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

Air-to-water heat pumps for heating, cooling and hot water

sequential | Posted in Mechanicals on

I am looking for feedback on an imagined hydronic heating, cooling, and domestic hot water system using existing radiators (or radiant floors), an air-to-water heat pump, dehumidifier(s), and a resistance heating coil and water tanks.

First and foremost, I am an architect, and so I am familiar with, but far, far from expert in HVAC systems. But this is one I dreamed up as an alternative to air-to-air heat pumps in retrofitted buildings with extant hydronic heating systems. I am looking for feedback, either big picture or detailed; anything that stands out, for good or ill. The system is obviously unusual and requires a slightly different ‘conditioning philosophy’, and so I will lay out what I see as positives and negatives here:

+ Adapts to existing hydronic systems—lower embodied energy in system replacement

+ Radiant system, usually a better ‘feeling’ heat and cool, and at smaller deltas

+ The hot water tank can be used as a heat battery for PV systems in the winter.

+/- On the coldest winter days, the heat pump return is pre-heated by the resistance heater in the water tank—powered either by grid power or excess PV. This could help increase heat pump efficiency (though I’m not sure the math necessarily works out here…)

+ Utilizes waste heat in summer for water heating, thereby…

+ …reduces the amount of waste heat expelled into the environment in summer.

+ Ventilation by opening a window is less wasteful than with air systems, as radiant systems can rely more on thermal mass—gives users easy option to control their own local comfort

– requires dehumidifier, either whole-building or in conditioned rooms, to prevent condensation on radiators in summer. Certainly an inconvenience

– supply and return to radiators will need to be insulated to prevent condensation occurring in unconditioned cavity spaces

– does not provide as much cooling ‘power’ as typical air-to-air systems. This could also be a positive, as it could be argued it provides sufficient though not excessive cooling. Personal taste.

– I think there might be an issue with how the heat exchange tank works in Winter.

– Not sure what’s the max return temperature the heat pump, indoors unit could receive in winter. Excess PV could heat the domestic hot water tank to quite a high temperature, in theory.

– all the temperatures I’m suggesting are super back-of-the-envelope. Many are likely not realistic

 

This concept relies on a couple ‘philosophical’ beliefs that I know all here may not agree with–but hopefully you can still provide feedback.

First, I rate thermal mass as more important than air-sealing.
I live in Germany, where natural ventilation is king, and buildings are usually built of masonry or concrete, and therefore have high thermal mass and slow swings in temperature. This is in some ways very different from that other german strategy, passivhaus, which is perhaps better suited to wood-framed low thermal mass structures. I have previously lived in New England, where this type is dominant, and I was a believer in air-sealing and lots of insulation. But I have come to appreciate that the ability to ventilate locally without relying on active air exchange systems is pretty important.

Second, this system would likely provide a ‘lower’ level of thermal comfort than the much easier and more widespread minisplit.
I worry the normal air-to-air heat pump is bringing cooling to places that never had it before, and once you have the capability you’re of course going to use it. So it may be more efficient, but there may be an overall increase in energy spent on conditioning. I also worry about what effect all this additional summertime heat rejection will do to air temperatures in cities. In addition, it doesn’t address hot water, nor PV storage, the latter of which is rapidly becoming the biggest hurdle in renewable energy adoption. Finally, though this is a less of the case in the US, where hydronic systems were often leapfrogged from steam straight to forced air, there are still lots of extant hydronic systems worldwide, and as with all green tech, embodied energy is usually the biggest determining factor in a system’s carbon payback period. Maintaining existing hydronic systems would be a significant coup in this respect.
So–though comfort is lower, it can still be sufficient, all while dramatically reducing energy, both direct and embodied.

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.

Replies

  1. Expert Member
    DCcontrarian | | #1

    Install two heat pumps: an air-to-water that serves the existing radiators, at reduced temperature, and an air-to-air that provides cooling in the summer and makes up the difference in heat output in the winter.

    1. Expert Member
      DCcontrarian | | #2

      Oh, and use a heat pump water heater for your hot water. There's your heat extraction in the summer.

  2. walta100 | | #3

    hydronic heating is absolutely great when you are visitor to a leaky uninsulated house that someone else is paying for the fuel to heat. The warm floors and hot radiators feel great and it is so quiet what a great luxury when fuel was so inexpensive.

    In the US installing maintaining and fueling hydronic systems is very expensive going with a hydronic system will at least double and likely triple your install budget. Mostly because hydronic cooling is more a theory than a reality if you live in a humid climate forcing one to install a second system for cooling.

    Today’s code requires you to build a tight well insulated house and suddenly it is simply impossible to get the luxuriously warm floor or radiators that you recall. Because if you put 100° floors under a tight well insulated house the air inside the home is going to be 90°maybe a little more. When you set the thermostat 69° the floors are more like 79° and are now cooler than your feet and don’t feel warm.

    In the US air to water heat pumps are as rare as gold and priced like they are made of gold. The systems are complex finding anyone competent to install it is no small feat getting someone else to service it later is almost impossible.

    Any air to water HP is bound to have a lower COP than a similar air to air system simply because it must make the water at least 30° warmer than the house needs to be. Look at the COP charts for any HP and you will see the larger the temperature difference the unit must overcome the lower the COP becomes.

    The only real upside is that it is quiet and takes up less space but a forced air systems can be made to run quietly if you make that the goal and the forced air system will be running all summer to provide cooling.

    I understand almost every building in Germany is heated with hydronic and few require cooling. If you truly believe the world is warming how long will it be before everyone is demanding cooling?

    Walta

    1. Expert Member
      DCcontrarian | | #5

      "Today’s code requires you to build a tight well insulated house and suddenly it is simply impossible to get the luxuriously warm floor or radiators that you recall. Because if you put 100° floors under a tight well insulated house the air inside the home is going to be 90°maybe a little more. When you set the thermostat 69° the floors are more like 79° and are now cooler than your feet and don’t feel warm."

      Just to add on to this a bit. As longtime readers may remember, I live in a bit of a mad scientist house of my own making. I wanted heated floors in my bathroom, but the house is tight and well-insulated (the typical GBA reader might consider it "average," the rest of the world probably would call it "extraordinary.") Aware of this limitation, I designed a multi-stage heating system, where the floors always come on first. If they can't keep up there are panels in the ceiling that come on next, and then an air handler; the air handler has four speeds. So six stages of heat in all, and separate for first and second floor.

      Right now it is very cold for DC. My design temperature is 22F, right now it's 20F in the heat of the day and it was 13F last night. I log the run time of the heating system, and the first floor spent most of the day yesterday between stages 2 and 3, which is about what I would expect. The second floor spent almost the entire day at stage 1 -- only the heated floors, in only the bathrooms -- with 1:13 of stage 2. The second floor is 1470 square feet, the two bathrooms have 107 square feet of heated area.

      This is kind of astonishing to me, and I think there must be a contribution from the floor below. My Manual J calculation is 8000 BTU/hr for the second floor at design temp, there's no way I'm getting that.

      But the broader point is that in a modern, tight house floor heat has to be used very sparingly to give the "warm toes" effect.

    2. sequential | | #7

      I think the benefits of hydronic come more from that a radiant heat source perhaps feels a bit warmer than a forced air source, when they are both outputting the same energy. (1)
      The massively overheating steam or hydronic radiators in the leaky house of old are not really the comparison nor the goal here.

      At the moment, air-to-water pumps are not really availble in the US, though since this was the case with air-to-air not that long ago, I consider this not really a problem for the theory. Perhaps they will never be adopted, but I guess I wouldn't feel overly-restricted by the current market. As we all know, that can change quite quickly.
      There are certainly commerical products available in the US, though these of course for much larger, more complex buildings that benefit greatly from economies of scale, and at least in my experience, are never the only HVAC system. In Germany there is already a consumer version for heating and cooling available (2), and who knows, perhaps it will make it's way over eventually.
      I was mainly just curious what a well-informed audience perhaps generally unfamilar with the product would make of it, and what flaws one sees in applying it to a climactic regions very different from Central Europe. I can't imagine this would make much sense at all in the many cooling-dominated regions of the US, but I figure there are enough densely-populated areas (NE, PNW) that have somewhat similar climates, and for which there might be an application. I guess the major difference remains that buildings in Central Europe are in general far more thermally-massive than in North America, which really benefits radiant systems.

      I think your point on COP could likely be quite true, though perhaps with a smaller necessary delta. I would have to get a lot more serious about numbers and do a carbon payback to see whether the carbon savings in reduced demolition and less new installation outweighed the efficiency difference, and for how long.

      Walta, thank you for your response.

      (1) https://www.researchgate.net/publication/309560937_Thermal_comfort_in_buildings_using_radiant_vs_all-air_systems_A_critical_literature_review

      (2) https://www.wolf-heiztechnik.at/Heizen/W%C3%A4rmepumpen/Split-LuftWasser-W%C3%A4rmepumpe-BWL-1S%28B%29 (in German but browser translation tools should give the gist)

      1. Expert Member
        DCcontrarian | | #15

        Summary of the linked article:
        "We performed a literature review to assess if radiant systems provide better, equal or lower thermal comfort than all-air systems. We included only peer-reviewed articles and articles published in proceedings of scientific conferences. The publications found have been classified based on research methods used. These include: (1) building performance simulation (BPS), (2) physical measurements (in laboratory test chambers and in buildings) and (3) human subject testing/occupant based surveys. This review identified eight conclusive studies: five studies that could not establish a thermal comfort preference between all-air and radiant systems and three studies showing a preference for radiant systems. Very few studies were based on occupant feedback in real buildings suggesting a significant research need. Overall, we found that a limited number of studies are available and therefore a solid answer cannot be given. Nevertheless, there is suggestive evidence that radiant systems may provide equal or better comfort than all-air systems."

        I love hydronics, but this seems like weak sauce.

    3. Expert Member
      DCcontrarian | | #14

      "Mostly because hydronic cooling is more a theory than a reality if you live in a humid climate forcing one to install a second system for cooling."

      I want to expand on this thought because it's important.

      There is a well-developed process for sizing HVAC equipment. However, AC is more than just cooling, it's responsible for dehumidifying the air as well, that's why it's called "air conditioning" and not cooling. There is no established process for sizing cooling equipment separately from dehumidification. The only safe way -- and probably the only code-compliant way -- to implement cooling using hydronic coils would be to install a conventional system as well, and only use the hydronic coils when the conventional system provides all of the dehumidification needed without providing all of the cooling needed.

      I know that in my climate, Washington, DC, that last bit basically never happens. It's not uncommon to run a dehumidifier in addition to air conditioning in the summer.

      If you live any place where dehumidifiers are sold in big box stores in the summer hydronic coil cooling is probably not for you.

    4. scsiguy | | #16

      > When you set the thermostat 69° the floors are more like 79° and are now cooler than your feet and don’t feel warm.

      This comes up often on GBA, and while it makes some logical sense, its not accurate. Yes, the floor temps (according to embedded sensors for my install) range from the mid 70s to low 80s, but this makes the floor noticeably "not cold". And in the bathrooms where the t-stats are set 4F higher than the rest of the house, the floors feel warm on the feet. Not hot like a radiant floor installed in an 80s era Colorado mountain condo operating in the winter, but nice. No matter the outside temperature, I'm comfortable walking around barefoot. I can't do that at my parents house which has a similar mix of slate and oak flooring, but baseboard heat. Whether this matters enough to you to outweigh the high upfront costs is a totally different question.

      > Any air to water HP is bound to have a lower COP than a similar air to air system simply because it must make the water at least 30° warmer than the house needs to be.

      This also feels too strong a conclusion. You are basically saying that the condenser temperature for a water based system is always higher than for something like a mini-split. My system targets 100F radiant temps at the outdoor design temp of 0F. But, for the majority of the heating hours during the winter, the target water temp is only in the mid-80s. This is for t-stats set between 68 and 74F. Are A2A systems really putting out air at much lower temps than 80-85F?

      You also have to take into consideration the whole system power usage, not just the heat pump COP. My system covers 3400sq/ft, across two floors, with 15 independent zones, all driven by one heat pump, and about 60w max of pump power for distribution (much less if only some zones are calling). It would have to be modeled, and the answer would vary greatly based on approach (e.g. number of heat pumps and heads/fan-coils), but it's not obvious that an A2A system will have better overall efficiency.

      Hydronics are expensive, and outside of commercial settings, cooling with hydronics is rare. Those are the big downsides. We should focus on these real challenges when answering questions in these forums vs. what feels more like speculation than grounded advice.

  3. Expert Member
    DCcontrarian | | #4

    Oh, and it never makes sense to use an outside heat pump for domestic hot water. Use a heat pump water heater instead.

    The outside installation costs a lot more for initial installation, saves nothing on operating costs, saves no room in the mechanical room, doesn't provide more hot water, and interferes with heating and cooling. Meanwhile the HPWH provides free cooling in the summer.

    A lot of hydronics guys think of heat pumps as just like boilers, only noisier. They aren't. This is an idea left over from boiler days that has no room in a heat pump world.

  4. Expert Member
    BILL WICHERS | | #6

    You're never going to see much cooling with a 65*F "cold side" supply water temperature to radiators, especially old-style radiators made for heating systems. For comparison, commerical chilled water cooling systems (I work with these a lot since they're common in large datacenter facilities) use 45*F supply temperatures on the chilled water loop, 55*F return, and the "radiators" are fin-tube type similar to car radiators, so LOTS more surface area to work with. You're going to get essentially zero dehumidification with 65*F supply water too, unless you're in a spectacularly damp enviornment, and even then, it won't be very effective.

    With heating, your only advantage here would be being able to retrofit onto existing hydronic heating systems. For new installs, it will be very expensive (as others have mentioned), and will likely be less efficient than forced air-type systems. Walta already did a pretty good explanation of explaining why you won't get that nice "it's all so warm" feeling in a modern structure too. Back in times of old, buildings were sided with planks (not sheet products like plywood or OSB), so loads of air gaps everywhere. No insulation at all, either, just air between the studs. This meant lots of air movement through the structure, so lots of energy loss. All the heat you put in essentially "moved through" the structure, so everything inside got warmed up. That was nice, but it was like running a woodstove for heat and controlling the temperature by keeping the windows open -- not very efficient, even though it might be pretty comfy. About the only remaining benefit would be the radiant warmth you'd get when near a radiator.

    If you were to do something like this, I'd try to keep it as a system for retrofitting existing structures so that the existing hydronic systems could condinue to operate. This may be a good idea in areas trying to electrify older structures that had boilers fired by, well, anything that burns :-)

    BTW, I'm not a fan of restricting things like cooling because it "might make people use it". We should NOT be de-prioritizing the comfort of building occupants in the name of saving energy. It's much better to find ways to keep the occupants comfortable while saving energy doing so by using better equipment and methodologies. I've never like the "do less with less" mentality, we should be trying to do more with less, or at least the same as we had with less energy.

    Bill

    1. sequential | | #8

      Yeah, this cold supply is likely pretty conservative, though the idea was it provides some cooling, and does not try to compete with the amount of cooling provided by a typical air-to-air heat pump.

      I agree that it is important not to totally sacrifice building comfort on the alter of efficiency, as it is sometimes more difficult to convince people to change lifestyle than spend money (though both are quite difficult ;)

      I would also argue, however, that ultimately inconsequential changes in lifestyle (changing setpoints from 68-75F to 64-82F) drastically reduce energy use without spending any carbon (or money) on new systems, insulation, complex tech etc. The level of comfort we aim for should probably be able to be extrapolated globally and still work from a carbon balance perspective; and I'm not so sure a Passivhaus with all the latest tech can be.

  5. TracyS | | #9

    I'll add a couple of my thoughts here on DHW and air to water heat pumps. i installed a chiltrix system on a new build house.

    Initially, I installed the side arm DHW heater tied into the air to water heat pump. Before we finished the build, i had removed that DHW and replaced it with a heat pump DHW. The negatives to having a DHW tied to the heat pump i found were:

    Complex and expensive - expensive three way valves to direct the heated water to the tank
    Low maximum temperature - The heat pump controller would limit the DHW to 120F. And then schedule an element in the tank to kick up the temperature periodically to 140F to mitigate legionella. The relay and the internal element was not included with the system. And at less than 120F, the shower farthest from the tank would not get warm enough for me.
    Maintenance - The indirect line in the DHW tank ran the water from the heat pump, which in a cold climate, has glycol for freeze protection. If you have to disconnect the DHW tank, you are capturing that glycol water from the DHW tank and then dealing with recharging the system afterwards.
    UL listing - The DHW tank that came with the system was not UL listed, and so regional building inspector rejected it. This was the final straw in replacing it with the heat pump system .

    1. Expert Member
      DCcontrarian | | #11

      I agree with all of your criticisms of the DHW from the heat pump. In addition, there's a problem that heat and AC are disabled when the water is heating, which can be a pretty long time because the output of the heat pump is not that big by water heater standards.

      The UL listing thing would have made me mad.

      1. Expert Member
        Akos | | #13

        I'm with TracyS. DHW off an air to water heat pump is a bad idea. Mine worked well till we hit a polar vortex at which point I had to disable it as it was taking too long to recover and we were loosing house heat.

        HPWH is your friend. Simple, cheaper BOM than indirect and much easier install.

    2. scsiguy | | #17

      There's more than one way to get DHW from an A2W heat pump. The Nordic ATW includes a desuperheater that runs anytime the heat pump is serving any other load. Using the feature drives up the installation cost though, since you need a second storage tank that sits before the DHW heater's tank. (There have been other posts on GBA about products that serve a similar function and can be added to the line sets for A2A heat pumps.)

      Also, since the Nordic is a split system, the indoor loop doesn't require glycol. So you can couple it to a side-arm tank without having to mess with glycol. But I always advise folks to just install a heat pump hot water heater.

      1. Expert Member
        DCcontrarian | | #18

        Agree on the heat pump hot water heater.

        Getting DHW from an AWHP can be summarized as "just because you can do something doesn't mean it's a good idea."

  6. TracyS | | #10

    As far as the cooling aspect, i am in Colorado with low relative humidity, cool nights and a few host days during the summer. I did switch the system to cooling mode during the middle of summer. On really warm days, i had the master bedroom thermostat call for cooling to take the edge off the heat. I don't think anyone who likes traditional AC would be happy with it. It does not have that instant cooling effect of standing in a cold air stream. But the dog likes it (cool floors). I would not use the radiant for cooling in a humid climate due to risk of condensation on the floor, or pushing a dehumidifier to keep the dew point low enough.

    As far as hydronic heating goes, i know there are arguments on this forum against it for low heat loss homes. We did 4" of polyiso on the exterior, and r-21 pink in the stud bays, r60 cellulose in the attic. But for me, the advantage of hydronic, in floor heating is the ability to easily zone individual rooms with dedicated thermostats. And not worry so much about where erv air flows and maintaining desired temperatures in each room. I know you can go with air handlers and electronic dampers. But my understanding is you ideally want a variable speed air handler to be able to deal with the number of dampers that may be open at any given time. I have 9 thermostat controlled zones in the house, plus one zone for snow melt on the porch stairs and the area of driveway near the north facing garage. I my opinion, the complexity and cost of a zoned air system starts to compete with a zone hydronic heating system. And i was more comfortable installing the pex heating lines than i would have been with zoned air ducts.

    1. Expert Member
      DCcontrarian | | #12

      One approach is to use a fan coil unit in each room. That allows you to put each room on its own thermostat. Hydronic systems handle small loads much better than any air-to-air systems, you can put a 3,000 BTU/hr air handler in a bedroom and run it at the lowest fan setting at maybe 10% output and get perhaps 150 Btu/hr. No air-to-air system would let you do that.

      The problem is that the fan coil units available for the US market right now are kind of disappointing, they're ugly and the controls tend not to work well.

      1. piperspace | | #19

        I used these for a remodel of a 1911 house in San Francisco. Control was indeed a challenge. I don’t find them ugly. Support from Toronto was good.

        https://www.jaga-canada.com/products/trench-convectors/

        1. Expert Member
          DCcontrarian | | #20

          I want to like Jaga, but I find their website so inscrutable it's impossible to tell what they sell.

          1. begreener | | #21
          2. Expert Member
            Akos | | #22

            I'm with DC, those are ugly, I would not want to look at those fan coils.

            I would want something that can be mounted inside a wall and come with a real air filter.

            Can't find it now, but a while back I came across a commercial hydronic coil that was air assisted. It was pretty close to something like a convector but you pipe the fresh air supply to it to boost the convective flow rate. It felt like the best of both worlds. You can use much smaller pipes you will already need for fresh air distribution and have something that can do both cooling and heating in a relatively small package without needing extra fan in it.

Log in or create an account to post an answer.

Community

Recent Questions and Replies

  • |
  • |
  • |
  • |