Green Party seeks critical infrastructure de-listing of data centres

Under current rules, data centres are classed alongside sectors such as energy, water and food as critical infrastructure, a status that can prioritise them for access to utilities during shortages. The Greens say that blanket treatment risks giving data centres precedence over households, public services and food production in droughts or other supply-constrained situations.

A single, large-scale data centre can have substantial resource demands: the party cites figures suggesting such a facility can use around 1.5M litres of water per day, roughly equivalent to the daily needs of a town of about 10,000 people. Construction activity in the sector is projected to increase sharply – the Greens say the value of data-centre projects could quadruple by 2029 – and much of that growth is concentrated in London and the South East, regions already flagged as water-stressed.

Green Party leader Zack Polanski, speaking today, called for an immediate policy change to remove the blanket critical-infrastructure label from new data-centre builds and to introduce stricter planning thresholds. He argued that the designation should be reserved for facilities demonstrably essential to national security or public services, rather than being applied to centres serving private technology firms.

“We need clear criteria to distinguish between data centres vital to national security or public services and those that simply serve private tech companies and private profit,” Polanski said, adding that local communities should have a say in planning decisions and that no data centre should be prioritised for water supply over households or essential services during droughts.

The party is also pressing for new environmental requirements for planning permission, including demonstrable compliance with high environmental standards and full community consultation. Its proposals include siting new facilities on brownfield land close to renewable energy sources, requiring centres to be capable of operating on 100% renewable power where possible, auditing and publishing annual water and electricity consumption figures, and investment in less water-intensive cooling and heat-recovery technologies to feed local heat networks.

Beyond environmental controls, the Greens have linked the debate to national strategy on artificial intelligence infrastructure. They argue the UK’s growing dependence on foreign-owned data infrastructure and opaque “black-box” AI systems increases strategic vulnerability and reduces transparency. Their proposals call for increased domestic investment in AI infrastructure, tighter regulation, enhanced transparency, protection for workers, and community-led decision making over new builds.

Reaction from the industry and government has not been published alongside the Green Party’s demands. Data-centre operators have repeatedly argued that they are essential to the digital economy and that modern facilities are increasingly efficient, but critics point to continued heavy water and energy usage in many designs and the geographic clustering of new sites in regions with constrained resources.

For infrastructure professionals and planners, the Greens’ proposals raise practical questions about how to balance digital connectivity and economic development with resource security and local consent. Key technical considerations include the choice of cooling technology (air, evaporative, or liquid cooling), availability and sustainability of water sources, grid capacity and resilience, opportunities to colocate with renewable generation, and the feasibility of heat capture and reuse.

Any move to restrict planning fast-tracking or to reclassify data centres would also have regulatory and commercial implications. Local planning authorities could gain greater leverage over siting and environmental conditions, potentially slowing deployment times. Conversely, tighter requirements on renewables integration, audited consumption reporting and mandatory heat-recovery systems would increase upfront costs for developers but could reduce long-term environmental impacts.

Environmental implications of data centres

The environmental impact of data centres is a growing concern as the rapid adoption of artificial intelligence (AI) drives unprecedented demand for electricity and water. While data centres are critical infrastructure for the digital economy, their physical and carbon footprints are expanding significantly, leading to increased scrutiny over their long-term sustainability.

Data centres are among the fastest-growing consumers of electricity globally. In the UK, they currently account for approximately 2.5% of national electricity use, a figure expected to quadruple by 2030. AI is the primary driver of this growth; while it currently contributes 5-15% of data centre energy use, it is projected to account for 35-50% by the end of the decade.

For grids still dependent on fossil fuels, this surge in demand directly increases greenhouse gas emissions . As Lalitha Krishnamoorthy explains in AI’s hidden energy crisis: How industry can build a more sustainable future.

Water usage and cooling demands

Cooling is a major environmental pressure point, as high-density AI workloads generate intense heat that traditional air cooling often cannot manage.

• Consumption Levels: A 1MW data centre can use up to 25.5M litres of water annually—equivalent to the daily consumption of roughly 300,000 people.
• Global Impact: Global data centre water use is projected to more than double to 1,200bn litres by 2030.
• Technological Shifts: To mitigate this, many operators are transitioning to closed-loop liquid cooling or dry cooling systems, which recirculate a fixed volume of liquid to minimise losses.

The rapid evolution of AI hardware also contributes to environmental costs through electronic waste and hardware retirement. Managing the full AI lifecycle, including the appropriate retirement of data and hardware, is considered essential for responsible deployment.

Mitigation and sustainable innovations

Industry leaders are exploring several strategies to balance growth with net-zero goals:

• Waste Heat Recovery: Data centres generate significant residual heat that can be redirected to warm local homes and businesses through district heating networks. For example, London’s data centres potentially release enough waste heat to warm hundreds of thousands of homes annually.
• Renewable Integration: Companies are increasingly sourcing 100% renewable power or co-locating facilities near clean energy sources like wind farms or nuclear plants.
• Grid Flexibility: Data centres can act as “active” grid assets by shifting non-urgent workloads (like AI model training) to periods when renewable generation is high, reducing the need for fossil fuel backup.

Like what you’ve read? To receive New Civil Engineer’s daily and weekly newsletters click here.