Risk based flood management of sewer systems

Abstract

Traditional sewer flood management is often norm-based. An alternative approach is risk-based sewer flood management. In risk-based sewer management it is possible to decide upon which measures to take in a more expedient way; it is possible to judge which measure is the best in reaching the goal of reducing risks due to flooding. A risk is defined as the chance that an event takes place multiplied with the effect of such an event. By analysing the type of effects and the frequency at which they occur, it is possible to compare different locations where flooding occurs. The main goal of this master thesis is to develop a method for using risk-based flood management of sewer systems. This method has to be appropriate for comparing the risks at different flooding locations and compare possible measures to prevent flooding based on their expediency. The norm used for optimizing the method is that it must result in a general and straightforward applicable method giving an unambiguous result. The following research question covers the goal of the master thesis: ‘How can risk-based flood management of sewer systems be implemented in a general applicable way?’ First an existing method for risk-based flood management is analysed. This method is used as a basis for the method developed in this thesis. In order to try to improve the existing method, the weak points of this method are altered in the method that is developed in this thesis. The method in this thesis determines the risk level using the severity score. The severity score is obtained by multiplying the occurring effects by their accompanying weights. These weights depend on the severity of effects as judged by the municipality. The risk level is then dependent on the severity score and the return period of the situation in question. Besides the risk level, also the annual expected severity score is calculated based on the severity score of a range of return periods, varying from small (0.5 years) to large (100 years). The annual expected severity score is useful to make a distinction between locations with the same risk level. In order to come to a well-founded choice regarding the design storms to use in the method, the Dutch design storms and composite design storms are compared. The underlying principles of both types of design storms are analysed and compared. Besides that, the difference between using a full precipitation series and the design storms is analysed for a couple of sewer systems. Based on these considerations, the composite design storms are used in the method. Based on a literature study the relevant effects of sewer flooding are obtained. The effects that are taken into consideration to determine the risk level are: flooding of buildings, risk of casualties, infection risk, traffic disruption and flooding of public space. For each of the effects, the way in which they can be quantified is determined. In order to be able to judge the applicability of the method, the uncertainty involved in using the 1D/2D-model and allocating the effect category weight are obtained. The model uncertainty is obtained by varying a couple of relevant model parameters and analysing the influence of this variation on the results. From this analysis it follows that the subcatchment area is the most influential parameter. The uncertainty in allocating the effect category weight is obtained by analysing the results of a questionnaire in which respondents are asked to give their judgment about the severity of the relevant effects. This analysis shows that mainly the judgments about intangible effects, like risk of casualties, infection risk and traffic disruption are very wide spread. A comparison of the influence of the model uncertainty and the uncertainty involved in allocating the effect category weights shows that the latter has a much larger influence on the risk level and annual expected severity score than the first, although the influence of model uncertainty is not negligible. The developed method in this thesis is a general applicable method for risk based flood management of sewer systems that gives unambiguous results. The main weak point of the method is the allocation of the effect category weights. In order to improve the method in future, attention has to be given to this aspect. To reduce the relative small influence of model uncertainty, a detailed investigation of the amount and characteristics of subcatchment area surcharging to the sewer system is recommended.