We are working on plan details for a new 2-story (plus basement) house in climate zone 6. Square footage is roughly 1300 per floor. The plan now is to use ICF for the foundation and open-cell spray foam or sprayed cellulose in the 2x6 framing on the main and second story.

What are some general thoughts on the ideal HVAC system for a new multi-story house? We plan to have a conventional high-efficiency, multi-stage/ECM natural gas furnace and also an HRV. I would entertain a mini-split system, although I don't believe the local HVAC companies have a lot of experience with air-source heat pumps, which makes me a little nervous about being a guinea pig.

From what I have read, there are mixed opinions on zoned systems. The consensus seems to be if they are done right, they are nice, but very few are done right. We had one in our last house (with ground-source heat pump), and I was not entirely happy with it.

Would installing two systems be preferable (one on the second floor with bedrooms, one in the basement for the basement and main level) to a single zoned system?

Or how about a single system with averaging t-stats on each floor?

Or perhaps one conventional system for the basement and main level, and mini-splits for the second story?

I know with this issue, as with all building issues, the devil is in the details, and a well performing system depends on the skill of the installer, the validity of the manual J/D calculations, commissioning, etc. I'll be meeting with the HVAC contractor recommended by our GC soon, and I'm trying to do my homework.

Thank you.

Without the heating & cooling load number it's a bit silly to start specifying the mechanical systems.

Almost all packaged gas fired HVAC systems (even multi-stage or variable speed) are ridiculously oversized for the loads of a 2600 foot 2015 code-min house. You're probably looking at a total heating load of about 30,000 BTU/hr @-10F or less if you take the time to optimize the building envelope, even at code-min.

A code min wall in zone 6 is 2x6/R20 + R5 continuous insulation, or 2x4/R13 + R10. With an ICF foundation it's probably going to be better to go with 2x4/R13 + R10 EPS, since you can then park the bottom plate of the studwall a half-inc back from the edge of the concrete (to accomodate the half-inch of sheathing), and have continuous 2.5" of EPS from the eaves to the footing of the foundation. That is sufficient dew point control for up to R15 of cavity fill (high density fiberglass or rock wool, or 0.7lb open cell foam), with no interior vapor retarders, which gives it a LOT of moisture resilience, and slightly better than code min performance.

If you want to keep with 2x6 framing, pull the framing back a full inch from the edge, and go with 3" (R12.6) of exterior EPS, and R20-R23 cavity fill, to have enough exterior R for dew point control at the sheathing to skip the interior vapor retarders. That would have about 1.4x code-min performance.

Don't trust the GC or the HVAC contractor to come up with the Manual-J. Hire a competent engineer or architect, and the loads down iteratively.

Zoning by floor usually makes the most sense. Including the basement as a zone for a first floor almost never makes sense, due to the fact that the heat & cooling loads of a basement are small and vary little with outdoor temps compared to what happens for first floors. Mitsubishi just released a half-ton cold climate mini-split (the MUZ/MSZ-FH06NA) that would more than adequately cover the loads most 1300' basements in zone 6: https://meus.mylinkdrive.com/files/MSZ-FH06NA_MUZ-FH06NA_ProductDataSheet.pdf

It's only $108 cheaper than the 3/4 ton, and modulates down to the same level as the 3/4 ton, but with with an ICF basement you'd have to have a lot of air leakage or a lot of basement windows to need the bigger one.

So, what's your location (got a ZIP code?), and what are your available fuel options/costs, and what are your electric rates?

Thank you, Dana.

Indeed, heat gain/loss calculations are needed to fully inform the discussion of HVAC options. We have not designed window layout, etc. yet, so those load calcs are unknown now.

Our zip is 56001. It is a small community, with only a few local builders, none of whom are experts in green building, and sadly most of whom have no interest. The GC we picked has an excellent reputation for honesty and craftsmanship, has a pleasant manner, and we trust him.

He has done ICF foundations for about 15 years. Another reason we are choosing ICF is the new 2015 MN building code, which requires R10 exterior foundation insulation. Trying to achieve this with a standard poured foundation and trying to successfully fasten EPS to it is not my idea of fun.

Ideally, I would like to do 3" of EPS outside the structural sheathing of the main and second story, and I am going to talk to our builder and his partner about this. But I know they have not done this before and am leery about being a test case. So my guess is we'll do the ICF foundation and have no exterior insulation above.

Local electric rates run about $0.107/kwh. NG is at the curb of our building lot and runs about $7.16/1000 cubic ft.

I like the idea of using mini-splits and will ask the HVAC contractors if they have experience installing any. Thanks, too, for the suggestion that zoning, one zone per floor, is likely the way to go. As I said, we had a 2-zone geothermal system in a past home, but I think the contractor ill-designed it. For one, the system was oversized by at least 2x (e.g., electric strip backup heat never came on once in the four years we lived there, and this was in climate zone 5 with real winters and some nights at -20 F). For another, they zoned the basement with the main level and, worse, used a basement TV room as a dump zone to balance the CFM flow. That room was freezing in the summer and hot in the winter.

So long story short, if we do go with a zoned system this time, I want to be sure the contractor does it right. For various reasons, I have ruled out hydronic in-floor heating. Again, I would be very interested in mini-splits but want to have confidence in the contractors' experience installing them.

Mankato's 99% outside design temp of -12F (see: https://articles.extension.org/sites/default/files/7.%20Outdoor_Design_Conditions_508.pdf) limits your options for mini-split solutions to the Mitsubishi H2i series and the Fujitsu _ _ RLS3-H, but it's at least possible to get there. If properly sized a ductless system will have a seasonal average COP between 2.5-3.0, maybe a hair better if all the stars align.

From an operating cost point of view, assuming an as-used COP of 2.5, the ductless will deliver 2.5 x 3412= 8530 BTU per kwh. Normalizing to million BTU (MMBTU), that's 1,000,000/8530= 117 kwh/MMBTU. At $0.107/kwh it comes in at $12.52/MMTU. If the COP averages 3.0 it would run (2.5/3) x $12.52= $10.43/MMBTU.

Assuming a typical 102,000 BTU/ccf or 1,020,000 BTU/kcf, burned in a 95% condensing furnace or boiler it comes it delivers 1,020,000 x 0.95= 969,000 BTU/kcf. Normalizing to MMBTU that's 1,000,000/969,000= 1.03 kcf/MMBTU, so the marginal cost is going to run about 1.03 x $7.16= $7.37/MMBTU, plus a bit for the power used by the controls, air handlers, pumps, call it $7.50/MMBTU, which is substantially less than with ductless heat pumps.

From a carbon footprint point of view gas-burner dumps 117lbs of CO2 per source-fuel MMBTU into the atmosphere, so net at 95% AFUE that's about 117/0.95= 123 lbs/MMBTU, plus a bit more for the power use- call it 130-ish.

Taking the worst-case 2.5 of 117kwh/MMBTU, and the 2020 business-as-usual projection by the EPA of Minnesota's grid emissions of 1658lbs/Mwh (= 1.685lbs/kwh, see: https://www3.epa.gov/airquality/cpptoolbox/minnesota.pdf ) the ductless system would be dumping 1.658lbs/kwh x 117kwh= 194 lbs/MMBTU, which is substantially more than condensing gas.

At the CPP's 2030 target of 1,213lbs/Mwh it would be emitting 1.213 lbs/kwh x 117 kwh = 142lbs/MMBTU, still slightly more than condensing gas, but if the COP is closer to 3 it's pretty much a wash carbon-wise, if not cost-wise.

Now that's the state-wide average, not necessarily the local utility average, so it's probably worth investigating options with your local utility if you're looking to green-up the power use. Minnesota is not an electricity deregulated state (http://www.thederegulationofenergy.com/wp-content/uploads/2015/02/us-dereg-map-v.png ), so you can't just buy lower-carb power to be delivered through the local grid. Rooftop solar net-metered could lower the carbon footprint of electricity, currently at a higher levelized cost than your current retail rates even after federal suisidies (that will likely change by 2025), so for the time being it'll be greener for you to go with a condensing gas solution.

With the small heating loads of new code-mins you'll get better efficiency out of hydronic heating systems. If panel radiators/radiant floors etc are deemed too expensive (they're comfy, but not super cheap), there are now quiet thin-profile fan coil solutions with efficient ECM drive blowers suitable for micro-zoning a house using a condensing hot water heater as the heat source, eg:

http://htproducts.com/fan-coil.html

Cheap fin-tube baseboards are much less effective at condensing temp, and most boilers won't modulate low enough to not short-cycle itself into low efficiency and an early grave on a micro-zoned system, but something like this backed up by a 75-125,000 BTU/hr tank type hot water heater or combi system can work pretty reasonably.

You'll need separate systems for air conditioning, but yours is a heating dominated climate. BadgerboilerMN (A Minneapolis area hydronic heating contractor who posts often on the radiant heating forum here) is able to cool his whole house adequately with a single 1.5 ton Mitsubishi mini-split with the head mounted near the top of the stairs. YMMV. The load numbers will tell if something like that works for you, or if a pair of mini-ducted mini-splits (one per floor) would be more appropriate.

Say you go with good mini-splits (mostly for cooling) and nat gas (for heating). Above ~32F (freezing), it becomes more efficient to use the mini-splits for heating.

EPS foam can be put into the form used for a poured foundation - no attachment issues (and this will increase the effectiveness of the concrete as thermal mass).

If you use Fujitsu, you might use an installer off of their list:

http://www.fujitsugeneral.com/contractors/search/index.cfm?int_radius=100&str_zip=56001

My understanding is that if they are listed on Fujitsu's site, they are trained and authorized to do the sale, install and service.

Does the code specifically require exterior insulation on the foundation? 

In Minnesota exterior foundation insulation is indeed prescribed by code for new construction, at a minimum of R10 (IIRC) with a total R of at least R15, with no more than half the R value being on the interior. (Varies considerably in detail from the basic IRC.)

If community solar takes off in MN as expected, it's possible to green-up your power-use footprint that way, but it's probably cheaper to get it installed on-site.

I was wondering what is the reasoning behind the exterior insulation as opposed to interior insulation?

As I understand it the concern in MN has been about frost-heaving the foundation if the footing doesn't get sufficient parasitic heating from the interior. I personally don't believe that concern is well founded if the footing is at the code-prescribed depth, and the foundation drainage is also built to code. This may be vestigial stuff left over from when shallower foundations were installed in wet ground, but I don't really know when/how the MN peculiarities around foundation insulation first entered the state codes.

It is always good to take professional suggestions, but if you are still confused then you can go through the reviews of the companies like comfort services, HVAC Services, amtek air reviews so that you can make up a good decision.

BBB listings Angies List whatever can cull the out & out crooks, but don't necessarily find you a contractor nuanced in the finer points of either building science or heating / cooling load calculations. For the calculation end of it it's better to pay an engineer or energy nerd whose product is the accuracy of the calculations rather than trusting the competence & diligence of the equipment installers.

Some HVAC contractors are competent to do that, but even among those who offer Manual-J calculations as a service a clear majority have big errors or grossly inaccurate assumptions on air leakage, and often are a mere pro-forma quickie using all code-min assumptions on R-values/U-factors, and code-max on air leakage, which is a LOUSY way to size the equipment, especially on better-than-code houses.

So true. The other issue is that the majority of available US heat sources are made for code minimum over-sized residences. Please let me know when you find a 20-40 kbtu mod-con boiler with 10:1 turn down...

Closest thing I've seen in the US are still twice that size. The 80K version of the condensing stainless fire tube Kiturami Homsys boiler/tankless HW heater (aka HTP UFT-080W, or Westinghouse WBRUNG-80W ) with a 10:1 turn-down is a pretty reasonable fit for 95% of the houses in the US, but a 40K version with a 10:1 turn-down would be nice, eh?

In Korea Kiturami is manufactures smaller variations on the theme for other markets, but I haven't found one that's 10:1, condensing, and stainless.

Perhaps many people looking for a 20 Kbtu boiler should be looking at a Chiltrix instead. But I haven't seen any actual-user reports.

I haven’t seen any third party testing reports on the Chiltrix either.

The Chiltrix is essentially an air-to-water heat pump system similar to the Altherma system by Daikin. The Chiltrix reportedly uses far less refrigerant and also uses an integrated/closed system approach that makes installation very simple without the need for a refrigeration technician...great for DIYers.

If one believes their claims, the Chiltrix has almost twice the COP as the Altherma. Chiltrix is affiliated with Wilco (e.g., same German company that manufactures the HR pump used in the NexGen electric boiler HR appliances) and uses a Toshiba compressor. In fact, I believe every component used in the Chiltrix is off-the-shelf from some other manufacturer. The Chiltrix is about twice as expensive as a comparable mod-con boiler. Most people who choose to go the mod-con boiler route have done so largely to avoid the significantly higher system complexity, associated higher acquisition cost and associated higher failure/maintenance risk of a heat pump system...and also have very low gas cost relative to electric cost.

If the Chilrix COP claims are indeed accurate and the system actually demonstrates high reliability, it could indeed become a game changer and likely a better solution than ground source heat pump systems. Of course, most everything is already a better solution than ground based heat pumps systems when you factor in their associated higher acquisition cost and associated higher failure/maintenance risk.

In high R houses the design air conditioning peak loads are often considerably higher than the space heating loads, even in cold climates, and it's not insane to consider doing it all with a pair of dedicated mini-ducted mini-splits (one per floor.) It's not as cushy as radiant floors but it's not terrible, and it's quieter than any oversized condensing hot air furnace would be.

A 1.5 ton mini-ducted Fujitsu is good for about 20K@ +17F and modulates down to about 3K @ +47F, but is unrated for temps below -5F. It'll still be putting out a fully specified amount of heat at -5F (consult the extended temperature capacity ratings). It keeps going at even lower temps, but if your 99% outside design temp is lower than -5F you might not get a "pass" from the code enforcement folks unless you install sufficient resistance heaters to fully cover the design load.