A Statistical Analysis of Natural Gas Pipeline Risk Factors and Failure Rates

摘要

In this paper, we compared the failure data from various pipelines to observe the trend regarding failure rates, causes of failure, aging characteristics and failure rate dependence on pipeline parameters. Statistical tools were used to develop an effective risk reduction strategy based on the need to reduce pipeline incident and failure rates. Five main safety risk factors: pipeline age, maximum allowable operating pressure (MAOP), time dependent factors (corrosion and material degradation) and time independent factors (third party damage, incorrect operation, material defect) were highlighted and statistically analyzed to demonstrate their effects on pipeline failure rate. An inverse transformation analysis of variance results for the failure rate model showed four risk factors A: Age (0.0001), B: MAOP (0.0003), D: Time Independent (0.0191) and E: Time Dependent (0.0022) to be significant since their obtained p-values was less than 0.05. From this we inferred that Factors A: Age, B: MAOP, D: Time Independent and E: Time Dependent are major contributing factors in the pipeline failure rates with R2 value - 0.83, prediction error sum of squares (PRESS) - 32.95 and coefficient of variance (CV) - 21.82. Relative scoring index approach and an aggregated weighting technique was applied to sort the grouped factors. The statistically developed safety plots showed that the effect on failure frequency of factors such as pipeline age, MAOP, corrosion (internal &external), and how the failure rates for pipelines above four decades of installation were affected by corrosion damage. This research identified the need for the urgent assessment of aging pipelines due to the high risk of corrosion and cracking. The model was validated using various case studies from history to demonstrate that the resulting interactions were similar to the hypothetical failure rate curve. The results of the interactions of the risk factors could provide risk and hazard prevention personnel with a better understanding of gas pipeline maintenance activities.