Analysis of Third-Party Damage Weights in Pipeline Risk Assessment System Based on Bayesian Networks

According to the Pipeline Risk Management Manual, weight relations between third-party damage index and its subordinate indices for most existing oil and gas pipeline assessment systems are invariable in any conditions. Bayesian network framework is established to analyze the uncertainty of relations between third-party damage index and its subordinate indices. Intelligent problem solving method is introduced to solve the above framework. The calculated weights with the Bayesian Network Proposed Frameworks are more practical value in China than the traditional weights in Pipeline Risk Management Manual. The risk assessment project for Karamay to Urumqi oil pipeline of Xinjiang Oilfield Oil Storage and Transportation Company is taken as an example. Compared with the traditional pipeline risk assessment system, the results calculated by the assessment system with the new third-party damage weights are more practical and provide reference for the pipeline operator.