Real-Time Mud Motor Stall Detection Using Downhole and Surface Data for Improved Performance Management and Failure Mitigation

Abstract

Mud motors are widely used in directional drilling and their failure during operation leads to costly non-productive time. There is currently no existing literature investigating the correlation between stalls detected using downhole sensors and concurrent signals produced in surface sensor data. Current motor stall detection algorithms using surface sensors are still rudimentary and error-prone. The objective of this study was to develop a robust stall detection algorithm using insights gained from correlating downhole and surface data. Previous studies have indicated that stalls are a major contributing factor to elastomer damage in a mud motor. Using downhole sensor datasets from multiple operators, we first identified all instances of motor stalls. These events are typically characterized by a sudden reduction in downhole vibration and rotation at the bit, accompanied by a sudden increase in torque. We then proceeded to map these stall events to time-synchronized surface data to identify the associated behavior of surface parameters during a stall, noting the differences in behavior during rotary and slide drilling periods. We analyzed 268 distinct stall events in the downhole data as well as several clusters of micro-stalls (characterized by a momentary spike in downhole torque coupled with a downward spike in downhole RPM that last less than one second, typically lasting for a few milliseconds at most). Mapping these events to the surface data helped identify a set of primary signals produced at the surface during every motor stall, and secondary signals that are produced in a majority of motor stalls. The primary surface signals we observed during a stall included a sharp spike in differential pressure and a sharp decline in weight on bit (WOB), typically within a 10-second window. Secondary surface signal observed in over 70% of motor stalls include decrease in rate of penetration (ROP). Statistical analysis of the downhole and surface signals demonstrated a strong correlation (p < 0.05) between the length of a motor stall and the magnitude of the differential pressure increase produced at the surface. Our analysis of stalls in downhole datasets demonstrated that the vast majority of surface-detectable stalls occurred during slide drilling, while rotary drilling contained significant clusters of micro-stall events that were too short to produce identifiable signals at the surface. This study builds upon existing literature and understanding of mud motor failure by correlating time-synchronized events in downhole and surface data. We have compiled our observations to create a comprehensive framework for detecting motor stalls at the surface using a set of surface signals. Our findings establish a robust way to detect motor stalls from surface data, which should be of significant value to lower-cost well construction operations and real-time monitoring of drilling operations.

Positive Displacement Motor Condition and Performance Prediction Using Surface and Downhole Sensor Data

“…The primary cause of elastomer damage is motor stalling; detecting those events provides information regarding the operating parameters and the corresponding motor response. Data-driven methods and machinelearning algorithms classify the failure modes and mitigate the risks [21,22].…”