Can anybody comment on the reasons of why we should not use ground fault currents in the IEEE Std 1584 equations?

I believe that the main reason is that the empirical data used for obtaining the curve fitting equations are based ONLY on three phase values of short circuit current. I think that no tests where done with single line ground faults.

Is this correct? Any additional reasons?

Thanks.

Yes, it is that simple. The thought was 3 phase will be worse than single phase and some single phase arc flash events may escalate to three phase. It doesn't always escalate but no way to know for sure. There is language in IEEE 1584 that says if you need single phase, use the three phase model and it will be conservative. (quite conservative)

I think Mike Lang's group over at Mersen did some of this and I've seen some other work done. For most typical equipment where you have the "standard" clearance distances between phases, single phase faults escalate into three phase faults within about 1 cycle (under 0.02 seconds) so there's not a lot of value in studying single phase faults except if the phase spacing is very large such as with overhead non-triplexed lines or actual single phase circuits. Either way, Mersen's results indicate that the difference between single and three phase arc flash isn't much most of the time. Plus even with ungrounded or resistance grounded systems video shows that you get A-ground-C arcing going on (2 arcs) to nearby structures (walls) even as B phase goes through a "zero" and from what I've heard it looks like the higher values for resistance grounded systems were basically an anomaly in the experimental data. So for the most part there just isn't much value in doing this.

Paul:

Do you happen to have a link to that study of Mike Lang's group?

Or an idea how I can search for it?

Thanks.

http://ep-us.mersen.com/resources/arcfl ... elp/?C=usa

Used to be a little bit easier. Scroll down to "Articles and White Papers" and you'll find it all right there. Read "Effect of Electrode Orientation"..., "Effect of Insulating Barrers'. All those "with permission of IEEE" are the articles I'm referring to from Mersen. The insulating barrier one in particular is the most relevant. The specific information that you seek is in there although the major topic of the article is somewhat different. You can also download/purchase it from IEEE directly but this is certainly a less expensive way to get the information!

One of the interesting results buried in those tests is the fact that the incident energy when a single phase arc flash actually occurred wasn't all that drastically different from three phase results which seems to suggest that the idea of some simple relationship such as "divide by 3" is actually a bad idea and that maybe single phase results are so close that there's not much value in determining them.

Personally I'm more keen on the idea of determining where arc extinguishment occurs. If we were indeed able to determine that then particularly with high resistance grounded systems, the arc would simply extinguish 90%+ of the time rather than escalating into a three phase arcing fault.