Several common partial discharge detection methods

Click: 312

1. Ultrasonic detection method

The ultrasonic sensor fixed on the transformer oil pillow wall can receive the ultrasound generated by the internal partial discharge of the transformer, thereby detecting the size and position of the partial discharge. The commonly used ultrasonic sensor is a piezoelectric sensor, with a frequency range of 7O-l50KHz, in order to avoid the magnetic noise of the iron core and the mechanical vibration noise of the transformer. Due to the low electrical interference caused by ultrasonic method and its widespread application in partial discharge localization, people have conducted in-depth research on ultrasonic method. However, the internal insulation structure of transformers is complex, and the attenuation of sound waves and the impact on sound speed by various sound media are different. The currently used ultrasonic sensors have poor resistance to electromagnetic interference and low sensitivity, which increases the difficulty of detection. In recent years, due to the improvement of the efficiency of acoustic and electrical energy conversion components and the development of electronic amplification technology, the sensitivity of ultrasonic testing has been greatly improved. Therefore, the development and application of this method is very promising

2. Photometry

It is performed using light radiation generated by partial discharge.

. In transformer oil, the wavelengths of light emitted by various discharges are different. Research has shown that, Usually at 500 Between~700mm. After photoelectric conversion, partial discharge identification can be achieved by detecting the photocurrent characteristics. Although significant progress has been made in analyzing partial discharge characteristics and insulation degradation mechanisms using optical measurement methods in the laboratory, due to the complexity and high cost of optical measurement equipment, low sensitivity, and the need for the detected substance to be transparent to light, it is not possible to be widely applied in practice.

3. Chemical detection

When partial discharge occurs in a transformer, various insulation materials will undergo decomposition and destruction, producing new products. By detecting the composition and concentration of the products, the state of partial discharge can be determined. At present, this method has been widely applied in online fault diagnosis of transformers. The pattern recognition system established based on the different types and degrees of faults, as well as the composition and concentration of gases, can achieve automatic fault identification. But until now, there has been no unified standard for judgment. Because it is more sensitive to detecting early latent faults, but cannot reflect sudden faults

4. Pulse current method

Pulse current method is used to obtain the actual discharge amount by detecting impedance, grounding wire of transformer casing, grounding wire of casing, grounding wire of iron core, and pulse current caused by partial discharge in winding.

. It is the earliest and most widely used detection method. International specialized agencies have also developed specialized standards for this, and the current sensor can usually be divided into narrowband and broadband according to the frequency band. Narrowband sensors are generally around 10KHZ, with advantages such as high sensitivity and strong anti-interference ability, but the output waveform is severely deformed. The broadband sensor has a bandwidth of around IOOKHZ, which has the advantage of high pulse resolution but low signal-to-noise ratio. The main drawbacks of this method are: firstly, the detection impedance and amplifier have an impact on the sensitivity, accuracy, resolution, and dynamic range of the measurement. Therefore, when the capacitance of the sample is relatively large, the sensitivity of the DC resistance fast tester is also limited by the coupling impedance. Secondly, the testing frequency is low, generally less than 1MHZ, thus containing less information. Thirdly, it has high sensitivity in offline state, while in the field it is susceptible to external interference noise and has poor anti-interference ability