Directional Sensing for Online PD Monitoring of MV Cables Wagenaars, P.; van der Wielen, P.C.J.M.; Wouters, P.A.A.F.; Steennis, E.F. Published in: Nordic Insulation Symposium, Nord-IS 05 Published: 01/01/2005 Document Version Publisher s PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication: A submitted manuscript is the author's version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website. The final author version and the galley proof are versions of the publication after peer review. The final published version features the final layout of the paper including the volume, issue and page numbers. Link to publication Citation for published version (APA): Wagenaars, P., Wielen, van der, P. C. J. M., Wouters, P. A. A. F., & Steennis, E. F. (2005). Directional Sensing for Online PD Monitoring of MV Cables. In Nordic Insulation Symposium, Nord-IS 05 (pp. 79-82). Trondheim. General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. Users may download and print one copy of any publication from the public portal for the purpose of private study or research. You may not further distribute the material or use it for any profit-making activity or commercial gain You may freely distribute the URL identifying the publication in the public portal? Take down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Download date: 05. Jan. 2019
Directional Sensing for Online PD Monitoring of MV Cables P. Wagenaars, P.C.J.M. van der Wielen, P.A.A.F. Wouters and E.F. Steennis Eindhoven University of Technology, P.O.Box 513, 5600 MB, Eindhoven, The Netherlands Abstract A new system to monitor partial discharges (PDs) online in MV power cables is being developed. For this system sensors are placed in the Ring-Main-Units (RMUs) at each end of the cable. In an RMU, the cable under test is not the only possible source of PDs. Also other MV cables and the MV/LV transformer may produce PDs. From the direction of each measured signal, it is known whether the PD comes from the cable under test or elsewhere. For the proposed solution, two current probes are installed at two locations in the RMU. The two measured currents are combined into one signal. PDs from other sources are removed from the resulting signal, while PDs from the cable under test remain unaffected. Measurements on a test set-up with two powered RMUs showed that this method works well. 2. Theory 2.1. RMU model In a cable and an RMU two different propagation channels can be identified: one between the sheath and three phases and one between two different phases [3]. This paper considers only the channel between the sheath and the three phases together. around MV cable, past last earth connection earth TCC transformer connecting cables (TCC) earth rail 1. Introduction A new system for the detection and localization of PDs in MV cables is being developed [1-4]. This online PD monitoring system detects PDs with inductive current sensors clamped around conductors in the RMUs at each end of the cable under test. An RMU consists of an MV/LV transformer and one or more MV cables (in most situations there are two MV cables). A schematic drawing of a typical Dutch RMU is depicted in figure 1. Usually, all MV cables and the transformer are connected to the MV rail. That implies that during online PD measurements not only PDs from the cable under test are detected, but also PDs from the transformer and other MV cables. For correct interpretation of the measured signal, PDs from these disturbing sources must be removed from the signal. PDs originating from the cable under test propagate from the cable into the RMU, while disturbing PDs propagate from the RMU into the cable under test. Therefore, it is possible to apply directional sensing to distinguish PDs originating from the cable under test from disturbing PDs. In the past, several methods to achieve directional sensing have been described [5-7]. Unfortunately, none of these methods can be used for application without modification of the RMU. This would violate the requirement that the diagnosed equipment must remain unaffected. The presented paper proposes a method of directional sensing that uses two inductive current sensors placed at two locations in the RMU. By combining the signals of both sensors directional sensing is achieved. 1 2 transformer MV cable earth connection Fig. 1 Schematig drawing RMU The equivalent circuit of the RMU depicted in figure 1 consists of several parallel impedances between phases and earth. The MV cables are modeled by their characteristic wave impedances Z 1 and Z 2. The transformer impedance is indicated with Z tr, and the capacitance of the three transformer connecting cables (TCC) with Z. If one of the MV cables produces a PD, it can be described as a current source producing a pulse. In figure 2, the equivalent circuits of the RMU are depicted for a PD coming from cable 1 and for a PD from cable 2. i three phase conductors together i tr i 1 i 2 i tr i i 1 i 2 Z tr Z Z 2 Z tr Z Z 1 (a) PD from cable 1 (b) PD from cable 2 Fig. 2 Equivalent circuits for PDs originating from cable 1 (cable under test) or cable 2 2.2. Directional sensing Consider the situation that cable 1 is the cable under test. In figure 2a it can be seen, that for a PD from this 79
cable, i 1 and i have an opposite direction, while for a PD from cable 2 (fig. 2b) the direction is the same. If the transfer function H from I to I 1 (with I en I 1 having the same direction) is known, it can be used to remove disturbing PDs (assuming that Z itself does not produce PDs). I1 I1 = 2I1 PD from cable1 I1 H I = (1) I1 I1 = 0 PD from cable 2 Where H = ±I 1 /I, + : PD from cable 2, - : PD from cable 1. Unfortunately, H usually differs for a PD originating from the cable under test and a disturbing PD. If a PD comes from cable 1 I1, Y + Y2 + Ytr H = H1 = = (2) I Y, with Y x is the admittance of impedance Z x. The subscript means: for a PD originating from cable 1. If a PD originates from cable 2 I Y1 H = H (3) Y 1, c2 2 = = I, c2 The subscript c2 means: for a PD originating from cable 2. Note that if the transformer produces a PD, the transfer function H = I 1 /I equals H 2. The problem of the different transfer functions for different PD origins can be solved by means of a correction transfer function H corr. If H in equation 1 is substituted with H 2, PDs from cable 2 (and from the transformer) will be removed from the resulting signal, PDs from cable 1 will be deformed. To correct the deformation, the resulting signal is multiplied by a transfer function H corr I filt = H corr ( I1 H 2 I ) (4) where I filt is the resulting filtered signal, free of disturbing PDs. In order to correct for the distortion, the transfer function H corr must be chosen such, that the filtered signal I filt equals I 1 for PDs from cable 1. Substituting I filt in equation 4 with I 1, yields: H corr = I1, H1 = I H I H + H (5) 1, 2 Note, that H corr = ½ if the two cable characteristic impedances are equal (H 1 = H 2 ) and dominant. Before any signal can be filtered, the transfer functions H 1, H 2 and H corr have to be determined. The most convenient way is to inject two calibration pulses and measure I 1 and I. To determine H 1 a pulse is injected in the 1 2 system as if it originates from cable 1. To determine H 2 a pulse is injected in the system as if it originates from cable 2. Note that this method only works correctly if Z does not produce PDs. Otherwise, a PD from Z would be misinterpreted as a PD from the cable under test. 3. Measurements 3.1. Description of the measurements At the KEMA premises in Arnhem, the Netherlands, a test set-up is designed with typical Dutch RMUs and two MV cables. In one RMU, current sensors are installed around the earth connection of the transformer connecting cables (see figure 1) to measure i. In the same RMU a second current sensor is installed around MV cable 1 past the last earth connection to measure i 1. There are two options for the injection of the calibration pulses: local injection, and injection at the far end of the cable. For local injection a coil is placed around MV cable 1 or 2 past the last earth connection to simulate a pulse originating from respectively cable 1 and cable 2. From the results, the transfer functions H 1 and H 2 are determined. The pulses are made as short as possible to cover the entire frequency range of interest. For far end injection a pulse is injected at equivalent positions at the other RMUs. For the calibration measurement, pulses are injected locally because local injection is more practical. Moreover, a pulse injected at the far end will have less high frequency content because high frequencies attenuate as the pulse travels through the cable. After applying the calibration pulses, several pulses are injected at the far end of cable 1 to simulate PDs from the cable under test and at the far end of cable 2 to simulate disturbing PDs. Finally, a pulse is injected locally with an injection coil around all three transformer connecting cables to simulate a disturbing PD from the transformer. For all test pulses i 1 and i are measured and filtered using equation 4. The measured current i 1 is compared with the filtered signal i filt. 3.2. Measurement results After calibration, the first test pulse is injected at the far end of cable 1. With cable 1 being the cable under test, this pulse should pass the filter undisturbed. Both the measured i 1 and the filtered signal i filt are plotted in figure 3. The small oscillations before the large pulse are picked up by direct radiation through air and are not actual currents in the cable. As can be observed in figure 3, the measured and filtered current match very well. The second test pulse is injected at the far end of cable 2. The current i 1 and the filtered signal i filt are plotted in figure 4. Because a pulse from cable 2 is a disturbing pulse, it should be removed from i filt. It can be seen that the first pulse is indeed almost completely 80
removed from i filt. The second pulse is a reflection. The original pulse from cable 2 travels into cable 1 and reflects back at the far end. Because it returns from the cable under test (cable 1) it is not removed from the signal. Apparently the equivalent circuit of figure 2 is not satisfactory for PDs from the transformer. In this circuit, only the impedances of the RMU components are considered. The inductances of the large loops in the RMU are neglected, which become important for the high frequencies involved with local injection. These inductances are introduced in the equivalent circuit of figure 6. In this equivalent circuit, the transfer function I 1 /I for a PD from the transformer is no longer equal to the transfer function H 2 for a PD from cable 2. Therefore, the pulse in the last experiment is not filtered from the signal. i tr i i 1 i 2 Z tr Z Z 1 Z 2 Fig. 3 - Pulse from cable 1, injected at far end Fig. 6 - Equivalent circuit with inductances 4. Alternative directional sensing method Fig. 4 - Pulse from cable 2, injected at far end The third injected pulse simulates the situation of a disturbing PD from the transformer. Both signals, i 1 and i, are plotted in figure 5. Since a PD from the transformer is considered a disturbing PD, it should be removed from the signal. The measurements show clearly that this is not the case. 4.1. Description PDs from the transformer are not removed from the filtered signal and have to be recognized by other means. One way is to look at the polarities of the first peak of the measured currents i and i 1. The polarities of the first peaks are equal for disturbing PDs, while for PDs from the cable under test the polarities are opposite (see figure 2). Using this property, the direction of PDs can be determined as long as the PDs can be clearly separated from each other and do not overlap. An additional advantage is that no calibration is needed. 4.2. Measurements In order to test this technique, positive and negative pulses are injected at the far ends of cable 1 and cable 2. Also pulses are injected as if they would originate from the transformer. In figure 7 the currents i en i 1 are plotted for test pulses from cable 1. In figure 8 the same currents are plotted for pulses from cable 2. Figure 9 depicts i and i 1 for a pulse from the transformer. Figure 7 indeed shows that the first peaks of both currents have opposite polarities. In figure 8 it can be observed that the first peaks of the disturbing pulses have equal polarity. Also disturbing pulses from the transformer (figure 9) have the same polarity. All measurements confirm the predictions from the model, that pulses from the cable under test have opposite initial polarities, while disturbing pulses have the same initial polarity. Fig. 5 - Pulse from transformer, local injection 81
6. References Fig. 7 - Positive and negative pulse from cable under test Fig. 8 - Positive and negative disturbing pulse [1] P.C.J.M. van der Wielen, J.Veen, P.A.A.F. Wouters and E.F. Steennis, On-line partial discharge detection of MV cables with defect localisation (PDOL) based on two time synchronised sensors, to be published on the 18th International Conference on Electricity Distribution (CIRED), Turin, Italy, June 6-9, 2005. [2] E.F. Steennis, P.C.J.M van der Wielen, J. Veen and P.A.A.F. Wouters, Aspects and implications of on-line PD detection and localization of MV cable systems, Conference Record of the International Conference on Insulated Power Cables (JICABLE), Versailles, France, 2003. [3] P.C.J.M. van der Wielen, On-line detection and location of partial discharges in medium-voltage power cables, Ph.D. thesis, Eindhoven University of Technology, Eindhoven, The Netherlands, 2005. [4] J. Veen, On-line signal analysis of partial discharges in medium-voltage power cables, Ph.D. thesis, Eindhoven University of Technology, Eindhoven, The Netherlands, 2005. [5] B.M. Oliver, Directional Electromagnetic Couplers, Proc. of the Institute of Radio Engineers, Vol. 42, 1954, pp. 1686-1692. [6] D. Pommerenke, T. Strehl, R. Heinrich, W. Kalkner, F. Schmidt and W. Weissenberg, Discrimination between Internal PD and Other Pulses Using Directional Coupling Sensors on HV Cable Systems, IEEE Trans. on Dielectrics and Electrical Insulation, Vol. 6 No. 6, Dec 1999, pp. 814-824. [7] D. Wenzel, U. Schichler, H. Borsi and E. Gockenbach, Recognition of Partial Discharges on Power Units by Directional Coupling, Proc. of the 9 th Int. Symp. on High Voltage Engineering, Graz, Vol. 5, 1995, paper 5626. Fig. 9 - Disturbing pulse from transformer 5. Conclusions The first directional sensing technique described in this paper uses two transfer functions to combine and filter currents measured at two locations in the RMU. This method achieves excellent results for rejecting signals coming from other cables. It fails to remove disturbing PDs originating from the distribution transformer. The alternative directional sensing technique managed to recognize all disturbing pulses. A disadvantage of this technique is, that consecutive PDs need to be clearly separated in time. An advantage is that no calibration is required. 82