Gas Flare Discrimination Using Spectral Feature Extraction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing gas analysis systems often mistakenly identify gas flares as gas leaks, leading to inefficient asset operation cessation and resource waste due to the inability to discriminate between the two.

Innovation Solution

A gas flare discrimination system comprising a gas detection sensor and a controller component that processes image data through acquisition, feature extraction, and discrimination models to accurately identify gas flares and initiate appropriate responses.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If gas detection systems use simple detection methods, then the system complexity is low, but the measurement precision is insufficient to discriminate between gas flares and gas leaks

Engineering Contradiction:
Improvegas flare discrimination accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system segments the gas detection process into multiple specialized components: an acquisition model for capturing spectral data, a feature extraction model for identifying gas characteristics, and a gas flare discrimination model for distinguishing flares from leaks. This segmentation allows each component to specialize in one aspect, improving overall measurement precision while making the complex system more manageable and maintainable

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system transitions from traditional single-dimensional gas detection to multi-dimensional spectral analysis by implementing hyperspectral imaging capabilities. This adds spectral dimensionality to the detection process, enabling the system to identify and discriminate gases based on their unique spectral fingerprints across multiple wavelengths, thereby achieving high measurement precision without excessive complexity

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Reliability

If gas detection systems mistakenly identify gas flares as gas leaks, then the detection sensitivity is high, but the loss of time and productivity occur due to unnecessary asset operation cessation

Engineering Contradiction:
Improvedetection accuracyVSAvoidasset operation efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The system implements feedback mechanisms where the discrimination model continuously refines its classification of gas events based on spectral pattern recognition. The system provides feedback loops that allow operators to verify detections and adjust parameters, ensuring high reliability in distinguishing flares from leaks while minimizing false alarms that would disrupt asset operations

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system performs preliminary spectral analysis and feature extraction before making final discrimination decisions. By pre-processing the spectral data and identifying key features early in the detection pipeline, the system can quickly and accurately determine whether a detected gas presence is a flare or leak, preventing unnecessary operational interruptions and maintaining productivity

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4660960A1Systems, apparatuses, methods, and computer program products for gas flare discrimination
Publication Date: 2025.12.10 REBELLION PHOTONICS
  • EP4660960A1 patent drawingFigure 1
  • EP4660960A1 patent drawingFigure 2
  • EP4660960A1 patent drawingFigure 3

AI summary

Systems, apparatuses, methods, and computer program products for gas flare discrimination are provided. An example gas flare discrimination system may comprise at least one gas detection sensor and at least one controller component. In some embodiments, the controller component is configured to obtain image data of a target area. In some embodiments, the controller component is configured to generate, by applying the image data to an acquisition model, gas channel data. In some embodiments, the controller component is configured to generate, by applying the gas channel data to a feature extraction model, gas feature data. In some embodiments, the controller component is configured to generate, by applying the gas feature data to a gas flare discrimination model, a gas flare identification flag. In some embodiments, the controller component is configured to initiate performance of one or more gas flare response actions based at least in part on the gas flare identification flag.