strange DGA test 1

[alehman]

A client has recently obtained an DGA test on a transformer (which has not been tested for many years). Ethylene is high (115ppm), hydrogen was 2ppm and acetylene was 0.

I understand ethylene indicates an overheating condition. Hydrogen also is an indicator of overheating, and as I understand, is formed starting at a lower temperature than ethylene. Another clue is the CO2/CO ratio is about 5, which I think indicates oxidizing cellulose. I plan to have it re-tested, but can anyone suggest a reason we would have ethylene but little hydrogen in the oil?

 

[electricpete]

I think overheating hot metal in contact with oil is predicted by key gas method. Looking at the example proportions they show for this fault in IEEE c57.104-1991 we see H2 - 2, CH4 - 16, C2H6 - 19, C2H4 - 63, C2H2 - 0. Ratio C2H4 / H2 = 63 / 2 is within a factor of 2 from your mentioned ratio 115/2. I would expect you should also see C2H6... did you?

One thing to consider is gas distribution and movement in the insulation itself and in the gas space.

Hydrogen is a lightest gas and least soluble in oil (lowest Ostwald coefficient) and therefor equilibrium concentration forces more of it into gas space. Anoter aspect H2 moves quicker into where the other gases take longer to move into gas space. So some scenario's:
1 - If the transformer does a lot of breathing due to thermal cycling, the high H2 removal rate from gas space drags down the amount in the oil
2 - if the gas was recently produced the H2 may have moved into the space and C2H4 not yet.

Another thing I have seen is that if a transformer with high concentration of combustibles is drained, vacuum processed and refilled, I have seen that the heavier gases can come back, apparently coming out of what was left in the oil soaked insulation during processing, but H2 is more mobile and completely sucked out by the vacuum process and doesn't come back.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[alehman]

Thanks for the great information. C2H6 is only 4ppm. Interfacial tension is very low - 17.9. The transformer serves general power circuits in an office building so it no doubt sees significant cycling. This building just changed ownership and there are no records of past life. It was made in 1930.