Surge arrester use improves transmission system performance in Malaysia's severe lightning environment.
Malaysia has a Very High Isokeraunic Level of Up to Thunderstorm Days Per Year. This high incidence of lightning activity causes multiple line trippings on extra-high-voltage (EHV) transmission lines managed and operated by the transmission division of Tenaga Nasional Berhad (TNB; Kuala Lumpur, Malaysia). These line trippings are responsible for about half of the utility's total number of system outages each year.
Since 1995, the TNB transmission division has installed gapped-type transmission line arresters (TLAs) to comply with the Malaysian grid code on system security standards. The TLAs have proven to be successful based on the reduction of lightning-caused outages since their installation.
However, gapped-type TLAs are costly. The high cost led the TNB transmission division, together with TNB Research, to conduct a pilot study on the use of gapless TLAs as an alternative to gapped-type TLAs. Three manufacturers participated in the pilot project, and 148 units of gapless TLAs were installed on a 12-km (7.5-mile) section of the 132-kV line routed between the Balakong substation and the Serdang substation.
With the arresters monitored for a year, this study has given TNB engineers valuable knowledge and experience on the future selection of lightning protection equipment for the transmission system.
In recent years, TNB has implemented various engineering solutions, such as enhancing its standard earthing design and installing overhead shield wires, to improve line lightning performance. Even though enhancements such as shield wires reduce the probability of direct lightning strikes to phased conductors, and measures such as reducing the tower footing resistance (TFR) lower the impact on the system, these modifications have proven to be incapable of completely protecting the transmission lines. The risk of a back-flashover is increased with the higher values of TFR encountered in rocky soil, whereas in softer soils (paddy fields, for example), the TFR value is much lower.
Therefore, TNB still regards the installation of TLAs as the optimum solution to its major lightning problem.
THE PILOT PROJECT
TNB launched a pilot project in May 2005 to study the performance of gapless TLAs as an alternative to the gap-type TLAs installed on the utility's transmission system. Three arrester manufacturers agreed to participate in the pilot project, and the evaluation was based on their equipment (referred to here as types A, B and C). Installation was arranged in stages for the three types of arresters, which were monitored for a full year from the installation date. A total of 148 arresters were installed from April 2005 to July 2005 (Fig. 2.)
The 132-kV Balakong-Serdang transmission line was selected for the project based on its high tripping records due to lightning and the geographical topography of the line (nearly 60% of the line route is through hilly terrain). The 12-km transmission line comprises 37 quadruple circuit towers with a 275-kV double circuit above a 132-kV double-circuit line.
A lightning simulation software package was used to determine the expected lightning performance of the lines. The package included a 3-D electrogeometric model where lightning strokes are generated and applied by ground wires or phase conductors in predivided sections.
The line performance without TLAs was simulated by applying 2000 lightning strokes. The result indicated the back-flashover rate was 2.96 flashovers/100 km/annum (4.76 flashovers/100 miles/annum) and the shielding-failure flashover rate was 0.79 flashovers/100 km/annum (1.27 flashovers/100 miles/annum). The total number of flashes and the flashover rate of the line without arresters were 33 flashes and 3.75 flashovers/100 km/annum (6.03 flashovers/100 miles/annum). However, the flashover rate reflects the lightning performance of the line as a whole (that is, the quadruple tower system with 275-kV and 132-kV circuits.).
The TLA data was inserted into a similar model, and the lightning performance was simulated with the TLAs installed in two different positions (refer to Fig. 1 for circuit identification and arrester-installation positions):
Three arresters on phase-circuit numbers 7, 8 and 9
Four arresters on phase-circuit numbers 7, 8, 9 and 12 (referred to as the L-shape configuration).
The results of the simulation studies confirmed that the overall lightning performance of the line showed a noticeable improvement with the TLAs installed.
Based on the results of the simulation studies, four arresters were installed on each tower using the L-shape configuration (Fig. 3), for example, TLA 1 to TLA 4 as shown in Fig. 1. The three types of arresters (A, B and C) were installed in three different sections of the 12-km transmission line. Field tests continued for more than a year, and the results of the monitoring and observation program on the three different gapless TLAs were presented to the TNB Technical Committee whilst the field tests and monitoring continued.
A lightning density map was generated using the Lightning Detection System available from TNB Research. From 2004 to 2007, the monitored lightning activity was judged to be fairly heavy with a recorded stroke density of 70 to 80 lightning flashes/sq km/annum (181 to 207 lightning flashes/sq mile/annum) near the Balakong substation. This means the area had generally recorded up to 23 to 27 lightning flashes/sq km/annum (60 to 69 lightning flashes/sq mile/annum).
(Editor's note: The most lightning-intense areas of U.S. are parts of Florida and Louisiana. The flash density in those areas is about one-fourth that of the Balakong substation area.)
Each arrester was equipped with a surge counter/sensor that measured the TLA surge and leakage currents. The counter readings were compared with the lightning map record to confirm its functionality, the surge date and time readings, and to assure the sensors had functioned correctly.
Using the surge counter records, three arresters were selected from each manufacture and returned to their respective manufacturer laboratory for repeated routine testing (voltage test, lightning impulse residual voltage test and partial-discharge test at a 1.05 maximum continuous operating voltage rating). All results were found acceptable except for arrester type A, where two of the three reference voltage values were higher than expected. Further examination will be imposed for this arrester type.
The tripping record for the 132-kV Balakong-Serdang transmission line (Fig. 4) shows that after the gapless TLA installation, the number of trippings due to lightning was reduced from six to two trippings for essentially the same period of time. The two trippings that occurred following the installation of the gapless TLAs were caused by a direct strike to the earth wire and, therefore, were considered unavoidable.
With regard to the circuit tripping record between January 1997 and April 2005, the number of trippings was reduced from eight trippings in the first three years to seven trippings in the next five years, January 2000 to April 2005. Following the installation of the gapless TLAs on all towers, there were no trippings recorded on the line except one incidence when a direct lightning strike caused the earth wires to snap and rest on the live conductors. Furthermore, on the 275-kV Pudu Ulu-Serdang line, the top section of the quadruple circuit was monitored to show the tripping rate was reduced significantly.
Throughout the duration of the project, all problems and failures that occurred on the transmission line and equipment were subjected to failure analysis. The main problems were associated with jumper failures, surge counter faults and insufficient clearance between a TLA and a smaller tower. The latter fault cause has prompted TNB to reconsider the installation of gapless TLAs on small towers.
Fig. 5 shows the disconnected line jumper found following a mid-span direct lightning strike that caused the earth wire to snap and rest on the 275-kV and 132-kV conductors. Subsequent investigation confirmed that all arresters were in good condition and the failure was caused by a weak connection of the line jumper.
The bolted-tee clamp used for the connection between the TLA line jumper (disconnecting device side) and the phase conductor created a weak mechanical point on the braided line jumper due to long-term vibration and fatigue (Fig. 6). Hence, the bolted-tee clamp connection is no longer used; it was replaced with a corona-free live-line clamp, which has a stronger connection.
Following the failure of more line jumpers, the line-wire jumper (braided type) was redesigned but still found to be too weak to withstand the inevitable vibration and fatigue. Flexible stranded copper wire is now the accepted standard for this connection.
In September 2007, it was noted that the majority of the surge counters were not giving any readings and data could not be retrieved by the sensor. Investigation showed there was evidence of moisture ingress, requiring a modification of weatherproofing the surge counter's main body and the handheld antenna.
This pilot project has provided invaluable information and benefits to TNB and, in part, the electricity supply industry. The benefits of the study are as follows:
Improved lightning performance and operation reliability of the 132-kV Balakong-Serdang line
Enhanced technical knowledge and experience on the use of gapless TLAs to mitigate lightning problems
Enhanced technical knowledge on international standards (for example, IEC 60099-4)
Product performance data on which TNB can base future specifications for equipment to mitigate transmission system outages attributable to lightning
Cost benefits, as the cost of a gapless TLA is less than that of a gapped-type TLA.
The project confirmed that the performance of transmission lines can be improved by the installation of gapless TLAs. However, the type A arrester used in the project is subject to further investigation, since it failed to confirm its suitability for circuits subject to Malaysia's weather conditions. There is considerable merit in using appropriate software packages in conjunction with lightning density data to determine the most suitable tower location for the gapless TLA to achieve optimum performance and protection from circuit outages due to lightning.
Iryani M. Rawi (email@example.com) earned a bachelor's degree in electrical engineering from University Technology of Malaysia in 2002. She is currently a senior engineer in the engineering department of the TNB transmission division, Tenaga Nasional Berhad. Her works covers the design, study and application of the overhead line transmission system. Rawi is a member of the Joint Industry Study on Lightning Protection System.
Mohd Pauzi Bin Yahaya (firstname.lastname@example.org) joined Tenaga Nasional Berhad in 1989 and is currently the senior manager of transmission at TNB Research. He was a member of the National Technical Committee for the Development of the Code of Practice on the Resistibility of Telecommunication Equipment to Overvoltages and Overcurrents. He was also a member of the Steering Committee for the review of MS 939: Protection of Structures Against Lightning. Currently, he is a member of the Joint Industry Study on Lightning Protection System.