Underground porous media are complex multiphase systems, where the behavior at the macro-scale is affected by physical phenomena occurring at the pore(micro)-scale. The understanding of pore-scale ...fluid flow, transport properties, and chemical reactions is fundamental to reducing the uncertainties associated with the dynamic behavior, volume capacity, and injection/withdrawal efficiency of reservoirs and groundwater systems. Lately, laboratory technologies were found to be growing along with new computational tools, for the analysis and characterization of porous media. In this context, a significant contribution is given by microfluidics, which provides synthetic tools, often referred to as micromodels or microfluidic devices, able to mimic porous media networks and offer direct visualization of fluid dynamics. This work aimed to provide a review of the design, materials, and fabrication techniques of 2D micromodels applied to the investigation of multiphase flow in underground porous media. The first part of the article describes the main aspects related to the geometrical characterization of the porous media that lead to the design of micromodels. Materials and fabrication processes to manufacture microfluidic devices are then described, and relevant applications in the field are presented. In conclusion, the strengths and limitations of this approach are discussed, and future perspectives are suggested.
The understanding of multiphase flow phenomena occurring in porous media at the pore scale is fundamental in a significant number of fields, from life science to geo and environmental engineering. ...However, because of the optical opacity and the geometrical complexity of natural porous media, detailed visual characterization is not possible or is limited and requires powerful and expensive imaging techniques. As a consequence, the understanding of micro-scale behavior is based on the interpretation of macro-scale parameters and indirect measurements. Microfluidic devices are transparent and synthetic tools that reproduce the porous network on a 2D plane, enabling the direct visualization of the fluid dynamics. Moreover, microfluidic patterns (also called micromodels) can be specifically designed according to research interests by tuning their geometrical features and surface properties. In this work we design, fabricate and test two different micromodels for the visualization and analysis of the gas-brine fluid flow, occurring during gas injection and withdrawal in underground storage systems. In particular, we compare two different designs: a regular grid and a real rock-like pattern reconstructed from a thin section of a sample of Hostun rock. We characterize the two media in terms of porosity, tortuosity and pore size distribution using the A* algorithm and CFD simulation. We fabricate PDMS-glass devices via soft lithography, and we perform preliminary air-water displacement tests at different capillary numbers to observe the impact of the design on the fluid dynamics. This preliminary work serves as a validation of design and fabrication procedures and opens the way to further investigations.
AbstractThe goal of this study is the quantitative characterization of the degree of natural alteration of marble samples by using image analysis for the automatic characterization and comparison of ...the pore structure of rock samples before and after weathering. The proposed methodology is based on a pore exploration path-finding algorithm for the identification of paths developing within the porous domain of marble samples in both natural conditions and after weathering. Along each identified path, the pore radius is measured, providing a thorough description of the pore space statistical distribution. The A* path-finding approach was developed and applied to binarized images obtained from two-dimensional (2D) thin sections of marble samples in both natural conditions and after 10 years of natural decay. The results are expressed in terms of 2D porosity and statistical distributions of the pore radius of the samples preweathering and postweathering. A comparison with the information obtained from standardized laboratory tests used for the physical and mechanical characterization of stone material is also provided. From a computational point of view, the presented approach is highly parallelizable. The presented approach works wells in complex porous structures characterized by high path tortuosity, pore-size heterogeneity, and pore surface roughness. Moreover, the methodology is less affected by small-scale pore features and noise produced during image binarization, compared with other algorithms for pore structure morphological analysis such as skeleton-based and maximal ball approaches.