Can Data Centres Be Green ?

Published on January 21

Squaring AI's Growth with Sustainability: Greener Data Centres and Their Global Impact

Artificial intelligence (AI) is transforming industries, economies, and our daily lives at an unprecedented rate. This surge in AI adoption has driven a parallel boom in the construction of data centres—facilities that house the computing power needed for AI applications. Globally, billions have been pledged toward expanding these centres. However, as data centres scale to meet AI's growing demands, they present a significant environmental challenge due to their immense energy consumption. The question arises: can we balance AI's growth with sustainability? Encouragingly, the answer lies in embracing greener energy solutions and innovative practices.



The AI Boom and Data Centre Expansion

AI's insatiable demand for data processing, training complex models, and supporting applications like ChatGPT and autonomous systems has placed data centres at the heart of the digital economy. Industry leaders like Microsoft, Amazon, and Google have committed billions to building state-of-the-art facilities. A report by MarketsandMarkets projects the global data centre market to reach over $350 billion by 2027, fueled in part by AI advancements.



Why Do Data Centres Consume So Much Energy?

The high energy consumption of data centres stems from two primary sources: powering servers and cooling them to prevent overheating.

  1. Server Power Demand:
  2. Data centres house thousands of servers that perform constant computations for AI training, data storage, and retrieval. AI model training, in particular, is computationally intensive, requiring significant electricity to perform trillions of calculations per second.
  3. Cooling Systems:
  4. Servers generate substantial heat, and cooling them requires energy-intensive systems like air conditioning and advanced liquid cooling. Studies suggest that cooling alone can account for up to 40% of a data centre’s energy use.

Without intervention, this energy demand is expected to grow exponentially as AI adoption increases.



Greener Energy and Solutions

Thankfully, the industry is rising to the challenge. Here are some promising strategies:

Renewable Energy Integration

Many companies are transitioning their data centres to run on renewable energy sources like wind, solar, and geothermal power.

  • Google: Operates data centres with 100% renewable energy.
  • Apple's Viborg Facility in Denmark: Runs entirely on wind energy, setting an example for other global operators.
  • Hydrogen Innovations: Hydrogen is emerging as a potential energy source for data centres. Microsoft has successfully tested hydrogen fuel cells to replace diesel backup generators, offering a cleaner alternative. In 2021, French IT company Atos partnered with HDF Energy to integrate hydrogen fuel cells into its data centres.

Energy Efficiency Innovations

  • Liquid Cooling Systems: Instead of traditional air conditioning, liquid cooling systems directly absorb heat from servers, using far less energy.
  • AI Energy Optimisation: AI is being deployed to optimise energy use in real-time, reducing power requirements during low-demand periods.

Waste Heat Recovery

Some cities and companies are finding ways to repurpose the heat generated by data centres. For example:

  • Stockholm, Sweden: Collaborates with data centre operators through Stockholm Data Parks, redirecting excess heat to warm thousands of homes.
  • UK Innovations: Cities like London and Newcastle have begun exploring similar solutions. Newcastle is piloting a scheme where heat generated by servers is redirected to heat water for local buildings, reducing reliance on traditional heating systems.

Jobs and Economic Growth

The transition to sustainable data centres is also creating significant economic opportunities.

Job Creation

Sustainable practices require expertise in renewable energy installation, facility construction, and advanced cooling technologies. A study by IRENA (International Renewable Energy Agency) projects that increasing renewable energy adoption could create 42 million jobs globally by 2050.

Case Study: Singapore

Singapore, which relies heavily on efficient data centres due to its small size and tropical climate, has introduced strict energy efficiency standards, sparking a wave of job creation in green technology. Companies and governments worldwide are following suit.



Countries Leading the Way

Several nations exemplify best practices:

  • Denmark: A pioneer in wind-powered data centres, exemplified by Apple’s Viborg facility.
  • Singapore: Combines government policy and industry innovation to create highly energy-efficient facilities.
  • United States: Tech giants like Google and Microsoft have committed to achieving carbon-neutral operations by 2030. Microsoft’s use of hydrogen fuel cells for backup power sets a benchmark for the industry.
  • Sweden: Innovators in waste heat recovery and renewable energy use, providing a roadmap for sustainable urban integration.

Conclusion

The rapid growth of AI and the accompanying expansion of data centres present a complex challenge: how can we meet rising energy demands while reducing environmental impact? Solutions such as renewable energy integration, hydrogen fuel cells, energy-efficient cooling, and waste heat recovery are already being implemented with measurable success. Countries like Denmark, Sweden, and Singapore demonstrate that it is possible to align economic growth, technological innovation, and sustainability.

Looking ahead, the global data centre industry must adopt these greener approaches at scale, supported by government policies, private investment, and public awareness.