Please use this identifier to cite or link to this item:
|Title:||Application of a semianalytical model to TNT transport in laboratory soil columns|
|Authors:||Louisiana State University (Baton Rouge, La.)|
Installation Restoration Research Program (U.S.)
Myers, Tommy E.
Townsend, Daniel M.
Hill, Benjamin C.
|Publisher:||Environmental Laboratory (U.S.)|
Engineer Research and Development Center (U.S.)
Abstract: Subsurface contamination by 2,4,6-trinitrotoluene (TNT) poses a threat to groundwater resources at many military installations involved in the manufacture and packing of TNT. Technical guidance for modeling the subsurface transport of TNT is needed to support Department of Defense goals for cleanup at these sites. Important aspects of this guidance include identification of significant processes involved, development of descriptors for those processes, and estimation of parameters used to quantify descriptors. This report describes application of the one-dimensional, semianalytical solute transport model to laboratory soil column (LSC) data for the analysis of TNT breakthrough curves (BTCs) and elution curves. The semianalytical model incorporates linear and nonlinear reaction terms into the one-dimensional advection-dispersion equation for solute transport and thereby allows more complicated process descriptors to be evaluated than do the available analytical models. Like the available models, the semianalytical model is readily implemented on desktop computers using commercially available mathematical software and is simpler to use than numerical models. The semianalytical model was applied to TNT BTCs obtained from four soils and to a TNT elution curve for a contaminated soil from a military installation. Process descriptors for sorption and transformation were investigated using the semianalytical model, and parameters for quantifying the process descriptors were estimated by fitting the semianalytical model to the LSC data. TNT BTCs for three of four soils were simulated adequately with equilibrium sorption and pseudo first-order transformation. Nonlinear descriptors for sorption (Freundlich and Langmuir) provided better fits to the data than a linear sorption descriptor, primarily because of the ability of the nonlinear models to capture tailing observed in the BTCs. Differences between the Freundlich and Langmuir fits were insignificant. The TNT BTC for one soil showed significant asymmetry on both rising and declining limbs that could not be adequately simulated by equilibrium sorption (linear and nonlinear) and pseudo first-order transformation. The elution curve for the contaminated soil could not be simulated satisfactorily with equilibrium sorption and pseudo first-order transformation alone. A descriptor for TNT dissolution was needed in addition to equilibrium sorption and psuedo first-order transformation to adequately simulate the contaminated soil elution curve. The one-dimensional semianalytical model is recommended for application to laboratory soil column data and the development of process descriptors from such data because the model is easy to use and allows for evaluation of complicated process descriptors. Although the model is one-dimensional and therefore has limited applicability to field problems, process descriptors developed from LSC data using the model are anticipated to be useful for structuring two- and three-dimensional numerical models of TNT subsurface transport.
|Appears in Collections:||Technical Report|