Description: Hydraulic fracturing is a widely used technique applied in unconventional reservoirs to generate large fracture networks. Interactions between hydraulic fracture (HF) and natural fracture (NF) can impact the fracture topology and thus the subsequent productivity. Despite a large number of studies on HF–NF interactions, the HF propagation path is normally judged based on ad-hoc criteria to decide whether crossing or deflection occurs and the mechanism behind has not yet reached a unified understanding. Here, we use a phase-field model (PFM), which is based on a unified fracture propagation criterion, to investigate the influence of in-situ stress, fracturing operational parameters and NF orientation and strength. We analyze the mechanism behind different propagation patterns resulting from different kinds of NFs—non-cemented and cemented ones under different conditions. In particular, we compare the total energies between the symmetric propagation and asymmetric propagation to verify the minimum energy propagation path. Our results indicate that a higher stress anisotropy more likely leads to HF–NF crossing and a less fracture complexity. Injection rate influences propagation speed and fracture complexity. Within a certain range (30°, 45°, 60° in this study), the larger the approaching angle is, the more complex the fractures become. With the increasing strength contrast between NF and rock matrix, the material heterogeneity increases, encouraging HF to form complex fractures. Opening more strongly cemented NFs, which act as a barrier for propagation, consumes more energy than HF propagation outside the interface. Lower stress anisotropy and higher injection rate lead to higher initiation pressure, requiring more energy for propagation.

Description: The accurate evaluation of pore pressure and injected volume is crucial for the laboratory characterization of hydromechanical responses of rock fractures. This study reports a series of laboratory experiments to systematically demonstrate the effects of external temperature and dead volume on laboratory measurements of pore pressure and injected volume in a rock fracture. We characterize the hydraulic aperture of the fracture as a function of effective normal stress using the exponential aperture model. This model is then employed to predict the pore pressure change and injected volume in the fracture without the influences of external temperature and dead volume. The external temperature changes in the cyclic loading test due to the Joule-Thompson effect for fluids. The effect of external temperature on pore pressure change in the fracture can be well explained by thermal pressurization of fluids. Our results also show that the external dead volume can significantly lower the pore pressure change in the fracture during the cyclic loading test under undrained conditions. The injected volume can also be substantially enlarged due to the external dead volume in a typical pore pressure system. Internal measurement of the pore pressure in the fracture using a fiber optic sensor cannot exclude the influences of external temperature and dead volume, primarily because of the good hydraulic communication between the fracture and pore pressure system. This study suggests that the effects of external temperature and dead volume on pore pressure response and injected volume should be evaluated for accurate laboratory characterization and inter-laboratory comparison.

Description: Induced seismicity associated with fluid injection into underground formations jeopardizes the sustainable utilization of the subsurface. Understanding the fault behavior is the key to successful management and mitigation of injection-induced seismic risks. As a fundamental approach, laboratory experiments have been extensively conducted to assist constraining the processes that lead to and sustain various injection-induced fault slip modes. Here, we present a state-of-the-art review on the emerging topic of injection-induced seismicity from the laboratory perspective. The basics of fault behavior, including fault strength and instability, are first briefly summarized, followed by the paradoxical stability analysis arising from the current theoretical framework. After the description of common laboratory methods and auxiliary techniques, we then comprehensively review the effects of fault properties, stress state, temperature, fluid physics, fluid chemistry and injection protocol on fault behavior with particular focus on the implications for injection-induced seismicity. We find that most of the shear tests are conducted under displacement-driven conditions, while the number of injection-driven shear tests is comparatively limited. The review shows that the previous work on displacement-driven rock friction and fault slip modes partially unravel the mystery of injection-induced fault behavior, and recent experimental studies on the injection-driven response of critically stressed faults provide complementary insights. Overall, laboratory experiments have substantially advanced especially our understanding of the roles of fault roughness, fault mineralogy, stress state, fluid viscosity, fluid induced mineral dissolution, and injection rate in injection-induced seismicity, which has been successfully used to interpret many field observations. However, there are still outstanding questions in this area, which could be addressed by future experimental studies, such as the feasibility of seismic-informed adaptive injection strategy for mitigating seismic risks, cold fluid injection into critically stressed faults under hydrothermal conditions, and fault friction evolution during cyclic injection spanning from undrained to drained conditions.