Adaptable pipework key to sustainable data centres

Victaulic is urging data centre operators to treat mechanical infrastructure choices, including pipe-joining methods, as a core part of long-term sustainability and resilience planning as power densities rise and cooling approaches shift.

It positions this as a gap between commonly used efficiency measures and the practical question of how facilities evolve over time. Power Usage Effectiveness (PUE), Water Usage Effectiveness (WUE) and Carbon Usage Effectiveness (CUE) remain standard indicators. At the same time, operators face pressure to maintain high availability while reducing energy and water consumption.

As IT loads change and refresh cycles shorten, adaptability in cooling and other mechanical systems is drawing more attention. Pipework and joint design can affect how quickly a site can be expanded, reconfigured or maintained, as well as how much waste and disruption upgrades and retrofits create.

Pipe joining is one example of an infrastructure decision that sits below headline sustainability targets but can have lasting operational effects. Victaulic argues that connections designed for maintenance and rerouting reduce the need to replace installed pipework when requirements change.

"Key mechanical system decisions, such as adopting adaptable pipe joining methods, can extend system lifecycles, reduce waste, and simplify upgrades," said Lissette Piedra van Dommelen, Territory Sales Manager for Iberia at Victaulic. "It's about building for the long term, ensuring a facility can adapt to meet future demands."

Metrics and continuity

Operators increasingly describe sustainability as an environmental, technical and economic discipline rather than a single efficiency number. This framing links energy and water targets to maintenance regimes, redundancy strategies and asset-life management.

"It's an approach that involves selecting low-energy technologies, designing optimised thermal architectures, and implementing solutions that ultimately reduce both PUE and WUE metrics," said José Farrona, Senior Solutions Engineer at Prime Data Centres.

"Equally important, however, is operational continuity through redundancy and predictive maintenance. This ensures uninterrupted service while avoiding inefficiencies," he said. "Operational sustainability demands the rational and optimised use of all resources, including support systems, human expertise, and physical infrastructure, to maximise asset longevity and minimise waste."

In practice, that places more attention on how physical systems are commissioned and modified. Data centres often expand in phases, adding halls or changing cooling distribution as demand shifts. Mechanical components that are difficult to alter can shorten maintenance windows and increase demolition and rework during upgrades.

AI and density

Rising interest in AI infrastructure is increasing scrutiny of power and cooling design as rack densities and heat loads grow. Operators are assessing whether existing mechanical layouts can cope, and how quickly changes can be made without affecting live environments.

"Data center infrastructures must be conceived with modularity, scalability, and operational adaptability at their core," said Pedro Muñoz, Global Switch Operations Director. "We now see the rise of AI-driven business growth challenging traditional infrastructures with shorter technology lifecycles and increasing power densities."

He also pointed to the need for "flexible, plug-and-play designs that allow for rapid, risk-free adaptation without interrupting live operations serving multiple clients."

Victaulic argues that modular mechanical systems reduce the need to lock in assumptions early in a build. It highlighted installation approaches that support phased construction and later rerouting, which can affect project timelines and long-term resource use. It also pointed to build methods that avoid hot works during assembly, changing labour and safety requirements on site.

Cooling transition

Cooling remains one of the sector's most material environmental issues, with operators balancing efficiency targets against equipment reliability. Many facilities have improved performance through air management and tighter control of setpoints, water flows and airflow. The industry is also experimenting with wider use of liquid and immersion cooling, particularly for higher-density workloads.

As these approaches evolve, pipework configuration becomes more consequential because the cooling method may change over a facility's life. Reconfigurable systems can affect the cost and complexity of moving from one approach to another, as well as downtime risk during retrofit projects.

Victaulic argues that adaptable pipework can support "circular" infrastructure practices, favouring reuse over replacement. It also links adaptability to the ability to respond to regulatory changes and shifting client requirements without large-scale reconstruction.

Organisational shift

The discussion is also extending beyond engineering into operations and workforce practices. Operators increasingly emphasise standardised processes and collaboration across sites, particularly among large providers running multiple regional facilities.

"Operational sustainability requires a cultural and organisational shift. Beyond infrastructure and technology, it involves programmatic initiatives to transform behaviours, foster collaboration across teams, and leverage synergies between sites and regions," said Jordi Vila Rotxés, Data Centre Operations Manager at Equinix Spain.

Operators and suppliers expect pressure on energy, water and carbon performance to remain high as demand grows. They also anticipate greater focus on designs that can be modified without extended outages as AI deployments and cooling technologies evolve.