I've been looking into geothermal heat pumps lately, self-educating. Came here and saw some of Dana1's posts--I had no idea that there were air-source heat pumps which could work reasonably effectively at such low temperatures. So I'd like to ask if there are any particular air products I should look at for my needs: - I need hot water out for radiant floor heat, ultimately. - Right now, to start, high-temp (140-150F) hot water out would be really helpful, since the radiant floor will have to come later as a separate project, and the baseboard units in this house are under-provisioned. - 3-ton would probably be a good size, based on historical heat usage. I know a lot more, but that summary should be good, without getting into the entire history of a very unusual house, with very unusual mechanicals ;-)

You packed a lot into a few paragraphs.

there are lots of products that are "reasonably effective" in lots of parameters. Most often I explain to my clients that virtually everything is a trade and we try to make budget and wish list come together.

Depending on where you are air to water heat pumps may cost as much or more as water to water. I'm not familiar with any high temp air to water systems. Better to design for lower temps anyway. You can look at radiant panels or high out put baseboards etc. Also incorporate outdoor reset so you only use highest temps when absolutley necessary.

Will take your word for the usage, but know that 3 ton heat pumps do not necessarily provide 36MBH.

Well, I guess the #1 questions is: air to water that still functions at low ambient temps? As for the tonnage, it's based both on historical BTU usage, and some fudge factor for low-temp performance drop-off, and the fact that I can probably get up to about 60,000BTU/h out of my wood-burning fireplace when needed.

The air-to-water options in the US are very limited.

The Daikin Altherma has a fully characterized output down to -20C/-4F, but if you're looking at 140-150F water temp requirements forget about it- both the efficiency & capacity will be pathetic. If you can fly with 90-100F water for space heating and your 99% outside design temps are above 0F it can do pretty well, but it's nowhere cheap as mini-splits.

There are CO2 refrigerant heat pumps that can deliver the higher temps with reasonable efficiency currently sold into Asian & Australian markets, some of which work well into negative single-digits F. The Sanden EcoCute (domestic hot water heater) and it's bigger combi cousin should become available in the US sometime soon, currently still working on US style safety certification issues. I know of a few of EcoCute versions being turned into combi-heaters for high-R houses in the Pacific Northwest (in locations with outside design temps in the 20s F), but the output capacity isn't up to snuff for a standard US code min type house.

Without knowing your heat loads or design temps it's tough to make reasonable recommendations. If 140F water is going to be necessary it's tough to get any efficiency or capacity out of three tons of geothermal too. Deciding on the heating system first, then figuring out the heat loads later is just the wrong order. If you have an existing heating system an a fuel use history on it it's possible to put a pretty good stake in the ground marking the heat load of the whole house, but that's just one small part of a heating system design.

Also note, geothermal pricing in many other parts of the US are nowhere near the stratospheric price points they seem to be in New England. Local markets matter (a lot!)

First off, thanks, you've given me exactly the information I needed.

Deciding on the heating system first, then figuring out the heat loads later is just the wrong order.

Here's the reasoning, short version: long-term I want radiant floors, but there's a lot of construction between now and that goal, and a lot of pieces to the whole puzzle. If I can start saving on our very high heating costs now, rather than 5 years from now, that would be nice. Until I saw some of your other posts I had no idea there were air-to-air heat exchangers that were functional at double-digit negative temps--I was still stuck in the mindset that they're less efficient below 40 and essentially useless below 30ish. I looked into the Fujitsu's, and, long story short, I am convinced that if I install them right now, I would get a 2-year payback with mild winters like what we just had, closer to 1 year in a cold winter. I could put them in, throw them out in 5 years, and be way ahead. But I'm certainly not going to do that without considering all the alternatives. (And I wouldn't actually throw them out after a few years--there's always craigslist, or keep them for A/C for the 2 weeks a year we could use it, if they can be actually be left off 11 months at a time.)

Functioning to -15F is not even 100% necessary. What that means, is that there is no time in the past 7 winters (how long I've been here), that pipes would have frozen had we not been home and this (but probably bigger than 3 ton) had been the only source of heat. I understand, roughly, the performance curve of these things, and I have no intention of depending on mini-splits keeping us comfortable when it's brutally cold, I'm fine with them on the rare coldest few days functioning at low capacity and efficiency, even not functioning for 12 hours at a time a couple of days every few years. We have the fireplace, and for now we still have the kW-guzzling electric boiler that could simply be set to a fail-safe lower temp.

If you're wondering, my shoulder season where I only need a little heat, and can't practically use the fireplace because it puts out too much heat, is about April through June and September through November in a normal year. In a mild year I might add March & December. So many many days with highs in 30s and 40s, lows in the 20s and 30s. Then, every once in a great while, the jet stream from Canada shifts over a bit east of the Continental Divide, and brrr... So just having them heat for the shoulder season, they would pay for themselves. I imagine that's unusual, but our micro-climate is very unusual. 3,600 feet above Denver, and whenever the news is all about extreme low temps in Denver, we'll be 10-15 *warmer*, but of course always 10-15 or more cooler in July & August (FYI, high today of 40).

And, BTW, I'm not going to do anything without at least talking to a dealer or two about geothermal. (We're on granite, good thermal characteristics to begin with, and you can't get down 200ft without hitting water flowing through fissures, at 600ft you'd probably see at least 15-20gpm flowing through. I've seen the horror stories of people where incompetent installers put in undersized loops. I suspect I'd have a hard time not winding up paying for an oversized loop.)

With W-W and radiant, thermal inertia becomes an issue, especially with a small 3 ton unit. Keep in mind that you need extra capacity to heat up the water, to heat up the buffer tank to heat up the radiators....then you start putting BTUs in the space.

At the end of the day you need an extra capacity to have reasonable responsiveness.

Excellent points, thank you--I was overlooking the buffer tank in particular. To be clear, I know that 2-ton is not enough and that 6-ton is too much. I'm not set on 3 ton for the final radiant system, I'm just using that as a first guess when comparing options initially, and bearing in mind that it might well need to be 4 or 5.

There are new 150F degree w-w heat pumps out there, which do not mind running at 90-100F at a later point as well.....

Right, first I found the new WaterFurnace that puts out 150, then Dana1's post about the air-to-air that works to -15F, which is what got me wondering about air-to-water that works in low ambient and/or puts out 150. But it seems there is no such thing, at least for now in the U.S. so my choices are ground loop water-to-water or air-to-air. I need to talk to contractors re cost, and do the hard work of more carefully characterizing my heating loads. Thanks to all, all comments have been useful. (Or I could go with the new Froling that burns wood chips--around here I could make it a summer job and get PAID to collect my fuel... But I think not. Too much work.)

I know a guy in Quebec heating with Fujistu RLS2H series mini-splits who has sailed through -25F and lower temps without any issues.  Even the 3/4 ton -9RLS2H or the newer more efficient -9RLS3H will deliver over 11,000 BTU/hr into a 70F room when it's -15F outside.  See the capacity tables beginning on page 15 (PDF pagination.)

An electric boiler slaved to a floor thermostat to keep the slab from being too cold and setting the room temps with mini-splits is a pretty cheap, comfortable and low-risk solution compared to an air-to-water or geothermal heat pump.  It cuts into the net efficiency, but that has to be weighed against the lifecycle cost of the alternative air-to-water or geo solutions, and maybe even against the lifecycle cost of PV solar to cover the difference in power use.  With an electric boiler + mini-split solution you basically choose the average efficiency by how much above room temp you set the slab. Even if they're set to the same temp it's more comfortable than an unheated slab, but the power used for keeping the slab at temp is miniscule.

Posted By Dana1 on 22 May 2015 04:11 PM
I know a guy in Quebec heating with Fujistu RLS2H series mini-splits who has sailed through -25F and lower temps without any issues.  Even the 3/4 ton -9RLS2H or the newer more efficient -9RLS3H will deliver over 11,000 BTU/hr into a 70F room when it's -15F outside.  See the capacity tables beginning on page 15 (PDF pagination.)

An electric boiler slaved to a floor thermostat to keep the slab from being too cold and setting the room temps with mini-splits is a pretty cheap, comfortable and low-risk solution compared to an air-to-water or geothermal heat pump.  It cuts into the net efficiency, but that has to be weighed against the lifecycle cost of the alternative air-to-water or geo solutions, and maybe even against the lifecycle cost of PV solar to cover the difference in power use.  With an electric boiler + mini-split solution you basically choose the average efficiency by how much above room temp you set the slab. Even if they're set to the same temp it's more comfortable than an unheated slab, but the power used for keeping the slab at temp is miniscule.

It might do so when it does not defrost. Hard to imagine that slab heat via an electric boiler is not beating the purpose here. May be you have a different definition of miniscule. Giving the exposure to outside conditions, plus bedroom and second floor heating issues, I am not sure why why you consider the mini splits a low risk alternative to geo. You