Digital Twins Revolutionizing Oil and Gas Industry Optimizing Drilling Operations with Physics Informed AI

Abstract Digital technologies have disrupted every industry in recent years, and the oil and gas industry are no exception. One of the most significant advancements is the utilization of digital rig technologies, which have revolutionized drilling processes, monitoring, and well management. By implementing physics-informed-AI based real-time Digital Twins operating at the edge, companies can maximize performance, efficiency, and cost reduction. In this study a digital twin with physics informed AI is presented. An efficient physics-based time domain model coupled with Machine Learning algorithms, forms the core of this innovative approach. The method utilizes performance data from small set of offset wells to continually optimize and update operating conditions. Its objective is twofold: minimizing predicted shock and vibrations and reducing mechanical specific energy while maximizing drilling rates. The real-time edge implementation has demonstrated significant improvements from well to well, including enhanced weight transfer to the bit, consistent top drive RPM performance, and reduced shock and vibrations. As a result, average drilling performance has improved by 40%, translating to expedited drilling operations and decreased overall drilling costs.