A METHOD TO SELECT REPRESENTATIVE ROCK SAMPLES FOR DIGITAL CORE MODELING.

X-ray computed tomography (CT) scanning method is the most accurate method to construct digital core, which can reflect the microscopic pore structure of real cores; therefore, it is widely used and researched by experts and scholars all over the world. However, there are few reports about how to select the CT scan core samples at present, and the current practice is to make CT scan samples by visually observing rocks or core columns to select a region that is considered representative or interesting, which can lead to a large difference between the selected sample and the whole rock and a digital core that cannot represent the real rock as a whole. In order to construct the digital cores that can reflect the whole rock structure and reservoir properties, combining with fractal theory, a scientific and reasonable method was proposed to select representative rock samples for digital core modeling. First of all, a core column is scanned by X-ray CT at a certain resolution and CT gray scale images are obtained and stored in the order of scan. Secondly, the fractal dimension (FD) of each image is calculated by box-counting method, and the calculated porosity of each image is achieved by the existing formula. Then, according to the size of the digital core to be constructed, the CT gray scale images are grouped, and the average FD and the average porosity of each combination are calculated by the derived equations. Finally, based on the proposed criteria the best image combination is selected and the preferred sample is determined accordingly. At the same time, a facile experiment was conducted to test the effectiveness of this method. The experimental results show that there are some errors between the subjectively selected cores and the long core in terms of permeability and porosity, and the petrophysical parameters of the core selected by the proposed method are close to those of the long core; as a consequence, the validity of this method was verified and it is feasible and practical to select the representative rock samples for digital core modeling by this method. [ABSTRACT FROM AUTHOR]