Multivariate Probability Distributions for Reliability Analysis of a Submerged Floating Tunnel

This dissertation investigates the use of probabilistic models, including copulas, vine-copulas, and Bayesian Networks, to enhance the understanding and reliability of submerged floating tunnels (SFTs). Despite their potential for challenging water crossings, no SFTs have been constructed beyond a prototype in Qingdao Lake, China, highlighting significant uncertainties related to their reliability under complex loading conditions. Through a case study of a hypothetical SFT in the Qiongzhou Strait, China, this research investigates the joint probability distribution of traffic and environmental loads, including wave height, wave period, and current velocities. The study develops methodologies for reliability analysis, focusing on the interactions between traffic loads, metocean conditions, and potential cascading failure scenarios. The findings demonstrate the importance of accounting for complex dependencies among variables to enhance the safety and stability of SFTs. The flexible models proposed in this research can be updated with new data, enabling continuous safety assessments and providing valuable insights for the design and planning of future SFT structures.