Not your usual "I am building a wine cellar" question. Advice appreciated

Retired Professional HVAC Engineer here, who is not offering professional advice, I practiced in a warm humid environment so I never came across this type of question before.

Here’s a decent discussion:

So I would ask a local Architect, not a professor.

1 Like

As I recall, Gold says that in the summer in Western Massachusetts the ground temp six feet down runs around 65F. That was surprising to me, but explains why my mostly-subgrade NY passive seller gets up around that level in the summer notwithstanding my overinsulation of the walls and ceiling.


Here’s a link to average ground temperatures in various locations.

I have a separate pdf of variation by season but can’t seem to post that here - so instead I’ll link to my previous posting of it (I hope)

1 Like


I can’t tell what depth the data is for. And is that the average over the course of the year?

It does tend to support Greg Tatar’s theory that the only reason 55F is regarded as ideal is because that’s the natural temp of cellars in England. :grin:


I was driving myself nuts trying to figure that out. I think it’s 6’ but I couldn’t find anything to confirm that.

It has got to be, so it’s not entirely useful as seasonal variations from the mean will differ markedly depending on location.
eg Kuala Lumpar’s average is 85F, I would suspect it’s the same for all 12 months.
Adelaide Australia listed at 64F, reckon that would be plus or minus 10F
Phoenix, Arizona listed at 73F, assuming Tuscon has a not dissimilar climate then we know highs are plus 20F from teh average.

Yeah my parents basement in WI seemed to be in the 50s with no cooling.

Trying to recreate my work from a decade ago - I found the full document here:

The ground temps are 30 feet below and are constant. And shallower depths they vary by season (see my second link above or p. 14 of above doc). But that seasonal variation I think is more illustrative than anything because it really depends on the variation in air temperature during the year.

John and Andrew, Purely based on the PHX numbers, I would guess it is a 10cm depth measurement.Purely a guess, but sources I looked at referenced a 10cm depth. It seems like the standard recorded depths are 10cm and then occasionally 50cm. These temps are annual average temp. For the AZ desert the monthly numbers are widely divergent from that annual mean.

Kent: precisely. I just found the 2023 data from the U of A in Tucson and there was a 50F swing between average 10cm soil depth temps here with Jan at 48F and July at 98F.


Thank you. I will show this to my contractor and see what he thinks. Very interesting.

Andrew, Thanks for that info. Very interesting

Thanks again everyone. This is exactly why I asked for help here. The breast hand depth of knowledge here is really special.

Much appreciated.

1 Like

Most importantly, it would seem, given that your are super insulating your room, is that you have high quality split cooling systems…and your duplication of them certainly seems like your are well prepared in case of a mechanical failure.

I am surprised, however, by how often you have had to replace them in addition to repairing them at significant cost.
I had a refrigeration company install mine, and over about 30 years I have had to replace my outside compressor once about 5 years ago (about $6-7000 I believe) and just replaced my inside evaporator/blower (about $5500] last year for the first time for some leaks in the “fins”. I have them do a maintenance check annually at the cost of a few hundred dollars. I understand that your unit(s) may be working harder in Arizona than mine in Seattle, but still….

Assuming this is long term, you should factor in climate change as those long-term averages are likely not what you will be facing and extended extreme heat events will be longer and hotter going forward.

1 Like

Agreed, It’s a lot. I think it is a safe bet to assume they are working harder here than in the PNW. That said, five times between two systems is a lot. My first two were likely poorly chosen units by a recommended “wine cellar expert” who was clearly not. Another two were when the compressors died and the outdated refrigerant systems were not compatible with new ones, so a new system was needed. One just system death. I’m hoping for more longevity with the new systems.

Here are some ideas, thoughts and recommendations as to your system(s).

  1. You really want someone to properly size it. That may sound obvious but I can’t begin to tell you how many smaller systems are based on “rules of thumb” or “our experience on past projects” rather than calculations. You may get pushback, particularly if it’s a company that specializes in wine rooms.

  2. Insist on commercial refrigeration equipment, installed by a company who has a long history of installing and maintaining commercial refrigeration systems, no residential mini splits or any other type of a system. I strongly suggest a non ducted spit system with the evaporator(s) mounted on the ceiling. Gravity (non-pumped) insulated PVC drain lines which means you’ll see them running horizontally in your room until you can run them down a wall, they can be within the wall cavity or better yet furred into a corner. I would use a minimum of a 1” drain and see if they can give you a removable section so you can periodically blow them out. I have a small section of flexible clear tubing between my evaporator and the drain line (it’s also insulated - flexible elastomeric type).

  3. Make sure the condensing unit(s) capacity is derated to account for for higher than design ambient condensing temperature (unit capacity is usually stated at 95 degrees) and adjusted for altitude.

  4. The evaporator delta T should be in the range of 8 - 13 degrees, when the evaporator is selected for the derated compressor capacity, this will help maintain a higher room humidity.

  5. Personally I would keep the controls super simple, no DDC controls. Each unit (assuming redundancy) has its own commercial electric thermostat with both an adjustable set-point and adjustable dead-band. The thermostat would control a solenoid valve located at the evaporator in what’s called a pump down arrangement (pressure switch at the condensing unit). Again assuming redundancy use an electro-mechanical alternator installed out at your condensing units to switch back and forth between the two systems. Install a separate monitoring system if you want remote monitoring/alarming.

  6. Redundancy. This is often easier conceptually than it is to implement. As an example, with two ceiling mounted evaporators and one of them off the one that is off is going to block airflow from the one that is on unless you have a bottom discharge. Even then the room flow won’t be uniform. I wouldn’t sweat it (this isn’t an OR) but it’s something to think about when the evaporator locations are chosen. Also you should have each system on separate electrical breakers. If they try to sell you on DDC controls this becomes a consideration too; like a data center yore trying to avoid a single point of failure so you would need two independent systems.

This isn’t meant to be comprehensive (long winded maybe), more to point out things to be considered.



Thank you Mike. That is hugely helpful.

1 Like

I think you should just hire Mike to install this.

1 Like

Bored Retired Engineer who has been house bound this week recovering from a mild case of Covid (my first), so I’ve had some time on my hands.

1 Like

Mike’s point is what I was driving at…commercial equipment installed by a firm experienced at installing and maintaining commercial equipment is important. At least I believe that is why I have only had to replace my system once over the past roughly 20-30 years.


Thanks again everyone, and especially Mike. Much appreciated.