Well Log Change Point Filtering for Accurate Depth Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for aligning well logs face challenges such as pathological alignment due to noise accumulation and require manual adjustments, especially when multiple logs show a particular subterranean feature at different depths, necessitating improved automated depth matching techniques.
Innovation Solution
A method using change point algorithms to detect and remove aberrant signal regions in well logs, including inconsistent and constant data, by identifying change points and adjusting well logs to a common depth reference, thereby enhancing alignment accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multi-pass cross-correlation measures are used to depth-shift well logs, then alignment between well logs is improved, but manual adjustments are still required and noise accumulation occurs
Solution Approach 1:
The patent extracts and removes aberrant signal regions from well logs before performing correlation-based depth matching. By identifying and eliminating noisy segments through statistical analysis and change point detection, the method prevents noise accumulation that would otherwise degrade alignment accuracy in subsequent processing steps.
Solution Approach 2:
The patent performs preliminary cleaning of well log data by detecting aberrant signal regions and removing them before the main alignment process. This preliminary action ensures that the subsequent cross-correlation or dynamic time warping operations work with clean data, eliminating the need for manual adjustments and preventing noise accumulation.
2Adaptability or versatility
If dynamic time warping is used to align well logs, then alignment flexibility is improved, but pathological alignment occurs due to noise accumulation
Solution Approach 1:
The patent extracts and removes aberrant signal regions from well logs before performing dynamic time warping. By identifying noisy segments through statistical analysis and change point detection, the method eliminates the noise that would cause pathological alignment, ensuring reliable results while maintaining the flexibility of DTW.
Solution Approach 2:
The patent performs preliminary cleaning of well log data by detecting and removing aberrant signal regions before applying dynamic time warping. This preliminary action ensures that the DTW algorithm operates on clean data, preventing noise accumulation and pathological alignment while preserving the method's adaptability.
3Productivity
If automated depth matching methods are used, then productivity is improved, but measurement precision deteriorates due to noise in the data
Solution Approach 1:
The patent extracts and removes aberrant signal regions from well logs before performing automated depth matching. By eliminating noisy segments through statistical analysis and change point detection, the method maintains high processing efficiency while significantly improving depth matching precision.
Solution Approach 2:
The patent performs preliminary cleaning of well log data by detecting and removing aberrant signal regions before automated depth matching. This preliminary action ensures that automated methods work with clean data, achieving both high productivity and precise depth matching without requiring manual intervention.
Data Source
Figure 1A~1D
Figure 2
Figure 3A
AI summary
A method includes receiving a well log having a signal. The method also includes identifying in the signal a first change point that demarcates a first signal region and a second signal region. The method also includes determining that the first signal region is inconsistent in comparison to the second signal region. The method also includes producing a modified well log by removing the first signal region from the signal in response to determining that the first signal region is inconsistent in comparison to the second signal region.