Fluid Flow Modeling in Fractured Rocks Using Human Mental Search Optimization

Abstract

Fluid flow in fractured rock systems presents a highly nonlinear and heterogeneous challenge, particularly in applications such as petroleum extraction, groundwater hydrology, and geothermal energy production. Traditional numerical models, including finite element (FEM) and finite difference (FDM) methods, often struggle with computational efficiency and accuracy in modeling rough rock fractures. In this study, we introduce a Human Mental Search (HMS)-based optimization framework to enhance fluid flow predictions in fractured rock systems. The HMS approach significantly reduces computational time while improving accuracy, as demonstrated by a 74% reduction in computation time compared to FEM (12.4 h for HMS vs. 48.2 h for FEM) and a lower mean absolute error (MAE) of 0.056 compared to 0.081 for FEM and 0.072 for FDM. Sensitivity analysis of fracture aperture sizes further validates the HMS approach, with flow rate predictions closely matching benchmark values, achieving an error margin below 1%. Moreover, HMS outperforms traditional methods in highly irregular fracture networks, maintaining a MAE of 0.086 for high-complexity networks, compared to 0.110 for FEM and 0.106 for FDM. These findings highlight the robustness of HMS in capturing intricate flow dynamics while ensuring computational efficiency. The proposed framework provides a novel and scalable approach for fluid flow modeling in fractured rock environments, offering improved predictive accuracy for real-world applications.