Bayesian approach for the reliability assessment of corroded interdependent pipe networks

Pipelines under corrosion are subject to various environment conditions, and consequently it becomes difficult to build realistic corrosion models. In the present work, a Bayesian methodology is proposed to allow for updating the corrosion model parameters according to the evolution of environmental conditions. For reliability assessment of dependent structures, Bayesian networks are used to provide interesting qualitative and quantitative description of the information in the system. The qualitative contribution lies in the modeling of complex system, composed by dependent pipelines, as a Bayesian network. The quantitative one lies in the evaluation of the dependencies between pipelines by the use of a new method for the generation of conditional probability tables. The effectiveness of Bayesian updating is illustrated through an application where the new reliability of degraded (corroded) pipe networks is assessed.