How I Built My Second Wine Cellar

My first wine cellar was in a suburb of Hartford, Connecticut. Then we moved to Indianapolis. The average annual temperature difference was marginal (Indy is 2 F warmer on average than Hartford, but has 12% less precipitation and half the snow). I needed a new wine cellar, and the lessons I learned in Hartford were directly applicable because of the temperature similarities. The big lessons learned regarded cooling and humidity. 

My experience in Hartford persuaded me that cooling units are not simply a waste of money, but they actually reduce the quality of storage because they cause the temperature to fluctuate too much. When I had the cooling units active I would get low temperatures as low as 53 degrees and as high as 72 F! Thus, cooling units are a problem to begin with, because they turn on and off (technology might develop, I suppose, where they stay continuously on and vary the amount of cooling they provide until they get it down to the right amount and then stay there as the newer luxury car heaters do; I will not hold my breath for this). On and off cycles imply that there is some level of temperature fluctuation that is unavoidable. But it seems that whatever fluctuation the units are programmed for, they face an additional problem in a room-size wine cellar. The temperature sensors of the cooling are hampered by the lack of air mobility. The units and the area around them gets cool, but that is a local phenomenon. A range of 53 to 72 F within weeks is disastrous. Let alone that 53 F means refrigeration, and the wine does not develop in the bottle (OK, I exaggerate, it develops, but slowly enough that the only ones to enjoy your wine will be your great-grandchildren), the real problem is contraction and expansion. The goal in the wine-cellar is not so much a low temperature as a steady one. Varying temperatures cause the wine in the bottle to expand and contract pushing old air or wine out of the cork and pulling in new fresh oxygen eager to oxidize my precious wines. So I turned the darned cooling units off, and temperature became a stable 66 to 72 year-round in my nicely insulated cellar. This was excellent and the insulation was not even R-30 as it should. Probably around R-20.

Temperature fluctuation having been solved, the remaining problem was humidity, or actually lack thereof. As you have learned from Gold's book (do not attempt to build a cellar without it; check it out at Winter cold means dry air. Dry air means dry corks, which eventually come to differ little from sponges. Sure the situation did get better once I dumped those pesky cooling units that dried the air, but I was still falling below 60% in the winter. I had initially installed a Hermidifier humidifier. This looked promising because it had a wall sensor/switch and was piped into the water system. It worked fine for a couple of years but then it got so much gunk that it broke. In other words, the maintenance-free humidifier had to have been cleaned regularly. All this despite that the water line into the humidifier had a filter.

Designing the new wine cellar produced two pleasant surprises, one on insulation and one on humidification.

I found an insulation contractor who can spray some foam polystyrene inside the walls which produces insulation of an R-30 value with only 4.5 inches of material. This is spectacular. It means I could get my R-30 wine cellar without losing wall space to inefficient fiberglass insulation. I should compare this method's claims with styrofoam, but I am fairly confident that even if styrofoam is equivalent in theory (it is, after all, extruded polystyrene) it would lose in practice because it does not form as tight a seal around the joists. After an entire Summer I can report that the temperature was at 66.5 F with one or two lapses to 67.5 in hot spells. Now that's extraordinary. Wait for the winter results. I would have been happy with a 12-14 F temperature variation, but now I am shooting for a variation of under 10.

The humidifying technology in forced-hot air systems has developed in the last few years, so that I now had available a drip screen system. The old forced hot air humidification technology involved a rotating drum, which was not so effective. Our old house routinely fell under 25% humidity when the temperatures outside fell below 20 F, while this new one stays consistently above 35%. So I used this system. It is linked to the water system, like the other, and it includes a fan to blow air through the screen. Of course, you need a remote thermometer/hygrometer to monitor the conditions from the comfort of your living spaces. There is one at Oregon Scientific for under $60, I think.

Anyhow, on to the pictures. Here is the empty space, the lines on the floor mark what were expected to be the inside edges of the walls.

The plumbing devices you see in the corner will be moved. Then we frame, and lose a few inches.

I hope all this makes some sense. Notice how the ceiling has gotten a new layer of framing bringing it down about 5 inches. Its purpose is to accommodate the ceiling insulation.

We can turn next to the pictures of the process of applying the new insulation. Here is the pillar seen in the picture immediately above and to the right. The outside wall has been drywalled and then the spray was applied: Note how this polysterene sticks onto the beams and the pillar without any gaps. It does not go on perfectly smooth. The thickest areas have about 5 1/2 inches of the stuff, the thinnest I suppose may come down to 4. With an R-value of about 7 per inch this takes me comfortably above R-30 on average.

Then drywall is applied on the inside walls. The drywall itself will add a little insulation. Then will start coming the racks. I had gone with premium redwood racks in the previous cellar. This time I'll go for pine resting on pressure treated lumber with a redwood kickboard or moulding at the bottom.

Redwood is desirable because of its resistance to rot. If we manage to keep the cellar at high humidity, one worries about condensation. According to Dr. Gold, condensation occurs only at well over 80%. It is most likely at the bottlom racks because the lower temperature there makes for less air capacity to carry humidity. 

I skipped the redwood racks because I never had a condensation problem in the previous cellar. Going with pine saves 30 to 40% on the cost of the racks. 

I keep, however, the resistance to rot at the floor. First, that is always a good idea in a basement. Moreover, when humidity used to drop too low I often would go in with a bucket and dump water on the floor. I hope I will not have to do this often here, but just in case I do, it is better that I avoid exposing sensitive pine to this treatment. Let us also see the finished product.

The above picture is from outside the door, which is marked also above in the second picture of this page. Notice that the ceiling lights are fluorescent so as not to increase the temperature if forgotten lit.

This view is from the far left corner of the previous picture (where the unopened cartons are. The peculiar device in the foreground is the humidifier. Finally, here is a fisheye view. I mostly buy cases and I like the ease of storing them unopened (along the wall on the right) or, opened in diamond bins (on the left). I have relatively little storage for individual bottles (far wall) and I do not regret it. Individual bottle storage hides the labels and is limited to regular size bottles. Magnums and half bottles do not fit, nor do some Champagne bottles. They all fit in diamond bins, though! 

Observe the light-bulbs. They are the spiral fluorescent kind, sold for their long life and low consumption. In the wine cellar, they offer a different advantage, that is much more important. They do not exude the heat of incandescent bulbs. Sooner or later, someone will forget to turn off the lights. Our builder had left incandescent bulbs in the cellar, and they were forgotten on for a few days. The temperature shot up to 78 degrees! Fortunately, that happened before the wine arrived. 

On the floor you might see a cream rug remnant. I lay a couple of those down in the winter in order to keep the temperature from dropping below 60 degrees. Three years later, my annual temperature variation is less than ten degrees, from lows of about 60 to highs of about 68 degrees. The humidity has never been below 60% and hovers at about 75%. 

Friends and visitors have asked for more details about the troubles I have had with cooling units. In case you are interested, here are some exchanges:

> Hi Nicholas;

> Just out of curiosity, why did you mention that cooling the cellar would be counterproductive?

> John


I dislike cooling units because they dry the air. In my Hartford wine-cellar, I started out with cooling units. As a result, our humidifier was running practically constantly and could not bring the humidity up to 60%. Moreover, the thermostats were very inconsistent, with a result that the min-max thermometers showed vast and clearly unacceptable temperature variations, from under 45 to over 70, if memory serves. Granted, the liquid in the bottles would experience a much smaller variation, but relying on an unmeasurable is silly. In-bottle temperature gauges do exist! Using them to trigger a thermostat ensures an even greater fluctuation of temperatures. The liquid changes temperature more slowly than the surrounding air. The result should be overshooting: by the time the bottle gets down to 55 degrees, the room may already be down to 35 degrees. Even if the thermostat turns off the cooling, the wine will keep getting cooler for a while.

As soon as I took out the cooling units, all problems disappeared. Temperature stayed steady and humidification was possible and consistent. The drawback is that the two locations in the wall where the units had been had cutouts and less insulation. Also, they had no wine-racks!

Once the cooling units were gone, neither my Hartford not my Indy cellar (which never had cooling) have more than 10 degrees of temperature variation over and entire year. Winter dips to 60 and Summer peaks at 68. Contrary to the lore, the primary role of the cellar is not cooling. It seems that if temperature can be kept under 68 or so, there is no damage. The principal role of the cellar is to provide constant temperature. The main concern is to avoid the wine from expanding and contracting in the bottle and pulling air through the corks. As an added bonus of the naturally-cooled cellar, you get a slightly faster evolution of the wines. Cooling them to 55 is tantamount to refrigeration, which means that they mature extremely slowly. Any infinitesimal gains in terms of taste in 40 years should be cancelled out but the extra twenty years of having to wait before enjoying the mature wine.

The lesson is that cooling units should only be installed if a collector finds that they are absolutely necessary despite the proper build of the cellar. As the cellar is influenced greatly by the temperature in the house, I would first advise lowering your AC thermostat in the house, and only if that does not work to put cooling units. My experience was a nightmare despite that I went with high-end units. The problem is not cooling capacity; the problem is thermostat consistency and narrow setting (to turn on at 56 degrees and off at 54; no cooling-unit information sheet discloses their on-off range). Moreover, the lack of air circulation in the cellar means that half the cellar might be at 45, but the cooling unit would still be running because its sensor is located at a pocket of 55 degree air. I think much of my problems was due to this, but I see no easy way to overcome it. Installing a fan to circulate air would produce heat, which would be counterproductive.

Cooling units in the northern states are manifestly unnecessary; if you put them in, you will just have to remove them later, so spare yourself the trouble and disfigurement of the cellar and the walls.

Cooling units in southern states may be necessary. Nevertheless, I would strongly suggest trying a year without them (and with strong house AC). If you get more than 17 degree variation in temperature--which I doubt--or an average temperature of over 67 degrees, you can resign yourself to them. If at all possible, avoid them. As your cellar fills up, your temperature variation will be reduced, because the volume of liquid in it will provide thermal inertia. An other idea is to fill the cellar with lots of water bottles; in the summer some of them can take a brief trip to the refrigerator or freezer so as to provide some additional cooling.



Back to my Other information page

Back to my Home page