Factory Witness Tests - Power Transformer - What Tests? 9

[dpc]

I'm looking for opinions on what tests are most important to actually witness during factory testing of power transformers if there are time constraints I'd assume the dielectric tests are the most critical, but how would you rank:

no-load and load loss tests
temperature rise test
turns ratio
resistances

Thanks,

Dave

 

[rbulsara]

voltage regulation?

Rafiq Bulsara

http://www.srengineersct.com

 

[dpc]

Yes, I left off a few. I'm just trying to establish a priority if time is limited.

 

[marks1080]

Ranking what you have listed I would say:

1 - resistance
2 - turns ration
3 - temp rise
4 - no load/load losses


But the question I would have for you is why would you want to rush this? You should be able to fully test a transformer within a couple days if under the gun... would you really need less time?

 

[marks1080]

ratio*****

 

[dpc]

You shouldn't assume it is me who is a rush. If the client says that they want to go to do a factory witness test but want to do it all in one day, sometimes they can't be persuaded otherwise. Then we do the best we can, since it is their money, not ours.

 

[rbulsara]

I have not done FWT of a transformer but I would ask the mfr to provide their standard FWT procedures and offering and would review it. That will be the basis of the testing agenda.

Anything requiring an extensive test set up may not get accomplished in a day. I have yet to see any mfr being ready with the test setup before we arrive on the floor! Any special test requires advance notice.

Assuming load setup is there and only speaking for electrical:
I would go with more or less, your list starting with dielectric, impedance, Voltage regulation and losses. If oil filled, is there a way to check integrity of sealing for leakage(like a pressure test?).

Review their routine test reports?

I would also be interested in verifying the accessories to be present as specified that are easier to get fixed at the factory than on the site.


Rafiq Bulsara

http://www.srengineersct.com

 

[jghrist]

Some tests will also be done in the field. These could be skipped in the factory witnessing because the field test will serve as a check. If the factory skips or bungles the turns ratio test, for instance, it will show up in the field. You don't want to find out that the turns ratio is wrong after delivery, but neither does the factory, so they aren't likely to screw that one up.

Frankly, unless you are concerned that the transformer manufacturer might not even do the tests without your being there to witness them, there isn't a whole lot of value in witnessing the tests. The tests are automated and computerized anyway, so you don't actually see anything but computer printouts. It's not like someone is reading values from an analog meter and writing them down so you can check on them.

 

[LionelHutz]

Then ask the client what they want to see during their visit. That way, you won't set-up to do certain tests and then have the client arrives and wants to see something else.

 

[dpc]

Thanks all. Unfortunately the client generally has little idea of what they want to see.

jghrist - I've never really seen the necessity of having someone witness the tests either in most cases. I think the best transformers I've ever purchased were done without a witness test. But if they want to go, but only want to spend 4 hours on something that should really take two days, we may as well try to make it as useful as possible.

The interest in factory witness tests seems to have a lot to do with the location of the factory.

 

[prc]

dpc, your last sentence tells all about the witness test !!

Any good manufacturer, worth his salt, will conduct all routine and type tests before the unit moves out of the factory whether witnessed by client or not.Most of them do not conduct internal test and then separate witness test.

In case you ask the bare minimum tests for witnessing, I will put as below, in that sequence. It does not mean that other tests are not important.I will ensure that manufacturer completed all LV tests themselves ie ratio,resistance,phase sequence measurements:

PD Test
No-load loss
Load loss
Temperature rise test
Impulse
Switching Surge
AC Induced overvoltage test with PD measurement.
Separate Source AC test.
Repeat No-load test
Leakage Test

 

[edison123]

I would rank the dielectric tests (separate source, induced overvoltage, tan delta etc.) at the top and then ratio test as minimum FAT. The rest like loss tests are already done and dusted and nothing can be done about it except may be penalize for losses beyond the contracted.

Muthu

http://www.edison.co.in

 

[xxjohnh]

Depends on the size and voltage ratings. Witnessed tests are usually paid for and are you using type tests for short circuit test values? Surge tests are important for large units and they some times fail. ANSI and IEC are similar except for impedance but you do need to get the standard your transformer is built to and study it before going.
Standard tests listed as factory tests should be the minimum.

 

[xxjohnh]

Tests should never be not performed in the factory to be done in the field. It is the other way around. Too late for repairs in the field. Transformers do fail factory tests. Nowdays with equipment being manufactured in new locations factory tests are important. The ANSI and IEC standards are the mandatory documents. Don't expect the manufacturer to give you a copy.

 

[marks1080]

Personally I would seriously consider not preforming field tests if the appropriate manufacturer test results were not provided.

 

[jghrist]

The interest in factory witness tests seems to have a lot to do with the location of the factory.

True! This also negates "buy American" predispositions. We had one client that was happy to accept GE Prolec transformers made in Monterrey, Mexico because the Utility Manager was witnessing factory tests.;-)

 

[rovineye]

I can tell you what I look at.
1. Dimensional check (overall size, maintenance clearances, attachments locations, bolting arrangement)
2. Auxiliary checks (Temperature monitor, leak detectors, anti-condensation heaters)
3. Winding resistance
4. Voltage ratio and phase displacement (important for us).
5.Impedance voltage, short-circuit impedance load loss.
6. No load loss and and magnetising current.
7. Separate source voltage withstand test (checks creepage clearance HV-LV and to earth).
8. Induced overvoltage withstand test.
9. Insulation resistance.
10. Temperature rise using simulated load method.
11. Sound level on first of type.
12. Weight on first of type.

Reference for most of these checks from IEC 726 & 76-1.

We witness tests as they are performed during the factories routine. Thye tell us the dates and duration, and I am fine with that. It takes our factories 1 day per xfmr generally (not oil cooled).

Location, location, location! For sure that drives the desire for industrial tourism.

 

[stevenal]

I agree that the dielectric tests are particularly important, since a failure of these tests might necessitate untanking to remedy. Impedance at the factory is important as a baseline to see if damage occurs during shipping. Loss testing is needed to see if the transformer meets spec or loss evaluation criteria.

The most value I get from witnessing is to see if the details in the spec are addressed. I skipped the last transformer's tests, since the same bidder got the order as the time before. By this time, we thought they should know the spec.

If place drives the decision, consider including travel costs of the witness in the bid.

 

[dpc]

Thanks for the tips - this helps me out.

Cheers,

Dave

 

[ausphil]

Our idea on witness testing is to firstly look at the test results on the transformer or cable or switchgear that is being tested to see if they are within spec, but also to ensure that the tests that are being performed are being performed correctly, and that the manufacturer has a system in place for eliminating a problem from happening again if something fails. You look to see that they have a system of calibrating their instruments, and that they are fit for the purpose for which they are to be used (ie not using a multimeter on resistance mode to measure the resistance of transformer windings).

Once you are happy with the test methods and that the results are correctly measured, then there really is no technical (as opposed to holiday) reason to be witnessing there again, except for some special circumstance.

To make witnessing worthwhile, you need to send people with test experience who know what tests need to be done, the test methods, and most importantly what can go wrong in measurements that can give false readings. Partial discharge testing is a common one where factories tend to eliminate noise from cranes, lights etc via a gating function. We have seen some manufacturers that have essentially gated out all the actual partial discharge, as well as the noise, so they record it as a pass, when in fact, the test has failed. In essence, to be of value, you must have people who know what to look for, rather than just looking at numbers produced by the manufacturer.

But to answer your original question, if there were time restrictions, then I would be there for dielectric tests (Applied, induced, PD, impulse), then I would rank the heat run next, mainly because the results of it are based on the loss measurements anyhow, but this is a time consuming one - you may want to come in on the end of this one when they are shutting it down and measuring the resistance when it is hot. If they are doing these correctly, you can take a fair bet that the resistance, ratio, impedance, oil tests are going to be done correctly. While you are there, look at the equipment they use for the other tests, see if it is in calibration. Ask the testers how they perform the other tests that you haven't witnessed. By the end of this, you should have a fairly good comfort level as to the correctness of the test methods and results.

ausphil