Abstract
X-ray Computed Tomography (CT) is a non-destructive and powerful imaging technique allowing scanned objects to be three-dimensionally inspected and characterized across a wide range of sample sizes (from meters to micrometres) and resolutions (millimetres to nanometres). For this reason, X-ray CT has been widely applied in many geosciences disciplines and its use grows exponentially with developments of imaging technologies and techniques. The majority of the applications concern petrophysical analysis of rock cores and plugs to estimate, from 3D images, petrophysical parameters such as porosity and permeability, the latter by means of flow simulations within the separated pore volume, without the need of laboratory analyses. Due to the huge impact that pores, pore throats, and fractures have on permeability, a critical step in the image analysis is the characterization of these features by measuring their size, shape, and orientation. This type of analysis is of particular interest in oil and mining industries, and disciplines such as hydrology, volcanology, and structural geology. However, a complete match between laboratory- and image-based measurements is often difficult to obtain because factors such as image resolution, feature size, presence of artifacts, and lack of specific imaging techniques limit and bias the analysis.
Imaging artifacts, such as the Partial Volume Effect (PVE), and the absence of specific techniques particularly complicates the analysis of fractures that, as opposed to pores, are difficult to individually extract and characterize. In addition, connected fractures are typically identified as a single object, thus measurements take into account the whole object which biases results. Furthermore, in a dual porosity system (where porosity has two distinct morphological characteristics), characterization of the pore volume depends on the mean size of the features of interest. Thus, the resolution of the image plays a key role since all features below the voxel size of an image will be unresolved and only the fraction above that size can be imaged and characterized.
In this thesis, the effects of these limitations are explored in three studies where X-ray CT was used to characterize both pores and fractures. Novel approaches are presented for more efficient measurement of the properties of fractures within a scanned rock. The algorithms developed allow definition of local fracture properties such as aperture and orientation in an automated way, and separation of connected fractures for individual analyses. In addition, the effect of poor image resolution on assessment of the petrophysical properties of triaxially-deformed tuff samples is investigated. It is shown that, even though a fraction of the pores below the voxel (volumetric picture element) size are not resolved, significant information in rocks with complex structures can be qualitatively (i.e. via image visualization) and quantitatively obtained for fractures and pores. Finally, high resolution data are used to define 3D vesicularity (i.e. porosity) and connectivity in volcanic lapilli, demonstrating how a well-resolved vesicularity provides insights into the processes that occurred within the magma conduit, and the state of the pre-fragmentation magma in a shallow-water volcano.