Fate of Escherichia coli experimentally injected in a drinking water distribution pilot system

Abstract Detection of coliforms and E. coli in drinking water distribution systems can be explained by accidental contaminations and/or growth of these bacteria in the network. This last point still remaining debatable, an experiment was carried out with two E. coli strains ( E. coli O126:B16 and one isolated from a drinking water distribution system) separately injected in a drinking water distribution pilot plant continuously fed with drinking water. The work aimed to study (i) the partition of the injected bacteria between the water phase and the indigenous biofilm phase and (ii) the kinetic of disappearance or growth of these bacteria. Depending on the strain, 1–50% of injected bacteria adsorbed within a few hours to the indigenous bacterial biofilm. During the first 5–7 days, both E. coli populations behaved similarly: the total number of injected bacteria decreased more quickly than the theoretical washout (death, lysis, predation, ...). Then the number of these bacteria slowly increased and at day 9–12, the number of recoverable injected bacteria became higher than the predictable number calculated from the theoretical washout. This result clearly proves that both E. coli strains are able to grow at 20°C in the absence of residual chlorine in a distribution network system largely colonized with an autochthonous population. However, colonization of the network by E. coli strains was only partial and transient.