Short circuit calculations (network protectors)

Status
Not open for further replies.

lielec11

Senior Member
Location
Charlotte, NC
We are reviewing a short circuit study and have received some push back on the results. The facility has three (3) MV feeds that step-down to 480V via three (3) 2500kVA transformers. All three of the 480V feeds then terminate in a common switchgear via three (3) 5000A MCBs (the drawings show fused/switches but the shop drawings actually have breakers).

The issue is with the withstand rating of the network protectors, which are close-coupled to the secondary of each transformer. They are only rated for 60kA, but due to all of the feeds terminating in a common switchgear, the fault currents from the other two services flow back into each NP. Therefore, the fault current at each NP basically triples from ~49k to ~145k. The study indicates they fail. We agree with this analysis and the results. However, the manufacturer and the EC believes the study is wrong and that the NPs are actually fine.

My question for this community is do you agree with the results of the study? I have attached an image of the original one-line diagram, as well as a screenshot of the SKM model as well. Any other insight would be greatly appreciated.
 

Attachments

  • NP eor dwg.PNG
    NP eor dwg.PNG
    187.4 KB · Views: 32
  • NP ssc.PNG
    NP ssc.PNG
    352.4 KB · Views: 29

paulengr

Senior Member
Why would they NOT believe it triples? A transformer as seen from a common voltage is just an impedance. So if we have two impedances in parallel, we get half the impedance thus short circuit doubles. Or in your case with three parallel impedances it triples. Only if the utility source becomes the limiting impedance would you expect a different result. At that point you get diminishing returns.

This is a common problem with network protectors and MTM schemes with the ties closed. You can add impedance or open ties or if you have lots of money, you can use triggered current limiters.
 

lielec11

Senior Member
Location
Charlotte, NC
I have NO idea but they are having trouble believing us for some reason, it's getting out of hand. I had to reach out to my industry colleagues here for a sanity check 😂.
 

JoeStillman

Senior Member
Location
West Chester, PA
Your SKM printout is very illustrative. Note the currents in each segment depend on whether the fault is upstream or downstream. A fault on any of the NP1, NP2, NP3 busses is 3x a single xfmr, but it flows from two different directions - 1x from upstream and 2x through the NP. The most any NP has to carry is 2x a single xfmr when the fault is upstream of it because the current is coming from the other 2 sources via the incoming bus. So they ARE over-duty, just not by 3x - only 2x.
 

lielec11

Senior Member
Location
Charlotte, NC
Your SKM printout is very illustrative. Note the currents in each segment depend on whether the fault is upstream or downstream. A fault on any of the NP1, NP2, NP3 busses is 3x a single xfmr, but it flows from two different directions - 1x from upstream and 2x through the NP. The most any NP has to carry is 2x a single xfmr when the fault is upstream of it because the current is coming from the other 2 sources via the incoming bus. So they ARE over-duty, just not by 3x - only 2x.

What if the fault was at the NP itself inside? Wouldn't that see 3x?
 

JoeStillman

Senior Member
Location
West Chester, PA
What if the fault was at the NP itself inside? Wouldn't that see 3x?

Line side and load side faults will be different. Remember that SC duty is the current the device must interrupt. For faults on the line side, interrupting duty is 2x. Faults on the load side will be 1X. Faults in between line and load will actually be 3X, but you will have already let the smoke out.

I don't know of any O/C device that is designed to protect against its own internal faults.
 

Hv&Lv

Senior Member
Location
-
Occupation
Engineer/Technician
I would assume with that large a network there are 50/51,67 on all the secondaries, maybe an 87 relay schemes? Possibly an 87 for each XF?
 

lielec11

Senior Member
Location
Charlotte, NC
I would assume with that large a network there are 50/51,67 on all the secondaries, maybe an 87 relay schemes? Possibly an 87 for each XF?

We don't have that information but regardless the short circuit current is what it is. IEEE doesn't let you use current-limiting devices in the calculation results.
 

Hv&Lv

Senior Member
Location
-
Occupation
Engineer/Technician
We don't have that information but regardless the short circuit current is what it is. IEEE doesn't let you use current-limiting devices in the calculation results.
I know. I agree with the study.

I was just curious.
 

jim dungar

Moderator
Staff member
Location
Wisconsin
Occupation
PE (Retired) - Power Systems
However, the manufacturer and the EC believes the study is wrong and that the NPs are actually fine.

Are the people, believing it is wrong, trained in electrical power systems analysis? Are any of them registered professional engineers? Many manufacturers have an 'independent group' which does their 'stamped engineering' work and is not part of their manufacturing operation, maybe you could get them involved.
 
Status
Not open for further replies.
Top