Rep. Alexandria Ocasio-Cortez presented muddy water samples to EPA officials, alleging that Meta's new data center in Morgan County, Georgia, has contaminated the local water supply. The incident highlights growing concerns about the massive water consumption of hyperscale data centers and their environmental impact as AI-driven expansion accelerates.
The recent congressional hearing featuring murky water jars from Morgan County, Georgia, has thrust data center environmental impacts into the national spotlight. Rep. Alexandria Ocasio-Cortez presented the contaminated samples to EPA Assistant Administrator for Water Jessica Kramer, prompting a commitment to investigate the alleged connection between Meta's data center construction and the community's deteriorating water quality.
The technical reality of data center water consumption paints a concerning picture. Modern hyperscale facilities typically consume between 1.8 to 3.6 million gallons of water per day for cooling purposes alone. Meta's Georgia facility, like other large-scale data centers, likely utilizes evaporative cooling systems that account for approximately 20-40% of a data center's total water usage. When compared to traditional office buildings that consume about 10 gallons per square foot annually, data centers require up to 100 times more water per square foot.
The water usage effectiveness (WUE) metric, which measures water consumption relative to IT equipment energy use, reveals the scale of the issue. Industry average WUE stands at 1.8 liters per kilowatt-hour, with some facilities exceeding 5 liters/kWh in hot climates. This translates to approximately 20,000-50,000 gallons per megawatt of IT load annually. For context, a single AI training cluster can consume over 1 million gallons of water monthly—enough to supply 3-4 average American households for a year.
The Morgan County situation highlights a critical technical challenge: data centers often locate in areas with abundant water resources, but their consumption can deplete local aquifers faster than natural recharge rates. The reported 29 million gallons siphoned by another Georgia data center in just 15 months represents approximately 0.1% of the area's annual groundwater recharge in some regions. While this may seem small, localized impacts can be significant, especially in rural communities relying on private wells.
From a supply chain perspective, water scarcity is becoming a critical factor in data center site selection. Major hyperscalers are increasingly investing in water recycling technologies, with some facilities achieving 80-90% water reuse through closed-loop cooling systems. However, these technologies require significant capital investment—increasing construction costs by 15-30%—and may not be feasible in all locations.
The market implications of these environmental concerns are substantial. With 70% of Americans now opposing data center developments near their homes, and over 20 states considering or implementing data center moratoriums, the industry faces significant regulatory headwinds. This opposition directly impacts the hyperscaler expansion plans that are critical to supporting AI infrastructure growth.
The tension between AI compute demands and environmental constraints is creating a new market dynamic. Companies that develop and implement water-efficient cooling technologies are gaining competitive advantage, with some achieving 40-60% water reduction through advanced liquid cooling and direct-to-chip cooling solutions. The market for sustainable data center technologies is projected to reach $45 billion by 2028, growing at 18% CAGR.
The EPA's investigation into Meta's Georgia facility will set an important precedent for how environmental regulations may be applied to data centers in the future. Should the agency establish stricter water usage guidelines or reporting requirements, it could significantly impact data center operating costs and construction timelines. Industry groups are already advocating for federal standards that would provide consistency across state regulations.
As the digital economy continues to expand, the data center industry must balance its growth with environmental stewardship. The Morgan County incident serves as a reminder that technical innovation in computing must be matched by innovation in resource management. The future of hyperscale infrastructure may depend not just on processing power, but on how efficiently these facilities can operate within their ecological constraints.
Comments
Please log in or register to join the discussion