A formation-response classifier for borehole deviation using conventional well log data
- 1School of Chemical Engineering, Vellore Institute of Technology, Vellore, Tamil Nadu, India
Abstract
Accurate and timely identification of borehole deviation is essential in drilling, as unexpected trajectory shifts serve as key indicators of localised geomechanical failure, formation anisotropy, and wellbore integrity degradation. This study presents a systematic machine learning (ML) framework for classifying borehole deviation as Normal or deviated, utilising seven petrophysical and geomechanical well-log features from a 301-sample single-well dataset. Five supervised classifiers, K-Nearest Neighbours (KNN), Logistic Regression (LR), Gaussian Naive Bayes (GNB), Multilayer Perceptron (MLP), and Histogram Gradient Boosting (HGB) were evaluated. Preprocessing incorporated standard scaling, median imputation, and class imbalance mitigation using balanced class weights and the Synthetic Minority Oversampling Technique (SMOTE). To eliminate spatial data leakage, models were evaluated within a 10-fold depth-blocked cross-validation framework across five random seeds, yielding 95% confidence intervals for all performance metrics. The HGB model achieved the highest macro-averaged F1-score of 97.88% (±2.05% CI) and an out-of-fold cross-validation accuracy of 98.00%, outperforming KNN (94.39%), MLP (86.04%), GNB (81.04%), and LR (77.03%). Permutation-based feature importance highlighted Delta T, Gamma-ray, and Resistivity as the primary indicators of deviation risk, aligning with geomechanical principles. These findings demonstrate that Histogram Gradient Boosting with leakage-safe preprocessing provides a robust offline proof-of-concept for assessing wellbore trajectory quality before field validation.
Related
- Presumptive detection of Giardia lamblia cysts in borehole water in Gombe metropolis, north-eastern Nigeria: a two-round cross-sectional study with implications for water safety
- A Comparison of Neutrophil Secretion of Cytokines and NET Formation in Response to Lipopolysaccharide and Group B Streptococcus (GBS) in Mice and Rats in vitro
- Deviation of bovine milk leukocyte functions against Escherichia coli and Staphylococcus aureus after co-stimulation with Weissella paramesenteroides
- Cytogenetic constraints on hybridization: a meta-analysis investigating the role of chromosome number in monocot hybrid evolution using a newly developed tool, the ploidy deviation index (PDI)
- Enhancing Clinical Classifier Robustness with Deep Representation Learning Based OOD Detection on DNA Methylation Data
BibTeX
@article{kumar2026formationresponse,
title = {{A formation-response classifier for borehole deviation using conventional well log data}},
author = {Naveen Kumar B and Aslam Abdullah Mohammed},
journal = {PLOS One},
year = {2026},
month = oct,
doi = {10.1371/journal.pone.0356220},
url = {https://doi.org/10.1371/journal.pone.0356220},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - A formation-response classifier for borehole deviation using conventional well log data
AU - Naveen Kumar B
AU - Aslam Abdullah Mohammed
PY - 2026
DA - 2026/10/09
JO - PLOS One
DO - 10.1371/journal.pone.0356220
UR - https://doi.org/10.1371/journal.pone.0356220
AB - Accurate and timely identification of borehole deviation is essential in drilling, as unexpected trajectory shifts serve as key indicators of localised geomechanical failure, formation anisotropy, and wellbore integrity degradation. This study presents a systematic machine learning (ML) framework for classifying borehole deviation as Normal or deviated, utilising seven petrophysical and geomechanical well-log features from a 301-sample single-well dataset. Five supervised classifiers, K-Nearest Neighbours (KNN), Logistic Regression (LR), Gaussian Naive Bayes (GNB), Multilayer Perceptron (MLP), and Histogram Gradient Boosting (HGB) were evaluated. Preprocessing incorporated standard scaling, median imputation, and class imbalance mitigation using balanced class weights and the Synthetic Minority Oversampling Technique (SMOTE). To eliminate spatial data leakage, models were evaluated within a 10-fold depth-blocked cross-validation framework across five random seeds, yielding 95% confidence intervals for all performance metrics. The HGB model achieved the highest macro-averaged F1-score of 97.88% (±2.05% CI) and an out-of-fold cross-validation accuracy of 98.00%, outperforming KNN (94.39%), MLP (86.04%), GNB (81.04%), and LR (77.03%). Permutation-based feature importance highlighted Delta T, Gamma-ray, and Resistivity as the primary indicators of deviation risk, aligning with geomechanical principles. These findings demonstrate that Histogram Gradient Boosting with leakage-safe preprocessing provides a robust offline proof-of-concept for assessing wellbore trajectory quality before field validation.
SN - 1932-6203
ER -