I have a garage/gym that I’d like to condition with a minisplit. Since I only use the gym 1 hour a day, I have the idea of cooling it on-demand instead of keeping it at temperature all the time. At the extreme, this could mean lowering the temperature from 90 to 64 degrees. Or in the winter, raising it from 50 to 62. I’d like this cooling to occur quickly so I can use the gym on-demand. Looking at the single-zone mini splits Mitsubishi sells, they range all the way from 9k to 42k BTU.
My questions:
- Will this work? Is there some limit where a larger machine can’t cool down the air any faster? Some pitfall I am not thinking of?
- How to compute the temperature drop time for a unit with x BTUs? Most calculations I find are about sizing an AC unit to maintain a temperature. I didn’t see anything about estimating how long it will take to reach a temperature. I want to make sure I understand the performance I’m going to get when I decide which machine to buy.
_
Attempt to answer question #1 about pitfalls
When I looked up information about oversized AC units, I saw two issues mentioned:
- Short-cycling could lower the life of the unit. I am guessing this would be less important when the unit is only used an hour a day.
- Less dehumidification would occur with a shorter cooling cycle, so the air could be damp. The units have a dry mode, which I could use to mitigate this issue.
I know about the cost, physical size, and beefier circuit required by larger machines.
—
Attempt to answer question #2 about computing temperature drop time
This is the formula I came to:
Edit: Don’t use this formula! It will make you buy an undersized unit!
( (Specific heat of air isobaric) * (density of air) * (volume of room) * (temperature drop) ) / (BTU/h of air conditioner) = time to reach temperature
Dropping values into Wolfram Alpha [1], I got:
( (1.006 kJ/(kg*K)) * (1.204 kg/m3) * (12 ft * 8 ft * 20 ft) * (26 degrees f) ) / (18000 btu/h) = 3 minutes
This formula ignores factors like solar heat gain or insulation, but my intuition says this may not affect the result much, because the temperature in that room does not rise meaningfully over 3 minutes.
If this is correct, the 18k BTU unit would be do a great job. However, I am not confident that I’m doing this calculation correctly. The result almost seems too good to be true - from my admittedly foggy memories with window/portable AC units, they did not cool down a hot bedroom room in a minute like this would seem to imply.
[1] Link. I clicked for Wolfram Alpha to interpret temperature as a temperature difference instead of an absolute temperature. The results seemed wrong if I left it as absolute temperature - it would give me a longer duration that was not sensitive to the temperature drop (56 minutes for 26 degree drop vs. 53 minutes for 1 degree drop).