Experimental and numerical investigations of pneumatic venting in unsaturated soils

Throughout the last century increased production of chemicals has led to numerous accidental spills or leakages of chlorinated solvents and mineral oil products to the soil. These chemicals (NAPLs) are only slightly soluble in water and are transported through soil as a separate phase. When NAPL infiltrates through soil, it leaves behind small volumes of NAPL, residual NAPL, trapped in the soil by capillary forces. When residual NAPL is retained in the unsaturated zone it poses risks of contaminating ground water or health risk to people living
in the vicinity of the contaminated area. Hence NAPL contamination needs to be restored. Soil venting is a very common in situ technique used in restoring NAPL spills located in the vadose zone of soils. However popular, the technique of soil venting possesses limitations which result in less efficient performance and prolonged clean up times. One of the reasons for limited performance efficiency is found when dealing with layered soils of variable permeability. In such soils the efficiency of SVE is often reduced due to airflow bypassing areas of relative low permeability. Problems of flow bypassing are the subject of this study.

In the present work a new venting technique, pneumatic venting, is proposed as an alternative to conventional soil venting, when dealing with restoration of layered soils. The technique is based on imposing substantial transient pressure fronts through the vadose zone of soil. These pressure fronts are imposed by repeatedly lowering the gas phase pressure of the soil and subsequently release pressure. In this study pneumatic venting was tested in 2D tank laboratory experiments. The experiments comprised venting experiments on homogenous and heterogeneous sand packs contaminated by Trichloroethylene. Experiments showed that changing pumping schemes from constant flow venting to pneumatic venting resulted in highly improved removal efficiencies. However experiments did not reveal the mechanisms controlling the effect of pneumatic venting. In order to determine the processes responsible for the effect of pneumatic pumping, the results of the laboratory experiments were tested numerically. By use of the numerical model the dominant removal mechanism during pneumatic venting was identified as gas expansion. When pressure is lowered, gas is expanding and thereby gas/VOC fluxes are mobilized within the entire plume area of the low permeable area.

Having identified the mechanism controlling pneumatic pumping, the numerical work was extended to comprise sensitivity analysis of the efficiency of the technique towards selected parameters. The parameters comprised initial NAPL distribution, absolute permeability of the low permeable area and finally the effect of the magnitude of the total pressure drop imposed on the system. Results showed that even though removal efficiency of pneumatic pumping was sensitive towards all of these parameters, the pneumatic venting performed better than constant flow in all scenarios tested. Removal was increased by factors in the range of 2.2 to 4.8 when using pneumatic pumping in preference to constant flow venting. The largest gain was found when reducing permeability of the low permeable area.

In summery the results of the present study suggest that pneumatic venting may be an attractive alternative to constant flow venting, however further research is needed to establish the applicability of the technique to field scenarios.