In Wook Yeo

Effect of shear displacement on the aperture and permeability of a rock fracture

The results of experiments using radial and unidirectional flow in a carefully described single rough aperture are reported and compared with numerical predictions. Aperture replicas of a natural sandstone fracture were made at 0, 1 and 2 mm shear displacements using silicone rubber, with a reproducibility of better than 2%. The experimental arrangement permitted shear displacement to be obtained without causing damage to the two displaced surfaces, which thus retained their original (and essentially matching) geometry. Both the number of contact points and the fractional contact area decreased with increasing shear displacement. With increasing shear displacement, mean aperture and standard deviation increased and the ratio of standard deviation to mean aperture increased slightly. Semivariogram studies indicated that as shear displacement increased in the direction normal to the roughness ridges, the aperture distribution became more closely correlated in the direction parallel to the roughness ridges than in the shear direction. Flow tests showed that with increasing shear displacement, the fracture became heterogeneous and anisotropic and became more permeable in the direction perpendicular to the shear displacement than in the direction parallel to the displacement. Simulations were made of flow through the fracture using the Reynolds equation and the actual aperture distribution; the measured hydraulic apertures were generally about 20% lower than those measured numerically.

Assessment of the validity of Stokes and Reynolds equations for fluid flow through a rough‐walled fracture with flow imaging

Understanding fluid flow through a rough-walled fracture is important in many problems such as petroleum and geothermal reservoir exploitation, geological storage of CO2, and sitting of radioactive waste repositories. In order to advance the understanding of fracture flow, we conducted the first direct measurement of flow velocity across rough-walled fractures at Reynolds number (Re) of 0.014 to 0.086. The results were used for an order of magnitude analysis to evaluate assumptions underlying the Stokes and the Reynolds equations, which are derived from simplifying the Navier–Stokes equations. Even at very rough subregions, viscous forces were at least 2 orders of magnitude greater than inertial forces, indicating that the Stokes equations are valid for Re < 0.1. However, the assumption made in the derivation of the Reynolds equation that ∂2ux/∂z2 is dominant over other viscous terms was not satisfied even at moderate roughness for Re < 0.1. The Reynolds equation overestimated flow rate.

Applicable range of the Reynolds equation for fluid flow in a rock fracture

The applicability of the Reynolds equation to fluid flow in rock fractures has recently been questioned. One issue is that the Reynolds equation overestimates the transmissivity for rough fractures. Another issue is to delineate the range of fracture roughness over which the Reynolds equation is reasonably accurate. In this paper, the analysis was undertaken to examine the relationship between fracture geometry and the accuracy of the Reynolds equation. This analysis leads to identifying a criterion parameter, (bm/Λ)(A/Λ), wherebm/Λ is the ratio of mean aperture to wavelength, representing roughness, andA/Λ is the ratio of amplitude of oscillated fracture roughness to the wavelength, representing tortuosity. We found that when (bm/Λ)(A/Λ) is less than 0.01, the Reynolds equation may yield accurate transmissivity predictions.

Solute dispersion in rock fractures by Non-Darcian Flow

This study aims at investigating the relationship between the ratio of dispersion coefficient to molecular diffusion coefficient (D L /D m ) and Peclet number (Pe) for a multi-solute system in non-Darcian flow regimes. Existing understanding on solute dispersion is primarily derived from Darcian flow regime with one solute. We find that solute dispersion in rock fractures can be characterized by the mechanism of both macrodispersion and Taylor dispersion, in non-Darcian flow domains. For the Darcian flow regime, different solutes lead to the same D L /D m - Pe relationship. However, as the flow becomes non-Darcian, solute with a higher molecular diffusion coefficient result in higher D L /D m at the same Pe than that with a lower diffusion coefficient.

Wettability-dependent DNAPL migration in a rough-walled fracture.

The effect of wettability on the migration of dense non-aqueous phase liquids (DNAPLs) through a rough-walled fracture was investigated. The migration characteristics of DNAPL were found to be strongly dependent on the wettability. For a fracture with a hydrophilic surface, DNAPL migrated through larger apertures as disconnected blobs when the groundwater flow regime was linear (Re=1). However, for non hydrophilic surfaces DNAPL did not migrate in the same way as for the hydrophilic surface. The intermediate-wet surface, with a contact angle of approximately 90 degrees , makes gravity pressure dominant over the capillary pressure, resulting in the fastest DNAPL migration. DNAPL was retained on the hydrophobic fracture, where the capillary barrier of larger apertures forced the DNAPL to migrate through the smaller apertures. In the nonlinear flow regime of Re=60, DNAPL generally migrated downward as a result of the inertial pressure of flowing water for all the wettability conditions, but the local downward migration paths were still determined by the capillary pressure, which resulted in the fastest and slowest migration on the hydrophilic and the hydrophobic fractures, respectively. This study implies that the hydrophilic and intermediate-wet surfaces will be favorable for DNAPL and oil recovery.

Tail shortening with developing eddies in a rough-walled rock fracture

Understanding fluid flow and solute transport in rough-walled fractures is important in many problems such as geological storage of CO2 and siting of radioactive waste repositories. The first microscopic observation of fluid flow and solute transport through a rough-walled fracture was made to assess the evolution of eddies and their effect on non-Fickian tailing. A noteworthy phenomenon was observed that as the eddy grew, the particles were initially caught in and swirled around within eddies, and then cast back into main flow channel, which reduced tailing. This differs from the conventional conceptual model, which presumes a distinct separation between mobile and immobile zones. Fluid flow and solute transport modeling within the 3-D fracture confirmed tail shortening due to mass transfer by advective paths between the eddies and the main flow channel, as opposed to previous 2-D numerical studies that showed increased tailing with growing eddies.

Influence of ambient groundwater flow on DNAPL migration in a fracture network: Experiments and simulations

[1] We consider the influence of ambient groundwater flow on the migration of DNAPL within a fracture network. In context of a modified invasion percolation (MIP) growth algorithm, we formulate a mechanistic model that includes capillary and gravity forces as well as viscous forces within the DNAPL and the ambient groundwater. The MIP model is verified against laboratory experiments, which show good agreement in DNAPL migration path through a two-dimensional fracture network. The results of both simulations and laboratory experiments suggest that ambient groundwater flow can be a significant factor controlling DNAPL migration path, velocity, and channeling pattern in a fracture network. INDEX TERMS: 5104 Physical Properties of Rocks: Fracture and flow; 5139 Physical Properties of Rocks: Transport properties; 5114 Physical Properties of Rocks: Permeability and porosity. Citation: Ji, S.-H., I. W. Yeo, K.-K. Lee, and R. J. Glass, Influence of ambient groundwater flow on DNAPL migration in a fracture network: Experiments and simulations, Geophys. Res. Lett., 30(10), 1504, doi:10.1029/2003GL017064, 2003.

Derivative-Assisted Classification of Fractured Zones Crossing a Deep Borehole

In this study, the derivative analysis using the derivative of drawdown with respect to log-time was utilized to determine candidates for hydraulic conductor domains (HCDs). At a 500-m deep borehole in the study site, the fractured rocks crossing the borehole were first classified in fractured and nonfractured zones by core logging and geophysical loggings, such as acoustic televiewing, density, and flow loggings. After conducting the hydraulic tests such as constant head withdrawal and recovery tests at the fractured zones and the nonfractured zones, the derivative analyses were carried out, of which the results were evaluated to determine the candidates for HCDs. For the nonfractured zones, the diagnostic plot has only a big hump indicating poor connection of the background fractures to the permeable geologic media, while those of the candidates for HCDs show various flow regimes. On the basis of these results, the candidates for HCDs among the fractured zones were determined. From discussion on the results, the combination of the spacing analysis and derivative analysis following a hydraulic test is recommended for determining the candidates for HCDs rather than other geophysical loggings.

Fluid flow through rough-walled rock fractures with hydrophobic surfaces

The effect of hydrophobic properties of surfaces on fluid flow was investigated through water flow experiments using rough-walled rock fractures. At low Reynolds number (Re < 10), the slip caused by drag reduction at hydrophobic rough-walled surfaces produced larger water flow than for fractures bounded by hydrophilic surfaces. The occurrence of drag reduction was only limited to the linear flow regime. An increased flow velocity due to slip made the flow nonlinearity stronger as the flow became nonlinear, which caused the smaller flow rate through the fracture with hydrophobic surfaces than that with hydrophilic surfaces. Results showed that the drag reduction of 7.92% at Re = 0.64 was changed to apparent drag enhancement of 8.92% at Re = 185 for the creosote-wetted rough-walled fracture. This changeover phenomenon was more pronounced with increasing the roughness of the fracture surfaces. This study implied a dependence of the operation efficiency on surface wettability and flow regime for DNAPL recovery from fractured bedrock and oil recovery or the sequestration of carbon dioxide in fractured petroleum reservoirs.