Method and assembly for detecting partial discharges of an electrical operating device

Abstract

The invention relates to a method for detecting partial discharges of an electrical operating device which comprises at least one switch (S1, S2, S3), in which for a plurality of potential locations for a partial discharge in the region of the operating device a correction factor table is determined in each case, wherein each correction factor table contains a plurality of correction factors, wherein each of the correction factors is assigned to a combination of switch settings of the operating device, by means of a sensor apparatus (16) electrical impulses are recorded, from among the correction factor tables a correction factor table matching the original location of the impulses is selected, the recorded impulses or a threshold value for the comparison with the impulses are corrected using the correction factor to which the present combination of switch settings of the operating device is assigned.

Description

[0001] 202309462 1 Description Method and arrangement for detecting partial discharges in electrical equipment The invention relates to the evaluation of partial discharge signals that occur in the area of ​​electrical equipment with at least one switch. Partial discharges are local electrical discharges that are caused, for example, in local inhomogeneities in the insulation material, for example in small air pockets, within the insulation of an electrical line as a result of strong inhomogeneities in an electrical field. Partial discharges can damage insulation locally and lead to its failure in the long term. For this reason, electrical machines and switchgear in particular are regularly checked for partial discharges. Partial discharges cause short electrical pulses with a typical pulse duration of less than 1 µs and frequency components well into the UHF frequency range.The pulses typically occur multiple times per sine wave of the mains voltage, for example 50 Hz. There are a variety of measurement methods for detecting partial discharges, some of which are standardized, for example, according to DIN EN 60270. These include offline measurements, which interrupt the regular operation of a piece of equipment and potentially incur high costs, for example due to the operational interruptions and the costs of the measuring devices. Such measurements can therefore only be performed occasionally. Online measurements during regular operation enable continuous monitoring of the equipment and, in particular, the detection of trends regarding partial discharge activity. The information obtained in this way can enable a reliable forecast that predicts the period during which equipment affected by partial discharges can still be operated reliably.To enable a reliable evaluation of the equipment, not only the measurement but also the interpretation of the measured values ​​is necessary. For this purpose, a system can be calibrated during commissioning, for example by providing and measuring an electromagnetic reference pulse. Such a reference pulse can, for example, be a short current pulse for which a charge quantity can be determined in a known manner as the integral of the current over time. A signal determined by the partial discharge measurement can then be associated with the determined charge quantity in terms of its height, area, or other characteristics. A threshold value can also be set; if this is reached or exceeded, a critical condition is recognized.To enable advanced methods such as condition-based maintenance, it is important that threshold values ​​are neither too low (= false alarm) nor too high (= potential destruction of the equipment). Particularly in the case of partial discharges in equipment with one or more switches, the correct selection of threshold values ​​or the correct interpretation of the measured values ​​is made more difficult by the possibility of different switch positions, which can significantly influence the pulses generated by partial discharges. The object of the invention is to provide an improved method and an improved device for detecting partial discharges in electrical equipment, with which increased fault tolerance with regard to the application of the threshold values ​​is achieved. The invention solves this problem by a method according to claim 1. Furthermore, a solution consists in the arrangement with the features of claim 6.202309462 3 In the method according to the invention for detecting partial discharges in an electrical device that comprises at least one switch, a correction factor table is determined for a plurality of potential locations for a partial discharge in the area of ​​the device, wherein each correction factor table contains a plurality of correction factors, wherein each of the correction factors is assigned to a combination of switch positions of the device. Furthermore, electrical pulses are detected by means of a sensor device. A correction factor table suitable for the source location of the pulses is selected from the correction factor tables, and the detected pulses or a threshold value for comparison with the pulses is corrected using the correction factor assigned to the existing combination of switch positions of the device.This advantageously achieves improved accuracy when comparing measured signals from partial discharges with a threshold value, and differences resulting from the current electrical situation, particularly switch positions, are better compensated for. Advantageous embodiments of the invention are evident from the dependent claims. The embodiment of the independent claims can be combined with the features of one of the subclaims or, preferably, with those of several subclaims.Accordingly, the following additional features can be provided: Calibration can be carried out for an initial switch position, in which case the detected pulses or a threshold value for comparison with the pulses are additionally corrected using the correction factor assigned to the initial combination of switch positions of the equipment. This advantageously takes into account the fact that a specific configuration of the switchgear is present even during calibration. In one embodiment of the invention, pulses from a partial discharge source are recorded before and after a change in the switch position from a first to a second switch position. Furthermore, a ratio of the signals before and after the change is then determined.For each of the correction factor tables, a ratio of the correction factors is then calculated from the correction factors for the first and second switch position, and the best match between the ratios of the correction factors and the ratio of the signals is determined. This has the advantage that the location of signals from a partial discharge can be determined by changing the switch position. It is not necessary for the change in the switch position to be brought about for measurement purposes. Furthermore, it is not necessary for the change in the switch position to isolate the partial discharge source from the sensor. Rather, adding or removing current paths that are not directly between the sensor and the partial discharge source is sufficient to determine the location.It is expedient to use vectors, each with a plurality of value pairs of frequency and scalar correction factor, as correction factors. The frequencies used in the value pairs are preferably in a range between 1 MHz and 2 GHz, with the ratio between a highest and a lowest frequency in a vector preferably being at most 100, in particular at most 10. The inventive arrangement for detecting partial discharges in an electrical device having at least one switch comprises a sensor for receiving electrical pulses and an evaluation device with a plurality of stored correction factor tables, each containing a plurality of correction factors.Each of the correction factors is assigned to a combination of switch positions of the equipment, and each correction factor table is assigned to a potential location for a partial discharge. The evaluation device is further configured to select a correction factor table from the correction factor tables that is appropriate for the location of origin of the pulses and to correct the detected pulses or a threshold value for comparison with the pulses using the correction factor that is assigned to the existing combination of switch positions of the equipment. The sensor is preferably a Rogowski coil or a high-frequency current transformer. The invention is described and explained in more detail below with reference to the exemplary embodiments illustrated in the figures.They show: Figure 1 an equivalent circuit diagram of a medium-voltage switchgear with three switches, a sensor for partial discharges and a partial discharge source, Figure 2 a version of the equivalent circuit transformed into a simpler form, Figure 3 schematically shows the sequence of a method for signal correction, Figure 4 schematically shows the sequence of a method for determining the location of signals from a partial discharge source. A partial discharge can be represented in simplified form by the discharge of a partial capacitance between two potentials. A further capacitance and the development of the partial discharge conductivity as a function of the voltage drop across the partial capacitance limit the maximum current and prevent its complete discharge. Like any electrical circuit, the partial discharge circuit must also be closed. This is achieved by the remaining components of the electrical equipment.Figure 1 shows an example electrical equivalent circuit 10 of a switchgear assembly, where a source of partial discharges is present. The resistances shown in Figure 1 are complex impedances 131…137, 141…149, 15. The partial discharge is symbolized by a current source 12. The equivalent circuit symbolizes a single-pole switchgear assembly with three switches S1…S3. The impedances 131…137 represent the impedances of the switchgear assembly and its busbar. The impedances 141…149 are impedances of the connected cables as representations of possible connected equipment. A sensor 16 for partial discharges is also shown. The equivalent circuit 10 shown in Figure 1 can be simplified to an equivalent circuit 20 according to Figure 2 using the applicable transformations. The complex impedances 21, 22 can be calculated from the equally complex impedances 131…137, 141…149.The equivalent circuit 20 shown in Figure 2 makes it clear that the complex impedances 21, 22 influence the reaction of the sensor 16 to partial discharges, since the current signal of a partial discharge is attenuated and filtered (bypassed the sensor 16) and thus does not reach the sensor unchanged, even if the partial discharge occurs locally close to the sensor 16. From the equivalent circuit 20 of Figure 2, a frequency-dependent correction factor K(f) can be determined for an occurring partial discharge. This correction factor can advantageously be applied to a measured signal in order to correct it before it is compared with a threshold value 101. For this purpose, for example, the correction factor K(f) can be multiplied with a measured signal in a frequency-resolved manner before the measured signal is converted into a single evaluation variable. Alternatively, a used threshold value 101 can also be corrected with the correction factor.This results in significantly different values ​​for impedances 21, 22 and for current source 12 for different switch positions of switches S1…S3. Experience has shown that this effect is significant in medium-voltage switchgear, since the impedance of the connected equipment can have a significantly greater influence than that of the relatively small busbar. It is therefore advantageous to use several correction factors K. n (f) where each correction factor K n(f) a specific switch position of the switches S1…S3 is assigned. A corresponding table 30A is as follows: Factor S1S2S3K1(f) OOX K2(f) XXX K3(f) XOX K4(f) OXX K5(f) - - O Where X stands for the switched-on (closed) state of one of the switches S1…S3 and O for the switched-off (open) state. In this example, the sensor 16 is placed away from the third switch S3. If the third switch S3 is open, the switch positions of the first two switches S1, S2 play only a very minor role. Therefore, there is no further case distinction, just the uniform correction factor K5(f). Calibration is typically performed to evaluate partial discharges. During calibration, a partial discharge is simulated, i.e., artificially introduced into the system. The amount of charge introduced is known; it results from the current integral over time.Sensor 16 determines the signal resulting from this artificial partial discharge, and based on the sensor response, a threshold is established at which a critical condition is assumed. In other words, real partial discharges are considered problematic and a warning is triggered. At the time of calibration, switches S1…S3 are in a specific position. Therefore, a correction factor is already present in the conversion between the measured signal and an evaluation value for comparison with the threshold. Since no further correction is necessary, the correction factor is K. n (f) to this switch position 1. For example, if all switches S1…S3 are closed during calibration, K2(f) = 1. The conversion obtained by the calibration can also be converted into the correction factors K n (f) are included, then all K n(f) <> 1. From the equivalent circuit diagrams 10, 20 it can be seen that the correction factors K n (f) also depend on where the partial discharge occurs and where the sensor 16 is located. The location of the sensor 16 is known and does not change. The location of the partial discharges, however, is not known in advance. Therefore, in this exemplary embodiment, further tables 30B, C, D of correction factors K are advantageously used for the invention. nm (f). In general, it is useful to use a table 30A…D of correction factors K nm(f) for each separate source of partial discharges. For each possible change in the power grid in the immediate vicinity, i.e. in this example each switch S1…S3, the partial discharge can occur on both sides of the switch. This results in six sources of origin for the three switches S1…S3. The three sources of origin on the busbar side correspond to one another, leaving four distinguishable 202309462 9 sources of origin. In addition to the area around the busbar, these are the cable areas beyond the three switches S1…S3. The area around the busbar is designated here with the index A. Since in the equivalent circuit diagram 10 of Figure 1, the current source 12, which represents the partial discharges, is located in the area of ​​the busbar, the correction factors K described above are n (f) also K n,A (f). This results in further correction factors K n,B (f), K n,C (f) and K n,D(f) for partial discharges occurring in the cable areas beyond switches S1…S3. It is understood that the correction factors and their relationships to one another can vary greatly, since the partial discharge signals are influenced very differently by the switch positions depending on their location. In other operating environments, however, the correction factors and especially their relationships to one another can also be very similar. In addition to Table 30A given above, further Tables 30B…D are therefore obtained: Assumed location of origin of the partial discharge: In the cable area beyond the first switch S1Factor S1S2S3K 1,B (f) OOXK 2,B (f) XXXK 3,B (f) XOXK 4,B (f) OXXK 5,B (f) - - O Assumed origin of the partial discharge: In the cable area beyond the second switch S2Factor S1S2S3K 1,C (f) OOXK 2,C (f) XXXK 3,C (f) XOX 202309462 10 K 4,C(f) OXXK 5,C (f) - - O Assumed origin of the partial discharge: In the cable area beyond the third switch S3Factor S1S2S3K 1,D (f) OOXK 2,D (f) XXXK 3,D (f) XOXK 4,D (f) OXXK 5,D (f) - - O It is possible that the correction factors are calculated based on the equivalent circuit diagram 20. Alternatively, it is also possible that some or all of the correction factors are estimated. In general, the correction factors are determined before the actual use and are thus taken from a database or other storage device during operation when measuring partial discharges. Figure 3 shows the further procedure schematically. If, during operation of the switchgear, a signal 100 of a partial discharge is measured in a first step 31, then in a second step 32 the signal can be corrected using the correction factors K n,m(f) in Tables 30A…D. For this purpose, a first correction factor K k,m (f) is determined, which is assigned to the switch state during calibration. Furthermore, a second correction factor K a,m (f) is determined, which is assigned to the current switch position. The partial discharge signal is then calculated with the ratio K a,m (f) / K k,m (f). The correction is conveniently carried out in a frequency-resolved multiplication, ie for all f i the signal TE(f i ) with K a,m (f i ) / K k,m (f i) multiplied. In real applications, it is expedient to only consider a section of the frequency spectrum, which typically lies in the range between 1 MHz and several GHz and whose width is typically significantly smaller than the stated range. For example, a specific sensor can consider the frequency range between 10 MHz and 200 MHz and, for this purpose, consider three individual values ​​at 10 MHz, 50 MHz, and 200 MHz. In this range, the signal is considered for a selection of frequencies, so that the signal TE(f) corresponds to a vector with value pairs of frequency and signal strength. For the correction factors, it is expedient to determine and save a similar vector with the same frequencies. The correction described requires a selection of the origin to be used for the partial discharge so that the correct K n,m(f) can be selected, especially with reference to the index m. If this is not yet known, the correction can still be carried out if the correction factors K n,m (f) or the respective relevant ratio are similar to each other. If the correction is carried out as an example with the second and third correction factors as K3 / K2, i.e., with a calibration with closed switches S1…S3 and with the second switch S2 currently open, then the correction is unproblematic if K 3,A / K 2,A ≈ K 3,B / K 2,B ≈ K 3,C / K 2,C ≈ K 3,D / K 2,DIn this case, one of the ratios can be used. After the correction, in a third step 33, the corrected signal 100 is compared with the threshold value 101. If the corrected signal 100 exceeds the threshold value 101, a critical condition is assumed, and in a fourth step 34, for example, a warning signal is triggered. If, on the other hand, the ratios such as K3 / K2 are different from one another, it is expedient to determine the point of origin of the partial discharge. The very fact that the ratios of the correction factors are different for the various assumed points of origin of the partial discharge can advantageously be used to detect the point of origin. This requires a switching action 103 of at least one of the switches S1...S3. Such a switching action 103 can be brought about for measurement purposes or occur during operation for other reasons.Regardless of the cause of the switching operation 103, this leads to a change in the partial discharge signal recorded by the sensor 16. The reason for this is that the current pulse of the partial discharge can take different paths after a switching operation than before, or in other words, in changed impedances in the equivalent circuit 20 of Figure 2. The procedure is shown schematically in Figure 4. Since partial discharges of a partial discharge source typically occur repeatedly, a signal TE. V (f) such a partial discharge before the switching operation 103 and a signal TE N (f) after the switching operation. In a first step 41, a frequency-dependent ratio of the two signals can be formed. As before, the signals are present in a real implementation, for example, as a vector of value pairs. The ratio can thus be formed per frequency, resulting in a ratio vector V(f) with V(fi ) = (f i , V i ) = (f i , TE N,2 (f i ) / TE V,2 (f i )), where TE N,2 (f i ) and TE V,2 (f i ) the signal components of the respective pair of values ​​at the frequency f i In a second step 42, ratios of the correction factors are determined that match the switching action 103, i.e., that represent the before and after state. If, for example, the second switch S2 is additionally closed from a state with the third switch S3 closed, then the corresponding ratio of the correction factors is K4(f) / K1(f). One (K 4,m (f) / K 1,m(f)), which differ in the assumed origin of the partial discharge. In a third step 43, this ratio is compared with the actual ratio of the signals. Those correction factors with the best agreement thus indicate the probable origin of the partial discharge. If the signal ratio V(f) agrees best with the ratio K 4,C (f) / K 1,C (f), then the point of origin of the partial discharge can be the one connected to the second switchgear. ter S 2 connected cables are accepted.

[0002] 202309462 14 List of reference symbols 10, 20 Equivalent circuit 12 Current source 131…137 Complex impedance 141…149 Complex impedance 15 Complex impedance 16 Sensor 21, 22 Complex impedance 30A…D Correction factor tables 31…34 Process steps 41…43 Process steps 100 Signal 101 Threshold

Claims

202309462 15 patent claims 1.Method for detecting partial discharges in an electrical device comprising at least one switch (S1...S3), in which - a correction factor table (30A...D) is determined for a plurality of potential locations for a partial discharge in the region of the device, wherein each correction factor table (30A...D) contains a plurality of correction factors, wherein each of the correction factors is assigned to a combination of switch positions of the device, - electrical pulses (100) are detected by means of a sensor device (16), - a correction factor table (30A...D) suitable for the origin of the pulses is selected from the correction factor tables (30A...D), - the detected pulses (100) or a threshold value (101) for comparison with the pulses (100) are corrected using the correction factor assigned to the existing combination of switch positions of the device. 2.Method according to claim 1, in which - a calibration is carried out for an initial switch position, - the detected pulses (100) or a threshold value (101) for comparison with the pulses (100) are additionally corrected using the correction factor assigned to the initial combination of switch positions of the equipment.

3. Method according to claim 1 or 2, in which - pulses (100) of a partial discharge source are recorded before and after a change in the switch position from a first to a second switch position, - a ratio of the signals (100) before and after the change is determined,. 202309462 16 - For each of the correction factor tables (30A...D), a ratio of the correction factors is formed from the correction factors for the first and second switch positions, - the best match between the ratios of the correction factors and the ratio of the signals (100) is determined.

4. Method according to one of the preceding claims, in which vectors each having a plurality of value pairs of frequency and scalar correction factor are used as correction factors.

5. Method according to claim 4, in which the values for the frequencies in the value pairs originate from a range between 1 MHz and 2 GHz. 6.Arrangement for detecting partial discharges in an electrical device comprising at least one switch (S1...S3), comprising - a sensor (16) for receiving electrical pulses (100), - an evaluation device with a plurality of stored correction factor tables (30A...D), each containing a plurality of correction factors, wherein each of the correction factors is assigned to a combination of switch positions of the device and wherein each correction factor table (30A...D) is assigned to a potential location for a partial discharge, configured to select from the correction factor tables (30A...D) a correction factor table (30A...D) suitable for the origin of the pulses (100) and to correct the detected pulses (100) or a threshold value (101) for comparison with the pulses (101) using the correction factor assigned to the existing combination of switch positions of the device. 7.Arrangement according to claim 6, wherein the sensor (16) is a Rogowski coil or a high-frequency current transformer.

Images