I know some of these questions have been asked many times before and I have read many of the postings but will ask some other specifics here.

I am building my own house in Vancouver, Canada. Our code requires R12.

The basement is basically a square with walk-out design. Two of the walls are esentially full depth bury. One wall is about 4' bury and the remaining (walk-out) wall fully exposed. Before backfilling the interior of the fully exposed wall i installed 2" XPS from top of slab down 24" figuring it was easy to do and would help cut off perimeter losses since the exterior is in the frost zone.

I understand there are basically 3 types of ridgid insulation available for this purpose; EPS, XPS, and Poly Iso. Can I get some information on the pros and cons of each and when not to use one or the other?

My details call for 3" XPS perimeter insulation, that part is easy enough.

It is the slab field insulation and edge (still under the slab) that i am thinking might be optimized with 3" of cheaper EPS field and 2" EPS & 1" XPS edge. Again advice sought.

Thanks

John

Only use foam rated for under slab. Dow makes it in a long range of load capacity. You may have to go to a large concrete supply house to get the higher ratings. Big box stores stock the minimum rating If you use different thickness at perimeter, it is very important that the top is 100% flat. Any ridge will leave a step on the under side and a crack will surely form.

Patrick T

XPS and polyisoc offgas and lose R value over time. they stabilize to values that are higher, but not that much higher, than EPS and they are less friendly between now and that day because of the offgassing. they are also more expensive, TYPICALLY though XPS can sometimes be similar in price.

the only time I recommend using them is when an extra R1 per inch or so is really important. that's not very often. that said, XPS is commonly stocked for slabs so it is often used for that reason.

XPS is all that is used in the Midwest and most this country and yours for that matter. I don't see off-gassing as an issue in foundation work or several miles of driveways we are now snow/ice melting.

On walk-out basements here in Minneapolis, we always insulate four feet out from the entry door as frost doesn't care if you insulate in, down or out. Naturally there is a thermal break between in and outdoor slabs.

The ROI is much lower below ground so extra attention, and insulation, should be given to the exposed wall and foundation.

Rob

Thanks for your info. Are there any workability issues with EPS? I know its more fragile. Does getting wet bother it? Around here XPS is about double the cost on a per inch basis!

I do like the idea of increasing the under slab insulation at the edge along the walk-out elevation (since its close to the frost line). Not sure what my best option is to do that though (I've already got 2" XPS vertical from the top of slab down 24"). Do I do that under slab edge in XPS or thicker EPS?

(Probably overkill but going with mesh on 10M (3/8") rebar on 24" centers, AND fiber reinforcement. The fiber allows me a 3" slab and is significantly cheaper than a 4" without it. I mention all this because I doubt it will crack and if it does there's not going to be any displacement.) Last but not least, what is my chair height? I presume I want the PEX mid-height in the slab so likely 1" chairs under the rebar.

Thanks

John

Price per inch is half as is the R-value. The cost per R-value is usually close.

The thinner the slab the less you will get from raising the PEX. We don't do it in a 4" slab.

The fiber allows me a 3" slab and is significantly cheaper than a 4" without it.

Every time I've done a 3" slab, I've been sorry. I've even been sorry doing 3-1/2" slabs. There just isn't that much saved in the long run and you get a better slab.

If you are doing a total of 3,000 sf of slab, the difference amounts to about 10 yards or one load. That's about a thousand bucks. You are already wasting much of that on your slab reinforcing. Using fiber, wire mesh and bar is truly overkill. What's your cost of fiber up there? Down here, it was running about $7/yd, so that's $220 for the thin slab, maybe up to $300 for the 4" slab. The wire mesh will run you $500 either way and that's without delivery or labor.

ICF has it exactly right.

We do many slabs for snow melting driveways as well as radiant heating slabs. Very few are specified with wire or re-bar, almost every one is 4"--if your lucky.

In most driveways we will use flat wire over PEX just to make handling concrete easier. Naturally it will help in crack control but we are confident that the fiber is sufficient and have had zero problem with fiber-only reinforcement in our extensive basement remodel and new basement slab radiant systems. I would rather invest in more insulation or efficient heat sources.

"Half" the R-value? EPS is a 4 per inch. XPS is a 5... before it offgasses. and there is nothing magic about a slab install that keeps a blower agent entrapped in the foam cells... it will offgas under a slab, in a warehouse, on a wall, whatever. EPS is a better value in almost any case.

I stand corrected. Just specifying some closed-cell foam over open-cell, which is in fact about twice the R-value. As for EPS vs. XPS, you have had me thinking for some time now. From the numbers it still EPS appears to be about 25% short of the equivalent inch of XPS--5.72 vs. 4.17 -- ( the ultra-fine point of off-gassing aside) and still a little structurally flimsy for my taste, but if the price is right, I will spec it on my next project (provided our staple guns will hold the PEX down long enough to pour the slab over it.

http://arlingtonpassivehouse.wordpress.com/tag/eps-v-s-xps/

the better the initial R, the more it drops with aging. Offgassing is not an "ultra fine point"... it erases the additional R value you just attributed to XPS, which is a significant amount of its R-value.

in the end you get 25% more R PER INCH with XPS. it typically costs a lot more than 25% more. so... why not just get more EPS, in most cases?

I honestly cant' say. But I will find out and get back to you...

Rigid Cellular Polystyrene (RCPS) performance is a very hotly debated topic and one can find much conflicting data put out by each side since this is a high $$$ business proposition…

According to Insulation Corporation of America: “Expanded polystyrene (EPS) is the winner! This was clearly shown in an independent, third-party test. Expanded polystyrene (EPS) maintains its R-value even after long-term exposure in northern climates. Extruded polystyrene (XPS) was shown to have lost R-value over time. In this 15 year side by side dual EPS outperforms XPS in both R-value retention and decreased water absorption. XPS insulation was reduced by half, while EPS boasts maintaining 94% of its specified R-value. Not only does XPS lose about half of their R-value it also costs about 50% MORE than EPS. So why would anyone pay 50% more for 50% less? EPS is manufactured in large blocks, it is cut into sheets or virtually any special size and shape. XPS is a final form and does not have the size flexibility that EPS offers. The product performance, size flexibility, recyclability, and the lower price of our “white stuff” clearly makes EPS the champion.”

The International Building Code (IBC) and International Residential Code (IRC) are the predominant building codes. These codes rely on the American Society of Civil Engineers (ASCE) 32-01 standard to determine appropriate thermal properties of XPS and EPS used for below-ground building applications for long-term protection against frost heave. ASCE 32-01 shows that nominal 5.0 R per inch XPS effectively becomes 4.0 after long-term exposure when used horizontally (retains 80% of original nominal R-value). Nominal 4.0 R per inch EPS effectively becomes 2.6 after long-term exposure when used horizontally (retains 65% of original nominal R-value). Nominal 5.0 R per inch XPS effectively becomes 4.5 after long-term exposure when used vertically (retains 90% of original nominal R-value). Nominal 4.0 R per inch EPS effectively becomes 3.2 after long-term exposure when used vertically (retains 80% of original nominal R-value).

So, relative to Rigid Cellular Polystyrene (RCPS) thermal properties, both products may be successfully used provided one uses the appropriate long-term R-value in the design. We use the ASCE 32-01 values. Of course one also needs to consider RCPS structural properties in the design to avoid building slab settling. This is always important and becomes increasing critical when you are designing a hydronic radiant heating system for a building in Anchorage for example with many inches of RCPS.

Correct me if I am wrong but the study quoted seems to be a combination of both off-gassing and water saturation R value degradation. Our code specifies the poly goes under the insulation so hopefully it stays dry.

XPS is seriously more expensive here. My local yard has good pricing even so:

3" EPS $0.83 sq.ft.
3" XPS $3.09 sq.ft. - yes 3.7 times as much!

Mesh $0.18 sq.ft.
10M (3/8") bar $0.25 ft.
32MPa concrete with fiber reinforcement ($15/m³) $125/m³ (in imperial units 4600 psi $96/cubic yard).

A coworker did his place with 10M bar on 24" centers, mesh on top. This kept the mesh from getting crazy bent with all the concrete workers on it. Reinforcement (on chairs) and PEX kept at the right height. Seems quite reasonable to me. Fiber is cheap insurance.

With this "overkill" reinforcement can someone give me specific issues with a 3" slab?

Thanks

John

You don't have to worry about the mesh getting trampled if you don't use it in the first place. Did you have pricing for the concrete without the fiber additive?

The thicker a slab is, the better it holds together. When you lay concrete, you are creating a composite. The composite isn't very good if the reinforcing parts aren't fully contained within the matrix.

Since nobody has mentioned it I'll state for the record, you should never (and I mean NEVER!!) use polyiso under a slab, or between a foundation-wall and soil. But it's OK to use on the interior side of open basements. Iso is mildly hyrgroscopic, and will wick & store moisture over time, if it gets regular exposure.

Initial moisture uptake of EPS is faster than XPS (due to the interstitial spacing between beads), but at comparable densities has higher fraction of fully closed cells, making it somewhat better at handling long term moisture exposure. Even at 7% moisture content it still retains most of it's R value, as indicated in sailawayrb's industry references.

Both EPS and XPS have significant derating/uprating curves with temperature. In the US the legally labeled R value of Type-II EPS is R4.2/inch when the average temperature through the material is 75F, but the tested R value goes up at lower temps, down with higher temps. In a heated slab where the bottom side of the slab is 85F and the subsoil temp is 55F, it's performance is somewhat better than labeled.

The derating/uprating curves for XPS are similar, but not as pronounced as with EPS, at least early in it's lifecycle. Freshly blown it's R-value is about R7/inch @ 75F, but that falls fairly quickly (months, a couple of years at most) to under R5.5/inch, with a much slower decline over time from there. It is allowed to be labeled at R5/inch by it's average performance over some presumed lifecycle (IIRC it's 25 years, but don't count on that.) It's fully depleted R-value of 1.5lb XPS is identical to Type-II EPS. In under 50 years any performance difference would be largely theoretical (and miniscule.)

In Europe use of HFC134a (the most common blowing agent used for XPS in the US) is banned for this application due to climate-change regulations, since HFC134a has about 1400x CO2 global warming potential (GWP). In Europe XPS is usually blown with CO2 (which strangely, has only 1x CO2 GWP :-) ). The blowing agent has similar thermal properties as the air that eventually (and fairly rapidly) displaces it, and CO2 blown 1.5lb XPS tests at R4.2/inch @ 75F in an ASTM C518 test, just like 1.5lb EPS.

In the US most EPS is blown with pentane with only 7x CO2 GWP. Pentane is a sufficiently small molecule that it's depletion rate has a half life in weeks, not months, and it pretty much hits it's fully depleted R value in well under a year, and is thus only labeled at that R4.2/inch @ 75F fully depleted value. Very early in it's life it would test slightly better, but nobody other than insulation designers would even bother to look at that- it's irrelevant from a lifecycle performance point of view.

A 3" slab meets code-min from a structural point of view but in a heated slab you'll have higher mechanical strain from the internal temperature differences, and without fiber reinforcement (or even with) is more susceptible to cracking

Brilliant! Thank you.

I have not found concrete people to be exceptionally accurate with slab depth, so pouring a 4" slab would likely give you a more realistic 3.5 average and I work with some very good concrete companies on our snow/ice melting jobs.

Yes, both incredibly informative and very thought provoking data Dana!

Don't mind me, I just make it up as I go along! 

(If you need online references for any of it, I could probably dig it up for you.)

Taking it even further off topic...

There has been a lot of development work over the past 5 years on low-GWP blowing agents for closed cell polyurethane to replace HFC245fa (the most common in use today, with over 1000x CO2 GWP) and the other similar products. Honeywell released it's trade-named Solstice blowing agent for polyurethane foams (GWP <5x CO2) late last year.  I'm not sure if DuPont has fully released it's FEA-1100 product (<2x CO2 GWP) in the US yet  (it seems to get more attention in Europe, which has much stricter regs), but if not it's only a matter of time.  These things don't gain market share overnight, and without regulation limiting GWP on blowing agents similar to what the Montreal Protocol did for ozone-depletion potential, HFC245fa will probably continue to command the lions share of the 2lb foam market in the US.  It's a reliable known product, with mostly good foam-production track record, were it not for those other pesky issues...

Changing over has learning curve and adjustment costs- I'm not holding my breath.  While HFC245fa was a HUGE overall environmental improvement over it's pre-Montreal Protocol CHFC predecessors, copious use of HFC-blown 2lb foam as insulation is a strong net-negative for the environment over any projected lifecycle.  It has many great properties that make it VERY useful in building assemblies, and it's not wise to throw the baby out with the bathwater, but where it can be designed down (or out), it's probably better to go with lower impact alternatives.

I've yet to get wind of a HFC134a replacement for XPS that delivers similar or better lifecycle-R performance without the onerous lifecycle GWP hit. Like 2lb polyurethane foam, if you can design it out/down, it's probably the right thing to do for the time being. Despite it's it's comparatively lower mechanical ruggedness, EPS is the obvious "environment-lite, 99.5 GWP-free" alternative in many applications, even 2lb polyurethane blown with HFC245fa is lower impact than US-style XPS, R-for-R. 

There are a very few 2lb spray foams out there using water as a blowing agent, but finding competent foam contractors who have it is a needle-in-haystack kind of deal.  Aloha Energy is a smaller regional manufacturer with only water-blown product, available in several densities.  Icycene makes a semi-permeable 2lb R5.2/inch water blown product, designation MD-R-200, but their other higher-R 2lb foam is blown with HFC245fa. 

From a building assembly designer's point of view I prefer Icynene's higher-permeance/lower-R 2lb foam, despite the somewhat lower thermal performance, since you can go higher-R in the foam layer without creating moisture traps. But it is still sufficiently low-perm to be useful in moisture control, unlike open cell foams.  It runs ~1.3 perms @ 3"/R15.6, compared to typical HFC blown 2lb foam at 1.2-0.8 perms @ 1"/R6-7.  Between studs or rafters, half the performance of R7/inch foam (as well as R5/inch foam) is robbed by the thermal bridging anyway, but from a dew-point control within the assembly point of view in a foam/fiber insulation stackup you can actually get there with the higher-perm stuff without closing off drying capacity toward the interior too much. But I s'pose most folks just look at R/inch, bigger is always better.

Factoid of the day:  In a 16" o.c. 2x6 wall assembly with a 25% framing fraction (typical, unless code requires fire blocking & seismic reinforcement) a shot of 5" of closed cell foam is R30 center cavity, but the average R of the assembly performs at only about R15 due to the thermal bridging of the framing.  That is a performance boost only ~R2 higher than if you filled it with bottom of the line low density fiberglass(!).   But if you put 2" of EPS on the exterior of the sheathing on the same crummy R19 low density fiberglass-insulated wall you're over R20- with 25% lower heat loss than with the heavy SPF wall, at much lower cost.  R8.4 foam on the exterior is sufficient for dew-point control up through and including US climate zone 5 and the very warmest edge of zone 6. That allows you to get by with standard latex paint as the interior vapor retarder- if don't use batts with facers or interior side poly you get drying rates at the sheathing more than 50x faster than with the 5" closed cell foam solution.  That's as much moisture purging in one week of warmer outdoor temperatures as a whole year of drying-weather with the closed cell cavity fill option!  As long as there is sufficient exterior R to not overload the sheathing during the winter resulting in high early-spring mold potential, it'll will tolerate far more moisture from any path than the cc foam solution. (Even R20 cavity/R7.5 foam is the minimum that works everywhere in zone-5, so R8+ buys you a bit of margin, especially with the cool-temp uprating curve for EPS. The foam/fiber R-ratio is key, but there are other factors in the model too.)

And if that's too thick a wall for you, a 2x4/R13 wall with 2" of exterior EPS would come in about R17-R18 after thermal bridging. That is still R2-R3 better than the 2x6 closed-cell wall with the same wall thickness, and dew-point control would be sufficient for any location in the US lower 48.  Swap the EPS with polyiso and you're at R20 (or just shy of it, in the colder climate zones, after cool-temp derating of the iso.)

OK, back to the regularly scheduled radiant slab discussion.

" Factoid of the day: In a 16" o.c. 2x6 wall assembly with a 25% framing fraction (typical, unless code requires fire blocking & seismic reinforcement) a shot of 5" of closed cell foam is R30 center cavity, but the average R of the assembly performs at only about R15 due to the thermal bridging of the framing. That is a performance boost only ~R2 higher than if you filled it with bottom of the line low density fiberglass(!). But if you put 2" of EPS on the exterior of the sheathing on the same crummy R19 low density fiberglass-insulated wall you're over R20- with 25% lower heat loss than with the heavy SPF wall, at much lower cost."

Now you tell me!