5:15 PM - 6:30 PM
[SSS08-P02] Experimental study on fluid injection-induced fault slip
Keywords:Induced seismicity, Strain, Acoustic Emission (AE)
In recent years, relatively large-scale earthquakes have occurred frequently in the vicinity of shale gas development sites, such as in North America. The earthquakes cause anxiety among residents and companies, and it becomes obstacles to safety subsurface development. As a cause of these induced seismicity, fluid injection into the subsurface formation possibly stimulates pre-existing faults and cause induced earthquakes (e.g., Ellsworth, 2013). When the pore pressure increases along the pre-existing fault due to fluid injection, the effective normal stress decreases. Then, the balance between friction and shear stresses collapses and the fault slides, leading to an earthquake. Numerical studies focus on reproducing observations and did not consider the mechanisms governing the frequency and magnitude of induced earthquakes. The models are often too complex and do not provide a detailed analysis of the basic processes leading to induced earthquakes. Therefore, it is necessary to consider the elementary processes of the injection-induced fault slip, such as the time dependence of the slip distribution and the change of the magnitude according to the stress conditions based on laboratory experiments.
In this study, we conducted the laboratory experiment to clarify the detailed process of fault slip caused by injection using an originally designed apparatus and large specimen.
The specimen used in this study is a 60 cm cube of granite with a 60 cm × 85 cm pre-existing fault on the diagonal. The surface of the fault plane was polished. We drilled the injection well in the center of the fault and set up a water injection point to apply pore pressure to the fault surfaces. Flat jacks are used to load differential stress to the specimen in two directions. In this way, we can load a differential stress, which is the driving force for fault slip, on the fault plane. In this condition, we inject water into the fault plane to reproduce the condition of fault slip caused by injection according to the Moll-Coulomb criterion. Thirty-nine strain gages were embedded in the fault surface to measure the shear strain along the fault plane. From this shear strain data, we can track the time variation of shear stress on the fault plane. In addition, a total of 12 AE sensors were installed on the specimen surface for measurement for micro-elastic waves (AE) generated by the fault slip.
The preliminary results show that the shear stress on the fault plane near the injection point decreases significantly with a rapid increase in the injection pressure. We interpret this drastic decrease of shear stress as a result of fault slip caused by injection. The shear stress on the fault plane decreased slowly during the gradual increase of the water injection pressure. This may correspond to fault slip without seismic waves (aseismic slip), as reported by Guglielmi et al. (2015).
We plan to study the slip behavior in more detail by combining the strain behavior with the information obtained from AE analysis, such as the hypocenter (slip position), magnitude, and their injection-pressure dependence.
In this presentation, we report the results obtained at present.
In this study, we conducted the laboratory experiment to clarify the detailed process of fault slip caused by injection using an originally designed apparatus and large specimen.
The specimen used in this study is a 60 cm cube of granite with a 60 cm × 85 cm pre-existing fault on the diagonal. The surface of the fault plane was polished. We drilled the injection well in the center of the fault and set up a water injection point to apply pore pressure to the fault surfaces. Flat jacks are used to load differential stress to the specimen in two directions. In this way, we can load a differential stress, which is the driving force for fault slip, on the fault plane. In this condition, we inject water into the fault plane to reproduce the condition of fault slip caused by injection according to the Moll-Coulomb criterion. Thirty-nine strain gages were embedded in the fault surface to measure the shear strain along the fault plane. From this shear strain data, we can track the time variation of shear stress on the fault plane. In addition, a total of 12 AE sensors were installed on the specimen surface for measurement for micro-elastic waves (AE) generated by the fault slip.
The preliminary results show that the shear stress on the fault plane near the injection point decreases significantly with a rapid increase in the injection pressure. We interpret this drastic decrease of shear stress as a result of fault slip caused by injection. The shear stress on the fault plane decreased slowly during the gradual increase of the water injection pressure. This may correspond to fault slip without seismic waves (aseismic slip), as reported by Guglielmi et al. (2015).
We plan to study the slip behavior in more detail by combining the strain behavior with the information obtained from AE analysis, such as the hypocenter (slip position), magnitude, and their injection-pressure dependence.
In this presentation, we report the results obtained at present.