Reliability Analysis of Corrosion Affected Underground Steel Pipes Considering Multiple Failure Modes and Their Stochastic Correlations

This paper presents a methodology for determining the probability of system failure of corrosion affected steel pipelines with multiple failure modes, considering the correlation of the load effect process at different time points and the correlation between different failure modes. The failure modes included fracture, strength failure, deflection and buckling and they were all random process and correlations. The first passage probability theory was employed to quantify the probability of pipe failure. The Monte Carlo simulation calculated the mean function μL(t) and the standard deviation σL(t) loading effect at each time point, and the correlation matrix contained the correlation coefficients between the different failure modes as a function of time, and calculated the failure probability of each failure mode, and then determined the failure probability of the system by considering the correlation between different failure modes. A case study was presented to illustrate the proposed methodology, followed by a sensitivity analysis to investigate the effects of key random variable son the probability of pipe failure. In many influencing factors, the multiplier constant k, the exponential constant n and the d thickness of the corrosion model had the greatest influence on the failure probability. With the increase of pipeline service life, n and d tend to increase. The upper limit of n value could reach more than 0.8 under different failure modes, and d could increase to more than ‒0.2; at the same time, k decreased year by year, and its lowest value could be reduced to about 0.2. The reliability algorithm considering time-varying factors is an effective tool for the failure evaluation of buried steel pipes under the influence of corrosion. The failure prediction of buried steel pipes under the action of corrosion and the determination of the autocorrelation coefficient of the load process at the relevant position of the pipeline in the time dimension and the correlation between different failure modes can improve the accuracy of pipeline failure probability prediction. © 2022, Chongqing Wujiu Periodicals Press. All rights reserved.