Circle of Blue reports that communities across the country are facing a "federal water infrastructure funding cliff" after the expiration of a bipartisan, multibillion-dollar cash infusion that Congress approved during the Biden administration. The article, by Abigail Mihaly, asks what comes next. For most systems, the honest answer depends on what each project was tied to. Read the Circle of Blue report on the expiring water funding before rebuilding any capital plan around it.
The usual mistake is to treat the program as a single line. A board hears that a program ended and asks how much the utility lost. That question cannot be answered from a program headline. The answer lives in each agreement, each reimbursement request, and each pending application, and those documents do not all close on the same day.
Sort Every Open Item Into One of Four Bins
Pull the file for each project that touched the expired program and place it in one bin.
Committed. A signed agreement exists, and work or purchase orders have been issued. The agreement's own end date and closeout terms govern, not the public summary of the program. Keep the executed copy in the asset file, with the award amount and the schedule for draws.
Requested but undrawn. The utility applied or was awarded, but has not yet submitted reimbursement requests for costs already incurred. This bin carries the most risk. The money may still be owed, but only if the paperwork moves before the agreement closes. Confirm the agency's closeout rules in writing, not from memory or from a colleague's summary.
Planned but unapplied. The project appeared in a capital plan as expected grant money. Its funding line now needs a replacement source, a deferral, or a reduced scope. The board should see that change as a decision, not as a quiet shift in a spreadsheet.
Unknown. Records that cannot say whether a project was ever submitted. Mark these as unknown. Do not fill them with an estimate, because an estimate in this bin will later be read as a fact.
Keep the Cost Trail Separate From the Forecast
A common error is to carry a planned grant into the rate model as though it had been received. Costs paid from local cash and later reimbursed create a receivable, and that receivable should be tracked on its own line. If it is not, a late reimbursement reappears as a revenue spike in a later budget year and distorts the comparison with prior years.
The rate model should carry the local-cash case as its base. A grant can appear as an upside only after a written award amount and a draw schedule exist. This is slower to present, and it will look conservative next to a plan that assumed the money. That conservatism is the point. A plan that depends on an unsigned award fails at the moment the award does not arrive, which is when the board most needs it to hold.
Weigh the Cost of Waiting
Deferral is often presented as the free option. It is not. A project pushed back to wait for replacement money carries its own costs: construction prices may move, the asset continues to age, and the risk of an unplanned repair on a deferred main or pump rises. Record the deferral as a decision with a stated reason, a review date, and the condition data that will trigger a reconsideration. If the condition data is thin, say so in the record.
Where Small Systems Carry the Paperwork
Systems with a small operating staff and no dedicated grants function often hold the most fragile files. Application packets, draw requests, and certifications may live in one person's email. When that person leaves, the record goes with them. The defensible steps are modest. Name a file owner, store the executed agreements and the draw log in one shared location, and note the agency contact who answers closeout questions. Whether a state technical assistance provider can help with this work is a question for the system's own agency, and the answer varies by state, so confirm it rather than assume it.
Write Around the Gap
When the funding picture is incomplete, the board document should say so in plain terms. Name the projects, the bin each occupies, the decision date, and the items still marked unknown. A forecast that shows a clean line where the file shows a question will be wrong in a way that is hard to notice until a contractor is waiting or a reimbursement window has closed. Uncertainty stated in the record is useful. Uncertainty hidden inside a smooth projection is not.