| dc.description.abstract | In low-lying areas such as the Netherlands the land is protected from floods and high tides by a flood protection system. The safety of the dikes part of this system are threatened by collapse, which can be caused by the dikes structural integrity under the increased loads imposed on it during a high water event. In particular, increased groundwater flow can lead to the formation of critical pore pressures that can either lead to internal erosion resulting in undermining of the dike by piping, or to slips in the inner or outer slopes of the dike. In order to assess the stability of dikes, information on groundwater flow patterns is thus essential. Under the imposed gradients, these patterns are strongly dependent on the geological structure and hydraulic properties of the subsurface. The variation in thickness of the subsurface layers and the continuity of these layers affect the pore pressure and the seepage of the groundwater.
This research investigates how the groundwater flow conditions around and underneath dikes, with a highly heterogeneous subsurface, influence the hydraulic head under normal and high water conditions in the river. Subsequently, the hydraulic head of the 3D model can be compared to the hydraulic head of the 2D model, which schematization has the same boundary conditions and input values as the 3D model. The aim of this research is to explore differences between 3D and 2D modelling in hydraulic heads and groundwater flow paths around a dike during characteristic conditions and a high water event. The investigation takes place on the dike between Wijk van Duurstede and Amerongen. The tool that will be used in this research is a 3D groundwater model in iMOD which is reduced to two dimensions.
The results indicate strong differences between the 3D groundwater model and 2D groundwater model. The hydraulic head distribution shows different behaviour in terms of dimensions:
- The behaviour changes with the distance from the river. A higher difference between the models is observed in the hinterland compared to the river.
- The higher the discharge, the more the behaviour can change around the river as well.
- The depth at which the results are observed show a higher difference deeper into the subsurface.
- The longer the model runs the higher the differences can become. This is mainly in the hinterland.
In terms of underestimation of the 2D model, which can provide false safety if used for dike stability assessments, the following findings were observed:
- Underestimation is highest in the aquifer on days with the high discharge.
- Impermeable clay with an hydraulic conductivity of 0.002 [m/d] has influence on the underestimation of the 2D model.
Concluded can be that a 2D groundwater model mainly overestimates which is OK for dike stability assessments. But the influence of the lithology on the underestimation of the 2D model has not yet been thoroughly determined therefore 3D groundwater model gives a better representation of the hydraulic head prediction in response to the influence in heterogeneity near and under river dikes than a 2D model. | |
