High DGA 11

[powerzizo]

Hi All
we have two identical generators transformers 420MVA(18/220kV). both transformers has almost the same service time 14 years and got high DGA and buchholz alarms one year in between.
both transformers are currently under inspection by OEM and initial findings were burnt cable in phase -W- HV leads to OLTC at 10CM distance from the HV winding.
another two similar transformers are still in service with almost the same operating conditions and service time.we are concerned if the in service transformers will suffer the same issue shortly. The routine DGA we do every six months and results comes clear for the in service transformers . the two faulty transformers most recent DGA before the failure were also normal one or two months before the buchholz activation.
are there other recommended tests should we do to predict this type of failures,noting that we do tan delta, DC resistance, turns ratio every year and results come normal?

due to long down time to repair any of the two faulty transformers, and considering long lead time to get new transformer, may an expert advise me the fastest way to find similar capacity transformer to minimize the shutdown time?

 

[ScottyUK]

On a transformer of that size and value the cost of a Hydran or a full online DGA system could be justified.

As for finding a transformer quickly, you need to be looking at the surplus market for a similar transformer unless you want to pay up front for a spare you will hopefully never need to use. You'll need to make some effort to preserve the spare transformer in a serviceable condition too.

 

[electricpete]

initial findings were burnt cable in phase -W- HV leads to OLTC at 0CM distance from the HV winding

Did you find the reason for the burnt cables? Are they overheated along entire length or just at one location? Is there a high resistance joint there… or something else?

are there other recommended tests should we do to predict this type of failures,noting that we do tan delta, DC resistance, turns ratio every year and results come normal?

You've listed most of the easy/productive tests.
The only other things that come to mind are
- if you don't install a hydran as per Scotty's good suggestion, consider increasing the frequency of your oil sampling
- Depending on what you know about the cause, I imagine eventually you'll want to drain the tank for internal visual inspection or correction? In the meantime also possibly increase the frequency of outages for off-line electrical testing.
- external thermography - we have been able to see high resistance contacts in LTC comparment this way. Ordinarily we don't hope to pick up anything on the outside wall of the big main tank that might indicate localized winding heating, but if you know exactly where to look based on those other transformers, then you might study for small changes/deviations (especially comparing against similar transformers or trending over time). Also you never know there might be some other indirect clue about transformer condition in infrared thermography.



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

 

[ScottyUK]

Many years ago a 3-phase single-tank 400MVA GSU transformer suffered a failure which appears to have some similarities with what you're describing. In the case of the transformer I'm thinking of, the problem was poor brazing of the leads of the HV tap winding which had resulted in porosity in the brazed joint and high resistance. The OLTC was a reversing type with 16 buck and 16 boost positions plus a neutral position. The failed transformer was repaired, and the sister transformer was found to have the same problem: on this latter transformer the tapchanging capability was disabled leaving the transformer on the neutral tap rather than undertake a long and expensive repair. I assume some commercial agreement was reached between the transformer manufacturer and plant owner to reflect the loss of reactive capability on that generating unit.

 

[bacon4life]

The OEM should be able to give you more details as to what caused the burning. Once you know whether it was slow warming or a sudden transient, you may be able to inspect the other 2 transformers. When you say similar, do you mean the OEM or the same ratings? If it is the same OEM, they should be able to explain whether or not other transformers are risk of similar failure.

Adding the online DGA to the transformer for the second half of the transformer's life as it increases in failure probability is great plan. Since the online DGA units have a much shorter life span than transformers, we are now specifying that all new transformers have appropriate fittings & valves for DGA but not actually installing the DGA units. For the online DGA units we did install, a huge cyber security challenge was to quickly get the data into the hands of someone both qualified and authorized to remove the transformer from service.

 

[prc]

You have to wait for the report from OEM or transformer expert. From this one can decide whether there is chance for similar fault in other healthy units. Once you identify the root cause of failure, then only we can decide the appropriate diagnostic or monitoring tests for healthy units.

 

[powerzizo]

Thank you all .
i attached some photos for the first faulty transformer findings during inspection by OEM who recommended insitu repair after skidding out for 2 m to enable de-tanking.Root cause not yet identified by OEM.
The second faulty transformer not yet inspected.
DGA results by OEM two weeks before the failure shows all gasses are normal but DBDS 200ppm. is it significant?
please have a look on the two DGA results two weeks before the failure and after the BUCHHOLZ activation and shutdown.
your thoughts will be highly appreciated

 

[prc]

DBDS-200 ppm? This is not shown in the test reports posted by you. If it were true, immediately arrange for testing oil samples from all transformers for potential corrosive sulfur in line with IEC 62535-2008 or ASTMD 1275B.
Since DGA did not reveal any abnormality before failure, there is no point in going for Hydran or any other on line monitoring. First point, find out the root cause. By this time it should be clear to an expert. Are you sure burning is only outside? Is there a joint in lead ( brazing or crimped) at the position of burning? If so, was it in sound condition? Some thing deep inside? Have you done all LV tests esp resistance, single phase excitation current at all tap positions on the faulty units? If not, it will help in identifying the problem.

 

[electricpete]

Sounds like a good call by prc. I hadn't heard of dbds, but as prc suggests sounds potentially destructive and that level sounds high

http://www.netaworld.org/sites/default/files/public/neta-journals/NWfall08-Lewand.pdf

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

 

[ScottyUK]

The fault must have developed quickly and aggressively for it to go from a clean DGA signature to outright failure within two weeks. It may have failed in a matter of hours or minutes, perhaps seconds. I agree that DGA is the wrong tool for faults which develop quickly, but I'm not certain I would dismiss continuous online DGA out of hand without understanding what took place here. It would certainly be interesting to understand the failure mechanism.

ePete,

That's a useful paper. Thanks for the link.

 

[powerzizo]

Thank you.
The two in service transformers had Irgmet passivator added before 7 years .the passivator content is being analyses every 6 months and still around 60ppm.However ,we are currently analyzing corrosive sulfur and waiting the results.
The OEM is busy detanking the transformer to check the extent of damage.
Do you think the drained oil with high DDBS and DGA can be treated and reused after repair?

 

[prc]

Thank you Pete for that link. Star for you.
powerzizo, When you added passivator in two transformers why you did not add it in the other two units. What was the amount of Ergamet-39 that you added initially? 100 ppm? Check the DBDS content in the working transformers with passivator.

The failures from corrosive sulphur is always like this- sudden and with out any indication in DGA. This cannot be forewarned by any of the diagnostic tests known so far. Normally failure will be inside the windings as inter turn faults at the top or bottom of the windings. Interesting point is these failures are seen only in GSU, Reactors or HV DC transformers.

Of course you can reuse the oil from failed units. But you must add passivator before that. But in case you have to use all new windings and insulation ( in case failure is from corrosive sulphur) why don't use good new oil. I had to handle such a case 10 years back. With Ergamet added I could arrest such failures noted in a batch of transformers.

When IEC 60296 -2003 Specifications for Transformer oil was revised in 2012 as ed4.0, under Table 2 -Specifications, it was added "DBDS - not detectable " with test method added. Also clause A2 of standard has to say " The only compound shown so far to be a potential Cu2 S forming agent and to be present in significant amounts in transformer oil is DBDS .Most oils found to be forming copper sulphide contain this substance"

 

[ScottyUK]

The failures from corrosive sulphur is always like this- sudden and with out any indication in DGA. This cannot be forewarned by any of the diagnostic tests known so far. Normally failure will be inside the windings as inter turn faults at the top or bottom of the windings. Interesting point is these failures are seen only in GSU, Reactors or HV DC transformers.

That's an interesting observation. Why does the failure occur so rapidly and without any forewarning? Also the last sentence is intriguing - any idea what the unique design features might be that make these types especially susceptible?

 

[prc]

The failure occurs from inter turn faults. Copper sulphide gets coated on inside layers of paper taping over the copper conductor -color aluminum grey to rain bow colors. This reduces the inter turn dielectric withstand strength and suddenly with out any heating or PD, breakdown occurs. More than 100 very large transformers and reactors of all make, around the world failed during 1995-2005.Failures were only in units with 'superior' oils,highly refined oils with sealed oil preservation system. These failures were arrested by adding metal passivators in such oils.
Why failures in units with constant current loading -I don't know. I could not find any logical explanation anywhere.

 

[ScottyUK]

Thanks prc, my little bit of learning for the day.

 

[powerzizo]

Thank you again for all.
The two faulty transformers were on the same site, the staff in charge could not confirm that passivator was added. Oil sample from faulty Transformer will be analyzed for corrosive sulfur and passivator content.
Inspection by OEM for both transformer revealed burning cables between the HV windings and the tap changer.The OEM started replacing the cables in the first transformer which already detanked , they told they can repair the second one without skidding out for detanking.
ScottyUK. Would you please explain how to disable tap changing, you mean just not to change the step by OLTC control or we have to physically disconnect the leads ?Apology for my poor knowledge.

 

[ScottyUK]

Physically link the neutral tap (assuming a reversing tapchanger) to the bushings, thus taking the entire tapping winding out of circuit. The tap winding is still energised but carries no current. For a generator transformer this is a fairly drastic step because the unit will struggle to reach its full reactive capability without the tapchanger to assist.

 

[powerzizo]

Hi ScottyUK and All.
This modification can be done from the tap changer side during shutdown or we will need to drain the oil and do some works from inside the transformer?

 

[prc]

powerzizo, are you planning to isolate the tap changer leads from service ? Or you want to isolate the tap changer completely? If so it can be decided by only OEM or by an expert after checking the connections physically. There are several alternatives available for designers so it will be difficult to make a guess. In some simple designs it will be possible to isolate leads by selecting particular taps only. In some others it will be forming loops at tap selector terminals and each tap is taken from individual loops. Then isolating leads is not possible. All this requires oil draining and change in internal connections.

 

[powerzizo]

Hi All,
Apology for late update. The transformer repair company has completed the replacement of HV leads to tap changer.according to them the root cause of burning the leads is uneven number of strands in each cable 7 and 14 in addition to poor crimping method.
The trans tanked after repair and moved to its bay.kill filled but unfortunately the Low freq heating machine had breakdown and no other one available. The repair company initially scheduled 15 days for drying with LFH but now mentioned it will take one month because no LFH.is where any other option to expedite the drying.can we back energize the transformer from HV 220kV to work as step down with LV side CB open just to have no load current,if this is possible,should it expedite the draining?