Welcome! Here are the website rules, as well as some tips for using this forum.
Need to contact us? Visit https://heatinghelp.com/contact-us/.
Click here to Find a Contractor in your area.
If our community has helped you, please consider making a contribution to support this website. Thanks!

3/8" vs 1/2" to panel radiators

Options
Blend9684
Blend9684 Member Posts: 14

Currently in the process of designing/installing new panel radiators in my house. I added a buffer tank to the design so that I can do per-room zoning. The house was built as all-electric and has insulation in all interior walls. The fireplace would also throw off a per-floor thermostat.

Using a 30-degree supply/return delta T, zone GPM ranges from .29 to .83 GPM. This puts the https://tools.uponorpro.com/calculator/ velocity at .97 to 2.77 ft/s for 3/8" Uponor hePEX. I read recommendations of 2-4 ft/s for air elimination, plus ease of installation and decided on 3/8" over 1/2" despite fitting availability, ironically increased cost, and increased head.

Should I have stuck with 1/2" despite velocity being lower? The radiators can bleed air from on top unlike baseboard.

Comments

  • EdTheHeaterMan
    EdTheHeaterMan Member Posts: 11,574
    edited November 8

    I'm gonna say that 1/2" would be the smallest PEX I would use to feed panel radiators. If you properly configure your system with the specifications of this book, Pumping Away, You do not need to entrain the air bubbles in order to vent them from the system. You will be using Boyle's law to remove the air over time. You just need to vent enough air to move some heat thru each panel radiator. As long as water is moving thru the system, air will be vented.

    Edward Young Retired

    After you make that expensive repair and you still have the same problem, What will you check next?

  • hot_rod
    hot_rod Member Posts: 26,602

    it is more about the length of the loop when you go with 3/ 8 pex

    If you sized with the Uponor data you will be fine

    The 4 fps is a guide to avoid velocity noise, at excessive velocity the air doesn’t move along with the water

    Bob "hot rod" Rohr
    trainer for Caleffi NA
    Living the hydronic dream
  • Blend9684
    Blend9684 Member Posts: 14
  • I've always sized with a 20°Δ which gives me some buffer. Half inch will supply up to a 15,000 BTU radiator easily and out West here, those are few and far between.

    Running three-eights tubing borders on play vs. work and I see the attraction. No reason not to unless length and output are not exceeded.

    8.33 lbs./gal. x 60 min./hr. x 20°ΔT = 10,000 BTU's/hour

    Two btu per sq ft for degree difference for a slab
  • EBEBRATT-Ed
    EBEBRATT-Ed Member Posts: 19,598

    I pulled up that calculator and at .85 gpm and 125 ' your looking at 16' of head. Pretty high.

  • Blend9684
    Blend9684 Member Posts: 14
    edited November 8

    Full math of velocity and head for 3/8" and 1/2" - maybe a mix of both sizes is appropriate?

    Radiator Qty

    Radiator Per Each 160F

    Home Runs

    Lg Loop (ft)

    3/8"

    1/2"

    GPM Per

    GPM Total

    Velocity

    Head per 100

    Total Head

    Velocity

    Head per 100

    Total Head

    2

    6246

    0.62

    1.25

    50

    2.08

    6.58

    3.29

    1.13

    1.54

    0.77

    2

    5730

    0.57

    1.15

    36

    1.91

    5.66

    2.04

    1.04

    1.33

    0.48

    1

    9149

    0.91

    0.91

    63

    3.05

    12.85

    8.10

    1.66

    3.00

    1.89

    1

    6246

    0.62

    0.62

    100

    2.08

    6.58

    6.58

    1.13

    1.54

    1.54

    1

    5252

    0.53

    0.53

    34

    1.75

    4.86

    1.65

    0.95

    1.14

    0.39

    1

    4775

    0.48

    0.48

    68

    1.59

    4.12

    2.80

    0.86

    0.97

    0.66

    1

    4298

    0.43

    0.43

    118

    1.43

    3.43

    4.05

    0.78

    0.81

    0.96

  • Kaos
    Kaos Member Posts: 848

    There is nothing wrong with running the smaller pipe. When I was working overseas I saw a lot of heat with either 8mm and 10mm pipe. If you don't need the GPM, the smaller pipe is fine. The panel rads are often run at higher delta T which further reduces return water temp without reducing output all that much which increases the efficiency of a modcon.

    The cost of the fittings is annoying, but the install is much easier.

    @EBEBRATT-Ed The op is only looking for 0.4GPM which is around 4' of head.

  • Blend9684
    Blend9684 Member Posts: 14

    Even moving to 20°Δ from 30°Δ gives me the ~4 ft on longer runs you mention. I'm thinking that I could go to 1/2" for the few that push head to 6-8.

    Head for home run loops is the largest not additive, correct? I can't find where I saw that.

  • Kaos
    Kaos Member Posts: 848

    In home run the highest loop resistance sets what your pump needs to supply. Standard deltaP circ is about 10' of head, so anything less than that should be OK as long as not too many extra fittings. The 9000 btu loop does feel a bit close, I would run that in 1/2.

  • hot_rod
    hot_rod Member Posts: 26,602

    Small diameter soft copper or PAP is very common in Europe for panel rads. I've seen panel radiator retrofits in some rooms of the old castles with tube as small as 10 mm.

    Caleffi makes 10,12,14, 15 & 16mm copper compression fittings. PAP up to18mm

    The first step is identifying how many BTU you need to move through the tube.

    30- 40° delta is not unheard of in panel radiator systems, which help minimize tube size as Kaos mentioned.

    Bob "hot rod" Rohr
    trainer for Caleffi NA
    Living the hydronic dream
  • Big Ed_4
    Big Ed_4 Member Posts: 3,364

    I would prefer to run 3/8" A Pex over 1/2" for the neatness and the better movement of air bubbles . I would not use the cheaper 3/8" crimp type Pex , if they make it for some dumb reason . If you do use the 1/2" only , the inside diameter of the 1/2" fitting is close to a 3/8" opening . The lack of 3/8" manifolds means we have to add adapters .

    There was an error rendering this rich post.

  • Blend9684
    Blend9684 Member Posts: 14
    edited November 8

    Uponor hePEX with the expansion fittings was my plan. It's worked out well for plumbing.

    Seems I need to use (1-1/2"?) copper (or I suppose iron/brass) between boiler and buffer tank and buffer tank and boiler header to zone circulators.

  • hot_rod
    hot_rod Member Posts: 26,602

    Pipe is always sized by the amount of flow you intend to move through it.

    When doing headers for multiple boilers, loads, the pipe sizes for the total gpm at around 2 fps.

    If you do a 2 or 3 pipe buffer, the connection at the tank needs to be sized up, as it is the hydraulic separation function. In this example at least 1-1/4" for the 10 gpm requirement.

    Screenshot 2025-11-08 at 10.30.57 AM.png

    The EngineeringToolbox has good flow sizers. for

    Screenshot 2025-11-08 at 10.36.56 AM.png

    e

    Bob "hot rod" Rohr
    trainer for Caleffi NA
    Living the hydronic dream
    Alan (California Radiant) Forbes
  • EBEBRATT-Ed
    EBEBRATT-Ed Member Posts: 19,598

    @Kaos the information is spotty. The original post said .29-.83 gpm with no length given. Second post said 125' @.4gpm.

    So what if the second longest length is 110' at .83.? I was looking at worst case. You need length and GPM to size the pipe and he said he was using 3/8 for everything