Sorry that I've not been following this thread, but the original question happens to be one that really bugs me. It seems to me that the CMP did one of those 'back handed' code changes where one part of the code is adjusted to correct an error that really should have been dealt with someplace else. An example is 310.15(B)(6), where a better load calculation would probably make more sense. In this case, the CMP is correcting an error in ampacity calculation; but rather than saying that NM cables with 90C temperature ratings have _different_ ampacity than THHN conductors in conduit with 90C temperature ratings, they simply say that NM cables must be used at their 60C ampacity.
The physics of ampacity calculation is given by the Neher-McGrath equation,
http://www.calcware.com/cwnmcalc2.htm . This equation is explicitly permitted to be used for ampacity calculations, under 310.15(C) if you have suitable 'engineering supervision'.
What the Neher-McGrath equation gives us is a relationship between the conductor temperature, the surrounding ambient temperature, the electrical resistance of the conductor, the thermal resistance between conductor and ambient, and the current being carried by the conductor. The conductor temperature is not set only by the current and ambient conditions, but is also set by the thermal resistance to ambient. Presumably any other sources of heat (other conductors) in the vicinity would be considered somehow as part of the ambient.
The tables permitted under 310.15(B) are based upon the Neher-McGrath equation with certain assumed parameters of thermal resistance and insulation characteristics. These parameters are never stated explicitly, however it is quite clear from the way the values of table 310.16 change with ambient temperature and permitted insulation temperature that the Neher-McGrath equation is being used.
This tells me that if a conductor has an ampacity of X amps, and I push X amps through that conductor, then if the presumed conditions of thermal resistance hold, then the conductor will heat up to its maximum temperature ( the temperature value at the top of the column). If I take three #8 THHN conductors, and bundle them together into a conduit, and run 55A through each of them, and if the ambient temperature is 30C, and <b>if the thermal resistance of my experimental setup matches that assumed when table 310.16 was generated</b> then the temperature of these conductors would hit 90C.
In particular, if the _derated_ ampacity of a conductor is based upon the 90C rating, then I expect the conductor to heat up to 90C when used at its ampacity. Going back to the example above, three #8 THHN conductors, in a conduit, in a 50C ambient condition. I run 45A through these conductors. I expect a conductor temperature of 90C, even though the conductors are being used at less than their normal 75C ampacity.
Unfortunately, I've not worked 310.15(B)(2)(a) into this thought process
Take 3 conductors at full capacity and call the heat output 1. 6 conductors at 0.8 capacity would have a total heat output of 1.28, 9 conductors at 0.7 would have a total of 1.47, 20 conductors at 0.5 would have a total of 1.67, 30 conductors at 0.45 would have a total of 2.03 and 40 conductors at 0.4 would have a total heat production of 2.13. Clearly the assumptions about thermal dissipation capability have to change as the number of conductors in a raceway goes up.
I am forced to assume that 310.15(B)(2)(a) was derived in a fashion similar to table 310.16, and that if I have a set of conductors with derated and _adjusted_ ampacity of X, and I run X amps through these conductors, that the temperature of the conductors would rise to the 'temperature rating' of the conductor as used in table 310.16 prior to derating and adjustment.
This problem with this whole approach, however, is that the thermal resistance in the real world is very unlikely to match that used to derive table 310.16. The thermal resistance numbers are presumably relatively conservative. However if, for whatever reason, the thermal resistance in a given situation were _higher_ than that assumed for table 310.16, then the temperature of the conductors would end up higher than expected.
My first guess is that the CMP decided to deal with the issue of 'NM cable buried in thermal insulation' by simply saying 'use the 60C ampacity'. This is probably a much simpler approach than having to develop an entire different set of tables for 'NM cable buried in thermal insulation' and trying to deal with the enforcement nightmares. But it hides the reality that a 90C conductor is a 90C conductor, designed to be used at up the point where the wires are actually at a temperature of 90C, near to the boiling of water.
The above leads me to a different guess. NM cable is generally permitted in flammable construction where the fuel load of the plastic cable is nothing compared to that of the rest of the building. I guess the thought is that 'wood is already flammable, why worry about a bit of plastic'. The kindling temperature of wood is very much dependant upon the water content of the wood. Heat wood up to 90C for an extended period of time, and you boil the water off and make it easier to ignite. Perhaps the 60C restriction is used to protect the building materials that NM is ordinarily used with.
-Jon