Summary: | Geologic mudrock formations can be used for different industrial applications such as natural gas production, CO2 sequestration, and radioactive waste storage. In natural gas production the rock is the source for gas, mainly methane, that is exploited as an energy source. In CO2 sequestration and radioactive waste storage the rock acts as a flow barrier preventing the release of CO2 or radionuclides into the environment. The release or retention of fluids and gases is controlled by a variety of flow and transport processes in the mudrock reservoirs. Pore scale flow processes play a crucial role determining fluid supply through pore arrangement, porosity, diffusion, adsorption/desorption of oil, gas and water. That is why prediction of flow and transport at the pore scale is relevant for productivity and long-term integrity estimations of reservoirs. Therefore, the detailed characterization of the mudrock microstructure is crucial. The microstructure however, is heterogeneous and mineralogy varies from a sub-μm to mm scale, and pore-sizes range from sub-nm to several μm. It is challenging to resolve this heterogeneity on all relevant scales with analytical methods. Imaging techniques provide the most detailed and localized insight into pore arrangement but fail to investigate statistically relevant sample volumes including pores < 50 nm in size. Fluid invasion and radiation measurement methods measure volume averaged bulk properties, and hence do not resolve local features. This results in a situation where experimental methods cannot be used to simultaneously study the local characteristics on all scales, and probe statistically relevant sample sizes. As a consequence, little is known about the systematic variation of features on the pore scale, over which scales it extends, and how this affects flow and transport mechanisms at larger scales. In this thesis small- and wide-angle X-ray scattering microscopy (SAXS-WAXS) is used as a novel method to overcome these current experimental limitations in the pore scale characterization of mudrocks. In the first experimental results section we demonstrate that a volume averaged but local characterization is achieved by focusing the X-ray beam to microscopic dimensions measuring 10 × 25 µm2. By applying the method on specially prepared mudrock thin sections of 10 – 30 µm thickness, microscopic rock volumes are probed below the scale of the rock’s structural heterogeneity. Operated in a scanning microscopy mode, thousands of consecutive measurements are collected at an effective scanning resolution of 10 × 10 µm2 over 2 × 2 mm2 sized sample areas. Then the measurable parameter space is investigated on rock samples oriented parallel and perpendicular to the bedding plane. We measure pore orientation, porosity and pore size distribution for pores 4 – 130 nm, and mineral features on a separate WAXS detector. We find that pores are aligned parallel to the bedding in rocks cut perpendicular to the bedding plane, while pores do not have a preferential alignment in rocks cut parallel to bedding. Local porosity values in selected matrix volumes in Opalinus Clay show considerable variation which could be related to differences in local mineralogy. In a multi-scale characterization effort, aiming at a full description of the entire pore spectrum, the method is combined with other high resolution imaging techniques including FIB-SEM, SEM and µ-CT. For a validation of the method measured pore size distributions from Opalinus Clay are compared to SANS, N2 physisorption and FIB-SEM data, showing a good agreement between all measurements. In the second experimental results section, several improvements are made to the experimental protocol achieving a higher scanning resolution of 5 × 5 µm2. By performing two separate SAXS and WAXS experiments at different sample detector configurations, a wide q-range is measured to resolve pore features from approximately 6 – 202 nm and mineral reflections in the angular 2θ-range of 4 – 24°. The investigated sample set includes two Posidonia, one Opalinus Clay and, two heterogeneous Eagle Ford mudrock samples, potentially yielding a variety of structural features. Samples were mounted perpendicular to the bedding plane and polished using an ion beam surface polish. Respective SAXS parameter maps show the pore orientation and scattering intensity for sample areas measuring 1 × 1.75 mm2. The measured degree of orientation – indicating the tendency of pore alignment – varies from 0.3 to 0.7 between the Eagle Ford and Opalinus Clay. These results indicate a much stronger preferential pore alignment in Opalinus Clay, which could be related to the high clay content in it. We also find that the degree of orientation changes as a function of pore size, which cannot be explained with the experimental data. From the intensity distributions I(q) in the SAXS patterns porosity is calculated using the Porod invariant, and the distribution is plotted over the entire sample areas. The maps reveal that local porosity variation in fine grained samples is determined by the distribution of large minerals in the matrix. In the more heterogeneous Eagle Ford sample more features are visible and the porosity distribution map is characterized through two different porosity domains in the matrix. Representative elementary volumes (REV) for the porosity for all samples are at volumes < 300 pixels (< 7.5 × 〖10〗^(-14) m3). Further local mineralogy was quantified from the WAXS data in microscopic sample volumes in the oriented of the Eagle Ford sample using Rietveld refinement. Corresponding mineral distribution maps mainly show the heterogeneous distribution of calcite, quartz, kaolinite, mica, bassanite, in a sub-area of 60 × 405 µm2 containing 927 patterns. The porosity and calcite distribution correlate, suggesting that the porosity in the matrix in that sample is largely controlled by the calcite content, but must also be influenced by other effects, e.g. clay minerals or microscopic grains. Thus, the results show that quantitative SAXS and WAXS measurement can be effectively combined to investigate characteristic relationships between mineral composition and pore distribution. These characteristic relationships may play as important role for upscaling scenarios, where flow properties can be predicted based on porosity and mineral composition.
|