Abstract: Nuclear magnetic resonance (NMR) simulations of a porous medium are provided based on physical data such as CT images. A method of evaluating the surface relaxation of the porous medium based on random-walk NMR simulations includes a novel algorithm to derive the absorption probabilities of the random walkers, and algorithms that can evaluate more accurately the surface area and reduce the voxel discretization effect. The methods can be applied to conventional porous media as well as unconventional porous media such as reservoir rocks with high surface-to-volume ratios, for example tight sandstones or shale.

Abstract: A computer-implemented method for deriving properties of a porous material, the method includes: a first stage including: obtaining a first image of the porous material on a first scale; extracting a first network of pores from the first image; and deriving a first set of properties of the porous material using a first network flow modeling based on the first network; and a second stage including: obtaining a second image of the porous material on a second scale larger than the first scale; extracting a second network of pores from the second image; and deriving a second set of properties of the porous material using a second network flow modeling based on the second network and the first set of properties.

Abstract: A computer-implemented method for deriving properties of a porous material, the method includes: a first stage including: obtaining a first image of the porous material on a first scale; extracting a first network of pores from the first image; and deriving a first set of properties of the porous material using a first network flow modeling based on the first network; and a second stage including: obtaining a second image of the porous material on a second scale larger than the first scale; extracting a second network of pores from the second image; and deriving a second set of properties of the porous material using a second network flow modeling based on the second network and the first set of properties.

Abstract: Nuclear magnetic resonance (NMR) simulations of a porous medium are provided based on physical data such as CT images. A method of evaluating the surface relaxation of the porous medium based on random-walk NMR simulations includes a novel algorithm to derive the absorption probabilities of the random walkers, and algorithms that can evaluate more accurately the surface area and reduce the voxel discretization effect. The methods can be applied to conventional porous media as well as unconventional porous media such as reservoir rocks with high surface-to-volume ratios, for example tight sandstones or shale.