United States
Environmental Protection
Agency
EPA 600/R-08/003 I January 2008 I www.epa.gov/ord
A Systematic Approach
for Evaluation of Capture Zones
at Pump and Treat Systems
FINAL PROJECT REPORT
Regional Flow
Target Capture Zone
Extraction
Well
Receptor
\
Actual Capture Zone
I'liirm-
Actual Capture Zone
Supply
Well
Ncar-Rivcr Mid-plume
ExtnnionWell t"
\ River I
Shallow
Aquifer
Source Area
Office of Research and Development
National Risk Management Research L
-------
EPA/600/R-08/003
January 2008
A for
at
Project Officer
David S. Burden
Ground Water and Ecosystems Restoration Division
U.S. EPA. National Risk Management Research Laboratory
Ada, OK 74820
National Risk Management Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
To download this and other EPA/ORD/NRMRL/GWERD publications visit http://www.epa.gov/ada/
Office of Research and Development
National Risk Management Research Laboratory I Ground Water and Ecosystems Restoration Division
-------
NOTICE
The U.S. Environmental Protection Agency through its Office of Research and Development and
the Office of Supcrfund Remediation and Technology Innovation funded and managed the research
described herein. This document was prepared by GeoTrans, Inc. for the U.S. Environmental Protection
Agency (U.S. EPA) under Tetra Tech Contract No. 68-W-02-034, Subcontract No. G9015.0.037.03.01,
and under Dynamac Contract No. 68-C-02-092, Subcontract No. 092580. It has been subjected to the
Agency's review and has been approved for publication as an EPA document. No official endorsement
should be inferred. Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
-------
FOREWORD
The U.S. Environmental Protection Agency is charged by Congress with protecting the Nation's land, air,
and water resources. Under a mandate of national environmental laws, the Agency strives to formulate
and implement actions leading to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, EPA's research program is providing data
and technical support for solving environmental problems today and building a science knowledge base
necessary to manage our ecological resources wisely, understand how pollutants affect our health, and
prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's center for investigation
of technological and management approaches for preventing and reducing risks from pollution that
threatens human health and the environment. The focus of the Laboratory's research program is on
methods and their cost-effectiveness for prevention and control of pollution to air, land, water, and
subsurface resources; protection of water quality in public water systems; remediation of contaminated
sites, sediments and ground water; prevention and control of indoor air pollution; and restoration of
ecosystems. NRMRL collaborates with both public and private sector partners to foster technologies that
reduce the cost of compliance and to anticipate emerging problems. NRMRL's research provides solutions
to environmental problems by: developing and promoting technologies that protect and improve the
environment; advancing scientific and engineering information to support regulatory and policy decisions;
and providing the technical support and information transfer to ensure implementation of environmental
regulations and strategies at the national, state, and community levels.
This document describes a systematic approach for performing capture zone analysis associated with
ground-water pump and treat (P&T) systems. The intended audience is technical professionals that
actually perform capture zone analyses (i.e., hydrogeologists, engineers) as well as project managers who
review those analyses and/or make decisions based on those analyses.
Stephen G. Schmelling, Director
Ground Water and Ecosystems Restoration Division
National Risk Management Research Laboratory
-------
This page left intentionally blank.
-------
CONTENTS
A. INTRODUCTION 1
B. A SYSTEMATIC APPROACH FOR CAPTURE ZONE ANALYSIS 5
C. SUMMARY. 31
D. GLOSSARY AND SELECTED ABBREVIATIONS 33
E. REFERENCES 35
APPENDIX A: EXAMPLES FOR THREE HYPOTHETICAL SITES
EXAMPLE A1 A1-1
EXAMPLE A2 A2-1
EXAMPLE A3 A3-1
APPENDIX B: EXAMPLES FOR TWO ACTUAL SITES
EXAMPLE Bl Bl-1
EXAMPLE B2 B2-1
-------
EXHIBITS
1. Six Steps for Systematic Evaluation of Capture Zones 4
2. Elements Associated with Step 1 (Prerequisites for a Capture Zone Evaluation) 5
3. Issues When Evaluating Horizontal Capture from Water Level Contour Maps 11
4. Using Specific Capacities From a Step-Drawdown Test to Estimate
Well Losses at Extraction Wells due to Turbulent Flow 17
5. Questions to Ask When Performing Simple Horizontal Capture Analyses 22
6. Items to Address after Actual Capture is Interpreted 26
7. Capture Zone Analysis - Iterative Approach 27
8. Possible Format for Presenting Results of a Capture Zone Evaluation 28
9. Examples of Summaries for Systematic Capture Zone Evaluations 29
TABLE
1. Topics Associated with Cited References 35
-------
FIGURES
1. Illustration of horizontal and vertical capture zones 1
2. Illustration of failed capture 2
3. Remedy objectives may or may not require complete hydraulic capture 7
4. Illustration of aTarget Capture Zone 8
5. Interpreting capture from water level maps 10
6. Drawdown and capture arc not the same 12
7. Gradient vector map example 14
8. Issues associated with well inefficiency and well losses at pumping well 15
9. Water level interpretation with measurement at pumping well versus near pumping well 16
10. Inward flow at boundary is hardest to achieve half-way between the pumping wells 18
11. Complication associated with water level pairs 18
12. Cross-section schematic illustrating water level pairs 19
13. Estimated flow rate calculation 20
14. Capture zone width calculation, one extraction well 21
15. Types of downgradient monitoring wells 25
-------
ACKNOWLEDGEMENTS
The U.S. EPA Office of Research and Development (ORD) wishes to express their appreciation to the
U.S. EPA Ground Water Forum for their initial conception of the idea for this document and their helpful
comments received during the writing. The Ground Water Forum was instrumental in the development
and review of this document along with ORD scientists Mr. Steve Acree, Dr. David Burden,
Dr. David Jewett, and Dr. Randall Ross.
-------
A. INTRODUCTION
This document describes a systematic approach for performing capture zone analysis associated with
ground-water pump and treat (P&T) systems. A "Capture Zone" refers to the three-dimensional region
that contributes the ground water extracted by one or more wells or drains (see Figure 1). A capture zone
in this context is equivalent to the "zone of hydraulic containment".
Illustration of Horizontal Capture Zone (Shaded) - Map View
Illustration of Vertical Capture Zone (Shaded) - Cross-Section View
Partially Penetrating ground surface
Extraction Well
Vertical capture does not encompass the entire aquifer thickness for this partially penetrating well. The top
figure does not convey this, which demonstrates the need to perform a three-dimensional analysis. If vertical
anisotropy is present (Kx > Kz), then the greater the vertical anisotropy, the shallower the vertical capture
zone will be.
Figure 1. Illustration of horizontal and vertical capture zones.
-------
If a contaminant plume is hydraulically
contained, contaminants moving
with the ground water will not spread
beyond the capture zone. Failed
capture, illustrated schematically on
Figure 2, can allow the plume to grow,
which may cause harm to receptors
and may increase the ultimate cost or
duration of the ground-water remedy.
The purpose of this document is to
present a systematic approach to
evaluating capture zones at P&T
sites. The intended audience is
technical professionals that actually
perform capture zone analyses (i.e.,
hydrogeologists, engineers) as well as
project managers who review those
analyses and/or make decisions based
on those analyses. The scope of this
document is limited to evaluating
capture in porous media and not necessarily karst
presented here may be used for such settings, but
Illustration of Failed Capture (Map View)
Target Capture Zone
Regional Flow
Extraction
Well
Actual Capture Zone
Plume
Escaped plume due to the gap between the capture zones
Figure 2. Illustration of failed capture.
or fractured rock settings. The methods and techniques
other more intensive techniques may also be required.
EPA places considerable emphasis on P&T performance and determination of whether or not these
systems are operating properly and successfully. As discussed in Elements for Effective Management of
Operating Pump and Treat Systems (U.S. EPA, 2002b), protection of human health and the environment
often requires hydraulic containment of contaminants. Capture zone analysis is the process of evaluating
field observations of hydraulic heads and ground-water chemistry to interpret the actual capture zone,
and then comparing the interpreted capture zone to a "Target Capture Zone" to determine if capture is
sufficient.
An optimization study (U.S. EPA, 2002a) of 20 "Fund-lead" P&T systems at Superfund sites concluded
that capture zones were not being adequately evaluated. At least 14 of the 20 sites did not have a clearly
defined Target Capture Zone. About half of the 20 sites had not attempted to interpret actual capture
based on water levels. Only eight of the 20 sites had a ground-water flow simulation model, and capture
zone analysis was found to be inadequate or incomplete at six of those eight. Overall, a recommendation
to improve the capture zone analysis was made for 16 of the 20 sites. The report also concluded there
was a need for improved guidance and training with respect to capture zone analysis. This document is
intended to partially address those needs.
This document is intended to be used as a companion document to Methods for Monitoring Pump-and-
Treat Performance (U.S. EPA, 1994, link provided in "References" section) when evaluating capture
zones. This document is intended to provide more detail regarding capture zone analysis, and includes
more complex examples, relative to the previous document. This document is not intended to be a
comprehensive reference for each topic presented herein nor is it a "how to" guide. However, a table
provided at the beginning of the "References" section helps guide the reader to sources of information
(cited within this document) according to specific topics.
-------
The approach presented here should be considered iterative since few sites, if any, begin the process with
sufficient field data to evaluate and confirm hydraulic containment. Monitoring wells and piezometers
are usually installed at sites to develop the site conceptual model and determine the nature and extent
of contamination. These sampling locations are typically installed prior to initiating a P&T remedy,
and may not be appropriate for evaluating plume capture. The systematic approach advocated here is
iterative in that it is advised that the practitioner obtain additional field information to address data gaps
and ambiguities if present. The completeness of the data set, including the locations and construction
of monitoring points for water levels and water quality, should be evaluated during remedial design and
throughout the performance monitoring period. Additional monitoring points should be installed to
address any data gaps that are identified.
This document primarily pertains to operating P&T systems. However, the concepts presented in this
document should also be considered during system design. In particular, an appropriate methodology for
evaluating plume capture, including requisite monitoring locations, should be developed as part of the
system design. Also, the implemented P&T system may differ substantially from the system that was
originally designed, and the following issues should be assessed:
• did the design account for system down time (i.e., when wells are not pumping)?
• did the design consider time-varying influences such as seasons, tides, irrigation, or transient
off-site pumping?
did the design account for declining well yields due to fouling, or provide for proper well
maintenance?
• did the design address geologic heterogeneities?
• did the design take into account other hydraulic boundary conditions such as a surface water
boundary or a hard rock boundary?
Such issues may impact the effectiveness of capture relative to the designed system, highlighting the need
to conduct capture zone evaluations for the operating P&T system.
Capture zone analysis should be included in plans for remedial action, Operations and Maintenance
(O&M), and/or long-term monitoring. Appropriate elements for inclusion in a performance monitoring
plan for capture zone evaluations are outlined in Section 2.5 of U.S. EPA (1994). The monitoring plan
should be evaluated and revised as appropriate as new data are collected and the site conceptual model is
improved based on interpretation of new data.
The appropriate frequency for capture zone evaluations is site-specific. Factors that should be considered
include changes in remedy pumping rates over time (and the associated time for the ground-water levels
to stabilize), the temporal nature of stresses (on-site and off-site), and the travel-time of contaminants to
potential receptors. Some examples of temporal stresses include off-site pumping wells (water supply
or irrigation), tidal influences, seasonal changes in surface water levels, and seasonal changes in net
recharge from precipitation or irrigation. Additional discussion of factors and strategies to consider when
specifying monitoring frequency for water levels and water quality is provided in Sections 2.2.1.4 and
2.2.6.3 of U.S. EPA (1994), respectively. Capture should be evaluated throughout the first year of system
operation, and on a routine basis thereafter as part of O&M. One or more capture zone evaluations per
year is appropriate at many sites due to changing conditions.
-------
This document highlights six key steps for systematically performing a capture zone evaluation (Exhibit 1).
Specific techniques to interpret the extent of capture achieved by the ground-water extraction are applied
in Steps 3 to 5. Each of these techniques is subject to limitations, and in most cases, no single line of
evidence will conclusively differentiate between successful and failed capture. Therefore, developing
"converging lines of evidence", by applying multiple techniques to evaluate capture, increases confidence
in the conclusions of the capture zone analysis. In some cases, modifications and additions to the
monitoring program may be required to provide sufficient data to conclusively differentiate between
successful and failed capture.
Exhibit 1
Six Steps for Systematic Evaluation of Capture Zones
Step 1: Review site data, site conceptual model, and remedy objectives
Step 2: Define site-specific Target Capture Zone(s)
Step 3: Interpret water levels
• potentiometric surface maps (horizontal) and water level difference maps (vertical)
• water level pairs (gradient control points)
Step 4: Perform calculations
• estimated flow rate calculation
• capture zone width calculation (can include drawdown calculation)
• modeling (analytical or numerical) to simulate water levels, in conjunction with particle tracking and/
or transport modeling
Step 5: Evaluate concentration trends
Step 6: Interpret actual capture based on Steps 1-5, compare to Target Capture Zone(s), assess
uncertainties and data gaps
These six steps for systematically evaluating capture, and the use of converging lines of evidence, are
illustrated in this document with five examples that vary in complexity. Appendix A contains three
illustrative examples, based on hypothetical sites which were developed for this document. These
hypothetical examples highlight some of the details associated with techniques for evaluating capture.
Appendix B presents example capture zone evaluations for two actual sites and demonstrates the
systematic application of the six steps. These examples are representative of many (but not all) sites. As
mentioned, this document does not apply to fractured or karst systems.
-------
B. A SYSTEMATIC APPROACH FOR CAPTURE ZONE ANALYSIS
Step 1: Review Site Data, Site Conceptual Model, and Remedy Objectives
The items listed in Exhibit 2 should be considered prerequisites for performing a capture zone analysis.
If the plume is not adequately delineated (width and/or extent), it may not be possible to establish a
meaningful Target Capture Zone (Step 2). Hydrogeologic data typically used as the basis for a capture
zone evaluation include information on stratigraphy, hydraulic conductivity (values and distribution),
hydraulic gradients (magnitude and direction), pumping/injection rates and locations, ground-water
elevations, and ground-water quality. Well construction information is important for interpreting some
of these data. In many cases, it is appropriate to review regional hydrogeologic data in addition to site-
specific data. If hydrogeologic information such as hydraulic conductivity distribution and hydraulic
gradient (magnitude and direction) are highly uncertain, then some of the techniques for evaluating
capture may be subject to an unacceptable degree of uncertainty, and additional characterization may be
appropriate.
Exhibit 2
Elements Associated with Step 1
(Prerequisites for a Capture Zone Evaluation)
Is the plume adequately delineated in three dimensions?
Is there adequate hydrogeologic information for performing capture zone evaluations?
* hydraulic conductivity values and distribution
* hydraulic gradient (magnitude and direction)
* aquifer thickness and/or saturated thickness
* pumping rates and locations
* ground-water elevation measurements
* water quality data and associated details
* well construction details
Is there a site conceptual model (not a numerical model) that adequately
* indicates the source(s) of contaminants
* describes geologic and hydrogeologic conditions
* explains observed fate and transport of constituents
* identifies potential receptors
Is the objective of the remedy clearly stated?
>• complete hydraulic containment of the plume, or
* partial hydraulic containment in conjunction with other remedies, such as Monitored Natural Attenuation
(MNA), for portions of the plume outside the Target Capture Zone
In order to develop remedy objectives and associated performance criteria for a P&T system, and
realistic means of evaluating capture zone performance with respect to these criteria, the Data Quality
Objectives (DQO) process (U.S. EPA, 2000) should be followed. The DQO process is a systematic
planning approach for data collection that is based on the scientific method. The DQO process involves
-------
identification of data gaps that may cause an erroneous decision to be made, and assessment of the cost-
benefit ratio of filling those gaps to reduce uncertainty. By using the DQO process, one can clearly define
what data and information about the remedy performance are needed and develop a data collection design
to help obtain the right type, quantity, and quality of data needed to make a sound decision about whether
or not the remedy is effective.
A site conceptual model (a text description, maps, and cross-sections that should not be confused with a
"numerical model", although a numerical model is based on a site conceptual model) should adequately
accomplish the following:
• indicate the source (s) of contaminants
• describe geologic and hydrogeologic conditions
explain observed fate and transport of constituents
• identify potential receptors
The objectives of the remedy regarding capture should then be established (Figure 3) so that an
appropriate Target Capture Zone can be specified (Step 2). Specifically, it should be determined if there
is a need for complete hydraulic containment (the definition of "capture" in this document), or if it is
acceptable to have an uncaptured portion of the plume that is below cleanup levels or is addressed by
another remedial technology. The type of remedy objective will dictate the specifics of the Target Capture
Zone.
Step 2: Define Site-Specific "Target Capture Zone"
The Target Capture Zone is defined herein as the three-dimensional zone of ground water that must
be captured by the remedy extraction wells for the hydraulic containment portion of the remedy to be
considered successful. This will depend on the site-specific remedy objectives (Step 1). The Target
Capture Zone should be clearly stated in site remedial action and monitoring plans, and illustrated on
maps and/or cross-sections when feasible. An example is schematically presented on Figure 4, with the
Target Capture Zone illustrated both horizontally and vertically.
The Target Capture Zone should be defined in terms of specific criteria, such as a specific concentration
contour or a geographical boundary along which an inward hydraulic gradient is to be established. If the
Target Capture Zone is based on a specific concentration contour, it may need to be updated over time
as concentrations change. If a variety of contaminants of concern are present, the Target Capture Zone
should consider each contaminant.
Step 3: Interpret Water Levels
Ground-water elevation measurements are used to:
• evaluate flow directions based on water level maps (horizontal or vertical)
• evaluate flow directions based on water levels at paired locations (gradient control pairs)
Both types of evaluation listed above can provide evidence regarding the extent of capture.
-------
Remedy Objectives May or May Not Require Complete Hydraulic Capture
Site 1 (Cross Section View)
Case 1: Complete Horizontal and Vertical Capture Case 2: Complete Horizontal Capture Only
Receptor Extraction Receptor
=
Horizontal capture requires
an inward gradient /
U U U
Regional
Flow
Plume
t^~~^~" — ^A
t
• • •
Vertical capture requires
an upward gradient
t Screened
Interval
Cross-Section View
portion of plume are below
cleanup levels and/or addressed
by other technologies
Case 1: Remedy objective is complete horizontal and vertical capture.
Case 2: Remedy objective is complete horizontal capture only, in conjunction with other remedial technologies for
the deeper aquifer
Site 2 (Map View)
Case 1: Capture for Entire Plume Extent Case 2: Capture for Portion of the Plume
Map View
Regional Flow
Receptor
Capture Zone
Map View
Regional Flow
Uncaptured portion below cleanup
levels and/or addressed by other technologies
Receptor
Capture Zone
Case 1: Remedy objective is horizontal containment of the entire plume
Case 2: Remedy objective is horizontal containment of the most contaminated portion of the plume, in
conjunction with other remedial technologies for the uncaptured portion
Notes:
Site 1 and Site 2 are distinct hypothetical sites and do not illustrate the same plume
Performance monitoring wells are not depicted on these schematics to maintain figure clarity
Figure 3. Remedy objectives may or may not require complete hydraulic capture.
-------
Illustration of a Target Capture Zone (Map View)
Regional
Receptor
Extractio:
Well
Target Capture Zone
Flow
Receptor
Illustration of a Target Capture Zone (Cross Section View)
Extraction
Cross-Section View
Well
Target Capture Zone
Semi-Confining Unit
Screened Interval
implies that an upward hydraulic gradient is required for this site
Flow
Figure 4. Illustration of a Target Capture Zone (map and cross-section views).
For most sites it is appropriate to analyze ground-water flow patterns in three dimensions (i.e., both
horizontal and vertical). The potential for vertical transport of contaminants to underlying or overlying
aquifers should be considered. Three-dimensionality of ground-water flow patterns in the vicinity of
pumping wells should also be considered. For instance, in the presence of partially penetrating wells
(see Figure 1), a flow divide will generally develop with respect to vertical flow, such that water at some
depth below the well screen does not flow to the extraction well. The depth of this flow divide for a
partially penetrating well depends on the vertical anisotropy. The greater the ratio of horizontal to vertical
hydraulic conductivity, the shallower the vertical capture zone will be.
When water levels are collected, it is good practice to provide the field technician with historical depth
to water data at each location, so that reasonableness of measurements can be evaluated in the field.
When anomalous data are observed, a plan to resolve discrepancies with historical data can be developed
-------
while the technician is still in the field. It is also good practice to periodically survey measuring point
elevations. For instance, changes in measuring point elevations can occur over time due to frost heaving.
In other cases, wells installed by different contractors at different times may be surveyed inconsistently.
Water Level Maps
Horizontal water level maps indicate interpreted contours of water levels within an individual
hydrostratigraphic unit. Vertical water level difference maps indicate vertical head differences or
gradients between hydrostratigraphic units. The extent of horizontal or vertical capture can subsequently
be interpreted on the basis of those maps (illustrated schematically in Figure 5):
• Horizontal Capture Analysis. Flow lines are interpreted as perpendicular lines to water
level contours (strictly valid only for isotropic systems). Horizontal capture is defined by
a bounding flow line, within which all other flow lines reach an extraction location. The
delineation of the capture zone in this manner is a derived interpretation, since water level
contours must first be interpreted from water level values.
• Vertical Capture Analysis. Water levels between adjacent hydrogeologic units are evaluated
to indicate zones of upward versus downward flow. The analysis can be based on vertical
head differences or vertical gradients (the head difference divided by the vertical distance
between measurements).
Note that "water level" and "head" are used interchangeably in this document. Contour maps interpreted
from water levels should generally include the following (some of which are not included on the
schematic illustrations within Figure 5):
the actual data values being contoured superimposed with the interpreted contour lines
(whenever feasible)
• labels for the contour lines
• an indication of any water level measurements made at extraction wells, and whether or not
they were corrected for well inefficiency and losses
locations of pumping and injection wells, ideally with rates indicated for the time period just
prior to the water level measurements
• enough basemap features to orient the reader, including the Target Capture Zone so the
success of capture can be evaluated, plus a north arrow and a scale
dashed (or otherwise identified) contour lines where data are sparse and contour lines are inferred
Interpreting horizontal capture from water level maps is subject to significant uncertainty. The issues
listed in Exhibit 3 should all be considered when interpreting horizontal capture from water level maps.
Many of these items also pertain to evaluation of vertical capture based on vertical head differences
or vertical hydraulic gradients between hydrostratigraphic units. In light of these uncertainties EPA
recommends using additional lines of evidence regarding capture to augment the evaluation of flow
directions interpreted from water level maps.
-------
Interpreting Capture From Water Level Maps
Potentiometric Surface Map: Horizontal Water Level Difference Map: Vertical
Extraction Well
A Monitoring Well
Horizontal Capture'. Can be interpreted from water level
contours by approximating the location of a "bounding
flowline", within which all other flowlines reach a
pumping well. In this example the entire plume is within
the interpreted horizontal capture zone, for the specific
hydrostratigraphic horizon evaluated.
Vertical Couture'. Can be evaluated by interpreting areas of
upward versus downward flow. In this example head
differences at well clusters were contoured, and the entire
footprint of the plume is within the area where upward flow
is interpreted. Note the number of well clusters is quite
limited.
In this example the Target Capture Zone corresponds to the plume boundary
Cross-Section Schematic for Illustrating Upward and Downward Head Differences
Area With Upward Flow Area With Downward Flow
Area With Downward Flow
Figure 5. Interpreting capture from water level maps.
-------
Exhibit 3
Issues When Evaluating Horizontal Capture from Water Level Contour Maps (Step 3)
Are the number and distribution of measurement locations adequate?
Contouring accuracy will generally increase as the number of data points increases.
Are water levels included in vicinity of extraction wells (and have well inefficiency and losses been
considered at extraction well locations)?
Ideally, water level data representative of the aquifer are obtained from piezometers located near extraction
wells. Water levels measured at an extraction well will be lower than in the surrounding aquifer material
due to well inefficiency and losses, which can lead to incorrect interpretations of capture.
Has the horizontal capture evaluation been performed individually for all pertinent horizontal units?
Care should be taken to avoid combining water level observations from multiple hydrostratigraphic units
to generate an overall water level map. Only observations collected from a specific unit should be used to
generate a water level map for evaluating horizontal capture in that unit.
Is there bias based on contouring algorithm?
Multiple interpretations of water level contours and associated flow directions are possible for one data
set by using a different contouring algorithm (or by having a different hydrogeologist contour the data
manually). The potential for alternate interpretations of water level contours should be considered when
evaluating capture based on the contours.
Is representation of transient influences adequate?
A water level map for one point in time may not be representative of water levels and flow directions at
other points in time, which may be impacted by seasons, tides, or other pumping wells with time-varying
pumping rates.
Has potential for vertical transport been neglected when evaluating horizontal capture?
Successful horizontal capture in one stratigraphic unit does not preclude impacted water from being
transported vertically to other stratigraphic units.
"Drawdown " Versus "Capture "
Drawdown is the change of water level due to ground-water extraction. It is calculated by subtracting the
water level measured under pumping conditions from the water level measured without pumping. The
"cone of depression" (i.e., the zone where drawdown is observed) caused by extraction from one or more
locations should not be confused with the capture zone associated with that extraction. As illustrated
on Figure 6, there are generally locations outside the capture zone where drawdown due to pumping is
observed. The difference between the "cone of depression" and the "capture zone" is due to the impact of
regional hydraulic gradients. The only case where the capture zone is the same as the entire area where
drawdown is observed is when the background hydraulic gradient is perfectly flat.
-------
Water Level
Contours
Drawdown and Capture are Not the Same
Drawdown Contours
K
Outline of the Cone of Depression
(zero drawdown contour)
Cap
re L jne
Cross-Section View: Difference Between Drawdown and Capture
This area has observed drawdown,
but is outside the capture zone
Downgradient Extent
of Capture Zone
Pumping
Well
Resulting Water Table
Due to Pumping
Drawdown is the change of water level due to pumping. It is calculated by subtracting water level under pumping
conditions from the water level without pumping.
Cone of Depression is the region where drawdown due to pumping is observed.
Capture Zone is the region that contributes the ground water extracted by the extraction well(s). It is a function of
the drawdown due to pumping and the background (i.e., without remedy pumping) hydraulic gradient. The
capture zone will only coincide with the cone of depression if there is zero background hydraulic gradient.
Figure 6. Drawdown and capture are not the same.
-------
Hand-Contouring versus Computer-Based Contouring
There are many different approaches to contouring measured water levels. The interpolation or mapping
approach is probably most common. Some prefer contouring by hand, while others prefer using
computer-based contouring algorithms (e.g., SURFER by Golden Software). In either case, vastly
different (yet reasonable) interpretations of flow direction and capture may be inferred from the same
water level data, based on the interpolations (between data points) and extrapolations (beyond data points)
associated with the evaluation. Whether contouring is performed by hand or is computer-based, the
results should be evaluated for hydrogeologic reasonableness.
An advantage of contouring by hand is that professional judgment and hydrogeologic insight (e.g.,
plume shape, orientation of hydrogeologic features) can be more easily incorporated into the contours.
However, hand-contouring can be time consuming, and is not very reproducible. One approach is to have
several different individuals contour the measured values, to potentially indicate different interpretations.
Computer-based contouring is generally faster and is more reproducible. Many different algorithms are
available, and each may yield different interpretations. If different algorithms cause different conclusions
regarding the success or failure of capture, it may suggest a need for additional water level measurement
points to resolve the uncertainties.
It is harder to incorporate professional judgment and hydrogeologic insight with computer-based
contouring, but it can be accomplished by augmenting the measured data with assumed values at "pseudo-
data points". The assumed values are used to force the computer-based algorithms to interpret the actual
data in a manner consistent with the insight of the user.
Contour maps should indicate (either on the map or in related text):
the software name and settings (if applicable) and specific algorithms applied
the locations and values for "pseudo-data points" where data values were assumed to augment
measured data
• any data distribution models (including trends and transformations) assumed or applied
Neither hand-contouring nor computer-based contouring using measured water levels strictly account
for the physics-based ground-water flow equation, and the underlying principles such as mass balance.
For instance, the contours interpreted from measured water level data may be inconsistent with known
or assumed spatial variation of hydraulic conductivity. The resulting hydraulic gradients (magnitude
and/or direction) interpreted from those contoured water levels may therefore also be inconsistent with
the known or assumed spatial variation of aquifer parameters. A properly constructed flow net, whether
obtained by hand or with software, leads to head contours that satisfy the ground-water flow equation and
the principle of mass balance, although only for a highly idealized situation. Basic information on flow
nets is given by Freeze and Cherry (1979) or Fetter (2001), and a more detailed presentation is provided
in Cedargren (1997).
One approach to improve interpreted contours (Wilson and Dougherty, 2002; Dougherty and Wilson,
2003; Tonkin and Larson, 2002) is to condition computer-based contouring (based on current water level
measurements and rates of extraction/injection) with assumed trends, or with the results of simulation
models of ground-water flow (analytical or numerical) that incorporate physical principles and are
consistent with the physics of ground-water flow.
-------
Gradient Vector Map Example
Hydraulic Gradient Vector Maps
Many computer-based contouring
programs can quickly generate hydraulic
gradient vector maps based on the
interpreted water level data. An example
is provided in Figure 7. Such maps can
make it easier to visualize flow directions
and gradient magnitudes.
A form of particle tracking can also be
performed based on computer-based
contours of measured water levels, but
such particle tracking is generally not
consistent with the physics associated with
the ground-water flow equation (unless the
computer-based contours are conditioned
with an underlying simulation model, as
discussed above).
Number and Distribution of Water Level
Measurements
Water level monitoring is conducted
within, at the perimeter, and downgradient
of the Target Capture Zone to interpret
ground-water flow patterns and the
associated capture zone. The number and
distribution of ground-water elevation
measurements are frequently not sufficient to interpret capture unambiguously. Additional water level
observation points, located appropriately, may be required to resolve the uncertainty.
As discussed in Section 2.2.1.3 of U.S. EPA (1994), the number of observations needed to evaluate
capture increases with site complexity and with decreasing hydraulic gradients around the perimeter
of the Target Capture Zone. However, there is no rule regarding the "correct" amount of water level
data. Contouring accuracy will generally increase as the number of data points increases. Installing
piezometers (used herein to indicate locations where only water levels are measured) is inexpensive at
many sites, and adding piezometers should be considered if the monitoring network is not sufficient to
construct water level maps with confidence.
Water Levels at Extraction Wells (Well Inefficiency and Well Losses)
The water level measured in an extraction well is typically lower than the water level in the adjacent
aquifer due to well inefficiency and well losses (see Driscoll, 1986 and Dawson and Istok, 1991). This is
schematically illustrated in Figure 8.
Well inefficiency can be caused by the following:
Figure 7. Gradient vector map example.
inappropriate drilling and/or installation of wells for the materials through which the well bore is
advanced
-------
Issues Associated with Well Inefficiency
and Well Losses at Pumping Well
Water level in piezometer
represents aquifer condition
Piezometer Screen <
N^ Water level in pumping well does not
represent aquifer condition
Cross-Section View
Figure 8. Issues associated with well inefficiency and
well losses at pumping well.
poor or inadequate development of new
wells
• biofouling (e.g., iron-fixing or sulfur-
fixing bacteria) and encrustation (e.g.,
scaling due to pH change or aeration)
for extraction wells that have been
operating for some time (usually months
or years)
Additional well losses may occur due to
turbulent flow inside the well bore and through
the well screen slots.
Using water levels at extraction wells can bias
the interpretation of capture, since the water
levels at the extraction wells used for contouring
may be much lower than water levels in the
aquifer material just outside of the well bore.
Thus, the capture zone may be interpreted to
be larger than it actually is when water levels
at the extraction wells are used for contouring.
This is illustrated in Figure 9. It can be equally
problematic to ignore water levels measured at extraction wells if no piezometers are located in the
vicinity of the extraction wells. In that case, the capture zone often is interpreted to be smaller than it
actually is. To avoid these problems, EPA recommends installing a piezometer near each extraction well.
It is also possible to install piezometers in the filter pack of extraction wells, although some causes of well
inefficiency (e.g., formation damage due to poor well construction) will not be mitigated by this approach.
If a piezometer is not available near a pumping well, a possible approach (until an appropriately located
piezometer is available) is to estimate aquifer water levels at the extraction well by correcting the
measured water level for well losses. Bierschenk (1964) and Hantush (1964) presented a graphical
method (see Exhibit 4) for determining head loss coefficients for well losses caused by turbulent flow
across the well screen, based on a plot of specific capacity versus pumping rate developed from a step-
drawdown test. However, this approach incorporates the assumption that all well inefficiency results
from turbulent flow near the well and in the well screen. Driscoll (1986) points out that other causes of
well inefficiency are not accounted for in this approach. Dougherty (2003) presents another well loss
estimation technique based on a recovery test in a pumping well. Note that well losses can change over
time due to well fouling, further complicating the issue. Again, locating piezometers near extraction wells
is much preferred to correcting water levels in extraction wells based on calculated well losses.
Vertical Head Differences versus Vertical Hydraulic Gradients
Vertical hydraulic gradient is the head difference divided by the vertical distance between measuring
points. Vertical gradients provide more information than head differences because they account for the
distance between measurements, but calculating vertical gradients can be confusing because the vertical
distance between measurements is generally not clear (because of the length of each well screen).
Also, the small numbers typically associated with gradients can be confusing. For those reasons, head
differences are often easier to work with.
-------
Head differences can be measured at specific well clusters, or they can be estimated over a region
by subtracting interpreted water level surfaces created for each hydrogeologic unit using contouring
software. Typically there are just a few locations at a site where clustered wells are available to provide
measurements of vertical head difference, and
contours based on just a few points are generally
Water Level Interpretation with Measurement
at Pumping Well versus Near Pumping Well
Measurement at Pumping Well
Using water level at the extraction well for developing
contours biases interpretation to indicate extensive
capture...
Measurement Near Pumping Well
(Not at Pumping Well)
MW-3
(120.52)
The piezometer near the extraction well indicates that
water level in the extraction well is not representative of
water level in the aquifer. Using water levels measured at
the extraction well will overestimate the capture. In this
example, no clear-cut capture is in fact apparent when the
water level measurement at the extraction well is excluded
and a water level measurement at a piezometer near the
extraction well is available.
Figure 9. Water level interpretation with measurement at
pumping well versus near pumping well.
subject to a high degree of uncertainty.
Other Potential Pitfalls Interpreting Capture
from Water Level Maps
If transient influences are present, water
level maps for multiple time periods may be
required to sufficiently evaluate the capture
zone. Examples of transient influences are
seasons (e.g., changes in net recharge due to
precipitation or irrigation), tides, or transient
pumping at other nearby wells. Such transient
influences can impact hydraulic gradients and,
hence, capture effectiveness.
Water level data measured in different
hydrostratigraphic units should generally not
be combined to generate an overall water
level map for the site. Only water level
measurements from a specific unit should be
used to generate a water level map for that unit.
Also, it is important to remember that successful
horizontal capture in one stratigraphic unit
does not preclude impacted water from being
transported vertically to other stratigraphic
units where horizontal capture may not be
achieved. Also, if extraction wells are partially
penetrating, then the potential limitations of
two-dimensional analysis of water levels should
be evaluated.
Water Level Pairs (Gradient Control Points)
Pairs of water level elevations on either side
of a boundary (horizontally or vertically) are
used to demonstrate inward flow relative to
that boundary. Examples include ground-water
elevations on either side of a hydrogeologic
boundary or property boundary, or stage
measured in a creek relative to the ground-water
elevation in the aquifer immediately adjacent
to the creek (a higher creek stage indicates no
discharge from the aquifer to the creek), or
water levels on either side of Target Capture
Zone boundary.
-------
Exhibit 4
Using Specific Capacities From a Step-Drawdown Test to Estimate
Well Losses at Extraction Wells
Ground water flow across the well screen is turbulent due to large hydraulic gradients. For this case Jacob (1950)
proposed the following expression for drawdown inside the well casing, sw:
s =BQ + CQ2
w
Where
sw = drawdown inside the well casing
SL = well loss
C = a "well coefficient", a measure of the head loss due to turbulent flow in the well screen and pump inlet
B = an "aquifer coefficient", a measure of the head loss due to laminar (Darcy) flow in the aquifer
Q = pumping rate
S 60
Q.
e
Bierschenk (1964) developed a
graphical method for determining
coefficients B and C. It is based
on a plot of the inverse of specific
capacity versus pumping rate
from a step-drawdown test, which
assumes that an equilibrium
drawdown in the pumping well
will be established during the
step-drawdown test for several
pumping rates. Rearranging the
equation provided above yields:
s /Q = CQ + B
The step-by-step
description of the procedure is as follows:
Pumping Rate Ql
\
Q2 Q3 Q4
Pumping Rate, Q (gpm)
Q5
Pumping Rate Q5
Pumping Rate Q4
Pumping Rate Q3
mping Rate Q2
Time (min)
Plot drawdown sw versus log(time) as
shown in the upper figure
For each pumping rate, record the
equilibrium drawdown at the pumping
well (sw)
Plot sw /Q versus Q on arithmetic scale as
shown in the lower figure. Fit a straight
line through the data and extend the fitted
line to a zero pumping rate. The slope of
the line is C and the y-intercept is B.
Calculate the well loss associated with a
specific pumping rate, SL = CQ2
-------
Inward Flow at Boundary is Hardest to
Achieve Half-Way Between the Pumping Wells
Outward flow at the boundary, but
flowline through the water level pair is
ultimately captured by a pumping well
Flowlines
Site Boundary
Figure 10. Inward flow at boundary is hardest to achieve
half-way between the pumping wells.
Complication Associated with
Water Level Pairs
94 1 A
94.7 A
River
94.3 A
94.6 A
94.6A
94.7
94.7 A Stages
A Water Levels
Specific water level pairs used for gradient
control points can appear to indicate a lack
of inward flow, even when capture is actually
achieved. This is illustrated for horizontal
capture in Figure 10 and Figure 11.
In Figure 10, water level pairs nearest the
pumping wells show inward flow relative to
the boundary, but water level pairs between
the pumping wells show outward flow relative
to the boundary. Nevertheless, the extraction
wells fully capture ground water between
locations A and A'.
In Figure 11, an adequate capture zone is
established to contain the plume. However,
water level pairs near the river show continued
discharge from the aquifer to the river because
the flow divide associated with the capture
zone occurs between the extraction well and
the river, and the water level pairs are located
downgradient of that flow divide.
Figures 10 and 11 illustrate that achieving
inward gradients at water level pairs used
to monitor gradient control near a boundary
generally requires more pumping than is
actually required to simply achieve adequate
capture. In each case, the water level pairs
illustrated would all show inward gradients
if the pumping rate was increased. That may
add confidence in the analysis of capture,
but may also increase the cost of treating and
discharging the water and/or potentially cause
other negative impacts (e.g., dewatering well
screens or wetlands).
Figure 12 is a cross-section view that
illustrates the types of interpretations that
can be determined using water level pairs
when enough water level data downgradient
of the extraction well are available. The top
schematic in Figure 12 definitively indicates
flow towards the river, but a specific flow
divide caused by the extraction well cannot
be interpreted (though a flow divide caused
by the extraction might still be present). The middle schematic in Figure 12 definitively indicates a flow
divide between the extraction well and the river. The bottom schematic in Figure 12 definitively indicates
inward flow from the river towards the extraction well. Note that measured water level at the extraction
well is not used in these interpretations, for reasons discussed earlier.
Inward flow near the capture zone
95
Interpreted
Capture Zone
97
99
River Stage
Measurements
Water Level
Measurements
Extraction Well
5 Water Level
Contour
Water level pairs may suggest lack of inward flow in
cases where capture is actually successful. In this case,
the flow divide caused by the well is located between the
well and river. Using water level pairs at the river may
cause an erroneous interpretation that capture is not
achieved. More pumping is required to cause inward
flow from the river to the aquifer, compared to the
pumping required to achieve capture of the plume.
Figure 11. Complication associated with water level pairs.
-------
Cross-Section Schematic Illustrating
Water Level Pairs
FLOW DIVIDE BETWEEN EXTRACTION WELL AND RIVER NOT CERTAIN
EW PZ-A PZ-B PZ-C
QROUND.WATER FLOW DIRECTION
EW EXTRACTION WELL
— ^—• WATER TABLE
FLOW DIVIDE BETWEEN EXTRACTION WELL AND RIVER
EW PZ-A PZ-B PZ-C
EW EXTRACTION WELL
QROUND-WATER FLOW DIRECTION
INWARD GRADIENT FROM RIVER
EW PZ-A PZ*
GROUND4HATER FLOW DIRECTION
EW EXTRACTION WE
PZ PEZOMETER
-•z-- WATERTABLE
Figure 12. Cross-section schematic illustrating
water level pairs.
If transient influences are present (e.g., seasonal
pumping, tides), water level pairs for multiple
time periods may be required for sufficient
evaluation.
Step 4: Perform Calculations
Specific calculations can be performed to add
additional lines of evidence regarding the extent
of capture, including the following:
simple horizontal analyses related to
capture, such as estimated flow rate
calculations and capture zone width
calculations
• modeling (analytical or numerical) to
simulate heads, in conjunction with
particle tracking and/or contaminant
transport modeling
Determining the appropriate types of calculations
to perform should be based on site complexity.
For instance, numerical simulation of heads for
evaluating capture may not be necessary for sites
with very simple hydrogeology and only minor
heterogeneity of aquifer parameters.
Simple Horizontal Analyses
The simplest (and most commonly applied)
horizontal capture zone analyses are estimated
flow rate calculations and capture zone width
calculations:
Estimated Flow Rate Calculations, illustrated in Figure 13, provide an estimate of pumping rate
required to capture the ground-water flux through the extent of the plume.
• Capture Zone Width Calculations, illustrated in Figure 14 for the case of one extraction well,
provide an estimate of capture zone width for a specific pumping rate.
Simplifying assumptions for these methods include the following:
• homogeneous, isotropic aquifer of infinite extent
• confined aquifer, uniform aquifer thickness
fully penetrating extraction well(s)
-------
Estimated Flow Rate Calculation
Assumptions:
• homogeneous, isotropic, confined aquifer of infinite
extent
• uniform aquifer thickness
• fully penetrating extraction well(s)
• uniform regional horizontal hydraulic gradient
• steady-state flow
• negligible vertical gradient
• no net recharge, or net recharge is accounted for in
regional hydraulic gradient
• no other sources of water introduced to aquifer due to
extraction (e.g., from rivers or leakage from above or
below)
Q = K-(b-w)-i- factor
Cross Section View
\7 Water table
Plume
(must use consistent units, such as "ft" for distance and "day" for time)
Where
Q
K
b
extraction rate
hydraulic conductivity
saturated thickness
w = plume width
i = regional (i.e., without remedy pumping) hydraulic gradient
factor'= "rule of thumb" is 1.5 to 2.0, intended to account for other contributions to the pumping well
such as flux from a river or induced vertical flow from other stratigraphic unit
Figure 13. Estimated flow rate calculation.
• uniform regional horizontal hydraulic gradient
• steady-state flow
• negligible vertical gradient
• no net recharge, or net recharge is accounted for in the regional hydraulic gradient
• no other sources of water to the extraction well (e.g., flux from rivers or from other aquifers),
except as represented by the "factor" in the estimated flow rate calculation
One or more of these simplifying assumptions will be violated at most sites. However, these simple
horizontal analyses can be performed in minutes, and force the practitioner to perform a basic assessment
of hydrogeologic data (e.g., hydraulic parameter values, variation of hydrogeologic parameters over space
and/or time). For those reasons, EPA recommends that these simple horizontal analyses be performed,
even though in most cases one or more of the assumptions will be violated and additional lines of
evidence from more sophisticated capture zone evaluation techniques will likely be appropriate to more
rigorously account for site-specific conditions.
-------
Capture Zone Width Calculation, One Extraction Well
Assumptions:
homogeneous, isotropic, confined aquifer of infinite
extent
• uniform aquifer thickness
fully penetrating extraction well(s)
uniform regional horizontal hydraulic gradient
• steady-state flow
• negligible vertical gradient
• no net recharge, or net recharge is accounted for in
regional hydraulic gradient
no other sources of water introduced to aquifer due
to extraction (e.g., from rivers or leakage from
above or below)
Well
-y
27ffi
tan| y
— or —
(Stagnation Point)
Where'.
Q
T
K
b
i
X
g
Ymax
YweU
I Q
X.=-QI2nTi • Y_=±Q/2Ti . _ _
(must use consistent units, such as "ft" for distance and "day" for time)
extraction rate
transmissivity, K- b
hydraulic conductivity
saturated thickness
regional (i.e., pre-remedy-pumping) hydraulic gradient
distance from the well to the downgradient end of the capture zone along the central line of the flow
direction
maximum capture zone width from the central line of the plume
capture zone width at the location of well from the central line of the plume
The above equation is used to calculate the outline of the capture zone. Solving the equation for x = 0 allows one to
calculate the distance between the dividing streamlines at the line of wells (2 • Ywell) and solving the equation for
x = co allows one to calculate the distance between the dividing streamlines far upstream from the wells (2 • Ymal). One
can also calculate the distance from the well to the stagnation point (Xg) that marks the downgradient end of the capture
zone by solving for xaty = 0. For any value ofy between 0 and Ymwc, one can calculate the corresponding x value,
allowing the outline of the capture zone to be calculated.
Figure 14. Capture zone width calculation, one extraction well.
The extraction rate "Q" in the estimated flow rate calculation incorporates a "factor" to account for other
potential contributions of water to the extraction location, such as water from a nearby creek or water from
an overlying or underlying unit. There is no scientific rule for assigning a value for the "factor", although
common practice is to assign a value between 1.5 and 2.0. Note that the variability in hydraulic conductivity
at many sites is as great or greater than the potential variability in this "factor". It is good practice to
perform the estimated flow rate calculation with several different values assumed for the "factor" (e.g., 1.0,
-------
1.5, and 2.0), to determine a range of values for the estimated flow rate required for capture. The extraction
rate "Q" in the capture zone width calculation does not account for any such "factor".
These calculations require an estimate of the regional hydraulic gradient, without the influence of
remedy pumping. Sometimes, the use of water level data obtained prior to the remedy is appropriate for
determining the regional hydraulic gradient. However, regional hydraulic gradients often change with
time. Accordingly, in some cases, the use of water level data obtained side gradient or even upgradient of
the contaminated area, collected during the remedy, may be more appropriate than pre-remedy water level
data for calculating regional hydraulic gradient.
Capture zone width calculations are most often performed assuming one extraction well. As illustrated in
Figure 14, the capture zone width calculation for one extraction well provides an estimate of capture zone
width near the extraction well (Ywdl) and far upgradient of the extraction well (Ymax), and also provides
the distance from the extraction well to the downgradient flow divide (XQ) that is often referred to as the
"stagnation point".
For cases with more than one extraction well, the capture zone width is often estimated by assigning
the total pumping rate at one centrally-located "equivalent well". Javandel and Tsang (1986) provided
solutions for one-, two-, and three-well extraction systems for cases where there is too much drawdown
if all of the extraction is applied at one well. These multi-well solutions assume the extraction wells are
located along a line perpendicular to the regional flow direction, and also assume that the total pumping
rate is divided equally among the extraction wells. Their solution provides the appropriate spacing for
such wells. However, the capture zone width far upgradient of the pumping wells (Ymax) will be nearly
identical to the case with all of the pumping assigned to one extraction well located in the center. Note
that the assumptions for this multi-well calculation are not met in many field situations. Grubb (1993)
provides a solution for a capture zone in an unconfined aquifer that can be utilized for multi-well
calculations.
Although the calculations associated with
these simple horizontal capture zone analyses
are quite easy, deciding on the actual values to
use for the calculations is not straightforward
when the parameters (e.g., hydraulic
conductivity, aquifer thickness, hydraulic
gradient) are not uniform. Performing
the calculations for reasonable ranges of
parameter values can provide upper and lower
bounds on the results.
The items highlighted in Exhibit 5 should be
addressed when using these techniques. If
answers to any of the questions in Exhibit
5 are "yes", then assumptions behind these
calculations are violated and other lines of
evidence regarding capture should be given
higher priority.
It should be stressed that these simple
horizontal capture zone calculations do not
pertain in any manner to vertical capture.
Exhibit 5
Questions to Ask When Performing
Simple Horizontal Capture Analyses (Step 4)
• Is there significant heterogeneity at this site, such
as a wide range of hydraulic conductivity due to a
buried paleochannel?
• Are there any other contributions of water to the
extraction wells (e.g., leakage from a river, leakage
from other stratigraphic units, clean water extracted
from outside the plume)?
• Do transient conditions and/or off-site stresses
exist, such as seasonal pumping at off-site
production wells?
If any answers are "yes ", the assumptions
associated with these methods are violated,
and additional lines of evidence should be examined.
-------
Modeling (Analytical or Numerical) to Simulate Heads In Conjunction with Particle Tracking
Different types of simulation models, ranging from analytical to numerical, can be applied to calculate
hydraulic heads and subsequently evaluate capture zones based on particle tracking. For instance, the
analytical-based code CAPZONE (Ohio State University) can be used to analytically construct ground-
water flow models of two-dimensional flow systems with isotropic and homogeneous confined, leaky-
confined, or unconfined flow conditions, using either the Theis equation or the Hantush-Jacob
equation. Particle tracking software (e.g., GWPATH by the Illinois State Water Survey) can then utilize
the simulated flow field to draw capture zones. WhAEM2000 (U.S. EPA) is another analytical-based code
that can be used for capture zone delineation. Numerical simulation codes such as MODFLOW (U.S.
Geological Survey) allow for simulation of more complex systems (three dimensional geometry including
aquifer heterogeneity and complex boundary conditions). Particle tracking codes (e.g., MODPATH by the
U.S. Geological Survey) can then utilize the simulated flow field to draw capture zones. Please note that
many codes are available for these types of applications, and the codes named herein are only mentioned
as examples.
A general reference for ground-water modeling and particle tracking is Anderson and Woessner (1992).
Ground water models should be calibrated to reasonably match field-measured heads and flow patterns.
Calibration is accomplished by varying parameter values, boundary conditions, and stresses until an
acceptable match with field-measured values is achieved.
Particle tracking based on simulation of heads can provide a precise delineation of both horizontal
and vertical hydraulic capture (not accounting for dispersion). Precision, however, should not be
confused with accuracy. The capture zone indicated by the particle tracking is only as accurate as the
underlying head predictions from the simulation model, which are subject to many types of uncertainty
(e.g., parameter values, boundary conditions). If the model inputs do not reasonably represent actual
conditions, there is potential for "garbage in - garbage out".
Ideally, the calibrated numerical model should subsequently be "verified" by simulating drawdown
responses to different pumping conditions, and comparing those predicted responses to field
measurements. This instills confidence that the model provides a reasonable representation of the
physical system.
Another way to "verify" the numerical model is to run forward particle tracking to show that particles
released at the location of the contaminant source reasonably account for the observed plume dimensions,
and that the downgradient plume extent is consistent with the amount of elapsed time since contaminants
were first introduced into the ground water. However, there are factors in addition to the prediction of
the flow system (including contaminant source locations and timing, dispersion, and adsorption) that
complicate such evaluations.
Tracking particles in reverse from initial locations around the extraction wells, to define the capture
zone, is a commonly used approach. However, it can lead to erroneous interpretations in two and three
dimensions. The apparent area of capture is highly controlled by the number of particles released, and
the specific locations (horizontal and vertical) where particles are released. For instance, if particles are
started at only one vertical location (e.g., the middle of the well screen) and tracked backwards, the results
may indicate that all water captured by the well comes from the aquifer screened by the well, and may not
show contributions to the well from water that may originate from aquifers above and/or below.
Another approach is to track particles backward from locations beyond the capture zone, to create a
"shadow" plot of particles that escape the capture zone (the blank area represents the capture zone).
-------
However, this method does not indicate the capture zone of each individual extraction well if there is
more than one extraction well.
Tracking particles forward in space and time from a large variety of starting points (horizontally and
vertically), and determining which of those particles reach each extraction location (i.e, wells or drains), is
generally a better way to assess the three-dimensional capture zone. The initial particle locations can be
plotted, using different symbols or colors to indicate the specific extraction location where each particle
is ultimately captured. In this manner, the capture zone of each individual extraction location can be
effectively illustrated. Producing multiple maps of this type, where each map illustrates particles starting
at a different vertical elevation, is an effective approach for illustrating the vertical capture zone of each
extraction location.
It is important to simulate pumping rates actually achieved with the P&T system, which in some cases
differ substantially from design values. It is not appropriate to simulate the maximum pumping rate if
that rate is not sustained in a continuous manner. In general, simulating the average pumping rate or
a range of pumping rates is more appropriate for evaluating current capture zones than simulating the
maximum extraction rate, except as a screening exercise.
It is also important to assess if the options selected for particle tracking, such as alternatives for removing
particles, are biasing the interpretation of capture. For instance, the particle tracking code may allow the
option of removing particles at "weak sinks" or allowing particles to pass through "weak sinks". A "weak
sink" is a model grid cell where some water is removed from the model, but where some water also flows
into one or more adjacent model cells. This could occur if an extraction well is located in a relatively
large grid cell and pumps relatively little water. The option selected by the user may greatly influence the
resulting interpretation of capture. One approach is to evaluate capture using both options, and determine
if the resulting interpretation is greatly influenced by the option selected. If so, both results can be
reported or finer model grid spacing can be considered.
In addition to providing a basis for particle tracking, ground-water modeling can bring a higher level of
understanding regarding the hydrogeology and contaminant transport at a site. Although ground-water
model results are subject to uncertainty, and ground-water modeling is not warranted at many simple
sites, EPA encourages the use of ground-water modeling at more complex sites as a tool for evaluating
and improving the site conceptual model, predicting capture zones, and evaluating alternate remediation
scenarios. However, actual field monitoring must be carried out in order to provide information necessary
to evaluate model predictions. Capture zone effectiveness is ultimately determined by field monitoring
that typically includes some combination of hydraulic head measurement and ground-water sampling and
analysis, in conjunction with field confirmation of remedy pumping rates.
Step 5: Evaluate Concentration Trends
Contaminant concentrations can be monitored at two types of locations downgradient of the Target
Capture Zone in an attempt to interpret capture (Figure 15):
sentinel wells are located downgradient of the Target Capture Zone and are not currently impacted
above background concentrations
• downgradient performance monitoring wells are located downgradient of the Target Capture
Zone and are currently impacted above background concentrations
For sentinel wells, contaminant concentrations should remain at background levels over time if capture
-------
Types of Downgradient Monitoring Wells
Uncap tured Portion Below Cleanup
Levels and/ or Addressed by Other Technologies
Downgradient
Performance
Monitoring Well
A
MW"2
Receptor
Sentinel Well
A
MW-3
MW
Regional Flow
Plume with
Continuous Source
Target Capture Zone
Concentration versus Time at
Monitoring Wells
- - MW-1 - MW-2 ....... MW-3
Within Capture Zone
Cleanup Standard
Downgradient Performance Monitoring Well
Non-Detect, plotted at half the detection limit
Sentinel Well
is successful. For downgradient performance
monitoring wells, contaminant concentrations
should decline to background levels (or below
cleanup levels) overtime if capture is successful.
The term "downgradient performance monitoring
well" is used herein to describe a specific subset
of "performance monitoring wells" that are
located downgradient of the Target Capture Zone
and are impacted by chemicals above background
concentrations. Other performance monitoring
wells might be located within the Target
Capture Zone, and might monitor hydraulic
performance and/or concentration trends. A good
understanding of contaminant release history
and plume dynamics is needed to successfully
position sentinel wells and "downgradient
performance monitoring wells".
A primary issue complicating the use of
concentration trends for evaluating capture
is illustrated in Figure 15, which presents
concentration versus time for the three
monitoring well locations (MW-1 to MW-3)
shown in the top portion of Figure 15.
This example pertains to a case with a
continuing source of dissolved contamination.
Concentrations at monitoring well MW-1
remain above background over time, because the
monitoring well is actually within the capture
zone (i.e., not downgradient of the Target Capture Zone), and therefore continues to be impacted by
contaminated water from the continuing upgradient source. However, since the actual extent of the
capture zone is not generally known, the concentration trend at MW-1 could be erroneously interpreted as
failed capture (because concentrations downgradient of the extraction well remain above background).
Interpretation of capture based on concentration trends at monitoring wells located downgradient of the
Target Capture Zone is complicated by several other factors:
there may be limited concentration data since monitoring ground-water concentrations is far more
expensive than monitoring water levels
• interpretations of concentration data related to capture may take years because ground-water flow
velocities (and associated concentration changes) are generally quite slow
• for sites with multiple hydrogeologic units that are impacted or may become impacted, it is
necessary to monitor wells in multiple hydrogeological units
• multiple releases of contaminants can result in multiple pulses in monitoring well concentration
data, which means that decreasing concentrations may be misleading if a second pulse has not yet
arrived
note: background concentration is "not detected"
Figure 15. Types of downgradient monitoring wells.
-------
In Figure 15, concentrations do not fall below cleanup standards at the downgradient performance
monitoring well (MW-2) until nearly eight years of monitoring have been completed. This is likely to
be an unacceptably long period of time for use as the sole indication of capture, highlighting the need for
other lines of evidence, such as those that measure hydraulic performance, to be used in conjunction with
this line of evidence.
Although these issues complicate interpretation of capture from concentration trends, the concentration
trends at these downgradient performance monitoring wells over time may ultimately provide the most
solid and compelling line of evidence that successful capture has actually been achieved. Therefore, both
hydraulic monitoring and chemical monitoring should usually be components of capture zone evaluations.
Estimates of capture performance based on hydraulic data allow relatively rapid assessments of system
performance that complement the more direct but longer term assessments provided by monitoring of
ground-water chemistry.
Such chemical data may also be used to assess consistency with the site conceptual model and/or with
other lines of evidence regarding interpretation of capture. In many cases, ambiguity associated with
the interpretation of such concentration data will indicate data gaps, which in turn may suggest a need to
collect additional field data to meet the data quality objectives of the capture zone evaluation.
Step 6: Interpret Actual Capture And Compare to Target Capture Zone(s)
Once multiple lines of evidence regarding capture have been evaluated, actual capture achieved by the
extraction wells should be interpreted, and the items in Exhibit 6 should be addressed. To avoid bias,
the actual capture should be interpreted independent of the Target Capture Zone (i.e., they should be
compared after the actual capture zone is interpreted).
Exhibit 6
Items to Address after
Actual Capture is Interpreted (Step 6)
Compare the interpreted capture zone to the Target Capture Zone
Does the current system achieve remedy objectives with respect to plume capture, both horizontally and
vertically?
Assess uncertainties in the interpretation of the actual capture zone
Are alternate interpretations possible that would change the conclusions as to whether or not sufficient capture
is achieved?
Assess the need for additional characterization and/or monitoring
Is there a need for additional plume delineation or additional piezometer locations to determine convincingly
whether or not actual capture is sufficient?
Evaluate the need to reduce or increase extraction rates
Should extraction rates, number of extraction wells, and/or locations be modified based on the results of the
capture zone analysis?
-------
Capture zone analysis should be an iterative process that includes the following:
• evaluate capture based on existing data
• identify any data gaps that create uncertainty in the conclusions of the capture zone analysis
• fill any data gaps that are identified (e.g., add new piezometers), and re-evaluate capture
• continue monitoring capture over time
• if capture is ever determined to not be sufficient, optimize the extraction system until capture is
sufficient
if capture is determined to be sufficient, continue routine monitoring and consider the potential to
optimize extraction locations and/or rates to reduce cost
The iterative process described above is illustrated in Exhibit 7. Increasing the pumping rates may add
confidence in the analysis of capture, but may also increase the cost of treating and discharging the water
and/or potentially cause other negative impacts (e.g., dewatering well screens or wetlands).
Exhibit 7
Capture Zone Analysis - Iterative Approach
Iterative
Evaluate capture using existing data
Are there data gaps that
make conclusion of capture
evaluation uncertain?
Complete capture zone
evaluation
No
Yes
Continue routine
monitoring
Optimize to reduce
cost
-------
Exhibit 8 presents one possible format for summarizing the results of a capture zone evaluation. It is
organized in a manner that provides the conclusions from different lines of evidence regarding capture,
and then presents overall conclusions regarding the adequacy of capture, uncertainties and data gaps, and
recommendations for future action. In most cases, the required level of detail cannot be presented in one
simple table such as Exhibit 8, and will actually be included in chapters or appendices of a report.
Exhibit 9 is an example of a similar format for summarizing capture zone evaluations with two cases: one
where capture is likely successful and one where capture is not convincingly demonstrated.
Exhibit 8
Possible Format for Presenting Results of a Capture Zone Evaluation
Line of Evidence
Is Capture Sufficient?
Comments
Water Levels
Potentiometric surface maps
Vertical head difference maps
Water level pairs
Calculations
Estimated flow rate calculations
Capture zone width calculations
Ground-water flow modeling
with particle tracking
Concentration Trends
Sentinel wells
Downgradient performance
monitoring wells
Overall Conclusion
Is capture sufficient, based on "converging lines of evidence"?
Key uncertainties/data gaps
Recommendations to collect additional data, install new monitoring wells, change current extraction rates,
change number/location of extraction wells, etc.
-------
Exhibit 9
Examples of Summaries for Systematic Capture Zone Evaluations
Example Where Capture Appears Successful
Step 1 : Review site data, site conceptual model,
remedy objectives
Step 2: Define "Target Capture Zone(s)"
Step 3 : Water level maps
Step 3 : Water level pairs
Step 4: Simple horizontal capture zone analyses
Step 4: Particle tracking
Step 5: Concentration trends
Step 6: Interpret actual capture and compare to
Target Capture Zone
Completed, all determined to be up-to-date and adequate
Updated based on revised plume map, illustrated on maps
Adequate monitoring well network exists to determine
capture; water levels indicate capture zone larger than the
Target Capture Zone, piezometers are available near each
extraction well for accurate water levels
Inward flow at all pairs along property boundary, and
vertical water level differences indicate hydraulic control
Estimated flow rate calculation suggests 50-100 gpm
should be sufficient, system currently at 100 gpm
Model calibration updated based on actual pumping
rates and drawdowns, particle tracks indicate successful
horizontal and vertical capture
Not relied upon for short term, concentrations do not
increase at sentinel wells in long term
Actual capture zone is interpreted to be larger than the
Target Capture Zone, all lines of evidence support that
conclusion, some reduction in pumping rates/locations
might be considered
Example With Many "Red Flags" - No Confidence That Capture is Successful
Step 1 : Review site data, site conceptual model,
remedy objectives
Step 2: Define "Target Capture Zone(s)"
Step 3 : Water level maps
Step 3 : Water level pairs
Step 4: Simple horizontal capture zone analyses
Step 4: Particle tracking
Step 5: Concentration trends
Step 6: Interpret actual capture and compare to
Target Capture Zone
Last plume delineation 5 years ago, unclear if remedy
objective is "cleanup" or containment
Not clearly defined, objective is simply "hydraulic
containment"
Inadequate monitoring well network exists to determine
capture, water levels indicate a "large" capture zone,
however, water levels are used at extraction wells with
no correction for well inefficiencies and losses (no
piezometers near extraction wells)
Vertical water level differences are not evaluated
Done during system design, estimated flow rate
calculation indicated 50-100 gpm would be required,
current pumping rate is 40 gpm
Not performed, no ground-water model being utilized
Evaluated but with inconclusive results
Not even possible since Target Capture Zone is not
clearly defined, conclusion of capture zone analysis
should be that there is a need to adequately address Steps
1 to 5, so that success of capture can be meaningfully
evaluated
-------
Tliis page left intentionally blank.
-------
C. SUMMARY
Capture zone analysis is the process of evaluating field observations of hydraulic heads and ground-
water chemistry to interpret the actual capture zone, and then comparing the interpreted capture zone
to a "Target Capture Zone" to determine if capture is sufficient. This document presents a systematic
approach to evaluating capture zones at P&T sites.
Six steps are suggested for a systematic capture zone evaluation:
Step 1: Review site data, site conceptual model, and remedy objectives
Step 2: Define site-specific Target Capture Zone(s)
Step 3: Interpret water levels
* potentiometric surface maps (horizontal) and water level difference maps (vertical)
• water level pairs (gradient control points)
Step 4: Perform calculations
* estimated flow rate calculation
• capture zone width calculation (can include drawdown calculation)
• modeling (analytical or numerical) to simulate water levels, in conjunction with particle
tracking and/or transport modeling
Step 5: Evaluate concentration trends
Step 6: Interpret actual capture based on steps 1-5. compare to Target Capture Zone(s), assess
uncertainties and data gaps
Specific techniques to assess the extent of capture achieved by the extraction wells are applied in Steps 3
to 5. Each of these techniques is subject to limitations, and in most cases no single line of evidence will
conclusively differentiate between successful and failed capture. Therefore, developing "converging lines
of evidence", by applying multiple techniques to evaluate capture, increases confidence in the conclusions
of the capture zone analysis.
The systematic approach advocated here is iterative in that it is advised that the practitioner obtain
additional field information to address data gaps and ambiguities if present. Along each step of the
process, the practitioner should evaluate the completeness of the data set and how to address uncertainty.
This document is intended to be used as a companion document to Methods for Monitoring Pump-and-
Treat Performance (U.S. EPA, 1994. link provided in "References" section) when evaluating capture
zones. This document is intended to provide more detail regarding capture zone analysis and includes
more complex examples than are discussed in that previous document. This document is not intended to
be a comprehensive reference for each topic presented herein nor is it a "how to" guide. However, a table
provided at the beginning of the "References" section helps guide the reader to sources of information
(cited within this document) according to specific topics.
-------
Some of the important "'big-picture" considerations when performing a capture zone evaluation include
the following:
• there should be a clearly stated remedy objective based on a site conceptual model, and an
associated three-dimensional "Target Capture Zone" that is clearly defined both horizontally and
vertically
* the evaluation should utilize as many '"converging lines of evidence" as practicable (i.e., use of
multiple techniques to evaluate capture)
the success of capture (relative to the Target Capture Zone) should be summarized in a format
that conveys the results of the different lines of evidence, identifies data gaps, and provides
recommendations (Exhibit 8 is an example of one such format)
• many of the components of a capture zone evaluation require hydrogeologic insight and expertise.
and practitioners should use the assistance of support personnel if they lack that expertise
The appropriate frequency for capture zone evaluations is site-specific. Factors that should be considered
include changes in remedy pumping rates over time (and the associated time for the ground-water levels
to stabilize), the temporal nature of stresses (on-site and off-site), and the travel-time of contaminants
to potential receptors. Capture zone analysis should be considered during system design, and should be
performed throughout the first year of system operation and on a routine basis thereafter as part of O&M.
One or more capture zone evaluations per year is appropriate at many sites due to changing conditions.
-------
D. GLOSSARY AND SELECTED ABBREVIATIONS
Capture Zone. A three-dimensional region that contributes the ground water extracted by one or more
wells or drains.
Converging Lines of Evidence. Applying multiple techniques to evaluate capture, such that confidence
in the conclusions of the capture zone analysis is increased.
Downgradient Performance Monitoring Well. Monitoring well located downgradicnt of the Target
Capture Zone and currently impacted by contaminants above background concentrations. This w7ell can
be used to monitor concentration reductions that are expected if capture is successful (note that the more
general term "performance monitoring well" may pertain to wells in other locations, such as within the
Target Capture Zone, that may monitor hydraulic performance and/or chemical performance).
Estimated Flow Rate Calculation. Used herein to refer to a calculation of ground-water flux that
requires capture, based on plume width and a set of simplifying assumptions.
Gradient Control Points. See "water level pairs".
Gradient Vector Map. A map that uses arrows to illustrate the direction of ground-water flow, and
optionally uses the length of the arrow to indicate the magnitude of the hydraulic gradient.
Head. Used interchangeably with "'water level"' in this document (see "'Water Level").
Hydraulic Gradient. The change in head over a distance (the "'magnitude") in the direction that
represents the maximum rate of head decline ("the direction").
MNA (Monitored Natural Attenuation). Refers herein to remediation of contaminants, actively
monitored, without use of an active remedy such as P&T.
Monitoring Well. Used herein to refer to a well which is used for measurement of water levels and
contaminant concentrations.
O&M. Operations and maintenance, used herein to refer to activities associated with operating and
maintaining a P&T system (does not refer to any specific period of time or regulatory status associated
with the remedy).
P&T. Pump and treat, used herein to refer to remediation systems where ground water is extracted and
treated and/or appropriately discharged
Particle Tracking. Tracing the movement of a particle of a conservative solute as it flows within the
ground-water flow system, not influenced by hydrodynamic dispersion.
-------
Piezometer. A well used to measure hydraulic water level in the subsurface, with a relatively short screen
or slotted interval such that the water level represents the specific location within the aquifer. Used in this
document to refer to a well which is used for measurement of water levels only (i.e., not for contaminant
concentrations).
Preferential Pathway. Used herein to refer to a continuous zone where ground water flows faster
relative to surrounding zones, due to aquifer heterogeneity.
Sentinel Well. Monitoring w7ell located downgradient of the Target Capture Zone that is not currently
impacted by contaminants above background concentrations.
Stagnation Point. Distance from the extraction well to the flow divide that indicates the downgradient
extent of the capture zone (in the direction of uniform background flow).
Target Capture Zone. The three-dimensional zone of ground water that must be captured by the remedy
extraction for the hydraulic containment portion of the remedy to be considered successful.
Water Level. Catenated by subtracting depth to water from a datum with a surveyed elevation, such as
the top of well casing.
Water Level Pairs. Pairs of water level elevations on either side of a boundary (horizontally or
vertically) that are used to demonstrate the direction of flow relative to that boundary.
Well Inefficiency and Well Losses. Refers to flic difference between flic water level in the extraction
well (lower) versus the water level in the aquifer immediately adjacent to the extraction well (higher) that
may result from a variety of factors.
-------
E. REFERENCES
Table 1 indicates the cited references in this document for specific topics.
Table 1. Topics Associated with Cited References
Topic
Previous EPA guidance
Flow nets
Water level maps/contouring
Well losses
Capture zone width calculation
Particle tracking based on numerical modeling
Cited
Reference #
15, 16, 17, 18
3,8,9
6, 14, 19
2,4,5,7, 11, 12
10, 13
1
The "Reference #" pertains to the number of the reference in the following "Cited References" section.
This is not intended as a comprehensive set of references for each topic, and some of the references cited
for one topic may also be valuable for other topics. This table is only intended to link the cited references
to the topics they were cited for in this document.
Cited References
1. Anderson, M.P. and W.W. Woessner, 1992. Applied Groundwater Modeling: Simulation of Flow and
Advective Transport, Academic Press, Inc.
2. Bierschenk, W.H., 1964. Determining Well Efficiency by Multiple Step-Drawdown Tests. International
Association of Scientific Hydrology, Publication 64, pp. 494-505.
3. Cedargren, H.R., 1997. Seepage, Drainage, and Flow Nets, 3rd Edition, John Wiley & Sons, Inc.
4. Dawson, K.J. and J.D. Istok, 1991. Aquifer Testing: Design and Analysis of Pumping and Slug Tests,
Lewis Publishers, Inc.
5. Dougherty, D.E., 2003. Assessing The Amount of Well Loss in A Pumping Well From
A Recovery Test, Subterranean Research, Inc. (http://www.subterra.com/downloads/
WellEfficiencyAndWellLosses.pdf)
6. Dougherty, D.E. and D.A. Wilson, 2003. Using On-Going Monitoring Data and Site Models to
Evaluate Performance of Remediation Systems, Conference Proceedings of MODFLOW and More
2003: Understanding through Modeling, International Ground Water Modeling Center.
-------
7. Driscoll, KG., 1986. Groundwater and Wells. Second Edition, Johnson Filtration Systems, Inc.
8. Fetter, C.W., 2001. Applied Hydrogeology. 4th Edition, Prentice-Hall.
9. Freeze, R.A. and J.A. Cherry, 1979. Groundwater. Prentice-Hall, Inc., Englewood Cliffs, NJ.
10. Grubb, S., 1993. Analytical Model for Estimation of Steady-State Capture Zones of Pumping Wells
in Confined and Unconfined Aquifers. Ground Water, vol. 31, no. 1, pp. 27-32.
11. Hantush, M.S., 1964. Hydraulics of Wells, in Advances in Hydroscience, V.T. Chow, Editor,
Academic Press, pp. 281-442.
12. Jacob, C.E., 1950. Flow of Groundwater, in Engineering Hydraulics (H. Rouse, Ed.), Wiley, New
York.
13. Javandel, I. and C.F. Tsang, 1986. Capture-Zone Type Curves: ATool for Aquifer Cleanup, Ground
Water, vol. 24, no. 5, pp. 616-625.
14. Tonkin, M.J. and S.P. Larson, 2002. Kriging Water Levels with a Regional-Linear and Point-
Logarithmic Drift, Ground Water, vol. 40, no. 2:185-193.
15. U.S. EPA, 1994. Methods for Monitoring Pump-and-Treat Performance, EPA/600/R-94/123, U.S.
EPA, ORD, R.S. Kerr Environmental Research Laboratory, Ada, OK. (http://www.epa.gov/oerrpage/
superfund/resources/gwdocs/per_eva.htnO
16. U.S. EPA, 2000. Data Quality Objectives Process for Hazardous Waste Sites, EPA OQ/G-4HW,
EPA/600/R-00/007. U.S. Environmental Protection Agency, Office of Environmental Information,
Washington, DC.
17. U.S. EPA, 2002a. Pilot Project to Optimize Superfund-financed Pump and Treat Systems: Summary
Report and Lessons Learned, EPA 542-R-02-008a.
18. U.S. EPA, 2002b. Elements of Effective Management of Operating P&T Systems, EPA 542-R-02-
009.
19. Wilson, D.A. and D.E. Dougherty, 2002. Optimizing Pump and Treat Systems: Data Assimilation,
Federal Remediation Technologies Roundtable Meeting, Arlington, VA, December 18, 2002 (http://
www.frtr.gov/meetings2.htm)
Codes/Software Cited (as Examples)
These codes have been cited in this document:
CAPZONE (Ohio State University)
GWPATH (Illinois State Water Survey)
MODFLOW(U.S. Geological Survey)
MODPATH (U.S. Geological Survey)
SURFER (Golden Software)
WhAEM (U.S. EPA)
-------
Mention of trade names or commercial products does not constitute endorsement or recommendation for
use.
EPA Web Sites With Links to Modeling Software
Center for Subsurface Modeling Support (CSMoS): www.epa.gov/ada/csmos.html
Center for Exposure Assessment Modeling (CEAM): www.epa.gov/ceampubl/ceamhome.htm
Council for Regulatory Environmental Modeling (CREM): http://cfpub.epa.gov/crem/
Other References
Bair, E.S., A.E. Springer, and G.S. Roadcap, 1991. Delineation of traveltime-related capture zones of
wells using analytical flow models and particle-tracking analysis. Ground Water, vol. 29, no. 3, pp.
387-397.
Bair, E.S. and G.S. Roadcap, 1992. Comparison of flow models used to delineate capture zones of wells:
1. Leaky-confined and fractured-carbonate aquifer. Ground Water, vol. 30, no. 2, pp. 199-211.
Blandford, T.N. and PS. Huyakorn, 1990. WHPA: An integrated semi-analytical model for delineation of
wellhead protection areas. U.S. EPA Office of Ground-Water Protection.
Faybishenko, B.A., I. Javandel, and PA. Witherspoon, 1995. Hydrodynamics of the capture zone of a
partially penetrating well in a confined aquifer. Water Resources Research, vol. 31, no. 4, pp. 859-
866.
Gorelick, S., 1987. Sensitivity analysis of optimal ground water contaminant capture curves. In: Proc.
NWWA Conference Solving Ground Water Problems with Models, Dublin, OH, pp. 133-146.
Javandel, I., C. Doughty, and C.F. Tsang, 1984. Groundwater Transport: Handbook of Mathematical
Models, American Geophysical Union Water Resources Monograph No. 10, Washington, DC.
Javandel, I., 1986. Application of capture-zone type curves for aquifer cleanup. Groundwater Hydrology,
Contamination, and Remediation, eds. R. Khanbilvard and J. Fillos, Washington, DC, Scientific
Publications, pp. 249-279.
Keely, J.F. and C.F. Tsang, 1983. Velocity plots and capture zones of pumping centers for ground water
investigations. Ground Water, vol. 21, no. 6, pp. 701-714.
Keely, J.F., 1984. Optimizing pumping strategies for contaminant studies and remedial actions. Ground
Water Monitoring & Remediation, Summer 1984, pp. 63-74.
Keely, J.F., 1989. Performance evaluation of pump-and-treat remediations, EPA/540/4-89-005, Robert S.
Kerr Environmental Research Laboratory, Ada, OK.
Lee, K.H.L. and J.L. Wilson, 1986. Pollution capture zones for pumping wells in aquifers with ambient
flow. EOS Transactions of the American Geophysical Union, vol. 67, p. 966.
-------
Shafer, J.M., 1987. GWPATH: Interactive ground-water flow path analysis. Bulletin 69, Illinois State
Water Survey, Champaign. IL.
Springer. A.E. and E.S. Bair, 1992. Comparison of methods used to delineate capture zones of wells: 2.
Stratified-drift buried-valley aquifer. Ground Water, vol. 30, no. 6, pp. 908-917.
Strack, O.D.L., 1989. GroundwaterMechanics. Prentice-Hall. Inc., Englewood Cliffs, NJ.
Yang, Y.J.. R. Spencer, and T.M. Gates, 1995. Analytical solutions for determination of non-steady-state
and steady-state capture zones. Ground Water Monitoring & Remediation, Winter, pp. 101-106.
Varljen. M.D. and J.M. Shafer. 1991. Assessment of uncertainty in time-related capture zones using
conditional simulation of hydraulic conductivity. Ground Water, vol. 29, no. 5. pp. 737-748.
Wilson, J.L., 1986. Induced infiltration in aquifers with ambient flow. EOS Transactions of the American
Geophysical Union, vol. 67, p. 966.
-------
APPENDIX A:
HYPOTHETICAL
-------
INTRODUCTION TO APPENDIX A:
EXAMPLES FROM HYPOTHETICAL SITES
To illustrate the systematic approach to capture zone analysis, and to highlight some of the details
associated with specific techniques for evaluating capture, examples for three hypothetical sites were
developed for this document. Each hypothetical site is represented by a numerical flow and transport
model that is intended to portray '"actual conditions" for operating P&T systems. Each example site has a
different degree of complexity, to represent a variety of real world conditions. In addition, each example
site highlights a specific issue related to capture zone analysis:
• Example Al: highlights complications of evaluating capture when significant vertical flow and
vertical contaminant transport are present
Example A2: highlights complications of evaluating horizontal capture when preferential
flow pathways arc present
* Example A3: highlights complications of evaluating capture when off-site stresses are
present
For each example, at least one pumping scenario that achieves successful capture and one pumping
scenario that does not achieve successful capture is presented.
The pumping scenarios for these examples should not be confused with prc-rcmcdy design options.
Instead, these scenarios are intended to represent different examples of operating P&T systems.
Example Al is presented in the greatest detail, and is used to demonstrate the entire systematic process
for capture zone evaluation. Example A2 and Example A3 are then used to illustrate specific aspects of
capture zone analysis.
-------
EXAMPLE Al
This example, for a hypothetical site, highlights complications of evaluating capture when significant
vertical flow and vertical contaminant transport are present.
Example Setup
The stratigraphy of the site contains two heterogeneous aquifers: a shallow aquifer and a deep aquifer
(Figure Al-1). These aquifers are differentiated based on geologic description, and there is no aquitard
separating these aquifers. In general, the deep aquifer material has a higher hydraulic conductivity than
the shallow aquifer material.
Figure Al-1
Supply
Well
North -
Vertical Scale (ft)
n20
Schematic Cross-Section
Near-River Mid-plume
Extraction Well Extraction Well Contaminant
Source Area
South
Upper Horizon
(Layer 1)
= 2.5-30ft/d
Middle Horizon (Layer 2)
K=1.5-20ft/d
Lower Horizon (Layer 3)
K=1.5-20fVd
Deep Aquifer (Layer 4)
K = 40 - 70 ft/d
The shallow aquifer is further divided into three distinct horizons (upper, middle, lower) to better
represent the following observations:
• there are partially penetrating wells
• only the upper horizon of the shallow aquifer is hydraulically connected to the river
• observed contaminant levels decrease with depth within the shallow aquifer
-------
For ease of presentation, the subsequent discussion refers to "layers" as follows:
Layer 1: shallow aquifer, upper horizon
Layer 2: shallow aquifer, middle horizon
Layer 3: shallow aquifer, lower horizon
Layer 4: deep aquifer
Figure A1-2 indicates the actual hydraulic conductivity distribution of each layer in this hypothetical
heterogeneous system. In reality, site managers would only have drilling logs, cores, slug test, or
pumping test data at a limited number of locations, and maps similar to those in Figure A1-2 might be
created from those data. Uncertainty in the hydraulic conductivity distributions is an important issue that
should be considered and addressed, as it directly affects ground-water flow, velocity vectors, and capture
zones.
The contaminant of concern is dissolved TCE. Figure A1-3 indicates the actual pre-remedy TCE
concentrations in each layer. The ground-water standard for TCE at this site is 5 ppb. There is a
continuing source of TCE (in the unsaturated zone) that has not yet been fully identified or delineated,
and it is being investigated further with the hope that the continuing source can be remediated in the
future.
Historically, dissolved TCE has discharged to the river from the south. In addition, there are two water
supply wells screened in the deep aquifer across the river from the contaminant source (i.e., north of the
river). These water supply wells have caused contaminated ground water to flow beneath the river, and
these water supply wells are impacted by TCE. The wells continue to pump and have wellhead treatment.
The TCE concentrations decrease with depth, and the lowest concentrations of TCE are observed in the
deep aquifer. For this hypothetical example, the distribution of TCE is known at all locations. In reality
site managers would only have data at locations where monitoring wells exist, and maps similar to Figure
A1 -3 would be created based on those sparse data using interpolation and/or extrapolation.
Scenarios Illustrated
Three pumping scenarios are evaluated with respect to capture (Figure Al-4).
* Scenario 1 has three remedy pumping wells screened in the upper horizon of the shallow
aquifer (Layer 1) near the river, with a total pumping rate of 22.5 gpm.
* Scenario 2 has five remedy pumping wells in the upper horizon of the shallow aquifer (Layer
1), with a total pumping rate of 44 gpm. Three of the five wells are located near the river, and
the other two are mid-plume or source-area wells.
• Scenario 3 has nine remedy pumping wells in the upper horizon of the shallow aquifer (Layer
1.), with a total pumping rate of 107 gpm. Five of the nine wells are located near the river.
and the other four are mid-plume or source-area wells.
None of the scenarios for this example have remedy pumping in the deeper layers.
-------
Figure Al-2
Hydraulic Conductivity Distribution (ft/day)
Layer 1
(ft)
2600-
Layers 2 & 3
IN
A
24oo- Water Supply Wells^/
(deep aquifer only) ~"
K (ft/day)
32
28
24
20
16
12
BOO I
Continuous Sources
(upper horizon only)
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
A
K (ft/day)
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
Layer 4
K (ft/day)
Note: The different colors
illustrate the distribution of 1+
the hydraulic conductivity
values, as indicated by the
scale bars. The values in
layer 4 are higher than
layers 1 to 3.
2400_ Water Supply Wells/
(deep aquifer only)
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
-------
Figure Al-3
Pre-Remedy Concentrations
Layer 1
Water Supply Wells/
(deep aquifer only)
Continuous Sources
(upper horizon only)
Layer 2
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
Continuous Sources
(upper horizon only)
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
Layer 3
Layer 4
(ft) (ft)
2400-
2200-
1800-
1600-
1400-
1200-
1000-
800-
/X^v-f N
Water Supply Wells /— /) A
(deep aquifer only) / f\
M
r
River
«
^
//
\\ I
%
\J
'
/l^ ysppb
X/ _ 50 ppb
i^ 100 ppb
Continuous Sources ~ 200 ppb
(upper horizon only) 500 PPb
2400-
2200-
2000-
1800-
1600-
1400-
1200-
1000-
800-
%~~~^t N
Water Supply Wells/J-^ A
(deep aquifer only) f\
, //I
7 /
River
u
W
\J
/[/* DSPF*
// _50ppb
// 100 ppb
Continuous Sources ~ 2oo ppb
(upper horizon only) Osooppb
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
-------
Figure Al-4
Pumping Scenarios
m
2600^
2400^
2200^
1800
1600^
1400-
1200^
1000-
800
12
Scenario 1 : 22.5 gpm
TCE 5ppb
+ Extraction Well (gpm) S\
1
River
/
7.5 7.5 7.5
\._
N
A
Ul
/' "~
Continuous Sources
(upper horizon only)
00 1400 1600
(ft)
2600-
2400-
2200-
2000-
1800 2000 2200 2400 2600 (ft)
'
=
1800-
1600-
1400-
1200-
1000-
800-
12
Scenario 2: 44 gpm
TCE 5ppb
•+• Extraction Well (gpm) S\
f 8 12 8
CJl
8
Continuous Sources
(upper horizon only)
N
A
•
00 1400 1600 1800 2000 2200 2400 2600 (ft)
Scenario 3: 107 gpm
(ft) &r
2400^
2200^
1800
1600-
1400-
1200^
1000^
800
12
TCE 5ppb
-*• Extraction Well (gpm) /^
/ I
f 7.3^ 4- "V +• -
„./ 13.4 11.8
River
8.2+- 420
1Q.2
1^.5
V/f— /
Continuous Sources
(upper horizon only)
00 1400 1600 1800 2000
N
A
[7.1
f
2200 2400 2600 (ft)
-------
Step 1 - Review Site Data, Site Conceptual Model, and Remedy Objectives
Is the plume delineated adequately in three dimensions?
The plume is bounded by monitoring wells with concentrations below standards in each
hydrostratigraphic unit, based on recent concentration data (for example, Figure A 1-5 illustrates
plume delineation in Layer 1). Thus, the plume delineation is considered adequate.
Is there sufficient hydrogeologic information?
Numerous well logs are available to evaluate stratigraphy. Slug test and pumping test data exist
for multiple locations (Table Al-1). Pre-remediation water level maps are available for calculating
regional gradients (Figure Al-5). Thus, the hydrogeologic information is considered sufficient.
Is there a site conceptual model that adequately explains the contaminant source and constituents?
Currently there exists a continuing source of contaminants to shallow ground water. Shallow
ground water flows north to the river, and the water supply wells on the other side of the river
cause some contaminated ground water to flow beneath the river to the supply wells.
Is the objective of the remedy clearly stated?
The remedy objective for this site is to prevent contaminants at concentrations above standards
from discharging to the river, and to prevent contaminants at concentrations above standards from
continuing to flow under the river towards the water supply wells.
Figure Al-5
Plume Delineation, Model Layer 1
Pre-Remedy Water Levels, Layer 1
-------
Table Al-1. Slug/Pumping Test Results
Layer
Layer 1 :
MW-6s
MW-9s
MW-15s
MW-20s
MW-31s
MW-33s
EW-1
EW-2
EW-3
Test Type
slug
pumping
slug
slug
pumping
slug
pumping
pumping
pumping
representative value
range from regional pumping test data
K Range (ft/day)
1.7-9
25-30
7-110
0.6 - 50
8-22
90 - 130
18-20
15-30
15-18
15
10-40
Layer 2:
MW-6m
MW-20m
MW-31m
slug
slug
slug
representative value
range from regional pumping test data
12-60
1.2-85
7-75
30
15-30
Layer 3 :
MW-3m
MW-12m
MW-17m
MW-26m
MW-37m
slug
slug
pumping
slug
slug
representative value
range from regional pumping test data
12-20
70- 115
20-30
1.2-28
0.7 - 12
20
10-35
Layer 4:
MW-3d
MW-12d
MW-18d
MW-21d
MW-28d
MW-37d
slug
slug
pumping
slug
slug
pumping
representative value
range from regional pumping test data
25 - 120
50 - 560
50-80
29 - 130
50 - 620
40-70
60
20 - 100
-------
Step 2 - Define Target Capture Zone(s)
As discussed in the main document, the Target Capture Zone can be defined based on: (1) complete
hydraulic capture of the entire plume (horizontal and/or vertical); or (2) capture of a specific portion of
the plume in conjunction with another remedial technology for the uncaptured portion. Two options for
defining the Target Capture Zone at this site are presented below and are illustrated in Figure A1-6:
Option 1 (More Conservative)
to prevent continuing discharge of TCE to the river, inward flow from the river to the aquifer
(i.e., to the south) will be established over the width of the plume exceeding 5 ppb
• vertical hydraulic containment is required south of the river within the 5 ppb plume, by
demonstrating upward flow to Layer 1 (where the continuing source is located) from
underlying hydrogcologic units
Option 2 (Less Conservative)
* to prevent continuing discharge of TCE to the river, a flow divide will be established between
the extraction wells and the river, over the width of the plume exceeding 5 ppb
• vertical hydraulic containment is not required south of the river as long as concentrations on
the north side of the river decrease below 5 ppb overtime as a result of the shallow extraction
south of the river in conjunction with Monitored Natural Attenuation (MNA)
For the purpose of illustrating the process of capture zone analysis, the discussion below will consider
both of these options for defining the Target Capture Zone. Note that the requirement for demonstrating
horizontal capture is more conservative for Option 1 than for Option 2. The requirement for
demonstrating vertical capture is also more conservative for Option 1. Other options for defining Target
Capture Zone are possible, but are not presented herein.
Step 3 - Interpret Water Levels
Horizontal and/or vertical capture can be interpreted from water level maps that are constructed from
measured water levels. In addition, flow directions can be analyzed from water level pairs (gradient
control points) to aid in interpretation of capture. The discussion below illustrates some of the analyses
that can be performed at this site based on water levels, for each of the three defined pumping scenarios.
to evaluate the extent of capture.
Water Level Maps: Horizontal Capture Analysis
The following figures present constructed water level maps for Layer 1 (the upper horizon of the shallow
aquifer where the extraction wells are screened), for each of the three pumping scenarios defined for this
example:
• Figure Al-7: Scenario 1 (22.5 gpm)
• Figure A1-8: Scenario 2 (44 gpm)
* Figure Al-9: Scenario 3 (107 gpm)
-------
Figure Al-6
Target Capture Zone, Option 1 - More Conservative
Supply
Well
North _
Vertical Scale (ft)
,20
Inward flow from
river to aquifer
Contaminant
Source Area
South
Target Capture Zone, Option 2 - Less Conservative
Water Supply Wells '
(deep aquifer only)
Plume
River
Continuous Sources
(upper horizon only)
Flow divide between
extraction wells and
the river in the upper
horizon of the
shallow aquifer
Note: Contamination in
deep aquifer is
addressed by other
technologies
1200 1400 1600 1800 2000 2200 2400 2600
-------
Figure Al-7
Horizontal Interpretation of Water Level Maps, Layer 1, Pumping Scenario 1 (22.5 gpm)
A. 40 Observation Points Without Stage Measurements B. 40 Observation Points With Stage Measurements
or Water Level Measurements near Pumping Wells
(ft)
2600
Plume
A
Continuous Sources 6 A
(upper horizon only)
TCE 5ppb
-f Extraction Well
A Monitoring Well
1200 1400 1600 1800 2000 2200 2400 2600 (ft) 1200 1400 1600 1800 2000 2200 2400 2600 (ft)
C. 40 Observation Points Plus Water Level
Measurements near Pumping Wells
D. 12 Observation Points Plus Water Level
Measurements near Pumping Wells
^ TCE 5ppb
4- Extraction Well
A Monitoring Well
Continuous Sources
(upper horizon only)
^ TCE 5ppb
4- Extraction Well
A Monitoring Well
Continuous Sources
(upper horizon only)
1200 1400 1600 1800 2000 2200 2400 2600 (ft) 1200 1400 1600 1800 2000 2200 2400 2600 (ft)
-------
Figure Al-8
Horizontal Interpretation of Water Level Maps, Layer 1, Pumping Scenario 2 (44 gpm)
A. 40 Observation Points Without Stage Measurements B. 40 Observation Points With Stage Measurements
or Water Level Measurements near Pumping Wells
(ft) (ft)
2600T ; r I , 1 7 1 2600
Continuous Sources
(upper horizon only)
Continuous Sources ,A >\
(upper horizon only) ?
r* TCE Sppb
4- Extraction Well
A Monitoring Well
1200 1400 1600 1800 2000 2200 2400 2600 (ft) 1200 1400 1600 1800 2000 2200 2400 2600 (ft)
C. 40 Observation Points Plus Water Level
Measurements near Pumping Wells
•* TCE Sppb
-*- Extraction Well
A Monitoring Well
Continuous Sources
(upper horizon only)
D. 12 Observation Points Plus Water Level
Measurements near Pumping Wells
TCE Sppb
4- Extraction Well
A Monitoring Well
Continuous Sources
(upper horizon only)
1200 1400 1600 1800 2000 2200 2400 2600 (ft) 1200 1400 1600 1800 2000 2200 2400 2600 (ft)
-------
Figure Al-9
Horizontal Interpretation of Water Level Maps, Layer 1, Pumping Scenario 3 (107 gpm)
A. 40 Observation Points Without Stage Measurements B. 40 Observation Points With Stage Measurements
or Water Level Measurements near Pumping Wells
(ft) (ft)
Continuous Sources A A
(upper horizon only)
TCE 5ppb
-f Extraction Well
A Monitoring Well
TCE 5ppb
•f Extraction Well
A Monitoring Well
Continuous Sources
(upper horizon only)
1200 1400 1600 1800 2000 2200 2400 2600 (ft) 1200 1400 1600 1800 2000 2200 2400 2600 (ft)
C. 40 Observation Points Plus Water Level
Measurements near Pumping Wells
Continuous Sources
(upper horizon only)
D. 12 Observation Points Plus Water Level
Measurements near Pumping Wells
Continuous Sources A
(upper horizon only)
1200 1400 1600 1800 2000 2200 2400 2600 (ft) 1200 1400 1600 1800 2000 2200 2400 2600 (ft)
-------
In these figures several potential water level maps are presented for each pumping scenario, to compare
the following situations:
• with/without water level measurements near the pumping wells
• larger/smaller amounts of water level data
• with/without river stage measurements
These different water level maps for Layer 1 illustrate how interpretations of horizontal capture can vary
depending on the availability of water level measurements near the extraction wells:
• with a large number of water level measurements, but no water level measurements near the
pumping wells (map "A" of Figures A1-7 to A1-9), horizontal capture is not clearly apparent
and successful capture is not interpreted even for Scenario 3 (107 gpm)
with the same number of water level measurements, plus water level measurements near the
pumping wells (map "C" of Figures A1-7 to A1-9), horizontal capture is apparent and the
extent of capture can be interpreted
Water levels at extraction wells are typically lower than water levels in the surrounding aquifer due
to well inefficiency and well losses. For this reason, water levels measured at extraction wells should
generally not be utilized to construct water level maps, because they potentially bias the interpretation of
capture (to be more extensive than is actually achieved). However, Figures Al-7 to Al-9 illustrate that if
no water level measurements are available near the extraction wells, interpretation of horizontal capture
will likely be biased towards an interpretation of poor capture. To avoid this problem, EPA recommends
installing a piezometer near each extraction well. However, if such piezometers do not exist, a possible
approach is to estimate aquifer water levels at the extraction wells by correcting the measured water levels
for well losses, until appropriately located piezometers are available. However, such calculations do not
account for all components of well inefficiency, and locating piezometers near extraction wells is much
preferred to correcting water levels in extraction wells based on calculated well losses.
For this site, the inclusion of river stage measurements in addition to the water level measurement is also
illustrated (map "B" on Figures Al-7 to Al-9), for the situation without water level measurements near
the pumping wells. These maps indicate that the addition of stage measurements at this site has little
impact on the interpretation of horizontal capture. For this site, including river stage measurements is not
as important as including water level measurements near the extraction wells. This is likely because of
the availability of water level measurements at several locations between the extraction wells and the
river at this example site. River stage measurements might impact the interpretation of capture more
substantially if these water level data from measuring points near the river were not available.
For Layer 1 in Figures Al-7 to Al-9, a water level map and associated interpretation of horizontal capture
are also provided for a case with many fewer water level measurements (map "D" on Figures Al-7 to
Al-9), and where water level measurements near the extraction wells are available. With fewer water
level measurements a larger zone of capture is interpreted. This illustrates that the number of available
water level measurements will impact the interpretation of capture. In general, the accuracy of the
interpreted capture extent will increase as the number of water level measurement locations increases. As
discussed earlier, availability of water level measurement points located near and around the extraction
wells are very important for accurately depicting capture zones from water level maps.
-------
For Scenario 3, the interpreted capture zones with water levels near the extraction wells (maps "C" and
"D" on Figure A1-9) indicate that the extraction wells south of the river are potentially capturing water
from north of the river. This suggests that the pumping rate associated with Scenario 3 (107 gpm) may be
more than is necessary to simply prevent water from discharging to the river from the south.
It is also observed that the capture zone widths on maps "C" and "D" of Figure A1-8 (total pumping of
44 gpm) are only slightly greater than the capture zone widths on maps "C" and "D" of Figure A1-7 (total
pumping of 22.5 gpm), despite the fact that the pumping rate is nearly double. The capture zone width
would be nearly double under very simplified hydrogeologic conditions, because the total extraction rate
is nearly double. However, that simple relationship does not hold when the hydrogeology is complex,
such as with this example, which includes a river that serves as a potential source of water to the
extraction wells and/or underlying strata that also provide water to the partially-penetrating extraction
wells.
In summary, a variety of potential water level maps can be constructed depending on the data available,
leading to a variety of potential interpretations regarding the extent of horizontal capture. For each of the
three pumping scenarios (Figures A1-7 to A1-9), at least one potential water level map is presented that
suggests potential for "failed" horizontal capture in Layer 1, and at least one potential water level map is
presented that suggests "successful" horizontal capture in Layer 1.
The most important consideration appears to be the existence of water level measurements near each
extraction well (or an estimate of water levels in the aquifer near each extraction well). Without such
measurements, horizontal capture is biased towards interpretation of poor capture. However, the number
of available data points away from the extraction wells also impacts the interpreted extent of horizontal
capture. The more water level points that are available, the more accurate the interpretation of capture.
Constructed water level maps for Layer 3 (lower horizon of the shallow aquifer) and Layer 4 (deep
aquifer) are presented on the following figures:
• Figure Al-10: Scenario 1 (22.5 gpm)
• Figure A1 -11: Scenario 2 (44 gpm)
• Figure Al-12: Scenario 3 (107 gpm)
The water level maps for Layers 3 and 4 do not indicate any discernable capture from the extraction wells
that are screened in Layer 1, for any of the pumping scenarios. That is because no water level
measurements or estimates for these depths are available in the vicinity of the extraction wells.
If piezometers screening these deeper horizons were installed in the vicinity of the extraction wells, some
horizontal capture in these layers would likely be apparent, even though the extraction wells are screened
only in the upper horizon of the shallow aquifer. This again illustrates the importance of having water
level measurements near extraction wells when interpreting horizontal capture using water level maps.
For a multi-aquifer or multi-layer problem, multi-level monitoring wells (or well clusters) near the
extraction wells are recommended.
-------
Figure Al-10
Horizontal Interpretation of Water Level Maps, Layers 3-4, Pumping Scenario 1 (22.5 gpm)
Layer 3: 12 Observation Points
Layer 4: 16 Observation Points
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
Figure Al-11
Horizontal Interpretation of Water Level Maps, Layers 3-4, Pumping Scenario 2 (44 gpm)
Layer 3:12 Observation Points
Layer 4: 16 Observation Points
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
I200 1400 1600 1800 2000 2200 2400 2600 (ft)
-------
Figure Al-12
Horizontal Interpretation of Water Level Maps, Layers 3-4, Pumping Scenario 3 (107 gpm)
Layer 3:12 Observation Points
Layer 4: 16 Observation Points
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
Water Level Maps: Vertical Capture Analysis
This site has a continuous source of TCE in the upper horizon of the shallow aquifer (Layer 1). By
evaluating head difference between the upper and lower horizons of the shallow aquifer (Layer 1 versus
Layers 2 and 3), the potential for downward flow from the contaminant source area can be evaluated. If
vertical gradients are upward at all locations, then dissolved contaminants will not be transported by
advection from the source area to underlying horizons. Note that at DNAPL sites there is a potential for
DNAPL to migrate downward even in the presence of upward hydraulic gradients. However, at this site
TCE concentrations at depth are too low relative to the solubility limit to be indicative of DNAPL at
depth.
Figure Al-13 illustrates interpretation of vertical head differences at existing well clusters, for each of the
three pumping scenarios:
• for Scenario 1 (22.5 gpm) there are downward vertical gradients interpreted near the contaminant
source and in the central portion of the plume south of the river, suggesting the potential for
downward advection of dissolved TCE
• for Scenario 2 (44 gpm) there is a much greater area where upward vertical gradients are
interpreted, largely due to the existence of the mid-plume extraction wells, although some areas
of downward vertical gradients are interpreted within the footprint of the plume south of the river
for Scenario 3 (107 gpm) upward vertical gradients are interpreted within the entire footprint of
the plume south of the river
-------
Figure Al-13
Water Level Difference Maps (Well Clusters In Layers 1 and 3)
Scenario 1: 22.5 gpm (ft) Scenario 2: 44 gpm
ft TCE Sppb
Extraction Well
-<>- Water Supply Well
(deep aquifer only)
A Monitoring Well
-- Downward Flow,
Head Difference > 0
— Upward Flow,
Head Difference < 0
A Continous Sources
(upper horizon only)
Upward Flow
^ TCE Sppb
Extraction Well
•fy Water Supply Well
(deep aquifer only)
A Monitoring Well
-- Downward Flow,
Head Difference > 0
— Upward Flow,
Head Difference < 0
•
Continuous Sources
(upper horizon only)
(ft)
>400 2600 (ft) 1200
Scenario 3: 107 gpm
2400 2600 (ft)
River
^ TCE Sppb
Extraction Well
<>- Water Supply Well
(deep aquifer only)
A Monitoring Well
-- Downward Flow,
Head Difference > 0
— Upward Flow,
Head Difference < 0
IN
A
Area With
Upward Flow
' Continuous Sources
(upper horizon only)
-------
These vertical head difference figures were created by contouring the measured head differences from
locations where water level data were available at multiple depths (i.e., clustered monitoring wells).
For Scenarios 1 and 2, the downward vertical gradients within the footprint of the plume suggest a
potential for dissolved TCE to migrate downward, and potentially beneath the river to the water supply
wells. However, the level of detail associated with this water level analysis cannot determine if such
transport beneath the river ultimately occurs, or if TCE transported downward near the contaminant
source area is ultimately captured (or adequately attenuated) by the shallow extraction wells near the river
(note there are upward vertical gradients near the river caused by the remedy extraction wells for all three
pumping scenarios).
Water Level Pairs (Gradient Control Points)
Two types of water level pairs are analyzed for this site:
river stage measurements are compared to water level measurements in the aquifer immediately
south of the river to evaluate horizontal hydraulic containment at the river (at this site the upper
horizon of the shallow aquifer is hydraulically connected to the river, so head differences between
the river stage and the aquifer water level indicate if flow direction is to or from the river)
• water levels are compared between the upper and lower horizons of the shallow aquifer (Layer 1
and Layer 3), to evaluate vertical hydraulic containment for the upper horizon of the shallow
aquifer that contains the continuing source of dissolved TCE
Figure Al-14 illustrates results for a large number of potential horizontal gradient control pairs (many
more than would typically be available at most sites) in the vicinity of the river, for each of the three
pumping scenarios. The use of so many data points is to highlight the details of the actual flow system
for this example, and should not be confused with the number of monitoring wells in this hypothetical
example illustrated earlier (i.e., many fewer locations, which is more realistic). The interpretations
regarding horizontal hydraulic containment at the river, based on Figure Al-14, are as follows:
for Scenario 1 (22.5 gpm), ground water in the aquifer discharges to the river at all locations, and
the downgradient extent of capture cannot be determined from these data
• for Scenario 2 (44 gpm), ground water in the aquifer discharges to the river at all locations, and
the downgradient extent of capture cannot be determined from these data
for Scenario 3 (107 gpm), the river discharges to the aquifer at all locations, indicating successful
hydraulic containment at the river
As discussed earlier (see Figure 11 of main document), the lack of demonstrated hydraulic containment
at the river in Scenarios 1 and 2 does not prove that a capture zone is not achieved. It is possible that
horizontal capture of the plume is successful but the flow divide associated with the capture zone is
established upgradient of the river. The hydraulic containment achieved at the river in Scenario 3 should
be viewed as a conservative measure of horizontal capture because achieving inward gradients at the
river along the entire plume width requires more pumping than simply achieving an adequate flow divide
between the extraction wells and the river.
-------
Figure Al-14
Horizontal Water Level Pairs (Gradient Control Points) for Scenario 1 (22.5 gpm)
2050-
2000-
1950-
1900-
1850-
N
A /
617.16 617.9 617.9 618.0 618.0 618.0 618.1 618.2 M8.3
/ T/ T. T, T. T. T, T. T. TV
River 618.4 618.3 618.2 618.2 618.2 618.3 618.4 618.5 611
P.
' ft
^ TCESppb Regional
-f Extraction Well Grouridwater
* Water Level Flow Direction
618.3 Measurement Continuous source located approximately 750 ft to the south
.7
1600 1650 1700 1750 1800 1850 1900 1950 2000 2050
2100 2150 2200 (ft)
Horizontal Water Level Pairs (Gradient Control Points) for Scenario 2 (44 gpm)
(ft)
2050-
2000-
1950-
1900-
1850-
N
A /
Wiy 617.9 617.9 617.9 618.0 618.0 618.1 618.2 618.2
; T/ T; T; T. T» T; T; T; ft
River 618'3 618'2 618i1 618i1 618i1 618i2 618i3 618'5 61
1 + + +
1 fl
^ TCESppb Reaional
* Extraction Well Grou^Zter
* Water Level Flow Direction
618.3 Measurement Continuous source located approximately 750 ft to the south
.7
1600 1650 1700 1750 1800 1850 1900 1950 2000 2050 2100 2150 2200 (ft)
Horjjzontal Water Level Pairs (Gradient Control Points) for Scenario 3 (107 gpm)
N
A
wi.y
f */
River
-------
Figures A1-15 to A1-17 illustrate results for a large number of potential vertical gradient control pairs
(again, many more than would typically be available at most sites), for each of the three pumping
scenarios, respectively. The interpretations regarding vertical hydraulic containment of Layer 1, south of
the river, are as follows:
• for Scenario 1 (22.5 gpm), downward flow exists in most of the area within the plume
footprint except near the extraction wells and the river, indicating potential for failed hydraulic
containment in the vertical (Figure Al-15)
* for Scenario 2 (44 gpm), upward flow occurs near the contaminant source area where the
mid-plume extraction wells are installed, but downward flow occurs in the central portion of
the plume, indicating potential for failed hydraulic containment in the vertical (Figure Al-16)
• for Scenario 3 (107 gpm) upward flow occurs within the entire footprint of the plume,
indicating successful hydraulic containment in the vertical (Figure A1-17)
Again, please note that at most sites a much smaller number of water level pairs are typically available for
the interpretation of horizontal and/or vertical containment than are illustrated in this example.
Step 4 - Perform Calculations
Specific calculations can be performed to add additional lines of evidence regarding the extent of capture.
including the following:
• simple horizontal analyses, such as estimated flow7 rate calculations and capture zone width
calculations
• modeling (analytical or numerical) to simulate heads, in conjunction with particle tracking
and/or transport modeling
The discussion below illustrates some of the calculations that can be performed at this site, for each of the
three defined pumping scenarios, to evaluate the extent of capture.
Simple Horizontal Calculations
Although the calculations associated with these analyses are quite simple, deciding on the actual values
to use for the calculations is not straightforward for this site because some of the parameters (e.g.. aquifer
thickness, hydraulic conductivity, magnitude of the hydraulic gradient) vary in space.
One complication pertains to aquifer thickness. The shallow aquifer is unconfined, so the thickness of the
aquifer is variable and depends on the actual water levels. Also, the extraction wells are only screened in
the upper horizon of the shallow aquifer (Layer 1), but the wells are partially penetrating wells and likely
draw water from all horizons of the shallow aquifer (Layers 1-3). and perhaps from the deep aquifer as
well because there is no aquitard separating the shallow and deep aquifers.
-------
Figure Al-15
Vertical Water Level Pairs (Gradient Control Points) for Scenario 1 (22.5 gpm)
(ft)
2100
2000
1900-
1800-
1700-
1600-
1500-
1400-
1300-
1200-
1100-
1000-
15
River
DO 1600
/ \ I
O O O O O
+ + +
00000
00000
®/r\ /^r\ f\\ /T\
d7 \£s \I? vl/
© © © © ©
ti? \±s \jj y? \j3
®/T\ /T\ /T\ /T^
^17 vl? vX? vI7
Ji
^T^i /Ts /^T\ /*Ts /T\
\J> ^X? ^X/ \J2 vt/
/T\ /T\ /T\ i^T\ /T\
vj> ^X? ^X/ \L/ ^I/
© © © © ©
© © © © ©
Continuous Sources ^ JCE 5ppb
© © /@ © ©
(^ /
4- Extraction Well
© Downward Flow
O Upward Flow
1700 1800 1900 2000 2100 2200 2300 2400 (ft)
-------
Figure Al-16
Vertical Water Level Pairs (Gradient Control Points) for Scenario 2 (44 gpm)
(ft)
2100
2000
1900
1800
1700
1600
1500
1400
1300
1200
1100-
1000
15
\
/ A
f
/
River
00 1600
A
00000
+ + +
00000
o o o o o
© © © © ©
®fT\ /T\ /T\ ST\
d7 \]7 \L/ v!7
®iT> /T\ /T\ /TN
^I/ \I? \3s ^1?
®i^T\ /TN /T\ /T\ fT\
vT7 \17 vI7 M7 M7
© 0 0 © ©
+
© O O O © 10
/
© o o o ©
Continuous Sources
cji i ^ TCE 5ppb
\ © o / o o ©
1 -f Extraction Well
V ¥-• © Downward Flow
\ O Upward Flow
1700 1800 1900 2000 2100 2200 2300 2400 (ft)
-------
Figure Al-17
Vertical Water Level Pairs (Gradient Control Points) for Scenario 3 (107 gpm)
(ft)
2100
2000
1900-
1800-
1700-
1600-
1500-
1400-
1300-
1200-
1100-
1000-
15
^^™
/
River
00 1600
S N
OOOOO A
A
o o o o o
OOOOO
n
OOOOO
OOOOO
OOOOO
01
0 0.0.0 0
OOOOO
OOOOO
o o + o o o
\
O O O O O 10
/
OOOOO
Continuous Sources
^ 0 0 /O+l 0 0 ^TCESppb
/ -f Extraction Well
y ^ m^* o ® Downward Flow
\ O Upward Flow
1700 1800 1900 2000 2100 2200 2300 2400 (ft)
-------
Some possible options for assigning aquifer thickness are:
~ 25 ft (saturated thickness of Layer 1)
• ~ 50 ft (saturated thickness of Layers 1-3)
~ 75 ft (saturated thickness of Layers 1 -4)
Based on the equations presented in Figure 13 (estimated flow rate calculation) and Figure 14 (capture
zone width calculation) of the main document, assigning a higher value for aquifer thickness would
suggest that more water flows through the plume, which would, in turn, require more pumping to achieve
a specific width of capture.
Another complication pertains to hydraulic conductivity. The slug test and pumping test data (Table
A1 -1) suggest that heterogeneities exist within each aquifer, and from aquifer to aquifer. However,
these simple calculations require a uniform value for hydraulic conductivity. Some possible options for
assigning hydraulic conductivity are:
• 15 ft/day (representative value for Layer 1)
* 30 ft/day (conservatively high value. Layers 1-3)
Based on the equations presented in Figure 13 (estimated flow rate calculation) and Figure 14 (capture
zone width calculation) of the main document, assigning a higher value for hydraulic conductivity would
suggest that more water flows through the plume, which would in turn require more pumping to achieve a
specific width of capture. Also note that lower hydraulic conductivity (at some sites) may dictate a need
for more wells, at a lower discharge rate per well, to achieve the required total flow rate due to potential
for lower sustained yields at each well.
A third complication pertains to regional hydraulic gradient. One of the assumptions of these simple
calculations is that regional hydraulic gradient is uniform, which is not true for most actual sites due to
aquifer heterogeneity, sources or sinks of water, or other factors. For this example, the magnitude of the
hydraulic gradient ranges from 0.01 ft/ft (near the contaminant source) to 0.025 ft/ft (just south of the
river). Forthe calculations presented below, only one value for hydraulic gradient (0.016 ft/ft) is utilized,
which is a simplification based on the hydraulic gradient in the immediate vicinity of remedy wells
located upgradient of the river.
To perform these calculations, a uniform value must also be assigned for the width of the plume requiring
capture. For this site, the width of the 5 ppb plume is approximately 550 ft (Figure Al-5).
Estimated Flow Rate Calculation
Some combinations of parameter values that can be inserted into the equation presented in Figure 13 of
the main document, for estimating the flow rate required for capture, are listed below:
K = 15 ft/day or 30 ft/day (discussed above)
• b = 25 ft or 50 ft (discussed above)
« w = 550 ft
• i = 0.016 (simplification, discussed above)
-------
The estimated flux (Q) through the plume, with '"factor" of 1.0 (i.e.. not accounting for other sources of
water to the extraction wells), for various combinations of parameter values, is given in Table Al-2:
Table A1-2. Flux Through Plume,
"Factor" = 1.0
K (ft/day)
15
30
b(ft)
25
50
25
50
Q (ft3/day)
3,300
6,600
6,600
13,200
Q (gpm)
17.1
34.3
34.3
68.6
Applying values for "factor" of 1.5 and 2.0, to attempt to account for other sources of water to the
extraction wells, increases the estimate of pumping required to capture the plume width, as indicated in
Tables A1-3 and A1-4.
Table Al-3. Estimated Pumping Required,
"Factor" = 1.5
K (ft/day)
15
30
b(ft)
25
50
25
50
Q (ft3/day)
4,950
9,900
9,900
19,800
Q (gpm)
25.7
51.4
51.4
102.8
Table Al-4. Estimated Pumping Required,
"Factor" = 2.0
K (ft/day)
15
30
b (ft)
25
50
25
50
Q (ft3/day)
6,600
13,200
13,200
26,400
Q (gpm)
34.3
68.6
68.6
137.1
These results illustrate how the assignment of parameter values impacts the results of this simple
calculation. It suggests that anywhere from 17.1 gpm to 137.1 gpm is required to capture the plume at
this site, given the simplification for uniform hydraulic gradient, based on different estimates for
hydraulic conductivity, aquifer thickness, and "factor".
-------
The results of the estimated flow rate calculations, which suggests anywhere from 17.1 gpm to 137.1 gpm
might be required to capture the plume at this example site, can then be compared to the pumping rate for
each of the three pumping scenarios defined for this example (Figure A1-4), to interpret whether or not
the pumping rate associated with each scenario is likely sufficient to provide capture:
• Scenario 1 (22.5 gpm): likely not enough pumping
Scenario 2 (44 gpm): possibly enough pumping
• Scenario 3 (107 gpm): likely enough pumping
These interpretations are semi-quantitative. They are quantitative in that they compare existing pumping
rates to calculated values for pumping that might be required for successful capture, but they are also
somewhat subjective because the calculated values are in the form of a range due to the significant
uncertainty in the underlying parameters (for the reasons discussed earlier). Note that these flow-rate
calculations do not provide any insight regarding vertical capture.
Capture Zone Width Calculation
As with the estimated flow rate calculations, a range of parameter values must be considered for
transmissivity (T = K * b) since a range of possible values for both hydraulic conductivity and aquifer
thickness are possible. For the calculations presented below, only one value for hydraulic gradient (0.016
ft/ft) is utilized, which is a simplification based on the hydraulic gradient in the immediate vicinity of
remedy wells located upgradient of the river (as discussed earlier with respect to the estimated flow rate
calculation).
Each of the three pumping scenarios defined for this example have more than one extraction well. The
capture zone width calculation is generally performed by assigning the total extraction rate to one
"equivalent well". The location of the equivalent well for each of the three pumping scenarios defined for
this example is illustrated in Figure Al-18. The location of the equivalent well is generally selected
visually so it is centrally located with respect to the plume width and/or extraction well locations,
and located at the most downgradient position of the actual extraction wells. This often represents a
significant level of simplification for a multi-well extraction system.
Capture zone widths calculated for the three pumping scenarios defined for this example, assuming one
centrally located extraction well, are illustrated in Figure Al-18.
As discussed in Section B of the main document (see Figure 14 of the main document), Ymax is the capture
zone width far upgradient of the equivalent well, and Yweu is the capture zone width at the location of
the well. Both should be considered. For the purpose of this example, the full capture zone that includes
both Ymax and Yweu is illustrated graphically (Figure Al-18), and calculations of Ymax are evaluated in
detail below. Since the plume width at this site is 550 ft, and Ymax is measured from the plume centerline,
Ymax should be greater than 275 ft to suggest successful capture.
Calculations for Ymax for each of the three pumping scenarios defined for this example are presented in
Tables Al-5, Al-6, and Al-7. These calculations account for some of the variations of parameter values
discussed previously. As noted in Figure 14 of the main document, these capture zone width calculations
require that consistent units be used. Therefore, pumping rate (Q) is converted from gpm to ft3/day prior
to the calculation.
-------
Figure Al-18
Capture Zone Width Calculations
Capture Zone Width Calculations, Scenario 1
(22.5 gpm)
30ft/day
Capture Zone Width Calculations, Scenario 2
(44gpm)
1200 1400 1600 1800 2000 2200 2400 2600 ("
15ft/day
25ft
15Wday
50ft
r
30ft/day
30tt/day
1200 1400 1i
Capture Zone Width Calculations, Scenario 3
(107 gpm)
30Wday
50ft
1200 1400 1600 1600 2000 2200 2400 2600 (ft)
-------
Table Al-5. Capture Width from Plume Centerllne
Scenario 1 (22.5 gpm, or 4,331 ft3/day)
K (ft/day)
15
30
b(ft)
25
50
25
50
T (ft2/day)
375
750
750
1500
Ymax (ft)
360
180
180
90
Table Al-6. Capture Width from Plume Centerline
Scenario 2 (44 gpm, or 8,470 ft3/day)
K (ft/day)
15
30
b (ft)
25
50
25
50
T (iWday)
375
750
750
1500
Ymax(ft)
706
353
353
176
Table Al-7. Capture Width from Plume Centerline
Scenario 3 (107 gpm, or 20,597 ft3/day)
K (ft/day)
15
30
b (ft)
25
50
25
50
T (ft2/day)
375
750
750
1500
Ymax (ft)
1,715
858
858
429
For each pumping scenario, the calculated value for Ymax can be compared to the target capture width on
either side of the plume centerline (275 ft). As with the estimated flow rate calculations presented earlier.
the calculated capture zone width for each of the defined pumping scenarios is actually a range, due to
the uncertainty in assigning a uniform value for some of the parameters. The following semi-quantitative
interpretations are made with respect to horizontal capture (based on the range of calculated values for
Y.
Scenario 1 (22.5 gpm):
Scenario 2 (44 gpm):
Scenario 3 (107 gpm):
likely not enough pumping (Ymax generally less than 275 ft)
possibly enough pumping (Ymax generally more than 275 ft)
likely enough pumping (Ymax always more than 275 ft)
Similar detailed evaluations could also be performed for Ywen
-------
Note in Figure Al-18 that the calculated capture zone boundary in some cases crosses the river, because
the analytical solution does not account for the river contributing water to the remedy well located south
of the river. This highlights one of the limitations of these simple calculations (i.e., they do not account
for other sources or sinks of water).
Exhibit Al-1 highlights questions that should be asked when performing these simple analyses, plus
answers to those questions, for this example. Based on the answers to those questions, other lines of
evidence are needed at this site to adequately assess capture for each of the three pumping scenarios.
Exhibit Al-1
Questions Asked When Performing Simple Horizontal Capture Analyses
Is using a single "representative value" for hydraulic conductivity adequate?
Probably not in this hypothetical example
Are other contributions to the extraction wells adequately considered?
Flux from a river: extraction wells are close to the river, there is high potential for the river to contribute
water to the extraction wells
Flux from other stratigraphic units: there is already uncertainty in thickness (b) to use for the shallow
aquifer, there is also potential for extraction well(s) to capture water from the deep aquifer because no
aquitard exists
Is potential for vertical transport of contaminants being considered by these methods?
No, and it is a potential concern in this hypothetical example due to the presence of water supply wells
screened in the deep aquifer
Particle Tracking Based on Numerical Ground-Water Flow Modeling
For this example, particle tracking was performed with the existing numerical flow model upon which
the example was generated (the "actual condition"). The following approach to particle tracking was
employed:
in each model layer, one particle was initially located in each model grid cell, at the vertical
midpoint of the layer, and tracked forward in space to the location where it was removed
(such as a well or the river)
for each particle removed by one of the remedy extraction wells, a symbol was plotted at the
initial location of that particle, to identify the specific well that captured the particle
This is a very effective particle tracking approach to illustrate three-dimensional capture zones.
The particle tracking results for each of the three pumping scenarios are presented in Figures Al-19 to
A1 -21. For each pumping scenario, capture zone maps are provided for particles starting at the vertical
-------
Figure Al-19
Particle Tracking Results, Scenario 1 (22.5 gpm)
Note: All the extraction wells are screened in layer 1
Layer 1
(ft)
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
Layer 2
A
r* TCE 5ppb
•f Extraction Well
(upper horizon only)
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
Note: When this figure is viewed in black-and-white,
the extent of the total capture zone is illustrated. When
this figure is viewed in color (such as from within the PDF
digital version), the colors additionally highlight the
capture zones of individual wells.
River
/
A
TCE 5ppb
* Extraction Well
(upper horizon only)
1200 1400 1600 1600 2000 2200 2400 2600 (ft)
-------
Figure Al-20
Particle Tracking Results, Scenario 2 (44 gpm)
Note: All the extraction wells are screened in layer 1
Layer 1
Layer 2
00 1600 1800 2000 2200 2400 2600 (ft)
(ft)
2400-
2200-
1800-
1600-
1400-
1200-
1000-
800-
~~ f
j
River
N
r
/
::!i[:!
';;::;::;:
ill
m li!':1jj!
A
i
* TCE 5ppb
• Extraction Well
(upper horizon only)
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
Note: When this figure is viewed in black-and-white,
the extent of the total capture zone is illustrated. When
this figure is viewed in color (such as from within the PDF
digital version), the colors additionally highlight the
capture zones of individual wells.
(ft)
2600-
2400-
2200-
1800-
1600-
1400-
1200-
1000-
800-
12
^
A
f /..*.**,+
River .ijiiii.
iilijiiii i:iiij:iiiiii
'1 IIP
)
TCE 5ppb
4- Extraction Well
(upper horizon only)
30 1 400 1 600 1 800 2000 2200 2400 26
30 (ft)
-------
Figure Al-21
Particle Tracking Results, Scenario 3 (107 gpm)
Note: All the extraction wells are screened in layer 1
Layer 1
Layer 2
A
Continuous Sources
(upper horizon only)
y
TCE 5ppb
Extraction Well
(upper horizon only)
1200 1400 1600 1800 2000 2200 2400 2600 (ft)
Layer 3
Note: When this figure is viewed in black-and-white,
the extent of the total capture zone is illustrated. When
this figure is viewed in color (such as from within the PDF
digital version), the colors additionally highlight the
capture zones of individual wells.
00 2000 2200 2400 2600 (ft)
-------
midpoint of Layer 1, Layer 2, and Layer 3 (i.e., each horizon of the shallow aquifer). No particles starting
at the vertical midpoint of Layer 4 (the deep aquifer) are removed by any of the remedy extraction wells.
Although these shallow extraction wells may help prevent contaminants from migrating down to the deep
aquifer in the future, the particle tracking results indicate that they will not remove contamination that has
already migrated down to the deep aquifer.
The particle tracking results for this example are summarized below:
for Scenario 1 (22.5 gpm), illustrated in Figure Al-19, the 5-ppb plume in Layer 1 is mostly
captured, but the 5-ppb plumes in Layers 2 and 3 are largely not captured, and TCE-impacted
ground water likely discharges to the river or flows beneath the river
for Scenario 2 (44 gpm), illustrated in Figure A1-20, effective capture is indicated for Layer
1, and nearly complete capture is indicated for Layers 2 and 3
for Scenario 3 (107 gpm), illustrated in Figure A1-21, complete capture is indicated for
Layers 1 to 3, and the capture extends well beyond the 5 ppb plume boundary, indicating that
there is more pumping than is actually required to capture the plume
The particle tracking results present a more comprehensive and precise illustration of the extent of
horizontal and vertical capture than the evaluations of water level maps or water level pairs. However,
the reliability of this line of evidence for interpreting actual capture depends on the reliability of the
model predictions, which are typically subject to uncertainty based on the presence of heterogeneity in
natural systems that can be difficult to characterize and represent in the model.
For the example presented herein, the particle tracking evaluations were performed with a model known
to be an accurate representation of ground-water flow conditions for the site. At actual sites, that will
never be the case. Ideally, a numerical model used for particle tracking should be "verified" by
reproducing measured drawdown responses to various pumping scenarios, increasing confidence in the
model's ability to accurately predict capture. Another way to "verify" the numerical model is to run
forward particle tracking to show that particles released at the source actually account for the observed
plume dimensions. However, there are factors in addition to the prediction of the flow system (including
contaminant source location and timing, dispersion, and adsorption) that complicate such evaluations.
Sensitivity analysis can be performed to evaluate how changes in model parameter values might impact
the particle tracking results. The best way to perform such sensitivity analysis is to make sure the
alternate runs are performed using parameter values in the ground-water flow model that still provide an
acceptable model calibration.
Step 5 - Evaluate Concentration Trends
At this site, the use of concentration trends at monitoring wells that are located between the extraction
wells and the river is difficult with respect to evaluation of capture. Sentinel wells (i.e., wells that are
currently clean) cannot be located between the extraction wells and the river, because that area is already
impacted by TCE. "Downgradient performance monitoring wells" should be located beyond the Target
Capture Zone, because monitoring wells located within the capture zone will often remain impacted if
there is a continuing source of contamination (see Figure 15 of the main document and associated
discussion). For this example, it is not clear that any monitoring wells between the extraction wells and
the river could be known to be located outside the Target Capture Zone, since the capture zone might
extend all the way to the river. However, declining concentrations at performance monitoring wells
-------
located north of the river could, overtime, provide evidence of successful vertical capture from the
extraction wells located south of the river (i.e., successfully preventing continued migration of
contaminants beneath the river). Other types of performance monitoring, such as stream bed wells or
pore water samplers, could be used to establish concentration trends in ground water immediately
adjacent to the river that would provide evidence regarding the success of horizontal capture in the
shallow aquifer.
Figure Al-22 illustrates the locations of several potential monitoring wells in Layer 1 (i.e., screened in the
upper horizon of the shallow aquifer), located between the extraction wells and the river (more than
would typically be available at most sites). Some are located closer to the extraction wells, and some are
located closer to the river. Also, some are located immediately north of the extraction wells (i.e., directly
downgradient of the extraction wells), while some are located between extraction wells (i.e., north of, but
not directly downgradient of, the extraction wells).
Figure Al-23 illustrates concentration trends observed at these locations, for each of the three denned
pumping scenarios.
Figure Al-22
Candidate "Downgradient Performance Monitoring Wells"
(ft)
91 no
£. IUU
2050
2000
1950
1900
A OCA
ToOU
•\of\r\
N
A x
^ TCE 5ppb
+ Extraction Well
A Monitoring Well
loUUn I " I
1600 1650 1700
River
A
^
MW-1 MW-4 MW-2 MW-5 MW-3
-fy. -§- -fy. -cj,- -^
MW-9 MW-7
-A- >
* * * ft
Groundwater
Flow Direction
Continuous source located approximately 750 ft to the south
1750 1800 1850 1900 1950 2000 2050 2100 2150 2200 (ft)
-------
Figure Al-23
Concentration Trends
A. Concentrations vs. Time:
Scenario 1 (22.5gpm)
B. Concentrations vs. Time:
Scenario 2 (44 gpm)
C. Concentrations vs. Time:
Scenario 3 (107 gpm)
-MW-1 -»-MW-2 -4-MW-3
-MW-4 -H-MW-5 -t-MW-7
-MW-9
Cleanup Standard
19 20 21 22 23 24 25 26 27 28 29 30
Time (year)
i
1
I
-•-MW-1 -"-MW-2 -t-MW-3
-K-MW-4 -«-MW-5 -i-MW-7
MW-9
MW-7 and MW-9, immediately north of pumping wells
Cleanup
Standard
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 26 29
Time (year)
-MW-1 -«-MW-2 -*-MW-3
-MW-4 -«-MW-5 -t-MW-7
-MW-9
Cleanup Standard
t i t i t i t : t i t : t i t : t
0 1 2 3 4 5 6 7 6 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Time (year)
-------
Interpretations of the contaminant concentration trends presented in Figure Al-23 are as follows:
for Scenario 1 (22.5 gpm), concentrations remain above 5 ppb at all the monitoring wells.
possibly because capture fails, but possibly because all are within the capture zone and are subject
to continued impacts as a result of the continuing contaminant source
• for Scenario 2 (44 gpm), concentrations over time decline at all of the monitoring wells, but only
decline below 5 ppb at some of the wells
• for Scenario 3 (107 gpm), concentrations decline below cleanup levels at all the monitoring wells,
likely because the upgradient source is controlled by mid-plume wells and clean water is being
pulled towards the monitoring wells from the river
At most sites only a subset of these data would actually be available, making interpretations regarding
capture even more difficult. Also, for monitoring wells where the concentrations do ultimately decrease
below 5 ppb, that result is not observed for a number of years (i.e., not a timely evaluation of capture
effectiveness). It is also apparent that monitoring wells can initially show a decline in concentrations but
then level off at a concentration higher than the cleanup level of 5 ppb, making it difficult or impossible
to make conclusions about capture based on declining concentration trends in early time periods.
Interestingly, the monitoring wells in this group that remain above the 5 ppb cleanup limit in Scenario 2
(MW-7 and MW-9) are those located closest to the extraction wells. This could be because those wells
are within the capture zone of the extraction wells and the wells closer to the river are not. It could also
be because that all the monitoring wells are within the capture zone of the extraction wells, and those
monitoring wells closest to the extraction wells experience higher concentrations than those located closer
to the river. The particle tracking results provide an additional line of evidence for determining which is
more likely.
Step 6 - Interpret Capture
Once multiple lines of evidence are developed based on technical evaluations such as those presented
above, the next step is to use "converging lines of evidence" to interpret the actual capture zone, and to
compare it to the Target Capture Zone. Exhibit A1 -2 provides a brief summary of each line of evidence
regarding horizontal and vertical capture, for each of the three pumping scenarios. Exhibit Al-2 also
provides an interpretation of capture effectiveness relative to the two options presented earlier for Target
Capture Zone. A summary table such as the one presented in Exhibit Al-2 is an effective way to
summarize a capture zone evaluation.
Target Capture Zone Option 1 is the more conservative option. It requires inward flow from the river to
aquifer, and also requires upward flow to the upper horizon of the shallow aquifer. Only Scenario 3
achieves these more conservative conditions.
Target Capture Zone Option 2is the less conservative option. It requires that a flow divide be established
between the extraction wells and the river, but does not require inward flow from the river to the aquifer
and does not require complete vertical containment (but will require monitoring in the deeper aquifer on
the north side of the river to make sure the remedy is allowing adequate attenuation of constituents
previously flowing beneath the river).
-------
Exhibit Al-2
Lines of Evidence Regarding Capture for Example Al, for Each Pumping Scenario
Method
Water Level Maps
Water Level Pairs
Estimated Flow
Rate & Capture
Zone Width
Calculations
Particle Tracking
with Ground- water
Flow Modeling
Monitoring Well
Concentration
Trends Between
the Extraction
Wells and the
River
Scenario 1: 22.5 gpm
• Horizontal capture may
or may not be achieved
• Using water levels
measured near
extraction wells,
appears horizontal
capture in Layer 1 is
achieved
• Downward flow from
Layer 1 near source
area and most of the
plume
• Discharge from aquifer
to river across entire
plume width between
the extraction wells and
river, presence of
divide cannot be
determined
• Downward flow in
most areas except area
near the extraction
wells
• Likely insufficient for
horizontal capture
• Nearly complete
capture in upper
horizon of shallow
aquifer, poor capture in
lower horizons of
shallow aquifer
• Concentrations do not
reach cleanup level of 5
ppbatany of the
monitoring wells
• Interpretation
ambiguous
Scenario 2: 44 gpm
• Horizontal capture may
or may not be achieved
• Using water levels
measured near
extraction wells, appears
horizontal capture in
Layer 1 is achieved
• Downward flow from
Layer 1 in the mid-
plume area
• Discharge from aquifer
to river across entire
plume width, presence
of divide between the
extraction wells and
river cannot be
determined
• Downward flow in
some portions of the
plume
• Potentially sufficient
for horizontal capture
• Complete capture in
upper horizon of
shallow aquifer, nearly
complete capture in
lower horizons of
shallow aquifer
• Concentrations reach
cleanup level of 5 ppb
at some monitoring
wells in 4-7 years
• Interpretation
ambiguous
Scenario 3: 107 gpm
• Horizontal capture may
or may not be achieved
• Using water levels
measured near
extraction wells,
appears horizontal
capture in Layer 1 is
achieved
• Upward flow within the
entire footprint of the
plume
• Discharge from river to
aquifer across entire
plume width, indicating
horizontal hydraulic
containment at river
• Upward flow in all
portions of the plume,
indicating vertical
hydraulic containment
• Likely sufficient for
horizontal capture
• Complete capture in all
portions of the shallow
aquifer
• Concentrations reach
cleanup level of 5 ppb
at all the monitoring
wells within
approximately 5 years
-------
Exhibit Al-2 Continued
Lines of Evidence Regarding Capture for Example #1, for Each Pumping Scenario
Method
Interpretation for
Target Capture
Zone Option 1
(More
Conservative)
Interpretation
for Target
Capture Zone
Option 2 (Less
Conservative)
Scenario 1: 22.5 gpm
• Horizontal condition is
not achieved
• Vertical condition is not
achieved
• Horizontal condition is
possibly, but not likely,
achieved
• Vertical condition is
likely not achieved, but
hard to evaluate without
transport modeling
and/or long-term
concentration trends on
the other side of the
river
Scenario 2: 44 gpm
• Horizontal condition is
not achieved
• Vertical condition is not
achieved
• Horizontal condition is
likely achieved
• Vertical condition is
possibly achieved, but
hard to evaluate without
transport modeling
and/or long-term
concentration trends on
the other side of the
river
Scenario 3: 107 gpm
• Horizontal condition is
very likely achieved
• Vertical condition is
very likely achieved
• Horizontal condition is
very likely achieved
• Vertical condition is
very likely achieved
Note that interpretations with respect to the two options for Target Capture Zone are based on all lines of
evidence presented, and the interpretation might be different if one or more of the lines of evidence was
not available.
For Target Capture Zone Option 2 (less conservative), based on all the lines of evidence presented, the
interpretations are as follows:
• For Scenario 1 (22.5 gpm), the horizontal condition possibly is achieved for the upper
horizon of the shallow aquifer (based on water level maps and particle tracking results), but
likely is not achieved for the lower horizon of the shallow aquifer because there are
downward hydraulic gradients in the shallow aquifer near the source area that will continue to
cause impacts to deeper ground water, and horizontal ground-water capture appears to be
incomplete in the deeper ground-water units primarily based on particle tracking results. The
vertical condition will require monitoring in the deeper aquifer on the north side of the river
to make sure the remedy is allowing adequate attenuation of constituents previously flowing
beneath the river. Monitoring will likely indicate failed capture due to the potential for
downward contaminant transport indicated by water level pairs and subsequent contaminant
transport beneath the river (it may take years of monitoring to reach a conclusion, and
transport modeling could augment the evaluation).
For Scenario 2 (44 gpm) the horizontal condition is likely met based on the water level maps
and particle tracking results. The vertical condition will require monitoring in the deeper
aquifer on the north side of the river to make sure the remedy is allowing adequate
attenuation of constituents previously flowing beneath the river, and that monitoring may or
-------
may not indicate successful capture (it may take years of monitoring to reach a conclusion.
and transport modeling could augment the evaluation).
• For Scenario 3 (107 gpm), horizontal capture is achieved based on water level maps, particle
tracking results, and especially based on inward flow at the river across the entire plume
width. The vertical condition will require monitoring in the deeper aquifer on the north side
of the river to make sure the remedy is allowing adequate attenuation of constituents
previously flowing beneath the river (it may take years of monitoring to reach a conclusion).
Monitoring is likely to indicate successful capture based on the upward head differences
observed south of the river within the footprint of the plume, which adds confidence that
remedy objectives will be achieved.
It should be noted that the interpretations of capture presented above are based on multiple lines of
evidence determined from a variety of technical analyses. If some of those lines of evidence were not
developed, the evaluation of capture might differ for one or more of the pumping scenarios. In all of the
cases discussed above, periodic water quality monitoring of the deep aquifer south of the river and north
of the river will be appropriate to confirm that concentrations decrease over time as a result of the
remedy. This is especially important given the presence of water supply wells that extract deep ground
water on the other side of the river. Also, in all cases stream bed wells or pore water samplers might be
considered to monitor concentration trends in pore water immediately adjacent to the river, which could
provide additional evidence regarding the success of horizontal capture.
Once the extent of actual capture zone has been interpreted, and that interpretation has been compared to
the Target Capture Zone, the following issues should be addressed:
• compare the interpreted capture zone to remedy objectives
assess uncertainties in the capture zone analysis
assess the need for additional characterization or monitoring
• determine if extraction (rates and/or locations) or monitoring should be modified
The remedy objective in this example is to prevent contaminants at concentrations above standards from
discharging to the river and/or flowing under the river. Scenario 3 appears very likely to achieve these
remedy objectives.
For Scenarios 1 and 2, it is difficult to establish if these remedy objectives are likely to be achieved. The
capture zone analysis for Scenario 1 (22.5 gpm) indicates that the remedy objectives will likely not be
achieved because downward hydraulic gradients are observed near the source area, and there is a high
potential (based on the particle tracking results) for contaminants in the deeper horizons of the shallow
aquifer to not be captured by the pumping wells. Nevertheless, it is possible that attenuation of TCE
concentrations due to the remedy, in conjunction with other attenuation mechanisms, may still help
achieve remedy objectives. The capture zone analysis for Scenario 2 (44 gpm) indicates the remedy
objectives will likely be achieved, particularly based on the particle tracking results that indicate nearly
complete capture of ground water within the plume footprint in all horizons of the shallow aquifer.
However, this assessment would be much more uncertain without the particle tracking analysis (or if there
was a high degree of uncertainty associated with the accuracy of the model upon which the particle
tracking results are based).
-------
There will always be uncertainties regarding aspects of the capture zone analysis. In general, the more
conservative the pumping strategy, the more certain each individual line of evidence is likely to be. For
this example, the interpretation of Scenario 3 (107 gpm) is subject to the least uncertainty, because it
satisfies even the more conservative requirements such as inward hydraulic gradients from the river to the
aquifer and upward flow at all available well clusters. The extra pumping associated with Scenario 3
(both in terms of the total rate and the use of mid-plume and source area wells) is more conservative, and
therefore reduces uncertainty in the capture zone analysis. In general, the reliability of the assessment of
capture often increases as the total pumping rate increases, even if uncertainty in some aspects of the
capture zone analysis remains.
At this example site, no need for additional characterization appears necessary. However, the evaluation
of water level maps illustrates that water level measurements (or estimates of water levels) near the
extraction wells are vital to interpreting capture from the water level maps, and if piezometers are not
available near the extraction wells, installing piezometers in those locations is strongly recommended.
If pumping for Scenario 1 is in place, the capture zone evaluation suggests that more pumping is likely
required mid-plume and/or near the source area, to improve control of downward gradients. Installing
remediation wells near the river with deeper well screens could also be considered, which allows water
to be drawn from a different (i.e., deeper) portion of the aquifer, which in turn might also sustain greater
pumping rates if designed properly. However, the deeper pumping near the river could risk drawing more
of the plume into the deeper layers, making more plume mass available for capture by water supply wells
located north of the river. Thus, pumping from deeper levels may not be prudent.
If pumping for Scenario 2 is in place, increasing pumping rate (mid-plume or near the river) could be
considered to improve confidence that remedy objectives are met, especially if cost impacts associated
with that action are reasonable (site-specific). If pumping for Scenario 3 is in place, consideration could
be given to reducing pumping rates, especially if the cost of operating individual wells and/or treating
more water is reasonably high (site specific).
-------
EXAMPLE A2
This example highlights complications of evaluating capture when preferential flow pathways are present.
For this hypothetical site, not all six steps associated with the systematic evaluation of capture are fully
demonstrated. Instead, specific items that demonstrate important aspects of capture zone analyses are
highlighted.
Example Setup
The area of interest is several square miles (Figure A2-1). The stratigraphy of this site consists of an
aquifer approximately 300 ft thick that overlies a competent aquitard. The aquifer primarily consists of
coastal plain sand. According to regional data, net recharge from precipitation at the site is expected to be
12 to 16 inches per year. Based on regional data and site-specific slug test data, the horizontal hydraulic
conductivity of the aquifer is heterogeneous and varies from 2 ft/day to 300 ft/day. Regionally,
preferential pathways associated with historic stream channel deposits are known to exist.
Figure A2-1
Actual Hydraulic Conductivity Distribution for One Model Layer,
and Pre-Remedy Water Table Map
1200CK
10000^
2000
4000
6000
8000
10000
12000 (ft)
-------
A river located west of the site flows from southwest to northeast. Pre-remediation measurements for the
water table are also illustrated in Figure A2-1, and they indicate that ground-water flow directions are
somewhat variable (to the northwest and the north). The river does not influence the head contour map
because the impacted aquifer is not in hydraulic connection with the river.
A ground-water flow model was used to generate this hypothetical example. The aquifer is represented
in the model with 12 layers, each of which has a heterogeneous hydraulic conductivity. Figure A2-1
indicates the assumed hydraulic conductivity distribution for one layer of this heterogeneous system.
There is a zone of generally higher hydraulic conductivity associated with a preferential pathway.
running north-south, that is apparent in Figure A2-1. In reality, site managers would only have slug test
or pumping test data at a limited number of locations and depths (plus information such as well borings,
well records from nearby wells, and regional hydrogeology reports), and the existence of this preferential
pathway might or might not be evident.
At this site the aquifer has been impacted by dissolved RDX, a contaminant associated with manufacture
of explosives. Dissolved RDX is mobile in the subsurface. The RDX leached into the ground water from
a drainage ditch over a 15-year period, after which the contaminant source was excavated.
Figure A2-2 indicates the pre-remedy concentrations of RDX in the model (maximum concentration at
any vertical depth within the aquifer). The remedy objective for this site is to hydraulically contain the
plume along the property boundary at all depths. The Target Capture Zone is to create a flow divide
between the remedy extraction wells and the property boundary for the entire aquifer thickness, across the
entire width of the plume (defined by the 2 ppb contour for RDX).
Figure A2-2
Fre-Remedy RDX Concentration
(ft)
1200O
1000O
800O
600O
400O
200&
Concentration in ppb
River
/ Property Line
2000
4000
6000
8000
10000 12000 (ft)
-------
Scenarios Illustrated
Figure A2-3 illustrates two pumping scenarios for this site:
• Scenario 1 has 5 pumping wells with the total pumping rate of 1,500 gpm.
• Scenario /"has 6 pumping wells with the total pumping rate of 1,800 gpm.
In each scenario, the pumping rate at each individual well is 300 gpm. The extraction wells are screened
over the entire thickness of the aquifer.
Figure A2-3
Pumping Scenarios
Scenario 1: 1,500 gpm Total
(5 wells, 300 gpm each)
Scenario 2: 1,800 gpm Total
(6 wells, 300 gpm each)
1000
"2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 (ft) 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 (ft)
Items Highlighted for this Example
For this example, the following items pertaining to capture zone are presented:
• the impact that the number of monitoring locations can have on plume delineation, which
impacts the width of the Target Capture Zone
the impact that a variable regional flow direction can have on capture zone width calculations
the impact that multiple extraction wells that are not oriented perpendicular to ground-water
flow direction can have on capture zone width calculations
-------
the impact that the absence of water level data at or near extraction wells can have on
interpretation of capture from water level maps
• the impact that heterogeneous aquifer conditions (e.g., preferential pathways) can have on
capture zone evaluation
• the use of gradient vector maps
Each of these items is presented below.
Plume Delineation
Plume delineation is associated with Step 1 of capture zone analysis. The plume delineation impacts the
definition of the Target Capture Zone (Step 2), which for this site is based on the extent of the plume
defined by the 2-ppb contour.
Figure A2-4 includes an illustration of the interpreted 2-ppb plume boundary, based on two different sets
of available monitoring data:
the figure on the left has fewer available points, such that the plume width is more uncertain (such
as in the northeast portion of the plume)
• the figure on the right has more available points, so the delineation of the plume is more certain
For this site, the interpreted plume based on fewer points has greater width, which in turn would increase
the size of the Target Capture Zone.
Figure A2-4
Plume Delineation
A. Plume Delineation: Less Points
B. Plume Delineation: More Points
2000 4000
10000 (ft)
2000 4000
8000 10000 (ft)
-------
Impact of Regional Flow Direction on Capture Zone Width Calculations
Capture zone width calculations are associated with Step 4 of a systematic capture zone analysis. One
assumption associated with these calculations is that regional hydraulic gradient is uniform (in both
magnitude and direction).
As illustrated in Figure A2-1, regional hydraulic gradient at this site is variable with respect to direction.
so it is not clear which direction is most appropriate for orienting the calculated capture zone for
comparison to the plume boundary. In this case, the best option might be to use the flow direction closest
to the toe of the plume where the extraction wells are located, which is more northerly than in some other
portions of the plume.
Impact on Capture Zone Width Calculations When Extraction Wells are not Oriented Perpendicular
to Ground-Water Flow
The estimated capture zone width for multiple extraction well scenarios will generally be similar to a
single "representative well" case if all wells are oriented perpendicular to direction of regional flow and
the pumping is evenly split between the wells. For this example, the pumping rate is split evenly among
multiple extraction wells, but the wells are not oriented perpendicular to regional flow direction (compare
water levels in Figure A2-1 with the orientation of extraction wells for each scenario in Figure A2-3). The
well locations are more closely aligned with the property boundary than with the direction of ground-
water flow. Therefore, a primary assumption of the capture zone width calculation for multiple wells is
violated. In this case, an "equivalent well" will never accurately represent the actual multi-well capture
zone due to the mis-alignment of the wells with the regional hydraulic gradient (in this case they are
instead aligned with the property boundary). This again illustrates that these simple calculations are often
of limited use because the simplifying assumptions they are based on do not allow the complexity of the
actual system to be adequately represented.
Impact of Water Level Data At/Near Extraction Wells on Interpretation of Water Level Maps
Interpreting capture based on water level maps is associated with Step 3 of a capture zone analysis.
Figure A2-5 illustrates two water level maps for each pumping scenario at this site. One of the water
level maps for each pumping scenario includes water level estimates at the extraction wells, and the other
water level map for each scenario does not include water level estimates at the extraction wells.
When water levels near the extraction wells are not available, capture is not apparent or easily interpreted.
Interpretation in those cases is biased towards an interpretation of poor capture. When water levels near
the extraction wells are available, a completely different interpretation of capture can occur.
Impact of Heterogeneous Conditions (e.g., Preferential Pathways) on Capture Zone Evaluation
For this site, a potential supporting evaluation of capture is particle tracking in conjunction with a
numerical model. Figure A2-6 illustrates simulated particle pathlines for each of the two pumping
scenarios. Particles are released at the plume boundary in different layers, and tracked forward. For
Scenario 1 (1,500 gpm) capture appears to fail in the area of the preferential pathway. For Scenario 2
(1,800 gpm), which includes an additional extraction well in that area, capture appears to succeed.
-------
Figure A2-5
Water Level Maps
Scenario 1: Observation Points without Water Level
Estimates at Pumping Wells
Scenario 1: Observation Points with Water Level
Estimates at Pumping Wells
^/Property/Line/ /
Interpreted
2000 3000 4000 5000 6000 7000 8000 9000 10000 11000(11) 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000(0)
Scenario 2: Observation Points without Water Level
Estimates at Pumping Wells
Scenario 2: Observation Points with Water Level
Estimates at Pumping Wells
Interpreted s
Capture Zone
2000 3000 4000 5000 6000 7000 8000 9000 10000 11000(ft) 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000(ft)
4- Pumping wells
* Monitoring wells
Water level contours
RDX 2ppb
-------
Figure A2-6
(ft)
11000
Capture Zone Width Calculation and Particle Tracking Results
Failed Capture, Scenario 1 (l,500gpm)
Successful Capture, Scenario 2 (1,800 gpm)
with Pumping rate 800 gpm
2000 3000 4000 5000 6000 7000 8000 9000 10000 11000(ft) 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000(ft)
For performing capture zone width calculation: + Pumping weiis --- capture zone width calculation
Q = 800 gpm or 1 ,500 gpm # Monitoring wells
K = 75 fWay (representative value) _ Particle patns
b = 300 ft
i = 0.013
Capture zone widths (Figure A2-6) were calculated for 800 gpm (arbitrarily selected) and 1,500 gpm
(same pumping rate as Scenario 1). This was done assuming one "equivalent well" and a regional flow
direction orientated slightly east of north. There is no specific method for selecting the location of the
"equivalent well", and it was somewhat arbitrarily selected. The regional flow direction was selected
qualitatively based on the measured water levels (Figure A2-1, which pertains to one point in time) plus
the orientation of the interpreted plume. Selecting other orientations for uniform flow direction would
lead to different orientations of the illustrated capture zones. Again, there is no easy or "correct" way to
determine the location of the "equivalent well" or the uniform flow direction to utilize, because of the
complexity of the actual system relative to the simplified assumptions associated with the calculation.
For Scenario 1 (1,500 gpm), the capture zone width from particle tracking is much smaller than the
corresponding capture zone width calculation, probably because the calculation of capture zone width
uses a uniform value of hydraulic conductivity. This does not accurately represent the aquifer,
particularly in the area of the preferential pathway. To be more conservative, the calculation of capture
zone width could use a hydraulic conductivity value at the high end of the expected range, but it still
might overestimate capture near the preferential pathway.
-------
Use of Gradient Vector Maps to Interpret Capture
Figure A2-7 presents a gradient vector map for the water level maps presented on the right-hand side of
Figure A2-5. These maps were produced using the same software that produced the water level contours
(i.e., the software produces the gradient vectors based on the contours). Note that the gradient vectors for
Scenario 1 (left-hand side of Figure A2-7) indicate a potential gap in capture, consistent with the particle
tracking results.
Figure A2-7
Gradient Vector Maps
Scenario 1: Gradient Vector Map (l,500gpm) Scenario 2: Gradient Vector Map (1,800 gpm)
(ft)
11000
2000 3000 4000 5000 6000 7000 8000 9000 10000 11000(11)
+ Pumping wells 1
* Monitoring wells - -
RDX2ppb
2000 3000 4000
Gradient vector
Interpreted capture zone
"\X \.\\A.\\i . "\ « \ . \ , \ \
\^. \ \ \ \ j \ \ \ \
\\x\\\\\\\y \\\\\N\N
> \ N \v\ ^ \ ^ \y^\ \ \ \ v o v
\N N ^s v ^r^ \ v > N
6000 7000 8000 9000 10000 11000(ft]
-------
EXAMPLE A3
This example highlights complications of evaluating capture when off-site stresses are present. For this
hypothetical example, not all six steps for capture zone analysis are fully demonstrated. Instead, specific
items that demonstrate important aspects of capture zone analyses are highlighted.
Example Setup
The location of hypothetical site is illustrated in Figure A3-1. Land surface is generally flat. The
stratigraphy consists of two aquifers (a surficial aquifer and a deeper aquifer) separated by an aquitard that
is regionally discontinuous. However, the aquitard has been identified in all site borings to date. The site
is not located close to any surface water bodies.
Based on regional data, the hydraulic conductivity in the surficial aquifer varies over a narrow range
(approximately 28 ft/day). Regionally, the net recharge to the aquifer from precipitation is estimated at 9
to 15 inches per year.
The contaminant of concern at the site is dissolved TCE (Figure A3-1). Regional ground-water flow
direction in both aquifers is to the north. Pre-remedy water levels in the surficial aquifer are illustrated in
Figure A3-1. The plume extends approximately 1000 feet downgradient (i.e., north) of the site boundary
in the surficial aquifer. The sources of contamination are located in the unsaturated zone, and the plume
primarily impacts the surficial aquifer. Several sources of the TCE have been identified, and those
sources have not yet been removed (further source area characterization is ongoing in advance of future
source area remediation).
The site currently has an interim remedy. The objective of the interim remedy for this site is to prevent
contaminants at concentrations above standards (5 ppb) from migrating beyond the property boundary in
the future using extraction wells located on the site property. Ground-water quality in the shallow aquifer
beyond the property boundary will continue to be monitored during the operation of the interim remedy.
No active remediation for the deep aquifer is anticipated as part of this interim remedy, and long-term
monitoring in the deep aquifer will continue to be performed to determine if concentrations in the deep
aquifer decline over time based on the performance of the interim remedy in the shallow aquifer.
Based on these remedy objectives, the Target Capture Zone for the interim remedy only applies to
horizontal capture in the shallow aquifer, and is defined for this site as a flow divide downgradient of the
extraction wells over the entire width of the 5-ppb TCE plume. One representation of a Target Capture
Zone is illustrated in Figure A3-2.
Scenarios Illustrated
This example is used to demonstrate the impact of off-site stresses on capture zone analysis for the
interim remedy. Two pumping scenarios are presented (Figure A3-3):
• Scenario 1 has 3 pumping wells screened in the surficial aquifer, with a combined pumping rate
of 21 gpm.
Scenario 2 has the same 3 pumping wells (21 gpm combined) as Scenario 1, plus an off-site
well pumping at 30 gpm that has recently been installed on a neighboring property (i.e., after
the remedial design for the interim remedy was implemented).
-------
Figure A3-1
Plume Maps and Pre-Remedy Water Levels
TCE Plume, Surficial Aquifer
Site
Boundary
5 ppb
20ppb
50 ppb
100 ppb
200 ppb
500 ppb
Continuous Sources
(Surficial Aquifer Only)
(ft)
2400
TCE Plume, Deep Aquifer
Site
Boundary
A /
V
Continuous Sources
(Surficial Aquifer Only)
BOO 1000 1200 1400 1600 1800 2000(ft)
BOO 800 1000 1200 1400 1600 1800 2000(ft)
Pre-Remedy Water Level Map, Surficial Aquifer
1000 1200 1400 1600 1800 2000(ft)
-------
Figure A3-2
Target Capture Zone
(ft)
LEGEND
+ Extraction Well
Plume
Site
Boundary
One version of a
Target Capture Zone
Continuous Sources
(Surficial Aquifer Only)
BOO 800 1000 1200 1400 1600 1800 2000(ft)
Figure A3-3
A. Well Locations and
(ft)
2200-
2000-
1800-
1600-
1400-
1200-
1000-
800-
600-
LEGEND
Pumping Scenarios
Rates (gpm), Scenario 1 B. Well Locations and Rates (gpm), Scenario 2
(ft)
Extraction Well & Rate (gpm)
9
J^^l^
r+ \ so
\ -*-
Site _^ _ T \ }
Boundary \A /y J Off.site
^V» c \\f f \^^jr OUDD|V wsll
\/
Continuous Sources
(Surficial Aquifer Only)
1600 1800 2000 (ft) 600 800 1000 1200 1400 1600 1800 2000 (ft)
-------
Items Highlighted for this Example
For this example, the following items pertaining to capture zone analysis are presented:
• the similarity of the capture zone width calculation to the particle tracking results at this site
the importance of monitoring well location with respect to interpretation of capture based on
water level pairs
• the impact an off-site stress can have on the capture zone (and associated capture zone
analysis)
an illustration of using a sentinel well to indicate failed capture, and the importance of
locating sentinel wells in critical locations
Similarity of the Capture Zone Width Calculation to the Particle Tracking Result at this Example Site
Capture zone width calculations are associated with Step 4 of a systematic capture zone analysis. Such
calculations are often of limited utility because one or more of the simplifying assumptions is typically
not met. For this site, however, the surficial aquifer is reasonably homogeneous, there is poor connection
to other aquifers, and there are no nearby surface water bodies to provide water to the extraction wells.
Therefore, a capture zone width calculation would be expected to be more reliable at this type of site.
Figure A3-4 illustrates the comparison of a capture zone width calculation for this site with the simulated
particle tracking results for pumping Scenario 1. To perform the particle tracking, one particle was
initially located in selected model grid cells in the model layer representing the surficial aquifer, vertically
in the middle of the layer, and tracked forward in space to the location where it was removed. For each
particle removed by one of the remedy extraction wells, a symbol was plotted at the initial location of that
particle, to identify the specific well that captured the particle.
For this site the capture zone width calculation approximates the capture zone reasonably well, for the
reasons discussed above.
Importance of Monitoring Well Locations for Water Level Pairs
Figure A3-5 illustrates water level measurements at monitoring well pairs for Scenario 1 (without off-site
pumping) and Scenario 2 (with off-site pumping), respectively. Two sets of water level pairs are
illustrated for each pumping scenario:
water level pairs along the northern property boundary, downgradient of on-site ground-water
extraction wells associated with the interim remedy
water level pairs along the eastern property boundary, between the on-site extraction wells
and the off-site well
For Scenario 1, with no off-site pumping, some of the water level pairs suggest inward flow at the
northern property boundary, and some of the water level pairs suggest outward flow at the northern
property boundary. Note that the locations indicating inward flow are those immediately downgradient of
the extraction wells, and the locations indicating outward flow are those in between the extraction wells.
-------
Figure A3-4
Particle Tracking versus Capture Zone Width Calculation, Scenario 1
(ft)
Regional
Flow
t
Capture Zone
Width Calculation
Boundary
Continuous Sources Y
(Surficial Aquifer qnly)
\f
LEGEND
+ Extraction Well
Plume
Capture Zone
Width Calculation
Continuous Sources
(Surficial Aquifer Only)
\
600 800 1000 1200 £400 1600 1800 2000(ft)
For performing capture zone width calculation:
• Q = 21 gpm
• K = 28 ft/day (based on average of slug test)
• b = 31 ft (based on well logs)
• I = 0.0033 (based on pre-pumping water level maps)
Capture zone width calculation results:
• Xo = -225 ft
•Ymll= 353ft
•Y = 706ft
500 800 1000 1200 1400 1600 1800
2000 (ft)
Note: When this figure is viewed in black-and-white,
the extent of the total capture zone is illustrated. When
this figure is viewed in color (such as from within the
PDF digital version), the colors additionally highlight the
capture zones of individual wells.
This illustrates that it is more likely to observe inward hydraulic gradients immediately downgradient of
the extraction wells. As illustrated in Figure 10 in the main document, the outward gradients at some
locations along the northern property boundary do not specifically indicate failed capture. It is possible
that water flows outward between the wells at the property boundary but eventually flows inward and
reaches the wells. Some potential interpretations regarding outward flow observed at a property boundary
are illustrated schematically in Figure A3-6. In that figure, the top interpretation represents failed capture,
but the bottom two interpretations represent successful capture despite the outward hydraulic gradients at
the boundary.
-------
Figure A3-5
Water Level Pairs
Head Differences at Water Level Pairs, Scenario 1
(ft)
1400-
1300-
1200-
1100-
1000-
900-
N
9.92 jit
10-J9-^9^71 * ^"~**\J0.30 §
10.26^--*^ + 10.24* ^
^--^*i1019* 10.19* ^~~~— -,
,--^fai8* 4" 10.74 *Jt 10.82
10.93 »_4 10.99
11.21 »,_* 11.26
11.44 »l 11.48
^H
11.70»,Vl1.73
^\. \ \ OUH
• JN. \ H \ ^---^— •— oite
•^^5s\ ^~~~^^' r Boundary
^*^^*--^^^^^ /
Continuous Sources /
(Surficial Aquifer Only) / » Water Level
/ 11l4s Measurement
4- Extraction Well
<- Flow Direction
900
1000
1100
1200
1300
1400
1500
1600
1700 1800 (ft)
Head Differences at Water Level Pairs, Scenario 2
1400-
1300-
1200-
1100
1000-
900-
9.72*
9.32
+ 9.50
9.50
"*~~^9.47
9.X
4- 9.74 ij 9.78
9.91 *A 9.93
10.14* * 10.14
\
10.38 »_i 10.37
10.62*_> 10.59
Off-site Well+
(Scenario 2 Only)
— Site
Boundary
Continuous Sources
(Surficial Aquifer Only)
* Water Level
11-48 Measurement
4- Extraction Well
<- Flow Direction
900
1000
1100
1200
1300
1400
1500
1600
1700
1800
-------
Figure A3-6
Potential Interpretations of Outward Flow Between Extraction Wells
Interpretation 1: Outward Flow Due to Failed Capture
Capture Zone
Property Boundary
Pumping Wells
Flow Direction
Flowline
Interpretation 2: Outward Flow Due to Flow Divide Between Wells and Property Boundary
Capture Zone
Property Boundary
Pumping Wells
Flow Direction
Flowline
Interpretation 3: Outward Flow Due to Flow Divide Downgradient of Property Boundary
Capture Zone
Pumping Wells
Flow Direction
Flowline
-------
The capture zone width calculation and the particle tracking results for this example site (Figure A3-4)
suggest that hydraulic containment is in fact achieved for Scenario 1, despite the outward hydraulic
gradients observed between some water level pairs at the northern property boundary (Figure A3-5). To
add conservatism, pumping rates could be increased to provide inward gradients at all water level pairs.
Also note in Figure A3-5, for Scenario 1, that inward gradients are established at every pair along the
eastern property boundary. In general, it is harder to achieve inward gradients at locations downgradient
of extraction wells compared to locations to the side of the extraction wells, due to the influence of the
regional hydraulic gradient.
For Scenario 2, where an off-site well is added after remediation is initiated, the data from the water level
pairs (bottom of Figure A3-5) indicate similar results along the northern property boundary, but also
indicate a potential for outward flow along the eastern property boundary. Particle tracking results for
Scenario 2 are illustrated in Figure A3-7, and add an additional line of evidence that horizontal capture
fails when the off-site pumping is added, because some on-site water reaches the off-site well.
This example illustrates that water level pairs are a useful line of evidence regarding capture, but
generally should be supplemented by other lines of evidence. Also, the analysis of water level pairs along
Figure A3-7
Particle Tracking Results, Scenario 2
(ft)
2400
2200-
2000-
1400-
800-
600-
LEGEND
+ Extraction Well
Site
Boundary
Capture zone of
remediation wells
Capture zone
of off-site well
Continuous Sources
(Surficial Aquifer Only)
600
BOO
1000 1200
1400
Note: When this figure is viewed in black-
and-white, the extent of the total capture zone
is illustrated. When this figure is viewed in
color (such as from within the PDF digital
version), the colors additionally highlight the
capture zones of individual wells.
1600 1800 2000 (ft)
-------
the eastern property boundary, indicating the impacts from the off-site well, is only possible if water level
pairs are located in that area. This illustrates that selection of locations for water level pairs, and the
schedule for evaluating water levels at those locations, should take into account the potential for current
and/or future off-site stresses.
Impact an Off-Site Stress Can Have on the Capture Zone (and Associated Capture Zone Analysis)
The particle tracking results for the two pumping scenarios (Figures A3-4 and A3-7) illustrate that the
addition of an off-site stress can impact the capture zone of an extraction system. This suggests that it is
important for site managers to stay abreast of developments at neighboring properties, or in some cases
changes in pumping at regional water supply wells. Similar impacts to capture zones can occur due to
transient influences such as irrigation pumping or irrigation recharge.
Illustration of a Sentinel Well for Evaluating Capture
Figure A3-8 illustrates several monitoring wells located beyond the property boundary, and concentration
trends overtime at those wells for each of the two pumping scenarios. Monitoring wells MW-ls and
MW-2s are already impacted prior to the remedy. MW-5s is a sentinel well, because it is located in an
area not yet impacted by the site.
For Scenario 1, the concentration trends at MW-ls and MW-2s are similar to those observed for the other
examples in this document. At the location closest to the extraction wells (MW-ls) the concentration
remains above cleanup levels, but without other lines of evidence it is hard to interpret if that is because
capture is incomplete, or if it is because MW-ls is within the capture zone. MW-2s, located further
downgradient, eventually cleans up below the 5 ppb cleanup level, but it takes more than 15 years for that
to occur. This example again illustrates the complications of evaluating downgradient performance
monitoring wells to demonstrate successful capture in the absence of hydraulic monitoring data. Each
line of evidence plays a role in the overall evaluation. In this case, the chemical monitoring data over a
long period of time support interpretations from the hydraulic monitoring. The chemical data provide
direct evidence that the interim remedy goals are ultimately achieved (based on long-term concentration
trends at MW-2s) and suggest that MW-ls is likely located within the capture zone.
For Scenario 2, there is a concentration increase above the cleanup level at sentinel well MW-5s
within one year. This type of monitoring would immediately indicate the potential for failed hydraulic
containment. It could also be due to an off-site contaminant source, and further investigation would be
appropriate. By using other lines of evidence regarding capture (such as the water level pairs illustrated
in Figure A3-5 and the particle tracking results illustrated in Figure A3-7), it seems likely that the
concentration increase observed in the sentinel well is due to failed capture. Also, MW-2s does not
reach the cleanup level for Scenario 2, likely because it is within the capture zone of the extraction wells,
whereas it is located downgradient of the capture zone in Scenario 1.
Note that the location of this sentinel well (MW-5s) is not immediately downgradient of the denned
plume, based on background-water levels. A sentinel well would only be placed in the vicinity of MW-5s
if site managers are aware of the potential off-site stress at the neighboring property. This again illustrates
the importance of being aware of potential off-site stresses that may impact the capture zone.
-------
Figure A3-8
Concentration Trends
(ft)
160O
A. Candidate Locations for
Downgradient Performance
Monitoring Wells
-*.
MW 9«! »
Downgradient Performance s
Monitoring Well <;-- — ^~
(impacted prior to remedy) ^\MW_IS
,^-^+ "~^^-~.
^^-- 4. ~~~^ MW-Ss
t
\|
Sentinel Well
(not impacted
prior to remedy)
900 1000 1100 1200 1300 1400 1500 1600 1700 1800 (ft]
B. Concentration vs. Time,
Scenario 1
I.
£
-MW-1S
-MW-Ss
Immediately downgradient of pumping wells
- Farther downgradient of pumping well!
Cleanup Standard
Between site boundary and ofrsite well
-(sentinel well)
18 19 20 21 22 23 24 25
C. Concentration vs. Time,
Scenario 2
1234567
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
Year
-------
APPENDIX B:
TWO ACTUAL
-------
INTRODUCTION TO B:
EXAMPLES TWO ACTUAL SITES
The following examples present capture zone evaluations for two actual sites to demonstrate the
application of the systematic approach presented in the main document. The example sites are as follows:
• Example B1: East Canal Creek Area (ECCA). Aberdeen Proving Grounds, Maryland
* Example B2: Milan Army Ammunition Plant. Operable Unit #4, Tennessee
The sole purpose of these examples is to demonstrate the application of the systematic approach to
capture zone analysis at actual sites by applying the suggested steps in the guidance document to the
sites using the existing site data, maps, and records. No attempt is made to present, reproduce, or discuss
all site data or all previous work products associated with these example sites. There was no evaluation
made of overall remedy performance or any portion of remedy performance beyond the hydraulic capture
being discussed. Statements regarding suitability or adequacy of the performance monitoring system and
sampling, achievement of performance goals, or any other aspects of the performance of the system are
expressly not made here. Inclusion of these examples does not imply any endorsement of the remedy or
of the performance monitoring of the remedy.
-------
EXAMPLE Bl
Example Capture Zone Evaluation, ECCA Site
SITE BACKGROUND INFORMATION
Location and Physical Setting
The East Canal Creek Area (ECCA) represents the eastern portion of the 700-acre Canal Creek Study
Area (CCSA), which is located within the Aberdeen Proving Grounds (APG). APG is a 72,000-acre
Army installation located in southeastern Baltimore County and southern Harford County, Maryland, on
the western shore of the upper Chesapeake Bay (see Figure Bl-1). The East Branch Canal Creek is
a small stream that flows southward in the western portion of the ECCA (see Figure Bl-2 for a more
detailed view of the ECCA). Kings Creek is located in the eastern portion of the ECCA. The land
surface is characterized as low-rolling terrain. The topographic elevation is near sea level in the vicinity
of Kings Creek, and is less than 30 feet above mean sea level (msl) throughout the ECCA.
Site Geology and Hvdrogeologv
The study area lies within the Atlantic Coastal Plain physiographic province. The regional geology
consists of unconsolidated sediments of sand, silt, clay, and gravel in a complex network of interbeds and
discontinuous lenses that thicken to the east. Crystalline basement rock occurs approximately 500 feet
below ground surface. The hydrostratigraphic units are as follows (see generalized cross section
presented in Figure Bl-3, and detailed west-to-east cross section presented in Figure Bl-4):
• Surficial Aquifer (discontinuous, up to 35 ft thick)
• Upper Confining Unit (10 to 5 0 ft thick)
• Canal Creek Aquifer (10 to 70 ft thick)
• Lower Confining Unit (35 to 65 ft thick)
• Lower Confined Aquifer
The Canal Creek Aquifer is the primary aquifer in the area, and as discussed later, it has been impacted by
site contaminants. In the general location of the East Branch Canal Creek, a paleochannel of Pleistocene
age eroded the Upper Confining Unit, and in that area the Surficial Aquifer is in direct contact with the
Canal Creek Aquifer (see Figure Bl-3). Based on Figure Bl-3, East Branch Canal Creek is a gaining
creek due to discharge of ground water from the surficial aquifer.
In the Surficial Aquifer, ground-water flow direction is generally away from topographic highs towards
the surface water bodies where ground-water discharges (i.e., within the study area illustrated in Figure
Bl-1, ground-water flow in the Surficial Aquifer is to the west in the vicinity of East Branch Canal Creek
and to the east in the vicinity of Kings Creek). In the deeper aquifers (Canal Creek Aquifer and the Lower
Confined Aquifer) ground-water flow direction in the ECCA is generally to the southeast, and does not
discharge to Kings Creek due to the presence of the Upper Confining Unit.
-------
Figure Bl-1. Location of Canal Creek Study Area
PENNSYLVANIA
Canal \Gunpowd
Creek
Peninsula
Scale in Feet
-------
u
t/3
cS
u
3
s
u
o
£
cS
K
^
£
u
u
M
K
a
11 ginnnn
MM
-------
o
•-P
u
-------
o
'•3
-------
Vertical hydraulic gradients are generally upward from the Lower Confined Aquifer to the Canal Creek
Aquifer. Vertical hydraulic gradients between the Surficial Aquifer and the Canal Creek Aquifer vary in
direction. In many areas hydraulic gradients are downward from the Surficial Aquifer to the Canal Creek
Aquifer. In the paleochannel area near the East Branch Canal Creek, however, hydraulic gradient is
generally upwards from the Canal Creek Aquifer to the Surficial Aquifer.
Transmissivity and hydraulic conductivity of the Canal Creek Aquifer in the vicinity of the extraction
wells were estimated from aquifer tests at each of the eight individual extraction wells (locations of
extraction wells are illustrated in Figure Bl-2). A representative value for transmissivity is approximately
2,200 ft2/d, and based on a typical thickness of 55 ft for the Canal Creek Aquifer in the
vicinity of the extraction wells, the average hydraulic conductivity is approximately 40 ft/d.
Transmissivity varies from this average value due to variations of hydraulic conductivity and aquifer
thickness. The aquifer testing yielded a transmissivity range of 965 ft2/d to 3,753 ft2/d.
Contaminants of Concern
Historically, the Canal Creek Area was a former manufacturing center of military-related chemicals and
agents. Previous ground-water investigations identified a large (approximately 5,000 ft long and 2,500 ft
wide) dissolved-phase chlorinated VOC plume in the Canal Creek Aquifer in the ECCA. Pre-remediation
concentrations for total VOCs are presented in Figure Bl-2. Chlorinated solvents
1,1,2,2-tetrachloroethane (1122-TeCA) and trichloroethene (TCE) were identified as the primary
contaminants. Elevated concentrations of VOC daughter products dichloroethene (DCE) and vinyl
chloride (VC) were also detected. The shape of the contaminant plume likely has been influenced by
historical water supply pumping from the Canal Creek Aquifer, which elongated the plume in an
east-west direction.
Contaminants are found throughout the vertical extent of the Canal Creek Aquifer, but concentrations are
lower near the bottom of the aquifer. For example, as presented in Figure B1-4, the 1122-TeCA
concentration at location 168 is 1,300 ug/1 in the middle of the aquifer, but only 9 ug/1 near the bottom of
the aquifer. Due to the upward hydraulic gradient from the aquifer below the Canal Creek Aquifer,
further downward migration is not expected. Existing monitoring wells in the deep aquifer confirm that
downward migration is not occurring.
Ground-Water Remedial System
The Record of Decision (ROD) was issued for the ECCA plume on July 17, 2000. It specified ground-
water extraction and treatment as the selected remedy for the main part of the plume, with institutional
controls and natural processes for the distal portion of the plume. Based on several site-specific
constraints, the primary objectives include the following: (1) maintain hydraulic capture of the 100 ug/L
composite VOC isocontour; and (2) provide mass removal for the VOC source area.
The extraction wells and treatment system were completed in 2002-2003. The ground-water treatment
plant (GWTP) is designed to handle a flow rate up to 305 gpm. The effluent is discharged to East Branch
Canal Creek. The system started operation on April 7, 2003.
The approved design utilized eight extraction wells with a total yield of 197 gpm. Extraction well
locations are illustrated in Figure Bl-2 (extraction well names are provided in Figures B1-6 to Bl-9,
which are presented later). The extraction wells are six-inches in diameter with depths ranging from 70 to
118 ft below ground surface (bgs). Well depths and screen intervals for two of the extraction wells (EW-2
and EW-6) are illustrated in Figure Bl-4. EW-2 has two screen intervals, one in the middle of the Canal
-------
Creek Aquifer and one near the bottom of the Canal Creek Aquifer. EW-6 has one screen interval, in the
lower portion of the Canal Creek Aquifer. In general, the extraction wells are screened in the lower
portion of the Canal Creek Aquifer. Each of the eight extraction wells are equipped with electric
submersible pumps that continuously pump ground water to the GWTP. The pumping rates, which vary
from 10 to 40 gpm at individual wells, are based on ground-water flow modeling that was conducted prior
to GWTP start-up. The design flow rate at each individual extraction well is listed in Table Bl-1.
Table Bl-1. Design Rate at Each Extraction Well
Extraction Well
EW-1
EW-2
EW-3
EW-4
EW-5
EW-6
EW-7
EW-8
Total
Design Flow Rate (gpm)
16
10
10
40
26
26
29
40
197
Extraction wells typically pump at the design flow rate, and when that occurs the total system extraction
rate is similar to the design system flow rate of 197 gpm. However, due to periodic downtime of
individual wells or the entire system, the average flow rate achieved over each quarter is generally less.
The average total flow rate for each quarter from mid-2003 to early-2005 is illustrated in Figure Bl-5.
Excluding first quarter of 2005, when considerable downtime was experienced, the long-term average
pumping rate actually achieved is approximately 150 gpm.
CAPTURE EVALUATION
Step 1 - Review Site Data. Site Conceptual Model, and Remedy Objectives
Initial aspects of the capture zone evaluation should determine if the following issues are adequately
addressed:
Is the plume adequately delineated in three dimensions?
• Is there adequate hydrogcologic information for performing capture zone evaluations?
* Is there an adequate site conceptual model?
* Is the objective of the remedy clearly stated?
The important conclusion is whether or not all of these issues are addressed to an extent mat allows the
remaining steps of the capture zone evaluation to be performed with an acceptable level of uncertainty.
-------
I
s
3
O-
05
o
o
H
u
M
•3
&c
to
C3
a
a
o
CO
a
o
rxj
a
o
a
CO
o
a
CO
o
CO
a
CO
C3
rxj
a
as
Q
(uidB)
MOJJ
-------
Is Plume Delineation Adequate?
It is important that the plume delineation be adequate so that a Target Capture Zone can be established in
Step 2. An interpretation of the pre-remedy plume for total VOCs in the Canal Creek Aquifer is
presented in Figure B1-2. The pre-remedy plume appears to have been well delineated with respect to the
100 ug/1 contour in the vicinity of the extraction wells, which is the hydraulic containment boundary for
this site. The exception may be at the extreme eastern portion of the plume, to the east of well
CCJ-104B. However, given that the ground-water flow direction is observed to be to the south and
southeast, the pre-remedy plume delineation appears adequate with respect to the 100 ug/L contour for
the purpose of defining a Target Capture Zone.
Figure Bl-6 presents the most recent available plume map that has been interpreted, corresponding to
approximately 10 months after system start-up. This includes a comparison of the interpreted 100 ug/L
contour for total VOCs before system start-up and after system start-up. Based on this comparison, the
interpreted 100 ug/L contour for total VOCs did not change substantially within 10 months of system
start-up. It should be noted that these contours are interpreted based on a limited number of data points,
and different interpretations are possible. However, for the purpose of conducting a capture zone
evaluation, the horizontal delineation of the total VOC plume to the 100 ug/L contour appears to still be
appropriate.
As discussed earlier, contaminants are found throughout the vertical extent of the Canal Creek Aquifer,
but concentrations are lower near the bottom of the aquifer. Due to the upward hydraulic gradient from
the aquifer below the Canal Creek Aquifer, further downward migration is not expected. Existing
monitoring wells in the deep aquifer confirm that downward migration is not occurring. Therefore,
vertical delineation is considered to be complete (i.e., the plume is assumed to extend to the base of the
Canal Creek Aquifer, but not below).
Is There Adequate Hydrogeologic Information?
The following brief summary is provided:
• The site has had extensive documentation of the geology and hydrostratigraphy.
• There have been numerous water level maps interpreted for the Canal Creek Aquifer (the
aquifer of interest) in the vicinity of the extraction wells, both with and without pumping.
Two water level maps, interpreted for two different time periods without pumping and
generally referred as "static conditions" by the site documents, are illustrated in Figure Bl-7
(March 2004, after six weeks with no pumping) and Figure B1-8 (April 2003, prior to system
startup). These figures provide information regarding the direction and magnitude of
the hydraulic gradient in the absence of remedy pumping. Note that the water level map
presented in Figure Bl-8 also highlights the interpretation of flow directions using arrows.
Based on other information, background flow conditions do not vary substantially by season.
Vertical hydraulic gradients have been evaluated between the Canal Creek Aquifer and the
overlying Surficial Aquifer and underlying Lower Confined Aquifer.
• Individual pump tests were conducted at each extraction well to estimate hydraulic
parameters.
-------
o
o
.0
'&
•3
o
U
1)
s
s
u
o
o
H
M
B
I
£
-------
o
o
§•
I
-
I
i.
tu
a
O
ea
2
I
S
-------
1 If
1
Tf
Q.
I
i/5
I
0>
>^
£
sS
I
I
i/5
00
TH
M
B
I
S
-------
• Water quality data have been collected prior to system start-up and after system start-up,
allowing for interpretation of spatial plume extent as well as concentration trends at
individual wells. The density of spatial measurements appears to be adequate to horizontally
delineate the 100 ug/L contour for the total VOC plume.
• Total pumping rate for the system is well documented over time. It is also documented that
wells operate at their design rates during operation, and those design rates have been
documented.
* Well construction data, including measuring point elevations and screen intervals, are
documented.
Although all of the details regarding each of the above items are not provided in this report, the amount of
available hydrogeologic information appears to be adequate.
Is There an Adequate Site Conceptual Model?
A site conceptual model (a text description, maps, and cross-sections that should not be confused with a
"numerical model") should adequately accomplish the following:
indicate the source(s) of contaminants
describe geologic and hydrogeologic conditions
• explain observed fate and transport of constituents
identify potential receptors
At this site, extensive work has been done to determine the sources of contaminants. For the ECCA, the
primary sources of contamination were sewer discharge points, located near wells CC-001 and CC-101.
Other sources of contamination exist to the north and northwest, and likely migrate towards the eight
extraction wells associated with the remedy. Although discharged wastes may have included DNAPLs,
no DNAPLs have been detected in ground water, and the dissolvcd-phasc concentrations of the individual
constituents in ground water are lower than those that typically exist when DNAPL is present.
Hydrogeologic conditions have been adequately defined, as discussed earlier. Contaminants in the Canal
Creek Aquifer, in the absence of pumping, flow to the south and southeast. The shape of the plume was
likely influenced by historical water supply pumping from the Canal Creek Aquifer. These water supply
wells trended on an east-west line across the ECCA, which likely caused contamination to spread more to
the east than would be expected based on the static (i.e., non-pumping) ground-water flow conditions.
The full thickness of the Canal Creek Aquifer is assumed to be impacted and targeted for horizontal
hydraulic containment. Vertical hydraulic gradients between the Lower Confined Aquifer and the Canal
Creek Aquifer are upward. Therefore, any dissolved contaminants present in the Canal Creek Aquifer are
unlikely to migrate downward to the Lower Confined Aquifer. Existing monitoring wells in the deep
aquifer confirm that downward migration is not occurring.
-------
There is observed degradation of the VOCs, based on the presence of daughter products of the primary
contaminants. It is also expected that the plume may be in a steady-state configuration towards the south
due to dilution from net recharge and dispersion. The ROD noted that there was no observed VOC plume
movement between investigations in the late 1980s and subsequent studies in the mid 1990s.
Currently, there is no potable use of ground water within the area impacted by the ECCA plume. An
ecological risk assessment indicated no unacceptable ecological risks.
In summary, there appears to be an adequate site conceptual model for performing a capture zone
evaluation.
Is Remedy Objective Clearly Stated with Respect to Plume Capture?
According to the ROD, "the goal of this remedy is to reduce the toxicity, mobility, and volume of
contaminated media in the East Canal Creek Area plume to meet Applicable or Relevant and Appropriate
Requirements (ARARs) in the plume by containing, capturing, and treating the contaminated
ground water in the main body of the plume and to eliminate exposure to the ground water through
implementation of institutional controls". The ROD also states that "implementation of the remedy
for the East Canal Creek Area plume would involve plume containment and capture and treatment to
reduce the toxicity, mobility, and volume of the contaminated media in the main body of the plume. The
downgradient portion of the plume will be evaluated and monitored to ensure that the natural processes
are protective and that downgradient contamination levels are being reduced as expected". The remedy
goal with respect to plume capture is to provide horizontal hydraulic containment of the 100 ug/L contour
for total VOCs in the Canal Creek Aquifer, for the entire thickness of the Canal Creek Aquifer.
Step 2 - Define Site-Specific Target Capture Zone(s)
The Target Capture Zone is the three-dimensional zone of ground water that must be captured by the
remedy extraction wells for the containment portion of the remedy to be considered successful. As
discussed above, the remedy goal with respect to plume capture is to provide horizontal hydraulic
containment of the 100 ug/L contour for total VOCs in the Canal Creek Aquifer, for the entire thickness
of the Canal Creek Aquifer. More specifically, the site documents generally refer to the Target Capture
Zone as the "static conditions" for the 100 ug/1 total VOC contour, based on summation of 2001-2002
concentrations for 1122-TeCA, TCE, cis-DCE, and VC (see Figures Bl-2 and Bl-7). The Target Capture
Zone extends the entire thickness of the Canal Creek aquifer. There is an awareness expressed in site
documents that the horizontal extent of a Target Capture Zone defined in this manner may change over
time, but interpretations to date indicate that the 100 ug/L contour for total VOCs has not changed
significantly since remediation pumping began, and the Target Capture Zone is still appropriate.
Step 3 - Interpret Water Levels
Potentiometric Surface Maps
Ground-water elevations have been routinely measured, and ground-water potentiometric maps have been
routinely constructed, after the system started operation. These water level maps indicate generally
consistent results. The water level map for October 20, 2003, approximately six months after system
operation began, is presented as Figure Bl-9.
-------
1
o
i
o
U
OH
I
I
1
1
I
OS
!— (
CQ
a
'• = = s
llllllll
Illlllll
iff
;••' -i :-• ;:
-------
Review of Figure B1 -9 indicates that there are not many water level measurement points available near
the extraction wells. Using water levels measured at extraction wells for constructing water level maps
can bias the interpretation of capture, since the water levels at the extraction wells may be much lower
than water levels in the aquifer material just outside the well bore. It can be equally problematic if no
water level measurement points are located in the vicinity of the extraction wells. To avoid these
problems, EPA recommends installing a water level measurement point near each extraction well.
However, if such measurement points are not available near the pumping well, a possible approach is to
estimate aquifer water levels at the extraction wells. The latter approach is utilized here. Figure B1-9
presents two water levels at each extraction well. One is a measured value, but that is not used for
contouring water levels. The other is an estimated value that is calculated by subtracting Theis-predicted
drawdowns from the static water levels (using superposition to account for drawdown due to pumping
from each of the extraction wells). The water level contours are based on the estimated value at each of
the extraction wells, rather than the measured value which is likely impacted by well inefficiencies and
well losses.
Other noteworthy features associated with the water level map presented in Figure Bl-9 include the
following:
measured water levels are posted
a Target Capture Zone is identified on the map (the 100 ug/1 contour)
interpreted flow directions are highlighted using arrows
pumping rates are identified
there is a scale and a north arrow
The interpretation of the water level map presented in Figure Bl-9, based on the arrows illustrated in the
figure, is that the capture zone extends beyond the Target Capture Zone. There is some level of
ambiguity regarding the interpretation, since much of it is based on the estimated values of water levels at
the extraction wells (which was necessary due to lack of water level measurement points near the
extraction wells). However, there are some water level measurement points near the extraction wells, and
quick inspection of these values (i.e., excluding the values at the extraction wells) suggests qualitatively
that the interpretation of the extent of capture is likely still valid. This can be evaluated more
quantitatively by looking at specific water level pairs.
Water Level Pairs (Gradient Control Points)
Pairs of water level elevations, located on either side of a real or conceptual boundary, can be used to
demonstrate inward flow relative to that boundary. For this demonstration site, this approach is
somewhat limited because of the relatively large distance between potential water level pairs. However,
calculation of water level differences between selected water level pairs, presented in Table Bl-2, still
provides for a useful line of evidence regarding capture.
Note some, but not all, of the water levels pairs presented in Table Bl-2 utilize estimated water levels at
extraction wells. More emphasis on the results for pairs that do not involve extraction well locations is
likely appropriate. In this particular case, all of the pairs yield a consistent interpretation of inward flow
relative to the Target Capture Zone boundary.
-------
Table Bl-2. Water Level Differences for Selected Water Level Pairs
Water Level
Pairs
Apr
2003
Jun
2003
Jul
2003
Aug
2003
Sep
2003
Oct
2003
Nov
2003
Dec
2003
Mar
2004
Apr
2004
Pairs that do not involve extraction wells:
003 to!04
009 to 167
009 to 166
106 to 005
106 to 004
0.87
0.22
0.51
NA
NA
1.25
0.40
1.18
NA
NA
1.21
0.44
1.19
NA
NA
1.37
0.57
1.57
NA
NA
1.23
1.23
1.23
NA
NA
1.19
0.41
0.41
1.49
3.39
1.09
0.81
1.19
1.57
3.68
1.23
0.44
1.18
1.56
3.63
NA
0.16
1.09
1.22
2.85
NA
0.30
1.18
1.24
3.37
Pairs that involve extraction wells (using estimated water levels at the extraction wells):
003 to EW-08
009 to EW-07
166 to EW-05
106toEW-04
3.84
1.61
0.63
NA
4.01
1.27
0.53
NA
3.91
0.71
3.23
NA
2.51
0.51
2.52
NA
3.35
1.46
3.84
NA
3.98
2.03
5.24
5.10
3.39
1.76
4.20
5.12
4.05
2.06
4.52
5.34
NA
5.01
7.59
7.50
NA
4.12
6.60
6.66
Note: Positive values suggest the inward flow relative to the Target Capture Zone.
Step 4 - Perform Calculations
Estimated Flow Rate Calculation
As discussed in the main document, the estimated flow rate calculation provides an estimate for the
pumping required to capture a plume, based on flow through the plume extent. This approach is
summarized in Figure 13 in the main document. Assumptions for this approach include the following:
• homogeneous, isotropic. confined aquifer of infinite extent
• uniform aquifer thickness
* fully penetrating extraction wcll(s)
* uniform regional horizontal hydraulic gradient
• steady-state flow
• negligible vertical gradient
* no net recharge, or net recharge is accounted for in regional hydraulic gradient
* other sources of water introduced to aquifer due to extraction are represented by the "factor"
-------
Assignment of specific values for these parameters is typically difficult, due to heterogeneities. For
instance, the hydrogeologic summary presented earlier indicates variation in aquifer thickness as well
as hydraulic conductivity determined with pump tests. Furthermore, based on the static water level map
presented in Figure Bl-7, the direction of background hydraulic gradient varies spatially, complicating
estimation of plume width. Therefore, the results from this line of evidence must be considered with
knowledge of these limitations. Nevertheless, it is useful to perform the calculation using best estimates
and/or ranges of values for specific parameters. For this demonstration, the following approach was
utilized:
transmissivity (hydraulic conductivity multiplied by thickness) was assigned as three
potential values:
Q 1,000 ft2/d (low estimate)
a 2,200 ft2/d (representative value)
Q 4,000 ft2/d (high estimate)
based on Figure Bl-7 (static water levels) a plume width of 4,000 ft was estimated based on
the width of the 100 ug/1 contour (that defines that Target Capture Zone) in the vicinity of the
southernmost extraction wells, in a direction perpendicular to the static ground-water flow
direction interpreted in that vicinity
• recognizing that hydraulic gradients are variable in space and time, the hydraulic gradient
was assigned as two potential values, 0.0007 ft/ft (representative value near extraction wells)
and 0.001 ft/ft (conservatively high value for vicinity of extraction wells), based on water
level contours for non-pumping conditions illustrated in Figures Bl-7 and B1-8
"factor" was assigned as three potential values (1.0, 1.5, and 2.0) to assess sensitivity of the
results to different degrees of potential capture of water from surface water and/or adjacent
aquifers
The flow rate calculation results, which estimate the amount of pumping that would be required to capture
a plume width of 4,000 ft based on the various combinations of parameter assignments, are presented in
Table B1-3.
These results are then compared to the actual pumping rate, which as discussed earlier, is typically 197
gpm when all wells are operating, with a long-term average of approximately 150 gpm. All of the
calculations of the pumping rate required to capture the plume are less than the 197 gpm the wells
typically operate at, indicating that a pumping rate of 197 gpm is likely more than enough for successful
capture. In fact, all but one of the calculated values is less than the long-term average of 150 gpm,
indicating capture is likely successful at the average long-term rate as well. The only exception is when
all the assigned parameters (transmissivity, hydraulic gradient, and the "factor") are assigned at the high
end of the range considered. By utilizing a range of values for the various input parameters, some of the
simplifications associated with this calculation are addressed. The consistent results for different ranges
of parameter values suggest that actual pumping at this site is likely sufficient for successful capture.
-------
Table Bl-3. Estimated Flow Rate Calculation*
Factor
1.0
1.5
2.0
Transmissivlty (ft2/
day)
1000
2200
4000
1000
2200
4000
1000
2200
4000
Hydraulic Gradient
(ft/ft)
0.0007
0.001
0.0007
0.001
0.0007
0.001
0.0007
0.001
0.0007
0.001
0.0007
0.001
0.0007
0.001
0.0007
0.001
0.0007
0.001
Estimated Flow Rate
(ftVday)
2,800
4,000
6,160
8,800
11,200
16,000
4.200
6,000
9,240
13,200
16,800
24,000
5,600
8,000
12,320
17,600
22,400
32,000
Estimated Flow Rate
(gpm)
14.54
20.78
32.00
45.71
58.18
83.12
21.82
31.17
48.00
68.57
87.27
124.68
29.09
41.56
64.00
91.43
116.36
166.23
'' based on estimated plume width of 4,000 ft
Capture Zone Width Calculation
As discussed in the main document, this line of evidence utilizes an analytical solution (illustrated in
Figure 14 in the main document), for a specific pumping rate, to determine if capture zone width is likely
sufficient. Assumptions for this approach include the following:
homogeneous, isotropic. confined aquifer of infinite extent
• uniform aquifer thickness
fully penetrating extraction well(s)
uniform regional horizontal hydraulic gradient
-------
• steady-state flow
• negligible vertical gradient
• no net recharge, or net recharge is accounted for in regional hydraulic gradient
• no other sources of water are introduced to aquifer due to extraction
Note that this calculation assumes no other sources of water are introduced to the aquifer due to induced
flow, such as from surface water or from an adjacent aquifer. This differs from the estimated flow rate
calculation, which accounts for other potential sources of water through the "factor" term.
When multiple extraction wells are present, this capture zone width calculation is typically applied by
assigning the total extraction rate to one "equivalent well". The location of the equivalent well is
generally selected visually so it is centrally located with respect to the plume width and/or extraction well
locations, and located at the most downgradient position of the actual extraction wells. This represents a
significant level of simplification for a multi-well extraction system.
For this site, the typical instantaneous pumping rate is 197 gpm, and the long term average pumping
rate (accounting for down time) is approximately 150 gpm. The conservative analysis of capture zone
width used here is based on 150 gpm, recognizing that much of the time a larger capture zone is present.
Calculations for Ywell ,Ymax, and XQ for different possible combinations of transmissivity and hydraulic
gradient values are presented in Table Bl-4.
Table Bl-4. Capture Width Calculation (150 gpm*)
Transmissivity
(ft2/day)
1000
2200
4000
Hydraulic
Gradient
(ft/ft)
0.0007
0.001
0.0007
0.001
0.0007
0.001
Distance
from Well to
Stagnation Point
X0 (ft)
6,565
4,596
2,984
2,089
1,641
1,149
*well
(ft)
10,313
7,219
4,688
3,281
2,578
1,805
Capture Zone
Width At Wells
(ft)
20,625
14,438
9,375
6,563
5,156
3,609
max
(ft)
20,625
14,438
9,375
6,563
5,156
3,609
Max Capture
Zone Width
Upgradient
(ft)
41,250
28,876
18,750
13,125
10,313
7,218
* consistent units are feet and days - pumping rate of 150 gpm is equal to 28,877 ft3/day
Figure Bl-10 illustrates the results. Note that results are illustrated based on the direction of background
hydraulic gradient. As discussed earlier, the background hydraulic gradient at this site varies spatially
(see Figure B1-7). The approach utilized herein was to orient the calculated capture zone based on the
approximate static ground-water flow direction in the vicinity of the southernmost extraction wells.
-------
e
o
1
_0
at
u
-^
•a
01
I
I
P5
u
L.
-------
It is apparent that the calculated capture zone widths are larger than the target capture width, which is
approximately 4,000 ft (as discussed earlier), except for the case where the transmissivity and hydraulic
gradient are both at the high end of the range. By utilizing a range of values for the various input
parameters, some of the simplifications associated with this calculation are addressed. Consistent results
that capture zone width is sufficient for many different ranges of parameter values suggest that actual
long-term average pumping rate at this site (150 gpm) is likely sufficient for successful capture. Also
note that calculated capture zone widths would be approximately 25% larger than presented in Table B1-4
if the typical instantaneous pumping rate of 197 gpm was used rather than the long-term average pumping
rate of 150 gpm.
Ground-Water Flow Model with Particle Tracking
The start-up of the GWTP and the extraction wells was implemented in a sequential fashion that consisted
of consecutive 2-day periods of well performance tests (step-testing and constant-rate 8-hour design flow
testing) at each well. A rigorous analysis of the hydraulic data was performed to correct for barometric,
tidal and competing pumping influences. The transmissivity estimates from the aquifer testing were
incorporated into a modified three-dimensional ground-water flow model, to predict the steady-state zone
of capture generated by the eight extraction wells. Manual water levels collected on July 28, 2003
(several months after pumping was initiated) were used as head targets during active pumping to
recalibrate the model versus observed drawdown. Simulation target summary statistics were compared
between the original and the final models to illustrate the model improvements.
Reverse particle tracking was performed, in conjunction with the revised model, to evaluate the simulated
capture zone for the typical instantaneous pumping rate of 197 gpm. The simulated capture zones
associated with 10 year time-of-travel and 20 year time-of-travel are summarized in Figure Bl-11. The
complete capture zone would be larger. This simulated capture zone encompasses the entire Target
Capture Zone, with the possible exception of a very small area at the extreme western edge. It should be
noted that a simulation performed with the long-term average pumping rate of 150 gpm would have a
smaller simulated capture zone than this illustration, which is based on 197 gpm.
It should also be noted that specific details about the model construction and particle tracking approach
were not provided in the summary report that was reviewed. It is likely that these details were provided
in other reports, but if not, an improved analysis would include simulations for the 150 gpm case and
more details regarding model construction and the particle tracking approach.
Step 5 - Evaluate Concentration Trends
Based on the water level map interpretation (Figure Bl-9) and capture width calculation (Figure B1-10) it
appears possible (or likely) that all of the monitoring wells located downgradient of the extraction wells
are within the capture zone of the extraction wells. Since there are continuing sources of ground-water
impacts, monitoring wells that are impacted and are located within the capture zone would be expected to
remain impacted. Therefore, this line of evidence would provide ambiguous interpretations (i.e., these
wells might not clean up over time whether or not capture is sufficient), and therefore this line of evidence
is not utilized for this capture zone evaluation.
-------
u
2
H
—
™
o
•a
s.
sS
U
s
M
-------
Step 6 - Interpret Actual Capture Based on Steps 1-5. Compare to Target Capture Zone(s). Assess
Uncertainties and Data Gaps
Based on evaluations of multiple lines of evidence discussed in Step 3 to Step 5, the actual capture
achieved by the extraction wells is interpreted in Step 6, and the following items are addressed:
• compare the interpreted capture zone to the Target Capture Zone
assess uncertainties in the interpretation of the actual capture zone
• assess the need for additional characterization and/or monitoring
evaluate the need to reduce or increase extraction rates
Table Bl-5 presents the summary of the capture zone evaluation for this site.
Table Bl-5. Summary of Capture Zone Evaluation
Step
Step 1: Review site data, site
conceptual model, remedy objectives
Step 2: Define "Target Capture
Zone(s)"
Step 3a: Water level maps
Step 3b: Water level pairs
Step 4a: Simple horizontal capture
zone analyses
Step 4b: Ground-water flow modeling
with particle tracking
Step 5: Concentration trends
Step 6: Interpret actual capture and
compare to Target Capture Zone
Summary/Conclusions
Completed, all determined to be up-to-date and adequate.
Clearly defined, illustrated on maps. Pertains to entire thickness of
Canal Creek Aquifer.
Interpreted capture zone is larger than the Target Capture Zone.
Estimated water levels at extraction wells are utilized when
constructing potentiometric surface maps due to lack of water level
measurement points near the extraction wells. This is an improvement
over using water levels measured at the extraction wells, but actual
water level measurements near the extraction wells would be preferred.
Inward flow at all pairs along the Target Capture Zone boundary.
Estimated flow rate calculation indicates the long-term average
pumping rate of 150 gpm is likely sufficient.
Capture zone width calculation indicates the long-term average
pumping rate of 150 gpm likely provides for sufficient capture zone
width.
Model calibration was updated after system operation based on
observed system performance. Particle tracking results indicate
successful capture for typical instantaneous pumping rate of 197 gpm,
but results for the long-term average pumping rate of 150 gpm were
not simulated. An improved analysis would include simulations for
the 150 gpm case and more details regarding model construction and
the particle tracking approach.
Not relied upon for short-term evaluation of capture.
The actual capture zone is interpreted to be sufficient relative to the
Target Capture Zone. Particle tracking was not performed for
long-term average pumping rate of 150 gpm, but all other lines of
evidence suggest that capture is sufficient. Adding a water level
measurement point near each extraction well would improve the
analysis.
-------
As discussed in Exhibit 8 of the main document, a summary of the following items is appropriate:
* Is capture sufficient, based on "converging lines of evidence "?
The capture zone analysis indicates that capture is sufficient, and the zone of capture is larger
than the Target Capture Zone, based on water level maps, gradient pairs, simple calculations
for estimated flow rate and capture zone width, and particle tracking results based on numerical
modeling. This provides a safety factor that accounts for uncertainties.
* Key uncertainties/data gaps
There is some uncertainty in the analysis of water levels due to the use of estimated water levels at
the extraction wells. Also, an improved analysis would include particle tracking simulations for the
150 gpm case and more details regarding model construction and the particle tracking approach.
However, because multiple lines of evidence regarding capture are available (as mentioned above),
none of these issues likely impacts the conclusion that capture is sufficient.
* Recommendations to collect additional data, install new monitoring wells, change current
extraction rates, change number/location of extraction wells, etc.
It is possible that adequate capture could be achieved with a lower total pumping rate. Further
evaluation to attempt to optimize pumping rates could potentially be considered, if it is determined
that a lower total pumping would significantly lower the cost of the remedy while providing an
adequate level of protection. Also, as stated earlier, adding water level measurement points near
each extraction well would improve the analysis.
-------
This page left intentionally blank.
-------
EXAMPLE B2
Example Capture Zone Evaluation, Milan OU4 Site
SITE BACKGROUND INFORMATION
Location and Physical Setting
The Milan Army Ammunition Plant (MAAP) is located in the western portion of Tennessee (see Figure
B2-1). The vicinity of the pre-remedy plume associated with Operable Unit 4 Region 1 (OU4) is
illustrated in Figure B2-2. Highway 104 is located just outside the plant boundary and Highway 77 is
located approximately 3,000 ft northwest of the plant boundary. There are no significant surface water
bodies in the immediate vicinity of the OU4 plume, although there are several small surface ditches or
creeks. Topography in the vicinity of the OU4 plume is relatively flat and slopes gently to the west.
Site Geology and Hydrogeology
The study area lies within the Gulf Coastal Plain physiographic province, on the eastern flank of the
Upper Mississippi River Embayment. The regional geology consists of sediments that include sand,
gravel, lignite, clay, chalk, and limestone. A generalized cross-section of the plant vicinity is presented
in Figure B2-3. The upper unit is the Memphis Sand, which is several hundred feet thick and consists of
sand with some layers of silt and clay. The Flour Island Formation serves as a confining clay layer below
the Memphis Sand. The plume associated with OU4 is located within the Memphis Sand.
In the aquifer of concern, ground-water flow is generally to the northwest. Vertical hydraulic gradients
are generally downward, and contamination is found to depths of more than 200 feet as a result
(contaminant distribution with depth is discussed in more detail later).
Hydraulic conductivity of the Memphis Sand in the vicinity of OU4 is within the range of 70 to 110
ft/day, based on a combination of aquifer test results and flow model calibration (both regional and local
flow models). Given an approximate aquifer thickness of 270 ft, the transmissivity of the Memphis Sand
is approximately 20,000 to 30,000 ft2/day.
Contaminants of Concern and Contaminant Distribution
The primary contaminants of concern for the OU4 plume are explosives. Water quality data are typically
presented for total explosives. The explosive with the highest concentrations is RDX.
Discrete depth sampling in conjunction with rotosonic drilling was performed to vertically delineate the
plume. Table B2-1 presents depth discrete sampling results for location MI-533 (between Route 77
and Route 104) and for location MI-527 (near Route 77). These data illustrate that the contaminants
are located deeper within the Memphis Sand towards the northwest due to downward vertical hydraulic
gradients. The interpreted plume depth, based on similar data from other locations, is illustrated in Figure
B2-4.
-------
Figure B2-1. Location of Milan Army Ammunition Depot
Milan Army
Ammunition Plant
Missouri . —-i-
Kentucky
Virginia
Arkansas ?# /^
'±1/
* Memphi s
Tennessee
\£hattanooga
North Caroli na
AH abama
Kentucky
Missouri
Arkansas
Milan Army
Ammunition Plant
-------
Figure B2-2. Vicinity of Milan OTJ4 and Extent of Explosive Plume
Legend
j
/ Plume Extent • Extraction Well
Monitoring Well
1125
SCALE IN FEET
2250
Note: "Plume Extent" based on site-specific concentration limit
-------
dnojg aujoqieio dnojg XOOHM
(id) UOUBA9I3
888
a
o
U
OS
&B
1
-M
05
J
CQ
a
(id) UO|1BA9|3
-------
ELEVATION IN FEET - MSL
a
03
I
o
u
I
es
u
-f
f
-------
Table B2-1. Discrete Depth Sampling for Plume Delineation
Depth
(ft)
Elevation
(ft MSL)
Lab I Results*
(ppb)
Lab 2 Results*
(ppb)
Location MI -5 33 (Between Ilwy 77 and Ilwy 104):
165-175
185-195
205-215
225-235
250-260
230-240
210-220
190-200
200
960
1725
ND
135
623
971
ND
Location MI-527 (Near Hwy 77):
135 -145
165-175
195-205
225-235
255-265
252-262
222-232
192-202
162-172
132-142
ND
ND
81
242
ND
ND
ND
73
212
ND
*Results represent concentrations of total explosives. Lab 1 was a contract lab,
Lab 2 was an on-silc lab.
Ground-Water Remedial System
A ground-water extraction and treatment system to address the plume of explosives was designed and
built in 2001, and began operation in June, 2002. Operation has been more or less continuous since that
time.
The extraction system consists of two lines of extraction wells (see Figure B2-2). A line of four
extraction wells, referred to as XP-1 through XP-4 and located along Highway 104, is intended to provide
hydraulic containment of the on-site part of the plume and prevent further off-site migration. A second
line of four extraction wells located closer to Highway 77, referred to as XP-5 through XP-8, is intended
to provide hydraulic containment of the off-site part of the plume that has been characterized.
The actual flow rate of each extraction well is presented in Table B2-2. Actual extraction rates total 1,135
gpm, which exceeds the original design flow rate of approximately 700 to 900 gpm. Higher flow rates
were implemented to add conservatism regarding capture, given that the extraction wells could produce at
least that much water. When the system was constructed, it was recognized that additional extraction
wells might be required further to the northwest to contain the remainder of the plume, but the extent of
contamination in this area was not fully characterized at the time.
-------
Table B2-2. Actual Flow Rate at Each Extraction Well
Extraction Well
XP-1
XP-2
XP-3
XP-4
XP-5
XP-6
XP-7
XP-8
Total
Actual Flow Rate (gpm)
100
190
190
100
100
190
190
75
1,135
CAPTURE EVALUATION
Step 1 - Review Site Data. Site Conceptual Model, and Remedy Objectives
Initial aspects of the capture zone evaluation should determine if the following issues arc adequately
addressed:
Is the plume adequately delineated in three dimensions?
• Is there adequate hydrogeologic information for performing capture zone evaluations?
Is there an adequate site conceptual model?
• Is the objective of the remedy clearly stated?
The important conclusion is whether or not all of these issues are addressed to an extent that allows the
remaining steps of the capture zone evaluation to be performed with an acceptable level of uncertainty.
Is Plume Delineation Adequate?
It is important that the plume delineation be adequate so that a Target Capture Zone can be established in
Step 2. An interpretation of the pre-remedy plume extent is presented in Figure B2-2. along with the
monitoring network. The actual concentration data from which the plume extent was determined is not
shown here. There arc some monitoring locations outside the interpreted plume map that provide a basis
for delineating the plume width, which is interpreted to be approximately 1,800 feet wide. These
monitoring wells were supplemented by numerous exploratory borings during the remedial investigation.
Also, depth discrete sampling was performed to vertically delineate the plume. Review of site records
indicates that the pre-remedy plume delineation is adequate for the purpose of defining a Target Capture
Zone.
-------
Is There Adequate Hydrogeologic Information?
The following brief summary is provided:
The site has had extensive documentation of the geology and hydrostratigraphy.
• A series of monitoring wells was installed relatively close to the extraction wells, which
allows for detailed characterization of water levels and water quality in the aquifer at
locations near the extraction w7ells.
For this capture zone evaluation, which was performed approximately one year after system
startup, additional water levels were collected as part of a planned short-term system
shutdown, as follows:
D first, water levels were collected during operation of the system at a total extraction
rate of 1,135 gpm
D then, water levels were collected after the system had been operating for
approximately 11 hours at a reduced total extraction rate of 775 gpm (to evaluate the
extent of capture for a potentially reduced total pumping rate)
n then, water levels were collected after the system had been shut down for a period of
72 hours
* Vertical gradients have been evaluated and are downward, consistent with the plume reaching
depths of more than 200 feet.
Hydraulic parameters were estimated from aquifer testing and ground-water flow model
calibration at both the regional scale and local scale.
• Numerous ground-water monitoring wells and exploratory borings with depth discrete
sampling were used to define the horizontal and vertical extent of the explosives plume
during characterization investigations.
* A baseline water quality sampling event for explosives concentrations in monitoring wells
was conducted prior to startup of the system. Data were then collected again in May and
June of 2003, approximately one year after extraction was initiated.
• Total pumping rate for the system is well documented over time, and exceeds the design
values.
• Well construction data, including measuring point elevations and screen intervals, are well
documented.
Although details regarding each of the above items are not provided in this report, review of site records
indicates that the amount of available hydrogeologic data is adequate.
-------
Is There an Adequate Site Conceptual Model?
A site conceptual model (a text description, maps, and cross-sections that should not be confused with a
"numerical model") should adequately accomplish the following:
• indicate the source (s) of contaminants
describe geologic and hydrogeologic conditions
• explain observed fate and transport of constituents
identify potential receptors
At this site, the source of contaminants is from a manufacturing facility located south of Highway 104,
where wastewater from operations was placed into surface ditches. Hydrogeologic conditions and
contaminant movement (horizontally and vertically) have been well documented. Contaminant transport
patterns are consistent with ground-water flow patterns. Institutional controls have been implemented to
prevent impacts to potential receptors. Although all of the details are not provided herein, review of site
records indicates that there is an adequate site conceptual model for performing a capture zone evaluation.
Is Remedy Objective Clearly Stated with Respect to Plume Capture?
A line of four extraction wells (XP-1 through XP-4) located along Highway 104 is intended to provide
hydraulic containment of the on-site part of the plume and prevent further off-site migration. A second
line of four extraction wells (XP-5 through XP-8) is intended to provide hydraulic containment of the off-
site part of the plume that has been characterized. The objective with respect to depth is to capture water
from the impacted depths, and not necessarily the entire thickness of the Memphis Sand. The extraction
wells were screened to include the most impacted portions of the aquifer with respect to depth (based on
the sampling with depth discussed earlier), and were screened across approximately 75% of the impacted
aquifer zone near the extraction wells (with respect to depth). Based on calculations and theory regarding
partially penetrating wells (not presented herein), given the sandy nature of the aquifer, the vertical extent
of capture would extend below the impacted portions of the aquifer.
Step 2 - Define Site-Specific Target Capture Zone(s)
The Target Capture Zone is the three-dimensional zone of ground water that must be captured by the
remedy extraction wells for the hydraulic containment portion of the remedy to be considered successful.
Based on the remedy goal with respect to plume capture described above, the Target Capture Zone is
stated as follows: "Provide horizontal hydraulic containment of ground water at each of the two lines of
extraction wells across the full width of the total explosives plume that is indicated in Figure B2-2". The
Target Capture Zone does not indicate any specific distance down-gradient from each line of extraction
wells that the capture zone needs to include, so long as a capture zone of appropriate width is achieved by
each line of extraction wells.
With respect to depth, there is no explicit Target Capture Zone that pertains to vertical hydraulic capture
(i.e., no specific depth where upward flow is required). However, as discussed earlier, design of the
extraction system took into account the increasing depth of the plume towards the northwest, such that
horizontal capture would be achieved for the depth intervals where the aquifer is impacted. Long-term
ground-water monitoring is being conducted to verify that further plume migration with depth does not
occur.
-------
Step 3 - Interpret Water Levels
Potentiometric Surface Maps
For this capture zone evaluation, which was performed approximately one year after system startup, water
levels were collected as part of a planned short-term system shutdown, as follows:
• first, during operation of the system at a total extraction rate of 1,135 gpm
• then, after the system had been operating for approximately 72 hours at a reduced total
extraction rate of 775 gpm
then, after the system had been shut down for a period of 72 hours
For this site, water levels collected for a specific range of aquifer depths are utilized to represent the flow
patterns for the overall aquifer. The site-specific details of this approach are not discussed herein. As
noted earlier, a series of monitoring wells was installed relatively close to the extraction wells, which
allows for detailed characterization of water levels in the aquifer at locations near the extraction wells,
which improves the ability to interpret water levels.
Figure B2-5 is a contour map of measured water levels that are based on the data for an extraction rate of
1,135 gpm. This map was constructed using a kriging algorithm. For this presentation, actual water level
values are not posted on the figure, and it is noted that a more complete presentation would include posted
water level measurements. Also shown in Figure B2-5 are vectors that depict the magnitude and
horizontal direction of the hydraulic gradient based on the water levels. These vectors were produced
by the software package that was used to develop the contours. The northern line of extraction wells
appear to be capturing more water than is necessary based on the flow vectors outside the plume that form
a trajectory toward the extraction wells. There appears to be a small area on the far eastern side of the
plume, at the southern line of extraction wells, where there may be a lack of capture. However, there is
some uncertainty regarding the quality of the measured water level at one location in that vicinity (i.e, east
of extraction well XP-1), and it is possible that this value is causing an erroneous interpretation regarding
water level contours. It should also be noted that the area potentially not captured at the southern line of
extraction wells is within the interpreted zone of capture for the northern line of extraction wells, and thus
within the capture zone of the overall system.
Figure B2-6 is a similarly-constructed contour map for water level measurements that were made when
the system was operating at 775 gpm. The results are generally similar compared to the results for 1,135
gpm, except that there appears to be less clean water captured by the northern line of extraction wells.
Once again, there appears to be a small area on the far eastern side of the plume, at the southern line of
extraction wells, where there may be a lack of capture, and this interpretation may be due to a
questionable water level measurement at one location east of extraction well XP-1.
Figure B2-7 is a similarly-constructed contour map for water level measurements that were collected
with the extraction system not operating. Comparing these water levels (without pumping) to the
previous two figures (with pumping) clearly illustrates the impact of pumping. Figure B2-7 also
highlights that some monitoring wells may have erroneous measurements or datums, because the flow
vectors indicate complications in the ground-water flow patterns that would not be expected in the
absence of pumping. Datums for these wells should be reviewed or re-surveyed; however it should be
noted that the magnitude of possible error is on the order of inches. Some of these errors could also be
-------
Figure B2-5. Interpreted Water Level Map, Current Pumping Rate of 1,135 gpm
-N-
Legend
/
/ Plume Extent • Extraction Well 9 Monitoring Well
/\ Vector Gradient ^sso^ Interpreted Water Level Contour
650
SCALE IN FEET
1300
Note: contours and vectors are interpreted from measured water levels
-------
Figure B2-6. Interpreted Water Level Map, Reduced Pumping Rate of 775 gpm
Interpreted Capture Zone
sOKVfs^
' x ^ \ \ \
\KA\\V
V
\ \ F^v K fe<*^ \
xr N ^.' • rt»\ s
\\
^\ -
\ i-.
Legend
/ Plume Extent • Extraction Well $ Monitoring Well
"/I Vector Gradient ^3«o— Interpreted Water Level Contour
650
SCALE IN FEET
1300
Note: contours and vectors are interpreted from measured water levels
-------
Figure B2-7. Interpreted Water Level Map, with No Pumping
Legenrf
x Plume Extent
"7 Vector Gradient
Extraction Well -^ Monitoring Well
so _^ Interpreted Water Level Contour
650
SCALE IN FEET
1300
Note: contours and vectors are interpreted from measured water levels
-------
due to 1) heterogeneity, 2) vertical head differences in well clusters, and 3) differing time frame for water
levels to reach equilibrium after pumping was terminated.
Noteworthy features associated with the water level maps presented in Figures B2-5 to B2-7 include the
following:
measured water levels are not posted. Posting the water levels would allow the reader to
better evaluate whether or not the interpreted water level contours are reasonable
the width of the total explosives plume that is the basis for the width of the Target Capture
Zone is identified on the map
• although total pumping rate is identified on each figure, pumping rates at the individual
extraction rates are not identified, and adding those would improve the presentation
• there is a scale and a north arrow
It is also noted that the use of vectors that are created by the contouring software makes flow directions
much easier to interpret.
Water Level Pairs (Gradient Control Points)
Pairs of water level elevations, located on either side of a real or conceptual boundary, can be used to
demonstrate inward flow relative to that boundary. For this demonstration site, a more sophisticated
approach using triangles was utilized. This approach utilized data from monitoring wells which were
installed near the extraction wells for this purpose. This method mathematically determines a hydraulic
gradient and flow direction from three water levels that form vertices of a triangle. Assumptions with this
method include a homogeneous aquifer between wells, a linear change in head between wells, and that
vertical head differences are small within the vertical interval from which water levels are used. A total
of 17 "triangles" formed by wells that appear to satisfy these assumptions were used to evaluate ground-
water flow directions for different rates of extraction.
The interpreted results for each of the three pumping scenarios are as follows:
• Figure B2-8 illustrates the flow directions derived from the two lines of extraction wells for a
total extraction rate of 1,135 gpm. According to Figure B2-8, ground-water flow is generally
towards the extraction wells, and flow divides downgradient of the extraction wells are indicated.
Figure B2-9 illustrates that the magnitude of velocity is less when the pumping rate is
lowered to 775 gpm, as depicted by the smaller arrows in Figure B2-9 compared to Figure
B2-8, but flow directions do not change appreciably, again indicating the creation of flow
divides downgradient of the extraction wells.
Figure B2-10 presents the results for the case with no pumping. As expected, the flow
directions change significantly when the extraction wells are turned off, with flow returning
to background conditions (flow toward the northwest).
These evaluations of water levels pairs, while only based on a few measurement points, can be used to
augment the conclusions from other lines of evidence regarding capture.
-------
Figure B2-8. Interpreted Water Level "Triangles", Current Pumping Rate of 1,135 gpm
350 700
/ Plume Extent • Extraction Well $ Monitoring Well
* Flow Direction
SCALE IN FEET
-------
Figure B2-9. Interpreted Water Level "Triangles", Reduced Pumping Rate of 775 gpm
350 700
/ Plume Extent • Extraction Well $ Monitoring Well
Flow Direction
SCALE IN FEET
-------
Figure B2-10. Interpreted Water Level "Triangles", with No Pumping
SCALE IN FEET
-------
Step 4 - Perform Calculations
Estimated Flow Rate Calculation
As discussed in the main document, the estimated flow rate calculation provides an estimate for the
pumping required to capture a plume, based on flow through the plume extent. This approach is
summarized in Figure 13 in the main document. Assumptions for this approach include the following:
• homogeneous, isotropic, confined aquifer of infinite extent
• uniform aquifer thickness
* fully penetrating extraction wcll(s)
* uniform regional horizontal hydraulic gradient
• steady-state flow
• negligible vertical gradient
* no net recharge, or net recharge is accounted for in regional hydraulic gradient
* other sources of water introduced to aquifer due to extraction are represented by the "factor"
Assignment of specific values for these parameters is typically difficult, due to heterogeneities. For
instance, the hydrogeologic summary presented earlier indicates variation in hydraulic conductivity.
Therefore, the results from this line of evidence must be considered with knowledge of these limitations.
Nevertheless, it is useful to perform the calculation using best estimates and/or ranges of values for
specific parameters. For this demonstration, the following approach was utilized:
* hydraulic conductivity was assigned as two potential values:
D 70 ft/d (low estimate)
o 110 ft/d (high estimate)
aquifer thickness was assigned as 270 ft, therefore transmissivity ranges from:
o 18,900 ft2/d (low estimate)
D 29,700 ft2/d (high estimate)
* based on Figure B2-7 (static water levels) a hydraulic gradient of approximately 0.0012 was
assigned
plume width of 1,800 ft was estimated based on the width of the explosives plume (that
defines that Target Capture Zone)
"factor" was assigned as three potential values (1.0, 1.5, and 2.0) to assess sensitivity of the
results to different degrees of potential capture of water from surface water and/or adjacent
aquifers.
-------
The flow rate calculation results, which estimate the amount of pumping that would be required to capture
a plume width of 1,800 ft based on the various combinations of parameter assignments, are presented in
Table B2-3.
Table B2-3. Estimated Flow Rate Calculation*
Factor
1.0
1.5
2.0
Transmissivity
(ft2/day)
18,900
29,700
18,900
29,700
18,900
29,700
Hydraulic
Gradient (ft/ft)
0.0012
0.0012
0.0012
0.0012
0.0012
0.0012
Estimated Flow
Rate (ft3/day)
40,824
64,152
61,236
96,228
81,648
128,304
Estimated Flow
Rate (gpm)
212
333
318
500
424
666
' based on estimated plume width of 1,800 ft
These results are then compared to the actual pumping rate. In this case, the actual pumping rate for each
line of extraction wells should be considered. For the southern line of extraction wells (XP-1 to XP-4) the
actual extraction rate is 580 gpm, and for the northern line of extraction wells (XP-5 to XP-8) the actual
extraction rate is 555 gpm. Based on this simple calculation, it would appear that the total rate at each
line of extraction is likely sufficient for successful capture, since the estimated flow rate required for
capture is generally less than these actual pumping rates. The only exception is for the high value of
transmissivity coupled with the high value of "factor". By utilizing a range of values for the various input
parameters, some of the simplifications associated with this calculation are addressed. The consistent
results for different ranges of parameter values adds confidence in the conclusion that the actual pumping
rate is likely sufficient.
For the reduced pumping rate of 775 gpm being evaluated as an option, the pumping rate for each line of
extraction wells should again be considered. For the southern line of extraction wells (XP-1 to XP-4) the
extraction rate during the "shut-down" test was 450 gpm, and for the northern line of extraction wells
(XP-5 to XP-8) the extraction rate during the "shut-down" test was 325 gpm. The results for estimated
flow rates calculated in Table B2-3 indicate that pumping at each line of extraction may or may not be
sufficient, because for several combinations of parameter values in Table B2-3 the estimated flow rate
required for capture exceeds these actual pumping rates. It should be noted that the hydraulic gradient
used at the northern line of the extraction wells (XP-5 to XP-8) may be an overestimate, because
pumping at the southern line of extraction wells would likely flatten the background hydraulic gradient
at the northern line of extraction wells. Thus, the simple analysis performed here may overestimate the
pumping required at the northern line of extraction wells.
Capture Zone Width Calculation
As discussed in the main document, this line of evidence utilizes an analytical solution (illustrated in
Figure 14 of the main document), for a specific pumping rate, to determine if capture zone width is likely
sufficient. Assumptions for this approach include the following:
-------
homogeneous, isotropic, confined aquifer of infinite extent
• uniform aquifer thickness
fully penetrating extraction well(s)
uniform regional horizontal hydraulic gradient
steady-state flow
negligible vertical gradient
no net recharge, or net recharge is accounted for in regional hydraulic gradient
no other sources of water are introduced to aquifer due to extraction
Note that this calculation assumes no other sources of water are introduced to the aquifer due to induced
flow, such as from surface water or from an adjacent aquifer. This differs from the estimated flow rate
calculation, which accounts for other potential sources of water through the "factor" term.
When multiple extraction wells are present, this capture zone width calculation is typically applied
by assigning the total extraction rate to one "equivalent well". The location of the equivalent well is
generally selected visually so it is centrally located with respect to the plume width and/or extraction well
locations, and located at the most downgradient position of the actual extraction wells. This represents a
significant level of simplification for a multi-well extraction system. For this site, a further complication
is that there are two lines of extraction. For this analysis, one "equivalent well" is utilized for each line of
extraction, and the capture zone width calculation is performed independently for each line of extraction
(ignoring potential interference between the two lines of extraction).
Calculations for Ywell ,Ymax, and XQ for different possible combinations of hydraulic gradient and
transmissivity values are presented in Table B2-4, for the current pumping rate of 1,135 gpm, with the
southern line of extraction wells (XP-1 to XP-4) at 580 gpm and the northern line of extraction wells
(XP-5 to XP-8) at 555 gpm. Figure B2-11 illustrates these results for the current pumping rates (1,135
gpm). Note that results are illustrated based on the direction of background hydraulic gradient. The
results indicate that this level of pumping provides sufficient capture at both lines of extraction, relative to
the plume width of 1,800 feet that defines the Target Capture Zone.
A lower pumping rate of 775 gpm was also evaluated. Calculations for Ywell ,Ymax, and X0 for different
possible combinations of hydraulic gradient and transmissivity values are presented in Table B2-5, for the
lower pumping rate of 775 gpm, with the southern line of extraction (XP-1 to XP-4) at 450 gpm and the
northern line of extraction (XP-5 to XP-8) at 325 gpm, based on rates during the "shut-down" test. The
results indicate that this level of pumping may not provide sufficient capture at the northern line of
extraction wells (XP-5 to XP-8), for the higher value of transmissivity (i.e., maximum capture zone width
of 1,755 ft upgradient of the extraction wells, versus plume width of approximately 1,800 feet). Figure
B2-12 illustrates the results for the reduced pumping rates. For the low value of transmissivity, capture
is sufficient at both lines of extraction wells. However, for the high value of transmissivity, capture is not
quite sufficient across the full plume extent at the northern line of extraction wells. This suggests capture
is likely sufficient for this pumping scenario, but with less certainty than at the current (higher) pumping
rates.
-------
Table B2-4. Capture Zone Width Calculation (Current Pumping)
XP-1 to XP-4 (580 gpm*)
Hydraulic
Gradient (ft/ft)
0.0012
TVansmissivity
(ft2M)
18,900
29,700
Distance
from Well to
Stagnation Point
" X0 (ft)
783
499
well
(ft)
1,231
783
Capture Zone
Width at Wells
(ft)
2,461
1,566
max
(ft)
2,461
1,566
Max Capture
Zone Width
Upgradient
"(ft)
4,923
3,133
"consistent units arc foci and days - pumping rale of 580 gpm is equal to 111, 658 ft /day
XP-5toXP-8(555gpm*)
Hydraulic
Gradient (ft/ft)
0.0012
TVansmissivity
(ft2/d)
18,900
29.700
Distance
from Well to
Stagnation Point
" x0 (ft)
750
477
well
(ft)
1,178
749
Capture Zone
Width at Wells
(ft)
2,355
1.499
max
(ft)
2,355
1.499
Max Capture
Zone Width
Upgradient
"(ft)
4,711
3,998
^consistent units are feet and days - pumping rate of 555 gpm is equal to 106,845 ff/Asy
Table B2-5. Capture Zone Width Calculation (Reduced Pumping)
XP-1 to XP-4 (450 gpm*)
Hydraulic
Gradient (ft/ft)
0.0012
TransmlssMty
(ft2/d)
18,900
29,700
Distance
from Well to
Stagnation Point
X0 (ft)
608
387
well
(ft)
955
608
Capture Zone
Width at Wells
(ft)
1,910
1,215
max
(ft)
1,910
1,215
Max Capture
Zone Width
Upgradient
(ft)
3,819
2,431
Consistent units are feet and days -pumping rate of 450gpm is equal to 86,631 ft1/day
XP-5toXP=8(325gpm*)
Hydraulic
Gradient (ft/ft)
0.0012
Transmissivity
(ft2/d)
18,900
29,700
Distance
from Well to
Stagnation Point
X0 (ft)
439
279
"well
(ft)
690
439
Capture Zone
Width at Wells
(ft)
1,379
878
max
(ft)
1,379
878
Max Capture
Zone Width
Upgradient
(ft)
2,758
1,755
Consistent units are feel and days - pumping rate of 325gpm is equal to 62,567 fP/Aay
-------
Figure B2-11. Results for Capture Zone Width Calculation, Current Pumping Rate of 1,135 gpm
Legend
/' Plume Extent % Extraction Well $ Monitoring Well
SCALE IN FEET
-------
Figure B2-12. Results for Capture Zone Width Calculation, Reduced Pumping Rate of 775 gpm
-N-
Leqend
f
/ Plume Extent 0 Extraction Well ^ Monitoring Well
1125 2250
SCALE IN FEET
-------
Ground-Water Flow Model with Particle Tracking
After the start-up of the P&T system, the ground-water flow model used to design the extraction system
was run to determine how well it predicted the drawdown response to known stresses. The two sets of
extraction rates (1,135 and 775 gpm) were input to the model to compute water levels and drawdown
(pumping versus no pumping). The model-computed output was then compared to the observed
drawdown data. It was determined that the model had a tendency to over-predict drawdown (see Figure
B2-13, part "a"), and the model was re-calibrated such that it predicted drawdown more accurately (see
Figure B2-13, part "b").
Capture zones derived by the model for each well, for the current extraction rate of 1,135 gpm, are shown
in Figure B2-14. This figure was generated by color-coding each cell in the model by the final
destination of a particle originating in that cell, as determined from the particle tracking. Although full
details are not provided herein, this particle tracking analysis did consider the three-dimensionality of the
problem. The initial particles were placed at different depths (i.e., model layers) where the aquifer is
impacted. Figure B2-14 illustrates the results for one such depth interval. These results indicate that
extraction of 1,135 gpm sufficiently captures the plume, plus a significant amount of clean water. Similar
results were achieved for other depth intervals where the aquifer is impacted (not presented herein). It is
also noted that a small area on the far eastern side of the plume may not be captured at the southern line of
extraction wells, but is subsequently captured at the northern line of extraction wells. This could
potentially be due to a slight discrepancy between the simulated ground-water flow direction in the model
versus the actual flow direction suggested by the shape of the plume outline.
At 775 gpm, the particle tracking results (Figure B2-15) indicate that the overall plume is still captured.
Again, Figure B2-15 illustrates the results for one depth interval where the aquifer is impacted, and
similar results were achieved for other depth intervals where the aquifer is impacted (not presented
herein). Again, there is a small area on the far eastern side of the plume that may not be captured at the
southern line of extraction wells, but is subsequently captured at the northern line of extraction wells.
Step 5 - Evaluate Concentration Trends
A baseline sampling event for explosives concentrations in monitoring wells was conducted prior to
startup of the system. Water quality data were collected again in May and June of 2003, approximately
one year after extraction was initiated, primarily to monitor progress of aquifer restoration. These data
are not relied upon for evaluating capture because this evaluation of capture was done so soon after
pumping was initiated. However, continued water quality monitoring will provide data from which
long-term trends can be determined and evaluated to provide additional evidence regarding capture.
Step 6 - Interpret Actual Capture Based on Steps 1-5. Compare to Target Capture Zone(s). Assess
Uncertainties and Data Gaps
Based on evaluations of multiple lines of evidence discussed in Step 3 to Step 5, the actual capture
achieved by the extraction wells is interpreted in Step 6, and the following items are addressed:
Compare the interpreted capture zone to the Target Capture Zone
• Assess uncertainties in the interpretation of the actual capture zone
Assess the need for additional characterization and/or monitoring
• Evaluate the need to reduce or increase extraction rates
Table B2-6 presents the summary of the capture zone evaluation for this site.
-------
Figure B2-13. Summary of Drawdown Response to Pumping: Original Model (a) and Re-Calibrated Model (b)
(a)
Computed vs Observed
1.5 2 2.5 3
Observed (ft msl)
775 gpm » 1138 gpm -Line of Equality
W
Computed vs Observed
2.5
Observed (ft)
775 gpm o 1135 gpm Line of Equality
-------
Figure B2-14. Summary of Particle Tracking Results, Current Pumping Rate of 1,135 gpm
Note: When this figure is viewed in
black-and-white, the extent of the total
capture zone is illustrated. When
this figure is viewed in color (such as
from within the PDF digital version),
the colors additionally highlight the
capture zones of individual wells.
Capture zone
of a public water
supply well
-N-
Legend
/ Plume Extent
Extraction Well
SCALE IN FEET
-------
Figure B2-15. Summary of Particle Tracking Results, Current Pumping Rate of 775 gpm
Note: When this figure is viewed in
black-and-white, the extent of the total
capture zone is illustrated. When
this figure is viewed in color (such as
from within the PDF digital version),
the colors additionally highlight the
capture zones of individual wells.
Capture zone
of a public water
supply well
-N-
Legend
f Plume Extent
Extraction Well
SCALE IN FEET
-------
Table B2-6. Summary of Capture Zone Evaluation
Step
Summary/Conclusions
Step 1: Review site data.
site conceptual model.
remedy objectives
Completed, all determined to be up-to-date and adequate.
Step 2: Define "Target
Capture Zone(s)"
Clearly defined horizontally and illustrated on maps, no vertical Target Capture Zone
specified (however, as discussed, extraction well screens were designed to sufficiently
provide horizontal hydraulic containment for all impacted depths).
Step 3a: Water level maps
Contours and flow vectors indicate successful capture for the overall system for the
current extraction rate (1,135 gpm) and the reduced extraction rate (775 gpm). There
appears to be a small area on far eastern side of the plume, at the southern line of
extraction wells, where there may be a lack of capture. This could be the result of an
uncertain water level measured east of extraction well XP-1. Also, that area is within the
interpreted zone of capture for the northern extraction wells, and thus within the capture
zone of the overall system. Water level measurements are available from locations near
the extraction wells so the evaluation is not biased by water levels at extraction wells.
Step 3b: Water level pairs
Actually uses triangles rather than pairs, results suggest inward flow and successful
creation of a flow divide for both the current extraction rate (1,135 gpm) and the reduced
extraction rate (775 gpm).
Step 4a: Simple horizontal
capture zone analyses
Estimated flow rate calculation suggests the current pumping rate of 1,135 gpm is likely
sufficient at both lines of extraction, and the reduced pumping rate of 775 gpm may or
may not be sufficient.
Capture zone width calculation suggests the long-term average pumping rate of 1,135
gpm is sufficient at both lines of extraction, and the reduced pumping rate of 775 gpm is
sufficient for the low value of transmissivity but potentially not sufficient for the high
value of transmissivity. This suggests capture is likely sufficient, but with somewhat
less certainty than with the current (higher) pumping rates.
Step 4b: Ground-water flow
modeling with particle
tracking
Model calibration was updated after the system began operating based on observed
system performance, and particle tracking results indicate successful capture for both the
current extraction rate (1,135 gpm) and the reduced extraction rate (775 gpm). A small
area on far eastern side of the plume may not be captured at the southern line of
extraction wells, but is subsequently captured at the northern line of extraction wells.
This could potentially be due to a slight discrepancy between the simulated ground-water
flow direction in the model versus the actual flow direction suggested by the shape of the
plume outline.
Step 5: Concentration
trends
Not relied upon for short-term evaluation of capture.
Step 6: Interpret actual
capture and compare to
Target Capture Zone
The actual capture zone is interpreted to be sufficient for the current extraction rate
(1,135 pm). However, there is some uncertainty regarding capture on the far eastern side
of the plume, along the southern line of extraction wells, for some lines of evidence (i.e.,
water levels and particle tracking).
Actual capture is nearly complete and may be sufficient for the reduced extraction rate
(775 gpm), although the capture zone width calculation indicates that capture may not be
sufficient for the high value of transmissivity. Again, there is some uncertainty
regarding capture on the far eastern side of the plume, along the southern line of
extraction wells, for some lines of evidence (i.e, water levels and particle tracking).
-------
As discussed in Exhibit 8 of the main document, a summary of the following items is appropriate:
• Is capture sufficient, based on "converging lines of evidence"?
The actual capture zone is interpreted to be sufficient for the current extraction rate (1,135 pm).
It appears that the zone of capture is larger than the Target Capture Zone at the northern line of
extraction wells, based on multiple lines of evidence. This provides a safety factor that accounts
for uncertainties, and is likely due to a flattening of the hydraulic gradient caused by the pumping
at the southern line of extraction wells. There is some uncertainty regarding capture on the far
eastern side of the plume, at the southern line of extraction wells, for some lines of evidence (i.e.,
water levels and particle tracking). Therefore, there is uncertainty as to whether or not the
Target Capture Zone is fully satisfied for the southern line of extraction wells. The causes of this
uncertainty are discussed below. However, multiple lines of evidence indicate that if there is a
lack of capture in that area at the southern line of extraction wells, the water in that area of the
plume would be subsequently captured at the northern line of extraction wells.
Actual capture is nearly complete and may be sufficient for the reduced extraction rate (775
gpm), although the capture zone width calculation indicates that capture may not be sufficient
for the high value of transmissivity. This issue is probably best resolved through calibration and
verification of the ground-water flow model. The results of the particle tracking analysis, based
on the ground-water flow model, indicate that the simulated extent of capture is greater than
indicated by the simple capture zone width calculation using the high value of transmissivity.
Thus, the high value of transmissivity used in the simple calculations is probably higher than the
calibrated value of transmissivity. Again, there is some uncertainty regarding capture on the far
eastern side of the plume for some lines of evidence (i.e, water levels and particle tracking).
• Key uncertainties/data gaps
The water level map constructed for the case with no pumping (Figure B2-7) indicated that some
monitoring wells may have erroneous measurements or datums, because the flow vectors indicate
complications in the ground-water flow patterns that would not be expected in the absence of
pumping. Datums for these wells should be reviewed or re-surveyed; however it should be noted
that the magnitude of possible error is on the order of inches.
As noted above, water level maps indicate some uncertainties in the capture zone evaluation
along the far eastern plume boundary at the southern line of extraction wells. In particular, there
is a potential water level data point east of extraction well XP-1 that might be errant. A potential
approach is to confirm that the measurement is accurate in the field. Another approach is to re-
contour the water levels without that value, or with a potentially different value, to determine if
the interpretation of water level contours east of extraction well XP-1 changes as a result.
Additional water level monitoring locations in that area might also help resolve this issue.
It was noted in the analysis of particle tracking results that a small area on the far eastern side
of the plume may not be captured at the southern line of extraction wells, but is subsequently
captured at the northern line of extraction wells. This could potentially be due to a slight
discrepancy between the simulated ground-water flow direction in the model versus the actual
flow direction suggested by the shape of the plume outline. This flow model could be evaluated
to determine if small changes to the model boundaries or parameter values might lead to a slightly
different simulated flow direction, such that the orientation of the simulated capture zones more
closely aligns with the interpreted plume shape.
-------
Recommendations to collect additional data, install new monitoring wells, change current
extraction rates, change number/location of extraction wells, etc.
The capture zone evaluation summarized herein concluded that the current system (1,135 gpm)
was pumping more water than was required for successful capture. It also concluded that the
reduced pumping rate of 775 gpm appeared to capture the plume in most areas (i.e.. nearly
complete capture), but with some uncertainty regarding capture near XP-1 (the easternmost
extraction well along Highway 104) and also some uncertainty at the northern line of extraction
wells. At this site, a recommendation resulting from the capture zone evaluation was to reduce
the total pumping rate from 1.135 gpm to 900 gpm to capture less clean water (specific rates at
individual wells were recommended, but those details are not presented herein). This provided a
"safety factor" relative to the 775 gpm scenario. The recommendation also suggested subsequent
capture zone evaluations after the new extraction rates were implemented, to verify with field
data that capture continued to be sufficient under the new pumping strategy. Long-term
evaluation of concentration trends at performance monitoring wells located downgradicnt of the
capture zone associated with each line of extraction wells will allow the success of the extraction
wells to prevent plume migration (horizontally and vertically) to be verified. Uncertainties
identified above could be addressed by reviewing and/or re-surveying the datums for several
wells where anomalous water levels were indicated, adding water level measurement locations
near extraction well XP-1. and potentially making slight modifications to the ground-water flow
model so that the simulated capture zones more closely align with the interpreted plume shape.
-------
&ER&
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGES FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
Recycled/Recyclable
Printed with vegetable-based ink on
paper that contains a minimum of
50% post-consumer fiber content
processed chlorine free
------- |