5:15 PM - 6:45 PM
[SCG40-P40] Laboratory crack-sealing, silica vein formation and permeability evolution driven by fluid pressure drop in the seismogenic depths
Permeability evolution of crust is important for understanding seismic cycles and is governed by cracking and its sealing with mineral vein formation (i.e., fault valve model; Sibson 1992). Various vein formation models have been proposed in nature, including the models associated with fluid pressure drop, however, there is little experimental validation of the vein formation processes and their time scales. Recent experimental studies have shown that silica precipitates upon fluid pressure drop (Amagai et al., 2019). Meanwhile, there are no examples of silica precipitation and formation of veins leading to closure of cracks. Here we reproduce crack sealing behavior by generating cracks and precipitates silica to form veins within the rock.
Hydrothermal flow-through experiments were conducted under a high confining pressure (Pc ) relevant to seismogenic depths (i.e., ≧100 MPa). Samples of Aji granite (diameter: 6 mm height: 12 mm) were introduced for thermal cracks at 600℃ with porosities of 1.1-2.8%. Experimental conditions were set at 200°℃ with a Pc of 100 MPa and an upstream pressure (Pup) of 50 MPa, and two downstream pressures (Pdown) of 0.1 MPa and 8 MPa. 300℃ experiments were set with two Pc of 100 MPa and 75 MPa, a Pup of 50 MPa, and a Pdown of 0.1 MPa. Temporal changes in the permeability of the samples were measured during the experiments, and the samples were analyzed by XCT, SEM, and EPMA.
In the 200℃ experiments, quartz and feldspars were mainly dissolved upstream under both conditions, while silica was not precipitated when Pdown = 8 MPa. Reticulate amorphous silica precipitated on the downstream surface in the experiment at Pdown = 0.1 MPa, and the permeability was constant at 1-5 × 10-19 m2 in both cases. On the other hand, at 300℃, the permeability decreased from 10-18-10-19 to 10-22-10-20 m2. Under both conditions, quartz and feldspars dissolved upstream, and chimney-like amorphous silica precipitated locally on the downstream surface. Furthermore, at the Pc = 100 MPa experiment, amorphous silica veins were formed in the interior of the sample near the downstream surface perpendicular to the fluid flow direction.
Although the permeability decreased significantly at 300℃, the precipitation expected from the solubility was 15 mg and 3.1 mg for the 200℃ and 300℃ 4-day experiments, respectively. These results suggest that the decrease in permeability cannot be explained by the amount of precipitated SiO2 alone, but likely be explained that reticulate precipitates at 200℃ did not seal, whereas the chimney-like precipitates at 300℃ contributed to the sealing.
At 300℃ and under high Pc, the amorphous silica veins were about 10-20 µm wide and were connected at both ends to the chimney-like precipitates. In this experiment, about two increases and decreases in permeability were observed, which may correspond to the formation of amorphous silica veins, suggesting the following sealing processes. First, the rock dissolved upstream, and then the fluid pressure drop caused a phase change to gas phase, resulting in the formation of chimney-like precipitates on the surface, and reduced the permeability. This increased the fluid pressure near the downstream surface, and the large differential stress caused the cracking perpendicular to the flow path, again increasing the permeability. Similar precipitation-cracking behaviors repeated, which again caused an increase or decrease in the permeability.
The crack sealing process was successfully reproduced by fluid pressure drop under high Pc of 100 MPa and 300℃, which is equivalent to the conditions of the seismogenic zone. Although this experiment was conducted under an extremely high fluid pressure gradient for a long time, mineral veins with a width of several tens of µm width, similar to those observed in the seismogenic zone at 300℃, were generated in four days. The results of this experiment may provide a constraint on a time scale for the fault-valve model.
Hydrothermal flow-through experiments were conducted under a high confining pressure (Pc ) relevant to seismogenic depths (i.e., ≧100 MPa). Samples of Aji granite (diameter: 6 mm height: 12 mm) were introduced for thermal cracks at 600℃ with porosities of 1.1-2.8%. Experimental conditions were set at 200°℃ with a Pc of 100 MPa and an upstream pressure (Pup) of 50 MPa, and two downstream pressures (Pdown) of 0.1 MPa and 8 MPa. 300℃ experiments were set with two Pc of 100 MPa and 75 MPa, a Pup of 50 MPa, and a Pdown of 0.1 MPa. Temporal changes in the permeability of the samples were measured during the experiments, and the samples were analyzed by XCT, SEM, and EPMA.
In the 200℃ experiments, quartz and feldspars were mainly dissolved upstream under both conditions, while silica was not precipitated when Pdown = 8 MPa. Reticulate amorphous silica precipitated on the downstream surface in the experiment at Pdown = 0.1 MPa, and the permeability was constant at 1-5 × 10-19 m2 in both cases. On the other hand, at 300℃, the permeability decreased from 10-18-10-19 to 10-22-10-20 m2. Under both conditions, quartz and feldspars dissolved upstream, and chimney-like amorphous silica precipitated locally on the downstream surface. Furthermore, at the Pc = 100 MPa experiment, amorphous silica veins were formed in the interior of the sample near the downstream surface perpendicular to the fluid flow direction.
Although the permeability decreased significantly at 300℃, the precipitation expected from the solubility was 15 mg and 3.1 mg for the 200℃ and 300℃ 4-day experiments, respectively. These results suggest that the decrease in permeability cannot be explained by the amount of precipitated SiO2 alone, but likely be explained that reticulate precipitates at 200℃ did not seal, whereas the chimney-like precipitates at 300℃ contributed to the sealing.
At 300℃ and under high Pc, the amorphous silica veins were about 10-20 µm wide and were connected at both ends to the chimney-like precipitates. In this experiment, about two increases and decreases in permeability were observed, which may correspond to the formation of amorphous silica veins, suggesting the following sealing processes. First, the rock dissolved upstream, and then the fluid pressure drop caused a phase change to gas phase, resulting in the formation of chimney-like precipitates on the surface, and reduced the permeability. This increased the fluid pressure near the downstream surface, and the large differential stress caused the cracking perpendicular to the flow path, again increasing the permeability. Similar precipitation-cracking behaviors repeated, which again caused an increase or decrease in the permeability.
The crack sealing process was successfully reproduced by fluid pressure drop under high Pc of 100 MPa and 300℃, which is equivalent to the conditions of the seismogenic zone. Although this experiment was conducted under an extremely high fluid pressure gradient for a long time, mineral veins with a width of several tens of µm width, similar to those observed in the seismogenic zone at 300℃, were generated in four days. The results of this experiment may provide a constraint on a time scale for the fault-valve model.