motor/gen HV testing. DC or AC HV? 1

[VLFit]

Like a lot of apparatus HV testing, it seems most DC hipot the object, like a stator coil, to perform the final evaluation. WHY? Besides the obvious answers of DC is less destructive and DC hipots are small and cheap, are there other reasons why an AC Withstand test is not performed? Maybe it's just economics??? maybe it's a lack of understanding the differences and why one or the other should be used. With an AC hipot, then partial discharge and power factor or tan delta can be measured too. ???

 

[electricpete]

Like a lot of apparatus HV testing, it seems most DC hipot the object, like a stator coil, to perform the final evaluation. WHY? Besides the obvious answers of DC is less destructive and DC hipots are small and cheap, are there other reasons why an AC Withstand test is not performed?

Those are the two big reasons. When you say less destructive, it carries at least two different aspects:
1 - dc test set trip allows much lower fault current during trip since the trip setpoing does not have to accomodate charging current, like an ac set.
2 - a single void can discharge only once during a dc test but repeatedly during an ac test. So you can accumulate a lot pd damage in an ac test than a dc test.
Watching an ac hi-pot conducted to destruction, you can get a feel sometimes quite a bit of action is evident (noise, flashes) before the test set would trip.

With an AC hipot, then partial discharge and power factor or tan delta can be measured too. ???

Typically power factor would be measured at nominal line to ground voltage and some lower voltage around 2kv or 20% to allow power factor tipup.
Typically partial discharge would be conducted by varying voltage to determine DIV / DEV but through a range up to normal line to ground voltage.
Above is typical practice and I think you will have the best luck evaluating data in this range since there is more data available for comparision. I don't know of any technical reason why you wouldn't measure power factor and pd above normal equipment line to ground voltage (as long as it's within test equipment limits), but it's not standard practice.

My comments are based on motor testing.


=====================================
(2B)+(2B)' ?

 

[VLFit]

Thanks for the reply.

Your points are good but,,,

My next question is to be expected, a motor or transformer or most everything operate under AC stress. The stresses and failure mechanisms are very different between AC and DC. Wouldn't you always want to test with AC if you could? What you say about the nature of a dc versus ac fault is true, but if there is a fault present, we will have to find it to fix it. Let the AC voltage make it a little more obvious. The other tests I mentioned I thought were done up to maybe twice voltage above normal operating voltage to see where pd might kick in and where a tip-up like you say may be when TD or PF testing. It jsut seems that many of hte motor shops I see and some I use, pull out their 30 year old 5 kVdc hipot and think they are doing a proper withstand test.

Thanks

 

[ScottyUK]

A high voltage AC test set can be fairly expensive because it has to deliver a fair amount of current if the load exhibits significant capacitance to earth. This is one of the reasons VLF test sets were developed for cable tests. VLF sets are fairly expensive too, unless you're a testing contractor and can get the equipment to earn its keep.

 

[VLFit]

You are correct about the high cost of a big AC test set, but many shops can benefit from smallish 30 kVac hipot to test up to 13.8 kV coils, like a 20 or 30 kVA set.
Regarding the VLF, you are right that most are used for cable testing but it was actually first developed for generator testing by GE and others in the60's. The first IEEE spec for using VLF for motors is IEE433-1974. Just upgraded to 433-2009. The VLF can be used as a source for overvoltage to do TD and PD testing. There are alternatives for motor shops if they would take a fresh look, but most, like other industries, stick with what they know, the ole dc hipotter. I was most wondering if testing is done with DC mostly because of the cost, unfamiliarity with technology like ac vs dc, and/or whether the motor engineers always write a test spec around dc. I am most familiar with the EASA type shows. What about the big generator OEMs and the huge independent repair companies, like a GE. I would assume they do all they can to test since the event of a field failure is so catastrophic.

 

[electricpete]

My next question is to be expected, a motor or transformer or most everything operate under AC stress. The stresses and failure mechanisms are very different between AC and DC. Wouldn't you always want to test with AC if you could?

I agree, AC may reproduce some stresses better, especially in the endwinding (when testing with one phase energized and others grounded).

*** DC provides an ability to do a simple stepped test or ramp test, plotting current vs voltage to provide some diagnostic insight beyond the go/no-go result from ac hi-pot. And may possibly (not guaranteed) allow the user to stop the test before the voltage is raised to the point of causing a failure.

What you say about the nature of a dc versus ac fault is true, but if there is a fault present, we will have to find it to fix it. Let the AC voltage make it a little more obvious.

One aspect mentioned - there is the pd which occurs during ac testing, much more than during dc. Normally we don't think of pd as something that would be damaging during the short interval of a test, but that concern about ac hi-pot of service aged windings which has been raised by people much more knowledegable than me (Greg Stone's book if I remember correctly).

The other aspect mentioned is that the damage from a fault during ac power frequency test is more extensive and requires more repair. How likely there is a significant difference... probably not very likely, I agree. You mentioned for some users in some situations of field testing, repairability without rewind may be a very important criterion (one could also argue they shouldn't be doing hi-pots).

Most importantly , as noted above, the stepped or ramped dc test may provide much improved ability to identify poor insulation and stop the test before the point of failure. In that case we are not comparing failure from ac test to failure from dc test, but failure from ac test to potentially avoided failure during dc test. A big difference when testing in the plant (vs shop).

DC has a lot of traction for historical reasons and cost reasons. If a machine owner uses dc testing in the plant, he may be more likely to specify the same for his repair shop.

There are pro's and con's as you've discussed. Personally I'm not advocating one over the other. We may use/specify one or the other or both, depending on the situation. I tried to respond in a manner that responded to your initial question and provided balance to your points.


=====================================
(2B)+(2B)' ?

 

[electricpete]

I forgot portability for dc, in case that wasn't mentioned already.

=====================================
(2B)+(2B)' ?

 

[VLFit]

Thanks for your time and valuable input. It's the dc vs ac issue, destructive vs non destructive, value of leakage currents vs go/no-go ac withstand, etc. to be continued

 

[VLFit]

Motor Testing High Voltage AC vs. DC. I am reposting this question in the hopes that there are more answers, or opinions, out there. The issue is important since many more motor shops want to perform PD and TD/PF testing and will need AC high voltage supplies. Thanks and Happy New Year to all.

 

[Addy71]

We use a DC Hipot tester (Phenix) as well as a power frequency tan delta tester (Tettex + Eltel) & a PD analyzer (PD Tech + HVPD). We have wondered about purchasing a VLF source to avoid the resonators; but I am unsure if the tan delta values plus PD values & patterns at power frequency would be replicated at VLF.

If yes; then I would appreciate if you can post some references/papers. Else; I do not see the benefit of VLF for generators (though I guess it is essential for cables).

Regards,

Aditya

 

[VLFit]

Thank you Aditya,

It seems you are quite familiar with the subject. Do you use the DC hipot for a withstand test or do you have an AC hipot for that? You have to use AC for the diagnostic tests so I figured you must use AC for the withstand test as well, although it will have to be higher in kV and KVA than what is required for the other two tests. The VLF could help there. The VLF as a withstand test is covered in IEEE 433-2009. As for using VLF for the PD and TD/PF testing, I will have to check. The numbers are different than what you will get at power frequency but would still be usable. There are several places in the standards that reference VLF for these diagnostic tests. I will have to look around. Of course a 60 Hz high voltage source is best but can be prohibitive, hence the push towards VLF. I'll get back to this forum with more info in a week or so.

Happy New Year.