Accuracy of Heat Loss Calculations
I’ve calculated our home’s heat loss using the method outlined by Dana Dorsett here:
I realised after the fact that not all of our gas readings are actual, some are estimated so I did a second calculation using a much longer period.
House is about 2600 sq/ft heating area, zone 7A. Home is over 100 years old but with retrofits. Heating is forced air gas. Water is on demand, gas.
99% design temp is -17F
Calc 1: (gas use estimated for one reading)
Jan 19 2022 to Feb 14 200
HDD65 – 1695
115.43 m3 gas used (adjusted for 97% efficiency on furnace 66K/64K BTU)
27.4K BTU/hr heat load at -17 F.
Calc 2: (actual readings, longer period)
Oct 18 2021 to April 13 2022
HDD65 – 8281
549.2 m3 gas used (adjusted for 97% efficient furnace 66K/64K BTU)
24.6K BTU/hr heat load at -17 F.
If we are breaking down these numbers to BTU/degree day, and we assume heat loss is linear, then how much confidence can I place in these numbers? I figure they are actually overstated as they include gas use for hot water which looks to be about 30 m3 per month (based on gas use in summer when heat is off) on average. If we assume this, then the heat load numbers would be about 25% less yet.
I’m finding this a bit hard to believe, although the home does compare quite well with those around us. See attached pic.
I’m trying to get a sense of whether an ASHP would make any sense down the line as we try to wean ourselves off gas.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
A couple of questions:
What do you set the thermostat to?
Do you use any other heat sources?
Is the 2600 sqft all above ground?
Is this a detached home?
Very efficient numbers! Easy to switch to a heat pump - if you feel any concern about the heat loss, you can easily incorporate a hybrid system with a gas furnace as backup.
You can usually get a pretty good estimate this way, but note that you can also easily take your own meter readings so that you can look at any arbitrary time period you want. All you do is read the dials, and you don't read the dial as a number until the pointer is PAST the number. That means wherever the pointer is pointing, you read it as the number BEFORE wherever the pointer is, unless the pointer is dead-on pointing at any particular number on the scale. Note that it's common for every other dial to rotate in opposite directions too. It sounds complicated but it's actually very simple to do once you've done it a few times.
If you take some readings over 24 hour periods, you can compare your gas use to outdoor temperatures and wind conditions to glean a little more info about your home's performance too.
Bill
Paul, thanks for the reply. Thermostat is at 22 C day, and 20 C night. My wife and girls are not happy below 22 C :-) Floor area, inside measurements are 1800 sq/ft on top two floors, and about 800 in basement. Home is fully detached. I have a hanging gas furnace in a 16x24 detached garage, but only use it when working in there.
You don't see too many heat pumps here due to the -17 F design temp, and likely just HVAC "inertia" for the local installers. The condensing gas furnace is pretty new...but yes, I'd like to look at the best way to integrate an ASHP to the existing furnace.
Bill, I did take a look at the gas meter and read readings (it's a metric version, showing m3) on one of our colder days at -7 F, but the hot water use I suspect really skews a day's data. I need a cold day (s) with zero hot water use. Part of my question was around the longer periods and accuracy vs checking over a day or two.
The installers spec'd a 66K Btu furnace which seems pretty drastic overkill, however from all accounts, that is not unusual in the trade. Although quite efficient at 97%, the bigger comparative difference came with reduced power use due to the ECM fan motor.
I installed that same basic furnace (2 stage) in 2006 in MSP and could not be happier with it. The ECM fan is the way to go. I think using metered gas usage is quite accurate in determining building performance. On a calm and cloudy steady state (temperature) day, timing furnace run times with a known efficiency furnace is going to get you very close to your hourly heat loss.
Doug
Doug, the furnace is a modulating Lennox. Other than an intake air pressure sensor issue on year 1 (middle of winter, zero heat), it’s been excellent. The only downside really is that the control/touchscreen is digital, so replacing it with something “smart” would actually dumb it down considerably. The CFM adjustments in particular via the service menu are very handy.
So if I can test the house with zero water use, over say 12 hours and calculate average HDD65, how does one account for wind speed? The worst case scenario would be -35 C (colder than our 99% design temp of -17 F) with 50-70km/h winds. Is this overthinking the heat load calcs?
I think you’re overthinking, as the HDD 65 vs fuel usage data isn’t using 0 mph wind, it uses average. With the furnace backup, I’d size aggressively. Cumulative heat loss isn’t evenly distributed across outdoor temps, so even if the heat pump only meets 90% of heat loss on the coldest day, across 12 months, it’ll provide much more than 90% of the heat.
I like Dana Dorsett's method because it relies on actual measurement rather than modeling, but if your house has significant solar gains that method will underestimate the heating loads.
Bopping into an old thread because it relates to my current situation...
This is a really interesting point. We have gotten an extremely suspect heat load estimate from an energy audit, so are more or less relying on Dana's method to size a heat pump. Do you have an idea of how we might estimate solar gain and take that into account?
This article can get you started:
https://www.greenbuildingadvisor.com/article/a-quantitative-look-at-solar-heat-gain
It has solar gain per square foot of window for 22 locations. If one of them is close to you use it. Basically you'd add the solar gain to the fuel usage.
Thanks! I appreciate your help. I've been mulling this some more. Since we're using real-world data, wouldn't solar gain have come into play during the measurement period, and thus already be factored in?
I guess what I mean is that we can expect the same solar gain as we saw during the measurement period, so I'm not sure why we'd add it.
Nope.
What you're trying to calculate is maximum heating load. The presumption is that heating load is directly proportional to the outside temperature, so for a time period you measure the average outside temperature and the heat consumed to get a coefficient, and then you apply that coefficient to the design temperature.
By ignoring solar gain, you're underestimating the heat consumed and thus underestimating the coefficient. But at the design temperature there is no guarantee there will be solar available -- in fact it's most likely to happen at night. So you have to use a coefficient without solar gain.
Paul, that's an excellent point. Makes perfect sense as the periods covered in my calcs would of course cover windy days too.
DC, we're not much for solar gain sadly as the house is mostly shaded by large trees, and is painted "Harris Cream" (yellow) to match the upper story Hardi siding. Our AC runs maybe 1-2 weeks per year due to proximity to Lake Superior less than 1 km away.