For colocation providers, sustainability is no longer a side initiative. It must now be embedded in everyday operations alongside resilience and uptime. Providers face the challenge of reducing energy intensity and water use, increasing renewable electricity adoption, while maintaining the service levels customers expect. At the same time, capital projects must be designed to deliver over 15–20 years, leveraging new or modular technologies to keep pace with evolving sustainability commitments.

Operators are increasingly under evidence-based scrutiny, and growth in demand is only intensifying this. The IEA estimates global data center electricity consumption at around 415 TWh in 2024 and projects could rise to about 945 TWh by 2030 in its base case, driven largely by AI and digitalization.

As the global data center footprint grows, regulators and customers want comparable information rather than broad promises. In Europe, that shift is reflected in structured reporting expectations under the Energy Efficiency Directive framework, Delegated Regulation (EU) 2024/1364, and Corporate Sustainability Reporting Directive (CSRD), which set out KPIs and reporting requirements.

Over the next three to five years, some of the biggest gains are likely to come from upgrading older energy-intensive and/or inefficient cooling systems and introducing newer methods, including liquid cooling for higher rack densities. Colocation providers must integrate new cooling systems within live environments without disruption or downtime. Sustainability therefore needs to be treated as a practical discipline, built into design choices and day-to-day operations, with the appropriate governance and willingness to evolve.

Build and upgrade choices that lower whole-life impact

When a facility is refurbished, one of the first questions should be what existing features can be retained. Embodied carbon and demolition waste can account for a large share of a site’s impact, so retrofitting can be a stronger option to avoid unnecessary reconstruction and preserve value already embedded in the structure, provided colocation requirements can still be met.

Telehouse South provides a practical example. In September 2024, we brought two additional floors online as part of our refurbishment programme. Retaining this structure delivered carbon offset advantages and reduced local disruption. In a live colocation environment, work of this kind still needs to be carried out while keeping operations uninterrupted.

Delivering upgrades in this kind of setting requires careful planning. In practice, that means:

Changes must be phased carefully, tested in sequence and carried out in a way that gives customers a clear migration path if moves between halls are required. In most occupied facilities, modular delivery is the practical option because it allows capacity to be added in smaller stages and aligned more closely with demand.

Equipment choices also shape ongoing impact and sometimes they create benefits beyond the facility itself. At Telehouse South, the cooling system operates with higher water temperatures to improve efficiency, with free-cooling chillers using ambient temperatures for low-carbon heat rejection throughout the year. The system does not use water for adiabatic cooling. More broadly, local conditions can sometimes support lower-impact cooling at city scale. Enwave’s Deep Lake Water Cooling system in Toronto is a good example, using cold water from Lake Ontario to cool hospitals, commercial buildings, residential buildings and data centers.

That does not mean every market has the same options. The broader point is that stronger sustainability outcomes usually come from designs that respond to local conditions instead of applying a standard template everywhere. Alongside cooling strategy, operators may also need to assess whether waste heat can be exported to district heating, because routing, redundancy, and available plant space often determine whether heat reuse can be introduced without adding unnecessary complexity.

Operational performance that stands up to scrutiny

Once the upgrade path is set, performance still depends on daily discipline. Airflow management, containment, and close control of systems all influence whether design efficiency translates into live operation. This is also where lower Power Usage effectiveness (PUE) and lower Water Usage Effectiveness (WUE) become practical rather than theoretical.

Colocation brings an extra challenge because contracted capacity and actual customer load do not always match. When a site is provisioned for much more power than customers are drawing, plant can run outside its most efficient range. Addressing this typically comes down to a combination of:

Transparency is equally important when it comes to renewable electricity. Claims need to be supported by clear evidence, including what has been procured and how it aligns with frameworks such as the EU Code of Conduct on Data Centre Energy Efficiency.Colocation brings an extra challenge because contracted capacity and actual customer load do not always match. When a site is provisioned for much more power than customers are drawing, plant can run outside its most efficient range. Addressing this typically comes down to a combination of:

Operational performance that Partnerships and governance that make sustainability real up to scrutiny

As scrutiny rises, focus often moves beyond the facility itself. Waste heat reuse is the clearest example, though it only works when engineering and governance are aligned. Heat must be available at a viable temperature, supported by the infrastructure and local demand to use it consistently.

In practice, projects become viable when local climate and temperature requirements allow meaningful heat transfer, when there is a district heating network and local heat consumer that can utilize heat consistently, and when the integration can be delivered without service disruption in a live environment. When those conditions are present, delivery works best with early agreement on interfaces, liabilities, and long-term operating responsibility, so the model remains financeable and operationally clear.

The most common obstacle is that the heat generated by a data center may not be sufficient for direct use by local heating systems. That is why heat reuse is often best delivered as part of a wider local energy plan, where responsibilities are aligned: the data center supplies a predictable source, and the network operator and local stakeholders look to deliver distribution and long-term obligations.

Community impact needs the same level of attention. A credible operator explains what is being built, engages early with local stakeholders, and works to reduce avoidable nuisance from traffic or standby generation. Governance is what keeps those commitments credible over time. It means aligning ESG priorities with business goals, keeping evidence behind public claims, and testing sustainability decisions alongside resilience and risk.

These priorities point to a clear conclusion. Over the next three to five years, progress in colocation is likely to depend less on broad commitments and more on consistent execution. This requires assessing contributing factors – such as needs and expectations, local conditions, and the best available techniques and technologies – within governance structures that enable flexibility and adaptability long term. More specifically, that means upgrading legacy systems without disrupting live environments, introducing new cooling approaches only where they deliver measurable value, tightening day-to-day operational control, and forming partnerships that are supported by credible governance. When handled in that way, sustainability becomes a way to align growth with efficiency, accountability, and trust.