Tackling Data Centre Water Use with Digital Realty

As AI continues to boom, data centres are feeling the heat and confronting power and energy consumption concerns.

Critically as part of this, water use is a challenging consideration, given that AI needs plenty of water to keep cool. Whilst the technology industry is eager to start developing cooling solutions, there are concerns that it could throw up sustainability challenges.

With this in mind, we hear from Aaron Binkley, VP of Sustainability at Digital Realty. Currently, the company is focused on integrating sustainability into the core of its business, with more than 300 data centres across 25 countries. The company recently debuted one in Crete using Schneider Electric solutions.

“As AI drives demand for data centres, it’s on us to innovate without compromising the planet – and it’s on me and my team to make sure we do it right,” Aaron says.

How do data centres use water?

Water has traditionally been used by the cooling systems in data centres. Some facilities still rely on evaporative cooling, which takes water and evaporates it to cool air and keeps server temperatures in an acceptable range. It’s effective and energy efficient but can use a lot of water, especially in large facilities with high computing demands.

Water use in data centres has evolved a lot in recent years. Many newer data centres – including those we’re designing – have moved away from water-based cooling systems altogether to more sustainable, water-free cooling systems that minimise water use and keep energy efficiency high. In fact, our standard data centre designs haven’t incorporated cooling systems that use water on a constant basis for over a decade. Air-cooled and closed-loop cooling systems – which recirculate cooling fluids within a sealed circuit – are gaining traction for good reason. Generally, they’re more efficient, dramatically reducing water waste by reusing it continuously. It’s a major step forward for sustainable cooling.

What is changing about data centre water use?

We have heard from communities concerned about potential data centre water usage, especially as AI and other high-performance computing needs grow. But things are evolving fast for the better: in fact, 75% of our global portfolio of more than 300 data centres don’t use water for cooling at all. And we're seeing major improvements in how water is used in those that do still rely on it.

AI is fast becoming a powerful tool for driving efficiency. Through our partnership with Ecolab, we’ve deployed an AI-powered water management solution that pinpoints opportunities to cut water use that aren’t immediately obvious to the human eye. It’s projected to save up to 15% annually, or around 126 million gallons, every year. That’s smart tech making a real-world impact.

Innovations like this are not just about reducing water or energy consumption; they’re also improving the overall efficiency of our data centres, which is essential as we look toward the future of digital infrastructure.

How does this still need to change?

There’s real progress being made – but there are always ways to go further. Some data centres haven’t adopted water-saving technologies or aren’t moving fast enough – especially in regions already under high water stress, where every drop counts.

As an industry, we need to go beyond just installing tech. We need a holistic approach to water management – one that factors in local water availability, prioritises efficiency in facility design, leverages the expertise of our data centre operations teams and uses every tool we have, to cut consumption.

Transparency matters too. It’s not enough to say we’re doing the right thing – we need to track it, measure it and share it, because customers, investors and communities deserve to see real progress.

It’s time for operators to set clear water-reduction targets, raise the bar on reporting and push harder on innovation. We’re only scratching the surface – especially with tools like machine learning, which hold massive untapped potential to drive even greater efficiency.

How can data centre operators optimise water use?

There are several ways to approach optimising water use. Choosing the right cooling system should be the first priority, with air-cooled systems – or closed-loop systems for those facilities that do still require water - quickly becoming the default choice. 

Beyond that, the use of AI and data analytics is vital and we’re already seeing huge benefits of AI for monitoring and optimising water use in real-time. That’s because of its impressive ability to analyse massive amounts of operational data and recommend specific actions that can lead to water savings and increased recycling of water. It’s probably the area where we see the greatest near-term potential.

Collaboration has to be an important part of the equation, too. We don’t have to figure this all out on our own: strategic partnerships give us access to specialised expertise that can help us drive innovation in water conservation, simplify adoption and accelerate the development of new solutions that can be of benefit to the whole industry.

Transparency and accountability are, in my view, make-or-break factors in how we tackle water use. When we set clear goals and openly track our progress, we create the pressure – and the momentum – to do better. That’s how we build a culture of continuous improvement across the entire sector.

Ultimately, optimising water use in data centres requires a careful combination of cutting-edge technology, smart decision-making and industry-wide collaboration. I certainly see it as an achievable goal.

Explore the latest edition of Data Centre Magazine and be part of the conversation at our global conference series, Tech & AI LIVE. 

Discover all our upcoming events and secure your tickets today.

Subscribe to our Weekly Newsletter for the latest insights, trends and expert analysis.

Data Centre Magazine is a BizClik brand