University of Twente Student Theses


Smart combinations : an alternative to dike reinforcements? Applicability to reduce flood risk in The Netherlands

Klein Wolterink, N. (2022) Smart combinations : an alternative to dike reinforcements? Applicability to reduce flood risk in The Netherlands.

[img] PDF
Abstract:The Dutch method to overcome flood risks is highly institutionalized, dikes are reinforced to reduce the flood risks hardly considering other options to prevent or mitigate floods. While other European member states have been able to implement more multi-layered flood safety – thereby improving their flood resilience – the Netherlands does not effectively use these other options. The other options considered in this research are smart combinations, which use a combination of measures from the different layers of the multi-layered flood safety approach, including the prevention of floods (layer 1), spatial planning (layer 2) and emergency response measures (layer 3). The problem addressed in this research is that the Netherlands focusses mainly on flood prevention, thereby missing the other options such as smart combinations to improve flood resilience. The Netherlands continues to have a strong capacity to resist flooding, compared to absorbing, recovering, transforming, or adapting to flood risks. While previous studies have indicated smart combinations that could be introduced in the Netherlands, it is not yet known where they can be implemented and what the concept of smart combinations can contribute to reducing flood risk. To improve the use of a wider range of flood measures, it is critical to determine whether these smart combinations can be effective, where they can be effective, and how much they can contribute to reducing the risks. To identify whether and where smart combinations can be effective, the characteristics of safety standard segments that allow for the implementation of smart combinations were found in literature. Segments where either the Vital Infrastructure (VI) is normative or where the Local Individual Risk (LIR) is normative, with a normative neighbourhood that has a significantly higher mortality than the other neighbourhoods (a LIR-hotspot). In total, there are 24 safety standard segments in the Netherlands that meet these criteria – 2 VI-normative segments and 22 LIR-normative segments. To quantify the effects of the smart combinations of decompartmentalization and improved evacuation practices at a suitable safety standard segment, a conceptual model was set up. Using only a digital elevation model, dike material characteristics, and outer water levels the water depth and rise rate over a dike ring had to be determined. To set up the model, two assumptions were done: the flow velocities were assumed to be zero and each compartment fills up from the lowest points upwards, irrespective of the breach or overflow location. The breach development over time was calculated with the Verheij-van der Knaap formula, after which inflow formulas were used to calculate the inflow volume into the first compartment. Once the water level in this first compartment exceeds the compartmentalizing dike, overflow to the next compartment is modelled. With the discretized water levels, and the time that it takes to move from one water level to the next, the rise rate could be determined. The results of the model include the maximum water depth maps and rise rate maps, which could be used to calculate the mortality, of which the median mortality per neighbourhood is used in the Local Individual Risk (LIR) calculation. To calibrate and to validate the model, a case study area was chosen: safety standard segment 27-2 – Tholen en St. Philipsland 2. This segment in the province of Zeeland was chosen based on three quantitative parameters that considered the ratio between length of dike and the hinterland, and the number of casualties. The dike characteristics of the dike at dike ring 27 and the normative outer water levels of the Eastern Scheldt were used to calibrate the model to match safety standard segment 27- 2. The water depths, rise rates and mortality maps that were obtained from the model were comparable to the data found in the database which contains all flood scenarios that were used to determine the norms in the Netherlands – the LIWO. Using the median mortality per neighbourhood, the Local Individual Risk differed less than 1% from the LIR values used in the norm derivations (16,742 years compared to 16,600 years for the alert value and 8,371 years compared to 8,300 years for the lower limit value). It was concluded that the model could be used to determine the effects of the two smart combinations on the mortality and the Local Individual Risk. The process of determining the water depths and rise rates with the model and using these results to calculate the (median) mortality and flood risk was repeated for three scenarios – the lowering of the compartmentalizing dike of 1 meter, the lowering of 2 meters and a complete removal of the dike – to determine the effects of various decompartmentalization strategies on mortality and the LIR. Parallel to that, a sensitivity analysis of a variation of evacuation percentages is done, to determine the effects of improved evacuation procedures. The Local Individual Risk for the different scenarios could be reduced by 5% (compartmentalizing dike one meter lower), 16% (compartmentalizing dike two meters lower) and 26% (removal of compartmentalizing dike) respectively. Combining the decompartmentalization with the improvement of evacuation procedures, the flood risks in the area could be decreased more. The evacuation percentage used in the norm derivation of safety standard segment 27-2 is conservative, estimated to be only 6%. Using the average evacuation percentage for Zeeland – 26% - the lower limit LIR for 27-2 can be reduced by one safety standard class if the measure is combined with decompartmentalizing the dike ring. Each improvement of 10% evacuation percentage results in roughly 10% flood risk reduction. Implementing the second- and third layer measures in other areas will require calculating the effects specific for that area, which can lead to other optimal smart combinations for the specific characteristics of the other areas. From the analysis, it could be concluded that there are 24 safety standard segments in the Netherlands where the smart combinations can be effective. Decompartmentalization of a part of dike ring 27 resulted in a decrease of flood risks with 26%, which could be improved by addressing the low evacuation percentage as well. The research has shown that – while the optimal implementation will differ for each segment – smart combinations such as decompartmentalization and improving evacuation procedures can effectively be used to reduce flood risk at 24 safety standard segments, as an alternative or addition to the standard procedure of reinforcing dikes.
Item Type:Essay (Master)
Faculty:ET: Engineering Technology
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