Multi-pore rock physics model: An intelligent approach for carbonate rocks

Abstract

Listen

Translate article

Despite the crucial role of the variability of pore type across a rock unit in determining the outputs of the corresponding rock physics model, existing models are known to consider a limited number of pore types. Expectedly, relatively high levels of uncertainty associated with such models have been reported in recent studies on some carbonate reservoirs. In this research, I proposed a modification to conventional rock physics models for carbonate rocks, where different pore types can be adequately addressed by means of an intelligent approach. For this purpose, I integrated a pore type classification scheme for carbonate rocks into a pattern recognition technique to come up with a comprehensive rock physics model that can accommodate so many (virtually unlimited) different pore types: multi-pore rock physics model (MP-RPM). Beginning by extending the porosity term of the Xu and White's rock physics model to a multi-pore-type domain, the proposed model estimated the pore geometry based on digital image analysis of images acquired from thin sections. The pore geometry herein refers to pore type and aspect ratio, which were obtained by implementing a multi-class classifier algorithm. As a reference, the pore geometry was further assessed by an expert geologist who developed a pore geometry catalog (reference database). A case study was performed by applying the modified model to a carbonate reservoir to estimate the P- and S-wave velocities, with the results compared with those of the conventional model as well as measured values. Results showed that, compared to the conventional model, the proposed modification provided for a better correlation to the ultrasonic and measured well-logging data. The most significant advantage of the proposed modification was the elimination of the restrictions on the different pore types that can be addressed appropriately, making the modified model able to capture the reality of sedimentary facies even more closely.