Subsurface transport of non-aqueous phase liquids: experimental and modeling studies

Abstract

The presence of a non-aqueous liquid phase in the subsurface often controls the rate and magnitude of groundwater contamination. This project focused on one dimensional infiltration of a non-aqueous phase liquid through an unsaturated zone. A bench scale model was developed in the laboratory by using a plexiglass box filled with sand. Two different types of oils were spilled on the soil surface in the center of the box and the movement of the oils was recorded as a function of time. It was observed that the density and viscosity dominate the transport of NAPL (Non-aqueous Phase Liquid) in the subsurface. The NAPL migration was affected by the volume of release, area of infiltration, elapsed time, properties of both NAPL and subsurface media, and subsurface flow conditions.

A #2 diesel oil spill at Lordstown, OH was also investigated to study contaminant transport in the vadose zone. A computer model called HSSM (Hydrocarbon Spill Screening Model), which was developed to study the LNAPL (Light Non-Aquesous Phase Liquid) transport in the unsaturated zone, was used to predict the impact of the spill on the groundwater. The model was run at two locations in the field, initially at the spill location and later at a sump which is 99 m away from the spill location. Predictions of the model at the sump were identical with the field conditions. The model was run at the spill location to predict the concentration of TPH (Total Petroleum Hydrocarbons) in nine monitoring wells. Site measurements at these nine wells in Nov. 1991 and Nov. 1992 showed no traces of TPH. Water table at the time was 4.2 m (14') deep.Model predictions of TPH matched the site measurements. The model predicts that the pollutant reaches the water table 26 months after the spill occurred. Later in October 1992, two new monitoring wells were installed. By that the time water table had risen 1.2 m, and the measured concentration of TPH was significantly higher. The model could not predict the concentration in these wells due to the site conditions. This model can be used as a screening model for emergency purposes and for initial site assessment studies. By estimating a few key parameters impact on the groundwater can be assessed.