ABSTRACT
Gas transmission pipeline network is of great importance to any country using natural gases in its various technological processes. However, the usefulness cannot overshadow the threat posed to people and property by the grid failures. In order to quantify these risks, several widely recognized pipeline incident databases have been established: OPS, EGIG, UKOPA, to name a few. Each database contains data about pipelines operated in different geographical regions with varying soil types, under different maintenance environments. In addition, definitions of incident vary across data samples. Hence, direct integration or pool of data into one analysis and sample raises suspicions about the validity of resulting inferences. The purpose of this paper is to move beyond the qualitative pipeline accident database comparison and to draw guidelines for quantitative integration of all available statistical information to improve reliability evaluation. We will review five natural gas transmission pipeline network accident databases and develop a methodology based on hierarchical Bayesian model of pipeline reliability data while taking into consideration varying incident criteria.
