FR3 fluid giving odd megger and Doble results?

[unclebob]

We have some transformers filled with FR3 liquid that we test annually. Problem is, with the aging of the units, the megger values are decreasing and the Doble results are getting worse. DGA doesn't show anything suspect, except that somewhat hi level of ethane (C2H6). These units are on different customers and were all built by different manufacturers.

The last one we tested Sunday : megger hi /lo+grd = 120 M, lo / hi+grd - 76 M. Doble on hi side : 2%

and five years ago, both megger values were about 1000 M (no Doble results there). I'm waiting the DGA results...

I feel like i'm testing some old PCB's units where results didn't mean nothing.

Anybody having this issue?

 

[prc]

In which country are these in service? Distribution or power transformer? Are they sealed units or with rubber membrane sealing in conservator?
Typical performance parameters with mineral oil and ester filled transformers are given in CIGRE Brochure 761-2019 Condition Assessment of Power Transformers- See Annexure tables.
In general you will get almost double winding power factors with FR3 compared to mineral oil. Thus 2 % on Doble test is equivalent to 1-1.5 % with mineral oil.
At present ABB,Spain is running an experiment with a 400 KV auto transformer bank in field, one phase transformer filled with mineral oil, others with natural and synthetic esters. It is hoped service experience of these units will give reliable comparison figures for different insulating fluids.

 

[jfpe]

Are you running oil tests beside DGA?

I would trend Water content (D1533) and dielectric breakdown (D877).

-JFPE

 

[unclebob]

Thanks for the answers!

PRC, all problematic transformers are sealed ones, no conservator.

JFPE, I'll be back with some figures tomorrow. But I don't recall having any issues with water content or dielectric strength parameters.