Flow rate info for pex and copper tube

hot_rod

The copper tube tables are from the engineeringtoolbox website. Lots of useful info there.
The pex calculator is at the PPI website.

Good to note 3/4 copper tube flow rates. 3/4 M or L can easily handle 6 gpm and fall within the 4 fps rule. So at a 20∆ that equates to 60,000 BTU/hr. With M copper you could probably run up to 7 gpm.

3/4 pex about 4.5 gpm, at 4.1 fps velocity, within the pex tube recommendations.

I'm running a demo in my shop at 6.1 gpm and no velocity noise at all.

33W on an ECM circ to move that flow rate, through a short 3/4 loop in my shop.

EdTheHeaterMan

You have all the fun toys

Alan (California Radiant) Forbes

Thanks, Hot Rod. Can you provide links to that info?

GroundUp

Awesome information Bob, thanks for sharing!

The only thing I'd like to add, is that "short loop" is very important. This tends to get minconstrued sometimes, and folks read it as "3/4 copper will flow 7 GPM". We've all seen zone valves and such with crazy high velocities making little to no sound, but the pumping power required to make such a velocity is often unreasonable anyway unless there is a bottleneck (such as the aforementioned zone valve).

Fo example, I have a loop of 25mm pex (approximately the same ID as 3/4" Type L) from my wood boiler heat exchanger (so it's pressurized to 12 psi) that runs approximately 240ft total underground and connected into a 3 foot loop of 3/4" Type L. There are two Grundfos 15-58s piped in RR parallel, one of which runs a 6 loop radiant floor with 250ft loops of 1/2" pex pulling through a 4.0 cV mixing valve. The other is piped directly through a 12x18x3.5" fan coil with 1/2" copper connections. The flowmeters on the radiant manifold show 4.2 GPM with only that 15-58 running on high. During that same draw, the inline Caleffi QuickSetter is showing approximately 3.0 GPM due to the recirc through the mixing valve, meaning roughly 1.2 GPM is being recycled- the QuickSetter is before the mixing valve so it does not see all radiant flow.

Now the fun part, which always takes me by surprise: when I add the fan coil to the equation using the other 15-58, (so both are running in parallel) the QuickSetter jumps only to 3.9 GPM. Even if I shut off the radiant, the whole loop still flows 3.7 GPM with only the single 15-58 pushing through the fan coil. Everything is silent, as velocity is well below 4.0 FPS, but I'd have certainly thought the two circs in parallel as piped would have driven the flow up closer to 6 GPM based on the curve and pressure drop using this very chart.

While it's a great guideline, sometimes I think both the performance curves and these charts need to be taken less literally and treated more as an educated theory. Radiant loops are a great example of just how inaccurate the information can be, if you actually sit down and digest the math.

Jamie Hall

A minor comment -- pumps in parallel, even if they are identical, don't always behave quite as one expects.

hot_rod

Alan (California Radiant) Forbes said:

Thanks, Hot Rod. Can you provide links to that info?

http://www.plasticpipecalculator.com/

Alan (California Radiant) Forbes said:

Thanks, Hot Rod. Can you provide links to that info?

https://www.engineeringtoolbox.com/piping-tubing-systems-t_6.html

http://www.plasticpipecalculator.com/

hot_rod

GroundUp said:

Awesome information Bob, thanks for sharing!

The only thing I'd like to add, is that "short loop" is very important. This tends to get minconstrued sometimes, and folks read it as "3/4 copper will flow 7 GPM". We've all seen zone valves and such with crazy high velocities making little to no sound, but the pumping power required to make such a velocity is often unreasonable anyway unless there is a bottleneck (such as the aforementioned zone valve).

Fo example, I have a loop of 25mm pex (approximately the same ID as 3/4" Type L) from my wood boiler heat exchanger (so it's pressurized to 12 psi) that runs approximately 240ft total underground and connected into a 3 foot loop of 3/4" Type L. There are two Grundfos 15-58s piped in RR parallel, one of which runs a 6 loop radiant floor with 250ft loops of 1/2" pex pulling through a 4.0 cV mixing valve. The other is piped directly through a 12x18x3.5" fan coil with 1/2" copper connections. The flowmeters on the radiant manifold show 4.2 GPM with only that 15-58 running on high. During that same draw, the inline Caleffi QuickSetter is showing approximately 3.0 GPM due to the recirc through the mixing valve, meaning roughly 1.2 GPM is being recycled- the QuickSetter is before the mixing valve so it does not see all radiant flow.

Now the fun part, which always takes me by surprise: when I add the fan coil to the equation using the other 15-58, (so both are running in parallel) the QuickSetter jumps only to 3.9 GPM. Even if I shut off the radiant, the whole loop still flows 3.7 GPM with only the single 15-58 pushing through the fan coil. Everything is silent, as velocity is well below 4.0 FPS, but I'd have certainly thought the two circs in parallel as piped would have driven the flow up closer to 6 GPM based on the curve and pressure drop using this very chart.

While it's a great guideline, sometimes I think both the performance curves and these charts need to be taken less literally and treated more as an educated theory. Radiant loops are a great example of just how inaccurate the information can be, if you actually sit down and digest the math.

As Jamie mentioned, the behavior of parallel pumps.

hot_rod

Really no good reason to oversize piping either.

Consider the rule of thumb of 40,000, or 4 gpm max. on 3/4 copper.
Jumping to 1" for 4-5 gpm puts you below 2 fps. This can make air removal tougher and you want turbulent flow through a heat emitter to transfer the energy from the fluid stream.

Plus the addition cost of tube and fitting when oversizing.

The charts for realistic flows exist, may as well put them to good use.

With pex, the opposite happens, it often gets undersized (OWF) as folks assume it moves flow similar to copper tube. It doesn't.

GroundUp

With the same ID, it does. 1-1/4" or 32mm pex is almost the same ID as 1" copper or iron and flows nearly identically under the same circumstances. Nominal for nominal, obviously not due to the smaller ID of the same nominally sized pex. To revert back to my original point regarding the "short loop", this was the example I had in my head. When people read that 1" can flow 12 GPM, they often assume loop length is irrelevant. In OWB applications, there is seldom such a thing as a short loop. So while 1" copper may flow 12 GPM with a 26-99 in a 10 foot loop, it's not going to do the same in a 500 foot loop of 1" pex.

hot_rod

Here is some more info regarding pressure drop.

Maybe it is a typo 26-99 to move 12 gpm in a 10' loop?

.040 psi/ft pressure drop in 1" L at 12 gpm flow


100' 1-1/4 pex at 12 gpm= 7.7 ft. head

500' of 1-1/4" pex at 12 gpm = 38' head. Better series a couple 26-99

mattmia2

also, expanding 1" and up pex by hand is not fun