NAPLs污染物指进锋面形态的指间距离预测

选取4种不同性质的NAPLs(非水相流体)污染物,分别应用分形维数方法和毛细压力函数方法对NAPLs迁移过程中有效界面张力和指进过程中指间距离的大小进行预测,同时建立二维可视化砂箱模型,模拟污染物在多孔介质中的非稳态指进迁移过程,获得指进迁移形态影像75幅. 结果表明,与试验观测值比较,对于相同迁移过程中的指间距离,毛细压力函数方法预测误差绝对值介于1%~93%,分形维数方法预测误差绝对值介于2%~25%,分形维数方法的预测误差小于毛细压力函数方法. 毛细压力函数方法对指间距离的预测随着NAPLs密度的增加而增大,对DNAPLs(重非水相流体)的预测误差明显增加,更适用于预测密度与水的密度相差不大的物质；分形维数方法的预测值与观测值呈显著线性相关(R2=0.957),相比于传统毛细压力函数方法,分形维数方法对指间距离的预测更准确,更适合用来预测不同密度NAPLs污染物的指进距离参数. 

Prediction of Finger Spacing for NAPLs Infiltrated in Porous Media

Four typical NAPLs were selected for vertical infiltration experiments conducted in two-dimensional visible sandbox apparatus. Totally 75images were captured by the digital camera to have the front infiltration morphological information. Fractal dimension was used to estimate the effective interfacial tension for finger spacing prediction. In addition, finger spacing predicted by the means of capillary pressure head was compared with the predictions by fractal dimension. The results showed that predictions by the capillary pressure head method deviated more from the observed finger spacing, with the absolute error values varied between 1%-93% while the prediction errors by fractal dimension were between 2%-25%. NAPLs denser than water have quite high prediction errors calculated by the capillary pressure head method. The predicted finger spacing showed a significant linear correlation with those observed in practice. Compared with the capillary pressure head method, the fractal dimension method predicted more precisely especially for DNAPLs, which could be used for unstable finger flow prediction in the contaminated soil risk management and in the effective design of recovery and remediation schemes.