1. Using machine learning for stuck pipe prediction as an early warning system for geothermal drilling operation in North Sumatra.
- Author
-
Sarwono, Lukas, Kartawidjaja, Maria Angela, and Wardana, Raka Sudira
- Subjects
- *
ARTIFICIAL neural networks , *SUPPORT vector machines , *GEOTHERMAL wells , *GAS well drilling , *DECISION trees , *MACHINE learning - Abstract
Stuck pipe in geothermal field is significant issues that can arise during drilling operations. This problem generates a substantial quantity of Non-Productive Time (NPT). Due to the fact that drilling in geothermal targets lost circulation zones, the likelihood of stuck pipe events increases. In numerous geothermal drilling projects, lost circulation events that contribute to stuck pipe events have become the most expensive NPT contributor. Although many observations have been made to develop a Machine Learning-based system for preventing stuck pipe incidents in hydrocarbon drilling operations, only few have been developed with drilling data from geothermal well. In this study, we propose a method to create a stuck pipe predictor using several machine learning algorithms such as Support Vector Machine (SVM), Artificial Neural Network (ANN) and Decision Tree (DT). The research showed that Machine Learning has the potential to help preventing stuck pipe incidents in geothermal drilling operation, even though the environment is different with oil and gas drilling operation. According to findings of this study, the Support Vector Machine Algorithm performed well in predicting stuck pipe, with 89% accuracy and 81% recall. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF