Research on prediction method for reliability growth of natural gas pipeline
In order to reduce the design and manufacturing defects of natural gas pipelines before commissioning,improve the service life,operational reliability and safety,and reduce the later maintenance cost,a prediction method for the reliability growth of natural gas pipelines was proposed by combining the statistical model with test data to achieve the prediction on the reliability growth level of natural gas pipelines.By analyzing the commonly used reliability growth models and combining the characteristics of repairable and small sub-samples of natural gas pipelines,the Duane model and the AMSAA model were ap-plied to natural gas pipelines.Combining the point estimation,trend test,goodness-of-fit test,interval estimation and standard deviation,the reliability growth model of natural gas pipelines was obtained,and the reliability growth was predicted and veri-fied by actual example.The results show that using the Duane model and the AMSAA model in the reliability growth test of natural gas pipeline is feasible.Among them,the AMSAA model is more advantageous,and the estimated instantaneous mean time between failure(MTBF)of natural gas pipeline can reach the test target value.The test time required for predicting the MTBF to reach 2 000 h is 317 954 611.3 h,and the future 51st failure prediction interval is[848.460 9,939.622 6]with an error accuracy of 1.25%.The method has good applicability and can effectively predict the reliability growth of natural gas pipelines.The research results can provide guidance for the design improvement of natural gas pipeline.
natural gas pipelinereliability growthDuane methodAMSAA methodmean time between failure(MTBF)
NETL
NSTL
万方数据
维普
