Casing Collar Locator Depth Control via Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for wireline operations in the oil and gas industry lack accurate and efficient methods for real-time tool-string depth estimation and depth control, leading to inaccuracies due to thermal and elastic deformations.
Innovation Solution
A system utilizing machine learning and Bayesian sensor fusion to automate casing collar detection, matching, and depth control, providing real-time tool-string depth estimates with associated uncertainty.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual depth correlation and matching is used, then operator control is maintained, but depth estimation accuracy deteriorates due to trial-and-error methods
Solution Approach 1:
The patent replaces manual operator control with an automated machine learning system that uses sensor fusion and pattern recognition to perform depth correlation and matching. The ML model automatically processes CCL signals, depth stamps, and prior information to generate accurate depth estimates without requiring operator intervention in the correlation process.
Solution Approach 2:
The system performs self-correction of depth estimates by automatically detecting casing collars, correlating them with prior information, and adjusting depth measurements in real-time. The automated framework continuously refines depth accuracy through sensor fusion without external intervention, making the system self-sufficient in maintaining measurement precision.
2Measurement precision
If integrated depth measuring wheel device is used, then continuous depth measurement is achieved, but measurement accuracy deteriorates due to thermal and elastic deformations
Solution Approach 1:
The system uses feedback from multiple sensors including CCL detectors, depth stamps, and prior information to continuously correct depth measurements. The ML model processes this feedback loop to identify and compensate for errors caused by thermal and elastic deformations, maintaining accurate depth estimation despite environmental factors affecting the measuring wheel.
Solution Approach 2:
The patent introduces an intermediary ML-based sensor fusion framework that mediates between the physical measuring wheel and the final depth measurement. This intermediary layer processes raw measurements through pattern recognition and correlation with prior information, filtering out errors introduced by thermal and elastic deformations before producing the final corrected depth estimate.
3Measurement precision
If automated machine learning framework is used, then depth estimation accuracy is improved, but system complexity increases
Solution Approach 1:
The patent implements a universal ML framework that handles multiple functions including CCL detection, depth correlation, pattern recognition, and error compensation within a single integrated system. This multi-functional approach consolidates what would otherwise require separate systems, managing complexity while maintaining high measurement precision through sensor fusion and automated processing.
Data Source
AI summary
Systems and methods for estimating tool string depth can include a framework that utilizes machine learning to determine depth and depth uncertainties for the tool string. The framework can take as inputs a casing collar locator signal, a depth, a cable speed, and a timestamp. Then a collar detector function, which can be a machine learning model, can detect a collar and output a certainty level associated with the detection. A collar identifier function can combine that certainty level with a prior collar map and other prior parameters to identify a particular collar and that collar's depth. Then a fusion function can output a depth and depth uncertainty for the collar or for the tool string.