Sandstone Core Properties, Hybrid PSO Coupled with MLR, and Averaging Techniques for Capillary Pressure and Relative Permeability Curve Assignment

Sandstone core properties play a crucial role in understanding the behavior of oil reservoirs. In a recent study, researchers investigated the properties of sandstone cores, including porosity, permeability, clay content, and wettability. Two different rock types were selected for analysis, with distinct properties reported in Table 1.

To estimate core scale relative permeability and capillary pressure curves for sandstone oil reservoirs, researchers employed a hybrid particle swarm optimization (PSO) coupled with machine learning techniques. By categorizing studies of similar lithologies, researchers were able to determine matching parameters using a reservoir simulator and the PSO algorithm.

The process involved several steps, including preparing the initial swarm, calculating relative permeability and capillary pressure values, creating include files, running MATLAB coupled with the reservoir simulator, and post-processing the results. By optimizing errors and creating a new swarm based on the best solutions, researchers were able to determine the optimized model parameters for each rock type.

Proposed correlations for relative permeability parameters were developed using multiple-linear regression techniques. These correlations were validated against reported experimental data, showing high accuracy and reliability for predicting relative permeability and capillary pressure functions at different salinity conditions.

Averaging techniques were also discussed to assign a single capillary pressure and relative permeability curve to grid blocks in the simulator. The Leverett J function was used to calculate the average capillary pressure, providing valuable insights into the behavior of sandstone reservoirs.

Overall, this study highlights the importance of understanding sandstone core properties and utilizing advanced techniques to estimate key parameters for reservoir simulation and oil recovery processes. The findings contribute to the ongoing research in the field of petroleum engineering and reservoir management.