Why wouldn't this work? (isolating mains sharing a single dry return)

Hap_Hazzard

We all know that, when dry returns are used in single-pipe steam, the returns should not be connected above the water line, as this allows steam to flow from one main to the other, making radiator balancing difficult at best. Unfortunately, this makes it necessary to run separate returns from the end of each header back to somewhere near the boiler where they drop into a common wet return leg.

But what if you used U-shaped traps to create water seals between the mains? The sketch below shows how this would work in the case of two mains.

As far as I can see, only one trap is required, but if you want to be really sure that steam doesn't fill the dry return (which it won't) you could put one on each main extension.

In operation, condensate will fill the trap up to the bull of the tee, and the water level will never go below the bottom of the bull. Condensate is allowed to drain from both mains, but steam won't be able to go from one to the other.

Steamhead

I've done that, and it works. Make sure your water seal is deep enough so 2-PSI steam won't blow it out. And install a drain so you can flush it out, and drain it when winterizing the building.

ethicalpaul

Won't even small pressures push the trapwater out unless you make a really deep trap?

Steamhead

ethicalpaul said:

Won't even small pressures push the trapwater out unless you make a really deep trap?

You'd need at least 3 feet of depth. It's best to have your vents before the loop seal, so they are closed by the time pressure starts to build. That way the ΔP across the loop is minimized, and there's much less chance of blowing the seal.

Hap_Hazzard

Hmmm… 2 psi is about 55" H₂O. Does it need to go that deep? My p'trol is set to 1.5, but it never cycles on pressure. I'm not even sure that the differential pressure would even get anywhere near the system pressure.

Initially, there'd be some pressure from the rising steam, but the vents are open. By the time the vents close, the condensate will already be running. The dead air in the dry return will expand some, which would tend to counteract the steam pressure in the mains…

¯\_(ツ)_/¯

Hap_Hazzard

Hey, @ethicalpaul, where did you get that Pyrex pipe?

Steamhead

Hap_Hazzard said:

Hmmm… 2 psi is about 55" H₂O. Does it need to go that deep? My p'trol is set to 1.5, but it never cycles on pressure. I'm not even sure that the differential pressure would even get anywhere near the system pressure.

Initially, there'd be some pressure from the rising steam, but the vents are open. By the time the vents close, the condensate will already be running. The dead air in the dry return will expand some, which would tend to counteract the steam pressure in the mains…

¯\_(ツ)_/¯

Right. So even if the boiler is at 2 PSIG, you won't have a 2-PSIG ΔP across the loop. So 3 feet of depth should be fine.

Hap_Hazzard

That's not so bad then. Still tempted to pipe it with pyrex so I can see if it moves. I think I'm going to at least put some manometer fittings on it.

ethicalpaul

Hap_Hazzard said:

Hey, @ethicalpaul, where did you get that Pyrex pipe?

Just a tri-clamp sight glass, not a whole pyrex pipe

I made a new video because I just moved my "return sight glass" to my waterline so I could see where it was on the return side just for fun:

https://youtu.be/qFCSPXd-8BQ

EBEBRATT-Ed

You making a water manometer, it will move. One steam supply will always have a different pressure than the other. The deeper the better. 28" drop/psi but yo need to make it deeper so you don't lose the seal and you should pipe it so you cant push water from i main to the other

Hap_Hazzard

How is the pressure different when both mains are connected to the same header?

Jamie Hall

Hap_Hazzard said:

How is the pressure different when both mains are connected to the same header?

Friction loss and varying condensation rates. Particularly if the mains are different lengths -- if they are, it is likely that that trap will see close to full boiler pressure differences at times.

SteamingatMohawk

I had exactly that situation and just ran both returns to the bottom and added a drain. It worked like a charm. See the pics.

Steamhead

Works great right at the boiler. But sometimes you have this situation at the other end of the basement with no way to lower the return line. The loop seal works.

SteamingatMohawk

It took me a while to come up with the idea. I think the reason is that all the sketches I saw in the books, discussions, etc. only showed one return (for simplicity) and I never thought about how my two returns coming together the way they did was significantly different.

dopey27177

Water seals are great they can cure many problems with banging economically.

Sketch taken from my book Steam The Perfect Fluid for Heating and Some of the Problems.

See attachment

SteamingatMohawk

Looks good. Is it possible for one of the returns to be at a higher enough pressure relative to the others to "blow" the water out of the seal" or what is the minimum seal height to prevent it?

archibald tuttle

Hap_Hazzard said:

That's not so bad then. Still tempted to pipe it with pyrex so I can see if it moves. I think I'm going to at least put some manometer fittings on it.

i did that, albeit polycarbonate, which can handle the temps but connection schemes you have to get creative. i was in 1" so i bought a 1/16" wall that was 1 1/8" on the outside from mcmaster-carr and sharked it with a 1" CTS shark x IPS fitting. Was one of the cooler things i've ever done and lived to tell about it.

https://www.youtube.com/watch?v=GCv7iKi0xgg

Hap_Hazzard

archibald tuttle said:

i did that, albeit polycarbonate, which can handle the temps but connection schemes you have to get creative.

That's great! It never ocurred to me to try polycarbonate. It's a lot cheaper, and pretty hard to break. It does absorb moisture, but that probably won't hurt anything.

ethicalpaul

Of course I love this kind of thing. Did you also do the one where the clear tube was put on the radiator supply, @archibald tuttle ?

Also I always have to say you have the greatest name here on HH

Hap_Hazzard

SteamingatMohawk said:

Looks good. Is it possible for one of the returns to be at a higher enough pressure relative to the others to "blow" the water out of the seal" or what is the minimum seal height to prevent it?

Since my mains are nearly the same length, I don't expect the pressure differential to be very high. Steamhead says 3' should be enough to cover it (see the discussion above), and I've never gone wrong when I've taken his advice. It would take about 1.3 psi to displace 3' of water, and my system never sees that much pressure.

Hap_Hazzard

I'm considering adding a water seal to both returns. Here's why.

The biggest concern is what would happen when the boiler hasn't been skimmed in a while and it starts surging. This would send pressure waves out both mains, and, since the mains aren't the same length, they wouldn't reach the water seal at the same time, so the water in the loop gets smacked one way then the other. If the first pulse forces half the water out of the loop, there's less resistance to the second pulse, and it could easily blast out the remaining water.

Adding a second water seal doubles the amount of fluid damping, so it might help to damp out these transient pulses. It also provides a backup, so steam won't flow from one main to the other unless both loops get blown out at the same time.

SteamingatMohawk

With a column that tall, it's pretty close to what I did by tying my two returns into the bottom of wet return. But then it's only a converted 2 family house, not some larger building.

Hap_Hazzard

In my situation it's the difference between running a single pipe back to the boiler or two. That's why I was looking for a solution like this. Since the ends of the mains aren't that far apart, it saves a lot of pipe if I can just run a single return, even if I have to use some to create the traps.

dopey27177

I was in the business for 35 years before I retired and started my consulting business and wrote my book.

To actually see the condensate return from the building nearly blew my mind.

I always did calculations in some buildings on how fast or how much water would return to the boiler in a given time to see if I had to do something special to overcome time lag.

Watching that amount of water drop in a vertical pipe made me realize this water is coming from a horizontal pipe flowing at 3-4 feet a second. This gave me some idea on high the water could stack up in the horizontal pipe and cause water hammer.

This visualization taught me something I all ready knew but never pictured it in my mind as today.

I think in pictures not words and here is a movie I never saww before.

Thanks for the show.


Jake

Hap_Hazzard

When I was a kid, there was this science fiction concept of "see-through metal." I wish I could repipe my system with that stuff, but here we are, half a century later, and they still haven't invented it!

Still waiting for my jetpack too, dammit.

archibald tuttle

@ethicalpaul glad you can appreciate the reference of my nom de plume. I can't claim credit for the clear radiator feed but now i want to go look, i assume maybe to illustrate counterflow or . . . ?
@dopey27177 i agree on the picture thing. this was a job where i wanted to ditch a complicated multi trapped returns to a condensate tank and pump. There had been a snowman that got replaced by noticeably shorter burnham atmoshperic. I think at vaporstat pressures it didn't even need the pumped return with the original boiler but gaining a foot and half I was sure of it according to theory. But I had to be 100% sure in real life before committing to extensive repiping that it would be expensive to walk back from so this was my empirical validation run, i.e. my picture - worth 1000 words. I think you can see i wasn't even a foot about the boiler water and I didn't have the camera on the vaporstat but i had a digital hooked up and at the high end i've got 16 or 18 inches of water but i'm going to assume that there is modest equalization around main such that i only experienced something like 12 inches of actual condensate height above the weir of the return (gifford style in this case)

@Hap_Hazzard i've got twin mains like this on boile, but they have F&T traps before the dry return because it runs to condensate tank with atmoshppheric vent (although this will be my next experiment for getting rid of unneccessary pumped return. But, in your case, if these are dry returns to a gravity hartford or gifford loop, does it really matter if they are separated If you aren't separately operating them with a zone valve so deliberately keeping one off and one on on occasion? is there really that much of a problem with a little cross talk on the steam or would it really be little different from a boiler that has two take offs that are teed in a drop header? So if one side is a little faster you might experience a little steam rounding from the shorter main into the longer until the flow was balanced. And you would get a little counterflow near the end of one main if it operated this way but with the sizes of pipe involved and the capacity of steam needed to be carried to those farthest extent of the mains is that really a problem?

so anybody who wants to contemplate any of this in Orlando I think we ought to have us a little 'Eatting and Cooling" get together. I'd say drinking and cooling, but not as good a pun and makes us sound like lushes - probably true but i didn't want to tell everybody.

Hap_Hazzard

@archibald tuttle The problem with "cross talk" is that the steam from the first main to close its main vent can cross over to the other main and close its vent before all the air has vented. If the air trapped between the steam fronts spans a radiator branch, the trapped air can only vent via that radiator's vent, putting it at a real disadvantage as far as getting steam,