Most failures of composite insulators originate from within the various interfaces. These interfaces could either be between different components (e.g. housing and hardware) which are macroscopic in nature and often visible to the unaided eye, or microscopic, i.e. within a particular component, such as internal to the rod or housing. Given the critical role played by interfaces in the long-term performance of composite insulators, it makes sense to study this subject in greater detail. At the same time, both users and manufacturers must know how effective existing standards are in screening out insulators which are likely to fail in service due to problems with design or manufacturing of their interfaces. This past INMR article from 2004, by Prof. Ravi S. Gorur, reviewed this topic.
The basic construction of composite insulators for lines and apparatus essentially involves three main components: a fiberglass core, a polymeric housing and metallic hardware. Perhaps this is an oversimplified description since it masks the numerous variations which exist in formulation as well as in processing and production techniques among the many different suppliers. Service experience has shown clearly that the various interfaces between the components of such insulators are the principal locations where problems tend to originate. This has already been recognized and there are steps outlined in the standards (such as IEC 61109) to ensure certain minimum requirements are satisfied. Incidentally, this particular Standard has de facto become the universal one being used today by insulator users who specify it as the basic requirement in their specifications for composite insulators.
As part of IEC 61109, some tests are to be performed on complete insulator assemblies while others are performed only on the various individual components. For example, there is a water immersion test conducted on the complete insulator and which is followed by a steep front impulse test as well as power frequency test. If there is a significant defect in the rod to housing interface, the insulator can be expected to fail internally in the steep front impulse test. Also, if there are major problems within the rod itself, the insulator could fail internally. Power frequency tests are performed to determine if there is any dramatic reduction in flashover voltage or if there are any punctures. Basically, this test evaluates the numerous microscopic interfaces in the housing and rod materials. It is quite likely that a diagnostic test such as power factor or tan delta would provide additional useful information. At the component level, there are tests for water absorption of the core material to ensure that there is no excessive intake of water through its constituents and the numerous interfaces within.
All these are type tests and design tests. They are performed on a limited number of samples and they need to be repeated only should there be a major change in materials, design or manufacturing process. The criteria for acceptance or rejection are clearly stated in the Standard, whose intent, incidentally, is not to provide a ranking based on the data measured from the various tests. Fig. 1 shows details of the hardware-rod-housing interface for two different composite insulators intended for the same application. This region is probably the most critical part of a composite insulator and the photo clearly illustrates the potentially large differences in how some suppliers design their products.
Both insulator types shown are presently being used by utilities, which means that they have both passed the relevant IEC 61109 tests. At the same time, however, it is clear that one has a more robust construction than the other, thereby providing a greater margin of safety (and comfort) for the user. Therefore, a more discriminating test than presently provided by 61109 would be useful so as to enable ranking important interfacial properties and assisting the process of insulator selection.
A past research project at Arizona State University evaluated the interfaces in the rod and housing materials used in composite insulators. Rods with variation in the type of glass (E glass versus ECR glass) and resin (epoxy, polyester and vinyl ester) were evaluated. Also included were silicone housings with variations in formulation (typically in the type and amount of inorganic filler as well as in the curing chemistry). One of the goals of this project was to establish a ranking of the various interfaces in these insulators.
Experiments Performed & Results
Samples of the fiberglass core were subjected to the dye penetration test as well as the water diffusion test as per IEC 61109. All samples passed the dye penetration test. In the case of the water diffusion test, the samples were individually sealed in glass tubes and placed into an oven. Fig. 2 shows the results from which it can be seen that there was a noticeable variation in the amount of moisture absorbed.
One group (samples 1 to 6) all had higher moisture absorption than did the other group (samples 7 to 9). The weight gain from this effect was highest immediately after completion of the test and then gradually reduced with time.
A voltage test was also performed in the manner described in IEC 61109. The current for all samples was less than 1 mA and therefore well within the specified limit. Fig. 3 shows scanning electron microscope photos of one sample taken from each of the two main groups exhibiting different levels of water absorption. The samples had been cut and polished in an identical manner. These photos revealed clear differences in interface quality. The micrograph depicting the rod with lower moisture absorption was characterized by relatively smooth interfaces between the glass fibers and the resin. The rod with higher moisture absorption, by contrast, displayed obvious shortcomings in quality, such as rough fibers, cracks in the resin and voids in the fiber-resin interface.
Samples of the silicone rubber housing material were then subjected to the water immersion test (boiling in 0.1 per cent NaCl solution for 42 hours) as prescribed in Section 126.96.36.199 of the Standard. Fig. 4 provides the results. While most materials displayed little moisture intake, there was one sample which showed a significant deviation from the rest. Nonetheless, all these materials would still be equally acceptable for housings.
The tests described above are fairly simple to perform and show significant differences in material properties which were not identified by the applicable Standard at the time. For applications on critical transmission lines, attention to such details in design and materials used in manufacturing composite insulators could make a significant difference in service experience.
Development of cost-effective diagnostic tools which can identify problematic interfaces in insulators both during production and later after being placed in service would probably contribute greatly to even wider application of composite insulators. While achieving this soon is perhaps still only a possibility, researchers must start by examining methods which can at least provide useful information in the laboratory. Candidates for consideration in this regard are methods employing ultrasonic devices, acoustics, vibration analysis, X-rays, MRI (magnetic resonance imaging) and possibly others.