A laboratory result for a trace compound is not a measurement of the water alone. It is a measurement taken against a floor the laboratory set in advance, and that floor moves with the method, the instrument, the calibration, and the quality control run on the day. Circle of Blue's report on a new study, dated October 23, 2026, turns on this point. Its summary begins with a mechanism every reader in this trade should be able to state: "When PFAS concentrations in drinking water fall below laboratory reporting limits" the reported figure no longer describes the water in the simple way a reader assumes.
What the floor is made of
A reporting limit is usually not a single physical boundary. In general practice it sits above the level an instrument can distinguish from background noise, because a number reported at that edge would carry too much uncertainty to use with confidence. The laboratory derives it from the lowest calibration standard it runs, from blank performance, and from the recovery and precision data it has demonstrated for that analyte in that sample matrix. Between the detection level and the reporting level, many programs allow a result to be flagged as an estimate rather than a clean quantity. Below the detection level, the report says not detected. That statement describes the method on that day. It is not a claim that the compound was absent.
Why an undercount claim needs its own record
When a study argues that concentrations may be underestimated, the argument usually runs through the floor. If many results sit below the reporting limit, a reported total can understate what the sample contained. Whether that holds for any particular dataset is an empirical question. The public summary does not give the full analytical design, and this article does not claim to know it. The reader's task is narrower: determine whether the laboratory records behind a figure can support the comparison being made with it.
What to log before the number is cited
Before a PFAS result moves into a board packet, a planning model, or a press response, the record should show, for each analyte, the following: the reporting limit in force on the sample date, not an older figure from a brochure; the lowest calibration standard and the date it was run; method blank results for the batch; recovery and duplicate precision for the batch; the qualifier codes used on the report and what each one means in the laboratory's own legend; and whether the reporting limit changed between the sampling year and the reporting year. A utility that tracks these fields once can answer a challenge in an hour. A utility that does not may spend weeks reconstructing them from email threads and archived reports.
Keep zeros and non-detects apart
Spreadsheets treat blank cells and zeros without any memory of what produced them. Averages built from those cells can mix a true zero, a non-detect at a stated limit, and an estimated value below the limit, then present the mean as if each contributed equally. Store the qualifier beside the value, carry the reporting limit as its own column, and set the averaging rule before the data arrives. The rule should state what happens to non-detects in a total, and that choice belongs in the method note, not in a decision made after the chart is drawn.
What this does not settle
A headline saying contamination may be underestimated does not establish how much, where, or in which systems. It is a prompt to examine the measurement chain. A reader who takes the claim seriously should pull the study's methods section, compare its reporting conventions with the laboratory reports already on file, and state in writing which findings apply to their own sources and which do not. Until that comparison exists, the study's standing for planning purposes remains an open question, and a complete record is the only way to make that status clear to a board, a regulator, or the public.