Hi again - would a 2.5MVA transformer count as a larger unit when it comes to varnish outgassing? Also - with solar duty, the transformer only operates under load for a few hours a day - could that extend the outgassing period?You don’t bother with the atmosphere on top. You grab an oil sample and test the dissolved gases. It’s cheap. It requires a special glass syringe to do it. This will tell you what is going on inside.
There should be pressure or vacuum on a transformer. During the day and as they warm up from operation you get a little pressure. If there is no pressure then the transformer is open to air. This allows moisture in which attacks the transformer insulation and structure. As that breaks down it creates acids that attack and destroy the oil. Moisture and oxygen need to be avoided at all costs with oil filled transformers.
With corn oil (FR3, bio temp, or whatever they call it), it gasses a lot more than traditional mineral oil. Also the varnish system on larger transformers will gas a lot as it runs for a couple weeks.
The bushings can sometimes be a problem. It’s just a draw lead nut pulling in the oil side with a rubber seal under the bushing/terminal interface. There is a certain amount of torque you need but not too much so sometimes these get bumped in shipping or are too loose abc leak.
Again, not my area, but a couple of things came out from that (and they may be obvious to others)--We have two, identical. hermetically sealed, mineral oil filled, 2.5MVA transformers, each fitted with a DMC protection relay. One transformer is functioning perfectly, the other is exhibiting the generation(?) of gas. The gas pressure builds up in the DMCR, forces the oil level in the DMCR below the oil-level trip threshold, and the transformer trips.
Great questions zbang. Since the oil level returns to its recent level when gas is released from the DMCR it must have gone somewhere and compressing an internal gas bubble makes the most sense - since, as you said, the oil is non-compressible. So, either air is getting into the transformer to form a bubble or is being generated in the transformer to form a bubble or the bubble is so large it's gradually bleeding out via the DMCR. So far, I like paulengr's comment about varnish outgassing. This transformer has had windings repaired and was refilled with oil so we can't say whether T1 and T2 have the same amount of oil in them.Again, not my area, but a couple of things came out from that (and they may be obvious to others)--
If they're properly hermetically sealed, no outside gasses should be able to get into the transformer. Does the gas pressure go negative when it cools after venting? Or... the gas has to be coming from somewhere, so if it's not being allowed into the transformer, it must already be there and is working its way out.
Oil is considered non-compressible, so gas pressure itself could not force the oil level down. It could force sidewalls out, lowering the level, but that ought to be unlikely. OTOH, if there are gas pockets somewhere in the transformer that should contain oil, that gas could compress and then the overall level would drop (like the classic u-tube manometer). With the transformer out of service, it might be interesting to pull a vacuum on the tank and see where it stabilizes and how the oil level changes.
(Were the transformers stored full? Was the oil changed before going into service? Did both transformers get about the same amount? Also gets back to the IR scan thing.)
We do not know the actual pressures I'm afraid, however the DMCR is set to trip at an overpressure of 0.3bar. We have not experienced an over pressure trip so farWhat is the XF pressure cool and at operating temp?
OK - gotcha - yes, that has been done. I've attached the DGA results. Nothing of any consequence was found.
Hi again - would a 2.5MVA transformer count as a larger unit when it comes to varnish outgassing? Also - with solar duty, the transformer only operates under load for a few hours a day - could that extend the outgassing period?
I have never ran an SFRA on a transformer that small..It can go on for months and be on and off. 2.5 MVA is kind of borderline. They all gas just some more than others. The big thing to watch is DGA results after it runs for a week or so. If the gassing does not slow down and/or you see an increase in dissolved gasses that indicates a problem with the transformer. The proportions of gases indicate what is happening. As an example the transformer coils could have been mounted too close to the tank wall or something shifted in transportation or maybe someone forgot to tighten a bolt internally. This will cause overheating of just one spot that will show up in the DGA test but you need an operating transformer to allow the oil to circulate and enough run time to allow gas to form. The standard start up tests (Megger, TTR, Kelvin bridge) just give you an indication of whether it was wired properly and whether it is safe to energize.
Things happen. I used to do a dozen transformers a year of that size. About one out of 20 had a defect, often something minor. Out of those tap changers are notorious for problems and the cheaper pad mount designs that utilities use with squarish coils often have hit spots where the coils get close to the tank walls.
Again, not my area, but a couple of things came out from that (and they may be obvious to others)--
If they're properly hermetically sealed, no outside gasses should be able to get into the transformer. Does the gas pressure go negative when it cools after venting? Or... the gas has to be coming from somewhere, so if it's not being allowed into the transformer, it must already be there and is working its way out.
Oil is considered non-compressible, so gas pressure itself could not force the oil level down. It could force sidewalls out, lowering the level, but that ought to be unlikely. OTOH, if there are gas pockets somewhere in the transformer that should contain oil, that gas could compress and then the overall level would drop (like the classic u-tube manometer). With the transformer out of service, it might be interesting to pull a vacuum on the tank and see where it stabilizes and how the oil level changes.
(Were the transformers stored full? Was the oil changed before going into service? Did both transformers get about the same amount? Also gets back to the IR scan thing.)
If you mean was the sample pulled before putting the unit in service or after it had been in service for some time it was the latter - the unit had been tripping for just over a week when we pulled the sample. If you mean was the unit energized at the time of pulling the sample the answer is no - the unit was disconnected when the sample was pulled.Did you pull it after energizing or before? The results will be meaningless if you pull them on a deenergized transformer.
Did you fill the transformer up with oil?No low-oil trip today - project site is 10 hours ahead of US
Not yet with the oil - hopefully tomorrow morning - will let you know.Did you fill the transformer up with oil?
who repaired the unit?
I have never ran an SFRA on a transformer that small..
would it show anomalies as well as it would larger 10-50MVA units?
If you mean was the sample pulled before putting the unit in service or after it had been in service for some time it was the latter - the unit had been tripping for just over a week when we pulled the sample. If you mean was the unit energized at the time of pulling the sample the answer is no - the unit was disconnected when the sample was pulled.
I require an SFRA before and after any transformer is moved, and we do a full battery of test every 5 years.Yes but I haven’t ever really used one. It’s one of those gadgets they push for 10 MVA and over but I question the value at any size since I’ve never seen one show anything the traditional tests don’t show and can’t seem to identify anything useful.
Good to know the sample is considered valid - thanks paulengr. HvLv suggested varnish off-gassing - that would be a nice conclusion - just not sure I understand the chain of events that lead to the DMCR gassing up and tripping. The DMCR has a tube extending about 4 inches into the transformer - presumably below the oil level - why should gas decide to go up the tube??Ok so valid sample. It’s showing you that there is no hot spots or arcing and given low moisture you don’t have any problems with seals. It’s just normal
Off gassing.
Because the oil level is low??Good to know the sample is considered valid - thanks paulengr. HvLv suggested varnish off-gassing - that would be a nice conclusion - just not sure I understand the chain of events that lead to the DMCR gassing up and tripping. The DMCR has a tube extending about 4 inches into the transformer - presumably below the oil level - why should gas decide to go up the tube??
OK - hopefully we'll add a couple of liters tomorrow. Appreciate all of your suggestion/comments/questionsBecause the oil level is low??
