PlainWater

Editorial & Corrections Policy

PlainWater turns the EPA's drinking-water compliance data into pages a resident, journalist, or local official can actually read. This page explains how those pages are produced, the standards we hold the data to, and exactly how to get an error fixed. We document the real process — including what is automated — rather than claim a level of hand-review we do not perform.

How Our Pages Are Produced

Every data page on PlainWater is generated from official EPA datasets through a documented pipeline, not written by hand and not generated from unsourced text:

  1. Acquisition — We download the complete EPA Safe Drinking Water Information System (SDWIS) bulk export from the ECHO program, together with the EPA UCMR5 PFAS occurrence files.
  2. Validation — Records are checked for structural integrity, unit consistency (counts, populations, concentrations), and plausible ranges before anything is published. Figures that fail a sanity check are held back, not guessed at.
  3. Derivation — We compute the added-value metrics the raw files do not contain — per-system and per-state violation rates, health-based shares, PFAS detection rates, and county/state aggregates — using methods documented on our methodology page.
  4. Presentation — Each system, county, state, and contaminant page renders directly from the validated database at request time, so what you see reflects the current published dataset, with its vintage shown in context.
  5. Editorial review — Templates, methodology, rankings logic, and guides are reviewed and maintained by the PlainWater editorial team. Individual entity pages are produced programmatically from the reviewed templates; we do not claim to manually review all 143,000+ water-system pages.

Sourcing Standards

  • Every figure traces to a named, public, primary source — principally the EPA SDWIS/ECHO downloads and the UCMR5 monitoring program.
  • We present reported figures as the source publishes them. Where we derive a metric (a rate, a share, a percentile), we say so and explain the calculation.
  • We do not republish unsourced statistics, fabricated figures, or scraped content. Where EPA has no data for a system (for example, a system not yet sampled for PFAS), we show the gap rather than fill it.
  • We accept no payment, sponsorship, or promoted placement from any water utility or company. Coverage and ordering are determined by the data alone.

Update Cadence

EPA refreshes SDWIS compliance data on a quarterly cycle and publishes UCMR5 PFAS results in periodic data releases. We monitor the federal release calendar and rebuild our database from each new release. Every data element carries the vintage of its source; violation counts are cumulative records held by EPA, not point-in-time safety readings.

Corrections Process

We want every number on PlainWater to match its EPA source. If you find a figure that looks wrong, tell us and we will investigate. Email our team with the page URL and the value in question. If our figure diverges from the current EPA release, we correct it on the next database rebuild; if it matches the source, we explain the discrepancy (often a vintage or denominator difference). Where EPA itself has since corrected a record, we note that the change originated upstream.