PEX Press fitting diameter question

Scott_Mountain_View_CA

The ID on the PEX Press fittings is 1/8" narrower than the FostaPEX tube. A 1/2" nominal PEX Press fitting has an actual 3/8" ID. How does this effect water flow? I've got dozens of 3/4" & 1/2" nominal fittings in my system; is there a cumulative impact to flow rate or is it a non-issue? My understanding that it's best to minimize elbows in general, so I would assume that adding an elbow AND narrowing the elbow ID to 3/8" on a 1/2" line would exacerbate the flow impact.

hot_rod

Yes you are correct that any fitting, including a reduced fitting size adds pressure drop to a circuit. You could calculate each fitting and foot of pipe, or simulate it with software and come up with a number.

Generally we turn all the fittings into EL equivalent length of pipe. then use that to determine the flow resistance.

Here is some of the calculations, as well as a link to a journal that may help seeing the entire design process and steps needed.

I doubt that the number of fittings you will be using will have a major impact of the flow you need to get the job done Crunching the numbers is the only way to really know.

I'm not positive what EL value to assign a pex insert fitting, somewhere between a 1/2 and 3/8 ell I suppose. Perhaps a fitting manufacturer has that data somewhere?


http://www.caleffi.com/sites/default/files/coll_attach_file/idronics_12.pdf

jonny88

In my experience with pex I will only run a homerun system,one line to each fixture.

Scott_Mountain_View_CA

Thanks. I hadn't considered the ID loss of PEX fittings, nor did it come up in my lengthy thread during the planning stages. The 1/2 tubing on each of my 6 rads drop down from the top of the wall betwen stud bays. Given the space constraints of 3.5" deep wall stud bays it is impossible to avoid at least two 1/2" nominal (3/8" actual) PEX elbows for input and return tubing inside the walls. That's four x 6 rads = twenty four 3/8" actual elbows. In addition, each of the 6 rad supply 1/2" and 6 return 1/2" will require a 3/4 x 1/2 x 3/4 tee....or another twelve 3/8" fittings. I've also got twelve 3/4" nominal (5/8" actual) elbows on the main supply/return.

I'll do the math this evening because I need to better understand the technology since this first hydronics project won't be my last.

It's been stated before that pump size can trump the flow issues....is that still true given the number of small ID fittings I'm using?

Edit: I just downloaded from the Viega web site both the FostaPEX tubing and fittings head loss tables. Since the fittings table takes into account the reduced ID, the calculations Rod suggested above should be straight forward.

Thanks again for the comments and help.

Scott_Mountain_View_CA

Oh man.

Attached is the Viega Press PEX fittings head loss table in equivalent pipe lengths. Big difference vs the copper Table 5-13 posted above by Rod. For example, a 1/2" nominal PEX Press Poly 90 elbow adds 12.6' of pipe vs Table 5-13 copper 1/2" adds only 1 foot. And on a 3/4 x 1/2 x 3/4, the the PEX adds 17.9' total for tee run & tee branch, while Table 5-13 shows that same fitting in copper adding only 2.4' of equiv. pipe.

When I do the math, my total system in equivalent pipe length is about 950', or about double what it would be if I had gone with an all copper system. I guess the big question is what size pump am I going to need. Total GPM all rads = 1.67

I'm having a hard time understanding how the flow resistance of the PEX fittings could be so much greater than copper. Am I missing something?

Bob Bona_4

The 90s don' have a smooth internal curve. But generally you use less of them. Usually a wash over copper.

Scott_Mountain_View_CA

The 90's I can understand the elbow...maybe, but the 3/4 x 3/4 x 1/2 tee is 17.6 equiv feet.....same fitting in copper is only 2.4', the delta doesn't make sense. And I've got 24 of those poly tees in my design. Those alone add 422 feet of equivalent pipe. Does that sound right to you? In the end it may not matter because my total rad GPM is only 1.67. Lot of guys here using PEX fittings, yet no one has mentioned the huge impact to flow rate which makes me wonder if I'm misunderstanding the Viega table.

If I use PEX tubing with copper 90 els or long els, I'm gonna be shelling out more $ in PEX-copper/copper-PEX adaptors than just buying copper tube all the way around.

hot_rod

copper fittings go over the outside of the tube, they are generally manufactured from the same tube and expansed, swaged, on the ends to accept the copper tube. So very low pressure drop via ells.

Remember when you calculate the pump head, just use the longest, most restrictive loop, not the entire 900 feet of piping. The gpm is the total of all the loads however for pump sizing.

The info to calculate all that is within Idronics 16.

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

Siggy did a seminar up at the Hydronics Expo in Toronto last month on High Efficiency Hydronics. One of the thoughts had to do with the use of bent tube instead of ells for hydronics. Here is a slide from his presentation.

The CDA claims a long radius bend like this does not add any pressure drop to a piping circuit.

That is my radiant garage floor you seen then pic

Scott_Mountain_View_CA

Rod, thanks for the response. I'm a bit confused by "Remember when you calculate the pump head, just use the longest, most restrictive loop, not the entire 900 feet of piping." I'm running a single loop 2 pipe parallel reverse system. The fittings are concentrated in he region of each of 6 radiators, the others being positioned at corners in as the mains wrap around the house exterior. Could you provide a little detail on what you meant?

Last night I did a detailed list of every fitting used, categorized by brass, pex, or copper and their associated head loss in equiv pipe feet. I'm late to get my wife to the airport, I'll forward my data later. Bottom line it indicates total head loss at 21! I'll need to convert part of my system to copper looks like.

Bob Bona_4

Bend as much as you can. I'm doing home runs to radiators in slab, 4" radius. No joints.

Scott_Mountain_View_CA

I'm bending where I can. Problem is I'm working inside 2x4 stud bays 16 oc, post and beam construction so also have to deal with 4x4 posts, window header & trimmers. It's too tight to make many bends in the Fosts-PEX. I had originally planned to use Uponor AqauPEX but was convinced by some on my original thread that wall noise would be an issue due to coefficient of expansion. What I failed to understand is how much more difficult it is to bend the Al tubing.

It looks like I have a few options:

1. Stick with my current plan and just get a bigger pump - Taco 10 would cover the head loss. Would add $250 to the project.

2. Run 100% copper main supply/returns, eliminating all 3/4" FostaPEX and replacing all the PEX poly tees and the majority of elbows with copper fittings. Haven't calulated the cost yet.

3. Stay with the 3/4 FostaPEX main supply/return which I've already purchased but replace all PEX supply/return tees/elbows with copper by inserting 2 pex-copper adaptors at each fitting.

I still plan to use the 1/2" FostaPEX inside the walls because of it's flexibility.

Do you understand what Rod was saying: "Remember when you calculate the pump head, just use the longest, most restrictive loop, not the entire 900 feet of piping." ? Again, I have a single loop parallel reverse return system...seems that I would need to use all fittings/tube in the head loss calculation?

The Viega PEX Press poly fittings head loss is astounding...shame on me for not researching this sooner. Their elbows are particularly bad.....I don't understand why they didn't design them with a sweep bend rather than a sharp 90. I'm calling them Monday to ask that question.

I looked for but couldn't find a head loss table for the Uponor fittings, to compare.

Harvey Ramer

Scott, their are very few places where you actually need to use a 90 fitting. For example, in a place where you can't readily make a bend due to the length of pipe and obstruction, you can usually cut the pipe, bring it in from both ends, leave the one pipe short, put a bend on the other pipe and join them with a straight coupler. Dramatically less pressure drop that way.

Hope that makes sense to you.

Harvey Ramer

As for the pressure drop on the branch of the tees,

So you are doing a parallel reverse return and you are taking 1/2" branches off of a 3/4" or 1" main.

The way to calculate it is as follows. Get the equivalent length of each branch run to the rads individually. Check them against their respective calculated flow rates (individually) to find the one with the highest pressure loss. Record the highest pressure loss and discard the rest.

Now calculate the equivalent length of your main in four sections.
Section. 1 Supply from boiler piping to first branch.
Section 2 Supply piping from first branch to end of main.
Section. 3 Return piping from boiler piping to first branch.
Section. 4 Return piping from first branch to the end of the main.

Calculator the respective pressure drops for each section of the main at the system calculated flow rate.

Compare the pressure drops of section 4 and section 2. Choose the highest pressure drop and discard the other. Now add this pressure drop to section 1 and section 3 and the pressure drop retained from your branch calculations.

The result will be an accurate indicator of your total system pressure loss.

Scott_Mountain_View_CA

Thanks Harvey...I'll do those calcs and post results. One question: Can I assume that for Items 1&2 you're traveling downstream from the boiler (i.e., with the water flow direction), and for Items 3&4 you're traveling upstream from the boiler (i.e., against the water flow direction). Is this correct?

Harvey Ramer

Correct.

Scott_Mountain_View_CA

One thing sticking in my craw is the difference in head loss between Uponor ProPEX fittings and Viega Press Pex fittings.

Example:
1/2" ProPEX poly 90 el - 3.7' equiv pipe length
1/2" Press PEX poly 90 el - 12.6 "

I do not understand how these two similar fittings could be that different in pressure loss. From an engineering and materials perspective I don't see how this difference could be possible without Viega designing sandpaper on the fitting inner walls. I'll be talking again to Viega tech support on Monday for an explanation.

Bob Bona_4

Uponor pipe and fitting OD is slightly bigger. If it helps, make up your Tees with copper/sweat pex adapters. I do this in toe kick installs with monoflo T/conventional T.

Scott_Mountain_View_CA

I pulled specs from Uponor and Viega sites. HePEX and FostaPEX tube ID's are identical.

Where Uponor has a big advantage is in their EP ProPEX fitting ID. Here's data for 1/2":

PEX Press 1/2" fitting: 0.315 ID; 1.9 GPM @ 8'/sec
ProPEX EP 1/2" fitting: 0.385 ID; 2.9 GPM @ 8'/sec

FostaPEX 1/2" tubing ID 0.475
HePEX 1/2" tubing ID 0.475

Bob Bona_4

Compare Uponor MultiCor and fittings for apples to apples PEX-AL-PEX. I would not use Hepex for "high" temp applications. Too much expansion.

rick in Alaska

Uponor fittings are expansion fittings, which have the same inside diameter of the pipe. Not sure if Multicor is an insert style, but if it is, then that would make the head pressure difference as there would be a restriction at that point. All I use is the expansion style of fitting due to the restriction in the fittings.
Rick

Scott_Mountain_View_CA

Bob Bona said:

Compare Uponor MultiCor and fittings for apples to apples PEX-AL-PEX. I would not use Hepex for "high" temp applications. Too much expansion.

Yup...which is why I abandoned my plan to use He and switched to Fosta. You may have been one of the fellas that warned me about movement noise in the walls.

Scott_Mountain_View_CA

Bob Bona said:

Compare Uponor MultiCor and fittings for apples to apples PEX-AL-PEX. I would not use Hepex for "high" temp applications. Too much expansion.

I looked at multipress fittings for MLC.....way too expensive for me, and I didn't see any poly alternatives.

hot_rod

I have used the Sioux Chief copper pex ells and plain copper crimp rings on every brand pex for probably 30 years now, including Viega and Uponor/ Wirsbo, no failures to date.

Here is an example of a 1/2 and 1" I'd consider these more of a copper ell for the pressure drop calc.

Granted the tube manufacturer want you to use their fitting, just saying.

I know Sioux Chief builds copper fittings, crimp rings and manifolds for many pex tube manufacturers.

Scott_Mountain_View_CA

Rod, I'm looking at using long turn copper ells with PEX Press adaptors rather than PEX Press ells

Bob Bona_4

We call them refrigeration elbows.

Scott, why again are you doing reverse return? It seems if you are on the ragged edge of design limitations re head loss, and are down to calc'ing every elbow (exhaustive), might it be time to step back and consider another distribution method such as remote manifolds?

hot_rod

Scott_Mountain_View_CA said:

Rod, I'm looking at using long turn copper ells with PEX Press adaptors rather than PEX Press ells

And sweating pex adapters into them? Probably less $$ and pressure drop in a long sweep pex ell.

Harvey Ramer

I am thinking that after you do correct head calculations you'll find that things aren't as bad as you think.

Scott_Mountain_View_CA

Bob Bona said:

We call them refrigeration elbows.



Scott, why again are you doing reverse return? It seems if you are on the ragged edge of design limitations re head loss, and are down to calc'ing every elbow (exhaustive), might it be time to step back and consider another distribution method such as remote manifolds?

Bob, I live in a 1,500 square Eichler home. Eichler homes are slab on grade with T&G roofing deck and no attic space. There is no realistic way to distribute supply within the heating envelope.

My 2 options:

1) Punch up from the boiler room and run piping over the flat T&G roof, punching branch piping down into the wall cavity at each of the 6 radiator locations (all 6 are on exterior walls). I didn't want to do this for several reasons including having to run all copper mains rather than PEX on the exposed roof.

2) Run piping in a soffit tight against the top of the exterior wall where it meets the horizontal roof decking eave. I chose this method.

My next decision was distribution method: series, direct return, or reverse return. I chose reverse return because of the balancing benefit. Had I chosen direct return I could have reduced total piping from 380' to 330' but would have sacrificed much of the balancing benefit.

Given my situation I see no other options available.

The majority of system head loss was due to 1) my particular design constraints requiring the use of numerous elbows, and 2) my original choice of fittings. PEX Press Poly fittings are flow hogs, particularly the tee's and elbows. I've eliminated all PEX Press Poly fittings from the design and replaced with copper and PEX brass adapters.

Because all of my branch input/outflow piping is contained within the exterior 2 x 4 wall cavity, I need to make a 90 degree turn from the angle valves at both ends of 6 radiators = 12 elbows. I had originally planned to use PEX Press bronze 1/2 NPT x 1/2 PEX ells but they too are flow hogs. I've eliminated all 12 from the design and replaced with copper 1/2 threaded male adapters, 1/2" copper refrigeration elbows, and PEX Press adapters.

I've also figured out how to eliminate ten 1/2 ells from the system altogether.

By this evening I'll have a revised head loss plan that hopefully will fit under either a Taco 007 or 008.

It's obvious that I'm learning a lot of this on the fly. I greatly appreciate the comments and suggestions from the forum contributors.

Scott_Mountain_View_CA

Harvey Ramer said:

I am thinking that after you do correct head calculations you'll find that things aren't as bad as you think.

Indeed. Thanks for that workup!