Two numbers, and their ratio is the whole answer
A cellar has a UA, which is how many BTU an hour cross the shell for each degree of difference, and a heat capacity C, which is how many BTU it takes to move everything in the room one degree. Divide C by UA and you get a time in hours. That time is the time constant, usually written tau, and it is the only number you need to answer every question on this page.
After one tau, the room has closed 63% of the gap to the surrounding temperature. After two, 86%. After three, 95%. It never quite arrives, which is why the calculator reports the hours to a temperature you name rather than a time to warm up.
| Hours elapsed | Share of the gap closed | A 55°F room in a 72°F house |
|---|---|---|
| 0.5 tau | 39% | 61.7 °F |
| 1 tau | 63% | 65.7 °F |
| 2 tau | 86% | 69.7 °F |
| 3 tau | 95% | 71.1 °F |
The bottles are the battery
The air in a 10 by 8 by 8 foot room weighs about 48 pounds and holds around 11.5 BTU per degree. Five hundred bottles hold about 870, and 300 lb of racking another 135. Together the contents are eighty-eight times the thermal mass of the air they sit in, and the time constant goes from 27 minutes empty to nearly 40 hours full. An empty cellar and a stocked one are different machines wearing the same shell.
This has a practical consequence that surprises people: adding bottles does not increase the running cost. Heat capacity does not appear anywhere in the steady-state heat gain, which is UA times the temperature difference and nothing else. Once the stock is at temperature it is free to keep there. What it buys you is time — through a power cut, through a compressor failure, through the week you are away and the unit trips.
Pull-down is slow and that is not a fault
The first cool-down of a loaded cellar is the same exponential run backwards, but the driving term is the unit capacity minus the standing gain rather than the full temperature difference. A 1,000 BTU/hr unit on a room with a UA of 25.5 can hold 72 minus 39, or about 33 degrees, so a 55 degree setpoint is comfortably inside its reach — but getting there from 72 takes about 23 hours with 500 bottles in the room, because the net cooling shrinks as the gap to the setpoint closes.
Two things follow. First, bring the room down empty if you can, then load it in batches, because each batch adds its own pull-down. Second, judge a unit by the temperature it can eventually hold, which is the surrounding temperature minus capacity over UA, not by how fast it gets there. The calculator reports both.
The door is usually the problem
A 21 square foot door at R-5 has a UA of 4.2. The remaining 427 square feet of shell at R-20 has a UA of 21.4. So the door is a sixth of the heat gain on under 5% of the area. Put a single glazed panel in that door at R-1 and its UA alone is 21 — the entire insulated shell over again, from one door. This is why cellar doors are gasketed, sweep-sealed and usually solid or double glazed, and why the calculator makes you enter the door separately instead of averaging it into the wall.
What this model does not do
It treats the room as one lump at one temperature. In reality the air responds in minutes and the wine in the middle of a rack responds over many hours, which is the entire reason thermal mass helps at all. It ignores latent heat, so a room that is also gaining or losing moisture will not behave exactly as shown. It assumes the surrounding temperature holds still during an outage. And it says nothing about whether the equipment is correctly sized, which is a load calculation — the walk-in cooler load calculator does that arithmetic for an insulated box in BTU per hour.
Related
The wine rack capacity calculator gives you the bottle count and the weight that feed the heat capacity here. The wine cellar humidity calculator handles the moisture the cooling coil pulls out. For the shell, the assembly R-value calculator gives the R that the wall actually reaches once the framing is counted, and the cellar insulation take-off lists what to buy. A comparable mass-and-loss calculation for a very different room is the greenhouse thermal mass calculator.
Questions people ask
How long will my wine cellar stay cold in a power cut?
For the default case here — a 10 by 8 by 8 room at R-20 with an R-5 door, 500 bottles and 300 lb of racking, sitting in a 72 degree house at a 55 degree setpoint — the time constant is just under 40 hours, the room is about 0.8 degrees warmer after two hours, and it takes 35 hours to reach 65. Change the bottle count and the racking to zero and the same room passes 65 in 24 minutes. Run your own numbers; the answer is extremely sensitive to how full the room is.
Does a fuller cellar cost more to run?
No. Steady-state heat gain is UA times the temperature difference, and heat capacity does not appear in it. Bottles cost you once, during the pull-down, when the unit has to take the heat out of them. After that they are free to hold and they make the room far more forgiving of an outage or a failure. The only running cost that changes is if the extra stock makes you open the door more often.
What specific heat should I use for a bottle of wine?
Around 0.6 BTU per pound per degree for the bottle as a whole. Wine is close to water at about 0.9, glass is about 0.2, and a typical 2.9 lb full bottle is roughly 1.65 lb of wine and 1.25 lb of glass, which averages near 0.6. Steel racking is about 0.11 and adds almost nothing; wood racking is 0.4 to 0.5 and does contribute if there is a lot of it.
Why is the pull-down time so long?
Because the driving force shrinks as you approach the setpoint. The unit runs against the temperature difference the whole way, so the net cooling rate at 56 degrees is much smaller than at 72. The calculation gives an exponential approach, not a straight line. A loaded room taking two or three days to come down for the first time is expected behaviour, not a fault, and it is why bringing the room down before loading it saves real time.
Can I use this for a root cellar, a curing chamber or a beer fridge?
The arithmetic is the same for any insulated box with mass in it — UA from the shell, C from the contents, tau from their ratio. What changes is that some of those rooms have significant moisture and respiration effects that this model ignores entirely, and none of them get any conditions guidance from this page. For anything storing food, the temperature and humidity you should be holding come from a tested source or a food safety authority, never from a calculator.