Drilling Digital Twin Predicts Drilling Dysfunctions and Performance in Real Time

Abstract

Abstract In the new digital age, improving drilling economics with tools like simulation and predictive analytics are key to enabling digital transformation. A novel real time digital twin has been developed to predict drilling dysfunctions and improve operational efficiencies. Real time surface (EDR) data with full physics time domain models are used to predict drilling dynamics and drilling dysfunctions in the bottom hole assembly at any depth. This provides a rare insight for drilling engineers to improve drilling performance and take predictive or corrective measures for reliability and operational efficiencies. In the study presented here, the digital twin uses real time surface along with the details of bottom hole assemblies and drillstring to predict downhole drilling dynamics responses. Detailed information of the formation, drill string, bottom hole assembly and drill bit mechanics are utilized. A fast-running time domain models based on mixed multi-body mechanics and finite element methods form the basis of the digital twin. The workflow is built to automatically recognize drilling rig states (rotary or slide drilling) and connection makeup to start and stop the predictive model. The real time integration has been tested over several wells for stability and performance metrics. With the integration of real time data, the digital twin systematically predicts ahead the true WOB, bit RPM and downhole MSE. The integration of surface data to real time models makes it a true digital twin.