1. A Surge in AI Data Centre expansion across the UK
The UK’s drive toward AI innovation has spurred an explosion in data centre infrastructure. With the government committing around £14 billion to tech hubs and designating data centres as critical national infrastructure, these facilities are now central to powering the AI future.
Forecasts suggest that by 2028, AI could account for nearly 19% of data centre power demand, with total energy usage expected to double by 2026.
To support this surge, the liquid cooling market is set to skyrocket—from about USD 165 million in 2024 to nearly USD 1.93 billion by 2035—a CAGR of over 25%.
2. Cooling the heat: Why Glycol-based fluids are vital
Liquid cooling: Beyond air methods
- Traditional hot/cold aisle air cooling strategies struggle to handle the heat density from AI workloads.
- Liquid cooling methods like direct-to-chip and immersion cooling offer far superior thermal management for high-density server arrays.
Glycol to the rescue
- Glycol-based fluids—namely propylene glycol (PG) and ethylene glycol (EG)—are popular for liquid cooling due to their high thermal conductivity, helping efficiently transfer heat away from critical components.
- These fluids are non-conductive and safe for immersion environments, making them ideal for both direct-to-chip and full immersion systems.
Efficiency and scalability gains
- Liquid-cooled data centres can enable up to 50% more compute power for the same—or even less—IT power compared to air-cooled setups.
- They also allow for higher operating temperatures in chilled water systems—up to 50/40 °C—helping reduce cooling energy and enabling effective heat reuse (e.g., district heating).
3. Sustainability, water use, and Glycol advantages
Reducing strain on water resources
With AI-heavy data centres consuming massive volumes of freshwater—potentially around 4.2–6.6 billion m³ nationwide by 2027, more than half of the UK’s total water withdrawal—sustainability is no longer optional.
Liquid cooling helps here: closed-loop systems using glycols drastically reduce water use compared to traditional evaporative cooling.
Environmental impact
Traditional fossil-derived glycols do have drawbacks—including high embodied carbon and energy-intensive production.
But bio-based alternatives like Susterra propanediol cut emissions by up to 84% and reduce pump power by around 32%, providing a greener path forward.
4. Industry momentum: From innovation to implementation
- Castrol has established a dedicated “Liquid Cooling Centre of Excellence” in Pangbourne, UK, for co-engineering immersion cooling fluids tailored to data centre demands. These efforts aim to reduce energy usage by ~30% and water use by ~80%, while supporting circular lifecycle strategies.
- Demand for immersion cooling fluid is outpacing supply, with lead times up from 6–8 weeks to 12–18, and price premiums rising 20–25% due to raw material constraints.
5. Why Glycol Cooling Matters for the UK’s AI Ecosystem
| Challenge | Glycol-based cooling benefits |
| Rising AI-driven compute density | High thermal conductivity => efficient heat removal |
| Energy and water sustainability | Closed-loop systems lower water usage and boost energy efficiency |
| Infrastructure scalability | Supports modular, dense deployments with future-ready flexibility |
| Environmental impact | Bio-glycols offer lower carbon and better pump efficiency |
| Industry validation | Research centres ensure reliability and promote adoption |
Wrapping up
As global AI demand grows, the UK’s data centre landscape is evolving rapidly—with liquid cooling technologies taking centre stage. Glycol-based fluids, especially those derived sustainably, offer effective, scalable, and environmentally responsible solutions for managing today’s heat-intensive workloads.
Whether deploying direct-to-chip or immersion setups, manufacturers and operators can tap into growing R&D, supportive infrastructure, and strong market momentum to build AI-ready, efficient, and sustainable data centres.
