An evaluation of two well insulated roof structures

Abstract: Moisture is considered to be a severe problem in outdoor ventilated attics in Sweden today (2015), and almost 60% of all single family houses and 10% of all multifamily houses have problems with mould growth correlated to moisture in the attics. Increased insulation layers and new heating systems in combination with a more humid climate have increased the risk of such problems. This study has focused on investigating the moisture and mould growth potential in a well-insulated outdoor ventilated attic and a cathedral roof in four different climate zones in Sweden; Lund, Stockholm, Borlänge and Luleå. The study was performed by modelling and simulating the structures in the hygrothermal software WUFI 5.3 with climate data from Lund, Stockholm, Borlänge and Luleå from the year 1990-1998 from SMHI. The structures were also investigated from a future-climate perspective, using modified data obtained from climate scenario RCP 8.5 considering parameters such as temperature, wind velocity and annual precipitation. The outputs of the critical parts of the structures were analysed for the risk of mould growth with four different mould growth prediction models: VTT-model, MRD-model, m-model and WUFI-Bio. The hygrothermal results showed that the cathedral roof has more fluctuating conditions compared to the attic. The relative humidity there is generally lower throughout the year, except during winter. The mould model results were inconsistent in their assessments, especially in Lund and Stockholm, while Borlänge and Luleå had fewer conflicting results. In general the cathedral roof performed better in cold climates, and the attic performed better in warmer climates. The future climate generated significantly worse situation for both structures in all locations. From our results general conclusions could be drawn. In Lund and Stockholm there was significant risks of mould growth in both structures, the recommendation is, based on accessibility, to construct an attic. In Borlänge it did not really matter what structure was chosen, the choice can be based on other preferences such as design, economics, etc. In Luleå, the cathedral roof was recommended, based on the results from the future climate. The spread of results from the mould predication models reflected the complicated processes of modelling mould growth. The study showed that it is important to be well aware of the models limitations and to treat the predictions as indications/suggestions more than truths.