No description available.

In nature, electrical signals have been recorded contemporaneously with volcanic and seismic activity, and have been proposed as precursors to earthquakes and volcanic eruptions. In the hydrocarbon industry, streaming potentials are used to investigate steam fronts, thus aiding enhanced oil recovery. There is therefore considerable current interest in electrical signals emanating from the Earth's crust and the mechanisms which give rise to them. Two of the theories that have been proposed to explain electrical signal generation are: The piezoelectric effect, caused by stress changes on piezoelectric minerals, such as quartz, which is found in many crustal rocks. The electrokinetic phenomenon, produced at a solid-liquid interface, where an electrokinetic current such as the streaming potential can be induced through a pressure, chemical or temperature gradient, resulting in electrical charge transport within the moving fluid. In order to investigate the possible mechanisms responsible for the generation of electrical signals in the Earth's crust, carefully controlled laboratory rock deformation and rock physics experiments have been performed under simulated shallow crustal conditions, where both electrical potential signals and acoustic emissions were measured. The deformation strain rate, confining pressure, pore fluid pressure, pore fluid chemistry and temperature were all varied systematically during conventional triaxial rock deformation tests on a range of rock types. Confining pressures were varied from 20 MPa to 100 MPa, pore fluid pressures from 5 MPa to 40 MPa, strain rates from 1.5 x 10"4 s"1 to 1.5 x 10"7 s"1 and temperatures from room temperature (25 C) up to 125 C. Over thirty five experiments were completed at room temperature on rock samples Clashach, Bentheim and Darley Dale sandstones and Portland limestone. More than ten experiments were done at elevated temperature on both dry and saturated samples of Clashach sandstone using a range of pore fluid chemistries. Significant developments in experimental apparatus were necessary for these latter experiments, including the design and construction of an electrical internal heater for the triaxial deformation cell. I identify that, for the temperature range between 25 and 125 C, that the primary sources of electrical potential signal generation are (i) piezoelectric in dry quartz-rich sandstone and (ii) electrokinetic in saturated samples of both sandstones and limestone. Factors that are found to influence the electrical potential signals during deformation include effective pressure, temperature, strain rate, pore fluid type and fluid flow. As failure is approached, both pre-seismic and co-seismic signals are observed with the magnitude of the signals varying with rock type. These observations can be explained by differences in the rock composition and variation in hydraulic and electrical pathways available for electric current flow during rock deformation. Variation in electrical potential difference can be seen during both the compactive and dilatant stages of deformation. At slower strain rates, local rock variation can be seen through changes in electrical potential signals which appear to be obscured at higher strain rates. The change in electrical and streaming potential signals during deformation reflect both the accumulating and accelerating damage identified by acoustic emission prior to fracture and the localisation of damage at dynamic fracture. After failure the potential decreases to a background value where it remains during constant frictional sliding at essentially constant stress. The presence of a crack or fault was identified to affect the electrical and streaming potential signals depending on their relative position with respect to the fault suggesting that electrical potential could be used as a method for fault location. An increase in temperature was found not to affect the mechanical properties within the range of experimental conditions explored. The effect of increased temperature on the electrical potential signals depends on the conditions applied to the rock such as thermal equilibrium times, deformation and ionic species within the pore fluid. If the rock is allowed to reach thermal equilibrium, the electrokinetic reactions between the solid-liquid interface are increased with an average electrical potential increase of 38 mV per 25 C. However, the new surfaces formed during rapid deformation cannot reach equilibrium, so that temperature has no effect on the electrical potential signals during compaction, dilatancy & failure. With this, the results as a whole suggest that in shallow crustal rocks, the change in electrical potential signal with temperature is below the background electrical noise level.

No description available.

No description available.

No description available.

No description available.

The geometric and spatial heterogeneity in fracture systems of natural fault zones reflect complex strain patterns and exert substantial influences on host rock properties. It has been recognized that general 3D strains will produce diffuse fracture orientation patterns and characteristic kinematics, which are proposed to vary towards the fault core. The bootstrapping methods and spatial correlation analysis [(semi-) variograms] were adopted to investigate the geometric and spatial hetero-geneity in fracture orientations collected systematically with reference to their spatial locations from three distinct fault zones. This relatively rigorous approach revealed that fracture orientation patterns display systematic spatial variations and high spatial correlations traversing fault zones. Factors related to the presence of pre-existing structures and lithologies can modify strains, creating complex fracture patterns and kinematics at different scales and spatial locations. The results suggest that spatially heterogenous fracture networks in subsurface can be highly connected as channels or barriers and will, in anyways, affect the fluid flow path in aquifers.

No description available.

This thesis is an experimental study of brittle rock fracture and frictional slip on preexisting faults, both of which are important in improving understanding of earthquake nucleation. The experiments featured Westerly granite samples under triaxial compression in which, full waveform Acoustic Emission (AE) data was continuously recorded. AE locations and source mechanism studies provide a non-invasive tool to resolve the temporal and spatial evolution of micro-fracturing, and understand the micro-scale processes that are operating within a sample. New observations of brittle fracture are presented, in which the transition from stable nucleation to unstable, dynamic fracture is demonstrated to be a three stage process consisting of. (1) the development of a fracture nucleus; (2) the sudden fracture propagation at a speed of 10's of mm/s; and (3) acceleration of the fracture to cause rupture. A consideration of fault length and time-to-rupture demonstrates that the fracture velocity must accelerate to an average rate of a few m/s in the 20-100 ms prior to rupture. Furthermore, by directly measuring the duration of the dynamic, high speed fracture propagation phase, a fracture speed of 175 m/s is inferred for the 0.6 ms prior to rupture. Source mechanism studies, b-values and velocity data are presented to further demonstrate the evolution of fracture. This research demonstrates similarities between fracture nucleation in intact rock, and the previous observations of frictional slip nucleation that were used to model earthquake nucleation. Laboratory stick slip events are thought to share the same mechanism as earthquakes, and so experiments were conducted to investigate the nucleation of stick slip events, and their premonitory AE. Stick slip characteristics are compared for a saw-cut (artificial) and a rough (naturally fractured) fault. As expected, the frictional coefficient for slip events on the smooth fault was lower (between 0.48 and 0.59) than the natural fault (0.72) and the number of premonitory AE were fewer (<120 compared to >3000). These AE were located, and source mechanisms and b-values calculated. For both fault surfaces, the first motion of each stick slip was recorded as a large-amplitude AE and was source located onto the fault surface. These represent the nucleation sites of the stick-slip events. The nucleation sites varied and were probably controlled by heterogeneity of stress or surface conditions on the fault. There is a close similarity between the seismic signatures of slip for both tests, appearing as impulsive events with no AE in the microseconds prior to slip. This contrasts with the gradual increase in amplitude observed for the fracture initiation tests, explained by the development of a process zone. The ability to identify stick-slip nucleation sites, and define the evolution of unstable fracture for intact rock fracture, has been demonstrated for the first time in this thesis, and has been used to improve the understanding of nucleation processes.

No description available.