Biofilm and microbial water quality were studied on 4 middle size full-scale distribution systems (DSs) (maximum
residence time from 23 to 160 h) deserving from 5,000 to 30,000 inhabitants. Three sampling
campaigns have been performed over a 1 year period on each DS. Biofilm was studied on cast
iron coupons incubated for more than one month in devices continuously fed with water from
the DS. Biofilm was quantified using the potential exoproteolytic analysis method (PEPA)
and by total (TDC) and heterotrophic plate counts (HPC) after detachment of the biofilm from
the support by sonication. Microbiological water quality (TDC and plate counts) was
estimated in parallel. In each DS, large variations of bacterial abundance in water were
observed: the ratio between minimum and maximum values in a given DS was on average
100 for the TDC and 1000 for the cultivable bacteria when all sampling points and campaigns
were considered. Very different biofilm densities in each DS were also recorded whatever the
method used to quantify the biofilm. The main factor controlling the biofilm was the
disinfectant residual: when all biofilm data were considered, microbial counts tended to
follow a logarithmic increase as the disinfectant residual decreased. A multiple linear
regression confirmed the impact of disinfectant residual on biofilm and also showed
significant effects on cultivable bacteria concentration in water. The maximum level of
biofilm measured during the present study by the PEPA method was higher than those
recently recorded in large full-scale DSs in Paris and Brussels. In order to study the impact of
water residence times, 4 groups (0-20h; 20-40h; 40-80h; 80-160h) were considered. A bellsharp
curve was observed in water for each microbial method whereas the bacterial
abundance in the biofilm tended to increase up to a maximum for the residence time class 80-
160h. The impact of temperature on microbial counts in water and biofilm was measurable
but quite limited. Includes 12 references, tables, figures.