Quantitative Identification and Application of Rock Types in Complex Carbonate Reservoirs in the Middle East

Porous carbonate reservoirs in the Middle East are affected by sedimentation and diagenesis, with characteristics of multiple reservoir genetic types and complex pore structures. The permeability corresponding to the same porosity can vary by tens to hundreds of times. How to establish an accurate relationship between porosity and permeability based on the identification of reservoir rock types has always been an international problem. At present, there are many classification methods for carbonate reservoir rock types, including Fock, Dunham, and other methods based on the geological genesis, as well as other methods such as R35, FZI, and FZI* based on mathematical models. The rock type classification method based on geological genesis emphasizes geological factors such as carbonate reservoir lithology, diagenesis, and pore characteristics in the study area, which is difficult to effectively determine the reservoir classification standard and evaluate reservoir quality. The rock type classification method based on a mathematical model emphasizes the quantitative relationship model between rock physical parameters and rock parameters such as porosity and permeability but ignores the control effect of geological genetic characteristics on reservoir quality differences. To quantitatively evaluate rock types by combining geological genetic characteristics with petrophysical characteristics, this paper proposes that the first step is to quantitatively classify and evaluate the thin section data by applying digital image processing technology based on computer vision and hierarchical clustering algorithm based on grid and determine the basic rock types. Secondly, the grid-based hierarchical clustering algorithm is used to analyze the mercury injection curve clusters corresponding to the basic rock types, and the quantitative identification standard of pore carbonate reservoir rock types is determined. Finally, different rock types are distinguished, and corresponding pore-permeability relationships are established. The accuracy is significantly improved compared with other methods. This technology has been successfully applied to the M reservoir of W oilfield in the Middle East, and the effect is remarkable.