Basement Insulation

[K2]

Hey everyone, you were a big help during my 1st floor renovation, I even had Terry's crew come install the Aquia 2 dual flush with soft close seat in the new bathroom, it's a great fixture!

I took a year off and now i'm hitting the basement. I read the Building Science articles and have read many posts about insulating basements. I guess my situation is a bit different than most of the articles I've read so i'm trying to get some thoughts on this:

I have a semi-daylight basement with a walk out. The foundation walls are about 5.5' high on the front of the house and the yard slopes down towards the back, so the foundation walls drop about 2' fifteen feet towards the back. This creates a bunch of stud bays on top of that foundation roughly 3-5 feet depending on front or back. Also, the framing is 2x4 but the concrete foundation walls are about 6-8 inches thick, so there is a 2-3" gap to make the framing plumb with the concrete walls.

All the articles talk about pushing XPS up against the foundation and framing in front of it. However If I was to do that I would end up with XPS adhered to the foundation for about 5.5', then a 2-3" gap between the rigid foam board and the stud bays all the way up to the joist bays. I have most of the stud bays filled with insulation already, but i'm wondering if anyone has run into this, or has thoughts on IF/how i should fill that space between the Rigid board and stud bays. Seems like I should double up on the XPS...

Thanks for taking the time to read!

k2

[Hotbacon]

I have a very similar situation. I left the batt insulation in the exterior wall stud bays (on top of the concret wall), and put XPS on the lower concrete wall. I then framed out my interior 2x4 wall in front of that. The gap between the interior wall and the concrete + XPS is at most 1", and the gap between the interior wall and the exterior framed wall is ~3". I will soon be putting batt insulation in the interior 2x4 walls before drywalling. I'm told that the 2 layers of batt insulation is not an issue since the kraft paper is not a vapor barrier, but rather a vapor retarder.

I played around with building some type of shelf above the 2 foot concrete wall section, but it ended up just being easier to frame the full interior wall.

[K2]

Thanks for the reply, I was toying with that idea, but I saw a few videos mention the XPS AS the vapor barrier, and taping all the seams for full coverage.

I like your idea, i'm just concerned I would then have to vapor barrier the entire wall after insulaton (I've used Ultratouch cotton batts throughout our house, which don't have the kraft paper facing...and this is what is currently in my exterior walls.)

Is there any harm in running that XPS straight up to my joist bays, and leaving about a 3" space between the back of the XPS and the exterior wall insulation? I suppose I could also chop another block of XPS and slide it behind the interior facing XPS to decrease the dead air space, but then my budget goes through the roof.

[Dana]

XPS @1" thickness is not a vapor barrier. Taping the seams doesn't make it a vapor barrier either, but it does make it into an AIR barrier, which stops the all-important convective transfer of moisture.

When we're talking vapor barriers/retarders, we're only talking about WATER vapor, and how easily it diffuses through the material when there are differences in VAPOR pressure from one side to the other. When talking about AIR barriers/retarders we're talking about how much air can move when there is a difference in AIR pressure from one side to the other.

They're not the same thing. A sheet of foil or poly in a wall that has large slices in it is still a very powerful vapor barrier, but a lousy air barrier.

Both are important for controlling moisture transfer in building assemblies, but air leaks are by far the greater cause of moisture problems in building assemblies.

At 1" the permeance of most XPS sheathing runs ~1.2 perms, which is considered only "semi-permeable". At 2 inches it's ~0.6 perms, and semi-IMpermable, but still more permeable than an asphalt loaded kraft facer (~0.4 perms).

In order to keep groundwater from wicking up to the foundation sill and rotting it out over time, you need the foundation insulation to be at least somewhat permeable- over half a perm is good, over 1 perm is better. That means no- more than 2" of XPS between the concrete & studwall, but importantly NO KRAFT FACERS OR ANY OTHER VAPOR RETARDER in the stud insulation- only unfaced fiber (batts or blown) or semi permeable half-pound foam (up to ~8"). The concrete must be allowed to dry toward the interior below grade. This means no facers, no poly no vinyl or foil wallpapers, only standard latex paints (2-5 perms) for an interior finish.

Going tighter than that on the below grade portion puts the foundation sill at risk unless you have a good capillary break between the concrete & sill such as metal flashing, membranes, or at the VERY least a foamy sill gasket. (Be sure to put a capillary break between the new studwall plate and the slab too, eh?)

Where there's a pony wall with a step in at the sill or concrete, cobble-in rigid foam to even up the sill to the concrete so that you can fill the gap between the new studwall and the original studwall with fiber insulation without putting the insulation in contact with he concrete. If you leave a gap in there you would have two issues:

1: Infiltration air would be able to pass through that gap unimpeded (this is called a "thermal bypass").

2: Air would be able convect between the batting in the new studwall and the cavity , lowering it's effective R value by at least 1/3 (more, as the temperature difference across it increases. Low & mid-density fiber insulation only achieves it's rated R if it has full air barriers on BOTH sides.

Even small gaps count- don't just jam rigid foam in there reducing the gap to an inch- ELIMINATE it by compressing tightly fitted batting in there. The gap between the foundation-top pony wall is most likely never going to come out at a standard batt depth, but it's just fine to compress UNFACED R11 or R13 batts in into a 2-3" gap. It's R per inch will rise when you do that but it'll be fewer inches, for a lower overall R value. (An R11 batt compressed to 2.5" is about R9. An R19 batt designed for 2x6 framing is R13 when jammed into 3.5" 2x4 stud bay, and go figure- it weighs the same as an R13 batt.) All batt installations need to be nearly-perfect, with no compressions or gaps, or it will not live up to rating. (Batts routinely underperform blown insulation at identical ASTM C 518 rated R values due to installation errors/defects. Perfection is difficult, but do your best.)

BTW: Below grade it's safer to use rock wool or fiberglass rather than cotton batting. Cotton is hygroscopic, and could absorb quite a bit of moisture before you detected it, should your foundation ever crack & leak, or when the once every 20 years rainstorm or hot water heater causes minor flooding. Rock wool and fiberglass would dry quickly and wouldn't wick moisture up the wall. Cotton & cellulose would. (I'm a big fan of cellulose in other areas though.)

The amount of foam you need on the above-grade portion of the foundation to avoid wintertime condensation problems inside the stud bay will vary by climate- it's the ratio of foam R to total R, and at what depth in the insulation the dew point of the interior conditioned space occurs. As a practical matter, as long as the interior surface of the foam at the average/mean winter temperature is above the dew point, there is almost no risk. In Seattle it takes almost no foam to achieve that end, since by convention (based on lots of field data) the dew point of winter interior air is rarely above 40F, and the mean January temp in Seattle is about 40F (on the EXTERIOR of the wall.)

If using XPS, a half inch is enough to be protective, but it's not much R, and it's cost effective from a utility savings point of view to go a bit fatter to achieve a higher total "whole-wall" R. The value-proposition would be to use 1-2" of unfaced EPS (beadboard, like a cheap cooler or coffee cup) for R4-R8. Adding that to a 16" o.c. studwall with batts would come to ~R15 (1" EPS) to R19 (2" EPS) after the thermal bridging of the studs & plates are factored in. Considering that 2x6 framing with a 20% framing fraction (typical for simple-framed walls with a few windows & doors, no bumpouts) and fiber insulated cavites comes in at about R14 whole-wall, going a whole lot more than R20 wouldn't make much sense.

You can glue the rigid foam to the wall with blobs of foam-board construction adhesive (standard adhesives have solvents that interact with polystyrene foam- get the right stuff.) Seal the seams & edges with 1-part foam or duct-mastic, not tape to make it air tight, then build the studwall snug up against it, trapping it even if the adhesive bond to the concrete eventually fails.

The SUPER value proposition is to find a local source for reclaimed sheets of roofing insulation. Fiber faced iso and unfaced EPS are pretty common, and typically sell at 1/4-1/3 the price of virgin-stock. A 2" sheet of iso runs ~R12, (R10, if its 1.5" iso, 0.5" perlite bonded to one side, a common insulation under torch-down roofs). Fiber faced iso will run from 0.5 to 1 perm, sometimes higher, but it's fine for this application. I did my basement with reclaimed 3" fiber-faced iso (R18), and NO studwall, trapped in place by furring through-screwed to the foundation, with the gypsum mounted onto the furring. I paid $20/sheet for 4'x8', and I've seen it since for less (virgin goods run $55-60/sheet.) Compare that to box-store pricing on 1.5" XPS + unfaced batting to achieve the same performance.

Other types can work as well, high density rigid fiberglass or rigid rock wool etc, as long as you don't go low-permeance on the facers (like steel clad) or wood/paper on both sides (you don't want wood facers against the concrete if you can help it, from a mold point of view.) Craigslist can be your friend:

http://seattle.craigslist.org/skc/mat/2785589513.html

http://seattle.craigslist.org/skc/mat/2785591503.html

^^^this stuff might work for you, since you're looking at ~65" on the concrete.^^^

But if it's a bunch of inconsistent dimensions it's probably not worth the cut'n'cobble pain.

4x8 and 4x4 roofing insulation deals DO come up, and they're cheap when they do. Searching on "rigid insulation" or "foam insulation" in the materials section occasionally never hurts. Virgin-stock unfaced EPS & fiber faced iso roofing insulation can usually be had from material-distributors who usually deal with commercial construction contractors, but the won't special order it for a homeowner on a project that small- you may have to wait if they're out of stock. You may be able to find reclaimed goods by calling roofing contractors who do flat-roofs/commercial roofing as well (or even demolition contractors.)

[K2]

OK! Thanks for all that information, I've read it about 6 times now, and I'm hoping I'm getting the gist of it. I think you're what you're saying isn't too far off from my original plan.

For the record, I planned on using 2" XPS, but just to be absolutely sure, i took a page out of the building science book to make sure we're talking the same thing.

In the picture, all the boxed highlighted text is what I believe to be your recommendations, can you confirm:

7) A capillary break beneath the sill plate. In the past I've used 30# felt paper, is this still an option in this situation?

8) Additional XPS to fill gap between 2&3(exterior stud and insulation), and 4(planned 2" XPS). (I'm assuming when you said: "Where there's a pony wall with a step in at the sill or concrete, cobble-in rigid foam to even up the sill to the concrete so that you can fill the gap between the new studwall and the original studwall with fiber insulation without putting the insulation in contact with he concrete." this is what you meant?)

9) Additional Unfaced Batt insulation (possibly rockwool/fiberglass) within 6(the planned interior stud wall). (I also read your comment - "The value-proposition would be to use 1-2" of unfaced EPS (beadboard, like a cheap cooler or coffee cup) for R4-R8. Adding that to a 16" o.c. studwall with batts would come to ~R15 (1" EPS) to R19 (2" EPS) after the thermal bridging of the studs & plates are factored in" as an alternative to this, considering, also, your comment about not using cotton below grade...)

You gave me a lot of valuable info, I hope I understand it correctly, pictures always help me, so here goes >>



Thank You!

[Dana]

Unless the place marked "Additional XPS to fill air gap" is perfectly dimesionen for the XPS, there will be air gaps. I was recommending using compressed batts which are more compliant for any depth anomalies.

In the picture it looks like the top of the concrete is chamfered toward the sill rather than leaving a flat step, which could be reasonably filled with a chunk of 1.5" XPS to even it up level with the sill (which is what I was talking about with that verbiage.)

To safely stuff that chamfered corner at the top of the foundation with batts, put a capillary break between the concrete & batt to keep the batts from wicking up moisture out of the concrete. Cotton is VERY hygroscopic, and can pull a lot of water before it feels wet, but it's a mold hazard well before it gets to that level, and even fiberglass will pull water somewhat. A sculpted chunk of XPS or whatever would do the trick,or even a minimal strip of poly sheeting draped over the chamfer by an inch, stapled to the edge of the foundation sill.

Trimming and compressing batts in there for a reasonably tight fit will impede air motion in all directions, whereas less than perfect fitting rigid foam would leave thin flat air channels that allow lateral & vertical air flow, possibly even convection over/under the top. Foam can be an perfect air barrier if perfect, but is a lousy air-retarder if it has gaps that aren't spray-foam sealed on all edges.

EPS would allow better drying toward the interior than XPS, if you don't already have the material purchased. You need not worry about interior moisture drives wetting the concrete via diffusion through EPS, but unless there's a good sill gasket there's always concern about ground water/rain-splash/dew loading up the concrete just below the sill if drying toward the interior is impeded. At 2" XPS is less than 1 perm, which is moderately vapor-tight (a class-II vapor retarder), but at 2" most grades of unfaced EPS are 2+ perms, which is ~3x the drying capacity. EPS is usually cheaper too. Just avoid anything with a foil or plastic facer, some of which are under 0.1 perms, true vapor-barriers.

Yes, only unfaced batts would go in the studwall, but if there's even a remote possiblity of a flood overtopping the bottom plate of the studwall rock wool or fiberglass would be a better choice than cotton batts, since cotton will wick much higher up the wall, and and can't dry quickly- you'd have to gut the thing completely. With rapid-drying materials in the event of minor flooding you can cut the bottom 6-12" of gypsum to let it dry, then patch the gypsum once it's dry enough.

30# felt isn't bad for the capillary break under the studwall plate but rigid foam insulation is better, and gives it a bit of clearance in a spill or minor flooding situation.

[K2]

Thanks for the reply!

I was taking some measurements tonight, and that chamfered edge isn't present in all circumstances, to be honest this whole house is just 'a little bit off' in every measurement and angle. So some places the sill sits flush on a flat concrete wall, in some places it's set back 1/2" from a chamfered edge, etc...etc...

I have some leftover batt material, so I'm hoping i have enough to do the first few walls to get a feel for it...just thinking out loud here, but I obviously need to make sure the batts cover the existing exterior studs toward the inside, or it will leave additional air space, correct?

I did order all the XPS, would it be advisable to not seal the sheets of XPS together with to get that interior dry potential? I had planned on sticking them together with the spray foam, that stuff is as good as any adhesive I've ever touched.

I also got some foam sill sealer and planned on using that under the PT sill plates, I will most likley double the sills so I can wedi/mud the floors per the other thread i got going here.

Anything there truly faulty in design?
Thanks again!
k2

thanks,
k2

[Dana]

Yes, you want to trim and install the batts so that there are no voids, even if it's a bit compressed. If the batts can be split depth-wise to thin them out a bit it might be useful if the depth it needs to compress is too much. Compressing it an inch isn't too tough, but 2" might be awkward with high-density cotton batts. I've never installed them, you'd know better than me. I'm more of a blown-insulation advocate where fiber insulation is concerned, but on small projects like this it's not worth renting a blower if you can be obsessive about fitting the batts perfectly an not creating gaps or compression-voids.

You can and SHOULD seal the seams & edges of the XPS to make it a good air-barrier. Keeping it air-leaky barely affects it's drying potential, since most drying is via vapor permeation through the material. If air is allowed to convect all the way from the room air to the concrete there will be more wintertime moisture drive in that direction. (But in Seattle, that's not nearly as big an issue as it would be in colder climates.) Sealing them with 1-part can-foam is fine. Low expansion "Windows & Doors" variants will have less trimming to do. If you end up with a big blob of expanded protruding into the stud bay or compressed batt side it can cause a batt to bunch & move, creating a void, cutting into the performance of the batt.

The stackup looks good- go for it!

If you're going to finish the floor, now's the time to insulate the slab with at least R5 (1" XPS), but R7.5 (1.5" XPS) is still long-term cost effective in your climate, even if you're heating with cheap natural gas or heat pumps. If you're going to insulate the floor, extend the floor insulation all the way to the wall-foam as a continuous capillary & thermal break, resting the studwall plate atop the floor foam. The 1" version of Wedi is only ~R4, but there's no reason you can't put half-inch XPS sheathing or something under the Wedi to fatten up R value. More than R7.5-8 wouldn't make financial sense unless you're installing radiant floor (in which case you'd be looking at R10 as an absolute minimum) or if you're heating with resistance-electricity/oil/propane, or unless you're using reclaimed foam at a huge discount below retail. (If reclaimed roofing foam use EPS or XPS only- no iso. Some high-density fiberglass panels might cut it, others not.)

With the amount of insulation you're putting in the wall the losses through the floor are going to dominate the heat loss of that space at Seattle's subsoil temperatures. The compressive strength of even lighter duty 1.5lb XPS (Foamular 150, etc.) is more than enough to handle the studwall, which isn't structural- it's not holding up the house, just the insulation. Even low-density EPS can handle any of the loads you're talking about here, particularly with the load distributed by the denser-stiffer Wedi + mud flooring. (Low density EPS can be used under wooden subflooring too.)

[K2]

Thanks Dana,

I have some R13, and R21 cotton batting left over from the first floor, so the R21 can definitely be split by depth.

I will be installing hydronic radiant heat (an extension of my 1st floor system) throughout the basement, so 2 follow on questions:

How important is it to run the foam floor to the foam wall? Only reason I ask is there are a couple walls that already have PT sill plates in place from a prior renovation, and I was hoping I could double up on the sills to get to the right height of finished floor.

Also, if the the sill plate is on the foam, i'm assuming we'd adhere the foam to the floor with some type of adhesive, then what is the recommended way to adhere the PT sills? I'd hazard a guess that you wouldn't want to drill through the foam to tap into the slab.

Thanks again, this is really helpful, and luckily we have enough snow here in Seattle that I can spend time researching before I feel guilty about not getting started.

k2

[Dana]

The heat loss out of the bottom of a doubled-up studwall plate isn't huge. The doubled-plate runs ~R3, and not being directly heated by the tubing. I wouldn't rip anything up to add foam under it as long as it has some sort of capillary break material between the wood and the slab. The critical thing with a radiant floor is getting sufficient R underneath the heated floor itself.

You actually DO want to drill through the plate & foam into the slab to secure the studwall. A Tapcon through-screwed through the bottom plate & foam into the slab every ~18-24" makes it very mechanically secure, and keeps the bottom of the studwall from ever migrating toward the room during minor earthquakes, etc.. Adhesive alone wouldn't cut it. The thermal bridging of the screws negligible, and you don't have perma-frost for soil under the slab either.

Thanks for the heads-up on the snow- I was intending to call my peops out in Port Orchard to see how it was going today.

[K2]

Ok, great, thanks for all the advice I appreciate it, I took the week off to get started and it's wednesday already and I haven't done ANYthing!

i'm anxious to get started!

[K2]

so is there a trick to the tapcon pre-drill? because the drill bit they gave me lasted about 5 holes and now it's completely useless, i can't drill any further than 1/2" into the concrete and the screws are 2 3/4", it's driving me crazy.

[Dana]

Are you using a hammerdrill/roto-hammer or did you try to burn your way through with a standard drill motor. The latter works OK on CMU/cinder block sometimes, but never on concrete or hard-brick.

A trip to the hardware store with the bit in hand you should be able to find a comparable carbide-tipped masonry drill. But you'd still need a hammer-drill.

[Terry]

For concrete, you will want a rotohammer. Bosch makes a nice one. One rotohammer bit should last for the entire job.

[jadnashua]

A regular drill motor will overheat the bit. It is designed to act more like a chipper than a cutter. WIth a normal drill, it's trying to cut. Concrete is too hard for that, it needs to be chipped. WIth the right drill motor, it's not much different than drilling into hardwood.