Ultrasonic testers can be used for many applications. I am familiar with using them for partial discharge (corona) detection in medium voltage cables. Although some defects can be detected with an ultrasonic test, most are missed. Therefore the detector should not be trusted to prevent failure.
Benjamin Lanz
Vice Chair of IEEE 400
Sr. Application Engineer
IMCORP- Power Cable Reliability Consultants
EA Technology have developed this into a working commercial product. It's incorporated within their 'UltraTEV' partial discharge monitor. The ultrasonics add to the detection capability of the instrument, being employed in conjunction with a TEV (transient earth voltage) sensor.
Ben - I'm interested in hearing your opinion of the TEV sensors used for PD location in handheld instruments. The principle itself seems pretty well established.
My area of expertise is cable so I will not make any statements about other electrical components. I am aware of word transient earth voltage (TEV) but, I have not seen the term used outside of marketing literature for cable. My understanding is that it is just another way to pick up PD signals. However, there is no magic here. On-line test (not on-line monitoring) are not assessed for sensitivity and do not asses the system during transient voltages.
Sensitivity
On-line tests do not include a method to verify the sensitivity of the test. There isn’t a way to measure the signal loss from the source of the signal (the defect) to the detector. Therefore, it is impossible to state with any confidence that no defect exists and the probability for failure is low.
Transient voltage
Your system experiences transients all the time. On-line tests only measure the apparent activity at the operating stress level. So, you measure the PD activity with a wand with an unknown sensitivity at the operating voltage, and find no ‘serious’ activity. Then you turn around, walk away, and your system experiences a transient voltage. The defect is ‘ignited’ and eats away at the last bit of insulation, and… bang you have failure.
This is no magic wand. On-line tests are a good option only if you can not take the system off-line and perform a thorough test.
Benjamin Lanz
Vice Chair of IEEE 400
Sr. Application Engineer
IMCORP- Power Cable Reliability Consultants
Our company recently bought one of the EA instruments and a few colleagues are treating it as a 'magic wand' which provides a convenient excuse to avoid the hassle and cost of taking a switchgear bay and cable out of service for testing. You've reinforced my instinctive feeling that it might be useful in finding a fault on the verge of failure, but is a poor substitute for an offline test on an isolated circuit. In all fairness to EA, they do offer an offline testing service too and recommend that method wherever it's possible to do so.
In response to the original question, Ultrasonic detection is a good method of detecting surface discharge on MV components e.g. bushings, dry terminations.
TEV is better for discharges deep inside the insulation.
Using both techniques will ensure that most types of discharge can be detected before failure particularly on 15kV switches and accessories. The techniques are less useful on cables except on terminations (where we find many problems).
Benlanz and ScottyUK are also both correct in saying that techniques are a no substitute for an offline direct PD measurement but it is a trade off between convenience of an on line test and the sensitivity of a an offline test. At 15kV, few companies would be doing regular offline PD measurements (except on cables) and many do neither offline or online (not reccommended!).
We have been performing offline IR/DDF and PD tests phase to earth and phase to phase on our 11kV switchgear for probably 30 years now, and have found many problems with both DDF and PD.
We have trialled the PDM (multichannel TEV PD monitor with a number of input channels and a few antenna channels to remove outside noise) and Micro TEV (handheld unit) instruments, and had mixed success with finding PD problems. Although there is no calibration possible, we tried to get a rough comparison between the dB readings, and an actual discharge. In most cases we found that there needed to be a discharge of around the 1000pC mark before the micro Tev got to an orange state, and that the dB levels in the PDM showed above the recommended values. Discharges at this level are of grave concern, and ones that would not go back in service without entry restrictions on the substation.
Our main concern is that whilst PD detection is a condition monitoring tool, we have found more issues with DDF rising over time than discharges in our gear, so we would be taking the group out of service to do DDF tests anyhow. Thermal runaway on the insulation in this type of switchgear, at the age of some of our gear can be a problem.
I cannot give any comment on the Mini TEV, however I believe that one of the drawbacks of the Micro TEV (just having a red, orange or green light for condition indication) has been improved to have the actual dB reading on screen. We hope to trial this unit soon, which we might use more as a safety indication for staff entering substations.
My belief is that these units (TEV monitors) would be a reasonable ranking tool, if you have limited test capability or staff, to plan which are the top (worst) say 10% of switchgear (from a PD point of view) you want to target with offline testing.
I agree with you but, I would like to add two caveats.
1. I do not recommend using the results from an uncalibrated sensor as a basis to assure reliability or protect human life.
2. I reiterate my caution with the use of uncalibrated measurement system to test MV-EHV power cable. 60% of all defects in shielded power cable are less than 50pC, which is below the threshold sensitivity of most on-line tests in the field.
As I said before, I do not have extensive experience with measuring PD outside of my area of expertise, MV-EHV shielded power cable. However, I can imagine that you have found some value in using uncalibrated sensors to detect PD activity in switchgear, buss bar, bushings and other unshielded equipment. This makes sense to me. If you find a problem you most likely have saved yourself a potential failure and paid for the PD survey service or equipment -a good investment.
Benjamin Lanz
Vice Chair of IEEE 400
Sr. Application Engineer
IMCORP- Power Cable Reliability Consultants
TBT,
We have had great sucess with finding faults with our Ultra Sonic detector.
We have found faults in metal-clad switchgear, open air disconnects, transformer tap changers, tracking through outdoor CT's, and splices, just to name a few.
It really makes a difference WHO is using the detector.
I would like to thank all of you for your input. I think we will give ultrasonics a try. It sounds like this technology, like many other technologies, depends on having the proper training in order for the program to be a success.
