Conceptual Frameworks for the Phytoremediation of Groundwater Contamination - Introduction to Phytoremediation of Contaminated Groundwater - page 347

Fig. 14.2 The water budget for a
contaminated groundwater site in
Florida (Modified from Halford
1998).

Fig. 14.3 The simulated effect
of existing trees and transpiration
(black arrows) on shallow and
deep groundwater flowlines that
discharge to Lake Druid, shown
as the flow of simulated particles,
at a VOC plume site in Florida
(Modified from Halford 1998).
One foot is equivalent to 0.305 m.

boundaries, called boundary effects, the model was extended
beyond that of the site property boundaries. Vertically, the
shallow aquifer was simulated as three layers, in order to
account for groundwater and surface-water interactions. The
model input for hydrogeologic characteristics was deter-
mined from an aquifer test conducted on site to determine
hydraulic conductivity (average was 33 ft/day (10 m/day)).
The discharge of groundwater to the lake also was simulated.
Even though MODFLOW is a groundwater-flow model,
surface-water features can be simulated. The lake and
streams at the site were simulated using the river package
of MODFLOW. All surface-water features were simulated
as gaining,

Because the area between the source of contaminants and
discharge was forested with native vegetation, MODFLOW
was used to simulate the impact that the trees would have on
groundwater flow. Evapotranspiration in this model was
simulated not by using the ET package but by assuming
that recharge was reduced in those areas that had trees
relative to areas that received recharge but did not have trees.
The model results suggest that the existing vegetation
removes only a small part of the overall groundwater flux
from the source area to the surface-water body. Groundwater
discharge to the lake was estimated to be about 20 gal/min
(75 L/min) and groundwater uptake by ET was simulated to
be less than 4 gal/min (15 L/min) (Fig. 14.3 ). indicating areas of groundwater discharge.