So I have been working through some scenarios for Arc Flash on padmount substations (particular the LV side) and I'm struggling with a few things I'm hoping to quiz the brilliant minds I've already seen in this forum.

One thing I'm noticing is the upstream protection device I'm using is the HV transformer fuses have a fairly slow clearing time (and that's just for the 3-ph fault, I know the single-grnd would be much slower due to the dy current conversions)

For the fault level I'm using infinite bus method using a default 4% transformer impedance, then performing arc flash calc on 15 scenarios, all with 1% increments to 85% of the Fault level. (Is this too conservative? I was thinking 90% would probably suffice, the worst case is usually the 85% scenario)

To date, I've used the default 1584 box measurements and conductor distance.
I have seen through these forums that a NFPA paper suggest 4cal/cm2 for LV side on a padmount substation.

I'm curious, what is the conductor distance in this case, would it be the distance between the LV busbar phases and would the dimensions be much bigger than the default 1584 guidelines? Thanks

Andrew