Risk-Informed, Reliability-Driven Decision Support for Building Basement Systems

Abstract

A lack of consideration for maintainability at the planning/design stage of a construction project often results in inadequate decision-making and selection of design alternatives, which will likely lead to unproductive operational regimes riddled with frequent defects. This paper attempts to anchor the quantitative risk analysis technique, Bayesian networks, as apt in predicting future operational defects based on the initial design alternatives by looking at specific case studies in building basement systems. Therefore, this paper aims to develop a comprehensive risk-informed, reliability-driven decision-making system. Central to this process is the implementation of a setup wherein a Bayesian Belief Network (BBN) operates as a powerful reliability engine. Concurrently, a risk matrix serves as a vital tool for facilitating risk-informed decision-making. This approach empowers stakeholders to make informed decisions regarding design choices while keeping maintainability at the forefront. It allows for maintainability to be practically used as an effective tool in buildings to lower the lifecycle cost. Furthermore, it is envisaged that this approach can be adopted in the construction industry globally for the prediction and diagnosis of defects, thereby facilitating risk-informed decision support. This study provides significant contributions to the building industry by providing a novel approach to a basement defect analysis, encouraging interdisciplinary collaboration among construction stakeholders, and assisting facility engineers and managers in properly determining maintenance requirements.