Ensuring Wellbore and Equipment Integrity via Data-Driven Adaptive Limit Modeling for Pressure Management and Stuck Pipe Prevention

Drilling automation can greatly contribute to maintaining process and equipment safety during well construction, paving the way to operational and financial targets. Real-Time estimation and automated monitoring of operational limits for equipment and processes enables early detection of and reaction to potential hazards, e.g. stuck-pipe or over-pressure. This paper discusses a data-driven approach to determine limits for hook-load (HKLD) and stand-pipe-pressure (SPP), and how they are used to automatically prevent abnormal conditions. As HKLD and SPP are constantly changing, the limits must adapt to them, which is a challenge. To achieve that, a statistical analysis of sensor data in real-time is performed to obtain safe operating envelopes (SOE) for both signals, where the outer boundaries of each SOE are henceforth referred to as limits. For HKLD these are derived by analyzing block movement, rig activity, and profile of the sensor signal over different time intervals. To include the drag in the system, pick-up (PU) and slack-off (SO) measurements are overlaid with the obtained limits which are updated after each transition to off-bottom. As soon as the HKLD exceeds a limit, a mitigation process is automatically triggered. The calculation of SPP limits follows a similar approach. However, changes in flow rate are typically quite fast resulting in steep changes of SPP. Hence, determining adaptive pressure limits is even more challenging. Despite the rapid changes in the signals, the automation system can adjust the limits accordingly. The algorithms have been verified and validated with simulated environments and in the field. Field tests have shown that both limits reliably adapt to the real-time signal. As a result, the monitoring algorithms detect overpulls as soon as the HKLD exceeds the upper limit and "set–downs" as soon as it falls below the lower limit. The SPP limits are determined in scenarios with consistent flow rates and used to reliably monitor for potential over-pressures. The limits quickly adapt to steep changes in SPP which is critical when automating the circulation system. In simulation environments where an automated drilling control system (ADCS) is configured by the drilling automation application, abnormal behavior of HKLD and SPP, such as overpulls or pack-offs, were successfully mitigated automatically. This enhances the process of rig automation, increases safety and ensures the integrity of the equipment. The presented approach yielded equivalent results to those of physics-based or machine learning-based approaches and can be made available in real time, which is crucial when an ADCS is used. In addition, other systems only account for static loads or pressure limits according to equipment specifications which does not consider operational constraints of the process. The presented approach provides dynamic limits derived directly from the monitored signals, thus reducing user interaction.