[Bacterial regrowth in drinking water. I. The upgrading of drinking water].

Seven dead-end water pipes were installed after each treatment step in a drinking water plant. During a period of 7 weeks the bacterial load of freshwater and stagnating water was investigated with different methods. A modified surface spread plate count (Plate Count Agar, 10-fold diluted, 14 days incubation at 20 degrees C) proved to be more effective than the traditional pour plate method, because it gave consistently higher colony counts and had a lower level of detection (0.001 CFU ml-1). The enumerating of electron-transport-system positive bacteria yielded higher numbers than the colony count methods, but is not recommended when recently oxidized water samples are to be investigated. Highest cell counts were attained when using epifluorescence microscopic counting, yet bacterial regrowth could not be monitored thus. The tendency of bacterial regrowth was highest in freshly ozonized water. In stagnating lake water no regrowth occurred after 1 and 3 weeks because of the balance of bacteria and their predators.