I'm building a passive cooling wine cellar in the basement of my Ottawa home. A passive cooling wine cellar is really just a super insulated closet relying upon ground temperature to maintain constant year round temperature. What is the best practice for vapour barrier(s) if using two layers of batt insulation between wood 2x4 against a pour concrete wall? I've read conflicting advice for/against vapour barriers/moisture barriers on the interior wall and/or against the concrete wall. There seems to be a conflict between keeping everything sealed and dry as opposed to allowing air/water movement in order not to trap moisture. I'm thinking about wrapping both sides of the double stud wall in 6m poly and leaving a 1" gap between the stud wall and foundation wall. Are there any issues with this approach? The house is 50+ years old and the basement has always been bone dry. There's ample space so wall thickness is not a concern.

Poly is far too vapor retardent to be used on both sides of a studwall- it's a moisture trap. Use unfaced 25-35mm of XPS rigid-foam sheathing between the studwall and the concrete (make that 50mm if any portion of the wall is above-grade, in Ottawa, or go with 100mm of unfaced EPS bead-board and a single layer of batting), and use UN-faced batts and NO vapor barrier (just an air barrier such as wallboard) on the conditioned-space side of the walls.

XPS slows down ground moisture migration into the studwall but still allows the foundation to dry toward the interior- as long as there is a vapor-path to the conditioned spaces. Poly vapor retarders on foundations as called out in Canadian building codes are a mistake IMHO, but the modeling of what it takes to keep studwalls mold-free isn't as simple as a blanket prescription. See:

http://www.buildingscience.com/documents/information-sheets/5-thermal-control/basement-insulation/files/bscinfo_511_basement_insulation.pdf

http://www.buildingscience.com/documents/information-sheets/5-thermal-control/crawlspace-insulation/files/bscinfo_512_crawlspace_edit.pdf

Blocking the ground moisture completely with poly drives the moisture higher in the foundation, risking the foundation sill & rim joists. In winter, above the frost line the risk of moisture accumulation from vapor-migration from conditioned-space air goes way up, but putting a vapor retarder on the warm side of the stackup traps summertime ground moisture in the studs. To stay mold free the wood needs to stay above the dew point of the room air, and the further toward the exterior/cold that wood extends, the higher the risk. By putting semi-vapor-retardent foam insulation on the cool side of the structure it raises the average temp of the wood limiting the number of winter hours it's below the dew point of the room air while limiting the rate of ground moisure getting in. How much foam it takes to get there is both climate and total-R value dependent. In Ottawa you'll generally be fine with 50% or more of the R value in exterior foam in above-the frost-line applications. The place in the assembly that hits the dew point of the interior air stays pretty much in the foam, where the air never reaches.

100mm of low density Type-I EPS (often sold as insulation for flat roofs) gives you ~R15 at 20C, but delivers closer to R20 at -20C, and would be only slightly more money than 50mm of XPS. Combining that with unfaced R12-R15 rock wool or fiberglass batts would give you a very decent average R and very low risk of mold & rot, as long as you skipped the poly vapor barriers. Going much more than 100mm of EPS or 50mm of XPS raises the risk to the foundation sill & rim joist.

Dana:

Your reply is very much appreciated.  Your explanation put the other information into context and allows me to make an reasoned and informed decision.

I was worried that the high relative humidity inside the cellar could result in moisture and rot effecting the stud wall.  As I understand your post, by moving the dew point inwards and letting moisture escape it shouldn't be an issue.  I'll probably go with the XPS.  It's now off to the building supply store to cost compare the options.

Thanks

 

Mind you, 2" of XPS is only ~ R10, and doesn't have as-favorable a temperature coefficient to raise the R of the foam with falling temps, but it should still get you there with a single layer of ~ R13 batting.

Box-stores don't generally carry unfaced Type-I EPS roofing or sheathing- you'd have to look at distributors who deal with contractors in the commercial building space, but it's pretty available and pretty cheap. Most accept cash or credit cards and will sell in small lots if you know EXACTLY what you're looking for. But they aren't into hand-holding or explaining things or recommending alternatives that they might have in stock. This is the kind of stuff you'd be looking at:

http://www.atlaseps.com/pages/TechData-geo-109.asp

http://www.achfoam.com/ACH/media/docs/Foam-Control/Foam-Control%20EPS%20Exterior%20Sheathing-ACH.pdf

The fact that EPS increases in R value significantly with lower temps is useful here, since it changes the ratio of F-foam/R-fiber in the right direction. At bigger delta-Ts rock wool and fiberglass batts fall in R value due to higher convection rates within the insulation. So when it's -20C on the cold side of your stackup an R13 batt might be only performing to ~R11, but if you started with R15 of Type-I EPS the total R would remain about the same, but it's now R17 Foam/R11 fiber, which keeps depth where the ~ 2-6C dew-point of the conditioned space air occurs firmly in the foam. The dew-point of 20C/30% air is ~ 2C, the dew point of 20C/40% air is ~6C. You normally want to keep it at 30%+ indoors for optimum comfort & health, but when it's -20C out it can be drier than indoors if you have a high enough ventilation rate. Average monthly temps for Ottawa in Jan/Feb is more like -10C, but you need some margin for the extended cold snaps where(as you undoubtedly know) it can stay consistently much colder for days on end.

Assuming most of it is below grade you can probably get away with R5 in foam due to the moderating effects of the soil, which isn't as cold as the daily lows by a long shot. But with substantial area above-grade you want to keep the foam/fiber interface winter average to be well above 5C, which means doubling the R-value of the foam at a minimum, but tripling it would be better. The more hours the condensation zone is in the fiber layer, the more moisture will accumulate in the stud-edges. As long as the moisture content of the woody-edge stays below ~28-30% the mold potential is low, but it won't/can't dry quickly in winter- keeping condensing condition duration to a minimum is the right way to go.

agree with GROUND SOURCE advice! 100% only addition would be IF you are considering addition of perimeter foundation insulation and / or fluid applied membrane for EXTERIOR foundation waterproofing.

Don't intend to hijack, but it appears that the OP has his solution and while everyone has their "wine cellar thinking caps" on, let me ask about this;

A 150 sf wine cellar in the corner of a 6" ICF below-grade basement in the Pacific NW where ground temps don't vary too much and where there isn't much if any frost depth.. The wine cellar shares two full ICF walls with hydronic slab heated basement space and two walls with ICF exterior foundation. The difference is that the wine cellar has no under-slab tubing or insulation. Also, 7" of Insuldeck over the top and 3" of heated slab on top of that.

Anyone else done this? Will the wine cellar be a cool box with respect to the rest of the (heated) basement?

Do I need to consider removing the foam from the exterior walls in order to open it up to some heat loss or will I get enough loss through the floor? Keep in mind that the ICF interior walls have footings as well, which reduces the area of the floor directly exposed to grade.

What do I need to do for the air in this space? Will it build up moisture as a result of cooling warmer house air that gets in?

It isn't built yet, so here is the opportunity to get it right.

My first reference on building my wine cellar was How and Why to Build a Wine Cellar by Richard M. Gold. He recommends insulating the entire exterior walls rather than leaving the bottom portion uninsulated. While this reduces the potential heat loss, it results in a more constant year round temperature. This approach worked for me in Kingston Ontario were the two exterior walls of my wine cellar were SIP with 6inches of polysterene and the interior walls 2x6 studs with rock wool batts. The interior temperature was 14C plus or minus 2 with seasonal fluctuations. Maintaining constant temperatures may be a good trade off for cooler temperatures.

I used 6 mil poly on the to make a complete vapour barrier around the cellar. I believe that was too tight as relative humidity sometimes reached 80%. This time I`ll use 1inch EPS on the interior of the ceiling. Hopefully that will keep the humidity at 70% or less. It`s easier to add moisture to the air than remove it.

Not sure how this will apply to the warmer and weter Pacific NW, but it worked well in southern Ontario.