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

Sizing a Minisplit System

ericbarker | Posted in Mechanicals on

I’ve been considering a minisplit system in my 1950s ranch for quite a few years and I’m finally ready to pull the trigger. The primary motivation is to provide AC since there isn’t any currently and only have casement windows, making window units a no-go. However, our heating source is a propane fired boiler with hydronic baseboard. I’d ultimately like to get off of propane entirely so I want to size the system to be adequate as a heating source in the future as well, likely pending additional envelope improvements.

The relevant information. House is a 1950s ranch in Zone 5 (NJ). The first floor walls are hollow CMU with 1/2″ EPS under vinyl siding. Crawl space with no insulation. There is a partially finished attic with a combination of R38 of closed cell foam and dense pack cellulose. Last blower door was on the order of 9.5 ach50, depending on whether or not I include the crawl space in the volume calculation. Quite a few skylights. In short: not great.

I had a manual J done which gave 62k Btu*h heating and 27k Btu* cooling. I did note that the calculation was done based on a R11 wall assembly which is very optimistic for the assembly. I get around R4.5 from what I’ve seen. I’ll have the recalculation done, but also using the R11 results as somewhat representative of what it would be if I were able to do exterior wall insulation in the future.

The first estimate that I got for a system recommended the following:

-Outdoor condenser: AOU36RLXFZH
-One Indoor ASU18RLF
-3 Indoor ASU7RLF1

The 18k unit would be mounted in the living room area which is essentially open to the entirety of the first floor. The 7k units would go in the three bedrooms on the first floor. My thoughts are the following:

-Total output is low to provide full heating supply.
-We have a conditioned attic with no dedicated supply currently. In the future we may turn this into living space, but I’m not looking to add a system to the attic until we decide on that. It stays fairly comfortable currently.
-The multi-split setup is likely giving up some efficiency to multiple single zone systems.
-Because of the low R value of the exterior walls, I don’t know that it’s a bad idea to have individual heads to supply each bedroom. I suspect room to room variation, especially for heat, could be significant.
-I will be attempting to improve the air sealing and at least the crawl space insulation situation in the next couple of years.

My first thought is to put in a single zone unit for the main floor and then, potentially, maintain the multisplit system for the individual bedrooms. I don’t really have a good mounting location for a second, large, single zone unit on the first floor so I don’t know about, say two 18k wall units instead of the current plan. Very interested to hear others thoughts.

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.

Replies

  1. Expert Member
    NICK KEENAN | | #1

    Have you looked at a ducted head for the second floor? One head per room is unlikely to be either comfortable or efficient.

    Given the wide spread between heating and cooling load you're unlikely to be able to do both with the same minisplit system. What about keeping the hydronic and using it to supplement the minisplits? If you want to go all-electric you could get an air-to-water heat pump. With the sealing improvements and the minisplits providing part of the load you could probably provide the rest of the heat with water temperatures in the low three digits, which is the sweet spot for air-to-water.

    1. ericbarker | | #2

      Ducted head for the second floor could be a possibility, although I'd really rather not put anything up there for the time being. It's essentially walk-in storage with a small seating area that we barely use. If we were to convert it to more useable space I'd be concerned that anything that we'd do now might be in the wrong place. I do understand that it's conditioned space and ideally would have a heating/cooling source. There is an electric baseboard heater up there that I've never turned on.

      The hydronic system definitely isn't going away any time soon, just trying to make reasonable decisions with this system that reduce the propane reliance. I like the suggestion of the air-to-water heat pump. I'll have to look into those further. At the back of my mind there's also the question of power outages...we've had multiple events of at least a few every 2-3 years since we've lived here. I can power my propane boiler electronics very easily using a portable Honda 2kW generator. An all electric house would be a different story.

      1. Expert Member
        NICK KEENAN | | #3

        I misread your original post. What I meant is a ducted head for the three bedrooms, not on the second floor.

        1. ericbarker | | #5

          That makes more sense. I really only have access from above for one of the bedrooms, although I do have access below all of the rooms. I'll look into that more, I'm not as familiar with the routing and grill locations.

  2. paul_wiedefeld | | #4

    How many propane gallons did you use over the past two years?

    1. ericbarker | | #6

      For 2018/2019 I used right around 2000 gallons total.

  3. ericbarker | | #7

    So I've some additional research and gotten an additional estimate. I decided to run my own load numbers using http://hvac.betterbuiltnw.com/, which resulted in the attached results. I ran a few different ways and got fairly consistent results (changing things like treating the crawl space as a basement with a modification factor for wall height, different air infiltration assumptions, etc). I got slightly higher overall cooling (32k vs 27k) and slightly lower heating (58k vs 62k) than the numbers that the insulation contractor had calculated. But still, pretty close.

    If I look at the first floor cooling load, I get around 18k for the open living area plus the three bedrooms. If I add the crawlspace, I get 21k, and then the total with the upstairs comes to the 32k we have four skylights in a small, finished room upstairs that we seldom use, that are a large contributor to the cooling load.

    The second contractor specced a AOU24RLXFZH/ASU9RLF,2-ASU7RLF combination for the three bedrooms and AOU24RLXFH/ASU24RLF for the open first floor space. Again, massively oversized given the load calculations.

    My thought at this point is to just install the AOU24RLXFH/ASU24RLF combination in the open area. It is, on its own, sized slightly over the first floor cooling load, and not that far under the calculated whole house cooling load. I could then live with this for a year, see how it performs during a heating and cooling season, and then consider whether or not a ducted mini, or other solution, for the bedrooms is actually necessary. Alternatively, would it even make sense to size down further, perhaps to one of the 18k units, if there's a likelihood of adding other heating/cooling in the future? I understand that we'll likely have significant temperature variations through the house given our low insulation levels and high air infiltration rates, but the areas that we care about the most (kitchen, living room, master bedroom) are pretty much in line with whether the single head would be mounted.

    Long post, appreciate thoughts on the approach.

    1. Jon_R | | #8

      Note that 2x "massively oversized" 1:1 mini-splits will modulate down to less than a "perfectly sized" single speed AC. And we know that the latter can provide good comfort and equipment life and acceptable efficiency. Summary: if there is a good use for some additional capacity (eg, heating), don't worry about up to 2x over sizing of the cooling load with a 1:1 mini-split. Avoid 5x oversizing.

      Pay attention to dehumidification/latent load. Many inverter units often do a lousy job, even when perfectly sized. Some even do better when oversized!

      1. ericbarker | | #9

        Thanks Jon. Point taken. Unfortunately I don't really have a good place for another 1:1 head unit on the main floor. Based on what you're suggesting, , it sounds like I'd want to go with the ASU24RLF / AOU24RLXFWH or even ASU30RLE / AOU30RLXEH, although that head unit would be gigantic in our living area.

        The modeling indicated a roughly 3700 Btu/h latent load. Do you have a good link for how to align this with the Pints/hour type of specs that are listed for these systems?

        1. Jon_R | | #10

          Unfortunately, specs usually list a single sensible heat ratio (SHR) or pints/hr. But it varies significantly with conditions, making this pretty much useless. You could consider a Daikin Quaternity, which will maintain good SHR over various conditions.

          3700 Btu/hr is about 3.7 pints/hr.

  4. Expert Member
    Dana Dorsett | | #11

    With loads that high have you considered spending half the HVAC money on improving the building envelope instead?

    With the possible exception of radiant heating, a tighter higher-R house buys more comfort than throwing more BTUs at it to cover the higher load.

    A half inch of EPS on an 8" CMU wall has a U-factor more than twice the IRC 2018 code max for mass walls in zone 5. See:

    https://up.codes/viewer/nevada/irc-2018/chapter/11/re-energy-efficiency#N1102.1.4

    https://up.codes/viewer/nevada/irc-2018/chapter/11/re-energy-efficiency#N1102.1.2

    Swapping out the half inch EPS for 2" of foil faced polyiso or 2.5" of reclaimed roofing polyiso would bring it up to code, and make a HUGE difference in the interior wall surface temperatures in winter.

    1. ericbarker | | #12

      Dana, thanks for the reply. I certainly have considered spending money to improve the envelope. In fact, I started to last year. I had cathedral ceiling spaces dense packed with cellulose and had closed cell spray foam installed behind knee walls and along the underside of the attic roof. We previously had an unconditioned walk in attic with the floor insulated with a combination of vermiculite, rodent droppings, and in many areas...nothing. That effort also sealed the top of the first floor walls in as much of the house as possible, which are unfilled CMU. I also built an insulated, sealed door to our basement, the opening to which was previously air sealed using a blanket hanging from a pipe and have done other air sealing around the house where possible. So, despite appearances, things could actually be worse!

      At this point, re-siding the entire house (while I haven't gotten an estimate) to add polyiso, I'd imagine, would be in the neighborhood of at least double our HVAC budget. There are multiple bay windows that would have to get rebuilt if we were to pad out the siding by an additional ~1.5", awkward doors in inside corners, insufficient overhangs, and potentially even windows/old sliding doors that I'd want to replace. Our crawlspace is also uninsulated, which would be low hanging fruit except that I have the ends of the first floor joists embedded in the CMU wall. That obviously complicates the job significantly.

      That said, I want to do all of these things. But in terms of priority, we've spent multiple years now living in a house that has summer indoor temperatures in the mid 80s, which is even less enjoyable now that I'm working from home part time. I would use window units as a stopgap except that we have exclusively casement windows and the portable AC unit that came with the house makes a ton of noise to reduce interior temperatures from, say 85 to 83.

      My thought is to size a mini split to cover a significant portion of the overall house cooling load, with the thought that future improvements to the walls and crawl space will allow us to use it to fill an increased portion of our heating needs as well as reduce room to room temperature variations. I'd definitely be interested in your thoughts about my "logic", even if they're harsh. Trying to do make the best compromises possible given a lot of constraints.

      And yeah, the walls are reaaaaaly cold in the winter. I think I've measured the interior surfaces down to around 50 degrees on the coldest days.

Log in or create an account to post an answer.

Community

Recent Questions and Replies

  • |
  • |
  • |
  • |