基于主分量法的管道腐蚀评价——以川气东送管道为例

Pipeline corrosivity assessment based on principal component analysis- Take the Sichuan-East Gas Transmission Pipeline as an example

川气东送管道沿线土壤环境复杂,由管道腐蚀造成的事故时有发生,对管道沿线进行土壤腐蚀评价能够有效地预测腐蚀情况,并及时采取措施。采用主分量分析法,将统计分析软件SPSS引入管道腐蚀等级评价中,从众多影响腐蚀的因素中剔除含重复信息或测量误差太大的因素,找到最佳的关键因素组合,得到土壤腐蚀性的控制因素,对沿线土壤腐蚀等级的评价结果的准确率可以达到80%以上,从而对油气管道腐蚀评价及日常管理维护提供技术支撑。主分量分析方法可应用于数据量较大、影响因素比较多的腐蚀等级评价中,特别是对于由于某种影响因素导致测量误差比较大的情况,该方法效果很好。

Soil conditions around the Sichuan-East Gas Transmission Pipeline are quite complicated. With accidents induced by corrosion of pipelines can be observed frequently, it is necessary to perform effective assessment of soil corrosivity around the pipeline to determine expected pipeline corrosion and adopt necessary countermeasures in timely manner. Through application of principal component analysis （PCA）, the statistical and analytical software, SPSS, has been deployed for assessment of pipeline corrosion. By eliminating repeated data or factors with significant errors from potential factors that may impact corrosivity, optimal combination of key factors can be highlighted to determine controlling factors for soil corrosivity. By using such factors, assessments for corrosion capacities of soil around the pipeline can be predicted with accuracy over 80%. These results may provide technical supports for assessment over corrosivity and routine maintenance of oil/gas pipelines. The PCA can be used for assessment of corrosivity involving huge quantity of data, multiple influencing factors. It is worth mentioning that the method displayed satisfactory performances in cases with significant errors induced by certain factors. （1 Figure, 8 Tables, 12 References）