Mobilization of Non-aqueous Phase Liquid by Flowing Gas Bubbles in Porous Media
Gas bubbles can significantly affect contaminant transport in both sediments and groundwater systems. In some circumstances, gas bubble mobilization can result in the transport of non-aqueous phase liquids (NAPLs) as coatings or films on the gas bubble surface, which could move a greater mass of volatile organic compounds (VOCs) through a porous medium than what might be transported as volatized, gas-phase mass. Despite this potential enhanced transport, whose outcomes include the appearance of oil sheens created as gas bubbles break at the water surface after being transported out of contaminated sediments, there has been little controlled laboratory study to determine factors that may control bubble-facilitated NAPL transport in porous media. Laboratory experiments conducted in thin, two-dimensional flow cells packed with medium sand showed that both trapped LNAPL (heptane, decane, soltrol 100) and DNAPL (creosote) could be mobilized when subjected to gas flow. In these experiments, LNAPL mobilization originated from the top of the source zone and moved along the gas flow pathway as a clearly visible finger (as high as 11.3 cm). When the emplaced upper sand layer was too thin (5.4 cm), the breakthrough of this LNAPL finger occurred, resulting in contamination of the clean water above the sand. LNAPL fingers occurred predominantly during relaxation and gas redistribution (slow discontinuous gas flow) after air injection was stopped in the 10 mL/min experiments. Observed pore-scale double displacements were considered as the main reason for the LNAPL finger development. In comparison, creosote was mobilized as either thin films on the bubble surface or tails below the bubbles. Petroleum hydrocarbon (PHC) concentrations measured above the clean sand indicated that creosote was constantly released out of the sand pack over 8 hours in the short-term experiments, with more mass being released through thinner and finer sand layers. In the long-term experiment, clearly visible evidence suggested significant depletion (~69%) of the DNAPL source zone after 30 days of gas injection. The similar ratios of PAHs in the transported and original source creosote confirmed the stable transport of creosote as a whole in the sand layer without breakdown rather than by mass transfer to the gas bubbles.