Drill Bit Selection and Drilling Parameter Optimization using Machine Learning

Abstract

Machine Learning (ML) Algorithms have demonstrated their tremendous application in optimizing and enhancing the performance of various complex operations in the field of science and technology. In this research work, ML is applied to address two of the most critical factors affecting the drilling performance in the Oil and Gas Industry, which are drilling bit selection and drilling parameters optimization. Rate of Penetration is a key performance indicator of drilling efficiency, higher ROP signifies higher drilling efficiency. In this research work, a hyperparameter tuned Random Forest Regressor algorithm with an accuracy of 0.73 based on the coefficient of determination i.e., R2 Score, is used to develop ROP prediction model and subsequently drill bit selection and drilling parameters optimization is performed using Particle Swarm Optimization. The developed model has practical applicability in the selection of drill bit and optimization of drilling parameters in the Oil and Gas field. Higher ROP results in less drilling time, which correspondingly results in less capital expenditure on the project.