Data Centres Friend or Foe
A single photo upload, card payment, video call, train app refresh or AI query can travel through a data centre before it reaches its answer. Most of the time, nobody notices. The building sits behind fencing on the edge of a town, humming quietly, while daily life depends on it.
That is why the debate around data centres can feel so split. One side sees them as the hidden engine of modern life. The other sees huge sheds that consume electricity, strain local grids, use land, and offer fewer local jobs than promised.
Both views contain truth. Data centres are neither saviours nor villains by nature. They are infrastructure. Like roads, reservoirs and power stations, they can serve the public well, or they can create costs that fall unfairly on nearby communities.
The real question is not whether society needs data centres. It does. The question is what kind of data centres we build, where we put them, and who benefits from them.
Why data centres have become essential
A data centre is a building designed to house computer servers, storage systems, networking equipment, cooling plant, backup power and security systems. That sounds dry until the services become visible.
They support:
Online banking and card payments
Hospital systems and digital records
Streaming, gaming and video calls
Retail websites and delivery tracking
Government services
Cloud storage for photos, documents and software
AI tools that need large amounts of computing power
In short, data centres are part of the plumbing of the internet. If they stop working, the effect can be immediate. Shops may struggle to take payments. Websites can go offline. Apps may fail. Businesses can lose access to files and systems.
For the UK, this matters because so much of public and private life now runs through digital services. A reliable digital economy needs physical places where data can be stored and processed. These places require power, water or other cooling systems, fibre connections, skilled staff and strong security.
Seen from this angle, data centres are a friend. They make remote work possible, help small firms use powerful software without owning their own servers, and support services people now rely on without thinking.
They also bring strategic value. Countries want control over digital infrastructure in the same way they care about ports, energy networks and transport links. If sensitive data, critical services and computing capacity sit too far outside national control, that can create risk.
So the argument for data centres is not weak. It is practical. Modern life already depends on them.
The costs are real and local
The problem is that digital services can feel weightless while their infrastructure is anything but. A data centre is a physical site with physical needs. It must draw power. It must reject heat. It must connect to the grid. It must be built somewhere.
That can create pressure in several ways.
Electricity demand can be hard to absorb
Data centres run around the clock. Servers need steady power, and cooling equipment often adds more demand. As AI use rises, some facilities need even more electricity because advanced chips produce significant heat.
This does not automatically mean a data centre is bad. Large hospitals, railways and factories also use large amounts of power. The issue is whether local and national grids can support the extra load without delaying homes, public services or other employers trying to connect.
In parts of the UK, grid capacity is already tight. New demand can require costly upgrades and long connection delays. If a data centre takes up capacity while local housing or renewable projects wait, the public may fairly ask who gets priority.
Water use can cause concern
Not every data centre uses the same cooling method. Some rely mainly on air cooling or closed-loop systems. Others use water-based cooling, which can be efficient for temperature control but raises concerns during dry periods.
The water question is very local. A site in a water-stressed area deserves more scrutiny than one with a secure supply and a system designed to minimise consumption. Communities should not have to accept vague promises. They need clear information about expected water use, peak demand and safeguards.
Land use can feel one-sided
Data centres often need large plots near power and fibre connections. They may compete with housing, farming, green belt land or industrial space that could support more jobs.
A large warehouse-style building can also change the feel of a place. Even when well screened, it may add fencing, security lighting, substations and service roads. Local residents may see disruption during construction but little direct benefit once the site is running.
That is where trust can break down. If a development looks private, distant and resource-hungry, people will resist it, even if they use the services it supports every day.
Friend or foe depends on design choices
The phrase Data Centres Friend or Foe suggests a simple vote. In practice, the answer sits in the details. A good data centre is not just a building full of servers. It is a set of choices about energy, heat, water, land, waste, resilience and community benefit.
Here are the choices that matter most.
Question | A better answer looks like | A worse answer looks like |
Where is it built? | On suitable brownfield or industrial land with grid access | On sensitive land with weak planning reasons |
How is it powered? | Clear plans for low-carbon electricity and grid support | Vague claims about being “green” |
How is it cooled? | Efficient systems matched to local climate and water limits | High water demand with little public detail |
What happens to waste heat? | Heat reuse explored for nearby homes, pools or industry | Heat vented with no local value |
Who benefits locally? | Training, apprenticeships, business rates and community gains | Short construction disruption and few lasting links |
How long will equipment last? | Repair, reuse and responsible recycling plans | Rapid replacement with poor e-waste control |
The best operators will be open about these points before planning permission is granted. The weakest will talk in general terms and hope the technical detail keeps the public quiet.
That approach no longer works. People understand that digital infrastructure has consequences. They may not know the exact cooling load of a server rack, but they know when a project feels overbearing, secretive or unfair.
A responsible data centre should be able to answer plain questions in plain English:
How much electricity will the site need at full capacity?
Will it require grid upgrades, and who pays for them?
What cooling system will it use?
How much water will it use on a normal day and during hot weather?
Can waste heat serve nearby buildings?
What will happen to old servers and batteries?
How many permanent jobs will there be?
What measures will reduce noise and light spill?
What local benefits are legally tied to the project, rather than simply hoped for?
If those answers are strong, the case for the data centre becomes stronger too.
The environmental issue is bigger than one building
It is tempting to look at one data centre and treat it as the whole problem. That misses the larger pattern. The demand comes from everyone who uses digital services, including households, businesses, schools, public bodies and entertainment platforms.
Every cloud backup, AI model, video stream and connected device sits somewhere. Deleting one proposed building does not delete the demand. It may move the work to a less efficient site elsewhere.
That does not excuse poor projects. It means the better goal is to shape demand and infrastructure together.
Cleaner power matters
A data centre connected to a greener electricity system causes less harm than one tied to fossil fuel-heavy power. In the UK, the electricity mix has become cleaner over time, but clean generation and grid capacity still need to grow.
Operators can sign power purchase agreements, support new renewable generation, use batteries to help manage demand, and design sites that can reduce load when the grid is under stress. These measures are useful only when they lead to real changes, not when they serve as marketing claims.
Efficiency still matters
The sector has improved efficiency in many areas, especially through better cooling, server use and monitoring. But growth can swallow efficiency gains. If each unit of computing becomes cleaner but total computing demand rises much faster, overall pressure can still grow.
This is especially relevant for AI. Some uses of AI may save time, reduce waste or improve research. Others may produce low-value outputs at high energy cost. The public debate should include both the supply side and the demand side.
Do all tasks need the largest model? Should companies measure and disclose the energy cost of heavy computing? Can software be written to do the same job with fewer resources? These questions belong in the data centre debate.
Waste heat should not be ignored
Servers turn electricity into heat. Often, that heat is removed and released into the air. In some locations, it can be captured and used in district heating, swimming pools, greenhouses or nearby industrial processes.
Heat reuse is not always simple. It depends on distance, temperature, pipework, demand and ownership. Yet it should be tested more often, especially for large sites near towns. Wasted heat is not just an engineering issue. It is a missed chance to make local benefit more visible.
Communities need more than reassurance
Many objections to data centres are framed as resistance to progress. That is unfair. Public concern often comes from poor engagement, unclear benefits and past experience with developments that promised more than they delivered.
A community may support digital infrastructure and still reject a bad plan. Those positions can sit together.
Good engagement means starting early, using clear language and publishing meaningful information. It also means admitting trade-offs. A data centre may bring business rates and some skilled jobs, but it will not employ as many people as a busy factory on the same footprint. It may improve digital resilience, but it may also add pressure to the grid. It may be cleaner than older facilities, but it still has an environmental cost.
People can handle honesty. What they tend to reject is a polished consultation that feels like a decision already made.
Local authorities also need the skills to judge proposals. Data centres are technical, and planning teams may face well-resourced applicants. Councils need access to independent advice on energy, cooling, noise, water, biodiversity and grid impacts. Without that, the balance of knowledge sits too heavily with developers.
National policy has a role too. If data centres count as critical national infrastructure, planning and grid rules should still protect local interests. A national need does not erase local impact. It raises the duty to manage it well.
A fair approach would include:
Stronger standards for energy and water reporting
Clearer rules on suitable locations
Better assessment of grid impacts
Serious review of waste heat options
Local training routes linked to long-term roles
Stronger conditions on noise, light and biodiversity
Public reporting after the site opens, not just before approval
The last point matters. Promises made during planning should not vanish once the building is live. Communities deserve to know whether predicted impacts matched reality.
So are data centres friends or foes?
They are friends when they support essential services, reduce risk, use cleaner power, minimise water demand, reuse heat where possible, and bring real local value. They are foes when they drain resources, hide behind vague claims, occupy land badly, and leave communities with costs but little benefit.
The better question is: who sets the terms?
If the terms are set only by speed, profit and demand for more computation, data centres will keep attracting resistance. If the terms include public value, environmental limits and local consent, they can become a necessary part of a cleaner, more resilient digital society.
The public should not be asked to choose between digital life and environmental responsibility. That is a false bargain. The task is to demand better buildings, better locations, better power choices and better accountability.
Data centres are here to stay. Whether they become friend or foe depends on whether they are built as private machines for endless demand, or as public-facing infrastructure with clear duties to the places that host them.
