Data center water use is a local infrastructure question as well as a national AI policy issue. This research guide brings together Gabriel Collins’s work on cooling, electricity supply, water demand, and public disclosure. The linked articles and presentations provide the underlying evidence and original publication details.
Start with the measurement questions
Comparisons should identify the facility, reporting period, water source, and cooling method, and distinguish water withdrawals from consumption. Regional water availability and the electricity needed to supply cooling matter alongside a campus’s annual total. These are central questions in the proposed ComputeFocus disclosure framework.
Core research and presentations
- The Data Center Industry Needs Its Own FracFocus-Style Transparency Initiative — Introduces ComputeFocus: regular facility-level electricity and water disclosures, cooling type, and water-source information to support informed public debate.
- Powering and Cooling the AI Boom in the United States — Presentation on the physical requirements and policy trade-offs of supplying AI infrastructure with power and water.
- Are Data Centers Just Another Industrial Asset? — Compares the datacenter buildout with earlier industrial expansions and examines Texas siting, land, utility, and public-acceptance questions.
- Permian Basin: The Next AI Data Center Frontier — Examines the regional electricity resources that could support new compute capacity and the infrastructure choices involved.
Connect facility demand to Texas water planning
Comparing Water and Electricity Load Growth in Texas provides a broader demand baseline. For the potential treatment and reuse of oilfield water, see the Texas Produced Water research guide. That research explores options; it does not establish that treated produced water is suitable for every cooling application or location.
Browse the Datacenters category for additional articles, or the complete Research Master Index.