Rock Fabric Analysis Based on Borehole Image Logs: Applications to Modeling Fracture Height Growth

Abstract

Abstract Completion and production stimulation engineers use isotropic or transversely isotropicwith a vertical axis of symmetry (TIV) mechanical earth models (MEM) as an input to predict hydraulic fracture (HF) height growth. Occasionally the MEM results in models that predict fracture heights at odds with the observed fracture height in the field (e.g., fracture hits, microseismic monitoring). A potential solution to this discrepancyis to add rock fabric into the HF modeling process to better match the containment boundaries observed in the field. Laboratory large-block experiments have shown the importance of rock fabric for HF height growth and containment. Rock fabric in organic-rich mudstones (e.g., Barnett) andhybrid sequences (e.g., Wolfcamp, Eagle Ford), results from the depositional and diagenetic history of the basin andcomprises different elements relevant to HF modeling. Key rock fabric elements include: a) thin, hard layers (named stringers); b) weak beds (e.g., ash beds); and c) weak interfaces (e.g., mineralized fractures, lithology contacts). The firstchallenge in usingrock fabric elements is identifying them with wireline logs (e.g., sonic logs). Because these features are thin, often only a few inches thick or less, they are less than the vertical resolution of most sonic logs. For thatreason, they are generally not considered in the creation of a MEM, which is the key input for modeling HF height growth. A novel rock fabric analysis workflow specifically targeted to HF height growth has been developed. It is based on high-resolution borehole images and the interpretation of the interaction of tensile drilling induced fractures with rock fabric. Qualitative flags (thickness and density of occurrence) are generated for all rock fabric elements found relevant to HF height containment, for subsequent integration into HF modeling. This offers a data-driven workflow to solve discrepancies between the hydraulic fracture heights predicted by models and those observed in the field. For weak interfaces, these flags are integrated into hydraulic fracture models as boundaries where the rock will undergo shear displacement and the fluid flow will experience complexity in its vertical path that will lead to effective containment. It is important to note that a single boundary might not stop fracture propagation, but rather a succession of boundaries is typically required to sufficiently contain a hydraulic fracture.