SEM and EDX analysis of The Electrode System for Partial Discharge Inception Voltage Measurement and Arcing Test of The Mineral Oil

Abstract

The aim of this paper is to describe the Scanning Electron Microscope (SEM) and the Energy Dispersive X - Ray (EDX) test results of the original and tested electrodes used for Partial Discharge Inception Voltage (PDIV) measurement, as well as for the arcing test of mineral oil. The experimental investigations were performed with the mineral oil, Nynas 4000x, with the water content not more than 10 ppm under room temperature. For PDIV test, the tungsten needle electrodes with the tip radius of 10µm, 20µm, and 40µm respectively were used as the high voltage electrode while the brass plane electrode with 75 mm diameter was used as the grounded electrode with a gap distance of 50 mm. The test circuit was set up according to IEC 60270. The test procedure was performed in accordance with IEC 61294. For the arcing test, the tungsten rod electrodes with the tip diameter of 1 mm, and 2 mm with the curvature of 0.2 mm were used as the high voltage electrode while the brass plane electrode of 75 mm diameter as the grounded electrode with the gap distance of 0.3 mm and 0.8 mm respectively. The test experiment was modified from IEC 60156. The erosion of electrodes used for the mineral oil testing was examined by SEM techniques. SE images, BSE images and EDX spectrum of the original and tested electrodes were produced. The topography, the morphology, and the EDX spectra of the examined electrodes are analyzed. From the test results, there was no evidence to show the erosion of the electrodes after they were used for PDIV and arcing test. It can also be argued that the investigated tungsten needles, rods and brass plane electrodes can be used for PDIV testing and for arcing test without the problem of erosion. In addition, carbon was the main contamination created at the surface of the tested electrodes. The development of carbon was highly possible from the degradation of mineral oil. [PUBLICATION ABSTRACT]

Headnote

Abstract: The aim of this paper is to describe the Scanning Electron Microscope (SEM) and the Energy Dispersive X - Ray (EDX) test results of the original and tested electrodes used for Partial Discharge Inception Voltage (PDIV) measurement, as well as for the arcing test of mineral oil. The experimental investigations were performed with the mineral oil, Nynas 4000x, with the water content not more than 10 ppm under room temperature. For PDIV test, the tungsten needle electrodes with the tip radius of 10µm, 20µm, and 40µm respectively were used as the high voltage electrode while the brass plane electrode with 75 mm diameter was used as the grounded electrode with a gap distance of 50 mm. The test circuit was set up according to IEC 60270. The test procedure was performed in accordance with IEC 61294. For the arcing test, the tungsten rod electrodes with the tip diameter of 1 mm, and 2 mm with the curvature of 0.2 mm were used as the high voltage electrode while the brass plane electrode of 75 mm diameter as the grounded electrode with the gap distance of 0.3 mm and 0.8 mm respectively. The test experiment was modified from IEC 60156. The erosion of electrodes used for the mineral oil testing was examined by SEM techniques. SE images, BSE images and EDX spectrum of the original and tested electrodes were produced. The topography, the morphology, and the EDX spectra of the examined electrodes are analyzed. From the test results, there was no evidence to show the erosion of the electrodes after they were used for PDIV and arcing test. It can also be argued that the investigated tungsten needles, rods and brass plane electrodes can be used for PDIV testing and for arcing test without the problem of erosion. In addition, carbon was the main contamination created at the surface of the tested electrodes. The development of carbon was highly possible from the degradation of mineral oil.

Keywords: Scanning Electron Microscope; Partial Discharge Inception Voltage; Arcing test; Mineral oil

1. Introduction

Mineral oil is predominantly used as liquid-insulating material for high voltage equipment, especially power transformers. The mineral oil provides not only an excellent electrical insulation but an outstanding cooling property also. The quality of such insulation oil plays an important role in the performance of the transformers which are expected to function reliably and efficiently for many years [1]. To assure that the mineral oil operates in good condition, the oil quality has to be examined regularly. Simultaneously, the good maintenance of the transformer needs to be continuously performed. Many test standards for verifying the mineral oil characteristics have been proposed such as IEC 60156[2], ASTM D 877[3] or ASTM D 1816[4]. Partial Discharge Inception Voltage (PDIV) is one of the test techniques used to investigate the mineral oil characteristics. Currently only the standard IEC 61294 is available for PDIV measurement of liquid insulation. IEC 61294 recommends needle - sphere electrode for such PDIV measurement [5].However, there are some questions about the validity and sensitivity of this PDIV measurement technique[6]. Recently, the effort to improve the PDIV measurement of liquid insulation has been discussed and carried out continually. Additionally, new concepts for PDIV measurement of the insulation liquids have been proposed as in [7-12]. Focusing on the electrode system for PDIV measurement of the mineral oil, a needle - plane electrode is also used by some research groups [9-15].This electrode type shows the good performance for PDIV testing of the mineral oil.

Unfortunately, there is a lack of study for the erosion or damaging of the needle - plane electrode system used for PDIV testing. Moreover, the erosion study of the arcing rod is very useful for evaluating the deformation of the electrode system used for arcing phenomena research.

2. Erosion of the electrode used for electrical testing of mineral oil

The erosion of the electrodes may occur after the electrodes are used for insulation tests for a certain time. The erosion process is accelerated by the persistent discharges or arcing of the electrode system. The erosion of the electrodes is caused partly by the evaporation of the metal from electrode spots during the arcing [16]. Examples of the needle tip erosion used for three standard lightning impulse breakdown testings, 80kV negative polarity and 120 kV positive polarity, of perfluoropolyether (PFPE) and transformer oil (AGIP ITE 360) had been reported in [17]. The shape of the needles is dramatically distorted compared to the original one after the impulse breakdown test as shown in Figure 1.Generally, electrode materials, arc discharge voltage, arc current, arc temperature, and the insulating medium characteristics are the main factors of the electrodeserosion. The appearance of electrode spots relates to the electrode erosion behavior, higher temperature and smaller spot size leading to more erosion. Unfortunately, the erosion quantification is a difficult task. Weight loss is not a good measure of erosion because mass transport and redeposit ion in regions away from the arc attachment obscure the effect of the arc's operation [18]. An example of arc erosion rates of different material tested by 12,000 ampere, 60 Hz, half cycle arc is illustrated in Figure 2[19]. Furthermore, arcing in the mineral oil generates bubbles, gases, carbonization and other consequences. The carbonization appears as some black deposits in the oil and finally degrades the oil characteristics.

3. SEM and EDX test techniques

To study electrode erosion, Scanning Electron Microscope (SEM) is a very powerful instrument. The general structure of SEM equipment is depicted in Figure 3[20]. With SEM, it permits the observation and characterization of materials on a nanometer to micrometer scale [21]. The topography of electrodes is acquired by reading the secondary electrons (SE images). Moreover, the image of the surface morphology of a specimen as well as the information of the surface composition can be identified by reading the backscattered electrons (BSE images) and the Energy Dispersive X - Ray (EDX) respectively. Figure 4 shows beam specimen interaction signals released by the tested specimen after it undergoes the electron beam. These beam specimen interactions are responsible for a multitude of signal types: secondary electrons, backscattered electrons, X-rays, auger electrons, cathadoluminescence. Details of SEM and EDX analyses can be found in [21-23].

4. Test procedure

The experiment can be divided into 2 parts: The former was the mineral oil testing experiment, the PDIV experiment and arcing test; the latter was the SEM and EDX investigation of the original and tested electrodes.

A. Mineral oil testing

A.1. PDIVtesting

The tungsten needle electrodes with the tip radius of lO!1m, 20!1m, and 40!1m respectively were used as the high voltage electrode while the brass plane electrode with 75 mm diameter was used as the grounded electrode. The gap distance of the electrode system was set up at 50 mm. The test circuit was set up according to IEC 60270 [25]. The test procedure was performed in accordance with IEC 61294 [5]. The test voltage was increased with a rate of 1 kV/s from zero until PDIV occurred. Then, PDIV was recorded. Each needle was tested ten times and nine needles of each needle tip radius were investigated. After that, the PDIV mean value of each electrode configuration was computed. The PDIV test circuit is shown in Figure 5.

A.2. Arcing test

The tungsten rod electrodes with the tip diameter of 1 mm, and 2 mm with the curvature of 0.2 mm were used as the high voltage electrode, while the brass plane electrode of 75 mm diameter was used as the grounded electrode. The gap distance of the electrode system was set up at 0.3 mm and 0.8 mm respectively. The test experiment was modified from IEC 60156[2]. The test voltage was applied to the electrode system from zero until a complete arcing occurred. Then the test voltage was kept at that arcing voltage level for 10 seconds. The arcing current and arcing voltage were recorded. The oil sample was tested with the same procedure six times for each arcing rod. Then, the mean value of the arcing voltage and arcing current of each electrode configuration was computed.The arcing test circuit is illustrated in Figure 6.

B. SEM and EDX Investigation

The original and tested electrodes, six needles, six rods, eight plane electrodes, were randomly selected to investigate by SEM and EDX test techniques. SEM and EDX investigation was performed by FELMI-ZFE, The Austrian Centre for Electron Microscopy and Nano Analysis. SE images, BSE images and EDX spectrum of the electrodes were created and analyzed.

V. Test results

A.PDIV and Arcing test results

According to the experiment, the PDIV of the mineral oil (UPDIV) tested with the needle - plane electrode system strongly depends on the needle tip radius. The PDIV test results of 10ìm, 20ìm, and 40ìm tip radius needles are shown in Table 1. For acing test, the mean value of the arcing voltage (Uarc) and the arcing current (Iarc) of the mineral oil are illustrated in Table 2.

B. SEM and EDX test results for original and PDIV tested needles

SEM and EDX test results of the needles, 10µm, 20µm, and 40µm tip radius, before and after they were used for PDIV testing of the mineral oil, were relatively similar. The examples of the 20 µm tip radius needle topography obtained from SE images and the needle morphology acquired from BSE images are illustrated in Figure 7.

BSE images revealed that both original and PDIV tested needles comprised of at least 2 types of material. To analyze the composition elements, the EDX examination was conducted. EDX spectra generated by the areas at the tip of the original needle revealed that tungsten was the major element and carbon was the minor element. The example of EDX analysis at the tip of the 10 11m tip radius needle is shown in Figure 8. Furthermore, oxygen, sometimes with high count rate, copper, calcium and aluminum were found, as well as silicon and magnesium.

C. SEM and EDX test results fororiginal and arcing tested rods

The SE images and BSE images of the original and arcing tested rods with a diameter of 1 mm, gap spacing of 0.8 mm, arcing current 132.9 rnA, arcing current density 16.93 A/cm2 are illustrated in Figure 9 a) -f) and for gap spacing of 0.3 mm, arcing current 389 rnA, arcing current density 49.53 rnA!cm2 as shown in Figure 9 g) -j) respectively.

The topography at the tip of the arcing tested rod as shown in Figure 9 h)has clearly changed from the original, Figure 9 g), because it was covered by the contamination layer generated from the arcing phenomena. However, there was no evidence to show that any part of the rod melted or eroded. The new composition was investigated by the EDX technique. Figure 10 represents the example of the composition elements of the area of interest, area 2 of Figure 9 j), of the tested rod. The main component was comprised of tungsten and carbon.

The SE images and BSE images of the original and arcing tested rods with a diameter of 2 mm with the curvature of 0.2 mm, gap spacing of 0.8 mm, arcing current 163.2 rnA, arcing current density 5.20 mA/cm2 are illustrated in Fig. lla) - f) respectively.

The main component of other arcing tested rods was also tungsten and the main component of the contamination was carbon as well.

D. SEM and EDX test results for original and tested plane electrodes

SEM and EDX techniques were employed for analysis the plane electrodes as well. The main component of the original plane electrode was copper and zinc. For the PDIV tested plane electrodes, the surface morphology was a bit changed. BSE images showed the clearly changing of the surface morphology of the arcing tested plane electrodes especially at the arcing point. However, there was no evidence to show that any part of the tested plane electrodes melted or eroded. Figure 12 and Figure 13 represent the BSE images of the PDIV and the arcing tested plane electrodes respectively.

BSE images revealed that the surfaces of PDIV and arcing tested plane electrodes comprised of at least 3 types of materials. EDX spectrum as depicted in Figure 14 shows that the major component consists of carbon, zinc and copper. Carbon was generated from the degradation of the mineral oil during the existence of partial discharge and the arcing process. Zinc and copper were the major elements for manufacturing the plane electrodes.

6. Conclusion

The SEM and EDX test results of the needle - plane and the rod - plane electrodes for the PDIV and the arcing test of the mineral oil may be concluded as the following:

1. There was no evidence to show the erosion or melting of the needles, rods and brass plane electrodes after they were used for PDIV and arcing test. It can be stated that the tungsten needles with tip radius of 10 11m, 20 11m, and 40 11m can be used for PDIV testing without the problem of erosion. Rods with 1 mm or 2 mm diameter including the brass plane electrode scan be used for arcing test also without the problem of erosion.

2. Carbon, very likely generated from the degradation of mineral oil, was found at the surface of electrodes; this was especially present at the tip of the arcing tested rods and at the arcing point of the tested plane electrodes.

3. The intensity of carbon depended on the arcing current density. The collected carbon on the tip of the arcing tested rod changed the topography and the surface morphology of the original rod which may affect the scattering of the arcing voltage and the arcing current of the mineral oil. At low current density, most carbon was found at the plane electrodes. With higher current density, carbon was found at the plane electrode and also the rod.

Acknowledgment

The authors would like to express their thanks to Dipl.-Ing. Dr.techn. Univ.-Doz. Peter Palt and Dipl.-Ing. Dr.techn. Stefan Mitsche for SEM investigation. The financial support for this research work is from BaurPrill- und Messtechnik GmbH.

References

Reference

[1] Marcelo M. Hirschler, Electrical Insulating, Materials: International Issues, ASTM STP 1376, March 2000, pp. 82 - 95.

[2] IEC 60156, Insulating Liquids - Determination of The Breakdown Voltage at Power Frequency - Test Method.

[3] ASTM D877, Standard Test Method for Dielectric Breakdown Voltage of Insulating Liquids U sing Disk Electrodes.

[4] ASTM D18l6, Standard Test Method for Dielectric Breakdown Voltage of Insulating Oils of Petroleum Origin Using VDE Electrodes.

[5] IEC 61294, Insulating Liquids - Determination of The Partial Discharge Inception Voltage (PDIV) - Test Procedure.

[6] X. Wang and Z. D. Wang, "Discussion on The Effectiveness of IEC 1294:1993, Insulating Liquids- Determination of The Partial Discharge Inception Voltage(PDIV)Test Procedure", 11th INSUCON International Electrical Insulation Conference, Birmingham, UK, 26-28 May 2009.

[7] M. Pompili, C. Mazzetti and R. Bartnikas, "Testing, Evaluation and Standardisation of Transformer Oils", ICDL 2005, 2005 IEEE International Conference on Dielectric Liquids, June, 26- July 1,2005, Coimbra, Portugal, pp.36l-364.

[8] Z. D. Wang, Q. Liu, X. Wang. P. Jarman, G. Wilson, "Discussion on Possible Additions to IEC 60879 and IEC 61294 for Insulating Liquid Tests", lET Electr. Power Appl., 2011, Vol. 5, Iss. 6, pp. 486-493.

[9] CIGRE 436, Experiences in Service with New Insulating Liquids, Working Group, A 2.35, October 2010.

[10] N. Pattanadech, F. Pratomosiwi, B. Wieser, M. Baur, M. Muhr, "The Study of Partial Discharge Characteristics of Mineral Oil Using Needle - Plane Electrode Configuration", ICHVE 2012: The 2012 International Conference on High Voltage Engineering and Application, Sep. l7, - 20,2012, China.

[11] N. Pattanadech, F. Pratomosiwi, B. Wieser, M. Baur, M. Muhr, "The Study of Partial Discharge Characteristics of Mineral Oil Using Needle - Plane Electrode Configuration Base on Partial Discharge Pulse Current Measurement", ICHVE 2012: The 2012 International Conference on High Voltage Engineering and Application, Sep. 17-20, 2012, China.

[12] N. Pattanadech, F. Pratomosiwi, M. Baur, M. Muhr, "The Influence of The test Methods on The Partial Discharge Inception Voltage Value of The Mineral Oil Using The Needle - Plane Electrode Configuration", CMD 2012: 2012 IEEE International Conference on Condition Monitoring, Sep. 23-27, 2012, Bali, Indonesia, pp 597-600.

[13] B. Dolata, H. Borsi, E. Gockenbach, "Comparison of Electric and Dielectric Properties of Ester Fluids with Mineral Based Transformer Oil", XV th International Symposium on High Voltage Engineering, Slovenia, August 27-31,2007.

[14] Massimo Pompili, "Partial Discharge Development and Detection in Dielectric Liquids", IEEE Transactions on Dielectrics and Electrical Insulation, Vol. 16, No.6, December 2009,pp.1648-l654.

[15] Chathan Cooke and Wayne Hagman, "Non - Destructive Breakdown Test for Insulation Oil", EPRI Substrations Diagnostic Conference, New Orleans, November 1994.

[16] Mazen Abdel- Salam, Hussein Anis, Ahdab EI-Morshedy, Roshdy Radwan, High Voltage Engineering, Theory and Practice; second edition, revised and expanded, p. 185.

[17] S. Patrissi, M. Pompili, H. Yamashita, E.O. Forster, "A Study of The Effect of Electrical Breakdown in Dielectric Liquids on The Needle Point Structure", IEEE: 11th International Conference on Conduction and Breakdown in Dielectric Liquids(ICDL),Baden-Dtittwil, Switzerland, July 19-23, 1993, pp 376-382.

[18] X. Zhou and J. Heberlien, "An Experimental Investigation of Factors Affecting Arc - Cathode Erosion", J. Phys. D: Appl. Phys. 31,1998, pp. 2577 -2590.

[19] W.R. Wilson, "High-Current Arc Erosion of Electric Contact Material", Transactions of The American Institute of Electrical Engineers, Part III: Power Apparatus and System, volume 74, Issue: 3, August, 1955, pp 657-664.

[20] http://www 4.nau. edulmicroanalysis/Microprobe-SEMlInstrumentation. html

[21] Joseph I. Goldstein et al, Scanning Electron Microscopy and X Ray Microanalysis, Third edition, 2003.

[22] Bob Hafner, Scanning Electron Microscopy Primer, http://www.charfac.unm.edu /sem_primer.pdf.

[23] Bob Hafner, Energy Dispersive Spectroscopy on the SEM: A Primer, http://www. charfac. unm. edu/instruments/eds on _ sem ~rimer.pdf.

[24] Asa Kassman Rudolphi, Scanning Electron Microsopy (SEM) and Scanning Probe Microscopy (SPM), Angstrom laboratoriet. http://www.cb.uu.se/~ewertiSEM SPM bildanalys. pdf

[25] IEC 60270, High Voltage Test Techniques - Partial Discharge Measurement, 2002-12.