Acoustic monitoring of laboratory hydraulic fracture growth under stress and pore pressure

Fluid-driven fracturing is used in a wide range of applications, including oil and gas extraction, geothermal energy recovery, and CO2 sequestration. In order to efficiently fracture the targeted rock formation, theoretical models provide estimates of the fracture size and shape. Carrying properly scaled laboratory experiments, on the other hand, allows to validate theoretical predictions by providing complete datasets of individual experiments performed under controlled conditions, and therefore to better understand the physics of fluid-driven fracturing. The DelFrac Consortium at TU Delft pioneered this field by building an acoustic monitoring setup inside a triaxial press applying three independent stresses on a cubic specimen (Groenenboom, 1998). At the Geo-Energy Lab, we intend to further investigate the solid and fluid mechanics of hydraulic fractures by building a novel experimental setup in our EPFL facility. The fracturing setup will consist in a triaxial frame designed to accommodate cubic-shaped specimens of up to 250 mm in length, and to apply up to 20 MPa independently on each axis. We will also have the ability to pressurize a pore fluid up to 5 MPa inside the frame in order to simulate in-situ conditions. Our current high-pressure pump can inject fluids with a maximum pressure of 51 MPa and a flow rate ranging from 1 μL/min to 90 mL/min through a high-pressure line and a cased wellbore inside the specimen. We will monitor the growth of the fracture with a combination of compressional and shear piezoelectric transducers for a total of 64 units, that can be used to both generate and measure acoustic energy. We use a function generator connected to a high-power amplifier to generate an excitation signal, which is then routed to one of 32 excitation transducers through a multiplexer. The other 32 transducers are connected to a high-speed acquisition board in order to simultaneously record the measured ultrasonic signals. By carefully placing the transducers on all six faces of the specimen to be fractured, we expect to record transmitted, reflected and diffracted acoustic events (Figure 1). We intend to use these three types of event in order to estimate the spatial extent of the fracture, as well as its thickness along raypaths. Transmission measurements, where the wave travels across the specimen, provide fracture thickness information (Groenenboom & Fokkema, 1998; Kovalyshen et al., 2014). Diffraction events, where the wave propagates to the tip of the fracture and then towards the side of the block, carry information about the tip position, and in turn give an estimate of the fracture size (Groenenboom et al., 2001). Reflection events, where the acoustic energy is reflected back toward the same side of the block, let us discriminate between dry and fluid-filled fracture in the case where a fluid lag is present at the fracture tip.