AI-Based Estimation of Poisson’s Ratio for Carbonate Formations Using Drilling Parameters in a Southern Iraqi Oil Field

Wellbore instability is a significant issue encountered during drilling operations. The mechanical properties of the formation are among the many factors that affect wellbore instability. Poisson's ratio is one of these mechanical properties and is a key factor in mechanical earth modeling (GEM). It is extremely important to minimize risks in drilling and production operations like sand output, collapse, tight holes, and pipe sticking. Poisson's ratio estimation contributes to optimizing hydro-carbon recovery and making important choices for a suitable field development plan. Poisson's ratio (υ) can be estimated both statically and dynamically. Static techniques measure the static properties in the lab, although static techniques are thought to be the most accurate way to determine the Poisson's ratio, they are costly, time-consuming, and unable to produce a continuous profile for Poisson's ratio. At the same time, dynamic methods compute the dynamic properties from well logging, such as density and the velocities of the compressional and shear waves, which are not always available. Thus, in this study, an artificial intelligence (AI) model is developed to estimate the Poisson's ratio for carbonate formation in the southern Iraqi oil field using available parameters during drilling. The dataset used in this study comprises over 451 data points, which range from depth of 2228 to 2453 m for the operations of training and testing. These data are including weight on bit (WOB), rotary speed (RPM), mud flow rate (FLW), Torque (T), standpipe pressure (SPP), and rate of penetration (ROP). The results indicate that new model can predict the Poisson's ratio with a high degree of accuracy (i.e., 93% correlation coefficients). Predicting rock Poisson's ratio from drilling data enables the early construction of a geomechanical model and saves cost and time compared to laboratory testing.