METHOD FOR PARTIAL DISCHARGE RECOGNITION IN HIGH VOLTAGE APPLICATIONS AND UNIT USING THE METHOD

Abstract

A method for partial discharge recognition in high voltage applications and a high voltage unit using the method, includes the steps of detecting a signal, transforming the signal from time to frequency domain, cutting frequencies above a defined threshold, and retransform the truncated signal from frequency to time domain. The information content of detected and truncated signals is determined and compared.

Claims

1. A method for partial discharge recognition in high voltage applications, comprising: detecting a signal; transforming the detected signal from time domain to frequency domain; cutting frequencies above a defined threshold resulting in a truncated signal; retransforming the truncated signal from frequency domain to time domain, and determining and comparing information content of the detected signal and of the truncated signal.

2. The method according to claim 1, further comprising: signals with comparable information content of the detected signal and the truncated signal as noise, and/or identifying signals with lower information content of the truncated signal compared to information content of the detected signal as a signal including partial discharge information.

3. The method according to claim 1, wherein the information content of the detected signals and the truncated signals are calculated, and/or comparison of information content is performed by subtraction and/or division.

4. The method according to claim 1, wherein determination and/or comparison of the information content of the detected signals and/or the truncated signals are performed in real time.

5. The method according to claim 1, wherein the method steps are repeated over a fixed time period, to increase reliability of results.

6. The method according to claim 1, wherein the determination of information content is performed by a calculation of Shannon entropy.

7. The method according to claim 1, wherein the detected signal is and/or comprises a voltage signal and/or a current signal.

8. The method according to claim 1, further comprising: producing and/or sending an alarm signal, at partial discharge recognition and/or identification.

9. The method according claim 1, evaluating a value of partial discharge by an iteration method, deleting signals identified as noise, and/or identifying a cut off frequency, above which signals are identified as noise.

10. A high voltage unit using a method according to claim 1 for partial discharge recognition, the unit comprising: an instrument transformer, a high voltage switch, a surge arrester, a bushing, and/or a medium or high voltage transformer.

11. The high voltage unit according to claim 10, wherein the unit comprises an isolator, or an isolator housing with rips on an outer surface.

12. The high voltage unit according to claim 10, wherein the unit is designed for voltages in a range of more than a kilovolt up to 1200 kV.

13. The method according to claim 2, identifying signals with equal information content of the detected signal and the truncated signal as noise.

14. The method according to claim 4, wherein determination and/or comparison of the information content of the detected signals and/or the truncated signals are performed in real time, automatically.

15. The method according to claim 4, wherein determination and/or comparison of the information content of the detected signals and/or the truncated signals are performed in real time, automatically, with a computer and/or online.

16. The method according to claim 5, wherein the method steps are repeated periodically over a fixed time period, to increase reliability of results.

17. The method according to claim 8, wherein the alarm signal is sent online via internet and/or mobile network.

18. The high voltage unit according to claim 10, wherein the unit is designed for voltages in a range of 15 kilovolt up to 1200 kV.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The present invention is further described hereinafter with reference to an illustrated embodiment shown in the accompanying drawings, in which:

[0029] FIG. 1 illustrates schematically a measured signal transformed to the frequency domain, with partial discharge signal, and

[0030] FIG. 2 shows schematically a measured signal transformed to the frequency domain, completely composed of noise.

DETAILED DESCRIPTION OF INVENTION

[0031] In FIG. 1 a measured signal of partial discharge, transformed to the frequency domain is shown. The signal is measured for example in form of voltage and/or current with time. An alternating voltage and/or current is introduced by partial discharge events for example between two electrical poles, with isolator in-between, particularly an isolator of a high voltage device respectively application/unit. As described before, the partial discharge signal is small compared to the high voltage between the poles, with high voltage for example in the range of some kilovolt up to 1200 kV, compared to some picocoulomb signals of partial discharge, i.e. small voltage and current signals. Partial discharge occurs for example across the outer surface of a solid isolator. A solid isolator for example consists of or comprises ceramic, silicone and/or composite materials. The isolator is particularly part of a high voltage unit, for example an instrument transformer, switch, bushing, transformer and/or search arrester. The isolator is for example designed in form of a housing and/or mounting structure of the high voltage unit for example in hollow cylindrical form with rips on the outer surface. The partial discharge also occurs for example in fluidic isolators, particularly oil of a high voltage units.

[0032] The partial discharge is a localized dielectric breakdown of the electrical isolation at high voltages between the electrical poles. The partial discharge at or in the isolator starts for example within voids, cracks, contaminants or inclusions, particularly at conductor-dielectric interfaces, and in liquids partial discharge starts particularly in bubbles, contaminants and/or inclusions. The partial discharge is only limited to a portion of isolation. The discharge is only partially bridging the distance between the poles respectively between electrical conductors. As the partial discharge is initiated, high frequency transient current pulses appear and persist for nano- to microseconds. Then the current pulses disappear and reappear repeatedly as for example a sinewave passes through the zero crossing. The partial discharge signals are short in duration and exhibit rise times of currents in the range of nanoseconds. High levels of electrical background noise, for example due to corona, crosstalk and other effects, make a correct measurement and simple recognition and determination of partial discharge within a measured signal of high voltage/currents difficult.

[0033] The measurements of a signal S, for example current and/or voltage signal, is performed with time. Due to the short duration and due to rise times in the range of nanoseconds of partial discharge currents in high voltage applications, a visualization and distinguishing between noise and partial discharge in voltage/current against time plots is difficult. To identify partial discharge and/or to determine a value of partial discharge, according to the present invention the information content I of a signal S is investigated. Measured signals with time S(t) are transformed from time to frequency domain for example by Fourier Transform and/or Discrete Cousin Transform. Frequencies f above a predefined frequency limit f.sub.lim are cut off and truncated signals are obtained, by removing the signal components at frequencies higher than f.sub.lim. The value of frequency limit f.sub.lim is for example defined every time at the beginning of a monitoring activity respectively measurement. In a step after removing frequencies f above a predefined frequency limit f.sub.lim, signals are transformed back respectively retransformed from frequency to time domain.

[0034] In FIG. 1 an example of a partial discharge signal S(f) in frequency domain is shown, i.e. after transformation from time domain, where each component of the signal S(f.sub.i) has an information Ii. For comparison, FIG. 2 shows a signal S(f.sub.i) completely composed of noise. There, each component S(f.sub.i) has no information, because the signal is completely casual. The signal information I is for example obtained as

[00001]${\displaystyle \underset{-\infty }{\overset{+\infty }{\int }}{\text{I}}_{\text{i}}df}=I$

i.e. by adding the information associated to each component. As seen in FIG. 1 and FIG. 2, partial discharge signals contain an information, while noise signals do not contain information. Noise signals are completely uncorrelated among themselves. By switching from the time domain to the frequency domain, each component has a part of the global information I.sub.i. By cutting off parts of the signal above frequency limit f.sub.lim, information is lost, if the measured signal comprises partial discharge and does not consist only of noise. According to the present invention the method respectively algorithm is to extract the information from the sampled signals, for example by calculating the Shannon entropy, and compare the information content of the sampled signal with the information content of the truncated signal, derived by cutting off frequencies of the signal above frequency limit f.sub.lim.

[0035] The information content of the measured respectively sampled signals is calculated, particularly by a computer locally and/or in the cloud, as described above for example by calculating the Shannon entropy, and the information content of the truncated signal respectively signal after cutting off frequencies is calculated, particularly by a computer locally and/or in the cloud, likewise as described above for example by calculating the Shannon entropy. Both calculations are for example performed in time with measurements or separately one after another, or in a predefined order. After removing frequency components, a time domain signal for the truncated signal is obtained for example by the inverse integral transform, particularly by Fourier Transform and/or Discrete Cousin Transform. The determination of information content of measured and truncated signals is analogous, for example in both cases by calculating the Shannon entropy.

[0036] A comparison of information content of the measured and truncated signals results according to the present invention in a recognition and/or determination of partial discharge. A comparison is performed for example manually and/or automatically, for example by a computer and/or in the cloud. Methods to compare signals include for example division and/or subtraction of the information content of the measured and truncated signals. For division a result is

[00002]$R=\frac{I_{truncatedsignal}}{I_{originalsignal}}<1$

smaller than one, if a measured signal comprises partial discharge. Without partial discharge is

[00003]$R=\frac{I_{truncatedsignal}}{I_{originalsignal}}~1$

a division result, which is exactly or mainly one, since a signal consists only of noise, and no information is lost by cutting off frequencies.

[0037] A recognition and/or determination of partial discharge has advantages as described above, for example a warning can be provoked and/or transmitted and further actions triggered, for example switching off voltage and/or disconnecting high voltage units from the grid, to avoid damage and/or destruction/failure of the units. Partial discharge measurements are also providing quality assessment and diagnostic results. The partial discharge is for example an indicator for the status of the apparatus under test. It is performed in factory quality tests, to show proper function of produced units. In operation in the grid, with a lot of background noise, an online monitoring on high voltage units, for example high voltage switches, bushings, and/or transformers, is possible.

[0038] A detection of signals is performed for example based on Transient Earth Voltages, i.e. induced voltage spikes on the surface of surrounding grounded metalwork. Further methods to detect and measure partial discharge use ultrasonic sound sensors, ultra-high frequency sensors/antennas, high frequency current transformer sensors, and/or directional couplers.

[0039] The above described embodiments of the present invention can be used also in combination and combined with embodiments known from the state of the art. For example, partial discharge can be detected by once performing the steps, particularly all steps as described before. A higher reliability is reached by performing the steps, particularly all steps as described before, repeated. An iteration method to determine partial discharge can include determining a value of frequency limit f.sub.lim. An iteration method can comprise repeating steps according to the present invention, and changing cut off frequency limit f.sub.lim, till a change of Shannon entropy occurs and/or no change of Shannon entropy occurs with changed frequency limit f.sub.lim.