Upscaling micro-heterogeneities in two phase flow in porous media: Theory and experiments

Characterisation of contamination of groundwater by nonaqueous phase liquids (NAPL) has been a topic of great interest for many years. These liquids are highly mobile and move as a separate phase in the groundwater system. Many soils and geological formations contain small-scale heterogeneities that have distinctly different multiphase flow properties than the main medium. These micro-heterogeneities play an important role in the spreading behaviour of non-aqueous pollutants. They also have significant effects on the amount and distribution of residual saturation once the flow of NAPL stops. They are often sources of localised pools of pollutants such as NAPL. Commonly, one is not interested in the details of fluid distribution in such a medium. Moreover, it is often computationally not feasible to discrete a compositional multiphase model at such small scales. Many upscaling approaches have been developed for and applied to heterogeneous media. In these methods, the porous medium is conceptually replaced with a homogeneous medium with equivalent multiphase flow properties. Many of these approaches focus only on the upscaling of saturated hydraulic conductivities and calculating an equivalent permeability for the heterogeneous media. There are some approaches that also include appropriate constitutive relationships of multiphase flow for the resulting homogeneous media. In this project, a novel method for the upscaling of saturated permeability, relative permeability and capillary pressure-saturation relationships will be developed. The approach is based on the identification of a Representative Elementary Volume (REV) for the heterogenous system. It will be applied to a soil with random microscale heterogeneities. Providing upscaled theories of multiphase flow is a scientific challenge that has to be met at various fronts. In particular, verification and testing of upscaled theories is crucial to their success. This must be done with the aid of laboratory experiments. In this project two-phase flow experiments in a centrifuge will be promoted as an attractive alternative to traditional laboratory experiments is carrying out displacement experiments in a centrifuge. This is a recent development in the study of two-phase flow and promises to be an important tool in scientific research on two-phase flow in porous media. The centrifuge facilities of GeoDelft are extremely will suited for this purpose. The use of centrifuge in two-phase flow studies is at an early stage. But GeoDelft is taking a leading role in this regard. In the framework of a recent EU-project NECER, preliminary tests have been performed in the GeoDelft centrifuge. Results, which have been analysed in cooperation with TUDelft, are very promising. The main objectives of this research work is to address open questions on the micro-heterogeneities and the rol of dynamic effects and effects on PcS-kr. An upscaling method will be developed for homogenisation of porous medium with micro-heterogeneities. The study includes theoretical and experimental parts. The experiments will be performed in the GeoDelft centrifuge facility. The experimental program is designed with the main goal of obtaining detailed quantitative data to study phenomena like dynamic effects and effects of micro-heterogeneities on Pc-S-Kr, and also to address scale problem in centrifuge experiments. The proposed research fits within the general framework of the TRIAS Call for Proposals and in particular subtheme "Soil heterogeneities and effects on transport process". In fact, heterogeneity and scale-up of phenomena plays a dominant role in all three sub-themes within the TRIAS-program "Subsurface Processes". In particular, when considering Natural Attenuation and two-phase flow, mass transfer from the pure contaminant phase to the aqueous phase is dominated by the geometry of the pure phase distribution. In this distribution, micro-heterogeneities play an important role. The results of this research will be of direct relevance to the design of remediation techniques and to the formulation of soil management policies. This project has strong links with TRIAS projects, TRIVA 15, 16 and 18 of Theme 3 (Reactive interfaces). Characterization of heterogeneities (either stochastics or deterministic) is the major issue in all these projects. They deal with dissolved pollutants and saturated flow whereas this proposed research deal with multiphase flow. Also, There are ongoing interactions among researchers active in these four projects.