September 18, 2014

Middlesex VT: Foam-free, WUFI-verified, and OSB-free high performance

The town of Middlesex, Vermont, located just north of Montpelier, has put itself on the high performance building map. Leading the way is the new “Whitchurch Passive House Cottage” – not only in terms of energy efficiency (Passive House certification is in the works), but also as a unique and adventurous example of “form follows function”. The project was executed with thorough commitment by the owner, Greg Whitchurch, along with CPHC Indigo Ruth-Davis and Passive House Certified Builder Chris Miksic of Montpelier Construction. Though simple in geometry, the house challenges conventional construction practices on a number of levels. Some of the most interesting innovations are hidden behind the cedar cladding, where advanced Pro Clima membranes and tapes provide state-of-the-art airtightness and moisture protection. Let’s take a closer look.  (Photos by Greg Whitchurch)

Whitchurch Passive House - with storefront windows show the post&beam structure behind

Whitchurch Passive House Cottage – with storefront windows showing the post & beam structure behind

The house employs some of 475 High Performance Building Supply’s best practices, including foam-free super-insulated TJI assemblies with interior and exterior airtight membranes to maximize moisture protection and insulation performance. (Visit DWG Construction Details for free access to our recommended high performance assemblies). The owner also wanted a flat roof that kept snow from sliding off, but he wanted to avoid using foam. What he got, according to a professional engineer familiar with the project, was a “non-vented cellulose dense-packed insulated flat roof assembly”. In the past, unvented flat roofs have routinely failed due to poor detailing and inadequate moisture control. We discuss flat roof issues in a couple of blog posts – see here and here. The game changer on this project was one product: INTELLO Plus “smart” vapor control membrane. Serving simultaneously as the airtight and variable vapor control layer, INTELLO made it possible to keep moisture transmission through the roof assembly within safe levels.

      Post & Beam
To verify and validate the performance of the project’s flat roof assembly, hygrothermal modeling was performed by 475 Technical Director and WUFI expert Floris Keverling Buisman. This analysis helped guide the final assembly configuration. The roof is free of sheathing and foam, and solely insulated by cellulose and wood fiberboard. It starts with an interior finish of tongue and groove white pine boards that are flush with the structure, followed by INTELLO Plus, 16″ I-joists, 2×6 hemlock, white pine 1x’s, wood fiberboard and EPDM. For this desired foam-free assembly, it was shown that INTELLO was required to control the moisture content of the roof and keep it below 20 M%, a safe level for wood sheathing elements in the assembly. The smart vapor retarding properties are required since the outward drying potential is zero due to the EPDM membrane roofing. The model did include a small amount of air leakage to verify that indeed the drying potential was larger than the (un-)anticipated airleaks/moisture intrusion in winter.

WUFI further showed that in Vermont’s climate it is beneficial to have a dark roof. The INTELLO Plus on the warm side of the assembly keeps wintertime moisture from the interior from diffusing out of the assembly, while allowing drying to the inside in summer. This process is aided by the highly absorptive dark roof. The graph below shows that the dark membrane plus INTELLO keeps the pine board sheathing below 20 M% in the first winter, though for a few weeks it exceeds 18%, the conservatively safe level for solid wood. After the 2nd winter, the moisture content goes down to very safe levels. The light-colored roof peaks higher and longer in the first winter (21 M%) and only in year 5 does it drop below 18 M%. Hence the black roof membrane leads to a roof with a high drying reserve and hence a very safe assembly. 475 worked closely with the team and a structural engineer to accept this assembly as a cold climate alternate solution. 

WUFI moisture % in pine boards above cellulose and INTELLO

WUFI moisture % in pine boards above cellulose and INTELLO  (Grey line is with white roof.  Dark line is black roof.)

Ultimately, a successful roof outcome for this project depended on careful airtight installation of the roof layer. Typically, INTELLO Plus is stapled to the bottom of rafters, then horizontal battens are fastened over the membrane to help distribute the load from blown-in insulation such as cellulose. In the Middlesex project, the interior of the roof assembly was built in reverse. Interior 2×8 pine tongue and groove boards were attached to the outboard side of the structure, over the traditional post and beam structure. This layer served as a sturdy backing for the densepack cellulose that followed, and the vapor open properties of the wood didn’t compromise the vapor intelligence of the INTELLO. It was then a breeze to achieve a perfect airtight layer with INTELLO across the top of the solid wood flat roof. Seams were made airtight with TESCON Vana solid acrylic tape.

SDC10770       INTELLO Plus airtight layer layed out over vapor open roof sheathing
To complete the roof assembly, 16″ TJIs were placed 3′ on center over the INTELLO, on top of which hemlock 2x6s were installed perpendicular to the TJIs. After another layer of vapor open pine boards, a layer of waterproof and vapor open SOLITEX Mento 1000 was added (temporary water/backup control). Last came fiberboard and 60 mil EPDM.

SDC10990       SDC11169
At the walls, DB+ replaced INTELLO Plus, but the overall airtight vapor control layer was installed in similar fashion, directly to the outside of the finished interior T&G cedar.

SDC10781       SDC11161
16″ TJIs were then cantilevered outward to create an uninterrupted thermal layer with densepack cellulose. Vapor open SOLITEX Mento Plus over the TJIs was integrated with deep plywood window bucks, followed by double strapping to support the vertical tongue and groove cedar cladding.

SDC11106       SDC11314
The owner installed sensors in the walls and roof to measure moisture in the assemblies – with a particular interest in the unvented roof.  We will update this project blog with moisture data as it becomes available.

SDC11433    SDC11354    WhitchurchPHPP  
The house is on track to meet the Passive House standard – and then some. At 0.28 ACH50, the blower door test comes in at less than half the leakage required for Passive House certification. This phenomenal result ensures that interior moisture will stay out of the walls and roofs, while keeping the annual heat demand around 3.4 kBtu/sf/yr for minimal heating and cooling bills.  Great job guys!

, , , , , , , , , , , , , , , , ,

Tweet about this on TwitterShare on FacebookShare on LinkedInPin on PinterestEmail to someone

9 Responses to Middlesex VT: Foam-free, WUFI-verified, and OSB-free high performance

  1. G Hawirko September 19, 2014 at 6:28 pm #

    I hate to burst your Bubble’s, but EPS is “Expanded Polystyrene” FOAM.

    • Ken September 29, 2014 at 10:45 am #

      yes, thank you – our focus is on above grade (or interior basement) solutions.

  2. Joel Baker September 25, 2014 at 9:32 pm #

    While this is truly an impressive effort, I have to wonder why we lead the story with the claim “foam free” and “OSB free” Cross section shows a fair bit of EPS foam used to isolate foundation , and the TJI webs appear to be OSB. What say ? Thx, JB

    • Greg Whitchurch September 29, 2014 at 3:21 pm #

      We did a lot of research looking for the “greenest” way to insulate below ground. We were unable to find another product here or overseas that performed nearly as well as EPS. While the other foams are blown w/ poisonous agents, EPS is blown w/ steam &, unlike PIC, XPS, etc., it is recyclable. Other “newer” products perform very well as insulating materials until they get wet. EPS has the ability to resist moisture take-up as well as to release the taken-up moisture & return to a relatively dry & stable state once again. About 15% of the foam we used was re-purposed (acquired from tear-downs). We used most of our scraps for insulating piers, drains, etc.

      We also researched the components of the OSB we used. All of it was contained in the TJI’s & we obtained relatively low-formaldehyde, low-VOC OSB-webbed 16″ TJI’s. Those TJI’s are on 3′ ctrs.: (9) 16’ers on ea. side & (9) across the 32′ roof. Again, all short- & long-term issues considered, we were unable to find a “greener” solution.

      However, we’d welcome any suggestions that might show us a better way – for next time!

      • Robert Haverlock June 14, 2016 at 9:14 pm #

        Again, foam glass and talk to Andrew Michler @ about his use of mineral wool at 7000 feet.

  3. Ken September 29, 2014 at 10:48 am #

    Yes, like above, we are referring to those assembly components that our products address – above grade air, thermal and vapor control. We’ll be more explicit next time….may even edit this post to be more succinct. Thanks.

  4. Robert Haverlock October 1, 2015 at 4:47 pm #

    Why not (cost) Foam glass for below ground? And, Passive-to-Positive builders used Mineral Wool in their below ground foundations…

  5. steve zerby March 21, 2016 at 11:40 am #

    Any updates on actual moisture readings?

    • Ken March 24, 2016 at 4:54 am #

      Hi Steve,
      Yes we are happy to report that the roof assembly continues to dry – as noted by Ted Cushman in the March issue of JLC. The owners have been faithfully collecting data and we hope to issue a comprehensive report on the roof status in the coming months. We have a snapshot of recent data noted at the top of this blog post: We look forward to following-up with more.

Leave a Reply