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

Heat pump in climate zone 5

raul4817 | Posted in General Questions on

I recently completed a manual J via the coolcalc software.  I am in climate zone 5. I am installing the equiptment in a conditioned attic space. This unit will only serve the newly constructed 2nd Floor. My wall assembly is an R26 with a cathedralized attic ceiling of R50. My windows are triple pane Double lo-e with argon U value .19 & SHGC . 36. I have 1100sf of floor space and approx. 13k cubic sq ft including conditioned attic space.  Coolcalc Software is giving me 40k +\-   Heating BTU’S & 20K+\-  Cooling BTU’S . 
So my question is does a heat pump make sense in my climate? Natural gas is still pretty cheap around here but I would at some point in the future like to include a grid tied solar array. I have looked at a goodman hybrid system that has a condensing furnace as backup since the heat pump is rated for down to -5f.  We do have a handful of days that it can dip below this. Based on my cooling demand a 16 seer 2ton heat pump would produce 24k heating btu’s i believe. This would work well i think if the temp stays 40f and above which is also the temp the efficiency of the heat pump would begin to degrade. 
I am also torn on sizing of the condensing furnace. There is a 2 stage 96% 40k btu or a modulating 98% 60k btu. The modulating version can modulate the gas valve down to a low fire output of 29k btu. 
Given the numbers and application would skipping the heat pump all together and investing the money into an 18seer 2ton  2 stage a/c condensor be a wiser investment. 
I have be happy with the price point of goodman and have had good experience with them so it was my go to. However i am not opposed to change. 

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.


  1. FluxCapacitor | | #1

    There are several mini splits that function quite well at -10F or less, which would negate the need for hybrid or back up furnace .

    Many posters here have had excellent low temp results with Mitsubishi and Fujitsu.

    1. raul4817 | | #2


      It has crossed my mind. Specifically the Mitsubishi hyper heat rated for down to -13f. However there are a few days it can and will drop below this. The other factor is of course price. The Mitsubishi ducted four zone is upwards of 5k. I priced out the Goodman with condensing furnace under 3k. My other concern is size of the heat pump. I would need a 3 ton minimum to get the required heating btus. But this would be grossly oversized for my cooling demands. Forgive me if I am incorrect. I have never even seen a heat pump installed here in Chicago. But after reading some on the site it has intrigued me to say the least. I like the idea of going all electric one day with a supplemental PV array. Electricity around here is around 2.5cents a kwh so doing that now can get pricey but I can imagine in the future electric getting cheaper and greener.

      1. Deleted | | #15


      2. Matthias_32 | | #51

        Do you mean 25 cents/kwh?

  2. Expert Member
    Dana Dorsett | | #3

    >"I have 1100sf of floor space and approx. 13k cubic sq ft including conditioned attic space. Coolcalc Software is giving me 40k +\- Heating BTU’S & 20K+\- Cooling BTU’S . "

    The 40KBTU/hr heating for only 1100' of space simply isn't a credible number, (its outlandish, in fact), especially for space with better than code windows & insulation, and the 20K number is suspect. For reference, my sub-code 2x4 framed ~2400' house with U-0.5ish windows plus 1600' of insulated basement comes in at 40K when it's 0F outside, 68F inside.

    There are plenty of cold-climate ductless & ducted heat pump solutions for a 40K load, but what you really need first is a more accurate load calculation. It's possible, even likely that your real heat load is about 1/3 that number.

    1. raul4817 | | #9


      I may need to consult with someone specific to my hvac sizing. My biggest concern around here is finding someone qualified enough to trust. Rules of thumb sizing in all too common. I originally used loadcalc which gave me a heating btu around 20k which is more in line with your estimation. I then proceeded to coolcalc and came up with a number in the high 30’sk. I added in erv ventilation and voila 40k. Im not sure why there would be such a big discrepancy. I may have made some input errors so i will look into that. Is there any software that i can use for say a per report basis that is reasonable?
      If my heating load is in fact closer to 20k then a 2 ton heat pump would fit the bill.

      1. this_page_left_blank | | #16

        You don't have to find someone locally. As long as you have a set of detailed plans you can email, there's no need for them to be geographically close to you.

      2. GBA Editor
        Martin Holladay | | #17

        Trevor is right. The person who calculates your heating and cooling loads doesn't have to live near your house. For more information on this issue, see "Who Can Perform My Load Calculations?"

  3. FluxCapacitor | | #4

    OK, I didn’t factor in your high (25cents/kW?) energy costs and the higher install costs of a Mitsubishi system Vs Goodman. With that in mind the mini-split system is a harder sell, unless you have a preference for it for other reasons.

    Also, 4 zones of Mitsubishi Hyper Heat for $5K (installed) is insanely cheap.

    BTW some of the Mitsubishi ducted air handlers, (i.e. PEA and MVZ), are available with supplemental resistance heat elements, which guarantees them to function to at least -19, and (as reported here on GBA) even lower? (Somebody help me with the absolute minimum?).

    You could consider “undersizing” a ducted mini-split so that on those handful of days it gets below, say -10F, you supplement with electric resistance heat. Somebody smarter than me (Dana?) probably could do the math if/when that becomes economical. This would also partially help with your concern with too much cooling capacity.

    I’m having a Mitsubishi system installed in a couple of weeks (New Jersey). It will be my only source of heat and cooling, other than occasionally firing up a direct vent gas fireplace. I could maybe save a little bit in energy cost if I added a condensing natural gas furnace, assuming current rates stay the same. But I need the space the gas furnace takes up, and I’m charmed with the mini-splits tech.

    1. raul4817 | | #7

      I the prices i posted are for the units themselves not including labor. I with do most of the leg work in regards to install such bringing power possibly gas to condensing furnace. I have an installer in mind who is willing to work with me on sizing, planning, and installing ductwork as well as installing the line set. I will however do as much i can to lower labor i.e. sealing, insulating ductwork, equiptment hauling, drilling as needed, etc. I really like the idea of a ducted mini split to distribute condition air evenly. The price on electricity is actually low this time of year since demand is lower. Utility companies in my area jack prices as demand fluctuates, high electric rates in the summer low gas. High gas in winter low in summer. Im sure this is common place in most places. Why producing a part of my energy from solar is so attractive,

  4. user-1072251 | | #5

    I build in northern NH, and used oil and gas furnaces for decades, but for the past decade have used only Mitsubishi minisplits, including converting my old New Englander home a few years ago. IMO none of the others come anywhere near close to the quiet, the consistency and year round comfort of the minisplits. Why, especially when you're thinking of installing solar panels that could power your house, would you install a fossil fuel system? To save two thousand dollars? You'll save a lot more than that when you aren't paying anything for fuel.

    1. raul4817 | | #12

      Solar is in the pipeline just not sure how far down. Money and time are both factors. I guess my reasoning was that if it did not happen for say 5-10 years i could use the condensing furnace when needed in and the heat pump predominately , electric rates permitting. When i do eventually have some free electric via my solar i could then switch my basement and 1st floor unit over to heat pump only and use the heat pump side of my hybrid set up exclusively. I have looked at the mitsubishi and am hesitant given the the price yes but as many of you know we did experience some brutally cold -23f temps a few weeks ago and i guess there is a concern to how the minisplit will fair. These temps are not common but -15 is very normal for jan and feb. I am aware there is the electric resistance backup route but without solar up and running that can get expensive quickly.

  5. Expert Member
    Dana Dorsett | | #6

    Seriously, 40,000 BTU/hr of heating load for an 1100' better-than-c0de space isn't possible with the windows closed. That's a load/area ratio north of 35 BTU/hr per square foot of conditioned space. That's the load/area ratio of a 2x4 framed house with no wall insulation and maybe R10 in the attic, leaky single-panes w/ clear glass storms, not the space described.

    1. raul4817 | | #11


      I hear you. Im gonna look at my calcs again. The coolcalc software is not the best when it comes to calculating on a floor by floor basis.

    2. FluxCapacitor | | #13

      I must admit I chuckle when Dana calls out potentially oversize mini-splits as “ludicrous”, “insane”, “outlandish”, and my favorite “grotesque”.

      Conjures up horror movie visions of an evil come-to-life wall unit….shuddering, and spitting out thick fog in a hissing frenzy.

  6. walta100 | | #8

    You made such great choices windows walls and attic insulation please do not shoot yourself in the foot now by putting your HVAC in the attic. It is a very poor choice that is likely to increase your operating costs by 20%. Ductless is great or see if you can fit the unit in a closet and put the ducts in the hallway under a dropped ceiling.

    My wild guess is being as heat rises the upstairs unit will do little heating but will do a lot of the cooling.

    If you want a heat pump to work in zone 5 I think variable speed compressor and electronic expansion valves are must have features. The installer is a more important than the brand name on the equipment.


    1. raul4817 | | #10

      The attic is insulated underside the rafters. I originally framed a utility closet to keep the furnace in my envelope however the wife wanted more floor space so i obliged(and also made her help me demo it). I spent the better half of a week installing vent baffles and air sealing them so i could install r50 of batts underside and essentially place my hvac inside the envelope. Given this is there still a 20% penalty with ducts up there? I still intend on insulating and air sealing ductwork as meticulous as possible.

      1. Expert Member
        Dana Dorsett | | #14

        Ducts in a conditioned attic are not a problem, but a 20% penalty would be on the LOW side if going ahead with a 40,000 BTU/hr non-modulating solution to a 10-13,000 BTU/hr problem.

        With more realistic load numbers you could prove it, but this is likely to be a decent fit for a Fujitsu 18RLFCD ducted mini-split. which would likely have decent margin on the real heating load, yet can modulate down to 3100 BTU/hr @ 47F. It doesn't come with a pan heater for clearing defrost ice during extended cold snaps, and would have to be monitored during Polar Vortex events, but it's way better than any 1-2 stage big air handler "dual fuel" solution with an oversized heat pump married to an oversized gas furnace, or a 1-2 stage ducted heat pump + auxiliary heat strips that does NOT use a vapor injection scroll compressor for extended temperature capacity & efficiency.

        An xxRLFCD series ducted minisplit has a rated output at -5F, but doesn't stop putting out heat at temperatures below that number, but the heat rate simply isn't specified below -5F. It uses the same compressor technology of Fujitsu's true cold climate heat pumps that have specified output at -15F- it's just missing the automated defrost ice management. The defrost ice management/pan-heater is the only real difference between Fujitsu's 15RLS3 and 15RLS3H cold climate version of their high-wall ductless, which is also likely to have margin on your true loads.

      2. walta100 | | #26

        You still get a penalty with a conditioned attic.

        To make the math ease let’s say the roof has a 12-12 pitch. If the addition is 24 feet wide that make the north roof 24 long and the south roof 24 feet long, you have doubled the surface area of ceilings thermal envelope when you move the insulation to the rafters. Again to make the math easy let’s say the walls are 12 feet tall and the whole attrition is over conditioned space. So your envelope is 1/3 larger.

        Now you have a conditioned attic you need to be sure to install supply and return registers in the attic, without ventilation mold becomes a real possibility.


  7. raul4817 | | #18

    The 18RLFCD may work well but what would need to be done in really cold times? Monitoring it but then what turn it off and rely on an auxillary heat source? If i were to say upgrade to something that has integrated defrost management what fujitsu model would do well? There is a AOU18RLXFZH that is rated for -15F. but this is a 2 zone unit. Could i use this with the ARU18RLF air handler or even 2 since its zoned for 2? Would 2 air handlers give me better distribution for the 3 bedrooms or will 1 work just the same without the added cost. I did the efficiency calculator to input my Kwh pricing and therm pricing into the fujitsu webpage and with my sf and assuming i used a condensing gas furnace with matching A/C It was still giving me a %9 savings for the year by going the minisplit route. If cold weather performance could hold up down into the negative teens and i do not have to monitor the defrost issues as closely. Given the price of the fujitsu units it becomes more and more attractive to go this route. Eventually I would like to do the same for the 1st floor and basment as well.

    1. Expert Member
      Dana Dorsett | | #19

      >"The 18RLFCD may work well but what would need to be done in really cold times? "

      If/when you see that the pan drain is plugged with ice, take a ~1000-1500 watt heat source like a hair drier (or install a lower wattage heat tape used for plumbing on the pan) and melt it all out, including any ice sheets that may have started build up in the rest of the pan. Fortunately there isn't much moisture in the air when it's -15F outside, but there is still some. Using a heat tape approach you can just plug it in and go back inside, checking on it every few hours then unplug once the ice has cleared.

      Failure due to ice buildup isn't very common even from those who don't monitor it, but that sort of damage would not be covered by warranty if/when it occurs, and it's not convenient to lose the mini-split on the coldest week of the year.

      Capacity-wise I suspect your actual load at -15F is still less than the (unspecified) capacity of the 18RLFCD. But there is a control output to the RLFCD cassettes for calling auxiliary heat from other sources when it's not keeping up. I have never really looked into how the aux-heat output works, but it's almost certainly not going to be necessary for 1100' of better than code upper floor even at temps well into the negative tees.

      >" There is a AOU18RLXFZH that is rated for -15F. but this is a 2 zone unit. Could i use this with the ARU18RLF air handler or even 2 since its zoned for 2?"

      The xxRLF cassettes are compateable with xxRLXFZH compressors, but the cassettes don't really modulate when tied to a multi-split. With most multi-splits there has to be a minimum of 2 cassettes or head hooked up for it to work at all- you can't just install a single 18RLF onto an 18RLXFZH compressor with nothing on the other zone. The non-modulating aspect of a multi-split solution reduces the efficiency by about 20%. That's anAHRI tested HSPF of 9.0 if using a pair of 9RLFs vs. 11.3 for the modulating single zone 18RLFCD. That's an average difference of only (9.0/11.3=) 0.796 x as much heat per kwh out of the multi-split solution.

      1. raul4817 | | #22


        I could easily defrost the unit by hand but my hesitation is when Im off at work for long stretches(sometimes as long as 18hrs). My wife is typically home but not sure this is something she can or wants to handle while I'm away. I can already see that conversation. "So your saying I have to go outside in the cold to check on the heat pump?"

        Ok so single zone is more efficient and should be more than capable to handle 3 bedrooms especially if using a ducted air handler.

        So then what I need or at least would like is a single zone minisplit with cold temp capabilities and auto defrost, that can be used with a compatible ducted air handler for equal distribution.
        I did search the fujistu website but could not find any single zone just the multi. Does they make one or will I need to look at another brand such as Mitsubishi? Do you have one in mind in the 18-24k btu range. I know that I need to look at my calcs again or employ someone with a little more knowlege.

        1. Expert Member
          Dana Dorsett | | #23

          >"I could easily defrost the unit by hand but my hesitation is when Im off at work for long stretches(sometimes as long as 18hrs). "

          Not a problem. It takes days, not hours, to produce enough defrost ice build up in the pan to interfere with the fan blade. If it's gone a handful of DAYS of continuous low temperatures it's good to take a peek. In most of zone 5 it is somewhat rare to go two weeks where it doesn't rise to the freezing point, but it doesn't have to get that warm to thaw the plug. The at the end of a defrost cycle the defrost water is several degrees above 32F when it drips off the coil into the pan.

          It's possible that your true loads are within the output range of a 1-ton- don't get zoned in on a 1.5 tonner. A cold climate Fujitsu 1.25 tonner is good for about 15,000 BTU/hr @ -15F, and over 18,000 BTU/hr cooling at it's maximum modulation (as opposed to it's "rated" modulation level at which it's efficiency was tested.)

          The single zone ducted Mitsubishis in the right size range don't use the cold-climate "hyper heating" compressors, and don't have pan heater options. The bigger air handler cold climate versions are either non-modulating multi-split. The their smallest fully modulating version is a 2.5 tonner that won't go below 18,000 BTU/hr @ +47F, so oversized it would not modulate with high efficiency until it's very cold out.

          1. raul4817 | | #24

            Thank you Dana,
            I clearly need to read some more into what is really feasible and efficient for me. I am now wondering if I bite the bullet and possibly mini split the entire house in one shot. 1st floor and basement is brick, uninsulated with combined with around 200osf of conditioned space. Would multi split make sense or two separate modulating minis? I'm thinking equipment pricing would be cheaper with a multi. But if it's not modulating then I pay on efficiency. I can go either way start with one and convert over the other half of the house at a later time. I do like the idea of only one heat pump out back vs 2. I appreciate everyone feedback and am feeling more at ease to hear that in zone 6 people are generally happy and problem free with these minis.

  8. user-1072251 | | #20

    I'll just say that dozens of builders in northern NH, which is Zone 6 and occasionally cold, have been using minisplits over the last decade with very few issues.

    1. Expert Member
      Dana Dorsett | | #21

      That's right- most of the RLS2 Fujitsus and FE Mitsubishis installed in New England over a decade ago without pan heaters are running just fine despite benign neglect. But it's important to flag the issue when recommending the RLFCD, to acknowledge that for the low percentage where it damaged the unit it is non-warranteed damage.

      I may be mistaken about the AOU18RLXFWH being capable of running with just a single zone hooked up. The NEEP database has performance data for running it with just a ASU18RLF down at -15F, and has both a mininmum and maximum output & efficiency data at several temperatures. The ASU18RLF is a 1.5 ton high-wall head, not a slim-duct cassette, but they indicate that in that mode it will modulate down to 7000 BTU/hr @ +47F, which is the minimum output of the compressor. That is more than twice the minimum modulation of an 18RLFCD, and probably more than half the heat load of this 1100' floor, an thus not a great choice even if it's possible to hook it up as a single zone. Multi-splits tend to shift output in half-ton steps rather than continuous micro-adjustment modulation the way single zone systems do.

  9. Expert Member
    Dana Dorsett | | #25

    >"Would multi split make sense or two separate modulating minis? I'm thinking equipment pricing would be cheaper with a multi."

    To the contrary- multiple single zone minisplits is usually cheaper than a multi-split of similar capacity, but the visual aesthetics and space requirements for a half dozen compressors can be a problem.

    Before buying any equipment it's important to get much more accurate load calculations than what your coolcalc run coughed up. Modulation ranges aren't infinite, and oversizing by more than 50% will usually be less efficient as well as less comfortable than right-sized equipment.

  10. Expert Member
    Dana Dorsett | | #27

    Raul: I've been informed via this thread that there are configurations with Mitsubishi MVZ air handers where they can be set up 1:1 with a single zone on the multi-split, if only via a branch box, but that configuration isn't tested for efficiency/capacity:

    It looks like the compressor will modulate with load in those configurations, but the MVZ air handler doesn't.

    The minimum modulation is a bit on the high side for your likely load upstairs, but it may do OK with separate MVZs for upstairs & down, once you get a better handle on the real load numbers.

  11. raul4817 | | #28


    I have done two more DIY load calcs
    Build it solar -23k heating 21k heating 14k cooling

    These both seem to be more in line with your original inclination. I did look at the mvz last night it did intrigue me. If I do end up going with a mini split or multi mini I defiantley see the importance of correct sizing and layout as well. I also noticed the cfm on most of the ducted air handlers do go beyond 700-800 cfm. Thus the the extra care needed in sizing and planning for ductwork(manual D). Beyond paper and pencil is there a software that you would recommend for manual j/s/d? I am fairly familiar with the concepts and know my construction details better than anyone seeing as I did the work. I Would not be opposed to even spending a few hundred dollars on software that could easily save me a thousand from paying a mechanical engineer or hers rater. My philosophy is any money saved in labor or in this case design is money spent on better equipment or other equipment like solar panels. Is this feasible or would calculating this myself be a lesson in frustration?

  12. Expert Member
    Dana Dorsett | | #29

    Loadcalc can get you close if you're aggressive on all assumptions. Most of the time if you subtract all of the infiltration and ventilation load from it's only 15-25% more than reality. A 21K load for 1100' of code-min space (19 BTU/hr per square foot of space) is high if your 99% outside design temp is around 0F, instead of -25F. How much of the load did the tool ascribe to infiltration? And, what is your 99% outside design temp?

    Wrightsoft is a popular pro-tool, but it's not just a few hundred USD. The basic residential package runs a bit less than a grand, and there is learning curve to it.

  13. raul4817 | | #30

    I used -10f for design temp.
    95f design temp for cooling @45% humidity

    I used 45cfm for fresh air infiltration/ventilation
    sensible 740
    latent 960
    heating 5733

    I used conditioned sf of 1150 closer to 1128 but i rounded up maybe i should round down
    cubic sf 8900 below drywall ceiling
    cubic sf 2700 condition attic space not sure if i should have omitted this one
    4 people
    2 appliances washer dryer
    triple pane low-e @ 185sf

    one thing to note is i used insulation under rafters @r-38. Highest it will allow me to go. I am at an R50.

  14. Expert Member
    Dana Dorsett | | #31

    >"sensible 740
    latent 960
    heating 5733"

    Those are the results from which tool?

    An outside design temperature of -10F is extremely rare for zone 5. Where are you?

    A design heating load/area ratio of 5-6 BTU/hr per square foot is possible with U0.19 windows and a limited amount of window area, but something seems amiss here.

    185 square feet of U0.19 window would be about 2500-3000 BTU/hr at an 80F indoors to outdoors temperature difference (-10F outside, 70F inside), something like half the total heat load calculated.

    An R50 roof with some thermal bridging typically comes in with a U-factor around o.025 BTU/hr per square foot per degree F. Continuous R50 would be under U0.020. (An R38 roof typically comes in around U0.030.) So assuming 1200 square feet of roof and an 80F temperature difference you'd be looking at roof losses of 0.025 x 80F x 1200'= 2400 BTU/hr.

    Adding window losses to the roof losses you're at 4900-5400 BTU/hr, which isn't leaving much wiggle room for wall losses.

    If using loadcalc or other online tools, set the air infiltration and ventilation to zero. But assuming no heat recovery 45cfm x 60= 2700 cubic feet per hour, x 80F x 0.018 BTU/ per degree per cubic foot comes to 3888 BTU/hr.

    Add that to the 5400 BTU/hr high-side estimate and you're at 9288 BTU/hr before accounting for the wall losses.

    Taking a WAG that including gables etc you have no more than 750 square feet of wall, and without knowing your actual stackup and thermal bridging for the "R26" walls let's call it U0.040 for wall losses of: U0.040 x 750' x 80F= 2400 BTU/hr

    Add it all up and you're at 11,688 BTU/hr, and that's probably a bit higher than reality, since I'm only guessing on the wall U-factor and area, and aren't properly modeling the non-linear aspects of the window U-factor at the lower -10F outdoor temperature. (U-factors for windows in the US are calculated at 0F outdoors.)

    The Fujitsu 12RLFCD delivers 15,000 BTU/hr @ -5F, and would almost certainly cover your load with some margin. The 18RLFCD delivers about 18,400 BTU/hr @ -5F, and would cover your load a -10F with about 50% of margin, even though they don't specify it's capacity at -10F. See the capacity tables on p16-P17 (PDF pagination) in this manual:

    1. raul4817 | | #32

      Those are the results from which tool? These are from

      Where are you? Just outside of chicago

      Taking a WAG that including gables etc you have no more than 750 square feet of wall? I have 1260sf of wall plus 120sf on the gables. I just added window sf and it is actually lower @150sf

      This morning i did a manual j on coolcalc Using a room by room Map&trace with all windows in there correct positions.
      @-10f/95f heating 25k cooling 8K 351cfm. no ventilation &Very Tight infiltration rates
      @0f/90f Heating 21k cooling 7K 310 cfm. no ventilation &Very Tight infiltration rates

      My 2nd heating demand seems to close to both build it solar & loadcalc. However my new cooling load has been cut in half?

      In regard to ducting Ideally i would like a register in each of the bedrooms and baths. 1 near the staircase/foyer and one in the laundry. Maybe just one for both foyer/laundry area?
      This would amount 7-8 registers in total and as many return as needed as well. With only needing low 300cfm's the ARU12RLF or ARU18RLF should have me covered. I plan on centering the unit so longest runs should be under 20ft

      Attached is a floor plan for reference

      Also for good measure I took a look at my original hvac design from my architect. Specs are calling for a 125k btu high efficiency condensing furnace with 2200CFM! "outlandish"


  15. Expert Member
    Dana Dorsett | | #33

    >"Where are you? Just outside of chicago"

    Chicagoland 99% outside design temps are in low negative to low positive single digits, not -10F. Any reasonable oversizing factor for the load at your ACTUAL design temperature would have you covered at -10F and beyond. Calculating the load at something other than the 99% design temp just makes the discussion confusing.

    With a better description of your wall and roof stackups ( all layers & materials, not just the insulation layers) we can come up with more accurate U-factors to use in an I=B=R type load calculation, which would be fine for sizing the equipment, and more accurate than selecting from less-appropriate menu options on loadcalc/coolcalc. (I usually maintain an IBR heat load spreadsheet on projects that I get involved with to keep track.) The basic methodology for running those numbers is explained in this now-classic GBA blog series:

    The parallel path method of calculating U-factors is explained here:

    and here:

    >"I took a look at my original hvac design from my architect. Specs are calling for a 125k btu high efficiency condensing furnace with 2200CFM!"

    That's enough furnace to theoretically heat my whole sub-code antique 2400' house + 1600' insulated (to-code) basement at temperatures cooler than -120F- temperatures not seen in my neighborhood since before the last ice age. Specifying that much heat for a better than code 1100' upstairs zone is pretty gonzo, but (alas) not uncommon.

    1. raul4817 | | #34


      I dusted off the calculator this morning after reading the i=b=r method.

      Here is what i came up with and my stack up from inside to out

      WALLS 1380sf
      2x6 16"O.C. 5/8 gypsum, 2.5" service cavity, continuous R3 eps foam board air sealed+ thermal break,2x6 r23 cavity insulation, 5/8" ply, tyvek wrb, vinyl siding.
      total R26
      R3 thermal brake at studs
      R8 foam board at all headers + overlapping r3 continuous foam board

      CEILING/ROOF 1200sf
      fabric type radiant barrier R1, R3+2x4 bonfig strips to deepen rafter bays & provide thermal break, r46 batts in deepened bays, R3 foam board vent baffles, 1.5" soffit to ridge air gap, 5/8" ply sheathing, Synthetic Roofing Underlayment, self adhered leak membrane last 4ft at edges of roof, asphalt shingles.

      3/4 PLY, R50 batts, 1/2 ply air barrier.

      at -10f design temp

      U temp
      Windows 150 sf .18 80F 2160 btu
      wall cavities 1050 sf .038 3192
      window headers 50 sf .07 280
      walls studs/plates 250 .10 2000
      cantilever 200 .02 320
      rafter bays 1075 .02 1720
      rafters studs 125 .05 500
      infiltration +3,888K
      12,260K total

      as far as infiltration I used your methodology
      "no heat recovery 45cfm x 60= 2700 cubic feet per hour, x 80F x 0.018 BTU/ per degree per cubic foot comes to 3888 BTU/hr."
      I am however looking at erv's not sure how this will affect my bottom line
      U factors where calculated by using 1/rvalue.
      For dimensional lumber rvalues i used R1.41 per inch and figured the wall assembly studs to account for aprox 20% of total sf after subtracting windows and their headers.

      My u factors may not be dead on and my infiltration rates possibly skewed but this puts me in the range of a 12k or 15k ducted mini.
      I still like the idea of the fujitsu XLTH line.
      in particular the AOU12RLS3H. max output @16k heating and modulating down to 3k I believe. But is this unit compatible with the ARU12RLF ducted cassette?
      If planning on using the ARU12RLF must i use the AOU12RLFC outdoor unit only?
      I gave fujitsu a call this morning but only got there voicemail.

      all of my calcs are based on a design temp of -10 but i do realize that at 0F my btu need be reduced.
      I feel like i am close to a realistic number now.

  16. Expert Member
    Dana Dorsett | | #35

    >"12,260K total"

    This is totally -18RLFCD territory, even though the capacity is not specified at -10F. The -12RLFCD may still have the capacity at -10F, but with less margin. Either one will still modulate down to 3,100 BTU/hr @ +47F.

    >"I still like the idea of the fujitsu XLTH line.
    in particular the AOU12RLS3H. max output @16k heating and modulating down to 3k I believe. But is this unit compatible with the ARU12RLF ducted cassette?"

    The AOU12RLS3H compressor is ONLY compatible with the ASU12RLS3 head.

    >"If planning on using the ARU12RLF must i use the AOU12RLFC outdoor unit only?"

    The ARU12RLF is compatible with any of Fujitsu's cold climate -RLXFZH multi-split compressors, but you will lose most of the modulation and quite a bit of efficiency if going that route. If going multi-split it will be more efficient if there are other zones with right-sized heads/cassettes on the system to reduce the amount of on/off cycling the compressor experiences at part load. To run the ARU12RLF on a single zone compressor you're limited to the AOU12RLFC.

  17. raul4817 | | #36

    so either I go with the 18RLFCD for second floor only and a possibly a 2nd 18RLFCD for 1st floor and basement, or go possibly a 36RLXFZH multi split for entire house zoned properly.

    Modulating on the rlfcd is at 3k
    What would each zone of the RLXFZH modulate down to at?
    from an efficiency perspective would having one RLXFZH unit with minimal cycling running be better than two RLFCD units with there low modulation?

    My numbers for first floor and basement are theoretical. I will be upgrading windows and insulating the basement within the next 5 years. I will run some numbers tonight as is and as upgraded. My first floor is un insulated plaster walls with a multi-wythe brick assembly. 3.5" depth on bricks with 1" cavity. What u-factor would i use for this wall?

    I'm thinking that multi zone for this space is almost a must given the conditioned sf is close to 1800sf. In theory i could go with the 18RLFCD for the second floor then separate the 1st and basement floors with a RLXFZH multi split. In cold weather events my RLXFZH will keep up and most the heat should convect upwards to the 2nd floor for some added comfort just in case the 18RLFCD starts the studder in -10F.

  18. Expert Member
    Dana Dorsett | | #37

    >"What would each zone of the RLXFZH modulate down to at?"

    The minimum modulation of the 18RLXZH and 24RLXZH a is 6200 BTU/hr @ +47F, but they modulate up/down in much bigger chunks than a single zone mini-split does. I'm not sure it they can split that 6200 BTU/hr evenly between two zones, but I suspect not. The minimum modulation of the 36RLXZH is 12,000 BTU/hr @ +47F.

    The max capacity of the 36RLXZH down at -15F is only about 22KBTU/hr, so it may not really be the right solution. You would almost certainly get more than 30K out of a pair of 18RLFCDs at that temperature, but since it's not tested and rated by the manufacturer at that temp we really don't know. At -5F a pair of 18RLFCDs is good for ~37,000BTU/hr, which is more than the 36RLXZH can deliver even at +5F (10F warmer!), according to data on the NEEP spreadsheet.

  19. raul4817 | | #38


    Wow, so the RFLCDs actually perform better than fujitsus RLXZH cold weather line up in the cold! So what makes the RLXZH better? The integrated drain pan heater?

    I ran some numbers last night on 1st floor uninsulated brick with the windows i propose one installing and the basement that will be insulated with new windows

    1st 21k
    basement 8k

    with that being said i would/could go with the AOU36RLXFZH splitting zones to 24k and 12k. or two more 18RLFCDs with one unit exclusively for the 1st floor and some ducts from the other unit directed to the 1st floor. That would however park 3 compressors our back! but i will cross that bridge when i get to it i guess.

    I took a look at the submittal sheets for the 18RFLCD and was trying to figure the energy consumption. I found this

    Cooling……………………………………………………………………..… 6.6 A
    Heating………………………………………...…………………...…............. 7.3 A
    Input Power
    Cooling…………………………………………………….……....………. 1.50 kW
    Heating……………………………………………………………..….…... 1.67 kW

    SEER…………………………………………………………………………... 19.7
    EER (cooling)…………………………………………......………….. 3.52 kW/kW
    COP (heating)…………………………………………….......………. 3.79 kW/kW
    HSPF (heating)……………………………………………….…..…… 11.3 Btu/hW
    Moisture Removal……………………….………………… 4.2 pints/h / 2.0 liters/h

    If running at max output for 24hrs @30days what would my kW usage look like? Just trying to assume the worst?
    I calculated like this 3.79kw x 24= 90.96 X 30days = 2728KW. I may be off im sure.


    1. Expert Member
      Dana Dorsett | | #41

      >"If running at max output for 24hrs @30days what would my kW usage look like? "

      You'd roast/frost yourself to DEATH if running it at max output for even 12-24 hours, let alone 30 days. The peak power numbers are only useful for determining what wire & breaker sizing is needed.

      >"Wow, so the RFLCDs actually perform better than fujitsus RLXZH cold weather line up in the cold! So what makes the RLXZH better? The integrated drain pan heater?"

      The pan heater is what makes it a "...cold weather..." multi-split, but it's still a multi-split. Nothing makes the RLXZH better. It's a multi-split, slave to multiple masters, with more internals to monitor and control. They use similar compressor technology in the RLFCD that gives it greater low-temp capacity & efficiency than simpler compressors, but the fact that it's a single zone system designed for a very particular cassette, not a wide selection of possible configurations it has less to do and can be fine-tuned by design.

      1. raul4817 | | #43

        "You'd roast/frost yourself to DEATH if running it at max output for even 12-24 hours, let alone 30 days." theoretical of course but still i was curious. My wife probably wouldn't mind. She is often cold with the t-stat set to 73f! we do however have some drafty spots throughout the 1st and basement so I compromise for now until i can seal up and upgrade windows and insulate the basement.
        I feel like if fujitsu made the XLTH line in a single zone or the RFLCD line with something as simple as an integrated pan heater it would maybe be more attractive to consumers in zone 5 and up. I not an engineer at fujitsu so may be something more complex.

        Your heat tape idea from earlier is very clever. I could easily install a switch or even a timer somewhere near my covered deck or perhaps even inside and solve this issue with very little leg work and the cost of the extra labor/material would be offset im sure in energy savings by using a single zone vs multi. I am seriously considering zoning each level with it own separate condenser. The condenser itself is already kind of slimlined vs traditional ac condensers that we are accustomed to so an extra few feet of space in the backyard may just be worth it to have a fine tuned hvac system.

        1. Expert Member
          Dana Dorsett | | #45

          The maximum setting in heating mode for most mini-splits is usually something like 85F, but if you hard-wired the system to run at maximum speed independent of the setting it can get much warmer than that.

          I too wish they would make an xxRLFCD-H series with integrated pan heaters, but they don't.

          Fujitsu's initial foray into "eXtremely Low Temperature Heating" was to slap a pan heater and crude controls onto their very popular and successful xxRLS2 mini-splits something like 7-8 years ago. Their pan heater control on that system was too crude and a power hog that ate into their HSPF numbers. About a year or two later they halted production on the RLS2 & RLS2H, and released the more efficient RLS3 & RLS3H. The RLS3H used a much better control scheme on the pan heater, cutting average power use by the heater by 5x or more compared the the RLS2H.

          A single zone mini-split per floor is a fairly common configuration. It usually allows you to turn the upstairs mini-split completely off during the heating season shoulder seasons, and the downstairs mini-split off during the cooling season shoulder seasons. Having that extra bit more load for the mini-split that stays on during those low-load weeks is more efficient than having two mini-splits doing a lot of on/off cycling.

  20. STEPHEN SHEEHY | | #39

    Raul- I think your maximum power use would be using the listed input of 1.67 kw, not the COP figure. I get about 1211 kwh. It's very unlikely you'd be anywhere near that in real life. Your average January temperature should be way above your 99% design. You'll never run full blast for thirty days.
    My design temperature is about the same as yours. Today is miserably cold, about 15° with a 30-40 mph wind. But it's sunny and my minisplit isn't running at the moment.
    I roughly estimate that my 1700 square feet of conditioned space uses about 3500 kwh per year for heat. Zone 6, 0° design temperature.

    1. user-723121 | | #40


      How many heating degree days for your location? You are somewhere in the 1 Btu/sf/hdd range, that is a great efficiency number. What efficiency gain are you getting with the minisplit, how much electricity would you use if you heated with electric baseboard only?

      1. STEPHEN SHEEHY | | #42

        Doug: Average heating degree days here is around 7300. I assume an average COP of about 3. If my estimate of around 3500 kwh is right, using resistance electric would mean around 10,000 kwh per year. In fact, our total electricity use is about 10,000 kwh per year. A hot tub and resistance water heater use quite a bit, but our solar panels generate around 8,000 kwh. House is 100% electric.

        1. Expert Member
          Dana Dorsett | | #44

          >"I assume an average COP of about 3.

          The HSPF of the 18RLFCD is 11.3, which is an average COP of 3.31.

          The HSPF of the 36RLXFZH running only ducted cassettes is 9.3, which is an average COP of 2.73.

          In real world use you can hit as-used numbers both above or below the rated HSPF, depending on how (sub) optimal the sizing was, and the average wintertime temperatures of the location. In most of US climate zone 6 it's possible to beat the nameplate label on efficiency by a small amount when sized optimally for the load.

        2. user-723121 | | #47

          Stephen and Dana,

          Thank you for that information, I am starting to get a feel for heat pumps and what they can do in a fairly cold climate. Turning 1 kWh into 3 is quite exciting and really changes the discussion on how to heat and cool off the natural gas grid (and maybe on the grid). Back in the day when we did our own heat loss calculations we would consider the number of occupants, type of lighting, glazing and all the rest. If one really pays attention to gas usage and electrical as well, changing all lights from incandescent to LED will show up as less kWh used but a slightly higher gas therm usage.

          To be fair, the source of electricity must be considered with loss at the generation plant along with transmission loss in getting to your house to really put a Green label on any residence. I am encouraged by all of the wind power coming online in the Midwest, the wind does blow a lot here and not all from leadership. Solar panels eliminate so much of the typical electrical loss, they have to be given strong consideration.

          Stephen; Any specifications you want to share on your house, walls, ceiling, foundation, windows, ACH50?

          Thanks, Doug

          1. Expert Member
            Dana Dorsett | | #48

            >"To be fair, the source of electricity must be considered with loss at the generation plant along with transmission loss in getting to your house to really put a Green label on any residence. I am encouraged by all of the wind power coming online in the Midwest, the wind does blow a lot here and not all from leadership. Solar panels eliminate so much of the typical electrical loss, they have to be given strong consideration."

            If digging around archive sites you can find the Clean Power Plan (CPP) documents it's possible to get at least some handle on it, as outlined in this bit o bloggery from three years ago:


            For most states the CPP was more of a stripe on the floor than a hurdle, and the landscape regarding renewables costs have shifted toward the cheaper, inducing private investment beyond state targets rendering the CPP's targets & mandates, and the CPPs assumed business-as-usual cases dirtier than what's likely to actually happen. Even in conservative Indiana some large utilities are seeking to build out wind power in excess of mandates to be able to avoid the cost of maintaining & fueling existing coal plants:


            Within the lifecycle of a heat pump the grid in most of the US is going to get dramatically cleaner.

          2. STEPHEN SHEEHY | | #49

            Doug: the house is quite tight (.59ach50) and well insulated. Under slab is about R- 18 (4" reclaimed XPS ); walls are double stud, around 12" thick, about R-42 (dense packed cellulose outer 8.5" , fiberglass inside inner stud wall); ceiling about R-70 or so ( cellulose).

            Intus windows, triple pane tilt/ turn. Approximately U-.12-15 (European method of showing U factor differs from ours. Center glass reported as .088).

            House is about 1700 square feet of conditioned space. 6.6 kw of rooftop solar.

    2. raul4817 | | #46

      3500kwh a year is encouraging to say the least. I definitely feel im going to either break even if not possibly save a few. Im also eager to see how my energy consumption will be reduce in the cooling season with the fujitsu. I currentley have an energy hog out back so savings should be significant. I made a call to my hvac salesman this morning to inquire about the minisplits he sells. He asks me “what are going to do if the grid goes down.” Well if i had a condensing furnace with no power i figure id be in the same boat. Cant wait to have the conversation with my tin snipper. Glad to hear you and so many others are having really positive results with the minis.

  21. user-723121 | | #50

    Stephen; If your COP is near 3 I think you are using less than 3,500 kWh annually to heat your house, here is why. 3,500 kWh x 3412 is 11,942,000 Btu's divided by 7,300 hdd and 1,700 SF is .96 BTU/sf/hdd. This times 3 or if heated by electric resistance is 2.88 BTU/sf/hdd. Your house is more efficient than that I would say by a factor of 2.

    Superinsulation for any cold climate and you easily meet the definition will consume about 1.5 BTU/sf/hdd. This was where the double wall homes I built in the 1980's came in. They were a bit larger than your home but only 1/2 as airtight. Could you be spending a bit more time in the hot tub than you are letting on?

    1. STEPHEN SHEEHY | | #52

      Doug -you might be right. It's hard to estimate our COP.

      We'll give up our ( mostly) solar powered hot tub when they pry it from our warm, wrinkled fingers.😊

  22. ERIC WHETZEL | | #53


    In case you're still getting quotes, I would contact:

    Mike (Compass Heating and Air) 630-504-8688

    He installed the Mitsubishi ductless mini-split system in our Passive House in Palatine (suburb of Chicago).

    He was helpful, knowledgable, professional, and we're really happy with our system. He even followed up with us before, during, and after the recent Polar Vortex to make sure we were alright.

    There's only a few subcontractors that we used during our build that we would feel comfortable recommending to family or friends and Mike is definitely one of them.

    Good luck with your project!

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |