Testing a 128KV or 345KV Porcelain Ring/Bell Insulator

"...........A man known as a wheeltapper was employed at large railway stations to check that the wheels on the bogies were sound and that the axle boxes were not hot. By using a long handled wheeltapping hammer he would strike the wheels of the bogie and hear if it 'rang true' (a wheel with a crack in it would give off a dull sound)....."

With apologies to the West Somerset Railway's website.

A megger test on a 345 kV porcelain insulator over its full length is of little value (unless the housing has a contaminated axial crack end to end). In order to evaluate the dielectric integrity of the insulator housing, high voltage or high frequency or a combination of both is generally required.

You don't mention whether the insulator is single piece, solid, station or line or whether the end hardware is installed. 345 kV implies 1050, 1175 or 1300 kV BIL and an overall length in the range of 7 - 8 feet, depending on the design and BIL.

One is generally concerned with two primary defects in wet process porcelain; cracks and porosity.

The ring test that you describe can detect firing dunts, cracks and other defects in wet process porcelain but it requires practice and experience. It isn't sufficient to tap one of the rings (more properly, weathersheds or petticoats). Couple of reasons, one is defects in the weathersheds are not as critical as those in the body and secondly, ringing one weathershed does not induce waves in the proper direction. The housing should be tapped over its length, between weathersheds or from the ends. Sound porcelain will have a bell-like ring (providing it is isolated from damping sources) whereas defective porcelain will ring with a duller tone. This method is not very scientific.

One can learn a lot about porcelain condition by visual inspection. Presumably your porcelain is glazed. Clean, glazed wet process porcelain should be shiny. If the glaze is dull, it is almost a certain indication of internal porosity in that area. Cracks can also be detected visually if they extend to the surface. In that case, penetrant die can be applied to verify and enhance cracks.

If the end hardware is not assembled, ultrasonic inspection can be used with a couplant on the end between the transducer and housing.

On hollow housings, electrical tests can be used to evaluate the housings; usually by use of a grounded screen expanded against the inner wall and energized electrodes applied between weathersheds on the exterior. A 60 Hz voltage gradient of 100 kV rms per inch is usually applied. High frequency and impulse voltages can also be used.

With hollow housings, hydrostatic pressure tests are often specified.

Mechanical tests can also be used; typically cantilever and, if done on a sampling basis with ultimate cantilever, pieces of the broken housing are often subjected to pressurized fuchsine dye tests; then broken for porosity examination.

You might want to refer to American National Standard (ANSI) C29.1, "Electrical Power Insulators-Test Methods".

Answer to Bluestone.

Your answer is far better that mine. I forgot all about the BIL test. And you quoted references which I should have. Good engineering

Snakers

In 1950 When I was at Lapp insulator works witnessing testing of 345 KV insulators for the proposed 345 KV PASNY (Power Authority State of New York) cross state power lines, we hi-potted them at double rated voltage + 1000 volts. The humidly and temperature was controlled. Flash over or puncture was considered failure.

The strings were also physically load tested on a dynamometer machine .

The bushings (for breakers and transformers) were also cantilever load stress tested for side loads.

Following this was a RI test (radio influence) test which was done by a "Stoddard" receiver and a calibrated antenna (volts per meter). There were no RI standards at that time. The RI data was used as a measure of merit between vendors.

Some were also tested for leakage current at rated voltage and power frequencies (capacitance)

Cornell University also had a test site and a 765 KV TIDD transformer and tested many insulators on there test site in Ithaca NY.

I believe this was all in accord with the IEEE test procedures in place at that time.

I do not recall any megger testing.

This was how it was done in the 50s. Probably but not very practical in your case

Snakers

Greetings,

You may wish to contact Prof Ravi Gorur (ravi.gorur@asu.edu) at Arizona State Univ who does insulator testing in his high voltage lab or can consult on test procedures.

Insulator testing is not only a fair weather test, they do contaminated testing also, and the altitude (RAD, Relative Air Density) the insulator will be installed at, loaded plus ice load testing at University of Quebec in Chicoutimi contact Prof.Jianhui Zhang jzhang@uqac.com

Also consider using a corona camera to view the insulator under test, corona on an insulator can lead to failure or be a sign of failure, www.corocam.com and www.corona-technology-course.com may also be of help. I am the North American CoroCAM distributor.

Dan Ninedorf

Tel: 800-531-6232

I would try a dye penetrant test they come in regular visible light and fluorescent types. It will take a little trial and error to properly interpret the results due to the surface variations. They will require a good cleaning afterward to avoid contamination failure.

Also consider x-ray, but it must be at the right incidence angle if you are looking for cracks and internal voids.

Also, a tap test with an audio spectrum analyser program on a pc may work. The tapper should be a fixed, consistent tap. A spring loaded tapper with a release mechanism in a fixture to hold everything will allow you to control several of the variables to help get consistent results. Collect a sample of units with defects to use for baseline sound signature for comparison testing of questionable units.

Good luck!