We are owner/builders constructing a retirement home near Olympia, WA.  TF Systems ThermoForm ICF walls - 5' conditioned crawl space,
and 10' main walls.  Crawl space walls are 10" concrete with 2.5" EPS inside, 4.5" EPS outside, 2" EPS under crawl space slab (rat floor).
Main walls are 6" concrete, 2.5" EPS inside, 4.5" EPS outside.  12-16" cellulose in ceiling, so around R40-R50. 
Planning to use triple pane casement windows, fiberglass frames. 
Also planning to install HRV and mini split ductless heat pump.  Overall, we are trying to get a really tight envelope with well above
average insulation.

My "rough" guess on HVAC sizing is 18k BTU cooling (probably really only needed 4-6 weeks a year, and then only
a few days a week during that time).  Maybe 30k BTU heating (again, probably only a few weeks a year with sub-freezing
nights).  We really like using a wood stove, so that will probably be the "main" source of heat - but you have to satisfy the
county building department...

Do you think one of the Mitsubishi single zone 24k or 30k units would keep the temperature fairly even throughout the house?
Or possible a dual zone 6k in the master bedroom, 18k or 24k unit in the middle of the house (open floor plan, great room,
kitchen/nook and dining room)? 

I would like to install the wall unit (1 or 2) on interior walls and would plan to run the linesets down into the conditioned crawl space
and then through the outside wall on the far side of the house.

First time post here, so sorry if I didn't follow all the correct "etiquette" yet!

It depends upon your layout and lifestyle. We have a small (1400 sq ft) ICF house, with a very open plan. The LR, DR, and kitchen are essentially one room. Our requirements are 9K BTU for A/C (in hot humid SC), and 12K heat. We have two Mits units; a 12K in the DR, and a 9K in the master suite. If we leave the doors open between rooms (and we always do), the temperature throughout the house is very even. We rarely use the bedroom unit except for very cold weather (single digits). In the summer, we run the main unit during the daym then turn it off, and run the smaller unit in the bedroom at night. The cost of electricity for HVAC runs about $200 for the entire YEAR! In spring and fall, we go for long stretches without running the A/C or heat at all. All in all, we A/C about 4 months out of the year, and heat for 2-3 months. This AM, it's 28 degrees out, the house is 65, and we haven't had the heat on in a couple weeks. This has been an unusually warm winter, though.

Make sure you do careful calculations for your actual load. I used HVACCalc ($50) to do our Manual J, being very careful with what I put in. The system has performed a bit better than predicted. I had a local 'pro' tell me I needed 2-1/2 tons of A/C over the phone, without even seeing the house!

Our house is 1400 sq ft, 9-1/2' ceilings, R-50 cellulose in the attic, Marvin double pane windows, and full ICF (Fox), including the basement.

Hi, jdebree -

     Thanks for your input!  The seasons are bit different here in the Puget Sound - mild summers (90+ is a heat wave, avg high about 80 in August)
and then cool fall and winter (but not too cold, avg low usually 30+ even in January with daytime highs about 40+ in January).  Most people
don't have air conditioning at all, or maybe one window unit.  So I'm sure the "heating" will determine sizing here.  As I said, we're really
interested in energy efficiency, but along with the efficiency comes comfort (even temps, no drafts, etc). 

     We also have a very open floor plan and always leave all the doors open (unless we have company and there's a messy bedroom!).  I am
glad to hear you have even temperatures throughout!  I am going to do a more complete Manual J calculation for peace of mind, but I also
have heard ridiculously "high" BTU requirements from some local contractors.  With these mini splits, my understanding is it's best to be a little
undersized (only struggling to keep up on the coldest days) as opposed to be way oversized.  Especially with all the wood available for the wood
stove, I really think we can do something like a 6k in the MBR and 18k or 24k in the Great Room - and probably only run the MBR unit on the
coolest (or the few very warmest) days like you do.  I'm also expecting that the mass effect will keep the temperature much more even
than "stick frame" builders are used to seeing.

Thanks again for sharing your real world experience!

How many SF is the floorpan? That sounds like a LOT of heat for the R values you are mentioning (R35 walls?) unless you are planning a huge house or have really huge windows. I am doing something similar up in Port Angeles R38 walls R60Ish ceiling and my heat requirement calcss for just under 2000 SF are significantly less than the numbers you have mentioned...

Hi, Ronmar -  The house is just about 2,100 sf.  Single story rambler with 9' interior walls, but on a 5' conditioned crawl space.  What did you come up with for heating requirement?  I thought the 24k - 30k for heating would be somewhat on the high side.  Are you in the 18k range?  I only did a very preliminary calculation
and probably didn't give myself enough credit for all of the energy features.  I really don't want to "oversize".  We don't have huge windows, and they are going to be double or triple pane.

Thanks for your input!

Yours actually looks a lot like ours minus the attached garage, same large covered front porch/patio. Ours is a 56X40 rectangle with a 7/12 pitch roof. We are also building into a hillside so We put ours on a full daylight basement with a single garage door down there. Also a mostly open floorplan with just under 9' ceilings(natural ICF stack height). We stuck to around the 15% window to floor area ratio(including aerogel skylights) but are using tripple pane/fiberglass casements with an overall U factor in the .24 range. My heat loss(done manually and calculated) at our 99% design temp comes in under 10KBTU We are planning on very tight with a HRV. Not planning on A/C. We have lived on the property for 14 years and I havn't encountered a situation we cannot get around with just ventilation up here.

Google Borst engineering. They have some calculators free for personal use. I like their heat loss calculator for a few reasons, fist off, I am cheap and it is free. Second is, its simple. The instructions are VERY detailed, but once you get your brain around what you are entering, it really is quite easy to manipulate to see the effects of design changes. It takes what I was doing by hand and refines it considerably, so I was tickled when I came across it. It dosn't take into account some things like skylights or an HRV, but you can average your ceiling R values based on the skylight area percentage and value, and manipulate the air exchange numbers to account for a 90% efficient HRV. I would reccomend it as a good learning experience.

Thanks for the additional info ronmar! I found the Borst website and the calculators - I too like "free" software! We have about 13% window/floor ratio, and I agree that practically speaking there will probably never be a real "need" for any A/C. The only reason I was trying to guestimate the cooling on the few hot days was because the electricity is so cheap around here that it seemed to make good sense to try and go with a heat pump. The mini split ductless seems to be in vogue right now, so I thought I would evaluate/price it out. Our 5' conditioned crawl space is mostly below finished grade (the lot slopes about 9-10%), so I think that will also help moderate the heat loss. If the calcs show anywhere near 12-16k BTU for heat loss, I would feel really good about being able to achieve an even temperature in the house throughout the year.

Of course, you are in the banana belt of Washington, so maybe you are a little bit milder climate!?  What windows are you using?  I looked at
Fibertec (from Canada) and I think Marvin makes a fiberglass triple pane.  I really need to spend more time on the windows and get them
ordered soon.  Are you doing a lot of work yourself?

The design temps are a little different. Ours are 28F and 75F and Olympia is 17F and 85F but I don't think our climates in general are all that different. That 17F temp will cost you a few more BTU/HR though... We get a little more coastal influence I guess.

We went with Duxton windows(canada) who use fiberglass pultrusions from inline I think(also Canadian) and cardinal glass units(US). We wanted turn and tilts which was one reason we went with them.

We are doing most all the work ourselvs. We used a local ICF guy for consult and the CC pours and I will bring in a few guys to hang the roof trusses in a week or so if the rain and snow will cut me some slack Might hire the drywall finish done as I have never been that patient with it...

Our 99% design temps are 20F and 90F. We've seen a range from 3F to 107F (!). We probably get a lot more solar gain, as SC is generally sunny. We have deciduous trees all around, so the house is shaded quite a bit during the warm months. The ICF really evens out the extremes, though. One feature we often use is the 'Dehumidify' mode, something I would think would be handy in your humid climate.

Surprisingly even though we are in the pacific NW, at least up here on the olympic peninsula we do not have too much trouble with humidity. In fact, I typically have trouble geting the house up to 50% in the winter months.

Oh pacific NW, that's great but yeah it really gone messy in the winter...

Posted By ronmar on 05 Mar 2017 02:03 PM
Surprisingly even though we are in the pacific NW, at least up here on the olympic peninsula we do not have too much trouble with humidity. In fact, I typically have trouble geting the house up to 50% in the winter months.

That's not surprising at all, since the outdoor dew points in the PNW are pretty low year-round, rarely north of 60F even in summer. From an air-dryness point of view it's pretty dry compared to most of the east coast.

You shouldn't be trying to bring the RH UP to 50% during any time of year, since that's the high end of human-healthy & comfortable range (30-50% RH @ 70F). In winter keeping it under 40% is advisable, since that reduces the moisture burden on the house. If you are actively humidifying the house, you are putting wood sheathed homes at higher mold risk.

Regarding the ductless heat pump system design, don't guess. Run an AGGRESSIVE Manual-J load calculation on a room by room basis. A room with a design heating or cooling load of less than 3000 BTU/hr is not a good candidate for getting it's own head or cassette.

The Mitsubishi multi-splits have a fairly high minimum -modulation on the compressor- possibly too high to actually modulate with load in your house. The FH06NA and FH09NA single-zone mini-splits can modulate down to ~1600-1700 BTU/hr and WILL modulate. By comparison, the MXZ-3C24NA2 (3-zone 2 ton), can only modulate down to 11,400 BTU/hr @ +47F in heating mode, which may be higher than your design heating load, and may literally NEVER operate in an efficient modulating mode (!)

When you have some real load numbers, and not just a WAG, you can work from there to find the most suitable ductless solution.

Unless this is some gia-normous house there's no way your heating load in that high-R ICF will be anywhere near 30,000 BTU/hr @ 23F (Olympia's 99% outside design temperature: https://articles.extension.org/sites/default/files/7.%20Outdoor_Design_Conditions_508.pdf ) There are many exceptions to prove the rule, but for a 2500' code min house in your climate you're probably looking at no more than 25,000 BTU/hr heating or cooling, and with a high-R ICF house it's likely to come in under 15,000 BTU/hr if you use heat recovery ventilation. Your latent cooling loads are actually negative, and unless you have a lot of west facing glass your peak cooling loads are going to be miniscule.

Posted By Dana1 on 06 Mar 2017 03:52 PM

Posted By ronmar on 05 Mar 2017 02:03 PM
Surprisingly even though we are in the pacific NW, at least up here on the olympic peninsula we do not have too much trouble with humidity. In fact, I typically have trouble geting the house up to 50% in the winter months.

That's not surprising at all, since the outdoor dew points in the PNW are pretty low year-round, rarely north of 60F even in summer. From an air-dryness point of view it's pretty dry compared to most of the east coast.

You shouldn't be trying to bring the RH UP to 50% during any time of year, since that's the high end of human-healthy & comfortable range (30-50% RH @ 70F). In winter keeping it under 40% is advisable, since that reduces the moisture burden on the house. If you are actively humidifying the house, you are putting wood sheathed homes at higher mold risk.

Yep, very wet, but not very humid:) From a personal comfort standpoint we find under 40% inside too dry. We like between 40-50. But when it does get cold here, we tend to drop into the lower 30% range...

Hi, Dana1 -

     Thanks for your input - I get a little more educated from each member!  I've started to look at some Manual J analyses, and I do think that the heat
load will come in around 14k - 18k.  As you say, the cooling load is tiny, but an added benefit of the mini split for those few hot days.  I looked at the
MXZ-2B20NA-1 (maybe a slightly older design) (2-zone 1.7 ton) and found that it can modulate down to 8,500 BTU/hr @ +47F in heating mode,
while the MZ-GL15NA can modulate down to 4,800 BTU/hr heating and 3,100 BTU/hr cooling.  The MZ-FH15NA can modulate down to 5,150 BTU/hr heating
but only down to 6,450 BTU/hr in cooling mode, but with the "hyper heat" capability.  In as much as the 99% design temerature is 23F, is the "hyper heat"
worth the slightly extra cost?  Is there any drawback to the "hyper heat" feature, other than a higher minimum cooling load?

     I like the idea of a single zone, but was initially worried about temperature gradient throughout the house.  I'm becoming convinced that with all
of the energy features, temperature should be fairly uniform.  On the other hand, I wouldn't mind installing a 2-zone system with 6K in the MBR
and a 9K or 12K unit in the Great Room.  I suspect that the 6K unit would get fairly infrequent use, but would allow some slight flexibility on those
few extended cold periods.  Since we will primarily be heating with an efficient wood stove, maybe that benefit goes away almost entirely.

     We do plan to install an HRV from the get-go.

The high minimum modulated output of the multi-splits are a problem at your low loads.

The Mitsubishi FH09NA can modulate down to 1600 BTU/hr out @ +47F, yet can still deliver more than 12,000 BTU/hr @ +17F, which is more than half your design heat load. A pair of them can modulate down to 32,000 BTU/hr, yet cover your peak loads with HUGE margin. The FH06NA has the same minimum modulation, but maxes out at about 10.7K @ +17F. It's maybe $200 cheaper than the FH09.

The combined minimum modulation of a pair of FH09s (or FH06) just a bit more than half the minimum output of an FH15, and would deliver better heat distribution (2 heads instead of 1), and even more capacity margin at low temp than a single FH15.

http://usa.mylinkdrive.com/uploads/documents/4560/document/MSZ-FH09NA_MUZ-FH09NA_Submittal.pdf

http://meus1.mylinkdrive.com/files/MSZ-FH06NA_MUZ-FH06NAH_ProductDataSheet.pdf

In my neighborhood a pair of FH09s installed by a pro comes in at about $5500-6K, all-in when put out to competitive bid. A pair of FH06 would be $300-500 less.

But get serious about calculating the actual room-by-room heat load numbers, stop guessing. Hire a certified P.E. or a RESNET rater that has experience with ICFs, which have somewhat lower peak loads at any steady-state U-factor than low mass buildings. Make sure they assume something like 1 ACH/50 air tightness, not the code-max 3ACH/50, and make sure they credit the HRV ventilation. Infiltration & ventilation loads can skew the results pretty wildly when you're in low-load territory.

The high minimum modulated output of the multi-splits are a problem at your low loads.

The Mitsubishi FH09NA can modulate down to 1600 BTU/hr out @ +47F, yet can still deliver more than 12,000 BTU/hr @ +17F, which is more than half your design heat load. A pair of them can modulate down to 32,000 BTU/hr, yet cover your peak loads with HUGE margin. The FH06NA has the same minimum modulation, but maxes out at about 10.7K @ +17F. It's maybe $200 cheaper than the FH09.

The combined minimum modulation of a pair of FH09s (or FH06) just a bit more than half the minimum output of an FH15, and would deliver better heat distribution (2 heads instead of 1), and even more capacity margin at low temp than a single FH15.

http://usa.mylinkdrive.com/uploads/documents/4560/document/MSZ-FH09NA_MUZ-FH09NA_Submittal.pdf

http://meus1.mylinkdrive.com/files/MSZ-FH06NA_MUZ-FH06NAH_ProductDataSheet.pdf

In my neighborhood a pair of FH09s installed by a pro comes in at about $5500-6K, all-in when put out to competitive bid. A pair of FH06 would be $300-500 less.

But get serious about calculating the actual room-by-room heat load numbers, stop guessing. Hire a certified P.E. or a RESNET rater that has experience with ICFs, which have somewhat lower peak loads at any steady-state U-factor than low mass buildings. Make sure they assume something like 1 ACH/50 air tightness, not the code-max 3ACH/50, and make sure they credit the HRV ventilation. Infiltration & ventilation loads can skew the results pretty wildly when you're in low-load territory.