I(sc) calculated based on NFPA 70E Annex D equation [D.2(a)]

I(sc) = {[MVAx10^6]/[1.732*V)]}*(100/%Z)

From my previous post

I(sc) = (1500000/831.36)*17.636684

I(sc) = 31821A or 32kA

Secondary I(sc) from transformer calculations

I(secondary side) = KVA*1000/(V*1.73) = 1500*1000/(480*1.73) = 1806A

I(sc) = I/%Z = 1806/0.0565 = 31970A

So, I'm pretty comfortable with the 32kA value. From the calculation, I know that this is the max short circuit amps for the system. Using the Annex D and equation [D.7.2(a)] I calculated the arching current at the switch gear using a gap of 32 from table D.7.2 and got 16.57kA. Which is the same as some of the online calculators have given me like:

http://www.arcadvisor.com/arcflash/ieee1584.html
So this is the arcing current that can enter the switch gear. The gear is fused by a 3000A fuse (that's all I know about it), so I can't calculate an acceptable flash boundary for this piece of equipement until I know the device opening time is (or use the standard 0.2s [as equation D.7.3(a) does], but the with a time delay fuse this could be longer, or shorter if it's current limiting), but I do know the fused discconnects for the first 4 MCC sections. All 1200A KRP-C-1200 Class L fuses.

Redoing the above arching current for an MCC section (gap changed to 25) the arcing current becomes 17.69kA. The average melt chart shows fuse breakdown in 0.02s to 0.03s (its log scale and a pain in the ass to read/guesstimate). Apparently I read it wrong must have been looking at 1600A fuse. Add 15% for melt time chart times up to 0.03s (found this on the arc flas site as well). Device opening time = 0.023s, at 85% of arc current (15.95kA) time from chart is still 0.03s.

The NFPA book discusses the opening time of a breaker with a saftey rating of 2 cycles, but this is a fuse and a time delay fuse at that, so if the arc is something less that 17 kA the system could very well have a longer arc event.

Distance from the main transformer is less than 50ft, so any voltage drop at this point of fault would almost be negliable (25' of 500kcmil, 6 per phase service is reduces the I(sc) by less than 2kA).

I'm pretty comfortable with my numbers, but I'm not comfortable with the time delay fuse and the NFPA book saying that "Additional time must be added if a time delay function is installed and operating", but no example of how much time and a chart showing 600V line side opening times of 60-120 cycles.