This paper presents a methodology for both the calibration and sensitivity analysis of the hydraulic simulation model relative to the C-Town distribution system included in the framework of the Battle of Water Calibration Networks. As regards the calibration, an automatic process based on the SCE-UA algorithm was initially performed then followed by a manual refinement. The subsequent sensitivity analysis allows for the quantification of the effect of the calibrated model parameters on the output variables of the hydraulic model; this latter analysis was performed following a technique based on the Monte Carlo-Latin Hypercube approach. The application of the calibration process leads to a set of parameters which enables the hydraulic model to properly reproduce the observed measures both in fire flow test conditions and in extended period simulation (one week). The sensitivity analysis highlights which of the input variables (pipe roughnesses, minor valve losses, pump speeds) most affect the output variables (reservoir water levels, node pressures and pipe flows), thus providing information useful within the context of system design, rehabilitation and real-time management.
Calibration and sensitivity analysis of the C-town pipe network model
ALVISI, Stefano;FRANCHINI, Marco
2010
Abstract
This paper presents a methodology for both the calibration and sensitivity analysis of the hydraulic simulation model relative to the C-Town distribution system included in the framework of the Battle of Water Calibration Networks. As regards the calibration, an automatic process based on the SCE-UA algorithm was initially performed then followed by a manual refinement. The subsequent sensitivity analysis allows for the quantification of the effect of the calibrated model parameters on the output variables of the hydraulic model; this latter analysis was performed following a technique based on the Monte Carlo-Latin Hypercube approach. The application of the calibration process leads to a set of parameters which enables the hydraulic model to properly reproduce the observed measures both in fire flow test conditions and in extended period simulation (one week). The sensitivity analysis highlights which of the input variables (pipe roughnesses, minor valve losses, pump speeds) most affect the output variables (reservoir water levels, node pressures and pipe flows), thus providing information useful within the context of system design, rehabilitation and real-time management.I documenti in SFERA sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.