Hi all... My house is more or less finished and I would like to measure BTU/FT2/HDD. I have a wattmeter measuring Wh's used by my mini-split; I have HDD's for a nearby weather station on wunderground.com; but I need to subtract some HDD's because I'm not living there and keep it a bit cool (around 60 degrees). What temp do the HDD calc's use for indoor set temperature? I realize the HDD's usually use the difference between 65 or 66 and the average daily temp. Then they probably assume that you gain a few more degrees from internal gains to get to 68 or 70...(?) So, since I don't live there, I'm also not getting benefit of internal gains (no water heater or refrigerator or other appliances or body heat etc.). I'd like to know a fair way of tweaking the HDD's to plug into the formula... Any help appreciated! Thanks, -michael

It would be more accurate if you could use a resistance heater for awhile - otherwise the COP of the HP isn't exactly known. At 60F, my guess would be to decrease a 65F base HDD figure by 5F (assuming the outdoor temp is remaining below 60F).

Hi Jonr..

I think subtracting from 65 sounds right to me, too. Maybe I was over-complicating it.

I did use the meter on a baseboard 220v heater before I started just to make sure the meter worked. But I'm not sure why I need to know the COP of the MSHP(?) It will l vary with outside temps anyway. Aren't we trying to measure the system as a whole is using?

Thanks,
-m

How will you know how many BTUs are being produced to calculate BTU/FT2/HDD? You could calculate "KWH into HP/FT2/HDD" but it would only be accurate at some outdoor temp value (because the HP gets less efficient as it gets colder outside).

Hi again..

You're right...I need to run the numbers over a whole season (or 2 or 3) before I have an accurate assessment of performance. But yes, I'm measuring the KWH's used by the MSHP and converting to BTU's......

Things look pretty good so far, but I'll have to wait for some colder temps...

Thanks,
-m

For new code-min (IRC2012 or 2015) houses base 65F is too high. Running it against base 60F HDD will likely be more accurate. Base 65F standard was closer to reality for 1950s homes with 2x4/R11 walls, R10-R20 in the attic, single-pane + storm windows, no foundation insulation, and air leakage north of 10ACH/50. B

Between electrical plug loads and mammalian occupancy most newer houses maintained between 68F-72F won't have a heat load needing to be served by the heating system until it's 60F or cooler outside, but clearly there's a range. If you're keeping it cooler than that indoors you'll have to use a correspondingly lower base number.

Most high-R houses would have to drop back to base 55F, at an indoor temp of ~70F, and many PassiveHouses balance at 50F.

For good weather history data sets at any arbitrary base temperature delivered in a .xls spreasheet format for easy analysis, try:

http://www.degreedays.net/

Be sure to select "daily" rather than the default "monthly" HDD numbers.

pbrane: would be interesting to look at day-to-day variability in your value and also compare your results to a Manual J. Perhaps compare the former to wind speed for the day.

> You're right...

I knew that if I kept posting, this would eventually happen.... :-)

Wind speed and solar gain both affect the instantaneous loads by quite a bit, but only at the extremes would it show up outside the sampling noise error in an energy use per HDD calculation.

And with a mini-split the raw COP efficiency changes pretty dramatically with both modulation level and outdoor air temperature, adding even more noise to the model. At 35F outdoor temp the efficiency at min-modulation can be 2x that of the efficiency at maximum output, using half as much input energy per BTU delivered.

I like Jonr's resistance heat for measuring. It eliminates much of that noise. Just don't know how practical it is.

Dana.. I've been trying to understand the idea of using lower base temps for tighter houses. I just can't help but think it's like moving the goalpost. I mean, I thought the BTU/FT2/HDD calc was used for comparing houses of different construction standards. For me, I was going to use it to see how I compare to passive houses etc (I've heard they're about 1 BTU/FT@/HDD, but I don't think that's an official way the passive house people use to qualify). But if I have to use fewer HDD's because of lower base temp, I'm making HDD a variable when it seems like that should stay fixed for making comparisons to other houses. So I need to be straightened out on this concept. Or maybe I should just use the "official" passive house requirement which uses annual heat load per area and skips the HDD's. But that seems almost useless to me (a lot of homes in a nice climate like parts of Hawaii? would be called passive??) Thanks for fielding my questions... -m

Jonr.. Yes it would be nice to be able to chart windspeeds against consumption etc, but I don't have access to good wind data. I've always wondered how wind affects a house...."house windchill?" But I think it just speeds up infiltration.... thanks, -m

If you use base 65F for a high-R house you would have days of negative heating energy use when outdoor temps were above it's natural balance point. It's ridiculous to use base 65F for comparing houses of dramatically different performance, but reasonable for comparing say, 1990 code-min house to a 2015 code min house.

And with a modulating heat pump of radically shifting efficiency with temp & load you can't even come up with more than 1.5 digits of precision on the amount of BTUs the thing delivered. When it's 40F outside and the thing is just idling along and it delivers 10,000 BTU compared to 20,000 BTU delivered when it's 20F outside isn't anywhere near a factor of two on the power used- it's probably more than a factor of three.

The BTU/HDD/per square foot number is close to useless even without the variable-efficiency error factor making it impossible to reliably estimate the actual BTU input.

If you record many HDD and temperature points, you could derive a HP efficiency curve - HDD vs energy use is supposed to be linearly proportional. But it isn't exactly and so I wouldn't make any bets about the accuracy.

On the other hand, real data and comparison metrics are important. As some say "You can't fix what you can't measure". And there can be rather large differences between simulations and actual.