Carolyn A. Morrow

The effect of mineral bond strength and adsorbed water on fault gouge frictional strength

Recent studies suggest that the tendency of many fault gouge minerals to take on adsorbed or interlayer water may strongly influence their frictional strength. To test this hypothesis, triaxial sliding experiments were conducted on 1 5 different single-mineral gouges with various water-adsorbing affinities. Vacuum dried samples were sheared at 100 MPa, then saturated with water and sheared farther to compare dry and wet strengths. The coefficients of friction, µ, for the dry sheet- structure minerals (0.2-0.8), were related to mineral bond strength, and dropped 20-60% with the addition of water. For non-adsorbing minerals (µ =0.6-0.8), the strength remained unchanged after saturation. These results confirm that the ability of minerals to adsorb various amounts of water is related to their relative frictional strengths, and may explain the anomalously low strength of certain natural fault gouges.

Low strength of deep San Andreas fault gouge from SAFOD core

The San Andreas fault accommodates 28–34 mm yr−1 of right lateral motion of the Pacific crustal plate northwestward past the North American plate. In California, the fault is composed of two distinct locked segments that have produced great earthquakes in historical times, separated by a 150-km-long creeping zone. The San Andreas Fault Observatory at Depth (SAFOD) is a scientific borehole located northwest of Parkfield, California, near the southern end of the creeping zone. Core was recovered from across the actively deforming San Andreas fault at a vertical depth of 2.7 km (ref. 1). Here we report laboratory strength measurements of these fault core materials at in situ conditions, demonstrating that at this locality and this depth the San Andreas fault is profoundly weak (coefficient of friction, 0.15) owing to the presence of the smectite clay mineral saponite, which is one of the weakest phyllosilicates known. This Mg-rich clay is the low-temperature product of metasomatic reactions between the quartzofeldspathic wall rocks and serpentinite blocks in the fault. These findings provide strong evidence that deformation of the mechanically unusual creeping portions of the San Andreas fault system is controlled by the presence of weak minerals rather than by high fluid pressure or other proposed mechanisms. The combination of these measurements of fault core strength with borehole observations yields a self-consistent picture of the stress state of the San Andreas fault at the SAFOD site, in which the fault is intrinsically weak in an otherwise strong crust.

Permeability reduction in granite under hydrothermal conditions

The formation of impermeable fault seals between earthquake events is a feature of many models of earthquake generation, suggesting that earthquake recurrence may depend in part on the rate of permeability reduction of fault zone materials under hydrothermal conditions. In this study, permeability measurements were conducted on intact, fractured, and gouge-bearing Westerly granite at an effective pressure of 50 MPa and at temperatures from 150° to 500°C, simulating conditions in the earthquake-generating portions of fault zones. Pore fluids were cycled back and forth under a 2 MPa pressure differential for periods of up to 40 days. Permeability of the granite decreased with time t, following the exponential relation k=c(10−rt). For intact samples run between 250° and 500°C the time constant for permeability decrease r was proportional to temperature and ranged between 0.001 and 0.1 days−1 (i.e., between 0.4 and 40 decades year−1 loss of permeability). Values of r for the lower-temperature experiments differed little from the 250°C runs. In contrast, prefractured samples showed higher rates of permeability decrease at a given temperature. The surfaces of the fractured samples showed evidence of dissolution and mineral growth that increased in abundance with both temperature and time. The experimentally grown mineral assemblages varied with temperature and were consistent with a rock-dominated hydrothermal system. As such mineral deposits progressively seal the fractured samples, their rates of permeability decrease approach the rates for intact rocks at the same temperature. These results place constraints on models of precipitation sealing and suggest that fault rocks may seal at a rate consistent with earthquake recurrence intervals of typical fault zones.

Experimental studies of compaction and dilatancy during frictional sliding on faults containing gouge

Transient strength changes are observed in fault gouge materials when the velocity of shearing is varied. A transient stress peak is produced when the strain rate in the gouge is suddenly increased, whereas a transient stress drop results from a sudden change to a slower strain rate. We have studied the mechanism responsible for these observations by performing frictional sliding experiments on sawcut granite samples filled with a layer of several different fault gouge types. Changes in pore volume and strength were monitored as the sliding velocity alternated between fast and slow rates. Pore volume increased at the faster strain rate, indicating a dilation of the gouge layer, whereas volume decreased at the slower rate indicating compaction. These results verify that gouge dilation is a function of strain rate. Pore volume changed until an equilibrium void ratio of the granular material was reached for a particular rate of strain. Using arguments from soil mechanics, we find that the dense gouge was initially overconsolidated relative to the equilibrium level, whereas the loose gouge was initially underconsolidated relative to this level. Therefore, the transient stress behavior must be due to the overconsolidated state of the gouge at the new rate when the velocity is increased and to the underconsolidated state when the velocity is lowered. Time-dependent compaction was also shown to cause a transient stress response similar to the velocity-dependent behavior. This may be important in natural fault gouges as they become consolidated and stronger with time. In addition, the strain hardening of the gouge during shearing was found to be a function of velocity, rendering it difficult to quantify the change in equilibrium shear stress when velocity is varied under certain conditions.

Effects of lithology and depth on the permeability of core samples from the Kola and KTB drill holes

Permeability measurements were conducted on intact core samples from the Kola drill hole in Russia and the KTB drill hole in Germany. Samples included granodiorite gneisses, basalts and amphibolites from depths up to 11 km. The tests were intended to determine the pressure sensitivity of permeability and to compare the effects of stress relief and thermal microcracking on the matrix permeability of different rock types and similar samples from different depths. The pore pressure Pp was fixed at the estimated in situ pressure assuming a normal hydrostatic gradient; the confining pressure -Pc was varied to produce effective pressures (-Pe - -Pc- -Pp) of 5 to 300 MPa. The permeability of the basaltic samples was the lowest and most sensitive to pressure, ranging from 10 -2o to 10-23m 2 as effective pressure increased from 5 to only 60 MPa. In contrast, the granodiorite gneiss samples were more permeable and less sensitive to pressure, with permeability values ranging from 10 -l? to 10 -22 rn 2 as effective pressures increased to 300 MPa. Amphibolites displayed intermectiate behavior. There was an abundance of microfractures in the quartz-rich rocks, but a relative paucity of cracks in the mafic rocks, suggesting that the observed differences in permeability are based on rock type and depth, and that stress relief/thermal-cracking damage is correlated with quartz content. By applying the equivalent channel model of Walsh gcl Brace (1984) to the permeability data of the quartz-rich samples, we can estimate the closure pressure of the stress-relief cracks and thereby place bounds on the in situ effective pressure. This method may be useful for drill holes where the fluid pressure is not well constrained, such as at the Kola well. However, the use of crack closure to estimate in situ pressure was not appropriate for the basalt and amphibolite samples, because they are relatively crack-free in situ and remain so even after core retrieval. As a result, their permeability is near or below the measurable lower limit of our apparatus at the estimated in situ pressures of the rocks.

Permeability differences between surface-derived and deep drillhole core samples

Laboratory tests reveal that the permeabil- ity of samples obtained from deep boreholes is often lower and more sensitive to pressure than the perme- ability of common surface-derived crystalline rocks re- potled in the literature. We attribute the differences in permeability behavior to the fact that surface rocks have histories of unloading, weathering and retrograde metamorphism which are not comparable to that of the deeper rocks. Weathering products that line cracks and pores of surface rocks and make these openings more difficult to close as pressure increases may account for the relatively low pressure-sensitivity of permeability. Stress-relief cracking in the borehole samples can also reduce the pressure sensitivity. These results have im- portant implications for models that incorporate as- sumptions about the transport properties of rock at depth, such as models of heat transport or fluid pressure buildup, because many models are based on the prop- erties of common surface-derived rocks. Other physi- cal properties that are controlled by cracks and pores, such as seismic velocity and electrical resistivity, may be similarly affected by differences between surface-derived and deep rocks.

A note on the frictional strength of laumontite from Cajon Pass, California

Laumontite mineralization is pervasive in joints and shear zones encountered in the Cajon Pass drillhole in southern California. In order to determine whether a gouge composed of this hydrated mineral affects shear strength in a manner similar to low-strength, day-rich fault gouges, frictional sliding experiments were performed under dry, saturated and high pore pressure conditions at effective pressures up to 450 MPa. Coefficients of friction ranged between 0.66 and 0.84, consistent with most crustal rocks and well above the values typical of clay-rich San Andreas fault gouges. Saturation state had no effect on strength or sliding stability. These results suggest that the presence of laumontite in shear zones at Cajon Pass will not affect the shear strength of the rock in a way that can account for the inferred low ambient shear stresses.

VELOCITY- AND TIME-DEPENDENT STRESS TRANSIENTS IN SIMULATED FAULT GOUGE

SYNOPSIS During the shearing of fault gouge, strength changes are often observed when the rate of strain is varied. A strain rate increase is characterized by a transient stress peak before reaching a final equilibrium value. Conversely, a stress drop and a gradual rise to the equilibrium strength accompany changes to slower rates. To study the phenomena, frictional sliding experiments were performed at normal stresses to 100 MPa on granite samples sawcut at 30° and filled with a layer of Ottawa sand. Pore water volume and shear stress were measured as the displacement rate along the inclined gouge layer was alternated between 2.2 × 10–3 and 2.2 × 10–5 mm/sec. Pore volume increased at the fast rate and decreased at the slower rate, verifying that gouge dilation is a function of strain rate. Pore volume changed until the critical void ratio of the granular material was reached for a particular rate of strain. Because of this, the degree of consolidation of the grains must also depend on rate. Soil mechanics studies have shown that densely packed sand typically exhibits a peak in shear stress before reaching a steady-state value, whereas the strength of loosely packed sand slowly rises to the final value, as we have also observed in these experiments. The dense sand is initially overconsolidated relative to the equilibrium level, whereas the loose sand is initially underconsolidated relative to this level. Therefore, the transient stress behavior must be due to the overconsolidated state of the gouge at the new rate when the velocity is increased and to the underconsolidated state when the velocity is lowered. Because time-dependent processes will also cause the gouge to compact, natural fault gouges may exhibit similar transient behavior as they become overconsolidated and stronger with time.

Low resistivity and permeability in actively deforming shear zones on the San Andreas Fault at SAFOD

The San Andreas Fault Observatory at Depth (SAFOD) scientific drill hole near Parkfield, California, crosses the San Andreas Fault at a depth of 2.7 km. Downhole measurements and analysis of core retrieved from Phase 3 drilling reveal two narrow, actively deforming zones of smectite-clay gouge within a roughly 200m wide fault damage zone of sandstones, siltstones, and mudstones. Here we report electrical resistivity and permeability measurements on core samples from all of these structural units at effective confining pressures up to 120MPa. Electrical resistivity (~10Ω-m) and permeability (10 21 to 10 m) in the actively deforming zones were 1 to 2 orders of magnitude lower than the surrounding damage zone material, consistent with broader-scale observations from the downhole resistivity and seismic velocity logs. The higher porosity of the clay gouge, 2 to 8 times greater than that in the damage zone rocks, along with surface conduction were the principal factors contributing to the observed low resistivities. The high percentage of fine-grained clay in the deforming zones also greatly reduced permeability to values low enough to create a barrier to fluid flow across the fault. Together, resistivity and permeability data can be used to assess the hydrogeologic characteristics of the fault, key to understanding fault structure and strength. The low resistivities and strength measurements of the SAFOD core are consistent with observations of low resistivity clays that are often found in the principal slip zones of other active faults making resistivity logs a valuable tool for identifying these zones.

Changes in permeability and fluid chemistry of the Topopah Spring Member of the Paintbrush tuff (Nevada Test Site) when held in a temperature gradient: summary of results

The permeability and groundwater chemistry results for the Topopah Spring Member are reported and compared with the results from the previous work on Bullfrog. Permeability measurements made on samples of the Topopah Spring Member of the Paintbrush Tuff at room-temperature and in a temperature gradient show that the initially high (3-65 {mu}da) permeabilities are little affected by heating to at least 150{sup 0}C. These permeability relationships are favvorable for the disposal of nuclear waste in this stuff in an unsaturated zone at the Nevada Test Site. The fluids discharged from the samples of tuff during the experiments are dilute, nearly neutral solutions that differ only slightly from the starting groundwater composition. 8 references, 10 figures, 5 tables.