This current paper outlines the investigation procedures used to calibrate and verify a new hydrologic model on five permeable pavement structures in two different climates: Birkdale, New Zealand and North Carolina, USA. A beta version computer modelling package designed by the United States (US) Environmental Protection Agency (EPA), the Storm Water Management Model for Low Impact Design (SWMM5-LID(beta)), was used for the procedure. The calibrated model for the Birkdale site demonstrated accurate prediction of the system's response for both individual storms and continuous simulation (long-term). The model accurately predicts the peak flow response of the system for the highly frequent events (> 30% exceedance probability) with the measured and modelled flow frequency curves matching over this range. Volume control prediction is less accurate as the model over predicts retention for events greater than 10% exceedance by an average 4.7 mm. The calibrated model for the four North Carolina pavements (except for CGP) demonstrated accurate prediction for storms between 10 mm and 30 mm. The sand fill of the CGP provided additional storage which reduced peak underdrain outflow and volume. Modelling of this relationship requires significant modification of the fixed parameters in the model, which was not attempted. Simulated outflow hydrographs for storms <10 mm followed a similar shape to the measured hydrographs but over predicted the response. Simulated outflow hydrographs for storms >30 mm again had very similar shapes to the measured hydrographs but could not predict peak flows. Two parameters governed the response of the permeable pavement model: the drain coefficient and the drain exponent. A drain exponent of 3 is used in all of the models. A higher drain exponent increases the accuracy of the model but the corresponding drain coefficient becomes smaller and smaller. A calibrated drain coefficient of 0.000006 is used for the Birkdale site, and for the sites in North Carolina (PICP1, PICP2, CGP and PC) 0.00024, 0.001175, 0.00013 and 0.000725 respectively, were determined.
展开▼