"The report found that spotty legal requirements for environmental review make it difficult to predict the potential impact of new data centers on the state's water system. In the South Bay, which already has the densest collection of data centers of any region in the state, the lack of clarity could hamper local planners' ability to navigate unpredictable environmental challenges, such as climate change and drought."
"Many data centers use water to cool sensitive electronic equipment, though the quantity varies widely between facilities - from virtually none to the equivalent usage of thousands of households. The report's authors said they reached out to every water provider in water districts where data centers are located, but none provided data about water use, either for individual developments or for the cumulative usage of such projects."
"The researchers are recommending the state strengthen its requirements for such reviews. "The accelerated concentration of these AI data centers and the plan to build them has put an increased pressure on the water planners to get a better handle on what they're actually getting into," Iris Stewart-Frey said."
More than a dozen data center projects are being developed in Santa Clara County, with local leaders promising thorough environmental impact assessments before facilities begin operating. A statewide survey of data centers conducted by Santa Clara University researchers and Next 10 finds that water impacts are hard to predict because legal requirements for environmental review are inconsistent and incomplete. The South Bay’s high concentration of data centers increases uncertainty for local planning, especially under climate change and drought conditions. Researchers recommend strengthening state requirements for environmental reviews. They note that data centers use water for cooling, with usage varying widely, but water providers did not supply information about water use for individual projects or cumulative totals.
Read at San Jose Spotlight
Unable to calculate read time
Collection
[
|
...
]