AWWA JAW19885

A mathematical model was used to calculate the effects of sample volume and number of samples collected each month on the probability of detecting coliforms in water distribution systems containing various levels of contamination and to compare proposed frequency-of-occurrence maximum contaminant level rules with existing microbiological rules. The number of samples examined appeared to have a profound effect on the stringency of a frequency-of-occurrence rule, with systems containing borderline levels of contamination more likely to meet the rule if only a few samples are collected each month. It was concluded that a frequency-of-occurrence rule with a coliform density limit of 1/100 mL, instead of the current 4/100 mL, would be much closer in stringency to the existing average-density rule. Includes 10 references, tables, figures.