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

How Replacing a Furnace Can Make You Less Comfortable

The problem arises when upgrading to a high-efficiency unit

Replacing an old furnace with a higher efficiency furnace can lead to comfort problems because of oversizing.
Image Credit: Energy Vanguard

Let’s say your trusty old furnace is at the end of its life. You’ve got to buy a new one, so you call your HVAC company and they rush over to make sure you don’t freeze during the next cold snap. They go and take a look at your furnace and find its capacity. They come back and tell you that you have a furnace rated for 60,000 BTU per hour and then talk to you about some of the options.

You could replace the old one with one that has the same efficiency, 80 AFUE (Annual Fuel Utilization Efficiency, essentially the percent efficiency over the whole heating season). Then they show you the price of their high efficiency furnace, which is 95 AFUE. It’s only $400 more and they convince you that since you’re going to have the furnace for another 15 or 20 years, you’ll quickly get that money back in savings on your gas bill.

So you tell them to go ahead and install the 95% furnace. Your old one, remember, was rated at 60,000 BTU/hour of input capacity. They look in their product catalog and find that the high efficiency model comes in a 46,000 BTU/hr model or a 69,000 BTU/hr model. There isn’t a 60,000 BTU/hr model, and they don’t want you to freeze, so they install the 69,000 BTU/hr model.

Looking at the wrong capacity

Hmmm. See a problem here? Many HVAC sales people and techs think in terms of input capacity for furnaces. To them, the 69 K model is a little bigger than the old 60 K model they’re replacing, but not too much bigger.

Actually, it’s worse than that.

The relevant capacity of the furnace isn’t the rate of BTU input. It’s how quickly those BTUs go into your home — the output capacity. Let’s look at what happened to those numbers when you replaced the old furnace. You started with 60,000 BTU/hr at 80% efficiency, so your old furnace supplied 60,000 x 0.8 = 48,000 BTU/hr. The new furnace has an output capacity of 69,000 x 0.95 = 65,550 BTU/hr. To summarize the numbers:

Input:    60,000 (old furnace) ==> 69,000 BTU/hr (new furnace)  —  15% increase

Output:  48,000 (old furnace) ==> 65,550 BTU/hr (new furnace)  —  37% increase

So, your new furnace is 37% larger in terms of its ability to provide heat to your home. If you had started with an older furnace, perhaps only 60% or 68% efficient, the amount of oversizing would be even worse.

Why it matters

How does that affect you? Your furnace will be on for shorter periods of time.

If you live in a high-performance home with a well-insulated and air-sealed building enclosure, you may not notice much of a difference. Well, let me rephrase that. You won’t notice much of a difference most of the time… when the furnace is off. But when it comes on, you’ll feel a blast of Sahara Desert heat making you long for the cool on the other side of the pillow. (RIP, Stuart Scott.)

If you live in an older home with air leakage and insulation problems, you may notice that you’re less comfortable. The furnace comes on and blasts you for a few minutes, possibly making you uncomfortably warm in some parts of your home. Then it goes off, and you feel the cold walls and the drafts.

What can you do?

If you’re getting a new furnace, ask the HVAC companies you get bids from how they’re planning to size the new system. There are a lot of ways to do it, and many of them are wrong. Some acceptable ways would be to do a Manual J heating and cooling load calculation, use manufacturers’ sizing software, or monitor the amount of runtime of your current system, especially at design conditions.

Probably the least reliable way to do it would be just to look at what size you have now and install that size or larger. To make it worse, just look at input capacity and ignore the effect of installing a new system with higher efficiency.

Bigger isn’t always better. When it comes to heating and cooling your home, it’s often worse.

Allison Bailes of Decatur, Georgia, is a speaker, writer, energy consultant, RESNET-certified trainer, and the author of the Energy Vanguard Blog. Check out his in-depth course, Mastering Building Science at Heatspring Learning Institute, and follow him on Twitter at @EnergyVanguard.


  1. AlanB4 | | #1

    Dual stage?
    I'm no expert so correct me if i am wrong but are not most high efficiency furnaces dual stage?
    I had a 80% eff 75k in (60k out), i replaced it with a 40k (26k low, 39k high) and have had no problems, but interestingly before upgrades HOT2000 had my heat load at 63k, and now with the new furnace and walls insulated the load is 45k, this week i have had temps within a few degrees of the -20C (-2F) design temp so i am playing with fire (though i plan to reduce air infiltration and insulate the attic this year which should bring my design load to 34k for next winter)
    I have noticed that the furnace seems to kick into high fire after about 5-10 mins then the thermostat shuts it down within 30 seconds, i wonder if its coincidence that in the last two days the rate of house heat loss vs furnace heat generation is balanced in a way that causing shutdown within 30 secs of high fire every time or its some problem somewhere (are there settings that can be adjusted wrong?) I hope this does not reduce furnace life.

  2. AlanB4 | | #2

    Thermostat was installed with the furnace
    Its a White Rogers 1F80-361, and i did a lot of furnace research so maybe i am remembering wrong but i think the furnace does the high fire on its own, its a Payne PG95XAT

  3. JIM BAERG | | #3

    Comfort after
    Timely article. Just to broaden the discussion, here are some other issues to think about. As it turns out, I just replaced my 120k Btu 80% Type I Lennox with a 80k But 95% Single stage. Over the last 8 years, I've been insulating my 100 year old house, so we were over capacity. I've never had much luck getting accurate numbers from auditing programs, partially because it is a high mass house (double wythe brick walls) and partially because we are rigorous set back, window opening/closing, drape opening/closing people. I just took the gamble that an 80k furnace would be large enough.
    I've timed the heat cycles with both furnaces after most of the insulating was done. The old furnace ran at 7 min on, 7 min off at 20 degrees below, and the new furnace has been 8 min on, 10/12 min off at -15 degrees. This is reassuring. I should note that the house is only moderately insulated, especially on the main floor. Blowing the void between the brick wythes only gave me about R-12, and our wall surfaces are still somewhat cool.
    The big difference is in delivery temperatures. The new furnace delivers much warmer temps which is great, but I'm noticing the temp swings in the house more. I'm sure the old furnace fan was oversized. Delivery air was cool, tested efficiency at 82%, and rusty flue pipe are the indicators. Ductwork is oversized, having been adapted from the original octopus coal furnace. Static pressure at .22 IWC, all on the return side.
    What is a bit unnerving about the new furnace is that it is a little noisy and I'm very conscious of it running ($ going out the window). I can't turn down the fan speed any more and stay within heat rise limits. Wish I had a multi-stage furnace. I'll just have to wait until I get a couple of utility bills to see what the difference in energy use is.
    The old thermostat was a round Honeywell which we manually set back; the new one is a brand X programmable from the hardware store. It seems to cycle about 2 degrees, but the high and low temps are very noticeable.
    So, my conclusion is that there are a lot of factors that contribute to efficiency, comfort and customer satisfaction. Furnace size, fan speed, thermostat settings and duct work all interact with building performance.

    It gets complicated......

  4. wjrobinson | | #4

    If I ever had hot air again I
    If I ever had hot air again I would get a variable speed unit with a large turn down. I too really do not appreciate hearing air blow. The turn down units run so quiet that you don't know they are running. Totally worth the price.

  5. Expert Member
    Dana Dorsett | | #5

    Even a 46K furnace is too big...
    Even a 46K furnace is too big for most homes in the US built since 1995. A typical ~2500' house built to IRC 2012 code minimums will have heat load in the low 20s or lower, making the 40K+ output of a 46K furnace 2x oversized. That's not a huge hit in comfort or efficiency, but it's not exactly ideal.

    One of my co-workers whose townhouse has a design heat load in the mid-teens ended up hacking the internal controls on his dual-stage gas furnace to disable the high fire rather than put up with the short hot-blast finish of most heating cycles. I've suggested that when the thing is at end of life replacing it with a hydro-air solution running off the hot water heater, a solution that can be more appropriately sized to the load.

    Until there is market or code demand for it, right-sized hot air furnaces for newer houses aren't going to be made. It's not rocket science to build them smaller, but the replacement equipment market is larger than for new installations, and codes generally do not require load calculations or right-sizing of replacement equipment (except in California, under Title 24.) As the state of the art of air source heat pumps marches on, they may eventually become the preferred replacement solution for those who care about comfort & efficiency. The EIA recently noted electricity is growing as heating-energy source is growing in every region of the US, and natural gas heating is falling everywhere in the US except in the northeast.

    Even at near record-low natural gas prices, better-class heat pumps have a lower operating cost than condensing gas in many US markets.

  6. AlanB4 | | #6

    @ RJP
    Thats a lot of dry reading you sent me ;) I'll try to have a look at it later today
    The temperature here is now at the low design temp and the furnace is kicking into high fire and staying at it longer so that confirms it was just at the unsweet spot yesterday. They should have designed it to stay on high fire for 2 mins or something like that once its started. I am also happy there is some wiggle room in the Hot2000 calculation, the furnace is firing many times an hour but does turn off if only for a few minutes at a time meaning the rate of heat loss is not maxing out the furnace yet. I suspect another couple degrees colder and it won't be able to keep up.
    The thermostat is placed in the living room in a reasonable location, its above an air intake but i have adjusted vents as well as i can and i am reasonably happy with the results.
    There does not seem to be any logic preventing too many runs an hour, i also raised the temp 1 degree while it was running a few days ago, apparently the flame had shut down but the fan was running because i immediately heard it reignite and fan of course kept going.

    I assume you meant ECM motor or is GE something else? I specifically insisted on ECM, the old furnace fan ran at 400W, this one runs at 100W low, 200W high fire (measured by 100W resolution whole house energy use monitor), and 400W A/C. I would like to turn the A/C fan level down if possible (its very weak, i suspect its low on refrigerant or just super small capacity), i will have to look into it,

  7. rjparker | | #7

    Alan B 2 Stage Furnace
    If the thermostat is shutting down within 30 seconds of high stage, I would investigate the thermostat's staging control (or lack thereof). A Honeywell Prestige or similar uses PID control logic to learn the operating environment and stage accordingly. Or the furnace logic board could be doing the staging independent of the thermostat, which is not always the most effective strategy. The best staging controls utilize indoor and outdoor temperature along with operating history. There are also parameters for maximum cycles per hour that can influence behavior. Thermostat placement is a slight possibility. If you are not proficient in the logic of both parts of the system, I would call out a factory representative for the furnace who knows the thermostat in question.

    Update: Based on the info posted below, you have a single stage thermostat with the furnace controlling the staging. Replacing the thermostat would help but more wiring is needed. See

    A major advantage of two stage or modulating systems is the GE variable speed indoor fan motor (typically called an ECM). In addition to adjusting for stages, it starts silently, ramping up to desired speed over 30 seconds or more. It uses less power than a standard motor at the same cfm, plus it will adjust itself to maintain cfm when slightly restricted due to less than optimum ducting. It's clearly superior with forced air heating and or cooling.

    Finally proper staging and control can allow a nominally oversized system to run at 50-67% output for extended periods, allowing full output when needed for very high loads and setback recovery.

  8. gtallan | | #8

    Right-sizing furnaces
    Dana, doesn't a modulating furnace design more or less solve the issue for right-sizing a hot air system?

    My own experience is that when I replaced the worn-out 2-stage 80% furnace in my own home, about 9-10 years ago, I was surprised that my HVAC contractors proposed ANOTHER 2-stage 80% unit! I had to do my own research to find that better options existed and ask for a different proposal. As I recall I paid $300 or so more for the 93% modulating option with ECM fan. I didn't know enough about sizing back then to question their calculations (or guesses?); 75kBTU may be oversized for a 1930's 2000sf house even in Minnesota - certainly more so as I use this site to help improve the building envelope! But even on the coldest days it just runs at a near-silent idle most of the time, with no real temperature swings. As aj says, it's totally worth the price for the lack of noise, as well as improved comfort.

  9. fitchplate | | #9

    We love our hydronics ...
    Doesn't this (scary) blog make the plumbers and process engineers in us very happy to contemplate outdoor resets, modcon boilers. and the pex and copper snaking through our houses,

  10. user-723121 | | #10

    Alan B. Two Stage Furnace

    I also have a 2 stage furnace (Lennox) and there is a dipswitch setting for either 10 or 15 minutes run time on low fire before the furnace goes to high fire. Mine came from the factory set for 10 minutes on low fire, I changed it to 15 minutes so now the furnace runs on low fire almost exclusively once the house recovers from the nightly setback. You can also set most thermostats for "cycles per hour", my Honeywell was originally at 5 and I now have it set on 3. This overides the thermostat "temperature differential" somewhat and makes for longer furnace run times less often. Your White Rogers may have a temperature differential adjustment setting, some do.

    Last evening we were at minus 8 (design temp. is -12) in Minneapolis and the furnace would run for 12 minutes on low fire, 3 times per hour. I completely disagree with the common wisdom of furnace sizing for 2 stage furnaces at 140% of design temperature heat loss and here is why.

    A 2 stage furnace should be sized at 140% for low fire output. This way, even at or near design temperatures the furnace will heat the house on low fire and the DC motor will run quietly at low speed, using less electricity. We use a deep nightly thermostat setback here from 70F daytime to 62F at night. On these cold mornings the furnace will run continuously on high fire (66,000 Btu) for almost 3 hours before shutting down. The low fire input is 45,000 Btu for the 95% furnace. Once the house is up to temperature the furnace will keep the house warm very quietly on low fire. The design temperature heat loss for our house is 27,720 Btu's. The output on low fire is 712.5 Btu's per minute so at design temperature the furnace will run for 13 minutes every 20 minutes and stay on low fire.

    A furnace sized for 140% of design temperature will not allow for much of a nightly setback so the efficiency in lowering the Delta T for 8 hours can't be realized. Not to mention the extra wear and tear on the furnace running 24 hours vs 16 hours daily.

  11. kevin_in_denver | | #11

    Short cycling shortens component life
    As a landlord responsible for about 40 furnaces, I've found that I have to replace components more often when the furnace short cycles a lot. Of course that sounds like common sense, but lately I've seen the correlation very clearly.

    Some of the components have limited lifespans in terms of total cycles before failure.

    The HSI will get a crack and fail sooner.
    The microswitch inside the pressure switch will fail.
    The high limit snapswitch can fail.

    Replacing the filters can help, but poor duct sizing is a common culprit. In the example above, the burner short cycles on high limit because they kept the old ductwork and it's not getting enough air through it.

  12. Expert Member
    Dana Dorsett | | #12

    Modulation has limits (#9 Graham Allan)
    Modulation is great, but turn-down ratios are not infinite. A typical 2-stage 40K furnace has a low-fire of something like 25K, (a ratio of less than 2:1) and is still oversized by nearly 2x for a house with a ~15K design load. Modulating furnaces are rarely more than 2:1, for the burner, even if the blowers are continuously variable ECM types.

    Modulating heat pumps can have a turn down ratio as high as 5:1 or more, and even crummier ones have 3:1. But not all are suitable for a MN climate.

    Doug's example of a low-fire output of 42K for a house with a design heat load of 28K is a prime example of how it is NOT a great solution. If you had a HIGH fire of 35K and a low-fire of 20K the thing could actually modulate during cold weather, stepping up & down between firing ranges. There's no point to a modulating or 2-stage unit if the operational mode is turning it on/off at it's lowest fire. You might as well just hard-wire the thing for low-fire (as my co-worker did.)

    The reduced heat loss from setbacks doesn't necessarily reduce fuel (or power) use with condensing multi-stage or modulating units. The highest efficiency is at low-fire. A 2 stage unit that runs say 96% efficiency at low fire will only get about 90-92% efficiency at high fire, and the efficiency of the air handler at high cfm is dramatically lower than at low cfm. The energy "saved" from the lower heat loss from the setback is mostly heat that would have been delivered at 96%, but the heat delivered during the recovery ramp is at 92%. Between the lower efficiency and higher power use during the recovery ramp, it takes a fairly long & deep setback before it's a significant cost or fuel savings. As an experiment to see if you're saving any fuel with setbacks, try running it in constant temp mode for a periods between meter-readings, then compare fuel use against heating degree days for that billing period compared to billing periods when you've been running it in setback mode.

    As Kevin's experience illustrates, most of the "wear & tear" on furnaces are on the ignition & start up cycles. If they can run continuously the number of burn cycles and motor start-up ramps falls off by orders of magnitude, and the thing will simply last longer. Running 24/7 REDUCES rather than increases the wear on the systems.

    But until the market demands it, more appropriately sized and modulating units with bigger turn down ratios won't be readily available.

  13. user-723121 | | #13

    Thermostat Setback Does Save Energy

    I did the setback vs non setback experiment about 5 years ago. Eliminating the 8 hour daily setback increased gas usage by about 5% on a year over year test for this house with this 2 stage furnace.. This was measured in Btu's /SF/HDD. To increase accuracy the total available solar minutes for the heating seasons would have to be compared, I could not find this information for Minneapolis. Before upgrading the furnace I did some air sealing, increased the attic insulation from R-38 to R-100 and added R-10 foundation insulation. The furnace choices were the 45k maximum or the 66k, I would not change a thing and really like how the 66k furnace has performed over the last 8 years. The furnace we replaced was a 28 year old 150k input gas furnace, now that was oversized but it did warm the house quickly in the morning.

    I believe a furnace will last longer running 16 hours per day versus 24. We heat our house on low fire and low fan speed which increases efficiency. I like the setback, we sleep better and the furnace does not run at night except during the very coldest weather. Controlling the cycles per hour allow longer run times less often and helps to lower the furnace cycles.

    A furnace sized at 140% of design temperature heat loss will not allow a deep setback in a cold climate, the long recovery time will be unacceptable.

  14. adam5532 | | #14

    Modulating furnaces
    Both Carrier and Trane now have fully modulating gas furnaces with a firing range of 40 - 100% in 1% increments. They also have fully variable speed ECM motors. The smallest is 60k BTU.

  15. Expert Member
    Dana Dorsett | | #15

    If the smallest is 60KBTU/hour... (response to Adam)
    ... at 40% it means the minimum firing rate is 24,000 BTU/hour.

    The min-firing rate is the approximate design heat load of a pretty good IRC 2012 code-min 2000 square foot house. That means it's never really going to modulate if installed in a new small to mid-sized house, and might as well be a 2-stage. 24K/60K, which would allow signficant setbacks without unduly long ramp up times.

    As a retrofit in an older house with a design heat load of 40-45K it might be a pretty decent fit, since it would spend a decent amount of time in a modulating mode.

  16. user-2588683 | | #16

    Heating a small (750s.f.) house.
    I am planning to replace my HVAC system this coming summer. I presently have an old 55K BTU forced air unit that nobody could ID by model number when the blower went out recently. I have to think it is pretty old.

    The house is in Central Illinois on a slab, pretty well insulated - 3 inches fiberglass in the ceiling, 7.5 inches Polyiso foam in the attic under the roof, 2 inches of spray foam and 1.5 inches Polyiso in the walls, new doors/windows. When I have used online load calculators, the results are in the 30K range.When the blower went out recently I had to use the oven (I know, I know) but I would heat the house to 70 at bedtime and it would be right at 50 in the morning with the outside temperature at 0 degrees.

    I have spoken to a couple contractors and they seem to be on the fence of a 2 stage 80% vs a high efficiency because my highest bill is in the 100.00 range and they question if its worth it. I am more interested in the 2-stage with an ECM - whatever efficiency - because the furnace I have doesn't run very long and it gets overly cool before it runs again. There is also the small selection of furnaces that small.

    I am intrigued with the idea of a hydronic coil in an air handler I would like to put it in the conditioned attic because space is at a premium. Buying a boiler seems like significant overkill to me, but I'm not opposed to an efficient water heater with a coil. However, after reading this and other information, I am very interested in the continuous water heater approach.

    The house is - obviously - only one bath and while I live here alone, I think it would be unusual for more than two to live here.

    I am looking for thought/opinions/feedback, if you're so inclined. I have two fears: 1) Such a setup wouldn't keep the house warm on days that are below 0 degrees and, 2) I wouldn't have the water for a long, hot shower. In fact, I built a nice walk-in shower and my occasional long shower is limited due to 40 gal water heater.

    Thanks for any thoughts.


  17. KeithH | | #17

    Sizing website and ha ha contractor sizing
    Ok, really, I have no relationship with this site (I've mentioned it a few times on GBA). I had a pro Manual D/J performed ($$$) on my house. I also used (intelligently) to calculate my house's load. Load calc managed to come up with a number within 10% of the manual J/D and my house is kind of complicated.

    If you, like me, have NEVER seen an hvac contractor perform or offer to perform a sizing other than farting out 30 btus/sf (a number which oversizes my house 200% over the manual D/J), you might want to try it out. Fair warning: even my pro manual D/J won't convince most hvac installers to size appropriately. They always want to fudge the D/J up one size.

    The last mini-split install conversation went like this:
    Me: "I want a mitsu mini split heat pump for this unducted installation"
    Contractor: "Ok, 380 sf ... let's call that 400 sf ... ok 30 btus/sf. So 12000 btus ... but you don't want to be cold ... so let's do the 15k unit.
    Me: "I have a manual D/J that says the load as built is 7k"
    Contractor: "Well, I guess we could do the 1 ton unit"
    Me: "I'm going to replace the windows in the next few years and add insulation above. Let's do the 9k unit"
    Contractor: "I can't do that. Here is my bid for the 1 ton unit"
    Me: "I'll get back to you on that".

  18. KeithH | | #18

    Dual stage furnace as the solution
    To the folks suggesting dual stage eliminates the oversizing problem, I'm like to remind you that a 80k load is for the design day. So if you oversized for the design day, you'll be radically oversized for the average days. Not to mention the fudge factor on the manual d/j plus upsizing a unit plus upsizing for a bigger ac coil and all of the sudden you have a furnace sized to 200% of your load (like my 70k/115k 95% efficient amana for a 50k house that only runs more than 5 minutes an hour on design days).

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |