Part 6 of 13

Water in the Desert: Where Would It Come From?

By Randy Salars · Source-review edition: September 7, 2026

An assurance such as “we will use very little water” should be the beginning of a conversation. For Grant County, the answer needs a source, a quantity, and an explanation of what the withdrawal would mean for existing users. A design that conserves water could make this proposal more attractive. We need enough detail to recognize that design and hold the operator to it.

The supplied air-permit notice does not provide a complete water budget or cooling design. It cannot settle claims of either minimal use or thousands of acre-feet annually. One acre-foot is approximately 326,000 gallons. As a hypothetical example, 100 acre-feet annually would equal about 32.6 million gallons, averaging roughly 89,000 gallons daily. That is a unit conversion, not an estimate for Site Layer 1, and an annual average would still leave peak demand unanswered.

Three different water numbers matter. Withdrawal is water taken from a source. Consumption is the portion evaporated or otherwise unavailable for immediate reuse. Recirculation is water passing through equipment again. Large recirculation volumes need not mean large new withdrawals. Low consumption alone does not tell us the pumping rate. The budget should show withdrawals, consumption, and discharges separately, including where discharged water goes. Source: USGS water-use definitions.

“Closed loop” describes only what happens within the specified loop. Water can circulate through computers repeatedly while a separate cooling tower evaporates water to release their heat. The Department of Energy illustrates that arrangement and explains the additional discharge used to control mineral buildup. Following the heat from the chips to its final release outdoors reveals which equipment actually needs replacement water. Source: DOE cooling guidance.

Low-water alternatives are credible. Microsoft's December 2024 announcement described a design using closed liquid loops and mechanical cooling to avoid evaporation for cooling. It acknowledged an electricity tradeoff and described planned pilots, with new sites expected online in late 2027. That announcement is evidence of a design pathway, rather than measured performance of an operating counterpart to Site Layer 1. Administrative uses such as restrooms still require water. Source: Microsoft's cooling-design announcement.

Here, the complete budget should cover computers, power equipment, treatment, maintenance refills, construction, and ordinary building uses. Fire protection needs its own storage and replenishment plan. Expected and maximum annual withdrawals should accompany peak-day requirements, seasonal differences, and full expansion. A convincing low-water proposal would state whether evaporative assistance is ever used, particularly during prolonged heat, and show the consequence if a preferred supply becomes unavailable.

Location matters at both ends of a pipe. Water could come from onsite wells, existing wells elsewhere, or a supplier. Reclaimed water or treated brackish groundwater could be options where suitable supplies and infrastructure exist. None is established as this project's chosen source in the reviewed documents. Buying water still requires identifying the supplying wells and their available capacity; treatment may introduce additional energy needs and waste streams.

One historical lead illustrates why records need careful reading. NMED lists Well 26B under the Phelps Dodge Hidalgo Smelter system. The system has status “I,” indicating inactive, while the individual well entry carries status “A.” Its recorded location is approximately 5.2 miles west-northwest of the proposed facility. Those entries do not establish whether it pumps today, who controls its rights, or whether this developer could use it. Source: NMED's Well 26B record.

The specific question is whether any former smelter wells, rights, or pipelines are being considered. If so, current ownership, rights, condition, and pumping information would matter. If not, the lead should be put aside. Nearby industrial history also does not establish that the proposed parcel itself is a former smelter site.

Silver City's water report identifies the Franks wells with the Gila–San Francisco basin and the Woodward and Gabby Hayes wells with the Mimbres basin. The reviewed documents do not demonstrate that this project would draw down those municipal supplies. They also do not identify its eventual supply wells. The next step is to locate those wells and evaluate their hydraulic connections. Claims of draining Silver City's water—or guarantees of complete separation—go beyond the evidence currently available. Source: Silver City's 2025 Water Quality Report.

A New Mexico Geological Society overview describes distinct regional basin systems, barriers, and possible connections, including generally northward groundwater flow through the Playas system. Published in 2000, it provides context rather than present-day predictions for proposed wells. Administrative basin boundaries and surface-drainage maps cannot identify those wells' physical connections either. Source: Kennedy, Hawley, and Johnson's groundwater overview.

Being downstream is not automatic protection from pumping. Withdrawals can lower groundwater levels around a well and change nearby flow. The reach depends on pumping and aquifer properties. USGS explains that sustained withdrawals can affect neighboring wells and exchanges with connected surface water. That explains the need for analysis; it does not establish a specific impact from this proposal. Source: USGS groundwater-depletion explanation.

For groundwater in declared basins, the Office of the State Engineer handles applications for new appropriations and changes to existing rights. The supporting rights and necessary approvals should be public alongside an assessment of physical supply. A legally authorized quantity and a well's dependable delivery under future conditions answer different questions. Source: New Mexico groundwater rules, 19.27.1 NMAC.

My confidence would increase with baseline measurements and a funded response if reality differs from the predictions. Nearby users should know who investigates falling water levels, when pumping must change, and who provides a justified remedy. A fine paid after a household loses its supply would offer little immediate help. Protection should begin before a problem becomes a crisis.

What a useful answer would include

The developer should supply a seasonal water balance, peak requirements, source locations, rights, and backup arrangements. A qualified groundwater specialist should assess cumulative drawdown with other pumping and drought assumptions. Regulators should identify required approvals, while an enforceable agreement assigns monitoring, warning thresholds, investigations, and funded remedies.

Questions worth asking

• What exact wells or suppliers would provide water, under which rights and approvals?
• What are the whole-facility annual and peak-day withdrawals and consumption at opening and full expansion, including hot-weather and drought conditions?
• What current evidence identifies the supplying aquifer and predicts cumulative effects on neighboring users?
• Are former industrial water assets involved, and what records establish their present condition, capacity, ownership, and available rights?
• What monitoring, pumping limits, complaint process, and funded remedies would protect existing users if the predictions prove wrong?

AI assisted most of the research and initial drafting under my direction. Sources are linked; corrections are welcome.