AI data centers use 2.7M gallons of water daily, stressing North Carolina communities

A single AI data center could consume as much water as a small town, with industry estimates showing a 250-megawatt facility using about 2.7 million gallons daily or nearly a billion gallons annually. This massive water demand is raising critical questions for North Carolina communities about resource planning, climate resilience, and whether the economic benefits of data centers justify their environmental costs.

What you should know: AI data centers require enormous amounts of water for cooling systems, and unlike other water users, much of this water is lost to evaporation rather than returned to the ecosystem.

• “It’s a meaningful chunk of water that would otherwise be reserved for residential or commercial growth,” said Ed Buchan, assistant director of Raleigh Water, comparing the demand to entire communities like Knightdale and Wendell.
• Most water utilities return nearly all treated water back to river basins, but cooling systems at data centers evaporate it, effectively removing it from the local water cycle.

Why peak usage matters: Water demand from AI facilities spikes during the hottest months, creating direct competition with residential users who also need more water in summer.

• “Understanding the peak usage is also important in addition to the annual total,” said Shaolei Ren, a University of California Riverside professor who studies AI’s water footprint.
• This timing creates particular stress on water infrastructure when resources are already stretched thin.

The scale problem: While data centers represent a small fraction of national water use compared to agriculture, their local impact can be substantial for individual communities.

• A Washington Post analysis found that a single ChatGPT query uses only about five drops of water, but these amounts multiply across billions of requests and massive cooling demands.
• Roughly half of data centers are not publicly disclosing their water usage, making resource management challenging for local utilities.

Innovation attempts: Some institutions are exploring creative solutions to reduce AI’s water footprint while maintaining computing capabilities.

• Duke University researchers are investigating ways to recycle heat generated by computing power: “We’ve looked at using the heat from GPU farms to warm up water that’s used across campus,” said Michael Pencina, Duke’s vice dean for data science.
• AI companies have pledged to become “water positive” by 2030, though transparency around actual usage remains limited.

Local infrastructure concerns: North Carolina water utilities are preparing for potential capacity challenges as more AI facilities consider the region.

• Raleigh has sufficient supply for one or two large facilities, but multiple data centers could accelerate the need for new reservoirs or other costly infrastructure projects.
• “If you had a significant number of them come into your service area, and then they were the type that evaporated a lot of water, then that might move up your need for a new water resource pretty quickly,” Buchan explained.

The economic tradeoff: Communities must weigh the substantial property tax benefits of data centers against their resource consumption and infrastructure demands.

• “Data centers clearly have benefits,” Ren noted. “On the other hand, they also use resources like power and water. The real question is whether those tradeoffs add up to a net benefit for the community.”