Abstract: The microscopic characteristics of carbonate reservoirs cannot be easily described due to the limitation of existing geophysical data scale. By using the critical porosity model as rock physical theory, the relationships between pore structures and elastic responses were built, and the key parameters were estimated through statistical method. Furthermore, the experiments were used to classify logging data with their critical porosities. The results indicate that the accuracy and stability of this new method (error less than 10%) meet the needs of carbonate reservoir exploration at Log YS1. It provides a new method to reduce the pore complexity and inversion uncertainty, and has reference value for geological interpretation.

Key words: carbonate reservoirs, rock physics modelling, critical porosity, Bayesian estimation