Casing corrosion prediction based on grey support vector machine

Casing corrosion is serious in Dongxin oilfield, aimed at many influencing factors and complex mechanism of casing corrosion, the influencing relationships between casing corrosion rate and injection water quality and soil parameter of Dongxin oilfield were calculated by grey relational analysis methods. The analysis results show that the main factors which influence casing corrosion of Dongxin oilfield are pH of injection water quality, pH of soil, oxygen level of injection water quality, saltness of soil and water content of soil. The support vector machine algorithm was used to establish casing corrosion rate predication model and the model parameters were optimized The result indicates that casing corrosion prediction model based on support vector machine has the highest estimation precision compared with other models, the maximum relative error is about 9% and the mean relative error is less than 6%, which confirms the feasibility of this method in predicting casing corrosion rate.