Comparison of 2‐dimensional simulation approaches for risk analysis in urban areas | Vergleich von 2-dimensionalen Berechnungsansätzen für die Gefährdungsanalyse in urbanen Gebieten

In urban areas, extreme rainfalls may result in threat to life or physical condition and cause considerable damage to public and/or private properties and infrastructures. In recent decades there has been an enormous effort to assess the risk of urban flooding in order to reduce the damages by making proper provisions. The present study aims at assessment of urban flooding while adopting and comparing three different modelling approaches: (1) the surface and sewer system are coupled bi-directionally using HYSTEM-EXTRAN 2D, which simulates overland flow and flow in the sewer system alternately. In a simplified approach, stormwater is directly applied to the sewer system. In case of overflow, water flows on the surface and may drain back to the system through street inlets, (2) differs from the second model in applying the stormwater. Here the rainfall is applied directly to the surface (except for building’s roofs which are connected to the sewer system) in order to simulate the resulting overland flow in advance of draining into the sewer system. The runoff parameters are set to be identical to the second approach, (3) the surface flow is simulated using HYSTEM-EXTRAN 2D without considering the sewer system. The effective rainfall is reduced compared to the other approaches to count for the drainage capacity of the sewer system.All models were set up for two case studies with steep and gentle slopes respectively. When comparing models (1) and (2), in both case studies similar results were observed in streets. However in model (2) a considerable amount of storm water remains on the surface where the backyard inlets are missing. It is concluded that in model (1) the amount of storm water draining into the sewer system is overestimated. The results of models (2) and (3) differ mainly in the streets, since in the later the manholes are missing and therefore the observed flood extent could be considered as inaccurate. In summary, the results show the importance of considering the bi-directional interaction of surface and sewer system when modelling urban drainage for the purpose of assessment of urban flooding and the traditional approach of connecting all areas directly to the sewer system leads to inaccurate results.