Transformer Temperature Problem

Last time I checked the primary was last August and it was within specs. Keep in mind that the temperature problem first bacame evident during June, while the capacitor exploded recently. What troubles me the most is that the other transformer has a much larger load with many more AC motor drives, without presenting any alarming temperature issues, despite the fact that the same maximum voltage (415~420V) is measured during the night.

If both transformers are producing an increased voltage at the same time, then that is likely caused by an increase in primary voltage. This would be consistent with a reduced night-time load on the power company grid. The capacitor explosion could be the result of gradual degradation culminating in catastrophic failure. What is the temperature behavior of the capacitor banks?

I'd advise checking the power quality. The THD reading is quite high, so that is a warning that either the transformer core is saturating (not likely with the load given) or you have a load generating harmonics. All it takes is one bad harmonic filter on a big motor to cause major headaches. I once talked to the maintenance tech at a bowling alley, who told me they had problems with erratic pin setter operation that were traced to harmonics on the power line generated at a foundry a mile away across the river!

If you find harmonics, you may be able to identify the source by correlating changes in harmonic content with varying motor speeds. Do you have the same distortion without the summer loads?.

The summer loads are basically two large cooling units 700A total, without AC drives, and yes the distortion is the same.

I cannot check the temperature behaviour of the capacitor banks because they are in an enclosure right next to the transformer.

I am thinking of removing all capacitors connected directly to the transformers secondary and only use the ones connected on the low voltage field which I can disconnect without the need to disconnect the transformer, given that my PF remains high. As for the harmonics I have contacted a local company specializing in power quality measurements to assist on the matter.

Temperature rise may be caused by several reasons...

But before you do anything check that the temperature indicator is correct.... (nothing like trying to find a non existent fault)..

Check also that the cooling system is functioning... or that the airways are not blocked.

2) If the TX is not overloaded check for harmonics in your circuit.

or

3) you could have an internal fault in the TX...

The cooling system is 6 fans 1800m3/h total supply located at the bottom of the transformer. One of them was broken and replaced without any noticeable effect on the temperature. A possible fault of the temperature sensor has also crossed my mind. But the transformer has three (one in each winding) and at high temperatures all three have the same indication