3PH transformer blows breaker on power up

[gar]

220330-1541 EDT

jim is sending you in the correct direction as one approach.

If you have a large resistive load, like full load, at the time of turn off, then you seem to have a lower probability of a big peak inrush, but not eliminated.

However, it it best and most successful if you reduce peak inrush directly. Meaning you start with a lower voltage, and/or high enough input series impedance to limit current.

.

 

[gar]

220330-1959 EDT

My post numbered #135 suggests adding resistance and impedance in series with each hot phase wire to the transformer. What I suggested is moderately inexpensive. This is relatively cheap to try, and may limit inrush enough so the breaker can hold. After the test the wire can be used for something else.

If this experiment reduces the problem, then it will be necessary to determine the next move.

I think FOG1 is just as capable to ask questions on this problem as virtually most electricians. This is because I don't think most electricians have much of an understanding of ferro magnetic transformer characteristics.

.

.

 

[Jraef]

51kVA Drive isolation transformer current at 240V = 128A, so inrush current when you energize it can therefore be upward of 1280A for a cycle or two. If the 200A main breaker has it's instantaneous trip set at 600%, that would do it.

 

[gar]

220403-2052 EDT

FOG1:

Have you run any recent experiments? If so, then what are the results?

.

 

[gar]

220403-2052 EDT

FOG1:

Have you run any additional experiments? If so, then what results did you get?

.

 

[gar]

220411-1600 EDT

It is now a week and several days since we last heard from FOG1. Has he been banned or what? That he is not an electrician is unimportant. The problem is an important one to study.

The problem he is presenting is a very important one. I would suggest that few electricians have any idea of what these inrush currents look like or how to measure them. Yet these can occur every time a ferromagnetic circuit is turned on. However, for random turn on occurrences the magnitude can vary greatly. Thus, many trials may be required to get an idea of the greatest magnitude.

If you knew the residual flux level and direction, and had a way to turn on the next excitation at the optimum point, a voltage zero crossing, and phase, then nearly every time you could produce a maximum current pulse. So it is either build circuity to control turn on time, or use many random trials to catch the right conditions for maximum peak current.

This problem ( inrush current ) is a condition that should not just be dropped. I have previously provided a link to my scope trace that shows a peak inrush current.

.