GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Heat Loss Calculations and Minsplit Sizing

rshuman | Posted in Energy Efficiency and Durability on

I am about to build a new house and have been playing around with heat loss calculations to gain some perspective on the size of mini-split I need before I start asking vendors for their estimates. I was hoping people here could give a once over of my results and answer a few questions that I have.

The 26’x36′ 1.5 story house is located in midcoast Maine just west of Belfast. It will be constructed using 2×4 double stud walls, with 4″ separation between the walls. I plan to insulate using dense pack cellulose, I have estimated a whole wall R value of about 32 for the stackup. The windows will be triple pane with a whole window R value of ~5. The ~R50 vented roof will be built using 2×12 rafters, rockwool will be placed between the rafters and 1″ of polyiso will be attached inboard of the framing members. A shed dormer will occupy about two-thirds of the south side of the second floor and will have the same amount of insulation in the walls and roof. The house is being built on a full unfinished basement. It is assumed the basement walls are insulated to ~R20, a couple inches of foam (~R10) will be placed under the slab.

Excluding exfiltration losses, my spreadsheet calculations indicate a heat loss rate of ~18500 BTU/h. This loss is dominated by losses through the basement walls and slab, which account for close to half of the total. Assuming a mechanical ventilation rate of 60 cfm, exfiltration losses are estimated at about 4500 BTU/h.

A little more background then some questions. My main source of heat is going to be wood. I keep doors to upstairs bedrooms closed throughout the winter, warming those rooms with heat moving up from the first floor. My current plan is to install a ductless mini-split on the first floor which will serve as a back up. I am contemplating installing baseboard electric on the second floor to keep the insurance company happy  and provide heat upstairs if it is needed.

First, I would love to get a sense from others if my results seem to make sense. And,, assuming they do, what they may translate into the size of the mini-split I should consider. Second, as stated earlier, losses in the basement dominate my non-exfiltration loss calculations. Should these losses carry the same weight as the losses from the first and second floors given the unfinished (and unheated) nature of that level? After all, heat does tend to rise. Any insight will be appreciated.

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. kyle_r | | #1

    What delta T are you assuming across the below grade walls?

  2. rshuman | | #2

    This occurred to me as I was posting my questions. I used a delta T of 69 (70-1) for all calculations. A more realistic delta T would cut losses from the basement by about a third.

  3. kyle_r | | #3

    It depends on how much of your basement is below grade, but I would argue probably cut in half.

  4. kyle_r | | #4

    For new construction I would not put a ductless mini split in. I would put in a ducted mini split. With the size of your house you can probably get a register in each main floor room with very little duct work in the basement. In your climate you probably want a unit with a pan heater. I would look at Carrier. Something like this.!/product/26449

    If you don’t want to heat the basement than don’t and disregard accounting for that heat load. You could just assume you will lose heat through your main floor, assume basement is at 55-60F.

    Unheated your upstairs might be ~5 degrees colder assuming your home is well air sealed. AC might be the bigger concern. If there is a closet you can rob a few square feet from you might be able to put another ducted unit upstairs with a central hall way return. Some of the mini split air handlers can be mounted vertically (carrier, Fujitsu).

    1. paul_wiedefeld | | #5

      100% agree!!!! Why not make the house as balanced as can be if it’s all on paper at this point? Plus AC will work so much better.

  5. rshuman | | #6

    Thanks for the comments/ideas to date.

    I will look into a single ducted mini-split for the main floor. I think it makes sense to modify my heat loss calculations to limit the importance of the basement losses and to account for first floor losses to the unheated basement via the floor.

    I am not a big fan of a second floor unit as I really don't think it will (ever?) be used given my heating strategy and tendencies. As a result, it isn't clear why I would want to spend a few grand on it. AC is not a concern as I have literally never used it in any house I have owned and don't see that changing any time soon.

    Again, thanks.

    1. paul_wiedefeld | | #7

      I wouldn’t do a separate unit for the second floor unless the layout makes it difficult. Just 1 24k btu cold climate ducted heat pump with ducts to both floors. That wouldn’t add any more expense besides supply/return ducting. You’d get AC for free, so use it or not, it’s there.

      Separately, which wood stove will you use? I don’t know much about low load wood stoves, but am interested in how you plan to not bake this well insulated house.

  6. rshuman | | #8

    Point taken with ducting the unit to both floors.

    I'm not sure which stove I will be using yet. I am looking on the low end of the admittedly loose 'capacity ratings' (like, less than ~1500 ft2, knowing I am in a better insulated house and colder climate than considered for the rating) but it is largely guess work. That said, if I discover I have nevertheless oversized the thing I can always have occasional hot fires as the conditions dictate.

    1. paul_wiedefeld | | #9

      Yeah the ducted system is the easy part here - plenty of heat pumps can fit that load and you can avoid backup baseboards. Those with variable capacity will be extra efficient and quiet. The wood stove seems harder - a ~20k btu load stove would probably be toaster oven sized? Maybe burn 3 logs per day on design day?

  7. kyle_r | | #10

    There are a lot of threads on GBA talking about small wood stoves. I would look at Stuv and Pacific Energy. Is the home going to be all electric? With a tight house think about back drafting.

    Consider whether you need makeup air for the stove, dryer, range hood, or any other combustion appliances.

  8. rshuman | | #11

    Paul - Come on, at least two toaster ovens.

    Kyle - Yes, the house is going to be all electric. I will bring air from directly outside to the stove but I may still need to consider back drafting under specific conditions.

  9. Expert Member
    Akos | | #12

    I would read through the excellent build log from Scott Goodyear, specifically take a look at the heating system:

    If you really don't need AC, something similar can be done for a reasonable budget. The nice part with his design is the buffer capacity of the storage tank can absorb a fair bit of the extra BTU from the wood stove.

    The part I really like about the setup is that there are practically no electronic controls. Everything just works.

  10. johngfc | | #13

    Have you considered adding exterior insulation to the upper portion of your below-ground basement wall? There's usually a very strong temperature gradient from the surface down, so an extra 2 or 3 inches of exterior foam, particularly on the N side (assuming you get sun elsewhere) can make a large difference in energy requirements use. I'm in CZ6 and am going to put 2" of exterior foam on the top 4' of the below-ground portion of my basement, with ~ R20 on the interior (so ~ R30-35 in the part with exterior form). This should make a substantial difference in winter, particularly on the N side where the top couple feet of ground will be well below freezing ~ 4-5 months/yr.

  11. rshuman | | #14

    I went through my "Manual J" calculations again, being a little more aggressive (and hopefully) realistic. In the update, I used an outdoor design temperature of 1 deg F for the walls of the basement, first, and second floors. I used an outdoor design temperature of 40 deg F for the basement slab, assuming that is the (constant) soil temp at a depth of about 4' below ground surface. I did not use this temp for the basement walls because a portion of them are abovegrade and I decided not to complicate matters by using depth-dependent outdoor temps. I used an indoor design temp of 70 deg F for the first and second floors of the house and 50 deg F for the basement. Losses were explicitly modeled from the first floor to the unheated basement.

    Making these changes, my estimated heat loss decreases from about 18,500 BTU/h (excluding exfiltration losses) to about 13,400 BTU/h. Including losses due to exfiltration (which remained the same at about 4500 BTU/h) gives me an estimated total loss of about 17,900 BTU/h.

    The losses excluding exfiltration are dominated by losses through the basement walls which account for about a third of the total. A more accurate accounting of the temps of the media adjacent to the outside wall surfaces would lower these losses, as would adding more insulation to the walls. I assumed R-10 worth of foamboard on the interior surfaces of the walls, all insulation will be inboard of the foundation. I expect I will add at least another R-10 to the interior walls of the basement but I decided not to include that assumption in my calculations at this time.

    In any event, as stated earlier, my current calculations suggest a total heat loss rate of 17,900 BTU/h. I would be interested in hearing suggestions for suitable mini-splits to meet these conditions. Is it simply a case of matching up the output of the units at my outdoor design temp (1 deg F) to my heat loss rate? How much should I oversize the unit given the (aggressive?) approach I have taken to modeling losses? Earlier discussions included whether I should simply install one (ductless) unit on the first floor and let the rising heat take care of the second floor (my position going in) or installing a single ducted unit to explicitly heat the second floor. In my current modeling, losses from the second floor are about 25% of the estimated loss, excluding exfiltration. Does the load required for the second floor change anyone's opinions about using a ducted unit? As stated originally, I generally do not heat bedrooms, which occupy about two-thirds of the upstairs.

    Finally, I should probably repeat that my primary source of heat will be a wood stove. That doesn't mean I can 'cheat' on the size of the mini-split (it still needs to be capable of heating the house when the stove is not operating) but it may factor into things somehow.

    Thanks for any and all (additional) opinions, advice, suggestions, etc.

  12. kyle_r | | #15

    Fujitsu makes nice ducted units (it’s what I have), but they do not have pan heaters for 1:1 ducted units. I don’t think Mitsubishi makes one small enough for your load. Carrier (rebranded Midea) is the only small ducted mini split with a base pan heater I am aware of. The one I linked to in my original post is what I would recommend or depending on how aggressive you think your calculations are you could bump up to the 24k unit. Since you said you don’t need AC upstairs, I think you’ll be fine without heating it. Check with the building department first to make sure they don’t require it. However I would get quotes both ways, you can always block those registers off, but it will very costly to add them later.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |