Ethanol flushing of gasoline residuals, microscale and field scale experiments

Abstract

Gasoline residuals formed by either removal of the gasoline in free phase during cleanup processes or by seasonal variation of the water table are a very common and lasting source of contamination of soil and groundwater. One method to remove these residuals at or below the water table is to pass or flush alcohol through the residuals and to remove the gasoline in the alcohol phase.

The study of the gasoline-water-ethanol interactions showed a relatively large 2-phase regino, with ethanol partitioning preferentially to the aqueous phase. The interfacial tension (IFT) is reduced as ethanol content increases and a miscible system, well suite to flushing, can always be obtained for ethanol concentrations above 88% wt.

Visualization experiments using glass micromodels showed that IFT differences are responsible for gasoline removal by a mechanism described as blob ejection, where the tip of the trapped blob or cluster of blobs that is in contact with ethanol has its IFT reduced, and the blob is ejected into the flowing ethanol rich fluid. This phenomenon facilities further dissolution of partially ejected blos into the flushing fluids.

Using sintered glass beads models, the horizontal direction was chosen for field trial of ethanol flushing. Pore scale instabilities were responsible for fingering development during the vertically downward flushing. Larger scale instabilities were responsible for fingering during vertically upward flushing. The model was completely cleaned up after 2-3 pore volumes of ethanol during lateral flushing.

During the field experiment, the density difference between water and ethanol caused the segregation of the fluids, with ethanol flowing upwards in the groundwater system. Although most of the injected ethanol missed the target region of residuals, 36% wt. of the emplaced gasoline was removed from the aquifer, and 45% wt. ethanol was recovered. Further laboratory experiments using a sintered glass beads model showed that the buoyancy effects can be used for the precise delivery of ethanol to the target regions by adjusting the level of the water table within the zone of residuals. By doing so, complete removal of the gaosline residuals was achieved in the laboratory after about 2 pore volumes of ethanol was injected.