As the rapid expansion of artificial intelligence infrastructure continues to strain municipal utilities, data centers have drawn mounting criticism from environmental advocates over their massive cooling and water consumption requirements. In response to these concerns, an AI infrastructure developer is attempting an unconventional engineering strategy in central Utah aimed at manufacturing fresh water directly on-site rather than depleting regional water resources.
Creekstone Energy recently announced a patent-pending system called Gigawater, designed to capture moisture produced during the combustion of fossil fuels. The proprietary process serves as the core technology for the company's Project Oasis, situated at its Delta Gigasite facility currently undergoing construction in Millard County. Under the proposed model, the operation will rely on natural gas piped in from Wyoming to simultaneously power operations and yield a usable water supply.
According to Creekstone founder and chief executive Ray Conley, the concept relies on fundamental chemical principles that drivers observe every day. Conley explained that when motor vehicles accelerate after idling at a traffic signal, liquid frequently drips from their tailpipes because moisture condenses within the exhaust system as a byproduct of fuel combustion.
The underlying chemistry is well established: when methane combusts, it breaks down into carbon dioxide and water vapor. However, applying this basic reaction at the scale required for utility-grade industrial operations poses significant engineering and environmental questions. The primary challenge rests on whether the system can function efficiently at a commercial scale without creating prohibitive ecological drawbacks.
To capture the vapor on a massive footprint, Project Oasis engineers plan to divert high-volume exhaust subterraneanly. The hot exhaust gas will be channeled underground through large beds of crushed rock positioned above an impermeable rubber liner. As the gaseous exhaust cools against the rock subterranean matrix, moisture will condense and collect along the protective rubber barrier for retrieval.
A crucial selling point emphasized by company leadership is self-sufficiency. Conley stated that the entire underground condensation apparatus functions without drawing external electricity, allowing the plant to synthesize a standalone water supply rather than siphoning water from nearby communities or fragile desert aquifers.
Despite the bold claims, the process remains purely conceptual at this stage. To transition the theoretical mechanics into practical application, Creekstone has partnered with researchers at the University of Utah to independently vet the engineering assumptions underpinning the Gigawater blueprint.
The collaborative research roadmap is structured across two distinct phases spanning roughly two years. The initial phase, which could take up to a full year, focuses on rigorous computational modeling, verifying baseline thermodynamics, and evaluating the chemical reactions. If those findings prove viable, the second phase will involve designing and constructing an experimental pilot facility during the following twelve months.
Preliminary engineering estimates released by Creekstone project that the Millard County installation could generate roughly 4,000 acre-feet of recovered water annually. Company executives claim this volume will easily surpass the operational water needs of the Delta Gigasite, allowing the complex to operate as a water-positive asset rather than a net consumer.
Developers envision several end uses for the synthesized water once it is captured. According to company releases, the harvested fluid could be routed toward agricultural irrigation, channeled into geothermal energy cycles, or treated through advanced filtration systems to serve as a potable drinking source for nearby areas.
Nevertheless, water conservation experts suggest keeping the scale of the output in perspective. While 4,000 acre-feet represents a substantial volume for a single commercial building, it constitutes a microscopic fraction when measured against regional demand. By comparison, statewide water metrics compiled by Utah State University recorded Utah's total statewide water supply at roughly 11.7 million acre-feet in 2024.
Logan Mitchell, a climate scientist and energy analyst with Utah Clean Energy, noted that while generating new water is preferable to depleting local wells, the actual impact on Utah’s overarching water security remains minimal. Mitchell pointed out that yielding 4,000 acre-feet amounts to approximately 1% of the state's agricultural water usage, characterizing the contribution as barely above zero in terms of transforming the broader hydrological landscape.
Mitchell framed the Gigawater initiative as merely one fragment of a much larger, highly intricate data center puzzle. While addressing water utilization solves a single operational concern, he stressed that a complete accounting of data center impacts requires evaluating multiple overlapping factors, including greenhouse gas emissions, grid integration, and broader state regulatory policies.
The core dilemma surrounding the project involves balancing water generation against the fossil-fuel emissions created by burning methane from Wyoming. Creekstone maintains that the data center sector does not need to force local communities to pick between regional economic investment and responsible environmental management. Whether Project Oasis can validate those claims will depend on the results of the university's forthcoming technical trials.
(0)