Predicting Rock Hardness Using Measurement While Drilling (MWD) Data in a Case Study of Rotary Drilling in an Open-Pit Mine

Abstract Rock hardness significantly impacts drilling performance and costs. Direct field measurement is challenging, so indirect methods using drilling data have been developed. Measurement while drilling (MWD) records parameters like weight on bit, torque, and penetration rate, which reflect rock properties and conditions. This paper presented a novel approach to predict rock hardness using MWD data, machine learning, and Azure’s Data Factory and Databricks services, utilizing Spark techniques. A large MWD dataset from the Sarcheshmeh mine was processed to extract relevant features. Machine learning models, including random forest, gradient boosting machines, and support vector machines, were compared to find the best predictor. Model performance was evaluated using precision, recall, and F1 score. Results demonstrate high accuracy in rock hardness prediction, providing valuable insights for optimizing drilling operations.