In this work, we develop and apply an offset well data analysis framework to generate a digital twin that is representative of bit state. We also strive to produce performance maps for well planning. Our workflow involves three major elements: 1) offset well data analysis to generate detailed depth-based and time-based statistics 2) computation of wear on bit and efficient weight-on-bit (WOB) versus depth for all the runs 3) automated machine learning to generate an accurate predictive model for bit dull grade to deploy for real-time operations. We define and calculate the efficient WOB as the minimum WOB needed to fully engage all cutters on the bit and maximize depth of cut. The efficient WOB changes with both bit state and formation strength, so an indication of bit wear can be used to predict formation strength. Deployment of the bit dull grade predictor enables the drilling crew to monitor the bit state while drilling and decide whether a reduction in ROP is due to a worn bit or other factors. Application to field cases from US Land yielded promising results. After computing the bit wear, performance maps for different bit states (e.g., sharp vs worn bits) are created. These performance maps help identify the optimal drilling parameters in well planning.

You can access this article if you purchase or spend a download.