Author(s): Sambito Mariacrocetta; Stefania Piazza; Gabriele Freni

Linked Author(s): Stefania Piazza, Gabriele Freni

Keywords: Bayesian approach; Brackish water infiltration; Lagrangian sensors; Optimal deploy; Urban drainage systems

Abstract: In the water sector, the problem of polluting source identification was mainly investigated regarding pressurized distribution networks respect to sewers systems as a drinking water contamination event determines an immediate alarm for public health, however, when an polluting discharge occurs in the sewer, various problems arise; in facts, the quality of wastewater acts both on the proper functioning of the sewer system, on the wastewater treatment plant (WWTP) and on the receiving water body in case of combined sewer overflow (CSO) activation. Therefore, the collection and analysis of real data is indispensable to control wastewater quality and identifying the origin of the pollution and the implementation of a chemical monitoring network is necessary to promptly detect and contain the event of contamination. In this study, it is proposed a Bayesian probabilistic approach to decide the optimal deploy of water quality sensors in urban drainage systems to detect the brackish water infiltration and reduce their costs assuming the deployment of Lagrangian sensors. Brackish water most often contains sodium chloride, sodium sulphate and sometimes magnesium chloride which cause damage to both the metal parts of the sewage system and the biological treatment unit of the wastewater treatment plant. Lagrangian sensors are monitoring platforms that do not require a stable installation, but they are deployed in the current and transported by the water. Sensors collect and transmit data in real time, and they are recovered at the end of pipe. The cost of the sensors is usually low as the risk of losing or damaging it is high. The proposed methodology incorporates several sources of information, including network topology, flows and non-formal “grey” information about the possible locations of contamination sources. For the solution of this problem, two main components are required: a calibrated model for hydraulic and water quality simulations in sewer systems (EPA SWMM model) and a Bayesian solver for likelihoods estimation and probability update. By using the Bayesian approach, new information, coming from the analysis, is incorporated in the approach allowing the operator to gain insight on the system once new contamination events are detected and identified. In this way, the approach is suitable for solving problems in which data are initially piecemeal and the operator plans to improve the monitoring strategy. The methodology is applied on the literature scheme from the Storm Water Management Model (SWMM) manual, Example 8, it is a combined sewer network consisting of 31 nodes, 29 pipes 7 sub-catchments, a pump, 2 outfalls (river and WWTP) and serving an area of 0.12 km2 (Gironàs et al., 2009).

DOI: https://doi.org/10.3850/IAHR-39WC252171192022432

Year: 2022