Get updated about news, events and contributors experiences

Joint HWBP and All-Risk effort to make maximum use of the results of the sheet pile test in Eemdijk

Posted at 31/01/2021 by Matthijs Kok

The paths of Matthijs Kok (program leader of All-Risk), Arny Lengkeek (advisor geotechnical engineering with 20 years of experience and PhD student in All-Risk at TU Delft) and Meindert Van (user in All-Risk and employed by Deltares) were joint by their interest into a sharper design of steel sheet piles for dike reinforcement. This blog reflects how the research into a large test data started, including some recommendations for future studies based on an informal talk between the three.

Contact details Matthijs Kok

Delft University of Technology

Left photo: Snapshots from the informal talk between Meindert Van (user in All-Risk and employed by Deltares), Arny Lengkeek (practitioner with 20 years of experience and PhD student in All-Risk at TU Delft) and Matthijs Kok (program leader of All-Risk). Right photo: Construction of a steel sheet pile as part of a dike reinforcement project (photo by Flood Protection program).

1) Involve users in the formulation of the study

Matthijs: Meindert, you formulated Arny’s research before the start of the All-Risk research program. Can you indicate how you came up with that?

Meindert: Steel sheet piles are already regularly used as dike reinforcement structures (right photo on the top). However, a faster and cheaper dike reinforcement design requires insights into how stronger a dike can be when reinforced with sheet piles and how great the deformations occur? At the Flood Protection Program (HWBP, in Dutch), we were going to do a large sheet pile test at Eemdijk, and the idea was to link this test to a PhD research. The sheet piling test is an HWBP project, and HWBP needed the conclusions under high pressure within a year and a half. However, my feeling was that there is simply a lot more data, which you could extract from such a large test. Apart from that, I spoke to Arny once. Arny liked to get involved in the test, and back then, he also said again: “I would even like to get a PhD on it”. That PhD initiative got delayed for a while, and when preparing the All-Risk proposal, Matthijs asked if I had any ideas, I brought the initiative back.

Photo after the collapse of one of the dikes at the Eemdijk test after applying water against the dike, infiltrating water and placing containers with a load on it.

Matthijs: Then that question came to you, Arny. And what did you think then, a great opportunity?

Arny: Indeed, an excellent opportunity! I already had 20 years of experience in the consultancy practice. However, the opportunities to join a test such as the Eemdijk in the Netherlands are quite limited. I had let myself slip into a meeting that a PhD research on that topic seemed nice to me, and Meindert picked that up.

At the test site in Eemdijk, Deltares and Witteveen + Bos, among other team members, constructed a circular test dike with 60 meters length and 5 meters high. Sheet piling has been installed on one side and not on the other. The dikes are loaded one by one, with different steps: by, among other things, applying water against the dike, infiltrating water and placing containers with a load on it. We are monitoring how and what point the dikes collapse. I think we have been able to measure a lot, and in the meantime have also been able to research a lot.

2) Sharper designs are possible

Meindert: What surprised you in this investigation, Arny?

Arny: The tests taught me that we tend to hedge uncertainties with all kinds of factors. What struck me when using advanced models is that if we calculate with all kinds of representative values for both strength and stiffness, we are therefore even further away from reality than with simple models. It is much better to feed a suitable model with expected parameters because we are not that far from reality. Because we are used to working with uncertainties and covering them all the time, we don’t understand what we are doing, and I thought that was an eye-opener for myself. We shouldn’t apply all kinds of extra margins, only where they matter.

Meindert: The sheet pile test itself has already yielded many results: shortly after the test, it was already calculated that it would make savings in tens of percentages, and Arny’s research will soon substantiated this much better.

The sheet pile test, before and after test conditions in the event of failure of an unanchored wall. Illustration by POV Macro stability available via the water forum website (

3) Practical experience really adds something to scientific research

Meindert: If you had started this research immediately after graduation instead of 20 years of experience in engineering work, could you have extracted what you have now achieved?

Arny: Twenty years of engineering work brought me a lot of experience in various areas, so both in the field of constitutive soil models and the field of constructions and the field of implementation. If I had to do that without experience, it would probably be a failure. Then, I would have based myself only on all kinds of calculation rules without having field experience. I did not learn the knowledge of constitutive soil models and soil-construction interaction at the TU. All these practical expertise, combined with reading a lot of literature, has helped me a lot. If I had started my PhD research straight from university now, I would have had different skills. And then I especially appreciate the programming skills to deal with data much faster. When I see how the students I have supervised do data analysis, I can’t beat that.

4- Fewer sheet piling with anchors in the future

Matthijs: Sheet piles are relatively expensive, compared to the so-called green dikes or dikes without steel reinforcement. Have these sheet piles become more attractive? For example, due to the Eemdijk test and your research and because these dikes do not need to be dimensioned as generously?

Arny: My personal opinion is that green dikes should be the preferred alternative, especially because they are easily adaptable, even with an ever-rising sea level. Sheet piles with anchors are much less flexible, and we should be careful with the needs for adaptation. I am currently seeing dike reinforcement projects where they want to place a sheet pile at every home or farm. And yes, there are quite a few. If a dike cannot be reinforced with soil due to lack of space, an unanchored sheet pile wall is often the best alternative. On the one hand, this alternative saves the costs and challenges associated with anchoring, and on the other hand, the sheet pile wall will still have a second life in the distant future with an anchoring.

Finally, in the Eemdijk sheet pile test, we saw that in the event of failure of an unanchored wall (see figure above), the wall continues to function as a membrane so that that breach formation will occur much less quickly. That slow down pace has a favourable effect on flooding and evacuation options, which we were not yet considering.

Meindert: Yes, in current design practice, we see a lot of sheet piling fitted with anchors, which we want to further optimise in the future and preferably not to use them when possible.

Last modified: 24/02/2021