Mixing and Balancing Multiple Zones

arcticcontrolsgoose

Looking to get advice.  

Going through design of a radiant floor system.  Numbers are adding up nicely, but now I’m working on controls.  

I’ve got 5 zones total; 4 of which are about the same avg. temp, one which is higher.  I was thinking of setting those 4 zones on the same dP circ branch, and then the 5th on a separate circ branch.

My design idea is to use a short/fat header and the boiler to inject since the internal boiler pump is too small.  See photo as closest illustration I could find (minus indirect).  The shorty/fat header and closely spaced tees on boiler means all circuits are hydraulically separated, correct?  So I don’t need to balance the two branches against each other (if main pipe sized well)? I would just focus on balancing the 4 manifold loops on the one circ…but the separation and variable pumps are doing their thing for the mains (theory check needed there).

Next.  The circ with the 4 zones is to be a lower temp from the other.  I was thinking of having the boiler to outdoor reset based on the hotter loop design, and then to somehow mix the branch return on the circ with the 4 zones.  I’ve hummed and hawed at how to do this (cheap and effective).  

…Thermostatic mixing valves, I don’t think I have the dT nor constant pressure on the variable pump to make this a reliable option.  

…then 3-way valve I would rather do manual then have a separate outdoor reset from internal boiler.  I see no value in adding a tekmar controller that does roughly the same as boiler but with the single added valve output for hundreds extra.  But with the manual it’s hard to picture what valves and things are needed around it. 

…But then I thought…if I have a BV on the return of this low temp zone, isn’t that directing a certain % of flow back to the circ suction? Is this a more reliable way to ensure flow from manifold circuits to main circ than 3-way mixing valve?  So I’d only need some tees.  And then the supply temp would in theory be somewhat proportionate to the boiler reset if, say, 70% of the return water was set to always go back to supply.  

If this were the case, I’ve done my equations.  The 4 zones have a dT of 10F.  Say this pump needs 12gpm of flow to the zones for this delta under design.  And it takes 3 gpm of the higher temp to “inject” the return temp to do this.  Then setting my BV to a high resistance to return 9gpm to suction would provide adequate flow?  It would also in theory require a smaller main header if only 3gpm is actually returned, plus whatever the other circ returns…and should be proportionate with varied calls for heat…

…am I missing something or does it sound workable?  If so would check valve on supply before and integral to circ be a good idea?  

Big Ed_4

Instal the zones with the largest load first on line . Reverse return the radiant loops off the manifold to help balance the radiation..

hot_rod

First off, what are the two temperature requirements? If they are with 10 degrees if one another, I would not bother with two temperatures 

A manual 3 way mix valve works nicely, it would track with the ODR zones. Very simple to pipe.

A motor could be added if it doesn’t control to your expectation 

arcticcontrolsgoose

Re: Big Ed

Thats another pump added.  Trying to find the most efficient, yet effective solution.  

Re: hot rod

They aren’t.  Delta is ~20 between the two branches. Seems just a little too high for combining, and a little too low for thermostatic valves…

How would you go about the 3-way?  I would assume the flow to want to take path of least resistance, would it be best to divert on return?  Is it really that simple? 

hot_rod

Do the manifolds have balance ability, you can fine tune flows there. 

Most systems could benefit from balance valves if you want to dial them in exactly. Engineers use balance valves extensively to get the results and efficiency they design for.
Variable speed circs can do a lot of flow adjusting, the boiler modulates temperature 

A proportional or “dumb” mix valve rides along with the reset curve.  From Idronics on mix valves

arcticcontrolsgoose

Thanks for the diagrams, helps a lot.  What you’re saying makes sense to me for ease of install and design.  I’m just a little surprised a BV isn’t needed on return to ensure proper mixed flow back to pump.  But I guess it draws what it can where it can…

Would you size the “antler” primary to be total flow of two branches ie 12gpm + other circ… or if I’m always only “borrowing” 3gpm, to size to 3gpm + other circ?  Saves a bit on not needed 1.5-2” pipe…but I also don’t want to have to redo it later… (to meet <2ft/sec, etc for separation).

My manifolds will have flow adjustments.  The loops within the manifolds will be of similar length, but the 4 zones are of all different sizes ranging from 2 loops to 7, so flows differ quite a bit to manifolds.

I’ve attached what I thought I’d need for BVs to balance the flows.  Overkill?  (Pic missing isolation and zone valves, for ease) I think this would make commissioning a breeze…

Also, could I pick your brain on thoughts of reserve return in either the manifolds or the zones within the branch… is one more effective than other?  Should both be done? Or if balancing proper is it not worth the extra piping?

Thanks!

hot_rod

The Caleffi 132 Quicksetter could double as an isolation valve and balance valve easy to use.

Here is how I would pipe it with a Sep 4 to handle air/ dirt/mag separation. The manual mix valve, a couple ECM circs, Webstone purge. Pipe sized to flow needed. Install boiler sensor 12" downstream of the upper right sep connection to monitor what is going to the distribution.

Rich_49

If these are indeed individual zones , your zoning is a solution to the problem which does not exist . Use the higher temp Supply water . Regardless of SWT , when a floor surface temp reaches X , the thermostat should be satisfied and zone valves / actuators should close therefor not allowing the surface temp to climb further . You don't have an issue unless something in your calcs is way out of whack

arcticcontrolsgoose

Would that not have the potential to cause short-cycling? 

Isn’t ideal circumstances for mod cons is low and slow.  Same with heating concrete.  So ideally I keep my zones to design to try and keep my return consistently low to get the most efficiency from system and cause least amount of stress on slab?

Short periods of high heat seems like it could cause headaches…but I’m learning here so I’m not confident I have the theory right…