A statistician's view of the U.S. primary drinking water regulation on coliform contamination

Abstract

The new U.S. regulation for monitoring possible coliform contamination of drinking water is based on a presenceabsence assay, in contrast to the previous regulation that was based on coliform density assays. A water is said to be in compliance if the fraction of 100-mL sample volumes that contain coliforms is judged to be less than 0.05. The rule is designed to allow only a small chance that the water will pass the test when, in fact, the fraction is greater than the so-called “protection reliability standard” of 0.10. Statistical analysis of the compliance criterion shows that the false pass rate is higher than intended and/or the protection reliability standard is higher than intended, and the false fail rate can be as high as 0.5. For purposes of comparison with the previous regulation, the new criteria are converted into corresponding coliform density values.