GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Audio Play Icon Headphones Icon Plus Icon Minus Icon Check Icon Print Icon Picture icon Single Arrow Icon Double Arrow Icon Hamburger Icon TV Icon Close Icon Sorted Hamburger/Search Icon

Community and Q&A

I need help choosing the correct minisplit and heat pump for my studio

DANIEL GOTTSEGEN | Posted in Mechanicals on

I have two proposals to install a single unit mini-split with an outdoor condenser for my studio building. The building has a 24 x 32′ foot print. The walls are 10′ high, with a cathedral ceiling that is 16′ high at the ridge. Walls are 2 x 6 (at least) with fiberglass batts. Roof rafters are 2 x10″ (I think) with fiberglass batts between, and then either 1 or 2″ rigid foam insulation skimming it (under the sheetrock). Building is tight. Floor also has fiberglass and rigid foam board between the 2 x 12″ joists.

The proposals recommend either the Mitsubishi MSZ (and MUZ) FE 18NA, which is the largest unit, or the MSZ (and MUZ) FH 15 NA (which is the largest unit in the FH series). The latter is a smaller unit obviously, but it has the three stage filtration, and one contractor says the FH is more “efficient” or something like that than the FE series. I want to be sure the unit is sized correctly. Just heard back from the guys who proposed the FH… now they say use the FE… guess I got my answer.
The building is an art studio (painting) and I do some carpentry in it (mostly not) so there can be some dust generation.

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. Stephen Sheehy | | #1

    Daniel: before the experts chime in, they'll probably want to know what part of the planet your studio is in, whether anyone calculated the design load, whether you need heat, A/C or both, and whether a blower door test has been done to determine how tight the building is.

  2. Expert Member
    Dana Dorsett | | #2

    The zip code (for 99% outside design temperature estimation) and the total window area/type would be important considerations before even taking a WAG at what the heating load might be.

    It's true that the -FH15 delivers about 15-20% more heat per kwh of power used than the FE18, testing at an HSPF of 12 rather than 10.3.

    But it's also true that the FE18 can deliver more BTU/hr at any temperature- it's simply a slightly bigger (but not as efficient) unit. That's 21,600 BTU/hr @ +5F instead of 18,000 BTU/hr @ +5F.

    So it matters what your 99% design temp is, and the heat load at that temp. If your heat load is 23,000 BTU/hr you'll probably still do OK with the FE18, but not the FH15. If your heat load is even 20,000 BTU/hr you're probably better off running with the FH15, despite being undersized.

    If you don't have a lot of window area and the 99% outside design temp is +15F or higher it's likely that even the FH15 is oversized, and maybe the -FH12 would do. A 24 x 32 footprint is only 768 square feet of conditioned space with a fairly simple geometry for less overall exterior surface. Even 2x4/R13 with R21 in the roof can come in under 12,000 BTU/hr @ +5F in an 800' building if the window area is fairly modest (and the windows are pretty tight). The FH12 puts out about 13,600 BTU/hr @ +5F. (Even fairly complex shaped homes in my neighborhood where the outside design temp is +5F regularly come in under 15 BTU/foot of conditioned space with only R20 2x6 construction even without the foam sheathing in the walls.)


    Thanks for the great responses (though some of it is way over my head... I will read it through carefully). I am in rural Vermont. It gets cold here!!! Last winter with the polar vortex, we had weeks of subzero temps, even getting down to -15 (and perhaps lower). There are not many windows on the building, because I need blank wall space to paint. Three 3 x 6' on the west gable end, and a couple of others. The windows and doors are tight and insulated glass (I know because I built the building with my own hands).The building is tight. I use little AC, maybe just a few days in the summer, so heating is the real (and heavy) demand.

  4. Expert Member
    Dana Dorsett | | #4

    I don't care one bit what the absolute lows were during the polar vortex event- it's irrelevant from a heat pump sizing point of view. Give us your location, from which we can estimate the 99th percentile temperature bin. If it gets down to -15F or lower at 1-2x/year over the past 20 years that might be a relevant number.

    The 99th percentile temperature bin is the temperature at which only 2190 hours out of all hours over the past 25 years were colder than that. See if there is a nearby city listed in the ACCA Manual-J compilation of 99% & 1% outside design temps:

    Since you are the one who built the place, what type of foam, and how thick? What is the stud spacing? What is the rafter spacing and insulation depth to that roof fiberglass? Is the insulated glass clear double-panes, or are there some low-E coatings?

    Then, what is the lowest interior temperature you would tolerate overnight in that building?

  5. Jerry Liebler | | #5

    I'd go with the larger unit! If you look closely at the specs you'll find that the larger unit's efficiency is a bit lower at full output BUT it is the same at the same output and can throttle down to the same lower limit as the smaller cousins. The extra capacity is there if you need it and the upfront cost of the upgrade is rather small. This is one case where an "oversized" unit does not result in lower efficiency, that is the real beauty of variable refrigerant flow and inverter technology.

  6. Expert Member
    Dana Dorsett | | #6

    Jerry- I suspect the FE18 is actually ~2x oversized for the actual load, and that the -FH12 is probably the "right" size. The FE 18 does not have an infinite turn-down ratio, but only about 4:1 between max & min modulation. If it's something 2x oversized it would spend more of the season cycling on/off, and only modulate during the coldest weather.

    The FH15 is slightly better at about a 4.6:1 turn down ratio, and would modulate quite a bit more, due to both a lower oversizing factor and a higher turn-down ratio.

    While efficiency at part-load is outstanding, cycling on/off all the time exacts a significant toll on average efficiency. Oversizing by more than 50% should be avoided whenever possible.

    The -FH12 has a turn down ratio of 5.7:1, so even if it's a bit oversized it should run nearly continuously.

    The blind walls w/ 2x6 framing with R19 batts and 1" of interior side polyiso under the sheet rock has a U-factor of about 0.053 BTU per square foot per degree F. If the windows are all in the gables above, that's 1120 square feet of wall. Assuming a 99% outside design temp of -5F and an interior design temp of +65F (assuming it's OK to let it get that cold overnight), that's a design delta-T of 70F. So the wall losses would be on the order of

    1120' x 70F x U 0.053= 4155 BTU/hr.

    Assuming you have about 25 square feet of U-0.5 door and about 100 square feet of U0.5 window (WAG, based on the descriptors), the window & door losses are about:

    125' x 70F x U0.5= 4375 BTU/hr

    Assuming R30 batts in 24" o.c. rafters and about 1000 square feet of roof (not a high pitched roof), that's a U-factor of about U0.04, so the roof losses run:

    1000' x 70F x U0.04= 2800

    That adds up to only 11,330 BTU/hr, which is about half the 21,600 BTU/hr capacity of -FE18 at +5F, and even the 18,000 BTU/hr capacity of the FH15 @ +5F would still be more than 50% oversized. At -5F they will have only about 85% of their rated capacity at +5F, but it's still a hefty oversizing factor. (85% of 18,000 BTU/hr is about 15,000 BTU/hr, to cover a presumed load of about 11,000 BTU/hr. making it fully 36% oversized, but with 4.6:1 oversizing the FH15 would modulate most of the season.

    If the actual 99% design bin is north of 0F it'll be substantially more oversized though and possibly running into cycling losses. Without more detail about ACTUAL location & ACTUAL construction who really knows? I suspect none of the contractors submitting proposals did even this rudimentary level of heat load calculation.

    With that information we'd be able to compare the load at the average mid-winter outdoor temp to estimate how much on/off cycling it would incur, but whnn in doubt, err to the small side, shoot for covering the AVERAGE mid-winter load rather than the 99% design load, even if you have to wear sweaters in there during the polar vortex events.


    HI Dana and Jerry, Thank you so much for taking this on. Much of your calculation and information is a bit over my head, but I will read it carefully. Dana, let me try to answer your questions as follows:
    I went to the document you provided a link to for the 99% figure. The closest is Barre, Vermont or Montpelier/ Barre (they are essentially the same). Montpelier is a bit north of me, a teeny bit colder perhaps, but also I am a bit higher, so it is closest. Burlington, which is farther north, is actually milder because of its proximity to Lake Champlain.

    My building is tight, but the roof is likely under insulated, I realize. I have probably 8' fiberglass batts in the roof cathedral ceiling, skimmed with 2' of rigid foam. I am not sure if the rafters are 16" oc or 12" oc (likely 16). Walls are 2 x 6' studs, 16" oc, with fiberglass batts. Floors are 12" joists (building is slightly off the ground on piers) likely 12" oc with 8" fiberglass and 1 or 2" rigid foam between the joists. There are 5 windows that are 32 x 60", double paned, w/o a low E coating, I think, and I don't believe they are argon filled. They are all toward one end of the building (3 on a gable end) that gets the most sun. The rest of the building has no windows to provide blank walls for painting. There is a double door, that is tight and has double paned, insulated glass, near the sunniest part of the building.

    There are two ceiling fans in the cathedral ridge. The building is currently heated with wall mounted LP heaters. As it is a work studio, I keep it at around 55 degrees when I am not working, heating it to around 68 degrees when I am working, which is five days a week, mostly daylight hours. I don't run the LP heaters much at all from June - October. We get a couple of days each summer hot enough that I run a window air conditioner. That is rare though, because I am at 1800' with broad exposure and cooling breezes.

    Does that give you the information you need to know? if not, I am happy to answer any other questions. One HVAC contractor who does a lot of these installs, spec'ed out the FE 18NA. The other one (from a different company) put down the FH unit (he put down the 18NA, but there is no such thing). His HVAC guy corrected it to the FE 18NA. but he still wondered if the FH 15 unit would be better. Thus my question.

  8. Expert Member
    Dana Dorsett | | #8

    So, your 99% outside design temp is really about -8F/-10F or so. The lowest setting most mini-splits can be programmed for is 60F (check the manuals- they're online.), so that's still about a 70F delta-T.

    You probably have R25 batts in the ceiling, and assuming the foam 2" of EPS (~R8) covers the rafter edges as well, the U-factor for the the roof is about 0.035.

    How high is the peak of the ridge compared to the average wall height?

    You have about 67' of U0.5 window.

    Assuming you have 2" solid wood doors they too will come in around U0.5. If they are panel doors with windows or insulated steel doors, let me know- it makes a difference. Don't be shy about mentioning storm doors either, eh?

    It's hard to gauge just how much ventilation you get under the piers, but we can treat it as if it's fully exposed, with no skirting to slow down the wind. Is the foam cut'n'cobbled between the floor joists, or does it cover the joist edges? And is it 1" or 2"?

    How thick is foam do you have on the walls and what type? (Drill a hole & measure it if you can't remember, then seal it with can-foam.)


    The door has large glass windows on both (it is a double door) to let in light. The glass is double paned, and I think it is low E... I will check. I believe I spent for good quality on the door because of all the glass, quality rubber gaskets, good glass. No storm door though. The rigid foam on the celling does cover the rafters, I think, and I will double check on its thickness. Walls are ten feet high. Ridge of cathedral ceiling is 16' high and runs the long length of the building (so 6' from top of walls to the ridge). There is no foam on the walls, just batts. One side of the building (south) is about 3' above grade at the bottom edge. The opposite is almost at grade, as the ground slopes up.There is a fair amount of stuff stored under there too. I had not put any skirting around the building. In the dead of winter here, there is no wind underneath usually, because of the snow depth. In fact it is an interesting thing that we all notice our houses warm up a bit when the snow is deep, as it insulated the base of the buildings. I had 4 feet of snow in front of my house through March (in my memory) last year. December can be somewhat more "open", less snow, though. Between the floor joists, I put Ix2" near the bottom edge as a kind of stop. Rigid foam board (continuous) sits on this to seal the bottom, then an 8" batt sits atop that filing the rest of the void. I don't know if the foam is 1 or 2" in the floor. I will get back to you with the foam depth in the ceiling. Thanks, Dana.

  10. Expert Member
    Dana Dorsett | | #10

    A double door is typically 45-50 square feet, lets call it 47'. If it's a solid wood door with double-paned lights it's pretty much a ~U0.5 door. So with 68' of U 0.5 window area that's 115' of total U0.5 surface. At a delta-T of 70F your windows & door losses are:

    115' x 70F x U0.5= 4025 BTU/hr

    With R19 batts in 16" o.c. 2x6 framing and NO-foam you're looking at a U-factor of about U0.07 BTU/hr per square foot. Adding in the gable area of ~144', and subtracting out the 115' of windows & doors you're at ~1150' of U0.07 wall, for wall losses of:

    1150' x 70F x U0.07= 5635 /BTU/hr.

    With a 6' rise on the gable that brings the total roof area to about 860'. Using the U-0.35 estimated in my prior post brings your roof losses to:

    860' x 70F x U0.35= 2107 BTU/htr

    I don't quite fully understand floor insulation, but it sounds like the foam is between the joists, and not covering the bottom edges of the joists? If that's the case you're looking at a U-factor of about U0.04, but probably at a delta-T of only 55F, between the stratification of the air indoors, and the warming effect of the ground. With 768' of floor area that makes the floor losses something like:

    768' x 55F x U0.04= 1690 BTU/hr

    Add it all up and you're looking at a heat load of about 13,457 BTU/hr @ -10F, which is about what the MSZ-FH15NA deliver at that temp.

    You could add a fudge-factor for the air infiltration losses, but since the place is backstopped with the propane heaters there isn't much point to that. At the COP 1.8-2 efficiency it runs @ -10F the per-BTU cost of the propane burners covering the shortfall aren't much more expensive than what the mini-split is delivering. If you look at data for either Barre or Montpelier, the mean January low temp is +8F, which would be only a 60F delta-T from your 68F fully-happy occupied temperarture:!dashboard;a=USA/VT/Barre

    At a 60F delta-T that heat load is only ~11,500 BTU/hr, and even if you added a 30% fudge factor for air infiltration losses, it would be within the 18,000 BTU/hr capacity the FH15 has at +8F outdoor temps.

    The mean overall January temps is about +17F. which would be an average delta-T of about 50F, yielding an average load of about 9300 BTU/hr, about half the output capacity of the -FH15 @ +17F, so it should modulate quite well delivering very good efficiency well into the shoulder seasons.

    Thus, methinks the FE18 is going to be overkill, but the FH15 probably isn't (or at least not such overill that it matters.)

    Note: When it drops below -18F outside these units may automatically stop until it rises to -13F, as noted in the double-asterisk fine print on page 2 the test submittal pages.

    So don't yank out the propane burners unless you install something else as the "Hail Mary" backstop to the FH15.

  11. DANIEL GOTTSEGEN | | #11

    Thanks, Dana. Your assumption about the floor joists was correct. I enjoyed reading your post, and I actually understood it enough that it was very clear to me. Thank you so much again for taking the time to help me with this.


Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |