B2-301 IMPROVING DOUBLE CIRCUIT TRANSMISSION LINE RELIABILITY THROUGH LIGHTNING DESIGN

Transcription

1 21, rue d'artois, F-7008 Paris B2-301 Session 200 CIGRÉ IMPROVING DOUBLE CIRCUIT TRANSMISSION LINE RELIABILITY THROUGH LIGHTNING DESIGN J. A. (TONY) GILLESPIE & GLENN STAPLETON Powerlink Queensland Australia KEYWORDS - Lightning, Earthing, Autoreclose, Transmission Line, Reliability 1. INTRODUCTION During the last twenty years, double circuit transmission lines in Queensland have been constructed, instead of a number of single circuit transmission lines. Double circuit transmission lines have the advantages of lower capital cost than constructing single circuits and better utilisation of easements. Experience on the Queensland transmission network has been that lightning is the major cause of forced outages on the 27 kv system. Since the use of double circuit construction reduces network security and reliability, the incidence of lightning induced outages is an important consideration in the design and operation of Queensland s network - especially double circuit outages on its high capacity, long distance transmission lines. This paper discusses the modelling of lightning outages on double circuit transmission lines and compares the results obtained from these models with the recorded performance of several major transmission lines in the Queensland network. The probability of occurrence of double circuit outages is identified, and the paper proposes guidelines for improving the lightning performance of transmission lines. 2. TRANSMISSION LINE PERFORMANCE A review of 27 kv transmission line performance has been performed using the Queensland transmission company s Forced Outage Database (FOD). Table I sets out the causes of forced line outages on the 27 kv network, comprising both double and single circuit lines. A total of 18 faults were considered in this study. Table II identifies the types of faults that have occurred on the 27 kv network. Based on the length of lines installed, the average lightning trip out rates for the network are 0.3 outages per 100 route km/yr for 27 kv circuits and 1.3 outages per 100 route km/yr for 132/110 kv circuits. Other useful data from the analysis showed that : % of faults were transient and autoreclosed successfully. The remainder were permanent faults % of faults were single circuit and.7% of faults were double circuit. Tgillespie@powerlink.com.au

2 1. TRANSMISSION LINE LIGHTNING DESIGN Power stations in Queensland are dictated by the location of coal deposits. The majority of generation in Queensland is located both in the Central region and in the western part of the Southern region. There are major load centres located 700 km 1000 km to the north and 700 km to the southeast of the state around the state capital, Brisbane. As a result, long interconnectors are required for the transmission of power to supply these distant load centres. Table I Causes of Forced Outages Table II Fault Types Leading on the 27 kv Network to 27 kv Forced Outages Fault Cause Faults (%) Fault Type Faults (%) Lightning/Storm 0. One Phase to Earth 7.7 Bush/ Cane fire 16.2 Two Phases to Earth 9. Polluted Insulation / Vegetation 16.2 Unknown 8.1 Unknown Cause 1. Phase to Phase.7 Line Equipment Failure 7. 3 Phase to Earth 2 Malicious rocks / projectiles 2 Natural Hazard animal contact 2 As shown earlier in Table I, lightning forms the predominant cause of outages on the 27 kv network. In the regions where high capacity transmission lines are installed in Queensland, published isoceraunic levels range from 30 to 0 thunder days per year, which equates to 1. to 2 lightning ground flashes/km 2 /yr. Sydney and Melbourne have ground flash densities around 1 and 0. flashes/km 2 /yr, respectively. Darwin has the highest ground flash density in Australia at approximately flashes/km 2 /yr. Typically, the design of a 27 kv transmission line specifies that the total lightning outage should not exceed 0.3 outages/100 km/year. Early generation 27 kv lines in Queensland relied on 1 normal disc (diameter 2 mm x 16 mm coupling distance) insulation, the minimum based on switching surge. However, recent 27 kv transmission lines have utilised 18 standard discs, or an 18 disc equivalent composite insulator, for improved lightning performance. At 132 kv, 8 discs are commonly used for acceptable lightning performance of transmission lines. In addition to increased insulation, sufficient additional earthing has been installed to improve the backflashover performance of lines. Typically, the line is tested to ensure that the majority of individual tower footing resistances do not exceed 10 ohms. In addition to this, the required footing resistance is specified not to exceed ohms for the first 2. km from a substation. This is to reduce the probability that backflashover will occur close to a substation. Surges resulting from backflashovers that occur further than 2. km along the transmission line will be significantly attenuated before reaching the substation terminals. 2. TRANSMISSION LINE LIGHTNING MODELLING SOFTWARE This paper details investigations into the lightning performance of three transmission lines, each with different structure types as follows :- D3S0A 330 kv suspension structures installed on a 208 km transmission line. D2S0K 27 kv suspension structures installed on a 336 km transmission line. D1S0A 132 kv suspension structures installed on a 100 km transmission line. All these structures are conventional double circuit lattice steel towers with twin overhead earthwires. Although not a high capacity interconnector, the third (100 km) transmission line 2

3 using the D1S0A 132 kv structure has been included in this discussion since it provides a unique case study for a high lightning area in North Queensland. Some 62% of the outages on this line were double circuit in nature, resulting in significant energy interruptions to northern load centres. Such a high proportion of double circuit outages warranted extensive investigation. Subsequent remedial work has reduced the lightning outages significantly, virtually eliminating double circuit outages. Three separate tools were used in the analysis of transmission line lightning performance :- 1. FLASH, written by Power Technology Incorporated (PTI), with source code modified to also include CIGRE stroke current probabilities and the Improved Electro-Geometric model. The software version utilised for this study was the early 1990 s MS-DOS Version. 2. TFLASH Version, produced by the Electric Power Research Institute (EPRI), A Mathcad program written by the authors. Results from these packages estimates the long term average lightning performance for the transmission line. For the transmission lines studied, these assumptions were made :- Ground flash density of 2 ground strikes/km 2 /year. Standard tower footing resistance of 10 ohms. Average span of 0 metres. 3. TRANSMISSION LINE PERFORMANCE Table III shows the simulation results for each of the transmission lines in this study modelled with each software design package. This table also includes for comparison the recorded lightning outages for each of the lines and categorises them in several different ways. There is reasonable agreement between calculated and actual performance bearing in mind that the programs provide a range for the expected long term performance. TFLASH appears to be significantly overstating shielding outages which flows on to total outages. This issue is currently being discussed with EPRI. 6. DOUBLE CIRCUIT OUTAGES Queensland s transmission network forms part of the National Grid in Eastern Australia that is used to transmit electricity for the Australian National Electricity Market. The security requirements of the National Grid are defined in the National Electricity Code and the National Electricity Market Management Corporation (NEMMCO) is responsible for managing security of the interconnected power system. It is a requirement that the National Grid can withstand any single credible contingency such as the outage of any single circuit transmission line without the loss of system stability or an interruption to customer load. Double circuit outages of transmission lines are so rare that they are not considered to be a single credible contingency, unless NEMMCO is advised that such an outage has become credible, eg. during a bush fire. This means that the National Grid is usually not operated to be able to safely withstand double circuit outages of transmission lines. Hence this can have an extreme impact on supply of loads and network stability. Major blackouts are likely to occur unless NEMMCO has been made aware of the risk and arranged for sufficient ancillary services (i.e. spinning reserve) to be scheduled. As was seen from the previous case studies, the probability of double circuit outages increase as the tower footing resistance increases. It was seen that even with 20 ohm footing resistance on the 27 kv structure, a double circuit outage could be expected to occur once in a year period. With the increased insulation of the 330 kv structure, the probability of a double circuit outage decreased to a 1 in 32 year event with a footing 3

4 resistance of 20 ohms. However, for such a low probability event, the consequences of such an event to a transmission network can still be significant. With the above in mind, Transmission Network Service Providers (TNSP s) are faced with the responsibility of deciding when to advise NEMMCO that a double circuit outage is a credible event, and should be taken into consideration in managing system security and the dispatch of the National Electricity Market. This decision must be made with awareness of the significant additional costs to Market Participants due to increased generation costs. In Queensland, the transmission utility has assessed the risks of double circuit outages occurring due to lightning, severe wind gusts, bush fire, cane fires, structural failure and insulator pollution flashover and has developed operating policies to declare double circuit outages as credible contingencies for each source of risk. Tower Type 132 kv D1S0 A Tower Footing (ohms) Table III Comparison of Lightning Outage Rates Calculated Actual Calculated Calculated Calculated Actual Total Lightning Outage Rate (Outages per 100 km/year) Total Lightning Outage Rate (per 100 km/yr) Shielding Failure Rate (Outages per 100 km/year) Total Backflashover Rate (Outages per 100 km/year) Double Circuit Flashover (Outages per 100 km/year) MC Flash TF MC Flash TF MC Flash TF MC TF Double Circuit Flashover Outage Rate ( per 100 km/yr) kv D2S0 K 330 kv D3S0 A Definitions: MC is User written Mathcad lightning design program; Flash is PTI Flash program; TF is TFLASH by EPRI. Calculated values are from the software packages. Actual is data recorded in the Forced Outage Database. 7. LIGHTNING PERFORMANCE IMPROVEMENT FOR TRANSMISSION LINES There are several options that can be adopted to improve the lightning performance of a transmission line. The concepts of increased line insulation, additional earthing, additional earthwires and auto-reclose strategies are considered further. At this point, surge arresters[] have not been applied to transmission lines in Queensland because of the high installation cost, and questionable reliability, especially when fitted to long lines.

5 7.1 INCREASED LINE INSULATION AND ADDITIONAL EARTHING Two of the easiest and most readily adopted methods for improving the lightning backflashover performance of a transmission line are to retrofit additional discs to line insulation, and to install additional earthing to the structures. Adding more discs increases the impulse flashover voltage of the insulation, thereby improving lightning performance. To understand the effect of tower footing resistance on backflashover performance, it is important to understand the backflash mechanism itself. When a lightning stroke hits an earthwire or structure, a steep fronted surge travels in each direction along the earthwire, as well as down each adjacent structure via steelwork to earth. The earthwire travelling wave induces a voltage on the phase conductors in addition to the power frequency voltage. The wave travelling down the structure to earth causes the tower steelwork to rise in potential above earth as a function of the tower surge impedance. Upon reaching the tower earth, part of the wave is reflected back up the tower due to the interface between the tower surge impedance and tower footing resistance to remote earth. Therefore, a voltage profile appears along the tower and a different voltage exists at each crossarm to earth. For large footing resistances, a significant current reflection occurs at the interface between the tower and its remote earth causing a larger voltage to be produced at the crossarms. When the voltage difference between the crossarm and conductor exceeds the impulse flashover level of the insulator, a backflash is said to occur when an arc originates from crossarm to conductor. Figures 1 to 3 show the way in which total outage rate varies with different levels of line insulation, and different tower footing resistances for D1S0A, D2S0K and D3S0A towers. There is a general trend for lightning performance to improve as transmission voltages rise from 132 kv to 27 kv to 330 kv. This is because the increased impulse flashover voltage for higher levels of insulation reduces the probability of backflashover. It can be seen from each of the curves that outage rates can be improved by adding two additional discs, and reducing the tower footing resistance. However, there is a practical limit on how many additional discs can be added to an existing structure, generally dictated by the need to maintain electrical clearances to the structure. D1S0A 132 kv Transmission Structure 3 30 Total Flashovers (outages/100km/annum) Tflash using 8 Discs Tflash using 10 Discs PTI Flash using 8 Discs PTI Flash using 10 Discs Tower Footing Resistance (ohms) Figure kv Outages as a Function of Footing Resistance and Line Insulation

6 D2S0K 27 kv Transmission Structure 2 Total Flashovers (outages/100km/annum) Tflash using 16 Discs Tflash using 18 Discs PTI Flash using 16 Discs PTI Flash using 18 Discs Tower Footing Resistance (ohms) Figure 2 27 kv Outages as a Function of Footing Resistance and Insulation Level Theoretically, best backflash performance would be for a footing resistance closest to 0 ohms. In practice, it is technically difficult and financially prohibitive to achieve a footing resistance of 0 ohms. As introduced in Section 3, earthing is installed on Queensland transmission lines to achieve a maximum of ohm footing resistance on structures within 2. km of substations, and a maximum of 10 ohms for the remainder of the transmission line. D3S0A 330 kv Transmission Structure 2 Total Flashovers (outages/100km/annum) Tflash using 18 Discs Tflash using 20 Discs PTI Flash using 18 Discs PTI Flash using 20 Discs Tower Footing Resistance (ohms) Figure kv Outages as a Function of Footing Resistance and Insulation Level 6

7 7.2 ADDITIONAL EARTHWIRES Each of the towers in this study has twin earthwires, and can be assumed to be perfectly shielded. However, the mechanism for backflashover is such that for a tower with a vertical conductor configuration, the farthest conductor from the earthwire has the smallest coupling with this earthwire. Due to this reduced coupling, the voltage across the insulation on the lowest phase has the largest voltage difference between the crossarm and conductor and is therefore the most likely to suffer backflash. By installing additional earthwires, coupling can be improved to the lower phases on the structure, thus reducing the backflashover rates of these phases. This paper considers the simple case of retrofitting an additional earthwire strung beneath the bottom most phase of the structure, along the centreline of the tower. This is an ideal position for stringing an additional earthwire, as it imposes the least increase in structural duty, and it is more likely that such an earthwire could be retrofitted to the structure without the need for a programmed double circuit outage. However, difficulties may arise in achieving statutory ground clearance with an under strung earthwire. The author s Mathcad program has been applied to analyse the effect of such an additional earthwire mounted along the centreline of the tower, and positioned below the bottom phase. The tower chosen for this comparison was the 132 kv D1S0A, since it was seen in Figure 1, earlier, to have the highest backflashover and double circuit rate due to its lower insulation level. Figure shows the Mathcad simulation results for the 132 kv D1S0A with 8 disc insulators, with and without an additional third earthwire. Effect of Additional Earthwires 30 2 Backflashover Rate (Outages/100 km/year) Discs & 2 E/W 8 Discs & 3 E/W Footing Resistance (Ohms) Figure Backflashover Improvement of 132 kv Structure using a Third Earthwire 7.3 AUTORECLOSE STRATEGIES The purpose of autoreclosing in network operation is to minimise transmission line outage time and maintain supply. It is an important tool to increase reliability of the transmission system and improve stability. The National Electricity Code requires autoreclose of transmission lines forming part of the National Grid, unless approved otherwise by NEMMCO. Details are not specified by NEMMCO. 7

8 For single pole autoreclose, a longer dead time is required than for three pole autoreclose because the current in the healthy phases will maintain the arc longer. Suggested minimum times are 00 msec for three pole and 1 second for single pole. Dead times should be less than 1 seconds for safety reasons in order to minimise the risk that a person could come into contact with the apparatus during this dead time. Queensland has adopted a dead time of 10 seconds for three pole autoreclose which allows oscillations to die down prior to reclosing. For single pole autoreclose, high speed autoreclose with a dead time of 1.2 seconds is adopted in order to minimise outage time thus improving system security. Such a small dead time is also beneficial in returning the feeder to service before subsequent lightning outages may occur. 8. CONCLUSIONS Operational experience of the Queensland transmission network indicates that lightning is the primary outage mechanism for its transmission lines. Due to the length of high capacity transmission lines in Queensland, modelling of lightning performance is an important tool in the design of these transmission lines. Three software packages have been applied in this study to review the lightning performance of 132 kv, 27 kv and 330 kv geometries utilised on transmission lines in the Queensland network. Taking the results from each package into consideration provided a useful range of simulated long term average data, which closely matched actual recorded performance data. Simulations also indicated that the probability of double circuit outage decreased with increased voltage level, due to increased line insulation. For a 20 ohm footing resistance, (which may represent a tower earthing situation that has since dried out from its original measured value of 10 ohms at construction), simulated double circuit outage rates range from 1 every months for the 132 kv radial transmission line, 1 every years for 27 kv structures, to 1 every 32 years for 330 kv structures. These quantities provide useful statistics to system planners. Although the probability of a double circuit outage is low for 27 kv and 330 kv high capacity transmission lines, the consequences - should such an event occur - can be severe. TNSP s must advise NEMMCO when they consider that a double circuit outage is a credible event so that NEMMCO can take appropriate action. Queensland has adopted increased line insulation and additional earthing in its network to improve lightning performance. The retrofitting of an additional under strung earthwire has been seriously evaluated for one transmission line in the Queensland network, but has not yet proceeded (early 2003). Surge arresters have not been implemented on transmission lines in Queensland to date, due to perceived doubts about arrester reliability, and the high installation and maintenance costs involved. 10. REFERENCES 1 TFLASH User Guide, Electric Power Research Institute, IEEE. Guide for Improving the Lightning Performance of Transmission Lines, IEEE P 123, IEEE Standards Department, Piscataway, NJ, CIGRÉ. Guide to Procedures for Estimating the Lightning Performance of Transmission Lines, WG 01 (Lightning) of Study Committee 33 (Overvoltages and Insulation Coordination), October Transmission Line Reference Book - 3 kv and Above (Chapter 12), J.G. Anderson, Electric Power Research Institute, Palo Alto, California, Second Edition, Guide for Application of Transmission Line Surge Arresters kv, Electric Power Research Institute, Palo Alto, California,