3-D fracture tracing for X-ray computed tomography data
Access full-text files
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
X-ray computed tomography (CT) is a nondestructive imaging method that shows differences in X-ray attenuation, which is a proxy for density. Sensitivity to density differences makes CT a good choice for imaging open natural fractures because of the significant density contrast between air and rock. The image data, however, are prone to artifacts and blurring, which makes taking accurate measurements challenging. As a first step to addressing this problem, a tool was created to quantify the spatial resolution of CT data with a point-spread function (PSF). The PSF tool permits accurate measurement of fine-scale features in CT data – critical for measuring fractures, which are often thinner in one dimension than the PSF size, in turn influencing measurement of aperture. The difference between the measurements given by the PSF method and a simple threshold value pick is shown to demonstrate a non-trivial improvement in accuracy.
In addition, a 3-D fracture network tracing algorithm was developed, for which the PSF is a necessary input, to characterize accurately the network’s attributes, such as fracture orientations, apertures, and roughnesses. To date, only single, isolated, relatively flat fractures have been characterized thoroughly in 3-D for CT data, and most numerical modeling has been conducted on 2D subsets. Additionally, research involving fracture networks is currently limited to simplified numerical models and simulations. This new work extends CT fracture characterization to the majority of fractured materials with 3-D networks, thus providing a source of real data for studying the difference between fracture networks and single fractures. A sample of Packsaddle schist, with a large number of thin fractures and complex, anastomosing bifurcations, was selected as a testing ground for the network tracing algorithms being developed. Preliminary analysis verifies the method’s efficacy.