Just wondering...Logix (and others I believe)make blocks with extra thick foam on one side. Would there be an overall heat loss advantage to use the thicker foam on the the north wall and less expensive 2.75" foam on other walls? This is assuming the transition from the wider to regular forms is workable. Dave

If it is colder on the north, than I suppose there would be a minimum benefit, during sunny days in the winter. When the sun is not out or it is not the heating season I suppose it would not help.

It is probably less expensive to just add 1/2" or 1" or more thickness of 4x8 sheets of EPS, XPS, or polyiso foam over the exterior after the wall is poured.

What will the exterior of your home be?

If you have a 3/4" air gap between the wall and the exterior siding, brick, etc., you can get an extra R2-R3 with a foil facer on the exterior foam. For instance, 1/2" polyiso is rated R3.0, but if you have a foil facer on it facing a 3/4" dead air space, the system R value is 5.8.

http://www.atlasroofing.com/tabbed.php?section_url=15&tab=31

I guess my question might also be... should one try to reach the same R value for all walls? Does heat flow rate increase on the less insulaed walls?

Arkie, we're planning faux stone, 2 to 3.5" thick, over the ICF. Using the polyiso is an interesting idea and would eliminate the transition to a narrower form. I didn't see any case studies at your referenced website showing how to install on ICF walls.

Am interested in cost effective way to increase R value of walls so I can minimize the HVAC eqpt. Dave

"should one try to reach the same R value for all walls? Does heat flow rate increase on the less insulaed walls?"

No, if you increase the insulation on one wall, the heat losses do NOT increase on the other walls.

Posted By Farmboy on 18 Dec 2011 04:31 PM
Just wondering...Logix (and others I believe)make blocks with extra thick foam on one side. Would there be an overall heat loss advantage to use the thicker foam on the the north wall and less expensive 2.75" foam on other walls? This is assuming the transition from the wider to regular forms is workable. Dave

What about QuadLock? It offers the ability to pick thicker EPS on the outside of the form. The standard ICF is R-22, which is 2.5" of EPS on each side with a 6" concrete corner. QuadLock allows one to go R30, R38 and even R45+


QuadLock Info

Lbear. Quadlock probably has the largest selection of options in this regard. Nudura does this with inserts (which I have actually never used). In my opinion, despite that fact that I am a wall guy, I think you are at a point of diminishing return with the walls (in most climates) since you are above a continous R-22 or so. If I wanted to this this I think I would go down the path of what arkie6 is suggesting above. Once you upgrade your walls you will put an even greater load on the roof and windows etc and your money may be best spent there.

Farmboy, If you take a hypothetical foam box with a perfect or ideal floor and roof (no heat transfer via floor or roof) make the walls 10 inches thick foam only (we'll say R-40) you will have uniformity of heat transfer through the walls. If you then make one wall one inch thick you will have a disproportionate heat transfer through the thinner wall and less transfer though each of the other walls respectfully. Therefore, the converse is also true. Make one wall extremely thick and you will have additional heat transfer through the other 3 walls. If this is hard to imagine make the wall 100 feet thick .

This comes back to heat is going to take the path of least resistance. Another point going back to the original hypothetical foam box. If you put one stud (say a 2x10") - in one wall only.... let's say it represents only 1% of the entire envelope. You cannot make the mistake of assigning one percent of the wall the R-value of wood and 99% the R-value of foam. A disproportionate amount of heat will be lost through the stud. If we say the 10" thick wood is R-10 and the foam is R-40 the wood will bring the actual R-value of the collective walls down to about 38.8. This comes from one over the sum of the U-values or rather Actual R-value =1/((.01/R10) + (.99/R40)). So here one percent of the envelope impacted about (40-38.8)/40 = about 3% or using U : (0.02577-0.025)/0.025 = about 3%.

Lbear, all this is to say -- seek out the lowest common denominator. At this point in your quest you may want to look into the windows. BTW -- I believe I used the same windows you were considering JeldWen. Regards.

Posted By TexasICF on 21 Dec 2011 10:05 AM
Farmboy, If you take a hypothetical foam box with a perfect or ideal floor and roof (no heat transfer via floor or roof) make the walls 10 inches thick foam only (we'll say R-40) you will have uniformity of heat transfer through the walls. If you then make one wall one inch thick you will have a disproportionate heat transfer through the thinner wall and less transfer though each of the other walls respectfully. Therefore, the converse is also true. Make one wall extremely thick and you will have additional heat transfer through the other 3 walls. If this is hard to imagine make the wall 100 feet thick .

Yes and no. If the heat input into the space remains constant then yes, the proportion of the heat will shift according to the R value, and the space temperature will change. But, if the temperature is held constant then increasing or decreasing the R value of one wall will not affect the quantity of heat moving through the other walls. Heat flow depends on delta T from inside to outside, so changes in one wall will not affect the heat flow through the other walls. The total heat flow from the space will change.

TexasICF-

Heat transfer rate equation:

q = k*A*(T2-T1)/L

q = heat transfer rate (W or Btu/hr)
k = thermal conductivity (W/(m degC) or Btu/(hr degF ft))
A = area (m^2 or ft^2)
T2 = higher temperature (deg C or deg F)
T1 = lower temperature (deg C or deg F)
L = thickness of material (m or ft)

Please notice that there are no terms in this equation for the heat transfer rate through other nearby walls!

You are getting confused between absolute and relative heat transfer rates. Making one wall so that it transfers no heat does not change the heat transfer rate through the other walls as long as the temperatures are maintained at the same values. See equation above.

At night when the total heat loss is highest, the exterior temperature of each of the walls will be pretty close to one another. This is also true on overcast or stormy days.

On sunny days the sun-exposed walls the average temp will be higher due to solar heating of the surface, but that also corresponds to higher air & radiation temp on the north walls but to lesser degree. The mis-balance occurs only occurs during sunny hours,. which is also when the heat load is significantly lower due to solar gains through the south facing windows. On average the difference is only a second-order effect, and there's no particular advantage to higher-R on the north wall than the other walls, or rather, not enough of an advantage to pay anything extra for in design, labor, or material.

But there IS a heating season advantage to disproportionate glazing area and types for the different sides of the house. At ~R20 and higher whole-wall Rs the heat loss from windows begins to exceed that of the walls at typical glazing fractions (unless you start going to ever higher-performance and more expensive windows.) But utilzing sing high solar gain double-glazing and more glazed area on the south side (even if it's 30-50% lossier than a super-window) yield a net heat gain in any US climate, whereas on the north side you're only looking at heat loss. Some solar tempered designs might warrant higher solar gain on the E facing windows as well, maybe even a small increase in glazed area, but not the W side, since that makes the house more prone to overheating during the cooling season, since the peak solar gains through the west side comes at a warmer part of the day.

In cooling dominated climates it can be important to minimize W facing glazing even with substantial S-glazing, since the solar angle is lower, and the window reflects less of the heat than on the S side in summer.

Posted By Lee Dodge on 21 Dec 2011 12:00 PM
TexasICF-

Heat transfer rate equation:

q = k*A*(T2-T1)/L

q = heat transfer rate (W or Btu/hr)
k = thermal conductivity (W/(m degC) or Btu/(hr degF ft))
A = area (m^2 or ft^2)
T2 = higher temperature (deg C or deg F)
T1 = lower temperature (deg C or deg F)
L = thickness of material (m or ft)

Please notice that there are no terms in this equation for the heat transfer rate through other nearby walls!

You are getting confused between absolute and relative heat transfer rates. Making one wall so that it transfers no heat does not change the heat transfer rate through the other walls as long as the temperatures are maintained at the same values. See equation above.

Lee, 

VERY INTERESTING.  Your making my brain hurt.  I believe your equation is correct but we do have two thicknesses L1 and L2 --- one for the 3 walls and one for the North wall and therefore two equations with q total = Q1 + Q2   Having given this additional thought I think dmaceld (above) is most correct with his answer of "Yes and No" now that I read it again I think he explains it quite well.  

Q = Q1 + Q2 = k*A1*(T2-T1)/L1  + k*A2*(T2-T1)/L2 

Since we changed only one wall  of the four walls then  A1 = 0.75A1 and A2 = 0.25A1

and if we made the North wall (L2) twice as thick as (L1) then:  L1 = 2*L1

so

q = k*(T2-T1) * ( 0.75A1/L1 + 0.25A1/(2*L1)) = 0.875 k (T2-T1) A1/L1

Therefore the total heat flow from the space will change by about 12%.

So (Yes) when you double the north wall thickness if the furnace is on (constant heat) while the overall heat transfer rate goes down then the temperature would go up and the other walls would experience additional burden.  And (No) If the heat source is managed by a thermostat and a constant temperature is maintained then you would have no additional heat traveling through the other walls.    Is that correct? 

It seems like the best answer might be No but it is also Yes because a maintained heat source will impact the other walls.  And if you flip it around and the heats trying to get in there's no outside thermostat.    Regards.

Only if the average temperature is higher would there be additional "burden" (more heat transfer, really) on the other walls. At a constant interior temp that wouldn't be the case. Decreasing the heat flux through one wall will always decrease the total, and losing temp more slowly is a GOOD thing.

It's usually easier/cheaper to achieve equivalent net reductions in heat transfer by shrinking the window area &/or lower-U glazing on the north side windows than doubling the R with EPS. At R22 and higher, focusing on increasing the R value of one (or more) walls is to ignore the bigger picture. As an example, a 10' x 40' north wall with five 10 sq/ft U0.34 windows will have:

350' of U0.045 wall, for a heat loss of (350 x 0.045=) 15.8 BTU/hr/F

and 50' of U0.34 window for (50 x0.34=) 17 BTU/hr/F...

which is more than half the (32.8 BTU/hr/F) heat transfer though the wall.

If you reduce the size of the windows 10% and go with pretty-good not too expensive U0.28 windows the heat transfer through the window drops to (45 x 0.28=) 12.6 BTU per F-delta, and the heat transfer through the wall increases to 16BTU/hr/F, for a total of 28.6BTU/hr/F, a reduction of ~13%.

To achieve that 13% reduction with fatter EPS would require reducing the wall heat flux from 15.8 BTU/hr/F down to 11.6 BTU/hr/F, or (11.6/150=) 0.033BTU/hr/F per foot, which is (1/0.033=) R30, and increase of R8 (2") above the original R22. Assuming roughly 10 cents/R/sq.ft. for EPS that's ($0.10 x 8 x 350=) $280 in cost adder for that wall, and the slightly-smaller slightly-better windows will often be a lower uptick in cost. (With ICFs the upcharge for going from R22 to R30 EPS may be more than that- my reference is typical costing on sheet goods.)

If you really want to sharpen the pencil on the big picture, you can now model the whole shebang using (at no charge) BeOpt:

http://www.greenbuildingadvisor.com/blogs/dept/musings/beopt-software-has-been-released-public (<

And if you add light-blocking shades with side seals to the windows in Dana's example and assume that they are closed 60% of the time (night-time) in winter, then the effective R-value for the windows increases from 3.6 (equiv. to U = 0.28) to 3.6 + 60% * 2.0 = 4.8, or U = 0.208, and the heat transfer through the windows is decreased to 9.4 Btu/hr F, and the whole wall drops to 25.4 Btu/hr F, a further reduction of heat losses by another 11%. Since it is colder at night than during the day, and the R=2.0 value is kind of a worst case, then the actual saving should be higher. The shades are expensive enough (~$100 a piece for windows of this size) that they are probably not economically justifiable unless you need some type of shades or curtains for privacy, but many times you do.

Thanks all for the illuminating discussion and thermodynamics lessons. Didn't realize how much in the dark I was on this topic. Still working thru the equations, but I see that in my Kansas climate it would be more effective to focus on orientation of the house, relook glazing area, placement and quality, sealing air leaks and a well rounded insulation plan for all parts of the house.

We will install Pella aluminum clad wood windows, casements for egress where required and fixed elsewhere. Most are triple paned with mechanical internal shades between the middle and inner glazing. U 27 to U 33. We plan to seal tight as a drum.   

Dana, I'll ck on the "uptick" for thicker Logix block and their Platinum Series (PS). PS is a regular thickness form, 2.75" each side, which uses, a new type of foam with "... millions of tiny graphite infrared absorbers and heat reflectors that reduce thermal conductivity." to get an R27.

Thanks for the info so far...keep it going. Dave

They seem to be very short on hard data when it comes to Neopor foam. When the best description uses "up to", be cautious.

Posted By Lee Dodge on 21 Dec 2011 07:22 PM
And if you add light-blocking shades with side seals to the windows in Dana's example and assume that they are closed 60% of the time (night-time) in winter, then the effective R-value for the windows increases from 3.6 (equiv. to U = 0.28) to 3.6 + 60% * 2.0 = 4.8, or U = 0.208, and the heat transfer through the windows is decreased to 9.4 Btu/hr F, and the whole wall drops to 25.4 Btu/hr F, a further reduction of heat losses by another 11%. Since it is colder at night than during the day, and the R=2.0 value is kind of a worst case, then the actual saving should be higher. The shades are expensive enough (~$100 a piece for windows of this size) that they are probably not economically justifiable unless you need some type of shades or curtains for privacy, but many times you do.

The quality of the side & bottom seals are very important, both from an energy efficiency point of view, as well as condensation issues.  A simple curtain may provide more creature-comfort for humans standing near the window, but by lower the temp of the window it provides a convective duct & and the force to drive the convection loop, and actually INCREASES energy use, as well as vastly increased condensation on the windows in winter.

For a decent discussion of the heating-season implications of different window enhancements, see:  http://cchrc.org/docs/reports/windo..._final.pdf