Manifold vs Pump Array?

RobinInCali

I am relocating a boiler panel and wanted some professional advice. If you had an issue with a bad tubing install (can't be changed at this point) which consists of three loops which are way-too-different lengths each, causing uneven heating, and you need also a Delta T of at 10 degrees, which means a higher flow rate:

Which would accomplish that best:

1) installing a Watts 1.5" M style manifold with adjustable loops (2-7 GPM each loop) and one Pump, or

2) installing a copper manifold with 3/4" outlets and individual pumps for each outlet/loop? Maybe adding to each pump/loop something like a Caleffi QuickSetter unit? or

3) insert your idea here :-)


Thanks for any advice.

Rich_49

Manifold with loop balancing capability

hot_rod

Do you have a gpm requirement for the various loops? One circulator with manifold and quicksteps. The Quicksetter windows do not fog or get cloudy if you need to confirm or change some day.

RobinInCali

I have three loops of staple up that need lots of flow to keep the house warm--1860s era Victorian--I am guessing 2-4 gpm per loop. I have a 1.5" Watts M Manifold I could use. Each loop adjusts 2-7.

The Quicksetters sound good with individual pumps. But the Watts manifold would be cheaper with less stuff to break.

I have two loops of slab radiant in the basement that work fine, such that they only need a couple hours a day at design temps to keep the basement cozy.

GroundUp

What size is the tubing and how long are the lengths? Achieving a 10* delta may be easier said than done

hot_rod

3 loops of what diameter pex and how long of a loop?

What is the floor build up over the tube.

The RPA RadPad suggest 1.2 gpm max. for a 500' loop of 3/4 pex.

Ideally you would find the load for the areas served, then work the numbers from the load to the tube layout, SWT and flow rate.

With a thick floor build up, inadequate tube and high load rooms, radiant floors may need supplemental heat of some sort.

RobinInCali

The tubing loops--as I said, it was a bad install by the former owner--is something like 250', 400' and 550' of 7/8" pex, spaced 16" on center. 1" sub floor with 5/6" oak parquet above.

If I could rip it out and do 1/2" pex with Thermofin C, 8" on center, I would. But I'd have to tear out a ceiling over 900 square feet of living space. $30k? $40k?

The system does work. It does heat the house, albeit unevenly and with too much waste. I just want to replumb it as efficiently as I can. Experience has proven that if I can get a Delta T of 15 or less, I'm good. I'd like to aim for 10.

Hope this info works.

In order to get Delta T of 10, what does the RPA RadPad suggest re: how much flow for each loop of 7/8" Pex?

RobinInCali

More clarification: I have three loops for the staple up Zone (off one Alpha Pump) and two loops for the slab Zone (off the other Alpha pump).

One mistake I've considered is (and I will correct this for the new panel I'm doing in the Boiler relocation) the slab zone does come first in the sequence. It is half the size, and pulls lots less than the staple up zone.

SuperJ

I think you need to think of heat output in terms of BTU/sqft or maybe linear foot of tubing rather than "delta T".

Delta T is just one part of the the BTU equation. Example going from a 15 to 10 deg F delta at the same BTU rate, requires 50% more flow.

BTU/hr = delta(f) x flow (gpm) x 500

Keep in mind adding flow almost always lowers your delta (thus negating significant BTU/rate gains). So it doesn't help you as much as you might hope it will.

If you want a to actually up the BTU output significantly while reducing the delta T you need even more flow. And as you add flow, you get incrementally less BTU per gpm added. Eventually you get to a point where adding more flow adds virtually no extra heat since the average water temp is basically even across the whole emitter.

I would clean out your system as best as you can, and pump it reasonably. You may not be able to get down to a 10deg F with a reasonable pumping arrangement. if you don't have enough heat at that point I would consider supplemental heat like panel rads. They can be installed much less intrusivally than tearing out ceilings.


Attached is some reading.

hot_rod

Actually the higher flow rate, tighter delta results in higher average loop temperature and higher output.

Good explanation of flow basis and how they effect heat transfer here.

https://www.caleffi.com/sites/default/files/coll_attach_file/idronics_16_na_0.pdf

This example shows a 1 and 2 gpm flow rate, in this 1/2 pex loop. About a 12% increase in output at higher flow, smaller delta.

I suspect you have 3/4 ID, 7/8" OD tube. At 2 gpm, 550 feet pressure drop around 15'.

Only so much you can do with the hand you are dealt, 16" spacing is another deal breaker.. Higher supply temperature, as much as the wood is comfortable with would be another way to increase output.

Of course the math here blows the constrained ∆T theory apart

RobinInCali

12% increase would make me one happy customer.

As I said, the system works well enough 10 months out of the year. In the deep cold of January (San Francisco can get all the way down to 40 degrees! Ha Ha) we do turn on space heaters to get from 66 to 68/69 in the house. 12% would get us back up to 70. That sounds awful warm, but not with 12 foot high ceilings.


Thanks everybody! This is great information!

SuperJ

Thanks Hot Rod, I agree with your math 100%. I'm just flinch at throwing more pump at a problem. (once your past a reasonable design point).

Speed 3 on the Alpha would get you pretty close to the operating point @Hotrod suggested.

hot_rod

SuperJ said:

Thanks Hot Rod, I agree with your math 100%. I'm just flinch at throwing more pump at a problem. (once your past a reasonable design point).

Speed 3 on the Alpha would get you pretty close to the operating point @Hotrod suggested.

Yes adding a lot more pump for a small increases the top of the graph shows... the juice may not be worth the squeeze.

if in fact he is only a few % from the comfort he desires, may be it is worth it.

RobinInCali

I agree that it's not worth throwing too much at this system. But since I'm relocating it anyway...

Do you also think that balancing the loops (in the staple up) will make a big difference? Before this thread, I was focusing on that as a larger issue than pumping. Cause I mean, how can it pump at all if the loops are balanced? More pumping is just going to move it more in the short loops and less in the longer ones, no?

SuperJ

Balancing is definitely worth while (or the long circuits are going to suffer the most), but to get the most out of balancing you need to make sure your pumping is adequate.
Right now you have all 3 staple up circuits on one pump.
It looks like you might have to pump that worst case 500ft circuit separately. (Existing Alpha 2 at speed 3). Need to figure out your flow and pressure drop targets for each other circuit.
You might be able to combine the 250ft and 400ft circuits onto one pump though if you at minimum put a QuickSetter on the shorter circuit.
Even at a fixed speed ECM pumps are still worth considering since they are much more efficient.

Don't forget to install the check valves if you end up with a bunch of parallel pumps.

This could be an opportunity to zone it out a little more since you have to pump some of your circuits separately anyways.

hot_rod

Yes you really need to balance loops that are that mismatched in length. If you use Caleffi Quicksetters you can balance and observe flow. That is a simple first step. You may need a higher head circulator, depends on what the current one will do after you balance all the loops.

140 SWT with 68 ambient, R4 floor gets about 17 BTU/ sq ft.

Bumping to 150 SWT will give you 21 BTU/ sq. ft. Again some ballpark numbers from the RadPad. Better number crunching could tighten up those estimates.

Although with that wide spacing, and high SWT expecting some striping.

RobinInCali

Caleffi quicksetters with individual pumps are better than one pump with a Watts 3 branch manifold with loop adjustments? Why?

SuperJ

Maybe not better or worse, but it depends on the pump you want to use. Your scenario is a bit different since you are trying to solve a pre-existing problem rather than design a new system.

As @Hotrod said you may need to look at higher head circulators to meet your needs. (like the 15-100, or possibly the 15-58). If you get a pump that handle your total flow at the worst case head pressure, than 1 pump can do it.nnA single pump that can move about 8 GPM (guess) at 14-15fthd could be tied to a manifold for easy balancing. Or you could pump each circuit separately as needed. It kinda depends on cost and complexity and materials already on hand. You just have to make an informed decision and go with it.

Starting from scratch the manifold with a single pump would be more elegant and probably give you a cleaner install.

There may be a hybrid approach where you use the manifold and a common lower head pump (sized for total flow, but only the head pressure of your shorter loops), and install a inline "booster" pump on your long circuit to make up the extra head pressure required.

Putting pumps in series causes their curves to stack vertically. (This was pretty common in geothermal loops when I was involved, often 1,2 or 3 26-99 pumps in series. We needed to get about 3gpm thru a bunch of 500' parallel 3/4" hdpe circuits)