View Full Version : Rejoin 1" in-slab radiant pipe after removing rusted section

11-02-2008, 05:26 PM

I have a nice old in-slab radiant heating system with 8 circuits in ~5000 sf that works great... until I sprung a leak last season. From what I can tell the system is actually in really good shape for its age. Anyway, my plumber used his infrared camera to help me locate the leak and I isolated that circuit. Since then I have cut open my slab and removed a rusted out section of the pipe. It looks as if there was a butt weld and just next to it on the underside the pipe rusted out over time. In the course of pulling out this section I got a look at an adjacent pipe and it looks fine. I also cut back far enough (about 18" to either side and found good solid pipe. So I cut out the rusted section, threaded both sides and am trying to join it back together.

Problem is that the pipes have absolutely zero give to them, either longitudinally or angularly. In order to make it up, I needed to use a length of iron pipe and two unions. It was not possible to use a coupling and one union. To do so I had to make the new section almost .6" shorter than actually required just top be able to install it over the rounded portion of the union fittings. I tried to then back off all 4 nuts each a little bit tpo make up the section, but alas it's not up to my standards. I can get it to hold 10 psi or so, but I can tell that it's just too fragile. I had to back the nuts too far away and each of them is leaking a tiny bit.

Question is what to do? I would prefer to stick with iron pipe, but can't think of a way to join it short of welding. That would work but will probably require more diging, of which I have had enough. I could go to 1" PEX, but I am afraid of burying those brass fittings into the new cement which I will have to pour. Do any of you pros have any suggestions? I suppose the PEX connection wil probably last 10 years or so, but I do not think it's good practice to bury that in concrete. I could figure out a way to encase it in a protective shroud, I guess. I could also go to copper somehow, but that is going to face similar problems with inflexibility of this situation as I ran into with iron.

Would appreciate any good ideas. Have photos if u want to see them.

Thanks a million,
Michael Baltay

11-02-2008, 05:58 PM
Coupling and a union. However if you have one leak, it's only a matter of time before more start springing up. You will find yourself chasing leaks until you finally replace the whole mess.

11-02-2008, 06:22 PM
Thanks for your quick response. I would appreciate any details you can provide in order to facilitate this reconnection the way you suggest.

I am not sure I will have the angular freedom to fit that up with a coupling and union. I can try to turn the coupling onto the existing pipe end sticking out of the slab, then turn half the union onto one end of the new pipe section, turn the other half of the union onto the other existing pipe, then try to turn the MPT end of the new section into the coupling. That last part is going to be a real bear. I already tried that quickly and there was not adequate give. I am relustant to heat the pipe which I guess I could do.

I can't disagree on your general comment that this system may have limited life remaining. On the other hand, I have the dough right now to fix this leak. I don't have the dough to upgrade to a new system right now, so I'll have to wing it. This system was installed by an aerospace engineer and he did my neighbor's as well. There are several thinks he did that are giving the system it's current longevity, so it may go for a few moe years.

Thanks again,

11-03-2008, 05:00 AM
Flat faced gasketed unions are about the only way to install a piece when there is no give on either side. That or Dresser type compression couplings with hot water resistant rubber seals.

master plumber mark
11-03-2008, 02:35 PM
Install the dresser coupling with heat resistant rings....

How old is this system anyway???

Why would someone install in steel??

11-03-2008, 10:31 PM

Thanks for all of your great ideas. Sorry to have gone a bit silent. I've been working on this repair.

So, what I tried in the end is a flexible metal hose. It's essentially a 12" piece od SST corrugated with a SSt braid over it and it has MPT 1" on both ends. That has plenty of give in it to easily make up the repair. The union itself must have a burr on it or something because no matter how tight I get it there is a tiny bit of air leaking out. Other than that it now seems pretty tight. I am still worried about having to bury two couplings and the union. I may redo it in the next day or so with PEX and just enclose the PEX fittings in good dusct tpe or some other waterproofing type material.

BIG problem now is that I hear a new small leak just upstream of this repaired section. NHMaster said it right off. That said, I really think this leaky area is all because of the primary leak I already found near that one bad weld, but over teh years there has been much more water and rus in and around this area. So now I need to sleep on the mess and decide if I am going to chase leak #2 down. I have been testing at 40psi air and I am worried that this is a bit strenuous. On the other hand I don;t want to replace the whole floor with new floor covering just to find I get steamy hot moisture seeping out again.

Thx. again,

I'll provide a status update when I get to the next milestone-


11-03-2008, 10:55 PM
Oh, I forgot to mention the age, which someone asked. This system was installed in 1952 and has a great efficiency and fairly rapid response. Like I said it was designed and installed by an aerospace engineer from MIT. The 1" iron pipe is pretty good in that it provides great flow at low pressures. So the system runs at 13 psi supply pressure or so under steady state. It does draw about 50 GPM, so I run a pretty big Grundfoss circulator and a 2" main supply and return to a header which is 80 feet away from the boiler. The iron pipe also gives great thermal conduction into the slab. PEX is great and I use that in the newer portion of my house, so I have a pretty good comparison between the two materials. Remember that PEX has a pretty heat transfer compared to iron, as a matter of fact, most tubing suppliers go far enough to claim thermal insulation as a benefit for hot water supply apps. In the end you do get most of the heat with either material, but the pex takes longer to respond, and gives more of an opportunity to lose heat to ground.

The system I have was installed with 6" of vermiculite / cement mix together, which forms a really nice compact insulation layer below. This is a cheap way to get good r value below the radiant. the pipes themselves seems to sit on gravel and they are only about 1 to 1 1/2 inches below the surface. They also provide a hell of alot of structural rigidity to the slab- essentially giant rebar. The pipes must have been heated and then bent to form the returns, because I don't see any fittings or angles, rather nice 12 OC loops. The seams were butt welded where pipe ran out. This is the source of ny current woes. I actually think there may be stress gradient crack just next to the weld. The leak seems to have emanated at the base of this weld. I wonder if over time this particular weld did not succumb to thermal expansion and contraction of both the pipe and slab so that it built up stresses and eventullly cracked just enough to start intermittent leaking. Could have been coupled with a poor weld in this instance. Will never know. If anyone wants to know more about this system, please do not hesitate to ask- for now more than you asked, but in its day, realy the best one could do and it has lasted 56 years and still gooing strong (atleast 7 of the 8 circuits) cross fingers-


master plumber mark
11-04-2008, 03:54 AM
56 years old, and one inch iron pipe in the slab....wow.....

I am surprised that the re-threading of the pipe did not
disturb the fragile sysem just enough to start another leak...

I hope you have a way to isolate off that one run
because you are walking on egg-shells

that is mean, mean, mean......

if you have threaded the ends, have you considered just
installing a couple of one inch dialectric unions on both ends
and sweat a piece of one inch L copper between them???

11-10-2008, 02:09 PM

Thanks for your post. Yes I considered dielectric unions, but they make me nervous to bury since they have rubber o-ring seals. I am leaning to use pex and protect the brass fittings with something.

Agree with you that this system may be on its last legs- That said, I have continued to work with this circuit (yes it is isolatable and I am running heat in the rest of the curcuits, so I can go like this for a long time) and it's in remarkably good condition. The iron pipe is nearly 100% intact as installed. In many cases the original red dye is still present on the outer surface. The pipes are quite well adhered to the cement and tapping seems not to really move them or bother them.

The original leak occurred at a wels, so I figure that was related to a poor initial job perhaps on that wels or a stress concentration fracture as the slab moved and cycled over the years.

All of that said, I have found two new leaks. Not sure if they were there all along or if I blew them open when I went to 100 psi air. The second leak was on a bend. Never dug it up (just jumped across it) so I do not really know what the cause was. It was just downstream of the big first leak, so it could be that water migrated in there and pooled up- will never know. Once I jumped that loop, this half of the circuit tests out and holds 30 psi air.

Th eother side has another small leak, which I am in the process of finding. (back to the thermal camera). Once I locate and repair that leak I am going to make a decision what to do.

I really like radiant, so I will have to either accept the liability of further leaks or try to pour a really thin radiant system on top of this one. Otherwise I have to go to radiators or forced air, neither of which is very appealing.

Thanks again for your comments, I really appreciate it.