Predicting PAHs removal from contaminated soil via subcritical water extraction using support vector regression

Abstract

Subcritical water extraction (SCWE) has been increasingly studied and applied in recent decades for the extraction of organic pollutants from contaminated soil. However, the efficacy of the SCWE technique for soil remediation depends not only on the operating parameters but also on the soil and pollutant properties. Models for predicting PAHs removal by the SCWE process are highly desirable for the process design and for facilitating a global understanding of the influence of each parameter for optimal remediation without numerous trials. In this study, a support vector regression (SVR) model was developed to predict PAHs removal from contaminated soil based on the SCWE operating parameters and soil and PAHs’ properties. Results revealed that the model exhibited a good predictability with the correlation coefficients of 0.99 and 0.93 for the training and testing datasets, respectively. The importance of input variables was in the following order: operating conditions > chemical properties > soil properties. The statistical analysis demonstrated the reliability and robustness of the developed model. Based on our findings, the SVR model is suitable for predicting the removal rate of PAHs by SCWE process.