REMEDIAL ACTION MODELING ASSESSMENT
WESTERN PROCESSING SITE, KENT, WASHINGTON
Draft
by
F. W. Bond
C. M. Smith
Battelle Project Management Division
Office of Hazardous Waste Management
Richland, Washington 99352
January 1985
Prepared for
U.S. Environmental Protection Agency
Region X
Seattle, Washington 98101
-------
prc
Planning Research Corporation
REMEDIAL ACTION MODELING
ASSESSMENT
WESTERN PROCESSING
DRAFT FINAL
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Waste Programs Enforcement
Washington, D.C. 20460
Work Assignment No
EPA Region
Site No.
Date Prepared
Contract No.
PRC No.
Prepared By
Telephone No.
36
X
10T16
February 4. 1985
68-01-7037
15-0360-49
Battelle
(Chris M. Smith)
(509) 943-6001
-------
DISCLAIMER
The information in this document has been funded by the U. S.
Environmental Protection Agency under Contract No. 68-01-6692, Task Order No.
107 to CHZM'HILL and Contract No. 68-01-7037 to PRC Engineering, with
Battelle Project Management Division, Office of Hazardous Waste Management as
the principal subcontractor. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
11
-------
ABSTRACT
The Western Processing Hazardous Waste Site consists of 13 acres near
Kent, Washington, which operated as an industrial waste recycling facility
from about 1960 through 1982. During 1982, the U.S. Environmental Protection
Agency (EPA) conducted surface water sampling around the site and found 26
priority pollutants, all of which were subsequently found on site. As a
result of these findings and subsequent efforts, the EPA initiated a series
of studies to characterize the site and evaluate remedial action
alternatives.
One of the efforts initiated by the EPA was to develop a groundwater
flow and contaminant transport model of the site to be used in evaluating
proposed"remedial actions. The development and calibration of this model and
its use in evaluating remedial actions is discussed in this report.
A conceptual model of the study area was formulated based on the
available hydrogeologic and contaminant data. The conceptual model formed
the framework for developing the groundwater flow and contaminant transport
model of the area around the Western Processing Site.
Once calibrated a limited sensitivity analysis was performed to
determine the sensitivity of the model results to changes in hydraulic
conductivity, porosity, recharge, dispersivity, and retardation factor. The
model was used to evaluate the effectiveness of remedial actions proposed by
CH2M HILL and the Potentially Responsible Parties, as well as minor
modifications. The remedial actions considered include: 1) no-action; 2)
source removal; 3) source removal and pump and treat; 4) capping and pump and
treat; 5) source removal and a slurry wall; and 6) source removal, slurry
wall, and pump and treat. Pumping rates of both 100 gpm and 200 gpm were
simulated in the model.
Of the remedial actions simulated, source removal combined with pump and
treat was found to provide the greatest reduction in mass of
iii
-------
trichloroethylene (TCE) from the groundwater flow system. Simulations showed
that all TCE was removed from the groundwater flow system in 15 and 5 years
at the 100 gpm and 200 gpm pumping rates, respectively. The slurry wall
around the site prevented the removal of TCE that was off site, but was
effective in reducing the lateral flow of clean water to wells during pumping
and containing contamination after pumping ceases.
IV
-------
CONTENTS
Disclaimer .]]
Abstract in.
Figures V1
Tables ..?
Acknowledgement xni
1. Introduction 1
2. Conclusions 3
3. Description of the Study Area '
Climate 7
Geology 9
Hydrology 10
Waste Disposal History 10
4. Model Development 17
Model Selection 17
Groundwater Flow Model Development 18
Structure 18
Boundary Conditions 17
Hydraulic Conductivity 21
Groundwater Potential 24
Hydraulic Stress 24
Porosity 24
Contaminant Transport Model Development 24
Contaminant Selection 25
Source Location 25
Source Area Concentration 26
Source Duration/Leach Rate 26
Sorpti on/Retardation 27
5. Model Calibration 28
Flow Model Calibration 28
Transport Model Calibration 32
Base Case Model Results 36
Numerical Dispersion 35
6. Sensitivity Analysis
Hydraulic Conductivity (Runs 1-4) 44
Porosity (Runs 5 and 6) ! ! ! ! 44-
Recharge (Runs 7 and 8) ! ! ! ! 45
Retardation (Runs 9 and 10) '.!!!! 45
Dispersivity (Runs 11 and 12) !".!!!! 46
Combination Runs (Runs 13-16) '.'.'.... 46
Sensitivity Analysis Summary .'.'..... 47
7. Assessment of Remedial Action Alternatives. ...'...,. 50
-------
CH2M HILL Proposed Remedial Actions 50
No Action (Extension of Base Case) 51
Source Removal 51
Source Removal combined with Pump
and Treat (100 gpm) 55
Source Removal Combined with Pump
and Treat (200 gpm) 58
Cap Combined with Pump and Treat (100 gpm). ... 58
Cap Combined with Pump and Treat (200 gpm). ... 60
PRPs Proposed Remedial Action 63
No-Action (Extension of the Base Case) 63
Source Removal and Slurry Wall 65
Source Removal, Slurry Wall, Pump and
Treat (100 gpm) 67
Source Removal, Slurry Wall, Pump and
Treat (200 gpm) 69
Source Removal, Pump and Treat (100 gpm),
No Slurry Wall 69
Source Removal, Pump and Treat (200 gpm),
No Slurry Wall 73
8 Summary 75
Base Case 75
Remedial Actions 76
References 79
Appendices
A. Calculation of Leach Duration and Amount from
Source Area 81
B. Recharge Calculations 89
C. Stream and Leakance Boundary Conditions 91
D. Calculation of Recharge Factor 96
-------
FIGURES
Number
la Comparison of the Total Mass of TCE Remaining
in the Groundwater Flow System for the CH2M HILL
Page
Ib
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Kemecnai Action oases
Comparison of the Total Mass of TCE Remaining
in the Groundwater Flow System for the PRPs
Remedial Action Cases ...
Western Processing Site Before 1984 Excavation
Smoothed Kriged Potential Surface for April, 1984
Western Processing Well and Reaction Pond Locations ....
Western Processing Model Area
Finite Element Grid of the Western Processing Site
Location of Surveyed Values for Mill Creek and
Drainage Ditch
Model -Predicted 1983 Top of Layer 1 Potential Surface
for the Base Case Simulation
Model -Predicted 1983 Top of Layer 2 Potential Surface
for the Base Case Simulation
Model -Predicted 1983 Top of Layer 3 Potential Surface
for the Base Case Simulation
Model -Predicted 1983 Top of Layer 4 Potential Surface
for the Base Case Simulation
Model-Predicted 1983 Bottom of Layer 4 Potential Surface
for the Base Case Simulation
Smoothed Kriged TCE Concentration Contours Fall, 1982
Model -Predicted 1983 Top of Layer 1 TCE Concentration
Contours for the Base Case Simulation ... ...
Model -Predicted 1983 Top of Layer 2 TCE Concentration
Contours for the Base Case Simulation
V
5
8
11
14
19
20
22
29
29
30
30
31
33
33
15
vn
-------
16 Model-Predicted 1983 Top of Layer 3 TCE Concentration
Contours for the Base Case Simulation 34
17 Model-Predicted 1983 Top of Layer 4 TCE Concentration
Contours for the Base Case Simulation 35
18 Model-Predicted 1983 Bottom of Layer 4 TCE Concentration
Contours for the Base Case Simulation 35
19 Model-Predicted Top of Layer 1 Potential Surface for the
Sensitivity Run with One-Tenth the Hydraulic
Conductivity 48
20 Model-Predicted Top of Layer 1 Potential Surface for the
Sensitivity Run with One-Half the Hydraulic Conductivity
and the Sensitivity Run with Double the Recharge 48
21 Model-Predicted 1993 Top of Layer 1 TCE Concentration
Contours for the No-Action Simulation 52
22 Model-Predicted 2003 Top of Layer 1 TCE Concentration
Contours for the No-Action Simulation 52
23 "Model-Predicted 2033 Top of Layer 1 TCE Concentration
Contours for the No-Action Simulation 53
24 Model-Predicted Top of Layer 1 Potential Surface for
CH2M HILL Source Removal and Pump and Treat (100 gpm)
Remedial Action 57
25 Model-Predicted Top of Layer 1 Potential Surface for
CH2M HILL Source Removal and Pump and Treat (200 gpm)
Remedial Action 59
26 Model-Predicted Top of Layer 1 Potential Surface for
CH2M HILL Cap and Pump and Treat (100 gpm)
Remedial Action 61
27 Model-Predicted Top of Layer 1 Potential Surface for
CH2M HILL Cap and Pump and Treat (200 gpm)
Remedial Action 61
28 Model-Predicted Top of Layer 1 Potential Surface for the
PRPs Slurry Wall and Pump and Treat (100 gpm)
Remedial Action 68
29 Model-Predicted Top of Layer 1 Potential Surface for the
PRPs Slurry Wall and Pump and Treat (200 gpm)
Remedial Action 70
30 Model-Predicted Top of Layer 1 Potential Surface for the
PRPs Pump and Treat (100 gpm) Remedial Action with no
Slurry Uall 72
vm
-------
31 Model-Predicted Top of Layer 1 Potential Surface for the
PRPs Pump and Treat (200 gpm) Remedial Action with no
Slurry Wall 74
B-l Locations of Regional Values Used to Determine
Boundary Conditions . 95
-------
TABLES
Number
1 Summary of the Model Results for the CH2M HILL
and PRPs Remedial Action Simulations
2 Average Monthly Precipitation, Potential
Evapotranspi ration (PET), and Actual
Evapotranspi ration (AET) for the Seattle Area ....... 9
3 Comparison of Shallow and Deep Well Potentials ....... 12
4 Summary of Soil and Groundwater Concentrations of
TCE and DCE (EPA, 1982) at the Three Suspected
Source Areas ........................ 15
5 "Parameters Used in the Final Calibrated Groundwater
Flow and Contaminant Transport Model ............ 23
6 Comparison of Observed to Model-Predicted Maximum TCE
Concentrations in the Groundwater at the
Three Source Locations ................... 36
7 Distribution of TCE in the Model Base Case Simulation ... 37
8 Summary of Sensitivity Runs ................ 39
9 Summary of Mass of TCE (Ib) Exiting the System for
the Sensitivity Runs .................... 40
10 Summary of Total Mass of TCE (Ib) in the Groundwater
System for the Sensitivity Runs .............. 41
11 TCE Concentration (ppb) at the Sources in 1983 for
the Sensitivity Runs .................... 42
12 Percent Change of Mass and Concentration of TCE for
the Sensitivity Runs Based on 1983 Results ......... 43
13 Model -Predicted Distribution of TCE for the
CH2M HILL No-Action Simulation ............... 54
-------
14 Model-Predicted Distribution of TCE for the CH2M HILL
Source Removal Action 56
15 Model-Predicted Distribution of TCE for the CH2M HILL
Source Removal Combined with Pump and Treat (100 gpm)
Remedial Action 58
16 Model-Predicted Distribution of TCE for the CH2M HILL
Source Removal Combined with Pump and Treat (200 gpm)
Remedial Action 60
17 Model-Predicted Distribution of TCE for the CH2M HILL
Cap Combined with Pump and Treat (100 gpm)
Remedial Action 62
18 Model-Predicted Distribution of TCE for the CH2M HILL
Cap Combined with Pump and Treat (200 gpm)
Remedial Action 62
19 Model-Predicted Distribution of TCE for the PRPs
No-Action Simulation 64
20 Model-Predicted Distribution of TCE for the PRPs
Source Removal with a Slurry Wall Remedial Action 66
21 Model-Predicted Distribution of TCE for the PRPs Source
Removal, Slurry Wall, and Pump and Treat (100 gpm)
Remedial Action 69
22 Model-Predicted Distribution of TCE for the PRPs Source
Removal, Slurry Wall, and Pump and Treat (200 gpm)
Remedial Action 71
23 Model-Predicted Distribution of TCE for the PRPs Source
Removal and Pump and Treat (100 gpm) Remedial Action
(No Slurry Wall) 71
24 Model-Predicted Distribution of TCE for the PRPs Source
Removal and Pump and Treat (200 gpm) Remedial Action
(No Slurry Wall) 73
A-l TCE Soil Concentrations Near Reaction Pond 1 82
A-2 TCE Soil Concentrations Near Reaction Pond III 82
A-3 TCE Soil Concentrations Around Well 21 83
A-4 Summary of Data Used to Make Total Mass Calculations at
all Three Source Locations 35
A-5 Summary of Estimated Total Mass of TCE Present at the
Three Suspected Source Areas 85
XI
-------
A-6 Time and Amount of TCE Leaching from the Unsaturated
Zone at Reaction Pond I 87
B-l Runoff Program Results,
90
C-l Stream Boundary Option Data Used to Simulate
Flux to Mill Creek 92
C-2 Stream Boundary Option Data Used to Simulate
Flux to the Ditch 93
-------
ACKNOWLEDGMENT
The authors wish to thank the EPA Region X staff, particularly Project
Manager Fred Wolf for supporting this project. The efforts of EPA Region X,
CH2M HILL (Bellevue Office, particularly Neil Geitner and Jeff Randall),
Hart-Crowser and Associates, and Gaynor Dawson, Joe English, and Jim Doesburg
of Battelle in helping to formulate the initial conceptual model and
providing technical assistance and review throughout the project are greatly
appreciated. Finally, the authors would like to thank Peggy Monter, Naney
Painter, Beth Eddy, and Dick Parkhurst of the Battelle Office of Hazardous
Waste Management who were instrumental in preparing this report. Our
successful completion of this project to a great extent has been possible
because of the efforts of the people acknowledged above.
XII 1
-------
SECTION 1
INTRODUCTION
The goal of this project was to evaluate remedial action alternatives
for the Western Processing Hazardous Waste Site in Kent, Washington, using a
calibrated groundwater flow and contaminant transport model. The specific
tasks of the study included:
review available data and identify deficiencies;
develop a groundwater flow and contaminant transport model of the study
area;
calibrate the flow and transport model with existing data;
perform a limited sensitivity analysis with .the final calibrated model;
and
evaluate remedial action alternatives for the site with the calibrated
model.
A conceptual model of the flow system was developed based on the
available hydrogeologic data. This conceptual model formed the framework for
developing the flow and transport numerical model of the site.
The Finite Element Three-Dimensional Groundwater (FE3DGW) code (Gupta et
al., 1979) was initially used to model the groundwater flow within an area at
and around the Western Processing Site. A finite element grid was developed
and the necessary data on geologic structure, boundary conditions, hydraulic
conductivities, and hydraulic stress were input in the code. Once the flow
model was calibrated, these data were input to the three-dimensional Coupled
Fluid, Energy, and Solute Transport (CFEST) code (Gupta et al., 1982) to
simulate the groundwater flow and contaminant transport. CFEST is an
extension of FE3DGW in that it uses the same hydrologic data structure and
finite element grid. In addition, CFEST includes the necessary parameters to
couple contaminant transport with groundwater flow.
-------
Once developed, the flow and transport portions of the model were
calibrated to observed 1982 through 1984 potentiometric and contamination
data provided by EPA Region X. In this phase of the project, the transport
modeling was based strictly on observed concentrations of trichloroethylene
(TCE) in the groundwater, surface water, and soil at the site. A limited
sensitivity analysis was performed with the final calibrated (base case)
model in order to test the sensitivity of the model results to variations in
the basic model input parameters.
The final calibrated flow and transport model was used to predict the
effectiveness of remedial action alternatives proposed for the Western
Processing Site by CH2M HILL and the Potentially Responsible Parties (PRPs).
The CH2M HILL proposed actions were: 1) source removal combined with pump
and treat; and 2) cap combined with pump and treat. The PRPs proposed action
(Landau Associates and Dames and Moore, 1984) consisted of a combination of
source removal, a slurry wall around the site, and pump and treat. In
addition to the basic remedial action runs, a no-action run (extending the
base case"'run into the future) was performed to provide a benchmark against
which remedial action results could be compared. Also, a few minor
variations (e.g., using different combinations of actions and variable
pumping rates) to the basic remedial action runs were simulated in order to
better understand the model results.
In all cases, the flow portion of the model was used to predict changes
in the groundwater potential (i.e., drawdown) and volumes of water removed.
The contaminant transport portion of the model was used to predict the mass
of TCE removed from the system and average concentrations up to 50 years into
the future.
-------
SECTION 2
CONCLUSIONS
A groundwater flow and contaminant transport model of the Western
Processing Site has been developed and calibrated. An acceptable calibration
was achieved both in terms of matching model-predicted to observed hydraulic
potentials and TCE concentrations, as well as accurately predicting the
groundwater gain of Mill Creek over the model region, and the concentration
of TCE in the creek. The model was also able to closely match the estimated
total mass of TCE in the groundwater flow system based on monitoring data for
the period 1982 through 1984.
A 4:imited sensitivity analysis was performed to determine the
sensitivity of "the model results to changes in hydraulic conductivity,
porosity, recharge, dispersivity, and retardation factor. The analysis
showed that all parameters tested have some impact on the results of the
final calibrated model, and no parameter can be changed without altering the
current calibration.
The final calibrated model was used to evaluate the effectiveness of
remedial actions proposed by CH2M HILL and the PRPs. All model runs have
simulated the transport of TCE. Other contaminants on site will behave
differently because they have different sorption properties, are present in
different quantities at the site, and/or have different source locations).
It is important to keep this in mind when extrapolating the model results for
TCE to a more comprehensive remedial action for all contaminants on site.
A summary of the results for all remedial action simulation runs is
shown in Table 1. A comparison of the total mass of TCE remaining in the
groundwater flc1" system for the CH2M HILL and PRPs remedial action cases are
shown in Figure la and Ib, respectively. The CH2M HILL PRPs base case and
no-action cases show essentially the same results. The only difference is
that the slurry wall elements in the PRPs case reduced the size of the
-------
TABLE 1. SUMMARY OF THE MODEL RESULTS FOR THE CH2M HILL AND PRPS
REMEDIAL ACTION SIMULATIONS.
to Hill Creek
TCE In System TCE In System 1983 - 1988
TCE Discharging TCE Discharging.'
X TCE
* TCE
Years to Average
to Hill Creek ' Discharging to Withdrawn Years to Remove Drawdown
1908 - 1993 Hill Creek by Pumping Remove 90% of Over Site
Cases in 1988 ( Ib) In 1993 ( Ib)
CH2H HILL Cases
No-Action 13.075 9.611
Source Removal 12,141 8,587
Source Removal,
Pump and Treat
(100 gpm) 1,511 593
Source Removal.
Pump and Treat
(200 gpm) 0 0
Cap, Pump and Treat
(100 gpm) 1,307 51M
Cap, Pump and Treat
(200 gpm) 0 0
PRPs Cases
No-Action 12.389 9.136
Source Removal 12.388 9,875
Source Removal,
Slurry Wall.
Pump and Treat
(100 gpm) 2.577 1.069
Source Removal,
Slurry Mall,
Pump and Treat
(?!)0 gpm) 0 0
Source Removal,
Pump and Treat
(100 gpm) 1.276 845
Source Removal,
Pi/mp dnd Treat
l?(X> gpm) 0 0
(Ib) ,
4,822
4,781
1.824
529
1.767
500
4.480
3.536
2,349
467
1.575
467
(Ib)
3.597
3.447
246
0
213
0
3.392
2.438
1.463
0
418
0
and Ditch
100
100
13
3
9
3
100
100
18
3
II
J
Wells
N/A
N/A
87
97
91
97
N/A
N/A
82
97
89
97
all TCE
50
50
15
5
15
5
-50
>50
15
5
20
5
the TCE
30
30
5
5
5
5
30
40
10
5
5
5
(ft)
N/A
N/A
4.0
8.5
4.5
9.0
N/A
N/A
6.0
12.0
4.0
8.5
N/A * Not Appl(cable
-------
E *>
QJ VI
at >>
13
*
) 01
C
20,000 ]
18,000
16,000
14,000
12,000
10,000
8,000
6,000
4,000
2,000
0
No Action
Source Removal and Pump Treat (100 gpm)
'Cap and Pump and Treat (100 gpm)
Source Removal and Pump and Treat (200 gpm)
'Cap and Pump and Treat (200 gpm)
1963
1S72
1983
1993
2003
2013
2023
2033
Year
Figure la. Comparison of the Total Mass of TCE Remaining in the Groundwater
Flow System for the CH2M Hill Remedial Action Cases
3
-0
1963
Source Removal, Pump and Treat(200 gpm),
with and without Slurry Wall
Source Removal and Slurry Wall
No Action
Source Reiroval, Pump and Treat (100 qpm)
Slurry Wall
Source Removal, Pump and Treat (100 gpm),
No Slurry Wall
2033
Figure Ib. Comparison of the Total Mass of TCE Remaining in the Groundwate-
Flow System for the PRPs Remedial Action Cases
-------
Reaction Pond III and Well 21 source area which in turn, slightly reduced the
mass loading to the system.
The results for the pump and treat where the wells were pumping 200 gpm
(758 1/min) showed that all of the TCE would be removed from the system in
the first 5 years. All cases where the wells were pumping 100 gpm (379
1/min) showed that 85% to 90% of the TCE is removed in the first 5 years and
the remainder is removed within the next 10 to 15 years. The slurry wall in
the PRPs cases prevented the pumping wells from removing TCE that had moved
off site, and therefore, the simulations with the wall were less effective at
removing the TCE.
There do exist some small differences in the results for the various
model simulations. But, for the most part, all remedial actions were about
equally effective in cleaning up the TCE in a 5 to 15 year time period.
-------
SECTION 3
DESCRIPTION OF THE STUDY AREA
The Western Processing Site is located within the City of Kent,
approximately four miles (6 km) north of the business district (Figure 2).
The facility occupies an area of about 13 acres (5 ha), and when in operation
consisted of a small laboratory, a solvent recycling plant, a fertilizer
plant, bulk storage tanks, drum storage areas, piles of flue dust,
construction debris, and large cement-block above-ground storage lagoons for
liquid wastes, cooling water, and process water (EPA, 1983). Mill Creek,
also known as King County Drainage Ditch No. 1, runs across the northwest
corner of-rthe site from south to north. A drainage ditch, bicycle trail, and
railroad tracks run along the eastern boundary of the site.
During the fall and winter of 1984, the site was partially excavated to
remove surface structures, waste piles, and some of the contaminated soil.
The modeling discussed in this document uses the pre-excavation ground-
surface elevation (averaging 25 ft above mean sea level) as a reference
point.
CLIMATE
The annual average rainfall at the Western Processing Site is 39 in.
(99 cm). There is a well defined dry season in the summer and a rainy season
in the winter. Table 2 shows the monthly average of precipitation, potential
evapotranspiration, and actual evapotranspiration. The amount of
precipitation that recharges the aquifer was estimated to range from 4 to
12 in./yr (10 to 30 cm/yr). Using a method described by Dunne and Leopold
(1978), a recharge of 8 in./yr (20 cm/yr) was obtained. A detailed
description of recharge calculations is contained in Appendix B.
-------
n MISCELLANEOUS DEBRIS
S&ATTU
BELLfVUE
RCNTON
KENT
PLASTIC COVERED
PILE OF |
CONTAMINATED SOIL
LAGOONS FILLED WITH
CONTAMINATED WATER
FERTILIZER PLANT
EMPTY DRUMS
FULL DRUMS
BATTERY CHIPS
STEEL MILL FLUE DUST
CONTAINING HEAVY METALS
NOT TO SCALE
V///A PONOED WATER
^^ ASPHALT
DRUMS OF ZINC CHLORIDE
Figure 2. Western Processing Site Before 1984 Excavation
-------
TABLE 2. AVERAGE MONTHLY PRECIPITATION, POTENTIAL EVAPOTRANSPIRATION
(PET), AND ACTUAL EVAPOTRANSPIRATION (AET) FOR THE SEATTLE AREA
PET,** in. AET,** in.
0.3 0.3
0.6 0.6
1.2 1.2
1.8 1.8
3.1 3.0
3.8 2.9
4.5 2.0
4.1 1.6
2.8 1.9
1.8 1.8
0.8 0.8
0.5 0.5
Month Precipitation,* in.
January
February
March
April
May
June
July
August
September
October
November
December
Annual
5.73
4.24
3.79
2.40
1.73
1.58
0.81
0.95
2.05
4.02
5.35
6.29
38.94
25.3 18.4
*(NOAA, 1974)
**(EH1s, 1984)
GEOLOGY
The Western Processing Site lies in the broad flood plain of the Green
River. Elevations in this valley average 20 ft (6 m) above mean sea level.
The sediments include alluvial fan deposits of sand, silt, peaty silt, and
clay more than 150 ft (45 m) thick, primarily derived from Mt. Rainer and
transported by the White River (Luzier, 1969).
The top 8 ft (2.4 m) underlying the Western Processing Site consists
primarily of unsaturated artificial fill material. Below the fill, the site
is underlain by a mixture of sand, silt, clay, and peat to a depth of
approximately 40 ft (12 m) (Unit 1). Intermittent clay lenses are present in
this layer (Hart-Crowser, 1984). From 40 to 150 ft (12 to 46 m) belov, the
surface (Unit 2) the material consists predominantly of fine to medium sand
with occasional layers of silty and/or gravelly sand (Hart-Crowser, 1984).
Underlying Unit 2 is a thick (greater than 218 ft) layer of silt which forms
an aquitard (CH2M HILL, 1984).
-------
Potential contours
1n feet above mean
sea level
FEET
> FLOW DIRECTION
Figure 3. Smoothed Kriged Potential Surface for April, 1984
11
-------
site. As a result of this action, disposal at the site ceased in 1982. In
1982 and 1983, the EPA installed a series of monitor wells and collected soil
and water samples. An initial remedial measure was implemented (limited
capping, excavation and removal, drum storage, etc.) to contain some of the
waste until a more permanent remedial action could be designed and
implemented.
In 1984, the EPA did more extensive excavation, removing all of the
structures and materials (drums, waste piles, buildings, etc.) on site and
removing some of the contaminated soil. Waste disposal records for the site
are sketchy; however, a review of the records that are available indicates
that TCE was one of the most common wastes received, and that it was received
from about 1960 to 1980. Sampling results confirm the widespread
distribution of TCE in on-site wells, although it has been detected (at low
levels) in only two of the off-site wells.
There is no evidence that trans-l,2-dichloroethylene (DCE) had ever been
received at the site, however, DCE has been found on site in both soil and
groundwater samples. In some cases the level of DCE approaches or exceeds
the level of TCE.
As a general trend, . it has been found that the occurrence of the
dichlorinated species increases downgradisnt from source areas. These
findings led to the conclusion that degradation of TCE is the likely source
of DCE. The most plausible mechanism for such a transformation is
biodegradation (Wood et al., 1981).
As a result of these determinations, the contaminant transport model
calibration assumed that this TCE to DCE transformation process occurs.
Because this transformation cannot be simulated in the model, the sum of TCE
and DCE concentrations was used to calibrate the model.
The high TCE concentrations measured in the soil and/or groundwater in
Wells 11, 15, 17, 20, 21, and 27 (Figure 4) indicates the existence of three
probable TCE source areas: 1) Reaction Pond I; 2) Reaction Pond III; and 3)
around Well- 21. A summary of the measured TCE and DCE concentrations in the
soil and groundwater at the three suspected source locations is shown in
Table 4. The sum of the TCE and DCE concentrations in the groundwater was
used for model calibration.
13
-------
South 196th ST.
)pC,i
1A.B 2 3
Reaction Pond
II
Western Processing
[Office Building
22A.B
WESTERN
PROCESSING
66 132 I98 264 330
MONITOR WELL
I-III REACTION PONDS !
PROPERTY BOUNDARY AND ' 24
PROPOSED SLURRY WASTE
LOCATION '
2SA, B
oil
WELL 30
Figure 4. Western Processing Well and Reaction Pond Locations
14
-------
TABLE 4. SUMMARY OF SOIL AND GROUNDWATER CONCENTRATIONS OF TCE AND DCE (EPA, 1982)
AT THE THREE SUSPECTED SOURCE AREAS
Associated Concentration TCE (ppb) Concentration DCE (ppb) Total TCE and DCE
Source
Reaction Pond I
Reaction Pond III
Around Well 21
Well Number
11
15
17
20
27
21
Soil
312
580,000
558,000
676
NS
1,520
Groundwater
80,000
210,000
42,000
1,100
140,000
170,000
.; Soil
0
0
0
0
NS
24
Groundwater
0
0
0
0
0
390,000
in Groundwater
80,000
210,000*
42,000
1,000
140,000*
560,000*
tn
NS = No soil samples taken.
Concentration matched in the model calibration process.
Soil samples may have been taken from either the saturated or unsaturated zone and
were not necessarily in contact with the groundwater.
-------
The maximum observed TCE concentration in groundwater is 210,000 ppb,
while the maximum observed soil concentration is 558,000 ppb. The EPA
priority pollutant human health criteria for TCE in water is 27 ppb at 10"'
cancer risk (EPA, 1980).
16
-------
SECTION 4
- MODEL DEVELOPMENT
A numerical model was developed to simulate groundwater flow and
contaminant transport at the Western Processing Site. The model was
developed in two steps:
1) a flow model was developed to describe the groundwater flow in the area
around the Western Processing Site; and
2) the flow model was used to form the basis of a transport model which
simulated the movement of contaminants in the groundwater.
Although the model was developed in two stages, the final result is a single
model which can be used to simulate groundwater flow and contaminant
transport at the site. Because the model was developed in a staged approach,
the flow and transport portions will be discussed separately.
MODEL SELECTION
A three-dimensional model was selected for the Western Processing Site
because it is able to simulate: 1) variations in permeability with depth; 2)
the vertical flow within the study area; 3) localized discharge to Mill Creek
and the drainage ditch; and 4) slurry wall and pumping depths in the proposed
remedial actions.
The numerical codes selected to model the Western Processing Site are
the FE3D6W flow code and -the CFEST transport code. The FE3DGW code simulates
groundwater flow while its companion code, CFEST, simulates contaminant
transport coupled with groundwater flow. The two codes are completely
compatible such that the simulation of transport phenomena using CFEST
proceeds directly from calibration of FE3DGW based on flow properties. Both
codes have been benchmarked against other numerical codes and have been
verified by solution of standard analytical problems (Gupta et al., 1979 and
1982).
17
-------
GROUNDWATER FLOW MODEL DEVELOPMENT
The flow model of the local area around the Western Processing Site was
developed based on the available hydrogeologic data. The model area is
2,790 ft (850 m) wide and 4,020 ft (1,225 m) long. The Western Processing
Site is located near the center of the model region (Figure 5).
A finite element grid was developed for the local model region to
properly represent the area! extent, boundary conditions, and primary
features of the hydrologic system. The grid consists of 311 nodes and 283
elements. The two-dimensional surface representation of the grid is shown in
Figure 6.
Data files were developed for the aquifer thickness and extent,
vertical and horizontal hydraulic conductivity, and hydraulic stress
(recharge and discharge) using data received from EPA Region X. The data
used in the final calibrated flow model are discussed below.
Structure...
The top 100 ft (30 m) below the water table was simulated in the model.
The top 30 ft (9 m) was simulated as a silt and fine sand material with
intermittent clay lenses (Unit 1 as defined in the previous section). The
intermittent clay lenses were simulated by reducing the ratio of vertical to
horizontal permeability. A coarser sand material (Unit 2 -as defined in the
previous section) was simulated between 30 and 100 ft (9 and 30 m) below the
water table.
Although the region was simulated with two geologic units, the vertical
dimension in the model was simulated as four layers. Layers 1 and 2 composed
Unit 1, while Layers 3 and 4 composed Unit 2. The layer thicknesses from top
to bottom were 13 ft (4 m), 17 ft (5 m), 10 ft (3 m), and 60 ft (18 m). This
subdivision of geologic units allowed for more vertical resolution in the
model, as well as allowing an accurate representation of pumping depth,
source removal depth, and slurry wall depth in the remedial action
simulations.
18
-------
WELL LOCHT10N MI MLMBt
Figure 5. Western Processing Model Area
19
-------
Boundary Conditions
The model boundaries were defined with the FE3DGW code's "leakance
boundary condition" option, which can be defined as a combination of flux and
constant head (held) boundary conditions. This option uses the distance from
the boundary to a known potential, the potential at that distance, and the
cross-sectional area associated with each node to calculate the flux and
potential at the boundary nodes based on the model conditions. In effect,
this option calculates the potential along the boundary based on regional and
local data. The regional data used to calculate the boundary conditions were
obtained from regional wells and Green River elevations. Elevations along
the Green River were assumed to be hydraulically connected to the groundwater
system.
Mill Creek and the ditch to the east of the site were simulated using a
stream boundary option. Rather than holding the groundwater potential at the
elevation of the creek and the ditch, this option allows the model to
calculate^ the groundwater potential based on the surface water elevation;
stream bottom elevation, cross-sectional area, thickness, and permeability;
and minimum stream depth. Surface water elevations at each node along Mill
Creek and the ditch, were interpolated or extrapolated from surveyed values
obtained in April, 1984 (Figure 7). A more detailed description of the
boundary conditions is provided in Appendix C.
Hydraulic Conductivity
Initially, horizontal (Kh) and vertical (Kv) hydraulic conductivities
were assigned to the two material types discussed above based on data in the
Hart-Crowser report (1984). These values were then adjusted in the
calibration process until a good match was achieved between model-predicted
and observed potentials.
The hydraulic conductivities used in the final calibrated model for each
material (Units 1 and 2) are shown in Table 5. The values were uniform
throughout each layer. For both material types, the final calibrated values
were within the range of values reported by Hart-Crowser (1984).
21
-------
' Well Location
A Survey Location and Elevation in i
Feet Above Mean Sea Level
II...
Figure 7. Location of Surveyed Values for Mill Creek and Drainage Ditch
22
-------
TABLE 5. PARAMETERS USED IN THE FINAL CALIBRATED GROUNDWATER
FLOW AND CONTAMINANT TRANSPORT MODEL
Model Layer
3~
Structure
Top (ft AMSL) 15 2 -15 -25
Bottom (ft AMSL) 2 -15 -25 -85
Thickness (ft) 13 17 10 60
Horizontal Hydraulic
Conductivity (ft/day) 3.5 3.5 50 50
Vertical Hydraulic
Conductivity (ft/day) 0.175 0.175 5 5
Kv/Kh 0.05 0.05 0.1 0.1
Effective Porosity 0.15 0.15 0.15 0.15
Longitudinal Dispersivity (ft) 25 25 25 25
Transverse Dispersivity (ft) 5555
Retardation Factor 4444
Lithology silty silty sand sand
sand sand
Surface Recharge = 8.0 in./yr
23
-------
Groundwater Potential
Due to a lack of transient data, it was assumed that the groundwater
system around the Western Processing Site is in steady state. A review of
the transient potential data that have been collected showed that, the
potentials do not change enough to significantly alter the flow field or flow
velocities. Therefore, the steady state assumption is considered to be
acceptable for the modeling effort.
A contour surface of the April, 1984 potential data was used to
represent the initial potential conditions (Figure 3). This surface was
prepared by kriging potential data from 36 wells on and around the site, and
measurements along Mill Creek and the ditch. Kriging is a statistical
technique used to estimate a surface from spatially-distributed data. The
model-predicted potential surface was compared to the kriged potential
surface in the model calibration process.
Hydraulic Stress
The only hydraulic stress considered within the model region was
recharge from precipitation. Recharge was assumed constant over the area at
8 in./yr (20 cm/yr). The only exception was in the asphalted (capped) area
on the site (Figure 2) where recharge was set at 0 in./yr. A detailed
description of the recharge calculations is contained in Appendix B.
Porosity
Measured values of porosity are not available for the Western Processing
Site. As a reasonable estimate, an effective porosity of 15% was used in all
layers of the model.
CONTAMINANT TRANSPORT MODEL DEVELOPMENT
The contaminant transport model" was developed using the calibrated flow
model, observed or estimated migration parameters, and estimates of source
loading on the groundwater system as a function of time. Data input files
were developed to define source concentrations, leaching rates, retardation
factors, and dispersivity. In most cases, these data were not specifically
known for the Western Processing Site. As a reasonable estimate, initial
values were selected from the literature and final values were derived in the-
24
-------
model calibration process. The data used in the final calibrated transport
model are discussed below.
Contaminant Selection
A review of the list of wastes received at the Western Processing Site
showed that TCE was accepted throughout the operating life of the site.
Also, high concentrations of TCE have been measured in many of the wells on
site and it is one of the ubiquitous contaminants. Therefore, TCE was
selected for use in calibrating the transport model, and for comparison of
the various remedial action alternatives simulated.
As discussed earlier, the DCE observed on site is believed to be a
degradation product of the TCE. Since the model is not capable of simulating
this degradation process, the initial mass (concentration) of TCE was assumed
to be the total mass (concentration) of TCE plus DCE. The total
concentration of TCE plus DCE at each of the three suspected source areas is
shown in-Table 4. These totals are the concentrations that were matched in
the model calibration process. From this point on, the total of TCE plus DCE
will be referred to simply as TCE.
Source Location
A review of the sampling results for TCE in the on-site wells (EPA,
1983) reveals three probable source areas: 1) Reaction Pond I; 2) Reaction
Pond III; and 3) around Well 21. The area of the finite elements used to
simulate Reaction Ponds I and III was set to represent the actual areas of
these ponds. Reaction Pond I was represented in the model by elements with
? 7
an area of 7,800 ft (750 m ). Reaction Pond III was represented in the
? p
model by elements with an area of 7,050 ft^ (655 m).
For an unknown number of years during the operation of the site, TCE was
mixed with fly ash in the area around Well 21. High concentrations of TCE in
both the soil and groundwater at the location of Well 21 indicate a probable
source. The area of this source is unknown, therefore, it was determined in
the model calibration process by the area which resulted in the desired
concentration and mass loading of TCE. In the final calibrated model the
2
source around Well 21 was represented by an element with an area of 1,600 ft
(149 m2).
25
-------
Source Area Concentrations
Leaching of TCE into the groundwater, rather than direct infiltration,
was the primary mechanism for introducing TCE to the model. Therefore, the
initial TCE concentration was defined at the elements representing the three
source areas, and the loading rate at each site was calculated as the
infiltration rate times the surface area of the source times the initial TCE
concentration at the source. The initial TCE concentrations at the three
sources were defined as follows: Reaction Pond I 6.0 x 105 ppb, Reaction
Pond III 1.9 x 106 ppb, and around Well 21 1.15 x 107 ppb. The
infiltration rate was constant over the entire model region at 8 in./yr (20
cm/yr). Using the areas for each source given above, the loading rates in
the model at Reaction Ponds I and III, and around Well 21 were 195 Ib/yr (88
Kg/yr), 559 Ib/yr (254 Kg/yr), and 769 Ib/yr (349 Kg/yr), respectively, over
the 20-year active disposal life of the site. The total model-predicted mass
of TCE disposal at the site was 1,523 Ib/yr (689 Kg/yr) or 30,460 Ib
(13,781 Kg) over the 20-year disposal period.
Source Duration/Leach Rate
The sources were assumed to be actively leaching TCE into the
groundwater for 20 years, from 1963 through 1983. The rate of leaching was
constant over the-20 years at the rates given above.
The mass of TCE currently in the unsaturated zone at the three source
areas was estimated to determine how long leaching into the saturated flow
system would continue (see Appendix A). This information was used to
determine how long to keep the sources active in the model. The results of
these calculations indicate that virtually all the mass of TCE in the
unsaturated zone beneath Reation Pond III and around Well 21 has already
leached into the saturated zone. Therefore, these two sources were turned
off in the model in 1983. The calculations show that Reaction Pond I would
continue to leach TCE into the groundwater for about 20 years, and that the
rat" decreases exponentially. Based on the calculations, the model input
concentration at Reaction Pond I was set at full strengh (6 x 105 ppb) for
the first five years into the future (1984 - 1988), and then it was reduced
by 75% at each of the next three time steps (1989 - 1993 = 50,000 ppb, 1994 -
26
-------
1998 = 37,500 ppb, and 1999 - 2003 = 9,400 ppb). After the year 2003, all
sources of TCE were turned off in the model.
SorptIon/Retardation
The CFEST model uses a single retardation factor for the entire
groundwater flow system (all layers), which was determined in the model
calibration process. The retardation factor used in the final calibrated
model was 4.0.
The retardation factor can be calculated from the bulk density, actual
and effective porosity, and the distribution coefficient (Kd) for the
material through which flow occurs. A discussion of these parameters and how
they relate to a retardation factor of 4.0 is contained in Appendix D.
27
-------
SECTION 5
MODEL CALIBRATION
The flow and transport models were calibrated by adjusting certain model
input parameters until a good match was achieved between model-predicted and
observed data. A brief description of the calibration process for both the
flow and transport models is provided below.
FLOW MODEL CALIBRATION
Once the data were input into the FE3DGW code, the model was run in the
steady-state mode to predict groundwater potentials. The model was
calibrated by comparing the model-predicted flow field to measured potential
data.
The difference between model-predicted and measured hydraulic potentials
was minimized by adjusting the following flow model parameters: the vertical
and horizontal hydraulic conductivity, the parameters controlling the flow to
Mill Creek and the drainage ditch (stream bottom permeability and thickness),
and the boundary conditions.
The final model-predicted potential surface for the water table (top of
Layer 1) (Figure 8) compares well with the kriged potential data (Figure 3)
and the conceptual model of the flow regime within the study area (i.e.,
localized flow to Mill Creek and the ditch, and regional flow to the
northwest). Potential surfaces for the model-predicted top of Layers 2, 3,
4, and the bottom of Layer 4 are shown in Figures 9, 10, 11, and 12,
respectively. The model-predicted groundwater flux to Mill Creek along the
reach within the study area is 0.45 cfs (1,101 m3/day). This value compares
well with a gain of 0.5 cfs (1,223 nr/day) along Mill Creek within the study
area as measured in May, 1982, by EPA Region X.
Regional groundwater flow as predicted by the model is to the northwest,
while localized flow is to Mill Creek. Discharge to Mill Creek dominates the
28
-------
ro
Figure 8. Model-Predicted 1983 Top of Layer 1
Potential Surface for the Base Case
Simulation
Figure 9. Model-Predicted 1983 Top of Layer 2
Potential Surface for the Base Case
Simulation
-------
to
o
Figure 10. Model-Predicted 1983 Top of Layer 3
Potential Surface for the Base Case
Simulation
Figure 11. Model-Predicted 1983 Top of Layer 4
Potential Surface for the Base Case
Simulation
-------
"71 f
UJ
Figure 12. Model-Predicted 1983 Bottom of Layer 4
Potential Surface for the Base Case
Slmulat Ion.
-------
flow patterns to a depth of about 30 ft (9 m) and its influence can be seen
at 100 ft (30 m). The model predicts that regional groundwater flow becomes
dominant somewhere between 30 and 50 ft (9 and 15 m) below the surface.
TRANSPORT MODEL CALIBRATION
Once the ^data were input into the CFEST code, the model was run in the
transient mode with five-year time steps from 1963 to 1983. The model was
calibrated by comparing model-predicted to measured TCE cpncentrations for
1983, TCE mass loading to Mill Creek, and total mass of TCE in the system in
1983.
The difference between model-predicted and measured data was minimized
by adjusting the retardation factor and source strengths in the model. Leach
time was assumed to be the period of active disposal (20 years). Because the
unsaturated zone is very thin, it was assumed that TCE entered the saturated
zone soon after disposal.
A sm'bothed kriged concentration contour plot of TCE within the model
study area is shown in Figure 13. The model-predicted TCE concentrations for
the top of Layer 1 (Figure 14) compare reasonably well with the kriged
values. The model-predicted TCE concentrations for the top of Layers 2, 3,
4, and the bottom of Layer 4 are shown in Figures 15, 16, 17, and 18,
respectively.
In addition to matching the observed location of the TCE plumes from the
three source areas, it was important to match the maximum observed TCE
concentration at these areas. This match was achieved through source
calibration. The measured and model-predicted concentrations at the three
i
source locations are shown in Table 6.
The concentration of TCE. in Mill Creek was calculated based on model
results and compared to the measured concentration. The model-calculated
concentration of 39 ppb based on a creek flow of 15 cfs (0.4 rn^/sec) is
comparable to the creek TCE concentration of 15 ppb measured in January 1984,
by EPA Region X.
32
-------
Figure 13. Smoothed Kriged TCE Concentration
Contours, Fall , 1982
Numerical
Dispersion
Numerical
Dispersion^1
Numerical
Dispersion
\
Numerical
Dispersion
Figure 14. Model-Predicted 1983 Top of Layer 1
TCE Concentration Contours for the
Base Case Simulation
-------
Figure 15. Model-Predicted 1983 Top of Layer ?
TCE Concentration Contours for the
Base Case Simulation
Figure 16. Model-Predicted 1983 Top of Layer 3
TCE Concentration Contours for the
Base Case Simulation
-------
CO
en
Figure 17. Model-Predicted 1983 Top of Layer 4
TCE Concentration Contours for the
Case Case Simulation.
Figure 18. Model-Predicted 1983 Bottom of Layer 4
TCE Concentration Contours for the
Dase Case Simulation
-------
TABLE 6. COMPARISON OF OBSERVED TO MODEL-PREDICTED MAXIMUM
TCE CONCENTRATIONS IN THE GROUNDWATER AT THE THREE
SOURCE LOCATIONS
Measured Model-Predicted
Concentration Concentration
Source Location (ppb) (ppb)
Reaction Pond I 210,000 212,000
Reaction Pond III 140,000 139,000
Around Well 21 560,000 557,000
The total mass of TCE (actually TCE plus DCE) in the flow system in
1983 as predicted by the model was 17,100 Ib (7,737 Kg). This value compared
well with the 18,000 Ib (8,144 Kg) of TCE plus DCE as estimated independently
by CH2M HILL (1985) and Landau Associates and Dames and Moore (1984) based on
the 1982 through 1984 measured concentration data. A list of the parameters
used in the final calibrated model is shown in Table 5.
BASE CASE MODEL RESULTS
The base case is defined as the 20-year simulation period from 1963
through 1983. Over this 20-year period, the model predicted that a total of
30,400 Ib (13,756 Kg) of TCE entered the groundwater flow system. Of this
total, 17,100 Ib (7,738 Kg) remained in the flow system in 1983. Of the mass
exiting the system 97% (13,300 Ib (6,018 Kg)) discharged to Mill Creek and
the remaining 3% discharged to the drainage ditch along the eastern boundary
of the site. No TCE entered the deeper, regional flow system which flows
northwest toward the Green River. The distribution of TCE in the groundwater
flow system for the base case is shown in Table 7.
NUMERICAL DISPERSION
Figures 14 and 15 have low concentrations of TCE upgradient from the
sources where it is not possible (from the conceptual model) for
contamination to have occurred from Western Processing. Occurrences of this
phenomena are evidenced by the closed loop 100 ppb contours west of Mill
Creek (which is upgradient from the creek in the top three model layers) and
36
-------
south and east of Well 21; these contour lines are identified in Figure 14.
In all cases, these upgradient occurrences of TCE are caused by numerical
dispersion. Numerical dispersion is inherent to the numerical solution of
the convective dispersion equation contained in the model code and,
therefore, cannot be avoided. Numerical dispersion can be reduced by
adjusting the model node 'spacing, however, it can never be completely
eliminated. In all TCE concentration contour plots (i.e., remedial action
simulation plots), upgradient occurrences of TCE have been attributed to
numerical dispersion'and should be disregarded.
TABLE 7. DISTRIBUTION OF TCE IN THE MODEL BASE CASE SIMULATION
Year
1968
1973
1978
1983
Total
TCE
Inflow fib)
7,601
7,601
7,601
7,601
30,404
TCE
Outflow (lb)
1,049
2,906
4,183
5,195
13,333
TCE Remaining
in Groundwater
System (lb)
6,552
11,247
14,655
17,071
37
-------
SECTION 6
SENSITIVITY ANALYSIS
After the model calibration was completed, several additional model runs
were made to test the sensitivity of the model results to changes in various
model input parameters. The sensitivity runs were performed with the base
case model (1963 - 1983) and the base case extended 20 years into the future.
Thus, the total simulation period was 40 years (1963 - 2003), and the
sensitivity runs were performed for the case where no remedial actions were
simulated in the model (base case and no-action case).
The model parameters varied in the sensitivity analysis were hydraulic
conductivity, porosity, recharge, retardation, and dispersivity. A summary
of the sensitivity runs performed is shown in Table 8. For the most part the
sensitivity analysis consisted of doubling and halving each parameter while
holding all other parameters constant. A summary of the values of the base
case model parameters is shown in Table 5.
The results of all the sensitivity runs are shown in Tables 9 through
12. In all sensitivity runs, the mass of TCE entering the groundwater flow
system from the source areas was not changed (the infiltration at the sources
was the same as the base case). Table 9 summarizes the mass of TCE exiting
the system at each time step (5-year intervals) to Mill Creek and the
drainage ditch east of the site. In all cases, virtually all (97%) of the
TCE exiting the system goes to Mill Creek. Table 10 summarizes the total
mass of TCE remaining in the system at each time step. Table 11 reports the
model-predicted TCE concentrations for 1983 at the three source areas, the
model-predicted groundwater flux to Mill Creek over the model region, and the
concentration of TCE in Mill Creek based on the creek loadings predicted by
the model. Table 12 summarizes the percent changes of mass and concentration
of TCE based on 1983 results.
38
-------
TABLE 8. SUMMARY OF SENSITIVITY RUNS
Run
Number Parameter(s) Adjusted
1 For all layers, the K is reduced by 0.5.
Layers 1 and 2: Kh = 1.75 ft/day, Kv = 8.75 x 10-2 ft/day.
Layers 3 and 4: Kh = 25 ft/day, Kv = 2.5 ft/day.
2 For all layers, the K is doubled.
Layers 1 and 2: Kh = 7.0 ft/day, Kv = 0.35 ft/day.
Layers 3 and 4: K = 100.0 ft/day, Kv = 10.0 ft/day.
3 The K for all layers is reduced by one-tenth.
Layers 1 and 2: Kh = 0.35, Kv = 0.0175.
Layers 3 and 4: Kh = 5.0, Kv = 0.5.
4 The K for all layers is increased ten times.
Layers 1 and 2: Kh = 35, Kv = 1.75.
Layers 3 and 4: Kh = 500, Kv = 50.
5 The effective porosity is reduced by 0.5 for all layers; porosity1
7.5%.
6 The effective porosity is doubled for all layers; porosity = 30%.
1 The recharge is reduced to zero.
8 The recharge is increased to 16 in./yr.
9 The retardation is reduced to zero.
10 The retardation is increased to 8.
11 The dispersivity is reduced by a factor of 5: D, = 5 ft, DT =
1 ft. L '
12
The dispersivity is increased by a factor of 5: D. = 125 ft,
n - -3c f+ L
DT = 25 ft.
13 The effective porosity is 30% for all layers and the retardation is 2.
14 The effective porosity of Layers 1 and 2 is 45% and the porosity of
Layers 3 and 4 is 40%.
15 For Layers 3 and 4: Kh = 25 ft/day, Kv = 2.5 ft/day; and porosity
= 25%
16 Effective porosity = 25%, retardation = 8.0.
K = hydraulic conductivity
39
-------
TABLE 9. SUMMARY OF MASS OF TCE (LB) EXITING
THE SYSTEM FOR THE SENSITIVITY RUNS
-F*
o
Time
Ste
1
2
3
4
5
6
7
8
Time
Step
1
2
3
4
5
6
7
8
p Year
1968
1973
1978
1983
1988
1993
1998
2003
Year
1968
Base
Case Run 1 Run 2
1,
2,
4,
5,
4,
3,
2,
2,
Base
Case
1,049
1973 2,906
1978 4,183
1983 5,195
1988 4,971
1993 3,708
1998 2,839
2003 2,000
049
906 1,
183 2,
195 3,
971 3,
708 3,
839 2,
000 2,
Run 9
3,003
6,672
7,380
7,404
4,759
1,621
720
140
604 1,768
823 4,514
770 6,004
506 6,836
616 5,454
045 3,048
689 1,729
419 814
Run 10
542
1,655
2,542
3,238
3,375
2,878
2,560
2,336
Run 3
175
632
1,126
1,513
1,676
1,513
1,284
1,130
Run 11
1,386
3,731
5,003
5,848
5,245
3,501
2,681
2,029
Run ft
4,493
8,160
7,508
7,879
3,139
244
61
15
Run 12
810
2,259
3,455
4,402
4,281
3,407
2,596
1,860
Run 5
1,859
4,706
6,181
6,949
5,418
2,897
1,573
707
Run 13
1,049
2,097
4,183
5,194
4,971
3,709
2,839
1,999
Run 6
542
1,655
2,543
3,238
3,375
2,877
2,560
2,336
Run 14
350
1,144
1,857
2,412
2,602
2,293
2,059
1,946
Run 7 Run 8
944 1,
2,645 3,
3,847 4,
4,824 5,
4,706 5,
3,636 3,
2,910 2,
2,192 1,
Run 15
969
2,677
3,783
4,655
4,453
3,428
2,893
2,405
143
131
466
491
162
723
725
780
Run 16
1,
1,
2,
2,
2,
1,
1,
310
029
691
202
373
100
869
762
-------
TABLE 10. SUMMARY OF TOTAL MASS OF TCE (LB) IN THE GROUNDWATER
SYSTEM FOR THE SENSITIVITY RUNS
Time
Step
1
2
3
4
5
6
7
8
Time
Step
1
2
3
4
5
6
7
8
Year
1968
1973
1978
1983
1988
1993
1998
2003
Year
1968
1973
1978
1983
1988
1993
1998
2003
Base
Case
6,552
11,247
14,665
17,071
13,075
9,611
6,833
4,848
Base
Case
6,552
11,247
14,665
17,071
13,075
9,611
6,833
4,848
Run 1
6,997
12,775
17,606
21,701
19,070
16,269
13,641
11,237
Run 9
4,598
5,527
5,748
5,945
2,161
784
125
0
Run 2
5,833
8,920
10,517
11,282
6,803
3,999
2,331
1,532
Run 10
7,059
13,005
18,064
22,427
19,827
17,193
14,694
12,373
Run 3
7,426
14,395
20,870
26,958
26,057
24,788
23,565
22,450
Run 11
6,215
10,085
12,683
14,436
9,966
6,709
4,089
2,075
Run., 4
3,
2,
2,
2,
Run
6,
12,
16,
19,
15,
12,
10,
8,
108
549
642
364
0
0
0
0
12
791
133
279
478
972
809
274
429
Run 5
5,742
8,637
10,057
10,709
6,266
3,613
2,101
1,409
Run 13
6,552
11,246
14,664
17,071
12,875
9,410
6,632
4,648
Run 6
7,059
13,005
18,063
22,426
20,026
17,393
14,894
12,573
Run 14
7,251
13,708
19,452
24,641
22,814
20,765
18,767
16,836
Run 7
6,657
11,613
15,367
18,144
14,413
11,021
8,172
5,995
Run 15
6,632
11,556
-15,374
18,320
14,842
11,658
8,826
6,436
Run 8
6,458
10,928
14,063
16,173
11,986
8,507
5,843
4,078
Run 16
7,292
13,864
19,774
25,173
23,774
21,918
20,104
18,354
-------
ro
TABLE 11. TCE CONCENTRATION (PPB) AT THE SOURCES IN 1983
FOR THE SENSITIVITY RUNS
Location Base Case Run 1 Run 2 Run 3 Run 4 Run 5
Reaction Pond I
Reaction Pond III
Well 21
Mill Creek
212,000
139,000
557,000
39
315,000
232,000
700,000
27
121,000 ;
72,000
368,000
51
441,000
403,000
822,000
11
21,000
10,000
80,000
60
218,000
132,000
620,000
52
Groundwater Flux Into
Mill Creek (cfs) 0.45 0.23 0.90 0.05 4.42 0.45
Location Base Case Run 6 Run 7 Run 8 Run 9 Run 10
Reaction Pond I
Reaction Pond III
Well 21
Mill Creek
212,000
139,000
557,000
39
184,000
133,000
448,000
24
207,000
161,000
718,000
36
213,000
120,000
416,000
41
207,000
117,000
622,000
56
184,000
133,000
448,000
24
Groundwater Flux into
Mill Creek (cfs) 0.45 0.45 0.44 0.46 0.45 0.45
Location Base Case Run 11 Run 12 Run 13 Run 14 Run 15 Run 16
Reaction Pond I
Reaction Pond III
Well 21
Mill Creek
212,000
139,000
557,000
39
485,000
222,000
911,000
43
53,000
50,000
207,000
33
212,000
139,000
557,000
38
160,000
122,000
373,000
18
212,000
144,000
538,000
34
154,000
119,000
354,000
16
Groundwater Flux into
Mill Creek (cfs) 0.45 0.45 0.45 0.45 0.45 0.42 0.45
-------
TABLE 12. PERCENT CHANGE OF MASS AND CONCENTRATION OF TCE
FOR THE SENSITIVITY RUNS BASED ON 1983 RESULTS
Run Mass
No. Exiting
1
2
3
4
5
6
7
8
9
10
11 ":"
12
13
14
15
16
-33
+32
-71
+52
+34
-38
-7
+6
+43
-38
+13
-15
0
-54
-10
-58
Total Mass
in System
+27
-34
+58
-86
-38
+31
+6
-5
-65
+31
-15
+14
0
+44
+7
+47
Reaction
Pond I
+49
-43
+108
-90
+3
+13
-2
0
-2
-13
+129
-75
0
-25
0
-27
Reaction
Pond III
+67
-48
+190
-93
-5
-4
+16
-13
-16
-4
+60
-64
0
-12
+4
-14
Well
21
+26
-34
+48
-86
+11
-20
+29
-25
+12
-20
+64
-63
0
-33
-3
-36
Mill
Creek
-31
+31
-72
+54
+33
-38
-8
+5
+44
-38
+10
-15
0
-54
-13
-58
Mill
Creek
Flux
-49
+100
-89
+882
0
0
-2
+2
0
0
0
0
0
0
-7
0
+ = increase
- = decrease
43
-------
A discussion of the results shown in the tables and a summary of all the
sensitivity analysis model runs is presented below.
HYDRAULIC CONDUCTIVITY (RUNS 1 - 4)
Four sensitivity runs were made where the base case horizontal and
vertical hydraulic conductivities (K) were changed by factors of 0.1, 0.5,
2.0, and 10.0. Decreasing K decreased the groundwater flux through the
system which:
1) decreased the TCE mass loading to the creek and ditch;
2) increased the total mass of TCE in the system;
3) increased the concentration of TCE at the three source areas; and
4) decreased the groundwater flux to Mill Creek.
Increasing K increased the flux through the system which had the opposite
effect of decreasing K as discussed above.
Decreasing the K by factors of 2.0 and 10.0 increased the water table
elevation'over the model region by about 1.0 and 5.0 ft (0.3 and 1.5 m),
respectively. Increasing the K by a factor of 10.0 decreased the potentials
by about 0.5 ft (0.15 m), whereas increasing K by a factor of 2.0 only
slightly decreased the potentials.
POROSITY (RUNS 5 AND 6)
Two sensitivity runs were made where the base case effective porosity
was changed by factors of 0.5 and 2.0. Decreasing the effective porosity
increased the groundwater velocity and decreased the amount of dilution. The
effect of this change was to:
1) increase the mass loading to the creek and ditch;
2) decrease the total mass in the system;
3) increase the concentration at 2 of the 3 source areas (Reaction Pond III
decreased slightly); and
4) maintain the same groundwater flux to the creek.
Increasing the porosity had the opposite effect of decreasing the porosity.
Changes in porosity had no impact on the groundwater potentials.
44
-------
DISPERSIVITY (RUNS 11 AND 12)
Two sensitivity runs were made where the base case longitudinal and
transverse dispersivity were changed by factors of 0.2 and 5.0. Decreasing
the dispersivity:
1) increased the mass loading to the creek and ditch;
2) decreased the total mass in the system;
3) increased the concentration at the three source areas; and
4) did not impact the groundwater flux to the creek.
Increasing the dispersivity had the opposite effect of decreasing it.
Changes in dispersivity had no impact on the groundwater potentials.
COMBINATION RUNS (RUNS 13 - 16)
Four runs were made in which either two parameters were changed
simultaneously or a single parameter was varied between layers.
In Run 13 the effective porosity was doubled and the retardation factor
was halved. These parameters both impact the groundwater velocity and the
changes offset each other, therefore, the results are virtually identical to
the base-case results.
In Run 14 the effective porosity in the upper and lower units was
increased to 45% and 40%, respectively. The impact of this increase is
similar to the impact of increasing porosity as discussed above.
In Run 15 the hydraulic conductivity in the lower unit (model Layers 3
and 4) was halved and the effective porosity in the lower unit was increased
to 25%. The impact of these changes was to:
1) decrease the mass loading to Mill Creek;
2) increase the total mass in the system; and
3) decrease the groundwater flux to Mill Creek.
The change in concentration and flux was small. The parameter changes made
only in Unit 2 (gravelly sand) changed the model results slightly (Table 12),
indicating that some contaminant is migrating through the lower unit of the
model to Mill Creek.
In Run 16 the effective porosity in all layers was increased to 25% and
the retardation was increased to 8.0. This case was intended to better match
the parameters used by CH2M HILL in their feasibility study. The effect of
these changes was to:
46
-------
1) decrease the mass loading to the creek and ditch;
2) increase the total mass in the system;
3) decrease the concentrations at the three source areas; and
4) maintain the base case flux rate to Mill Creek.
The groundwater potential surfaces were virtually unchanged by the
changes made in the combination runs.
SENSITIVITY ANALYSIS SUMMARY
Only three of the sensitivity runs showed significant changes in the
potential surfaces: those where the K was decreased by factors of 2.0 and
10.0, and where the recharge was doubled. The potential surfaces for the top
of Layer 1 for the case where K was decreased by a factor of 10.0 is shown in
Figure 19. The potential surface for the case where the K decreased a factor
of 2.0 and for the case where recharge was doubled are virtually the same and
are shown in Figure 20.
Chan-gjng the hydraulic conductivities changes the groundwater flux
through the system as well as the velocity, whereas changing the porosity or
retardation factor only impacts the velocity of the contaminant. Changing
recharge has a small impact on groundwater flux, whereas dispersivity has no
impact on flux. Therefore, the groundwater flux to Mill Creek is almost
completely controlled by the hydraulic conductivity. Since the base case
model accurately predicts the measured value of groundwater flux to Mill
Creek over the model region the hydraulic conductivity used in the model is
probably reasonable.
The mass of TCE exiting the system is about equally controlled by the
hydraulic conductivity, the porosity, and the retardation factor. If any one
of these parameters is changed by a factor of two, the results are altered by
a factor of about plus or minus 35%. This degree of sensitivity supports the
values currently used in the model.
Dispersivity has the greatest influence on concentration at the three
source areas. Dispersivity is also the parameter in which we have the least
confidence. The other parameter which influences the source concentrations,
is the hydraulic conductivity. Changes in porosity, retardation, and
recharge have little influence on the concentrations at the sources. Since
47
-------
-
00
WP 8EN HUN 14 TOT CTOP LRYEB 1)
Figure 19. Model-Predicted Top of Layer 1 Potential
Surface for the Sensitivity Run with One-
Tenth the Hydraulic Conductivity
VP 8EHHUH 2PCT
LflYER 1)
Figure 20.
Model-Predicted Top of Layer 1 Potential
Surface for the Sensitivity Run with One-
Half the Hydraulic Conductivity and the
Sensitivity Run with Double the Recharge
-------
dispersivity> is the primary factor controlling source concentration, and the
model-predicted concentrations match the measured values, the base case
dispersivity is probably reasonable.
The limited sensitivity analysis illustrated how every parameter has
some influence on the groundwater potential, the flow to Mill Creek, the TCE
flux to Mill Creek, and/or the total mass of TCE in the system. Changing any
of the parameters will affect the current calibration.
49
-------
SECTION 7
ASSESSMENT OF REMEDIAL ACTION ALTERNATIVES
Remedial action alternatives for the Western Processing Site have been
proposed by CH2M HILL and the PRPs for site restoration. The purpose of the
remedial action assessment performed in this study was to simulate these
proposed alternatives in the final calibrated model in order to evaluate and
compare their effectiveness.
The base case consisted of running the final calibrated model for 20
years, from 1963 through 1983. In order to simulate the remedial action
alternatives, certain model parameters were adjusted in 1983 and the model
was run for an additional 50 years to the year 2033. When the pump and treat
alternative was simulated, water was withdrawn for the first 30 years (1983 -
2013) in the CH2M HILL cases and for the first 5 years (1983 - 1988} in the
PRPs cases. The model results during the prediction period were used to
determine which action would be most effective in reducing the level of
contamination at the site.
The following section describes the remedial action alternatives and how
they were simulated in the model, and provides the results of each
simulation. The alternatives are discussed in two sections, those proposed
by CH2M HILL and those derived from the alternative proposed by the PRPs. A
summary of the results is also provided in Section 2.
CH2M HILL PROPOSED REMEDIAL ACTIONS
CH2M HILL requested that three basic simulations be performed with the
model:
1) no-action;
2) source removal combined with pump and treat; and
3) cap combined with pump and treat.
50
-------
The pump and treat cases were run with two different pumping rates. In order
to better understand the results, a simulation involving source removal only
was also performed. These three cases and variations are- discussed below.
No Action (Extension of the Base Case)
The no-action simulation consisted of running the base case'model (final
calibrated model presented earlier) 50 years into the future. The purpose of
this simulation was to determine the predicted extent of TCE contamination if
no remedial measures are implemented at the site. This simulation served as
a benchmark against which the CH2M HILL proposed remedial actions could be
compared.
The potential surfaces for the no-action case are the same as for the
base case (Figures 8 to 12). TCE concentration contours at the top of Layer
1 in the years 1993, 2003, and 2033 (10, 20, and 50 years into the future)
are shown in Figures 21 through 23, respectively. The total mass of TCE in
the flow system, and the total mass of TCE discharging to Mill Creek and the
drainage ditch over 5 year intervals are shown in Table 13.
Table 13 shows that of the 31,699 Ib (14,378 Kg) that entered the
groundwater flow system between 1963 and 2003, 15% and 4% remain in the
system in the years 2003. and 2033 (20 and 50 years in the future),
respectively. Of the 17,071 Ib (7,743 Kg) remaining in the flow system in
1983, about 88% and 99% discharges to Mill Creek and the ditch by the years
2003 and 2033, respectively. As for the base case, about 97% of the TCE
exiting the system discharges to Mill Creek and the remaining 3% discharges
to the ditch.
Source Removal
The source removal action assumes that 6 ft (2 m) of unsaturated
material is excavated and removed from the site. The effect of this action
would be to remove the source of TCE that leaches from the unsaturated zone
into the saturated flow system.
Source remova" was simulated in the model by eliminating further input
of TCE beyond 1983. Because Reaction Pond I is the only source that
contributes TCE to the system after 1983 (see Appendix A), source removal was
accomplished in the model by eliminating this source after 1983.
51
-------
UP CF RUN 20 CONC (TOP UTTER 1)
Figure 21. Model-Predicted 1993 Top of Layer 1 TCE
Concentration Contours for the No-Action
Slmula11 on
UP CF RUN 20 CONC OOP LAYER 1)
Figure 22. Model-Predicted 2003 Top of Layer 1 TCE
Concentration Contours for the No-Action
S f milla t ion
-------
en
CO
VP CF HUN 20 CONC CTOP LflYCT 1J
Figure 23. Model-Predicted 2033 Top of Layer 1 TCE
Concentration Contours for the No-Action
Simulation
-------
TABLE 13. MODEL-PREDICTED DISTRIBUTION OF TCE FOR
THE CH2M HILL NO-ACTION SIMULATION
Time
Step
1*
2*
3*
4*
5
6
7
8
9
10
11
12
13
14
Total
End
Year
1968
1973
1978
1983
1988
1993
1998
2003
2008
2013
2018
2023
2028
2033
TCE
Inflow (Ib)
7,601
7,601
7,601
7,601
975
244
61
15
0
0
0
0
0
0
31,699
TCE
Outflow (Ib)
1,049
2,906
4,183
5,195
4,971
3,708
2,839
2,000
1,288
824
531
379
290
240
30,403
Total TCE in
System (Ib)
6,552
11,247
14,665
17,071
13,075
9,611
6,833
4,848
3,560
2,736
2,205
1,826
1,536
1,296
*Base case
Inflow is mass entering the groundwater flow system and
outflow is the mass exiting the system.
54
-------
The groundwater potential surfaces for the source removal case are
identical to those for the base case (Figures 8 to 12) because the soil below
the water table was undisturbed. The source removal action removes so little
mass of TCE from the system, that the TCE concentration contour plots are
essentially the same as those for the no-action case (Figure 21 through 23).
The total mass of TCE in the flow system, and the total mass of TCE
discharging to Mill Creek and the ditch over five-year intervals are shown in
Table 14. The source removal action removes about 1,300 lb (588 Kg) of TCE
from the groundwater flow system. This action results in a slight reduction
in the mass of TCE remaining in the system and exiting to the creek and
ditch.
The source removal action is not effective in cleaning up the site
because most of the TCE has entered the saturated flow system by 1983.
Source Removal Combined with Pump and Treat (100 gpm)
This remedial action assumes that the entire site is excavated to a
depth of 6 ft (2 m) (source removal) followed by the withdrawal of
contaminated groundwater (pump and treat).
The source removal portion of this alternative was simulated in the
model as discussed above. Pumping and treatment was simulated in the model
by withdrawing water from a system of 38 wells spaced across the area of the
site. Nodes of the finite element grid (Figure 6) were used to represent
wells, and the wells were situated in 3 rows running north - south along the
east, central, and west portions of the site. It was assumed that the wells
were drilled to a depth of 30 ft below the water table and that the total
withdrawal rate was 100 gpm (379 1/min) (2.6 gpm (10 1/min) from each well).
Pumping was simulated for the first 30 years of the 50 year prediction
period.
The potential surface for the top of Layer 1 for this case is shown in
Figure 24. The average drawdown over the site due to pumping was about 4 ft
(1.2 m).
The total mass of TCE in the flow system, and the total mass discharging
to Mill Creek and the pumping wells over 5-year intervals are shown in
Table 15. The model results show that 91% of the TCE is removed from the
flow system in the first 5 years, and that all the TCE is removed in 15
55
-------
TABLE 14. MODEL-PREDICTED DISTRIBUTION OF TCE FOR
THE CH2M HILL SOURCE REMOVAL ACTION
Time
Step
1 - 4
5
6
7
8
9
10
11
12
13
14-...
Total
Total
End
Year
(Same
1988
1993
1998
2003
2008
2013
2018
2023
2028
2033
(1983 - 2033)
(1963 - 2033)
TCE
Inflow (Ib)
as for Base
0
0
0
0
0
0
0
0
0
0
0
TCE
Outflow (Ib)
Case)
4,929
3,554
2,587
1,734
1,054
623
355
217
135
99
15,287
28,620
Total TCE in
System (Ib)
12,141
8,587
6,000
4,266
3,212
2,589
2,234
2,017
1,882
' 1,783
56
-------
en
Figure 24. Predicted Top of Layer 1 Potential Surface
for CH2M HILL Source Removal and Pump and
Treat (100 gpm) Remedial Action
-------
years. After 1983, 87% of the TCE exiting the system is removed by the
pumping wells; the remaining 13% discharges to Mill Creek and the ditch.
TABLE 15. MODEL-PREDICTED DISTRIBUTION OF TCE FOR THE
CH2M HILL SOURCE REMOVAL COMBINED WITH
PUMP AND TREAT (100 GPM) REMEDIAL ACTION
Total TCE in
System (Ib)
1,512
593
0
0
0
Time
Step
1 - 4
5
6
7
8
14
Total
End
TCE
. TCE Outflow Mb)
Year Inflow Mb) Mill Creek
(Same
1988
1993
1998
2003
2033
as Base Case)
0
0
0
0
0
0
1,880
254
59
0
0
2,193
Pumping Wells
13,679
663
534
0
0
14,876
Source Removal Combined with Pump and Treat (200 gpm)
This remedial action is similar to the previous action except that the
38 wells were pumped at a total rate of 200 gpm (758 1/min) instead of
100 gpm (379 1/min).
The potential surface for the top of Layer 1 for this case is shown in
Figure 25. The average drawdown over the site due to pumping was about 8.5
ft (2.6 m).
The total mass of TCE in the flow system, and the total mass discharging
to Mill Creek and the pumping wells at 5-year intervals are shown in
Table 16. The model results show that all the TCE is removed from the system
in the first 5 years. After 1983, 97% of the TCE exiting the system is
removed by the pumping wells; the remaining 3% discharges to Mill Creek and
the ditch.
Cap Combined with Pump and Treat (100 gpm)
The capping action assumes that the site is covered with a low
permeability material in order to eliminate the infiltration of water through
the unsaturated zone and the resulting leaching of TCE into the saturated
58
-------
en
POT hor una n
Figure 25. Predicted Top of Layer 1 Potential Surface
for CH2M HILL Source Removal and Pump and
Treat (200 gpm) Remedial Action
-------
Time
Step
1 - 4
5
6
14
Total
End
TCE
Year Inflow (Ib) Mil
(Same
1988
1993
2033
as Base Case)
0
0
0
0
TCE
Outflow (Ib)
1 Creek Pumping Wells
545
0
0
545
16,526
0
0
16,526
flow system. The capping action would be followed by the installation of
wells to pump and treat contaminated groundwater.
TABLE 16. MODEL-PREDICTED DISTRIBUTION OF TCE FOR THE
CH2M HILL SOURCE REMOVAL COMBINED WITH
PUMP AND TREAT (200 GPM) REMEDIAL ACTION
Total TCE in
System (Ib)
0
0
0
The cap was simulated in the model by reducing the recharge rate over
the site "'to 0.8 in./yr (2 cm/yr) or one-tenth the estimated annual average.
The pump and treat action was simulated as discussed in the previous case
where the pumping rate was 100 gpm (379 1/min).
The potential surface for the top of Layer 1 for th-is case is shown in
Figure 26. The average drawdown over the site due to pumping was about
4.5 ft (1.4 m).
The total mass of TCE in the flow system, the mass entering from the
unsaturated zone, and the mass discharging to Mill Creek and the pumping
wells at 5-year intervals are shown in Table 17. The model results show that
92% of the TCE is removed from the system in the first 5 years, and that all
of the TCE is removed in 15 years. After 1983, 88% of the TCE exiting the
system is removed by the pumping wells; the remaining 12% discharges to Mill
Creek and the ditch.
Cap Combined with Pump and Treat (200 gpm]
This remedial action is similar to the previous action except that the
38 wells were pumped at a total rate of 200 gpm (758 1/min) instead of
100 gpm (379 1/min).
The potential surface for the top of Layer 1 for this case is shown in
Figure 27. The average drawdown over the site due to pumping was about 9 ft
(2.7 m).
60
-------
VP CF NUN M TOT CTDf UHBI 11
wป CF iua
Figure 26. Model-Predicted Top of Layer 1 Potential
Surface for the CH2M HILL Cap and Pump
and Treat (100 gpm) Remedial Action
Figure 27.
Model-Predicted Top of Layer 1 Potential
Surface for the CH2M HILL Cap and Pump
and Treat (200 gpm) Remedial Action
-------
5
6
7
8
14
Total
1988
1993
1998
2003
2033
98
25
7
0
0
130
1,822
220
47
0
0
2,089
14,040
528
530
0
0
15,105
TABLE 17. MODEL-PREDICTED DISTRIBUTION OF TCE FOR THE
CH2M HILL CAP COMBINED WITH PUMP AND TREAT (100 GPM)
REMEDIAL ACTION
End TCE TCE Outflow fib) Total TCE in
Year Inflow (1b) Mill CreekPumping WeTTs System (1b)
1-4 (Same as Base Case)
1,307
584
0
0
0
The total mass of TCE in the flow system, the mass entering from the
unsaturated zone, the mass discharging to Mill Creek, and the mass removed by
pumping wells at 5-year intervals are shown in Table 18. The model results
show that virtually all of the TCE is removed from the system in the first 5
years. The only TCE remaining in the system after 5 years is the small
amount that enters from the unsaturated zone. After 1983, 97% of the TCE
exiting the system is removed by the pumping wells; the remaining 3/ป
discharges to Mill Creek and the ditch.
TABLE 18. MODEL-PREDICTED DISTRIBUTION OF TCE FOR THE
CH2M HILL CAP COMBINED WITH PUMP AND TREAT (200 GPM)
REMEDIAL ACTION
Total TCE in
System (Ib)
0
0
0
0
0
Time
Step
1 - 4
5
6
7
14
Total
End
TCE
Year Inflow (Ib) Mi
(Same
1988
1993
1998
2033
as Base Case)
97
25
0
0
0
TCE Outflow (Ib)
11 Creek
519
0
0
0
519
Pumping Wells
16,649
25
0
0
16,674
62
-------
PRPS PROPOSED REMEDIAL ACTION
The PRPs proposed remedial action is discussed in detail in the report
by Landau Associates and Dames and Moore (1984). In summary, they proposed a
combination of source removal, a slurry wall around the site, and pumping and
treatment. In order to more thoroughly analyze their proposal in the model,
a series of six model runs were performed.
The first model run was a no-action simulation to provide a benchmark
against which the remedial action runs could be compared. The second model
run simulated a slurry wall combined with source removal. The third and
fourth model runs simulated the combination of slurry wall, source removal,
and pump and treat, at total pumping rates of 100 and 200 gpm (379 and
758 1/min). To evaluate the effectiveness of the slurry wall, the fifth and
sixth model runs were similar to the third and fourth runs except that the
slurry wall was removed.
No Action (Extension of the Base Case)
The PRPs no-action simulation is basically the same as the no-action
case for the CH2M HILL simulations, except that the finite element grid was
modified for the PRPs simulations to contain elements which represent the
slurry wall. In the no-action case the slurry wall elements were assigned
the same permeability as surrounding elements. For cases involving the
slurry wall, these elements were assigned a permeability of 2.8 x 10 ft/day
(10 cm/sec). The purpose of this simulation was to verify that the same
model solution was achieved for the finite element grid with the normal
permeability slurry wall elements as for the grid without the slurry wall
elements.
The potential surfaces for the PRPs no-action case are the same as for
the original base case (Figures 8 to 12). The TCE concentration contours are
nearly identical to those for the CH2M HILL no-action case (Figures 21
through 23).
The total mass of TCE in the flow system, the mass entering the system,
and the mass discharging to Mill Creek and the ditch over 5-year intervals
are shown in Table 19. The numbers in this table are different than those in
Table 13 (CH2M HILL no-action case) because the areas of the Reaction Pond
III and Well 21 sources are slightly smaller (the slurry wall elements
63
-------
TABLE 19. MODEL-PREDICTED DISTRIBUTION OF TCE FOR
THE PRPS NO-ACTION SIMULATION
Time
Step
1*
2*
3*
4*
5
6
7
8
9
10
r:.. 11
12
13
14
Total
End
Year
1968
1973
1978
1983
1988
1993
1998
2003
2008
2013
2018
2023
2028
2033
TCE
Inflow (Ib)
7,034
7,034
7,034
7,034
975
244
61
15
0
0
0
0
0
0
29,431
TCE
Outflow (Ib)
919
2,612
3,815
4,757
4,619
3,497
2,688
1,909
1,231
788
505
358
270
223
28,191
Total TCE in
System (Ib)
6,115
10,537
13,756
16,033
12,389
9.136
6,509
4,615
3,384
2,596
2,091
1,733
1,463
1,240
*Base case.
64
-------
removed a 3 ft (1 m) strip along western edge). After the first four time
steps, when sources around Reaction Pond III and Well 21 are no longer
active, the amount of TCE entering the system from Reaction Pond I is
Identical to the base case, thus, it appears that adding the extra elements
does not change the model results significantly. As a result, all PRPs cases
were simulated with the slightly reduced area of the source around Reaction
Pond III and Well 21.
The results of the PRPs no-action run served as a benchmark against
which the PRPs remedial action simulations could be compared. The results
showed that 28,136 Ib (12,730 Kg) of TCE entered the groundwater flow system
between 1963 and 1983, and that an additional 1,295 Ib (586 Kg) leached in
from the unsaturated zone over the next 20 years. By 1983, 12,103 Ib
(5,476 Kg) were discharged to either Mill Creek (97%) or the ditch (3%), and
16,033 Ib (7,254 Kg) remained in the system.
Source Removal and Slurry Wall
This "model run simulated a combined source removal and slurry wall
remedial action. The source removal action assumes that the unsaturated
material contaminated with TCE is excavated and removed from the site. This
action was simulated in the model by eliminating all input of TCE to the
groundwater flow system after 1983.
The slurry wall was simulated in the model as a series of 3 ft (1 m)
wide elements around the perimeter of the site (Figure 4). The slurry wall
elements were assigned a permeability of 2.8 x 10 ft/day (10 cm/sec) and
the wall extended to a depth of 50 ft below the surface (40 ft below the
water table). The intent of the wall was to contain the contamination on
site, reduce the lateral flow to the wells, and increase upward vertical
flow.
The potential surfaces for this case are essentially the same as those
for the base case (Figure 8 - 12). The potential on site averages about
0.6 ft (0.2 m) higher than the base case potential due to the impact of the
slurry wall.
The distribution of TCE in the flow system at 5-year intervals is shown
in Table 20. Comparing these results to the results for the no-action case
65
-------
TABLE 20. MODEL-PREDICTED DISTRIBUTION OF TCE FOR THE PRPS
SOURCE REMOVAL WITH A SLURRY WALL REMEDIAL ACTION
Time
Step
1 - 4
5
6
7
8
9
10
11
12
13
14
Total
End
Year
(Same
1988
1993
1998
2003
2008
2013
2018
2023
2028
2033
TCE
Inflow (Ib)
as for Base
0
0
0
0
0
0
0
0
0
0
0
TCE
Outflow (Ib)
Case)
3,645
2,513
2,232
1,712
1,265
864
608
457
368
309
13,973
Total TCE in
System (Ib)
12,388
9,875
7,643
5,931
4,666
3,802
3,194
2,737
2,369
2,060
66
-------
shows that the slurry wall prevents the TCE from migrating off site, thereby
increasing the mass of TCE on site and reducing the TCE loading to Mill
Creek. All TCE exiting the system either goes to Mill Creek, (97%) or the
ditch (3%).
Source Removal, Slurry Wall, Pump and Treat (100 gpm)
This model run simulated the remedial action proposed by the PRPs. It
consisted of a combination of source removal, pump and treat at 100 gpm
(379 1/min), and a slurry wall around the perimeter of the site.
The source removal and slurry wall actions were simulated in the same
manner as discussed in the previous case. As stated earlier, for this
remedial action involving pumping from shallow wells, the purpose of the
slurry wall was not only to contain the contamination, but to reduce the
lateral flow of water from off site to the wells, and induce upward flow
through the highly contaminated near-surface materials on site. Pump and
treat was simulated in the same manner as in the CH2M HILL cases. Water was
withdrawn from a network of 38 wells evenly distributed over the site. The
wells were drilled to 30 ft (9 m) below the water table and the total
withdrawal rate from all wells was 100 gpm (379 1/min). The major difference
between the CH2M HILL and PRPs cases is that in the PRPs cases, pumping was
simulated, for the first 5 years of the 50-year simulation period instead of
for the first 30 years as in the CH2M HILL simulations.
The potential surface for the top of Layer 1 for this case is shown in
Figure 28. The average drawdown over the site due to pumping is about 6 ft
(1.8 m).
The distribution of TCE in the flow system for this case is shown in
Table 21. The results show that 84% of the TCE presently in the flow system
(1983) will be removed after 5 years of pumping. All of the TCE is removed
from the system in 15 years (after the first 5 years it discharges to Mill
Creek because pumping is stopped). Of the TCE exiting the system in the
first 5 years, 84% is withdrawn by the pumping wells, and the remaining 16%
goes to Mill Creek or the ditch.
67
-------
en
oo
UP CF HUN 32 POT (TOP UtYEH 11
Figure 20. Model-Predicted Top of Layer 1 Potential
Surface for the PRPs Slurry Wall and
Pump and Treat (100 gpm) Remedial Action
-------
TABLE 21. MODEL-PREDICTED DISTRIBUTION OF TCE FOR THE
PRPS SOURCE REMOVAL, SLURRY WALL, AND
PUMP AND TREAT (100 GPM) REMEDIAL ACTION
Time
Step
1 - 4
5
6
7
8
14
Total
End
TCE
TCE Outflow (Ib)
Year Inflow (Ib) Mill Creek Pumping Wells
(Same
1988
1993
1998
2003
2033
as Base Case)
0
0
0
0
0
0
2,153
1,508
1,069
0 '
0
4,730
11,303
0
0
0
0
11,303
Total TCE in
System (Ib)
2,577
1,069
0
0
0
Source Removal, Slurry Wall, Pump and Treat (200 gpm)
This remedial action is similar to the previous action except that the
38 wells 'were pumped at a total rate of 200 gpm (758 1/min) instead of 100
gpm (379 1/min). The 200 gpm pumping rate is twice the rate proposed by the
PRPs.
The potential surface for the top of Layer 1 for this case is shown in
Figure 29. The average drawdown over the site due to pumping is about 12 ft
(3.6 m).
The distribution of TCE in the flow system for this case is shown in
Table 22. The results show that all of the TCE exits the flow system in the
first 5 years. Of the total leaving the system, 97% is withdrawn by the
pumping wells and 3% discharges to the creek or ditch.
Source Removal, Pump and Treat (100 gpm), No Slurry Wall
In order to evaluate the benefit of a slurry wall, this case simulated
source removal and pump and treat (total of 100 gpm) without the slurry wall.
All parameters were identical to the previous case which simulated the PRฐs
proposed remedial action, except that this case did not include the slurry
wall.
69
-------
UP CF HUM 33 POT (TOP LflYEB 1)
Figure 29.
Model-Predicted Top of Layer 1 Potential
Surface for the PRPs Source Removal, Pump
and Treat (200 gpm), and Slurry Wall
-------
TABLE 22. MODEL-PREDICTED DISTRIBUTION OF TCE FOR THE
PRPS SOURCE REMOVAL, SLURRY WALL, AND
PUMP AND TREAT (200 GPM) REMEDIAL ACTION
End
TCE
Year Inflow (Ib) Mil
(Same
1988
1993
2033
as Base Case)
0
0
0
0
TCE
Outflow (Ib)
1 Creek Pumpinq Wells
481
0
0
481
15,552
0
0
15,552
Total TCE in
System (1b)
0
0
0
The potential surface for the top of Layer 1 for this case is shown in
Figure 30. The average drawdown over the site due to pumping was about 4 ft
(1.2 m). The drawdown was greater in the case with the slurry wall (6 ft)
because the wall cut off much of the lateral flow.
The--;distribution of TCE in the flow system for this case is shown in
Table 23. The results show that 92% of the TCE in the system in 1983 will be
removed after 5 years of pumping and all the TCE exits the system after 20
years (after the first 5 years it goes to Mill Creek because the pumps are
turned off). Of the TCE exiting the system in the first 5 years, 89% is
withdrawn by the pumping wells and the remaining 11% discharged to Mill Creek
or the ditch.
TABLE 23. MODEL-PREDICTED DISTRIBUTION OF TCE FOR THE
PRPS SOURCE REMOVAL AND PUMP AND TREAT (100 GPM)
REMEDIAL ACTION (NO SLURRY WALL)
Total TCE in
System (1b)
,276
845
276
0
0
Time
Step
1 - 4
5
6
7
8
14
Total
End
TCE
TCE Outflow (Ib)
Year Inflow (Ib) Mill Creek
(Same
1988
1993
1998
2003
2033
as Base Case)
0
0
0
0
0
0
1,623
431
569
276
0
2,899
Pumpinq Wells
13,134
0
0
0
0
13,134
71
-------
ro
CF HLJNM POT flOP UWCT
Figure 30. Predicted Top of Layer 1 Potential Surface
for the PRPs Pump and Treat (100 gpm)
Remedial Action with no Slurry Wall
-------
Source Removal, Pump and Treat (200 qpm), No Slurry Wall
A simulation was made with source removal and pump and treat (total of
200 gpm) without the slurry wall. All model parameters were identical to the
PRPs proposed remedial action with the 200 gpm pumping rate except that the
slurry wall was eliminated.
The potential surface for the top of Layer 1 for this case is shown in
Figure 31. The average drawdown over the site due to pumping was about 8.5
ft (2.4 m). The drawdown was greater in the case with the slurry wall (12
ft) because the wall cut off much of the lateral flow.
The distribution of TCE in the flow system for this case is shown in
Table 24. The results show that all the TCE exits the flow system in the
first 5 years. Of the total exiting the system, 97% is withdrawn by the
pumping wells and 3% discharges to Mill Creek or the ditch.
TABLE 24. MODEL-PREDICTED DISTRIBUTION OF TCE FOR THE
"' PRPS SOURCE REMOVAL AND PUMP AND TREAT (200 GPM)
REMEDIAL ACTION (NO SLURRY WALL)
End TCE TCE Outflow (Ib) Total TCE in
Year Inflow (Ib) Mill Creek Pumping Wells System (Ib)
(Same as Base Case)
1988 0 481 15,552 0
1993 00 00
2033 0. 0 0 0
0 481 15,552
73
-------
UP CF RUN aO POT (TOP UTTER 1)
Figure 31. Model-Predicted Top of Layer 1 Potential
Surface for the PRPs Pump and Treat (200
gpin) Remedial Action with no Slurry Wall
-------
SECTION 8
SUMMARY
A summary of the base case and .remedial action simulation results is
discussed in this section.
BASE CASE
TCE was introduced to the groundwater flow system in the model over a
period of 20 years (1963 - 1983) at three source locations: 1) Reaction
Pond I; 2) Reaction Pond III; and 3) around Well 21. Over this 20 year
period the model predicted that 30,400 Ib (13,789 Kg) of TCE was spilled at
the site. Over the same period, about 13,300 Ib (6,033 Kg) discharged to
Mill Creek or the railroad drainage ditch, with 17,100 Ib (7,756 Kg)
remaining in the flow system -in 1983. This number compares well with the
estimated mass of TCE in the flow system of 18,000 Ib (8,165 Kg) calculated
independently by CH2M HILL .(1985) and Landau Associates and Dames and Moore
(1984) based on the 1982 through 1984 chemistry data for the site. Of the
30,400 Ib (13,789 Kg) of TCE that was spilled at the site, the model
indicated that 13%, 37%, and 50% originated in the area of Reaction Pond I,
Reaction Pond III, and Well 21, respectively.
A summary of the results for both the CH2M HILL and PRPs remedial action
simulations is shown in Table 1 (Section 2). A comparison of the total mass
of TCE remaining in the groundwater flow system for the CH2M HILL and PRPs
remedial action cases is shown in Figure la and Ib, respectively (Section 2).
For both the CH2M HILL and the PRPs simulations, the base and no-action
case results in Figures la and Ib are essentially the same. The results
would be identical except that the slurry wall elements in the PRPs cases
reduced the size of the Reaction Pond III and Well 21 source areas in the
model, thereby slightly reducing the mass loading of TCE at these areas.
75
-------
The model results show that Mill Creek has been and will continue to be,
if no remedial actions are taken, the primary discharge point for TCE
migrating from the Western Processing Site. In the base case (1963 - 1983)
and the no-action case (1983 - 2033), about 97% of the TCE exiting the system
discharges to Mill Creek, and the remaining 3% discharges to the drainage
ditch along the eastern boundary of the site. According to the model, no TCE
enters the regional groundwater flow system which flows toward the Green
River; all the TCE remains in the local flow system controlled by Mill Creek
and the drainage ditch. By 1983, a little less than half (44%) of the TCE
that was estimated to have entered the flow system during site operation had
exited to Mill Creek and the ditch. For the no-action simulation, the model
predicted that 89% and 96% of the total mass of TCE that entered the system
discharged to Mill Creek or the ditch by the years 2008 and 2033 (25 and 50
years into the future), respectively.
REMEDIAL ACTIONS
Two sets of remedial action runs were simulated with the model: those
based on the remedial actions proposed by CH2M HILL; and those based on the
remedial actions proposed by the PRPs. The model runs based on the CH2M HILL
proposed actions were:
1) no-action (extend the base case into the future);
2) source removal;
3) source removal and pump and treat (100 gpm);
4) source removal and pump and treat (200 gpm);
5) cap and pump and treat (100 gpm); and
6) cap and pump and treat (200 gpm).
The model runs based on the PRPs proposed actions were:
1) no-action;
2) source removal combined with a slurry wall;
3) source removal, slurry wall, and pump and treat (100 gpm);
4) source removal, slurry wall, and pump and treat (200 gpm);
5) source removal and pump and treat (100 gpm) (no slurry wall); and
6) source removal and pump and treat (200 gpm) (no slurry wall).
76
-------
The no-action simulations consisted of running the base case (1963 -
1983) 50 years into the future in order to provide a benchmark against which
remedial action cases could be compared.
The source removal action (excavation of contaminated unsaturated soils)
alone is ineffective in cleaning up the TCE on site because apparently very
little TCE remains in the unsaturated soils. According to the calculations
shown in Appendix A, only about 1,300 Ib (590 Kg) of TCE currently exists in
the unsaturated zone on site. Of this total, 99% is present in the
unsaturated soils in the area around Reaction Pond I. Most likely, the
source removal action would be effective for constituents with high affinity
for adsorption such as metals and other highly sorbed contaminants.
Therefore, this action should not be construed to be universally ineffective
at the Western Processing Site based on the predictions for TCE.
In the CH2M HILL cases, the source removal and capping actions combined
with pumping and treatment achieved about the same results; the action
involv.ing...source removal is slightly more effective in reducing the TCE mass
loading to Mill Creek, but a slightly greater contaminant mass remains in the
system than for the action involving capping. These results are similar
because both actions essentially eliminate the leaching of TCE from the
unsaturated zone to the saturated groundwater system. The cap eliminates
only about 5 gpm (19 1/min) of recharge over the site, a small number
compared to the 100 gpm and 200 gpm (379 and 758 1/min) pumping rates.
In the PRPs cases, the remedial actions without the slurry wall were
more effective in reducing the total mass of TCE in the system and the mass
of TCE exiting to Mill Creek and the ditch, than the cases with the slurry
wall. The reason for this is that the slurry wall prevented the pumping
wells from removing contamination which is outside the slurry wall. The
slurry wall is effective in reducing the lateral flow of water from the creek
to the puming wells, and thereby reducing the total quantity of water that
requires treatment. Also, the slurry wall is effective in containing
contamination once the pumping ceases. The slurry wall would be more
effective if a cap were also installed over the site to reduce the amount of
recharge. The cap would reduce the leaching and transport of contaminants
through the system and under the wall. Even without the cap, however, the
rate of migration of contamination through the site and under the wall could
77
-------
be low enough that the concentrations would be di-luted to acceptable levels
in the groundwater flow system.
In all cases involving pumping, withdrawal of water at the 200- gpm
(758 1/min) rate removed all the TCE from the system in the first 5 years.
About 97% of the TCE was removed by the pumping wells with the remainder
discharging to Mill Creek (and a trace to the drainage ditch).
In the CH2M HILL cases and the PRPs case with the slurry wall, pumping
at the 100 gpm (379 1/min) rate removed all the TCE in the first 15 years.
In all three cases, most of the TCE (about 90%) exits the sytem in the first
5 years. Over the 15-year period, between 85% and 90% of the TCE is removed
by the pumping wells for the CH2M HILL cases. In the CH2M HILL cases,
pumping was simulated for the first 30 years (1983 - 2013), whereas in the
PRPs cases, pumping was simulated for just the first 5 years (1983 - 1988).
Therefore, in the PRPs cases, all TCE exiting the system after the first 5
years discharges to Mill Creek or the ditch.
In the PRPs case of source removal and pump and treat at 100 gpm without
the slurry wall, 20 years were required for all the TCE to exit the system.
Ninty-two percent of the TCE exited the system during the first 5 years of
pumping. Of this amount, 89% was removed by the pumping wells, and the
remainder discharged to Mill Creek and the ditch. After the pumping ceased
an additional 15 years was required for the TCE to discharge to Mill Creek or
the ditch.
The drawdown over the site for all cases involving pumping ranged
between 4 and 12 ft (1.2 and 4.6 m). The 100 gpm (379 1/min) pumping rate
resulted in average drawdowns between 4 and 6 ft (1.2 and 1.8 m) whereas the
200 gpm (758 1/min) pumping rate resulted in average drawdowns between 8 and
12 ft (2.4 and 3.7 m). For the same pumping rate, the cases with the slurry
wall had about 2 to 3 ft (0.6 to 0.9 m) more drawdown because the wall
eliminated lateral flow to the wells.
In summary, to a large degree the CH2M HILL and PRPs remedial action
cases performed about equally. As expected, the modeling results indicated
that the 200 gpm (758 1/min) pumping rate performed better than the 100 gpm
(379 1/min) rate. These data and other differences which are discernable in
the model results should be considered in the selection of the remedial
action alternative to be implemented at the site.
78
-------
REFERENCES
CH2M HILL. 1985. Feasibility Study Western Processing Kent, Washington.
Draft. Seattle, WA.
CH2M HILL. 1984. Remedial Investigation Western Processing Kent,
Washington. Final Draft. EPA WA 37-0116.1, Seattle, WA.
Dunne, T., and L. B. Leopold. 1978. Water in Environmental Planning. W. H.
Freeman & Co., San Francisco, CA.
Ellis, R. 1984. Personal Communication. Soil Conservation Service,
Spokane, WA.
Environmental Protection Agency. 1984. Western Processing Alternatives
Assessment Study, 1983 Data. Environmental Services Division, EPA Region X,
Seattle, -WA.
Environmental Protection Agency. 1983. Investigation of Soil and Water
Contamination at Western Processing, King County, Washington. Environmental
Services Division, EPA Region X, Seattle, WA.
Environmental Protection Agency. 1980. "Water Quality Criteria Documents;
Availability." Federal Register, pp. 45, 231, 79318-79379, Friday,
November 18.
Gupta, S. K., C. T. Kincaid, P. R. Meyer, C. A. Newbill and C. R. Cole.
1982. A Multi-Dimensional Finite Element Code for the Analysis of Coupled
Fluid, Energy and Solute Transport (CFEST).PNL-4260.Pacific Northwest
Laboratory, Richland, WA.
Gupta, S. K., C. R. Cole and F. W. Bond. 1979. Methodology for Release
Consequence Analysis Part III, Finite-Element Three-Dimensional Ground-
Water (FE3DGW) Flow Model. Formulation Program Listing and User's Manual.
PNL-2939, Pacific Northwest Laboratory, Richland, WA.
Hart-Crowser and Associates. 1984. Hydrologic Assessment Western Processing
Kent. Washington. Seattle, WA.
Landau Associates and Dames and Moore. 1984. Weste.-n Processing Remedial
Action Plan, Phase II. Volumes I, II, and III, Edmond, WA.
Luzier, J. E. 1969. "Geology and Ground-Water Resources of Southwestern
King County, Washington." Water-Supply Bulletin No. 28, Department of Water
Resources, Olympia, WA.
79
-------
National Oceanic and Atmospheric Administration. 1974. CTimates of the
United States - Volume II - Western States. U.S. Department of commerce,
Washington, D.C.
Richter, R. 0. 1981. Adsorption of Trichloroethylene by Soils from Dilute
Aqueous Systems. Contract No. F49620-79-C-0038, Air Force Engineering and
Services Center, Environics Division, Environmental Engineering Branch.
Wood, P. R., F. Z. Parsons, J. Demarco, H. J. Harween, R. F. Long, I. L.
Payan and M. C. Ruiz. 1981. Introductory Study of the Biodegradation of the
Chlorinated Methane, Ethane, and Ethene Compounds. Draft Report prepared for
Water Supply Research Division, EPA MERL, Cincinnati, OH. Summary paper
presented at the Annual AWWA Conference, St. Louis, MO, 1981.
80
-------
APPENDIX A
CALCULATION OF LEACH DURATION AND AMOUNT FROM SOURCE AREAS
Three sources of TCE have been identified at the Western Processing
Site, based on the concentration levels measured in the soil and water.
These three sources are: 1) Reaction Pond I; 2) Reaction Pond III; and 3)
around Well 21. Calculations were made to estimate the mass of TCE present
in the unsaturated zone (both in soil and water) of these three locations
based on 1982 soil concentration measurements (EPA, 1983 and 1984). These
mass estimates were then used to estimate the time required for TCE to
completely leach the unsaturated zone into the saturated zone. Time of leach
calculations were based on the average annual recharge at the site (8
in./yr). The results of these calculations were used to determine the number
of years past 1982 to keep the sources active in the model, and to estimate
the source strength for each time step simulated.
SOURCE AREA DATA
The data used to estimate the mass of TCE in the unsaturated zone of
each source area are summarized below.
Reaction Pond I
The total volume of contaminated unsaturated soil at Reaction Pond I was
estimated to be 46,250 ft3. This number is based on a surface area of
2
9,250 ft and an average depth of unsaturated contaminated soil of 5 ft. The
surface area is the actual area associated with Reaction Pond I. The average
depth of contaminated soil was estimated from the soil sampling depths and
concentrations reported in Table A-l. The TCE concentration in the soil was
estimated to be 600,000 ppb based on the measured soil concentrations from
the borings at Wells 15 and 17 (Table A-l). In order to represent a worst
81
-------
case, the concentration was assumed to be the maximum observed soil
concentration and to be uniform over the area of the disposal site.
Table A-l. TCE Soil Concentrations Near Reaction Pond I
Soil Concentration (ppb)
Depth (ft)
3
6
9
Well 15
U
580,000
180,000
Well 17
U
558,000
350,000
U = Not Detected
Reaction Pond" III
The total volume of contaminated, unsaturated soil at Reaction Pond III
was estimated to be 35,650 ft . This number is based on a surface area of
p
7,125 ft and an average depth of contaminated soil of 5 ft. The surface
area is "-"the actual area of Reaction Pond III. The average depth of
contaminated soil was estimated from the soil sampling depths and
concentrations reported in Table A-2. The TCE concentration in the soil was
estimated to be 700 ppb based on the measured soil concentrations from
Well 20 (Table A-2). In order to represent a worst case, the concentration
was assumed to be the maximum observed soil concentration and was assumed to
be uniform over the area of the disposal site.
Table A-2. TCE Soil Concentrations Near Reaction Pond III
Soil Concentration (ppb)
Depth (ft) Well 20
3 U
6 M
9 676
12 544
U = Not Detected
M = Present but below minimum quantifiable limit.
82
-------
Area Around Well 21
The total volume of contaminated unsaturated soil in the area around
Well 21 was difficult to estimate because the area of disposal is not known.
Therefore, the volume was assumed to be the same as that of Reaction Pond
III, 35,650 ft3. The TCE concentration in the soil was estimated to be 1,600
ppb based on the measured soil concentration from Well 21 (Table A-3).
Table A-3. TCE Soil Concentrations in Well 21
Soil Concentration (ppb)
Depth (ft) Well 21
3 U
6 U
9 116
12 1,520
U = Not Detected
CALCULATION OF SOURCE AREA MASS
The calculations of mass in the unsaturated zone for each of the three
source locations are presented below.
For all source areas, the actual and effective porosities of the
unsaturated material were assumed to be 0.25 and 0.15, respectively. The
distribution coefficient (Kd) of TCE in the unsaturated soils was set at 0.2
at all source locations.
The equation used to calculate the mass of TCE at each source location
is:
= {mass in soil} + {mass in water}
- (Cs)(Vols)(ys) + O)(Cw)(Vols)(Yw) (A-l)
where
total mass of TCE in the soil (Ib)
actual porosity
83
-------
8 = moisture content-
C$ = TCE concentration (ppb) in the soil
Vols = total volume of unsaturated soil (ft3)
Ys = density of the soil (lb/ft3)
Cw = TCE concentration (ppb) in the water fcw = ^ (C$)}
Yw = density of the water (lb/ft3)
W
Kd = distribution coefficient
The density of the soil was calculated as:
Ys = (1 - 6A)(Y) = 121.5 lb/ft3
where:
SA = actual porosity (0.25)
Y = dry density of sand (162 lb/ft3)
The TCE soil concentrations were based on measured soil concentrations
in the unsaturated zone at wells around the source areas. The TCE water
concentrations were estimated to be five times greater than the soil
concentrations based on a distribution coefficient (Kd) of 0.2. The ppb
Q
notation is equivalent to Ib per 10 Ib in the total mass calculation.
Reaction Pond I
Based on the data in Table A-4, the total mass of TCE in the unsaturated
zone at Reaction Pond I was calculated to be 4,670 Ib. The total mass in the
soil and water was calculated to be 3,370 Ib and 1,300 Ib, respectively.
84
-------
Table A-4. Summary of Data Used to Make Total Mass Calculations
at all Three Source Locations
Reaction Pond I Reaction Pond III Well 21
8 0.15 0.15 0.15
0A 0.25 0.25 0.25
Cs (ppb) 600,000 700 1,600
Cw (ppb) 3,000,000 3,500 8,000
Vol$ (ft3) 46,250 35,650 35,650
Ys (lb/ft3) 121.5 121.5 121.5
\ (lb/ft3) 62.4 62.4 62.4
Kd 0.2 0.2 0.2
Reaction Pond III
Based on the data in Table A-4, the total mass of TCE in the unsaturated
zone at Reaction Pond III was calculated to be 4.2 Ib. The total mass of TCE
in the sot! and water were calculated to be 3.0 Ib and 1.2 Ib, respectively.
Around Well 21
Based on the data in Table A-4, the total mass of TCE in the unsaturated
zone around Well 21 was calculated to be 9.6 Ib. The total mass in the soil
and water were calculated to be 6.9 Ib and 2.7 Ib, respectively.
The estimated total mass of TCE present in the unsaturated zone (based
on October, 1982, soil analyses) at the three suspected source areas is
summarized in Table A-5.
Table A-5. Summary of Estimated Total Mass of TCE Present
at the Three Suspected Source Areas
Mass of TCE (Ib)
In Soil In Water Total
Reaction Pond I 3,370 1,300 4,670
Reaction Pond III 3.0 1.2 4.2
Around Well 21 6.9 2.7 9.6
85
-------
DURATION AND AMOUNT OF LEACHING FROM SOURCE LOCATIONS
The preceding calculations show that Reaction Pond I is the only source
location which contains a significant quantity of TCE in the unsaturated
zone. Calculations of the time required for TCE to leach out of the
unsaturated zone (duration), and of the amount that would leach out each
year, were made for Reaction Pond I. This information was used in the
modeling to determine the length of time into the future that Reaction Pond I
would remain active, and the quantity of TCE that would leach from the
unsaturated to the saturated zone at each time step.
Reaction Pond I
The calculations were based on the concentration of TCE in the water of
the unsaturated zone and the rate of water movement through the unsaturated
zone from annual recharge. Equation A-l can be used to calculate the TCE
concentration in the water (C ) based on the total mass of TCE remaining in
w
the system at any given point in time. For convenience, the equation can be
rewritten as follows:
MassTCE
w Vol (Kd Y. + 6 y ) (A-2)
9 J W
The Kd in equation A-2 is equivalent to C./C .
5 W
The volume of water (Vol ) moving through the unsaturated zone from the
W
average annual recharge was calculated as follows:
Volw = Rhg x A = 6,167 ft3/yr
where
Rhg = average annual recharge (0.67 ft/yr)
A = area of Reaction Pond I (9,250 ft2)
This number was used to calculate the mass of water (Mass ) passing through
W
the unsaturated system, as follows:
86
-------
Massw = Volw x yw = 384,800 Ib/yr Y
The initial mass of TCE in the system was 4,670 Ib. For each year a
mass of TCE removed from the system was calculated as Cw times Massw- A new
mass of TCE in the system was calculated as the previous mass minus the mass
removed, and the procedure was repeated. This iterative process continued
until virtually all of the TCE was leached from the system. The results of
this calculation are shown in Table A-6.
Table A-6. Time and Amount of TCE Leaching from the
Unsaturated Zone at Reaction Pond I
Mass TCE (Ib) Mass TCE (Ib)
Year
1
2
3-....
4
5
6
7
8
9
10
Remaining
4,670
3,516
2,647
1,992
1,500
1,129
850
640
482
363
Total Removed
1,154
2,023
2,678
3,178
3,541
3,820
4,030
4,188
4,307
4,397
Year
11
12
13
14
15
16
17
18
19
20
Remaining
273
206
155
116
88
66
50
38
29
22
Total Removed
4,464
4,515
4,554
4,582
4,604
4,620
4,632
4,641
4,648
4,653
Table A-6 shows that virtually all of the TCE in the unsaturated zone
above Reaction Pond I is leached into the saturated zone in 20 yr. Over each
5-year period (the time step used in the model) about 75% of the mass of TCE
remaining is flushed from the system. Based on these calculations, the
Reaction Pond I source was allowed to leach for 20 years into the future in
the model, and the source strength was reduced by 75% at every time step.
Reaction Pond III and Area Around Well 21
The total mass of TCE in the unsaturated zone above Reaction Pond III
and in the area around Well 21 was so small that it was not necessary to make
calculations of the leach duration and amount. As a result, these sources
87
-------
were allowed to leach up to the present, and then turned off for all model
predictions into the future.
-------
APPENDIX B
RECHARGE CALCULATIONS
Recharge due to precipitation was calculated using the water balance
formula:
Recharge = Precipitation - Actual Evaporation - Runoff (B-l)
Average annual precipitation and actual evapotranspiration for the study area
are about 39 in./yr (99 cm/yr) and 18 in./yr (46 cm/yr), respectively (NOAA,
1974). Runoff was calculated using a method developed by the U.S. Soil
Conservation Service and modified by Dunne and Leopold (1978). The technique
is based on a simplified infiltration model of runoff, daily precipitation
events, and empirical approximations which consider such factors as soil
type, land use, vegetative cover, and storm separation interval to determine
the antecedent soil moisture conditions.
A program developed at Battelle and based on the Soil Conservation
Service method "was used to calculate runoff for the Western Processing study
area. The calculations were made using daily precipitation data for 1982 and
1983. The output from the program is a list of runoff estimates for a range
of runoff curve numbers. A runoff curve number of 70 was selected for the
study area based on the soil type (Group B), land use (residential area with
one acre lots), and total impervious area (20%) (Dunne and Leopold, 1978).
The curve number 70 corresponds to curve number 85 if normally wet antecedent
moisture conditions prevail, as might be expected for the area around Kent.
The results of the model for the two runoff curve numbers at several
storm separation intervals for the year 1982 and 1983 are shown in Table B-l.
89
-------
TABLE B-l. RUNOFF PROGRAM RESULTS
Storm Separation Runoff (in,
Interval
(days)
0
1
2
3
1982
CN 70
0.3
6.8
7.9
11.4
CN 85
2.5
14.8
16.9
20.5
./yr)
1983
CN 70
0.3
7.6
16.0
19.2
CN 85
2.9
17.3
25.1
28.1
CN = Curve Number
Using equation B-l, storm separation intervals of 0, 1, 2, and 3 days,
and averaging the runoff over two years yields estimated recharges of about 5
in./yr (13 cm/yr) and 12 in./yr (30 cm/yr) for runoff curve numbers 70 and
85, respectively. The curve number that applies to the area around Kent is
probably between 70 and 85. Therefore, the actual recharge was estimated to
be about 8 in./yr (20 cm/yr).
In the final calibrated model a recharge value of 8 in./yr (20 cm/yr)
was applied uniformly over the local model region except for paved areas on
the Western Processing Site (Elements 128, 129, 130, 143, 144, 145, 146, 152,
153, 154, 163, 164, and 175) where the recharge was set to 0 in./yr.
90
-------
APPENDIX C
STREAM AND LEAKANCE BOUNDARY CONDITIONS
The boundary conditions in the FE3DGW model were defined using the
stream boundary options to describe flux to Mill Creek and the ditch, and the
leakance boundary option to describe flux across the perimeter boundaries.
This appendix provides a more detailed discussion of the data used in the
model to implement these options.
STREAM BOUNDARY OPTION
Surface water bodies are often expressions of the water table. This
phenomenon is often treated in a groundwater model by holding the groundwater
elevation at the level of the surface water. The stream option in the FE3DGW
code allows the potential to fluctuate above or below a stream, and
calculates a flux (to or from the stream) based on the potential difference
between the elevation of the stream and that of the groundwater. The data
required by the model to make this calculation are the stream surface
elevation; the stream bottom elevation, cross-sectional area, thickness, and
permeability; and minimum stream depth. These data were entered into the
model for each node along Mill Creek and the drainage ditch. The model
calculates the flux to (gaining) or from (losing) each node using Darcy's
Law. The data used to implement the stream option in the final calibrated
model for Mill Creek and the ditch east of the site are provided in
Tables C-l and C-2, respectively.
The surface water elevation at nodes along Mill Creek and the ditch were
interpolated and extrapolated from measurements at five locations along the
Creek and two along the ditch (Figure 7). The measurements were made on
April 10, 1984 by EPA Region X.
91
-------
TARLE C-l. Stream Boundary Option Data Used to Simulate Flux to Mill Creek
ro
NUUC
NUMBED
2
12
22
33
34
35
47
58
75
92
122
134
150
159
177
186
204
214
230
239
251
263
275
287
286
289
290
301
309
286
285
272
271
270
L w C r K
Fl t'VATtn J
1
1
I
I
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
I . U6
! ,ซ0
i.75
I .70
1 .67
1.63
1 .61
.60
.59
.57
.565
.558
.55
.5?
.5
.49
.4M
.46
.27
.07
10.87
10.67
L. * r. c i\
LF.'iGTM
246.0
J 6 9 . 0
27'>.0
246.0
246.0
246.0
1 4 * . 0
96.0
140.0
131.0
100. 0
1 >) ft . 0
7'J.O
50.0
K2.0
9H.O
1 1 "I . 0
130.0
165.0
215.0
246.0
29-5.0
10.47 2^5.0
10.35
10.3
10.27
1 0 . 1 a
10.1
10.05
10.21
10.17
10.13
10.09
\ 't 0 . 0
277.0
29'i.i
J24.0
560.0
215.0
1 'to . 0
2 1 :i . o
<> h ? . 0
S^M.O
1 0 . ofc 3/>ft.O
? K P. 10.01
?9 S
ซ*.
.<**
360.0
t 9 ft . t)
U '1 1 1 K
wlOTH
5.0
5.0
5.0
5.0
5.0
5.0
5..)
5.0
5.n
5.0
5.0
5 . 0
* . o
5.0
5.0
5.0
5.0
5.')
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5,0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
s.o
ELEVATION
10.86
10. ft
10.75
10.70
10.67
10.63
10.61
10.60
10.59
10.57
10.565
10.558
10.55
0.52
0.5
0.49
0.48
0.46
0.27
0.07
9.87
9.67
9.47
9.55
9.3
9.27
9.1H
9.1
9.05
9.21
9.1 7
9.13
9.09
9.04
9.01
ซ.9fl
THICKNESS PERME
o.
o.
o.
o.
o.
o.
0.
o.
o.
0.
o.
o.
o.
o.
o.
o.
o.
0.
0.
o.
o.
0.
0.
o.
o.
0.
0.
o.
0.
o.
o.
o.
o.
o.
o.
o.
o.
0.
o.
o.
o.
0.
0.1 0.
O.I 0.
0.1 0.
0.1 0.
0.1 0.
0.1 0.
0.1 0.
0.1 0.
0.1 0.
0.1 0.
0. 0.
0. 0.
0. 0.
o. o.
o, o.
ABILITY
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
142
1 12
run unccrv
DEPTH
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.2S
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
-------
TARLE C-2. Stream Boundary Option Data Used to Simulate Flux to the Ditch
N'mt
NUMBER
51
65
62
99
1 1 1
1 15
1 29
1 11
157
166
1 M
193
209
219
234
LMtt K
ELEVATION
13.
13.
3.
i.
3 .
3.
3.
13.
1 3.
13.
1 3.
13.
13.
12.
1 2.
57
52
19
<45
U2
1
39
37
31
32
3
I -5
0
t\
6
LHt tr,
LENGTH
66. 0
I 25.0
1 >) ft . 0
^2.0
57.0
57.o
60.0
7n.i
(> ti . 0
57.0
5 ' . 0
7 i . 0
b2.n
125.0
73."
C.K
Hi
2
2
2
2
2
2
?
2
2
2
2
2
2
2
2
OTH ELEVATION
.ti 12.57
.0
.0
.0
.0
.0
. 0
. o
. 0
. o
.0
.')
.0
.0
. >*
2.52
2.19
2.15
? i^ 2
2.1
2.39
2.37
2.31
2.32
2.3
2.15
2,o
t.e
1.6
C*EtK
THICK
0.
o.
0.
o.
o.
o.
0.
0.
o.
0.
o.
o.
o.
o.
o.
NESS PERMC
i
i
i
i
i
i
i
i
i
i
i
i
i
l
i
o.
o.
o.
o.
o.
o.
o.
o.
o.
o.
o.
o.
o.
o.
o.
ABILITY
112
112
112
112
112
112
112
112
112
112
112
112
112
112
112
11N tซtEK
DEPTH
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
-------
LEAKANCE BOUNDARY OPTION
The "leakance boundary condition" option of the FE30GW code allows
flexibility in defining external boundaries of the model region. Rather than
a constant flux or held potential at the boundary, the leakance option allows
the potential and flux to be varied depending on the conditions which exist
within the study area.
The data required by the model to make this calculation arethe distance
from the boundary to a known potential, the potential at that distance, and
the cross-sectional area of the boundary. These data are entered into the
model for each node along the boundary, including both surface nodes and
nodes at depth. The model calculates a boundary flux at each node using
Darcy's Law, which is then used to calculate the potential at the boundary.
A map depicting the regional potentials used to calculate the
groundwater potential at certain distances from the boundaries is shown in
Figure C-l. This map shows the distances to the extended boundary and the
gain or lass in potential elevation out to these distances.
The Green River elevations were interpolated from three measurements
taken by EPA Region X in April, 1984 :
1. east of benchmark 32 (southwest of site) - 9.8 ft (3.0 m) AMSL;
2. east of benchmark 22 (west of site) - 8.8 ft (2.7 m) AMSL; and
3. Tukwilla Gauge (north of site) - 7.9 ft (2.4 m) AMSL.
94
-------
Regional Potential Value in Feet
LQCflL MODEL RESIGN
+3 ft
VU1M4 P1VCEB8DC SITE
Figure C-l. Locations of Regional Values Used to Determine
Boundary Conditions
95
-------
APPENDIX D
CALCULATION OF RETARDATION FACTOR
The retardation factor, K, can-be calculated by the formula
K = 1 + B Kd f0'1)
where B is defined as the bulk density divided by the effective porosity, and
Kd is the distribution coefficient. The dry density of the silty sand at the
Western Processing Site was estimated as the bulk density of sand
(2.6 giti/cnr) times the quantity 1.0 minus the actual porosity (1.0 - 0.4 =
0.6); or"1.6 gm/cm . Kd's for TCE have been reported in the range of 0.1 to
1.0 cnr/gm (Richter, 1981), depending on the soil type. A Kd of 0.3 cnr/gm
was used to represent the silty sand material at the Western Processing Site.
Using the bulk density for silty sand of 1.6 gm/cm and an effective
porosity of 15%, yields a value of 10.7 for B. Substituting B and Kd into
equation D-l yields a K of 4.2, which means that the TCE travels about four
times slower than the groundwater.
This value of K was used as a guide in determining the retardation
factor to use in the final calibrated CFEST model.
96
-------
ANALYSIS OF THE WESTERN PROCESSING MODEL SENSITIVITY
USING LATIN HYPERCUBE SAMPLING
To better understand Battelle's groundwater flow and contaminant
transport model of the Western Processing Site, a sensitivity analysis was
performed using Latin Hypercube Sampling (LHS). LHS is a constrained
sampling scheme which selects values within a specified range and
distribution (Iman and Shortencarier, 1984). These values can then be used
as model input parameters so that the correlation between the tested input
parameters and the model results can be determined.
A Latin Hypercube Sample is a multiparameter (multivariate)
sample composed of a number of replications of individual realizations of
the multiparameter set. Each individual parameter is assigned to a
probability distribution (normal, lognormal, uniform, loguniform,
triangular, beta, or user-defined). The range of each probability
distribution is partitioned into a number of equally probable intervals; a
uniform distribution is divided into a number of intervals of equal length.
LHS involves selecting a parameter value for each of the equally probable
intervals (according to the conditional distribution of that interval) and
then randomly permuting the orders of intervals in order to introduce the
proper degree of correlation between parameters. The main advantage of LHS
is that the entire range of the parameter is sampled in an efficient
manner.
The LHS technique was applied to analyze the sensitivities c"
five parameters with respect to the conditions at the Western ฐrocessing
Site. These parameters are: the hydraulic conductivity of Unit 1; t^e
hydraulic conductivity of Unit 2; the effective porosity; the recharge; and
the retardation factor for trichloroethylene (TCE). The sampling range and
distribution for each parameter are shown in Table 1. These distributions
were input into the LHS program which generated 25 realizations of these
parameters (Table 2). These parameter realizations were then input into
CFEST to create 25 realizations of the model output. Note that the source
strength and dispersivity were not analyzed in this study.
-------
Table 1. Range and Distribution of Parameters
Used in LHS Analysis
Parameter
Hydraulic Conductivity
(Unit 1)
Hydraulic Conductivity
(Unit 2)
Porosity
Recharge
Retardation Factor
Range
1 to 10 ft/day
10 to 100 ft/day
10% to 35%
6 in./yr to 12 in./yr
2 to 8
Distribution
Uniform
Uniform
Uniform
Uniform
Uniform
-------
Table 2. Western Processing Sensitivity (Latin
Hypercube Sample Input Vectors)
Run No.
1
2
3
4
5
6
7
1.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
K*
Unit 1
x (1)
1.40
6.26
6.80
9.31
4.07
5.30
6.47
3.62
1.00
2.42
9.27
7.72
7.24
1.95
3.08
4.86
7.98
5.34
8.85
9.88
2.59
3.47
4.57
5.73
8.52
K*
Unit 2
x (2)
60.0
21. 9
47.5
96.8
55.1
75.5
19.3
84.5
90.1
16.6
93.9
70.4
28.5
37.4
73.2
12.8
64.7
52.6
41.0
27.9
45.3
86.5
62.4
79.0
33.9
Porosity
x (3)
0.318
0.147
0.169
0.227
0.306
0.261
0.320
0.280
0.123
0.237
0.172
0.246
0.135
0.218
0.186
0.159
0.252
0.194
0.3*1
0.107
0.206
0.119
0.298
0.286
0.339
Recharge
in/yr
x (4)
11.1
10.5
10.6
7.24
8.91
7.82
6.47
8.52
10.8
9.50
9.28
7.05
8.26
6.70
6.01
8.66
6.96
9.71
11.5
9.92
7.66
11.6
10.2
8.16
12.0
Retardation
x (5)
7.54
7.85
4.30
4.78
4.04
7.13
3.58
6.49
4.57
2.90
5.31
7.41
6.29
5.38
2.20
5.63
6.62
5.87
5. 03
3.70
3.21
2.99
6.94
2.34
2.66
* ft/day
-------
The MINITAB statistical package was used to analyze the
relationship between parameter input and model results. The C
classifications used in MINITAB are shown in Table 3. Histograms (which
show the distribution of the model results) and descriptive statistics of
the model outputs (1983 concentrations, flux to Mill Creek, and total mass
of TCE in the groundwater system) are shown in Tables 4 and 5,
respectively. The histograms and descriptive statistics indicate that the
overall changes in input result in less than an order of magnitude change
in model-predicted 1983 source concentrations. But these changes do result
in order of magnitude changes in Mill Creek flux and predicted 1983 Mill
Creek concentrations.
The model results were regressed upon values of the input
parameters (Apppendix A). The regressions were analyzed by comparing the
t-statistics (t-ratio) for the regression equations. A summary of the
regression t-statistics is shown in Table 6. A higher absolute value of
the t-statistic indicates a greater correlation between the parameters.
For example, in Table 6 the highest t-statistic for the flux to Mill Creek
is for the hydraulic conductivity of Unit 1. Therefore, the flux to Mill
Creek is primarily controlled by the conductivity of Unit 1. In this
example, the conductivity of Unit 2 has the next highest t-statistic and
therefore is the secondary controlling factor. T-statistics with absolute
values less than two are considered insignificant. Examination of t-
statistics for the input parameters indicates that the parameter with the
greatest affect on the model results is the hydraulic conductivity of Unit
1. The results of the regression analysis for each of the model output
variables' are discussed below.
The predicted 1983 concentration at Well 21 is primarily
controlled by the hydraulic conductivity of Unit 1, followed by the
retardation factor, effective porosity, and recharge. The predicted 1983
concentration at Reaction Pond I is almost entirely controlled by the
hydraulic conductivity of Unit 1 with little or no influence from the other
parameters. The dominance of the Unit 1 hydraulic conduct! vi'ty indicates
that the hydraulics (the conductivity and gradient) are dominating the
-------
Table 3. Key to C Classifications
Cl = 1'983 TCE Concentration at Well 21 (ppb)
C2 = 1983 TCE Concentration at Reaction Pond I (ppb)
C3 = 1983 TCE Concentration at Reaction Pond III (ppb)
C4 = Flux to Mill Creek (cfs)
C5 = 1983 TCE Concentration in Mill Creek (ppb)
C6 = Hydraulic Conductivity of Unit 1 (ft/day)
C7 = Hydraulic Conductivity of Unit 2 (ft/day)
C8 = Porosity
C9 = Recharge (ft/day)
CIO = Retardation
Cll = 1968 Total Mass of TCE in Groundwater System
C12 = 1973 Total Mass of TCE in Groundwater System
C13 = 1978 Total Mass of TCE in Groundwater System
C14 = 1983 Total Mass of TCE in Groundwater System
C15 = 1988 Total Mass of TCE in Groundwater System
C16 = 1993 Total Mass of TCE in Groundwater System
C17 = 1998 Total Mass of TCE in Groundwater System
C18 = 2003 Total Mass of TCE in Groundwater System
-------
Table 4. Histograms of Model Results
P1TB > HI37 Cl
Histogram of Ci N = 25
(1983 TCE Concentration at Well 21)
Midnoint Count
7 *******
i17B > HI37 C4
Histogram of C4
(Flux to Mill Creek 1
Kirfooint Count
N = 25
-w-t-c-i'*.'
******
0
3 ***
N7B > HIST C2
Histo_prarn of C2 N = 25
(1983 TCE Concentration at Reaction
Micaoint Count Pond I)
ฃ0888
80888
1 1710800
120088
148888
1&8088
1 80888
280888
223888
248888
2&088C
258000
5
ฃ
4
4
0
2
c!
1
i
0
ft
0
1
*****
******
****
****
**
**
*
*
8. 1
8.2
ฃi. 3
8. 4
e.s
8. ฃ
0.7
0. a
8.9
1. 0
I. i
"17B > HI
1
1
3
4
4
3
4
1
4
1
2
37 C5
Histogram of C5
(1983 TCE
PI idooi nt
20
48
ฃ8
88
180
120
148
*
*
***
*
*,***
***
****
*
****
*
**
N = 25
Concentration in Mill Creek)
Count
1
4
2
9
8
ฃ
2
*
****
**
*********
******
**
180
(1983 TCE Concentration at Reaction
Kicaoint Co MI-IT: Pond III)
ฃ'Z!'2'0i3 ' 4 **** '
IB'!'8 8 8 ฃ ******
288C88 0
24|Z|iZ100 5 *****
2E8888 8
Histogram of Cll
(1968 Total Mass of TCE in System)
tfidooint Count
4888
5288
5ฃ'?8
ฃ888
ฃ400
ฃ888
7208
7ฃ88
1
8
1
3
4
10
5
1
*
*
***
#***
*****
*****
*
-------
Table 4. (Cont'dK
S 1 n .* ;
Histogram of C12 N = 25
(1973 Total Mass of TCE in System)
Xidsoint Count
ฃ0.00
7000
3ฃ:20
5'Z'00
183133
11000
12000
1 3000
14000
15000
1 *
0
0
1 *
7 *******
0
8 ********
A ****
3 ***
1 *
MTB > HIST C13
Histocram
(1978 Total
^licooint
ฃ000
,_B000
10000
12000
1A000
16000
18000
ฃ0000
22000
MTB > HIST
Hist oar am
(1983 Total
M iCDOint
6000
8000
10000
-ฃ'200
'. -000
1 ฃ000
13000,
2'?ฃ30
2ฃ.ฃ'i'0
ฃA^0 1?
26000.
28000
of C13 N = 25
Mass of TCE in System)
Count
1 *
0
0
7 *******
1 *
7 *******
5 *****
3 ***
1 *
CIA
of CIA N = 25
Mass of TCE in System)
Count
1 *
0
?
2 **
5 *****
1 *
0
g *******.**.
2 **
3 ***
1 *
1 *
:ซ.~3 > nIST Cl5
Histogram of CIS N = 25
(1988 Total Mass of TCE in System!
tficooint Count
0 1 *
0
6
1208)0 2
10001? 3
ฃ0(2 01? 2
1
******
**
*********
*+
****
*
tfTB > HIST C16
Histogram of CIS N = 25
(1993 Total Mass of TCE in System)
^.iciacint Count
0 1 *
4^00 3 ***
8000 5 *****
12000 7 *******
160C0 3 ***
ฃ0000 5 *****
ฃ4000 0
ฃ8000 1 *
MTB > HIST C17
Histogram of C17 N = ฃ5
(1998 Total Mass of TCE in System)
Midooint Count
0 1 *
4000 7 *******
3000 6 ******
12000 5 *****
j C. 71 iTl '21 ', *
-''.~~.B.... .t ,--~'.'L2-. Z lo
~.iฃ".'- crarfi OT ClS .- - _-j
(2003 Total Mass of TCE in System)
l^iciDoirit Count
f\ = 2
5 *****
_ 'ฃ.' C
^AJi?
3 *-ป*-
*
-------
[able 5. Descriptive Statistics for Model Results
REACTION REACTION MILL CREEK MILL CREEK
WELL 21 POND I POND; III FLUX CONCENTRATION
N
MEHN
MEDlfYN
T MLlflNJ
STDEV
GEMEON
KAX
MINI
Q3
Ql
Cl
ฃ5
331701
332803
374252
1 38385
27677
703421
225303
474029
272914
2
ฃ5
111-454
94962
106443
54375
10875
ฃ83913
54238
138981
70451
C3
ฃ5
159548
132970
152643
71768
14354
389244
8Q675
199807
102434
TMEAN.= Mean of the trimmed
SEMEAN = STDEV/ /N~
Q3 = Third quartile (7SX.)
Ql = First quartile (2^)
CD
C'=-
i^J
0. 638
0. ฃ40
0.637
0. ฃ83
0. 057
1. 140
0. 139
0.909
0. 440
sample (less the upper and lower 5%)
C7
"ปtr
C.U
88. 1
81. 6
86. 8
38. 3
7. 7
188.4
. 17.0
120.8
60. 6
Total Mass in the System
1968
Cll
1973
N
VEPN
MEDIflN
TKCfiN
STDEV
BEMEON
MAX
MIN
Q3
. _ , f
ฃ643
6831
6678
576
115
750ฃ
4976
7046
-." f^
; 1650
12170
1 1760
1864
373
14722
6244
1 ฃ958
1978
C13
ฃ5
15555
16377
15701
3426
685
ฃ1595
6145
17973
1 ฃ633
1983
C14
ฃ5
18640
19664
18785
5083
1017
ฃ8164
5777
ฃฃฃ80
1 4LT'c'B
1988
CIS
ฃ5
15371
16233
15433
6201
1ฃ40
ฃ7936
1385
19829
-6 :!..;
1993
C16
ฃ5
1ฃ434
1ฃ818
1S395
6584
1317
27005
0
17157
:>?." /
1998
C17
~tET
C. wJ
9706
9869
6647
1329
ฃ6002
0
. 1 4629
377?
2003
CIS
PS
8ฃ79
6998
7910
649E
1298
ฃ5039
0
1ฃฃ93
c3'
-------
Table 6. Sumtrary of T-Statistics for Regression Equations
K Unit 1
K Unit 2
Porosity
Recharge
Retardation
Factor
1983
Well 21
Concentration
r!2.72
1.16
-4.33
-3.35
-7.27
1983
Reaction
Pond I
Concentration
-8.98
0.08
-1.59
0.03
-1.13
1983
Reaction
Pond III
Concentration
-9.73
0.89
-2.90
0.37
-2.89
Mill
Creek
Flux
32.59
6.47
1.29
-1.25
0.39
1983
Mill
Creek
Concentration
11.64
0.56
-7.33
1.25
8.95
1983
Total TCE
Mass in
Svstem
-11.40
-0.76
7.34
-1.14
8.90
-------
transport at this source and the retardation has very little effect. The
predicted 1983 concentration of Reaction Pond III is also dominated by the
hydraulic conductivity of Unit 1, while the effective porosity and
retardation factor affect the concentration to a significant but to lesser
degree.
The hydraulic conductivity of Unit 1 dominates the predicted flux
to Mill Creek. To a lesser degree, Mill Creek flux is also influenced by
the conductivity of Unit 2. The predicted 1983 TCE concentration in Mill
Creek is about equally dependent on the conductivity of Unit 1, the
retardation factor, and the effective porosity, with the hydraulic
conductivity having a slightly greater influence.
Regressions were also performed on the predicted total mass of
TCE -remaining in the groundwater system and the input parameters. (Note
that the mass input was constant but the mass exiting the system is
variable.) The total mass in the system is controlled by the hydraulic
conductivity of Unit 1 followed by the retardation factor and effective
porosity. The conductivity has more control after the sources are no
longer active. Comparison of the predicted 1983 concentration at three
sources and the total mass remaining in the groundwater system indicates
that the total mass of TCE is controlled primarily by Reaction Pond I,
followed by Reaction Pond III. Well 21 has no effect on the total mass of
TCE in the system. The differences between the contributions of tne source
areas is probably due to the fact that the Reaction Pond sources discharge
to '-1'ill Creek while the Well 21 source may nave li'tle e'~ect :n ''-:~"
Creek.
In summary, the LHS analysis has shown that, of the parameters
tested,- the model is most" sensitive" to'the hydraul ic' "conductivi'ty of Unit
1. The flux and contaminant loading to Mill Creek, the total mass
remaining in the system, and the peak concentration at the source areas are
all controlled by the Unit 1 conductivity. Because the hydraulic
conductivity of Unit 1 is the major controlling factor, confidence in the
conductivity values used as input to the nodel will 'esult in the Creates:
reliability in model results. The conductivities used in the mode"! have
-------
been verified by field testing (at least in the northern portion of the
site), therefore much confidence can be placed in the flow-portion of the
model.
-------
APPENDIX A
MINITAB Regression Equations
-------
Cl
C6-C10
The renression ecuation is
Cl ป 993017 - 39189 C6 i- 366 C7 - 486237 C8 - 15677 C9 - 3
Pred ictor
Constant
C6
C7
ca
C3
C i|?i
Coef
338017
-39133
3ฃฃ. A
-486287
-15ฃ77
-33334
Std. Coef
ฃ3531
3082
315. 8
11ฃ33ฃ
4573
4677
t-rat 10
16.47
-12. 72-
1. IB
4, 33
-3. 35
-7=27
3 = 4ii'4c, R-so = 33.37C
Ana lysis of Variance
ource
=
-------
MTB > REGRESS C3 5 C6-C10
The rearessiori eauation is
C3 - 364534 - ฃฃ327 C6 * ฃ08 C7 - ฃ42655. C8 * 1275 C3 - 10048 C10
Predictor
Constant
C5
C7
C8
C3
C10
Coef
364534
-22327
208. 1
-242655
1275
-10048
Std. Coef
45095
2294
235. 1
63607
3478
3481
t-rat io
8. 38
-9.73
0.89
-2.90
0.37
-ฃ. 89
S = 23805 R-sq = 86.346 R-so(adj) = 82.753
Analysis of Variance
Source DF ~ S3 MS
Regression 5 1.0674E+11 2.1347E+10
Error 19 1.687SE+10 853303925
Total 24 1.2362E+11
MTB > REGRESS C4 5 C6-Cxtf
The regression eauation is
C4 = - 0.0335 * 0.104 C6 * 0.00211 C7 + 0.149 C8 - -3. .?:ฃ'?- C3
^red ict or
Icr star,.;
C6
C7
H5
ri0
:: = 0.^41,
Coef
-0. 03353
U' . 1 'ซ. ^. s & /
2.2021131
C. 143Z
^1 t \ฃl lฃ. _ Q ^ Q
* 3 -r-_
Std. Coef
D. 05256
0. 003187
C. 0003255
0. 1152
Spoils?
58. 293 ?-sc
t-rat io
-0. ฃ3
32. 53
S. 47
:"t!
c'. 39 v
Analysis of'Variance
Source DF SS PO
Regression.-- 5- . ^1,63276..- .-C. 37ฃ5ฃ
Error 13 ฃ..23253 D. 0^172
ToTai 24 1. 31537
-------
The regression eauation is
C5 = 126 + 10.0 C6 -*- 0.0496 C7 - ฃ30 2e * 1.64 ซ-3 - 11.7
Predictor Coef Std.Coef t-ratio
7. 41
11.64
3.5E,
-7- 33
j.
-a.
S =. li.il 3-so = 33.ฃฃฃ R-so(ad.i) = 91.411
Anaivsis of Variance
Source DF S3 MS
Renression 5 32753.3 6550.7
Error 19 ฃ388.9 125.7
Total ฃ4 35142.3
Constant
C6
C7
CO
C9
C10
125. S0
10. 0436
0. 04957
-ฃZ2i. 41
1 . 64 1
-_ 1 . 723
16.37
0. 6630
0. 08844
31.45
1 . 309
1.303
MTB > REGRESS Cll 5 C6-C10
The regression ecmation is
Cll = 5973 - 138 Cฃ - 0.49 C7 - 33S5 C3 - ฃ4.1 19 - 154
ฐredictor
Constant
C8
C7
28
u Z?
21 '2i
Coef
5973. 3
*"" ฑ ^ f w>wf
-ป. 486
2334. 3
-24. 07
154. 45
Std. Coef
378, 3
13. ฃ4
1 . 37ฃ
701. 4
ฃ3. 18
ฃ3. ฃ0
t-rat io
15.75
-7. 15
~ฃ . ill!
5. ฃ3
G. 33
cr '-IQ
_j . c. y
= 35.100 R-sa(adi) = 81.173
Analysis of Variance
Source DF SS " ' 'MS
Regression 5 6764025 1356805
Error 13 '1187763 62514
Total ฃ4 7371733
-------
The renression ecuation is
C12 = 3ฃ98 - 453 C6 - 3.51 C7 + 12133 C8 - 31.0 C9 + 553 C10
Predictor
Constant
Co
C7
C8 .
i* ""
.""' ' ">
* x1
Coef
3698
-459. 25
-3.513
12133
-91.123
=;^T "?~7
<_f -Jwซ j r
Std.Coef
1019
51. 85
5.31A
1890
78. 63
73. ฃ3
t-rat 10
5.51.
-8. So
-0.ฃ6
6. 42
-1. 1&
7.03
S = ฃ73.3 '-so = 33. ฃ51 Ft-sc(adi) = Sฃ. 3ฃ7
5 is of Variance
a DF SE
Error
"oral
j
13
rcS
14343548
45233:.
6334333
>-
ci
The reqression eouation is
C13 = 12002 - 8ฃ7 CS - ฃ.ฃ0 C7 t- 21ฃ35 C8 - 156 C3 + 1043 C10
Pred ict or
Const ar.~
Cฃ
C7
^~ G
C3
C10
3 = ^13
S -, o E v 5 ' = " "
'E/jrce'""' ' '
;.5GrS = 5 j.0.''
ป 'r" *" O v"
" ;- ai -
Coef
12Q02
-3ฃฃ. 33
-ฃ. ฃ'23
21ฃ35
-155. 7
1043. 3
~,-50
var ia.r.c =
' r'r '
5 15
13 2
24 28
Stc. Coef
lฃS7
85. 73
8. 732
3127
130. 1
130. 2
~> - ~ "-ป> 3
_/ - . -Jll-'t "
ฃS
B 1-3 11 04 .5
3312534
1773ฃ3ฃ
t ra~ 10
7. 1;L
-10. 10
-0.75
ฃ. 32
-1.20
8. 01
"" "" 3 ""
a
rฃ '
1 ฃ32220
124ฃ7ฃS
-------
PITS > REGRESS C14 5 C6-C10
The rsaression squat' ion is
C14 = 13407 - 1322 C6 - 3. 0 C7 + 31008 C3 -
C9 + 1566 C10
Predictor
Const ant
C6
C7
C8
C9
C10
Ccef
13407
-1322.4
-9.03
3100S
-200. 2
1565.6
Std.Coef
2280
116.0
11.88
4227
175.8
176.0
t-rat io
' 5.88
-11.40
-0.76
7.34
-1. 14
8.90
S = 1537 R-SQ = 33.042
Analysis of Variance
DF
Source
Regression 5
Error 19
Total ฃ4.-
SS
57ฃ30ฃ344
43144264'
ฃฃ0051200
R-sq(adj) = 91.211
MS
115381332
2270751
MTB > REGRESS CIS 5 C6-C10
The renression equation is
CIS = 8906 - 1667 C6 - 7.1,ฃ7
3red
/*
(~O
Co
C7
CB
C3
ictor
ns-cant
~~ J.
-
Coef
89'Ziฃ
667.0
-7. 10
36106
131. 5
Std
. Co
ฃ5
1 ~? 1
13.
ef
Sฃ
fcJ
48
t-rat
-
4794
Cl? %V 1307. 1
c
17,09
V51S Of
Source
^s
.pr
35 si on
Er<"or
" " .
: a
;
Var
SF
. ,,,5
-9
14
R-SQ =
iance'
s .. ,.as,7
crer
*-i -l
33. 387
SS
3.223.68
488452'
133
139
..4
. ฃ
- R-SQ
(adj
MS
ซj
I ^!.
-0.
7.
-C.
9.
) =
i
ซ.
ฃ.
ji
i.
6
53
c-
^J
g
"^1
4.
56
92. 435
.. ... :.. s .-v.
.. .17.3464.480 ... . . , ,, - ,,-.
2920445
32231 aais. '
-------
MTB > REGRESS C16 5 CฃZ10
The redression eauation is
Clฃ = 5543'- 1826 Cฃ - 0. ฃ C7 + 35863 CS - 105 C3 + 1373 ClS?
Prea ictor
Constant
Cฃ
""7
Co
C3
ซ - I"*
Coef
5543
-182ฃ. 1
-0. ฃ1
356S3
-105. 0
1373.2
Std. COST
2344
144. ฃ
14.82
5272
213. 3
213.5
t-ra~ i.o
3 c*
-12. ฃ2
-0. 04
ฃ.30
-ฃ:. 43
3. 02
S =1330 R-SQ = 93.548 R-sa(adj) =
Analysis of Variance "
Source DF SB , Kฃ3
Rscre = sior: 5 3731'?24o4 13452:i?435
Error 13 ฃ711333ฃ 3532537
Total 24 1040221824
MTB > REGHES3 C17 5 C6-C10
The rearession equation is
C17 = 2325 - 1884 C6 + 6. 3 C7 + 33023 C8 + 4 C3 + 1320 C10
Prsd ictor
Constant
u- O t&
C7 " '
ca
C3
C10
S = 2182
Coef
2325
-1383. 7
8.8ฃ
33023
4. 5
-1323. 3
R-SQ: =
Std. Coef
3302
iฃ7. 3
17.21
6121
254. 7
254. 8
31,4ฃ8 R-so
t-rat io
0.89
-11.22
0. 51
5. 33
0. !?ฃ
7.54
(ad.i)- = 8-3. 223
Analysis of Variance
Source DF SS MS
Regression .5 ... 97.00i345.44 ... ,194000912,.
Error "" ''" 13' ' "9'047845ฃ 47ฃ2024
"j-al 24 10ฃ0453^C'3
-------
C6-C10
he rsrression sauation is
Ci8 = 783 - 1372 C6 + 21.0 C7 -t- 28375 CS + 143 C3 * 1766 C10
Predictor
Constant.
C6
C7
C8
C3
C10
S = 2335
Pinalvsis of
Source
Regression
Error
Total
V.~B > RE3RE
Coef Std.Coef t-rat
783 3624 . 0.
-1871.6 184.3 -10.
21.00 18.83 1.
28375 6713 4.
143.4 273.5 '0-
176.6.0 273.7 6.
y
R-sa = 89.225 R-sq(adj) =
Variance
DF SS MS
5 302565760 180513152
19 108937848 5736729
24 10115ฃ3584
35 C14 Z Ci-13
io
2ฃ
15
11
31
51
31
86. 389
fiie regression equation is
C14 = 16232 +0.0009 Cl + 0.308 C2 - 0.203 C3
Predictor
Constant
Cl
C2
C3
3 = 32M
":=.'-<=.'-= '-.'-"
vr.,'~~5i,_r]
Tot al
Source
Cl
L. j
Coef Std.Coef t-rat
16292 1355 8.
0.00085 0.01201 ' 0.
0.30807 0.06225 4.
-0.20253 0.05336 -3.
R-sa = 65.323 R-sq(adj) =
v a r i a r. c s
D- SS MS
24 620051200 " '"
DF Seq SS
1 23078470
T ฃ52764656 ' -'"'
1 1.1 31 3-217 1ฃ
io
33
07
35
41
: 60. 363
,- .,. ..-.- . ....,-.--- - ..;/> <_;-., -| "
-. A
-------
he regression eauation is
C14 = 30298 - 134 C5 * 22. B C6
--sdic'cov' Coef S~a. 3oef tratio
C :-=: an- 30293. 4 32.7 326.98
-5 -133.786 . 1.221 -109.60
'-= 22.79 17.54 1.30
S = 173.8 R-sa = 33.893 R-sa(adj) = 99.883
Analysis of Variance
Source DF SS MS
Regression 2 619386752 309693376
Error 22 664470 . 302123
DF Sec SS
1 619335744
51012-7
1
MTB > RE3RESS C14 ฃ C4 C5
The regression eouation is
C:4 = 312317 - ฃ7 C4 - 133 C5
^r-ecictor Coef S.td.Coef t-ra-io
Constant 32i31ฃ.6 iiZlia. 0 3i?3. e-4
C4 ฃ=,. 7 1Sฃ. 9 I3.il
C5 -133. i35ฃ l.ฃ0ฃ -1HZI.G6
-3 = 1T(5. 7 P-SG = 99. ฃ23 R-SG(acj) = -rr. B~
Analysis of Variance
Source J)~ ' ' ...S3 :'. ........ flS ....,..-.".
Rfecre'ssiofi ' :":ฃ " ' 'ol934i 1ง4 ' : ' 3ฎ'3&7C59ฃ'
32275
Erป'or
To-tal
Source
C4
C3
c!ci!
24
DF-
1
1
71 '3257
620I2512S4
SBQ- SS
224095344
395245792
-------
WESTERN PROCESSING
HYDROLOGIC CHARACTERIZATION
by
C. M. Eddy
J. M. Doesburg
Battelle Broject Management Division
Office of Hazardous Waste Management
Richlandj Washington 99352
March, 1985
Prepared for
Environmental Protection Agency
Region X
Seattle, Washington 98101
-------
WESTERN PROCESSING HYDROLOGIC CHARACTERIZATION
Slug tests and borehole dilution tests were performed at the
Western Processing Site (Figure 1) to better characterize the groundwater
flow system. The tests were performed during the period February 11-13,
1985, by Jim Doesburg, Mary Lilga, and Chris Eddy of Battelle's Office of
Hazardous Waste Management. The results indicate that hydraulic
2
conductivities of the zones tested range from 5 x 10" cm/sec (142 ft/day)
to 4 x 10 . cm/sec (1 ft/day). The test procedures and results are
discussed below. The field forms and plots used to calculate the hydraulic
conductivity are included in Appendices A and B, respectively.
SLUG TEST PROCEDURES
The hydraulic conductivity of an aquifer can be determined by
instantaneously changing the water level in a well and observing the
recovery. For the Western Processing Site, a 3.5-in. diameter slug was
used to change the water level. The slug was designed to displace the
water 5,408 cm , which should result in a 66-cm change in water level in a
4-in. diameter well. The recovery was measured after both inserting and
removing the slug using an electrical tape.
The change in water level over time was then plotted on semi-log
paper and the method described in Bouwer and Rice (1976) was used to
analyze the results.
SLUG TEST RESULTS
A summary of the slug test results is shown in Table 1. Four
zones of permeabilities could be discerned. Conductivities vary between
the shallow and deep portions of the aquifer, and the north and south
portions of the site. The shallow portion of the aquifer in the north of
the site has a conductivity of about 10" cm/sec (3 ft/day), while the deep
_2
portion has a conductivity of about 4 x 10 cm/sec (113 ft/day). The
-------
Boring Locations
N
A
t 13
South 196th Street
./"
Legend
.17 Qround Wซlซf WปIU
80' Monitoring Well
-- 60' Dual Monitoring Wซll
0 150 300
=i =
Scale in Feet
.4
v:!
//4-84-5
.11
.K
* ~^84-4
.24
.25
x-
4-1
.7 x
is
^Fence
/ Line
^84-3
Darrves A Moore
Figure 1. Western Processing Well Locations
-------
opposite is true of the southern portion of the site; the deep portion of
the aquifer has a conductivity of about 3 x 10 cm/sec (8.5 ft/day), while
f)
the .shallow portion is approximately 2 x 10" cm/sec (57 ft/day). The
values for conductivity can be quantitatively verified by comparing the
calculated values with the geology of the screened interval (Table 2). The
2
fine-to-medium sand has a conductivity of 3 x 10 cm/sec (85 ft/day),
while the presents of peat, clay, and/or silt decreased the conductivity to
6 x 10"4 cm/sec (1.7 ft/day).
Table 1. Summary of Slug Test Results
Hydraulic
Conductivity
Well No. (on/sec)*
1A
IB
17A
84-1B
84-2
84-3
84-4A
84-4B
84-5A
84-5B
2 x 10"?
8 x io;J
6 x 10 ;
2 x io;j
9 x 10 I
8 x 10",
2 x 10":
4 x 10";
4 x ID",
5 x 10"
*Average of the results for
dropping and removing the
slug (see Appendix B).
Table 2. Comparison of Hydraulic Conductivity and Lithology
Conductivity
Well No. Depth (cm/sec) Lithology
1A- 15 ft 2 x IQ'l Sand with Silt/Clay Lenses
17A 15 ft 6 x 10"; Silt and Clay
84-4A 25 ft 2 x 10"f Fine-to-Medium Sand
84-5A 25 ft 4 x 10~T Peat/Fine-to-Medium Sand
IB 30 ft 8 x 10"Z Peat/Fine Sand
84-1B 50 ft 2 x ID"); Fine-to-Medium Sand
84-2 50 ft 9 x 10".: Fine Silty Sand
84-3 50 ft 8 x 10"^ Fine-to-Coarse Sand
84-4B 50 ft 4 x 10"; Sandy Silt
84-5B 50 ft 5 x 10 Fine-to-Medium Sand
-------
The values for conductivity used in Battelle's groundwater flow
and contaminant transport model of the Western Processing Site were
1.2 x 10"3 cm/sec (3.5 ft/day) for the upper 30 ft, and 1.8 x 10"2 cm/sec
(50 ft/day) for the portion of the aquifer below 30 ft. The values closely
match the values obtained from the slug tests in the northern portion of
the site, while these values vary an order of magnitude in the southern
portion of the site. Since the majority of the contaminants of interest in
the model were disposed of in the northern portion (Reaction Ponds I and
III) of the site, the difference between the model predicted and measured
hydraulic conductivities will probably have little effect on the overall
transport of contaminants.
BOREHOLE DILUTION TEST PROCEDURES
Borehole dilution testing as described by Freeze and Cherry
(1979) has been used extensively in Europe on a means of determining
groundwater velocity. The theory is that the horizontal velocity of
groundwater through a well-bore can be determined by measuring the change
in concentration over time of a specific ion introduced into the borehole.
This measurement is performed by packing off a known portion of a screen,
introducing a known ion, and measuring the changes in concentration over
time. Figure 2 is a schematic of the apparatus used at the Western
Processing Site.
BOREHOLE DILUTION TEST RESULTS
The borehole dilution tests met with mixed results. Several
equipment problems and difficult field conditions resulted in tests being
performed on only two boreholes. Both Battelle pumps refused to work (both
were lab tested prior to field work), and the conductivity meter would not
function properly. The conductivity meter gave anomolous measurements when
the probe was moved near the lower mixing tube. When the probe was moved
near the upper mixing tube, the anomolous measurements ceased (see
-------
Strip Chart Recorder
T_ Conductivity Meter
=0-
Screen <
Injection Port
Pump
Ground Surface
-*-Well Casing
i-* Packer
Outlet
Conductivity Probe
Inlet
Figure 2. Borehole Dilution Test Apparatus
-------
Figure 2). We obtained a third pump from the EPA Manchester lab which did
work. Further modifications have been made to the test apparatus to
eliminate the.probe problem in the future.
The tests that were performed did show the expected change in
conductance with time. Because of the errors in the initial readings, the
calculated velocities were suspect. The borehole tests indicated that
horizontal groundwater flow near Well 17A was much less than that of Well
IB. Unfortunately, the length of time that was required to complete the
test at Well 17A was longer than the time available on site.
REFERENCES
Bouwer, H., and R. C. Rice. 1976. "A Slug Test for Determining Hydraulic
Conductivity of Unconfined Aquifers with Completely or Partially
Penetrating Wells." Water Resources Research, Vol. 12, No. 3, pp. 423-428.
Freeze, R. A., and J. A. Cherry. 1979. Groundwater. Prentice-Hall, Inc.,
Englewood Cliffs, NJ.
-------
APPENDIX A
SLUG TEST FIELD FORMS
-------
SITE Western Processing
Page 1 of
Date 2/11/85 10:30
SLUG TEST FIELD FORM
Investigator J. Doesburg
C. Eddy
M. Lllga
Borehole I
Radius of borehole 4 in./lO cm
Radius of casing 2 in./5 cm
Depth of well 443 cm
Length of screen 3 ft./91 cm
Reference point for
water level measurments Top of PVC
Casing
Elevation Steel Casing 23.38 ft. AMSL 713 cm
Static water level 250 cm (2/ll/85)-246.5 cm (2/13/85)
Static water level elevation M66 cm
Ground surface elevation _NA
Slug volume 330 cu. in. /5408 cm3
Anticipated displacement 26 in./66 cm
Remarks: * Stop.watch had to be restarted at approximately
30 minutes.
This well was installed with a backhoe (F. Wolf verbal
communication). There is a large gravel pack around
the screen.
Water level in well dropped 21 cm overnight. Reading
8:30 a.m.-2/12/85. Puddles which had been 6" deep in
vicinity were gone in morning.
T
-------
WELL '# 1A
SLUG TEST FIELD FORM
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
TIME
DROPP
0:00
0:05
0:15
0:25
0:35
0:45
0:60
1:15
1:30
1:45
2:00
2:15
2:30
2:45
3:00
3:30
4:00
4:30
5:00
DEPTH
TO
WATER
NG SLUG
250
221
229
232
"234
237
237.5
238.5
239
240
240.5
241
241.5
241.5
242
242.5
242.5
242.5
242.75
STATIC
MINUS
TEST
0
29
21
18
16
13
12.5
11.5
11
10
9.5
9
8.5
8.5
8
7.5
7.5
7.5
7.25
#
20
21
22
23
24
25
26
27
28
29
1
2
3
4
5
6
7
8
9
10
11
TIME
6:00
7:00
9:00
11:00
15:00
20:00
25:00
* SEE
35:00
45:00
2:52:00
REMOVI
0:15
0:30
0:40
0:50
1:00
1:10
1:20
1:30
1:40
1:50
2:05
DEPTH
TO
WATER
242.75
243
243.5
244
244
245
244.5
REMARK
244.5
244.5
250
G SLUI
267
257
252.5
251
249
148
246.5
245.5
244.5
244
242.5
STATIC
MINUS
TEST
7.25
7
6.5
6
6
5
5.5
5.5
5.5
0
17
7
2.5
1
-1
-2
-3.5
-4.5
-5.5
-6
-7.5
.
-------
Paqe _3_ of 5_
WELL # 1A
SLUG TEST FIELD FORM
1
^^^^^M
12
13
i^-
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
TIME
2:15
2:30
2:45
3:00
3:15
3:30
4:00
4:30
5:00
5:30
6:00
7:00
8:00
10:00
12:00
14:00
17:00
22:00
28:00.
47:30
57:00
DEPTH
TO
WATER
242.5
242
241
241
241
241
240
239
238
237.5
236.5
235
234.5
234
234
234
234
233.5
233.5
235
234
STATIC
MINUS
TEST
-7.5
-8
-9
-9
-9
-9
-10
-11
-12
-12.5
-13.5
-15
-15.5
-16
-16
-16
-16
-16.5
-16.5
-15
-16
#
34
35
36
37
TIME
1:10:0
l:29:/)l
1:46:01
.8:26:01
DEPTH
TO
WATER
) 235
) 236
) 236
1 257
STATIC
MINUS
TEST
-15
-14
-14
7
-------
Page 4_ of J_
WELL # 1A
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
. 22
RET
TIME
0:00
0:05
0:16
0:24
0:31
0:39
0:47
0:57
1:10
1:22
1:35
2:00
2:20
2:40
3:00
3:30
4:00
4:30
5:00
5:45
6:30
7:15
ST
^^^^^^^
DEPTH
TO
WATER
246.5
217:5
225
228
230
232
233.5
235
237
238
239
240
241
242
242
242.5
243
243.5
244
244
244.5
245
i -
STATIC
MINUS
TEST
0
29
21.5
18.5
16.5
14.5
13
11.5
9.5
8.5
7.5
6.5
5.5
4.5
4.5
4
3.5
3
2.5
2.5
2
1.8
SLUG TES
.
.
T FIELE
# .
23
24
25
26
27
28
29
30
1
2
3
4
5
6
7
8
9
10
11
12
13
"
) FORM
i . i i
TIME
8:00
9:09
10:00
12:00
14:00
17:00
20:07
3:30:39
SLUG Rl
0:16
0:27
0:40
0:55
1:13
1:30
1:45
2:00
2:20
2:40
_
3:00
3-20
-
3' 40
M
DEPTH
TO
WATER
245
245.2
245.3
245.5
245.8
245.9
245.9
246
MOVED
259.5
259.5
259.5
259
257.5
256.5
256.5
255.5
255
254
253.5
"
OC 0
ฃ30
DCO C
CJฃ . J
STATIC
MINUS
TEST
1.5
1.3
1.2
1.0
.7
.6
.6
0
13
13
13
12.5
11
10
10
9
8.5
7.5
7
6r
. 0
V^BMI^^^^MH^^M
. .
-------
Page 5_ of 5
WELL # 1A
SLUG TEST FIELD FORM
#
14
15
16
17
18
19
20
21
22
23
24
25
26
TIME
4:00
4:30
5:00
5:45
6:30
7:15
8:00
9:00
10:00
11:00
13:40
15:00
17:00
DEPTH
TO
WATER
25.2
251.5
251
250.5
249.5
249
248.5
248
248
247.5
247
247
246.7
STATIC
MINUS
TEST
5.5
5
4.5
4
3
2.5
2
1.5
1.5
1
1
1
.2
#
TIME
DEPTH
TO
WATER
STATIC
MINUS
TEST
-------
page i or
SITE Western Processing
Date 2/11/85 10:00
SLUG TEST FIELD FORM
Investigator J. Doesburq
C. Eddy
M. Lilga
Borehole # IB
Radius of borehole 10 cm
/////////7
Radius of. casing 5 cm
Depth of well 914 cm - -
Length of screen 91 cm
Reference point for
water level measurments Top of PVC
Casing
Elevation steel Casing 788 cm AMSL
Static water level 276 cm
Static water level elevation -^512 cm
Ground surface elevation NA
Slug volume 330 cu. in. 75408 cm3
Anticipated displacement 66 cm
rif ป
*- t.*i
2j?w
1
r
yj
/ ii n
snfnt
-J-
i
L
1
j
Remarks: Cold, wet. Intemittent rain
-------
Page 2_ of 2
WELL # IB
SLUG TEST FIELD FORM
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
TIME
DROPPI
0:00
0:05
0:15
0:25
0:35
0:45
0:55
1:05
1:15
1:25
1:35
1:45
1:55
2:05
2:20
2:40
2:55
3:10
3:25
DEPTH
TO
WATER
NG SLUG
276
217
218
220
222
224
228
228
230
231
233
234
235
237
239
242
243
244
246
STATIC
MINUS
TEST
0
59
58
56
54
52
48
48
46
45
43
42
41
39
37
34
33
32
30
#
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
TIME
3:40
4:00
4:30
5:00
5:30
6:00
7:00
8:00
9:00
10:00
12:00
14:00
16:00
18:00
20:00
22:00
24:00
26:00
28:00
DEPTH
TO
WATER
247.5
249
251
253
254.5
256
259
261.5
264
265
268
269
271
?72.5
273.5
275
275
?75.5
276
STATIC
MINUS
TEST
28.5
27
25
23
21.5
20
17
. 14.5
13
11
8
7
5
3.5
2.5
1
1
.5
0
-
-------
SITE Weytena Processing
Date 2/11/85 09:00
Page 1 of 1
SLUG TEST FIELD FORM
Investigator J. Doesburq
C. Eddv
M. Lilqa
Borehole I HA
Radius of borehole 10 cm
Radius of casing 5 cm
Depth of well 457 cm
Length of screen 91 cm
Reference point for
water level measurments
Casing
Top of PVC
Elevation Steel Casing 767 cm AMSL
Static water level 284 cm
Static water level elevation -M83 cm
Ground surface elevation NA
Slug volume 330 cu. in. /5408 c
Anticipated displacement 66 cm
rrrr,
\snnc
Remarks: Recovery in about 5 seconds.
Maximum displacement about 8 .cm.
Using 8 cm as yo and .1 and 5 sec. at y0and t gives a K of
2X10-1 cm/sec.
Well 11B PVC casing is broken.
-------
SITE Western Processing,
Page 1 of
Date 2/12/85 08:50
Investigator J. Doesburg
C. Eddy
M. Lilga
SLUG TEST FIELD FORM
Borehole #
Radius of borehole 10 cm
Radius of casing 5 cm
Depth of well 524 cm
Length of screen 91 cm
Reference point for
water level measurments Top
Casing-East Side(PVC Cut on Slant)
Elevation Steel Casing 741 cm AMSL
Static water level 196 cm (2/11/85) 205 cm (2/12/85)
Static water level elevation ^536 cm
Ground surface elevation NA
Slug volume 330 cu. in. 75408 cm3
Anticipated displacement 66 cm
Remarks: 2/11/85 Raining
2/12/85 Sunny
Well 17B has screwed joints.
15)
^L
t
*-*-JA
h
T
L
\
!
\
-------
WELL # 17A
SLUG TEST FIELD FORM
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
TIME
DROPP
0:00
0:05
0:15
0:30
0:40
0:45
0:55
1:10
1:25
1:35
1:45
2:00
2:10
- 2:20
2:35
2:50
3:00
3:10
3:30
DEPTH
TO
WATER
NG SLUC
205
142
147
151.5
153.5
156
158
162
164.5
167
168
170
171
173.
176
178
178.5
180
181
STATIC
MINUS
TEST
.
0
63
58
53.5
51.5
49
47
43
41.5
38
37
35
34
32
29
27
26.5
25
24
#
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
TIME
3:40
3:55
4:10
4:25
4:35
4:55
5:15
5:25
5:40
6:00
6:30
7:00
7:30
8:00
8:30
9:00
9:30
10:00
11:00
12:00
13:00
14:45
DEPTH
TO
WATER
182
184
185
186
187
190
191
191.5
192
193
195
196
197.5
198
198.5
200
200.5
201
202
202.5
203
203
STATIC
MINUS
TEST
23
21
20
19
18
15
14
13.5
13
12
10
9
7.5
7
6.5
5
4.5
4
3
2.5
2
2
!
-------
Page 3_ of 4
WELL # 17A
SLUG TEST FIELD FORM
#
42
43
44
45
46
47
48
49
50
1
2
3
4
5
6
7
8
9
10
11
12
TIME
16:00
17:00
18:00
20:00
22:00
25:00
30:00
35:00
51:00
REMOV
0:08
0:25
0:40
0:50
1:00
1:15
1:30
1:45
2:00
2:10
2:30
2:45
DEPTH
TO
WATER
203.5
204
204
204
204
204
204
"204
204
NG SLUG
274
269
267
267
266
266
265
263
262
261.5
260.5
259.5
STATIC
MINUS
TEST
1.5
1
1
1
1
1
1
1
1
69
64
62
62
61
61
60
58
57
56.5
55.5
54.5
#
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
TIME
3:00
3:20
3:50
4:10
4:30
5:00
5:30
6:00
6:30
7:15
8:00
9:00
10:00
11:00
12:00
13:00
14:00
15:30
17:00
18:30
20:00
22:00
DEPTH
TO
259
258.5
256.5
255.5
255
253
252
250.5
249
247
246
243.5
241
240
237
235.5
234
231
229
227.5
226
224
STATIC
MINUS
TEST
54
53.5
51.5
50.5
50
48
47
45.5
44
.42
41
38.5
36
35
32
30.5
29
26
24
22.5
21
19
-------
Page _4_ of 4
WELL I UA
SLUG TEST FIELD FORM
#
35
36
37
38
39
40
41
TIME
24:00
27:00
30:00
35:00
40:00
50:00
1:46:0
Stop '
DEPTH
TO
WATER
221.5
219
216.5
214
212
209
) 205
/atch re
STATIC
MINUS
TEST
16.5
14
11.5
9
7
4
at 11:3
set at t
i actual tim
me 10:53, s
#
t
top wat
TIME
:h time
DEPTH
TO
WATER
1:04
STATIC
MINUS
TEST
r
-------
oITc _U
Page 1 of
nn
SLUG TEST FIELD FORM
Date
Investigator j. Doesburq
r. Eddv
M. Liloa
Borehole # I?B
Radius of borehole IQ cm
Radius of casing 5 cm
Depth of well 914 cm
Length ;6f screen 91 cm
Reference point for
water level measurements Top of PVC
Easing
Elevation Steel Casing 736 cm AMSL
Static water level 235 cm
Static water level elevation -\, 451 cm
Ground surface elevation NA
Slug volume 330 cu. in./54Q8 cm3
Anticipated displacement 66 cm
I
Remarks: Joints are screwed together so slug won't
fit in hole.
-------
SITE Upstern Processing,
Page 1 of
Date y/13/85
Investigator c. Eddv
H. L11g?
Borehole # 84-1-A
SLUG TEST FIELD FORM
Radius of borehole 10 cm
Radius of casing 5 cm
Depth of well 762 cm
Length of screen 152 cm
Reference point for
water level measurements Top of PVC
Casing
Elevation NA
Static water level 302
cm
Static water level elevation
Ground surface elevation N
Slug volume 330 cu. in. /54DR cm3
Anticipated displacement 66 cm
I/////
Remarks: Unable to slug test due to presence of screws in casing
-------
Western Processing
Date 2/13/85
Investigator c. Eddy
M. Lilga
Borehole # 84-1-B
Page 1 of
SLUG TEST FIELD FORM
Radius of borehole 10 cm
Radius of casing 5 cm
Depth of well 1524 cm
Length of screen 152 cm
Reference point for
water level measunnents Top of PVC
Casino
Elevation MA
Static water level 321.5 cm
Static water level elevation NA
Ground surface elevation NA
Slug volume 330 cu. in./54Q8 cm3
Anticipated displacement 66 cm
Remarks:
STATIC. U
-------
WELL I 84-
SLUG TEST FIELD FORM
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1
2
3
4
5
6
TIME
INilJ
0:.00
0:08
0:21
0:30
0:38
0:47
1:05
1:34
2:32
2:49
4:00
5:41
5:54
6:30
8:11
REMOVI
0:11
0:25
0:33
0:37
1:00
1:44
DEPTH
TO
WATER
AL WATE
321.5
347
324
322.5
322
322
322
322
322.5
322.5
322.5
323
323
322.5
322
NG SLUG
322.5
324
323.5
322.5
322.5
322.5
STATIC
MINUS
TEST
0
25.5
2.5
1
.5
.5
.5
1
1
1
1
1.5
1.5
1
.5
.5
2
1.5
.5
.5
.5
DEPTH TO
WATER
INCREASES
#
7
8
9
TIME
2:30
3:27
4:10
RETEST
0:10
0:17
0:24
0:36
1:07
2:07
REMOVI
0:00
0:12
0:24
0:34
0:40
0:53
1:22
2:09
3:57
4:45
DEPTH
TO
WATER
322.5
322.5
322.5
355.5
327
323.5
322
322
322
^G SLU
322
325
324
323
322.5
322.5
322.5
322.5
322.5
322.5
STATIC
MINUS
TEST
.5
.5
.5
13.5
5
1.5
0
0
0
0
3
2
1
.5
.5
.5
.5
.5
.5
-.
.
-------
SITE
Page 1 of
no
SLUG TEST FIELD FORM
Investigator j. Doesburo
Eddv_
M. Liloa
Borehole # 84-2
Dames and Moore wel>sir,ale well, middle of eastern side
of site adjacent to fence.
Radius of borehole IQ cm
Radius of casing 5
cm.
Depth of well 1524 cm
Length of screen 152 cm
Reference point for
water level measurements TOD of PVC
Casino
Elevation
Static water le-vel 266.5
cm.
Static water level elevation
Ground surface elevation i\
Slug volume 330 cu. in. /5408
Anticipated displacement 66 cm
Remarks:
STATIC,
H
-------
WELL * 84-2
SLUG TEST FIELD FORM
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
TIJ1E
INITIA
0:00
0:05
0:15
0:25
0:35
0:45
0:55
1:05
1:15
1:25
1:35
1:45
1:55
2:10
2:20
2:30
2:40
2:50
3:00
3:15
3:30
3:45
DEPTH
TO
WATER
. WATER
266. 5
208
212.5
215.5
218.5
220.5
222.5
"225
227
229
231.5
232.5
235
237
239
240.5
242
242.5
244
245.5
246.5
248.5
STATIC
MINUS
TEST
LEVEL
n
58.5
54
51
48
46
44
41.5
39.5
37.5
35
34
31.5
29.5
27.5
26
24.5
24
22.5
21
20
18
1
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
TIME
4:00
4:15
4:30
4:45
5:00
5:20
5:45
6:05
6:30
7:00
7:30
8:00
9:00
10:00
11:00
12:00
13:00
14:00
16:00
18:00
DEPTH
TO
WATER
249.5
251
252.5
253.5
255.5
255.5
257
257.5
259
260
261
261.5
262.5
263.5
264.5
265
265.5
265.6
266
266.5
STATIC
MINUS
TEST
17
15.5
14
13
11
11
9.5
9
7.5
6.5
5.5
5
4
3
2
1.5
1
.9
.5
0
1
-------
Page _2_ of
WELL # 84-2
SLUG TEST FIELD FORM
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Ifl
19
20
TIME
REMO
0:20
0:45
1:00
1:10
1:20
1:25
1:35
1:40
1:50
2:00
2:10
2:20
2:30
2:55
3:25
3:45
4:00
4:15
4:30
5:00
DEPTH
TO
WATER
flNG SLt
321.5
319
314
312
310
309
308
306
305
303
302
300
298.5
295
291
289.5
?R7.5
?Rfi
285
283
STATIC
MINUS
- TEST
G
55
52.5
47.5
45.5
43.5
42.5
41.5
39.5
38.5
36.5
35.5
33.5
32
28.5
24.5
23
?1.
1Q.R
18.5
16.5
#
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
TIME
. 5-: 30
6:00
6:30
7:00
7:30
8:00
9:00
10:00
11:00
12:00
13:00.
14:00
15:00
16:20
17:20
DEPTH
TO
HATER
281
278.5
277
276.5
275
274
271.5
270.5
269.5
268.5
268
267.5
267
267
267
STATIC
MINUS
TEST
14.5
12
10.5
10
8.5
7.5
5
4
3
2
1.5
1
.5
.5
.5
-------
SITE. uPS|0rn Processing
Date p/17/85
Investigator c. Eddy
M- Liloa
Borehole # 84-3
page 1 of 3
SLUG TEST FIELD FORM
Radius of borehole 10 cm
Radius of casing 5 cm
////////
Depth of well 1524 cm
Length o'f screen 152 cm
Reference point for
water level measurments Top of PVC
Casin
Elevation MA
Static water level 270 cm
Static water level elevation NA
Ground surface elevation NA
Slug volume
330 cu. in. 75408 cm3
Anticipated displacement 66 cm
Remarks:
-T777/
-------
Paqe J_ of J_
WELL # 84-3
SLUG TEST FIELD FORM
#
.
1
2
3
4
5
1
2
3
4
5
6
7
8
9
10
11
1
2
3
4
TIME
"INITI
0:00
0:10
0:30
0:55
1:15
REMOV
0:13
0:25
0:35
0:45
1:00
1:15
1:30
1:45
2:00
2:20
2:40
DROPPI
0:20
0:30
0:40
0:50
DEPTH
TO
WATER
IL WATEf
270
289
270.5
270
270
NG SLUC
288
278
274
273
272
271.5
271
271
270.5
270.5
270
NG SLUG
273
270
-269
268
STATIC
MINUS
TEST
LEVEL
0
1 19
.5
0
0
18
8
4
3
2
1.5
1
1
.5
.5
0
3
0
-1
-2
DEPTH TO
WATER INCRE
DEPTH TO
WATER INCR
#
5
6
USES
7
8
9
10
1
2
3
4
5
6
7
8
9
10
11
12
1
2
TIME
1:00
1:15
1:30
1:45
2:00
2:20
REMOV I F
0:1?
0:24
0:30
0:40
0:55
1:05
1:20
1:40
2:00
2:20
2:40
3:00
DROPPI
0:08
0:21
DEPTH
TO
WATER
269
269
269
269.5
269.5-
270
G SLUC
285
277
275
273
272.5
272
272
271
271
270.5
270.5
270
NG SLU
281
277
STATIC
MINUS
TEST
-1
-1
-1
-.5
-.5
0
15
7
5
3
2.5
2
2
1
1
.5
.5
0
s*
b
11
7
DEPTH TO
WATER INCR.
-------
WELL #
SLUG TEST FIELD FORM
#
3
4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
TIME
0:29
0:39
REMOV
0:00
0:18
0:28
0:34
0:41
0:48
0:58
1:10
1:20
1:30
1:40
1:50
2:00
2:20
2:40
3:00
3:30
4:00
DEPTH
TO
WATER
271.5
270
[NG SLU(
270
276.5
274.5
274
273
273
272.5
271.5
271
271
272
271.5
271.5
271
271
271
270,5
270.5
STATIC
MINUS
TEST
1.5
0
0
6.5
4.5
4
3
3
2.5
1.5
1
1
2
1.5
1.5
1
1
1
.5
.5
-
1
TIME
DEPTH
TO
WATER
STATIC
MINUS
TEST
-------
SITE Western Processing
Page 1 of
Date 2/12/85
Investigator c-
M. Lilga
Borehole # 84~4A
SLUG TEST FIELD FORM
Radius of borehole 10 cm
Radius of casing 5 cm
Depth of well 762 cm
Length of screen 152 cm
Reference point for
water level measurments TฐP of pvc
Casing
Elevation
NA
Static water level 268 cm
Static water level elevation
Ground surface elevation "A
Slug volume 330 cu. in. /5408 cm3
Anticipated displacement 66 cm
Remarks:
z*
t
H
-------
Paqe _2_ of J_
WELL # 84-4A
SLUG TEST FIELD FORM
#
1
2
3
4
5
6
7
R
9
10
11
12
13
14
1
2
3
4
5
6
7
TIKE
INITIA
0:00
0:10
0:20
0:35
1:00
1:15
1:35
1 !45
1:55
2:05
2:15
2:30
2:45
3:30
REMOV
0:22
0:35
0:46
0:55
1:05
1:15
1:25
DEPTH
TO
WATER
. WATER
268
255
267
268
266
266
266.5
7fi7
267
267
267
267
267
267
NG SLUG
287
277.5
274
273
271.5
270.5
270
STATIC
MINUS
TEST
LEVEL
0
13
1
0
2
2
1.5
1
1
1
1
1
1
1
19
9.5
6
5
3.5
2.5
2
#
8
9
10
11
12
13
Id
15
16
1
2
3
4
5
6
7
8
9
10
11
TIME
1:35
1:50
2:00
2:15
2:30
3 .- nn
v^n
4:15
5:00
DROPPI1
0:30
0:40
0:50
1:00
1:10
1:20
1:40
1:50
2:05
2:25
3:00
DEPTH
TO
WATER
269.5
269.5
269
269
269
?sq
?fiQ
268.5
268
G SLU(
263.5
263
264.5
265
265.5
265.5
266
267
267.5
267.5
267
STATIC
MINUS
TEST
1.5
1.5
1
1
1
1
i
.5
0
4.5
5
3.5
3
2.5
2.5
2
1
.5
.5
1
-------
Pa.qe J_ of 3
WELL # 84-4A
SLUG TEST FIELD FORM
1
12
.13
1
2
3
4
5
6
TIME
3:30
4:00
REMOVI
0:00
1:05
1:20
1:45
2:00
2:15
2:35
DEPTH
TO
WATER
267.5
268
NG SLUG
268
274
269
269.5
269
269
268
STATIC
MINUS
TEST
.5
0
0
6
1
1.5
1
1
0
#
-
TIME
DEPTH
TO
WATER
STATIC
MINUS
TEST
f
-------
SITE Western Processing
Page 1 of _4_
Date 2/12/85
SLUG TEST FIELD FORM
Investigator c. Eddv
M Lilaa
Borehole # 84-4B
Radius of borehole 10 cm
Radius of casing 5 cm
Depth of well 1524 cm
Length "of screen 152 cm
Reference point for
water level measurements
Top of PVC
Casing
Elevation MA
Static water level 269 cm
Static water level elevation NA
Ground surface elevation NA
Slug volume 330 cu. in./54Q8 cm3
Anticipated displacement 66 cm
17777
Remarks: "Fizzing" noise noted in well during first part
of slug test.
-------
Page _ฃ_ of 4
WELL # 84-4B
SLUG TEST FIELD FORM
1
.^^ ^
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
, Ifl
19
20
2J
22
TIME
TMTIA
0:00
0:08
0:20
0:35
0:45
0:55
.1:10
1:25
1:40
1:50
2:00
2:15
2:30
2:45
3:00
3:30
3:50
4:15
4:40
5:00
5:20
5:40
DEPTH
TO
WATER
. WATER
269
205
208
209
210
211
2,12
212
213
213.5
213.5
214.5
215
215
216.5
217.5
218.5
220
221
221.5
222
223
STATIC
MINUS
TEST
LEVEL
0
64
61
60
59
58
57
57
56
55.5
55.5
54.5
54
54
52.5
51.8
50.5
49
48
47.5
47
46
#
23
24
25
26
27
28
29
30
31
32
33
34
35
36
.37
38
39
40
41
42
43
44
TIME
6:00
6:20
6:40
7:00
7-: 30
8:00
8:35
9:00
9:45
10:30
11:15
12:00
13:00
14:10
15:15
16:30
18:00
19:30
21:00
23:00
25:00
27:45
DEPTH
TO
WATER
224
225
225.5
226
227
228.5
228.5
230
231
232
233
234.5
236
237.5
239
240.5
242.5
245
246
248
249
251
STATIC
MINUS
TEST
45
44
43.5
43
42
40.5
40,5
39
38
37
36
34.5
33
31.5
30
28.5
26.5
24
23
21
.20
18
-------
Paqe JL of A
WELL # 84.aa
SLUG TEST FIELD FORM
#
45
46
47
48
49
50
51
52
53
54
55
-55
57
58
59
60
61
62
63
64
65
66
TII1E
30:05
33:00
36:00
40:00
45:00
50:00
1:00:0
1:10:0
1:20:0
REMOVI
0:00
0:10
0:15
0:30
0:40
0:50
1:00
1:10
1:20
1:35
.1:45
2:00
2:20
DEPTH
TO
WATER
252.5
255
256
258
260
261.5
] 264
) 266.5
3 268
JG SLUG
269
3?7
328
326
324
323
321.5
320
319.5
317
316
314.5
310.5
STATIC
MINUS
TEST
16.5
14
13
11
9
7.5
5
2.5
1
0
58
59
57
55
54
52.5
51
5C.5
48
47
45.5
41.5
#
67
68
69
70
71
72
73
74
75
7&
77
78
79
80
81
82
83
84
85
86
87
88
TIME
2:35
2:50
3:00
3:15
3:30
3:45
4:00
4:20
4:40
5:00
5:30
6:00
6:30
7:00
8:00
9:00
10:00
11:00
12:00
14:00
16:00
18:00
DEPTH
TO
WATER
310
308.5
307.5
306
304
303
301.5
300
298
296
294
292
290
289
285.5
282
280
278
277
274.5
272.5
271.5
STATIC
MINUS
TEST
41
39.5
38.5
37
35
34
32.5
31
29
27
25
23
21
20
16.5
13
11
9
8
5.5
3.5
2.5
-------
Page 4_ of 4
WELL # 84-4B
SLUG TEST FIELD FORM
#
89
90
91
92
93
94?
TIME
20:00
22:00
25:20
28:00
32:00
33 mir
1:10:0(
DEPTH
TO
WATER
270
270
270
270
270
utes is
271,
1
STATIC
MINUS
TEST
1
1
1
1
1
3:49 p.n
at 4:26
p.m.
#
TIME
DEPTH
TO
WATER
STATIC
MINUS
TEST
. ___
-------
Page 1 of 3
SITE Upstprn Processing
Date ?/i?/85
Investigator
SLUG TEST FIELD FORM
C. Eddy
M. Lilga
Borehole # 84-5A
Radius of borehole 10 cm
Radius of casing 5 cm
Depth of well 762 cm
Length of screen 152 cm
Reference point for
water level measurements Top of PVC
Casing
Elevation
NA
Static water level 241 cm
Static water level elevation NA
Ground surface elevation NA
Slug volume 330 c'u. in. /54Q8 cm3
/
Anticipated displacement 66 cm
Remarks:
-------
Page _ฃ_ of J_
WELL # 84-5A
SLUG TEST FIELD FORM
#
1
2
3
4
5
6
7
8
9
10
11
12
1-3
14
15
16
17
18
19
20
21
22
~-
TIME
INIT:
0:00
0:12
0:27
0:39
0:51
1:05
1:15
1:25
1:35
1:45
2:00
2:15
2:30
2:45
3:00
3:20
3:40
4:00
4:30
5:00
5:30
6:00
DEPTH
TO
WATER
AL WATE
241
201
197
202
202.5
204
204.5
205
206
206.5
207
208
209
210
211
212.5
213
214
215
216
217
218
STATIC
MINUS
TEST
R LEVEL
0
40
44
39
38.5
37
36.5
36
35
34.5
34
33
32
31
30
.28.5
28
27
26
25
24
23
#
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
1
2
3
4
TIME
6:45
7:30.
8:15
9:00
10:00
11:00
12:30
14:00
15:30
17:00
18:30
20:00
22:00
24:00
27:00
30:00
35:00
REMOVI
0:05
0:14
0:22
033
DEPTH
TO
WATER
220.5
221
222.5
223
225
227
228
229.5
231.5
232
233
234
236
237.5
238
239
241
NG SLU
303
301
300
297
STATIC
MINUS
TEST
20.5
20
18.5
18
16
14
13
11.5
9.5
9
8
7
5
3.5
3
2
0
3
62
60
59
54
-------
Page _1 of f_
WELL # 84-5A
SLUG TEST FIELD FORM
#
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
TIKE
0:44
0:50
1:10
1:20
1:40
1:50
2:05
2:15
2:30
2:45
3:00
3:20
3:40
4:00
4:30
5:00
5:30
6:00
6:45
7:30
8:15
9:02
DEPTH
TO
WATER
295
294
292
290
288
286
285
284
282
280.5
279
277
275
273.5
271.5
269.5
267.5
265.5
262.5
260
259
257
STATIC
MINUS
TEST
54
53
51
49
47
45
44
43
41
39.5
38
36
34
32.5
30.5
28.5
26.5
24.5
21.5
19
18
16
#
27
28
29
30
31
32
33
34
35
37
38
TIME
10:00
11:00
12:00
13:30
15:00
17:00
19:00
22:00
25:00
29:00
33:07
DEPTH
TO
WATER
254
253
251.5
249.5
247
246
245
244
243
242
241
STATIC
MINUS
TEST
13
12
10.5
8.5
6
5
4
3
2
1
0
-------
SITE intern Processing
Page 1 of _2_
Date 2/13/85
SLUG TEST FIELD FORM
Investigator C. Eddy
M. Lilga
Borehole # 84-5B
Radius of borehole 10 cm
Radius of casing 5 cm
Depth of well 1524 cm
Length of screen 152 cm
Reference point for
water level measurments Top of PVC
Casing
Elevation NA
Static water level 259.5 cm
Static water level elevation NA
Ground surface elevation NA
Slug volume 330 cu. in. /54Q8 cm
Anticipated displacement 66 cm
Remarks:
I
It
H
-------
Page _ฃ of 2
WELL # 84-5B
SLUG TEST FIELD FORM
#
1
2
3
4
5
6
1
2
3
1
2
3
4
5
1
2
3
4
5
TIME
WATER
0:00
0:16
0:25
0:35
0:49
l:-5
REMOV
0:OS
0:20
0:50
DROPP
0:10
0:18
0:30
0:45
1:01
REMOV If
0:12
0:19
0:25
0:35
1:10
DEPTH
TO
WATER
LEVEL ]
259.5
263.5
259.5
259
259
259.5
NG SLUC
" 259
259.5
259.5
NG SLUG
306
262
259.5
259.5
259.5
!G SLUG
261
260
259.5
259.5
259.5
STATIC
MINUS
TEST
NCREASES
0
4
-.5
-.5
0
.5
0
0
464.5
3.5
0
0
0
1.5
.5
0
0
0
#
1
2
3
4
5 .
6
1
2
3
4
WATER LEVEL
INCREASES 5
6
7
TIME
DROPPII
0:02
0:12
0:19
0:27
0:33
0:41
REMOV If
0:12
0:22
0:30
0:36
0:45
0:53
1:02
DEPTH
TO
WATER
G SLU<
256
277
261
260
259.5
259.5
G SLU(
260
259.5
259.5
259.5
259.5
259.5
259.5
STATIC
MINUS
TEST
3.5
17.5
1.5
.5
0
0
.5
0
0
0
0
0
0
"
-------
APPENDIX B
PLOTS USED TO CALCULATE HYDRAULIC
CONDUCTIVITY FROM SLUG TESTS
-------
For the EPA Wells:
Length of Screen = 91 cm
Borehole Radius = 10 cm
Well Radius = 5 cm
A* = 1.8
B* = 0.2
For the Dames and Moore Wells:
Length of Screen = 152 cm
Borehole Radius = 10 cm
-Well Radius = 5 cm
A* = 2.0
B* = 0.2
From Bouwer and Rice (1976).
-------
WELL Ifl DROPPING SLUG (1)
Hydraulic Conductivity
Height of Water Column
YO
Y
t
3 x 10
181 cm
29 cm
1 cm
205 sec
0.00
150.88
301.75
452.63
TIME IN
603.50
SECONDS
754.38
905.25
1055.13
-------
WELL Ifl REMOVING SLUG (1)
10"4 cm/sec
Hydraulic Conductivity
Height of Water Column
YO
Y
t
8 x
181 cm
15 cm
5 cm
255 sec
127.5
255.0
382.5 510.0
TIME IN SECONDS
637.5
755.0
822.5
-------
WELL Ifi DROPPING SLUG (2)
Hydraulic
Height of
Conductivity
Water Column
YO
Y
t
= 1 x 10" 3
= 181 cm
= 21 cm
= 1 cm
= 469 sec
cm/ sec
337.5
675.0
1012.5
TIME IN
1350.0
SECONDS
1687.5
2025.0
2362.5
-------
WELL IB DROPPING SLUG (1)
Hydraulic Conductivity
Height of Water Column
YO
Y
t
8 x 10
699 cm
59 cm
10 cm
532 sec
217.5
435.0
652.5
870.0
1087.5
1305.0
1522.5
TIME IN SECONDS
-------
WELL 17R REMOVING SLUG (1)
Hydraulic Conductivity = 2 x 10
Height of Water Column = 313 cm
YO = 65 cm
Y = 30 cm
t = 736 sec
-4
cm/sec
o_
CO
or:
LJ
CO
o
0.0
375.0
750.0
1125.0
TIME IN
15CO.O
SECONDS
1875.0
zso.o
-------
WELL 17R DROPPING SLUG (1)
Hydraulic Conductivity
Height of Water Column
YO
Y
t
1 x 10
313 cm
60 cm
2 cm
723 sec
382.5
765.0
1147.5 1530.0
TIME IN SECONDS
1912.5
2225.0
2S77.5
-------
WELL 84-IB DROPPING SLUG (1)
Hydraulic Conductivity
Height of Water Column
YO
Y
t
3 x
1,263 cm
26 cm
1 cm
24 cm
_2
10 cm/sec
0.000
61.375
122.750 184.125 245.500
TIME IN SECONDS
306.875
368.250
429.625
-------
WELL 84-IB DROPPING SLUG (2J
o_
LJ
LJ
O
Hydraulic Conductivity
Height of Water Column
YO
Y
t
2 x
1,263 cm
13.5 cm
1.0 cm
21 sec
_2
10 cm/sec
0.000
15.875
31.750 47.625 63.500
TIME IN SECONDS
79.375
95.250
111.125
-------
WELL 84-IB REMOVING SLUG (2)
Hydraulic Conductivity
Height of Water Column
YO
Y
t
1 x
1,263 cm
3.0 cm
0.5 cm
35 sec
_2
10 cm/sec
o_
CO
a:
DJ
cu
o
o.
-a
0.000
35.625
71.250
108.875
142.500
178.125
213.750
249.375
TIME IN SECONDS
-------
WELL 84-2 DROPPING SLUG (1)
Hydraulic Conductivity
Height of Water Column
YO
Y
t
o_
CO
a:
LJ
t->
LJ
LJ
o
1 X
1,313 cm
57 cm
2 cm
623 sec
10"3 cm/sec
0.0 1:5.0 270.0 405.0 540.0
TIME IN SECONDS
675.0
810.0
945.0
-------
WELL 84-2 REMOVING SLUG (1)
-4
10 cm/sec
Hydraulic Conductivity
Height of Water Column
YO
Y
t
8 x
1,318 cm
55 cm
10 cm
386 sec
130.0
280.0
390.0
520.0
650.0
780.0
910.0
TIME IN SECONDS
-------
WELL 84-3 REMOVING SLUG (1)
Hydraulic Conductivity
Height of Water Column
YO
Y
t
1 x 10 cm/sec
1,315 on
18 cm
1 cm
53 cm
o_
CO
LJ
LJ
O
0.0
20.0
40.0
60.0 80.0
TIME IN SECONDS
100.0
120.0
110.0
-------
WELL 84-3 REMOVING SLUG (2)
10 cm/sec
Hydraulic Conductivity
Height of Water Column
YO
Y
t
9 x
1,315 cm
16 cm
1 cm
60 sec
22.5
45.0
67.5
90.0
112.5
135.0
157.5
TIME IN SECONDS
-------
WELL 84-3 REMOVING SLUG (3)
10 cm/sec
Hydraulic Conductivity
Height of Water Column
YO
Y
t
4 x
1,315 cm
6.5 -cm
1.0 cm
93 sec
30.0
60.0
90.0 120.0
TIME IN SECONDS
150.0
180.0
210.0
-------
NELL 84-4R DROPPING SLUG (1)
Hydraulic Conductivity = 4 x 10
Height of Water Column = 555 cm
YO = 13 cm
Y = 0.1 cm
t = 20 sec
-2
cm/sec
0.00
25.25
52.50 78.75 105.00
TIME IN SECONDS
131.25
157.50
153.75
-------
WELL 84-4R REMOVING SLUG (1)
Hydraulic Conductivity = 5 x
Height of Water Column = 555 cm
YO = 19 cm
Y = 1.0 cm
t = 98 sec
10"3 cm/sec
37.5
75.0
112.5
TIME IN
150.0
SECONDS
187.5
225.0
2S2.5
-------
WELL 84-4B ~ DROPPING SLUG (1)
-4
10 cm/sec
Hydraulic Conductivity
Height of Water Column
YO
Y
t
2 x
1,316 cm
60 cm
10 cm
2,200 sec
600.0
1230.0
1800.0 2400.0
TIME IN SECONDS
3000.0
3600.0
4200.0
-------
WELL 84-4B REMOVING SLUG (1)
Hydraulic Conductivity
Height of Water Column
YO
Y
t
5 x
1,316 cm
60 cm
10 cm
613 sec
-4
10 cm/sec
to
en
LJ
E-H
LJ
OJ
o
0.0 525.0 1050.0 1575.0 21X.O
TIME IN SECONDS
2625.0
3150.0
3675.0
-------
NELL 84-Sfi DROPPING SLUG (.11
Hydraulic Conductivity- 3 x 10
Height of Water Column = 582 cm
YO = 40 cm
Y = 10 cm
t = 802 sec
cm/sec
CO
<ฃ
CJ
Hi
CJ
"0.0 262.5 525.0 787.5 1050.0
TIME IN SECONDS
1312.5
1575.0
1837.5
-------
WELL 84-5R REMOVING SLUG (1)
-4
10 cm/sec
Hydraulic Conductivity
Height of Water Column
YO
Y
t
5 x
582 cm
60 cm
3 cm
1,086 sec
0.0
247.5
495.0
742.5 890.0
TIME IN SECONDS
1237.5
1485.0
-------
WELL 84-5B REMOVING SLUG (1)
-2
Hydraulic Conductivity = 3 x 10" c cm/sec
Height of .Water Column = 1,325 cm
YO = 0.5 cm
Y = 0.1 cm
t = 9 sec
(O J
a:
CJ
ฃ-
CiJ
CiJ
O
>
\
-B-
0.00 6.25
12.50
18.75
TIME IN
25.00
SECONDS
31.25
37.50
43.75
-------
WELL 84-5B DROPPING SLUG (2)
Hydraulic Conductivity
Height of Water Column
YO
Y
t
_2
6 x 10 cm/sec
1,325 cm
47 cm
0.1 cm
18 sec
o_
CO
s
E-ซ
O
0.000
7.625
15.250
22.875
30.500
38.125
45.750
53.375
TIME IN SECONDS
-------
APPENDIX D
CALCULATION OF LEACH DURATION AND AMOUNT FROM SOURCE AREAS
Three sources of TCE have been identified at the Western Processing Site
based on the concentration levels measured in the soil and water. These
three sources are: Reaction Pond I; Reaction Pond III; and the area around
Well 21. A known disposal site is not located in the vacinity of Well 21,
but the high concentrations of TCE in soil and water samples from Well 21
indicate that disposal occurred in this area. Calculations were made to
estimate the mass of TCE present in the unsaturated zone (both in soil and
water) of these three locations based on 1982 soil concentration measurements
(EPA, 198.2). These mass estimates were then used to estimate the time
required for TCE to completely leach out of the unsaturated zone into the
saturated zone. Time of leach calculations were based on the average annual
recharge at the site (8 in./yr - Appendix A). The results of these
caluclations were used to determine the number of years past 1982 to keep the
sources active in the model, and to estimate the source strength for each
time step simulated.
SOURCE AREA DATA
The data used to estimate the mass of TCE in the unsaturated zone of
each source area are summarized below.
Reaction Pond I
The total volume of contaminated unsaturated soil at Reaction Pond I was
estimated to be 46,250 ft . This number is based on a surface area of 9,250
o
ft and an average depth of unsaturated contaminated soil of 5 ft. The
surface area is the actual area associated with Reaction Pond I. The average
depth of contaminated soil was estimated from the soil sampling-depths and
concentrations reported in Table D-l. The TCE concentration in the soil was
estimated to be 600,000 ppb based on the measured soil concentrations from
Wells 15 and 17 (Table D-l). In order to represent a worst case, the
-------
concentration was estimated to be about the maximum observed soil
concentration and was assumed to be uniform over the area of the disposal
site.
Table D-l. TCE Soil Concentrations Near Reaction Pond I
Soil Concentration (ppb)
Depth (ft)
3
6
9
Well 15
U
580,000
180,000
Well 17
U
558,000
350,000
U = Not Detected
Reaction Pond III
The total volume of contaminated unsaturated soil at Reaction Pond III
was estimated to be 35,650 ft . This number is based on a surface area of
2
7,125 ft_.; and an average depth of contaminated soil of 5 ft. The surface
area is the actual area associated with Reaction Pond III. The average depth
of contaminated soil was estimated from the soil sampling depths and
concentrations reported in Table D-2. The TCE concentration in the soil was
estimated to be 700 ppb based on the measured soil concentrations from
Well 20 (Table D-2). In order to represent a worst case, the concentration
was estimated to be about the maximum observed soil concentration and was
assumed to be uniform over the area of the disposal site.
Table D-2. TCE Soil Concentrations Near Reaction Pond III
Soil Concentration (ppb)
Depth (ft) Well 20
3 U
6 M
9 676
12 544
U = Not Detected
M = Present but below minimum quantifiable limit.
-------
Area Around Well 21
The total volume of contaminated unsaturated soil in the area around
Well 21 was difficult to estimate because this area is not a known disposal
area. Therefore, the volume was assumed to be the same as that of Reaction
Pond III, 35,650 ft^. This volume is based on the same surface area and
average depth as those used for Reaction Pond III. The TCE concentration in
the soil was estimated to be 1,600 ppb based on the measured soil
concentration from Well 21 (Table D-3).
Table D-3. TCE Soil Concentrations Near Well 21 Source
Soil Concentration (ppb)
Depth (ft) Well 21
3 U
6 U
9 116
12 1,520
U = Not Detected
CALCULATION OF SOURCE AREA MASS
The calculations of mass in the unsaturated zone for each of the three
source locations are presented below.
For all source areas, the actual and effective porosities of the
unsaturated material were set at 0.25 and 0.15, respectively. The
distribution coefficient (Kd) and retardation factor for sorption of TCE on
the unsaturated soils at all source locations were set at 0.2 and 4.0,
respectively.
The equation used to calculate the mass of TCE at each source location
is:
MassT-E = {mass in soil} + {mass in water}
= (1 - 9A)(Cs)(Vols)(Ys) + (9E)(Cw)(Vols)(Yw) (D-l)
Where:
-------
MassTCE = total mass of TCE in the soil (Ib)
QA = actual water porosity
% = effective porosity
Cs = TCE concentration (ppb) in the soil
Vols = total volume of unsaturated soil (ft )
Ys = density of the soil (lb/ft3)
Cw = TCE concentration (ppb) in the water {Cw = -^ (C$)}
Yw = density of the water (lb/ft3)
The density of the soil was calculated as:
Ys = (1 - GA)(Y) = 121.5 lb/ft3
where:
0A = actual porosity (0.25)
Y = dry density of sand (162 lb/ft3)
The TCE soil concentrations were determined based on measured soil
concentrations in the unsaturated zone at wells around the source areas. The
TCE water concentrations were estimated to be five times greater than the
soil concentrations based on an distribution coefficient (Kd) of 0.2. The
Q
ppb notation is equivalent to lb/10 Ib in the total mass calculation.
Reaction Pond I
Based on the data in Table D-4, the total mass of TCE in the unsaturated
zone at Reaction Pond I was calculated to be 3,830 Ib. The total mass in the
soil and water was calculated to be 2,530 Ib and 1,300 Ib, respectively.
-------
Table D-4. Summary of Data Used to Make Total Mass Calculations
at all Three Source Locations
Reaction Pond I Reaction Pond III Well 21
0E 0.15 0.15 0.15
0A 0.25 0.25 0.25
Cs (ppb) 600,000 700 1,600
Cw (ppb) 3,000,000 3,500 8,000
Vo1s (ft3) 46,250 35,650 35,650
Ys (lb/ft3) 121.5 121.5 121.5
Yw (lb/ft3) 62.4 62.5 121.5
Kd 0.2 0.2 0.2
Reaction Pond III
Based on the data in Table D-4, the total mass of TCE in the unsaturated
zone at Reaction Pond III was calculated to be 3.5 Ib. The total mass in the
soil and_water were calculated to be 2.3 Ib and 1.2 Ib, respectively.
Around Well 21
Based on the data in Table D-4, the total mass of TCE in the unsaturated
zone around Well 21 was calculated to be 7.9 Ib. The total mass in the soil
and water were calculated to be 5.2 Ib and 2.7 Ib, respectively.
The estimated total mass of TCE present in the unsaturated zone (based
on October 1982 soil analyses) at the three suspected source areas is
summarized in Table D-5.
Table D-5. Summary of Estimated Total Mass of TCE Present
at the Three Suspected Source Areas
Mass of TCE (Ib)
In Soil In Water Total
Reaction Pond I 2,530 1,300 3,830
Reaction Pond III 2.3 1.2 3.5
Around Well 21 5.2 2.7 7.J
-------
LEACH DURATION AND AMOUNT FROM SOURCE LOCATIONS
The preceeding calculations show that Reaction Pond I is the only source
location which contains a significant quantity of TCE in the unsaturated
zone. Therefore, calculations to determine the time required for TCE to
leach out of the unsaturated zone (duration), and the amount that would leach
out each year were only made for Reaction Pond I. This information was used
in the modeling to determine the length of time into the future to keep
Reaction Pond I active, and the quantity of TCE to leach from the unsaturated
to the saturated zone at each time step.
Reaction Pond I
The leach duration and amount calculations were based on the
concentration of TCE in the water of the unsaturated zone and the rate of
water movement through the unsaturated zone from annual recharge. Equation
D-l can be used to calculate the TCE concentration in the water (C ) based on
w
the total mass of TCE remaining in the system at any given point in time.
For convenience, the equation can be rewritten as follows:
MassTCE
w Vols {Kd (1 - QA) Ys + % Ywl (D-2)
The volume of water (Vol ) moving through the unsaturated zone from the
W
average annual recharge was calculated as follows:
Vol = Rhg x A = 6,167 ft3/yr
W
where
Rhg = average annual recharge (0.67 ft/yr - see Appendix A)
A = area of Reaction Pond I (9,250 ft2)
This number was converted to a mass of water (Massw) passing through the
unsaturated system so the units are consistent.
Mass = Vol x v = 384,800 Ib/yr
" W *W
------- |