pump placement in a radiant floor circuit


  #1  
Old 10-12-08, 11:36 AM
S
Member
Thread Starter
Join Date: Sep 2008
Location: Central Iowa
Posts: 42
Upvotes: 0
Received 0 Upvotes on 0 Posts
pump placement in a radiant floor circuit

I am installing a radiant floor heat (RFH) circuit into a radiator system. The RFH loop will be fed by a Webstone purge tee put into the boiler primary loop. I know the tee must be placed carefully relative to primary loop bends (≥ 6 pipe diameters). From the purge tee to and from the manifold will be 1/2" copper, with a feed pump (Grundfos, can post the # if useful) for the hot side of the manifold.

For placing the pump between the purge tee and the RFH manifold, does it matter how close to one end (purge tee) or the other (manifold)? Total run to the manifold is about 10 feet, 5 or 6 elbows. I can conveniently place the pump close to the purge tee or very close to the manifold.

Is it OK to have elbows right away on the copper off the purge tee, or is there a similar requirement about a certain number of pipe diameters?

Anything else I should be thinking about?
 
  #2  
Old 10-12-08, 10:27 PM
Who's Avatar
Who
Who is offline
Member
Join Date: Jan 2006
Location: San Jose
Posts: 2,066
Upvotes: 0
Received 2 Upvotes on 1 Post
Was the radiant floor heating designed to use the same water temperatures as the radiators? If so, then you could just pipe it as a parallel circuit, use a simple pump relay like a Taco SR503, wire a t-stat and that floor could run its own zone with no need for fixing. If it was designed to exactly replace the BTU emission of the previous radiators or current heat load for that space, then you could conceivably divert some flow throw it from the current pump.

If the radiators will need higher temperatures than the floor, then you will have to put the circ out in the loop with the emitter on something like a Taco 5000 which is actually more fro going on a water heater, to be sure that someone doesn't get scalded. Also to keep the temperature from rising on a pressure drop between hot and cold.

How was the radiant sized?
 
  #3  
Old 10-12-08, 11:20 PM
S
Member
Thread Starter
Join Date: Sep 2008
Location: Central Iowa
Posts: 42
Upvotes: 0
Received 0 Upvotes on 0 Posts
Thanks for the response!

The radiant was sized for heat loss from the zone. The 2 rads it is replacing were capable of more BTUs but the RFH should suffice; there was a slight remodel so there is also more air flow now to the RFH zone than before. In fact the calculations said "radiant should just do it here" so this is a bit of an experiment - I may be adding back a radiator or some baseboard down the road. Hope not.

So the radiant is designed to call for heat when it wants it, and take it from the front end of the primary loop, I think almost exactly as you suggested Who:



Look OK?

My question is pretty mundane, it is about pump P3. Is it better for P3 be real close to the boiler, or to be real close to the manifold (there are three PEX loops off the manifold)? Or does it just not matter as long as it is in the copper line?
 
  #4  
Old 10-12-08, 11:54 PM
Who's Avatar
Who
Who is offline
Member
Join Date: Jan 2006
Location: San Jose
Posts: 2,066
Upvotes: 0
Received 2 Upvotes on 1 Post
P3 is fine where it is, as long as both radiators and radiant are working at designed for the same water temperatures. I'd suggest one change. Run the radiant as a parallel heating circuit so that there is only P/S point. Your way the rads may get colder or hotter water depending on if the radiant circuit is calling for heat.
 
  #5  
Old 10-13-08, 04:40 PM
NJT's Avatar
NJT
NJT is offline
Member
Join Date: Oct 2006
Location: USA
Posts: 19,710
Upvotes: 0
Received 8 Upvotes on 6 Posts
Ideally, I would say about a foot of straight pipe ahead and after the pump.

The idea behind close spaced tees is for as close to zero pressure difference between them. putting elbows immediately at the ports is probably not a good idea. On the other hand, as long as BOTH ports were treated the same way, you would probably be OK ...

I think I understand that what Who is saying is that the cool water returning from the radiant will dilute the hot water going to the radiators... I realize that it won't be a heckuva lot through that 1/2" circuit, but he's probly right...

How many BTU you counting on for the radiant ? You sure that 1/2" is big enough to feed three loops ?
 
  #6  
Old 10-13-08, 07:57 PM
S
Member
Thread Starter
Join Date: Sep 2008
Location: Central Iowa
Posts: 42
Upvotes: 0
Received 0 Upvotes on 0 Posts
OK thanks guys, I really appreciate the feedback.

Originally Posted by NJ Trooper
Ideally, I would say about a foot of straight pipe ahead and after the pump.
I like shooting for ideal. Of my alternatives with convenient access, one of them places the pump in the middle of a 6 foot run, sounds like I've got a decision made there.
Originally Posted by NJ Trooper
The idea behind close spaced tees ...
OK how about I go with 6" of 1/2" copper on one branch, 7" on the other, before I place an elbow in each line, sound OK?
Originally Posted by NJ Trooper
... the cool water returning from the radiant will dilute the hot water going to the radiators...
Yes, I see that is a good point. Here is the rationale: the radiant may want to run hotter than the rest, that is the basis for putting the tee there. FYI it is a high mass system, converted gravity flow with 3"-4" in the basement near the boiler and a lot of 2" in the walls. It rarely runs very hot, usually 125* (135* peak last year), I think by calculations the radiant needs to run 140* on peak demand. Also, if I return somewhere else I won't be using the closely spaced tee, I guess that is useful for ghost flows?
Originally Posted by NJ Trooper
How many BTU you counting on for the radiant ? You sure that 1/2" is big enough to feed three loops ?
Well no I'm not sure but they are kinda small loops. Heat loss calculation for the space is almost exactly 10k BTU. 3 loops total 500 feet of 1/2" PEX snapped into heavy gauge aluminum plates. Is the 1/2" copper big enough? The contractor who designed the system (those aren't my drawings!) thinks so.
 
  #7  
Old 10-14-08, 04:11 PM
NJT's Avatar
NJT
NJT is offline
Member
Join Date: Oct 2006
Location: USA
Posts: 19,710
Upvotes: 0
Received 8 Upvotes on 6 Posts
I think even on the low speed, that 15-58 will pump more than you need...

I think 6" would probably be OK ... but, why not just turn the webstone 45° on the pipe, and use a 45° at the other end, if you've got the room ? (without seeing the 'physical', I can't visualize what the intent is though... so this may not work for you)

I believe you will find that 140° will be too hot if those plates are stapled to the underside of the floor. You may find that you end up installing a tempering valve and dropping the temp in the radiant, and adding that length of fin-tube that you don't want to use!

What Who was trying to say was to pipe the radiant in parallel with the other heating loop, utilizing the same set of closely spaced tees. Tee off the supply loop upstream of the other pump, after the CSTs, and return upstream of the return CST. You would need flow checks on both pumps. Install the tempering valve whether you think you will need it or not. If you don't need it, you can always set it at a higher temp, or at least LIMIT the temp to the radiant. Save the bucks on the purge tee and put it toward the tempering valve.

You can move 15K BTU through 1/2" pipe with an acceptable flow rate. What I fear might happen is that when that small flow in the main run hits the manifold and splits into three, the flow will be unacceptably slow in each individual loop. You want to maintain something like 2 FPS in order to be able to remove air from the loops. Why not choose 3/8" for the radiant loops ?
 
  #8  
Old 10-14-08, 08:43 PM
S
Member
Thread Starter
Join Date: Sep 2008
Location: Central Iowa
Posts: 42
Upvotes: 0
Received 0 Upvotes on 0 Posts
The reason for the 6" is so I can go vertical with the copper, to get it up and out of the way to where I want to hang the pump. I mean, there is room to go straight, but I wanted to leave that space empty.

Here, below is an old pic, the wiring is no longer like that (this forum helped me fix that!) but you can see the piping. The manifold is also not in this old picture but it is mounted on the right hand wall.



So I could run 1/2" copper L-R and go pretty much straight from the webstone to the manifold, but my plan was to keep the space to the right of the boiler empty, seems a shame to block it with pipe if I don't have to.

So the webstone goes in the supply pipe, then short 1/2" copper runs with elbows to get up above the boiler, then a L-R horizontal run up high. That straight run would be the place to hang the new pump, above and slightly behind the existing pump. Get copper to the wall and run down to the manifold.

Sensical?

p.s. no comments on the decrepit ceiling please
 
  #9  
Old 10-18-08, 10:02 PM
S
Member
Thread Starter
Join Date: Sep 2008
Location: Central Iowa
Posts: 42
Upvotes: 0
Received 0 Upvotes on 0 Posts
I'm not following the combination of these two quotes:

Originally Posted by NJ Trooper
I think even on the low speed, that 15-58 will pump more than you need...
...
What I fear might happen is that when that small flow in the main run hits the manifold and splits into three, the flow will be unacceptably slow in each individual loop. You want to maintain something like 2 FPS in order to be able to remove air from the loops.
Couldn't I just turn up the 15-58 to medium speed?

How do I decide how fast to run the 15-58?



Too late to go to 3/8" plates now, the 1/2" plates are installed.

I can't easily skip the purge tee because I think there is simply not enough pipe length to drop a feed between the CSTs and the first pump. See photo in my last post.

Yes the 1/4" aluminum plates are fastened to the subfloor. Does it help that the finish floor is tile (large thermal mass, correct?), as far as hitting it with 140* water? And I was worried it would not get warm enough!

I'm thinking I'll try learning the hard way on the tempering valve. The way I see it, replumbing later for a Taco 5000 would be cheap and easy, but right now the valve is more than $100 so I'd rather wait and see if I need it. So I'll leave it out for now. At all reasonable?

Would outdoor reset help with the mixing valve issue? My Munchkin 140M (yes I know it may be oversized, again too late for that, it went in a couple years before I found this site) did not come with ourdoor reset. Right now I am dropping the max T manually as the weather gets colder.
 
 

Thread Tools
Search this Thread
 
Ask a Question
Question Title:
Description:
Your question will be posted in: