DISSOLUTION OF TRAPPED LIGHT NON-AQUEOUS PHASE LIQUID IN THE PRESENCE OF TRAPPED GAS
MetadataShow full item record
The dissolution of residual, or trapped, light non-aqueous phase liquid (LNAPL) is an important process that controls, in part, many LNAPL contaminated site remediation. Smear zones can be created in LNAPL source zones by water table fluctuation, trapping both residual LNAPL and gas below the water table to create a three-phase gas-LNAPL-water system. Two sets of laboratory experiments were conducted in this study to investigate the effect of trapped gas on the distribution and dissolution of trapped LNAPL. The first set involved visualization experiments where heptane was spilled in a two-dimensional (2-D) cell. Observations suggested that both heptane and air were trapped during an experiment with a large water table fluctuation, and consisted of singlets, doublets and multi-pore ganglions. Heptane was distributed as very thin films around most of the trapped air, and as discontinuous lenses, most of which were connected to air. The second set of experiments consisted of toluene dissolution experiments in short one dimensional (1-D) columns. Results showed that toluene dissolution was faster in an air-toluene-water system than in a toluene-water system. Toluene concentrations in the three-phase system were high during the early period of the dissolution, and dropped rapidly after 150-250 pore volumes, followed by long tailing concentrations. The increased toluene-water interfacial area associated with toluene films or layers in the three-phase system might be responsible for the observed dissolution behaviour. Toluene concentrations during dissolution from the two-phase system declined more slowly, and were consistently higher than those in the three-phase experiments. Modelling showed that existing two-phase empirical dissolution models, when applied to a three-phase system, over-predict early toluene concentrations, and significantly under-predict the later toluene concentrations and the time required for dissolution. The lumped mass transfer coefficient, based on the best fit of a general power law model, was found to be orders of magnitudes higher for dissolution in the three-phase system than that for the two-phase system. In conclusion, trapped LNAPL dissolution in the presence of trapped gas is substantially different than that which occurs without any gas phase, and predictions of natural attenuation and remediation performance made using two-phase models will be misleading.