Our AT3 -400MVA autotransformer is mounted at 400/220kV SS and it is in servise with two another autotransformers AT1(1983) and AT2 (1987) with same rated Power

Two years in service we did not take any oil samples for gas analyzing, but at 24 march 2006 we did send the oil sample to the laboratory and from them we have this results:

Dissolved Gas in Oil Analysis of transformer oil

AutoTransformer: AT3
Manufacturer: Siemens-Koncar
Date of production: 2003
Power: 400MVA
Voltage: 400/220kV
No load Tap changer 3 pos.
Cooling system: OFAF
First group ventilators aktivation at 530C
Second group At 63 0C

Temperature of oil (top) when the samples taken was T=340C

The results of Oil analysis:

Hidrogen H2 303 ppm
Methane CH4 795 ppm
Acetylene C2H2 17 ppm
Ethylene C2H4 1139 ppm
Ethane C2H6 151
Carbon monoxide CO 176
Carbon Dioxide CO2 345 ppm

I am asking to you: It is our transformer in dangers, wat can we do to save Autotransformer. It is problem indicated because of cooling system , because the first group ventilators activate when the temperature oil is 53 0C. that mean the pump will activated with ventilators.
The firs group has 9 ventilators and second group 9
With respect
Gazmend Kabashi
Transmission and Dispatch maintenance engineer

Dis your laboratory give you a diagnosis ?

Try this site for some assistance

http://www.nttworldwide.com/diag.htm

I agree, ask your oil lab. There are quite a large number of expert programs that will check Roger's ratio's Dorenburg ratio's to give a recommendation. Complete analysis would also look at the trends.

A simpler method, not so sophisticated or exact is the key gas method. Roughly speaking, look at the highest gas and see what is tells you. Your highest gas is ethylene. That is a hot metal gas indicating hot metal in presence of oil. I would suspect a hot internal connection. Tap changer contacts are always a suspect. You have NLTC contacts and also I believe under load contacts (are under load contacts in the same compartment?). Best option is draindown inspection but not always practical. Next best thing, you might briefly take the unit off-line, change the NLTC position, put back on-line and check for change in gassing rate. (while the unit is off-line you might want to also check resistances and ratio's).

Also a long-shot check but possibly beneficial would be infrared survey. We have had success identifying hot contact in separate-comparment LTC unit based on LTC unit being hotter than main tank and also showing abnormal temperature gradient bottom to top.

I don't see indication of cooling malfunction. That would be indicated by predominance of CH4 and possibly H2. Also 34C at time of sample is fairl cool. Can you explain more why you think you have a cooling system problem?

I agree with Pete on this as well.
Have a look at the site above, as they have an on-line gas analysis program ( free ).

I also think there is internal heating/overloading, but without an excess of acetylene, it does not indicate an highly active arcing condition. I would focus on internal connections and tap changer contact condition, which as Pete said requires a drain and/or de-tank for inspection.

For monitoring, as well as thermography, try using a contact ultrasonic probe on the transformer shell ( if safe to do so ) and around the NLTC mechanism. Listen for a cracking sound for evidence of internal contact arcing. Compare sound levels and nature of sound with a similar unit under similar operating conditions.

Good luck