Frequently Asked Questions
1. Can recirculation be demonstrated?
This question often comes up because density-driven convection
(DDC) technology relies on multiple treatment steps (multiple
passes through the well) to achieve a high degree of treatment. For
example, at the 10th Street Site
in Columbus, Nebraska, the
theoretically calculated recirculation ratio was 3.10. That
is, a typical water molecule entering the treatment zone
of the well from upgradient would pass through the well and
the treatment zone 3.10 times (on average) before escaping
downgradient.
Since the stripping efficiency in the well was 75% (measured
twice, at 74.6% and 75.3%) recirculation was critical to
achieve a higher overall removal efficiency. Monitoring
results from a well-designed system of monitoring wells indicated that the overall
removal efficiency actually achieved across the treatment zone (upgradient concentrations
of TCE versus downgradient concentrations) was 90.1%. The only way to
get an overall reduction greater than the single-pass stripping efficiency is
for the water to pass through the well, on average, more than once. These
reduction rates indicate an average of 3.02 trips (versus the 3.10 predicted
from theoretical calculations) through the DDC well for each water molecule entering
the treatment zone. Thus, a typical water molecule would enter the treatment
zone and pass through the DDC well, and then return to the DDC well twice before
escaping downgradient. These data demonstrate recirculation in the 10th
Street Site DDC system.
For most projects, demonstration of recirculation is hampered
by lack of a sufficient monitoring system. Only pilot studies have sufficient wells in a well-designed
configuration to allow demonstration of recirculation.   has
performed three pilot studies with adequate monitoring systems. The three
pilot studies were at the Former Naval Ammunition Depot (NAD) in Hastings, Nebraska,
at the 57th & N. Broadway Site in Wichita, Kansas, and at the 10th
Street site mentioned above. In all three cases, recirculation was demonstrated. At
the NAD and the 57th & N. Broadway sites, dye tracer studies were used to
demonstrate recirculation.
DDC, Blowerless Air Sparging, and Blowerless In-Well Stripping
technologies utilize the several passes the water makes
through a recirculating well to increase overall effectiveness.
n-Situ Groundwater Remediation, Floating Product Removal,
and Dissolved Oxygen Enhancement technologies do not rely on
recirculation to achieve adequate removal of dissolved contaminants. For
ISGR and FPR systems using granular activated carbon treatment,
for example, removal of the contaminants typically is complete
(to non-detect levels) in a single pass through the well. DOE
technology generally adds sufficient oxygen to the water on
a single pass to support complete mineralization of the contaminants
(up to 3.5 ppm hydrocarbons) to CO2 and H2O, but can saturate
the water with oxygen on each of the several passes through
the well.
2. Since a recirculating well establishes a closed-loop
zone of circulation around the well, how can water approaching
the treatment zone from upgradient enter the well? Why doesn’t
the water from upgradient simply go around the recirculating
well’s zone of circulation?
This question comes up because flowlines do not cross, and
it would seem that the three-dimensional pattern of flow
lines established around a recirculation well would effectively
bar the flow of “new” water from upgradient
entering the treatment zone of a recirculation well.
The answer lies in understanding the upgradient/downgradient
flow regime around a recirculating well. The figure below
is a plot of the results of modeling efforts for the Massachusetts
Military Reserve site on Cape Cod. This was a near sand-box
aquifer at the location modeled.
.

Figure 1 - Vertical Cross-Section of Treatment Zone - Click
to enlarge
The water exiting the upper screen flows outward in all
directions from the recirculating well, including upgradient. The
resultant obstruction to flow at the upper level of the aquifer
can be seen in the plot on the left. The water approaching
and entering the treatment zone from upgradient cannot flow
through the recirculating zone; but, it can flow under it,
as it is pulled toward the well by the low pressure induced
at the inlet screen. This is the manner in which the “new” water
enters the treatment zone, is mixed with other (recirculating)
water in the well, and begins its recirculation.
3. Is rebound a problem with recirculating well technologies?
Rebound happens when a remediation technology fails to remove
or degrade all of the contamination, as when a reagent such
as potassium permanganate, or a biological agent such as
a particular strain of bacterium, or a nutrient such as molasses,
or a physical removal method such as air sparging, fails
to reach all parts of an aquifer. Residual contamination
then diffuses out of un-remediated zones once the remediation
technology is turned off or ceases to function. Any
technology that leaves residual contamination in the aquifer
is likely to suffer rebound. Also, any site in which
there are continuing sources of contaminants, either in source
areas upgradient, in the vadose zone, in low-flow zones within
the aquifer, or in LNAPL or DNAPL layers or zones within
the aquifer, will experience rising concentrations once a
remediation system is turned off or is no longer functioning
(e.g., exhaustion of a nutrient).
At most sites where we have
installed recirculating well systems, it was not the role
of the systems to deal with concentrated source areas. Rather,
our projects have tended to be in downgradient portions of
the plume, dealing exclusively with dissolved contamination. In
this “gate-keeper” function,
a remediation system can only remediate the contamination
that reaches it. As long as contaminants continue to
arrive at the recirculating well system from upgradient,
the system will still be needed and will have to be operated
continuously. If the system is shut off with contaminants
still moving into the area from upgradient, concentrations
are going to rise.
In cases where there is residual contamination in the vadose
zone that is still leaching downward into the groundwater,
no amount of remediation of the groundwater (removal of dissolved
constituents from the groundwater) is going to remove the
source in the vadose zone. Until the source is addressed, by soil vapor
extraction, excavation, soil heating, or other means, the continuing release
of contaminants to the groundwater will necessitate continuing operation of the
recirculating well system.
In cases where there is LNAPL or DNAPL contamination, groundwater
remediation will have to continue until the LNAPL or DNAPL
is removed. Floating Product
Removal (FPR) technology can be used to remove floating product and other floating
contaminants, while treating the groundwater to non-detect levels. However,
if an FPR system is shut off before all the floating product is removed, rebound
is quite likely.
Many sites have contaminants that are trapped in silt layers,
clay layers, and other low-flow regimes within the aquifer. The contaminants entered those
zones perhaps years before, when contamination levels were higher, and are slowly
diffusing out of those zones into the higher flow regions that are now at lower
concentrations. One of the difficulties with pump-and-treat systems and
air-sparging systems is that they tend to treat water only from those layers
and zones of the aquifer that yield water most readily, the higher conductivity
zones. Yet the higher conductivity zones are often the least contaminated. Removing
water from the lower-flow, more stagnant zones of the aquifer would be preferable,
once the permeable pathways have been cleaned, but there is no way to redirect
flows with pump-and-treat or air sparging technologies.
Pump-and-treat systems, in effect, preferentially remove
water from the least contaminated zones of an aquifer and
continue to do so even after those zones have been cleaned. These systems rely primarily on diffusion to remove
contaminants from the more stagnant zones in the aquifer. While diffusion
eventually will move the contaminants to the higher conductivity zones, where
pumping can remove them, diffusion processes are very slow when compared to convective
flow processes. While waiting on diffusion, it is necessary to pump and treat
and release enormous quantities of water, at very low contamination levels, over
many years. This problem is exacerbated for contaminated
slit and clay layers that are relatively thick.
Recirculating well systems circulate water through all
zones of the aquifer by inducing vertical gradients and flows
over a large treatment area. By removing
water from one level of the aquifer and releasing it to another level of the
same aquifer, recirculating wells induce vertical gradients. The vertical
gradients induce vertical flows through the more stagnant parts of the aquifer
and flush contaminants into higher flow zones. By pushing water through
the more stagnant zones of the aquifer, recirculating wells do not rely on diffusion
alone to move trapped contaminants. Convection moves contaminants faster
than diffusion, often orders of magnitude faster.
Because recirculating well technologies move all of the
water in an aquifer, they can remove contaminants that pump-and-treat
and air sparging systems must trust to diffusion to move. In
this way, recirculating well technologies can more rapidly
remove these trapped contaminants that will, if not removed,
lead to rebound when the remediation is ended.
4. How was the projected capture width or radius of influence
calculated for my site? Has this theory ever been demonstrated
in the field?
The calculations are performed according to a method developed
several years ago, and that is commonly used with recirculating
wells. The calculations require about three hours to
perform with a pencil and calculator.
At the 10th Street
Site in Columbus, Nebraska,
a transducer study was performed to measure the actual capture
width of
the pilot well, so it could be compared to the predicted
capture width. The theoretical calculations indicated an
expected capture width of 192 feet. The transducer
study results were used to determine the gradient imposed
by the pumping of the recirculating well. The imposed
gradient and the natural gradient, were entered into a
proprietary particle-tracking model developed by   to
permit modeling using actual aquifer pressures generated
by a recirculating well system. The results of the
particle-tracking model are presented in the figure below. The
result was that the capture width of the recirculating
well, was 216 feet.

Figure 2 - Particle Tracking Model
Result - Click to enlarge
|