit: approximately 143 acres and up to 720 megawatts, within reach of thousands of homes and public buildings that will need low-carbon heat in the decades ahead.

A campus of that scale could produce more heat than the surrounding area could consume.

The proposed development may also include on-site gas generation. Heat recovered from burning gas cannot automatically be treated as clean heat, and no credible appraisal should obscure that fact.

But this complexity strengthens the case for examining fuel, power, carbon, heat recovery and local demand within one accountable model.

It does not justify looking away from the relationship altogether.

Nobody appears to have been required to place the scale of the proposed heat output and the surrounding need for low-carbon heat into the same decision-making frame.

That sentence contains the argument of this article in miniature.

What developers, councils and national government can do before the opportunity is lost

Developers and operators should design the relationship with place into the project

Commission the whole-place picture before fixing the masterplan.

Model heat recovery and local benefit as part of the core design and commercial strategy, not as an appendix produced during consultation. Hundreds of days of planning uncertainty cost more than understanding the place properly at the beginning.

Local authorities and housing providers should bring their existing assets to the negotiating table

Treat your heat maps, housing estates, public buildings, energy centres and existing networks as part of the negotiating table.

Ask for the asset to be modelled both in isolation and as part of the place before consent, not afterwards. Infrastructure you already own may be the readiest opportunity in the region.

National decision-makers should require every major campus to show what it could create for its host community

Require every nationally significant data centre proposal to demonstrate what it could create for the place that hosts it.

That does not mean obliging every campus to supply heat regardless of feasibility. It means requiring the relevant relationships, scenarios and trade-offs to be examined before the opportunity is designed out.

There is an old principle, and our company is named for it, that a decision should be judged by its effects on the seventh generation to come.

Heat networks can hold that horizon. They are long-lived assets. Pipes laid by one generation can become the opportunity into which the next generation connects.

The future of infrastructure will not be judged solely by what it delivers to a national system. It will also be judged by what it gives back to the places that host it.

The test is simple: not whether a town tolerates its data centre, but whether it would fight to keep it.

The railways made towns. The motorways made warehouses. The AI age is choosing right now, in documents most people will never read.

Imagine what Britain could build if we chose deliberately.

The next campus, the next local plan, the next line in the National Policy Statement are all being decided now. If you are shaping any of them, the moment to map the place, model the asset twice and choose is before the red line is drawn. That is exactly the work HAUK is built for, and the conversation we would most like to have.

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?

Contact Oliver Riley at oliver@generation7.co.uk.