Impact of fracture coatings on fracture/matrix flow interactions in unsaturated, porous media

Abstract

Groundwater flow in unsaturated, fractured rock is often assumed to be dominated by the porous matrix component. This is frequently the result of reasoning that water flowing in the fractures is rapidly imbibed into the rock matrix by capillary suction forces with negligible resistance to uptake at the fracture/matrix interface. However, the existence of a low‐permeability mineralized layer or coating at this interface may substantially reduce matrix imbibition and could consequently result in fracture‐dominated flow. To test this idea, four tuff samples containing natural fractures were obtained from tuff formations in southern Nevada. By performing imbibition experiments into the matrix rock, across a mineralized fracture face and then across a fresh uncoated fracture face, water uptake as a function of time and coating was measured. A model combining Darcy's law and the Washburn equation has been used to describe the imbibition behavior. Before testing the tuff samples, the experimental technique and imbibition model were first tested using ordered sphere packings with known pore size and structure. From the tuff imbibition data, the ratio of the permeability through the mineralized face to the permeability through the uncoated face was found to be 0.3 (relatively little effect of the mineralized layer) for two samples, 0.01 for one sample, and 10−7 (uptake strongly restricted by the mineralized layer) for one sample. Using a simple fracture flow model and the imbibition model mentioned above to simulate matrix uptake from the fracture, numerical simulations indicate that the existence of fracture coatings could significantly decrease the travel time and increase the travel depth of water flowing through fractures.

Modeling Macroporous Soils with a Two-Phase Dual-Permeability Model

“…The initiation of macropore flow depends on initial conditions and rainfall intensity (e.g., Zehe and Flühler 2001;Weiler and Naef 2003). The mass transfer between matrix and macropores is controlled by the properties of the macropore surface and can be strongly inhibited, especially when the macropores are affected by lining and coating (Thoma et al 1992;Gerke and Köhne 2002). All these complex processes should be covered by numerical models to simulate water flow in macroporous soils (macropore flow).…”

Surface‐zone flow along unsaturated rock fractures

“…Hillel and Gardner [1970] presented an analysis of transient infiltration into a crust-covered soil, based upon the Green-Ampt model. A similar study on water imbibition into fracture coatings and underlying rock was performed by Thoma et al [1992], based on the Washburn [1921] approach. Both of these analyses showed that during the later stages, when the wetting front is within the underlying porous medium, the infiltration rate could be predicted based upon S or the hydraulic diffusivity (D) of the second medium.…”

“…This process, which we term "fracture surface-zone flow," can occur in fracture coatings, skins, or microcracked damage zones which have higher permeabilities than the underlying bulk rock matrix, when the local flux of water is high enough to sustain satiated water contents within fracture surface zones. Under such conditions, water is expected to flow preferentially within the fracture surface zone (Figure l Previous studies concerned with the hydrologic importance of the fracture surface zone (fracture skin) appear to have focused solely on matrix imbibition of water from fractures [Thoma et al, 1992;Chekuri, 1995]. Such studies have addressed the importance of permeability contrasts in controlling rates of water flow from fractures into rock.…”

Modifications and additions to selected TOUGH2 modules

“…However, there is evidence that fracture-matrix interaction in the unsaturated zone at Yucca Mountah is limited due to fingering flow in fractures (Glass et al, 1996) and mineral coatings on the fracture surfaces, and in fact there may be only part of fracturematrix interface areas that contributes fracture-matrix inter flow. Details of the conceptual models to reduce the fracture-matrix interface area have been proposed by Thoma et al (1992), Altman et al (1996), Ho (1997), and Bodvarsson et al (1997). In TOUGH2 the input data specified in CONNFi introduces information for the connection between elements (Pruess, 1987).…”