Operations and maintenance: The engineering powerhouse behind next-gen data centres

Digital infrastructure is expanding faster than anyone imagined possible. From AI training and streaming, to cloud and edge computing, the demand for constant connectivity and compute power has driven data centres into a new era of complexity. 

Yet, while front-end technologies get the spotlight, it's the often-overlooked operations and maintenance (O&M) behind the scenes that keep everything running smoothly.

As many as 55% of data centre operators globally have reported suffering an outage in the past three years, and with these disruptions costing an average of £7,000 per minute, performance failures have become far more than just technical glitches. Now, they pose serious threats to uptime guarantees, customer satisfaction, security and business continuity.

Global demand for data centre capacity could reportedly more than triple by 2030, and this rapid growth places enormous stress on both power and cooling infrastructure. This surge is driven in part by the rise of gigawatt-scale hyperscale colocation centres, essentially data centre 'campuses', whose unprecedented energy density and compute loads are pushing traditional cooling solutions to their limits and demanding new levels of efficiency and innovation.

We're seeing this market grow first-hand, with our own customers, and the associated challenges that inevitably come along with it. To meet those demands, data centre cooling infrastructure must be re-engineered - not just for energy efficiency or space optimisation, but for long-term serviceability, upgradeability and resilience.

Cooling as a strategic priority for data centres

Cooling has long been a challenge in data centre design, as servers generate significant heat in dense configurations. Maintaining optimal operating temperatures is critical to ensure performance, reliability and hardware longevity, but systems often struggle to keep up with rising thermal loads - especially as rack densities increase and energy efficiency becomes a top priority. 

However, as data centres grow more diverse and geographically distributed, cooling systems must adapt in real time to changing workloads, ambient temperatures and equipment stress. This will ensure efficiency, resilience and performance across varied environments.

Meeting these demands calls for a new generation of engineered systems. Cooling infrastructure must be able to grow organically with the data centre, maintaining high efficiency across a broad range of operating conditions. 

But more importantly, it must be serviceable. Engineers need rapid access to components, maintenance cycles need to be short and non-disruptive, and system monitoring needs to be granular, predictive and actionable.

The latest figures show that 80% of data centre operators believe their most recent significant downtime incidents could have been prevented with better management, processes or configuration. Furthermore, 70% are caused by human error.

This highlights a critical truth for the data centre industry: O&M isn't just a post-deployment concern. Instead, it must be embedded into the design strategy from the outset. 

Designing for maintainability and modularity

The prefabricated and modular data centre market in the UK and EU was valued at £3.1 billion in 2024 and is expected to grow at a CAGR of 15.39%. When compared to traditional data centres, modular designs can be deployed 60% faster - offering both cost and time efficiencies.

From fan layouts and coil placement to materials and mechanical-digital interfaces, systems must be modular in both capacity and structure. This enables upgrades, replacements or expansions to take place without downtime. 

For example, cooling units with modular fan walls and scalable coils allow tailored deployment across data centre footprints - from dense hyperscale to compact edge sites. Our new Fan Wall Unit (FWU) - ApX Series reflects this shift toward modular design. With a scalable capacity range of 100kW to 1000kW, it is designed to adapt to changing thermal loads and layout requirements.

Its use of high-efficiency EC fans, compact footprint and configurable coil options supports efficient airflow management and space optimisation in increasingly dense computing environments. These systems can be integrated seamlessly into diverse layouts, adapting to changing power densities, airflow requirements and expansion plans. 

According to research, data operators in the US typically spend 40% of their total energy consumption on cooling systems, so improving efficiency is essential when it comes to reducing operating costs and carbon footprint. If the US trend is anything to go by, driven by the continued commitment to AI, then it's likely these are patterns we'll soon see here in the UK and across Europe.

The advantages of air cooling methods

Equally, market adoption of liquid cooling is currently at about 10%, yet is expected to triple by 2028. Air cooling, however, continues to be the backbone of most data centre operations globally, with as many as 80% of data centres still relying on these methods as they remain more practical, scalable and cost-effective for the vast majority of workloads.

But to continue to succeed, the air-cooling systems of tomorrow must be smarter than those of the past. High-efficiency electronically commutated (EC) fans, advanced coil designs and custom airflow configurations are essential tools in meeting rising performance demands. These must be delivered within a system architecture that is easy to install and flexible enough to evolve as workloads shift.

By listening to operators and learning from real-world deployment challenges, engineers can build cooling infrastructure that not only works on day one but continues to deliver value over the life of the facility. Facilities that can quickly scale, self-diagnose issues and recover from faults with minimal disruption are more agile, more resilient and, ultimately, more profitable. 

Proactive O&M as a competitive advantage

Recent research from McKinsey & Company shows that predictive maintenance can reduce unplanned downtime by 50%, demonstrating that proactive O&M is now a strategic advantage for next-generation data centres. 

As workloads become increasingly dynamic and complex, adopting this forward-thinking approach isn't just beneficial - it's essential. O&M must be embraced as a continuous journey, evolving seamlessly through every upgrade and performance review. 

This relentless dedication is what keeps data centres agile, resilient and primed to pivot with rapidly shifting technologies and business demands. In today's fast-paced industry, complacency is a risk no one can afford. 

Cooling infrastructure must be engineered with foresight, growing alongside the operators it supports. That's why we work closely with our customers to develop solutions tailored to their specific needs - designed with an engineering-first mindset that prioritises adaptability, performance and long-term value.

This ensures that cooling infrastructure provides the flexibility and scale needed not only to handle tomorrow's challenges, but also to drive continuous innovation and become the foundation for lasting success in the fast-changing world of digital infrastructure.