It got down to the design temperature of minus 3 degF (-19 C) for my town last night, so I decided to perform a test to measure the cooldown rate for the house. To perform this test, I started with the house at 67 F (19.4 C), the normal thermostat setting for the house in the day and early evening, and at 10:45 PM, I turned the thermostat down to 50 F (10 C). I set up the computer and a one-wire temperature sensor to log the indoor temperature every minute through the night as the house cooled down.

After 8 hours, the house had cooled by 8.0 degrees F (4.4 C) to 59.0 F (15 C). Meanwhile the outdoor temperature had increased from -3 F (-19.3 C) to +3 F (-16.1 C) by daybreak. So the average differential temperature over 8 hours was 63 F (35 C). Maybe we could call it 1 degF (0.56 C) per hour for a differential of 63 F (35 C). The house has some solar tempering, but during the day before the cooldown test, there was little sun, so minimal solar thermal storage.

The cooldown rate should be a function of the thermal inertial of the house inside the insulated envelope, the insulation level, and the infiltration rate. The house tested is single-story, wood-frame, with conditioned crawl space, R-19 crawlspace walls (mostly below ground), ~R-29 walls, ~R-4 to R-5 windows including the cellular shades, and R-60 ceiling. Floors are mostly hardwood with some tile, but no "extra" thermal inertia features except for 8" (203 mm) of gravel at the bottom of the crawl space, which might be considered a bit indirect. There is R-5 insulation under the gravel, except near the edges where it is very roughly R-29. Infiltration was measured at 2.45 ACH at 50 Pa. (See website in signature block for further details.)

It would be interesting to see how houses with high thermal inertia (high mass) might compare in terms of cooldown rate to this wood-frame, standard production house. I also think that the cooldown rate might be helpful in verifying thermal models of houses. I hope to run the test again after a clear, sunny day to see if that impacts the results.

Would also be interesting to measure the heat needed to maintain temperature under similar conditions and then put a number on the thermal mass (since the two other factors are now pretty much known, especially if you measure pressure differential).

I've noticed that wind seems to be a bigger than expected factor, too.

Posted By jonr on 10 Dec 2012 06:02 PM
Would also be interesting to measure the heat needed to maintain temperature under similar conditions and then put a number on the thermal mass (since the two other factors are now pretty much known, especially if you measure pressure differential).

I agree, but I do not know of an easy way to measure the heat needed to maintain a constant temperature near design conditions. The natural-gas fired, hot air furnace that I have is modulating, so it would probably take a measurement of gas flow to determine the heating rate, which is not very feasible. In fact, you can see it in the temperature profile when I turned the thermostat back up in the morning. The temperature rises more rapidly at first when the "error" between desired and actual was 8 degF, and then the rise rate drops as the set point is approached. Not real straightforward.

I could set up an electric heater that would be easy to measure energy input to, but hard to circulate the heat. I suppose that I could try to use the furnace fan to recirculate the air with the thermostat turned down to keep the furnace from firing.

Posted By ICFHybrid on 10 Dec 2012 06:09 PM
I've noticed that wind seems to be a bigger than expected factor, too.

The heating design temperature is approached on the day after a front blows through. So high winds on day/night #1, but the temperature drops lower on day/night #2 when the winds drop down. In fact, the lowest temperatures (i.e., heating design temperature) usually occur only when the winds are light. The winds were light during my test.

I agree, electric heaters + circulation fan may be the only good way to continuously measure short term heat load.

Can you read your gas meter??

Posted By woodgeek68 on 11 Dec 2012 05:16 AM
Can you read your gas meter??

Yes and no. Perhaps you could help. I have an all-analog meter such as shown at http://en.wikipedia.org/wiki/Gas_meter So there are the four top dials (labeled 1,000,000, 100,000, 10,000, and 1,000) used to obtain the gas usage that I am billed for, and I know how to read and understand them. Using those four dials is not nearly precise enough for a good measurement. In December, I typically use about 33 CCF for the month, or about 1 CCF per 24-hour day. Therefore, over the 8-hour measurement period, I would read 0, 1, or maybe even 2 CCF, not precise enough to be usable.

That brings us to the two lower dials that I do not know how to read, but that may provide the precision needed. One is labeled "1/2 foot" and the other "2 feet." Do you know if that means each division of the "2 feet" dial represents 2 cubic feet, or if one revolution is 2 cu. ft.? Since there are 10 divisions, if it is 2 cu.ft. per division, that would imply one revolution would be 20 cu. ft. That would be 1/5 of a single unit on the "1,000" dial. The 1/5 of a single unit would be 1/50 of a revolution, and that may or may not be possible for me to read. I just about need a vernier on that tiny scale.

Suggestions?

According to the site http://www.correllglassstudio.com/Calculating%20fuel%20consumption.pdf, the "2 feet" dial goes around one complete revolution for 2 cubic feet of gas. Therefore, it cannot be used with the other less sensitive dials to get more precision than the "standard" measurement using the four, lower resolution dials. The 1/2 foot and 2 feet dials are only used to measure rates real time, and to check for leaks. Since a modulating natural-gas furnace ramps the firing rate up and down, it would not be possible to measure the "average" fueling rate over a period of time.

Looks like you would have to install something like this .

Good find. That would be a great solution. Do you think the gas company would mind reading my usage in cubic meters rather than cubic feet? I could give them the conversion factor!

Or maybe they would just under-charge you! More seriously, I'd splice it in just before the furnace.

Good point. I had been thinking in terms of replacing my existing meter, and I knew that would not work. I'll check into it.