Authors : Ichenwo John Lander; Marvellous Amos

Volume/Issue : Volume 11 - 2026, Issue 3 - March

Google Scholar : https://tinyurl.com/bdfxednv

Scribd : https://tinyurl.com/2kkuemne

DOI : https://doi.org/10.38124/ijisrt/26mar184

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Abstract : Rate of Penetration (ROP) optimization is one of the challenges faced in petroleum operations, especially in presence of heterogeneous lithologies. The purpose of this study is the proposal of a machine learning based framework combining Principal Component Analysis (PCA) and regression modelling to quantify the lithological control on ROP. Measurements obtained while drilling logs from five wells in Nigeria’s Middle Benue Trough consisting gamma ray, resistivity and bulk density together with ROP measurements from an extensive database were used for shale, sand and carbonate intervals. The PCA extracted three principal components indicating that the three principal components explain 100% variance. 68.2 % for combined lithological effects of the reservoir was captured by PC1. PC2 with 28.8 means relationship of porosity-permeability and PC3 with 3.0 meaning residual formation properties were revealed. The regression model using PCA produced an R² of 0.891, while the RMSE was 2.35 m/h. The model produced a classification accuracy of 94.2% and an F1-score of 0.96. This was a 67% error reduction as compared to the mean predictor. These findings demonstrate the viability of PCA for real-time optimization of drilling in complex sedimentary settings, with direct application to adjusting parameters, selecting bits, and planning drilling programs.

Keywords : Rate of Penetration, Principal Component Analysis, Machine Learning, Lithology Classification, MWD Logs, Drilling Optimization, Middle Benue Trough.

References :

1. Altindal, M.C., et al. (2024). Anomaly detection in multivariate time series of drilling data. Geoenergy Science and Engineering, 234, 212589.
2. Bourgoyne, A.T., & Young, F.S. (1974). A multiple regression approach to optimal drilling and abnormal pressure detection. SPE Journal, 14(04), 371-384.
3. Ebrahimabadi, A., & Afradi, A. (2024). Prediction of Rate of Penetration (ROP) in Petroleum Drilling Operations using Optimization Algorithms. Rudarsko-geološko-naftni zbornik, 39(3), 119-134.
4. Elkatatny, S. (2021). Real-time prediction of the rate of penetration while drilling complex lithologies using artificial intelligence techniques. Ain Shams Engineering Journal, 12(1), 917-926.
5. Khan, S. H. (2025, August 7). Advanced Hybrid Transformer LSTM Technique with Attention and TS Mixer for Drilling Rate of Penetration Prediction. arXiv.org. https://arxiv.org/abs/2508.05210
6. MDPI (2024). Real-Time Lithology Prediction at the Bit Using Machine Learning. Geosciences, 14(10), 250.
7. Wang, Y., Lou, Y., Lin, Y., Cai, Q., & Zhu, L. (2024). ROP Prediction Method Based on PCA–Informer Modeling. ACS Omega, 9(21), 22456-22468.
8. Xiong, M., et al. (2024). A rate of penetration (ROP) prediction method based on improved dung beetle optimizer. Nature Scientific Reports, 14, 25047.