Abstract

Accurate permeability prediction is very critical to the successful operation and management of hydrocarbon reservoir. Support Vector Regression (SVR) has been employed as a computational intelligence technique for accurate prediction of reservoir permeability. However, optimization of SVR hyperparameters is highly essential and has predominantly been achieved through a trial-and error approach. This paper introduces the implementation of the real-coded genetic algorithm (RCGA) for optimizing Support Vector Regression (SVR) hyperparameters. This marks the first instance of employing such an approach in predicting the permeability of hydrocarbon reservoirs. RCGA is a heuristic optimization algorithm widely employed in many optimization problems. The RCGA-SVR model is evaluated against two commonly used SVR models: one with hyperparameters manually optimized through a trial-and-error approach (SVR) and another with randomly-optimized hyperparameters (RAND-SVR). The comparison utilizes real-life industrial datasets obtained from petroleum exploration, specifically gathered from four distinct oil wells in a Middle Eastern oil and gas field. The results indicate that RCGA-SVR surpasses both ordinary SVR and RAND-SVR. Notably, RCGA-SVR showcases a reduction in root mean square error (RMSE) of 14%, 26%, 12%, and 8% for well-A, well-B, well-C, and well-D respectively in comparison to ordinary SVR. The results of this paper show that heuristic algorithms such as RCGA has excellent potential in the optimization of SVR hyperparameters employed for modeling and characterization of hydrocarbon reservoirs result in enhanced correlation and error performance. Hence, RCGA is suggested for optimizing SVR hyperparameters, aiming for precise predictions of hydrocarbon reservoir permeability, given its reported excellent performance.

Copyright of the paper named above is hereby assigned and transferred to the Nigerian Journal of Applied Science and Innovative Technology published by College of Engineering, Bells University of Technology, Nigeria.