The mud-rich sediments that comprise the majority of genetic units like levees and hemipelagites are typically characterized as high-quality seals, with very low permeability and high capillary entry pressure. However, they are often disrupted by flow units such as coarse-grained layers, injectites and open fractures. Such flow units could intersect one another to form connected flow networks and can, therefore, impact the overall fluid flow of the particular genetic unit. This paper considers an upscaling workflow for quantifying such impacts and for deriving robust effective permeability to be used in flow simulation at larger scales. The workflow involves stochastically constructing representative models of those genetic units by superimposing the flow units one after another on the top of a background matrix of mud-rich sediments; then carrying out flow simulations on the models to derive their effective permeability values. The results from a simple numerical experiment, which was designed for verifying the workflow, are presented. An analysis on the results shows that the overall along-layer flow is controlled by layer permeability and continuity, with the matrix permeability distribution being a secondary factor; and that the overall across-layer flow is mainly controlled by the matrix permeability distribution in the absence of any connected network of layers and conduits, but otherwise by the connected network.
|Title of host publication||IMA Conference on Modelling Permeable Rocks With Special Focus on CO2 Storage|
|Publisher||The Institute of Mathematics and its Applications (IMA)|
|Number of pages||4|
|Publication status||Published - 1 Apr 2010|