Prediction of Young’s Modulus for Reservoir Rocks using Machine Learning

Young’s Modulus is an important mechanical property of rocks in hydraulic fracturing design, as it determines their stiffness and indicates the extent to which they will deform elastically under uniaxial compressive stress. Young’s Modulus for reservoir ro ck can be determined via laboratory testing or using well log geophysical data. However, the laboratory testing depends on the availability of core samples, while data analysis based on well log data requires appropriate conversion formulas, which demand significant time, cost, and complexity. The approach presented in this study is based on the fact that real time drilling data, such as weight on bit (WOB), torque on bit (TQR), standpipe pressure (SPP), rotary speed (RPM), rate of penetration (ROP), and drilling fluid flow rate (FLOWIN), are re adily available in the early stages of the drilling process without incurring additional costs. Two machine learning algorithms were used to correlate drilling data with the dynamic Young’s Modulus : Random Forest and Decision Tree. Two different datasets were used in this study, the first dataset was utilized to develop and train the model, while the other dataset was used to validate the developed models. The first of the two methods employed demon strated a remarkable match between the given values and the predicted values. The correlation coefficients ranged from 0.71 to 0.95, with the average absolute percentage error being less than 5%. Based on the obtained results, the use of drilling data combined with artificial intelligence models for predicting Young’s Modulus proves to be a viable approach.