Mini split sizing

Hello,

I have received a couple of different quotes and suggestions for size so I thought I would look for advice on here.

I have a 250 sq foot room, 18x14. It is a bonus room that is connected to our house and has two openings into the house through 5 foot wide french doors. The doors are usually open. The roof is a flat roof and the 18 foot wall has three 43"x60" single pane windows and a 5 foot wide french door which is 70% glass.

The room also faces the west. The room starts to gradually receive direct sun at 2:00 and probably has full sun hitting the windows by 3:30 or so. I also live in Florida.

I am not worried about heating of course, just cooling. Load calcs that I have found calls for around 12K btu worse case. This is when I plug in poor insulation, single pane windows, west facing, etc.

I am looking at either a 12K but or 18K. No one lives in the room, it is where we have the dog crates, ping pong table, etc. I'm trying to avoid the unit running at full capacity a lot but also don't want to oversize.

I appreciate any feedback or more expert load calculations. Like I said, I received recommendations on both sizes in the bids.

 

Go with the 12k unit if its has Inverter Technology

 

Is there a reputable link someone can send me for a room load calc?

 

If you're aggressive enough in your air-leakage and R-value assumptions, the freebie online Co0lCalc Manual-J tool does a reasonable job. LoadCalc is more dumbed down and more likely to oversize by 30%(sometimes more), but it's also not terrible. Run both tools, take the one that gives the SMALLER number, and don't up-size it from there.

It's key with either of them to be aggressive rather than conservative on your R-value and other guesses- even when done perfectly per the Manual-J instructions, taking into account any and all factors that might lower the load number, it'll still have some margin.

MANY 3/4 ton mini-split heat pumps have 11-12,000 BTU/hr of cooling capacity at maximum speed, even though it's "rated" (= tested for efficiency) at a modulation level of 9000 BTU/hr. For a room that size, even with the flat (presumably insulated?) roof and west facing single pane slider it's still more likely to fit the capacity of a 3/4 tonner. With modulating systems for best comfort it's important to keep track of the minimum modulated output rather than focusing solely on the maximum. If oversized it'll cycle on/off more, with bigger variations in room temp & air movement. When right sized with a low minimum output it should run nearly continuously.

It's highly unlikely to have a cooling load anywhere near 18K, and the minimum output of an 18K mini-split might even exceed the average (not 1% design condition) load of the room. Modulation isn't infinite, and if it only modulates for couple hours out of the day you're giving up both efficiency and comfort.

If you punt on the load numbers and go with a 3/4 tonner that doesn't quite keep up when the sun is beating through the west facing slider, adding heat rejecting window film is a reasonably cheap & easy DIY for cutting the contribution from that glass in half. Most load tools will give you a breakdown on the contributions from windows/roof/walls/floors separately.

 

Thanks. I did try cool calc but had trouble for some reason. I got one more quote and they are initially saying an 18k but are going to look at it more. They were concerned about the 18 foot wall covered with 60% of glass. A calc I found online on a manufacturer website showed 12k as the max load. The 12k unit I looked at ramps up to 13k max

 

This proposed unit may help cool the area outside of those open doors too. Does that area need some cooling help? You could close the doors, and think about that.

 

It does not need to cool the outside areas. I would rather it not so the house air runs as it should. I want the bonus room air to just blend in with the house.

 

my concern is that if it cools the outside rooms too much it could cause my central unit to run less causing bedrooms to get warm

 

I did a calc on cool calc and it is surprisingly only showing 6500 btu for a 250 sq foot room, with large west facing windows and a flat room. The software does not ask you for the square footage of the windows, just low, medium or high and I selected high. I even added a make believe south facing wall with windows and the btu calc was 7500. I'm a bit surprised. Does that sound right?

If I make it a living room with 2 occupants it bumped it to 9000. I think I had it as an "other" room with 1 occupant. It is calling for 415 cfm and 9000k btu. How many cfm's are in one ton, I thought it was 400? If I get a 12K btu unit will that be at least 415 cfm?

 

The manufacturer-based calculators regularly overshoot by 50%, sometimes more.

What was the minimum output of the 12KBTU/hr unit you looked at?

A load of 6500 BTU/hr for a 250 room is a ratio of less than 500' per ton. The whole house average is likely to be closer to 1500' per ton, so it's not really surprising at all, even if not totally accurate if the glass area & type wasn't accurately specified.

There is no hard & fast cfm per ton rule. For high SEER ducted central air 500cfm or more is often necessary to meet it's SEER numbers. When tuning a system for a lower sensible heat ratio (= better dehumidification) dialing it back to 350 cfm per ton is usually legit, but a lot lower than that can run into evaporator coil icing issues. With mini-splits there is even less of a correlation between cfm & capacity.

Most 1-ton ductless units don't put out more than 400 cfm at max speed, even when delivering north of 13,000BTU/hr of cooling. Mini-splits usually modulate the compressor speed, condenser fan speed, and the indoor head's blower speed to max out the effiency, and to avoid icing up the evaporator coil in the head.

 

I had read that it is not healthy for a mini split compressor to not run at or near full capacity for a certain amount of time each day. Is that correct?

I am looking at having an 18k dual zone installed but one of the heads may not ever hit full capacity and the other should come close, but combined they might only need 15k. One of the heads won’t run but 3-4 hours a day also.

I am thinking 12k for the 250 sq foot bonus room with lots of glass and 9k for a bedroom. The compressor ramps down to 7k. I don’t think it will run at 18k much if at all. Probably only need 13-15k mostly.

is that harmful to the compressor?

 

Nope. If it weren't "healthy" for the compressor to run it at max speed they would have to shorten the warranty period. It's less EFFICIENT to run at max speed compared to the middle-speed sweet spot, but it's still (usually) more efficient than always cycling on/off due to oversizing.

TERRIBLE IDEA!!

The modulation range and turn down ratios of multi-zone compressors are smaller than single zone units, and as a rule the minimum modulated output of the compressor is higher- it's usually the MAX output of a half-ton head. A typical multi-split has a turn down ratio of 1:2.5 to 1:3, not more. A typical 18K unit's minimum output is about 6-8KBTU/hr, which is the calculated 1% design load. That means it will pretty much never modulate with load, and will almost always be running a duty cycle on/off/repeat.

The cooling load of a typical bedroom is usually less than 4KBTU/hr. It's a rare bedroom that actually needs a 9K head to stay cool, though some single-zone 9K mini-splits can modulate to under 2K, which is a reasonable fit. (eg: Mitsubishi MSZ/MUZ FH09NA that can drop back to ~1600 BTU/hr, or LG's LAN/LAU 090HYV3 that throttles back to nearly 1000 BTU/hr at high efficiency. Almost no bedrooms need the 7K min-output of the multi-split you were looking at unless it has floor to ceiling "sunset view" windows for a west wall.

It's not harmful to the compressor to never hit max speed, or to cycle on/off at minimum speed, but it reduces both the comfort and efficiency. Unless you spring for a wall-remote, mini-split heads sense the room temp from the intake air on the head itself, and when it's cycling on/off the temperature swings in the room are much higher. Spinning up both the head and the compressor from zero several times an hour takes more energy than idling along at low to mid speed for that hour.

 

Another possible consideration is that if you are on a realtime power plan, you could either have your AC cut off automatically if you signed up for that at a big savings, or if you manually cut off your AC when the 5-minute realtime price spikes really high. In that case, quicker recovery would have some value.

I am on a non-automatic realtime plan. No big spikes this summer, but there were some last summer.

 

The unit I was looking at has a smallest indoor of 9000. I think it said it can ramp down to 3000 to 4000. The bedroom does have three big windows facing the west but right now I have some reflective insulation on the outside which is blocking the sun. I need to get air in there and would rather not do a window unit.

is there a recommendation on my best options? The other room is 250 ft.² with big windows facing the west also. For that I feel pretty comfortable with a 1 ton. It has a flat roof and the 18 foot wall is mostly glass. So I was thinking the 9000 in the bedroom and the 12,000 in the other room

 

I read the data again on the dual zone compressor and it ramps down to 4200. The bedroom is 150 sq feet with big west facing windows although I currently have them covered with reflectix insulation.

the bigger room will not use the mini split at night. If only the bedroom is drawing from the compressor at night and the compressor ramps down to 4200 would that work and be somewhat efficient and healthy for the unit?

 

What does LoadCalc come up with for cooling loads for those rooms?

How many square feet of west facing window? Is it clear glass (no low-E), single pane, double pane?

Many 2-zone multi-splits will support a 6-7K head, but even that's crazy-oversized for a 150 square foot room with less than 100 square feet of unshaded west facing window.

For yuks I just ran a quick LoadCalc for a 10' x 15'= 150 square foot corner room in Orlando with 10' ceilings, slab on grade, with the long axis facing west, assuming R11 in the walls, R25 in the attic, no slab edge, no overhangs or shading factors, clear glass double-panes w/ no shades. With 100 square feet of west facing window (basically that huge 10' x 10' floor to ceiling "sunset view" ), with two people, and it came in at 9603 BTU/hr, 7300 BTU/hr of which was glazing load. Dial the window size back to 50 square feet (adding the 50 square feet back to the wall size) and it comes up with 5835 BTU/hr. Mind you, this tool regularly oversizes by 30% or more compared to what a professionally done aggressive Manual-J delivers. Reality (if there is such a thing as "reality" for a fictitious room ) is probably under 5000 BTU/hr in that scenario, even with the monster-sized picture window.

With a retrofit using DIY window-film the solar gains can can be cut by about half without significantly impacting daylight. (I recently installed some Gila Titanium Heat Control on the storm windows of about 24-28 square feet of window on a west-facing dormer in my 1.5 story house- it's not too dark, and does make a noticable difference, but not a huge difference. With more window area it would have a bigger net impact.)

The half-ton Mitsubishi FH06NA is a couple hundred cheaper than the FH09, and would handle your cooling load in the bedroom unless you live in an unisulated tarpaper shack with leaky "glass optional" single pane windows. There are several lower cost Chinese single-zone 3/4 tonners that might be cheaper than a 2-zone that are pretty good. Midea's DLCSRAH09AAK / DLFSHAH09XAK throttles down to about 3000 BTU/hr in cooling mode, runs an SEER of about 25 It's probably sold under a number of other vendors' nameplates (including Carrier). I'm probably more aware of the higher-end Chinese units that have good low-temp "hyper heating" capacity with more sophisticated compressor technology given my local market. There are probably cheaper but still pretty-good models with reasonably low minimum-modulation levels using simpler compressors.

 

That’s great info, thank you. The bigger room has about 55 ft.² of windows facing the west and 25 ft.² of glass on the door facing the west. Windows are cheap single pane.

The smaller room also has 55 ft.² of windows facing the west that are single pane. Both rooms do have solar shades. The door in the bigger room does get open often throughout the day and that room also has two big openings into the house. So I want to make sure the unit blends in well with the house.

I was looking at using pioneer. I think made by Midea. My searches have shown good customer satisfaction and when I called technical support twice with questions they answered immediately. The smallest they have is a 9000 btu head.

if the outdoor unit ramps down the 4000 is it still not a good idea to put a 9000 in the room?

 

Both rooms also have a flat roof with 30 year old insulation that is probably ineffective

 

Some Pioneer single zone 9000 BTU/hr mini-splits can't modulate below 5800 BTU/hr in cooling mode, eg: The WS009AMFI19HLD or WS009GMFI19HLD . It's important to find out from tech support what the minimum output of the individual multi-zone compatible 9K heads are. If the pair of head can evenly split the 4K output of the compressor it's not terrible. If the minimum output of a given head tied to the multi-zone compressor is 4K, it's not great, but better than 5.8K when serving a peak load of 5.8K.

In FL it's pretty safe to dense pack an unvented flat roof with cellulose or fiberglass, as long as it doesn't have a "cool roof" material up top. See Table 3 of this document, in particular the "1A Miami" and "2A Houston" rows, "Spray fiberglass (1.8pcf) " and "cellulose" columns. Note that with a light metal "cool roof" there is a slightly elevated risk in the Houston simulation, but that was simulated for a north facing pitch, not a flat roof, which gets more drying help from the sun.

It's fine to dense-pack over/under existing blown or batt insulation, letting it compress flat. Assuming 2x10 rafters (9.25" nominal cavity depth) it would result in about R38, if fiberglass, R35 if cellulose. But the R35 cellulose would actually outperform R38 fiberglass due to it's favorable "thermal diffusivity", which has a measurable thermal mass effect. The higher specific heat of the cellulose fiber inserts a few hours of delay for the peak ceiling temp after the peak roof deck temp, resulting in lower peak & average ceiling temperatures. (For non-technical graphic explanation, see this bit of advertising fluff from a European fiberboard insulation manufacturer. Note the differences in diffusivity between glass wool and cellulose shown on that page.) Even 2.5 -3 lbs per cubic foot cellulose would work in your climate- no need for the 4 lbs density they often shoot for in northern New England. That's doable with a single stage rental blower, if you're up for it.

But R38 using 1.8lbs density fiberglass would also be pretty good compared to some mouse-eaten low density R25s or whatever. Fiberglass is somewhat translucent to infra-red radiation, even more so for low density fiberglass and it's performance drops the hotter the roof deck gets due to the radiated portion of the heat transfer through the batt. At higher density that radiated fraction is very small. By contrast, cellulose is completely opaque to IR.

When re-roofing, as little as R5 rigid insulation above the roof deck does an even better job at keeping the moisture content of the roof deck low, and would make the unvented roof closer to what the IRC codes prescribe for IECC climate zones 1 (south FL) and 2 (most of FL.) That could be 1.25" of Type VIII (or denser) EPS, which is "walkable" without damaging the insulation, or 1" of 2lbs density roofing polyiso. Don't use 1" XPS in that application, since it's only warranteed to R4.5 @ 20 years, as it's climate damaging HFC blowing agents slowly diffuse out. At 1" it would settle down to R4.2 at full depletion. EPS is the same polymer, but doesn't have the early-years performance boost of the HFCs. It's blown with much more benign isopentane, most of which escapes the foam at the factory where it is recaptured, not vented to the atmosphere. Polyiso uses similar low-impact hydrocarbon blowing agents- usually a proprietary mix- often variations of pentane as the primary component.

 

I’ve beeb able to see what’s in the flat roof from our attic and it isn’t much. It wasn’t even R 19 and whatever it is it is 35 years old. It is batts though.

let me throw another wrench in this. The 150 sq ft bedroom does have a supply vent that I ran with the flex duct from our Plenum. It is a long run And it is picking up a lot of heat load. If I can solve that problem then I probably don’t need a unit in there but so far I have not been able to solve it. The air is cold enough in the evening that you don’t need to run the unit at night but during the day that room needs a little extra cold air. So I am wondering if a mini split is the right option. The room just needs a little kick during the hottest part of summer days

I’m guessing it could do a lot of short cycling during the day with the supply vent. The supply vent isn’t cold cold air during the day but it is cooler air, maybe 65 degrees out of the vent on a very hot day.