xfmr loading

[958]

Our company recently lost a 37.5MVA transformer. Upon investigation,the paper was ruined , piles of rust was found on top of the core we think due to a leaking pressure relieve valve.There had been a steady increase in moisture and dissolved gases over the last two years. Nobody can point to just one thing.The company hired to test and evaluate the xfmr,said that we had been overloading it over a prolonged period of time. The xfmr has been in standby mode for the last two years ,thus allowing the moisture settle in. My question is this. The average load on this xfmr had been about 23MW. Could this really overlaod it?
I have always understood that 80% of full load rating was a good, safe figure to use in loading.
This is a GE (1972) forced air and pumps,with a LTC.

 

[DanDel]

What was the PF of the 23MW load?

 

[patrick7]

Also, what is the rating of the transformer
high and low side voltage? Also, what are the actual voltages impressed on the trasformer?

PF of the load would be very helpful.
also, is the transformer physically located in an area that does not allow for ample airflow for cooling? What are the historical reading for operating temperature? Did your testing show any thermal degradation of the winding? Since you said that the paper was ruined.

If this transformer is oil filled; witch is what I'm assuming. How do you get a pile of rust on top of the core if the oil level is correct?

Since the ave load comprises both higher and lower values the possibility that the transformer has seen overload conditions exists. The exact amount of overload and the duration of the overload all take their toll on transformer live. The loss of life during each overload is cumlative.
and non recoverable.

The transformer has been in standby mode for several years. Does that mean the transformer has been deenergized?
The tests that show increasing moisture and desolved gases indicate that there may have been a turn to turn fault within the winding that continued over a period of time until finally the short was sufficient to trip the equipment out of service.

 

[958]

Sorry about lack of info on "xfmr loading". The xfmr is rated 115KV high side and 35.4KV low. The average PF was around .9 lagging.The oil level was fine.The xfmr has been setting energized for two years,with little or no load on it. It is located in a power plant switchyard with fairly good air flow. I am not a engineer, I am a substation foreman of 30yrs. Thank you. 958.

 

[wfowfo]

I would think the load on the unit was less of a factor than the moisture. You spoke of paper degradation.....was this evaluated from the gas analysis? Was the nitrogen blanket kept positive pressure?

 

[NorthSlope]

Overloading would NOT account for rust on the top of the core. That's a function of water - only! 23 MW on a 37.5 MVA transformer is well within loading norms. How lightly loaded, or unloaded, lets moisture "settle in" to a sealed tank is beyond me.
Depending on what "paper was ruined" means - it seems to me that the transformer leaked to atmosphere no matter what the loading.
I'd find a new transformer testing outfit.

 

[edison123]

It is the end-of-life (34 years) failure helped (or not helped) by water entry. Your loading had nothing to do with the failure.

I agree with northslope - get another testing firm.


* Basically, I would like a full-time job on part-time basis *

 

[DanDel]

It sounds like there was definite water intrusion. Overload conditions may have existed at some time in the past, but shouldn't have been p[resent at the 23MW, PF of 0.9.

I'm assuming the transformer faulted internally when you say you 'lost' it; this would cause visible internal damage to the windings. Was this present?

I'm still curious about the paper issue. Typically, with this kind of fault, a furfural test is done on the paper to determine what extent of overload the transformer has been exposed to in it's history. Was this done?

 

[Zogzog]

I agree that water intrusion was probally the cause but I wouldnt rule out corrosive sulfurs in the oil. ASTM 1275 has a new alternate test method (B) for testing for corrosive sulfurs, wouldnt hurt to get an analysis done just to be sure, not that it will help you now.