Data centres long benefited from being both essential and largely unseen. That invisibility is now fading. As sites become larger, energy demand becomes more politically sensitive, and grid constraints tighten, data centres are increasingly visible in the places that host them. 

When infrastructure becomes visible, the public conversation changes with it. The question is no longer only whether a project is economically valuable, but whether it is responsible, proportionate and integrated into its surroundings.

The issues that surface are rarely cosmetic. They are the practical questions that recur in planning committees, consultations and political debate because they map onto real constraints and lived impacts: pressure on electricity networks, operational carbon intensity, water use during dry periods, backup generation, noise and local air quality. If those fundamentals are not addressed credibly, no sustainability narrative carries much weight. Heat reuse is not a shortcut around them.

Even where the fundamentals are handled well, the sector’s sustainability story often remains defensive and abstract: minimising harm, meeting standards, contracting for renewable power and satisfying reporting frameworks. Necessary, certainly, but often detached from the question communities and local authorities ask more directly - does this development only draw from the local system, or does it contribute something the place can recognise and value?

Why heat, and why now

That is where heat becomes interesting, but not because every data centre will become a district heating anchor. Most will not. The economics are highly location-specific. Viable heat export depends on proximate demand, a credible network route, workable commercial structures and an operational model capable of supporting an interface safely and predictably. Many schemes remain pilots for precisely those reasons.

What is changing, however, is the way heat networks themselves are being framed. Increasingly, they are treated not as niche sustainability add-ons but as long-lived civic infrastructure. Once that framing takes hold, the relevant question shifts away from whether a site will export heat immediately and towards whether the asset is being designed so future integration remains possible. I have started to think of that as the ‘second connection’.

The first connection is obvious - electricity makes a data centre operational. The second connection is whether the asset can connect into wider civic infrastructure over time rather than functioning as a sealed asset sitting apart from the place around it. Heat is not the only possible form of second connection, but it is one of the few outputs a data centre produces that is local, physical and publicly legible. Where it works, it can help decarbonise nearby buildings and strengthen the viability of heat networks that might otherwise remain marginal.

Designing for optionality

Crucially, second connection thinking is about optionality. Even where heat export is not commercially viable on day one, preserving the ability to connect later may still matter because it avoids designing the option out altogether. In practice, that often turns on decisions that appear minor until they become irreversible, including safeguarding space for future interface equipment and metering, protecting workable routes to the site boundary, ensuring security layouts and servicing strategies do not sterilise future corridors, and maintaining a land and rights position capable of supporting future connection if circumstances change.

Planning is where precision becomes especially important. There is a meaningful difference between describing a scheme as capable of facilitating future heat export and committing to obligations that later become commercially uncertain or operationally unrealistic. Vague commitments can harden into years of friction as policy evolves and expectations shift. A more durable approach is to be explicit about what is being safeguarded (space, routes, allowances and technical capability) while being equally clear about what is not being promised, such as immediate connection, guaranteed heat output or funding third-party infrastructure. Done properly, this is not about avoiding responsibility; it is about ensuring consent remains deliverable and financeable over the life of the asset.

Where optionality is quietly lost

Real estate documentation is another place where second connection optionality is often quietly lost. Leases, easements, covenants and lender controls are frequently drafted around a sealed operational model and can make later integration difficult even where all parties support it commercially. If heat export may become viable during the life of the asset, documentation needs to preserve the ability to accommodate it, including by avoiding unintentional barriers to granting rights, permitting third-party access, or installing additional plant and metering.

The commercial structure matters as well. Heat export raises questions around ownership of interface assets, operational priority, performance standards, outages, metering, access and curtailment rights. Those are not reasons to avoid integration. They are the reason the strongest projects tend to approach heat as infrastructure requiring careful allocation of rights and responsibilities rather than as a loosely framed sustainability gesture.

Integration, not imposition

Done well, the second connection can make a data centre feel integrated into a place rather than imposed on one. It turns waste into visible local value and, in the right locations, shifts the sustainability debate from what the asset takes to how it integrates and contributes over time.

Practical steps for preserving the second connection

• Safeguard physical space and routes for future heat interface equipment and metering at the design stage, before layouts become fixed.
• Ensure security perimeters and servicing strategies do not sterilise potential connection corridors to the site boundary.
• In planning submissions, distinguish clearly between what is being safeguarded (space, routes, technical capability) and what is not being committed to (immediate connection, guaranteed output or third-party funding).
• Review leases, easements, covenants and lender controls for provisions that may unintentionally prevent future heat export, third-party access or installation of additional plant.
• Where heat export is a realistic prospect, address ownership of interface assets, operational priority, performance standards, outage protocols, metering and curtailment rights through dedicated commercial documentation rather than informal undertakings.
• Maintain a land and rights position capable of supporting a future connection, including the ability to grant necessary access and infrastructure rights without requiring retrospective consent from third parties.