Strange phenomenon at a power transformer

Oh I don't like the sound of that, it sounds like a flash-over of some sort during transformer inrush at switch on. Where does the flash occur - on the circuit breaker or the transformer?

You really should find out for sure what is going on before running this system as your equipment may be damaged (faulty manufacturing, transport damage, damaged winding or bushing, etc) and could fail prematurely or when the transformer output load is connected.

Can you supply more information including posting a picture indicating the location of the flash, and the transformer VA rating.

Dear All,

There is a flash at the transformer every time at the moment it is energized.

We checked transformer bushings, and found no damages.

The flash runs between flanges of the top over and the transformer tank.

There is a gasket between both flanges, so the top over and the tank are not directly touched.

There are around 60 sets of bolt and nut to tighten the flanges. These bolts and nuts are of stainless steel and insulated from top or tank flanges. The Insulation is done with insulation sleeves and washers. Transformer vendor says, the insulation for the bolts are to avoid corrosion between steel flanges and stainless steel bolts.

The top over and the tank is bonded with two bonding wires at two locations.

Does anybody know the reason of this phenomenon?

Regards

Based on what you are saying it sounds like the top and bottom of the transformer tank are completely insulated from each other (like the two plates of a capacitor) and you are getting a flash-over.

First, are you sure the bonding wires form a low resistance path between the top and bottom of the tank? No corroded metal or broken wires breaking the circuit?

Dear all,

Yes, top and tank are insulated except for two bonding jumpers.

We measured the resistance, with a multimeter, between top and tank, and result is nearly zero Ohms.

To remove the insulation from the bolts may be a solution, and we are preparing to do so. However, we would like to know what is cause of the flash over between flanges.

Regards

I would suspect the bonding attachments

Clean all mating points of the bonding system, including retapping the bolt holes.

It is a possibility that the bonding is not sufficient to conduct all of the charging current. After all you replaced the connections of quite a few bolts with only 2 conductors while creating a large capacitor.

Please check the transformer body earthing. Is it connected properly? If connected properly please check the earth resistance.

As another measure please enrgise the transformer from 11kv side keeping the 33 kv breaker OFF. Observe any spark at the flange.

Secondly the 33kv insulator may apparently look Ok but may be there a small creapage which might be giving way at the time of energisation due to inrush of current during switching.

You have said both the transformers are identical.Have u removed the insulation of the bolts in the transformers where there is no flash?if not removing may not be a solution.What u say as flash is an arcing and needs more detailed investigation and I think the manufacturer should be involved .

Dear all,

Thank you everybody.

Both transformers are identical, they are with insulated bolts for top and tank flanges. We have not removed the insulating material from the bolts. Transformer manufacturer has agreed to supply replacement bolts and we will replace the bolts with new ones but without insulation.

After the bolt replacement, top and tank flanges will be electrically connected at many points where uninsulated bolts are installed. Induced current in top and tank can freely circulate without producing voltages, we guess, there will be no more flash.

However, we still have a question. Where or how the voltage, which causes a flash, is induced between top and body flanges at the moment the transformer is energized.

Regards

Reply to Akihito Shigeno

Could it be contact bounce when you close the breaker?

I'm grasping at straws here but I have an idea not mentioned by anyone else here. The brief higher than normal inrush current that charges the transformer primary might make the transformer housing act as an additional transformer secondary wiring itself. The chassis could then produce sufficiently high voltages across the bolt insulation to cause a spark. Try different phase wirings as an experiment.

Best of all for the group would be a posted picture of the arc site.

did you do pf test on hi side, to top, and hi to tank? in addition to all xfmr tests prior to commisioning the unit? what are results?

Hello Akihito,

Few questions,

a) What is the vector group of the transformer? is it Dyn11?

b) If it is Dyn11 , has the neutral of the low side grounded?

c) Is the body of the transformer bonded to a good earth?

d) I would recommend isolating the transformer and conducting insulation test on HV/lv windings.

cheers.

Thank you all,

a) What is the vector group of the transformer? is it Dyn11? Yes, Dyn11.

b) If it is Dyn11 , has the neutral of the low side grounded? Yes, LV side neutral is grounded.

c) Is the body of the transformer bonded to a good earth? Yes, Transformer tank is grounded to the grounding network.

d) I would recommend isolating the transformer and conducting insulation test on HV/lv windings.

Hi-potential test for both HV and LV side was conducted before power cables are permanently connected.

I will report you the result after energizing the transformer without insulation at bolts.

Best Regards

I have been watching this discussion & have extra questions

Can you provide details of the differences in the transformer circuits.

Are there surge arrestors on one transformer but not the other, are there different circuit breakers?

What is the length of line feeding each transformer & from the CB to the transformer.

Are the two transformers at the same site or are they in different locations?

Are you energising on-load or no-load?

The results of the bolt changes will be good to understand but if it corrects the problem - then you have to question if the other unit will develop the same problem over time.

Dear all,

This is the final report for this subject.

We replaced insulation at the bolts and nuts with washes called "contact washers".

With the contact washers, transformer body and top flanges are bonded at many places (more than 50) around the flange. There is no more flash after that.

I still have the original question, where the voltage which caused the flash came from. Anyway, thank you very much.

Regards