A new analysis method for evaluating reliability of oil pipelines

The evaluation for reliability of long-distance oil pipelines is urgent requirement for pipeline's owner.The traditional methods for evaluating the reliability of oil pipeline are based on the theory of reliability and supposal of failure rate,regarding the pipeline as a system of series connection or parallel connection.The failure rate is used to represent pipeline's reliability.The reliability calculated using statistic data of failure rates is unpractical for design,inspection and maintenance of pipeline's.A new concept of integrated reliability is put forward in consideration of more factors related with design, operation and maintenance of pipeline.A new model is programmed with artificial neural network.The new method for evaluating reliability is more practical than the traditional ones.The approach and steps to establish analysis model are given.