USGS Just Made Data Center Water Use the Subject of Its National Water Research Competition. Four Awards, $462,500, November 12 Deadline.

The Water Resources Research Act program is one of the quieter corners of federal science funding — a 1964-vintage federal-state partnership that funds 54 Water Resources Research Institutes, one in each of the 50 states plus the District of Columbia, Puerto Rico, the US Virgin Islands, and Guam.

It does not usually make news. This cycle it should.

The FY2026 104(g) General national competitive solicitation — opportunity number G27AS00164, Assistance Listing 15.805, posted September 30, 2026 and closing November 12, 2026 at 5:00 PM Eastern — has pointed its research priority squarely at the water consumption of data centers and other water-intensive industries that draw on public supply systems.

That is a small pot of money attached to one of the most consequential measurement problems in American infrastructure policy right now.

The numbers, precisely

Do the multiplication before you draft a budget: four awards at the $120,000 cap would total $480,000, which is $17,500 more than the posted program funding. The implied average award is about $115,625.

That is not an accounting error on anyone's part — agencies routinely post a ceiling and a total that do not multiply cleanly, and the expected award count is an estimate. But it tells you something practical: do not assume the ceiling. A proposal budgeted at exactly $120,000 with no articulated scaling path is asking the program officer to either fund fewer projects or negotiate your scope down after selection. A budget that identifies which year-three tasks are severable, and what the project still delivers without them, is a materially easier award to make.

The match compounds this. A $120,000 request requires $120,000 in non-federal funds, making the real project cost $240,000. On a program with roughly $462,500 in federal money, the four winning institutes will collectively bring nearly half a million dollars of their own to the table.

Only 54 organizations in the country can apply

This is the eligibility fact that most reporting on the program gets wrong, and it is the one that determines whether this opportunity is relevant to you at all.

Applications must come from a Water Resources Research Institute established under the Water Resources Research Act of 1984 and located at a university in one of the eligible states and territories. An individual faculty member cannot submit to grants.gov directly. A nonprofit cannot apply. A consulting firm cannot apply. A utility cannot apply.

If you are a researcher who wants to work on this, your path runs through your state's institute, and the operative deadline is almost certainly not November 12. Several institutes run their own internal pre-proposal calls to select which projects they will forward — the institutes at UMass Amherst, the University of Vermont, Texas A&M, and the University of Kentucky all publish internal 104(g) processes. Those internal deadlines sit weeks ahead of the federal one, and in a cycle where the NOFO posted on September 30 for a November 12 close, several have likely already closed or are closing now.

Call your institute director this week. With 29 days to the federal deadline on a solicitation that requires institutional submission plus a documented 1:1 match, the realistic question is not whether you can write a proposal but whether your institute still has room to carry one.

Note also the subject-matter exclusions: the program does not fund health-effects research involving human subjects, nor oceanography — except estuarine research, which remains in scope. That estuarine carve-out matters for anyone proposing work on coastal public supply systems, saltwater intrusion, or discharge from cooling systems into tidal waters.

Why the research priority is the real story

The NOFO's framing is worth reading closely. It asks for work that advances observational and analytical approaches capable of reducing uncertainty and strengthening confidence in publicly available information about water use by data centers and other water-intensive industries served by public supply.

Unpack that and you get a precise statement of what USGS believes is broken.

Data center water consumption is genuinely, structurally opaque in the United States, for reasons that are methodological rather than conspiratorial:

• Public supply hides it. A facility that buys municipal water does not report a withdrawal — its consumption disappears into a utility's aggregate commercial or industrial account. The national water-use data system was built around tracking withdrawals by source and sector, and a hyperscale campus on city water fits that architecture badly.
• Withdrawal and consumption diverge sharply. Evaporative cooling consumes a large fraction of what it takes in. Once-through and closed-loop systems have radically different consumptive profiles for the same nameplate capacity. Reporting one number without specifying which one it is renders comparison meaningless.
• Self-reported figures carry trade-secret protection. Facility-level operational data is frequently withheld as competitively sensitive, and water-use intensity is treated as proprietary.
• Siting outpaces measurement. Capacity is being announced faster than any monitoring methodology can be standardized around it.

The NOFO does not ask for an advocacy position on any of this. It asks for measurement methodology — observational and analytical approaches that shrink the error bars on numbers the public can already see. That distinction should govern how a competitive proposal is framed. A project that estimates how much water the AI buildout will consume in your state is an impact study. A project that produces a defensible method for distinguishing data-center consumptive use from other commercial use within a public supply system, and validates it against a real utility's records, is responsive to what was actually solicited.

The strongest proposals here will almost certainly involve a named utility partner with a data-sharing arrangement already in hand. That partner can plausibly serve double duty: as the source of the records that make the method validatable, and — if the contribution is structured and valued correctly under the applicable cost principles — as a contributor toward the non-federal match. Confirm the match treatment with your institute's sponsored programs office before you build it into a budget; third-party in-kind contributions have specific documentation and valuation requirements, and a match that does not survive audit is worse than a smaller one that does.

The timeline oddity, and the companion NOFOs to watch

Look again at the dates. This is the FY2026 solicitation, posted on September 30, 2026 — the last day of fiscal year 2026 — closing in November 2026, with an anticipated start date of May 1, 2027.

That is roughly seven months between the close of the competition and the start of the work, and it is a pattern worth recognizing across federal science right now. FY2026 money is being competed late and obligated into FY2027 performance periods. NSF closed FY2026 carrying over more than a billion dollars in unobligated research funds (our breakdown), and the compressed-NOFO, delayed-start shape here is a smaller instance of the same fiscal-year dynamics.

For an applicant, the practical consequence is bridge planning. If your proposal depends on a graduate student, that student needs support from roughly November through the following May regardless of whether you win. Build the staffing plan around a May 2027 start, not around the award notice.

Two companion solicitations are also in the pipeline. USGS has indicated that the FY2026 104(b) Base and FY2026 104(g) PFAS NOFOs have moved to their final level of approval, but has not published timelines. The 104(b) base grants are the formula-style allocations that flow to every institute; the PFAS competition is a separate national competitive track. If your work sits closer to contaminant transport than to industrial water demand, the PFAS NOFO — not this one — is the opportunity to watch, and the absence of a posted date means checking the program page rather than waiting for an announcement.

The short version

If you are at a Water Resources Research Institute: this is a four-award competition closing November 12 with an implied average award near $115,625, a mandatory dollar-for-dollar match, and a research priority that rewards measurement methodology over impact estimation. Secure a utility data partner and scope a budget that degrades gracefully below the cap.

If you are a researcher who is not at an institute: your deadline is your institute's internal pre-proposal date, and it may already have passed. Call this week.

If you are watching federal water policy generally: a national competitive program just made the water footprint of the AI buildout its central research question. The $462,500 is small. The signal is not.