[ long and detailed discussion omitted for the sake of brevity.
It comes down to: You can really run exposed NM cable a.k.a. Romex
in unfinished basements on the walls. ]
You made me go down to my basement (currently being finished!) and get
the NEC, and look up all the references. You are indeed completely
correct. For some reason, I had stuck in my mind that the 7' rule
applied everywhere, and it really is only for attics.
That having been said: If I were a building inspector, and I saw
exposed NM on a flat surface in a wood shop (where people move sheets
of plywood and long wood pieces around, and push heavy cast iron tools
on mobile bases), I would insist on protecting the NM cable.
Unfortunately, article 336-6(b) never defines what "physical damage"
Here is a slightly scary story. When talking to the chief building
inspector (the plan checker), he suggested that I use explosion-proof
wiring for my wood shop in the basement finishing project. He even
pointed out that wood dust in the air does indeed make my 200 sqft
shop qualify to be a class II group G hazardous location --- which
would have made wiring it just about impossible for an amateur. He
relented when I pointed out that (a) all the wiring will be in stud
walls, behind drywall, (b) a small residential shop is not a hazardous
location (in the NEC sense), and classifying it as such would be at
least very unusual. Then he suggested that I get a dust collector to
exhaust the dust to the outside; and I showed him that a location for
that dust collector was already identified in the plan, and that a
dedicated circuit for it will be provided. He really liked that idea.
In this climate of extremely eager (one might even say over-eager)
building inspectors (a.k.a. the authority having jurisdiction), I'd
rather be safe than having to redo the job.
By the way, on another topic: There was a short discussion of box fill
calculations. From my experience: A box that is too full or close to
the limit according to the NEC calculation is also no fun to wire.
Doing it the first time, with all new wire, is doable but tough. But
if you ever have to redo it later (add more stuff, change the
connections), it gets to be absolutely no fun. After trying to jam
too many wires into boxes a few times, I gave up, and got a lot of
larger boxes, and even retrofitted them in a few places.
But also, the box fill calculation in the NEC is a little too
simplistic. Here are two examples on opposite ends of the spectrum.
A wire that simply runs straight through a box counts for one
allowance (typically 2.25 cu in for 12-gauge). That's silly, because
a wire that goes straight through uses up next to no room, sits flat
against the back of the box, and doesn't get in the way. Similar
argument for internal clamps, which also cost one allowance, yet the
typical Romex-clips that are found in metal outlet boxes use up nearly
no room, sit on the edge of the box, and don't get in the way.
On the other hand, a device on a yoke counts for two allowances,
independent of what size it is. For a non-grounded light switch
(which is very thin) or a normal side-wired 120V outlet, this makes
sense. But the same rule applies to really big and fat devices, for
example a GFCI with 5 wires attached (2xline, 2xload, ground), or a
4-way light switch (also 5 wires), or a L14-30 outlet (4 wires, each
10-gauge). Trying to push those big devices into a box that's already
near the fill calculation limit is not fun, and is really asking for
trouble (wires and fingers usually get pinched). In those cases,
really deep boxes (2 1/8 deep), or even the oversize boxes (I think
they are 4 3/4" square rather than 4" square) are your friends.
Extension rings are a second-best solution (because their flange gets
in the way).
I'm also not sure I understand the progression of volume allowances.
It is 2 cu in for 14-gauge, and only goes up to 3 cu in for 8 gauge.
But to me, it seems that 14 gauge is so much easier to handle, in
particular because you take a bundle of them, even with a wirenut at
the end, and bend them into a nice twisted loop to store them in a
box. I'd hate to work in a box that's filled to the NEC limit with
solid (not stranded) 10-gauge and 8-gauge conductors, that must be
The address in the header is invalid for obvious reasons. Please
reconstruct the address from the information below (look for _).
Click to see the full signature.