GROUND-WATER FLOW DIRECTIONS AND
CONTAMINANT SOURCE AREA EVALUATION
FOR THE PLAQUEMINE AQUIFER
REGION 6
-------
GROUND-WATER FLOW DIRECTIONS AND
CONTAMINANT SOURCE AREA EVALUATION
FOR THE PLAQUEMINE AQUIFER
October 7,2004
Prepared For:
The Louisiana Department of Environmental Quality
Office of Environmental Assessment
Prepared By;
Scott Ellinger, M.S., P.G.
U.S. Environmental Protection Agency Region 6
Multimedia Planning and Permitting Division
-------
EXECUTIVE SUMMARY
GROUND-WATER FLOW DIRECTIONS AND CONTAMINANT SOURCE AREA
EVALUATION FOR THE PLAQUEMINE AQUIFER
SCOTT ELLINGER, M.S., P.G.
U.S. ENVIRONMENTAL PROTECTION AGENCY REGION 6
MULTIMEDIA PLANNING AND PERMITTING DIVISION
October 7,2004
This report presents the U.S. Environmental Protection Agency Region 6, (EPA) ground-
water model and contaminant source area evaluation performed during 2003-2004 related to
ground-water contamination near Plaquemine, Louisiana. This report contains information on
local ground-water flow directions, discusses the likely significant source area for ground-water
contamination, and also contains general recommendations with respect to future water well
pumping.
In 1997 and 1998, the Louisiana Department of Health and Hospitals (LDHH) sampled
water wells at the Myrtle Grove Trailer Park in Plaquemine, and detected levels of vinyl chloride
and cis 1,2 dichloroethylene, which exceeded Maximum Contaminant Levels (MCLs) (2 ppb for
vinyl chloride; 70 ppb for cis 1,2 dichloroethylene). The two chemicals were detected again
during the following LDHH sampling period, which is every three years for public water supply
wells, in March 2001. Since 2001, extensive ground-water sampling data has been collected by
the combined efforts of the Louisiana Department of Environmental Quality (LDEQ) and EPA,
to delineate the extent of contamination, determine the source location, and to assess whether
contaminant concentrations are changing over time. Sampling and analysis have also been
performed by DOW Chemical Company in accordance with LDEQ guidance and oversight.
Since becoming aware of the vinyl chloride contamination, LDEQ has taken significant
actions to ensure that no one is drinking contaminated water, and has conducted door-to-door
visits, public meetings, and provided general information on the extent of contamination to
citizens. Additional work by EPA has included the coordination of reviews of the vinyl chloride
analytical laboratory data with hydrologists and other scientists at the EPA National Risk
Management Research Laboratory (NRMRL) in Ada, Oklahoma. In addition, the Agency for
Toxic Substances and Disease Registry (ATSDR) conducted a Health Consultation evaluating
the health effects of past exposures to vinyl chloride and other substances detected in Myrtle
Grove Trailer Park wells. The city of Plaquemine has installed monitoring wells to serve as
advance warning devices if contaminants are detected in the area of the city's backup water
supply wells.
Those past efforts, however, have not provided any conclusive evidence regarding the
most outstanding Myrtle Grove contamination issue to date: determining the contaminant source
location. The purpose of EPA's ground-water modeling and contaminant source area evaluation
is to answer certain fundamental questions related to complex ground-water flow directions, and
-------
Plaauemine Aquifer Report
10/7/04
Page: 3 of 43
to evaluate potential and most likely contaminant source areas within this complex ground-water
flow regime. One of the main reasons why ground-water modeling was used is because a
computer has the ability to simulate the direction of net ground-water flow under complex site
conditions.
In June 2001, LDEQ requested technical assistance from EPA Region 6 and NRMRL
regarding contaminant degradation and source identification. Additional ground-water modeling
objectives were provided by LDEQ in February 2003. Project planning included discussions on
the modeling project approach between individuals at EPA, LDEQ, and NRMRL. Project
objectives focused on determining the directions of local flow in the Plaquemine aquifer upper
sand unit, understanding how flow is affected by the Mississippi River and by pumping wells,
and using flow directions to help determine the likely significant source area location. The
resulting computer model captured the fundamental ground-water flow attributes of the affected
part of the Plaquemine aquifer upper sand unit, and is used for estimating the range of ground-
water flow conditions within the range of site complexity and available historical and physical
data.
Below is a discussion of how each modeling objective was met, and a conclusion based
on the data related to each objective:
Objective 1: To determine potential directions of ground-water flow in the Plaquemine
aquifer upper sand unit over the project area.
The net direction of ground-water flow in the Plaquemine aquifer upper sand unit without
pumpage was determined by using the Modpath program and particle tracking pathlines over a
recent 600-day period. The modeling results show that the net flow direction, without pumpage,
is primarily to the west. The net direction of ground-water flow in the Plaquemine aquifer upper
sand unit calculated by the modeling is consistent with previously published information from
the U.S. Geological Survey and the Louisiana Geological Survey.
Objective 2: To understand how ground-water flow is affected by aquifer interaction with
the Mississippi River, pumping wells, and by possible regional ground-water
flow gradients.
The results of the modeling indicate that under the majority of the modeling conditions
with pumpage, the gross direction of ground-water flow in the plume area is westerly. Flow
directions on very localized scales may be different, because flow is directly related to well
locations, variable pumping rates and schedules, and river stage level. Pumping wells have the
greatest influence over ground-water flow directions when the river stage is low. However, this
modeling evaluation suggests that a gross westerly flow direction still exists in the plume area
even when high capacity wells are pumping.
-------
Plaauemine Aquifer Report
10/7/04
Page: 4 of 43
Objective 3: To the degree possible with flow modeling, evaluate the likelihood of possible
contaminant source locations, and whether multiple source locations are
possible.
In order to meet Objective 3, and draw conclusions about the potential sources of
contamination in the Plaquemine aquifer, an evaluation of the modeling results and conclusions
was combined with an evaluation of potential source areas in the modeling domain.
The main modeling results and conclusions that were evaluated include the net flow direction
without pumpage, flow as affected by river stage changes, and flow with well pumpage. To
supplement the evaluation of the likely significant location for a contaminant source based on the
modeling, other historical information was considered, which included a previous evaluation by
LDEQ of many potential source areas.
Based on the ground-water flow modeling, including observations about the local
geology, hydrogeology, river hydrology, combined with information about known chemical
releases and LDEQ's previous investigation of potential source areas and releases, only one
geographic area was identified that is a likely significant source area for the contamination in the
Plaquemine aquifer. This particular area is near the northeastern edge of the existing
contamination plume and along the western side of the Mississippi River. Previous chemical
analyses at this location indicate the presence of compounds that have the potential to degrade
into vinyl chloride in soil, sediments, and ground-water. In addition, historical information
indicates that perchloroethylene was discharged to the Mississippi River at this location in the
past, and the river is in direct contact with the contaminated part of the Plaquemine aquifer.
Perchloroethylene is one of the chemicals that can degrade to vinyl chloride.
A scenario of the mechanism of transport of the contaminants from the likely significant
source area was also constructed to ensure that the results of these evaluations fit the physical
facts of transport in this river-influenced transport regime. Surface contamination at the likely
significant source area probably became scoured, eroded, and transported southward by river
water causing the original source area to become enlarged. Soluble contamination was
transported downward into the Plaquemine aquifer, and could have passed through sediments
and preferential pathways along the courses of existing industrial pipelines. Contamination
reaching the Plaquemine aquifer upper sand unit would then have been transported and dispersed
from the river in a westerly direction, moving underneath Myrtle Grove trailer park and other
impacted locations of the Plaquemine, Louisiana area.
-------
Plaouemine Aquifer Report
10/7/04
Page: 5 of 43
Objective 4: To estimate age and location of contaminant release(s) to degree possible
with flow modeling, and plume movement with flow and/or contaminant fate
and transport modeling, if possible.
The age of the contaminant release was estimated by using an average ground-water
velocity of Ift/d, and by back calculating, based on the net flow direction, the amount of time it
would have taken the contamination to travel from the likely significant source area to the current
plume location. A 20-40 year time frame was estimated for contamination to reach the Myrtle
Grove wells and the plume area from the likely significant source area, or from a zone of possible
source area spreading, located adjacent to this source area. This time estimate contains a number
of variables, however, it is considered to be adequate for the purpose of gaining an idea of the
order of magnitude of the time that it would take for a contaminant released from the likely
significant source area to reach the Myrtle Grove wells.
Uncertainties discussed in the report include modeling uncertainties, and uncertainties
related to the conclusions. Uncertainties were addressed by using all available site-specific data
and published regional information, and by filling data gaps using the professional judgement of
members of the modeling team and peer reviewers. The principal uncertainties in the ground-
water flow model include the specification of general head boundaries, the determination of
aquifer property values for use in the model, and the method of inclusion of water well pumping
rates and schedules. Other uncertainties related to conclusions involve rates of erosion, and the
rates of transport for sediment and contamination affected by fluvial processes. Modeling
sensitivities are discussed with regard to the principal sensitivities affecting ground-water flow
directions. Sensitivities were identified for boundary conditions, aquifer property values, and
water well pumping data.
Developing a calibrated contaminant fate and transport model was outside the capability
of existing data. However, the ground-water flow modeling described in this report may serve as
a basis for a future calibrated fate and transport model, if more complete fate and transport data
become available.
-------
Plaauemine Aquifer Report
10/7/04
Page: 6 of 43
TABLE OF CONTENTS
EXECUTIVE SUMMARY 2
TABLE OF CONTENTS 6
LIST OFFIGURES 8
I. BACKGROUND 9
A. Introduction 9
B. Objectives 11
H. MODEL CONSTRUCTION SUMMARY 12
A. Model Grid and Layering 13
B. Boundary Conditions 14
C. Calibration 16
EL MODELING RESULTS AND SOURCE AREA EVALUATION 20
A. Potential Directions of Ground-water Flow 20
B. Aquifer/River Interactions, Pumping Wells, and Regional Flow Gradients 23
C. Flow Directions and Possible Source Locations 27
1. Observations and Results 27
2. Previous Source Area Investigation and Evaluation 28
3. Current Source Area Identification 31
4. Likely Significant Source Area Mechanism of Transport 33
D. Age of Contaminant Release and Plume Movement 34
IV. MODELING SENSITIVITIES AND UNCERTAINTIES 36
A. Ground-Water Flow Modeling Sensitivities 36
1. General Head Boundary Sensitivity 36
2. Aquifer Property Sensitivity 37
3. Sensitivity to Site-Specific Pumping Data 37
B. Ground-Water Flow Modeling Uncertainties 38
1. General Head Boundary Uncertainty 38
2. Uncertainty in Setting Aquifer Property Values 38
3. Uncertainty in Water Well Pumping Parameters 39
4. Source Area Identification Uncertainty 39
-------
Plaauemine Aquifer Report
10/7/04
Page: 7 of 43
V. CONCLUSIONS AND RECOMMENDATIONS 40
A. Directions of Ground-Water Flow 40
B. Likely Significant Source Area 40
C. General Recommendations Including Water Well Pumping 41
VI. REFERENCES 42
APPENDIX 43
A: Technical Background Document and Conceptual Model
B: Combined Quality Assurance Project Plan and General Workplan
C: Correspondence-Model Reviews and Refinements
-------
Plauuernine Aquifer Report
10/7/04
Page: 8 of 43
LIST OF FIGURES *
Figure 1: Project area location map 9
Figure 2: Location of Myrtle Grove water wells 10
Figure 3: Locations of most of the wells and piezometers used in the report 12
Figure 4: Mississippi River Docks I and n showing plotted river stage 13
Figure 5: Locations of general head boundaries 14
Figure 6: Model west-east cross-section 15
Figure 7: Multiple well time-series graph. 18
Figure 8: Approximate ground-water flow at high river stage (no pumping) 21
Figure 9: Approximate ground-water flow at low river stage (no pumping) 21
Figure 10: Net direction of ground-water flow 22
Figure 11: Pumping wells in the model domain by type 24
Figure 12: Flow at high river stage based on wells pumping at a low rate 25
Figure 13: Flow at low river stage based on wells pumping at a low rate 25
Figure 14: Net flow (all river stages) 26
Figure 15: Example particle tracking pathline 26
Figure 16: Locations of multiple sites, areas, and facilities evaluated 30
Figure 17: Likely significant contaminant source area location 32
Figure 18: Detections of vinyl chloride and cis-1,2 dichlorethene 35
These figures were compiled using a variety of computer programs. Due to related
complexities, map scales have not been included with figures.
-------
Platiuemine Aquifer Report
10/7/04
Page: 9 of 43
I. BACKGROUND
A. Introduction
This document presents the results of ground-water modeling and contaminant source
area evaluation performed by the U.S. Environmental Protection Agency Region 6 (EPA),
Multimedia Planning and Permitting Division, for a portion of the Plaquemine aquifer upper
sand unit in eastern Iberville Parish and southern West Baton Rouge Parish, near Plaquemine,
Louisiana (Figure-1). The modeling was performed'so that EPA could evaluate the ground-water
flow in this portion of the Plaquemine aquifer, and to make certain observations and conclusions
about the likely significant source of contamination. The Plaquemine aquifer is described in the
Technical Background Document and Conceptual Model, Appendix A.
M.O..O.C
..... X
»
Figure 1: Project area location map.
-------
Plaquemine Aquifer Report
10/7/04
Page: 10 of 43
In 1997 and 1998, the Louisiana Department of Health and Hospitals (LDHH) sampled
water wells at the Myrtle Grove Trailer Park in Plaquemine, eastern Iberville Parish, Louisiana
(Figure-2), and detected levels of two contaminants, vinyl chloride and cis 1,2 dichloroethylene,
which exceeded Safe Drinking Water Act standards called Maximum Contaminant Levels
(MCLs). These two chemicals were detected again during a March 2001 sampling event of the
same wells. LDHH conducts ground-water sampling and analysis every three years for public
water supply wells.
Figure 2: Location of Myrtle Grove water wells.
Following these detections, Myrtle Grove trailer park wells were immediately removed
from service, and trailer park residents were provided with alternative water supplies.
Concurrently, the EPA, the Louisiana Department of Environmental Quality (LDEQ), the city of
Plaquemine, and the LDHH took steps to ensure that human health was protected, and to
investigate the nature and extent of the contamination. Since 2001, the EPA and LDEQ have
conducted extensive ground-water sampling to collect data and assess the nature and extent of
the contamination. Sampling and analysis have also been performed by DOW Chemical
Company. EPA Region 6, EPA's National Risk Management Research Laboratory (NRMRL),
and LDEQ have used these data to delineate the extent of contamination, evaluate potential
source locations, and assess whether contaminant concentrations are changing over time so that
drinking water wells can be protected. In addition, LDEQ has conducted door-to-door visits,
public meetings, and provided general information on the extent of contamination to citizens.
-------
Plauuemine Aquifer Report
10/7/04
Page: 11 of 43
The city of Plaquemine has installed monitoring wells to serve as advance warning, if
contaminants are detected in the area of the city's backup water supply wells.
During 2001, LDEQ requested technical assistance from EPA and NRMRL regarding the
degradation of contamination and the identification of the source of contamination. Additional
specific ground-water modeling objectives were provided by LDEQ during February 2003, and
discussions followed between EPA, NRMRL, and LDEQ related to overall project development
and modeling project goals.
In May 2004, the Agency for Toxic Substances and Disease Registry (ATSDR) released a
Health Consultation report for the Myrtle Grove Trailer Park water well system, concluding that
residents were exposed to contamination; however, the levels of chemicals detected were below
levels likely to result in adverse health effects. In addition, ATSDR recommended continued
monitoring of the vinyl chloride to better characterize the plume and determine its source(s) and
to assist in the evaluation of potential exposures to vinyl chloride-contaminated ground-water in
areas surrounding the Myrtle Grove Trailer Park site.
The emphasis of this project has been on determining local ground-water flow directions
in the Plaquemine aquifer upper sand unit and observing how flow is affected by external
influences including interactions with the Mississippi River and water extraction from local
pumping wells. The project has also resulted in observations about the likely significant
contaminant source location for ground-water contamination in the Plaquemine aquifer upper
sand unit.
B. Objectives
The following modeling objectives were established during project planning discussions
between EPA, NRMRL, and LDEQ, based on LDEQ's initial request and subsequent input. The
following objectives are also listed in the "Combined Quality Assurance Project Plan and
General Workplan " (See Appendix B):
1. Determine potential directions of ground-water flow in the Plaquemine aquifer over the
project area;
2 . Understand how ground-water flow is affected by aquifer interaction with the Mississippi
River, pumping wells, and by possible regional ground-water flow gradients;
3. To the degree possible with flow modeling, evaluate the likelihood of possible
contaminant source locations, and whether multiple source locations are possible; and
4. Estimate age and location of contaminant release(s) to degree possible with flow
modeling, and plume movement with flow and/or contaminant fate and transport
modeling, if possible.
-------
Plaquemine Aquifer Report
1W7/04
Page: 12 of 43
II. MODEL CONSTRUCTION SUMMARY
Computer models representing both steady-state flow conditions and transient flow
conditions have been developed for this project. Steady-state models represent water level
measurements and hydraulic boundary conditions for a single time period (i.e., single day), and
transient model conditions represent variable water level measurements and variable hydraulic
boundary conditions covering multiple time periods. Steady-state models were used mainly as a
test and evaluation tool to evaluate how the model reacted to initial model setup including model
layering, boundary condition options, and ranges in aquifer property values (including hydraulic
conductivity and aquifer storage); and the transient model was used to evaluate ground-water
flow conditions under more complicated flow scenarios consistent with fluctuations in
Mississippi River stage. The transient model encompasses a time period of approximately 600
days, extending from October 2001 until June 2003. Data collected as late as March 2004 was
used for model validation. The 600-day transient model time period encompasses several river
stage cycles, which are consistent with the stage cycles as described in the previous report by
Whiteman, 1972, entitled "Ground Water in the Plaquemine-White Castle Area. Iberville Parish.
Louisiana".
During this 600-day period, data collection included Plaquemine aquifer upper sand unit
water level measurements taken from observation wells (Figure-3) on seven day intervals, and
Mississippi River stage measurements, which were collected from two nearby gauging stations
Dockn
NX*
,off ,|«f«
"^1. ^^V
"-"•" .„.--
"V
tat
A " ' " '"-
Wells Screened in Plaquemine
Aquifer Upper Sand Unit
• Well Locations
Water Feature
Dow Chemical
Plaquemine, LA
N
1
SOUTCM:
EPA R.gwr 6 PO.
IMOTIOER/UMFMM
TJ&
Figure 3: Locations of most of the wells and piezometers used in the
report.
-------
Plaquemine Aquifer Report
10/7/04
Page: 13 of 43
every seven days (Figure-4). In the transient model, hydraulic boundary conditions and
observation well water levels change every seven days.
Figure 4: Mississippi River Docks I and II showing plotted river
stage fluctuations.
A. Model Grid and Layering
Model layering is representative of the major stratigraphy including the Plaquemine
aquifer upper sand unit and lower sand, an intervening clay/slit layer between the upper sand unit
and lower sand, and overlying natural levee deposits (i.e., top stratum). The model's total
vertical depth is 560-feet, where 0-feet represents model bottom and 560-feet represents land
surface. This layering is derived from EPA lithologic logs and from published cross-sections
such as those provided by Whiteman (1972) and Saucier (1969). The model top stratum extends
from the surface (560-feet) downward to the top of the Plaquemine aquifer upper sand unit at
460-feet. The upper sand unit extends from 460-365 feet, the intervening silt/clay layer from
365-315 feet, and the lower Plaquemine aquifer from 315-70 feet. From 70 feet to 0 feet are
cells that represent deeper sands, silts, and clays containing saltwater.
-------
Placiuemine Aquifer Report
10/7/04
Page: 14 of 43
B. Boundary Conditions
Model boundary conditions were specified to represent the main boundaries of the
physical hydrogeologic system. The main model boundaries include the Mississippi River,
which was treated as a constant head boundary to the east and north; and the western and
southern edges of the model domain within the Plaquemine aquifer upper sand unit, which were
treated as general head boundaries. Inactive model cells/no flow boundaries were included for
the section of the model domain extending to the east of the Mississippi River, because no
detailed water level data is available for that area. However, it is recognized that some recharge
to the Plaquemine aquifer may exist from the upland terraces still further eastward near the city
of Baton Rouge, which is outside the model domain. A map view of the model grid, constant
head boundary, general head boundaries, and inactive model cells are illustrated in Figure 5.
Figure 5: Locations of general head boundaries along the
west and south domain (green), constant head
boundary along the river (red), and inactive
model cells east of river (purple).
The Mississippi River was treated as a constant head boundary because the river is in
direct contact with the Plaquemine aquifer upper sand unit, similar to the example cited in the
MODFLOW users manual (page 2-27) where an aquifer is in direct contact with a major surface
water feature (McDonald and Harbaugh, 1988). Constant head boundary values are based on
-------
Plaauemine Auuifer Report
10/7/04
Page: 15 of 43
DOW Chemical Dock I and Dock n river stage measurements. As the river meanders southward
through the model domain, boundary head values uniformly decrease downstream, with Dock I
and n measurements being used as points of reference. The constant head boundary extends
through the surficial model layer to a depth at which the constant head boundary is in direct
contact with the Plaquemine aquifer upper sand unit. Figure 6, a model cross-section, shows the
spatial relationship shared by the constant head boundary, the Plaquemine aquifer upper sand
unit, and the general head boundary to the west. Due to the river meander curve across the north
of the model domain, the constant head boundary essentially forms most of the northern
boundary for the Plaquemine aquifer upper sand unit.
The western boundary was treated as a general head boundary (GHB) in the Plaquemine
aquifer upper sand unit. A GHB can be applied to many different modeling scenarios, and can
provide a link between local-scale ground-water flow and regional flow. Limited data from
Clean Harbors, Plaquemine (located to the southwest of the model domain), indicates that a
Figure 6: Model west-east cross-section showing boundary
conditions.
relatively small amount of fluctuation (about 3-feet) could occur approximately 4 miles from the
Mississippi River. The GHB was placed approximately 4 miles from the river so that minor head
fluctuation in the GHB would not significantly affect the model. The model time schedule for
GHB fluctuation is different than the model time schedule for river fluctuation, based on data
from Clean Harbors and on the work by Whiteman, 1972. Whiteman, 1972, explains the
relationship in head fluctuation time in the Plaquemine aquifer by comparing the head
-------
Plaauemine Aquifer Report
10/7/04
Page: 16 of 43
fluctuation in the river with head fluctuation for locations to the west. Water wells located near
the river reflect even small changes in river stage within a few minutes or hours; and that farther
away, in backswamp areas (approximately 10-miles from the river), the time lag between
changes in river stage and corresponding changes in water levels increases to a week or more,
and only reflect major changes in river stage.
The southern domain boundary has similar characteristics to the western boundary, and
was set as a general head boundary for the Plaquemine aquifer upper sand unit. The GHB
boundary values were set by using a combination of data from Clean Harbors, Plaquemine, with
observations and adjustments made for calibration.
The lower Plaquemine aquifer is not the focus of this model, however, an attempt was
made to include the lower Plaquemine aquifer to the degree possible. The reason for including
the lower Plaquemine aquifer is because certain individual wells are screened in the lower
Plaquemine aquifer, and because some wells are screened in both the lower Plaquemine aquifer
and in the upper sand unit. Due to a lack of data for the lower Plaquemine aquifer, GHB values
for the lower Plaquemine aquifer on the west and south were set to the same values as the
Plaquemine aquifer upper sand unit boundaries on the west and south. A constant head boundary
was used on the east for the lower Plaquemine aquifer. Table 1 gives the model layer structure
along with ranges of values used and evaluated during the model calibration process for the
general head boundaries.
C. Calibration
Calibration and calibration targets have been defined by the American Society for Testing
and Materials (ASTM) Standard Guide: D 5611-94. Calibration is defined as the process of
refining the model representation of the hydrogeological framework, hydraulic properties, and
boundary conditions to achieve a desired degree of correspondence between the model
simulations and observations of the ground-water flow system. Calibration targets are defined as
measured, observed, calculated, or estimated hydraulic heads or ground-water flow rates that a
model must reproduce, at least approximately, to be considered calibrated. For this model, the
main calibration targets are the water levels (hydraulic heads) measured in water wells during the
period extending from October 2001 until June 2003.
The main goal of the calibration of the ground-water flow model was to ensure that the
ground-water flow directions in the Plaquemine aquifer upper sand unit over the 600-day period
were as accurate as possible. This goal was met by having modeling runs that exhibited good
matching between simulated and observed water levels. After numerous model iterations and
design refinements, and before a detailed systematic formal calibration procedure had been
developed, model runs and output calibration graphs indicated that good matching between
-------
Plaauemine Aquifer Report
10/7/04
Page: 17 of 43
simulated and observed water levels had been obtained. An example of a time-series plot
indicating good water level matching is provided in Figure 7. The graph in Figure 7 is an
example of a time-series graph for three wells, lOlc, PZ-32, and PZ-40, overlain in one graph
(well locations are underlined in red on Figure 3). These three overlain graphs are representative
of calibration graphs for the remaining water wells in the data set. Since good matching was
already obtained prior to the development of a detailed systematic formal calibration procedure, a
Table 1: Model layer structure and general head boundary values.
Model
Layer
1
2
3
4
5
6
7
8
9
10
11
12
Model Layer
Feature
Top Stratum
Upper
Plaquemine
aquifer
Upper
Plaquemine
aquifer
Upper
Plaquemine
aquifer
Upper
Plaquemine
aquifer
Upper
Plaquemine
aquifer
Upper
Plaquemine
aquifer
Aquitard
(clay/silt
layer)
Lower
Plaquemine
aquifer
Lower
Plaquemine
aquifer
Lower
Plaquemine
aquifer
Inactive layer
Conductance
of West
GHB (ft'/day)
N/A
67-33
67-33
78-39
78-39
72-36
83-41
N/A
300
300
300
N/A
General Head
Specified on
West GHB (ft)
N/A
539-538
539-538
539-538
539-538
539-538
539-538
N/A
540
540
540
N/A
Conductance
of South
GHB (ft2/day)
N/A
26-13
26-13
30-15
30-15
28-14
33-16
N/A
300
300
300
N/A
General Head
Specified on
South GHB (ft)
N/A
547-541
547-541
547-541
547-541
547-541
547-541
N/A
540
540
540
N/A
Thickness
(ft)
100.2
4
16
18.7
18.7
17.4
20.0
49.4
72.1
89.5
84.0
70
Elevation
of top of layer
(ft)
560.0
459.8
455.8
439.8
421.1
402.4
385.0
365.0
315.6
243.5
154.0
70
-------
Plaquemine Aquifer Report
10/7/04
Page: 18 of 43
decision was made that the remaining calibration would be more effective if it evaluated the
ranges of input parameters producing good matching between simulated and observed water
levels. Therefore, a less formal calibration approach was used, which also helped to streamline
the modeling project for meeting project objectives and schedules. There is some overlap
between the calibration and the sensitivity analysis for this project. Sensitivity is discussed in
Section IV.
101C(Obs.)*tead
pz40(Obs.)JHead
101C/101OHead
pz40fez40*tead
Figure 7: Multiple well time-series graph. Graph shows close agreement between
transient calculated heads and observed heads.
The initial part of the calibration process, identifying the parameters to which the
rimulated ground-water elevations are sensitive, was followed by making refinements in
parameter values to achieve matching between simulated and observed ground-water levels. To
help accomplish this task, EPA requested the assistance from two organizations, Waterloo
-------
Plaquemine Aquifer Report
10/7/04
Page: 19 of 43
Hydrogeologic, Inc., and NRMRL. Waterloo Hydrogeologic, Inc., a ground-water model
consulting and software development firm, assisted by reviewing the boundary conditions and
ranges of aquifer parameters early in the model development phase to identify steps to improve
calibration, and by running the model in steady-state and transient modes to check model outputs
with revised parameters. For reference, model outputs and comments about calibration
parameters and values from Waterloo Hydrogeologic, Inc. are provided in Appendix C.
NRMRL, with assistance from an on-site technical support services contractor (Shaw
Environmental and Infrastructure, Inc.), conducted an examination of a more recently revised
computer model, and further examined and approximately calibrated several important model
parameters. NRMRL used a structured approach that minimized calibration error. The NRMRL
examination included the boundary conditions of the east domain boundary of the lower
Plaquemine aquifer; the hydraulic conductivity, specific storage, vertical hydraulic conductivity
of the aquitard between the upper and lower Plaquemine aquifer units; and conductance of the
general head boundary. The examination approach involved matching the observed and
calculated ground-water levels for both the peak and minimum parts of the hydrograph, assuming
that the general head remained unchanged, while systematically changing other parameters over
reasonable ranges. This approach indicated that simulation results are sensitive to the examined
model parameters. The examination specified the ranges of parameters evaluated and provided
time-series graphs for each model run to indicate relative parameter sensitivity. See Appendix C
for details.
Following the identification of sensitive parameters and value ranges as described above,
EPA further evaluated boundaries and parameters by performing model run iterations to examine
the match between observed and simulated water levels. These iterations provided a way to
reinforce decision making on the model parameters and ranges of parameters, and a way to
determine which combinations of parameters provided a reasonably close match between
simulated and observed water levels. Table 2 provides the ranges of aquifer property values
applied to the model during calibration.
Table 2: Representative values of aquifer hydraulic properties applied to the model
Model
Layers
1
2-7
8
9-11
12
Hydraulic Conductivity Ranges (ft/d)
Kx
0.05
80-500
0.05
80-500
N/A
Ky
0.05
80-500
0.05
80-500
N/A
Kz
0.005
8.0-50.0
0.005
8.0-50.0
N/A
Specific
Storage
(I/ft)
0.0001
0.0001
0.0001
0.0001
N/A
Specific
Yield
0.2-0.275
0.2-0.275
0.2-0.275
0.2-0.275
N/A
Effective
Porosity
0.2-0.275
0.2-0.275
0.2-0.275
0.2-0.275
N/A
Total Porosity
0.25-0.275
0.25-0.275
0.25-0.275
0.25-0.275
N/A
-------
Plaguemine Aquifer Report
10/7/04
Page: 20 of 43
III. MODELING RESULTS AND SOURCE AREA EVALUATION
This section presents the results of the modeling of the Plaquemine aquifer upper sand
unit as it relates to each of the objectives listed in Section I. B. of this report. All figures
depicting ground-water flow are from model layers 4, 5, or 6, representing the approximate mid-
section of the Plaquemine aquifer upper sand unit, and the flow within these three layers was
virtually the same.
A. Potential Directions of Ground-water Flow
Objective 1: To determine potential directions of ground-water flow in the Plaquemine
aquifer upper sand unit over the project area.
In order to derive net ground-water flow, the Modpath program, which computes ground-
water flow pathlines and the positions of simulated particles at specified points in time (i.e.,
particle tracking), was utilized. Modpath, which was developed by the U.S. Geological Survey,
uses output obtained from MODFLOW (a three-dimensional finite-difference ground-water flow
modeling program) for each model stress period. A stress period represents a block of days when
specified aquifer stress parameters remain constant, such as the general head and constant head
boundaries during each 7-day period. The modeling used 168 stress periods over the 600-day
time frame. For a transient simulation, Visual Modflow, the pre and post data processor, uses the
specified boundary conditions to determine the length of each stress period; the modeler cannot
directly specify stress period length. Each stress period is capable of producing different ground-
water flow directions because the combinations of conditions driving the ground-water flow vary
from one stress period to the next. Steady-state model runs compute output from only a single
stress period.
By reviewing results from stress periods that represent high and low Mississippi River
stage levels, a range in flow directions for the Plaquemine aquifer upper sand unit was
determined. Figure 8 represents ground-water flow in May, 2002, when the river stage is
relatively high (31-34 feet above sea level) under non-pumping conditions. Figure 9 represents
ground-water flow in September, 2002, when river stage is relatively low (5-6 feet above sea
level) under non-pumping conditions. The westerly flow direction shown in Figure 8 is
representative of flow in the majority of stress periods, although some low stage flow directions
(Figure 9) appear to cause a southerly or even easterly flow direction. The frequency of
simulated easterly and southerly flow directions is relatively low, encompassing only a few
model stress periods.
The potential direction of ground-water flow in the Plaquemine aquifer upper sand unit is
illustrated in Figure 10. Results show that the net flow direction, without pumpage, is primarily
to the west. Assuming that any future pumpage does not substantially influence the Plaquemine
aquifer upper sand unit, then the net westerly flow can be considered the potential (future) flow
-------
Plaauemine Aquifer Report
10/7/04
Page: 21 of 43
Figure 8: Approximate ground-water flow at
high river stage (no pumping).
Figure 9: Approximate ground-water flow at low
river stage (no pumping); easterly flow
also possible at low stage.
-------
Plaquemine Aquifer Report
10/7/04
Page: 22 of 43
direction in the Plaquemine aquifer upper sand unit. Pumping simulations are in the next
section.
The approximate westerly direction of ground-water flow in the Plaquemine aquifer
upper sand unit calculated by the model is consistent with previously published information.
Both the U.S. Geological Survey Regional Aquifer System Analysis program study (Martin, et.
al, 1989), and a previously published study conducted by the Louisiana Geological Survey, which
discusses ground-water in the Plaquemine-White Castle Area (Whiteman, 1972), indicate an
approximately westerly direction of ground-water flow. These two published reports utilize map
scales that cover a much wider area than the area included in the model domain for this report,
but they are still useful for making comparisons and specifically for comparing local and regional
flow directions.
Figure 10: Net direction of ground-water flow over 600-day transient model run
period. Yellow arrows have been added for visual clarity between
particle pathlines (red).
-------
Plaauemine Aquifer Report
10/7/04
Page: 23 of 43
B. Aquifer/River Interactions, Pumping Wells, and Regional Flow Gradients
Objective 2: To understand how ground-water flow is affected by aquifer interaction with
the Mississippi River, pumping wells, and by possible regional ground-water
flow gradients.
Pumping simulations were conducted to evaluate how the ground-water flow field could
react to various pumping conditions and river stages, and to compare the flow field under
pumping stress with non-pumping conditions. The wells chosen for these pumping simulations
are listed in the State registered water well database maintained by the Louisiana Department of
Transportation and Development (DOTD). Since the past pumping conditions are mainly
unknown, pumping simulations were conducted using a range of pumping rates and well
configurations, to reduce the uncertainty in the results. These simulations ranged from using the
maximum number and types of wells available in the DOTD database and relatively high
pumping rates, to using a subset of these wells and relatively low pumping rates.
In order to show how the flow field is affected by pumping at low pumping rates, and
how the flow field under low pumping rates is affected by river stage, a subset of wells was
chosen from the DOTD data base in the model domain. Figure 11 shows the types of wells in the
model domain and the subset of wells that were used in the low pumping, flow field and river
stage analysis. Some of the wells in this subset are screened in the Plaquemine aquifer upper
sand unit, and some of the wells are screened in the lower Plaquemine aquifer. Figures 12 and
13 show the flow field produced by pumping this subset of wells at a constant low pumping rate,
at high and low river stages respectively. The model stress periods used for Figures 12 and 13
are the same stress periods previously used for Figures 8 and 9 (May, 2002, and September,
2002).
Results shown in Figures 12 and 13 show a westerly flow in the Plaquemine aquifer
upper sand unit within the plume area. These simulations also indicate that clusters of wells
screened in the Plaquemine aquifer upper sand unit influence water flow directions in the upper
sand unit the most. Flow directions in the Plaquemine aquifer upper sand unit do not appear to
be affected by wells screened in the lower Plaquemine aquifer. The wells at DOW Chemical are
screened in the lower Plaquemine aquifer. Flow directions in the Plaquemine aquifer upper sand
unit are most affected by pumping wells when the river stage is low.
In order to show how the flow field could react to increased pumpage (i.e, a larger
number of wells with higher pumping rates), a simulation was conducted using all the wells
available in the DOTD database for the model domain, with higher pumping rates than those
used in the Figure 12 and 13 modeling simulations. Instead of showing flow directions at just
high and low river stages, this simulation shows flow over all model stress periods by using
-------
Piaguemine Aquifer Report
10/7/04
Page: 24 of 43
particle tracking pathlines (Figure 14). Again, the flow appears to be in a predominantly westerly
direction.
In addition to the effect of the river stage on ground-water flow direction, the effect of the
river stage on ground-water velocity was also evaluated in order to meet Objective 2. Figure 15
shows a particle tracking pathline with velocity time markers as computed by the Modpath
computer program. This figure illustrates both flow direction change and velocity change in the
Plaquemine aquifer upper sand unit. This example pathline, which has been enlarged, represents
a total horizontal distance of approximately 150 feet, with time markers shown every 7-days. As
C kern leal Predict
Fin Protcttoi
Clem teal Prodic*
Commercial
Cltinleal Prodic*
Cfemiul
I**
200
XIO
200
ChtmfcalPiodick
Sippiv
Petrol* i m Piodic*
Fire Protctloi
. Filer Edge
M,'-5is
nr-dbn
-m*
CIVot Plaqumiit W«lk
uncpai
-Irrlgattoi
Irrlgafloi
Figure 11: Pumping wells in the model domain by type, including
modeling subset
shown on Figure 15, flow can move from the Plaquemine aquifer upper sand unit eastward
toward the river at low stage intervals, but this flow reversal toward the Mississippi River
appears to last for only brief periods, and thus does not have a large effect on the net particle
velocity in the westerly direction, away from the river. This conclusion is consistent with the
1972 Whiteman report (Whiteman, 1972), which states that the volume of water involved in the
cyclic movement into and out of the river is less than 1% of the total volume of water in the
aquifer.
-------
—'-^-A-^-,Af^-"v /
K^^S^y /
V—A"^y*frii(>
DOmCHEJMCAL
(APPROXIMATE
BOUNDARY)
/ \- / /.-, V-—A—V--t'
//Xs- ttepcnr&toif: - -\-"'\^- -e,
J'/j ,yys/^, /~^»SL
/ / / /;x^.'/ V'';''V -^x
/ / >' ' *-'-^
^v
/
'V -•" / A/
OTYOFPLAQUEM1ME
WATER WELLS
Figure 12: Flow at high river stage based on wells
pumping at a low rate.
Plauuemine Aquifer Report
10/7/04
Page: 25 of 43
DOW CHEMICAL
(APPROXIMATE
BOUNDARY)
i / / / ,• <
/ //.'//
/ » / / / /
CITY OFPLAQUEMINE
WATER WELLS
Figure 13: Flow at low river stage based on wells
pumping at a low rate.
-------
Plaouemine Aquifer Report
10/7/04
Page: 26 of 43
Figure 14: Net flow (all river stages) based on
all wells pumping at high pumping
rates. River is in blue.
Example Particle Tracking Pathline
Velocity Time Markers
N
Figure 15: Example particle tracking pathline
(enlarged) showing changes in flow
direction and velocity over time.
-------
Platiuemine Aquifer Report
10/7/04
Page: 27 of 43
In summary, the results of the modeling indicate that under the majority of the modeling
conditions, the direction of ground-water flow in the plume area is westerly. The flow in the
plume area is affected by the Mississippi River, which is a large driver of ground-water flow in
the Plaquemine aquifer. During a small number of low river stage stress periods, the Mississippi
river did appear to affect the direction of ground-water flow and cause the direction to be in a
more easterly direction in the plume area. This in turn changed the ground-water velocities and
gradients for these particular stress periods. When net ground-water direction and net ground-
water velocity were calculated, these relatively few periods of reverse ground-water flow at low
river stage did not have a large affect, and thus did not change the evaluation that the ground-
water is flowing westward. Effects of pumping alone on the wells located in the Plaquemine
aquifer upper sand unit, though more evident during low river stages, appear to be
inconsequential to the overall gross direction of ground-water flow in the plume area.
C. Flow Directions and Possible Source Locations
Objective 3: To the degree possible with flow modeling, evaluate the likelihood of possible
contaminant source locations, and whether multiple source locations are
possible.
In order to meet Objective 3, and draw conclusions about the likely significant source of
the contamination in the Plaquemine aquifer, an evaluation of the modeling results and
conclusions was combined with an evaluation of potential source areas in the modeling domain.
A scenario of the mechanism of transport of the contaminants from the likely significant source
area was also constructed to ensure that the results of these evaluations fit the physical facts of
transport in this river-influenced transport regime.
1. Observations and Results
The following observations and results were evaluated in the identification of the likely
significant source area.
- The net flow direction without pumpage in the Plaquemine aquifer upper sand
unit is to the west over the 600-day model period. Flow direction reversal is short
in duration.
- Direct hydraulic communication between the Mississippi River and the
Plaquemine aquifer upper sand unit is the most significant controlling hydrologic
influence over flow in the Plaquemine aquifer upper sand unit. The river stage
cycle frequency, duration, and magnitude within the model time period is
representative of river stage cycles in the past for the Plaquemine, Louisiana area.
-------
Plaguemine Aquifer Report
10/7/04
Page: 28 of 43
— Total water demand in the past for all municipal, industrial, agricultural,
irrigation, and residential uses of ground-water was variable over time depending
on water supply needs (i.e, daily, monthly, and seasonal variations). The
influence of related pumping wells screened in the Plaquemine aquifer upper sand
unit, therefore, was not constant in the past.
— The hydraulic influence of the range of Mississippi River stages cycles on the
Plaquemine aquifer upper sand unit was constant, and is constant today.
- Pumping simulations during high and low stage, under a range of pumping rates
and conditions, shows a consistent net flow to the west in the contaminated part of
the Plaquemine aquifer upper sand unit.
- A hydraulic divide caused by the Mississippi River, discussed by Whiteman,
1972, would require that any source location be west of the divide. In addition,
there are no known possible contaminant source locations east of the Mississippi
River.
- Contaminated source material (liquid or solid) near and within the river channel,
especially at the deepest part of the channel, would have a direct pathway into the
Plaquemine aquifer upper sand unit. The deepest part of the channel is the west
side.
— Ground-water contamination spreading to the west from an elongated source near
the edge of the Mississippi River is consistent with approximate north-south
dimensions of contamination in the Plaquemine aquifer upper sand unit.
- A source area (river bank location) near the northeastern edge of the existing
plume, could be scoured, eroded, and transported southward by river water and
could cause the source area to be elongated to provide source dimensions
consistent with the north-south plume dimensions.
2. Previous Source Area Investigation and Evaluation
To supplement the evaluation of the location of a contaminant source based on the project
observations and results, EPA considered available information on potential source areas that
were previously investigated and evaluated by LDEQ. From May 2001, until June 2001, LDEQ
performed surveillance inspection activities relating to vinyl chloride contamination at the Myrtle
Grove trailer park. The LDEQ reviewed information on all locations on the Louisiana Source
Water Protection Area Inventory list within a 1-mile radius of the Myrtle Grove trailer park (a
total of 47 facilities). According to LDEQ, the primary focus of these activities was "to
-------
Plaauemine Aquifer Report
10/7/04
Page: 29 of 43
determine the potential for any existing possible sources of vinyl chloride contamination to soil
and/or ground-water within a 1-mile radius of the Myrtle Grove trailer park." No sampling was
conducted as part of the activities, although a visual inspection of each location was conducted to
determine the presence of hazardous materials and/or wastes, focusing on chlorinated chemical
compounds known to degrade to vinyl chloride, and vinyl chloride itself. The conclusion of the
memorandum on the surveillance inspection activities of these locations is that no areas of
concern were discovered (LDEQ, 2001).
From mid-2001 through 2003, LDEQ also conducted investigations into additional
facilities and sites, in order to consider the possibility of historical contaminant sources and in
order to assess sources in a wider radius than the surveillance investigation activities area. These
additional facilities and sites were selected by LDEQ for further review based on reviews of
computer files and facility files; reviews of historical aerial photographs; information provided
by citizens and requests from citizens; and observations by LDEQ personnel while they were in
the field. Figure 16 is a map produced by LDEQ showing multiple locations of sites, areas, and
facilities which were evaluated; but additional sites were evaluated as part of the surveillance
inspection activities, which are not labeled on this map. LDEQ also evaluated the possibility of
potential sources being located in areas along the Mississippi River, Bayou Jacob, Bayou
Plaquemine, Highway 1, the railroad tracks, and area pipelines.
Additional sites not labeled on Figure 16 but still investigated and evaluated by LDEQ
include dry cleaners and laundromats. LDEQ identified dry cleaners by reviewing historical
telephone book records dating from 1946 to 1991, and found that all dry cleaner sites were
located south of Bayou Plaquemine and therefore, south of the plume area. The Source Water
Protection Area Inventory did identify one laundromat located on Highway 1, north of Bayou
Plaquemine, but found that this location was never used for dry cleaning purposes, so it did not
use chemicals that degrade to vinyl chloride.
The LDEQ evaluated these facilities, sites, and areas using all information available at the
time, including sampling data that it obtained during some of the investigations. The LDEQ
determined that many of the sites, which were investigated and evaluated, did not cause the
contaminant plume, or were not of the magnitude to have contributed to the plume significantly.
However, LDEQ was not able to identify any likely significant source locations because the
ground-water flow direction was not yet known.
-------
Plaquemine Aquifer Report
10/7/04
Page: 30 of 43
Figure 16: Locations of multiple sites, areas,
and facilities evaluated by LDEQ
during 2001-2003; Map does not
include additional locations
evaluated during surveillance
inspection activities.
1. Nadler Industries
2. Old Iberville Motors
3. Louisiana Vacuum Service
4. Former Solid Waste Landfill
5. Bayou Jacob fill area
6. New Iberville Motors
7. Consolidated Companies
8. Alleged Truck Burial Site
9. Motion Industries
10. Former Welding Shop
11. Alleged burial area behind trailer park
12. Former Myrtle Grover Sugar Mill
13. DOW Lighthouse Road Landfill
14. Industrial Haulers
15. DOW Warehouse
16. Former ChemVac site
17. Former DOW injection well
18. DOW PCE spill area
19. DOW Block 49 landfill
20. DOW Australia Point landfill
-------
Plaaueniine Aquifer Report
10/7/04
Page: 31 of 43
3. Current Source Area Identification
Ground-water flow modeling completed for this project confirmed that most potential
source locations evaluated and ruled out by LDEQ in Figure 16, could not have significantly
contributed to the contamination in the Plaquemine aquifer upper sand unit. These locations,
both individually, or as a group which could have formed a larger source area, are either too far
west or too far south for the net westerly ground-water flow to have transported the contaminants
to the current plume location and to have made the current plume size and shape. Additional
areas that were ruled out as significant sources of area ground-water contamination based on both
LDEQ's investigation and net flow direction are areas west of the Mississippi River such as
Highway 1 and the railroad, and areas to the south, including Bayou Jacob and Bayou
Plaquemine.
Based on the observations and results made from the ground-water modeling for this
project, including observations about the geology, hydrogeology, river hydrology, known
chemical releases; and LDEQ's previous investigation of potential source areas and releases, only
one geographic area was identified that is a likely significant source area for the contamination in
the Plaquemine aquifer upper sand unit. This area is located near the northeastern edge of the
existing contamination plume and along the western side of the Mississippi River. (See Figure
17). At the area located near the northeastern edge of the existing contamination plume and
along the western side of the Mississippi River (the "likely significant source area"), a number of
major natural features and man-made circumstances coincided, creating a scenario with both a
series of contaminant sources and a pathway to the Plaquemine aquifer upper sand unit. This
scenario, which did not exist elsewhere in the overall project area, is the only known scenario,
which could have caused the significant and broad contamination of the Plaquemine aquifer
upper sand unit. This conclusion is described in more detail in this section, and in the following
section on contaminant transport mechanisms.
A review of historic chemical use and past releases of contaminants in the likely
significant source area shows several incidents or locations where chemicals that degraded to
vinyl chloride contamination were released. One such location is site #13 on Figure 16, the DOW
Lighthouse Road Landfill. Chemical analyses conducted by DOW Chemical Company as part of
a 1993-1994 site investigation at site #13 identified seven compounds in one soil boring, which,
have the potential to degrade into vinyl chloride. The fact that these contaminants were found in
the soil means that they likely were released to the shallow ground-water also. These seven
individual compounds detected at the likely significant source area are: 1,1,1-Trichloroethane,
1,1,2-Trichloroethane, 1,1,2,2-Tetrachloroethane, 1,2-Dichloroethane (EDC), Hexachloroethane,
Tetrachloroethene, and Trichloroethylene (TCE).
Past spills of perchloroethylene (also called tetrachloroethylene, which is a compound
that can degrade to vinyl chloride) at the DOW Chemical Company facility in the immediate
vicinity of site #13, occurred during 1964, and 1993, according to LDEQ. These spills may also
-------
Plaquemine Aquifer Report
10/7/04
Page: 32 of 43
have contributed to the contamination in the Plaquemine aquifer upper sand unit. Informal
discussions between EPA and LDEQ staff indicate that a quantity of approximately 10,000
pounds of perchloroethylene was spilled at Dow Chemical in 1964, but the exact location of the
spill, and whether or not it was released to the Mississippi River is not fully known. It is known,
however, that of the approximately 761,000 pounds of perchloroethylene spilled in 1993, at Dow
Chemical Company (approximately 500,000 pounds at the tank farm, and approximately 250,000
pounds in the canal), approximately 1,036 pounds reached the Mississippi River near the likely
significant source area.
Several geologic and hydrogeologic conditions exist in the likely significant source area
in the vicinity of site #13 that would enable contamination released in this area to easily enter the
Plaqeumine upper sand aquifer unit. One of the most important hydrogeologic conditions is that
there is direct geological and hydrogeological connection of the likely significant source area to
the Plaquemine aquifer upper sand unit. Geologic cross-sections referenced in the Technical
Background Document (Appendix A) indicate that site #13 is at or very close to a point, which is
a direct pathway into the Plaquemine aquifer upper sand unit, either through the interface
between the Mississippi River and the aquifer, or through preferential man-made pathways such
as along industrial pipelines penetrating the top stratum and the upper sand unit of the aquifer.
Site #13 and the area adjacent to it are also located on the cut bank of the Mississippi River,
Figure 17: Likely significant contaminant source area
location.
-------
Plaauemine Aquifer Report
10/7/04
Page: 33 of 43
where fluvial processes, including stream bank erosion and material transport, take place. These
processes are described in the next section.
4. Likely Significant Source Area Mechanism of Transport
The mechanisms of contaminant transport into the Plaquemine aquifer upper sand unit
from the likely significant source area include transport through fluvial processes and transport
along preferential pathways. Fluvial processes most likely elongated the original source area in a
north-south direction, causing a zone of possible source spreading (See Figure 17). These
processes facilitated the movement of any contaminants released at or near site #13 into locations
where they could enter the Plaquemine aquifer either along the surface and/or within the river
bottom. It is probable that source material in both liquid and solid phases was transported
southward along the river edge by river flow, and transported into the river along the river
channel bottom by turbulent river flow associated with lateral stream cutting. Lateral stream
cutting tends to be prominent in streams flowing in winding courses on low slopes (Spencer,
1983), such as the Mississippi River. As described by Spencer, at each turn in a stream channel,
water is shifted towards the outside of the turn, where turbulence and erosion are concentrated.
The result of this flow is to deepen the stream channel and steepen the bank on the outside of the
curve. This causes the outside bank to become unstable and slump into the stream. Any
turbulent river flow at the likely significant source area probably moved some contaminated
material into deeper parts of the Mississippi River, which is in direct contact with the
Plaquemine aquifer upper sand unit.
Possible preferential pathways into the Plaquemine aquifer upper sand unit may be along
pipelines located near the likely significant source area. At least two subsurface pipelines (i.e.,
the Texaco and Air Products pipelines), penetrate the surface within the likely significant source
area, extending downward from the surface through overlying silts and clays, to reach a
subsurface depth at which they pass beneath the Mississippi River. These pipelines extend
through the Plaquemine aquifer upper sand unit beneath the likely significant source area.
Contaminants reaching these preferential pathways would have an accelerated descent into the
Plaquemine aquifer upper sand unit. A diagram showing the locations of these pipelines is in
Appendix A.
Once contaminants moved into the Plaquemine aquifer upper sand unit they were
transported and dispersed away from Mississippi River in a westerly direction. Contaminant
transport processes including advection and dispersion, which likely contributed to plume growth
and spreading, are not described in this report because the inclusion of these mechanisms is
beyond the scope of this project.
-------
Plaauemine Aquifer Report
10/7/04
Page: 34 of 43
D. Age of Contaminant Release and Plume Movement
Objective 4: To estimate age and location of contaminant release(s) to degree possible
with flow modeling, and plume movement with flow and/or contaminant fate
and transport modeling, if possible.
The age of the contaminant release was estimated by using an average ground-water
velocity of Ift/day under a non-pumping scenario as the rate of contaminant movement, and by
back calculating, based on the net flow direction, the amount of time it would have taken the
contamination to travel from the edge of the Mississippi River. A 20-40 year time frame was
estimated for contamination to reach the Myrtle Grove wells and the plume area from the river
edge. The Mississippi River was used as a reference point since it is the location of the area of
possible source spreading and the interface between the river and the Plaquemine aquifer upper
sand unit.
The time frame is a rough estimate in that it only considered the ground-water velocity
and the direction of ground-water flow. The actual rate of contaminant movement is likely
different than the ground-water velocity due to complicated chemical processes and reactions that
may have occurred within the plume, including reductive dechlorination. The inclusion of these
chemical processes was beyond the scope of the modeling performed for this analysis. However,
this time estimate is considered to be adequate for the purpose of gaining an idea of the order of
magnitude of the time that it would take for a contaminant released into the Plaquemine aquifer
upper sand unit to reach the Myrtle Grove wells. This time frame also fits with the time of the
release of contaminants from the 1964 spill, while it does not rule out that contaminants from the
1993 release could also be contributing to the current ground-water plume.
Considering that there has been continuous water movement within the Plaquemine
aquifer upper sand unit, it is possible that the original source area has been flushed to such a
degree that it may no longer be recognizable as the original source area by chemical sampling
and analysis of soil and ground-water alone, and the center of contaminant (plume) mass may
have shifted away from the source area to the west. In fact, the highest concentration of vinyl
chloride in ground-water (97 ppb) was detected in well VV to the west of the likely significant
source area. Similarly, the possible zone of source area spreading may have been eroded to a
degree that contaminated material is no longer identifiable on the surface. Figure 18 shows the
location of the likely significant source area and the zone of possible source spreading relative to
the ground-water contamination.
Developing a calibrated contaminant fate and transport model was outside the capability
of existing data. However, the ground-water flow modeling described in this report may serve as
a basis for a future calibrated fate and transport model, if more complete fate and transport data
are obtained.
-------
Plauuemine Aquifer Report
10/7/04
Page: 35 of 43
Figure 18: Detections of vinyl chloride and cis-1,2 dichlorethene. Brown
shaded area indicates approximate plume outline (from LDEQ
2002).
Likely significant source area in black, and area of possible
spreading indicated by dotted pattern.
-------
Plaauemine Aquifer Report
10/7/04
Page: 36 of 43
IV. MODELING SENSITIVITIES AND UNCERTAINTIES
A. Ground-Water Flow Modeling Sensitivities
The purpose of this section is to present the principal sensitivities in the ground-water
flow model, and more specifically, the sensitivities affecting ground-water flow directions under
non-pumping conditions. The level of sensitivity analysis conducted was consistent with EPA's
suggested graded approach for defining the level of Quality Assurance effort needed, based on
the intended use and project scope and magnitude (EPA, 2002; see Table 2). The sensitivity
analysis included the ranges of information contained in Tables 1 and 2 of this report.
A brief summary of methods to conduct a sensitivity analysis for a ground-water flow
model are contained in the American Society for Testing and Materials (ASTM) Standard Guide:
D 5611-94. While the techniques presented by ASTM are more formal than necessary given the
scope of this project, they do provide an established rationale for classifying types of sensitivity
and their significance. ASTM recommends classifying the sensitive inputs, which may cause
either a Type I, n, HI, or IV sensitivity, depending on whether the changes to the calibration
residuals and modeling conclusions are significant or insignificant (of no concern) when these
inputs are changed. Type I, n, and HI sensitivities were identified for this modeling. These
sensitivities are discussed below.
According to the ASTM classification system, Type I and Type n sensitivities occur
when, regardless of the value of a certain input parameter, the model output and thus the
conclusion from the modeling, remains virtually the same. For example, a parameter that causes
a Type I or n sensitivity could be changed, and the modeling results would still yield the same
direction of ground-water flow. According to the ASTM classification system, Type HI
sensitivities occur when variations in a certain input parameter cause the model to become
uncalibrated. Values of the input parameters in these ranges are not usable. Therefore, as long as
they are identified and the values that cause the model to lose calibration are not used, Type HI
sensitivities are of no concern. Type IV sensitivities occur when additional data collection to
decrease the range of possible values of an input parameter is needed. No Type IV sensitivities
were identified for this ground-water flow model.
1. General Head Boundary Sensitivity
A sensitivity was identified for general head boundaries. This sensitivity was evaluated
mainly during model construction and after calibration by examining the main computed outputs:
time-series transient calibration graphs, steady-state calibration graphs, equipotential surface, and
flow direction vectors. Significant variations in general head boundaries decreased matching
between simulated and observed water levels, and so those variations were deemed unrealistic
for use in the model. Thus the general head boundary sensitivity was determined to be Type ffl.
-------
Plaauemine Aquifer Report
10/7/04
Page: 37 of 43
2. Aquifer Property Sensitivity
The model showed sensitivity to changes in values for aquifer storage, where some
storage values caused model predictions to lag observation data, indicating that those storage
values were too large. Changes in model output from variation in aquifer storage values were
examined on transient calibration time-series graphs, equipotential surface maps, and flow
direction vector maps. Even though model predictions lagged observation data as observed on
time-series graphs, the cyclic nature of water level fluctuation magnitude and frequency appeared
to remain generally consistent, and water flow directions remained generally the same overall.
This sensitivity is considered Type n sensitivity. Appendix C contains additional details about
sensitivity to hydraulic conductivity and aquifer storage.
Values for hydraulic conductivity were largely based on literature values, including
Whiteman, 1972, other publications on regional hydrogeology as discussed in Appendix A, and
on the professional judgement of members of the modeling team including peer reviewers.
Because of the lack of site-specific hydraulic conductivity data, modeling output was evaluated
to see whether it was sensitive to changes in the hydraulic conductivity. Over the duration of the
project, many model iterations were conducted using a wide range of hydraulic conductivity
values. Model flow velocities seemed sensitive to changes in hydraulic conductivity, but the
water level calibration and the directions of flow were consistent, even with large hydraulic
conductivity changes. Because the variation of hydraulic conductivity caused insignificant
changes to flow directions, this is considered a Type I sensitivity.
3. Sensitivity to Site-Specific Pumping Data
There is a lack of site-specific data on pumping parameters for wells in the Plaquemine
upper aquifer sand unit. The main issue is whether the gross model flow field would change
dramatically in the area of contamination with variable pumping rates and schedules, and with
variable numbers of pumping wells. In order to determine whether the model is sensitive, the
model was run with variable pumping rates and schedules and with variable numbers of wells in
various locations. The results showed that the gross ground-water flow directions in the
contaminated area were not very sensitive to these changes in simulated pumping. However,
changes in flow did occur around individual pumping wells and clusters of pumping wells.
Therefore, the pumping parameters exhibited a Type I or n sensitivity. The uncertainty
associated with the lack of information on pumping parameters is discussed later in this section.
-------
Plaauemine Aquifer Report
10/7/04
Page: 38 of 43
B. Ground-Water Flow Modeling Uncertainties
The uncertainty discussion is included in order to list the most outstanding flow modeling
uncertainties that were encountered and how they were addressed. Uncertainties in the modeling
were mainly caused by data gaps, which necessitated making assumptions during the modeling.
Uncertainties were addressed by using all available site-specific data and published regional
information and by filling data gaps using the professional judgement of members of the
modeling team and peer reviewers. The principal uncertainties in this ground-water flow model:
i) the specification of general head boundaries; ii) the determination of aquifer property values
for use in the model; and, iii) the determination of water well pumping rates and schedules, are
discussed in this section.
1. General Head Boundary Uncertainty
Uncertainty exists in the specification of the general head boundary conditions, because
only limited measured data were available to utilize in this process. Inputs for general head
boundaries include the reference heads, the boundary conductance values, and the time schedule
(i.e., a head fluctuation schedule). The reference heads are the water level values along the
length of the boundary. Conductance is a specified numerical parameter which represents the
resistance to flow between a general head boundary and ground-water. The time schedule shows
the times of changes in the reference heads, which are used to match data to the model stress
periods. Except for the limited data from Clean Harbors, Plaquemine, as mentioned in Section n
of this report, there were no measured data available along the length of the general head
boundaries to allow these boundary condition inputs to be set to site-specific values. Due to this
uncertainty in setting the general head boundaries, it is possible that anomalous boundary flow
effects may occur along these boundaries during some model stress periods. The fact that it is
uncertain how closely the model boundary reference heads, conductance, and fluctuations in
water levels match actual field conditions, and that there could be anomalous boundary flow,
however, did not adversely affect the modeling results because output calibration graphs still
showed good matching between simulated and observed water levels.
2. Uncertainty in Setting Aquifer Property Values
Another area of uncertainty in the modeling was in the setting of aquifer property values.
Site-specific aquifer test data were not available because no aquifer tests had been conducted on
the monitoring wells that were installed in 2002 and 2003 by LDEQ and EPA just prior to the
commencement of the modeling. Due to the lack of site-specific data, ranges of values for
hydraulic conductivity and storage were established based on published literature and on the
judgement of the modeling team, as previously mentioned. Because the ranges of aquifer
property values were not measured for this model, the sensitivity of the model to changes in
aquifer properties was evaluated.
-------
Plaauemine Aquifer Report
10/7/04
Page: 39 of 43
3. Uncertainty in Water Well Pumping Parameters
A third category of uncertainty in the modeling is in regard to aquifer pumping
parameters. Uncertainty exists because of the lack of information on the historical pumping
conditions (well locations, pumping rates, and pumping duration) in the Plaquemine aquifer
upper sand unit. In addition, the ground-water flow model calibration data was collected during
a 600-day period that was typically under local non-pumping conditions, so it is not known how
well-calibrated the model would be under pumping conditions. To address these pumping
parameter uncertainties, flow direction output was determined by performing model simulations
in which the pumping rates and well locations varied, as described in Section ffl. B of this report.
Ground-water flow directions near the Mississippi River in the vicinity of the ground-water
contamination remained consistent, even with the use of ranges of simulated pumping conditions,
so the assumed values for the pumping parameters are considered to have produced reasonable
modeling results.
4. Source Area Identification Uncertainty
There are two areas of uncertainty related to the mechanism of contaminant transport
from the likely significant source area, which are listed here for the purpose of facilitating any
related future work. One uncertainty is regarding the rate and degree of erosion that has probably
taken place along the cut bank of the Mississippi River. The rate and degree of stream sediment
and stream bank erosion may need to be distinguished from the rate and degree of erosion of
surface contamination. Similarly, the transport of solute, and possibly dense non-aqueous phase
liquids, within the Mississippi River, is an area suggested for future study. Clarification of these
mechanisms would help define the degree of source area spreading that has probably taken place,
in terms of time, distance, volume, and area of spreading, and may provide additional
information on the plume shape and distribution of contaminants in the Plaquemine aquifer.
-------
Plaguemine Aquifer Report
10/7/04
Page: 40 of 43
V. CONCLUSIONS AND RECOMMENDATIONS
A. Directions of Ground-Water Flow
Ground-water modeling showed that the net flow direction without pumpage in the
Plaquemine aquifer upper sand unit is primarily to the west. Flow directions to the south and
east may occur at low river stage, but these deviations from westerly flow extend over relatively
short periods of time. Pumping simulations during high and low Mississippi River stage, under a
range of pumping rates and conditions, shows a consistent net flow to the west over the
contaminated part of the Plaquemine aquifer upper sand unit. Direct hydraulic communication
between the Mississippi River and the Plaquemine aquifer upper sand unit is the most significant
controlling hydrologic influence over flow in the Plaquemine aquifer upper sand unit. The river
stage cycle frequency, duration, and magnitude used in the modeling is representative of river
stage cycles in the past for the Plaquemine, Louisiana area.
B. Likely Significant Source Area
Based on flow directions determined during ground-water modeling, and other related
observations about the geology, hydrogeology, river hydrology, area history, documented
chemical releases, and LDEQ's previous investigation of potential source areas and releases; only
one geographic area was identified that is a likely significant source area for the contamination in
the Plaquemine aquifer upper sand unit. This area is located near the northeastern edge of the
existing contamination plume and along the western side of the Mississippi River. The majority
of the ground-water contamination detected in the vicinity of Plaquemine, Louisiana, including
Myrtle Grove trailer park, most likely originated from this source location.
From this likely significant source area, contamination probably spread southward along
the edge of the river and migrated downward into the Plaquemine aquifer upper sand unit.
Contamination reaching the Plaquemine aquifer upper sand unit would have been transported and
dispersed from the river in a westerly direction, moving underneath Myrtle Grove trailer park and
other impacted locations of the Plaquemine, Louisiana area.
-------
Plauuemine Aquifer Report
10/7/04
Page: 41 of 43
C. General Recommendations Including Water Well Pumping
Evaluate pumping rates and schedules for high capacity wells in the Plaquemine area,
such as the city of Plaquemine water supply wells, in order to better assess the effects of
pumping on ground-water flow, and better evaluate the possibility of wells drawing in
contaminated water.
Maintain records of pumping rates and schedules, so that any future modeling
assessments of ground-water flow and plume movement may be performed with more
accurate information.
Continue collecting contaminant concentration data from the existing monitoring well
network to help evaluate changes in contaminant concentrations over time.
Consider obtaining a more detailed delineation of areas of the plume above MCLs, and
evaluate what treatment options may be applied to those areas to reduce contaminant
concentrations to below MCLs.
-------
Plaguemine Aquifer Report
10/7/04
Page: 42 of 43
VI. REFERENCES
American Society for Testing and Materials (ASTM), D5611-94, Annual Book ofASTM
Standards, Philadelphia, Pennsylvania.
LDEQ, 2001, Memorandum from Joseph B. Pecot to Bobby Mayweather, Capital Regional
Office; RE: Myrtle Grove Trailer Park, Plaquemine, Louisiana, Vinyl Chloride
Contamination of Water Well.
Martin, Angel Jr., and Whiteman, C.D., Jr., 1989, Geohydrology and Regional Ground-Water
Flow of the Coastal Lowlands Aquifer System in Parts of Louisiana, Mississippi,
Alabama, and Florida-A Preliminary Analysis, U.S. Geological Survey, Water
Resources Investigations Report 88-4100.
McDonald, M.G., and Harbaugh, A.W., 1988, A Modular Three-Dimensional Finite-Difference
Ground-Water Flow Model, Book 6, Modeling Techniques, U.S. Geological Survey.
National Risk Management Research Laboratory (NRMRL), March 19, 2004 Memorandum to
U.S. EPA Region 6.
Saucier, R.T., 1969, Geological Investigation of the Mississippi River Area, Artonish to
Donaldsonville, Louisiana, U.S. Army Corps of Engineers, Technical Report S-69-4.
Spencer, Edgar, W., 1983, Physical Geology, Addison-Wesley Publishing, 611 p.
U.S. Environmental Protection Agency (EPA), 1994, Ground-Water Modeling Compendium, 2nd
Edition, Model Fact Sheets, Descriptions, Applications, and Cost Guidelines; Office of
Solid Waste and Emergency Response, EPA-500-B-94-004.
U.S. Environmental Protection Agency. 2002. Guidance for Quality Assurance Project Plans for
Modeling QA/G-5M. EPA/240/R-02/007. Office of Environmental Information.
Washington, D.C.
Whiteman, C.D. Jr., 1972, Ground Water in the Plaquemine-White Castle Area, Iberville Parish,
Louisiana, Louisiana Geological Survey, in cooperation with U.S. Geological Survey,
Water Resources Bulletin 16.
-------
Plaquemine Aquifer Report
10/7/04
Page: 43 of 43
APPENDIX
A: Technical Background Document and Conceptual Model
B: Combined Quality Assurance Project Plan and General Workplan
C: Correspondence-Model Reviews and Refinements
-------
APPENDIX: A
Technical Background Document
and Conceptual Model
-------
TECHNICAL BACKGROUND DOCUMENT AND
CONCEPTUAL MODEL
POTENTIAL GROUND-WATER FLOW
DIRECTIONS AND CONTAMINANT FATE AND
TRANSPORT IN THE PLAQUEMINE AQUIFER
OF IBERVILLE PARISH AND WEST BATON
ROUGE PARISH, LOUISIANA
February 17,2004
Prepared Bv
Scott ELlinger \ Nancy Fagan2, James Harris', and Eric Adidas:
(1) Corrective Action/Waste Minimization Section,
(2) State/Tribal Oversight Section, (3) UST/Solid Waste Section
Multimedia Planning and Permitting Division
EPA Region 6
-------
Technical Background Document
Date: Feb. 17,2004
Page: 2 of 45
EXECUTIVE SUMMARY
A. BACKGROUND
The purpose of this document is to present technical background information necessary to
support a numerical (computer) ground-water model of the Plaquemine Aquifer. The numerical
model is intended to assist with understanding and interpreting general ground-water flow
conditions and contaminant movement in the Plaquemine Aquifer over the project area in a basic
capacity. This modeling project is further described in the Combined Quality Assurance Project
Plan and General Work Plan (EPA, August 2003). Detailed technical and scientific reviews
about model conceptualization and setup have been conducted by scientists at EPA Region 6, and
by scientists (including hydrology specialization) at the EPA National Risk Management
Research Laboratory (NRMRL), by scientists and modeling experts at Shaw Environmental and
Infrastructure, Inc. (providing technical support services to NRMRL), and also by Waterloo
Hydrogeologic Inc.(developer of Visual Modflow) through a support services agreement with
EPA Region 6. Additional limited discussions were held with U.S. Geological Survey staff,
providing in-house support to EPA Region 6, on issues related to contaminant fate and transport
in the Plaquemine aquifer.
EPA Region 6 has been assisting the Louisiana Department of Environmental Quality
(LDEQ) with a comprehensive ground-water investigation in Plaquemine, Louisiana. This
investigation is a result of vinyl chloride contamination discovered at the Myrtle Grove Trailer
Park early in 2001. Since becoming aware of the vinyl chloride contamination, the LDEQ has
taken actions to assure that no one is drinking water from the aquifer, and has conducted door-to-
door visits, public meetings, and provided general information on the extent of contamination to
citizens. Extensive sampling of ground-water wells in the area has been conducted since 2001 in
an effort to delineate the contamination, determine the source, and to see if contaminant
concentrations are decreasing with time due to natural attenuation. The Agency for Toxic
Substances and Disease Registry (ATSDR) conducted a Health Consultation and released a
report in June 2002 stating that past exposures to vinyl chloride and all other substances detected
in well water at the Myrtle Grove Trailer Park were too low to produce any adverse health
effects. The City of Plaquemine has installed a system of sentinel wells to serve as advance
warning, if contaminants show up in the area of the City's backup water supply wells.
EPA Region 6 has participated in the ground-water investigation effort by coordinating
ongoing studies of the vinyl chloride plume with NRMRL, and Region 6 staff experts in
hydrogeology. The general area of the plume has been delineated, but our ground-water experts
believe that existing geochemical data is not sufficient for making conclusions about the source
location(s).
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 3 of 45
EPA and LDEQ have been collecting ground-water elevation data from monitoring wells
installed by EPA, and local non-drinking water wells and piezometers installed by Dow
Chemical. Since ground-water flow in the aquifer is affected by the water levels in the
Mississippi River, pumpage, recharge, and geologic and hydrogeologic factors, it is necessary to
collect data that can be integrated in order to obtain a complete picture of the local ground-water
regime. With this information and published information from other agencies such as the U.S.
Geological Survey, our EPA Region 6 staff experts are developing this conceptual model upon
which to base a numerical model to simulate this complex ground-water flow system. Region 6
is documenting this modeling effort by utilizing a Quality Management Plan, a Combined
Quality Assurance Project Plan and General Work Plan, this Technical Background Document
and Conceptual Model, and ultimately a final report will be produced.
B. MODEL CONSTRUCTION SUMMARY
The modeling process will involve developing steady-state models and a transient model
covering the period extending from October 2001 until June 2003. During this 600-day period,
aquifer water level measurements were collected from approximately 50 observation wells and
piezometers, and Mississippi River stage level measurements were collected from two gauaging
stations every seven days. Steady-state models, representing water level measurements and
hydraulic boundary conditions for a single day, will be used mainly to facilitate complex
transient model construction by narrowing down model aquifer property data and hydraulic
boundary conditions and parameters. Once reliable results are achieved under steady-state
conditions, a more complicated transient model will be developed. In the transient model, the
hydraulic boundary conditions and observation well water levels will change according to field
measurements over the 600-day period. Water level calibration will be attempted by making
adjustments to input data within reasonable ranges, and by statistical treatment and graphing of
calculated heads versus observed heads.
During the modeling process, it is expected that different configurations of the hydraulic
boundaries will be tested for sensitivity. Some of the boundary conditions that are likely to be
used include a river boundary, general head boundary, constant head boundary, and no-flow
boundaries. To the degree supporting data are available, the model can include water extraction
from local pumping wells to help understand man-made stresses on the Plaquemine aquifer, and
how historical well extraction may have influenced ground-water flow directions. However, it is
very likely that more wells exist (or existed in the past) than for which data are actually available
today; therefore, certain assumptions may need to be made for historical pumping rates and
schedules.
If possible, chemical fate and transport modeling will be included with steady-state and/or
transient flow models. But due to limited geochemical data, such modeling may only include the
processes of advection and dispersion, and calibration will likely be very limited.
-------
Technical Background Document
Date: Feb. 17,2004
Page: 4 of 45
TABLE OF CONTENTS
EXECUTIVE SUMMARY 2
A. BACKGROUND 2
B. MODEL CONSTRUCTION 3
TABLE OF CONTENTS 4
LIST OF FIGURES 5
I. GEOLOGICAL FRAMEWORK 6
A. REGIONAL SUMMARY 6
1. PHYSIOGRAPHY 6
2. REGIONAL STRUCTURAL FEATURES 6
3. SURFACE GEOLOGY 7
B. LOCALGEOLOGY 12
1. DEPOSITIONAL ENVIRONMENTS 12
2. STRATIGRAPHY 12
3. THE BATON ROUGE FAULT 13
H. HYDROGEOLOGY 21
A. REGIONAL AQUIFER SYSTEMS 21
B. LOCAL HYDROGEOLOGY 34
1. SEASONAL/CYCLIC MOVEMENT; SALT WATER FLUSHING 34
2. GROUND-WATER LEVEL MEASUREMENTS DURING 2001-2003 ... 34
3. THE BATON ROUGE FAULT 35
HI. COMPUTER MODEL SETUP OVERVIEW 39
A. MODEL GRID AND LAYERING 39
B. HYDRAULIC PROPERTIES 39
C. HYDRAULIC BOUNDARY CONDITIONS 40
1. Mississippi River 40
2. General Head Boundaries 41
3. No-Flow Boundaries and Inactive Model Cells 41
D. WATER WELL (PUMPING WELL) DATA 41
IV. REFERENCES 42
APPENDIX 45
-------
Technical Background Document
Date: Feb. 17,2004
Page: 5 of 45
LIST OF FIGURES
FIGURE-1: LOCATION MAP AND MODEL DOMAIN 8
FIGURE-2: REGIONAL SUBSURFACE GEOLOGY 9
FIGURE-3: MAJOR REGIONAL STRUCTURAL FEATURES 10
FIGURE-4: SURFACE GEOLOGY 11
FIGURE-5: HYDROSTRATIGRAPHY (EAST-WEST CROSS SECTION) 15
FIGURE-6: HYDROSTRATIGRAPHY .16
FIGURE-7: CROSS SECTIONS 17
FIGURE-8: TOP OF THE PLEISTOCENE 18
FIGURE-9: CROSS-SECTION C-C' 19
FIGURE-10: LOCATION OF MAJOR FAULTS NEAR BATON ROUGE 20
FIGURE-11: MAJOR REGIONAL AQUIFERS 23
FIGURE-12: COASTAL LOWLANDS AQUIFER SYSTEM 24
FIGURE-13: MISSISSIPPI RIVER VALLEY ALLUVIAL AQUIFER 25
FIGURE-14: DIAGRAM OFZONES A-E 26
FIGURE-15: TOTAL THICKNESS OF SAND IN UPPER PLEISTOCENE AQUIFER .. 28
FIGURE-16: PERCENTAGE OF SAND IN UPPER PLEISTOCENE AQUIFER 29
FIGURE-17: MEASURED 1980 WATER-LEVEL ALTITUDES 30
FIGURE-18: PREDEVELOPMENT FLOW IN UPPER PLEISTOCENE AQUIFER 31
FIGURE-19: GROUND-WATER FLOW IN 1980, UPPER PLEISTOCENE AQUIFER .. 32
FIGURE-20: AERIAL DISTRIBUTION OF PUMPAGE 33
FIGURE-21: DEPTH OF OCCURRENCE OF FRESHWATER 36
FIGURE-22: NORTH-SOUTH CROSS SECTION ; 37
FIGURE-23: 3-DIMENSIONAL DIAGRAM 38
-------
Technical Background Document
Date: Feb. 17,2004
Page: 6 of 45
I. GEOLOGICAL FRAMEWORK
A. REGIONAL SUMMARY
1. PHYSIOGRAPHY
This ground-water modeling project covers the geographic area indicated by Figure-1.
The project area and the Plaquemine aquifer lie within the Gulf Coastal Plain physiographic
province. The Gulf Coastal Plain physiographic province covers the entire State of Louisiana
(Renken, 1998). As reported by Meyer and Turcan (1955), the Baton Rouge area is divided
approximately by the Mississippi River into two sections of the Gulf Coastal Plain; the
Mississippi Alluvial Plain and the East Gulf Coastal Plain. The Mississippi Alluvial Plain has a
relief of approximately 20 feet measured from the crest of the natural levee to the lowest back-
swamp surface. The East Gulf Coastal Plain, to the east of the Mississippi River, is a moderately
dissected area of low relief, with altitude ranging from approximately 120 feet above mean sea
level (MSL) in the northern part of the area to approximately 30 feet MSL in the southern part
(Meyer and Turcan, 1955).
The major geological formations underlying the Gulf Coastal Plain consist of sedimentary
rocks that range from unconsolidated to poorly consolidated clastic rocks which are Jurassic to
Quaternary in age (Renken, 1998). These formations dip gently southward towards the Gulf of
Mexico Geosyncline or towards the Mississippi Embayment. Younger Gulf Coastal Plain rocks
of late Eocene to Pliocene do not extend as far north into the embayment as do older strata, but
instead crop out as a belt that parallels the coastline and dip gently southward towards the Gulf of
Mexico. Figure-2 illustrates the southward dip and relative ages of Gulf Coastal Plain geologic
units and includes generalized illustrations of growth faults and salt dome structures. No salt
domes appear to be located in the immediate area of Plaquemine, Louisiana. For more detailed
information on salt-dome locations in the Gulf Coastal Plain, the reader is referred to Beckman
(1990), who presents the locations of 624 salt domes on a 1:1,500,000 scale base map.
2. REGIONAL STRUCTURAL FEATURES
Martin and Whiteman (1989) present a diagram of major regional structural features of
Louisiana and adjacent States (Figure-3). Regional structural features relative to southern
Louisiana include the axis of the Mississippi Structural Trough and the Miocene axis of the Gulf
Coast Geosyncline. On a local scale, a series of east-west trending faults are known to exist in a
belt over seventy miles long and up to eight miles wide extending between Slidell and Baton
Rouge (Durham and Peoples, 1956) and include the Baton Rouge fault zone.
-------
Technical Background Document
Date: Feb. 17,2004
Page: 7 of 45
3. SURFACE GEOLOGY
A surface geological map is included as Figure-4. This is a portion of the map prepared
by Saucier and Snead (1989) describing the Quaternary Geology of the Lower Mississippi Valley
at a scale of 1:1,100,000. As indicated for the Plaquemine area, on the western side of the
Mississippi River Holocene backswamp deposits occur and consist of overbank deposition not
directly affected by meandering stream channels. On the eastern side of the river are Holocene
meander belt deposits which include channel deposition related to lateral migration of past and
present river courses. In terms of subsurface geology, deposits of Pleistocene age underlie
Holocene sediments and form uplands to the east of theMississippi River. The Plaquemine
aquifer includes only Holocene and late Pleistocene age material.
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 8 of 45
FIGURE-1: LOCATION MAP AND MODEL DOMAIN
i
•
•'/"
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 9 of 45
FIGURE-2: REGIONAL SUBSURFACE GEOLOGY
(From Renken, 1998)
-^H^
, <-^3nP^^ ^
-----^>:i^l IT -,v •--tr-T-,. *" ''ovvx,
:-" >'^BurTT>Vr--
^:^^^\r
?>~^CX:^A;
EXPLANATION
Geologic units
| I Quaternary sedimentary deposits
[ '• .~ I Late Tertiary sedi mentary rocks
Early Tertiary sedimentary rocks
Late Cretaceous sedimentary rocks
Early Cretaceous sedimentary rocks
Jurassicsedimentary rocks
Triassic and Permian sedimentary rocks
Pennsylvanian sedimentary rocks
Mississippian, Devonian, Silurian, Ordovican,
and Cambrian sedimentary rocks
Precambrian igneous and metamorphic rocks
LJthologic units
Crystalline rocks
Limestone and dolomite
Sandstone, siltstone, and shale
Novaculitic and cherty sandstone and shale
Marl, limestone, sandstone, limey clay and shale
Clay and/or s hale
Sand
Evaporite (salt)
Fault—Arrows show relative movement
Boundary of physiographic province
Boundary of physiographic section
Coastal Plain rocks dip gently toward the Gulf at Mexico geosynclme or toward the center of the
Mississippi Embayment Growth or listnc faults, which formed during and after these rocks were deposited,
displace some strata. Diapiricflowage of salt strata, which is caused by the salt being overloaded by thick
accumulations of younger sedimentary strata, has resulted in the formation of intrusive salt domes. Flat-
lying carbonate rocks dominate much of the Ozark Plateaus Province in northern Arkansas, whereas
intensely folded and faulted shale, sandstone, and chert-nova cu lite underlie the Ouachita Province
-------
Technical Background Document
Date: Feb. 17,2004
Page: 10 of 45
FIGURE-3: MAJOR REGIONAL STRUCTURAL FEATURES
(From Martin and Whiteman, 1989)
94-
08-
I
OKLAHOMA •
34-
32-
30-
MEXICO
28-
6 20 Vo 60 80 l6o KI.OMETERS
EXPLANATION
— — Boundary el th» regional study arc*
—— Boundary o( tho coaalal Lowland* iludy araa
X Axla of aynollna
^ Axla ol antlollna
I I Fault-—Hachura* an downthrown alda
PROJECT AREA
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 11 of 45
FIGURE-4: SURFACE GEOLOGY
(From: Saucier and Snead, 1989)
Modeling project area outlined by the square. (Hb) refers to Holocene backswamp deposits-areas
of overbank deposition not directly affected by meandering channels. (Hmml) refers to Holocene
Mississippi River meander belt deposits-areas of overbank deposition related to lateral migration
of past and present river, also point bar deposits.
-------
Technical Background Document
Date: Feb. 17,2004
Page: 12 of 45
B. LOCAL GEOLOGY
1. DEPOSITIONAL ENVIRONMENTS
The depth of the freshwater bearing part of the Plaquemine aquifer at the project area
extends to approximately 560 feet below land surface. This approximate depth is indicated on
maps and cross sections by Cardwell and Rollo (1960) and Whiteman (1972). Below this depth
is brackish water or salt-water bearing sand. The geologic descriptions in this section are focused
on strata extending from the surface to approximately 560 feet below land surface.
Cardwell and Rollo (1960) refer to the area south of Baton Rouge as a deltaic plain or
complex, and that these deposits consist of three general depositional units: (1) older deltaic
Pleistocene deposits, (2) an alluvial sequence of late Pleistocene and Holocene, and (3) Holocene
deltaic deposits. The main body of late Pleistocene alluvial material lies west of the river but is
only of limited extent east of the river due to the present river course marking the eastern limit of
river migration (meandering). Cardwell and Rollo (1960) describe the Holocene deltaic deposits
as being thin deposits which are interbedded with the fine grained upper part of the river
alluvium and consist predominantly of clay and sandy clay.
2. STRATIGRAPHY
The 1972 study by C.D. Whiteman subdivided the upper 600 feet of sediments into
different units based on age, depositional environment, and lithology. Whiteman subdivided the
sequence into two general depositional units: (1) alluvial deposits of Pleistocene and Holocene
age and (2) older deltaic depositsiof Pleistocene age. Whiteman refers to a "top stratum"
overlying the Holocene alluvial deposits, representing natural levee deposits consisting of silts
and clays. These natural levee deposits are further indicated on the geologic map by Saucier and
Snead (1989), given in Figure 4 as baekswamp deposits (Hb).
Figures 5 and 6 present Whiteman's 1972 depiction of hydrostratigraphy over the
modeling project area. Section A-A' (Figure 5) presents an east to west profile and C-C' (Figure
6) presents an northeast to southwest profile. The wells nearest the project area providing
geologic control points for these published cross-sections are wells Ib-181, well Ib-191, and well
57104. In the vicinity of these three wells, Whiteman (1972) indicates that the top stratum
extends to approximately 100 feet below land surface. The top of the Plaquemine aquifer begins
at the base of the top stratum and extends to approximately 560 feet below land surface. The
aquifer is divided into an upper sand unit and lower sands which are separated by a silt or clay
layer extending from approximately 200 to 250 feet below land surface. Whiteman (1972)
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 13 of 45
indicates that the base of the Mississippi River extends downward to approximately the same
depth as the top of the Plaquemine aquifer.
During 2002^ EPA Region 6 sub-contractors completed 5 subsurface borings in the area
between the City of Plaquemine and DOW Chemical (EPA Work Assignment No. R06757,
Corrective Action Support to the Louisiana Department of Environmental Quality). These
borings were completed to the base of the upper sand unit. Boring logs are contained in the
Appendix.
EPA lithologic logs are consistent with published cross-sections. EPA logs indicate the
contact between the base of the top stratum and the top of the Plaquemine aquifer to be
approximately 100 to 110 feet below land surface, which is consistent with the contact elevation
reported by Cardwell and Rollo (1960), Saucier (1969), and Whiteman (1972). EPA logs
describe the upper 100 feet as consisting of silty fine sand and wood, with green to brown to gray
clays. From 100 feet to the base of the upper sand unit, the sediments were described as dense,
gray, fine to medium to coarse sand, and clay, with medium gravels near the bottom of the unit.
Generally there is a downward coarsening sequence which also remains consistent with
previously published work.
Whiteman's 1972 cross section C-C' depicts a southward dipping local strata which
appears to be consistent with regional southward dip of the Coastal Lowlands Aquifer System.
Saucier (1969) presents a map of the top of the Pleistocene dipping southward beneath the
project area, but this dip appears to represent a local depositional surface rather that the dip of the
Coastal Lowlands Aquifer System (Figure 8).
Figure 8 presents surface geology: consisting of backswamp deposits west, point bar
deposits, and natural levee deposits. In cross-section, Saucier (1969) places the top of the
Pleistocene at an elevation that is close to Whiteman's (1972) base of the Plaquemine aquifer
upper sand unit. As noted on Saucier's cross-section, the amount of coarse grained material
increases near the bottom of the upper sand unit (Figure 9).
3. THE BATON ROUGE FAULT
The Baton Rouge fault was recognized by Fisk (1944) on an aerial photograph showing
displacement of an abandoned floodplain, and was later described by Durham and Peoples (1956)
in an abstract prepared for the Louisiana Geological Survey. McCulloch (1991) reports that the
surface of East Baton Rouge Parish is traversed by at least two, active east-west striking faults,
the Baton Rouge and Denham Springs-Scotlandville faults which may be broken by a number of
other older and inactive but similar faults. The locations of these faults over East Baton Rouge
Parish, as prepared by McCulloch (1991) is given in Figure 10. Durham and Peoples (1956)
reported that the Baton Rouge fault may be traced from the Mississippi River floodplain in south
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 14 of 45
Baton Rouge eastward into Livingston Parish, a distance of twenty-file miles.
The Baton Rouge fault has a dip angle of 70 degrees from land surface to about 3000 feet,
then decreases from 60 to 55 degrees from 4000 feet to 6000 feet, then ultimately to dip at
approximately 45 degrees at depths near 8000 feet (Durham and Peoples, 1956). Fault
displacement apparently varies with depth, ranging from a few tens of feet near the surface to a
few hundred feet of displacement at much greater depth intervals. McCulloch (1991) reports that
displacement of the Baton Rouge fault at the surface in East Baton Rouge Parish averages about
20 feet, and displacement increases with depth as a result of movement contemporaneous with
sediment accumulation. Surface topography near the Baton Rouge fault exhibits a "reverse drag"
or "roll over" effect as documented by Durham and Peoples (1956) and by McCulloch (1991).
This has caused a reversal of surface dip from southward to northward within a one-half to one
mile wide belt south of the Baton Rouge fault.
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 15 of 45
FIGURE-5: HYDROSTRATIGRAPHY (EAST-WEST CROSS SECTION)
(From Whiteman, 1972)
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 16 of 45
FIGURE-6: HYDROSTRATIGRAPHY
(NORTHEAST TO SOUTHWEST CROSS SECTION)
(From Whiteman, 1972)
-------
Technical Background Document
Date: Feb. 17,2004
Page: 17 of 45
HGURE-7: CROSS SECTIONS
(From Whiteman, 1972)
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 18 of 45
FIGURE-8: TOP OF THE PLEISTOCENE
(From Saucier, 1969)
LEGEND
NATURAL LEVEE
POINT BAR
POINT BAR AREAS CONTAINING
i UNUSUALLY LARGE AMOUNTS OF
FINE-GRAINED MATERIALS
BACKSWAMP
THIN BACKSWAMP DEPOSITS l<30 FTI
OVERLYING BURIED MEANDER BELT
DEPOSITS
] ABANDONED DISTRIBUTARIES
• INDEFINITE CONTACT
• CORPS OF ENGINEERS BORINGS
0 BORINGS BY OTHER AGENCIES
y FAULT AFFECTING NEAR
0 SURFACE DEPOSITS
_ ELEVATION OF TOP OF PLEISTOCENE
DEPOSITS IN FEET MSL
Qtp PLEISTOCENE PRAIRIE TERRACE
BORINGS USED TO CONTOUR BURIED
TERTIARY OR PLEISTOCENE SURFACE
-------
Technical Background Document
Date: Feb. 17,2004
Page: 19 of 45
FIGURE-9: CROSS-SECTION C-C'
(From Saucier, 1969)
L.EGE NIP
jil.N V I R O N M E N T 3. .O F PE P O 5 IJT ION
^^= ^^ ALLUVIAL APRON
TOPSTRATUM -
LIT HO LOG 1C TYPE 3
CPLeiSTOCEME DE POSITS)
1" _~^~ ^ S 1 LT V C U AV
•=•:•:•'-••-:. \ SAND
3AMOY CLAV
SAMD 1_ f—"~ 1 J SAND (.SP, Sw)
SA NO a. OR AVt L I I 1 GRAVEL CdP, Gw)
1 (E-^^K I S.UTV SAMO (S1V,5
SA rsiDY C L/
-SILT-
SAFMC.Y SILI
-------
Technical Background Document
Date: Feb. 17,2004
Page: 20 of 45
FIGURE-10: LOCATION OF MAJOR FAULTS NEAR BATON ROUGE
(From McCulloch, 1991)
KNOWN FAULT. TEETH ON
DOWN-THROWN SIDE
POSSIBLE FAULT BASED ON SUGGESTIVE
SURFACE AND/OR SUBSURFACE
INDICATOnS: TEETH ON OOWNTMROWN
SHJE
-"- CONJECTURAL FAULT, EVIDENCE
UNKNOWN: TEETH ON OOWNTHROWN SIDE
PRINCIPAL SOURCES:
I PARSONS (1967)
3 VAN SICLEN (1971)
3 DEPT. PUBLIC WORKS (197«)
4 ROLAND AND OTHERS (1901)
3 THIS REPORT
MILES
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 21 of 45
n. HYDROGEOLOGY
A. REGIONAL AQUIFER SYSTEMS
The purpose of this section is to summarize and explain the regional hydrogeology and
describe how the regional aquifers relate to the Plaquemine aquifer. The regional aquifers
summarized here are the Coastal Lowlands aquifer system and the Mississippi River Valley
alluvial aquifer. The locations of these major aquifers are given in Figure-11, map scale
1:2,000,000. The locally named Plaquemine aquifer is related to the lithology of the Mississippi
River valley alluvial aquifer but is associated with the hydrology of the Coastal Lowlands aquifer
system.
The Coastal Lowlands aquifer system consists of a gulf-ward thickening, heterogenous,
unconsolidated to poorly consolidated wedge of discontinuous beds of sand, silt, and clay, with
permeable zones which consists of interbedded sand and clay (Renken, 1998). An idealized
diagram of the aquifer system is included as Figure-12. As reported by Martin and Whiteman
(1989), the degree of sediment and aquifer heterogeneity is significant which is reflected by
major lithologic changes that occur over short distances vertically and horizontally. Stratigraphic
dip is to the south and ranges from approximately 10 to 50 feet per mile in the outcrop area to
over 100 feet per mile in the southern part of the aquifer system and at depth.
The Mississippi River Valley alluvial aquifer is the upper aquifer of the Mississippi
embayment aquifer system, and extends southward from the head of the Mississippi Embayment
to the Gulf of Mexico (Ackerman, 1996). The map by McGee (1997), Figure 13, illustrates the
freshwater bearing part of the aquifer which extends as far south as Saint Mary, Saint Martin, and
Assumption Parish. In describing ground-water flow, McGee cites the work conducted by
Whitfield (1975) and states that ground-water flow adjacent to streams is nearly perpendicular to
stream flow, discharging from aquifers into streams during dry periods and moving a short
distance from streams into an aquifer during periods of high stream stage.
Ackerman (1996) reports that the Mississippi River Valley alluvial aquifer is laterally
equivalent to the upper part of the Coastal Lowlands aquifer system in east-central Louisiana.
Renken (1998) reports that the Quaternary alluvial deposits and deltaic deposits of the lower
Mississippi River Valley are lithologically similar to and in good hydraulic connection with the
underlying deposits of the Coastal Lowlands aquifer system. Therefore, the Quaternary sands
and gravels of the Mississippi River Valley alluvial aquifer continue on to the Gulf of Mexico
but are included as part of the Coastal Lowlands aquifer system.
The Plaquemine aquifer has been included as part of the Coastal Lowlands aquifer
system. The Plaquemine aquifer has been grouped with other Holocene-upper Pleistocene
aquifers in Louisiana as part of the U.S. Geological Survey Regional Aquifer-System Analysis
-------
Technical Background Document
Date: Feb. 17,2004
Page: 22 of 45
(RASA) program. The broad objective of the RASA program is to assemble geologic,
hydrologic, and geochemical information, to analyze and develop an understanding of the aquifer
system, and to develop predictive capabilities that contribute to effective management of the
aquifer systems (Martin and Whiteman, 1999).
Under the RASA program, the Coastal Lowlands aquifer system has been divided into
five overlapping regional permeable zones, Zones A through E, to assist in quantifying the
system. Figure-14 is a regional north to south cross-section showing the relative positions of
Zones A-E. The Plaquemine aquifer is included in Zone A, consisting of Holocene-upper
Pleistocene deposits. See Table-1 for the grouping of aquifers in the five regional permeable
zones. Figures 15-20 illustrate sand thickness and percentage, predevelopment ground-water
flow, ground-water flow under 1980 conditions, and the distribution of ground-water pumpage in
Zone A.
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 23 of 45
FIGURE-11: MAJOR REGIONAL AQUIFERS
(From Renken, 1998)
Wai er-yitiding rocks of Segment 6 that crop out can be
grouped into four major aquifer systems, two minor aquifer
«y*tems, and two minor aquifers The Western Interior Plaint
confining system Is pan of a widespread, geologically complex, and
poorly permeable sequence. Individual geologic units or pans of
units within the confining system locally yield water to wells The
outcrop extent of these hydrogeologic units Is shown here.
EXPLAKAT1ON
rtqjor aquifer systems
Surflclal aquifer system
Coastal lowlands aqutfer system
Mind jippi •mbaymantaqutfer tystem
Ozark Plata* ws «qurftr fytttfn
Minor aquifers and aquifer systems
Aquifers «nd aquifer systems in rocks of
Cretaceous age
Southeastern Coastal Plain aquifer system
(B lack Wa rrio r River «q ulfer)
Tokio-Woodbine aquifer
Edwards-Trmily aqutfenystem (Trinity aquifer)
Ouachh^ f*y nt*in* aquifer
Confining systems and confining units
Western Interior Plains confining system
(locally a minor aquifer)
Confining unit
Base mod if wd from U.S. Geological Survey dig rtaldata,
1:2.000,000, 1972 Alters Equal-Area Conic projection
Standard parallels 2fl°30'ind 4E°30',central meridian -96°00'
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 24 of 45
FIGURE-12: COASTAL LOWLANDS AQUIFER SYSTEM
(From Martin, 1989)
R«ch»rf« n£'
«rt» i
I
£:*
N /#y/7// g|3
• /'//" '/ 23 • J2
tr 111
« c i • £
a 1' o 9
Sr c i i- c
* c ' " P
gljsl
5s 5 2
a | a
Diicharg* ant
s
QULF OF MEXICO
EXPLANATION
Sind
Clay
Direction of ground-water flow
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 25 of 45
FIGURE-13: MISSISSIPPI RIVER VALLEY ALLUVIAL AQUIFER
(From McGee, 1997)
93
ARKANSAS 92'
i
91'
I
, Stufyaru.wlure
•*tm®, tlM Mississippi
||» Rlmalluvial
^||r aqulereontalns
' ~jjj Intlwattr
l~i?f""*' * T-- xl-?1™
^1;Nv^ sit A J^ 7
:% '^?T iV r-"^'
^MAflY^ X'~* /"' v-.
>
TERflEBONNE
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 26 of 45
FIGURE-14: DIAGRAM OF ZONES A-E
(From Martin, 1990)
TPNN
OKLA.I ARK
2IE
<\z
£ -
— 'iy;
=rl ...
-5^'' '*''-'•'•*! •' •
-------
3"S?
« £
S3
Gulf Coast RASA
regional
permeable zone
Zone A
(Holocene-upper
Pleistocene
deposits)'
ZoneB
(Lower
Pleistocene-upper
Miocene deposits)
ZoneC
(Lower Pliocene-
upper Miocene
.deposits)
ZoneD
(Middle Miocene
deposits)
ZoneE
(Lower Miocene-
upper Oligocene
Model
layer
number
2
3
4
5
6
Aquifers in Louisiana
Upland terraces Alluvial "400-foot" sand
"Shallow sands" Plaquemine of Baton Rouge
Chicot Grainercy ("200-foot" "600-foot" sand
"200-foot" sand sand of New Orleans) of Baton Rouge
of Lake Charles Norco ("400-foot" Atchafalaya
"500-foot" sand sand of New Orleans)
of Lake Charles Gonzales-New Orleans
"700-foof'sand ("700-foot" sand
ofLakeCharles ofNewOrleans) Ponchatoula (npper)
"1,200-foot" sand of New Orleans Ponchatoula (lower)
"800-foot" sand of Baton Rouge Big Branch
"1,000-foot" sand of Baton Rouge Kentwood
"1,200-foot" sand of Baton Rouge Abita
"1,500-foot" sand of Baton Rouge Covington
"1,700-foot" sand of Baton Rouge SlideB
Blounts Creek
Evangeline (upper)
"2,000-foot" sand of Baton Rouge Evangeline flower)
"2,400-foot" sand of Baton Rouge
Tchefuncta
Hammond
Zone 2 of Florida parishes and Williamson Creek
PointeCoupee (upper) Jasper (upper)
Zone 2 of Florida parishes and Jasper (lower)
PointeCoupee (lower) Camahan Bayou
Zone 3 of Honda parishes and
PointeCoupee
"2,800-foot" sand of Baton Rouge
Anrite
Ramsay
Franklinton
Catahoula
Aquifers in Mississippi
Undifferentiated aUuvial and
terrace
Mississippi River alluvial
Terrace
Citronelle
Miocene (upper)
Miocene (middle)
Miocene (lower)
Aquifers in
Alabama
Absent
Absent
Alluvial
Terrace
Citronelle (upper)
Citronelle (lower)
Miocene (upper)
Miocene (lower)
Aquifers in
Florida
Absent
Absent
Absent
Sand and
gravel (lower)
Sand and
gravel (tower)
b. 17, 2004
of 45
Table-1 : REGIONAL PERMEABLE ZONES (RASA-STUDY)
(From Martin, 1999)
t
a
a
1
i
f
1
i
•t.
'Grubb, 1987, p. 110-113. RASA, Regional Aquifer-System Analysis. Because of regional structure and the way the regional permeable zones were defined, the zones cross
chronostratigraphic lines in places; the greatest discrepancies occur in western Florida and southwestern Alabama, where zones D and E consist of sediments of Holocene to middle
Miocene age.
-------
Technical Background Document
Date: Feb. 17,2004
Page: 28 of 45
FIGURE-15: TOTAL THICKNESS OF SAND IN UPPER PLEISTOCENE AQUIFER
(From Martin and Whiteman, 1989)
DOWHDIf LIMIT
OF AQUIfCR
EXPLANATION
-200- LINE OF EQUAL THICKNESS.
tnt«rv«l 100 (••!
0 20 40 60 80 100MLES
I I t —L 1 J MEXICO
0 20 40 60 .80 100 KILOMETERS
-------
Technical
Date: Feb. 17, 2004
Page: 29 of 45
FIGURE- 1 6:
PERCENTAGE OF SAND IN UPPER PLEISTOCENE AQUIFER
(From Martin and Whiteman, 1989)
\
DOWNDIP LIMIT
OF AQUIFER—
o
EXPLANATION
-60- LINE Of EQUAL SANO
PERCENTAGE, Interval
20 percent
X
0 20 40 60 80 100 MIES
h^r-T1-^—' * MEXICO
0 20 40 60 80 tOO KILOMETERS
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 30 of 45
FIGURE-17: MEASURED 1980 WATER-LEVEL ALTITUDES IN
UPPER PLEISTOCENE AQUIFER
(From Martin and Whiteman, 1989)
DOWHDIf LIMIT
OF AOUIFIK
EXPLANATION
-100- POTCNTIOMETMIC CONTOUR
--•how* iltltud* tt which wittr
would hiv* Hood In tightly c«i«a
w*ll(. Contour Interval 26 (••!.
Dilum !• ••« l«v»l
• CONTROL POINT
gm AMA WHtRE WATtM UVCL 18
EH NOT CONTOUMO
0 ?0 40 60 80 100 MILES
| , ' , ,' , ', ' ' MEXICO
0 20 40 60 80 100 KILOMETERS
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 31 of 45
FIGURE-18: PREDEVELOPMENT FLOW IN UPPER PLEISTOCENE AQUIFER
(From Martin and Whiteman, 1989)
EXPLANATION
Rate and direction of flow,
In cubic feet per day.
0 TO 100.000
- 100,000 TO 500.000
> 500.000 TO 1.000,000
1.000.000 TO 5.000.000
DOW NO If LIMIT
OF AQUIFER
Flow through top of aquifer
( I Area with downward vertical flow
HI Area with upward vertical flow
20 40 60 80
I ' I .' I '| L
OOMLES
0 20 40 60 80 100 KILOMETERS
MEXICO
-------
Technical Background Documpnt
Date: Feb. 17, 2004
Page: 32 of 45
FIGURE-19: GROUND-WATER FLOW IN 1980, UPPER PLEISTOCENE AQUIFER
(From Martin and Whiteman, 1989)
-fc
EXPLANATION
R«t» and direction of flow,
In cubic feet per day.
• 0 TO 100.000
«• 100,000 TO 500.000
> 500.000 TO 1,000.000
Flow through top of aquifer
_J Area with downward vertical flow
Area with upward vertical flow
MEXICO
n 20 40 60 80 100 KILOMETERS
-------
Technical Background Document
Date: Feb. 17, 2004
Page: 33 of 45
FIGURE-20: AERIAL DISTRIBUTION OF PUMPAGE FROM
UPPER PLEISTOCENE AQUIFER
(From Martin and Whiteman, 1989)
•fc.
<%•
"ooaaaoaoaoo«•»<£
* »***aa*aaaaa«*a*«F
o * *»ooaoaaoooo»v
»««OOOO«OOO *
EXPLANATION
LIMIT OF
AQUIFER "
PUMPAQE RATE.
IN MILLION
GALLONS PER DAY
+ 0-1
0 1-6
D 6-10
B 10-16
• 16-20
• QrMtw th*n 20
T
>.
%.
0 20 40 60 80 100 MILES
0 20 40 60 BO 100 KILOMETERS
••%.»
MEXICO
«».
-------
Technical Background Document
Date: Feb. 17,2004
Page: 34 of 45
B. LOCAL HYDROGEOLOGY
1. SEASONAL/CYCLIC MOVEMENT; SALT WATER FLUSHING
Water movement in the Plaquemine aquifer is described by Whiteman (1972). Whiteman
reports that there seems to be 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. Whiteman (1972) also
reports that although water levels normally provide the clearest indication of the direction of
movement of water in an aquifer, seasonal movement of water into and out of temporary storage
in the Plaquemine aquifer is so large that it masks the effects of much smaller net long-term
movement. In such cases Whiteman (1972) states that indirect evidence must be used to
determine the direction of net water movement. Whiteman (1972) also reports that there is
chemical evidence indicating that there is net movement towards the river from water entering
the relatively high upland terrace deposits northeast of the project area.
Whiteman (1972) reports that at high river stages water enters the Plaquemine aquifer
from the river and moves outward from both sides of the river. Under maximum hydraulic
gradients of about 5 feet per mile that occur when the river is rising rapidly, water would have a
velocity of less than 1-foot per day in the upper sand unit. Therefore, Whiteman (1972) states
that water moves outward only a short distance during the seasonal periods of high river stage,
and as the river begins to fall the direction of water movement is reversed and moves from the
aquifer towards the river. Whiteman (1972) also states that the amount of water involved in the
cyclic movement into and out of the aquifer is large because of the large area involved, but is
actually less than 1% of the volume stored in the aquifer.
As reported by Whiteman (1972), most of the sands making up the aquifer probably
contained salt water at some time. Whiteman (1972) reports that fresh water west of the
Mississippi River entered the aquifer from the river and through the top stratum, and a flushing
process has been underway since the change in the course of the Mississippi River from the west
side of the alluvial valley to its present position near the east side, and is probably still in
progress at a slow rate. The limit of flushing at the time of Whiteman's 1972 report is indicated
by the position of the fresh-salt water interface given by Figure 21, showing a south-southwest
direction of saltwater flushing. This flushing may reflect a net long-term movement of ground-
water in the Plaquemine aquifer.
2. GROUND-WATER LEVEL MEASUREMENTS DURING 2001-2003
Cross sections by Whiteman (1972) (Figures 5 and 6— this document) show that the
-------
Technical Background Document
Date: Feb. 17,2004
Page: 35 of 45
Mississippi River cuts through the top stratum and into the aquifer upper sand unit, to a depth of
approximately 100 feet below land surface. Just to the south of the City of Plaquemine, Saucier
(1969) illustrated that the Mississippi River cuts to approximately 65 feet below land surface
(Figure 9-this document). Deep parts of the river channel provide direct hydraulic contact
between the Plaquemine aquifer and the river, and additional connection may be provided by
point bar deposits. Ground-water occurs under artesian conditions throughout the Plaquemine
area where wells may be flowing or non-flowing depending Mississippi River stage.
For this modeling project, ground-water level measurements are available for 51-wells
and piezometers located in the general project area completed in the upper sand unit of the
Plaquemine aquifer. For many of the wells, measurements have been collected every 7-days
during the period from October 15, 2001 until June 2, 2003. Data collection from other wells
(EPA owned wells and other piezometers) began later than October 15, 2001 depending on
installation date of the respective wells and piezometers. Data shows significant water level
fluctuations in the Plaquemine aquifer, and the ground-water fluctuations are generally
consistent with the time of river stage rise or fall. The reader is referred to published work by
Whiteman (1972), and Meyer and Turcan (1955), for a discussion of lag time between river
fluctuation and ground-water rise or fall, and magnitude of ground-water rise or fall with distance
from the Mississippi River.
3. THE BATON ROUGE FAULT
Tomaszewski (1998) reported on the hydraulic effects of the Baton Rouge fault during a
study of the effects of pumpage on the 1500-foot sand south of the Baton Rouge fault, near
Brusly, J^ouisiana, and states that previous investigations have shown the Baton Rouge fault to
act as a leaky barrier to the flow of water across the fault. Tomaszewski (1998) reports that the
Baton Rouge fault is an important hydrogeologic control, and fault displacement ranges from a
few feet at land surface to as much as 350 feet at the top of the 2000-foot sand.
Figure 22 is a generalized north to south hydrogeologic section showing the Baton Rouge
fault, the Mississippi River alluvial aquifer and deeper aquifers, and fault displacement in each
aquifer. Figure 23 is a similar diagram by Kuniansky (1989). Tomaszewski (1998) further states
that at the base of the Mississippi River valley alluvial aquifer, displacement of the fault is only
about 30-feet, but below the aquifer displacement gradually increases with depth.
-------
Technical Background Document
Date: Feb. 17,2004
Page: 36 of 45
FIGURE-21: DEPTH OF OCCURRENCE OF FRESHWATER
(From Whiteman, 1972)
RI2E.
HI*.
its.
xfxt -
-------
Technical Background rw..
Date: Feb. 17, 2004
Page: 37 of 45
FIGURE-22: NORTH-SOUTH CROSS SECTION THROUGH BATON ROUGE FAULT
(From Tomaszewski, 1998)
Section 8 - B'
WBR-53
Baton Rouge
fault
N<
1 KILOMETER
-------
Technical Background Document
Date: Feb. 17,2004
Page: 38 of 45
FIGURE-23: 3-DMENSIONAL DIAGRAM SHOWING
FAULT AND MAJOR SAND UNITS
(From Kuniansky, 1989)
EXPLANATION
U BAND AND QPIAVEI
OCLAY
AND 8H.T
UX3ATIOM HAP
-------
Technical Background Document
Date: Feb. 17,2004
Page: 39 of 45
HI. COMPUTER MODEL SETUP OVERVIEW
The purpose of this section is to describe the preliminary conditions being considered for
setting up main components of the steady-state and transient flow models. It is likely that during
the course of model development and calibration, there will be changes to these preliminary
conditions as deemed necessary to improve model function and reliability.
A. MODEL GRID AND LAYERING
Model layering will represent major stratigraphy including the Plaquemine aquifer upper
sand unit and lower sand, an intervening clay/slit layer between the upper sand unit and lower
sand, and overlying natural levee deposits (i.e., top stratum). The model total vertical depth
represents 560-feet, where 0-feet represents model bottom and 560-feet represents land surface.
This layering is derived from EPA lithologic logs (in Appendix) and from published cross-
sections such as those provided by Whiteman (1972) and Saucier (1969). The model top stratum
extends from the surface (560-feet) downward to the top of the Plaquemine aquifer upper sand
unit at 460-feet. The upper sand unit extends from 460-365 feet, the intervening silt/clay layer
from 365-315 feet, and the lower Plaquemine aquifer from 315-70 feet. From 70 feet to 0 feet
are cells that represent deeper sands, silts, and clays containing saltwater. Additional layering
will probably be included to correspond with well and peizometer screen intervals.
An appropriate grid cell size will be determined during the course of model development.
A refined grid will cover the main area of interest (i.e., the area of the vinyl chloride plume),
which might be approximately 100' x 100' for the transient model, and even smaller for the
steady-state models (i.e., 50' x 50'). Larger cells may be needed for the transient model in order
to keep model run time manageable. The entire grid will be large enough that any model
boundary flow effects will not affect results for the main area of interest.
B. HYDRAULIC PROPERTIES
Tabulated local aquifer test data are not readily available to EPA for this modeling
project. Therefore, property values will be based on published data and on best professional
judgement. Martin and Whiteman (1999) report that lateral hydraulic conductivity of sand beds
within the regional permeable zones were determined by examining 3077 aquifer tests, and report
that Martin and Early (1987) described the compilation, summarization, and statistical analysis of
1001 aquifer test results in Louisiana, some which were in the Coastal Lowlands aquifer system.
Martin and Whiteman (1999) summarize data for lateral hydraulic conductivity of sand beds for
permeable Zone A as including 1,126 aquifer tests conducted, with an arithmetic mean of 163
ft/d, a standard deviation of 152 ft/d, a harmonic mean of 49 ft/d, the average of harmonic and
arithmetic mean as 106 ft/d, and a geometric mean of 101 ft/d.
-------
Technical Background Document
Date: Feb. 17,2004
Page: 40 of 45
C. HYDRAULIC BOUNDARY CONDITIONS
1. Mississippi River
The Mississippi River appears to be the most important controlling hydraulic feature in
the model domain. Numerous river stage measurements have been taken at DOW Chemical
docks I and It; dock I is located on the southeastern edge of the DOW Chemical facility and dock
n is located along the northern edge. For the period of October 2001 to June 2003, river stage
measurements show considerable variability over time, with low river stages apparently
occurring in late Summer and Fall and high stages occurring in the Winter and Spring. During
this period, river stage measurements vary as much as 30-feet between high and low river stages.
Measurement variability is also evident between docks I and n even if measurements are taken
on the same day. For example, on November 26, 2003, the dock I measurements was 4.51 feet
above sea level and dock n was 5.38 feet above sea level; and on June 10, 2002, dock I was at
33.32 feet above sea level and dock JJ was measured at 36.23 feet above sea level. This type of
variability is probably due to the natural decrease in water level along the distance of the river
between the two docks. The full set of river stage measurement data and piezometer data is
located in the Appendix.
River stage measurements and piezometer readings for the Plaquemine aquifer upper sand
unit show that river stage affects water levels in the upper sand unit. Water level profiles by
Whiteman (1972) indicate that the range of fluctuations is directly related to distance from the
river, where the greatest fluctuations occur in wells located near the river. Wells near the river
reflect even small changes in river stage within a few minutes or hours, and wells distant (several
miles) from the river reflect smaller water level fluctuations with additional lag time. For
detailed information on fluctuations near the project area, the reader is referred to Whiteman
(1972), and Meyer and Turcan (1955). Whiteman (1972) also reports that pressure changes in
the aquifer that accompany a change in river stage are transmitted rapidly through the aquifer, but
the actual movement of water is very slow (< Ift/d under a maximum natural hydraulic gradient
of 5-feet per mile).
For computer modeling purposes, the river may be simulated as a river hydraulic
boundary which requires a numerical value for conductance:
r-KLW/
C~ /M
Where: K = hydraulic conductivity of river bed
L = length of reach through a cell
W = width of river through a cell
M = thickness of river bed
-------
Technical Background Document
Date: Feb. 17,2004
Page: 41 of 45
Alternatively, if supporting data are unavailable to calculate C, then a constant head
boundary may be tested, and if determined to produce reasonable results, substituted for the river
boundary. Either type of boundary can be set up in transient mode with river stage fluctuations
representing the actual measurements collected from docks I and H
2. General Head Boundaries
If necessary, a general head boundary can be added along the northern, western, and
southern model domain help provide links between local-scale ground-water flow and regional
flow. There seems to be very little available supporting data for upper sand unit hydraulic heads
in outlying areas. To help determine appropriate boundary conditions, Agency files and any
associated facility ground-water information will be utilized as sources of information.
3. No-Flow Boundaries and Inactive Model Cells
No-flow boundaries or inactive model cells will be utilized to improve the function of the
model as necessary. It is possible that such boundaries will be used for grid cells representing
deep sands, silts, and clays containing saltwater, and grid cells east of the Mississippi River.
D. WATER WELL (PUMPING WELL) DATA
This model may include water extraction to the extent reliable pumping data are
available. Pumping well data sets would need to include well name, x and y coordinates, screen
ID number, screen top elevation, screen bottom elevation, well stress period stop time, and
pumping rate. Some of this data may be available from the Louisiana Department of
Transportation and Development or LDEQ. It is likely that actual data for historical pumping
rates and schedules may not be available for many local water wells, and that many more wells
existed in the past than are in the current listed on the LaDOTD water well database. The
LaDOTD database contains 24-categories of water wells for Iberville Parish and West Baton
Rouge Parish. These categories included abandoned wells, chemical products, commercial,
destroyed, domestic, fire protection, food products, miscellaneous, inactive, institutional,
irrigation, monitoring, multipurpose, municipal, other, petroleum, plugged, power generation,
recovery, rigg supply, rural, stock, supply, and unknown. The total number of wells in this
database for Iberville Parish is 1,712, and the total number for West Baton Rouge Parish is 1,128,
although only a fraction of these are located in the model domain..
-------
Technical Background Document
Date: Feb. 17,2004
Page: 42 of 45
IV. REFERENCES
Ackerman, D.J., 1996, Hydrology of the Mississippi River Valley Alluvial Aquifer, South-
central United States: U.S. Geological Survey Professional Paper 1416-D, 56 p.
Beckman, J.D., and Williamson, A.K., 1990, Salt-Dome Locations in the Gulf Coastal Plain,
South-central United States: U.S. Geological Survey Water-Resources Investigations
Report 90-4060, 44 p.
Boniol, D., Autin, W.J., and Hanson, B., 1988, Recharge Potential of Louisiana Aquifers, A
Supplement to the State Aquifer Recharge Map and Atlas Plates, Louisiana Geological
Survey, Open-File Series No. 88-07.
Buono, Anthony, 1983, The Southern Hills Regional Aquifer System of Southeastern Louisiana
and Southwestern Mississippi, United States Geological Survey, in cooperation with
U.S. Environmental Protection Agency, Water Resources Investigations Report 83-4189.
Cardwell, G.T., and Rollo, J.R., 1960, Interim Report on Ground Water Conditions Between
Baton Rouge and New Orleans, Louisiana, Louisiana Geological Survey, in cooperation
with U.S. Geological Survey, Water Resources Pamphlet No. 9.
Durham and Peoples, 1956, Pleistocene Fault Zone in Southeastern Louisiana, Abstract,
Transactions of the Gulf Coast Association of Geological Societies
EPA, August 2003, Combined Quality Assurance Project Plan and General Work Plan, Potential
Ground- Water Row Directions and Contaminant Fate and Transport in the Plaquemine
Aquifer of Iberville Parish and West Baton Rouge Parish, Louisiana, EPA Region 6
Multimedia Planning and Permitting Division.
Fisk, H.N., 1944, Geological Investigation of the Alluvial Valley of the Lower Mississippi River,
Mississippi River Commission, U.S. Army Corps of Engineers.
Kuniansky, E.L., 1989, Geohydrology and Simulation of Ground-Water Flow in the 400-foot,
600-Foot, and Adjacent Aquifers, Baton Rouge Area, Louisiana
Martin, Angel, Jr., and Whiteman, C.D., Jr., 1985, Generalized Potentiometric Surface of
Aquifers on Pleistocene Age, Southern Louisiana, 1980, Water-Resources Investigations
Report 84-4331.
Martin, Angel Jr., and Whiteman, C.D., Jr., 1989, Geohydrology and Regional Ground-Water
Flow of the Coastal Lowlands Aquifer System in Parts of Louisiana, Mississippi,
-------
Technical Background Document
Date: Feb. 17,2004
Page: 43 of 45
Alabama, and Florida-A Preliminary Analysis, U.S. Geological Survey, Water Resources
Investigations Report 88-4100.
Martin, Angel, Jr., and Whiteman, C.D., Jr., 1990, Calibration and Sensitivity Analysis of a
Ground- Water Flow Model of the Coastal Lowlands Aquifer System in Parts of
Louisiana, Mississippi, Alabama, and Florida, U.S. Geological Survey Water Resources
Investigations Report 89-4189.
Martin, Angel, Jr., and Whiteman, C.D., Jr., 1999, Hydrology of the Coastal Lowlands Aquifer
System in Parts of Alabama, Florida, Louisiana, and Mississippi, U.S. Geological Survey
Professional Paper 1416-H.
McCulloch R.P., 1991, Surface Faults in East Baton Rouge Parish, Open-File Series No. 91-02,
Louisiana Geological Survey.
McGee, Benton, D., 1997, Occurrence of Nitrate in the Mississippi River Alluvial Aquifer in
Louisiana, June through December, 1993, Louisiana Department of Transportation and
Development, in cooperation with U.S. Geological Survey, Water Resources Technical
Report No. 61.
Meyer, R.R., and Turcan, Jr., A.N., 1955, Geology and Ground-Water Resources of the Baton
Rouge Area, Louisiana, U.S. Geological Survey Water Supply Paper 1926.
Morrison, R.B., 1991, Quaternary Nonglacial Geology: Conterminous U.S., The Geology of
North America, Volume K-2, Chapter 18, Quaternary Geology of the Lower Mississippi
Valley, The Geological Society of America, p. 547-582.
Renken, R.A., 1998, Ground-water Atlas of the United States; Regional Summary—Arkansas,
Louisiana, and Mississippi, HA-730-F, U.S. Geological Survey.
Renken, R.A., 1998, Ground-water Atlas of the United States; Surficial Aquifer System--
Arkansas, Louisiana, and Mississippi, HA-730-F, U.S. Geological Survey.
Renken, R.A., 1998, Ground-water Atlas of the United States; Coastal Lowlands Aquifer
System- Arkansas, Louisiana, and Mississippi, HA-730-F, U.S. Geological Survey.
Renken, R.A., 1998, Ground-water Atlas of the United States; Mississippi Embayment Aquifer
System- Arkansas, Louisiana, and Mississippi, HA-730-F, U.S. Geological Survey.
Saucier, R.T., 1969, Geological Investigation of the Mississippi River Area, Artonish to
Donaldsonville, Louisiana, U.S. Army Corps of Engineers, Technical Report S-69-4.
-------
Technical Background Document
Date: Feb. 17,2004
Page: 44 of 45
Saucier, R.T., and Snead, J.I., 1989, Quaternary Geology of the Lower Mississippi Valley,
Louisiana Geological Survey, Map Scale 1:1,100,000.
Tomaszewski, D.J., 1996, Distribution and Movement of Saltwater in Aquifers in the Baton
Rouge Area, Louisiana, 1990-1992, Louisiana Department of Transportation and
Development, in cooperation with U.S. Geological Survey, Water Resources Technical
Report No. 59.
Tomaszewski, D.J., 1998, Hydrogeology and the Effects of Pumpage on the 1500-Foot Sand
South of the Baton rouge Fault, Near Brusly, Louisiana, 1996, Louisiana Department of
Transportation and Development, in cooperation with U.S. Geological Survey, Water
Resources Technical Report No. 65.
Whitefield, M.S., Jr. 1975, Geohydrology and Water Quality of the Mississippi River Alluvial
Aquifer, Northeastern Louisiana, Louisiana Department of Public Works Water
Resources Technical Report no. 10, 29.
Whiteman, C.D. Jr., 1972, Ground Water in the Plaquemine-White Castle Area Iberville Parish,
Louisiana, Louisiana Geological Survey, in cooperation with U.S. Geological Survey,
Water Resources Bulletin 16.
-------
Technical Background Document
Date: Feb. 17,2004
Page: 45 of 45
APPENDIX
-------
Source: Lighthouse Road RECAP Assessment
And Corrective Action Plan, DOW Chemical,
January, 2004
tff™.\^ -/a5?fecsa?Sx ;
-------
Source: Lighthouse Road RECAP Assessment And
Corrective Action Plan, DOW Chemical, January, 2004
-------
VteSh
m EPA-I
ORK ASStQNyEKt Ma ?mj$7
GORRECtlVS ACTION, SUPPORT
PLAOUEMINE, LOUISIANA
P
SOIL DESefilPTKDW
.18,1. ft. .^AVD..
'
EJnj-wn s) t
A-r-'r-/ j
N
to
-so
I!
50-Kr
GtQy stlty
— greenish—grq y
frpm 24-0 lo Z§:0 «,
f
from 3i:.G to; 39& 'ft.
:Q:.3.0
0.-18
O.S2
C-S9
60 -
siity clay
-------
TQ&
ray, siliy fin-e son<3
-------
• LDQ OF
.«M
SL'
i^f-:--.
;..-"•.—;
!: M^Jiijf^ :ti^fis^ gray siity firta
• Medium dense :.gr0y mediuro1
S '
I-:.-:-3-
: »s
,-
"
i- tci
-------
-------
fTOINS WELL
"\::;EBM
f§MgisiT »
^;:Ai^
1^4- ^ b:g$
6'^ ft feg
tit) ft. ti
7S5. ?ia?j« ST.
SjlIE:
-------
-------
LOG OF BORING NO. EPA-2 (con't)
WORK ASSIGNMENT NO. R06757
CORRECTIVE ACTION SUPPOm*
PLAQUEMINE. LOUISIANA
/.
E
-------
* aft
1
•&.•
rs son<3.
•f^yYV-l/Jrj
_M9ti3
&Jiss grc-y'Sns to;
denrse grq.jf :siiiy fine sancf.
car -v. #HI.
No ?*;P/
-------
MONfTORING WELL
EPA-2
WORK ASSIGNMENT NO. R08757
CORRECTIVE ACTION SUPPORT
PLAOUEMINE, LOUISIANA
Lacking Cover-
2 inch Sf.ainless S'te.
-------
TYPE8OfiS&£
U3G; OP BORING MO. EPA-3;
, m®?&?
:syppoFiir
LOUISIANA
LOCATION- .
S: 5?
I
.£?
itff.
a
v J- EietATK>W .t&jJLftJJAVO
g
i Qark gmy silty- cia.y; v?/sh.eU fcc.grTjeni® (fill)
lo 4Q.Q ft.
b~.1"3?'3"S3
O.17
:0:r
0.11
3,45
oL
a;(^ spa;
-------
-------
£3
-------
Locking; Cover
v^.i 3urj;og^:
igQ :ffe,- bgs
7G ft., fogs
: ft. fogs
satt too
fx
-------
-------
SQJL
:!
:<~f.
fe
LM
5?#>rirri:; 'if-'***n iPi-1 'f^'"f^*» ^t/ ':n f:T
2>.C»Jj I; I: i-(-\i/»*.-t V-.I-;.^? j-.V -\?-£.;3~f i J}.
she}'
2.51
gr0y sllty :ffn,e sane.
^,
I ' I
7S8 MOR1M -Pt:
. SU8W i»
S«iii W'Mi
-------
£
t-i
I
:fiae
*
*Jl
fl-U
:i&
s.k
vi.
^
NS
m
M.eatum clsase
'fjras tp rneGiuirrt sairiid: :w:/v&Ggei$k>;n<3J coarse ;ss?vd
Medium, dfeoss. :gray ifine to coarse: sorrd- w/o.cca.sTonqi: fine Grove.:
-i&^
•&*&»
-------
LOO OF fiORWa WO. BPA-4 (con't)
WORK ASSIGNMENT NO. RO6757
CORRgCWE ACTION SUPPORT
PLAQUBAINE. LOUISIANA
-------
.SS,.ft. has
ft
\m ft, bgs
£0/40
-------
A 8 .. -IDIEIFIG HllljlK L
1
2
3
4
5
6
S
9
10
1 1
~
13
14
15
16
17
18
19
20
2_
_23_
?6
IF
29
30
~
34
^2'
36
"1
3
4
4
"I
"7
"3
4
"J
"J
5
5
VM
P4-7
>M-2
PA-3
PA-4
•PA-S
DDP1
DDP2
DDP3
DDP4
NNS
01C
702C
104C
705C
p»27
p»28
p*29
X \
648784.71
469273.72
668684.64
6493J4.77
6692J8.23
670456.36
648983.85
668476.00
669401.57
668164.33
669805.72
669460.77
669782,79
669328.36
670042,81
669631.69
669219.79
pl30 | 666408.64
PI31 668469.40
ZlfiW
3353089.96 790.00 Rig
3354475.77
3352437,72
3354191.95
3355145.69
3354953.35
3356022.52
3355597.70
3357080.79
3356012,90
3356089.18
3355330,93
3355330,95
3354767.24
3354329.75
3354508.97
3354975.56
3354412.77
pz32 , 668727.56 3353837.67
p;33 468784,87
£134
Doc*/
Doc* /'
647802,92
670450.66
669168.40
666999.45
p»34 I d47577.39
pl37 I 477090.44
pr38
pt3i
0140
P142
pt43
pr44
pz47
pita
ptSO
P2S1
p2S2
P2S3
pl!4
£>?£.' '
ptS7
TW-I'D
671309.37
549226.02
669221.92
668729.07
668729.03
669492.74
670642.61
648482,00
469404.41
468991.54
667633.79
587561.09
669318.16
669478.94
669034.11
669537,90
668428.49
668471.37
668471.37
449414,43
469507.67
SW-2S 449507.24
JfvTJir 6~49455.43
3353877.38
3353936.96
3354609.45
3357933.26
3349447.77
~3350366.27'
3350708.37
t
m
-f^fjflllljp-
180.00 jjjggH
TOC alle 8OF BOP attar 10/16/01 10/16/01 ] 10/22/01
11/06/02 11/02/02 JLEVEL
1
\
J
714.00 I 18.76! ••' 09
109.00 {SB!
104.60 if- :
720.00 JSS,
111: SO M"5(t
"'"'0.00 £pi!
109.00 j§|?
770.70 £8&?-
r?TooH|
'"•"pSs
776.50 K
MS-001I
^ 19.1 5,19
I
I
I
'15.00 ipii
»'4.00_^'
liiMm
1
TiToolll
3357027.90 179.00 Hi*
3354527.45
3354510.52
3353823.70
3353827.45
3354877.66
3355824,19
3354077.59
3355602.43
3355780.65
3355298.83
3356298.83
3353640,95
3353475.85
3353474.87
1352797.76
188.00 fHs
152.00 g|il
768.50jR|Ri
m.ooB
"8.20 H
120.00 jggj
122.20 US
'22,50 H
,'2r-"m
123.00'S5|
r 121. 20^
ril31o~l|
3353367.42] 122.00 f|fe
3355378,47 222.00 Sis
3356075.~20 204.00 l|i
~3355634.67 202.00 i§|
3357783.43
257. 00 US
3357774,78 *91.00_^
~33Sl7S4'.3j"' 251. i6~jm
SW-4S 669659.29 3357757.73 190.20 *&%
I
1
i
I
m
8
i
1
1
1
15,911
17.63]
14 34
17,15
4.88
6.46
4.19
6. sin
17.13 6.TT
17 84
1751
26.07 2676 20.15
2625 2692 20.96,
25.96 2669 21.35
19.35 20:09 15.06
2102 2172 16.67
23.21 2393 18.86
2006' 2077 16,34
I i 13,21
6.04
5. 52
6.14
5,74
5.27
4,73
4.85
4,77
4.43
4i4l"
4682 40.27 6,46
7.65
20.77
23-. 80
24.72
1 27.57
27.52
26.97
24,12
24,01
26.21
23.45
10/22/01
LEVEL
"i
j
i
1
16.73J 8.42"l
17,36]
" foo"
6.93
6.29
157i; 7T
13,02
iT'ijf
5.51
7.87
7.58
15. 76] 8,12
"1553| " 7.5
18. 09i 8.2
18.98 7,72
19.55
1354
15.19
1J29
14.95
11.99
38.11
7.66 1010
7.07
6.26
6.33
6.34
5.82
5.63
8.62
10.1
•• — "t - —
.). ..
I '
21.45
2404 24.27
[1 22,06
2267
26.43
27,40
L 25.09
29,72
21,95
24.35
21.47
24,32
T
- -•
'•s_ ^ — r
10/29/01.
10/29/01
LEVEL
1579
16,19
12,68
14,66
11,92
U.tJT
14,63
1442
17,15
18,00
18.55
12.51
14,20
16,35
14,02
11.06
37.36
11.00
|
9.36
8.1
7.71
8.45
6.61
8.95
8.69
9.26
8.61
9.14
8,7
8.07
7.28
7,32
7.28
6.75
6.56
9.37
11
.; : 1 ;
i i
....
\ "- ~
— r •
I
f
I
;;{
i
i ..
j i '
r
T
i _ .
i
-0
11/6/01
15,16
P
11/5/01
.EVEL
9.97
15991 8.3
1282J 7.77
1422
jTe|
8.89^
~
13,54 ».43
13,59
14.21
13.97
16,55
1749
18,07
12,11
13,80
15,92
1362
1068
36,59
11.73
f
!- -
I
9.16
9.67
9.06
9,74
9.21
8.66
7.68
7.72
7.71
7,15
6.94
10.14
11,73
r-
Q
11/12/01
17.38
17.29
1396
15,78
R
11/12/01
.EVEL
7.77
7
6.63
7.33
s
11/19/01
19,27
1884
1545
17,51
11/19/01
.EVEL
5.66
8.46
5.14
6.6
12,6 5,B3| 13.81] 4.72
15.35
15.26
16.09
15,67
187
19.39
19.59
13.08
1476
17.05
14.41
11,3
39,3
8,7
" ' 1- '
P ..
' : ' ; !
F'
,.., ....
t T ~
• • ----- - - ---• • - -
.:L:"V""::.":'.TM:"
L ... . . .
| 1
. . L.
7,62
7.48
7.79
7.36
7.59
7.31
7,03
8.71
6.76
6,58
6.36
6.32
7.43
8.7
17.22 5.75
1707] 6.67
18
17.48
205
21,05
21,19
14 45
16,07
18.42
15.59
12.39
41,16
6,5
6,68
6.66
6.79
s. ell
5.43
5.34
6.45
6.21
5,18
5.23
5.57
6.6
11/26/01
19.79
19.68
16.3
18.3
LEVEL
6,36
4.61
4.29
4,81
14,66 3.86
18.1
17.96
1895
18 38
21 46
22
22
1525
169
19,25
16.43
13.2
42.22
5.38
3
ill
4.78
4.93
4.66
4.83
4,7
4.62
4.54
2/3/01
15,47
16,61
13.74
.• Y
12/3/01
LEVEL
12/10/01
•
|
i
I
9,66
7.48
6,85
14.9 8.21
12.56] S.97
142
N4.3
L83
1466
16 97
18 01
1878
13,05
4.62J 14,71
4.38
4.34
4.42
4.51
5.38
16,74
14 56
8.77
8,44
9,05
8.35
446
8.76
6.17-
582
5.04
4 19
4.71
4 54
511
9.32j 619
8.69I 8,18
7,84
10,03
6,74| 625
6.81
6.89
6.21
11.691 6.93
368) 9.93
11,5) 11.5
_ .- .. —
!
L ..
8,05
9,56
8,63
6,47
2469
246
":j"~ _:jv:.-[_r -.L.. _
-i
!
I
T
J_
.. ..J .....
f I
J
~ -L— .-
,
[
r .. . ...
— ±. .. _£ — T.... . :
i
L.
: — r :T ::v
.
r
" ' " I I [
.... . . T_ ..
t r"" XI " L i ""
J
4-—. -- .
|._ ...
j*.
-------
-1
2
3
4
5
6
7
8 .
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
4!
46
47
48
49
50
51
52
53
.EVEL
20.67
15.53
14.42
17.29
13.49
18,78
18.03
19.34
17,92
20.1
18.52
16.59
13.54
13,47
14.07
12.14
11.16
22.04
24.6
2.85
6.26
3.24
3.41
1.78
2.1
2.49
2.5
2.86
4.63
6.24
7.76
3.4
5.25
6.99
8,74
3.43
23.73
25.5
.EVEL
22.3
18.03
17.35
19.7
16.75
20.87
20.26
21.38
20.15
21.66
20.46
18.86
16.39
16.27
16.64
12.03
14.19
23
25.6
TOC
3.32
0.69
TOC
TOC
TOC
TOC
TOC
TOC
0.86
2.69
4.57
0.75
2.6
4.14
3.23
1.09
19.45
30
.EVEL
20,97
19.90
25.43
24.01
22.06
19.04
16.92
19.49
17.64
16.53
27.28
30.00
12/31/01
TOC
1.06
TOC
TOC
TOC
TOC
TOC
TOC
'OC
TOC
0.57
2.37
TOC
TOC
1.91
0.97
TOC
17.7
31.9
12/31/01
LEVEL
23.23
26.13
24.26
21.72
19.8
29.03
31.9
1/7/02
0,87
3,09
0.23
0.67
TOC
•QC
TOC
TOC
0.21
2.45
3.73
4.77
TOC
1.63
3.49
1.93
TOC
22.27
27.5
1/7/02
.EVEL
24.28
21,2
20,36
22.44
22.62
23.84
22.97
21.85
19.89
20.14
16.84
24.46
27.5
1/14/02
"7,62
7,63
4.25
5.27
3.8S
5.39
5.32
6.05
5.72
8.98
9.64
9,66
3.5
5.25
7,54
5.04
2
29.8
19.07
1/14W2
.EVEL
17.8;
16,6i
16.3.
17;84
17.68
17.4;
17.8;
17.3'
17.3'
17.0
16.7
16.2
16.2'
16.01
16.7
16.6
16.9
19,0'
i/21/02
13,02
12,23
6.61
10.9
6.8
10.56
10,4
11,37
10.68
14.4
14.87
14.71
7.72
9.4
11.85
8.92
5.64
35.51
13,04
'
1/21/02
.EVEL
12.13
12.06
11.98
12.21
11.73
12.41
12.34
12.61
12.36
11.89
11.83
11.91
12,07
12.12
11.78
11.86
11.98
11.22
13.04
1/28/02
12.03
12.28
8.94
10.65
7,39
10.05
10.08
10.79
10.36
13.4
14,17
14.53
8.16
9,85
12.1
9.57
6,45
33.89
14,82
1/28/02
.EVEL
13.12
12.01
11.66
12.46
11.14
12.92
12,66
13.09
12.67
12.89
12.63
12.09
11.63
11.67
11.63
11,2
11.17
12.84
14.82
2/4/02
3.14
6,35
3.78
3.83
2.99
2.39
2,9
2.89
3.31
4.76
6.3
7.9
3.53
6.25
6.94
5.63
3.21
23.66
25,7
.EVEL
22.01
17.94
16.81
19.28
16.64
20.68-
19,84
20.99
19,72
21.53
20,4
18,72
16.26
16.27
16.69
16.14
14,41
23.07
25.7
roc ' ~
2.23
roc
TOC
TOC
roc
roc
roc
roc
0.23
1.88
3.57
roc
1,33
2,92
1.84
TOC
19.08
30.4
.EVEL
22.06
26.06
24.82
23.06
20,19
20.71
18.93
27.65
30.4
roc
3.26
roc
0.95
TOC
roc
TOC
0.29
roc
2.82
4.07
5.03
TOC
1.73
3.67
1,96
TOC
22.69
26.5
—"AT
2/18/02
.EVEL
21.01
22,16
23.69
23.47
22.63
21,59
19,79
- 19.96
18.81
24.04
26.5
AU
5.6
6.55
3.34
4.33
1.92
3.78
3.39
4.43
4.23
7.05
8
8.57
2.77
4.52
6.61
4.46
1,63
27.34
21.6
-------
2
.a
s
6
7
a
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
2'
25
26
27
2!
"2S
30
31
32
3C
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
60
51
52
53
EVEL
19.5S
17.74
17.25
| 18.78
16.61
19.19
18.B5
19.4S
18.8
19.24
18.7
16.05
17.02
17
17.02
16.31
15.99
19.39
21.6
9.6
9.41
5.89
7.81
4.52
7.24
7.23
8
7.53
10.96
11.63
11.69
5.13
6.85
9.19
-6.54
•3.41
31.85
16.8
.EVEL
15.SS
14.88
14.7
15.3
14.01
15,73
• 15,51
15.88
15.5
15.33
15.07
14.93
14.66
14.67
14.44
14.23
14.21
14.88
16.8
3/11/02
11.54
11.31
7.81
8.21
6.16
5.33
9.22
10.02
9.49
12.89
13.51
13.62
7.04
8.73
11.07
8.38
5.23
33.67
15.01
3/11/02
.EVEL
13.61
12.98
12.78
14.9
12.37
R&MEfiSSSI
-------
2
3
4
5
b
7
a
9
10
11
12
13
14
15
16
17
18
19
20
21
22
24
25
26
27
28
29
30
31
32
33
34
"3i
36
37
38
39
40
41
42
43
44
45
46
47
48
49
5!
51
52
5:
6VEL
20.007
23,714
22.208
23
26.04
26
23.74
21.92
27.16
29.9
— 3W
roc
TOC
roc
4.62
TOC
0.26
TOC
TOC
TOC
TOC . •
TOC
TOC
TOC
TOC
0.49
1.74
TOC
TOC
0,71
TOC
31.5
.•
23.30
24.03
26.21
24.88
22.92
28,70
31.5
TOC
TOC
TOC
3
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC .
TOC
TOC
34.07
-2m
TOP.
TOF
-ff.7'9
0.-32
TOF
TOF"
TO'F
TO'F
TOF
TOF •'
24.92
31.20
34.07
18.69
19.73
TOC ' "
TOC
•JA
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
36.65
0.63
TOP
33.62
36.66
TOC
TOC
TOC
0.98
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
. , 13.5
19.46|TOC
20.66 £'11
26.73 TOF
26.65 TQF
TOF
.0.26
TOF
TQF
TO'F
.TOF.
iiTOF
22.41 T.OF
TPF'
-TOF
TOF
TOF
T'&F
TOF,
TOF
TOF
TOF
TOF
26.94
33,32
8HH
21.69
TOC
TOC
TOC
3.62
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
TOC
0.71
TOC
TOC
TOC
TOC
31.1
0.12
2.19
TOC
0.86
TOC
0.17
TOC
TOC
TOC
24,30
26.91
28.41
31.1
19.97
20.91
26.98
25.76
T:0'F
T:QF
ifbc
ITOC
TOC
•/—cc — r
2.79
4.13
1.8
7.77
2.12
'" 4" 17
1.3
1.99
TOC
0.89
1.18
1.6
1.47
3.62
4.75
5.78
0.87
2.46
4.16
TOC
26.25
2.09
4,73
2,63
4.93
5.67
5.1
3.69
3.52
4.03
TOC
1.42
2.41
0.17
0.79
4.15
CD 1
17.81
20.89
19.39
20.16
23.03
20.12
19.29
21.12
22.01
21.66
22.36
21.66
22.67
21,95
20.84
18,92
19.06
19.47
23,80
26,25
18.00
16.37
21.20
21.93
20.86
20.65
19.51
19.68
21.44
19.32
21.08
3
6.89
' 3.35
10.31
5.8
5.92
2.43
4.28
" ' 0.65
3.68
3.65
4.39
3.89
7.25
7.93
8.1
1.74
3.49
5.76
TOC
>1.'i7
6.02
9,35
: "8,'S7
'Si*?
"611B
:: 3ll9
17.40
16,13
17.84
17,61
19.36
18,37
18.16
18.83
17.88
19.29
19.09
19.49
19.14
19,04
18.77
18.62
18.06
18.03
17.87
18.97
| 2117
16.66
17,00
18,01
18,22
18.65
18.56
17.87
17.83
18.77
19,38
18.16
18.80
8.22
13.19
6,59
15.99
13.5
11.72
7,87
10.7
5.9
10.59
10.25
11,45
10.45
14.69
14.9
14.33
6.81
8.6
11,15
4.48
: 3S;S7
12.8
••• 7;?
• ,12168
15^-9
: 11'&1
: ISJ
• 'Jin
21.17:. 'M't 18.97: j/BSSS
17.64| 1
17.83- ,5,05
I 17.33
17.62
"; ;'~t$ii
12.18
11.83
12.60
11.93
11. «
12.67
12,72
12.41
12.63
12.38
12.49
12.43
12.58
11,60
11.80
12,29
12.98
13.02
12.48
13.14
10.96
12.80
13.67
13.70
12,03
11.46
TifsT
12.33
12.51
12,47
12.32
11.96
12.74
12,39
12,19
13.1S
' 1124
10.32
15.72
11.24
18.95
15.75
14.32
10.65
13.23
8,25
13,14
12.85
13.97
13.02
16.95
17.32
16.94
9.62
11.3
13,65
7.32
'. 37,98
105
:• 1277
14.52
1TM9
II 7.78
17 SB
";: rak*
iM
1 Mas
-its
"• as?
10.06
9.30
9.9S
8.97
9,40
9.97
9.94
9.88
10.28
9.83
9.89
9.91
10.01
9.34
9.38
9,68
10.17
10,22
9,78
10,30
8.84
11,12
11.03
9.51
9.08
9.74
12.13
17.4
13.01
20.22
17.5
16.15
12.42
15.07
10,17
14.98
14.65
15.79
14.81
18,7
19,04
18.27
11.4
13.08
15.61
9.13
13.93
16.22
T27
7.621
8,18
7.70
7.65
8,14
8.17
8.04
8.36
7.99
' 8.09
8,0)
8,22
7.69
7.66
6,39
8,44
8.02
6.49
7.32
8.6
9,30
9,17
7.61
7,45
«54 7.98
isl'SB- 7.99
9.81 1&69 8.03
9.76
9.79
' •ill
8.01
8.01
9.50 ' . IS341. 7.73
9.99 1:3v2S 8.20
9.62
T73
10.22
10.30
:mS2
13.43
8.02
7.91
8.44
8.55
CM
17.8
16.7
12.94
15.52
10.57
15.37
15.04
16.16
15.23
19,06
19.48
19.25
12.1
13.78
16,29
9.93
39'.7B
8.31
leiss
2E>,0'6
i6.68
-I.K09
I!?'!
16.'5
iV.42:
-------
2
3
4
s
£
1
6
9
10
11
12
13
14
15
16
17
13
19
'2C
21
22
23
2'
25
26
27
28
29
30
3-
32
33
3'
35
36
.37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
5j
s:
7/28/02
7.36
7.S9
7,65
7.59
7,96
7.60
7.70
7.72
7.80
7.23
7.22
7.37
7.6i
7.74
7.34
7.52
7.69
7.04
8.31
8.34
8.27
7.18
6.94
7.44
7.44
7.36
7.S3
7.60
7.36
7.66
7.S4
7.59
7.62
' 7.79
C6^
13.81
18.S5
14.6
21.46
18.4
17.51
13.96
16.36
11.75
16.2
15.94
17
16.12
19.64
20.12
20
13.04
14.68
17.11
14.17
11.02
40;2it
7.8
•13;3B
15.78
17.52
ZH17
alias
So'zs
ifesa'
'•'?I3!&
$39-
• ism
~T4i8
.•#.44
.1:5:42
11.7
15.22
6.59
6.47
6,59
6.46
6.75
6.78
6.63
6.73
6,78
6.77
6.80
6.88
6.91
6.65
6.58
6.62
6.75
6.84
6.52
6.60
6.60
6.61
7.80
7.41
7.32
6.51
6.40
6,67
6.68
6.54
6.61
6.73
6.66
6.65
6.70
6.66
6.63
6.76
14.2
18.92
14.97
21.95
18.65
17.88
14.3
16.61
12.22
16.43
16.18
17.25
16.32
19.91
20.42
20.35
13.43
15.1
17.51
14.59
11.38
40:55-
7.38
. 1'4
•17.05:
17.89
21.S
Ziitf
4
&&
4T?4
tm
i&3?5
•M;£
15592
.4^*8
-2tf98
22.35
2r:«
2S.«7
1'S.47
i* — i
9/3/02
3.44
4.33
3.76
4.02
6,23
' 3.96
3.46
4.33
3.51
3.70
4.49
4.86
4.46
5.04
4.71
4,16
3.56
3.74
3.71
3.37
3.25
6.63
6.7
4.03
3.90
4.47
4.82
4.29
3.86
3.72
3,67
4,47
6.10
3,46
4,29
4.18
3.12
3.22
4.46
I 4.35
3.94
4.46
3.48
' CY
. 9/9/02
15.24
19.11
16.62
22.39
. 18.35
18.4
15.04
16.92
13.25
16.5
16.4
17.24
16.78
19.67
20.34
20.6
14.2
15.86
18.12
15.56
12.6
3996-
8.06
! 1465
- 1*5*
18.3
2152
21 fe
: 20.8B
: 1tf£f?
' 11*
: ^
". -&?
• $53»
! 1788
: feo's
is**
i iTW
20-36
] '20.84
?946
23:8
IS S3
cz
9/9/02
6.16
6.91
6.67
5.63
6.SO
6.89
5.66
6.19
6.28
6.47
6.34
6.64
6.25
6,62
6.36
6.02
6.61
5.66
5.61
5.21
5.02
6.86
8.06
5.12
5.22
6.73
6.06
6.10
6.09
5.66
5.61
4.31
6.65
6.63
6.16
6.03
6.10
6.27
6.07
| 6.86
5.63
5.92
6.27
9/16/02
16.15
16.82
'. ' 20.7
24.11
20.53
19.76
16.26
18.5
14.29
18.32
18.1
19.13
18.43
21.72
22.23
22,19
15,29
16.94
19.36
16.47
13.28
4? 19
5,45
is
1*06
2ft?7
23522''
'23.
22M7
18,85
i i^a
! **»
i *m
i?®
«'«
17 ft
14!73
-------
-------
14-SS
1707
1382
1052
4058
783
1367
181,9
51-1-2
2Q55
1998
1727
IT 19
19 tS
19.00TOC
12.S3 1<4
16.72 1 OS
20.06JXI3C;
21.19 iTtJe1
1448
1717
1513
1203
15,54
1954
2086
tags'
2307
20.22
25.28
26.69
23,67
16,56
12.7-2
1*54
18.21
20.40
20.59
22.02
18.26
-------
-------
1
2
3
4
T"
6
7
6
9
10
11
12
13
14
15
16
17
18
19
2C
21
22
23
24
25
26
27
IT
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
GC
7/7703
7.4
6.98
3.9
5.35
1.36
4.85
4.68
5.53
4.98
i .m
! :
7/7/03
17.7S
17.31
16.69
17.76
17,17
18.12
18.06
18.35
18.06
17.60
17.18
17.00
17.05
16.97
18.86
16.54
16.52
16.91
19.18
15.18
16.38
1S.4S
17.21
17.21
16.87
16.63
17.48
18.07
17.32
17.89
16.60
16.78
16.85
16,60
16.73
16.47
16.41
17,36
17.35
17.82
16.20
16.24
18.35
16.32
-------
APPENDIX: B
Combined Quality Assurance
Project Plan and General Work Plan
-------
COMBINED QUALITY ASSURANCE PROJECT
PLAN AND GENERAL WORK PLAN
POTENTIAL GROUND-WATER FLOW
DIRECTIONS AND CONTAMINANT FATE AND
TRANSPORT IN THE PLAQUEMINE AQUIFER
OF IBERVILLE PARISH AND WEST BATON
ROUGE PARISH, LOUISIANA
August 15, 2003
Prepared By:
Scott Ellinger
Multimedia Planning and Permitting Division
EPA Region 6
Telephone No. (214) 665-8408
-------
GROUP A: PROJECT MANAGEMENT
Al TITLE AND APPROVAL SHEET
TITLE: POTENTIAL GROUND-WATER FLOW DIRECTIONS AND
CONTAMINANT FATE AND TRANSPORT IN THE PLAQUEMINE
AQUIFER, IBERVILLE PARISH AND WEST BATON ROUGE PARISH,
LOUISIANA
PREPARED BY:
Name: Scott Ellinger
Title: Geologist, Arkansas/Louisiana Section
Signature: ^Lr^fST' 7/*^X>? Date: tr//r"«fl3
-t ^
/r"/«fl
PEER REVIEWED BY:
Name: Michael Bechdol
Title: Environmental Scientist, Ground-Water/UIC Section
Signature: /^Mx^^»-^ .]^"~w y Date:
-------
APPROVED BY CEP A):
Name: Ben Banipal
Title: Chie£ Arkansas/Louisiana Section
& M
Signature: >** < M Date
APPROVED BY(LDEQ):
Name: Tim BCnight
Title: Administrator, Environmental Technology Division
.^•f^-rfc/7
Date:
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 4 of50
A2 TABLE OF CONTENTS
Section Page
GROUP A: PROJECT MANAGEMENT 2
Al TITLE AND APPROVAL SHEET .... 2
A2 TABLE OF CONTENTS 4
A 2.1 LIST OF FIGURES 5
A.2.2 LISTOFTABLES 5
A3 DISTRIBUTION LIST 6>
A4 PROJECT AND TASK ORGANIZATION 8
A5 DEFINITION AND BACKGROUND OF PROBLEM 11
A6 PROJECT/TASK DESCRIPTION AND SCHEDULE 15
A7 QUALITY OBJECTIVES AND CRITERIA FOR MODEL INPUT/OUTPUT 18
A.7.1 MODEL DEVELOPMENT AND QUALITY REVIEW CRITERIA 18
A.7. La. MODELING OBJECTIVES AND DATA REQUIREMENTS 19
A.7.1.b CONCEPTUAL MODEL DEVELOPMENT 20
A.7.I.C. FIGURES AND TABLES 21
A.7. Ld REVIEW CONSIDERATIONS FOR CONCEPTUAL MODEL
FORMULATION 22
AJ.l.e. MODEL (CODE) SELECTION 22
A.7. l.f. MODEL CONSTRUCTION AND CALIBRATION 23
A.7.1.g. SENSmVITY/UNCERTAINTY ANALYSIS 27
A8 SPECIAL TRAINING REQUIREMENTS/CERTinCATION 28
A9 DOCUMENTATION AND RECORDS 29
GROUP B: MEASUREMENT AND DATA ACQUISITION 30
B7 MODEL CALIBRATION 30
B9 NON-DIRECT MEASUREMENTS (DATA ACQUISITION REQUIREMENTS) 33
BIO DATA MANAGEMENT AND HARDWARE/SOFTWARE CONFIGURATION 35
GROUP C: ASSESSMENT AND OVERSIGHT 38
Cl ASSESSMENT AND RESPONSE ACTIONS 38
C2 REPORTING TO MANAGEMENT 40
GROUP D DATA VALIDATION AND USABILITY 41
Dl DEPARTURES FROM VALIDATION CRITERIA 41
D2 VALIDATION METHODS 41
D3 RECONCILIATION WITH USER REQUIREMENTS 41
REFERENCES 49
-------
A 2.1 LIST OF FIGURES
Figure 1. Project Organization
Figure 2. General Location Map
Figure 3. Aerial Photograph of Study Area
A.2.2 LIST OF TABLES
Table I. Project Schedule
Table 2. Major Steps in Modeling Evaluation Procedures
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15, 2003
Page 5 of SO
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. IS, 2003
Page 6 of 50
A3 DISTRIBUTION LIST
After final approval of this QAPP and General Work Plan, the Project Manager will transmit hard
copies via U.S. Mail to the organizations and individuals listed below.
U.S. Environmental Protection Agency, Region 6
Donald Johnson
Regional Quality Assurance Manager
Management Division (6MD)
Charles Ritchey
Division Quality Assurance Officer
Multimedia Planning and Permitting Division (6PD)
Mark Potts, Chief
Hazardous Waste Enforcement Branch
Compliance Assurance and Enforcement Division (6EN-H)
Philip Dellinger, Chief
Ground- Water/UIC Section
Water Quality Protection Division (6WQ-SG)
Olivia Balandran, Team Leader
Office of Environmental Justice (6RA-D)
U.S. Environmental Protection Agency-National Risk Management Research Laboratory
Steven Acree, Hydrologist
Ground-Water and Ecosystem Restoration Division
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. IS, 2003
Page 7 of SO
U.S. Department of Health and Human Services
Danielle Langmann, Environmental Health Scientist
Agency for Toxic Substances and Disease Registry
Louisiana Department of Environmental Quality
Steve Chustz, Senior Scientist
Office of Environmental Assessment
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15, 2003
Page 8 of 50
A4 PROJECT AND TASK ORGANIZATION
AH environmental monitoring and measurement efforts mandated or supported by the U.S.
Environmental Protection Agency (EPA) are subject to a centrally managed quality assurance (QA)
system. The EPA Quality System defined in EPA Order 5360.1 A2 (EPA, 2000d), Policy and Program
Requirements for the Mandatory Agency-wide Quality System, includes coverage of environmental data
produced from models. Environmental data includes any measurement or information that describe
environmental processes, location, or conditions; ecological or health effects and consequences; or the
performance of environmental technology. For EPA, environmental data includes information collected
directly from measurements, produced from models, and compiled from other sources such as databases
or literature. The EPA Quality System is based on an American National Standard, ANSI/ASQC E4-
1994.
Consistent with the National Standard, E4-1994, Section §6.a.(7) of EPA Order 5360.1 A2 states
that EPA organizations will develop a Quality System that includes approved Quality Assurance Project
Plans (QAPP), or equivalent documents defined by the Quality Management Plan, for all applicable
projects and tasks involving environmental data with review and approval having been made by the EPA
QA Manager (or authorized representative defined in the Quality Management Plan). More information
on EPA's policies for QA Project Plans is provided in Chapter 5 of the EPA Manual 5360 Al (EPA,
2000a), EPA Quality Manual for Environmental Programs and Requirements for Quality Assurance
Project Plans (QA/R-5) (EPA, 2001). This guidance helps to implement the policies for models defined
in Order 5360.1 A2.
Any party that generates data under the QA program is responsible for implementing minimum
procedures to ensure that the precision, accuracy, completeness, sensitivity, comparability, and
representativeness of its data are known and documented. Each party must prepare a Quality Assurance
Project Plan (QAPP) for each environmental data collection effort. In response to this requirement, the
EPA Project Manager has prepared this QAPP which presents the overall project description, project
organization and responsibilities, and QA objectives associated with the ground-water flow modeling to
be conducted. This project-specific QAPP complies with all relevant elements of "EPA Guidance for
Quality Assurance Project Plans for Modeling" (EPA QA/G-5M) (December 2002) and has undergone
peer-review.
To complete this modeling project, EPA Region 6 Multimedia Planning and Permitting Division,
will develop a ground-water model to represent a portion of the Plaquemine Aquifer, in Iberville Parish
and West Baton Rouge Parish, Louisiana. Simulating subsurface phenomena, such as ground-water flow
and contaminant fate and transport, 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. To
facilitate major aspects of model development and model peer reviews, the Multimedia Planning and
Permitting Division will utilize a multi-organizational team approach. The team will include technical
experts from the EPA Multimedia Planning and Permitting Division, the EPA National Risk Management
Research Laboratory (NRMRL) Robert S. Kerr Environmental Research Center in Ada, Oklahoma, the
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15, 2003
Page 9 of 50
State of Louisiana - Louisiana Department of Environmental Quality (LDEQ), and limited contractual
support involving field data collection, literature research, and software support.
Figure-1 shows the project organization. Overall project supervision lies with the Management
of the Multimedia Planning and Permitting Division. Management of the Multimedia Planning and
Permitting Division provides direction to technical staff which is responsible for developing the
conceptual model and the numerical model. The Region 6 Quality Assurance Manager and Division
Quality Assurance Officer provide guidance and support during QAPP development and processes for
model peer reviews to ensure that the Agency's Quality Assurance requirements are being met. Two
Federal agencies will conduct model reviews-the EPA NRMRL and the U.S. Geological Survey. These
two organizations will serve to provide technical assistance on the conceptual model, model set up, model
calibration, sensitivity analysis, and interpreting modeling results. The Louisiana Department of
Environmental Quality will review the model mainly to determine whether the State's modeling goals are
being/have been achieved. Certain sets of technical data being used in the model have been generated
during a recent agency Contractor Work Assignment to enable Region 6 to provide assistance to the
Louisiana Department of Environmental Quality. The Work Assignment Manager maintains specific data
files and coordinates data needs with the other modeling team members. The EPA Region 6 Library
assists by conducting literature searches and ordering and obtaining supplementary information.
The overall purpose for this model is to evaluate and simulate potential ground-water flow
directions and contaminant fate and transport 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). In the case that insufficient data is available to support contaminant fate and
transport, then only the ground-water flow sections of the QAPP will be in effect and considered as part
of the modeling effort. Fate and transport modeling may have to be conducted at a later date under a
modified or separate QAPP. The reader is referred to the Multimedia Planning and Permitting Division
for any additional issues related to the QAPP and General Work Plan, and to the members of the
modeling team which are available to discuss specific model input data. The official, approved Quality
Assurance Project Plan and General Work Plan will be maintained in the files of the Mulitmedia Planning
and Permitting Division—Model Development Project Manager. The Technical Background
Document/Conceptual Model contains detailed model input data and will be attached to the official,
approved QAPP and General Work Plan upon completion. During the course of this project certain
conditions, processes, and procedures inherent to modeling may change. If such changes cause any
significant changes to the QAPP, the Region 6 Quality Assurance Manager will be notified.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 10 of 50
PROJECT ORGANIZATION
OfcnllPnjttlSipmsioi
SKilta Cilnii, Arociafc Dir«!« hi RCRA
Bti Bail A, Ciirf, Ctmdirc Aelui/Wadt Miiiiiiidti Sttlia
, MiHiouluPlaiiii(ailP(»ittii{Diriaii
Rqim ( Qulilj Assirute Maijir
Daild IAKH
MaijemnlDirisiot
Ui((IDnda«
< KIMIl(Hi
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 11 of 50
A5 DEFINITION AND BACKGROUND OF PROBLEM
Figures 2 and 3 depict the general study area. 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 the March 2001 sampling event.
LDEQ has been conducting a phased ground-water investigation since April 2001. The objective
of LDEQ's investigation has been to identify the source of vinyl chloride contamination, and to delineate
the extent of the plume and ensure protection of human health and the environment. The events listed
below are the main elements of LDEQ's investigation to date.
Neighborhood/local business survey of water wells, review area water uses
Review potential contamination sources
Review of DOW Chemical monitoring data
Research reductive dehalogentation of chlorinated solvents/develop sampling strategy
Phase 1 sampling, April 2001, 11 water wells
Phase 2 sampling, May 2001, split samples with DOW Chemical on 9 boreholes and
confirmed Phase 1 results
Phase 3 sampling (June 2001) sampled 21 Wells
Received assistance from EPA-NRMRL lab in Ada, Oklahoma
Received EPA-NRMRL on July 26, 2001 detailing response to State questions on plume
chemistry and ground-water flow regime
Phase 4 (June 2001), split sampling with DOW Chemical on 6 boreholes
Phase 5 (July 2001) testing of DOW Chemical fire water wells screened in Plaquemine
aquifer and split samples on 3 additional borings installed by DOW Chemical
Phase 6 (August 2001) sampling Shintech wells and 7 private wells
Phase 7 (September 2001) DOW Chemical 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 Chemical piezometers)
LDEQ oversight of installation of additional piezometers by DOW Chemical and split
samples on additional borehole installed by DOW Chemical
LDEQ sampled additional well installed by EPA and two additional sentinel wells
installed by the City of Plaquemine
Collection of water level data weekly, Establishment and sampling of a long term
monitoring network.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 12 of 50
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 13 of 50
Figure-3. Aerial photograph of study area.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 14 of 50
In February 2003, the Multimedia Planning and Permitting Division began to plan ground-water
modeling activities for the Plaquemine aquifer. During the planning phase the following modeling
objectives for this project were determined.
I. Determine potential directions of ground-water flow in the Plaquemine aquifer over the project
area.
2. Understand how ground-water flow is affected by aquifer interaction with the Mississippi River,
pumping wells, and by possible regional ground-water flow gradients.
3. To the degree possible with flow modeling, evaluate the likelihood about possible contaminant
source locations, and whether multiple source locations are possible.
4. Estimate age and location of contaminant release(s) to degree possible with flow modeling, and
plume movement with flow and/or contaminant fate and transport modeling, if possible.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 15 of 50
A6 PROJECT/TASK DESCRIPTION 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). ASTM is a private organization that publishes consensus standards for a variety
of fields, including ground-water modeling. The ASTM Subcommittee D18.21 on ground-water and
Vadose Zone Investigations has five standards related to ground-water flow modeling. These standards
have been written in the form of guides (not rigid standards) and include the following publications:
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-56U Standard Guide for Conducting a Sensitivity Analysis for a dround-Water Flow Model
Application; and
D-5718 Standard Guide for Documenting a Ground-Water Flow Model Application
The application of a ground-water model ideally would follow several basic steps to achieve an
acceptable representation of the hydrogeologic system and to document modeling results for the end-user,
decision-maker, or regulator. These steps include:
Identify and define modeling objectives;
Develop Project Work Plan;
Develop Quality Assurance Project Plan;
Collect, organize, and interpret available information and data;
Prepare a conceptual model;
Set up numerical model;
Calibrate model;
Validate model;
Run model for flow simulations;
Run model for fate and transport simulations;
Perform post-simulation analysis;
Evaluate overall modeling effectiveness;
Determine whether objectives have been/are being met;
Determine preliminary results and level of accuracy and error;
Reiterate modeling steps as necessary to improve quality and accuracy; and
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. IS, 2003
Page 16 of 50
Final results and report preparation.
The final modeling report for this project is likely to include written and graphical representations
of model assumptions and objectives, the conceptual model, code description, model construction, model
calibration, predictive simulations, sensitivity analysis, and conclusions. The following list is
representative of the Table of Contents typical of many modeling reports.
Title page
List of tables
Executive summary
Introduction
Model objectives
Hydrogeologic characterization
Conceptual model
Code evaluation
Input parameters and model framework
Model calibration
Sensitivity analysis
Simulations performed
Conclusions and recommendations
References
Tables
Figures
Well data
Additional data
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15, 2003
Page 17 of SO
Project Schedule
(
^
c
c
Activity-Assessment and Planning Phase
Establish Modeling Objectives; Literature Search; Evaluate Data Completeness
Develop Test/Interim Model
Develop Draft Combined QAPP/General Work Plan
Peer Review QAPP and General Work Plan
Prepare Draft Conceptual Model (Including Specific Model Input)
Obtain QAPP and General Work Plan Signature Approval
Peer Review Conceptual Model (Including Specific Model Input)
Activity-Implementation and Evaluation Phase
Set Up Numerical Model
Brief EPA and LDEQ Management
LDEQ and EPA Public Outreach (on plans, procedures, conceptual model, etc)
Perform Modflow and Modpath Simulations
Perform Mass Transport/Reactive Transport Simulations
Assess Overall Modeling Effectiveness; Make Necessary Modifications
Verify all QA/QC Requirements Met
Brief EPA and LDEQ Management on Modeling Results
LDEQ and EPA Public Outreach on Modeling Results
Prepare Final Modeling Report
Month
Feb.
July
July
July
July
Aug.
Aug.
Month
Aug.
Aug.
Aug.
Sept.
Sept.
Sept
Sept.
Oct.
Oct.
Oct.
Schedule Is approximate and contingent upon fulfilling the data quality requirements of this QAPP.
Schedule may be modified or extended as necessary to reflect time requirements for any additional data
•ollection or analysis activities. Prior draft schedule indicated August 15 as final report date; due to
•hanges in project scope, the current projected report date has been reflected as given above.)
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15, 2003
Page 18 of 50
A7 QUALITY OBJECTIVES AND CRITERIA FOR MODEL INPUT/OUTPUT
A.7.1 MODEL DEVELOPMENT AND QUALITY REVIEW CRITERIA
The EPA data quality objective (DQO) process is a systematic planning tool designed to ensure
that the measurement data collected are of the type, quantity, and quality that are the most appropriate for
supporting the decisions that will be based on these data (EPA I999a; I999b). Data quality depends on
the intended use of the data and decisions that are to be based on the data. For projects that require data
collection or environmental data produced from models, EPA's DQO process will be followed (EPA
1993; 1999a; 1999b, 2002). Environmental data includes any measurement or information that describe
environmental processes, location, or conditions; ecological or health effects and consequences; or the
performance of environmental technology. For EPA, environmental data includes information collected
directly from measurements, produced from models, and compiled from other sources such as databases
or literature. The DQOs of this project are to:
- Provide quality, reliable data enabling Modflow, Modpath, and MT3D/RT3D computer
codes to be employed to determine generalized net ground-water flow directions and
contaminant fate and transport in the Plaquemine Aquifer that can be used as follows:
Assisting with making observations and conclusions about short-term flow, long-
term flow, and net flow within the flow system for the Plaquemine aquifer,
Help evaluate the effects of pumping wells (public, private, industrial, etc.) on
ground-water flow;
To the degree possible with flow modeling, evaluate the likelihood about
possible contaminant source locations, and whether multiple source locations are
possible.
To meet the DQO's stated above, systematic modeling guidelines for meeting data quality will be
followed when acquiring, generating, and handling data to develop the Plaquemine Aquifer ground-water
model. These guidelines are summarized from existing EPA guidance. (See "Documenting Ground-
Water Flow Modeling at Sites Contaminated with Radioactive Substances", EPA-540-R-96-003, U.S.
EPA, Office of Air and Radiation (OAR), January 1996). This guidance describes the activities and
thought process that should be a part of a model application, documentation, and review of ground-water
modeling. The contents within this guidance may be applied to a wide variety of ground-water modeling
projects, including the Plaquemine Aquifer model. Not all elements of the guidance are strictly
applicable to this modeling project. The guidance is divided into a series of elements which are typical of
most modeling exercises. These elements are described here in general terms. Specific model input data
is included with the Project Work Plan.
I. Modeling Application Objectives (Section A.7.l.a)
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 19 of 50
2. Conceptual Model (Section A.7. l.b)
3. Figures and Tables ( Section A.7. l.c)
4. Review Considerations for Conceptual Model Formulation (Section A.7.1 .d)
5. Model (code) Selection (Section A.7. l.e)
6. Model Construction and Calibration (Section A.7. Lf)
7. Sensitivity/Uncertainty Analysis (Section A.7.1 .g)
A.7. La. MODELING OBJECTIVES AND DATA REQUIREMENTS
The objectives of a modeling study should be clearly specified up front, considering applicable
regulatory and policy issues. AH assumptions incorporated within the modeling objectives should be
reviewed with respect to reality and their potential impacts on project objectives. The level of model
complexity and, in turn, the type of model required (e.g., numerical model, analytical model, or graphical
techniques) should be documented as part of the modeling objectives.
The definition of modeling objectives is important. It is necessary to give the peer reviewer a
clear understanding about what the model results will be used for and how these results fit into the
development of the model.
• The purpose and scope of the modeling exercise should be clearly
indicated.
• The modeling objectives should be identified in this section. In the
summary and conclusions of the report, each of these objectives should
be discussed separately in the context of how the modeling was used to
meet the objective and the degree to which the objective was met.
• The data required to construct a conceptual model should be described
and the relevance of the data to ground-water flow and fate and transport
should be discussed.
• The source of the data should be presented. That is, which data will be
or was collected in the field, versus taken from the literature and/or
model calibration.
• The uncertainties associated with the data should be discussed. Are
some field collection methods better than others? How reliable are
literature values? A probable range in which the parameters will fall
should be assigned, prior to the modeling analysis.
• The general sensitivity of the data to the determination of ground-water
flow and fate and transport should be discussed. This discussion should
enable the field characterization program to be more focused.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 20 of 50
Limitations and weaknesses in the data base should be presented as well
as plans to enhance the data base.
Recommendations should be presented, detailing additional data needed
to increase confidence in the modeling results.
A.7.l.b CONCEPTUAL MODEL DEVELOPMENT
Prior to documenting the type of model used and how it was constructed, the report should
contain a thorough discussion of the conceptual model that is the foundation of the mathematical model.
The conceptual model does not necessarily need to restate all of the information known about the region
being modeled. Rather, the conceptual model should be described in terms of the assumptions made to
simplify the system. The conceptual model should also list data gaps and their impact on the modeling
results. Typical information that should be provided with respect to the conceptual model includes the
following:
• The hydrogeologic system should be described in detail, including
lithologic contacts, facies changes, discrete features, and spatial
variations of geologic units and their hydraulic properties. The rationale
for the variability of the properties should be explained (e.g.,
depositional history);
• The boundaries of the system should be described in a water budget
analysis (evapotranspiration, runoff, pumping and recharge rates). The
methodology for determining individual components of the water budget
should also be included;
• The geometry of the system should be presented in three dimensions
with a rationale for possible simplification. For example, the analysis of
the unsaturated zone may be reduced to two dimensions;
• The rationale for any simplifications (e.g., steady state) made to the
conceptual model should be presented;
Uncertainties in the conceptual model should be presented and related to
earlier discussions of data limitations and uncertainties.
• The contaminant source term should be described.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 21 of 50
A.7.I.C. FIGURES AND TABLES
The following are illustrations and tables that should be included to describe the conceptual
model. Some figures may not be required; however, justification for conceptual model assumptions
should be given if figures and tables are omitted.
• Map showing location of study area.
• Maps and cross sections showing the thickness of the unsaturated zone.
• Geologic map and cross sections indicating the areal and vertical extent
of the system.
• Topographic map indicating surface water bodies.
• Contour maps showing the tops and/or bottoms of the aquifers and
confining units.
• Isopach maps of hydrostratigraphic units.
• Maps showing extent and thicknesses of stream and lake sediments.
• Maps indicating any discrete features.
• Maps and cross sections showing the unsaturated zone properties.
• • Potentiometric surface maps of aquifer(s) showing hydraulic boundaries.
• Maps, cross sections, or tables showing storage properties of the aquifers
and confining units.
Maps, cross sections, or tables showing hydraulic conductivity of the
aquifers, confining units, and stream and lake sediments.
• Maps, hydrographs, and/or tables of water-budget information, including
evapotranspiration, runoff, ground-water recharge, ground-water
pumping, and gains/losses between ground-water and surface water.
• Maps, cross sections, or tables indicating effective porosity of the
aquifers (required for particle tracking).
• Maps and cross-sections indicating transport parameters;
• Areal and cross-sectional isoconcentration maps of primary contaminants in soil and
ground-water;
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 22 of SO
• Time-series graphs of contaminant concentrations;
Relevant source term inventory information.
A.7.Ld. REVIEW CONSIDERATIONS FOR CONCEPTUAL MODEL FORMULATION
• Is the conceptual model consistent with the field data?
• Are conceptual model simplifications justified?
• Are sufficient data available to meet the modeling objectives?
• Have database deficiencies been clearly identified and modeling
implications discussed?
Have the natural boundaries of the aquifer system been described?
A.7.1.e. MODEL (CODE) SELECTION
The selected model (code) should be described with regard to its flow, contaminant transport and
transformation processes, mathematics, hydrogeologic system representation, boundary conditions, and
input parameters. The reliability of the model (code) should be assessed including a review and listing of:
• Peer reviews of the model's theory (e.g., a formal review process by an individual or
organization acknowledged for their expertise in ground-water modeling or the
publication of the theory in a peer-reviewed journal);
• Verification studies (e.g., evaluation of the model results against laboratory tests,
analytical solutions, or other well accepted models), being able to address PCE/TCE
degradation;
• Relevant field tests (i.e., the application and evaluation of the model to site-specific
conditions for which extensive data sets are available);
• The model's (code) acceptability in the user community as evidenced by the quantity and
type of use;
• Full model documentation; and
• Publication and peer review of model code testing.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. IS, 2003
Page 23 of 50
The assumptions in the model (code) should be analyzed with regard to their impact upon the
modeling objectives and site-specific conditions. Any and all discrepancies between the modeling
requirements (i.e., as indicated by study objectives, conceptual model, and available data) and the
capabilities of the selected model should be identified and justified. For example, the implications of the
selected code supporting 1-, 2- or 3-dimensional modeling; providing steady versus unsteady state
modeling; or requiring simplifications of the conceptual model should be discussed. Other criteria that
should be documented include:
• Selection criteria should be clearly presented for the selected code(s);
• The general features of the code should be discussed, including whether the code is a
proprietary version of the code used for modeling, solution methodologies for the flow
and transport equations, hardware requirements, degree of code testing, and availability
of source code and documentation;
• The assumptions and limitations should be described, particularly those pertaining to the
conceptual model; these would include code dimensionality, ability to simulate
heterogeneities, and flow and transport through the unsaturated zone;
• The basis for regulatory acceptance should be discussed which may include a history of
use, particularly for applications in a similar regulatory context, and;
• Documentation on the source code should be included, with an executable version of the
code and data sets relevant to the problem.
A.7. l.f. MODEL CONSTRUCTION AND CALIBRATION
Model construction includes the design of the model grid for numerical models, selection and
positioning of boundary conditions, and definition of hydraulic and chemical properties. The model
report should document the assumptions and reasons that form the basis of model construction.
For numerical models, generally acceptable rules of grid design and time step selection should be
applied to meet the modeling objectives. The grid chosen for each modeling study should be justified and,
if possible, grid convergence analyses should be documented.
When a numerical model is used, the mapping of the location of the boundary conditions and
other geometric details (e.g., wells, repository, and contaminant sources) on the grid should be evaluated.
If arbitrary or artificial boundaries are used, justification for their use should be given and evidence
presented to demonstrate that their use does not adversely impact the model results within the area of
interest.
The input estimation process whereby data are converted into model inputs (e.g., spatial and
temporal interpolation, extrapolation or Kriging, or averaging) should be described. This description
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 24 of SO
should include a map or table containing the spatial location and the associated values of data used to
perform the interpolation. Important considerations include:
Layering and Gridding
• The grid should be presented as an overlay of a map of the area to be modeled.
• The rationale for the selection of the grid spacing, number of model layers, and the
resulting number of nodes and elements should be given.
• The grid should be refined as needed to properly define boundary conditions such as
rivers and locations where the aquifer is stressed.
• A vertical cross section of the modeled area which displays the vertical layering of the
model with respect to its hydrogeology should be included.
• Horizontal and vertical grid coordinates and elevations should be identified clearly.
Boundary and Initial Conditions
The report should clearly identify assigned boundaries and initial conditions in figures
and tables.
• Selection of all boundaries and initial conditions should be justified.
• Uncertainty surrounding boundaries and initial conditions should be discussed.
• The boundaries should be positioned to ensure that future simulations will not be
adversely affected by pumping wells or other features that stress the system.
• Ground-Water flow boundaries should coincide with natural features and/or hydraulic
controls (e.g., ground-water divides).
• The areal recharge should not exceed the saturated hydraulic conductivity of the surficial
soil through which it must travel; otherwise ponding would occur.
• Potentiometric lines on streams that are gaining water should point upstream, whereas the
lines should point downstream along losing streams.
• Ephemeral streams generally should not be modeled as constant head boundaries.
Transient boundaries should be clearly identified.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 25 of 50
Streams are frequently modeled as ground-water divides, that is, all ground-water
flowing towards the stream is assumed to be captured by the stream. The modeler should
justify this assumption, as not all streams fully penetrate the aquifer.
Ln the natural system, boundaries may shift with time, and the effect that these positional
changes may have on the results of modeling should be considered.
Surface-water/ground-water interactions should be discussed.
The transient nature of boundaries should be described.
Recharge and evapotranspiration are difficult to determine, and therefore, recharge as a
flux boundary is often used as a calibration parameter. The method for determining
recharge should be presented.
Interpretation and extrapolation methods (e.g., Kriging) should be described.
Boundaries between two types of porous media should always coincide with grid and
layer boundaries.
Calibration
Decision process flow diagrams should show the approach that was taken to calibrate the
model.
The calibration process should be described in detail, including any assumptions and
limitations.
The objectives of calibrating heads and flows should be presented.
/
The sources and magnitudes of errors should be described, particularly the potential
effects on the predictive simulations which will be performexl later (e.g., risk assessment).
Modifications to the parameter values, boundary conditions, and imposed hydraulic
stresses should be discussed in detail, particularly focusing on the response of the
modeled system to the altered values and the rationale for the changes.
The rationale for the convergence criterion for the heads and concentrations should be
presented, in addition to a discussion of the overall mass balance results.
Problems that arose due to failure of the code to converge or numerical instabilities
should be described.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 26 of 50
The calibrated parameter values should be compared with the initial range of these
parameters. Particular emphasis should be placed on parameters that fall outside their
originally estimated range.
If both steady-state and transient calibrations are performed, their similarities and
differences within the results should be discussed. The rationale and selection of time
steps for the transient calibration should be discussed.
The mass-balance results should be discussed.
The calibrated model should be a good match with the conceptual model, such as flow
directions and parameter values.
The results should meet the calibration targets.
The water balance error should be less than one percent.
The calibrated parameters, especially hydraulic conductivity, should not appear patch
worked. Unless there is evidence indicating that hydraulic conductivity values change
substantially from one grid block to the next, it should be assumed that large percentages
of the modeled area are relatively homogeneous.
Areal recharge should be uniform unless there is sufficient justification to vary the
recharge rates locally.
Well logs and aquifer stress test data should be reviewed to ensure that the hydraulic
conductivities assigned to that area are compatible.
The volume of water entering or exiting local streams, lakes, or rivers should be
consistent with the field data.
It should be kept in mind that head and concentration values computed at a node are
representative of an area rather than a point. Model calibration over a short period of time
where there is a large variation in hydraulic heads, such as during a pumping test, should
be avoided.
Vertical gradients within an aquifer in which the well is not fully penetrating should be
considered when the model is calibrated.
A list and a figure indicating the final calibrated values for parameters and boundary
conditions should be included.
The match to the calibration targets should be shown in figures as well as in tables.
Sections within the model should be outlined and discussed according to their "goodness
of fit" to the calibration targets.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 27 of SO
• Particle tracking should be shown in planar and cross-sectional views.
Sensitivity Analysis
• The approach undertaken for the sensitivity analysis should be detailed.
• The rationale for selecting parameters for the sensitivity analysis and for determining
whether there were sufficient simulations investigating single or multiple parameters
should be presented.
• The sensitivity of model calibration quality and model predictions to variations in
parameter values, including grid spacing, time steps, and boundary conditions, should be
discussed, emphasizing parameters in which there is a large degree of uncertainty and the
results are very sensitive.
• The relevance of the overall uncertainty and sensitivity with respect to the objectives of
the predictive simulations should be discussed.
• The results of the sensitivity analysis should be displayed in a graph as well as in
narrative form.
A.7.l.g. SENSmVITY/UNCERTAINTY ANALYSIS
Many of the modeling scenarios will involve parameters that can vary over a considerable range
and field measurements of many parameters are lacking. For this reason, the sensitivity of model
predictions to key model parameters should be documented. Documentation should include the
following:
• A range tested for selected parameters and how they were chosen;
• A number of model simulations conducted for each parameter tested;
• How sensitivity coefficients or other measures of model sensitivity were computed.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 28 of 50
A8 SPECIAL TRAINING REQUIREMENTS/CERTIFICATION
EPA, States, and the regulated community employ ground-water models for a variety of purposes
including at least the following: evaluations of corrective action options and remedial studies; risk
assessment; performing wellhead assessments; evaluating possible leachate migration from solid non-
hazardous waste landfills; mine closure planning and acid mine drainage problems; understanding
contaminant fate and transport at hazardous waste sites; supporting State risk reduction programs;
evaluating natural attenuation; mass balance geochemical modeling; and uses of models by permit
applicants. Ground-water modeling may be a formidable task due to the complexity of the sciences
underlying ground-water modeling and because the type and level of specialized expertise needed to carry
out the array of modeling related tasks. While no formal Federal licenses/certifications are necessary for
EPA to develop the upper sand unit model, it is nevertheless important for EPA to demonstrate modeling
competency to the public and regulated community. The technical complexities and difficulties
associated with ground-water modeling have led to the development of a project modeling group with the
appropriate skill mix. Members of the modeling group, or individuals with expertise directly available to
the group, have education and experience in geology and 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.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15, 2003
Page 29 of 50
A9 DOCUMENTATION AND RECORDS
Documentation of the modeling process is crucial for assuring the defensibility of the modeling
application and for providing enough information so that a thorough peer review may be conducted. The
EPA Project Manager will maintain and archive all modeling files (hard copy and electronic) in
accordance with Agency records management requirements. In general, modeling files are expected to be
categorized as follows:
Files for Conceptual Model;
Files for Water Level Data;
Files for River Stages;
Files Related to Contaminant Concentrations;
Files for Computer Model Setup;
Files for Initial Conditions and Calibration;
Model Output Files;
- Report Files;
Processor Software Manuals;
MODFLOW, MODPATH, and MT3D/RT3D
Model Review/peer Review Files; and
QAAP Files.
For electronic files, the size of any particular file and ability to access the information during
model development will determine the optimum electronic file storage device and backup file location
(e.g., floppy disk, computer hard drive, EPA network drive, or compact disk). During model
development, the computer hard drive (EPA computer ID # A38458) will contain the files necessary to
develop, run, and calibrate the model and will be stored under the file name C: PLAQUEMDME-
FLOWMODEL. VMF. A backup copy on the same computer and disk drive will be named C:
PLAQUEMINE-FLOWMODEL.VMF.BACKUP. A second backup copy will be placed on the Region 6
network disk drive under the filename F:\USER\SELLINGE\PLAQUEMINE-
FLOWMODEL.VMF.BACKUP. When the flowmodel and contaminant transport model are fully
operational, a third backup copy will be written to a compact disk. This compact disk will be labeled
"Plaquemine Ground-Water Model 2003" and will be maintained by the project manager.
-------
Combined EPA QAPP and
Work Man
Revision No. 1
Date Aug. 15,2003
Page 30 of 50
GROUP B: MEASUREMENT AND DATA ACQUISITION
The sections below address Group B, Sections B7, B9, and BIO, which are referenced by
"EPA Guidance for Quality Assurance Project Plans for Modeling" (EPA QA/G-5M) (December 2002)
as being especially relevant for modeling. QAPP The remaining Group B Sections, Sections B1 -B6, and
B8, are addressed by Booz, Allen, and Hamilton (June 2003) in: Draft Quality Assurance Project Plan,
Myrtle Grove Ground- Water Investigation, Semi-Annual Ground-Water Sampling Program. Plaquemine,
Louisiana. The June 2003 QAPP will be attached to the final signed copy of this QAPP and retained by
the project manager. Copies will be available to reviewers upon request.
B7 MODEL CALIBRATION
The purpose for calibrating this model is to produce simulated water level and contaminant
transport results that are generally consistent with field measurements. Model calibration procedures will
be accomplished by 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. Any identified deficiencies in calibration will be resolved to the extent
possible by adjusting model input parameters, initial conditions, and boundary conditions so that the
model simulates the aquifer system to a desired level of accuracy. The degree of success in calibration
will be presented in the final modeling report.
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
user manual (Waterloo Hydrogeologic, Inc., 2000).
The Mean Error is defined by the equation:
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15, 2003
Page 31 of 50
I
MeanError =—
«
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.
1 ° .
MeanAbsoluteError =—^JiXcalc — Xobs]i
n =
The Standard Error of the Estimate (S.E,E.) is provided by (this error estimate is also referred to as the
calibration residual:
_ •-Xobs)-i -\ > {Xcalc- Xobs),
S.£.E.= U±d
n - 1
(.
); - ^
V <-«
The Root Mean Squared (RMS) is given by:
RM'S =- MXcalc-Xobs),
The Normalized Root Mean Squared error (Normalized RMS) is given by the RMS divided by the
maximum difference in the observed head values:
NormalizedRMS =
(Xobs)mK -
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.
-Xcalc) [ (Xobs-Xaveobs)1
I*I,Xob,*-
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. IS, 2003
Page 32 of 50
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
tune, residuals versus time, normalized residuals versus time, and error versus time. In terms of
calibration for contaminant fate and transport, graphs are also available for calculated versus observed
concentrations and concentration versus time.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 33 of SO
B9 NON-DIRECT MEASUREMENTS (DATA ACQUISITION REQUIREMENTS)
Data obtained from non-measurement sources will include published and unpublished
information obtained from literature searches, information from EPA Region 6 RCRA facility files, and
information obtained from an existing computer database. Collecting, organizing, and interpreting quality
data is critical to the success of this modeling effort.
Two organizations are involved with conducting the literature searches. (I) the EPA Region 6
Library staffed by ASRC Aerospace Corporation, and (2) Booz Allen Hamilton Inc. The EPA 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,
naps, and other literature material critical to this modeling project. Booz Allen 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 Hamilton Inc. has provided a bibliography relating to the Plaquemine aquifer
and currently under review by the project manager. Generally, only information obtained from peer-
reviewed, published, and authoritative scientific information sources will be utilized in the model in order
to ensure an acceptable level of data quality.
The Region 6 RCRA file room contains facility files for the DOW Louisiana Operations facility.
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 (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.
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. A spread-sheet list of registered water wells and related data for
Iberville Parish and West Baton Rouge Parish has been obtained from the Louisiana Department of
Environmental Quality, which obtained their data from DOTD. For the LDEQ/DOTD data, Region 6
cannot independently make any statements about the quality data and information obtained from DOTD.
However, Region 6 will try and confirm well data providing that confirmatory information are available.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. IS, 2003
Page 34 of 50
For EPA owned wells and DOW Chemical wells, confirmatory information includes obtaining
Professional Land Surveyor reports, and global positioning system (GPS) location checks performed by
LDEO Staff durinp fh<»_ir nvwall
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15, 2003
Page 35 of 50
BIO DATA MANAGEMENT AND HARDWARE/SOFTWARE CONFIGURATION
This section introduces computer modeling programs (MODFLOW and MODPATH,) and data
processing programs which may be utilized in this project (Visual Modflow and Ground-Water Modeling
System). For a discussion on data management, the reader is referred to Section A10: Documentation and
Records. Many of the essential elements, properties, and numerical values that will enable these
computer programs to run are described in the Conceptual Model. Certain sections of the following
discussion about MODFLOW were taken from the U.S. Geological Survey public domain internet
website (www.water.usgs.gov)
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 ground-water flow
modeling program because of its ability to simulate a wide variety of ground-water 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
regulatory agencies, universities, consultants, and industry for ground-water investigations, development
of remedial designs, and is accepted as suitably reliable for use in legal proceedings. 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 unconfmed 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 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
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 36 of 50
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
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 ground-water system.
MQDPATH 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.
Program code dealing with contaminant fate and transport to be used in the model are MT3D
(mass transport in three-dimensions) or RT3D (reactive transport in threeniimensions). The selection of
which code will depend on conceptual model and whether reactions from PCE/TCE to vinyl chloride will
need to be simulated. MT3D is a modular three-dimensional transport program for simulation of
advection, dispersion, and chemical reactions of contaminants in ground-water. MT3D is intended for
use with MODFLOW or any other finite-difference flow model, and is based on the assumption that
changes in the concentration field will not substantially affect the flow field. RT3D is based on MT3D, is
for simulating reactive multi species transport in three-dimensional ground-water aquifers.
To assist with running the MODFLOW, MODPATH, and contaminant fate and transport
programs, data processing programs will be used. Possible program choices in the Multimedia Planning
and Permitting Division, depending on availability, are Visual Modflow and/or Ground-Water Modeling
System (GMS). Visual Modflow is a proprietary modeling program produced by Waterloo
Hydrogeolpgic Inc., and is designed to facilitate model development, data input, calibration, and the
visualization of model output. It is anticipated that at least the initial ground-water flow modeling will be
conducted with Visual Modflow. GMS is generally considered to be a more sophisticated and
comprehensive ground-water modeling program that will be utilized if/when specific chemical reactions
need to be simulated. Details on data management within Visual Modflow are given below as referenced
from the user's manual.
Visual Modflow is considered a fully-integrated ground-water 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 ground-water 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/Wmdows 98/Windows NT 4.0 (Service Pack 3).
Visual Modflow has three main modules: the Input Module, Rim Module, and Output Module.
The Input Module allows the user to graphically assign all of the necessary input parameters for building
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 37 of 50
a three-dimensional ground-water flow model. The input menus represent the basic model building
blocks for assembling a data set for MODFLOW, MODPATH, and ZoneBudget. The menus are
displayed in logical order and guide the modeler through steps necessary to design a ground-water 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.
Numerical model data management is an integral component of Visual Modflow. Visual
Modflow stores all data as a set of data files. Input files are ASCII files, however some output files are
binary. If any formatting mistakes are in the input file, Visual Modflow will not process the data.
Appendix-A of the Visual Modflow User's Manual lists all data files and describes their formats, and the
reader is referred to the manual for detailed information. The file extension .vmf contains the basic
project file. The model being developed will be given the filename: Plaquemine-flowmodeLvrnf. Upon
completion and approval of the final computer model, all the Visual Modflow files will be copied to a
"closed" CD for safe keeping.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 38 of 50
GROUP C: ASSESSMENT AND OVERSIGHT
Cl ASSESSMENT AND RESPONSE ACTIONS
Element Cl describes the different types of assessments and model performance evaluations to be
performed during the model development process. These assessment and evaluation activities, described
below, will ensure that the quality objectives and criteria for model input/output set in Section A7 Model
Development and Quality Review Criteria are being fully achieved.
Objectives and Data Requirements
The first step in the model development process is the complete assessment of available input
data needed to build the Conceptual Model and subsequently a numerical model that will meet the project
objectives stated in Section AS. Data review will include historical ground-water level data, river flow
and stage data at various point along the section of the Mississippi River to be modeled, pumping
schedules of extraction wells, recharge data within the model area, and contaminant concentration/plume
chemistry information. Aquifer test data, tithologic profiles from soil borings will also be reviewed to
assess if the aquifer can be characterized adequately. The review process will determine whether the data
are sufficient to support a transient ground-water flow model that can include the periodic movement of
water to/from the Mississippi River to the upper sand unit of the Plaquemine Aquifer. The assessment
should evaluate data uncertainty, limitations, and weaknesses. For a transient flow model, the time frame
for model calibration and validation will be established after review of temporal and spatial distribution of
input data For calibration purposes, data spanning a multiple year period should be available. A
rigorous calibration of ground-water surface\ water interface will require adequate flow and water level
data within the near-river zone.
After complete review of available data, the project should either move forward to building the
Conceptual Model, or a recommendation should be made for collecting additional data needed to ensure
that a model can be developed that meets the project objectives. Any recommendations for additional
data needs will be provided to the Chief of the Arkansas/Louisiana Section as indicated on the project
organizational chart contained in Section A4.
Review Considerations for Conceptual Model Development
If the data assessment process concludes that available input data are acceptable and adequate for
modeling purposes, the next phase would be building the Conceptual Model. During this phase, the
assessment process will evaluate:
• Whether the hydrogeologic system can be adequately described with available data to
meet the objectives listed in Section A5;
• Whether water budget analysis is adequate to describe the boundaries of the system
including the river, and
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. IS, 2003
Page 39 of 50
Uncertainties in the conceptual model and their possible effect on model output.
Once the Conceptual Model is complete, the assessment process will evaluate if the Conceptual
Model meets the criteria listed in Section A7.1 .d. The response action would be to determine whether a
numerical model based on the Conceptual Model will meet the project objectives, or whether a
recommendation will be made for collecting additional data needed for an adequate Conceptual Model. If
the Conceptual Model is satisfactory and meets the criteria listed in Section A.7.1.d, the project should
move forward towards building the numerical model. If the Conceptual Model does not meet the criteria
listed in Section A7. l.d, the Chief of the EPA Arkansas/Louisiana Section and the Region 6 Quality
Assurance Manager shall be notified.
Code Selection
The selected codes MODFLOW, MODPATH, and MT3D/RT3D are public domain, industry
standards and have been used extensively for many years. Therefore, a rigorous assessment of the
selected codes is not required. However, the assessment process should evaluate whether the selected
models, with their underlying assumptions and limitations, are capable of meeting the project objectives
outlined in Section A5.
Model Construction and Calibration
Once construction of the numerical model is underway, several assessments will be performed
throughout the model development process to ensure that model development and calibration criteria
established in A.7. l.f are being satisfied. The model may require calibration to both steady-state and
transient conditions. An initial steady-state model assuming average conditions may be calibrated to
estimate input parameter distribution. A transient calibration may follow to improve parameter estimation
such as aquifer hydraulic conductivity and riverbed conductance.
If preliminary model results do not satisfy the target calibration criteria and validation
requirements, all possible errors and accuracy of input data, model formation, and field observations will
be thoroughly investigated. If adjustments to calibration criteria or model objectives are needed, they will
be fully investigated and documented, and revisions to the combined QAPP and Work Plan will be issued
through the formal QA process.
Once a satisfactory calibration is achieved, the model will be validated over a time period for
which observed data such as river flow, river stage, and ground-water elevation are available.
Model validation will apply a weight of evidence approach that will include comparison of time
series plot of observed and simulated values of fluxes and water level data. The model and observed data
comparison process will recognize the inherent error and uncertainty in both the model and the
observations.
Simulation of Scenarios.
Once the model is calibrated and validated, ground-water flow and contaminant fate and transport
modeling will be performed. The simulations will be assessed to ensure DQOs established in Section
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 40 of 50
A.7.1 are achieved. Particle tracking will be performed to assess generalized flow directions and flow
rates. The ability of particle tracking to illustrate change in flow direction and flow rate under transient
conditions will be assessed.
Sensitivity Analysis
Sensitivity of model output to key model input parameters over their expected range of variability
will be assessed and documented in the final stage of the numerical modeling process. Sensitivity to
water level and river flow fluctuations, recharge, and pumping rate of wells on model simulations will be
studied. The sensitivity analysis should evaluate how uncertainty in model output may be reduced in a
cost-effective manner during future data gathering efforts.
The modeling group will conduct model assessments processes described above continually
during the model development process. Performance audits will consist of comparison of model results
with observed historical data, and general evaluation to ensure reasonable model behavior for output
lacking historical data. The assessment process will analyze output data and determine possible
anomalies or departures from assumptions made during the planning phase. Any problems with output
data quality and usability that cannot be resolved with appropriate corrective actions will be reported to
the Chief of the Arkansas/Louisiana Section and to the Region 6 Quality Assurance Manager. Any
corrective actions will be documented in writing and also provided to the Chief of the Arkansas/Louisiana
Section and to the Region 6 Quality Assurance Manager.
C2 REPORTING TO MANAGEMENT
According to the Region 6 Quality Assurance reporting criteria, one-time projects of 12-months
duration or less will require only a final QA report. Therefore, since the schedule contained in Section
A6 extends from February - October 2003, no formal written QA progress reports are anticipated during
the course of this modeling project. The final report will be prepared by the modeling team.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 41 of 50
GROUP D DATA VALIDATION AND USABILITY
Dl DEPARTURES FROM VALIDATION CRITERIA
The review and validation processes for this project are elaborated in Section Cl. Section Cl
also addresses how departures from calibration and validation criteria will be addressed during model
development, and the necessary response actions if acceptance criteria are not met.
D2 VALIDATION METHODS
Model validation will be mainly performed on an ongoing basis during model development and is
discussed in Section Cl. After completion of model development and simulation, a final assessment will
be made on whether the model and its outputs satisfy the user requirements following the criteria in
Section A7.
D3 RECONCILIATION WITH USER REQUIREMENTS
Upon completion of numerical modeling incorporating assessment procedures outlined in Section
Cl, a draft document will be prepared for internal and external peer review, as indicated in A4. The
document will provide a detailed description of short-term and long-term flow characteristics and
contaminant fate and transport within the Plaquemine Aquifer based on the numerical model. The report
will describe data review, verification, and validation processes described in Section C to confirm the
steps of modeling process were followed correctly to produce the model outputs and that the results meet
project objectives. The report will address all A5, A7, and group B elements and present results that meet
the project objectives, and will describe and justify deparatures from any criteria set in the Work Plan or
QAPP. The report will discuss if outputs are the right type, quality, and quantity needed to meet project
objectives and will describe limitations of the output data that may impact usability.
During preparation of the final report, the following table will be used as a checklist to ensure
major steps in the modeling evaluation procedure have been completed. Contents of the final report will
be reflective of the Table of Contents contained in A6.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15, 2003
Page 42 of 51
Reviewed by: Scott Ellinger
Date: September 30. 2004
Major Steps in Modeling Evaluation Procedures
(appraisal refers to only a basic flow model and its informal characteristics)
MODELING AND EVALUATION CRITERIA
APPRAISAL
Yes
No
Comments
OBJECTIVES AND DATA REQUIREMENTS
Are the purpose and scope outlined?
Are the objectives consistent with decision-making needs?
Are the objectives satisfactory ?
Are a site description and waste disposal history provided?
Are the data requirements for the proposed modeling outlined?
Are the sources of data adequately presented?
Are data uncertainties discussed?
Is the probable sensitivity of the future modeling results presented
for the data?
Are the potential data limitations and weaknesses provided?
Are the plans to resolve data limitations discussed?
/
/
/
/
/
V
limited
for basic flow
model only
N/A
N/A
CONCEPTUAL MODEL DEVELOPMENT
Is the physical framework discussed in detail?
Both regional and local?
Is the hydrogeologic framework described in detail?
Both regional and local?
Are the hydraulic boundaries described in detail?
Is the conceptual model consistent with the field data?
Are the uncertainties inherent in the conceptual model discussed?
Are the simplifying assumptions outlined?
Are the assumptions justified?
/
V
/
/
/
/
only as possible,
very limited
data
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15, 2003
Page 43 of 51
MODELING AND EVALUATION CRITERIA
Are the following figures and/or tables1 included-
Map showing location of study area
• Geologic map and cross sections indicating the
area! and vertical extent of the system.
• Topographic map with the surface water bodies.
• Contour maps showing the tops and/or bottoms of
the aquifers and confining units.
* Isopach maps of hydrostratigraphic units.
• Maps showing extent and thicknesses of stream
and lake sediments.
• Maps indicating discrete features (e.g., faults), if
present.
• Maps and cross sections showing the unsaturated
zone properties (e.g., thickness, K^,,).
• Potentionietric surface maps of aquifers) and
hydraulic boundaries.
• Maps and cross sections showing storage
properties of the aquifers and confining units.1
• Maps and cross sections showing hydraulic
conductivity of the aquifers, confining units and
stream and lake sediments.
• Maps and hydrographs of water-budget
information.
APPRAISAL
Yes
/
/
V
/
V
V
No
/
/
/
V
/
Comments
on geologic
maps
for top of
Pleistocene
not available
not available
n/a
river or
constant head
boundary
not available
not available
not available
MODEL APPLICATION
Section
SCOPING ANALYSIS
Are scovins analyses performed? .
/
In some instances tabular representation of the data may be appropriate.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. IS, 2003
Page 44 of 51
MODELING AND EVALUATION CRITERIA
Do scoping results lead to proposed modeling approach?
SITE CHARACTERIZATION MODELING
Code Selection
Is the rationale for the selection clearly presented for
proposed code(s)?
Are the general features of the code(s) presented?
Are the assumptions and limitations of the code(s)
presented and compared to the conceptual model?
Is the basis for regulatory acceptance presented?
Does the code have a history of use?
Is the code well documented?
Is the code adequately tested?
Are the hardware requirements compatible with those
available?
Model Construction
Layering and Gridding:
Is the domain of the grid large enough so that the
boundaries will not interfere with the results?
Do the nodes fall near pumping centers on existing and
potential future wells and along the boundaries?
Is the grid oriented along the principal axes of hydraulic
conductivity?
Is the grid discretized at the scale appropriate for the
problem?
Are areas of sharp contrasts (e.g., hydraulic conductivity,
concentration, gradient) more finely discretized?
Do adjacent elements vary in size by a distance less than
a factor of 1.5?
APPRAISAL
Yes
/
V
/
/
/
/
/
/
/
/
/
No
Comments
n/a
n/a
limited
boundary
effects
possible
no supporting
data
finer grid in
plume area
gradational
layering as
possible
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15, 2003
Page 45 of 51
MODELING AND EVALUATION CRITERIA
Are strong vertical gradients within a single aquifer
accommodated by multiple planes or layers of nodes?
If matrix diffusion is important, are the confining units
adequately discretized in the relevant regions of the
model?
Is the grid irtore finely spaced along the longitudinal
direction of simulated contaminant plumes?
Is the aspect ratio less than 100:1 ?
Are the following figures included:
• Grid presented as an overlay of a map of the area
to be modeled.
• A vertical cross section(s) which displays the
vertical layering of the model grid.
Boundary and Initial Conditions
Is justification provided for the selection of all boundary
and initial conditions?
Are model boundaries consistent with natural hydrologic
features?
Are the boundary and initial conditions consistent with
the conceptual model?
Are the uncertainties associated with the boundaries and
initial conditions addressed?
Are the boundaries far enough away from any
pumping/injection centered to prevent "boundary
effects"?
Are transient boundaries discussed?
Is the rationale given for simplifying the boundaries from
the conceptual model discussed?
Are the values for the assigned boundaries presented?
Model Parameterization
APPRAISAL
Yes
/
/
/
/
/
/
/
/
/
I
No
Comments
gradational
layering as
possible
n/a
n/a
n/a
as possible for
general head
limitations
discussed as
possible
limited effects
possible
limitations
discussed as
possible
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 46 of 51
MODELING AND EVALUATION CRITERIA
Are data input requirements fully described?
Is the discussion of the data well founded with respect to
Objectives and Data Review Section ?
Are the interpretation and extrapolation methods (e.g.,
Kriging) adequately presented?
Do the figures and tables completely describe the data
input with respect to discrete components of the model?
Are the model parameters within the range of
reported or measured values?
MODEL CALIBRATION
Has calibration been attempted?
Is the rationale for model calibration approach
presented?
Are the calibration procedures described in detail?
Are the calibration criteria presented?
Does the calibration satisfactorily meet specified criteria?
Is the rationale presented for selecting convergence
criteria?
Are code convergences and numerical instabilities
discussed?
Do the calibrated parameters fall within their expected
ranges?
APPRAISAL
Yes
V
/
/
/
/
/
/
/
/
No
/
Comments
n/a
as reported
values are
available
only basic
calibration for
screening level
model
informal
procedures
main
targets/water
levels
for water
levels
discussed with
model
reviewers
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 46 of 51
MODELING AND EVALUATION CRITERIA
Are data input requirements fully described?
Is the discussion of the data well founded with respect to
Objectives and Data Review Section ?
Are the interpretation and extrapolation methods (e.g.,
Kriging) adequately presented?
Do the figures and tables completely describe the data
input with respect to discrete components of the model?
Are the model parameters within the range of
reported or measured values?
MODEL CALIBRATION
Has calibration been attempted?
Is the rationale for model calibration approach
presented?
Are the calibration procedures described in detail?
Are the calibration criteria presented?
Does the calibration satisfactorily meet specified criteria?
Is the rationale presented for selecting convergence
criteria?
Are code convergences and numerical instabilities
discussed?
Do the calibrated parameters fall within their expected
ranges?
APPRAISAL
Yes
V
/
/
/
/
/
/
/
/
No
/
Comments
n/a
as reported
values are
available
only basic
calibration for
screening level
model
informal
procedures
main
targets/water
levels
for water
levels
discussed with
model
reviewers
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. IS, 2003
Page 47 of 51
MODELING AND EVALUATION CRITERIA
Are discrepancies explained?
Has the calibration been tested against actual field data?
Are the differences between steady-state and transient
calibrations presented?
Could other sets or parameters have calibrated the code
just as well? Is this discussed?
Are areal and cross-sectional representations of the final
calibrated results included for both hydraulic heads ?
Does calibration of the model take into account the
inconsistency between point measurements at wells and
areal averages of model output?
Is the match between the calibration targets and final
parameters shown diagrammatical^?
Were calibrating errors presented quantitatively
through the use of descriptive statistics?
If particle-tracking was performed, are these results
shown?
Is the calibrated model consistent with the conceptual
model?
Are any changes to the conceptual model discussed and
justified?
Is non-uniform areal recharge applied? Is this approach
justified?
Does the calibration properly account for vertical
gradients?
APPRAISAL
Yes
/
V
/
V
/
/
/
No
/
Comments
discussed with
model
reviewers
discussed with
model
reviewers
unlikely;
informal
sensitivity
checks
suggests
limited ranges
available in
model
(electronic)
option not
available in
software
version
n/a
n/a
no available
supporting
data
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15, 2003
Page 48 of 51
MODELING AND EVALUATION CRITERIA
Is the calibrated hydraulic conductivity field consistent
with the geologic logs and aquifer stress tests?
Are the convergence criteria appropriate?
Was a mass balance performed?
Is the water-balance error less than 1%?
Are the mass balance results for the calibrated model
discussed?
Is the model's water balance consistent with known flows
of rivers and levels of lakes?
SENSITIVITY ANALYSES
Was a sensitivitv analysis performed?
Is the approach to the sensitivity analysis detailed?
Were all input parameters selected for investigation?
If not, was rationale presented for excluding parameters?
Was a sensitivity analysis performed on the boundary
conditions?
Are the ranges of parameters appropriate?
Were sufficient simulations performed? Was justification
provided?
Was the relevance of the sensitivity analysis results to the
overall project objectives discussed?
APPRAISAL
Yes
V
/
/
/
/
/
No
/
Comments
no available
supporting
data
Informally
discussed with
lab
limited general
head
supporting
information
n/a
n/a
no supporting
data
informal only
main
parameters
only
during
calibration for
general head
boundaries
determined on
during
calibration
numerous
simulations
performed
for flow
directions only
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15, 2003
Page 48 of 51
MODELING AND EVALUATION CRITERIA
Is the calibrated hydraulic conductivity field consistent
with the geologic logs and aquifer stress tests?
Are the convergence criteria appropriate?
Was a mass balance performed?
Is the water-balance error less than 1%?
Are the mass balance results for the calibrated model
discussed?
Is the model's water balance consistent with known flows
of rivers and levels of lakes?
SENSITIVITY ANALYSES
Was a sensitivitv analysis performed?
Is the approach to the sensitivity analysis detailed?
Were all input parameters selected for investigation?
If not, was rationale presented for excluding parameters?
Was a sensitivity analysis performed on the boundary
conditions?
Are the ranges of parameters appropriate?
Were sufficient simulations performed? Was justification
provided?
Was the relevance of the sensitivity analysis results to the
overall project objectives discussed?
APPRAISAL
Yes
V
/
/
/
/
/
No
/
Comments
no available
supporting
data
Informally
discussed with
lab
limited general
head
supporting
information
n/a
n/a
no supporting
data
informal only
main
parameters
only
during
calibration for
general head
boundaries
determined on
during
calibration
numerous
simulations
performed
for flow
directions only
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. IS, 2003
Page 49 of 51
MODELING AND EVALUATION CRITERIA
Are the results presented so that they are easy to
interpret?
Were sensitivity analyses performed for both the
calibration and the predictive simulations?
APPRAISAL
Yes
V
/
No
Comments
informal only
only
informally
The internal peer review of the preliminary draft report will be conducted by:
• EPA Region Multimedia Planning and Permitting Division
• EPA Robert S. Kerr Environmental Research Center
Following the internal peer review, a revised report will be prepared for external peer review,
following additional modeling simulations if needed. The external peer review will be conducted by:
• Louisiana Department of Environmental Quality
• State of Louisiana Geological Survey
• United States Geological Survey
A final report will be issued after addressing all external peer review comments.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 50 of 51
REFERENCES
American Society for Testing and Materials. 2002. Annual Book of Standards Volume 4.08.
Philadelphia, Pennsylvania.
Booz, Allen, and Hamilton, June 2003, Draft Quality Assurance Project Plan, Myrtle Grove Ground-
Water Investigation, Semi-Annual Ground-Water Sampling Program, Plaquemine, Louisiana,
prepared under EPA RCRA Enforcement, Permitting, and Assistance Contract, Work
Assignment R06804.
Harbaugh, A.W., and McDonald, M.G. 1996. User's documentation for MODFLOW-96, an update to
the U.S. Geological Survey modular finite-difference ground-water flow model. U.S. Geological
Survey, Washington, D.C.
McDonald, M.G., and Harbaugh, A.W. 1988. A modular three-dimensional finite-difference ground-
water flow model: U.S. Geological Survey, Washington, D.C.
Pollock, D. W. 1994. User's guide for MODPATH/MODPATH-plot, version 3: A particle tracking post-
processing package for MODFLOW, U.S. Geological Survey Finite-Difference Ground-water
Flow Model. U.S. Geological Survey, Washington, D.C.
U.S. Army Corps of Engineers. 1999. Engineering and Design - Groundwater Hydrology. Manual No.
1110-2-1421, Washington, D.C.
U.S. Environmental Protection Agency. 1993. Data Quality Objectives Process for Superfund, Interim
Final Guidance. EPA540-R-93-071. OSWER. Washington, D.C.
U.S. Environmental Protection Agency. 1994. Report of the Agency Task Force on Environmental
Regulatory Modeling: Guidance, Support Needs, Draft Criteria and Charter. EPA 500-R-94-
001. Office of Solid Waste and Emergency Response. Washington, DC.
U.S. Environmental Protection Agency. 1996. Documenting Ground-Water Modeling at Sites
Contaminated with Radioactive Substances. EPA 540-R-96-003. Office of Air and Radiation,
Washington, DC.
U.S. Environmental Protection Agency. 1998a. EPA Guidance for Quality Assurance Project Plans:
EPAQA/G-5. EPA/600/R-98/018. Quality Assurance Management Staff. Washington, D.C.
U.S. Environmental Protection Agency. 1999a. Data Quality Objectives Process for Hazardous Waste
Site Investigations. EPA QA/G-4HW Peer Review Draft.
-------
Combined EPA QAPP and
Work Plan
Revision No. 1
Date Aug. 15,2003
Page 51 of 51
U.S. Environmental Protection Agency. 1999b. Guidance for Data Quality Objectives Process, Final.
EPA QA/G-4 Peer Review Draft.
U.S. Environmental Protection Agency. 2000a. Policy and Program Requirements for the Mandatory
Agency-wide Quality System. EPA Order 5360.1 A2. Washington D.C.
U.S. Environmental Protection Agency. 2000b. EPA Quality Manual for Environmental Programs
(Order 5360 Al). Washington, D.C.
U.S. Environmental Protection Agency. 2001. EPA Requirements for Quality Assurance Project Plans
for Environmental Data Operations QA/R-5. EPA/240/B-01/003. Washington, D.C.
U.S. Environmental Protection Agency. 2002. Guidance for Quality Assurance Project Plans for
Modeling QA/G-5M. EPA/240/R-02/007. Office of Environmental Information. Washington,
D.C.
Waterloo Hydrogeologic, Inc. 2000. Visual MODFLOW - integrated groundwater modeling
environment for MODFLOW,MODPATH, MT3D, RT3D, and WinPEST.
-------
APPENDIX: C
Correspondence and
Model Reviews and Refinements
The contents of this appendix are arranged in chronological order. This appendix
provides a record of model reviews, refinements, and overall construction. All of the review
comments herein have been carefully considered and utilized to refine the model, when
appropriate and to the extent possible with available supporting information and actual field
measurement data. In a few cases, existing information and data limitations have restricted the
degree to which some review comments could be utilized.
The section on Model Manipulations contains calibration tests for ranges of critical model
parameters. Due to testing procedures, not all model output resembles calibrated data due to test
values being potentially out of calibration range (for example, figures 12f, 14e, 14f, lOd).
-------
Correspondence
-------
UNITED STATES ENVIRONMEKTAL PROTECTION AGENCY
\ REGION 6
| 1445 ROSS AVENUE. SUITE 1200
DALLAS. TX 75202-2733
APR 06
MEMORANDUM
SUBJECT: Follow-Up on Evaluation and Recommendations for Ground- Water Flow
Modeling, Plaquemine Aquifer, Iberville Parish and West Baton Rouge Parish,
Louisiana
FROM: Scott Ellinger, Geologist
Corrective Action and Waste Minimization Section
Multimedia Planning and Permitting Division
TO: Steven D. Acree, Hydrologist
Applied Research and Technical Support Branch
Ground-Water and Ecosystems Restoration Division
I want to express my appreciation to the Applied Research and Technical Support Branch,
Ground-Water and Ecosystems Restoration Division for their technical support during
development of the Plaquemine ground-water flow model. I really enjoyed working with you,
Dr. Wang, Mr. Earle, and Mr. Ahsanuzzaman. The valuable reviews and suggestions provided
by the Center for Subsurface Modeling Support during the course of this very challenging and
complex modeling project were very helpful. These reviews and suggestions were very helpful
in constructing the model.
The suggestions provided on March 15, 2004 were considered for inclusion in the model,
and utilized to the degree allowed by existing supporting data and/or by best professional
judgement. Accordingly, the model manipulation results and observations were reviewed and
considered relative to a transient water level calibration. The modeled and observed
potentiometric surfaces were compared for the range of current river conditions, as recommended
by the laboratory, by using a new set of data representing low and high river conditions measured
during Fall 2003 and Spring 2004. This validation step produced flow field conditions generally
similar to the flow field directions derived from the original calibration data set.
I also appreciate your willingness to be available in the future, if related needs arise on
this or other complex sites. Thanks a lot for your assistance.
Internet Address (URL) • http://www.epa.gov
Rocyded/RoeycUbta • Printed with Vegetable OB Based Inks on Recycled Paper (Minimum 25% Postconsumer)
-------
, ..
\ UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
NATIONAL RISK MANAGEMENT RESEARCH LABORATORY
GROUNOWATER AND ECOSYSTEMS RESTORATION DIVISION
p °-BOX1198 •ADA-OK 7482°
March 19, 2004
MEMORANDUM
OFFICE OF
RESEARCH AND DEVELOPMENT
SUBJECT: Plaquemine Aquifer Contamination, Plaquemine, Louisiana (01RC06-001)
Contaminant Evaluation, Lighthouse Road Landfill
FROM: Steven D. Acree, Hydrologist
Technical Assistance & Technology Transfer Branch
TO: Nancy Pagan
U.S. EPA, Region 6
As requested, analytical results from soil samples obtained from boring LR-27 at the
Lighthouse Road Landfill site in Plaquemine, Louisiana, were reviewed by Daniel F. Pope, Bruce
E. Pivetz, and Kelly Hurt of Dynamac Corporation. Dynamac is an off-site contractor providing
technical support services to this laboratory. The data and other available information
concerning the site were reviewed to determine possible sources for vinyl chloride contamination
observed in the Plaquemine Aquifer and to evaluate whether sufficient information exists to
provide a screening-level estimate of the concentrations of vinyl chloride that may be produced
under ideal conditions.
Seventeen individual compounds were detected in the soil samples from boring LR-27.
A search of the literature was conducted to determine which of these compounds are known to
have potential to degrade to vinyl chloride in the environment. Of these compounds, the
following compounds were found to have significant evidence indicating that they could degrade
to vinyl chloride in soil, sediments or ground water. No references were found indicating that
the other compounds were likely to degrade to vinyl chloride.
1,1,1- Trichloroethane
1,1,2 - Trichloroethane
1,1,2,2-Tetrachloroethane
1,2-Dichloroethane (EDC)
Hexachloroethane
Tetrachloroethene
Trichloroethylene (TCE)
The concentration of vinyl chloride in ground water that may be produced from the
degradation of these compounds is highly dependent on many factors including, but not limited
to: the redox setting, site microbiology, and ground-water flow parameters. Geochemical
Internet Address (URL) « http://www.epa.gov
Recycled/Recyclable « Printed with Vegetable OH Baaed Inks on 100% Postconiumer, Process Chlorine Free Recycled Paper
-------
conditions suitable for biotransformation and subsequent accumulation of vinyl chloride must be
present as well as the necessary microflora for the required transformations. In addition, any
meaningful estimate of the concentration of a contaminant in ground water would require
information concerning the rates of ground flow and estimates of the biotransformation rates.
The document entitled, "Revised Draft, Lighthouse Road Corrective Action Plan", dated
March 25> 1998, was reviewed to determine whether sufficient information was available to
allow screening-level estimates of the possible range of vinyl chloride concentrations that may be
produced in ground water from the degradation of these compounds. However, data suitable for
evaluating redox conditions arid ground-water flow parameters were not reported. Therefore, it
is felt that estimates of the range of possible concentrations of vinyl chloride would require many
assumptions to be made without supporting information and would not provide meaningful
guidance for use during contaminant transport modeling. In general, it is recommended that
actual site data (i.e., measured concentrations of vinyl chloride in ground water at the landfill
site) be used to the extent possible, rather than calculations of source values, to help bound
possible source strengths for use in a contaminant transport and fate model.
If you have any questions concerning this review, please do not hesitate to call me at your
convenience (580-436-8609). We look forward to future interactions with you concerning this
and other sites.
cc: Scott Ellinger, Region 6
-------
Breakdown of analytical results from (soil) LR-27 (Boring drilled on 7/19/93 and
11/3/93)
Depth Interval
r-21
2' -3'
4' -5'
5' -6'
7' -8'
8' -9'
Chemical Contaminant
Hexachlorobenzene
Hexachlorobenzene
1,2-dichloroethane (EDC)
Hexachlorobutadiene
Hexachlorobutadiene
Bis (2-chloroethyl) ether
Hexachlorobenzene
1 ,4-Dichlorobenzene
1,2-Dichloroethane (EDC)
1 , 1 ,2i2-Tetrachloroethane
Tetrachloroethene
Hexachlorobutadiene
Bis (2-chloroethyl) ether
Hexachlorobenzene
Hexachloroethane
1,2-Dichloroethane (EDC)
Tetrachloroethene
Hexachlorobutadiene
Hexachloroethane
Hexachlorobutadiene
Bis (2-chloroethyl) ether
Hexachlorobenzene
1,2-Dichloroethane (EDC)
Tetrachloroethene
Hexachlorobutadiene
Hexachloroethane
Hexachlorobutadiene
Bis (2-chloroethyl) ether
Hexachlorobenzene
Hexachlorobutadiene
Hexachloroethane
Analytical Results
5.732 ug/g
64.915 ug/g
0.3364 ppm
2.2494 ppm
16.296 ug/g
4.74 ug/g
124.433 ug/g
0.05 ppm
0.31 ppm
0.13 ppm
0.27 ppm
580.384 ug/g
9.581 ug/g
60.026 ug/g
24.231 ug/g
170.210 ppm
122.069 ppm
863.009 ppm
32.3 14 ppm
3.998 ug/g
2.899 ug/g
37.033 ug/g
32. 1474 ug/g
22.9996 ppm
587.0649 ppm
4.0018 ppm
356.005 ug/g
15.401 ug/g
29.03 ug/g
789.0 ppm
49.3084 ppm
Date Analyzed
7/20/93
7/20/93
8/4/93
8/4/93
7/20/93
7/20/93
7/20/93
8/28/93
8/28/93
8/28/93
8/28/93
7/20/93
7/20/93
7/20/93
7/20/93
7/27/93
7/27/93
7/27/93
7/27/93
7/20/93
7/20/93
7/20/93
8/4/93
8/4/93
8/4/93
8/4/93
7/20/93
7/20/93
7/20/93
9/22/93
9/22/93
LR Table 1/15/04
-------
Depth Interval
9'- 10'
8' - 10'
10' - 12'
Chemical Contaminant
Hexachlorobutadiene
Bis (2-chloroethyl) ether
Hexachlorobenzene
Hexachloroethane
1,2-Dichloroethane (EDC)
Trichloroethylene (TCE)
1,1,2 - Trichloroethane
Tetrachloroethene
1,1,2,2 - Trichloroethane
1,3 - Dichlorobenzene
1,4 - Dichlorobenzene
1 ,2-Dichlorobenzene
Hexachlorobutadiene
Hexachloroethane
Hexachlorobenzene
Hexachlorobutadiene
Hexachloroethane
1,4 - Dichlorobenzene
1,2-Dichloroethane (EDC)
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethene
1,1,2 - Trichloroethane
Trichloroethene
Hexachlorobenzene
Hexachlorobutadiene
Hexachloroethane
Chlorobenzene
Chloroform
1 ,2-Dichlorobenzene
1 ,3-Dichlorobenzene
1,4 - Dichlorobenzene
1,2-Dichloroethane (EDC)
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethene
1, 1,2 - Trichloroethane
Trichloroethene
Analytical Results
1339.70 ug/g
73.443 ug/g
83.564 ug/g
203. 165 ug/g
544.4243 ppm
26.6042 ppm
46.9863 ppm
174.2637 ppm
26.7307 ppm
3.8497 ppm
3.25 19 ppm
2.092 ppm
1048.568 ppm
153.728 ppm
44.7 ppm
523.9 ppm
42. 1 ppm
3.29 ppm
585.4 ppm
12.02 ppm
188.8 ppm
28.9 ppm
S.llppm
188.0 ppm
1,589 ppm
44.0 ppm
2.71 ppm
4.22 ppm
5.70 ppm
6.54 ppm
20.9 ppm
1,713 ppm
108.6 ppm
630.6 ppm
229.4 ppm
150.1 ppm
Date Analyzed
7/20/93
7/20/93
7/20/93
7/20/93
8/05/93
8/05/93
8/05/93
8/05/93
8/05/93
8/05/93
8/05/93
8/05/93
8/05/93
8/05/93
1 1/1 1/93 (preped on
1 1/09)
ll/ll/93(prepedon
11/09)
LR Table 1/15/04
-------
Depth Interval
12' - 14'
14' - 16'
16' - IS-
IS' 20'
20' - 22'
23' - 25'
Chemical Contaminant
Hexachlorobenzene
Hexachlorobutadiene
1,4 - Dichlorobenzene
1,2-Dichloroethane (EDC)
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethene
1,1,2 - Trichloroethane
Trichloroethene
Hexachlorocyclopentadien
e
1 ,2-Dichlorobenzene
1 ,3-Dichlorobenzene
1,4 - Dichlorobenzene
1,2-Dichloroethane (EDC)
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethene
1,1,1- Trichloroethane
1,1,2 - Trichloroethane
Trichloroethene
Hexachlorobutadiene
1,2 - Dichlorobenzene
1,2-Dichloroethane (EDC)
Tetrachloroethene
1,1,2 - Trichloroethane
Trichloroethene
Hexachlorobutadiene
1,2-Dichloroethane (EDC)
Tetrachloroethene
Hexachlorobutadiene
1,2-Dichloroethane (EDC)
Tetrachloroethene
Hexachlorobutadiene
Hexachlorobenzene
Analytical Results
66.9 ppm
430. ppm
7.03 ppm
38.7 ppm
8.31 ppm
357.7 ppm
28.25 ppm
3 1.99 ppm
8.0 ppm
0.05 ppm
0.06 ppm
0.19 ppm
33.63 ppm
0.1 3 ppm
8.44 ppm
0.05 ppm
4.53 ppm
1.85 ppm
2.66 ppm
0.055 ppm
5.96 ppm
0. 17 ppm
0.12 ppm
0.44 ppm
0.44 ppm
0.07 ppm
0,07 ppm
1.72 ppm
0.04 ppm
0.05 ppm
10.20 ppm
0.47 ppm
Date Analyzed
1 1/1 l/93(preped on
1 1/09)
1 1/1 l/93(preped on
11/09)
ll/ll/93(prepedon
11/09)
ll/ll/93(prepedon
11/09)
ll/ll/93(prepedon
11/09)
1 1/1 l/93(preped on
11/09)
LR Table 1/15/04
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
NATIONAL RISK MANAGEMENT RESEARCH LABORATORY
GROUNOWATER AND ECOSYSTEMS RESTORATION DIVISION
P.O. BOX 1198 • ADA, OK 74820
March 15, 2004
OFFICE OF
MEMORANDUM RESEARCH AND DEVELOPMENT
SUBJECT: Plaquemine Aquifer Contamination, Plaquemine, Louisiana (01RC06-001)
Evaluation and Recommendations for Groundwater Flow Modeling
FROM: Steven D. Acree, Hydrologist
Applied Research & Technical Support Branch
TO: Scott Ellinger
U.S. EPA, Region 6
The transient groundwater flow model provided on February 13, 2004, has been reviewed
by Dr. Mingyu Wang, Rob Earle, and Noman Ahsanuzzaman of Shaw Environmental &
Infrastructure, Inc. Shaw is an on-site contractor providing technical support services to this
laboratory in our Center for Subsurface Modeling Support. This review focused on the technical
adequacy of the flow model construction and calibration. Files related to contaminant transport
and predictive simulations using the calibrated flow model were not provided for review. In
addition, the model was also manipulated by examining and calibrating several important
parameters using a structured approach and the observed data in a particular period. Through this
procedure, it is believed that the calibration error is minimized.
In general, the flow model was constructed using a comprehensive modeling process and
accepted practices. This model fundamentally captures the geologic and hydrogeologic attributes
of the principal modeling components for the upper Plaquemine Aquifer and the aquitards. The
calibrated parameters were able to generate good matching between the observed and calculated
water heads in the upper Plaquemine Aquifer. Incorporation of the recommendations provided
below, including changes in the boundary conditions used for the lower Plaquemine Aquifer,
should result in a product that is useful for estimating the possible range of previous and future
flow conditions within the limitations imposed by the site complexity and available historical
and physical data. Following calibration, it is recommended that modeled and observed
potentiometric surfaces be compared for the range of current river conditions to verify the
accuracy of predictions under the current conditions.
Based on this review and manipulations, the following recommendations are provided.
Please note that some of the following comments address issues related to the inclusion of
significant pumping stresses in predictive simulations. In addition, the model manipulation
results and the observations from those results are attached.
Internet Address (URL) • http://www.epa.gov
Recycled/Recyclable « Printed with Vegetable OH Based Inks on 100% Postconsumer, Process Chlorine Free Recycled Paper
-------
*.
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
NATIONAL RISK MANAGEMENT RESEARCH LABORATORY
GROUNDWATER AND ECOSYSTEMS RESTORATION DIVISION
P.O. BOX 1198 • ADA. OK 74820
October 17,2003
MEMORANDUM
OFFICE OF
RESEARCH AND DEVELOPMENT
SUBJECT: Plaquemine Aquifer Contamination, Plaquemine, Louisiana (0 1 RC06-00 1 )
Initial Recommendations for Groundwater Flow Modeling
FROM: Steven D. Acree, Hydrologist
Applied Research & Technical Support Branci
TO: Scott Ellinger
U.S. EPA, Region 6
Based on a preliminary review of the groundwater flow model under development and
our discussions on October 8, 2003, the following recommendations regarding model
construction and improvement are provided for your consideration by Noman Ahsanuzzaman
and Robert Earle of Shaw Environmental & Infrastructure, Inc. Shaw is an on-site contractor
providing technical support services to this laboratory. No recommendations are currently
provided regarding construction of a contaminant transport model for this site. At present, it
appears that insufficient data may be available to fully support development of a calibrated
numerical model of contaminant transport. Additional recommendations concerning appropriate
options for evaluating transport will be provided following review of the calibrated groundwater
flow model.
Recommendations
I. Make all cells in the model active cells except for the bottommost layer.
2. Add layers beneath the Mississippi River and in the Piaquemine aquifer. It is
recommended to add the first layer close to the river bottom (e.g., 5 ft) and then gradually
increase the layer thickness by not assigning thickness larger than two-times the
neighboring layers. Also, consider eliminating the top layers that represent the alluvial
sediments so that the model simulates only the Plaquemine aquifer. We recommend
combining the top six layers into one layer with equivalent hydraulic conductivity (i.e.,
keep two layers above the Plaquemine aquifer).
3. Eliminate the constant head boundaries at the upgradient and downgradient edges of the
model. Run the model with only the river boundary condition to see if it will converge
and give a reasonable solution in the study area. Note that the initial head for all model
cells (in all layers) needs to be very close to the actual initial observation while running
Internet Address (URL) • http://www.epa.gov
Recycled/Recyclable -Printed with Vegetable Oil Based Inks on 100% Postconsumer, Process Chlorine Free Recycled Paper
-------
the model with only the river boundary. To start, the average head for the wells near the
western edge could be assigned to all cells (underneath the water table) as initial head. If
the model does not converge or gives unrealistic output, add a general head boundary
(GHB) along the far western edge of the model. Calibrate the head value of the GHB with
respect to any nearby wells and/or known water features. The initial conductance value
for the GHB can be estimated from the cell size and the hydraulic conductivity value
from slug or pumping test data (for a nearby well).
4, Refine the model grid so that the smallest cells in the model study area (where the plume
is located) are 20'x20f or possibly smaller. The dimensions of the smaller cells may be
determined by calculating the particle travel distance of interest based on the seepage
velocity and the time of interest. The distance water travels during this time period
should be greater than or equal to the distance from the center of one cell to the next.
5. A zone (in Zone Budget model) in the layer immediately underneath the river can be
assigned so that fluxes through the bottom of the river can be calculated.
6. Make the Mississippi River feature one continuous river feature in the model by joining
the two major segments along the edge of the model with a single line of river cells.
7. Determine the river hydraulic gradient from the Dock I and Dock2 measurements and the
longitudinal distance (along the river) between the two stations. Divide the river boundary
into several linear segments and calculate the head values at the two edges of each
segment from the hydraulic gradient. River boundaries for each linear segment can then
be applied as a line boundary. The model pre-processor (Visual MODFLOW) would
automatically interpolate the head values for each cell between the start and the end
points of each linear segment.
8. Use the zone budget model to calculate the river discharge (preferably, the flow velocity)
at any section along the river. Check model validity by comparing this value to the actual
river discharge. Adjustment of the river hydraulic gradient could be needed to validate.
Note that if the zone budget model fails to give reasonable output, 'stream boundary'
could be used to represent the river instead of the river boundary. The advantage of the
'stream boundary' is that the user can assign the river flux as an input.
9. Select the datasets for calibrating the flow model. Consider selecting the datasets with the
high and low peaks (all peaks) of the hydrograph. This is necessary to get an average
gradient for the entire modeling period. If required, some datasets in between the high and
low peaks can be ignored to reduce the modeling effort. In case of data unavailability for
some wells near the river (mostly during high peaks), wells away from the river can be
used, if available, for model calibration.
10. Calibrate the flow model for each dataset (one snapshot) under steady-state conditions.
Once the steady-state models for all datasets within the modeling period are calibrated, a
-------
transient model can be developed by inputting the transient boundaries from the steady-
state models. Note that this transient model can be used for particle tracking analyses
within the modeling period only. To extrapolate the particle tracking, the modeling period
for the transient model needs to be extended.
[f you have any questions concerning these comments, please do not hesitate to call me at
your convenience (580-436-8609). We look forward to future interactions with you concerning
this and other sites.
cc: Nancy Pagan
-------
State of Louisiana
Department of Environiflental Quality
f^
M. J- "MIKE" FOSTER, JR. February 20, 2003 QHA@BOHUNGER
GOVERNOR -pr SECRETARY
Mr. Steve Gilrein rn co
1445 Ross AMenue gt ££
Dallas, Texas ^
-o ^s.
RE: Ground Water Modeling Objectives o 9?
A.N. Wilbert & Sons (Myrtle Grove Trailer Park); AI # 81438 ^ ^
Iberville Parish 3
Dear Mr. Gilrein:
I would like to thank you for the support and assistance that EPA Region VI and EPA's
National Risk Management Research Laboratory (NRMRL) in Ada, Oklahoma have
provided in our ongoing efforts to identify the source of vinyl chloride and cis-1,2-
dichloroethene in the vicinity of Plaquemine, Louisiana. Your efforts have also been
instrumental in ensuring the protection of potential receptors.
The next phase of this ongoing cooperative effort is the ground water modeling phase.
As part of this effort, we have been requested by your staff to provide our ground water
modeling objectives. [ have attached these objectives to this letter and request that your
staff contact us immediately if additional information is needed.
The Louisiana Department of Environmental Quality is relying upon EPA Region VI and
EPA NRMRL to perform ground water modeling and geochemical interpretation and
requests an expedite4 project schedule if at all possible. We feel that projects such as this
show the value of having EPA available for technical expertise and assistance to the
State.
If you have any questions or require additional information please let us know as soon as
possible.
Thank you for your assistance and cooperation as we attempt to resolve the complex
technical issues related to this contaminant plume.
Sincerely,
w
recycled paper
James H. Brent, PhD
Assistant Secretary
TBK/stn
Cc: Nancy Pagan, EPA Region VI
Scott Ellinger, EPA Region VI
Steve Acree, EPA NRMRL
OFFICE OF ENVIRONMENTAL ASSESSMENT
P.O. BOX 82178 - BATON ROUGE, LOUISIANA 70884-2178 • TELEPHONE (225) 765-0355 « FAX (225) 765-0617
AN EQUAL OPPORTUNITY EMPLOYER
-------
Model Objectives
r1
A plume of ground water contamination consisting of vinyl chloride and cis 1,2-
dichloroethene has been identified in the Mississippi River Alluvial Aquifer in Iberville
Parish, Louisiana, The Mississippi River Alluvial Aquifer is encountered at a depth of
approximately 100 feet below ground surface'at this location. Plume size is
approximately 2 miles in a North/South direction and approximately 1 mile in an
East/West direction. Modeling Objectives are as follows:
I. Determine direction and rate of ground water flow through rising and falling
river stages to assist in determining possible source locations and if multiple
source locations are likely;
2. Estimate age, quantity and duration of the release(s);
3. Determine size, fate and transport of contaminant plume to ensure protection
of potential receptors;
4. Determine fate and transport of contaminant plume to estimate plume size and
concentrations into the future;
5. Evaluate the effect of historical pumping conditions on ground water flow
direction;
6. Determine impact of pumping Cfty of Plaquemine Backup supply wells on
plume movement and the potential contaminant concentrations that may enter
the backup wells should wells be used; and,
7. Determine if historical contaminants discharged into the Mississippi River
* may be a source of the identified contaminant plume.
-------
May 28, 2002
MEMORANDUM
SUBJECT: Plaquemine Aquifer Contamination, Plaquemine, Louisiana (01RC06-001)
Interpretation of Geochemical Indicator Parameters
FROM: Steven D. Acree, Hydrologist
Technical Assistance & Technology Transfer Branch
TO: Nancy Pagan
U.S. EPA, Region 6
As requested, analytical results from samples obtained during December 2001 and
February 2002 for parameters indicative of geochemical conditions and contaminants at the
referenced site have been reviewed by Dr. Daniel Pope of Dynamac Corporation. Dynamac
Corporation is an off-site contractor providing technical support services to this laboratory, hi
general, the plume geochemistry appears to be conducive to biodegradation of both cis-1,2-
dichloroethene and vinyl chloride, though degradation of vinyl chloride may be relatively slow
in some parts of the plume. Long-term monitoring at locations representative of the different
environments within the plume would be required to provide useful estimates of contaminant
attenuation rates. With respect to the sources for this contamination, the geochemical data did
not provide any conclusive evidence regarding the possible number or location of source(s).
Additional information concerning the ground-water flow field and interactions with the river is
necessary to allow interpretation of possible source location(s).
The geochemical environment within and adjacent to the contaminant plume in the
Plaquemine aquifer varies from aerobic to methanogenic. There is no nitrate present in any of
the samples, indicating that nitrate is either depleted or, more likely, simply not found at
significant levels. Dissolved manganese is found at concentrations as high as 1350 ug/1,
indicating that manganese reducing conditions have occurred in the aquifer. Ferrous iron is
observed at concentrations up to several mg/1, indicating that iron reducing conditions are also
present. Both observed contaminants, cis-l,2-dichloroethene and vinyl chloride, can be
biodegraded under manganese and iron-reducing conditions. Sulfate is found at concentrations
as high as 40 mg/1, but is depleted in some locations. High sulfate concentrations may reduce
reductive dechlorination by competing as an electron acceptor, though this may not be an
important factor at this site. However, it is notable that sulfate is depleted in most locations
where methane is elevated, and those are the locations where neither contaminant was detected.
Sulfate depletion coupled with methanogenesis may be a significant factor in degrading both
contaminants in some locations at the site.
-------
As the geochemistry varies significantly in different portions of the Plaquemine aquifer, it
is possible that attenuation rates also vary, assuming biodegradation of the contaminants is a
significant component of attenuation. Both cis-l,2-dichloroethene and vinyl chloride can be
degraded under anaerobic and aerobic geochemical regimes observed at the site. However, it is
possible that vinyl chloride may exhibit some accumulation under particular anaerobic
conditions. This may explain the presence of vinyl chloride under conditions where the possible
parent products (tetrachloroethene and/or trichloroethene) have been completely degraded.
1 If you have any questions concerning these comments, please do not hesitate to call me at
your convenience (580-436-8609). We look forward to future interactions with you concerning
this and other sites.
cc: Steve Chustz, LDEQ
-------
VW.Z
«4j|f^
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
NATIONAL RISK MANAGEMENT RESEARCH LABORATORY
SUBSURFACE PROTECTION AND REMEDIATION DIVISION
p. O_ BOX 1198- ADA, OK 74820
July 26, 2001 OFFICE OF
RESEARCH AND DEVELOPMENT
Tim Knight, Administrator
Environmental Technology Division
Office of Environmental Assessment
Louisiana Department of Environmental Quality
P.O. Box 82178
Baton Rouge, LA 70884-2178t
Dear Mr. Knight:
As requested by your office through the Multimedia Planning and Permitting Division of
USEPA, Region 6, available information regarding ground-water contamination near the Myrtle
Grove Trailer Park, Plaquemine, Louisiana, has been reviewed by Dr. John Wilson and Steve
Acree of this division and Drs. Daniel Pope and Kelly Hurt of Dynamac Corporation. Dynamac
is an off-site contractor providing technical support services to this division. The request for
review included the following questions regarding the vinyl chloride contamination at this site:
1. What was the original material spilled?
2. What is the mechanism of the degradation?
3. How old is the spill?
4. Is it expected that this is a single release or multiple sources?
5. Can any conclusions be drawn from the large area, depth, and shape of the plume?
6. What is the hydrogeologic model of the origin, fate, and transport of this plume?
7. Are there any suggestions for alteration of the current protocol?
In general, definitive answers tpthese question? are not possible from the available data.
However, hypotheses may be formulated and an initial conceptual model developed. Such
hypotheses should be tested during future characterization. The objective of this review is to
provide as much insight as possible concerning the source(s), transport, and fate of this
contamination. The following evaluations and recommendations are keyed to the questions that
were posed, as listed above.
1. Characterization of the original source material(s) for the observed ground-water
contamination is not possible from the available data. Based on experience at similar sites, the
original material was probably a more highly chlorinated parent compound of the observed cis-
1,2-DCE and vinyl chloride, such as PCE or TCE, or^mixtures of such compounds that may have
included c/.y-l,2-DCE and vinyl chloride as components. The observed contaminants probably
Recycled/Recyclable . Printed with Vegetable Oil Based Inks on 100% Recycled Paper (40% Postconsumer)
-------
result mainly*from the degradation of the more highly chlorinated parent compounds.
Experience also indicates that'some of the source materials were probably released as dense
nonaqueous phase liquids. In this setting, such liquids would be expected to migrate laterally as
well as vertically through the heterogeneous shallow sediments, possibly forming relatively
broad source areas for continued ground-water contamination. The observation of increasing
contaminant concentrations with depth, in borings 8 and 9 may indicate DNAPL contamination
also exists within the upper sand unit upgradient of these locations.
2. The mechanism for the degradation of the parent compounds is probably reductive
dechlorination. The available literature indicates the aquifer is naturally under reducing
conditions as evidenced by reported iron concentrations in solution and elevated organic matter.
Under siich conditions, reductive dechlorination has been observed at other sites.
3. The hydrogeologic information was not sufficiently site specific-to constrain estimates of
net ground-water seepage velocity or the age of this plume(s) within useful limits. Available
literature states that river stage has a major influence on hydraulic gradient and that this area may
be transitional between gaining and losing river loaches. Therefore, significant temporal
variability in flow direction due to variability in river stage and associated dispersion of dissolved
contaminants would be expected to occur. Based on this information and the size of the
contaminated area, potential plume ages ranging from less than 5 years to greater than 40 years
are possible for this plume(s). The estimates cannot be further constrained using evaluations of
the observed contaminant concentrations. Site-specific information concerning ground-water
velocity in the contaminated zone(s) would be required to refine these estimates.
4. The spacing between well locations is too large to allow identification of meaningful
trends in the concentration data, if present, and, accordingly, firm conclusions concerning the
number of sources or releases responsible for the current plume. In addition, the high degree of
interaction between the aquifer and the river probably results in significant dispersion or
smearing of the plume downgradient of source areas. Based on the size of the plume and
previous experience at other sites, it is highly probable that multiple sources for this
contamination are present.
5. Although few firm conclusions can be drawn, hypotheses may be formed based on the
observed contaminant data. Such hypotheses may be tested during future investigations. Several
characteristics of this data set appear to be pertinent with respect to hypotheses concerning
contaminant sources, transport, and fate. First, the observed contamination occupies a large area
in which concentrations do not vary rapidly with distance. Using the current data, the extent of
the aquifer with observed vinyl chloride concentrations above the MCL of 2 ug/1 is
approximately 1.85E6 m2. Based on depth discrete ground-water samples obtained from borings
8 and 9, contaminants in this area may be migrating predominantly in the lower portion of the
upper sand unit. Assuming homogeneous conditions and that the plume exists between depths of
approximately 140 ft to 190 ft below land surface, the contaminated volume of the aquifer would
-------
be approximately 2-83 E7 m3. If it is assumed that the average concentration of vinyl chloride in
the water is 10 ug/1 and the average aquifer porosity is 30%, approximately 85 kg of vinyl
chloride would be present in the dissolved plume. It should be noted that this is probably an
underestimate as significant dilution occurs when samples are obtained from wells screened
across contaminated and uncontaminated portions of the aquifer. This mass of vinyl chloride
may be produced through reductive dechlorination of approximately 180 kg of TCE or 225 kg of
PCE. This represents a minimum estimate of potential release volumes as it uses highly
conservative assumptions such as complete dissolution and degradation of parent materials and
no loss of vinyl chloride following production. Previous experience indicates that estimates of
source mass in this range probably represent releases from an industrial facility or other user of
large quantities of such materials.
Although the lateral variation in contaminant concentrations throughout the plume is
relatively low, the data indicate that higher concentrations may be present near the northern
extent of the study area and may decrease in a south to southwest direction. This may indicate
net ground-water flow in the contaminated zone(s) is generally in a south to southwest direction
and that sources for ground-water contamination may exist north of the study area. In similar
fashion, contaminant concentrations may increase, decrease, and then increase again in an west to
east transect across the northern boundary of the study area. This may indicate the existence of
multiple source areas for ground-water contamination. However, firm conclusions cannot be
drawn in these respects as the data density is to low to support identification of definitive trends.
Additional'data from the northern portion of the study area and north of the current study area
would be required to better define spatial contaminant trends and potential source locations.
As noted above, the vertical profiles of samples obtained from locations I through 9
indicate that contaminants in the area currently under study may be migrating mainly near the
bottom of the upper sand unit. This may be indicative.of several conditions such as the existence
of an upgradient DNAPL source trapped at this depth within the aquifer, the existence of
sediments with higher hydraulic conductivity at this depth than overlying sediments, or of
significant leakage from the upper sand unit to the lower sand unit in this vicinity. Additional
characterization of ground-water flow parameters and contaminant distribution would be
required to determine the cause of the observed trend with respect to depth.
6. The available information may be used to develop an initial conceptual model for site
hydrogeology and contaminant origin, transport, and fate. As noted above, the tenets and
hypotheses that form this initial model should be tested using site-specific data obtained during
future investigations, remediation, and long-term monitoring activities. Net ground-water flow
direction in the depth range where contaminants are observed may be generally toward the south
to southwest based on regional hydrology and contaminant distribution. However, the net flow
direction may be expected to vary locally in response to pumping and in response to the influence
of the river and other factors. It may also vary between the shallow sediments and deeper sand
units used for water supply. Regional hydrogeologic information support this conceptual model
-------
of ground-water-flow as this area is characterized as transitional between gaining and losing river
reaches. Other factors that support the conceptualization of a south to southwest net flow
direction are the location of the site adjacent to an eroding bank of the river and the presence of
multiple water production wells in this area. As noted, the available information was not
sufficient to constrain estimates of potential ground-water flow rates within useful bounds.
Information from literature indicates that average seepage velocities may be less than 1 ft/d.
Ground-water and associated contaminants within this regime migrate mainly through advection
with significant dispersion due to river/aquifer interactions;
Given the dimensions of the contaminated area, it appears that there may be more than
one source area contributing contaminants to the plume(s). These plume(s) may be the result of
muS.tijple releases and; continuing sqwgces for gfoi<|id-water contj^jk*^Q?i:Xe.g., DNAPLs); The
potential for a DNAPL source or a zone of increased hydraulic conductivity in the lower portion
of the upper sand unit is also indicated by the vertical profiles of contaminant distribution along
the northern study boundary. The observed contamination probably results, in large measure,
from the degradation of more highly chlorinated parent compounds, such as PCE and TCE,
through reductive dechlorination. More detailed contaminant and geochemical characterization
of potential source areas and the resulting plumeS" would be required to provide definitive
information on sources and degradation processes. Insufficient data are available to estimate
contaminant attenuation rates. Accordingly, future plume behavior cannot be reliably predicted.
Based on the.potential net ground-water flow direction, wells used by the City of Plaquemine for
water supply may be at future risk. A more detailed characterization of site hydrology,
contaminant distribution, and temporal behavior would be required to define plume behavior and
allow evaluation of risk management options.
7. The following suggestions are provided for consideration during future investigations.
Review of historical information concerning use and disposal of chlorinated solvents by local
industries and examination of historical aerial photographs may provide information for
development of a more refined conceptual model of potential source areas. Once potential areas
are identified contaminant distributions upgradient and downgradient of these areas may be
obtained to identify source areas that currently contribute contaminants to the plume(s).
Information concerning the size, nature arid vertical distribution of contaminants within active
source areas will allow better evaluations of remedial options and the potential costs and
effectiveness.
It is recommended that a more detailed characterization of net ground-water flow
directions and rates within the contaminated portions of the aquifer be undertaken. This
\mformation will be useful in identifying source areas for the observed plume(s) as well as
evaluating risk to receptors and risk management options. Techniques for evaluating net
hydraulic gradients in this setting include monitoring of ground-water elevations in piezometers
at relatively high monitoring frequency (e.g., daily) through one or more seasonal cycles and,
potentially, use of dedicated ground-water velocity sensors such as the one produced by
-------
Hydrotechnics, [nc. This probe provides a direct measure of ground-water flow velocity that can
be recorded using an associated data logger. A few such instruments placed within the plume
should provide sufficient data to estimate net gradients.
Addition of certain parameters to the sampling program may provide information to
better assess whether multiple sources are present and whether significant biotransformation of
the vinyl chloride may be occurring. Parameters pertinent to investigations of the natural
attenuation of chlorinated solvent plumes are discussed in detail in Wiedemeier and others
(1998) (Technical Protocol for Evaluating Natural Attenuation of Chlorinated Solvents in
Ground Water, EPA/600/R-98/128, U.S. EPA, Cincinnati, OH). A companion publication
describing.application of-the protocol entitled "Evaluation of the Protocol for Natural
Attenuation of Chlorinated Solvents, £PA/600/R-I>1/025" is also available. These documents
may be obtained free of charge from our website (www.epa.gov/ada/pub5/reports.html). In
particular, measurement of ethene, ethane, and methane may be useful at this site. Comparison
of spatial trends in these data may allow identification of signatures from different sources.
Measurement of dissolved hydrogen in monitoring wells constructed of PVC may be useful in
determining whether the environment for significant dechlorination of the vinyl chloride exists
within the dissolved plume. In addition, measurement of carbon isotope ratios as described in
Bloom and others (2000) (Carbon isotope fractionation during microbial dechlorination of
trichloroethene, cis-l,2-dichloroethene, and vinyl chloride: Implications for assessment of natural
attenuation, Environmental Science and Technology, 34:2768-2772) may provide direct evidence
that reductive dechlorination of contaminants in the plume is presently occurring.
As the conceptual model is refined through incorporation of additional data, use of
mathematical models such as BIOCHLOR (Aziz and others, 2000) may provide insight into the
sensitivity of projected plume behavior to assumptions and ranges of parameter values
incorporated into the conceptual model. This model and, associated documentation are also
available from our website (www.epa.gov/ada/csmos/models.html). However, hydrogeologic,
contaminant, and geochemical data should be obtained to better constrain estimates of ground-
water flow and contaminant attenuation rates in order to allow meaningful evaluations.
It is recommended that a permanent well network for periodic monitoring be established
to provide more detailed information for these evaluations and predictions of plume behavior.
Future studies should continue to use vertical profiling of contaminant distribution to identify
appropriate depths for permanent wells. It is also recommended that additional ground-water
sampling and hydrogeologic investigation occur to better define the plume in the direction of any
other potential receptors, such as the City of Plaquemine wells. Information concerning
hydraulic communication between the upper and lower sand units near the city wells and the
extent of this plume may be used to site permanent monitoring wells to evaluate future plume
migration as well as potential threats to these receptors.
-------
I .
If you have any questions concerning this evaluation or we can be of further service,
please do not hesitate to call us at your convenience (Acree: 580-436-8609, Wilson: 580-436-
8534).
Sincerely,
Steven D. Acree, Hydsologist
Technical Assistance and Technology Transfer Branch
John T. Wilson, PhtD., Microbiologist
Subsurface Remediation Branch
cc: Ric|c Ehrhart, USEPA, Region 6
Jerry Jones
-------
State of Louisiana
Department of Environmental Quality
M J. "MIKE" FOSTER. JR. ,,. n,vl.E CIVENS
GOVERNOR SECRETARY
June 14,2001
Mr. Gregg A. Cooke, Regional Administrator
USEPA Region VI
1445 Ross Avenue, Suite 1200
Dallas, TX 75202-2733
DearMr.Ga^ke: ^^
The Louisiana Department of Environmental Quality (LDEQ) is currently investigating
groundwater contamination that was discovered by the Louisiana Department of Health
and Hospitals in the drinking water wells at the Myrtle Grove Trailer Park, Plaquemine,
Louisiana. LDEQ has collected groundwater samples from numerous private water
wells in the area and has split samples with Dow Chemical Corp. (Plaquemine) on
several new geoprobe type sampling points.
Although the results to date are inconclusive, it appears the main contamination is the
result of some type of chlorinated hydrocarbon and its degenerative daughter products.
Also, the scattering of concentration levels seems to suggest that we are dealing with
non-linear degeneration with the possibility of multiple source locations. These
preliminary findings very closely match the type of chlorinated hydrocarbon groundwater
contamination that was found at England Air Force in Alexandria, Louisiana. With help
from the technical staff (Dr. John Wilson, among others) of EPA Kerr Lab, a plausible
model for the groundwater contamination at England AFB was developed.
With this in mind, LDEQ is requesting technical assistance of EPA Region VI/Kerr Labs
(Dr. John Wilson) regarding contaminant degradation and source identification in the
modeling of the groundwater contamination at the Myrtle Grove Trailer Park site.
W OFFICE OF THE SECRETARY P.O. BOX 82263 BATON ROUGE, LOUISIANA 70884-2263
,„,„. TELEPHONE (504) 765-0741 FAX (504) 765-0746
/oeopapec AN FOU A (.OPPORTUNITY EMPI.OYFR
-------
Mr. Greg Cook
June 14,2001
Page 2
Please have Dr. Wilson or a member of his staff contact David Beatty or Steve Chustz of
my staff to coordinate this project. They may be reached at (225) 765-0585.
Thank you for your assistance.
Sincerely,
J.DaleGiv
Secretary
tbk
c: David Beatty, ETD
Steve Chustz, ETD
Laurie Peacock, ETD
-------
Model Reviews and Refinements
-------
VALIDATION TEST
March 31, 2004
PROCEDURE
Compare modeled and observed potentiometric surfaces
for the range of current conditions
Utilize new set of data not previously used in model
calibration
New data represents high and low river conditions
Data selected was for March 2004 (high) and November
2003 (low)
Data verified through 4th quarter of 2003, and updated
through 3/22/04. 1st quarter 2004 data not yet verified.
-------
VALIDATION DATA (HIGH STAGE)
POTENTIOMETRIC SURFACE (HIGH STAGE)
'///
S11SSS/S// / /
xwx^ /•////
SSX-'/'/'// / / / /
//////IJ
-------
VALIDATION DATA (LOW STAGE)
POTENTIOMETRIC SURFACE (LOW STAGE)
-------
RESULT
• Both validation and calibration data sets
produce similar ground-water flow
directions.
• RMS slightly increased over calibration
data sets
- May result from any recently activated
pumping wells
• Good test result overall.
4
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
1 NATIONAL RISK MANAGEMENT RESEARCH LABORATORY
? GROUNDWATER AND ECOSYSTEMS RESTORATION DIVISION
?* P.O. BOX 1198 • ADA, OK 74820
March 15, 2004
OFFICE OF
MEMORANDUM RESEARCH AND DEVELOPMENT
SUBJECT: Plaquemine Aquifer Contamination, Plaquemine, Louisiana (0 1 RC06-00 1 )
Evaluation and Recommendations for Groundwater Flow Modeling
FROM: Steven D. Acree, Hydrologist '^O^^ f^ — ~ ---
Applied Research & Technical Support Branch
TO: Scott Ellinger
U.S. EPA, Region 6
The transient groundwater flow model provided on February 13, 2004, has been reviewed
by Dr. Mingyu Wang, Rob Earle, and Noman Ahsanuzzaman of Shaw Environmental &
Infrastructure, Inc. Shaw is an on-site contractor providing technical support services to this
laboratory in our Center for Subsurface Modeling Support. This review focused on the technical
adequacy of the flow model construction and calibration. Files related to contaminant transport
and predictive simulations using the calibrated flow model were not provided for review. In
addition, the model was also manipulated by examining and calibrating several important
parameters using a structured approach and the observed data in a particular period. Through this
procedure, it is believed that the calibration error is minimized.
In general, the flow model was constructed using a comprehensive modeling process and
accepted practices. This model fundamentally captures the geologic and hydrogeologic attributes
of the principal modeling components for the upper Plaquemine Aquifer and the aquitards. The
calibrated parameters were able to generate good matching between the observed and calculated
water heads in the upper Plaquemine Aquifer. Incorporation of the recommendations provided
below, including changes in the boundary conditions used for the lower Plaquemine Aquifer,
should result in a product that is useful for estimating the possible range of previous and future
flow conditions within the limitations imposed by the site complexity and available historical
and physical data. Following calibration, it is recommended that modeled and observed
potentiometric surfaces be compared for the range of current river conditions to verify the
accuracy of predictions under the current conditions.
Based on this review and manipulations, the following recommendations are provided.
Please note that some of the following comments address issues related to the inclusion of
significant pumping stresses in predictive simulations. In addition, the model manipulation
results and the observations from those results are attached.
Internet Address (URL) • http://www.epa.gov
Recycled/Recyclable « Printed with Vegetable OH Based Inks on 100% Postconsumer, Process Chlorine Free Recycled Paper
-------
Review Comments and Suggestions
1. Much higher conductance values were specified for the general head boundaries of the
top layer of the upper Plaquemine Aquifer, which had a much smaller layer thickness
(Table 1), than the other layers. This indicates that a much higher average hydraulic
conductivity was assumed for the layer. Unless there is a particular reason to do so, it is
recommended that the same average hydraulic conductivity value be used in all the layers
of the upper Plaquemine Aquifer for the computation of the conductance of the general
head boundaries.
2. In the reviewed flow model, general heads varied with time. Was the general head a
calibrated model parameter? If not, what was the underlying role applied in specifying
the general head values for the general head boundary conditions in the different time
periods? It was noted that the lower general heads for the western general head boundary
were specified corresponding to the higher river stage periods. Is there any particular
reason for that representation? If the varied general head is necessary, it is suggested that
lower general head be specified for periods with lower recharge and higher general head
for periods of increased recharge.
3. The imposition of significant pumping stresses in the lower Plaquemine Aquifer will
necessitate the use of a general head condition for the west and south boundaries of this
aquifer unit in the predictive model that will be constructed based on the calibrated flow
model. Considering that, it is better to specify a general head condition for the west and
south boundaries of the lower Plaquemine Aquifer in the calibration stage. Specifications
of the parameters, including general head and conductance, for those general head
boundaries in the lower Plaquemine Aquifer could refer to those in the corresponding
boundaries in the upper Plaquemine Aquifer. In addition, a specified/constant head
condition is suggested for the east domain boundary of the lower Plaquemine Aquifer
beneath the Mississippi River based on the model manipulation results presented below.
4. It appears that better matching between the observed and calculated water heads in the
lower Plaquemine Aquifer would be obtained with slightly lower specified/constant head
values for the east domain boundary of the aquifer beneath the Mississippi River than the
river stages. However, the flow model with pumping stresses used for long-term
predictions could be specified with a conservative boundary condition compared with the
calibrated flow model to make the results more defensible. For example, for estimation
of a conservative or overestimated transport time (i.e., longer than the actual) from the
northern part of the flow domain to the contamination area of interest, it should be
appropriate to specify the same values for this specified/constant head boundary as the
river stages. In this way, more defensible conclusions may be drawn from the simulation
results.
5. The lower Plaquemine Aquifer (with a large thickness of about 250 ft) was not
discretized into smaller sub-layers in the reviewed model. This should be acceptable for
the model calibration stage in which very limited observation data were available and
pumping stresses were not placed in the aquifer. However, it may be inadequate for
properly simulating the scenarios using pumping wells with different screened intervals
-------
in the model prediction stage. Therefore, it is suggested that at least two more layers be
discretized in the lower Plaquemine Aquifer in the flow model used for predictive
scenarios. The screened intervals of the most significant pumping wells in the lower
Plaquemine aquifer should be considered during the specification of these layers.
6. After the model calibrations are completed, the inactive cells should be activated for all
active layers in the southeast corner of the model domain to reduce the boundary effect
for the scenarios with pumping in the predictive simulations.
7. The principal uncertainties in the model include the specification of appropriate boundary
conditions and historical groundwater extraction rates. It is recommended that predictive
simulations be run using the probable range of conditions to evaluate the impact of these
uncertainties on the model results.
If you have any questions concerning these recommendations, please do not hesitate to
call me at your convenience (580-436-8609). We look forward to future interactions with
you concerning this and other sites.
Attachment
cc: Nancy Pagan, Region 6 (without attachment)
-------
Table 1. Layer structure and attributes obtained from the Visual MODFLOW files of the
reviewed flow model
Layer
1
2
3
4
5
6
7
8
9
10
Layer
Feature
Aquitard
Upper
Plaquemine
Aquifer
Upper
Plaquemine
Aquifer
Upper
Plaquemine
Aquifer
Upper
Plaquemine
Aquifer
Upper
Plaquemine
Aquifer
Upper
Plaquemine
Aquifer
Aquitard
Lower
Plaquemine
Aquifer
Inactive
Layer
Conductance
of West GH
Boundary
(ft2/day)
380.5
67.14
78.34
78.34
72.7
83.9
General
Head
Specified
on West
GH (ft)
,538.8-
542.7
538.8-
542.7
538.8-
542.7
538.8-
542.7
538.8-
542.7
538.8-
542.7
Conductance
of South GH
Boundary
(ft2/day)
150
26.5
30.9
30.9
28.7
33.1
General
Head
Specified
on South
GH (ft)
540-556
541.2-
557.2
541.7-
557.7
541.2-
557,2
540.8-
556.8
540.8-
556.8
Thickness
(ft)
100.2
4
16
18.7
18.7
17.4
20
49.4
245.8
69.8
Elevation
of the top
of the
layer (ft)
560
459.8
455.8
439.8
421.1
402.4
385
365
315.6
69.8
-------
ATTACHMENT
Model Manipulations
Objectives
To examine model calibrations using a transient flow condition, and compare calibration results.
Procedure
1) Add general head boundary conditions to the west and south flow boundaries for the
lower Plaquemine Aquifer.
2) Select the time period from t=350 to t=500 days in which the hydraulic head was
relatively stable for a long period and included high peak and low minimum values.
3) Assess the maximum distance west of the Mississippi River at which significant effects
of river stage fluctuation on groundwater head may exist. It was estimated that this
distance may be less than 20,000 ft from the Mississippi River based on the observed
groundwater head data.
4) Estimate the general head value to be applied for the selected calibration period based on
hydraulic gradient using general head reference points 10,000 ft outward from the
domain boundaries.
5) Manipulate the flow model by examining and calibrating the model parameters including
the boundary conditions of the east domain boundary of the lower Plaquemine Aquifer,
the hydraulic conductivity (K), specific storage (Ss), vertical hydraulic conductivity of
the aquitard between the upper and lower Plaquemine Aquifer units, and conductance of
the general head boundary. This can be done through matching the observed and
calculated water heads for both the peak and minimum parts of the hydrograph assuming
that the general head remained unchanged in this period. The reason for using the shorter
time period to calibrate the model, instead of the whole observation period, was to reduce
the "noise" from the seasonal fluctuation of regional groundwater head which was
difficult to determine and to which the variation of the general head was related.
Therefore, the calibrations of other important parameters, particularly the hydraulic
conductivity, could be focused by minimizing disturbances related to potential changes in
the general head during the relatively short time period.
Results
The locations of the observation points for groundwater head are shown in Figure 1.
Figures 2-15 present the model manipulation and calibration results. It should be noted that
Figure 2 shows a better match, while Figures 6 and 7 also exhibit good matching between the
observed and calculated water heads. In addition, the comparisons between the observed and
calculated water heads are presented for the time period between t=350 and 500 days except for
Figs.2j through 2m which show the match between the observed and calculated water heads for
-------
the whole observation period. Data from the entire period was used to examine whether a
constant general head may be appropriate in a long-term prediction model.
Please note that the screened intervals of the observation wells used in this analysis were
located in the upper Plaquemine Aquifer except for wells SW-1D and SW-3D which are
screened in the lower Plaquemine Aquifer. In addition, all results presented in the figures below
were obtained with a specified/constant head condition on the east boundary of the lower
Plaquemine Aquifer except for Figures 14 and 15 in which a no-flow condition was placed on
the aforementioned boundary. Furthermore, the same water head was specified for the
specified/constant head boundary condition of both the upper and lower Plaquemine Aquifer
units based on the Mississippi River stages.
Observations
1. The model simulation results were sensitive to the examined model parameters, including
the hydraulic conductivity, specific storage, vertical hydraulic conductivity of the
aquitard between the upper and lower Plaquemine Aquifer units, and conductance and
general head of the general head boundary for the selected time period. This was a critical
model response for adequate model calibrations. Therefore, it is believed that a transient
flow condition provides a good capability for calibrating the flow model by implementing
an appropriate calibration approach.
2. It appears that a specified/constant head would be appropriate for the boundary condition
of the east domain boundary of the lower Plaquemine Aquifer beneath the Mississippi
River based on the observed water head in the aquifer. This may be particularly important
for representations of the Plaquemine Aquifer in the long-term prediction model
incorporating pumping stresses. However, it appears that the calibration results for the
upper Plaquemine Aquifer, which do not include groundwater extraction, would not be
very dependent on the lower Plaquemine Aquifer.
3. There was a good match between the observed and calculated water heads in all the
presented scenarios at observation point PZ-27, located near the Mississippi River, and at
which the groundwater head was dominantly controlled by the river stage. This further
confirmed that the constant/specified head boundary is appropriate for the river
characterization.
4. Although the scenario with a horizontal hydraulic conductivity (K) of 120 ft/day for the
upper and lower Plaquemine Aquifer displays a better match between the observed and
calculated groundwater heads (Fig. 2), the matching was also good in the other two
scenarios with horizontal K=80 and 160 ft/day for the Plaquemine Aquifer (Figs. 6 and
7). A hydraulic conductivity value of 160 ft/day was used in the reviewed model. Based
on these results, a model uncertainty analysis should be performed using the predictive
model. A range from approximately 80 to 160 ft/day could be considered for the
uncertainty analysis with respect to the horizontal hydraulic conductivity of the upper
Plaquemine Aquifer.
-------
5. In those scenarios with a good match between the observed and calculated groundwater
heads (Figs. 2, 6, and 7), the same specific storage and hydraulic conductivity (horizontal
and vertical) values for the aquitard between the upper and lower Plaquemine Aquifer
layers were specified as those used in the reviewed model.
6. A good match between the observed and calculated groundwater heads could be achieved
by applying the fixed general head in different periods or seasons (Figs. 2j-2m), while it
is understood that a better match could be obtained by adjusting the general head in
different periods or seasons.
-------
11860
Fig. 1. Locations of observation points for groundwater head.
-------
Fig. 2a. Observation point PZ-27.
Fig. 2. Comparisons between the observed and calculated water heads for the scenario:
aquifer horizontal K=120 ft/day, vertical K=12 ft/day, and Ss=0.0001/ft; aquitard horizontal
K=0.05 ft/day and vertical K=0.005 ft/day; general head of the west boundary=538 ft; general
head of the south boundary=544 ft; conductance of the west general head boundary=l 11.4
ft2/day for a west/east cell face with area=3,713 ft2 (layer 3); and conductance of the south
general head boundary=19 ft2/day for a north/south cell face with area=3,800 ft2 (layer 3). Figs.
2a through 2i use data from t=300-500 days. Figs. 2j through 2m use data from t=0-580 days.
Note:
Figs. 2a through 2m: A good match between the observed and calculated groundwater
heads was obtained with aquifer horizontal K=120 ft/day, vertical K=12 ft/day, and
Ss=0.0001/ft; aquitard horizontal K=0.05 ft/day and vertical K=0.005 ft/day; general head
of the west boundary=538 ft; and general head of the south boundary=544 ft.
-------
Fig. 2b. Observation point PZ-32.
-------
Fig. 2c. Observation point PZ-34,
-------
Fig. 2d. Observation point PZ-36.
-------
Fig. 2e. Observation point PZ-37.
-------
Fig. 2f. Observation point SW-2S.
10
-------
Fig. 2g. Observation point SW-4S.
11
-------
Fig. 2h. Observation point SW-1D.
12
-------
Fig. 2i. Observation point SW-3D.
13
-------
Fig. 2j. Observation point PZ-27.
14
-------
Fig. 2k. Observation point PZ-32.
15
-------
Fig. 21. Observation point PZ-34.
16
-------
Fig. 2m. Observation point PZ-36.
17
-------
2000
1000
10000
118(50
Fig. 3a. Layer 3.
Fig. 3 Contours of simulated water head and flow directions (t=434 days, at hydrographic peak)
in the Plaquemine Aquifer from the same scenario as that used for Fig. 2 (aquifer horizontal
K=120 ft/day, vertical K=12 ft/day, and Ss=0.0001/ft; aquitard horizontal K=0.05 ft/day and
vertical K=0.005 ft/day; general head of the west boundary=538 ft; general head of the south
boundary=544 ft; conductance of the west general head boundary=111.4 ft2/day for a west/east
cell face with area=3,713 ft2; and conductance of the south general head boundary=19 ft2/day for
a north/south cell face with area=3,800 ft2).
18
-------
2000
4000
£000
118(50
Fig.3b. Layer?.
19
-------
8000
10000
11860
Fig. 3c. Layer 9.
20
-------
11860
Fig. 4a. Layer 3.
Fig. 4. Contours of simulated water head and flow directions (t=454 days, a time point between
the hydrographic peak and valley) in the Plaquemine Aquifer from the same scenario as that used
for Fig. 2 (aquifer horizontal K=120 ft/day, vertical K=12 ft/day, and Ss=0.0001/ft; aquitard
horizontal K=0.05 ft/day and vertical K=0.005 ft/day; general head of the west boundary=538 ft;
general head of the south boundary=544 ft; conductance of the west general head
boundary=111.4 ft2/day for a west/east cell face with area=3,713 ft2; and conductance of the
south general head boundary=19 ft2/day for a north/south cell face with area=3,800 ft2).
21
-------
8000
10000 11860
Fig.4b. Layer?.
22
-------
11860
Fig. 4c. Layer 9.
23
-------
8000
10000
11860
Fig. 5 a. Layer 3.
Fig. 5 Contours of simulated water head and flow directions (t=469 days, at hydrographic valley)
in the Plaquemine Aquifer from the same scenario as that used for Fig. 2 (aquifer horizontal
K=120 ft/day, vertical K=12 ft/day, and Ss=0.0001/ft; aquitard horizontal K=0.05 ft/day and
vertical K=0.005 ft/day; general head of the west boundary=538 ft; general head of the south
boundary=544 ft; conductance of the west general head boundary=l 11.4 ft2/day for a west/east
cell face with area=3,713 ft2; and conductance of the south general head boundary=19 ft2/day for
a north/south cell face with area=3,800 ft2).
24
-------
11860
Fig. 5b. Layer 7.
25
-------
11860
Fig. 5c. Layer 9.
26
-------
Fig. 6a. Observation point PZ-27.
Fig. 6. Comparisons between the observed and calculated water heads at six observation wells
for the scenario: aquifer horizontal K=160 ft/day and vertical K=16 ft/day, Ss=0.0001/ft;
aquitard horizontal K=0.05 ft/day and vertical K=0.005 ft/day; general head of the west
boundary=538 ft; general head of the south boundary=544 ft; conductance of the west general
head boundary=l 11.4 ft2/day for a west/east cell face with area=3,713 ft2 (layer 3); and
conductance of the south general head boundary=19 ft2/day for a north/south cell face with
area=3,800 ft2 (layer 3).
27
-------
Fig. 6b. Observation point PZ-32.
28
-------
Fig. 6c. Observation point PZ-34.
29
-------
Fig. 6d. Observation point PZ-36.
30
-------
Fig. 6e. Observation point SW-2S.
31
-------
Fig. 6f. Observation point SW-1D.
32
-------
Fig. 7 a. Observation point PZ-27.
Fig. 7. Comparisons between the observed and calculated water heads at six observation wells
for the scenario: aquifer horizontal K=80 ft/day and vertical K=8 ft/day, Ss=0.0001/ft;
aquitard horizontal K=0.05 ft/day and vertical K=0.005 ft/day; general head of the west
boundary=538 ft; general head of the south boundary=544 ft; conductance of the west general
head boundary=l 11.4 ft2/day for a west/east cell face with area=3,713 ft2 (layer 3); and
conductance of the south general head boundary=19 ft2/day for a north/south cell face with
area=3,800 ft2 (layer 3).
33
-------
Fig. 7b. Observation point PZ-32.
34
-------
Fig. 7c. Observation point PZ-34.
35
-------
Fig. 7d. Observation point PZ-36.
36
-------
Fig. 7e. Observation point SW-2S.
37
-------
Fig. 7f. Observation point SW-1D.
38
-------
Fig. 8a. Observation well PZ-27.
Fig. 8. Comparisons between the observed and calculated water heads at six observation wells
for the scenario: aquifer horizontal K=120 ft/day and vertical K=12 ft/day, Ss=0.0005/ft;
aquitard horizontal K=0.05 ft/day and vertical K=0.005 ft/day; general head of the west
boundary=538 ft; general head of the south boundary=544 ft; conductance of the west general
head boundary=111.4 ft2/day for a west/east cell face with area=3,713 ft2 (layer 3); and
conductance of the south general head boundary=19 ft2/day for a north/south cell face with
area=3,800 ft2 (layer 3).
39
-------
Fig. 8b. Observation well PZ-32.
40
-------
Fig. 8c. Observation well PZ-34.
41
-------
Fig. 8d. Observation well PZ-36.
42
-------
Fig. 8e. Observation well SW-2S.
43
-------
Fig. 8f. Observation well SW-1D.
44
-------
Fig. 9a. Observation point PZ-27.
Fig. 9. Comparisons between the observed and calculated water heads at six observation wells
for the scenario: aquifer horizontal K=120 ft/day and vertical K=12 ft/day, Ss=0.00001/ft;
aquitard horizontal K=0.05 ft/day and vertical K=0.005 ft/day; general head of the west
boundary=538 ft; general head of the south boundary=544 ft; conductance of the west general
head boundary=111.4 ft2/day for a west/east cell face with area=3,713 ft2 (layer 3); and
conductance of the south general head boundary=19 ft2/day for a north/south cell face with
area=3,800 ft2 (layer 3).
45
-------
Fig. 9b. Observation point PZ-32.
46
-------
Fig. 9c. Observation point PZ-34.
47
-------
Fig. 9d. Observation point PZ-36.
48
-------
Fig. 9e. Observation point SW-2S.
49
-------
Fig. 9f. Observation point SW-1D.
50
-------
Fig. lOa. Observation point PZ-27.
Fig. 10. Comparisons between the observed and calculated water heads at six observation wells
for the scenario: aquifer horizontal K=120 ft/day and vertical K=12 ft/day, Ss=0.0001/ft;
aquitard horizontal K=0.5 ft/day and vertical K=0.05 ft/day; general head of the west
boundary=538 ft; general head of the south boundary=544 ft; conductance of the west general
head boundary=l 11.4 ft2/day for a west/east cell face with area=3,713 ft2 (layer 3); and
conductance of the south general head boundary=19 ft2/day for a north/south cell face with
area=3,800 ft2 (layer 3).
51
-------
Fig. lOb. Observation point PZ-32.
52
-------
Fig. lOc. Observation point PZ-34.
53
-------
Fig. lOd. Observation point PZ-36.
54
-------
Fig. lOe. Observation point SW-2S.
55
-------
Fig. lOf. Observation point SW-1D.
56
-------
Fig. 11 a. Observation point PZ-27.
Fig. 11. Comparisons between the observed and calculated water heads at six observation wells
for the scenario: aquifer horizontal K=120 ft/day and vertical K=12 ft/day, Ss=0.0001/ft;
aquitard horizontal K=0.05 ft/day and vertical K=0.005 ft/day; general head of the west
boundary=543 ft; general head of the south boundary=544 ft; conductance of the west general
head boundary=111.4 ft2/day for a west/east cell face with area=3,713 ft2 (layer 3); and
conductance of the south general head boundary=19 ft2/day for a north/south cell face with
area=3,800 ft2 (layer 3).
57
-------
Fig. 1 Ib. Observation point PZ-32.
58
-------
Fig. 1 Ic. Observation point PZ-34.
59
-------
Fig. lid. Observation point PZ-36.
60
-------
Fig. lie. Observation point SW-2S.
61
-------
Fig. 1 If. Observation point SW-ID.
62
-------
Fig. 12a. Observation point PZ-27.
Fig. 12. Comparisons between the observed and calculated water heads at six wells for the
scenario: east boundary condition specified as no-flow instead of specified head; aquifer
horizontal K=120 ft/day and vertical K=12 ft/day, Ss=0.0001/ft; aquitard horizontal K=0.05
ft/day and vertical K=0.005 ft/day; general head of the west boundary=538 ft; general head of
the south boundary=544 ft; conductance of the west general head boundary=l 11.4 ft2/day for a
west/east cell face with area=3,713 ft2 (layer 3); and conductance of the south general head
boundary=19 ft2/day for a north/south cell face with area=3,800 ft2 (layer 3).
63
-------
Fig. 12b. Observation point PZ-32.
64
-------
Fig. 12c. Observation point PZ-34.
65
-------
Fig. 12d. Observation point PZ-36.
66
-------
Fig. 12e. Observation point SW-2S.
67
-------
Fig. 12f. Observation point SW-1D.
68
-------
SW-1D
& SW-2S
Tested hydraulic
connection window
between upper and lower
Plaquemine aquifers
Fig. 13. Location of hydraulic connection window between the upper and lower Plaquemine
Aquifer units.
69
-------
Fig. 14a. Observation point PZ-27.
Fig. 14 Comparisons between the observed and calculated water heads at six observation wells
for the scenario: hydraulic connection window between the upper and lower Plaquemine
Aquifers with horizontal K=120 ft/day, vertical K=12 ft/day and no-flow condition for the
east boundary of the lower Plaquemine Aquifer; aquifer horizontal K=120 ft/day and vertical
K=12 ft/day, Ss=0.0001/ft; aquitard horizontal K=0.05 ft/day and vertical K=0.005 ft/day;
general head of the west boundary=538 ft; general head of the south boundary=544 ft;
conductance of the west general head boundary=l 11.4 ft2/day for a west/east cell face with
area=3,713 ft2 (layer 3); and conductance of the south general head boundary=19 ft2/day for a
north/south cell face with area=3,800 ft2 (layer 3).
70
-------
Fig. 14fr Observation point PZ-32.
71
-------
Fig. 14c. Observation point PZ-34.
72
-------
Fig. 14d. Observation point PZ-36.
73
-------
Fig. 14e. Observation point SW-2S.
74
-------
Fig. 14f. Observation point SW-1D.
75
-------
Fig. 15a. Observation point PZ-27.
Fig. 15. Comparisons between the observed and calculated water heads at six wells for the
scenario: placement of a hydraulic connection window between the upper and lower
Plaquemine Aquifers with horizontal K=120 ft/day, vertical K=120 ft/day, and no-flow
condition for the east boundary of the lower Plaquemine Aquifer; aquifer horizontal K=120
ft/day and vertical K=12 ft/day, Ss=0.0001/ft; aquitard horizontal K=0.05 ft/day and vertical
K=0.005 ft/day; general head of the west boundary=538 ft; general head of the south
boundary=544 ft; conductance of the west general head boundary=l 11.4 ft2/day for a west/east
cell face with area=3,713 ft2 (layer 3); and conductance of the south general head boundary=19
ft2/day for a north/south cell face with area=3,800 ft2 (layer 3).
76
-------
Fig. 15b. Observation point PZ-32.
77
-------
Fig. 15c. Observation point PZ-34.
78
-------
Fig. 15d. Observation point PZ-36.
79
-------
Fig. 15e. Observation point SW-2S.
80
-------
Fig. 15f. Observation point SW-1D.
81
-------
testing Propose^
Refinements on Tran^j6ht
Flo
Test Approach
General arder of teating:
1, ;Rivef to Constant Head Bdundairy
2. Conductivity Changes
3." General Head Bounilary
4. Grid and Layer Reifinernefit
5.; Particle Path Travel Time Check
6* Pumping Weil 2one of Influence
-------
Observations & Conclusidns
flow directions fy^
even with suggested modifications (in
non-pumping s6|nafid)
^Adjusted K-yaJues reStilted jn mc)fe
reasonable particle s Iragkjiig
Frdrn previous Bf^ld wn S
-PurhfDiing well zone of infiuence^appearsv
reduced
©bse rvations &
o stops
wher^mitstitutecllor rlyerpouhda^ ?
boundary geometry jslrriportant
General head iounclary on south domain
helps Calibrate j$^
No apjsiarent different results from grid and
layer refinement
-------
Methodology
• Slides A1-A2
- Shows jOrigin^l Bruits fpr£orr»parison
- Includes 4-r6presetttatiye lE^u|ppten|ial Maps
Includes 4^represeiltative fl
-------
Slide A2: Original Time^S0ries
Methodology
>lidesB1-B2;
coristant ..'
changes according totfeid rneasyfements
'• Cortstant head extends 'into Haquetpine'faqtiffer
Conductivity adjustment
. » Kxy upper, S/and I units 275, ^Kz^SS
;• Kxy elayilit = .0028,"Kz = ,0(?628
Wo ftoyv (inactive cells) east elf river all layers
(Sritjl cell jsizethe same (236' #23§')
isayerifig f pr rnajpr stratigraphy only; except eliminated layer
3 ancii fcombined layers 4'iind 5
No pumpage
-------
Slide 81 :Equipotentials
-------
Slides C1-C2
-i Same as 81 ^
• Kxy uppfer sandjunit-;25Q;HKz = 25
Ky clay/slit ^0^5;jKzi|).
Slide Siiifeg ulpoteiitJals
Stress Period 80
6
-------
Methodology
Slides D1-D
fCdecreasecl £gan \r\ tipper sand juriit
.~
0.00$
-------
§lide D "1; Egu(potentials
Stress Period 80
Stress Period 160
Slide O2S Tirniaseries
-------
• Slides E1-E2
• Test^or increased K Jn gjay/stit
-Includes overlying patyralJeyee cjejDosits
- 1neli4deS||ay/slit layer between upper and
lower
PfOduees anomalous results; poor
«_•_• •.• =__
-------
IVlethoidology
Added on Souh Domain
Set as transient bpuridary
As matter of testing practicality-no curveature added
Inactive sells^addeid;east;of^;Pz ^7-^8 {§p(J!|yhieastern
quadrant) and south of river
readjusted;
Kxy ypper sand unit.?? 169. ;Kz = 15
Kxy clay/slit = 0.05,' Kz^ 0.005
10
-------
11
-------
Methodology
• Slides X31-G2
•Grid $nd1_ay§
- ;Five Additional l^^rs |c>r Upper ^nd;Unit
- Grid Celt Size p0|r|a5ed Iroiii isise' ^^3S') Jo (60
x 6Q^)QV6r irtssn Sfsa of ?ifit6ir0st
«'jJnsuffjpienfAyatlable M|rnpry for Mpd.path at 60 X 60
• AbhoirMa! Huni T^rfnination
Size Changedlo 118' X 118'
Modpath Run Compiet
12
-------
Slide G2: Time-Series
.'•. /pDP3
,1 / '•• ! l\ '
,\l
letNodcilogy
1: Parte Velocity Ch
Gomparisori between 0,i new parc
tracing veloities
Time markers ;set at every
New results show marKecl deer0a$^ in
|j1}iec3 grond-waerjyppjty
13
-------
Slide HI: Particle Velpciiy
New Particle Velocity
X=717
7174-7099 / 50 = 1.5 ft/d
(probably represents max
particle velocity)
V Slides jJj-J2
- Xestiinfluenee^f-Pumping Wells {visual only)
•Slide ill
- Punjping 3-rnuniQJpal we^s^l^QO gp,ni,co
fpr Entire 600-clay run ffi
•Slide J2
- Slide ^:pr©sents;i0bservati
nearest tH^
effects £t stance rorrj
14
-------
15
-------
Testing Proposed
Refinements on Day-364
Steady-State Flow Model
January 28, 2004
Methodology
• Slide A: Original flow result (for comparison)
• Slide B: Grid refinement to 20 x 20
• Slide C:
- Kx,Ky=175;Kz = 17.5
- Layering for major stratigraphy only (eliminated layer 3,
combined layers 4 and 5)
- No flow boundaries east of river in each layer
- GHB same
- River boundary same
- No,pumpage
-------
Methodology
Slide D:
- Slide C conditions except for;
- GHB geometry changed to represent river profile,
same head and conductance value
- Calibration for wells 35 and 36 appear further off
due to GHB profile change
Slide E:
- Cross section (zoomed) showing layer refinement
based on well screen intervals
- Few gradational boundary change layers added
Methodology
• Slide F:
- Flow result with added layer refinement
• Slide G:
- River boundary converted to constant head
- Constant head entered as line decreasing head
downstream
-------
Methodology
Slide H:
- Same as slide G except;
-Kx,Ky=160;Kz = 30
Slide I:
- Same as slide H except;
-Kx, Ky=175,Kz = 25
Test Observations
(NO Pumpage)
Changing river to constant head shows little effect on flow
Addition of no flow boundaries shows little effect on flow
GHB profile change makes calibrating wells 35 and 36 more
difficult
Model relatively sensitive to K, especially at lower ranges
Layer refinement shows little effect on flow (but may with
pumpage) given that only approximate bulk conductivity values
are available
-------
Test Observations
(No Pumpage)
Overall, no dramatic change in flow
direction
Gradients appear practically the same
Water wells more readily calibrated with
original boundary configuration
-------
-------
-------
-------
-------
Steven Acree To: Scott Ellinger/R6/USEPA/US@ EPA
01/14/2004 05-05 PM CC: NanCy Fagan/R6/USEPA/US@EPA, Mingyu
u i rivi Wang/ADA/USEPA/USOEPA, Rob Earle/ADA/USEPA/US@EPA, Abu
Ahsanuzzaman/ADA/USEPA/US @ EPA
Subject Modeling Recommendations
Scott,
The updated flow model has been reviewed by Dr Mingyu Wang, Rob Earle, and Noman Ahsanuzzaman
of Shaw Environmental & Infrastructure, Inc. Shaw is a contractor providing technical support services to
this laboratory. The following comments and recommendations are provided for your consideration.
1. Several important factors should be considered in determining the layer discretization for a
numerical groundwater model construction. One of them is the distribution of the depths of the intervals for
the groundwater sampling and water level measurements. Whether the current layer discretization is
adequate for the Plaquemine Aquifer, particularly, the upper Plaquemine sand aquifer, should be further
examined. Specifically, since most of the groundwater head measurements were made using piezometers,
which represent water head for small intervals, the layer discretization should effectively correspond with
the different measurement depths of those piezometers. Particularly, more attention should be paid to
these areas close to the river where the vertical flow component could be significant. Moreover, the layer
thickness should not sharply change, if possible^ between the neighboring layers. In the reviewed model,
the upper Plaquemine sand aquifer was divided into three layers labeled as layers 3, 4 and 5 with layer
thickness values of 5.6, 42.6, 46.8 ft, respectively. The thickness difference between layers 3 and 4
appears to be too large. In addition, if computation resources and time are allowable, a layer discretization
with more than three layers should be considered for the upper Rlaquemine sand aquifer and the layer
thickness should be accordingly adjusted appropriately (for example, five layers with equal thickness or
gradually increasing thickness downward).
2. A river boundary condition was applied to only layers 1 and 2 in the model. If the river package
was considered appropriate to handle the Mississippi River, the river boundary condition should be
specified also for layer 3 (even layer 4 if the thickness of layer 3 is small), which is based on the statement
that "the Mississippi River cuts through the top stratum and into the aquifer upper sand unit, to a depth of
approximately 100 feet below land surface" (Ellinger et al., 2003, page 39).
3. Based on the contact between the river and the upper Plaquemine aquifer (see recommendation
2) and available hydraulic head measurements, it appears that it would be appropriate and easier to treat
the river as a constant (specified) head boundary. Accordingly, the calibration of the river conductance
would not be needed. The constant or specified head condition would be applied to the top three layers
including the top layer of the upper Plaquemine sand aquifer. The appropriateness of considering the river
as a constant or specified head boundary is further supported by the observation that the hydraulic
gradient between the river (Dockl) and PZ27 is similar to that between PZ28 and PZ31 (Fig. 1). It should
be noted that the average head difference between PZ28 and PZ31 was approximately three times as
much as that between Dockl and PZ27, while the distance between PZ28 and PZ31 is about three times
as long as that between Dockl and PZ27. In fact, the distance betweerr PZ27 and the actual hydraulic
contact point between the river bottom and the upper Plaquemine aquifer could be larger than that
between Dockl and PZ27. These observations indicate that a direct hydraulic contact exists between the
river and the aquifer. Although the Mississippi River could be alternatively treated as a river boundary, as
in this model, river conductance would need to be calibrated and should be very large. It should be noted
that all calibrated parameters include errors or uncertainty. Thus, eliminating the need for calibration of a
parameter may enhance the calibrations of the critical model parameters, such as hydraulic conductivity.
Therefore, treating the river as a specified/constant head boundary, instead of a river boundary, appears
to be suitable and defensible.
4. In the reviewed model, the general head boundary condition was applied to layers 3 and 4. and a
-------
no flow condition was applied to all other layers along the west boundary. In addition, the no flow condition
was applied to all layers of the north and south boundaries. It is agreed that it could be suitable to apply
the no flow condition to all layers along the north boundary. However, it is suggested that the south
boundary be specified as a general head condition for all layers of the upper Plaquemine sand aquifer.
Also, the regional groundwater fluctuation far from the Mississippi River, such as at the Safety Kleen
facility, should be examined to properly select the general head values. The general head could be varied
spatially but should correspond to the slope of the river stage. However, it may be fixed with time by
appropriately selecting the reference head. Please note that it is better to appropriately specify the
reference head based on the regional groundwater level and calibrate the conductance that could be
changeable spatially but fixed with time for the general head boundary. Moreover, it is suggested that all
layers, other than those with the constant head or river boundary, be specified as no flow boundary along
the Mississippi River within the current model domain. Furthermore, it would be suitable to place the
domain boundary around the southwest corner of the model domain with a profile corresponding to the
river course to facilitate specification of the reference head and calibrate the conductance for the general
head boundary. Please refer to Figure 2 below for the suggestions on the model boundary conditions.
5. It is noted that the calibrated parameters in the reviewed model included the general head, river
conductance, and specific storage. It is suggested that in addition to the conductance for the general head
boundary and specific storage, the model calibration should focus on the hydraulic conductivity field.
Zonation of the hydraulic conductivity field should be considered based on any available hydraulic tests or
other information such as borehole soil or geophysical logs. The calibration of hydraulic conductivity should
include the horizontal and vertical hydraulic conductivities, particularly, the horizontal hydraulic conductivity
for the upper Plaquemine aquifer and vertical hydraulic conductivity for the aquitard between the upper
and lower Plaquemine sand aquifers, if sufficient data are available. If it is possible to calibrate the vertical
hydraulic conductivity, the continuity of the aquitard could be evaluated, which may be very important in
controlling the hydraulic connection of the groundwater system.
6. The hydraulic conductivity value specified for the aquifers in the reviewed model appears to be too
high (500 ft/day) based on the fact that the geometric mean of hydraulic conductivity for the permeable
zone A that includes the Plaquemine aquifers was estimated as 101 ft/d and the mathematic mean 163 ft/d
during previous investigations (Ellingeret al., 2003, page 26).
7. It is reasonable to first calibrate the steady state flow models and then the transient flow models as
was performed in this study. However, it should be noted for the calibrations of the steady state flow
models that it is critical to appropriately select typical periods in which the flow field may represent an
approximate steady state condition. Based on the observed groundwater head variation with time, it is
possible to assess which periods could be roughly considered as steady state flow and then determine
representatively the average water head for steady state model calibrations. Figure 3 shows a
groundwater head hydrograph for PZ27 from which groundwater flow with an approximate steady state in
some periods can be judged, for example, Period A. It can be seen that groundwater head in this period
exhibited relatively little fluctuation.
8. It is recommended that the following modeling process be followed: 1) Select the proper periods
representing steady state flow; 2) Calibrate steady state flow models; 3) Determine the initial head based
on a steady state run immediately before a transient model calibration; 4) Perform transient flow model
calibrations; 5) Validate the calibrated model using the data that are not used in the model calibrations; 6)
Conduct sensitivity analyses; 7) Estimate the typical variation pattern or historic changes of the river
stages in the past several decades or longer, and construct the past flow field beginning at a time point of
interest. This constructed flow field would be a foundation for further investigations of the flow paths and
contaminant fate and transport.
9. Sensitivity analyses should be performed for evaluating the importance and model uncertainty of
different parameters and boundary conditions, such as vertical recharge, hydraulic conductivity, general
head and its corresponding conductance. Specifically, the effect of different boundary conditions along the
-------
upper right boundary and the top left boundary on model calibration and simulation results should be
carefully examined.
10. It is understood that the large-scale groundwater extraction did not take place within the
Plaquemine Aquifer during the calibration periods. However, extraction may have played a significant role
in controlling the groundwater flow field for some periods in the past. Understanding the attributes of these
man-made stresses, such as pumping rates and depths of the screened intervals, will be helpful in
determining the layer discretization for the groundwater modeling and constructing the past groundwater
flow field, if contaminant fate and transport simulations are planned.
Reference
Ellinger, Scott, Nancy Pagan, James Harris and Eric Adidas. USEPA Region 6, September 29, 2003.
TECHNICAL BACKGROUND DOCUMENT AND CONCEPTUAL MODEL POTENTIAL
GROUND-WATER FLOW DIRECTIONS AND CONTAMINANT FATE AND TRANSPORT IN THE
PLAQUEMINE AQUIFER OF IBERVILLE PARISH AND WEST BATON ROUGE PARISH, LOUISIANA.
Steven D. Acree, Hydrologist
Robert S. Kerr Environmental Research Center
P.O. Box 1198 / 919 Kerr Research prive
Ada, OK 74821
(580) 436-8609 (voice) / (580) 436-8614 (FAX)
-------
October 30
Improving the Transient
Calibration
Part 1
Issues with Initial Calibration
Poor initial heads used
<* Constant heads used throughout layers
•* Should reflect aquifer conditions
Models predictions lags' observation data
<* Indication of storage being too large
-------
Steps to Improve Calibration
Adjust initial heads
^ Run model in steady state using boundaries from
first stress period
•* Save steady state heads
* Run model in transient mode using steady state
heads as initial conditions
This improved the early time calibration, but
there was still a pronounced lag
-------
Steps to Improve Calibration
To reduce the lag in the model I reduced
specific storage
^ Iterative process / run and check calibration
A value of 0.0001 seemed to work well
Improved Transient Calibration
I like to see time-series graphs for transient
calibrations
The following slides show the original time-
series graph and then the improved graph
•* The model appears to match observations at all
wells reasonably
* In the graphs the line with small boxes is the
model prediction / the large boxes are observed
water levels
-------
J* i* £•• - -—- a-
& a *«' a - e
-"'""^™
|lH|
wj wUS
}
I
i<
1 ?.'*.*. \ -', K «6^EB
li
J\ 1
•i s .n
. ! -
MJ V
1 Mi
• V
i <'
•
• •
A
»A_
HMdmTinM
. '
fill
J '.. «j
p ill
-
t ?,
\ '1 '
i •,
i 1 |
J 1 1
1 • J J
X« » 7 "!•/
7*.-^ *b] .»• t
-J ^
: j
^ r%.
\ i
y
5
1
"•
-------
Improved Calibration
-------
Initial Calibration
Improved Calibration
-------
Initial Calibration
Improved Calibration
?£i -•• !j - ' .-H-d«™»
~S!s ]!' '
a ******* i.
?
a
8.
•
,ff -jii
l\-i\
/i ; W i 1, i!
p ,\ •• i
i • Ai\;
J •>
'
A '* J
A .. v- / i i
/\ i
j \ •• : -:
:.i
-------
Improved Calibration
8
-------
Transient model matches time series data
well
Model boundary conditions and hydraulic
conductivity values were not changed to
improve calibration
We should go through the model together
early next week
17
-------
I Waterloo
| hydrogeologic
SOFTWARE • CONSULTING • TRAINING
October 23
Improvement of Calibration for
185 day Case
Initial Calibration
Poor match to overall gradient
Gradient too small
Heads low near river, heads high near
western boundary
-------
Initial Calibration
Improving the Calibration
Need increased heads near the River
* Increase the connection between the river and the
groundwater system
>< Increase the conductance in the River Boundary
> For this case, a value of 13000 ft2/day (original
value = 2592 ft2/day)
-------
Improving the Calibration
Need lower heads near western boundary
Decrease reference heads along boundary
Original: 544 to 545 feet
•* New reference heads: 539 to 538 feet
New Equipotentials / Flow Directions
-------
New Calibration Graph
Calibration Comments
Gradient well matched
Absolute mean residual = 0.3 feet
Normalized RMS = 7%
Mean error = -0.002 feet
Overall, very good calibration statistics
-------
Velocity Estimate
Darcy Flux = (K) x (Gradient) = 1 ft/day
Velocity = Darcy Flux / porosity = 3.6 ft/day
Calculation completed at river near well NN5
Comments
No recharge in model
Transient calibration should be considered
Model is well behaved - converges quickly
May need more discretization for transport
-------
October 23
Improvement of Calibration for
364 day Case
Initial Calibration
Very similar problems as with the 185 day
case
<* Shallow gradient
Poor match to high and low heads
-------
Initial Calibration
Improving the Calibration
Using same conductance that achieved a
good calibration in the 185-day case at the
river
* River bed conductance should remain constant in
models
Decrease the GHB on the western boundary
slightly
^ Original: 545 to 543 feet
* New reference heads: 543 to 541 feet
-------
New Equipotentials / Flow Directions
New Calibration Graph
-------
Calibration Comments
Gradient well matched
Absolute mean residual = 0.2 feet
Normalized RMS = 8%
Mean error = -0.05 feet
Overall, very good calibration statistics
Velocity Estimate
Darcy Flux = (K) x (Gradient) = 0.2 ft/day
Velocity = Darcy Flux / porosity = 0.7 ft/day
Calculation completed at river near well NN5
-------
Comments
No recharge in model
Transient calibration should be considered
Model is well behaved - converges quickly
May need more discretization for transport
------- |