Application of Machine Learning for Bit-formation Matching in Drilling Operations
Petroleum Science and Engineering · 29.05.2025 · 9 (1) · 38-47
DOI: 10.11648/j.pse.20250901.14 ↗Kurz gefasst
Fachartikel in der Zeitschrift Petroleum Science and Engineering, online erschienen am 29.05.2025. Verfasst von Shedrach Igemhokhai und 5 weiteren Autor:innen. Die Arbeit nennt 17 Literaturangaben.
Abstract (Originalsprache)
Efficient bit formation matching is imperative for the success and cost-effectiveness of drilling operations and emissions reduction to provide energy solutions. Currently, drill bit selection predominantly depends on historical data and experiential knowledge. While machine learning, particularly Artificial Neural Networks (ANNs), has gained prominence in bit selection, other diverse and impactful algorithms such as XGBOOST and Random Forest (RF), are often overlooked. This paper involves the systematic application and comparative analysis of XGBOOST, RF, and ANN, alongside an optimization approach using Genetic Algorithm. The study comprehensively considers various influential factors including formation properties, drilling fluid characteristics, bit design, and operational parameters. In this study, we achieved promising results with the highest classification accuracy for bit selection recorded at 0.97 using the XGBOOST model, while RF and ANN yielded accuracies of 0.91 and 0.93 respectively. Additionally, we obtained impressive R squared values of 0.991, 0.975, and 0.953 for predicting the Rate of Penetration using the XGBOOST, ANN, and RF models respectively. These algorithms, coupled with the optimization techniques, aim to establish a robust framework for nuanced and accurate bit-formation matching. The results obtained hold significant potential for minimizing costs and optimizing resource allocation and utilization during the planning and execution of drilling projects in the oil and gas industry.
Zitieren
APA
Igemhokhai, S., Bello, K., Olaoye, A., Momodu, E., Adejumo, A., & Akindayini, O. (2025). Application of Machine Learning for Bit-formation Matching in Drilling Operations. Petroleum Science and Engineering, 9(1), 38-47. https://doi.org/10.11648/j.pse.20250901.14
Vancouver
Igemhokhai S, Bello K, Olaoye A, Momodu E, Adejumo A, Akindayini O. Application of Machine Learning for Bit-formation Matching in Drilling Operations. Petroleum Science and Engineering. 2025;9(1):38-47. doi:10.11648/j.pse.20250901.14
BibTeX
@article{Igemhokhai2025,
author = {Igemhokhai, Shedrach and Bello, Kelani and Olaoye, Abiodun and Momodu, Ehigie and Adejumo, Abayomi and Akindayini, Oladele},
title = {Application of Machine Learning for Bit-formation Matching in Drilling Operations},
journal = {Petroleum Science and Engineering},
year = {2025},
volume = {9},
number = {1},
pages = {38--47},
doi = {10.11648/j.pse.20250901.14}
}
EUVIMED