On the validity of modeling concepts for the simulation of groundwater flow in lowland peat areas – case study at the Zegveld experimental field

Abstract

Abstract. The groundwater flow models currently used in the western part of The Netherlands and in other similar peaty areas are thought to be a too simplified representation of the hydrological reality. One of the reasons is that, due to the schematization of the subsoil, its heterogeneity cannot be represented adequately. Moreover, the applicability of Darcy's law in these types of soils has been questioned, but this law forms the basis of most groundwater flow models. With the purpose of assessing the typical heterogeneity of the subsoil and to verify the applicability of Darcy's law, geo-hydrological fieldwork was completed at an experimental field within a research area in the western part of The Netherlands. The assessments were carried out for the so-called Complex Confining Layer (CCL), which is the Holocene peaty to clayey layer overlying Pleistocene sandy deposits. Borehole drilling through the CCL with a hand auger was completed and revealed the typical heterogeneous character of this layer, showing a dominance of muddy, humified peat which is alternated with fresher peat and clay. Slug tests were carried out to study the applicability of Darcy's law, given that previous studies suggested its non-validity for humified peat soils due to a variable horizontal hydraulic conductivity Kh with head differences. For higher humification degrees, the experiments indeed suggested a variable Kh, but this appeared to be the result of the inappropriate use of steady-state formulae for transient experiments in peaty environments. The muddy peat sampled has a rather plastic nature, and the high compressibility of this material leads to transient behavior. However, using transient formulae, the slug tests conducted for different initial groundwater heads showed that there was hardly any evidence of a variation of the hydraulic conductivity with the applied head differences. Therefore, Darcy's law can be used for typical peat soils present in The Netherlands. The heterogeneity of the subsoil and the apparent applicability of Darcy's law were taken into account for the detailed heterogeneous model that was prepared for the research area. A MODFLOW model consisting of 13 layers in which 4 layers represent the heterogeneous CCL was set up for an average year, assuming steady-state conditions; and for the winter of 2009 to 2010, adopting transient conditions. The transient model was extended to simulate for longer periods with the objective of visualizing the flow paths through the CCL. The results from these models were compared with a 10 layer model, whereby the CCL is represented by a single layer assuming homogeneity. From the comparison of the two model types, the conclusion could be drawn that a single layer schematization of the CCL produces flowpath patterns which are not the same but still quite similar to a 4 layer representation of the CCL. However, the single layer schematization results in a considerable underestimation of the flow velocity, and subsequently a longer travel time, through the CCL. Therefore, a single layer model of the CCL seems quite appropriate to represent the general flow behavior of the shallow groundwater system, but would be inappropriate for transport modeling through the CCL.