Could anyone help me to elaborate this description?

"In certain cases, such as for switchboards, it is possible that the fault may occur or propagate to the line-side of the local protective device."

By the way, how do you deal with the arc flash study for a project only contains single phase service? Even it could be a 200A 240V service?

Thanks!

What their getting at is that the plasma in the arc flash is conducting and the arc itself can travel down the bus. If an arc flash is large enough (a.k.a. huge fireball) it could reach the line side of the protective device and continue arcing ahead of the protective device - even if the device itself trips.

This is why the "limiting" over current device (i.e. the one that you consider will trip to end the event) is a device upstream in a separate enclosure. Unless the equipment is designed so the arc does not propagate to another area i.e. arc resistant equipment.

For low voltage single phase, there are still no official equations yet.

Thanks a lot for the response. Is there anyway to calculate the incident energy for this kind of case? How accurate it is?

I don't know what software package you are using, if any, but the one I use has an option to "Exclude Main Device" which allows it to ignore the main bkr/fuse and look at the next one upstream to determine the IE at the bus.

I am using PTW32, what's yours? Do you treat all your buses in this way? Not all distribution panels have barrier between main device and branch breakers.

I am using EasyPower and I usually treat the panels this way.

I do this for all locations except, 'drawout' compartmentalized switchgear using power circuit breakers and MCC's that have their main device in a barriered compartment.