Downhole Pipe Defect Segmentation via Time-Frequency Spectrograms

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for monitoring downhole oil pipe corrosion in wellbores with multiple concentric casings are inefficient, as they rely on manual judgment and lack automated tools for accurate defect detection and quantification, particularly from within the innermost pipe.

Innovation Solution

A downhole logging tool with transmitter and receiver coils generates time-frequency spectrograms, using geometric active contours and level set methods to automatically quantify pipe defect attributes like length and width, enabling precise defect detection and monitoring.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If manual judgment methods are used for defect detection, then device complexity is reduced, but measurement precision and reliability deteriorate

Engineering Contradiction:
Improvesystem complexityVSAvoiddefect detection accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent replaces manual visual inspection with automated image processing and machine learning algorithms. The system uses spectrogram analysis, geometric active contours, and level set methods to automatically detect and quantify pipe defects, eliminating the need for human workers to manually interpret measurements and significantly improving detection accuracy and consistency.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If automated defect detection is implemented, then measurement precision improves, but device complexity increases

Engineering Contradiction:
Improvedefect detection accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs self-analysis through automated image processing algorithms that independently detect, segment, and quantify defects without human intervention. The machine learning models automatically interpret the spectrogram data, calculate defect attributes, and generate results, making the complex system self-sufficient and reducing the need for manual operation.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent segments the defect detection process into distinct computational stages: spectrogram generation, image preprocessing, geometric active contour segmentation, level set method refinement, and attribute extraction. This modular segmentation of the automated system makes the complex process more manageable and implementable through coordinated software modules.

Inventive Principle:
Principle #1Segmentation

3Ease of operation

If workers perform manual defect analysis, then ease of operation is maintained, but productivity deteriorates

Engineering Contradiction:
Improveoperational simplicityVSAvoidinspection efficiency
Core Design Contradiction:
Ease of operationVSProductivity

Solution Approach 1:

The patent replaces the manual inspection process with automated computational analysis. The system rapidly processes spectrogram data and generates defect assessments without human intervention, dramatically increasing inspection throughput and productivity while maintaining ease of use through automated result generation.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

4Device complexity

If manual defect quantification is used, then device complexity is reduced, but loss of information increases

Engineering Contradiction:
Improvesystem complexityVSAvoiddefect attribute accuracy
Core Design Contradiction:
Device complexityVSLoss of information

Solution Approach 1:

The patent replaces manual estimation of defect attributes with automated image analysis that precisely measures defect length, width, area, and other characteristics from the spectrogram. The geometric active contours and level set methods accurately quantify defect geometry, preserving all relevant information without human error or subjective judgment.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This approach allows for automated, accurate detection and quantification of pipe defects, improving the monitoring of downhole corrosion and enhancing the integrity of well operations by reducing manual error and enhancing the reliability of defect analysis.

Implementation Method 1

A downhole logging tool with transmitter and receiver coils generates time-frequency spectrograms

Methodology Applied
Scientific EffectElectromagnetic induction: Electromagnetic Induction

Data Source

PatentUS11333013B2Segmentation of time-frequency signatures for automated pipe defect discrimination
Publication Date: 2022.05.17 HALLIBURTON ENERGY SERVICES INC
  • US11333013B2 patent drawing
  • US11333013B2 patent drawing
  • US11333013B2 patent drawing

AI summary

A method for taking measurements in a wellbore including lowering a downhole logging tool in the wellbore; obtaining, via the logging tool, a plurality measurements of a downhole element; processing the plurality of measurements to obtain a plurality of processed measurements; calculating a contour model based on the plurality of processed measurements; and determining a parameter of the downhole element based on the contour model.