Quantification of sewer leakage by a continuous tracer method.

Water authorities interested in the evaluation of the structural state of a sewer must quantify leakage to plan strategic intervention. However, the quantification of the exfiltration and the localisation of structural damage are challenging tasks that usually require expensive and time-consuming inspections. Herein, we report one of the first applications of the QUEST-C method to quantify the exfiltration in a continuously operating sewer by dosing two chemical tracers, sodium bromide (NaBr) and lithium chloride (LiCl). The method was applied at the catchment scale in a 14-year-old sewer in Rome, Italy. Preliminary laboratory tests, field measurements, and numerical simulations showed that reliable results require the QUEST-C method to be applied to sewers without lateral inflows, during periods of quasi-steady flow, and that the travel time of the NaBr tracer is minimised. Three sewer reaches were tested and the estimated exfiltration, as a fraction of the dry weather flow (DWF), increased from 0.128 in the agricultural area to 0.208 in the urban area. Although our estimates are at the lower end of the range given in the literature (0.01-0.56 DWF), the exfiltration was not negligible, and interventions should focus on the sewers in urban areas. This illustrates the capability of the QUEST-C method to guide strategic intervention at low cost and without an interruption of sewer operation. However, careful interpretation of the results is recommended for sewers with many lateral inflows, where leakage may be overestimated.

Quantification of sewer system infiltration using delta(18)O hydrograph separation.

The infiltration of parasitical water into two sewer systems in Rome (Italy) was quantified during a dry weather period. Infiltration was estimated using the hydrograph separation method with two water components and delta(18)O as a conservative tracer. The two water components were groundwater, the possible source of parasitical water within the sewer, and drinking water discharged into the sewer system. This method was applied at an urban catchment scale in order to test the effective water-tightness of two different sewer networks. The sampling strategy was based on an uncertainty analysis and the errors have been propagated using Monte Carlo random sampling. Our field applications showed that the method can be applied easily and quickly, but the error in the estimated infiltration rate can be up to 20%. The estimated infiltration into the recent sewer in Torraccia is 14% and can be considered negligible given the precision of the method, while the old sewer in Infernetto has an estimated infiltration of 50%.