In SKM, what fault current do we use to set OCPDs? Bolted fault or arcing current? I am having trouble figuring out what to use.

I'm not sure I understand your question but I'll try to answer anyway.

Normally, coordination is the priority for setting breakers. You need to full range of the curves up through the maximum fault current available for the devices you are trying to coordinate. SKM automatically cuts the TCC's off at their max fault current.

Now then, if you have say, an adjustable 200A feeder breaker set at "MAX" that feeds a 200A lighting panelboard full of 1P20A breakers and the arcing fault current is somewhere in the short-time range of the 200A feeder breaker and you can lower the instantaneous pickup setting of that breaker such that the arcing fault current is within the instantaneous range of the new, lower, setting then you probably want to be mindful of that and lower the setting. That will likely not affect coordination between the 1P20's and the 200A upstream feeder breaker but will probably dramatically lower arc flash incident energy at the lighting panel.

So for coordination of LV CB, the fault current has no effect on our coordination? We should plot the curves and as long as there is no overlap between curves then coordination is ok?

On another note, why is the trip delay max at 2 secs for SWBD-4? It is clear from the TCC that the ATS FDR and even the GEN BRK trip well before two seconds at the max arcing fault?

Yes and no. Fault current is important for coordination if you are trying to get no overlap in the instantaneous region (often impossible). If the downstream device maximum fault current is much lower than the upstream sometimes it is possible to set the upstream instantaneous pickup higher than the maximum available current of the downstream device. Often it is not. There is often no instantaneous delay capability for breakers that allow you to build in a delay to achieve coordination that way. You are likely to have a lot of overlap in the instantaneous region on any given project. That is just how it is.

I suggest you watch the Easy Power 4-part series on coordination. I know you are using SKM but the Easy Power videos are pretty good at explaining various coordination issues and how to solve/resolve them.

You are also welcome to go to my YouTube channel (Brent Baumer) and watch my videos on using SKM for power system studies. If you still need help, I am willing to do a short ( <30 minute) Teams call free of charge.

It is not a delay. Your arcing fault current of 2.9kA is withing the short time region of GEN BRK and ATS FDR, not tripping until about 60 seconds and 170-ish seconds respectively. On your arc flash study you have selected the "2 second rule) so it cuts off at 2 seconds (the max time it would take for someone to be blown away from the fault or run/dive/crawl away from the fault.

It looks like SWBD 4 MCB should trip in the instantaneous region at 2.9kA just looking at your graph but I don't know exactly where 2.9kA falls. It may be in the short time no-man's land and not trip within 2 seconds.

You might be able to lower the breakers INST down a notch or two or 3 but it looks like you're going to effectively lose the short time pickup and delay if you go much more than than unless you also lower the LTPU, which will basically reduce your generator capacity.