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

Sizing a furnace

mtaylor12345 | Posted in Mechanicals on

I was reading an article about HVAC sizing with Manual J & D.

Here is my question:

If you have data for an existing home — heating-only therms, furnace run-time (from smart thermostat), HDD, etc for the coldest days of the year for multiple years — can you calculate furnace sizing in a different way?

For example, I used 62 heating therms in Feb 2015 (after subtracting summer non-heating gas use), so I put in 6.2 million btus. I have a basic furnace with one blower speed installed in 1989, so it’s probably 75% efficient (as a guess). So the house required 4.7 million btus of heat (1.5 million lost via combustion).

The furnace ran for 61 hours that month so about 77,000 btus/hr heating input.

The average hourly loss was 7,000 btu/hr (672 hrs that month), but on the coldest day, I know it ran 4.25 hours or 327,000 btus input that day, which is about 13,600 btus/hr per hour heat loss. I could continue this process to create a scatterplot and fit a curve to the record cold temp in our area + 5%.

I imagine I’m not the first person to think along these lines…Is there anything to this? Can it be extended to finding the right sized furnace?



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. Expert Member
    Dana Dorsett | | #1

    Indeed it can be used for sizing the replacement furnace, but you're not tackling it from quite the right direction. Heat load is fairly linearly proportional to the difference in indoor & outdoor temperature, and you're looking only at fuel use over time, not temperature. To make it temperature dependent you also have to factor in the average temperatures.

    Here's my take on estimating loads and furnace sizing based on fuel use:

    It's really coming up linear constant of heat load for every degree F below the heating/cooling balance point temperature, and projecting what the load will be at any arbitrary low temperature.

    Using the load at the historical 99th percentile temperature bin (87 hours out of every year is cooler than that) is preferable to the all time record low for sizing the equipment. It doesn't take much of an oversizing factor to be able to cover the load at the lower temperatures, and temperatures typically won't dwell below the 99% temperature for long enough to matter within the lifecycle of the equipment, or, even during Polar Vortex events. ASHRAE recommends oversizing by no more than 1.4x the 99% load.

    If you want to dig deeper and better estimate the heating/cooling balance point as well as the load you COULD use a regression analysis tool:

    For simple furnace sizing problems that's a bit overkill- the higher accuracy doesn't buy you much, when furnace capacity comes fairly hefty steps, not a continuum.

  2. Jon_R | | #2

    Note that such methods don't properly account for peak winds - which, in a leaky house, can be a significant factor in peak energy use. On the other hand, ~1.4x over-sizing compensates for significant errors.

  3. Expert Member
    Dana Dorsett | | #3

    Such methods also don't account for hot water heating, which can cancel out quite a bit of wind, or solar gain, which can be problematic in sunny more temperate climates.

    Subtracting out summer use isn't necessarily very accurate, since incoming water temperatures increases fuel use per gallon of water in winter, and there is usually more hot water volume used in winter. These methods also don't factor in solar gain, which can be a large percentage error in temperate sunny locations.

    But 62 therms total heating use for the month of February implies a pretty low heat load. (Where is this- the California coast or something?) It's doubtful that there is a hot air furnace that can be properly sized for your actual load. But small hydronic air handlers or coils running off the hot water heater can serve it up pretty nicely, at higher comfort & overall efficiency than separate hot water & space heating appliances.

    As an example a 1-ton air hydronic handler like a FirstCo CW4 ( ) specifies the output at three different flow rates, and four different entering water temps.

    It'll deliver 11,800 BTU/hr at 1 gpm at an entering water temperature (EWT) of 120F.

    Pump it at 3 gpm and that rises to 14,500 BTU/hr.

    Crank the storage temp on the tank up to 140F and it's good fo 16,600 BTU/hr @ 1 gpm, 20,300 BTU/hr @ 3 gpm, which is probably enough for this situation. (Run the fuel use load calculation and find out) .

    That's enough to heat my sub-code ~2400' 2x4 framed antique (+ ~1500' of insulated basement) at an outdoor temperature of +25-30F.

    A typical atmospheric drafted 50 gallon water heater has a burner output between ~30-35,000 BTU/hr, so some up-sizing might be called for if it turns out your 99% heat load is over 20,000 BTU/hr, or you may face some pretty long recovery times after even a fairly short shower during the coldest hours of the season. But condensing tank hot water heater with a 76,000 BTU/hr burner would have PLENTY of capacity for handling both the heating and hot water loads.

    By giving the water heater more load, it spends less time in standby, for lower standby losses. Those losses can be pretty significant for atmospheric drafted units, which is why the hit only 55-65% efficiency in an EF test, even though the steady state combustion efficiency is usually ~78-82%. With condensing water heaters the standby losses are much lower, but still not super-low, and putting it to use for space heating is still an net efficiency improvement.

  4. mtaylor12345 | | #4

    Thanks for the responses. This is northern Utah - plenty of cold and snow. My highest usage is usually in January, up to 150 therms or so. 2650 ft2 brick with no wall insulation and only some (flat) roof insulation.

  5. Expert Member
    Dana Dorsett | | #5

    Got a ZIP code? (To estimate outside design temperature, and for weather data). The 99% outside design temp in Ogden is +11F, but in Logan it's +2F, so it kinda matters WHERE you are in northern Utah.

    Have you run the fuel-use based load calculation outlined in the blog piece?

    Hydronic air handlers come in many sizes, and at your 150 therms in January fuel usage it may STILL be a better choice than a hot air furnace.

    eg: Last January Logan UT saw ~1500 HDD65, so you used about 0.1 therms per degree-day, or (/24=) 0.0042 therms per degree-hour which is 417 BTU per degree-hour. At 80% efficiency that means the heat going into the ducts was 0.8 x 417= 334 BTU/degree-hour.

    The design temperature of +2F is 63F below the presumptive 65F heating/cooling balance point, for an implied load of

    63F x 334 BTU/degree-hour = 21,042 BTU/hr.

    Using your real location's data and meter reading dates & usage it's possible to be more precise about it, but it's pretty clear that you'd be able to get there with a water heater and hydronic air handler, but it may need to be one size step up from the smallest of the line FirstCo CW4, or not- it really depends on where the calculations come out. The burner capacity of the water heater would have to be verified that it's up to snuff too.

    If the low usage is only due to super-deep overnight setbacks and a low thermostat setting, the calculations would have to use something other than 65F for the base temperature, but then projected back to 65F or so (to be sure the heating system can still cover the code-minimum 68F at the 99% outside design temperature.)

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |