A probabilistic-based model for predicting pipeline third-party hitting rate

Abstract

Third-party damage (TPD) is a severe threat to the integrity of the in-service oil and gas pipeline. This work aims to develop a probabilistic model to predict the hitting rate of pipelines by third-party excavation under a dynamic, uncertain process, which is subjected to the impact of the effectiveness of various preventative practices. In this work, a Bayesian network (BN) was utilized to predict the hitting rate on the pipeline by third-party excavations, in conjunction with the Markov process. The effectiveness of typical industry prevention practices (e.g., one call system) against third party damage was considered. The uncertainty associated with survey samples for the qualitative assessment of the effectiveness of prevention practices was addressed using the bootstrap technique. The presented methodology can provide a complete probabilistic description of the uncertainty associated with pipeline hitting.