GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X 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

Sizing a Single-Zone Minisplit

matt2021 | Posted in Energy Efficiency and Durability on

I know I am abusing folks’ expertise and patience with my questions related to my porch-to-four-season conversion, and I apologize for that.  Yet, determining the correct size of the minisplit that will be added to the room seems to go beyond my abilities.  I tried to use various tools, including one recommended, in a past discussion, by Dana Dorsett; yet, I am still uncertain between 12,000 BTUs and 15,000 BTUs as the correct load calculation.  My problems, I think, originate from not knowing how to best characterize the level of insulation of my new room’s walls, ceiling, and windows.   So, let me summarize what seem to me the relevant details, and ask what the correct load/minisplit sizing is, or I should calculate that myself.

The single-zone equipment I have been looking at is Mitsubishi’s 12,000 BTU Hyper-Heat system; specifically: MSZ-FS12NA [indoor unit] and MUZ-FS12NA [outdoor unit]

Here are the room’s details (though many on this forum might already have them):

Zone 5

Room’s dimensions: 23’x14.5′; lowest wall is 7′, highest wall 10.5′

R values: ceiling (R38 between the rafters plus R5 or R6.5 polyiso board under the rafters); exterior walls (R6 Zip System plus R21 in the cavities of the top walls — that is, above the windows; R6 Zip System plus R15 in the cavities of the knee wall); floor (R35); windows (R3.8) and patio door (R3.7).

The windows and patio door combined, it turns out, are 60% of the wall surface (which I know is a lot, and indeed more than what I had originally calculated and shared, here, in a different post).

The room will be connected to the rest of the house (which has its own, properly sized––and not oversized––forced air system) by a six-foot door opening.

The three sides of the room are approximately North-West-South, in an area that is partially shaded by tall trees.

Thanks to some helpful advice I received on this forum, I am investigating the possibility of adding storm windows.  Down the line, I will also certainly add cellular shades, hopefully with tracks, to the windows.  Yet, none of that (and maybe even some of the details I offered above) should be considered in calculating the load, and in sizing the system.

When trying to use the tool to calculate the load, most often I landed in the neighborhood of 12,000 BTUs; yet, when indicating that my walls are not very well insulated (given the large % of wall that is occupied by glass, it’s been my thinking), the numbers seemed to lean more towards 15,000 BTUs.  For whatever is worth it, my hunch is that the above-mentioned Mitsubishi system would be adequate; yet, a certain fear of it being undersized does remain.

Thanks to anyone who will be able to shed some light on this!

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.

Replies

  1. onslow | | #1

    matt2021,

    I ran a quick check on the described space and R values. At a design point of 0F and a delta T of 65F I am seeing a total heat loss value of a hair over 6,000 btu/hr. I would be curious to see what parameters are pre-embedded in the program you are using. Your floor, roof and remaining wall area comprise about 1900 btu and the window/slider is about 4300 btu. I am ignoring the shared house wall as it would be neutral to the load. I de-rated the R values to R35 roof, R30 floor and R18 wall. I gave the windows the R3.8.

    Even going to a delta T of 70F only adds about 600 btu, which still won't get you close to 12k. Perhaps the program builds in a truly horrendous air exchange rate. In any case, these numbers are your maximum projected heat load, so most of the year it will be a much smaller number. Daytime gain will further reduce the demand during the shoulder seasons, though daytime gain will increase your cooling needs.

    I am seeing your biggest problem being as finding a small enough unit that can dial back output and still provide the cooling you will need. How close is the current heating system to actual demand? You would likely see lots of short cycling even with the 12k unit.

    As a check, from your description I see 333.5 sf floor, 345 sf roof, 415 sf wall minus 60% for 166 sf in play and 249 sf of window/slider. Are you including overhang area by chance? Also, I am curious about the different batt assumptions above and below the windows. Sketches would help.

    1. matt2021 | | #4

      Onslow, Thank you very much! Some responses and comments below:

      - What you say further demonstrates my ignorance in the matter. I have no idea how loads for walls, etc. are calculated. Nor do I know whether priority should be given to loads for heating or, instead, to loads for AC; I suspect that in my case it's the latter. If so, then maybe a VARIABLE unit like the one I indicated is precisely the way to go, don't you think? From the specs, it seems that, at 47 F, it has a minimum capacity of 3,700 BTU/h (for cooling, 2500 BTU/h at 95 F). Is that the meaning of the specs? That is, can I conclude that a unit like this, while rated to give a lot of BTUs if need be, can also properly perform at lower levels? (I am attaching the specs sheet in case you want to look at it.)

      - The online tool I was referring to is found here: https://hvac.betterbuiltnw.com/. I should mentioned that, in other discussions, Dana Dorsett himself warns against taking online tools too literally; they are certainly not a replacement for a professional report.

      - "Are you including overhang area by chance?"

      No.

      - "Also, I am curious about the different batt assumptions above and below the windows"

      The current plan is for the walls above the windows to have three inches of closed-cell spray foam (R-21), and for the kneewall under the windows to have R-15 mineral wool. In both cases, to those values, the R-6 provided by the Zip System R6 must be added.

      Sorry for not having an easy way to produce sketches right now. I hope the above descriptions help. Thanks again!

      1. nynick | | #10

        I'm going to take the time to do this HVAC sizing tool. Thanks

  2. sayn3ver | | #2

    Not apples to apples, my house is 758 sqft. R15 batts in the walls. R38 batts between ceiling joists and r15 batts ran perpendicular across the joists in the unconditioned attic.

    House sits due south along the long side (30x25)

    Have been heatng and cooling with a single 15k Mitsubishi hyper heat in zone 4a for 5 years now.

    Has kept the house plenty warm in the winter.

    Summertime it keeps up but is at it's designed manual j limit cause on 98+ days it'll hold 75f in the house during the day and work it's way back down to 70f at night.

    I'm personally adding a 6k unit in each bedroom just so we can close the doors in the next year. We have a kid now and while the single zone has worked for our needs, the ability to close a door at night and have retain comfort has taken priority.

    The advice I was given on here, regarding my own scenario years ago was spot on.

    I was in a similar scenario with calculations putting me on the edge. These units need to run a long time for any meaningful dehumidification.

    This site is an true resource.

    1. matt2021 | | #5

      Thank you, Sayn3ver! Even if it's not apples to apples, the comparison you draw is still helpful to me.

      I very much agree about the side being a great resource.

  3. nynick | | #3

    I will be building a very similar room in about a year. The architect has already designed the house. Our room will be 12x24 with a flat ceiling, knee walls and one slider. Since it will be new construction we can pay close attention to air sealing....I hope.
    For estimating purposes, the GC had his HVAC sub do a heat load calculation (which I have yet to see). I specified this room should probably be on its on Zone, since it has such a large expanse of glass and can be separated from the main house by a 6 foot door, just like yours.
    The estimate came back with an 18K unit for this room alone, which I think is crazy. Plus, he specified another 24K for the first floor and another 18K for the second floor. All Mitsubishis. Total sf for the house is 2500sf.

    I've got an upcoming 'fight' on my hands.

    Make sure you calculate your floor losses. Good luck.

    1. matt2021 | | #6

      Nynick, I am not a professional but even to me those numbers seem off.

  4. walta100 | | #7

    Let’s think about the cooling load when the August sun is pouring thru that glass.

    Walta

    1. matt2021 | | #8

      Thanks Walta. Indeed, more than once, regarding similar situations, I've read the advice "go by your cooling (vs. heating) load." The problem I am having is on how to calculate such load. I am not sure I am able to use the online tool correctly, or correctly understand its results.

    2. nynick | | #9

      Agree. I'm specify low SGHC windows just for this reason.

      1. matt2021 | | #11

        I myself went for the low e coating that reduces solar heat gain (Marvin calls it Low E3). On the windows, the SHGC will be .23 (with a U factor of .26). The next level up (which was not proposed to me by the vendor -- and after all my "studies" I failed to inquire about) would have been Low ERS: its SHGC would .22 (so, no big gain there?), but the U factor would have been a more attractive .22. I wonder whether that would have been cost prohibitive for me; but the point is moot now, and I already placed the order.

  5. onslow | | #12

    To Matt first and Saynever second,

    The simplest formula to get btu/hr over a wall, roof, floor or any area for that matter is;

    (Total square feet of area) x (total degree difference between outside and inside) divided by the R value of the wall, roof, floor, window, etc. in question. The result is btu/hr.

    It is a bit more common to use the - (U value)x Area x delta T = btu/hr

    It gets into a much more extended discussion if I explain why I de-rated your apparent R values.

    Glad to have it, just not now.

    I went to the BuildBetterNW site and declined involvement once I read the use agreement. Not sure what secret info they have to protect, but I don't want to get cross wise with them if making observations about the software output. If you consistently came up with 12k load, there are two choices. Your inputs are at fault or their software is.

    Re-run the surface areas yourself to check my inputs. I guessed the shed roof pitch is why your end walls are 10.5' on one side and 7' on the other. Use the outside faces of surfaces because that is where the heat leaves last. Use the rough opening dimension for windows and doors.

    Walta is correct that you will most likely be facing higher cooling load than heating load. The description of the head capacity shows a max cooling of 13,000+ btu, so without a detailed sun load review I can only say it is probably fine. Any tree shade that protects the windows will be to your favor. The real question mark is heating and the units minimum output of 3700 btu.

    Run your heat loss at a delta T of 40 or 45 and you will see your loss rate is right there at minimum unit capacity. It represents 30-25F outside. For CZ 5 you probably see this temperature range much more than 0F. The unit you have chosen will tend to short cycle a bit spring and fall when ambient is 50F. Come summer time, it will reverse the behavior, load peaking in July or August when it hits 90F or worse. Don't forget to plan for the head drain.

    More later.

    Saynever,

    I don't think you will be happy putting 6k heads in the bedrooms. Not sure if you are expanding head count on the existing outdoor unit or installing a second outdoor unit. If you are currently keeping the munchkin happy now, but want to close the door at night, simple resistance heat from cove heaters is much less costly and directly responsive to the rooms needs. For answers on the AC side of the equation, I recommend posting a new question. People more on the ball than I, can explain how the head count mates with the outdoor unit to divide capacity. Maybe a ducted unit can feed both more successfully IF you can fit it in.

    1. sayn3ver | | #13

      I didn't mean to hijack the OP. I would be installing two 6k single zones for essentially night time use in the summer. I have no desire to install a ducted unit in my unconditioned attic. I'd have to see specs on the air handler to see if it would physically fit in the 14.5" wide access hatch to my attic and or my crawlspace.

      I understood the challenges going in initially in trying a single source heating/cooling solution.

      I got my 608 tech cert last winter when I was laid off. I do commercial electrical work as my job and already own all the tools now after having to fix my contractors installation of my original mini. The installation labor only costs me my time.

      Worst case is I run one or two of the 6k's for my swing season load and shut off the 15k. I could use a whole house dehumidifier for the swing seasons to be perfectly frank. I know it isn't a perfect solution but those single zone 6k's modulate down pretty low (down to 1700 btu for cooling).

      I already have an electric resistive fan unit in each bedroom as a backup source (which we needed to utilize when the mini split failed mid winter two years ago due to the initial contractor's poor installation).

      I treat them for emergency use which I don't think is discussed enough regarding air source heat pumps.

      1. matt2021 | | #15

        Sayn3ver,

        No worries at all, about highjacking the post. I am very glad that my post could be helpful you to you as well. For certain questions especially, when I post them, I am hoping they be of relevance or interest to others!

    2. matt2021 | | #14

      Onslow, thank you SO much! This is so educational. I’ll run all the numbers according to the formula you gave me.

      As for your derating the insulation, my guesses are: thermal bridging (which I’m trying to address) or insulation losing its value overtime (I know foam does that; mineral wool might be the most stable, right?); or both of course. Am I on the right track?

Log in or create an account to post an answer.

Community

Recent Questions and Replies

  • |
  • |
  • |
  • |