Accurate numerical simulation of the movement of dissolved chemicals (solutes) in groundwater requires definition of chemical properties such as dispersion, sorption and species parameters; as well as identification of transport boundaries to add solute to the flow system, initial contaminant conditions, and a flow solution. The flow solution provides groundwater velocities to calculate advection, which is typically the dominant transport process for most groundwater systems. However, dispersion, diffusion and sorption all affect the movement of a contaminant plume.
These processes are most often modeled for dissolved solutes using MT3DMS. Visual MODFLOW Flex provides the interface to assign: transport boundaries for sources or sinks of contamination, distributed initial concentration values, and rock/soil properties such as porosity or dispersivity.
Contaminants subject to sorption mechanisms are temporarily retained in the soil, so they move at a lower velocity than groundwater. The retardation coefficient in groundwater refers to the ratio of the distance travelled by a dissolved chemical (solute) in relation to the distance travelled by water within the same time interval. The higher Rf, the lower the plume will spread. There are three equilibrium sorption isotherm models available in MT3DMS to describe this process: Linear, Langmuir and Freundlich.
These isotherms describe the partition process between sorbed and dissolved phases, with the linear model being the most commonly used because groundwater typically has low concentrations. The linear model uses a parameter called the distribution coefficient (Kd) to calculate the retardation of the solute front due to sorption as it passes through the soil matrix, as well as to quantify the amount of mass in dissolved and sorbed phases.
Solutes in groundwater have a wide range of soil-to-solute-specific distribution coefficients. If a contaminant source contains multiple solutes, then each will have a retardation factor typified by the Kd for that solute and there will be a number of solute fronts (Fetter, 2001). The higher the Kd value assigned for a specific species, the more mass will be adsorbed to the soil and the slower the solute will travel compared to groundwater (i.e. it will have a higher retardation value). MT3DMS calculates the dissolved solute concentration I in each node; as a result, if you have a Kd that is higher than it should be, it may appear that the contaminant plume is moving more slowly than expected. In addition, if your model has a low hydraulic gradient it may even appear that the chemicals are not moving at all.
Generally, if you have a contaminant transport model and the contaminant plume is not migrating as fast as you expect it to, check your Kdvalues, they may be too high. The most common reason for an incorrect Kd is errors in the conversion of units when calculating Kdvalues. When calculating Kd, where Kd= Foc*Koc, be careful to use the correct units. The units must be consistent and converted to the model units before entering the Kd value into the model. Modifying the Kd value can substantially affect plume migration in your ground water model. Koc values come in many different units such as L/mg or L/μg while models typically use m3/Kg (in Visual MODFLOW you can choose other units that are later transformed into consistent units for MT3DMS).
So, if your plume is not moving, you probably made a mistake in converting units which led to a Kd that is too high.
Often site-specific values are not available for use in groundwater models; reviewing and using peer-reviewed values from literature is a recommended best-practice to alleviate some of the uncertainty in your model (US EPA, August 1999).
References:
Fetter, C.W. 2001. Applied Hydrogeology. New Jersey: Prentice-Hall, 600 pages
US Environmental Protection Agency. August 1999. Understanding Variation in Partition Coefficient, Kd, Values. Washington, D.C.