Calculating how much air needed to even out temps for minisplit
I’m building an 800sf 2 story house, R20 under slab, R40 walls, R60 roof, as airtight as possible, but likely just double pane windows (due to cost). 12% glazing, mostly on south side, w/ 4″ slab on first floor. Using Marc Rosenbaum’s heat loss spreadsheet, I get around 4000BTHU building load here in Portland Oregon at the 97% coldest day of 23F and a design temp of 70F, so a delta T of 47F.
I’m considering heating/cooling with ductless minisplit on open main floor, and HRV with recirculation mode for fresh air, and to potentially help even out the temps between rooms when in recirc mode. I’m wondering how I could calculate how much air would need to be moved into each room to equalize temps.
Assuming open doors would give 50-100cfm of air movement. There are 2 10×12 bedrooms next to a hallway at the top of the stairs, plus a 5×8 bathroom. Ceilings are 8′.
So bedroom 1 has 120 sf floor, and 80sf+96sf of wall shared with the warm area. It then has 96sf ext wall at R40 and lets say 16sf windows at R3. Then there is 80sf dividing wall between it and the other bedroom. Finally there is a 120sf ceiling at R60. There is 960cuft of volume.
Bedroom 2 has 120 sf floor, 96sf wall shared with other bedroom, and 96+80sf outside wall, with 16sf windows, and 120sf ceiling. There is 960cuft of volume.
Bath would have 40sf floor and ceiling, 40sf shared with warm hall, 64sf shared with bedroom, and 40+64sf ext wall. lets say a 4sf window in there too. 320cuft volume
I believe I need 30cfm for fresh air, and was thinking that could be achieved for part of the time, and then swtich to recirc mode for the rest of the time to even out temps.
I’ve read that HRVs cannot be used to mix air, but seems like one that has a recirculation mode and higher CFM should be able to do just that. I think heating would be less of an issue with the warm air rising and mixing by itself. Cooling however, seems like it would need 100% active mixing to bring any cool air upstairs.
1) is this a crazy idea?
2) if not, am I on the right track with the math?
3) if so, would an auxiliary air mixing vent system make any sense?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Forget about the volume of the rooms- the volume is irrelevant. What counts is the actual heat load of the room.
The constant you need to work with is the heat-capacity of air. To raise a cubic foot of dry air 1 degree F takes about 0.018 BTU.
You can't heat a room by replacing it's air with the same temperature you're pulling from it- you have to define the maximum temperature delta that can be tolerated in the design, and work from there.
For the time being let's assume you're OK with a doored-off room being 5F cooler than the space heated with the mini-split, and say that the heat load of that room is 500 BTU/hr.
At a delta-T of 5F with 0.018 BTU/ degree-ft^3 , for every cubic foot per hour you get:
5F x 0.018 BTU/ft^3 = 0.09 BTU
So to deliver 500 BTU/hr it takes 500 / 0.09= 5556 cubic feet per hour.
In a 60 minute hour that's a flow rate of 5556 / 60= 93 cfm.
With different heat loads and different delta-T constraints you'll come up with different numbers, but that's the basic napkin-math version, which is at LEAST as accurate as Marc' Rosenbaum's spreadsheet heat load tool.
Tom, I have a very similar situation with a house I just built. Staying clear of the number crunching for the moment, let me try address your assertion that "I think heating would be less of an issue with the warm air rising and mixing by itself." Whether it's stack effect or just plain convection, don't assume that this will help you substantially with heat distribution. In last year's cold winter, the heat from my one minisplit would not make the trip up the stairs, through doors, and into the two small bedrooms. I had to break down and install the second head upstairs (which was called for by the Manual J calcs anyway--I just wanted to test it....) You could design in some cold air returns, or go ahead and figure out a good way to run the HRV ducting. I'd like to hear how your design works out, and how closely it conforms to the calculations. I'll be trying HRV-recirculating this winter, last winter it hadn't been fully set up yet.
Dana- thanks for that. Exactly what I was looking for. A few followup questions (aren't there always?!)...
My thought was that since the air on the main floor (with the minisplit) was a different temperature than the air in the upstairs rooms, wouldn't exchanging those 2 air masses effectively change the temp in those rooms, and then the minisplit would have to work a bit harder to warm (or cool) this new colder (or warmer) air? That's not the same as replacing the air with the same temp I'm pulling from it.
Assuming I'm just confused about the above, if I need 93cfm (or whatever my actual math ends up at) to deliver 500BTUH to heat that room- then I'd need to exchange 93cfm of air with the heated space below to reach that 5F difference equilibrium?
Another question- on another thread someone suggested that I didn't understand what recirculate function does on an HRV, and that I was confusing it with bypass function- which would bypass the heat exchange core and bring unconditioned fresh air straight in. I'm assuming that I was correct in my understanding of the recirc function, but at 93cfm per room, I doubt it's cost effective to find an HRV large enough to provide that amount of recirculation in that mode.
Andy- I'm still at the design phase, and may very well end up just putting some back-up resistance heaters upstairs and calling it good, but hoping for a more elegant solution that would allow the upstairs to benefit from the cooling the minisplit can do.
It doesn't matter where that air is coming from- all that matters is that the exit air on the ventilation is 5F warmer/colder than the air it's exhausting to provide the heating/cooling.
Figuring out what that exit temperature on the recirculating ventilation is going to be has too many variable that are site & system specific, but if it's constantly recirculating it'll be something like the average temp in the house, not the temp of the warmest room. If your ventilation system only draws from the space with the mini-split the exit air it's dumping into the remote rooms will be pretty much the temp of the air in the mini-split room. If it's both supplying and drawing directly from each room it will be more of an averaging effect.
And yes, since it will be then distributing cooler air into the warmer rooms & conversely, it adds to the heating & cooling load seen by the mini-split, which is EXACTLY what you want.
It takes a pretty big delta-T or a ridiculous cfm number to provide much heating & cooling with a air recirculation system, but in low-load homes like that it's less ridiculous than in code-min homes.
The HRV might help even not in recirc mode, if you plan the inlets and outlets right. If it delivers fresh air near the minisplit head, and exhausts it from the parts of the house you need to get heat to, it will push the warmer air towards those spaces.
Also, you will get some heat flow simply through your uninsulated floors and walls, so that helps a little.
I think you could install a very simple system to boost your circulation, just a powered vent from a 1st floor room to a second floor room right above it. Maybe reverse the fan direction summer vs. winter? Consider roughing that in in case it is needed.
For air conditioning, note that a big part of what that does is remove moisture, which I'd guess is easier to even out. If the whole house is dry, a little warmer upstairs is probably not a problem.
Charlie,
You wrote, "The HRV might help even not in recirc mode, if you plan the inlets and outlets right. If it delivers fresh air near the minisplit head, and exhausts it from the parts of the house you need to get heat to, it will push the warmer air towards those spaces."
Many people have speculated that your approach is valuable, but the theory is undermined by actual calculations of the heat capacity of the air. In an article called Choosing HVAC Equipment for an Energy-Efficient Home, I discussed this issue with building scientist John Straube.
Straube said, “Ventilation air doesn’t do much to move around heat. ....Ten cfm of 72 degree air to a 65 degree bedroom won’t make any difference to the temperature in the bedroom at all. Open doors work better than HRV ducting.”
Martin, I agree...I only said "might help," not "will solve the problem". Sorry if I implied it would do more than it does. Dana's numbers show that 10 CFM would make 0.5 degree difference for his scenario. My point was more that recirc mode vs. HRV mode isn't critical. In think in practice, main advantage of recirc mode is that you could run it at higher CFM without as much energy penalty.
One other thing worth noting:
A big part of why people have success with wood stoves is that they often heat the living room to well above 72 F. If you've been sitting reading next to a wood stove, and then you go upstairs into a cold bedroom and get under a thick comforter, you likely to be fine with it being less than 65 in the bedroom. Having regions of the house at different temperatures can be seen as a feature, not a bug, as it lets occupants warm up or cool down by wandering to the appropriate spot. It might seem horribly wasteful to heat the living room to 76 F, but particularly if the heat source is central, away from outside wall, it probably actually takes less heat to do that and let the fringes of the house be cooler than 65 than to heat every last corner to 68 F.
Open doors work well. If you have to close doors for extended periods (say 8 hours), thermal mass (+ uninsulated wall transfer) might be enough to coast through that period without too much temperature change - during most weather.