Substation Transformer Predictive Maintenace: Partial Discharge Detection
SUMMARY
Pre-processing frequency translation algorithm choice
Partial Discharge Characterization
Partial Discharge Pattern Fingerprinting
WideBand Vs. Narrowband transducer options.
In the first part of this white paper, examples of efficient detection and recognition technologies for partial discharge (PD) sources using acoustic emission (AE) method with advanced ultrasound signal processing were presented.
High-resolution Multiple Signal Classification algorithm, our best choice for the frequency analysis, effectively detects in registered AE signal (usually sunk in noise) frequency components corresponding to partial discharges.
On the basis of selected digital signal processing methods and complex model research, a database with PD type classification with different Patterns was shown in order to prepare Predictive Maintenace models.
An important conclusion of this characterization and modeling research is that each of the investigated insulation defects (PD sources) generates repeatable, characteristic and unique acoustic emission signals that can be analyzed as a fingerprint.
The increase of different PD pattern detection and recognition efficiency can be improved by simultaneous application of narrow (20÷100 kHz) and wideband (100÷900 kHz) ultrasound sensors.
On the second part of this paper, a proposal for a proof of concept combines Acoustic emission sensors (ultrasound) for partial discharge analysis with NIR (Thermal image sensors) for thermal stress analysis, with other -more common-sensors in order to provide a Sensor Fusion analysis.
Click on the link to get the Substation Transformer Partial Discharge Paper:
