The Real Numbers

In the court of public opinion, data centers are increasingly being tried for their thirst. Headlines warn of AI "drinking" our reservoirs, often citing that a simple conversation with a chatbot consumes a bottle of water. As water professionals, however, we have a duty to move beyond clickbait and engage with the hydro-engineering reality.

The standard industry metric is Water Usage Effectiveness (WUE), calculated as liters of water used per kilowatt-hour of energy consumed. In their latest environmental reports, major hyperscalers reported significant water consumption: Google consumed approximately 6.4 billion gallons globally in 2023, while a 2024 Lawrence Berkeley National Laboratory report estimated U.S. data centers consumed 17 billion gallons directly through cooling—with projections that this could double or quadruple by 2028.

The "bottle of water per AI query" statistic is often misunderstood. Research from UC Riverside estimates that while training a model like GPT-3 consumed roughly 700,000 liters of freshwater, the operational water cost per query is significantly lower—a conversation of 20-50 questions consumes about 500 milliliters, meaning a single short query is closer to 10-25ml depending on cooling efficiency.

The Location Problem: Where Water Stress Meets Digital Demand

The tension lies not in the global volume of water used, but in the local timing and source of that use. Data centers are hyper-localized stressors. According to S&P Global analysis, 43% of data centers globally are operating in areas of high water stress in the current decade—and this is projected to rise to 45% by the 2050s.

If a hyperscale facility is built in a water-stressed basin like the American Southwest or parts of Spain and relies on evaporative cooling during peak summer months, it competes directly with municipal supplies. A low global WUE is meaningless to a local utility manager if the specific facility in their jurisdiction is drawing millions of gallons of potable water during a drought.

Direct vs. Indirect: The Hidden Water Footprint

A critical distinction often missed in public discourse is the difference between direct and indirect water consumption. Direct water use encompasses cooling at the data center itself—space humidification, direct adiabatic cooling, and heat rejection in cooling towers.

However, the International Energy Agency estimates that 60% of data center water consumption is from indirect use—water consumed by the power plants generating their electricity. Lawrence Berkeley Lab estimated that in 2023, U.S. data centers consumed an additional 211 billion gallons of water indirectly through electricity generation—12 times greater than direct cooling use.

"Data centers are essential infrastructure, much like hospitals or water treatment plants. They drive the very digital tools—smart metering, leak detection AI, and digital twins—that we use to conserve water." — Ashwin Dhanasekar, Brown & Caldwell

Engineering Solutions: The Shift to Water-Positive

The good news is that the industry is not standing still. We are seeing a rapid shift toward "water-positive" commitments and a transition in cooling technologies.

Immersion & Liquid Cooling: The Future

As chip density increases for AI workloads, air cooling is becoming insufficient. The industry is moving toward direct-to-chip and immersion cooling, supported by funding from the U.S. Department of Energy's COOLERCHIPS program. These technologies involve submerging servers in dielectric fluids that capture heat far more efficiently than air—often consuming near-zero water.

Immersion cooling is growing at an 18% CAGR and offers significant advantages: it requires only about one-third of the space of air-cooled configurations and can support rack densities exceeding 380 kW (compared to 5-7 kW for traditional air-cooled cabinets). Because immersion systems rely on synthetic dielectric fluids rather than evaporative processes, they dramatically reduce the freshwater burden.

Wastewater Reuse: The Sensible Path Forward

Perhaps the most "sensible" path forward is decoupling data centers from potable supplies entirely. Leading examples include facilities in Loudoun County, Virginia and Douglas County, Georgia, which utilize reclaimed wastewater for cooling towers.

However, in the U.S., alternative water sources contribute less than 5% of water used by the industry, according to the 2024 Lawrence Berkeley National Laboratory report. The quality of cooling water affects equipment performance—reclaimed water can cause more corrosion, scaling, and microbial growth than potable water, requiring additional treatment infrastructure.

Regulatory Landscape: What's Coming

Regulation is tightening globally. Key developments include:

A 2025 poll across five European countries found that 72% of respondents support mandating new data centers only be built if new renewable energy sources are created to power them, with strong support for water-use restrictions in stressed regions.

A Call for Context

We must remain vigilant but rational. Demonizing data center water use ignores the efficiency gains they enable elsewhere. Instead of panic, we should advocate for:

• Transparency: Better reporting on consumptive water use (lost to evaporation) versus withdrawal (returned to source)
• Location-aware siting: Demanding that data centers in high-stress areas utilize non-potable water or dry-cooling technologies
• Integrated planning: Engaging water utilities in hyperscale siting decisions before permits are granted

The "sensible" approach is not to stop the growth of digital infrastructure, but to integrate it intelligently into our water systems. We have the technology to cool our servers without draining our future—we just need the engineering will to deploy it.