Browsing by Author "Vasin, Milos"

Now showing 1 - 1 of 1

Open Access
Influence of the soil structure and property contrast on flow and transport in the unsaturated zone
(2010) Vasin, Milos; Neuweiler, Insa (Prof. Dr. rer. nat.)
The unsaturated zone represents a transition zone for contaminants spilled on the ground surface, which by reaching an underlying aquifer cause groundwater pollution. The complexity of processes occurring in the unsaturated zone and highly heterogeneous structures, which are never known in detail are the main reasons why it is challenging to make predictions of processes under unsaturated conditions. Instead of resolving the exact distribution of heterogeneities,which would cause enormous computational and field effort, flow and transport in the unsaturated zone are very often modeled in an average sense, where the input parameters of the model are spatially averaged (effective parameters). Derivation of those simplified models is called upscaling. In this study, modeling of flow and transport in the unsaturated zone, when upscaled models are used has been investigated. This study focuses on upscaled models for flow and transport in the unsaturated zone derived by means of homogenization theory. Those two derived models can not be used in general since they are derived under certain assumptions, which are necessary when homogenization is used. Therefore, many questions rise, when those models are used considering their limitations. One of the major assumption of upscaled models derived by means of homogenization is domain periodicity. In this case the effective parameters could be derived explicitly as the structure is known in detail. However, in nature the structure of domain is usually unknown and effective parameters have to be estimated. The derived upscaled models could be only considered as reliable when the effective parameters are estimated in an adequate and effective way, capturing the influence of heterogeneities on the smaller scale. Additional to the periodicity and difficulty with estimation of effective parameters, mentioned models could be derived either for equilibrium or non-equilibrium conditions, dependent from the parameter contrasts between materials. However, in order to distinguish if equilibrium or nonequilibrium model is more suitable for modeling of flow and transport processes, typical time scales have to be estimated. In order to investigate above mentioned challenges with regard to effective parameter estimation, assumption of upscaled models and time scale analysis, three lab experiments have been performed. The experimental data have been compared with numerical simulations or analytical solutions. The experiments done here have been performed under well controlled conditions with artificial heterogeneous structures. As a result, the conclusions of the experiments are specific for these typical conditions. The first part of this study has been focused on flow in the unsaturated zone under equilibrium conditions, meaning that the upscaled model has been derived using a small soil parameter contrast. Different structures, with significantly different connectivity (periodic and random structure) have been investigated in order to gain a better knowledge of the structural influence on the estimation of effective parameters. Additionally, the applicability of the mentioned upscaled flow model under ideal and non-ideal conditions has been assessed such that the domain does not fulfil the assumption of periodicity, but also of small parameter contrasts needed in the case of equilibrium model. It has been shown that the estimated parameters used in the upscaled 1D model performed well. Estimated parameters based on only rather limited information were sufficient to predict the drainage process very well. The flow in the unsaturated zone under non-equilibrium has been investigated in the second part of the study. This implies that the parameter contrast between soil materials used in the experimental study was large. Different options for estimation of typical time scales have been presented and discusses as they are decisive in order to chose, which upscaled model (equilibrium or non-equilibrium) is more appropriate to be used. The obtained time estimates have been further compared with the experimental and numerical findings. It has been shown that the water capacity was the crucial parameter in order to make good drainage time predictions. The system in this example has reacted with the fastest predicted time scale. During the third part of this research, solute transport under equilibrium and non-equilibrium has been investigated. The goal was to observe if equilibrium or non-equilibrium of solute transport could be predicted by using time scale analysis. The estimated time scales have been compared with experimental results. The equilibrium and non-equilibrium have been obtained during the experiments leading to tailing and retardation of tracer. Both equilibrium and non-equilibrium conditions could be predicted by the time analysis. Model assuming equilibrium would give bad predictions of solute transport in case of experiment, where nonequilibrium occurred.