A new $60,000 ISSE seed project will examine how the rapid growth of artificial intelligence data centers could affect communities across Tennessee by considering factors like energy demand, water use, noise, land use, and local property markets.
The project, “The Local Footprint of AI Data Centers: Noise, Water, and Land-Market Tradeoffs in Tennessee,” is led by UT researcher Yuefeng Hao and runs from Aug. 15, 2026, through June 30, 2027. Over the span of the project, Hao and his team will work to develop a scenario-based model that will help in evaluating where future data centers may be most appropriately located.
As the rapid growth of artificial intelligence continues, Hao expects demand for the infrastructure that supports it to continue to increase as well.
“If we’re going to use more AI, we’re going to need more data centers,” Hao said.
With this anticipated growth, a number of questions arise within the communities where these new facilities may be developed. Data centers can place significant demands on electricity and water resources, while cooling equipment can generate noise that may affect nearby neighborhoods. Hao also plans to consider broader environmental and land-use factors, including water-resource conditions and the possible relationship between data-center development and surrounding property values.
Energy demand is one of the project’s primary considerations. During periods such as summer heat waves, electricity use is already elevated as homes and businesses rely more heavily on air conditioning. In adding the demand of large data centers, additional pressure is placed on local energy systems.
Another factor Hao is concerned with is water use. Data centers require substantial water resources for cooling, which places a heightened focus on local water availability and conditions as an important part in the evaluation of potential locations—Hao’s background in hydrology will help guide that portion of the research.
The project will bring these factors together in a scenario-based decision model. Rather than considering a proposed data-center location through a single measure, the model will incorporate variables such as land use, economic conditions, community factors, policy considerations, climate conditions, water resources, and electricity infrastructure.
The long-term goal is to translate that model into a web-based decision-support tool. Users could select a potential location and evaluate the issues most important to them, such as water, energy, or housing impacts. The tool could then provide an assessment or map illustrating the potential tradeoffs associated with developing a data center in that area.
Hao envisions several groups using the research. Policymakers could use the model to support decisions about where data-center development may or may not be appropriate. Members of the public could use the tool to better understand the potential impacts within their own communities. Researchers could also benefit from the data collected through the project.
Building the research dataset itself may become an important contribution. Hao noted that detailed information about data-center operations can be difficult to obtain because of the rapidly developing nature of the industry alongside some of the operational data not being publicly available. His team hopes to supplement existing information by visiting facilities, establishing industry connections, and directly measuring factors around known data-center locations.
For Hao, the project brings together three areas that have shaped his research: hydrology, agriculture and land systems, and optimization.
“Before, I was making maps and wasn’t always sure how people could use them,” Hao said. “With this kind of website, the research can become something people can actually use.”