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

Community and Q&A

Calculating Heat Loss

nybhh | Posted in Energy Efficiency and Durability on

I want to calculate ACTUAL (not hypothetical) building heat loss.

I recently purchased a wifi thermometer/monitor for a well insulated, wood stove heated outbuilding on our property.   The app shows these really cool charts of the temperature fluctuations over the course of the day and this data can even be exported in csv format at various increments down to 1 minute for any time frame desired.

This got me thinking that there should be some formula to calculate the ACTUAL heat loss of a building rather than the hypothetical heat loss simply by tracking the temperate change of a building relative to the outdoor temperature.   My semi-exhaustive search on Google yielded no results so I wanted to ask if anyone here was familiar with a method to do so?

Ideally, one could simply place one of these $20 devices inside and one outside, export the data over a 24 hour period and plug that into an Excel spreadsheet to spit out an actual heat loss number, ideally in BTUs or some unit that would be usable for AC and boiler sizing, etc.  The more data you fed it, the more accurate it would be but it seems like even a few manual measurements over the course of a few hours could give pretty accurate results.

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. charlie_sullivan | | #1

    When it's heated with a wood stove, it's really hard. How fast it cools down depends on the insulation and the thermal mass, so you can't tell how much of it's caused by which without further data. If you knew the efficiency of the stove weighed the wood, and measured its moisture content you might be able to do something with that, but that's pretty iffy. If you have electricity there, you could heat it with electric space heaters only for 24 hours and measure their electric energy consumption and then be able to calculate pretty precisely what the heat loss is.

  2. nybhh | | #2

    Thanks Charlie. There is electricity, its a wood shop and I will be adding a boiler at some point prior to turning on the water supply but just haven't gotten around to it yet as it is a big expense and I haven't really needed it yet with the wood stove . Point is, I have and could use one of those portable oil-filled electric radiators to do this if there is a formula I can use.

    My original thought process however was to not use the wood stove during the test due to the complications of tracking btus off the stove as you suggest, and just tracking the heat loss change over a period of time vs the outdoor temperature but that may be a much more complicated math exercise than what you suggested. The building is very well insulated and there is only a few degrees of temp swing overnight with a 30-40 degree outdoor temp delta.

  3. tallpinescabin | | #3

    Absolutely, not that hard depending on fuel method.
    In my own home, I've read the gas meter a couple times over an extreme cold snap, and made an assumption about other uses. You can pretty quickly figure out heat loss that way, but that was for a forced air gas furnace.

    Electric wouldn't be too bad to do either. Wood stove would be more complicated, but doable. I've done that in a relative's cabin....weighed the wood I used over a specific time period, measured the moisture content, and made some assumptions to arrive at a BTU loss number.

    I think the important thing to get this right is to do it over a long enough time period that you don't have to worry about thermal mass effects as much, etc. Say over 2 weeks in really cold weather, you get more cycles of data input/output.

  4. nybhh | | #4

    The more I think about this, the more it seems critical to me that the outdoor air temperature get factored in for an accurate calculation of heat loss, and that is also constantly in flux also which would need to sort of multiple samples over the same period of time. How do you guys account for that?

    It will require a lot more internal BTUs to maintain a 68-degree indoor temp when it is 20-degrees out than it will when it is 65-degrees out.

    1. charlie_sullivan | | #5

      Yes, you definitely need the outdoor temp. And it's tricky if it's varying. If the variation is a steady cycle, you can factor that out by just measuring for 24 hours, and using the average outside over that 24 hours. If varying a lot day to day as well then you are better off averaging over a week or two. But if you look at a forecast you can probably find two days where the high temperature is pretty close, and plan to do it from say 3 PM one day to the same temperature at 3 PM the next day.

      You'd actually want to run for two days: The first 24 hours you are just getting the inside temperature stable. Then you make no change to the temperature, but start monitoring the energy use. If all goes well, 24 hours later you are back to the same outdoor temperature and you can record the energy use, and the average outdoor temperature over those 24 hours. That's all the info you need.

  5. Expert Member
    Michael Maines | | #6

    Brandon, I think the most accurate approach is to compare what you use for heating fuel to the annual heating degree days, adjusted to the baseline temperature you keep your shop heated to. Then divide that by the delta-T at any given time to estimate hourly demand.

    1. charlie_sullivan | | #9

      Michael, that approach is great if you are heating is to a steady temperature with a known amount of fuel. If you are heating an outbuilding with a wood stove you aren't keeping it at a steady temperature, and you don't know the amount of fuel consumed or the efficiency very accurately.

      1. Expert Member
        Michael Maines | | #10

        Ah, good point--I should have read more carefully.

  6. Jon_R | | #7

    I collect hourly furnace burner runtime, indoor temp, outdoor temp, wind and sky condition (to eliminate solar gain periods). From this I fit the data to something like:

    btu used = deltaT * 701.00 - 2710 (internal) + deltaT * 12.00 * wind

    Note that to get a design load, you need the deltaT and the often neglected wind.

    As Charlie said, with metered electric heat (perhaps a single day, preferably much more), you could do something similar. Or do like most and trust a Manual J.

  7. BirchwoodBill | | #8

    Look for Dana Dorset’s method for using your utility bill. I used that to downsize my boiler.

    1. Expert Member
      Dana Dorsett | | #14

      That works for a whole-house load estimate, but becomes more difficult for just a an outbuilding, shop or garage, unless one runs a test using separately metered electric space heaters keeping the space a constant temperature over a cold week or more.

      For reference on fuel-use load calculations, see:

      Wood stoves are all over the place on actual combustion efficiency, as is the BTU content of the wood itself, which adds a large error to the load numbers, even if one were able to keep the indoor temps constant. An I=B=R load calc using a mere WAG for air infiltration would be more accurate than a calculation based on cord wood. An I=B=R load calc (see: ) would also be good enough for sizing any future heating equipment. Unless the building is pretty large the odds are good that a single ductless mini-split (of a type appropriate for the local climate) would be comparable or cheaper in up-front installations costs as a propane or oil fired boiler solution, but in most markets would be cheaper (sometimes substantially cheaper) to run than a boiler, and easier to right size for the load.

      All oil fired boilers would be too large, and only the smallest propane fired condensing boilers are likely to modulate low enough to provide reasonable comfort & efficiency. There may be right-sizable propane fired wall furnaces that would work, but we won't know until we get a better handle on the actual load numbers.

  8. nybhh | | #11

    Hi all,
    Thanks for the suggestions, particularly the formulas.
    I feel like there is probably some way to calculate this with no heat inputs actually simply by monitoring the lag in temperature changes compared to outdoor air temps over a few days but the math may just get too complicated, too quickly to be of any real use. Thanks again for the recommendations. I may give it a go with a small electric space heater but I'm not sure 1500 watts (5118 BTU/hr) is going to cut it in an 800 SF outbuilding.

    1. charlie_sullivan | | #13

      An option is to just leave the 1500 W heater on full blast without any thermostat control, and see what temperature rise you get, even if it's not as much as you'd like. If you do that over time and log inside and outside temperatures, there is then enough information to extract both the thermal mass and the thermal resistance.

    2. Expert Member
      Dana Dorsett | | #15

      >"I'm not sure 1500 watts (5118 BTU/hr) is going to cut it in an 800 SF outbuilding."

      1500 watts is 5118 BTU/hr. That would cover a load/square foot ratio of about 6 BTU/hr per square foot. For a 2x4/R13 type structure with clear glass double-panes and R20 in the attic/roof that should cover the load down to an outdoor temp in the low 40s F, perhaps lower if the total window & door area is low, and for a 2x6/R20 with low-E windows it would be good down to the low 30s F or lower.

      Of course pair of them would cover even lower temps.

      What is your 99% outside design temp?

      1. nybhh | | #16

        Ok cool, part of this is to size a a boiler so I’m prepared to jump through some hoops for this and enjoy this sort if thing.

        I could not find a 99% temp for my exact area but I would guess it is about 3 degrees F. I’m in the mid-Hudson valley of NY at about 1000 ft of elevation (41.793710, -74.222961). The building is very tight, R21 in the walls, R60 in the cathedral ceiling assembly, R10 below the slab (on grade) and along the inside of the frost walls. Windows are argon-filled double paned Pellas. There is PEX in the 6” slab, 4 loops of about 175 - 200 ft each and we have a buried 250 gallon propane tank with a line feeding the building. My plumber wants to go with the smallest Navien combo boiler unit for heat and domestic hot water but I’m concerned it will be oversized.

        I have one of those Watt-o-Meters that I can use with the radiator to get an exact energy use number as well. Is there an article here that explains exactly how to do this?

        1. Expert Member
          Michael Maines | | #17

          Brandon, just curious--why don't you trust standard heat loss calculations that are typically used for this situation, assuming the inputs are correct?

          1. nybhh | | #18

            Intellectual curiosity? Once you have the actual building, why not test it and see how it stacks up against the calculations? Spray foam in the ceiling has the same R value as batts but seals better against air infiltration, no calculations takes into account me spray foaming all the gaps in the sheathing before insultating the walls, or taking extra time to cut rigid foam for all the odd corners prior to exterior sheathing. How much heat am I losing through that chimney, or the plumbing vent, or the 8” exhaust duct for my dust collection system, that huge south facing window has to help with solar gain right?... etc...

          2. Expert Member
            NICK KEENAN | | #22

            Manual J uses a simple linear model where heat loss is directly proportional to the temperature difference. I suspect that simple model is overly simplistic.

            However, empirically it seems to give reasonable results.

        2. johngfc | | #19

          The 2 common ways to get the 1% temperature are using either hourly or daily data. I think the hourly method is preferred but not sure. For your location, here are the 1% temps using daily data from Jan 1 1980 to Dec 31 2019. It's getting warmer, so these are conservative for heating and optimistic for cooling.
          daily min temp - 1% -2.2
          daily max temp 99% 93.2

          You can get the data from here:

          1. nybhh | | #20

            Nice link, thank you! 👍

          2. Expert Member
            Michael Maines | | #23

            Heat loss IS directly proportional to the temperature difference, assuming an airtight assembly and not a lot of thermal mass. Manual J includes inputs for air infiltration, which a blower door can help you determine. I'm interested in alternative approaches that may be more precise or just plain different but would like to see them compared to an accurate Manual J to see how far off they are.

          3. charlie_sullivan | | #24

            Michael, even with a perfectly air tight assembly, linear is an approximation. A very good approximation, but not exact. Window U-factors are different at different temperatures, and insulation R-value changes with temperature. For most insulation the effects are mild--only polyiso has a variation there that is worth considering in practice. So it's only an academic point, but it in fact a simplification to say it's only proportional to temperature difference.

  9. bfw577 | | #12

    An electricity monitor can be used to get an extremely accurate heat loss number. I connected one to my oil furnace and logged an entire winter before converting to mini splits. It was as simple as seeing how long my 80k btu oil furnace ran each hour. I then made a chart off the outside temperature. Example at 20 degrees my oil furnace ran for 15 mins every hour. 15 mins of runtime was about 20k btu.

    Using this technique I nailed sizing my mini splits perfectly.

    1. Expert Member
      NICK KEENAN | | #21

      This is the way to go. Do you have a link to the monitor you used?


      1. bfw577 | | #25

        I use an efergy engage hub. It cost around $125 with one set of sensors. I bought extra sensors to monitor my solar and mini splits. It logs data in 15 second intervals.

        Here is a random screenshot of the ap of my dryer running for example. You would see a similar on off rise with your heat source per hour.

        1. CollieGuy | | #29

          We have three Efergy Engage TPM monitors, one for our main feed and one for each of our two ductless mini-splits. That gives us a pretty accurate picture of their energy requirements, but no measure of the heat they supply nor, by extension, our home's heat loss. For that, I ran our oil-fired boiler as our sole source of heat for a one week period in February, 2018 and tracked its consumption with our MazoutMan monitor pictured below, and our indoor and outdoor temperatures with our Netatmo weather station.

          Over the course of that one hundred and sixty-eight hour period, we used 112.87 litres of fuel oil, with an average indoor and outdoor temperature of 20.1°C and -1.9°C respectively. That works out to be an average of 0.67 litres per hour or 5.95 kWh(e) of heat at an AFUE of 83 per cent. If those numbers are more or less correct, that puts our home's heat loss in the order of some 0.270 kW per °C (52 year old, 2,700 sq. ft. Cape Cod with no passive solar gain to speak of due to heavy shading).

  10. nybhh | | #26

    I wanted to thank everyone who contributed to this. I'm currently running a calculation based on this excellent article kindly written and posted by Dana.

    I'm going to continue to run it for a few more days to collect more data but my internal temperature with one 1500W space heater has stabilized around 65F with no wood stove inputs. There is a weather station (KNYKERHO8) very close by so I am using their data and the last two days have been 36HDD @ 65Fand 31 HDD @ 60F.

    Assuming the electric space heater is 100% efficient and outputting 5,118 BTU/hr (122,832 BTU/day), my calculations are currently estimating an implied heating load of 9,575 BTU/hr at 60F balance point and 8,956 BTU/hr at 65F balance point but that is making no corrections for the fact that 1500W is holding an average temp of 65F instead of 68F and I'm a little unsure how to compensate for that in the calculations.

    The article does discuss but I'm a little confused on exactly how to implement it.

    "For example, if you normally keep the thermostat at 62°F rather than 68°F, subtract 6 F° from the temperature bases to get the BTU/degree-hour constant, but add 6 F° to the total heating degrees when you run the final number to be sure it meets code when sizing the equipment."

    Since my internal temp is averaging 65 degrees instead of 68, I have a 3 degree offset but I'm not sure exactly how to offset that 3 degrees from my "temperature base". If it means that the implied BTU's/hr @ 60F balance point is = (2F is my 99% design temp)
    (60 - 3 - 2) * BTUs per degree hour @ 60
    then it actually gives me a lower implied head load per hour which is opposite of what intuition tells me if it were holding at 68F instead of 65F.

    Perhaps the formula is ((60 + 3 - 2) * BTUs per degree hour @ 60 or perhaps it means I download 63F and 57F data from the weather station? Anyway, a little clarification here would be of great assistance while I collect more data.

    My plumber wants to install a Navien NCB-150E which is a combo boiler/DHW unit since this building is plumbed with a full bathroom (w/ shower) and a kitchenette (w/ DW). I'm using the building as a wood shop but it is designed to also function as a guest house if needed. The Navien NCB-150E has a minimum BTU/hr input of 12,000 BTU's/hr @ 95% AFUE. Provided my data remains in this same range, do people think this unit appropriately sized? They have a model NHB-55 which goes down to 8,000 BTU/hr but it does not have DHW also so we would need to add a separate hot water heater (probably electric). If it weren't for the shower, that might make sense. I'll honestly probably leave the thermostat at 60 degrees most of the winter and use the wood stove when I need it warmer. There is PEX in the 6" concrete slab so the original plan was to deliver heat to the building via radiant floor so I'm not really as concerned with temperature swings as much as I am with short-cycling the unit.

    Thanks again everyone for all the help.

    1. charlie_sullivan | | #28

      You should be sizing the boiler based on its maximum output not its minimum output. Both of those are drastically oversized. You could get around the short cycling problem with a buffer tank, but that gets expensive.

      Here's a discussion of very small boilers.

      A minisplit is the straightforward solution, but there's also the Chiltrix air-to-water heat pump, which is oversized for your application but not as severely.

  11. Jon_R | | #27

    > wants to install a Navien NCB-150E

    Consider a mini-split heat pump instead. But the amount of use effects the economics.

  12. nybhh | | #30

    I have nothing against mini-splits for cooling but forced air heating isn’t very common here so its just not something I am familiar with, especially in a shop with sawdust. Since I already have propane to the building and PEX installed in the slab, I’m more inclined to try and find a solution that takes advantage of that although I’m not going to rule it out. I do have Reynaud’s syndrome in my hands and feet so a heated slab with no draft does help with that but it is a secondary consideration.

    Mini-splits also don’t address DHW so I’ll still need a solution for that as well and I only have 100amps to this building. Two electric heating appliances would take a big chunk out of that and I have several 20 amp 240V tools hooked to a 30 amp dust collection system so limited excess electric.

    we considered a propane fired water heater for both also but with dust collection, there is potentially negative building pressure and my contractor liked the Navien because it has its own fresh air intake and he was not aware of a water heater that did.

    Edit: The link to the discussion about tiny electric boilers is interesting. I’ll investigate some if those options as well, thanks for the link. The Electro Industries unit might be just about perfect but still needs a DHW solution..

  13. charlie_sullivan | | #31

    Chiltrix only draws 13 A and can use s 20 A, 240 V circuit. It can supply heat to your floor without needing a buffer tank and can supply heat to a DHW tank. But at $4300, it is pretty expensive.

    You might be able to get away without a buffer tank with 12,000 BTU/h minimum firing rate boiler by using the thermal mass of the floor as a buffer. The downside is that if you charge the floor with more heat than you need, you overheat the space. Whereas if you charge a tank with more heat than you need, the heat can stay in the tank. It can probably work fine but getting the controls on it right is tricky.

    Plenty of boilers and water heaters have fresh air intake. Search words are "sealed combustion" or "direct vent". I'm a little worried about your plumber if he thinks that's rare and hard to find.

  14. nybhh | | #32

    Thank you Charlie. I just took a quick look at the Chiltrix unit and that obviously checks a lot of boxes as well, other than entry cost. I'll need to dig into that a deeper tomorrow but that electrical draw is very doable as well and I have a spare dedicated 10/3 pulled to the the utility closet already.

    I can't dig up the drawing right not but I think there is about 800-1,00 LF of 1/2" PEX A in the slab (9-10" on center, 4 loops) so that works out to be only about 7-9 gallons of water/glycol in the slab so not sure if that is enough to get around a buffer tank or not?

    As far as the sealed combustion and direct vent, I was specifically referring to hot water heaters as an option for both the floor and DHW when I brought it up to the plumber and to be fair, he didn't seem thrilled by the idea so it might have something to do with that. He's worked with me on this property since we purchased it almost 10 years ago and has done a LOT of work here on several different structures and project types and has been great but I do agree that his strengths seem to lie with maintaining existing systems and regular plumbing and not the design of new HVAC systems. I haven't practiced in many, many years but I have an architectural background so I'm somewhat comfortable taking up some of the slack on the design side but certainly not all of it. Thank you.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |