Oil and gas pipelines are likely to fail resulting from corrosion because of the interactions between pipe materials and service environment,which seriously impacts the safety operation of pipe systems.The processes of corrosion initiation and growth have the nature of nonlinearity and complexity,which make the prediction of corrosion failure likelihood(CFL)to be difficult.In order to overcome this difficulty,an analytic model for predicting CFL of oil and gas pipelines was developed by means of fuzzy logic approach.In the model,corrosion damage factor,inspection effectiveness and inspection numbers are considered as the key influencing factors affecting CFL of pipes.The fuzzy rules between the factors and CFL were determined according to the influence degrees of the factors.The model was applied to a natural gas pipeline to find out the pipe that is most likely to fail due to corrosion.The results can provide effective reference for working out pipe maintenance plans.