PROJECT WORK PLAN
POTENTIAL GROUND-WATER FLOW
DIRECTIONS IN THE UPPER SAND UNIT OF
THE PLAQUEMINE AQUIFER, IBERVILLE
PARISH, LOUISIANA
(DRAFT)
March 25,2003
Prepared by:
Scott Ellinger
Multimedia Planning and Permitting Division
EPA Region 6
Telephone No. (214) 665-8408
-------
TABLE OF CONTENTS
Section Page
1. Introduction 2
a. Project Background 2
i. Scope of Work 2
ii. Purpose of Model 3
iii. Capabilities and Limitations 3
b. Organizations and Modeling Team 4
c. Project Milestones and Schedule 5
2. Modeling Objectives 8
3. Information and Data Collection 11
a. Sources 11
i. Literature Search 11
ii. Louisiana DOTD Information : 12
iii. EPA Region 6 Facility Files 13
iv. LDEQ Ground-Water Investigation 14
4. Conceptual Model Development 15
a. Elements to Research and Evaluate 15
b. Perform Integrated Interpretation 16
c. Documenting Uncertainty 16
d. Presenting the Conceptual Model 17
5. Numerical Model Development 18
a. MODFLOW AND MODPATH 18
b. Numerical Model Framework 21
i. Model Domain 21
ii. Model Layers 24
iii. Ground-Water Wells 24
iv. Hydraulic Properties 25
v. Boundary Conditions 27
c. Model Output 28
i. Output Options 28
d. Sensitivity Analysis 32
6. Final Modeling Report 34
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 2 of 34
1. Introduction
a. Project Background
i. Scope of Work
The area to be modeled encompasses the approximate area between DOW Chemical Company
(Louisiana Operations facility) and the City of Plaquemine, Louisiana, and lies along the western side of
the Mississippi River in Iberville Parish. In this immediate area, a vinyl chloride plume has recently
been discovered in the Plaquemine aquifer upper sand unit as deep as approximately 200 feet below
ground surface. In 1997 and 1998, the Louisiana Department of Health and Hospitals (LDHH) sampled
water wells at the Myrtle Grove Trailer Park in Plaquemine, Louisiana, and detected levels of vinyl
chloride and cis 1,2 dichloroethylene which exceeded Maximum Contaminant Levels (2 ppb for vinyl
chloride; 70 ppb for cis 1,2 dichloroethylene). The two chemicals were detected again during March
2001. LDHH failed to notify residents of the chemical detections until after the March 2001 sampling
event.
The Louisiana Department of Environmental Quality (LDEQ) has been conducting a phased
ground-water investigation since approximately April 2001. The objective of LDEQ's investigation has
been to identify the source of vinyl chloride contamination. The events listed below are the main
elements of LDEQ's investigation to date.
Neighborhood/local business survey of water wells
Review of DOW monitoring data
Research reductive dehalogentation of chlorinated solvents/sampling strategy
Phase 1 sampling, April 2001, 11 water wells
Phase 2 sampling, May 2001, split samples with DOW/confirmed Phase 1 results
Phase 3 sampling (May-June) sampled 21 wells
Received assistance from EPA-NRML lab in Ada, Oklahoma
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 3 of 34
Received response from EPA-NRML on July 26, 2001
Phase 4 (June 2001), split sampling with DOW on 6 boreholes
Phase 5 (July 2001) testing of fire water wells screened in Plaquemine aquifer
Phase 6 (August-Sept) sampling Shintech wells and 7 private wells
Phase 7 (Sept) DOW installation of 7 piezometers
Phase 8 (beginning of EPA support with contractor assistance (December-February
2002) 42 sampling locations (32 private wells, 4 new EPA wells, 2 new City sentinel
wells, 4 DOW piezometers)
The EPA Region 6 Multimedia Planning and Permitting Division and the EPA Robert S. Kerr
Environmental Research Center have previously provided technical support at the request of LDEQ for
specific elements of the State's investigation, and on February 20, 2003, LDEQ formally requested
ground-water modeling assistance from the Multimedia Planning and Permitting Division. The goal for
performing ground-water modeling is to integrate all available ground-water related information to
obtain a better overall understanding of short-term and long-term net ground-water flow directions in the
upper sand unit.
ii. Purpose of Model
To complete this modeling project, staff of the U.S. Environmental Protection Agency (EPA)
Region 6, Multimedia Planning and Permitting Division, will develop a ground-water flow model of
limited scope and detail to represent a portion of the upper sand unit of the Plaquemine Aquifer, in
Iberville Parish, Louisiana. The overall purpose for this model is to evaluate and simulate potential
ground-water flow directions in the upper sand unit, given that ground-water flow directions are
influenced by Mississippi River stages and other local and regional aquifer stresses (e.g., pumping wells
and possible regional water movement).
iii. Capabilities and Limitations
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 4 of 34
Similar to all ground-water flow models, this model will represent a simplification of the actual
ground-water flow system. This model is intended to provide a computer generated simulation of
potential ground-water flow directions in the upper sand unit of the Plaquemine Aquifer covering the
project area. The model is not intended to simulate the transport and transformation of contaminants
released from a source area. EPA Region 6 does not believe that adequate information, data, and
understanding of the plume geochemistry is available at this time to attempt chemical transport modeling.
However, the model should assist with delineating possible source areas for the upper sand unit vinyl
chloride plume. This model is not intended to determine remedial measures for the contaminated area.
This modeling effort will simplify and integrate all available ground-water flow information and
data of the upper sand unit, enhance one's overall understanding of the local ground-water flow system,
facilitate hypothesis testing of field data, subsurface physical properties, and boundary conditions,
evaluate short-term and long-term ground-water flow directions, and simulate changing (transient)
hydrologic conditions overtime. The model should help evaluate the effects of pumping wells (public,
private, industrial, etc.) on local ground-water flow. The model will employ the MODFLOW and
MODPATH computer codes; a chemical transport code will not be used.
b. Organizations and Modeling Team
Simulating subsurface phenomena, such as ground-water flow, is a complex process involving
development of a conceptual model of the system, selection of a computer code that is capable of
performing the simulation, transforming aspects of the conceptual model into their mathematical
counterparts, and evaluating the results. Because of the technical complexities associated with
developing a flow model of the upper sand unit, the Multimedia Planning and Permitting Division has
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: • 5 of 34
decided to form a modeling project team containing personnel with an appropriate technical skill mix.
The modeling team will include technical experts from the EPA Multimedia Planning and Permitting
Division, the EPA Robert S. Kerr Environmental Research Center in Ada, Oklahoma, the State of
Louisiana—Louisiana Department of Environmental Quality, and limited contractual support.
Contractual support involves field data collection, literature research, and software support. Members of
the modeling team, or other individuals with expertise directly available to the team, have education and
experience in geology, hydrogeology, hydrology, engineering, mathematics, chemistry, applied ground-
water modeling, and software and data processor training. If, during the course of the this modeling
project additional skills, training, and continuing education are needed, the Agency will seek to fulfill
these additional requirements as appropriate.
Two contract organizations are involved with conducting the literature searches. (1) the EPA
Region 6 Library staffed by ASRC Aerospace Corporation, and (2) Booz, Allen, and Hamilton, Inc.
Software (data processor) support will be obtained from Waterloo Hydrogeologic, Inc. These software
support services are expected to assist with model setup, boundary conditions, calibration procedures,
sensitivity analysis, and final model review.
c. Project Milestones and Schedule
The Multimedia Planning and Permitting Division will perform the modeling related tasks as
indicated below. These tasks are consistent with guidance from a number of organizations, including the
U.S. EPA Office of Solid Waste and Emergency Response (EPA 500-B-94-004, 1994), the U.S. Army
Corps of Engineers— Manual 1110-L-1421 (February 28, 1999), and the American Society for Testing
and Materials (ASTM). The ASTM Subcommittee D 18.21 on Ground-Water and Vadose Zone
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 6 of 34
Investigations has approved six standards related to this modeling project. These standards have been
written in the form of guides (not rigid standards) and include the following publications:
D-5447 Standard Guide for Application of a Ground-Water Flow Model to a Site-
Specific Problem
D-5490 Standard Guide for Comparing Ground-Water Flow Model Simulations to Site-
Specific Information
D-5609 Standard Guide for Defining Boundary Conditions in Ground-Water Modeling
D-5610 Standard Guide for Defining Initial Conditions in Ground-Water Modeling
D-5611 Standard Guide for Conducting a Sensitivity Analysis for a Ground-Water Flow
Model Application
D-5718 Standard Guide for Documenting a Ground-Water Flow Model Application
As summarized by EPA, ASTM, and other information sources, the application of a ground-
water flow model would ideally include several milestones. Milestones specific to the modeling project
at hand are reflected as discrete modeling events as presented in Table 1.
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 7 of 34
Table 1. Project Schedule
Activity
Establish modeling objectives
Obtain approved work plan
Quality Assurance Project Plan
Collect, organize, and interpret available information and data
Prepare conceptual model
Set up numerical (computer) model
Calibrate model
Perform Modflow/Modpath simulations
Conduct post-simulation analysis
Evaluate overall modeling effectiveness
Determine preliminary results
Reiterate model simulations as necessary
Final results and report preparation
Dates
Feb
March
March
April
May
May
May
May
June
July
July
July
August
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 8 of 34
2. Modeling Objectives
The objectives of a modeling study should be clearly specified up front, considering applicable
regulatory and policy issues. Similar to all ground-water flow models, this model will represent a
simplification of the actual ground-water flow system. As stated earlier in the Work Plan, the overall
purpose for this model is to evaluate and simulate potential groundwater flow directions in the upper
sand unit, given that ground-water flow directions are influenced by Mississippi River stages and other
local and regional aquifer stresses (e.g., pumping wells and possible regional water movement). The
resulting flow directions should help determine where contamination came from and where it may go in
the future, and at what approximate rate of movement.
The model is not intended to simulate the transport and transformation of contaminants released
from a source area. The model will employ the MODFLOW and MODPATH computer codes; a
chemical transport code will not be used.
From the modeling purposes given above, criteria can be established which place boundaries on
the modeling objective. First, project objectives should describe exactly what will be modeled,
simulated, under what conditions, and over what time frame. Objectives must remain within the
capabilities of the MODFLOW and MODPATH computer codes, and within the capabilities of the data
processing software package (Visual Modflow). Objectives must be consistent with overall model
framework, modeling approach, model construction, calibration, use, and intended use of results. Data
limitations and weaknesses, and how they are reflected in the computer model should be reflected in the
objectives. Finally, objectives must be consistent with Agency analysis and decision-making needs.
From the above criteria, the specific modeling objectives for this project can be written as
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25, 2003
Page: 9 of 34
provided in Table 2. If necessary, these objectives may be slightly modified during the course of the
project to address any aspects of unforseen data limitations or weaknesses.
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 10 of 34
Table 2. Modeling Objectives
Objective
Goal
Purpose
Determine direction and rate of ground water flow
in the upper sand unit over the defined project area
through rising and falling river stages, through
periods of river level stability, and with possible
regional groundwater flow effects.
To assist in making conclusions
about short/long term net flow,
contaminant source locations,
and if multiple source locations
are likely.
B Evaluate the hydraulic effects of actual and/or
historical groundwater withdrawals from water
wells, as data are available, on upper sand flow
directions over a defined period of time.
Evaluate the level of
significance pumping wells
have on the flow system.
Determine zone of influence of pumping City of
Plaquemine backup water supply wells.
Assess capture zone and help
evaluate risk of wells being
contaminated.
D Determine if historical contaminants discharged
into the upper sands at specific locations along the
course of the Mississippi River may be source areas
of the identified contaminant plume.
To evaluate whether
contaminants may have entered
the aquifer from a position near
the riverbank.
Estimate age, location, and duration of contaminant
release(s) to degree possible with flow modeling.
Helps understand spatial and
temporal relationships of the
plume with water movement.
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 11 of 34
3. Information and Data Collection
Collecting and organizing information and data is critical to the success of this modeling effort.
Sources of information and data to be collected will include published and unpublished information
obtained from a literature search, information derived from the on-going LDEQ phased ground-water
investigation, information from existing EPA Region 6 facility files, and information from the Louisiana
Department of Transportation and Development list of registered water wells.
a. Sources
i. Literature Search
A literature search is currently underway. The main purpose for conducting the literature search
is to acquire written materials that will provide the basis for the conceptual model. The types of
information to be collected include: (a) any relevant regional and local hydrogeological reports, (b) any
previous investigations specifically on the Plaquemine aquifer, (c) available information on groundwater
use including purpose, quantities, and future projections, (d) boring log data and cone penetrometer log
data, (e) monitoring well data, (f) production well data, (g) well construction characteristics, (h)
geophysical data, (i) geologic, hydrologic, and topographic maps and cross-sections, (j) aerial
photographs, (k) land use maps, (1) soil maps, (m) climatic data, (n) Mississippi River stage data, and any
other information that seems relevant to developing this ground-water flow model.
Two organizations are currently involved with conducting the literature search, the EPA Region
6 Library (staffed by ASRC Aerospace Corporation), and Booz, Allen, and Hamilton, Inc. The EPA
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 12 of 34
Library (and Library Network), established in 1971, includes libraries in the Agency's Washington, D.C.
Headquarters, all 10 Regional Offices, and Agency laboratories located throughout the United States.
The combined Library network collection contains a wide range of general information on environmental
protection and management; the basic sciences such as biology and chemistry; the applied sciences such
as engineering and toxicology; and extensive coverage of topics featured in legislative mandates such as
hazardous waste, drinking water, pollution prevention, and toxic substances. The Region 6 Library, at
the request of the project manager, has been providing results from ongoing literature searches for
specified subjects related to the Plaquemine Aquifer, and has already provided valuable reports,
documentation, maps, and other literature material critical to this modeling project. Booz, Allen, and
Hamilton, Inc., under work assignment R06804, RCRA Corrective Action Support for Region 6 States,
has been requested to complete other related literature research activities and make specific inquiries to
the U.S. Army Corps of Engineers, the U.S. Geological Survey, and the Louisiana Department of
Transportation and Development. Booz, Allen, and Hamilton, Inc. has provided a bibliography relating
to the Plaquemine aquifer and is under review by the project manager.
ii. Louisiana DOTD Information
The Louisiana Department of Transportation and Development (DOTD), Water Resources
Section, in cooperation with the Louisiana District of the United States Geological Survey (USGS), has
established and maintains the State's Water Well Registration Program. This program entails a Statewide
inventory of all registered existing and newly-drilled water wells, monitor wells, etc. This program helps
to ensure that (a) wells are properly constructed and sealed to protect against surface contaminants such
as flood water, spills, etc.; (b) ensures clean water for rural residents; (c) provides a means for the State
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 13 of 34
to collect, catalog, store, and disseminate water well construction and drilling data to the general public;
and (d) helps the State with its water management responsibilities.
A list of registered water wells can be generated by choosing a Parish and the Owner's Name, or
by choosing a Township and Range within that Parish. The wells in each list are sorted by owner's
name or by section number then by and owner's name. Information in each listing includes owner's
name, dotd well number, owner's well number, well depth in feet below land surface, well use,
casing size in inches, date of drilling, water level in feet below land surface, date water level was
measured, geologic unit (aquifer name) ,well coordinates (i.e., latitude and longitude in degrees,
minutes, and seconds) and available information codes. From this DOTD information, pertinent
information will be selected for model input data, and the information will undergo a computer file
structure modification process to transform the basic DOTD data into model input data. This process
entails converting DOTD well data into Microsoft Excel, then converting Excel into a text file (.txt) and
space delimited format, then importing .txt file into the data processor. Where data is available, the
following information will then be used by the data processor: well name or ID number, x and y well
coordinates, screen ID number, screen top elevation, screen bottom elevation, pumping schedule (stress
period start/stop time), and pumping rate.
iii. EPA Region 6 Facility Files
The Region 6 RCRA fileroom contains facility files for a large number of industrial
facilities in Region 6, including the DOW Louisiana Operations facility at Plaquemine. These official
Agency RCRA files are subdivided into 4-categories: RCRA Technical Files (TE), RCRA Permit Files
(PE), RCRA Part- B Files (PB), RCRA Enforcement Files (EN), and Confidential Business Information
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 14 of 34
(CBI). RCRA Files for DOW Chemical or any other nearby facilities considered to have pertinent
information will be acquired from the RCRA file room and evaluated. CBI will not be used for this
modeling project.
iv. LDEQ Ground-Water Investigation
The Louisiana Department of Environmental Quality (LDEQ) has undertaken measures to try
and identify the source of the vinyl chloride contamination in the Plaquemine Aquifer. LDEQ's work,
on-going since approximately April 2001, has involved multiple well sampling events, chemical
analyses, installing wells and piezometers, evaluating water quality information from DOW chemical
and other industrial facilities, and working with members of the Plaquemine, Louisiana community.
During the course of LDEQ's investigation, LDEQ requested that the Multimedia Planning and
Permitting Division provide technical support, which Region 6 has supplied through EPA Work
Assignment R06084, RCRA Corrective Action Support to Region 6 States. This Work Assignment has
resulted in contractor deliverables containing valuable information that will be reviewed for information
and data related to model construction and calibration. The EPA Work Assignment Manager (WAM)
retains contractor deliverables per records management requirements, and the WAM will be the primary
source of information and data resulting from EPA's previous support to LDEQ.
Information and data files independently developed and maintained by LDEQ will be made
available as necessary to support this modeling effort. Further, any related information retained by the
EPA Robert S. Kerr Environmental Research Center in Ada, Oklahoma, resulting from its support to
LDEQ, will also be made available for review purposes.
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 15 of 34
4. Conceptual Model Development
The main purpose of the conceptual model is to present the hydrogeologic framework and flow
system of the upper sand unit, based on all available existing information and data, in simplified
qualitative terms that can be efficiently translated into the numerical model. The steps involved in
conceptual model development are be: (a) researching and evaluating various aspects about the physical
hydrogeologic system, (b) performing an integrated interpretation of all relevant information, (c)
documenting weaknesses and uncertainty of the conceptual model, and (d) presenting the conceptual
model.
a. Elements to Research and Evaluate
The conceptual model does not necessarily need to restate all of the information known about the
region being modeled. The conceptual model may be described in terms of the assumptions made to
simplify the system, including data gaps and their impact on the modeling results. Information that will
be evaluated with respect to the conceptual model includes the following depending on information
availability.
The geologic and hydrogeologic system will be researched in detail. This includes the regional
and local hydrogeological framework, the relationship and extent of hydrogeologic units, including
lithologic contacts, facies changes, discrete features, and spatial variations of geologic units and their
hydraulic properties. The hydraulic boundaries of the system will be evaluated, and if available, a water
budget analysis (evapotranspiration, runoff, pumping and recharge rates) will be reviewed. Other
information to be researched and evaluated includes aquifer material properties (porosity, hydraulic
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25, 2003
Page: 16 of 34
conductivity, storativity, isotropy, degree of aquifer heterogeneity).
A number of supporting figures and graphics may be evaluated and included with the conceptual
model depending on availability. These include the following:
(1) Study area location map
(2) Geologic map and cross sections
(3) Topographic maps
(4) Maps of tops and bottoms of aquifer and confining units
(5) Isopach maps of hydrostratigraphic units
(6) Maps showing extent and thickness of river bottom sediment
(7) Maps of river levels and depths to bottom
(8) Maps indicating any discrete geologic features affecting water flow (e.g.,
salt domes or faults)
(9) Potentiometric surface maps of upper sand and hydraulic boundaries
(10) Maps and cross sections showing hydraulic conductivity of upper sand
and confining unit
b. Perform Integrated Interpretation
Following the review and assessment of the types of information described above, a integrated
interpretation will be performed. The purpose of making this interpretation is to combine collected
information and data to produce an accurate interpretation of site characteristics. Decision making
relative to the integrated interpretation will be based on best professional judgement.
c. Documenting Uncertainty
When developing this model, it will be important to document the quality, quantity, and
completeness of information and data upon which model is based. For the conceptual model, there will
be aspects of the research into the upper sand unit flow system that lack adequate definition. Thus a level
of uncertainty will likely be introduced into the conceptual model. Where these uncertainties exist in the
conceptual model, they will be explained and documented to an appropriate degree.
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 17 of 34
d. Presenting the Conceptual Model
The conceptual model will be completed and presented by preparing a written description with
complementary graphical illustrations. Most of these illustrations will be reproduced from published
research. Possible graphics include simplified hydrogeologic cross sections, potentiometric surface
maps, structure maps, multi-dimensional graphics, and water balance diagrams. Part of the narrative
description will likely include discussing the role of modeling and hypothesis testing relative to the
completeness of the conceptual model.
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 18 of 34
5. Numerical Model Development
a. MODFLOW AND MODPATH
This section introduces three computer programs (MODFLOW, MODPATH, and Visual
Modflow) which will be utilized in this project.
Many of the essential elements, properties, and numerical values that will enable these three
computer programs to run are described in the Project Work Plan. MODFLOW is a Modular Three-
Dimensional Finite-Difference Ground Water Flow Model that was developed by the U.S. Geological
Survey (McDonald and Harbaugh, 1988; Harbaugh and McDonald, 1996) during the early 1980s.
MODFLOW is the world-wide standard groundwater flow modeling program because of its ability to
simulate a wide variety of groundwater systems, its extensive publically available documentation, and its
rigorous USGS peer review. MODFLOW does not contain a mass transport component. When properly
utilized, MODFLOW is the standard model used by courts, regulatory agencies, universities, consultants,
and industry.
MODFLOW is designed to simulate aquifer systems in which (1) saturated-flow conditions
exist, (2) Darcy's Law applies, (3) the density of ground water is constant, and (4) the principal
directions of horizontal hydraulic conductivity or transmissivity do not vary within the system. These
conditions are met for many aquifer systems for which there is an interest in analysis of ground-water
flow and contaminant movement. For these systems, MODFLOW can simulate a wide variety of
hydrologic features and processes. Steady-state and transient flow can be simulated in unconfined
aquifers, confined aquifers, and confining units. A variety of features and processes such as rivers,
streams, drains, springs, reservoirs, wells, evapotranspiration, and recharge from precipitation and
irrigation also can be simulated. At least four different solution methods have been implemented for
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 19 of 34
solving the finite-difference equations that MODFLOW constructs. The availability of different solution
approaches allows model users to select the most efficient method for their problem.
MODFLOW simulates ground-water flow in aquifer systems using the finite-difference method.
In this method, an aquifer system is divided into rectangular blocks by a grid. The grid of blocks is
organized by rows, columns, and layers, and each block is commonly called a "cell." For each cell
within the volume of the aquifer system, the user must specify aquifer properties. Also, the user specifies
information relating to wells, rivers, and other inflow and outflow features for cells corresponding to the
location of the features. For example, if the interaction between a river and an aquifer system is
simulated, then for each cell traversed by the river, input information includes layer, row, and column
indices; river stage; and hydraulic properties of the river bed.
MODFLOW uses the input to construct and solve equations of ground-water flow in the aquifer
system. The solution consists of head (ground-water level) at every cell in the aquifer system (except for
cells where head was specified as known in the input data sets) at intervals called "time steps." The head
can be printed and (or) saved on a computer storage device for any time step. Hydrologists commonly use
water levels from a model layer to construct contour maps for comparison with similar maps drawn from
field data. They also compare computed water levels at individual cells with measured water levels from
wells at corresponding locations to determine mode error. The process of adjusting the model input
values to reduce the model error is referred to as model calibration.
In addition to water levels, MODFLOW prints a water budget for the entire aquifer system. The
budget lists inflow to and outflow from the aquifer system for all hydrologic features that add or remove
water. Other program output consists of flow rates for each model cell. MODFLOW can write the flow
rates onto a computer storage device for any hydrologic feature in a simulation. These cell-by-cell flow
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 20 of 34
rates commonly are read by post-processing programs for detailed analysis of the simulated ground-water
system.
In addition to MODFLOW, a program called MODPATH (Pollock, 1989) will be utilized for
particle tracking. MODPATH is a particle tracking post-processing package designed to work with
MODFLOW. Output from steady-state or transient MODFLOW simulations is used in MODPATH to
compute paths for imaginary "particles" of water moving through the simulated groundwater system.
MODPATH also keeps track of the time of travel for particles moving through the system. By carefully
determining the starting position of particles, it is possible to use MODPATH to perform a wide range of
analyses, such as delineating capture and recharge areas or drawing flow nets.
To assist with running the MODFLOW and MODPATH programs, a data processor will be
utilized called Visual Modflow. Visual Modflow is a proprietary modeling program produced by
Waterloo Hydrogeologic Inc., and is designed to facilitate model development, data input, calibration,
and the visualization of model output. Visual Modflow is considered a fully-integrated groundwater
modeling environment which allows the user to graphically design the model grid, properties and
boundary conditions, visualize the model input parameters in two or three dimensions, run the
groundwater flow, and pathline simulations. The hardware requirements for running Visual Modflow are
a Pentium-based computer, 32 MB or RAM (64 is recommended), CD ROM drive, a hard drive with at
least 100 Mbytes free, and Windows 95/Windows 98/Windows NT 4.0 (Service Pack 3).
Visual Modflow has three main modules: the Input Module, Run Module, and Output Module.
The Input Module allows the user to graphically assign all of the necessary input parameters for building
a three-dimensional groundwater flow model. The input menus represent the basic model building
blocks for assembling a data set for MODFLOW, MODPATH, and ZoneBudget. The menus are
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25, 2003
Page: 21 of 34
displayed in logical order and guide the modeler through steps necessary to design a groundwater flow
model. In the Run Module, the user parameters and options which are run-specific. These include
selecting initial head estimates, setting solver parameters, activating the re-wetting package, specifying
output control, etc. Each of these menu selections has default settings which may be changed by the
modeler as warranted. The Output Module allows the user to display modeling and calibration results,
and allow the user to select, customize, and overlay various display options for presenting modeling
results.
b. Numerical Model Framework
i. Model Domain
The platform upon which the model will be constructed will be a high-quality aerial photograph
(bitmap) of the Plaquemine, Louisiana area. This bitmap, serving as the model domain, will
encompasses the approximate area between DOW Chemical Company and the City of Plaquemine,
Louisiana along the western side of the Mississippi River in Iberville Parish. The exact area the domain
will encompass will be determined by the modeling team.
The bitmap image will be supplied by the Region 6 GIS support group. Once the image is
acquired and determined by the modeling team to encompass the appropriate area, it will then be
prepared by the project manager for use within the data processor. The preparation involves image
refinement in terms of resizing, rotating, and aligning the domain (as necessary) within the data
processor, and geo-referencing the image to a coordinate system. Geo-referencing will be accomplished
by selecting three physiographic features on the image and determining their real world coordinates. The
real world coordinates will be entered into the data processor for each physiographic feature.
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 22 of 34
From previous work, the coordinates for three physiographic features are already available and
may be used for this modeling project. These feature coordinates are indicated in Table 3.
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 23 of 34
Table 3. Geo-reference Coordinates
Feature
1
2
3
UTM-X
664958.976229
673307.304083
668409.562597
UTM-Y
3351685.132390
3350383.535420
3354828.151360
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 24 of 34
The data processor requires only two geo-reference points to establish domain coordinates. However, a
third geo-reference point will be utilized as a cross-check and verification tool to ensure the coordinate
system is operating correctly.
ii. Model Layers
The hydrostratigraphic unit to be modeled is the Plaquemine Aquifer upper sand unit. This unit
may be represented as a single layer within the model. However, the conceptual model will ultimately
determine the number of layers to be used within the numerical model.
iii. Ground-Water Wells
The data processor is capable of importing information on existing water wells including well
depth, pumping schedule, screened interval, pumping rates, and x-y ground coordinates. The Louisiana
Department of Transportation and Development (DOTD), Water Resources Section, in cooperation with
the Louisiana District of the United States Geological Survey (USGS), has established and maintains the
State's Water Well Registration Program. This program entails a Statewide inventory of all registered
existing and newly-drilled water wells, monitor wells, etc. This program helps to ensure that (a) wells are
properly constructed and sealed to protect against surface contaminants such as flood water, spills, etc.;
(b) ensures clean water for rural residents; (c) provides a means for the State to collect, catalog, store,
and disseminate water well construction and drilling data to the general public; and (d) helps the State
with its water management responsibilities.
In addition to water well location information available from DOTD, more detailed information
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 25 of 34
over the project area may be contained within files located at LDEQ. During the March-June 2001time
frame, LDEQ performed a neighborhood/local business survey of water wells within the area affected by
the vinyl chloride plume at that time. Any information from this survey will be utilized for modeling.
From DOTD's database, a list of Registered water wells can be generated by choosing a Parish
and the Owner's Name, or by choosing a Township and Range within that Parish. The wells in each list
are sorted by owner's name or by section number then by and owner's name. Information in each listing
includes owner's name, DOTD well number, owner's well number, well depth in feet below land surface,
well use, casing size in inches, date of drilling, water level in feet below land surface, date water level
was measured, geologic unit (aquifer name) ,well coordinates (i.e., latitude and longitude in degrees,
minutes, and seconds) and available information codes. From this DOTD information, pertinent
information will be selected for model input data, and the information will undergo a computer file
structure modification process to transform the basic DOTD data into model input data. This process
entails converting DOTD well data into Microsoft Excel, then converting Excel into a text file (.txt) and
space delimited format, then importing .txt file into the data processor. Where data is available, the
following information will then be used by the data processor: well name or ID number, x and y well
coordinates, screen ID number, screen top elevation, screen bottom elevation, pumping schedule (stress
period start/stop time), and pumping rate.
iv. Hydraulic Properties
The data processor allows the input and editing of certain hydrogeological properties to model
layers and zones. These properties are hydraulic conductivity, specific storage, specific yield, effective
porosity, and initial heads. Values or ranges of values for each property will be determined as part of the
conceptual model. A brief description of each property is below.
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 26 of 34
Hydraulic conductivity (K) is a measure of a porous materials ability to transmit fluid. It has
relatively high values for sand and gravel and relatively low values for clays and rock. Sediments with
high K-values are more able to conduct water than sediments with low values. Typical values are 10~2
cm/sec for a medium sand, 10"9 cm/sec for an unfractured clay, 10~n cm/sec for unfractured granite, and
10"10 cm/sec for plastic liners. Hydraulic conductivity is used in conjunction with applications of Darcy's
Law for determining ground-water velocity. Darcy's law is a derived formula representing the flow of
fluids through a porous material. The real or seepage velocity (V^ is calculated by the equation below
where nef= effective porosity, K = hydraulic conductivity, and / = hydraulic gradient:
T7 ~K •
VR = - /
Effective porosity («e/) is used by MODPATH to determine the average linear ground-water
velocities for use in time-dependent functions such as time markers along pathlines. Effective porosity is
not used in MODFLOW simulations.
Specific storage (5"s) of a saturated aquifer is the volume of water that a unit volume of aquifer
releases from storage because of expansion of water and compression of the aquifer under a unit decline
in hydraulic head. Using (Ss), the data processor determines the primary storage coefficient used by
MODFLOW. The primary storage coefficient is calculated to be equal to the specific storage multiplied
by the layer thickness.
Specific yield (SY) is the storage term used for unconfined aquifers. SY is the ratio of the volume
of water that drains from saturated material due to gravity to the total volume of the material. For sand
and gravel aquifers the specific yield may is generally equal to the porosity. This is a unitless value with
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 27 of 34
typical values for sands and gravels being 0.2 - 0.35.
Fw
V tota
For a transient simulation, MODFLOW needs a starting head distribution, i.e., initial heads or
starting heads. Values for initial heads will be determined from the conceptual model. These heads are
used for head calculation only in the first time step, but may be saved and used to calculate well
drawdown, the difference between the starting head distribution and some later head distribution.
v. Boundary Conditions
Boundary conditions defined in the numerical model will result from interpretations made from
the conceptual model. The most obvious boundary condition will represent the Mississippi River. For
the transient river boundary, boundary condition criteria! will be specified. These criteria are: start/stop
time, river stage elevation for a starting location, river stage elevation for an ending location, the
elevation of the river bottom, and conductance of water to/from the river.
Other boundary conditions designated in the model will possibly include aquifer recharge,
natural or man-made discharge points/drains, evapotranspiration, and constant head or general head
boundaries.
vi. Particle Tracking
Particles used by MODPATH will be assigned in various locations within the model domain as
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 28 of 34
necessary to determine net flow directions and travel time. Both forward and backward particle tracking
will be performed as necessary to bolster the modeling study. Particle discharge times will be
determined during the actual numerical model construction.
c. Model Output and Calibration
i. Output Options
Following MODFLOW and MODPATH runs, the data processor allows a number of output
options. These include:
(1) General contouring options
(a) Head equipotentials (head values in each cell)
(b) Head difference (differences in head between selected layers)
(c) Head flux (fluxes of water between adjacent layers)
(d) Drawdown (differences between initial head and calculated
head)
(e) Elevation (elevations of cell bottoms or tops)
(f) Net recharge (specified recharge values minus calculated
evapotranspiration)
(g) Water table (head values in uppermost active cell)
o
(2) Ground-water velocity-vector options
(a) Projection (view velocity projections)
(b) Direction (flow directions with not-to-scale velocity vectors)
(c) Magnitude (flow directions with scaled vectors
(3) Pathlines
(a) Time related for transient system
(b) Show in all layers
ii. Calibration
The purpose for calibrating this model is to produce simulated water level results that are
generally consistent with field measurements. Model calibration procedures will be accomplished by
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 29 of 34
utilizing software functions integral to the data processor, producing statistically derived graphs and
plots, and by making adjustments through model iterations to minimize differences between simulated
and .observed values. Data sources for calibration include published and unpublished water level data,
regional and local water level data, any available data from municipal and private industry sources, and
recent water level data collected by EPA and LDEQ. Water level data collected from certain wells in the
field will be selected to represent observation wells within the data processor to facilitate model
calibration.
Following a Modflow run, the head equipotential option will be selected as model output and
head equipotential contours will be displayed along with a calibration plots dialog box. Within the
calibration plots dialog box the user can select head observation wells and the type of calibration
statistic/graph to view. Available graphs include a calculated versus observed head graph, a residual
distribution graph, and a calculated/observed heads overlay. The calculated/observed heads overlay
allows the data processor user to visualize the differences between calculated and observed values while
viewing the plan view of the model domain. Scaled symbols are used to display calibration residuals
whereby the size and color of each symbol will depend on the difference between the calculated value
and observed value.
The purpose of the calculated versus observed graph is to a graphically represent the quality of
fit between observed data and the calculated results from the model. This graph provides an indication of
how well the modeling effort has simulated observed field conditions. Several calibration statistics may
be produced from this plot including the mean error, the mean absolute error, the standard error of the
estimate, the root mean squared (RMS), normalized RMS, interpolated versus extrapolated, and the 95%
confidence interval. The following statistical equations are presented in more detail in the data processor
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 30 of 34
user manual (Waterloo Hydrogeologic, Inc., 2000).
The Mean Error is defined by the equation:
1 "
MeanError = — V (Xcalc - Xobs)i
n±f
where Xobs is the observed value and Xcalc is the calculated value for a data series. The Mean Absolute
error is the same as the Mean Error except that the absolute values of each calculated and observed head
difference, are summed.
MeanAbsoluteError = — ^ [Xcalc -
i=l
The Standard Error of the Estimate (S.E.E.) is provided by (this error estimate is also referred to as the
calibration residual:
(Xcalc - Xobs)] - £ (Xcalc - Xobs)t
S.E.E.= \
n-\
The Root Mean Squared (RMS) is given by:
1 "
RMS = - jy (Xcalc - Xobs)2
The Normalized Root Mean Squared error (Normalized RMS) is given by the RMS divided by the
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 31 of 34
maximum difference in the observed head values:
NormalizedRMS =
RMS
Viewing the 95% Confidence Interval allows the user to see a range of calculated values for each
observed value. Within this interval, the modeler can be 95% confident that simulation results are
acceptable for a given observed value.
D 4.1 o*
Range = ±1.96 •
- Xcalc)2
- +
(Xobs - Xaveobs)2
'
n-2
+ 1
The residual distribution graph displays the residual distribution for selected observation wells.
This graph depicts the population, frequency, or relative frequency of observations for specified intervals
of normalized calibration residual values. The head versus time graph displays the head versus time for
selected observation wells. This graph presents a time series plot of observed and calculated heads for
each observation point selected. The statistics versus time graph include the normalized RMS versus
time, residuals versus time, normalized residuals versus time, and error versus time.
Regional data will serve to tie in the model boundaries with any regional flow system that may
encompass and exceed the model domain, as documented by the conceptual model, and will provide a
general range of water levels expected to lie within the model domain. It is anticipated that published
regional groundwater flow data may show a general trend for higher head values to the northeast, and
decreasing head values towards the south or southwest. Examples of this trend are contained in McGee
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25, 2003
Page: 32 of 34
(1997) and Whiteman (1972). McGee's work, Figure 2, Table 2, shows selected wells and water levels
completed in the Mississippi River alluvial aquifer and adjacent aquifers in Louisiana, with a general
trend of decreasing water levels from Northern Louisiana to Southern Louisiana. Whiteman identified
and described a small net movement of water toward the river from the northeast and a net movement of
water away from the river to the southwest within the area of his 1972 study.
Within the model domain calibration will be more precise. Calibration to historical water levels
will be achieved through an iterative process of comparing observed to simulated values, adjusting values
for hydraulic conductivity, storage, specific yield, effective porosity, and total porosity, to minimize
differences between observed and simulated values. Calibration will be checked during each model
stress period.
d. Sensitivity Analysis
Some of the modeling scenarios for the upper sand unit will involve input parameters that can
vary over a considerable range. For this reason, the sensitivity of model predictions to variations of
parameter values should be evaluated. Once parameter values/ranges of values are determined, the
modeling team will decide on an approach for conducting the sensitivity analysis. The sensitivity
analysis is likely to involve the following:
Determining the rationale for selecting parameters for the sensitivity analysis; emphasis
will be given to parameters for which there is a large degree of uncertainty
Determining the range to be tested for each parameter and determining the number of
model simulations to be conducted for each parameter
Determining whether sufficient simulations to investigate parameter values and ranges
of values have been conducted
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 33 of 34
Evaluating the sensitivity of model calibration quality and model predictions to
variations in parameter values, including grid spacing, time steps, and boundary
conditions
Assessing the relevance of the overall uncertainty and sensitivity with respect to the
objectives modeling project.
-------
Plaquemine Aquifer Ground-Water Flow Model
EPA Region 6 Work Plan
Draft Date: March 25,2003
Page: 34 of 34
6. Final Modeling Report
Suggested contents
i. Title page
ii. Table of contents
iii. List of figures
iv. List of tables
v. Executive summary
vi. Introduction
vii. Model objectives
viii. Hydrogeologic characterization
ix. Conceptual model
x. Modflow/Modpath evaluation
xi. Input parameters and model framework
xii. Model calibration
xiii. Sensitivity analysis
xiv. Simulations performed
xv. Conclusions and recommendations
xvi. References
xvii. Tables
xviii. Figures
xix. Well data
xx. Additional data
------- |