Abstract
Coal seams are characterised as highly anisotropic and heterogeneous geological formations with extremely low porosity and permeability at reservoir conditions. Despite the advances in digital rock physics with the introduction of micro-computed tomography (micro-CT) over the last decade, characterising coal remains a challenge. One such challenge is processing micro-CT images to extract fracture features such as aperture, length and orientation which cannot be directly resolved using micro-CT images. We aim to present an automated fracture identification technique using grayscale images obtained from micro-CT imaging of coal. This technique uses pattern recognition and computer vision to recognise the presence of a fracture, even those that are below the resolution of micro-CT data. This approach will then generate probability maps of fractures while alleviating the dependence on a single segmented image. Later, a three-dimensional fracture system for coal will be developed from the above approach, and this can be used as an input for flow simulations and estimation of permeability. Overall, this technique will enable us to develop multiple realisations of the coal fracture system using a probabilistic method and prove to provide an alternative path to resolve the problems caused by manual segmentation.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have