海上稠油热采井喷事故风险动态演化模型OA

To accurately identify the key nodes along the risk propagation path of blowout accidents during offshore heavy oil thermal recovery,this study proposed a dynamic risk evolution analysis meth-od that integrates a knowledge graph(KG)with the SIRE(Susceptible-Exposed-Infectious-Recovered)epidemic model.Based on the knowledge graph,a causation network for blowout accidents was con-structed,the SIRE model was used to characterize the propagation and evolution of risk in the accident network,and node importance was quantitatively evaluated.Specifically,cumulative risk value,maxi-mum risk value,and time to peak were used to evaluate the importance of initial event nodes,whereas exposure probability,failure probability,and recovery probability were used to analyze the importance of risk evolution nodes under different failure scenarios.The results show that,among the 49 initial event nodes,reduced steam injection efficiency and tubing connection leakage are of relatively high impor-tance.Under three typical scenarios,namely wellbore integrity failure,steam injection system failure,and fluid-solid coupling interface failure,risk evolution nodes such as wellhead pressure control failure and high-pressure fluid ingress into the wellbore exhibit stronger risk-dominating effects.Robustness tests of risk evolution nodes under parameter perturbation indicate that the stability of node-importance ranking in the improved SIRE model is 9.95%higher than that in the traditional SIRE model.The KG-SIRE model can significantly improve the accuracy of key-node identification and the effectiveness of risk intervention,providing a structured,quantifiable,and inferable analytical method for safety risk prevention and control in offshore heavy oil thermal recovery.