Is this the place for a multisplit?
I’m deciding on a minisplit for a new build and having never done a specification before, I’d like to check my theory against the intuition of an experienced designer. Several articles on this site suggest that multiple, single-head mini splits operate more efficiently in practice than a multisplit. However, my project seems to me to be a good candidate for a single minisplit with two heads for the two zones in the project, because the zones have roughly equivalent heating loads. So my question boils down to this: are multiple, single-head minisplits always the best solution, or is this a case where a two-head multisplit really does make sense?
Building Information:
600sqft apartment with 200sqft loft (Zone A) above 600sqft conditioned garage / woodshop (Zone B), separated by R30 floor
Climate Zone 6b
~8000 Heating Degree Days
Design Temperature 7*F
Peak Heating Load for Zone A: 7,000 btu/hr
Peak Heating Load for Zone B: 6,000 btu/hr
I intend to keep Zone B at a lower temperature most of the time, perhaps 50*F.
Option 1: Two separate, single-head minisplits, likely sized at 9,000 btu/hr, one for each zone.
Option 2: A single 18 kbtu/hr condenser driving two 9,000 btu/hr heads, one for each zone.
My second question is if the pre-charged “DIY” units are worth consideration against established companies like Mitsubishi. On one hand, it’s hard to trust a brand titled “Mr. Cool”. On the other hand, the installed cost would be significantly lower since HVAC labor is extremely expensive in my area. And perhaps the pre-charged linesets and quick-connect fittings actually lead to a more reliable installation?
Any expert perspective on these two questions would be much appreciated!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Multi split is almost never the right answer. About the only time it is worth it is if you are tight on outdoor space.
Besides the lower efficiency, one issue with them is they are always sending refrigerant to indoor units when heating. The problem with your setup is that since you won't be heating the garage most of the time (or barely heating it), that head would still get refrigerant even when off, this will make the COP of heating upstairs much lower most of the time.
As for DIY, except for the ducted units, I don't think any of them are hyper heat. You want something with decent capacity near you design temp for this application. I've had good luck with Midea units, better value than most out there.
Where is the data that supports the statement that a multisplit is less efficient than two minisplits? I have always been under the impression that multisplits are more efficient than multiple minisplits.
single zone:
https://ashp.neep.org/#!/product/25822/7/25000/95/7500/0///0
multi:
https://ashp.neep.org/#!/product/25830/7/25000/95/7500/0///0
HSPF and SEER is better on the single zone. If you check other manufactures, it will usually be the same. There are some outliers where the multizone outdoor unit is much larger, so it has better heat exchanger, but overall this is the trend.
This is also nameplate values. The multi needs to be running with both zones heating at the same time to get that HSPF value. This rarely happens in the real world.
Thank you for the information. I did a little more reading and apparently multisplits are typically not more efficient than minisplits but it can depend on the design and the equipment but in general minisplits will be a little more efficient. Good to know.
An oversized multi split is a certain recipe for unhappiness.
No, you are not an exception I don’t think the lower cost of the multi splits are worth it, you lose controllability efficacy and redundancy.
Check the instruction many miny systems will not let one set the temp below 55°.
Understand there are only a few big assembly lines that make all the mini systems to the spec for that order and then the different labels get applied, so the quality is pretty much the same.
I installed a DIY MRCOOL it worked great the only real down side is what to do with the excess tubing. Move the unit or coil it up behind it.
Walta
I have a Mr. Cool minisplit DIY version and there is a FP (Freeze Protection) mode that sets it down to 46F. I don't know if they all come with that function or not though.
I should mention, the next setting up is 62F though, so it goes from 46F to 62F with nothing in between.
It was relatively easy to install. It's quiet. I think it works down to -13F. I would use these again since I'd have to pay double or more for an HVAC installer to install it.
While I do think there are good applications for multisplits, especially ones like the Mitsubishi Smart Multi that can close their zone valves further down and have better minimum outputs, this application (very different usage patterns and different solar gains) seems better for two single splits. If you told me the two spaces were connected by open stairs, and would be at the same setpoint all the time, then I'd say, sure, multi could work fine.
As for brands, Mitsubishi is great, but as you noted, very expensive, and I think in simple single split configurations there isn't a huge benefit. I have Mitsubishi on my current home, but it is also because it is the only source of heat for my home, and is a fairly complex system that benefitted from a sophisticated installer.
Gree or Midea both make units with very good low temp performance ("hyperheat", though I think that is a Mitsubishi trade term), and nearly all of the DIY options you'll see out there are rebranded Gree or Midea.
To narrow the options a bit, I'd recommend looking at the Pioneer Hyperformance (a rebranded Midea, I believe, but stocked out of Florida):
https://www.pioneerminisplit.com/collections/wyf/products/pioneer-hyperformance-9-000-btu-ductless-mini-split-inverter-wi-fi-enabled-air-conditioner-hyper-heat-pump-full-set-230v?variant=40534494412842
https://ashp.neep.org/#!/product/93205/7/25000/95/7500/0///0
Or Pioneer (Midea) Diamante Ultra: https://www.pioneerminisplit.com/collections/wyt/products/pioneer-diamante-ultra-9-000-btu-wi-fi-ductless-mini-split-inverter-air-conditioner-heat-pump-full-set-with-16-ft-kit-230v?_pos=2&_fid=e1b5bc541&_ss=c
https://ashp.neep.org/#!/product/64671/7/25000/95/7500/0///0
Or the Gree Sapphire: https://www.greecomfort.com/assets/our-products/sapphire/documents/sapphire-sales-brochure-a.pdf
https://ashp.neep.org/#!/product/25405/7/25000/95/7500/0///0
All have good performance at or below -13F, and hold good COPs. If you use the Advanced Sizing Calculator on the NEEP website you can input your zip code and see more detail on how the unit will run vs. outdoor temperatures.
As Walta mentioned, check the manuals to see how far down the setpoint will go. My Mitsubishis will only go down to 59F in normal heating mode, and a "smart set" feature that drops it to 50F, but this varies a lot between companies and models.
EDIT: Forgot to mention, if you plan your rack well, you can stack the two units on top of each other, and run one electrical circuit with a small subpanel to the location, with one service outlet and two breakers. That makes the installation difficulty virtually the same as running a multisplit.
Thanks everyone, this is extremely helpful perspective!
I'm getting a clear message that a multisplit isn't the way to go, and I trust everyone's perspective. Yet from a theoretical perspective, I'm still finding it hard to dismiss multisplits. In my build, for example, I calculated that even with a lower indoor setpoint of 50* vs 68*, Zone B would still have a significant heating load of about 60% of that of Zone A under a typical winter temperature of 20*F. My intuition was that a multisplit would be able to modulate these two heat flows. (This imbalance between zone loads becomes greater at more mild temperatures - say 50* when Zone B would need nothing but Zone A would still need perhaps 1500btu/hr - but these mild temperatures represent a small fraction of the total annual energy consumption of the system.) Is there any data, or design guidelines, on how closely balanced zones need to be for a multisplit to perform efficiently? The fact that so many multisplit systems exist, and the simplicity of a single condenser when it can easily meet the total needs of the system, make it counterintuitive to me that multiple, single head systems are considered best practice.
Sam, thanks for the specific unit recommendations! The units you linked have much better COPs at the relevant temperatures than the similarly priced systems I was looking at, so that makes it an easy decision.
Thanks to everyone who took the time to chime in!
Jimmy
I'm not aware of any detailed studies on the matter - there are so many models and different home configurations that it is hard to know where to dig in. However, I'd say that while you are probably right that during low temperature conditions things will probably be fine, in the sense that the outdoor unit will be operating close to its maximum, and the indoor units will be modulating within the upper 50% of their ranges, and things will work well.
The problem is actually more in the shoulder months. Because the multisplit outdoor units have much lower minimums, you can have the situation where one of the indoor units needs to run, but the heat demand is higher than the minimum of the outdoor unit, so the outdoor unit will cycle considerably. This isn't great for efficiency, but it can affect comfort too, because the rooms can fluctuate temperature more than when modulating, and one room can find itself overheating from the bleedover refrigerant circulating in the system.
You aren't in a big cooling area, but this can be even worse in cooling situations, the unit minimum is too high, so it can't get enough runtime to dehumidify well.
This is all to say that they can still work, and sometimes they have to work, but you both pay more for one with better modulation like the Mitsubishi Smart Multis, and/or live with some of the design compromises.