Geesch! I think I get it, as Gfretwell described it. The point of transition must be the point at which the wiring coming from the MCC first encounters the furnace area. So, if the furnace is 10' wide or less, it would be less than 10% of a long run of 100'. The relatively long run would act as a heat sink. Then you could apply the higher allowable ampacity value to the portion of the circuit beyond that transition point..or, is it the next transition point where the wiring leaves the furnace area? It wouldn't make sense to allow a wire to cook behind a furnace and still be fully loaded per 75 degree table!
I guess I'm still confused as to where that transition point should be.
George, your example seems to be allowing someone to immediately go back to the #3, without any cool-down distance consideration! I think the whole point of this article is to keep the current/resistance induced heat from adding to the ambient induced heat, thereby cooking the insulation. If you allow the wire to cool down for a sufficient length after being exposed to a high temp area, then you won't run into trouble. Your example also ignores the usual branch circuit loading of only 80%. You were describing the load on that #3 as 100 Amp, and that would be putting a 100% load on that #3. But, I'm sure you knew that, and were just simplifying the example.
