The use of integrated modelling in an urban water system and its influence on flood estimation

Heesbeen, Jeroen (2011) The use of integrated modelling in an urban water system and its influence on flood estimation.

Abstract:In an urban catchment there are several systems present: a sewer system (a combined or separate system), a receiving water body and surface flood pathways created during extreme events. Regulating urban flooding is very complex because of the interactions between the sewer system and the surface water system. Presently each of the policies regulates a different part of the urban water system, but there are no policies that consider the urban water system as a whole. In practice this can easily lead to problems. If the water levels of the receiving water body are too high there can be no free discharge of combined sewer overflow, resulting in backwater in the sewer system. These backwater effects can cause flooding in the urban area if the storage of the sewer system is filled by a rainfall event. Furthermore combined sewer overflows can have a significant effect on the peak discharge of the receiving water body. To determine the impact of flooding models are used which calculate a scenario that occurs on average once every few years, the so called design-event. If flooding is caused by the interaction of the sewer system and the receiving water system it is unclear which design-event should be used to assess the flood risk and the return period of the event, because the flooding can be caused by multiple types of events. Although it is possible that interactions occur between the subsystems in an urban water catchment, water systems in the urban area are often modelled without little cross-reference to each other, although in more complicated situations this may be required. To determine whether integrated modelling should influence the design of the urban water system a case study was performed. The study area is the industrial area De Vosdonk in Etten-Leur, a town in the Dutch province Noord- Brabant. Near the Vosdonk are frequent problems with flooding due to the limited discharge capacity of the receiving water body and the insufficient capacity of the sewer system. The objective of this study is to compare different methods of integrated modelling and determine the influence of integrated modelling on return periods of flooding of the urban water system near the Vosdonk. There are different methods of integrated modelling available, each with their own strengths and weaknesses:  The models can be run separately and allow for interaction using level hydrographs, this method is easy to set up, but because a number of iterations are required to increase the accuracy of this method, the calculation times are higher than with the other methods.  Coupling of existing models using a modelling framework. A relatively new modelling framework is the Open Modelling Interface (OpenMI). At the moment, only a few case studies have been done using OpenMI and none of them used the models that are used in the Netherlands.  Fully integrated models. Fully integrated models are difficult to create because of software limitations, but once they are finished they are easy to use. The main difficulty in creating fully integrated models is that current models often cannot be used because they were not built with integration in mind. To determine which integrated modelling method is preferable to determine the flood risk of the urban water system, the different methods of integrated modelling are evaluated on a number of criteria. Since the calculation times for the method using level hydrographs are too high and the fully integrated model gave memory errors for long simulation periods, the model framework OpenMI was the preferred method in this study. The influence of integrated modelling on the calculated number of floods is determined using a timeseries of precipitation data of measurement stations near the study area. Using fifteen years of precipitation data, two types of calculations were performed. The first one is a stand-alone calculation of the sewer system and the receiving water body and the second one a calculation using the integrated model in OpenMI. The number of floods in the sewer system that were calculated in both methods was compared and there was a small, but no significant, increase in the number of floods calculated with the integrated model. Although the number of floods did not change much using integrated modelling, the flood depth increased compared to the stand-alone calculations. The “stand-alone” model of the receiving water body already uses a simple schematization of the urban area and this proved a good estimation of the water levels in the receiving water body. From the results of the integrated model it can be concluded that integrated modelling is of no added value for the sewer system in this case when the policy is only based on the frequency of flooding, since there is no significant increase in the number of flooding events. However, when the policy is based on flood risk, integrated modelling is of importance, since the calculated flood depths increase when using the integrated model. The added value of integrated modelling for the receiving water body is limited since the stand-alone models often incorporate a simple schematization of the sewer system which can give a good insight in the outflow from the sewer system to the receiving water body. Policy makers which incorporate flood risk in their plans should determine the need for integrated modelling in their plans. A first insight whether it is necessary to create an integrated model is to determine how sensitive the urban water system is for interactions between the subsystems. Since this study showed that integrated modelling influences the total flood risk, this will be likely the case in other urban areas as well, especially those with similar characteristics as the Vosdonk. Future research should focus on whether there is an effect of integrated modelling on the flood risk in the urban water system for urban areas with other characteristics, such as a smaller urban catchment compared to the total catchment size, as well.
Item Type:Essay (Master)
Faculty:ET: Engineering Technology
Subject:56 civil engineering
Programme:Civil Engineering and Management MSc (60026)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page