Infiltration of organic liquids in unsaturated sands: Comparison of experimental measurements with scaled and unscaled analytical solutions.

Abstract

This study evaluated how infiltration behavior varied as a function of liquid and soil properties; how precisely four infiltration models originally developed for water described the infiltration of organic liquids; and how precisely scaling techniques generated the model parameters for organic liquid-air systems from those for water-air systems. To achieve these goals, a fully instrumented lab-scale sand-column was constructed, to allow all parameters of the infiltration equations to be measured during the flow experiments. The column was equipped with a fast-counting gamma-ray apparatus to evaluate the saturation and to track the position of the wetting front versus time; with tensiometers to measure the capillary pressure as a function of position and time; and an electronic system to maintain a constant height of ponding and to measure the rate of liquid delivery. Infiltration experiments for water, n-hexanol, ethylene glycol and 4-chlorotoluene, were run in three air-dry sands. Brutsaert's model was considered to be the best of those examined here because its predictions were most accurate and it is explicit in position of the front instead of explicit in time. For 4-chlorotoluene, Brutsaert's model predicted the data fairly well despite the unique flow behavior of this liquid. For 4-chlorotoluene, the moisture profile was saturated behind the wetting front by only 85% and the hydraulic conductivity was 62% smaller than the saturated hydraulic conductivity. Prediction of the infiltration of organic liquids using parameters obtained by scaling water data, using the Miller, Youngs and Prince, and Youngs scaling technique, compared well with the experimental data yielding relative deviations of 6-14%. Larger deviations of 21-35% were observed for two experiments that had significant air entrapment. Air entrapment seemed to be correlated with liquid mobility. Organic liquid mobility larger or much smaller in comparison to that of water appears to be associated with significant air entrapment, resulting in smaller precision in the prediction of infiltration based on water data.