Dynamic reliability model for subsea pipeline risk assessment due to third-party interference

Abstract

The accidents of subsea pipelines due to third-party interference often result in catastrophic impacts, therefore, risk assessment has progressively become substantial to ensure the safety and reliability of the systems. However, the current risk analysis approaches are unable to minimize the uncertainties in the analysis due to the high demands of the qualitative inputs. The Bayesian network approach is believed to be able to provide answers to such a problem. The main advantage of this technique is that it allows the inference model and predictive analysis for constructing the current and future performance of the system based on the observed evidence. These can be achieved by introducing the subsea pipeline’s accident history and operational data in the model for developing the conditional probability distribution of each variable in the analysis. This paper proposes a dynamic reliability model for subsea pipeline risk assessment due to third-party interference based on the Bayesian approach. This technique is combined with fault tree and the finite element models for producing a reliable risk assessment framework for subsea pipelines. It is expected that the proposed model will be able to minimize the number of qualitative inputs in the analysis and also provides dynamic results for estimating the risk level of the subsea pipeline throughout its service life.