Locating Partial Discharges in Power Transformers with Convolutional Iterative Filtering

Area of Science:

• Electrical Engineering
• Materials Science
• Condition Monitoring

Background:

• Transformer insulation failure is a major cause of electrical grid disruptions.
• Partial discharge (PD) is a key indicator of insulation degradation.
• Accurate localization of PD events is essential for timely maintenance and preventing catastrophic failures.

Purpose of the Study:

• To develop and validate advanced algorithms for pinpointing the location of partial discharge (PD) events within transformers.
• To improve the accuracy and reliability of PD source localization using ultra-high frequency (UHF) sensor data.
• To offer a more effective diagnostic tool for transformer health assessment.

Main Methods:

• Review of existing signal arrival time determination and source localization techniques for PD events.
• Proposal of a novel Convolutional Iterative Filter with Thresholding (CIFT) algorithm for precise signal arrival time detection.
• Development of a reference table of travel times for accurate PD source location resolution.
• Experimental validation using laboratory-generated PD events and simulated data from operational transformers.

Main Results:

• The proposed CIFT algorithm demonstrates superior performance in determining PD signal arrival times.
• The new localization approach significantly outperforms existing methods in accuracy.
• Location accuracy improvements are particularly pronounced (up to 3.7X) for PD events originating far from the UHF sensors.
• The methodology proves effective for both laboratory and simulated real-world transformer conditions.

Conclusions:

• The developed algorithms offer a substantial advancement in the capability to locate partial discharge sources in transformers.
• This enhanced localization accuracy enables more effective and targeted maintenance strategies, reducing transformer downtime.
• The proposed method represents a significant step forward in non-invasive transformer condition monitoring and fault prediction.