Capillary Trapping Following Imbibition in Porous Media: Microfluidic Quantification of the Impact of Pore‐Scale Surface Roughness. Issue 11 (27th November 2019)

• Record Type: Journal Article
• Authors: Mehmani, Ayaz; Kelly, Shaina; Torres‐Verdín, Carlos; Balhoff, Matthew
• Is Part Of: Water resources research. Volume 55:Issue 11(2019)
• Page Start: 9905
• Page End: 9925
• Publication Date: 2019-11-27
• Subjects: imbibition -- capillary trapping -- microfluidics -- surface roughness -- porous media; Hydrology -- Periodicals; 333.91
• DOI: 10.1029/2019WR025170
• Languages: English
• ISSNs: 0043-1397

Abstract

Abstract: Due to diagenesis, pores in subsurface rocks such as sandstones exhibit varying degrees of surface roughness in the forms of authigenic cement coatings and mineral dissolution. Previous work describing capillary trapping in porous media has primarily focused on pore‐space geometry, wettability, and fluid viscosity contrast, while acknowledging, but not quantifying, the potential impact of surface roughness. We introduce a method to implement surface roughness with controlled variation of hillock density and heights into glass microfluidic chips and investigate surface roughness impacts on gas trapping following imbibition of water into air. We demonstrate that surface roughness with hillock height‐to‐pore‐depth ratios (herein called Ω ) less than a media‐dependent threshold ( Ω = 6%–10% in the micromodels) does not promote nonwetting phase (gas) trapping. By contrast, rougher micromodels with Ω values larger than the aforementioned roughness threshold show a dramatic increase in the saturation of trapped gas (gas saturation values up to 64%) due to an observed change in imbibition dynamics from binary filling to pendular‐ring formation within pore throats as well as capillary pinning within pore bodies. Furthermore, when the micromodel intermediate capillary number results are compared to Land's model, only the roughest microfluidics chips ( Ω > 10%) fall within the literature‐described values of the characteristic trapping constant, C, implying that surface roughness is also a key gas trapping control, independent of or in addition to pore‐space geometry and wettability. An a priori menisci stability criterion and a heuristic explanation based on local contact angle variations are proposed to explain surface roughness‐induced trapping. Key Points: Surface roughness is incorporated in a controlled manner into flow paths within glass micromodels that replicate granular porous media At a critical threshold region, micromodel roughness induces significant changes in subpore‐scale flow regimes, from binary filling to capillary pinning and the occurrence of pendular rings; this threshold is described in terms of hillock height‐to‐pore‐depth ratio A meniscus stability criterion is proposed for the formation of capillary bridges on roughness hillocks; below this criterion, pendular ring formation and capillary pinning are not favored A pseudo‐pore network model approach is used to demonstrate that capillary pinning and not percolation effects due to subpore‐scale geometrical randomness results in an exponential increase in capillary trapping as surface roughness increases