A Serious Challenge to a New Building Technology
Every new building technology comes with new problems along with its benefits, and structural insulated panels (SIPs) are no different. SIPs are made of rigid foam insulation sandwiched between sheets of oriented strand board (OSB).The sheets of OSB act as load-bearing members, so there is no need for framing, which saves wood. SIPs can produce a very tight and wellinsulated building envelope when properly sealed (see “The Lowdown on Structural Insulated Panels,”HE Jan/Feb ’02, p. 38). But therein lies the problem. When the joints are not properly sealed, air and water vapor infiltration, especially in cold climates, can lead to disaster.
The catastrophic failure of 109 SIPsbuilt residential units in Juneau,Alaska highlights the challenges associated with building with SIPs. Failures that I’ve identified since March 2001 include roof and wall failures; the structures involved range from single-family houses to15- unit apartment buildings. A study of 37 roof failures pointed to poor installation as the major culprit (see “A Problem with Poor Installation?”). But as a project manager in identifying the 109 failures and in erecting 21 replacement roofs, I’ve discovered problems with SIP manufacturing as well.
One of the failures of Premier Industry panels in Juneau was on a project where the manufacturer had a representative on-site at least part-time. All but three of the units that were built with Insulspan panels that failed were installed by factory crews or under factory supervision. Several contractors were involved in the actual building of the failed units in Juneau; some of these were trained by the manufacturers in proper procedures.
Panels that had apparently good seals failed in the same fashion as panels that had obvious construction defects. The installation defects compounded the design and manufacturing defects in the panel system joints.The manufacturing problems with SIPs include out-ofsquare panels and routings for the splines that are significantly larger than the spline material, making it impossible to achieve an effective seal.The seal is contained inside the panel skins, which are connected with splines that completely hide the sealant. Once the panels are in place, visual inspection would reveal only gross defects in the seal.When the sealant is compressed against the foam core rather than being bonded to the OSB on both sides of the joint,water vapor can pass through microscopic voids.Thus, failure is not detected until it reaches catastrophic proportions, as happened in Juneau.
The manufacturer’s sealing instructions provide for a one-shot barrier to moisture penetration.To be effective, the barrier must keep water vapor on the warm side of the dew point. Moisture vapor can penetrate the joints as long as the outside temperature is at or below the dew point of the inside air. Locations that have less cool weather than in Juneau could be expected to experience the same type of damage— but not as quickly. Juneau has about 95% cool or cold weather, but my company recently received photographs of a house in central Ohio that had a damaged SIP roof and walls. The house is the same age as the average unit in Juneau, and the panels appeared to be about half as damaged as typical units in Juneau. According to the information we received, the panels were manufactured by one of the manufacturers involved in Juneau and were erected by a related company.
Other factors, such as occupant load and solar exposure, also affect the rate of damage by affecting the amount of moisture in the inside air and the rate of condensation in the SIPs. The manufacturers have suggested that the absence or inadequate functioning of the heat recovery ventilators (HRV) is to blame for the damage observed, but an HRV is designed primarily to ensure occupant comfort; it is not meant to protect the structure.
What Will Help?
Building Science Corporation’s Joe Lstiburek noted in a report on Juneau roof failures that the makers of Insulspan panels, which were on the roofs that his people inspected, needed to redesign their joints.The report stated,“New joint designs can be developed that rely on gaskets and interlocking joints.” We suggest that they also need to redesign their system to include
• a continuous vapor retarder of significantly lower perm rating than the OSB skin on the warm surface of the panels, which in Juneau is the inside surface;
• specific instructions on how to achieve effective seals in the panel joints with the provided sealants, including weather, temperature, and surface moisture limitations, and exact placement instructions;
• detailed instructions on the installation of the vapor retarder and design of the heat recovery ventilator (HRV) system; and
• a cold-roof design that is mandatory.
A properly installed vapor retarder of significantly lower permeability than the OSB serves as a first line of resistance to vapor or air penetration of the panel joints. Properly installed sealants, adhered to both adjacent surfaces at the proper locations, provide a second line of resistance to moisture penetration to the level in the joint where condensation can occur. A cold roof installed over the panels provides a damage control mechanism in the event that the first two mechanisms fail.
While we would never sanction depending on the HRV to protect the structure, a properly designed HRV can be beneficial to the structure by preventing the accumulation of warm moist air in the upper levels of the structure, and by maintaining a slight negative pressure on the structure, which helps to minimize the vapor pressure. Designing an HRV to benefit the structure requires increasing air flow and changing inlet/outlet locations—a more costly alternative than one that simply addresses occupant comfort. Over the life of a structure, it is unlikely that the vapor retarder will remain intact. Multiple layers of vapor resistance, comparable to the multiple load paths in structural engineering, provide the best chance for survival of the SIP structure.
In many respects, Juneau is the Challenger disaster of the SIP industry and, to a lesser extent, of code certification agencies such as Building Officials and Code Administrators International, Incorporated (BOCA), and the International Conference of Building Officials Evaluation Service (ICBO ES). The SIPs were tested and certified for their structural performance, but apparently they were never evaluated by the code agencies in a system-oriented process that would have tested the joint design. However, the code agency certifications were interpreted to cover all of the manufacturer’s installation instructions and recommendations, not just the structural issues, and the process leading to the failures was irreversibly set in motion. Whether the SIP industry learns its lessons and applies them as NASA did is the big question confronting the industry and the buying public.
Enter your comments in the box below:
(Please note that all comments are subject to review prior to posting.)