GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Hydronic ceiling installation (DIY)

Brad_S | Posted in Energy Efficiency and Durability on

Climate Zone: Marine 4 (Seattle, WA)
Ceiling size: 900 sq/ft
Ceiling Drywall: 5/8″
Joists: 24″ o.c. 2×4
Attic: unconditioned

Current plan is to remove all old attic insulation and wood shingle roofing scraps thrown in during a previous roofing job. Then replace old knob and tube wiring with something more modern. Finally, air seal and add R-39 blown cellulose.

As long as I’m going that far I started wondering if I might also add PEX for hydronic heating. Seems like it would be easy to install when everything is stripped out. However, I’m trying to figure out how to do it correctly.

I was thinking 1) barrier PEX in direct contact with drywall, 2) covered w/ R-39 cellulose. However, should I add radiant barrier over the PEX to direct radiation down? Or should I sheath the ceiling joists with rigid foam with cellulose above that and leave empty space around the PEX (which might also help with cellulose weight due to 24″ o.c. joists)?

Or, are hydronic ceilings a bad idea and I shouldn’t do it at all?

Thank you.

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. Brad_S | | #1

    Also, is running cool water through these lines in the summer generally an effective cooling system or a bust?

  2. user-2310254 | | #2

    If you are doing a significant retrofit, you might be able to meet your heat (and cooling) needs with a conventional HVAC or split mini system.

    R-49 is code for Zone 4. Do you have enough space at the exterior edges for that depth of insulation?

  3. Brad_S | | #3

    Steve, there's not enough room at the edges to reach R-49. It will take 1.5-2 feet before there's enough room.

    I currently have a 4 year old natural gas furnace but my wife doesn't like forced air.

  4. user-2310254 | | #4


    Have you considered installing rigid foam on top of the roof? You could install all the insulation (assuming a simple roofline) or do a combination of foam (at least 20%) and air peamable insulation. See this article for details:

  5. GBA Editor
    Martin Holladay | | #5

    Here is a link to a design guide for hydronic ceilings: Radiant Ceiling System Design.

    Note that I am not endorsing this approach, nor am I endorsing this design guide.

    Hydronic ceilings can work, but these systems are often more complicated than necessary. A simpler approach -- emphasizing air sealing, high R-values, and simpler HVAC equipment -- is usually preferred.

    I don't recommend that you try to circulate chilled water through these pipes for cooling. This approach often results in condensation problems. You don't want a damp drywall ceiling.

  6. Brad_S | | #6

    Thanks, Steve, I'll check that out. What's the advantage of rigid foam on the roof vs blown cellulose with weak insulation along the edges of the roof?

    Martin, thanks for the radiant ceiling link. From the feedback it doesn't sound like a great idea to install a hydronic heating in the ceiling.

  7. user-2310254 | | #7


    The exterior insulation would allow you to achieve the full R-49 over the entire roof line. But probably the biggest benefit is using exterior foam would enable you to seal the attic and bring it inside the home's envelope. In one of his articles, Martin also notes the following:

    - Rigid foam above the roof sheathing interrupts thermal bridging through the rafters.
    - Rigid foam above the roof sheathing keeps the sheathing warmer and drier than it would be if all the insulation were on the interior side of the roof sheathing.

    Read more:

    Of course, you have to weigh costs and benefits of all the advice you receive on this site.

  8. Brad_S | | #8

    Thanks, Steve

  9. user-2310254 | | #9


    JLC has a video on radiant cooling you might be interested in viewing. The video also points to additional resources:

  10. Brad_S | | #10

    You're messin' with my mind, Steve. Now I want to add hydronic again. "High temperature cooling, low temperature heating; 60F water to cool, 90F water to heat."

    Hottest day in Seattle this year was on August 4th, 2017. 93F max, 66F min. Dew point 59F. If I understand correctly that means 60F water flowing through my ceiling would not condense moisture from the air. Although, I might feel comfortable with water warmer than 60F.

    Hottest day on record in Seattle was on July 29, 2009 (I remember that day, I was dying.) Max 105F, min 73F. However, dew point was still only 60F.

    Further, the new smart thermostats might help in this regard. If they can grab dew point from WiFi and make it available to the cooling system, the hydronic would always stay above dew point and never condense water vapor. Then you wouldn't need the dehumidifier -- though, you might not need a dehumidifier anyway in Seattle summers because the air is dry. Austin dew point on Aug 4 2017 was 69F. If they were running 60F water through that ceiling in the YouTube video it would definitely condense.

    Interesting to think about.

  11. Jon_R | | #11

    In most climates, radiant cooling will require a dehumidifier to maintain the recommended interior %RH. At most loads (<= 10F delta T), this also suffices to prevent condensation.

    But running a dehumidifier can negate the energy efficiency benefits of radiant. Plus the issues involved with distributing dehumidified air. And producing and distributing cool water.

  12. user-2310254 | | #12


    Note that Risinger calls this install a "work in progress." Even if they can make it work without dripping or driving up energy costs, Austin is probably a more appropriate climate than the Pacific Northwest. Installing this type of system definitely puts you in guinea pig territory.

  13. Brad_S | | #13

    I think the most cost efficient solution in this climate is to open windows in the front and blow a fan out the back during the cool nights. Or a whole house fan. But the whole house fan would need good insulation for the winter. Or even an indoor A/C unit because the really bad days are pretty rare.

    I'll bet the hydronic would work here w/o a dehumidifier. The problem would be complexity as Martin pointed out. Your thermostat would need to measure dew point then keep water temp above the danger zone. Might be more of a rich person's solution.

  14. Expert Member
    Dana Dorsett | | #14

    Brad S has it right- Seattle's low summertime dew points translate into NEGATIVE latent load about 99% of the time, yielding low over night temperatures suitable for a nighttime ventilation strategy for cooling.

    But those same low dew points also meant that mechanical dehumidification should not be necessary to be able to use radiant cooling without condensation on ceiling.

    Unless the chilled water distribution plumbing is tightly insulated everywhere the controls have to be able to keep the distribution plumbing at or above the indoor dew points to avoid condensation on the plumbing. That's part & parcel of any chilled water cooling system design. Chilled water cooling with absorptive chillers is common in commercial construction, but the design skills and controls are out there.

    Doing it on the cheap you'd be pretty safe in Seattle if you simply set the chilled water temp to a fixed 65F. The number of hours the outdoor dew points exceed 65F per decade in Seattle is but a handful, and artifact of being downwind of an ocean with surface temperatures in the low 50s. Even at the peak humidity weeks of July/August dew points exceed 60F less than 5% of the time. See:

    For those few hours per decade there might be a slight fogging of the ceiling paint, but even during the peak cooling weeks the average temperature at the ceiling for the gypsum to dry overnight.

    With stamped sheet metal heat spreaders coupling the PEX thermally to the gypsum board even 65F water would probably be enough to chill the room to the low 70sF at Seattle's 82F 1% outside design temperature. Without a Manual-J and the full system design it's hard to specify the water temp requirements with certainty, but unlike most of the right half of the US map, radiant cooling should be pretty effective, with low moisture accumulation risk here.

  15. Brad_S | | #15

    Well, I'm interested in this project again, Dana.

    65F was the magic number I was thinking in my head too without risking 60F.

    Do people install hydroponic chillers (aquariums) on systems like this?

  16. Expert Member
    Dana Dorsett | | #16

    Before picking a chiller, you need to first run both the heating and cooling load numbers, and determine the water temp requirements that would cover the design cooling (and heating) load(s). The "right" solution for both heating and cooling might be a reversible chiller (or not). Without heat transfer plates thermally coupling the PEX to the ceiling gypsum the water temp requirement would be insane.

    "Design by web-forum" isn't usually going to yield optimal (and often not even functional) results. The best you can get out of this is an outline of what direction to steer, but this is something that in the end really requires doing at least some of the math.

  17. Brad_S | | #17

    OK, thank you, Dana

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |