Internal Insulation – four systems in one historic residential building

Abstract

Abstract Reducing the heat loss from European buildings must include thermal retrofitting of the historic buildings. In these cases, where buildings have facades worthy of preservation, internal insulation is the only solution although external insulation is better hygrothermal solution. Inspections in buildings with internal insulation have in multiple cases shown mold growth problems at the intersection between the original wall and the insulation. Thus, there is a need for real-life testing; i.e., raising the level of reality by using professional craftsmen to apply internal insulation to existing buildings inhabited by ordinary users. Building owners are more likely to accept these solutions when they have been tested in reality, and not only in the lab or with simulations, on efficiency, robustness, and moisture safety. In Denmark, the older, worthy-of-preservation, historic buildings are commonly built with solid masonry facades. Consequently, internal insulation might involve hygrothermal risks for the external walls. The present study describes a case study of a typical historic building built in 1837. In specific, the paper focuses on the top (5th) floor of the building, which is a residential area (a commune). On the 5th floor, four different types of insulation materials were installed in eight different rooms. The performance of the four insulation materials was determined by monitoring the temperature and relative humidity at the intersections between the insulation and the existing wall, as well as the indoor and outdoor climate. Results of the measurements of the available period showed that relative humidity is increasing for most of the sensors, with an exception of a few sensors that appear to have decreased values compared to the first month. Furthermore, the risk of mold growth was calculated with the VTT mold growth model based on performed measurements and indicated minimum to non-existent mold growth risk so far.