Experimental Investigation of Seawater Intrusion in Sewers by Means of Lagrangian Sensors

In the water sector, the problem of polluting source identification was mainly investigated regarding pressurized distribution networks respect to sewers. The same monitoring network may be used to investigate other issues that may impact negatively on sewer networks and wastewater treatment plants such as groundwater infiltration and ex-filtration. Groundwater infiltration acts as a diluent for sanitary flows increasing dry weather flows and decreasing organic load. Large infiltration flows may surcharge CSOs producing dry weather overflows having a negative polluting impact on receiving water bodies. In coastal areas, groundwater infiltration may have additional negative impacts due to saline concentrations that may produce rapid deterioration of mechanic parts (pumping stations, mixers, etc.) and may reduce the biological efficiency of waste water treatment plants (WWTP). In order to facilitate early detection and efficient containment of groundwater intrusions in sewers, the present work aims to develop an experimental campaign for data collection and analysis through a mobile type water quality sensor. The device used is Eawag`s SQUID that is a mobile sensor with high quality probes that can measure temperature, pH, Conductivity and Oxido-Reduction Potential (ORP). Furthermore, through a microSD card installed in the SQUID it is possible to transfer the data collected on a laptop. The sensor is floating, and it can be deployed in any manhole of the network following the flow until it is recovered, usually at a pumping station or at the WWTP. The case study chosen is a portion of the urban catchment of Mondello in Palermo (Italy). It is a combined sewer system with HDPE pipes (Manning’s roughness coefficient equal to 0.0125 s/m1/3) and the area has an extension of 2.7 km2; Fig. 2 shows the scheme of the drainage network which consist of 112 pipes and 113 nodes (including the outfall). The analysis was based on two phases: the first was based on modelling results and optimization aiming to design the minimum number of sensors and deployments to investigate the whole network; the second phase was based on experimental verification of the designed monitoring campaign. Finally, the areas of the network where infiltration most probably took part were highlighted and visually inspected to confirm the results of the indirect analysis.