Conceptual Frameworks for Phytoremediation to Achieve Hydrologic Goals - Introduction to Phytoremediation of Contaminated Groundwater - page 193

8.2.4 Groundwater Discharge and Potential
Evapotranspiration

In most cases, this evaluation of the removal of water by
ET p in relation to groundwater flux can occur assuming
steady-state conditions, especially for newly installed
phytoremediation systems characterized by small trees. As
the plants grow and interact more directly with the water
table, however, conditions are no longer at steady state but
are transient, as groundwater levels decline. The effect of
storage on the balance between removal of water by ET p and
addition by recharge no longer can be ignored.
Transient simulations presented in Matthews et al. (2003)
indicate that in shallow water-table aquifers that have sig-
nificant storage, a noticeable, long-term (not just one day)
drawdown may take longer to be observed than in an aquifer
with less storage. This finding has implications for the use of
groundwater fluctuations as a master criterion to determine
whether or not plant and groundwater interaction is occur-
ring at a site. For example, in aquifers with high storage
coefficients, the lack of a diurnal drawdown in a planted area
does not necessarily mean that the trees are not removing
groundwater.
The relation between ET p and groundwater flux through a
phytoremediation site ultimately will determine the size of a
phytoremediation planting, for instance when the goal is to
stop groundwater flow from occurring past a particular
boundary. In general, the larger the volume of groundwater
flux in the contaminated aquifer, the larger the planting area
should be. The planting size of a phytoremediation system is
linearly related to groundwater flux. Matthews et al. (2003)
reported that for a sandy aquifer, an increase in anisotropy,
as a ratio of flow in the horizontal relative to vertical direc-
tion from 0 to 200, resulted in an increase in planting area.
Essentially as hydraulic conductivity, K , increased, ground-
water flux increased and the planting size needed to be
increased. For a less permeable aquifer sediment, the planted
area also increased, but at a slower rate.
These and other fixed variables constrain the maximum
groundwater uptake rate at all sites where phytoremediation
is being evaluated. The ET p is a measure of the maximum
potential ET for a given area at a given location. Groundwa-
ter flux within this area also is not going to exceed the
maximum amount of flow, if conditions are at steady state.
The removal of groundwater by plant uptake also is rela-
tively fixed for trees at different ages and will approach a
maximum amount under closed-canopy conditions. To cap-
ture a given amount of groundwater flux within a fixed area
in some cases will require planting older trees that each have
a higher transpiration rate, because using smaller trees
would require installing a greater number of trees, at closer
spacing, or across a larger area than site property boundaries
permit. Or, ground covers would need to be planted to
reduce infiltration so that the trees could remove groundwa-
ter from upgradient areas only. Each of these options has
limitations, such as the increased expense for installing older

In general, for hydrologic control to occur at a site where
phytoremediation has been implemented, the volume of
groundwater that flows through the planted area in a given
time period, or flux, must be less than or equal to the flux of
water lost by ET p . If the groundwater flux exceeds that of
ET p , then a component of groundwater will be unaffected
by trees.
Darcy's Law (Chap. 4) provides a fundamental frame-
work to evaluate the groundwater flux that discharges
through a given cross-sectional area of saturated media and
forms the basis of the evaluation of hydrologic control by
phreatophytes. In his laboratory, Darcy showed that the
specific discharge, v , of water through sand-filled columns
was directly proportional to the discharge, Q , and indirectly
proportional to the cross-sectional area of the sand, A . The
specific discharge also is proportional to the head gradient,
D h , but indirectly proportional to the distance over which the

head gradient was affected,

D l . This reduces Darcy's Law

into the equality Q kiA .

In groundwater systems, the flow of water occurs only
through the interconnected pores. As such, the interference
of the soil matrix affects the volume and rate of flow, and
this discharge of groundwater has to be normalized by divid-
ing Q by the porosity, n , of the cross-sectional area of aquifer
material to obtain the true seepage velocity of groundwater,
as Q v . Porosity and seepage velocity are inversely related.
We saw above how to calculate the volume rate of ground-
water flow as

¼

Q ¼ Aik = n (8.2)

and that the Darcy velocity, Q v , is ik / n , therefore

Q ¼ AQ v

(8.3)

The vertical flow of groundwater in a shallow water-table
aquifer is dependent upon the resistance to vertical flow
relative to horizontal flow, or anisotropy, as defined in
Chap. 4. Many aquifers near the land surface that become
contaminated consist of sediments that were deposited in
horizontal, flat planes. Similarly, fining-upward sequences
of sediments associated with aquifers composed of
meandering alluvial sediments also impede vertical flow on
account of the difference in hydraulic conductivity
associated with the different sediments. If groundwater in
such sediments becomes contaminated, the same equations
can be used, but a retardation factor, f , must be considered to
account for the slower movement of a contaminant relative to
the bulk flow of groundwater on account of chemical, physi-
cal, and biological processes; this is discussed in Chap. 13.