Author(s): Notaro Vincenza; Freni Gabriele; La Loggia Goffredo; Oliveri Elisa
Linked Author(s): Gabriele Freni
Keywords: Performance; Statistical analysis; PI; Reliability; Sewer system; GLUE
Abstract: In the last decades, many definitions have been proposed to explain the concept of performance of hydraulic systems and many methodologies have been defined in order to quantify performance in the most objective and general way (Notaro et al, 2005). This aim is usually pursued by the definition of Performance Indicators (PIs) synthetically representing the behaviour of the system. As all modelling procedures, however, performance analysis is a conceptualized representation of the real drainage system behaviour and of the physical processes occurring in it with the aim of evaluating the present and the future reliability system. Therefore, its predictions will be affected by a degree of inherent uncertainty that can be comparable with model parametric one (Beven & Binley, 1992; Gupta et al., 2005). According to the previous considerations, a new statistical approach to the performance analysis has been proposed (Notaro et al. 2005, Freni et al. 2005), aimed to reduce the performance analysis predictions uncertainties and to increase the robustness and objectivity of the performance analysis procedure by reducing the impact of subjective operator choices during the performance assessment procedure. The statistical approach has been compared to the classical literature one by the mean of the Generalised Likelihood Uncertainty Estimation (GLUE) approach proposed by Beven and Binley (1992). The proposed analysis has been applied with specific reference to urban flooding and taking the sewer system of Parco d’Orleans in Palermo (Italy) as case study.