Case Studies of a Novel Digital Twin System for Real-Time Early Identification and Warning of Pipe Stuck

Abstract

Abstract Pipe stuck serves as one of the primary downhole accidents in drilling industry with consequences ranging from minor inconvenience to severe complications. Despite the variation on its cause and mechanism, this downhole accident could be prevented or avoided based on early identification of its unique symptom or data behavior, provided intervention or remediation measures are taken promptly. While conventional prediction or recognition methods, which rely on the fluctuation amplitude and curve trends of surface drilling parameters, are subject to user experience and human error, it's often the case that their timeliness cannot be guaranteed. A novel digital twin system, based on the deviation of hook load, torque and SPP (Standpipe Pressure) from theoretical modeled values and ROC (Rate of Change) of these three actual parameters, is developed for real-time early identification and warning of pipe stuck while drilling. A PSRI (Pipe Stuck Risk Index) ranging from 0%-100% is proposed based on ROC and deviation for comprehensive evaluation regardless of drilling state. Both T&D and wellbore hydraulics models are adopted to model the theoretical hook load, torque and SPP within preset computing cycle. ROC and deviation thresholds, corresponding to three different risk levels as low (green), medium(yellow) and high(red), were obtained by using RFA (Random Forest Algorithm) based on data of historical wells with and without pipe stuck accidents. Preventive or intervention measures are generally mandatory for on-site personnels or engineers once high-risk (red) alert is identified by the system. Several case studies are outlined to exhibit that the occurrence of pipe stuck accidents could be detected typically between 30 minutes and 2 hours ahead during back-reaming and tripping operations, which proved its potential for early identification and prevention of this downhole accident in real-time drilling. Cost containment and operational efficiency could be created with system implementation as it helps to reduce analysis time and provide timely proactive action in real-time monitoring, which demonstrates that digital twin driven technique could better capture and utilize digital data to minimize NPT and optimize drilling.