Journal of ELECTRICAL ENGINEERING, VOL. 56, NO. 1-2, 25, 31 35 HOW GOOD ARE ELECTRICAL DIAGNOSTIC TESTS TO PREDICT WATER TREES IN HV CABLES AN EVALUATION USING OPTICAL MICROSCOPY Motaheruddin Ahmed Mohammad A. Al-Ohali Mohammad A. Garwan Khaled Y. Soufi Zakariya M. Hamouz DC Current, DC Conductivity and Partial Discharge tests were performed as diagnostic tools to indicate the presence of water trees in 15 kv, 35 kv and 69 kv underground field-aged high voltage cable samples from the Eastern Province of Saudi Arabia. The insulation micro-sections of these cable samples were then observed under an optical microscope for visual detection of water trees. The DC Current and the DC Conductivity tests showed insulation deterioration in the 15 kv, 35 kv and the 69 kv cable samples, with the possibility of existence of water trees in them. Optical microscopy confirmed the presence of water trees in the 35 kv and 69 kv cable samples, but not in the 15 kv sample. The Partial Discharge test carried out on one 35 kv cable sample of sufficient length also showed insulation deterioration. However, optical microscopy could not detect any water tree in this cable sample. K e y w o r d s: DC current test, DC conductivity test, partial discharge test, optical microscopy 1 INTRODUCTION Polymeric insulation in power cables was first introduced over 5 years ago. However, premature degradation of the HV cables still remains a costly problem for the utility sector. A major cause of this premature degradation can be attributed to a phenomenon called water treeing in the cable insulation [1]. Diagnostic testing for the condition assessment of power cables is increasingly becoming important. A fundamental interest of power utilities is to increase the total reliability at a minimum cost by improved maintenance routines. Diagnostic testing of installed cross-linked polyethylene (XLPE) power cables is of high interest because of the large number of old cables in service with high probability of failure caused by water tree degradation. In a study undertaken by the investigators to analyze water trees in underground HV cables in the Eastern Province of Saudi Arabia using nuclear microscopy [2], a large number of cable samples of three voltage ratings and from three different locations were collected. Before embarking on visually detecting water trees in the insulation micro-sections of these cable samples, a painstaking and labour-intensive process, it became necessary to screen the cable samples for possible existence of water trees. Several electrical diagnostic tests have been tried by various researchers. The present study is an attempt to evaluate the effectiveness of three such tests: DC Current, DC Conductivity and Partial Discharge. 2 EXPERIMENTAL METHODOLOGY 2.1 DC Leakage Current Test The experimental set-up for the measurement of the DC current is shown in Fig. 1. The set-up consists of the cable sample, HV DC source, insulating holder, and a measuring circuit. The HV source was rated 14 kv DC, 5 kva. The insulating holder was used to mount the cable above the ground level. The measuring circuit consisted of a stable resistor, a protecting circuit and a digital voltmeter. The whole set-up had one ground, ie, the HV source, cable sheath and the measuring circuit were all grounded to the same point. 2.2 Procedure for DC Leakage Current Measurement DC voltage was raised in steps of 1 2% of the test voltage. The reading of the voltmeter was recorded after 15 seconds from the time of applying the DC voltage. This elapsed time was found to be sufficient for the capacitive current to decay when the voltmeter reading became constant. Then the voltage was raised and again the voltmeter reading was recorded after the elapse of the discharging time of the capacitive current. The process of applying a voltage and recording the voltmeter reading continued until the testing voltage was reached. The cable DC current is calculated as the voltmeter reading divided by the measuring resistance. This procedure was repeated for all suitable cables, unused and used. In the literature [3] it has been reported that if the slope of King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia, Address for correspondence: Dr M. Ahmed, Box 196, Dhahran 31261, Saudi Arabia, E-mail: mahmed@kfupm.edu.sa ISSN 1335-3632 c 25 FEI STU
32 M. Ahmed M. Al-Ohali M. Garwan K. Soufi Z. Hamouz: HOW GOOD ARE ELECTRICAL DIAGNOSTIC TESTS... C x Cable sample Cable Conductor HV Cable Sheath Coupling capacitor C HV Source Insulating Holder R V HV Source Z A PD Detector Coupling quadripole Fig. 1. Experimental set-up for DC current measurement. 8 6 4 2 Fig. 2. Experimental set-up for PD measurement. DC leakage current (na) 14 12 1 8 6 4 2 1 2 3 4 5 6 7 1 2 3 4 5 6 Fig. 3. Currentvoltage characteristic for an unused 35 kv cable sample. 6 5 4 3 2 1 Fig. 4. Current-voltage characteristic for a used 15 kv cable sample. 4 35 3 25 2 15 1 5 1 2 3 4 5 6 7 1 2 3 4 5 6 7 Fig. 5. Current-voltage characteristic for a used 35 kv cable sample. Fig. 6. Current-voltage characteristic for a used 69 kv cable sample. the current-voltage characteristic changes from a linear to non-linear relation, then this is a sign that the insulating material is deteriorating. 2.3 DC Conductivity Test The circuit set-up for the DC conductivity measurement was the same as that for the DC current test in Figure 1. The procedure followed for measuring the DC conductivity was similar to DC current measurement. The only difference was in the way of applying the DC voltage. In this case, a short circuit was made after the application of the DC step voltage. The DC current and time parameters were measured after the application of the DC step voltage and the short circuit. Different DC step voltages were tried and the corresponding current-time characteristics were compared. It has been reported [4, 5] that if the decay time is quite long, then this is a sign of the deterioration of the cable insulating material and, therefore, there is a possibility of water trees in them. 2.4 Partial Discharge (PD) Test The set-up for the PD measurement is shown in Fig. 2. It consisted of an AC voltage source, a Coupling Capaci-
Journal of ELECTRICAL ENGINEERING 56, NO. 1 2, 25 33 7 6 5 4 3 2 1 16 14 12 1 8 6 4 2 5 1 15 2 25 3 5 1 15 2 25 3 Fig. 7. DC conductivity current-time characteristic for the same used 15 kv cable sample for which the DC current test result (Fig. 4) also indicated insulation deterioration. Fig. 9. DC conductivity current-time characteristic for the same used 69 kv cable sample for which DC current test result (Fig. 6) also indicated insulation deterioration. 25 2 15 1 5 5 1 15 2 25 3 Fig. 8. DC conductivity current-time characteristic for the same used 35 kv cable sample for which the DC current test result (Fig. 5) also indicated insulation deterioration. tor, a Coupling Quadrupole (measuring impedance), and a PD Detector and the cable sample. 2.5 Procedure for Partial Discharge Measurement After connecting the cable to the PD set-up, the cable was excited from a 6 Hz voltage supply. The voltage was increased to 1.75U where U is the rated power frequency voltage between the conductor and the earth or metallic screen for which the cable is designed. Any PD signal above 5 PC was taken as a measure of cable deterioration as reported in the literature [6, 7]. 2.6 Optical Microscopy For optical microscopy to visually detect water trees, the first step was to remove the insulation of the cable samples from all the usual components of the cables. The bare insulations were microtomed into 1-micrometer thin sections using a rotary microtome. These microslices were then scanned under a high-power trinocular microscope at a typical magnification of 1. Images of water trees observed were recorded using a video camera attached to the microscope and stored in a PC. 3 RESULTS AND DISCUSSIONS 3.1 DC Current Test Before the field-aged cable samples were tested, an unused cable sample was measured as a check on the experimental conditions. The current-voltage characteristics for a 35 kv unused cable sample is shown in Fig. 3. A linear relationship can be seen, an indication that the cable insulation is undamaged and can be taken to be free of water trees. Figures 4 to 6 show the current-voltage characteristics for field-aged 15 kv, 35 kv and 69 kv cable samples, respectively. A non-linear relationship was observed for the three voltage ratings when the applied DC voltage was raised to about 4 kv. This non-linearity became very sharp when the voltage was raised above 5 kv. This result indicates a deterioration of the cable sample with a possible presence of water trees. 3.2 DC Conductivity Test DC conductivity measurements on an unused cable sample showed a very fast decay of DC current, which indicated that these cables were not electrically damaged. This has been found to be in conformity with the results obtained from the DC current characteristics. The DC current- time characteristics for the same used 15 kv cable sample is shown in Fig. 7. A long decay time can be seen and, as reported in the literature [4, 5], this is a sign of the deterioration of the cable insulating material possibly due to water trees. Similar results were obtained for the used 35 kv and 69 kv cable samples as shown in Figs. 8 and 9, respectively. From the above results, it is clear that the DC leakage current and conductivity tests data are in full agreement with each other. The same cables that were tested by the two techniques gave similar indication of the dielectric deterioration of the cable insulation. Therefore, either of these two tests can be used as a possible diagnostic tool for screening a large number of field-aged cables for the presence of water trees.
34 M. Ahmed M. Al-Ohali M. Garwan K. Soufi Z. Hamouz: HOW GOOD ARE ELECTRICAL DIAGNOSTIC TESTS... Fig. 1. Some typical bow-tie water trees in 1- µm thin insulation sections of the 35 kv cable sample. Fig. 11. Some typical bow-tie water trees in 1- µm thin insulation sections of the 69 kv cable sample. 3.3 Partial Discharge Test Partial Discharge (PD) measurement was made on the only available 35 kv cable sample of sufficient length ( 4 m) in accordance with international standards [8]. This cable sample is different from the 35 kv cable sample tested for DC current and conductivity tests discussed above. The PD measured value for this cable at 25 kv applied voltage was 85 pc, which is a very high value indicating that the cable sample has insulation deterioration. According to published literature [8], the PD value for a good cable should be lower than 5 pc. The high PD level in the cable sample tested can be attributed to either small cavities inside the XLPE insulation or the presence of water trees. 4 OPTICAL MICROSCOPY No water tree was observed in the 15 kv cable samples that were diagnosed to have insulation deterioration by the DC current and conductivity tests. The absence of water trees indicates insulation deterioration of these cable samples due to causes other than water tree growth. Typical water trees seen in the 35 kv cable samples are shown in Fig. 1. The DC current and conductivity tests also predicted the presence of water trees in these cables. Only bow-tie water trees were observed, no vented water tree was detected. The density of the water trees seen was estimated to range from 1cm 3 to 3cm 3. The sizes of the trees observed varied from about 1µm to about 4µm. Typical water trees observed in 69 kv cable samples are shown in Fig. 11. No vented tree was seen in these cable samples also. The DC current and conductivity tests indicated insulation deterioration these cable samples possibly due to water trees. The number of bowtie trees observed had an estimated density ranging from 5cm 3 to 1cm 3. The sizes of the trees observed varied from about 5µm to about 2µm. Compared with the water trees seen in the 35 kv cable samples, the trees found in these 69 kv samples were in general less densely distributed, smaller in size and lighter in colour. Although the Partial Discharge test on the only suitable 35 kv cable sample of sufficient length indicated a possible existence of water trees, optical microscopy failed to detect any water tree in this sample. However, no definite conclusion can be drawn about the efficacy of the PD test for diagnosing of the presence of water trees on the basis of results from only one sample. 5 CONCLUSIONS The DC Current and Conductivity tests were successful to reveal the presence of water trees in 35 kv and 69 kv cable samples confirmed by visual inspection using optical microscopy. These tests also showed insulation deterioration for the 15 kv sample. However, no water tree was detected in this sample. Nevertheless, we feel that these tests can be useful as diagnostic tools for water
Journal of ELECTRICAL ENGINEERING 56, NO. 1 2, 25 35 trees when a large number of field-aged cable samples are required to be screened. This can save time and labour needed to prepare and visually inspect such large numbers of cable samples by optical microscopy. The usefulness of the Partial Discharge test for this purpose remains doubtful however. Acknowledgements This work forms part of a research project funded by King Abdulaziz Center for Science & Technology, Riyadh, Saudi Arabia. It was carried out using the facilities at the Research Institute, King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia. References [1] SHAW, M. T. SHAW, S. H.: Water Treeing in Solid Dielectrics, IEEE Transactions on Electrical Insulation EI-19 (1984), 419452. [2] AHMED, M. Al-OHALI, M. GARWAN, M. HAMOUZ, Z.: A Study of Water Trees in Underground High Voltage Cables Using Nuclear Microscopy, A Research Project in progress at King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia. [3] NOBILE, P. LAPLANTNEY, C.: Field Testing of Cables: Theory and Practices, IEEE Transaction on Industry Applications IA-23 (1987), 786795. [4] TOYODA, T. MUKAI, S. OHKI, Y. LI, Y. MAENO, T.: Estimation of Conductivity and Permittivity of Water Trees in PE from Space Charge Distribution Measurements, IEEE Transactions on Dielectrics and Electrical Insulation 8 (21), 111116. [5] PATSCH, R. PAXIMADAKIS, A.: Correlation between Electrical Signals, Tree Size and Breakdown Voltages of Water-Treed Model Cables, Proc. 4th International Conference on Conduction and Breakdown in Solids, pp. 363 367, 1992. [6] MASHIKIAN, M. S. LUTHER, R. McIVER, J. C. JUR- CISIN, J. SPENCER, P. W.: Evaluation of Field-Aged Crosslinked Polyethylene Cables by Partial Discharge Location, IEEE Transaction on Power Delivery 92 (1994), 62628. [7] Electrical Test Methods for Electric Cables, Part 2: Partial Discharge Tests, IEC Report No. 885-2, 1987. [8] Electrical Test Methods for Electric Cables, Part 3: Test Methods for Partial Discharge Measurements on Length of Extruded Power Cables, IEC Report No. 6885-3. Received 27 May 24 Motaheruddin Ahmed obtained his PhD degree in experimental nuclear physics from University of Birmingham, England and is now a Professor at Physics Department, King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia. His present research interests include studies of material using atomic and nuclear techniques. Mohammad A. Al-Ohali obtained his PhD degree in Nuclear Physics from Duke University, USA. In 1993. He is an Associate Professor in Physics Department, King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia. Currently he is holding the position of Deanship, Graduate Studies at this university. Mohammad A. Garwan obtained his PhD in Nuclear Physics from the University of Toronto, Canada, in 1992. He is presently the Dean of Dammam Community College, King Fahd University of Petroleum and Minerals (KFUPM), and Associate Professor in the Physics Department, KFUPM. Previously he was the Director for Applied Physical Sciences, KFUPM and Chairman of the Physics Department, KFUPM. Khaled Y. Soufi is a Research Engineer III in the Center for Engineering Research, King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia. Mr Al-Soufi is supervising the High Voltage Laboratory. He has actively participated in many research projects related to high voltage and power systems. His research interest includes High Voltage testing and Applications, Power system analysis. Zakariya M. Hamouz received his PhD degree in Electrical Engineering from King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia in 1994 and is currently an Associate Professor in the same university. Dr Al-Hamouz is a Member of the Electrostatic Processes Committee of the IEEE Industry Applications Society.