Data centers consumed 222 billion liters of water directly for cooling in 2025, but that figure could nearly triple to just under 644 billion liters by 2030 without water-saving measures in place. As artificial intelligence adoption accelerates worldwide, the infrastructure powering these systems faces a critical constraint that rarely makes headlines: water. While the industry focuses on securing electrical power for GPU-intensive workloads, the cooling systems that keep servers from overheating are quietly becoming one of the most pressing environmental challenges facing tech companies.

The scale of this problem is staggering. Rystad Energy, a leading energy intelligence firm, estimates that without active mitigation strategies, global water consumption by data centers could rise to 644 billion liters per year by 2030. However, more water-efficient pathways could bring this down to 543 billion liters in a moderate case or as low as 388 billion liters in an aggressive water-saving scenario. These projections underscore a critical reality: the trajectory of data center water use is not predetermined. Technology choices, regulatory frameworks, and investment decisions made today will determine whether the industry faces a water crisis or manages sustainable growth.

Why Does Data Center Water Consumption Vary So Dramatically?

Data centers use water primarily to remove the heat generated by computing equipment, but consumption varies significantly depending on the cooling technology deployed and the geographic location of the facility. While artificial intelligence servers generate substantially more heat than traditional computing infrastructure, the rise of AI does not necessarily mean a proportional increase in water consumption. Rack-level liquid cooling, where coolant circulates directly through server hardware, can reduce the heat burden on facility-level systems and enable greater use of dry cooling technologies.

The industry measures water efficiency using a metric called water-use effectiveness, or WUE, which quantifies water used per unit of IT energy. This metric can also account for the water consumed to generate the electricity used by a data center, capturing both direct and indirect water footprints. Some cooling technologies can reduce on-site water use but require more electricity. For example, dry cooling can save around 2.15 liters of water for every kilowatt-hour of IT load, but requires an additional 0.30 to 0.74 kilowatt-hours of electricity.

The indirect water footprint can be material relative to direct consumption, but its contribution varies significantly depending on the water intensity of the electricity supply. In the United States, the indirect portion of water consumption by data centers can be more than twice that of direct water consumption. This means that a data center's overall water footprint depends on both its on-site cooling system and the water intensity of the electricity it consumes, creating a complex web of environmental impacts that extend far beyond the facility itself.

How Are Major Tech Companies Approaching Water Efficiency?

Major data center operators have reported widely varying water-use effectiveness metrics, reflecting significant differences in cooling technology, regional portfolio composition, and measurement methodology. Amazon Web Services reported an average direct site WUE of 0.12 liters per kilowatt-hour in 2025, while Meta reported 0.19 liters per kilowatt-hour for 2024, and Microsoft reported 0.27 liters per kilowatt-hour for fiscal 2025. Digital Realty reported 0.59 liters per kilowatt-hour and Equinix 0.91 liters per kilowatt-hour in 2025. These values should not be read as a simple ranking, given differences in the treatment of leased and colocation facilities and variation in regional portfolios.

AWS's own 2025 regional values ranged from 0.02 liters per kilowatt-hour in Stockholm to 2.85 liters per kilowatt-hour in Jakarta, a spread of more than 100 times within one operator's reporting framework. This dramatic variation underscores how strongly geography and cooling architecture influence the metric. Major hyperscalers including AWS, Google, Microsoft, and Meta have committed to using water more sustainably, with all four setting targets to become "water positive" by 2030, meaning they aim to replenish more water than they consume.

"Forecasting the water impact of data centers is far from straightforward, with consumption varying significantly depending on cooling technology, climate and whether we measure water withdrawn or water actually consumed," explained Dr. Minh Khoi Le, Global Head of Data Center and Hydrogen Research at Rystad Energy.

Dr. Minh Khoi Le, Global Head, Data Center and Hydrogen Research at Rystad Energy

These companies are working on projects to restore freshwater supplies and improve water efficiency, including in sectors such as agriculture. However, these efforts vary by location and watershed, with greater focus often placed on areas facing high water stress. As a result, the impact on local water resources will vary significantly depending on where data centers are located and how aggressively companies pursue restoration projects.

What Regulatory Frameworks Are Emerging to Address Data Center Water Use?

Water-use effectiveness regulation has started to emerge globally, but often stops at reporting requirements rather than performance mandates. The European Union Commission is planning a Data Center Energy Efficiency package that would introduce a labeling system and performance standards. Singapore's Green Data Center Roadmap aims for less than 2 cubic meters per megawatt-hour, equivalent to 2 liters per kilowatt-hour. Regulations are developing in the United States but vary widely by state.

Texas, for example, recently imposed a moratorium on new data center approvals pending audits of the tax breaks, power, water, and cooling use associated with these facilities. Water stress is also becoming an increasingly important consideration for data center regulation, as water demand is set to grow in regions that already face significant constraints on availability. Regions with high and extremely high water stress are projected to account for 34 percent of the data center sector's total global direct water consumption by 2030.

Steps to Reduce Data Center Water Consumption

• Deploy Rack-Level Liquid Cooling: Implement advanced cooling systems that circulate coolant directly through server hardware, reducing the heat burden on facility-level systems and enabling greater use of dry cooling technologies that consume significantly less water.
• Adopt Dry Cooling in Appropriate Climates: Utilize dry cooling technologies in colder geographic locations where they are most effective, as these systems can save approximately 2.15 liters of water for every kilowatt-hour of IT load compared to traditional water-based cooling.
• Implement Water-Positive Restoration Projects: Invest in freshwater supply restoration and water efficiency improvements in local watersheds, particularly in regions facing high water stress, to offset direct water consumption and support community water security.
• Establish Water-Use Effectiveness Standards: Adopt transparent WUE measurement and reporting practices that account for both direct cooling water consumption and indirect water embedded in electricity supply, enabling better comparison and continuous improvement across facilities.

Requiring the adoption of the least water-intensive cooling technologies could reduce consumption in water-stressed regions by 45 percent, highlighting the potential impact of technology standards in these areas. Some of the most exposed areas globally include Jamnagar and Thane in India and Reeves County in Texas, where data center water consumption is high relative to local water stress.

The convergence of AI infrastructure expansion and water scarcity creates an urgent imperative for the technology industry. Without deliberate action, data center water consumption will continue its upward trajectory, straining local water resources in regions already facing significant constraints. However, the wide variation in water-use effectiveness metrics across operators demonstrates that substantial efficiency gains are achievable through technology adoption and operational best practices. The next three years will be critical in determining whether the industry embraces water-efficient cooling technologies and regulatory frameworks, or whether AI's infrastructure boom becomes synonymous with environmental strain in water-stressed regions worldwide.