A Method for Detecting and Solving Auto Driller Instability: A Real-Time Control System Solution

Abstract Multiple tool failures on projects across the globe led a multidomain team to collaborate with an operator to explore unconventional sources for causes and solutions. This is the narrative of the development, implementation, validation, and deployment of a novel solution to reduce auto driller instability. These systems are prone to "hidden" instability where large variations in parameters such as penetration rate, WOB, differential pressure, torque, and rotary speed are observed. The main roles of an auto driller are to control drilling parameters, reduce operational variations, and optimize drilling. System instability reduces drilling performance and the lifespan of surface and downhole tools. Service company and operator subject matter experts worked together to develop a method to detect instability. The resulting algorithm is used to monitor all the parameters affected by the auto driller system. Data from multiple sources, such as surface, downhole, and control system settings are aggregated and passed to a software plugin that enables system dysfunction to be identified. The user is presented with a wide range of alarms, flags, and key performance indicators (KPIs). Multiple wells were monitored in real time using this algorithm. Potential dysfunction flags were passed to the operator's and service company's real-time operations centers (ROCs) to validate the output before suggestions were sent to the rig. This workflow identified multiple cases of instability and system shutdowns that went unnoticed for several minutes on the rig. In some cases, as much as 10% of drilling time per stand could be recovered from the system inefficiency. The deployment results were reviewed by a team of domain experts to validate the events that were caught and those that were missed, and then the algorithms were improved iteratively. The system worked well with a fit-for-purpose real-time monitoring workflow producing near perfect results. Because it would be distracting to send the raw signals directly to the drillers at this point, more work needs to be done to make the interface available to the drill teams directly at the wellsite. It was found that reducing peak stress on components contributed to fewer downhole failures. The first trials were too short to quantify the effect on rig equipment life but reducing the peak torque on the top drive and motor current on the drawworks should lead to reduced failure rates for these tools. This information can also be sent to the rig contractor to quantify the value of upgraded control systems.