Modelagem numérica de escoamento não saturado em pavimentos permeáveis com presença de drenos de fundo

Abstract

One of the main reasons for the increased frequency of flooding in urban areas is the increase in impermeable surfaces. Sustainable urban drainage practices (Sustainable Drainage Systems, SuDS), such as Permeable Pavements (PP), are alternatives that reduce the volume and peak flow of surface runoff. Few studies presenting computational numerical modeling have considered the physical process of unsaturated water flow through PP layers. The objective is to calibrate the hydraulic parameters exclusively through inverse simulations. The model used was the Hydrus- 1D, and the parameters optimization algorithm was the Levenberg-Marquardt. The model input data were measured in experimental modules of PP (plot scale). The PP experimental modules comprise the wear layer, bedding layer, base layer, and bottom drains. Water inflows and outflows through the PP were monitored on a sub-hourly scale for 12 rainfall events. The results show excellent fit with NSE values> 0.89 and NSE > 0.95 for PP with conventional and porous wear layers, respectively, small mean error (MAE and RMSE < 0.34 cm) and over and underestimation (-0.73 ≤ Pbias ≤10,52) of the model concerning the measured drainage data. The PP with porous wear layer showed better adjustments concerning the applied metrics (-0.83 ≤ Pbias ≤5.02). It was possible to observe that the coating is the PP layer that predominantly influences drainage. Therefore, despite the emphasis on sizing the base layer, more specific criteria need to be adopted for the choice of materials for the PP wear layer is necessary.