Curiosity as Method
RDH’s wildfire work began with practical questions about enclosure performance. Published research provided part of the picture, but many questions about how materials, assemblies, and details respond to wildfire exposure could only be answered through observation and testing. Rather than rely on assumptions, James Bourget and Robin Urquhart began building small-scale test assemblies and conducting burns to better understand that behavior.
Building Our Own Evidence
To investigate how common window systems perform during wildfire exposure, RDH began conducting standardized window burn testing. Robin Urquhart built a California State Fire Marshal (SFM) 12-7A test rig in Atlin, British Columbia, allowing the team to evaluate commercial window systems, glazing configurations, and candidate protection measures under consistent conditions.
Window testing has revealed important differences in how glazing systems respond to wildfire exposure. Annealed glass fails within the first minute of exposure, while tempered glass lasts longer and can still break within the eight-minute test. The location of the tempered pane within an insulated glazing unit also affects performance. Standard residential vinyl frames distort under heat, while thicker European multi-chamber vinyl performs comparatively well. Borosilicate glass, the same material
used in oven doors, has remained intact throughout the full test.
As the work expanded, the test rig moved to Thompson Rivers University in Kamloops, where it supports academic partnerships, student research, and ongoing testing with local firefighters. RDH continues to use the facility while expanding its testing through collaborations with Cal Fire, industry partners, and other research organizations.
Scaling the Questions
As new questions emerged, so did the scale of the testing. RDH expanded from window testing to full-scale wall burns, evaluating complete enclosure assemblies under representative wildfire exposures.
One series of tests subjected one-hour-rated wall assemblies to a 50 kW/m² fire load at the base of the wall, representing a structure-to-structure fire exposure.
Following testing, the Tyvek and framing behind a wall assembly clad with mineral wool and fiber cement remained in pristine condition. A second assembly, consisting of gypsum sheathing beneath vinyl siding, protected the framing side of the wall through gypsum’s endothermic reaction, even as the cladding and strapping sustained damage. In both cases, post-test examination documented how fire moved through the assemblies and confirmed the continuity of the wildfire-resistant barrier.
The same approach has since been applied to residential projects, helping evaluate enclosure assemblies under representative fire loads before construction.
Beyond the Burn
Testing is one way RDH develops expertise. Modeling and post-fire investigations provide additional ways to explore enclosure performance and compare observations across different conditions.
RDH’s building physics team has adapted finite-element tools commonly used for hygrothermal analysis, including THERM, Flixo, and Siemens NX, to model fire loads across roof assemblies and evaluate heat flow into building structures. Building-scale wildfire modeling further examines how radiant heat and flame impingement from surrounding fuels interact with site conditions, vegetation, neighboring buildings, and topography.
Site investigations following wildfires provide another important source of evidence. RDH staff have documented enclosure performance in Jasper, Paradise, and Los Angeles, observing what burned, what survived, and why. In a Jasper campground where twenty buildings survived a fire that destroyed the surrounding forest, the exposed building surfaces provided valuable insight into how enclosure assemblies performed under real wildfire conditions and where they approached their limits.
Together, testing, modeling, and field observations allow each question to be explored from multiple perspectives. Physical testing reveals how assemblies perform under controlled conditions. Modeling extends those observations to additional scenarios. Field investigations compare those findings with the performance of real buildings, helping identify new questions and guide future investigation.