United States
Environmental Protection
Agency
Optimal Well Locator (OWL)
A Screening Tool for Evaluating
Locations of Monitoring Wells
User's Guide
Version 1.2
-------
EPA600/C-04/017
February 2004
Optimal Well Locator (OWL)
A Screening Tool for Evaluating Locations of Monitoring Wells
User's Guide
Version 1.2
Prepared by:
Ponniah Srinivasan
CertainTech, Inc.
20695 Settlers Point Place
Sterling, VA 20165
Daniel F. Pope
Dynamac Corporation
3601 Oakridge Blvd.
Ada, OK 74820
Elise Striz
Ground Water and Ecosystems Restoration Division
U.S. Environmental Protection Agency
Ada, OK 74820
Prepared Under EPA Contract Number 68-C-02-092
Project Officer
David S. Burden
Ground Water and Ecosystems Restoration Division
United States Environmental Protection Agency
Ada, OK 74820
NATIONAL RISK MANAGEMENT RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OH 45268
-------
Notice
The U.S. Environmental Protection Agency (EPA) through its Office of Research and Development
funded and managed the research described here under contract to Dynamac Corporation and its
subcontractors. It has been subjected to the Agency's peer and administrative review and has been
approved for publication as an EPA document. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
All research projects making conclusions or recommendations based on environmental data and funded
by the U.S. Environmental Protection Agency are required to participate in the Agency Quality Assurance
Program. This project was conducted under an approved Quality Assurance Project Plan. The procedures
specified in this plan were used without exception. Information on the plan and documentation of the
quality assurance activities and results are available from the Project Manager.
When available, the software in this document is supplied on an "as is" basis without any guarantee or
warranty of any kind, express or implied. Neither the United States Government (U.S. Environmental
Protection Agency, Office of Research and Development, National Risk Management Research
Laboratory, Ground Water and Ecosystems Restoration Division), Dynamac Corporation or its
subcontractors, nor any of the authors accept any liability resulting from the use of this software.
Disclaimer
Neither the EPA nor Dynamac Corporation or its subcontractors GeoTrans and CertainTech (hereinafter
collectively referred to as Dynamac and subcontractors), will be liable under any circumstances for
the direct or indirect damages incurred by any individual or entity due to this software or use thereof,
including damages resulting from loss of data, loss of profits, loss of use, interruption of business,
indirect, special, incidental or consequential damages, even if advised of the possibility of such damage.
This limitation of liability will apply regardless of the form of action, whether in contract or tort,
including negligence.
The EPA and Dynamac and subcontractors do not provide warranties of any kind, express or implied,
including but not limited to any warranty of merchantability or fitness for a particular purpose or use, or
warranty against copyright or patent infringement.
The entire risk as to the quality and performance of the program is with the user. Should the program
prove defective, the user assumes the entire cost of all necessary servicing, repair, or correction.
The mention of a trade name is solely for illustrative purposes. Neither the EPA nor Dynamac and
subcontractors do hereby endorse any trade name, warrant that a trade name is registered, or approve a
trade name to the exclusion of other trade names. Neither the EPA nor Dynamac and subcontractors give,
or imply, permission or license for the use of the trade name.
IF USER DOES NOT AGREE WITH TERMS OF THIS LIMITATION OF LIABILITY,
USER SHOULD CEASE USING THIS SOFTWARE IMMEDIATELY AND RETURN
IT TO THE EPA. OTHERWISE, USER AGREES BY THE USE OF THIS SOFTWARE
THAT USER IS IN AGREEMENT WITH THE TERMS OF THIS LIMITATION OF
LIABILITY.
This document is being published electronically for distribution with the associated software via
downloading from the World Wide Web or on CD-ROM. To access the software electronically, the user's
system must be connected to the World Wide Web.
-------
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 these mandates, 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 is the Agency's center for investigation of
technological and management approaches for reducing risks from pollution that threatens human health
and the environment. The focus of the Laboratory's research program is on methods for the 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 scientific
and engineering information to support regulatory and policy decisions; and providing the technical
support and information transfer to ensure effective implementation of environmental regulations and
strategies at the national, state, and community levels.
The Optimal Well Locator ( OWL) program is a simple screening tool to evaluate and optimize the
placement of wells in long term monitoring networks at relatively simple sites. OWL uses typically
available site data such as routine water level measurements and well locations in a simple algorithm to
estimate the change in the direction/magnitude of the ground-water flow field overtime. By viewing the
potential variation in the ground-water flow field direction and magnitude over time, users will develop
a greater understanding of the dynamic nature of ground-water flow at their site instead of assuming a
static ground-water flow field, which is seldom the case. In addition, OWL uses minimal contaminant
plume characterization data in a simple contaminant migration algorithm to estimate the potential plume
migration paths at a site as a function of the change in ground-water flow field direction/magnitude over
time. OWL combines the possible plume migration paths into a composite plume to allow users to clearly
visualize the possible locations where the plume can be present. In addition, OWL can produce a map
which combines plume migration with monitoring well coverage to pinpoint locations where the plume
concentration is to predicted to be high and the well density to be low. This information will allow users
to assess the location of existing wells in their networks and better position future wells to intercept
the plume. OWL is specifically designed for users who have minimal ground-water and contaminant
modeling experience. Site managers can use OWL as a screening tool to evaluate their site monitoring
well networks and potentially reduce monitoring costs by optimizing well performance.
Stephen G. Schmelling, Djrector
Ground Water and Ecosystems Restoration Division
National Risk Management Research Laboratory
-------
IV
-------
Table of Contents
Section Page
Notice/Disclaimer ii
Foreword iii
Acknowledgements vii
1.0 Introduction 1
1.1 Program Objectives 1
1.2 Program Features 2
1.3 Program Data Requirements 3
1.4 Program Theory 4
1.4.1 Ground-Water Flow Field Variation 4
1.4.2 Plume Migration Path Variation 5
1.4.3 Composite Plume 5
1.4.4 Well Locations and Monitoring Well Coverage 6
1.4.5 Well Optimal Location Factor 6
1.5 Program Application and Limitations 6
2.0 Installing OWL 9
2.1 Hardware Requirements 9
2.2 Software Requirements 9
2.3 Installation Steps 9
2.4 Uninstalling OWL 15
3.0 Program Layout and Screen Appearance 21
3.1 OWL Layout in the Windows Environment 21
3.2 Work Outline and Map Layers Windows 22
3.2 Map View Window 23
3.3 Message View Window 24
4.0 Guided Tour and Field Case Study Tutorial 25
4.1 Site Selection and Characterization 26
4.2 Opening OWL 27
4.3 Opening a New Project File 27
4.4 Setting the Measurement Units 29
4.5 Import the Site Basemap 29
4.6 Set Basemap Scale 29
4.7 Import Well Location Data 32
4.8 Import Ground-Water Elevation Data 34
4.9 Set Plot Limits 37
4.10 Inputting Model Data 40
4.11 Choosing the Source Location 41
4.12 Run Linear Regression 42
4.13 View Raw Output Files 42
4.14 View Water Level Contours 44
4.15 Transport Run 46
-------
Section Page
4.16 Export Model Results 47
4.17 View Plume Contours 47
4.18 View Minimum Distances 49
4.19 Well Optimal Location Factor (WOLF) 52
4.20 Project Data Management and Printing 55
5.0 Glossary 57
6.0 References 61
Appendix A. Field Case Study Site Characterization Data Collection and
Evaluation for Application of OWL A-
AppendixB. Program Algorithms B-
Appendix C. Site Basemap Formats C-
Appendix D. Spreadsheet Formats for Well Location and Ground-Water Elevation Data D-
Appendix E. Guidance to Edit Ground-Water Elevation Data for OWL E-
VI
-------
Acknowledgements
This project was funded by the U.S. EPANRMRL Ground Water and Ecosystems Restoration Division,
under Contract Numbers 68-C-99-256 and 68-C-02-092 to Dynamac Corporation, and Subcontract No.
91517 to GeoTrans, Inc.
The OWL presented in this guide is an integrated Windows-based program based on the concepts and
methodology authored by Dr. John Wilson and Dr. Elise Striz of EPA. The OWL program was integrated
into the Windows environment using a Graphical User Interface (GUI) developed by Ponniah Srinivasan
of CertainTech, Inc.
vn
-------
vm
-------
1.0 Introduction
The Optimal Well Locator ( OWL) program was developed and designed to be a screening tool to
evaluate and optimize the placement of wells in long term monitoring networks at relatively simple sites.
The OWL program enables users to approximate the change in ground-water flow field over time at their
site and then use this variation to forecast the potential location of the plume. In addition, the program
combines this predicted plume location with existing monitoring well coverage to allow the user to assess
if their existing wells are in the optimal location and determine where to place new wells to intercept the
plume. This information will allow site managers and others to make informed decisions on monitoring
well locations using site specific data.
The OWL program is designed for users who have limited modeling experience and require a simple
transparent tool to evaluate their monitoring well placement. It is not intended to replace more advanced
models used to evaluate and optimize monitoring well locations. OWL is also not a comprehensive
ground-water flow and contaminant transport model. It should not be used to try to reproduce the
existing plume at a site. Instead, OWL is a screening program to evaluate well placement using typically
available site data and simple algorithms to predict the possible plume locations at a site as a function
of the variation in ground-water flow direction and magnitude over time. It is therefore subject to many
limitations, which are provided in this manual to ensure its appropriate application.
1.1 Program Objectives
The first objective of the OWL program is to allow the user to visualize the change in ground-water flow
direction and magnitude at a site over time. To accomplish this, OWL uses typically available site data
including routine water level measurements and well locations in a regression algorithm to produce a
linear plane approximation of the ground-water flow field. OWL provides an individual map of the
ground-water potential for each measurement event. By viewing the variation in the ground-water flow
direction and magnitude over time in these potential maps, users will be able to see and develop a greater
understanding of the dynamic nature of ground-water flow at their site.
The second objective of the program is to allow the user to visualize the influence of the changes in
ground-water flow over time on plume migration at the site. OWL uses minimal contaminant plume
characterization data in a simple contaminant migration algorithm to forecast the potential plume
migration path for each approximate ground-water flow field. OWL provides a separate plume
concentration map for each ground-water flow field to demonstrate to users how variations in ground-
water flow magnitude and direction can impact the movement of contaminants over the site. In addition,
OWL combines all of the predicted plume migration paths into a composite plume concentration map.
This map will show users the forecasted locations where the plume is likely to be present at their site,
giving them a full picture of the sweeping coverage of the plume.
The third and final objective of OWL is to allow the user to assess the location of existing wells and to
better position future wells. OWL performs this task by producing a map which combines the composite
plume concentrations with a simple measure of monitoring well coverage into a Well Optimal Location
Factor (WOLF). The WOLF map has two purposes. First, it will enable users to identify locations where
the plume concentration is predicted to be high and the well density to be low. Such a spot would be an
optimal location for a new well. Second, if plume concentration is forecast to be low and well density is
high in an area, it will alert users that it may be possible to eliminate redundant existing wells.
1
-------
1.2 Program Features
The OWL program provides a user-friendly graphical user interface (GUI) program environment for easy
data entry, program execution and graphical presentation of results. The OWL program allows the user to:
1. Display a user-provided site map (basemap). OWL is capable of importing an electronic site
map and displaying overlays of data such as well locations, ground-water head contours, and
contaminant plumes on this basemap.
2. Approximate the ground-water flow direction and magnitude at the site (the ground-water
"flow field"), using user-supplied ground-water elevation data from the site and display
on the basemap. For each sampling event, OWL will fit a linear plane surface through the
ground-water levels using a least squares regression and display the water level contour lines
on the basemap. OWL supplies a coefficient of determination (R2) for each approximation,
to give the user an indication of how well the linear plane model approximation fits
the observed water level data for each measurement date. If the linear plane provides a
satisfactory fit, the user can compare the displays on the basemap to visually assess the
changes in ground-water flow direction and magnitude at the various sampling events.
3. Calculate and display on the basemap the contaminant plume migration paths using a
constant source concentration for each approximated ground-water flow field over a user-
chosen time. The user can then compare the predicted plume migration paths to see how
variation in the ground-water flow direction and magnitude can cause the predicted location
and spread of the plume to vary for the same time of travel and source concentration.
4. Calculate and display on the basemap a composite plume which displays the average
concentration contours of all plume migration paths combined. This provides users
with a visual demonstration of the possible areas which may be swept by the plume as a
consequence of the variation in the ground-water flow.
5. Calculate and display contours of the minimum distance to the nearest monitoring well on the
basemap to interpret the well coverage at the site. This map shows users the areas of the site
where there is minimal monitoring well coverage.
6. Calculate and display contours on the basemap of a Well Optimal Location Factor (WOLF)
which combines the monitoring well coverage and the predicted location of the composite
plume. Large values of the WOLF indicate locations where the monitoring well density is
relatively low and contaminant concentrations are forecast to be high, i.e., good locations for
new wells. Small values alert the user to regions where monitoring wells may be redundant,
i.e., high monitoring well density and low predicted plume concentrations.
7. Display and print tables of the calculated data used to make the graphical displays in items
2-6 above. These tables can be used to export the calculated data to other programs for
further modeling, graphical display, or for checking results. OWL provides the tables in text
format, which enables them to be used by many word processing, database, spreadsheet, and
modeling programs.
-------
8. Display and print the site basemap with all of the described graphical displays with legends.
1.3 Program Data Requirements
The OWL program requires data which are typically available at most ground-water monitoring sites
including:
1. A Scaled Basemap : The OWL program requires the user to have a scaled basemap of the site
in electronic form. The basemap should be oriented with the y-axis parallel to north. This
basemap is imported into the program and used as the basis for all graphics. It is also used
to define the well locations (x,y) in consistent distance units on a rectangular grid. Most
sites have a basemap in electronic or hard copy form. If the basemap is already in electronic
form it may be directly imported if it has the correct format. OWL will import basemaps
in electronic formats including raster (bitmap) images and vector format files. Raster maps
(such as an aerial photo or scanned image) should be in Windows or OS/2 Bitmap (.bmp)
format. Vector maps (CAD files) should be in AutoCAD Version 12 or 14 DWG format.
OWL can accept other formats; see Appendix C Site Basemap Formats and Scaling. If the
map is only in hard copy form, it may be scanned into electronic form using a scanner. When
scanning, the user should select a *.bmp format. The lowest resolution for the scanned bitmap
which still maintains image clarity should be used. By using a lower bitmap resolution, the
OWL program will be able to display its many graphics more efficiently. An example of how
to prepare and import a basemap is given in Appendix A.
2. Well Locations: OWL requires that all of the well locations be in consistent (x,y) coordinates
from a rectangular grid based on the map dimensions. The map distance units must be in feet
or meters or converted to these units. If the basemap is already in electronic form, a file will
exist which provides the (x,y) values of the well locations. If one was fortunate enough to
have or to obtain a survey, the well locations will be in standard mapping coordinates, such
as UTM or State Plane, which can be used in the program. Sometimes well coordinates have
been measured in terms of an artificial grid (i.e., distances from some reference origin to
each well at the site) which is also satisfactory. For most sites, however, the basemap is only
available in hard copy and shows well locations but no (x,y) coordinates. In this case, the
user must define the well locations themselves by creating their own rectangular grid of the
site dimensions based on the map scale bar. This may be accomplished by simply overlaying
a grid on the map by hand, measuring the well locations with a ruler and employing the
scale bar dimension to convert to distance units. The method by which the rectangular grid
is defined is not important, as long as a consistent set of (x,y) coordinates is obtained for
all of the wells which matches the basemap dimensions. An example of how to assign well
locations using the grid method is given in Appendix A.
3. Ground-Water Elevations: The OWL program requires routine measurements of ground-
water levels (preferably monthly or quarterly) from a monitoring well network demonstrating
good spatial coverage of the site. All monitoring wells must be sampled at approximately
the same time at each sampling event; e.g., a quarterly sampling event that takes a month
will not represent the ground-water flow field at a specific point in time. Sampling events
should be regularly spaced; e.g., quarterly or semiannually to produce the best record of
ground-water flow field variation over time. The program requires that all the ground-water
levels be corrected to elevations from a fixed datum such a mean sea level. They must also be
corrected for any free product. The measurement units for the elevations must be in the same
-------
units (feet, meters) as the well locations (x,y). An example of how to convert from ground-
water levels to elevations and correct for free product is given in Appendix A.
4. Subsurface Geology: The monitoring wells used to provide water level data for OWL must be
screened in the same aquifer at the site. The aquifer should be homogeneous, isotropic and of
constant thickness to meet the assumptions inherent to the contaminant migration algorithm
used by the program. Site data on the subsurface geology from cores and well completions
can help the user select wells and aquifers that meet these restrictions. Although this data is
not explicitly entered into the program, it is essential to review the subsurface geology at the
site to ensure it does not violate the program assumptions. An example of how to determine if
the aquifer at a site meets these requirements is given Appendix A.
5. Site Characterization: The contamination and hydrologic characteristics of the aquifer at the
site must entered into the OWL program. This information includes:
contaminant source width
contaminant source concentration
contaminant retardation factor
contaminant half-life
aquifer hydraulic conductivity
effective porosity
longitudinal dispersivity
transverse dispersivity
The contaminant width and contaminant source concentration should be measured at the
site. The contaminant retardation factor may be estimated from site data, or estimated using
values from literature references for the contaminant and the site aquifer matrix. Contaminant
half-life should be estimated from site data. Hydraulic conductivity should be determined
from site data. Effective porosity may be estimated using literature reference values or
simple relationships; e.g. longitudinal dispersivity = 10% of plume length, and transverse
dispersivity = 10% of longitudinal dispersivity. Appendix A provides specific information on
how to obtain the site characterization data.
1.4 Program Theory
OWL is intended for users who have limited modeling and programming experience and require a simple
tool to assess monitoring well placement. It, therefore, employs an elementary step-wise approach to meet
these objectives as described in the following sections. The user is referred to Appendix B for a detailed
description of the algorithms used to perform these tasks.
1.4.1 Groun d- Water Flow Field Variation
The OWL program can be used to approximate the ground-water flow field for each round of ground-
water elevation measurements. Ground-water elevations should be recorded as hydraulic head values,
z, for every well location, (x, y), defined as a point on a rectangular grid overlain on the site basemap
or in standard mapping coordinates (e.g., UTM, State Plane). For each round of ground-water elevation
measurements entered into the program, OWL performs a least squares regression analysis of the water
level values, (x,y,z) and produces the equation of the linear surface which best fits the observations. OWL
-------
uses the equation to calculate the magnitude of the hydraulic gradient and the direction of flow given as
an azimuth angle from North for each set of observations.
In addition, the OWL program provides an estimate of the goodness of fit of the linear surface to the
water level observations through the coefficient of determination, R2. The coefficient of determination is
the square of the Pearson correlation coefficient. An R2 value close to one indicates that a linear model
is a good approximation to the data, whereas a value less than one indicates that some variability in the
data is not explained by the linear model. For example, an R2 value of 0.8 indicates that the linear surface
model explains 80% of the variation in the water level data, whereas a value of 0.5 indicates the model
would only explain 50% of the variability. Values of R2 more than 0.5 are preferred; values less than
0.5 are considered to be poor, indicating that the calculated linear surface does not fit the actual ground-
water surface well. The user should examine the data to determine if particular data points contribute
disproportionately to low R2 values. Such data points may relate to recharge, discharge or other locations
that adversely affect the linearity of the ground-water elevation surface.
A comparison of the simplified ground-water flow fields derived from each round of ground-water level
measurements provides the user with a visual assessment of the variation of ground-water flow at the site.
These fluctuations may be an outcome of seasonal variability, such as changes over the years in response
to wet or dry years, and recharge and discharge. This variation in the ground-water flow field over time
will directly impact the migration path of any contaminant plume.
1.4.2 Plume Migration Path Variation
The OWL program allows the user to examine the possible migration paths of a contaminant plume
arising from a single continuous source as a function of each ground-water flow field. In particular,
OWL uses the linear hydraulic head surface from the regression (to define the flow field), user-entered
contaminant source data, and a user-chosen simulation time in the Domenico ID analytical transport
solution (Domenico, 1987). This solution of the one-dimensional advection-dispersion equation produces
a two-dimensional estimate of contaminant plume shape for each ground-water flow field. The program
also reports contaminant concentrations for each node on a grid of user-chosen spacing. The estimates
can be plotted on a map of the site for display, and can be exported in a space-delimited text file for use
by other programs.
1.4.3 Composite Plume
The OWL program creates a separate plume shape estimate for each set of ground-water elevation
data (i.e., each round of water sampling). In addition to allowing viewing of the plume migration paths
separately, the OWL program also averages the estimated contaminant concentrations from each plume
shape to generate a composite plume shape based on several sets of sampling data (i.e., several sampling
rounds). The composite average plume shape can be compared with the actual plume shape as revealed
by sampling in order to assess the validity of the site conceptual model (e.g., source size and location,
ground-water flow directions and rates, and attenuation). This composite plume can be used as a measure
of the potential variation in the plume location at the site over time caused by the changing ground-water
flow field.
1.4.4 Well Locations and Monitoring Well Coverage
OWL provides spatial measure of monitoring well coverage at the site based on contours of "the distance
to the nearest monitoring well." The program calculates the distance from each node of the user defined
-------
grid to the nearest monitoring well, and plots the results as contours on the site basemap. By examining
these "distance to nearest monitoring well" contours, the users can visually identify regions of high and
low monitoring well coverage.
1.4.5 Well Optimal Location Factor
OWL uses the "distance to nearest monitoring well" contours, along with the composite plume which
represents the potential variation in all the plume migration paths, to calculate a "Well Optimal Location
Factor" (WOLF). WOLF is denned as the product of the composite plume concentration at a grid
node and the square of the "distance to the nearest monitoring well" at the grid node. Contours of the
WOLF values are plotted on the site basemap for display. High WOLF values indicate areas where
the contaminant concentrations are expected to be high and the monitoring well coverage is low. Such
locations may be considered optimal positions for new wells. In contrast, low WOLF values indicate areas
where concentration is expected to be low and monitoring well coverage high. These areas would be good
locations to assess if existing monitoring wells are redundant.
1.5 Program Application and Limitations
The OWL modeling algorithms are intentionally very simple, which means the program is subject
to many assumptions and limitations. OWL first assumes that the ground-water surface in a single
homogeneous, isotropic aquifer at a site can be represented as a linear plane. The program uses a linear
regression technique to create a planar ground-water surface for each set of water level measurements.
The presence of substantial surface relief, recharge or discharge zones, surface water features, pumping
wells or ground-water divides on or near the site may cause deviation of the ground-water surface from a
plane and violate this modeling assumption. Therefore, the user will find that OWL is mainly restricted
to application at relatively small plume sites, where the user is already aware that the ground-water flow
field can be approximated by a linear plane. The larger the site, the less likely this restriction will be met.
The Domenico contaminant migration algorithm used by OWL is also very simple, and is only applicable
to sites with well defined uncomplicated contaminant sources. An example of an appropriate site would
be an underground storage tank (UST) site with a single contaminant source and uniform hydrogeology.
The contaminant source should exhibit minimal variability in its distribution throughout the vadose zone
and saturated zone. Given these restrictions, OWL will generally be limited to LNAPL sites where most
of the LNAPL is held in a small volume of vadose zone, with no more than a small pool of LNAPL on
the water table. Long plumes of LNAPL on the water table, and large smear zones would be problematic.
DNAPL sites might be considered if there is no evidence for movement of the DNAPL very far down
into the aquifer. Sites where DNAPL has pooled on retarding layers should be excluded. It will be
noted that, in general, these restrictions mean that the program is best used for UST sites, where fuels
are the contaminants, the fuel has not moved far from the tank as a NAPL, most of the source has been
removed, and the hydrogeology is very simple (homogeneous, isotropic, with no recharge, divides, etc. on
the sitethis practically means water table aquifers in a sand matrix underlain by thick continuous clay
layers).
The OWL program assumes that the contaminant source is linear and constant along a vertical plane
below the line. Sites with multiple or complexly-shaped sources, significant water table variations
where smear zones or contaminated vadose zones are present, or other conditions where the source may
vary significantly from the model assumptions should be assessed with more complex models. OWL
is also not applicable where ground-water flow is very slow, such as in clay media, because the model
does not consider molecular diffusion. When using OWL, the source characteristics (source width and
concentration) can be changed by the user to test sensitivity of the predictions to variations in the source.
-------
The OWL model assumes linear equilibrium adsorption and first-order decay. These assumptions are
mainly applicable to certain organics, especially BTEX and chlorinated solvents, under the limited
organic matter content typical of the subsurface. Inorganics, or reactive contaminants, may deviate
markedly from these assumptions.
OWL allows the user to input values for several site parameters (e.g., conductivity, dispersivity, as
listed above). The user should determine a range of these values that convey the variability in the site
parameters, and input several values of each parameter into OWL to determine how the results are
affected by variations in the site parameters. Hydraulic conductivity, solute (contaminant) half-life,
source width and source concentration are examples of factors likely to significantly affect model results.
In particular, the user should input a range of values of conductivity (k) to determine the sensitivity to
k. This would show variability in the longitudinal aspect (how long the plume gets). This would be
important if the program were used to locate guard wells because the length of plume would be expected
to vary significantly due to differing k values in the subsurface.
The plume concentration and WOLF functions in OWL are based on a two-dimensional analytical model
(the Domenico analytical solution of the one-dimensional advection-dispersion equation (Domenico,
1987) which assumes a constant source and constant hydrogeological properties throughout the modeled
area. The aquifer and flow field are assumed to be homogeneous and isotropic; a uniform conductivity
(k) is assumed throughout the modeled volume. OWL is not suitable where vertical flow gradients
significantly affect contaminant transport, or where hydrogeologic conditions change significantly over
the modeled part of the site. OWL does not and cannot replace professional judgement for siting wells
and assessing plume behavior, but assumes that the user is able to adequately assess the applicability of
the program results to particular site conditions.
In summary, the application of the OWL program will be subject to the following limitations:
1. The ground-water hydraulic head must be reasonably represented by a linear plane surface.
There should be no ground-water divides, significant surface water bodies, pumping/injection
wells, or anything else which could cause the ground-water heads across the site to deviate
significantly from a flat surface.
2. The contaminant plume must be derived from a single continuous constant contaminant
source in a well defined location. A continuous source provides a constant stream of
dissolved contaminant to maintain the contaminant plume.
3. The monitoring wells used to provide data for OWL must be screened in a single aquifer.
The screens should be of similar length and placement in the aquifer. The aquifer must be
homogeneous and isotropic, and of constant thickness. Homogeneous and isotropic means
that the structure or composition of the aquifer matrix is uniform throughout and has the same
characteristics in all directions. For instance, a uniform sand aquifer which had the same
hydraulic conductivity, whether measured in the horizontal or vertical direction, could be
considered homogeneous and isotropic.
4. The monitoring wells used to provide data for OWL must provide good spatial coverage of
the plume and surrounding aquifer. Too few wells, in scattered locations, will not provide
enough uniformly-distributed ground-water elevation data to allow accurate estimation of the
ground-water flow field and contaminant plume shape.
-------
5. The locations of the monitoring wells must be known using a fixed two-dimensional
coordinate system so that the OWL basemap of the site can be calibrated. At most sites, the
well locations have been determined by survey, so the coordinates are available in UTM, State
Plane, or a local coordinate system. If not, the user must create a rectangular grid derived
from the basemap scale bar dimensions and assign well locations using this grid by hand.
6. The OWL model assumes linear sorption and first-order decay. These assumptions of linear
sorption and first-order decay are mainly applicable to certain organics, especially BTEX
and chlorinated solvents, under the limited organic matter content typical of the subsurface.
Inorganics, or reactive contaminants, may deviate markedly from these assumptions.
-------
2.0 Installing OWL
The steps described in this section are required for successful installation of OWL. Because the OWL
installation program creates files and subdirectories on your hard disk, you must have write-privileges
to use the installation program. Most users should have no problems with installation, but if installation
ends before it is complete, or if there are any error messages during installation, consult the system
administrator in your organization for help.
2.1 Hardware Requirements
The OWL software requires the following hardware for reasonable performance:
A PC capable of using Microsoft Windows 95, 98, NT, ME, 2000 or XP as the operating system
At least 32 MB of RAM
40 MB of free hard disk space for the user directory
A CD ROM drive for installation, if you receive OWL on a CD. A CD drive with at least 4X
speed is recommended for reading the OWL Setup CD.
2.2 Software Requirements
The following software should be installed or made available through a network on the computer running
OWL:
Microsoft Windows 95, 98, NT, ME, 2000 or XP
Spreadsheet software (Excel or Lotus) for creating water level and well locations data worksheets.
2.3 Installation Steps
Run Windows 95, 98, NT, ME, 2000 or XP.
Insert the OWL CD ROM in the CD-ROM drive, and close the drive tray.
Click on the "" button at the lower left-hand corner of the screen, then click on the ""
menu item. The Run dialogue will appear.
-------
Open:
Type the name of a program, folder, document, or
Internet resource, and Windows will open it for you.
I 3
Cancel I Browse,.,
From the dialog command line, type D:setup.exe (D is assumed to be your CD ROM drive letter. If your
CD is a different drive letter, use the appropriate letter). Alternatively, you may click the ""
button and browse to the drive containing the OWL CD, and click on the setup.exe file on the CD. Click
"" to start the setup process. The following setup screen will appear.
Welcome
Welcome to OWL Setup program. This program
will install OWL on your computer.
It is strongly recommended that you exit all Windows programs
before running this Setup Program.
Click Cancel to quit Setup and close any programs you have
running. Click Next to continue with the Setup program .
Cancel
Click "" to continue to the next dialog as shown below.
10
-------
Choose Destination Location
Setup will install OWL in the following folder.
To install into a different folder, click Browse, and select
another folder.
You can choose not to install OWL by clicking Cancel to exit
Setup.
r Destination Folder
CAProgram FilesSOWL
Browse..
< Back |f""||iegt7"""]| Cancel
The default folder shown in the dialog may be left as is, or if desired, click "" to browse to
another folder.
>..Y-
CJAiMC. \u!<.>t*>, '
OK
Cancel
:xit
< Back Next >
Cancel
When the desired folder is chosen, click "" to continue. The dialog box as shown below will
appear.
11
-------
*pl Backup Replaced Files
This installation program can create backup copies of all files
replaced during the installation. These files will be used when
the software is uninstalled and a rollback is requested. If
backup copies are not created, you will only be able to uninstall
the software and not roll the system back to a previous state.
Do you want to create backups of the replaced files?
Cancel
It is strongly recommended that you choose "" to backup any files replaced by the OWL setup
routine. By default the backup is created in a subfolder BACKUP under the OWL folder (or the user may
choose a different folder). Click "" and a dialog will appear prompting for a default menu group
in which to place the OWL start icon.
JES Select Program Manager Group
Enter the name of the Program Manager group to add OWL
icons to:
OWL
Accessories
Administrative Tools
< Back
Next>
Cancel
12
-------
It is not necessary to choose a menu group, but the user may choose a preexisting group or type in a name
for a new group. Click "" to continue.
i Start Installation
You are now ready to install OWL.
Press the Next button to begin the installation or the Back
button to reenter the installation information.
" for the dialog Start Installation to continue. Follow the instructions of the setup program
as they appear on the screen.
The setup program may take several seconds to install and verify the location of necessary Windows
system files. You will see a progress bar dialog indicating progress of installation.
Installing
Current File
Copying file:
C:\..AData\site1 \other_formats\site1 basemap_acad12.dxf
All Files
Time Remaining 0 minutes 13 seconds
Cancel
13
-------
The installation program will then update the system configuration. After installation is complete, you
will see the dialog box Installation Complete. Click "" to close the dialog box.
Installation Complete
OWL has been successfully installed.
Press the Finish button to exit this installation.
Finish >
It may be necessary to restart the computer after installation is complete. If a dialog box appears
prompting a restart of the computer, press "" to proceed.
To run OWL, click on the "" button, select the Programs folder, then choose the OWL folder.
Select the program item OWL to run the program. The initial screen as shown below will appear. The
program is now ready to operate.
D & y a * * e ;. * J. -;,
MinDisI conic
'/QLFccrtot.
I
14
-------
2.4 Uninstalling OWL
To uninstall the OWL program, select "" from the menu on the task bar, choose ","
then choose "." Double click on the icon "." A dialog box
titled "Add/Remove Program Properties" will appear, similar to that shown below.
GH Add/Remove Programs
Currently installed programs:
Sortbylfiame
Scroll through the list of software and select OWL as shown above, then click "." Next
the dialog box Select Uninstall Method will appear. Choose "" method (this is probably
already chosen as the default selection), then press "."
Select Uninstall Method
Welcome to the OWL uninstall program.
You can choose to automatically uninstall this software or to
choose exactly which changes are made to your system.
Select the Custom button to select which modifications are to
be made during the uninstall. Select the Automatic button for
the default uninstall options. Press the Next button to continue.
(" Automatic
P Custom
Next>
Cancel
15
-------
On the next dialog box, the uninstall routine gives the user a choice of including a "rollback" in the
uninstall process. A rollback will use the system files stored during the initial installation of OWL (if that
option was chosen during installation). A rollback reinstalls the stored system files that were originally
on your computer, while removing the system files that the OWL installation added. If other programs
have been installed after OWL was installed, a rollback could cause the removal of system files added to
your computer during the installation of the other programs, possibly causing problems. If other software
has been installed since originally installing OWL, it may be necessary to consult your PC support person
before uninstalling OWL.
Perform Rollback
2
Cancel
If no other program has been installed after OWL was installed, the user may click "," then press
"" to perform an uninstall with rollback.
If other programs have been installed after OWL was installed, click "" and then "" to
proceed with uninstallation without rollback.
16
-------
Perform Uninstall
You are now ready to uninstall the OWL from your system.
Press the Finish button to perform the uninstall. Press the Back
button to change any of the uninstall options. Press the Cancel
button to exit the uninstall.
< Back
Finish
Cancel
The Perform Uninstall dialogue box will appear: click "" to proceed with the uninstall.
In a moment the setup program will uninstall all the OWL files, and restore the Windows system files
(DLL, OCX, and so on) that may have been replaced during OWL installation.
The uninstall program may prompt the user to choose whether to remove specific system files. Removing
these files saves space on your computer, but may cause problems if the files are needed for another
program. It is usually safe to select "" to this dialogue box.
Remove Shared Component
The system indicates that the following shared file is no longer used by
any programs and may be deleted.
C: SWIN N T \System32Wb5db. dll
If any programs are still using this file and it is removed those programs
may not function. Leaving this file will not harm your system. If you are
not sure what to do, you should select the No to All button. Do you want
to remove the shared file?
Yes
No
Yes to All
No to All
17
-------
Once "" is chosen, a message box showing the progress of uninstallation appears briefly, and
then the uninstall process will finish.
For Custom Uninstall
If the "" process is chosen (when the "" method or "" choice is presented), a series of dialogue boxes appear, allowing the user to choose specific
files to be uninstalled. The "" process is unlikely to be useful to the average
computer user, but may be helpful to computer system administrators.
Select Private Files to Remove
2
Cancel
Select System Files to Remove
The following files were copied to your Windows/System
directories during the installation. Use caution when removing
these files, they may be in use by other programs.
C: \WI N N T \System32\Corndlg32. ocx
C: SWIN N T \System32\D bgrid32. ocx
C: \WI N N T \System32\M fc42. dll
C: \WI N N T \System32\M scomctl. ocx
C: \WI N N T SSystem32SM sflxgrd. ocx
C: \WI N N T \System32\M sstdf mt. dll
C: SWIN N T \System32\M svcrt. dll
C: \WI N N T \System32\F! ichtx32. ocx
C: SWIN N T \System32\T abct!32. ocx
C: \WI N N T \System32\D bcdbf32. dll
C: \WI N N T \System32\D bcdib32. dll
C: \WI N N T \S vstem32SD bclst32. dll
Select All !
Select None
< Back
Next>
Cancel
18
-------
Select Directories to Remove
The following directories were created during the installation.
Selecting a directory will remove it and all files and directories
that are contained within it.
C:\Program FilesSQWL
CADocuments and Settings\dpope\Start MenuSPrograms\0\W
il
Select All
±J
Select None
< Back
Next>
Cancel
Select Registry Keys to Remove
The following Registration Database Keys were created during
the installation. Select those keys that you want to remove.
S oftware\M icrosof t\Windows\CurrentVersion\U ninstall\OWL \
S of(ware\M icrosof t\Windows\CurrentVersion\U ninstallSOWL \
li
Select All
J ±J
Select None
< Back
Next>
Cancel
19
-------
1. Input data
-[I] 1.1 Set units
[I] 1.2 Base map
O 1.3 Map scale
Q 1.4 Well locations
D 1.5 Water levels
D 1.6 Plot limits
D 1.7 Model data
Q 1.8 Source location
2. Regression
Q] 2.1 Run linear regression
O 2.2 View raw output files
O 2.3 Water level contours
3. Transport run
O 3.1 Run analytical solution
O 3.2 Export model results
1 ' 3.3 View plume contours
3.4 View MinDist contours
3.5 View WOLF contours
3.2 Project Steps and Map Layers Windows
Upon startup, there should be two windows (Project Steps and Map Layers) on the left side of the main
"OWL" window. The window at the top left (Project Steps) displays the steps which should be followed
in a session for a smooth interaction with OWL. As each step is completed, the corresponding box at
left will be checked. Use this window as a reminder or guidance as progress through the OWL session is
completed.
The window at the bottom left of the OWL window (Map Layers) lists the possible layers that can be
seen in the Map View window. Layers are additions to the basemap, such as ground-water elevation
contours, plume shapes, etc., whose display can be turned on or off without affecting the basemap. A layer
is displayed on the screen over the basemap so that the layer appears to be part of the basemap; i.e., a
ground-water elevation contour layer shows the contours on the basemap. If a layer is not yet available
for display, its name will be shown using dimmed letters. The check box next to the layer name may be
used to display or hide a layer selectively.
22
-------
3.3 Map View Window
Clicking on the "" checkbox in the Work Outline window causes an Open Basemap File
dialogue box to appear for opening a basemap file. The user can browse to and open a basemap file. The
basemap can be made to appear in the Map View window by clicking the "" button at the right-
hand bottom of the dialogue box. The Map View window will fill most of the OWL window.
On the tool bar, there are several buttons for manipulating the view of the basemap shown in the
Map View window.
The function of each button is described below:
The first button allows the user to zoom the basemap view in order to
enlarge the view.
The second button allows the user to zoom to a selected window of the
map. Just select this button and a crosshair cursor will appear.
Move the crosshair to the top corner of your window selection using
the mouse. Click and hold left mouse button. Drag the mouse across
the map to select the window you wish to zoom in on. Click the left
mouse button again to zoom in on the window.
The third button allows the user to zoom out on the basemap view in
order to reduce the view.
The fourth button allows the user to return to the previous view.
The fifth button allows the user to return back to the entire basemap.
At the bottom right of the Map View window, the coordinate display is provided. The numbers displayed
here represent the X and Y coordinates of the cursor position on the basemap. The coordinates are used
for calibrating the basemap, locating wells, and other operations where specific positions on the basemap
must be located in reference to positions at the site.
When a bitmap format basemap is first opened, the XY coordinates shown in the coordinate display are
in pixel units, and the top left corner of the bitmap image is considered the origin (0,0 point). If the user
calibrates the bitmap basemap to real coordinates (in the Set Map Scale dialogue box), the XY coordinates
shown in the coordinate display are based on the real coordinates system, and are in the units of the real
coordinates system. For example, if during the calibration to real coordinates dialogue, the user set the
lower left corner of the bitmap to (0, 0) and the upper right corner to (1000, 1000), moving the cursor
over the bitmap would cause the XY coordinates shown in the coordinate display to vary between (0, 0)
and (1000, 1000), depending on where the cursor was located on the screen display of the bitmap.
X:+500.000000 Y:+3ZO.OOOOOO
23
-------
3.4 Message Window
A message window appears on the lower-right corner of the OWL application window. Messages
regarding file open, file save, and any error conditions are displayed in this window.
Message
$ aved project file: C: \Program FilesMDWLAD ata\site1 \site1. prj
Opened base map file: site"! basemap.bmp)
24
-------
4.0 Guided Tour and Field Case Study Tutorial
The Guided Tour and Field Case Study Tutorial (Case Study/Tutorial) provides a step-by-step tutorial to
introduce the user to OWL. A description of the site used in the Case Study/Tutorial is found in Appendix
A of this document. Notes on site characterization and basemap preparation are also found in Appendix A.
The user should read Appendix A before using this Case Study/Tutorial, and refer to Appendix A for more
site information while using the Case Study/Tutorial.
This Case Study/Tutorial will guide the user to:
produce a project file,
input site data, including a basemap of the site,
carry out regression and Domenico solution calculations on the site data,
apply the results (ground-water elevation and contaminant concentrations contours) for viewing on
the site basemap, and
calculate and view Minimum Distances and WOLF contours on the basemap.
For comparison of your results with an analysis prepared by professional modelers, the OWL installation
program provides a folder containing a completed Case Study project file (with all data input and calcula-
tions completed). This completed Case Study project file is found in the folder C:\Program Files\OWL\
Data\sitel, named 'casestudprf (assuming that you used the default installation locations for OWL).
Using OWL - Basic Workflow
Using OWL involves four basic steps:
Provide Site Data, Set Units, and Choose Map Presentation Settings
The user inputs site data, including an electronic map of the site (site basemap), well locations,
ground-water elevations, contaminant source location, and model data (site parameters: e.g.,
hydraulic conductivity, effective porosity) . The user chooses the area of the basemap to include
in calculations and presentation of program output ("Plot Limits").
Calculate Ground-Water Flow Field
OWL uses the well locations and ground-water elevation data to run a linear regression in order
to calculate the ground-water flow field for each sampling event. The ground-water elevation
contours can be plotted on the site basemap.
Predict Contaminant Plume Shape and Size
OWL uses the input data to run the Domenico solution to predict contaminant transport/plume
development. The plume contours can be plotted on the site basemap.
Calculate Well Location Factors
OWL calculates and plots the Minimum Distance function and the WOLF values. The Minimum
Distance function shows the variation in well density across the site. The WOLF combines
monitoring well coverage and predicted contaminant concentration into a single term which can
be mapped. The WOLF can be used to decide where additional monitoring wells could be placed,
and which existing monitoring wells might be providing redundant data.
25
-------
4.1 Site Selection and Characterization
The first step in using OWL requires gathering and analyzing site data to ensure that the site
characteristics meet OWL requirements. As explained in the Introduction, the OWL program is a
site screening program for use in helping to determine possible locations for new monitoring wells,
and assessing the usefulness of existing monitoring wells. OWL provides only a cursory assessment,
and is meant to replace neither professional judgment nor more advanced mathematical models
for use in assessing monitoring well locations. It should only be applied to sites which possess the
characteristics outlined in the Introduction Sections 1.3 and 1.5, and in Appendix A: Field Case Study Site
Characterization Data Collection and Evaluation for Application of OWL.
The data needed for using OWL include:
an electronic map ("basemap") of the site (See Appendix C for basemap specifications)
well locations based on a rectangular grid system (x,y) in an acceptable spreadsheet (see
Appendix D for spreadsheet formats). These may be easting, northing, UTM, State Plane or from
a user-defined grid system based on basemap scale bar dimensions (see Appendices A and C for
more details)
ground-water elevation data for each well at each sampling date in an acceptable spreadsheet (see
Appendix D for spreadsheet formats)
contaminant source width transverse to ground-water flow (if free product is present, may be
estimated as diameter of circle circumscribing free product area, otherwise can be estimated
based on soil or ground-water concentrations)
contaminant source concentration (dissolved concentration near the source)
contaminant retardation factor
contaminant half-life
hydraulic conductivity
effective porosity
longitudinal dispersivity
transverse dispersivity
The contaminant source width and contaminant source concentration should be measured at the site. The
contaminant retardation factor may be estimated from site data, or estimated using values from literature
references for the contaminant and the site aquifer matrix. Contaminant half-life should be estimated
from site data. Hydraulic conductivity should be determined from site data. Effective porosity may
be estimated using literature reference values for the site aquifer matrix (e.g., gravel, sand, silt, clay).
Longitudinal dispersivity and transverse dispersivity are usually estimated using literature reference
values or simple relationships; e.g. longitudinal dispersivity = 10% of plume length, and transverse
dispersivity = 10% of longitudinal dispersivity.
26
-------
4.2 Opening OWL
OWL can be started by choosing OWL in the Program menu section of the Start menu.
liEHSSH?"
iHfflH
OWL opens displaying a "Project Steps" window, a "Map Layers" window, and a "Message" window
(see Section 3.0 Program Layout And Screen Appearance}.
The Project Steps window assists the user to proceed in an orderly fashion; each consecutive step in using
the program is presented in order in the Project Steps window, with convenient checkboxes to allow the
user to initiate each step, and keep up with which steps have been finished.
The Map Layers window shows the user which "layers" (overlays; additions to the site basemap)
have been activated. For instance, when monitoring well locations are plotted on the basemap, the
locations are part of an overlay which can be turned on and off (made visible on the map, or invisible)
by checking or unchecking the "" checkbox in the Map Layers window. The site basemap is
an electronic copy of a site map which can be displayed in the OWL program. An example site map
(sitelbasemap.bmp or sitelbasemap_acadl4.dwg) comes with the program for use in the tutorial; the user
can import maps for their sites from many GIS programs, or they can scan a paper map and save it as a
bitmap or TIP file for importing into OWL.
The Message window shows the user messages about what the OWL program is doing. For example,
when the user opens the Case Study/Tutorial example project file, the Message window shows the
message "Opened project file:sitel.prj."
4.3 Opening a New Project File
After OWL is started, click on the "" item in the Menu Bar.
27
-------
%| Optimal Well Locator (OWL) 1
File Edit View Run Tools Win
New Project Ctrl+N
Open Project.,, Ctrl+O
Close Project Ctrl+W
5ave Project Ctrl+5
Save Project As.,,
Open Base Map...
Export Model Results
Print Ctrl+P
Exit
A drop-down menu will appear. Click on "." A message box will appear.
Assign Project File
2SJ
In order for OWL to handle data entry properly, one needs to
assign a project file. It is best that you save the project file
in the same directory as other files (sitemap, well and water
level data) you will need for this project. This will also
ensure easy transfer of all relevant files of a project to
another computer/file location later,
Note that you will still need to click [5ave Project File] button
(or choose Save option under File menu) frequently to save
your data to hard disk.
Click [OK] to assign a project file now.
OK
Cancel
After reading the message, click on "." A Save Project File dialogue will appear.
save Protect File
Savejn:
sltel
_JTemp
_Sl Data
jj «- (t)
1
*]
Save
! Save as type: j Project file (*.prj)
Cancel
Double-click on the "Sitel" folder to open that folder, type in the desired file name (sitel.prj is suggested)
and click "."
28
-------
4.4 Setting the Measurement Units
Click on the checkbox beside "" in the Project Steps window.
1. Input data
0 1.1 Set units
[I] 1.2 Base map
H 1.3 Map scale
A dialogue box will appear.
Set Units
Select basic units here. These units
will be used for entering/displaying
data throughout the project.
Distance units
Time units
Concentration units
Help
Click on the drop-down menu for each item (",
-------
Set Map Scale
Calibrate map from
I Two points
two calibration points [preferred]
Image coords Real coordinates (ft)
X Y X Y
Locate point 1
Locate point 2 592 F
0
640
600
Calibrate map to real coordinates
OK
Cancel
Help
Two Calibration Points
For the tutorial, choose the default choice "." Click "." A
crosshair cursor will appear and the "Set Map Scale" dialogue box will be temporarily hidden behind
the Map Window. Move the crosshair cursor down to the origin at the bottom left corner of the basemap.
Click on the origin. The "Set Map Scale" dialogue box should reappear with the image coordinates
X: +50.0000 and Y:+576.000. Please note that there may be some slight variability in these numbers
depending on how you placed the crosshair, and it is not critical to reproduce them exactly. Next, enter the
real site coordinates X=0 and Y=0 into the Real Coordinates X and Y locations in the dialogue box.
Click "." The crosshair cursor will appear again and the dialogue box will be
temporarily hidden behind the Map Window. Move the crosshair cursor up the top right corner of the
basemap and click on the tickmark just below the top right corner of the map. The "Set Map Scale"
dialogue box should reappear with the image coordinates X: +592.0000 andY+82.000. Please note that
there may be some slight variability in these numbers depending on how you placed the crosshair, and it
is not critical to reproduce them exactly. Next, enter the real site coordinates "X = 640" and "Y = 600"
into the Real Coordinates X and Y locations in the dialogue box.
Click "" and "." The X-Y coordinate numbers shown at
the bottom right of the Map Window are now calibrated to the actual site coordinates on the map instead
of the image coordinates.
A Scale Bar
If desired, the user may also register the map to its field units using the scale bar on the map if available.
This method is not recommended because it is less accurate than the two point calibration. There are
several pop-up boxes which warn the user about the lack of accuracy.
30
-------
Set Map Scale
Calibrate map from | a scale bar (less accurate)
r Map origin-
Locate origin , ...
mi^J Image coordinates
[50
r Scale bar-
LocatepointlJ Image coordinates
J-ocatejMinljJ |mage coordinates
Enter distance between the points
Calculated scale 100 pixels = |H76470588 ft
X Y
J430 [548
fssT
iT2cT
jCalibrate map to real coordinates!
OK
Cancel
Help
To use this method, first click on the checkbox beside "
-------
Click "" and "." The X-Y coordinate numbers shown at
the bottom right of the Map Window are now calibrated to the actual site coordinates on the map instead
of the image coordinates.
4.7 Import Well Location Data
Click on the checkbox beside "" in the Project Steps window. The first of a series of
dialogue boxes will appear; the first box describes the data format for the Well Location data.
)0
0MW-21
0MW-2f
OMW-26
/-27O
0MW-2E
- Introduction
For importing into OWL, well data must be prepared using
| spreadsheet software (Lotus 123, Microsoft Excel) in the
format illustrated below. Please refer to the OWL
documentation for more information.
Well ID column
X and Y coordinate columns
B
J\A
NODE EAST NORTH
MV-1 250.135 465.635
MV-2 285.44 428.26
HV-3 313.5 332.465
Well names must be in text format. Use numeric values
forX,Y coordinates (No missing values or text allowed).
Help
| hJext
Cancel
The sample Well Location file for the Case Study/Tutorial is in the correct format, so click "" to
go to the Well Locations Step 2 dialogue box, where the user can select the file format and select the data
file.
Well Locations Step 2: Specify File
Specify well data file
If you do not have the data file ready, use spreadsheet or
database software to prepare the data in the format
illustrated in Step 1. Acceptable file formats are shown in
the "Select a file format" list below.
Click to select the file format of the data file first. Then
select the file using the folder button at the lower-right.
Select a file format (Excel 97/2000/2002 J
Select file by clicking on the folder button at the right.
|well_locs.xls
Help
Next > I Cancel
32
-------
The Case Study/Tutorial sample well location data file is in Excel 97 format, the default choice, so leave
that selection unchanged in the first drop-down menu. Click on the "" button in the dialogue box
to bring up a dialogue box for finding the well location data file. Browse to the correct folder if necessary
(the Case Study/Tutorial sample data file should be in the sitel folder), select the well_locs.xls file, and
click "." Click "" in the Well Locations Step 2 dialogue box.
The next dialogue box in the series allows the user to specify the fields in the spreadsheet or database
which contain the well location data.
Well Locations Step 3: Specify
S3:
ui ss
i~- CM C't
3= 3 S
\
Field Names
Spreadsheet files may contain several [worksheets of
data. Please select below the sheet for well locations
data. Also identify the columns for the data parameters.
Select sheet: I'welllocs'S
WELL ID column/field: WELLID
X coordinate column/field: X
Y coordinate column/field: |Y
Number of records found: 15
Read Data
Help
feel-
Cancel
The spreadsheet "sheet" (worksheet) that contains the Case Study/Tutorial data is 'well Iocs '$, so leave
the first drop-down menu unchanged. Likewise, the Well ID column/field is WELLID, the X coordinate
column/field is X, and the Y coordinate column/field is Y, so leave those drop-down menus unchanged.
Click "" so that OWL will read the spreadsheet and bring the data into the program for use.
A message should appear at the bottom left of the dialogue box indicating, "File is successfully read."
The number of records found should be 15. Click "."
The final dialogue box in the Well Locations data series should appear, allowing the user to view the
imported data.
'ell Locations Step 4: Data
2
-------
The user cannot edit the well location (or ground-water elevation) data in OWL. If data changes,
corrections or additions are needed, the user should edit the data in the original program used for the data
(e.g., Excel). The user should make changes to the data file using the original program, save the changes
to a file, and then read the file using OWL to incorporate the changes.
If, while editing the data file, problems such as "file access-violation" errors are encountered, the user
should:
(1) close the OWL session (after saving the project),
(2) modify the data file using another software program (e.g., Excel), and save the data file,
(3) close the other program (e.g., the program used to modify the data), and
(4) start an OWL session, and read the modified versions of the data files.
The widths of the data columns shown in the dialogue box may be adjusted by clicking and dragging the
column borders to make editing and viewing easier.
Note that in the Well Locations Step 4 Data dialogue box, the user can:
Scroll horizontally if the spreadsheet exceeds the window width
Scroll vertically if the spreadsheet exceeds the window height
Change column width by clicking and dragging on the line between columns
Change row height by clicking and dragging on the line between rows
Click "
-------
Click "." The second dialogue box in the series allows the user to select a file format and a file
for the ground-water elevation data to be used.
Water Level Data Step 2: Specify File
X Microsoft Excel
File Edit View
Arial
Specify water level data file-
If you do not have the data file ready, use spreadsheet
software to prepare the data in the format illustrated in
Step 1. Acceptable file formats are shown in the "Select
a file format" list below.
Click to select the file format of the data file first. Then
select the file using the folder button at the lower-right.
Ai
Select a file format (Excel 97/2000/2002 J
Select file by clicking on the folder button at the right.
Ql
wl data.xls
Help
< Back
Next>
Cancel
The sample water level data file in Excel 97 format, the default choice, so leave that selection unchanged
in the first drop-down menu. Click on the folder button in the dialogue box to bring up a dialogue box for
finding the well level data file. Browse to the correct folder if necessary (the sample data file should be in
the sitel folder), select the wl_data.xls file, and click Open.
Click "" in the Water Level Step 2 dialogue box. The third dialogue box in the series allows the
user to select the worksheet in the spreadsheet which contains the ground-water elevation data to be used.
Water Level Data Step 3: Specify Fieli
o
I CO
i- Field Names-
O 01 CJ
O O CM
O
a.
5
c
10
(N (N (N £
_ CD
*T CD
_ f
cn m
Spreadsheet files may contain several [worksheets of
data. Please select below the sheet for water levels
data.
Select worksheet: 'wl transpose's J*J
ID OO
rj o
Number of wells: |10
*- CM
Number of sample dates: |7
Number of records found: 7 Read Data
Help
< Back
Cancel
35
-------
The worksheet should be 'wl transpose $' in this Case Study/Tutorial. Click Read Data so that OWL will
read the spreadsheet and bring the water level data into the program for use. A message should appear at
the bottom left of the dialogue box indicating "File is successfully read." The number of wells should be
10, and the number of sample dates should be 7.
Click "." The final dialogue box in the series will appear, showing the imported water level data.
fater Level Data Step 4: Data
t--
01
Qi "
o
OJ O
° i "
1 ^
00 |£ £3
22
< iu ^
i^
T"
- OJ
5
01
p
tN
C"4
tN
ID
22
IN
(N
n
ID
(N
C-4
CJ
01
01
r-4
03
o
o
-r
01
ID
C3
ID
ID
TO
01
O4
03
§
oi
un
01
01
fvi
(D
01
03
22
CD
0
0
0
0
h
fj
r
r
c
h
DAT E| MW-1
'25/1 994 j 23.96
12/1 995 i 22.94
^72§7TI| 2035^
MW-2|
22. 998 1
22.028]
MW-3
21.966
21.012
^J2Z22| 21.25
TT54751 21529
-9999!
-9999!
-9999!
llJJ
Help < Back
21.43
T&78
17.73
MW-4
22.39
21.82
22.14
21749
22.39
20.77
19.77
MW-E
23. 7E
22. E
23.57
22.572E
24. 5E
"231
20.17
Jj
i Finish !j
Zancel
Note that the convention in OWL for denoting missing water level data is -9999. The user cannot edit
the ground-water elevation (or well location) data in OWL. If data changes, corrections or additions are
needed, the user should edit the data in the original program used for creating the data (e.g., Excel). The
user should make changes to the data file using the original program, save the changes to a file, and then
read the file using OWL to incorporate the changes.
If, while editing the data file, problems such as "file access-violation" errors are encountered, the user
should:
(1) close the OWL session (after saving the project),
(2) modify the data file using another software program (e.g., Excel), and save the data file,
(3) close the other program (e.g., the program used to modify the data), and
(4) start an OWL session, and read the modified versions of the data files.
Note that in the Water Level Data Step 4 dialogue box, the user can:
Scroll horizontally if the spreadsheet exceeds the window width
Scroll vertically if the spreadsheet exceeds the window height
Change column width by clicking and dragging on the line between columns
Change row height by clicking and dragging on the line between rows
Note also that the water levels data must be in a specific format, without any other data on the same
worksheet. No other columns of data can be present on the same sheet. The data must be in the format as
shown in Appendix D.
Click "."
36
-------
4.9 Set Plot Limits
Click on the checkbox beside "" in the Project Steps window. The Plot Limits dialogue box
will appear.
Plot Limits
Map extents | Water level contours | Concentration contours
-Map Extents
Specify the map extents for all calculations and display.
Borders of both water level and concentration contours must
fall within these extents.
Intially, these extents will be assigned to the physical extents
of the map you have opened.
Left (-60.4
Right (650.7
Bottom |-29.1
(698.2
Help
The Plot limits dialogue box has three tabs: Map Extents, Water Level Contours, and
Concentration Contours. The Plot Limits dialogue box allows the user to set the area of the
basemap where results of OWL calculations are displayed.
Note that OWL uses all the available well data for regression calculations, regardless of any plot
limits that are chosen from any of the three tabs in the Plot Limits dialogue box. That is, all the
data that have been read into OWL from the Well Location and Water Level Data dialogue boxes
are used in OWL calculations. The various plot limits options affect only the display of results on
the basemap.
The user should be certain that the chosen plot limits include all the wells for which sampling
data is imported into OWL. If the Plot Limits area does not include all wells for which data is
used in the calculations, errors could occur. Also, for display purposes, if the well falls outside
the raster map, it may cause an error.
Map Extents
The Map Extents tab in the Plot Limits dialogue box allows the user to "Specify the map extents for all
calculations and display." By typing in the XY coordinates of Left, Right, Top and Bottom locations of
a rectangular area on the basemap, all calculations and display of water level contours, plume contours,
Minimum Distance contours and WOLF contours are constrained to that chosen area on the basemap.
The XY coordinates of the desired locations can be determined by placing the cursor on the desired
location on the basemap, and noting the XY coordinates displayed at the lower right of the Map View
window.
37
-------
For this Case Study/Tutorial, type in:
Left:
Bottom:
0
0
Right:
Top:
640
692
k
J
Water Level Contours
The Water Level Contours tab in the Plot Limits dialogue box allows the user to select a rectangular area
of the basemap where water level contours will be plotted. (The data for these contour lines come from
the water levels worksheet imported in the Water Levels window.) The area selected must be within the
area previously selected under the "Map Extents" tab of the Plot Limits window.
Plot Limits
.x]
Map extents Water level contours Concentration contours
-Water Level Contours -
Select the extents of the border for plotting water level
contours. These values must lie within the map extents
you chose before.
Left (0
Bottom (0
Right (640
Top 1(82"
OK
Help
For this Case Study/Tutorial, type in:
Left:
Bottom:
0
0
Right:
Top:
640
692
J
Concentration Contours
The Concentration Contours tab in the Plot Limits dialogue box allows the user to "Select the extents
used for creation of a virtual grid for superposition of transport model results." The user can select a
rectangular area of the basemap where a grid of points will be located, so that results of calculations
in the transport model (the Domenico model used in this program to calculate contaminant fate and
transport) can be plotted on the basemap. The transport model calculation results are the contaminant
concentrations and locations; these results are used to draw the estimated contaminant plume on the
basemap.
38
-------
In general, a smaller grid (i.e., a smaller rectangular area specified under the Concentration Contours tab)
would calculate faster.
Plot Limits
Map extents | Water level contours Concentration contours
i Concentration Contours
Select the extents used for creation of a virtual grid for
superposition of transport model results.
W iSet the limits the same as that for water levels.;
Leftfr
Right (640
Bottom fo T°P (692
Grid spacing along X and Y (3
Number of grids: X J214
Y [231
Draw Limits on Map
OK
Help
There is a checkbox so that the rectangle extents can be automatically set the same as that for the water
levels contours, as set under the "Water level contours" tab. For this Case Study/Tutorial, click the
checkbox.
The Concentration Contours tab in the Plot Limits dialogue box also allows setting the grid spacing along
the X and Y axis. The grid space is the length and width of one square in the grid, in whichever distance
units were selected from the Set Units dialogue box. The grid spacing is used by OWL to calculate the
number of grid nodes (points) along the X and Y axis on the basemap. Smaller grid spacings mean more
grid nodes. More grid nodes provide more detailed plume maps, and involve longer calculation times.
When the grid spacing is entered (and the user tabs away), the "number of grids" (the number of nodes or
points in the grid) on the X and Y axes is automatically listed below in the boxes. The user cannot type
data into the boxes where the "number of grids" is shown.
The grid spacing also determines the spacing of the lines that OWL uses to render the calculated
contaminant plume on the basemap ("Plume contours" checkbox in the Project Steps window). Grid
spacing should be chosen to provide about 200-300 grid lines along each direction in the virtual grid.
If the user changes the grid spacing after running the analytical solution ("Run analytical solution"
checkbox in the Project Steps window), the analytical solution must be rerun to use the new grid spacing.
For this Case Study/Tutorial, set the grid spacing to 3 ft. Tab down or click on the "Number of Grids: X"
data box, and the grid numbers will be automatically set at X = 214 and Y = 231.
Click on the "" button to show the chosen area on the basemap.
Click the "" button to close the dialogue box.
Click on the "" checkbox in the Map Layers window to display or hide the plot limit
rectangles.
39
-------
4.10 Inputting Model Data
Click on the "" checkbox in the Project Steps window. The Transport Model Input Data
dialogue box will appear.
Transport Model Input
Hydraulic conductivity p14 ft/yr
Effective porosity 0.45
Longitudinal dispersivity 30
Transverse dispersivity 13
Retardation factor |l
Simulate first-order decay
Solute half-life
Simulation time 6
Source width 30
Source concentration 125
Close
ft
mg/L
Help
The following site data are needed for calculation of the Domenico solution to contaminant transport (i.e.,
plume development). Note that the units used are based on the units chosen in the Set Units dialogue box.
Hydraulic conductivity
Effective porosity
Longitudinal dispersivity
Transverse dispersivity
Contaminant retardation factor
Simulation time (time allowed in the modeling calculation for the contaminant to move from
the source to form the plume)
Source width (width of source area transverse to ground-water flow)
Source concentration (dissolved concentration near the source)
For this Case Study/Tutorial use the following values:
Hydraulic conductivity (K) = 514 ft/yr
Effective porosity = 0.45
Longitudinal dispersivity = 30 ft
Transverse dispersivity = 3 ft
Contaminant retardation factor = 1
Simulation time = 6
Source width = 30
Source concentration =125
40
-------
Note that the contaminant source width and contaminant source concentration should be
measured at the site. The contaminant retardation factor may be estimated from site data, or
estimated using values from literature references for the contaminant and the site aquifer matrix.
Contaminant half-life should be estimated from site data. For this Case Study/Tutorial we assume
that no degradation is taking place at the site. Hydraulic conductivity should be determined from
site data. Effective porosity may be estimated using literature reference values for the particular
site aquifer matrix. Longitudinal dispersivity and transverse dispersivity are usually estimated
using literature reference values or simple relationships; e.g. longitudinal dispersivity = 10% of
plume length, and transverse dispersivity = 10% of longitudinal dispersivity.
OWL allows the user to input values for several site parameters (e.g., conductivity, dispersivity,
as listed above). The user should determine a range of these values that convey the variability
in the site parameters, and make model runs with several values of each parameter to determine
how the model results are affected by variations in the site parameters. Hydraulic conductivity,
solute (contaminant) half-life, source width and source concentration are examples of factors
likely to significantly affect model results. In particular, the user should input a range of values
of conductivity (k) to determine the sensitivity to k, because site values of k are likely to vary
more than any of the other site parameters used in the model. Using a range of values for k will
cause the model to calculate different plume lengths. Knowledge of potential variability of plume
lengths should help the user assess the impact of program results on site decisions. For example,
this would be important if the program were used to locate guard wells because the length of the
plume would be expected to vary significantly due to differing k values in the subsurface.
Simulation time is the time that will be simulated for contaminant plume travel from the source.
The simulation time starts at the date of the sampling data used (the sampling round), continues
for the selected time, assumes that there is no plume existing before the sampling date used, and
assumes that the ground-water conditions prevailing at the time of the sampling data continue for
the full simulation time. The estimated plume shape at the end of the simulation time is displayed
on the basemap.
The Solute Half-Life is the attenuation half-life of the soluble contaminant that makes up the
plume. The half-life is the time that it takes for contaminant concentration to decrease by one
half.
The Source Width specifies the width of the source transverse to ground-water flow. The source
line is always arranged perpendicular to the calculated ground-water flow direction for each
sampling date. As ground-water flow varies over the various sampling periods, the source line
varies in its orientation to the map coordinates.
4.11 Choosing the Source Location
Click on the "" checkbox in the Project Steps window. The Source Location dialogue
box will appear.
41
-------
Source Location
Click the center of the source location on the base
map or enter its X,Y values below.
Y 475
Close
Help
The Source Location dialogue box allows the user to set the location of the contaminant source on the
basemap. The user can click the "
-------
Regression output file: results.out
£Q:
46
*-~. 17 | riw-i
,^ i The well location is 250.1352222000000 465.6350000000000
\^ 17 The head value listing for all dates is 23.96000000000000 22.94000000000000
lt 1 23.43000000000000 22.38750000000000 23.64000000000000 22.67000000000000
" 17 20.35000000000000
33 £ M₯-2
17 The wel1 location is 285.4422222000000 428.2600000000000
.£ i The head value listing for all dates is 22.99800000000000 22.02800000000000
^ 17 22.22000000000000 21.54750000000000 -9999.000000000000 -9999.000000000000 -
%1 i 9999.000000000000
« 17 MW-3
=1 3 The well location is 313.5000000000000 332.4650000000000
^ 17 The head value listing for all dates is 21.96600000000000 21.01200000000000
, 21.25000000000000 20.52900000000000 21.43000000000000 18.78000000000000
^ 17.73000000000000
S m~4
JJ The well location is 338.5700000000000 450.1100000000000
*'J The head value listing for all dates is 22.39000000000000 21.82000000000000
lt 22.14000000000000 21.49000000000000 22.39000000000000 20.77000000000000
[ " 19.77000000000000
MW-5
The well location is 291.7655555000000 511.4050000000000
For each well, re suits.out gives:
1) The well location coordinates ("The well location is...") and
2) The ground-water elevations data for each date when the well was sampled ("The head
value listing for all dates is...")
For each sampling date, results.out gives:
1) The sampling date ("The date for this regression is..."),
2) The regression equation with coefficients ("The regression equation is...")
3) The correlation coefficient to show how well the regression fits the data; i.e., the R-square
value ("the regression coefficient for the equation fit is...")
4) The value for the calculated ground-water gradient ("The magnitude of the ground-water
gradient is...") and
5) The difference in degrees of ground-water flow direction from the north ("The
direction of the ground-water flow from north is...).
Regress.out gives a summary table of the regression results:
1) The sampling date for the data used in each regression equation ("date"),
2) the x coefficient of each regression equation ("xcoeff')
3) the y coefficient of each regression equation ("y coeff')
4) the constant of each regression equation ("constant")
5) the correlation coefficient of each regression equation ("R2")
6) the value for the calculated ground-water gradient ("mag") and
7) the difference in degrees of ground-water flow direction from the North ("dir").
43
-------
Final, out gives:
Number of wells (nl), Number of dates (k)
The following information is given for each well:
Well name, X, Y coordinates of location
Water elevations for all the sample dates
The following results are shown for each date of measurement:
Sample date
Regression matrix details
Matrix dimension (mm, same as number of wells)
Final adjusted matrix
Coefficient matrix
Right-hand-side matrix (rhs)
Rank of matrix
Parameters
Residuals
Variance of residuals
Regression coefficient
Hydraulic gradient
Flow direction (degrees from the north)
Click the "" in the upper right corner of each output file window to close each window.
The output files are opened in the user-specified text editor during OWL installation (for example,
Notepad for small text files, and Wordpad for large text files). The user can check which programs are
specified (and change the specified programs) by clicking on " Options>" on the OWL menu
bar.
4.14 View Water Level Contours
Click on the "" checkbox in the "Project Steps" window to bring up a dialogue
box allowing the user to see a spreadsheet view of the linear regression results, and to set and view the
water level contours on the basemap.
Water Level Contours
[Hate
j 1/25/1994
| 2/12/1995
| 0/30/1998
12/4/1998
2/21/1998
UJ=L
Xcoeff
-0.01355
-0.007342
-0.007783
-0.005596
0.0022
Select a sample date
Contour leve
Line w
) Show contour aut
Ycoetf
0.007089
0.009284
0.01085
0.01046
0.02145
Const
23.92
20.25
20.09
18.73
12.89
irst in the above matrix
s: Min |l5 Ints
dth p
smatically when a date is
t 11
H2
0.9813
0.3812
0.9644
0.3726
0.7728
srval [i
selected
|
Magnitud
0.0153
0.01184
0.01335
0.01186
0.02156
X
Directio ^ii
117 ||
141 I
144 I
151 I
-174^JI
Jj 1
_ ^ [24 |
Show Contour j |
Clear Contours | j
_Close_J
Help |
44
-------
The first frame at the top of the dialogue box shows a spreadsheet-like view of the X coefficient (Xcoeff),
Y coefficient (Ycoeff), constant (Const), correlation coefficient (R2), magnitude, and direction for each
sampling date.
The regression equation is:
{GW head} = [Xcoeff] {X coordinate} + [Ycoeff] {Y coordinate} + Const
The term "R2" gives an indication of how closely the regressed plane fits the ground-water elevation data.
Values of R2 closer to 1.0 show a better fit of the regression to the sampling data. "Magnitude" gives
the calculated average change in ground-water elevation along the flow direction; i.e., the ground-water
gradient. "Direction" gives the angle of ground-water flow in degrees clockwise from the north (north is
0 degrees; OWL assumes that north is at the top of the screen on the basemap).
The R2 term is the square of the Pearson Product Moment Correlation Coefficient. It can be thought of as
the portion of the total variation in the data that can be explained by the regression. The values of R2 in
the case study vary from 0.7728 to 0.9813. These are particularly good fits. The regression explains from
77% to 98% of the variation in the data. These data fit the assumption in OWL that the water table is a
plane; at other sites, the fit to the assumption may not be as good. Appendix E provides recommendations
for editing the data set when R2 is low.
If necessary, horizontal and vertical scrollbars are present so that all the results can be viewed by
scrolling. Column width and row height of the spreadsheet view can be changed by clicking and dragging
on the line between the columns or rows, in order to fully view the contents of each cell.
Just below the frame containing the spreadsheet view in the "Water Level Contours" dialogue box is a
frame showing the "Mapping Results." This part of the dialogue allows the user to show the calculated
ground-water level contours on the basemap, for a chosen sampling date. The user can select a sampling
date in the spreadsheet view by clicking on a cell in the desired row, or by moving up and down among
the sampling dates using the up and down arrow keys on the keyboard.
The range of ground-water levels to be plotted as contour lines on the basemap can be entered ("Contour
levels"); the minimum value of ground-water head to be plotted as a contour line on the basemap (Min)
and the maximum value of ground-water head to be plotted (Max) can be typed into the entry boxes. The
contour interval can also be entered; this is the difference in ground-water elevations between successive
contour lines. The width of the contour lines to be plotted can be entered; wider lines are more easily
visible on the basemap. The width is entered in screen units (pixels). That is, when the width is 2, lines
that are two pixels wide are displayed on the map.
The default contour interval is 1 foot or 1 meter, depending on whichever distance unit was selected
from the Set Units dialogue box. If the hydraulic gradient across the site is very flat, a contour line may
not displayed at a contour interval of 1. In this case set the contour interval to 0.1 foot or 0.03 meter.
Although depth to water data are often read and recorded with a precision of 0.01 foot with respect to the
top of the well casing, the survey of the elevation of the well casings is usually accurate to 0.1 foot.
A checkbox is provided to show the contour lines on the basemap automatically when a sampling date is
chosen by clicking on the spreadsheet view.
45
-------
Two buttons are provided in the "Water Level Contours" dialogue box; one displays the contours for the
chosen sampling date and contour parameters on the basemap ("Show Contour"), and the other button
removes the display of contours from the basemap ("Clear Contours").
Note that the contours are plotted as though the ground-water surface was a plane; ground-water
flow at specific locations on the actual site may differ significantly from the flow directions
indicated by the contour lines shown in this program. These differences in flow could be
caused by ground-water recharge or discharge, or other processes or structures that could cause
deviations from a planar ground-water surface.
Click on the "" to bring it to the front for viewing the contours without the dialogue box
covering the map. The dialogue box can be brought back to the front by clicking on "" (on
the Menu Bar), and clicking on ""
The differences in ground-water flow direction and gradient at the different sampling dates may be
examined by choosing sampling dates and plotting the contours.
Click the "" button to close the dialogue box.
The basemap with contour lines may be printed when the basemap window is active. Click ""
(on the Menu Bar), "," enter a map title if desired in the dialogue box that appears, and click
"" to print. If Print is selected when the basemap window is not active, an error box appears.
Click "" on the error box, click the basemap to make it active, and choose "" from the
Menu Bar again.
4.15 Transport Run
Click on the "" checkbox in the Project Steps window to run the Domenico
model calculations for solute transport in ground water. The model calculations are based on the well
46
-------
location data, water level data, and source location data entered earlier, and the Transport Model input
data entered after clicking on the Model Data checkbox in the Project Steps window. A message will be
shown in the status bar at the bottom right of the OWL window indicating the model was successfully
run ("Analytical simulation finished"). Appendix B gives details on the implementation of the Domenico
solution used in OWL.
4.16 Export Model Results
Click on the "" checkbox in the Project Steps window to save the model results
in a text file suitable for exporting to other programs. A Select File for Model Results dialogue box
appears, allowing the user to save the model results in a text file to the desired directory. The default
filename is OWLResults.out, but the user can type in a new name if desired. If a file by that name is
already in the directory, a warning message is displayed. The file may be quite large, depending on the
amount of sampling data input to OWL as well locations and ground-water elevations.
4.17 View Plume Contours
Click on the "" checkbox in the Project Steps window to make the View Plume
Contours dialogue box appear.
'lume Contours
- From model output
Concentration ranges from all the
model simulations are:
Cmin:1.1E-18
Cmax:12Q.397
u enerale levels
Contour levels: M in J|0.001
Max |100.
Number of levels |5
Generate Levels
W Use Log scale
Cmin Cmax
0.001 0.01
0.01 0.1
0.1 1.
1. 10.
10. 100.
Color
W Legend Locate
Date: 1 1/25/1 934 j-J
Line width )3
Clear I Draw Plume
Close I Help
The first frame in the "Plume Contours" dialogue box (on the top left hand side of the box) "From model
output" shows the minimum and maximum calculated contaminant concentration ranges from all the
model simulations. These values are automatically placed in the "Generate levels" part of the dialogue
box discussed below.
On the bottom left hand side of the "Analytical Model Results" dialogue box ("Generate levels") the user
may enter minimum and maximum values for the solute (contaminant) contour levels to be contoured on
the basemap, and the number of contour levels desired to present. The "Number of levels" is the number
47
-------
of contours that the contaminant concentration range is to be divided into to be displayed as contours on
the basemap. The maximum number of levels is five.
An option is provided to use a log scale for the plume contours. Clicking the "
-------
Click on the "" button to display the plume contours (and legend, if chosen) on the
basemap. Click on the "" button to clear the currently-drawn plume (and legend, if chosen).
The basemap with plume contours (and legend, if chosen) may be printed by choosing ""
from the Menu Bar. The basemap window must be active, or an error message will appear. When Print
is selected, a dialogue box will appear allowing the user to enter a map title to be printed on the bottom of
the page.
Click the "" to close the dialogue box.
The user can step through the plots of each plume to view the impact of the change in ground-water
flow direction and magnitude. If there is a large variation in the ground-water flow field overtime, the
difference in the plumes created for the same source concentration and time of travel can be striking,
as it is in this case study. In addition, the user can view the "" plume, which is a composite
of all the possible plume migration paths. This composite plume provides the user with all the possible
locations where contamination may exist at the site, given the historical ground-water flow variation.
For this case study, the composite plume is predicted to be present over a large region and the existing
monitoring wells clearly cover only a small portion of that region.
4.18 View Minimum Distances
OWL provides the capability to calculate distances from each grid point to the nearest well and present
these distances as contours on the basemap. These contours give the user a visual indication of the
49
-------
monitoring well coverage across the site. The minimum distance data will be used in the next section to
calculate a Well Optimal Location Factor (WOLF), to indicate locations where new monitoring wells
might be needed and where existing wells may be redundant.
Click on the "" checkbox in the Project Steps window. The "Minimum
Distance Contours" dialogue box appears.
\> ' " . , '
1 1 Dmin
Range of distances to the g g
nearest well are i i __ n
bu. U
DISTiiiiM. 30.93 ...-,-, ...
DISTmax: 309.92 I 1F|gn
r i i 1 ^40 0
Select contour intervals
r Lege
Contour levels: Min |0
Number of levels |5 L
Generate Levels | | | Qearj
Dmax Cole
60.0
120.0
180.0
240.0
300.0
nd Loca
ine width |3
*1
>r
tej
-
Draw contours 1
II
Close I He
Ip 1
The first frame (on the top left hand side of the dialogue box) in the "Minimum Distance Contours"
dialogue box is "Range of Values." This frame reports the range of distances from the node points on
the superimposed grid to the nearest well; the minimum distance (DISTMin) and maximum distance
(DISTMax). This range is for locations within the map limits (as set before in the Plot Limits dialogue
box). Zero and the DISTmax values are automatically entered into the "Generate levels" Min and Max
part of the dialogue box discussed below.
On the bottom left hand side of the "Minimum Distance Contours" dialogue box ("Generate levels") the
user may enter minimum and maximum values for the distance contour levels to be contoured on the
basemap, and the number of contour levels desired to present. The "Number of levels" is the number of
contours that the minimum/maximum distance range is to be divided into to be displayed as contours on
the basemap. For this Case Study/Tutorial, enter Min = 0.0, Max = 300, and Number of levels = 5. Click
the "" button to use the chosen contour min, max and levels.
The colors used for contours can be changed by the user, if desired. The user may click the ""
column in the spreadsheet view on the top right hand side of the "Minimum Distance Contours" dialogue
box to show the "Select a color" palette. Click the appropriate cell under the Color column in the
spreadsheet view, then click the desired color in the "Select a color" palette. The selected color will
appear in the spreadsheet cell. Continue clicking a spreadsheet cell and a color until colors are selected
for all the levels shown in the "Minimum Distance Contours" dialogue box. Click on the ""
button to shut the "Select a color" dialogue box.
50
-------
OWL provides an option to place a legend showing the minimum distances associated with colors of the
contours. Check the box labeled "
-------
Click the "" button to close the Minimum Distance Contours dialogue box.
The user can look at the color contours of "Minimum Distance" and develop an understanding of the
existing well coverage at the site. For this case study, the dark blue circles highlight the region where
a monitoring well is close, within 60 ft. As you move away from the wells, the contour colors indicate
that there is less and less coverage by existing monitoring wells. At some point, it is clear that there are
regions where the nearest monitoring well is so far away that it represents essentially no coverage. This
"Minimum Distance" is useful by itself, because the user can begin to have a meaningful measure of
"void space" where the site lacks monitoring well coverage.
4.19 Well Optimal Location Factor (WOLF)
OWL provides the capability to calculate a number that combines the distance from each grid point to the
nearest well (minimum distance) and the average concentration from the composite plume, and present
these numbers as contours on the basemap. This number, known as the Well Optimal Location Factor
(WOLF), is based on the following calculation:
WOLF = (average concentration at each grid node) * (square of the minimum distance to the
nearest well)
WOLF contours give the user a visual indication of how the predicted contaminant concentration at a
location as a function of the change in ground-water flow overtime can be combined with monitoring
well coverage to highlight areas of concern in a monitoring well network. Such areas would be locations
where monitoring well coverage is high, but contamination concentration is forecast to be low. Or vice-
versa, contaminant concentration is predicted to be high, but monitoring well coverage is low. Thus the
WOLF becomes the yardstick, by which the user can evaluate the location of existing monitoring wells
and select locations for new monitoring wells.
Click on the "" in the Project Steps window and the WOLF Contours dialogue
box appears.
VOLF Contours
r
r rom model output
Range of WOLF factor values
from the model simulations are:
WOLFmin:3.0E-14
WOLFmax: 9.6E+4
i- iii
Contour levels: Min j|l.
Max J1.0E+5
Number of levels p
Generate Levels
P" Use Log scale
Export grid data
Frnin Fmax
1. 10.
10. 100.
100. 1000.
1000. 1.0E+4
1.0E+4 1.DE+5
Color
L., ...
W Legend Locate
Line width R
Clear J Draw contours
Close Help
52
-------
The first frame (on the top left hand side of the dialogue box) in the WOLF Contours dialogue box is
"From model output." This reports the range of calculated WOLF values from the model simulations, the
minimum WOLF (WOLFMin) and maximum WOLF (WOLFMax). This range is for locations within
the map limits (as set previously in the Plot Limits window). The WOLF for a grid node is the average
predicted contaminant concentration at that grid node multiplied by the square of the minimum distance
from the node to the nearest well. The WOLFmin and WOLFmax values are automatically entered into
the Min and Max part of the "Generate levels" frame discussed below.
On the lower left of the "WOLF Contours" dialogue box ("Generate levels") the user may enter minimum
and maximum values for the WOLF contour levels to be contoured on the basemap, and the number of
contour levels desired to present. The "Number of levels" is the number of contours that the WOLFmin/
WOLFmax range is to be divided into to be displayed as contours on the basemap. An option is provided
to use a log scale for the WOLF contours. Clicking the "Use Log scale" checkbox causes the WOLF
contours to be divided based on a logarithmic scale rather than the usual linear scale. This option is
helpful when WOLF values vary over several orders of magnitude, as is usually the case. For this Case
Study/Tutorial, check the "
-------
The WOLF data may be exported in a text file by clicking on the "" button in the
lower left of the dialogue box. (There is no warning for overwriting a previously existing file of the same
name).
The basemap with WOLF contours (and legend, if chosen) may be printed by choosing File, Print from
the Menu Bar. The basemap window must be active, or an error message will appear. When Print is
selected, a dialogue box will appear allowing the user to enter a map title to be printed on the bottom of
the page. The file should be printed to a color printer.
Click the "" button to close the dialogue box.
The user can now look at the contours of the WOLF on the site basemap and begin to pinpoint areas of
concern for the performance of the monitoring well network. For this case study, using a time of travel
of six years, the red region in the WOLF plot indicates that there are high concentrations predicted based
on the variation in historical ground-water flow and low monitoring well coverage. This red region,
especially west of most of the existing wells, may therefore be a good place to consider putting a new
monitoring well. The WOLF values will change depending on the time of travel and continuous source
concentrations used to create the plumes. So the user can vary these values, rerun the plume migration
and see how the monitoring well network performs for different timelines. But the objective is still clear,
to assess the monitoring network by combining the predicted concentrations produced from the historical
variation in the ground-water flow field with the coverage of the existing monitoring wells.
54
-------
4.20 Project Data Management and Printing
The project file containing all currently entered data can be saved by clicking on ""
to bring up the Windows Save dialogue box. The project file can be saved under a new name by clicking
on "" to bring up the Windows Save as dialogue box.
To delete a file, the user should go to the hard drive location where they created the project file and delete
the project file name. For this case study, the user was asked to create the project files in C:\program files\
owl\data\ sitel\ directory. The case study name suggested was sitel.prj. If the user goes to this directory,
they can select the sitel.prj file and delete it from the hard drive.
Any of the basemap contour plots may be printed by choosing "" from the Menu Bar. The
map window must be active to print, or an error message will appear. When "" is selected, a
dialogue box will appear allowing the user to enter a map title to be printed on the bottom of the page.
The maps are best to view when printed in color.
55
-------
56
-------
5.0 Glossary
adsorption process by which molecules collect on and adhere to the surface of an adsorbent solid because
of chemical and/or physical forces
attenuation a lessening in concentration, degree, or mass
basemap an electronic map of a site, used in OWL for display of site data and results of OWL calculations
constant thickness aquifer aquifer thickness is assumed to be essentially uniform in the study area. For
example, the aquifer should not be twenty feet thick at one point, and one foot thick at
another point
coordinate system for use in OWL, a Cartesian coordinate system for describing the location of
monitoring wells at the site. For example, the Universal Transverse Mercator (UTM)
system, or the State Plane (state plane coordinate system (SPCS)) may be used by
surveyors to describe locations. Alternatively, the OWL user may describe well locations
in terms of an arbitrary Cartesian coordinate system based on a grid established on a site
map
coefficient of determination (R2) a statistic used in least squares regression to describe the goodness of fit
of an assumed model to the observed data. The coefficient of determination is the square
of the Pearson correlation coefficient. OWL calculates a coefficient of determination
to describe the fit of a linear plane to ground-water elevation data. An R2 close to one
indicates that the linear surface model explains almost all of the variation in the ground-
water level data. If the value of R2 is 0.500, the assumption that the ground-water surface
is a plane explains half the variation. As a general rule of thumb, elevation data sets
that produce a value of R2 that is less than 0.500 indicate that a linear plane is not an
acceptable model for the ground-water elevation surface
diffusion process by which ionic and molecular species move from a region of higher concentration to a
region of lower concentration
dispersion phenomenon by which a solute in flowing ground water mixes with uncontaminated water,
becoming reduced in concentration. Dispersion is due both to differences in water
velocity at the pore level and differences in the rate at which water moves through
different strata. Also refers to statistical measures of how widely a set of data vary
dispersivity property that quantifies dispersion in a medium
Domenico solution a mathematical model used to simulate multidimensional transport of a decaying
contaminant species. For use in OWL, the Domenico solution is modified to simulate
two-dimensional transport. See Appendix A for a detailed description
57
-------
first-order decay attenuation or degradation (i.e., of a dissolved contaminant) according to a first-order
rate law; the rate of a first-order reaction depends on the concentration of the reactant.
A chemical reaction in which an increase (or decrease) in reactant concentration results
in a proportional increase (or decrease) in the rate of the reaction. The equation can be
written as dc/dt = -kc, where c = contaminant concentration, and k is the rate constant
fully penetrating well well in which the screened length is equal to the saturated thickness of the aquifer
geological log a description of the underground features (depth, thickness, matrix, type of formations)
found while drilling a well
ground-water divide site location, on either side of which ground water flows in opposing directions
half-life time required for a concentration or mass of contaminant to diminish by one-half
head (elevation head) elevation of the ground-water table above a specified point, usually a standard
reference point such as mean sea level
homogeneous aquifer structure or composition of the aquifer matrix is uniform throughout the aquifer
hydraulic conductivity relative ability of soil, sediment, or rock to transmit water; a coefficient of
proportionality describing the rate at which water can move through a permeable medium
hydraulic gradient the change in total hydraulic head divided by the change in distance in a given
direction
hydraulic head sum of the elevation head, the pressure head, and the velocity head at a given point in an
aquifer. Also referred to as the total head
hydrostratigraphic unit in which the geologic materials have similar hydrologic properties
isotropic aquifer the aquifer matrix has the same characteristics in all directions; e.g., the hydraulic
conductivity is the same horizontally and vertically
ground-water head contours lines on a map representing equal ground-water elevations
ground-water flow field the direction of flow and elevation changes of ground-water across a site
line source a compact source of solute (i.e., contaminant) not elongated along the length of the ground-
water plume
minimum distance function a calculation in OWL that determines the minimum distance from a grid
point to the nearest monitoring well. Used to calculate the WOLF values
porosity the ratio of void volume to total volume of a rock or sediment
sorption movement of a substance from the dissolved phase to the solid phase, whether by adsorption,
absorption, fixation or precipitation. Sorption may be reversible or irreversible
58
-------
transmissive zones subsurface units where ground-water flow is constrained or bounded by lower
hydraulic conductivity materials (i.e., geologic impediments to flow) or hydrologic
barriers (e.g., hydraulic head boundaries). Transmissive zones may be bounded by
components as obvious as the water table or units with low hydraulic conductivity, or by
conditions as subtle as small differences in grain size, sorting, and packing of seemingly
uniform sands
uncertainty reduction of confidence in a conclusion when more than one estimate is available for a
variable
vadose zone zone between the ground surface and the water table
water-table aquifer an unconfined aquifer
Well Optimal Location Factor (WOLF) calculated factor used to help determine possible locations for
new monitoring wells, and eliminate redundant wells. See Appendix A for a detailed
description
59
-------
60
-------
6. References
Domenico, P.A. 1987. An analytical model for multidimensional transport of a decaying contaminant
species. Journal of Hydrology, 91(1987):49-58.
Mace, R.E., Fischer, R.S., Welch, D.M., and Parra, S.P. 1997. Extent, Mass, and Duration of
Hydrocarbon Plumes from Leaking Petroleum Storage Tank Sites on Texas, Geological Circular
97-1, Bureau of Economic Geology, Austin, TX.
U.S. EPA. 1998. Technical Protocol for Evaluating Natural Attenuation of Chlorinated Solvents
in Ground Water. United States Environmental Protection Agency, Office of Research and
Development, Washington DC 20460. EPA/600/R-98/128, September 1998.
U.S. EPA. 2000. BIOCHLOR: Natural Attenuation Decision Support System User's Manual, Version 1.0.
EPA/600/R-00/008, January 2000.
U.S. EPA. 2003. Performance Monitoring for Natural Attenuation Remedies in Ground Water. Office of
Research and Development and Office of Emergency and Remedial Response, Washington, DC
20460. OSWER 9355.4-25, PB 2003 103270, EPA 540-R-03-004, March 2003.
61
-------
62
-------
Appendix A. Field Case Study Site Characterization,
Data Collection, and Evaluation for Application of OWL
The following material provides data for a Case Study/Tutorial to demonstrate application of OWL for
analysis of site data. The material is designed to be used in conjunction with Section 4.0 Guided Tour
and Field Case Study Tutorial of the main document, which explains how to input and analyze the Case
Study/Tutorial data in OWL.
1. Appropriate Application of OWL Program
The OWL program is a simple screening program which is subject to numerous limitations on its use.
It should only be applied to sites which possess the characteristics outlined below. The first step in
using OWL requires gathering and analyzing site data to ensure that the site characteristics meet OWL
requirements.
1.1. Required Site Characteristics
a. The ground-water elevation data and site characteristics should indicate that the water table
or piezometric surface of the aquifer of interest can be represented by a linear plane. The
assumption of a linear hydraulic head surface means that flow in the aquifer is a consequence
of a constant horizontal gradient and no vertical gradients are present.
b. There should be good spatial coverage of monitoring wells screened in the aquifer of
interest across the site. The locations of the source and wells should be known using a fixed
coordinate system (see Appendix C for more details on the coordinate system). At least three
widely-spaced wells (in a triangular configuration; i.e., not in a straight line) are required to
define a linear plane reasonably well, but more wells improve the quality of OWL predictions.
c. There should be several sets of measurements of ground-water elevations from the monitoring
wells in the aquifer at different times. Routine ground-water elevation measurements on a
monthly or quarterly basis are preferred. If free product such as LNAPL exists in the wells, its
depth must be measured so that ground-water elevations may be corrected for its presence.
d. The aquifer of interest should be homogeneous and isotropic, and be of constant thickness.
e. There should be a single constant contaminant source which is well delineated, i.e., the three-
dimensional extent of the source is known. All sampling points (e.g., monitoring wells) used
in the OWL calculations should be outside the source area.
2. Data Collection and Evaluation Steps for a Field Case Study
The appropriateness of the application of the OWL program to a site is dependent on the site's ability to
meet the criteria outlined in Section 1 above. The following is a step by step description of data collection
using a field case study to teach the user to simultaneously evaluate the site and prepare the input data
for the OWL program. This material is to be used in conjunction with Section 4.0 Guided Tour and Field
Case Study Tutorial of the main document, which explains how to input and analyze the Case Study/
Tutorial data in OWL.
A-l
-------
2.1 Initial Data Collection for Site
a. Collect all available site characterization data including:
1) site history, well types and locations,
2) well completion information, core or drilling logs,
3) historical ground-water elevations,
4) hydraulic conductivity measurements,
5) well historical contaminant concentrations,
6) historical source area contaminant concentrations,
7) site basemaps, and
8) site plume maps.
b. At a minimum, the following information must be available to use the OWL program
properly:
1) well types/locations/completions,
2) description of aquifer of interest based on core/drilling logs,
3) aquifer hydraulic conductivity measurements,
4) source area location, and
5) routine measurements of ground-water elevations across the monitoring well
network.
Field Case Study
The site of interest is an inactive gas station which previously dispensed regular leaded,
unleaded plus and supreme grades of gasoline. The fuels were stored in four single-walled steel
underground storage tanks which had been in place for 18 years. When the tanks were excavated,
four soil samples at the excavation indicated total BTEX concentrations ranging from 8.37 to 266
mg/kg. Ground-water quality measurements in the source area of the excavation indicated total
BTEX concentrations up to 81.4 mg/L. BTEX concentrations remain high. Measurements of
total BTEX in the monitoring wells down gradient of the source area established the presence of a
significant plume emanating from the tank field.
Based on these findings, further site characterization was undertaken, including installation
of several monitoring wells in and around the plume area. Source area soil and ground-water
contaminant concentrations were measured. Soil and sediment core samples were taken during
construction of the monitoring wells to determine the subsurface geology and help delineate
plume shape and position. The core data were used to construct fence diagrams to display the
subsurface geology along the apparent centerline of the existing plume, and perpendicular to the
centerline. Screened intervals of the wells were also shown on the fence diagrams. A hard copy
basemap was drawn to show the contaminant source locations and the monitoring well locations.
Contaminant concentration contour maps were prepared to depict plume location and shape.
Slug tests were used to measure hydraulic conductivity of the contaminated aquifer in several
locations to determine whether the aquifer was isotropic and to provide hydraulic conductivity
measurements for use in OWL. Ground-water elevations were measured in the monitoring wells
seven times over a five year period.
A-2
-------
2.2. Prepare a Scaled Basemap of Site for OWL
a. Prepare or obtain a scaled basemap of site with significant features, including location of
source, monitoring wells, surface water, pumping wells, etc.
b. The best way to prepare a basemap is to have a professional surveyor conduct a survey to
determine well locations/elevations and other features such as buildings. A scaled basemap
on which the surveyed coordinates of the well locations are known can then be prepared (see
Appendix C for more details on the coordinate system).
c. If a hard copy basemap exists, it can be scanned into electronic form and scaled by the OWL
user based on the map scale bar, or on three points on the map whose location is known.
OWL accepts basemaps in both raster (*.bmp) and vector format (*.dwg). Appendix C has
information on basemap requirements.
d. If possible, the mapping convention of north being coincident with the y-axis should be used
for all basemaps.
e. If the basemap indicates the presence of significant surface water bodies, pumping wells, or
any other features which produce vertical gradients, the site may be inappropriate for OWL.
Field Case Study
For the field case study, a hard copy basemap with scale was available. This hard copy was
scanned into electronic form (*.bmp). . The basemap for the case study is shown in Figure 1
with distances defined in feet. As can be seen in the figure, there are no apparent features (surface
water, pumping/injection wells) which would preclude one from using OWL on this site. There is
also good spatial coverage of monitoring wells across the site.
2.3. Determine monitoring well locations based on map scale
a. Use the scaled basemap to determine the (x, y) coordinates of each well location. Be sure
to identify the units (ft, m) associated with the coordinate convention. The well coordinates
may have been assigned using an origin and scale created by a mapping specialist, or they
may have been assigned using a professional survey which may be in UTM, State Plane, or
in other mapping convention coordinates. If a scaled basemap is only in hard copy form and
no well locations are assigned, the user must use the scale bar on their map and determine the
well locations by hand. It is only important that the same scale be used for all well locations (
i.e., one cannot mix UTM with State Plane coordinate system).
b. Enter the well locations into a spreadsheet program so they may be imported into the OWL
program. The necessary data format is described in Appendix D.
A-3
-------
600-
500-
400-
200-
100-
Figure 1. Site basemap.
Field Case Study
For the field case study site, the basemap shown in Figure 1 was only available in hard copy. It
was possible to use the hard copy map scale and a ruler to determine the locations of the wells.
For the case study, the well locations based on the assigned grid are shown in Table 1. The user
then scanned the basemap into electronic *.bmp form, so it could be imported into OWL.
A-4
-------
Table 1. Monitoring Well Locations in Feet Based on Site Basemap Grid
Well x(ft) y(ft)
MW-1
MW-2
MW-3
MW-4
MW-5
MW-6
MW-7
MW-8
MW-9
MW-11
MW-12
MW-18
250.135
285.44
313.5
338.57
291.765
392.62
240.705
321.21
331.21
319.825
359.17
261.52
465.635
428.26
332.465
450.11
511.405
334.65
406.41
331.80
410.55
302.335
324.02
444.82
2.4 Characterize the hydrogeology of the aquifer of interest
a. Collect all core/drilling log information for the site. Also collect well completion information
such as depths and screen locations.
b. Prepare fence diagrams which show the subsurface geology denned at the monitoring wells
and correlate the thickness of the formations between wells.
c. Determine the thickness of the aquifer across the site from the fence diagrams.
d. Identify the wells completed into the aquifer of interest. Do not use data from wells not
completed into the aquifer of interest.
e. Collect available hydraulic conductivity measurements on aquifer. Perform well testing and
slug testing if no measurements are available.
Field Case Study
OWL assumes that the site aquifer can be represented as a single homogenous isotropic aquifer
of constant thickness. In reality there are no such aquifers, but many aquifers come close to
this definition. The best way to characterize an aquifer is to obtain drilling logs or cores which
describe the subsurface geology. It is then possible to draw fence diagrams which represent
this geology and assess the aquifer lithology and thickness. In addition, it is essential to have
completion reports on every well to determine which monitoring wells are screened in the aquifer
of interest.
For the field case study, Figures 2 and 3 display site fence diagrams which were used to determine
the location and thickness of the aquifer at each available monitoring well. These figures clearly
demonstrate the presence of a sandy confined aquifer which is overlain and underlain by clay
confining units. Although fence diagrams indicate that the aquifer is not defined as purely
homogeneous, it appears to be sufficiently similar to warrant that assumption.
A-5
-------
adv NoisaiavHO
& ft
b E
o
<
§
6
5
z
I
Q
~Z.
O
0
PLINGEVEN"
S
S,
g
i
1
Q
E.
If
ru.
Xn_
m Q-
LJJ
TT
^~. u
"GJ "5"
m m
g^
m
I
C
1
P
^
Q
Jl
LJJ
Q
1
1
§
Q
3E SEPARATE
<
^
s
60
o
60
1
60
O
£
o
C/3
O
T3
o
o
(/J
S/j
E
-------
o
60
_O
"o
o
60
I
60
o
O
p
o
'
s
S/j
(Id) NOI1VA313
-------
Figures 2 and 3 also display the thickness of the aquifer across the site. One can measure this
thickness to determine if it remains essentially constant. Table 2 shows the thickness of the
aquifer at each monitoring well. Although not constant, the thickness ranges between 6-11 feet
which is a small variation for this problem.
Finally, the fence diagrams show that MW-8 and MW-12 are not completed in the aquifer of
interest. These wells should therefore be dropped from the list of wells to be used to perform the
regression to compute the linear hydraulic head surface for the aquifer since they do not measure
the head in the aquifer of interest.
Table 2. Range of Aquifer Thickness at Field Case Study Site
Elevation of Bottom Elevation of Gray
Dark Plastic Clay Brown Silty Clay Aquifer Thickness
Well (msl) Bottom (msl) (ft)
MW-1
MW-2
MW-6
MW-8
MW-11
MW-12
MW-1 3
MW-14
MW-16
MW-17
20
21
23
25
23
23
22
22
23
20
9
10
16
17
16
16
16
15
14
14
11
11
7
8
7
7
6
7
9
6
Several measurements of hydraulic conductivity were made for the field case study site. Rising
head slug tests were performed on three monitoring wells (MW-6, MW-9, and MW-11) which
were all completed into the aquifer of interest. The hydraulic conductivities measured at these
wells were 2.21 ft/day, 0.77 ft/day, and 1.66 ft/day, respectively. The geometric mean of these
conductivities is 1.41 ft/day (514 ft/yr) which can be used in the OWL program. An effective
porosity of 0.45 was used based on literature values for the subsurface matrix present at the site.
2.5 Collect and correct all ground-water elevation measurements for the wells
a. Collect all ground-water elevation measurement data for the monitoring wells. OWL requires
that ground-water elevation data from the monitoring well network be collected over a
sufficient time period to capture any variation in the ground-water flow field. The best data
will be taken on a routine schedule, either monthly or quarterly.
b. Adjust all ground-water elevation measurements to mean sea level (msl) and correct for free
product. All data must be in converted to msl before it is used in the program to insure it
is all referenced to the same constant datum. Site data may record ground-water elevations
as feet below ground surface, or below the top of the well casing (TOC). Such data must
be converted to msl as indicated in the equations below. The conversion of ground-water
elevation data to msl depends on whether free product such as LNAPL is present in the well
A-8
-------
(see below for method to correct for free product). See Figure 4 for a typical well diagram
showing top of casing (TOC), ground surface, free product depth, ground water depth, and
well screen location.
t
Depth to jS \
Free '
Product
Depth to
Ground
Water
i >
t
^m
\
^m
\, w
r
^m
^^m
^
^
^
1 Top of Casing
Ground Surface
Top of Free Product
Top of Ground Water
Top of Screen
Bottom of Screen
Figure 4. Well completion and fluid level definitions.
c. Enter the ground-water elevation measurements into a spreadsheet for entry into the OWL
program. The format shown in Appendix D must be used.
d. If a ground-water elevation measurement is missing at a well for a particular date, enter a null
value of -9999 for that data.
Field Case Study
For the field case study there were eight separate dates when ground-water elevation
measurements were taken from 10 wells. Unfortunately, the measurement dates were not on a
quarterly or monthly basis, but they did cover five years and two seasons. For this reason they
were considered to be marginally representative of the seasonal variation in the ground-water
flow field. Future measurements at the site should be taken at least quarterly, and the sampling
dates placed at times representative of seasonal variation.
The ground-water elevation measurements were taken as feet below ground surface (bgs). Several
of the wells also had free product present, and the thicknesses of the free product had been
measured. It was therefore necessary to correct all the ground-water elevations to a consistent
datum of msl and also for free product. The corrections may be made as follows:
A-9
-------
Free product in the wells indicates that the wells are in the
source area. In such cases the O WL assumption that the
source is a line should be carefully examined to see if the
OWL program is really a valid application for the site.
1. No Free Product Present
If no free product is present and the ground-water elevation is measured as below ground surface
(bgs), it is necessary to have the top of the casing (TOC) in msl to determine the ground-water
elevation as follows:
TOC (msl) - GWelevation (bgs) = GWelevation (msl)
For the example site, MW-1 with a TOC of 30.63 ft msl, displayed a ground-water elevation
reading of 6.67 ft bgs on Jan 25, 1994. No free product was present in the well. Therefore the
ground-water elevation in msl at MW-1 on that date was:
30.63ft msl - 6.67 ft bgs = 23.96ft msl
2. Free Product Present
If free product such as LNAPL is present in the well, it is necessary to convert the thickness of the
free product to an equivalent water thickness to determine the corrected ground-water elevation in
msl. This is accomplished using the following equation:
Corrected GW Elevation (msl) = GW Elevation (msl) +
(LNAPL specific gravity) * (thickness of free product)
For the case study site, monitoring well MW-2 with a TOC of 30.63 ft displayed a depth to
ground water of 9.01 feet and a depth to free product of 8.50 ft on Dec 12, 1995. The thickness of
the free product is therefore:
Free Product Thickness = 9.0lft - 8.50ft = 0.51 ft
The ground-water elevation in msl is :
30.63ft (msl)- 9.01 ft (bgs) = 21.62ft msl
If the specific gravity of the free product (density of fluid/density of water)= 0.8, the corrected
ground-water elevation is:
Corrected GW Elevation (msl) = 21.62ft msl +
0.8(0.51 ft FP) =22.0 3 ft
All wells which demonstrate free product thickness must have their ground-water elevations
corrected in this manner before their data may be used in the OWL program.
A-10
-------
Once all the ground-water elevations for the case study are corrected for any free product and
converted to mean sea level, they should be entered in a spreadsheet like that shown in Table
3 for the example site data (see Appendix D for the specific format). A null value of-9999
should be substituted for any missing values. This null value is used as a flag in the program.
The spreadsheet can then be imported into OWL or the data may be hand entered into the OWL
program spreadsheet.
Table 3. Site Monitoring Well Locations and Corrected Ground-Water Level Measurements (msl)
1/25/94 12/12/95 10/30/98 12/4/98 12/21/98 1/11/99 3/29/99
Well x(ft) y(ft) Head (ft) Head (ft) Head (ft) Head (ft) Head (ft) Head (ft) Head (ft)
MW-1
MW-2
MW-3
MW-4
MW-5
MW-6
MW-7
MW-9
MW-11
250.135
285.44
313.5
338.57
291.765
392.62
240.705
331.21
319.825
465.635
428.26
332.465
450.11
511.405
334.65
406.41
410.55
302.335
23.96
23.00
21.97
22.39
23.76
21.14
23.52
22.13
21.81
22.94
22.03
21.01
21.82
22.90
20.64
22.19
21.49
20.70
23.43
22.22
21.25
22.14
23.57
20.86
22.55
21.81
20.96
22.39
21.55
20.53
21.49
22.57
20.14
21.49
20.88
20.16
23.64
-9999.00
21.43
22.39
24.55
21.98
22.60
22.26
18.63
22.67
-9999.00
18.78
20.77
23.50
19.24
20.64
19.92
16.46
20.35
-9999.00
17.73
19.77
20.17
17.23
20.15
18.36
15.91
MW-18 261.52
444.82 23.51
22.48
22.94
21.93 -9999.00 -9999.00 -9999.00
The OWL program assumes that the hydraulic head in the aquifer can be represented by a linear
surface. If pumping/injection wells, ground-water divides or significant surface water features are
present, the ground-water surface probably will not meet the linear surface assumption. The user
should examine the ground-water elevation data to determine if there are any obvious deviations
from a plane surface. If there are wells where the ground-water elevations are significantly lower
or higher than the surrounding wells it can indicate the surface is not linear. Such deviations can
violate program assumption, making OWL inappropriate to use. However, it may not be possible
to be certain whether the linear surface assumption is appropriate merely by examining ground-
water elevation data. For instance, hidden features such as leaking water mains with ground-water
mounds are not readily determined from ground-water elevation data. To be certain that the linear
surface assumption is appropriate, the user should run the linear regression in the OWL program
to determine how well the linear surface fits the data. An R2 value of 0.90 or higher is considered
good (see the Introduction Section 1.2.1). For the Case Study/Tutorial data, it was clear that the
data did not display any unusual ground-water elevation highs or lows.
2.6 Collect data on contaminant source location and contaminant transport properties
a. Delineate the contaminant source location using site characterization data to establish if it
can be represented as a single source. Estimate source width and the maximum time the
contaminant source was present.
A-ll
-------
b. Estimate transport properties of aquifer using existing plume to establish dispersion. A rule of
thumb for a screening model is
longitudinal dispersivity = 10% of plume length and
transverse dispersivity = 10% of longitudinal dispersivity.
c. Estimate fate parameters for adsorption and decay rate information if desired.
Field Case Study
For the field case study there were numerous measurements of contaminant concentrations in the
ground-water and soil/ground-water contamination in the source area. Figure 5 shows the benzene
plume contours and delineated source area on the site basemap for December 1995. From the
figure, it is clear that the source of contamination is limited to the area where the fuel tanks were
located. If one looks at the ground-water concentrations in wells in the source area they are
consistent over time which supports the OWL model assumption of a continuous source. The
width of the source was estimated to be about 30 feet using a ruler on the scaled basemap. The
tanks were installed in 1970. Although it is impossible to know when the tanks may have begun
leaking, the most conservative estimate of leakage duration would be 31 years (installation date
of the tanks). The length of the plume is about 300 ft so the estimated longitudinal dispersivity is
30 ft and the estimated transverse dispersivity is 3 ft. For the case study and most other studies,
fate parameters such as adsorption and decay are unknown. Therefore a retardation factor of one
(no adsorption) was chosen and the decay rate was set to zero (no decay); these choices provide
conservative estimates of plume extent in OWL. The user should always check the effects of
using different parameter values in OWL to determine the effect of changes in the parameters
on the OWL model results. In particular, the user should input a range of values of hydraulic
conductivity (k) to determine the sensitivity to k, because site values of k is likely to vary more
than any of the other site parameters used in the model. Using a range of values for k will cause
the model to calculate different plume lengths. In addition, the user may vary the simulation
times in OWL to determine how they impact the plume length. The plume will achieve steady
state depending on the different parameter and time values used. Knowledge of potential
variability of plume lengths should help the user assess the impact of program results on site
decisions.
3. Run the OWL Program
Once the data have been collected, corrected and checked to see if they meet the requirements of the
program limitations, it is possible to run the OWL Program. Please proceed to 4.0 Guided Tour and Field
Case Study Tutorial of the user's manual to see how to enter the data for the field case study, execute the
program and view the results.
A-12
-------
- -^^^^^^^^^^^^NSIEUATlONMyE '
Figure 5. Contours of benzene concentration for field case study.
A-13
-------
A-14
-------
Appendix B. Program Algorithms
The algorithms used in OWL are best described using the schematic illustration in Figure 1. The input
to the program includes ground-water level data for several sample dates. Linear regression is used to
fit a planar ground-water surface for each sample set. Based on these surfaces, ground-water elevations
are interpolated on the nodes of a virtual uniformly-spaced grid. Scalar data for contaminant transport
simulation (for defining the flow field, dispersivity terms, decay terms, source location, etc.) are supplied
by the user. The Domenico analytical solution is then used to perform contaminant transport simulation
for each set of data (Domenico, 1987). Concentration values are calculated at the each of the grid nodes
based on the analytical solution. An average concentration matrix is then calculated for each grid node,
from the concentration values of that node for all sample dates. For each grid node, distance to the nearest
ground-water well is determined. By multiplying a square of this minimum distance with the average
concentration for the node, a Well-Optimal Location Factor (WOLF) is calculated for each node. Each
algorithm is described in the following.
1. Linear Regression Algorithm
OWL uses least squares linear regression to fit the following function describing a planar ground-water
surface to each set of ground-water level data:
z=Ax + By + C (I)
where z regressed ground-water level at the node location (x,y)
x x coordinate of the node (x,y)
y y coordinate of the node (x,y)
A Regression coefficient for x direction
B Regression coefficient for y direction
C Regression constant
Once the coefficients for the best fit plane for each set of observations on a particular date are determined,
OWL uses equation 1 to predict the ground-water elevation, z, for any location (x, y) and plots linear
contours of these elevations on the basemap for the given observation date. OWL calculates the
magnitude of the ground-water gradient, dz/ds, from the slope of this plane:
= +
(2)
If the y coordinate axis is coincident with north as required by the program, OWL provides the ground-
water flow direction, 9, in degrees from north as:
\dz I dz
\
_ *
6 =arctan\
dx dy
The least squares linear regression in OWL uses observations of ground-water levels from each available
sampling date to determine the best-fitting plane representing the ground-water surface. For a particular
date, the ground-water elevations are assumed to fit the following linear equations:
B-l
-------
II
It
£. ««
D) c
co 2
CD =
T3 E
CD
i-f
u
>H"
"V
CD
II
T3
C «
§ ,'§?&
ag*
8 £:§
'§9 *
§ =2 g,-2
Q a g §
.2 I fe.
.a
3
O
X »
s-s
'H 'S S o
§ ^ aj f
-------
MW1: z^Cj+Axj+Byj+ej
MW2: z2 = C2+ Ax2 + By2 + e2
(4)
MW.: z = C. +Ax. + By. + e.
l l l l ? l l
MW: z =C +Ax + By +e
n n n n / n n
where the residual errors, ep ey .... e..., en are assumed in be independent and normally distributed with
zero mean and constant variance. In least squares regression, the strategy is to solve for the coefficients,
A, B, and C, by minimizing the sum of the square of the residual errors, Sr, for all of the available data:
(z.-Ax.-By.-C)2
1=1 i=l
Because the approach is to minimize the sum of the squares, it is called least squares fitting. Minimizing
Sr will produce the best-fit plane for the given set of observations (xf yf z), by minimizing the distance
from each ground-water level observation to the regressed plane.
There are many different solution algorithms which may be used to determine a set of coefficients, A, B,
and C which will minimize the sum of the squares, Sr. Most solutions are based on a system known as the
normal equations. First, the sum of the squares of the residuals is differentiated with respect to each of the
unknown coefficients:
dS
r
= -2
(z.-C- Ax.-By.)
^ i i 'i'
dc
dS x"1 (f.\
r - ->> x.(z.-C- Ax.-By.) ( '
dA
dS
^ = -2> v.fz.- C- Ax.-By.)
i/V J '
dB
These partial derivatives are set to zero to give a system of simultaneous linear equations which when
solved produce the coefficients that minimize the sum of the residuals, Sr:
x.C + \x2A + V ' x.y.B = VJT.Z.
l Z_^ ! Z_^ I'' I ^j I I
B-3
-------
These equations may be written in matrix form as:
C
A
B
n
Z*
z*.
-1
JC.Z
(8)
where the [ ]-1 denotes the inverse of the matrix. When the inverse exists, the matrices may be solved and
the solution is unique. If the inverse does not exist, an infinite number of solutions exist.
Unfortunately, the above normal equation approach may fail when it is applied to the least squares
problem of fitting a plane to ground-water elevation data. This is because the well locations are often in
UTM or State Plane coordinates which may be quite large (more than six digits) and only vary by the last
two digits. In addition, water level elevations from well to well may only change by small fractions. This
produces a set of observations which appear very similar numerically. As the least squares systems of
equations are formulated for such ground-water elevation data, it is apparent that the modeling functions
appear to be linearly dependent. Linearly dependent equations produce a singular coefficient matrix which
leads to a solution to the system of equations that is not unique. Kahaner et al. (1989) noted that the
normal equation solution to the general least squares problem is not the best approach since it performs
poorly with overdetermined, underdetermined, and singular systems. Instead, they suggested a method of
matrix factorizations known as orthogonal matrices to solve singular systems like those produced from
attempting to fit a plane to ground-water elevation data. It is this least squares solution method which
is used in OWL. A description of the orthogonal factorization solution to the least squares problem is
beyond the scope of this document and the reader is referred to Kahaner at al. (1989) for a complete
discussion.
The OWL program provides a coefficient of determination, R2, to judge the goodness of fit of a linear
plane to fit ground-water elevation data. The equation for R2 is:
, S - S
R2=_L r_
(9)
t
where St is given by:
(10)
The value of R2 from the regression results compares the amount of variation of ground-water elevation in
wells one from another to the amount of variation of elevation in wells from the plane fit through all the
elevation data by the regression equation. When there is perfect match between the measured elevations
and the elevation of the plane, then the assumption that the ground-water surface is a plane explains all
the variation between the individual wells and the value of R2 is 1.000. If the value of R2 is 0.5000, the
assumption that the ground-water surface is a plane explains half the variation in elevation between wells.
If the value of R2 is less than 0.5000, the assumption that the ground-water surface is a plane explains less
than half the variation. As a general rule of thumb, elevation data sets that produce a value of R2 that is
less than 0.5000 are not appropriate for use in OWL.
B-4
-------
It should be noted by the user that the coefficient of determination, R2, provided by OWL is not an
adjusted R2. It does not account for sample size (number of wells and observations) and should not be
used to compare the fit of the data between different sampling dates. It can only be used as a measure of
goodness of fit of the plane model to the observations for the sample set from which it was calculated.
The least squares linear regression program code for OWL was written in FORTRAN. It calls the SQRLS
subroutine of Nash (1987). SQRLS solves an overdetermined, underdetermined, or singular system of
linear equations in a least square sense using an orthogonal factorization solution. SQRLS contains the
Linpack Subroutine SQRDC. This code was substantially modified by the authors to address the specific
problem of fitting a plane to ground-water elevation data posed by OWL. The performance of this
algorithm was verified using the linear regression package available in Mathematica.
2. Analytical Transport Solution
The Domenico analytical solution for ID contaminant transport is used in OWL (Domenico, 1987). This
solution is used to simulate the multidimensional transport of a decaying contaminant species. For the
purpose of OWL, the Domenico solution is adapted to a two-dimensional transport. In this solution, the
concentration at a distance x downstream of the source and distance y off centerline of plume at time t is
specified as
C(x,y,t) = -£-exp
erfc
a -2
1-J1
erf
-erf-
y
where
K-i
v =
Definition of the terms:
C(x,y,t)
C0
Y
x
y
t
a*
a
(11)
(12)
Concentration at a distance x downstream of the source and distance y off
centerline of plume at time t (Concentration units)
Concentration in the source zone at time t = 0. (Concentration units)
Line source width (L)
Distance downgradient of source (L)
Transverse distance from the plume centerline (L)
Simulation time (T)
Longitudinal ground-water dispersivity (L/T)
Transverse ground-water dispersivity (L/T)
First-order degradation rate (1/T)
B-5
-------
v Ground-water [retarded] seepage velocity
K Horizontal hydraulic conductivity (L/T)
/ Hydraulic gradient (L/L)
0 Effective soil porosity
R Constituent linear retardation factor
The equation is similar to the Domenico analytical model used in the Bioscreen (U.S. EPA, 1996)
program, except the equation has been reduced to two dimensions. For further discussions of the
applicability of the solution and assumptions, the reader is referred to the Bioscreen User's Manual (U.S.
EPA, 1996). The Domenico analytical solution in OWL was programmed in Visual Basic. It was verified
using MATHCAD.
In OWL, the user should note that the (x,y) values used in the Domenico solution represent the distance
traveled down the centerline of the plume and transverse to this centerline, respectively. They are
therefore along the direction of ground-water flow and transverse to it. This is standard notation for the
Domenico solution which was maintained for consistency with the reference. As the ground-water flow
field magnitude and direction changes for the regressed ground -water surface plane produced from each
set of ground-water level observations, OWL automatically rotates the line source to be perpendicular to
the direction of flow and maintains a constant source width Y. This ensures that the Domenico solution
uses the same source but produces a plume which is unique to each ground-water flow field.
The following assumptions are inherent to the application of the Domenico analytical solution:
The aquifer and flow field are assumed to be homogeneous and isotropic.
Adsorption is assumed to be a reversible process simulated by linear equilibrium isotherm.
The ground-water flow field is assumed to be uniform and one-dimensional.
The ground-water velocity is fast enough that molecular diffusion in the dispersion terms can be
ignored. This may not be appropriate for transport through clays.
It is assumed that the degradation can be effectively simulated using first-order degradation (by
specifying a half-life value).
A line source of specified width extending perpendicular to the flow direction is assumed. The
source is also fully mixed in the vertical direction.
The orientation of the "line" in the line source is assumed to be based on the ground-water flow
direction, and not any other physical extent or shape.
3. Average Concentration Plume Algorithm
For each grid node, an arithmetic average of the concentration values for all the sample dates is calculated
as shown below.
. _j_
n ' (13)
where:
C(i,j, t) Concentration value for node (i,j) at time t
n Number of sample dates
Cav (i,j) Average concentration value for node (i,j)
B-6
-------
The average plume concentration algorithm for OWL was programmed in Visual Basic. Its performance
was verified by spreadsheet.
The following assumptions are inherent to the application of the average plume concentration calculation:
Ground-water flow field interpreted from each sample date is given equal weight to that from any
other sample date. For example, if three sample dates fall within one year, followed by a sample
date 2 years later, taking a simple arithmetic average may not be the best approach.
Contaminant transport based on the flow field set by each sample set, is assumed to take place
independently of the transport based on other sample sets. This is in contrast with a typical
numerical transport model where the effects of varying ground-water field are sequentially
applied.
The arithmetic average technique works best when the concentrations from different sample sets
do not vary significantly from each other. If the concentrations vary by orders of magnitude, a
geometric average may be better.
The average concentrations should not be used as predicted concentrations for design purposes.
4. Well Optimal Location Factor (WOLF) Algorithm
The Well Optimal Location Factor (WOLF) is calculated for each node of the grid using the function
WOLF(i,f) = AninO'J)2 ' QvgO',7)
(14)
where:
Distance from node (i,j) to the nearest monitoring well
Cav (i,j) Average concentration value for node (i,j)
WOLF(i,j) WOLF factor value for node (i,j )
Nodes with a higher value of the WOLF value as compared to neighboring nodes, denote a likelihood of
a significant plume concentration where there are no nearby monitoring wells. The relative WOLF values
can therefore be used to select optimal locations to place new monitoring wells. Note that this approach
is designed to be used as an initial screen tool for locating wells.
The following assumptions are inherent to the calculation of the Well Optimal Location Factor:
The application of relative WOLF values to place new monitoring wells is based on numerous
assumptions and is a simple initial tool for locating wells.
Several other factors such as geology, uncertainty of the source extent, etc. should be evaluated to
assess the usefulness of the WOLF.
B-7
-------
REFERENCES
Domenico, P. A., 1987. An analytical model for multidimensional transport of a decaying contaminant
species, Journal of Hydrology, 91(1987):49-58.
Kahaner, D., Moler, C., and Nash, S., 1989. Numerical Methods and Software, Prentice Hall, Englewood
Cliffs, NJ.
Newell, C.J., McLeod, R.K, and Gonzales, J.R., 1996. Bioscreen Natural Attenuation Decision Support
System, User's Manual, Version 1.3, EPA/600/R-96/087, August 1996.
Nash, Stephen, 1987. George Mason, University.
http://iris.gmu.edu/~snash/nash/software/NMS single/samples/sqrls.f
B-t
-------
Appendix C. Site BaseMap Formats and Scaling
The Optimal Well Locator (OWL) can import a digitized site map (the "basemap"). OWL can import
a scanned or electronically generated image of a site map and display the map in the background of the
Map View window, for use in overlaying other data such as well locations and plume contours.
The following image types are recommended for use in OWL:
1) Raster maps (such as an aerial photo or scanned image): Use Windows or OS/2 Bitmap * .bmp
format. OWL does not accept the * jpg format.
2) Vector maps (CAD files): Use AutoCAD Version 12 or 14 DWG format.
OWL may also accept the following raster and vector formats. However, use these optional formats with
caution, as there are strict limitations (compression type, monochrome, etc.) with each of the formats
below:
Raster formats:
.RLE: Windows Bitmap Compressed (.RLE); (not CompuServe .RLE)
.RLC: Files RLC (.RLC) in two colors
.TIP: Files TIFF (.TIP) OWL may accept monochromatic non-compressed, PackBits
compressed, or Group 3 Modified Huffman compressed, but not LZW compressed.
Vector formats:
.DXF files
ESRI ShapeFiles (.SHP)
.WMF files (Windows MetaFile)
AutoCAD DWG formats are recommended for vector-formatted OWL basemaps. Typically a vector-
format basemap has several types of objects (text labels, lines, polygons, arcs, etc.), but an ESRI Shapefile
can store only one type of graphic object. Because OWL can only import one Shape file, the Shape file
format is not recommended for use as an OWL basemap. Instead, Shape files may be imported into
AutoCAD and saved as a single DWG or DXF file.
The basemap should be to scale, and should show significant site features, including location of
contaminant source, monitoring wells, surface water, pumping wells, etc. At most sites the well
locations have been determined by survey, so the coordinates are available in UTM, State Plane, or
similar coordinate system. If not, a professional surveyor should conduct a site survey to determine
well locations/elevations and other features such as buildings. A scaled basemap on which the exact
coordinates of the well locations are known can then be prepared. If possible, the mapping convention of
north being coincident with the y-axis should be used for all basemaps.
C-l
-------
Map Calibration
Many site maps already have Cartesian coordinate systems (XY coordinates such as northings and
eastings) associated with the map, either printed on the sides of a paper map, or stored in the file with the
electronic map. The XY location values for known points on the map can be used for the real coordinates
in the Set Map Scale dialogue box in OWL. If the site map does not have a pre-existing XY coordinate
system, the user can make up their own coordinate system assigning an origin (a 0, 0 point) to a location
on the site (perhaps at the lower left-hand corner of the site paper map) and laying out an XY coordinate
system along the sides of the map. Then the user can locate known points on the site map (e.g.,
monitoring wells, buildings, benchmarks) and determine the XY coordinates of these points in relation to
the origin point for use as the real coordinates in OWL.
Calibration Using Two Known Points
When choosing two points on the basemap to calibrate the map in OWL, the user should choose two
points as far from each other as possible, and located on a diagonal to the XY coordinate system. If the
two points are close together, the error in calibrating distance can be very large. If the two points are
along the same coordinate line (e.g., both points on a line parallel to the X axis), OWL will not be able to
calibrate to distances along the other axis.
Click on the checkbox beside "
-------
Calibration using a scale bar
The basemap may also be calibrated by clicking on opposite ends of an object such as a scale bar on the
basemap and specifying the distance between the points. This method of calibration may be much less
accurate than using two calibration points as in the preceding method.
Identify a scale bar on the raster map and specify another point on the map as the origin of the X,Y
coordinate system. Click on the "" button to specify the map origin. Now zoom in to a
section of the map where the map origin is to be placed, and click on the center of the origin.
Next, zoom in to a section of the map where the scale bar is displayed. Identify two scale markings and
enter the distance between these two points. For example if the scale bar markings are at 0, 200, 400 ft,
the marking for 0 ft and 400 ft may be selected.
On the dialog box, click on "," and locate a marking point on the scale bar and click on
it. Then click on "," and locate the other marking point on the scale bar and click on it.
The two points of a scale bar usually fall on a line parallel to the X axis; this is acceptable.
Enter the requested distance data as shown below:
Enter the distance between the points:
Type the distance between the points 1 and 2 in real map coordinates. Enter the values in the distance
units requested on the dialog box.
Calculated scale 100pixels =:
After entering the distance value, the calculated scale will be displayed in this field.
Once the map origin, the two scale bar points, and the distance between them are entered, check the
"" check box. The map is now calibrated. The status bar of the
user interface should now read "Raster map calibrated. Coordinates = ft." Press "" to close the
dialog.
C-3
-------
C-4
-------
Appendix D. Spreadsheet Formats for OWL Data Files
OWL supports import of well location and ground-water elevation data from common spreadsheet
software file formats, including Excel 4.0, Excel 5.0, Excel 97, Excel 2000, Excel 2002, Lotus WK3, and
Lotus WK4. The data must be in a specific layout format in the spreadsheet, as described below.
Spreadsheet Format for Well Location Data
The spreadsheet format for the well location data is shown below.
1 A
1 IWELLID
2 JMW-1
3 JMW-2
'4 JMW-3
5 lMW-4
6 MW-5
7
8
9
10
11
12
13
14
15
IT
MW-6
MW-7
MW-9
MW-1 1
MW-13
MW-1 4
MW-1 5
MW-1 6
MW-1 7
MW-1 8
B
X
250.135
285.44
313.5
338.57
291.765
392.62
240.705
331.21
319.825
214.715
28.53
372.84
566.385
339.605
261.52
C
Y
465.635
428.26
332.465
450.11
511.405
334.65
406.41
410.55
302.335
579.025
293.02
688.275
301.3
172.27
444.82
The well location data should be in the following format. There should be three columns:
Column 1: contains the well identification
Column 2: contains the X coordinate for the well locations
Column 3: contains the Y coordinate for the well locations
Each column should have a title or label, such as "Well ID, X Coordinate, Y Coordinate." The rows
and columns should be contiguous; there should be no empty rows or columns between the data rows
and columns. The well names should be in text format. Numeric values must be used for the X,Y
coordinates. No missing values or text are allowed in the X,Y coordinate cells.
Survey data on the location of the monitoring wells may not be available to you. You can use OWL to
help create the spreadsheet that is needed to Import Well Location Data from information in the basemap.
Create a blank spreadsheet listing wells and locations, and print the spreadsheet. After you have
calibrated the basemap to real coordinates, the x and y coordinates of the cursor are listed in left hand
side of the gray bar below the basemap. Position the cursor over each monitoring well, and write the
coordinates on the hardcopy of the spreadsheet. Then enter the data in the spreadsheet from the hardcopy.
D-l
-------
Spreadsheet Format for Water Level Data
The spreadsheet format for the water level data is shown below.
^J^^^A^_
TjDATE '
~Jj 1/25/1994
3
4
5
6
7
8
12/12/1995
10/30/1998
12/4/1998
12/21/1998
1/11/1999
3/29/1999
MW-1
23.96
22.94
23.43
22.39
23.64
22.67
20.35
c
MW-2
23.00
22.03
22.22
21.55
-9999.00
-9999.00
-9999.00
D
MW-3
21.97
21.01
21.25
20.53
21.43
18.78
17.73
E
MW-4
22.39
21.82
22.14
21.49
22.39
20.77
19.77
, F
MW-5
23.76
22.90
23.57
22.57
24.55
23.50
20.17
lG__
[iviwi
21.14
20.64
20.86
20.14
21.98
19.24
17.23
H
MW-7
23.52
22.19
22.55
21.49
22.60
20.64
20.15
MW-9
22.13
21.49
21.81
20.88
22.26
19.92
18.36
J
MW-11
21.81
20.70
20.96
20.16
18.63
16.46
15.91
K ..
MW-1 8
23.51
22.48
22.94
21.93
-9999.00
-9999.00
-9999.00
The water level data must be in the specified format as shown in the above screenshot. No other data can
be present on the same worksheet; i.e., only the column titles, a date for each row, and the water level data
can be on the worksheet. The rows and columns should be contiguous; there should be no empty rows or
columns between the data rows and columns.
Water level data from each well are in a column below the well name. Data from each sampling date are
in a row right of the sampling date. Missing data are indicated by the value -9999.0.
If the available water level data is formatted as sampling dates across the columns and well IDs
are down the rows, use the Transpose operation (in Excel) to convert the data to the above format. To do
so, highlight and copy the data range first, then use Edit > Paste Special and select Transpose to paste the
data on a different worksheet.
D-2
-------
Appendix E. Guidance to Edit
Ground-Water Elevation Data for OWL
OWL assumes that the ground-water surface at a site is a plane and uses linear regression of the available
water elevation data to fit a plane through the data points. OWL provides the regression results in the
Water Level Contours dialogue box to alert the user to water level data sets which are not well fitted by
a linear plane. The ground-water elevations in particular wells at a site will vary one from another. The
ground-water elevations in particular wells may also vary from the hypothetical plane that was fit by the
regression through all the elevations. The value of R2 from the regression results compares the amount
of variation of water elevation in wells one from another to the amount of variation of elevation in wells
from the plane fit through all the elevation data by the regression equation. When there is a perfect match
between the measured elevations and the elevation of the plane, then the assumption that the ground-
water surface is a plane explains all the variation between the individual wells and the value of R2 is
1.000. If the value of R2 is 0.5000, the assumption that the ground-water surface is a plane explains half
the variation in elevation between wells. If the value of R2 is less than 0.5000, the assumption that the
ground-water surface is a plane explains less than half the variation. As a general rule of thumb, elevation
data sets that produce a value of R2 that is less than 0.5000 are not appropriate for use in OWL. If the fit
is poor, it is important that the user review the ground-water elevations within the questionable data sets
and determine if they are truly not suitable for a plane fit or if there are anomalous points which cause the
deviation.
When reviewing the points in a data set, the user should be aware there is an important difference between
the data requirements for OWL and the data requirements for a transport and fate model. Statisticians
recognize two types of error in drawing an inference from a data set. Type I error draws an inference
from the data set when the inference is not true. Type II error fails to draw an inference from the data
set even though the inference is true. In other words, type I error sees something that is really not there,
while type II error fails to see something that really is there. Transport and fate models are used to shape
the conceptual model of a plume, and make predictions about its behavior. Data to calibrate a transport
and fate model should be selected to guard against type I error. The data that are available to calibrate the
model should be screened, and data that are at all questionable should be excluded. OWL is intended to
identify the best locations to look for a plume. Actual monitoring data from the wells that are installed
at the best locations are used to shape the conceptual model of the plume, and make predictions about its
behavior. Data for use in OWL should be selected to guard against type II error. As much as possible,
the data available for OWL should be retained. This is particularly true if the available data are limited.
1. Sources of Error in Ground-Water Elevations
Several factors or events can influence the relative variation in the elevation of the ground-water
in monitoring wells at a site. There may be a simple error in reading the tape or transcribing the
measurement. There are several models of depth-to-water meters, and the tapes on some of them are
difficult to read. Only fractions of a foot are marked, and the foot increments are only labeled every
foot. It is easy for an inexperienced field technician to associate the mark for a fraction-of-a-foot with
the wrong mark for a foot. For example, the technician may read 8.4 feet as 9.4 feet. Some meters will
give a false reading if there is moisture condensing on the side of the riser or screen above the free water
surface. Individual measurements of depth to water may have significant errors.
E-l
-------
In areas where the ground freezes, frost heave can raise the elevation of the top of casing by several
inches. If the elevation of the ground-water in a well is consistently a few inches higher than its
neighbors, there may be an error in the elevation of the top of the casing.
If the well is not properly sealed, drainage from rain or snowmelt can infiltrate the aquifer in the area of
the well and raise the ground-water level. Are anomalously high ground-water elevations associated with
periods of precipitation? Water from leaking water mains and sewer lines can also raise the water levels.
Compare the map of monitoring wells to an as-built map of utilities.
A particular well may be screened in a different hydrologic unit than the unit containing the plume of
contamination, and there may be differences in head between the units. If there are marked differences
in elevation between neighboring wells, consult the well logs. Use wells to calibrate OWL that have
narrow screens in the contaminated depth interval, and exclude wells with wide screen intervals. If water
in one well is clean and the water in a neighboring well is contaminated, keep the elevation data from the
contaminated well and exclude the data from the clean well.
If the well with anomalous ground-water elevations is in or near a cluster of other wells with more
representative water elevations, the anomalous well can be excluded with little damage to the conceptual
model of water flow in the aquifer. When the regression statistics are calculated, the constant will
change (the average water level elevation for that sampling date), and R2 should have a larger value,
but the direction (of flow) and magnitude (the hydraulic gradient) should change relatively little. If the
anomalous well has no close neighbors or is at the margin of the well field, it will have a strong influence
on the direction and magnitude of flow that is extracted by the regression. In this case it is best to have an
external reason to exclude the well from the data set used in OWL. At a minimum, compare runs of OWL
with and without the anomalous data, to see if its presence makes a practical difference.
2. Site Example
The data used in OWL should be evaluated systematically. This process will be illustrated with data from
a site in Southern California. A spill of gasoline formed a plume of contamination in ground water in
a shallow aquifer in silty fine sands and clean sands. The sand aquifer was confined above by 15 to 20
feet of clays and silts that extended to the land surface. The ground-water surface was near the contact
between the sands and clays. Monitoring wells were screened across the first 10 to 15 feet of the sandy
aquifer.
The site basemap was scanned, imported and calibrated to real map coordinates in OWL. Data was
imported on elevations in 14 wells on each of 14 sampling dates. After OWL ran the regressions, the
following information was provided in the Water Level Contours dialogue box.
E-2
-------
Date
4/16/1999
7/26/1999
10/15/1999
1/11/2000
4/18/2000
7/28/2000
10/25/2000
1/17/2001
4/3/2001
7/9/2001
10/5/2001
1/25/2002
4/17/2002
7/15/2002
Xcoeff
-0.0003305
-0.002093
-0.001118
-0.001646
-0.001298
-0.001482
0.00008204
-0.001422
-0.001532
-0.001403
-0.001969
-0.001973
-0.001508
-0.001428
Ycoeff
0.000158
0.003013
0.001617
0.001682
-0.000609
0.001704
0.003496
0.001168
0.002036
0.002069
0.002483
0.003078
0.002351
0.001996
Const
8.052
9.649
9.491
8.9
9.205
9.478
8.646
9.148
8.951
9.39
9.531
9.366
8.635
9.072
R2
0.003971
0.09294
0.6465
0.7303
0.1055
0.6654
0.1673
0.6411
0.6389
0.6197
0.0997
0.1288
0.6515
0.6651
Magnitude
0.0003665
0.003668
0.001966
0.002354
0.001433
0.002258
0.003497
0.00184
0.002547
0.0025
0.003169
0.003656
0.002793
0.002455
Direction
1156
1452
1454
1356
6487
1 39
181 3
129.4
1 43
1459
141.6
147.3
147.3
144.4
Sampling dates with R2 less than 0.5000 are highlighted in red and are bolded. At first examination, the
regression was not useable for 6 of the 14 sampling dates.
To determine the cause of the low values for R2, the contoured elevations were compared to the actual
elevations in the ground-water elevation data file. To facilitate the comparison, the basemaps with the
contours were printed from OWL and the actual values of elevation were hand written on the printed
maps next to the location of the monitoring wells. The example below compares the actual data and
contours for data collected 10/5/2001. The value of R2 for the regression to elevation data collected
on that day was only 0.0997. In the contour map below, the wells are labeled by name followed by the
ground-water elevation. There was reasonable agreement between the contours and the actual data except
for one well. The other wells range in elevation from 9.23 to 9.91 feet. The anomalous well has an
elevation of 11.81 feet (red on the contour map). The anomalous well is near the center of the site, and is
surrounded on all sides by wells with much lower elevation.
MW-9 9.42
MW-1 9.91
MW-2 9.55 "
MW15 9.41 , MW-13 9.26
4- *
,f MVV-14 11.81
^
MW-3 9.38
4; X
MW-6 9.39
}:'." MW-16 9.22
MW-7 9.33"*
MW-10 9.39
> > MW-12 9.23
10/5/2001
MW11 9.33
When the elevation of MW-14 in the ground-water elevation data file was replaced with -9999.00 (null
value), and the data were contoured again, the new value of R2 was 0.5536. In the new contour map
E-3
-------
below, notice that the average elevation was lower, and that the hydraulic gradient was steeper when the
questionable datum was removed.
MW-1 9.91,
." MW-9 9.42,,
MW-2 9.55,
4- MW-15 941 +
MW-13 926
MW-3 938- I
9.3B
MW-6 9 39
MW-10 939,
MW-16 9.22
'*-',- *
,. MW-7 933 >,.' KlW-12 923
r* ,*-
f ,
*. f
'MW-11 9 33- '' *
The original ground-water elevation data file is presented below. The process of comparing the elevation
data to the contoured data revealed a highly suspect datum in five of the six dates with low values of R2
in the regression equation. The dates and suspect data are colored red and bolded in the table below. The
suspect data deviated by at least 1.0 foot from a more representative value.
Three of the five suspect ground-water elevations were associated with the same monitoring well (MW-
14). Although comparison of the elevation data to the contours made it easier to recognize the suspect
data, they could also be easily distinguished by comparing the elevations in a well for a particular date
to the elevations on the date before and the date after the particular date, and by comparing elevations in
adjacent wells on the date before and date after. However, this process of comparison did not reveal any
suspect data in the elevations collected on 4/16/1999. The suspect data were replaced with the code for
missing data, and the file was imported into OWL, and the regression was run a second time. The edited
ground-water elevation data file is presented below.
E-4
-------
L. t
M
- Tfc
M
B i KWSflOOP'
~n ".. r
TTt :'.>'!
II ' " .' MJ
i2
15
-33B50SIJ n!
-'"1 4-'V 4' ,
jT-rTT- /V
9 ' '.I -S999 W
Presented below are the statistics from the Water Level Contours dialogue box after the regression was
run with the edited data set. The values of R2 in the regressions for the dates where the suspect datum was
deleted are now all greater than 0.5000, and are now acceptable to use in OWL.
Date
4/16/1999
7/26/1999
10/15/1999
1/11/2000
4/18/2000
7/28/2000
10/25/2000
1/17/2001
4/3/2001
7/9/2001
10/5/2001
1/25/2002
4/17/2002
7/15/2002
Xcoeff
-0.0003305
-0.001327
-0.001118
-0.001646
-0.00129805
-0.001482
-0.001612
-0.001422
-0.001532
-0.001403
-0.00133
-0.001347
-0.001508
-0.001428
Ycoeff
0.000158
0.001648
0.001617
0.001682
-0.001149
0.001704
0.002623
0.001168
0.002036
0.002069
0.001346
0.001964
0.002351
0.001996
Const
8.052
9.514
9.491
8.9
9.035
9.478
9.102
9.148
8.951
9.39
9.419
9.256
8.635
9.072
R2
0.003971
0.6599
0.6465
0.7303
0.5901
0.6654
0.5931
0.6411
0.6389
0.6197
0.5536
0.6375
0.6515
0.6651
Magnitude
0.0003665
0.002116
0.001966
0.002354
0.001665
0.002258
0.003078
0.00184
0.002547
0.0025
0.001892
0.002382
0.002793
0.002455
Direction
1156
141 2
145.4
1356
1336
139
1434
1294
143
1459
1353
1456
147.3
144.4
The magnitude (slope of the regression) for data collected on 4/16/1999 is low, about 10% of the
magnitude on other dates. The magnitude of the regression is the hydraulic gradient. If the gradient is
truly flat, then variations in elevations are dominated by errors of measurement and by minor variations
caused by spatial heterogeneity. If there is no contribution to the variations caused by the hydraulic
gradient, then a regression to fit a plane would not explain the variations in elevation and the value of R2
should be low. If the ground-water surface is truly flat, the data collected on 4/16/1999 should be retained
to use in OWL.
To determine if the gradient was truly flat, the range in ground-water elevations on 4/16/1999 was
compared to the range in elevations on the other dates. If the ground-water surface is truly flat, and other
sources of variation are the same, the range in variation in elevations on 4/16/1999 should be less than the
E-5
-------
ranges in variation on other dates. This was not the case. The range in variation across all the wells on
4/16/1999 was 1.49 feet. After the suspect data were deleted, the range in variation across all the wells
on each of the other dates extended from a high of 1.07 feet to a low of 0.30 feet, with an average of 0.65
feet. The data collected on 4/16/1999 was more variable than the other data, and the very flat gradient
extracted by the regression was probably circumstantial. Data collected on 4/16/1999 was therefore
excluded from the final OWL run.
The average direction of flow was 141 + 6.1 degrees on the eight sampling dates when the value of R2
was greater than 0.5000 without editing. The direction of flow predicted from the original data collected
on 4/18/2000 was 64.9 degrees and the direction of flow predicted on 10/25/2000 was 181.3 degrees.
When the regressions were calculated with the edited data sets, the direction of flow was much more
representative of the other dates. The direction of flow on 4/18/2000 was 133.6 degrees and the direction
of flow on 10/25/2000 was 148.4 degrees. The regression equations extracted from the edited data set
were more internally consistent. The possible plume contours generated by OWL were confined to a
smaller area when OWL was run with the edited data set.
E-6