Hi, I am running an arc flash study for a system with a 480V main switchboard which is fed from a pad mount transformer. The total bolted fault current was 55.70kA which is the combined output of the transformer and downstream motor contribution. 48.94kA of bolted comes from the pad mount and the arcing current (low arcing tolerance used) was 24.97kA for the current flowing through the pad mount's fuse which is the upstream protection. On another main switchboard which is fed from the same type of pad mount but does not have downstream motor contribution the total bolted fault current was 48.95kA with an arcing current (low arcing tolerance used) of 26.43kA for the current flowing through the pad mount's fuse.

Now my thinking would be, because the bolted fault current from the pad mount is within .01kA of each other, their arcing fault currents should essentially be the same however the difference in the arcing current is 1.46kA. This difference in arcing current causes the fuses to trip differently by roughly half a second (1.2s vs 1.7s). After double checking the modeling and ensuring everything else was identical I went and dug into the 1584 standard and I think I understand what happened. Based on my understanding of the 1584 standard, the arcing current is split between the sources based on their percentage of total bolted fault contribution. So taking 48.94kA/55.70kA for the first switchboard roughly equates to 88% of the total arcing current calculated vs the 100% of the total arcing for the second switchboard. This would explain why the software is providing two different arcing currents for essentially the same bolted current when the only difference between the two is the addition of motor contribution.

This to me would imply that based on having another source, in this case the motor contribution, would affect the arcing current through the fuse from the utility contribution despite the bolted fault current value from the utility not changing. This does not make sense to me as my understanding would be the fuse would act upon the current it senses which would be the current through the pad mount transformer which should be unaffected by motor contribution downstream. Am I wrong for thinking the arcing current from a contributor should be based on the bolted fault current through the contributor rather than a percentage split of the total arcing current? The equation to convert bolted to arcing current uses exponentials making it nonlinear so I'm unsure how that works with a percentage split based on the original bolted fault current.

If anyone could shed light on where I am going wrong it would be greatly appreciated, thank you.

If you can post a screenshot of your model with fault currents and component information displayed I'll take a shot at it.

Sure, I have attached screenshots of the two systems as well as the arc flash values I got. Please note I pulled these lineups from the larger model so the values changed slightly but are still showing the differences described in the original post. The lineups are the same in the utility scenario besides the motor contribution on MSB-01. I believe the arcing current difference is from the IEEE 1584 arcing current split so SKM seems to be functioning as intended. However, I do not understand from the physical perspective why the motor contributing from the loadside would impact the arcing current flowing through the pad mount fuse as they are two separate paths to the fault.

Not sure what happened to your other replies, or mine. Was going to look at this again today but I guess not.