« Peak export point | Main | Deciduous tree shading on my PV array »



Feed You can follow this conversation by subscribing to the comment feed for this post.

Ed Davies

I wonder if using HDD65 is realistic for a well insulated house. When the outside temperature is 65 °F you'll probably get 70 °F indoors even without any use of the heat pump, just heat leakage from the domestic hot water tank, sunlight through the windows, light bulbs, human and dog heat, etc.

If you have the raw data then it might be worth working out your kWh/HDD for, say, each month. Using those figures play with the HDD base to determine the best fit: the base which gives the most consistent kWh/HDD.

Still, 280 Wh/HDD Fahrenheit is 21 W/K (watts per degree Kelvin, Centigrade or Celsius) which is pretty good. Even with a COP of 3 on the heat pump that's 63 W/K - I'm estimating 45 W/K for a house with half the floor area. Still, a smaller house will have proportionally a larger surface area and so, per square foot or metre of floor area, a larger loss so your numbers validate that I've got reasonably realistic numbers. Thanks for that.

Marc Rosenbaum

Hello Ed, thanks for your comment. I'm not sure what the balance point is of this house, it's likely 60F or less I'm guessing. And I'm pretty sure that the building heat loss coefficient is more than 63W/K - it's hard to go from consumption directly to heat loss coefficient. I do know that through the coldest weather last winter - lowest outdoor temp was 7F - the HP used about 0.9 kW. This indicates either a better COP or a lower heat loss coefficient than I think - a lot closer to the number you calculate.


Mini-split COP is extremely variable with outdoor temperature and compressor & fan speeds- at 35F outdoor temp the COP of a 12RLS at lowest comp hits around 4.5, but at highest speed is barely half that. (Ecotope has been studying this in some depth for the BPA- the condensed version relevant to this discussion can be found at www.nrel.gov/docs/fy11osti/52175.pdf - see the COP graphic on p18.) It's tough to fully model the heat loss of your house without the room by room temp and solar data, but it's conceivable that as-operated your average COP could easily exceed 3.

This makes a case for oversizing mini-splits by 1.5-2x for the load. As fully modulating systems with a significant turn-down ratio, they would not short-cycle into low efficiency or an early grave at 2x oversizing, but would bump the average operating speeds down for higher average COP. If most track similarly to the 12RLS (and they should) it would yield a huge net jump in average operating efficiency- a half-COP point or more.

IQF freezer

Excellent Blog! I really admire your thinking and the way you have put these information in this post. Thanks for sharing an informative post.

Marc Rosenbaum

Dana, thank you for your comment. I completely agree with your points and have been following the Ecotope work. The one place we're not oversizing so much is with ducted units because it drives duct size up.

Jijm Shipsky

I don't understand the Fujitsu spec, which lists an outdoor design temp of 47 F DB. Does this mean the unit produces no heat below 47 F outside temp? It sound like yours produced heat even at 7 F outdoor temp.

Marc Rosenbaum

Heat pumps are rated in heating nominally at 47F. Fujitsu's sizing is great for predominantly heating climates. The 12RLS (now the 12RLS2) is a nominal one ton unit (12,000 BTU/hour is a ton), and at standard rating conditions for cooling it is one ton. It's heating output is much higher - at 47F it is rated at 22,000 BTU/hour, and at -5F it is rated at 15,000 BTU/hour. So it's really sized well for small efficient houses in a heating climate, where peak heating exceeds peak cooling load.

Gerry Owen

I live on a Gulf Island (Gabriola) on the coast of British Columbia. This is a predominantly heating climate with moderated extremes so it never gets too cold (20 F would be a rare cold day). In response to government sponsored energy grants for high efficiency systems a group on this island (Sustainable Gabriola) has arrange to bring in and install close to 100 of the 12RLS and the new 12RLS2 in this small community. This provides an opportunity for checking reliability and customer satisfaction over a wide range of conditions. The experience with these units has been remarkably good. The few complaints have been traced to non-equipment related issues (like one resident putting a screw through a between wall refrigerant line and loosing charge). One woman who weaves raw wool complaint=ed that she was getting very little heat from her system. She had never cleaned her filters and they were caked with a half inch of wool fibre!)

I am heating a 1200 square foot 2 bedroom '60's single story home with one unit and despite my yet to be replaced single pane windows it keeps the place comfortable (68 F). After the window replacement this summer I expect next winter will be even better. This seems to be a very efficient very reliable product.


Thanks for your contribution, Gerry, good to hear others have had success. I find that even without raw wool around it makes sense to clean the filters 2-3 times a heating season.

The comments to this entry are closed.