I need to work on the bathroom sink and would like to use the valves under the sink in furtherance of that objective. However, these will not move even under a gentle beating with a light hammer.
I happen to have some dry ice at the moment and was wondering if by shrinking the metal in the valves, that may they not become more compliant?
Is there is chance this could work? What are the risks, if any?
You’ll probably wind up freezing the water inside the pipes and causing them to burst…OTOH, I know some people will use liquid nitrogen to freeze a section of a water pipe so they can work on it without shutting off the water.
My suggestion would be to shut off the water to the house and then replace the valve, or at least dismantle it and see if you can figure out what the problem is.
I was thinking of evacuating the line first and relying on the shrinkage of the metal parts to disrupt whatever corrosion is preventing their normal operation.
I could simply replace the valves, but this would save me time and money if it works. Thank you for taking the time to respond however. 
If you can access another shut-off the best bet really is replacing them. If something goes wrong, how are you going to stop the flow?
If they are stuck there must be corrosion or mineral deposits, and though freezing them may break them free it may just plain break them. If they are stem valves with washers, the washers are probably toast. If they are ball valves they probably still won’t seal properly anymore even if you unstick them.
There should be a main shut-off. Use it.
Yes I know all of that - if you read my post I clearly said there are valves just for the bathroom.
I also realize the valves are “toast” as you put it. The only thing I care about is whether or not this might work. If you are also saying that the cold temps would destroy non-metal components, I find that highly doubtful and would like to know upon what information that opinion is based.
These valves are OEM so to speak which means early 60’s as that is when the house was built. It was built by a contractor for his own family so I know he used the best of everything available at the time.
I wouldn’t have to bring the valves all the way down to -70F necessarily. I can make attempts at loosening them once I see that the laser thermometer is reading anything below 20-30F.
I think what Fluffy is trying to tell you is that in the plumbing world a stuck valve usually means it’s time for a new valve. Once a valve has seized, so to speak, they tend to self destruct upon un-freezing. Of course, IIRC, it’s not that they’ll leak when in the open position (though they can, water can come up around the stem) it’s that they won’t shut off properly.
If you really don’t want to replace them, shut off the water to the house and remove the packing nut, the nut around the stem…I assume we are talking about stem valves as opposed to quarter turns. The innards should lift out and you may be able to assess the damage. If it looks repairable take it to a plumbing store and see if they have a replacement stem.
From what I understand, most metals contract while being cooled. If trying to remove a tightly bound metal on metal contact, heat usually allows the outer device to expand when heat is applied externally. I don’t know why dry ice will help this particular situation.
I appreciate that suggestion and I have done precisely that in other situation. I’m not sure why it didn’t occur to me this time :smack: - probably because I was excited about having a decent question for GQ. Fuckin’ over-eager, wide-eyed newbies. 
I would still like to pursue this for anyone who might have something additional to add. But again, thank you.
My thought was that all of the metal would be cooled together and that in the process of shrinking would either loosen some of the corrosion making it possible to turn the valve or would create enough slack between whatever parts are sticking that I could accomplish the same end.
I like the trick of heating the outer metal at a faster rate than the interior sections. With a high enough temperature heat source, you should be able to get significant expansion of the external parts with minimal expansion internally.
In this case though, the temp difference is only about 150F - not the couple thousand F degrees you might get with a torch.
This. It’s usual to use heat, like a blow torch, for freeing sticking nuts as the heat will expand the nut, freeing it from the unexpanded bolt. This should apply equally for your undersink valve.
Thank you. I can work with a flame (nitrous, propane/oxygen) but it’s not my first choice in a confined space. Also I have absolutely no idea of how much heat to use, how long etc. I would like to try it, but I think it is one of those things you need to have a “feel” for - something that only comes with experience.
I guess if I ran the water while doing this it would reduce the odds of going overboard though. What do you think? It’s counter productive but would make me feel better if you think it still might work.
I’m not much of a plumber, but would spraying them with PBBlaster or another penetrating oil help? It’s pretty rare that I run across a siezed bolt that PBBlaster won’t loosen, but I don’t know what these valves are like.
I’m kind of interested to see if dry ice will help and hope you proceed with the experiment for the good of humanity. On the other hand, you need special valves on CO2 tanks or else they will freeze and stick, so … maybe not.
thank you for taking this in the intended spirit - an experiment.
Again, I have to apologize for not having thought this through. I’m not familiar with that brand, but I do have some liquid wrench. Should I try that? do you think PBB is available in the NE US? If I try WR and it doesn’t work, will it interfere with PBB?
Thanks.
It sound like you’re having trouble turning the valves to the off position. Using a torch you could just end up destroying the washer inside the valve. There usually a cap over the valve that the stem runs through. If you can remove that, and the lever or wheel on the end of the valve stem, you can shoot some lubricant inside there and get it in the threaded part of the valve. Sometimes there a flatten section of the valve stem you can grab with a wrench to get more torque on the stem. If you see green corrosion from copper, there are some acids to help dissolve that. I couldn’t give you a recommendation on that though. Maybe a local plumbing supply could help.
Good stuff Tripolar. I’ll check that out - maybe tonight since I don’t seem sleepy yet and nothing sooths the soul like power tools (yes I know they’re not called for - before you start harshing my buzz).
You can cold soak the back part of the valve with a wet rag and heat the rest. That’s how I keep soldered joints from remelting if they are close to one I’m working on.
They make a do-hickey that does just that.
I’m more worried about being extremely clumsy and setting fire to the cabinet, hamper, bottles of unknown solvents, etc. (some exaggeration - not a lot)
Oh man that is so friggin’ cool. Shit - just saw the “cool” part, but I’m not changing it.
I have virtually no use for something like that but I want one. If I had it, I could FIND a use - right?
Please, don’t enable my addiction. :eek: 
I would first try to use a big wrench or pliers on the handle. Maybe apply some heat safely with whatever you have-dryer, heat gun or torch. Only try the torch if you have complete access and no chance of causing a fire.
These valves get stuck when people open a valve completely allowing the stem to bind against the back of the fitting. When water is turned on, the best thing is to open a valve all the way and then turn it back in half a turn or so. You lose no pressure and the valve will not stick.
In general, if you have two metallic parts which fit snugly one inside the other, cooling both equally will make the fit tighter, while heating both equally will make the fit looser. So for something like a ball valve, cooling the valve itself will be exactly the wrong thing to do.
With any valve of any design, you have to consider that the amount of heat you can apply with typical DIY tools is more than enough to destroy any gaskets, seals, o-rings, etc inside the valve. Prety well guaranteeing that the valve will no longer shut off, and may well leak directly to the outside world, e.g. around the valve stem.
It’s fun to experiment, but under a sink cabinet and playing with a water flow you may not be able to turn off upstream sounds a bit like “experimenting” with bear-baiting.
A couple years ago I was exploring into a shower valve with the house supply still on. I just wanted to get a part number so I could buy it some new guts; I didn’t intend to disassemble it far enough to trigger a leak. I *juuust *touched one screw of what I thought was a fascia plate and the plastic innards broke in two, triggering a flood. This was inside an enclosed shower stall, so you wouldn’t have thought it was too bad, right?
Well it took me probably a minute to get to the house shutoff (an indoor ball valve I knew worked well). But in that short time the high speed spray in wild directions had soaked a 5x8ft patch on the ceiling and much of the interior of the wall behind the shower enclosure.
Didn’t end up with mold in the wall, but I did spend a couple man-hours trying to mop out that space as best I could. And a couple days with a fan and a hair dryer in there to dry out the rest. And no shower in that bathroom for 3 days.
It was a good experiment in that much knowledge was gained. But it’s not one I’m inclined to repeat. Nor any of its close cousin experiments.
LSL - interesting. I’m trying to understand though. Let’s make this simple for illustration. Let’s say you have a 1" OD pipe with a wall that is 1/8th (less say about 3/1000ths) and you slide into that another pipe with 7/8ths OD (make the wall whatever works for the example).
Now we’ll subject the pipes to corrosion “fusing” them together.
If I cool the assembly to -70F, you’re saying that that the outer pipe will contract more than than the inner pipe and make the fit even tighter?
That would make sense if the smaller size (mass?) of the the smaller pipe meant it couldn’t contract as much. But is that really how it works? I would have guessed that they both would have contracted proportionally. If so, shouldn’t the worst case be that they shrink an equal amount and the tolerance of 3/1000 remains the same?
I spent a while on this and realized that the walls would shrink in all 3 dimensions and I had no idea how to model that.