1
Monitoring of HV Generator Circuit Breaker Contact Ablation Based on Acoustic Emission Abhisek Ukil, Senior Member, IEEE, Martin Zlatanski, Moritz Hochlehnert
Abstract High voltage (HV) generator circuit breaker (GCB) is a key protective element for isolating HV generators in case of faults, or operational reasons. GCB typically has a double contact system. During the breaking operation process of the GCB, at first the nominal contact opens, after a while, the arcing contact opens, with a minimum time-elapse between the two operations. An arc ignites when the arcing contact opens. This arc should concern the arcing contact only, because the nominal contact is not designed to withstand arcs. However, after each operation the arcing contact is slightly ablated because of the arc. The ablation due to arcing makes the contact shorter and shorter. Therefore, the time interval between the two opening operations also gets shorter. For proper operation of the GCB a minimum time interval for the commutation of the current from the nominal contact system to the ablation contact system has to be assured. In this paper, a noninvasive direct ablation measurement method based on acoustic emission is presented. This is aimed at monitoring the delay in the time instant of the start of arcing contact touch as it gets shorter due to ablation, from the acoustic emission signals. Test results carried out on real breakers substantiate the proposed approach.
Index Terms Ablation, accelerometer, arcing contact, CB diagnostics, condition monitoring, electrical contact Manuscript IM-12-6440 revised on November 29, 2012. This work was supported by the Sensors & Signal Processing program, ABB Corporate Research. A. Ukil (corresponding author) is with ABB Corporate Research, Segelhofstrasse 1K, Baden-Daettwil, 5405, Switzerland (tel: +41 58 586 7034, fax: +41 58 586 4006, email:
[email protected]). M. Zlatanski is with ABB Corporate Research, Segelhofstrasse 1K, Baden-Daettwil, 5405, Switzerland (tel: +41 58 586 7561, email:
[email protected]). M. Hochlehnert is with ABB Power Products High Voltage, GCB division, Zurich, Switzerland (tel: +41 58 588 3985, email:
[email protected]).
January 18, 2013
DRAFT
2
erosion, HV breaker, piezoelectric sensor, vibration monitoring, online diagnostics, time delay measurement.
I. I NTRODUCTION High-voltage (HV) generator circuit breaker (GCB) is located between the generator and the step-up transformer in power plants such as fossil-fired, nuclear, gas turbine, combined cycle, hydro and pumped storage, etc. Main duties of a GCB are as follows [1], •
interrupt fault currents,
•
interrupt system-source and generator-source short-circuit currents,
•
carry and interrupt load currents,
•
synchronize the generator with the grid,
•
separate the generator from the grid.
IEEE Std C37.013-1997 [2] covers the requirements applicable for GCB, namely, rated maximum voltage, continuous current, short-circuit current, interrupting time, closing time, etc. With modern SF6 technology and self-blast principle for circuit breakers (CBs), GCBs with rated shortcircuit current up to 250 kA for generating units with ratings up to 2000 MVA are available [1]. Fig. 1-plot (a) shows a typical GCB, and plot (b) the layout of a power plant with GCB in between the generator and the main transformer (MT) [1]. Following the self-blast principle, the interrupting chamber of the GCB has double contact system, a nominal contact and an arcing contact. A schematic has been shown in Fig. 2. The breaking operation process of the GCB is described below. •
At first the nominal contact opens.
•
After a while, the arcing contact opens.
•
In the time interval between the two operations, the current flow commutes fully to the arcing contact, which is still closed.
•
A certain time period is needed for the commutation of the current into the ablation contact system in which the current can be interrupted by use of self-blast principle.
•
After each operation the arcing contact is slightly ablated because of the arc.
•
The ablation due to arcing makes the contact shorter. Therefore, the time interval between the two opening operations also gets shorter.
January 18, 2013
DRAFT
3
Fig. 1.
(a) Generator circuit breaker, (b) layout of power plant with GCB [1].
Fig. 2.
Cross-sectional view of the interrupting chamber of GCB.
Therefore, ablation of the arcing contact is a critical parameter for condition monitoring of the GCB [3]. In the state of the art dynamic contact resistance measurement (DRM) method [4], the dynamic resistance of the GCB is measured while the contacts move. Typically for this, the breaker has to be taken out of service. To overcome that, new DRM methods are proposed at rated speed without dismantling the breaker, but the interpretation of the resistance curve is very ambiguous [5]. Therefore, a direct, noninvasive contact ablation measurement method is of great interest.
January 18, 2013
DRAFT
4
In this paper, a noninvasive direct ablation measurement method based on acoustic emission (AE) is presented. This is aimed at monitoring the delay in the time instants of the start of arcing contact touch as it gets shorter due to ablation. The remainder of the paper is organized as follows. In section II, the principle of the proposed method, basics about acoustic emission, associated challenges, and review of previous works are presented. Section III describes the sensors used, experimental setup, and the different tests. Application results are presented in section IV. Discussions on the proposed method and the results are described in section V, followed by conclusions in section VI. II. P ROPOSED M ETHOD A. Principle At nominal closing velocity, the overlapping length translates in a timing difference of several ms [6]. The overlapping length of the double contacts is equivalent to the time for commutation. Therefore, the time difference between the start of the nominal and the arcing contact is directly indicative of the status of the arcing contact [6]–[7]. A representation of the different sequential contact times is shown in Fig. 3, for a closing operation. For an old breaker, as the arcing contact is shorter, there will be time delay for the arcing contact instant, hence the time difference would be shorter. In practice, the nominal contact would not be ablated much, therefore as shown in Fig. 3, the nominal contact instant tN om would practically remain constant, while the arcing contact instant tArc would be delayed, during a closing operation. During opening, the order would be reversed. However, during opening there will be electrical arc, which would expectedly heavily affect the measurement. Therefore, measurement during opening is not preferred. The method for arcing contact ablation monitoring investigated in this paper consists of measurement of the AE of the GCB contacts through a piezoelectric sensor or the mechanical vibration by the use of an accelerometer. The acquired data can be used in two ways. One consists in comparing the AE/vibration pattern acquired during an opening/closing operation with a reference record and quantifying the difference. Another way is to use it for direct time interval measurement, i.e., the time elapsed between the touching of the arcing and the main contacts.
January 18, 2013
DRAFT
5
Fig. 3.
Timing difference of arcing and nominal contacts in a GCB.
B. Acoustic Emission, Vibration Monitoring AE sensors typically acquire the elastic waves with frequency content from several tens of kHz up to several MHz [8]–[10]. In this frequency range, change in material density, elasticity or cracks can be detected [11]–[13]. Generally, time-domain deviations in signal magnitude, pattern sequence and timing are monitored as well as changes in the frequency content [8]–[10]. Fig. 4 shows the principle of AE along with the frequency classifications of AE and vibration, which are utilized for different machine health monitoring. The signature monitoring can be applied to a vibration pattern, which is useful as an online monitoring tool [11]–[14]. For GCB, deviations in the AE/vibration signature would indicate the degraded condition. C. Challenges The challenges of the implementation deal with the following aspects. •
Choosing the most appropriate sensor: the sensor should operate in the frequency range of interest and its resonant frequency must be situated outside the considered bandwidth. The frequency response inside this band is preferred to be linear for easier interpretation of the results. The sensor must be able to withstand the peak signal levels, without being damaged and at the same time provide high sensitivity.
January 18, 2013
DRAFT
6
Fig. 4.
•
Principle, frequency classification and applications of acoustic emission in machine condition monitoring [8].
Choosing the most appropriate position for the sensor: the sensor head should be adequately positioned in order to limit the useful signal attenuation and restrain the influence of the other AE/vibration sources. Moreover, special attention should be paid to the coupling at the sensor-breaker interface.
•
Linking the GCB acoustic signature change and the aging of the arcing contact: tests involving arcing contacts with different level of ablation should be performed in order to experimentally quantify the correlation between the contact aging and the change in the AE signature.
The AE signature can be recorded during opening or/and closing operation of the GCB. A closing operation is more suitable since the AE/vibration related to the arc from the current interruption, as well as to its extinguishing, during opening may lead to the triggering of false alarms.
January 18, 2013
DRAFT
7
D. Previous Works Vibration monitoring is in general applied for diagnostics purpose for the mechanical parts in equipments like transformer, circuit breaker, rotating machines [14]. Prior arts related to CBs are mentioned in [15]–[21]. Short summaries and related discussion are described below. Dewulf et al. [15] described different monitoring parameters of power CB using ‘CBWatch2’ device. For mechanical operations, they monitor primary contact separation speed, buffering, over travel, friction, breakage, spring fatigue, etc. However, contact ablation is not measured. Hoidalen and Runde [16] described continuous condition monitoring of HV CB using vibration monitoring. This is a relevant work, however, their conclusion was that vibration patterns (magnitudes) are interesting for mechanical diagnosis, but do not signify contact specific diagnostic. Lai et al. [17] analyzed mechanical failures in CBs, but they concluded that ablated contact abnormalities are difficult to observe. The work shows typical vibration signal, but does not quantify any contact ablation. Landry et al. [18] applied vibration monitoring techniques to diagnose problems like distortion in spring drive, loosening of auxiliary contacts, leakage of oil in the damper, etc. Regarding arcing contact, they concluded that no specific variation in acceleration signal was distinctively visible. Runde et al. investigated about acoustic diagnostic for CB, showing interesting characteristics vibration signal in [19]. But the authors concluded that the shown delay is “normal deviation defined by manufacturers”. Any quantitative measurements on direct ablation measurement were absent. Runde et al. further investigated for puffer type HV CB about different mechanical problems like lubrication problems, incorrect assembled crank shaft, sliding friction between the arcing contact members, incorrectly adjusted moving contact, etc in [20]. They also applied reference contact travel curve. However, measurement of the ablation of the arcing contact was not mentioned. In the paper [21], Runde et al. investigated arcing contact wear. However, in the presented result in [21], start of the contact before and after the endurance test, has not been characterized in conjunction with a reference like travel curve. So, the time delay quantification might not be precise, e.g., in Fig. 8 in [21], for the top part which has a lobe similar to the bottom plot, starting around 52 ms. The origin of the signal part between 46–52 ms in the top plot (refer to Fig. 8 in [21]) is not clear in absence of a reference signal. January 18, 2013
DRAFT
8
In general, any quantitative measurements of the contact ablation in GCBs by AE techniques with proper reference signals are yet to be presented. In this paper, we intend to fill that gap, by demonstrating quantitatively the feasibility of the AE-based technique for direct, noninvasive arcing contact ablation measurement. III. E XPERIMENTS A. Sensors After detailed survey, the SDT1-028K piezofilm sensor [22] was chosen for the test. The SDT1 sensor is based on piezofilm technology, being very light weight. It has linear characteristics up to 100 kHz, with resonant frequency points above 10 MHz [22]. This is relatively high compared to other sensors, which typically show linear characteristics up to about 20 kHz. This allows to record AE events in high frequency range as well, even though AE from mechanical parts typically take place in low frequency (< 10 kHz) zone [8]–[10]. It has a sensitivity of 15–20 mV/g [22]. It does not require any power supply and signal amplifier. R Among other sensors that we considered, ACH-01 [23], 7202A [24], Piezotron⃝[25], ADXL001
[26] models are significant. However, for brevity purpose, results obtained using SDT1-028K would be reported in the following sections. B. Test Setup In the planned tests, the AE sensors are placed on the surface of the GCB, in order to sense the elastic waves due to the mechanical movements (especially for the contacts). Due to contact ablation, the arcing contact gets shorter, hence it is expected that the arcing contact instant gets delayed as the GCB ages. The measurement setup, shown in Fig. 5, consisted of three poles of the HGI3 GCB [1], at ABB GCB manufacturing unit, Zurich, Switzerland. Contacts in the three poles (marked as A, B, C in Fig. 5) of the GCB are in three different ablation states. As it is a test GCB, the contact overlap lengths and the resistance values are measured previously. Less contact overlap length indicates more ablation. Pole B is in relatively good condition, while pole A is medium ablated, and pole C highly ablated. For the tests, a rotary travel sensor [27] (see Fig. 5) was mounted on the rotary rod of the mechanical drive (HMB type [1]), to provide a reference signal for correlating the contact instants. It is very important to correlate the contact instants from the AE January 18, 2013
DRAFT
9
signals with a reliable independent reference, which was often a missing factor in previously reported works (see section II.D).
Fig. 5.
Measurement setup for the three poles of GCB.
Fig. 6 shows the mounting of the SDT1 sensor at different positions on the surface of the GCB. Due to this requirement of several tests, the sensor was not permanently fixed at particular positions, rather fixed temporarily. This is not ideal, however as the study is concentrated on feasibility, it was acceptable. Nevertheless, extensive care was taken to ensure uniform mounting for each case. As seen in Fig. 5, the sensor was interfaced to an oscilloscope [28] via a low-noise preamplifier [30]. The SDT1 sensor does not inherently need signal amplifier. However, the preamplifier was used for experimental reasons like signal amplification in certain cases, and particularly for filtering noise in certain frequency bandwidths, e.g., in our cases below 400 Hz. The oscilloscope WaveJet 354A [28] has four analog channels. For the tests, an auxiliary signal from the control panel was used as the trigger signal. The AE signal from the sensor(s) and the travel sensor signals were synchronized at the scope. The breaker was operated with the mechanical drive, January 18, 2013
DRAFT
10
using the Megger (Programma) TM1600 circuit breaker analyzer system [29], as shown in Fig. 5.
Fig. 6.
Positions of acoustic emission sensor for measurements on the GCB poles.
C. Experiments The GCB is located at test lab, so it is not connected between a generator and transformer in a plant. So, the tests are conducted at off-power conditions. Nevertheless, the GCB is operated with the mechanical drive as done in plant. In the tests, the Programma TM1600 breaker analyzer system was used to trigger the GCB closing and opening operation. Data were recorded during the closing operation, multiple times at each position. From the plot (ii) of Fig. 6, it can be noticed that in position 1 (i.e., top of GCB), we also used two sensors simultaneously on two poles, synchronized at the scope. IV. R ESULTS Figs. 7–11 show the synchronized travel sensor signal and the AE signals obtained using the SDT1 sensor for the five operations across four positions (refer to Fig. 6) during one of the closing January 18, 2013
DRAFT
11
operations. In Figs. 7, 9–11, the signals for different poles were acquired separately, while in Fig. 8, signals were acquired from poles B and C simultaneously. Fig. 12 demonstrates the good repeatability of the AE signal, showing for example, three different signals for three different closing operations at position 1 (refer to Fig. 6). The variance in the time delay measurement for several measurements is less than 0.1% of the normalized travel time. The signals were synchronized at the scope, sampled at a sampling frequency of 2.5 MHz. The X-axis in all the plots in Figs. 7–12 shows the normalized time, 0% indicating the start and 100% the end of the movement. This is done to normalize any breaker specific effect. A. Measurement from Rotary Travel Sensor In the top plot in Figs. 7–11, the travel curve from the travel sensor is shown. For the mechanical drive movement of the GCB, the travel sensor measures the angular movement, the rotary travel curve span being measured in the range of 7–13.5 V in oscilloscope, as shown in Figs. 7–11. From previously performed DRM, the reference angle values for the start of the arcing contact (for each of the poles A, B, C), and the nominal contact are known. From that, one can easily calculate the voltage levels of the rotary travel sensor, making sure the travel sensor is mounted always in the same way. From those marked reference points for the arcing and the nominal contacts for the different breaker poles, vertical lines are drawn in Figs. 7–12 to indicate the independent reference for the AE signal. B. Measurement at Position 1 For the GCB, position 1 (refer to Fig. 6) is best in terms of sensor proximity to the contact positions in the breaking chamber. From Fig. 12, we can see that the AE signals are very repeatable. From Fig. 7, we can notice the direct correlation of the AE signal with the arcing contact. The big signal lobes in the plots (ii–iv) of Fig. 7 mark the start of the arcing contact, as confirmed by the reference travel position from the DRM values projected onto the measured travel curve. In Fig. 7, the corresponding normalized travel time values (in X-axis) are tracked in the AE signal for a distinctive pattern, e.g., the start of the big signal lobe. The matching normalized travel time reference values (in X-axis) in the AE signals in Fig. 7, correspond to the arcing contact. As shown in Fig. 7, the observed time delays conform to the different ablation January 18, 2013
DRAFT
12
Fig. 7. GCB, operation at Position 1 (Fig. 6): plot (i) travel curve, and AE signals acquired from (ii) pole A, (iii) pole B, (iv) pole C, using AE sensor during closing operation.
status of the breaker poles, as we expect delay in contact touch as the poles get more ablated (i.e., shorter in overlap length). Quantitative comparisons of the relative ablation levels and the travel time delays from the AE measurements (Fig. 7) are provided in Tables I and II respectively. From the comparative figures in Tables I and II, it is evident that the time delay (as % of total travel time) in the measured AE signal is linearly correlated with the ablation status of the poles.
January 18, 2013
DRAFT
13
Fig. 8.
GCB, operation at Position 1, two poles simultaneously (Fig. 6): plot (i) travel curve, and AE signals simultaneously
acquired from (ii) pole B, (iii) pole C, using two AE sensors during closing operation.
C. Measurement at Position 1: Two Poles Simulatenously Fig. 8 shows the simultaneous measurement of the pole B (good condition) and pole C (highly ablated). Like Fig. 7 measurements, it can be seen that the simultaneous AE measurements are also correlated with the arcing contacts. From the contact instants in Fig. 8, we notice a travel time delay of 4.1% (= 62.2% −58.1%) for pole C, with respect to (w.r.t) the pole B, matching quite good with Table II figures.
January 18, 2013
DRAFT
14
Fig. 9. GCB, operation at Position 2 (Fig. 6): plot (i) travel curve, and AE signals acquired from (ii) pole A, (iii) pole B, (iv) pole C, using AE sensor during closing operation.
D. Measurement at Position 2 Fig. 9 corresponds to the position 2 (operation iii, in Fig. 6), where the sensors were placed on the top insulator, the contacts being directly inside that. From Fig. 9, we can also see the travel delay in the AE signals at this position. However, because it is a different material than the metal part in position 1, we see a different signal type. We can also notice higher signal fluctuation between 20%–50% travel time, i.e., during the initial drive movement.
January 18, 2013
DRAFT
15
Fig. 10.
GCB, operation at Position 3 (Fig. 6): plot (i) travel curve, and AE signals acquired from (ii) right of pole C, (iii)
between pole B & C, (iv) between pole A & B, (v) left of pole A, using AE sensor during closing operation.
E. Measurement at Position 3 Fig. 10 corresponds to the position 3 (operation iv, in Fig. 6), where the sensors were placed (indicated by the arrows) on low-voltage (LV) metal base plate where all the breaker poles are connected. As we move away from the breaking chamber, we lose the characteristics AE signal at this position, as shown in the plots (ii-v) in Fig. 10, which correspond to the four possible positions. Instead, we see mainly the low frequency mechanical vibrations as we get closer to the mechanical drive. From the plot (iv) of Fig. 6, we can notice that the drive is not located in January 18, 2013
DRAFT
16
Fig. 11. GCB, operation at Position 4 (Fig. 6): plot (i) travel curve, and AE signals simultaneously acquired from (ii) rotating rod between pole A & B, (iii) rotating rod between pole B & C, using two AE sensors during closing operation.
the middle, rather towards pole B & C. Therefore, at the left of pole A, we notice very weak signal (see plot (v) of Fig. 10). F. Measurement at Position 4 Fig. 11 corresponds to the position 4 (operation v, in Fig. 6), where the sensors were placed on the LV rotating mechanical rod (at the end of which the rotary travel sensor was mounted, see Fig. 5). We measured the vibration signals simultaneously from the two positions, between January 18, 2013
DRAFT
17
Fig. 12.
GCB operation at top of breaker position: AE signals acquired from pole B, using AE sensor during three different
closing operations.
poles A & B, and between poles B & C. In Fig. 11, the period between 20%–40% travel time corresponds to the start of drive movement, while the period between 75%–100% travel time corresponds to the end of drive movement. As the drive is connected closer to poles B & C (see Fig. 5), one can see a stronger vibration signal in plot (iii). However, in either position, the contact instants are not distinctly visible. V. D ISCUSSIONS The following comments are cited on the proposed method, and the results. 1) From the results shown in Figs. 7–11, Tables I–II, we can note linearly correlated direct measurement of the arcing contact ablation levels from the time delay (in % of total travel time) measurements from the AE signals. Availability of the reference travel sensor and the DRM values ensure consistent interpretation of the AE signals. January 18, 2013
DRAFT
18
TABLE I D IFFERENT ABLATION LEVELS OF THE THREE POLES OF GCB
Pole
Relative contact
Status
Overlap length
overlap length
difference w.r.t pole B
(%)
(%)
B
100
Good condition
-
A
65
Medium ablated
35
C
46
Highly ablated
54
TABLE II A RCING CONTACT TIME DELAY MEASUREMENT FROM AE SIGNAL
Pole
Arcing contact
Status
Travel delay
instant
w.r.t pole B
(% of total
(% of total
travel time)
travel time)
B
57.9
Good condition
-
A
60.5
Medium ablated
2.6
C
62.1
Highly ablated
4.2
2) The single phase (e.g., Fig. 7) and simultaneous two phase measurements (e.g., Fig. 8) in the GCB did not show any cross-phase influence on the AE signals, or additional time delay due to synchronization of the two sensors. 3) From the measurements in Figs. 7–11, we see that the position 1 (refer to Fig. 6) is possibly the best one, being in proximity to the contacts in the breaking chamber. In Figs. 9–11, as we move away from the position 1, we gradually lose the characteristic AE signal, and instead start sensing more the mechanical vibration coming from the drive and the structure. 4) In the present analysis, a rough signal characterization is done on the AE signals, mostly based on what is readily visible. In principle, one would need a data conditioning, like proper filtering of the time domain AE signal (e.g., for the high frequency (HF) glitches around 35% travel time in plot (i) of Figs. 7–11), and processing like automatic change detection of the levels for marking the ablation contact, storing in buffer and comparative January 18, 2013
DRAFT
19
computations of travel delay, etc. However, with the relatively good signature of the characteristic signal amplitude change in the AE signals, subsequent processing would be easier. 5) If one uses absolute time, variation of the mechanical drive might influence the measurement, namely, by introducing additional confusing time delay or advancement. However, to tackle that, the travel sensor has been used as an external time reference. 6) The HF glitches seen around 35% travel time in plot (i) of Figs. 7–11, come from the trip coil energization. This was validated separately by doing HF emission tests with and without involving the GCB, and then considering the difference of the tests to localize the source. This does not influence the signals coming from the contact operations. Similar glitches could also be observed in Figs. 7–12, especially during the initial drive movement part (between 0%–40% travel time), as the measurements were done in an industrial environment exposed to different HF origins. 7) For direct measurement of the arcing contact ablation, one would need an AE sensor, preferably with linear characteristics in the frequency range of interest. The SDT1 sensor [22] is possibly a good candidate. The AE sensors are not prohibitively expensive. Therefore, this could potentially be a cost-effective solution for direct contact ablation measurement in a noninvasive way. 8) An important issue is the mounting of the AE sensors. Due to the requirement of doing number of experiments, we had to use power tape for mounting the sensor. In practice, one has to use superglue or similar methods, ensuring as uniform mounting as possible. Screw adjustable mounting is also a good possibility, e.g., for the 7202A sensor [24]. Nevertheless, the uncertainty of the uniform mounting was included in the tests, and the promising results point that this could anyway further be improved. A. Future Directions Following the technology feasibility, the following future actions are foreseen. 1) Tests are planned under powered condition, the GCB being placed between the generator and main step-up transformer. 2) Tests are planned predominantly at position 1 (see Fig. 6), with fixed, uniform mounting.
January 18, 2013
DRAFT
20
3) For tests under powered, HV condition, the oscilloscope might not be used. Because, the HV might damage the oscilloscope, or the insulation criteria (e.g., interfacing the oscilloscope with properly insulated cables, etc.) might not be fulfilled. Hence, embedded data acquisition system might be needed. VI. C ONCLUSION Generator circuit breaker is a key protective element for isolating HV generators in case of faults, or operational reasons. However, the ablation due to arcing makes the GCB contact shorter. Therefore, the arcing contact ablation is a key electrical parameter for monitoring. In this paper, a noninvasive direct ablation measurement method based on acoustic emission is presented. The AE sensors are placed on the surface of the GCB, in order to sense the elastic waves due to the mechanical movements of the contacts. Due to contact ablation, the arcing contact gets shorter, hence it is expected that the arcing contact instant gets delayed as the GCB ages. The SDT1 [22] AE sensor has been used for tests on the GCB, at different positions. The AE signals capture the signature of the arcing contact instant, demonstrating a linearly correlated time delay (in % of total travel time) with the differently ablated poles, as expected from the principle. A rotary travel sensor is used with known DRM values, as independent reference to mark the start of arcing and nominal contacts, ensuring consistent interpretation of the AE signals. ACKNOWLEDGMENT The authors would like to thank Detlef Pape, Kai Hencken, Martin Lakner, Andrea Andenna, Yannick Maret for their support. R EFERENCES [1] ABB, “Brochures, manuals, documentations for Generator Circuit Breakers - HECS, HGI series,” 2012. Available: http://www.abb.com/gcb [2] IEEE, “IEEE Standard for AC High-Voltage Generator Circuit Breakers Rated on a Symmetrical Current basis,” IEEE Std C37.013-1997, 1997. [3] IEEE, “IEEE Guide for the Selection of Monitoring for Circuit Breakers,” IEEE Std C37.10.1-2000, 2001. [4] M. Landry, O. Turcotte, F. Brikci, “A Complete Strategy for Conducting Dynamic Contact Resistance Measurements on HV Circuit Breakers,” IEEE Transactions on Power Delivery, vol. 23, no. 2, pp. 710–716, 2008. [5] M. Landry, et al., “A New Measurement Method of the Dynamic Contact Resistance of HV Circuit Breakers,” In proc. IEEE T&D Conf. Latin America, 2006.
January 18, 2013
DRAFT
21
[6] G. Krzysztof, L. Zehnder, R. Vogelsang, T. Schoenemann, M. Stanek, “Method for determining contact wear in a heavy-duty circuit breaker,” Patent application WO 2008/000105 A1, 2006. [7] B. Rusek, G. Balzer, M. Holstein, M. S. Claessens, “Timings of high voltage circuit-breaker,” Electric Power Systems Research, vol. 78, pp. 2011–2016, 2008. [8] D. J. Yoon, Fundamentals of Acoustic Emission, Asia-Pacific Smart Structure Tech., Korea, 2008. [9] M. G. Duncan, J. W. Whittaker, “Acoustic emission calibration instrumentation ,” IEEE Transactions on Instrumentation and Measurement, vol. 38, no. 3, pp. 827–831, 1989. [10] G. Muravin, Inspection, Diagnostics and Monitoring of Construction Materials and Structures by the Acoustic Emission Method, Minerva Press, London, 2000. [11] Z. Jun-Hong, P. K. Chee, Z. Zhao-Wei, F. L. Lewis, “Tool Wear Monitoring Using Acoustic Emissions by Dominant-Feature Identification,” IEEE Transactions on Instrumentation and Measurement, vol. 60, no. 2, pp. 547–559, 2011. [12] L. Ruoyu, D. He, “Rotational Machine Health Monitoring and Fault Detection Using EMD-Based Acoustic Emission Feature Quantification,” IEEE Transactions on Instrumentation and Measurement, vol. 61, no. 4, pp. 990–1001, 2012. [13] V. Hanel, W. Thelen, “Monitoring screws under tensile load using acoustic emission analysis,” IEEE Transactions on Instrumentation and Measurement, vol. 45, no. 2, pp. 547–550, 1996. [14] F. Brikci, “Vibro-acoustic Testing Applied on Tap Changers and Circuit Breakers,” Tech. Report, Zensol Automation Inc., 2010. Available: zensol.com/Articles/Article-TechCon-2010.pdf [15] J. A. Dewulf, T. Jung, J. P. Dupraz, G. F. Montillet, “A Development and application of Circuit Breakers Diagnostic and Monitoring,” In proc. IEEE T&D Conf. Exp., 2003. [16] H. K. Hoidalen, M. Runde, “Continuous Monitoring of Circuit Breakers Using Vibration Analysis,” IEEE Transactions on Power Delivery, vol. 20, no. 4, pp. 2458–2465, 2005. [17] M. L. Lai, S. Y. Park, C. C. Lin, H. Naidu, “Mechanical Failure Detection of Circuit Breakers,” IEEE Transactions on Power Delivery, vol. 3, no. 4, pp. 1724–1731, 1988. [18] M. Landry, F. L´eonard, C. Landry, R. Beauchemin, O. Turcotte, F. Brikci, “An Improved Vibration Analysis Algorithm as a Diagnostic Tool for Detecting Mechanical Anomalies on Power Circuit Breakers,” IEEE Transactions on Power Delivery, vol. 23, no. 4, pp. 1986–1994, 2008. [19] M. Runde, T. Aurud, L. E. Lundgaard, G. E. Ottesen, K. Faugstad, “Acoustic Diagnosis of High Voltage Circuit Breakers,” IEEE Transactions on Power Delivery, vol. 7, no. 3, pp. 1306–1315, 1992. ¨ en, “Vibration Analysis for Diagnostic Testing of Circuit-Breakers,” IEEE [20] M. Runde, G. E. Ottesen, B. Skyberg, M. Ohl´ Transactions on Power Delivery, vol. 11, no. 4, pp. 1816–1823, 1996. ¨ en, “Vibration Analysis for Periodic Testing of Circuit-Breakers,” In proc. 11th Int. Symp. [21] M. Runde, B. Skyberg, M. Ohl´ HV Engg., 1999. [22] measurement SPECIALITIES, “Datasheet, documentations for SDT Shielded Piezo Sensors,” 2009. Available: http://www.meas-spec.com [23] measurement SPECIALITIES, “Datasheet, documentations for ACH-01 Accelerometer,” 2008. Available: http://www.measspec.com [24] measurement SPECIALITIES, “Datasheet, documentations for 7202A Accelerometer,” 2009. Available: http://www.measspec.com R Emission Sensor, type 8152Bx1x, 8152Bx2x” [25] Kistler Instrumente AG, “Datasheet, documentations for Piezotron⃝Acoustic
2011. Available: http://www.kistler.com
January 18, 2013
DRAFT
22
[26] Analog Devices, “Datasheet, documentations for ADXL001 Accelerometer,” 2010. Available: http://www.analog.com [27] Novotechnik, “Rotary Transducer as part of Megger (Programma) CB Testing Systems,” 2012. Available: http://www.novotechnik.com [28] LeCroy, “Oscilloscope WaveJet 354A,” 2011. Available: http://www.lecroy.com/oscilloscope/oscilloscopemodel.aspx?modelid=2000 [29] Megger, “TM1600 circuit breaker analyzer system,” 2010. Available: http://www.megger.com [30] Stanford Research Systems, “Low-noise Pre-amplifier model SR560,” 2011. Available: http://www.thinksrs.com/downloads/PDFs/Manuals/SR560m.pdf
B IOGRAPHIES
Abhisek Ukil (S ′ 05 − M ′ 06 − −SM ′ 10) received the bachelor of electrical engineering degree from the Jadavpur University, Calcutta, India, in 2000 and the M.Sc. degree in electronic systems and engineering PLACE PHOTO HERE
management from the University of Bolton, Bolton, UK in 2004. He received his Ph.D. from the Pretoria (Tshwane) University of Technology, South Africa in 2006, working on power systems disturbance analysis with Eskom. From 2000 to 2002, he worked as a software engineer at InterraIT, India. After joining in 2006, currently
he is a Principal Scientist at ABB Corporate Research, Switzerland. He is author/coauthor of 48 refereed papers, a monograph, two book chapters, and inventor/co-inventor of 8 patents. His research interests include power systems, signal processing, condition monitoring, sensor-based embedded systems.
Martin Zlatanski received the M. Sc. and the Ph.D. degrees in electrical engineering from the University of Strasbourg, Strasbourg, France in 2007 and 2011, respectively. Since 2011 he is with ABB Corporate PLACE PHOTO
Research in Switzerland, where he is involved in the development of condition monitoring systems for power products.
HERE
Moritz Hochlehnert is currently group leader of secondary technology development for generator circuit breaker and working for ABB Power Products High Voltage, Generator Circuit-Breaker division, Zurich, PLACE PHOTO
Switzerland since 2007. He obtained his Diploma in electrical and electronics engineering in 2007 from the University of Karlsruhe, Karlsruhe, Germany.
HERE
January 18, 2013
DRAFT
