Corona investigation via ultrasound

[lielec11]

I'm working on a job to develop procedures for a large facility that includes 15kV switchgear. Going through the latest version of NETAs testing and maintenance manual one of the visual inspections for switchgear reads as follows:

Inspect physical, electrical, and mechanical condition including evidence of moisture or corona.

I've heard of the corona effect in the past but not knowing very much I started doing some research. I came across an article (http://www.bradyinfrared.com/news/NEWS21194.pdf) from a testing company that mentions utilizing ultrasonic testing to track potential corona issues.

Is anyone here familiar with this, or has anyone actually performed this kind of testing before? I would like to learn some more about it before I decide whether or not to recommend it to my client.

Thanks..

 

[ATSman]

Corona Detection

Corona Detection

I have not read your article but a trick that us old-timers used on 4160v and 12kv back then was to take a small transistor radio and tune it off of a station (dead zone static) and just move it around from cubicle to cubicle front and back. If corona discharge was detected the volume of the static increased and gets louder the closer you move the radio to it. It seems that the frequency of the arcing was in the same range as the AM/FM band. When arcing occurs the air around it becomes ionized and the Oxygen molecule O2 breaks apart and forms O3 molecules (Ozone) which has a distinct odor. Like what you smell in the air after a lightning storm. Also if the environment is quiet enough then you can hear the arcing; sounds exactly like bacon sizzling in a hot pan. If indoors turn out the lights and just crack open the enclosure door and you may see a blue glow around the corona. Remember to use your eyes, ears and nose!
We found a lot of problems in MV gear before they turned into disasters.

http://www.canada-training-group.ca/Newsletter/Entering-Outdoor-Substations,-Part-5-of-8

 

[steve66]

I didn't read the link, but I scanned it a little.

I would be very skeptical believing anything from a single source with no real credentials or credibility.

Especially when that source is offering the same services they are promoting.

 

[ghostbuster]

I have also used this cheap "radio sniffer" to determine if arcing is present.

The best way however is to use an ultrasonic detector with a parabolic dish and headphones to accurately locate (from a safe distance) the exact source of this arcing.

 

[junkhound]

+1 on AM radio

Overall, the OP link set off my BS detector, so did read the entire article.

a. ionization of air as the nitrogen in the air breaks down -- misleading, the Paschen cure for air, nitrogen alone, oxygen alone, argon alone, are all different. It is not just the N2 that ionized - how else to get oxone.

b.
......present without current flow.... - it may only be mA, but there IS current flow

Other than those 2 tiems, fairly accurate general description, although it does imply that ultrasound (or straight audio) is as sensitive as other detection methods.

Did a study of corona detection a few years ago, and determined that any sound detection at sea level (poorer detection with altitude obviously) is much less sensitive than dark room visual, CT 1-20 MHz measurements, or RF detection (e.g AM radio)

That said, ultrasound could be useful, but is inferior to other detection methods.

 

[big john]

...I've heard of the corona effect in the past but not knowing very much I started doing some research. I came across an article (http://www.bradyinfrared.com/news/NEWS21194.pdf) from a testing company that mentions utilizing ultrasonic testing to track potential corona issues.

Is anyone here familiar with this, or has anyone actually performed this kind of testing before? I would like to learn some more about it before I decide whether or not to recommend it to my client...

We do a lot of MV testing and corona issues are very common. Often we find them during de-energized maintenance where you can see there has been tracking or discoloration from the discharge on insulation surfaces.

Occasionally the corona is so severe that you can smell it as ozone in the energized cubicles or hear it, this is especially true of high-voltage overhead structures.

As part of our online maintenance we do a basic PD survey at the same time we do thermographic scans. We use the UltraTEV+ which is a good front-line tester. It's not a guarantee because it can be difficult to detect PD within metal-clad gear, but it's pretty common to find suspect equipment on MV gear, especially in old, dirty plants.

What I like about the UltraTEV+ is that it doesn't rely on any single mode of detection: You're using both ultrasonic and surface voltage detection in conjunction so it helps rule out false positives caused by outside interference. I've never tried the radio trick, that's not a bad idea, I think I'll take a little pocket transistor radio with me next time I go out and see if that corroborates what my detector says.

 

[lielec11]

Thanks for all of your responses. It seems to be a legitimate method of corona detection. I'm assuming the "radio" method and using actual ultrasonic testers has to be done while the gear is energized? Or can you detect the corona effect while de-energized? I understand that you may be able to smell the ozone or see the white residual dust if the gear is de-energized which helps. Just not sure if you can use the test equipment while de-energized as well.

Thanks!

 

[GoldDigger]

Both ultrasound and RF methods require not only that the equipment be energized but also that the corona be occurring (if it is weather dependent, etc.)

 

[zog]

Ultrasound can detect corona but only after it is to the point of near failure and IMO it's not any better than a visual inspection. PD detection is going to find issues much earlier than any acoustical methods. As Brian mentioned the Ultra TeV is a good basic PD detector.