# Minimum capacity of multisplits

| Posted in Mechanicals on

I have a question about what’s important to look at when comparing single splits to multisplits

I have read through most articles on this and what it sounds like is that the key to minimum energy consumption for cold weather heat pumps is to look for systems with as low as possible *minimum capacity* @ 47F, the argument being that most of the time we are in the shoulder season and the heatpump will be asked to do very little.

So one should focus on the min capacity at around 47F, as once it gets down to 17F/5F etc. the heatpump will not be running at minimum capacity anymore, so then its all about the COP at those temps and the *rated capacity*.

If anything the extreme temps can be “ignored” as that only happens a few days a year and its ok to just supplement with resistive heating with COP of 1.

I’m seeing the lowest minimum capacity at 47F for most 12kbtu (small singlesplits) is about 4kbtu, while its about 50% of max/rated capacity for most multisplits. i.e. 24k for a 48k multisplit.

I was under the impression that single splits can ramp down MUCH lower? as in closer to 1:10?

Another thing is, multisplit heads can be 6k each, while singlesplits are usually minimum 12k (some 9). So for multiple bedrooms (assume we HAVE to have heads in each room), wouldn’t it be preferable to have 6k heads connected to a multisplit instead of 12k heads each to its own compressor?

Gree has an interesting one – its a “24k” (multi) unit that can ramp down to 7k at 47F, but goes back up to minimum 18k when it gets cold. Isn’t this exactly what we want? I can’t find any mitsu multisplits that can go this low at 47F

https://ashp.neep.org/#!/product/30252/7/25000/95/7500/0///0

## Join the leading community of building science experts

### Replies

1. | | #1

The heads of a multisplit might be smaller (depends on brand) but the modulation of the outdoor portion of the multisplit divided by number of heads is often larger than the minimum of any one head.

The 6kbtu Mitsubishi heads turn down to 1.6kbtu.

1. | | #3

whats the lowest that a 36k 3 zone system can turn down to in comparison?

2. | | #9

I don't think people realize how low some of these single zones can turn down. That Mitsubishi will run at an incredible 100-150 watts at minimum heating/cooling. Even if it short cycled the electricity consumption would be neglible.

I verified with an Emporia Vue all the minimum electricity consumption listed on the NEEP website for the MUZ-FS06NA. It's pretty incredible watching the Emporia and seeing it using only like 130 watts while pumping out decent heat at min capacity.

https://ashp.neep.org/#!/product/34425/7/25000/95/7500/0///0

Edit- Meant to reply to Paul above.

2. Deleted | | #2

Deleted

3. Expert Member
| | #5

There is simply not enough information out there how these units turn down and work. This makes it very frustrating to figure out what works.

On some, the min modulation changes with the number of heads connected, you can start with something with decent modulation but by the time you add 4 heads, it might only have a 2:1 turndown.

It is also not clear how the min capacity is split between heads. On my multi split, I find that the heads run pretty much on/off and the outdoor unit ramps up and down to match how many are calling for heat.

There is also refrigerant bypass in heating between all indoor units so a unit off is still getting some heat. Depending on heat load, this bypass heat is enough to overheat something like a bedroom. Lot of bypass also decreases efficiency, the COP of the units running with a single head tends to be lower then all.

Overall, I would stay away from a multi split unless it you are running something like two large ducted air handlers. Even there, go for one to one as they will work better and provide some backup for about the same BOM cost.

4. | | #6

Good multisplit will do 1/3 of rated power - not btu. So if unit is rated at 6 kw it will turn down to 2kw consumption. With all the bypasses and various size heads trying to figure out turn down in btu is a lost cause

5. | | #7

Hi, my recent observations on a Fujitsu AOYG24kbta3 leads to this:
In heating, the units will blow at a minimum temperature of 100°F. The minimum air flow of indoor units is quite high compared to some Mitsubishi units for example:
9-12k wall units: 340m3/h
9-12k slim-duct: 450m3/h
9k-12k floor unit: 270m3/h
The outdoor unit can modulate very low, the limit comes from the indoor uni: the minimum heating power is determined by minimum air flow times temperature delta.
On other brands like Mitsubishi, I can see even lower air-flows on indoor units (below 200m3/h in silent mode), the limitation could then come from the outdoor unit. However, the "auto" fan speed mode seems to select a low-medium speed most of the time, meaning that it might be necessary to manually select a "quiet" speed to match the load to the production in some rooms, and to change this setting as the load increases. In that mode however, heat tends to stay near the ceilling with wall units...
To my view the turn down ratio is connected to two other things:
- comfort: on wall units, cycling leads to lower comfort because the ambiant air temperature can swing a lot, I find it less a problem with ducted systems. So with wall units it is more desirable to have a continuous heating.
- efficiency and COP: cycling comes with a penalty. On my multisplit, when indoor units have a low duty cycle, units will work one at a time very often and the efficiency seems significantly lower in this case.
Therefore a multi-split system should be able to work with many units running simultaneously most of the time for the best efficiency and comfort. Having one wall unit in each room will likely not meet this criterion.

6. | | #8

I'm convinced that single-splits are the way to go.

But I'm noticing that Manual J isn't useful for sizing minisplits at all because it only gives the loads for the extremes. E.g. my design temp is 7F, so for a (large) space that needs 18k, that's at 7F, which is only useful for if I need to size the equipment to be able to keep temp at that extreme. But it seems minisplits are designed to work with supplemental heat (for the extreme days) and we really want to size for the rated capacity at the "nominal" winter temp, which in my case is about 20F. But my manual J does not have a heating load @ 20F.

Do I need to do a manual J using 20F as the design temp, and then match that load with the rated capacity of a minisplit at 20F? (or 17F on NEEP)

I originally started looking at matching the MAX capacity at 5F to my manual J load number, but then realized that I just need enough baseboard heaters to make up the difference for the few days a year that's 5F outside. So whats the point in looking at the manual J @5F at all!?

1. | | #10

Manual J is just a number, how you design your system around that is up to you. In practice, if you're going with multiple single-split units, you are very likely oversized in total capacity anyways - however, that doesn't mean you are necessarily oversized from a practical perspective. Single-zone splits have superior turndown (up to 10-1) and you can additionally turn entire condensers/systems off, leading to even better turndown. The additional outdoor heat exchanger area also works in your favor as you'll likely be operating at a more efficient COP than rated capacity.

Given all of that, I think it's a good idea to design your system to be completely sufficient on splits alone. Even if that isn't important to you, maybe the next homeowner will appreciate it and not decide to rip it all out for a fossil system because "heat pumps don't keep up".

• |
• |
• |
• |