Beam Pump Optical Leak Detection on the Polished Rod
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing beam pumps suffer from unreliable mechanical solutions for detecting oil leaks, leading to potential contamination and damage, with existing automatic systems being prone to false alarms and inefficiencies.
Innovation Solution
A computer-implemented method using a camera to capture an image sequence of the linkage during pumping movements, identifying fouling by comparing with a known clean state, determining fouling location and variation, and checking for exceeding a threshold to detect oil escapes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual inspection rounds are used to check for oil leaks, then the system is simple and easy to operate, but the detection is not continuous and may miss incipient leaks
Solution Approach 1:
The patent replaces manual mechanical inspection with an automated optical detection system using a camera and image processing algorithm. The camera captures images of the stroke region, and computer vision algorithms automatically identify oil leaks by analyzing pixel variations and comparing against reference images, eliminating the need for human inspectors while providing continuous monitoring.
Solution Approach 2:
The system performs self-diagnosis by automatically analyzing captured images to detect oil leaks. The evaluation unit autonomously processes the visual data, identifies fouling patterns indicative of leaks, and generates alerts without requiring external inspection personnel, enabling the system to monitor itself continuously.
2Reliability
If automatic sensor-based systems are used to detect oil leaks, then continuous monitoring is achieved, but false alarms occur and the system becomes complex
Solution Approach 1:
The patent replaces contact-based mechanical sensors with a non-contact optical detection system. The camera-based approach uses image processing and pattern recognition algorithms to identify oil leaks by analyzing visual characteristics of the stroke region, avoiding the false alarms associated with mechanical sensors while maintaining continuous monitoring capability.
Solution Approach 2:
The system creates a visual copy (image) of the stroke region and analyzes this copy to detect oil leaks. By capturing and processing images of the area, the system can identify leak patterns without physical contact, reducing false alarms from sensor interference while maintaining detection accuracy.
3Measurement precision
If the camera captures the entire stroke region continuously, then complete coverage is achieved, but the data processing load and storage requirements increase
Solution Approach 1:
The patent extracts and focuses analysis only on the relevant stroke region where oil leaks are most likely to occur. By directing the camera to capture specifically this area and using image processing to isolate the stroke region from the rest of the image, the system achieves complete coverage of the critical area while minimizing unnecessary data capture and processing of irrelevant regions.
Solution Approach 2:
The system segments the image data by identifying and focusing on the stroke region specifically. The evaluation unit divides the image processing task into segments, analyzing only the relevant portions containing the stroke region and linkage components, thereby reducing overall data volume while maintaining detection precision for the critical areas.
Data Source
AI summary
A computer-implemented method for detecting an oil leak in a beam pump, wherein, during pumping movement of the rod linkage through the wellhead, a stroke region is formed, the method including a) capturing images of an image sequence containing at least the stroke region of the rod linkage, b) detecting soiling in the stroke region of the rod linkage in comparison with a known clean state for an image in question of the image sequence, c) ascertaining the position of the detected soiling for an image in question of the image sequence by the associated position of the rod linkage, d) determining the change in the soiling at the ascertained position over at least one pumping cycle from the images of the image sequence, e) checking if the change in the soiling exceeds a predefined limit value, and if applicable, outputting a warning of an oil leak.