Reconstruction of three-dimensional porous media using a single thin section

Area of Science:

• Geosciences and Petrophysics
• Computational Modeling and Simulation

Background:

• Accurate reconstruction of disordered media, like porous rocks, is crucial for predicting flow and transport properties.
• Current methods often rely on low-order statistics, failing to capture the complex morphology of heterogeneous porous media.
• Reconstructing 3D porous media from 2D thin sections is challenging due to data limitations and computational complexity.

Purpose of the Study:

• To develop a novel method for reconstructing 3D porous media from a single 2D image using multipoint statistics.
• To accurately represent the morphology, connectivity, and flow properties of porous media.
• To address the limitations of existing reconstruction techniques for heterogeneous materials.

Main Methods:

• Utilizes a 1D raster path inspection of the 2D digitized image.
• Combines cross-correlation functions, grid splitting techniques, and Shannon entropy for heterogeneity assessment.
• Employs an adaptive technique to identify optimal hard data points for reconstruction.

Main Results:

• The proposed method successfully reconstructs 3D porous media from 2D thin sections, preserving long-range connectivity.
• Quantitative statistical tests show good agreement between reconstructed properties and experimental data for Berea sandstone and carbonate rock.
• The method demonstrates computational efficiency.

Conclusions:

• The multipoint statistics-based reconstruction method effectively overcomes limitations of traditional approaches for heterogeneous porous media.
• Accurate prediction of morphological and transport properties is achievable from limited 2D data.
• This technique offers a viable solution for reconstructing complex 3D geological structures.