Are air-to-water heat pumps about to boom?
I think we’re about to see a boom in the popularity of air-to-water heat pumps (AWHP’s).
First, air-to-water is fundamentally more cost-effective than air-to-air
One of the reasons minisplits have taken off in recent years is that in most cases it is easier and cheaper to run a thin copper tube around a house than ducting to move the same amount of heat. An air-to-water heat pump paired with a hydronic air handler is essentially equivalent to a minisplit, except that water is being used as the medium instead of refrigerant.
· Water is cheaper than refrigerant, both for the medium and for handling
· PEX is cheaper than copper
· Crimping is cheaper than brazing.
You don’t need an EPA license to fill a hydronic heating system!
Right now, air-to-water equipment is expensive. But in other parts of the world it’s very popular, I’ve read that air-to-water heat pumps are outselling boilers in Germany, and two million a year are sold in China. If you go on Alibaba there are literally thousands of AWHP’s for sale, some for less than a thousand dollars. Conceptually they’re no more complicated than minisplits; once they are adopted in large numbers there’s every reason to believe the cost will come down to where they are competitive.
In terms of operating cost, sellers of AWHP’s claim they have better efficiency than minisplits. That may or may not be true, but there’s no reason to believe that the performance would be any worse.
Second, hydronics can deliver more comfort.
The key to comfort is delivering heating or cooling where and when you want it. Hydronics can deliver heating or cooling in smaller and finer increments than other heat delivery methods. The simplest emitter for a hydronic system is an air handler unit (AHU). These are conceptually similar to the heads on a minisplit, but unlike a minisplit they come in sizes small enough for a small bedroom, and running them at very low outputs is not a problem. It is rare to see a minisplit head smaller than about 6,000 BTU/hr. AHU’s come in sizes down to 3,000 BTU/hr, and can be run at ultralow fan speed to produce just a few hundred BTU/hr.
What makes this possible is that AWHP’s will have a buffer tank inside. For example, Chiltrix provides a 17 gallon buffer tank. When set up as Chiltrix recommends this can capture 12 minutes of the heat pump running at its minimum capacity, so regardless of the demand for heat the heat pump won’t cycle. It’s true that the tank takes up space inside the house, but it’s only the size of about four 5-gallon buckets stacked.
With hydronics, it’s easy to mix and match, so beyond AHU’s you can add conventional radiators, baseboards, and floor or ceiling radiant heat. AWHP’s are particularly well suited to radiant heat systems because their ideal working temperatures are in the radiant sweet spot. With an AWHP you can run all of those devices off of one heat pump, and you can control them all centrally.
In commercial buildings hydronic heating and cooling has been the norm for a long time. The AWHP’s we’re seeing today are a marriage of the small compressor technology of minisplits with the distribution of large commercial systems, a lot of the AHU’s are based on commercial designs.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
No boom until they can provide 160- ish degree water in sufficient volume. They will become popular when they can replace baseboard boilers and unless you have specific low temp baseboard (which most folks don’t have) 130 degree water simply is no going to cut it in majority of retrofit scenarios. In their current state the aswhp are a boon for well designed radiant heat, but then again- retrofit is not easy and only new constructions have radiant…
Just my 2 cents
The baseboard you'd use at 130F is the same as the one for 160F, you just need more of it.
Maybe! They have strengths but are expensive at the equipment level, supporting hardware, and labor level. Also, the majority of Americans don’t have hydronic heating and don’t seem to want it. Ducting has major advantages for cooling and has won out in America.
That said - they’re better than a multi-split in many ways, so that’s a good market to target.
Paul, in cold climates (zone5-up) they are fairly common with millions of houses using baseboard heat. As I mention in my post response above (below?) baseboard heat is superior to forced air (ducted or ductless) in terms of comfort in cold climate houses that are not super tight (i.e., the vast majority of existing houses). For this reason they are still regularly being built around me (NW NJ), paired with central air (rarely heat pumps) or mini splits for cooling.
I agree they’re more common in certain regions, I’ve lived with a few forced water systems. But I believe the total hydronics share is <10% of existing houses and likely much less for new houses. Most people don’t want them in the US.
Hydrosolar has their house brand units on sale right now. And they have a 1/4 ton AHU for just under $700 now. Pair a couple of those with the hydronic unit and you have a good solution for the bedroom distribution issue that mini splits have. For a buffer tank, it would be nice to just throw in a big 40 gallon electric water heater for $400, but that's a lot of glycol you'd need.
Interesting article!
I have oil fed multi zone baseboard and Fujitsu low temp mini splits now. I love the Fujitsu units because they have cut my oil bill dramatically. However, from a comfort perspective the baseboard is superior (in our slightly above average 1980's house) because it is more uniform, less stratified heat delivered closer to where we want it (e.g., the floor of the bedrooms and bathrooms). We are right now going through a cold spell where the temp here in NW NJ will hit -6F tonight. My Fujitsu's can handle it but my wife will insist on leaving the boiler/baseboard on in the bedrooms, bathrooms and a few other places because of this.
Unfortunately, as greenright notes above, baseboard is typically designed for 160-180F water and replacing it all with 120-130F high efficiency baseboard is expensive.
However, I wonder if there is a way to replace my boiler with an A2W HP to run the baseboard at 120F, coupled with the Fujitsu mini splits as the primary source of heat ? Like most houses with baseboard, my baseboard is somewhat oversized based on 160-180F water temp. And if the baseboard is only the secondary source, with the mini splits being primary, I wonder if 120F water will throw off enough btu's in the right places to add just enough comfort where we want it. I would also have to figure out how to use the thermostats to keep both systems running at the same time, instead of one or the other. Alternatively, I wonder if I can just splice in a few high btu European radiators or hydronic air handlers in the existing baseboard run, instead of replacing all of the baseboard..... and then use the baseboard as primary heat, with the mini splits coming on at the coldest temps.
BTW, I was looking at a Mitsubishi VRF catalogue the other day and they show VRF units that run both air handlers and water units that provide heat and domestic potable hot water. These are very common in commercial spaces. They are expensive but if they come down a little in price I could see a single VRF system as a solution for residential heat and potable water .
An AWHP must output water at least 20° warmer than the air from AAHP that 20° of needless work cuts the COP by .5 or so making AWHP at least 10% less efficient. That extra operating cost plus the cost of a cooling\ dehumidifying system means AWHP will always be a luxury niche product costing 2X or more to install and 10% more to operate.
Other luxury like stone counter tops have gone main stream inexpiably but I have a hard time imaging folks showing off their AWHP and people being impressed.
Walta
Chiltrix claims a seasonal EER of 22.2 and a COP of 3.93 for their AWHP. I know that the measurement methodology is different than for AAHP so I'm not sure how it stacks up. But they claim it's more efficient than AAHP and comparable with geothermal.
https://www.chiltrix.com/AWHP-technology/
Take a look at the NEEP index all the COP numbers change with the temp often by more than 50% from +47 to -5°.
Marketing numbers always put a product in its best light.
Walta
As a person in a non-cooling climate (we don't need it) the cooling downside is not a downside.
I don't know if ATWHP's will explode, but I certainly can see their benefit in certain climates. I lived in a hydronically-heated home for two years that burned down during 2020's fire season. I am presently designing and building a home that will be heated with an ATWHP instead of the mod-con natural gas boiler I had before.
I am in Zone 5 dry, and should be the perfect climate for mostly heating days and a few cooling days (and cooling only certain areas of the home). I will make hot water via solar and also generate electricity via PV. I drank the proverbial Kool-Aid into hydronics after reading John "Siggy" Siegenthaler's books and articles.
Siggy stresses that one 3/4" water line can carry the approximate BTU as an 18" duct. Add appropriate insulation to those two tubes and it's quite clear how revolutionary a concept this should prove to designers and installers now that we have the ability to make chilled water in addition to hot. I do think new builds will be much more appropriate canvases for this new "revolution" as the old way of thinking (and insulating) isn't already baked into the home.
IMHO, hydronics are unparalleled for comfort, quietness, and efficiency. Are they more expen$ive to install and operate? If one wants large radiant panels in the floors, ceilings, and/or walls (mine will have all three), then perhaps it is. If one wants low-temp modern radiators controlled via PRVs and ducted or ductless fan coils, then I don't see how it possibly can be. Truly.
Ductwork is expensive, labor intensive, and takes up massive amounts of cubic feet within a building envelope compared to hydronics. And despite installers' best efforts, it often leaks and no one notices. Balancing an air system is challenging, and zoning nigh impossible without much more expense and redundancy. Air takes more energy to heat and cool than water and is more inefficient to move long distances. Dehumidification in many climates is critical, but as the OP pointed out, commercial buildings have been using hot and chilled water to heat and cool buildings for generations in ALL climates. Residential is simply stuck in the stone age, as are many designers and installers.
Many of these folks are starting to embrace ductless mini-split systems. They need to imagine a system that does NOT need refrigerant pipes strung all over the inside and outside of homes and NOT needing head units in every bedroom and living space. The revolution is coming... it's simply a matter of who will embrace it and how long it will take to amass a following. We still use the Imperial systems of measurement here, so the US is just very change-adverse in many ways, it seems. *sigh*
That's my $0.02, folks.
I am very confused. An air to water heat pump still needs refrigerant, doesn't it? I can't see how a system that is comprised of just water is going to pump heat anywhere.
Yes. As I understand it, the condenser pumps hot refrigerant to a heat exchanger that is either attached to and made a part of the condenser unit outside (monoblock) or is a separate unit indoors. This heat exchanger also has a pump system. The heat exchanger causes the hot refrigerant to heat water that is then pumped to the various types of emitters; for example: baseboard, radiant floor or ceiling loops, Euro style radiators or fan activated hydronic units that appear much like floor or wall mounted air to air mini split heads.
What interests me most are the mono block systems. Because the refrigerant handling systems are one unit, there is no need to run, vacuum down and charge linesets so the units can be DIY without the hiring of an HVAC guy (prices for A2A systems are exploding around me -if you can get an HVAC guy)... The down side is that outside water lines must be somehow protected from freezing .
Yes, but it uses a much smaller "charge" (amount) of refrigerant since all the refrigerant is in the compressor package, and there are fewer connectors so less chance for leaks. Regular mini/multi split system run long copper refrigerant lines between system components since the refrigerant moves the heat energy through the system.
I've only seen air to water heat pumps used to retrofit existing hydronic systems though, and I've not seen them used for cooling, only heating, in that case. DC, are you talking about systems that would use only water for main heating/cooling loop instead of refrigernant, with the refrigerant ONLY in ONE part of the system (not between pieces)?
BTW, hydronic heating systems are far more common in the Northeastern US, and in areas with a larger stock of older homes. Hydronic systems predate ducted systems. Hydronic systems have a few big advantages: then tend to be perceived as more comfy (less issues with drafts and feeling chills), and they can be relatively easily adapted to different energy sources for the heat supply. The big downside is that hydronic heating systems typically cannot be easily adapted to also be a cooling system. I even know of homes that have a hydronic system for more comfy/cozy heating in the winter, and a ducted system solely for cooling purposes.
Bill
I think he's proposing units such as these for cooling and some of the heating. They come in sizes as low as 1/4 ton rating, so a few of these in bedrooms is a nice option that minisplits can't control quite as well.
https://www.chiltrix.com/fan-coil-units-pix.html
Yes. They are very much comparable to minisplit heads.
Those are one of the few types sold in the US. To see the potential, go to Alibaba and search for "hydronic fan coil units." Hundreds and hundreds of types, many under $100 per unit:
https://www.alibaba.com/trade/search?fsb=y&IndexArea=product_en&CatId=&SearchText=hydronic+fan+coil+unit
There are three main ways of configuring the heat pump.
Monoblock systems like the Chiltrix have the heat exchanger in the compressor unit outdoors. For those, you have to fill the entire system with antifreeze, which gets expensive, propylene glycol is over $25/gallon in bulk these days. They won't freeze while in use, but if you lose power they can freeze and burst if not protected. Note that anywhere power is unreliable you should be running antifreeze anyway to keep the system from freezing in an outage.
Arctic makes a monoblock system where the heat exchanger is attached to the compressor but on an extension. You mount it against a wall, then cut a hole in the wall so the heat exchanger sits inside the building envelope and is protected from freezing.
There are also systems where the heat exchanger is inside and is connected to the compressor with refrigerant lines like a conventional HVAC. I don't like those because part of the appeal of air-to-water is not messing around with refrigerant.
A while back, I came across the training manual for the Daikin AWHP system:
https://apps.goodmanmfg.com/training/files/54aeac3a49b97TB-A101-Altherma-Install-Commissioning.pdf
If you read through it, you'll see that this will never work here. There no way an installer can do a proper install unless it is all they are doing. Even for someone used to working with complicated systems, this is simply too much.
Unless you are looking for a personal project, maybe, I can't see anything like this ever going mainstream.
I'll admit that I didn't read the entire 553 pages.
But all of the design section is stuff you should be doing for any HVAC system -- needs analysis, room-by-room load calculations. The post-commissioning analysis -- measuring actual performance and comparing against projections -- should ideally be done on any system as well and I doubt is done very often.
I would argue that hydronics much more tolerant of mis-sizing than other technologies. If you oversize a hydronic emitter -- radiator or AHU -- there is no penalty, the buffer tank absorbs any short-cycling. If you miscalculate the load for a particular zone it's no big deal to replace the emitter with a larger one, so long as it's not in-floor radiant. Compare that with trying to fix an undersized duct. The guys who were installing hydronic systems 100 years ago weren't doing computer models. It worked for them because it is so forgiving.
I wish there would be an air to water HP solution that would work with cast iron radiators. I have a 1980s-era natural gas boiler, and we installed a ducted minisplit for cooling upstairs, which is adequate for cooling/dehumidifying the entire space, but our 112-year-old place is leaky as heck and I worry that it will be even less comfortable with a full minisplit solution. We may end up there eventually, though.
Maybe, I think units, design simplicity and cost still have to catch up.
I like the concept of the Monoblock system, so that you can run 1/2" Uponor to the AHU's. As someone mentioned already, you're not dealing with refrigerant lines, it makes life easier.
I'm not a fan of radiant only, I like air circulation, integrated ventilation and filtration. So if they come up with a slim duct ADU, and cost comparable to current AWHP's single unit systems, it would be a no brainer.
Well I shouldn't say no brainer, because jumping into a type of system that no one knows how to install or maintain presents its own problems.
Here's hoping,
Jamie
They exist, they're not in wide distribution in the US right now. They're extremely popular in Europe and Asia. Here's a few:
Horizontal concealed fan unit:
http://www.ice-air.com/product/horizontal-concealed-fan-coil-unit/
Vertical concealed fan unit:
http://www.ice-air.com/product/vertical-concealed-fcu/
In-ceiling cassette:
https://www.carrier.com/commercial/en/us/products/airside/fan-coils/42wkn/
Ductless:
https://www.siglersocalengineering.com/ductless.html
In post #18 I put a link to Alibaba where you can really see the breadth of variety available.
The Hydrosolar site mentioned in comment #3 seems to have a lot of good info on it.
https://hydrosolar.ca/
They are located near Montreal, Quebec, so you know they enjoy the cold. :)
They have an article on something called "Enhanced Vapor Injection" which supposedly helps air to water heat pumps perform better in cold climates.
https://hydrosolar.ca/blogs/advanced-technical-zone/enhanced-vapor-injection-for-air-to-water-heat-pumps
There's also mention of a hybrid solar panel that combines solar PV with solar thermal water heating that looks interesting (never heard of that).
https://hydrosolar.ca/collections/hybrid-pvt-photovoltaic-and-thermal-solar-panels
https://hydrosolar.ca/blogs/advanced-technical-zone/vacuum-tube-and-hybrid-pvt-panels-comparison-for-north-america1
I've been thinking about what exactly I would like and what these systems should be. I think manufacturers try to make something that simply does too much.
I just want a box that connects up to a standard hyper heat mini split outdoor unit. This box has 1 digital in from a thermostat, an outdoor temperature sensor and and a dial for selecting outdoor rest curve. That is it. No configuration, no software, no system design outside of getting it about right sized. Simple box that just makes hot water.
I'm also pretty sure this will never happen.
EDIT:
I will make one concession. A 2nd digital input to reverse the unit for cooling.
Akos,
That would be my ideal system too. That these units are presently more popular outside North America may work in our favour?
The Chiltrix does that one better -- zero inputs. You set the water temperature*. It adjusts the compressor speed and the circulator speed to try and maintain that water temperature. If it is able to do that, it does. If the needed output is less than about a quarter of the rated output, it can't modulate down that low. In that case it cycles -- it runs until the water temperature is 3.6F above the set point, then shuts down until the water temperature is 3.6F below setpoint. With a 17 gallon buffer tank and no load that's about 12 minutes of runtime for the 2 ton unit, if there is a load then the runtime is going to be longer.
If you have conventional emitters -- radiators or radiant panels -- you control each zone with a zone valve and a thermostat. If you have an air handler the fan turns on when the thermostat tells it to. If you want a zone you can put one on but you don't need one.
*(If you would like to adjust the water temperature depending on outside temperature you can program in a reset curve).
Edited in response to your edit: yes, there is an input for heat/cool/off.
The Taco System M appears to be a good solution for a retro fit to replace a boiler. In time people will realize that bring refrigerant into their house is not a good thing. Most people are unaware of the comfort factor of hydronics until they live in a house at -20F below zero and don’t have to wear a sweater and slippers. You can’t argue with the physics and math.
Once you start to zone heating and cooling, hydronics makes things much simpler. But most people are used to being un comfortable.
Curious: Why do you say "bring refrigerant into their house is not a good thing?
Most refrigerants require some EPA certification to install. While non toxic, they still need special handling. Would rather install PEX on a monobloc — keeping the refrigerant outside and pump water through the house rather than non toxic chemicals moving through a brazed line that requires specialized skills. Would rather deal with polypropylene glycol and water.
Also, a monobloc uses a lot less refrigerant, so it's cheaper and if there is a leak the impact isn't as great. You're less likely to have a leak or even a miscalculation by the installer resulting in an improper refrigerant level which kills efficiency.
The Taco system M outputs under 20kbtu at 17 degrees, that’s not good enough for the price. Hydronics is great but it’s niche! Not many Americans live in -20 degree areas and amongst those, not many want hydronics.
Well this last 24 hours has been a perfect laboratory for me to test our hydronic baseboard heating system versus our mini split systems. As I menitoned above, most of our heat comes from Fujitsu Low Temp single head units:(ASUG12LZAH1) in the basement and main rooms. However, when it gets very cold, oil boiler/hydronic baseboard heat kicks on in the bedrooms and baths. Even as the temperature dipped well below -0-F last night the Fujitsu's worked great, keeping the house plenty warm for my taste. However, my wife insisted on cranking up the hydronic baseboard heat in the bedrooms/baths and I had to admit to her that the baseboard heat did feel a bit more comfortable: more even and closer to the floor. For me, I would be happy to rely solely on the mini splits, but folks like my wife (a big part of the population I suspect) would certainly opt for hydronic heat on cold days.
I think the advantage of a AWHP would be the potential for a drop in replacement for a hot water boiler. It is a bit surprising someone hasn't come up with an ardiuno hack to turn a minisplit into a air to water unit.
It's a little bit more than an Arduino hack but people have posted a couple of DIY hacks here. My recommendation today though would be to order a unit off of Alibaba, they're cheap.
The SanCo2 is essentially that, except for hot water.
https://www.eco2waterheater.com/product-info
"The SANCO₂ heat pump water heater system consists of two parts. The Heat Pump Unit, where the hot water is produced, using the CO2 refrigerant to extract heat from the ambient air, and the insulated Storage Tank, available in 43, 83, 119, 200, 285, 360, 455 and 505 Gallon Sizes, where the hot water is stored."
Heat transfer is regular copper pipe: no refrigerant work required to install (a big advantage).
The SANCO2 is limited to 8KBTU/hr in heating applications, which kind of takes the bloom off of it. See: https://www.smallplanetsupply.com/sanco2-for-combi-systems
There's also some other strange restrictions: "It is important with this set up to note that efficiency is required that domestic hot water usage be greater than 25 gallons per day." They also say, "Combi systems should ideally be designed and used in climates with a design winter temperature 22F, which will ensure you don’t see a reduction in radiant floor heat output on the days when it matters most. "
I'd like to learn more about those restrictions to find out what they really mean.
On the plus side they say it produces hot water down to -25C (-12F).
DC,
I reviewed the Sanco system technical literature after suggesting it might be a possible alternative system for a poster asking about going from tankless gas to tankless electric for space heating. I think Akos pointed out the limitations to me. I am unclear on rules about attaching other peoples literature, so if curious, just go to the Sanco site and download the tech literature.
The system literature says it can provide hot water down to -20F, (Page 4, shows operation down to -20F or -25F) and output water temps of 145F+. On Page 8, it does indeed show the limiting factors for heating applications. Specifically, the system can only be used for heating if DHW usage is a minimum of 25 gal. per day AND the minimum design ambient temp is 27F or higher. Maximum heating capacity used must be under 8,000 btu/hr. Maybe a home system limitation, the multi residence systems seem much bigger. Perhaps not cost effective at that level.
When used for domestic hot water, the system's in-feed water will fall around 45-55F every time it cycles, so the potential delta T becomes about 100F. If the throughput of the pump is maxed at 20.5 gal/hr, then the 171 lbs of water x 100 = 17,100 btu/hr. This is a bit higher than the stated nominal capacity of 15,400 btu/hr, so either I missed something or they are being conservative. In any case, the system primarily uses water in an open loop manner. The higher set point for the tank ensures very good recovery rates as a mixing valve essentially extends the capacity.
I believe that the issue with use for heating could be the high return water temperatures present if used in a directly connected heating loop. (This may not be allowed.) A closed loop means warm infeed water monkeys with the rise potential. If the infeed water temp is 100F, available rise may drop to 50F. At a 20.5 gal/hr flow rate and a 50F delta T that is 8550 btu/hr. (171 lbs water x 50) Why the added restriction of a 27F ambient limit is unclear to me.
Placing a sealed heat exchange loop in the middle part of the stratified tank might work, but a falling temperture would trigger the sensor activating the pump, which in a static tank, means the mixing of the layers in the tank would limit btu input due to the same falling delta T issue. The minimum usage of 25 gal per day may be required to keep the tank in some degree of stratification.
Maybe a 500 gallon tank or some intermediate tank holding return water from a floor loop could increase the available heating capacity for floor heat. Sorting through the physics of that is above my pay grade, so I will defer to the engineers here.
What you'll sometimes see in wood boiler systems is a hot-tank/cold-tank setup. The cold tank has room temperature water, when the boiler is running water is pumped out of the cold tank, heated, and stored in the hot tank. When heating is needed water is taken from the hot tank, run through the radiators until it's room temperature, and returned to the cold tank. Both tanks are open to the atmosphere so their fill level can vary without creating big pressure swings. This way you always get a big delta T across the heating source.
If maintaining delta T is important you could do this with the Sanco2. It is a fair bit of trouble, but the CO2 technology is interesting. If it could really produce 17K BTU/hr that begins to put it on the edge of viability for residential heat. If it can produce water at 145F that puts it at the edge of viability for boiler replacement in radiator systems.