This is Part 4 of a blog series describing construction of the Orchards at Orenco project in Oregon. The first installment was titled The Largest Passivhaus Building in the U.S.

A highly insulated and airtight roof assembly tops off the Orchards at Orenco building enclosure. Overall the roof construction was very straightforward, although the parapets complicated things somewhat. Once the roof trusses were placed, they were sheathed with plywood. At the edge of the roof, we carried out a simple detail that ultimately helped us achieve a very high level of airtightness.

Experience on previous projects has taught us that the joint between the wall and roof is one of the most problematic areas for maintaining air barrier continuity. We therefore spent a considerable amount of time early during the design phase to discuss this issue and to develop a conceptual approach for the wall-to-roof joint (see Image #2, below). Typically, parapet walls on wood-frame buildings are framed as an extension of the roof truss framing; however, this standard approach prohibits a simple, constructable, and effective detail for transitioning air barrier from the wall sheathing to the roof sheathing.

Working with the design team, we decided that the parapet walls would be framed as separate components placed on top of the roof sheathing and fastened to the roof framing, but this would be done only after the air barrier had been transitioned from the wall to roof at the sheathing planes (see Images #3 and #4, below). The final detail followed the initial concept very closely.

Air sealing details

Once the roof framing was complete, including the parapet walls, we then installed a self-adhered rubberized asphalt membrane (Firestone V-Force) over the roof sheathing and several inches up the faces of the parapets (see Image #7). This membrane serves multiple functions: it’s a vapor barrier and air barrier for the roof assembly, but also a temporary roof to help keep the building interior dry until we could install the insulation and permanent roofing.

All of the service penetrations of the roof (such as plumbing vents) were sealed to the self-adhered membrane with wet sealant (see Image #8).

Exterior rigid foam insulation

Three layers of 4-inch polyisocyanurate rigid insulation were then installed over the self-adhered membrane (see Image #10). The base layer was screwed into the roof trusses and the upper layers were then adhered to each lower layer, thus minimizing the impact of thermal bridging through the fasteners.

Joints in the insulation boards were staggered to further prevent any potential thermal leakage paths. Once the insulation had been laid, a coverboard was installed and then the 80 mil single-ply roofing membrane (Firestone UltraPly TPO XR 135) was fully adhered to the coverboard (see Image #11). The total R-value of the roof assembly is R-78.

Two types of siding over exterior mineral wool

The building is clad in fiber-cement siding and brick veneer (see Image #12).

Before installing any cladding, the masons and siders first installed a 1 1/2-inch-thick layer of 8 lb. density mineral wool insulation over the water-resistive barrier (WRB), which on this project consists of Tyvek CommericalWrap. The masons used Rockboard 80 mineral wool and the siders used ComfortBoard IS mineral wool (see Image #13). Both products are manufactured by Roxul.

Stainless-steel masonry anchors were installed over the housewrap. These anchors penetrate the exterior insulation layer, creating minor thermal bridges. At the siding areas, ¾ inch thick x 3 inch wide treated plywood vertical furring strips were installed over the continuous insulation layer (see Image #14). The siders then nailed the fiber-cement planks to the furring strips with stainless steel siding nails (see Image #15).

The structural engineer of record for the project — Scott Nyseth of Stonewood Structural Engineers — designed the attachment of the furring strips to the backup wall: #12 hot-dipped galvanized screws, 4 ½ inches long, fastened at 16 inches on center vertically in line with the studs.

Keeping the furring strips co-planar

Initially the siders encountered difficulties in terms of getting the furring to lay flat on the mineral wool (see Image #16). It took some time for the crew to get a good feel for the proper amount of drive on the screws to achieve a co-planar application. On early mockups of the exterior wall, we had identified this issue, noting that plywood furring in particular was susceptible to compression into the mineral wool layer. We found that 1×4 furring when installed over the high-density mineral wool was more rigid and did not exhibit the compression problem in the same way as did the plywood furring strips. However, the siders were concerned about splitting with the 1×4 material and pushed for using the plywood instead; this was agreed to by the design team.

Roxul has issued installation guidelines and now recommends 2×3 or 2×4 furring for siding installations over semi-rigid mineral wool. Although 2x material is not necessary for structural attachment of the siding, it would certainly help with maintaining a flatter (more co-planar) application of the furring and limiting the potential for splitting.

Cutting mineral wool

Mineral wool boards can be difficult to cut to fit tightly around windows and penetrations. With a bit of research we located a tool made specifically for this application and offered it to the subcontractors.

This tool, the SkärBord insulation cutting table, is manufactured in Sweden and is sort of like a large miter box for insulation material (see Image #17). It allowed the trades to make quick, accurate cuts of the material, which helped us achieve a very high quality installation. (For more information on mineral wool cutting tables, see this GBA article: “Mineral Wool Insulation Isn’t Like Fiberglass.”)

Supporting the brick veneer

Brick veneer generally only occurs at the base of the walls, extending four feet or so up the facade. In typical construction, brick veneer is supported on a ledge that is cast into the foundation; however, this approach creates a large thermal bridge. To minimize thermal bridging at this project, the brick veneer is supported on a continuous 4”x4” galvanized steel ledger angle that has been thermally isolated from the primary structure, using off-the-shelf steel FAST brackets (manufactured by Fero Corporation), installed at 3’-0” on center along the concrete foundation around the base of the building (see Image #18).

By using these brackets, we were able to apply the expanded polystyrene insulation in a continuous layer over the exterior side of the concrete foundation (see Image #19). After the steel brick ledge was attached (see Image #20), the brickwork began (see Image #21).

Blown-in fiberglass between the studs

Once the building was dried in, we began interior finish work. Insulation netting was installed over the exterior wall framing and fiberglass insulation was then blown into the 2×10 wall cavities (see Image #22). A smart vapor retarder (MemBrain) was then stapled up on the walls over the insulation netting (see Image #23).

Before the drywall was hung, the insulation was inspected by both the local building department and Earth Advantage, the PHIUS+ rater for the project. After factoring in the advanced framing methods and details, the 9 ¼ inches of blown fiberglass, and the 1 ½ inch of mineral wool exterior insulation, the total R-value of the exterior walls is 39.

An important quality control issue arose during construction: How to best verify the quality of the blown fiberglass installation in the 9 ¼” wide wall cavities? The specifications called for the insulation to be installed at a certain density; however, you can’t determine density visually. To some degree the density can be checked with a simple “mattress test,” by placing one’s hand on the netting and pushing on the installed insulation to “feel” the density, but this is a subjective assessment.

Travis Moore, our clever project engineer, developed a crude but effective measuring device to provide a more objective assessment: a sampling box constructed of sheet metal and tape that allowed for a number of random inspections of the installed density (see Image #24). The box was pushed through the netting and insulation, and then pulled back out of the wall cavity, thereby removing one cubic foot of installed insulation (see Image #25). The box was then weighed on a scale to determine the as-installed density (see Image #26). A log was developed to record the sampling done during the course of the installation (see Image #27). Through a combination of visual inspection and random sampling with Travis’ box, we identified a number of wall areas that needed additional insulation and the insulators returned to blow in more fiberglass.

Vapor-permeable materials encourage drying

Several design measures improve the moisture performance and durability of the exterior walls.

As required by the Passivhaus standard, a high level of airtightness (maximum 0.6 ach50) was specified to improve the building’s thermal performance and to limit the potential for condensation within the walls.

The wall materials were chosen with any eye to vapor permeance. This was done to maximize drying capacity: a “flow-through” design concept, if you will.

Roxul insulation was installed on the exterior to limit thermal bridging but also to help keep the temperature of the wood framing and sheathing above the dew point for nearly the entire year. A mineral wool product was specified rather than XPS or EPS due to its high vapor permeance.

The housewrap (Tyvek CommercialWrap) has a very high vapor permeance (28 perms).

Plywood sheathing was chosen rather than OSB due to its higher wet-cup perm rating (10 perms).

Lastly, the MemBrain vapor retarder was specified for the interior side of the wall — rather than polyethylene sheeting or kraft paper — in order to facilitate drying to the interior should there be any moisture remaining in the wall upon completion of construction, or should small amounts of moisture enter the wall in service.

A monitoring set-up will track moisture performance

Early on during construction, we worked with Building Science Consulting Inc. (BSCI) and Roxul to establish a program to monitor the moisture performance of the exterior walls. As the interior finish cycle began, BSCI came to the site and installed sensors, pins, and other instrumentation to serve the monitoring study.

Relative humidity (RH) sensors and moisture content pins were placed in southward-facing wall cavities at two adjacent apartments on the third floor of the building. The same application went into place at north-facing wall cavities in the two apartments across the hall. Sensors were also placed on the face of the housewrap (WRB) and within the exterior insulation layer at the same wall locations.

Sensors and pins will measure moisture content, temperature, and RH. Data collection is planned for a period of one to two years. Analysis reports, findings and conclusions will be available from Roxul at a future date.

A forthcoming blog will discuss the blower-door test at the Orchards at Orenco project.

Mike Steffen is a builder, architect, and educator committed to making better buildings. He is vice president and general manager of Walsh Construction Company in Portland, Oregon.