GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted
Building Science

Planning a Furnace-to-Ducted Heat-Pump Retrofit, Part 2: Data Collection

Answer these questions before making the switch from a gas or oil furnace to a heat pump

A Mitsubishi air handler works with outdoor minisplit units to provide ducted heating and cooling in whole-house situations. Image Credit: Carl Seville

Replacing fossil-fuel furnaces with fully ducted cold-climate heat pumps is a scalable strategy for electrifying and decarbonizing tens of millions of housing units in the U.S. In Part 1 of this series, I discuss the advantages of ducted “vertical” and “multi-position” air handlers (“air handlers” for the sake of brevity) compared to ductless minisplits. Here, I will explain five categories of data needed to plan a successful furnace-to-air handler retrofit: customer goals, building envelope, electric service and loads, ductwork, and physical space.

Assessing customer goals

Like any home improvement project, a successful heat-pump retrofit begins with clear goals. I like to ask open-ended questions to better understand what my customers hope to accomplish. Questions include:

GBA Prime

This article is only available to GBA Prime Members

Sign up for a free trial and get instant access to this article as well as GBA’s complete library of premium articles and construction details.

Start Free Trial


  1. charlie_sullivan | | #1

    Another great article. We are going to need hundreds of thousands of people to learn the skills you are teaching here.

    Your comment about addressing the duct sizing issue simply by replacing an oversized furnace with a correctly sized heat pump is spot on. There can be a tendency for the degree of oversizing to creep up over the years. For example, a house with no air sealing, minimal attic insulation and single-pane windows starts out in 1960 with a 60 kBTU/h input, low efficiency furnace. Over the years, it gets replaced with slightly larger units by contractors who want to be conservative, and the new units are also higher efficiency so the output goes up even more than the nominal size. Meanwhile double pane windows, better attic insulation, and some air sealing decrease the load. 60 years later, it's 3-4 X oversized.

    In one house I looked at, the HVAC contractors' first reaction was that it couldn't possibly work with the existing ducts, because the heat pump would need more air flow. But the oversized furnace meant the ductwork already had more flow than it should (and was annoyingly noisy). A right-sized heat pump would solve that problem, not create a problem.

    Another strategy that I've considered (but not done) is to devote the ducts to just one portion of the house and install minisplits for the rest, so that the main trunk ducts and the returns are plenty big. This particularly makes sense if some of the ducts are poorly sealed and insulated and are hard to access. For example, run from the basement to the second floor through 2x4 exterior walls. Have you considered or taken that approach?

    1. GBA Editor
      Kiley Jacques | | #2

      This is all such good information, Charlie. Thank you for sharing it. You have given me an article idea. I really value the conversations that come out of Jon's contributions.

    2. Jon_Harrod | | #5

      Thanks, Charlie. I do think mixing ducted and ductless can be a great problem-solver in some situations, including the one you describe. A lot of our older farmhouses have hardly any ductwork to the second floor. This generally doesn't cause much of a problem in heating mode but doesn't work at all well for cooling. A ductless head in the primary bedroom can often make a huge difference.

      Whenever possible, I like to design this type of system with multiple single-zone outdoor units, rather than a single multizone system. This gives some redundancy and also reduces short-cycling.

    3. doug_horgan | | #8

      Good comments, just wanted to chime in that a lot of houses in my area (DC) have a second ducted system added to the upstairs, mainly to handle air conditioning which the long duct runs from the basement often fail to deliver. So it's certainly an approach that addresses practical issues with legacy systems.

  2. luke_p | | #3

    I had my bi-energy electric resistance 15Kw + oil fired 77000 btuh furnace replaced with a ducted 2 ton ultra low temp carrier heat pump with 10Kw aux heat, fully variable speed. 1000 sqft ground floor very old semi-detached duplex in Montreal Quebec, but with some insulation upgrades (ranges between R20 and R10 depending on which room, no insulation on the basement walls yet) and an apartment above us with it's own heating system. Just had a night with a low of -25C and it performed just fine. No issues using the existing ducts.

  3. ajc_electric | | #4

    how about the having and existing gas furnace as a back up to a heat pump unit ?
    Primary being the heat pump condenser and back up being the gas/oil fired unit?

    1. Jon_Harrod | | #6

      This is absolutely a viable approach. Nate Adams and others have argued that these "dual fuel" systems will be a great path to electrification for many households. Sized and set up correctly, dual fuel systems can reduce fossil fuel use by 70% or more. Other advantages include extra capacity, redundancy, and possibly a lower bar for customer acceptance.

      On the flip side, dual fuel is a path to reduce, not eliminate, fossil fuels. With these, you still need to keep your gas meter/propane tank/oil tank and all the other fossil fuel infrastructure. My preference would be to eliminate fossil fuels entirely. Even in cold climates in the Northern US and Canada, cold climate heat pumps are available that will work for most houses.

      1. luke_p | | #10

        Here in Quebec we call it bi-energy. It is very common and even has special electrical rates that are quite advantageous. A typical set up is either a conventional heat-pump of electric resistance heating system that supplies heat until -12C. Below that temperature, the electric meter has a temperature sensor that sends a signal to the heating system, which tells it to switch over to the 2ndary heat source. Typically oil or gas. At the lower temperatures the electric rate goes up by a factor of 5x.
        I chose to completely get rid of fuel all together in my house since our electric supply is mainly hydro powered, plus I wanted to remove the fuel infrastructure of chimney, oil tank and the associated maintenance.

      2. tekjunkie28 | | #13

        The problem is affordability. Thats out of the price range of most. Apartments are not going to install a heatpump.

        My house is all electric and the cost in insane for all electric too. Ive spent over $18k in 7 years on electric at 11.4 cents per KW. Another problem I see is that people in my area are adopting NG and getting rid of there HP. Its not cold enough here for that IMO.

        I just flat out costs the consumer too much money.

        1. arossti | | #14

          This is why Climate Stability requires Finance Solutions. $18k buys a lot of kit.

  4. adam5532 | | #7

    Thanks for a great series of articles. I plan to convert to a ducted heat pump system powered by solar panels in the future. I did a Manual J calculation in Cool Calc and the CFM per ton comes out to about 600. But it looks like air handlers normally can't be set over 450 CFM / ton, or 1350 CFM for the 3 tons I need for heating. The Manual J calc -- 89 deg: 21,561 BTU/hr and 1045 CFM; 95 deg 25,153 BTU/hr and 1227 CFM; 116 deg 33,799 BTU/hr and 1664 CFM; 122 deg 34,785 BTU/hr and 1764 CFM. We are in Portland OR. The design temp is 89 deg, but we like fresh air and saving energy, so it is unlikely that we will even use the air conditioning until it is 95 or so. It got up to 116 degrees last summer -- this is where we need it to perform. A 3 ton unit has the BTU output, but the CFM requirement is not matched by any actual equipment. The Man J says that the cooling load is almost entirely sensible (SHR 0.961). I know that I need to reduce static pressure in the system as well. Is my Manual J calculation wrong, or does this mean that a heat pump won't actually be able to cool the space down to 75 degrees when the outside temp is over 95 degrees in our climate?

    1. doug_horgan | | #9

      There are complicated answers to your question (I'll leave to actual experts) but one simple one is that it's not standard practice to size systems for extremely anomalous weather. "Design temperatures" are normally set to the 99% point, i.e. temperatures in a typical year will be higher than that for about 1% of the time, nearly four days worth. But these calculations are not especially precise, there is an element of oversizing built in to Man J, and in typical scenarios the house stays at or very near the set point.
      Sizing to a one-time (we hope!) weather event like that heat wave would likely result in oversized equipment, which has a number of downsides.
      I'm also a little curious about the SHR you're coming up with. Was humidity low during that crazy heat wave? Is your house really tight for air leakage? I know dry-climate air conditioning is a whole other beast with 450-600cfm/ton airflows being effective, kind of like what your Man J is showing--and people I know sometimes install mismatched equipment to achieve these flow rates (requires caution and expert knowledge). I would have thought Portland would be pretty damp even in a heat wave?

      1. charlie_sullivan | | #11

        The Pacific Northwest has fairly constant dew point all year, which means high relative humidity in the winter and low relative humidity in the summer. It's a very different summer scenario than in the South and East of the US.

        1. adam5532 | | #15

          Thanks Douglas and Charlie for the feedback. Even for the less extreme temps, it still wants very high CFM per ton. What then will be the result of running a slower fan speed than calculated? The air exiting the registers will be cooler, and more humidity will be extracted than desired, making the air too dry?

          1. charlie_sullivan | | #16

            I don't know enough about CoolCalc to know why it's giving you that unusual recommendation. But I think you will be fine using the BTU/h numbers it gives you and ignoring its CFM recommendations, and instead going with the equipment's CFM capabilty. If you stay within the equipment's capability, you won't get any equipment performance problems.

  5. arossti | | #12

    We've initiated this process using a hybrid gas/ASHP system retrofit into our existing ductwork. Article here: - when our conventional AC dies - we will replace it with the ASHP from the Coleman Echelon series - so our backup/supplemental heat will remain gas, but the ASHP will do most of the annual heating. We also plan to undertake new insulation, airtightness and window improvements in the interim - further reducing future ASHP load. I see almost no information out there on hybrid approaches. It would be great if you could mention them since the controls are important - so the call preference can go to the ASHP rather than the gas burner.

    1. Jon_Harrod | | #17

      I'll definitely plan a future article on hybrid/dual fuel systems. There are some interesting controls issues that come up around optimizing the balance point/switchover temperature. A good starting point is Allison Bailes' article here:

    2. aaron55 | | #20

      I think this hybrid approach may make sense for me. I have a 95% eff. gas furnace now that is 12 yrs old. I don't have AC but want to get it. So, if my current furnace is compatible with a heat pump, i could add that instead of getting a traditional AC unit. Might help to lop off some more of my emissions when I use the heat pump for more of my shoulder season heating needs. If my furnace isn't compatible with a heat pump, I guess I'll have to decide whether to just get an AC for now or replace it all. I also need to do a lot of work on air sealing and insulating, so hopefully i can get a lot of that done before I make a decision this year.

  6. Deleted | | #18


  7. aaron55 | | #19

    It seems like converting old, under- or uninsulated houses from fossil fuels to heat pump air handlers is going to cost too much to be adopted. I think I'd have to redo my shoddy ducts and add in second floor returns (or put a system in my attic and in my basement) to do this. To redo my ducts alone I got a quote for $16k, which seems about what I'd expect given the time involved and the pricing in my area. Then, I'd have to add in the air handler/install, which will probably push the total cost into the $25-30k range, right? Most people aren't going to do this are they?

Log in or become a member to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |