9:00 AM - 9:15 AM
[SCG63-01] Effect of geothermal mining along the vertical fracture: Case study for seaward side of Japan Trench
Keywords:Fluid flow along normal fault, heat flow, Japan Trench
Higher heat flow values than expected for the seafloor age of the incoming Pacific plate have been observed between the outer rise and the Japan Trench. Overlapping the broad anomaly, local variations at a scale of several kilometers were detected through concentrated measurements along lines perpendicular to the trench.
Several horsts and grabens exist trending parallel to sub-parallel to the trench axis on the seaward side of Japan Trench. The broad high heat flow zone can be attributed to pore fluid circulation in a permeable layer developed within normal faults of the horsts and grabens. Park et al. (2021) report anomalously high helium isotope ratios in sediment pore water, suggesting fluid infiltration into the upper mantle and subsequent outflow through bend-faults seaward of Japan trench. In order to construct a testable hydrological model along the fault system, we attempt to obtain closely-spaced heat flow measurements with piston core sampling across these normal faults.
During the KH-20-8 and KH-22-6 cruises, we conducted dense heat flow measurements along the E-W transect crossing 3 normal faults (39deg20’N, 38deg53’N, and 38degN). The widths of fault scarps range 300 to 1000m, and the fault offsets range ~100 to 250m. Amazingly, we observed a local high heat flow anomalies at all 3 sites with their peak located near the base to the mid-slope of the fault scarps. In more details, the heat flow outside the fault scarp ranges 65-85 mW/m^2, and is 90-100 mW/m^2 on the fault scarp with its width probably narrower than 500m. Note that the ‘basal’ heat flow values themselves are higher than expected from the age of the Pacific Plate.
Using a commercial software COMSOL multiphysics, we conducted 2D-FEM numerical simulations for the pore fluid flow and the thermal structure. We assume that the driving force for the flow is the geothermal gradient in the formation only. Since the surface heat flow can be affected by the geometry (fault scarp), we compared cases with and without surface undulation. We set higher permeability in the fault zone (200m wide and 2km deep) and a lower permeability for the surface sediment (300m thick).
For a uniform permeability with surface geometrical undulation, the surface heat flow varies by ~80% of the basal heat flow, but the calculated envelope does not match the observation. For the permeable fault with impermeable sediment cap model without geometrical change, a certain fluid flow occurs dirven by a given basal heat flow (50 mW/m^2) and the surface heat varies from 40 mW/m^2 next to the fault to 120 mW/m^2 within the fault zone. The model with both the permeable fault and the surface geometry may overestimate the heat flow contrast between the hanging wall (low heat flow due to downward fluid flow ) and foot wall sides (high heat flow due to upward fluid flow). We noticed that the surface heat flow also depends on the elapsed time since the onset of faulting (i.e. onset of fluid flow).
Further study is necessary. Still, we confirm that the heat and fluid mining from the base of the simulated system (2km deep) can occur with the permeable fault zone, impermeable sediment cap and the fault escarpment.
Several horsts and grabens exist trending parallel to sub-parallel to the trench axis on the seaward side of Japan Trench. The broad high heat flow zone can be attributed to pore fluid circulation in a permeable layer developed within normal faults of the horsts and grabens. Park et al. (2021) report anomalously high helium isotope ratios in sediment pore water, suggesting fluid infiltration into the upper mantle and subsequent outflow through bend-faults seaward of Japan trench. In order to construct a testable hydrological model along the fault system, we attempt to obtain closely-spaced heat flow measurements with piston core sampling across these normal faults.
During the KH-20-8 and KH-22-6 cruises, we conducted dense heat flow measurements along the E-W transect crossing 3 normal faults (39deg20’N, 38deg53’N, and 38degN). The widths of fault scarps range 300 to 1000m, and the fault offsets range ~100 to 250m. Amazingly, we observed a local high heat flow anomalies at all 3 sites with their peak located near the base to the mid-slope of the fault scarps. In more details, the heat flow outside the fault scarp ranges 65-85 mW/m^2, and is 90-100 mW/m^2 on the fault scarp with its width probably narrower than 500m. Note that the ‘basal’ heat flow values themselves are higher than expected from the age of the Pacific Plate.
Using a commercial software COMSOL multiphysics, we conducted 2D-FEM numerical simulations for the pore fluid flow and the thermal structure. We assume that the driving force for the flow is the geothermal gradient in the formation only. Since the surface heat flow can be affected by the geometry (fault scarp), we compared cases with and without surface undulation. We set higher permeability in the fault zone (200m wide and 2km deep) and a lower permeability for the surface sediment (300m thick).
For a uniform permeability with surface geometrical undulation, the surface heat flow varies by ~80% of the basal heat flow, but the calculated envelope does not match the observation. For the permeable fault with impermeable sediment cap model without geometrical change, a certain fluid flow occurs dirven by a given basal heat flow (50 mW/m^2) and the surface heat varies from 40 mW/m^2 next to the fault to 120 mW/m^2 within the fault zone. The model with both the permeable fault and the surface geometry may overestimate the heat flow contrast between the hanging wall (low heat flow due to downward fluid flow ) and foot wall sides (high heat flow due to upward fluid flow). We noticed that the surface heat flow also depends on the elapsed time since the onset of faulting (i.e. onset of fluid flow).
Further study is necessary. Still, we confirm that the heat and fluid mining from the base of the simulated system (2km deep) can occur with the permeable fault zone, impermeable sediment cap and the fault escarpment.