GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Air seal conjoined trusses?

tundracycle | Posted in General Questions on

Some of the lines of melting snow and ice dams on our roof (previous post here: appear to be along trusses and in particularly double or triple conjoined trusses. The top cords in most of these are 2×4 and 2×6 with a few 2×8 and infrequently 2×10.

The top cords of all trusses appear to be fully enclosed with closed-cell foam so thermal bridging would seem to not be significant.

Our builder believes the problem with these (approx 7 of 13 lines of melting snow) is warm moist indoor air being pushed up between the conjoined trusses. He says that these are known to be difficult/impossible to air seal due to the gussets making it impossible to join the trusses together tight enough and that expansion and contraction will cause caulk or spray foam to crack and so not seal them.

First photo is a triple. The second is the bottom cord of two double conjoined trusses.

Thoughts? Ideas?


GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. Expert Member
    BILL WICHERS | | #1

    I’d caulk them. Use a good urethane caulk or acoustic caulk between them prior to assembly. Since I think your framing is done already, all you can do is caulk any exposed gaps. Urethane caulk is pretty durable, and others on here have had good luck with the “big stretch” caulk. It’s recommended to air seal between sistered rafters so the trusses should be treated similarly.


    1. tundracycle | | #2

      I believe they did use something on them when they joined them together but not sure what. I'll have to check on that. Our builder is of the opinion that nothing works to seal these or similar conjoined framing members in walls.

      1. Expert Member
        BILL WICHERS | | #3

        They might have just used construction adhesive to make structural bond between them. It’s common to do that when sistering joists too — I’m doing that right now myself to stiffen up a bouncy floor in preparation for tiling! :-)

        I think construction adhesive between trusses that are also fastened together with nails will probably make a pretty decent air seal. I’m positive that acoustic caulk would make a good air seal since it stays gooey so some slight movement will have no effect on its sealing abilities over time. I’m pretty confident that polyurethane caulk would last too because it is tough stuff. Caulk is about your only option if the trusses are already up though, unless you bury the entire upper chord of the truss in spray foam to seal any possible air pathways in or out.


        1. tundracycle | | #8

          Thanks Bill. The builder said that they used construction adhesive on them but that anything more gooey would drip down and make a mess.

  2. benwolk | | #4

    Either caulk the joints or use a good air-sealing tape such as SIGA Rissan/Wigluv or Proclima Tescon Vana and run that over the face of the truss members. You can get a wide tape that will bridge the entire ganged truss. I would pair this with Prosoco R-Guard Joint and Seam Filler at the sharp interior corners and where the trusses meet the spray foam. You can slather the joint and seam filler all over the intersection between the spray foam and them trusses and then tape over the top of that and run a nice line of tape right down the members of the truss. Those products will move with the truss and won't crack or split to ensure that you have a complete air seal for the life of the house.

    1. tundracycle | | #7

      Thanks Ben. How will this work where there is a bit of foam covering the top cords?

      1. benwolk | | #12

        The liquid flashing can bridge any of those gaps and if you spread it over any cracks it will seal those and seal to the foam. You can use a thick nap paint roller to roll on Prosoco Fast Flash, but you have to trowel on their joint and seam filler product.

  3. gary__b | | #5

    If the builder's theory is correct, it'd seem that melting could only occur where a conjoined truss aligns with a seam in the roof sheathing. Obviously you can't tell exactly where those seams are because of the foam, but in cases where you have multiple conjoined trusses near each other, it should be pretty easy to tell (if they all have it, they can't all have sheathing seams).

    That the top chords are enclosed in foam makes it really unlikely to me (not an expert) this is air leakage caused. The apparently (from pics) slight coverage of the top chord with foam should stop air, but there's probably still decent thermal bridging (if there's a half inch of foam, you've only added ~R3 to the wood). I'm guessing you had it right.

    1. benwolk | | #13

      I agree with Gary, W Ramsay. The issue is probably more likely thermal bridging through the joists since the path of least resistance is still directly through the top chord of the trusses. You only have R-3.5 to 5.5 at those locations even if you have R-30 in the bays. Your roof looks like a standard wall when frosted over. Try fastening some rigid foam insulation over the bottom of the top chords of a number of trusses and see if your issue goes away. Or get your hands on a thermal camera if you can to properly diagnose the location of the thermal bridging.

  4. tundracycle | | #6

    Here's the truss engineering for the triple truss in my first photo. I believe the photo is the portion to the right. It's a 2x8 top cord so 7.25" actual. There is Durovent in the rafter bays which is 1.25" so Durovent + code minimum 7" foam s/b about 8.25" or 1" beyond the top cord?

    Does it look like 1" beyond the top cord in the photo? To me it looks like there may be 1" over the top cord above the gusset but only about level with the top cord in the bay? Below the gusset looks like maybe 1/8" over the top cord as you can clearly see the 3 trusses?

    I'm hoping to crawl in there this morning and probe to the top cord to get a measurement.


    1. tundracycle | | #14

      Just measured and it’s mostly about 1/4” over the 2x8 so we’re closer to 6” closed cell rather than 7”.

  5. MattJF | | #9

    It would really only be the exposed components leading into the foam that can leak. How about a small holes, like 3/16", to create some space in the seams. Then polyurethane caulk injected into the hole to create a dam, and caulk the exposed seams up to the foam.

    I would blower door to verify this is a meaningful leak path, but I have mentioned the blower door and fog before...

    1. Expert Member
      BILL WICHERS | | #10

      I think you'll have a pretty minimal amount of leakage through the assembly you show, especially if they builders used construction adhesive between the members. The leakage path is long, and small, which won't allow much air to move past.

      I have had some success using a foam gun (the great stuff "PRO" gun), and the small plastic tips to inject foam into little voids in places like this. Stick the tip as far in as you can to the biggest part of the gap, then hold down the trigger to inject foam as you slowly pull the tip out. The foam is orange, and will expand in the void so that you can see where it goes. Sometimes this is easier than caulking, but it does need a big enough gap to fit the tip in. Have some paper towels handy too since some foam always dribbles out as you extract the tip from the void in the wood.


      1. MattJF | | #11

        I agree it seems like it should be a small leak there. That is why I think a blower door would be useful. It seems like there is a good chance there is leakage into the soffit or something big like that. Pulling some of the soffit vent if easy and looking at the blocking that was done might be useful.

        What does the truss design looking like for the second photo?

        1. tundracycle | | #15

          I’m at the site now. I’ll post that truss as soon as I get home.

  6. 730d | | #16

    Is that attic conditioned space? Is the attic part of the building envelope ?
    How warm is it up there ?

    1. tundracycle | | #18

      The attic is finished conditioned space within the building envelope. There is an open stairway going up to there from the second level and heat rises... The house is currently kept at about 55-60°f but the loft is usually around 65-70.

      The HVAC system that serves the loft (and south end of the 1st & 2nd levels) isn't running yet. Once it is then temps throughout the house should stabilize.

      1. tundracycle | | #19

        Open stairwell from the loft looking down to the 2nd level:

        1. 730d | | #20

          I like Ben Wolk #13 idea of adding some test foam to a part of it.
          Also, I would poly your stairway of and drop the temp up there, as much as you can if only as a test. Getting lots of new snow today and Saturday. Good time to test.

          1. tundracycle | | #21

            Hi Mike, what would the purpose of cooling it down up there be? I'd guess that we'd definitely see less melting on the roof but I don't know that that would tell us much about the cause?

      2. tundracycle | | #22

        BTW, this evening the loft was 60°f, about the same temp as the rest of the house. There is a small electric heater that on the second level near the loft stairs that they use occasionally and when that is on the loft will heat up quite a bit but apparently not otherwise.

  7. tundracycle | | #17

    I believe this is the two doubles in the second photo. The photo is of the portion to the right.

    1. 730d | | #23

      Its a bit intimidating here as there are a lot of very knowledgeable bright people here. I think your temp will be getting warmer up there when your HVAC is complete. You need more separation of the inside temps and exterior roof temp. The warmer it is up there the worse you melt. Also, I don't trust your ventilation is adequate. It is installed as much to exhaust heat as it is to keep things dry.

      1. 730d | | #24

        I meant roof ventilation. Between the insulation and roof sheeting.

  8. tundracycle | | #25

    Followup to Matt on blower door.

    We have done three blower door tests for a total of 6 days.

    1) After insulation and prior to rock. DPIS (Eric Boyd, Eagan MN) filled the house with smoke and then pumped up the pressure and stated that they could not find any leaks. They then brought it neg pressure (I believe -80pa IIRC) and couldn't find any leaks so stated that it looked good to sheetrock. We asked them for CFM50 and ACH50 tests and ACH calc came back at 4.8 ACH50. So much for their no leaks... And our confidence in them.

    2) The insulator (who previously bragged that he would get us to below 1.0 ACH50) spent three days with his blower door, smoke and IR. He stated that he had reduced it to 1.8 ACH50 which was the best that he could do and that for any further reduction we'd need to Aeroseal which we determined (thanks to Dana) was not worth the cost.

    3) Spent a day with another energy person and their blower door, smoke and IR cam two weeks ago specifically trying to solve the roof melt problem. He found two leaks that we filled with foam but these did not fix the roof melt problem (more below). He ran a CFM50 and got 1918 and calc'd a 1.36 ACH50 which is about what I would expect given the 1.8 we'd gotten prior to sheetrock.

    However, when I calc'd ACH50 I got 1.9, not 1.36. I learned that he had used the building volume calculation from an earlier test which used floor to floor to calculate volume rather than floor to ceiling so there is some confusion about how to calculate volume.

    In any case, with all of these blower door/smoke/ir tests we've not found the source of our problems.

  9. tundracycle | | #26

    As I noted above we spent a day (10 Dec) with an energy consultant (plus blower door, smoke, ir cams) trying to solve our roof melt problems. This is a guy who we determined was good at running basic blower door tests and finding common frequently seen problems but not a building science person.

    We cleared the roof of snow and then got a new 5" to see that the problems persist. Most seem to be in line with trusses though I've not yet been able to determine that for sure. If trusses then two potential sources of problem; 1) warm indoor air running up between conjoined trusses which is what our builder believes is the problem or 2) thermal bridging.

    As seen above and discussed in my R-42 vs R-49 thread there are some trusses with the top cord somewhat exposed (covered with only 1/8" or less foam). To what extent might 1" or more of closed cell foam fully encasing these top cords reduce thermal bridging or indoor air penetration to the sheathing? (note that had the insulation contractor done the code required 7" of foam the top cord in the photo above would have had at least 1" over it.)


    1. MattJF | | #27

      Thanks for the update on the blower door work. It is disappointing that they were not able to ID anything to fix. The only tricks I know of are to crank the pressure way up and have sufficient smoke density, which might be an issue in a house of your size.

      After checking everything obvious, I think we circle back to the question do you really have a problem? All hot roofs will develop some ice dams under certain weather conditions. It is really just a matter of do they get bad enough to lead to other problems.

      On my roof, which is R51 in the bays and R17 at the rafters, I still get notable "striping" at the framing. My neighbors' roofs, which are cathedral and something like R30 fiberglass squished with some sort or partial venting, they melt out much faster than my roof. They don't get ice damns of any concern though because they actually totally melt out regularly enough with our snow loads.

  10. tundracycle | | #28

    I did a post on the issues:

    I'd appreciate any comments on anything I got wrong and of course any further recommendations on correcting this. I'm pushing our builder to bring in someone with more knowledge, such as someone from BSC, so any recommendations on who might be good would also be welcomed.

  11. CMObuilds | | #29

    Im in central WI, so same climate. It seems to me this isnt an air leakage problem as much as a thermal problem. You cant put a thin layer of foam over top chords and expect it to perform, too much heat transfer to the roof deck and the snow starts to melt, that in my opinion is why I would never build using conditioned attic if I can help it. Houses like yours always have these details and they seem to always battle problems like this due to the designs.
    Possible you have install issues with the spray foam given small channels for interior air to contact your roof deck.
    Given the size and shape of your house, your ACH 50 number is very good.

    1. Expert Member
      MALCOLM TAYLOR | | #30


      I agree.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |