Air-to-Water Hydronic Heating for Cold Climates
rockies63
| Posted in General Questions on
So Matt Risinger did a Youtube video on electric Spacepak heat pumps for radiant floor heatiing.
https://www.spacepak.com/solstice-air-to-water-heat-pumps
Basically, they were coupled with Warmboard panels to distribute the heat but the Spacepak can also use low temperature radiators and provide DHW.
Has anyone used this system, either with warmboard or radiators, and if so, how did it function?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Search and download construction details
Replies
I don't have the SpacePak but I have what I think is a similar system, a Chiltrix air-to-water heat pump. I heat my house with a combination of radiant coils in the floor and ceiling, and hydronic air handlers. For the radiant coils I have a water set point of 95F and for the air handlers I use 113F.
I only installed the system in April so I don't have a full season of heating yet, when I do I plan to review it for GBA. It also does cooling and we made it through the summer with no issues. The 99% design temperature in Washington is 21F, it was 24 when I left the house this morning and every zone was at its set point with only three of the four zones firing. The real test will be on Saturday, the forecast is for a low of 11F and a high of 20F.
Low temperature hydronic is different. It doesn't feel hot. There are no radiators that are too hot to touch like in an old-school system. The whole house is just at a very pleasant and even temperature.
The SpacePak unit on the video was the Inverter extreme unit which is being introduced this fall. Warmboard sells the SpacePak SIM036 or the SIM060 in their current lineup.
My current house has Warmboard and it provides even heat throughout the house with little to no noise.
Attached is a preliminary design of a hydronic system that does heating/cooling/DHW/hot tub using a A2WHP for my PGH design I am working on. It also has a NG backup boiler because I don’t trust the grid. (Trees falling on power lines).
Good luck shopping for a heat pump!
The problem with all of this is the diagram you sketched has a BOM cost north of $30k. I can comfortably heat and cool a reasonable sealed and insulated place with a pair of $1800 ducted minisplits.
It is fun to play with hydronics, I have lots of it in my home, but I rather put that $25k in my pocket for a new build.
Yes, the costs get up there along with the labor costs. Hoping to do most of the work that I can. But I grew up with hydronics and love the quiet, comfortable evenly distributed heat.
"comfortable evenly distributed heat" This is definitely the case in older uninsulated homes. Places with rads were comfortable despite the fact that there was zero air sealing and almost no insulation. Places with central air was a combination of a burst of heat and followed by slow freeze between furnace cycles.
Once you fix the air leaks, have decent insulation and use fully modulating equipment, this does not happen. You would be amazed on how even, comfortable and quiet a super insulated house. Since the HVAC will run on low modulation most of the year, there is practically no noise out of the ducts. In most cases you can't even feel the air blowing from the registers unless you put your hand right in front of it.
Most well sealed houses will need an HRV/ERV for fresh air. This means ducting already. Since your envelope is low loss, the airflow required to deliver fresh air is almost the same as you would need to heat, so you are not increasing the noise by using a heat pump.
If you must have hydronic, I would look at this blog. This is one of the cleanest, efficient and well though out hydronic installs:
https://www.greenbuildingadvisor.com/article/an-introduction-to-the-flatrock-passive-house
First place you can cost save is that you will not need a full coverage of warmboard. I would actually run the numbers of your heat loss and output from the emitters. Most places only need about 1/3 to 1/2 of the floor area covered. You would put them only under high traffic areas and by larger windows/doors.
Simplifying your setup, there is no point in heating you domestic hot water with A2W, a standard heat pump water heater is way cheaper only a bit less efficient. If you want more efficient, install drain water heat recovery and you are still less than the cost of indirect+controls+plumbing.
If you want real backup, you are looking at a wood burning stove. If you are only looking at outages of a day or so, a well sealed place will simply coast through. My home looses about 4F in 6hours with the heat off in the winter. If you want heat for longer, upsize the buffer tank, much cheaper than the additional gas burner.
Hydronic cooling is hard and expensive. A slim ducted mini split is about the same cost as your hydronic air handler and more efficient. Here we are back to my initial point, anything that can cool the place can also fully heat it using the same ducts and air handler, so why even mess around with hydronics....
The ventilation requirement for a tight house is in CFM = [House Area in Sq. Ft. x 0.03] + [(Number of Bedrooms +1) x 7.5]
Which gives 90 CFM for a 3-bedroom 2,000 SF house and 157.5 for a 4-bedroom 4,000 SF house, for example. The rule of thumb is 400 CFM per ton for HVAC so that would equate to 2700 to 4700 BTU. You'd have to be super efficient to heat and cool a house of that size just using the ventilation airflow. With an open floor plan, in a tight house, you can get away with one heat emitter per floor, but with a floor that is cut up into bedrooms that doesn't work so well. Where minisplits fall down is bedrooms, they just don't come small enough for the loads seen in bedrooms.
If you want good HVAC distribution you're looking at ductwork. And yeah, the hydronic components are expensive, but right now ductwork is crazy expensive where I am. I just heard from a neighbor how was quoted $70K to replace her oil-fired boilers and radiator with a heat pump and duct work. That's not a typo, $70K. Hydronic installation is work that a person of ordinary abilities (by GBA standards) can do.
Hydronic takes up a lot less space in the house. Most of what I have is 1/2" PEX, which can run anywhere. Houses in my neighborhood are selling for $500 per square foot, if I can save 50 square feet by not needing an air handler room that's $25K.
Main suite ventilation rate is typically 25CFM. Last one I did with a bit above conde min walls and quality double panes needed 34CFM for heat. Not exactly ventilation but close enough that you could use the same duct.
You won't heat an entire house with 90CFM of airflow, but you can do a reasonable place with 350CFM which is not far off from the boost airflow rate on a larger ERV.
The only time I've seen hydronic on anything new built is for snow melt. Outside of that, ducts are the norm.
DC confirmed the ERV sizing calculations:
Qfan=0.01Afloor+7.5 (bedrooms+1) = 0.01 * 3771 + 7.5 * (5 + 1) = 90 CFM.
I did a room by room analysis of both the heating and cooling load for the new house design and laid out the circuits using the equations in Modern Hydronics.
One of problems with hydronics in a PGH is the low flow rates of the fluid. And yes, the floor temperature will only be in the 75f to 78F range. For the lower level slab, PEX around the perimeter should meet the low load.
During the last power outage of 4 hours, our current 80’s house that was retrofitted with Warmboard and ccSPF only dropped about 1 degree per hour.
The failure mode is for things breaking and a shortage of HVAC technicians or parts. Where things are down for weeks. Boilers have a high MTBF, don’t know the numbers on a heat pump.
In terms of cooling a combination of an ERV and dehumidifier should work most of the summer up here in Zone 6A. Too bad residential DOAS are not readily available. A small ducted unit could handle the load. We both agree that you need to design things for low loads.
The Chiltrix site is great at showing the derated numbers for the FirstCo units. Don’t see anything like that from SpacePak. Preliminary calculations should allow me to increase the water temperature to the right comfort level.
(Replying to Bill's post #14)
I have tubing in the floors of the bathrooms, ceilings of other rooms. With my water temperature set at 95F the bathroom floors have a surface temperature of 76F or so, which I think is perfect, it's very comfortable to stand on.
On the second floor I have two bathrooms and three bedrooms, the bathrooms have floor heat and the bedrooms have ceiling heat. I have a two-stage thermostat for the entire story, set so that when the thermostat is set for 70F the floors go on at 70.5 and go off at 71.5, and the ceilings go on at 69.5 and off at 70.5. Most of the time the floors provide all of the heat for the whole floor.
I log run times for each zone. So far for the month of December the floors have been on an average of 15:23 per day and the ceilings have been on an average of just under 25 minutes. So the floors are on most of the time and the ceilings only when it's very cold out. I really like this setup.
So I have a question, how are you running the pumps when the power goes out? Some kind of UPS just for the boiler controls and circulator? A generator? And if a generator which one (because controls often need clean power)?
No specifications yet on the generator, but it needs to able to power the boiler, pumps and bare minimum. Will probable go with a Kohler with Modbus, so I can collect the runtime data.
They talked for an hour and the guy from Texas never said a word about cooling.
Hydronic cooling is a nasty bag of worms often ends up being a separate forced air system.
Walta
Cooling is a big gap in Warmboard current offering. It sounds like they are working on a solution. Which is why there is an air handling unit in my PGH design. Love the idea of radiant ceiling panels, but the low cooling load could not justify the dollars, in Zone 6A.
Looks like a couple of weeks of my own labor to install the ductwork.
There's a reason they call it "air conditioning" and not cooling -- the air needs to be dehumidified as well as cooled. For that you need to move air over a cool surface, you need an air handler.
I wouldn't say hydronic cooling is nasty, it's simple: you run cold water through a fan unit. It's very common in commercial space. The real problem is that it offers only slight benefits over a minisplit system in a residential setting, mainly that it's easier to deliver very small amounts of cooling. And you're probably looking at about five times the cost of a minisplit system for the pieces.
I had a look at the Chilltix system - it looks pretty good.
One thing I like about these air to water systems is that it can also produce DHW (which mini-splits can't) and with the Chilltix system it can be run exclusively on solar PV.
DC-Contrarian: Did you add the Chilltrix Psychroligix Dynamic Humidity Control and the V18 Varible Power Backup Heater to your system? What size system did you go with - CX34 or CX50 and which models of air handlers? What did you do for DHW?
I passed on the DHW. I thought it was gimmicky. Basically, when you're in cooling mode it switches to heating mode, heats up the water and then switches back to cooling mode. There's no heat recovery. I have a heat pump water heater, which contributes cooling and dehumidification during cooling season.
I have two of the CS34, the CX50 wasn't available when I designed my system. I don't have the psychrologix. I did go for the V18 but I actually haven't hooked it up yet, I may be sorry tomorrow night.
I have five of the CXI air handlers, three 6,000 btu and two 3,000 btu. One of the things I like is that you can run them on very, very low. You can put the 3,000 btu unit in a bedroom and run the fan on ultra-low and get maybe a couple hundred btu of heating or cooling. That's kind of the knock on mini-split systems.
FWIW the new SpacePak Solstice Extreme is more efficient and has more low-temp capacity than the Chiltrix, but it's also substantially more expensive, partly due to lack of competition at that performance level. (They can be as expensive as lower-end ground source heat pump systems.) The limited turn-down ratio calls for a decent sized buffer to keep it from short cycling during the shoulder seasons. It only comes in one size, and is "right sized" for homes with design heat loads in the ~30-40KBTU/hr range.
But the new cold climate SpacePak doesn't NEED expensive radiant floors like WarmBoard or low temp panel rads. Even with 130F output it had decent capacity & efficiency. That is an output temp that can work reasonably with cheap fin-tube baseboard or antique cast iron rads if you have enough of it. See:
https://www.spacepak.com/Themes/SpacePakTheme/images/gallery/ILAHP%20130%20leaving%20water.svg
With the right controls and a reasonable system design most A2W heat pumps can provide some latent cooling without lots of dripping on plumbing or air handlers, but it can't be left as an afterthought if your average outdoor dew points are north of 60F in summer.
I've been in discussions with a self described "HVAC nerd" in the Boston area who has installed quite a few of the SpacePak Solstice systems and is fully up to the challenge of managing the latent cooling and designing for low temps. His proposed retrofit to my house are still being tweaked, but is a real contender. Another system I'm looking at for my home is the LGs Multi-Vs + Hydro kit, which would take care of the latent loads like any other heat pump. (Samsung has a similar small full- VRF system, but isn't shipping the hydronic modules to the US, so... ) It's a bit harder to find support for LG in my area, whereas SpacePak's factory & company HQ is about an hour's drive from my house, which in this era of supply chain complexity might turn out to matter.
I was looking at the installation manual for the Carrier/Midea dual fuel setups. It seems that all they do is use the Midea thermostat interface module to drive the outdoor unit and remove the TXV from the Carrier indoor unit.
My thinking is that nothing is stopping me from swapping in a liquid to liquid coil for the air handler and using the exact same setup. Based on the outputs, I can get 110F water at rated capacity and a bit less with warmer water.
Still a bit of a science experiment, the cost is not too bad for a split air to water setup that uses pretty much all off the shelf bits and controls and runs down to -22F. Something that is still the realm of fixable if it ever breaks down.
I've seen two guys on the internet do things similar to what I think your saying. Both were using the absolute cheapest minisplits, like sub $1,000, probably ordered from Amazon. One had the refrigerant running through a flat plate heat exchanger with water running through the other side. The other just had about 50' of 3/8" copper tubing that the refrigerant ran through sitting in a 50 gallon drum of water. Unfortunately neither one had a detailed description of how they did. And I'm not licensed to handle refrigerant which kind of takes some of the appeal away.
But they did leave the impression that there's nothing inherently complicated about an air-to-water heat pump.
However, a big part of the appeal of minisplits is how they handle modulation. I don't know whether there needs to be communication between the head and the compressor, or whether the compressor is able to look at the flow of refrigerant coming back and figure it out.
The issue with a random mini split is that you need the cirucit board from the indoor unit to run the setup. I can't see doing this in a way that doesn't look like a cludge.
The Midea unit combined with the thermostat module is much less science fair. Apparently there is still modulation, the thermostat module has an intake temperature sensor that it uses for this. I guess if I was running it as air to water, I would mount the intake sensor somewhere in the living space.
I'm sure it won't be as efficient as real air to water but under $2k for a 2 ton unit is a good deal. I'm also thinking of using a 2 zone multi split unit with one zone for the hydronic heat and the other one for cooling + supplemental heat.
So I like this kind of mad science.
My understanding is the Chiltrix began life as a hacked minisplit, although they seem to have moved up to something that is custom manufactured. My observation on how it works: there is a variable speed compressor, a flat-plate heat exchanger, a variable speed circulator pump and temperature sensors on the water intake and outlet. The microprocessor adjusts the flow on the circulator to try and keep about a 10F rise across the heat exchanger. At the same time, the microprocessor adjusts the compressor speed to try and keep the outlet water about 5F above the setpoint. If conditions are right the compressor runs continuously, putting out exactly as much heat as is being consumed by the emitters. There are two exceptions, if the compressor can't put out as much heat as required, or if the compressor can't modulate low enough to put out as little as required. If the compressor can't put out enough the compressor is just run full out. If the compressor can't modulate down to put down little enough the compressor has to cycle. Chiltrix recommends a minimum 17 gallon buffer tank, if the load is lower than the lowest setting of the compressor the buffer tank temperature will creep up. When it gets 2C above the setpoint the compressor shuts off, and rests until the tank is 2C below the setpoint. Based on observations I'd say the minimum modulation is about 20% of total capacity, or about 4800 BTU. It takes about 1,000 BTU to raise the buffer tank from the low point to the high point, or about 12 minutes of runtime.
With something like an Arduino a control circuit for this would be straightforward. The only unknown part to me would be interfacing with the compressor, what the protocol is for communicating the power setting. If it's any kind of standard protocol, Arduino has libraries for almost every one.
It would be a cool project.
DC,
That is excellent info.
I'm really trying to avoid any DIY control bits here. If this will work in the long term, it needs to be built with off the shelf parts and no software. I already have a PLC running my ERV, I really don't want to have more software in the house.
Another bit of mad science to think about:
There is a device called a "CoolBot" that provides the control electronics to use a window air conditioner to create a cold room. See: https://www.storeitcold.com/
The way that it works is that it has a heating element that you tape to the temperature sensor in the air conditioner to trick it. The control board turns the air conditioner on by heating the sensor and letting its regular control mechanism think it's in a warm room. I wonder if you could control a minisplit in a similar way.
Akos, is there a model number or link to these units?
I'm interested if these are any different than what Thermatlantic already does with its DX2W modules.
Thanks for letting me know about the Thermatlantic system. From some quick Googling it sounds interesting. John Siegenthaler has a review here:
https://www.pmmag.com/articles/96692-a-unique-air-to-water-heat-pump-system-from-a-business-entrepreneur
Here is the instructions for the relay module. The interesting bit is pairing a 38MARB outdoor unit (rebadged Midea hyper heat) with a standard fan coil on P.8.
https://ca-manufacturer-documents.s3.amazonaws.com/CarrierDocs/IM-KSAIC03-07.pdf
I think is should work with something like the DX2W module.
>whereas SpacePak's factory & company HQ is about an hour's drive from my house, which in this era of supply chain complexity might turn out to matter.
Funny you should say that, Chiltrix is in Virginia, I drove to their HQ to swap out some equipment not that long ago, it was just easier that way.
I looked into the LG Multi Vs + Hydro kit about a year ago, and the answer I got was it wasn't really available. More specifically, LG wouldn't sell it to anyone who hadn't had factory training, and that the training wasn't happening. Has that changed?
I also recall that the minimum system size was pretty big, maybe five tons?
>More specifically, LG wouldn't sell it to anyone who hadn't had factory training, and that the training wasn't happening. Has that changed?
They still require the training, which has recently been resumed (according to one as-yet uncertified contractor I have been talking to.)
>I also recall that the minimum system size was pretty big, maybe five tons?
Five tons is the LARGEST single phase Multi V.
But the smallest Multi-Vs that can still work with the low-temp Hydro Kit is 4 tons (which is still reasonably sized for my loads. YMMV)
Note that the smallest, largest and ONLY SpacePak Solstice cold climate hydronic heat pumps is also nominally rated 5 tons for cooling. But it's cold climate max capacity is less than 4 tons at the outdoor temps that matter for climate zones 5 & up.
The small Chiltrix is two tons.
hi dana,
i'm currently trying to identify a system to install in a north-east cz-6 new build, and trying to find hvac nerds installers in the region, would you be able to recommend your associate, or others in the area?
DC: Did Chilltrix mention how small or large of a space can be heated using one of their CX34 unites, or how many CXI handlers are needed per zone/size of zone?
Every house is different.
There aren't any rules of thumb that work reliably, to size heating and cooling equipment you need to have a Manual J done. Once I had a Manual J I sent it to Chiltrix and they were very helpful with my system design.
If you want to get an idea of your system size, this article shows you how to do it using your fuel usage history:
https://www.greenbuildingadvisor.com/article/replacing-a-furnace-or-boiler
For my upcoming build, cost is not surprisingly becoming a factor in the planning of our cold climate build.
I couldn’t help but notice that the third post on this discussion mentioned a 25k difference in price between two ductless minisplits, and an air to water heat pump. If the ATWHP was only used to heat a single zone concrete slab, without a buffer tank or any sort of domestic water heating, would the up charge still be as significant?
I only ask because, for me, 25k is a significant amount.
Thanks!
Ben
Ben,
It depends on who is doing the hydronic install. If you can DIY, a simple system like you propose can be done on the cheap. Even if you sharpen your hydronic layout pencil, you are still looking at a BOM cost of at least $10k. If you compare that to a multi position air handler heat pump with ducting, the material cost is about 1/2. Plus you still don't get cooling for that cost.
If you're only doing a single zone there's not much advantage to hydronics. Where hydronics really shines is in a building with a lot of rooms with varying loads, you can get small amounts of heat to lots of spaces.
The BOM discussion was based on the number of zones, costs go up as you add zones. Zones can add comfort and evenly distribute heat/cooling through a house - depending on the geometry.
The heat loss for the garage slab calculation identified three circuits of approximately 275 long connected to a manifold, a circulator or two, plus injection controls. You also need to add in the insulation under the garage and the labor. Yes, it does add up.
The cost to heat a hot tub - is the heat exchanger, insulated PEX, a circulator, labor, and a penetration to get to the hot tub. So, you need to make the economic trade-off of using resistance heat vs. a heat-pump.
For a great room design, I can install a SpacePak fan coil and mount in high in the wall - similar to a mini-split - then run two insulated PEX lines to the basement. This is just plumbing and does not require refrigerant skills. That is labor I can do myself.
This is all wonderful info, I thank you all. Admittedly I am not the greatest at DIY, but my experience as a snowmaker gives me confidence that I can run water through pipes, and take precautions against freezing.
I might have to learn more about the SpacePak. Is the fan coil something that gets integrated with a ducted air delivery system? The option to heat a slab to me seems to indicate that one could put PEX in the slab, but not commit to finishing the hydronic system first thing, allowing for effort to be put elsewhere.
My goal with the single zone hydronic system is to use the slab as a thermal battery in case of power outages in our rural area. We are unlikely to have a wood stove in our all electric build
Thanks again,
Ben
The idea that concrete can act as a "thermal battery" is way, way oversold. Let's say your slab has a footprint of 1000 square feet and is 4" thick, that's 333 cubic feet or about 33,000 pounds. Water has a heat capacity of 1 BTU per pound per degree F, concrete is about half that, so your slab has a heat capacity of 15,000 BTU per degree F. To pick a number, let's say the heating load is 30,000 BTU per hour, so a 10F drop in slab temperature gives five hours of heat. Which is not nothing. The problem is actually getting that heat out of the slab, in order for the heat to come out there has to be a temperature differential, you need big swings in the interior temperature to get that heat out.
When the heat isn't out that slab works to make your house less comfortable. It takes time to heat up and cool off, which means when the heat clicks on there's a delay before the slab warms up and starts putting out heat, and a delay when the thermostat cuts off before the slab stops putting out heat. With the slab in the example above, if you have a properly-sized boiler of 30,000 BTU/hr capacity it's going to take the boiler five hours to raise the slab temperature by 10 degrees. So you get overshooting. The big selling point of systems like Warmboard is that they have low heat capacity and can respond quickly to the thermostat.
Note also that the rest of the house -- the drywall, framing, tile, flooring and other materials -- have significant heat capacity, typically quite a bit more than the concrete slab. In most cases if you want to build a house that holds its heat you're better off adding more insulation than adding more heat capacity. Adding insulation has the benefit that it increases comfort and efficiency all the time.
John Siegenthaler outlined an air to slab design recently. You can store a lot of BTU in a concrete slab. It is more applicable to commercial sites than residential. But for basements and garages you have somewhat of a battery.
I guess it depends on what you mean by "a lot." Fifteen tons of concrete holds less BTU than a gallon of heating oil.
For Minnesota, a lot would be 8 hours of no power, no pumping, no BTUh. A heated basement slab evens out the bumps in the road. For a garage, it means opening the garage door at -29F and not having the thermostat click on immediately.
I just think that people way over-estimate the heat capacity of concrete. I know my house has more heat capacity in the drywall than in the basement slab.
Those numbers make a lot of sense. My girlfriend will appreciate the cost savings, and it doesn’t take much to convince us to add more insulation.
Thanks again,
Hope your holidays are going well,
Ben
@DC_Contrarian, are you available for a couple of questions regarding your Chiltrix setup? I've been looking at that system for a couple of years now and would appreciate your comments. My email is matthewwilliamson55 [at] gmail.com if you wouldn't mind reaching out.
A2W systems baffle me. If you type in "Fujitsu General" in your search engine you will see references to Fujitsu's large product catalogue-for international sales. One third of the catalogue highlights air to water heat pump systems called "Waterstage". And yet the air to water Waterstage equipment does not seem widely available here to the US public at large. It is as if that is a different HVAC world across the Atlantic and Pacific. But Fujitsu must be doing well enough in Asia and Eurpoe with the Waterstage system to be so prominently highlighting them. Why is it so hard to get information about these systems here?
In the video linked in the original post, the claim is made that in Germany air-to-water heat pumps are now outselling fuel-burning boilers, and that in China two million air-to-water heat pumps are sold a year.
I will say that air conditioning is much more common in the US than in other places.
Interesting....I don't understand why an outside condenser used for A2W in Winter can't also run A2A for air conditioning in Summer... The Waterstage system looks like it uses a standard Fujitsu low temp condenser attached to a hydronic internal pump unit. I don't see why a large multi condenser couldn't be designed to run a few A2A interior units as well for A/C...
One of the big fundamental advantages of A2W over A2A is the ability to modulate down. It's hard to get A2A to modulate much below a quarter of full capacity. When you put multiple heads onto one compressor this becomes a problem to try and get a little bit out of just one head. In A2W you can have a water tank in the house that absorbs the excess capacity and stores it for a few minutes. Mixing A2A with A2W you would lose that benefit.
A2W cooling is the norm in commercial buildings in the US. While a house might have a 2-ton compressor, an office building might have ten, twenty or fifty tons in an air-to-water chiller. I believe if there were widespread adoption this has a fundamental advantage over minisplits, in that it's cheaper and easier to pipe cold water around the house than refrigerant. But right now in the US air-to-water is a niche product and minisplits are cranked out by the millions. You can basically buy a whole minisplit today for what the fittings would cost to add a zone of hydronic cooling. I don't believe that this is something inherent to a2w; in fact if you compare a minisplit with a hydronic cooler the minisplit is much more complicated. It's just economies of scale.