Team looking to tap underground ‘thermal batteries’ to cool AI Data centers and save water

Researchers from the Prairie Research Institute at the University of Illinois Urbana-Champaign propose utilizing aquifer thermal energy storage (ATES) to cool AI data centers across the United States. Led by Illinois State Geological Survey researchers Yu-Feng Lin, Andrew Stumpf, and postdoctoral researcher Upasana Pandey, their study, published in Groundwater, suggests that this system can significantly reduce both electricity consumption for cooling and water usage.

ATES involves pumping cool groundwater from an aquifer through subsurface pipes to absorb heat from data centers via a heat exchanger. The warmed water is then returned underground for storage, with the process reversible for seasonal heating and cooling. This method leverages the Earth's stable underground temperatures, offering substantial energy savings compared to adjusting for drastic outdoor temperature swings.

Illinois is identified as an ideal location for ATES due to its distinct seasonal temperature variations, abundant aquifers, and suitable subsurface geological properties. The researchers emphasize that the water used can come from non-potable sources like deep, salty aquifers or contaminated groundwater, addressing water scarcity concerns.

While ATES systems have higher upfront costs, they offer lower long-term operating expenses over their 20- to 40-year lifespan. The primary barriers are financial evaluation metrics, which often prioritize shorter-term returns, rather than scientific or technical feasibility, as drilling expertise already exists within related industries.