US data centers tripled annual water consumption to 17B gallons
forgeeks.net
August 25, 2026
6 min read
45/100
Summary
US data centers consumed an estimated 17 billion gallons of water in 2023, roughly three times the estimated total in 2014, according to a Congressional Research Service report published by the Library of Congress. The estimate includes water used directly for cooling and indirectly to generate the electricity powering facilities. More than 80% of the total water footprint came from electricity generation, while direct consumption accounted for a relatively small share of overall US water use. The federal government does not systematically track data-center water consumption, so the figure is an estimate assembled from incomplete state and local data rather than a facility-by-facility national accounting. Cooling choices shift demand between water and electricity. Evaporative systems generally use less energy but consume substantial water, while air-cooled systems reduce on-site water use while generally requiring more power. Municipal drinking-water systems supplied 97% of water used by US data centers in the CRS estimate, creating local competition for constrained supplies. The report says AI infrastructure is increasing electricity demand through denser specialized hardware, and consumption may have risen since its data ended in 2023. US water reporting remains fragmented because states and local utilities set requirements, and some municipal service agreements limit disclosure. Members of Congress have introduced bills on reporting and water reuse, though most remain at the introduction stage. European Union rules require data-center operators to report annual freshwater consumption.
Key Takeaways
- US data centers consumed an estimated 17 billion gallons of water in 2023, about three times the estimated amount used in 2014, according to the Congressional Research Service.
- Electricity generation accounted for more than 80% of US data centers’ estimated water footprint, including water consumed outside the facilities themselves.
- The US has no uniform federal system for measuring water consumption across all data centers, and some utility agreements restrict public disclosure of customer volumes.
- Evaporative cooling generally lowers energy use but consumes water, while air-cooled systems reduce direct water demand and generally require more electricity.
- Municipal systems supplied 97% of the water used by US data centers in the CRS estimate.
What the discussion said
The thread spent less time on the headline’s 17 billion gallons than on whether that number means anything without a denominator. Many commenters argued that data-center water use is tiny beside irrigation, livestock, golf courses, and other established demands; to them, treating AI cooling as an environmental emergency is numeracy theater, especially when far larger uses receive less outrage. Several also stressed that cooling water is often evaporated rather than dumped, and that closed-loop, hybrid, or site-specific designs can trade water against electricity. The pushback was sharper than a simple comparison fight. Critics said AI is not replacing almond groves or cattle; it is an additional industrial load, so the relevant comparison is water demand before and after the AI buildout, particularly in drought-prone regions. They also rejected food-versus-AI comparisons as a dodge: agricultural water has obvious human utility, while the value of mass-market LLM output remains disputed. Commenters converged on one practical point: local hydrology matters far more than a national aggregate. Evaporative cooling may be sensible where water is abundant, but using potable aquifers where recharge cannot keep pace is reckless. Skeptics further complained that operators conceal facility-level figures, forcing the public to debate estimates rather than verified withdrawals and consumption.
Where opinion split
The central dispute is whether AI data-center water use is a negligible distraction or a meaningful new environmental burden. Defenders say national-scale comparisons show it is dwarfed by agriculture and leisure uses, while efficient cooling can save substantial electricity. Critics answer that every new AI facility increases total demand, aggregate figures hide local aquifer stress, and questionable AI utility does not justify opaque withdrawals from scarce water supplies.
Community Sentiment
Positives
- Compared with agricultural irrigation and livestock, the reported national AI cooling total looks modest, undercutting claims that data centers alone are driving U.S. water scarcity.
- Evaporative cooling can remove server heat with far less electricity than dry alternatives, making water use a real efficiency trade rather than automatic waste.
- Closed-loop and hybrid cooling designs offer a path to sharply lower on-site withdrawals when developers invest in them and choose sites intelligently.
- Several commenters favor matching cooling technology to local water abundance instead of imposing one nationwide rule that could waste more power.
Concerns
- National comparisons can hide the real harm: a new AI campus is an added load that can drain a stressed local aquifer even if its U.S. share is small.
- Evaporated cooling water leaves the local basin, and mineral-heavy blowdown can become wastewater, so calling it endlessly recyclable glosses over the constraint communities face.
- Comparing AI demand with food production or golf courses dodges the harder question of whether low-value LLM output warrants scarce water at all.
- Operators' reluctance to publish facility-level water data fuels suspicion that estimates understate the burden and blocks informed local siting decisions.
