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 Bad Sizing?

bmul23 | Posted in Mechanicals on

I have a ~2500 sq foot colonial with no AC and hydronic nat gas based heating. We’ve decided to put in one or two heat pumps with ducted air handlers. Both air handlers would be in unconditioned spaces (basement and attic) and cover the respective adjacent floors. We got several quotes from different contractors and all of them sized our house with a 5 ton multi-zone Mitsubishi MXZ-SM60 unit (non hyper heat). After reading the great resources on this website and getting feedback from a local non-profit I suspected the heat load was far less. None of the contractors will do a full room by room manual J so I took it into my own hands. I did a manual J using an online tool (load-calc) and used my natural gas usage data with the boiler sizing article ( from here. All come out to between 38 and 40kBTU/hr at the design temperature in my area. 

To be clear, my goal isn’t to 100% replace my hydronic system but rather reduce it’s use to colder weather and hot water (indirect tank). I suspect the natural gas numbers align with the load calculation because they don’t consider that the new ducts and air handlers will be located in unconditioned space. I’ve read that the losses from the air handlers through the ductwork can be 15 to 40%. 

Question 1) Is the 5 ton unit correct because of these losses? The calculation I did, did not ask about location of ducts.

Question 2) The turn down ratio (modulation) of the MXZ-SM60NAM (60kBTUh to ~16kBTU) vs MXZ-SM60NAM2 (60kBTUh to ~30kBTUh) seems to have dramatically changed for the cooling according to NEEP . Why is this? 

Question 3) Because of the above info and the concern about AC not being able to modulate in the fringe months, I’ve considered instead going with a 4ton hyper heat multi unit. 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.


  1. paul_wiedefeld | | #1

    Bad sizing for sure. Your fuel use calcs are more accurate.

    You could install 1 system per floor, like the below 24kbtu hyper heat option. Combined, it turns down to 16kbtu, but individually each can turn down to 8kbtu. The 24kbtu has much better dehumidification than the 18kbtu option, so that’s why you’ll be slightly oversized. The minimum turndown for the 24kbtu is lower than the 18kbtu so it doesn’t matter.

    You’ll be installing new ductwork, make sure it’s tight and well insulated.

    1. bmul23 | | #3

      I'll look into that configuration. I think our main concern is that the cost goes up dramatically for the two systems and the modulation on the 48kBTU hyper heat goes down to ~16kBTU when cooling. Our cooling needs heavily favor the second floor due to sun exposure while the first floor has a porch that shields it from the sunny side. One of the reasons I like the bigger unit is that we could split that 16kBTUh cooling at the low end to perhaps 4kBTUh on the first floor and 12kBTUh on the second. If it was a separate system that met our heating needs then even 8kBTU might cause cycling on the first floor. Does that make sense?

      1. paul_wiedefeld | | #5

        That's surprising to hear actually! Here, the equipment is actually the same price, and both installations involve two indoor air handlers. So the second outdoor unit installation is driving up the cost that much?

        1. bmul23 | | #8

          For non hyper-heat units it's about 1k per unit which I believe is because of electrical work that needs to be done in our panel. We don't have a lot of free slots so it's going to be a bit of musical chairs.

          1. Expert Member
            Akos | | #12

            I'm with Paul and the two sperate outdoor units. Same BOM cost, better performance and a bit of redundancy.

            You can even save a bit of cost by not getting a hyper heat for upstairs as a fair bit of the heat will be provided by the main floor unit and the 2nd floor one will be doing mostly cooling.

            Check if you panel supports tandem breaker, I've use something like a Seimens Q21520CTNC quad when tight on space.

            It is very hard to have a basement that is not part of the conditioned space, wether you like it or not, it is part of your thermal envelope. Basements are one of the best locations for HVAC equipment.

            Keeping it out of the attic is hard in case of retrofit, I would like at least having the indoor unit inside the house and only running the ducting through the attic. This is much easier to seal up and makes service and filter changes a breeze. Some of the slim ducted units can be mounted vertically so it only takes about a 1'x3' footprint in the back of a closet.

            All the residential Mitsubishi SM units are the same physical unit. The only difference between the sizes is different firmware, doesn't matter which one you get really.

  2. walta100 | | #2

    You can use fuel usage and weather data to very accurately size the heat pump for the heating load.

    Over sizing is a cooling issue in that the larger equipment doesn’t run long enough to remove enough moisture so the home feels cold and clammy.

    If you select 2 outdoor units my guess is they will have enough modulation run without cycling.

    If at all possible avoid putting the equipment in the attic.

    1. bmul23 | | #4

      Unfortunately we can't avoid putting them in the attic but they will be insulated. Is short cycling really that bad for efficiency or is is mainly just for humidity?

    2. orange_cat | | #7

      BEcause I was not sure of my manual J calculations, I did a little research on the impact of incorrect heat pump sizing and performance. It is more complicated than the simple logical answers - at least in studies. For example, there are studies that show that oversized heat pumps may handle humidity better:

      "Another somewhat surprising result is the impact that oversizing has on space humidity levels. The number of hours exceeding 55 % RH decreases when the unit is oversized, especially in Houston. This reduction is primarily due to effect of duct leakage: the larger unit size reduces the fan runtime thereby reducing the amount of moisture brought into the house via return air leaks. A secondary factor in lowering humidity levels is the increases sensible gains (or heat) provided by the increases fan power for the oversized unit. "

      "Field testing and simulation analysis have been used to assess the impact of over-sizing on energy use and space humidity levels. Sonne et al. (2006) changed out oversized air conditioner units in four Florida houses and replaced them with units sized according to ACCA Manual J (ACCA, 2011a). Detailed performance data was collected both before and after the right-sized unit was installed. Their study found mixed results in terms of seasonal energy use and space humidity levels. In some houses energy use was higher, in some it was lower, and in others the results were inconclusive. Similarly, relative humidity (RH) appears to be either slightly higher and or unchanged after the right-sized unit was installed. They
      also speculated that duct leakage impacts were greater for the right-sized unit since longer periods of system operation were required to meet the same load. More duct leakage increases the thermal losses to the attic (supply ducts are colder for longer ‘on’ periods) and brings in more fresh air into the system.
      Both these effects increase the sensible and latent loads imposed on the system.
      A simulation study by Henderson et al., (2007) also confirmed the modest and somewhat unexpected impact of oversizing. They found that, when 20 % duct leakage was factored into the simulations, both energy use and space humidity levels were only slightly affected, even when both latent degradation effects and part load cyclic efficiency losses were considered. For example, oversizing by 30 % in Miami for the HERS Reference house increased energy use by only 2 % and actually resulted in slightly lower space humidity levels."

      (If you want to go down this rabbit hole go to and search for heat pump sizing, etc).

      1. bmul23 | | #11

        What did you end up doing with your house (load calculation estimate vs install)?

  3. walta100 | | #6

    There are a lot of people that “feel” and” think” short cycling is uses much more electricity but I have not seen laboratory testing that would show it to be a fact. When I look at charts showing power usages the startup power spikes are so short, they are almost nonexistent and irrelevant in a 24-hour chart.

    I do feel equipment with long run cycles will last much longer than equipment with short run cycles.


    1. bmul23 | | #9

      That's what I suspected, as It's kind of like amplitude vs duty cycle modulation. I understand that short cycling would proportionally dehumidify less than constantly running but I've always struggled with the power reasoning.

  4. bmul23 | | #10

    My preference is to go with a 48kBTU hyper heat multi-zone unit as it seems more appropriately sized and can modulate down to 16kbtuh. However, I'm concerned that with duct loss (because they're in unconditioned spaces) I'll be under the heat load of my house. Any thoughts on this?

    1. paul_wiedefeld | | #14

      Eh. It’s definitely real, but it’s usually harped on by the hydronic and ductless installers. That’s why you get huge ranges. Your existing system also has “duct losses” they’re just in pipe form! Are they insulated? Are you using high temperature water? The ducts will likely be more insulated and use a lower temp, although with higher surface area.

      1. bmul23 | | #15

        That's a good point. Is there a ball park loss I should assume in the ductwork? Does manual J include this?

        1. paul_wiedefeld | | #18

          If the fuel lose calcs come in at 40kbtu, a 20% buffer (which is probably 25+% due to the boiler’s inefficiency and piping losses) should be more than enough.

          1. bmul23 | | #19

            Thanks Paul. I just did a detailed room by room calculation out on cool-calc with exact window measurements and such (including ALE from blower door test). Heating load came out to 42.5kBTU/hr including a 15.5% loss due to ducting while the cooling came in at 48.8kBTU/hr including a 13.3% loss due to ducts. I'm hoping this is as reasonable as possible

  5. walta100 | | #13

    Almost every post from an unhappy mini split owners have 2 things in common the systems have multiple heads and are oversized.

    I thought the heat pump was a supplemental heating system and the current boiler would take care of the coldest days?


    1. bmul23 | | #16

      That's still correct, but if the cost difference is minimal to be able to theoretically use exclusively the heat pump system I would prefer that.

  6. peter766 | | #17

    it's totally valid to be concerned about the size of the unit! 😅 the 5 ton might be a tad overkill based on your manual J, but considering duct losses in unconditioned spaces, they might be onto something. it's often that professionals play it safe with a bit larger units to avoid underperformance, but energy efficiency can get compromised.
    oh, the turn down ratio thing can be tricky, huh? 🤔 it seems like the different modulation in the two models might be related to updated tech or different compressor characteristics. checking with the manufacturer might clear that up. it’s kinda important to make sure that your system is capable of handling both mild and peak conditions efficiently.
    considering a 4-ton hyper heat does make sense in some scenarios. 🧐 balancing between ensuring enough heating capacity and not oversizing too much for the cooling season is the tricky part. if you’re living in a climate where winters can be fierce, the hyper heat can be a decent move to ensure warmth without completely relying on the gas heating.
    always ensure the system is capable to provide comfort year-round and it might be worth bringing in an independent consultant for a second opinion on the size and type of unit, especially when you’re getting varied advice! 🕵️‍♂️ good luck navigating through this HVAC puzzle! 🍀

    1. bmul23 | | #20

      Thanks Peter. Yeah, it's frustrating trying to find a unit that meets all the requirements. I would be fine with the 5 ton unit except that when they went from MXZ-SM60NAM to MXZ-SM60NAM2 the cooling modulation became terrible at the low end. That's the main driver for the 4 ton hyper heat.

  7. bmul23 | | #21

    What I can't reconcile is why the numbers for the MX-SM60NAM2-U1 from NEEP (max/min output) don't line up with the Mitsubishi numbers at all?
    Mitsu - Page 90/91
    Attached file or:\M_MXZ-SM_ENGINEERING.pdf

    1. bmul23 | | #22

      It seems like Mitsubishi lists Max, 75%, 50% 25% and then Min capacity. The 25% is blank because it can't modulate that low but the minimum is much lower than the 50% capacity. It appears NEEP lists the lowest % number, in this case 50% instead of the actual min.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |