PD and TD testing motors and generators - what to use?

[VLFit]

Finding the right AC hipot needed to apply the voltage in easy, but what vendors supply a reasonably priced PD measurement device for elevated voltage off-line testing? Same for a Tan Delta device.

 

[electricpete]

We use the old standards Qualitrol/Iris partial discharge and Doble power factor and have been for years. Haven't done any comparison shopping so I don't have much to say unless you're looking for comments other than price.

fwiw I suspect you will find a lot cheaper options for power factor / tan delta and can get good value from those less expensive options. But partial discharge is a trickier beast and I think the vendors tweak their parameters and presentation such that comparing results among vendors is not particularly easy. If you're trending your own machines over a period of time that's not a problem. But if you are testing for others they may have preferences (as a plant customer of motor repair shops, when they test our motors I'd prefer them to use Qualitrol/Iris)

 

[edison123]

PD testers are expensive and testing methodology varies from supplier to supplier. Despite the hype from Doble for years, I don't put much stock on the efficacy of PD testing since it is historic trending and results are subjective interpretations at best.

Tan delta kits are much cheaper and testing methodology is the same for all suppliers. The results are not subjective and standards have specific values as guideline unlike PD testing.

Of late, I see that Tan delta testing is also losing its sheen among my clients.

No one test or combination of tests can predict the residual life of electrical winding accurately in my experience. They are all guestimates at best.



Muthu

http://www.edison.co.in

 

[VLFit]

Thank you for the replies.

Although accuracies in tests may be suspect, and the repeatability of results over time depends largely on the consistencies in the methods used and various comparative technicians abilities, both are still useful indicators of winding degradation. The TD test is good for time trending and comparative analyses and the PD is the same but also useful for spot checking to catch problems. There are not many other diagnostic alternatives more precise or repeatable. Since the PD test can be rather crude, looking for maximum allowable levels compared to the precision of cable PD testing, the equipment and the methods can be less advanced, like the IRIS PDTech DELTAMAXX products I assume you are referring to. One problem is that many of the systems available are far more sophisticated than needed, having been designed for cable testing and not apparatus testing. One positive development that helps both is the gaining acceptance (IEEE433-2022) of using VLF AC hipots for the voltage source rather than power frequency. Both PD and TD can be performed at 0.10 Hz, as well as overvoltage withstand testing of course.

We'll keep looking and trying. Maybe there will be something new at the EASA show next week.

Thanks for your input.

 

[electricpete]

We have VLF cable tester and EPRI gave a presentation awhile back on using our VLF cable tester for motors. In my understanding, it was claimed there was some ability to test without determinating the cable, and still distinguish somewhat between cable problems and motor problems by performing successive tests with cable shields lifted and landed. On the surface that sounds like an attractive option to avoid determinating the power cable itself (which saves work and potentially avoids introducing problems if terminations are not properly remade) but the details seemed rather complex and I was under the impression more work is needed to firm up the recommendations for interpretation of results.

 

[VLFit]

What you say is true, however, your testing scenario introduced cable testing into the equation, not just generator coil testing: much more complex in distinguishing what’s going on where. VLF is great for testing those high uF loads where a Parallel Resonant set or huge AC Dielectric Test Set would be needed, like with cable testing. VLF works fine for TD testing anything, especially when you have little choice regarding test equipment for field applications. Same for PD although the VLF waveform comes into play in terms of the accuracy and repeatability of measurements. It's good we have the many tools we have to select what works best for different situations. VLF has been used quite successfully for over twenty years for withstand cable testing as well as the two diagnostics we're talking about.

 

[edison123]

Good to know VLF testing is gaining acceptance by standards. Are these test kits cheaper though? And will the clients accept it as an alternative to power frequency testing?

Muthu

http://www.edison.co.in

 

[VLFit]

I assume you are referring to clients that are testing motors and generators. For cable testing we have used VLF for withstand and TD and PD testing for over twenty years, with standards existing for almost as long. That is old news and beyond question.

If you are referring to only rotating machinery testing, the first IEEE433 spec was published in 1974 following work done by GE in creating their own system for testing their large generators. That standard has since been upgraded twice, now to IEEE433-2022. Many are using VLF for this application as there is little reason not to, mostly for field and maintenance testing.

Hope this helps

 

[edison123]

Yes, I am aware of 433 std for rotating machines and have the 1991 version. Unlike the power frequency hipot, VLF hipot has not been made mandatory or even mentioned as alternative test in other rotating machines standards and hence none of my clients accept it as proof testing.



Muthu

http://www.edison.co.in

 

[VLFit]

That's a pity. Why is the IEEE standard not good enough for your clients? Worldwide standards have existed for twenty years for using VLF for testing cables and terminations. Rotating machinery is a far less complicated and exacting an application than cable testing.
There is nothing magical or mysterious about VLF. It is simply a reduction in the frequency while maintaining a suitable waveform that conforms to standards everywhere. Their argument is even less valid when you consider they use DC voltage and think that is ok. ????

 

[Gr8blu]

A large chunk of the eccentricities in partial discharge measurement is a result of the way a specific stator coil is manufactured. The coil geometry, the connection between coils, the connection to the circuit rings, the type of lead and lead routing within the stator frame - all affect the partial discharge results. This is why the PD testing on a single machine can be trended with reasonable accuracy, but why a standardized pass/fail criteria cannot be applied across all (or even most) machines.

Also note that IEEE 1434 deals with partial discharge measurement in rotating machines. Latest version I have access to is 2005, but there should be something newer.

Converting energy to motion for more than half a century

 

[edison123]

When VLF testing is offered as an alternative for rotating machines power frequency testing in the rotating machine standards, then the clients will accept it. Otherwise, they have every right to demand we follow the present testing method. No point getting huffy about it.

Muthu

http://www.edison.co.in

 

[VLFit]

I understand, but probably many would opt for VLF is they understood the economics and convenience of using VLF over a multi-ton power frequency set. For withstand testing it should not be objectionable.

 

[edison123]

Yes, I agree lugging a much lighter test kit would be a weight off my shoulders.

Muthu

http://www.edison.co.in