Particle.news

Study Finds U.S. Drinking Water PFAS Levels May Be About Twice as High as Reported

Models show standard reporting treats measurements below laboratory minimum reporting limits as zero, which can halve estimated average PFAS concentrations.

Overview

  • The peer-reviewed paper published Oct. 23 in Cell Reports Sustainability used censoring-aware statistical models on the EPA’s UCMR5 dataset to analyze about 1.86 million PFAS measurement records from nearly 10,000 community water systems.
  • The team found roughly 98% of those individual PFAS results were reported as below the laboratory minimum reporting limit (MRL), a common lab threshold that does not mean the chemical is absent.
  • When below-limit results are counted as zero, conventional summaries captured about 44% of the model’s estimated average for 25 PFAS; for the six PFAS covered by the EPA rule the conventional average was ~2.6 ng/L versus a modeled ~5.2 ng/L.
  • The authors emphasize these numbers are model-derived estimates rather than new measurements of concentrations below MRLs and do not by themselves show water supplies are unsafe, and they urge agencies to report non-detects with context and uncertainty.
  • The study calls for clearer public reporting, lower laboratory reporting limits, and targeted follow-up testing to reduce uncertainty and to improve exposure estimates, regulatory decisions, and communication with affected communities.