Research On Dynamic-Bayesian-Network-Based Resilience Assessment Method Of Offshore Oil And Gas Equipment

Development of offshore oil and gas has become an important way to ensure national oil security.As an important infrastructure for offshore oil exploitation,the safe and efficient production of offshore oil and gas equipment is the important basis of offshore oil exploitation.Resilience is understood as the ability of an engineering system to react and recover to its original state after being attacked by unexpected disturbances or events.The resilience evaluation of offshore oil and gas equipment can quantitatively evaluate the resistance ability and recovery ability of the system,which is of great significance for the safe and stable operation of the equipment.The work proposes a dynamic-Bayesian-network-based resilience evaluation methodology.Resilience of oil and gas equipment and facilities,such as blowout preventer(BOP),subsea oil and gas pipeline,and subsea wellhead connector are evaluated and analyzed.Problems such as insufficient data in the modeling process and insufficient coupling law of multiple factors in the mapping process are solved.Control system is one of the most critical components of the deep-water BOP,which directly determines the performance of BOP and the safety of the deep-water drilling system.Failure factors of the engineering system are not single,and the influence rules of their influencing factors usually do not have accurate mathematical models.There are problems that are difficult to evaluate and model.In order to solve the problems such as strong coupling and difficult modeling of various external factors,a novel method for constructing the resilience model of offshore oil and gas equipment control system by mapping the physical model into dynamic Bayesian networks(DBNs)is proposed.The external shock model is established by mapping from actual physical models of influencing factors,the failure rate of a single component is obtained from the established disaster model,the system performance is obtained through the combination of Markov model and DBNs.The performance of system degradation and recovery process is simulated in real-time,and resilience of power supply and control systems of subsea blowout preventers under the influence of external disasters are evaluated.The multifactor physical mapping model and the DBN model of the system under various shocks are established.The influences of categories and strengths of disasters,and system configurations and maintenance strategies on the system resilience are researched.Structure system is almost the most important part in an engineering system.In the harsh working environment,the performance degradation of structure systems is caused by both natural and external factors.Failure of structure system is caused by the cumulative damage of the fatigue properties of components,and it is closely related to the mechanical properties.As system performance parameters,reliability or availability is the macro performance of the system.The real failure mechanism and failure process of the structure system cannot be reflected.In order to solve the problem,a resilience evaluation metric based on remaining useful life and its corresponding evaluation method of offshore oil and gas equipment structure system are proposed.The failure process of subsea oil and gas pipelines can be truly modeled,and resilience evaluation results are more reasonable and realistic.According to the actual situation of structure system,the degradation process is divided into two parts,named natural degradation and accelerated degradation.The two parts are modeled separately,and then the degradation model of the system is established by using DBNs.The crack depth propagation and the remaining useful life changing law of pipeline under the influence of external impact are simulated.In the recovery process,the model is established by combining Markov and DBNs.The crack depth and remaining useful life of the system in the recovery process are simulated,and the resilience of subsea oil and gas pipelines under influence of external impact is evaluated according to the resilience evaluation metric.When evaluating the resilience of the system,the actual physical and mechanical properties of the structure should be fully considered,such as the shape,size,assembly relationship,equivalent stress and strain of the structure itself.All the data needed in the DBN modeling process cannot be obtained according to the expert experience and historical data when mapping the physical models onto DBNs model.In order to solve the problem of insufficient training data in the resilience modeling process,a resilience evaluation method combining finite element analysis and DBNs is proposed.Finite element model of the structure system and finite element analysis results are directly converted into a DBN model to constructed the degradation model.The geometrical,structural,mechanical and material factors are mapped to the corresponding nodes in DBNs,and parameters of geometrical,structural,mechanical and material are used to construct the parameter model of DBNs.Recovery process is composed of fault diagnosis stage,resource allocation stage and maintenance stage.Physical models for different recovery stages are mapped into DBN sub-models,and sub-models are integrated to form a whole recovery model.The resilience of wellhead connectors is evaluated under the impacts of different inter wave velocities and different inter wave action times.The resilience variation law of the wellhead connector is studied.