Screws as condensation points
Maybe I’m over thinking again…
When 4″ of XPS is held in place with 5.5″ screws many screws will end up going through the sheathing (missing the studs).
Another case I am thinking of is when I install a flat roof with 3.5″ of polyiso screwed down.
Then a membrane and 1/4″ polyiso (special high compression strength polyiso) glued down (R1). So screw heads have R1 protecting from cold – like standing snow. In this case there is about 1 screw per 1 sq ft.
At 1 screw per 1 sq ft, the cross sectional area of the screws is only about 1/3000 of the assembly.
The cross section of the screws is small enough to not significantly affect the overall R value of the assembly, but don’t the protruding screw tips provide condensation points?
If they do provide tiny condensation points is this a problem and why or why not?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Mark,
what is your location?
have you seen this?
http://sustainable.cchrc.org/docs/REMOTE_Manual.pdf
see page 36
Thanks John. So I am not over thinking!
On page 36 it shows the exposed screw tips being covered from the interior. But the link shows new construction (or gut retrofit).
This is not an option for me, because it is a retrofit, and the whole point is to avoid damaging the interior finish.
Suggestions?
I'm on Long Island (winter to 0F summer to 90F and plenty of rain).
Mark,
suggestions?....wooden screws...just kidding
At least you are not in Fairbanks......
I do not think they mentioned your concern in this Retrofit article
https://www.greenbuildingadvisor.com/homes/old-house-gets-superinsulation-retrofit
the article may still be interesting for you if you have not seen it
That retrofit was closer to a gut retrofit. So they had the option of capping any protruding screws from the interior.
(Yes I've read that one)
I cannot touch the interior finish.
One mitigating factor that I have is that rather than just fiberglass batts in the 2x4 wall I have Air Krete too. But that doesn't help me for the roof which will have screws at 1 per sq ft and cannot possibly hit the rafters all the time.
I like the wooden screws suggestion! I'll take 500 #12 6" long please ;-)
I've thought about this quite a bit and have not yet come up with an easy to install solution.
About the least bad idea that I've had is to counter sink the screw heads deeply into the furrings (which are only 3/4 or 1" thick) and then put a dab of spray foam over the heads.
So I'm still looking for a good solution.
Could the paper facing (vapor retarder) on the fiberglass batts be my savior here by reducing vapor flow into the assembly enough to stop condensation? After all risk of condensation occurring shouldn't be much more than on the current sheathing.
However, if there is condensation on the screw tips the kraft paper will make it harder to dry out.
(Now my two threads overlap so posted in both)
I looked up thermal conductivity of steel, and it is more than an order of magnitude less than copper, but nearly two orders more than wood.
So what would really satisfy me....
1) 3D thermal analysis of a steel screw in 4" of XPS with a 40F temperature gradient
2) Experimental data
3) Data from houses build this way, where moisture sensors were located close to screws.
I'm not likely to get either by tomorrow.
So here is what I did...
My own less than scientific experiment, but enough to give a good indication...
1) Put a 4" steel screw across two cups in my kitchen, fill one cup with ice water (~32F), lean head of the screw into ice water.
2) Put a second 4" steel screw into 3" of XPS with the head in a cup of ice water and the tip protruding 1/4".
3) Cook leftover turkey soup in kitchen with the extractor fan on low, and within 8' of screws. This is a worst case humidity level, and maybe unrealistic, but the roof in question is in a kitchen area. Note it is important that the soup is turkey soup.
4) Eat soup.
5) Go back and check screw tips for condensation and second cup for any drips.
So far using an infrared temperature probe I can clearly see a temperature gradient in the screw without XPS. Also I can see condensation within 2" of the head, but there is only a glossy sheen of water, growing to nearly enough to form 1 drip after an hour. Whereas the outside of the ceramic cup with the 32F water is soaked. After nearly 2 hours the condensation is over 3" of this screw, but still not enough for 1 drip to form and hard to see in the 3rd inch. This appears to be a steady state.
The tip of the screw embedded in 3" of XPS has clearly cooled, but is much closer to the temperature on the warm side of the XPS than the cold side (measured with temp probe) it has no condensation so far on the tip.
Here is what I think based on what I've read and my kitchen counter experiment.
* In my climate (Long Island average low 20F in Jan) this might be a problem if only 2" of XPS is used.
* With 4" of XPS it should not be a problem in this climate, especially in non kitchen / bathroom areas.
* In an extreme climate like Alaska it might be a problem if only 4" of XPS was used. Thus the suggestion that I referenced ( http://sustainable.cchrc.org/docs/REMOTE_Manual.pdf - p36 ) of capping the protruding screw tips, if you have access to interior of sheathing, this is a good idea in Alaska.
Thanks to all those that chimed in.
- Mark
Love your experiment Mark. I've been watching this closely, but since I haven't had anything to add to this conversation, I've stayed out so far. We're interested because we're doing some very similar retrofits in Maine (7,500 DD), so it could be a real issue for us. We're not Alaska, but we're colder than you.
THERM should be able to analyze this condition, but I haven't learned how to use it with any confidence yet. Anyone else?
Jesse,
I got the most helpful replies over on
http://forums.jlconline.com/forums/showthread.php?t=49255&page=2
Especially from Thorsten, Pete Engle, John B, and NW Architect - great thanks to them.
I think that the average low temperature for January is more important that heating degree days.
I managed to get some very low VOC adhesive for the roof, so I'm going fasten 16" OC into rafters for the field, and where I should have more like 12" oc for perimeter (for better wind uplift) I will still use 16" oc but some glue also. If we miss the rafters, a bit I won't worry or make life hard for the crew since I have R1 protection for the screw heads and my climate should be just fine.
For the walls we will aim for the studs too, but again won't make life hard for the crew if there are a few missed, I won't worry about missed studs because with 4" in my climate it won't be a problem. In the case of screws that are backed out I'll either use foam in a can or tape to patch any holes from the outside
Rather than software simulation -- which can be quite excellent, you could repeat my experiment for temperatures down to about 0F.
Now I'm quite confident that with 4" of XPS things are good down to 20F average low, and maybe a little below that. If you want to confirm for yourself for lower temperatures you can use a salt solution rather than water.
To do this set your freezer to a little below your average low temp for January. Mix up a strong salt water solution in a paper cup or other container. Brine freezes at 0F. Chill the salt solution down the the desired temperature, and use it just like I did but with salt solution instead of iced water. For me with an average low of about 21F I thought 32F water would do, but for 10F or 0F you might prefer to use the salt solution. If your freezer is already set below your average low, just keep testing the salt solution until it reaches the desired temperature.
Of course if you wait until January there will be other means to do the experiment.
Also do remember to cook _turkey_ soup!
If you try this please post the results in the forum (and email me mark at net zero energy dot org )
Mark,
I applaud your geekosity. I've wondered the same. We're currently doing an essentially new house (reusing some of the foundation), so we'll have the chance to view missed screws from the inside. We opted for 3" ship-lap poly-iso for the walls above grade, followed by Tyvek, vertical strapping & siding.
For the roof we're going with 2-layers of 2" XPS taped + staggered on top of the sheathing, followed by high temp ice+water, then 2-layers of x-furring to provide venting & anchor for a metal roof. The 1st layer of furring is parallel to the pitch & hopefully embedded in the TJI rafters at 24" OC. Again we'll have the chance to view from the interior attic.
A reminder to the crew: mark w/ sharpie from the exterior where strapping/furring screws miss the studs/rafters - it will then be easier to find those locations on the inside.
If indoor RH is controlled and there is a reasonable air/vapor retarder on the interior of the wall and ceiling assemblies, then condensation should not be an issue.
The real issue is the significant degredation of insulating value of foam when penetrated by hundreds of nails, particularly those that are not embedded in wood framing, since steel is about 1000x as thermally conductive as rigid foam.
Using adhesive between framing members is a good option.