Which brings up an interesting point. Do we get to use the ISC of the module, or do we need to have an adjusted ampacity greater or equal than the string fuse?
In NEC2011, the article 690.8 had a three point checklist for conductor ampacity. In summary:
A: 1.56*Isc for terminal temperature ampacity, usually 75 C
B: 1.25*Isc/total derate for wire temperature ampacity, usually 90C
C: OCPD where required shall protect the wire as sized. Thus both the terminal ampacity and the derated conductor ampacity must at least be large enough to round up (240.4(B)) to the as-built OCPD rating, and not correspond in an exact match to the previous standard OCPD.
Part A and B still exist in 690.8 for 2014, and are the same calculation, just with different wording and arrangement. Part C is mysteriously absent. I'm not sure if this removal was intended to remove this calculation, or if it was removed because it is implied elsewhere in the NEC. More likely because it is implied in the general articles for overcurrent protection.
And there are numerous examples where part C governs. Because OCPD is sized by 1.56*Isc, and then rounded up to a size that is available. Adjusting it for derate factors, it is is significantly larger than 1.25*Isc/total derate, even when you can take credit for the non-intuitive 240.4(B).
With string fuses, every integer ampere below 15A could be a standard OCPD rating for 240.4(B). So using a 15A fuse, that means you need at least 14.000000001 Amps worth of wire ampacity at conditions of use.