[Quantitative analysis of transformer oil dissolved gases using FTIR]

Area of Science:

• Analytical Chemistry
• Spectroscopy
• Infrared Spectroscopy

Context:

• Transformer dissolved gas analysis (DGA) traditionally relies on chromatography, which has limitations including carrier gas requirements, regular calibration, and safety concerns.
• Existing DGA methods face challenges with small gas volumes, multiple components, detection limits, and interference from other gases.
• The need for a safer and more reliable online monitoring system for transformer health is critical.

Purpose:

• To develop an online dissolved gas analysis system for transformers utilizing Fourier transform infrared spectroscopy (FTIR).
• To address the limitations of chromatographic methods in DGA, such as carrier gas dependency, calibration needs, and safety issues.
• To establish a robust quantitative analysis model for characteristic gases in transformer oil.

Summary:

• A novel DGA system based on FTIR was developed to overcome the drawbacks of chromatography.
• A quantitative analysis model was created using sparse partial least squares, piecewise section correction, and feature variable extraction with improved TR regularization.
• The system effectively analyzes characteristic gases like methane (CH4), ethane (C2H6), and carbon dioxide (CO2) with high accuracy.

Impact:

• The FTIR-based DGA system meets essential requirements for transformer monitoring, offering improved safety and efficiency.
• This technology provides a viable alternative for real-time transformer health assessment, reducing operational risks.
• The developed quantitative model demonstrates the potential of FTIR in complex gas mixture analysis for industrial applications.