Predicting fracture patterns in sandstones in fold-thrust belts

点击翻译按钮获取中文摘要

Sandstones are important reservoirs for hydrocarbons and are likely to provide most of the opportunities for underground carbon storage. Many reservoirs are hosted in folds formed by compressional tectonics. However, the successful exploitation of subsurface sandstone reservoirs demands understanding of the fractures patterns that form as a consequence of the compression - they can greatly alter the reservoir performance. Significant uncertainty exists - can models of the structure on a large scale predict the distribution, concentration and orientation of fracture damage through this structure? It is this question that this CASE studentship is designed to answer, using an example study from outcrop. This project partners the University of Aberdeen with Midland Valley Exploration Ltd (MVE), world leaders in the provision of structural geology software to the subsurface geoscience industries. The partnership will provide the student with excellent training in commercially-relevant structural geology. The results will immediately benefit MVE and their extensive client base. MVE will meet the standard CASE requirements of a stipend supplement and internship, together with the entire cost of fieldwork and will provide their Move software with technical support without charge. The project involves several components: field structural mapping; 3D model-building; structural restoration; forward modelling; computation of stress, strain and fracture patterns; combination of multiple model outputs; and testing model predictions against natural fracture patterns. The student will therefore be exposed to the entire workflow followed by industrial structural geologists. All of these activities are supported within MVE's Move software. The field case study is located in the southern the Moine Thrust Belt, NW Scotland, where well-exposed and accessible fold-thrust structures, developed in Torridonian sandstones, provide excellent analogues (scale and rock type) for structures in the subsurface. The student will map these digitally in the field using by MVE's Move software. Mapping will be facilitated by using geological and terrain data available through collaboration with BGS. The mapping provides the foundation for creating a 3D model (in Move) for the structure of selected folds. The models in turn will be restored to their undeformed (unfolded) state and then forward-modelled in Move to form evolutionary histories. The history of folding will then be used to model the stress evolution of the constituent beds using Move's mass spring solver. When populated with theoretical rock properties, different loading conditions and fracture criteria, an array of model predictions for fracture patterns will be created that can be combined to form fracture probability maps. These predictions will be tested against direct measurements of fracture patterns made in the field. Existing predictions of fractures through large-scale fold structures use in situ examples with point measurements from subsurface wells and drill holes up-scaled and interpolated to surrounding rocks. Yet up-scaling is fraught with uncertainty - commonly the predictions are not supported by subsequent drilling campaigns. By using an outcrop-based test the student will resolve fundamental questions such as whether fracture patterns correlate with final (observed) geometry of the hosting large-scale fold (e.g. curvature) or reflect the total strain history of the structure. The approach will also delineate which parts of structures are particularly sensitive to model input parameters (rheology, imposed fracture criteria, loading conditions) thereby risking predictions of fracture patterns. Training will be provided in all relevant aspects of structural geology both in the field and through modelling, both in Aberdeen and within MVE offices in Glasgow. The student will complete their PhD well-placed to follow a career in industry or to continue in research.