Has anybody got experience of attaching roof trusses to ICF walls where the external foam is extra thick?

To get good R values I am going to have the outside layer of foam much thicker than the inside one.  The outside foam will probably be six and a half inches thick.  The concrete will be six inches and then the inside foam will be two and a half inches.  Ideally, I would like to have the top of the wallplate flush with the top of the foam.

My understanding is that the truss has to sit on the wallplate where the bottom chord joins the top chord (or at least very close to it).  The picture below tries to show the set-up.

 
                        TTTTTTT
                      TTTTTTT
                   TTTTTTT
              TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT
          TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT
      TTTTTTFFFFFF wwwww  FF
   TTTTTTT FFFFFF ccccccc FF
TTTTTTT    FFFFFF ccccccc FF
                FFFFFF ccccccc FF
                FFFFFF ccccccc FF

T - Truss
W - Wallplate
F - Foam
C - Concrete

The roof has a 45 degree pitch and the outside foam is obviously going to get in the way of top chord of the truss as it comes down over the eaves.  The logical thing to do would seem to be to cut channels in the outside foam to allow the attic trusses to sit properly.

Is this the standard way to do it?

edit - Forgot to say that the block will probably be Nudura (if this makes any difference).

hello aa_uk. I think you're going overkill on t your foam thickness, but no matter. Just take your plans to your truss people and have them design a raised-heel truss with load bearing designed for where it is needed in the truss. Shouldn't cost much more, & you'll get a step towards the efficiency you seek. FM

UK friend, I'm impressed with your graphical tools -- or at least your creativity in description . Anyway, i would just cut the slots in the foam and not mess with redesign cost and etc.

I must agree with Colorado however that you probably don't need the foam thickness whereever you are (in the UK? North Pole?). Is this building going to have any windows or doors? Keep in mind that although you may choose to increase your r-value off the chart you are dealing with a series of weakest links and heat is going to find them. How, exactly are you doing your attic (or roof) and your windows and doors.

First let me say that my hat's off to you since your going ICF so sorry 'bout the rant. Second, I have a more or less similar "conversation" with builders here....

FYI:  Around the world, R-values are given in SI units, typically square-metre kelvins per watt or m²·K/W (or equivalently to m²·°C/W). In the United States customary units, R-values are given in units of ft²·°F·h/Btu.

So... r-value is about what heat gets through (or lack of heat - known as cold).  If you had a mystical r-1000 material between your 2x4 studs of your house you will still have 15%-20% of your home performing to the r-value of wood due to the studs, headers and etc. Thus your average r-value would be about 800. Sounds good ---- but 20% of your structure is still performing to about r-1 per inch or approximately r-4 at best (close to that of a very high end window). So when you do the math you'll find that your building still can't beat an ICF performanc of continuous r-22 plus.

Why bore you with this story? What is your plan is for the rest of the building -- windows and roof. Unless your windows and roof are "way up there" and they don't allow connections inside to out with wood or steel --- you may may be wasting your money. Basically, ICF as it stands (typically around r-22 or so) already puts the burden right back on the roof in a big way and on the windows in an even bigger way.

If you are trying to meet code in the UK (and i have read a thing or two about some new extreme rules for r-value in the UK - is that correct??? ) --- don't feel bad, our government and rule makers in the area of thermal performance, etc. don't and haven't understood, r-value very well either (or to be fair, choose not to understand). Thermo is not just about conduction. Regards.

Thanks very much for your comments. Cutting the foam is my least preferred option (there's enough to do already!) so I will talk to the architect and truss company about the raised heel trusses. The truss design isn't done yet so if the additional cost isn't too great then I will go that way.

The rules for r-values have got quite a bit tighter here over the last few years and they are going to get a lot tighter still. The current building regulation is that your walls have to be r-21 or better. In 2013 that will change to about r-31 and by 2016 the government wants all new builds to be "zero carbon".

My personal goal is just to try and get my heating bills as low as possible! I live in Shetland which is a group of islands about 120 miles off the north coast of Scotland (lat 60 N) and it does get a bit wild in the winter. We don't get really, really cold temperatures, just lots of wind.

I am aiming at about r-38 for the floor, walls and roof. The floor will be a concrete slab on top of about a foot of EPS. The roof will probably be attic trusses with about 6 inches of PIR foam between the rafters and another 2 inches underneath. To be honest I would prefer a SIPS roof but I think that it would cost about 50% more than trusses.

Anyway, thanks again for your very helpful comments.

Andrew,
Have your trusses mfgr for cantilever equal to the thickness of the outside foam. It is no problem for the truss people to do this, and the added cost should only be a couple dollars (pounds?, euros?) per unit. When dealing with the truss people, be sure that you are both talking about the same dimensions (out to out of foam, or out to out of load bearing) They should be able to provide a detailed drawing of the truss before final ordering. Look it over carefully before proceeding with the order.

Thanks very much, Wes. (and it's still pounds here!)

Hi aa_UK. As Texas pointed out, once you have great insulation your windows will be your greatest weakness, after air leakage. Your walls and roof should not have an air leakage issue, just be certain that where they meet is well detailed. Your building will almost certainly require an ERV or maybe an HRV; at least plan for one so that installation is easy. FM

Thanks very much. It's going to have an HRV system.

Agreed, insulation is just part of the solution to windy and cold - you need to put extra effort into sealing. For example, house wrap or taped foam outside plus internal plastic vapor barrier to create two air sealed surfaces. Then do blower door testing. Consider thermal window shutters too.

Posted By jonr on 05 Apr 2010 06:31 PM
house wrap or taped foam outside plus internal plastic vapor barrier to create two air sealed surfaces.

Oh, not good, but not all bad either. You do not want sealed surfaces, instead you want vapor permeable surfaces. House wrap is vapor permeable, plastic is not. With today's understanding of vapor movement through, and moisture entrapment in, walls, the conclusion is you want a vapor barrier only in the coldest regions of Northern US and Canada. But I'm not sure you want that with an ICF wall even in cold Canada. Good discussions of this issue are at buildingscience.com. You do want a good air barrier. House wrap outside and wall board and latex paint inside provides that. It's a given that all gaps penetrating the wall need to be sealed.  Air tightness is one of the biggest advantages of ICF.

The problem with two sealed surfaces is you create an almost tight container. With daily temperature swings the pressure inside the container rises and falls and you end up with a vapor pump. During the cool phase air is drawn into the container as the internal pressure drops. This draws in air with higher humidity than during the hot phase. During the hot phase pressure increases in the container and expels air. The problem is the air inside the container is warmer than during the cold phase and thus holds a greater amount of vapor. Subsequently the system tends to always pull in more vapor than it expels. You eventually reach saturation during the cold phase and moisture condenses. The condensed moisture doesn't evaporate completely before the next cold phase starts to bring in more vapor.

The more humid the climate the greater the problem. In the humid south even vinyl wall paper is sufficiently vapor tight to cause mold to build up in the wall immediately behind it. aa_uk is surrounded by water so I would guess he is in a fairly high humid climate, although the humidity may not be terribly high in the cold winter. In any case, the last thing he would want is to create a vapor trap.

Another option, try using a raised heel truss OR bottom chord bearing with the bottom chord serving as the soffit which allows vertical clearance over the wall for insulation and a connector such as this one:

http://www.strongtie.com/products/connectors/H-LTA1_MASONRY.asp There are many similar to this that we use in the Caribbean high wind zones, find one that matches your engineering requirements

and eliminate the top wood plate altogether.

Thanks very much, atitagain.  I will speak to Simpson.

As far as vapour barriers go, I am intending to put one in the roof.  From the top, construction will be:

Slates
Wooden battens (horizontal - to attach the slates to)
Wooden counterbattens (vertical - to stop water build up behind the horizontal battens)
Breatheable membrane
Sarking board (Sheets of plywood nailed to the rafters to provide racking strength)
6 inches insulation between rafters (with 2 inch air gap between the top of the insulation and the sarking boards)
2 inches insulation under rafters
vapour barrier
drywall
plaster skim coat

Given that I will have an HRV system I am not too worried about moisture build up on the house.  Our local building control group are adamant that you must have a 2 inch air gap between the top of the insulation and the bottom of the roofing boards to stop moisture build up in the roof so that should keep that side of the vapour barrier dry.  On the inside there is just the drywall and I am confident that the HRV system will do its job.

Thanks for all your comments.