Power Transformer Inrush Calculation

[Robert789]

I'm looking at the existing protection for a bank of three single-phase, 2500 KVA, 44 KV to 13.2/7.62 KV transformers, connected delta-wye, at a rural substation.

Protection for this transformer bank was originally supplied by 125 Amp Southern States Type PF fuses. About 2 years ago (before I began working here), protection was changed to 100 Amp Type PF fuses at the recommendation of an outside engineering firm.

The 100 Amp fuse curves are to the right of the traditional inrush points, 12x FLA at 0.1 second and 25x FLA at .01 second. However, the sofware I'm using (Milsoft LightTable) plots a transformer inrush curve instead of just plotting inrush at two points. This curve seems to indicate that proper coordination is not achieved between between inrush and the fuses in the range of about 200 thru 500 amps (2x thru 5x FLA).

I'm not finding any documentation describing the software's inrush assumptions, but it does appear that the curve may be a creation designed to cover both inrush and cold-load pickup. The curve passes through 25x FLA @ .01s, 12x @ .1s, 6x @ 1s, 3x @ 10s, and 2x at 100s.

At 2x FLA, fuse minimum melt is about 20 seconds and 'inrush' curve is 100 seconds. At 3x FLA, fuse min melt is 5.25 seconds and 'inrush' is 10 seconds. At 6x FLA, fuse min melt is about 1.2 seconds and 'inrush' is 1 second.

So, here's my attempt to get some of you to help me make a judgement call... Given that the minimum melt curve passes to the right of the inrush curve as plotted between 2x and 5x FLA, would you replace the 100 Amp fuses with 125 Amp fuses or just let it go. Curves for both fuses lie entirely to the left of the transformer damage curve. Peak load on this substation is approximately equal to the OA rating of the transformer bank.

Thanks,

Robert

 

[dpc]

Based on experience, if you're OK at 12X @ 0.1 sec, and the fuses have not actually melted on inrush, I'd probably leave well enough alone.

Transformer inrush is notoriously difficult to predict and seems to vary even for very similar transformers.

BTW, when looking at transformer damage curves for this delta-wye transformer, be sure to consider the impact of a secondary line-to-ground fault - the damage curve needs to be shifted to the left - 58% of the "normal" damage curve. You're probably aware of this, but just in case....

 

[prc]

dpc,can you explain your last sentence little more. I did not follow.Are you considering line current vs phase current?

If the FLA current of a trf is 100 A,what rated fuse is to be used? 100 A or higher rated?

 

[Robert789]

prc,

For a line to ground fault on the secondary side of a delta-wye transformer bank, the fault current will be split between two phases on the primary side of the transformer.

Fault current seen by each fuse will be {1/sqrt(3) = 0.58} times the fault current in a wye-wye connection.

Robert

 

[dpc]

As Robert789 says, the issue is the transformer damage curve for through faults. For a delta-wye transformer, the line current on the primary side is only 58% of the winding current in the secondary side for a ground fault. So each fuse sees only 58% of the current that is actually flowing through the secondary winding. To account for this when determining if the primary protection is providing full protection against through-fault damage, the damage curve can be shifted to the left on the TCC by a factor of 0.58.

The IEEE Red Book covers this pretty well.

 

[prc]

dpc,it seems there is some confusion.The primary line current will be 100% and phase current through windings will be 58%.So fuse shall be for 100 % and winding withstand on primary side will be 58%,but secondary winding current will be 100 %.So trf overall withstand will remain 100%.Am I wrong ? I will have a look at Red book.

dpc,can I have your expert view on my second question?

 

[dpc]

What second question?

The fuse will see only 58% of the current that the secondary winding is seeing so it will not trip as quickly as might be believed by simply looking at the standard damage curve. By shifting the damage curve, this is compensated for.

 

[prc]

dpc,let me beg to differ.Fuse if put in line will see full 100 % current.But phase current in winding will be 58 % .That is normal with the delta connection.

My second query was if transformer rated line current is 100A ,what should be the rating of fuse-100A or higher ?

 

[dpc]

We are talking about line-to-ground faults on the secondary side. Draw it out or look at the Red Book. I'm obviously not explaining it very clearly.

I thought already answered that question. If the 100 A fuse clears 12X @ 0.1 sec and has never caused a nuisance trip on inrush, I wouldn't bother taking an outage just for changing it.

I did not really evaluate the adequacy of the 100A fuses to carry load current. That's another issue.

 

[waross]

Hi guys;
I think the confusion stems from looking at a line to ground fault on a wye secondary and the resulting primary current. Only one phase is loaded with a wye secondary ground fault. Hence the 58% factor.
With a three phase secondary fault, the 100% factor is of course correct.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[TZilla]

Minimum-melt curves for power fuses are typically plotted at 25C with no initial load. Although on paper it may appear that the 100A fuses are okay, but when the fuse element is heated due to load current or higher ambient operating temperatures, the fuses will respond sooner. This is most likely to become an issue when you have a momentary on the 44 kV system, and the fuses need to accomodate the magnetizing inrush current when they pickup the transformers.

Some software programs allow you to put a "setback" allowance on the minimum-melt curve to see how it's really going to respond. Also the 12X multiplier is pretty conservative, and it's likely to be much smaller due to source impedance and transformer size.

In short, it appears that the 125A fuses were the proper size for this application.

 

[jghrist]

I have seen the inrush limits at 6x, 3x, and 2x referenced before, and used in software. The inrush values are probably conservative for larger transformers. You might have a 100A fuse blow on inrush, but probably not. I wouldn't take an outage to replace the fuses.

Check the 125A fuse against the transformer damage curve. If it is OK, then have some available to use as replacements if the 100A fuse ever blows on inrush.