This picture was taken in March 2003 in Indianapolis. I posted it here to generate this discussion. I did not inspect this work, and have no knowledge of the installation.
If the fine print note to iwires reference was reviewed it would show that an expansion fitting would be OK, and or additional lengths (slack) in the feeders and laterals to allow for movement caused by settlement or frost.
Exposed runs of RNMC would have to comply with the comments related to that in Article 352 and I agree.
I am happy with all of the comments here and see where we all have something to add that helps the new persons understand.
Really is better than arguing!
The wireway "Gutter" is one designed for use in this wet location. I have a close up of the label, and will post it here when I can find it, the trouble with me is that I am always moving my pictures around on my computer.
On another note before "expansion fittings" were introduced into the NEC we were always allowed to use "Expansion Joints" but now I see that they are not allowed by 300.7(B) and the FPN calls attention to the RNMC article too.
300.7(B) Expansion Fittings (used to say Joints).
Raceways shall be provided with expansion fittings (used to say joints) where necessary to compensate for thermal expansion and contraction.
FPN: Table 352.44(A) provides the expansion information for polyvinyl chloride (PVC). A nominal number for steel conduit can be determined by multiplying the expansion length in this table by 0.20. The coefficient of expansion for steel electrical metallic tubing, intermediate metal conduit, and rigid conduit is 11.70 × 10-6(0.0000117 mm per mm of conduit for each °C in temperature change) [6.50 × 10-6 (0.0000065 in. per inch of conduit for each °F in temperature change)].
Personally, I would still allow the "expansion joint" we made using a flexible raceway like greenfield or sealtight between the runs of RMC, for example at the building expansions.
How about you other AHJ's?