The transformer design doesn't matter. When you look at arc flash incident energy, obviously one area of concern is inside the transformer housing itself if it's a dry type, or at/around the bushings if it's a sealed design (generally oil filled). In this case, you look at two places. First, you look at the available fault currents, X/R ratio, and so forth on the primary side. Generally this is not the major issue and would only be separate anyways if you have two separate air termination cabinets like what is common on a substation-style transformer with ducts on the front and rear.
Second, you look at the available bolted fault current on the secondary side. You already know the impedance so getting to a bolted fault current is pretty easy, and so would the X/R ratio. These give you the necessary values in order to calculate incident energy. Granted there are several taps on the secondary winding but since arc flash incident energy is normally calculated as a worst case, you can safely ignore all the intermediate taps and focus on the overall secondary coil.
In terms of a drive, this is slightly more complicated. There is some material published on DC but not a lot. There is a new second in Annex D of 70E for 2012 that gives a simplified DC arc flash calculation which is close enough. This gives you available fault current on the DC bus, assuming you've got a traditional (converter/inverter) design and not a cycloconverter/matrix design where there is no DC bus.
On the output side, you need to look at available fault current from the drive. One consideration that drives this is that the semiconductors themselves which are supplying power at any given time are particularly sensitive to heat. This means that they are fused with very high speed fuses (trip in 1/4 cycle or less) with the exception of Siemens. So although it is certainly possible to exceed the maximum output of the drive, for all intents and purposes you can look at this situation by simply modelling the maximum output current of the drive itself plus the protection which is configured in the drive. During operation you've got to remember that at any given time up to 2 IGBT's can be switched on at any given time and there are places in the timing where it is even possible to have a short circuit across the drive internally.
Siemens is not putting semiconductor fuses in all of their drives. This may be an issue. Six drive installations in Australia have reported that when a drive module fails, it literally explodes and throws shrapnel around. The design has not changed. The following presentation still advertises a lack of fuses and is from a company which is now part of Caterpillar. I would have to take a hard look at exactly how much total energy is stored in the DC bus capacitor or inductor if I bought a Siemens SiBAS traction drive or similar due to this potential issue.
See this presentation, page 71:
http://www.wmea.net/Technical%20Papers/AC%20IGBT%20Drive%20System%20for%20Draglines%20-%20Nov%2006.pdf
