Poor water planning could leave data centres AI and dry

Sweco author: Andy Smith, Technical Director

Artificial intelligence (AI) is changing the way we do things and driving significant acceleration in digital innovation. Central to this innovation is the development of data centres – the heart of the physical infrastructure needed to support AI’s growing demands for cloud storage and processing. However, a critical element of the required existential growth in said infrastructure that requires immediate attention is the substantial water consumption that can be associated with cooling these facilities.

As a consultant providing specialist technical water environment advice for some of the world’s largest AI data centres, I have a unique perspective on the mounting pressures this industry places on an already scarce and vital resource. Robust water planning will be essential to ensure that the very technology propelling us forward does not conflict with the sustainable development agenda.

The confluence of AI, data centres, and water usage

Data centres are the backbone of AI development, housing the servers and infrastructure that power machine learning models and large-scale computations. These facilities require massive amounts of energy, much of which is used for cooling.

In many cases, water is a preferred cooling medium because it is more energy-efficient than air cooling, reducing energy consumption and associated carbon emissions. However, this efficiency comes at a significant cost: almost unprecedented water demand.

To illustrate, a 2024 Environmental Report from one major data centre operator outlines that they have an annual cooling total water consumption that exceeds 20 billion litres, primarily due to water cooling needs. As AI adoption surges, this volume is set to rise exponentially, raising critical questions about resource allocation and environmental sustainability.

Balancing technological progress with environmental sustainability

The benefits of AI are undeniable, from improving healthcare outcomes to enabling sustainable urban planning. However, the pursuit of these advancements cannot come at the expense of our environment. It is imperative that water use, availability, and alternative cooling technologies are considered from the inception of data centre projects. Failure to do so risks solving one problem – energy efficiency – while exacerbating another: water scarcity.

Data centre operators and planners must integrate sustainable practices into every stage of DC development. The planning system must also evolve to address these challenges, ensuring a harmonious balance between technological progress and environmental preservation.

Strategies for sustainable water management in data centres

1. Location

Operators must consider water availability within the catchment when determining the most appropriate location for their data centres, this must be evaluated as a key element of the due diligence prior to land purchase / option agreements.

2. Reducing Water Use Intensity

Operators must focus on minimising the water used per megawatt of IT capacity. Advanced cooling technologies, such as closed-loop systems, can help achieve this goal.

3. Replenishing Water in Stressed Catchments

Data centres should replenish more water than they consume, particularly in areas already facing water scarcity. This could involve investing in ecosystem restoration projects or improving local water quality.

4. Adopting Alternative Water Sources

Non-potable and freshwater alternatives, such as recycled wastewater or used industrial water should be prioritised where feasible. These options reduce pressure on potable water supplies and can be integrated into innovative cooling solutions.

5. Site-Specific Solutions

Every water catchment area has unique challenges, requiring tailored approaches to water management. Operators must conduct detailed hydrological assessments to identify the most sustainable solutions for each site.

6. Enhancing Ecosystem Health

Data centres should go beyond water neutrality, actively improving the health of ecosystems in the areas they are located. This could include restoring wetlands or investing in biodiversity initiatives.

7. Knowledge Sharing

Collaboration is key to addressing global water challenges. Data centre operators should share tools, technologies, and best practices to drive industry-wide improvements in water efficiency and sustainability.

The Role of Policy and Planning

To ensure the integration of these sustainable practices, it is essential to enhance the planning system. Robust policies and regulatory frameworks should be established to enforce public sustainability and environmental commitments within the data centre industry.

This approach will facilitate the alignment of technological advancements with environmental stewardship.

Conclusion

The expansion of AI technologies presents both opportunities and challenges. By proactively addressing the water consumption associated with data centre operations through strategic planning and sustainable practices, we can support the growth of AI while preserving our precious water resources and promoting environmental sustainability.

—

For water consultancy support visit https://www.sweco.co.uk/services/water-energy-industry/water-consultancy/

For data centre consultancy visit https://www.sweco.co.uk/services/data-centre-design/