Contaminant Fate and Reactive Transport in Groundwater - Dealing with Contaminated Sites: From Theory Towards Practical Application - page 869

Fig. 19.7 Geometry, conceptual model and parameters for an example of toluene transport in
groundwater

fringe have been identified in some field studies (e.g. Anneser et al. 2008 ; Lerner
et al. 2000 ) and, recently, demonstrated in a series of laboratory experiments (Bauer
et al. 2008 , 2009a , b; Rolle et al. 2008b , 2009 ).
Figure 19.7 also reports the principal parameters used in the simulations. The
extent of the model domain was chosen to represent a typical contamination sce-
nario for a shallow unconfined aquifer. To evaluate a suitable spatial discretization
for the model, a range of different grid resolutions was tested. One of the major
concerns was that artificial mixing due to numerical dispersion would cause an over-
estimation of the fringe biodegradation reaction (Cirpka et al. 1999 ). Therefore, a
fine vertical grid discretization (

0.125 m) and the HMOC solver for advec-
tion (hybrid method of characteristics, Zheng and Wang ( 1999 )) were selected to
limit numerical dispersion and to be able to accurately capture the thickness of the
reactive fringe zone. The flow simulations were carried out for steady-state condi-
tions using the numerical code MODFLOW (McDonald and Harbaugh 1988 ). A
constant hydraulic head boundary condition was used at the downstream boundary
and a constant groundwater flow condition was used at the upstream boundary.
The numerical code selected for the transport simulations was PHT3D (Prommer
et al. 2003 ). It combines the transport simulator MT3DMS (Zheng and Wang 1999 )
with the geochemical model PHREEQC-2 (Parkhurst and Appelo 1999 )forthe
computation of coupled transport and reactive processes. As initial condition for the
transport simulation, it was assumed that the aquifer was uncontaminated. The con-
taminant originates from the upper-left boundary with a constant concentration of

z =