Posted By toddm on 17 Oct 2018 06:09 PM
Routing lines instead of ducts makes it fairly easy to do but every house is a unique challenge. I have a wall mount conditioning 1,000 square feet downstairs and a second in a chase conditioning 600 feet upstairs. The downstairs benefits from slab on grade. The cold air sinks to the slab and rolls though my open-plan house. Closed doors and unusual heat loads --west wall, windows -- would toss a wrench in this approach. For upstairs bedrooms you'd want one head per room.

Consider an online purchase if you are adventurous. Midea is a Chinese OEM producing mini splits for Carrier, Lennox, Trane and a host of private labels, and making a push of late to sell under its own name. HVAC contractors understandably aren't happy to lose the hardware markup so you'll need to find a willing tradesman.

Posted By Dana1 on 17 Oct 2018 07:24 PM
It's definitely worth pursuing, but do your homework first. Getting the head sizing correct for the room/zone load(s) is more important for multi-zone ductless than it is for fully modulating one-head-per-compressor versions. Most multi-splits don't modulate in a continuous fashion, instead only changing in step-functions depending on which heads are calling for cool/heat, and to ensure performance there is always a bit of "extra". To keep liquid refrigerant from entering the compressor on the intake line that extra gets distributed to the other heads on the system even when those heads are inactive or even "off". When there is a gross mis-match between room load and minimum output even a nominally "off" heat will be enough to heat or cool the room. The zone load to head size has to be at least roughly proportional across all zones for this to not become a comfort problem, in either cooling or heating mode. A recent example of this came up on the GBA blog site (overheating, in this case):

https://www.greenbuildingadvisor.com/question/mitsubishi-heating-way-past-setpoint

So use a couple of the freebie online cooling/heating calculators to get a good handle on the oversizing factors on each zone before specifying or buying any equipment.

http://www.loadcalc.net/

https://www.coolcalc.com/

Be aggressive in the air infiltration assumptions (assume it's air-tight) and with the R-values to get good numbers, and don't rely on just one of the freebie calculators. Try to get them to agree withing 15% of each other if you can. As long as zone to zone ratios are the same, it'll be good enough for specifying head sizes.

toddm: Do you have verification that Carrier has switched to Midea? (They used to have Gree units under the paint, but that changed a year or so ago.)

Posted By 1201 on 18 Oct 2018 02:01 AM
Thank you. I didn't realize the multizone units were that complicated. I'll have to do the load calcs. I've never done one but it would be good to know. How would I account for my white roof and walls? Also, do you have any good multi zone units you would recommend? Or any resources I can study on installing a multi zone mini split? What are your thoughts on just doing three single zone units?

Posted By toddm on 18 Oct 2018 11:55 AM
https://www.carrier.com/carrier/en/tr/news/news-article/carrier__midea_launch_residential_ductless_hvac_joint_venture_in_north_america.aspx

Also Toshiba, Samsung, LG, Goodman, Kenmore. Solves the repair issue for online purchase, although you'll want to compare specs.

https://www.mideaminisplit.com/USA/cat2354120_4069043.aspx

Mine are badged Senville. Sogoodtobuy.com altho check prices on Amazon.

Posted By Dana1 on 18 Oct 2018 06:16 PM

Posted By 1201 on 18 Oct 2018 02:01 AM
Thank you. I didn't realize the multizone units were that complicated. I'll have to do the load calcs. I've never done one but it would be good to know. How would I account for my white roof and walls? Also, do you have any good multi zone units you would recommend? Or any resources I can study on installing a multi zone mini split? What are your thoughts on just doing three single zone units?

The white walls are "in the noise", of error on an insulated 2x4 wall, at most the cooling equivalent of R1-R2 exterior insulating sheathing. A high solar reflective index might be worth a bit more, unless you have more than R25 in the attic.

Single mini-splits have the advantage that they modulate up/down with load and won't interact with other zones, and it's often cheaper too.

But you still have to run the load numbers to size them optimally- the modulation range isn't infinite. Some only have a 3:1 turn-down ratio- if you oversize one of those by 2.5x for the peak load it won't modulate at all during the average load, and won't hit it's SEER numbers. The online calculators seem to oversize by 25% even when being aggressive on the input assumptions, so don't be tempted to take the next size up if your calculated load comes in 5-10% over the stated capacity of the unit. Be sure to look at the extended temperature capacity tables, which show the max capacity at different indoor & outdoor temps. The AHRI "rated" nameplate capacity in a submittal sheet is at 95F outdoors, 80F indoors (at a specified indoor & outdoor wet bulb temps too, which I don't recall off hand.) Your local 1% outside design temperature could be quite a bit above that, but the extended temperature capacity tables will allow you to still size it correctly.

Posted By Dana1 on 18 Oct 2018 09:08 PM
What sort of insulation do you have in the walls? What type of siding?

The "whole-wall R" of a typical 2x4 wall with empty wall cavities is about R4, after factoring in the thermal bridging of the studs, and adding in the R-values of the wallboard/siding/sheathing/air films, etc. Filling the cavities with R11-R13 batts increases that to about R10-R11 whole-wall. Adding an R2 factor to an R10 wall would reduce the conducted gain by 15-20%. If the high solar reflective index paint is reducing the surface temp by a LOT it could be a bit more than that, but I'd be surprised to see anywhere near 50%.

If the walls are uninsulated the effect of the high SRI wall paint would be much more pronounced.

Solar gains and cooling loads tend to be dominated by windows and roofs. When playing around with the freebie Manual-J load tools compare the cooling loads with clear glass single panes, clear double-panes, and low-E windows. The west facing windows have a bigger effect on peak load than south facing windows, due to the fact that the high sun angle at mid-day reflects a substantial amount of the incident light, whereas in the evening the sun is pretty much straight-at the windows with much less external reflectance. East windows have the same gain, but since those gains happen in the AM after the house and outdoor temperatures are cooler the effects on the peak load isn't as high as west windows, even if the effect on energy use is about the same.

Price out the ductless both ways, but in my area it's usually pretty much a wash, often cheaper to go with 2-3 separate minisplits than a 2 or 3 ton 3 zone multi-split. The raw hardware might be slightly cheaper for the multi-split, but the quoted installed price tends to be higher. Being able to modulate with load and no interactions with other zones is a big plus for going with separate minisplits. Having the visual noise of a row of condensers outside is the down side of that approach.

If you're using them for cooling-only it's pretty cheap to go with separate units. If using heat pump versions it's an upcharge. In my area (New England) it really pays to go with a "hyper heating" or "cold climate" version which has higher capacity and efficiency at low outdoor temps than plain-jane types, but that may not be the case in TX. Does it drop below 15F very often?

The primary difference is the compressor technology- cold climate versions use some clever vapor injection schemes to optimize it for capacity at the temperature extremes, which has the additional benefit of somewhat better overall efficiency even in the middle ranges. Midea's's "Hyper" versions have SEERs north of 20, whereas their other units test at SEER 15.

Posted By Dana1 on 18 Oct 2018 10:56 PM
The wall surfaces never hit anywhere near the peak temperatures of roof surfaces. This is because vertical surfaces hotter than the ambient air temperature see vastly more convection cooling than the roof does, and the high intensity mid-day sun comes at walls at an oblique angle, but more directly on south facing roof pitches.

In 1987 R13 was the standard for almost all 2x4 construction, even in the south. R11s & R7s were common enough during the 1960s when oil was $8/bbl. Assume it's R13 unless you KNOW it's something lower.

Average kwh use isn't a good measure of peak cooling loads, and it gets even weirder when you're looking at systems with uninsulated or leaky ducts in the attic above the attic insulation, which on some days can account for half the energy use but less than half the peak load. Unbalanced & or leaky ducts can also drive infiltration loads.

Just run the online cooling loads assuming no infiltration, and at the most-optimistic R-values that might be credible (R13s, not R10s, etc) and see where the numbers come out, room by room, zone by zone, with no discounting for high SRI finishes. The high SRI roof makes a much bigger difference when the ducts are in the attic due to the high parasitic load. The load you care about is the load in the rooms, which see some benefit, but not the huge benefit reaped by systems stuck up in the attic.

Posted By 1201 on 19 Oct 2018 04:49 PM
you are correct about that. the wall was always about 10 degrees cooler than the roof, even though there is more air movement at the roof vs walls. peak I ever measured at roof was 150. vs 140 peak for the walls. but the walls do get quite hot too. on a 75 degree day here I still measured walls in full sun getting to 118 degrees. I also agree that a lot of the benefit was because the attic where the ducts are, was not getting as hot in the midday sun. I didn't measure it but Ive heard that attics can get to 170 degrees. I ran a loadcalc and the three bedrooms upstairs all came in around 2-3k btu. the smallest units I saw were 9k btu. would a 9k btu unit work for a 100 sq ft room? ie would it be too grossly oversized?

Yes, most would be grossly oversized, but some can work. There are no 3/4 ton mini-splits sold in the US that would be ideal for a 2-3KBTU/hr load. The LG "Artcool" 3/4 tonners can modulate down to about 1,000 BTU/hr ((models LA090HSV5 or LA090HYV1 same innards, different finish on the indoor unit), which wouldn't be bad for a 2-3K peak load. Most others have minimum outputs that are considerably higher- higher than your peak load even . (eg: The minimum output of a 3/4 ton Midea is about 4500 BTU/hr.)

The solution for clusters of rooms near each other all with low loads is often a mini-duct cassette. Most can be mounted below the ceiling in a closet to serve 2-4 adjacent rooms/spaces with a short plenum splitting into appropriately sized duct runs to the rooms. Some mini-duct cassettes have more power for moving air than others, and the lower the power, the shorter the duct runs must be, and the more careful the duct design has to be. This is of course more time consuming & complicated to install than a wall-blob type, but despite lower tested SEER & HSPF they will run more efficiently than a wall type that is cycling on/off most of the time rather than modulating. A modulating 9K mini-duct serving a 6-8000 BTU/hr peak load will usually do VERY well on both efficiency & comfort. Running the ducts to remote rooms in soffits below ceiling level is usually possible, but it means more construction work. It's not particularly hard work, and as a DIY it's pretty affordable. (The Midea MCHSU-09PHH2 + MEHSU-09CHD2 hyper-heating 3/4 ton mini-duct cassette modulates down to 3K out, and can be had for under a grand online, not counting linesets, condenser pad, ducts, and other ancillary items, if you're leaning that way.

A Fujitsu -9RLFCD also modulates down to about 3K,out has much better efficiency, and the advantage that it (unlike all others) can be mounted both horizontally or vertically, and MUCH higher static pressure specs (= longer duct runs). Mounted vertically in a "mechancical closet) with a common return grill below and the duct plenum above it'll take up about 5-6 square feet of floor area and easy access for service, filter changes, etc. About $1800 online.

This is a 1.5 ton mini-ducted Fujitsu mounted in just such a micro closet, with some of the duct runs in below the ceiling soffits:

https://uploads.disquscdn.com/images/2ffa6e108a7ded9f51130ff14126239b275b1244b7d53138beb63b4182d68f13.jpg?w=800&h=1200

https://uploads.disquscdn.com/images/7843213f27734395e6ede8ea696552a8eafd3a2dd7f62c2b61241bb23189a293.jpg?w=800&h=1200

Posted By 1201 on 19 Oct 2018 05:04 PM
thanks a bunch toddm. I'll check them out. I looked on amazon and I couldn't find any branded midea but I did see some affordable 9000btu units for about $600. I would need three of those if I go with individual units.


question- I read somewhere that the indoor units run the fan continuously, even at set temp??

is that correct? I ask because it could make humidity a problem.

thanks