Study On Quantitative Risk Assessment And Decision Making Of Third-party Damage For Oil And Gas Pipelines

With the rapid process of urbanization,third-party damage poses an increasingly serious threat on the safety of oil and gas pipelines.Proper risk assessment and decision making can help to make an optimized arrangement of risk-prevention resources by exploring the relation between risk factors and potential accidents.The risk of third-party damage is variety and complex.Besides,as human behavior cannot be simulated accurately,traditional risk analysis methods find it difficult to apply to the third-party sabotage.Therefore,Bayesian theory and game theory are applied to analysis different types of damage.The research is arranged as follows.1.In order to achieve a comprehensive identification of third-party risk factors,3S_BN theory system is proposed based on statistics,scenario analysis,safety barrier analysis and Bayesian network.According to the 3S_BN,a risk identification model with multi-state risk factors is established and its performance is tested in a case study of a typical pipeline leakage accident.Compared to the previous papers,the optical fiber broken and fire and explosion are included in the consequences of pipeline failure,followed by the definition of risk factors in the post-incident.2.When third-party damages are made without intention,a quantitative risk assessment model is proposed based on the reasoning mechanism of Bayesian network and expert judgments.In the proposed model,D-S evidence theory is adopted in terms of its performance for the computation of conditional probability table in the Bayesian theory in order to improve experts' belief.The model is validated for different situations of leakage,fire and explosion and optical fiber broken.Through sensitivity analysis and posterior probability reversal,the key influencing factors and possible paths of incidents are confirmed.Finally,the application of the proposed model is discussed.3.It is difficult for traditional risk decision-making methods to simulate the strategic interaction between third parties and pipeline company.The game theory is used to analyze the interaction process,considering different types of sabotage attack targets,attack strategies,and differences in revenue calculations.Under the assumption of complete rationality and common knowledge,the Bayesian-Stackelberg model is established.The case study shows that this approach can give advice to the optimization of defense resource allocation to achieving the purpose of reducing operation and maintenance costs by comparing the expected losses of different defense strategies by assignment calculations.4.Traditional game theory is based on the analysis of pipeline company and third-party sabotage who fully understand each others' strategies,which is a one-time game.In view of this,the evolutionary game theory is used to analyze the long-time dynamic strategy adjustment of both parties.The mental model and cognitive rules are used to simplify the complex solving process of the traditional Wright manifold theory.What's more,prospect theory is adopted to change the expected payoff matrix into a payoff sensing matrix,so that the analysis is conducted under limited rationality completely.The analysis results presents a threshold and four equilibrium conditions,which give pipeline company a flexible reference of defense strategy adjustment.