.ufi Bit
.- ..;. v. ,. my.

15".“ .
. 6.5..

: ...
3.. .1

 

 

....w.:l.av.fi, .V . . ‘
4 _, ., , ‘ . fiihéfié»

.. 5;.-

 

lllllllllllllllllllllllllllllllllllllllllllllllll

3 1293 014201

This is to certify that the

thesis entitled
GROUNDWATER FLOW IN A FRI-:CTURED POROUS

MEDIA AT PALOS FOREST PRESERVE, ILLINOIS
presented by

Michael L. Kohn

has been accepted towards fulfillment
of the requirements for

MASTERS degree in GEOLOGY

 

: prof/essor

Date % _ {4.4 /’2/ /C/~’?;Sfl

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE N RETURN BOXto monthl- chockout from your rocord.
TO AVOID FINES Mum on or bdoro out. duo.

DATE DUE DATE DUE DATE DUE

 

 

w 330 g
0 ~- 4 iv I

 

 

 

 

.5 , -
’ Li, 0 ".

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Wan-9.1

 

GROUNDWATER FLOW IN A FRACTURED POROUS MEDIA
AT PALOS FOREST PRESERVE, ILLINOIS

By

Michael Leon Kohn

A THESIS

Submitted to
Michigan State University
in partial fiilfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Geological Sciences

1995

ABSTRACT

GROUNDWATER FLOW IN A FRACTURED POROUS MEDIA
AT PALOS FOREST PRESERVE, ILLINOIS

By

Michael Leon Kohn

Low-level radioactive waste generated from the reactors was buried on-site at the Palos
Forest Preserve in six foot deep trenches until late 1949, when it was excavated and removed.
In 1973, tritium was detected in the dolomite aquifer at the site. The Red Gate Woods
stream is suspected to have become loaded with tritium from waste burial during reactor
operation and transported it to the unconsolidated materials below. Seasonal fluctuations of
tritium concentrations in the dolomite aquifer at well FP5167 proximal to the picnic area
were unexplained. Seasonal fluctuations were hypothesized to be the result of fluctuations
in recharge through the glacial drift bringing varying amounts of tritium into the dolomite
aquifer or varying fluxes of groundwater flow in the dolomite aquifer diluting a constant flux
of tritium from the drift and varying the concentrations of tritium in the dolomite aquifer.
A three-dimensional fmite-difference groundwater flow model and a three-dimensional
contaminant transport model were constructed to test the hypotheses. Modeled tritium
concentration data for the two hypotheses was graphically compared and contrasted against
measured tritium concentration data from well FP5167. Based on the results of graphical
comparison, fluctuations in recharge are most likely the major control on the fluctuations in
tritium concentrations in the dolomite aquifer proximal to well F P5167, but the variations

in groundwater flux are probably also a contributing factor.

ACKNOWLEDGEMENTS

I would like to thank the MSU Department of Geological Sciences Faculty and Staff for their
support during the achieving of my masters degree. My major advisor, Professor Grahame Larson,

has provided valuable guidance and a wide scope of knowledge that helped me to finish my thesis.

I would also like to thank Jim Nicholas of the US. Geological Survey for providing the data on

which this thesis is based, and for the numerous efforts to facilitate the completion of my thesis.

Heartfelt thanks to my parents, Leon and Rosemary Kohn, who have provided the limitless

encouragement and support throughout my entire educational process.

The deepest thanks of all I extend to my lovely and understanding wife Louise. She has supported
and stood by me throughout the 1.5 years after graduate school when week-nights and week-ends
were devoted to my thesis. Her love and help were freely given even when my writing was causing
us not to have a life. It is solely the result of her love and support that l have achieved what I have

today.

iii

TABLE OF CONTENTS

List of Tables .......................................................... vi
List of Figures ........................................................ vii
1.0 Introduction ........................................................ 1
1.1 Statement of Purpose ............................................. 8

1.2 Conceptual Approach ............................................ 8

1.3 Previous Models ................................................ 9

2.0 Geology .......................................................... 11
2.1 Quaternary .................................................... 11

2.2 Silurian ...................................................... 13

3.0 Hydrogeology ..................................................... 15
3.1 Surface Water ................................................. 15

3.2 Groundwater .................................................. 16

3.2.1 Drift .................................................... 16

3.2.2 Dolomite ................................................ 18

3.2.2.1 Dolomite Hydrogeological Parameters .................. 20

4.0 Tritium Migration .................................................. 23
5.0 Groundwater and Contaminant Transport Models ......................... 27
5.1 MODFLOW .................................................. 27

5.2 Aquifer Characterization ......................................... 30

5.3 Initial Conditions ............................................... 33
5.3.1 Boundary Conditions ...................................... 33

5.3.2 Initial Parameters ......................................... 34

5.3.2.1 Recharge ......................................... 34

5.3.2.2 Leakance ..................................... '. . . . 36

5.3.2.3 Top and Bottom Elevations ........................... 37

5.3.2.4 Grid Spacing ...................................... 37

5.3.3 Assumptions ............................................. 39

5.4 MT3D ....................................................... 40

iv

TABLE OF CONTENTS (CONT'D)

6.0 Calibration and Sensitivity Analysis .................................... 40

6.1 Flow Model Calibration ......................................... 40

6.2 Sensitivity Analysis ............................................. 44

6.3 Transport Model Calibration ...................................... 45
7.0 Testing of Hypotheses ............................................... 52
8.0 Summary and Conclusions ........................................... 59
Appendix I

Precipitation Records for the Palos Forest Preserve ........................ 67
Appendix II

Measured Tritium Concentrations at Dolomite Aquifer Wells ................ 71
Appendix III

Water Levels in Dolomite Aquifer Wells ................................ 75
List of References ...................................................... 77

Table 1:

Table 2:

Table 3:

Table 4:

Table 5:

Table 6:

Table 7:

Table 8:

Table 9:

Table 10:

Table 11:

LIST OF TABLES

Summary of Dolomite Hydrogeological Parameters ................... 20
Initial Aquifer Property Values .................................... 36
Summary of Model Statistics from Initial Calibration Run .............. 44
Comparison of Groundwater Flow Model

Statistical Analysis Results ....................................... 46
Final Calibration Statistics for the Groundwater

Flow Model ................................................... 49
Dolomite Well Construction Elevations ............................. 51
Final Groundwater Flow Model Parameter Values ..................... 52

Measured and Modeled Tritium Concentrations at

Forest Preserve Well FPS 167 ..................................... 60
Precipitation Records for the Palos Forest Preserve Area ................ 67
Measured Tritium Concentrations at Dolomite Aquifer Wells ............ 71
Water Levels in Dolomite Aquifer Wells ............................ 75

vi

Figure 1:
Figure 2:
Figure 3:

Figure 4:

Figure 5:

Figure 6:

Figure 7:
Figure 8:

Figure 9:

Figure 10:

Figure 11:
Figure 12:
Figure 13:
Figure 14:
Figure 15:

Figure 16:

LIST OF FIGURES

Site Location Map ............................................. 2
Site Map ..................................................... 3
Model Area Map ............................................... 4

Groundwater Tritium Isocon Map for Dolomite Aquifer
(December 1983) .............................................. 5

Tritium Concentrations in Shallow Drift Materials Near the
Red Gate Woods Stream ........................................ 6

Seasonal Fluctuations in Tritium Concentrations at

Well FP5167 .................................................. 7
Site Geological Cross Section ................................... 12
Tritium Concentrations in Drifi Core Moisture Below Plot M .......... 17
Tritium Concentrations in Red Gate Woods Stream Waters ............ 25

Comparison of Fluctuations in Tritium Concentrations at

Well FP5167 with Precipitation .................................. 28
Model Geological Cross Section A-A' ............................. 31
Potential Evapotranspiration Amounts in Illinois .................... 35
Model Area Map with Grid Overlay ........................... g. . . 38
Recharge Node Values for Groundwater Flow Model ................. 42
Potentiometric Surface Map for Layer 1 ........................... 53
Potentiometric Surface Map for Layer 2 ........................... 54

vii

Figure 17:
Figure 18:
Figure 19:

Figure 20:

LIST OF FIGURES (CONT'D)

Groundwater Tritium Isocon Map for Layer 1 ....................... 55
Groundwater Tritium Isocon Map for Layer 2 ....................... 56
Annual Precipitation at Palos Forest Preserve ....................... 58
Comparison of Measured Versus Hypothesis Concentration Curves ..... 62

viii

1.0 Introduction

In 1943, the US. Army Corps of Engineers built three of the world's first nuclear reactors
in the Palos Forest Preserve, Illinois (Figure 1). Low-level radioactive waste generated from
the reactors was buried on-site at Plot M in six foot deep trenches until late 1949, when it
was excavated and removed (Figure 2). In 1973, tritium was detected in Forest Preserve
Well 5167 (Figure 3), which is located approximately 1200 feet down gradient from the
burial site, and a groundwater sampling program was implemented. Monitoring wells were
installed in the glacial drift and underlying Silurian dolomite to study the migration of tritium
in the groundwater. Tritium enters the dolomite aquifer at Plot M through the drifi and
lessens in concentration northward along the flowpath until it disappears below detectable
levels after approximately 730 feet. The tritium concentrations reappear in the groundwater
approximately 470 feet further along the flowpath and rise to peak concentrations that are
five times greater than those in the groundwater beneath the original source at Plot M (Figure
4) before diminishing again. A proposed interpretation for this unexpected concentration
gradient is that shortly after the burial of the radioactive waste, the Red Gate Woods stream
became loaded with tritium and transported it into the Red Gate Woods picnic area (Figure
5). Here the stream bed materials are more permeable and the stream water entered the
glacial drift below the stream and left behind high concentrations of tritium in the drift
materials. Still unexplained are the seasonal fluctuations of tritium concentrations in the
dolomite aquifer below the Red Gate Woods picnic area (Figure 6), which is the problem that

this study will address.

 

 

 

SITE A/PLOT M

   
    
   

CH iCAGO
LAKE

MICHIGAN

  
 
 

lLLlNOlS

 

Q‘
m
2
C

R

(OX RIVER

 

Lou nee R

   

 

     

ARGONNE-
Efif T

   
   

Farm 5
PALOS FOREST

"5551!; -

      

O 5 10

 

SCALE IN MILES

 

 

 

 

Reference: ‘Environmental Restoration and Waste Management. Site Specific Plan: Vol. 9, Site
A/Plot M”, US. Department of Energy, Chicago Field Office. DOE/CH-9227

Figrnel:SiteIncationMap

 

 

 

 

Cook County

 

 

 

    

 

  
 

  
 
 
    
  
    
   

 

 

LKChicago
IL L1 N O l S
ct“'
n ,-
ic'flw. '
Du Panel .. .. 0‘, ‘
COe .. 0' .... “u“ .
- Cook Co. .-
e‘ ,. - .
0.6.3 - Red State Woode
5319.... ' g '\ Picnic Area
6 -' -
e“ . )
a .t'. .:
o“ ' 0 Plot M
5'.» . area 0A
¢¢°°n - O ”o
/ 0‘s .. , o
' c" '~ Red Gate
’ ' my." Woods Stream
Avenue Site . 00 ®
A ‘1
PALO-s FOREST smash/kw

$ . ..; . g.ac ...

9 1 1090 2190 FEET
FT 27.0 550 METERS
Topographic contour interval so teat. NOW) or 1929.

 

 

 

 

“mnefiSiteMap

 

 

 

iLLlNOlS AND MICHIGAN CANAL
A

 

 

5215

 

 

N
SIS. SI"
‘ ‘
ARC’_R AVENUE \

 

 

 

 

 

DH 7
A
on 4 A
A iOO 200 SUD
L__L_1_J
I Fear
1—
<
O.
’— E"P “‘A'iOh
O A LAN 1 i
0
LL 5215
‘ Fore-I Prelerve 'eil
OH 5
‘ Dole-ire Hole leii
A
p\
< ' A' / Line of Cron Section
47
DH 2 I
a A

 

 

 

 

 

 

IfiguretklflodeLArealflap

 

 

 

0'2 \
5153

.(o.4 1i

   

 

Red Gate Wooda

Picnic Area . (:03)
DH.('20.2) DH‘G e

(<03); Dl-i-7 4'

’———

 

 

 
   
  

0.2
l<o.2)
czree
N
1 o 250 FEET
i-———-i
o 75 METERS
EXPLANATION

-5.0- LINE OF EQUAL TRiTiUM
CONCENTRATION. IN NANOCURIES
PER LiTER-- interval in variable

(4.9) WELL AND NUMBER-- Well number (tritium

concentration. In nanocuriee per titer in
5‘“ December 1983)

Reference: Nicholas and Healy (1988)

 

Figure 4: Groundwater 'l‘rifiumlsocon Map for-DolomiteAquifer (December 1983)

 

 

 

 

 

  
 
 
 

Red Gate Woods
Picnic Area

 

 

 

Phi M "wh ;_:

0 75 FEET
F—r“
0 2O METERS

. Red Gate

teed: Stream

  

3
///' zso FEET
li————~
o 75 METERS

 

 

 

 

Reference: Nicholas and Healy (1988)

Figure5:1rfihnnCXnnxnnrafionshaShaflonflDrfitnhunrhflsPkmmthelhui(kueVVoods

Sueanr

 

EXPLANATION

I DRIFT CORED MOISTURE
ANALYZED FOR TRITIUM-- The
highest concentration at each
location In contoured

TRITIUM CONCENTRATION. IN
NANOCURIES PER LITER

10—103
102-103
3
10-10
ratios
5

more than 10

Hill”

 

 

 

 

 

 

 

 

 

2 E

Q l—

l— : 1‘ r T T I i i i T

E 35 watsm

'2 CL 12 -
LIJ

o {1’3

2 E 8 .
O 3

o o ‘ __ .
2 o

E E o L I L I I J

— z

E z 1976 1978 1980 1932 '1934

Reference: Nicholas and Healy (1988)

Figure 6: Seasonal Fluctuations in Tritium Concentrations at Well FP 5167

 

 

 

1.1 Statementnfhnncse

The purpose of this study is to determine if the fluctuations of tritium concentrations in the
dolomite aquifer at Forest Preserve Well 5167 are the result of (1) fluctuations in recharge
through the glacial drift bringing varying amounts of tritium into the dolomite aquifer or (2)
varying fluxes of groundwater flow in the dolomite aquifer diluting a constant flux of tritium
from the drift and varying the concentrations of tritium in the dolomite aquifer. The first
hypothesis represents a variation in localized recharge while the second hypothesis represents
a variation in regional recharge. These two hypotheses are based on the above mentioned
assumption that the major source of tritium to the dolomite aquifer in the study area is the

tritium present in the glacial drift materials below the stream.

1.2 MW

To address the above mentioned hypotheses a through review and investigation was made
of published material regarding the site. Available data was compiled to be used in the
construction of a three-dimensional groundwater flow model of the site. As the available
data was compiled and analyzed, the geological and hydrogeological concepts of the site
were formed and enhanced. A three-dimensional groundwater flow model was constructed
for the site to compute the parameters needed for the contaminant transport model (eg-
groundwater flow rates, recharge through the drift). Assumptions were made to facilitate the
characterization of hydrogeological conditions at the site with mathematical equations. The
groundwater flow model was calibrated using yearly averages of hydraulic heads and

precipitation. The model parameters obtained from the background review of published

9

material were adjusted during the calibration process to create the best representation of

physical parameters with mathematical equations.

The output parameters from the groundwater flow model were used as the input parameters
for the contaminant transport model. Parameter values for the groundwater flow model were
adjusted with regards to the contaminant transport model that was calibrated under the same
conditions as the groundwater flow model. The contaminant transport model was run, one
time for each hypothesis, for the months of three years that were wet, normal, and dry with
regards to precipitation. The numerical results from the transport modeling of each
hypothesis at Forest Preserve well 5167 were plotted and compared to the measured tritium
data of the same time periods to evaluate which hypothesis produces the best match. The
hypothesis that produced the best graphical match for the graph of the measured tritium data
was thereby most correct. Conclusions were formed and presented as an enhanced

conceptual hydrogeological model for the site.

1.3 mm

Previous groundwater flow models have been constructed for the study area by Olympio
(1980). The models that Olympio constructed were regional in size and consisted of a two-
dimensional finite-difference areal model, a three-dimensional finite-difference areal model,
and a two-dimensional finite difference cross-sectional model. The areal groundwater flow
models encompassed an area of 11.7 square kilometers and had a grid spacing that varied

from 100 x 100 meters to 200 x 400 meters. The three models were designed to test the

10

conceptual hydrogeological model that Olympio had constructed for the region. The data
set that Olympio used to construct the flow models was small and incomplete, necessitating
the formation of numerous assumptions in order to construct the groundwater flow models.
The two-dimensional model was run at steady state to test the assumption that recharge to
the area was uniform, and to calculate a leakage rate from the drift to the dolomite. The one
layer used in the two-dimensional model was the underlying Niagaran dolomite that
represented a leaky-confined aquifer. The three-dimensional groundwater flow model was
developed to simulate the steady state distribution of groundwater in both the unconfined and
confined aquifers. The two layers used in the three-dimensional model were the upper
unconfined glacial drift and the lower confined dolomite. The value for transmissivity of the
dolomite aquifer was assumed to be homogeneous and isotropic throughout the system.
Finally, a two—dimensional cross-sectional model was constructed to determine the pathways

of groundwater flow through the two layer system.

None of the flow models could be calibrated due to the lack of data and the assumptions
needed to create them. A sensitivity analysis was conducted on the three-dimensional model
that indicated that the most sensitive parameters were the recharge rate and the transmissivity
of the dolomite. The results of the models indicated that a difference of 1 meter or less
between the modeled and measured head values was sufficient, and that further investigation
was needed to obtain additional data on bedrock topography and hydrogeological parameters

of the aquifers beneath the study area to refine and calibrate the flow models.

11

2.0 Geology
The following description of the study area geology is from Nicholas and Healy (1988). The

study area consists of a clayey drift overlying jointed Silurian dolomite (Figure 7).

2-1 Quatcmant

The drift is made up of Quaternary materials deposited during the Wisconsinan Stage
glaciation that are highly variable both laterally and vertically with respect to their lithology,
structure, and degree of sorting. It is composed of the Malden and Wadsworth Members of
the Wedron Formation which range in thickness over the study area from 1 to 170 feet with
a thickness of approximately 140 feet beneath Plot M. The Wadsworth Till Member is a
dense, clayey silt that becomes progressively better sorted with depth. Thin layers of sand
and gravel are numerous in the upper 25 - 35 feet of this member. The layers range in
thickness from 1 to 6 inches, dip to the northwest, and are laterally continuous for at least
300 feet beneath Plot M. Subvertical fractures that are filled with sand are also present in
the upper 10 to 15 feet. The underlying Malden Till Member is composed of sandy silt and
gravel which is well sorted and appears to have been reworked by glacial meltwater beneath
Plot M. Between the Maiden Till Member and the underlying Silurian dolomite is a thin
layer (1-2 feet) of drift composed chiefly of sand and gravel, except in areas where a clayey
silt layer is located on top of the drift-dolomite interface. Two valleys that have been incised
by glacial meltwater into the drift of the study area are located to the north and west. ‘ To the
north, the Des Plains River valley contains deposits of sand and gravel with occasional

boulders. The Sag valley to the west contains thick deposits of peat and has a low gradient

 

 

ALTITUDE. IN FEET

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

North South
75° w.u—-\
DH-1
Well
DH-Z
E
a Wall Plot M b
no a 7, 0”" Well
5 g DH-3
3 : mu
W911 DI'I-B
e e
‘ g 5157 Glacial drift
= g _
850‘ :- ~
g E
«a 3
o e
2 2 w."
2 g 5159 Approximate bedrock
‘ 5 surface and top of L.
000‘ 0 ‘ Subreaionai /" weathered zone
_ Joints \ ‘ \ \ _
qu- ‘ — Q/ l\ __ //_‘\ '-
1?. L1 1 1;.— I
1 TI 1: 1 ./ ll "
55° // ii I i i: ii ii
Regional\lolnta EL ll 1 : {Siluriani : dolomite 4”
i I ' r I it
II I I j j I I
H H "
I l
- ' l l . -
500 .
0 1000 FEET VERTICAL EXAGGERATION x 19
(T 300 METERS

Reference: Nicholas and Healy (1988)

Figure 7: Site GeologicalCmssSection

 

 

13

with no natural stream.

22 Silurian

The Silurian dolomite in the study area is composed of the Racine, Sugar Run, Joliet,
Kankakee, and Elwood Formations (in descending order). It is the dolomite of the Racine
Formation whose groundwater flow and tritium transport processes that this study will focus
upon. The dolomite of the Racine Formation averages 80 feet in thickness and is a light-
gray, silty and cherty, well-bedded dolomite that ranges from pure to shaley. This dolomite
has been extensively weathered by solution. Outcrops of dolomite scattered along the edge
of the Des Plains River and the Calumet Sag Channel indicate the dolomite is thin bedded,
argillaceous, highly fractured, and contains many clay filled solution cavities that measure

up to 30 x 60 feet in size. The top portion of dolomite bedrock is weathered and fiactured.

The study area is located near the crest of the Kankakee arch and the strata dip slightly
towards the east and southeast. Jointing in the brittle dolomite was caused by tensile stresses
from subsidence of the Michigan basin and uplift of the Wisconsin and Kankakee arches.
This stress caused two sets of vertical joints and a horizontal joint set that was enlarged by
dissolution along bedding planes. Vertical fractures are common in the weathered zone at
the top of the dolomite, but decrease in frequency and aperture below the upper few feet of
dolomite (Nicholas, 1988). Horizontal joints along bedding planes have been observed in
outcrops and borehole geophysical logs (Figure 7). The horizontal joints are at least

hundreds of feet long and are enlarged by solution with some up to several inches wide, as

14

seen at outcrops along the northern side of the Des Plains River valley. Acoustic and caliper
logs of wells in the study area clearly show the horizontal joints. The jointing and lithology
correlate between wells and appear to be continuous for at least 1600 feet. The US. Army
Corps of Engineers collected rock cores during the construction of the Calumet Sag Channel
that show horizontal joints having apertures up to 2 feet, some of which are filled with clay-

to-sand sized particles of glacial origin.

For the purpose of this study, the four major horizontal joints in the upper 80 feet of the
dolomite will dominate the groundwater flow system in the dolomite aquifer. They are
aerially extensive with only canals or sinkholes interrupting them. The following description
of the major fractures beneath the study area comes from Nicholas and others (unpublished)
and Shapiro and Nicholas (1989). Interpretation of borehole geophysical logs (caliper and
acoustic televiewer) indicated that there were four major horizontal fractures or fracture sets
within the upper 80 feet of dolomite that range from 0.1 to 0.75 feet in aperture. The
uppermost fracture set occurs in the weathered zone of the dolomite at approximately 566
feet and consists of three or more horizontal fractures that may be connected by vertical
fractures. The second fracture set appears on geophysical logs as a single large aperture
fracture at approximately 544 feet. Rock cores however indicate that it may consist of 5-6
small aperture fractures along bedding planes. A large aperture fracture at approximately
522 feet comprises the third fi'acture set, and a small aperture fracture at approximately 511
feet constitutes the fourth fracture set. Hydraulic tests using packers indicated that there are

no natural hydraulic connections between horizontal fractures in the study area (Nicholas et.

15

a1, 1988), but since many of the dolomite hole (DH) wells are uncased below the dolomite-
drift interface and penetrate the depth of the above described fractures there are now
anthropogenic pathways linking horizontal fractures. Beneath the dolomite, a sequence of
shales and dolomites separate the upper dolomite aquifer from the lower dolomite aquifer
and form a lower impervious boundary (aquitard) for the dolomite aquifer that this study will

focus on.

3.0 Hydrogeology
The hydrogeology of the study area is composed of the two key components of Surface

Water and Groundwater.

3.1 Smfaslflater

The main components of the surface water system in the study area are the Red Gate Woods
stream, Chicago Sanitary and Ship Canal, Illinois and Michigan Canal, and the Calumet Sag
Channel (see Figure 2). The channels and canals are recharged both by groundwater and
surface water. They remain filled with water throughout the year and exhibit a relatively

constant stage.

The Red Gate Woods stream is an ephemeral stream that flows, seldom other than early
spring, from near Site A towards it's discharge area into the Illinois and Michigan Canal
(Nicholas and Healy, 1988). The source of the streamflow comes in the forms of snowmelt,

precipitation, and groundwater discharge. Proximal to Plot M, the stream is a gaining stream

l6

and then becomes a loosing stream in its lower reaches. The large degree of surface relief in
the Plot M area serves to enhance the groundwater discharge into the stream. Even during
periods of high flow, the stream water seldom reaches as far as Archer Avenue (Nicholas and
Healy, 1988). Evidence of the gaining and loosing properties of the stream can be seen as
the isocons of tritium in the streambed materials beginning proximal to Plot M and ending
at the Red Gate Woods picnic area (Figure 5). During the summer months,
evapotranspiration removes a sufficient amount of water from the underlying drift to lower
the water table below the stream bottom. The stream occasionally flows in the fall and

winter when the ground is not frozen.

3.2 Cnnundrrtater

The groundwater flow system in the study area is divided into two regions; the saturated
portion of the drift, and the dolomite aquifer beneath the glacial drift. Groundwater in the
drift region recharges the dolomite aquifer. The focus of this study is on the dolomite aquifer

so the description of the saturated deposits of the drift will be brief.

3.2.1 Dfifi

The hydrogeology of the saturated unconsolidated deposits that make up the drift is complex
and variable. The complex stratigraphy of the drift is composed of multiple layers with
different hydraulic conductivities. Nicholas and Healy (1988) divided the hydrogeOlogy of
the drift into three zones with the following descriptions in descending order: (1) an upper

perched zone (2) a variably saturated zone (3) and a lower, fully saturated zone (Figure 8).

17

 

 

 

  
  
 
 

 

710
.. - Red Gate
88-35 53 9 88 0 Woods
690‘ '-
67:): 1° -

 

  
 
  
 
 
 
 
 

  

perched

 

  

 

 

 

 

 

 

650-1 I, zone I-
630‘ 'v.r|.bly_
/ saturated

’ zone

  
 
 
  

Lower iuliy

ALTITUDE. IN FEET

 

 

 

 

 

 

 

590-4
saturated
zone
ETD-I /
j ’ Driit
. _ — fl '—
550- Dolomite
63° 0 50 FEET
I , VERTICAL EXAGGERATION x3
0 10 METERS
EXPLANATION

LINE OF EQUAL TRITIUM CONCENTRATION.
IN NANOCURIES PER LITER "'
Dashed where approximate. Contour
interval Is logarithmic

-100-

eeeeeeeee SAND LAYER

Reference: Nicholas and Healy (1988)

Figure 8: Tritium Concentrations in Drift Core Moisian'e Below Plot M

 

 

18

The upper perched zone is composed of saturated sand layers that are surrounded by layers
of clayey silt. Groundwater flow in this zone is nearly horizontal due to the lower hydraulic
conductivity of the surrounding clayey silt. Water levels in this zone fluctuate due to

seasonal changes in recharge, which are also reduced by the overlying concrete cap of Plot

M, to the drift.

The variably saturated zone is approximately 35 feet thick and is located between the upper
perched zone and the lower, fully saturated zone. The saturated materials of this zone are
composed of sandy silt and gravel. This zone has a hydraulic conductivity greater than that
of the bottom of the upper perched zone, thus allowing it to drain at a higher rate. This
higher rate of drainage is evidenced by the piezometers installed in this zone only

occasionally having water in them, hence the term "variably saturated".

The lower, fully saturated zone is composed of saturated materials that consist of silty sand
and gravel. Groundwater levels in piezometers installed in this zone have a constant amount
of water with levels that vary only over a very small range. The groundwater tritium isocons
(Figure 8) of this zone indicates that groundwater flow is downward towards the dolomite

it interfaces with.

32-2 Dolomite

The groundwater flow system in the dolomite aquifer is controlled by drainage areas; the Des

Plains River (parallels the Illinois and Michigan canal along the study area) to the northwest,

19

and the Calumet Sag channel to the southwest. Groundwater flows in the dolomite aquifer
fi'om the center of the Palos Forest Preserve toward the valley discharge areas (Nicholas and

Healy, 1988). Leakage from the overlying glacial drift recharges the dolomite aquifer.

The dolomite has both a primary porosity which consists of the void spaces between
crystalline grains, and a secondary porosity which consists of fractures and solution enlarged
joints. The primary porosity of the dolomite is mostly insignificant in terms of groundwater
flow, except as a storage reservoir (Nicholas and Healy, 1988). The major conduit for
groundwater flow in the dolomite is the solution enlarged joints, especially those along
bedding planes. There is a distinct and important relationship observed between the fiactures
and lithologic changes within the bedrock (Nicholas and others, unpublished). The major
fractures in the bedrock occur above argillaceous layers which would have a significantly
lower hydraulic conductivity and solubility. The lower hydraulic conductivity and solubility
would retard the vertical component of groundwater flow across the argillaceous layers
which would in turn enhance the horizontal component of groundwater flow and favor matrix
dissolution above and along the top of the argillaceous layers. As the horizontal fractures
are increased in aperture by this preferential process, they become preferred groundwater
flow pathways and increase the effects of the dissolution process perpendicular to their axes

which increases their aperture.

20

3.2.2.1 chmueflxttosmlngicaflflmtm

The groundwater flow system within the dolomite aquifer can be divided into three different
types of zones: (1) an upper weathered bedrock zone (2) dolomite rock-matrix and (3)
fractures/joints within the dolomite bedrock. A large variety of borehole geophysical tests
have been conducted on dolomite wells in the study area to determine the hydrogeological
parameters of these zones. Horizontal joints located below an altitude of 570 feet form a
regional groundwater flow system throughout the study area. A summary of the

hydrogeological parameters of these three zones is presented in Table 1.

Table 1: Summary of Dolomite Hydrogeological Parameters

 

 

 

 

 

F

Zone Transmissivity Storativity Porosity Hydraulic
Conductivity

Weathered 120.96 ftZ/day 1.6 x 104 0.200 NA
Zone
Dolomite NA NA 0.074 0.9504 ft/day
Matrix
544 Fracture 26039.23 2.2 x 10‘5 0.800 NA

i ftz/day

 

 

 

 

 

 

 

 

The weathered zone is composed of weathered and fractured dolomite that extends from the
drift-dolomite interface to an average depth of 565 feet. Within the weathered zone is a set
of horizontal fractures, 3 to 4 fractures, that are connected by vertical fractures (Nicholas,
1988). Water levels in the weathered zone responded to pumping like a leaky confined
aquifer, because this zone is hydraulically connected to the overlying drift (Nicholas, 1988).
Inflatable packers were used to isolate the weathered zone from an altitude of 565-575 feet.

Results of the test were presented by Nicholas and others (unpublished) which indicated a

21

transnrissivity of 1.4 x 10‘3 ftZ/sec. Storativity was not included so the value of 1.6 x 10“ was
obtained from uncredited analysis plots of the pumping test data. A porosity of 0.20 was
chosen from ranges of porosities for different geologic materials that was published by

Dominico and Schwartz (1990).

The dolomite matrix that was tested for hydrogeological parameters is composed of more or
less competent crystalline dolomite with possible small fractures. The altitude of the
dolomite matrix extended from approximately 544-565 feet. The vertical hydraulic
conductivity of the dolomite matrix was measured from laboratory analysis of rock cores
which yielded values of 1.3 x 10'8 to 1.5 x 10'10 ft/s. The rock cores measured by the
laboratory did not include fractures, so the dolomite matrix may yield a higher value of
hydraulic conductivity than the range indicated by laboratory analysis (Nicholas, 1988).
Recovery measured during a pumping test of the packed off dolomite matrix (555-565)
indicated a hydraulic conductivity of 1.1 x 10'5 ft/sec (Nicholas and others, unpublished).
During the first two seconds of pumping the matrix, water levels dropped 15 feet in the well,
indicating that much of the water from this zone comes from storage. Laboratory analysis
of dolomite matrix core samples collected at 123 locations yielded values of porosity that
ranged from 1.7 to 18.7% (Nicholas and others, unpublished). The mean porosity was 7.4%,
which falls within the range of porosity values for dolomite (0-20%) published in Physical

and Chemical Hydrogeology (Dominico and Schwartz, 1990).

Three more major horizontal joints sets in the upper 80 feet of dolomite, that were

22

continuous throughout the study area at altitudes of 544, 523, and 513 feet, were identified
by Nicholas (1988). Nicholas and Healy (1988) estimated the hydraulic gradient of this
system to be 6.25 x 10“. The most important and transmissive of the three above mentioned
horizontal fracture is the fracture at 544 feet. The following description of the fracture at an
altitude of approximately 544 feet comes from Nicholas (1988), Nicholas and others
(unpublished), and Shapiro (1989). The dolomite fracture at 544 feet pumped like an infinite
confined aquifer and had a transmissivity greater than any of the other fractures that were
measured, which is approximately equal to the transmissivity of the entire section of
dolomite measured during open-hole (packers not used) pumping tests. Analysis of the
pumping tests of this fracture by Shapiro (1989), indicated a transmissivity of 0.301 38 ftz/sec
and a storativity of 2.2 x 10's. The hydraulic connection of horizontal fracture sets by
existing uncased boreholes in the dolomite was proven during open-hole tests with a flow
meter that measured vertical flow. Within the first seconds of the pumping test, water
flowed vertically in the observation boreholes to the altitude of the fracture being pumped.
Water quality data indicates that there is minimal mixing between the water from the fracture
at the altitude of 522 feet and the fracture at 544 feet. Tritium concentrations in the fracture
at 544 feet are ten times less than the those of the weathered zone, which may indicate the
downward flow of groundwater from the weathered zone to fractures below through Open
wells. Under non-pumping conditions in open wells, groundwater also flows from the
weathered zone to the fracture at 544 feet. Nicholas and others (unpublished) states that the
fracture at 544 feet is a flow divide, above the most argillaceous layer in the dolomite

section, whose high transmissivity contributes to the low relief of the potentiometric surface

23

measured across the site. The average width of the fracture at 544 feet is approximately
0.336 feet (Keys, 1986). Values of porosity for this fracture were not found during the
literature review, so a value of 0.80 was assumed based on the likely hood that the fracture

would not have a constant width nor parallel plates as an assumption of 1.0 would indicate.

4.0 Tritium Migration

Tritium (H) is a radioactive isotope of hydrogen that has a half-life of 12.43 years. Tritium
concentrations are traditionally reported as tritium units (TU), with 1 TU corresponding to
1 atom of 3H in 1018 atoms of 1H (Fontes, 1980). Tritium concentrations for this study are
reported as nanocuries per liter (nCi/L) in groundwater at the site. One curie is the amount
of material undergoing 3.7 x 10‘0 disintegrations per second. The measurement errors
associated with the analysis of tritium at concentrations of 1.0 nCi/L or less in groundwater
total i 0.1 nCi/L (Nicholas and Healy, 1988). Tritium occurs naturally in the atmosphere
and moves with the water cycle by the tritium atom substituting for the hydrogen atom in the
water molecule (Reilly et. al., 1994). In this way, the tritium atom follows the path of water
through the groundwater flow system. Nuclear weapons testing in 1953 increased the
concentrations of tritium in the atmosphere, thereby increasing the concentrations in
precipitation. During the early 1960's, tritium in precipitation reached three orders of
magnitude greater concentrations than normal under natural conditions (Reilly et. al., 1994).
Since the atmospheric nuclear weapons testing ban in 1963, tritium levels in the atmosphere
have been steadily declining. Tritium concentrations in atmospheric precipitation are

influenced by factors such as latitude, distance from the ocean, and seasons.

24

Elevated tritium concentrations at the study area are the result of interaction by precipitation,
surface water, and groundwater with radioactive waste buried in the vicinity of Plot M during
the reactor operation from 1943 to 1949. Golchert and Sedlet (1978) reported that tritiated
groundwater had discharged fiom the glacial drift proximal to Plot M into the Red Gate
Woods stream since at least 1954. The general trend of tritium concentrations in stream
water (Figure 9) indicates that upgradient tritium concentrations in the stream are usually at
the background level of 0.2 nCi/L, but increase proximal to Plot M and then decrease
downstream as a result of dilution (Nicholas and Healy, 1988). Concentrations of tritium
above background levels have been measured downstream as far as where the stream
discharges into the Illinois and Michigan canal. Samples of stream water proximal to Plot
M in 1983 had tritium concentrations of 3 1 .6 to 425 nCi/L which suggest that the discharge
of tritiated groundwater into the Red Gate Woods stream may occur at localized areas where

the stream bottom intercepts sand lenses in the glacial drift (Nicholas and Healy, 1988).

Concentrations of tritium in groundwater in the drift below Plot M are presented in Figure
8, which indicate a downward movement of tritiated groundwater through the drift to the
drift-dolomite interface. Also present is a minor lateral component of tritiated groundwater
flow which may be the result of a preferred migration pathway along bedding planes and
unconsolidated materials of higher hydraulic conductivity. Data from the moisture in drift
cores indicates that the movement of tritiated groundwater beneath Plot M is progressing at
a very slow rate (Nicholas and Healy, 1988). Data presented by Olympio (1980) indicated

that the travel time for tritiated groundwater in the drift from Plot M to the drift-dolomite

 

 

STREAM - SAMPLING LOCATION

7

k

L

10 N

L
T

 

T 7

1000

> ---- February 25. i976
.. February 26. I976

 

100

 

...... ... April i9. i978
—— MI] '6. 1979
_..-— January 27. I980
May 27. 198I
-..-...... March l8. 1982
—----— January 25. I983

A l

L

a 4 a aaaal

 

I

TRITIUM CONCENTRATION, IN NANOCUFIIES PER LITEH

 

a anaaaal

 

I 1

 

 

o zoo 400 600 . 300 1000 1200
DISTANCE, IN FEET
r
10-131 10/100 N
- 71109 1
”200‘ .)
j - 71109
/- \. e srarmsmmc LOCATION-
. / i Tritium concentration.
01403 ' 5 if' - / In nanocuries per liter.
{/ N. 5,1 1. m: ...... .... ....
'\ ‘4’124 o 150 FEET
-\ 31jugs/1:70 o 40 METERS
425
/‘ 212.4
110.2/'

 

 

 

Reference: Nicholas and Healy (1988)

Figure 9: Tritium Concentrations in Red Gate Woods Stream Waters

 

 

26

interface was 1 1 16.66 days. Olympio (1984) concluded that the initial downward movement
of tritiated groundwater occurred before the concrete cap, which cuts off the drift below Plot

M from recharge (precipitation), was constructed in 1956.

Tritium concentrations in the dolomite aquifer beneath the study area are the focus of this
study. Tritiated groundwater enters the dolomite at two different areas (Figure 4), forming
two distinct groundwater tritium plumes (Nicholas and Healy, 1988). Concentrations of
tritium and the size of the tritium plume are considerably less beneath Plot M than beneath
the Red Gate Woods stream. This difference is unexpected since the concentrations of
tritium and the size of the source area would seem to be much greater at Plot M rather than
at the Red Gate Woods picnic area. The hydrogeological characteristics of the weathered
bedrock zone make it a conduit for tritium migration from the drift into the dolomite aquifer.
Wells open to the weathered bedrock zone that are along the groundwater flow path from the

source beneath the stream yield the highest tritium concentrations.

Nicholas and Healy (1988) observed that elevated concentrations of tritiated groundwater
have not been detected in wells open to subregional dolomite joints (see Figure 7). The
regional horizontal joints are major conduits for the migration of tritiated groundwater in the
vicinity of the Red Gate Woods plume. Nicholas and Shapiro (1986) report that tritium
concentrations in dolomite solution joints decrease with depth and range from 0.2 to 30
nCi/L. The tritium concentration in the adjacent dolomite matrix is less than that of the

solution joints and has a maximum measured concentration of 10 nCi/L. Forest preserve

27

well (F P) 5167 is open to the major horizontal joints. Tritium concentrations in F P 5167
have fluctuated seasonally since measurement began in 1973, ranging from background
levels of 0.2 nCi/L in the summer to about 10 11ch in the winter (Nicholas and Healy,
1988). Nicholas and Healy (1988) noticed a lag time between major precipitation events and
tritium fluctuations (decreasing concentrations) of 20 to 40 days in FP 5167. Figure 10
presents graphs of tritium concentrations and precipitation amounts over time at the same
time scales. The matching of these graphs led Nicholas and Healy (198 8) to conclude that,
"Variations in the concentration of tritium in well PP 5167 are caused by variations in

recharge to the dolomite."

5.0 Groundwater and Contaminant Transport Models

5.1 MODELQW

MODFLOW (McDonald and Harbaugh, 1988) is a three-dimensional finite-difference
computer program that was used to simulate groundwater flow conditions in the study area.
The MODFLOW program simulates three-dimensional flow by using block-centered finite-
difference equations that can be solved by using either the Strongly Implicit Procedure or
Slice Successive Overrelaxation iterative solutions. The program uses subroutines (modules)
that are grouped into packages and procedures. The packages allow the incorporation of
internal and external influences on the model such as wells, rivers, recharge, drains, and

evapotranspiration.

Other researchers (Huyakom et al, 1983; and Bibby, 1981) have chosen finite element

 

 

P

 

b

.. ARGONNE NATIONAL LABORATORY II

J l r l L ' I

 

Ii‘é‘t‘tsnauo

 

 

CUMULATIVE MONTHLY
RECIPITATION IN INCHES

I T T T T o

I r
WELL 5167

13

 

 

 

r r r I l 1 1

1976 1978 1980 1982 1984

 

TRITI UM CONCENTRATION.
IN NANOCURIES PER LITER P
ADJUSTED FOR DECAY
as

Reference: Nicholas and Healy (1988)

Figure 10: Comparison of Fluctuations in Tritium Concentrations at Well FP 5167
With Precipitation

 

 

29

models to better represent an aquifer in which fractures and jointing are present. Finite
element models use irregularly spaced grids, represent such tensorial concepts as
transmissivities which do not coincide with coordinate axes of the model, and are more
flexible in representing boundaries than finite difference models (Kinzelbach, 1986). The
finite-difference model can accurately represent boundaries and capture hydrologic details
in a small area by keeping the grid size small. Tensorial concepts that do not coincide with
coordinate axes can be handled by aligning the model axes with the tensors of the aquifer to

be modeled.

The finite-difference program MODF LOW was chosen to model the hydrogeologic
conditions of the study area because of the many advantages it offered. MODFLOW uses
a modular structure of subroutines to simulate specific features of the hydrologic system
(sinks, rivers, drains, wells, recharge, etc.). The division of the program into modules
provides the capability of examining specific hydrologic features independently, which
allows for feature independent sensitivity analysis and statistical calibration modeling runs.
The modular structure also allows for the incorporation of new packages without rewriting
the entire model code. The advantages of a finite element model are equaled by constructing
the MODFLOW finite-difference model with the above mentioned considerations. Since the
testing of the hypotheses was dependent upon contaminant transport modeling, it was also
important to choose a flow model that could communicate with and support a transport
model. MODF LOW is the industry standard finite-difference groundwater flow model and

is widely supported by a variety of post-processing models. By using the industry standard,

30

linking with post-processing models is usually pro-established and free of de-bugging errors
that may not be readily apparent and may affect model results in a negative way. The post-
processing contaminant transport program MT3D (Zheng, 1992) has a pre-established link

with MODFLOW.

5-2 AgarfELClraractermLtron

The focus of the geologic and hydrogeologic investigation at the site has been on the
dolomite bedrock rather than the overlying glacial drift. The quantity and quality of
hydrogeological information gathered for the fractured bedrock zone is excellent, while only
limited information has been gathered for the overlying glacial drift. Both of the proposed
hypotheses focus on the hydrogeological properties of the bedrock zone and the transport
properties within it. Due to the focus of the study and the lack of complete knowledge of the
overlying drift, the conceptual hydrogeological model created focuses solely on the dolomite

bedrock beneath the study area.

Figures 3 and 11 present the plan and cross-sectional views of the conceptual
hydrogeological model respectively. According to Nicholas (1988), "The geology and
hydraulic properties of the dolomite suggest a conceptual model of flow that is analogous
to that in a layered-aquifer system. Horizontal fractures or fracture sets, such as the
weathered zone are analogous to aquifers, and the dolomite matrix between the horizontal
fractures is analogous to confining layers." Nicholas and Shapiro (1986) described the

dolomite flow system where, "Each solution joint is hydraulically analogous to a infinite

 

.<.< 838m 220 838.... a8: “2 ER

 

31

 

 

 

ruua
00' CON 0
XON I 20.k<rm0(xm J<U_kaw>
mJ(Ufl J(FZ°N.KOI
Emma
One One
man L 1 men
one g 1 one
nnn T .n am><qus3»0(au ULIOJOD 1 non
Ora I r Ova

RDRDRDNOQSDNOQQ >1<0230a BOJu oz \hkNQKQRbR QSQNQSQNQ“\Q5R5QSQNQNQRBQNORFQXDRhQV

 

 

are I r or»
0mm I .w02(¥(w..: X.¢.r(2 NEIOJOD L Onn
mom 4 1 man

00“ 1 0°“

    

“on I “on

.- (">(I: “ZON OUKUIP4‘u;

One 1 One

 

 

“km 1 “BO

WJZW>(
CW10K<

 

 

 

 

0.“ 0.“
15.300 IPKOZ

 

 

 

 

 

32

confined aquifer that is bounded above and below by the dolomite matrix which is assumed
to be impervious. However, the upper most solution joint set responds like an unconfined
formation due to it's hydraulic connection with the weathered zone at the glacial drift -
dolomite contact". Moffet and others (1986) stated that, "The dolomite matrix has high
storage and low hydraulic conductivity relative to the fractures, which are major conduits for
fluid flow." Previous descriptions of the hydrology of Silurian dolomite in northeastern
Illinois haven't differentiated the properties of the matrix or unfractured rock mass, from
individual fractures or zones of multiple fractures. Instead, they have assumed the aquifer
to be homogeneous for the scale of the investigation (Nicholas and others, unpublished).
Researchers have interpreted potentiometric, transnrissivity, and water quality data as if the
dolomite were a homogeneous, isotropic, porous medium. An assumption of this magnitude
may be valid for studies of water supply; but at the smaller field scales, usually associated
with contaminant transport where a few major fractures dominate solute transport, such an
assumption is probably inappropriate. Shapiro (1989) presented mathematical solutions and
interpretations for oscillatory pumping test data from dolomite wells in the study area, that
were a better match for projected solutions by the dual porosity model which described the
dolomite fractures and matrix separately, rather than the equivalent-porosity model which
treated them as a single homogeneous unit. In constructing a groundwater flow model for
this study area, the hydraulic and solute-transport pr0perties of individual fracture will need

to be described.

The model descritization focused on the dolomite bedrock where groundwater flow and

33

tritium transport would be modeled. The weathered bedrock zone was selected as the upper
model layer and the major bedrock fracture at 544 feet as the lower model layer. The
dolomite matrix between the above mentioned zones was represented as a leakance value
between the two zones. Communication between the major fracture at 544 feet and the lower
dolomite matrix and fractures has been stated to be minimal based on groundwater quality
data The lack of communication with zones below it and the argillaceous layer underlying

the fracture at 544 feet, make the bottom of the lower model layer a no-flow boundary.

Although the overlying glacial drift was not included in the model descritization, the
recharge needed to achieve calibration of the flow model to measured hydraulic heads of
dolomite wells will be representative of the groundwater contribution from the glacial drift,

Red Gate Woods stream, and precipitation.

53 Initialflunditiuns

53-1 W

The boundary conditions for the site groundwater flow model consisted of no-flow and
constant head boundaries which were set at the same value for each model layer. The north
side (Illinois and Michigan canal) and the south side (DH-2) of the model area were treated
as constant heads. Data for the Illinois and Michigan canal hydraulic head was unavailable
so an estimate of 578 feet was used (Nicholas, communication). The constant head along
the southern model boundary was chosen to equal the measured hydraulic head of DH-2 for

the time period of interest. The eastern and western sides of the model area were designated

34

as no-flow boundaries. Potentiometric surface maps of the study area indicated that
groundwater flow in the model area moved approximately perpendicular to the Illinois and
Michigan canal. The no-flow boundaries were located a sufficient distance away from the
focus area of the model so as not to create an influence on the hydrogeological parameters
that was not representative of actual groundwater flow characteristics. Golchert (1993)
indicated that forest preserve well FP5215 would be a good flow boundary for the model area
due to the absence of tritium in that well throughout the years of sampling, and it's side-
gradient location to the groundwater migration pathway beneath Plot M and the Red Gate

Woods stream.

5.3.2 InitiaLRaranreiers

The initial aquifer parameter values input into the groundwater model are presented in Table
2. The origin and explanation of many of the parameter values are discussed in previous
sections. The parameters of leakance, recharge, top and bottom elevations and the topic of

grid spacing will be discussed in further detail in the following sections.

5.3.2.1 Regharge

Recharge is the property used to represent the contributions from precipitation, the glacial
drift, and the Red Gate Woods stream to the dolomite aquifer. Values for recharge were
obtained from the cumulative monthly precipitation measured at the nearby Argonne
National Laboratory from 1980-1988. Monthly precipitation values were adjusted for

potential evapotranspiration occurring in Illinois (see Figure 12). The precipitation value

 

 

(we) NOIlVHIdSNVHiOdV/AB 'MiNEiocI

m.~

woo

bmwm

a?

gfiflggflgggflgfiag

$85 conch ”Scouomom

0mm pi 5o —
:2. cza. >52

 

 

’
ID

 

 

‘9
[x

 

 

 

Nd.

 

 

 

 

 

5N?

 

 

 

«.9

 

 

 

 

 

 

/z<:zua

 

 

n
(SEHONI) NOLLVHIdSNVHiOcIV/G Tqusuos

 

 

 

 

 

 

 

36

remaining after the subtraction of the potential evapotranspiration value was then used in the
groundwater flow model as recharge. Different magnitudes of the recharge value were
applied to different portions of the flow model (ie. higher at the stream to represent increased
contribution from the stream and the higher permeability of the streambed deposits). A
summary table of yearly precipitation values and the calculation of recharge values is

included in Appendix I.

Table 2: Initial Aquifer Property Values

 

 

Model Top Bottom Transmissivity Storativity Leakance
Layer Elevation Elevation
Layer variable variable Tx = 120.96 ftzlday 0.00016 0.09504
1 556-600 546-590 Ty = 120.96 ftz/day day‘l
ft ft T2 = 40.32 ftz/day

 

544 ft 543.7 ft Tx = 260392.32 ftZ/day 0.000022 not

Ty = 260392.32 ftZ/day applicable
T2 = 26039.232 ftz/day

 

 

 

 

 

 

5.3.2.2 Leakage;

Leakance is a property for which a value is assigned to represent the conductivity between
model layers. In the case of the groundwater flow model constructed for the study area, the
upper weathered dolomite zone (layer 1) and the lower dolomite fracture at 544 feet (layer
2) are separated by a layer of dolomite matrix that is not included as a separate layer in the
flow model. Instead the hydraulic conductivity of the dolomite matrix was divided by its'
thickness yielding a value with units of days". To the flow model this parameter represents
a communication value between model layers. The leakance value is assigned to layer 1

because the bottom of layer 2 is treated as a no-flow boundary across which communication

37

is not made with lower layers.

5.3.2.3 Warts

Model layer top and bottom elevations were created for each model node with data compiled
from Nicholas and Healy (1988), Nicholas (1988), Keys (1986), and from well construction
specifications. The top of the weathered bedrock zone (layer 1) was interpreted from
bedrock maps and well construction specifications. The bottom of this layer was obtained
by subtracting 10 feet, the approximate thickness of the weathered zone across the site, from
the top elevation at that nodal location. The top of layer 2 was the top of the major fracture
(544 feet) whose bottom elevation was obtained by subtracting the fracture thickness,

interpreted from acoustic televiewer logs to be 0.37 feet on average, from the top elevation.

5.3.2.4 oncoming

The grid system created for the three-dimensional groundwater flow model is presented in
Figure 13. Spacing between grid lines was kept small in the focus area of the model and then
expanded at maximum increments of 1.5x when moving away from focus areas. Grid
spacing ranged fiom approximately 50 to 150 feet. The purpose of keeping the grid spacing
small was to capture the hydrogeological details of the focus area and to facilitate a
contaminant transport process that modeled the concentration fluctuations at a single well
within the study area (F P5 167). The axes of the finite-difference grid were aligned parallel
and perpendicular to the groundwater flow direction and the major directions of

transmissivity at the site.

 

 

 

Figure 13: Model Area Map With Grid Overlay

 

 

39

5.3.3 Assumptions

The following assumptions were made to facilitate the characterization of hydrogeological

conditions with mathematical equations:

1.

Small grid spacing will facilitate the accurate representation of boundaries and
hydrogeological details by the finite-difference groundwater flow model.
The aquifer values obtained through hydrogeological testing and laboratory

analysis are representative of actual conditions at the site.

. No-flow boundaries were placed far enough away from the model focus area to

avoid creating an influence that was not representative of actual groundwater flow
characteristics.

The hydraulic head of the Illinois and Michigan canal did not vary over the time
periods being modeled, and was an accurate representation of the hydraulic head
of the dolomite aquifer at that location.

Recharge to groundwater from precipitation and the Red Gate Woods stream can
be represented by average monthly precipitation values, adjusted for potential
evapotranspiration, with increased precipitation values beneath stream nodes in
the flow model.

The major source of tritium to the dolomite aquifer in the study area is from the

tritium present in the drift below the stream.

40

5.4 M1312

MT3D (Zheng, 1992) is a modular three-dimensional transport model for simulation of
advection, dispersion, and chemical reactions of contaminants in groundwater systems. The
MT3D transport model is based on the assumption that changes in the concentration field
will not affect the flow field measurably. MT3D uses the hydraulic heads and various flow
and sink/ source terms saved by the groundwater flow model, and automatically incorporates
the specified hydrogeologic boundary conditions. The model uses a mixed Eulerian-
Lagrangian approach to the solution of the advective-dispersive-reactive equation, based on
a combination of the method of characteristics (MOC) and the modified method of
characteristics (Zheng, 1992). This approach combines both the strength of the MOC for
eliminating numerical dispersion and the computational efficiency of the modified MOC.
MT3D is a contaminant transport model that was developed for use with any block-centered

finite-difference flow model, but came with an pre-established link for MODFLOW.

6.0 Calibration and Sensitivity Analysis

6.1 W

Once the hydrogeological parameters had been compiled and the groundwater flow model
had been constructed, the calibration process was started. To begin the calibration process,
the groundwater flow model was run under steady state conditions using the average
hydraulic heads from 1984 for dolomite aquifer wells as calibration targets and the canal and
DH-2 as constant head boundaries. Recharge was added to the model as a series of

transitional steps over three model runs, with each successive run adding to the recharge of

41

the prior run. The first run was made without recharge to establish a baseline potentiometric
surface in the absence of recharge to compare later runs against. For the second run, the
average monthly recharge for 1984 was added to each model cell and termed "area recharge".
Recharge nodes beneath the position of the Red Gate Woods stream were assigned an
increased recharge value (10x normal) for the third run, to represent the added recharge
contribution from the stream and the increased permeability of the streambed deposits (see

Figure 14).

Recharge added during runs 2 and 3 did not bring the water levels at the northern end of the
model area high enough to match the average hydraulic heads measured from dolomite wells.
Area recharge was increased by 1 order of magnitude increments in an effort to create a
hydraulic gradient that was more representative of the measured hydraulic heads. Recharge
that was increased by 5 orders of magnitude greater than normal still did not bring the water
levels at the northern end of the model area high enough to match the average hydraulic
heads measured from dolomite wells. The corresponding increases in the greater recharge
contribution from the stream nodes raised water levels of well nodes proximal to the stream
higher than the measured heads and created potentiometric surfaces that were not realistic
for the hydrogeological conditions at the site. Also, the increase in recharge values was far

beyond what was reasonable for an average monthly recharge value for any year on record.

To address this problem, the hydraulic gradient of the site had to be reduced so recharge

could have more of an effect. The current hydraulic gradient exhibited by the model was

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I I \I\ r I I I i I I I I 2 I U I I I
O
\- I \ I I I I I \J/(TH I i I r ‘1) I I I
/
MG {\gw I 151 (N I O.
r I I A - i“-1"! I I r r (a Ia’lr I r
I I I r "T/ I I I I II2 I r 9 I m I I
I I I I I 1 I V I I I I I (2 I E r I
I I 51?? i I lDI'IIIIE/l In II; Al I I NI: TV I r
I I [Ar 1 I I A 1 AU II I r I 4 I\ I I I
I I I I I I (PEA " I I-II ‘ r I I 4 Vi I r I
I I I I EIH 13‘ I nIH 14153 I I I I/ A/fl I I I
r I I I I It I1 [5: T l r I lkh 4K] 1 r I I
I I I r I /‘Q @UT {I I I a I II I I I
r I I IDH 81/“ IDHIB l \K 2 K I r I I I
I I I I 45 it} I; l I I I IW\4 I r .4/ I I r
- I r I I //I I I ,// I I I I/j I / I I I
I I 3H 17 / I l// I I I r /l/4 I V/l r I I
i\\t\/lfi/ I ///I I I I I I I //r I I I l
I r r I / I I I I l I A I / i I I I I
I I I I I I r I I I I I I I I I I
f/
I I I I l I I I I I /4 I I I I I l I
r I I I I I I I i a] I I I I I r I I
x'fl—‘\ / EXPLANATION
‘ m// I I... am VeL-
I I [’7'\1—-‘-afl 3 4 4 I I I l 4 HohRecha-oeVeIue
/ A Hernia-tram
I 3/-—<\ I I I i I I I I I I a m 2% p53
/Ifir J3 I I I I I I I I I L
3 f
I I/ I I I I I I l I I I
\
\ I

 

Figure 14: Recharge Node Values For Groundwater Flow Model

 

 

 

 

43

3 x 10'3 and the average hydraulic gradient indicated by Nicholas and Healy (1988) was

6 x 10“. Since the constant head value for the Illinois and Michigan canal was based on an
estimate, it was raised (two feet) to 580 feet above MSL and recharge values were reduced
to their original values of the third calibration run. The resulting hydraulic gradient was 1
x 10'3 whose decrease proved a much better match for the target heads measured from
dolomite wells throughout the study area and closer to the value indicated by Nicholas and
Healy (1988). Potentiometric surface maps from further runs indicated that an increased in
recharge of 2 orders of magnitude at stream nodes from north of Plot M to the picnic area
provided a better match for target heads proximal to the stream along this interval. This
increased level of head matching provided a more accurate representation of the streams
increased recharge contribution and the higher permeability of streambed sediments. The
new potentiometric surface resulting from this modified recharge pattern also provided a
groundwater migration pathway from the stream to forest preserve well FP5167, which was

needed to facilitate the modeling of the hypotheses.

Once the water levels from the groundwater flow model were close to matching the average
hydraulic heads measured for dolomite aquifer wells in 1984, statistical calibration runs were
performed on the groundwater flow model. The initial calibration statistics from the first
calibration run are presented in Table 3. The statistics from calibration runs are measures
of calibration quality and are computed using residuals. Residuals were calculated by
subtracting the model head from the target (measured) head. The absolute residual mean,

which is an average of the absolute values of all residuals, provides a measure of the average

44

total error of the groundwater flow model.

From the results of his three-dimensional regional groundwater flow model of the study area,
Olympio (1980) indicated that a difference of 1 meter or less between measured and model
hydraulic beads was satisfactory because the seasonal water level changes in the dolomite
were often less than one meter. The maximum residual for the initial calibration of the
groundwater flow model was 2.60 feet which is less than the 1 meter (3.28 feet) determined

by Olympio, so this degree of accuracy was acceptable as a starting point.

Table 3: Summary of Model Statistics from Initial Calibration Run

 

 

 

 

 

 

 

 

 

 

 

 

Statistical Parameters Initial Flow Model Calibration
Residual Mean 0.485612
Residual Standard 1.054892
Deviation

Residual Sum of Squares 12.137548
Absolute Residual Mean 0.891335
Minimum Residual -0.905691
Maximum Residual 2.605039
Observed Range in Head 1.848633
Residual Standard 0.570634
Deviation /Range in Head JL

 

 

 

6.ZS"'!l'

With the completion of the groundwater flow model calibration process, a sensitivity
analysis was performed on the groundwater flow model. The sensitivity analysis was

conducted by varying each model parameter independently by a factor of 1.2x and

45

computing calibration statistics for each variation. Separate sensitivity analysis runs were
conducted for each parameter and for each model layer (ie. storativity for layer 1, storativity
for layer 2, etc.). The parameters varied were transmissivity (x,y,z), leakance, recharge, and
storativity. The results of the sensitivity analysis indicated that the most sensitive model
parameters were recharge and transmissivity in the x and y direction for model layer 1.
Leakance and transmissivity in the x and y direction for model layer 2 were sensitive to a
lesser extent, and the variation of the other parameters for each model layer produced a
negligible effect on the statistical results. Results of the sensitivity analysis were concurrent
with Olympio's (1980) observation that the hydraulic heads in the dolomite was most
affected by variation in recharge rate and transmissivity of the dolomite. Some of the
parameters that were varied during the sensitivity produced groundwater flow model results
that were statistically a better match for the target heads. A set of calibration heads was also
added for layer 2 for a more comprehensive calibration analysis. The results of the altered

parameter values on the statistical analysis are presented in Table 4.

6.3 When

The final calibration process was started, after the sensitivity analysis was completed, by
initial modeling of contaminant transport for the site under steady state conditions.
Contaminant transport model runs were performed using MT3D and the output parameters
from the steady state groundwater flow model for average conditions during 1984. The
results of each run were plotted as potentiometric surface and tritium isocon maps for each

model layer of each model run. The resulting four maps for each contaminant transport run

46

were evaluated against the groundwater flow parameters used to create the flow model and
with regards to the hypotheses that assumed tritium transport from the Red Gate Woods

stream to forest preserve well FP5167 with groundwater flow. For the hypotheses to be

Table 4: Comparison of Groundwater Flow Model Statistical Analysis Results

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Statistical Parameters Initial Flow Final Flow Layer 1 Layer 2
Model Model Final Final
Calibration Calibration Calibration Calibration
Residual Mean 0.485612 0.2921 18 0.074686 0.509690
Residual Standard 1.054892 0.805015 0.771954 0.778184
Deviation
Residual Sum of Squares 12.137548 13.201607 5.413422 7.778184
Absolute Residual Mean 0.891335 0.706828 0.652895 0.760760
Minimum Residual -0.905691 -1.193472 -1.193472 -0.669058
Maximum Residual 2.605039 1.710400 1.243420 1.710400
Observed Range in Head 1.848633 2.180969 1.604248 2.180969
Residual Standard 0.570634 0.369109 0.481 194 0.356806
Deviation /Range in Head |

 

 

tested, contaminant transport of tritium from recharge proximal to the stream had to migrate
beyond FP5167 and short of F P5215. Tritium was added with recharge to the model at the
stream nodes earlier identified as having recharge values of 2 orders of magnitude greater
than the area recharge applied over the entire model area. The MT3D units of concentration
were lbs/f9, due to flow model parameter unit choices of days and feet, and the amount
added to the recharge was 100 lbs/fi3. The results of the first few contaminant transport
model runs indicated that limited migration was progressing outwards from stream nodes in

layer 1, but tritium migration with groundwater flow was occurring beneath the stream nodes

47

in layer 2 as small isolated pockets instead of migrating towards well FP5167.

The concentration of tritium in the recharge applied to the stream nodes was the first
parameter to be varied to address the migration problem. Over a series of transport runs the
tritium concentration was increased to a final value of 1,000,000 lbs/fi3. The increase in
concentration increased tritium isocon values in layers 1 and 2, but failed to broaden the
extent of migration in either layer. The next step was to increase the recharge associated
with tritium concentration by a factor of 3x in an effort to increase the hydraulic impetus for
migration, but this method also failed to increase the extent of tritium migration in either

model layer.

With variations in recharge and concentration failing to increase the extent of tritium
migration, variations in the aquifer parameters were implemented in a series of contaminant
transport model runs. Since leakance was the key parameter linking groundwater migration
pathways between the two model layers, it was the first parameter to be varied. The most
important leakance nodes were beneath the stream where tritium migration would progress
from the source atop layer 1 into layer 2 below. Values of leakance were increased at stream
node locations over a series of transport runs. Figure 5 was used as a guide for choosing the
stream nodes at which the highest amount of leakance should occur to facilitate tritium
transport to well F P5 167. Golchert and Sedlet (1985) indicated that a sand and gravel lense
is known to underlie the stream proximal to the locations of DH-9 and DH-lO and may be

connected to the dolomite in this area. This connection is evidenced by the inability to

48

properly seat the casing of DH-9 to the drift-dolomite interface and by the highest tritium
values being detected in DH—9 and DH-lO which would require a supply of water with high
tritium concentrations from the source area in the drift materials beneath the stream. Stream
nodes in this area were given a leakance value of higher magnitude to represent this
condition. The results of increasing the leakance values beneath stream nodes did increase
the extent of tritium migration in layer 2, but the maximum distance of migration only

extended about half the distance from the stream to well FP5167.

Until this point in the contaminant transport model calibration process, variation of model
parameters had focused on sensitive parameters whose determined values had a variability
built into them due to the method of measurement and functional representation. The only
sensitive parameters left to be varied were the transmissivities of layers 1 and 2. The degree
of unknown fracturing and interpreted thicknesses of layer 1, compared with the detailed
pumping test information obtained for layer 2, made layer 1 the first choice for variation.
Through a series of separate transport runs, the transmissivity (x,y,z) of layer 1 was both
increased and decreased. The results of this variation indicated that an increase in the x and
y transmissivity by 2 orders of magnitude and the z transmissivity by 1 order of magnitude
facilitated the migration of tritium from the stream to well FP5167. Variation of the x and
y transmissivity of layer 2 showed that increasing these parameters by 3 times the original
value extended the migration of tritium past well FP5167 and short of well FP5215’, thus
meeting the earlier stated criteria by Golchert for boundary conditions. Recharge

concentrations of tritium were reduced back to the original value of 100 lbs/fl3 with no effect

on the extent of tritium migration. An observation node was placed at the location of well
FPS 167 in layer 2 to record values of tritium concentrations during the contaminant transport
modeling process. Increased concentrations (1,000 to 100,000) are recommended for the
testing of hypotheses because concentrations measured at the observation node were of

smaller magnitudes that would make entry, evaluation, and manipulation difficult and

visually challenging.

A final statistical analysis was performed on the groundwater flow model whose parameters
had been altered during the calibration of the contaminant transport model. The results of

this calibration are presented and compared to the last statistical analysis of the flow model

in Table 5.

49

Figure 5: Final Calibration Statistics for the Groundwater Flow Model

 

 

 

 

 

 

 

 

 

 

Deviation /Range in Head

 

 

 

Statistical Parameters Final Flow Model Final Transport
Calibration Model Calibration
Residual Mean 0.2921 18 0.349744
Residual Standard 0.805015 0.772095
Deviation
Residual Sum of Squares 13.201607 12.932129
Absolute Residual Mean 0.706828 0.704379
Minimum Residual -1 . 193472 -0.894644
Maximum Residual 1.710400 1.555310
Observed Range in Head 2.180969 2.063965
Residual Standard 0.369 109 0.374083

 

 

50

Comparison of the two sets of statistical analyses indicates that the parameters resulting from
the final calibration of the contaminant transport model are a better match for the measured
hydraulic heads of dolomite aquifer wells. The residual sum of squares, maximum residual,
and the absolute residual mean were all reduced indicating that the final alteration of
parameters formed a better match for measured heads and that the amount of model error was

reduced.

The wells completed within the dolomite aquifer at the study area were constructed by
similar methods, but to different depths (see Table 6). The wells (DH and FP) were cased
ofi' to the drift-dolomite contact and then continued on as open boreholes into the dolomite
aquifer. With the different depths of open boreholes intersecting different and multiple zones
within the dolomite bedrock, water level data from a few wells was difficult to match with
the hydraulic gradient. Potentiometric surface maps constructed for the area by the
groundwater flow model often could not match measured hydraulic heads for these few
wells. Contours of measured hydraulic heads that included these wells ofien included bulges
and sporadic variations that are not characteristic of groundwater flow within a well
developed fracture system, such as the one found at the study area. Golchert (1988) reported
that dolomite aquifer wells DH-6 through DH-9 and DH-10 are open to the overlying drift.
These wells would not be representative for hydraulic head and tritium concentration
measurements of the dolomite aquifer. Statistical analysis of calibration runs that included
the hydraulic head values for these few wells produced higher residual values. The resulting

sum of squares from statistical analysis of calibration runs that didn't include these few wells

51

would be a lower number reflecting the greater degree of accuracy created by the

groundwater flow model in matching the measured heads of representative wells.

Table 6: Dolomite Well Construction Elevations

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l——=I= =fi=================r==n

Well 1]) Top of Casing (ft) Top of Bedrock (ft) Borehole Bottom
(ft)

DH-l 743.5 572 527

DH-2 721.2 568 519

DH-3 679.5 556 505 I

DH-4 674.6 565 394

DH-S 659.6 585 558

DH-6 656.5 583 572

DH-7 665.6 587 578

DH-8 658.2 586 570

DH-9 656.3 581 578

DH-10 645.9 565 545

DH-l 1 657.0 573 426

DH-12 658.2 569 427

DH-l3 658.9 571 I 428

DH-14 653.2 571 427

DH-15 660.7 573 489

DH—16 657.0 572 485

DH-l7 656.0 571 484

FP5167 651.7 567 505

 

 

52

Figures 15 and 16 present the final potentiometric surface maps for layers 1 and 2 resulting
from the final calibration of the contaminant transport model. The final extent of tritium
migration for layers 1 and 2 are presented in Figures 17 and 18. The resulting final

parameter values for the groundwater flow model are presented in Table 7.

Table 7: Final Groundwater Flow Model Parameter Values

 

 

 

 

Model Top Bottom Transmissivity Storativity Leakance
Layer Elevation Elevation
Layer variable variable Tx = 12096.0 ftZ/day 0.00016 variable
1 556-600 546-590 ft Ty = 12096.0 ftz/day 0.04572 -
ft T2 = 403.2 ftz/day 1000 day‘I
Layer 544 ft 543.7 ft Tx = 7811760 ftz/day 0.000022 not
2 Ty = 7811760 ftz/day applicable
T2 = 26039.232
fiz/day

 

 

 

 

 

 

7.0 Testing of Hypotheses

Two hypotheses were proposed to account for the tritium fluctuations at forest preserve well
FP5167: (1) the result of fluctuations in recharge through the drift bring varying amounts of
tritium into the aquifer or (2) varying fluxes of groundwater flow in the dolomite cause the
dilution of a constant flux of tritium from the drift and lessen the concentrations of tritium
in the aquifer. Hypothesis number 1 was modeled by using the hydraulic gradient and
average precipitation for each month to represent the fluctuations in recharge through the
drift. Hypothesis number 2 was modeled by using the average monthly precipitation for the
year to represent a constant flux of tritium fi'om the drift and allowing the hydraulic gradient
for each month to represent the variation in fluxes of groundwater flow. Three years were

chosen to model the two hypothesis; wet (1983), normal (1985), and dry (1986) conditions

 

 

 

 

 

 

 

 

 

 

 

 

 

ILLINOIS AND MICHIGAN CANAL
580 .0
5'59
4‘
/‘p\\\
m
O
O
7
A
/N——— m
581
3»
N We NEW “34
5160‘ %' '0 a
7
5167 3
A ”"2 DH I4
DR‘Is
on ISA DH ‘0“ 9
‘ H
582 .4 on Is
A w . 0.. .
A
DH 5
A
OH 7
A
DH 4
A
583. I _
I" O 100 290
<
a. LL44
*5 Fear
0
LL
on 3 EXPLANATI
//,DK SISQ
(O: ‘ ‘oren Preaorvo loll
A7 ' 0:5 Dole-Ito Hole I011
550 0 Ground-clot Con-Ion IFTI
583. 5
/ 0H 2
A

 

 

 

 

Figure 15: Potentiometric Surface Map ForLayer 1

 

 

 

 

 

 

 

 

ILLINOIS AND MICHIGAN CANAL
580.0
5158
5159
‘ A
/RCHER AVENUE \
5 81 .O

 

 

 

La \
'o
o
7
.4
QM ‘“
é
. O
o
3»
’o/IRK/
581 7 N “‘“

 

 

 

  
 

 

 

 

 

 

 

 

 

 

,ww Ti “
VT Io r331
5167 7i
A m": DH 14
ON IS
on 13‘ OH 9
DH II
5215 ”"6 on I7 I
‘ ‘
582.4
100 200 300
l.__L__L__J
FEET
583.I
EXPLANATION
5|67
‘ For." Prouerve Ioll
DH 5
‘ Dole. to Hole Nell
580.0 Ground-over Contour lFTl
lfigune16£PounnunnehicSmdhcehhuprrIawerz

 

 

 

 

ILLINOIS AND MICHIGAN CANAL

 

 

$215

5159

5150

 

ARCHER AVENUE

PARKING
A
SIOO
sun
A
on 7
A
on 4

c:
I .
a

Q FOOTPATH .

AREA
mIz
A
DH IS
mlfil‘ ‘,
W11
DH 16
we ”6
A
ms
A

\

 

 

—-os—-
a”
A

ms
A

CONIOUI

 

Expicnorion

 

lrivnna Inocan (lb/fr 1
Forest Protervo loll

Dole-II. Hole Ioll

3
INIEMAL - z Huh I

 

 

 

Figure 17:Groundwater 'h-itiumlsoeon Map ForLayer l

 

 

 

 

 

 

ILLINOIS AND MICHIGAN CANAL

SIS!

    
  

    

.. Wk (1-
.. .1 @®\

, II|
C‘I I.I

at

, e
.1 , ’I

511!

I Do 200 300

 

EXE I omit I on
3
—° 5— YrIIIll lunch (LI/Fl 1
5:, FVPUII 'vnlnrvo loll
ON 5
‘ Deleon. .I. loll

 

3
(:0le INTERVAL - 0.7 “.3157 I

 

 

 

 

Figure 18:Groundwater 'Drifiumlsocon Map For-layer2

 

 

S7

with regards to yearly precipitation amounts (see Figure 19). The resulting model
concentrations of tritium at well F P5167 were recorded and graphically compared and
contrasted against the measured tritium concentrations (see Appendix II, Measured Tritium
Concentrations at Dolomite Aquifer Wells) at dolomite aquifer well FP5167. The
concentration graph of the hypothesis that best matched the graph of measured tritium data

thereby became the hypothesis that was most correct.

Two model nms were performed under transient conditions for each month that was modeled
to determine the tritium concentration at well FP5167 for each hypothesis. The resulting
hydraulic heads and tritium concentrations for each month and hypothesis were saved and
used as initial heads and concentrations for the model runs of each hypothesis for the next
month. For each month and hypothesis, the transient groundwater flow model was run to
obtain groundwater flow parameter values, which were then input into the run of the transient
contaminant transport model for each month and hypothesis. At the end of each transport
model run hydraulic heads and tritium concentrations were saved for the next months
models, and the concentration at the observation node (well F P5167) was recorded.
Modeling was conducted for each month of the three years mentioned above that had the

corresponding hydraulic head, precipitation, and measured tritium data.

The Illinois and Michigan canal hydraulic head was fixed at 580 feet for each model run and
the southern constant head boundary was represented by the measured hydraulic head of DH-

2 (see Appendix III - Water Levels in Dolomite Aquifer Wells) for the month of interest.

 

ggméauouflnafimisgfigfiq

 

 

mag .59. owe. mmmr vmmr mam? Nam? Ems one.

 

(we) uoIIeIIdIoaJd lenuuv

 

 

 

 

 

 

 

 

 

59

Constant head values were the same for each of the two model runs for the month in
question. Monthly precipitation values were altered to account for evapotranspiration and
monthly yearly averages were computed (see Appendix I - Precipitation Records for the
Palos Forest Preserve Area). Precipitation was added to the model at varying locations and
magnitudes as discussed in earlier sections of calibration and sensitivity analysis. The
average precipitation value for the month was used for each model run of hypothesis number
1, and the average monthly precipitation value for the year was used for each model run of
hypothesis number 2. The tritium concentration used in the recharge was set at 100,000
lbs/ft3 to achieve concentration numbers of higher magnitudes at well FP5167 that were

easier to manipulate, evaluate, and visualize.

The resulting tritium concentrations for modeling each hypothesis and the monthly average

measured tritium concentration at forest preserve well FP5167 are presented in Table 8.

8.0 Summary and Conclusions

Low-level radioactive waste buried near Plot M has contaminated the soils and groundwater
beneath the Palos Forest Preserve with tritium. Since 1976, seasonal fluctuations of tritium
concentrations have been observed at forest preserve well FP5167. Tritium has migrated
from the Red Gate Woods Stream below Plot M into the dolomite aquifer to the area
proximal to well FP5167. Two hypotheses have been proposed to explain the seasonal
fluctuations of tritium in dolomite aquifer well FP5167: (1) the result of fluctuations in

recharge through the glacial drift bringing varying amounts of tritium into the dolomite

Table 8: Measured and Modeled Tritium Concentrations at Forest PreserveWell FP5167

60

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Month-Year Measured Tritium Hypothesis # l Hypothesis # 2 J
Concentration Concentration Concentration
Jan 83 1.00 2002.3 1950.9 ll
Feb 83 1.65 2018.5 1950.9
Apr 83 1.65 1849.2 1950.9
May 83 1.45 1809.8 1950.9
July 83 0.28 2023.8 1950.9
Aug 83 0.55 2030.9 1950.9
Sept 83 0.39 1835.8 1950.9
Nov 83 4.85 1768.6 1950.9
Dec 83 4.10 1913.8 1950.9
Jan 85 1.7 2040.6 2000.1 ll
Mar 85 1.2 1784.8 2000.1 1
Apr 85 0.1 2044.5 2000.1
May 85 0.18 2015.6 2000.1
Jun 85 0.28 2027.0 2000.1
July 85 0.16 2007.4 2000.1
Aug 85 0.37 2059.0 2000.1
Sept 85 0.65 2036.4 2000.1
Oct 85 0.55 6892.6 2000.1
Nov 85 2.40 2113.7 2000.1
Jan 86 0.22 2055.8 2070.9
Feb 86 0.21 1946.5 2070.9
Mar 86 2.3 2111.6 2070.9
Jul 86 0.36 2023.0 2070.9
Oct 86 3.15 1951.6 2070.9
Dec 86 2.85 2113.6 2070.9

 

 

 

 

 

61

aquifer or (2) varying fluxes of groundwater flow in the dolomite aquifer diluting a constant
flux of tritium from the drift and varying the concentrations of tritium in the dolomite
aquifer. The first hypothesis represents a variation in localized recharge while the second

hypothesis represents a variation in regional recharge.

A three-dimensional finite-difference groundwater flow model was constructed to model the
hydrogeological conditions at the site. The model parameters obtained from the background
review of published material were adjusted during the calibration process to create the best
representation of physical parameters with mathematical equations. The output parameters
from the groundwater flow model were used as the input parameters for the contaminant
transport model. Parameters for the groundwater flow model were adjusted with regards to
the contaminant transport model that was calibrated under the same conditions as the

groundwater flow model.

The results of the contaminant transport modeling of each hypothesis are graphically
compared against the measured tritium concentrations at forest preserve well FP5167 in
Figure 20. Measured and modeled tritium values were normalized to facilitate the graphical
comparison process. Based on the assumption that each hypothesis was accurately
represented by the choice of model parameters, the hypothesis that is the best graphical
match for the measured tritium data would also be the most correct one. Since the graph of

hypothesis number 1 represents the varying of tritium concentrations with recharge and the

 

 

 

A A A A A A

 

 

1 : .. 1. .. ..
II
0.8 . ----------------- [I ---------------------------------------
O
.\

------- ------.. ......................................
I

 

   

 

o . . . . R . . . . . . .
Ian's: May83 Sept83 Jan‘ss Mayss Aug85 'Jan'86' ‘IuIsIs

[+ Measured Values + Hypoth. 2 Values]

 

 

 

 

 

 

0.8 - .................

J
I
I
0.64 ................ (IL/N.

0.4. ..................... I ..............................
h I I /\
024.0 ....... I ...... I ...... I ................. PHI ........ L .....
I , I ,

\9/5; I A” I _,.

o I ' : ‘ . . ' .3 : ' '
Jan83 May83 Sept83 Jan85 May85 Aug85 Jun“ Jul“

L-o- Measured Values + Hypoth. 1 Values]

 

 

 

 

 

 

figmflQmpafisonofMeamedvaHypotheeisConcenu-aflonCurves

 

 

63

graph of hypothesis number 2 represents the varying of tritium concentrations with
groundwater flow, graphical comparison suggests the variation of modeled tritium
concentration fluctuations is better explained by variable recharge rates rather than variable
groundwater flow rates. The trend of rising tritium concentration values for hypothesis
number 2 do roughly parallel the general rising trend of measured tritium concentrations,
indicating that variable groundwater flow rates may also be a contributing factor to the

tritium fluctuations at well FP5167.

It is not the magnitude (numerical values) of the graph curves and peaks in Figure 20 that are
important for comparison and evaluation. The shape of the curves and location (month-year)
of the peaks determine which hypothesis best matches the measured tritium fluctuations at
well F P5167. The concentration curve for hypothesis number 2 presents a series of three
plateaus, which represent constant tritium concentrations for each year. The absence of
fluctuation in model results for this hypothesis indicates that varying groundwater flow rates
are not the sole cause for fluctuations in tritium concentrations in the dolomite aquifer. Since
the same value for recharge was used for each month of each specific year, the constant
values of tritium concentrations indicate that variations in recharge amounts are most likely
the major control with regards to tritium concentrations in the dolomite aquifer. This
conclusion agrees with the statement by Nicholas and Healy (1988) that "Seaso‘181

fluctuations in tritium concentration in well 5167 are caused by variable of recharge I0

 

dolomite."

64

The overall graphical shape of hypothesis number 1 is a fair match for the overall graph
shape of the measured tritium data fi'om well FP5167. The normal precipitation year (1985)
provides the best match between measured tritium values and modeled tritium values for
hypothesis number 1. Some of the peaks and valleys are slightly offset which may be
attributed to differences in modeled and actual tritium migration times. The slight offsets
may also be due to the effect of initial heads and concentrations from a month not directly
preceding the modeled month (because of a lack of complete data for the prior month). The
area of the graphs corresponding to May of 1985 to December of 1986 present an area of
fairly constant values followed by peaks and valleys that are roughly similar between the

measured tritium data and the modeled tritium data from hypothesis number 1.

The changing of measured aquifer parameter values during the groundwater flow and the
contaminant transport model calibration process was performed to create a set of three-
dimensional models that facilitated the modeling of tritium migration from the Red Gate
Woods stream to the dolomite aquifer in the vicinity of forest preserve well F P5167.
Statistical analysis of the groundwater flow models ability to match measured hydraulic
heads of dolomite aquifer wells in the study area was performed to provide a control on the
accuracy of the models with regards to error. Error is inherit in the methodology of
parameter measurement so variation of measured parameter values is not necessarily a false
representation of hydrogeological conditions at the site. Some of the parameters obtained
from the literature review were varied in excess of two orders of magnitude. However, many

of these variations were performed in order to represent a set of circumstances that were not

65

accounted for during parameter measurement. Statistical analysis indicated that after the
final variation (calibration) of model parameters, the groundwater flow model had greater
accuracy with regards to measured hydraulic heads and less error as a result of these changes

to hydrogeological parameter values.

Measured hydrogeological parameter values were also varied to achieve transport of tritium
from the source area to the area proximal to well FP5167. The omission of the overlying
glacial drift, due to the lack of needed measured parameters and data, could have removed
a possibly significant portion of the tritium migration pathway. If the drift had been included
with its lack of measured physical data, calibration would have been based on unknown
values and error control would have been severely compromised. The missing contribution
of tritium migration in the drift had to be accounted for by adjustment of the hydrogeological
parameters of the Silurian dolomite. With a complete data set for the glacial drift and its
inclusion in the groundwater flow model, the peaks and valleys of modeled trititnn
concentration data may have more closely matched the location (month-year) of the
measured tritium concentration data. The alteration of the hydrogeological parameters of the
dolomite to compensate for the missing glacial drift was an assumption that had to be made
to maintain the calibration accuracy of the flow model and should not have significantly
affected the ability of the flow and transport models to adequately address the above

mentioned hypotheses.

An interesting observation can be found by looking at Figures 10 and 20 and comparing

66

recharge amounts with corresponding tritium concentrations at FP5167. Comparison
indicates that increased amounts of recharge result in lower concentrations of tritium in the
dolomite aquifer. It would be reasonable to theorize that if recharge was the source and
migration pathway for tritium into the dolomite aquifer, that increased amounts of recharge
to the dolomite aquifer should bring increased concentrations of tritium. A migrational time
delay does not seem to be the answer due to the fact that tritium concentrations following
1983, a very wet year, are lower than those following prior years that were had less total
recharge. A more probable explanation is that the ability for the tritium source to mix with
recharge is associated with a kind of reaction rate that when exceeded, the excess recharge
not involved with the reaction serves to dilute the tritium that has gone into solution. Years
of precipitation that meet or fall below the limitations of the reaction rate bring recharge to

the dolomite aquifer that is more concentrated with respect to tritium.

It must be noted that the conclusions formed in this section are based on the assumption that
the groundwater flow model created was an accurate representation of the hydrogeological
parameters at the site and that the modeling method used to represent each hypothesis was
also an accurate representation of the hydrogeological conditions that the hypotheses are

based upon. In groundwater flow modeling no one solution is unique, there is an infinlte

number of parameter value combinations that could form a working solution. The statiSticaI
analysis and large number of model runs performed as part of this study were conducted to

reduce potential error and provide a control on the representativeness and accuracy of the

model created.

APPENDICES

APPENDIX I

Precipitation Records for the Palos Forest Preserve

 

67

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

88¢ 88¢ 8~ 88m

88¢ £~m¢ 88 88%

88¢ 88¢ 2.3 8-....-

88¢ 82¢ 8.... 8.28

88¢ 82¢ 8.8 8-3..

28¢ 28¢ 8s 8-..?

8-¢ 88¢ 8.2 8-8.2

88¢ 38¢ 8.. 8-8“.

~32 R2¢ m2~¢ 28¢ 8s 88..
88¢ 28¢ .28 8.8g

83¢ .. .88 8.8 8-32

88¢ 82¢ 8... 8-80

88¢ m2~¢ Mo 8-8m

82¢ 88¢ 88.: 8.84

88¢ 85¢ 8... 8-..;

22¢ 88¢ 3.2 8-8..

88¢ 28¢ 8.8 8%....

88¢ 88¢ 8.: 8-..?

88¢ 88¢ 3 8.85.

82¢ 82¢ 88 8-8“.

2.8 88¢ 88¢ 38¢ 28 8-8..
82¢ ~2~¢ I88 8-8a

88¢ 38¢ 3 8-32

32¢ 38¢ 2... 8-60

88¢ 88¢ 8.2 888

88¢ 22.8 8.3 8-8.3

88¢ 28¢ a: 8-.....

88¢ 88¢ 8... 88¢

88¢ 88¢ 2... 8.8...

88¢ 8¢~¢ 8.8 8-8....

88¢ 82¢ 8+ 8.8...

88¢ 22¢ 8... 88".

8.8 ¢82¢ 88¢ 88¢ 2..~ 8-8..
.5358 88:8; .598... .92 258.... 8.4.8... 3.2.5 ..Sflcogaaaa 89.83.58... 23

 

mammoi 820d mafia on. .2 2:83. 5:28.32... um flan...

 

68

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0000.0 02402.0 0 00.5...

0000.0 0NmN.0 00 N 00.85.

0000.0 33.0 on m 00-5.4.

005.0 050.0 20 N_. 00.55.

58.0 vmmNd 0N 0 00-00“.

K20.- 0090 £000 59.0 on m 00-2.2.
205.0 va0 Nv N 5.85

052.0 SNNd K 0 v0-52

_.0N_-.0 mNmN. 0 0 v0-60

2 2 3 88 ¢ 8 a 8-8m

0000.0 03. _. .0 mm m $534.

0000.0 $8.0 _. 0 v8.2.

0000.0 0002.0 02 m 3.5:.

302.0 0300 v 5 v0.82

0N0; 3.8.0 3 02 v0-22.

005.0 oomNd N0 N V0-85.

34.00 mm 3 0 mm m V0-30...

00.0w Nero 008.0 0000.0 N0 N v0-2.2.
an? 0 oomNd no N 00-80

Ema 0 vmrv 0 0 NF 00.52

vaN 0 mnom 0 N 3 00-80

88 ¢ 82. o E 2 8.86

88¢ ~82 o 8. m 8-8....

0000.0 980 0 0 02 00.2.

0000.0 vam 0 3 02 00-0....

88¢ 28 o m 2 8-8.2

mowed .53. 0 mm. 0.. 00-56.

EN 0 oomN 0 v0. 0 8-85.

200.20 02.2.0 SUN 0 00-8“.

onvNNF 35.0 2000.0 300.0 00 F 00-02.
.28 .0 008.0 00.: Nwéoo

Bde 53.0 00.: Nm->oz

008. 0 002. 0 05v N080

.Eoszm .3005 ..ac>E. £902.". .m>< 25:05: at. £095 ..€< .ona>m_ coo: 5283395 0.03 cos—£5005 Baa

 

 

.8880 a 2.5

 

 

69

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$000 0000.0 00.0 00-00....

0000.0 0000.0 00.0 00.00".

00.00 000 0.0 0000.0 0000.0 00..- 00-00..
00000 030.0 00.0 00.80.

0000.0 0000.0 00.0 00-52

0000.0 0000.0 3.0 00.60

0000.0 0000.0 00.0 00000.

0300 000.1 0.000 00.03.

0000.0 00.0.0 .000 00.3

0000.0 0000.0 00.0 00.00..

0000.0 300.0 00.0 00-00....

030.0 0000.0 00.0 00..0<

0000.0 0000.0 3.0 00-3....

0000.0 0000.0 00.0 00.00“.

00.000 003.0 2: 0.0 0000.0 00.0 00-00..
0000.0 0000.0 00.0 00-80

000.0 000 0.0 00.0 00-52

0000.0 0000.0 00.0 00-80

0000.0 0000.0 00.00 00-000

0000.0 0000.0 00.0 00-004.

0000.0 0000.0 00.00 00-....-

0000.0 0000.0 00.0 00.5..

0000.0 0000.0 0.0 00.00...

0.00.0 0.00.0 0.0 00-..?

00.00 0000.0 00.0 00-5.2

00000 030.0 00.0 00.00”.

00.00 0000.0 0000.0 0000.0 00.0 00-00..
0000.0 0000.0 00.0 00-80

0000.0 0000.0 00.00 00-82

000.0 003.0 00.00 00-60

0000.0 0000.0 000 00.00%

0000.0 0000.0 00.0 00-02

- 0000.0 0000.0 00.0 00-.....

...-.3800 .9020. 03> 30.0.300- .m>< 25:22 90.3020 .25 .agm cow: 00.08.2020. .53 5.3.3.020! E

 

A3080 0 030.0

 

70

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0000.0 58.0 F 0 00-000
00000.0 3.00.0 00 00 00.32
000N0 00Nv0 00 N0 00-000
0000.0 0000.0 00 0 00-000
0000.0 0NON.0 00 0 00-030
0000.0 00NN.0 00 0 00-3..
0000.0 0000.0 .00 N 00-02.
0000.0 3000 N0 0 00:00.2
NON0. 0 N05. 0 0N. 0 00-090
ESE—.0 .3005 000>E§020 .m>< 25:05. 3:. £00.."— ..€< .0003“: 300: 00.08.3000... :00. 5.303.000... 20o

 

00.0080 0 2000

 

APPENDIX II

Measured Tritium Concentrations at Dolomite Aquifer Wells

71

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.0 0.0 0.0 0.0 0.0 0.0 00-00M.
00.0 00.0 00.0 00.0 00.0 0.0 0.0 00-03.
00.0 00.0 E 00.0 0.0 0.0 0.0 00-.....
00.0 00.0 0.0 0.0 0.. 0.0 00-000
0.0 0 0.0 00-00.).

0.0 00.0 0.0 0.0 0.0 0.0 0000.0
0.0 00.0 0.0 0.0 0.0 0.0 00-00....
0.0 0.0 0.0 0.0 0.0 0.0 00-000
0.0 0.0 0.0 00-00..

0.0 00.0 00.0 0.0 0.0 0.0 00.80
0.0 0.0 0.0 00-062

0.0 00.0 0.0 0.0 0.0 0.0 00-80
0.0 0.0 0.0 0.0 0.0 0.0 00-00m.
0.0 0: 0.0 0.0 0.0 0.0 00.03.
0.0 00.0 0.0 0.0 0.0 0.0 00-30
0.0 3 0.0 0.0 0.0 0.0 00-000
0.0 0.0 00.0 0.0 0.0 0.0 5.000..
0.0 E 0.0 00-00...

0.0 0.0 00.0 0.0 0.0 0.0 00.00....
00.0 0: 0.0 0.0 0.0 0.0 00-000
0.0 0.0 0.0 00-00..

00.0 00.0 00.0 0.0 0.0 00.80
0.0 00.0 00.0 P 0.0 00-32
00.0 E 00.0 0.0 0.0 00-80
0.0 0: 00.0 0.0 0.0 00-000.
0.0 0.0 00-00<

0.0 00.0 00.0 0 0... 00-.....
0.0 0: 00.0 3 0.0 00-000
0.0 00.0 0.0 0.0 0.0 00-00.).
0.0 00.0 0.0 0.0 0.0 0 0000.0
0.0 8.0 00.0 0.0 0.0 0000.2
0.0 0: 00.0 0.0 0.0 0.0 00-000
00 0.0 0.0 00-00..

03.0 0-:0 0.10 0.00 0.00 0.10 3.0 0.20 0-20 00.0 00.0 00.0 0.00

 

0__0>> 000.090 00:00.00 .0 0000000000000 0000...... 00000005. .00 0.00.0

72

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

000 :0 0.0 0.0 00.0 0.0 0.0 00.0 00.0 00.0 00.0 0000 00-000
0.0 0.00 0.0 00.0 0.0 0.0 00.0 00.0 00.0 00.0 00.0 00.00.).
3.0 00 0.0 0.0 00.0 0.0 0.0 00.0 0.0 0.0 03 0.0 0000.0
00.0 0.00 0.0 0.0 00.0 0.0 0.0 0.0 00.0 00.0 00.0 0000.2
00.0 0.0 000.0 00.0 00-000

50 0.00 0.0 0.0 0.0 0,0 0.0 00.0 0.0 0.0 00.0 00.0 00-000
00.0 0.00 0.0 0.0 00.0 0.0 0.0 00., 0.0 0.0 000.0 0.0 00-000
00 000.0 00.0 00-002

0: 0.00 0.0 0.0 00.0 3.0 0.0 00.0 0.0 00.0 000.0 0000 00.000
00.0 00.0 0 0 00-000

000 :0 0.0 0.0 00.0 0.0 0.0 00.0 0.0 00.0 00.0 0 0 00-000.
E00 0.00 0.0 0.0 00.0 0.0 0.0 Be 00.0 0.0 0.0 0.0-.00
010.0 0.00 00.0 0.0 00.0 0.0 0.0 00.0 0.0 0.0 0.0 0.000
00.0 0.00 0.0 00.0 00.0 0.0 0.0 00.0 00.0 0.0 00.0 0.0 00-005.
0.0 0.00 0.0 0.0 0.0 00.0 0.0 0: 00.0 00.0 0.0 0.0 0000,...
00.0 0.0 0.0 0000.2

: 0.00 0.0 0.0 0.0 0.0 0.0 00.. 00.0 0.0 0.0 0.0 00-000
00.0 0.0 3 00-00..

0.00 0.00 0.0 0.0 0.0 0.0 00.0 0.0 00.0 5.0 0.0 3 00-000
0.3 0.00 0.0 0.0 0.0 00.0 0.0 00.0 00.0 00.0 0.0 0.0 00-002
00.0 0.0 3 00-000

0 2.0 0.0 0.0 0.0 00.0 00.0 0.0 0.0 00-000
0.00 0.0 0.0 0.0 0.0 00.0 0.0 0.0 00-00<

0.00 0.0 0.0 0.0 0.0 00.0 0.0 0.0 00000

0.0 0.0 00-000

0.00 0.0 0.0 0.0 0.0 0.0 0.0 0.0 00.00.).

0.00 0.0 0.0 0.0 0.0 0.0 5 0.0 00.00...

0.0 0.? 0.0 00.00....

0.: 0.0 0.0 0.0 0.0 0.0 00.0 0.0 0.0 0.0 5 00-000

0.0 0.1 0.0 00.0 0.0 0 00-000

0.0 0.0 0.0 0.0 00.0 0.0 00.0 0.0 0.0 0.0 00-000

0.0 0.0 0.00 00-82

0.00 0.0 0.0 0.0 0.0 0.0 0: 0.0 0.0 0.0 0.0 00.05

00.10 0000 0.20 0-:0 0.:0 0.20 100 0-20 0.:0 00:.- 00.0 - 003. 000

0200008000 E220... 0050005. ”or 0.00..-

 

73

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.0 20.2 0.0 0000.2

0.0 0.0 0.0 0.0 0.2 0.0 0.0 000.2 0.0 00-000

0.0 0. 20 0.0 0.0 :3 00.0 00-000
0.0 20.0 000.0 00-000

0 00 2.0 0.0 000.0 00.0 00-82
0 0.00 0.0 0.0 000.0 0.0 00.80
0.0 0.00 0.0 0.0 0 0.0 0.0 3 0.0 0.0 000.0 00: 00-000
0.0 0.0 0 00-0000

0 :0 3 0.0 020.0 0000..
0.0 0.00 0.0 0.0 3 0.0 0.0 0.2 0.0 0.0 000.0 00.0 0.000
0.0 000.0 000.0 00.00....

0.0 00 5 0.0 20.0 0.2 00.000
0.0 0.00 0.0 0.0 2 0.0 0.0 0.2 0.0 00.0 00.0 2.0 0000.2
0.0 0. 00 0.0 0.0 3.0 00.0 0.000
0.0 00.0 0.0 00-000

I00 0.0 0.0 0: 20.0 0.0 00.2 0.0 00.0 00.0 00.0 00-000
0.0 000.0 0.0 00-002

0.0 0.0 0.0 00.0 0.0 0.0 00.2 0.0 0.0 20.0 02.0 00.000
0.0 000.0 20.2 00.000_

00.0 20.0 0.0 00.0 00.0 00-00<
00.0 00.00 0.0 0.0 R0 0.0 0.0 00.2 0.0 00.0 000.0 000.0 00000
0.0 00.0 0.0 00-000

:0 00.00 00.0 0.0 000.0 000.0 00.00.).
00.0 000.0 00-.0<

3.0 00.00 0.0 0.0 00.0 0.0 0.0 20.2 0.0 0.0 000.0 0.0 0000.2
W00 00.00 0.0 0.0 00.0 0.0 0.0 0: 0.0 0.0 00.0 0.0 00-000
00.0 0.0 0.0 00.0 0.0 0.0 00.2 0.0 0.0 00.0 000.0 00-000
0.0 00.0 0.1 00-000

00.: 0.00 0.0 0.0 00.0 0.0 0.0 00.2 0.0 :00 000.0 0.0 00-32
0.02 0.0 0.0 00.0 0.0 0.0 :2 00.0 2 00-000
0.02 2.00 0.0 0.0 00.0 0.0 0.0 00.2 00.0 00.0 0.0 00.0 00-000
0002 0.20 0.0 0.0 20.0 0.0 0.0 00.2 0.0 0.0 3.0 000.0 00-00<
v0.0 0 0.0 0.0 00.0 00.0 0.0 E 0.0 3.0 20.0 00.0 0.0-.00
2.20 0-20 0-...0 E00 0.10 0-:0 3.0 0....0 0.10 0000 0000 00.0 0000

 

 

 

A3008 2 030%

74

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

00000-0000 0000.000 000000 0000:0350 000 0z< .000 0000.
0.0 00.0 000.0
0.0 0.: 0.0 0.0 0 0.0 0.0 0.0 0.0 0.0 00.0 000.0
0.0 0.: 0.0 0.0 00.0 00.0
0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 00.0 0.0
0.0 000.0
0.0 0.00 0.0 0.0 0 0 0.0 0.0 0.0 0.0 000.0 0.0
0.0 00.0 0.0
0.0 00.0 0.0
00-10 0-10 0-:0 0.10 0.10 0.10 0.20 0.20 0.10 00 00 0000 0000

 

 

 

00.00000 00 2000

APPENDIX III

Water Levels in Dolomite Aquifer Wells

75

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

00.000 00.000 00.000 0.000 . 00-030
0.000 00.000 00.000 00.000 00.000 00.000 0200 0.000 00.000 00.000 00.3..
00.000 00.000 00.000 0.000 00-00.).
00000 0.000 00.000 0.000 00.000 00.000 00.000 00.000 00.000 0000.0.

00. 000 00.000 00.000 0.000 0. 000 00. 000 00.000 00.000 00.000 00-000

00. 000 00.000 00.000 00.000 00. 000 0. 000 00.000 00.000 00.000 00-000

00.000 00.000 00.000 00.000 00. 000 00.000 00.000 00.000 0 0 .000 00-002

00.000 00.000 00.000 00.000 00.000 00.000 00.000 00.000 00.000 00.00

00.000 00.000 00.000 00.000 00.000 00.000 00.000 0.000 00.000 00-000

00.000 00.000 00.000 00.000 00.000 00.000 00.000 0.000 0.000 00.03

00.000 0.000 0.000 00.000 0:00 00.000 00.000 00.000 00.000 00-50

00.000 00.000 00.000 00.000 00.000 0.000 00.000 0.000 00.000 00-000

00.000 00.000 0.000 0.000 00.000 00.000 00.000 00.000 0.000 00.00.).

00.000 00.000 00.000 00.000 2.000 00.000 00.000 00.000 00.000 00-.0<

0.000 00.000 00.000 00.000 0.000 00.000 00.000 00.000 00.000 00.00.).

00. 000 0 0 .000 00.000 00.000 00. 000 00. 000 00.000 00.000 00.000 00-000

00.000 00.000 0.000 00.000 00.000 00.000 00.000 00.000 00.000 00-000

00.000 00.000 00.000 00.000 00.000 00.000 00.000 00.000 000 00.000

0200 00.000 00.000 00.000 00.000 :00 00.000 00.000 00.000 0.0-ma

00. 000 00.000 00.000 00.000 00. 000 00. 000 00.000 00.000 00.000 00.00

00.000 00.000 0.000 0.000 00.000 5.000 00.000 5.000 00.000 00-000

00.000 00.000 00.000 00.000 00.000 00.000 00.000 00.000 00.000 00.00.).

00.000 00.000 00.000 00.000 00.000 00.000 00.000 00.000 00.000 00.000.

0 0.000 0 0 .000 00.000 00.000 3.000 00.000 00.000 00.000 0.000 00-000

00.000 00.000 00-000

00.000 00.000 00-002

00.000 00.000 00-000

0. 000 000 00-000.

00.000 00.000 00-30

00.000 00.000 00-005.

0.000 0.000 00.000

00.000 00.000 00-000

00.000 00.000 00-000

07:0 07:0 07:0 07:m 07:0 0-:0 0-:0 0-:0 0-:0 0-:0 0.:0 0.:0 0-:0 0.0-<0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0__0>> 000.090 2.00200 5 0.05.. 00.02, H: 0.000.

 

76

8007003 006.09 0:800. 800.2200 000 ._z< so: 0000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mvdmm 20.0mm wmdmm 90.0mm 5.0mm mm.mmm Nmmmm @9QO mmdmm 5.0mm mNmm mmNmm mmNmm mm->oz
vwdmm demm fiomm Nowm omdmm :Nmm vmdmm mommyb
hwdmm demm hmdmm 20.0mm mndmm mmdmm mmdmm mNmmm mmdmm mmdmm mnNmm N©.Nmm 5.QO mm-m:<
mm. Pmm mm. wmm NN. ..wm mm. 5m 5m. 5m 3..QO mvémm mo.vmm on. mm m. 0mm m.mmm mnmmm 70.7mm 09:3...
30.7mm Nvémm mm. 0mm mmémm mm.vmm mhémm mmfimu

mmdmm mm.~mm mNNmm mmNmm wm.Nmm mm.mmm vamm mmémfi
Némm 9.5m moémm Nwémm hmémm NmNmm mmemm hm->oZ
3. 5m mmdmm mwdmm umdmm VN. 5m vamm m 0. 0mm >980
um. me mm. 5m mu. me mm. Pmm :Nmm mmmmm mm.mmm unmmm mm. me monm vmm Nmmwm mmémm nmdow
nmdmm mwdmm wndmm demm m 0. 5m mommm mmdmm .542.
NV. Pmm mm. ..mm mm. 0mm 5m. 5m Nm. Em mm.mmm vmm mmdmm av. me mm. me mmdmm mm.mmm mnmmm No-05...
mm. 5m mm. 5m mm. rmm vmdmm NN.mmm m ...Nwm hm..a<
mm. 5m E. «mm mm. 0mm mm. ..mm mm. 5m vm.mmm Nodmm madam 0.. 5m mm. 5m m.mmm @QO 2.me Nana—2
2.5m 2.0mm voémm mwémm v._.mm mm.Nmm mémm hobo“.
mm. SE 0.0. 0mm mwNwm 0.0.QO mmdmm mm. 5m vm. me Nmmmm vmmm vodwm mmémo

mmdmm N0. me mvdwm N.mmm tmmm 9. 0mm mm. Pmm mmmmm vmmmm 90.0mm mméoO

NTID mwurm mtio v v-10 9710 m..:0 $.10 #10 $.10 m-ID v.10 ”ID «-10 m._.<D

 

 

 

 

 

 

 

 

 

 

 

.5280 : 2000

 

LIST OF REFERENCES

77

LIST OF REFERENCES

Bibby, Robert. "Mass Transport of Solutes in Dual-Porosity Media." W
ResearchVol. 17, No. 4 (1981): 1075-1081.

Dominico., P.A., and F.W. Schwartz. PW. New York:
John Wiley & Sons, 1990.

Fontes, J .C., "Environmental Isotopes in Groundwater Hydrology", Hamil"
BMW, Vol. 1, New York: Elsevier, 1980.

Golchert, N.W., and J. Sedlet. "Radiological survey of Site A and Plot M." W
W, DOE/EV-OOOS/7 (1978).

Golchert, N.W., Sedlet, J., and Hayes, K.A.. "Environmental Surveillance of the Palos
Park Forest Preserve." W, ANL-83-6 (1983).

Golchert, NW. and Sedlet, J.. "Site Surveillance and Maintenance Program for Palos
Park - Report for 1982," Argonne National Laboratory, (April 1984).

Golchert, NW. and Sedlet, J.. "Site Surveillance and Maintenance Program for Palos
Park - Report for 1983," Argonne National Laboratory, (June 1984).

Golchert, NW. and Sedlet, J.. "Site Surveillance and Maintenance Program for Palos
Park - Report for 1984," Argonne National Laboratory, (1985).

Golchert, NW. and Sedlet, J.. "Site Surveillance and Maintenance Program for Palos

Park - Report for 1985." WWW, ANL-86-25 (1986).

Golchert, N.W.. "Site Surveillance and Maintenance Program for Palos Park - Report for
1986." W, ANL-87-8 (1987)-

Golchert, N.W.. "Site Surveillance and Maintenance Program for Palos Park - Report for

1987." W, ANL-33-12 (1983)-

Golchert, N.W.. "Site Surveillance and Maintenance Program for Palos Park - Report for

1987." W, ANL-88-12 (1988)-

78

Golchert, N.W.. "Site Surveillance and Maintenance Program for Palos Park - Report for

1988." W, ANL-39/7 (1939)-

Golchert, N.W.. "Site Surveillance and Maintenance Program for Palos Park - Report for
1992." WWW, ANL-93/4 (1993)-

Huyakom, P.S., Lester, B.H., and CR. Faust. "Finite Element Techniques for Modeling
Groundwater Flow in Fractured Aquifers." WW Vol. 19, No.
4 (1983): 1019-1035.

Jones, D.M.A.. "Variability of Evapotranspiration in Illinois." WW
Circular Vol. 89, (1966).

Keys, Scott. Written Communication. 1989.

Kinzelbach, Wolfgang. mm ‘
inBASIQ. Amsterdam: Elsevier, 1986.

 

McDonald, M. G. and A. W. Harbaugh. "A Modular Three-Dimensional Finite Difference

Ground-Water Flow Model " W
W Book 6, Modeling Techniques, Chapter A1 (1988).

Moffet, D.L., Nicholas, J.R., and Shapiro A.M., "Theory and Application of Hydraulic
Testing in a Fractured Dolomite Near Chicago, Illinois", paper from the Proceedings
of the Mid-West Groundwater Conference, 1986.

Nicholas, James R. Personal Communication. 1992 - 1996.

Nicholas, J .R.. "Overview of a Hydrogeological Study of Fractured Dolomite in
Northeastem Illinois", paper presented at the International Conference on Fluid Flow
in Fractured Rock, Georgia State University, Atlanta, May 16-18, 1988.

Nicholas, J .R., and R.W. Healy. "Tritium Migration From a Low-Level Radioactive-

Waste Disposal Site Near Chicago, Illinois." W
Page; 2333 (1988).

Nicholas, J. R, and Shapiro, A. M., "Hydraulic Characteristics of a Jointed Dolomite
Beneath a Low-Level Radioactive-Waste Disposal Site. " ,EOS Transactions of AGU,
Vol. 67, (1986).

Nicholas, J.R., Shapiro, A.M., and de Vries M.P.. "Hydraulic and Tracer Tests in a
Fractured Dolomite, Northeast Illinois", unpublished.

79

Olympio, J. C, "Low- Level Radioactive Waste Burial at the Palos Forest Preserve, Illinois,
Part 1. Preliminary F mite-Difference Models of Steady State Ground- Water Flow"

W No 80-- 775 (1980)

Olympic, J .C., "Low-level radioactive-waste burial at the Palos Forest Preserve, Illinois;

Geology and hydrology of the glacial dn'fi" mmmmammmmmm
Bauer, No. 2226 (1984).

Reilly, T.E., Plummet, L.N., Phillips, P.J., and E. Busenberg. "The use of simulation and
multiple environmental tracers to quantify groundwater flow in a shallow aquifer."

Womb, Vol. 30, NO- 2, pp- 421-433 (1994)-

Shapiro. A.M.. "Interpretation of Oscillatory Water Levels in Observation Wells During
Aquifer Tests in Fractured Rock." W, Vol. 25, No. 10, (1989).

Shapiro, A.M., and Nicholas, J. R.. "Assessing the Validity of the Channel Model for
Fracture Aperture Under Field Conditions." WW, Vol. 25, No.
5, (1989).

Zheng, Chunmiao. "MT3D A Modular Three-Dimensional Transport Model for Simulation
of Advection, Dispersion and Chemical Reactions of Contaminants in Groundwater
Systems," S.S. Papadopulos & Associates, Inc. (1992).

 

"lillllilllllllllllllll“