Best pipes to survive freezing?

[ITMFA]

which pipe can best survive a freeze. plastic, copper or steel?

[RBM]

PEX

[Eigenvector]

Yes indeedy - EXCEPT for the joints

[Red]

The worst is pvc. It doesn't just split, it shatters for 6-8 ft.

[jtnospam]

I live in SoCal, so I'm not speaking from experience, but copper will stretch much more than the different plastics and iron (almost all "steel" plumbing pipe is really iron, not steel). When it does rupture, it will be less catastrophic and a less drastic leak than plastic or iron. Help yourself by insulating it well.-Jitney

[Edwin Pawlowski]

Except for PEX. It will take much more than copper.

[dpb]

Besides the other points made, also depends heavily on wall thickness

-- all are made in varying "schedules" -- copper "L" and/or "M", for example, Sch 30/40 as examples for pvc and/or galvanized/black iron...

What's the application and/or reason for the question?

[Karl S]

There has been a lot of real-world testing out here on the frigid, flat, frozen, windy, treeless tundra of Western Alaska.

The answer is PEX.

[George E. Cawthon]

Agree, who better to answer than an Alaskan. But yes copper does stretch, right apart, just like iron pipe.

[Dan_Musicant]

: :ITMFA wrote: :> which pipe can best survive a freeze. :> plastic, copper or steel? : :Besides the other points made, also depends heavily on wall thickness :-- all are made in varying "schedules" -- copper "L" and/or "M", for :example, Sch 30/40 as examples for pvc and/or galvanized/black iron... : :What's the application and/or reason for the question? So, what will withstand freezing better? L copper or M copper?

[dpb]

In order of _decreasing_ wall thickness, copper is K, L, M ...so M (of a given diameter) is most likely to burst first, K withstand the most

[George E. Cawthon]

That is what one would expect based on the strength of the pipe. However, I suggest a test by filling 12" lengths with water capping both ends (no airspace) and then setting them outside at night for 8 hours in 15 degree weather. If that is too cold for you then set them outside when the temperature is 25 degrees. I would expect that if one pipe breaks then all pipes will break. You could throw in an iron pipe and see how that fares.

OTOH, there are lots of variable that could be expected, so one has to be very precise about the freezing temperature and length of period to predict anything meaningful.

[dpb]

If you'll read the post, all that was compared was the expected _relative_ likelihood which is the question asked and there's no doubt at all that thicker-walled tubing/pipe of any type will be more resistant than thinner walled which, again, was the only point being addressed.

[cavedweller]

Well, no, actually. So-called iron pipe is really steel. Steel is simply iron with a little bit of carbon, after all. "Cast iron" piping comes in larger sizes and is used for DWV.

[dpb]

...

The volume expansion owing to the phase change to ice from water occurs at the freezing point so the atmospheric temperature doesn't make any difference on that score. Once solid ice is formed, the thermal coefficient of expansion is positive so as it is cooled further it actually would contract slightly. (In fact, it's about three times that of copper ~51E-6 vs 17E-6 in/in/C) so would actually reduce the pressure on the pipe wall as the temperature were lowered as it would contract more rapidly than the copper pipe itself).

[George E. Cawthon]

Unfortunately "survive a freeze" is rather ambiguous. Does the OP mean air temp drop below

32 or does he mean that the all of water in the pipes actually freezes. If the latter, then one would expect little difference relative to wall thickness. In the real world where pipe temperatures are such that all of the water doesn't freeze, the shape of the pipe, the orientation, joints, the wall thickness, etc. make a difference. Possibly number of joints, types of joints, and smoothness of the joints is more important than the wall thickness.

[George E. Cawthon]

You seem to just want to argue. No one cares about this stuff. We are talking about the real world experience where the weather gets colder (atmospheric temperature drops), and I would assume the pipes are in a house.

The point of specifying a time and a temperature was to make sure that the water actually freezes. So if you like, just put the pipes in a freezer and you won't have to worry about the effect of the atmosphere.

[dpb]

"In the real world" it's highly unlikely the OP would be asking about other than a condition where sufficient water freezes that a portion of the pipe is solid so that, for all practical purposes, the entire pipe may as well frozen.

Again, whether it actually causes a pipe burst in a real plumbing orientation in a house, for example, has a tremendous amount to do with length between joints, hot or cold supply line, etc., etc., etc., granted. That's why I asked OP quite some time earlier for the motivation behind the question in the first place.

Again, all I pointed out was that for comparative purposes, a K or M copper pipe is much more robust to freezing than an L pipe of the same diameter, length, orientation and so on. That is still true and is also just as true for Sch 40 vis a vis Sch 30 pvc or galvanized. No more has been claimed, but no less, either...

As a practical "real world" thing, I don't recall ever having had a break of an "M" piece of piping, but several in "L". That's not to say "M" won't burst when frozen and I have never said such a thing and I'm sure there are many instances where it has. But, I can certainly state that one disadvantage in using "L" for the initial cost savings is that it will certainly be far more susceptible to a break from freezing than will "M" -- again, in the same configuration/environment. That it would be guaranteed to hold, no; I'm not nor have I claimed that--only that odds are are significantly improved.

[dpb]

No, you are the one that seems to want to argue -- all I said was that thicker walled pipe is more substantial than thinner and you started the argument of trying (it seemed to me) of reading far more into that than what I said -- If not, fine, let's let it drop. I do have a penchant for not wanting what I post distorted nor taken from context, however.

But, as I noted elsewhere, we can assume all we want about OP's actual purpose in asking the question but we don't know -- I asked earlier and didn't get it answered so still don't know.