Identification and characterization of individual fractures in 3D networks of microtomography – a first step towards multi-scale analysis of reservoir fractures

Abstract:

Fractures provide significant conduits for fluid flow in tight (low porosity) reservoirs. Hydraulic fracturing is often used to create fractures and thus to increase permeability and enhance hydrocarbon recovery. Although such technique is commonly used in the petroleum and geothermal industry, the relationships between reservoir rock, stress and fracture formation are not well understood partly because the three-dimensional (3D) geometry of subsurface fractures is difficult to image directly at the resolutions required. Microtomography enables the observation of 3D internal structures (both pores and fractures) of rocks at micro-scale. Fractures at micro-scale show similarity with those at macro-scale and can be described by power-laws based on previous two-dimensional (2D) studies of fractures. Aiming to establish the scaling law of fractures in 3D space, we characterize fractures in microtomographic images in this study. In our workflow the first crucial step is to identify individual fractures in the 3D network. Starting from 2D, percolation theory is used to detect the connectivity of fractures, and a modified moving window method is used to detect the strike of a fracture – by changing the placement of the moving window following the intersection of the fracture and the boundary until the end point of the fracture is found. The 3D topology of a fracture is determined by the analysis of the connectivity of fractures in 2D slices. Once individual fractures are identified and registered, the characterization of fractures can then be achievable. Direct characterization parameters include the position of each fracture, the size (in voxels), orientation, and dimensions in three principal orientations. Derivative parameters include the density of fractures, the density of intersections, and the statistics of the direct parameters. This technical progress promises further development of the multi-scale analysis of reservoir fractures.