This project showcases a data-driven approach to optimize the Rate of Penetration (ROP) in drilling operations, leading to significant cost savings and improved efficiency. By leveraging machine learning models, specifically Gradient Boosting and XGBoost, on data from four Flemish Pass exploration wells, the project achieved a high accuracy in predicting ROP.
The models were used in conjunction with Multi-Objective Particle Swarm Optimization (MOPSO) to determine the optimal drilling parameters for various geological formations. This resulted in a simulated cost reduction of over $5.58 million. The study highlights the potential of machine learning to modernize drilling operations in the oil, gas, and geothermal sectors.
The code for this project is attached as ML_AppInDrilling.