Hygrothermal performance of insulated, sloped, wood-framed roof assemblies

Roofs are the single enclosure element common to artificial shelters constructed by all cultures in all climate regions. The hygrothermal performance of insulated, sloped, wood-frame roof assemblies has long been of interest to building scientists and building codes alike. Requirements for the ventilation of roof assemblies have been included in building codes for more than 50 years, however moisture problems still occur. Unvented roof assemblies have been suggested as an alternate and potentially superior solution. While unvented roofs have become relatively common in the warmer southern climates of North America, some technical issues remain unclear and the need exists for further study and demonstration of the hygrothermal performance of both ventilated and unvented roof assemblies in northern climates.;This thesis seeks to improve the understanding of the hygrothermal performance of ventilated and unvented, insulated, sloped, wood-framed roof assemblies using field measurement, analytical calculation and computer simulations.;A review of existing literature and current industry experience and practice was prepared. Five categories of research needs were identified: roof leaks, air leakage, vapor diffusion, roof temperature and ice dams. Air leakage, vapor diffusion and roof temperature are addressed by the research work described in this thesis.;The connection between attic moisture problems and air leakage was confirmed through field investigation. Analysis suggests that the benefits of improved ceiling air tightness may not be realized until extremely low levels of air leakage are achieved. Problems may be more easily mitigated through control of indoor relative humidity levels. Attic ventilation can assist in mitigating moisture problems, but it is not as effective as the preventative measures.;An unvented cathedral ceiling roof assembly was constructed at a test house in Vancouver BC and monitored for several years. The assembly was insulated with air impermeable, vapor permeable sprayed polyurethane foam insulation and without the use of a vapor barrier. The assembly performed adequately, however a strong connection between indoor humidity levels (i.e. vapor pressure) and sheathing moisture content was identified. When strict control of indoor humidity is not possible, modeling should be conducted to establish the vapor permeance required reduce to winter time vapor diffusion so that the moisture content of the North-facing roof sheathing does not exceed 20%.;A side-by-side field monitoring study of six ventilated and unvented roof assemblies was undertaken at a test hut in Coquitlam BC. The test demonstrated that both ventilated and unvented roof assemblies can perform satisfactorily; frost and condensation are not a concern when the indoor humidity is reasonable, the ceiling plane allows minimal air leakage, roof temperatures do not remain cold for extended periods and the ventilation is straightforward. Under these conditions the unvented cathedralized roof assemblies exhibited only slightly higher peak wintertime sheathing moisture contents. The application of two heavy coats of latex paint as a vapor control layer on the inside surface of the foam insulation was demonstrated to be ineffective.;Monitored sheathing and shingle temperatures of four pairs of ventilated and unvented roof assemblies in an Atlanta GA test hut confirmed that peak summertime sheathing and shingle temperatures are reduced by ventilation, however the study also suggested that there is little difference between the sheathing and shingle temperatures recorded in a conventional (ventilated) cathedral ceiling roof assembly and an unvented cathedral ceiling assembly.;The WUFI 4.1 Pro software program was used to predict the performance of the unvented roof assembly employed at the Vancouver BC test house. Good agreement was achieved between predicted and measured sheathing temperatures and moisture contents, demonstrating the potential to predict the performance of unvented roof assemblies. Further work should be done to assess the impact of moisture storage in the wood framing, material property assumptions and boundary conditions. The WUFI 4.1 Pro software should considered for the simulation of ventilated attic roof assemblies.