GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter 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

Help with air-source heat pump sizing

llsrvd | Posted in General Questions on

We are looking into getting minsplits to take care of most of our heating and provide some cooling. We currently have no AC and have a steam boiler which also heats our water. We live in Concord, MA, west of Boston. My concern is about correct sizing of the units. Our house is 1462 square feet not including the basement. Main floor is about 800 square feet so upstairs is about 600. Upstairs is a finished attic with probably average insulation. Walls are 2×4 construction and are insulated. House was built in 1950. We have 2 pane low e windows. Contractors so far are not using Manual J. Proposals so far are for 1 unit downstairs and a smaller unit upstairs. I would be grateful for any guidance. Thanks!

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
    Dana Dorsett | | #1

    Get a RESNET rater or a certified P.E. to run the Manual-J for you, and stress that it's for a heat pump, and needs to use the most aggressive assumptions (rather than conservative), since oversizing is going to have a bigger hit on heat pump efficiency/comfort than for gas burners.

    The whole house load of this place if the basement is NOT insulated will still be no more than about 30,000 BTU/hr, and could easily be less than 25,000 BTU/hr. If the basement is insulated and well sealed the heating load could be closer to 20,000 BTU/hr (but probably still above that mark, in most cases. Insist that they assume very tight construction, using heat recovery ventilation, or the default numbers for air leakage will exceed reality.

    Have TWO separate 1 ton or 1.5 ton heat pumps would probably get you there with warm climate heat pumps. If a cool climate modulating mini-split (eg Fujitsu xxRLFCD series) you'd likely drop down to a 3/4 tonner, since they don't come smaller than that. According to NEEP's spreadsheet the -9RLFC delivers 15,000 BTU/hr (max) @ +5F, 13,500 BTU/hr @ -5F, which is proabably more than the heat load of an individual floor. For comparison shopping (after getting the room by room, floor by floor load numbers) see:

    http://www.neep.org/sites/default/files/ColdClimateAir-SourceHeatPumpSpecificationProductListing-Updated6.22.18_0.xlsx

    The 99% outside design temp for Concord MA is somewhere between +5F and +10F (Framingham's is +6F, Worcester's is +5F, Boston's is +12F) but it does (rarely) hit -5F. With standard heat pumps resist the urge to oversize to cover the loads at lower temps- use auxiliary heat strips to supplement the heat pump at sub-zero. With the cool-climate Fujitsus it's OK to bump it up one in the (unlikely) event that a 3/4 tonner is marginal for the load of one floor, but not two. Oversizing a standard heat pump by 1.5x to cover loads at temperatures 10F colder than your design temp results in excessive cycling losses during most of the heating season, resulting in heating season efficiency substantially lower than the nameplate HSPF- burning a bit of heat strip power to cover the difference for fewer than 1% of the hours is more efficient overall. But with heat pumps that modulate over a wide range (like many mini-splits) 1.5x oversizing can still be fine.

  2. GBA Editor
    Martin Holladay | | #2

    Lyell,
    Dana is right: The first step in designing an HVAC system is perform your load calculations. Here is a link to an article you should read: "Who Can Perform My Load Calculations?"

  3. Yupster | | #3

    I perform a lot of heat load calculations and I see plenty of two level houses around that ft² size in cold climates with much higher loads than that when they have minimal insulation. Most of the houses around here in built in the 1950's era have R-8 to R-12 batts, and R-20 in the attic. No basement insulation. They often have loads of 50,000 btuh's or more. If the basement was somewhat insulated I might hazard a guess closer to 30,000 btuhs. I do live in Ontario, Canada and we use a slightly different calculation method called CSA F280-12. But I would guess much higher for a 1950's two-level house with poor insulation. Of course, that's why it is important to have the numbers run. You just can't guess at it, houses are all so different.

  4. llsrvd | | #4

    Thank you all for your advice! First of all I'm a little confused about how to name the size of the units. If a unit lists 18K BTU/H for Heat and 15K BTU/H for cooling what size unit is it? What has been proposed so far Mitsubishi MSZ-FH15NA for downstairs and MSZ_FH12NA for upstairs and MXZ-3C30NAHZ for outside. The other quote was for Mitsubishi MSZ-GL24NA first floor and MSZ-GL09NA upstairs and the same outdoor unit as the first quote. On my own I did a calculation using LoadCalc. I'm trying to avoid hiring some one to do it, although i understand that is the most thorough way to do it. Regarding the house, the basement is not insulated except for some fiber glass I added along the sill plate. Dormer was built in the 70s and batt insulation is in there- probably not high R value. We had fiber glass blown into the first floor which had no insulation at all. The attic is insulated and the knee wall but not optimally. So isn't Ontario colder than here? Would you estimate 50K for our climate?

  5. Expert Member
    Dana Dorsett | | #5

    The "size" is usually the rated cooling capacity under AHRI. TheAHRI heating capacity isn't closely coupled to the AHRI cooling number.

    Worse still, the AHRI heating heating capacity (which it must be able to deliver at +17F), isn't necessarily the MAXIMUM capacity of a modulating heat pump, or it's capacity at your +7F-ish outside design temperature, but rather the modulated output level at which it's efficiency was tested.

    With the ductless Mitsubishi FH-series the "rated" capacity would also be the full-speed capacity at +5F outdoors, 70F indoors. That's close enough to your 99% outside design temperature that it's "good enough" for sizing, once you have the calculated load numbers.

    The heating capacity @ +5F of the FH15 is 18,000 BTU/hr, (and probably twice the actual heat load of the first floor):

    http://meus1.mylinkdrive.com/files/MSZ-FH15NA_MUZ-FH15NA_Submittal.pdf

    The heating capacity @ +5F of the FH12 is 13,600 BTU/hr, and probably close to 1.5x the load of the upstairs (TBD):

    http://meus1.mylinkdrive.com/files/MSZ-FH12NA~MUZ-FH12NA_Submittal.pdf

    But when tied to an MXZ-3C30NAHZ compressor they do not modulate with load, which makes getting the sizing right critical! If going with separate mini-splits they would at least modulate, adjusting it's output with load for at least part of the season, even if oversized.

    The GL24 has even more capacity @ +5F than the FH15, and thus probably approaching ridiculous oversized for the first floor:

    https://nonul.mylinkdrive.com/files/MSZ-GL24NA-U1-MUZ-GL24NA-U1_ProductDataSheet.pdf

    The GL09 delivers 7600 BTU/hr @ +5F, and might be close to right sized for the upstairs (or not- TBD):

    https://nonul.mylinkdrive.com/files/MSZ-GL09NA-U1_MUZ-GL09NA-U8_Submittal.pdf

    At 28,600 BTU/hr @ +5F the MXZ-3C30NAHZ multi-split may not be too ridiculously oversized for the whole house load, but was probably selected to support the oversized heads for the first floor.

    http://meus1.mylinkdrive.com/files/MXZ-3C30NAHZ_Submittal.pdf

    The MXZ-2C20NAHZ or MXZ-3C24NAHZ might be more appropriate, if the heads are better tuned to the actual loads,without a large oversizing factor:

    http://meus1.mylinkdrive.com/files/MXZ-2C20NAHZ_Submittal.pdf

    http://meus1.mylinkdrive.com/files/MXZ-3C24NAHZ_Submittal.pdf

    Better still (for both comfort and efficiency) would be to use two separate and completely modulating compressors, not a multi-head.

    A pair of 3/4 ton Fujitsu -9RLF3H units would probably cover the loads, delivering 15,000 BTU/hr each @ +5F, 12,000 BTU/hr each @ -5F, and probably cheaper than a MXA-3CNAHZ solution:

    http://www.fujitsugeneral.com/us/resources/pdf/support/downloads/submittal-sheets/9RLS3H.pdf

    (The+5F & -5D max capacity numbers are from the NEEP spreadsheet.)

  6. Yupster | | #6

    The design temperature for where I live is only -9°F. Lowell near you has a design temp of 1°F. A 10 degree difference is signficant but the kind of house you are describing usually has a lot of missing insulation in awkward corners in the finished attic and low ceiling insulation. Air sealing is also pretty bad usually. So my guess would be in the ~50,000 range. But you have load calcs, so you don't need my guess. :) What did they come out as?

  7. llsrvd | | #7

    I haven't hired any one to come in and calculate my load. I just did something on line at loadcalc.net. It came out as 33K BTU heating and 14.5 BTU cooling. I also had an assessment done by giving details about my house and 6 years worth of oil bills to someone who put them into a model developed by Tufts University. that came out as needing a 12 K -15K unit for the WHOLE house. So I am a little confused. How much does it cost to hire someone to do a load calc?

  8. GBA Editor
    Martin Holladay | | #8

    Lyell,
    Once again, I urge you to read this article: "Who Can Perform My Load Calculations?"

    The article includes detailed advice and estimated cost information.

  9. Expert Member
    Dana Dorsett | | #9

    Yupster: Actually Lyell Slade does NOT have the load calcs (yet), which is THE problem right now. One proposal indicates the contractor thinks the load of the upstairs is under 7,600 BTU/hr, the other is presumes more than that, but under 13,600 BTU/hr.

    Both could be wrong. Using a stupid BTU per square foot rule of thumb at 15 BTU/ft^2 the 600' space would come in at 9K, at 20 BTU/ft it would be 12K. But if there isn't a lot of window area up there it could even be 6K.

    If I were to play "busy contractor" and just take a WAG at it, the 10,900 BTU/hr capacity @ +5F of the FH09 would probably do it, and the 15,000 BTU/hr capacity of the 9RLS3H (which is more than the output of an FH12) would almost SURELY do it (with the windows closed, anyway. :-) ) Even if the load is 6K, with a dedicated modulating compressor to let it adjust it's output with load the ridiculous oversizing factors won't matter, since ether the FH09 or 9RLS can modulate low enough that it would still modulate most of the season.

    But if tied to multi-split the heads have to be sized a bit more precisely to hit the efficiency numbers (and to stay comfortable.)

    A load of 50KBTU/hr @ 1F for a '50s vintage 1500' house would only be possible if there were NO insulation, and single-pane-only windows (no storms.). That might have been the original heat load, but with low-E double-panes and wall insulation, even with only R13 in the roof /attic it would be under 30K.

    Concord is roughly equidistant between Lowell and Framingham, but closer to the moderating effects of the ocean than either of those locations, and slightly lower in elevation than Framingham where the 99% outside design temp is +6F. It's unlikely that the 99% outside design temperature is colder than +5F. Either +6F or +7F is more likely. Further inland and 500' higher in elevation, Worcester's 99% design temp is +5F or +6F (depending on whether you're looking at ACCA vs. ASHRAE's published numbers.)

    https://www.captiveaire.com/catalogcontent/fans/sup_mpu/doc/winter_summer_design_temps_us.pdf

    https://articles.extension.org/sites/default/files/7.%20Outdoor_Design_Conditions_508.pdf

  10. Expert Member
    Dana Dorsett | | #10

    The oil use model can be useful.

    I could believe the 15K heat load number if the basement is insulated and the house is pretty tight, but 12K heat load doesn't seem likely unless you have a much lower than "typical" window/floor area ratio than most circa 1950 houses and a VERY tight house.

    What Martin said- pay for a PROFESSIONAL load calculation, performed by somebody who makes their living & reputation off the accuracy of their numbers (as in "not an HVAC installer", who makes a living installing and maintaining equipment.)

    The loadcalc online tool tends to oversize by 30-35% unless you select "very tight" construction & ducts, and no ventilation. At that point it still oversizes compared to an aggressive Manual-J using a pro tool, but not nearly as much. When the loadcalc comes up with 33K, reality is likely to be about 25K in a "real" Manual-J, (which is still slightly higher than a "15 BTU per square foot" idiot's rule of thumb that works most of the time in eastern MA.)

  11. Yupster | | #11

    I stand corrected Dana, I ran a quick calc on a building roughly the size specified, with some average amount of windows and insulation thrown in and it came out at ~35,000 btuhs. I guess I shouldn't try to guess at climates I don't live/work in. :D

  12. llsrvd | | #12

    I really appreciate all the advice. I have now read Who can perform my load calculations? And it sounds like it's actually not that easy to find someone to do a small already built house. However I sent email to Mike Duclos in Newton (mentioned in the article) to see if he would be willing to do it.

  13. llsrvd | | #13

    While I wait to hear back from Mike Duclos, I had one other question about floor plan and how that affects sizing. The main floor is an open L shape that includes the dining room, living room and kitchen. There is a small bedroom and bathroom also. Upstairs is a master bed and bath and one other small bedroom. We really only want 2 units (would like to avoid added cost) and realize there will be rooms that are underserved. Does that configuration change the load recommendations?

  14. Expert Member
    Dana Dorsett | | #14

    The load is what it is. Oversizing the hell out of a head in a space next to a doored-off room doesn't heat the doored-off room any better or worse than sizing it for the whole floor. Where you place a ductless head/cassette and how it is oriented to blow can be optimized, but short of a ducted mini-split (more expensive to install), there will probably be some heat distribution issues for the upstairs. Odds are pretty good that optimal placement of a wall-coil type could work downstairs (but if you have a full basement in which to run ducts a ducted mini-split may be a better option even for the downstairs.)

    Fujitsu's xxRLFCD slim-duct cassette systems mentioned in response #1 are pretty good, but how easy it is to install & run ducts fully inside of conditioned space will vary. A 9RLFCD (12KBTU/hr max @ +5F) would probably cover your upstairs load, an 12RLFCD (16K max @ +5F) almost surely would, but the installed cost would be considerably more than wall-coil type mini-split. A detailed floor plan of the upstairs might help decide if that's worth contemplating.

    http://portal.fujitsugeneral.com/files/catalog/files/9RLFCD.pdf

    http://www.fujitsugeneral.com/us/resources/pdf/support/downloads/submittal-sheets/12RLFCD.pdf

  15. llsrvd | | #15

    That makes sense. thanks for the info on the ducted system. I think that's beyond where we want to go right now. If we can get the 2 heads sized correctly, at least we're improving on our energy efficiency over the oil boiler.

  16. llsrvd | | #16

    Dana, re your earlier post, the basement is not insulated (except for the sill plates), is not well sealed and overall the house is not tight. And I would guess we have average window area for this age, size house.

  17. llsrvd | | #17

    Hi Dana, with regard to your answer, number 5 above, I am trying to understand the difference between having dedicated outdoor units for the 2 indoor heads and a single outdoor unit. I understand that it is not as efficient to have a multi split. I believe I read elsewhere on here that they can be 20% less efficient. Is that correct? In your answer above you mention that if the outdoor unit is smaller (2C20 or 2C24) that could help, but I notice that the 3C30 and those 2 all only modulate down to about 7000BTU/hr. I'm definitely out of my depth here. Can you help me understand why the smaller outdoor units would be better and why the dedicated unites would be even better? Thanks!

  18. Expert Member
    Dana Dorsett | | #18

    Lyell:

    >" I understand that it is not as efficient to have a multi split. I believe I read elsewhere on here that they can be 20% less efficient. Is that correct?"

    I wouldn't hazard a general percentage- many things affect the operating efficiency, but yes, within a manufacturer's line up the tested HSPF efficiency (which has a presumed load relative to the size of the compressor) can be more than 10% lower, even as much as 20% lower. But a lot depends on the size of the load relative to the size of the compressor, and the size of the heads relative to the zone loads. Oversizing the compressor comes with an efficiency penalty, and your peak load isn't necessarily going to match HSPF test conditions. Oversizing the heads for the zones drives the overall duty cycle down, reduces the likelihood that zone calls for refrigerant will overlap to keep the compressor spinning, and the compressor will cycle on/off even more often, all of which takes a toll on efficiency.

    >"In your answer above you mention that if the outdoor unit is smaller (2C20 or 2C24) that could help, but I notice that the 3C30 and those 2 all only modulate down to about 7000BTU/hr. I'm definitely out of my depth here. Can you help me understand why the smaller outdoor units would be better and why the dedicated unites would be even better?"

    Spinning up a heavier compressor & blower wastes more energy than spinning up a smaller one, so there is an as-used additional penalty (and extra up front cost) for oversizing the multi-split compressor when the heads are sized to guaranteed on/off cycling. Mulit-split compressors don't modulate the output to track the load- they step up/down depending on the number & size of the zone heads that are calling for refrigerant.

    By contrast, single zone mini-splits DO modulate the up/down to track the load. At part load the large size of the coil relative to the immediate load makes the heat transfer more efficient, and the overall efficiency soars compared to running at high speed and duty-cyling on/off.

  19. llsrvd | | #19

    Thank you for all the work you do on here helping people with their questions. And I 'm sorry I keep asking all these questions. So if have an FH09 that modulates down to 1600 Btu for heat and an FH12 that goes down to 3700 but I put them on an multi split unit that only goes down to 7000, do I get the benefit of that lower range? I apologize if this is a stupid question!

  20. Expert Member
    Dana Dorsett | | #20

    The answer is, NO, there is no advantage to the modulation range when tied to a multi-zone compressor, since they won't modulate. This is exactly the problem I've been trying to describe.

    When married to a multi-split compressor the FH12 will only run at it's nominal 13,600 BTU/hr rate...

    ...and the FH09 will only run at it's nominal 10,900 BTU/hr.

    http://meus1.mylinkdrive.com/files/MSZ-FH09NA_For_MXZ_MULTI-ZONE_SYSTEMS_ProductDataSheet.pdf

    http://meus1.mylinkdrive.com/files/MSZ-FH12NA_For_MXZ_MULTI-ZONE_SYSTEMS_ProductDataSheet.pdf

    For the MXZ compressor to even throttle down to it's minimum 7200 BTU/hr rate one of the zones would have to be an GL06 or GE06 head (the nominal output of the FH06 is higher than that, at 8700 BTU/hr).

    http://meus1.mylinkdrive.com/files/MSZ-GL06NA-U1_For_MXZ_MULTI-ZONE_SYSTEMS_ProductDataSheet.pdf

    http://meus1.mylinkdrive.com/files/MSZ-GE06NA-9_For_MXZ_MULTI-ZONE_SYSTEMS_Submittal.pdf

    http://meus1.mylinkdrive.com/files/MSZ-FH06NA_For_MXZ_MULTI-ZONE_SYSTEMS_ProductDataSheet.pdf

    The only time these heads CAN modulate is when they're married to a dedicated single-zone compressor, at which point, yes, the FH09 can throttle back to 1600 BTU/hr , the FH12 down to 3700BTU/hr, etc.

    You can change up the blower speeds of the heads with the remote when married to a multi-split compressor, but that doesn't change the net output much, and when there's excess refrigerant due to running a head at low blower speed it gets distributed to the other heads, whether their blowers are running or not. With ridiculous oversizing of the heads to their room loads this excess refrigerant flow can put enough excess heat/cool into those zones to create a comfort problem even WITHOUT the blower running.

    It's often/usually cheaper to install two separate mini-splits than a 2 or 3 zone multi-split, and it's more efficient & comfortable to boot. But even with separate mini-splits it's important to size them appropriately for the loads- the modulation range is not infinite, and a bigger compressor that is cycling on/off during periods of low load is still wasting more power than a smaller compressor that's cycling on/off.

  21. llsrvd | | #21

    Hi again Dana, in looking over the submittal sheet, I noticed this:

    Operating Conditons (Indoor Intake Air Temp.) (Max./ Min.)
    Cooling5
    90o F (32o C) DB / 67o F (19o C) DB
    Heating
    80o F (27o C) DB / 70o F (21o C) DB

    Does this mean that the temp inside in the winter has to be at least 70F for it to operate? We keep our heat at 63F in the daytime and 61F at night. Will we not be able to do that with these units?

    Thanks!

  22. Expert Member
    Dana Dorsett | | #22

    Those conditions listed in the submittal sheets are the "Rating Conditions per AHRI Standard" conditions at which it's efficiency and capacity was tested.

    The full range of indoor operating conditions is MUCH wider than that, spelled out in the operating & engineering manuals.

    The lowest east to set setpoint for a Mitsubishi unit in standard heating mode is 60F, but it can be set even lower (as low as 50F) if you go into "SMART SET" mode on the remote. Setting back a couple of degrees won't have a huge impact, on efficiency, but for modulating minisplits it's generally more efficient to "set and forget", letting the system ramp up and down in micro-increments as the load changes. When you raise the settings by 2F it will run at a much higher speed until the new setpoint is reached. That higher speed is less efficient, and will more than erase any energy savings that accrued from the lower indoor temperature. (A 5F setback will take a significant bite out of the as-used efficiency.)

  23. llsrvd | | #23

    Thanks for your answer on the multi split versus mini split. I have a much better understanding now and plan to ask for 2 dedicated outside units. And for the answer regarding the indoor temp range. That makes a lot more sense. And thank you for the important advice to set and forget. We'll definitely follow that once we're up and running.

  24. kenorakq | | #24

    I have been reading/lurking for a long time here and I must say...that is the most informative thread I've read on the whole multi vs mini split topic...thanks to Dana and Lyell

  25. GBA Editor
    Martin Holladay | | #25

    Yes. Dana Dorsett deserves a round of applause for generously sharing his expertise with GBA readers.

    Thanks, Dana.

  26. Expert Member
    Dana Dorsett | | #26

    I'm no expert- I just make it up as I go along! :-)

    (Sometimes I'm lucky, and manage to be right, but not always...)

  27. ohioandy | | #27

    Lyell, a couple weeks ago in comment #15 you concluded that a mini-ducted was probably beyond where you wanted to go, due to the higher install costs. I wonder if others can chime in about this from a comfort perspective. I built (and live in) a house that uses a Fujitsu 9k ceiling cassette in the open-plan main floor, plus a second 9k Fujitsu mini-ducted in an upstairs hallway ceiling feeding two bedrooms up there. After a few winters with this arrangement, I'm inclined to think that a mini-ducted in the basement (i.e. in the floor of the first level, feeding the first level) would be more comfortable and efficient than the single ceiling cassette (or a unit mounted high on a wall.)

    My new house is quite tight and stratification shouldn't be a big issue, but with the heat sources located quite literally near the top of the building, things are a bit too cool down low. (I have some backup resistance heat in the basement, which is well insulated, but I'm loathe to run that too much.)

    So now I'm also about to retrofit an old house with minisplits, and it's a similar size to yours, with two levels and an open basement. After all the air-sealing and insulation work, I think spending up a little on the labor of a mini-ducted install--instead of a single wall-mounted head--will pay huge comfort dividends on the main floor.

    I know HVAC installers probably avoid the mini-ducted because they typically involve, well, ductwork, plus some carpentry and finishing. How have others weighed this HVAC design choice?

  28. llsrvd | | #28

    I have been working with a contractor who has agreed to the 2 outside units. However he feels that we really need the Mitsubishi FH12NA upstairs even though that space is only about 550 square feet. Here is what he says "The 09 unit is only rated for up to about 400 square feet, that unit is on a second floor too that will have a larger cooling load during summer temperatures." Thoughts? Thanks!

    1. Expert Member
      Dana Dorsett | | #33

      Lyell: The FH09 is good for 10,900 BTU/hr heating @ +5F, and 12,000 BTU/hr cooling (max- was tested at 9000 for SEER efficiency). I can't imagine the loads of your 550 square foot zone exceed that. What did your loadcalc tinkering come up with for heating & cooling loads on that space?

  29. walta100 | | #29

    I think you will find when you start using the minis you will find the upstairs unit will do almost all the cooling and the down stairs unit will do almost all the heating. If you are keeping the steam system consider installing just the upstairs unit and see how things work. You may decide the second mini is something you can live without.

    Walta

  30. kjmass1 | | #30

    Are you looking to add AC, reduce your energy footprint, offset your steam usage, or a little bit of them all?

    I was in a similar situation- 1940s home with old steam boiler and no AC. Good insulation and windows, about 1800 sqft outside of Boston. Open first floor and three BRs on second floor with open landing. Ideal for wall splits. I went with TWO Fujitsu multi's 24ks, and in hindsight should have done things differently. I have a 12K on each floor (3), and a 7K in the master BR (I like it cold at night). A couple thoughts:

    At least with Fujistu, a wired thermostat is mandatory in my eyes if you want to try and heat an entire floor. You just have to be able to sense the temperature from far away. Sensibo unit also can work if you are tech inclined.

    Like Walter said, my second floor unit barely runs in winter, maybe 20% of the 1st floor unit.

    My single 12K can keep the whole house pretty close to 71 by itself in to the low 30s. Below that it'll start defrosting and slowly start to drop in output (not XLTH). You need a clear shot for the air to circulate so hallways and walls aren't ideal. I have steam heat as a backup, and anything below 20 degrees I just turn off the heatpumps and enjoy the radiators. The heatpumps really shine in Spring/Fall.

    This past summer, the 12K on the second floor, which had a clear shot to blow down the steps in swing mode, kept the entire house at 71 and <45% humidity. If you close the bedroom doors they can creep up 3-5 degrees at night. A couple of the multiday heatwaves I would turn the first floor on dry mode. I also run the master 7k just to help out and I'm a little picky with sleeping temperature. I think even during the 2-3 day heatwaves ~100 that we had in July, if I only had the second floor 12K, it would still keep the house under 73-75. Maybe that's acceptable to you, but like I said I'm more sensitive and need a consistent 71 and dry.

    If I had to do it over, I would have done a single modulating 12K on the first floor primarily for heating, and figured a way to get the other 3 units all on the same multi. I just didn't want linesets running all over the house.

    Don't skimp on an installer, trust me. Use a certified one.

    I didn't notice the issues Andy mentioned, but with the wall units you can direct the air more towards the floor.

    Between splits and radiators- enjoy the silence! You can't beat it.

  31. Expert Member
    RICHARD EVANS | | #31

    Lyell,

    The mini split installer may truly mean well. I don't think they will make much more from selling you the FH12NA compared to the FH09na. The installer may be genuinely trying to adhere to Mitsubishi protocols (whereby they want a unit in every room).

    After dealing with countless HVAC contractors- all of whom wanted to oversize my mini split by 5x, I had to resort to trickery to have them install a right-sized unit. I finally told them that I was installing a 140k btu wood stove and that the mini split would just be for the cool summer nights here in NH. It worked. Some of them finally agreed to install the right-sized unit...

    You might have to get creative to make this work. Good Luck!

  32. user-6842208 | | #32

    This is a great discussion! I am right now getting ready to install mini splits in my new construction house that is 1260sf ranch with basement insulation, 2x6 walls with cellulose and 1.5” exterior Roxule insulation, VERY good attic insulation, and decent windows. I have worked hard to keep it as air sealed as I can with my limited knowledge...we have had a LOT of trouble finding anyone around here that knows anything about building efficiently so I’ve taken on most of the details myself. So...we are slated to install a 36,000 btu outdoor unit with a 9,000, 15,000, and two 6,000 btu heads indoors. There is a head in every room. By the sounds of it, this is WAY overkill for our house. The plans have been passed by our HERS rater and they didn’t say it was ridiculous. The HVAC guy did a load calculation and he actually sized this system SMALLER than another HVAC guy I spoke with. So, now I see that I should probably have someone else do a load calculation but I am pretty much OUT of time. The system should be installed in the next couple weeks! Any advise on getting a hold on this fast?

  33. walta100 | | #34

    user-6842208

    Please tell us your name and where your home will be.

    Do you want to start your own thread or hijack this one?

    If I were to be bad and use a rule of thumb for a well built house would be 1500-2000 square feet per ton for AC.

    I think you will find you will get the best bang for your buck with a conventional 17 seer heat pump with the duct work in your basement.

    Walta

    1. Expert Member
      Dana Dorsett | | #35

      1200-1500 feet per ton is the WAG rule of thumb number I'd use for a smaller house for cooling. A ton per 2000 might be possible- a lot depends on the window orientation & type. A 1260' high-R rancher could even have as much as 1.5 tons of cooling with high-gain windows, but I doubt it.

      A 1.5 ton Fujitsu fully modulating mini-ducted unit in the basement would probably cover the cooling load, and delivers over 18,000 BTU/hr of heating @ -5F, which may or may not be enough, depending on the actual 99% outside design temp. Since it modulates down to 3100 BTU/hr in cooling mode (and at +47F in heating mode) it's not a disaster if you only have 1 ton of cooling.

      A 3 ton non-modulating multi-split does indeed feel like extreme overkill.

      Run your own quick & dirty Manual-J using loadcalc.net and you'll see how ridiculously oversized a 3 ton system would be. The loadcalc number will be bigger than what a pro tool would spit out, and you won't be able to input your actual U-factors/R-values of a higher performance house, but see where it comes out.

Log in or create an account to post an answer.

Community

Recent Questions and Replies

  • |
  • |
  • |
  • |