A DISCRETE BLOB MODEL OF CONTAMINANT TRANSPORT IN GROUNDWATER WITH TRAPPED NONAQUEOUS PHASE LIQUIDS

Citation
Xq. Yang et al., A DISCRETE BLOB MODEL OF CONTAMINANT TRANSPORT IN GROUNDWATER WITH TRAPPED NONAQUEOUS PHASE LIQUIDS, Chemical engineering communications, 154, 1996, pp. 33-57
Citations number
53
ISSN journal
00986445
Volume
154
Year of publication
1996
Pages
33 - 57
Database
ISI
SICI code
0098-6445(1996)154:<33:ADBMOC>2.0.ZU;2-Q
Abstract
The present work describes a model developed for multiphase transport in the subsurface. The system under consideration comprises three phas es including the immobile solid phase composed of soil grains, the aqu eous phase flowing through the bed of soil grains, and the non-aqueous phase of liquids (NAPLs) that may consist of several mutually soluble compounds. Numerous field data and experimental results have indicate d that the residual nonwetting fluids of the NAPLs in groundwater syst ems are trapped, i.e., completely surrounded by the wetting aqueous ph ase. In the model, therefore, the NAPLs are treated as discrete blobs with generalized local-size distribution, while the aqueous phase is a ssumed to be a continuum. In addition, the model takes into account th e surface area-to-volume ratio and the ratio of the aqueous contacting area to the overall surface area of NAPL blobs. Interactions between the two liquid phases manifest themselves in the governing equations o f the aqueous phase as an integral over all NAPL blobs. The rate-contr olling factors in the transport process have been analyzed. By resorti ng to a semi-implicit finite-difference algorithm, numerical studies h ave been carried out to examine characteristics of the system, such as time-dependent concentration profiles, remaining quantities of the co ntaminant, and variations of the NAPL blob size. The results of simula tion reveal that the initial blob size of the NAPLs and hydrodynamic c onditions profoundly affect the rates of dissolution and transport; de sorption and dissolution occur simultaneously, but the dissolution is completed first; and the NAPL attenuates via a moving front in the soi l bed under convection-dispersion control. The model predictions for t oluene dissolution in a glass bead column agree well with the availabl e experimental data.