I'm designing my "dream house" which I hope to build in northern Michigan next year.  I want net zero with roof top PV.
It's about 2400 sq ft main floor with a full basement.  I plan r60 ceiling, r45walls, r25 basement and r10 under slab and good triple pane windows with lots of attention to air sealing.  I'll use a ducted ERV for ventilation ( I've chosen a way oversize "commercial" unit because of it's very low energy use and to better handle adding a heat exchanger).  My design heat loss is 12kBTU/h at -10f.  I'm undecided on how to heat it.  One plan is use a minisplit.  Another is an air to water heat pump.  Another is 'geothermal' water to water heat pump.  The air to water or water to water heat pump options could also handle DHW & cooling.  Choosing other than the air to air minisplit allows radiant floors which adds to efficiency and comfort.  The efficiency gain results in fewer solar panels but no where near enough to justify the cost of going to radiant floors.  With low total heat loss I see no need for multiple zones other than one for the main floor and one for the basement, which could as well remain unheated.   I think it is without question, the lowest cost way to achieve net zero is a low temperature minisplit with resistance baseboards as back-up and to provide some supplementary heat during very cold spells.  The absolute lowest energy use, and by far the most expensive, would be a water to water (geothermal heat pump) and radiant floors.  Slightly higher energy use hence slightly larger PV array, but far less costly, would be an air to water heat pump (with resistance heat back up) and radiant floors. A less costly but similar performance system would use the same heat pump,  eliminate the radiant floors heating with the "cooling" water to air heat exchanger in the ERV return air.  I don't like the idea of adding "fan coils" to water output heat pumps which is the usual way of getting cooling from them so I've "invented the heat exchanger in the over sized ERV method.   However even if I choose the minisplit I think it is prudent to "plumb" the slab of the basement with PEX as it's very expensive to add later.  I'm not sure but suspect installing the Pex under the sub floor with aluminum spreaders is the easiest and lowest cost main floor radiant option, this method seems practical even as a "retrofit".    Help sell me on radiant floors.  My floor "system" is "trimjoist" 24" OC & all the under subfloor details I've seen have 16" OC framing.   I don't see this as a problem with 8" tube spacing, I'll just have 3 tubes between trusses.  I would not use conventional manifolds with either  floor but instead "header" pipes to connect the loops in parallel.
For the basement slab I plan 12" nominal loop spacing.   With any heat pump hydronic heating system I believe a rather large 'buffer' tank is needed.  The buffer tank  needs to store the full output of the heat pump at the highest heating temperature for the minimum " run" time of the heat pump with a 'reasonable' temperature rise (I choose 5 deg F and 6 minutes).  What am I missing?

One of the problems with radiant slabs in a home with very low heat loss is that your slabs become very low temperature emitters and lose that warm to the toes feeling. Have you run some engineering to see what happens with 8" tubing and low heating requirements? That may be way overkill with your envelope.

One way I would consider going there is to use just a minisplit and put resistance radiant in a few key locations, like the bathroom floors. The money saved over full radiant installation would buy a lot of PV, which is more towards your overall goal. Also gets you some cooling in Summer. The resistance floors would help you feel warm when all else was cold and provide some additional BTUs during cold spells. An efficient, sealed fireplace or wood stove might also be a good idea as insurance against unexpected deep cold or electrical outages.

If you do heat the slabs, I'd use more than R10, which is probably code minimum there.

If truly your "dream" house I would go for comfort and long-term cost payoff since hopefully you'll be there a long time (versus net zero only). I determined for my dream house (also in design) to go with the expensive geothermal + radiant floor heat plus a limited air duct system for cooling, erv air exchange and dehumidification. Like you I'm concerned primarily about heating (Minnesota).

Getting 300%-500% efficiency with geo hopefully will pay for itself as well (mini split air-air heat exchanger maybe 200% efficiency plus won't heat your hot water).

Radiant floor heat is fairly "future proof". The water can be heated with geothermal, elec resistance, wood boiler, etc. Not true with minis. Even if you decide to use minis though, I would run PEX like you said where you think you may want it in the future (at least for the basement or garage where you'll have concrete).

The heating comfort of a radiant floor can't be beat with any other system IMHO especially with tall ceilings like I have (10').

I'm going down the ICF route including floors. Having ICF floors is more expensive for sure (say $15 sq ft) but easy to add radiant heat along with the benefit of thermal mass and minimal noise between levels. Also helping to justify ICF floors and radiant heat a bit more is that I have seasonal allergies so I don't want any carpet. Wood floor and\or stamped concrete with strategic rugs is what I want.

Getting an annualized COP of 5 out of geo at northern MI subsoil temps isn't very likely, even though folks in (much warmer subsoil temps) western Europe can sometimes pull it off in very low-temp slabs. I wouldn't assume being able to achieving better 3.5 for average COP going that route- there's some design risk on both output capacity and efficiency, since every system is a custom system.

If it's N.MI, (and not the U.P.) the typical January mean temps in the fingers of the mitt is about 18-19F and a reasonably sized decent efficiency mini-split would deliver average seasonal efficiencies in the mid to high-2s, even though though the 0F efficiency is less than 2. The Mitsubishi MUZ-FE18NA (a popular well-regarded 1.5 ton) can deliver more than 12,000BTU/hr with a bit of margin at your -10F design load, and typically costs less than $4500, installed. It's +17F COP is 2.77 when modulating along at 11.6KBTU/hr out, and your +17F heat load will be lower, getting slightly higher efficiency. You could even be looking at the low-3s for a seasonal average.

Most of the eastern Idaho installations in this ongoing field study where efficiency is measured in-situ have comparable mean-January temps (only ~2F warmer than N. MI), and the group averaged a COP of 2.81, but also have higher average loads relative to the output of the mini-split, which would result in a lower as-used average. See Table 24, p39 (p52 in .pdf pagination):

http://neea.org/docs/reports/ductless-heat-pump-impact-process-evaluation-field-metering-report.pdf?sfvrsn=16


For the difference in cost between that and a lowest-tonnage GSHP with the buffer tanks and radiant etc, you can buy at least 2kw of grid tied PV, maybe even 3kw to offset a substantial fraction of the heating power use, which is something close to half-way to net-zero. Marc Rosenbaum heats with a somewhat more efficient 1-ton Fujitsu in a house with ~R30 whole-wall values at an outside design temp of +10F, and manages to hit Net Zero with a 4.76kw grid tied array: http://blog.energysmiths.com/

There are now getting onto a dozen Net Zero houses in central MA (design temps around 0-5F) being heated with mini-splits. The PV arrays tend to be 6-8kw (not everybody is a miserly as Marc on power use :-) ) but that approach to Net Zero has become a well-trodden path. Assuming you have the roof area for 6-7kw there's no reason you couldn't get there too.

Dana1,
I'll have roof area for a 10kw array of PV. The MUZFE-18NAA & MSZFE-18NAA is the mini split without question. From another thread an all in one heat pump water heater is the best way to get DHW. I've about talked myself into a small masonry heater as "back up" heat" with no need for electricity.

One of the issues with glowing pieces in a home with very low heated loss is that your pieces become very low heated range emitters and reduce that heated to the feet sensation.

I have the Mitsubishi Dana refers to and use the radiant floors for comfort. Sub-floor radiant systems with aluminum plates - something we design and install weekly - usually require design temperatures higher than optimum for a GSHP. Radiant ceiling by contrast will typically operate a few degrees above ambient with better response and easier installation in some cases.

A designer should review the room-by-room heat loads and deliver an design water temperature before you decide the optimum heat source. Domestic hot water should also be designed with some experience especially if the water heater will serve space and DHW needs.

I'm gonna go out on a limb here and guess that your design-condition heat load is more than 2x the output of that mini-split at that temp, and that your whole-wall R is less than half the R45 Lieber has specified, with lower performing windows too.

Liebler is building  to a superinsulated level with less than U0.25 window where the room-to-room temperature differences even at design condition won't be very high most of the time, even WITHOUT optimizing the ventilation-air flow from the warmer space toward the cooler rooms via jump ducts etc.,(but better with a reasonably designed ERV ducting that takes temperature balancing into account.)  In a house like that the required BTU/ft output of radiant floor would be so low that it would be in imperceptible boost in comfort for 95% of the hours (and probably 99% of waking-hours) every heating season.

Sub-floor radiant systems with aluminum plates...usually require design temperatures higher than optimum for a GSHP

Is this true for tile floors? My calculations indicate that aluminum plates outperform concrete in this case.

Liebler - have you ruled out open loop geo? If you only use it for heating, hardness in the water is not much of a concern. I agree - if you have a concrete floor planned, throw some PE-RT tubing in it - for efficiency and flexibility reasons.

I would like to re-open this thread. I am considering GEOTHERMAL with radiant for my new home which is my dream home! It will not be crazy inculated and will not be crazy efficient but I am trying to make the best decision possible for longevity and affordability.
I would like to know more from BadgerBoilerMN about the boiler he uses?
Additionally I would like to know about radiant under hardwood as well as under tile. I am in southern indiana in the Bloomington area so that's the climate.

Like the rest of it, any good heating system design using radiant floors calls for doing heat load calculations up front. The ratio of the BTU per square foot of radiant floor (and the floor covering/type) determines the required water temps, which in turn defines the system efficiency.

Low temp radiant floors can be amazingly efficient with heat pumps, and more comfortable than ducted-air. It's more expensive up front than a ducted air (or ductless air source heat pump). From a comfort point of view, in high-R houses that never need more 4BTU/ft out of the radiant floor, the increase in floor temp can be too subtle to notice or care at the average heat load. Whether a comfort uptick that subtle "worth it" in those situations is highly subjective- it's not the same as radiant floors retrofitted into a 1970s code-min house, where the floor temps will average well above the room temp.

If your whole-house heat load is under 25,000 BTU/hr @ Bloomington's +5F outside design temp you can probably make a Daikin Altherma air-to-water heat pump work, usually at a lower price point and comparable efficiency to geothermal in your climate. The Altherma uses varable-speed mini-split compressor technology for the outdoor unit, and refrigierant-to-water heat exchanger on the interior side. A radiant designer comfortable with fossil-fired boilers should be able to design the radiation piece around these without much trouble.

"If your whole-house heat load is under 25,000 BTU/hr"
this is the ultimate IF to get the benefit of these new high efficiency heat sources, and you need to build your house to a certain level to get there. Without triple glazed windows, without thermally broken, high R wall insulation, the chances of getting to these levels decreases. You're at the point where someone needs to "run the numbers" to determine the load with the features you have decided on like the basement insulation and the roof insulation. Use those calculations to aid your decision making about going further. These are calculations you can do with computer programs that have been discussed here, or you can hire someone to do them Either way, now is the time.

I love my radiant floors and my ICF house but my design temp is -32C

To go radiant AND geo, man put the cash into the envelope and heat it with a candle.

You can put the money into the most efficient heating system you can find or the most efficient envelope. Seldom do you need to do both.

I understand BobI thanks to your help I have finalized the basement foundation plan. Now, I need someone here that can help with the manual J loads. Need to finalize my wall stack up first!!!

You're much better off finalizing it after/while you are working on the heat loss calculations

Wrote you an email! Wall stack up is next step!

Posted By Bob I on 13 Jan 2014 06:39 PM
You're much better off finalizing it after/while you are working on the heat loss calculations

Absolute gospel!

When anticipating high-ticket items like geo and radiant floor, it's good to compare/contrast the cost of the mechanical systems relative to the cost of reductions to the heat load.  Be prepared to pay a premium per "ton" of reduced load, since

A: The insulation does not have an operating cost and

B: The lifecycle of the insulation and other building components are  3-5x longer than the mechanical systems.

When you get down to a design condition load of 5 BTU/foot for a the radiant you might look long and hard at what that radiant costs relative to reducing the heat load to 4BTU/foot.  For concrete slab heating the radiant is still going to be pretty cheap, but for other types of radiant floor there is still significant cost, but at heat loads that low you need to figure out if the upcharge is really going to be worth it.

The downloadable freebie tool BeOpt is pretty flexible, and designed specifically for playing the cost/benefits of different approaches against each other on an energy-savings basis (and beats a Manual-J or I=B=R approach for heat load estimates.)  The underlying energy modeling features embedded in BeOpt is DOE2, which is also a downloadable freebie useful in it's own right, but doesn't have the cost comparison features- it's strictly an energy use estimator.  Sometimes the triple-panes are more expensive than another inch or two of foam on the walls for achieving the same load reductions, other times not- depends on your designs, and your actual quotes for the upgrades.  (I keep hoping vacuum-insulated glass will make it into residential windows soon, rendering triple-panes  obsolete, but I'm not holding my breath.  It HAS to be cheaper to make than triple-panes and beats them on performance, but hasn't been fully released to the market yet.)

"Sometimes the triple-panes are more expensive than another inch or two of foam on the walls"
but you'll feel warmer sitting near triple pane windows, since the interior pane will be closer to indoor air temperature. Comfort is important.

Posted By Bob I on 14 Jan 2014 01:37 PM
"Sometimes the triple-panes are more expensive than another inch or two of foam on the walls"
but you'll feel warmer sitting near triple pane windows, since the interior pane will be closer to indoor air temperature. Comfort is important.

Agreed, comfort counts, but how much comfort and at what cost, in this moderate climate?

There are some pretty good mid-U0.2s double panes out there, and at Bloomington's January mean temp of about +30F I'm not sure the comfort argument for triple-panes works as well as it does in a zone 5 or zone 6 climate. Bloomington IN is still zone 4.

It makes a comfort difference at the 99% design temp (+5F) and lower but that's by definition only 1% of the hourly bins.  Whether it's worth the upgrade in this case is sort of like the low-load radiant floor situation- highly subjective.