Diagnosing leaking pipe - condensate vs supply

sreja · April 2013

Hi folks!

Got a question i could use a quick answer to.

We have a supply line to a radiator and the condensate return line from the same radiator running parallel, and buried in concrete.

One or both of these lines is leaking.

The clue is this:

When the radiator is turned off (we have a trv on it), the leaking stops (or nearly stops?).

I initially thought that this MUST mean that the leak is in the condensate return line, since the supply line would be filled with steam whether the radiator is on or off, and the condensate line would only have water if the radiator is on.

But now i'm not so sure my logic is correct. Based on the heat felt from the concrete, i'm starting to think that when the radiator is off, the steam supply to it is filled with air not steam, and so in this case BOTH the supply and condensate lines are essentially "empty" of steam/water.

Meaning that there is no clue to be had about whether the leak is in the supply or condensate line.

Thoughts?

Rod · April 2013

Leak?

Hi- You didn't mention what the indications you have that there actually is a leak?

- Rod

sreja · April 2013

re: Diagnosing leaking pipe - condensate vs supply

actually dripping hot water

this is actually an area between two floors of a 4 story brick+masonry building built in 1929. they were fans of encasing the steam pipes in the 10 inch thick concrete floors...

we've already begun the pain of opening up the floors and jackhammering through concrete..

that there is a leak is undebateble..

the question is.. does the fact that closing the valve on the radiator stops the leak tell us anything about whether the leak is in the supply line or condensate return line..

nicholas bonham-carter · April 2013

Where's the leak

If you close off the steam inlet valve and the leak stops, then the leaking pipe must be on the return. This is unusual, because the condensate is pure water, with no oxygen, therefore no corrosion conditions.

So sorry I had not read your posters name, old contact that you are with our shared Peerless 211a connection!

You may have to finally fix those leaks under the lobby --NBC

sreja · April 2013

re: Diagnosing leaking pipe - condensate vs supply

NBC, my old friend. It's always a pleasure when i see you reply to one of my posts!

What you're saying is what I was thinking originally too.. (I was actually thinking originally that if there was a hole in the supply line then the leak would get WORSE if the radiator was shut off).

But then I remembering everything that Dan and everyone here has written about air having to be moved out of the system before steam can move in, has made me question that logic, and consider whether the lack of air flow through the radiator would mean that the supply pipe to the radiator would not be filled with steam when the radiator is off, making the conclusion that the leak must be in the return line unfounded...

ps.

this is a separate leak than the one under the lobby, and a much easier test case.. and if it's any indication i'd say the lobby leak is looking more difficult than ever to tackle!

Dave in QCA · April 2013

Location of leak

sreja, I think your original conclusion is correct. If there was a leak in the steam supplly line, it would allow anything inside that pipe to leak out, including air. So, I suspect that during a long cycle, the air would leak out, steam would arrive, and steam and some condensate would leak out into the slab. The fact that when you turn off the radiator the leaking stops is a strong indicator that the leak is in the condensate line. And, actually since the condensate in a 2-pipe system intermixes with air, it does come oxygenated and also, because of mixing with the carbon dioxide in the air, becomes very slightly acidic with carbonic acid. In large process applications, one of the treatments added to the boiler is a volatile amine that travels out of the boiler with the steam and helps to keep the condensate in a positive pH. Such treatments are rarely used in steam heating systems.

My building is constructed much the same as yours except they did not run heating piping in the floor. The first floor is piped up from the basement. On the second and 3rd floor, the risers run in chases in the walls and enter the room through runouts that are just above the floor and connect to the radiator. The returns use a side outlet trap with the line running behind the radiator and then back into the wall. Is there any way you could do something like that instead of breaking up the floor?

I am attaching a couple of photos. The first photo is uncropped and unaltered, showing the room with a side piped radiator. The second is cropped and light enhanced photo that shows an up close view of the piping in the upper floors of our building.

sreja · April 2013

re: Diagnosing leaking pipe - condensate vs supply

Thanks Dave (and NBC) -- my steam guy said same thing as you guys, that the original conclusion was correct and that it would suggest a return condensate line leak..

The really odd thing is that we (he) tore up the concrete floor yesterday (10+ inch thick concrete slab with supply and return encased within it, and we ran the steam system last night and spotted only one very small leak from the now-exposed pipes -- a small SUPPLY leak at a joint in the supply line. The location of the supply line leak matches our best guess+evidence of the location of the long term leak before we tore open the concrete.

The condensate line was (nearly) completely exposed and after 8 hours of producing steam we saw no leaks from it..

Conclusions:

1. The clue about leak stopping when radiator turned off tells us the leak is in the return line.

2. The direct evidence that we can after 1 night tells us the leak is in the supply line.

Solution:

Repipe new alternate lines for BOTH the supply and return and leave the old half-buried lines in place. That seems the most sensible solution even though it sure is tempting to just try to replace the visibly leaking joint..

Bonus:

If i had to guess, based on the evidence of where the leak might be before breaking up the concrete, and the condition of the concrete, and then the visible leak in the supply line and no evidence of a leak in the condensate line, i would guess that the only leak has always been at that one SUPPLY line joint and that there is NO condensate line leak at all. If true, something must be logically wrong with our interpretation of this clue about turning off the radiator. Either we are not thinking about the clue properly, or something.. Food for thought or experiment..

Anyway, I really appreciate the feedback and suggestions as always -- it's a real adventure working with old steam systems!

sreja · April 2013

re: Diagnosing leaking pipe - condensate vs supply

Stop the presses -- I just had another theory!

On this particular radiator (about half in the building are piped like this), both the supply and return connect IDENTICALLY to the radiator, but on opposite sides of it. They both come UP from floor and 90 at the bottom of the radiator into it.

So.. If the radiator isn't pitched really well, i wonder if condensate isn't flowing right back down through the supply line!

That would explain the symptoms i think.

Dave in QCA · April 2013

I was ALL wrong!

Ok my theory was all wrong. Aparently the leak is slow enough that it may not leak enough air to get the steam up to the pipe and/or the condensate issue that you have suggested.

My old system was also a very early Dunham type system and as in their early drawings, they showed both the supply and the return connections at the bottom of the radiator. The supplies are a bit on the large size, although not as large as on a 1-pipe system. The returns, (most of them) are connected through an eccentric bushing. The eccentric bushing is abosolutly necessary to drain the condensate when the supply is also at the bottom. If you have a regular bushing, then ALL of the condensate will exit back out the supply pipe. Of course, even if you have an eccentric bushing and the pitch is not correct, as you have pointed out, the condensate will flow back out the supply pipe.

sreja · April 2013

re: Diagnosing leaking pipe - condensate vs supply

Thanks for the follow up Dave..

I wish we had spent a bit more time thinking before tackling a solution -- as it is we opted to repipe both pipes and have run into some trouble with fittings. I'll report more when it's done. If i knew then what I knew now we would have opted to try to replace a smaller section. But hindsight is always 20-20.

sreja · April 2013

re: Diagnosing leaking pipe - condensate vs supply

Just a quick post to summarize the work after completion, for what it's worth.

It took 2 days to expose the supply and condensate mains and fittings, then almost a day getting a broken pipe thread out of the supply main tee. eventually the decision was made to route new supply+return pipes to the radiator -- which was not trivial because instead of running under the tiled bathroom floor they we basically routed mostly in a wall void created by removing bricks.

I do now believe that the leak was located in a 90 degree elbow of the supply line -- we could see the leak after the pipe was exposed.

The initial clue that turning off the radiator stopped the leaks led us all to believe the leak was in the condensate line. But as I posted recently, I now believe that was in error. In reality i believe the radiator was pitched wrong and condensate was actually training from the radiator to the supply line. It may have even been water hammer that exascerbated the leak.

The one lesson I believe i took from this is regarding rerouting the pipes. That turned out to be a scary choice, since the pipe broke off in the main fitting, and was not easy to fix, and carried a real risk of turning into a bigger nightmare, in the future i believe that a wiser course of action when working with such old stuff, would be to give oneself the most chances.. that is, start with the easier repair and see if that fixes it, and avoid eliminating options. Another lesson is to be suspicious of clues and question your assumptions.. steam always seems to hold some surprises.. Last lesson is: always check radiator pitch..

Diagnosing leaking pipe - condensate vs supply

Comments

Categories