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
Building Science

Planning a Furnace-to-Ducted Heat Pump Retrofit, Part 1: Needs Assessment

The advantages of using an air handler when replacing a gas furnace with heat-pump technology

A ducted heat pump with air handler installed in an upflow configuration (NYSERDA).

Installation of cold-climate heat pumps is increasing rapidly as states pursue decarbonization goals and homeowners recognize the environmental, health, and economic benefits of moving away from fossil fuels. Cold-climate heat pumps offer a huge degree of design flexibility; options for indoor units include high- and low-wall-mounted ductless heads, ceiling cassettes, and low-profile “compact-ducted” units. One design, in particular, has tremendous potential to scale: It is what many manufacturers refer to as the “multi-position air handler” or “vertical air handler.” These units (referred to as air handlers for the rest of this article) have a form factor—or hardware design—similar to a conventional natural gas or propane furnace. They can be installed in upflow, downflow, and horizontal configurations.

According to EIA data from 2015, about 53.8 million housing units in the U.S. have forced-air systems burning natural gas, propane, or fuel oil as their primary heating source. This represents about 79% of the housing units that use fossil fuel and about 46% of all housing units. Converting these furnaces to heat-pump air handlers presents a viable path to electrifying a large segment of our housing stock.

natural-gas furnace and a heat-pump air handler
Before and after: A natural-gas furnace (left) and a heat-pump air handler (right). Photo courtesy of Sarah Cox, Halco Energy.

Advantages of air handlers over ductless minisplits

For homes with existing ductwork, air handlers are often the best choice. Compared to a strategy of putting ductless heads in multiple rooms, air handlers offer several advantages, including:

The ability to deliver heated and cooled air to every room. With ductless systems, it is usually cost-prohibitive to install a head in every room in the house. It is also a poor design strategy since even the smallest ductless heads can be oversized for small rooms. A typical ductless design leaves some rooms without their own source of heated and…

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. norm_farwell | | #1

    Thanks for this Jon, I look forward to the follow up articles.

    Re Maintenance:
    It sounds like you recommend annual cleaning of a ductless indoor head. I know this recommendation is common in the industry. But from what I’ve see, in most situations (kitchen areas being the exception) the blower wheel and indoor coil can go 3-5 years before a service call might be necessary. This is with Mitsubishi equipment which I understand has control features to minimize dust loading of the blower wheel. I don’t know whether that feature is standard on other makes or not.

    On outdoor units, I don’t advise clients that an annual cleaning is necessary either, on the basis of this GBA article from a couple of years ago:

    I’d be grateful for your thoughts. How do you advise clients about this? Am I wrong in taking a minimalist approach?

    1. Jon_Harrod | | #3

      Hi Norman,

      I agree that in many cases, there isn't a lot of buildup on the coils/blower after a year or two. It seems to depend a lot on dust loads in the home and how much folks run the AC. I do think the annual service is a good generic recommendation. In high-humidity areas like the Gulf Coast, some companies push twice-a-year cleanings. Outdoor coils generally seem to stay pretty clean as well, though I have seen some badly clogged with cottonwood fluff.

      In addition to the cleaning, the annual visit is a chance to check the drain pans and lines for clogging, the surge protector, and the delta T (heating) and delta H (cooling) to make sure the equipment is operating as it should. Although there's significant a cost associated with annual service, I think it's worth it to keep the system operating at peak levels, reduce the chance of emergencies, and ensure that warranty will be honored.

      1. norm_farwell | | #5

        Thanks Jon, I appreciate the advice.

        1. jameshowison | | #9

          FWIW we had terrible trouble keeping dust off the blower wheels on the Mitsubishi one-way ceiling cassettes. They needed to be cleaned frequently, like every 6 months or the wheels would get unbalanced and make all sorts of noise. In discussions it did seem that other one-way cassettes had had similar problems, lack of clarity on sealing from above (dusty attics) and difficulty fitting the cover on out of level ceilings (level, not flat), leading to by-passing what filter there is.

          Eventually we had them all replaced with the multi-position air handler and ducts with a 4" filter box, not a single issue since then. I can see the blower wheel when changing the filter and it's clean.

          Just an anecdote that suggests the experience of high wall units and dust might not carry over to the ceiling cassettes.

  2. irin | | #2

    Speaking from experience : we took out old gas furnace and even older air conditioner and all ducts and installed two minisplits in our raised ranch house (built in 1954). Our old duct system was contaminated with mice droppings and other nasty things - we tore them out and never looked back. There were no way to clean that mess. Minisplit heads actually allow much better control of air movement than old “hole in the ceiling or floor” manual register. It is true that there was a learning period when we experimented with the fan speeds and temperature settings but it is not that difficult :)

  3. d_w | | #4

    I would find it to be helpful to review the SEER difference between ducted air handler and ductless units as part of this comparison. A big part of converting from gas for me is specifically climate-related, and pursuing the most efficient solution with mini splits (with some cost in comfort) is important to that end.

    1. Jon_Harrod | | #6

      Hi Daniel, I will talk about SEER and HSPF when I get into equipment selection in a later article. My take is this: the big cuts in emissions come from getting rid of on-site combustion of fossil fuels, choosing the lowest carbon electricity available, and addressing the cost-effective envelope issues. Once we do those things, the difference in emissions between an 18 SEER ducted and a 21 SEER ductless is going to be quite small.

      It's also important to consider that SEER and HSPF ratings are based on standard testing conditions, and that real-world performance is going to vary based on local climate and specifics of design, installation, and operation. For example, this study by Steven Winter Associates ( found nearly threefold differences in real-world efficiency for equipment with similar ratings. My approach is to pick cold-climate equipment from the NEEP-approved list (, then focus on best practices to get as close as possible to rated performance.

      1. d_w | | #7

        Looking forward to the article, thanks for the reply

  4. riderpetch | | #8

    Thanks for the article. I look forward to your next one. I am a proponent of heat pumps for societal decarbonization and efficiency. I just swapped my NG furnace (ducted house system) for an air handler and ccASHP system. In your forthcoming article, here are a few things worth considering in determining the effectiveness of these 'box swaps' from NG furnace to an air handler should include: (1) Be sure to determine how many speeds your 'variable speed air handler' has. Mine appears to have only 3 speeds (some have 5 or more) and at the lowest speed, there is more nosie in the existing metal ducts then I have preferred. What noise mitigation exists? (2) Is the proprietary heat pump's brain compatible with AI thermostats (and without extra costs/equipment)? My new interface does not allow scheduling, so I need to upgrade it. At first glance, it looks like it will be up to $800 (CAD) to upgrade the themostat to a better proprietary model, compared to just $350 for an AI unit. My proprietary brand is not yet compatible with some AI thermostats like NEST, but is apparently compatible with ECOBEE. However, if I want to get an AI thermostat, I will need to buy a $600 proprietary commincation adaptor tool. (3) My unit is pretty quiet in the cooling phase, but it is located outside my bedroom and the noise in the winter is not unbearable, but is more then I like. What effective noise mitigation exists? Mass Loaded Vinyl and/or 5/8 drywall on resilience strips sound like options. Are there people with successful and simple noise mitigation DIY solutions? Also, we should understand STC numbers (see wiki) coming off these outdoor heat pumps so we know if our standard wall designs will hide the noise. Recommended distance from bedrooms? (4) The final one is how easy is it for a homeowner to get technical support? I called the tech support (with regards to the technical features of my thermostat, and they will only talk to my HVAC consultant (not homeowners), so homeowners have limited support to help operate these units. I feel having the capacity to incorporate an AI-based thermostat interface is critical. Again, I am a proponent of my ccASHP system and it throws great heat in cold temperatures and good cool air in the summer, but these are questions to get ahead on so we can be successful. Thanks!

    1. Jon_Harrod | | #10

      Thank you. These are great ideas. The issue of proprietary controllers has been on my mind a lot lately. People really want the scheduling, interface, and AI features of thermostats like Nest and Ecobee, but connecting to these third-party thermostats adds expense and often comes with a loss of functionality.

      Noise mitigation probably deserves an article of its own. I'm not aware of any formal guidance about distance from bedrooms, but following the manufacturer's specs minimum line set length, avoiding wall brackets in favor of ground-mounting (at least on wood walls), and avoiding placement outside bedrooms are all things I recommend.

      1. riderpetch | | #15

        Glad to hear these points prove useful. There are likely a few articles here! I agree that AI thermostats seem like the way to go with how they optimize the home heating/cooling scheduling. They seem to be cheaper than proprietary options too. As mentioned, heat pumps that allow fan speed adjustments and duct static pressure monitoring/control may help reduce duct and register noise.
        One other thought on the duct noise is that new protocols for air handler location and duct installations may be required that favour the homeowner comfort vs ease of access for the installers. For instance, in my 1250 sf rancher house (on a crawl), the air handler is in the crawl on the right side of the house (perhaps because it happened to be easier to install it here). The short duct runs on the right side of the house have much louder air noise at the registers compared to the registers on the left side of the house which have much longer duct runs (from the air handler to the registers). So I guess shorter metal rund make more noise at the registers. So perhaps installation guidance should consider putting air handlers in a centralized to create more even duct runs.
        The City of Vancouver (Development and Building Services) which is going heavily in support of low carbon space and water heating options like heat pumps have a Neighbourly Installation Guide for Heat Pump and Noise located here: So perhaps they could be contacted for noise mitigation ideas if you do an article on helpful noise mitigation strategies.

      2. john_heckendorn | | #18

        "People really want the scheduling, interface, and AI features of thermostats like Nest and Ecobee, but connecting to these third-party thermostats adds expense and often comes with a loss of functionality."

        Jon, would you say that at this point in time, there's a fundamental mismatch between traditional and modern heat-pump HVAC models? Traditional systems designed around cheap and abundant energy consumption in suboptimally-insulated environments have generally operated through on-off cycling to adjust for gradual changes in ambient temperature, and their thermostats -- including Nests, Ecobees, etc. -- have been designed to operate in concert with this older model.

        Instead of cycling, the new HVAC model favors very gradual modulation in climate-controlled environments, and may run constantly sipping rather than periodically gulping energy. In most climate zones, they may almost imperceptibly transition from heating to cooling and back to heating over 24 hours. If it requires less energy and cost to maintain an optimal climate, rather than needing to schedule operation cycles that raise/lower temperatures over wide ranges to compensate for shortcomings with the HVAC or the building as a whole, it's easy to see how fitting square pegs (AI designed for traditional HVAC) into round holes (modern HVAC) might be challenging.

        IMO, in a residence that has been appropriately designed and constructed, the control interface for modern HVAC should almost never need to be used, or even consulted, since the intelligence for climate control has been applied to that environment as a whole (with a working passivhaus perhaps the ultimate achievement of this). As a resident, the only things I want to set are my optimal temperatures when awake and asleep, and have the building execute with minimal energy use, cost, and maintenance to achieve those.

        1. Jon_Harrod | | #20

          This is an interesting point; thermostats designed for conventional systems (even ones with two- or three- stage operation) depend on on/off logic rather than continuous modulation. They achieve efficiency through temperature setback, whereas inverter-based systems achieve efficiency through optimizing the compressor speeds and often lose efficiency if deep setbacks are used.

          1. john_heckendorn | | #25

            Hi, Jon, thanks for responding. I've been told by heat pump HVAC people that there are add-on interface products to connect Nests, etc., with HP systems like Fujitsu and Mitsubishi, but in doing so, you lose the modulation features, which are the raison d'etre of the best HP HVAC solutions. I wonder how many people have done HP retrofits to residences that are unsuitable for them, and to pile error upon error, have spent additional money so that their Nest or Ecobee thermostats will work with them.

  5. dennisgray | | #11

    Many regions of the country still rely heavily on coal burning power plants. Going electric before the utilities convert to more renewable energy source is kind of defeating the green purpose. Maybe with solar panels you always be good.

    1. Jon_Harrod | | #21

      I think that this was once the case; but between the improved efficiency of heat pumps and the trend toward decarbonizing the electricity sector, I'm convinced that heat pumps make sense in all grid regions. As David Roberts pointed out in his Vox essay "Electrify Everything," a heat pump installed now will see its annual emissions continue to drop as the grid improves. A 2020 study by the Sierra Clubs seems to bear this out:

  6. jimhammel | | #12

    For 15 years my HVAC providers have been telling me about the unreliability of heat pumps due to the complexity of their controls. Any unit can have problems, but has there been an inflection point in technology that makes new heat pumps more reliable?

    1. Jon_Harrod | | #22

      Hi Jim,

      My impression is that quality control, especially for top-tier manufacturers, is very high. Most of the failures I've seen have been due to problems with design, installation, or maintenance. A big part of my mission in writing about heat pumps is to try to prevent these problems, which could easily cause heat pumps to have an ongoing reputation for unreliability. There are also some logistical gaps that need to be filled--especially related to parts supply chains, but this is always the case with newer technology.

  7. ZenHomes | | #13

    Probably so, but the bigger reality is that the old-school manufacturers and their dependent installers are no longer able to hold back the tide of better technology.

  8. bje11 | | #14

    Great article, Jon. Thank you!

  9. Elizabeth_H | | #16

    Great article. Like others I am keen to find a carbon friendly and efficient heating and cooling system to replace the 40 year old electric furnace. ASHP makes great sense. My application is a bit unusual with no cooling in the house and a combination of main floor forced air and upper floor electric radiant heat. There is no other mechanical ventilation. As I improve building envelope, I am concerned about ventilation and humidity management. The ductless systems have some capacity for humidity management but local codes are now requiring ERV or HRV. It seems that this challenge of getting to the right combination of heating, cooling, air handling and moisture management is outside the expertise of the contractors. Any suggestions are appreciated!

    1. Jon_Harrod | | #23

      Hi Elizabeth,

      One idea to consider in this application is a mixed system: air handler for the main floor and ductless for the upper floor. Another option would be a compact-ducted air handler in the attic.

      Ventilation air can be brought into the main-floor duct system via an ERV, HRV, or ventilating dehumidifier. If you'd like ventilation for the upstairs rooms, through-the-wall room ERVs may be a possibility.

      The bigger issue is finding a team that can design and install this system for you. Your contractors may be able to get some technical support through their manufacturer's rep; alternately, it might be worth reaching out directly to the manufacturer (Mitsubishi, LG, Samsung, Carrier etc.) to see if them can recommend any local engineers or contractors who can help.

  10. kearsarge | | #17

    Thanks Jon. Another use case to consider. Currently, we have HWB via oil burner. I'm looking at replacing my central AC with a heatpump next go around for usage in at least the shoulder seasons of greater Boston! Seems like an easy choice for downstairs where ductwork is in basement but not sure on upstairs unit where ductwork is only partially buried in loose fill insulation.

    1. Jon_Harrod | | #24

      Hi Patrick,

      Ductwork in an unconditioned attic is always worth looking at carefully, both to avoid energy losses and problems (winter ice dams, summer condensation). As long as the ductwork is well-sealed and fully insulated (with at least a couple inches of a vapor-impermeable material like closed cell spray foam), it can work. Getting the ductwork up to standards may require partial removal and reinstallation of the loose fill insulation.

  11. user-6750330 | | #19

    About three years ago we moved to a two-story house in Philly where floor space was at a premium. We eliminated a utility room on the ground floor by replacing the water heater with an on-demand unit and replacing a bulky gas furnace with an air-source heat pump. We conditioned a space in the attic and moved the heat pump up there. The house had masonry wall with no additional insulation, so we added stud walls with fiberglass and had room to reroute the ductwork in the new partitions. So far, the system has worked very well in winter and summer. Solar panels we added to the roof keep electric costs down. Very glad we made the change.

  12. dtcaswell | | #26

    Hi Jon, Were doing a complete renovation of an old (1750) Cape style home to bring up to code and insulation standards. Completely gutted now and was initially planning on using propane fueled furnace/AC ducted system. After reading your article might rethink fossil fuel for heat pump system. Only thing installed so far is a 500 gallon in ground propane tank, but will be using a gas stove and possibly a gas freestanding stove for backup heat so will still need the tank for them. We're in coastal NH just north and in Zone 5.

  13. jrhemstad | | #27

    @Jon Harrod

    Could you comment on "dual-fuel" heat-pump setups with respect to backup heat?

    I live in Northern Minnesota where we're guaranteed to have more than a handful of -10F days or lower.

    I currently have a traditional natural gas furnace + air conditioner in a 1970s built home that I've done my best to retroactively air seal (e.g., spray foamed some but not all of the rim joist) and would love to update to a heat pump, but I don't feel comfortable with having only electric resistance heat as the backup heat source. Am I being unreasonably concerned?

    If not, I had done pretty considerable research into high-efficiency "communicating" dual-fuel systems with a natural gas furnace and air-source heat-pump that work in tandem to decide which unit provides the heat. Much of what I've read about these systems does not make me optimistic as they sound quite fussy to setup and maintain the communication brains of the system.

    Curious on what your take is for what someone should do in my situation.

  14. andy_c | | #28

    It would be great to hear people’s experience of these heat pumps in cold climates. I’m looking to replace an old propane furnace in a large property with ducting… I want to put a heat pump in but all the local suppliers say these things don’t work below -10C (14F). Whereas the specs say they’re good down to -4F or sometimes lower.

    Their “real world” view is somewhat off-putting as we do have long winters here… but I’m normally inclined to believe the spec over opinions. Any recommended low temperature whole house units?

    I also considered running a refrigerant coil around our wood stove chimney and past the heat pump so we could warm it’s area up in closer temps… any thoughts on that concept?

    1. Jon_Harrod | | #29

      Hi Andy,

      Just to share a bit about my own experience. I live in Upstate NY; our local design temp is 0F/-18C, and we get a few winter mornings colder than that. My Mitsubishi heat pump has performed great for the last two winters with no backup electric resistance heat.

      I would stay away from any attempt to draw heat from the chimney using refrigerant. Heat pumps and wood stoves operate at very different temperatures, and this could really wreak havoc on the heat pump system, not to mention voiding any warranty. A better strategy would be to use the wood stove as an independent heat source to complement the heat pump system.

      1. andy_c | | #30

        Hey thanks for your reply. What Mitsubishi model do you have?

        Just to be clear the wood stove is in operation as a supplementary heat source inside the property - it’s the main heat source at present… my concept was to use the warmth in the outdoor chimney when the stove is running, as the chimney is adjacent to where I’d put the outdoor heat pump. Conceptually a heat coil that takes its energy from the chimney could elevate the ambient air temperature around the heat pump by a few degrees - for free. I thought it was a pretty clever concept that would avoid issues on super cold days and allow the heat pump to operate.

        1. Jon_Harrod | | #31

          I have an SUZ-KA18NAHZ outdoor unit with an SVZ-KP18NA air handler. So far it's been a solid and trouble-free system.

Log in or become a member to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |