Reliability of quantitative and qualitative assessment of air leakage paths through reductive sealing

Cardoso, Vitor E.M. [Universidade do Porto] Ramos, Nuno M.M. [Universidade do Porto] Almeida, Ricardo M.S.F. [Universidade do Porto] Almeida, Manuela [Escola de Engenharia da Universidade do Minho] Pereira, Pedro F. [Universidade do Porto] Prignon, Martin [UCL]

A full characterization of a building air leakage is labour intensive. As results of laboratory and mock-up experimentation rarely portray in situ conditions, the assessment of real case studies bring added value. Still, the results of experimentation of the latter face more challenges than the former. In this work a full quantitative and qualitative assessment of air leakage paths is performed, using a light steel framing (LSF) modular building with structural insulated panels (SIPs) as case study. Blower-door measurements undergo for a sealing campaign of eleven steps, a technique often described as reductive sealing. Additionally, smoke tracer measurements were carried out to visually identify the air leakage locations. The application of three regression methods resulted in different uncertainty estimates. Less than 7% of the total air leakage was not attributed to one of the considered types of air leakage paths. Assessing less impacting leakage paths first and placing similar types of air leakage paths in a consecutive sealing order seems to be the most correct strategy when using the reductive sealing technique. On average, at a reference pressure difference of 4 Pa, the sealing step uncertainty averaged, 9.9%, 18.8%, and 27.5%, depending on the method used for regression of the blower door test results. Despite the highest calculated uncertainty, literature shows that the application of the method leading to it, Weighted Line of Organic Correlation (WLOC), provides the results in closer agreement with the observed uncertainty of measurements.