Building With Nature: A System Analysis of Dynamic Double Dike Systems

Abstract

Sea level rise and land subsidence increasingly put more pressure on Dutch flood protections. Additional to regular dike reinforcements, new, more nature-based, solutions are considered, such as a double dike system (DDS). Such system, consisting of two parallel dikes in combination with an interdike area in the form of a salt marsh, offers the opportunity to adapt via hydro-morphological processes to sea level changes. An increased bed level of the interdike area of a double dike system potentially positively affects wave attenuation. Furthermore, if the second (landward) dike breaches, the elevated salt marsh of the interdike limits breach growth and discharge through the breach. However, previous studies show that it is hard to prove the feasibility and effectiveness regarding flood safety of a DDS relative to a single dike and an in-depth hydraulic engineering substantiation for flood risk reduction is lacking. The objective of this research was to investigate which factors determine the effectiveness of a DDS, with the focus on the dynamic development of the system. This leads to the following research question: Under which circumstances can a double dike system be an effective alternative to a regular dike reinforcement? It is possible to distinguish three distinct systems based on the function of the first (seaward) dike: dynamic open DDS, dynamic closable DDS, and static closed DDS. This distinction is primarily based on the degree to which water is permitted to pass through the first dike under both normal and extreme conditions via an opening in the dike. This interaction between the outerdike area and interdike is unobstructed for a dynamic open DDS and almost nonexistent for a static closed DDS. To achieve a more hybrid approach, dynamic closable DDS utilizes adjustable structures like tidal culverts to regulate the inflow. Under normal tidal conditions, the permitted inflow governs the sediment supply and the potential for accretion of the bed level in the interdike area. Depending on the design of the system, during extreme storm conditions, the first dike can act as primary flood defense or as breakwater that provides shelter for the second (landward) dike. This research focused on an open dynamic DDS system, allowing for maximum potential bed level accretion of the interdike area. In doing so, the first dike functions as a breakwater and failure of the system occurs when the second dike is breached. Due to the unrestricted in- and outflow, the assumption is made that water levels are uniform and equal to the hydraulic conditions outside the system. The double dike systems are analysed with the help of two approaches: failure probability approach and local individual risk (LIR) approach. The first approach compares systems based on equal probability of failure. The second approach compares systems based on an equal LIR of 10−5 per year and includes the expected mortality in case of flooding. Systems have been compared to each other with the help of calculated dike heights and total dike volumes. Dikes were tested for failure due to overtopping or overflow. Therefore, a transmission model was applied to describe the transmission of waves through the system: wave transmission over the first dike, wave propagation in the interdike area and wave overtopping over the second dike. The probabilistic calculation method FORM was used to find the failure probability and design point. For the LIR approach, this model was extended with a 1D flood model to determine the flood characteristics and subsequent mortality fraction if failure of the dike occurs. Firstly, a simplified case study was performed based on conditions in the Western Scheldt. Opening the existing dike in the current hydraulic conditions necessitates the construction of a second dike that is only 1.7 meter lower than the single dike system. A sea level rise (SLR) of 0.4 meter in 2070 and 1.0 meter in 2120 was implemented. In addition, morphology scenarios were applied, ranging from no accretion to maximum accretion. In the scenario with maximum accretion, rapid accretion to Mean High water takes place, after which the bed level accretion follows SLR. Comparing systems based on equal failure probability, a single dike should be raised identically to the SLR. In the case of a double dike system and 1.0 meter SLR, this increase in height of the second dike should be approximately 1.15 meter to maintain equal failure probability. Sedimentation in the interdike area has a negligible impact on failure probabilities due to the dominant influence of water depth, meaning the water depth is too high for waves to be affected by the bottom of the interdike area. However, this comparison changes when based on equal LIR. With 1.0 meter SLR, the additional required height of a single dike increases to 1.7 meters. Applying this approach to a double dike system, the effect of the height of bed level of the interdike area on the breach flow does have significant influence. Analysing the effectiveness of a double dike system in comparison to a single dike system based on volume increase, these differences are magnified. However, generally, it can be noted that the extreme storm surge in the Western Scheldt is dominant and a second dike with substantial dimensions is needed to comply to its retaining function. Secondly, a sensitivity analysis was performed on four different parameters and system configurations: 1) height of the first dike in combination with the length of the interdike area, 2) hydraulic conditions with lower water levels, 3) varying area size of the hinterland and 4) implementation of vegetation. The effect of the first dike on the failure probability is not dominant, since the required increase of the height of the second dike is not proportional to the applied decrease in height of the first dike, increasing its effectiveness in terms of dike volume. This effectiveness increases also for hydraulic conditions with lower extreme water levels. In addition, there is a positive relation between the effectiveness and the area size of the flooded hinterland. Vegetation does not have any influence on the dike height due to the high water depths under design conditions. In conclusion, the following three location specific circumstances are important for an open DDS to be an effective alternative to a regular single dike reinforcement: • Sufficient sediments present to allow for a higher rate of sediment accretion compared to sea level rise. • Hydraulic conditions in which waves are more important than storm surge. • Large hinterland area. It is more advantageous to employ the local individual risk approach in comparative studies assessing the effectiveness of double dike systems. The significant positive impact of bed level accretion on flood impact is disregarded when only considering failure probabilities. It is recommended to improve the flood model, implementing a 2D model and considering the effect of the foreshore on breach growth, to gain a more comprehensive understanding of the potential of an elevated bed level of the interdike area on the flood impact, so it can be effectively incorporated into the Dutch flood risk approach. In addition, it is recommended to consider the implementation of an open dynamic DDS in coastal areas with high sediment concentrations and suitable hydraulic conditions (sufficient tidal range and mild storm surge). Furthermore, the wave damping function of the first dike is not critical for interdike areas with sufficient length and reduced depth. It is more efficient to prioritize the flood safety of the second dike. This allows for more effective design and management decisions concerning the height and safety requirements of the first dike. Finally, it is advisable for areas with high water levels during extreme conditions to consider a closable system. This entails the ability to close the gaps in the first dike under specific conditions, such as utilizing tidal culverts, limiting the hydraulic loads on the second dike. However, implementing a closable system necessitates larger investments in modifications to the first dike. Additionally, the maintenance of the first dike becomes more crucial due to its retaining function first. Therefore, it is recommended to conduct a comprehensive cost-benefit analysis to compare the open dynamic double dike system with the closable double dike system.