The development and stimulation of oil and gas fields are inseparable from the experimental analysis of reservoir rocks. Large number of experiments, poor reservoir properties and thin reservoir thickness will lead to insufficient number of cores, which restricts the experimental evaluation effect of cores. Digital rock physics (DRP) can solve these problems well. This paper presents a rapid, simple, and practical method to establish the pore structure and lithology of DRP based on laboratory experiments. First, a core is scanned by computed tomography (CT) scanning technology, and filtering back-projection reconstruction method is used to test the core visualization. Subsequently, three-dimensional median filtering technology is used to eliminate noise signals after scanning, and the maximum interclass variance method is used to segment the rock skeleton and pore. Based on X-ray diffraction technology, the distribution of minerals in the rock core is studied by combining the processed CT scan data. The core pore size distribution is analyzed by the mercury intrusion method, and the core pore size distribution with spatial correlation is constructed by the kriging interpolation method. Based on the analysis of the core particle-size distribution by the screening method, the shape of the rock particle is assumed to be a more practical irregular polyhedron; considering this shape and the mineral distribution, the DRP pore structure and lithology are finally established. The DRP porosity calculated by MATLAB software is 32.4%, and the core porosity measured in a nuclear magnetic resonance experiment is 29.9%; thus, the accuracy of the model is validated. Further, the method of simulating the process of physical and chemical changes by using the digital core is proposed for further study.

Dynamic displacement experiments and numerical two-phase flow estimation program are presented in this study. Unsteady-state core-flooding system was designed and utilized for oil-water dynamic displacement experiment. Two in situ sandstone core plug samples were used to investigate relative permeability via diesel and 3 wt % brine as non-wetting and wetting phase fluids. The transient data (pressure drop and produced fluid volume) were collected automatically for oil-water relative permeability computation. For relative permeability analysis, an inverse method program was constructed with four components: (a) an IMPE finite-difference numerical simulator of the flow through the core; (b) functional Corytype power law model of relative permeability in terms of a set of adjustable parameters found by minimizing an objective function; (c) the objective function formed by the sum of the square of the differences between the observations and calculated data; (4) the Gauss-Newton with Levenberg-Marquardt modification procedure for the leastsquares problem to minimize the objective function definition. All the above processes are embodied in relativepermeability calculation program, RCP, which is constructed in this study using FORTRAN language.