A statistical predictive model to estimate the time dependence of metal loss (ML) for buried pipelines has been developed considering the physical and chemical properties of soil. The parameters for this model include pH, chloride content, caliphate content (SO), sulfide content, organic content (ORG), resistivity (RE), moisture content (WC), clay content (CC), plasticity index (PI), and particle size distribution. The power law-based time dependence of the ML was modeled as P = ktv, where t is the time exposure, k is the metal loss coefficient, and v is the corrosion growth pattern. The results were analyzed using statistical methods such as exploratory data analysis (EDA), single linear regression (SLR), principal component analysis (PCA), and multiple linear regression (MLR). The model revealed that chloride (CL), resistivity (RE), organic content (ORG), moisture content (WC), and pH were the most influential variables on k, while caliphate content (SO), plasticity index (PI), and clay content (CC) appear to be influential toward v. The predictive corrosion model based on data from a real site has yielded a reasonable prediction of metal mass loss, with an R2 score of 0.89. This research has introduced innovative ways to model the corrosion growth for an underground pipeline environment using measured metal loss from multiple pipeline installation sites. The model enables predictions of potential metal mass loss and hence the level of soil corrosivity for Malaysia.ud
展开▼
机译：考虑到土壤的物理和化学特性，已经开发出一种统计预测模型来估计埋藏管道的金属损失（ML）的时间依赖性。该模型的参数包括pH值，氯化物含量，哈里发酸盐含量（SO），硫化物含量，有机含量（ORG），电阻率（RE），水分含量（WC），粘土含量（CC），可塑性指数（PI）和粒度分布。 ML的基于幂律的时间依赖性建模为P = ktv，其中t是时间暴露，k是金属损耗系数，v是腐蚀增长方式。使用诸如探索性数据分析（EDA），单线性回归（SLR），主成分分析（PCA）和多元线性回归（MLR）等统计方法对结果进行分析。该模型显示，氯化物（CL），电阻率（RE），有机物含量（ORG），水分含量（WC）和pH值是对k影响最大的变量，而哈里发盐含量（SO），可塑性指数（PI）和粘土含量（CC）似乎对v有影响。基于实际数据的预测腐蚀模型已得出合理的金属质量损失预测，R2值为0.89。这项研究引入了创新的方法，该方法使用来自多个管道安装地点的测量到的金属损失来模拟地下管道环境的腐蚀增长。该模型可以预测潜在的金属质量损失，从而预测马来西亚的土壤腐蚀性水平。 ud
展开▼
