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

Choosing a Minisplit System

aleaf | Posted in Mechanicals on
Hi, I’m wondering if anyone can help provide some direction here. We have a 1,450 square ft, two story home in Madison, WI built in 1953, which has no cooling or ductwork. Currently the house is heated by original hydronic baseboards and a 15 year old, high efficiency gas boiler. We’d like to add cooling to the house and ideally, electric heat that would allow us to drop our gas service. We actually like the silent, well-distributed heat of the baseboards, so an ideal heating solution would be a higher temperature heat pump to replace the boiler. Unfortunately (and correct me if I’m wrong), it looks like this isn’t currently a practical option. So as a possible solution we are looking at minisplits.
The house is 770 sqft on the first floor; 680 on the second, with a small, unfinished attic above and unfinished basement below. Design heating load for the house (at -6F; based on utility bills) is 29,000 btu/hr. A blower door test yielded 1457 CFM50, but this included the unfinished basement which has since had several leaks sealed. A manual J calculation using CoolCalc with the blower door results, and default assumptions for the home construction period (except for the windows, which are new), yields design heating loads of 16,400 btu/hr for the first floor and 14,100 btu/hr for the second floor. The layout of the house is mostly open, except each floor has a smaller bedroom that would have the door shut at night. Design loads for the two bedrooms are 3,500 and 2,500 btu/hr.
Given this configuration and a budget of less than $15k (ideally closer to $10k, although we’re willing to pay more for a lasting solution that works well in terms of efficiency and comfort), how would you approach this problem? Some alternatives I am considering, and potential pitfalls based on what I have read:
1) A 4 head multi-split system:
Having separate heating zones in the open areas AND the two isolated bedrooms would be great in principal, but from everything I have read (for example, here on GBA), almost no one who seems to know what they’re talking about recommends multi-splits, mostly because of an inability to modulate down to meet lower loads, both at the house level and in small bedrooms like the two here. One of the contractors I spoke with proposed a Mitsubishi MXZ-4C36NAH for $16k. According to the NEEP database, the minimum output of this system is 22,500 btu/hr, meaning it would be cycling on and off for 70% of our typical winter heating load. A Daikin Aurora 4MXL36TVJU, which has a minimum heating output of 6,600 btu/hr at 47F, might be a better fit, but still prone to the issue of grossly oversized heads in the bedrooms (I believe the minimum head size is 7k). How bad is it really to have oversized heads that cycle in bedrooms? The HVAC people I’ve talked to (who are granted, trying to sell something) all claim that customers love these systems and they get very few complaints. How is this cycling any different in terms of comfort from a conventional forced air system that cycles (and is more or less comfortable)? The COP for the MXZ-4C36NAH at its minimum output ranges from 4-6 at the temperatures/loads under which it would be cycling for our house. How much would cycling be expected to decrease average COP?
2) Two independent single-zone systems:
For example, a Mitsubishi MSZ-FS15NA on the first floor and a MSZ-FH12NA on the second floor. This would solve the modulation problem and be a simpler (hopefully cheaper) installation that would also provide some redundancy. My concern with this setup, however, is comfort in those isolated bedrooms. Maybe they’d be fine with the doors open (maybe not), but I’m guessing with the doors closed they would need additional space heat to be comfortable. We could easily install some kind of resistance heater in each, but if the resistance heat ends up carrying most of the load for the bedrooms, would the overall heating efficiency for the house be any better (or worse) than a multi-split system that was cycling? Say we achieve an average COP of 3 with the mini-splits but the resistance heat makes up 10-20% of the load; this would knock the overall COP for the house down to 2.16 to 2.5. I could also imagine the bedrooms getting stuffy in a summer cooling situation, but maybe this wouldn’t be so bad if the whole house was dehumidified.
I know people use short-run ducts with mini-splits to balance loads among small rooms, but this seems like it would be expensive, as it would require adding ductwork to the basement and attic, which would also probably impact efficiency, even if the ducts were insulated. None of the contractors who’ve seen the house seem to think this is a good idea (and it’s also not clear that they ever do this for minisplit installs, so how can I be confident that this would be done “right”?).
Am I missing anything? Thanks in advance for your input!

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. yesimon | | #1

    What's the floor plan like? Are bedrooms in corners? Do you have more bedrooms? Does heat float up to the second floor? I wouldn't worry too much about cycling if you're within range of your fuel-use heat loss calculation.

    I would lean towards 2 systems. One for downstairs open area, and one for bedrooms + upstairs - probably multizone unless you can get a deal on concealed duct. As long has you have 4 spaces in your electrical panel.

    Since cost a big consideration - I strongly recommend to start with installing a single DIY unit for the open downstairs/upstairs, while still keeping baseboards. Most downstairs heat leaks "upstairs" anyways. That will cover replace a large fraction of your fuel consumption without breaking the bank.

    1. aleaf | | #2

      Hi yesimon, thanks for the reply. Attached is a crude drawing of the floor plan (sorry, I'm a total noob at Sketchup). After reading some more articles on this site I'm also leaning towards 2 systems with one head on each floor. The downstairs (15k btuh) head would be in the open space that includes the dining/living/kitchen area. Some of this heat would leak upstairs through the open stairway, which would be a good thing. The only concern would be getting heat to Bedroom 3. Upstairs, the 12k btuh head would be in the large master bedroom, positioned to hopefully blow air across the hall into Bedroom 2 and the bathroom. Minimizing the temperature gradient to bedroom 2 would also be a concern. I'm thinking we could see how it goes and install inexpensive wall heaters (or even radiant panels if it wasn't too difficult) to make up any heating deficit in bedrooms 2 and 3. The doors to these would be open for at least half the day, so that would also help. Since we also want this system to cool, I think we need a head on each floor. I've priced it out and it seems like the market rate for two single-zone systems around here is about $12k. Not cheap, but within our budget.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |