Infiltration in sewer systems: comparison of measurement methods

For 20 years, several methods for the estimation of infiltration have been developed in various countries. These conventional methods are subject to considerable uncertainties due to their underlying assumptions and general principles which are not estimated. Two extended comparative studies of the conventional methods have been made in order to assess the variability in infiltration estimations and associated uncertainties according to the method used. The choice of method is not critical when the objective of a sewer diagnostic study is to define the spatial distribution of infiltration contributions at subcatchment scale. Nevertheless, the methods based on the analysis of the minimum night flow that are generally applied during one dry weather day should be applied during 8 to 10 dry weather days in order to provide estimations with acceptable uncertainties.