TFSystems vertical ICF systems allow you to place as much EPS on the outside that you want and zero EPS on the inside.

So, if it is solid concrete on the inside, how do you run your electrical wiring and outlets that are required by code on any wall space at least 2' wide and at least one receptacle every 12' of wall?

You might investigate poured concrete walls with foam dropped into the form.

Arkie6: boxes and conduit in the cmu cavities as the walls go up. I'm betting you can see the end result at the school nearest you. Cmu is a common residential building practice in NM. Could be houses have thin boxes in furred out drywall, but parging plaster over the blocks is quicker, and in my case, cheaper.

Check the TFSystems website. You can either put the conduit in the form before the pour or you can use their inside wall cavity option and plumb/wire like you would for 2x. You can also put pex in the form and heat the walls.

I amittedly had to chuckle over a number of "facts" presented on the website including statements like "the effective R-30 value occurs during times of strong winds accompanied by frigid temperatures."

All testing appears to have been done by simulation by a gentleman whose claim to fame is being a Meteorologist. But who knows maybe he knows his stuff.

To get R-30 anyway, with vertical grout-lines fully bonded inside to outside and horizontials almost bonded is hard to imagine. Also the design of the foam is such that most of the middle foam is wasted as the concrete curves around it to make a beautiful thermal by pass and the parts on each side of it only appear to be about 1 inch thick.

That said, if there is a place to use this product, it would be NM. Compared to standard cmu, this would be a step up.

With regards to OP's original question, if you are going to build a detached garage, you would build the cmu wall on the house anyway, so why not attach the garage with the cmu wall as the fire separation. If it works out on the lot, putting the garage on the south might further shelter the house from heat gains.

I would agree that one should never blindly take it on faith that total R values published by ICF companies are accurate. It is always best to calculate the total R value yourself using accurate R values for the known assembly materials.

Posted By sailawayrb on 23 Nov 2013 06:32 PM
I would agree that one should never blindly take it on faith that total R values published by ICF companies are accurate. It is always best to calculate the total R value yourself using accurate R values for the known assembly materials.

Yep- don't get sucked into ICF marketing fluff referencing the R4.5/inch 40F R value for Type-II EPS.  While that is a valid number, it's only valid when the mid-point between the interior & exterior of EPS is at 40F.  In an even-balanced ICF that true only when the concrete is 40F.

If the  conditioned space interior is 70F, it takes a sustained/average +10F outdoor temp to get there, which isn't exactly the climate where most people in the US live.  Above that temp the R-value of the EPS is less. The labeled R for Tyep-II EPS is R4.2/inch @75F average center temp in an ASTM C 518 test. They can get away with citing R4.5/inch only because they're selling wall system, not insulation. If it were an insulation-only product it could be considered fraud under current FTC rules. 

SIP vendors have similar issues, but they don't have the luxury of playing games with the dynamic mass benefit "equivalent R" the way ICF vendors can.  I've yet to see a SIP vendor deal directly & honestly with the (quite real) thermal bridging of the bottom & top plates and other bracing- it's always a clear-wall infinite-sheet type number in the literature.  Yes, the thermal bridging is significantly lower than studwalls, but it' far greater than zero.

I'm designing a high thermal mass house

I'd be interested in seeing an independent analysis of the expected savings. Preferably one that uses data from your exact location and in the case of internal mass, includes how much variation in indoor temp you are willing to tolerate (no variation means you will have greater energy use due to the elimination of setbacks).

With internal mass, I'd consider some type of automated whole house ventilation. Otherwise you are looking at opening/closing windows 2 to 4 times/day for ~6 months. At least if you want to optimize the use of the mass.

NRG typically saves over 60% per sq ft over conventionally insulated buildings.

IMO, they are scammers.

Dana with all due respect, you're lumping all SIPs into the lumber based variety. Remember steel SIPs require no splines, and what's analogous to the "bottom and top plates" are only as thick as 18 gauge sheet metal - not nearly as substantial a thermal bridge as the lumber variety.

Posted By Jelly on 25 Nov 2013 03:04 PM
Dana with all due respect, you're lumping all SIPs into the lumber based variety. Remember steel SIPs require no splines, and what's analogous to the "bottom and top plates" are only as thick as 18 gauge sheet metal - not nearly as substantial a thermal bridge as the lumber variety.

Since steel has more than an order  of magnitude higher thermal conductivity than wood, it surely DOES matter!  We're talking ~30 Btu/(ft.hr.^oF) for steel, 0.08 Btu/(ft.hr.^oF) for framing type species of wood. That's about a 375x difference, so unless the steel structural webbing spanning the insulation is less than 1/375th the thickness of the equivalent wood member, it'll have the same or greater thermal bridging. (reference)

18 gauge steel is about 0.05" thick (sometimes thicker, depending on plating).  Multiply by 375 x and that's the equivalent thermal conductivity of an 18-19" wide beam of hemlock or yellow pine.

To make it even worse, steel web tied to steel sheeting on both sides as in a SIP is a funky-shaped heat-sink, an I-beam (if in the middle of the SIP) or channel- beam (top/bottom plates- window & door cuts, etc), worse bridging than a steel stud with gypsum or wood on either side, due to the much lower conductivity of the sheathing planes.  Steel studwalls at 16" o.c. with standard sheathing and wallboard insulated to the same cavity-R as wood studwalls built identically, stud-for-stud, plate for plate, suffer about a ~30% lower "clear-wall" R value, despite the dimensional thinness of the thermal bridges.

So while steel SIPS may not have the same spline features of OSB type SIP, the thermal bridging in a steel SIP still matters (and quite a bit.)

Dana, relative to your comments about ICF “effective R-value” (i.e., an R-value that attempts to consider both dynamic thermal resistance and capacitance effects of the materials used in assembly), you are exactly right.  While high mass ICF walls can significantly outperform low mass walls having a similar “standard R-value” (i.e., an R-value as one would normally calculate just considering steady-state thermal resistance of the materials used in the assembly), this is ONLY true when the outdoor temp cycles significantly BOTH ABOVE and BELOW the indoor temp within a 24 hour period. One needs to be fully aware that ICF companies often exaggerate their “effective R-value” numbers and these numbers are also completely invalid if the building is NOT located in this very specific diurnal climate.

Posted By sailawayrb on 25 Nov 2013 05:38 PM
Dana, relative to your comments about ICF “effective R-value” (i.e., an R-value that attempts to consider both dynamic thermal resistance and capacitance effects of the materials used in assembly), you are exactly right.  While high mass ICF walls can significantly outperform low mass walls having a similar “standard R-value” (i.e., an R-value as one would normally calculate just considering steady-state thermal resistance of the materials used in the assembly), this is ONLY true when the outdoor temp cycles significantly BOTH ABOVE and BELOW the indoor temp within a 24 hour period. One needs to be fully aware that ICF companies often exaggerate their “effective R-value” numbers and these numbers are also completely invalid if the building is NOT located in this very specific diurnal climate.

One NEEDS TO BE FULLY AWARE that some posters do not live in ICF houses (;-)
My hands on and not theoretical info shows that both you statements are incorrect in Calgary area. Temp seldom swing enough for that to come into effect yet in four years my heating appliances run considerable less then full time when approaching design temps (which Dana has pointed out to be colder then ashrae numbers). Heat loss using R50 for my ICF walls and standard for all other components indicate a heat loss of 50,000 yet my boiler is rated at 50000. There is little intrinsic heat in this house as there is only two of us in over 6000 square feet of heated space.

Posted By sailawayrb on 25 Nov 2013 05:38 PM
Dana, relative to your comments about ICF “effective R-value” (i.e., an R-value that attempts to consider both dynamic thermal resistance and capacitance effects of the materials used in assembly), you are exactly right.  While high mass ICF walls can significantly outperform low mass walls having a similar “standard R-value” (i.e., an R-value as one would normally calculate just considering steady-state thermal resistance of the materials used in the assembly), this is ONLY true when the outdoor temp cycles significantly BOTH ABOVE and BELOW the indoor temp within a 24 hour period. One needs to be fully aware that ICF companies often exaggerate their “effective R-value” numbers and these numbers are also completely invalid if the building is NOT located in this very specific diurnal climate.

ICF's in the desert southwest are a no brainer building method but ICF's are not an ineffective building system in the colder interior climate zones. According to the 2012 IEEC they recognize ICF/mass walls in Zones 5- 8 and an ICF wall only needs to have a R-21 steady state rating to pass code. That means 95% of ICF's out there are allowed in Zones 5 - 8. The 2.5" EPS x 6" concrete x 2.5" EPS forms will pass code all the way up to the arctic Zone 8 climates. At the same time the 2012 IEEC requires a wood framed home to have R-20 interior plus R-5 exterior OR R-13 interior plus R-10 exterior to pass code in Zones 6 - 8.

ORNL did their DBMS tests many years ago and used the "hot box" in a limited way and even then, they stated that there is an "effective thermal mass benefit" to ICF's. Of course climates like the desert SW work better but colder interior climates like Zones 6 - 8 can still benefit from it.

Exactly right, it is the companies and folks who have little theoretical background and often little integrity that proclaim the magical heating abilities of ICF. However, when someone with a theoretical background actually looks into the claim and actually takes the time to take exact measurements, they always find that the building heat loss is perfectly consistent with heat transfer theory. When outside temps stay continuously above or below the inside temp, the ICF behaves precisely like a "standard R-value" wall assembly. Frankly, I wouldn't live in anything but an ICF house these days. However, there is little need to make false claims to justify doing so.

""When outside temps stay continuously above or below the inside temp, the ICF behaves precisely like a "standard R-value" wall assembly.""

If this was the case and you did the design, you would be installing an oversized boiler.
Apart from the odd chinook, our outside temp values never get close to in our inside temps seven months of the year. Yet heating experts here in Calgary routinely use R-50 for ICF walls when doing heat loss clac's. Even those who sell heating equipment. Why would they want to sell small units and make less money when according to a number of posters on this forum they should only use R23 and sell bigger units?

Instead of saying that ICF people make false claims, spend time examining why ICF perform at least twice as good as their standard R-value.
To steal a line from one of our Geo friends "One measurement is worth a thousand expert opinions."

Posted By FBBP on 25 Nov 2013 07:10 PM

Posted By sailawayrb on 25 Nov 2013 05:38 PM
Dana, relative to your comments about ICF “effective R-value” (i.e., an R-value that attempts to consider both dynamic thermal resistance and capacitance effects of the materials used in assembly), you are exactly right.  While high mass ICF walls can significantly outperform low mass walls having a similar “standard R-value” (i.e., an R-value as one would normally calculate just considering steady-state thermal resistance of the materials used in the assembly), this is ONLY true when the outdoor temp cycles significantly BOTH ABOVE and BELOW the indoor temp within a 24 hour period. One needs to be fully aware that ICF companies often exaggerate their “effective R-value” numbers and these numbers are also completely invalid if the building is NOT located in this very specific diurnal climate.

One NEEDS TO BE FULLY AWARE that some posters do not live in ICF houses (;-)
My hands on and not theoretical info shows that both you statements are incorrect in Calgary area. Temp seldom swing enough for that to come into effect yet in four years my heating appliances run considerable less then full time when approaching design temps (which Dana has pointed out to be colder then ashrae numbers). Heat loss using R50 for my ICF walls and standard for all other components indicate a heat loss of 50,000 yet my boiler is rated at 50000. There is little intrinsic heat in this house as there is only two of us in over 6000 square feet of heated space.

How did you model the solar gains?

"Very little intrinsic heat" isn't a hands-on measurement of anything, and doesn't even rise to the significance of a theory. Instrument the house, model the solar gains, and you'll see why your average fuel use is what it is.

For the record, what is your average power-use rate, and how much of that power is used inside rather than outside the thermal envelope?  Average plug loads are way bigger than zero, even for off-grid electricity sipping houses.  A typical US home averages about a bit more than a kilowatt, an energy conscious 2-person family might average 1/4 that if they usually live in the dark, have only one tiny refrigerator, don't watch TV, and only power up their computer & internet connections for an hour a day.  Even at an average of 275 watts that's ~1000BTU/hr of background energy gain.  A pair of sleeping humans is good for 500 BTU/hr- awake but sedentary that bumps to 800BTU/hr. On your feet and walking (not running) around it's over 1000.   Even if you have a bare couple thousand BTU/hr of internal gains that's already 4% off your calculated 50K heat load.

Even a cloudy-sky day will  be good for at least a couple thousand BTU/hr during daylight hours (unless you have no windows, and a  high solar reflectance roof and wall-paints), and in full sun it'll be north of 10,000BTU/hr, pre-charging the thermal mass of the walls in an ICF to a higher temp, with a relevantly long decay time that does not occur with  low-mass walls.

The thermal mass of the walls (and other materials in the house) DOES have a significant impact on reducing (primarily) your peak loads, but modeling the ICF as a low mass R50 even for treating peak loads is a cruder-than-crude model, with plenty of potential for paradoxical results. Solar gains will usually be a bigger factor than thermal mass on average loads.

DOE2.2 (the model underneath BeOpt, both freebie downloads from the US DOE, with weather data for many locations in Canada) does a reasonable job of modeling both solar gain and mass effects.  Both effects are real, and both make a difference in your actual energy use.

The "effective R-value" of mass walls really is climate/shading factor/design-specific, no matter what the marketing hype states, and now matter what equivalent-R your getting out of your particular (un-modeled) build.  The modeling tools aren't perfect, but they're pretty good- WAY better than a WAG, but importantly, offering far more accuracy & insight than cherry picked marketing numbers.

Well FBBP, we prefer to not make claims that can’t be fully explained and substantiated and that are also inconsistent and conflict with well established heat transfer theory. Believe what you will if it makes you happy. And you certainly should be happy if your building is performing at or better than your forecast design performance and objectives. However, differences between forecast design performance and actual performance can nearly always be fully explained by design errors and design omissions, and are rarely the result of newly discovered physics. So if you have a wall assembly that is performing at twice the “standard R-value” as you claim, you should consider documenting your findings and seek a patent invention.

Dana, your explanation has much merit as always.

Taking a not very accurate heat load estimate for an entire house, comparing it to actual use and then trying to back calculate the R value of the walls is ludicrous. You could just as well attribute the difference to the color you painted the ceiling, eg "yellow paint doubles the ceiling R value".

It's also important to differentiate between seasonal energy savings and dampening. A high mass house can store heat (say from the furnace) and release it when temps drop, thereby reducing peak loads and required system size. This effect happens even when outdoor temps remain below comfortable. But using this for reduced system sizing will fail if the outdoor temperature drops to design temp and stays there for too long.

My statement of “very little intrinsic heat” is simply to say there is nothing extra in this house that would account for additional heat. There is a fair amount less energy used then would be used if this house were occupied by a family of five which is the least you would model a house of this size to.

Well Dana – even if all you say is true, it is all stuff that should have been taken into account in any heat loss calculation anyway. The only thing we are changing here is the r-value of the walls. Using R23 gives us a boiler that has significant unneeded capacity. Using R50 give us a safe margin.
Your grasping at straws to discredit the R50 is so unlike you.
The last minus 32 period was overcast and stormy for the better part of a week. The roof is over a vented attic. Walls have a light coloured Hardie over a vented ½” air gap.

Sailor – What makes me unhappy is people who refuse to look beyond, to see that maybe something in the textbooks is not quite right. This is not a new physics; it is just something that the present methods of testing are not accounting for.

Jon – this is not “a not very accurate heat load estimate for an entire house” The heat loss calc’s where done by a gentlemen certified to use Wrightsoft software and done on the latest version at that time. Wrightsoft might not be the best software but in the hands of a professional, it is close. Also I am not trying to “back calculate”. All I am saying is that the professional doing my heat loss calc’s used R-50 as the factor for the ICF wall portion of the building and after 4 years it has worked marvelously.

The company doing the heat calc’s has used R50 on all their ICF houses in the Calgary area and had no complaints in the past 10 years. If a supplier uses R23 and sells the client an over sized unit, the house is comfortable and no one complains. If he sells an undersized unit and the client is cold, the whole world will know. This is not a one off. It is not some new physics. Competent ICF installers and their customers have been saying the same thing since Greenbuildingtalk first started. There was always somebody to shout them down, so they just went away. Take the time to reread the ICF forum and that is what you will find. Because they got tired of listening to people say that there is no way that ICF can perform at those levels, they just took the message to where it would be appreciated and GBT is the poorer for it.
Another interesting thing about GBT is that there is always someone to say that suppliers don’t bother doing calc’s properly and just sell the client oversized units. When they do it wright, they are wrong too. Poor guys just can’t win.

Your unwillingness to at least explore what many people have been saying for a number of years is not a service to GBT readers. For what its worth, I believe you are seeing the difference of insulated mass as compare to naked mass which is what OP is talking about. His is not an ICF system.