Assessing life cycle impacts of data centre cooling

USA: Microsoft has published a report charting life cycle assessments of four data centre cooling technologies – air cooling, cold plates, one-phase immersion and two-phase immersion.

The report quantifies for the first time how much energy and water are consumed and greenhouse gas emissions produced across the entire lifespan of the data centres. It also digs into what was required to produce all the virtual machines, chips, servers, cooling and other support equipment, such as extracting raw materials, manufacturing components, transportation at different points in the process and even eventual disposal. 

The study, which took more than two years to produce, found that switching from air cooling to cold plates that cool data centre chips more directly could reduce greenhouse gas emissions and energy demand by around 15% and water consumption by 30% to 50%. 

Cold plates are a newer technology that Microsoft is deploying in its data centres. Considered a type of direct-to-chip cooling, the coolant is pumped in a loop to a flat container that sits right on top of the chips in a server rack.

The study found that cold plates and the two immersion cooling technologies reduce greenhouse gas emissions by 15% to 21% over their entire life cycles, energy demand by 15% to 20% and water consumption by 31% to 52% compared to air cooling. 

The team expected the three liquid methods to outperform air cooling on carbon emissions and water and energy consumption, but those benefits hadn’t been quantified across the entire life cycle of those technologies.

Microsoft maintains that two-phase immersion has potential for reductions in all areas, but it currently uses liquid PFAS substances, which are under regulatory scrutiny in the EU and the USA.  

The report published in Nature covers cooling chips for general compute, or CPUs, not the specialised chips designed to handle AI workloads. The Microsoft team is reported to be working on a follow-up to examine life cycle impacts of AI chips and expects to see similar improvements with advanced cooling methods.

Open research

Microsoft plans to use the findings to inform new data centre designs and cloud operations and to help meet its broader sustainability goals. It is also making the methodology available to others in the industry through an open research repository, enabling anyone in the industry to plug in their own data and scenarios to conduct a life cycle assessment of their own operations.

“A lot of people do life cycle assessments after the fact,” to understand a data centre’s environmental impact after it’s built,” said Husam Alissa, director of systems technology in cloud operations and innovation at Microsoft and leader of the lifecycle assessment study. 

“When we’re trying to make future design decisions, we typically look at total cost of ownership, performance, sustainability and other factors. We’re advocating in this paper for the use of life cycle assessment tools to guide engineering decisions early on and also sharing the tool with the industry to make adoption easier.

“Our intention is not to say, ‘this is the right technology’. They all could be. There are different circumstances that make you use a technology,” Alissa said. “What we’re trying to do here is tell the industry, ‘Here’s how you build an end-to-end life cycle assessment that takes cooling into account. And here is a tool for you that you can customise to your specific needs and then make a decision’.”

Data centre operations depend on external factors such as local energy grids. The Nature paper also quantified how much energy, water and greenhouse gas emissions could be saved by switching from a typical energy grid to 100% renewable sources of energy, finding that greenhouse gas emissions could be reduced by 85% to 90% regardless of what cooling technologies were used.

“It was interesting to see that cold plates could be as good as the two immersion cooling methods,” said Teresa Nick, Microsoft’s director, natural systems and sustainability for cloud operations and innovation and co-author of the paper.

Complicated

While calculation of design factors such as total cost of ownership, availability, time to market and even reliability is mostly straightforward to quantify and compare, Microsoft recognises that sustainability impacts can be hard to define and calculate across an entire supply chain and data centre ecosystem.

Getting information about how raw materials were obtained, as well as the carbon, water and energy involved in manufacturing, can be difficult, the report authors admit. The researchers pressed suppliers to divulge such data though not all of them participated, then produced formulas so they – and others – could estimate the figures in the future. 

“Having the embodied emissions known, public and shared in databases could help accelerate life cycle assessment efforts,” Alissa added.

Life cycle assessment can also be used to inform all aspects of data centre structure and function, including how to optimise and run a data centre most efficiently. One technology might do better on one measure, another better on a different one, with none of them outperforming in every area.

“In a nutshell, we’re trying to understand the trade-offs,” Nick said. “You’re trying to understand the context of what you’re doing and what the impacts are.”