KATWIJK AAN ZEE
Text by Mark Voorendt
Katwijk aan Zee is located near the original mouth of the Rhine, which in early days formed the northern boundary of the Roman Empire. The original mound of this river, which was situated a bit more to the North than the present mouth, was blocked by siltation in 1163 after the Saint Thomas flood.
After the damming of the Rhine near Wijk bij Duurstede, no water flowed through the old Rhine any more, which enabled siltation of the original river mouth. The natural river mouth near Katwijk was excavated again between 1804 and 1807 and meanwhile an inner and outer dewatering sluice was constructed. A steam-driven pumping station was added in 1881 to enable dewatering of the hinterland even during high water levels at sea. In 1954 this pumping station and the old inner sluice were replaced by a modern, electrical pumping station, at present known as the Koning Willem Alexandergemaal, named after the king of the Netherlands.
At the end of the twentieth century, part of the town of Katwijk, located along the coast, South of the dewatering canal, appeared to be not sufficiently protected against storm surges, threatening 3000 inhabitants who in fact lived in an unembanked area (figure). It was the last weak link along the Dutch coast, as reported by Rijkswaterstaat in 2003. Several alternative designs were made to improve the flood protection of Katwijk, meanwhile dealing with growing parking problems along the boulevard. The final design, which was made by Arcadis, is described in the following section, followed by another section about the alternative design made by CURNET, the Municipality of Rotterdam and TU Delft.
The final design
The weak part of the dunes has been reinforced from October 2013 to February 2015 with a dike embedded in the dunes. A sub-soil parking garage for 663 cars has been constructed between the dike and the boulevard. Meanwhile, the dune area was widened and newly arranged. The dike-in-dunes is constructed along the part of the boulevard that is too low to retain critical water levels that could occur on average once in 10 000 years, which is the norm for dike ring 14 (figure). This is the part where the boulevard is lower than NAP + 10,00 m, a stretch of about 900 m. The total erosion volume of the dune and beach in cross-shore direction is leading for the evaluation of the flood-protectiveness of dunes. It is not really important whether this volume is present in the height or in the width, so for aesthetic reasons (view form the boulevard) it was decided to make the dunes lower and to put extra sand volume on the beach. To achieve an even lower dune, a ’hard structure’ was needed to prevent further erosion. The total width of the dunes over the dike, from boulevard to dune toe, is about 120 m. This is 90 meter wider than in the original situation.
The dike has a sand core and is covered by basalton blocks on top of a filter layer and geotextile (figure). The crest level of the dike could be as low as NAP + 7,50 m, but for again for aesthetic reasons the dike is covered by sand, which brings the top of dunes to a level of aboutNAP + 8,00 m. At locations where dunes present before the start of the project were already higher than 7,50 to 8,0m, the original dune top level was maintained4. The crest of the dike has a width of 5,0 m. The dike shall be exposed to wave attack when the sand on and in front of it would have eroded. The sand in front of the dike will then sufficiently reduce wave overtopping. It can relatively easily be adapted in future.
The design methodology of the ’hybrid’ dike-in-dune structure consists of the following main elements:
Methods to determine the crest level of the Katwijk hybrid structure are not prescribed, so a custommade design had to be made. The starting point of the design was the prescription of the Dutch Regulations Assessment of Flood Defences 2007 (VTV2007), intended for the assessment of common dunes, that after the occurrence of a governing storm, a dune profile consisting of only sand should remain with a height of at least 1,00 m above storm surge level and a sand volume of at least 53 m3/m1. In the case of Katwijk, this corresponds with a remaining crest level of NAP + 7,50 m. If the dike is considered as the required remaining dune profile, it could have the same crest level, because no arguments can be found to justify a lower level, regarding the retaining height. As far as the volume is concerned: the remaining profile of a sand dune should be sufficient to retain the water, but the dike slope is also exposed wave attack. The dike, however cannot erode or become unstable, so wave overtopping becomes a major design aspect for the dike. The freeboard is about 1,70 m, which is sufficient to reduce the wave overtopping discharge to acceptable amounts. Lowering the crest height of the dike would imply that more sand will have to be present on the beach, but in that case the flood defence would become too sensitive to relatively small deviations in the assumed hydraulic boundary conditions.
To compensate settlements that will occur during the first few years after construction, the construction height of the dike will be 0,10 m higher than the required crest level at the end of its design life time (50 years).
The parking garage is situated between the dike-in dune and the boulevard (Figures 2.46 and 2.47). It is not part of the flood defence, although it is covered by the same sand as the dike and therefore gives the impression that it is integrated. The main starting points for the design of the parking garage are:
The development of the parking garage, therefore, does not resort under theWater Act and is therefore not included in the project plan for the reinforcement of the coastal defence. The new zoning plan (bestemmingsplan) ’Kustwerk Katwijk’ enables the construction of the parking garage. The reasoning behind the design of the garage is therefore explained in the environmental effect report (milieueffectrapportage, MER) and the zoning plan. The location of the garage is included in the integrated layout plan that is used as an underlay for the zoning plan and the project plan.
The 'Wall-in Dunes' alternative design
In an early stage of concept development, several alternative designs have been made for the improvement of the coastal defence of Katwijk. One of these designs has been made by Delft University of Technology, TNO, Rotterdam’s municipal engineering department, CUR building and infrastructure, and other agencies. It comprised a parking garage in the dunes, but no dike. The seaward wall of the garage was a flood-retaining diaphragm flood wall with a depth of 15 to 20 m (figures).
The idea was that the diaphragm wall would have to resist the waves after erosion of the 30 metres wide dune in front of it. The experience is hat dune erosion - if it occurs - can proceed quite rapidly: about 80 to 100 metres in a few hours, so complete erosion of a 30 m wide dune can reasonably be expected. With help of a computer simulation it was calculated that these 30 m would be completely eroded after 15 h. Waves would than directly hit and overtop the wall.
The amount of the wave overtopping volume that was still considered acceptable was determined by the storage capacity of the area behind the flood defence, rather than the structural integrity of the flood wall itself. This storage capacity was estimated at 50 000m3. A 25mwide bed protection should have prevented scour in front of the flood wall. A side-effect of this bed protection was that it would have acted as a shallow bermduring a storm, thus reducing wave overtopping volumes.
Three types of bed protection were considered: a flat layer of rubble, rubble on a slope and a breakwater. The first solution appeared to result in too much overtopping volumes for the maximum crest level of NAP + 7,50 m. The second alternative was much too expensive (the costs of the rubble was estimated at 30 million euro), but the third alternative with the breakwater was the least expensive (about 10 to 15 million euro for the rubble). The breakwater crest was designed at NAP + 6,50 m with a width of 5,00mand an outer slope of 1:3,5.
The parking garage was designed at the land side of the flood wall. The flood wall had a double function: next to retainingwater, it provided stability to the garage structure. The floodwallwas sufficiently strong and stable on its own, so a possible collapse of the parking garage would not affect the flood protection. The same applies to some restaurants adjacent to the flood wall, at the beach side. They are not needed for the flood defence function. This alternative design was finally rejected by the client for several reasons.