GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X 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 radiant ceiling for heat and cooling

NoCoFarmer | Posted in Mechanicals on

Hi all,

I recently inherited (as in, moved across the country to take possession of) a 1800 sqft ranch with a back wing, in No Colorado built in 1975. Currently heated with a 80% cast iron boiler and baseboards with an additional separate DHW heater. The insulation is not great and I will be upgrading wherever I can. The walls are 2×4 with fiberglass batt. The main living wing of the house is a vaulted ceiling (peak at 12 ft) with beams at 48″ oc. (that is 2×6 with fiberglass batt on top of that). The attic’d portion of the house will eventually get r39+ of blown in cellulose. The main dilemma is that we love radiant heat (its quiet, no blowing vents, radiant is comfortable), but hate the baseboards (they take up space and in these walls I could feel the heat coming through to the outside on a cold night. )

I spent some time looking for radiant replacements that would make up for the baseboards but couldnt find enough wallspace or btus to make up for them. We considered underfloor staple up but are set on solid plank wood flooring and probably area rugs in places so I think the transmission would be poor. Then I discovered this article: and thought I found the solution to my dilemma. I have done a room by room heatloss calculation estimating some final insulation numbers (35kbtu at -10F ) and have been lurking around various threads to understand the parameters such a low mass radiant system would need. I also have a cramped space for mechanicals (actually trying to recover some from the behemoth we currently have. ) So I think a HE mod/con boiler with a lower min burn (like this: ) and a sidearm (like this: for thermal mass might be the right choice for my situation. Then I have some freedom to zone it in a sensible way and have plenty of DHW for the 2 adults and 3 growing kids (read bathtub fills).

The other aspect here is the climate in NoCo (5B), where we have large and rapid temperature swings. One needs an agile system able to adapt quickly. The house does not have any AC or duct work of any kind except for a whole house fan and the number of days one needs cooling is limited to a few late summer afternoons. The proponents of radiant ceiling (for heat and cooling) have intriguing arguments and demonstrations. The number of negative comments about the principle and the performance are very few and the proponents are passionate about it. But the number of outfits who can even speak to them seem fewer. So I wanted to get some live discussion from GBA about whether I’m on the right track.


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. Expert Member
    Akos | | #1

    Floor heat is probably the cheapest option if you can DIY. Instead of simple staple up, I like to use Ultrafin, much quicker install and less piping and keeps the pipes away from floor staples. Ceiling should work as well but you have to put up a new celing which is way more work than floor heat.

    You are close to 20btu/sqft which is doable with floor heat. With hardwood flooring you are limited to 85F surface temperature, which is still 30btu/sqft. As long as the carpeted are is not a big portion of the house floor, it should not be an issue with floor heat.

    With a correct sized boiler modulating boiler, you don't need a buffer tank. Just the water in the pipes of a typical floor heat setup is more than sufficient to keep a 12k min fire unit from short cycling.

  2. walta100 | | #2

    In my mind you need to do your improvements in order if not you double your work and buy the wrong equipment.

    Start with window and doors if they are changing only after air sealing should you insulate then finally should you update the HVAC.

    If you improve buildings envelope you will need fewer BTU and less base board. You may want to have a look at some of the modern hydronic baseboards they can be quite thin.

    I do not think you think you find hydronic ceiling cooling to be a low cost DIY friendly project. On the plus side you do happen to find yourself in one of the few places where it is dry enough to actually work well.


  3. NoCoFarmer | | #3

    Thanks for your comments. Actually I forgot to mention that my cooling btus are free due to an open irrigation well I have that has quite chilly water in it. I realized I can easily run water from and back through a plate exchanger and manually switch the system over to it with diverting valves on the few afternoons I need it. My heatloss calcs are already making assumptions about envelope improvements. The current envelope is poorly sealed (crawl space sills, door jams let light through, attic currently has 3 inch FG batts, etc. etc.)

    I should mention I do plan to DIY atleast the radiants. Quotes I have gotten are outrageous for the labor (reading between the lines.) And I have a 9 and 11 yearolds who are champin at the bit to help daddy with the power tools. The attic'd portion will go on top of the existing ceiling with the basic style plates glued down. Ceiling then gets reinforced with sisters (the joists are 2x4 24 oc.) Then I have about 28x24 of the vaulted to do with a panel jack in the full seigenthaler prescription. That is actually the most critical to do as ceiling since it is the south side room and gets hot in the summer afternoon. The south wall atleast is getting better insulation as the sheetrock is already off.

    My remaining items I want to solidfy are the right choices on pumps and exchangers. I was thinking of a deltaT pump like this:

  4. RedDenver | | #4

    I live in the Denver area and have been thinking about radiant ceilings for both heating and cooling. Please let us know what you decide.

    I've been looking at the Tekmar 406 ( with a heatpump as primary using a single tank for heat and cooling with electric for backup heat as shown on page 12 of the installation guide (minus the DHW).

    1. hydronicnerd | | #7

      Buy the Tekmar 406! I have it installed for radiant heating AND cooling coupled with an oversized 3.5Ton geo heat pump. Works like a charm. Even wife happy. Issue-buy and install the BIGGEST buffer tank you can fit. It takes the load off the HP and prevents short cycling. Also, do your Man J twice, insulate and you will be a happy camper. Max use of HP is 45 minutes/hr which shows excess capacity in HP.

      1. Expert Member
        NICK KEENAN | | #12

        This sounds like my dream system. How are you managing humidity?

        1. hydronicnerd | | #13

          Unico air handler to move & dry air. Run it on low fan speed with Tekmar 557 to monitor/control both temp AND humidity. Next iteration will include DOAS HRV. Tekmar is pricey but it's plug & play and it WORKS (I do not work for Tekmar). Bear in mind, the Tekmar 406 will only deliver ~68F chilled water to radiant floors when the room temp is > 70 degrees (no puddles allowed on the floor).

          1. RedDenver | | #33

            Thanks for the feedback.

            Looking at the 406 specs, it should be able to deliver chilled water below 68F when there's a humidity sensor installed.

  5. Expert Member
    NICK KEENAN | | #5

    I've also been wondering about ceiling hydronic, and I keep thinking, "what's the catch?"

    Because it sounds like it solves all the problems with floor hydronic. You can run it hotter because you're not in contact, and ceilings are just a lot less complicated than floors. It seems like it beats dedicated radiators in every aesthetic sense, except maybe as a place to dry your mittens. Reading the Seigenthaler article it doesn't sound like a lot of work -- especially if you're going to drywall the ceiling anyway.

    In terms of cost, BaseRay, which I consider the premium baseboard hydronic, runs about $50/linear foot and puts out 550 BTU/lf. So 9 cents/BTU.

    48 feet of PEX @20 cents/foot=$9.60
    Sheet of 4x8x3/4 Poly-ISO= $18
    48 feet of aluminum transfer plates @60 cents/foot= $28.20

    Total: $55.80, or $1.75/square foot. At 30 BTU/SF that's less than 6 cents per BTU, or about two thirds the cost of BaseRay and about $2100 for the 35K BTU specified. This is apples to oranges because the BaseRay is rated at 180F, the hotter your water the lower the cost per BTU. Of course there are other costs but those are going to be there no matter what your system.

    If you have to redo the drywall, I'm paying about $2 a square foot for drywall hanging and finishing, all-in. But you get a new ceiling in the bargain.

    1. Expert Member
      Akos | | #10

      The catch is that you get much less heat out of ceiling panels than from floors.

      You CAN get more heat, but you need very hot water. For a modcon to supply that, you'll loose a fair bit of efficiency. A heat pump will struggle in colder climates with water that hot.

      If you keep the temperatures reasonable, say 100F supply, 90F mean surface temperature, you are getting 15BTU/sqft. You can get 15btu/sqf out of floor running at 77F temperature, or you can reduce the area the heated floor (and your cost by 1/2) by bumping up the floor heat to 85F. This temperature is still doable with heat pumps and keeps a modcon well into condensing range.

      1. Expert Member
        NICK KEENAN | | #11

        I'm trying to think why that would be. When you look at formulas for heat transfer, they care about two things: the ambient temperature and the temperature of the heated surface. For the same water temperature I would think the ceiling would be hotter than the floor because drywall has less insulating capacity than typical flooring (not that either have a whole lot).

        1. Jon_R | | #17

          The right formulas also account for airflow. Which is different with floors/walls/ceilings. Don't be fooled by thinking this is 100% radiant heat - there is a lot of convective transfer.

          Do consider radiant walls (up to 120F) for greater heat transfer than a ceiling or floor, often without direct losses to the exterior.

          1. Yupster | | #21

            The heat transfer coefficient used for estimating output on a ceiling is 1.6, as opposed to 2.0 on a floor. A wall is 1.8. It can be used in the formula Output = Coefficient x Area x (Panel temperature - air temperature). But since you can run a ceiling radiant panel at way higher temperatures (120°F compared to 84°), you can get more heat out of them than a floor. They also aren't affected by sofas or rugs or cabinets, giving you more area to use in general.

          2. Expert Member
            Akos | | #22

            From the Uponor radian design manual, the transfer coefficients are:

            • Radiant floor thermal transfer
            coefficient = 2.0 BTU/h/ft2/°F
            • Radiant wall thermal transfer
            coefficient = 1.4 BTU/h/ft2/°F
            • Radiant ceiling thermal transfer
            coefficient = 1.1 BTU/h/ft2/°F

            So a ceiling heat system puts out about 1/2 the heat of a floor heat using the same temperature water.

            This is for heat only.

          3. Expert Member
            NICK KEENAN | | #24

            The info about HTC is very helpful. Thanks Jon, Yupster and Akos. Now I just have to figure out which of the three numbers is right!

            Having a higher HTC on the ceiling for cooling may actually work out quite well, I can use warmer water which means less concern about condensation.

          4. Jon_R | | #26

            When it comes to condensation, note that humidity isn't a constant with location. Ie, a ceiling is likely to be more humid* than an interior wall.

            * - due to stack effect inducted infiltration and that humid air rises

        2. tommay | | #36

          The heat transfer coefficient is one of those tricky things. It is not absolute. You have to extrapolate it's value based on air temperatures and air flow and even then it is not absolute. So one cannot put a set value on it since it changes constantly depending on the changing conditions.

    2. Expert Member
      Dana Dorsett | | #15

      >"In terms of cost, BaseRay, which I consider the premium baseboard hydronic, runs about $50/linear foot and puts out 550 BTU/lf. So 9 cents/BTU."

      That's more like comparing apples to figs than to oranges.

      It's only 550 BTU/lf at an AWT of 180F (say, 190F out of the boiler, 170F return). At condensing temps it's about 200 BTU/ft, or about 25 cents per BTU-out capacity.

      In my house I have a couple of micro-zones using BaseRay @ 120F AWT, and they work great. But it was only made cheap by using salvaged goods. Used/reclaimed radiation recommended where possible, since it's is a lot greener than virgin stock goods, since no new mining or metals refining going on, only transportation and re-painting. They clean up easily, and at low temp the surfaces stay well within the operating temperatures of standard interior latex paint. There are miles of 9A and 6A BaseRay in scrap yards at $15-20 per foot (sometimes cheaper) all over New England, but I'm not sure how difficult it would be to find used BaseRay in Colorado. I've never paid more than $12/foot for used BaseRay on Craigslist.

      If using salvaged cast iron baseboard try to find fully assembled sections in the right sizes, with the cast-in foot on each end. While the new push nipples needed for disassembly/reassembly are available, it's a lot more work, and requires buying/renting the special tool for pressing sections together. (It's a PITA.) The sheet metal end caps / valve covers can be purchased online.

      1. Expert Member
        NICK KEENAN | | #25

        I think we're actually agreeing, lowering the water temperature makes PEX coils look even better.

  6. Expert Member
    NICK KEENAN | | #6

    In the summer, do your cold water pipes sweat? The big issue with hydronic cooling is humidity control. If the dew point of your air is higher than the incoming water temperature, you will get condensation which will lead to water damage and mold and all sorts of bad things. In a conventional forced air HVAC that condensation happens on the cooling coil in the air handler where it can be captured and drained.

    If you normally get sweating pipes you will have to dehumidify your air in order for hydronic cooling to be practical.

    1. hydronicnerd | | #8

      Duh. You need not only outdoor reset but humidity control. Tekmar 406 coupled with heatpump and you are good.

      1. Yupster | | #14

        You have to insulate all the pipes with insulation with a sealed vapour barrier. You also need to insulate valves, circulator volutes, etc. The high quality manufacturers like Caleffi make foam shells for insulating most of their valves. You also need dewpoint control.

        1. Expert Member
          NICK KEENAN | | #16

          If the pipes and controls are all inside the envelope and you have dewpoint control I don't see the need for insulating everything.

          Dewpoint control is the key.

          1. Yupster | | #20

            If you keep the room temperature at 75°F and 50% RH, standard design conditions here in Canada, the dewpoint is 55°F. Guess what the typical temperature is for chilled water in a radiant cooling system? 55-58°F
            And there is no way you can guarantee that the RH isn't ever going to go up, hit 55, 60%. Dewpoint controls are for radiant panel SURFACE temperatures, not the chilled fluid passing through them. You definitely need to insulate and vapour barrier all the components.

  7. BirchwoodBill | | #9

    I removed all my baseboard and installed Warmboard throughout the house. So basically all radiant heat from the floor using water between 80 to 100F. The heat comes from a simple boiler and 80 gallon indirect tank kept at 140f through a heat exchanger. There is no need for a buffer tank or fancy controls. I do like the Tekmar and it is best of breed. Will use on the next house. In retrospect I should have used a heat pump for both heating and cooling. Read Dana Dorset’s recommendation he help me with my load calculations.

  8. Jon_R | | #18

    No need for outdoor reset - unless you care about condensing boiler or heat pump efficiency.

  9. jameshowison | | #19

    When we were considering this the question of whether we could adequately control humidity was a big sticking point for us, we just didn't have experience with it and a fairly leaky old house combined with southern humidity had us worried. We need not have worried, we use an UltraAire 98H and humidity control has been wonderful. Sure it would be better/easier had we been really able to tighten things up more, but even with our leaky old windows and some big holes not fixed yet, the dehu keeps up at very reasonable costs.

    Probably not important in CO, but relevant to anyone considering radiant cooling.

  10. Yupster | | #23

    Re: Akos, Comment #22
    Interesting! I get my figures from Caleffi's Idronics technical magazine, as well as Siggy's Modern Hydronic Heating book. I'd be willing to bet there is a difference in the assumptions used to make those coefficients, since they rely on a number of standard operating conditions and construction. Interestingly, the Uponor design manual uses the same 1.6 number as I do later in the same manual, with no explanation. I'm inclined to use the idronics number of 1.6 since it is a more recent publication with a great technical explanation and uses the same construction I would use. You can find it here if you are interested:

    Regardless, reduced capacity at the same water temperature for sure.

    1. Expert Member
      Akos | | #27

      Lot of those coefficients depend on the temperature you are running at, so unless one wants to crawl through the equations, they will always only ballpark. That is a big spread though. I've seen 0.8 for ceiling radiant heat somewhere as well.

      In heating dominated climates, I still fail to see what ceiling panels buy you. About the only way it makes sense if you have high temp heat from a wood fired boiler.

      1. Expert Member
        NICK KEENAN | | #28

        >"In heating dominated climates, I still fail to see what ceiling panels buy you."

        When you heat the floor you have to run in a narrow range of temperatures, too warm and it gets uncomfortable. In the ceiling you can run in a much broader range, and you can make it hot if you need to.

        Ceilings are unaffected by the placement of furniture and rugs.

        If you spring a leak, most floor materials aren't easily repaired. Drywall is very easy to cut open and patch.

        The ceiling is better protected from physical damage than the floor. Nobody's walking on top of your pipes. Nobody's using an air gun to shoot flooring down onto your pipes.

        The big drawback of the ceiling is that the HTC is lower so the area has to be bigger. But if you're doing cooling, that allows you to use a warmer water temperature, which makes condensation less of an issue. Let's take the HTC on the ceiling at 1.9 for cooling and 1.1 for heating. If your heating and cooling BTU needs are the same you could run the cooling water at 60F (20F delta at 80F ambient) and your heating water at 107F (35F delta at 72 ambient) and do that. Those look like really reasonable numbers.

        1. Expert Member
          Akos | | #29

          Unless the heating is significantly undersized, the floor never gets too hot. The usual problem is that people put in too much floor heat and the floor never gets warm enough.

          Temperature limits are not an issue, I have hardwood over my heated floors and it never gets hot enough to cause issues.

          My wife constantly re-arranging our living space, never had an issue with furniture placement. It does feel warmer under a couch though.

          Leaks with pex happen at fittings. So no fittings in the field, no leaks.

          Nails are not an issue if you use the right setup (pipes not against subfloor) or put down click flooring.

          Most well insulated houses without silly window orientation/sizing have very little sensible cooling load compared to heat load. I doubt the cooling efficiency matters much.

          I've have done practically every flavor of heated floors and I can tell you the setup shown for the ceiling radiant is labour intensive. If you are going to DIY, it is doable. If you are paying for someone to put that up, that is a lot of money, never mind the fact that you need to install more of it.

          1. Yupster | | #30

            As a note, you don't need to install more with ceiling radiant heating panels, you could actually install less but at the expense of a higher water temperature. Radiant cooling makes great sense where the other option is installing a radiant heated floor AND a duct A/C system. But definitely a luxury solution...not unlike radiant floors.

        2. Jon_R | | #32

          > your heating water at 107F

          See the link in #17 for "limit ceiling temperatures to a nominal 95 deg F".

          1. Yupster | | #35

            You can go up to 120°F, where the gypsum board then becomes the surface temperature limitation. The 95°F or 81°F as seen elsewhere is based on radiant asymmetry, so it's really a limitation based on customer preference. If the customer likes that "warm fire" feeling of one side hot, the other side, er, not, then raising panel temperatures could be a great solution to minimize cost and panel area.

      2. Yupster | | #34

        Hey Akos, I found a great explanation on how HTC are calculated on Robert Bean's site. Here is the link. Might come in useful for's quite straightforward.

        1. Expert Member
          Akos | | #37

          Thanks for the link. Don't know why the radiant manufactures don't include a graph like that in their literature. Simple and easy to use.

        2. Expert Member
          NICK KEENAN | | #38

          What's interesting is that he settled on walls as the optimal surface. Isn't a basic principle that you want to put heat sources closest to areas of heat loss? It seems like that's going to be easier on the floor or ceiling than on a wall, the windows are right where you want to put your heating.

          Plus, that's the surface in the house that is most commonly drilled or nailed into. "To prevent panel damage the wall was blueprinted and photographed prior to boarding. All dimensions and details recorded in the Operation and Maintenance Manual for the next lucky homeowner. " I can only envision how that's going to work out.

  11. cr0ntab | | #31

    Subscribing to topic - I'm looking into radiant ceiling cooling myself.

  12. NoCoFarmer | | #39

    Nice thread folks, thanks for all the comments. Further questions:

    how necessary is freeze protection in a system like this? I am a bit leary of to much glycol messing with viscocity and specific heat and such. I started looking at a battery backup for the circ. opinions?

    If I go with a combi boiler like the navien, I then have DHW direct and only the space heating on the side arm? How much will this affect the short cycling issues? It does eliminate the heat exchanger and associated flow meter complexity. what other boiler/tank combinations could I consider given my desired footprint? (~14-16" from wall 8ft ceiling+attic)


    1. tommay | | #42

      No problem using glycol. Just be sure you purchase glycol that is designed for heating systems. Regular anti freeze, like what you use in your car, will work but the heating types have more detergents to keep you system clean. It can be diluted down to 50/50 but I usually start with near 100% just in case in the future you have to service the system and lose a little. If you can find Cryotek at your local plumbing supply, use that, it is non-toxic as well. Remember glycol is used in cars because it absorbs heat and can deal with higher temperatures better than just plain water. Water in some cases can be more detrimental depending on its quality and conditions.

  13. Expert Member
    Akos | | #40

    Why are you looking at glycol? Unless you are running an outdoor chiller setup, you don't want glycol. Glycol is a maintenance item, you have to keep on top of it as it can get acidic and corrode pretty much everything.

    With the combi boilers, DHW comes off a plate heat exchanger inside the unit, there is no side arm.

    Short cycling with these can be an issue as you need a lot of BTU for DHW, the min fire even with good turn down on many of them is pretty high. If you have a low load place with low mass baseboards, this can definitely cause issues. With any high mass setup with reasonable zoning, this should not be an issue.

    1. NoCoFarmer | | #41

      Akos, on the boiler choice, yes I'm putting up low mass ceiling, hence the buffer tank (I guess I'm using the term sidearm incorrectly).

      And for freeze protection, then a battery backup would be enough? there are a couple of tight places where the tubing has to be routed by lesser insulation I want to make sure it's protected in power outages.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |