The issue really is that when you are doing the initial design, it can be months or even years away from the actual installation, and you could have little data.
I always try to size cables and equipment for the maximum fault level that's likely to occur, and the protection such that everything still works if the fault level is much lower, but this is difficult as you don't want to over-specify as this really puts the cost up.
Even at the later stages you often still won't have genuine figures from the dno, they usually just supply stock figures.
So unfortunately you are in a position of ensuring a cost effective design, which is suitable for fault levels you don't know!
With site transformers there is often another issue, and that is you have to assume a primary fault level and a size/voltage impedance of the transformer, it has happened to me in the past where I have specified a transformer size, only for the installer to use something larger (and lower impedance) as they often take it as a minimum. (As this is normal for kit sizing, - we can use 10kA mcbs in place of 6kA, 10 way board rather than 8 way, a 3kVA 230/110V transformer is ok to use instead of 2kVA etc)