Electrical Imaging Logging for Accurate Shale Laminaset Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for identifying formation interfaces, particularly bedding planes, in electrical imaging logging are inaccurate due to subtle gray scale changes and weak edges, and are time-consuming, while methods like Hough transform and core analysis are laborious and limited to cored intervals.
Innovation Solution
A method involving preprocessing electrical imaging data to generate polar plate images, using K-means++ clustering, circularly connected component labeling, and sinusoidal bedding interface fitting with a modified Levenberg-Marquardt algorithm to accurately extract and identify shale laminaset types.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If image processing methods such as edge detection, image segmentation, and edge tracking and matching are used to identify formation interfaces, then the interfaces can be detected automatically, but the extraction accuracy is poor due to subtle gray scale changes and weak edges
Solution Approach 1:
The patent transforms the electrical imaging logging data from grayscale intensity values to hue values in the HSV color space. This parameter transformation changes the basis of interface detection from subtle gray scale variations to more distinct color hue variations, thereby improving the accuracy of formation interface extraction while maintaining automatic detection capability
Solution Approach 2:
The patent replaces traditional edge detection algorithms with a hue-based clustering approach. Instead of relying on mechanical edge detection operations that fail with weak edges, the method uses unsupervised learning clustering on hue values to identify interfaces, achieving both automation and high accuracy
2Measurement precision
If Hough transform method is used to extract bedding interfaces, then the trajectories can be identified, but the process is time-consuming and laborious and cannot handle low-quality images
Solution Approach 1:
The patent replaces the Hough transform method with an unsupervised learning clustering algorithm that operates directly on hue values. This substitution eliminates the time-consuming nature of Hough transform while maintaining the ability to identify bedding trajectories, and additionally handles low-quality images more effectively through the robustness of clustering methods
Solution Approach 2:
The patent changes the input parameter from grayscale intensity to hue values, and replaces the transformation-based Hough method with a clustering-based approach. This parameter and method change reduces computational complexity and processing time while improving robustness to image quality variations
3Measurement precision
If type division is achieved by means of cores, microslices, and X-ray diffraction, then accurate classification can be obtained, but the process is time-consuming and laborious and can only describe cored intervals
Solution Approach 1:
The patent creates a virtual classification system that uses electrical imaging logging data (a non-invasive copy of formation properties) to identify laminaset types, replacing the need for physical core samples. This copying approach maintains classification accuracy while dramatically improving productivity by eliminating the time-consuming core sampling and laboratory analysis processes
Solution Approach 2:
The patent replaces physical core analysis methods (microslices, X-ray diffraction) with an automated image-based clustering system. This substitution eliminates manual laboratory work and extends classification capability from only cored intervals to the entire logging interval, significantly improving efficiency
Data Source
AI summary
The present application relates to a method for identifying a type of a shale laminaset based on electrical imaging logging, includes loading and decoding electrical imaging data, and preprocessing the electrical imaging data to generate electrical imaging polar plate image data. An image center cluster is initialized by K-means++ algorithm, finding a new cluster center by iteration according to a mean value after calculating a Euclidean distance, and completing clustering after a threshold or a limit of times is reached. N clustered values are binarized to obtain an end-to-end connected matrix conforming to a principle of wellbore imaging, then a circularly connected component labeling algorithm is carried out, and sinusoidal bedding interface fitting is performed after extracting pixels of laminae by a bedding clustering algorithm. The laminae of a shale gas reservoir is then divided based on electrical imaging logging clustering response characteristic value ranges for different laminasets.