This contribution by Rao Thallam, Jim Hunt and Albert J. Keri of Arizona-based utility, Salt River Project, is based on a paper delivered by Mr. Hunt at the 2013 INMR WORLD CONGRESS in Vancouver.
Switching Surge Factor is defined as the ratio of a power system’s maximum switching surge voltage to maximum operating crest voltage (line to neutral). It is also stated that switching surge factor is a design factor that shall determine selected clearances. NESC specifies that adequate data to support these design factors shall be available, e.g. for a 500 kV system, maximum crest operating voltage (line to neutral) = (550/sqrt (3)) * sqrt (2)=449.1 kV. If switching surge (line to ground) voltage is 808 kV, then switching surge factor is 808/449.1=1.8 pu
Table 124.1 of the latest NESC (IEEE C2-2012) gives clearances for switching surge factors ranging from 1.8 to 2.7 pu for 500 kV systems.
Switching Surge Voltages
Switching overvoltages occur because voltages at the instant of power system switching are not the same as final voltages. When a breaker is closed, there will be transition period during which voltages can overshoot by as much as 300 percent or more. Front time of the initial transient can be from a few hundred microseconds up to one to two milliseconds. EHV line energization and/or reclosing overvoltages have significant impact on line design as well as on equipment at terminal stations.
Line switching overvoltage levels along the length of a transmission line are statistical in nature and depend on factors such as instant of breaker closing on the voltage wave, relative closing times of the various poles, length of the transmission line, parallel lines in the same right of way, ground path characteristics, etc. When a transmission line is switched open, (trapped) charges remain on the line and that slowly decay to zero. Overvoltage levels are normally higher for ‘reclosing’ the line with the trapped charge compared to line ‘energization’ without the trapped charge. Line design is usually based on withstanding a 98% value.
But line energization is not the only cause of switching surge overvoltages on a transmission line. Other possible causes include:
• Switching shunt capacitor banks;
• Switching series capacitor banks;
• Switching shunt reactors;
• Fault occurrence & fault clearing;
• Other switching operations.
Study of Transmission Line Energization Overvoltages
Transmission line energization overvoltages cannot easily be calculated by analytical equations. Before the advent of digital computers, these were calculated by Transient Network Analyzers. In a TNA study, the power system is physically modeled on a smaller scale. Transmission lines are modeled as a distributed inductors and capacitors while generators are modeled as equivalent sources. The breaker, usually modeled by a computer-controlled switch, is repeatedly closed hundreds of times and the results are statistically analyzed to compute maximum overvoltages.
Description of EMTP Studies
TNA studies are now obsolete, replaced by Electromagnetic Transient Type Programs such as ATP, EMTP and PSCAD/EMTDC. Fig. 1 shows an overall view of Salt River Project’s (SRP) and adjacent 500 kV transmission systems. In order to calculate switching surge factor for SRP transmission lines, the system bounded by the substations represented in red were modeled using the PSCAD/EMTDC program. Fig. 2-1 shows a more detailed representation of the boundary substations along with the lines of interest. The system equivalents at the boundary locations are calculated using short circuit database. The calculated system equivalent parameters include positive, negative and zero sequence impedances at each boundary bus and between all boundary buses.
The lines of interest where switching surge factors are calculated, are represented with the frequency dependent (Phase) model. This line model is a distributed RLC traveling wave model that incorporates the frequency dependence of all parameters. Lines that are not the subject of detailed analysis are modeled as coupled Pi-Line sections.
Line Arrester Model
396 kV or 420 kV rated line arresters are utilized at various locations along the line in order to limit the voltages within acceptable limits. Preference has been given towards the 420 kV rated line arresters.
Series Capacitors at Coronado – Silver King Line Terminals
Series capacitor installations at the Coronado-Silver King line terminals are modeled in detail. The gaps across the series capacitors are modeled to short out the series capacitors for specified MOV current and energy settings (see Fig. 2-2).
Line clearances for all 500 kV lines were measured using Light Detection and Ranging (LiDAR) technology. An SRP contractor provided this service, in which an aircraft mounted remote sensing module was used along the power line corridor. Relative distances were measured by illuminating the objects in the corridor with a laser and analyzing reflected light. A 3-D point cloud was then interpreted to create a model of the transmission line, topology of the terrain, vegetation and structures in the right-of-way. Once the points were classified and a model created, transmission design software could be used to evaluate various scenarios, such as: conductors carrying rated current and heated to maximum operating temperature; or conductors displaced laterally by wind of a specified force per square foot. The transmission design program then produced a summary of any clearance concerns from energized wires to ground, to power lines that cross the corridor, buildings, signs, poles, bridges, etc.
The NESC requires that adjustments be made for local conditions such as elevation above mean sea level. It also allows for reduced clearance to ground in a ‘pedestrian only’ area that is not traversed by large vehicles or horseback riders. For higher line voltages (generally 230 kV AC and above), if an engineering study has determined that switching surge factors are lower than default values, certain clearance requirements can be reduced.
Switching surge overvoltages were computed for all 500 kV lines shown in Fig. 2. Detailed results are presented for the 500 kV Coronado-Silver King line that runs 180 miles and transmits power from Coronado (CO) generating station to Silver King (SI) receiving station. Springerville generating station is connected to Coronado by a 50-mile tie line. Hence, the CO-SI line plays an important role in transmitting generated power to the metro Phoenix area.
Results of switching (reclosing) surge overvoltages for the CO-SI line are presented in Figs. 3 and 4. The reclosing operations from Coronado are in 60 cycles and from Silver King are 35 cycles from line opening times. As explained, switching surge overvoltages are statistical in nature and the values shown are 2% values. When a line is switched 100 times, voltages exceed the stated values in 2% of the cases. In these figures, switching surge factors are computed on 550 kV (1 per unit) base value. Original study values were converted from 525 kV base to compare with NESC numbers.
In Figs. 3 and 4, target values are based on measured line clearances and the switching surge factor based on NESC specifications. The CO-SI 500 kV line has arresters installed at Coronado and Silver King stations to limit switching overvoltages at the line ends. But with arresters at line ends only, switching overvoltages exceed target values at several locations along the line. The solution adopted was to install 420 kV rated transmission line arresters (TLAs) at three locations, i.e. at 60, 108 and 137 miles from Coronado and a study showed that this would limit the overvoltages and meet NESC criteria.
Some iterative fine-tuning was required to select TLA locations that met the criteria. For example (see Fig. 4), in Case 9c the SSF still exceeded the limit at mile 41 along the line. The adjustment in Case 9d improved the result at this point but resulted in a small excess SSF at mile 66. In Case 9e the targets were satisfied at all critical locations.
In another sensitivity case, it was shown that all of the Coronado-Silver King criteria would be met with the same TLA locations as Case 9c if the TLAs were rated 396 kV rather than 420 kV. Installing the lower protective level arresters would have reduced the SSF at mile 41 by 0.10 pu. However, as indicated, SRP prefers the 420 kV rating because the arresters are less likely to be hit by voltage surges that will cause the metal oxide varistor blocks to conduct and this results in a less demanding duty cycle.
Results of a similar SSF study for another SRP transmission line (Coronado-Sugarloaf) are shown in Fig. 5. For this line, the switchings (energizations) are conducted long after opening, without trapped charges. In the pre-TLA case, the calculated SSF at Coronado was pessimistic based on the assumption that an existing line-end reactor, which is protected by arresters, might be switched off. It was found that two TLAs were sufficient to meet the target SSF levels for this line, located 9 and 25 miles from Coronado
Selection of Transmission Line Arresters
Until recently, the preferred method of limiting switching overvoltages on transmission lines included:
1. Installing breakers with pre-insertion resistors;
2. Synchronous switching of breaker poles.
Breaker switching modification, as above, is considered to reduce breaker reliability. Switching surge voltage reduction is along the entire line and not at any specific location.
Arresters can be installed on transmission lines to reduce lightning or switching overvoltages on lines. There are two types of Transmission Line Arresters: Externally-Gapped Line Arresters (EGLAs) and Non-Gapped Line Arresters (NGLAs) and their relative Pros and Cons are discussed below:
Externally-Gapped Line Arresters
No disconnectors are needed;
Arrester is not stressed with continuous voltage and hence there is freedom to design the arrester voltage.
Insulation coordination is not precise and difficult because gap sparkover is not precise;
No control of switching overvoltages;
Arrester has to be individually designed for each situation, i.e. tower, conductor location;
Difficult to identify failed arresters.
Non-Gapped Line Arresters
Precise insulation coordination;
Flexible design for installation – tower, conductor etc.;
Easier to identify failed arrester.
Disconnector needed. Disconnectors for high voltage arresters are complex;
Arrester stressed continuously and hence higher rated voltage needed than EGLA.
In the end, Salt River Project selected gapless metal oxide arresters housed in polymeric housings. These were installed on selected 500 kV lines at strategic locations based on extensive switching overvoltage studies. Since line insulation after failure is self-restoring, the arresters were provided with disconnectors. This facilitated the line being energized immediately in the event of arrester failure, without replacing it. Operation of the disconnector makes it easier to identify any failed arrester.
Line energization for switching surge calculation showed energy discharged in the line arrester was not significant for a standard station class arrester. However, line energization is only one of many situations that could stress the arrester and these were conservatively specified with the same energy rating as a typical station class arrester. SRP specifies 420 kV duty cycle rating arresters for 500 kV systems and has had good operating experience. Considering the high reliability required for line arresters, they are specified with 420 kV duty cycle rating.
Highlights of SRP’s TLA specification are as follows:
Maximum Continuous Operating Voltage (MCOV) kVrms 335
Duty Cycle Rated Voltage, kVrms 420
Arrester Energy Rating
Energy Handling Capability*: 13.0 kJ/kV of MCOV
(*as described in IEEE C62.11, Draft 9, Nov. 2011, Energy class G)
Vendor shall provide type test data as proof that the arrester meets the energy handling capability specified above.
Figs 6, 7 & 8 depict the installation methodology and the finished product. A crane with an insulated boom, known as a Condor, was used due to the height of the wires. Attachment hardware was simple and consisted of readily available standard items. SRP elected to use armor rods rather than clamping directly to the aluminum outer strands of the conductor. The device lent itself to ‘hot’ installation and performing the work with the line energized helped to greatly simplify the process since 500 kV line outages are challenging to arrange. At one location, arrester condition monitors were installed, mounted on the lattice tower in series with the ground lead. Files can be downloaded via a wireless USB communications device and this will provide intelligence regarding arrester functionality as well as number of surges experienced by each arrester.
Switching surge overvoltages can be reduced by installing transmission line arresters (TLAs) along a line. Gapless metal oxide arresters in polymeric housings are preferable over externally-gapped line arresters (EGLA) for this application. They provide more precise insulation coordination and flexible installation. Switching overvoltage studies, using EMTP type programs, provided the information needed to select line locations that provide required switching surge reduction where line clearances are critical. Standard station class arrester energy rating specification is recommended for TLAs.