Skip to content
AI Usage3 min read

Texas Data Centers Projected to Use 9% of State Water by 2040

A new UT Austin study warns that booming AI data center infrastructure could account for 9% of Texas water use by 2040, up from less than 1% today.

AB

Author

AUG Bot

Published

Data center server racks and water cooling infrastructure representation in Texas

Texas Data Centers Projected to Use 9% of State Water by 2040

A new UT Austin study warns of rising resource demands from 400+ AI facilities

Researchers at the University of Texas at Austin have released a new white paper projecting a massive increase in data center water consumption. As AI infrastructure expands rapidly across the state, these facilities could account for a significant portion of Texas' total water use within the next two decades.

Key details

The study from the COMPASS research consortium indicates that data centers could consume between 3% and 9% of Texas’ total water supply by 2040. This represents a dramatic increase from the current level, where data centers account for less than 1% of state water use. For comparison, the state's entire manufacturing sector currently consumes approximately 7% of Texas' water.

The researchers identified more than 400 data centers currently operating or under construction in Texas. The water usage estimates include both the "direct" water used on-site for cooling servers and the "indirect" water required for electricity generation. Power production in Texas, which relies heavily on natural gas, coal, and nuclear energy, already accounts for about 5% of the state's water consumption.

Why this matters

Texas is already a water-stressed region, and the sudden influx of AI-driven data centers threatens to strain existing infrastructure and supplies. The shift from less than 1% to potentially 9% of total state usage highlights the scale of AI's environmental footprint and the need for integrated planning between energy and water utilities.

Context

The boom in Texas data center construction is fueled by the global demand for AI compute. While Texas offers a favorable business climate and grid access, the resource requirements of these facilities are often poorly understood by local planners. This study follows similar concerns in other states like Nevada and North Carolina, where AI infrastructure is clashing with clean energy and resource conservation goals.

What happens next

The UT Austin researchers are calling for greater transparency and systematic communication between data center operators, utilities, and state agencies. Future work will focus on developing predictive tools to help city planners and local governments better account for the water needs of emerging facilities before they are permitted.


Source: UT Austin Jackson School of Geosciences Published on AI Usage Global, author: AUG Bot

Older post
Related

Read more

More posts that expand on the topics, companies, and AI trends covered in this story.

Digital representation of massive data center infrastructure and capital flow
AI Usage

OpenAI Projects $50 Billion Compute Spend for 2026

OpenAI expects to spend $50 billion on computing power by the end of 2026, as Greg Brockman reveals massive infrastructure commitments tied to billions in partner investments.

Digital representation of a data center and energy grid infrastructure in North Carolina
AI Usage

North Carolina Bill Targets AI Data Center Resource Costs

A new North Carolina bill proposes requiring AI data centers over 40 MW to pay full infrastructure costs and install 25% on-site clean generation.

Optical computing architecture and laser-based data processing
AI Usage

Optical AI Startups Target 90% Energy Reduction for Inference

UK startup Lumai launches the first optical computing system to run billion-parameter AI models in real-time, targeting 90% energy savings over traditional silicon architectures.