Research On Leak Prediction And Risk Assessment Method Of Subsea Gas Pipeline

Subsea gas pipelines are used as the main tool of gas transportation.Once the leakage accident occurs,it will affect the stability of offshore platforms,and even cause fire,explosion.Therefore,aiming at problem of difficulty of leakage detection and unclear gas leakage dispersion trajectory,research on the leak prediction of subsea gas pipeline,the study of underwater gas released from seabed soil considering current and wave coupling effect and the probabilities risk assessment of leakage disaster has been carried out.This paper establishes a leak prediction and risk assessment method of subsea gas pipeline,which provide technical and theoretical support for ensuring the steady and safety operation of subsea gas pipelines.Based on the above,the main research work and content are as follows:(1)Research on leak prediction of subsea gas pipeline using data-drivenBased on the relevant data of subsea gas pipeline,e.g.,process parameters,detection parameters,environmental parameters,the leak of subsea pipeline is simulated by Computational Fluid Dynamics dynamic to obtain crucial parameter.Pressure,mass flow rate,gas leak velocity and temperature of the pipeline inlet and outlet are used as inputs,and the leakage location,leakage size and leakage pressure are determined as the target output.Machine learning algorithm is used to establish a data-driven model for subsea gas pipeline leak prediction to achieve accurate and effective prediction.As a result,the proposed leak prediction model has a minor error and R~2=0.99337.It turned out that the error of leak location increases with the increase of the leak size.Conversely,the error of the pressure at the leak point is lager when the leak size decreases.The proposed model further improves the prediction accuracy on the basis of achieving rapid prediction.(2)Research on migration behavior of underwater gas released from buried subsea pipeline in marine environmentConsidering environment factors,e.g.,sea current,wave and seabed soil,a simulation model has been established to investigate the dispersion behavior of underwater gas released from the seafloor soil under combined action of current and wave.The numerical model is validated by comparing with small-scale experiments to express the validity.The migration behavior of released gas in seafloor soil and seawater environment under wave-current coupling was studied.The effect of release rate,current velocity and soil porosity on gas plume is investigated.The parameters,e.g.,rising time,horizontal dispersion distance and surfacing area are estimated.The results indicate that the presence of seafloor soil has a significant effect on the migration shape and trajectory of underwater gas plume.The combined action of current and wave increases the horizontal dispersion distance of gas plume.This study can support risk assessment and emergency planning of subsea gas release accidents.(3)Research on probabilities risk assessment of leak disaster of subsea gas pipelineA methodology mapped by BN into BRANN is proposed for risk assessment of subsea gas pipeline leak accidents,which can capture the nonlinear and the non-sequential features of accident escalation.BN is used to model the accident scenario from causations to consequences considering the conditional dependencies among accident contributory factors,and the probabilities of basic event are determined.The crucial influence factors are extracted.A probabilities model of subsea gas pipeline leak accident is established.It is observed that BRANN model with a minor error(MSE=1.98E-09)and perfect fit performance is superior in the network performance and prediction accuracy(R~2=0.9413).This methodology can help to perform more effectively dynamic risk assessment of leak accidents.