Eric Tenthorey

Evolution of strength recovery and permeability during fluid–rock reaction in experimental fault zones

Abstract Physical and chemical fluid–rock interactions are implicated in controlling earthquake nucleation and recurrence. In particular, interseismic compaction, sealing and healing of fractured fault rocks can lead to strength recovery and stabilisation of fault zones. In contrast, these same processes can also assist increases in pore fluid pressures and consequent destabilisation of faults. Here, we present high-temperature, hydrothermal experiments designed to assess the evolution of strength of fault zones in previously intact rock, and also characterise the associated changes to porosity and permeability. Cores of Fontainebleau sandstone were initially loaded to failure in a high-pressure gas–medium apparatus. The failed specimens were then hydrothermally reacted at 927°C for variable duration under isostatic conditions, and subsequently re-fractured to determine the ‘interseismic’ strength recovery. In the most extreme case, hydrothermally induced gouge compaction, cementation and crack healing resulted in 75% strength recovery after reaction for 6 h. Isostatic hydrothermal treatment also resulted in dramatic reduction in porosity and permeability. Strength of the fault zone following hydrothermal reaction appears to be closely correlated to porosity, consistent with previous studies on brittle failure of porous aggregates. The experimental results show how hydrothermal reactions in fault zones may lead to two competing, time-dependent effects; fault strengthening due to increased cohesion in the fault zone and fault weakening arising from elevated pore pressures within a well cemented, low-permeability gouge layer.

Cohesive strengthening of fault zones during the interseismic period: An experimental study

[1]xa0There is widespread evidence indicating that faults regain a portion of their strength during the interseismic period. Here, we present experiments designed to understand and quantify the interseismic cohesive strengthening resulting from fluid-rock reactions in fault zones. The triaxial experiments consisted of fracturing cores of Fontainebleau sandstone under dry conditions, forming a localized shear failure zone (stage 1). The specimens were then reacted hydrothermally under isostatic conditions, allowing the fault damage zone to compact, consolidate and strengthen (stage 2). Following reaction, the specimens were then reloaded to failure under nominally dry conditions, so that the increase in cohesive strength of the fault could be measured (stage 3). Experiments show that cohesion increase is positively correlated to temperature and pore pressure during reaction. After 6 hours of reaction at the highest temperatures (927°C) and pore pressures (200 MPa), cohesion increases by as much as 35 MPa. Microstructural examination of the specimens showed that the gouge particles within the fault compacted and cemented together, exhibiting textures typical of pressure solution and that fractures in the surrounding damage zone had healed. A theoretical treatment of the data was conducted using these experiments in combination with results on time-dependent changes in fault cohesion presented by Tenthorey et al. (2003). We find that the rate-controlling process in our experiments has an activation energy (Q) of approximately 70 kJ mol−1. We use this information to develop a model for time-dependent cohesive strengthening in fault zones within the continental seismogenic regime. We conclude that significant cohesive strengthening of fault zones can occur during the interseismic period of medium to large earthquakes given the presence of reactive pore fluids.

Precipitation sealing and diagenesis: 1. Experimental results

During burial and diagenesis of granular aggregates, significant permeability reduction may be induced by the formation of low-temperature, authigenic minerals. To quantitatively assess the importance of this process, we have conducted a series of hydrothermal flow-through experiments using deionized water and labradorite/quartz sand. All experiments were conducted in a modified triaxial apparatus, configured to allow continuous permeability measurements. Under most of the conditions tested, significant permeability reduction is observed with no concurrent decrease in porosity. The overall permeability reduction sometimes exceeds 1 order of magnitude over 4 days and is positively correlated to temperature and deviatoric stress. Scanning electron microscope observations together with data from additional experiments show that the observed permeability reduction is entirely a result of secondary mineral growth. Si and Al concentrations in the postexperiment fluids are also correlated to temperature and stress, confirming the link between the chemical state of the system and permeability behavior. In all experiments, permeability reduction is fastest early and levels off in the late stages. To explain the permeability behavior as a function of time, a conceptual model is developed in which precipitation of authigenic minerals is rapid at early times while dissolution of quartz and labradorite is most active. As the system approaches equilibrium, the components necessary for secondary mineral formation are liberated at a lower rate, thereby causing precipitation to slow. Although authigenic mineral formation does not reduce total pore space in these experiments, there is a reduction in effective porosity, which results in permeability reduction.

Composition of fluids during serpentinite breakdown in subduction zones: Evidence for limited boron mobility

We experimentally investigated the trace element compositions of fluids released during breakdown of subducted serpentinites. Serpentinites contain significant amounts of fluid-mobile elements such as B, Cs, As, and Ba, which during seafloor alteration are incorporated into mantle rocks. During the later high-pressure breakdown of the serpentinites, these trace elements are redistributed among the residual olivine, orthopyroxene, minor chlorite, and fluid. We find that B is far more compatible in these minerals than previously assumed; it has a fluid/residue partition coefficient (F/R D ) of 3–5. Most other fluid-mobile elements (Cs, As, Ba, Pb) are strongly enriched in the fluid and exhibit expected F/R D values of 30–250. Serpentinites are possibly the most important sink for B in subduction zones; the experimental results suggest that significant B is recycled into the deep mantle. Furthermore, high B concentrations in mantle olivines might be a fingerprint for previous metasomatism or serpentinization.

Reaction-enhanced permeability during serpentinite dehydration

Fluid mobilization during prograde metamorphic reactions is a poorly understood, yet crucial, phenomenon that has implications for a number of geologic problems. Here we present the first experiments to characterize permeability evolution during dehydration of a natural rock. Dehydration of serpentinite specimens resulted in the generation of pore space and the rapid development of an interconnected pore network. During breakdown of antigorite to olivine + talc + H 2 O, permeability increased rapidly by at least three orders of magnitude. In natural systems, such an increase in permeability would be transitory, because high confining stresses would rapidly eliminate the porosity generated during reaction. Such transitory, reaction-enhanced permeability of serpentinite provides a mechanism by which fluids can migrate and possibly facilitate subduction-zone seismicity and contribute to partial melting of the mantle wedge. The results presented also have implications regarding middle- to deep-crustal fluid flow.

Permeability evolution in quartz fault gouges under hydrothermal conditions

[1]xa0The permeability (k) of fine-grained quartz aggregates were measured in situ during hot pressing (HPing) experiments to explore the evolution of fluid transport properties of fault zones during the interseismic period. Experiments were conducted at temperatures of 150°C and between 700 and 850°C, with confining and pore water pressures of 250 and 150 MPa, respectively. Significant permeability reduction was observed between 700 and 850°C, with permeability reduction rates (r = (1/t) ln (kto/kt)), ranging from approximately 6 × 10−5 s−1 at 700°C to a maximum of approximately 7.4 × 10−4 s−1 at 850°C. Permeability decreased exponentially with time, and the permeability reduction rate increased with increasing temperature, increasing differential stress, and decreasing grain size. Analysis of the permeability-porosity relationships indicates that permeability in the simulated gouge at high temperature shuts off at a critical porosity of 0.045 ± 0.004. The presence of microstructures, such as grain interpenetration, grain shape truncation, arrays of fluid inclusions, and development of quartz overgrowths on grains, indicate that k reduction was controlled by dissolution-precipitation creep processes. Extrapolation of the permeability reduction rates, measured in this study, to temperatures typical of the continental seismogenic regime highlights the strongly time-dependent nature of permeability in natural fault wear products at depths of nucleation of major earthquakes. Within the recurrence time of large earthquakes, quartz-rich fault zones in the fluid-active midcrustal to lower continental crustal regimes can evolve from high-permeability conduits to low-permeability seals. Episodic changes in the fluid transport properties of faults during the interseismic period are likely to impact on the pore pressure evolution of fault wear products.

Mapping secondary mineral formation in porous media using heavy metal tracers

[1]xa0A new technique is presented which allows easy identification of secondary minerals, formed during experimental diagenesis. Reacting fluids are doped with Ba and Sr; heavy metals which substitute into the alkali-bearing authigenic minerals that precipitate because of alteration of primary labradorite and quartz. The main secondary phases observed are a Ba-rich zeolite mineral, strontianite, and a Ba-rich calcite. When polished sections of postexperiment specimens are viewed using scanning electron microscope backscatter imaging, secondary phases appear as bright patches due to their higher density and average atomic number. The more difficult method of X-ray elemental mapping confirms that these zones represent secondary minerals rich in Ba and Sr. This technique allows the spatial distribution of secondary minerals to be mapped out at the specimen scale in a short time frame. Image analysis shows that observed reductions in permeability are caused primarily by mineral precipitation in narrow pore throats and in areas containing an abundance of fine-grained material. Secondary mineral formation is concentrated near the top of the sample and is interpreted to migrate through the sample as a reaction front. Mineral coverage is then quantified and used to explain the observed permeability evolution.