Lumped or partially distributed urban drainage models have been used over the past three decades by drainage engineers for evaluation of control options. During that period, many modelling tools have been developed, however, it is generally recognized that considerable experience is required to properly apply models in the analyses. With the more recent shift in control options towards source control and increasing availability of geospatial data at small scales, existing tools such as USEPA SWMM are being applied to study smaller areas in more detail, and are frequently used as uncalibrated models to evaluate source control options. Therefore, guidance for development of such models is still being developed and this study hopes to contribute to this goal.
This research evaluates and compares modelling single lot with and without low impact development (LID) implementation using USEPA SWMM to build distributed models and lumped models. A hypothetical data was examined by developing several grid based models with different grid sizes. The models were examined under two soil types and different time steps. The results from the hypothetical data were tested using runoff data from a 1.65 hectare developed urban area. The case study examined three different models configurations: 1) one catchment, 2) grid model of hundred metre square area and 3) homogenous areas model where every building, backyard, front lawn and streets were presented separately as single catchments. SWMM was run using local rainfall data for six months and for two short events. The results of the models were compared and evaluated based on the total runoff volume, peak flow rate and infiltration volume. Conclusions are drawn regarding appropriate model structure for small catchments, with and without application of LID practices.