Minisplits in Extreme Cold
I’m still trying to figure out if for me a mini split system will be best in my situation. I live in se Wisconsin and we had about 3 weeks below zero this year. Not constantly below zero but it’s been extremely cold.
Figured I’d check with others on how there systems faired.
Also I notice there’s seems to be a lot of people saying not to use multi head mini splits in there systems. And I have been trying to search for information that describes this issue but haven’t been very successful. In my case the mast is complete opposite side of the house would definitely need its own heat and the rest of the house living room kitchen wide open but two more bedrooms down a hallway may be hard to get them to heat or cool. Typically in our area all the hvac guys recommend using a Mitsubishi down til about 13 then switching over to either electric base board or radiant heat etc.
Appreciate any opinions on this.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Hi Wiscoguy,
I’m in Milwaukee and getting a 18k btu Mitsubishi minisplit (single head) with hyperheat installed to an upstairs bedroom for heating and cooling duties. I won’t really know about the very cold winter performance until next winter, but I actually expect to improve comfort — the furnace doesn’t reach the bedroom in winter with heat, due to the distance, and I expect I can turn down the gas furnace themostat even more in winter and heat the rest of the house less.
For distributing heating and cooling better, I’m looking into converting that bedroom’s door to a transom window style frame, but sizing the “window” to fit a standard window fan with two motors. I know this won’t be as much air movement as a furnace blower, but I think it’ll help with drying out the upstairs in summer (no central AC, I’ve relied on window AC units for years.)
I can report back in a year, but that probably won’t help you make a decision now.
I've been through three winters now. It never gets super cold here, but certainly a lot of time in the sub 13F range, a little less frequently below 0F, and rare excursions to as low as -10. The Fujitsu heat pump has never failed to pump out heat. So the idea that you'd need to switch to other heating at 13F is wrong. It would have been true a couple of decades ago, but the cold climate heat pumps of today are different. Just make sure you get a cold climate model. That's not to say that you shouldn't also have back up heat, because you definitely should.
Plenty of units out there that are way more efficient than baseboard or fossil fuel heat even at -22. As long as the refrigerant part was done properly these units will deliver heat what they are rated at.
This 18k Carrier can still deliver a COP of 2.85 at 5 degrees and 2.11 st -22. Those are effective efficiencies of 285% and 211% at 5 and -22. Even at -22 its still more than twice as efficient as any fossil fuel source of heat.
https://ashp.neep.org/#!/product/26721
Thanks for this is carrier a good brand for mini splits I was always told Mitsubishi was the best
PHIUS uses a 2.8x de-rate on grid electricity to account for the efficiency of the gird... so your 2.85 COP is about 100% efficient with source efficiency- still very good, but not 2x.
The issue is often the capacity, not the efficiency with heat pumps in cold weather. Even the hyperheat models start losing compacity at ~5°F, so just make sure you have enough for your 99% ASHRAE design day, it may require upsizing the unit.
Or backup heat for those exceptional days. Just as long as you aren’t running resistance baseboard heaters all winter.
What if your on net metered solar? Right now my panels are directly powering both my mini splits and still sending almost all their power out to the grid. They would achieve those efficiencies correct? What about at night when I am using my credited power I sent to the grid?
its a difficult question. The way PHIUS seems to handle it is to only give credit to solar USED onsite... i.e. no credit for net metering. Mostly because your net metering credit is only financial, not actual. There is no battery on the grid storing YOUR energy, and you are not providing the same value (economic or environmental) to the grid at noon when everyone has excess solar available as when you draw down those credits at 5pm. The peaker plants are 20-35% efficient and that's what your drawing from at 7am and 5pm.
So you always get that efficiency, but is that input energy coming from your solar or a coal plant depends on time of day.
I live in the far northeast corner of Iowa in Decorah. I heat 2/3 of my house with a Fujitsu 36RLXFZH connected to 2 12k floor mounts and a 15k wall mount. Our lowest temp a few weeks back was -24 or so, and we had several days where it struggled to get above zero. I also have a two zone forced air furnace for backup, and heated the remaining 1/3 of the house with it. 1800 square feet of the building is an 1850s solid brick house with no wall insulation (solid masonry walls, plastered right to the inside course) and the other 1100 square feet is a new addition we added last year. We also supplement with a Jotul 3 wood burner, but that's mostly in the evenings.
The heat pump worked well and was able to mostly keep up. I did notice that the high fan speed setting was throttled beginning at about -15 degrees. I suspect that since the unit isn't putting out very hot air at that point, the fan speed is lowered so that what is passed through the unit comes out hot?
In terms of energy use, the unit used about 1400kwh in January and 1600kwh in February. Not bad at all, especially given all the talk of how much electricity other multis use. When the temp is say 15 it'll consume anywhere between 25 and 50kwh a day, and on the coldest of cold days we were averaging about 100kwh/day.
I have a 9.5kw solar array and very favorable net metering. I went into the winter with 5500kwh banked and burned through that by about the 10th of February. My February electric bill was $140. For March my production and consumption will be about equal, and from there on out it's all down hill. $200 in annual heating for a 2900 square foot house- not bad.
I was going to ask about that with multi heads not being as efficient. My new house is kind of cut up due to lots size three different bedrooms. I was think g a unit with four separate heads to run everything in my house but I keep hearing how these units aren’t very efficient. Plus there would end up being some long runs to the outdoor unit. Appreciate any thoughts on this.
Shockingly I can give a short report from Austin TX on Mitsu performance down to 5°F. From Austin TX. I don't have energy use or efficiency data (why don't the units report this through their apps? So lame that they don't). But the hyper heat units we have purred along (we were lucky enough not to lose power). Now we had to set the thermostat to 68° per request of the grid operator, but we were warm throughout.
I snapped a pic of > 8" of snow under our units, justifying my idiosyncratic request for the Quicksling stands for our outside units :) Had we not had those stands, I'm convinced we would have had problems. We had some icing in the pan, but never were all the drainage holes filled. The units defrosted without issue, my impression is that they did so less often that in corner season humidity.
Judging by online posts, heat pumps were freezing up all over Austin, lacking pan heaters, going to AUX heat strips at the worst time for the grid. I'm not sure how the mitsu non hyper heat units would have performed.
Of course we still had over-heating in some rooms due to the multi-split bleed over (I never had "thermal off" turned on on some of the heads), and that is pretty frustrating.
Of course I was happy to have these working, but I did still deeply regret covering the chimney, given that neighborhoods nearby were without power for 5 days and we would have had zero backup heat if our power went out.
I've been using a non-hyper heat unit with a base pan heater for heating a studio for a couple of years now. No issues with the cold except the expected drop in output on those 5F days. Since it is sized to to heat with the reduced output, kept the place nice and comfortable at 70F without issues. Measured around COP of 2 (hard to get exact air flow on these, so big error there but ballpark) on those cold days, still way better than a resistance heater. Well worth the money.
Same here, non-hyper heat unit but no base pan heater in Zone 6. 15 deg 99% design temp. Lots of days of 10-20 degrees, average temperature has been 32 for months. Minisplits didn't have any issues - except cleaning the ice build up once.
That being said, Electricity prices have been up, Gas prices have been down. So the balance point has shifted to run more gas this year. But that's probably a fluke. My differential natural gas is 76 cents, electricity 13 cents - so I did end up shifting more to my legacy gas system than I expected by the end of the winter. (I believe that's $1.85 therm for COP=2 on a mini split.)