What is the minimum horizontal length of PVC conduit that can be run before an expansion couping is required by code? Code reference please. Thanks

It depends on the temperature change; for a 200 degF change in temperature, as little as 3 feet, but that's obviously not typical. See 352.44, and table 352.44(A).

You'd also have to look at the coefficient of expansion of the structure, beit wood or metal. PVC expands/contracts about 6x greater than concrete or steel and about 20x more than wood. Humidity changes in a wood structure may far exceed thermal expansion of the structure or the PVC, and should also be taken into account.

0.0000380 %/F, PVC
0.0000060 %/F, concrete (varies by type)
0.0000065 %/F, Steel
0.0000020 %/F, Wood (varies by type)
0.5000000 %/% moisture for Fir

352.44 requires an expansion fitting if this is going to be more than 1/4".

[This message has been edited by SteveFehr (edited 10-29-2006).]

I think the NEC has adequately provided the guidelines for installing expansion fittings for raceways be it metallic or non metailic. If the building is shifting is one thing but if the temperatures are causing structural materials to expand or contract they have more of an impact on the PVC than the other materials so I would not be concerned with anything except the PVC. The expansion of wood or concrete is nothing in comparision. I would ignore it. It is also possible in most cases to install a raceway system in such a fashion using elbows etc and not have to worry about expansion fittings at all.

In my area I like to figure an exposed outside run of PVC is exposed to a temperature change of 100F.

Given that temp change I need an expansion fitting where 6' or more of PVC is installed between fixed points.

I get a lot of grief for using 100 F. People ask how often do you get a 100 F day or how often do you get a -10 F day?

IMO how often is not relevant, what is relevant is it does happen. When I see exposed PVC runs that people have not installed enough expansion fittings in I either see the fittings pulled out or the PVC looking like an ocean wave.

I also see people forget that PVC support spacing is not the same as EMT.

6" PVC requires a support every 8'

3 1/2" to 5" PVC requires a support every 7'

2 1/2" to 3" PVC requires a support every 6'

1 1/4" to 2" PVC requires a support every 5'

1/2" to 1" PVC requires a support every 3'

I agree and that is a great point to bring up.

It is probably a better design to use turns to make up for expansion.

I see turns used to compensate for expansion on steam lines and gas lines.

Usually if it is a long straight run they put some bends in the middle, kind of a "U".

Right, left, left, right, if that makes it any clearer.

Quote:
It is also possible in most cases to install a raceway system in such a fashion using elbows etc and not have to worry about expansion fittings at all.

I agree with you George but wouldn't you have to put one in anyway for code?

Eddy- It's possible you might still have to install an expansion fitting even if you used elbows and the like to take up the expansion/contraction but not necessarlily per code. I'd like to add to what iWire said because he knows it and just didn;t say it. What is just as important and maybe more important than we realize is that we use the supports that are designed for PVC. I'm talking about the ones that allow the conduit to slide back and forth during temperature change. So many jobs I look at the contractors use standard one hole straps and they end up having to change them.

Per 352.44:
"Expansion fittings for RNC shall be provided ..... in a STRAIGHT RUN between securly mounted items such as boxes, ..."

So if you run the 90s (up, over, down) no expensive expansion fittings are required. Plus, it is easier to pull the wire in.

I think we need to keep in mind that the problem is that the PVC WILL expand- and a lot more than anything it's likely to attach to! You simply can't fet around this fact.

Whenever you see a drooping, or wavy, PVC run, you can be sure the pipe was not purchased, or installed, that way. What you see are small bends caused by expansion, aided by the weight of the wires within.

Plumbers can use paired elbows, simply because their elbows are threaded. Pipe movement actually results in the threads tightening / loosening. With glued-up pipe, the angle of the bend will change.

Support the pipe more often, and you will get a series of small bends- rather than a few big ones. Within limits! Too much expansion, and you just might start popping anchors loose.

As for what temperature to use in your calculations: what matters is the temperature of the PVC. A run exposed to the sun will heat up well above the air temerature. You really ought to consider this in your ampacity calculations, as well. With PVC this is especially important, because the material becomes much more pliable with even a small change in temperature.

Thanks for the replies. The run I'm talking about is installed under a mobile home.(feeder conduit) it is about 8'in length and is between a 90 and an LB.