SF6 Partial Discharge
SF6 Partial Discharge
SF6 Partial Discharge
G.A.S. Gesellschaft fr analytische Sensorsysteme mbH, TechnologieZentrumDortmund, Emil-Figge-Str. 76-80, D-44227 Dortmund, Germany Institut fr Spektrochemie und angewandte Spektroskopie (ISAS), Bunsen-Kirchhoff-Str. 11, D-44139 Dortmund, Germany
Abstract Although gas insulated switchgear and substations (GIS) belong to the most reliable components of electric power networks, failures still occur in their long term operation. Some of these failures may be attributed to small variations of gas compositions within the shielded equipment. Thus the prediction of a breakdown should be attainable on the basis of sensitive analytical monitoring of the insulating gas. Such a monitoring system for GIS and SF6-filled circuit breakers is realised on the basis of continuous sensing of the insulating gas using ion mobility spectrometers (IMS). Especially, partial discharge IMS provides power utilities a diagnostic tool able to perform on-site monitoring and event controlled inspections of the SF6 gas. The results of on-site monitoring and on-line analysis of SF6 contained in a circuit breaker of a GIS in Germany, before and after operation are described. Conclusions of Investigations of different disconnecting switches in a substation in Germany are presented, indicating that a full recombination of decomposition products was not achieve in all cases. Finally, the structure of a system for automatic on site sensing of SF6-decomposition in the hydroelectric power plant of Foz do Areia, State of Paran, Brazil, is discussed. Introduction During the past 30 years the use of sulphur hexafluoride in insulated switchgears in high voltage substations increased considerably. This due to their enhanced lifetime, reduced required area and volume and longer maintenance cycles. Very high standards in engineering and manufacturing provide a high reliability of these substations. However, during long term operation failures still occur which can cause severe damages and be costly to repair. A closer view shows that partial discharge (over a long period), sparcs and power arcs which belong to the operation of high voltage switches lead to decomposition of the insulating gas SF6 [1-3]. Due to long maintenence cycles, decomposition products may accumulate over long times within the switchgear. Thus, a relatively small partial discharge activity over a long period may lead to a noticeable amount of corrosive and toxic by-products. Accordingly, insulating spacers and other components could be exposed to a corrosive medium for long time periods. Hence, the establishment of a diagnostic tool, which allows the on-site and on-line analysis of the insulating gas, is of considerable interest for the user of SF6 filled equipment. Ion mobility spectrometry is a technique designed for sensitive environmental monitoring by using a low cost, rugged and highly mobile instrument. Using this technique for supervision of the insulating gas in GIS during operation offers the opportunity to achieve an event-based maintenance and detect possible failure. In the following paper results of an super vision of a substation will be summarised with respect to the question that full recombination of decomposition products in a switchgear will be achieved or additional maintenance will be necessary.
92
Ion Mobility Spectrometry Ion mobility spectrometry is a technique which was designed for the detection of trace compounds within a gas, for example gaseous pollutions in air [4-5]. It combines high sensitivity and relatively low technical expenditure with a high speed data acquisition [6]. The main advantage with respect to other common detection principles is the fact, that the instrument can work on-line, continuously and even without operator [7]. Such an instrument is based on the drift of ions at ambient pressure under the influence of an electric field. The ions experience a separation process that is based on different drift velocities due to different masses and charges or different geometrical structures. Collecting these ions on a Faraday plate delivers a time dependent signal corresponding to the mobility of the arriving ions. Such an ion mobility spectrum contains information about the nature of the different trace compounds present in the sampled gas. For generation of ionic species some additional components are required, such as an ionisation source (In the case of SF6 a Partial-Discharge Ionisation source is needed.) and an ion shutter grid. Digital signal processing may provide a better readability of the obtained spectra or may automatically derive key characteristics from the spectrum as the peak position or signal amplitude [5]. It is possible to examine the change in the SF6 gas characteristics by comparing the change in the ion mobility spectra with the reference spectrum of IMS Reference pure SF6 gas. Figure 1 shows the results obtained Sparks Partial by laboratory analysis of a sparked sample and a Discharges sample aged through partial discharges compared with a reference spectrum. The shift of the peak towards shorter or longer drift times can be seen changes in the spectrum that comply with the known formation of a variety of decomposition Drift Time / ms products. The peak shift is caused by ions with different mobilities, while a broadening of the peak is the result of a greater variety of ions formed by Figure 1: Ion Mobility Spectra for SF6 aged gas. impurities produced during the discharges. [8-9]
Graph2 27.05.1999 12:43 J.I. Baumbach, Institut fr Spektrochemie und Angewandte Spektroskopie, ISAS, Dortmund
25
20
Signal / mV
15
10
30
35
40
45
50
55
Results and conclusion For investigations on SF6, a portable IMS for inspections and an IMS for continuous monitoring with partial discharge as ionisation source was developed (More details are summarised at WWW.GAS-Dortmund.de.) [10-12]. In initiated inspection programs these PD-IMS operated in several SF6-insulated substations in Germany during normal operation. The IMS was connected to the SF6-filled gas compartments with a thin polyethylene tube, providing a small flux of SF6 passing through the instrument, adjustable with a needle valve. To ensure comparable conditions, also reference spectra of pure SF6 were taken on site, which exhibited no significant difference to the spectra taken in the laboratory. Of special interest for a possible user is the data acquisition time of some minutes (including data processing), which was achieved in this experiment under real conditions. The flux was adjusted at about 2 L/h at ambient pressure. The total measurement time was about 5 min. The connection to the IMS was installed at the filling valve with an adapter and a Teflon line of 1 mm inner diameter. A needle valve made the adjustment of a constant small flux through the IMS possible. Several circuit breakers, current divider, voltage divider and other SF6-filled compartments of different substations are investigated by the procedure mentioned. In Figure 2 results from disconnecting switches are presented. The figure 2 a) shows the results of measurements at 3 phases of a disconnecting switch with a small peakshift of around 0.5 ms correlating to a low concentration of decomposition products inside. The length of the peakshift is correlated to the mean concentration of the decomposition products. In opposite to a) the results of investigations at another disconnecting switch shown in figure b) indicates a peakshift of 5.4 ms in one phase
93
correlating to a very high concentration of decomposition products. The identity and origin of the by-products may be investigated using more selective analytical methods like FTIR and GC-IMS. The performance of the monitoring system, however, is demonstrated through detection of the peak shifts. Directly after the measurements the disconnecting switch was opened by the user of the GIS and a heavily damaged spacer inside the switch was identified as possible error source. Accordingly the IMS-Measurement prevent a failure in the disconnecting switch.
1,2
Signal / a.u.
1,2
H:Stadtwerke Bochum ABB ABB F3F4 Lngstrenner Graph5 03.04.01 15:58 P. Pilzecker, G.A.S. Dortmund, Germany J.I. Baumbach, ISAS Dortmund, Germany
Peakshift
Graph6 01.08.01 14:56 P. Pilzecker, G.A.S. Dortmund, Germany J.I. Baumbach, ISAS Dortmund, Germany
1,2
Graph11 01.08.01 14:59 P. Pilzecker, G.A.S. Dortmund, Germany J.I. Baumbach, ISAS Dortmund, Germany
Peakshift of 5.4 ms !!
1,0
1,2 1,1
0,8
1,0
1,0
0,6
Signal / a.u.
0,4
0,2
a)
0 10 20 30 40 50 60 70 80 90 100
0,9
1,0
0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0,0 0 20 40 60 80 100
b)
Driftzeit / ms
Signal / a.u.
Driftzeit / ms
Signal / a.u.
0,8
0,0
0,8
0,6
0,6
0,4
0,4
0,2
0,2
0,0 30 32 34 36 38 40 42 44 46 48 50
0,0 30 35 40 45 50
Drift Time / ms
Drift Time / ms
Figure 2: IMS spectra obtained from two different disconnecting switches a) with small amount of decomposition products b) with a high amount of decomposition products (failure)
Inside a different GIS in Germany measurements has been carried out at a circuit breaker under operation before, while and after switching with load. Some seconds after switching a peakshift of 0.4 ms was detected as result of the switching (Figure 3). A
Peakshift of 10 different circuit breakers (3 phases)
DT 43,6 ms DT 44 ms DT 45 ms
1,0
1,0
0,8
E01 LS Phase c
E01 LS c
E01 Phase A
E02
E03
E04
E05
E06
E07
E08
E09
E10
1 1,5 1,8 1,5 1,2 1,4 1,4 1 1,6 1,6 1,7 1,3 1,5 1,3
0,8
Signal / a.u.
0,6
0,4
0,2
Phase B
30 40 50 60 70
0,0
20
Signal / a.u.
0,6
Drift Time / ms
Phase C
0,4
0,7 0,9 1,2 0,9 1,8 1,6 1,6 1,4 1,6 1,3
Remaining decomposition products in terms of products by a single switch with load Colour
0,2
Peakshift
Peakshift for switching once is
Switchings
0,0 40,0
0,4 ms !!
42,5 45,0
47,5
50,0
0,4 ms !!
Drift Time / ms
Figure 3: Spectra obtained from a circuit breaker before and after switching (with load) in correlation to pure SF6
Figure 4: Classification of all circuit breaker of a GIS depending on the peakshift (virtuel amount of switchings)
comparison of this result with measurements at the other circuit breakers in the same GIS shows, like displayed in Figure 4, that the detected peakshift are as wide as it would be switched with load several times. Some of the measured circuit breakers have a peakshift as it would be switched more then 4 times (See Figure 4). This leads to the conclusion that despite of the molecular sieves installed in the compartments a significant amount of by-products is still present in the gas of the circuit breakers after switching. The presumption of the manufacturer of circuit breakers that the decomposition products formed by the arc while switching recombine completely after switching is not true in all cases. Although molecular sieves are used, a total recombination of the decomposition products inside the circuit breaker was not achieved sometimes.
94
Employing IMS as presented here allows the set-up of an automated monitoring system. The structure of such a system, installed in the SF6insulated substation in the hydroelectric power plant of Foz do Areia (owned by COPEL), Brazil, is drafted in Figure 5. Crucial compartments of the substation are equipped with an IMS connected to the gas inside the compartments as shown in Figure 5. The acquired spectra were transferred by a local area network to the control room where the IMS server is placed. Here the spectra are evaluated through comparison with reference data, so that a database of decisions of maintenance becomes available. With an installed LAN-connection of Figure 5: the operating company the IMS can work without supervision by service personnel.
GAS-00-P1
Maintenance Group
GAS-00-P1
Control Group
GAS-00-P1 IMS-server Internet access for diagnosis and consulting
G.A.S.
Installation and Structure of an IMS-Monitoring System for automatic on-site sensing of SF6 -decomposition in the hydroelectric power plant of Foz do Areia.
References [1] Baumgartner, R., Fruth, B., Lanz, W., Petterson, K.,Partial Discharge - Part X PD in gasinsulated substations - measurements; practical considerations; IEEE Electrical Insulation Magazine 8 (1992) 16-27 [2] Kusumoto, S., Itoh, S., Tsuchiya, Y., Diagnostic technique of gas insulated substation by partial discharge detection. IEEE Transactions on Power Apparatus and Systems PAS-99 (1980) 1456-1463 [3] van Brunt, R.J.: Processes Leading to SF6 Decomposition in Glow-type Corona Discharges. SPIG 1988 (1988) 161 [4] Roehl, J.E., Environmental and Process Applications for Ion Mobility Spectrometry. Appl.Spectrosc.Rev.26 (1991)1-57 [5] Eiceman, G.A., Advances in Ion Mobility Spectrometry: 1980-1990. -Crit.Rev.Anal.Chem.22 (1991)17-36 [6] Eiceman, G.A., Karpas, Z.: Ion Mobility Spectrometry, CRC Press, Boca Raton, Ann Arbor, London, Tokyo, 1994 [7] St.Louis, R.H., Hill, H.H., Ion mobility spectrometry in analytical chemistry. Crit.Rev.Anal.Chem.21 (1990)321-355 [8] Baumbach, J.I., Pilzecker, P., Trindade, E., Meinders, J.: Diagnosing the Health Of SF6 Switchgear Transmission &Distribution Jan. 2000 [9] Pilzecker, P., Baumbach, J.I., Trindade, E.: Proc. of the IEEE International Symposium on Electrical Insulation Anaheim, CA, USA April 2-5, 2000; 400-403 [10] Held, M., Kurrat, M., Rickert, F., Baumbach, J.I., Segundo, S.M.A., Soppart, O., Klockow, D., Versuchsaufbau zur Analyse von Zersetzungsprozessen im Isoliergas SF6 durch Teilentladungen. - Elektrizittswirtschaft 96(1997)290-293 [11] Soppart, O., Baumbach, J.I., Alberti, S.M., Klockow, D., On-site Quality Assessment of Sf6 using Ion Mobility Spectrometry. - Conference Proceedings 10th Internat.Symp. on High Voltage Engineering, Montreal, Canada, August 25 - 29, 1997, vol. 4, p. 147-150 [12] Baumbach, J.I., Klockow, D., Kurrat, M., Segundo, S.M.A., Soppart, O., Verfahren zur berwachung des Qualittszustandes des Fllgases Schwefelhexafluorid in gasgefllten Anlagen. - Patent DE 195 28 290 (3.8.1995)
95