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 Radiant Heat Design

paulneumann | Posted in General Questions on

I know this question has been debated before and that there are a lot of mixed opinions about the SANCO2 HPWH, but want to approach the topic again. I understand it requires substantial lift to maintain the COP and that it takes some planning or a “science project” to make it work, but we are a group of engineering students designing a 634 SF home. It will eventually be lived in but the owners are pushing for a radiant system even if it doesn’t have the ROI.

That being said, why can’t we just run shorter loops with a larger spacing on center and use a higher temperature. Currently with the envelope we’ve designed and modeled we have a peak heat loss of about 8000 btuh. So about 12.6 Btu/SF. Yes small but when the DHW load matches if not surpasses the heating load it seems like sense to combine the systems.

I am new to hydronic radiant design (besides the basic principles and equations) but have been reading Modern Hydronic Heating along with a few other resources. Any other recommendations welcome.

I would love some advice or to have someone help me in the right direction. For instance the SANCO2 can supply 18.5 GPH with a lift of 95°F. So a ∆T=(145-50). Lets say we then supply the floor with 100 °F water and want to maintain the space at 75°F then the flow rate would be .64 GPM. Since technically the SANCO@ can only supply .31 GPM would we want to match the flow rate to the water heater or alter length of the loops/runs if that makes sense. Because you can always tackle it from there ways right? Loop spacing/ runs, supply temp, and flow rate?

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
    NICK KEENAN | | #1

    The limiting factor in floor water temperature is peak temperature, an under-foot temperature above about 85F or so starts feeling uncomfortable and there have been reports of circulation problems in people's feet at even lower temperatures. In order to get your heat output up you need the average temperature to be high, and if your peak temperature is capped that means your average temperature has to be close to your peak temperature. Which is a long way of saying that your tubing needs to be close together.

    An average heat output of 12.6 BTU/sf is going to be pretty high. You'll probably find that due to things like bathtubs and kitchen counters only about half the floor area is actually usable for heating surface. Generally the smaller the house the lower the ratio is. So in the places where you have tubing you're going to be needing to get more like 25 BTU/sf, which is going to put you up against the comfort limit.

    1. Deleted | | #3


    2. paulneumann | | #38

      I understand what you mean now thank you. Had not realized that. Definitely starts to shift radiant system design because I am understanding now why we would start to have high temperatures.

  2. Expert Member
    NICK KEENAN | | #2

    " For instance the SANCO2 can supply 18.5 GPH with a lift of 95°F. So a ∆T=(145-50). Lets say we then supply the floor with 100 °F water and want to maintain the space at 75°F then the flow rate would be .64 GPM. Since technically the SANCO@ can only supply .31 GPM would we want to match the flow rate to the water heater or alter length of the loops/runs if that makes sense. Because you can always tackle it from there ways right? Loop spacing/ runs, supply temp, and flow rate?"

    What is your outside design temperature? That should be in your heating load calculations and you have to use that in your calculations. I can't believe it's 50F. Let's say it's 20F. So 95F of lift gives 115F as your water temperature on the design day. Can the heat pump provide the needed BTU's at that temperature? If not, it's game over. At 8000 BTU/hr and 0.31 GPM the heat pump needs to see a temperature drop of 52F. With water going out at 115F that means water coming in at 63F. That's not going to work for space heating, although it sounds about right for DHW. Are you sure that 0.31 GPM rating is for space heating?

  3. Expert Member
    NICK KEENAN | | #4

    Re-reading what your wrote, 18.5 GPH @ 95F lift sounds like a water heater spec, with the lift on the incoming water temperature. That works out to 14.5K BTU/hr which should be plenty. What you need to find out is what kind of lift you can get over outside temperature in space heating mode.

  4. paulneumann | | #5

    Sorry think I should have drinken another cup of coffee before posting that the first time because it didn't make too much sense.

    It's not the Manual J we did—but see the attachment for a quick building envelope load calc along with the design temperatures. All together were about a little over 8000 BTUH for peak heating but I have been waiting to Re-Do the manual J until our framing/envelope system is fully decided.

    So Outdoor Design Temp= 0°F. T

    Yes the 18.5 GPH is the water heater spec (rated at 15400 btuh). That being said I get a little mixed up with all the different water heater rating factors. UEF vs EF vs FHR vs UHFR. I had been trying to determine kwh/gal to get an understanding of the radiant system because I am trying to determine the peak load that is being consumed by the hydronic radiant system. Like worst case scenario ( coldest day of the year, no sun is out, lots of showers/hot water being used) what is the peak load.

    So for instance the way I look at it is that if we can use the FHR or that 18.6 GPH from the HPWH and simply fluctuate the temperature and keep the flow rate low then the electrical load will be significantly lower. So for a COP of lets say 4 with a rating of 15,400......=15400/4 = 3850 btuh and 3850/3412 = 1.128kW to heat the 18.6 gallons of water. When we factor in the pump power then that would give us the total electrical load correct?

    Where I get mixed up is the heat transfer between the slab and the room. I understand the HX for a cooling/heating coil and the GPM/CFM needed for that using the "universal HX equations" but lets say we wanted to sustain 8000btu/h for one hour. Can I really assume that it is truly just based off that equation even for a radiant system? I mean I guess there will be some heat loss depending on slab insulation but I still just thought I couldn't make that assumption.

    1. Expert Member
      NICK KEENAN | | #13

      In hydronic systems it's really common to combine heat and hot water. What is often done is "priority zoning," where if there is demand for hot water the heating coils are shut off. The thinking is that the need for hot water is sporadic and intense and it won't hurt to turn the heat off for a few minutes.

      Usually priority zoning is done in tankless systems. Now, the heat pump you're looking at doesn't have enough capacity to run even a single shower, a modest shower at 1 gpm at 50F rise is about 25K BTU/hr. So what you would have to do is have a tank. Then the question of how many BTU'/hr the hot water tank consumes is kind of up to you -- you could have a tank that holds 24 hours worth of hot water and size the zone to heat it evenly over 24 hours. You could just figure out how much capacity all of your other heating needs will take, then devote the rest to hot water, and from that figure out how big a tank you use.

      1. paulneumann | | #22

        That helps a lot to think about it that way thank you! Hadn't really considered how high the flow rate for a shower would be but helps to think about it that way especially for sizing the tank. Is it a good assumption to take the gallons for a load of laundry as an assumption for usage ( I know it would very based on hot or cold cycle) and same with the dishwasher just based off 120°F and then would be mixed?

        Still learning the pluming aspect of it all but trying to learn quick!

        1. Expert Member
          NICK KEENAN | | #23

          The way tank water heaters are rated is in first-hour capacity. That's how much hot water it can deliver in one hour, starting with a full tank of heated water. It's the capacity of the tank plus the heating ability of the heater.

          Let's say you have a 40 gallon tank and devote 6,000 BTU/hr to heating water. If your water needs to be heated 50F that's 15 gallons per hour. So you have a first-hour capacity of 40+15=55 gallons.

          Then you have to figure out what your peak hour of hot water usage is going to be. Usually this is just assuming all of the occupants take showers in the same hour, and you figure on 20 gallons per occupant for showers. So a 40-gallon tank is probably right for two occupants.

          Note that if you use 55 gallons of hot water your heat pump is only providing 15 gallons per hour so it will take almost four hours to refill the tank. So you don't want to be doing priority zoning because that's a long time for the heat to be off.

  5. paul_wiedefeld | | #6

    It’s not just the radiant floor ROI, it’s the low floor temperatures that make it a poor match for efficient buildings.

    As to the delta T question, I suggest reading the Caleffi Idronics series. Mixing temperatures is what you’ll want to do here. You can use a small delta T for the floor with a lower temperature supply and a large delta T on the heat pump side.

    For what this will end up costing, I think exploring using an electric boiler/tank and skipping the Sanden in favor of better insulation and/or solar PV will be compelling.

    1. Expert Member
      NICK KEENAN | | #7

      If you're going with electric resistance heat I don't see any point in a boiler. Just put the heating elements in the room.

      1. paul_wiedefeld | | #8

        Definitely, depends on how badly they want a radiant floor.

        1. Expert Member
          NICK KEENAN | | #15

          Even electric resistance heat embedded in the floor is cheaper and simpler than a boiler and tubing.

    2. paulneumann | | #9

      Hey Paul thanks so much for the comments and advice and ill check out Caleffi,

      Im an Architectural Engineering Student (about to take the FE) and focusing on the mechanical side of buildings but still learning. Could you elaborate more on why the low floor temp makes it a bad match because clearly I am just missing something? Just in terms of like the amount of time it takes to transfer the heat/ fluctuate temperature or? In my mind even with a lower floor temp you still achieve a better MRT than any sort of forced air/ perimeter heating and thus more efficient because you can achieve a better comfort level with less heat.

      So maybe my thinking is wrong but the home is a small "carriage house" so 634 SF above a small 2 car "somewhat" uninsulated garage. Part of the thinking was that we could lose less heat to the garage with a radiant system as opposed to forced air/ convectors like baseboard resistance heaters. We even explored hydronic baseboard heaters as well but quickly turned away from that due to the higher temperature needed (unless I am mislead).

      So what would be an ideal floor temperature to make it efficient? I guess that's where I am getting confused or mixed up. Because I mean couldn't we just alter the flow rate/temp of the water like you mentioned with mixing to achieve a greater ∆T or just calculate it based off the volume of the slab and the density?

      Also the main reason for the Sanden is using a refrigerant with a GWP of 1. Yes I know that with smart practices, proper installation, and responsible recycling of refrigerants it can be mitigated but still like both the higher lift and lower operating temp CO2 provides as a refrigerant along with the lower GWP.

      Were looking at an off grid storage option especially for essential loading so trying to stick away from an electric boiler/ tankless even with PV due to the higher peak pull.

      1. paul_wiedefeld | | #12

        The floor temp is more of an expectation thing. The floors at 12btu/sqft peak will be about 6 degrees over room temp, 1% of the year. The eventual owners will have to decide if it’s worth paying for that vs. more underfloor insulation.

        As to the efficient temperature for the floor, lower is better. The constraint is that delta T across floor loops is usually kept at 10 degrees though, so the higher delta T for the Sanden will have to be a different loop, which isn’t a big deal. Is the Sanden rated down to 0 degrees for space heating?

        Off grid entirely? Resistance space heating would only be 2.34kw peak and DHW can be rationed but the economics drastically change if you’re off grid using a battery, possibly in favor of the Sanden but probably more in favor of better insulation.

        1. paulneumann | | #17

          The intent right now is to be 100% off grid capable (if we can make it possible and logical). Because of that I am not too worried about the daytime because one through passive strategies along with an ERV and high internal gains but at night here in Colorado the Diurinal Temp swing is significant and trying to limit the Peak load even at 2.32kW at night would be ideal. Less worried about the storage amount needed and more about limiting the peak electrical load if that makes sense. So that's one of the biggest pushes towards a Combi-System is with an oversized hot water heater we could use it as a small "thermal" bank at night and only be supplying the pumping power. Yes it will need to regenerate but if we kept out flow rate of the radiant system to match the flow rate of the FHR of the SANCO2 the 18.5 GPH or .31 GPM it could make some sense right?

          Not sure if I did the best job explaining but that was the thought.

          In terms of insulation I wish we could just forget the mechanical systems and add more. Unfortunately due to setbacks were pretty constrained by the size of the wall and can't make the 634 SF smaller as its a very small home to begin with. Additionally low carbon materials are hard to find (at least low cost) with a high R/inch so we're trying to find a bit of a sweet spot at the moment. We are looking at using hempcrete actually if we can get the cost right especially for it's fireproofing properties.

          I just checked the SANCO2 specs (see below). It is not rated down to 0°F so thank you for bringing that up. I mean in a sense it is because the hot water system can produce 15,400 btu/hr for DHW down to -25°F but for the maximum 8000 BTU/Hr of space heating it's not. Furthermore the intention was to place the split system within the garage and since the door is south facing using a windowed door to benefit from solar gain on the concrete foundation because yes we will be cooling down the garage if its kept in there. Ideally though the garage will never below 27°F but maybe thats not a good assumption to make and show do some load calcs.

          The thought was to maybe have a summer bypass damper to benefit from the split system for cooling but that maybe adding too much complexity.

          Right now our peak load is about 8000BTU (the maximum sanco2 is rated for space heating) but ideally we will get it down a little is the goal to help use this system. Thats where I get a little mixed up is that an 83 gallon tank seems more than enough to supply lets say 1 hour of peak heating along with 40 gallons of DHW usage but I could be doing some calculations wrong as well.

          1. Expert Member
            NICK KEENAN | | #19

            100% off grid is a really heavy lift. It would be good engineering homework to look into the practicalities of it. Basically it's really expensive to store energy, either in batteries or as heat in some sort of container. The cost of capturing, storing and accessing energy is many multiples of the cost of that energy from other sources. Also the amount you have to store to be 100% off grid -- prepared for any contingency -- is enormous and most of that capacity never gets used.

            For example, let's say you look at weather records and decide you're going to design for the worst-case day in the past 100 years. You'd be building a system that has excess capacity 36,499 of 36,500 days.

      2. Expert Member
        NICK KEENAN | | #16

        There's no reason to believe that any particular heating system is going to have higher losses through the floor. If anything, a radiant floor might have higher losses because the temperature gradient across the insulation is higher since the heat source is in the floor.

        There's no reason baseboard radiators or convectors won't work with water in the 105F range, you just have to make them bigger than you would at 180F. But it's a lot cheaper and simpler than in-floor.

        1. paulneumann | | #20

          The goal would be to be somewhat off grid capable but still grid connected for those really rare days. So yes still being able to be grid connected but with most of the power generated and stored for daily use. Yes just going net zero and properly sizing a pv system to meet the loads is the best pound for pound. But you're still pulling half your power from fossil fuels each night. Our goal is to be "off-grid capable" so that we aren't pulling or sending much power through the grid.

          For instance in California with the amount of solar and the new proposals for Net Metering caps. We are less concerned with the dollar "cost" of the energy but the carbon "cost". That being said I am not a huge proponent of Li-Ion battery storage for the life-cycle embodied cost of them.

          1. Expert Member
            NICK KEENAN | | #29

            I'm going to make a blanket statement and say any kind of on-site storage of energy is going to be impractical. On a macro level that's not how we're going to decarbonize our energy delivery infrastructure. In fact, on these very pages there is spirited debate as to whether rooftop solar does more good than harm, and even a few skeptics of community-level solar.

        2. paulneumann | | #31

          Honestly I appreciate the opposing point of view. I am passionate and curious but also a young engineer and often overlook something. I am new to GBA but wish I had learned about it earlier because I have already shifted some view on some things.

          I will check out some of the threads. I agree with you point however I think in some ways if solar is already there (or enforced by code like here in boulder) then storing power where it is created rather than sending it to a grid is beneficial. I mean T & D losses aside, the issues with wildfires, electrifying everything on the grid and running more power through the lines, the cost of the utility installing new transformers or updating service entrances, etc.

          In that sense I think still creating energy on site and not moving it is beneficial. Would still love to hear any opposing view though and will read more on GBA into those discussions.

          1. Expert Member
            NICK KEENAN | | #36

            There just is no existing cost-effective home energy storage technology. For that matter, they don't really exist at scale either. The two most promising are reservoir storage -- using surplus power to pump water back up into the reservoir in hydroelectric facilities -- and molten salt. Neither of these is something you're going to install in the utility room in the basement.

            For a deep dive I recommend "Power After Carbon" by Peter Fox-Penner. He's been a GBA contributor in the past and he really knows his stuff. He believes we can decarbonize our economy by 2050. It's going to mean much greater reliance on the grid, because everything we now burn fossil fuels for on an individual basis -- vehicles, home heating, small engines -- is going to have to be replaced by electricity. He is skeptical of rooftop solar and downright dismissive of things like the Tesla PowerWall.

          2. paulneumann | | #37

            Definitely good points and decarbonizing likely won't be cheap. Can't argue that there isn't an existing cost-effective electrical energy solution currently and maybe I am just naive but in some ways even if it's not cost effective there's some carbon cost savings which could be beneficial in my mind.

            I appreciate the recommendation and will definitely read it because I am very interested in that. I am currently taking an embodied carbon course and doing Life Cycle Analysis. The more I have been on this site though the more I really value the knowledge shared so thank you. I myself hate tesla power walls and all Li-Ion Batteries so I already want to read this.

            Curious though any opinions on home thermal storage systems in terms of energy?

            I know Viessmann's Vitolator uses a unique ice storage system ( no idea the price though) but I find it very interesting.

            Also been curious about BlueFrontier

            I have also been tinkering with and working on a design and building a small thermal storage "battery" using calcium carbonate and a solar thermal collector.

            Been very interested in what SunAmp is doing as well

            Also TNO and their Heat Battery.

          3. Expert Member
            NICK KEENAN | | #39

            The thermodynamics of phase-change or some sort of chemical reaction is much more favorable than trying to store heat by just heating something up and cooling it off. That said, I'm not sure any of those solutions are really available right now at a practical price point.

            The big problem with thermal storage is how much do you store? I've never seen a system that was practical for season-to-season storage, most just have enough capacity to get through the night after a sunny day in winter. But solar is notoriously unreliable. What if you get a four-day blizzard and your solar collectors are blocked? What if it's followed by a cold snap? When you look at how expensive thermal storage is, and how cheap either grid electricity or liquid hydrocarbons, it becomes really difficult to justify thermal storage.

            There's also a philosophical question. Should every house have its own well, septic system and landfill? If those services are best provided at centralized facilities, why is energy any different? I believe that the most environmentally sound way to live is at high density in small amounts of land, and that's only really possible with centralized water, energy and waste disposal.

    3. paulneumann | | #11

      Really appreciate the Caleffi resource as well wow exactly what I was looking for! Hopefully will give me some more context as well

    4. user-2890856 | | #24

      Could I ask why low surface temp requirements make something a poor choice ?

      1. paul_wiedefeld | | #25

        It's a great choice if the low surface temperature is understood by the owner.

        1. user-2890856 | | #26

          If the NEUTRAL surface temp is not understood by the owner someone did not do their job or does not actually understand the technology . The floor is still going to feel warmer than with any other heating system , correct ?

          1. paul_wiedefeld | | #27

            Yes/No/Maybe? The floor will be warmer, I cannot speak to whether a particular person will feel it and if they can, whether it's often enough to matter.

      2. Expert Member
        NICK KEENAN | | #28

        People hear heated floors and they think of something that's like heating-pad warm, near body temperature. When they realize it's only a half-dozen degrees above room temperature and barely perceptible they feel misled. But in a tight house that's generally all you need.

        1. user-2890856 | | #30

          The industry should never have used the Warm floors garbage . Although when radiant was in it's infancy here , warm floors was not misleading because houses sucked in general

        2. paulneumann | | #32

          So how do you feel about warm board and products like that with the aluminum embedded? Ideally you want something that can conduct heat quickly but also can have thermal mass right?

          We are using a hemp-crate slab for our radiant system about 2". My thought was to embed it at the top of the slab and to use aluminum plates to help with heat conduction. The thought was to benefit from some heat going into the slab but also conducting some heat to the wood flooring.

          Curious if you have any opinion about that?

          1. user-2890856 | | #33

            Thermal mass can and will bite you in the ass everytime . Mass should be in the heat source . I prefer radiant ceilings quite honestly . You still have to insulate the slab and as already determined , it will not feel warm . Radiant ceilings allow higher surface temp if needed , will never have furniture or carpets on them and work similarly to the sun . You can use a percentage of the space to save some up front costs or use 100% and hit a grand slam by using the lowest temp supply water as possible optimizing WHATEVER heat source you're using .

            I have heard of folks using ground aluminum in their pour , this would really spread greater than any other way . Skip the plates and learn how to use ODR and that mass of the floor , or not

          2. BirchwoodBill | | #34

            We replaced all of the Subfloors in our house with Warmboard and micro-zoned the installation. The Warmboard goes under the shower and bathtubs. The system is run off our DHW heater using a heat exchanger, which is on outdoor temperature reset from 75 to 95 degrees. Warmboard has published a set of curves that you can use as part of an emitter design.

            We keep the bathroom tile floors set to 72F, wood floors to 70F. Bedrooms are at 69F.
            For Warmboard to work properly, you need R20 insulation under the floor.

          3. Expert Member
            NICK KEENAN | | #35

            The biggest mistake in the history of radiant heat was the adoption of the notion that "thermal mass" is beneficial. Most early systems were poured concrete, and the people installing them confused two different properties of concrete. Concrete has pretty high thermal conductivity, which is what you want in a radiant surface because the more conductive it is the more even the surface temperature is which makes it work better and avoids hot spots.

            However, concrete also has some heat capacity -- not as much as people tend to think, but some -- and the early adopters somehow thought that heat capacity was a good thing and what lead to even heating. They even made up a pseudo-scientific name for that quality and called it "thermal mass." The reality is you want your radiant surface to have as little heat capacity as possible, to be as responsive to the thermostat as possible. An ideal radiator warms up to heating temperature when the thermostat clicks on, and cools to room temperature when the thermostat clicks off. Modern systems like WarmBoard boast about "high responsiveness," which is what you want in any heating system. The WarmBoard system has a continuous layer of aluminum under the floor, which has very high conductivity and very low heat capacity. If your system has significant heat capacity you get lags and overshooting.

            There was just a guy here last week posting about his concrete-floor radiant system that was installed in the 1990's. He was finding that it took over a day to satisfy the thermostat when he first turned it on and wondered if that was normal. I'm sure he'll also find that on the first warm day in the spring his house is a sauna until that slab has had a day to cool off.

  6. paulneumann | | #10

    additionally to give a bit more context. The project is both a real project that will eventually be lived in but also part of a competition. One of the main goals besides NZE is a significantly reduced embodied carbon total. That being said I don't have as much practical real life experience so really appreciate all the advice and input because sometimes as a student it's easy to get mixed up with a cool idea versus the practical side/ maintenance issue/ who is going to repair it/ etc.

    1. Expert Member
      NICK KEENAN | | #14

      I really like the technique outlined in this article for minimizing concrete use in on-grade construction:

      1. paulneumann | | #18

        Awesome thank you for that!

  7. user-2890856 | | #21

    Look into doing radiant ceilings . You can use higher surface temps , use anywhere from 40 - 70% of the space of floors and nothing you put on the floor will lessen your radiator size or output .

  8. gusfhb | | #40

    I love radiant heat, and did my last house with it. I also plumbed my downstairs in this house with it. It is a great way to heat a house and can be done DIY pretty darn cheaply. PEX, plywood, tablesaw, viola, 'warmboard'

    I never hooked up the radiant downstairs here


    Once I put wood floors down, the floors are not cold.

    Oh, and the other realization, after spending 14 grand on a condensing boiler for the existing hydronic heat:

    Do you ever need AC?

    There is the kicker. My house is one head unit away from not needing a 'heating' system at all. I just didn't realize it 13 years ago.

    I would have saved ~12 grand by buying one more mini split and forgoing hydronic heat altogether

    So, all this complexity, all the calculations, all the money, if you are going to hang a minisplit anyway, use it for heat

    Yes, I love radiant heat, and all the worry about it in a well insulated building is, IMHO, hooey.

    But why?

    I will never use it again

    If you will ever need AC, you already have a heating system there, why pay for a second one?

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |