Literature Review on Windows Airtightness Performances: Insights and Gaps

Old windows make a major contribution to the authentic look of a façade, and maintaining those elements whenever possible is essential for the conservation of our architectural heritage. However, those are often leaky and are consequently responsible for high energy losses, acoustical and thermal discomfort, and incorrect sizing of the ventilation systems. Having a better knowledge of windows airtightness performance is crucial in assessing need for an intervention based on the balance between costs and impact. In this paper, we review 43 studies published between 1930 and today reporting on window airtightness measurements. The results show that the lack of standardized way of expressing the results and the small number of reported results and the bad repartition among studies make it difficult to draw solid conclusions, especially for in-situ measurements. Despite those limitations, certain trends can be observed based on the reviewed studies: (1) in-situ testing provides worse results than laboratory tests, which is explained by the window-wall interface, the deterioration over time and the deterioration during installation; (2) since the arrival of weatherstripped windows, there is no clear improvement of the windows airtightness over time, for both laboratory and in-situ testing; (3) among the different opening mechanisms, sliding windows are found performing worse than others; (4) among different materials for window frames, steel and aluminium are found performing the worse and wood the best. However, those conclusions were strongly hindered by the identified limitations; (5) existing windows are notably leakier than newly installed windows and the performance seems to decrease over the years, but this was based on a very few numbers of windows tested. Further work should focus on acquiring data from in-situ measurements on existing windows. Those study should report the results in terms of leakage characteristics and minimum information should be the opening mechanism, window frame materials, the period of construction of windows and the year of testing. Additionally, those trends were observed by analysing considerable number of studies with different objectives. Each of the trends observed here should be confirmed and confronted through dedicated studies.