Estimating Mean Coliform Densities of Water Distribution Systems

Current drinking water regulations set a limit on the average number of coliforms allowed in samples that are collected during distribution system monitoring. This article presents evidence that the average number of coliforms in monitoring samples is a poor estimate of the mean coliform density of the distribution system. The evidence includes variance in coliform counts, data truncation caused by limitations of the examination methods, changes in coliform densities during transport of the samples to the laboratory, and false results. A regulation based on the fraction of monitoring samples with coliform densities greater than a preselected limit may be a feasible alternative to the current microbiological maximum contaminant level.