|Trapping Phenomena for Two-Phase Flow in Porous Media and its representation in Upscaling Procedures|
|Project manager:||Prof. Dr.-Ing. Rainer Helmig|
|Research assistants:||Dipl.-Ing. Hartmut Eichel, M.Sc.|
|Duration:||1.9.2003 - 31.12.2004|
|Funding:||This research is supported through a European Community Marie Curie Fellowship|
|Project Partners:||The resarch was carried out in cooperation with
C. J. v. Duijn and I. S. Pop from the Applied Analysis Group at the TU Eindhoven.|
Abstract:Two-phase flows in porous media are of utmost importance for
environmental engineers trying to remediate sites contaminated with
e.g. hydrocarbons in the subsurface. The presence of rock
heterogeneities as occurring in most natural formations influence the
flow behaviour significantly and hence complicate the computations.
If the medium is heterogeneous, trapping of one phase (DNAPL) may
occur at interfaces separating two homogeneous layers having different
permeabilities. This process, which is mainly due to capillary
pressure effects, has been revealed by experiments.
The aim of this project is to establish and investigate a
multi-dimensional two-phase porous media flow model that includes
interface trapping phenomena. Once the model is clearly stated, a
simplified situation will be studied: a lens embedded into a
homogeneous medium (the cell problem).
First we look for steady state solutions where the DNAPL stays trapped
at the interface, and in particular we search for a maximal solution
that would give the maximal amount of trapped DNAPL.
Next we continue our investigations by seeking non-steady solutions of
the standard cell problem. This knowledge will be used for a
homogenization based upscaling procedure. The main goal is to propose
an effective equation for two phase flow in a porous medium that
contains many lenses. This model should account for the trapping
effects at the micro-scale.
In recent years, the necessity for incorporating time-depedent terms
in the saturation--capillary pressure--relationship has been
proposed. Up to now, most of these relationship are established
through experiments and subsequent parameter fitting.
Therefore another aim of this research is to investigate the framework
in which time dependent terms may occur during a homogenization
procedure, and do enhance the physical understanding of this terms.