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

Forced air is forced air … isn’t it?

fitchplate | Posted in General Questions on

All this talk about ductless minisplits is making me dizzy. If a forced air furnace gives a bad atmosphere in a building, doesn’t a minisplit per room just give many bad atmospheres?

Dana reminded me on another post that a forced air furnace increases infiltration/exfiltration. Well, why doesn’t a minisplit heater or cooler do the same?

When air blows past my skin, it makes me cool due to evaporation. Heated air still causes evaporation. Ductless minisplits blowing air into the room from above means air blowing into my face from above. Just not pleasant, even if the diffusion technology is much better (than conventional forced air), the air in the room is moving and there is an atmosphere of abnormal mechanical convection and cross current flow.

If ductless minisplits – to be able to handle cold temperatures – cost several thousand $’s per ton, how can they possibly be a better buy than a good old (yuck, I hate them) ducted forced air that cost much less for the same BTU’s?

I am not getting it. With such high capital outlay costs, what is the attraction? Why not go with zoned electrical radiant space heating solutions in each room? Heating things still remains a better strategy for building performance and occupant comfort than heating air.

Are we not conspicuously overrating a technology (ductless minisplits) that can be beat $ for $ and comfort feature for comfort feature by other solutions?

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. GBA Editor
    Martin Holladay | | #1

    You are free to choose any heating and cooling system you prefer. Ordinary forced-air systems have several advantages. They are readily available, relatively inexpensive, and easily serviced. For more information on forced air systems, see All About Furnaces and Duct Systems.

    Ordinary forced air systems also have several disadvantages. Furnace manufacturers don't yet make models that are small enough for the needs of a superinsulated house, so builders of high-performance homes end up selecting models that are oversized.

    The main reason that conventional forced-air systems increase infiltration and exfiltration rates (as Dana Dorsett mentioned) is that their duct systems leak, and are often located outside of the home's thermal envelope. Even when ducts are located in "interior" chases, they can increase infiltration or exfiltration rates if the ducts are leaky and if the chase includes cracks that communicate with the outdoors.

    These problems associated with leaky ductwork do not occur with ductless minisplits.

    I have no idea what you mean by "If a forced air furnace gives a bad atmosphere in a building, doesn't a minisplit per room just give many bad atmospheres?" It sounds like you are implying that forced air furnaces and ducless minisplit heat pumps are responsible for IAQ problems. If that is what you are implying, you are wrong. If the air in your house is stuffy -- or has what you call a "bad atmosphere" -- then your heating system is unlikely to be the cause of or the solution to your problems. The solution is to remove indoor pollutants (cigarettes, sources of moisture, and sources of formaldehyde) and to install a mechanical ventilation system.

    But, it's a free country. If you like electric-resistance heating systems, go ahead and install one. Of course (as I'm sure you know), operating an electric-resistance heating system will cost two or three times as much as operating a ductless minisplit system.

  2. Expert Member
    Dana Dorsett | | #2

    Mini splits aren't moving air from one part of the house to the other, so no room-to-room pressure differences are created to drive air infiltration. The pressure difference created by the the blower head is extremely local to the head, and completely inside the pressure boundary of that room.

    Air handlers on ducted systems are never completely & perfectly balanced, and will always create room-to-room pressure differences, pressure differences that are affected by whether the room doors are open or closed, etc. These pressure differences can become quite significant in systems where the a room has a supply register but not a return register, or if the duct design isn't well balanced. In well designed systems those pressure differences can be made small, but never completely eliminated.

    But even the best designed duct system has pressure differential issues when the ducts leak.

    In a US climate zone 6 climate a better-class mini-split will average a COP of about 2 (300% efficiency). In a US zone 3 climate it'll average a COP of about 4 (400% efficiency.)

    In any climate a resistive element (radiant or baseboard) is limited to at best a COP of 1.

    If electricity had low/no cost (environmental or cash) it wouldn't really matter. In a home that beats PassivHaus standards on energy use it may not matter (much). But in most existing homes (even new ones) in US in zones 4 or higher, replacing resistance heating of a large zone heated with a mini-split has a sufficiently high IRR due to the higher efficiency to pay off the capital investment in about 3 years, sometimes less. In a high performance house it might take more like 5-6 years, but that's still well within the lifecycle of the mini-split, and a far better ROI than a photovoltaic system of sufficient size to cover the operating cost difference.

    The air volumes and velocity of mini-splits at low speed (where it spends most of it's time) are quite low, with negligible wind-chill, and exit air temperatures more than 10F higher than human body temp. While you don't want to place the ductless head where it's blowing right in your face while sitting at the dining table, there isn't very much in the way of wind chill, just as you don't really feel the wind chill of the blower when your refrigerator is running. (They're quieter than your refrigerator too.)

  3. jackofalltrades777 | | #3

    I currently have a ducted HVAC system and even though it was supposedly professionally designed and installed, the system has the problems Dana mentioned above. I have yet to be in a home with a ducted HVAC system that doesn't have problems with pressure balancing and dealing with room temperature issues. When my system activates it gets me so cold that I have to pull a sheet over me to keep me from shivering. It then shuts down and minutes later it gets hot again and I have to remove the sheet to keep from sweating. Then the system kicks back on and this game is played all summer long. Welcome to modern ducted HVAC.

    That is why my future home will have ductless mini splits. I've had it with these ducted systems. To each his own but the energy numbers don't lie. Ducted systems are nowhere near as efficient as ductless mini splits are.

  4. wjrobinson | | #4

    Dana, I sit at a lunch stop where 30' away the mini split AC blows to me and cools me with a breeze which I desire on a 90 degree day. There is a breeze most definitely when on full speed.

    That said, I like mini splits.

  5. fitchplate | | #5

    So why not just put airtight and heat loss prevention quality into the distribution, and comfort and efficiency into the design by using digital controls for a room by room or zoned forced air system.

    Are you guys saying that all things being equal (zoned and controllable, the cost of fuel, and quality of the distribution system), including ECM motors, that the cost per BTU of a heat pump is cheaper than the cost of burning hydrocarbons on site?

    And that retiring the capital cost debt over the useful life time of the system nets out in the favor of the minisplit.

  6. GBA Editor
    Martin Holladay | | #6

    Q. "So why not just put airtight and heat loss prevention quality into the distribution?"

    A. GBA has always advocated that approach. See my article on the topic: All About Furnaces and Duct Systems.

    Q. "Why not put comfort and efficiency into the design by using digital controls for a room by room or zoned forced air system?"

    A. Because it's hard to zone a forced-air system. For more on this topic, see The Achilles’ Heel of Zoned Duct Systems.

    Q. "Are you guys saying that all things being equal (zoned and controllable, the cost of fuel, and quality of the distribution system), including ECM motors, that the cost per BTU of a heat pump is cheaper than the cost of burning hydrocarbons on site?"

    A. If you are talking about fuel oil or propane (common fuels in New England), the answer is a definite yes. If you are talking about natural gas, probably not.

    Q. "Are you guys saying that retiring the capital cost debt over the useful life time of the system nets out in the favor of the minisplit?"

    A. Absolutely. Carter Scott provides heating and cooling for his two-story homes in Massachusetts with two ductless minisplits, which he installs for less than $6,000. (See Just Two Minisplits Heat and Cool the Whole House.) There is no way to beat that equipment cost (unless, perhaps, you use noisy PTACs).

  7. Expert Member
    Dana Dorsett | | #7

    In New England climate & electricity pricing (higher than the US national average) mini-splits are a cheaper way to heat on a full-lifecycle levelized cost basis than anything but gas-fired hot air. See Figure 2 of this document:

    You'll note in the graphic that the marginal operating cost of heating with heat pumps overlaps the marginal cost of heating with gas (it can be cheaper on the utility bills than heating with gas in many locations) but heat pumps come with a higher homeowner capital cost. Even at the high-end of the electricity price assumptions, the ENTIRE cost of heating with heat pumps (including the capital costs) is less than the MARGINAL cost of fuel oil at the low-oil-price assumption. Propane is more expensive per BTU than #2 oil in most markets.

    Built into the LCOE of gas in that analysis is some amount for extending the gas-grid, costs that are usually rate-based, with all gas consumers paying for those expansions, which is based on gas-grid costs for projects under way in Connecticut. On a gas-grid that isn't undergoing expansion the lifecycle costs of heating with gas is going to be lower than heating with mini-splits.

    Mind you this analysis for predominantly climate zones 5 & 6 region (with bits of zones 4 & 7). In zones 3 & 4 the fossil fuel prices aren't necessarily different, but the efficiency of ductless mini-splits goes up significantly. Whether mini-splits beat gas on LCOE is highly dependent on the local electricity & natural gas market pricing over the next 20 years.

    While natural gas is currently at or near historical low pricing (and has nowhere to go but up), the looming tsunami of very low cost grid tied photovoltatics (PV) is going to put a cap on how far electricity prices can rise. Sanford Bernstein (the investment bankers) are predicting electricity price deflation beginning in about 10-12 years. Several others in the financial sector (notably Citi-Group, Bloomberg, and Barclays) all have similar analyses pointing to downward pressure on electricity pricing during the 2020s due to rapid growth of low-cost PV. The wild card is how quickly plug-in electric cars gain market share, and whether it will be fast enough to increase total electricity sales in the face of increasing efficiency measures. Electricity sales in the US have been flat or falling for about 7 years now, and that's BEFORE market penetration PV on the customer's side of the meter has reached point of being even 1% of all power generated on US grids. The exponential growth of privately owned PV is just starting to be felt by competing power generators in a few local US markets, but is clearly going to be affecting electricity pricing in a big way long before your mini-split is toast.

    It's up to local utility regulators to restructure it in a way that benefits all stake holders though, but before 2030 retail electricity prices will have to stop rising (or drop), or it will become cost-effective for many ratepayers to defect from the grid entirely.

    So, even if natural gas might be cheaper than mini-splits today, it's not entirely clear that it will be over the full lifecycle of the equipment- there are many active variables in the equation. It seems unlikely that propane or heating oil would ever catch up though- worldwide demand for oil as transportation fuel isn't going to stop growing, and new sources of oil are expensive to exploit. The ~$100/bbl price point of recent years is volatile, but it would have to drop below $50 to be competitive with ductless heat pumps, and at that price the oil-sand & shale-oil producsrs would be losing $15-30/bbl.

  8. fitchplate | | #8

    You know, most products eventually go down market as a result of their natural life cycle but energy seems the exception and has always been very well protected (permitted but clearly illegal anti-trust behavior). Seven years of lowering electricity demand explains why nuclear plants are not being built in the US but even with the James Bay development, Canada is still investing in new reactors.

    I am not so sure that the quantity of photovoltaics coming on line are sufficient to hurt the retail electricity markets well protected monopoly. I realize it takes only a few percent of loss of gross sales to make a relatively large dent in utility’s profits, though. But I see them getting regulatory approval, state by state, for increasing inter-tie fees and reducing the $/kw returns to the home owner and farmer PV producers.

    I was raised to shy away from electricity for heat, as most of us were. And NY has so far successfully kept out fracking. But now I wonder, if you’re analysis pans out, Dana, if I should exchange out my Triangle Tube Prestige Excellence for an electric hot water heater to deliver DHW and hydronic heating?

    Since I am off the natural gas corridor, I have been hit hard by last year’s propane price rise in western NY. I love wood but it’s a lot of work and dirty and I am not a spring chick anymore. The $/BTU is attractive but the distribution of wood heat is not so efficient.

    So if I want radiant heat, is it better to change out to electrical baseboards or electric hot water?

  9. GBA Editor
    Martin Holladay | | #9

    Q. "If I want radiant heat, is it better to change out to electrical baseboards or electric hot water?"

    A. If you want electric resistance heat -- and I'm not sure why you would prefer electric resistance heat to a ductless minisplit, unless you are determined to keep your equipment costs to a bare minimum -- then electric resistance baseboards make much more sense than a hydronic system connected to an electric boiler or an electric resistance water heater.

  10. Expert Member
    MALCOLM TAYLOR | | #10

    Or perhaps radiant cove heaters?
    With not much of an upgrade over what is often called "code minimum" insulation levels here, it is fairly easy to build a house in the PNW with a monthly heating cost of under $50. With loads that small I wonder where the line is where a mini-split gains the advantage over resistance electric?

  11. Expert Member
    Dana Dorsett | | #11

    Flithc Plate: At recent and projected propane pricing it's worth heating at least one large zone with a mini-split in western NY. But at current NY electricity prices resistance electricity is still more expensive than condensing propane in most locations. So while an electric boiler might drop right in to your system, it isn't going to save ANYTHING on marginal heating costs in the near term, and will likely be more expensive.

    You may be aware that the regulatory environment for the utilities is undergoing a MAJOR restructuring right now in NY to be able to accommodate high penetration of distributed renewables, distributed storage, micro-grids, etc, but retail electricity pricing is likely to even rise in the near-term (due to rising natural gas contract prices with the power generators), and won't start falling for another decade.

    I'm not sure what you're paying for propane or electricity in Western NY but near me it's been running about $3/gallon. In a 95% efficiency boiler that breaks down this way:

    91,600 BTU/gallon x 0.95 = 87,000 BTU/gallon, or 1,000,000/87,000= 11.5 gallons / MMBTU At $3/gallon you're paying $34.50/MMBTU

    A kwh of electricity in an electric boiler deliver 3412 BTU, or 1,000,000/3,412= 293kwh / MMBTU. To be at parity with $3/gallon propane in your boiler electricity would have to be priced $34.50/293= 11.8 cents/kwh. If your electricity is cheaper than that, the electric boiler would be marginally cheaper. If it's more expensive, the Triangle Tube would be cheaper.

    The average residential retail price of electricity in NY state is currently north of 20 cents, but that's heavily skewed by NYC & Long Island prices:

    EIA data indicates you might have been paying $4/gallon for propane at the end of the winter, but you graph it over time you'll see the running 5-year average has been about $3:

    BTW: Another recently published investment bank analysis has just joine the chorus mourning the imminent death of the centralized power generation grid model:

    If you have radiant floors for radiation you can get mini-split efficiency out of the Daikin Altherma air-to-hydronic heat pump, provided you have sufficiently low peak water temp requirements. They're pretty efficient at cranking out 100-110F water even when it's 0F outside, but if you need 140F water, forget about it. It's a big chunk of change though, and unless you live in a high-performance house you may not have sufficient capacity.

    If you have radiant floors you could control the boiler with a floor thermostat, set it to say 72-73F during the winter, then use a mini-split to control the ROOM temp and still save quite a bit of money. A72-73F floor is still very cushy under foot, but during higher load periods the mini-split would be carrying the lion's share of the load.

    Malcolm: With $50/month electric heating bills using resistance heating you'd be saving about $400/year with a mini-split, provided that that your minimum fixed charges aren't $25 or something even more ridiculous (which is the case, in some areas.) If you saved even $300/year with a mini-split that cost $3000 more to install than an electric baseboard solution, that's still an IRR of 10%, which isn't a terrible. It's not hard math- assume you'll get an average COP of at least 3.0 with the mini-split in your area (3.5 is more likely west of the Cascades) and the cost about $3500-4000/ton, installed. Each ton of mini-split puts out as much heat as 3.5 kw of baseboard/cove or electric boiler.

    An electric boiler/hydronic solution would cost more than mini-splits due to the higher cost of the radiation, but it can be micro-zoned. An electric baseboard solution might cost half as much as a mini-split- much cheaper hardware, but considrably more electrician-time. The lifecycle costs would have to be estimated too- the electric boiler would only last about as long as the mini-split (say 20-25 years), but the radiation would be good for at least 2-3x that. Baseboards/cove heaters are probably good for 30-50 years.

  12. Expert Member
    MALCOLM TAYLOR | | #12

    Thanks Dana. My post was a bit misleading. I meant that the additional electricity used for heating is around $50 during the months we need to heat here in the PNW. So the total spent on heat for the whole year is around $400. The possible annual saving from going to a mini-split would presumably be about $250. I'm wondering where that leaves me?

  13. wjrobinson | | #13

    Lowest cost compromise, Wood stove with free carpentry waste and trees for temperatures below a mini splits COP 4 temperatures, and let the mini provide all the heat at temps creating a COP of 4.

    The benefit is only having to feed the wood stove once a day and only during cold snaps or when desired verses must have to.

  14. fitchplate | | #14

    I feel like the guy who just spent his wad on a new horse and buggy the day the Model T was brought to market. The capital cost of retooling makes mini splits for heating is a poor proposition for someone like me. I will probably kick the bucket before my current system is ready for refurbishing.

    I already have the propane fired TT boiler pushing 4 four zones to 14 Stelrad wall hung rads and for the summer I have two Mitsubishi MUY-GE18NA Mr. Slims. I found this summer that the open clerestory windows, 30" eaves and lower windows open on the north side create an amazing cooling breeze beyond my expectations.

    Lucky that I can heat 1100 sq ft with vaulted ceilings for about the same as Malcolm. I need 13K BTU's/hr at design temp, located on the south shore of Lake Ontario. We have a small efficient 28K BTU wood stove and we can buy a full cord of ash and maple split for about $225. Propane and electricity are expensive here as you point out Dana, but mean low temperatures do not often go much below 18 F for Jan and Feb. Haven't moved in yet but I am imagining less than $75/mo for prime heating season (3-4 months).

  15. Expert Member
    Dana Dorsett | | #15

    Malcolm: It's still easy math-calculate the differences in upfront costs for resistance heaters vs. a mini-split solution, and divide your annual heating bill by 3 (if you're east of the Cascades) or 4 if west of the Cascades to come up with the IRR of the investment.

    If you would have paid $400/year heating with baseboards you're really probably going to burn $100-130 with the mini-split. But even if it's only $250/year savings for that $3000 cost difference, that's still more than 8% marginal rate of return, for something that's going to be in service for 20 years or more.

    If you get out your accountant's pencil you 'll also note that savings are not taxed, but the utility bills are paid for in after-tax dollars. Don' t forget to include the avoided cost of window air conditioners in that up front cost calculation too.

    Though it's unlikely to pencil out favorably this year, you might also take a peek at how much grid-tied PV it would take to deliver that $250/year in electricity, see if that beats a mini-split. At the rate PV costs are falling 2025 the PV solution probably would beat a mini-split, if you're still allowed to net-meter at retail then. It's pretty clear that by the time PV that's cheap most locations will be compensating solar on a value-of-solar-tariff best case (similar to how it's done in Austin TX or the state of MN) , or onerous monthly fees in worst case, but dumb net-metering can't/won't last forever in the age of buck-a-watt PV, or even two-buck PV.

    Flitch Plate: I feel your pain- the upcharge for heat pump versions rather than cooling only mini-split would not have been huge in the grand scheme of things. I'm wondering why your cooling loads are so high in an 1100' house that it takes 3 tons of ductless to cool it though(?). In a tight 1100' house without a lot of west facing glass I would have expected the total cooling load to come in between 1-1.5 tons. (A single MSZ-FH15NA could probably swing both the heating & cooling loads in an open floor plan house that size.)

  16. fitchplate | | #16

    Dana, I think you’re right but touch optimistic. I relied on someone else’s conservative calc’s. Although my vaulted ceilings add about 5000 cubic ft over the same area with an 8ft ceiling, I would figure I am 1 ton over the safe margin for the high humidity location of south shore Lake Ontario climate zone 5A (moist) average summer humidity levels in mid-day + 80%.

  17. fitchplate | | #17

    I have (don’t tell Martin) 4 X 4” passive airlets which (as I think Martin might have pointed out once) can get “air locked” and not deliver a flow of any volume without a mechanical exhaust operating (i.e. stove hood externally vented; bath fan). Rather than use the mechanical exhaust-only hybrid approach, in order to save some juice I would like an effective passive draw and these venturi stack caps are promising … particularly when I hear your (Dana’s) numbers on flow.
    This European design looks interesting too: but more complex and expensive.
    Keith … I am back and forth (Buffalo-Toronto) monthly and could bring a couple of Twin Comfo’s across if you order them online for HD store pickup.

  18. fitchplate | | #18

    Opps, gettin' my posts crossed.

  19. Expert Member
    MALCOLM TAYLOR | | #19

    Thanks for the reply, I'm almost with you except for a couple of things. Isn't the difference in the initial capital costs between the mini-split and baseboards also in after tax dollars, negating some of that benefit? Similarly, shouldn't that difference be seen to be generating income for me, say at a rate of 5%, so that $3000 dollar capital difference also swings the equation in favour of the baseboards by another $150 a year?

  20. Expert Member
    Dana Dorsett | | #20

    Malcolm: If you can run a spreadsheet, you can run your own NPV on the various options using different assumptions on what you'd expect your money to earn in a similarly safe investment, electricity price inflation, etc. (Ten year US treasure bills or bank CDs are paying only about half that 5% rate.)

    You can make the financial model as complex & detailed as you like, but I'm not sure it's worth taking it to the n-th degree.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |