Conflicting minisplit sizing
![](https://secure.gravatar.com/avatar/02310ead9754b7a8d0b0764b7088f976?s=52&d=https%3A%2F%2Fimages.greenbuildingadvisor.com%2Fapp%2Favatars%2F200x200%2Fm.png&r=g)
Hello. Can anyone help me get the correct size for my mini split? I live in NE Pennsylvania. Between talking to Mitsubishi, Home Depot , a Mitsubishi contractor and a searching online, Ive gotten answers from 6k to 18k units. Everyone here seems to be more knowledgeable than my previous contacts. My basement has an insulated slab with 2” XPS underneath it and 1” xps on the slab perimeter. The floor is exposed concrete. Half of the room ceiling is 9’ insulated and other half is 9’ exposed. Entire basement is underneath conditioned space. North and south walls are fully in the ground w 3” polyiso against the wall. North and south rim joists are also insulated with the same. Those two walls are 22’ each. The east and west walls also have 3” polyiso and are 26’. There is an uninsulated room on either side of the east and west wall. They both have an insulated steel door on each wall. If someone can point me in the right direction, that’d be much appreciated. Thanks. -Matt
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
![](https://www.greenbuildingadvisor.com/app/themes/greenbuildingadvisor/dist/img/modules/detail-library/thumb1.png)
![](https://www.greenbuildingadvisor.com/app/themes/greenbuildingadvisor/dist/img/modules/detail-library/thumb2.png)
Replies
A big, important unknown is air infiltration, from both stack effect and/or high winds. For heating, 9K (max output at low temp, not nominal size) is a good guess for never being cold. Much less for AC.
One option to get you into the ballpark is to plug in a couple of space heaters into a set of cheap power meters or wifi plugs and check what it takes to keep the space heated on a cold night. Based on the specs, my guess is a 6k unit is already over sized for the space.
But even if you scale it up to account for the difference between the cold sample night and seasonal lows, it still won't be accurate wrt to windy day air infiltration. For example, try this on the basement using the "loose" ACH value and 0F outside. Unlikely in a basement, but air infiltration alone could exceed 6k. Add about 3K (yes, I ran the numbers - using two exposed walls) for conduction losses and that's 9K (requiring about a 6K nominal HP).
Are you asking about just the basement?
Yes
Maybe I don’t even need anything besides a dehumidifier down there in the summer? Maybe a small NG vented heater to get it to 70 in winter?
If I understand this correctly, this is a 22' x 26 (= 572 square foot) BASEMENT, with minimal window area, and possibly NO door area, with R15+ walls and R8+ under the slab?
Unless you're holding polka parties with a dozen 200 lb Czechs dancing with kegs on one shoulder and need something bigger for cooling, the Mitsubishi -FH06NA is more than adequate (and probably crazy oversized) for the (nearly-nonexistent) heating load. In fact, with the upstairs heated to 70F it's unlikely the basement would ever drop below 65F even without active heating.
How much total window area is there? Are they low-E double panes?
Is there a gas water heater or furnace or boiler in the basement?
>"Maybe I don’t even need anything besides a dehumidifier down there in the summer?
A heat pump water heater would cover most of the dehumidification needs unless it's leaking quite a bit of outdoor air.
>"Maybe a small NG vented heater to get it to 70 in winter?""
Even the smallest NG heaters would be sub-optimally oversized.
As a ball park heat load estimate, sort of worst-casing it assume there is 3' of above grade exposure on the foundation + bandjoist on all 4 sides (even though it's only 2), for about 300 square feet of U0.05-ish wall, and call it 30 square feet of U0.5 ish window, and an outside design temp of 0F, indoor design temp of 70F (a 70F difference.)
300' x U0.05 x 70F= 1050 BTU/hr
30 x U0.5 x 70F= 1050 BTU/hr.
That's 2100 BTU/hr plus infiltration, plus another smaller number for the below grade wall & slab losses. Even doubling it to account for those other losses comes in at 4200 BTU/hr. The Mitsu FH06 puts out nearly twice that at 0F. If a window were left open it might call for an FH09.
>"There is an uninsulated room on either side of the east and west wall. They both have an insulated steel door on each wall."
To get any closer to reality takes a better description of the east & west walls and the spaces exterior to those walls, the total amount of window area, and the actual above-grade exposures, etc. But it surely won't take anywhere near the output of a 1-ton mini-split, and the half tonner is probably overkill.
The partition walls are insulated (not just the doors), and those rooms are also under conditioned space?
If there's a gas water heater in the basement it's standby losses will be providing some of that heat, but for the cost of about 20' - 25' of fin-tube baseboard, an expansion tank, a potable-rated plate type heat exchanger and a couple of small circulator pumps , a zone relay and a thermostat one could set up a fairly conventional looking hydronic heating solution that would cover the load (with margin), sipping heat out of the water heater.
For a lot less upfront money 1500 watts of electric resistance heater (baseboards, cove heaters or panel radiators) would cover the calculated heat load.
Dana, how many polka spins per minute are you estimating to come to those results?
That was the best example of heat load I've read on this site so far 😂
Lol. I love it. There aren’t any windows in the “finished” portion. The room to the east is a small mechanical room with a forced hot air furnace and on demand hot water heater. The west room is storage with two tiny egress windows. The entire basement is below my conditioned house. The main trunk of my HVAC runs right down the middle of the finished part. The entire basement is just a big rectangle. So, the finished part is just one room.