Air-to-water heat pumps in cold weather

stanwagon

I live at 9500 feet in Colorado, and temps often dip below 0 F. My house is well-insulated (2008 build) and heated with in-floor fluid via Triangle Tube gas boiler. I have two Viessman thermal panels that do a good job for hot water when panels are not snow covered.

My contractor seems not certain whether an air-to-water heat pump would work to heat the house. A local conservation organization tells me that heat pumps can definitely work well up here, except perhaps for a few hours a year on the coldest day (it was -22 last week) when an electric backup will kick in. But it is possible, isn't it, that there is a difference between air-to-air heat pumps (which I imagine supply 70-80 degree air to the house) and air-to-water, which must heat the glycol fluid to about 120. And if it is used for hot water as well, then 130 would be needed.

Quote from conservation guy:

"Most cold climate heat pumps are rated to work down to -14 F. Some folks may find they would like supplemental heat for when temperatures get that low – but ideally that supplemental heat only turns on a few hours/year. However, cold climate heat pumps are not yet widespread here and there’s hesitancy among HVAC contractors to carry and install them…due to past models not working in cold climates. But I assure you, heat pumps are successfully being used in Minnesota which is also climate zone 7!"

Being a mathematician I enjoyed learning about the "ton" unit: melts a ton of ice in a day. Probably a 5-ton unit would work if it works at all. I called Bosch but they would not talk to me. I am also a little confused on this point: 5 tons is 60000 BTU/hour. My boiler is rated at about 134000 BTU/hour. But I have a feeling that a direct comparison like this is not appropriate.

There is never air conditioning used here, so I realize it is not an ideal heat-pump application. But I do want to try to eliminate the burning of natural gas if it can be reasonably done. Is it reasonable to use a heat pump in this situation?

Comments on any of this are welcome. Thanks in advance.

fentonc

Air-to-water heat pumps just seem to be extremely rare in the USA, which is probably a practical issue if you need to have something installed, serviced, etc, and you're not an ultra-DIY type. Here is an example: https://www.chiltrix.com/air-to-water/. In regards to your boiler, it's probably oversized (most seem to be) - mine is 140K BTU/hr input, but has never run more than ~25% duty cycle over the course of a day.

Hot_water_fan

It is reasonable and would work, but Air-to-water heat pumps are rare currently in the US. You can easily incorporate backup (electric, gas, etc.) which can greatly limit your gas usage.

A few notes:
1. You need a heat loss, here's one method: https://www.greenbuildingadvisor.com/article/out-with-the-old-in-with-the-new.
2. you most likely don't need 120 degree water in a new house for in-floor heating, so that's not a constraint. Probably 100 degrees will work just fine.
3. That boiler is most likely extremely, embarrassingly oversized, so you don't need a one-to-one capacity replacement.
4. Caleffi has an Idronics series about this topic.

Jamie Hall

Hmm... just tried a couple of methods for estimating heat loss. A Manual J (Slant/Fin) and a fuel usage based calculator. And the real world. Manual J and the real world came out pretty close. Fuel based came out about 50% undersized.

hot_rod

Here is some good non biased info on A2WHP, including actual experience at below  0 conditions 

Know also that 5 ton rating will drop as the outdoor temperature does, see the charts in this Idronics

The key to good performance is running them at 120F or lower. Pretty much radiant, or large sized panel rads

https://www.caleffi.com/sites/default/files/file/idronics_27_na.pdf

stanwagon

Many thanks, all. I will examine the links and info. Yes, I believe our boiler does generally heat the liquid to less than 120 for the floor. But the hope is to have this work for the hot water too (when the solar thermal is not taking care of it), and that would require perhaps 130. Thanks again.

Hot_water_fan

Hmm... just tried a couple of methods for estimating heat loss. A Manual J (Slant/Fin) and a fuel usage based calculator. And the real world. Manual J and the real world came out pretty close. Fuel based came out about 50% undersized.

Can you elaborate
@Jamie Hall? Fuel usage is real world to me (ie I’ve seen exactly how much fuel is used in the coldest day of the past few years and it matched up with the prediction). How do you define it? 

Jamie Hall

I may have done the math wrong. Wouldn't be the first time. I used the mean fuel usage -- 0.483 gallons per day per degree day, based on some 20 years of data. The standard deviation on that figure is huge -- 0.14 gallons per day per degree day, which somewhat undermines my confidence. If I were to use the absolute maximuum figure for those 20 years (a particularly nasty week in February of 2016) -- 1 gallon per day per degree day -- the result appears to be closer to what I regard as reality -- which is Cedric being able to hold the temperature with an 80% duty cycle (running about 90 minutes on, shut off for about 12 minutes, etc. (design delta T is 70 F) (I'll freely grant that neither Cedric, not the radiation which Cedric powers, is normal -- Cedric is not oversized for the radiation (he rarely gets up to his 6 ounce cutoff, even after an hour of running) nor is the radiation oversized for the structure; it can keep up at design, but only just barely).

The biggest single problem that I see is that the heat loss of Cedric's home is very heavily influenced by wind. Most days, of course, this isn't that much of a problem -- but it is not uncommon for the coldest days to also have the highest wind loads, but they seldom last for more than a few hours -- but the building has to be heated during those few hours, but that extra load (at least a third again the heat loss) won't be reflected in an average over a day, never mind a week.

I think that the average fuel burn approach probably does work reasonably well for average houses in average settings, and I think it is a useful giggle check for any structure -- since it is so quick. But honestly I'd not use it to select a boiler or furnace, particularly since the various on-line or downloadable Manual J applications are also reasonably quick -- although they do require one to know something about the structure.

Hot_water_fan

@Jamie Hall Share the data! Sounds extremely interesting. I’ve included wind before, it makes a surprisingly small difference for an old stone house with single panes and no insulation. 

mrhemi

"Quote from conservation guy:"
I don't know what the 'conservation folks' are like in your area, but where I am from, they remind me of a quote from Ronald Reagan (I believe), "We're from the government. We're here to help you"
What could possibly go wrong?

Hilex007

Recently I installed HydroSolar 5 Ton Air to Water heat pump for a client. Old house without insulation. I used existing cast iron radiators for heating. I installed 2 spacepak fan coils for cooling and extra heat. Old system operated @ 180 F. New system runs @ 125 F. With 3 way valve and brazed plate heat exchanger heat pump provides domestic hot water ( 80 gal @ 126 F ). It all comes to heat loss and gain. Calculate domestic hot water demand and that is it.

Jamie Hall

Sometimes it would work. Sometimes... not so much. Do your due diligence.

zac_2

Really look at the cold weather efficiencies carefully, not all units are the same. I installed an Enertech Advantage recently and it was getting a COP of 2.2 @ -24C with no backup. Chiltrix is awful at low temps, their capacity and efficiency fall right off when it gets below -10C or so. Some other units to look at with good cold weather specs are: Stiebel Eltron WPL, Gree Versati IV and Taco System M. All these units can do 100% domestic too. As always, high COPs with heat pumps are only unlocked when system temp is low so always <6" radiant tube spacing and short loop lengths (I always keep 1/2" loops below 250ft) to facilitate high flow rates and tight delta T.

hot_rod

Keep the triangle tube for back up if it still works, a hybrid system

looks like maybe 6 more brands showing a2whp
Here at AHR this year