Factors to consider heating with one ductless head per floor
Are there guidelines on how well a ductless head will perform based on square footage of area trying to cover? It seems like there is a lot more to it then heating load of the entire floor. Type of layout, volume of space, cfm of blower, thermostat location, ability of heat to go through doorways, oversizing of head, etc. Assuming insulated basement foundation walls, are the square footage limits on first floor living areas heated by one ductless head?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
I can say that 1000sqft of well insulated open space works no problem. The area by windows is cooler, best is a bit of extra heat there (resistance floor heat or baseboard).
You can heat larger and less open spaces just need a bit of additional electric heat in rooms around the perimeter. In this case the minisplit would still be doing the bulk of heating with doors open.
Cooling is a different story, as soon as you have rooms, it is better to go with a ducted unit so that each space can get the right amount of cooling.
I have a 12K serving around 1150 sf on the first floor. It's undersized for the load with an uninsulated, 50% above grade basement below. If I got that insulated I could get pretty close to my design load for my 1st floor. A small addition in the back might need supplemental heat.
I've had pretty good luck with cooling- My second floor 12k will keep 2000sf cool by itself outside of a multiday heatwave, when we'll turn on additional units. Bedroom doors need to stay open and we'll run ceiling fans.
Keeping the bedroom doors open is not an option for every household, this is why a ducted unit for the upstairs area is always better. It is more cost to put in but will be much better comfort.
The upstairs unit provides very little of the house heat, you can save a bit of money by not going with a hyper heat unit for it.
In BSC's study of MSHPs in Transformations' houses in Western MA, we found problem cases when we were going over ~1100 sf per head. This is not meant to be a hard-and-fast rule, but just an observation. Of course, cutting up the floor plans, open/closed doors, etc. make a huge difference.
Minisplit Heat Pumps: Lessons from the Field
Exactly what I was looking for, thank you. A tough comparison as these were all superinsulated homes, but informative to see the pain points. Square footage to wall area ratio seems like it could be a good general rule of thumb. Looks like all of these homes were sub 900sf/head as well.
Is "thermal bouyancy" in the 2 floor samples something about distribution? Because our upstairs is always warm.
3 12k head units in 2 floor, 1 upstairs and 2 downstairs so <900 sq ft per head unit. But the 2nd floor is always warmer than the 1st floor ...
The report talks about redistribution systems...
I am using 2 12k units each heating 900 sqft. My house is a 1800sq ft 1958 split level with significant air sealing and insulation. Everything is wide open between floors so I get excellent heat transfer.
I'm using a 9k minisplit for about 1500 square feet on the main floor, and it's also providing about 80% of the heat for the 1000 square feet second floor. Open floor plan except for the bathroom and utility room. Those rooms are slightly cooler on cold days, but not enough for me to do anything about. The only thing I wish I had was a remote thermostat, but it was about $500 extra for the part alone. Bottom line is that for me, with a mostly open floor plan, just using the heat loss calculation worked out fine.
That's impressive! I imagine this is a new home build and not retrofit?
Yes, a super insulated house. The temperature throughout the house is pretty uniform most of the time irrespective of the heat pump, which is most likely attributable to Passivhaus standard windows, low infiltration and high insulation.
I think the bigger issue with a single head is air leakage more than insulation.
Insulation helps with keeping temperature difference low but unless you have a very tight building, infiltration will make it uncomfortable.
Typically houses leak near the perimeter at ground level, all that cold air will just tend to stratify and settle there. A central mini split can do some mixing of the air (floor mount probably better on this front), but I don't think it can do enough to overcome this.
So to make the single head work, your really need to focus on your air leaks.
I agree with Akos.
I have single head on serving about 1200sqft. About half of which is a single open area. It does fine capacity wise, but there is more temperature difference than I would like. The outside corners of my 1950's bungalow feel less comfortable than the did with the furnace and central ductwork. If I crank up the fan to even out the temperatures, the air movement feels uncomfortable. Makes me appreciate ductwork more.
If I could do it again I'd put in a bit of ductwork to move the supply to the perimeter, since my unfinished basement would've made a couple of 6inch runs no problem.
If you want remote temperature sensing on the cheap, buy some wire, remove the existing thermistor and cut the wire in half. Splice in a long length, make some kind of custom enclosure for the thermistor that is vented top and bottom to the room, purely for aesthetic purposes. That's pretty much what I did, after a misadventure with a PAC-SE41TS-E Mitsubishi part that was supposed to be compatible with the minisplit I have. When I measured the resistance of the new thermistor, it was all wrong. So I just removed the old one from the minisplit and replaced the one in the new wall mounted casing with it.
The resistance is a little off, probably due to something I did wrong with the splice or the fact that I added 40' to what was a 6 inch long wire, electrical isn't my strong suit yet. But I just adjust the thermostat a couple degrees and call it a day. Maybe I'll hunt down the problem someday...when I have time...
You can also just get the kumo unit for the head, and a connected stat for the room. Less than $200 total.
I wasn't familiar with kumo before looking it up. It appears to just be an app, which doesn't do anything to address the issue. A connected thermostat, which you kind of gloss over as a trivial matter, is what is needed and it's what is labour intensive or expensive.
You can get wireless, battery operated temperature & humidity sensors that connect to the kumo system. It's not clear to me whether they override the internal thermistor or not. But they provide some kind of feedback anyway...I'll have to install one someday and find out.
That may worth with a Mistubishi, but it definitely won't work with my Fujitsu.
Sort of... Kumo is an app, but also a system. You can get a ~$300 remote thermostat; you can also get a <200 contoller and wireless room thermostat; both allow control of the unit from the Kumo app, but more importantly allows the temp to be read in the room, not above the unit, which smooths out the temperature and runs more efficiently.
I seem to remember some discussion of a similar system for Fujitsu, but I could be wrong.
Yes, this is on my list of things to do eventually. It's low on the priority list because I've still got an unfinished bathroom and countless other projects, not to mention the fact it will void the warranty I still have on the install.