Automated Mud Check with an AI-Enhanced Automated Mud Skid: Results of a Long-Term Permian Field Trial

Abstract Accurate, consistent, and frequent measurement of circulation fluid properties is essential for optimum well construction. A full mud check is conducted by mud engineers typically twice a day using equipment that is now many decades old, providing data that is often of questionable quality and reliability. Sporadic and inaccurate mud check information can be the cause of fluid-related non-productive time and increased drilling costs associated with e.g. wellbore instability, lost circulation, well control and hole cleaning problems. As an alternative to manual measurements, an AI-Enhanced Automated Mud Skid (AMS) is introduced here. The AMS digitalizes and automates mud measurements, monitoring, and reporting. The current design is able to provide more than 30 mud checks at elevated temperature during a 24-hour period. The AMS uses the data obtained from various inline sensors (i.e., Coriolis flowmeter, pressure and temperature sensors, oil-water cut analyzer, etc.) together with machine learning algorithms to report automated mud check results for water and oil-based drilling fluids. During a field trial in the Permian basin with an independent operator to evaluate the overall reliability and accuracy of the system, the unit provided real-time measurements of non-Newtonian YPL viscosity, rheological parameters (3, 6, 100, 200, 300 and 600 rpm readings), friction factor, critical Reynold's number, pressurized density, and oil-water ratio with temperature control up to 150°F. The system also provides remote equipment control, system cleaning, health monitoring, real-time read-out of mud properties, and the connection to a real-time hydraulics software for ECD calculation and drilling margin control. During the field trial, the unit performed more than 500 WBM tests at 120°F and 1000 OBM tests at 150°F with an average accuracy of 95% to estimate rheological properties, density, and oil/water ratio of drilling fluids. Overall operational reliability was calculated as 93% taking into account mechanical failures and lessons learned from the initial field trial phase. The AMS provided benefits by avoidance of excess chemical mud treatment and dilution, efficiency gain with increased ROP from improved hydraulics, decreased fluid-related non-productive time (NPT) and trouble cost, improved consistency and accuracy of drilling fluid measurements, and novel ways of working for the mud engineer. The technology described here opens the door to fully automated fluid monitoring, maintenance, and solids control in well construction operations. This was shown to benefit well construction efficiency (e.g., lowering drilling costs by better cuttings transport), quality, productivity (e.g., reducing reservoir impairment from unwanted solids), and safety (e.g., lowering the incidence of well control and lost circulation events due to better equivalent circulating density (ECD) management in narrow margin drilling environments).