I have gotten mixed reviews of Geothermal systems. I am in Western NY so we are primarily a heating climate. We are building a new house and don't really intend to have AC at all. We likely will need to go Off Grid with our electricity too. Would a Geothermal system be worth the cost? I have heard that mini-splits can be more efficient. I have heard that the systems struggle to provide enough heat in my cold climate. I would love to have radiant floor heating, which I have heard is even more costly. Would like some guidance here, thank you!

Going off grid with a Geo system could work but the battery bank would have to be enormous. Heat pumps operate best when they run slow and steady.

Being in western NY, least amount of sun when you need the most for heating.

Off grid sounds great and all but it's not easy or cheap. When the solar panels can't re-charge the batteries you have to use a generator. Run the batteries down too far too many times and they don't last long.

Off grid and radiant... get a big propane tank, hot water for DHW, warm water for radiant, fuel for the generator. Try to have as little draw on the electric loads as possible so as to have the smallest battery bank.

I agree - use propane or a wood fired boiler for heating with off-grid power. Maybe some co-generation (heat from a propane fueled generator).

What about the mini-split v. geothermal argument. Which is better?

Geothermal will cost more initially but be more efficient (in btu per kwh).

If you really want to run a heat pump from off-grid power, then I'd look at designs that run only when the sun shines and store the heat.

What about if we end up on-grid? Still worth the cost? Seems like it would take a while to recoup 15k for a water furnace system vrs a 2-3k boiler

Bueller??

Minisplit is clearly cheaper to install, but has only one distribution point. They loose efficiency and capacity at very low outside temps, but mainly their COP goes down to a COP of 1 at the coldest days, which is not the case with geo.

So their winter peak power demand is much higher than geo heat pumps, not helping you with the off the grid plan. If you go off the grid, choose a 7 series variable speed to ensure that you keep your startup amperage low.

At the current stage of storage technology, your best shot would be a a geosystem connected to the grid, especially if you do not have access to natural gas.

Do you plan an open space floor plan? Tell us more about your house and your exact location. Are you at elevation?

No issue with heat capacity with geo in cold climate,

First, if you're going off grid there is no way in hell ANY heating system that uses electricity (even for pumping hot water) makes any sense at all, when you consider the size of the PV array and the size & enormous cost of the batteries necessary to get you through a snowy or cloudy winter week. Seriously- forget about it, forget propane boilers, forget heat pumps of any type. Stick with wood-burners, or find a propane wall furnace that doesn't need power (there are a few models out there suitable for off-grid living.)

Or, stay on the grid, and go with heat pumps.

The COP of cold climate mini-splits running at full speed is still over 1.5 at -10F outside temps (about 1.8 in bench tests). It has to be heluva lot colder than that before it drops to 1.

In a previous thread ( http://greenbuildingtalk.com/Forums/tabid/53/aff/12/aft/84081/afv/topic/afpgj/3/Default.aspx ) the order of magnitude of the heat load of this 1600' house with a 1600' ICF basement is likely to come in around 35,000 BTU/hr, considerably less if it can be made more air tight than most log homes.

This can be covered with 2.5-3 tons of ductless (maybe even 3 or 4 heads, if you like) for roughly $10-12K, total installed cost. The seasonal average COP in that location (Livonia, NY) would run between 2.5-3.0, compared to ~4.0 for a better-than-average geo installation, maybe even 4.5 for a perfectly designed perfectly implemented low temp hydronic GSHP system.

At NY type electricity pricing there may be geothermal solutions that still make financial sense compared to mini-splits on a lifecycle basis, but to do that it has to be considerably less expensive than it is in my neighborhood if it's going to beat mini-splits plus enough PV solar to cover the difference in annual power use.

Fixed broken link that Dana referenced and created URL:

Need a cheap to install radiant flooring system

I do have a few customers on the grid with GSHPs, but indeed their battery is rather large, and they really wanted to be off grid, and costs were secondary. If you are cost conscious stay on the grid.

If you go with heat pumps, the most important thing to understand is that the operating temps of the floor have a direct impact on the efficiency of the heat pump systems. The lower the better, which really should only leave you with top of the floor aluminum systems (preferably protruded aluminum) and/or embedded in concrete/gypsum. That way you can get the operating temp down to around 90F which will make your radiant very efficient.

If you are looking for cheap radiant, which requires higher supply temps, get a boiler.

In terms of mini split efficiency, usually multihead are significantly less efficient.

Dana, could you refer me to the testing or report where cold climate heat pumps run better than a COP of 1.5 at -10F in the field. I could not find anything.Or even some bench testing would help.

See Figure 9 on p14 (p22 in PDF pagination) for a bench test.  That's the older Mitsubishi FE.  The newer Fujitsu RLS3 and Mitsubishi FH series cold climate mini-splits are comparable, and may be slightly better than that, but probably not nosing into the 2s by much, if at all.

Bench testing tends to be less error prone in the measurements than in-situ field testing due to more/better instrumentation and better control over the environment variables.

Multi-split efficiency is hampered by the high minimum-modulated output of the compressors relative to the min/max output of the heads.  A 3-ton compressor will typically only modulate down to 6000-7000 BTU/hr @ +47F output temperature, which is roughly the maximum output of a half-ton ductless head.  So when it's warm enough out that the heads are cycling rather than modulating, when only a single half-ton head is calling for refrigerant the blower on that head has to run at it's maximum speed at much lower efficiency in order to move the amount of heat it's being served, or else it begins to short-cycle the compressor. Either way it takes a huge hit, at temperatures where it would be running a COP of 4 or better if the compressor had enough load to be modulating, with the heads still running at low speed.  Some single-head mini-splits have fairly high min-modulation that gets in the way of efficiency too, particularly when oversized for the loads (the formerly very popular FE18NA is one of those).   The Mitsubishi FH09NA is a real sweetheart for cold-climate low-load zones, since it can dial back to 1600 BTU/hr @ 47F, yet still delivers over 10K @ 0F.

In just the past week or so I ran a power use analysis against heating degree-day data on house heated with pair of 2 ton Fujitsu multi-splits (the all too common "ductless head for every room" mistake).  Minimum-speed output of about 6K per, for over 12K total at min compressor speed, and looking at the Manual-J the compressors were being forced into cycling whenever temps exceed ~35F.  It's pretty clear from the kwh/HDD ratios that efficiency during the late winter and early spring (March-April) was quite a bit higher than the much more temperate weather of late April into May.  With the hit from cycling the estimated as-used seasonal average COP looks to be about 2.6-2.7 The submittal HSPF for that unit  is 9.5 (= COP of 2.8), whereas if more optimally sized and a lower minimum modulation it could have been north of 3, in a location with a 99% outside design temp in the +12F to +13F range.  A couple of dedicated 1.5 ton Fujitsu mini-duct units would have been a much better choice in that house, but it is what it is...

http://www.nrel.gov/docs/fy11osti/52175.pdf

I looked at that report before, and the numbers did not add up.

If you look at figure 9 you cite (and figure 10), the COP at high fan and high comp power is similar 1.8 (at -10F) versus 1.9 (at +3F), which does not make sense. Also the COP at 22F, 32F and 42F is exactly the same (around 2.3). Does not make sense at all that you do not have a much higher heat extraction out of the air at higher temperatures, not shifting to a higher COP, especially when the capacity is increasing significantly according to figure 8.

Now if you look at figure 8, the capacity line seems correct. The capacity drops from 10,000 BTUs at +3F to 7,000 BTUs at -10F. Since COP is defined as energy delivered divided by energy used, this would mean that the compressor must have drawn much less energy at colder outside air temperature, namely (7,000/3.412)/1.8 =1,140 watts at -10F versus (10,000/3.412)/ 1.9 = 1,542 watts at +3F. The increase in compressor energy use is simply too much at warmer temperatures. One would expect the 15%, but not 35%.

The data is very fishy, to the point where I call it not trustworthy.

The last time I put a power meter on a FE12NA at -10F it was drawing 1,942 watts very consistently, at 70F inlet temp. I could not measure output, but if the output number of 7,000 watts is correct, you have a COP of 1.06.

Take the Mitsubishi reported data in figure 10 and continue the COP line to -10F, and the COP would also be a 1.0
COP is pretty much a straight line (linear function) in relation to outside temperature, the data points in this study you are citing are all over the place.

Maybe the first question should be what utilities are available to you, and at what cost? Also, what size space are we talking about? If you're stuck using propane for heating, then geo absolutely could make sense, on or off grid. If you have natural gas available to you, geothermal is likely not cost effective.

IMO, one could do well off-grid in most weather with PV solar, Chiltrix heat pumps, minimal batteries, low temp radiators and a large water storage tank. But sometimes it would be necessary to run a propane generator.