The line-lightning performance and mitigation studies of shielded steelstructure

Transcription

1 The line-lightning performance and mitigation studies of shielded steelstructure distribution lines ASNAWI MOHD BUSRAH, MALIK MOHAMAD Energy System Group TNB Research Sdn Bhd No 1, Lorong Ayer Hitam, Kajang, Selangor MALAYSIA Abstract: - The over voltage on the electric power line is the function of line design and the lightning incidence. The occurrence of lightning in Malaysia is amongst the highest recorded in the world and it is beyond our control. However the designs of power line are within the engineer s control. As overhead distribution lines and their associated equipments are most prone to insulation failure due to lightning and very much affected in terms of reliability, a study was intended to determine the line lightning performance and to analyze available mitigation options focusing on medium voltage overhead distribution line. In this paper, simulation using commercially available software to determine the lightning performance of the chosen lines will be discussed. A Monte Carlo statistical method is applied for the simulation of lightning activity, while a three-dimensional Electro-geometric model is adopted for the determination of stroke termination. Shielding from the nearby objects is also taken into account. Linear and non-linear tower footing resistance (TFR) representation, counterpoises, leader propagation flashover model are among standard features used in the simulations. Simulations are carried out to study the effects of critical flashover (CFO) voltage, tower footing resistance (TFR) and ground flash density (GFD) on the line lightning performance. In addition, several mitigation steps to improve the lightning performance are also recommended. Key-Words: -Line lightning performance, Critical flashover voltage, Tower footing resistance, Ground flash density 1 Introduction One source of over voltages on the equipments comes from the overhead line. The lightning strike on the overhead line causes over voltage to be present and at the same time, that over voltage can travel from the line to other equipments connected to it. Based on breakdown statistics of Malaysia power utility, Tenaga Nasional (TNB) from 2001 to 2007, almost half of the total faults occurring on the medium voltage overhead line are caused by lightning and transient. Furthermore, resent observations performed by the Malaysian Meteorological Department indicate that in average, thunders occur 200 days a year in Malaysia. For the period from 2004 to 2007, nearly 9 million lightning strikes were detected in the Peninsular Malaysia using TNB owned Lightning Detection System (LDS) currently being maintained and operated by TNB Research, a subsidiary of TNB [1]. As overhead lines and their associated equipments are most prone to insulation failure due to high frequency of lightning occurrences in Malaysia, a study was initiated to determine the line lightning performance and also to analyze available mitigation options. The study will focus on medium voltage overhead distribution lines since they are very much affected as far as reliability is concerned. Many works in determining lightning performance for transmission lines have been carried out in recent years where the focus is more on the shielding effects [2-6]. Research on lightning performance of distribution lines in past years addressed more on the effect of CFO apart from induced voltage effect [7-11]. This paper will discuss the simulation in determining the lightning performance of the chosen lines using commercially available software. Simulation was also carried out to analyse the effects of critical flashover (CFO) voltage, tower footing resistance (TFR) and ground flash density (GFD) on the line lightning performance. ISSN: ISBN:

2 2 Data Collections and Site Measurements A 33kV bare shielded overhead distribution line was chosen as case study. The line which is approximately km runs through the palm oil estate along the coastal area of a northern state of Peninsular Malaysia. The ground profile is relatively flat between 9 m to 22 m above sea level. 2.1 Line Parameters The line consists of 106 lattice towers with heights between 23 m to 25 m. There are basically four types of lattice tower on the line: intermediate tower (L-type), section tower, angle tower and terminal tower. The line is installed with shield wire with 0 shielding angle. Only nine of the towers are installed with lightning arrester and the towers without lightning arrester are installed with arcing horns on the phase wire. Two lightning arresters are installed at each substation terminal tower. The other arresters are distributed throughout the line with heavy concentrations between towers 76 to Site Measurement There are two important parameters measured at site which are soil resistivity and tower footing resistance (TFR) Soil-resistivity The result of soil resistivity measurements around selected towers along the line are summarized in Table 1. The sample measurements were analyzed and summarized into the soil model based on layer. The models of the soil were obtained using CDEG (Current Distribution, Electromagnetic, Grounding and soil analysis) software. Table 1: Soil resistivity values and the soil model 2.2 Tripping data The history of breakdown records for the line was gathered from the site office. Based on the past tripping data, the actual line lightning performance is calculated as 52.4 flashover per 100 km-years. 2.3 Lightning parameters The line lightning parameters were obtained from TNB LDS. The GFD is between 10 to 23.3 flashes/km 2 -year, with average of 16.6 flashes/km 2 - year. A comprehensive Ground Stroke Density Map for the area surrounding the line is shown in Fig. 1. Although the suggested model consists of three layers, but there was limited information about the actual earthing system connected to the tower. Furthermore, the software will only allow input to the soil model up to two layers. In the simulation, only one layer soil model is used. This is important for the simulation when simulating high discharge current striking the lines. At high discharge current, soil ionization may occur and this may lower the TFR Tower Footing Resistance (TFR) Besides soil resistivity measurements, actual TFR measurements were also taken. The summary of TFR measurements are tabulated in Table 2. Table 2 shows the line is having very good TFR which is below 10 Ω as suggested by [17], that if CFO is less than 200kV, TFR must be less than 10 Ω. Fig. 1 Ground Stroke Density Map of the Line ISSN: ISBN:

3 Table 2: Tower Footing Resistance Measurement Pole No. TFR (Ω) Lightning Performance Assessment Sigma SLP software was used for the simulation. The software package was originally developed in order to enable design of the lightning protection of transmission and distribution lines, with special reference to the application of line surge arresters. In the determination of line lightning performance, a Monte Carlo statistical method is used for the simulation of lightning activity, while a threedimensional Electro-geometric model [12-14] is adopted for the determination of stroke termination. Shielding from the nearby objects is also taken into account. Linear and non-linear tower footing resistance (TFR) representation, counterpoises, leader propagation flashover model, linear or upward concave stroke front, initial voltages and etc are standard features used in the simulations. Electromagnetic transients on the line are computed by the multiphase travelling wave method. Transients on the line are computed separately from the transients on the towers, corresponding to the connections being performed by Thevenin equivalent. were assumed to be 12m from the line for both sides. A few GFD values were also tested in the simulation. From the literatures [15-17], it is assumed that there are three strokes per flash. Based on actual lightning data obtained from the LDS, it was found that there is about 1.2 to 1.6 stroke per flash. The GSD obtained were then converted to GFD using either 3 strokes per flash or 1.2 strokes per flash in order to be conservative in the simulation. For the TFR readings, a few values were chosen in the simulation based on the measured values. One of those values was selected in order to obtain a conservative simulation result. At the same time, using the existing line data, the value of TFR varied from 3 Ω to 200 Ω to see the effect of TFR versus flashover rate as shown in Fig. 2. The results were obtained using GFD = 16.6, assuming two arcing horns were installed on the line (CFO = 165 kv). Fig. 2 Flashover rate versus TFR 3.1 Simulation Results and Discussions A few different settings for lightning were used in the simulations; include different peak current distribution suggested by the software: the original CIGRE peak current distribution and the ground CIGRE peak current distribution. The argument on the ground CIGRE distribution is that the original CIGRE measurements were distorted because of the tower used to capture the measurements. Since the tower structure may attract higher discharge current to hit the tower, the actual peak discharge current distribution should be corrected to the ground level without any tower structures. The effect of nearby object was also taken into account in the simulation since they are quite consistent. The palm oil trees with height of 10m From Fig. 2, TFR play an important role in the line lightning performance with the shielded line. By changing the TFR from 1 Ω to 100 Ω, the flashover rate almost doubles. There are a few tower installed with two arcing horns, one arcing horn, and no arcing horn. The present of arcing horns across the insulator reduces the effective CFO on the line. Therefore, different CFO was used in the simulation to represent the arcing horn being installed on the line. Based on the tower design and configuration, the CFO value is calculated as 341 kv without arcing horn. With the installation of one arcing horn, the CFO reduces to kv and adding another arcing horn will reduce the CFO to 165 kv. Besides arcing horn, contamination on the pin insulator can also reduce the actual CFO of the line. However this effect ISSN: ISBN:

4 cannot be determined without proper lab testing on the selected pin insulator from the line. The important results are summarized in Table 3 where different CFO and GFD values were applied, while maintaining the TFR with 10 Ω. Table 3: Simulation results under various CFO and GFD using ground and original CIGRE lightning peak current distribution CFO (kv) (no arcing horn- CFO=341kV) (1 arcing horn- CFO=236.5kV) (2 arcing horn- CFO=165kV) Flashover rate (flashes/100kmyear) (CIGRE Ground) 0 (GFD=16.6) 0 (GFD=6.4) 3.93(GFD=16.6) 0.51(GFD=6.4) 24.7(GFD=16.6) 10.70(GFD=6.4) Flashover rate (flashes/100kmyear) (CIGRE Original) 0.71(GFD=16.6) 0.27(GFD=6.4) 0 (GFD=16.6) 11.87(GFD=16.6) 4.58(GFD=6.4) 2.37(GFD=16.6) 51.51(GFD=16.6) 19.86(GFD=6.4) 16.85(GFD=16.6) 4 Mitigation Options Analysis In this section specific recommendations are made in order to improve the line lightning performance of the case study line. Please bear in mind that since each individual line is different from each other, the recommendation will also be different and tailor made to unique characteristic of the lines concern. As shown earlier, the simulation results show quite a good match with actual data at site in terms of line lightning performance. In obtaining the simulation result, the line is assumed to be installed with two arcing horns. The installations of arcing horn bring down the CFO between the phase conductor and the ground from 341kV to almost 165kV. Therefore, the obvious mitigation step can be taken in order to improve the line lightning performance is to improve the CFO of the lines. 4.1 Improving CFO Simulation results with the improving CFO are tabulated in Table 4. Table 4: Flashover rate versus CFO 3.2 Comparison with the Actual Data The simulation results were compared with the actual data obtained from site. The simulation based on the original CIGRE peak distribution current, TFR value equal 10 Ω, GFD value of 16.6 and two arcing horn, resulted in total flashover rate equal to This value was quite close with the actual flashover rate which was If the arcing horn is not included, the simulated line lightning performance drops to Bear in mind that even without arcing horn, CFO can still drop because of contamination. Since the line is along the coastal area, contamination may play a role in bringing down the actual CFO value on the line insulation. If the ground CIGRE peak distribution current was used, while maintaining the two arcing horn, the simulated line performance drops to The original CIGRE is more appropriate than the ground CIGRE since the line is higher than the rest of the palm oil tree. Therefore, the line will attract more lightning. Based on the simulation results, the CFO of the line improve significantly by taking out the arcing horn installed along the line. However, in practical this step may jeopardize the protection of the pin insulators should extremely high fault current pass through them. Adding pin insulator on the line can also improve the CFO of the line. In the line design, there are a lot of spaces to put more pin insulators. For straight through tower, the distance between two cross arms is about 2.7m. The three pin insulators take about 0.62m from the top cross arm. Therefore, there is still about 2.08m from the middle cross arm to the top phase conductor. By installing another pin insulator between phase conductor and cross arms will not compromise the striking distance between the top conductor and the middle cross arm. ISSN: ISBN:

5 Contamination on the pin insulator can also bring down the actual CFO of the line. Regular cleaning of pin insulators from contamination is recommended as it can improve the CFO of the lines. If the frequency of cleaning is prohibitive in term of interruption or maintenance cost then, the step can be taken is to install more pin insulators provided the tower design permit it. With more pin insulators, the CFO after contamination will settle down at higher value than the CFO for 3 pin insulators. 4.2 Improving TFR Another mitigation step is to improve the tower footing resistance. Simulations were repeated with different TFR and the results are tabulated in Table 5. The TFR values were varied from 10Ω to 0Ω and the flashover rate drops from 51.5 to The assumption made is that all the tower having a same TFR which is not practical in real lines. Table 5: Flashover rate versus TFR TFR(Ω) Flashover rate per 100km -yr This mitigation step may not be that practical since the tower footing resistance used in the simulation was already low which 10Ω. It is quite difficult to achieve a low tower footing resistance. Furthermore, the TFR is not consistent and vary over time and also weather conditions. 4.3 Installation of Lightning Arresters The third mitigation step which can be taken is to install more lightning arrester (LA) along the lines. The simulations were carried out using TFlash software. The simulations start without any LA, until all towers were installed by LA. The simulation results with different percentage of LA installed along the line are represented in a graph as shown in Fig. 3. Fig. 3 The flashover rate versus different percentage of LA installed along the line From the simulation, even at 75% of the line was installed with lightning arrester, may result only about 40% improvement on the line lightning performance. Only if the arresters were installed on every tower then the line lightning performance approach zero flashover rates assuming there is no LA failure, as proved in [17]. 5 Conclusion In this study, simulation result of the selected line s lightning performance is quite comparable with the actual flashover rate at site. Apart from estimating lightning performance, simulations were also carried out to analyse several mitigation steps. Based on the simulation studies, it is best to improve the CFO of the line first before any other mitigation steps can be taken. Testing can be done on the selected pin insulators in order to properly estimate any reduction on the actual CFO cause by contamination. If the actual CFO is lower than expected CFO, then cleaning on the pin insulator may be required. If the frequency of cleaning is prohibitive in term of interruption or maintenance cost then it is suggested that extra pin insulators be added to improve the line CFO. Since each line is unique, the mitigations will also be different and tailor made to unique characteristic of the lines concern. There is still need to study individual lines and individual cases before proper solution can be implemented. Acknowledgement: An acknowledgement is made to Distribution Division of TNB for their support, valuable inputs and assists especially in data collection. Special mention also goes to Universiti Tenaga Nasional for ISSN: ISBN:

6 their cooperation and collaboration in completing the research project. References: [1] N. Abdullah, M.P. Yahaya, N.S Hudi, Implementation of Lightning Detection System Network in Malaysia, 2nd IEEE International Conference on Power & Energy, Johor Bahru, Malaysia,2008. [2] Brown, G. W., and Whitehead, E. R., Field and Analytical Studies of Transmission Line ShieldingÑ II, IEEE Transactions on Power Apparatus and Systems, vol. 88, pp , [3] Love, E. R., Improvements on Lightning Stroke Modeling and Applications to Design of EHV and UHV Transmission Lines, M. Sc. Thesis, University of Colorado, Denver, [4] Mousa, A. M., and Srivastava, K. D., Effect of Shielding by Trees on the Frequency of Lightning Strokes to Power Lines, IEEE Transactions on Power Delivery, vol. 3, pp , April [5] Mousa, A. M., and Srivastava, K. D., The Lightning Performance of Unshielded Steel- Structure Transmission Lines, IEEE Transactions on Power Delivery, vol. 4, pp , Jan [6] Mousa, A. M., and Srivastava, K. D., The Distribution of Lightning Strokes to Towers and Along the Span of Shielded and Unshielded Power Lines, Canadian Journal of Electrical and Computer Engineering, vol. 15, pp , [7] Jacob, P. B., Grzybowski, S., and Ross, E. R., An Estimation of Lightning Insulation Level of Overhead Distribution Lines, IEEE Transactions on Power Delivery, vol. 6, no. 1, pp , [8] Jacob, P. B., Grzybowski, S., Libby, L., and Barsley, P. K, Experimental Studies of Critical Flashover Voltage on Distribution Line Construction, IASTED, International Journal on Energy Systems, no. 1, pp , [9] Ross, E. R., and Grzybowski, S., Application of the Extended CFO-Added Method to Overhead Distribution Configurations, IEEE Transactions on Power Delivery, vol. 6, no. 4, pp , Oct [10] Shwehdi, M. H., Investigation and Analysis of Lightning Impulse Strengths of Multiple Dielectrics Used in Electrical Distribution Systems, Ph.D. thesis, Mississippi State University, [11] Shwehdi, M. H., and El-Kieb, A. A., Lightning CFO of Multiple Series Dielectrics Used on Distribution Systems, IASTED Proceedings, Power High Tech 89, Valencia, Spain, pp , July 4 7, [12] Eriksson, A. J., An Improved Electrogeometric Model for Transmission Line Shielding Analysis, IEEE Transactions on Power Delivery, vol. 2, pp , July [13] Dellera, L., and Garbagnati, E., Lightning Stroke Simulation by Means of the Leader Propagation Model Parts I and II, IEEE Transactions on Power Delivery, vol. 5, no. 4, pp , Oct and discussion in vol. 6, no.1, pp [14] Rizk, F. A. M., Modeling of Transmission Line Exposure to Direct Lightning Strokes, IEEE Transactions on Power Delivery, vol. 5, pp , Oct [15] Hileman, A. K., Insulation Coordination for Power Systems, Marcel Dekker, Inc., pp 214. [16] McDermott, T. E., Short, T.A., and Anderson, J. G., Lightning Protection of Distribution Lines, IEEE Transactions of Power Delivery, vol. 9, no. 1, pp , Jan [17] IEEE Power Engineering Society, IEEE Std , IEEE Guide for Improving the Lightning Performance of Electric Power Overhead Distribution Lines, [18] CIGRE Working Group 01 (Lightning) of Study Committee 33 (Overvoltages and Insulation Coordination), Guide to Procedures for Estimating the Lightning Performance of Transmission Lines, Paris, Oct ISSN: ISBN: