PubAg

Main content area

Transient multicomponent gas-phase transport of volatile organic chemicals in porous media

Author:
Amali, S., Rolston, D.E., Yamaguchi, T.
Source:
Journal of environmental quality 1996 v.25 no.5 pp. 1041-1047
ISSN:
0047-2425
Subject:
adsorption, air, benzene, equations, groundwater, mass transfer, models, porous media, prediction, sand, soil, sorption isotherms, vapors, volatile organic compounds
Abstract:
Liquid mixtures of volatile organic chemicals (VOCs) contaminating many soil and groundwater sites give rise to multicomponent mixtures of their vapors. Column studies were conducted to characterize multicomponent diffusive transport of TCE and benzene vapors through air-dry sand. The vapors diffused from their liquid mixture source at one end of the column to the other end where their concentration was maintained at zero by flowing clean air across the column. Simulations based on Stefan-Maxwell equations of vapor diffusion predicted that the TCE and benzene multicomponent steady-state flux densities should be 6.7 and 5.5% higher than the predictions based on Fick's law, respectively. The measured steady-state multicomponent mass flux densities were higher than Fick's law estimates by 4.5 to 10.6 and 3.9 to 10.0% for TCE and benzene, respectively. These results show that multicomponent equations should be used to predict the steady-state flux densities in a multicomponent VOC vapor mixture. During the transient phase of the experiments, use of Fick's law led to adequate predictions of flux density and concentration. A transient transport model based on Fick's law and a multicomponent, nonlinear adsorption model predicted measured TCE and benzene concentrations along the column in both experiments more accurately than a model based on a single-species linear isotherm, which underpredicted vapor concentrations for both species. The magnitude of the underestimation was less for benzene, which is less strongly adsorbed at low concentrations than TCE. The linear adsorption isotherm adequately predicted concentrations at the early stages of the experiment when low concentrations were predominant.
Agid:
2645465