EXPERIMENTAL STUDY ON FAILURE PROCESS OF TIGHT SANDSTONES AND SHALES BASED ON STRAIN ENERGY AND KEY STRESS THRESHOLDS

Rock failure research has been conducted extensively in recent years due to its crucial role in the production enhancement of unconventional resources. This dissertation focuses on quantifying rock failure process for shales and tight sandstones through laboratory experiments. We have proposed several parameters that can be used to describe rock failure process and crack development better. Brittleness is widely used for drilling optimization and reservoir stimulation. However, it is not always a reliable parameter for designing hydraulic fracturing. More effective methods are needed to assess the fracability of tight sandstones and shales. We have performed triaxial compression experiments to rupture eleven tight sandstone and shale samples. The samples show systematic behavior from the measured stress-strain curves. The strain energy method is developed to characterize rock failure process quantitatively. We find that the brittle failure exhibits a higher percentage of elastic strain energy and residual strain energy, but a lower percentage of dissipated strain energy than the ductile failure. We also use key stress thresholds, including crack initiation stress (Ci), crack damage stress (Cd), and compressive strength (Cp), to explore progressive failure stages using the experimental data of the measured samples. The results show positive relationships between key stress thresholds and Young’s modulus. The ratios of Ci/Cp and Cd/Cp can better describe the ductile and brittle damage modes for tight rocks with different crack development processes. Both ratios can also evaluate rock damage behavior before and after failure. Another major component of this research is to study rock failure process by carrying out triaxial failure tests on 28 organic shale samples. We develop the strain energy evolution approach to detect crack initiation stress. A low Ci/Cp indicates that the organic shale samples are relatively ductile with slow crack growth, consistent with a long crack development process, which leads to complicated microfracture distribution. It reflects that conventional brittleness is not relevant to fracture development for the measured samples. In addition, crack initiation stress is introduced to determine elastic stage and predict in-situ stress.