Reconstruction Of Sandstone 3D Model From 2D SEM Sample Micrographs

The sandstone is one of the most important reservoirs in our country. It is necessary to investigate the sandstone because of its unique grain and pore microstructure and reconstruct sandstone geology models to reveal the petrophysical properties. This paper reports on an investigation concerning the theory and technique of reconstructing and displaying the three dimensional(3D) grain and pore microstructure from 2D SEM micrographs and measuring the 3D parameters.In analyzing the 2D SEM micrographs, this paper adds a well structure on the slice structure and improves the general auto throat-segment and manual throat segment method and makes the throat-segment more efficient. The 2D analysis software can analyze both grain and pore medium.In sandstone stochastic modeling, this paper has researched 3D Voronoi modeling method and a stochastic modeling method based on morphology and compared several stochastic modeling methods' advantage and disadvantage. Finally, the stochastic modeling method based on stochastic morphology to reconstruct the sandstone sample is select. The 3D pore models using this reconstructing method not only have good vision effect but also have high correlation with 2D statistical data.In order to analysis the 3D parameter, this paper adopts regression and correlation method to analysis data correlation and run-length method to analysis the connectivity of 3D model. A new method expanding 2D morphology to 3D morphology is proposed to measure pore connectivity, pore coordination number, throat length and throat width which are in relation to sandstone sample's permeability.In 3D display, this paper discusses the projection display method, analyzes and compares the merits and shortcomings of some display methods, such as projection display, entity display and Virtual-Reality display and so on. Due to the conglutination of many grains in the reconstructed 3D grain model, an improved Virtual-Reality display method is proposed. Its good display effect is validated by the experiments.