The Real vs. Imagined Problems with Data Centers’ Water Use
The article discusses the real and perceived problems with data center water use across the United States, highlighting the lack of transparency from companies and the differing views of researchers and environmental advocates. It explores various cooling methods and their water demands, using specific examples like Microsoft's data center in Mount Pleasant, Wisconsin, where a lawsuit was filed for water use projections. The piece concludes that while data centers aren't the largest water users nationally, local impacts are significant and often fuel community opposition.
The debate surrounding data center water consumption is a significant concern for communities, despite researchers noting it's an understudied topic. While some, like Purdue University's Assistant Professor Yi Ding, warn that data centers can significantly threaten local water supplies, others like Oxford University researcher David Mytton argue that clean energy integration is a more critical issue than water use.
Large tech companies are actively addressing the public relations challenges. Google and Amazon have both announced commitments to be "water positive" by 2030, aiming to replenish more water than they consume. However, a widespread lack of transparency from most data center operators regarding their individual water usage exacerbates public alarm.
Cooling technologies play a crucial role, with water-intensive evaporative cooling contrasting with more efficient closed-loop systems. The article points out that even closed-loop systems have trade-offs, like increased energy consumption and chemical use. A notable instance of the transparency issue occurred in Wisconsin, where an environmental group sued the City of Racine to obtain water use projections for Microsoft's data center campus in Mount Pleasant. The projections, released under court order, showed substantial water usage despite Microsoft's reliance on closed-loop cooling.
Despite national data indicating that data centers are not the leading drivers of water scarcity compared to thermal power and agriculture, the article emphasizes that the issue is acutely felt at the local level. Concerns about dwindling drinking water supplies in specific communities, particularly those facing drought conditions, make the hyperlocal impact of data centers a top priority for residents. Experts suggest that substantial infrastructure upgrades, potentially costing billions, might be necessary to meet future data center water demands and that greater transparency is essential for credible research and community acceptance.