In the past 20 years, the internet has exploded, becoming an integral part of our daily lives. From streaming movies to online shopping, making our online lives bigger than ever. But behind the scenes, data centres – massive buildings filled with racks of servers and tech gear, quietly powering our digital world – have grown even bigger and more complicated.

Data centres have grown larger and more complex, housing a multitude of Information and Communications Technology (ICT) devices, from powerful servers and central processing units (CPUs) to memory units and storage devices. Keeping all this stuff cool is a big job too, with cooling systems components such as chillers and pumps working hard.

But here’s the catch: the environmental impact. As concerns about climate change continue to rise, so does the need for sustainable solutions. Data centres are being heavy consumers of energy, consuming massive amounts of electricity to power and cool their equipment. That’s where smart, energy-efficient data centres cooling system come in.

Did you know that around 40% of a data centre’s energy goes into cooling? That’s a lot! But there’s hope. Enter liquid cooling systems – a promising alternative to traditional air-cooling methods. These systems use liquids, like water or coolants, to absorb and dissipate heat more efficiently, leading to significant energy savings. From immersion and spraying techniques to direct-to-chip cooling methods, there’s a variety of liquid cooling solutions on the market, each with its own benefits and challenges.

The Wasa Zero Emission Data Centre (WSTAR) is a research facility we’re constructing in Vaasa, Finland, the energy capital in Finland. It’s part of the Technobothnia research infrastructure and offers direct and near connections to the energy industry. WSTAR also supports research on data processing and storage. The project aims to revolutionise how data centres operate. Learn more at the Wasa Zero Emission Data Centre WSTAR .

That’s where my research comes in. My research focuses on taking the data centre liquid cooling system to the next level. By integrating cutting-edge technologies like Model Predictive Control (MPC) and deep reinforcement learning (DRL), I’m developing a novel control strategy MPC-DRL that optimises energy usage while minimising environmental impact. Think of it as a smart thermostat for data centres – constantly adjusting settings to maintain optimal performance and efficiency.

But it’s not just about saving energy, we can make data centres greener and more efficient. My research also explores innovative ways to reuse waste heat generated by data centres. By capturing and repurposing this heat for tasks like district heating or greenhouse heating, we can further reduce our carbon footprint and create a more sustainable energy ecosystem.

Moreover, this control strategy isn’t just good for the planet – it’s good for business too. By improving the reliability, latency, and cost-effectiveness of data centre operations, we’re helping businesses maximise their resources and stay competitive in today’s fast-paced digital landscape.

And the best part? I’m not just theorising – I’m putting my research to the test. Through rigorous real-world experimentation in WSTAR data centre, I aim to demonstrate the effectiveness of the proposed control strategy in optimising liquid cooling systems and enhancing overall data centre performance.

So, while the internet keeps growing, let’s make sure we’re keeping our planet green too. With smarter, more energy-efficient data centres, we can build a greener future for generations to come.

Hussain Kahil

Doctoral Researcher
School of Technology and Innovations
University of Vaasa, Finland