COMPARATIVE EVALUATION OF SUPERVISED MACHINE LEARNING ALGORITHMS FOR PETROPHYSICAL PROPERTY PREDICTION IN 'FSH FIELD', ONSHORE NIGER DELTA

Authors
  • O. E. Oyanameh

    Anchor University, Lagos, Nigeria.

    Author

  • O. Abiola

    Federal University of Technology, Akure

    Author

  • J. O. Amigun

    Federal University of Technology, Akure

    Author

Keywords:
Array, Array, Array, Array, Array
Abstract

Accurate prediction of reservoir properties is fundamental to improved reservoir characterization, production forecasting, and field development planning in hydrocarbon-bearing basins. Conventional geostatistical approaches in the Niger Delta often struggle with nonlinear subsurface heterogeneity, sparse well control, and uncertainty propagation, resulting in reduced predictive robustness. This study presents a performance comparative analysis of four supervised machine learning (ML) algorithms comprising Random Forest (RF), XGBoost (XGB), Support Vector Machine (SVM), and Gaussian Naïve Bayes (GNB)for enhanced prediction of reservoir properties using well-log datasets from the FSH Field, Niger Delta. The objective is to identify the optimal ML technique capable of improving the accuracy, stability, and reliability of geostatistical reservoir modeling workflows.A comprehensive dataset consisting of gamma ray (GR), resistivity (RT), neutron porosity (NPHI), density (RHOB), and sonic logs (DT) was preprocessed through normalization, depth alignment, outlier correction, and feature engineering. Models were trained on split well-log data and evaluated using standard regression metrics including the Coefficient of Determination (R²) and Root Mean Square Error (RMSE), while GNB was assessed using prediction accuracy.Results demonstrate that Random Forest and XGBoost significantly outperform the other algorithms, each achieving perfect predictive performance with R² = 1.000 across all evaluation cases, reflecting excellent nonlinear mapping capability and noise resilience. RF achieved the lowest average error (RMSE = 0.0005), followed closely by XGBoost (RMSE = 0.00225). SVM recorded moderate-to-high performance, with R² values ranging from 0.696 to 0.981 and an average RMSE of 0.054, indicating sensitivity to nonlinear patterns and the need for advanced kernel optimization. Gaussian Naïve Bayes, as expected for a classification-based algorithm applied to continuous data, yielded comparatively lower predictive accuracy (0.775 average accuracy) and was not competitive for reservoir property regression tasks.Overall, the findings confirm that ensemble-based supervised ML algorithms particularly Random Forest and XGBoost provide superior predictive accuracy and robustness for reservoir property estimation in data-limited Niger Delta fields. Integration of these ML models into geostatistical workflows can therefore enhance reservoir modeling fidelity, reduce subsurface uncertainty, and strengthen decision-making for field development and production optimization in the FSH Field.

Author Biographies
  1. O. E. Oyanameh , Anchor University, Lagos, Nigeria.

    Department of Earth Sciences, Anchor University, Lagos,Nigeria.

  2. O. Abiola , Federal University of Technology, Akure

    Department of Applied Geophysics, Federal University of Technology, Akure, Nigeria

  3. J. O. Amigun , Federal University of Technology, Akure

    Department of Applied Geophysics, Federal University of Technology, Akure, Nigeria

References

Adeogun, O. Y., Abdulwaheed, M. O., Adeoti, L., Allo, O. J., Fasakin, O. O., and Okunowo, O. O. (2025). Machine learning approach in predicting water saturation using well data at “TM” Niger Delta. Scientific African, vol. 27, e02596. https://doi.org/10.1016/j.sciaf.2025.e02596

Ahmadi, M. A., et al., (2019). Connectionist neural network and decision tree paradigms for predicting hydrocarbon saturation in heterogeneous reservoirs. Journal of Petroleum Science and Engineering, vol. 174, p.550-563.

Akinlabi, I. A., and Abiodun, A. O. (2024). Integration of seismic and well log data

For reservoir characterization in offshore Niger Delta, Nigeria. Journal of Geography, Environment and Earth Science International, vol. 28(6), p.1–11.

https://doi.org/10.9734/jgeesi/2024/v28i6774

Akpoyovbike A. A. (1978). Tertiary lithostratigraphy of Niger Delta. AAPG Bulletin,

Vol. 62(2), p.295–300. https://doi.org/10.1306/C1EA482E-16C9-11D7-8645000102C1865D

Alao, O. A., Ata, A. I., and Nwankwo, C. N. (2013). Lithofacies and depositional

environment of the Paleogene sequences in parts of the Niger Delta basin, Nigeria. Journal of Environment and Earth Science, vol. 3(11), p.101–110.

Anifowose, F., Labadin, J., and Abdulraheem, A. (2017). Ensemble machine learning applications in reservoir characterization. Journal of Petroleum Science and Engineering, vol. 151, p.158-167.

Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32. Kluwer

Academic Publishers.https://doi.org/10.1023/A:1010933404324

Caers, J. (2011). Modeling uncertainty in the Earth sciences.Wiley-Blackwell. 229p.

Chen, T., and Guestrin, C. (2016).XGBoost: A scalable tree boosting system. In

Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (p.785-794).

Chicco, D., Warrens, M. J., and Jurman, G. (2021). The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation. PeerJ Computer Science, vol. 7, e623. https://doi.org/10.7717/peerj-cs.623

Chilès, J.-P., and Delfiner, P. (2012). Geostatistics: Modeling spatial uncertainty (2nd

Ed.). John Wiley and Sons. 736p. https://doi.org/10.1002/9781118136188

Deutsch, C. V., and Journel, A. G. (1998). GSLIB: Geostatistical software library and

user's guide (2ndEd.). Oxford University Press. 369p.

Dougherty, J., Kohavi, R., and Sahami, M. (1995). Supervised and unsupervised

discretization of continuous features. Machine Learning Proceedings 1995, p.194-202.

Dubrule, O. (2003). Geostatistics for seismic data integration in earth models.

Society of Exploration Geophysicists and European Association of Geoscientists and Engineers. (279p.)https://doi.org/10.1190/1.9781560801962

Ejedawe, J. E. (1981). Patterns of incidence of oil reserves in Niger Delta basin. AAPG Bulletin, vol. 65(9), p.1574–1585.

https://doi.org/10.1306/2F919BC4-16CE-11D7-8645000102C1865D

El-Sebakhy, E. A., et al.(2012). Support vector machines framework for predicting

petroleum reservoir properties. Neurocomputing, vol. 84, p.1-13.

Eze, C. I., Okengwu, K. O., and Onuoha, K. M. (2020). Reservoir characterization and sequence stratigraphy of "Alpha" Field, Onshore Niger Delta Basin, Nigeria. Journal of Petroleum Exploration and Production Technology, vol. 10(6), p.2235–2252. https://doi.org/10.1007/s13202-020-00898-w

Fagbemi, O. I., Olayinka, A. I., Oladunjoye, M. A., (2024). Focused reservoir characterization: analysis of selected sand units using well log and 3-D seismic data in ‘Kukih’ field, Onshore Niger Delta, Nigeria. Scientific Reports, vol. 14, p.13763. https://doi.org/10.1038/s41598-024-56100-7

Fang, Z., Ba, J., Carcione, J. M., Xiong, F., and Gao, L. (2024). Permeability

prediction using logging data from tight reservoirs based on deep neural networks. Journal of Applied Geophysics, vol. 229, p.105501.

Goovaerts, P. (1997). Geostatistics for natural resources evaluation. Oxford

University Press. (483 pp.)ISBN: 978-0-19-511538-3

https://doi.org/10.1093/oso/9780195115383.001.0001

Haldorsen, H. H., and Damsleth, E. (1990). Stochastic modeling. Journal of Petroleum Technology, vol. 42(4), p.404–412.

Hand, D. J., and Yu, K. (2001). Idiot's Bayes—not so stupid after all? International

Statistical Review, vol. 69(3), p.385–398. Oxford, UK: Wiley-Blackwell. https://doi.org/10.1111/j.1751-5823.2001.tb00465.x

Hastie, T., Tibshirani, R., and Friedman, J. (2009). The elements of statistical learning: Data mining, inference, and prediction (2nd Ed.,). New York, NY: Springer.745p.

He, M., Gu, H., and Wan, H. (2021). Log interpretation for reservoir properties prediction based on XGBoost algorithm. Journal of Petroleum Science and Engineering, vol. 205, p.108786. https://doi.org/10.1016/j.petrol.2021.108786

Hossin, M., and Sulaiman, M. N. (2015). A review on evaluation metrics for data

classification evaluations. International Journal of Data Mining and Knowledge Management Process, vol. 5(2), p.1-11.

Ibe, A. A., Oturu, F. K., and Anyanwu, J. (2024). Optimizing reservoir

characterization: insights from integrated data analysis. Discover Geoscience, vol. 2, p.70 https://doi.org/10.1007/s44288-024-00076-4

Iboyitie, O. (2024). Machine learning-based rock facies classification for improved reservoir characterization in Niger Delta. FUPRE Journal of Scientific and Industrial Research, vol. 13(2), p.20–27.

Isaaks, E. H., and Srivastava, R. M. (1989). An introduction to applied geostatistics.

Oxford University Press. 561p.

Journel, A. G., and Huijbregts, C. J. (1978). Mining geostatistics. Academic Press. 600p.

Koray, A.-M., Bui, D., Kubi, E. A., Ampomah, W., and Amosu, A. (2024). Machine learning based reservoir characterization and numerical modeling from integrated well log and core data. Geoenergy Science and Engineering, vol. 243, p.213296 https://doi.org/10.1016/j.geoen.2024.213296

Kuhn, M., and Johnson, K. (2013). Applied predictive modeling. Springer. 600p

https://doi.org/10.1007/978-1-4614-6849-3

Li, X., and Zhang, D. (2018). Data-driven geostatistical modeling using machine learning. Computers and Geosciences, vol. 118, p.1–12.

https://doi.org/10.1016/j.cageo.2018.05.009

Li, J., and Zhao, X. (2020). Machine learning applications in well log data analysis: A review and comparison of algorithms. Journal of Applied Geophysics, vol. 181, p.104168. https://doi.org/10.1016/j.jappgeo.2020.104168

Moriasi, D. N., Arnold, J. G., Van Liew, M. W., Bingner, R. L., Harmel, R. D., and Veith, T. L. (2007). Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. Transactions of the ASABE, vol. 50(3), p.885–900. https://doi.org/10.13031/2013.23153

Moriasi, D. N., Gitau, M. W., Pai, N., and Daggupati, P. (2015). Hydrologic and water

quality models: Performance measures and evaluation criteria. Transactions of the ASABE, vol. 58(6), p.1763–1785. https://doi.org/10.13031/trans.58.10715

Nadin, P. A., Kusznir, N. J., and Cheadle, M. J. (1995). Early Tertiary uplift and Eocene subsidence in the Northern North Sea Basin from 2D forward and reverse stratigraphic modelling. Journal of the Geological Society, vol. 152(5), p.833–848. https://doi.org/10.1144/gsjgs.152.5.0833

Nancy, C., and Raj, S. A. (2018). Advanced machine learning paradigms for subsurface log evaluation and lithofacies prediction. Journal of Petroleum Science and Engineering, vol. 168, p.420–432.

Obaje, N.G., 2009. Geology and Mineral Resources of Nigeria. Springer Dordrecht

Heidelberg London, New York, 221p.

http://dx.doi.org/10.1007/978-3-540-92685-6

Orife, J. M., and Avbovbo, A. A. (1982). Stratigraphic and unconformity traps in the Niger Delta. AAPG Bulletin, vol. 66(2), p.251–265. https://doi.org/10.1306/03B59A7C-16D1-11D7-8645000102C1865D

Rahimi, M., Asghari, O., and Hajizadeh, F. (2022). Machine learning and geostatistical residual kriging integration for spatial modeling of subsurface property distributions. Journal of Petroleum Science and Engineering, vol. 215, p.110620. https://doi.org/10.1016/j.petrol.2022.110620

Reijers, T. J. A. (2011). Stratigraphy and sedimentology of the Niger Delta. Geologos, vol. 17(3), p.133–162. https://doi.org/10.2478/v10118-011-0008-3

Sun, A. Y., Chen, J., and He, J. (2021). Physics-informed machine learning for geoscience applications. Earth-Science Reviews, vol. 214, p.103515.

Tariq, Z., Elkatatny, S., Mahmoud, M., Abdulraheem, A., & Ali, A. Z. (2021). A novel systematic and evolved approach based on XGBoost-firefly algorithm to predict Young's modulus and unconfined compressive strength of rock. Applied Soft Computing, 111, 107691.

https://doi.org/10.1016/j.asoc.2021.107691

Uzkeda, H., VidalRoyo, O., and Amilibia, A. (2023). Using AI tools to fill an incomplete well log dataset: A workflow. Journal of Applied Geophysics, vol. 212, p.104992 https://doi.org/10.1016/j.jappgeo.2023.104992

Vapnik, V. (1995). The Nature of Statistical Learning Theory. Springer-Verlag, New York, 2nd Ed. 314p.

Cover Image
Downloads
Published
2026-12-31
Section
Articles