Pump sizing (8 loops, 5 zones), 1/2" pex, in Boston climate

ChootarLaal

I have a radiant heat application which is a sandwich type (grooves routed into plywood subfloor and solid flooring on top) and spray foam insulated.
Currently I have 3 zones and 5 loops (<300 ft lengths) but I'm re-doing my basement and will be adding zones and loops, so the total will be 8 loops and 5 zones of <300 ft lengths. I went with 7-8" oc spacing with 1/2" pex and ~1500 sq ft total.
I have a Taco 007 circulator pump but when all of the zones are open, the flow meter is kind of low and I'm thinking it needs some help pushing all that volume of water when I DO tie in the rest of the loops.

Also, I'm looking for a ECM, or varying speed type of pump so I dont have to touch it. The loops are activated off of Taco zone relay and a radiant manifold actuator (24v-2wire) operates each loop. Any recommendations on how to calculate how much gpm and head feet I will need based on my specs below?

The way I understand low gpm should be applied since you want the BTUs to be absorbed by the flooring but another radiant contact said much higher gpm should be applied because if the hot water isn't circulating, then the BTUs aren't transferred, so basically he said,

" more gallons per minutes, more pounds of water, equals more BTUs transferred into area, higher return temperature, but more heat transferred."

1/2" pex
7-8" oc spacing
8 loops, 5 zones, 250-300ft legth max each loop
~1500 sq ft
open cell spray foam insulated
plywood subfloor and solid flooring

SWEI

Any of the residential ECM cirs will handle this. Their curves are all within a few percent of each other and roughly match a 008.

What's your longest existing loop? Try to keep the loops at 200-250 feet if possible.

ChootarLaal

2 of the loops are <100 ft, but the other 6 are very close to 250-280' each.
But looking for recommendation on a particular type or if there exist any ft head or gpm calculators online, based on my inputs.

Jamie Hall

The BTU in a radiant loop may be calculated in exactly the same way as any other hydronic calculation -- BTU equals gpm times delta T times 500. That's for the loop. Now the BTU which can effectively be delivered by that loop to the space is simply the area of the space times the BTU per square foot which can be emitted -- which in turn is related to the way the system is built and the tolerable floor temperature, which in turn is related to the temperature of the water entering the loop.

So you start by figuring out what your particular radiant loop can deliver. Then you juggle the hydronic formula to figure the delta T from end to end of the loop and the gpm in the loop. More gpm, lower delta T which translates into more uniform floor temperatures.

Then once you know what the gpm requirment is, and good head loss computation for the pipe size and fittings (bends, valves, what have you) will give the head loss.

SWEI

I use the tables in the back of the Uponor CDAM for calculating PEX flows: 1/2" SDR9 PEX flowing 1 GPM of 120°F water will create a resistance of 3.47 feet per 100' of tubing, so a 280 foot loop would lose 9.7 feet.

The Grundfos Alpha, Taco VR1816F, Wilo Stratos, B&G Ecocirc 1914, etc. are more than adequate for that load, and yours will be lower due to lower flows and some of the loops being shorter. Given that you can't currently buy a smaller ΔP circulator in North America, pick your favorite color, set the pump on a medium constant or proportional pressure mode and let 'er rip.

Once you've trimmed the manifolds for even floor temps and dialed in your ODR curve, look at the boiler ΔT and adjust the pump setpoint until it's somewhere in the 10-15°F range.

hot_rod

The RPA sells this handy slide rule calculator for quick number crunching. The suggest.6 gpm for 1/2 pex, 300 foot lengths.

Gordy

Agree with all above comments.

Couple other things. Radiant floor loops should be in the 10* delta range for more even panel heating.

Pump head is the highest head loop (longest usually) plus S/R piping to that loop. If the pump can over come that head then it will be fine for the rest of the shorter loops.

Also with zone valves you should be using a delta p circulator. This type will adjust flow based on how many zones are open, or closed. A standard circ such as the 007 would require a differential bypass be used.

Did you use plates in your sandwich method?

ChootarLaal

Do plates really make a difference? I haven't used plates in 2 out of 3 floors but can use them if they are worth it. I'm doing the last floor (lowest, garden level) currently.
I believe there is a differential bypass with the 007 currently, is the diff bypass the small loop which helps alleviate the dead head pressure on the circulator pump?

What's a delta p circulator?

Gordy

Plates do help. See some of @hot rod thermal imagery. Of different plates, and radiant assemblies.

Yes differential bypass alleviates excessive velocitie noise, slamming zone valves,when all zones are not calling, and dead heading.

Delta p is differential pressure. Delta p circ is what you want. It eliminates the differential bypass by performing the same function only better. The ecm circ speeds up, or slows down based on pressure differential saving electricity.

With your present setup the 007 is always drawing the same wattage.

ChootarLaal

Any type or gauge of AL plates will help?
Is Grundfos Alpha 15-55 a delta p circ?

Gordy

Heavier gauge is best with the omega style channel.

Yes the 1555 is .

ChootarLaal

So 0.5-1.0 gpm per loop would be adequate?

Gordy

Depends on your heat loss technically. However .5 is a good starting point. Increase, or decrease flow to individual loops to get a loop delta in the 10 range.

SWEI

The amount of flow you actually need depends on the heat loss and the panel construction (plates or not, quality of plates, insulation, and tube spacing.)

ChootarLaal

"Loop delta in 10 range" means the supply and return temp difference is 10F?

ChootarLaal

SWEI, are you saying there is a 9.7 ft head requirement for each loop, if it's a 280 ft loop? And if I have 6/8 loops at that length then I should be figuring 9.7 * 6, ~60 ft head?

SWEI

I'm saying worst case, if you push 1.0 GPM through a 280 foot length of 1/2" PEX, that it will present 9.7 feet of head. When you pipe in parallel the head stays the same but you add the flows (GPM) together. So if all eight loops were 280' and pushing 1.0 GPM apiece, the pump would have to deliver 8 GPM at 9.7 feet. Make some of those loops shorter and the math quickly gets complicated, but the result will be smaller. Lower the flow by even 0.1 GPM and the number also drops. So 8 GPM and 9.7 feet forms a worst possible case, and given that the pumps can handle that, you don't have to worry about the pump sizing.

Gordy

In a nutshell step one (heat loss) controls the whole design.

Gordy

ChootarLaal said:

"Loop delta in 10 range" means the supply and return temp difference is 10F?

Yes. However this number will vary with load to the loop(s). 10 would be at design outdoor temp. That loop delta would narrow slightly as loop load decreases.

rick in Alaska

Is that 9.7 feet of head accurate? I don't have my chart available, but that seems high.
Rick

ChootarLaal

SWEI,
im still confused. i only have 1 pump for the 8 loops. so when you say 8 gpm and ~10 ft head, im now assuming it would be 1gpm/loop for 8 loops. unless you are assuming i have 1 pump per loop.

Gordy

You take the total flow the 1 pump provides divided by the 8 loops. So avg 1 gpm. IF each loop is exactly the same length. Could be some higher if shorter than longest loop.

SWEI

rick in Alaska said:

Is that 9.7 feet of head accurate? I don't have my chart available, but that seems high.

Piqued my curiosity, so I dug in a bit. Count me confused.
I have a document titled PressureLossTables_H191_1210.pdf that says 3.47 ft/100ft for 100°F water.
My 2011 CDAM says 4.05 ft/100ft for 100°F water.
The 2016 CDAM I just downloaded says 20.61 ft/100ft for 100°F water. ****?

Gordy

Their Getting more pipe out of the same amount of pex. The draw machine is in high speed. that is a bit mysterious seems the 2016 is way out there.

ChootarLaal

Laws of fluid mechanics have drastically been altered in 2016?

ChootarLaal

https://www.uponorpro.com/~/media/Extranet/Files/manuals/PressureLossTables_H191_1210.aspx?sc_lang=en