Advanced Detection of Sucker Rod Pump Faults Using Computer Vision and Dynamometer Card Analysis

Abstract This research explores a new method for monitoring and diagnosing oil wells in artificial lift systems, particularly those using sucker rod pumps (SRP). By combining dynamometer cards, computer vision, and deep learning techniques, we aim to enhance traditional dynamometer card analysis. Our approach involves advanced computer vision and deep learning algorithms to automate interpretation and improve accuracy in identifying issues such as fluid pound, gas interference, and tubing leaks.This method goes beyond relying solely on textual data, which is often limited. We use state-of-the-art deep learning techniques in computer vision to quickly and accurately identify issues, enabling real-time monitoring and predictive maintenance. Testing has shown promising results, with a fault detection accuracy rate of 75-94% in various fault types from dynamometer card images.