Dissolved Gas Analysis on Oil-Air type transformers

[shockking]

I'm reading IEEE C57.104 and reviewing the DGA test results for a few dozen pad mount oil-air (OA) type transformers (mineral oil). These transformers are are only sort of sealed, they have a bleed valve that opens if pressure gets high enough. It seems to me that if there were any "fault gases" (CH4, C2H2, etc.) that they would bleed off eventually and not necessarily stay dissolved in the liquid. Is that not the case?

 

[Elect117]

I think taking samples over time (for us it is year over year) and comparing the results is really what you are looking for. Yes, they probably bleed a little. That is also why testing regularly is a part of the analysis. If you have access to historical, you can see the values increasing over time.

 

[shockking]

Thanks, makes sense. For what it's worth, I was mistaken and most of these reports are for sealed devices, which are a relatively small portion of our fleet. There's only a couple reports for OA type units, which were an exception. Of course, now it's got me wondering if we should start testing the rest of the fleet that are all OA...

 

[Elect117]

We DGA test our 66kV transformers and advise HV customers do theirs on a cycle.

I think we impedance test most other transformers and oil level check on a maintenance cycle. I don't think we send samples out for them though. It is usually a cost benefit type of situation.

If you don't have spares on hand and you need to plan the replacement of the transformer a year or two in advance to have one custom made, then I would probably include testing.

Most pad mounts are easily replaced. Depends on the type and circuit configuration.