I just saw a study that had 166 cal/cm^2 and an arc flash boundary of 30'. This is on the primary side of the service disconnect for a facility.

How realistic is that 30'?

It all depends on how the study was done. Did they model from the utility primary fuse in with the primary fuse as the protective device for the main service equipment? Did they use a 2 second cutoff for the clearing time? Did they use available short circuit current from the utility or was it an infinite bus value? Let's start with those questions prior to determining validity of result.

Yes, utility primary fusing was modeled and is the protective device for main service equipment (xmfr is 3,250 kva)
Yes, a 2 second cut off time was used.
The short circuit current from the utility was 4,300 A at 12470 V.

I thought I read somewhere (and cannot find) that the equations can over-estimate arc flash boundary to something that is not realistic so wanted to reach out to see if that were the case.

166 calories with a 30-foot AF boundary is realistic for the secondary of a 480V, 3750 kVA transformer, but not the primary.
If the label is on the transformer itself then it makes sense because the secondary is exposed when opening the cover.
John M