Evaluation of Reservoir Porosity and Permeability from Well Log Data Based on an Ensemble Approach: A Comprehensive Study Incorporating Experimental, Simulation, and Fieldwork Data

• Published in: Natural resources research (New York, N.Y.) Vol. 34; no. 1; pp. 383 - 408
• Main Authors: Nyakilla, Edwin E., Guanhua, Sun, Hongliang, Hao, Charles, Grant, Nafouanti, Mouigni B., Ricky, Emanuel X., Silingi, Selemani N., Abelly, Elieneza N., Shanghvi, Eric R., Naqibulla, Safi, Ngata, Mbega R., Kasala, Erasto, Mgimba, Melckzedeck, Abdulmalik, Alaa, Said, Fatna A., Nadege, Mbula N., Kasali, Johnson J., Dan, Li
• Format: Journal Article
• Language: English
• Published: New York Springer Nature B.V 01.02.2025
• Subjects: Artificial neural networks; Back propagation networks; Core analysis; Fieldwork; Gaussian process; geophysical logging; Machine learning; Mathematical models; Mean square errors; Multivariate analysis; Neural networks; normal distribution; Performance evaluation; Permeability; Porosity; prediction; Reservoirs; Support vector machines; Well logs
• DOI: 10.1007/s11053-024-10402-9
• ISSN: 1520-7439

Permeability and porosity are key parameters in reservoir characterization for understanding hydrocarbon flow behavior. While traditional laboratory core analysis is time-consuming, machine learning has emerged as a valuable tool for more efficient and accurate estimation. This paper proposes an ensemble technique called adaptive boosting (AdaBoost) for porosity and permeability estimation, utilizing methods such as support vector machine (SVM), Gaussian process regression (GPR), multivariate analysis, and backpropagation neural network (BPNN) for prediction based on well logs. Performance evaluation metrics including root mean square error, mean square error, and coefficient of determination (R2) were used to compare the models. The results demonstrate that AdaBoost outperformed GPR, SVM, and BPNN models in terms of processing time and accuracy, achieving R2 values of 0.980 and 0.962 for permeability and porosity during training, respectively, and 0.960 and 0.951 during testing, respectively. This study highlights AdaBoost as a robust and accurate technique that can enhance reservoir characterization.