REAL-TIME AUTOMATIC PREDICTION, DETECTION, AND MITIGATION OF FLUID LOSS DURING DRILLING OPERATION EMPLOYING ALONG STRING MEASUREMENT (ASM) DATA ALONG WIRED DRILL PIPES BY USING DIGITAL TWIN IN NORWEGIAN CONTINENTAL SHELF

To automate drilling operations, reliable real-time downhole data, as well as knowledge to extract information about the bottom hole environment, are essential. Numerous dynamics, such as drilling bits, fluid flow, and drill string dynamics, can be employed to monitor the downhole environment. Among the many drilling parameters measured by ASM and relevant to these dynamics, fluid pressure is the key measurement, conveying valuable information regarding drilling hydraulics in real-time operations. Using ASM data, the current study uses the digital twin technique for continually monitoring downhole conditions to discern the onset and progression of fluid loss. An authentic real-time approach for monitoring and control of downhole fluid hydraulics should incorporate both model physics and downhole measurements. An innovative physics-based data analytics model has been developed to simulate fluid flow dynamics during drilling operations. Physical models are formulated as differential equations of mass and momentum conservation, which are discretized along the borehole utilizing finite element method. As a data analytics tool, the finite element solution is integrated with Kalman filtering to estimate downhole parameters in real-time. The combination of finite element and Kalman filtering algorithms (FE-KF) offers the most efficient way of dealing with fast and continuous streams of noisy ASM data. Using this approach in conjunction with high-frequency ASM data, flow rate and pressure along the borehole domain can be predicted in real-time. Estimates of flow rate and pressure are compared to baseline values provided by surface facilities. By identifying all contributing factors to pressure drop across the annulus, an estimate of the quantity of fluid lost can be calculated online. In this approach, a series of ASM data from a well on the Norwegian Continental Shelf (NCS) is used to identify and monitor the initiation and progression of fluid loss far in advance of detection on the surface. Current industry solutions for monitoring downhole pressure are based on assumptions (such as mono-bore wells, simple drill strings, and Newtonian drilling fluids) that result in an unrealistic perception of the downhole environment. Furthermore, all monitoring strategies are delayed and only informative to a single point, namely the location of the pressure sensor. As a complement to such approximate solutions, these methods often make use of unreliable surface data. An intriguing aspect of this research is using digital twin to introduce an algorithm capable of overcoming the shortcomings of previous methodologies while also employing an intelligent and delicate application of ASM data to predict and mitigate downhole drilling events in real-time.