The index is built from filings utilities submit to their own state regulator. Nothing here is a survey, an estimate of an estimate, or a proprietary normalization. Every figure is either as-filed or computed from as-filed values, and every page says which.
Sources
Texas. Water loss audits filed with the Texas Water Development Board. A retail public water utility with more than 3,300 connections, or with an active financial obligation to TWDB, files annually; below that threshold, every five years. Since January 2025, utilities with an active obligation or applying for financial assistance must also have the audit validated, so an unvalidated recent audit is itself a flag we display.
Massachusetts. Unaccounted-for water and residential gallons per capita per day as published by MassDEP from Annual Statistical Reports, required of municipal suppliers averaging 100,000 gallons a day or more.
Presentation scale is detected per file and normalized on ingest. Every volume in the index is million gallons. This sounds pedantic and it is the single most common transcription error in regulator exports.
The closure check
A water audit is an identity, not an opinion:
system input = billed authorized + unbilled authorized + apparent loss + real loss
We evaluate it on every filing that publishes all five terms, with a tolerance of 0.5 percent of system input to absorb rounding in the filed figures. Three verdicts, never two:
- closed — the components sum to reported input within tolerance.
- failed — they do not, and we state the signed residual in million gallons and as a share of input. That volume is either mis-stated or sitting in a category the audit does not name.
- insufficient — the filing does not publish enough terms to evaluate the identity at all.
The third verdict is the reason this index is worth reading. A check that cannot fail is not a check, and a system that reports an unevaluable filing as passing is worse than one that reports nothing. Most filings land here, because published summaries commonly carry totals and percentages rather than the components a balance is made of. We name which terms were missing.
Do not aggregate insufficient with closed. In the bulk data they are separate values for exactly this reason.
The recoverable figure
Apparent loss is valued at the retail rate, because it is water that reached a customer and was never invoiced — revenue, recoverable inside a budget year, no capital work. Real loss is valued at variable production cost, because stopping a leak saves treatment and pumping, not retail price.
Where a filing carries no retail rate, a state median is substituted and the figure is marked modeled everywhere it appears, including in the CSV. Where a filing publishes total loss without splitting apparent from real, the whole volume is valued at variable cost and labelled as a deliberate floor — the revenue portion cannot be estimated without the split.
A recoverable figure is an order of magnitude, not a collections target. It says whether the problem is worth an afternoon or a quarter. Which accounts, which zones, and which meters requires your billing history, your meter reads, and your production data reconciled against each other, and no public filing contains that.
What we do not do
We do not rank by loss percentage. A small percentage on a large system is worth more than a large percentage on a small one, and percentage rankings punish small rural systems for arithmetic.
We do not recompute a utility’s filed figures to make them close. If a filing does not balance, the page says so and shows the residual. Adjusting inputs until an identity holds is how a check becomes decoration.
We do not publish a figure we cannot source. An absent value renders as an em dash, never as a zero.
Corrections
If a number here is wrong, it is wrong in the filing or wrong in our reading of it. Both are worth knowing. Email [email protected] with the system and the year; we correct the page and note what changed and when.