The expansion of artificial intelligence data centers and Bitcoin mining is placing new attention on a resource that extends beyond electricity. As governments, utilities, and local communities review new computing projects, water availability has become an important consideration because both industries depend on cooling systems that require water resources.
Water consumption varies significantly between data centers because it is influenced by factors such as cooling technology, climate, seasonal weather, electricity sources, and regional water availability.
Traditional evaporative cooling generally uses more water while consuming less electricity, whereas closed-loop and immersion cooling can reduce water demand. Electricity generation also contributes to indirect water consumption, meaning two facilities with similar computing capacity can have very different environmental impacts depending on where they operate.
Research cited in a report estimates that U.S. data centers consumed about 228 billion gallons of water in 2023, compared with approximately 2.9 trillion gallons used for lawn irrigation.
Transparency Emerges as a Key Policy Issue
Community opposition has grown in locations including Tucson, Arizona, and Aragón, Spain, where proposed data center developments have faced inspection over concerns about local water supplies.
According to the report, policymakers often lack standardized facility-level information because reporting requirements remain inconsistent. In the United States, responsibilities are divided among federal, state, and local agencies, as well as utilities, while voluntary disclosures remain common.
In the European Union, legislation adopted in 2024 classifies facility-level data center information as private, limiting public access despite broader environmental transparency commitments.
The report recommends mandatory facility-level reporting, standardized disclosure rules, smart metering, and better methods for measuring indirect water used in electricity generation. It also calls for greater adoption of less water-intensive power sources, including wind and solar, particularly in regions facing water constraints.
The water discussion also connects to Bitcoin mining, which relies on large-scale computing infrastructure with similar power and cooling requirements to AI data centers.
The overlap has become increasingly relevant as Bitcoin miners face growing economic pressure. Capriole Investments founder Charles Edwards highlighted another challenge affecting miners. He noted that Bitcoin miner transaction-fee revenue has fallen to levels last seen in 2019, when Bitcoin traded near $5,000, despite the network now being valued at roughly 13 times that level.
The data also shows subdued on-chain activity compared with the strong network usage recorded during the 2017 and 2021 bull markets. Edwards said the trend helps explain why some miners are moving toward AI-related computing infrastructure while raising questions about Bitcoin network security if the pattern continues.
AI data centers use water directly for cooling systems and indirectly through the electricity generation needed to power their computing infrastructure.
Future expansion may depend on transparent water reporting, efficient cooling technologies, and access to sustainable energy resources.
Both industries process large volumes of computational workloads, requiring high-performance computing facilities supported by substantial electricity, cooling systems, and related infrastructure.
