A literature review of probabilistic hygrothermal assessment for building envelopes

This study presents a state-of-the-art review of the evolution of probabilistic hygrothermal assessments of building envelopes. First, relevant literature was identified by employing a systematic methodology, including scoping and systematic searches, quality assessment, and study analysis. The analysis, based on 143 selected publications, provides statistical insights into uncertainty propagation and sensitivity analysis methods, geographical distributions, climatic conditions, sustainability perspectives, uncertain parameter groups, employed metamodels, hygrothermal analysis tools, model dimensions, building usage, and moisture-related damage models. The results illustrate that the majority of probabilistic hygrothermal assessment studies, exceeding 95 %, are not explicitly incorporating sustainability perspectives, including environmental, economic, and societal. Also, to enhance the accuracy in this field, it is important to integrate hygrothermal functions as uncertain input variables, given their potential to significantly impact hygrothermal performance compared to single material properties. While 90 % of studies are confined to 1D modelling, extending research to 2D and 3D is advised to capture more complicated details. Moreover, case studies in tropical climates, where moisture- related issues can be prevalent, are noticeably lacking. Sensitivity analysis faces limitations due to its transient nature and variable interactions, necessitating identification of methods capable of non-linear and interactive analyses with low computational costs. Furthermore, several studies neglect ambient conditions and building usage considerations as stochastic variables, highlighting the need for comprehensive analyses. Lastly, understanding the merits and limitations of metamodeling methods is crucial for achieving reliable and robust hygrothermal analyses. The paper concludes by outlining the identified gaps and offering recommendations for future research in this field.