Bayesian Network revealing evidence-based strategies to enhance the performance of building envelope openings subject to wind-driven rain

Abstract

Severe storms and tropical cyclones bring destructive winds and heavy rain. While building structural performance has significantly improved in the last few decades due to higher regulatory requirements, some non-structural elements, such as windows, external doors, roof coverings and attachments such as guttering, fascia and eaves, remain subject to minor failure, causing loss of amenity and damage to structural building components over time. Enhancing the performance of buildings has become imperative to mitigating the impacts of tropical cyclones and storm events. Damage investigations conducted after tropical cyclones and severe storms have consistently revealed that windows and external glazed doors are affected by wind-driven rain, causing leakage into the cavity and interior of the building. This research study focuses on repeated water ingress through windows and external glazed doors. Wind-driven rain can penetrate undamaged windows and external doors, gaps around the window seals or doors, and waterproofing elements, thereby allowing water to enter buildings. A qualitative expert interview research approach was applied to identify several factors affecting the performance of openings (windows and external glazed doors). Subsequently, a Bayesian Network model was developed according to the determined parameters and expert workshops. The Bayesian Network scenario analysis enabled the researchers to identify the best combination of management interventions to enhance the performance of openings to water ingress from tropical cyclones and severe storms. The study findings provide evidence-based support for industry and government authorities to develop effective strategies for enhancing the performance of openings subject to wind-driven rain from tropical cyclones and severe storms.