Embankments are one of the key aspects of flood protection programs. As the sea levels are expected to rise the coming decades coastal protection is becoming increasingly important. To better prepare for the future, increased understanding of embankment breaching is desired. An area of embankment breaching that has been relatively left untouched is the impact of a foreland during a breaching event. This study examines the different effects caused by the presence of a foreland, namely reduced flow rate through the breach, limited breach growth and an elliptical erosion pattern in the foreland.
Embankments are generally more erodible than forelands, and erosion of the foreland occurs after the adjacent embankment has been breached to its base level. As water continuous to flow over the foreland, a plunging jet will form, falling from the high foreland into the breach bottom. This causes undermining in the foreland and subsequent headcut erosion. A distinction is made between cohesive and non-cohesive forelands. Cohesive soil layers experience rotational failure and non-cohesive layers experience sliding failure.
A model is presented that calculates headcut erosion in a foreland, flow characteristics over the foreland and erosion shape of the foreland. The erosion shape of the foreland affects the discharge through the breach and is captured in the non-dimensional constant m. The presented model is implemented as a module into the existing breaching model BRES. The added functionality will increase the accuracy BRES model results.
Two experiments were performed to verify the model. The first experiment examined soil failure of a foreland in dry conditions by pushing an extracted foreland soil layer from a table to observe the failure type and the effects of vegetation on the strength of the soil. The second experiment was performed in collaboration with the University of Antwerp. A soil layer was extracted from a foreland and placed in a large flume. The foreland was subjected to a plunging jet to simulate a breaching event. The outcomes of the model were in line with observations made during both experiments.
The presented model serves as a significant first step towards implementing forelands in flood protection programs, but further research is suggested. The uncertainties stemming from assumptions in the model and the underlying sediment transport relations are significant. Most importantly uniformity of the soil is assumed, whereas in reality root systems in fertile forelands can lead to significant local strengthening. The assumptions in this model are conservative, but additional studies can further increase the accuracy of the foreland- and BRES model.