Análise de sensibilidade de redes de drenagem urbana a parâmetros hidrológicos utilizando modelagem hidrodinâmica e dimensionamento automatizado

Abstract

The hydrodynamic models allow a better characterization of the flow in urban rainwater systems, including the backwater effects. Thus, when used in sizing, they produce more efficient and adequate solutions when compared to those obtained from traditional sizing, such as, for example, from the Manning equation. However, they make the process iterative, since changes in the network propagate both upstream and downstream causing each change to be evaluated individually. This work proposes the automatic method of sizing the drainage network called ECO-SWMM that uses the Storm Water Management Model (SWMM) for hydrodynamic calculation of the network following a set of sizing rules in accordance with the usual drainage calculation rules urban, based on the reports generated by each simulation by SWMM. On the other hand, the designers, due to the scarcity of information from the Brazilian watersheds, especially the urban ones, dimension the water runoff drainage system with input variables whose real values are uncertain, compromising the own design. Previous studies such as Allasia and Villanueva (2007) have already evaluated the impact of uncertainty using the cost of the network as an analysis variable, facilitating the understanding of the results by the responsible one. However, none of them evaluated the impact of uncertainty on a network dimensioned using hydrodynamic modeling, due to the complexity of its dimensioning. Thus, this work has the objective of evaluating the uncertainties in the estimation of the hydrological parameters (CN coefficient, watershed area, watershed concentration time and position of the rainfall peak) in the drainage network design using hydrodynamic modeling. The methodology consisted in varying each parameter individually for hydrological simulation in the Modelo Hidrológico Ecotecnologias (MHE), generating the inlet flows for each node of the network for hydraulic simulation in the SWMM by means of the proposed automatic method (ECO-SWMM), which in turn showed is effective for the design of networks that operate exclusively by gravity and with great potential to helping designers in the sizing of urban drainage. On the other hand, the results of the uncertainty analysis showed results similar to those obtained by Allasia and Villanueva (2007); however, with slightly lower values, indicating the use of hydrodynamic models, even in networks dimensioned by gravity, impacts positively in the reduction of uncertainty. The most sensitive parameter was the CN indicating that changes of 10% in its value, alter approximately 30% in the peak flow and 15% the costs. The watershed area and the watershed concentration time had similar results, presenting around 10% of change for peak flow and 5% of costs for typical changes of their values. Changes in the position of the rainfall peak did not significantly change the flows and cost of the network.