Buried copper piping leaks

dfru

I am troubleshooting an existing building which is experiencing pipe leaks in a buried copper heating hot water distribution system. The piping is insulated and buried in sand a foot or so below a concrete slab. The main pipe run is 2", and appears to run 100 ft straight, without any provisions for expansion or movement (except perhaps some movement within the insulation - which is a rigid foam). There are a series of 2"x2"x1" tees in the pipe, with 1" feeds to heating coils. The failure was at a 1" elbow which is connected to the tee prior to running to the heating coil. I have calculated that the 100 ft run of pipe will expand approximately 1.5 inches (at a 120 degree delta tee). My initial thought is stress related to this. Any thoughts or opinions would be much appreciated.     

icesailor

Underground Copper Leaks:

It is my personal experience that burying copper heat pipes, even in insulation can be a recipe for disaster, even with expansion joints. The expansion pressure as the pipe gets hot will move mountains. As sand replaces the air space from the expansion, along comes the contraction. Which will also move mountains. It will also pull a copper tube right out of a properly soldered joint. If the 2" insulated mains have 1" risers, the expansion will try to break the 1" risers out of the main.

At least that is what I have seen, personally.

It could be a never ending fix.

furnacefigher15

I agree with Ice

Try to find a way to reroute piping over head/above the concrete some how. maybe a soffit. Connections underground are always going to go bad eventually.

Steve Whitbeck

Expansion

Build an expansion loop into the main and put a swing joint in the latteral. should solve most of your troubles ( works best if your leak is in the center of the main)

Steve Whitbeck

loop

How I would build an expansion loop for the main is - (if You have the room ) buy part of a roll of 2 inch soft copper. don't straighten it out. break out enough concrete to put this under the floor. I have done many repairs this way and have never had to re repair one of these.

Mark Eatherton

Got pictures???

Copper has numerous means of failure, including stress from expansion and contraction, It can also fail from excessive velocity, known as hydraulic erosion corrosion.



BTW, if you do decide to install an expansion joint, you MUST anchor the main on the opposite ends of the system in order to direct the expansion/contraction potentials TOWARDS the expansion acceptance device. Otherwise, you are wasting your time and money. And as has been noted, if the branches are not properly set up, you will simply transfer the stress to another part of the system. Outdoor reset controls, and continuous variable speed pumping can go a long way towards alleviating these stresses. They can't eliminate them completely, but they can be alleviated significantly.



ME

Paul Pollets

2"

I'd be using PEX for the main, considering the cost of copper.

SWEI

2"

Given the choice, I'd specify 63mm SDR11 Climatherm.  Far less thermal expansion than PEX plus uber-reliable fusion-welded joints.



Assuming it can't reasonably be dug up, does anyone make a 2" burster head?  You could pull in some HDPE or PEX, then do small floor cuts at the tee locations.

dfru

Velocity

I looked at potential velocity erosion as a factor, but I do not believe it to be the case. Velocities in this system are low - in the 3 ft per second range. The failure was at the heel of the 1" elbow. Crack about 1/4 of the way around the circumference.

dfru

Loop

I was thinking of adding some sort of flexible connection at each of the 1" tees. Any suggestions as to what I might use?

bob young

ex. joint

HOW  ' bout a braided bronze expansion joint. we used them on hi-rise domestic h.w. risers & pump connections.

icesailor

Buried Copper:

You can't bury copper flex joints. They will pull apart.

You can't fool Mother Nature. You can try, but you might fail.

RJ_4

leaks

another thing to check is if the copper pipe was reamed properly, pipe that is not reamed can cause a wearing out of the solder joint.

dfru

Reaming

Thanks. But in this case I do not see internal pipe erosion or velocity as a factor.

dfru

Reaming

Thanks. But in this case I do not see internal pipe erosion or velocity as a factor.

Mark Eatherton

Need more information about the method of failure...

Was the branch coming off the top or the bottom of the main?



Was the elbow pointed parallel to the main, or at a right angle?



Where was the crack in relation to the flow line? In other words, at the beginning of the fluid entrance to the el, or at the end?



You describe a "coil". Is this a fan coil unit, or an in cementitous radiant floor coil?



How long is the 1" branch, and how is it anchored to avoid movement?



Your options of expansion joints are fairly limited. You can do the long legged U tube, or the radi type, but can not use a mechanical joint (trombone with O rings) nor should you use a corrugated type of flex connector below grade. And you definitely can't use any type of mechanical joint below concrete, unless access in the future is guaranteed, i.e. an access panel at the actual device.



As I previously noted, in order for an expansion joint to work correctly, the pipe must be anchored such that the growth potential is direct towards the expansion joint.



Lastly, the only way to determine wether the failure was stress or erosion is to split the pipe and fitting length wise and look for signs of erosion. Remember that horses always walk up stream.



ME

dfru

Pipe leaks

Was the branch coming off the top or the bottom of the main? The branch was coming off of the top of the main.



Was the elbow pointed parallel to the main, or at a right angle? The elbow was at right angles to the main.



Where was the crack in relation to the flow line? In other words, at the beginning of the fluid entrance to the el, or at the end? The crack was at right at the heel (bend) of the elbow. Perpendicular the direction of flow. 



You describe a "coil". Is this a fan coil unit, or an in cementitous radiant floor coil? The coil is at a varaible air volume box which is in a separate vault below the slab.



How long is the 1" branch, and how is it anchored to avoid movement? The 1" branch is insulated, and does not appear to be anchored, except it may be embedded in concrete near the box vault.



Your options of expansion joints are fairly limited. You can do the long legged U tube, or the radi type, but can not use a mechanical joint (trombone with O rings) nor should you use a corrugated type of flex connector below grade. And you definitely can't use any type of mechanical joint below concrete, unless access in the future is guaranteed, i.e. an access panel at the actual device.



As I previously noted, in order for an expansion joint to work correctly, the pipe must be anchored such that the growth potential is direct towards the expansion joint.



Lastly, the only way to determine wether the failure was stress or erosion is to split the pipe and fitting length wise and look for signs of erosion. Remember that horses always walk up stream.

Mark Eatherton

Hmmm...

That doesn't fit what I thought the failure mode would be based on expansion growth potentials. If the main were moving 1.5", I'd have expected the fault to be parallel to the flow, not at a right angle to the flow. That mode of failure would indicate to me significant branch horizontal movement at a right angle to the main.



Have you experienced more than one failure?



How old is the system?



Why the 120 degree DT? 70 to 190?



Is there no outdoor reset on this system?



There is always the possibility that the fitting was defective right out of the box.



ME

icesailor

Underground leaks:

ME,

I have seen, more than once where copper heating mains (2") were placed in the ground and covered with clean, screened sand. After a few years, leaks started. Because the expansion pressures were tremendous. They put thrust blocks on large water service ends. As the expansion occurs, it pushes through the sand in the forward direction. When the pipe cools, it contracts back. While expanded, the sand ahead is compressed. The sand behind falls in behind. After a period of time, the contraction force is equal to the expansion force. And the pipe will pull out of the socket like a frozen pipe is pushed out of a fitting. If there is a riser on the main, it can break off like repeatedly bending a copper wire.

That's what I have seen in relation to the discussion.

And you don't ever want to know how I found out about it. I shouldn't be here for finding out. Foolish. Someone should have known better. And it wasn't me. I just did it.

Along with the fact that you can solder a copper pipe with fuel oil in it because the oil has a higher boiling point than 50/50 solder so you can still solder like you can't with water and steam.

I've done too many stupid things when I didn't know differently.

Mark Eatherton

You're right...

I DON'T want to know what you did in your (hopefully) younger days. Especially if it has to do with fuel oil, torches and copper pipe. ;-)



Just can't figure why a fuel oil line would have such substantial expansion contraction differential, unless the fuel oil was heated due to viscosity.



Copper can do some really strange things, as can all pipes.





ME

icesailor

Sorry,

Two completely different learning experiences.

I'll relate it but my flight is in 10 minutes. I don't type that fast.