9:45 AM - 10:00 AM
[HSC05-04] Coupled chemical osmosis and rock deformation: Numerical investigations and their comparison with experimental results
Keywords:chemical osmosis, poroelasticity, mudstone, seal layer
Groundwater flow and solute transport in mudstone is generally considered to be very slow, and hence, mudstone formations are expected to perform as effective seal layers for carbon capture and storage (CCS) project. In the low-permeable, muddy formations, coupling processes such as chemical osmosis and poroelastic behavior can affect groundwater flow, solute transport, and rock deformation. Some mudstones are known to behave as semipermeable membranes. This is because the negative charges on the surface of clay minerals retard the transport of charged particles in narrow pores. In such mudstones, chemo-osmotic flow occurs by the concentration gradient (e.g., Marine and Fritz, 1981), and in some cases, pressure buildup can reach to about 20 MPa (Neuzil, 2000). In addition, pore pressure change may induce the deformation of porous medium. The purpose of this research is to evaluate the effect of chemical osmosis on the deformation and pressure behavior in mudstone based on both laboratory experiments and numerical simulations.
In the laboratory experiments, core samples with 50 mm diameter and 30 mm height were prepared from the mudstone of the Oganomachi Formation, collected from Saitama prefecture. The core samples were kept in NaCl solution with the same total dissolved ion concentrations of pore water (less than 0.1 g/L) under hydrostatic pressure at the beginning of the experiments. The lateral and the upper surfaces were sealed by silicon rubber while the lower surface was in contact with the NaCl solution of which concentration was kept at about 10 g/L. The circumferential strains of the sample were measured by the optical fiber strain sensor attached helically on the lateral surface of the sample. The measurements were continued for more than 200 hours and the temperature was controlled almost constant throughout the experiments.
A numerical model to calculate chemo-osmotic and poroelastic processes was used to compare numerical simulation results with the experimental ones. The governing equations were based on a set of equations describing groundwater flow and diffusion including chemical osmosis process (Malusis et al., 2012) and those describing poroelastic effect (e.g., Wang, 2000), and pore pressure, concentration, and rock deformation were calculated. The calculated domain, boundary conditions and initial conditions were set to be the same as those of the laboratory experiments, for example, calculated domain was a cylinder with 50mm in diameter and 30mm in height, the upper and lateral surfaces were set to be no-flow and no solute flux boundary, and the lower surface was set to be constant pressure (0 kPa) and constant concentration (about 10 g/L). All the surfaces were set to be strain-free boundaries.
In the laboratory experiments, the continuous contraction was observed, and was interpreted to be caused by the osmotic flow and induced deformation by pressure decrease in the sample. The amount of contraction near the bottom was larger than that near the upper surface. The maximum 130 με was measured near the lower surface.
The effects of the parameters variation on the strain behavior were evaluated through numerical simulation, and the appropriateness of the model to explain the overall behavior of the samples was discussed by comparing the laboratory experimental results and the numerical simulations.
In the laboratory experiments, core samples with 50 mm diameter and 30 mm height were prepared from the mudstone of the Oganomachi Formation, collected from Saitama prefecture. The core samples were kept in NaCl solution with the same total dissolved ion concentrations of pore water (less than 0.1 g/L) under hydrostatic pressure at the beginning of the experiments. The lateral and the upper surfaces were sealed by silicon rubber while the lower surface was in contact with the NaCl solution of which concentration was kept at about 10 g/L. The circumferential strains of the sample were measured by the optical fiber strain sensor attached helically on the lateral surface of the sample. The measurements were continued for more than 200 hours and the temperature was controlled almost constant throughout the experiments.
A numerical model to calculate chemo-osmotic and poroelastic processes was used to compare numerical simulation results with the experimental ones. The governing equations were based on a set of equations describing groundwater flow and diffusion including chemical osmosis process (Malusis et al., 2012) and those describing poroelastic effect (e.g., Wang, 2000), and pore pressure, concentration, and rock deformation were calculated. The calculated domain, boundary conditions and initial conditions were set to be the same as those of the laboratory experiments, for example, calculated domain was a cylinder with 50mm in diameter and 30mm in height, the upper and lateral surfaces were set to be no-flow and no solute flux boundary, and the lower surface was set to be constant pressure (0 kPa) and constant concentration (about 10 g/L). All the surfaces were set to be strain-free boundaries.
In the laboratory experiments, the continuous contraction was observed, and was interpreted to be caused by the osmotic flow and induced deformation by pressure decrease in the sample. The amount of contraction near the bottom was larger than that near the upper surface. The maximum 130 με was measured near the lower surface.
The effects of the parameters variation on the strain behavior were evaluated through numerical simulation, and the appropriateness of the model to explain the overall behavior of the samples was discussed by comparing the laboratory experimental results and the numerical simulations.