I just have a question about how to model a long busbar. I have a busbar that is about 600' long and has breaker tapped into it every 20 ft. That would be 30 taps into a 600' long bus. I am trying to model this in a way that takes into account the impedance of the bus. I would also like to know if perhaps the impedance of this bus does not matter for short circuit or arc flash calcs?

My only though is that maybe I can just add the 600' length to the cable that is feeding the bus.

If the impedance is significant, model it as 30 buses with 20 ft of bus impedance (you might have to call it cable) in between each. The impedance will matter, but it may not affect the short circuit current or arc flash IE enough to affect the evaluation.

Busways are nothing more than cables with a different shape. I treat busways like cables. Busway impedance's are typically very low. I calculate the Ibf at the feeder end then at the opposite end of the bus, run the IE numbers and see if there is a significant difference of HRC. If so, I take the lowest Ia of the two numbers which gives the longest PD clearing time. Most times unless it's a very long busway run there isn't enough difference in IE numbers from one end to another to change PPE levels or go crazy with labels every 20 or 30 ft. I found the safest and simplest way it label the entire busway based on the highest IE. Beside, a 4"x6" label on a busway 20' in the air is impossible to read. For one customer I color coded the busways with 12"x12" plastic engraved placards. Each placard had large engraved font with the PPE level. Green cat 1, yellow cat 2, orange cat 3 & red cat 4. Over cat 4 was a red placard with PPE with a circle and line through it. Easily readable from 30' floor level.

As mentioned earlier, sometimes impedance matters, sometimes not. The capacitance can cause all kinds of strangeness like having higher fault current at the remote end (in an impedance grounded system).
http://www.cdc.gov/niosh/mining/pubs/pdfs/deoco.pdf

Another big problem is that especially if the bus is connected very close to the transformer, the reduction in fault current may get low enough that the main breaker takes excessively long to trip and thus the incident energy increases paradoxically as you get further away from the transformer.

Generally I like to think of these things like a parabola. You are usually on either the rising or falling side of the curve. It takes at least three data points to determine where the parabola might fall and whether it affects your results or not.

If you want the electrical properties of the bus bar they can be found here: http://www.hubbellpowersystems.com/literature/connectors/AEC-41.pdf
With this you should be able to calculate the GMR based on the spacing then the impedance - it wont be pretty. I would go back to the manufactuer and request the info.

Each mfg has data sheets for their busways. GE, Siemens, Westinghouse, Sq D are the major guys. As pointed out by engrick... busway impedance is very low microhms/ft so unless its a 1500' run of alum conductor 400A busway there is typically little change in Ibf from one end to the other.

I went ahead and modeled it as 30 individual buses with 20' sections of busway inbetween them. Unfortunately I had to model it in a different online because I hit our software license bus limit. The results were Interesting but the worst case scenario at any point along the busbar remains the same so we will just label the whole bar as such. But it does have an interesting affect on the available fault current at the different points along the busbar which we will have to look into more and possibly upgrade some breakers in the future.