Low flow rate through PEX

myqhenryt

I'm currently working on the design for a radiant system for an existing home. Initially, and through the use of LoopCAD, looks like I need five heating zones -- 1 low-temp radiant (Warmboard), 1 mid-temp radiant (Joist Trak), 1 higher-temp radiant (Joist Trak), 1 fancoil (supplemental), and 1 baseboard (basement). Warmboard is already installed and Joist Trak will be installed sometime in the future. Based on the feedback of others in a post last week, it seems that I can get away with combining the radiant zones due to the range of temps. Maybe into one or two zones...I'm still figuring out if I should just simplify things and go with one radiant zone.

Anyway, with a 10* delta-t and 1/2 in pex loops (max 300 ft) LoopCAD has some circuits at what appears to be a very slow flow rate -- .10 GPM for example. My understanding is that velocities should be kept within 2 ft/s and 4 ft/s. That said, should I be concerned with loops running at such a low rate? If so, any recommendations on how to correct this and keep the system working efficiently while keeping the home properly heated?

Here are some other details that may be important:
Boiler: Lochinvar Knight WH-055
Mixing: To be decided -- I'm looking to simplify this as much as possible. Ideally I'd like to use 'dumb' valves. But Lochinvar's mixing control looks attractive.
Manifolds: SS Mr Pex or Uponor
Pumps: To be decided -- I'm still trying to wrap my head around P/S piping if that is the best solution vs. using a hydrolic separator or another setup.
Indoor design temp: 70*
Outdoor design temp: 10*
Heat loss: 45,000 btuh for the entire house (3300 sqft), 26,000 btuh for the radiant areas (1900 sqft). Basement is 1400 sqft
Floor coverings: Mixed -- the majority of the home has R 1.5. There are areas with R3, but it looks like I can properly heat these areas by adjusting flow rates and/or by breaking up into smaller circuits or by having three pipe runs per joist bay.

bmwpowere36m3

Seems normal for radiant floor application, i.e. lower than "optimal" flow rates of 2-4 fps. You could also consider using 3/8" PEX, but your loop lengths will need to be less.

myqhenryt

bmwpowere36m3 said:

Seems normal for radiant floor application, i.e. lower than "optimal" flow rates of 2-4 fps. You could also consider using 3/8" PEX, but your loop lengths will need to be less.

Didn't realize that those low flow rates would be considered normal. I recall reading somewhere here that .10 GPM is too low. Assuming 3/8" PEX is out of the question, the flow rates can be increased but the result would be a higher delta-t, correct?

bmwpowere36m3

.1 GPM are you sure... that's 500 BTU/hr

myqhenryt

.1 GPM per loop, for a handful of them. Attached is a screen cap from LoopCAD with the circuit info. Note this doesn't include the Warmboard loops.

Zman

I see no advantage to such low flow rates. If I recall correctly, you start running into laminar flow issues below about .35 gpm. I would also try to reduce the loops lengths to more like 200 with a max of 250. You will end up with more even heating and efficient circulation (wire to water). Hopefully then you could use a delta p ecm circ like the grundfos alpha. Typically you should be able to be right around .75 gpm per loop. 2-4 ft/sec is hard to achieve with radiant loops, this is generally not a concern.
Carl

myqhenryt

Great feedback, thanks!

Another area that I'm trying to wrap my head around is the use of delta-p and delta-t pumps with the boiler and radiant system. Is there a good "rule of thumb" here or best practices?

SWEI

Thumb-shaped rule around here is to use a ΔP circ when there are zone valves or TRV's downstream of it and a ΔT if it's a single zone (or a large emitter, like an MUA coil) with no control valves.

bmwpowere36m3

Maybe I'm seeing things or not understanding… but for example: kid's bedroom 1, heat loss 1063 btus/hr, 0.1 GPM and delta t 10*. According to the hydronics formula you should have a 0.21 GPM… still small.

2 fps in 1/2" PEX is about 1.2 GPM (per Modern Hydronic book). @Rich commented in my inquires (wrt to 3/8" PEX):

Rich said:

.55 gpm is about 1.835 ft/s & .7 is 2.33 ft/s . I would not worry too much about it though since when the TRV is near closed and as it is closing and opening you will often find your fluid lower than the standard which honestly needs to be brought up to date and recognize the tech we regularly employ at present . That rate was probably necessary with old air scoops but nowadays air bubbles hardly escape a quality air separator . Many radiant circuits also flow VERY LOW flows and I have not experienced an air problem in quite some years .

Raising the GPM will simply lower your delta t. Your design temps are already below condensing (<130*). So running a small delta t really doesn't have a big impact.

Those loops have really small loads… @SWEI knows better, but if it were me I'd be using 3/8" pex.

Zman

SWEI said:

Thumb-shaped rule around here is to use a ΔP circ when there are zone valves or TRV's downstream of it and a ΔT if it's a single zone (or a large emitter, like an MUA coil) with no control valves.

I agree completely.

I am not sure what you have going in loop cad.
I would not have a 144 loop and a 294 loop in the same system.
The max delta numbers are way too high. Someone familiar with that software should be able to get you on the right track.
The goal is to keep it simple. Keep the flow rates about the same, the temps about the same and the loop lengths about the same. Balance the output by varying the spacing and size of the heating panels in each area based on the heat loss. If a room has a very low heat loss, you could only put tubing in the areas you need it. Areas with high heat loss would have tighter spacing, typically near the windows.
Carl

myqhenryt

M3, I'm now looking into 3/8" PEX. Although the cost of pipe and whatnot goes down, I'm concerned about the additional amount of loops that will be needed due to limitations on length. I'll see how things turn out in LoopCAD.

Carl, my understanding was that inconsistent loop lengths are less of a concern when using manifolds fitted with balancing valves. I thought I caught that in Siggy's book. But maybe if I go with M3's recommendation of 3/8" PEX that will give me more flexibility where I can get the lengths somewhat even with additional loops. As far as LoopCAD, there is a setting where you can force delta-T's -- I'll try using that. My only concern with adjusting tube spacing is having inconsistencies with the floor temp where there are hot and cold spots. Maybe this is less of an issue with extruded plates between joists.