Hydraulic Fracture Propagation through Preexisting Discontinuity Monitored by Acoustic Emission and Ultrasonic Transmission

Hydraulic fracturing is critical to enhance hydrocarbon production from ultra-low permeability unconventional reservoirs, and is the common completion methodology for tight formations around the world. Unfortunately, these reservoirs are often highly heterogeneous and their heterogeneity imparts a degree of geometrical complexity in hydraulic fractures that is poorly understood. Fracture complexity (e.g. branching) results in higher surface area and could be beneficial to production provided it remains conductive. Understanding the sources and consequences of fracture complexity is thus of high importance to completion and production operations. In this study we postulate that textural complexity in tight heterogeneous formations induces fracture complexity, and that the main sources of textural complexity are associated with veins, bed boundaries, lithologic contacts, and geologic interfaces. We thus study the effect of interfaces on hydraulic fracture propagation under laboratory conditions by Acoustic Emission (AE) and Ultrasonic Transmission (UT) monitoring techniques. The experiments were conducted on low permeability sandstone blocks of 279 x 279 x 381 mm length with saw cut discontinuities oriented orthogonally to the expected direction of fracture propagation. The rock is loaded in a poly-axial test frame to representative effective in-situ stress conditions of normal and deviatoric stress. Hydraulic fracturing was initiated by injection of silicon oil into a borehole drilled off center from the block. Acoustic emission (AE) events were continuously monitored during testing using nineteen P-wave sensors. Additional sensors were installed to periodically monitor ultrasonic transmission (UT) along various directions oblique and perpendicular to the fracture and the interface. The AE and UT data were recorded using a Vallen AMSY-6 system, with 16-bit amplitude resolution and 5 MHz sampling rate. Detailed analysis of AE localizations allowed us to identify various stages of hydraulic fracturing, including fracture initiation, fracture interaction with the preexisting interface, fracture propagation and fracture closing. Observed variations in UT amplitudes and velocities, during fracture propagation, were related to fracture propagation, as well as to the penetration of fracturing fluid to the conductive interface. Analysis of the combined AE and UT data provides additional insight into the fracturing process and significantly improved our understanding of the dynamics of hydraulic fracture propagation. Detailed post-test 3D mapping of the final fracture allowed us to make an independent comparison of actual fracture and the fracture monitored by the AE and UT measurements. This comparison is essential for validating the interpretation of microseismic monitoring during hydraulic fracturing in the lab and in the field.