Hi We are building a new home in Denver, Colorado and are looking to use a ceiling mini split unit ad the main source of heat and cooling in the house. It is a ranch style house, 1350 sf w a full basement, 9 ft ceilings, 3 bed 2 back on main level. It is a rectangle shape, w the kitchen/great room being 17' x 37', and a main hallway with the master on one side and the other two rooms on the other side. Master is 10 x 14 w a 5 piece bath, other two rooms are 10 x 11 with a jack and jill bathroom. What size unit would be the best and where would be the best location for the unit? Thanks!

Have you done a room by room as well as a whole building heat loss study?

Hi yes I have the calculations for the RES check I did for the building permit - not with me now but I can post later when I get home. It is a well insulted house w/ 2x6 walls w/ r-19 and all windows are double pane and low E.

Denver is in US climate zone 5.

An R19 fiberglass 2x6 wall assembly doesn't even meet IRC 2009 code min, but it did make IRC 2006 code min.  This may meet bare-minimum legal construction standards, but hardly what performance builders would call "well insulated"  for a place with a 99% outside design temp of -3F.    Compressed into a 5.5" wall cavity an R19 batt performs no better than R18 (if perfectly installed), and assuming a ~25% framing fraction (typical for 16" o.c. framing) and R1 for the gypsum + sheathing + siding that comes in at a "whole wall" average performance of R13.6.

If it were 2x6 R23 rock wool w/R8 of exterior insulating sheathing it would be fairly well insulated, with a whole-wall R of about R23.2, which loses about 40% less heat per square foot of wall area than an R19 wall.  That makes point-source or multi-point-source heating such as ductless mini splits a lot easier, but still not a slam-dunk.

If by "We are building..." you mean this place is still in the planning phase there are lots of ways to improve the prospects for heating with mini-splits. If not, we'll have to see where you stand. (Is there at least an insulated foundation?)

If it's too late to make any real changes, that's not to say it can't be heated/cooled primarily with ductless mini-splits, only that it'll be more difficult. A key concern is the individual heat loads of the rooms doored off from the main space, so list the room-by-room loads and we'll see where it stands.  Any doored off room with a design heat load more than 800 BTU/hr is going to need some sort of heat source, but if the load isn't at least 5000 BTU/hr it's not worth giving it it's own ductless head. 

You can do a lot for the doored off rooms without burning a lot of power by using radiant cove heaters sized for the room loads, and controlled with a combination of occupancy sensor and thermostat, letting those rooms run cool when unoccupied.  By use of occupancy sensor control it'll sip power, since it only heats up when occupied, but the hit in comfort is low with the use of radiant cove heaters, since they come on fast and 99% of the heat transfer is via direct radiation heating up the objects in the room directly (including the humans), and raising the average radiant temp in the room indpendent of the air temperature.  Average radiation temp is more important to human comfort than air temperature, which is why on a clear calm +10F day humans can be comfortable in T-shirts if standing in direct sun.  Radiant cove heaters can make it comfortable while the room is coming up to temp, even if the room is starting out at 60F or lower.  As long as the radiant cove is big enough to heat the room to 68F at the 99% outside design temp it will meet code doing it that way.  Total installed cost per room should be under $500 (under $250 as a DIY) as opposed to $1500-2000 for an oversized ductless head that can't run at it's optimal efficiency due to extreme oversizing.

Any analysis of "can I get by without a heat source in this room" also requires an answer to "how much can it drop below the heated rooms" and "will the door be kept closed".

Thank you for your detailed reply. Let me get those numbers when I get home and I will get this post updated. I like where this is going. Again, thanks for your thoughtful input.

Posted By Dana1 on 08 Apr 2014 04:54 PM
which is why on a clear calm +10F day humans can be comfortable in T-shirts if standing in direct sun.

True as long as I'm inside @ 70F looking out a window :)

Ok so im having a little difficulty determining the heat loss of the rooms - i cant find the r-factor for the windows. Anyway, the master is 10 x 14 with one exterior wall. The window on that wall is 5x5 (here is the link to the exact window for that room, as well as the other two rooms:

http://www.lowes.com/pd_319625-40446-748171612737_1z0w6xo+1z11pos__?productId=3119817&Ns=p_product_qty_sales_dollar|1&pl=1¤tURL=%3FNs%3Dp_product_qty_sales_dollar%7C1%26page%3D1&facetInfo=

One wall is adjacent to the garage, which is also a 2x6 wall insulated with drywall on both sides. The other two walls are interior walls, 9' flat ceiling insulated to r-38.

Room 2 is 10 x 11, with the 5x5 window on the exterior wall, one wall is adjacent to the garage, which is 2x6 insulated with drywall on both sides. Other two walls are interior walls, 9' flat ceiling.

Room 3 is 10 x 11, has the 5x5 window on the exterior wall, and has three interior walls, 9' flat ceiling.

Dana1, the house is in the rough framing stage, we are getting the shingles on and beginning rough plumbing. Only sheathing is on the walls. Exterior walls are 2x6 at 24" O.C. The house has a full basement, with a structural floor in the basement and a crawlspace under the structural floor. Both basement and crawlspace will be insulated. I know that Denver is in Zone 5, but we do have moderate winters here. It will get below freezing maybe a handful of nights out of the year. I feel we are on the edge between zone 5 and 4. Don't know if that makes a difference with anything but thought id throw that in there since you made the point.

jonr, as stated in the previous paragraph the basement will be insulated, and for the time being with our two young kids, all the doors will be left open. Obviously that will change with them down the road but for the time being we will assume all doors are left open.

TLP, im with you - im not standing outside in 10F weather either!

Again, thank you all for your detailed responses and advice. If needed, i can send a PDF of the house plan.

Posted By fordman460 on 09 Apr 2014 01:13 AM
Ok so im having a little difficulty determining the heat loss of the rooms - i cant find the r-factor for the windows. Anyway, the master is 10 x 14 with one exterior wall. The window on that wall is 5x5 (here is the link to the exact window for that room, as well as the other two rooms:

http://www.lowes.com/pd_319625-40446-748171612737_1z0w6xo+1z11pos__?productId=3119817&Ns=p_product_qty_sales_dollar|1&pl=1¤tURL=%3FNs%3Dp_product_qty_sales_dollar%7C1%26page%3D1&facetInfo=

Those windows have a listed U value of 0.3.

R value = 1/U value; therefore, R value for that window = 1/0.3 = 3.33.

"add reply" returns do not work in Chrome. Void.

Lets see if my returns work better in chrome using quick reply...

Your in a risky situation of not having a home that is minisiplit ready, especially in bedrooms doors closed that lack airflow and are not insulated well enough and air tight. What ACH @ 50 are you shooting for and ventilation? The question of sizing and placing your unit is a little late. or at least having an idea where your lines will be and/or ducts, air handlers, etc. It is already difficult with mass envelopes that are air tight. Mineral wool will mitigate risk. Perhaps a wool insulated wire and plumbing chase furred inside your 2x6 wall that eliminates insulation breaks of the R23 and/or wool board outside.

Wool has a high perm rating, makes a nice retarder inside and out. Lattice over it for a rain cavity then attach siding. Here are some good vapor retarder for Z5 that vary with humidity, they are an air barrier too, become more vapor open in summer when relatively humidity is high and vapor closed in winter when the relative humidity is low. Thus protecting the structure against vapor diffusion in winter, while allowing it to dry inwards in summer - - something a conventional vapor barrier can't offer. If it works with the roof planes for continuous put it against the chase furs-to-2/6 face. Just added R14 to your R23 2x6 wall, and two insulated thermal breaks for the studs.

http://www.foursevenfive.com/index.php?main_page=index&cPath=70_71_98_97_101_76&zenid=c0s23m7q2tokt4qhnkjqaj69f3
http://www.certainteed.com/products/insulation/mold-prevention/317391

About all I can think of at this point... It's been hit and miss with the calculations, models, mfg specs, etc, from what I read. I'm looking at field test results trying to figure a way to install the unit and system last once I understand my actual's.

Posted By fordman460 on 09 Apr 2014 01:13 AM
Ok so im having a little difficulty determining the heat loss of the rooms - i cant find the r-factor for the windows. Anyway, the master is 10 x 14 with one exterior wall. The window on that wall is 5x5 (here is the link to the exact window for that room, as well as the other two rooms:

http://www.lowes.com/pd_319625-40446-748171612737_1z0w6xo+1z11pos__?productId=3119817&Ns=p_product_qty_sales_dollar|1&pl=1¤tURL=%3FNs%3Dp_product_qty_sales_dollar%7C1%26page%3D1&facetInfo=

One wall is adjacent to the garage, which is also a 2x6 wall insulated with drywall on both sides. The other two walls are interior walls, 9' flat ceiling insulated to r-38.

Room 2 is 10 x 11, with the 5x5 window on the exterior wall, one wall is adjacent to the garage, which is 2x6 insulated with drywall on both sides. Other two walls are interior walls, 9' flat ceiling.

Room 3 is 10 x 11, has the 5x5 window on the exterior wall, and has three interior walls, 9' flat ceiling.

Dana1, the house is in the rough framing stage, we are getting the shingles on and beginning rough plumbing. Only sheathing is on the walls. Exterior walls are 2x6 at 24" O.C. The house has a full basement, with a structural floor in the basement and a crawlspace under the structural floor. Both basement and crawlspace will be insulated. I know that Denver is in Zone 5, but we do have moderate winters here. It will get below freezing maybe a handful of nights out of the year. I feel we are on the edge between zone 5 and 4. Don't know if that makes a difference with anything but thought id throw that in there since you made the point.

jonr, as stated in the previous paragraph the basement will be insulated, and for the time being with our two young kids, all the doors will be left open. Obviously that will change with them down the road but for the time being we will assume all doors are left open.

TLP, im with you - im not standing outside in 10F weather either!

Again, thank you all for your detailed responses and advice. If needed, i can send a PDF of the house plan.

 
Give me an exact ZIP code for estimating the design temperature.

If it's still in the rough framing stage you still have the possibility of making changes that will affect the total thermal & moisture performance of the house. 

Start by defining the exterior sheathing as the wall portion of the primary air barrier: Caulk the sheathing to the framing in EACH stud bay, caulk the seams between doubled-up top plates, and caulk the bottom plate to the subfloor, using an acoustic sealant type caulk.  When it comes time to run the rough electrical, seal every framing penetration, both lateral & vertical with can-foam.  When it comes time to install the windows, use foam backer-rod and a bead of flexible "windows & doors" type can-foam to air-seal around the windows at/near the sheathing, and pack fiber insulation into the rest. Air sealing is by far the cheapest best bang/buck performance enhancement you'll ever make.

Instead of IRC 2006 code min, shoot for at least IRC 2012 and then some for "whole wall" R.  Adding 2" of EPS foam on the exterior would be sufficient R value to protect the sheathing from wintertime moisture accumulation with up to R23 in the wall cavities without an interior side vapor barrier, and would add a lot of resilience to the structure.  It'll be a cost adder of about 80 cents to a buck a square foot for the exterior foam, but it's a cost that still has a pretty good internal rate of return on heating (and some cooling) energy cost savings, as well as some savings on HVAC mechanicals.  The labor costs of installing bottom-of-the-line R19s isn't any more than the labor cost of installing R23 rock wool or R21 HD fiberglass, but the performance uptick is more than the mere whole-wall R value, since it's sufficiently air-retardent that it impedes convection currents within the fiber layers at the outdoor temperature extremes (and actually increases in R value with lower outdoor temps/bigger temperature differences), unlike R19s whose performance is degraded by convective loop heat transfer at high delta-Ts.  Damp sprayed cellulose would end up at about R20, but would be even more air-retardent- whether that is higher or lower cost than R23 rock wool or R21HD fg depends on your local market, but SHOULD be considered, since sprayed/blown goods fill in perfectly around wiring/plumbing etc, and will fill even very narrow gaps at jack studs etc to near perfection. 

If you add exterior foam, don't use any LESS than R7.5, or it will not provide sufficient dew point control.  At 1.5"  R7.5 XPS is a bit thinner, but more expensive than 2"  R8.4 Type-II EPS, but it loses R over time, and uses environmentally damaging blowing agents, making it overall less desirable. At 1.5" R9/R10  polyiso is also thinner than R8 EPS, but has a severe derating at colder temps, but would be an acceptable alternative if the extra wall thickness is an issue.  It's average performance in a zone 5 location would be no better than R8.4 EPS. The performance of EPS increases significantly with lower temperature, as does XPS, but don't try to cheat the system with 1.5"/R6.3 EPS, even though that will sometimes work.

The attic/roofing framing should be sufficient to allow at least 15" of blown insulation above the top plates of the wall assemblies, which will likely require an "energy heel" truss. If it's only set up for R38 batts between 2x12 joists it's worth making some modifications.

It's probably too late, but EPDM sill gaskets between the foundation and sill plates is a pretty cheap & effective air seal and capillary break, far outperforming the cheap foamies.  If the latter is what is installed, it's worth air-sealing and insulating the band joists & foundation sills with an inch of closed cell spray polyurethane (about a buck a square foot).   Insulating the foundation walls with 1.5-2" of EPS trapped in place by a 2x5 studwall with cheap unfaced or kraft-faced R13s is cost effective for zone 5.  Put an inch or two EPS between the bottom plate and the slab (or rat-slab, in the crawlspace) as both a capillary and thermal break. Since it's not structural you can use a single top-plate- it only needs to hold the insulation & wallboard.  Seal the seams of the EPS with fiber reinforced duct mastic, and the edges with can foam, to make the foundation air-barrier continuous with the above grade air barrier (band-joist & sheathing.)  It's worth putting 2" of EPS under the slab(s), if it's not already too late, and float the slabs at the edges with an inch or two of foam as a thermal break against the foundation wall.

You may not want to go whole hog on all of it, but the long term cost effectiveness limits are "whole-assembly" R-values somewhere in the neighborhood of what's shown in Table 2 on p10 of this document.  Read at least the whole first chapter.  In cheaper labor/material markets with high energy pricing you can go a bit more than that, but in high labor/material low energy cost areas maybe not as much.  You'll note that the 2x6 R23 w/ R8.4 foam I've recommended comes in around R24-R25, which is below the ~R30 starting point for zone 5 walls in Table 2.

I put this on Martin's blog on GBA " Residential Commisioning" HVAC reminded me of this thread.....perhaps someone can straighten me out

Everywhere you look now the trend and movement is to replace central forced air with single or multi point ductless minisplits. As you know, there is a risk in designing a minisplit/ventilation system as two separate systems compared to a well-known centralized forced air system with a lot of design guides and empirical data. Size and placement of the minisplit system, ventilation rates per ASHRAE 62.2 or BCS-01 or whatever is required. The combination of which can get pretty confusing. Just the ventilation debate alone has caused a lot of confusion.

Also a lot of confusion exist out there on minisplit sizing and placement based on the effects of envelope air sealing, insulation levels, moisture, material properties, air flow, room temps, duct CFM, and fan CFM. There are a lot of variables here that can change drastically between climate zones and building geometry.

I’m guessing typical load analysis by hand or math, or, modeling software like REScheck or REM are not going to work here, there are too many dynamics to determine how far a single source minisplit at some location can affect temperature differences in some number of rooms, whether it be a single story, basement, multi-story. WUFI or some combination of it and a loads model might be better, but complex.

The idea attempting to DIY integrate these system, hot/cold air handlers exhaust with HRV/ERV fresh air intake to perhaps a distribution plenum box, then to a distribution trunk line of some with spider branches of some diameters, to rooms may be a good way to commission comfort complaints of many clients a builder has to please. Spot exhaust the bathroom and kitchen. Problem is low CFM of the HRV and perhaps long trunk and duct branch runs and increase energy cost due to inefficient ducts. Makes one wonder if we are back to a conventional central air forced system in need of HRV.

Unless I’m missing something and there are minisplit-ventilation integrated systems on the market? Or, a better way to deal with rooms that after a build have too large of a temperature difference, or, let’s just say do not meet a customer’s expectation or what a builder may have predetermined inaccurately? Perhaps you could point to a simple software program and/or calculations a builder could use accurately that has proven to predetermine by field build(s) to satisfy a design model or analysis? Last thing I want to do is guess, use the wrong design guide(s), modeling, and find out when my walls are closed I need to run additional lines, ducts, air handlers, blowers, swap out the unit for a bigger one.

a risk in designing a minisplit/ventilation system as two separate systems

My understanding is that the low risk approach to ventilation is to design it as a separate system (supply and return), no matter what you do for heating and cooling.

Hi Dana1

The zip code is 80003

Thanks

Posted By jonr on 10 Apr 2014 03:34 PM

a risk in designing a minisplit/ventilation system as two separate systems

My understanding is that the low risk approach to ventilation is to design it as a separate system (supply and return), no matter what you do for heating and cooling.

If you were in two different buildings trying to accomplish heating/cooling in one and ventilation in another that reasoning would work. When you are in the same building integration exist and perhaps an opportunity to solve weakness in this case closed door room temp differentials are a common problem. Some say 2-3F degree is acceptable but how do you design to that? What model what method? If you do not hit that target how do you commission your design?

I've yet to hear of a mini-split that was integrated with a ventilation system. Mini-duct cassettes arguably could be, but there may be some design/operation conflicts there.  From a total design perspective it's usually better to separate the "V" out of the HVAC if you want it all to run optimally.

The minimum cfm rates & duty cycles required for ventilation are far too low to do a lot of heat transfer at delta-Ts that would be tolerable for a doored-off high-loss room.  But when rooms are nearly self-heating with one or two 250-300BTU/hr live humans, installing only HRV exhaust registers in doored off rooms and supplying the ventialtion air to that space via jump-ducts to the common area can be enough to make a difference.  But the key factor is is to keep the room load down- this approach is really only viable in high-R houses with low-U windows, and not a lot of glazed area in doored off spaces.

A crude napkin-math method of estimating the delta-T without the ventilation air help can be found in response #3 of this GBA thread which  is:

"To crudely estimate the approximate delta, assume the U-factor of a partition wall with half-inch gypsum on both sides to be about U0.4 BTU per degree-F per square foot. It'll be about U0.2 for a ceiling or floor that is fully conditioned on the other side.

Say a 13' x 12' bedroom with only attic above and crawlspace below has a heat load at the full 70F interior design temp of 2750 BTU/hr, and shares 25' of 9' tall wall+ door with a fully conditioned space. One sleeping human deliver about 250BTU/ reducing the load to about 2500 BTU/R.

That's 225 square feet of space, or 2500 / 225= 11 BTU per square foot of partition. At a U-factor of 0.4 for the partition that would impart a delta-T of about 11/0.4= 27.5F, not very acceptable.

 But if the unoccupied heat load for the room were reduced to 800 BTU/hr, the ~250BTU/hr of one sleeping human lowers that to 350 BTU/hr, or 550/225= 2 BTU per square foot on the partition wall, and 2/0.4= 5F, which in most cases WOULD be an acceptable delta at the 99% outside design condition."

If you want to fine-tune it with the ventilation air assuming a jump-duct supply, air has a specific heat of about 0.018 BTU per degree-F. so if you assign a cfm number to the ventilation air that you'll be designing in, and a maximum tolerable delta-T you can come up with how much heat that delivers. For instance, say you're delivering 20cfm by design, and will accept a 5F delta. That 20 cfm is 600 cubic fee per hour, and a heating capacity good for  600cfh x  5F x  0.018 BTU/ft³-F-degree= 108 BTU/hr.

That's not a huge fraction of the total load in the prior example, but it's still a double digit percentage.

Bottom line, unless it's a fairly low-load room you can't expect much heating out of the ventilation system unless you really crank up the rate or can accept much larger temperature differences.

Hi all took a little time to decide what i wanted to do here. I have decided to add 2" polyiso to the outside of the exterior sheathing, and am going with dense packed loose cellulose for the 2x6 wall cavity insulation which should give me around r-35 for the walls. Attic insulation will be bumped up to an r-49. I will seal all cracks in walls, floors, ceilings, etc. with spray foam and/or sealant before any insulation or drywall is put on. I had company do a heat load on the whole house and it came out to 26887 required for the whole house. They had my building construction as average, which i do not agree with as i think it is going to be a very tight house. Individual room breakdown is as follows: 4411, 4306, 3113, 15057. Since this thread has gotten a little long, a reminder of what im trying to acheive = a single minisplit heat pump in the main room, with backup heat in the three individual rooms. Backup/alternate heat will be provided by electric infrared heat coves, and the bathrooms small inwall cadet fan heaters.

Please let me know your thoughts.

Thanks!

The MSZ-FH15NA   delivers about 18KBTU/hour @ +5F, and about 14.5K @ -13F.

The AOU-15RLS2-H has comparable output, but a slightly lower nameplate efficiency.  (HSPF of 10.3 vs. 11.0 for the-FH15NA)

Either would be enough to heat the whole house at the mid-winter average heat load, and fully cover the 15K heat load of the main area.

With a few 1 kw - 1.3 kw resistance heaters for the doored off spaces you're there!

Dana1 can you give me a recommendation for the resistance heaters for the rooms??

Thanks!