Application of Bayesian Networks for Geotechnical Risk Analysis: Case Study in Algeria

Abstract

The management of geotechnical risks in construction projects is becoming increasingly indispensable due to the growing complexity of civil engineering works. In addition, the random nature of soils forces practitioners to use tools that deal with the concepts of chance and probability. Bayesian networks are considered as effective tools for addressing issues related to the uncertainty and reliability of systems. The structure of this type of network allows great flexibility of use, and this is the reason why a large number of applications have emerged in recent years. However, the complexity of such networks limits their use in the professional field. The purpose of this article is to design a Bayesian network that could be exploited by all project stakeholders to better manage geotechnical risks. This method could be used in the concrete case of the highway viaduct project of the penetration link from the town of Ghazaouet to the East–West Highway in Algeria. This is expected to provide an effective decision support tool that will certainly allow for a better management of the major risks associated with construction. The results obtained during our study allow us to visualize the most likely scenarios for the occurrence of major risks and thus to be able to intervene in due time.