Truss Roof Assembly for an ICF House
I am looking to build a ICF house and sub it out myself. I like the advantages of roof trusses and like the idea of parallel chord trusses on ICF houses because the concrete creates a firm attachment point that is relatively resist to push out forces that these trusses could be prone to. I also like the idea of an over roof or umbrella roof that deals with water and an under roof with an air barrier, although I am open to its criticisms. A cor-a-vant would be in the soffit and a ridge vent would be included. This is my back of the napkin/SketchUp schematic of a roof assembly and I would appreciate feedback. Feedback could range from obvious problems in design to wasteful expense. The house will go on a hilltop above the Arkansas river in Tulsa Oklahoma, so the disaster resistance features of ICF will be advantageous. The roof system will either be standing seam or Tesla. The actual truss web design, other than it being parallel, would be done by the truss manufacture. Please pick it apart if you have time.
Thanks
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
this is zoomed in a little
jason_stratten,
- You don't need or want that foam board layer. It impedes drying into the vented channel above, adds very little R-value and makes the assembly harder to build.
- If you are planning to completely fill the cavities with cellulose, I would suggest a layer of sheathing on top of the trusses as you have shown, (or permeable roof underlay, which saves one layer of sheathing, and gives improved drying), then frame the vent channels with 2"x4"s on the flat, sheathed with plywood above.
https://www.greenbuildingadvisor.com/article/building-a-vaulted-high-performance-and-foam-free-roof-assembly
- If your air-barrier is the underside of the trusses, there is no need for spray foam over the walls.
- One small point: You don't ever want to rely on the ICF walls to resist outward forces from the trusses. Those loads need to be either taken up by a ridge beam, or by the design of the trusses themselves.
Thanks for the feedback.
Point 1. I see your point and will consider.
Point 2. I don't like the idea of having the air barrier on the inside of the cellulose. I was planning on having zip on the outside of the trusses and retaining mesh on the underside that is designed for cellulose. I just feel that this will do a better job protecting the cellulose from changing humidity and nature in general. In your opinion, is it okay to leave exposed zip as the bottom of the air channel and exposed OSB or plywood as the top surface of the air channel(top sheathing)with high temp ice and water shield above that?
Point 3. I agree, but in Tulsa I really don't like the idea of of all that cellulose essentially open to outside air. Also, the field mice here are an issue. I'll think it over and let me know if I am misunderstanding something.
Point 4. Fair, I will have the trusses engineered and the house approved by an engineer so if I need further tension cables or something for outward pressure it should be accounted for.
Thanks for your advice.
Jason_stratten,
Point #2. One weakness of above sheathing vent channels is that any moisture in the cellulose has to diffuse though the lower sheathing before it can be vented away. Zip is only about 2 perms. That's not as forgiving as my second suggestion of using an underlayment, but yes your stack-up should work.
Good luck with your build!
The above deck venting also won't meet code (at least here), the vent channel needs to be formed under the roof deck.
As Malcolm suggested earlier, a sarking membrane is very common install on the other side of the pond, I would look at those details ie:
https://www.proctorgroup.com/products/roofshield
If your code does allow above deck vent, I would not built it with OSB. Plywood has much higher permeability and I would add still add a largish diffusion vent near the peak:
https://buildingscience.com/documents/guides-and-manuals/gm-2101-guide-building-conditioned-unvented-attics-and-unconditioned
thanks for the feedback.
seeking clarification on a few issues. The assembly is top to bottom:
1-metal/tesla roof
2-ice and water shield(will be required by most roofers)
3-osb/plywood- I take your point that Plywood would be better with water vapor although everything should dry to the vent space.
4-vent space from eaves to ridge(2x4s standing or flat)
4.5-ridged foam(which i think i agree with Malcolm, and will remove from the plan)
5-zip system taped to create a air barrier to the house
6-20 inches of cellulose between parallel cord trusses.
7-net to hold the cellulose
8-1.5" air space
9- 5/8 drywall
I think it will meet code here because the vent isn't even required and a hot roof would meet code. I think the vent is under the roof deck but maybe I am misunderstanding something.
My fear of having an air barrier at the bottom of the trusses and a fabric/membrane at the top comes from my location in a long grass prairie. If there is a space, a field mouse will eventually find its way in. I worry a bit about them chewing through the zip, and that would be a huge stinking problem, but with a membrane over cellulose, if they get into the vent space, they will find the perfect nest. maybe I am overreacting, but my current builder grade house is filthy with them and a constant chore.
I was planning a ridge vent and core-a-vent at the eaves or something similar.
My idea is that zip down dries to the house, zip up dries to the vent space. That vent space when combined with the Tulsa sun will have excellent drying capacity.
Thanks for the comments and any more feedback would be appreciated,
Jason
Jason_stratton,
I don't know any codes that allow un-vented roofs with permeable insulation like cellulose, although you can use vapour diffusion ports in climate zones 1, 2 and 3.
The worry is that having the Zip in-between the truss space and the vent channel, and no interior air-barrier, means the cellulose may take on more moisture from below that it can exhaust by diffusion. I don't know enough about y0ur climate to say if that is a big concern.
I guess I'm confused. The way I see it, the trust space is essentially open to house air, complete with multiple heat pumps and a dedicated dehumidifier. The lowest perm material on the bottom side of the truss space would be latex paint. In the north, north of me, you need to worry about house air condensing out of the air on cold days inside the wall cavity. This is largely not an issue in places that only get to freezing a couple times a year, but the original idea behind the Ridged foam was to have the condensing point, happen inside the foam, this would seem to eliminate the issue and seemed like a belt and suspenders approach to me since we don't have a frost line, don't use frost free bibs, and unfortunately often don't even consider the possibility of a freeze when building a house (A complete mistake). A common wall system here is to use something like vapor impervious aluma flash, then OSB, then fluffy insulation of some type and then drywall.
Thanks for the feedback, i appreciate it.
Jason
This video has Matt and positive energy discussing vapor impermeable house wraps.
https://www.youtube.com/watch?v=xPXcnioVBFw
It would have been helpful if you'd mentioned the climate info earlier. Aren't there parts of Oklahoma that get cold in the winter?
I did mention the location. All of Oklahoma is zone 2 or 3. It can get cold here. That is why I criticized the fact that most builders here don't even think about freezing. When Texas froze solid 2 winters ago(including Matt's area) we did as well. I had a copper pipe against the sheathing, in my builder's grade house, that lead to a pipe burst causing 15k in damages. Honestly, Oklahoma is rife with bad building science. The majority of the houses have blow-in on the attic floor and ACs and furnaces in the attic space with ducts. This works because energy is cheap and and the inside to outside delta is always lower in hot climates. Better building methods here tends to be closed cell against the roof sheathing. Closed cell is easy, and effective because it creates your vapor and water barrier. I was hoping for something a bit smarter , but maybe I am outsmarting myself 😏.
My plan looks a lot like 1. in the below article.
https://www.greenbuildingadvisor.com/article/sandwiching-roof-sheathing-two-impermeable-layers
Improper unvented assemblies are a problem in cold climate, in warmer climate it is much easier.
In zone 2 or 3 you can go with a simple unvented assembly with fluffy insulation and a diffusion vent at the ridge as per the link earlier. You don't need any rigid or vent channels above, this just adds complication. Unvented assemblies also hold up much better in extreme weather events if you are in an area with very high wind.
If you don't like the diffusion vented idea, your initial plan of a bit of exterior rigid will also work. In your climate you only need R7 above the roof for condensation control for your 20" of dense pack (table 1 bellow):
https://buildingscience.com/documents/building-science-insights-newsletters/bsi-100-hybrid-assemblies
Again, no point for any top venting, top venting is only needed in very heavy snow country.
I agree. On a hilltop in Tulsa, wind is always a problem. The scrub oak that is up there, looks like it has been through a war.
Is there a reason the article doesn't talk about cellulose and specifically only talks about fiberglass or mineral wool? I am undecided between fiberglass and cellulose.
Either way, I think I am more comfortable with 1.5 inch ridged, compared to vented vapor diffusion. If I go with an unvented assembly with the ridged, do I still need the ports?
I am still uncertain on a couple of the details. The first is how much eaves to add. looks wise, there wouldn't be eaves, and weather wise eaves are probably in issue for winds, but a benefit for water. I like the idea of the umbrella roof assembly(a term I borrowed from Matt Risinger) for air tightness, because with zip and tape you can create a really tight box without eaves. probably cheaper to go with raised heel truss and bring the spray foam up to cover the eave connection, use ridged on top. Lots of people seem to put standing seam on diagonal spacers anyway for venting below the metal. If i go with a tesla roof they will likely want a flat substrate and their own ice and water shield.