Data Centers’ Thirst: The Growing Strain on Water Resources

The rapid expansion of artificial intelligence and cloud computing is creating a hidden crisis for community water systems across the United States. A novel study reveals that these systems may struggle to meet the surging water demands of data centers, particularly during peak usage on the hottest days of the year.

Billions Needed for Infrastructure Upgrades

Research conducted by a team at UC Riverside, in collaboration with Caltech, indicates that community waterworks will require substantial investment – estimated between $10 billion and $58 billion – to upgrade infrastructure and accommodate the increasing water needs of data centers. This estimate depends on the pace of data center growth.¹

The study projects that without improved water efficiency, data center cooling systems could require an additional 697 million to 1.45 billion gallons of peak water capacity per day within four years – a volume comparable to the typical daily water supply of New York City. Even with optimistic reductions in water utilize, the demand could equal half of New York City’s annual supply.¹

Evaporative Cooling and Peak Demand

Data centers rely heavily on evaporative cooling systems, which are energy-efficient but require significant water usage. Even as annual water consumption figures often receive attention, the study highlights a critical, often overlooked issue: the ability of municipal utilities to supply enough water during peak demand, particularly on hot summer days.¹

Warehouse-sized data centers can experience a six to ten-fold increase in daily water demand during hot weather, and some planned facilities could see spikes exceeding 30 times average usage. This forces local governments to invest in infrastructure capable of handling these peaks, even if the capacity remains largely unused throughout the year. Some facilities under construction have been allocated up to 8 million gallons of water daily, enough to supply multiple small towns.¹

Challenges Beyond Funding

Shaolei Ren, an associate professor at UC Riverside’s Bourns College of Engineering, who led the research, emphasizes that funding isn’t the only hurdle. “Even if you have money, the water source is another challenge,” Ren said. “In many cases, the water is naturally replenished by snowpack and reservoirs. But reservoirs and snowpack are limited. You may have money to build treatment plants and pipes, but money can’t buy more snowpack.”¹

The nation’s existing water and wastewater infrastructure already faces trillions of dollars in funding needs for upgrades and maintenance over the next two decades, according to the Environmental Protection Agency.¹

Recommendations for a Sustainable Future

The researchers propose several solutions to mitigate the growing strain on water resources:

• Report Peak Water Use: Data center developers should report peak water usage, not just annual averages, to provide a more accurate picture of demand.
• Public-Private Partnerships: Data center companies should collaborate with local communities to fund water infrastructure upgrades.
• Offset Water Use: Data centers should invest in water capacity improvements to offset their own consumption and ensure sufficient supplies for future community growth.
• Adaptive Cooling Methods: Data centers should adjust cooling methods based on grid and water system stress, utilizing water-based cooling when the power grid is strained and switching to dry cooling when water supplies are limited.

UC Riverside and Google Cloud Collaboration

The University of California, Riverside (UCR) is working to modernize its infrastructure through a three-year agreement with Google Public Sector, gaining access to Google Cloud’s computing resources at a fixed subscription rate.² This collaboration aims to facilitate innovation and provide researchers with scalable resources.²

RAISE@UCR and AI Research

UCR’s Artificial Intelligence ReSearch and Education Institute (RAISE@UCR) is dedicated to pioneering AI technologies and addressing challenges across various fields, including environmental monitoring.¹

The study, titled “Small bottle, massive pipe: Quantifying and addressing the impact of data centers on public water systems,” was authored by Shaolei Ren, Yuelin Han, Pengfei Li, and Adam Wierman.¹

“People recognize power as a constraint for data center growth,” Ren said, “but most of them haven’t realized water is a hidden and even more binding constraint in many communities.”¹