GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Audio Play Icon Headphones Icon Plus Icon Minus Icon Check Icon Print Icon Picture icon Single Arrow Icon Double Arrow Icon Hamburger Icon TV Icon Close Icon Sorted Hamburger/Search Icon

Community and Q&A

Airtight Sheathing & Thermally Isolated Double-Stud Walls?

Daniel Ernst | Posted in Energy Efficiency and Durability on

Has anyone out there designed / built a thermally isolated double-stud wall AND used the airtight sheathing approach in a TWO-STORY residential project?

I’m thinking of the two-story, rectangular box as the thermal envelope. It gets wrapped on all five sides with plywood sheathing, which acts as the primary air barrier. Then it gets capped with an unoccupied, fully vented, stick built roof. Make sense?

Now, combine that concept with a curtain wall—to eliminate the thermal bridges in the framing. That’s the idea anyway.

I know of examples that use airtight sheathing combined with “interior” double-stud walls, but in those houses, the exterior walls support the floor and roof frames. In these designs, the floor joists and rim joists act as thermal bridges—and they are located where they are most exposed to moisture accumulation and deterioration.

Perhaps the thermal bridges can be modeled to quantify the heat loss, but the “durability risk” is a little murky.

See this link for an example from some GBA advisors and regular forum participants (architects Jesse Thompson & Phil Kaplan, builder Dan Kolbert):

Back to a thermally isolated frame w/ airtight sheathing:

How would you treat the junction between the balloon framed curtain wall and the platform framed structure (account for the differential shrinkage in the wall heights)? How would you reliably tie the wall sheathing to the attic floor sheathing?

Is this idea worth pursuing? Or is it too much trouble for the added benefit?

This is for a mixed-humid climate zone (4A, in close proximity to zone 5).

1) Eliminates (reduces) thermal bridging.
2) Reduces the chance of wintertime moisture accumulation in critical rim joist area / frame.
3) Airtight sheathing throttles solar vapor drive in the summer.
4) Eliminates fuss of ADA methods (i.e. acoustical sealant), theoretically a lower ACH50 result.

1) New construction methodology—unintended consequences?
2) Airtight sheathing approach consumes more resources.

Thanks for the discussion!

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. Riversong | | #1


    I think this is an idea well worth pursuing, and I don't understand why double-wall builders are still penetrating the thermal envelope with floor framing.

    In the link you provide, the inner stud wall bears only on the 4" of slab-edge insulation and hence has no load-bearing capacity. It's easy enough to remedy that, either by pouring a wider grade beam (and creatively hiding the top of the slab-edge foam) or by using the ThermoMass foundation system which can place up to 4" of XPS midline inside a 12" thick concrete walll (4"/4"/4" CIC), which is better yet.

    Differential shrinkage can be resolved by stacking the same cross-grain elements in both inner and outer wall (in other words, a band joist between both storeys in both walls), and the storey walls can be cavity isolated from each other while providing floor shear continuity to the outer walls by plywood double top plates or by extending the subfloor to the outer band joist.

    1) Eliminates (reduces) thermal bridging.

    2) Reduces the chance of wintertime moisture accumulation in critical rim joist area / frame.
    This is an issue if there is no interior air barrier.

    3) Airtight sheathing throttles solar vapor drive in the summer.
    No it doesn't. Airtight sheathing only stops air movement, not vapor diffusion which is the mode of moisture transport when the sun strikes the cladding. And, of course, neither does a vapor-open WRB.

    4) Eliminates fuss of ADA methods (i.e. acoustical sealant), theoretically a lower ACH50 result.
    Yes, but places the air barrier on the wrong side during the heating season when exfiltration into the thermal envelope can cause condensation (such as at band joists). Air channels through the thermal envelope don't have to terminate outside - they can be internal loops from floor to floor or electrical outlet to electrical outlet through wiring chases.

    1) New construction methodology---unintended consequences?
    There's nothing unusual or difficult about this strategy. The primary unintended consequences of double wall systems are 1) excessive materials, 2) ambiguous load paths, 3) hidden thermal bridging (such as floor joists) and 4) differential shrinkage.

    2) Airtight sheathing approach consumes more resources.
    Perhaps, but if the attic sheathing can do double duty as the floor of storage space that's accessed from an airtight pull-down stair, "hayloft" doors on the gable ends (with, perhaps, an exterior staircase that doubles as a second floor fire escape), then the only issue would be sufficient insulation depth below that air barrier layer.

    I did that once by cross-hatching 2x8 joists 24" oc in each direction, but that's a lot of materials. It could also be done with site built or factory parallel-chord trusses or deep TJIs.

    I've done several house plans that, while not incorporating an exterior air barrier, create a double wall without second floor thermal bridging and could easily be modified for an exterior barrier approach.

  2. Riversong | | #2

    In reading through the JLC article on the Kolbert House, a few things stand out.

    The first is that the relative impermeability of the Huber Zip Wall (less than 1), combined with taped seams and, in this case, an unvented roof covered with self-adhering bituminous membrane makes this a house assembly that cannot dry to the exterior.

    The second, and very revealing, point is that with all the taping and caulking and impermeable air barrier materials, the house tested at 2.2 ACH50 before insulation and at 0.77 ACH50 after insulation and drywall. This clearly demonstrates that the exterior air barrier is mediocre, while the lions share of air sealing was done by the dense-pack cellulose and the drywall (even with no special efforts to make the drywall an air barrier.

    And the house ends up with a wrong-side vapor barrier with very little drying in the dominant direction, and an unvented roof that cannot dry and is sheathed with an OSB product that the builder admitted had already experienced some edge swelling during construction.

    To top it off, his HVAC contractor insisted on an ERV rather than HRV, which will return moisture to the interior environment in the winter when it's critical to remove it in a very tight smallish house.

    I also wonder about the decision to save money by installing very minimal 2-loop radiant heat system (tubing 4' away from slab edge and irregularly laid out) on the first floor only. I'd be surprised if there was uniform warmth throughout the house.

    The house incorporates a small solar thermal system (supplemented by electric), but doesn't appear to be well-designed to take advantage of passive solar heat. Some south and west windows are shaded with fixed awnings that look like afterthoughts and some are not. Judging by the sun shadows, the south windows are undersized and over shaded, while the west windows are oversized and undershaded.

    I've never grokked the two dissimilar siding aesthetic (except the Queen Anne shingled gables), but this house really throws a curve by piercing the transition with various height windows.

  3. TJ Elder | | #3


    I have a project in the works that takes a similar approach to what you describe. Plywood sheathing is required structurally for lateral loading, and it serves as air barrier. Walls are not double framed in the sense of a single combined volume of insulation, because the sheathing wraps the inner (loadbearing) structural framing.

    Walls have 2x6 studs, the roof has 2x10 joists, and roof sheathing joins the wall sheathing without interruption. Above the roof joists there's 2x8 overframing, which frames the overhangs and allows additional insulation exterior of the sheathing. Exterior insulation is rockwool, 5" at the roof and 3.5" for the walls. Siding goes over furring that's supported with long screws through the rockwool using small wood spacer blocks.

    There are no vapor impermeable layers, so all assemblies can dry in both directions. There's venting below the siding and below the upper roof sheathing (which supports the roof membrane but is secondary to the lower air barrier sheathing). Thermal bridging is well controlled in the walls, as the rockwool has minimal interruption and adds R-14 over the structural frame. At the roof, the 2x8 overframing mostly runs perpendicular to the structural framing beneath, so that helps minimize thermal bridging. At the foundation, the deep rockwool layer continues over the slab edge (switching to a version intended for use below grade).

    This is also for climate zone 4 (PNW), about 4500 HDD. Walls total R-34 and the roof is R-54. This includes cellulose in the 2x6 walls and 2x10 roof framing. A significant fraction of the total R-value is exterior to the sheathing and the structural frame, which should help with longevity of the structure. Floors and roofs have beams for the longer spans so the 2x10 joists don't span more than 10'-0".

  4. Interested Onlooker | | #4

    To carry this approach to its (logical?) limit would see a construction in which the inner and outer frames were separate entities linked only at the penetrations for windows and external doors.

    The external frame would support the dead loads of roof and snow and react the wind loadings on walls and roof. It would be braced in shear by sheathing which would support the WRB and siding and roof coverings. The sheathing might serve as an air barrier if this were thought advisable.

    The inner frame would support the ceilings and floor and the internal live loads. It could be braced in shear by shear panels of cross-bracing. The gap between inner and outer shells would be filled by insulation and an internal air barrier, if preferred, could be established by an air-tight drywall detail.

    If the foundation were of the Thermomass type in which internal and external leaves are separated by insulation then the separation of functions and structures could continue right to the foundations.

  5. Interested Onlooker | | #5

    shear panels or cross-bracing...

  6. Daniel Ernst | | #6

    Robert - Thanks for the response. A few points:

    Regarding the solar vapor drive, plywood does not stop diffusion, but it does retard diffusion. If it's considered a vapor semi-permeable material (anywhere from 1 - 10 perms, depending on its moisture content), why wouldn't it throttle the summertime solar vapor drive? I'm thinking about some typical weather patterns, summer showers followed by a baking sun . . .

    Since this is for a mixed-climate zone, I'm trying to balance the summer vs. winter vapor drives. The airtight sheathing approach SEEMS like a good idea. Dense packed cellulose should retard any convective loops in the double-stud wall cavity.

    On the JLC article, it appears that Kolbert used ZIP sheathing for walls and roof, but it wasn't wrapped completely around the enclosure (note the blocking used between the rafters). That may account for the initial ACH50 reading.

    I don't plan to use OSB, and I expect to have a completely open and vented attic (a little cramped for anything but storage, so yes, the attic floor sheathing could provide a benefit there).

    I'm playing with some variations on the framing cross-section, trying to make it effective AND simple. ;) Thanks again.

  7. Daniel Ernst | | #7

    Thomas - Thanks for the input and updates.

    Regarding your rockwool outsulation project, did you figure out your support system for the 2 x 4 "furring strips" used to hold the siding and provide the exterior insulation cavity? Will you be adding another layer of sheathing to this assembly? Innie or outtie windows?

    Do you have a blog, or other place to share pictures when you start? It would be great if you could share your progress!

  8. John Brooks | | #8

    Daniel and All.
    A lot of good thoughts and questions on this thread.

    I would like to see more drawings posted or linked to.

  9. Daniel Ernst | | #9


    Not sure about your need for anonymity . . . but, yes, that's the question. Is it logical? Is it worth the extra effort and resources?

    What you say makes sense. The idea of separating structure and skin is similar to concepts used by timber framer Ted Benson. He's taken that concept from his timber framing operation into his modular building program.

    In "How Buildings Learn," Stewart Brand explores this concept in depth, discussing the different layers of a building with architects Frank Duffy and Chris Alexander:

    Brand further divides and prioritizes Duffy's "shearing layers" into six S's:

    Site (Eternal)
    Structure (Long Lasting)
    Skin (Changed in as little as 20 years)
    Services (7 - 15 years)
    Space (As little as 3 years)
    Stuff (Every day!)

    It's an interesting concept.

    As a society, we often think of buildings as assets. Really they are liabilities (albeit necessary). The trick is to build one that is less a liability---with regards to efficiency, durability, maintenance, and adaptability.

    Separating the layers might just do this very thing . . .

  10. Interested Onlooker | | #10

    Both sets of framing would come under Brand's heading of 'structure' and be regarded as 'long lasting. However, the outer frame, sheathing and roof constitutes a weather-proof entity so if they could be left undisturbed then the possibility might exist of radical modifications to the inner frame, floors and ceilings without compromising the outer shell.
    And you could carry out the work in the warm and dry...
    This might argue in favor of an air barrier on the outer shell rather than ADA but I also find persuasive Robert Riversong's argument that the visibility and fixability of an ADA is important .

  11. Daniel Ernst | | #11

    John - Right now I don't have anything set-up for posting drawings, but I cansend you an AutoCAD file of some variations---if you're interested. Email: djncu(@) . . .( remove the parenthesis).

  12. TJ Elder | | #12

    Daniel, to answer your question about rockwool on the walls--the strategy is to use 3.5" thick rockwool in 24" wide panels oriented upright. The wall framing follows a consistent 24" horizontal module. Furring goes over the rockwool rather than fitting between, to minimize cutting the rockwool and to cover the joints. The furring is 5/4 x 4 laid flat, basically a deck plank, and is attached through a section of 2x4 that's set within the rockwool. This requires a 7" long SIP screw to get good embedment in the 2x6 stud. The 2x4 spacer prevents crushing the rockwool, which is a complication for using rockwool rather than a less compressible plastic foam board.

    There's only one layer of wall sheathing, over the structural frame and under the rockwool. This is one of the main reasons for using rockwool, that it goes inside a deep rainscreen cavity (4.75" from sheathing to siding). This avoids having to add sheathing and then add more furring strips for a code-required rainscreen gap. In this case the gap is 1.25", the thickness of the furring strips. There may be some loss of thermal value due to wind washing, but the site is not windy.

    Windows are "innie", attached to the structural frame and without plywood window boxes. Window detailing really isn't complicated because everything exterior to the sheathing is effectively part of the cladding system. It wouldn't make sense to run peel & stick over the rockwool at the jambs, for example, and instead the window simply seals at the sheathing like an outie window. The siding (cedar T&G) returns at the jambs like an outside corner and back-vents into the rockwool. There is a metal profile at the sill going all the way past the siding, and z-trim at the head. I will document all this during construction, and a blog sounds like a good idea.

  13. Riversong | | #13

    Regarding the solar vapor drive, plywood does not stop diffusion, but it does retard diffusion... why wouldn't it throttle the summertime solar vapor drive?

    Perhaps we have a different understanding of the term "throttle". I think of it as "strangle" or close off completely.

    But, while winter vapor diffusion drive and liquid diffusion drives are in opposite directions and relatively weak (which is why we no longer consider winter vapor diffusion a significant problem), in summer all moisture drives are inward and significantly higher, particularly when the sun comes out after a soaking rain.

    Both liquid diffusion and vapor diffusion are temperature driven (vapor diffusion is a function of vapor pressure which is a combined factor of temperature and humidity).

    Typical winter outward vapor drive might be close to 900 Pa. If the cladding is saturated and warms to 120° in the summer sun, the inward vapor drive could be almost 3000 Pa and at 140° almost 4000 Pa.

    Placing insulation behind the cladding limits heat transfer inward and so keeps the cladding at a higher temperature when the sun shines.

    Inward summer vapor drive becomes more problematic with the use of very low perm interior skins and even more so with AC and cool inside surfaces. Plus summer condensation is more problematic because ambient temperatures are perfect for rapid mold growth.

  14. Riversong | | #14

    And I should add that an unvented drainscreen can become an oven (or, more accurately, a sauna) in the summer. Though I'm not much of a fan of rainscreens, if they are used they should be fully vented.

  15. Riversong | | #15

    everything exterior to the sheathing is effectively part of the cladding system

    Unless you're considering the 3.5 inches of mineral wool as nothing more than a thick drainage membrane, then it's part of the thermal envelope not the cladding system.

    I would continue to caution you about your SIPS screw and spacer block system. Though it's not carrying any structural or even window loads, it still places the entire cladding system in cantilever to the structural wall and small, spool-like spacers may not be sufficient to resist the shear over time. With no bottom support, the cladding assembly may creep over time as both the spacers and the framing they're tied to shrink or compress across the grain.

  16. Garth Sproule | | #16

    Robrt & Thomas
    A detail I have been thinking about would be to attach the furring to the rafter tails using some sort ofappropriate metal straps for connection. This would effectively transfer the load of the siding to the stud walls. of course, this requires that the rafters align with the studs...not sure what to do on gable walls though...

  17. TJ Elder | | #17


    There would also be a discontinuity at window openings, so you'd need to hang furring strips from below the sills. What I have in mind for the 2x4 spacer is to run the grain vertically like a discontinuous stud. To avoid sagging over time, maybe there should be additional support at the top of each furring strip. That could be an angle bracket or a longer segment of spacer block with two screws.

  18. Riversong | | #18


    I'm not sure exactly what system you're referring to now, but when I build my Riversong Truss wall, in addition to all vertical parts of the exoskeleton bearing directly on the extended foundation (whether wide grade beam or 12" thick ThermoMass wall), I side nail them to the rafter tails at the eaves (which are offset from the studs) and to a nailer attached to the underside of the lookout rafters that support and are perpendicular to the gable fly rafters.

  19. Garth Sproule | | #19

    I was referring to furring over outsulation. What you describe for gables would easily work here too. Nice detail.
    Unless your windows are very wide, I don't see much discontinuity of load, as the siding below the window, would normally be spanning the window opening, thereby transferring its load to the furring adjacent to the openings.

  20. Garth Sproule | | #20

    One more thought...if the furring could be securely attached at the roof framing, would spacers even be necessary??

  21. TJ Elder | | #21


    In my scheme the furring is just a 5/4 board laid flat over the rockwool, so it would get noodly to hang more than a few feet. If you were talking about a 2x4 oriented like a stud, then it could span a floor height and the siding would brace the weak axis. That would be a curtain wall approach.

  22. TJ Elder | | #22

    Garth, to respond to #19, you'd want to support the siding with the furring and not vice versa. Any window more than 22.5" wide will interrupt at least one furring strip, and that needs to hang from below the sill. Start with all the furring securely attached to the structure, then add the siding.

  23. Riversong | | #23

    The funny thing about supporting the strapping with the overlying siding is that this is what happens with structural framing on some old buildings, which seem to continue to stand just out of habit and stubbornness.

    I dismantled an two-storey old carriage house in Boston years ago for material resuse, and it was disconcerting how unstable the frame became as we removed sheathing boards from roof and sidewalls.

    I also jacked up the back wall of another old, but attached, carriage house on the side of a New England farmhouse some years back during a December snowstorm to replace the rotting sills. I discovered that the wall had no framing at all, but was just plank on plank with offset joints (the planks being wide and about 1¼" thick).

    Not only that but, at sometime, two 3'x5' rectangles were cut out of the wall for double-glazed picture windows and then a second floor for bedrooms and bathroom was built on top of it. The wall, with all that weight on top, was slowly sagging into the rotting sill, and the picture windows were sunk 3/8" into their pine sills, apparently the primary support for the second floor, and without breaking their seals.

  24. User avatar GBA Editor
    Martin Holladay | | #24

    Okay, I figured out how to to this. Dick Russell sent me these two details, along with the following message:

    "Riversong replied that there is no real reason the joists must go all the way out to the outer wall. In post #8, John Brooks said he'd like to see more drawings. I would have posted a pic on the joist part of the discussion, but I see no way to do it on the forum. I've attached a couple of drawings showing this detail, or at least the way I did it on my new house. Feel free to use them.
    Dick Russell"

    Here are the illustrations he sent:

  25. David Meiland | | #25

    Do you need the I joist on top of the 2x4 wall, and can you move the rim to the inside?

  26. John Brooks | | #26

    We certainly need illustrations in order to discuss stuff like this.
    Do you have comments about this drawing or is it just a test?

  27. John Brooks | | #27

    I am having a hard time visualizing your "system"
    Have you actually built this way?
    do you have any photos?
    Or wall sections?
    It sounds to me like you have all the detailing headaches of the other outsulation stratagies only you are trying to use mineral wool in place of foam?

  28. User avatar GBA Editor
    Martin Holladay | | #28

    Check out the information above the illustrations -- I just edited my post.

  29. John Brooks | | #29

    Thanks Martin & Dick
    Dick, nice drawings
    do you have a drawing of your wall to roof detail?
    Do you have a firestop between the first floor & the second floor?

  30. Riversong | | #30


    Can the rest of us attach illustrations? That would make this forum far more useful.

  31. User avatar GBA Editor
    Martin Holladay | | #31

    Unfortunately, GBA does not yet have the resources to re-write our Web site software to permit photo posting. It's a feature I'd love to have. Some day we will.

    In the meantime, GBA readers are always welcome to send illustrations to me at my e-mail address:
    martin [at] greenbuildingadvisor [dot] com

    To the extent that my schedule allows, I'll do my best to post photos or drawings when GBA readers ask for my help. If the volume of requests becomes too great, I may not be able to fulfill all requests, but I'll do my best.

  32. Riversong | | #32

    Dick's details are very similar to what I've designed for several clients, but I would suggest a couple changes.

    The 1x3 gussets would not be necessary unless the two walls were assembled together on the deck and raised as a unit (which would be very heavy), and standard 1x3 spruce strapping is too narrow, thin and brittle to take that many nails without splitting. The only place such a gusset would be necessary in his details are where the exterior deck puts a torsion force on the 2x6 studs, but 3/4" CDX or 2x4s would be more appropriate. Also, if the gussets are cut to the full width of the wall assembly, they are useful to control spacing and would provide a wider nailing area.

    But the other detail I don't like is the Maine Deck Brackets. These are aluminum I-beam sections, with the outer flange shorter than the inner, to space a deck away from the house for drainage. They require that the WRB and sheathing be interrupted because the bracket must be bolted directly to framing (with either through bolts or lags into 6" of solid wood!), creating an unnecessary penetration through all layers of exterior weather protection that is sealed with caulk. Also, the distance that these brackets cantilever the deck puts a considerable torque on the wall which requires engineering to properly resist.

    The practice of spacing exterior deck ledgers from the house is one, like the rainscreen, that I believe is overused and often unnecessary. A PT ledger bolted directly to the sheathing (which requires nothing more than Ledgerlok screws) can be properly flashed, just like any other interruption in the cladding, with WRB over lapping the upper flashing leg. This makes a far more secure and more weathertight junction than either washered lag bolts or Maine brackets. The best flashing I've found for the ledger is Rhino Flashing, which is a heavy duty plastic Z-flashing that is impervious to all PT compounds (and virtually everything else).

  33. User avatar GBA Editor
    Martin Holladay | | #33

    From Johns Brooks:
    Johns calls this "In Progress Detail posted only for discussion"

    Click to enlarge photo

  34. User avatar GBA Editor
    Martin Holladay | | #34

    From Interested Onlooker:

    The attached illustration shows a concept for a double-stud wall construction in which the two stud frames are entirely separate except where they are linked by window reveals and external doorways.

    The gap between sheathing and drywall would be filled with dense-pack cellulose - hence no firestop within the cavity.

    The top plate of the lower 2x6 stud also acts as the toe plate for the upper 2x6 stud (the studs are in line) and its position up the wall is set solely by the length of the lower stud.

    The outer frame carries the roof and snow dead loads and the wind loading on the roof and exterior walls.

    The inner frame carries the internal live loads of the household and supports the upper floors, ceilings, drywall etc.

    Both frames require shear bracing. The outer is braced by the sheathing. The inner would need to be braced by shear panels, X-bracing or sheathing. The shear panels or sheathing would need to be arranged so that they did not impede the cellulose fill process.

    The air control layer could be at either the outer sheathing or the inner drywall. If it were placed at the outer sheathing then its effectiveness could be checked (and altered if necessary) before the inner drywall and cellulose were put in place. Placing it at the inner drywall would be more fiddly but would prevent winter exfiltration into the cavity and could be easily checked and repaired.

    I would value any comments (pro or con) that GBA forum folk might care to give.
    Interested Onlooker

    Click to enlarge photo

  35. John Brooks | | #35

    I. Onlooker,
    your detail is much like Dick's except it dose not tie together at the floor deck.
    The outer studs if not full height would have "hinge effect"
    Have you worked on a wall-to-roof detail?

  36. John Brooks | | #36

    My original thoughts were that with airtight sheathing there would really be little need for airtight drywall as well.
    I have since changed my opinion and now believe that it would be wise to also make the drywall "tight" too.

  37. Riversong | | #37


    A couple problems with placing the outer wall plate at an arbitrary height is:
    1) that you'll have to build the wall one stud at a time rather than the typical raising from the deck fully sheathed,
    2) and the sheathing has to break on the plate for shear bracing - there will not be enough nailing surface for two courses of sheathing to meet at a single wall plate (CA now requires all shear joint framing to be solid 3x).

    If the outer and inner walls are attached at each floor or ceiling level (by even ½" CDX), then there is no need for interior wall shear bracing. All interior loads are vertical gravity loads, except the dynamic live loads which can cause shimmy. Drywall is more than enough to brace interior bearing walls, but the floor membrane is what resists lateral wind loading in the mid-transverse plane of the exterior walls, so it has to be tied to the outer walls.

    Either a CDX double top plate beneath the floor assembly or an extension of the subfloor above the floor framing will work to create this lateral shear connection.

    Also, separating the lower and upper wall cavities will make it easier for most cellulose installers to get proper density, and will create a better firestop between stories.

  38. Riversong | | #38

    I should add that the easiest double wall tying option for most conventional framers to accept is to have the outer wall top plate at the same height as the second floor framing so that the subfloor can extent out to make the connection. This allows the second storey walls to be framed, sheathed and erected from the deck as is typically done. But it will require very long (tall) sheathing panels to span from sill to top plate on the first floor walls or intermediate blocking.

    With the wall tie at the first floor double top plate plane, 4x9 or 4x10 plywood would suffice, but the second storey wall would have to be dropped in place as it's raised.

  39. John Brooks | | #39

    Robert R,
    Do you think that floor deck and densepak cellulose alone would meet code as firestop?

  40. John Brooks | | #40

    I have never seen a 3x4 or a 3x6
    I know they are used in CA
    They would be useful at panel joints for shear bracing and airtightness.
    Concerning airtight sheathing how about using "sill-seal" as a gasket?

  41. Riversong | | #41


    National Fiber, the cellulose company I use, had their material third-party tested and approved as a firestop.

    "National Fiber’s cellulose insulation has been approved as a fire blocking material under Section 708.2.1, Item 1, of the UBC, Section 716.2.1 of the IBC, and is permitted as an alternate to the fire blocking in Section R602.8, Item 1, of the IRC by Omega Point Laboratories, Report for Project No. 16094-11638. When installed in a dry or spray application to a depth of 14.5 inches, cellulose outperforms conventional wood fire blocking in fire blocking tests."

    As for sill seal as a sheathing gasket, I don't think the standard bubble-type sill seal is that good at stopping air movement to the degree that we now aspire to, and it would not compress enough for a flat joint.

    EPDM gaskets, I believe, are much more reliable and durable, but they too would create a lump. The one time I used ADA EPDM gaskets (from Conservation Technology) under my exterior wall plates, I didn't realize how much they raised the walls until I rolled joists from the outer walls to a center bearing wall (without gaskets) and had to shim every joist at the center wall.

    I don't know of any product that works as well as Tremco, except perhaps a good subfloor adhesive like PL Premium urethane, which is excellent at gap bridging but makes disassembly or re-adjustment a nightmare.

  42. John Brooks | | #42

    I don't know of any product that works as well as Tremco, except perhaps a good subfloor adhesive like PL Premium urethane, which is excellent at gap bridging but makes disassembly or re-adjustment a nightmare.

    Robert, are you talking about the expansion gap?
    I remember a builder warning me about allowing construction adhesives in "the gap" of exterior sheathing because the construction adhesive would cure hard and not compress and thus defeat "the gap".

  43. Dick Russell | | #43

    Regarding those two diagrams Martin posted for me, the gussets between inner and outer walls were something I threw in early on, not knowing how the framers might prefer to do things. As it turned out, the outer walls were framed and erected first, followed by the inner walls, so there are no gusset ties.

    The diagrams pertain to the connection between the two stories where there is framing on the lower level wall. The house is set on a hill, with full-height foundation walls on the uphill side and part way downhill on the sides. The downhill side is totally framed, walkout, as is half of one side. The rest of the sides are pony walls.

    There are double sills all around the foundation (inner and outer). Where there is full-height foundation, the upper level outer wall rests on an the outer sill, and the floor joist rim will be insulated as part of the upper level wall cavity. Where there is framing above and below the floor, as in the two diagrams, the floor joist rim will be insulated as the top part of the lower wall cavity.

    Originally, the outer wall was envisioned as 2x4, overhanging the foundation by 2" to cover the exterior foundation foam layer. The inner wall was thought to be the bearing wall for the floor and roof. Late in the game, consideration of deck attachment and selection of the Maine deck brackets for use led to the creation of an "outer rim" on the walls where there is deck attached, for mounting the brackets. With this rim let into the studs, to keep the sheathing plane uninterrupted, use of a 2x4 stud would have left only 2" of stud width at the top of the wall, not a lot of meat. The engineer suggested making the outer wall perimeter all 2x6 and have it bear the roof load. There were some lateral shear concerns as well, although I'm not sure they were valid. Anyway, that's the way the walls finally went up.

    At the upper level wall/attic floor joint, there was supposed to be a 3/4" plywood plate connecting the inner and outer wall top plates, with the ceiling joists resting on and beyond this. Somehow this tie plate got left out, creating a 3" gap between top plates. The joists serve to keep the top of the inner wall in place.

    David (post #25), were you asking about the two diagrams, or something else? If about my diagrams, I'm not sure what you are suggesting.

  44. Daniel Ernst | | #44

    Thanks to everyone that contributed to this discussion (and Martin for posting the drawings).

    I've been playing with some iterations very similar to Dick's 2nd option, but resting the outer "rim joist" on a top plate, same as the interior wall. Then the subfloor sits on top and ties both rim joists together. This method would allow both walls to shrink evenly. My main concern is whether or not the outer rim joist is supported well enough without blocking (which would defeat the point by creating a thermal bridge).

    On Dick's drawing, is there any concern that the let-in rim joist would shrink compared to the remaining stub of 2 X 6 below the top plate?

    Interested Onlooker - Differential shrinkage is my primary concern with this method. If you take this concept up to the roof by drawing three levels of floor / ceiling joists, then you'll see what I mean (assumes a crawl space - 1st Floor, 2nd Floor, Attic Floor joists).

    The only other option I've discovered is to balloon frame the outside wall---use using a single stud to span the entire wall height. This would connect to an attic floor joist (and join the wall and attic floor sheathing together), but it would remain separate from the platform framed structure. In essence, you would have a ceiling joist below the attic floor joist --- taking the parallel wall concept into the attic area. The only thing tying the frames together would be gussets.

    The outer wall would support the roof, the inner wall the floors. With this method you end up with repetitive headers, but it separates the frames more completely. Material intensive?

    Again - thanks for the discussion and input!

  45. Interested Onlooker | | #45

    Many thanks to everybody for their helpful sharing of experience and expertise and also to Martin for posting the drawings. I rely on GBA to educate me in the nuances of green building so that I can speak to architect and builder with at least some clue of what I'm talking about. As ever, the answers have thrown up more questions for which I beg your indulgence.

    #35 ....your detail is much like Dick's except it dose not tie together at the floor deck.
    The outer studs if not full height would have "hinge effect"

    Yes, I had overlooked that. Duh.

    It is becoming clear that what is needed is some detail that links the outer studs to the inner studs and which :

    transfers the loads which push or pull the face of the outer wall inwards or outwards and

    does not transfer loads (e.g. differential shrinkage) which act in the plane of the outer wall and

    does not constitute too much of a thermal bridge.

    I know how I would do that in my day job (aerospace) but what is best low-tech solution?
    #37 ....2) and the sheathing has to break on the plate for shear bracing - there will not be enough nailing surface for two courses of sheathing to meet at a single wall plate.

    I can see that would be a problem for the horizontal joints. How is it usually addressed at the vertical joints where two panels of sheathing meet at a single stud?

  46. User avatar GBA Editor
    Martin Holladay | | #46

    From Daniel Ernst:

    "Some different iterations for the 'airtight sheathing and thermally isolated' wall Q&A session."

    Click to enlarge photo

  47. John Brooks | | #47

    glad you posted drawings
    I am a visual person and I could not follow your written description without the visual aid
    I have the same problem with all of the written descriptions of complex wall/roof sections
    this is friggin great

    Option A(far left) is going to add 2 ft vertical compared to a normal house.
    An extra 10%+/- of siding, sheathing,studs ..yada..yada

    Option B (1 ft taller than normal) is not-so-bad except for the variation in shrinkage that you mentioned..hmm
    this differential shrinkage is complex with both sides of the thick wall/roof "seeing" extremely different thermal/moisture conditions

    Option C .. I see the structural weakness that you mentioned

  48. John Brooks | | #48

    Should we really worry about a thermal bridge such as the "floor deck"
    3/4" thermal bridge?
    plywood on edge has R-value ? eh?
    maybe not-so-bad?

  49. Riversong | | #49

    Robert, are you talking about the expansion gap?

    No, that comment was in the context of wall plate to deck air sealing.

    The expansion gap required around plywood and OSB sheathing cannot be filled, so tape remains the only "solution" to air sealing. But how many framers actually leave that gap, particularly those who are using the sheathing as the air barrier?

  50. Riversong | | #50

    A point of order (literally): Referring to post numbers is not necessarily helpful, since not all of us have the same numbering system. On my screen, I have the choice of ordering the posts from "newest to oldest" (which I do) or from "oldest to newest", and the posts are displayed 10 to a page with each page numbered from 1 to 10.

    That's why I use either the blockquote HTML tag to repost the question or statement I'm commenting on. This is sometimes a hassle if you don't get the formatting exactly right, but the other option is to simply "quote" the statement you're responding to.

  51. User avatar GBA Editor
    Martin Holladay | | #51

    From Thomas Jefferson:

    "The attached image shows the wall construction I described in the Airtight Sheathing & Thermally Isolated thread."

  52. Riversong | | #52

    My main concern is whether or not the outer rim joist is supported well enough without blocking (which would defeat the point by creating a thermal bridge).

    If the subfloor is extended out to the outer rim joist, then no blocking is required and the floor shear membrane is tied to the exterior walls to resist lateral wind loading. (And, no, that negligible amount of thermal bridging is a small price to pay for structural integrity).

    On Dick's drawing, is there any concern that the let-in rim joist would shrink compared to the remaining stub of 2 X 6 below the top plate?

    Of course it would. Best to use laminated rim joist material, the same as against the I-joists. A KD SPF 2x12, assuming it's installed at the 19% MC that is allowed for KD lumber at the time of milling, and equilibrates to 10% MC, will shrink ¼" across the grain.

    Differential shrinkage is my primary concern with this method. If you take this concept up to the roof by drawing three levels of floor / ceiling joists, then you'll see what I mean (assumes a crawl space - 1st Floor, 2nd Floor, Attic Floor joists).

    Differential shrinkage is a problem on a conventionally-framed house as well, and one that is often overlooked. A platform-framed house, with 2x12 1st floor joists, 2x10 second floor joists and 2x8 ceiling joists can shrink at the outer, load-bearing walls as much as 13/16", including cross-grain shrinkage of the plates. If there is a different stacking of structural members in the center bearing wall (such as a sub-joist built-up wooden beam), then the floors may end up sagging toward center (or vice versa).

    But the issue certainly becomes more important with a double exterior wall bearing different loads.

    The only other option I've discovered is to balloon frame the outside wall---use using a single stud to span the entire wall height.

    You'll be hard-pressed to find studs that long and code will not allow such unsupported spans for load-bearing vertical framing.

    That's why I settled on my Riversong (modified Larsen) Truss system, which puts all the loads (except wind) on the primary (inner) wall and then add an exterior truss chord, gussetted 24" oc, to support only siding and windows. Granted, this system won't work in coastal wind or seismic zones, but should be code-approved everywhere else. Because the exterior skeleton is not load-bearing (though I support it on the foundation and tie it into the rafter tails), it doesn't require continuous members, but can be assembled from shorter elements cleated together at the junctions.

  53. Riversong | | #53

    It is becoming clear that what is needed is some detail that links the outer studs to the inner studs and which :
    - transfers the loads which push or pull the face of the outer wall inwards or outwards and
    - does not transfer loads (e.g. differential shrinkage) which act in the plane of the outer wall and
    - does not constitute too much of a thermal bridge.

    For those who choose (or are required) to use 2x6 exterior framing and exterior sheathing, the least problematic option is to make the outer exterior wall roof-bearing (this transfers roof wind and unbalanced snow loads to the shear walls), make the inner exterior walls floor-bearing (requires no more shear bracing than drywall provides), and use either identical cross-grain lumber in both walls or engineered (no shrink) lumber in the inner wall, while tying the two walls together at the floor assemblies with the subfloor.

  54. Riversong | | #54


    The "additional" vertical height of the framing system is necessary in any case to get full insulation depth at the eaves. This requires either a rafter plate on top of ceiling joists (with appropriate metal tie downs) or some kind of "high-healed" truss.

    It's easy enough to compensate for the extra attic height by eliminating those wasteful and unnecessary 9' and 10' ceilings that seem to be all the rage. I use 8' ceilings on the first floor and sometimes 7½' ceilings on the bedroom floor, which makes a far more intimate and cozy space. That's a problem only with ceiling fans or chandeliers. I'll sometimes use no ceiling lights at all upstairs (to avoid additional ceiling penetrations) and simply wall sconces and switchable receptacles for table lamps. The side lighting makes the space even more intimate.

  55. Riversong | | #55


    Of your three options (I wish I could read the text), the approach in number one I believe makes the most sense but with some modifications. It makes sense because it isolates the roof and wind loads onto the outer exterior (sheathed) wall and places the floor loads on the inner exterior wall, while providing continuity of air barrier.

    But the balloon framing is not practical, because of length of lumber and difficulty in raising such a tall sheathed wall and the wall needs an intermediate shear member at the second floor level.

    So I would suggest plating the outer wall at the second floor deck height and tying it with the subfloor. This also makes insulating easier with some cavity isolation at least in the vertical dimension.

    You could either use lumber rim joists in both walls (as in option 3) or engineered lumber in the floor systems and eliminate the gussets, since the subfloor is sufficient lateral tie.

    To raise the rafters above the air barrier on a secondary plate would, of course, require engineered tension ties to prevent out-thrust. These could be metal straps tying the rafters back to the joists (but those would penetrate the air barrier), or additional rafter ties above (in the lower third of the attic height)) which might interfere with storage space, or some other creatively engineered solution.

    The upstairs ceiling would not require any structural connection to the rafter ties (joists) above but could be 2x4s if gusseted vertically to the upper joists or 2x6s if not (depending on span) to reduce the excessive and redundant materials. The double ceiling allows full insulation depth below the air barrier and avoids the need for vent baffles.

  56. Riversong | | #56


    Your wall system looks like it's solved a number of issues, but I wonder about how air-open the mineral wool batts are and how much R-value degradation there might be both from convection and from moisture.

    Of course, there are compromises with every system, it seems counterproductive to put a thermal layer outside both the air barrier and the weather barrier. You could use a secondary weather barrier between the batts and the strapping but that would probably make batt installation impossible.

    I would also caution you about tongue and groove siding. I don't think that's any more weather tight than ship-lap and it makes any vertical adjustment impossible. While that's difficult to do even with shiplap, it's not that hard to rework the lap if necessary (I've had to to that at some point in most applications).

    Perhaps the flat look is the aesthetic you want, but some kind of shadow line makes the cladding both more visually interesting and can offer better drainage, such as with pattern 105 novelty drop siding. I once had a load of novelty siding delivered with a mix of ship-lap and T&G and the T&G was much more problematic to work with, having to be driven together like a hardwood floor and having considerably less overlap than the ship-lap.

  57. TJ Elder | | #57

    Robert, as you say every system has to balance advantages and disadvantages. Thermal loss due to moisture should be near zero because sorption is similar to XPS (for Thermafiber RainBarrier). They make two densities, 3 pcf and 4.5 pcf, and it's possible the higher density has greater air flow resistance. I really can't say whether air flow is a problem thermally, but I'd guess you could add some thickness to compensate just like with loose fill cellulose in an attic.

    The system in my illustration avoids having cold sheathing and it requires less framing lumber than most double wall designs. Secondary sheathing and then extra rainscreen furring has been a sticking point and a reason to place rockwool within the rainscreen gap. I'll repeat that this has long been standard practice for cavity walls in Europe, where foam board insulation is less popular.

    I also prefer the single bearing frame with shear ply to the exterior. It works with an insulated slab edge and a simple 6" wide concrete stem wall.

  58. John Brooks | | #58

    Daniel & All,
    This is a damn good thread.

  59. Riversong | | #59


    Can you say "damn" on a public PG-13 rated thread?

  60. TJ Elder | | #60

    Here's an example of a "curtain wall" scheme by some folks I know:
    There's a 2" gap for a thermal break behind the outer 2x4 frame, and that frame stands on 4x6 blocking that's bolted into the foundation at the bottom.

    There is a brace at each floor level that includes plywood sandwiched into the outer frame like an extended subfloor, but it's not continuous with the interior subfloor. Some of the systems described here earlier suggested tying the two frames with the subfloor, but that would compromise the air barrier design.

    The assembly totals about R-41 at 14" overall thickness. My version would have about the thickness and R-value if the rockwool layer were increased to 5.5". Note that for either design the 2x6 frame and plywood are required for the structure, due to height and a small footprint in a seismic zone.

  61. Riversong | | #61

    The graphic you linked to claims R-36.

  62. TJ Elder | | #62

    The graphic you linked to claims R-36.

    It's 11" of cellulose. Maybe R-36 is supposed to be a whole-wall value.

  63. John Brooks | | #63

    Can you say "damn" on a public PG-13 rated thread?

    Robert, If I read between the lines
    you are saying that there is something different about this thread.
    There has been some extra effort to be polite and to tiptoe around other's feelings?
    Is that so bad?
    maybe it is because it somehow slows down the "flow"?

    I think it's a great thread because many people have contributed AND we have visual aids!
    It also happens to be my favorite subject "Airtight Sheathing"
    Yet now I am starting to question my own obsession.
    Maybe the primary airtightnes SHOULD be at the drywall as Robert has been drumming?

    I see that I am not the only "Rube Goldberg" here and "we" are going to extremes to make this "work"

    Maybe all this effort is justified in the very extreme climates?
    Is it really needed in North Texas (me) or Virginia(Daniel) or Scotland (I.Onlooker)?
    Thomas Elder what is your climate/city?
    Dick Russell ... your climate?... Garth ..Canada (extreme cold)?

  64. John Brooks | | #64

    I am starting to get a picture of your wall..
    The Field of the wall does not seem too-complicated
    But what happens when you get to the windows?
    Does someone have to field fabricate a piece of furniture around every door and window opening?
    (This is the problem I see with many of the BSC/BuildAmerica Details)

    And what happens at the wall to ceiling/roof?
    How do you "close" the bottom of the wall?

  65. Riversong | | #65

    Maybe all this effort is justified in the very extreme climates?

    Igloos have worked well for thousands of years. Quinzees, too. Skin tents in the summer.

  66. Garth Sproule | | #66

    If, like me, you believe that climate change is upon us, then yes, I think the efforts are not only justified, but morally obligated.

    I wonder how the people of the north dealt with that period of time each fall, when it was extremely cold but not enough snow was yet available to make their igloos...

  67. John Brooks | | #67

    you miss my point
    I believe in the effort
    I am guessing that your climate needs the extreme thermal break
    I am thinking my climate and Virginia ... maybe this is overkill?

  68. Dick Russell | | #68

    John, I'm in central NH, a zone 6 region, upwards of 7500 HDD. Naturally, my interests in building science are focused on this climate, and fortunately the building of a good wall for up here is fairly straightforward, or ought to be. I shudder to think about what folks on the gulf coast have to consider in building a wall with the possibility of 80 F dew point air outside and air conditioned low to mid 70s inside.

  69. TJ Elder | | #69

    John Brooks,

    To answer your earlier question, I haven't previously built the wall in the illustration. It's a design for a house in Portland, Oregon, to be built in the spring. The other detail that I linked is a passive house under construction here. They installed the exterior layer of cellulose and the skin while the interior is still a shell, so there's already a pretty good thermal enclosure for the job site. The owners are two architects and they are finishing the interior themselves.

    I'd like to think that rockwool outsulation makes detailing easier at wall openings and terminations compared to rigid foam, because there's not so much concern of water getting trapped behind it. The perm rating is high even though the moisture sorption is so low. It's not necessary to try and capture all the edges and joints with peel & stick membranes. I think of it kind of like a bear's fur. Of course I also haven't tried this so we'll have to see whether that still makes sense after some mockups.

    My design intent with the window details is to minimize interruption of the rockwool layer, so the siding basically folds into the opening at the jambs with a miter joint, same as the building corners. The furring strip adjacent to the opening just loses an inch of width. The sill is a deep metal profile, similar to what's often done with innie windows in Germany and Austria. The head has a strip of perf metal across the vent gap. At the foundation there's a continuation of thick rockwool going underground, and wool also folds over the (lower) roof sheathing. The roof is double sheathed, with the lower layer serving as air barrier and structural diaphragm, joining the wall sheathing around the perimeter. That joint can be taped, and there should be some benefit to having R-14 protecting the tape. Roof overframing allows overhangs without interrupting the air barrier, and creates space for outsulation over the top.

    Obviously more drawings are in order, so maybe I will have to set up a blog and post details and photos as they're available.

  70. Riversong | | #70

    I wonder how the people of the north dealt with that period of time each fall, when it was extremely cold but not enough snow was yet available to make their igloos...

    Farley Mowat, perhaps the first European to live with the Inuit through all the seasons, said that he discovered why the inland Eskimo's deer hide shelters were often so decrepit and full of holes - they wore their primary shelter on their bodies: two layers of deer skin, fur out over fur in (the original Polar Fleece).

    In fact, one of the reasons that both polar fleece and mukluks are now so popular is that Will Steger and Paul Schurke, who led the first unsupported dog-sled expedition to the North Pole in 1986, had fleece clothing and mukluks made by their wives (each of whom started businesses from this).

    I had the pleasure of leading dog-sled trips a few years later for Outward Bound with Steger's retired 10-year-old lead sled dog, the best dog I ever worked with (we carried fleece booties for the dogs, too).

  71. John Brooks | | #71

    is this sort-of-similar to the window detail you are thinking about?
    plus the mineral wool

  72. TJ Elder | | #72

    John, that metal sill is a good example, but the window is different because it has that high German frame profile for the sill piece to fold up against. There's another picture (somewhere) of a similar window mockup showing the sill profile more clearly, and the sill tucks under a metal lip on the window cladding. What's odd about that window is the exposed joint around the in-swing window leaf, because it looks like there'd be a dribble of water if you opened the window after a rain.

    A more important difference is that in the picture the innie window is set back into a double frame wall with cellulose, so it becomes essential to shed water forward around the window recess. That's where the innie becomes more difficult to flash. With my setup the drainage plane is where the asphalt felt is shown, at the sheathing. The window sits out an inch over that plane. It's not so pressing to shed water outward from the window when the drainage plane is somewhat behind the window.

    My version of the sill profile folds up under the window and gets a sealant joint between the metal sill and the fiberglass window frame. The profile has a few inches of vertical surface (below that up-turn under the window) against the sheathing where it's attached to the wall.

  73. Daniel Ernst | | #73


    Although I'm in Zone 4, data from NOAA shows the nearest town has 5,447 HDD. We're not New England, but that's still plenty. And the summertime humidity can be downright oppressive too ;)

    I guess I'm just sick of shoddily built 2 X 4 walls with poorly placed fiberglass insulation and NO air sealing measures whatsoever. These kinds of buildings create a huge future liability.

    There is a better way---and builders can create better homes without breaking the budget. But it does take knowledge, understanding, experience, and effort (a lot of study).

    ADA does have a number of benefits (visibility and repairability on top of the list). But it seems to me that creating the air barrier at the exterior provides more benefits for the southern climates, especially since the risk of wintertime moisture accumulation is a moot point. Case in point, go back to Robert's calculations of winter vs. summer vapor pressures. Or look at Martin's blog on solar vapor drive.

    Anyway, yes - a good discussion. Informative and on topic. Thanks to all who have contributed.

  74. Riversong | | #74


    The majority of summertime moisture transport into or through a wall assembly is in the form of water vapor, expedited by both high ambient humidity and solar radiant drive. So what is needed in a hot/humid climate is a vapor retarder on the exterior, not an air barrier.

    The WRB is designed to control wind-driven liquid moisture, and no other wind barrier should be necessary.

    Most routine air movement into or through a building envelope is from the constant stack effect pressures that occur in a heated building, proportional to the indoor/outdoor temperature difference and the building height. So the air barrier is primarily to protect from exfiltration during the heating season. For that reason, it makes more sense to have it on the inside in any climate.

    The primary difference in wall assembly order between cold, mixed and hot climates is in the location of the vapor retarder layer or the relative permeance of the various wall elements.

  75. John Brooks | | #75

    your climate is much colder than I had guessed
    It looks like I am the one with the mildest climate 2400 HDD 2600 CDD

  76. Riversong | | #76

    t looks like I am the one with the mildest climate 2400 HDD 2600 CDD

    Geez, you could live in the back of your truck in that climate.

    I've got 8500 HDD and don't worry about cooling. It's 27° and falling with a windchill of 17°.

    But my hot tub has a fire in its belly, I've got bourbon in mine and the moon is full, so life is good.

  77. User avatar GBA Editor
    Martin Holladay | | #77

    From Lucas Durand:

    "I thought I'd throw my hat into the ring with this idea I had for a cold-climate double wall envelope that uses air-tight sheathing. It is similar to something I will build in the spring. In this case air barrier continuity relies on seams taped with mastic tape and strategically located gaskets.

    "The weakest part of the AB assembly that I can see is where the ceiling sheathing butts up against the top plate of the load bearing wall. If the sheathing could be butted tightly against the top plate with a foam gasket between before nailing I thought this might make a tight seal. Tape could afterwards be applied over this same joint as insurance.

    "I'm curious what you all think of the way I drew the roof structure... The WRB makes a nice continuous transition to the wall, but obviously there is no real ventilation through the attic space beneath the WRB. It is a "cold roof" however in that there is ample ventilation between the WRB and roof tin."

  78. Riversong | | #78


    I think the top plate transition would be easier to detail with a double top plate and a gasket sandwiched between the two. But I wonder about the longevity of taped and mastic seams in plywood which requires room for expansion movement at the seams.

    And your outer wall, at least the second storey, would have to be built in place one stud at a time since each one is screwed to the roof truss.

    And then you're forced to use batt insulation, because blowing dense-pack would require penetrating the air barrier (I ran into this with my first true Larsen Truss house), and batts never fill the cavities like blown insulation.

  79. Lucas Durand | | #79

    Thanks for your comments.
    I wonder about the longevity issue of tape too. Would this not also be an issue with proprietary air-tight sheating systems too? How much better is the tape used on Zip systems?
    The idea of installing the studs one-by-one doesn't bother me too much since I tend to think of them more like truss members than studs...
    Wall insulation (in my case) will probably be mineral-wool batts since I cannot find a trustworthy cellulose contractor in my area. If I was more confident in my technique I would consider doing dense pack cellulose myself... Roof insulation will be loose-fill cellulose however.
    Do you think that there is a ventilation issue in the roof as I have drawn it?
    I am really trying to come up with a "cold roof" assembly that encapsulates the insulation on all six sides...
    I remember John Brooks talking about the roof being a hat...
    I like that idea of segregating the thermal and air control layers from the weather shedding layers...

  80. Lucas Durand | | #80

    I should mention re: possible ventilation issues that the WRB would be Tyvek or similar from roof peak all the way down.

  81. Riversong | | #81


    I have my suspicions about the Zip System with no additional WRB. I can't imagine it being a reliable long-term barrier, and Huber won't divulge a perm rating for the installed system (while evidence from a third party suggests a perm of less than 1).

    I think your unvented attic is fine - a large air void over the insulation allows drying even without air movement. And I think your "cold roof" is also fine, as long as there is some vertical air movement over the strapping (such as with corrugated or ribbed roofing). But I would question the use of plastic WRB as roofing underlayment (it wouldn't meet IRC standards) and would avoid Tyvek, which is the most permeable of all polymeric WRBs - far more so than is either necessary (min. 5 perm) or desireable (since it allows inward vapor movement as well). And all plastic WRBs can trap liquid water on the wrong side. In spite of the tight air barrier below, there is some potential for condensation on the roof decking/WRB.

  82. Lucas Durand | | #82

    The idea of installing the studs one-by-one doesn't bother me too much since I tend to think of them more like truss members than studs...

    Eventualy I'll just end up copying your truss system Robert ;)

  83. Riversong | | #83

    Eventualy I'll just end up copying your truss system Robert

    Oh good. Royalties. ;-)

  84. John Brooks | | #84

    I like your sketch
    Especially that you included "THE RED LINE"

  85. John Brooks | | #85

    Eventualy I'll just end up copying your truss system Robert ;)

    Lucas, I have had the same thought.
    The problem for me is that I can not find clients who are willing to live in homes with "smart" simple shapes.
    I personally like simple Architecture.
    I can not "sell" simple Architecture.

  86. Lucas Durand | | #86

    Thanks John.
    There's something satisfying in connecting "the red line".
    I'll post another sketch soon that I'd like some opinions on. It's a different wall/roof detail that I think wears a better "hat".

  87. Lucas Durand | | #87

    John, what is simple architecture to you? Is there a threshold between simple and complex? Are there specific features or forms that don't lend themselves to thick stick building?

  88. Garth Sproule | | #88

    Nice drawings! Are those done by hand? You mentioned trying " to come up with a cold roof assembly that encapsulates the insulation on all six sides." I'm sure you have already considered this but maybe parallel chord roof trusses would work for you...and add some drama to your upper floor ceilings at the same time. The problem of course, is how to get the cellulose into that space...

    I like your choice of the ThermoMass system...could you include a drawing of your footer/insulation details?

  89. John Brooks | | #89

    Simple is ONE Ridge or NO ridge
    If there are 2 stories... the 2nd floor is NOT offset

    Simple-Not is Dormer

    and then the insanity continues after dormers

    But I am also attracted to Not-So-Simple Architecture
    Because it can be art
    good art makes me feel good

  90. Interested Onlooker | | #90

    Can you disguise (and thereby sell) simple architecture with skillful use of site, light, planting, porches, balconies, breezeways, detached garages, massing, proportion, color and all the more subtle tools of the architect? Or do your customers just want fancy doo-dads like they have next door?

  91. John Brooks | | #91

    you make good suggestions

    Almost every client I have had comes to me with a predetermined image
    a style
    a certain fashion
    The clients who are most concerned about Energy Use can be the most hard headed about style.

  92. John Brooks | | #92

    The typical building site is problematic as well
    A certain number of square feet must be achieved to make the house "appraise"
    The buildable area is limited and often determines the footprint of the house.

    detached garages ha,ha Most would never hear of such a thing
    The Garage wags the design of almost every house

    And don't get me going about Fireplaces in Texas!!!!!!
    most of my clients are beyond my influence

  93. Interested Onlooker | | #93

    When I suggested a detached garage I was thinking of one that was not integral to the main mass of the house rather than, God forbid, one where the owners might have to breathe fresh air and risk getting wet between garage and house. The recent Passivhaus design in Salem OR shows what can be done to disguise the "thermal brick". Of course, if your customers want dormers and Addams family rooflines you're sunk...

  94. User avatar GBA Editor
    Martin Holladay | | #94

    John Brooks:

  95. John Brooks | | #95

    I was actually intending to post this detail on another thread...
    but this is related to airtight leave as is

    the resolution is not very clear
    If anyone wants wants a PDF or high res JPG you can email me at
    housedesign (at) verizon (dot) net

  96. John Brooks | | #96

    I am attempting to avoid "gooey-stuff"
    My thinking is that Airstop Tape can be used on the conditioned side and gaskets would be better on the not-so-conditioned side

    the plywood is joined to the top plate by a gasket
    the plywood is joined to the drywall with airstop tape

  97. Riversong | | #97

    I personally like simple Architecture. I can not "sell" simple Architecture.

    You've just got to play to the American consumer's mind. Design a lovely, decorated cube and market it as the Taj MaBox (perhaps selling it through the Big Box stores).

  98. Riversong | | #98

    The clients who are most concerned about Energy Use can be the most hard headed about style.

    AAARRRGGGHHH! I was designing a small 640 SF off-grid cabin ½ mile into the woods to be built by my first Yestermorrow class, The Sun-Tempered Super-Insulated Home. It was to incorporate my Riversong truss with native lumber and be as passive solar as possible (which requires 2' overhangs on the south and lots of south glazing).

    BUT - she had a picture in her mind of the traditional, historic New England cape, with no roof overhangs and a solid wooden plank front door with no glass.

    Fortunately, and because she wanted my class to at least start the project, I was able to convince her to toss at least some of her pre-conceived ideas and design the house around her goals (which is the only way I'll design any house).

  99. TJ Elder | | #99

    Design a lovely, decorated cube

    There have certainly been some trends toward boxiness, including boxy vehicles like the Element and the original Scion, but mainstream Americans have somehow held onto nostalgic images of houses. If you were to target the young hip crowd, they like boxy architecture. Just browse through Dwell. Unfortunately they also celebrate impractical features--no overhangs, walls of glass, no concern for thermal bridging. Boxy architecture also makes designers fixate on witty siding materials, so skin dominates the design.

  100. Riversong | | #100

    designers fixate on witty siding materials, so skin dominates the design.

    Yes, beauty in our culture is rarely more than skin deep.

    But do "witty siding materials" make passersby chuckle?

  101. Riversong | | #101

    Is this an example of witty design?:

    Here is an example of a very simple house geometry with high curb appeal:

    And here's one with an even simpler geometry that's bound to turn heads:

    And, finally, an example of the epitome of the airt-tight outsulation approach:

  102. Lucas Durand | | #102

    I do alot of drawing now that it's cold and dark at 5pm. I wish my wife would let me whittle in the house.

    I'm sure you have already considered this but maybe parallel chord roof trusses would work for you...and add some drama to your upper floor ceilings at the same time.

    I have, but I like flat ceilings. I even have my wife interested in flat ceilings now ;)
    I have become preoccupied with this idea of a cold-roof that sits over the thermal/air control layer like a hat. I will post another drawing tomorrow that better shows what I mean.

    I like your choice of the ThermoMass system...could you include a drawing of your footer/insulation details?

    I sure like Thermomass too. It seems purpose designed for double walls/ truss walls.
    Unfortunately, I don't think Thermomass is in my future. Local concrete contractors reluctance foreshadows beaucoup $$$. The drawing I posted was sort of an "ideal" idea I had for airtight sheathing (also not in my future).
    I am curious however, how an FPSF foundation would be detailed with Thermomass. In my mind I see a thermal "short circuit" between the wing insulation and the foam core - through the outter concrete whyte. I wonder if some vertical exterior foam would still be required with Thermomass in this case?

  103. Riversong | | #103

    And then some designers just can't abstain from making a simple geometry complicated:

  104. Riversong | | #104

    I am curious however, how an FPSF foundation would be detailed with Thermomass. In my mind I see a thermal "short circuit" between the wing insulation and the foam core - through the outter concrete whyte. I wonder if some vertical exterior foam would still be required with Thermomass in this case?

    The vertical and horizontal insulation must be continuous, so I don't see how the ThermoMass system would make sense for a FPSF unless the midline insulation was in addition to the outer vertical and wing insulation, which might make sense if a very cold climate. A 2x6 outer stud wall could cover 2" of foam and still have good bearing on the foundation and this would allow for a 14" thick double stud wall assembly for an R-50+ wall.

    One advantage of the ThermoMass wall when meeting a slab, is that the inner whythe of concrete can be stepped down the height of the combined slab and subslab insulation thickness, so the subslab foam meets the vertical mid-line foam to create continuity, and the inner whythe can also be poured over a 4" wide strip of foam that isolates it from the footing (this is how I did my current project).

  105. Lucas Durand | | #105

    and the inner whythe can also be poured over a 4" wide strip of foam that isolates it from the footing (this is how I did my current project).

    Robert, could the same approach be applied on the exterior side of the foundation? After the forms are removed, horizontal wing insulation for FPSF could be butted up against the foam strip thereby creating continuity with the foam core.
    Totally off the topic of this thread but I can't resist exploring this.

  106. Riversong | | #106


    I considered that, in response to your earlier query, but the loads on the exterior wall, if bearing the roof and snow loads, may be too much for a 4" wide strip of 25 psi XPS. Perhaps it's possible with the higher density XPS, but there would have to be rebar through that foam to tie the outer whythe to the footing and that would undermine the thermal break.

  107. User avatar GBA Editor
    Martin Holladay | | #107

    Robert Riversong:

  108. Garth Sproule | | #108

    The foam under the outer whythe should work in your case, as your major loads are all carried on the interior whythe...

  109. Riversong | | #109

    With the ThermoMass system, at least one of the two concrete wythes must be tied to the footings with rebar dowels max 72" oc. Since the advantage of the ThermoMass wall is that all interior concrete can be thermally isolated from the exterior (with the exception of the steel form ties, though fiber ties are available), it makes more sense to keep all steel wall-to-footing ties in the outer, cold wall.

    The foundation contractor on my current project loved steel, so he put dowels in both wythes, which meant we needed to drill holes in the 4" bottom foam thermal break and slide the pieces down over the dowels. The one advantage of that is there needs to be some mechanism for preventing the bottom foam from floating into the concrete and wire-ties on the dowels above the foam block can do that.

  110. Riversong | | #110

    I should add: least one of the two concrete wythes must be tied to the footings with rebar dowels and bearing directly on the footing.

    The fiberglass ties effectively turn the two wythes into a single structural wall; but since each wythe has only 4" of bearing width, I think it's essential that at least one of the two have direct bearing on the footing.

  111. Daniel Ernst | | #111

    Hi John,

    The passive house is coming along, slowly but surely. I've attached a picture taken a couple of weeks ago. The stainding seam roof is complete; the rainscreen siding is coming soon.

    This thread helped me form some decisions on the structural aspects of the house. Frankly, I couldn't wrap my head around the thermally isolated wall system:

    * Load sharing issues (two structural walls?)
    * Sheathing plane placement
    * Penetrations / flashings

    I think it would work great for your single story concept, but translating it to a two-story structure is much more difficult than I anticipated. Perhaps a larsen truss would be the ticket? EVERYTHING is a compromise ;-)

    I did use the airtight sheathing concept, via StoGold (liquid applied air barrier / WRB) over plywood. Although the application was tedious, it provided a vapor permeable treatment to the sheathing, AND extended exposure time for my lengthy construction schedule. September was one of the rainiest on record, and the system performed exceptionally well in keeping out the elements.

    Regarding the thermally isolated aspect - my main concern was not only the thermal benefit, but the durability issue of having cold sheathing in a double-stud wall system. John Semmelhack ran models with both THERM and WUFI. He didn't find any significant problems (caveat - in this mixed climate - 4A-such a system in colder climates might not be so forgiving). To provide some reference for the moisture risk, he modeled the construction details against more generic wall systems in our local area.


  112. John Brooks | | #112

    thanks for the update and posting the photo....
    hope you post more photos/details when you get a chance

    Have you done a blower door test yet?
    I like the Outie Windows

  113. John Brooks | | #113

    Greetings Daniel Ernst....
    I just thought I would give your Thread a "Bump"
    This was a great thread and an historic thread....
    this was one of the(or perhaps THE) first threads in the Q&A with attached illustrations.
    This was the GBA Q&A at it's best

  114. Daniel Ernst | | #114

    No blower door test yet. Have to wait a little while to see if I'm on the naughty or nice list ;-)

    I'm dealing with envelope penetrations now. I'll have to start a new thread.

  115. Aj Builder, Upstate NY Zone 6a | | #115

    Daniel, great looking home and site. The wrapping of the porch roof looks so right.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |