The water challenge of data centres in England: concentration and peak demand

A small sector in total volume, but with localised impacts

Analysis of consumption data shows that 67% of data centres use less than 1,000 m³ per year, figures comparable to those of a medium-sized office building. At the opposite end of the spectrum, a small group—3% of the facilities analysed—exceeds 50,000 m³ per year, with an average consumption within this group of 115,800 m³ annually.

The report estimates that, if available data are extrapolated across England, the sector may be abstracting close to 1.9 hm³ per year from the public water supply. Although this volume is not significant at national scale, the study emphasises that impacts must be assessed locally and, in particular, in terms of peak demand.

One of the principal findings is the marked seasonality of consumption, with increases during the summer months coinciding with higher ambient temperatures and therefore greater cooling needs. This pattern concerns water companies, not because of the total annual volume, but due to the coincidence of data centre demand peaks with maximum domestic consumption during heatwaves.

Average monthly water consumption of data centres compared with maximum ambient temperature. Source: Water Efficient Data Centres (WRc for MOSL, 2026).

Cooling at the centre of the debate

The report confirms that water use is concentrated almost exclusively in cooling systems, as humidity control has only a marginal impact.

The technology employed is decisive. In England, most data centres use cooling systems with little or no significant water consumption, or hybrid systems that only use water under high-temperature conditions. According to sector analysis, approximately 5% of facilities may be using hybrid cooling systems that significantly increase consumption during summer.

The study highlights the indicator Water Usage Effectiveness (WUE), which relates water consumption to the energy used by IT equipment. At European scale, the average WUE stands at 0.58 l/kWh, although differences between Member States are considerable. Some countries report values clearly above the average, while others are well below, depending on climate, average facility size, and prevailing cooling technologies.

Average WUE of data centres by EU Member State. Source: Water Efficient Data Centres (WRc for MOSL, 2026), based on EED data.

When WUE is analysed by facility size, the pattern also changes: larger-scale data centres tend to record higher average values, which the report links to a greater presence of water-based cooling technologies in high-capacity installations.

Average WUE by data centre size. Source: Water Efficient Data Centres (WRc for MOSL, 2026), based on EED data.

Rapid growth and misalignment with water planning

One of the clearest messages of the report is that sector growth—driven by digitalisation and the expansion of artificial intelligence—is not yet fully integrated into water resource planning.

The National Framework for Water Resources 2025 explicitly mentions data centres as new significant demand sources and warns of the need for greater transparency and collaboration to understand their present and future requirements.

The mismatch between five-year water resource planning cycles and the rapid pace of new data centre development makes it difficult to anticipate demand scenarios, especially in regions experiencing water stress or network constraints.

Recommendations: mandatory reporting and a common benchmarking framework

Following a review of existing regulatory frameworks, the report concludes that England should move towards a mandatory and centralised reporting system, aligned with the European Energy Efficiency Directive (EED).

Key recommendations include:

• Mandatory reporting of metrics such as total water consumption, potable water consumption and Water Usage Effectiveness (WUE).
• Disaggregation by water source (potable, reused, abstracted, etc.) and by end use.
• Separate reporting of annual average consumption and peak-month consumption.
• Consideration of minimum performance standards for new facilities.

The study also suggests evaluating the adoption of an efficiency labelling system, similar to that proposed by the European Commission, integrating both energy and water indicators and facilitating benchmarking between installations.

Greater collaboration and focus on alternative water sources

Finally, the report identifies as a priority the strengthening of collaboration between the data centre sector, water companies, and regulators, particularly regarding the use of treated effluent as an alternative to potable water.

Although potable water remains the predominant source, there are technical opportunities to use non-potable resources with additional treatment, thereby reducing pressure on the public supply.

The core message is clear: data centres are not currently major water consumers in aggregate terms, but their growth, geographic concentration, and peak-demand profile require stronger integration into water resource planning and the development of a common framework for transparency and efficiency.