Insulator sets are among the most common components in transmission networks and their quality and performance play a vital role in ensuring a reliable electricity supply. A key part of that is their behaviour when exposed to a power arc. Poor design with respect to power arc behaviour can significantly reduce the lifetime of the insulator set and increase the likelihood of supply interruptions.
There are a wide variety of insulating material options and environments in which insulator sets are used. As such, insulator set designs vary considerably and may even need to be tailored for specific applications. Even the smallest differences in design can affect the power arc behaviour of a string. Physical testing is the most reliable way to assess a string’s power arc behavior and therefore utilities increasingly demand that insulator sets undergo power arc testing before they commit to installing them on their networks.
Over the last several years, KEMA Laboratories, Prague carried out over 160 power arc tests on insulator sets of various types. These tests were performed according to IEC 61467:2008. In addition to revealing the most common failure modes for different types of insulator sets, this overview contributed by Robert Jech highlights a potential issue with how IEC 61467:2008 is currently defined. This is an issue that could impact the value utilities derive from today’s tests. The insight gained from this will provide insulator set manufacturers with useful input into how best to design strings to increase the chances of first-time success in testing and thereby reduce the time and money spent on redesigns. It may also stimulate discussion around the specification of the standard to ensure it meets the needs of manufacturers and utilities equally.
Description of Power Arc Tests
Power arc testing of insulator sets at KEMA is performed according to the IEC 61467:2008 standard. This standard applies to insulator strings or sets comprised of ceramic, glass or composite materials that will be mounted on metallic poles or towers and used in AC overhead lines with a nominal voltage above 1000 V.
The standard defines methods, parameters, circuits etc. for power arc tests on both insulator sets and short strings. Different test arrangements are allowed within the standard and choice depends on the final application of the insulator sets, according to customer requirements. The test circuit and series should be chosen based on factors such as geometry and type of insulator, its position on the line and type of tower.
A set’s intended position on a line determines whether it should be tested with a balanced or unbalanced supply circuit and, depending on line parameters, short-circuit current to be used during testing. If the insulator set will be positioned in the first or last 5% of a line, it should be tested with an unbalanced supply circuit while sets to be located between the 5% and 95% points of the line require a balanced supply circuit for testing. Similarly, sets for the ‘middle’ section (i.e. from 24% to 76% of the line length) require only 20% of the rated short-circuit current of the network. Sets for use at locations between the 5% and 24% points – or 76% and 95% points –require 50% and sets in the first or last 5% require the full short-circuit current of the network.
Meanwhile, type of tower governs choice of a balanced or unbalanced return circuit during testing. Sets to be used in the center phase window of a tower require testing with a balanced return circuit (referred to as test series X). For outer positions on the tower or where there is no center phase window, an unbalanced return circuit is used (referred to as test series Y). Furthermore, the number of tests to be performed depends on whether the customer intends the same type of set to be used throughout the line (in which case the complete test series X or Y is performed) or in just along part of the line. Fig. 1 shows a typical test arrangement for a V-string set with composite insulator This test arrangement comprises balanced supply and return circuits. The arrangement pictured simulates an insulator set position between the 5% and 95% points of the line in the center phase window of the tower.
Depending on type of insulator set, IEC 61467:2008 standard may also require that it undergo verification tests in addition to the power arc test, the main being the mechanical failing load test (MFLT). This is performed on the insulator units to ensure they can withstand mechanical forces after a power arc is applied. Insulator sets may also be required to undergo a dry power frequency flashover (DPFF) test to ensure the insulator does not suffer punctures at voltages below the flashover voltage. Additional electrical tests may be performed on the fittings and conductors within the insulator set to verify withstand capabilities.
Statistical Overview of Tests
Analysis of results of tests carried out on 162 insulator sets over a five-year period is shown below:
Types of Insulator Sets & Components
Insulator sets come in a number of different design types and insulating materials, depending on the intended application. The insulator sets tested during the five-year period comprised:
• 92 suspension sets
• 49 tension sets
• 17 “V” strings
• 4 cross-arm insulators
In terms of the insulating material used and insulator design, insulator sets tested included:
• 81 composite
• 34 glass, cap & pin
• 12 glass, cap & pin (short string)
• 27 porcelain, long rod
• 8 porcelain, cap & pin
As shown in Fig. 2, 66% of components tested successfully completed a full test sequence. Only one set suffered a separation failure during the arc test. DPFF tests were not performed on an additional 35 sets and no MFLT was performed on an additional 10 sets. These 45 sets are not included among those sets that successfully completed testing even if no failure occurred during those tests.
IEC 61467:2008 specified an MFLT for all 162 insulator sets tested. One set failed before the MFLT could be performed and the results for the remaining 161 sets are shown in Fig. 3. Of these, MFLTs were not required by the customer for 10 sets. Consequently, 151 MFLTs were performed during the period covered, with 145 sets passing and 6 failing.
Similarly, the standard specified DPFF tests for 54 of the sets, and the results are summarized in Fig. 4. DPFF tests were not required by the customer for 36 sets and therefore 18 DPFF tests were carried out with no failures.
Fig. 5 summarizes results for the sets required to undergo mechanical tests on the fittings and conductors. The 20 sets not included in this chart comprise: 1 that failed during power arc testing, 6 that failed during the MLFT and 13 where the standard did not require such tests. Of the remaining 142 sets, the tests were not required by the customer for 106 sets. In total, 36 successful tests were carried out and 4 sets failed.
The figures above give an indication of the quality of insulator sets available today and the sustained need for testing before installation. However, perhaps more illuminating is the insight into the principal failure modes for insulator sets. This insight highlights areas that manufacturers should pay particular attention to when designing new insulator sets.
Long-rod Insulator Sets
For long-rod sets, three particular potential issues were identified. The metallic parts of the insulator unit can be prone to melt if not well designed. Furthermore, poor design can lead to the sheds near these metallic parts breaking. Finally, the arc can cause metal to evaporate from the set’s protective fitting, causing ‘puddling’.
Composite Insulator Sets
The silicone rubber used in composite insulators exhibits high resistance to power arc testing. However, the critical point for these sets is the where the fiberglass core connects with the metal end fittings.
Cap & Pin Insulator Sets (Glass or Porcelain)
Cap & pin insulators exhibit high mechanical resistance after power arc tests. The main areas of concern for these sets are the possibility of sheds breaking and, in the case of glass sets, cooling of the cap & pin units.
Protective Fittings: Load-Bearing Fitting Protection & Arc Direction Fittings
Power arc tests put much greater mechanical and thermal strain on these fittings than short-circuit tests. This must be taken into account when designing the protective fittings of insulator sets of all kinds.
There is also risk of material from these fittings melting onto the insulator units. Care should also be taken to ensure that protective fittings do not move during arcing, which can lead to the arc root sitting on the load-bearing fittings.
Protective Fittings: Corona Rings, Grading Rings
These fittings are not primarily designated for arc currents but influence the position of the arc root and how it moves. As with other protective fittings, risk factors here include material melting onto the insulator units and movement of the fittings causing the arc root to move away from its intended position. In addition, changes to the contours of these fittings can lead to excessive corona discharge and radio noise.
Power Arc Behaviour During Tests
Carrying out the tests reviewed above provide an opportunity to closely study the behavior of power arcs during testing. Two distinct basic behaviors were observed. In the first case, the power arc runs between the protective fittings designed to define its path (i.e. the arcing horns, arcing rings etc.). This is the intended behavior of the power arc and means that most of the stress from the arc is exerted on the protective fittings and insulator sets with only limited impact on the insulator material itself. Crucially, this kind of behavior has little impact on the tower and conductors.
However, a second behavior can occur where the arc travels along the line conductor or the tower. In this case, the impact on the insulator set is much less with much lower stresses on the protective fittings and insulator units. However, there is a much greater negative impact on the tower or conductor that is carrying the arc.
Currently, IEC 61467:2008 does not specify the behavior of the arc during testing. It is also not typically specified in the tender documents for grid installation or upgrade projects. This could present a real issue for utilities and manufacturers of insulator sets. If the design of the insulator set leads to the arc behaving in the second of the two ways described above, it is likely that the insulator set will pass the power arc test as the arc is not close to the insulator and therefore cannot damage it. However, if such a set is used in the field, any arcs experienced would travel along the tower or line conductor causing damage. Moreover, damage to these components would be more expensive to repair and lead to longer outages.
Statistical study of results of power arc tests allowed the critical points of various types of insulator sets to be identified. This offers valuable insight for manufacturers when designing new insulator sets.
In addition, the experience gained from these tests highlighted a potential issue with the IEC 61467:2008 standard as it is specified today. As the standard does not specify the behavior of the arc during testing, it could allow potentially dangerous insulator sets to pass testing. This could give utilities a false sense of security in the insulator sets they are purchasing. To ensure the standard meets the needs of utilities, insulator material providers and insulator fittings manufacturers, this issue should be addressed. To do this, arc behavior during testing should be specified either through an update to IEC 61467:2008 or routinely in tender documents from utilities. The latter course would require all utilities to become more aware of this issue.