Episode 5: Castrol’s Nick Barrett on Liquid Cooling at Scale

Liquid cooling has moved past the pilot stage. The question now is whether the rest of the industry – the designers, the contractors, the facility teams and more – actually understands what they're maintaining.

In this episode of the Data Centre Podcast, host Ben Craske sits down with Nick Barrett, Data Centre OEM Liaison at Castrol, to talk thermal management, liquid cooling deployment and what it takes to scale high-density AI infrastructure without costly mistakes. Nick brings over 15 years experience at Castrol and a background in automotive engineering, and has worked exclusively on data centres since 2021.

🎧 Listen and subscribe on Spotify here

▶️ Watch on YouTube here

In this episode we explore:

• Why liquid cooling is already running at scale, not just in pilot projects
• Why education, not just engineering, is the real barrier to safe deployment
• Why cooling fluids need planning from day one, not as an afterthought
• Why ecosystem collaboration is critical to reducing risk and accelerating adoption
• What megawatt racks mean for the next generation of cooling infrastructure

From automotive engineering to data centres

Nick's path into data centres came via more than 15 years in fuels and lubricants at Castrol, originally trained as an automotive engineer. He joined Castrol's data centre initiatives in 2021, bringing the company's heritage in thermal management and fluid technology into a sector that was only just starting to take liquid cooling seriously.

The education challenge behind liquid cooling

Deploying liquid cooling successfully takes more than trained operational engineers, Nick explains. Designers, consultants, contractors, facility teams and end users all need a genuine understanding of how these systems work and how they should be maintained. As rack densities keep climbing, that shared understanding is what stands between long-term reliability and expensive mistakes.

The reality of scaling liquid cooling

While much of the industry still talks about liquid cooling as something on the horizon, Nick is clear that it's already operating successfully at a significant scale. Fluid management, commissioning, system flushing, monitoring and end-of-life planning all matter, and cooling fluids shouldn't be an afterthought bolted on late in the process. They need to be considered from the earliest stages of data centre design.

Why ecosystem collaboration matters

No single organisation solves this alone. Hardware manufacturers, cooling providers, research institutions and engineering partners all need to work together from the earliest design stages. Validation and material compatibility testing, done collaboratively, are what reduce risk and speed up adoption across the industry.

Castrol's role in the future of data centres

With more than a century of experience in thermal management and lubrication, Castrol is bringing expertise from industries such as automotive, motorsport and advanced manufacturing into the data centre sector. Nick explains how the company is evolving beyond supplying fluids to delivering end-to-end support, including commissioning, fluid monitoring, maintenance and lifecycle management. This growing role reflects the increasing importance of thermal management as data centre infrastructure becomes more complex.

Preparing for the era of megawatt racks

As AI workloads push rack densities past 300kW, then 600kW, and eventually toward a megawatt per rack, thermal management is set to become one of the defining challenges in data centres. Nick points to research programmes and Open Compute Project working groups already preparing the industry for what's coming — because the infrastructure of the next few years won't look like the infrastructure of today.

Subscribe and Listen
Episode 5 is brought to you by Data Centre Magazine.

Want to hear more? Check out Episode 4 of The Data Centre Podcast, to hear AVK's Charlotte Berry-Selwood discuss talent, leadership and diversity in the data centre industry.