Mustafa Rezaei

Permeability prediction is essential for reservoir characterization, commonly derived from core analysis and mercury injection capillary pressure (MICP) data. Many conventional models, often calibrated for sandstones, are based on parameters such as porosity or specific mercury saturation points, which limits their accuracy in carbonate reservoirs due to differing rock properties. This study addresses these limitations by integrating parameters from 12 existing permeability estimation models and selecting three reliable models for carbonates: Swanson’s, Winland’s, and Pittman’s models. These models incorporate Swanson’s parameter (maximum Sb/Pc, where Sb is non-wetting phase saturation and Pc is capillary pressure), pore-throat radius at specific saturations (Winland’s rx), and porosity (Pittman’s model). An integrated framework was developed using data from 50 carbonate samples and validated with 20 additional samples and log data. The permeability values range from 0.01 to 450 millidarcies (mD), with porosity from 1% to 30%. Multiple linear regression established robust relationships between permeability, porosity, and Swanson’s parameter, improving prediction accuracy. Validation using MICP and Stoneley wave data confirmed the model's reliability, demonstrating significant advancements in permeability estimation for heterogeneous carbonate systems. This approach offers a comprehensive and accurate tool for reservoir characterization.