Assessment of Sheet Pile and Cofferdam Reinforced Dykes in PLAXIS 2D Based on the Dutch Guidelines

Abstract

The rise in sea levels and changes in safety standards necessitate reinforcing earthen dykes to comply with the permissible probability of failure. Traditional methods of raising and widening dykes, such as berms, present significant challenges in densely populated areas due to space constraints. As an alternative, integrating structures like sheet piles into dykes offers a viable solution. However, despite conservative design guidelines for these solutions, excessive deformations near the toes of dykes have been observed in several instances, resulting in damage to nearby properties. In this paper, we utilize the finite element method to model the soil-structure interaction in dykes reinforced with unanchored sheet piles and cofferdams (two sheet pile walls connected with anchors). We focus on hypothetical cases typically found in the Netherlands, while using typical Dutch soft (organic) soil data as obtained from the Eemdijk test. The models focus on determining the long-term deformations near the dyke during high water levels, and the factor of safety for macro-stability based on deterministic approach and the semi-probabilistic method outlined in the Dutch guidelines (POVM, 2020a). The model shows the benefit of constructing a cofferdam as dyke reinforcement when it is not possible to use inclined anchors. The use of cofferdams as an independent structure decreased lateral displacement at the top of the wall by 40% and increased the stability by 16% when compared to unanchored sheet pile wall. The effect of stress path dependent shear strength on the factor of safety is demonstrated. Additional deformations resulting from soil predrilling and liquefaction during installation are also evaluated. These effects are reflected in the properties of the interface elements, represented by its strength reduction factor. The influence was most prominent at the inner toe, where lateral displacements doubled when decreasing the interface strength reduction factor from 0.8 (no predrilling) to 0.5 (with predrilling).