DRC-2010-002786 - 0901a0688017ab89

DENrsoNDddfiillNt$
ilR C r 20 1q;fi,fi,47*8 S
l0$0 l?tt $trpoq $uils g$l
Oenver, CO 80283
u$A
Tel : 303 82&t?98
Fax : 303 389*fl2!
wsr.dcnlronmf noo"com
VIA HAND DELIVERY
March 31, 2010
Mr. Dane L. Finerfrock
Executive Secretary
Utah Radiation Control Board
State of Utah Department of Environmental Quality
168 North 1950 West
Salt Lake city, UT 84114-4950
Dear Mr. Finerfrock
Re: White Mesa Uranium Mill State of Utah Groundwater Discharge Permit No. UGW370004-
Revised Infiltration and Contaminant Transport Modeting Report
Pursuant to Part I.H. I I of the White Mesa Mill's Groundwater Discharge Permit, please find enclosed two copies of the
Revised Infiltration and Contaminant Transport Modeling Report prepared by MWH Americas Inc.
Please contact me at 303-3894132 or Harold Roberts at 303-389-4160 if you have any questions or require any further
information.
Yours very ffily,
DnxrsoN MINES (USA) CoRp.
W
Jo Ann Tischler
Director, Compliance and Permitting
Encl.
cc: Ron F. Hochstein
Harold R. Roberts
David C. Frydenlund
David E. Turk
Denison Mines (USA) Corp.
Revised Infiltration and Contaminant Transport
Modeling Report, White Mesa Mill Site,
Blanding, Utah
March 2010

REVISED
INFILTRATION AND CONTAMINANT TRANSPORT
MODELING REPORT
WHITE MESA MILL SITE
BLANDING, UTAH

DENISON MINES (USA) CORP.

March 2010

Prepared for:

Denison Mines (USA) Corp.
1050 17th Street, Suite 950
Denver, Colorado
80265

Prepared by:

MWH Americas, Inc.
10619 South Jordan Gateway, Suite 100
Salt Lake City, Utah
84095

i
TABLE OF CONTENTS

SECTION PAGE

EXECUTIVE SUMMARY ES-1

1.0 INTRODUCTION 1-1
1.1 Objectives of Infiltration and Contaminant Transport Model 1-1
1.2 Permit Requirements 1-1
1.3 General Modeling Approach to Address Permit Requirements 1-4 1.4 Infiltration and Contaminant Transport Model Report History 1-5 1.5 Document Organization 1-6

2.0 BACKGROUND 2-1
2.1 Site Overview 2-3 2.1.1 Facility Description 2-3 2.1.2 Tailings Cover Design 2-5
2.1.3 Tailing Cell Liner Systems 2-6
2.1.4 Characteristics of Tailings 2-8
2.2 Site Characteristics 2-9 2.2.1 Climate 2-9 2.2.2 Summary of Site Geology 2-10
2.2.3 Hydrogeology of the Perched Aquifer System 2-11
2.2.4 Groundwater Quality of the Perched Aquifer System 2-12
2.2.5 Vadose Zone Hydrogeology and Geochmistry of the Unsaturated Bedrock 2-13 2.3 Conceptual Model of Water Flow (Infiltration) Through the Tailings
Cell Cover and Potential Contaminant Transport Through the
Vadose Zone 2-14
2.3.1 Unsaturated Flow 2-14 2.3.2 Contaminant Transport in the Unsaturated Zone 2-18

3.0 METHODOLOGY 3-1
3.1 Overall Modeling Approach 3-1
3.2 HYDRUS-ID and HP1 Computer Codes 3-2 3.2.1 HYDRUS-ID 3-3
3.2.2 HP1 3-4
3.3 Infiltration Model of Tailings Cell Cover 3-5
3.3.1 Domain 3-5
3.3.2 Finite Element Node Spacing 3-5 3.3.3 Boundary Conditions 3-5
3.3.4 Input Parameters 3-8
3.3.5 Initial Conditions 3-9
3.3.6 Duration of Simulations and Time Steps 3-9
3.3.7 Sensitivity Analysis 3-9

ii
TABLE OF CONTENTS
(continued)

SECTION PAGE
3.4 Flow and Contaminant Transport Model of the Bedrock Vadose Zone 3-10 3.4.1 Domain 3-10
3.4.2 Finite Element Node Spacing 3-11
3.4.3 Boundary Conditions 3-11
3.4.4 Input Parameters 3-12 3.4.5 Initial Conditions 3-15 3.4.6 Duration of Simulations and Time Steps 3-15
3.4.7 Sensitivity Analysis 3-16

4.0 RESULTS 4-1
4.1 Tailings Cell Cover System Infiltration Modeling 4-1 4.1.1 Model-Predicted Water Flux Rate 4-2
4.1.2 Evaluation of Build-up of Water in Tailings 4-3
4.1.3 Sensitivity Analysis 4-4
4.2 Bedrock Vadose Zone Flow and Contaminant Transport Modeling 4-5 4.2.1 Cells 2 & 3 4-6 4.2.2 Cells 4A & 4B 4-10
4.2.3 Cell 1 4-11
4.2.4 Evaluation of Closed-Cell Cover System Performance
(Potential Impacts to Groundwater) 4-12 4.2.5 Sensitivity Analysis 4-13 4.3 Uncertainty and Assumptions 4-16

5.0 CONCLUSIONS AND POST-AUDIT MONITORING PLAN 5-1
5.1 Conclusions 5-1 5.1.1 Model-Predicted Water Flux Rate for Tailings Cell Cover
System 5-1
5.1.2 Evaluation of Build-up of Water in Tailings 5-2
5.1.3 Bedrock Vadose Zone Flow and Contaminant Transport
Modeling 5-3 5.1.4 Summary of Closed Cell Cover System Performance 5-7
5.2 Post-Audit Monitoring Plan 5-8

REFERENCES R-1

iii
LIST OF APPENDICES
• Appendix A – Laboratory reports with results of vadose zone mineralogical testing
and properties of stockpiled soil.
• Appendix B – Laboratory report with unsaturated and saturated hydraulic properties of the bedrock core samples.
• Appendix C – Bedrock sampling to characterize hydraulic and geochemical
properties of the vadose zone
• Appendix D – Vegetation evaluation for the evapotranspiration cover.
• Appendix E – Comparison of cover designs based on infiltration modeling.
• Appendix F – Evaluation of the effects of storm intensity on infiltration through evapotranspiration cover.
• Appendix G – Sensitivity analysis comparing infiltration rates through the
evapotranspiration cover based on cover vegetation, biointrusion, and precipitation.
• Appendix H – Radon emanation modeling for the evapotranspiration cover.
• Appendix I – Tailings hydraulic conductivity evaluation.
• Appendix J – Tailings cell dewatering modeling.
• Appendix K – Statistical evaluation of tailings pore water chemistry and
identification of source term concentrations.
• Appendix L – Evaluation of potential water flow through the tailings cell liners.
• Appendix M – Geochemical model and reactive transport modeling of flow and transport through the vadose zone.
• Appendix N – Predictive simulation input and output files in electronic format only
(on CD).

iv
LIST OF TABLES
(Tables at end of respective section)

TABLE
NO. TITLE
3-1 Saturated and Unsaturated Hydraulic Properties of the White Mesa Mill Tailings Cell Cover Infiltration Model and Bedrock Vadose Zone Contaminate Transport
Model
4-1 Model-Predicted Chloride Concentrations at the Bottom of the Bedrock Vadose
Zone After 240 Years for Cells 2 and 3 Evaluated as Part of the Sensitivity Analysis
4-2 Model-Predicted Sulfate Concentrations at the Bottom of the Bedrock Vadose
Zone After 240 Years for Cells 2 and 3 Evaluated as Part of the Sensitivity
Analysis
4-3 Model-Predicted Depth Within Bedrock Vadose Zone at Which Uranium Concentration is Equal to the Minimum Groundwater Compliance Limit After 240 Years for Cells 2 and 3 Evaluated as Part of the Sensitivity Analysis

LIST OF FIGURES
(Figures at end of respective section)

FIGURE
NO. TITLE
2-1 Site Map
2-2 Generalized Cross Section with Modified Cover Design
2-3 Annual Precipitation at Blanding, Utah (1905-2005)
2-4 Daily Precipitation at Blanding, Utah (1905-2005)
2-5 Piezometric Surface Contours Perched Aquifer (December 2009)
3-1 Modeling Approach, Model Domain and Boundary Conditions
4-1 Model-Predicted Water Flux Rate Through Tailing Cell Cover (Typical 57-Year Period)
4-2 Tailings Saturated Thickness (Head on Liner) During the Operational,
Dewatering, and Post-Closure Steady-State Timeframes for Cells 2 & 3
4-3 Model-Predicted Water Flux Rate at the Bottom of The Bedrock Vadose Zone
(Entering the Perched Aquifer) During the Operational, Dewatering, and Post-Closure Steady-State Timeframes for Cells 2 & 3
4-4 Model-Predicted Volumetric Water Content Throughout the Bedrock Vadose
Zone During the Operational, Dewatering, and Post-Closure Steady-State
Timeframes for Cells 2 & 3

v
LIST OF ACRONYMS

µg/l micrograms per liter
µm micrometer
AE actual evaporation ADE advection-dispersion equation
ANP acid neutralization potential
AT actual transpiration
bgs below ground surface
cm/sec centimeters per second
cm/d centimeters per day
D & D decommissioning and deconstruction DRC Division of Radiation Control

ET evapotranspiration
FML flexible membrane liner ft/ft feet per foot

GCL geosynthetic clay liner
GWCL Ground Water Compliance Limits GWQS Ground Water Quality Standards
HDPE high-density polyethylene
HFO hydrous ferric oxide
ICTM infiltration and contaminant transport modeling

K d partition coefficient

L length
M mass
mg/L milligrams per liter
mm/yr millimeters per year
MSL mean sea level MWH MWH Americas, Inc.

PE potential soil evaporation
PET potential evapotranspiration
pH potentiometric hydrogen ion concentration PT potential transpiration

vi
PVC poly vinyl chloride
SCK·CEN Belgian Nuclear Research Centre

T time U.S. EPA United States Environmental Protection Agency
USGS United States Geological Survey

ES-1
EXECUTIVE SUMMARY
This document presents the results of infiltration and contaminant transport modeling to
support Denison Mines (USA) Corp.’s Ground Water Discharge Permit (Permit No.
UGW370004 revised version dated 20 January 2010) (the “Permit”) for its White Mesa
uranium milling and tailings disposal facility (the “Mill”). As described in Part I.H.2 of
the Permit, Denison is required to prepare an infiltration and contaminant transport
modeling (ICTM) report.
The primary objectives of the infiltration and contaminant transport models are to
demonstrate the long-term ability of the tailings cells cover system to adequately contain
and control tailings contaminants and protect nearby groundwater quality of the
uppermost aquifer.
This Revised ICTM Report was prepared based on comments received from the Utah
Division of Radiation Control (DRC) on 2 February 2009 on the November 2007 ICTM
Report and discussions at meetings held between the DRC, Denison, and MWH on 31
March 2009 and 2 September 2009.
BACKGROUND INFORMATION
Facility Description
The White Mesa Mill is located in southeastern Utah, approximately six miles south of
Blanding, Utah. The Mill includes a mill facility and tailings cells located south of the
Mill. The tailings cells comprise the following:
• Cell 1 – 55 acres, used for the evaporation of process solutions
• Cell 2 – 65 acres, used for storage of barren tailings sands
• Cell 3 – 70 acres, used for storage of barren tailings sands and evaporation of
process solutions
• Cell 4A – 40 acres, used for storage of barren tailings sands and evaporation of process solutions
• Cell 4B – currently being permitted (approximately 40 acres to be used for
storage of barren tailings sands and evaporation of process solutions).

ES-2
The tailings cells generally were excavated into the underlying Dakota Sandstone and are
separated by dikes composed of compacted earthen materials. In the vicinity of the
tailings cells, the perched water table is approximately 75 to 115 ft below ground surface,
which is 40 to 90 ft below the bottom of the tailings cells.
Proposed Tailings Cover Design
The construction of a monolithic evapotranspiration (ET) cover is proposed to cap the
entirety of all tailings cells. The proposed ET cover would be 2.84-m (9.3-ft) thick and
would consist of (from top to bottom):
• 15 cm (0.5 ft) of a gravel-amended topsoil admixture to promote revegetation
and provide for protection against erosion and frost damage
• 107 cm (3.5 ft) of random fill soil (sandy clayey silt) placed at 85% of
Standard Proctor dry density to serve as a water storage, biointrusion, and
radon attenuation layer
• 162 cm (5.3 ft) of random fill soil (sandy clayey silt) composed of 2.8 feet of
random fill compacted to 95% of Standard Proctor dry density over 2.5 feet of
random fill placed at 80% of Standard Proctor dry density, to serve as grading
(platform fill) and radon attenuation layers.
A monolithic ET cover is the preferred design to minimize infiltration and meet the radon
attenuation standard. The proposed cover design will be sufficient to provide adequate
thickness to protect against frost penetration, provide adequate water storage capacity to
minimize the rate of infiltration into the underlying tailings, and provide long-term
moisture within the cover to attenuate radon flux.
GENERAL MODELING APPROACH
To evaluate performance of the cover system, a model of the cover system was
constructed to predict potential infiltration of water through the cover to the tailings.
Several cover designs were tested with the cover system infiltration model including the
currently permitted rock cover design and a proposed monolithic ET cover (and several

ES-3
variations). Minimizing infiltration through the cover and preventing build-up of
leachate head within the tailings are required for compliance with Part I.D.8 of the
Permit.
An additional requirement of Part I.D.8 is that the final design construction and operation
of the cover system will ensure that the groundwater quality at the compliance
monitoring wells does not exceed the Ground Water Quality Standards (GWQS’s) or
Ground Water Compliance Limits (GWCL’s). However, the cover system infiltration
model cannot evaluate impacts to groundwater quality. To evaluate potential impacts to
groundwater, a vadose zone model was constructed to predict potential flow and
contaminant transport through the bedrock vadose zone beneath the tailings cells. To
address specific requests of the DRC, the operational and dewatering phases of the
tailings cells were included in the modeling in addition to the 200-year regulatory
timeframe after the cells are closed (with the cover in place). For these simulations,
potential water flux rates through the liners were estimated for the operational phase,
dewatering phase, and post-closure steady state based on water levels estimated in the
tailings cells. The bedrock vadose zone model evaluates the potential impacts of the
tailings cell system as a whole (liner system, dewatering system, and cover system) on
groundwater for the project lifecycle, including the operational phase (without cell cover
system), the dewatering phase (with an interim cover only), and the 200-year regulatory
post-closure period (with complete cover system, but with some limited water remaining
in the tailings). For the 240-year period modeled, the potential flux rate and contaminant
transport through the underlying bedrock vadose zone is dominated by the effect of the
operational phase when the cells were not covered. As a result, the bedrock vadose zone
model including the operational phase is not a reliable indicator of performance of the
closed-cell cover system. However, even with the operational phase, model-predicted
contaminant concentrations in vadose zone pore water entering the perched aquifer did
not exceed the GWQS’s or GWCL’s for any downgradient monitoring wells, thus
demonstrating compliance with Part I.D.8 of the Permit.
Other modeling and calculations were performed to support initial conditions and
boundary conditions used in the cover system infiltration model and the bedrock vadose

ES-4
zone flow and transport models. Specific details of the modeling are presented
throughout the remainder of this report with supporting information provided in the
appendices.
Following conceptual-model development, numerical modeling was completed with the
following two basic models:
1. Cover Model: Infiltration modeling with HYDRUS-1D of the tailings cell
cover system with daily precipitation and evapotranspiration to estimate
potential long-term average infiltration rates to the tailings.
2. Bedrock Vadose Zone Model: Vadose zone flow and potential contaminant
transport modeling with HP1 (HYDRUS-1D coupled with the geochemical
program PHREEQC) through the bedrock vadose zone to the underlying
perched aquifer during the operational phase, dewatering phase, and 200-year
regulatory post-closure steady-state timeframes. Vadose zone properties were
based on the results of a detailed sampling program performed to characterize
geochemical properties of the bedrock. HYDRUS-1D was used to confirm
the results for flow and transport of a conservative solute (chloride) predicted
by HP1.
The contaminants modeled with HP1 included pH, major cations and anions necessary to
achieve charge balance (aluminum, calcium, carbonate, chloride, magnesium, potassium,
sodium, and sulfate), and selected trace elements (arsenic, cadmium, copper, iron, nickel,
uranium, vanadium, and zinc). The most dependable indicators of site water quality and
of potential cell failure are uranium and sulfate, due to their predominance, and chloride,
due to predominance and mobility. In particular, chloride will migrate unretarded and act
as a conservative tracer and thus would be expected to be detected before all other site
contaminants. Uranium was included because it is one of the primary contaminants of
concern.
To evaluate the potential “worst case” for build-up of water in the tailings (“bathtub
effect”), the model-predicted long-term average water flux rate through the tailings cell

ES-5
cover system was used to estimate the total amount of water entering the tailings during
the 200-year regulatory timeframe. By assuming a completely impermeable liner system
(i.e., no water flow through the liners; all water that infiltrated through the cover was
accumulated in the cells), the total amount of water entering the tailings through the
cover would be accumulated in the cell. By dividing this total water flux by the tailings
porosity, the potential rise in water levels in the tailings was calculated for this worst-case
scenario. Under this scenario, there would be no impacts to groundwater, because no
water would be leaving the cells.
MODEL RESULTS
The HYDRUS-1D infiltration model was used to predict potential water fluxes through
the tailings cell cover system. The HP1 bedrock vadose zone contaminant transport
model was used to predict the potential flow and transport of conservative (chloride) and
nonconservative (sulfate, uranium, and other trace elements) solutes through the bedrock
vadose zone to the perched aquifer. Sensitivity analyses were performed to evaluate the
impacts that uncertainty in parameter input values have on model results.
Model-Predicted Water Flux Rate for Tailings Cell Cover System
The model-predicted average long-term water flux rate through the proposed monolithic
ET tailings cell cover, assuming a historical climate record (based on climatic data
recorded between 1932 and 1988), was 0.45 mm/yr. The average long-term water flux
rate corresponds to approximately 0.1% of the average annual amount of precipitation
recorded at the Blanding weather station. This is in contrast to an average long-term
infiltration rate of 34 mm/yr predicted for the currently permitted rock cover design. The
increased performance and reduction of infiltration for the ET cover relative to the
original rock cover design, is attributed to the presence of vegetation and associated root
water uptake via transpiration. The model-predicted water flux rate through the
monolithic ET cover indicates that the available storage capacity of the cover should be
sufficient to significantly reduce infiltration, and the ET cover should function properly
as designed.

ES-6
A monolithic ET cover is the preferred design to minimize infiltration necessary to meet
the Permit requirements (Part I.D.8) and meet the radon attenuation standard. The
material thicknesses for the different cover layers were based on the results of radon
attenuation modeling to achieve the State of Utah’s long-term radon emanation standard
for uranium mill tailings (Utah Administrative Code, Rule 313-24). Furthermore, the
proposed cover design will be sufficient to provide adequate thickness to protect against
frost penetration and biointrusion, provide adequate water storage capacity to minimize
the rate of infiltration into the underlying tailings, and provide long-term moisture within
the cover to attenuate radon flux.
Evaluation of Build-up of Waters in Tailings
To evaluate the potential for build-up of water in the tailings (“bathtub effect”), the long-
term average water flux rate through the tailings cell cover system (predicted with the
infiltration model) was used to calculate the amount of water entering the tailings during
the 200-year regulatory timeframe specified by the Permit. The amount of water
expected to migrate through the cover and enter the tailings cells (i.e., assuming all
recharge to the tailings can act to increase the amount of head on the liner) was then used
to calculate the maximum potential rise in water levels in the tailings assuming no water
flow through the liners (i.e., all water that infiltrated through the cover was accumulated
in the cells). The assumptions for evaluating the “bathtub effect” result in an end-
member scenario expected to produce a conservative estimate of closed-cell cover system
performance.
The amount of water calculated to enter the tailings after 200 years is equal to 90
millimeters (0.3 feet) of water. Assuming a tailings porosity of 57%, the calculated
water-level rise on the liner is approximately 160 millimeters (0.53 feet). Consequently,
a significant build-up of water (“bathtub effect”) within the cells is not anticipated and
the leachate head within the tailings is not predicted to rise above or over-top the
maximum liner elevation (which typically is greater than 20 feet above the bottom of the
cell), meeting the requirement of the Permit (Part I.D.8).

ES-7
Bedrock Vadose Zone Flow and Contaminant Transport Modeling
The bedrock vadose zone flow and contaminant transport model was used to predict
potential flow rates and contaminant transport rates through the bedrock vadose zone to
the perched aquifer during the operational, dewatering, and post-closure steady-state
timeframes. Solute transport models were developed for the bedrock vadose zone
beneath Cell 1 (contingency cell identified for the potential disposal of decommissioning
and deconstruction debris), Cells 2 & 3, and Cells 4A & 4B. For simplicity, a vadose
zone thickness of 12.8 meters (42 feet) was assumed for all of the simulations. This is a
conservative assumption given that the average vadose zone thicknesses beneath Cell 2,
Cell 3, and Cell 4A are 19.2 m (63 ft), 20.1 m (66 ft), and 17.1 m (56 ft). HP1 was used
to simulate potential solute transport of conservative (chloride) and nonconservative
(sulfate, uranium, and other trace elements) solutes through the bedrock vadose zone
beneath the tailings cells.
Potential water flux rates through the primary liner installed beneath Cells 2 & 3 and the
secondary liner installed beneath Cells 4A & 4B were calculated using the Giroud-
Bonaparte Equation. Estimates of potential water flux rates through the liners were used
as an upper boundary condition (time-dependent flux) for the HP1 model used to predict
flow and solute transport through the bedrock vadose zone to the perched aquifer during
the operational, dewatering, and post-closure steady-state timeframes. The average long-
term water flux rate through the ET cover (predicted with the infiltration model) was used
as an upper boundary condition (constant flux) for Cell 1 to represent the post-closure
steady-state period. The bottom of Cell 1 (if constructed) will contain a soil liner
compacted to achieve low permeability, but this layer was not included in the modeling,
which yields conservative estimates of solute transport through the bedrock vadose zone.
The calculated potential water flux rates through the liners were multiplied by the
average solute concentrations measured in the tailings slimes drains to yield a time-
dependent mass flux rate applied as an upper boundary condition to the top of the
bedrock vadose zone. The average solute concentrations were used as input to represent
the source term solution chemistry of the tailings pore water.

ES-8
Cells 2 & 3 Model-Predicted Water Flux Rate. The potential water flux rate at the
bottom of the bedrock vadose zone (immediately above the perched aquifer) is predicted
to reach a maximum value of approximately 7.5 mm/yr after 25 years of tailings cell
operation (note that tailings cells are not covered during this period). The potential flux
rate is then predicted to rapidly decline in response to decreased head (saturated
thickness) that occur in the tailings during the dewatering phase, ultimately reaching a
long-term steady state value of approximately 0.7 mm/yr during the 200-year regulatory
post-closure period. There is considerable evidence that the cells are not leaking.
Consolidation of fine-grained tailings and deposition of tailing slimes, coupled with the
chemical nature of the pore water (e.g., precipitation of gypsum and amorphous mineral
phases), is anticipated to essentially seal some of the defects, which would act to decrease
the potential flux rates through the liners.
Cells 2 & 3 Model-Predicted Chloride Concentration. The model-predicted increase
in chloride concentrations at the bottom of the bedrock vadose zone beneath Cells 2 & 3
after 240 years (including operational, dewatering, and post-closure periods) of transport
is 0.01 mg/L. The chloride concentration at the bottom of the vadose zone represents the
model-predicted addition of chloride as a result of the potential flux from the tailings
cells. While there is naturally-occurring chloride in the vadose zone, the modeling
assumed no initial chloride for simplicity, and because there is a lack of data concerning
background chloride concentrations and the distribution of chloride within the vadose
zone. Furthermore, the model-predicted chloride concentration is the solute
concentration in vadose zone pore water that will reach the perched aquifer; however, the
predicted concentration is not equal to the concentration in groundwater. A model was
not constructed to determine the actual (diluted) concentration in groundwater because
the chloride concentration predicted at the bottom of the vadose zone was orders of
magnitude less than the minimum GWCL for chloride, which is 10 mg/L. The minimum
GWCL (for chloride and all other solutes modeled) was selected from the list of
monitoring wells located immediately downgradient from the tailings cells (i.e.,
monitoring wells MW-5, MW-11, MW-12, MW-14, MW-15, MW-23, MW-24, MW-28,
MW-29, MW-30, and MW-31; GWCL’s for these wells are specified in the Permit).

ES-9
Cells 2 & 3 Model-Predicted Sulfate Concentration. The model-predicted sulfate
concentration at the bottom of the bedrock vadose zone beneath Cells 2 & 3 after 240
years of transport is 0.014 mg/L. The distribution of sulfate within the bedrock vadose
zone is controlled by the amount of gypsum that may precipitate from solution. The
sulfate concentration at the bottom of the bedrock vadose zone represents the model-
predicted addition of sulfate as a result of the potential flux from the tailings cells. A
model was not constructed to determine the actual (diluted) concentration in groundwater
because the sulfate concentration predicted at the bottom of the vadose zone was orders
of magnitude less than the minimum GWCL for sulfate, which is 532 mg/L for
monitoring wells located immediately downgradient from the tailings cells.
Cells 2 & 3 Model-Predicted Uranium Concentration. Uranium is not predicted to
reach the bottom of the bedrock vadose zone beneath Cells 2 & 3 during the 240-year
timeframe. Adsorption of uranium onto the surface of hydrous ferric oxide (HFO)
present in the bedrock vadose zone limits the transport distance below the liner. The
depth at which the model-predicted uranium concentration is approximately equal to the
minimum GWCL (0.0049 mg/L) after 240 years is 2.3 meters (8 feet) below the tailing
cell liner system; a minimum of 10.5 meters (34 feet) above the perched water table. The
uranium concentration within the bedrock vadose zone represents the model-predicted
addition of uranium as a result of the potential flux from the tailings cells. HFO is the
only solid phase that serves as a potential sorption site of uranium and other trace
elements, which is a conservative assumption because other phases (e.g., hematite,
quartz, clays, etc.) also participate in surface complexation reactions.
Cells 2 & 3 Model-Predicted Concentration of Other Trace Elements. The sorption
of uranium was competitive because additional trace elements were modeled. Solutes
included in the model were based on their elevated concentrations in the tailings pore
water as compared to the GWCLs. Transport of the following trace elements was
modeled: arsenic, cadmium, copper, nickel, vanadium, and zinc. Similar to uranium,
these solutes were predicted to migrate a limited distance below the liner (e.g., a few
meters).

ES-10
Cells 4A & 4B Model-Predicted Water Flux Rate. The calculated potential flux of
water through the secondary liner beneath Cells 4A & 4B for the maximum head within
the leak detection system during the operational and dewatering periods is approximately
8 x 10-5 mm/yr. The potential flux rates predicted at the end of dewatering are assumed
to equal the rate during post-closure steady state because the increase in water levels is
anticipated to be minor. Therefore, the model-predicted water flux rate at the bottom of
the bedrock vadose zone (immediately above the perched aquifer) during post-closure
steady-state is 8 x 10-5 mm/yr.
Cells 4A & 4B Model-Predicted Concentrations. For all practical purposes, chloride is
not predicted to reach the bottom of the bedrock vadose zone during the 12-year
operational and 200-year post-closure periods (the chloride concentration predicted to
reach the water table at 212 years was 5 x 10-14 mg/L). The chloride concentration is not
predicted to exceed the 10 mg/L minimum GWCL anywhere in the vadose zone because
of the diminutive chloride mass flux rate entering the vadose zone. Considering that
chloride is a conservative tracer, and that transport is not affected by sorption or mineral
precipitation reactions, coupled with the fact that the model predictions demonstrate
nearly zero impact, additional model predictions of solute transport for nonconservative
contaminants (sulfate, uranium, other trace elements) was considered unnecessary.
Cell 1 Model-Predicted Water Flux Rate. If Cell 1 is constructed for decommissioning
and deconstruction disposal, it will include a soil liner compacted to achieve low
permeability and will be covered with the monolithic ET cover. The cover design will be
the same as the monolithic ET cover proposed for the other cells. Consequently, the
long-term average infiltration rate would be equivalent to the value presented for the
other cells. The model-predicted water flux rate at the bottom of the vadose zone
(immediately above the perched aquifer) during 200-year post-closure steady-state is
predicted to be approximately 0.5 mm/yr.
Cell 1 Model-Predicted Concentrations. The source term of the decommissioning and
deconstruction debris is assumed to equal the concentrations assigned to the tailings pore
water, which is anticipated to lead to conservative predictions that over predict the

ES-11
potential impacts. For all practical purposes, chloride is not predicted to reach the bottom
of the bedrock vadose zone during the 200-year transport timeframe (the chloride
concentration predicted to reach the water table at 200 years was 7 x 10-9 mg/L).
Considering that chloride is a conservative tracer, and that transport is not affected by
sorption or mineral precipitation reactions, coupled with the diminutive transport
distance, additional model predictions of solute transport for nonconservative
contaminants (sulfate, uranium, other trace elements) was considered unnecessary.
CONCLUSIONS
The assumptions used to construct the numerical models to predict infiltration through
the cover and potential impacts to the perched groundwater system, generally were either
conservative or based on anticipated conditions. As a result, the predictions are
considered to be conservative. The proposed monolithic ET cover will minimize
infiltration into the tailings, will prevent build-up of leachate head on the cell liner, and
will be protective of groundwater quality; contaminant concentrations are not predicted to
exceed the GWCS’s or GWCL’s at the compliance monitoring wells specified in the
Permit, thus demonstrating compliance with the Permit. Furthermore, the results of the
radon attenuation modeling demonstrate that the proposed monolithic ET cover will
attenuate radon fluxes thereby achieving the State of Utah’s long-term radon emanation
standard for uranium mill tailings (Utah Administrative Code, Rule 313-24).

1-1
1.0 INTRODUCTION
This document presents the results of infiltration and contaminant transport modeling to
support Denison Mines (USA) Corp.’s Ground Water Discharge Permit (Permit No.
UGW370004 revised version dated 20 January 2010) (the “Permit”) for its White Mesa
uranium milling and tailings disposal facility (the “Mill”). As described in Part I.H.2 of
the Permit, Denison is required to prepare an infiltration and contaminant transport
modeling (ICTM) report.
Denison has engaged MWH Americas, Inc. (MWH) to work with Denison personnel to
develop the assumptions and data for the infiltration and contaminant transport models
and interpret the model results.
1.1 OBJECTIVES OF INFILTRATION AND CONTAMINANT TRANSPORT
MODELS
The primary objectives of the infiltration and contaminant transport models are to
demonstrate the long-term ability of the tailings cells cover system to adequately contain
and control tailings contaminants and protect nearby groundwater quality of the
uppermost aquifer.
1.2 PERMIT REQUIREMENTS
Part I.H.2 (Infiltration and Contaminant Transport Modeling Work Plan and Report) of
Denison’s Permit presents the requirements for infiltration and contaminant transport
modeling, as summarized below.
An infiltration and contaminant transport modeling report that demonstrates the long-
term ability of the tailings cells cover system to adequately contain and control tailings
contaminants and protect nearby groundwater quality of the uppermost aquifer must be
submitted to the Utah Division of Radiation Control (DRC) for Executive Secretary
approval. This report shall demonstrate how the tailings cell engineering design and

1-2
specifications will comply with the minimum performance requirements of Part I.D.8 for
Closed Cell Performance Requirements] of the Permit.
The infiltration and contaminant transport modeling report must describe:
• Applicable and pertinent historic studies and modeling reports relevant to the
tailings cell cover design and tailings cell system performance.
• Information necessary for infiltration and contaminant transport modeling,
including representative input values for vadose zone and aquifer soil-water
partitioning (Kd) coefficients, tailings source term concentrations, tailings
waste leach rates, vadose zone and aquifer velocities and dispersivity,
contaminant half-life or other rates of decay, etc. If any required information
is not currently available, conservative assumptions can be used for the model
input.
• Computer models that will be used to simulate long-term performance of the
tailings cells cover system. Specific information on model design, including
governing equations and their applicability to site conditions, grid design,
duration of simulation, and selection of time steps must be described.
• The conceptual models used and justification why they are representative or
conservative of actual field conditions at the site. The conceptual models will
identify the physical domains and geometries simulated including the tailings
cell design and construction, all boundary and initial conditions assigned in
the models, and the shallow aquifer locations where future potential
contaminant concentrations have been predicted.
• How the infiltration and contaminant transport problem has been
conceptualized, planned, and executed to demonstrate compliance with the
requirements of Part I.D.8 of the Permit.

1-3
• Model results, model calibration, steady state conditions, sensitivity analyses,
post-model audit plan.
Additionally, Part I.D.8 (Closed Cell Performance Requirements) of the Permit presents
requirements regarding performance requirements for closed cells at the facility, which
impacts both actual infiltration at the site as well as how this infiltration will be modeled,
as follows:
• Before reclamation and closure of any tailings disposal cell, the Permittee
shall ensure that the final design, construction, and operation of the cover
system at each tailings cell will comply with all requirements of an approved
Reclamation Plan, and will for a period of not less than 200 years meet the
following minimum performance requirements:
- Minimize infiltration of precipitation or other surface water into the
tailings, including, but not limited to the radon barrier.
- Prevent the accumulation of leachate head within the tailings waste layer
that could rise above or over-top the maximum flexible membrane liner
(FML) elevation internal to any disposal cell, i.e., create a “bathtub
effect”.
- Ensure the groundwater quality at the compliance monitoring wells does
not exceed the Ground Water Quality Standards (GWQS’s) or Ground
Water Compliance Limits (GWCL’s) specified in Part I.C.1 and Table 2
of the Permit.
Further, Part I.C.1 (Permit Limits) of the Permit includes the following:
• The Permittee shall comply with the following GWCL’s – contaminant
concentrations measured in each monitoring well shall not exceed the
GWCL’s defined in Table 2 of the Permit. Groundwater quality at the site
must at all times meet all the applicable GWQS’s and the ad hoc GWQS’s

1-4
defined in R317-6 even though the Permit does not require monitoring for
each specific contaminant.
Part I.H.2.f also states that “Upon Executive Secretary approval of the final infiltration
and contaminant transport report, the Reclamation Plan may be modified to accommodate
necessary changes to protect public health and the environment.”
The infiltration and contaminant transport modeling report has been prepared to comply
with the Permit as described above.
1.3 GENERAL MODELING APPROACH TO ADDRESS PERMIT
REQUIREMENTS
The Permit specifically states that the purpose of the infiltration modeling is to evaluate
the closed-cell cover system performance. To evaluate performance of the cover system,
a model of the cover system was constructed to predict potential infiltration of water
through the cover to the tailings. Several cover designs were tested with the cover system
infiltration model including the currently permitted rock cover design and a proposed
monolithic evapotranspiration (ET) cover (and several variations). Minimizing
infiltration through the cover and preventing build-up of leachate head within the tailings
are required for compliance with Part I.D.8 of the Permit.
An additional requirement of Part I.D.8 is that the final design construction and operation
of the cover system will ensure that the groundwater quality at the compliance
monitoring wells does not exceed the GWQS’s or GWCL’s. However, the cover system
infiltration model cannot evaluate impacts to groundwater quality. To evaluate potential
impacts to groundwater, a vadose zone model was constructed to predict potential flow
and contaminant transport through the bedrock vadose zone beneath the tailings cells. To
address specific requests of the DRC, the operational and dewatering phases of the
tailings cells were included in the modeling in addition to the 200-year regulatory
timeframe after the cells are closed (with the cover in place). For these simulations,
potential water flux rates through the liners were estimated for the operational phase,
dewatering phase, and post-closure steady state based on water levels estimated in the

1-5
tailings cells. The bedrock vadose zone model evaluates the potential impacts of the
tailings cell system as a whole (liner system, dewatering system, and cover system) on
groundwater for the project lifecycle, including the operational phase (without cell cover
system), the dewatering phase (with an interim cover only), and the 200-year regulatory
post-closure period (with complete cover system, but with some limited water remaining
in the tailings). For the 240-year period modeled, the potential flux rate and contaminant
transport through the underlying bedrock vadose zone is dominated by the effect of the
operational phase when the cells were not covered. As a result, the bedrock vadose zone
model including the operational phase is not a reliable indicator of performance of the
closed-cell cover system. However, even with the operational phase, model-predicted
contaminant concentrations in vadose zone pore water entering the perched aquifer did
not exceed the GWQS’s or GWCL’s for any downgradient monitoring wells, thus
demonstrating compliance with Part I.D.8 of the Permit.
Other modeling and calculations were performed to support initial conditions and
boundary conditions used in the cover system infiltration model and the bedrock vadose
zone flow and transport models. Specific details of the modeling are presented
throughout the remainder of this report with supporting information provided in the
appendices.
1.4 INFILTRATION AND CONTAMINANT TRANSPORT MODELING
REPORT HISTORY
The original Permit specified that a work plan must be submitted and approved before the
ICTM report could be prepared. Denison submitted a work plan to the DRC in a letter
dated 3 September 2005. However, the DRC did not review this work plan and removed
this requirement from the Permit as stated in a letter from the Executive Secretary to
Denison dated 3 November 2006.
The ICTM report was submitted to the DRC for Executive Security approval on
21 November 2007. The DRC reviewed the report and submitted review comments and a
request for additional information in a letter to Denison dated 2 February 2009. To
facilitate discussion and provide clarification regarding the DRC’s comments, a meeting

1-6
was held between the DRC, Denison, and MWH on 31 March 2009 at the DRC’s office
in Salt Lake City, Utah. A follow-up meeting was held on 2 September 2009 also at the
DRC’s office in Salt Lake City. Meeting minutes for these two meetings were prepared
and approved by the DRC. On 1 December 2009, Denison submitted a memorandum
prepared by MWH that provided preliminary responses to the DRC’s comments and
request for additional information. Subsequently, a deadline of 31 March 2010 was
established for submittal of the revised ICTM report. The revised ICTM report, and
supporting documentation contained within the appendices, is submitted here in its
entirety. The 2010 ICTM report submitted here supersedes the 2007 ICTM report.
1.5 DOCUMENT ORGANIZATION
The remainder of this report includes the following sections:
• Section 2.0 – Site Background; descriptions of the site including tailings cell
cover and liner designs, as well as tailings chemical and physical
characteristics, site geology and hydrogeology, conceptual model of water
flow (infiltration) through the tailings cell cover, and conceptual model of
water flow and potential contaminant transport through the vadose zone
• Section 3.0 – Methodology; descriptions of the tailings cell cover infiltration
model, vadose zone flow and transport model, tailings cell dewatering model,
input parameters and boundary conditions, and modeling assumptions
• Section 4.0 – Results; descriptions of the results of the tailings cell cover
infiltration model, vadose zone flow and transport model, tailings cell
dewatering model, and sensitivity analysis
• Section 5.0 – Conclusions; summary of the conclusions of the tailings cell
cover infiltration model, and bedrock vadose zone flow and transport model,
along with recommendations for a post-audit monitoring plan
• Section 6.0 – References

1-7
• Appendix A – Laboratory reports with results of vadose zone mineralogical
testing and properties of stockpiled soil
• Appendix B – Laboratory report with unsaturated and saturated hydraulic
properties of the bedrock core samples
• Appendix C – Bedrock sampling to characterize hydraulic and geochemical
properties of the vadose zone
• Appendix D – Vegetation evaluation for the evapotranspiration cover
• Appendix E – Comparison of cover designs based on infiltration modeling
• Appendix F – Evaluation of the effects of storm intensity on infiltration
through evapotranspiration cover
• Appendix G – Sensitivity analysis comparing infiltration rates through the
evapotranspiration cover based on cover vegetation, biointrusion, and
precipitation
• Appendix H – Radon emanation modeling for the evapotranspiration cover
• Appendix I – Tailings hydraulic conductivity evaluation
• Appendix J – Tailings cell dewatering modeling
• Appendix K – Statistical evaluation of tailings pore water chemistry and
identification of source term concentrations
• Appendix L – Evaluation of potential water flow through the tailings cell
liners
• Appendix M – Geochemical model and reactive transport modeling of flow
and transport through the vadose zone
• Appendix N – Predictive simulation input and output files in electronic format
only (on CD).

2-1
2.0 BACKGROUND
This section provides information on the:
• Site background including descriptions of the White Mesa Mill facility,
proposed tailings cell cover design, tailings cell liner systems, and tailings
chemical and physical characteristics;
• Site characteristics including descriptions of climate, geology, hydrogeology
of the perched aquifer system, groundwater quality of the perched aquifer
system, and vadose zone hydrogeology and geochemistry of the unsaturated
bedrock;
• Conceptual model of water flow (infiltration) through the tailings cell cover;
and
• Conceptual model of water flow and potential contaminant transport through
the vadose zone.
Site-specific studies and reports reviewed to prepare this modeling report included:
• Engineering Report, Tailings Management System, White Mesa Uranium
Project, Blanding, Utah (D’Appolonia Consulting Engineers, Inc., 1979)
• Construction Report, Initial Phase – Tailings Management System, White
Mesa Uranium Project, Blanding, Utah (D’Appolonia Consulting Engineers,
Inc., 1982)
• Cell 4A Lining System Design Report for the White Mesa Mill, Blanding, Utah
(Geosyntec Consultants, 2006a)
• Stockpile Evaluation Tailings Cell 4A, White Mesa Mill - Technical Memo
submitted to International Uranium (USA) Corporation (Geosyntec
Consultants, 2006b)

2-2
• Cell 4B Design Report, White Mesa Mill, Blanding, Utah (Geosyntec
Consultants, 2007a)
• Revised Construction Drawings, DMC White Mesa Mill, Cell 4A Lining
System (Geosyntec Consultants, 2007b)
• Analysis of Slimes Drains for White Mesa Mill, Cell 4A (Geosyntec
Consultants, 2007c)
• Hydraulic Testing at the White Mesa Uranium Mill Site, near Blanding, Utah
during July 2002 (Hydro Geo Chem, Inc., 2002)
• Site Hydrogeology and Estimation of Groundwater Pore Velocities in the
Perched Zone, White Mesa Uranium Mill Site near Blanding, Utah (Hydro
Geo Chem, Inc., 2009)
• Revised Background Groundwater Quality Report: Existing Wells for Denison
Mines (USA) Corp.’s White Mesa Mill Site, San Juan County, Utah (INTERA,
Inc., 2007a)
• Revised Addendum Evaluation of Available Pre-Operational and Regional
Background Data Background Groundwater Quality Report: Existing Wells
for Denison Mines (USA) Corp.’s White Mesa Mill Site, San Juan County,
Utah (INTERA, Inc., 2007b)
• Revised Addendum Background Groundwater Quality Report: New Wells for
Denison Mines (USA) Corp.’s White Mesa Mill Site, San Juan County, Utah
(INTERA, Inc., 2008)
• Summary of Work Completed, Data Results, Interpretations, and
Recommendations for the July 2007 Sampling Event at the Denison Mines,
USA, White Mesa Uranium Mill, near Blanding, Utah (Hurst and Solomon,
2008)
• Reclamation Plan, White Mesa Mill, Blanding, Utah, Source Material License
No. SUA-1358, Docket No. 40-8681, Revision 3.0 (IUC, 2000)

2-3
• Reclamation Plan, White Mesa Mill, Blanding, Utah, Radioactive Materials
License No. UT1900479, Revision 4.0 (Denison, 2009)
• Hydrogeological Evaluation of White Mesa Uranium Mill (TITAN
Environmental Corporation, 1994)
• Tailings Cover Design, White Mesa Mill, Blanding Utah (TITAN
Environmental Corporation, 1996).
Complete citations for these and other sources cited throughout this document are
provided in the References section.
2.1 SITE OVERVIEW
2.1.1 Facility Description
The White Mesa Mill is located in southeastern Utah, approximately six miles south of
Blanding, Utah. The Mill includes a mill facility and tailings cells located south of the
Mill (see Figure 2-1). The focus of this report is the tailings cells; for information
concerning site history or milling operations, see the Reclamation Plan (IUC, 2000;
Denison, 2009).
The tailings cells comprise the following:
• Cell 1 – 55 acres, used for the evaporation of process solutions
• Cell 2 – 65 acres, used for storage of barren tailings sands
• Cell 3 – 70 acres, used for storage of barren tailings sands and evaporation of
process solutions
• Cell 4A – 40 acres, used for storage of barren tailings sands and evaporation
of process solutions

2-4
• Cell 4B – currently being permitted (approximately 40 acres to be used for
storage of barren tailings sands and evaporation of process solutions).
The tailings cells generally were excavated into the underlying Dakota Sandstone and are
separated by dikes composed of compacted earthen materials. The tailings cells are lined
as described in Section 2.1.3. In the vicinity of the tailings cells, the perched water table
is approximately 75 to 115 ft below ground surface, which is 40 to 90 ft below the bottom
of the tailings cells.
The White Mesa Mill is a zero-discharge facility, thus all liquids must be eliminated
through evaporation. Currently, Denison is actively evaporating process waters from
Cell 1, Cell 3, and Cell 4A. Cell 1 is currently used as an evaporation pond only and will
not be used to hold solid tailings. During site closure the solution in the Cell 1 pond will
be evaporated dry and the evaporite crystals, sediment, geomembrane liner, and any
contaminated underlay (foundation) material will be relocated to another cell. Disposal
of the decommissioning material in Cell 1 is identified as a contingency in case other
cells (e.g., Cell 4B) do not have adequate storage for such material. The cover system
constructed above Cell 1 would be identical to the design proposed for the other tailings
cells.
Cell 2 is no longer receiving tailings and has been covered with approximately three feet
of soil. Cell 3 is near to being full of tailings and is in the process of being covered. The
interim soil cover is placed to facilitate site closure and will be used as platform fill to
achieve sufficient grading and provide a stable working surface. Water removed from
Cells 2 & 3 by the dewatering systems will be discharged to Cell 1 and subsequently
evaporated. Cell 4A is currently receiving tailings and will eventually be filled with
tailings and covered during site reclamation. Cell 4B is currently being permitted and
would be operated in a manner similar to Cell 4A. Descriptions of the proposed tailings
cover system and constructed liner systems are provided in the sections below. The
proposed cover system would be constructed across all of the tailings cells.

2-5
2.1.2 Proposed Tailings Cover Design
The construction of a monolithic evapotranspiration (ET) cover is proposed as part of this
ICTM report to cap the entirety of all tailings cells. The proposed 2.84-m (9.3-ft) thick
monolithic ET cover design (see Figure 2-2) would consist of (from top to bottom):
• 15 cm (0.5 ft) of a gravel-amended topsoil admixture to promote revegetation
and provide for protection against erosion and frost damage
• 107 cm (3.5 ft) of random fill soil (sandy clayey silt) placed at 85% of
Standard Proctor dry density to serve as a water storage, biointrusion, and
radon attenuation layer
• 162 cm (5.3 ft) of random fill soil (sandy clayey silt) composed of 2.8 feet of
random fill compacted to 95% of Standard Proctor dry density over 2.5 feet of
random fill placed at 80% of Standard Proctor dry density, to serve as grading
(platform fill) and radon attenuation layers.
A monolithic ET cover is the preferred design to minimize infiltration and meet the radon
attenuation standard. The proposed cover design will be sufficient to provide adequate
thickness to protect against frost penetration, provide adequate water storage capacity to
minimize the rate of infiltration into the underlying tailings, and provide long-term
moisture within the cover to attenuate radon flux.
Details regarding the short-term establishment and long-term sustainability of the
vegetative component of the ET cover are summarized in Appendix D. Empirical data
regarding the ecological characteristics of the species mix (rooting depth and root
distribution) and established plant community (percent cover) were summarized from the
literature and nearby lysimeter studies to develop a conceptual model of the vegetative
component for the ET cover system. The empirical data were then used to parameterize
the infiltration model and predict the ET cover’s performance over the long term (see
Appendices E and G).

2-6
The monolithic ET cover was tested with the infiltration model as described in Sections
3.0 and 4.0. A sensitivity analysis was performed to test variations in the ET cover
design and the proposed design and variations demonstrated significantly improved
performance over the currently permitted rock cover design (see Appendix E for details
of this comparison). Results of modeling the emanation of radon-222 from the top
surface of the monolithic ET cover are presented in Appendix H. The proposed cover
design replaces the top surface of the cover; the side slope design may include rock
armoring, as in the original design (TITAN Environmental, 1996).
2.1.3 Tailings Cell Liner Systems
Cells 2 & 3. The tailings liner systems for Cells 2 & 3 are identical and consist of a
slimes drain collection system overlying a single liner (see Figure 2-2). The design
consists of (from top to bottom):
• slimes drain system (cell bottom only)
• liner protective blanket
• 30-mil (0.03-inch) poly vinyl chloride (PVC) flexible membrane liner (FML)
• 6-inch compacted bedding material
• prepared subgrade with limited leak detection system (i.e., a single pipe at the
toe of the southern dike).
Cells 4A & 4B. The tailings liner system for Cell 4A is double lined, and consists of a
slimes drain collection system overlying a primary liner, leak detection system, and
composite secondary liner (see Figure 2-2). A composite liner is defined as a
geomembrane liner underlain by a low-permeability soil (e.g., naturally compacted soil or
geosynthetic clay layer). The design for Cell 4B is currently under review, but
preliminary drawings indicate a design identical to that of Cell 4A, with minor
deviations. The design consists of (from top to bottom):

2-7
• slimes drain system (cell bottom only)
• 60-mil (0.06-inch) high-density polyethylene (HDPE) geomembrane (primary
liner)
• geonet drainage layer (leak detection system)
• 60-mil (0.06-inch) HDPE geomembrane (secondary liner)
• geosynthetic clay liner (GCL)
• prepared subgrade.
Slimes drain systems are installed in Cells 2, 3, 4A, and 4B. The slimes drains in Cells 2
& 3 include both 1.5-inch and 3-inch diameter slotted PVC pipe installed in a 1-ft thick
clean sand layer above the protective blanket. These lateral drains are installed on 50-ft
centers parallel to the southern edge of the tailings cells and cover an area that is
approximately 400 ft (north-south) by 600 ft (east-west). The slimes drains in Cells 4A
& 4B are on 50-ft centers and are located beneath the entirety of the cells. Leak detection
systems are installed under the cells and are monitored weekly. Details of the liner
systems are provided in D’Appolonia Consulting Engineers (1982) for Cells 2 & 3, in
Geosyntec Consultants (2006a) for Cell 4A, and in Geosyntec Consultants (2007) for
Cell 4B.
There is strong evidence to suggest that no significant leakage has occurred through the
liner systems beneath Cells 2 & 3 over the past 30 years. Evidence that Cells 2 & 3 are
not leaking includes:
• No significant leakage indicated by the leak detection systems
• No leakage indicated by the perched aquifer water table surface elevations

2-8
• No observations of contamination (e.g., acid leaching, dissolution of
carbonates, gypsum precipitation, staining) were recorded during drilling of
monitoring wells installed between and adjacent to the cells during spring
2005
• Total uranium was detected at background levels in bedrock core samples
collected while drilling monitoring wells between and adjacent to the cells
(see Appendix A)
• No contaminants detected in groundwater at levels above natural background
concentrations (INTERA, Inc., 2007a; 2007b; 2008), which is corroborated by
the finding that the groundwater age beneath the tailings cells is dominated by
water that is at least 50 years old (Hurst and Solomon, 2008)
• No contaminants detected in groundwater as evaluated through stable isotopes
(Hurst and Solomon, 2008).
2.1.4 Characteristics of Tailings
The tailings are generally silty sand but heterogeneous due to the placement process.
Based on grain-size analyses performed on the tailings, sand-sized particles are dominant
(57 percent on average) with the remainder being silt- and clay-sized particles. Grain
size distribution data for the White Mesa Mill tailings are compared to data collected at
other uranium mill tailings facilities (see Appendix I). The saturated hydraulic
conductivity of the tailings assumed for White Mesa was based on measured values
reported for the Cotter Corporation’s Canon City Mill tailings impoundment (see
Appendix I). The mill tailings at Canon City are considered to be representative of the
mill tailings at White Mesa because the average grain-size distributions between the two
sites are similar.
The tailings are initially saturated when placed but are dewatered through evaporation
and pumping from the slimes drains system. The solution chemistry of the tailings pore
water, as represented by samples collected from the Cell 2 slimes drain, was assumed to

2-9
be identical for all of the cells (see Appendix K). Tailings pore water in the slimes drains
(i.e., immediately above the tailing cell liners) is considered to be more representative of
solutions that would remain in the tailings cells during operations and at closure given
that these solutions would have had sufficient time to equilibrate with the tailings.
Furthermore, water extracted from the slimes drains, as opposed to samples grabbed from
surface ponds, is not affected as much by evaporation/evapoconcentration and
addition/recirculation of mill process water; evaporation and recirculation of mill process
water would tend to create a variable source-term solution chemistry that is dissimilar to
and not representative of the long-term pore water chemistry in the tailings.
2.2 SITE CHARACTERISTICS
2.2.1 Climate
The climate of the Blanding area is semiarid with average annual precipitation of 13.3
inches (Utah Climate Center, 2007). Most precipitation falls in the form of rain, with
about one-quarter of the precipitation falling as snow. There are two separate rainfall
seasons in the area: a late summer season when monsoonal moisture from the Gulf of
Mexico leads to thunderstorms and a winter season related to fronts from the Pacific.
The average annual Class A pan evaporation rate is 68 inches.
Climatological data are available for the weather station near Blanding, Utah (420738),
located approximately six miles north of the White Mesa Mill at an elevation of 6,040 ft
above mean sea level (ft above MSL). The White Mesa Mill is located at an elevation of
5,600 ft above MSL. Data are available for the period December 1904 through
December 2006; however, large gaps in the dataset (i.e., missing precipitation and/or air-
temperature measurements) occurred during 1905, 1910 to 1912, 1915, 1916, 1917, 1927,
1929, 1931, 1989, and 2005. Data for the period between 1932 and 1988 are nearly
continuous.
The long-term average annual precipitation at the Blanding weather station was
13.3 inches with a standard deviation of 3.9 inches. Annual precipitation for the period
1905 through 2005 is presented in Figure 2-3. The greatest annual precipitation was

2-10
measured in 1909 (24.5 inches), but other years that exceeded 20 inches include 1906
(23.6 inches), 1957 (22.4 inches), 1941 (21.5 inches), 1908 (20.2 inches), 1997
(20.2 inches), and 1965 (20.1 inches). Daily precipitation for the period 1905 through
2005 is presented in Figure 2-4. The largest daily precipitation event was 4.48 inches,
which occurred on 1 August 1968.
The mean annual temperature for Blanding, Utah is 52°F, based on the period 1971-2000.
January is typically the coldest month, with a mean monthly temperature of about 30°F.
July is generally the warmest month, with a mean monthly temperature of 76°F. Daily
ranges in temperatures are typically large.
Winds are generally light to moderate (less than 15 miles per hour) at the site during all
seasons, with winds prevailing from the south. Strong winds are associated with summer
thunderstorms and frontal activity during the late winter and spring.
2.2.2 Summary of Site Geology
The White Mesa Mill is located within the Blanding Basin of the Colorado Plateau
physiographic province. The average elevation at the site is 5,600 ft above MSL. The
site is underlain by unconsolidated alluvium overlying sedimentary bedrock consisting
primarily of sandstone and shale. The unconsolidated deposits are primarily aeolian silt
and sand and range from 1 to 30 ft thick (these deposits have been removed where the
tailings cells are located). The bedrock underlying the site is relatively undeformed and
horizontal (generally dips are less than 3 degrees). Cretaceous Dakota Sandstone and
Burro Canyon Formation are at or near the surface; these sandstone units have a
combined thickness of 100 to 140 ft at the site. Beneath the Burro Canyon Formation is
the Morrison Formation, which is primarily shale. The Brushy Basin Member is the
uppermost member of the Morrison Formation and is composed primarily of bentonitic
mudstones, siltstones, and claystones. The contact between the Burro Canyon Formation
and Brushy Basin Member dips slightly to the south. Beneath the Brushy Basin Member
are the Westwater Canyon, Recapture, and Salt Wash members of the Morrison
Formation. Beneath the Morrison Formation are the Summerville Formation, Entrada

2-11
Sandstone, and Navajo Sandstone. For more detailed descriptions of the geologic setting
see the Reclamation Plan (IUC, 2000; Denison 2009).
2.2.3 Hydrogeology of the Perched Aquifer System
Groundwater beneath the site is first encountered as a perched zone within the Burro
Canyon Formation. The low-permeability Brushy Basin Member of the Morrison
Formation acts as an aquitard and forms the base of the perched aquifer. Monitoring
wells at the site are screened across the saturated portion of the Burro Canyon Formation
and generally extend down to the contact with the Brushy Basin Member. The saturated
thickness of the perched zone ranges from less than 5 to as much as 82 ft beneath the site,
assuming the base of the Burro Canyon Formation is the base of the perched aquifer. The
water table of the perched aquifer was 13 to 116 ft below ground surface (bgs) at the
facility in 2007. The perched water table is shallowest near the wildlife ponds (13 ft in
piezometer P-2), east of the Mill and tailings cells. Groundwater within the perched zone
generally flows south to southwest beneath the site (see Figure 2-5). Recharge to the
perched aquifer is primarily from areal recharge due to infiltration of precipitation and
seepage from the wildlife ponds on the eastern margin of the site. Discharge from the
perched aquifer is believed to be to springs and seeps along Westwater Creek Canyon and
Cottonwood Wash to the west-southwest and along Corral Canyon to the east of the site.
The discharge point located most directly downgradient of the tailings cells is believed to
be Ruin Spring in Westwater Creek Canyon, a tributary to Cottonwood Wash,
approximately two miles from the tailings cells.
The horizontal hydraulic gradient in the perched aquifer downgradient and in the vicinity
of the tailings cells ranges from approximately 0.01 to 0.04 feet per foot (ft/ft) and is
generally to the south and southwest with local variations in magnitude and direction (see
Figure 2-5). Recharge from the wildlife ponds causes localized mounding of the water
table.
The hydraulic conductivity of the perched aquifer (generally within the Burro Canyon
Formation) has been characterized through aquifer pumping tests, slug tests, packer tests,
and laboratory analysis of core samples. Based on tests performed in perched zone

2-12
monitoring wells downgradient of the tailings cells (MW-3, MW-5, MW-11, MW-12,
MW-14, MW-15, MW-17, MW-20, MW-22, and MW-25), the geometric mean
horizontal hydraulic conductivity in this area ranges (based on several analysis methods)
from 0.064 to 0.12 ft/day (2.3 x 10-5 to 4.3 x 10-5 cm/sec) (Hydro Geo Chem, 2009).
Based on these hydraulic conductivities, a porosity of 18 percent, and an average
hydraulic gradient of 0.013 ft/ft, the average linear velocity of groundwater downgradient
of the tailings cells was calculated to be 0.005 to 0.009 ft/day (1.7 to 3.2 ft/year) (Hydro
Geo Chem, 2009).
Beneath and immediately upgradient of the tailing cells, the geometric mean hydraulic
conductivity (based on tests from wells MW-23, MW-25, MW-27, MW-28, MW-29,
MW-30, MW-31, MW-32, TW4-20, TW4-21, and TW4-22) was 0.08 ft/day (3 x 10-5
cm/sec). Based on gradients in the vicinity of each well, the hydraulic conductivity at
each well, and the estimated effective porosity of 18 percent, the geometric mean linear
velocity of groundwater was calculated to be 0.012 ft/day (4.5 ft/year) (Hydro Geo Chem,
2009).
The vertical hydraulic conductivity of the underlying Brushy Basin Member of the
Morrison Formation is significantly lower and demonstrates that it acts as a perching
layer. Cores from the Brushy Basin Member had vertical hydraulic conductivities of 2.1
x 10-7 to 25.4 ft/day (7.3 x 10-11 to 5.9 x 10-4 cm/sec) with a geometric mean of 3.4 x 10-5
ft/day 1.2 x 10-8 cm/sec (IUC, 2000).
2.2.4 Groundwater Quality of the Perched Aquifer System
Groundwater quality in existing and new wells completed in the perched aquifer has been
used to establish background concentrations and determine GWCLs. For additional
details regarding groundwater quality and the determination of GWCLs, see the Revised
Background Groundwater Quality Report: Existing Wells for Denison Mines (USA)
Corp.’s White Mesa Mill Site, San Juan County, Utah (INTERA, Inc., 2007a); Revised
Addendum Evaluation of Available Pre-Operational and Regional Background Data
Background Groundwater Quality Report: Existing Wells (INTERA, Inc., 2007b); and

2-13
Revised Addendum Background Groundwater Quality Report: New Wells (INTERA, Inc.,
2008).
2.2.5 Vadose Zone Hydrogeology and Geochemistry of the Unsaturated Bedrock
The vadose zone is the zone between the ground surface and the perched water table.
The vadose zone beneath the White Mesa Mill is within the unconsolidated deposits
(removed during construction of the tailings cells), the Dakota Sandstone, and Burro
Canyon Formation. The vadose zone thickness was calculated by taking the difference
between the elevation of the cell bottom and the distance to the water table (see
Appendix C). The minimum vadose zone thicknesses beneath Cells 2 & 3 and Cell 4A
are approximately 42 ft and 40 ft, respectively. As a comparison, the average vadose
zone thicknesses beneath Cell 2, Cell 3, and Cell 4A are 63 ft, 66 ft, and 56 ft. For the
vadose zone transport models, the vadose zone thickness beneath Cells 2 & 3 and Cells
4A & 4B was assumed to be 42 ft (12.8 m).
Samples of bedrock from the vadose zone between and immediately adjacent to the
White Mesa Mill tailings cells were collected and characterized for hydraulic and
geochemical properties. The original laboratory reports are included in Appendix A and
B and statistical analyses of the data and identification of hydrogeochemical units are
included in Appendix C. Hydraulic properties are used to predict the flow of water
through the vadose zone, while geochemical properties are used to predict water/rock
chemical reactions as the tailings pore water potentially migrates beneath the tailings
cells. Geochemical properties tested include mass concentrations of hydrous ferric oxide
(HFO) and acid neutralization potential (ANP). The mass of ANP is used in the vadose
zone reactive transport model to predict the consumption of alkalinity (as a neutralization
front) as low-pH tailings pore water potentially migrates beneath the tailings cells, while
the mass of HFO is used to predict surface complexation (adsorption) reactions. Soil
water retention and unsaturated hydraulic conductivity curves are presented and used to
identify hydrologic units, while a statistical analysis of the geochemical data is presented
and used to identify geochemical units. Lithologic data combined with the hydrologic

2-14
and geochemical data form the basis for assigning hydrogeochemical stratigraphic units
within the vadose zone (see Appendix C).
2.3 CONCEPTUAL MODEL OF WATER FLOW (INFILTRATION) THROUGH
THE TAILINGS CELL COVER AND POTENTIAL CONTAMINANT
TRANSPORT THROUGH THE VADOSE ZONE
This section presents the conceptual model for water flow (infiltration) through the
tailings cell cover and potential contaminant transport through the vadose zone. Details
of the implementation of the conceptual model into the numerical model as well as
parameter values, boundary conditions, and initial conditions used in the modeling are
described in detail in Section 3.0. Results of the numerical modeling are presented in
Section 4.0.
2.3.1 Unsaturated Flow
Unsaturated Flow Governing Equation. Unsaturated flow through the vadose zone can
be described with a modified form of the Richards Equation. The Richards Equation is
derived by combining the Darcy-Buckingham equation with the mass continuity
equation. The governing flow equation for one-dimensional vertical isothermal flow of
liquid water (as an incompressible fluid) in a variably saturated rigid porous medium,
assuming that the air phase plays an insignificant role in the liquid flow process, is given
by the following modified form of the Richards Equation (Simunek et al., 2009):
)()]1)(([)(hSz
hhKzt
h −+∂
∂
∂
∂=∂
∂θ
where:
θ = volumetric water content [L3L-3]
h = pressure head of soil water [L]

2-15
S = sink term, volume of water removed from a unit volume of soil per unit
time (e.g., uptake by plants) [T-1]
z = spatial coordinate in the vertical direction [L]
t = time [T]
K = unsaturated hydraulic conductivity [LT-1].
The unsaturated hydraulic conductivity (K) is a function of the volumetric water content
(θ) and pressure head (h), and as a result can vary in both space and time. The pressure
head and volumetric water content may be used interchangeably as the independent
variable. Hydraulic properties of unsaturated porous media (i.e., θ(h) and K(h)) are
nonlinear functions of the pressure head (h), and a solution to the Richards Equation is
commonly solved numerically with a computer program.
Unsaturated Hydraulic Conductivity. To solve the above equation, it is necessary to
specify the relationships of unsaturated hydraulic conductivity (K) versus the effective
water saturation (Se), and of pressure head (h) versus volumetric water content (θ).
The relationship of unsaturated hydraulic conductivity versus effective water saturation,
assuming the pore-size-distribution model presented in Mualem (1976), is described by
the following equation (van Genuchten, 1980):
[]2/1 )1(1)(mm
e
l
es SSKhK−−=
where:
K = unsaturated hydraulic conductivity [LT-1]
Ks = saturated hydraulic conductivity [LT-1]
Se = effective saturation [dimensionless].
l = empirical pore connectivity parameter [dimensionless]
m = empirical shape parameter [dimensionless].

2-16
The effective saturation is equal to:
rs
r
eS θθ
θθ
−
−=
where:
Se = effective saturation [dimensionless]
θ = volumetric water content [L3L-3]
θr = residual volumetric water content [L3L-3]
θs = saturated volumetric water content [L3L-3].
Soil Water Retention. The relationship of pressure head (h) to water content (θ),
assuming the pore-size-distribution model presented in Mualem (1976), is described by
the following equation (van Genuchten, 1980):
()
⎪⎩
⎪⎨
⎧
≥
<+
−+=
0
0]1[
h
hhh
s
mn
rs
r
θ
α
θθθθ

where:
θ = volumetric water content [L3L-3]
θr = residual volumetric water content [L3L-3]
θs = saturated volumetric water content [L3L-3]
h = pressure head of soil water [L]
α = empirical fitting parameter [L-1]
n = empirical fitting parameter [dimensionless]
m = empirical shape parameter [dimensionless].
The fitting parameters (α, n, and m) are considered to be empirical coefficients that affect
the shape of the hydraulic functions used to describe variations in water content and
hydraulic conductivity for different soil water pressures. For unsaturated porous media,

2-17
the pressure head of soil pore water is negative (i.e., less than atmospheric pressure) and
is commonly referred to as matric potential or soil-water tension (negative). The
unsaturated hydraulic conductivity is a function of the saturated hydraulic conductivity,
pressure head, and moisture content. As a result, the unsaturated hydraulic conductivity
in the vadose zone can vary through time. In an unsaturated system, the advective
velocity is largely controlled by variations in soil moisture content because the
unsaturated hydraulic conductivity and effective porosity varies through time as moisture
contents vary. The saturated and unsaturated hydraulic properties are listed in Section
3.0 and Appendix C.
Plant-Water Uptake. The sink term in the Richards Equation is defined as the volume
of water removed from a unit volume of soil per unit time. This accounts for plant-water
uptake and can be defined in terms of soil water pressure head (h) as described by the
following equation (Feddes et al., 1978):
pShzhS)(),(α=
where:
S = sink term, volume of water removed from a unit volume of soil per unit time
(e.g., uptake by plants) [T-1]
α = root water uptake water stress response function [dimensionless]
Sp = potential root water uptake rate [T-1].
The root water uptake water stress response function (α) is a dimensionless function that
ranges between 0 and 1, and is dependent on the soil water pressure head and vegetation
type. For example, when conditions are extremely dry or extremely wet, plants cease to
take up water. A plant root distribution function can also be used to account for variable
plant water uptake with depth. The following equation can be used to describe conditions
that involve spatially variable root density (Simunek et al., 2009):
ppTzbS)(=

2-18
where:
Sp = potential root water uptake rate [T-1]
b = normalized root water uptake distribution (root density) [L-1]
Tp = potential rate of transpiration [LT-1].
The root water uptake distribution is normalized to ensure that b(z) integrates to unity
throughout the rooting depth (Simunek et al., 2009). Spatially variable root density has
been observed for grasses, in which grass roots are usually most dense near the ground
surface and decrease with depth (see Appendix D).
2.3.2 Contaminant Transport in the Unsaturated Zone
Contaminant Transport Governing Equation. Contaminant transport can be described
by the advection-dispersion equation (ADE). The governing equation for unsaturated
zone contaminant transport with advection, dispersion, mineral precipitation/dissolution
reactions, and surface complexation reactions (sorption/retardation) of contaminants is
(Jacques and Simunek, 2005):
i
iiwi Rz
qC -z
CD z t
C −∂
∂⎟⎠
⎞⎜⎝
⎛
∂
∂
∂
∂=∂
∂θθ
where:
θ = volumetric water content [L3L-3]
Ci = aqueous concentration of species [ML-3]
t = elapsed time [T]
z = spatial coordinates in the vertical direction [L]
Dw = hydrodynamic dispersion coefficient in the liquid phase [L2T-1]
q = Darcy flux [LT-1]
Ri = general source/sink term for geochemical reactions [ML-3T-1].

2-19
Hydrodynamic Dispersion. The equation used to describe the hydrodynamic dispersion
coefficient in the liquid phase is given by (Bear, 1972; Simunek et al., 2009):
wwL
w tDqDDθθ+=
where:
θ = volumetric water content [L3L-3]
Dw = hydrodynamic dispersion coefficient in the liquid phase [L2T-1]
Dl = longitudinal dispersivity in the liquid phase [L]
q = Darcy flux [LT-1]
Dw = molecular diffusion coefficient in free water [L2T-1]
tw = tortuosity factor in the liquid phase [dimensionless].
While the equation used to describe the tortuosity factor in the liquid phase is given by
(Millington and Quirk, 1961; Simunek et al., 2009):
2
37
s
wt θ
θ=
where:
tw = tortuosity factor in the liquid phase [dimensionless]
θ = volumetric water content [L3L-3]
θs = saturated volumetric water content [L3L-3].
Dispersion versus Diffusion. The hydrodynamic dispersion coefficient includes the
effects from mechanical dispersion and molecular diffusion. These two processes act to
dilute and spread contamination as it is transported by advection. For saturated systems,
mechanical dispersion tends to dominate over molecular diffusion because advective
velocities are high; as a result, effects due to diffusive transport of mass may be ignored
in a saturated system with high velocities. Conversely, for unsaturated systems,
molecular diffusion tends to dominate over dispersion because advective velocities are

2-20
low; as a result, effects due to mechanical dispersion of mass may be ignored in an
unsaturated system with low velocities (Bear and Verruijt, 1987; Fetter, 1998).
The relative contribution of mechanical dispersion to diffusive transport of mass can be
evaluated by calculating the dimensionless Peclet number within the vadose zone. The
Peclet number is given by the following equation (Bear and Verruijt, 1987; Fetter, 1998):

wD
vdPe=
where:
Pe = Peclet number [dimensionless]
v = velocity [LT-1] [Darcy flux divided by volumetric water content]
d = average grain size diameter of vadose zone material [L]
Dw = molecular diffusion coefficient in free water [L2T-1].
Diffusion tends to dominate over dispersion for Peclet numbers less than 0.4 (Bear and
Verruijt, 1987; Fetter, 1998). For the White Mesa Mill vadose zone, the following
assumptions were considered in calculating a Peclet number: (1) the water flux is
assumed to equal the highest potential water flux rate calculated to migrate through the
tailing cell liners (5 x 10-3 cm/day; see Appendix L); (2) the volumetric water content in
the vadose zone was assumed to be at 75 percent saturation (0.14; see Appendix C); (3)
the average grain size diameter was assumed to equal 0.05 cm, which is characteristic of
fine- to medium-size sand grains (see Appendix C); and (4) the molecular diffusion
coefficient for chloride was assumed to equal 1.75 cm2/day (Li and Gregory, 1974). The
calculated Peclet number, assuming these values as input, is equal to 0.001, which
indicates that diffusive transport of mass is going to dominate, and the effects of
mechanical dispersion can be ignored because of the low mass transport velocities.
Sorption and Retardation. Chemical reactions between dissolved constituents in
vadose zone pore water (e.g., metals and radionuclides) and the vadose zone bedrock
(Dakota Sandstone and Burro Canyon Formation) often dictate spatial and temporal
variations in contaminant-plume transport and mobility in the subsurface by controlling

2-21
the degree of adsorption-desorption of aqueous complexes to surface assemblages and
mineral precipitation-dissolution reactions. The amount of HFO and ANP in the vadose
zone is used to determine surface complexation and acid neutralization reactions (see
Appendix C). The geochemical model and reactive transport modeling of flow and
transport through the vadose zone are described in Appendix M. The geochemical and
solute transport properties are listed in Section 3.0 and Appendix C and M.

(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
#Ç
#Ç
#Ç
#Ç
#Ç
0
MILL SITE
Wildlife Pond
Wildlife Pond
Wildlife Pond
(currently being
permitted)
Cell No. 4B
Cell No. 3
Cell No. 2
Cell No. 1
Cell No. 4A
P5
P4
P3
P2
P1
MW-5
MW-4
MW-3
MW-2
MW-1
MW-32MW-31
MW-30
MW-29
MW-28
MW-27
MW-26
MW-25
MW-24
MW-23
MW-22
MW-21
MW-20
MW-19
MW-18
MW-17
MW-16
MW-15
MW-14
MW-12
MW-11
TW4-22
0 1000
Feet
µEXPLANATION
Test well
Monitoring well
Piezometer
File: Fig 2-1_Denison site map_0310.mxd SLC 22/03/10
Base map adapted from USGS 7.5 Minute
Topographic maps of Black Mesa Butte,
Blanding South, No-Mans Island, and
Big Bench, Utah Quadrangles.
Coordinates are UTM Zone 12,
NAD 1927 meters.
DENISON MINES (USA) CORP.
WHITE MESA MILL
FIGURE 2-1
SITE MAP@A
#Ç
0
9.
3

f
e
e
t
30 MIL PVC FLEXIBLE MEMBRANE LINER UNDERLAY (CRUSHED SANDSTONE)
UNDERLAY (CRUSHED SANDSTONE)
0.5 FT. (MIN)
1 FT.
1 FT.
SLIMES DRAIN COLLECTION SYSTEM (CLEAN SAND WITH PERFORATED PVC PIPE; IN LIMITED AREAS ONLY)
LINER PROTECTIVE BLANKET (NATIVE SILTY-SAND SOIL)
DETAIL 2 - TAILINGS CELL LINER SYSTEM
Cells 2 and 3
0.5 FT. (MIN)
SLIMES DRAINS
GEONET DRAINAGE SYSTEM (PRIMARY LEAK DETECTION)60 MIL HDPE GEOMEMBRANE
60 MIL HDPE GEOMEMBRANEGEOSYNTHETIC CLAY LINER (GCL)
Cells 4A and 4B
VEGETATION (PRIMARILY GRASSES)
TAILINGS
WATER TABLE
DAKOTA
SANDSTONE
BURRO CANYON
FORMATION
SANDY CLAYEY SILT UPPER 2.8 FT. COMPACTED TO 95% STANDARD PROCTOR DRY DENSITY AND LOWER 2.5 FEET PLACED AT 80% STANDARD PROCTOR DRY DENSITY FOR GRADING (PLATFORM FILL) AND RADON ATTENUATION
5.
3

F
T
.

(
M
I
N
)

3
.
5

F
T
.

TOP-SOIL WITH GRAVEL FOR EROSION AND FROST PROTECTION
SANDY CLAYEY SILT PLACED AT 85% STANDARD PROCTOR DRY DENSITY FOR WATER STORAGE, BIOINTRUSION, AND RADON ATTENUATION
DETAIL 1 - TAILINGS COVER DESIGN
0.5 FT. (MIN)
30

f
e
e
t
42

f
e
e
t

m
i
n
i
m
u
m
DENISON MINES (USA) CORP.
WHITE MESA MILL
GENERALIZED CROSS SECTION
WITH MODIFIED COVER DESIGN
FIGURE 2-2
FI
L
E

Fi
g

2
-
2

G
e
n
e
r
a
l
i
z
e
d

m
o
d
i
f
i
e
d

C
r
o
s
s

S
e
c
t
i
o
n
_
0
2
1
0
.
a
i

0
3
/
3
0
/
1
0

S
L
C
Note: Cross section represents minimum separation distance between
tailings and water table based on data from monitoring wells.
0
5
10
15
20
25
1905 1915 1925 1935 1945 1955 1965 1975 1985 1995 2005
Year
An
n
u
a
l

P
r
e
c
i
p
i
t
a
t
i
o
n

(
i
n
c
h
e
s
)
FILE Fig 2-3 Denison Annual precip_Chart_807.ai 08/30/07 SLC
DENISON MINES (USA) CORP.
WHITE MESA MILL
ANNUAL PRECIPITATION AT BLANDING, UTAH
(1905 TO 2005)
FIGURE 2-3Note: years with incomplete data not shown.
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
1905 1915 1925 1935 1945 1955 1965 1975 1985 1995 2005
Year
Pr
e
c
i
p
i
t
a
t
i
o
n

(
i
n
c
h
e
s
)
FILE Fig 2-4 Denison Daily precip_Chart_807.ai 08/30/07 SLC
DENISON MINES (USA) CORP.
WHITE MESA MILL
DAILY PRECIPITATION AT BLANDING, UTAH
(1905 TO 2005)
FIGURE 2-4
(A
(A
(A
(A
(A
(A
(A
(A(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A
(A (A
(A (A
#Ç
#Ç
#Ç
#Ç
#Ç
0
G
G
G
G
G
G
G
G
G
G
G G
G
G
G
G
G
G
G
G G GGG GG
GGGG
G G
GG
G G
G
G
G
G
RuinSpring
MILL SITE
Wildlife Pond
Wildlife Pond
Wildlife Pond
(currently being permitted)
Cell No. 4B
Cell No. 3
Cell No. 2
Cell No. 1
Cell No. 4A
MW-5
MW-4
MW-3
MW-2
MW-1
MW-32
MW-31MW-30
MW-29
MW-28
MW-27
MW-26
MW-25
MW-24
MW-23
MW-22
MW-21
MW-20
MW-19
MW-18
MW-17
MW-16
MW-15 MW-14
MW-12
MW-11
TW4-22
5510
55
20
5500
5490
5530
5480
5470
5
5
6
0
5
5
8
0
5550
5
5
7
0
5460
5
5
4
0
5
5
9
0
5
6
0
0
5610
5
6
2
0
55 9 0
5 6 0 0
5520
5610
0 2000
Feet
µEXPLANATION
Water level contour line, dashed where uncertain
Test well
Monitoring well
Piezometer
Temporary monitoring well
File: Fig 2-5 Piezo surface contours_0310.mxd SLC 03/30/10
Base map adapted from USGS 7.5 Minute
Topographic maps of Black Mesa Butte,
Blanding South, No-Mans Island, and Big Bench, Utah Quadrangles.
Coordinates are UTM Zone 12,
NAD 1927 meters.
DENISON MINES (USA) CORP.
WHITE MESA MILL
FIGURE 2-5
PIEZOMETRIC SURFACE CONTOURS
PERCHED AQUIFER
(DECEMBER 2009)@A
Note: Water level contours from
Hydro Geo Chem, Inc. (2009)
#Ç
0
G

3-1
3.0 METHODOLOGY
This section provides information regarding the conceptual and mathematical models
used to:
• Predict potential infiltration rates through the tailings cell cover
• Predict potential flow and contaminant transport from the tailings cells
through the underlying vadose zone to the perched groundwater.
Detailed descriptions of the modeling effort are provided in the remainder of this section.
The overall modeling approach is described in Section 3.1. The HYDRUS-1D and HP1
codes are described in Section 3.2. The methodology for modeling potential infiltration
through the cover is described in Section 3.3. The methodology for modeling potential
flow and contaminant transport through the bedrock vadose zone to the underlying
perched groundwater is described in Section 3.4.
3.1 OVERALL MODELING APPROACH
Following conceptual-model development, numerical modeling was completed with the
following two basic models (see Figure 3-1 for model domains):
1. Cover Model: Infiltration modeling with HYDRUS-1D of the tailings cell
cover system with daily precipitation and evapotranspiration to estimate
potential long-term average infiltration rates to the tailings.
2. Bedrock Vadose Zone Model: Vadose zone flow and potential contaminant
transport modeling with HP1/HYDRUS-1D through the bedrock vadose zone
to the underlying perched aquifer during the operational phase, dewatering
phase, and 200-year regulatory post-closure steady-state timeframes.
HYDRUS-1D was used to confirm the results for flow and transport of a
conservative solute (chloride) predicted by HP1.

3-2
To evaluate the potential “worst case” for build-up of water in the tailings (“bathtub
effect”), the model-predicted long-term average water flux rate through the tailings cell
cover system was used to estimate the total amount of water entering the tailings during
the 200-year regulatory timeframe. By assuming a completely impermeable liner system
(i.e., no water flow through the liners; all water that infiltrated through the cover was
accumulated in the cells), the total amount of water entering the tailings through the
cover would be accumulated in the cell. By dividing this total water flux by the tailings
porosity, the potential rise in water levels in the tailings was calculated for this worst-case
scenario. Under this scenario, there would be no impacts to groundwater, because no
water would be leaving the cells.
3.2 HYDRUS-1D AND HP1 COMPUTER CODES
The computer code HYDRUS-1D was used to predict potential infiltration through the
proposed tailings cell cover, while the computer code HP1 (HYDRUS-1D coupled with
the geochemical program PHREEQC) was used to predict potential flow and contaminant
transport through the bedrock vadose zone beneath the tailings cells. HYDRUS is a
finite-element model that simulates water flow and solute transport in
unsaturated/saturated porous media, and was developed by the U.S. Salinity Laboratory
in collaboration with the Department of Environmental Sciences at the University of
California at Riverside (Simunek et al., 1998; Simunek et al., 2005; Simunek et al.,
2009). HYDRUS-1D was selected because it is capable of simulating the dominant
processes affecting infiltration and contaminant transport given the semiarid conditions
and multiple hydrogeologic layers that must be simulated at the site. HP1 (Jacques and
Simunek, 2005) is a reactive transport code that combines the infiltration, unsaturated
flow, and multicomponent contaminant transport modeling capabilities of HYDRUS-1D
(Simunek et al., 2005) with the equilibrium geochemical model PHREEQC (Parkhurst
and Appelo, 1999). HP1 was selected because it has the capabilities of HYDRUS-1D,
but can also model geochemical (water-rock) interactions necessary to predict the
transport of nonconservative solutes that may participate in surface complexation
(adsorption) and mineral precipitation/dissolution reactions.

3-3
3.2.1 HYDRUS-1D
The program can be used to analyze water and solute movement in unsaturated, partially-
saturated, or saturated porous media. HYDRUS allows for spatial and temporal variation
in soil properties, allowing for simulation of a heterogeneous soil profile under variably-
saturated, unsteady-flow conditions. HYDRUS can simulate one-dimensional advection,
dispersion, retardation (sorption), and degradation of contaminants. HYDRUS-1D is one
of the few, commercially available, frequently tested models that can simulate both
unsaturated flow and contaminant transport in the vadose zone (including layered
stratigraphy) with a variety of initial and boundary conditions. Consideration of
discontinuities in capillary and unsaturated hydraulic conductivity is very important for
layered systems because travel times and storage of water and contaminants in the vadose
zone is complex (due to potential capillary-barrier effects). The model provides accurate
results when appropriate spatial discretization for the finite-element domain is
established.
HYDRUS has been used to simulate deep percolation beneath final-closure designs for
radioactive-waste management at the Nevada Test Site, flow around nuclear-subsidence
craters at the Nevada Test Site, and influences of a capillary barrier at the Texas low-
level radioactive waste disposal site. A comparison of HYDRUS to other codes
(CHAIN, MULTIMED-DP, FECTUZ, and CHAIN 2D) was prepared by the
U.S. Environmental Protection Agency to evaluate each code’s ability to predict
radionuclide fate and transport in the unsaturated zone (Chen et al., 2002). Of the codes
evaluated by Chen et al. (2002), HYDRUS was the most comprehensive, containing the
greatest number of physical processes. Scanlon et al. (2002) performed a comparison of
codes for simulation of landfill covers in semiarid environments. In addition to
HYDRUS, the evaluation by Scanlon et al. (2002) included the codes HELP, Soil-Cover,
SHAW, SWIM, UNSAT-H, and VS2DT1. This evaluation indicated that Richards-
Equation-based codes such as HYDRUS-1D are more appropriate for simulating near
surface water balance than those using a water-balance approach such as HELP. Only
HYDRUS-1D, SWIM, and VS2DT1 could simulate a seepage face. Of these VS2DT1,
did not simulate the upper atmospheric boundary conditions as well as HYDRUS-1D.

3-4
The HYDRUS-1D program numerically solves the Richards Equation for
saturated/unsaturated water flow and the Fickian-based advection-dispersion equation for
heat and solute transport. HYDRUS-1D incorporates unsaturated soil-hydraulic
properties using the van Genuchten (1980), Brooks and Corey (1964), or modified van
Genuchten-type (Vogel and Cislerova, 1988) analytical functions. The water flow
portion of the model can incorporate (constant or time-varying) prescribed head and flux
boundaries, as well as boundaries controlled by atmospheric conditions. Soil surface
boundary conditions may change during the simulation from prescribed flux to prescribed
head-type conditions. The code also allows for internal sinks such as plant-water uptake.
The solute transport portion of the model can incorporate (constant and time-varying)
prescribed concentration and concentration flux boundaries. The dispersion tensor
includes a term reflecting the effects of molecular diffusion and tortuosity. The transport
equation is coupled to the flow equation through the velocity term.
3.2.2 HP1
The HP1 model was developed by the Belgian Nuclear Research Centre (SCK•CEN) in
collaboration with the U.S. Salinity Laboratory and Department of Environmental
Sciences at the University of California at Riverside. HP1 couples the HYDRUS-1D
variably-saturated water flow and multicomponent contaminant transport model with the
PHREEQC geochemical code. The HP1 code retains all of the features documented in
HYDRUS (as described above) but incorporates additional modules capable of
simulating a broad range of low-temperature geochemical reactions in water, soil, and
groundwater systems. HP1 can simulate multicomponent reactive transport under mixed
equilibrium/kinetic geochemical reactions, including interactions with minerals, gases,
exchangers, and sorption surfaces, based on thermodynamic equilibrium, kinetics, or
mixed equilibrium-kinetic reactions. Neutralization of the infiltrating tailings porewaters,
sorption of solutes, and mineral precipitation/dissolution reactions within the bedrock
vadose zone were determined using HP1.

3-5
3.3 INFILTRATION MODEL OF TAILINGS CELL COVER
3.3.1 Domain
The tailings cell cover model consisted of a one-dimensional conceptual representation of
the planned cover design and was 284 cm (9.3 ft) thick extending from the cover surface
to the top of the tailings (see Figure 2-2 and Figure 3-1).
3.3.2 Finite Element Node Spacing
The finite-element nodes were discretized in the vertical direction to simulate layers in
the tailings cell cover system. Construction of the finite-element mesh is dependent on
surface and bottom boundary conditions and represented heterogeneities due to layering
(Simunek et al., 2009). As a result, node spacing was finer than the tailings cell cover
layers to simulate steep hydraulic gradients which result from transient wetting
(precipitation and infiltration) and drying (evapotranspiration) fronts. Fine-grid spacing
is necessary to accurately simulate water flow (infiltration) through the unsaturated cover
system because hydraulic properties (soil water retention and unsaturated hydraulic
conductivity) may change very rapidly during very short timeframes in a nonlinear
manner. Because hydraulic properties vary much faster and on a finer scale near the land
surface due to rapid changes in atmospheric conditions (daily variations in precipitation
and evapotranspiration were modeled), the node spacing varied between 0.1 and 1 cm
near the top of the cover model domain representing the tailings cell cover system. To
reduce errors due to numerical dispersion, the ratio between neighboring elements did not
exceed 1.5 (Simunek et al., 2009).
3.3.3 Boundary Conditions
An atmospheric upper boundary condition was applied across the top of the model
representing the tailings cell cover to simulate meteorological conditions and was a
function of precipitation and potential evapotranspiration, as described in the paragraphs
that follow. Free drainage (i.e., unit gradient) was assumed for the lower boundary
condition of the model representing the tailings cell cover. Because of the one-

3-6
dimensional nature of the model, the sides of the domain are implicitly assumed to be
zero-flux boundaries.
Atmospheric Boundary Condition. Daily precipitation and air-temperature
measurements were obtained for the Blanding weather station and used as inputs to the
model to determine boundary conditions (Utah Climate Center, 2007). Given the flat
nature of the cover (0.2 percent slope), no runon- or runoff-based processes were
assumed to occur. As a result, precipitation applied to the upper boundary was removed
through evaporation or transpiration, retained in the soil profile as storage, or transmitted
downward as infiltration (potential recharge or drainage to the tailings). The 57-year
period between 1932 and 1988 was selected for use in the vadose zone model because it
contained (see Figure 2-3 and Figure 2-4):
• a nearly continuous time series
• a mixture of the largest annual and daily precipitation events
• consecutive wet years.
The third and fourth wettest years on record (1957 and 1941; 22.4 and 21.5 inches,
respectively) are within the time series selected, and are approximately 9% and 14% less
than the maximum annual precipitation of 24.5 inches recorded during 1909. The largest
daily precipitation event of 4.48 inches, which occurred on 1 August 1968, is represented
in the time series selected. Also, the climate record included the period 1978-1987,
which is a 10-year timeframe characterized by above-average amounts of annual (15.2
inches) and winter precipitation. Increased precipitation was modeled as part of the
sensitivity analysis.
Some interpolation was necessary to construct a continuous time series between 1932 and
1988. Missing precipitation measurements were left blank but accounted for only a small
subset of the population (10 days out of 20,820 days). Air-temperature measurements
were interpolated between missing data points, but overall accounted for a small subset
(55 days out of 20,820 days) of the time series.

3-7
A combination temperature-based and solar-radiation-based approach was used to
calculate daily potential evapotranspirative (PET) fluxes using the Hargreave’s Equation.
This approach was selected because long-term meteorological data (e.g., wind speed)
were not available, and the Hargreave’s Equation can be used as a substitute for the
Penman-Monteith Equation. The calculations assume a hypothetical grass reference crop
with sufficient access to water such that the amount of PET is controlled by site-specific
climatic conditions (e.g., air temperature, day of year, solar declination). PET was
calculated for each day from measured maximum and minimum air temperatures in
addition to estimated radiative fluxes following the methodology outlined in the work of
Allen et al. (1998). The average annual PET between 1932 and 1988 was 46.5 inches.
Potential evaporation (PE) from the soil surface and potential transpiration (PT) from
roots were partitioned from the PET assuming 40% vegetative cover; and the winter
months that included December, January, and February were assigned a transpiration rate
of zero, and only evaporation was simulated in the HYDRUS-1D model (see
Appendix D). The percent of vegetative cover was varied as part of the sensitivity
analysis.
Transpiration. Root water uptake will vary as a function of the soil water pressure head
within the rooting zone, the normalized plant root distribution function (i.e., density of
roots), and the rate of potential transpiration. The rate of potential transpiration is
assigned as part of the atmospheric upper boundary condition for the cover model, which
HYDRUS then uses to compute the actual transpiration (AT) as a function of time and
space within the rooting zone. For example, when conditions are extremely dry (i.e., less
than the wilting point) or extremely wet (i.e., near saturation) plants cease to uptake
water, and the actual transpiration would be zero. At intermediate soil water conditions,
the actual transpiration would be a fraction of the potential transpiration. The water
stress response function for grass was selected from the default database in HYDRUS.
The database does not distinguish between different species of grass.
The model assumed an anticipated scenario with a maximum rooting depth of 107-cm
and an anticipated root density distribution (see Appendix D details). For grasses, roots
are usually denser near the ground surface and decrease with depth. A nonlinear decrease

3-8
in root density with depth was assumed for the root-water-uptake function (i.e., assumes
vegetation removes more water near the ground surface and less with depth). The
maximum rooting depth and root density distribution was varied as part of the sensitivity
analysis. The root-water-uptake function is a dimensionless number proportional to the
root distribution or root density. The Feddes et al. (1978) water-uptake model with
water-response functions for grass was selected from the default database in HYDRUS.
The database does not distinguish between different species of grass. The assumed
wilting point pressure was varied as part of the sensitivity analysis.
Evaporation. The rate of potential evaporation is also assigned as part of the
atmospheric input file. In HYDRUS, the potential evaporation rate is reduced to an
actual evaporation (AE) rate if a specified pressure head is reached at the surface. The
pressure head at which this occurs is controlled by equilibrium conditions between soil
water and atmospheric water vapor. The model assumed a minimum surface pressure
head of -15,000 cm, which is the recommended value by the program. When the pressure
head at the surface reaches -15,000 cm the program calculates a reduced, actual
evaporation rate. The minimum surface pressure head at the upper boundary was varied
as part of the sensitivity analysis.
Climate Record. The 57-year climate record comprised of measured precipitation and
calculated potential evaporation and potential transpiration was repeated to establish a
synthetic atmospheric record for greater durations. Generation of a concatenated
atmospheric record assumes that historic meteorological conditions are considered
representative for the future.
3.3.4 Input Parameters
Hydraulic properties required for the cover model include vertical saturated hydraulic
conductivity, residual soil water content, saturated soil water content, and the soil water
retention empirical curve-fitting parameters. Unsaturated hydraulic properties for the
tailings cell cover materials were estimated using grain-size and geotechnical data for
these materials with the soil-properties database in HYDRUS (details regarding
derivation of hydraulic properties are provided in Appendix E). Hydraulic properties

3-9
used in the model are presented in Table 3-1. The van Genuchten-Mualem single-
porosity soil hydraulic property model was selected to characterize the soil hydraulic
properties.
3.3.5 Initial Conditions
Initial conditions for the infiltration model were determined by evaluating a long-term
simulation that used the concatenated atmospheric input file as an upper boundary
condition (i.e., the 57-year climate record repeated twice). The pressure head distribution
for the final time step of the 114-year simulation was used as the initial condition for the
transient simulations used to predict water infiltration rates through the cover. The
methodology implemented to establish the initial conditions for the site is a commonly
accepted approach for solving hydrogeologic modeling problems. For all HYDRUS-1D
simulations, initial conditions were prescribed as pressure heads (as opposed to water
content) to facilitate model convergence.
3.3.6 Duration of Simulations and Time Steps
Climatological data for the 57-year period 1932 through 1988 were repeated to generate
the necessary duration of input data. Climatic data were input on a daily basis for the
tailings cell cover model.
The minimum and maximum time-step lengths were 1 x 10-6 day (0.09 seconds) and 0.5
days for the HYDRUS-1D models. The maximum number of iterations per time step was
40. In HYDRUS-1D, solution efficiency is maximized by incorporating adaptive time-
step adjustments based on criteria described in Simunek et al. (2009).
3.3.7 Sensitivity Analysis
To test the importance of simulating reduced performance of the vegetative component of
the cover system, and how increased precipitation could influence the transport of water
through the monolithic ET cover, the HYDRUS-1D infiltration model was run using
different assumptions aimed at characterizing a lower bound, base case, and upper bound

3-10
scenario. Rates of model-predicted water flux entering the tailings cells were compared
between simulations using different input assumptions. The effects on moisture content
by the parameters used to assess establishment of vegetation and root water uptake were
also evaluated to determine whether moisture contents that were input into the radon
model are conservative. Input variables incorporated into the sensitivity analysis
included percent vegetative cover, maximum rooting depth and root density, wilting point
pressure head, minimum surface pressure head, and precipitation. A complete
description regarding the sensitivity analysis and comparison of infiltration rates through
the cover based on cover vegetation, biointrusion, and precipitation is included in
Appendix G.
3.4 FLOW AND CONTAMINANT TRANSPORT MODEL OF THE BEDROCK
VADOSE ZONE
Solute transport models were developed for the bedrock vadose zone beneath Cell 1
(contingency cell identified for the potential disposal of decommissioning and
deconstruction debris), Cells 2 & 3, and Cells 4A & 4B.
3.4.1 Domain
The bedrock vadose zone model extended from the base of the tailings cell liner systems
through the Dakota Sandstone and Burro Canyon Formation to the perched water table
surface (see Figure 3-1). The vadose zone thickness was calculated by taking the
difference between the bottom elevation of the cell and the distance to the water table for
individual monitoring wells. The minimum vadose zone thickness beneath Cells 2 & 3
and Cell 4A was approximately 12.8 m (42 ft) and 12.2 m (40 ft), respectively (based on
2007 water level data). As a comparison, the average vadose zone thickness beneath Cell
2, Cell 3, and Cell 4A are 19.2 m (63 ft), 20.1 m (66 ft), and 17.1 m (56 ft). A minimum
vadose zone thickness of 12.8 meters (42 feet) was assumed for all of the simulations of
solute transport beneath the cells (see Appendix C for a discussion of vadose zone
thicknesses and a summary table of vadose zone thickness beneath the tailings cells).

3-11
3.4.2 Finite Element Node Spacing
The finite-element nodes were discretized in the vertical direction to simulate layers in
the bedrock vadose zone. The bedrock vadose zone model had a uniform node spacing of
5 cm. In order to reduce numerical errors due to discretization, grid spacing was based
on recommendations provided by Jacques et al. (2006).
3.4.3 Boundary Conditions
Variable specified mass flux rates (flux multiplied by the concentration) were applied to
the upper boundary of the bedrock vadose zone.
Cell 1. For Cell 1, which was assumed to contain no saturated materials, the average
long-term water flux rate through the ET cover (predicted with the infiltration model)
was used as an upper boundary condition (constant flux) to represent the post-closure
steady-state period. The bottom of Cell 1 (if constructed) will contain a soil liner
compacted to achieve low permeability, but this layer was not included in the modeling,
which yields conservative estimates of chloride transport through the vadose zone.
Cells 2 & 3. Potential water flux rates through the liner systems for Cells 2 & 3 were
calculated using the Giroud-Bonaparte Equation (Giroud and Bonaparte, 1989; Giroud, et
al., 1992) as described in Appendix L. The predicted saturated thickness of the tailings
during the operational phase, during active dewatering, and during post-closure steady
state was used in the Giroud-Bonaparte Equation to calculate the potential flux rate
through the liner for use as an upper boundary condition in the flow and contaminant
transport model of the bedrock vadose zone. Groundwater flow modeling with
MODFLOW of Cells 2 & 3 was performed to estimate tailings-dewatering rates through
time and average water levels (saturated thickness) that will remain in the tailings after
dewatering (described in Appendix J). In addition to the maximum saturated thickness of
the tailings during operations, the number of potential liner defects, and their impacts on
potential water flux through the liners, were evaluated as part of the sensitivity analysis
(see Appendix L for details).

3-12
Cells 4A & 4B. Potential flux through the liner for Cells 4A & 4B was calculated using
the Giroud-Bonaparte Equation as described in Geosyntec Consultants (2006a; 2007a).
Dewatering predictions for Cells 4A & 4B used in the Giroud-Bonaparte Equation to
calculate the potential flux through the liner for Cells 4A & 4B were from Geosyntec
Consultants (2006a; 2007a).
The average solute concentrations used as inputs represented the source term solution
chemistry of the tailings pore water and were also varied as part of the sensitivity analysis
(see Appendix K for a discussion of source term chemistries). The lower boundary at the
base of the domain was assumed to be fully saturated (i.e., water table conditions with a
constant pressure head equal to 0 cm [atmospheric pressure]), representing the water-
table surface of the perched aquifer. A zero concentration gradient was specified at the
lower boundary for solute transport. Because of the one-dimensional nature of the
model, the sides of the domain are implicitly assumed to be zero-flux boundaries.
3.4.4 Input Parameters
Water Flow. Hydraulic properties required for the vadose zone flow model include
vertical saturated hydraulic conductivity, residual soil water content, saturated soil water
content, and the soil water retention empirical curve-fitting parameters. The saturated
and unsaturated hydraulic properties were measured for cores from the Dakota Sandstone
and Burro Canyon Formation (see Appendix B for original laboratory report). Bedrock
core sample collection methodologies, presentation of soil water retention and
unsaturated hydraulic conductivity curves, and selection of hydrologic units are discussed
in Appendix C. Hydraulic properties used in the model are presented in Table 3-1.
The vadose zone model assumed a single set of hydraulic properties consistent with the
test results reported for the Dakota Sandstone. This assumption is considered appropriate
because the saturated and unsaturated hydraulic properties of the samples are quite
similar to one another (see Appendix C). Assignment of a single set of hydrogeologic
properties should not significantly affect the model results given the similarity in
unsaturated hydraulic properties [θ(h)] and [K(h)] for all samples (i.e., there were no

3-13
large differences in soil water retention curves or unsaturated hydraulic conductivity
curves for the materials tested). The hydraulic properties (and dry bulk density) from
MW-23 (55.5-56.0 ft) were used as input to the model because the hydraulic functions
are intermediate as compared to the other samples. Unsaturated hydraulic conductivity of
the vadose zone was not included in the sensitivity analysis because the unsaturated
hydraulic conductivities vary to match flux rates under a unit hydraulic gradient.
Contaminants Modeled. The contaminants modeled included pH, major cations and
anions necessary to achieve charge balance (aluminum, calcium, carbonate, chloride,
magnesium, potassium, sodium, and sulfate), and selected trace elements (arsenic,
cadmium, copper, iron, nickel, uranium, vanadium, and zinc). Trace elements included
in the model were based on their elevated concentrations in the tailings slimes drains as
compared to the GWCLs. Aluminum was included and used to obtain charge balance.
These solutes are the most dependable indicators of site water quality and of potential cell
failure due to their predominance (uranium and sulfate) and predominance and mobility
(chloride). In particular, chloride will migrate unretarded and act as a conservative tracer
and thus would be expected to be detected before all other site contaminants. Uranium
was included because it is one of the primary contaminants of concern.
Source Term Concentrations. The average solute concentrations were used as input to
represent the source term solution chemistry of the tailings pore water (see Appendix K
for a discussion of source term chemistries). The average concentration of chloride,
sulfate, and uranium were 3,221 milligrams per liter (mg/L), 62,847 mg/L, and 24.3
mg/L, respectively. No source degradation, treatment, or dilution was assumed: that is
concentrations were held constant through time. As part of the sensitivity analysis, the
initial solute concentrations were varied: with the maximum reported values used for an
upper bound and the mean minus one-half standard deviation used for the lower bound.
Geochemistry. Geochemical properties of the vadose zone included the amount of acid
neutralization potential (ANP) and mass of hydrous ferric oxide (HFO) present in the
bedrock vadose zone. The amount of ANP and HFO were based on measured values
obtained from core. The sampling methodology, results, and statistical analysis of the

3-14
data, in addition to a discussion regarding the selection of hydrogeochemical units, are
summarized in Appendix C while the original laboratory data are contained in Appendix
A. As part of the sensitivity analysis, the amount of ANP was varied with the geometric
mean plus one geometric standard deviation used for an upper bound and the geometric
mean minus one geometric standard deviation used for the lower bound. The amount of
HFO did not vary significantly within the bedrock vadose zone and was not included in
the sensitivity analysis.
The partial pressure of oxygen was fixed in the model assuming a dissolved oxygen
concentration in vadose zone porewater equal to 2 mg/L. The partial pressure of carbon
dioxide was fixed in the model assuming 10-2.0 atmospheres of pressure, but was varied
as part of the sensitivity analysis to 10-1.0 atmospheres of pressure used for an upper
bound and 10-3.0 atmospheres of pressure used for the lower bound. Redox conditions
were controlled by the oxygen couple. The following minerals were allowed to
participate or dissolve, depending on their saturation indicies: gypsum, calcite (ANP),
amorphous aluminum hydroxide, and amorphous iron hydroxide (ferrihydrite or HFO).
The mass of HFO allowed to participate in surface complexation reactions was fixed
according to measured values in bedrock (geometric mean), and HFO that precipitated
from solution did not add to the available sorption sites. The number of available
sorption sites was based on assumptions implicit with the Dzombak and Morel (1990)
diffuse layer model database. Additional details regarding the geochemical and reactive
transport model are summarized in Appendix M.
Diffusion. Tortuosity, and its effect on molecular diffusion, was explicitly modeled
during contaminant transport modeling by incorporation of a tortuosity factor for the
liquid phase (Simunek et al., 2009). Given the extremely low advective velocity,
mechanical dispersion was assumed to be negligible relative to molecular diffusion (see
Section 2.0).
Degradation and Production. No degradation or production of chloride, sulfate,
uranium, or other trace elements was assumed. Radioactive decay of uranium is
considered to be relatively minor due to the slow processes involved (e.g., the half-life

3-15
for natural uranium, which is predominantly U-238 [~99.3 %], is 4.5 x 109 years).
Although uranium and other trace elements can be removed from solution through
microbial processes, to yield more conservative model predictions, these processes were
not simulated.
3.4.5 Initial Conditions
Water Flow. Initial soil water pressure heads within the bedrock vadose zone were
estimated by applying a constant flux boundary using ~1% of average annual amount of
precipitation. For all HP1 simulations, initial conditions were prescribed as pressure
heads (as opposed to water content) to facilitate model convergence.
Geochemistry. Solution concentrations in the bedrock vadose were estimated by
assuming equilibrium of calcite with the HFO surface. Only calcium and carbonate were
included as aqueous species. The modeling assumed no initial concentrations of other
solutes in the vadose zone for simplicity.
3.4.6 Duration of Simulations and Time Steps
Simulations were run to evaluate solute transport during the operational phase,
dewatering phase, and post-closure steady-state timeframes equal to a total duration
simulation of 240 years. The operational and dewatering phases (see Appendix L for
details) were followed by 200 years following closure as required by the Permit.
The minimum and maximum time-step lengths were 1.04 x 10-2 day (900 seconds) and
180 days for the HP1 model. The maximum number of iterations per time step was 40.
In HP1, solution efficiency is maximized by incorporating adaptive time-step adjustments
based on criteria described in Simunek et al. (2009).
3.4.7 Sensitivity Analysis
A sensitivity analysis was performed to quantify the model-prediction uncertainty due to
estimating solute transport input parameters. Three values were selected for each input

3-16
parameter, representing three different scenarios corresponding to a lower bound, base
case, and upper bound. Input variables incorporated into the sensitivity analysis included
source term solution chemistry of the tailings (see Appendix K for details), maximum
tailings saturated thickness during operations (see Appendix J for details), number of
potential liner defects (see Appendix L for details), ANP of the bedrock vadose zone (see
Appendix C for details), and partial pressure of carbon dioxide gas within the bedrock
vadose zone (see Appendix M for details). Because neither tailings saturated thickness
nor the number of potential liner defects are explicitly simulated in the bedrock vadose
zone model, variability in these parameters was incorporated in the sensitivity analysis by
varying the potential flux rates through the liner.
TABLE 3-1
SATURATED AND UNSATURATED HYDRAULIC PROPERTIES
OF THE WHITE MESA MILL TAILINGS CELL COVER INFILTRATION MODEL AND BEDROCK VADOSE ZONE CONTAMINANT TRANSPORT MODEL
Residual Soil Saturated Soil Saturated Hydraulic
Model Purpose Thickness Thickness Water Content Water Content Conductivity
Layer z (ft) z (cm)θr (-)θs (-)α (cm-1)n (-)Ks (cm/d)
Monolithic ET Cover
1 erosion protection and frost penetration 0.5 15 0.045 0.254 0.0145 1.406 5.6
2 water storage, biointrusion, and radon attenuation 3.5 107 0.055 0.404 0.0145 1.406 7.4
3 grading and radon attenuation 2.8 86 0.046 0.334 0.0229 1.261 3.6
4 grading and radon attenuation 2.5 76 0.059 0.439 0.0125 1.461 10.4
Bedrock Vadose Zone
1 -42 1280 0.003 0.184 0.0103 1.386 9.37
Note: Derivation of hydraulic properties for the cover are described in Appendix E. Derivation of hydraulic properties of the bedrock vadose zone are described in Appendix C.
Curve Fitting Parameters in the Soil Water
Retention Function
284 cm
(9.3 feet)
914 cm
(30 feet)
1280 cm
(42 feet)
Perched Aquifer
HY
D
R
U
S
-
1
D

M
O
D
E
L

O
F

TA
I
L
I
N
G

C
E
L
L

C
O
V
E
R
HP
1
/
H
Y
D
R
U
S
-
I
D

M
O
D
E
L

O
F

BE
D
R
O
C
K

V
A
D
O
S
E

Z
O
N
E
Free Drainage Lower Boundary Condition
Variable Specified Flux Upper Boundary Condition
Liner
Contaminant Concentrations
Fixed Pressure Head Lower Boundary Condition
3,221 mg/L chloride
62,847 mg/L sulfate
24.3 mg/L uranium
Water table (pressure head = 0 cm [atmospheric])
Potential Flux Rate through Tailing Cell Liner Calculated
with the Giroud-Bonparte Equation
Atmospheric Upper Boundary Condition
Precipitation Evapotranspiration
TAILINGS
(UNSATURATED)
TAILINGS
(SATURATED)
DAKOTA
SANDSTONE
BURRO CANYON
FORMATION
DENISON MINES (USA) CORP.
WHITE MESA MILL
MODELING APPROACH,
MODEL DOMAIN
AND BOUNDARY CONDITIONS
FIGURE 3-1
FI
L
E

Fi
g

3
-
1

M
o
d
e
l

D
o
m
a
i
n
_
B
C
S
_
0
3
1
0
.
a
i

0
3
/
2
3
/
1
0

S
L
C

4-1
4.0 RESULTS
This section presents the results from the tailings cell cover infiltration model and
bedrock vadose zone flow and contaminant transport model. The HYDRUS-1D
infiltration model was used to predict potential water fluxes through the tailings cell
cover system, the results of which are presented in Section 4.1. The HP1 bedrock vadose
zone contaminant transport model was used to predict the potential flow and transport of
conservative (chloride) and nonconservative (sulfate, uranium, and other trace elements)
solutes through the bedrock vadose zone to the perched aquifer, the results of which are
presented in Section 4.2. A sensitivity analysis was performed to evaluate the impacts
that uncertainty in parameter input values have on model results. The results of the
sensitivity analysis for the tailings cell cover infiltration model and bedrock vadose zone
contaminant transport model are described in each section separately. Key modeling
assumptions and model uncertainty are discussed in Section 4.3. For all HYDRUS-1D
and HP1 simulations the water and mass balance errors did not exceed 1%. As a general
rule-of-thumb, mass balance errors that do not exceed 3% are considered acceptable.
4.1 TAILINGS CELL COVER SYSTEM INFILTRATION MODELING
The HYDRUS-1D infiltration model was used to predict potential water fluxes through
the proposed monolithic ET cover system assuming atmospheric boundary conditions and
a cover design as presented in Figure 2-2. The construction of a monolithic ET cover is
proposed as part of this ICTM report to cap the entirety of all tailings cells. As described
in Section 3.0, the model did not include runoff and 100% of the precipitation was
allowed to evaporate or infiltrate into the top layer of the cover; no runoff was assumed to
occur. Water not removed through evapotranspiration, or stored within the cover system,
was transported downward as drainage that could potentially recharge the tailings cells.
The modeling approach, model domain, and boundary conditions are described in
Figure 3-1.

4-2
4.1.1 Model-Predicted Water Flux Rate
The model-predicted water flux rate through the tailings cell cover during the anticipated
57-year climate record (between 1932 and 1988) is shown on Figure 4-1. The model-
predicted water flux rate varies during the 57-year period from a minimum rate of
0.17 millimeters per year (mm/yr) to a maximum rate of 1.1 mm/yr, with an average
long-term flux rate through the cover system of 0.45 mm/yr. The average long-term
water flux rate corresponds to approximately 0.1% of the average annual amount of
precipitation recorded at the Blanding weather station. The average long-term infiltration
rate predicted to enter the top of the tailings cells was reduced from 34 mm/yr for the
currently permitted rock cover design to 0.45 mm/yr for the proposed monolithic ET
cover design (see Appendix E for details of this comparison). The increased performance
and reduction of infiltration for the proposed ET cover, relative to the original rock cover
design, is attributed to the presence of vegetation and associated root water uptake via
transpiration.
The model-predicted water flux rate through the monolithic ET cover indicates that the
available storage capacity of the cover should be sufficient to significantly reduce
infiltration, and the ET cover should function properly as designed. The transport of
water below the rooting zone and into the tailings material would occur when the storage
capacity of the overlying soil materials is exceeded; for example, during multi-
consecutive years, or longer, that receive above average amounts of annual or winter
precipitation. Breakthrough of water through the bottom of the cover (i.e., drainage),
beginning at about year 48 (see Figure 4-1), results from the occurrence of three
consecutive years that received above average amounts of winter precipitation followed
by another seven years that received above average amounts of annual precipitation.
The model-predicted infiltration rates for the monolithic ET cover are consistent with
data reported from lysimeter and infiltration modeling studies of other vegetated ET
covers (e.g., Albright et al., 2004; Bolen et al., 2001; Fayer and Gee, 2006; Gee et al
1994; Scanlon et al., 2005). Furthermore, results from the nearby uranium mill tailings
lysimeter at Monticello (Waugh et al., 2008) also agree with model predictions for the

4-3
proposed cover system at White Mesa. For example, the average infiltration rate
measured at Monticello during a seven year monitoring period was approximately
0.6 mm/yr, which corresponds to approximately 0.2% of the average annual amount of
precipitation (recorded at the Monticello weather station) occurring as recharge. At
Monticello, nearly all of the measured drainage occurred during an exceptionally wet
winter-spring season and no drainage was measured during the first four years of the
monitoring period (Waugh et al., 2008). Overall, the comparison between model
predictions for the proposed ET cover system at White Mesa, with results reported for
lysimeter studies constructed with similar cover system designs and located in similar
semiarid climates, suggests that only a fraction of the average annual amount of
precipitation (e.g., less than 0.5%) can be expected to occur as recharge to the underlying
tailings.
In summary, a monolithic ET cover is the preferred design to minimize infiltration
necessary to meet the Permit (Part I.D.8) and meet the radon attenuation standard. The
material thicknesses for the different cover layers were based on the results of radon
attenuation modeling to achieve the State of Utah’s long-term radon emanation standard
for uranium mill tailings (Utah Administrative Code, Rule 313-24). The results of
modeling the emanation of radon-222 from the top surface of the monolithic ET cover are
presented in Appendix H. Furthermore, the proposed cover design will be sufficient to
provide adequate thickness to protect against frost penetration and biointrusion, provide
adequate water storage capacity to minimize the rate of infiltration into the underlying
tailings, and provide long-term moisture within the cover to attenuate radon flux.
4.1.2 Evaluation of Build-up of Water in Tailings
To evaluate the potential for build-up of water in the tailings (“bathtub effect”), the long-
term average water flux rate through the tailings cell cover system (predicted with the
infiltration model) was used to calculate the amount of water entering the tailings during
the 200-year regulatory timeframe specified by the Permit. The amount of water
expected to migrate through the cover and enter the tailings cells (i.e., assuming all
recharge to the tailings can act to increase the amount of head on the liner) was then used

4-4
to calculate the maximum potential rise in water levels in the tailings assuming no water
flow through the liners (i.e., all water that infiltrated through the cover was accumulated
in the cells). The assumptions for evaluating the “bathtub effect” result in an end-
member scenario expected to produce a conservative estimate of closed-cell cover system
performance.
The amount of water calculated to enter the tailings after 200 years is equal to 90
millimeters (0.3 feet) of water. Assuming a tailings porosity of 57%, the calculated
water-level rise on the liner is approximately 160 millimeters (0.53 feet). Consequently,
a significant build-up of water (“bathtub effect”) within the cells is not anticipated.
Therefore, the proposed cover design will prevent the accumulation of leachate head
within the tailings that could rise above or over-top the maximum liner elevation (which
is typically greater than 20 feet above the bottom of the cell), meeting the requirement of
the Permit (Part I.D.8).
4.1.3 Sensitivity Analysis
To test the importance of simulating reduced performance of the vegetative component of
the cover system, and how increased precipitation could influence the transport of water
through the monolithic ET cover, the HYDRUS-1D infiltration model was run using
different assumptions to evaluate the effects of lower bound, and upper bound scenarios,
in addition to the base case. Rates of model-predicted water flux entering the tailings
cells were compared between simulations using different input assumptions. The effects
on moisture content by the parameters used to assess establishment of vegetation and root
water uptake were also evaluated to determine whether moisture contents input into the
radon model are conservative. Input variables incorporated into the sensitivity analysis
included percent vegetative cover, maximum rooting depth and root density, wilting point
pressure head, and precipitation. A complete description regarding the sensitivity
analysis and comparison of infiltration rates through the cover based on cover vegetation,
biointrusion, and precipitation is included in Appendix G.
The lower bound, anticipated, and upper bound long-term average water flux rates
entering the tailings were 0.19 mm/yr, 0.45 mm/yr (see Section 4.1.1), and 2.4 mm/yr,

4-5
respectively, which would result in a total of approximately 38 mm (0.12 ft), 90 mm (0.3
ft), and 480 mm (1.6 ft) of water entering the tailings during the 200-year regulatory
timeframe, respectively, and corresponding to an increase in saturated tailings thickness
of 67 mm (0.21 ft), 160 mm (0.53 ft), and 840 mm (2.8 ft), respectively.
The results of the sensitivity analysis demonstrate that the design and construction of a
monolithic ET cover will be sufficient to minimize infiltration into the tailings and
prevent the formation of a bathtub effect for a broad range of conditions used to represent
the establishment of vegetation, root water uptake by vegetation, and amount of
precipitation that may occur at the site, thereby meeting closed cell performance
requirements specified in the Permit (Part I.D.8.a and Part I.D.8.b). The results of the
sensitivity analysis for the broad range of conditions described above, also demonstrate
that the monolithic ET cover will have sufficient long-term moisture to attenuate radon
fluxes thereby achieving the State of Utah’s long-term radon emanation standard for
uranium mill tailings (Utah Administrative Code, Rule 313-24). Overall, all of the
simulations, including the simulation with the lowest average flux rates, demonstrate that
the amount of moisture predicted with the infiltration model exceeds the amount of
moisture used in the radon attenuation model, which indicates that the predictions of
radon emanation at the surface are conservative.
4.2 BEDROCK VADOSE ZONE FLOW AND CONTAMINANT TRANSPORT
MODELING
The bedrock vadose zone flow and contaminant transport model was used to predict
potential flow rates and contaminant transport rates through the bedrock vadose zone to
the perched aquifer during the operational, dewatering, and post-closure steady-state
timeframes. The modeling approach, model domain, and boundary conditions were
described in Figure 3-1. Solute transport models were developed for the bedrock vadose
zone beneath Cell 1 (contingency cell identified for the potential disposal of
decommissioning and deconstruction debris), Cells 2 & 3, and Cells 4A & 4B. For
simplicity, a vadose zone thickness of 12.8 meters (42 feet) was assumed for all of the
simulations (see Appendix C for discussion of vadose zone thicknesses). This is a

4-6
conservative assumption given that the average vadose zone thicknesses beneath Cell 2,
Cell 3, and Cell 4A are 19.2 m (63 ft), 20.1 m (66 ft), and 17.1 m (56 ft). HP1 was used
to simulate potential solute transport of conservative (chloride) and nonconservative
(sulfate, uranium, and other trace element) solutes through the vadose zone beneath the
tailings cells. Conservative and nonconservative solute transport was predicted for the
vadose zone beneath Cells 2 & 3. Because of the difference in potential mass transport
rates (discussed below), only the transport of conservative solutes (chloride) was
predicted for Cells 4A & 4B and Cell 1.
Potential water flux rates through the primary liner installed beneath Cells 2 & 3 and the
secondary liner installed beneath Cells 4A & 4B were calculated using the Giroud-
Bonaparte Equation (see Appendix L for equations used and list of assumptions).
Conservative estimates of potential water flux rates through the liners were used as an
upper boundary condition (time-dependent flux) for the HP1 model used to predict flow
and solute transport through the vadose zone to the perched aquifer during the
operational, dewatering, and post-closure steady-state timeframes. The average long-
term water flux rate through the ET cover (predicted with the infiltration model) was used
as an upper boundary condition (constant flux) for Cell 1 to represent the post-closure
steady-state period. The bottom of Cell 1 (if constructed) will contain a soil liner
compacted to achieve low permeability, but this layer was not included in the modeling,
which yields conservative estimates of transport through the vadose zone.
The calculated potential water flux rates through the liners were multiplied by the
average solute concentrations to yield a time-dependent mass flux rate applied as an
upper boundary condition. The average solute concentrations were used as input to
represent the source term solution chemistry of the tailings pore water (see Appendix K
for a discussion of source term chemistries).
4.2.1 Cells 2 & 3
Head on Liner. The head above the single liner beneath Cells 2 & 3 was used as input to
calculate the potential rate of fluid migration through the liners into the underlying
vadose zone. For Cells 2 & 3, operational data (see Appendix L for details) and

4-7
predictions with the MODFLOW dewatering model (see Appendix J for details) were
used to estimate the saturated thickness of the tailings through time. For simplicity, the
average operational period for Cells 2 & 3 (23 years) was used in the flux calculations.
For modeling purposes, the head on the liner was assumed to increase linearly for 13
years from approximately zero to fully saturated conditions. The average saturated
thickness across the entirety of the cell when the cell was entirely full, 5.82 m (19.1 ft),
was used as the maximum head during operations. Then the cells were assumed to
remain fully saturated for an additional 10 years for a total operational period of 23 years,
at which point active dewatering was assumed to be initiated. The MODFLOW
dewatering model predicted that, as a result of dewatering activities, that the tailings
would draindown nonlinearly through time reaching an average saturated thickness of
1.07 m (3.5 ft) after 10 years (i.e., total operational phase plus dewatering phase equal to
33 years). The saturated tailings thickness (head on liner) assumed in the model during
the operational, dewatering, and post-closure steady-state timeframes is plotted in
Figure 4-2.
Model-Predicted Water Flux Rate. The model-predicted water flux rate at the bottom
of the vadose zone (immediately above the perched aquifer) at various times during the
operational, dewatering, and post-closure steady-state timeframes are plotted on Figure 4-
3. The potential flux rate predicted to enter the perched aquifer closely resembles the
flux rate assumed to enter the top of the bedrock vadose zone with some minor
differences that result from equilibration of initial conditions and storage of moisture
within the bedrock. The flux rate predicted at the water table reaches a maximum value
of approximately 7.5 mm/yr (compared to the 8.3 mm/yr maximum flux rate applied at
the upper boundary) after 25 years of operations. The flux rate then rapidly declines in
response to decreased head (saturated thickness) that occur in the tailings during the
dewatering phase (see Figure 4-2), ultimately reaching a long-term steady state value of
approximately 0.7 mm/yr (see Figure 4-3).
Model-Predicted Volumetric Moisture Content. The model-predicted volumetric
water content throughout the vadose zone beneath Cells 2 & 3 during the operational,
dewatering, and post-closure steady-state timeframes is plotted in Figure 4-4. The

4-8
volumetric water contents represent synoptic timeframes after 13 years (beginning of
maximum head conditions in Cells 2 & 3), 23 years (at the end of maximum head
conditions), 33 years (at the end of dewatering of Cells 2 & 3), and 100 & 240 years
(post-closure steady-state). The volumetric water content profiles after 100 and 240
years are identical, indicating that steady state flow conditions have developed in the
bedrock vadose zone.
Model-Predicted Chloride Concentration. The model-predicted chloride concentration
at the bottom of the vadose zone beneath Cells 2 & 3 after 240 years (including
operational, dewatering, and post-closure periods) of transport is 0.01 mg/L. The
chloride concentration at the bottom of the vadose zone represents the model-predicted
addition of chloride as a result of the potential flux from the tailings cells. While there is
naturally-occurring chloride in the vadose zone initially, the modeling assumed no initial
chloride for simplicity, and because there is a lack of data concerning background
chloride and distribution of chloride within the vadose zone. Furthermore, the predicted
chloride concentration is the solute concentration in vadose zone pore water that will
reach the perched aquifer; however, the predicted concentration is not equal to the
concentration in groundwater. The chloride mass flux (water flux multiplied by the
concentration) entering the perched aquifer system will mix with groundwater resulting
in a reduced concentration compared to the value predicted to occur at the bottom of the
vadose zone. A model was not constructed to determine the actual (diluted)
concentration in groundwater because the chloride concentration predicted at the bottom
of the vadose zone was orders of magnitude less than the minimum GWCL, which is 10
mg/L for chloride. The minimum GWCL (for chloride and all other solutes modeled)
was selected from the list of monitoring wells located immediately downgradient from
the tailings cells (i.e., monitoring wells MW-5, MW-11, MW-12, MW-14, MW-15, MW-
23, MW-24, MW-28, MW-29, MW-30, and MW-31; GWCL’s for these wells are
specified in the Permit) (see Figure 2-5). Chloride transport predicted with HP1 was
confirmed with HYDRUS-1D.
Model-Predicted Sulfate Concentration. The model-predicted sulfate concentration at
the bottom of the vadose zone beneath Cells 2 & 3 after 240 years of transport is

4-9
0.014 mg/L. The distribution of sulfate within the bedrock vadose zone is controlled by
the amount of gypsum that may precipitate from solution. The sulfate concentration at
the bottom of the vadose zone represents the model-predicted addition of sulfate as a
result of the potential flux from the tailings cells. While there is naturally-occurring
sulfate in the vadose zone initially, the modeling assumed no initial sulfate for simplicity.
Furthermore, the predicted sulfate concentration is the solute concentration in vadose
zone pore water that will reach the perched aquifer; however, the predicted concentration
is not equal to the concentration in groundwater, which will be diluted. The sulfate mass
flux entering the perched aquifer system will mix with groundwater resulting in a reduced
concentration compared to the value predicted to occur at the bottom of the vadose zone.
A model was not constructed to determine the actual (diluted) concentration in
groundwater because the sulfate concentration predicted at the bottom of the vadose zone
was orders of magnitude less than the minimum GWCL, which is 532 mg/L.
Model-Predicted Uranium Concentration. Uranium does not reach the bottom of the
vadose zone beneath Cells 2 & 3 during the 240-year transport timeframe. Adsorption of
uranium onto the surface of HFO present in the bedrock vadose zone limits the transport
distance below the liner. The depth at which the uranium concentration is approximately
equal to the minimum GWCL (0.0049 mg/L) is 2.3 meters (8 feet) below the liner
system; a minimum of 10.5 meters (34 feet) above the perched water table. The uranium
concentration within the vadose zone represents the model-predicted addition of uranium
as a result of the potential flux from the tailings cells. While there is naturally-occurring
uranium in the vadose zone initially, the modeling assumed no initial uranium for
simplicity.
Model-Predicted Concentration of Other Trace Elements. The sorption of uranium
was competitive because additional trace elements were modeled. Solutes included in the
model were based on their elevated concentrations in the tailings pore water as compared
to the GWCLs. Transport of the following trace elements was modeled: arsenic,
cadmium, copper, nickel, vanadium, and zinc. Similar to uranium, these solutes were
predicted to migrate a limited distance below the liner (e.g., a few meters).

4-10
4.2.2 Cells 4A & 4B
Head on Secondary Liner. The head within the leak detection system above the
secondary liner beneath Cells 4A & 4B was used as input to calculate the potential rate of
fluid migration through the secondary liner into the underlying vadose zone. For Cells
4A & 4B, operational data (see Appendix L for details) and dewatering predictions
(Geosyntec Consultants, 2007a; 2007c) were used to determine the length of the
operational and dewatering timeframes. For simplicity, the average operational period (6
years) and dewatering period (6 years) for Cells 4A & 4B was used in the flux
calculations. The maximum head on the secondary liner during operations is assumed to
equal 0.004 m (0.01 ft) for Cells 4A & 4B (Geosyntec Consultants, 2007a; 2007c).
Significantly reduced head on the secondary liner for Cells 4A & 4B, as compared to
Cells 2 & 3, is due to a more extensive slimes drain collection system, the upper primary
liner, and pumping of the leak detection system, thus reducing the head on the secondary
liner. The maximum head on the secondary liner was assumed to remain constant
throughout the operational and dewatering periods (total of 12 years). The actual head on
the secondary liner during the majority of the operational and dewatering periods is
expected to be less than 0.004 m.
Model-Predicted Water Flux Rate. The calculated flux of water through the secondary
liner beneath Cells 4A & 4B for the maximum head within the leak detection system
during the operational and dewatering periods is approximately 8 x 10-5 mm/yr. The
potential flux rates predicted at the end of dewatering are assumed to equal the rate
during post-closure steady state because the increase in water levels is anticipated to be
minor (see Section 4.1). Therefore, the model-predicted water flux rate at the bottom of
the vadose zone (immediately above the perched aquifer) during post-closure steady-state
is 8 x 10-5 mm/yr.
Model-Predicted Volumetric Moisture Content. The model-predicted volumetric
water contents throughout the vadose zone beneath Cells 4A & 4B at various times
throughout the operational, dewatering, and post-closure periods are nearly identical to

4-11
the long-term steady-state profile (at 100 and 240 years) beneath Cells 2 & 3 as shown on
Figure 4-4.
Model-Predicted Chloride Concentration. For all practical purposes, chloride is not
predicted to reach the bottom of the vadose zone during the 12-year operational and 200-
year post-closure periods (chloride concentration predicted to reach the water table at 212
years was 5 x 10-14 mg/L). The chloride concentration is not predicted to exceed the 10
mg/L minimum GWCL anywhere in the vadose zone because of the diminutive chloride
mass flux rate entering the vadose zone.
Model-Predicted Concentration of Other Solutes. Considering that chloride is a
conservative tracer, and that transport is not affected by sorption or mineral precipitation
reactions, coupled with the fact that the model predictions demonstrate nearly zero
impact, additional model predictions of solute transport for nonconservative contaminants
(sulfate, uranium, other trace elements) was not considered necessary.
4.2.3 Cell 1
Head on Compacted Soil Liner. The bottom of Cell 1 (if constructed) will contain a
soil liner compacted to achieve low permeability. The compacted soil liner was not
included in the bedrock vadose zone model as a simplification. Furthermore, if necessary
Cell 1 would be constructed to contain demolition debris generated during
decommissioning and deconstruction (D & D) of the mill, and all debris would be dry in
an unsaturated state (negative soil water pressure head). Therefore, the build-up of head
on the compacted soil liner is not anticipated.
Model-Predicted Water Flux Rate. If Cell 1 is constructed for D & D disposal, it will
be covered with the monolithic ET cover design. The design will be the same as the
cover proposed for the other cells (see Figure 2-2). Consequently, the long-term average
infiltration rate would be equivalent to the value presented for the other cells (see Section
4.1 and Figure 4-1). Therefore, the model-predicted water flux rate at the bottom of the
vadose zone (immediately above the perched aquifer) during the 200-year post-closure
steady-state is predicted to be approximately 0.5 mm/yr.

4-12
Model-Predicted Volumetric Moisture Content. The model-predicted volumetric
water content throughout the vadose zone during long-term steady conditions beneath
Cell 1 is nearly identical to the long-term steady state profile beneath Cells 2 & 3 (at 100
and 240 years) plotted in Figure 4-4.
Model-Predicted Chloride Concentration. The source term of the D & D debris is
assumed to equal the concentrations assigned to the tailings pore water, which is
anticipated to lead to conservative predictions that over predict the potential impacts. For
all practical purposes, chloride is not predicted to reach the bottom of the vadose zone
during the 200-year transport timeframe (chloride concentration predicted to reach the
water table at 200 years was 7 x 10-9 mg/L). The bedrock vadose zone depth at which the
chloride concentration is approximately equal to the minimum GWCL (10 mg/L) is 4.65
meters (15.3 feet), approximately 8.1 meters (26.7 feet) above the water table.
Model-Predicted Concentration of Other Solutes. Considering that chloride is a
conservative tracer, and that transport is not affected by sorption or mineral precipitation
reactions, coupled with the diminutive transport distance, additional model predictions of
solute transport for nonconservative contaminants (sulfate, uranium, other trace elements)
was not considered necessary.
4.2.4 Evaluation of Closed-Cell Cover System Performance (Potential Impacts to
Groundwater)
To evaluate the potential impacts to groundwater as a result of closed cell cover system
performance, all infiltration migrating through the cover to the tailings was assumed to
pass through the tailings and the liner system into the underlying bedrock vadose zone.
Model predictions of chloride, sulfate, uranium, and other trace elements (arsenic,
cadmium, copper, nickel, vanadium, and zinc) at the bottom of the vadose zone do not
exceed the GWCLs for monitoring wells located immediately downgradient from the
tailings cells (i.e., MW-5, MW-11, MW-12, MW-14, MW-15, MW-23, MW-24, MW-28,
MW-29, MW-30, and MW-31) (see Figure 2-5). Therefore, the proposed cover design
will be protective of groundwater quality; contaminant concentrations are not predicted to

4-13
exceed the GWCS’s or GWCL’s at the compliance monitoring wells specified in the
Permit, thus demonstrating compliance with the Permit (Part I.D.8).
4.2.5 Sensitivity Analysis
A sensitivity analysis was performed to quantify the model-prediction uncertainty due to
estimating model input parameters. Three values were selected for each input parameter,
representing three different scenarios corresponding to a lower bound, base case, and
upper bound. Input variables incorporated into the sensitivity analysis included source
term solution chemistry of the tailings pore water (see Appendix K for details), maximum
tailings saturated thickness during operations (see Appendix J for details), number of
potential liner defects (see Appendix L for details), acid neutralization potential of the
bedrock (ANP) vadose zone (see Appendix C for details), and partial pressure of carbon
dioxide gas within the bedrock vadose zone (see Appendix M for details). A complete
description regarding the sensitivity analysis and results for the bedrock zone
contaminant transport modeling is included in Appendix M. Based on the results for
conservative transport of chloride (i.e., limited transport distance) within the bedrock
vadose zone beneath Cells 4A & 4B and Cell 1, the sensitivity analysis was only
evaluated for Cells 2 & 3.
For the transport of conservative solutes, rates of model-predicted chloride concentrations
at the bottom of the vadose zone (entering the perched aquifer) are presented; for
nonconservative solutes, rates of model-predicted sulfate concentrations at the bottom of
the vadose zone (entering the perched aquifer) are presented, while for uranium, bedrock
vadose zone depths at which uranium concentrations approximately equal the minimum
GWCL are presented. Results between simulations using different input assumptions are
compared to evaluate the effect of parameter uncertainty on predictions of contaminant
transport through the bedrock vadose zone.
Chloride. The model-predicted chloride concentrations at the bottom of the bedrock
vadose zone beneath Cells 2 & 3 after 240 years of transport are summarized in Table 4-1
for the seven model simulations. The sensitivity analysis assessing the range in chloride
concentrations predicted to occur at the bottom of the bedrock vadose zone considered a

4-14
range in solute concentrations, maximum tailings saturated thickness, and number of
potential liner defects. The input variables including ANP and partial pressure of carbon
dioxide gas within the bedrock vadose zone were not included because these parameters
would only affect the transport of nonconservative solutes. The results for the base case
scenario were described in Section 4.2.1. The upper bound model-predicted chloride
concentration at the bottom of the vadose zone was 18 mg/L, which is slightly greater
than the minimum GWCL of 10 mg/L. However, mixing of vadose zone pore water with
groundwater in the perched aquifer system would dilute this concentration below the
minimum GWCL. The lower bound model-predicted chloride concentration at the
bottom of the vadose zone was essentially zero (9.1 x 10-6 mg/L). Assuming all other
variables are equal, the model-predicted chloride concentrations are least sensitive to the
source term chemistry and most sensitive to the number of potential liner defects (which
affects the potential liner flux rate), while the maximum tailing saturated thickness during
operations has an intermediate effect (see Table 4-1, response variable statistic column).
Sulfate. The model-predicted sulfate concentrations at the bottom of the bedrock vadose
zone beneath Cells 2 & 3 after 240 years of transport are summarized in Table 4-2 for the
nine model simulations. The sensitivity analysis assessing the range in sulfate
concentrations predicted to occur at the bottom of the bedrock vadose zone considered a
range in solute concentrations, number of potential liner defects, ANP within the bedrock
vadose zone, and partial pressure of carbon dioxide gas within the bedrock vadose zone.
Based on the results for chloride transport discussed above, the maximum tailings
saturated thickness was excluded from the sensitivity analysis assessing nonconservative
solute transport. The results for the base case scenario were described in Section 4.2.1.
The upper bound model-predicted sulfate concentration at the bottom of the vadose zone
was 45 mg/L, which is less than the minimum GWCL of 532 mg/L. The lower bound
model-predicted sulfate concentration at the bottom of the vadose zone was essentially
zero (1.0 x 10-5 mg/L). The distribution of sulfate within the bedrock vadose zone is
controlled by the amount of gypsum that may precipitate from solution (see
Appendix M). Assuming all other variables are equal, the model-predicted sulfate
concentrations are least sensitive to the ANP of the bedrock vadose zone and most
sensitive to the number of potential liner defects (which affects the potential liner flux

4-15
rate), while the source term chemistry and partial pressure of carbon dioxide gas within
the bedrock vadose zone have an intermediate effect (see Table 4-2, response variable
statistic column).
Uranium. The model-predicted bedrock vadose zone depth at which the uranium
concentration approximately equals the minimum GWCL (0.0049 mg/L) after 240 years
of transport beneath Cells 2 & 3 is summarized in Table 4-3 for the nine model
simulations. The sensitivity analysis assessing the range in uranium concentrations
predicted to occur within the bedrock vadose zone used the same input variables as for
sulfate. The upper bound model-predicted depth at which uranium approximately
equaled the minimum GWCL was 3.9 meters. The base case scenario was described in
Section 4.2.1. The lower bound model-predicted depth at which uranium approximately
equaled the minimum GWCL was 1.3 meters. None of the sensitivity runs predicted that
uranium, or other trace elements (arsenic, cadmium, copper, nickel, vanadium, and zinc;
see Appendix M) would reach the perched aquifer in the 240 year period simulated.
Assuming all other variables are equal, the model-predicted uranium transport depths are
least sensitive to the source term chemistry and most sensitive to the number of potential
liner defects, while the ANP and partial pressure of carbon dioxide gas within the
bedrock vadose zone have an intermediate effect (see Table 4-3, response variable
statistic column). The distribution of uranium is controlled by sorption onto the surfaces
of HFO within the bedrock vadose zone. Factors that influence the transport and sorption
potential of uranium are discussed in Appendix M.
For the broad range of input values of different geochemical variables, model-predicted
concentrations of chloride, sulfate, uranium, and other trace elements (arsenic, cadmium,
copper, nickel, vanadium, and zinc; see Appendix M) in the perched aquifer are not
predicted to exceed the GWCLs for monitoring wells located immediately downgradient
from the tailings cells (i.e., MW-5, MW-11, MW-12, MW-14, MW-15, MW-23, MW-24,
MW-28, MW-29, MW-30, and MW-31) (see Figure 2-5). Therefore, the proposed cover
design will be protective of groundwater quality; contaminant concentrations are not

4-16
predicted to exceed the GWCSs or GWCLs at the compliance monitoring wells specified
in the Permit, thus meeting the permit requirements (Part I.C.1 and Table 2; Part I.D.8.c).
4.3 UNCERTAINTY AND ASSUMPTIONS
The numerical modeling presented in this report was based on fundamental biological,
ecological, physical, and geochemical assumptions concerning the mechanisms
controlling infiltration through the tailings cell cover model and contaminant transport in
the bedrock vadose zone. However, as with all numerical models, the model only
replicates the actual physical system to the extent that it is based on an accurate
conceptual model that describes the site hydrogeology, boundary conditions, and initial
conditions. The goal of the conceptual model is to describe these conditions (e.g.,
vegetation, stratigraphy, hydraulic properties, transport mechanisms, and boundary
conditions) with a sufficient level of detail to address the objectives of the study.
Because the subsurface environment is heterogeneous, simplifying assumptions are
required so that the characteristics of the system can be quantified and incorporated into
the numerical model.
Some of the simplifications include assuming the bedrock vadose zone thickness is equal
to the minimum separation distance between the bottom of the tailings cells and the top
of the perched aquifer. In the model, the vadose zone (distance between the liner beneath
the cells and the perched aquifer water table) was assumed to be 12.8 meters (42 feet) for
all of the cells. This vadose-zone thickness is the minimum depth to the water table
(measured in nearby monitoring wells), which only occurs in one small area. The depth
from the bottom of the cells to the perched aquifer water table is up to 27 meters (90 feet)
in some areas. The assumption of a minimum bedrock vadose zone thickness is
conservative as it results in shorter travel times for contamination to reach the water
table. Actual travel times are likely to be much greater than predicted, particularly for
transport beneath the western half of Cells 2 & 3 where the vadose zone thickness is
much greater than 12.8 meters.
There is considerable evidence that the cells are not leaking. However, potential flux
rates through the tailings cell liner systems were calculated using empirical equations and

4-17
assumptions developed for landfills. In reality, the tailings may limit the transmission of
water, thus actual flow rates for a given hole size would be less than the calculated flow
rates through the liners. Leakage rates computed for a tailings facility are expected to be
less than the measured leakage rates for landfills because tailings are likely to have a
limited capacity to transmit all available water. Consolidation of fine-grained tailings
and deposition of tailing slimes, coupled with the chemical nature of the pore water (e.g.,
precipitation of gypsum), is anticipated to essentially seal some of the defects, which
would act to decrease the potential flux rates through the liners.
A significant amount of gypsum, amorphous aluminum hydroxide, and amorphous iron
hydroxide (ferrihydrite or HFO) was predicted to precipitate within the shallow bedrock
vadose zone, which would be expected to modify liquid phase saturation and effective
porosities, resulting in decreased water flux rates. It is likely that a layer of mineral
precipitates would act to seal any holes in the liners, further reducing contaminant
transport mobilities and transport distances.
Leakage from the unlined wildlife ponds have resulted in significant impacts to the
perched water table surface (see Figure 2-5), which is not evident beneath the tailings
cells. Considering the significantly larger footprint of the tailings cells, compared to the
wildlife ponds, combined with the fact that the tailings cells have contained tailings at
nearly fully-saturated conditions for long periods of time, if leakage were significant, it is
likely that evidence would have appeared at this point.
The vadose-zone model assumed no lateral flow, only vertical flow. This ignores the
impacts that horizontal heterogeneities may have on migration in the vadose zone.
Because there is little information concerning vadose-zone heterogeneities, a two- or
three-dimensional model was not constructed. However, given that hydraulic gradients
in the vadose zone are strongly vertical, flow is primarily vertical, and thus a one-
dimensional model is adequate for vadose zone flow and transport.
The bedrock vadose zone flow component of the contaminant transport model cannot be
calibrated because there are no moisture content or pressure head data available for the
vadose zone. Quantifying moisture fluxes through desert vadose zones is very difficult

4-18
due to the small magnitude of fluxes and the very long response times (Walvoord et al.,
2002). The bedrock vadose zone modeling does not account for vapor transport. Under
natural conditions, water transport in thick desert vadose zones is dominated by upward
vapor transport over very long time periods (Walvoord et al., 2002). Modeling
performed by Walvoord et al. (2002) indicates that most thick desert vadose zones are in
a slow drying process that is on the order of tens of thousands of years. Upward vapor
transport would act to slow downward contaminant migration.
The assumptions used to construct the numerical models to predict infiltration through
the cover and potential impacts to the perched groundwater system, generally were either
conservative or based on anticipated conditions. As a result, the predictions are
considered to be conservative.
TABLE 4-1
MODEL-PREDICTED CHLORIDE CONCENTRATIONS
AT THE BOTTOM OF THE BEDROCK VADOSE ZONE AFTER 240 YEARS FOR CELLS 2 AND 3 EVALUATED AS PART OF THE SENSITIVITY ANALYSIS
Response Variable Evaluated Response Variable Statistic
Chloride Concentration at the Bottom Change in Chloride Concentration at the Bottom
of the Bedrock Vadose Zone at 240 yr of the Bedrock Vadose Zone at 240 yr
(mg/L)(mg/L)
1 Base Case Base Case Base Case 0.0096 0
2 Upper Bound Base Case Base Case 0.012 0.0024
3 Lower Bound Base Case Base Case 0.0087 -0.00090
4 Base Case Upper Bound Base Case 0.25 0.24
5 Base Case Lower Bound Base Case 0.00030 -0.0093
6 Base Case Base Case Upper Bound 18 18
7 Base Case Base Case Lower Bound 0.0000091 -0.010
a Model run 1 is the base case scenario.
b The base case assumed the mean concentration, while the upper bound assumed the maximum concentration and the lower bound assumed the mean minus one-half standard deviation.
c The base case assumed the average saturated thickness, while the upper bound assumed the maximum saturated thickness and the lower bound assumed the the average value minus the
difference between the upper bound and base case saturated thicknesses.
d The base case assumed one small hole and one large hole defect per acre, while the upper bound assumed one small hole and three large hole defects per acre and the lower bound assumed
one small hole defect per acre.
Input Parameter Varied
Model Runa Chloride
Concentrationb
Maximum Tailings
Saturated Thickness During
Operationsc
Number of Potential
Liner Defectsd
TABLE 4-2
MODEL-PREDICTED SULFATE CONCENTRATIONS
AT THE BOTTOM OF THE BEDROCK VADOSE ZONE AFTER 240 YEARS OF TRANSPORT FOR CELLS 2 AND 3 EVALUATED AS PART OF THE SENSITIVITY ANALYSIS
Response Variable Evaluated Response Variable Statistic
Sulfate Concentration at the Bottom Change in Sulfate Concentration at the Bottom
of the Bedrock Vadose Zone at 240 yr of the Bedrock Vadose Zone at 240 yr
(mg/L)(mg/L)
1 Base Case Base Case Base Case Base Case 0.014 0
2 Upper Bound Base Case Base Case Base Case 0.017 0.00303Lower Bound Base Case Base Case Base Case 0.012 -0.0020
4 Base Case Upper Bound Base Case Base Case 0.034 0.020
5 Base Case Lower Bound Base Case Base Case 0.0085 -0.00556Base Case Base Case Upper Bound Base Case 45 45
7 Base Case Base Case Lower Bound Base Case 0.000010 -0.014
8 Base Case Base Case Base Case Upper Bound 0.014 0
9 Base Case Base Case Base Case Lower Bound 0.015 0.0010
a Model run 1 is the base case scenario.
b The base case assumed the mean concentration, while the upper bound assumed the maximum concentration and the lower bound assumed the mean minus one-half standard deviation.
c The base case assumed a partial pressure of 10-2.0 atmospheres, while the upper bound assumed a partial pressure of 10-1.0 atmospheres and the lower bound assumed a partial pressure of 10-3.0 atmospheres.
d The base case assumed one small hole and one large hole defect per acre, while the upper bound assumed one small hole and three large hole defects per acre and the lower bound assumed one small hole defect per acre.
e The base case assumed the geometric mean, while the upper bound assumed the geometric mean plus one geometric standard deviation and the lower bound assumed the geometric mean minus one geometric standard deviation.
Input Parameter Varied
Model Runa Sulfate
Concentrationb
Partial Pressure of Carbon
Dioxide Gasc
Number of Potential Liner
Defectsd
Acid Neutralization
Potential (ANP)e
TABLE 4-3
MODEL-PREDICTED DEPTH WITHIN THE BEDROCK VADOSE ZONE AT WHICH URANIUM CONCENTRATION IS EQUAL TO THE MINIMUM GROUNDWATER COMPLIANCE LIMT AFTER 240 YEARS FOR CELLS 2 AND 3
EVALUATED AS PART OF THE SENSITIVITY ANALYSIS
Response Variable Evaluated Response Variable Statistic
Bedrock Vadose Zone Depth at Which Uranium Change in Depth at Which Uranium Concentration
Concentration Equals the Minimum GWCL at 240 yr Equals the Minimum GWCL at 240 yr
(meters)(meters)
1 Base Case Base Case Base Case Base Case 2.30 0
2 Upper Bound Base Case Base Case Base Case 2.50 0.20
3 Lower Bound Base Case Base Case Base Case 2.15 -0.15
4 Base Case Upper Bound Base Case Base Case 3.90 1.60
5 Base Case Lower Bound Base Case Base Case 2.15 -0.15
6 Base Case Base Case Upper Bound Base Case 3.70 1.40
7 Base Case Base Case Lower Bound Base Case 1.30 -1.00
8 Base Case Base Case Base Case Upper Bound 2.20 -0.10
9 Base Case Base Case Base Case Lower Bound 1.55 -0.75
GWCL = groundwater compliance limit
a Model run 1 is the base case scenario.
b The base case assumed the mean concentration, while the upper bound assumed the maximum concentration and the lower bound assumed the mean minus one-half standard deviation.
c The base case assumed a partial pressure of 10-2.0 atmospheres, while the upper bound assumed a partial pressure of 10-1.0 atmospheres and the lower bound assumed a partial pressure of 10-3.0 atmospheres.
d The base case assumed one small hole and one large hole defect per acre, while the upper bound assumed one small hole and three large hole defects per acre and the lower bound assumed one small hole defect per acre.
e The base case assumed the geometric mean, while the upper bound assumed the geometric mean plus one geometric standard deviation and the lower bound assumed the geometric mean minus one geometric standard deviation.
Input Parameter Varied
Model Runa Uranium
Concentrationb
Partial Pressure of Carbon
Dioxide Gasc
Number of Potential Liner
Defectsd
Acid Neutralization
Potential (ANP)e
Time (years)
Wa
t
e
r

F
l
u
x

R
a
t
e

(
m
m
/
y
e
a
r
)
0.0
1.0
2.0
3.0
0 102030405060
DENISON MINES (USA) CORP.
WHITE MESA MILL
MODEL-PREDICTED WATER FLUX RATE
THROUGH TAILINGS CELL COVER
(TYPICAL 57-YEAR PERIOD)
FIGURE 4-1
File: Fig 4-1 Water Flux Rate_0310.ai 03/23/10 SLC
Time (years)
Ta
i
l
i
n
g
s

S
a
t
u
r
a
t
e
d

T
h
i
c
k
n
es
s

(
m
)
0
2
4
6
8
0 100 150 20050 250
DENISON MINES (USA) CORP.
WHITE MESA MILL
TAILINGS SATURATED THICKNESS (HEAD ON LINER) DURING
THE OPERATIONAL, DEWATERING, AND POST-CLOSURE
STEADY-STATE TIMEFRAMES FOR CELLS 2 & 3
FIGURE 4-2
File: Fig 4-2 Tailings Saturated Thickness_0310.ai 03/23/2010 SLC
Time (years)
Wa
t
e
r

F
l
u
x

R
a
t
e

(
m
m
/
y
e
a
r
)
0
2
4
6
8
10
0 50 100 150 250200
DENISON MINES (USA) CORP.
WHITE MESA MILL
MODEL-PREDICTED WATER FLUX RATE AT THE BOTTOM OF
THE BEDROCK VADOSE ZONE (ENTERING THE PERCHED AQUIFER)
DURING THE OPERATIONAL, DEWATERING, AND POST-CLOSURE
STEADY-STATE TIMEFRAMES FOR CELLS 2 & 3
FIGURE 4-3
File: Fig 4-3 Model-predicted water flux rate_0310.ai 03/23/2010 SLC
Volumetric Water Content (-)
Be
d
r
o
c
k

V
a
d
o
s
e

Z
o
n
e

D
e
p
t
h

B
e
n
e
a
t
h

L
i
n
e
r

S
y
s
t
e
m

(
c
m
)
1500
1000
500
0
0.0 0.05 0.10 0.15 0.20
23 Years (beginning of maximum water level conditions during operational period)
Water Table(fully saturated conditions)
Top of Bedrock Vadose Zone Beneath Tailings Cell Liner System
33 Years (end of dewatering period)13 Years (beginning of maximum water level conditions during operational period)
100 & 240 Years (post-closure long-term steady-state conditions)
DENISON MINES (USA) CORP.
WHITE MESA MILL
MODEL-PREDICTED VOLUMETRIC WATER CONTENT
THROUGHOUT THE BEDROCK VADOSE ZONE
DURING THE OPERATIONAL, DEWATERING, AND
POST-CLOSURE STEADY-STATE TIMEFRAMES FOR CELLS 2 & 3
FIGURE 4-4
File: Fig 4-4 Volumetric Water Content_0310 03/30/2010 SLC

5-1
5.0 CONCLUSIONS AND POST-AUDIT MONITORING PLAN
This section summarizes the results of infiltration and contaminant transport modeling
performed to support Denison’s Ground Water Discharge Permit for the White Mesa
uranium milling and tailings disposal facility and provides recommendations for a post-
audit monitoring plan.
5.1 CONCLUSIONS
The HYDRUS-1D infiltration model was used to predict potential water fluxes through
the tailings cell cover system. The HP1 bedrock vadose zone contaminant transport
model was used to predict the potential flow and transport of conservative (chloride) and
nonconservative (sulfate, uranium, and other trace elements) solutes through the bedrock
vadose zone to the perched aquifer. Sensitivity analyses were performed to evaluate the
impacts that uncertainty in parameter input values have on model results.
5.1.1 Model-Predicted Water Flux Rate for Tailings Cell Cover System
The model-predicted average long-term water flux rate through the proposed monolithic
ET tailings cell cover, assuming a historical climate record (based on climatic data
recorded between 1932 and 1988), was 0.45 mm/yr. The average long-term water flux
rate corresponds to approximately 0.1% of the average annual amount of precipitation
recorded at the Blanding weather station. This is in contrast to an average long-term
infiltration rate of 34 mm/yr predicted for the currently permitted rock cover design. The
increased performance, and reduction of infiltration for the ET cover relative to the
original rock cover design, is attributed to the presence of vegetation and associated root
water uptake via transpiration. The model-predicted water flux rate through the
monolithic ET cover indicates that the available storage capacity of the cover should be
sufficient to significantly reduce infiltration, and the ET cover should function properly
as designed.

5-2
A monolithic ET cover is the preferred design to minimize infiltration necessary to meet
the Permit requirements (Part I.D.8) and meet the radon attenuation standard. The
material thicknesses for the different cover layers were based on the results of radon
attenuation modeling to achieve the State of Utah’s long-term radon emanation standard
for uranium mill tailings (Utah Administrative Code, Rule 313-24). Furthermore, the
proposed cover design will be sufficient to provide adequate thickness to protect against
frost penetration and biointrusion, provide adequate water storage capacity to minimize
the rate of infiltration into the underlying tailings, and provide long-term moisture within
the cover to attenuate radon flux.
5.1.2 Evaluation of Build-up of Water in Tailings
To evaluate the potential for build-up of water in the tailings (“bathtub effect”), the long-
term average water flux rate through the tailings cell cover system (predicted with the
infiltration model) was used to calculate the amount of water entering the tailings during
the 200-year regulatory timeframe specified by the Permit. The amount of water
expected to migrate through the cover and enter the tailings cells (i.e., assuming all
recharge to the tailings can act to increase the amount of head on the liner) was then used
to calculate the maximum potential rise in water levels in the tailings assuming no water
flow through the liners (i.e., all water that infiltrated through the cover was accumulated
in the cells). The assumptions for evaluating the “bathtub effect” result in an end-
member scenario expected to produce a conservative estimate of closed-cell cover system
performance.
The amount of water calculated to enter the tailings after 200 years is equal to 90
millimeters (0.3 feet) of water. Assuming a tailings porosity of 57%, the calculated
water-level rise on the liner is approximately 160 millimeters (0.53 feet). Consequently,
a significant build-up of water (“bathtub effect”) within the cells is not anticipated and
the leachate head within the tailings is not predicted to rise above or over-top the
maximum liner elevation (which typically is greater than 20 feet above the bottom of the
cell), meeting the requirement of the Permit (Part I.D.8).

5-3
5.1.3 Bedrock Vadose Zone Flow and Contaminant Transport Modeling
The bedrock vadose zone flow and contaminant transport model was used to predict
potential flow rates and contaminant transport rates through the bedrock vadose zone to
the perched aquifer during the operational, dewatering, and post-closure steady-state
timeframes. Solute transport models were developed for the bedrock vadose zone
beneath Cell 1 (contingency cell identified for the potential disposal of decommissioning
and deconstruction debris), Cells 2 & 3, and Cells 4A & 4B. For simplicity, a vadose
zone thickness of 12.8 meters (42 feet) was assumed for all of the simulations. This is a
conservative assumption given that the average vadose zone thicknesses beneath Cell 2,
Cell 3, and Cell 4A are 19.2 m (63 ft), 20.1 m (66 ft), and 17.1 m (56 ft). HP1 was used
to simulate potential solute transport of conservative (chloride) and nonconservative
(sulfate, uranium, and other trace elements) solutes through the bedrock vadose zone
beneath the tailings cells.
Potential water flux rates through the primary liner installed beneath Cells 2 & 3 and the
secondary liner installed beneath Cells 4A & 4B were calculated using the Giroud-
Bonaparte Equation. Estimates of potential water flux rates through the liners were used
as an upper boundary condition (time-dependent flux) for the HP1 model used to predict
flow and solute transport through the bedrock vadose zone to the perched aquifer during
the operational, dewatering, and post-closure steady-state timeframes. The average long-
term water flux rate through the ET cover (predicted with the infiltration model) was used
as an upper boundary condition (constant flux) for Cell 1 to represent the post-closure
steady-state period. The bottom of Cell 1 (if constructed) will contain a soil liner
compacted to achieve low permeability, but this layer was not included in the modeling,
which yields conservative estimates of solute transport through the bedrock vadose zone.
The calculated potential water flux rates through the liners were multiplied by the
average solute concentrations measured in the tailings slimes drains to yield a time-
dependent mass flux rate applied as an upper boundary condition to the top of the
bedrock vadose zone. The average solute concentrations were used as input to represent
the source term solution chemistry of the tailings pore water.

5-4
Cells 2 & 3 Model-Predicted Water Flux Rate. The potential water flux rate at the
bottom of the bedrock vadose zone (immediately above the perched aquifer) is predicted
to reach a maximum value of approximately 7.5 mm/yr after 25 years of tailings cell
operation (note that tailings cells are not covered during this period). The potential flux
rate is then predicted to rapidly decline in response to decreased head (saturated
thickness) that occur in the tailings during the dewatering phase, ultimately reaching a
long-term steady state value of approximately 0.7 mm/yr during the 200-year regulatory
post-closure period. The calculations used to determine the potential flux rates through
the liners were based on conservative assumptions, and there is considerable evidence
that the cells are not leaking. Furthermore, consolidation of fine-grained tailings and
deposition of tailing slimes, coupled with the chemical nature of the pore water (e.g.,
precipitation of gypsum and amorphous mineral phases), is anticipated to essentially seal
some of the defects, which would act to decrease the potential flux rates through the
liners.
Cells 2 & 3 Model-Predicted Chloride Concentration. The model-predicted increase
in chloride concentrations at the bottom of the bedrock vadose zone beneath Cells 2 & 3
after 240 years (including operational, dewatering, and post-closure periods) of transport
is 0.01 mg/L. The chloride concentration at the bottom of the vadose zone represents the
model-predicted addition of chloride as a result of the potential flux from the tailings
cells. While there is naturally-occurring chloride in the vadose zone, the modeling
assumed no initial chloride for simplicity, and because there is a lack of data concerning
background chloride concentrations and the distribution of chloride within the vadose
zone. Furthermore, the model-predicted chloride concentration is the solute
concentration in vadose zone pore water that will reach the perched aquifer; however, the
predicted concentration is not equal to the concentration in groundwater. A model was
not constructed to determine the actual (diluted) concentration in groundwater because
the chloride concentration predicted at the bottom of the vadose zone was orders of
magnitude less than the minimum GWCL for chloride, which is 10 mg/L. The minimum
GWCL (for chloride and all other solutes modeled) was selected from the list of
monitoring wells located immediately downgradient from the tailings cells (i.e.,

5-5
monitoring wells MW-5, MW-11, MW-12, MW-14, MW-15, MW-23, MW-24, MW-28,
MW-29, MW-30, and MW-31; GWCL’s for these wells are specified in the Permit).
Cells 2 & 3 Model-Predicted Sulfate Concentration. The model-predicted sulfate
concentration at the bottom of the bedrock vadose zone beneath Cells 2 & 3 after 240
years of transport is 0.014 mg/L. The distribution of sulfate within the bedrock vadose
zone is controlled by the amount of gypsum that may precipitate from solution. The
sulfate concentration at the bottom of the bedrock vadose zone represents the model-
predicted addition of sulfate as a result of the potential flux from the tailings cells. A
model was not constructed to determine the actual (diluted) concentration in groundwater
because the sulfate concentration predicted at the bottom of the vadose zone was orders
of magnitude less than the minimum GWCL for sulfate, which is 532 mg/L for
monitoring wells located immediately downgradient from the tailings cells.
Cells 2 & 3 Model-Predicted Uranium Concentration. Uranium is not predicted to
reach the bottom of the bedrock vadose zone beneath Cells 2 & 3 during the 240-year
timeframe. Adsorption of uranium onto the surface of hydrous ferric oxide (HFO)
present in the bedrock vadose zone limits the transport distance below the liner. The
depth at which the model-predicted uranium concentration is approximately equal to the
minimum GWCL (0.0049 mg/L) after 240 years is 2.3 meters (8 feet) below the tailing
cell liner system; a minimum of 10.5 meters (34 feet) above the perched water table. The
uranium concentration within the bedrock vadose zone represents the model-predicted
addition of uranium as a result of the potential flux from the tailings cells. HFO is the
only solid phase that serves as a potential sorption site of uranium and other trace
elements, which is a conservative assumption because other phases (e.g., hematite,
quartz, clays, etc.) also participate in surface complexation reactions.
Cells 2 & 3 Model-Predicted Concentration of Other Trace Elements. The sorption
of uranium was competitive because additional trace elements were modeled. Solutes
included in the model were based on their elevated concentrations in the tailings pore
water as compared to the GWCLs. Transport of the following trace elements was
modeled: arsenic, cadmium, copper, nickel, vanadium, and zinc. Similar to uranium,

5-6
these solutes were predicted to migrate a limited distance below the liner (e.g., a few
meters).
Cells 4A & 4B Model-Predicted Water Flux Rate. The calculated potential flux of
water through the secondary liner beneath Cells 4A & 4B for the maximum head within
the leak detection system during the operational and dewatering periods is approximately
8 x 10-5 mm/yr. The potential flux rates predicted at the end of dewatering are assumed
to equal the rate during post-closure steady state because the increase in water levels is
anticipated to be minor. Therefore, the model-predicted water flux rate at the bottom of
the bedrock vadose zone (immediately above the perched aquifer) during post-closure
steady-state is 8 x 10-5 mm/yr.
Cells 4A & 4B Model-Predicted Concentrations. For all practical purposes, chloride is
not predicted to reach the bottom of the bedrock vadose zone during the 12-year
operational and 200-year post-closure periods (The chloride concentration predicted to
reach the water table at 212 years was 5 x 10-14 mg/L.). The chloride concentration is not
predicted to exceed the 10 mg/L minimum GWCL anywhere in the vadose zone because
of the diminutive chloride mass flux rate entering the vadose zone. Considering that
chloride is a conservative tracer, and that transport is not affected by sorption or mineral
precipitation reactions, coupled with the fact that the model predictions demonstrate
nearly zero impact, additional model predictions of solute transport for nonconservative
contaminants (sulfate, uranium, other trace elements) was considered unnecessary.
Cell 1 Model-Predicted Water Flux Rate. If Cell 1 is constructed for decommissioning
and deconstruction disposal, it will include a soil liner compacted to achieve low
permeability and will be covered with the monolithic ET cover. The cover design will be
the same as the monolithic ET cover proposed for the other cells. Consequently, the
long-term average infiltration rate would be equivalent to the value presented for the
other cells. The model-predicted water flux rate at the bottom of the vadose zone
(immediately above the perched aquifer) during 200-year post-closure steady-state is
predicted to be approximately 0.5 mm/yr.

5-7
Cell 1 Model-Predicted Concentrations. The source term of the decommissioning and
deconstruction debris is assumed to equal the concentrations assigned to the tailings pore
water, which is anticipated to lead to conservative predictions that over predict the
potential impacts. For all practical purposes, chloride is not predicted to reach the bottom
of the bedrock vadose zone during the 200-year transport timeframe (the chloride
concentration predicted to reach the water table at 200 years was 7 x 10-9 mg/L).
Considering that chloride is a conservative tracer, and that transport is not affected by
sorption or mineral precipitation reactions, coupled with the diminutive transport
distance, additional model predictions of solute transport for nonconservative
contaminants (sulfate, uranium, other trace elements) was considered unnecessary.
5.1.4 Summary of Closed Cell Cover System Performance
The assumptions used to construct the numerical models to predict infiltration through
the cover and potential impacts to the perched groundwater system, generally were either
conservative or based on anticipated conditions. As a result, the predictions are
considered to be conservative.
Part I.D.8 (Closed Cell Performance Requirements) of the Permit states:
“Before reclamation and closure of any tailings disposal cell, the Permittee shall ensure
that the final design, construction, and operation of the cover system at each tailings cell
will comply with all requirements of an approved Reclamation Plan, and will for a period
of not less than 200 years meet the following minimum performance requirements:
• Minimize infiltration of precipitation or other surface water into the tailings,
including, but not limited to the radon barrier.
• Prevent the accumulation of leachate head within the tailings waste layer that
could rise above or over-top the maximum flexible membrane liner (FML)
elevation internal to any disposal cell, i.e., create a “bathtub effect”.

5-8
• Ensure the groundwater quality at the compliance monitoring wells does not
exceed the Ground Water Quality Standards (GWQS’s) or Ground Water
Compliance Limits (GWCL’s) specified in Part I.C.1 and Table 2 of the
Permit.”
The bedrock vadose zone model evaluates the potential impacts of the tailings cell system
as a whole (liner system, dewatering system, and cover system) on groundwater for the
project lifecycle, including the operational phase (without cell cover system), the
dewatering phase (with an interim cover only), and the 200-year regulatory post-closure
period (with complete cover system, but with some limited water remaining in the
tailings). For the 240-year period modeled, the potential flux rate and contaminant
transport through the underlying bedrock vadose zone is dominated by the effect of the
operational phase when the cells were not covered. As a result, the bedrock vadose zone
model including the operational phase is not a reliable indicator of performance of the
closed-cell cover system. However, even with the operational phase, model-predicted
contaminant concentrations in vadose zone pore water entering the perched aquifer did
not exceed the GWQS’s or GWCL’s for any downgradient monitoring wells, thus
demonstrating compliance with Part I.D.8 of the Permit.
Based on the model results, the proposed monolithic ET cover will minimize infiltration
into the tailings, will prevent build-up of leachate head on the cell liner, and will be
protective of groundwater quality; contaminant concentrations are not predicted to exceed
the GWCS’s or GWCL’s at the compliance monitoring wells specified in the Permit, thus
demonstrating compliance with Part I.D.8 of the Permit. Furthermore, the results of the
radon attenuation model demonstrate that the proposed monolithic ET cover will
attenuate radon fluxes thereby achieving the State of Utah’s long-term radon emanation
standard for uranium mill tailings (Utah Administrative Code, Rule 313-24).
5.2 POST-AUDIT MONITORING PLAN
To check the accuracy of the model predictions, a post-audit can be performed, often
referred to as model verification. Additional data are collected and after a specified

5-9
period, the model is rerun with new input data and the results are compared to field-
measured data for the same period. Given difficulties associated with data collection and
the time-scale on which processes occur in the vadose zone, a post-audit of the
HYDRUS-1D and HP1 models is not practical. Given the time-scale on which the
model-predicts contaminants could potentially reach the perched aquifer, post-audit
monitoring should include ongoing groundwater level measurements and groundwater
sampling, but at a reduced frequency and at a limited set of wells relative to that currently
used to establish background levels. Sampling should focus on the closest downgradient
monitoring wells.
A post audit of the MODFLOW model for the tailings cell dewatering (presented in
Appendix J) is described below. For post-audit monitoring of the dewatering system,
water levels in the tailings and pumping rates and volumes should be measured and
recorded monthly. The model predictions should be compared to these data. If the
dewatering rates predicted by the model are considerably different than actual measured
rates, the MODFLOW model should be recalibrated by adjusting terms such as areal
recharge, hydraulic conductivity of tailings, storage parameters, and/or slimes drain
conductance to match dewatering rates and measured water levels.

R-1
REFERENCES

Albright, W.H., C.H. Benson, G.W. Gee, A.C. Roesler, T. Abichou, P. Apiwantragoon,
B.F. Lyles, and S.A. Rock, 2004. Field Water Balance of Landfill Final Covers.
Journal of Environmental Quality 33:2317-2332.
Allen, R.G., L.S. Pereira, D. Raes, and M. Smith, 1998. FAO Irrigation and Drainage Paper, No. 56, Crop Evapotranspiration (Guidelines for Computing Crop Water
Requirements), pp. 300, FAO - Food and Agriculture Organization of the United
Nations, Rome, Italy.
Bear, J., 1972. Dynamics of Fluid in Porous Media, American Elsevier, New York, pp 764.
Bear, J., and A. Verruijt, 1987. Modeling Groundwater Flow and Pollution, D. Reidel
Publishing Company, Boston, pp. 414.
Bolen, M.M., A.C. Roesler, C.H. Benson, and W.H. Albright, 2001. Alternative Cover
Assessment Program: Phase II Report. Geo Engineering Report 01-10, Department of Civil and Environmental Engineering, University of Wisconsin-Madison.
Brooks, R.H., and A.T. Corey, 1964. Hydraulic Properties of Porous Media. Hydrology
Paper no. 3. Colorado State University, Fort Collins, CO.
Chen, J-S., R.L. Drake, X. Lin, and D.G. Jewett, 2002. Simulating Radionuclide Fate
and Transport in the Unsaturated Zone: Evaluation and Sensitivity Analyses of
Select Computer Models. U.S. Environmental Protection Agency EPA/600/R-02/082. July 2002.
D’Appolonia Consulting Engineers, Inc., 1979, Engineering Report, Tailings
Management System, White Mesa Uranium Project, Blanding, Utah.
D’Appolonia Consulting Engineers, Inc., 1982. Construction Report, Initial Phase – Tailings Management System, White Mesa Uranium Project, Blanding, Utah. Prepared for Energy Fuels Nuclear, Inc.
Denison Mines (USA) Corp., 2009. Reclamation Plan, White Mesa Mill, Blanding, Utah,
Radioactive Materials License No. UT1900479, Revision 4.0, November 2009.
Dzombak, D.A., and F.M. Morel, 1990. Surface Complexation Modeling: Hydrous
Ferric Oxide, John Wiley & Sons, New York, NY.
Fayer, M.J., and G.W. Gee, 2006. Multiple-Year Water Balance of Soil Covers in a
Semiarid Setting. Journal of Environmental Quality 35:366-377.
Feddes, R. A., P. J. Kowalik, and H. Zaradny, 1978. Simulation of Field Water Use and
Crop Yield, John Wiley & Sons, New York, NY, pp. 188.

R-2
Fetter, C. W, 1998. Contaminant Hydrogeology, 2rd Ed., Prentice Hall, Upper Saddle River, New Jersey, pp. 458.
Gee, G.W., P.J. Wieringa, B.J. Andraski, M.H. Young, M.J. Fayer, and M.L. Rockhold,
1994. Variations in Water Balance and Recharge Potential at Three Western
Desert Sites. Soil Sci. Soc. Am. J. 58:63-72.
Geosyntec Consultants, 2006a. Cell 4A Lining System Design Report for the White Mesa
Mill, Blanding, Utah. Prepared for International Uranium (USA) Corporation,
January 2006.
Geosyntec Consultants, 2006b. Stockpile Evaluation Tailings Cell 4A, White Mesa Mill,
Technical Memo submitted to International Uranium (USA) Corporation, 23 January 2006.
Geosyntec Consultants, 2007a. Cell 4B Design Report, White Mesa Mill, Blanding, Utah.
Prepared for Denison Mines (USA) Corp., December 2007.
Geosyntec Consultants, 2007b. Revised Construction Drawings, DMC White Mesa Mill,
Cell 4A Lining System. June 2007.
Geosyntec Consultants, 2007c. Analysis of Slimes Drains for White Mesa Mill - Cell 4A, Computations submitted to Denison Mines, 12 May 2007.
Giroud, J.P., and R. Bonaparte, 1989. Leakage through Liners Constructed with
Geomembrane Liners, Parts I, II, and Technical Notes, Geotextiles and
Geomembranes 8:27-67, 71-111, 337-340.
Giroud, J.P., K.B. Tweneboah, and R. Bonaparte, 1992. Rate of Leakage through a Composite Liner due to Geomembrane Defects, Geotextiles and Geomembranes
11:1-28.
Hurst, T.G., and D. K. Solomon, 2008. Summary of Work Completed, Data Results,
Interpretations, and Recommendations for the July 2007 Sampling Event at the
Denison Mines, USA, White Mesa Uranium Mill, near Blanding, Utah., University of Utah Geology and Geophysics Department. Prepared for the Utah Division of
Radiation Control, May 2008, pp. 62.
Hydro Geo Chem, Inc., 2002. Hydraulic Testing at the White Mesa Uranium Mill Site,
near Blanding, Utah During July 2002. Prepared for International Uranium (USA) Corporation. August 2002.
Hydro Geo Chem, Inc., 2009. Site Hydrogeology and Estimation of Groundwater Pore
Velocities in the Perched Zone, White Mesa Uranium Mill Site near Blanding,
Utah. Prepared for Denison Mines (USA) Corp., December 2009.
Hydro Geo Chem, Inc., 2003. Site Hydrogeology and Estimation of Groundwater Travel
Times in the Perched Zone, White Mesa Uranium Mill Site, near Blanding, Utah. Prepared for International Uranium Corporation. January 2003.

R-3
INTERA, Inc., 2007a. Revised Background Groundwater Quality Report: Existing Wells
for Denison Mines (USA) Corp.’s White Mesa Uranium Mill Site, San Juan County,
Utah. Prepared for Denison Mines (USA) Corp. October 2007.
INTERA, Inc., 2007b. Revised Addendum Evaluation of Available Pre-Operational and
Regional Background Data Background Groundwater Quality Report: Existing
Wells for Denison Mines (USA) Corp.’s White Mesa Mill Site, San Juan County,
Utah, November 2007.
INTERA, Inc., 2008. Revised Addendum Background Groundwater Quality Report: New
Wells for Denison Mines (USA) Corp.’s White Mesa Mill Site, San Juan County,
Utah, April 2008.
International Uranium (USA) Corporation, 2000. Reclamation Plan, White Mesa Mill,
Blanding, Utah. Source Material License No. SUA-1358, Docket No. 40-8681
Revision 3.0. July 2000.
Jacques, D., J. Simunek, D. Mallants, and M.Th . van Genuchten, 2006. Operator-
splitting errors in coupled reactive transport codes for transient variably saturated
flow and contaminant transport in layered soil profiles. Journal of Contaminant
Hydrology, 88, p. 197–218.
Jacques, D., and J. Simunek, 2005. User manual of the multicomponent variably
saturated transport model HP1: Description, verification and examples. Version 1.0.
BLG-998. SCKڄCEN, Mol, Belgium, pp. 79.
Li, Y-H., and S. Gregory, 1974. Diffusion of Ions in Sea Water and Deep-Sea Sediments. Geochemica et Cosmochemica Acta 38:703-714.
Millington, R. J. and J. M. Quirk, 1961. Permeability of Porous Solids, Trans. Faraday
Soc., 57, 1200-1207.
Mualem, Y., 1976. A New Model for Prediction the Hydraulic Conductivity of
Unsaturated Porous Media, Water Resources Research 12:513-522.
Parkhurst, D.L and C.A.J. Appelo, 1999. User’s guide to PHREEQC User's Guide to PHREEQC (Version 2)--A Computer Program for Speciation, Batch-Reaction,
One-Dimensional Transport, and Inverse Geochemical Calculations, U.S.
Geological Survey Water-Resources Investigations Report 99-4259, pp. 312.
Simunek, J., K. Huang, and M. Th. van Genuchten, 1998. The HYDRUS Code for
Simulating the One-Dimensional Movement of Water, Heat, and Multiple Solutes in
Variably-Saturated Media. Research Report No. 144. U.S. Salinity Laboratory,
Agricultural Research Service, U.S. Department of Agriculture, Riverside, CA.
Simunek, J., M. Th. van Genuchten, and M. Sejna, 2005. The HYDRUS-1D Software
Package for Simulating the One-Dimensional Movement of Water, Heat, and
Multiple Solutes in Variably-Saturated Media, Version 3.0. Department of

R-4
Environmental Sciences, University of California – Riverside, Riverside, CA. pp. 240.
Simunek, J., M. Sejna, H. Saito, M. Sakai, and M. Th. van Genuchten, 2009. The
HYDRUS-1D Software Package for Simulating the Movement of Water, Heat, and
Multiple Solutes in Variably Saturated Media, Version 4.08, HYDRUS Software
Series 3, Department of Environmental Sciences, University of California
Riverside, Riverside, California, USA, pp. 330.
Scanlon, B.R., R.C. Reedy, K.E. Keese, and S.F. Dwyer, 2005. Evaluation of
Evapotranspirative Covers for Waste Containment in Arid and Semiarid Regions in
the Southwestern USA. Vadose Zone Journal 4:55-71.
Scanlon, B.R., M. Christman, R.C. Reedy, I. Porro, J. Simunek, and G.N. Flerchinger,
2002. Intercode Comparisons for Simulating Water Balance of Surficial Sediments
in Semiarid Regions. Water Resources Research 38:1323-1339.
TITAN Environmental Corporation, 1994. Hydrogeological Evaluation of White Mesa
Uranium Mill. Prepared for Energy Fuels Nuclear, Inc.
TITAN Environmental Corporation, 1996. Tailings Cover Design, White Mesa Mill,
Blanding Utah. Prepared for Energy Fuels Nuclear, Inc. September 1996.
Utah Climate Center, 2007. Daily climate records were extracted from the COOP
database from http://climate.usurf.usu.edu/products/data.php for the Blanding, Utah
weather station (420738) on 25 January 2007.
van Genuchten, M.Th., 1980. A Closed-Form Equation for Predicting the Hydraulic
Conductivity of Unsaturated Soils. Soil Sci. Am. Jour. 44:892-898.
Vogel, T., and M. Cislerova, 1988. On the Reliability of Unsaturated Hydraulic
Conductivity Calculated from the Moisture Retention Curve. Transport in Porous
Media 3:1-15.
Walvoord, M.A., M.A. Plummer, and F.M. Phillips, 2002. Deep Arid System
Hydrodynamics: 1. Equilibrium States and Response Times in Thick Desert Vadose
Zones. Water Resources Research 38.
Waugh, W.J., M.K. Kastens, L.R.L. Sheader, C.H. Benson, W.H. Albright, and P.S.
Mushovic (2008), Monitoring the performance of an alternative landfill cover at
the Monticello, Utah, Uranium Mill Tailings Disposal Site, Proceedings of the
Waste Management 2008 Symposium, Tucson, Arizona, 24-28 February 2008.

APPENDIX A

LABORATORY REPORTS WITH RESULTS OF VADOSE ZONE
MINERALOGICAL TESTING AND PROPERTIES OF
STOCKPILED SOIL

ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Analytical
Report
MWH America's Inc.
10619 S. Jordan Gateway Suite 100
Salt Lake City, UT 84095
ACZ Project ID: L62140
Doug Oliver:
Enclosed are the analytical results for sample(s) submitted to ACZ Laboratories, Inc. (ACZ) on April 20, 2007.
This project has been assigned to ACZ's project number, L62140. Please reference this number in all future
inquiries.
All analyses were performed according to ACZ's Quality Assurance Plan, version 11.0. The enclosed results
relate only to the samples received under L62140. Each section of this report has been reviewed and approved
by the appropriate Laboratory Supervisor, or a qualified substitute.
Except as noted, the test results for the methods and parameters listed on ACZ's current NELAC certificate
letter (#ACZ) meet all requirements of NELAC.
This report shall be used or copied only in its entirety. ACZ is not responsible for the consequences arising
from the use of a partial report.
All samples and sub-samples associated with this project will be disposed of after May 27, 2007. If the samples
are determined to be hazardous, additional charges apply for disposal (typically less than $10/sample). If you
would like the samples to be held longer than ACZ's stated policy or to be returned, please contact your Project
Manager or Customer Service Representative for further details and associated costs. ACZ retains analytical
reports for five years.
If you have any questions or other needs, please contact your Project Manager.
Doug Oliver
April 27, 2007
cc: Ryan Jakubowski
Project ID: 1004-A0002-87430-OM/
MWH America's Inc.
P.O. Box 6610
Broomfield, CO 80021
Accounts Payable
Report to:Bill to:
REPAD.01.06.05.01
ACZ Sample ID:L62140-01
Sample ID: MW-30 37.5-38.0
Sample Matrix:Leachate
MWH America's Inc.
Project ID:1004-A0002-87430-OM/
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:04/20/07 00:00
Date Received:04/20/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Metals Analysis
XQ
Aluminum, dissolved M200.7 ICP 04/25/07 21:09137 mg/L 0.2 djt0.03*
Calcium, dissolved M200.7 ICP 04/25/07 0:4053.5 mg/L 1 djt0.2*
Iron, dissolved M200.7 ICP 04/25/07 21:09295mg/L 0.05 djt0.02*
Magnesium, dissolved M200.7 ICP 04/25/07 0:4059.6 mg/L 1 djt0.2*
Manganese, dissolved M200.7 ICP 04/25/07 0:408.440 mg/L 0.03 djt0.005*
Uranium, dissolved M200.8 ICP-MS 04/24/07 1:150.0156 mg/L 0.0005 scp0.0001*
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Sample ID:L62140-02
Sample ID: MW-30 43.0-43.2
Sample Matrix:Leachate
MWH America's Inc.
Project ID:1004-A0002-87430-OM/
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:04/20/07 00:00
Date Received:04/20/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Metals Analysis
XQ
Aluminum, dissolved M200.7 ICP 04/25/07 21:1369.90 mg/L 0.2 djt0.03*
Calcium, dissolved M200.7 ICP 04/25/07 1:0168.4 mg/L 1 djt0.2*
Iron, dissolved M200.7 ICP 04/25/07 21:1338.30 mg/L 0.05 djt0.02*
Magnesium, dissolved M200.7 ICP 04/25/07 1:0132.5 mg/L 1 djt0.2*
Manganese, dissolved M200.7 ICP 04/25/07 1:010.057 mg/L 0.03 djt0.005*
Uranium, dissolved M200.8 ICP-MS 04/24/07 1:210.0109 mg/L 0.0005 scp0.0001*
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Sample ID:L62140-03
Sample ID: MW-30 43.2-43.5
Sample Matrix:Leachate
MWH America's Inc.
Project ID:1004-A0002-87430-OM/
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:04/20/07 00:00
Date Received:04/20/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Metals Analysis
XQ
Aluminum, dissolved M200.7 ICP 04/25/07 21:1858.30 mg/L 0.2 djt0.03*
Calcium, dissolved M200.7 ICP 04/25/07 1:0553.5 mg/L 1 djt0.2*
Iron, dissolved M200.7 ICP 04/25/07 21:1825.30 mg/L 0.05 djt0.02*
Magnesium, dissolved M200.7 ICP 04/25/07 1:0526.1 mg/L 1 djt0.2*
Manganese, dissolved M200.7 ICP 04/25/07 1:050.070 mg/L 0.03 djt0.005*
Uranium, dissolved M200.8 ICP-MS 04/24/07 1:270.0078 mg/L 0.001 scp0.0002*
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Sample ID:L62140-04
Sample ID: MW-23 53.0-53.5
Sample Matrix:Leachate
MWH America's Inc.
Project ID:1004-A0002-87430-OM/
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:04/20/07 00:00
Date Received:04/20/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Metals Analysis
XQ
Aluminum, dissolved M200.7 ICP 04/25/07 21:30176 mg/L 0.2 djt0.03*
Calcium, dissolved M200.7 ICP 04/25/07 1:0976.5 mg/L 1 djt0.2*
Iron, dissolved M200.7 ICP 04/25/07 21:30304mg/L 0.05 djt0.02*
Magnesium, dissolved M200.7 ICP 04/25/07 1:09106mg/L 1 djt0.2*
Manganese, dissolved M200.7 ICP 04/25/07 1:094.370 mg/L 0.03 djt0.005*
Uranium, dissolved M200.8 ICP-MS 04/24/07 1:450.0156 mg/L 0.0005 scp0.0001*
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Sample ID:L62140-05
Sample ID: MW-23 74.0-74.3
Sample Matrix:Leachate
MWH America's Inc.
Project ID:1004-A0002-87430-OM/
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:04/20/07 00:00
Date Received:04/20/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Metals Analysis
XQ
Aluminum, dissolved M200.7 ICP 04/25/07 21:4340.70 mg/L 0.2 djt0.03*
Calcium, dissolved M200.7 ICP 04/25/07 1:1424.7 mg/L 1 djt0.2*
Iron, dissolved M200.7 ICP 04/25/07 21:4319.10 mg/L 0.05 djt0.02*
Magnesium, dissolved M200.7 ICP 04/25/07 1:1428.4 mg/L 1 djt0.2*
Manganese, dissolved M200.7 ICP 04/25/07 1:140.069 mg/L 0.03 djt0.005*
Uranium, dissolved M200.8 ICP-MS 04/24/07 1:500.0112 mg/L 0.0005 scp0.0001*
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Sample ID:L62140-06
Sample ID: MW-23 82.5-82.7
Sample Matrix:Leachate
MWH America's Inc.
Project ID:1004-A0002-87430-OM/
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:04/20/07 00:00
Date Received:04/20/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Metals Analysis
XQ
Aluminum, dissolved M200.7 ICP 04/25/07 21:4815.20 mg/L 0.2 djt0.03*
Calcium, dissolved M200.7 ICP 04/25/07 1:1811.3 mg/L 1 djt0.2*
Iron, dissolved M200.7 ICP 04/25/07 21:4814.50 mg/L 0.05 djt0.02*
Magnesium, dissolved M200.7 ICP 04/25/07 1:1812.7 mg/L 1 djt0.2*
Manganese, dissolved M200.7 ICP 04/25/07 1:180.049 mg/L 0.03 djt0.005*
Uranium, dissolved M200.8 ICP-MS 04/24/07 1:560.0122 mg/L 0.0005 scp0.0001*
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Sample ID:L62140-07
Sample ID: MW-23 99.8-100.0
Sample Matrix:Leachate
MWH America's Inc.
Project ID:1004-A0002-87430-OM/
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:04/20/07 00:00
Date Received:04/20/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Metals Analysis
XQ
Aluminum, dissolved M200.7 ICP 04/25/07 21:5229.50 mg/L 0.2 djt0.03*
Calcium, dissolved M200.7 ICP 04/25/07 1:2219.1 mg/L 1 djt0.2*
Iron, dissolved M200.7 ICP 04/25/07 21:5274.60 mg/L 0.05 djt0.02*
Magnesium, dissolved M200.7 ICP 04/25/07 1:229.0 mg/L 1 djt0.2*
Manganese, dissolved M200.7 ICP 04/25/07 1:220.222 mg/L 0.03 djt0.005*
Uranium, dissolved M200.8 ICP-MS 04/24/07 2:140.0147 mg/L 0.0005 scp0.0001*
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Sample ID:L62140-08
Sample ID: MW-23 103.0-103.3
Sample Matrix:Leachate
MWH America's Inc.
Project ID:1004-A0002-87430-OM/
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:04/20/07 00:00
Date Received:04/20/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Metals Analysis
XQ
Aluminum, dissolved M200.7 ICP 04/25/07 21:5624.50 mg/L 0.2 djt0.03*
Calcium, dissolved M200.7 ICP 04/25/07 1:2614.4 mg/L 1 djt0.2*
Iron, dissolved M200.7 ICP 04/25/07 21:5615.50 mg/L 0.05 djt0.02*
Magnesium, dissolved M200.7 ICP 04/25/07 1:269.8 mg/L 1 djt0.2*
Manganese, dissolved M200.7 ICP 04/25/07 1:260.229 mg/L 0.03 djt0.005*
Uranium, dissolved M200.8 ICP-MS 04/24/07 2:190.0105 mg/L 0.0005 scp0.0001*
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Sample ID:L62140-09
Sample ID: MW-23 103.0-103.3DUP
Sample Matrix:Leachate
MWH America's Inc.
Project ID:1004-A0002-87430-OM/
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:04/20/07 00:00
Date Received:04/20/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Metals Analysis
XQ
Aluminum, dissolved M200.7 ICP 04/26/07 6:1223.50 mg/L 0.2 djt0.03*
Calcium, dissolved M200.7 ICP 04/26/07 6:1212.7 mg/L 1 djt0.2*
Iron, dissolved M200.7 ICP 04/26/07 6:1215.20 mg/L 0.05 djt0.02*
Magnesium, dissolved M200.7 ICP 04/26/07 6:129.4 mg/L 1 djt0.2*
Manganese, dissolved M200.7 ICP 04/26/07 6:120.224 mg/L 0.03 djt0.005*
Uranium, dissolved M200.8 ICP-MS 04/24/07 2:250.0105 mg/L 0.0005 scp0.0001*
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Sample ID:L62140-10
Sample ID: PBS
Sample Matrix:Leachate
MWH America's Inc.
Project ID:1004-A0002-87430-OM/
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:04/20/07 00:00
Date Received:04/20/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Metals Analysis
XQ
Aluminum, dissolved M200.7 ICP 04/26/07 6:160.15 mg/L 0.2B djt0.03*
Calcium, dissolved M200.7 ICP 04/26/07 6:160.2 mg/L 1B djt0.2*
Iron, dissolved M200.7 ICP 04/26/07 6:160.04 mg/L 0.05B djt0.02*
Magnesium, dissolved M200.7 ICP 04/26/07 6:16mg/L 1U djt0.2*
Manganese, dissolved M200.7 ICP 04/26/07 6:16mg/L 0.03U djt0.005*
Uranium, dissolved M200.8 ICP-MS 04/24/07 2:31mg/L 0.0005U scp0.0001*
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Report Header Explanations
Batch A distinct set of samples analyzed at a specific time
Found Value of the QC Type of interest
Limit Upper limit for RPD, in %.
Lower Lower Recovery Limit, in % (except for LCSS, mg/Kg)
MDL Method Detection Limit. Same as Minimum Reporting Limit. Allows for instrument and annual fluctuations.
PCN/SCN A number assigned to reagents/standards to trace to the manufacturer's certificate of analysis
PQL Practical Quantitation Limit, typically 5 times the MDL.
QC True Value of the Control Sample or the amount added to the Spike
Rec Amount of the true value or spike added recovered, in % (except for LCSS, mg/Kg)
RPD Relative Percent Difference, calculation used for Duplicate QC Types
Upper Upper Recovery Limit, in % (except for LCSS, mg/Kg)
Sample Value of the Sample of interest
QC Sample Types
AS Analytical Spike (Post Digestion)LCSWD Laboratory Control Sample - Water Duplicate
ASD Analytical Spike (Post Digestion) Duplicate LFB Laboratory Fortified Blank
CCB Continuing Calibration Blank LFM Laboratory Fortified Matrix
CCV Continuing Calivation Verification standard LFMD Laboratory Fortified Matrix Duplicate
DUP Sample Duplicate LRB Laboratory Reagent Blank
ICB Initial Calibration Blank MS Matrix Spike
ICV Initial Calibration Verification standard MSD Matrix Spike Duplicate
ICSAB Inter-element Correction Standard - A plus B solutions PBS Prep Blank - Soil
LCSS Laboratory Control Sample - Soil PBW Prep Blank - Water
LCSSD Laboratory Control Sample - Soil Duplicate PQV Practical Quantitation Verification standard
LCSW Laboratory Control Sample - Water SDL Serial Dilution
QC Sample Type Explanations
Blanks Verifies that there is no or minimal contamination in the prep method or calibration procedure.
Control Samples Verifies the accuracy of the method, including the prep procedure.
Duplicates Verifies the precision of the instrument and/or method.
Spikes/Fortified Matrix Determines sample matrix interferences, if any.
Standard Verifies the validity of the calibration.
ACZ Qualifiers (Qual)
B Analyte concentration detected at a value between MDL and PQL.
H Analysis exceeded method hold time. pH is a field test with an immediate hold time.
U Analyte was analyzed for but not detected at the indicated MDL
Method References
(1) EPA 600/4-83-020. Methods for Chemical Analysis of Water and Wastes, March 1983.
(2) EPA 600/R-93-100. Methods for the Determination of Inorganic Substances in Environmental Samples, August 1993.
(3) EPA 600/R-94-111. Methods for the Determination of Metals in Environmental Samples - Supplement I, May 1994.
(5) EPA SW-846. Test Methods for Evaluating Solid Waste, Third Edition with Update III, December 1996.
(6) Standard Methods for the Examination of Water and Wastewater, 19th edition, 1995.
Comments
(1) QC results calculated from raw data. Results may vary slightly if the rounded values are used in the calculations.
(2) Soil, Sludge, and Plant matrices for Inorganic analyses are reported on a dry weight basis.
(3) Animal matrices for Inorganic analyses are reported on an "as received" basis.
REPIN03.02.07.01
Inorganic
Reference
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Inorganic Extended
Qualifier Report
ACZ Project ID:L62140MWH America's Inc.
ACZ ID PARAMETER QUAL DESCRIPTIONMETHODWORKNUM
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPManganese, dissolvedL62140-01 WG223614
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPManganese, dissolvedL62140-02 WG223614
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPAluminum, dissolvedL62140-03 WG223639
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPIron, dissolved
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPManganese, dissolvedWG223614
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPAluminum, dissolvedL62140-04 WG223639
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPIron, dissolved
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPManganese, dissolvedWG223614
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPAluminum, dissolvedL62140-05 WG223639
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPIron, dissolved
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPManganese, dissolvedWG223614
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPAluminum, dissolvedL62140-06 WG223639
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPIron, dissolved
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPManganese, dissolvedWG223614
REPAD.15.06.05.01
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Inorganic Extended
Qualifier Report
ACZ Project ID:L62140MWH America's Inc.
ACZ ID PARAMETER QUAL DESCRIPTIONMETHODWORKNUM
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPAluminum, dissolvedL62140-07 WG223639
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPIron, dissolved
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPManganese, dissolvedWG223614
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPAluminum, dissolvedL62140-08 WG223639
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPIron, dissolved
M3 The accuracy of the spike recovery does not apply because
analyte concentration in the sample is disproportionate to
the spike level. The recovery of the method control sample
was acceptable.
M200.7 ICPManganese, dissolvedWG223614
REPAD.15.06.05.01
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Certification
Qualifiers
ACZ Project ID:L62140MWH America's Inc.
Metals Analysis
The following parameters are not offered for certification or are not covered by NELAC certificate #ACZ.
Aluminum, dissolved M200.7 ICP
Calcium, dissolved M200.7 ICP
Iron, dissolved M200.7 ICP
Magnesium, dissolved M200.7 ICP
Manganese, dissolved M200.7 ICP
Uranium, dissolved M200.8 ICP-MS
REPAD.05.06.05.01
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Analytical
Report
MWH America's Inc.
1801California Street Suite 2600
Denver, CO 80202
ACZ Project ID: L64240
John Mahoney:
Enclosed are the analytical results for sample(s) submitted to ACZ Laboratories, Inc. (ACZ) on August 03,
2007. This project has been assigned to ACZ's project number, L64240. Please reference this number in all
future inquiries.
All analyses were performed according to ACZ's Quality Assurance Plan, version 11.0. The enclosed results
relate only to the samples received under L64240. Each section of this report has been reviewed and approved
by the appropriate Laboratory Supervisor, or a qualified substitute.
Except as noted, the test results for the methods and parameters listed on ACZ's current NELAC certificate
letter (#ACZ) meet all requirements of NELAC.
This report shall be used or copied only in its entirety. ACZ is not responsible for the consequences arising
from the use of a partial report.
All samples and sub-samples associated with this project will be disposed of after September 10, 2007. If the
samples are determined to be hazardous, additional charges apply for disposal (typically less than
$10/sample). If you would like the samples to be held longer than ACZ's stated policy or to be returned, please
contact your Project Manager or Customer Service Representative for further details and associated costs.
ACZ retains analytical reports for five years.
If you have any questions or other needs, please contact your Project Manager.
John Mahoney
August 10, 2007
cc: Ryan Jakubowski
Project ID:
MWH America's Inc.
PO Box 6610
Broomfield, CO 80021
Accounts Payable
Report to:Bill to:
REPAD.01.06.05.01
ACZ Sample ID:L64240-01
Sample ID: L61917-01
Sample Matrix:Soil
MWH America's Inc.
Project ID:
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:08/03/07 09:55
Date Received:08/03/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Soil Analysis
XQ
Acid Neutralization
Potential (calc)
M600/2-78-054 1.3 08/08/07 16:221t CaCO3/Kt 5 calc1
Neutralization
Potential as CaCO3
M600/2-78-054 3.2.3 08/04/07 9:350.1 %0.5B lwt0.1*
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Soil Preparation
XQ
Crush and Pulverize USDA No. 1, 1972 08/03/07 14:00 lwt
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Sample ID:L64240-02
Sample ID: L61917-02
Sample Matrix:Soil
MWH America's Inc.
Project ID:
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:08/03/07 09:55
Date Received:08/03/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Soil Analysis
XQ
Acid Neutralization
Potential (calc)
M600/2-78-054 1.3 08/08/07 16:221t CaCO3/Kt 5 calc1
Neutralization
Potential as CaCO3
M600/2-78-054 3.2.3 08/04/07 10:070.1 %0.5B lwt0.1*
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Soil Preparation
XQ
Crush and Pulverize USDA No. 1, 1972 08/03/07 14:03 lwt
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Sample ID:L64240-03
Sample ID: L61917-03
Sample Matrix:Soil
MWH America's Inc.
Project ID:
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:08/03/07 09:55
Date Received:08/03/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Soil Analysis
XQ
Acid Neutralization
Potential (calc)
M600/2-78-054 1.3 08/08/07 16:220t CaCO3/Kt 5 calc1
Neutralization
Potential as CaCO3
M600/2-78-054 3.2.3 08/04/07 10:39%0.5U lwt0.1*
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Soil Preparation
XQ
Crush and Pulverize USDA No. 1, 1972 08/03/07 14:07 lwt
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Sample ID:L64240-04
Sample ID: L61917-04
Sample Matrix:Soil
MWH America's Inc.
Project ID:
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:08/03/07 09:55
Date Received:08/03/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Soil Analysis
XQ
Acid Neutralization
Potential (calc)
M600/2-78-054 1.3 08/08/07 16:234t CaCO3/Kt 5 calc1
Neutralization
Potential as CaCO3
M600/2-78-054 3.2.3 08/04/07 11:110.4 %0.5B lwt0.1*
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Soil Preparation
XQ
Crush and Pulverize USDA No. 1, 1972 08/03/07 14:11 lwt
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Sample ID:L64240-05
Sample ID: L61917-05
Sample Matrix:Soil
MWH America's Inc.
Project ID:
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487(800) 334-5493
Inorganic Analytical
Results
Date Sampled:08/03/07 09:55
Date Received:08/03/07
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Soil Analysis
XQ
Acid Neutralization
Potential (calc)
M600/2-78-054 1.3 08/08/07 16:230t CaCO3/Kt 5 calc1
Neutralization
Potential as CaCO3
M600/2-78-054 3.2.3 08/04/07 11:43%0.5U lwt0.1*
Parameter EPA Method Result Units MDLQual AnalystDatePQL
Soil Preparation
XQ
Crush and Pulverize USDA No. 1, 1972 08/03/07 14:15 lwt
REPIN.02.06.05.01 * Please refer to Qualifier Reports for details
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Report Header Explanations
Batch A distinct set of samples analyzed at a specific time
Found Value of the QC Type of interest
Limit Upper limit for RPD, in %.
Lower Lower Recovery Limit, in % (except for LCSS, mg/Kg)
MDL Method Detection Limit. Same as Minimum Reporting Limit. Allows for instrument and annual fluctuations.
PCN/SCN A number assigned to reagents/standards to trace to the manufacturer's certificate of analysis
PQL Practical Quantitation Limit, typically 5 times the MDL.
QC True Value of the Control Sample or the amount added to the Spike
Rec Amount of the true value or spike added recovered, in % (except for LCSS, mg/Kg)
RPD Relative Percent Difference, calculation used for Duplicate QC Types
Upper Upper Recovery Limit, in % (except for LCSS, mg/Kg)
Sample Value of the Sample of interest
QC Sample Types
AS Analytical Spike (Post Digestion)LCSWD Laboratory Control Sample - Water Duplicate
ASD Analytical Spike (Post Digestion) Duplicate LFB Laboratory Fortified Blank
CCB Continuing Calibration Blank LFM Laboratory Fortified Matrix
CCV Continuing Calivation Verification standard LFMD Laboratory Fortified Matrix Duplicate
DUP Sample Duplicate LRB Laboratory Reagent Blank
ICB Initial Calibration Blank MS Matrix Spike
ICV Initial Calibration Verification standard MSD Matrix Spike Duplicate
ICSAB Inter-element Correction Standard - A plus B solutions PBS Prep Blank - Soil
LCSS Laboratory Control Sample - Soil PBW Prep Blank - Water
LCSSD Laboratory Control Sample - Soil Duplicate PQV Practical Quantitation Verification standard
LCSW Laboratory Control Sample - Water SDL Serial Dilution
QC Sample Type Explanations
Blanks Verifies that there is no or minimal contamination in the prep method or calibration procedure.
Control Samples Verifies the accuracy of the method, including the prep procedure.
Duplicates Verifies the precision of the instrument and/or method.
Spikes/Fortified Matrix Determines sample matrix interferences, if any.
Standard Verifies the validity of the calibration.
ACZ Qualifiers (Qual)
B Analyte concentration detected at a value between MDL and PQL.
H Analysis exceeded method hold time. pH is a field test with an immediate hold time.
U Analyte was analyzed for but not detected at the indicated MDL
Method References
(1) EPA 600/4-83-020. Methods for Chemical Analysis of Water and Wastes, March 1983.
(2) EPA 600/R-93-100. Methods for the Determination of Inorganic Substances in Environmental Samples, August 1993.
(3) EPA 600/R-94-111. Methods for the Determination of Metals in Environmental Samples - Supplement I, May 1994.
(5) EPA SW-846. Test Methods for Evaluating Solid Waste, Third Edition with Update III, December 1996.
(6) Standard Methods for the Examination of Water and Wastewater, 19th edition, 1995.
Comments
(1) QC results calculated from raw data. Results may vary slightly if the rounded values are used in the calculations.
(2) Soil, Sludge, and Plant matrices for Inorganic analyses are reported on a dry weight basis.
(3) Animal matrices for Inorganic analyses are reported on an "as received" basis.
REPIN03.02.07.01
Inorganic
Reference
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Inorganic Extended
Qualifier Report
ACZ Project ID:L64240MWH America's Inc.
ACZ ID PARAMETER QUAL DESCRIPTIONMETHODWORKNUM
RA Relative Percent Difference (RPD) was not used for data
validation because the sample concentration is too low for
accurate evaluation (< 10x MDL).
M600/2-78-054 3.2.3Neutralization Potential as CaCO3L64240-01 WG229660
RA Relative Percent Difference (RPD) was not used for data
validation because the sample concentration is too low for
accurate evaluation (< 10x MDL).
M600/2-78-054 3.2.3Neutralization Potential as CaCO3L64240-02 WG229660
RA Relative Percent Difference (RPD) was not used for data
validation because the sample concentration is too low for
accurate evaluation (< 10x MDL).
M600/2-78-054 3.2.3Neutralization Potential as CaCO3L64240-03 WG229660
RA Relative Percent Difference (RPD) was not used for data
validation because the sample concentration is too low for
accurate evaluation (< 10x MDL).
M600/2-78-054 3.2.3Neutralization Potential as CaCO3L64240-04 WG229660
RA Relative Percent Difference (RPD) was not used for data
validation because the sample concentration is too low for
accurate evaluation (< 10x MDL).
M600/2-78-054 3.2.3Neutralization Potential as CaCO3L64240-05 WG229660
REPAD.15.06.05.01
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Certification
Qualifiers
ACZ Project ID:L64240MWH America's Inc.
Soil Analysis
The following parameters are not offered for certification or are not covered by NELAC certificate #ACZ.
Neutralization Potential as CaCO3 M600/2-78-054 3.2.3
REPAD.05.06.05.01
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Sample
Receipt
ACZ Project ID:
Date Received:
Received By:
Exceptions: If you answered no to any of the above questions, please describe
NANOYES
X
X
X
X
X
X
X
X
X
X
X
X
X
1) Does this project require special handling procedures such as CLP protocol?
2) Are the custody seals on the cooler intact?
3) Are the custody seals on the sample containers intact?
4) Is there a Chain of Custody or other directive shipping papers present?
5) Is the Chain of Custody complete?
6) Is the Chain of Custody in agreement with the samples received?
7) Is there enough sample for all requested analyses?
8) Are all samples within holding times for requested analyses?
9) Were all sample containers received intact?
10) Are the temperature blanks present?
11) Are the trip blanks (VOA and/or Cyanide) present?
12) Are samples requiring no headspace, headspace free?
13) Do the samples that require a Foreign Soils Permit have one?
Contact (For any discrepancies, the client must be contacted)
Shipping Containers
Cooler Id Rad (µR/hr)Temp (°C)
Notes
Receipt Verification
8/3/2007
L64240
N/A
N/A
NA4111 22.3 15
Client must contact ACZ Project Manager if analysis should not proceed for
samples received outside of thermal preservation acceptance criteria.
MWH America's Inc.
Date Printed: 8/3/2007
REPAD.03.11.00.01
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Sample
Receipt
ACZ Project ID:
Date Received:
Received By:
Sample Container Preservation
SAMPLE R < 2 G < 2 BK < 2 Y< 2 YG< 2 B< 2 O < 2 T >12 N/A RAD
8/3/2007
L64240MWH America's Inc.
CLIENT ID ID
L64240-01 XL61917-01
L64240-02 XL61917-02
L64240-03 XL61917-03
L64240-04 XL61917-04
L64240-05 XL61917-05
Abbreviation Description Container Type Preservative/Limits
BLUE Sample Container Preservation Legend
B Filtered/Sulfuric BLUE pH must be < 2
BK Filtered/Nitric BLACK pH must be < 2
G Filtered/Nitric GREEN pH must be < 2
O Raw/Sulfuric ORANGE pH must be < 2
P Raw/NaOH PURPLE pH must be > 12 *
T Raw/NaOH Zinc Acetate TAN pH must be > 12
Y Raw/Sulfuric YELLOW pH must be < 2
YG Raw/Sulfuric YELLOW GLASS pH must be < 2
N/A No preservative needed Not applicable
RAD Gamma/Beta dose rate Not applicable must be < 250 µR/hr
R Raw/Nitric RED pH must be < 2
Sample IDs Reviewed By:
* pH check performed by analyst prior to sample preparation
REPAD.03.11.00.01
BD[
Page 1 of 13
BD[
Page 2 of 13
BD[
Page 3 of 13
BD[
Page 4 of 13
BD[
Page 5 of 13
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Report Header Explanations
Batch A distinct set of samples analyzed at a specific time
Found Value of the QC Type of interest
Limit Upper limit for RPD, in %.
Lower Lower Recovery Limit, in % (except for LCSS, mg/Kg)
MDL Method Detection Limit. Same as Minimum Reporting Limit. Allows for instrument and annual fluctuations.
PCN/SCN A number assigned to reagents/standards to trace to the manufacturer's certificate of analysis
PQL Practical Quantitation Limit, typically 5 times the MDL.
QC True Value of the Control Sample or the amount added to the Spike
Rec Amount of the true value or spike added recovered, in % (except for LCSS, mg/Kg)
RPD Relative Percent Difference, calculation used for Duplicate QC Types
Upper Upper Recovery Limit, in % (except for LCSS, mg/Kg)
Sample Value of the Sample of interest
QC Sample Types
AS Analytical Spike (Post Digestion)LCSWD Laboratory Control Sample - Water Duplicate
ASD Analytical Spike (Post Digestion) Duplicate LFB Laboratory Fortified Blank
CCB Continuing Calibration Blank LFM Laboratory Fortified Matrix
CCV Continuing Calibration Verification standard LFMD Laboratory Fortified Matrix Duplicate
DUP Sample Duplicate LRB Laboratory Reagent Blank
ICB Initial Calibration Blank MS Matrix Spike
ICV Initial Calibration Verification standard MSD Matrix Spike Duplicate
ICSAB Inter-element Correction Standard - A plus B solutions PBS Prep Blank - Soil
LCSS Laboratory Control Sample - Soil PBW Prep Blank - Water
LCSSD Laboratory Control Sample - Soil Duplicate PQV Practical Quantitation Verification standard
LCSW Laboratory Control Sample - Water SDL Serial Dilution
QC Sample Type Explanations
Blanks Verifies that there is no or minimal contamination in the prep method or calibration procedure.
Control Samples Verifies the accuracy of the method, including the prep procedure.
Duplicates Verifies the precision of the instrument and/or method.
Spikes/Fortified Matrix Determines sample matrix interferences, if any.
Standard Verifies the validity of the calibration.
ACZ Qualifiers (Qual)
B Analyte concentration detected at a value between MDL and PQL. The associated value is an estimated quantity.
H Analysis exceeded method hold time. pH is a field test with an immediate hold time.
U The material was analyzed for, but was not detected above the level of the associated value.
The associated value is either the sample quantitation limit or the sample detection limit.
Method References
(1)EPA 600/4-83-020. Methods for Chemical Analysis of Water and Wastes, March 1983.
(2)EPA 600/R-93-100. Methods for the Determination of Inorganic Substances in Environmental Samples, August 1993.
(3)EPA 600/R-94-111. Methods for the Determination of Metals in Environmental Samples - Supplement I, May 1994.
(5)EPA SW-846. Test Methods for Evaluating Solid Waste, Third Edition with Update III, December 1996.
(6)Standard Methods for the Examination of Water and Wastewater, 19th edition, 1995 & 20th edition (1998).
Comments
(1)QC results calculated from raw data. Results may vary slightly if the rounded values are used in the calculations.
(2)Soil, Sludge, and Plant matrices for Inorganic analyses are reported on a dry weight basis.
(3)Animal matrices for Inorganic analyses are reported on an "as received" basis.
For a complete list of ACZ's Extended Qualifiers, please click:http://www.acz.com/public/extquallist.pdf

REPIN03.02.07.01
Inorganic
Reference
Page 6 of 13
BD[
Page 7 of 13
BD[
Page 8 of 13
BD[
Page 9 of 13
BD[
Page 10 of 13
BD[
Page 11 of 13
Page 12 of 13
Page 13 of 13
BD[[[[
Page 1 of 20
BD[
Page 2 of 20
BD[
Page 3 of 20
BD[
Page 4 of 20
BD[
Page 5 of 20
BD[
Page 6 of 20
BD[
Page 7 of 20
BD[
Page 8 of 20
BD[
Page 9 of 20
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Report Header Explanations
Batch A distinct set of samples analyzed at a specific time
Found Value of the QC Type of interest
Limit Upper limit for RPD, in %.
Lower Lower Recovery Limit, in % (except for LCSS, mg/Kg)
MDL Method Detection Limit. Same as Minimum Reporting Limit. Allows for instrument and annual fluctuations.
PCN/SCN A number assigned to reagents/standards to trace to the manufacturer's certificate of analysis
PQL Practical Quantitation Limit, typically 5 times the MDL.
QC True Value of the Control Sample or the amount added to the Spike
Rec Amount of the true value or spike added recovered, in % (except for LCSS, mg/Kg)
RPD Relative Percent Difference, calculation used for Duplicate QC Types
Upper Upper Recovery Limit, in % (except for LCSS, mg/Kg)
Sample Value of the Sample of interest
QC Sample Types
AS Analytical Spike (Post Digestion)LCSWD Laboratory Control Sample - Water Duplicate
ASD Analytical Spike (Post Digestion) Duplicate LFB Laboratory Fortified Blank
CCB Continuing Calibration Blank LFM Laboratory Fortified Matrix
CCV Continuing Calibration Verification standard LFMD Laboratory Fortified Matrix Duplicate
DUP Sample Duplicate LRB Laboratory Reagent Blank
ICB Initial Calibration Blank MS Matrix Spike
ICV Initial Calibration Verification standard MSD Matrix Spike Duplicate
ICSAB Inter-element Correction Standard - A plus B solutions PBS Prep Blank - Soil
LCSS Laboratory Control Sample - Soil PBW Prep Blank - Water
LCSSD Laboratory Control Sample - Soil Duplicate PQV Practical Quantitation Verification standard
LCSW Laboratory Control Sample - Water SDL Serial Dilution
QC Sample Type Explanations
Blanks Verifies that there is no or minimal contamination in the prep method or calibration procedure.
Control Samples Verifies the accuracy of the method, including the prep procedure.
Duplicates Verifies the precision of the instrument and/or method.
Spikes/Fortified Matrix Determines sample matrix interferences, if any.
Standard Verifies the validity of the calibration.
ACZ Qualifiers (Qual)
B Analyte concentration detected at a value between MDL and PQL. The associated value is an estimated quantity.
H Analysis exceeded method hold time. pH is a field test with an immediate hold time.
U The material was analyzed for, but was not detected above the level of the associated value.
The associated value is either the sample quantitation limit or the sample detection limit.
Method References
(1)EPA 600/4-83-020. Methods for Chemical Analysis of Water and Wastes, March 1983.
(2)EPA 600/R-93-100. Methods for the Determination of Inorganic Substances in Environmental Samples, August 1993.
(3)EPA 600/R-94-111. Methods for the Determination of Metals in Environmental Samples - Supplement I, May 1994.
(5)EPA SW-846. Test Methods for Evaluating Solid Waste, Third Edition with Update III, December 1996.
(6)Standard Methods for the Examination of Water and Wastewater, 19th edition, 1995 & 20th edition (1998).
Comments
(1)QC results calculated from raw data. Results may vary slightly if the rounded values are used in the calculations.
(2)Soil, Sludge, and Plant matrices for Inorganic analyses are reported on a dry weight basis.
(3)Animal matrices for Inorganic analyses are reported on an "as received" basis.
For a complete list of ACZ's Extended Qualifiers, please click:http://www.acz.com/public/extquallist.pdf

REPIN03.02.07.01
Inorganic
Reference
Page 10 of 20
BD[
Page 11 of 20
BD[
Page 12 of 20
BD[
Page 13 of 20
BD[
Page 14 of 20
BD[
Page 15 of 20
BD[
Page 16 of 20
BD[
Page 17 of 20
BD[
Page 18 of 20
Page 19 of 20
Page 20 of 20
BD[
Page 1 of 27
BD[
Page 2 of 27
BD[
Page 3 of 27
BD[
Page 4 of 27
BD[
Page 5 of 27
BD[
Page 6 of 27
BD[
Page 7 of 27
BD[
Page 8 of 27
BD[
Page 9 of 27
BD[
Page 10 of 27
BD[
Page 11 of 27
BD[
Page 12 of 27
BD[
Page 13 of 27
BD[
Page 14 of 27
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Report Header Explanations
Batch A distinct set of samples analyzed at a specific time
Found Value of the QC Type of interest
Limit Upper limit for RPD, in %.
Lower Lower Recovery Limit, in % (except for LCSS, mg/Kg)
MDL Method Detection Limit. Same as Minimum Reporting Limit. Allows for instrument and annual fluctuations.
PCN/SCN A number assigned to reagents/standards to trace to the manufacturer's certificate of analysis
PQL Practical Quantitation Limit, typically 5 times the MDL.
QC True Value of the Control Sample or the amount added to the Spike
Rec Amount of the true value or spike added recovered, in % (except for LCSS, mg/Kg)
RPD Relative Percent Difference, calculation used for Duplicate QC Types
Upper Upper Recovery Limit, in % (except for LCSS, mg/Kg)
Sample Value of the Sample of interest
QC Sample Types
AS Analytical Spike (Post Digestion)LCSWD Laboratory Control Sample - Water Duplicate
ASD Analytical Spike (Post Digestion) Duplicate LFB Laboratory Fortified Blank
CCB Continuing Calibration Blank LFM Laboratory Fortified Matrix
CCV Continuing Calibration Verification standard LFMD Laboratory Fortified Matrix Duplicate
DUP Sample Duplicate LRB Laboratory Reagent Blank
ICB Initial Calibration Blank MS Matrix Spike
ICV Initial Calibration Verification standard MSD Matrix Spike Duplicate
ICSAB Inter-element Correction Standard - A plus B solutions PBS Prep Blank - Soil
LCSS Laboratory Control Sample - Soil PBW Prep Blank - Water
LCSSD Laboratory Control Sample - Soil Duplicate PQV Practical Quantitation Verification standard
LCSW Laboratory Control Sample - Water SDL Serial Dilution
QC Sample Type Explanations
Blanks Verifies that there is no or minimal contamination in the prep method or calibration procedure.
Control Samples Verifies the accuracy of the method, including the prep procedure.
Duplicates Verifies the precision of the instrument and/or method.
Spikes/Fortified Matrix Determines sample matrix interferences, if any.
Standard Verifies the validity of the calibration.
ACZ Qualifiers (Qual)
B Analyte concentration detected at a value between MDL and PQL. The associated value is an estimated quantity.
H Analysis exceeded method hold time. pH is a field test with an immediate hold time.
U The material was analyzed for, but was not detected above the level of the associated value.
The associated value is either the sample quantitation limit or the sample detection limit.
Method References
(1)EPA 600/4-83-020. Methods for Chemical Analysis of Water and Wastes, March 1983.
(2)EPA 600/R-93-100. Methods for the Determination of Inorganic Substances in Environmental Samples, August 1993.
(3)EPA 600/R-94-111. Methods for the Determination of Metals in Environmental Samples - Supplement I, May 1994.
(5)EPA SW-846. Test Methods for Evaluating Solid Waste, Third Edition with Update III, December 1996.
(6)Standard Methods for the Examination of Water and Wastewater, 19th edition, 1995 & 20th edition (1998).
Comments
(1)QC results calculated from raw data. Results may vary slightly if the rounded values are used in the calculations.
(2)Soil, Sludge, and Plant matrices for Inorganic analyses are reported on a dry weight basis.
(3)Animal matrices for Inorganic analyses are reported on an "as received" basis.
For a complete list of ACZ's Extended Qualifiers, please click:http://www.acz.com/public/extquallist.pdf

REPIN03.02.07.01
Inorganic
Reference
Page 15 of 27
BD[
Page 16 of 27
BD[
Page 17 of 27
BD[
Page 18 of 27
BD[
Page 19 of 27
BD[
Page 20 of 27
BD[
Page 21 of 27
Page 22 of 27
Page 23 of 27
Page 24 of 27
Page 25 of 27
Page 26 of 27
Page 27 of 27
BD[
Page 1 of 26
BD[
Page 2 of 26
BD[
Page 3 of 26
BD[
Page 4 of 26
BD[
Page 5 of 26
BD[
Page 6 of 26
BD[
Page 7 of 26
BD[
Page 8 of 26
BD[
Page 9 of 26
BD[
Page 10 of 26
BD[
Page 11 of 26
BD[
Page 12 of 26
BD[
Page 13 of 26
BD[
Page 14 of 26
BD[
Page 15 of 26
BD[
Page 16 of 26
BD[
Page 17 of 26
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Report Header Explanations
Batch A distinct set of samples analyzed at a specific time
Found Value of the QC Type of interest
Limit Upper limit for RPD, in %.
Lower Lower Recovery Limit, in % (except for LCSS, mg/Kg)
MDL Method Detection Limit. Same as Minimum Reporting Limit. Allows for instrument and annual fluctuations.
PCN/SCN A number assigned to reagents/standards to trace to the manufacturer's certificate of analysis
PQL Practical Quantitation Limit, typically 5 times the MDL.
QC True Value of the Control Sample or the amount added to the Spike
Rec Amount of the true value or spike added recovered, in % (except for LCSS, mg/Kg)
RPD Relative Percent Difference, calculation used for Duplicate QC Types
Upper Upper Recovery Limit, in % (except for LCSS, mg/Kg)
Sample Value of the Sample of interest
QC Sample Types
AS Analytical Spike (Post Digestion)LCSWD Laboratory Control Sample - Water Duplicate
ASD Analytical Spike (Post Digestion) Duplicate LFB Laboratory Fortified Blank
CCB Continuing Calibration Blank LFM Laboratory Fortified Matrix
CCV Continuing Calibration Verification standard LFMD Laboratory Fortified Matrix Duplicate
DUP Sample Duplicate LRB Laboratory Reagent Blank
ICB Initial Calibration Blank MS Matrix Spike
ICV Initial Calibration Verification standard MSD Matrix Spike Duplicate
ICSAB Inter-element Correction Standard - A plus B solutions PBS Prep Blank - Soil
LCSS Laboratory Control Sample - Soil PBW Prep Blank - Water
LCSSD Laboratory Control Sample - Soil Duplicate PQV Practical Quantitation Verification standard
LCSW Laboratory Control Sample - Water SDL Serial Dilution
QC Sample Type Explanations
Blanks Verifies that there is no or minimal contamination in the prep method or calibration procedure.
Control Samples Verifies the accuracy of the method, including the prep procedure.
Duplicates Verifies the precision of the instrument and/or method.
Spikes/Fortified Matrix Determines sample matrix interferences, if any.
Standard Verifies the validity of the calibration.
ACZ Qualifiers (Qual)
B Analyte concentration detected at a value between MDL and PQL. The associated value is an estimated quantity.
H Analysis exceeded method hold time. pH is a field test with an immediate hold time.
U The material was analyzed for, but was not detected above the level of the associated value.
The associated value is either the sample quantitation limit or the sample detection limit.
Method References
(1)EPA 600/4-83-020. Methods for Chemical Analysis of Water and Wastes, March 1983.
(2)EPA 600/R-93-100. Methods for the Determination of Inorganic Substances in Environmental Samples, August 1993.
(3)EPA 600/R-94-111. Methods for the Determination of Metals in Environmental Samples - Supplement I, May 1994.
(5)EPA SW-846. Test Methods for Evaluating Solid Waste, Third Edition with Update III, December 1996.
(6)Standard Methods for the Examination of Water and Wastewater, 19th edition, 1995 & 20th edition (1998).
Comments
(1)QC results calculated from raw data. Results may vary slightly if the rounded values are used in the calculations.
(2)Soil, Sludge, and Plant matrices for Inorganic analyses are reported on a dry weight basis.
(3)Animal matrices for Inorganic analyses are reported on an "as received" basis.
For a complete list of ACZ's Extended Qualifiers, please click:http://www.acz.com/public/extquallist.pdf

REPIN03.02.07.01
Inorganic
Reference
Page 18 of 26
BD[
Page 19 of 26
BD[
Page 20 of 26
BD[
Page 21 of 26
BD[
Page 22 of 26
BD[
Page 23 of 26
BD[
Page 24 of 26
Page 25 of 26
Page 26 of 26
BD[
Page 1 of 25
BD[
Page 2 of 25
BD[
Page 3 of 25
BD[
Page 4 of 25
BD[
Page 5 of 25
BD[
Page 6 of 25
BD[
Page 7 of 25
BD[
Page 8 of 25
BD[
Page 9 of 25
BD[
Page 10 of 25
BD[
Page 11 of 25
BD[
Page 12 of 25
BD[
Page 13 of 25
BD[
Page 14 of 25
BD[
Page 15 of 25
BD[
Page 16 of 25
BD[
Page 17 of 25
BD[
Page 18 of 25
ACZ Laboratories, Inc.
2773 Downhill Drive Steamboat Springs, CO 80487 [phone]
Report Header Explanations
Batch A distinct set of samples analyzed at a specific time
Found Value of the QC Type of interest
Limit Upper limit for RPD, in %.
Lower Lower Recovery Limit, in % (except for LCSS, mg/Kg)
MDL Method Detection Limit. Same as Minimum Reporting Limit. Allows for instrument and annual fluctuations.
PCN/SCN A number assigned to reagents/standards to trace to the manufacturer's certificate of analysis
PQL Practical Quantitation Limit, typically 5 times the MDL.
QC True Value of the Control Sample or the amount added to the Spike
Rec Amount of the true value or spike added recovered, in % (except for LCSS, mg/Kg)
RPD Relative Percent Difference, calculation used for Duplicate QC Types
Upper Upper Recovery Limit, in % (except for LCSS, mg/Kg)
Sample Value of the Sample of interest
QC Sample Types
AS Analytical Spike (Post Digestion)LCSWD Laboratory Control Sample - Water Duplicate
ASD Analytical Spike (Post Digestion) Duplicate LFB Laboratory Fortified Blank
CCB Continuing Calibration Blank LFM Laboratory Fortified Matrix
CCV Continuing Calibration Verification standard LFMD Laboratory Fortified Matrix Duplicate
DUP Sample Duplicate LRB Laboratory Reagent Blank
ICB Initial Calibration Blank MS Matrix Spike
ICV Initial Calibration Verification standard MSD Matrix Spike Duplicate
ICSAB Inter-element Correction Standard - A plus B solutions PBS Prep Blank - Soil
LCSS Laboratory Control Sample - Soil PBW Prep Blank - Water
LCSSD Laboratory Control Sample - Soil Duplicate PQV Practical Quantitation Verification standard
LCSW Laboratory Control Sample - Water SDL Serial Dilution
QC Sample Type Explanations
Blanks Verifies that there is no or minimal contamination in the prep method or calibration procedure.
Control Samples Verifies the accuracy of the method, including the prep procedure.
Duplicates Verifies the precision of the instrument and/or method.
Spikes/Fortified Matrix Determines sample matrix interferences, if any.
Standard Verifies the validity of the calibration.
ACZ Qualifiers (Qual)
B Analyte concentration detected at a value between MDL and PQL. The associated value is an estimated quantity.
H Analysis exceeded method hold time. pH is a field test with an immediate hold time.
U The material was analyzed for, but was not detected above the level of the associated value.
The associated value is either the sample quantitation limit or the sample detection limit.
Method References
(1)EPA 600/4-83-020. Methods for Chemical Analysis of Water and Wastes, March 1983.
(2)EPA 600/R-93-100. Methods for the Determination of Inorganic Substances in Environmental Samples, August 1993.
(3)EPA 600/R-94-111. Methods for the Determination of Metals in Environmental Samples - Supplement I, May 1994.
(5)EPA SW-846. Test Methods for Evaluating Solid Waste, Third Edition with Update III, December 1996.
(6)Standard Methods for the Examination of Water and Wastewater, 19th edition, 1995 & 20th edition (1998).
Comments
(1)QC results calculated from raw data. Results may vary slightly if the rounded values are used in the calculations.
(2)Soil, Sludge, and Plant matrices for Inorganic analyses are reported on a dry weight basis.
(3)Animal matrices for Inorganic analyses are reported on an "as received" basis.
For a complete list of ACZ's Extended Qualifiers, please click:http://www.acz.com/public/extquallist.pdf

REPIN03.02.07.01
Inorganic
Reference
Page 19 of 25
BD[
Page 20 of 25
BD[
Page 21 of 25
BD[
Page 22 of 25
BD[
Page 23 of 25
Pa
g
e

2
4

o
f

2
5
Pa
g
e

2
5

o
f

2
5

APPENDIX B

LABORATORY REPORT WITH UNSATURATED AND
SATURATED HYDRAULIC PROPERTIES OF THE BEDROCK
CORE SAMPLES

APPENDIX C
BEDROCK SAMPLING TO CHARACTERIZE HYDRAULIC AND
GEOCHEMICAL PROPERTIES OF THE VADOSE ZONE

C-1
APPENDIX C
BEDROCK SAMPLING TO CHARACTERIZE HYDRAULIC AND
GEOCHEMICAL PROPERTIES OF THE VADOSE ZONE
The purpose of this appendix is to describe the approach used to collect samples of
bedrock material from the vadose zone beneath (immediately adjacent to) the White
Mesa Mill tailings cells for characterization of hydraulic and geochemical properties.
Hydraulic properties are used to predict the flow of water through the vadose zone, while
geochemical properties are used to predict water-rock chemical reactions as the tailings
pore water potentially migrates beneath the tailings cells. Geochemical properties tested
include mass concentrations of hydrous ferric oxide (HFO) and acid neutralization
potential (ANP). The mass of ANP is used in the vadose zone reactive transport model to
predict the consumption of alkalinity (neutralization front) as low-pH tailings pore water
potentially migrates beneath the tailings cells, while the mass of HFO is used to predict
surface complexation (adsorption) reactions. Soil water retention and unsaturated
hydraulic conductivity curves are presented and used to identify hydrologic units, while a
statistical analysis of the geochemicical data is presented and used to identify
geochemical units. Lithologic data combined with the hydrologic and geochemical data
form the basis for assigning hydrogeochemical stratigraphic units within the vadose zone.
VADOSE ZONE CORE
Samples of bedrock in the vadose zone were obtained from core that is stored in core
boxes in a storage shed at the White Mesa Mill. Core drilling was performed in 2005
prior to installing monitoring wells. The sample locations and depth intervals selected for
characterization were based on the location with respect to the tailings cells and the
availability of recovered core material. Monitoring wells that were located immediately
adjacent to or between cells 2 & 3 were selected preferentially over core from more
distant locations.

C-2
Availability of Core
There were five wells that had core available in the vicinity of the tailings cells: MW-23,
MW-24, MW-28, MW-30, and TW4-22 (see Figure C-1 for locations). Intervals of
available core are listed in Table C-1. No samples were selected from MW-28 because of
the paucity of recovered core. A cross section illustrating the monitoring wells with
available core and the depth intervals of samples selected for hydrogeologic and
geochemical characterization is plotted in Figure C-2. Sample interval depths are
measured in feet below ground surface (ft bgs). Core was not retrieved from
approximately 0 to 20 ft bgs because the material consisted of native unconsolidated
surficial soils, native unconsolidated windblown silt and sand, and unconsolidated soils
used to construct the tailings cell berms. Recovery of core greater than this depth varied
between holes. All core measured 2 inches in diameter.
Lithologic Descriptions
Rock core from wells in the vicinity of the tailings cells were logged for rock type, grain
size, color, bedding/lamination, staining, and induration. Geologic logs and lithologic
descriptions of samples submitted for hydrogeologic and geochemical characterization
are summarized in Table C-2. Core included material from the Dakota Sandstone, Burro
Canyon Formation, and Brushy Basin Member (shale) of the Morrison Formation;
however, during inspection of the core material, a distinct contact between the Dakota
Sandstone and Burro Canyon Formation could not be identified. Therefore, the objective
of the characterization program was to select samples at appropriate intervals to quantify
subsurface heterogeneity in terms of hydraulic and geochemical properties.
Geologic logs indicate that the predominate lithology between the bottom of the tailings
cells (~30 ft bgs) and the top of the Brushy Basin Member shale (~103-127 ft bgs) is
sandstone (Dakota Sandstone and Burro Canyon Formation) with one or two minor
lenses of siltstone and conglomerate. On average, there is approximately 3 feet of
siltstone and 4 feet of conglomerate (see Figure C-2). Interpretation of the geologic logs

C-3
for the monitoring wells suggests the presence of four geologic layers within the vadose
zone:
• Upper sandstone (~30-56 ft bgs)
• Conglomerate or gravelly sandstone (56-60 ft bgs)
• Siltstone (60-63 ft bgs)
• Lower sandstone (63 ft bgs-bottom).
The geologic layers are included here to facilitate the statistical evaluation of the
geochemical properties, and identification of hydrogeochemical stratigraphic units.
VADOSE ZONE DEPTH BENEATH THE TAILINGS CELLS
The vadose zone thickness beneath the tailings cells is tabulated in Table C-3. The
vadose zone thickness is calculated by taking the difference between the bottom elevation
of the cell and the distance to the water table. The minimum vadose zone thickness
beneath Cells 2 & 3 and Cell 4A is approximately 42 ft and 40 ft, respectively. As a
comparison, the average vadose zone thickness beneath Cell 2, Cell 3, and Cell 4A were
63 ft, 66 ft, and 56 ft. For the vadose zone transport models, the vadose zone thickness
beneath Cells 2 &3 and Cells 4A & 4B was assumed to equal 42 ft (12.8 m).
VADOSE ZONE CORE SAMPLING METHODOLOGY & LABORATORY
ANALYSIS METHODS
Sampling for Geochemical Properties
Samples of the core for geochemical analysis were collected at approximately 5-foot
centers beginning from approximately 30 ft bgs (selected to correspond with the
approximate base of the tailings cells) and extending toward the water table. At depths

C-4
greater than about 65 ft bgs, samples were collected approximately every 7 to 10 feet.
The exact spacing varied depending on availability of core and the necessity to
characterize different rock types based on grain size distribution and degree of sorting.
Generally, the core samples collected for geochemical characterization were 0.2 feet
long. Geochemical properties, including the amount of HFO and ANP, were evaluated
from laboratory analyses of 34 core intervals (34 primary samples with four duplicates
for 38 total analyses reported) selected from four monitoring wells (MW-23, MW-24,
MW-30, and TW4-22).
The following numbers of samples were collected and analyzed for geochemical
properties from the four geologic layers observed in the cores above the Brushy Basin
shale (described above):
• 18 within the upper sandstone unit
• 4 within the conglomerate unit
• 3 within the siltstone unit
• 9 within the lower sandstone unit.
The arithmetic average for each of the four sample intervals analyzed in duplicate was
used in the statistical analysis. The upper sandstone unit contained more samples because
of its proximity to the bottom of the tailings cells. In general, more sandstone samples
were collected because this lithology was the dominant rock type, with the upper and
lower sandstones having average thickness of 26 feet and greater than 37 feet,
respectively, relative to the conglomerate and siltstone units which average 4 feet and 3
feet thick, respectively.
Mass of HFO. The mass of hydrous ferric oxide present in bedrock core samples was
determined via chemical extraction with hydroxylamine-hydrochloride (HH) solution.
The procedure consisted of subjecting samples of crushed bedrock to short-term leaching

C-5
to completely dissolve amorphous-mineral phases (e.g., ferrihydrite/hydrous ferric oxide)
and partially dissolve some crystalline minerals (e.g., goethite). The solution acts as a
reducing agent converting iron from the solid phase (Fe+3) to an aqueous phase (Fe+2).
The leaching procedure used to obtain the mass of HFO was similar to the approach
adopted for the Naturita UMTRA Site as prepared by the U.S. Geological Survey for the
U.S. Nuclear Regulatory Commission as part of NUREG/CR-6820 (Davis and Curtis,
2003; Davis et al., 2004). The approach used in NUREG/CR-6820 to estimate the mass
of HFO was also similar to an approach adopted by the U.S. EPA Environmental
Research Laboratory (Loux et al., 1989).
The procedure consisted of drying the core at 34°C followed by crushing to remove
material larger than 3 mm. The HH solution (100 mL) was then added to 10 grams of
crushed rock in a 250 mL bottle and placed in a shaking-water bath at 50°C. Aliquots of
extracted solution were withdrawn after 96 hours and filtered (<0.45 μm) prior to
analysis.
The concentration of dissolved iron reported in the leachate (in milligrams per liter) was
then converted to the mass of iron (in milligrams of iron per kilogram of rock) originally
present in the rock sample by multiplying the solution concentration by the water to rock
proportion (0.1 liter divided by 0.01 kilograms). The concentration of iron in the rock
can then be converted to the concentration of HFO in the rock by multiplication of the
ratio of the molecular weight of HFO divided by the molecular weight of iron (89 grams
per mole divided by 55.8 grams per mole). A statistical analysis of iron concentrations in
the leachate is presented as part of the results below (the original laboratory data sheets
are contained within Appendix A).
HFO is the only solid phase that is credited as a potential sorption site of uranium and
other trace elements, which is a conservative assumption because other phases (e.g.,
hematite, quartz, clays, etc.) also participate in surface complexation reactions.
Mass of ANP. The mass of acid neutralization potential (assumed to be calcite) present
in the bedrock was measured directly using the methodology outlined in U.S. EPA

C-6
method M600/2-78-054. The test involves placing pulverized material into an acid bath,
which is heated until the reaction ceases as noted by the absence of bubbling from the
release of CO2 gas. The neutralization potential is then determined by titrating with a
base to determine how much of the acid was neutralized by fast-reacting calcium-bearing
carbonate minerals. The data are reported as grams of calcite per kilogram of rock. The
ANP data are considered to be representative because the test only measures fast-reacting
carbonate minerals. A statistical analysis of the ANP data is included as part of the
results below (the original laboratory data sheets are contained within Appendix A).
Hydraulic Properties
Five sandstone samples were analyzed for hydraulic properties. One sample, considered
to represent the transition from siltstone to sandstone (MW-23 74.3-74.6 ft), was also
analyzed for hydraulic properties. No conglomerate layers were analyzed for hydraulic
properties because the core samples from this rock type generally consist of irregular
shaped, angular pieces with variable sorting and clast sizes. Furthermore, it is likely that
the conglomerate behaves hydraulically very similarly to the sandstone because the
matrix is sandstone and the clasts are generally small gravel in low percentages (less than
30%). The core samples collected for hydraulic characterization varied between 0.2 to
0.5 feet long.
Soil water retention characteristics for the vadose zone samples were determined by a
variety of methods including hanging column, pressure plate, water activity meter, and
relative humidity box methods to cover a range of pressure heads from 0 cm (saturated
water content) to -851,000 cm (residual water content). The saturated hydraulic
conductivity of the samples was measured using a flexible wall permeameter. The test
methods, and original laboratory report, are included in Appendix B.

C-7
VADOSE ZONE CORE SAMPLE RESULTS
Geochemical Properties
The analytical results of the geochemical test data for ANP and HFO (represented as
dissolved iron concentrations in the leachate) are posted on Figures C-3 and C-4,
respectively. Statistical analyses of the geochemical data for both ANP and HFO,
including number of data points, minimum and maximum values, median, arithmetic
average (mean), arithmetic standard deviation, geometric mean, and geometric standard
deviation are tabulated in Table C-4. The statistical data are grouped by different
attributes to determine whether any statistical trends can be identified. The groupings
included categorization on a well-by-well basis, geologic layers, by depth assuming three
arbitrary 25-foot thick zones, and for the entire population.
An analysis of the distribution of the data indicated that the geochemical properties
within the vadose zone are distributed log normally. Therefore the geometric mean,
rather than the arithmetic mean, should be used to guide the comparison between attribute
groups. Overall, the geometric mean ANP did not vary significantly between the
different attribute groups. The geometric mean HFO did vary slightly between lithologic
groups and between wells with higher concentrations reported for the siltstone unit and
for MW-30 and TW4-22. To simplify the conceptual model, the geometric mean of the
entire population was selected as the base case value for both ANP and HFO.
To support the sensitivity analysis, and determine a range of values for the amount of
ANP, the geometric mean plus one geometric standard deviation was selected as an upper
bound, while the geometric mean minus one geometric standard deviation was selected as
the lower bound. The geometric mean plus one geometric standard deviation of the log
transformed data corresponds to approximately 68% of the observations. The amount of
HFO was not included in the sensitivity analysis because the three sets of parameters
were not significantly different (upper bound equal to 59.8 mg/L iron; base case equal to
55.4 mg/L iron; lower bound equal to 51.0 mg/L iron).

C-8
Hydraulic Properties
The unsaturated and saturated hydraulic properties for the vadose zone samples were
estimated from laboratory measurements and through optimization. Parameter values
measured in the lab on vadose zone samples included the residual water content (θr,),
saturated water content (θs), water contents at intermediate soil water pressures θ(h), and
the saturated hydraulic conductivity in the vertical direction (Ks). Parameter values
determined through optimization included the empirical fitting parameters (n and α). The
fitting parameters are considered to be empirical coefficients that affect the shape of the
hydraulic functions used to describe variations in water content and hydraulic
conductivity for different soil water pressures. The unsaturated hydraulic properties
[θ(h)] and [K(h)] are highly nonlinear functions of the soil water pressure (h).
Unsaturated hydraulic properties (parameters θs, α, and n) were determined by fitting van
Genuchten’s (1980) analytical model to the water retention data using the RETC
computer code developed by the U.S. Salinity Laboratory for the U.S. EPA (van
Genuchten et al., 1991). RETC utilizes a nonlinear, least-squares parameter optimization
method to estimate the unknown variables. During parameter optimization, the program
is run many times in succession, each time incrementally varying the unknown variables
so as to minimize the sum of squared residuals until convergence is reached and the
measured data are matched. Data collected from MW-30 (44.0-44.5 ft) were not included
because the core experienced swelling and cracking after the saturated hydraulic
conductivity test.
A comparison between the measured and model-predicted soil water retention curves for
the vadose zone samples are plotted in Figure C-5. Overall, there is good agreement
between the measured and optimized parameter values used to describe θ(h). The
hydraulic properties of the vadose zone samples are included in Table C-5. Justification
for selection of hydrogeologic parameters for the vadose zone is discussed below.

C-9
A single set of hydraulic properties for the vadose zone assuming the properties of
sandstone was used as input to the model. This assumption is considered appropriate
because the saturated and unsaturated hydraulic properties of the samples are quite
similar to one another (Figures C-5 and C-6) with the exception of MW-23 (74.3-74.6 ft).
MW-23 (74.3-74.6 ft) had a smaller storage capacity and a slightly lower saturated
hydraulic conductivity compared to the other samples. Assignment of a single set of
hydrogeologic properties should not significantly affect the model results given the
similarity in unsaturated hydraulic properties [θ(h)] and [K(h)] for all samples (i.e., there
were no large differences in soil water retention curves or unsaturated hydraulic
conductivity curves for the materials tested). The hydraulic properties from MW-23
(55.5-56.0 ft) were used as input to the model because the hydraulic functions are
intermediate as compared to the other samples. Unsaturated hydraulic conductivity of the
vadose zone was not included in the sensitivity analysis because the unsaturated
hydraulic conductivities vary to match flux rates under a unit hydraulic gradient.
CONCLUSIONS
The statistical evaluation of the geochemical data supports the assignment of one
geochemical unit within the vadose zone. For ANP, the geometric mean of the
population was used to establish the base case value, while the geometric mean plus one
geometric standard deviation, and minus one geometric standard deviation, were used to
establish the upper and lower bounds. For HFO, only the base case scenario will be
simulated because the upper and lower bounds, calculated using the same approach as for
ANP, were similar. Evaluation of the soil water retention and unsaturated hydraulic
conductivity curves for the core samples indicated that the one set of parameters could be
used to describe water flow through the vadose zone. Taken together, one
hydrogeochemical unit will be used as input to characterize the flow and transport of
solutes through the vadose zone.

C-10
REFERENCES
D’Appolonia Consulting Engineers, Inc., 1982. Construction Report, Initial Phase –
Tailings Management System, White Mesa Uranium Project, Blanding, Utah.
Prepared for Energy Fuels Nuclear, Inc.

Davis, J.A. and G.P. Curtis, 2003. Application of Surface Complexation Modeling to
Describe Uranium(VI) Adsorption and Retardation at the Uranium Mill Tailings Site
at Naturita, Colorado, Report NUREG CR-6820, U. S. Nuclear Regulatory
Commission, Rockville, Maryland, pp. 223.

Davis, J.A., D.E. Meece, M. Kohler, and G.P. Curtis, 2004. Approaches to surface
complexation modeling of Uranium(VI) adsorption on aquifer sediments, Geochimica
et Cosmochimica Acta, 68, 18, 3621-3641.

Geosyntec Consultants, 2007. Revised Construction Drawings, DMC White Mesa Mill,
Cell 4A Lining System. Report prepared for Denison Mines (USA) Corp. June 2007.

Loux, N.T., D.S. Brown, C.R. Chafin, J.D. Allison, and S.M. Hassan, 1989. Chemical
speciation and competitive cationic partitioning on a sandy aquifer material,
Chemical Speciation and Bioavailability, 1, 3, 111-125.

van Genuchten, M.Th., 1980. A closed-form equation for predicting the hydraulic
conductivity of unsaturated soils, Soil Sci. Am. Jour., 44, 892-898.

van Genuchten, M.Th., J. Simunek, F.J. Leij, and M. Sejna, 1991. The RETC code for
quantifying the hydraulic functions of unsaturated soils, Version 6.0., EPA Report
600/2-91/065, U.S. Salinity Laboratory, USDA, ARS, Riverside, California.

C-11

TABLE C‐1. AVAILABLE INTERVALS OF CORE FOR MONITORING WELLS ADJACENT TO THE TAILINGS 
CELLS. 
MW‐23 MW‐24 MW‐28 MW‐30 TP4‐22 
49‐86 20‐71 54‐60 27‐46 20‐63 
87‐90 73‐98 78‐80 50‐60 68‐70 
96‐110 99‐105 84‐90 ‐ 76‐80 
111‐132 108‐120 91.5‐110 ‐ 82‐100 
‐ ‐ ‐ ‐ 105‐115 
Note:  All depths are measured in feet below ground surface.   

C-12

TABLE C‐2. GEOLOGIC LOGS OF SAMPLES SUBMITTED FOR HYDRAULIC AND GEOCHEMICAL CHARACTERIZATION. 
Monitoring 
Well 
Depth 
Interval 
(ft bgs) 
Sample 
Collection 
Date 
Rock Type Grain Size Sorting Induration Color Comments Analysis 
MW‐23 49.3‐49.5 May‐09 Qtz Sandstone fine to medium well well very pale brown 10YR 7/3 laminations Chem. 
MW‐23 53.0‐53.5 Feb‐07 Qtz Sandstone fine to medium 
(predominately fine) moderately well very pale brown 10YR 8/3 cross bedding, Fe‐Mg grains 5% Chem. 
MW‐23 55.5‐56.0 Feb‐07 Qtz Sandstone fine to medium 
(predominately fine) moderately well very pale brown 10YR 8/3 cross bedding, Fe‐Mg grains 5% Hyd. 
MW‐23 59.3‐59.5 May‐09 Qtz Sandstone fine to medium well well very pale brown 10YR 7/3 laminations Chem. 
MW‐23 63.8‐64.0 May‐09 Qtz Sandstone fine to coarse poorly well very pale brown 10YR 8/3 small weathered feldspar grains (5‐
10%), small iron stains Chem. 
MW‐23 68.9‐69.3 May‐09 Conglomerate fine to coarse sand and 
fine gravel poorly moderately very pale brown 10YR 8/2 clasts of chert (qtz), weathered 
feldspar, and limestone (?) Chem. 
MW‐23 74.0‐74.3 Feb‐07 Qtz Sandstone very fine to fine moderately well white 10YR 8/1 Fe‐Mg grains 1%, iron staining Chem. 
MW‐23 74.3‐74.6 Feb‐07 Qtz Sandstone very fine to fine moderately well white 10YR 8/1 Fe‐Mg grains 1%, iron staining Hyd. 
MW‐23 82.5‐82.7 May‐09 Qtz Sandstone fine to medium 
(predominately fine) moderately well white 10YR 8/1 Fe‐Mg grains <1% Chem. 
MW‐23 82.7‐82.9 Feb‐07 Qtz Sandstone fine to medium 
(predominately fine) moderately well white 10YR 8/1 Fe‐Mg grains <1% Hyd. 
MW‐23 89.9‐90.0 May‐09 Qtz Sandstone fine to coarse sand and 
fine gravel 
poorly 
sorted moderately very pale brown 10YR 8/2 small weathered feldspar grains (5‐
10%), small iron stains Chem. 
MW‐23 99.8‐100.0 Feb‐07 Qtz Sandstone very fine to fine moderately not well 
very pale brown where stained 
10YR 7/4 and light gray where 
not stained 10YR 7/2 
Fe‐Mg <1%, much iron staining Chem. 
MW‐23 103.0‐
103.3 Feb‐07 Qtz Sandstone fine to medium moderately well white 10YR 8/1 Fe‐Mg <1% Chem. 
MW‐23 103.3‐
103.5 Feb‐07 Qtz Sandstone fine to medium moderately well white 10YR 8/1 Fe‐Mg <1% Hyd. 
MW‐30 31.3‐31.5 May‐09 Qtz Sandstone fine to coarse poorly well yellowish brown 10YR 5/4 
large weathered feldspar and chert 
clasts (10%), matrix is 50% 
weathered mafics 
Chem. 
MW‐30 35.5‐36.0 Feb‐07 Qtz Sandstone 
fine to medium 
(predominately 
medium) 
moderately well yellowish brown 10YR 7/4  ‐‐ Hyd. 
MW‐30 37.5‐38.0 Feb‐07 Qtz Sandstone 
fine to medium 
(predominately 
medium) 
moderately well yellowish brown 10YR 7/5 Fe‐Mg grains 5% Chem. 
MW‐30 43.0‐43.2 Feb‐07 Qtz Sandstone fine to medium 
(predominately fine) moderately well white 10YR 8/2 Fe‐Mg grains <1%, some iron 
staining Chem. 
MW‐30 43.2‐43.5 Feb‐07 Qtz Sandstone fine to medium 
(predominately fine) moderately well white 10YR 8/2 Fe‐Mg grains <1%, some iron 
staining Chem. 

C-13
TABLE C‐2. GEOLOGIC LOGS OF SAMPLES SUBMITTED FOR HYDRAULIC AND GEOCHEMICAL CHARACTERIZATION. (continued) 
Monitoring 
Well 
Depth 
Interval 
(ft bgs) 
Sample 
Collection 
Date 
Rock Type Grain Size Sorting Induration Color Comments Analysis 
MW‐30 44.0‐44.5 Feb‐07 Qtz Sandstone fine to medium 
(predominately fine) moderately well white 10YR 8/2 Fe‐Mg grains <1%, some iron 
staining Hyd. 
MW‐30 50.0‐50.2 May‐09 Qtz Sandstone very fine w. silt well well yellow 10YR 8/6 ‐‐ Chem. 
MW‐30 53.9‐54.0 May‐09 Siltstone silt w. fine sand  ‐‐ moderately weak red 10R 5/3  ‐‐ Chem. 
MW‐30 59.8‐60.0 May‐09 Qtz Sandstone/
Conglomerate 
fine to coarse w. small 
clasts poorly well very pale brown 10YR 7/4 chert and weathered feldspar 
clasts (very fine conglomerate) Chem. 
MW‐24 29.7‐29.9 May‐09 Silty Sandstone very fine sand w. silt well well white 7.5YR 8/1 grading to 
reddish yellow 7.5YR 7/6 
sample is iron stained and grades 
from white to orange Chem. 
MW‐24 34.9‐35.1 May‐09 Qtz Sandstone fine to medium well well very pale brown 10YR 8/2 mafic grains (5‐10%) Chem. 
MW‐24 40.0‐40.2 May‐09 Qtz Sandstone fine to coarse poorly well very pale brown 10YR 8/2 weathered mafic and feldspar 
grains (5%) Chem. 
MW‐24 44.7‐44.9 May‐09 Qtz Sandstone fine to coarse poorly well white 10YR 8/1 numerous large weathered 
feldspar grains (10%) Chem. 
MW‐24 49.8‐49.9 May‐09 Qtz Sandstone fine to medium well well white 10YR 8/1 weathered feldspar grains (5%) Chem. 
MW‐24 56.0‐56.2 May‐09 Siltstone silt w. fine sand  ‐‐ well light greenish grey 5GY 7/1  ‐‐ Chem. 
MW‐24 63.4‐63.5 May‐09 Siltstone silt w. fine sand  ‐‐ well pale yellow 5Y 8/3 ‐‐ Chem. 
MW‐24 73.0‐73.2 May‐09 Qtz Sandstone fine to medium well well white 10YR 8/1 
small zone of poorly sorted sand to 
very coarse w. weathered 
feldspars 
Chem. 
MW‐24 80.0‐80.3 May‐09 Qtz Sandstone very fine to fine well moderately white 10YR 8/1  Chem. 
TW4‐22 34.0‐34.2 May‐09 Qtz Sandstone fine to medium well well very pale brown 10YR 7/4 ‐‐ Chem. 
TW4‐22 39.4‐39.6 May‐09 Qtz Sandstone fine to medium well well very pale brown 10YR 7/4 prominent cross bedding (5‐10 
degrees) Chem. 
TW4‐22 45.6‐46.0 May‐09 Qtz Sandstone medium grained well well very pale brown 10YR 7/4 faint cross bedding (5‐10 degrees) Chem. 
TW4‐22 50.0‐50.3 May‐09 Qtz Sandstone medium to coarse well well light yellowish brown 10YR 6/4 weathered mafic grains, faint cross 
bedding Chem. 
TW4‐22 55.5‐55.7 May‐09 Qtz Sandstone/
Conglomerate fine to coarse poorly well very pale brown 10YR 8/2 large weathered feldspar clasts Chem. 
TW4‐22 60.2‐60.3 May‐09 Qtz Sandstone/
Conglomerate fine to medium moderately well very pale brown 10YR 8/2 weathered feldspar clasts Chem. 
TW4‐22 69.8‐70.0 May‐09 Qtz Sandstone medium to coarse poorly moderately very pale brown 10YR 7/4 weathered feldspar clasts Chem. 
TW4‐22 84.0‐84.3 May‐09 Qtz Sandstone fine grained/silty poorly poorly‐
moderately pale yellow 5Y 8/3 includes silt and med‐coarse sand 
(5‐10%) Chem. 
Notes: (1) All depths are measured in feet below ground surface. (2) Samples were collected during two site visits by MWH staff: February 2007 and May 2009. (3)Qtz = 
quartz; Hyd. = Hydraulic Sample; Chem. = Geochemical Sample.  (4) Color classified according to the Munsell soil color system. 

C-14

TABLE C‐3. VADOSE ZONE THICKNESS BENEATH THE TAILINGS CELLS. 
Tailings 
Cell Location 
Bottom 
Elevation of Cell 
(ft above MSL) 
Nearest 
Well 
Water Table 
Elevation 
(ft above MSL) 
Vadose Zone 
Thickness 
(ft) 
Cell 2 Cell 2 NW corner 5602 MW‐24 5506 96 
Cell 2 Between Cells 2&3 5592 MW‐29 5511 81 
Cell 2 Cell 2 N side 5595 MW‐28 5541 54 
Cell 2 Cell 2 NE corner 5605 TW4‐20 5553 52 
Cell 2 Between Cells 2&3 5582 MW‐30 5535 47 
Cell 2 Between Cells 2&3 5588 MW‐31 5542 46 
Cell 2 Cell 2 N side 5600 TW4‐22 5571 29 
Cell 3 Cell 3 SW corner 5585 MW‐23 5495 90 
Cell 3 Cell 3 S side 5585 MW‐12 5500 85 
Cell 3 Cell 3 S side 5577 MW‐05 5502 75 
Cell 3 Between Cells 2&3 5585 MW‐29 5511 74 
Cell 3 Cell 3 S side 5582 MW‐11 5518 64 
Cell 3 Cell 3 SE corner 5592 MW‐25 5535 57 
Cell 3 Between Cells 2&3 5585 MW‐31 5542 43 
Cell 3 Between Cells 2&3 5577 MW‐30 5535 42 
Cell 4A Cell 4A S side 5562 MW‐14 5494 68 
Cell 4A Cell 4A SW corner 5557 MW‐15 5493 64 
Cell 4A Cell 4A N side 5570 MW‐11 5518 52 
Cell 4A Cell 4A NE corner 5575 MW‐25 5535 40 
Notes: 
1. Units for elevation are referenced to feet above mean sea level (ft above MSL). 
2. Bottom elevations for Cells 2 & 3 from D'Appolonia (1982). 
3. Bottom elevations for Cell 4A from Geosyntec (2007).    
4. Average water table elevations from 2007 Water year. 
5. The vadose zone thickness was calculated as the difference between the cell bottom and the water 
table elevation. 
6. The average vadose zone thickness for Cell 2 (excluding TW4‐22), Cell 3, and Cell 4A were 63 ft, 66 
ft, and 56 ft, respectively.  TW4‐22 excluded because this well is located upgradient of Cell 1. 

C-15
TABLE C‐4. STATISTICAL ANALYSIS OF THE GEOCHEMICAL DATA BASED ON DIFFERENT ATTRIBUTES. 
Mineralogical Property & 
Statistical Metric MW‐23 MW‐24 MW‐30 TW4‐22 Upper 
Sandstone Conglomerate Siltstone Lower 
Sandstone 
29‐54 
ft bgs 
54‐79 
ft bgs 
79‐104 
ft bgs 
Entire 
Population 
ANP (g CaCO3/kg rock) 
Count 10 9 7 8 18 4 3 9 16 12 6 34
Minimum 0.5 2 0.5 2 1 2 6 1 0.5 0.5 4 0.5 
Maximum 182 27 69 36 69 182 9 27 69 182 27 182
Arithmetic Mean 22.6 7 13.1 11.1 10.1 48.5 7.7 7.7 10.3 21.3 7.8 13.8
Standard Deviation 56.1 7.7 24.9 11.4 16.6 89.0 1.5 9.1 17.7 50.9 9.4 32.4 
Geometric Mean 5.7 5.2 3.7 7.5 5.0 9.7 7.6 4.3 4.8 6.4 5.5 5.4
Geometric Standard 
Deviation 4.3 2.1 5.4 2.5 3.2 7.5 1.2 3.2 3.4 4.2 2.2 3.4 
Median 4.0 4 5 7.5 4.5 5.0 8.0 4.0 4 6 4 4.0
Geo Mean Plus 1 Geo 
Standard Deviation 10.0 7.2 9.1 10.0 8.2 17.1 8.8 7.5 8.1 10.5 7.7 8.8 
Geo Mean Minus 1 Geo 
Standard Deviation 1.4 3.1 0 5.0 1.8 2.2 6.3 1.1 1.4 2.2 3.3 2.1 
HFO (mg/L Fe) 
Count 10 9 7 8 18 4 3 9 16 12 6 34
Minimum 3.85 3.85 25.3 16.3 4 4 42 9 3.85 3.85 8.72 3.85
Maximum 304 226 503 509 509 276 503 109 509 503 105 509
Arithmetic Mean 93.1 44.8 179 238 163 87.3 257 48.1 170 124 49.0 132 
Standard Deviation 106 71.2 183 192 168 128 232 42.5 175 153 41.1 155
Geometric Mean 43.1 18.9 102 149 70.3 30.9 169 30.8 71.3 51.8 32.4 55.4
Geometric Standard 
Deviation 4.2 3.8 3.3 3.3 4.9 6.1 3.5 2.8 5.1 4.6 2.9 4.4 
Median 46.9 10.4 72.9 195 69.8 34.6 226 19.1 69.8 47.6 45 59.8
Geo Mean Plus 1 Geo 
Standard Deviation 47.3 22.7 105.7 152.4 75.1 37.1 172.3 33.7 76.4 56.3 35.3 59.9 
Geo Mean Minus 1 Geo 
Standard Deviation 38.8 15.0 99.1 145.7 65.4 24.8 165.2 28.0 66.3 47.2 29.5 51.0 
Notes: (1) ANP = acid neutralization potential; HFO = hydrous ferric oxide; ft bgs = feet below ground surface; Fe = iron; CaCO3 = calcite.(2) Two samples analyzed for ANP were 
reported at the practical quantitation limit (PQL) of 0.5 g CaCO3 per kilogram of rock. These values were assumed for the statistical analysis.  (3) The arithmetic average for each of 
the 4 sample intervals analyzed in duplicate was used in the statistical analysis.  (4) The conversion between milligrams of iron per liter of water in the leachate to milligrams of HFO 
per kilogram of rock is described in the text.  

C-16

TABLE C‐5.  SUMMARY OF UNSATURATED AND SATURATED HYDRAULIC PROPERTIES OF BEDROCK 
VADOSE ZONE CORE SAMPLES. 
Well ID and 
Core 
Interval 
(ft bgs) 
Residual soil 
water 
content 
θr 
(% vol) 
Saturated soil 
 Water 
content 
θs 
(% vol) 
Curve fitting parameters in the 
soil water retention function a 
Saturated hydraulic 
conductivity in the 
vertical direction 
Ksat 
(cm/d) 
Dry Bulk 
Density 
Ρb 
(g/cm3) α 
(cm‐1) 
n 
(‐) 
MW‐30 
35.5‐36.0 0.004 0.199 0.0266 1.348 69.9 1.98 
MW‐23 
55.5‐56.0 0.003 0.184 0.0103 1.386 9.37 2.03 
MW‐23 
74.3‐74.6 0.016 0.122 0.0003 1.354 2.47 2.33 
MW‐23 
82.7‐82.9 0.003 0.160 0.0069 1.336 14.9 2.10 
MW‐23 
103.3‐103.5 0.006 0.205 0.0287 1.349 263 1.84 
MW‐30 
44.0‐44.5b 0.032 b 0.264 b 0.0081 b 1.201 b 0.707 2.23 
Notes: 
All depths are measured in feet below ground surface (ft bgs). 
aThe van Genuchten‐Mualem single‐porosity soil‐hydraulic‐property model was selected to characterize 
the soil‐hydraulic properties. 
bWater retention parameters based on volume adjusted values because core cracked and swelled after 
conductivity testing. 

C-17
Figure C-1. Monitoring well and test well locations near the tailings cells 2 and 3 with
core available. Locations include MW-23, MW-24, MW-28, MW-30, and TW4-22.

C-18
Figure C-2. Generalized cross section of monitoring wells and available core in the vicinity of the tailings cells. Lithologic information and sample intervals selected for geochemical and hydraulic characterization are also
identified.

C-19
Figure C-3. Acid neutralization potential (ANP) results from core samples.

C-20
Figure C-4. Amount of hydrous ferric oxide (HFO) from core samples (represented as dissolved iron concentrations in the leachate).

C-21
Figure C-5. Comparison between the measured and model-predicted soil water retention
curves for the vadose zone samples.

C-22
Figure C-6. Log hydraulic conductivity as a function of water content for the vadose
zone samples.

APPENDIX D
VEGETATION EVALUATION FOR THE
EVAPOTRANSPIRATION COVER

D-1
APPENDIX D
VEGETATION EVALUATION FOR THE EVAPOTRANSPIRATION COVER
This appendix provides an evaluation of vegetation that would be used as an integral part
of an evapotranspiration (ET) cover proposed for reclamation of tailing cells at the White
Mesa Mill Site. A critical component of an ET cover is the plant community that will be
established on the cover and will function over the long term to provide protection from
wind and water erosion and assist in removing water through the process of transpiration.
In this appendix, issues related to the short-term establishment and long-term
sustainability of vegetation proposed as part of the ET cover are addressed. These issues
include: plant species selection, ecological characteristics of species (i.e., longevity,
sustainability, compatibility, competition, rooting depth and root distribution),
characteristics of the established plant community (i.e., percent plant cover and leaf area
index [LAI]), and soil requirements for sustained plant growth.
Empirical data regarding the ecological characteristics of the species mix (rooting depth
and root distribution) and established plant community (percent cover) are summarized
from the literature and nearby lysimeter studies to develop a conceptual model of the
vegetation component for the ET cover system. The empirical data were used to
parameterize the numerical model and predict the ET cover’s performance over the long
term. A range of parameter values intended to correspond to a base case (anticipated or
expected) scenario and a worst case scenario is presented. These values were used in the
predictive simulations performed using HYDRUS. The range in data values are included
to determine which parameters may be more sensitive in predicting flow through the
cover. Results of model simulations are presented in Appendix G.
PROPOSED SPECIES FOR ET COVER RECLAMATION
The following 12 species (10 grasses and 2 forbs) are proposed for the ET cover system
at the White Mesa Mill Site. These species were selected for their adaptability to site
conditions, compatibility, and long-term sustainability. Species were also selected based

D-2
on the assumption that institutional controls will prohibit grazing by domestic livestock.
The proposed species are:
• Western wheatgrass, variety Arriba (Pascopyrum smithii)
• Bluebunch wheatgrass, variety Goldar (Pseudoroegneria spicata)
• Slender wheatgrass, variety San Luis (Elymus trachycaulus)
• Streambank wheatgrass, variety Sodar (Elymus lanceolatus ssp. psammophilus)
• Pubescent wheatgrass, variety Luna (Thinopyrum intermedium ssp. barbulatum)
• Indian ricegrass, variety Paloma (Achnatherum hymenoides)
• Sandberg bluegrass, variety Canbar (Poa secunda)
• Sheep fescue, variety Covar (Festuca ovina)
• Squirreltail, variety Toe Jam Creek (Elymus elymoides)
• Blue grama, variety Hachita (Bouteloua gracilis)
• Common yarrow, no variety (Achillea millefolium)
• White sage, variety Summit (Artemisia ludoviciana).
These species are described in more detail later in this appendix.
PROPOSED SEEDING RATES
Given a mixture of the species listed above, Table D-1 presents broadcast seeding rates
for each species. Seeding rates were developed based on the objective of establishing a

D-3
permanent cover of grasses and forbs in a mixture that would promote compatibility
among species and minimize competitive exclusion or loss of species over time. The
proposed seeding rate is based on pounds of pure live seed per acre (lbs PLS/acre).
ECOLOGICAL CHARACTERISTICS OF PROPOSED SPECIES AND
ESTABLISHED PLANT COMMUNITY
Longevity and Sustainability
All of the species proposed for reclamation of the tailings cells are long-lived, except for
slender wheatgrass (Elymus trachycaulus) and squirreltail (Elymus elymoides). Slender
wheatgrass is a perennial bunchgrass that is short-lived (5 to 10 years) but has the ability
to reseed and spread vegetatively with rhizomes. Squirreltail is also a short-lived
perennial but has the ability to establish quickly and is highly effective in competing with
undesirable annual grasses. Both of these species are included in the proposed seed
mixture because of their ability to provide quick cover for erosion protection and to
effectively compete with annual and biennial species that cannot be relied upon to
provide consistent and sustainable plant cover. The use of these species will facilitate the
establishment of the remaining long-lived perennials that have been documented to be
highly adapted to the elevation, climate, and soil conditions found at the White Mesa Mill
Site (Monsen et al., 2004; Alderson and Sharp, 1994; Wasser, 1982; Thornburg, 1982).
The perennial grasses and forbs in the proposed seed mixture include species that develop
individual plants that are long lived (30 years or more) and are able to reproduce either by
seed or vegetative plant parts like rhizomes and tillers. The use of these species in
reclamation of the tailing cells will ensure a permanent or sustainable plant cover because
of the highly adapted nature of these species to existing site conditions, their tolerance to
environmental stresses such as drought, fire, and herbivory, and their ability to effectively
reproduce over time.
The use of a mixture of species for the ET cover also contributes to longevity and
sustainability. The establishment of a diverse community has many advantages over a

D-4
monoculture for sustained plant growth. The use of a variety of species ensures that
diverse microsites that may exist over a seeded site are properly matched with species
that are adapted to those specific environmental conditions. In addition, a mixture of
species reverses the loss of plant diversity and enhances natural recovery processes
following impacts from insects, disease organisms, and adverse climatic events. Finally,
mixtures provide improved ground cover and surface stability, along with reducing weed
invasion by fully utilizing plant resources such as water, nutrients, sunlight and space.
Weeds in this context are typically annual or biennial plants considered to be undesirable
or troublesome, especially growing where they are not wanted.
Compatibility
Reclamation research and its application have been ongoing in the U.S. since the early
1900s. First with the reseeding of millions of acres following the dust bowl of the 1930s.
Then, improvements of large tracts of arid and semiarid rangelands between the 1960s
and 1980s following more than a half a century of rangeland exploitation through
overgrazing. In 1985 the U.S. Department of Agriculture Conservation Reserve Program
was implemented which resulted in the conversion of more than 40 million acres of
marginal farm land to permanent grasslands through an extensive seeding program.
Finally, there have been tens of thousands of acres of mined lands reclaimed across the
U.S. with the implementation of federal and state rules and regulations governing mine
land reclamation. Over this time period, there have been thousands of reclamation
publications in the form of books, scientific journal articles, symposium proceedings, and
government publications. Many publications have reported on the performance of
individual species and mixtures of species under semiarid conditions similar to
southeastern Utah (e.g., Plummer et al., 1968; Monsen et al., 2004). All of this work has
led to a knowledge base about species compatibility. Species that are seeded together in
mixtures must be compatible as young, developing plants or certain individuals will
succeed and others will fail. The species proposed for the ET cover at the White Mesa
Mill Site are all compatible with each other and seeding rates will be used to prevent
overseeding species that may be aggressive [e.g., pubescent wheatgrass (Thinopyrum

D-5
intermedium)] and could potentially dominate the site (Monsen et al., 2004). These
species are commonly seeded together and many studies have shown excellent
interspecies compatibility (e.g., DePuit et al., 1978; DePuit, 1982; Redente et al., 1984;
Sydnor and Redente, 2000; Newman and Redente, 2001). Finally, to increase
compatibility and to reduce competition among seeded species, sites would be broadcast
seeded as opposed to drill seeded. According to Monsen et al. (2004), drill seeding
causes species in a mixture to be placed in potentially competitive situations, while
broadcasted seeds are not placed in as close contact with each other as with drilling and
therefore are less likely to be negatively impacted from competition.
Competition
There are two ways to view competition. In the context of establishing an ET cover on
the tailing cells, the use of seeded species to compete with weeds or woody plants is a
desirable attribute. However, competition among seeded species with the potential loss
of any of these species is undesirable. Therefore, as stated earlier, the proposed seed
mixtures is comprised of species that can coexist and also fully utilize plant resources to
keep weeds or woody species from colonizing and excluding seeded species. The
establishment of weeds, especially invasives (i.e., non-native species whose introduction
causes economic and environmental harm) is unacceptable because of the potential loss
of seeded perennial species and the subsequent reduction in species diversity, plant cover,
and overall sustainability. The establishment of deep rooted woody plants is
unacceptable because of the potential for biointrusion through the cover and into the
tailings material. Once established, the proposed seed mixture will produce a grass-forb
community of highly adapted and productive species that will effectively compete with
undesirable species, including shrubs native to the area. Paschke et al. (2003) present a
literature review on shrub establishment on mined lands and conclude that one of the
primary reasons that shrub establishment does not occur in mined land reclamation is
because of competition from herbaceous species. This finding is also supported by
DePuit et al. (1980), DePuit (1988), Munshower (1994), and Monsen et al. (2004).
Because of the highly adapted and competitive nature of the species that will be seeded,

D-6
the invasion of indigenous woody species will be inhibited, and intrusion into the cover
below the water storage layer (top 4 feet of the cover) from their roots is not anticipated
to occur. Woody species in this environment are slow-growing and not nearly as
competitive for water and nutrients as the proposed grass and forb species (Monsen et al.,
2004). In addition, species like sagebrush, piñon pine, and Utah juniper have become
dominant components of the regional flora primarily because of decades of overgrazing
that has removed more palatable grasses and forbs and allowed less palatable woody
species to establish and expand their range (Dames and Moore, 1978; Ellison, 1960).
This process is referred to as retrogression (Holechek et al., 1998). These conditions will
not occur on the tailing cells cover and therefore will not be a factor favoring the
establishment of woody species.
Percent Plant Cover and Leaf Area Index
Monitoring of an alternative cover at the Monticello, Utah, Uranium Mill Tailings
Disposal Site showed that the plant cover performed well over a seven year period. Plant
cover ranged from 5.5% during the first growing season to nearly 46% in the seventh
growing season (Waugh et al., 2008). A total of 18 species were seeded at the Monticello
Site and of these 18 species, eight species contributed 70% of the total plant cover.
Approximately one half of the species proposed for the White Mesa Site were seeded at
Monticello and of the eight best-performing species, four of these species are in the
White Mesa mixture. High performing species used at Monticello that are not proposed
for White Mesa include three introduced species that can be highly competitive (i.e.
smooth brome, crested wheatgrass, and alfalfa) and were not considered acceptable for
the White Mesa Site. Based on these results and the similarity in environmental
conditions between Monticello and White Mesa, a plant cover estimate of 40% was
determined to be a reasonable estimate for a long-term average, while a percent plant
cover of 30% was assigned as a worst case scenario under drought conditions. The
percent vegetative cover at White Mesa is expected to be slightly less than what would be
found at Monticello because the average annual precipitation at White Mesa is
approximately 13 inches compared to 15 inches at Monticello and the average annual

D-7
maximum/minimum air temperatures are 64/37oF for White Mesa and 59/33oF for
Monticello. The slightly greater precipitation and lower temperatures at Monticello are
due to its slightly higher elevation of 7,000 feet compared to 5,600 feet at White Mesa.
Long-term average plant cover for the tailing cells along with monthly leaf area index
(LAI) values were estimated for the proposed ET cover at the White Mesa Site. Three
primary publications were used to estimate monthly LAI for the ET cover, including:
Groeneveld (1997), Scurlock et al. (2001), and Fang et al. (2008). Table D-2 presents a
compilation of LAI values based on North American data sets that were focused on
semiarid herbaceous plant communities. Months with a LAI of zero were assigned a
transpiration rate of zero, and only evaporation was simulated in the HYDRUS-1D
model. It is important to note that the proposed species for the ET cover include both
cool- and warm-season species. This combination of species will maximize the length of
the growing season and transpiration from early spring to late fall. Cool-season species
are more productive and use more water during the cooler times of the growing season,
while warm-season species are more productive and use more water during the warmest
period of the year.
The formation of desert pavement and potential impact on plant cover has been raised as
an issue for discussion. Desert pavements are armored surfaces composed of angular or
rounded rock fragments, usually one or two stones thick, set on or in a matrix of finer
material (Cooke and Warren, 1973). These surfaces form on arid soils through deflation
of fine material by wind or water erosion due to a lack of protection by surface vegetation
(Cooke and Warren, 1973). Desert pavements are not common in semiarid regions and
do not occur where either wind or water erosion are controlled by plant cover (Hendricks,
1991), as would be the case for the White Mesa cover system. In addition, there is no
evidence of desert pavement formation either on the White Mesa Site or areas
surrounding the site. Even with the use of a topsoil layer amended with gravel, there is
no supporting evidence to indicate a potential for desert pavement formation or an
associated decrease in plant cover over the long term.

D-8
Rooting Depth and Distribution
The effective rooting depth would be 3.5 feet (107 cm). Six primary publications were
used to estimate root densities by depth for the plant community that would establish on
the ET cover, including: Hopkins (1953), Bartos and Sims (1974), Sims and Singh
(1978), Lee and Lauenroth (1994), Jackson et al. (1996), and Gill et al. (1999). Table D-
3 and Figure D-1 present an estimate of effective root densities by depth for the ET cover
system proposed for the White Mesa tailing cell cover. Root densities are presented for
anticipated and worst case scenarios. The root densities were interpolated at depth in
order to parameterize the HYDRUS-1D model (e.g., 0 to 15 cm interpolated from 0 to 1.9
grams/cm3). Rooting depths for species included in the proposed seed mixture for the
White Mesa site are presented in Table D-4.
BIOINTRUSION
Based on a review of the wildlife survey data from the 1978 Environmental Report
produced for the White Mesa site (Dames and Moore, 1978), and a thorough literature
review of burrowing depths and biointrusion studies, the maximum depth of on-site
burrowing would be approximately one meter or slightly over three feet. Wildlife survey
data for the site indicate that burrowing mammals include deer mice, kangaroo rats,
chipmunks, desert cottontails, blacktailed jackrabbits, and prairie dogs. Other burrowing
mammals, such as pocket gophers and badgers have not been observed in the area of the
White Mesa site (Dames and Moore, 1978). Of the list of burrowing mammals that may
occur on the site, the prairie dog is the species capable of burrowing to the greatest depth.
Studies by Shuman and Whicker (1986) and Cline et al. (1982) conducted in southeast
Wyoming, Grand Junction, Colorado and Hanford, Washington, document maximum
burrowing depths of prairie dogs between 2.0 and 3.2 feet. Based on this empirical data
and the potential species that may use the site as habitat, any burrowing activity that may
occur would be limited to about three feet below ground surface. Burrowing animals
would not significantly perturb the surface of the cover, and thus would not encourage
growth of undesirable species. The effects of biointrusion on water flow and radon
transport through the ET cover was evaluated as part of the sensitivity analysis for the

D-9
infiltration model (see Appendix G), as well as the radon attenuation model (see
Appendix H).
SOIL REQUIREMENTS FOR SUSTAINABLE PLANT GROWTH
There are two key components to establishing an ET cover with a sustainable plant
community. The first is to select long-lived species that are adapted to the environmental
conditions of the site. The second is to provide a cover soil that will function as an
effective plant growth medium over the long term by supplying plants with adequate
amounts of water, nutrients and rooting volume.
There are a number of soil characteristics that are particularly important to achieve long-
term sustainability in semiarid environments and include the following: pH, electrical
conductivity (EC), sodium levels, percent organic matter, texture, bulk density, cation
exchange capacity, macronutrient concentrations, available water holding capacity, and
soil microorganisms. Table D-5 presents levels for most of these soil properties that are
considered necessary for long-term sustained plant growth. In addition, the table includes
soil property levels from soil samples of potential cover soil collected from stock piles at
the White Mesa Site in May 2009.
The soil properties of the potential cover soil that are acceptable for sustaining long-term
plant growth include: pH, EC, sodium adsorption ratio (SAR), percent clay content, and
extractable phosphorus. Those soil properties that appear to be deficient and would need
improvement include: percent organic matter, total nitrogen, and extractable potassium.
Cation exchange capacity was not measured in the potential cover soil, but it is believed
that the cover soil will have an acceptable level for sustained plant growth based on the
percent clay content and a recommendation that an organic matter amendment be added
to the soil during the reclamation process. Bulk density of the emplaced cover material
will be specified in the cover design and will be controlled during the construction
process to be within the sustainability range shown in Table D-5.

D-10
In order for the potential cover soil to function as a normal soil and provide long-term
sustainable support for the vegetation component of the ET cover, it will be amended to
improve organic matter content, nitrogen and potassium levels. An organic matter
amendment will also improve available water holding capacity and cation exchange
capacity. The source of organic matter will depend upon availability in the region and
could either be composted biosolids or a combination of manure and hay to provide a
material that has the appropriate carbon to nitrogen ratio for sustained plant growth.
Such an organic matter amendment will also provide a source of soil microorganisms that
will function to cycle nutrients over time and ensure sustainable plant growth.
UPDATED RECLAMATION PLAN AND ENGINEERING SPECIFICATIONS
Upon Executive Secretary approval of the Infiltration and Contaminant Transport
Modeling (ICTM) Report, and proposed conceptual cover design included as part of this
report, the Reclamation Plan would be modified to accommodate necessary changes to
protect public health and the environment. The Reclamation Plan would include most of
the details presented within this appendix, as well as vegetative design criteria that must
be met (e.g., percent cover, exclusion of woody and invasive species). The Engineering
Cover Design will include material specifications and a comparison to the data used to
support the modeling presented in the ICTM Report.
ECOLOGICAL CHARACTERISTICS OF PROPOSED SPECIES
Important ecological characteristics for each species proposed for reclamation are
provided in the paragraphs that follow. Species information was obtained from Monsen
et al. (2004), Alderson and Sharp (1994), Wasser (1982), and Thornburg (1982). The
proposed species are adapted to the elevation (5,600 feet), precipitation (13 inches per
year on average), and soil textural ranges (loam to sandy clay) that are well within the
environmental conditions of the White Mesa Site. Table D-4 presents a summary of the
ecological characteristics discussed in the following paragraphs.

D-11
Western wheatgrass, variety Arriba (Pascopyrum smithii)
Western wheatgrass is a native, rhizomatous, long-lived perennial cool season grass.
It grows well in a 10 to 14 inch mean annual precipitation zone and is adapted to a
wide range of soil textural classes at elevation ranges up to 9,000 feet. Western
wheatgrass has been an important species for restoring mining related disturbances,
for erosion control and for critical area stabilization in semiarid regions because of its
ease of establishment and ability to grow successfully in pure or mixed stands of both
warm and cool season species. Western wheatgrass is fire tolerant and regenerates
readily following burning. The variety of Arriba is known for rapidly establishing
seedlings and high seed production. The combination of its ability to spread
vegetatively and reproduce by seed ensures long-term sustainability of this species.
Bluebunch wheatgrass, variety Goldar (Pseudoroegneria spicata)
Bluebunch wheatgrass is a native, cool season perennial bunch grass. Bluebunch
wheatgrass grows on soils that vary in texture, depth and parent material. It is one of
the most important and productive grasses found in sagebrush communities in the
intermountain west. Bluebunch wheatgrass is fire tolerant and regenerates
vegetatively following burning. This species is well adapted to a 12 to 14 inch mean
annual precipitation range and is considered to be highly drought resistant.
Bluebunch wheatgrass performs well in mixtures with other species and grows at
elevations up to 10,000 feet.
Slender wheatgrass, variety San Luis (Elymus trachycaulus)
Slender wheatgrass is a native, cool season, perennial bunch grass that occasional
produces rhizomes. It is a short-lived species (5 to 10 years) but it reseeds and
spreads well by natural seeding, exceeding most other wheatgrasses in this
characteristic. Slender wheatgrass can serve as an important pioneer species; its
seedlings are vigorous and capable of establishing on harsh sites. In addition, it is
able to establish and compete with weedy species. Slender wheatgrass is commonly

D-12
seeded in mixtures with other grasses and forbs to restore disturbances and
rehabilitate native communities. It is adapted to a wide variety of sites and is
moderately drought tolerant. It performs best at sites with an annual precipitation of
15 inches or more, but can grow on sites with precipitation levels as low as 13 inches.
Streambank wheatgrass, variety Sodar (Elymus lanceolatus ssp. psammophilus)
Streambank wheatgrass is considered to be part of the thickspike wheatgrass (Elymus
lanceolatus ssp. lanceolatus) taxa. Variety Sodar is a native, perennial sod grass that
is highly rhizomatous and adapted to the western intermountain area. It is highly
drought tolerant and performs well in mean annual precipitation ranges between 11
and 18 inches. It grows on a wide range of soil textures, from sandy to clayey.
Streambank wheatgrass is commonly used in mine land reclamation and is best
known for its ability to control erosion and compete with annual weeds. Its highly
rhizomatous nature ensures long-term sustainability of this species.
Pubescent wheatgrass, variety Luna (Thinopyrum intermedium ssp. barbulatum)
Pubescent wheatgrass is a long-lived sod forming perennial introduced from Eurasia.
It is highly drought tolerant and grows where the mean annual precipitation is 12
inches or more. It is adapted to a wide range of soil textures, from sand to clay.
Pubescent wheatgrass is a highly persistent species, should be seeded at low densities
to avoid competition with native species and has been found to be effective in
reducing the establishment of woody plants.
Indian ricegrass, variety Paloma (Achnatherum hymenoides)
Indian ricegrass is a native, cool season, perennial bunchgrass with a highly fibrous
root system. Indian ricegrass is one of the most common grasses on semiarid lands in
the west and is one of the most drought tolerant species used in mine land
reclamation. It generally occurs on sandy soils, but is found on soils ranging from
sandy to heavy clays. It grows from 2,000 to 10,000 feet in areas where the mean

D-13
annual precipitation is 6 to 16 inches. Indian ricegrass is slow to establish, but highly
persistent once it becomes established.
Sandberg bluegrass, variety Canbar (Poa secunda)
Sandberg bluegrass is a native, cool season perennial bunchgrass that is adapted to all
soil textures and is highly resistant to fire damage. Sandberg bluegrass is one of the
more common early-season bunchgrasses in the Intermountain area. It grows at
elevations from 1,000 to 12,000 feet and can be successfully established in areas with
a mean annual precipitation of 12 inches or more. Established plants are not overly
competitive, and therefore highly compatible with other native species.
Sheep fescue, variety Covar (Festuca ovina)
Sheep fescue is a short, mat-forming native perennial that grows well on infertile soils
in areas with a mean annual precipitation of 10 to 14 inches. It is long-lived and
highly drought tolerant. Sheep fescue is a cool season species that greens up early in
the spring. The proposed variety, Covar, was introduced from Turkey and is
commonly used in mine land reclamation for long-term stabilization and erosion
control. This variety was selected because plants are persistent, winter hardy, and
drought tolerant.
Squirreltail, variety Toe Jam Creek (Elymus elymoides)
Squirreltail is a short-lived perennial that is selected for its ability to establish quickly
and to effectively compete with undesirable annual grasses. It grows along an
elevation range from 2,000 to 11,000 feet and on all soil textures in mean annual
precipitations zones of 8 to 15 inches. Squirreltail is fairly tolerant of fire because of
its small size.

D-14
Blue grama, variety Hachita (Bouteloua gracilis)
Blue grama is a low-growing perennial warm season bunchgrass. Blue grama
produces an efficient, widely spreading root system that is mostly concentrated near
the soil surface. Blue grama is adapted to a variety of soil types, but does best on
well-drained soils and once established, is highly drought tolerant. This species is
commonly found with cool-season species and is highly compatible with other native
perennials.
Common yarrow (Achillea millefolium)
Yarrow is a common native forb species that is rhizomatous and found growing from
valley bottoms to timberline. It is commonly used in mine land reclamation,
establishes easily from seed and is highly persistent. It grows on a variety of soil
textures and found in a mean annual precipitation range between 13 and 18 inches.
White sage, variety Summit (Artemisia ludoviciana)
White sage is considered to be a pioneer rhizomatous forb species that establishes
quickly on disturbed sites and is highly compatible with perennial grasses. It does
best on well-drained soils, but can be found growing on a wide range of soil textures.
It is adapted to sites above 5,000 feet in elevation and to sites with a mean annual
precipitation above 12 inches.

D-15
REFERENCES
Alderson, James and W. Curtis Sharp. 1994. Grass Varieties in the United States. U.S.
Department of Agriculture, Agriculture Handbook No. 170. Washington, D.C.

Bartos, Dale and Phillip Sims. 1974. Root dynamics of a shortgrass ecosystem. J. of
Range Management 27:33-36.

Brady, N. C. 1974. The Nature and Property of Soils. 8th ed. MacMillian Press. New
York, NY.

Cline, J. F., F. G. Burton, d. A. Cataldo, W. E. Shiens, and K. A. Gano. 1982. Long-
term biobarriers to plant and animal intrusions of uranium mill tailings. Rep. PNL-
4340. Pacific Northwest Lab. Richland, WA.

Cooke, R. V. and A. Warren. 1973. Geomorphology in Deserts. University of
California Press. Berkeley, CA.

Coupland, R. T. and R. E. Johnson. 1965. Rooting characteristics of native grassland
species in Saskatchewan. J. of Ecology 53:475-507.

Dames and Moore. 1978. Environmental Report—White Mesa Uranium Project, San
Juan County, Utah. Prepared for Energy Fuels Nuclear, Inc.

DePuit, E. J. 1982. Cool-season perennial grass establishment on Northern Great Plains
mined lands: status of current technology. Pages B1-B24 In Proceedings:
Symposium on surface Coal Mining and Reclamation in the Northern Great Plains.
Montana Agricultural Experiment Station Research Report 194. Bozeman, MT.

DePuit, Edward J. 1988. Productivity of reclaimed lands—rangeland. Pages 93-129 In
Hossner, Lloyd (ed.) Reclamation of Surface-Mined Lands Vol II. CRC Press. Boca
Raton, FL.

DePuit, E. J., J. G. Coenenberg, and C. L. Skilbred. 1980. Establishment of diverse
native plant communities on coal surface mined lands in Montana as influenced by
seeding method, mixture and rate. Montana Agricultural Experiment Station
Research Report 163. Bozeman, MT.

DePuit, E. J., J. G. Coenenberg, and W. H. Willmuth. 1978 Research on Revegetation of
Surface Mined Lands at Coalstrip Montana: Progress Report 1975—1977 Res. Rep.
127. Montana Agricultural Experiments Station, Bozeman, MT.

Ellison, L. 1960. Influence of grazing on plant succession of rangeland. Botanical
Review 26:1-78.

D-16
Fang, Hongliang, Shunlin Liang, John R. Townshend and Robert Dickenson. 2008.
Spatially and temporally continuous LAI data sets based on integrated filtering
method: Examples from North America. Remote Sensing of Environment 112:75-
93.

Foxx, T. S. and G. D. Tierney. 1987. Rooting patterns in the pinyon-juniper woodland. pp.
69-79 In Everett, R. L. (ed.). Proceedings—Pinyon-Juniper Conference. USDA Forest
Service. Intermountain Forest and Range Experiment Station. General Technical Report
INT-215.

Gill, Richard, Ingrid Burke, Daniel Milchunas and William Lauenroth. 1999.
Relationship between root biomass and soil organic matter pools in the shortgrass
steppe of Eastern Colorado. Ecosystems 2:226-236.

Groeneveld, David. 1997. Vertical point quadrat sampling and an extinction factor to
calculate leaf area index. J. of Arid Environments. 36:475-485.

Harding, R. B. 1954. Surface accumulation of nitrates and other soluble salts in California
orange orchards. Soil Science Society of America Proceedings. 18:369-372.

Hendricks, David M. 1991. Genesis and classification of arid region soils. Pages 33-79 In
Skujins, J. (ed.) Semiarid Lands and Deserts. Marcel Dekker, Inc. New York, NY.

Holechek, Jerry L., Rex D. Pieper, and Carlton H. Herbel. 1998. Range Management
Principles and Practices. Prentice Hall, Upper Saddle River, NJ

Hopkins, Harold 1953. Root development of grasses on revegetated land. J. of Range
Management 6:382-92.

Jackson, R., J. Canadell, J. Ehleringer, H. Mooney, O. Salsa and E. Schulze. 1996. A
global analysis of root distributions for terrestrial biomes. Oecologia 108:389-411.

Lee, C. A. and W. K. Lauenroth. 1994. Spatial distributions of grass and shrub root systems
in the shortgrass steppe. The American Midland Naturalist 132:117-123.

Ludwick, A. E. and J. R. Rogers. 1976. Soil test explanation. 502 Service in Action.
Colorado State University Agricultural Extension Service. Fort Collins, CO.

Monsen, Stephen. B., Richard Stevens and Nancy L. Shaw. 2004. Restoring Western
Ranges and Wildlands. U.S. Department of Agriculture. Forest Service. General
Technical Report RMRS-GTR-136-vol 1-3. Rocky Mountain Research Station. Fort
Collins, CO.

Munshower, Frank. 1994. Practical Handbook of Disturbed Land Revegetation. CRC
Press. Boca Raton, FL.

D-17
Newman, G. J. And E. F. Redente. 2001. Long-term plant community development as
influenced by revegetation techniques. J. Range Manage. 54:717-724.

Paschke, M. W., E. F. Redente, and S. L. Brown. 2003. Biology and establishment of
mountain shrubs on mining disturbances in the Rocky Mountains, USA. Land
Degradation & Development 14:459-480.

Plummer, A. Perry, Donald R. Christensen, and Stephen B. Monsen. 1968. Restoring
Big-Game Range in Utah. Utah Division of Fish and Game. Publication No. 68-3.
Utah Division of Fish and Game, Ephraim, UT.

Redente, E. F., T. B. Doerr, C. E. Grygiel, and M. E. Biondini. 1984. Vegetation
establishment and succession on disturbed soils in northwest Colorado. Reclamation
and Revegetation Research 3:153-166.

Scurlock, J. M. O., G. P. Asner, and S. T. Gower. 2001. Worldwide Historical Estimates
of Leaf Area Index, 1932-2000. Oakridge National Laboratory. ORNL/TM-
2001/268.

Shuman, R. and F. W. Whicker. 1986. Intrusion of reclaimed uranium mill tailings by
prairie dogs and ground squirrels. J. Environmental Quality 15:21-24.

Sims, Phillip and J. S. Singh. 1978. The structure and function of ten western North
American grasslands. J. of Ecology 66:573-597.

Spence, L. E. 1937. Root studies of important range plants of the Boise River
watershed. J. of Forestry 35:747-754.

Sydnor, Russell S. and Edward F. Redente. 2000. Long-term plant community
development on topsoil treatments overlying a phytotoxic growth medium. J.
Environmental Quality 29:1778-1786.

Thornburg, Ashley A. 1982. Plant Materials for Use on Surface-Mined Lands in Arid and
Semiarid Regions. USDA. Soil Conservation Service. SCS-TP-157. EPA-600/7-79-
134. U.S. Government Printing Office. Washington, D.C.

USDA. 2009. http://plants.USDA.gov. Accessed on 18 August 2008.

Wasser, Clinton H. 1982. Ecology and Culture of Selected Species Useful in Revegetating
Disturbed Lands in the West. U.S. Department of Interior. Fish and Wildlife Service.
FWS/OBS-82/56. U.S. Government Printing Office. Washington, D.C.

Waugh, W. J., M. K. Kastens, L. R. L. Sheader, C. H. Benson, W. H. Albright, and P. S.
Mushovic. 2008. Monitoring the performance of an alternative landfill cover at the
Monticello, Utah, Uranium Mill Tailings Disposal Site. Proceedings of the Waste
Management 2008 Symposium. Phoenix, AZ.

D-18
Weaver, J. E. and F. E. Clements. 1938. Plant Ecology. 2nd Edition. McGraw-Hill.
New York, NY.

Wyatt, J. W., D. J. Dollhopf, and W. M. Schafer. 1980. Root distribution in 1 to 48 year
old stripmine spoils in southeastern Montana. J. Range Management 33:101-104.

D-19
Table D-1. Species and seeding rates proposed for ET cover at the White Mesa Mill
Site.

Scientific Name Common Name Variety Native/
Introduced
Seeding
Rate (lbs
PLS/acre)†
Grasses
Pascopyrum smithii Western wheatgrass Arriba Native 3.0
Pseudoroegneria spicata Bluebunch wheatgrass Goldar Native 3.0
Elymus trachycaulus Slender wheatgrass San Luis Native 2.0
Elymus lanceolatus Streambank wheatgrass Sodar Native 2.0
Elymus elymoides Squirreltail Toe Jam Native 2.0
Thinopyrum intermedium Pubescent wheatgrass Luna Introduced‡ 1.0
Achnatherum hymenoides Indian ricegrass Paloma Native 4.0
Poa secunda Sandberg bluegrass Canbar Native 1.0
Festuca ovina Sheep fescue Covar Native 2.0
Bouteloua gracilis Blue grama Hachita Native 1.0

Forbs
Achillea millefolium Common yarrow No
variety
Native 6.0
Artemisia ludoviciana White sage Summit Native 3.0

Total 30.0
†Seeding rate is for broadcast seed and presented as pounds of pure live seed per acre (lbs PLS/acre). ‡Introduced refers to species that have been ‘introduced’ from another geographic region, typically outside
of North America. Also referred to as ‘exotic’ species.

D-20
Table D-2. Leaf area index for the ET Cover at White Mesa Mill Site.

Month
Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec
0 0 0.3 0.7 0.6 0.6 1.8 2.4 2.6 0.8 0.1 0

Table D-3. Root densities (anticipated case and worst case) for the White Mesa Mill
Site.

Depth (cm) Root Density (grams cm-3)
Anticipated Case
Root Density (grams cm-3)
Worst Case
0-15 1.9 1.3
15-30 6.2 4.3
30-45 1.7 0.8
45-60 0.8 0.5
60-75 0.6 0.3†
75-90 0.6 0.0
90-107 0.4 0.0
†Maximum rooting depth under worst case scenario would be 68 cm.

D-21
Table D-4. Summary of ecological characteristics of plant species proposed for the ET cover at the White Mesa Mill Site.

Sp
e
c
i
e
s

Or
i
g
i
n

An
n
u
a
l

o
r

Pe
r
e
n
n
i
a
l

Me
t
h
o
d

o
f

S
p
r
e
a
d

Ea
s
e

o
f

Es
t
a
b
l
i
s
h
m
e
n
t
a
Co
m
p
a
t
i
b
i
l
i
t
y

w
i
t
h

Ot
h
e
r

S
p
e
c
i
e
s
a
Lo
n
g
e
v
i
t
y
a
An
n
u
a
l

Pr
e
c
i
p
i
t
a
t
i
o
n

R
a
n
g
e

(i
n
c
h
e
s
)

El
e
v
a
t
i
o
n

R
a
n
g
e

(f
e
e
t
)

So
i
l

T
e
x
t
u
r
e
b
R
o
o
t
i
n
g

D
e
p
t
h

(c
m
)

So
i
l

S
t
a
b
i
l
i
z
a
t
i
o
n
a
Dr
o
u
g
h
t

T
o
l
e
r
a
n
c
e
a
Fi
r
e

T
o
l
e
r
a
n
c
e
a
Western
wheatgrass Native Perennial Vegetative 4 3 4 10-14 ≤9,000 S,C,L 109d 4 4 4
Bluebunch
wheatgrass Native Perennial Seed 4 4 4 12-14 ≤10,000 S,C,L 122e 4 4 4
Slender
wheatgrass Native Perennial Seed 4 4 2 13-18 ≤10,000 S,C,L 109d 2 2 2
Streambank
wheatgrass Native Perennial Vegetative 4 4 4 11-18 ≤10,000 S,C,L 165f 4 4 3
Pubescent
wheatgrass Introduced Perennial Vegetative 4 2 4 12-18 ≤10,000 S,C,L 185d 4 4 3
Indian
ricegrass Native Perennial Seed 3 4 4 6-16 ≤10,000 S,L 84g 2 4 2
Sandberg
bluegrass Native Perennial Seed 4 4 4 12-18 ≤12,000 S,C,L 45h 2 3 4
Sheep fescue
Native Perennial Seed 4 2 4 10-14 ≤11,000 S,C, L 56e 3 4 2
Squirreltail
Native Perennial Seed 3 4 3 8-15 ≤11,000 S,C,L 30c,i 2 4 3
Blue grama
Native Perennial Vegetative 2 4 4 10-16 ≤10,000 S,L 119g 4 4 4
Common
yarrow Native Perennial Vegetative 4 3 4 13-18 ≤11,000 S,C,L 105h 4 3 2
White sage
Native Perennial Vegetative 4 4 4 12-18 ≥5,000 S,C,L 20c,i 3 3 2
aKey to Ratings—4 = Excellent, 3 = Good, 2 = Fair, 1 = Poor bSoil Texture Codes—S = Sand, C = Clay, L = Loam cDepth represents minimum depth; no information in the literature on average or maximum depth could be found dWyatt et al., 1980; eWeaver and Clements, 1938; fCoupland and Johnson, 1965;gFoxx and Tierney, 1987; hSpence, 1937;iUSDA, 2009

D-22
Table D-5. Soil properties and their range of values important for sustainable plant
growth, along with analytical results of soil available for ET cover construction at the
White Mesa Mill Site.

Soil Property Level for
Sustainability
Reference Levels for
On-Site Soil
pH (units) 6.6 to 8.4 Munshower (1994) 7.7 to 8.1
EC (mmhos/cm) ≤4.0 Munshower (1994) <1.5
Sodium adsorption ratio ≤12 Munshower (1994) <0.5
Organic matter (%) 1.5 to 3.0 Brady (1974) 0 to 0.4
Texture (%) 35 to 50% clay Brady (1974) 36 to 50% clay
Bulk density (g/cm3) 1.2 to 1.8 Brady (1974) 1.59 to 1.99†
Water holding capacity
(cm H2O/cm soil)
0.08 to 0.16 Brady (1974) 0.084-0.14†
Cation exchange
capacity (meq/100g)
5 to 30 Munshower (1994) Not measured
Total nitrogen (%) 0.05 to 0.5 Harding (1954) 0.02 to 0.05
Extractable phosphorus
(mg/kg)
6 to 11 Ludwick and Rogers
(1976)
10 to 57
Extractable potassium
(mg/kg)
60 to 120 Ludwick and Rogers
(1976)
11 to 36
†Calculated values

D-23
Figure D-1. Root density profile for a semarid grassland community for the White Mesa ET
cover.

APPENDIX E
COMPARISON OF COVER DESIGNS BASED ON
INFILTRATION MODELING

E-1
APPENDIX E
COMPARISON OF COVER DESIGNS BASED ON INFILTRATION
MODELING
To compare the potential performance of different conceptual cover designs, infiltration
modeling with HYDRUS was performed to predict infiltration rates through the tailing
cell cover for each design. Four cover designs were evaluated and included one rock
cover and three evapotranspiration (ET) covers of variable thicknesses and layering. The
material thicknesses for the different cover designs were based on the results of radon
attenuation modeling to achieve the State of Utah’s long-term radon emanation standard
for uranium mill tailings (Utah Administrative Code, Rule 313-24). Results of the radon
attenuation model are presented in Appendix H. Rates of model-predicted water flux
entering the tailings cells were compared between the various simulations and used as the
basis to guide the selection of a cover design for the White Mesa tailings cells.
As specified in the Groundwater Discharge Permit (Part I.H.2.f), the Permittee may
include supplemental information to justify modification of certain Permit requirements,
including tailings cell cover system engineering design and construction specifications.
Upon Executive Secretary approval of the ICTM report and proposed conceptual cover
design included as part of this report, the Reclamation Plan would be modified to
accommodate necessary changes to protect public health and the environment.
CONCEPTUAL COVER DESIGNS
Four different cover designs were simulated with the HYDRUS infiltration model to
evaluate the range in model-predicted water flux rates expected to flow into the tailings.
The four conceptual cover designs simulated were:
1. Monolithic ET cover design
2. ET cover design with a compacted clay layer
3. ET cover design with a gravel capillary break layer
4. Conventional rock cover design with a compacted clay layer.

E-2
Cover 1 is a 2.84-m (9.3-ft) thick monolithic ET cover that consists from top to bottom
of:
• 15 cm (0.5 ft) of a gravel-amended topsoil admixture to promote revegetation
and provide for protection against erosion and frost damage
• 107 cm (3.5 ft) of random fill soil placed at 85% of Standard Proctor dry
density to serve as a water storage, biointrusion, and radon attenuation layer
• 162 cm (5.3 ft) of random fill soil comprised of 2.8 feet random fill
compacted to 95% of Standard Proctor dry density over 2.5 feet of random fill
placed at 80% of Standard Proctor dry density, to serve as grading (platform
fill) and radon attenuation layers.
The first cover is based on a monolithic ET cover design with a water storage layer
thickness that would provide sufficient soil volume to allow the establishment of
vegetation and protection against intrusion by burrowing animals (see Appendix D).
Based on empirical data published in the literature, and the potential species that may use
the site as habitat, any burrowing activity that may occur would be limited to the upper
one meter of the cover, with the remainder of the cover (1.84 m) not impacted.
Previously, TITAN Environmental (1996) completed a freeze/thaw evaluation based on
site-specific conditions which indicated that the anticipated maximum depth of frost
penetration was 6.8 inches (0.6 ft). Therefore, the entire soil-gravel admixture layer and
upper few centimeters of the underlying water storage layer will provide adequate
protection against frost penetration. The lower platform fill is based on the assumption
that a minimum of 3 feet of random fill has already been placed above Cell 2 and that the
lower 2.5 feet was placed at 80% while the upper 0.5 feet of that material will be
compacted along with 2.3 feet of additional random fill to 95%.
Cover 2 is a 2.72-m (8.9-ft) thick ET cover design with a compacted clay layer that
consists from top to bottom of:

E-3
• 15 cm (0.5 ft) of a gravel-amended topsoil admixture to promote revegetation
and provide for protection against erosion and frost damage
• 107 cm (3.5 ft) of random fill soil placed at 85% of Standard Proctor dry
density to serve as a water storage, biointrusion, and radon attenuation layer
• 31 cm (1 ft) of compacted clay compacted to 90% of Modified Proctor dry
density to serve as a radon attenuation layer
• 119 cm (3.9 ft) of random fill soil comprised of 1.4 feet random fill
compacted to 95% of Standard Proctor dry density over 2.5 feet of random fill
placed at 80% of Standard Proctor dry density, to serve as grading (platform
fill) and radon attenuation layers.
The second cover is based on an ET cover design that contains a compacted clay layer
meant to provide added protection to minimize radon fluxes. Cover 2 differs from Cover
1 in that it includes a clay layer between the water storage layer and the platform fill
layers. The lower platform fill is based on the assumption that a minimum of 3 feet of
random fill has already been placed above Cell 2 and that the lower 2.5 feet was placed at
80% while the upper 0.5 feet of that material will be compacted along with 0.9 feet of
additional random fill to 95%.
Cover 3 is a 3.14-m (10.3-ft) thick ET cover design with a gravel layer that consists from
top to bottom of:
• 15 cm (0.5 ft) of a gravel-amended topsoil admixture to promote revegetation
and provide for protection against erosion and frost damage
• 107 cm (3.5 ft) of random fill soil placed at 85% of Standard Proctor dry
density to serve as a water storage, biointrusion, and radon attenuation layer

E-4
• 31 cm (1 ft) of gravel to serve as a capillary break to inhibit vertical migration
of water
• 162 cm (5.3 ft) of random fill soil comprised of 2.8 feet random fill
compacted to 95% of Standard Proctor dry density over 2.5 feet of random fill
placed at 80% of Standard Proctor dry density, to serve as grading (platform
fill) and radon attenuation layers.
The third cover is based on an ET cover design that contains a gravel layer. Cover 3
differs from Cover 1 in that it includes a gravel layer between the water storage and
platform fill layers to create a capillary break. The lower platform fill is based on the
assumption that a minimum of 3 feet of random fill has already been placed above Cell 2
and that the lower 2.5 feet was placed at 80% while the upper 0.5 feet of that material
will be compacted along with 2.3 feet of additional random fill to 95%.
Cover 4 is a 1.91-m (6.25-ft) thick rock cover design with a compacted clay layer that
consists from top to bottom of:
• 8 cm (0.25 ft) of rock to serve as armor/riprap to provide for long-term surface
resistance (layer not included in numerical model)
• 61 cm (2 ft) of random fill soil compacted to 95% of Standard Proctor dry
density to serve as a frost barrier and radon attenuation layer
• 31 cm (1 ft) of compacted clay compacted to 90% of Modified Proctor dry
density to serve as a radon attenuation layer
• 92 cm (3 ft) of random fill soil comprised of 1 foot random fill compacted to
95% of Standard Proctor dry density over 2 feet of random fill placed at 80%
of Standard Proctor dry density, to serve as grading (platform fill) and radon
attenuation layers.

E-5
The fourth cover design (IUC, 2000; Denison Mines, 2009) is based on the currently
permitted cover design. The rock cover design with a compacted clay layer was based on
technology previously recommended by the U.S. Nuclear Regulatory Commission
(NRC). However, recent advances in cover design technology have emphasized the
construction of vegetated, monolithic ET covers for minimizing infiltration through
engineered cover systems, particularly in arid and semiarid regions. The presence of
vegetation in an ET cover is expected to enhance evapotranspiration and to significantly
reduce infiltration of water into the tailings. The rock cover is not anticipated to become
vegetated; therefore infiltration is expected to be much higher for the rock cover
compared to the ET cover designs.
The conceptual model designs and hydraulic properties for the different cover materials
used to parameterize the HYDRUS models are summarized in Table E-1. The soil water
retention and hydraulic conductivity curves for the different cover materials are plotted in
Figures E-1 and E-2. Original laboratory datasheets used to support the hydraulic and
geotechnical properties are tabulated in Attachment E-1.
INFILTRATION MODEL CONFIGURATION
Grid Spacing
The finite element nodes were discretized in the vertical direction to simulate layers in
the different cover systems. Construction of the finite element mesh is dependent on
surface and bottom boundary conditions and represented lithologic heterogeneities due to
stratigraphic layering (Simunek et al., 2009). As a result, node spacing was finer than the
material layers in order to simulate steep hydraulic gradients which result from transient
wetting (precipitation and infiltration) and drying (evapotranspiration) fronts. Fine grid
spacing was necessary to accurately simulate water flow through the unsaturated zone
because hydraulic properties vary significantly as a function of moisture content and
pressure head. Because hydraulic properties vary much faster and on a finer scale near
the land surface due to rapid changes in atmospheric conditions (daily variations in
precipitation and evapotranspiration were modeled), the node spacing varied between 0.1

E-6
and 1 cm near the top of the domain representing the cover system. Due to the large
contrast in hydraulic conductivity for the compacted clay and gravel layers, the node
spacing immediately above and below these layers decreased in a step-wise manner to
0.35 cm. In order to reduce errors due to numerical dispersion, the ratio between
neighboring elements did not exceed 1.5 (Simunek et al., 2009).
Initial Conditions
Initial conditions for the infiltration model were determined separately for each scenario.
The initial pressure head distribution was determined by evaluating a long-term (114-
year) simulation that used the concatenated atmospheric input file as an upper boundary
condition (i.e., the 57-year climate record repeated twice). The pressure head distribution
for the final time step of the 114-year simulation (Figure E-3) was used as the initial
condition for the transient simulations used to predict water infiltration rates through the
cover. The model was then rerun using these initial conditions, and the long-term water
infiltration rates were averaged during a second 114-year simulation. The methodology
implemented to establish the initial conditions for the site is a commonly accepted
approach for solving hydrogeologic modeling problems.
Boundary Conditions
The upper surface of the model domain was simulated with an atmospheric boundary
condition, while the lower boundary of the cover system was simulated as a unit gradient.
The amount of precipitation was based on the 57-year climate record 1932-1988.
Development of the climate record for the site is discussed in detail within the main body
of the report. All model simulations assumed an anticipated scenario with a maximum
rooting depth of 107-cm and an anticipated root density distribution, as supported by data
presented in Appendix D. Concatenation of the atmospheric boundary file considered an
anticipated scenario assuming 40% vegetative cover. The winter months that included
December, January, and February were assigned a transpiration rate of zero, and only
evaporation was simulated in the HYDRUS-1D models. Cover 4 was the only model that
did not simulate transpiration; only evaporation was simulated because this cover design

E-7
is not expected to become vegetated. The sensitivity of the infiltration modeling results
to the parameter values assigned to represent the vegetation, reduced performance due to
biointrusion, and the amount of precipitation that may occur at the site, was evaluated as
part of Appendix G.
The HYDRUS models did not include runoff and 100% of the precipitation was allowed
to evaporate or infiltrate into the top layer of the cover. Each scenario was modeled with
daily precipitation and ET input using the 57-year climate record. The use of daily rather
than hourly input yields nearly identical results, as evaluated in Appendix F.
INFILTRATION MODEL RESULTS
Water Flux
The model-predicted water flux rate through the four different tailings cell covers during
a typical 57-year climate record are plotted in Figure E-4. The model-predicted flux rates
are plotted on linear and semilog scales to illustrate differences in infiltration rates
between the four different cover designs. The infiltration rates plotted on the linear scale
demonstrate significantly improved performance of the three ET covers as compared to
the currently permitted rock-cover design. The infiltration rates plotted on the semilog
scale demonstrate that all three ET covers were predicted to behave in a similar manner,
with only nominal differences in model-predicted flux rates between designs. The
average infiltration rate predicted to enter the top of the tailings cells for all four
simulations is summarized in Table E-2. Overall, the three infiltration models for the ET
covers showed significantly improved performance compared to the currently permitted
rock cover design. For example, the average model-predicted long-term infiltration rates
were reduced from 9.2x10-3 centimeters per day (cm/d) for the original rock cover design
(Model 4) to 1.2x10-4 cm/d for the monolithic ET cover design (Model 1) and 3.1x10-5
cm/d for the ET cover design with a capillary break (Model 3), a reduction of 75 to 300x,
respectively. The increased performance and reduction of infiltration for the ET covers is
attributed to the presence of vegetation and associated transpiration. Simulations for the
three ET covers produced two instances that resulted in temporary short-term surface

E-8
ponding, while simulations of the fourth cover design (due to decreased saturated
hydraulic conductivity of the surface layer and increased subsurface moisture contents)
produced many instances in which surface ponding occurred; however, as noted above,
runoff was not simulated so that estimates of infiltration through the cover, via surface
ponding, would be representative of the expected flat nature (0.2% slope) of the surface.
For all HYDRUS simulations the water mass balance errors did not exceed 1%. As a
general rule-of-thumb, mass balance errors that do not exceed 3% are considered
acceptable.
SELECTION OF PREFERRED COVER DESIGN
The fourth cover design can be eliminated as a design possibility because the model
predicted much higher rates of infiltration; furthermore, because the three ET cover
designs (Models 1, 2, and 3) contain a soil-gravel admixture layer to minimize long-term
surface erosion, the construction of a rock cover to cap the tailings is not advantageous
and would lead to increased rates of water flow through the tailings.
Comparison of Model 1 and Model 2 indicates that the inclusion of a compacted clay
layer does not significantly reduce infiltration rates compared to the monolithic cover.
Furthermore, performance of the second cover design does not account for reduced
performance of the compacted clay layer. Reduced performance of the compacted clay
layer could occur as a result of shrinking and swelling of the clay particles due to
repeated wetting and drying, or desiccation and cracking during installation. Cracking
within the clay layer could lead to desiccation thereby affecting long-term moisture
contents and the materials effectiveness as a radon attenuation barrier. Cracking could
also lead to preferential flow and increased rates of infiltration comparable to results
predicted for the first cover design. The clay layer is not necessary for radon attenuation
(see Appendix H). Uncertainty of the materials performance, and potential similarity to
infiltration rates as compared to the monolithic ET cover (assuming reduced
performance), effectively eliminates consideration of the second cover design.

E-9
Comparison of Model 1 and Model 3 indicates that the inclusion of a capillary break
results in a moderate reduction of infiltration compared to the monolithic cover.
However, performance of the third cover design does not account for reduced
performance of the gravel layer. Reduced performance of the gravel layer could occur if
fines migrate into the material. Plugging of pore spaces would tend to produce a material
type more similar to the overlying soil and lead to increased rates of infiltration
somewhat comparable to results predicted for the first cover design (assuming reduced
performance). The potential migration of fines could be minimized by the inclusion of
different filtering media (soil layers); however, inclusion of such materials would
significantly increase the difficulty to construct and lead to uncertainty in performance.
Given the simplicity, the monolithic ET cover can be constructed with greater certainty
and quality assurance. The difference in performance between the covers is nominal, and
the addition of any filtering media only adds more complexity and increased uncertainty
with only a minimal amount of potential benefit.
Therefore, the design and construction of a monolithic ET cover is the preferred
alternative for infiltration control. The model-predicted water flux through the
monolithic ET cover indicates that the available storage capacity of the cover should be
sufficient to significantly reduce infiltration, and the ET cover should function properly
as designed. The transport of water below the rooting zone and into the tailings material
would occur when the storage capacity of the overlying soil materials is exceeded; for
example, during multi-consecutive years or longer that receive above average amounts of
annual or winter precipitation. For the monolithic ET cover (Model 1), breakthrough of
water through the bottom of the cover, beginning at about year 48, resulted from the
occurrence of three consecutive years that received above average amounts of winter
precipitation followed by another seven years that received above average amounts of
annual precipitation.
Moisture Contents for the Preferred Cover Design
Moisture content variations at five depth intervals (50, 100, 150, 200, and 284 cm) during
a typical 57-year climate record are plotted in Figure E-5 for the monolithic ET cover.

E-10
Within the rooting zone, at 50-cm and 100-cm depths, the model-predicted volumetric
water content varies from a low of roughly 11% up to a high of 36%; while below the
rooting zone, at 150-cm and 284-cm depths, the model-predicted volumetric water
content varies from a low of roughly 14% up to a high of 25%. Throughout the year,
within the rooting zone, the amount of moisture increases during the winter months
because the amount of precipitation generally exceeds the amount of evapotranspiration;
and then, during spring and summer, the amount of moisture decreases in response to
increased evapotranspirative fluxes. The time series plots verify that within the rooting
zone the moisture content has large year-to-year variability, and that below the rooting
zone the amount of moisture does not significantly vary through time due to efficient root
water uptake within the water storage layer. Overall, the amount of moisture in the cover
predicted by the infiltration model generally exceeds the amount of moisture used in the
radon attenuation model (12.6%). Therefore, the design and construction of a monolithic
ET cover is the preferred alternative for radon control.
CONCLUSIONS
The construction of a monolithic ET cover is the preferred alternative to minimize
infiltration and meet the radon attenuation standard. The proposed cover design will be
sufficient to provide adequate thickness to protect against frost penetration, provide
adequate water storage capacity to minimize the rate of infiltration into the underlying
tailings, and provide long-term moisture within the cover to attenuate radon fluxes.
REFERENCES
Advanced Terra Testing, Inc. (1996), Geotechnical lab report prepared July/August 1996.

Chen and Associates, Inc. (1978), Earth Lined Tailings Cells, White Mesa Uranium
Project, Blanding, Utah, Report prepared for Energy Fuels Nuclear, Inc. on 18 July 1978.

Chen and Associates, Inc. (1979), Soil Property Study, Proposed Tailings Retention
Cells, White Mesa Uranium Project, Blanding, Utah, Report prepared for Energy Fuels Nuclear, Inc. on 23 January 1979.

E-11
Chen and Associates, Inc. (1987), Physical Soil Data, White Mesa Project, Blanding
Utah, Report prepared for Energy Fuels Nuclear, Inc. 1987.

D’Appolonia Consulting Engineers, Inc. (1979), Engineering Report, Tailings
Management System, White Mesa Uranium Project, Blanding, Utah, Report prepared for Energy Fuels Nuclear, Inc. June 1979.
D’Appolonia Consulting Engineers, Inc. (1982), Letter Report, Section 16 Clay Material
Test Data, White Mesa Uranium Project, Blanding, Utah, Report prepared for Energy
Fuels Nuclear, Inc. on 8 March 1982. Denison Mines (USA) Corp. (2009), Reclamation Plan, White Mesa Mill, Blanding,
Utah, Radioactive Materials License No. UT1900479, Revision 4.0, November 2009.

Dwyer, S.F. (2003), Water balance measurements and computer simulations of landfill covers, Ph.D. Dissertation, University of New Mexico at Albuquerque, pp. 250.
Geosyntec Consultants (2006), Stockpile Evaluation Tailings Cell 4A, White Mesa Mill -
Technical Memo prepared for International Uranium (USA) Corporation on 23
January 2006. International Uranium (USA) Corporation (2000), Reclamation Plan, White Mesa Mill,
Blanding, Utah, Source Material License No. SUA-1358, Docket No. 40-8681,
Revision 3.0, July 2000.
Rogers & Associates Engineering Corporation (1996), Report of radon diffusion coefficient measurements [and density of solids for clay borrow source material].

Simunek, J., M. Sejna, H. Saito, M. Sakai, and M. Th. van Genuchten (2009), The
HYDRUS-1D Software Package for Simulating the Movement of Water, Heat, and Multiple Solutes in Variably Saturated Media, Version 4.08, HYDRUS Software
Series 3, Department of Environmental Sciences, University of California Riverside,
Riverside, California, USA, pp. 330.

TITAN Environmental Corporation (1996), Tailings Cover Design, White Mesa Mill,
Blanding Utah, Report prepared for Energy Fuels Nuclear, Inc. September 1996.

Western Colorado Testing, Inc. (1999), Additional Clarifications to the White Mesa Mill
Reclamation Plan, Report prepared for International Uranium (USA) Corporation and
submitted to the U.S. Nuclear Regulatory Commission 15 April 1999.

E-12

TABLE E‐1. COVER MODEL DESIGNS AND HYDRAULIC PROPERTIES USED TO PARAMETERIZE THE HYDRUS MODELS 
Model 
Layer Purpose Thickness 
(cm) 
Residual soil 
water content 
θr 
(% vol) 
Saturated soil 
water content 
θs 
(% vol) 
Curve fitting parameters in the soil 
water retention function a 
Saturated hydraulic conductivity 
in the vertical direction 
Ksat 
(cm/d) 
Source 
α 
(cm‐1) 
n 
(‐) 
Model 1: Monolithic ET Coverb 
1 Erosion Protection 15 0.045 0.254 0.0145 1.406 5.6 1
2 Water Storage & Radon 
Attenuation 107 0.055 0.404 0.0145 1.406 7.4 2 
3 Upper Platform Fill (High 
Compaction) Radon Attenuation 86 0.046 0.334 0.0229 1.261 3.6 2 
4 Lower Platform Fill (Base Grade) 
Radon Attenuation 76 0.059 0.439 0.0125 1.461 10.4 2 
Model 2: ET Cover with a Compacted Clay Layerb 
1 Erosion Protection 15 0.045 0.254 0.0145 1.406 5.6 1
2 Water Storage & Radon 
Attenuation 107 0.055 0.404 0.0145 1.406 7.4 2 
3 Clay Radon Attenuation 31 0.08 0.391 0.0122 1.386 0.032 3
4 Upper Platform Fill (High 
Compaction) Radon Attenuation 43 0.046 0.334 0.0229 1.261 3.6 2 
5 Lower Platform Fill (Base Grade) 
Radon Attenuation 76 0.059 0.439 0.0125 1.461 10.4 2 
Model 3:  ET Cover with a Gravel Capillary Break Layerb 
1 Erosion Protection 15 0.045 0.254 0.0145 1.406 5.6 1
2 Water Storage & Radon 
Attenuation 107 0.055 0.404 0.0145 1.406 7.4 2 
3 Capillary Break 31 0.03 0.33 2.8 2.5 112,300 4
4 Upper Platform Fill (High 
Compaction) Radon Attenuation 86 0.046 0.334 0.0229 1.261 3.6 2 
5 Lower Platform Fill (Base Grade) 
Radon Attenuation 76 0.059 0.439 0.0125 1.461 10.4 2 
Model 4:  Conventional Rock Cover with a Compacted Clay Layerc 
1 Frost Barrier & Radon 
Attenuation 61 0.046 0.334 0.0229 1.261 3.6 2 
2 Clay Radon Attenuation 31 0.08 0.391 0.0122 1.386 0.032 3
3 Upper Platform Fill (High 
Compaction) Radon Attenuation 31 0.046 0.334 0.0229 1.261 3.6 2 
4 Lower Platform Fill (Base Grade) 
Radon Attenuation 61 0.059 0.439 0.0125 1.461 10.4 2 

E-13

(continued on next page) 
Footnotes: 
a. The van Genuchten‐Mualem single‐porosity soil‐hydraulic‐property model was selected to characterize the soil‐hydraulic properties. 
b. The terminal rooting depth was located at 107‐cm depth below ground surface. 
c. Original cover design.  No transpiration was simulated.  Only evaporation was assumed to occur.  
Sources: 
1. Erosion protection layer hydraulic properties were estimated by correcting for the volume and mass of gravel added to the random fill material. The residual water content was corrected on 
a volume basis, while the saturated hydraulic conductivity was corrected on a mass basis. The hydraulic parameter values for the soil were adjusted assuming the gravel‐soil mixture was 
composed of 25% gravel by weight, and that the soil dry bulk density with no gravel was 1.59 g/cm3 while the gravel density of solids was 2.65 g/cm3. The calculated volume percent of 
gravel, assuming no rock porosity, was 19%.  The saturated water content was adjusted to correspond to values used in the radon attenuation model (porosity equal to one minus ratio of in‐
place dry bulk density to density of solids).  The density of solids was assumed to equal 2.67 g/cm3. 
2. Hydraulic properties of the water storage (placed at 85% Standard Proctor), upper platform fill (compacted to 95% Standard Proctor), and lower platform fill (placed at 80% Standard 
Proctor) layers were predicted using the soil‐properties database in HYDRUS.  The average grain size distribution of stockpiled random fill (loam to sandy clay) from 32 samples was 44% 
sand, 36% silt, and 20% clay, while the maximum Standard Proctor dry bulk density was 1.87 g/cm3 (Attachment E‐1).  The in‐place dry bulk density was taken as a percentage of the 
maximum density for each layer.  The saturated water contents were then adjusted to correspond to values used in the radon attenuation model (porosity equal to one minus ratio of in‐
place dry bulk density to density of solids).  The density of solids was assumed to equal 2.67 g/cm3. 
3. Compacted clay (90% Modified Proctor) hydraulic properties were predicted using the soil‐properties database in HYDRUS.  The average grain size distribution of 16 samples obtained from 
the clay source identified at Section 16 was 28% sand, 37% silt, and 35% clay, while the maximum Modified Proctor dry bulk density (from 4 samples) was 1.71 g/cm3 (Attachment E‐1).  
Samples collected from 0 to 5 feet depth below ground surface were not included as part of the averaging because the material was slightly more sandy.  The in‐place dry bulk density was 
taken as a percentage of the maximum density.   The saturated water content was then adjusted to correspond to values used in the radon attenuation model (porosity equal to one minus 
ratio of in‐place dry bulk density to density of solids).  The density of solids was assumed to equal 2.53 g/cm3.  The saturated hydraulic conductivity was assumed to equal 3.7x10‐7 cm/s. 
4. Gravel hydraulic properties were taken from measurements of pea gravel reported by Dwyer (2003). 

E-14

TABLE E‐2. AVERAGE INFILTRATION RATE PREDICTED TO ENTER THE TOP OF THE TAILINGS CELLS FOR THE FOUR 
COVER DESIGNS MODELED 
Model Cover Design Water Flux 
(cm/d) 
Water Flux 
(mm/yr) 
Water Flux 
(% of Average Annual 
Precipitation) 
Amount of water 
entering tailings after 
200 years 
Model 1 Monolithic ET cover 1.2 x 10‐4 0.45 0.14% 0.09 m (0.29 ft) 
Model 2 ET cover with a 
compacted clay layer 5.4  x 10‐5 0.20 0.062% 0.04 m (0.13 ft) 
Model 3 ET cover with a gravel 
layer 3.1 x 10‐5 0.11 0.036% 0.02 m (0.066 ft) 
Model 4 Rock cover with a 
compacted clay layer 9.2 x 10‐3 34 11% 6.7 m (22 ft) 
Note: The average annual precipitation for the 57‐year climate record was recorded 1932‐1988. 

E-15
Figure E-1. Semilog plot of the soil water retention curves (pressure head as a function of water content) for the different cover materials. The placement or compaction density
for the random fill (RF) as a percentage of the Standard Proctor (SP) maximum dry
density is noted for clarification.

E-16
Figure E-2. Semilog plot of the unsaturated hydraulic conductivity curves (log hydraulic conductivity as a function of water content) for the different cover materials. The
placement or compaction density for the random fill (RF) as a percentage of the Standard
Proctor (SP) maximum dry density is noted for clarification.

E-17
Figure E-3. Initial pressure head distributions for the four tailings cell cover designs.

E-18
Figure E-4. Model-predicted water flux rates through the four tailings cell cover designs during a typical 57-year climate record. The model-predicted
flux rates are plotted on linear (upper figure) and semilog (lower figure) scales
to illustrate differences in infiltration rates between the different cover
designs.

E-19
Figure E-5. Moisture content at nested intervals within the monolithic ET cover (Model 1) during a typical 57-year climate record. Water content
variations within the rooting zone (upper figure) and below the rooting zone
(lower figure) are plotted. The bottom of the rooting zone is located at 107-
cm depth, while the bottom of the cover corresponds to 284-cm depth.

APPENDIX F
EVALUATION OF THE EFFECTS OF STORM INTENSITY ON
INFILTRATION THROUGH EVAPOTRANSPIRATION COVER

F-1
APPENDIX F
EVALUATION OF THE EFFECTS OF STORM INTENSITY ON
INFILTRATION THROUGH EVAPOTRANSPIRATION COVER
The purpose of this appendix is to evaluate the sensitivity of the infiltration modeling
results to storm intensity and ponding of water on the monolithic evapotranspiration (ET)
cover surface. In the modeling presented in the Infiltration and Contaminant Transport
Modeling (ICTM) Report, precipitation was input on a daily basis, thus ignoring storm
intensity. However, summer monsoon precipitation events are characterized by severe
thunderstorms of short duration. To test the importance of simulating storm intensity and
ponding, the HYDRUS model of the monolithic ET cover was run using both hourly and
daily precipitation rates. Model-predicted soil moisture content and water flux within the
proposed monolithic ET cover were compared for the simulations using hourly versus
daily precipitation data as input.
To evaluate the effect of storm intensity on infiltration, two scenarios were modeled:
• a 10-day period with a single intense storm event of 4.4 cm
• a 92-day period representing the summer monsoon season (July 1 through
September 30) with 19.2 cm of precipitation (recorded in 1987), the greatest
recorded precipitation for the period July 1 through September 30.
The hourly scenarios were modeled with hourly precipitation and ET input, while the
daily scenarios were modeled with daily precipitation and ET input. The July 1 through
September 30 time period was simulated because severe thunderstorms occur most
commonly during this period (summer monsoon).

F-2
CONCEPTUAL COVER DESIGN
The proposed conceptual cover design consists of a 2.84-m (9.3-ft) thick monolithic ET
cover as described in Appendix E.
INFILTRATION MODEL CONFIGURATION
The hydraulic properties of the cover material, and grid spacing of the HYDRUS-1D
model, are summarized in Appendix E.
Initial Conditions
The initial water content within the cover material was taken from the base case model
simulation (i.e., assuming 40% vegetative cover, anticipated rooting depth/distribution,
and anticipated climate). The initial water content for the top 50-cm portion of the cover
is plotted in Figure F-1. The change in water content at a depth of 15 cm is a result of the
abrupt change in material properties between the erosion control layer and water storage
layer.
Boundary Conditions
The upper surface of the model domain was simulated with an atmospheric boundary
condition. The lower boundary of the cover system was simulated as a unit gradient.
The HYDRUS models did not include runoff and 100% of the precipitation was allowed
to evaporate or infiltrate into the top layer of the cover.
Generally, surface ponding and runoff would be expected to occur for conditions in
which the precipitation rate exceeds the infiltration capacity (e.g., saturated hydraulic
conductivity) of the surface soil. If water accumulates on the surface of the ET cover
(i.e., ponding), evaporative losses from the surface layer are accounted for by applying a
net infiltration rate, and the flux of water along the boundary is governed by the hydraulic

F-3
potential until all the water infiltrates. Runoff was not simulated, but would act to reduce
infiltration.
Precipitation. For the 10-day time series, one precipitation extreme was considered
which corresponded to the 100-year average recurrence interval (ARI) for a precipitation
event lasting 1-hour. For the Blanding, Utah, weather station, this amounted to 4.4 cm of
precipitation distributed during a 1-hour storm. Precipitation frequency estimates were
obtained from the National Oceanic and Atmospheric Administration (NOAA)
Hydrometeorological Design Studies Center Web site
(http://www.nws.noaa.gov/oh/hdsc/index.html). The maximum amount of precipitation
recorded on any given day at the Blanding weather station during the period-of-record
(1904-2005) from July through September only exceeded 4.4 cm on three occasions. The
10-day simulation that used hourly input data assumed that all of the precipitation (4.4
cm) fell between 13:00 and 14:00 on the first day, which was aimed at reproducing a
severe rain event. The base case simulation had precipitation input on a daily basis for
the first day (4.4 cm for the day). The remaining days were assumed to lack
precipitation.
For the 92-day time series, precipitation data from 1987, the year with the greatest
recorded amount of monsoon season precipitation (19.2 cm from July 1 through
September 30) were used as model input. For the 92-day simulations that used hourly
input data, all precipitation was assumed to occur between 13:00 and 14:00 for each day.
Daily precipitation for the 1987 monsoon season is plotted in Figure F-2. There were
three storm events that received 2 to 4 cm of precipitation and four storm events that
received 1 to 2 cm of precipitation. The maximum amount of precipitation received on
any given day during the 1987 season was approximately 4 cm, which corresponded to a
50-year ARI 1-hour duration precipitation event. The maximum precipitation event of
3.99 cm occurred on August 23 (day 54), the day after a 1.57-cm event and was followed
by 2.51-cm and 0.89-cm events, which amounts to nearly 50% of the precipitation
received during the entire 1987 monsoon season.

F-4
Precipitation data recorded during 15-minute increments were obtained for the 1987
monsoon season from the NOAA National Climatic Data Center Web site
(http://www.ncdc.noaa.gov/oa/ncdc.html) for the Blanding weather station. The 15-
minute precipitation data were obtained in order to compare the actual and modeled
(hourly input) storm intensities. The modeled storm intensities using hourly input
generally exceeded the actual storm event intensities. Furthermore, the finer-resolution
precipitation data revealed that the afternoon storm events that occurred between 12:00
and 16:00 generally consisted of a handful of short-duration storms in which
approximately 0.3 to 0.8 cm of water were distributed during 15 and 30 minutes,
respectively, at intermittent times throughout the day. The maximum precipitation event
of 3.99 cm received nearly 90% of the rainfall throughout a 1.75-hour storm. As a result,
modeling with hourly input of precipitation rates, which assumed that all of the
precipitation occurred during a 60-minute storm event, simulated more intense storms
compared to what was actually recorded during the 1987 monsoon season.
Evapotranspiration. Simulations with hourly input data required a finer temporal
resolution of potential evaporation (PE) and potential transpiration (PT) than the values
used in which precipitation was input on a daily basis. The cumulative potential
evapotranspiration (PET) for any given day during July, August, and September was
0.65, 0.52, and 0.40 cm, respectively. The hourly variation in PET for the monsoon
season is plotted in Figure F-3. The amount of PE and PT during the hours 0:00-6:00 and
18:00-24:00 was assumed to represent 1% of the total daily value. The rates of PE and
PT between the hours of 6:00 and 18:00 were simulated as a step-increasing function
with a maximum value occurring at 12:00. The rate of PE and PT was not varied during
the 10-day time frame, and was assumed equal to the rate determined for July.
INFILTRATION MODEL RESULTS
Ten-day Period
The daily precipitation input example did not produce any surface ponding because the
precipitation intensity did not exceed the capacity of the soil to transmit the water. The

F-5
saturated hydraulic conductivity of the surface layer of the ET cover used in all model
simulations was 6.5 x 10-5 cm/sec (5.6 cm/day; 0.23 cm/hr), which is greater than the
amount of precipitation input on a daily basis (4.4 cm/day). However, the hourly
precipitation input example did result in temporary ponding for a 10-hour period between
13:02-23:00 with a maximum ponding depth of 3.2 cm. The formation of ponding for the
hourly precipitation input example was expected because the saturated hydraulic
conductivity of the ET cover surface (5.6 cm/day; 0.23 cm/hr) was about twenty times
lower than the precipitation rate (4.4 cm/hr).
Differences in subsurface soil moisture and water flux were compared for the daily and
hourly input examples to yield insight into potential differences that may arise by
simulating infiltration through the cover with daily input data rather than hourly data.
The moisture content distribution within the upper 50-cm of the cover after day two and
day six is plotted in Figure F-4. The water content between the two simulations are
similar, with the soil profile being slightly more wet on day two for the simulation that
used hourly input data. The difference is attributed to the ponding which would transmit
slightly more water deeper into the profile as compared to a scenario that does not lead to
ponding. However, after day six, the differences in water content diminished (Figure F-
4) and the instantaneous water flux within the cover for the simulations using hourly and
daily input were nearly identical (Figure F-5). The infiltration pulse from the
precipitation event was not transmitted below 50-cm depth, which demonstrates that the
approach using daily or hourly input data would yield similar flux rates through the cover
system into the underlying tailings cell. Based on this evaluation, the model
simplification of using daily input rather than hourly input does not affect the predictive
results.
Monsoon Season
Model predicted water content at 30-cm and 50-cm depths during the 1987 summer
monsoon season (19.2 cm of precipitation) using daily and hourly input data are plotted
in Figure F-6. The simulation using hourly input data produced eight instances in which
temporary ponding occurred. The maximum surface pressure head (i.e., ponding depth)

F-6
was 2.8 cm, and saturated conditions at the surface remained for approximately 8 hours
following the August 23 storm event of 3.99 cm. Runoff was not included in the
simulations in order to be conservative. A similar change in water content was noted at
30-cm and 50-cm depths for the simulations that used daily and hourly input data, which
suggests that despite greater variations in water content within the shallow subsurface
(e.g., the top 10 cm), using daily or hourly input data result in the same flux rates
following back-to-back, high-intensity rainstorm events. Little variation in water content
occurred beneath 75-cm depth for the daily and hourly input. Therefore, the method of
modeling storm events through daily or hourly input has little effect on the prediction of
infiltration through the ET cover (Figure F-7).
CONCLUSIONS
The results of the sensitivity analysis provide justification that daily inputs of
precipitation predict conditions that are representative of field conditions that would
occur during high-intensity monsoon rainstorm events (as modeled using hourly input
data). Hourly input allows for surface ponding, which leads to differences in the
distribution of soil moisture in the short term in the upper 50 cm of the cover relative to
the modeling with daily input. However, little variation in water content and water flux
occurred deeper in the ET cover between the simulations using daily and hourly input.
This was true for both the 10-day scenario in which a single 100-year ARI storm of 4.4
cm was simulated and the 92-day scenario in which the precipitation input was set based
on the maximum recorded monsoon season of 1987 in which 19.2 cm of precipitation
occurred between July 1 and September 30. Based on this evaluation, the model
simplification of using daily input rather than hourly input does not affect the predictive
results. The exclusion of runoff, and simulation with a one-dimensional model, results in
representative estimates of infiltration through the cover.

F-7
Figure F-1. Initial water content for the top 50-cm portion of the monolithic evapotranspiration (ET) cover.

F-8
Figure F-2. Precipitation recorded during 1987 monsoon season (July 1 through September 30). The 1987 monsoon season recorded the maximum amount of
summertime precipitation (19.2 cm).

F-9
Figure F-3. Atmospheric boundary condition representing daily values of potential evapotranspiration (PET) during July, August, and September. The amount of potential
evaporation (PE) and potential transpiration (PT) were calculated assuming 40%
vegetative cover.

F-10
Figure F-4. Water content within the upper 50-cm of the cover system after day two
(upper figure) and day six (lower figure) for the simulations that used hourly and daily
input. Both simulations reproduced a 100-year average recurrence interval 1-hour long
precipitation event.

F-11
Figure F-5. Water flux within the upper 50-cm of the cover system after day two (upper
figure) and day six (lower figure) for the simulations that used hourly and daily input.
Both simulations reproduced a 100-year average recurrence interval 1-hour long
precipitation event. Negative and positive water flux rates correspond to the downward
and upward movement of water, respectively. Note the change in scale for flux between the two plots.

F-12
Figure F-6. Model predicted water content at 30-cm and 50-cm depths during the 1987 monsoon season (July 1 through September 30), which recorded the maximum amount of
summertime precipitation (19.2 cm), for the simulations that used daily and hourly input.

F-13
Figure F-7. Water flux within the upper 100-cm of the cover system at the end of the 1987 monsoon season for the simulations that used daily and hourly input. Negative and
positive water flux rates correspond to the downward and upward movement of water,
respectively.

APPENDIX G
SENSITIVITY ANALYSIS COMPARING INFILTRATION RATES
THROUGH THE EVAPOTRANSPIRATION COVER BASED ON
COVER VEGETATION, BIOINTRUSION, AND PRECIPITATION

G-1
APPENDIX G
SENSITIVITY ANALYSIS COMPARING INFILTRATION RATES THROUGH
THE EVAPOTRANSPIRATION COVER BASED ON COVER VEGETATION,
BIOINTRUSION, AND PRECIPITATION
To test the importance of simulating reduced performance of the vegetative component of
the cover system, and how increased precipitation could influence the transport of water
through the monolithic evapotranspiration (ET) cover, the HYDRUS model was run
using different assumptions aimed at characterizing an anticipated scenario and reduced
performance scenarios. The reduced performance scenarios were based on conservative
assumptions that are expected to over predict the potential impacts (e.g., lead to increased
fluxes for water flow); while the anticipated scenarios were based on professional
judgment that reflect assumptions considered to be representative of expected conditions.
Rates of model-predicted water flux entering the tailings cells were compared between
simulations using different input assumptions. The effects on moisture content by the
parameters used to assess establishment of vegetation and root water uptake are evaluated
in this appendix to determine whether moisture contents that are input into the radon
model are conservative. Impacts to hydraulic properties of cover material and water
infiltration rates due to biointrusion of animals were also evaluated.
SENSITIVITY ANALYSES
Vegetation Establishment
Percent Cover and Rooting Depth/Distribution. Empirical data used as model input
for the HYDRUS model regarding the ecological characteristics of the vegetation
(rooting depth and root distribution) and established plant community (percent cover)
were obtained from the literature and nearby lysimeter studies (Monticello) Background
information that supports the empirical data and are included as part of Appendix D.
To evaluate the effect of vegetation establishment, four scenarios were modeled:

G-2
• an anticipated scenario with a maximum rooting depth of 107-cm and an
anticipated root density distribution
• a reduced performance scenario with a maximum rooting depth of 68-cm and a
reduced root density distribution
• an anticipated scenario assuming 40% vegetative cover
• a reduced performance scenario assuming 30% vegetative cover.
The two rooting depths and root density distributions are illustrated in Figure G-1. A
larger percent cover results in a higher rate of partitioned transpiration relative to
evaporation, while a smaller percent cover results in a lower rate of partitioned
transpiration relative to evaporation. A lower percent cover results in less root water
uptake (i.e., transpiration) and more evaporation, relatively speaking, which would lead
to increased fluxes and increased moisture contents. In regard to rooting depth and root
density distribution, a greater rooting depth and distribution function results in more root
water uptake (i.e., transpiration), which would lead to reduced fluxes and reduced
moisture contents within the rooting zone.
Root Water Uptake and Evaporation. Root water uptake will vary as a function of the
soil water pressure head within the rooting zone, the plant root distribution function (i.e.,
density of roots and rooting depth), and the rate of potential transpiration (PT). The rate
of PT is assigned as part of the atmospheric upper boundary condition, which HYDRUS
then uses to compute the actual transpiration (AT) rate as a function of time and space
within the rooting zone. For example, when conditions are extremely dry (i.e., less than
the wilting point) or extremely wet (i.e., near saturation) plants cease to uptake water, and
the AT would be zero. At intermediate soil water conditions, the AT would be a fraction
of the PT. The water stress response function for grass was selected from the default
database in HYDRUS. The database does not distinguish between different species of
grass, and transpiration is assumed to cease at soil water pressures below the assumed
wilting point of -8,000 cm. However, plants in many arid and semiarid environments

G-3
(many of which were selected for the ET cover) commonly maintain transpiration at
significantly lower (more negative) soil water pressures. For example, crested
wheatgrass can survive in soil water conditions where the soil water pressure ranges
between -20,000 and -40,000 cm (Chabot and Mooney, 1985; Brown, 1995).
The rate of potential evaporation (PE) is also assigned as part of the atmospheric input
file. In HYDRUS, the PE rate is reduced to an actual evaporation (AE) rate if a specified
pressure head is reached at the surface. The pressure head at which this occurs is
controlled by equilibrium conditions between soil water and atmospheric water vapor.
All simulations have assumed a minimum surface pressure head of -15,000 cm, which is
the recommended value by the program. When the pressure head at the surface reaches
-15,000 cm the program calculates a reduced, actual evaporation rate.
To evaluate the effect of these two assumptions, and potential impacts on subsurface
water content and fluxes through the ET cover, two scenarios were modeled:
• a scenario with a vegetation wilting point of -30,000 cm and a minimum surface
pressure head of -150,000 cm combined with 40% cover and the anticipated
rooting density/depth distribution
• a scenario with a vegetation wilting point of -30,000 cm and a minimum surface
pressure head of -150,000 cm combined with 30% cover and the anticipated
rooting density/depth distribution.
A more negative soil water pressure head specified at the upper boundary would result in
decreased soil water pressures and reduced moisture contents within the upper 1 to 2
centimeters of the cover; while a more negative soil water pressure head specified for the
wilting point would result in drier conditions within the rooting zone and reduced fluxes
at the bottom of the cover system. The objective of these two scenarios is to determine
whether the moisture contents that are input into the radon attenuation model are
conservative. Unless otherwise noted, all simulations assume the default wilting point of
-8,000 cm and a minimum surface pressure head of -15,000 cm.

G-4
Biointrusion
The probability of reduced performance due to burrowing animals is evaluated here
through an order-of-magnitude calculation, even though to date the site has experienced
only minor problems with burrowing animals (Denison, 2009). Impacts to the cover
from invading woody species was not evaluated because such plants are not expected to
populate the vegetative community, as discussed in Appendix D. Based on empirical
data published in the literature, and the potential species that may use the site as habitat,
any burrowing activity that may occur would be limited to the upper one meter of the
cover, but would not impact the remainder of the cover (1.84 m). Effects from a
burrowing prairie dog are used to evaluate changes in porosity of cover material and
water flux through the cover.
The average prairie dog burrow diameter is reported to be 11 cm, resulting in a cross-
sectional area of one burrow equal to 95 cm2. The typical burrow frequency is 12
burrows per acre (Burns et al., 1989; Cheatheam, 1977). Assuming a cylindrical burrow
that is 100-cm deep, the volume of void space for one burrow would equal 9,500 cm3,
which is equal to a total void space of 114,000 cm3 on a per acre basis. As a comparison,
the total void space for a 100-cm layer of soil, with no burrows, and a porosity of 0.40
would equal 1.6 billion cm3 on a per acre basis. Based on these estimates, the void space
added by burrowing animals would increase the porosity by approximately 0.007%.
Therefore, it is reasonable to conclude that if burrowing animals invade the cover, the
potential impacts to porosity of the cover materials would be minimal.
To evaluate potential impacts from burrows on water flux rates, the amount of
precipitation on a daily basis was assumed to directly recharge the cover system through
the circular holes. Using the long-term average amount of annual precipitation recorded
at the Blanding weather station (1932-1988), the amount of precipitation that would
recharge the cover on a per acre basis through the burrows would equal 98 cm3/d
(assuming burrows are 11-cm diameter and there are 12 burrows per acre). The average
model-predicted long-term infiltration rate through the cover (1.2x10-4 cm/d; see
Appendix E) would equal 4,856 cm3/d of water on a per acre basis minus surface areas

G-5
exposed by burrows. As a conservative estimate, if precipitation were to directly enter all
of the burrows and recharge the lower portion of the monolithic ET cover, the long-term
infiltration rate could be expected to increase approximately 2% from 0.45 to 0.46
millimeters per year. Therefore, it is reasonable to conclude that if burrowing animals
invade the cover the infiltration rate through the cover would not be significantly
affected. Surface runoff into burrows is not expected as burrows usually are surrounded
by a berm of soil.
Precipitation
The anticipated climate record for the White Mesa Mill was taken from historic data
recorded at the Blanding weather station between 1932 and 1988. Development of the
climate record for the site is discussed in detail within the main body of the report. To
evaluate the effects of long-term accumulation of water in the water storage layer and ET
cover performance, the three wettest years on record were inserted into the climate
record. Inclusion of consecutive-wet years is the recommended procedure for evaluating
the effects of increased precipitation on infiltration rates through an ET cover (Khire et
al., 2000).
To evaluate the effect of increased precipitation on a climate-record basis two scenarios
were modeled:
• an anticipated scenario using the 57-year climate record between 1932 and 1988
• an increased precipitation scenario using the 57-year climate record with the three
wettest years consecutively inserted in place of three average years.
The three largest precipitation years in the climate record were 1957 (56.9 cm), 1906
(59.9 cm), and 1909 (62.2 cm). These years replaced precipitation values measured in
years 17 through 19 (between 1946 and 1948), which were 29.4, 35.3, and 35.1 cm,
respectively, and are close to the long-term average. The two climate scenarios are
illustrated in Figure G-2. All of the simulations include the period 1978-1987 (years 47

G-6
through 56), which is a 10-year timeframe characterized by above-average annual
precipitation (38.6 cm).
INFILTRATION MODEL CONFIGURATION
The proposed conceptual cover design consists of a 2.84-m (9.3-ft) thick monolithic ET
cover as described in Appendix E. The hydraulic properties of the cover material, and
grid spacing of the HYDRUS-1D model, are summarized in Appendix E.
Initial Conditions
Initial conditions for the infiltration model were determined separately for each scenario.
The initial pressure head distribution was determined by evaluating a long-term (114-
year) simulation that used the concatenated atmospheric input file as an upper boundary
condition (i.e., the 57-year climate record repeated twice). The pressure head distribution
for the final time step of the 114-year simulation was used as the initial condition for the
transient simulations used to predict water infiltration rates through the cover. The model
was then rerun using these initial conditions, and the long-term water infiltration rates
were averaged during a second 114-year simulation. The methodology implemented to
establish the initial conditions for the site is a commonly accepted approach for solving
hydrogeologic modeling problems.
Boundary Conditions
The upper surface of the model domain was simulated with an atmospheric boundary
condition, while the lower boundary of the cover system was simulated as a unit gradient.
The amount of precipitation was based on the 57-year climate record 1932-1988, or the
modified climate record with increased precipitation, as explained above. The winter
months that included December, January, and February were assigned a transpiration rate
of zero, and only evaporation was simulated in the HYDRUS model. The HYDRUS
models did not include runoff and 100% of the precipitation was allowed to evaporate or

G-7
infiltrate into the top layer of the cover. Each scenario was modeled with daily
precipitation and ET input using the 57-year climate record.
INFILTRATION MODEL RESULTS
Water Flux
The average infiltration rate predicted to pass through the ET cover and enter the top of
the tailings cells during the 57-year climate record for all simulations is summarized in
Tables G-1 and G-2. The simulations include scenarios that account for differences in
the establishment of vegetation (percent cover; rooting depth and distribution), root water
uptake (wilting point), evaporation at the surface (minimum surface pressure head), and
amount of precipitation. Simulations for the ET cover produced a few instances in which
surface ponding occurred; however, runoff was not simulated so that estimates of
infiltration through the cover, via surface ponding, would be conservative. For all
HYDRUS simulations the water mass balance errors did not exceed 1%. As a general
rule-of-thumb, mass balance errors that do not exceed 3% are considered acceptable.
Overall, for simulations that received the same amount of precipitation, the average
infiltration rate is least sensitive to the percentage of vegetation covering the surface and
most sensitive to the rooting depth and root density distribution, while the wilting point
and minimum surface pressure head allowed at the surface have an intermediate effect
(see Figure G-3 and Table G-1). For example, assuming the anticipated rooting
depth/distribution and the anticipated climate record, a change in the percent vegetative
cover from 40% to 30% increased the flux rate from 0.45 millimeters per year (mm/yr) to
0.53 mm/yr. A change from the anticipated to a reduced rooting depth/distribution (for
40% vegetative cover and the anticipated climate record) increased the flux rate from
0.45 mm/yr to 2.3 mm/yr. And finally, a change from the default wilting point to the
decreased wilting point (for 40% vegetative cover, anticipated rooting depth/distribution,
and anticipated climate record) decreased the flux rate from 0.45 mm/yr to 0.19 mm/yr.
The scenario that leads to the highest infiltration rate (2.4 mm/yr) consists of the

G-8
following assumptions: 30% vegetative cover, reduced rooting depth/distribution,
anticipated climate record, and default wilting point.
Assuming all other variables are equal, the average infiltration rate increased for the
simulations that modeled increased precipitation (see Figure G-4 and Table G-2). For
example, assuming the anticipated rooting depth/distribution and 40% vegetative cover,
the change from an anticipated to an increased precipitation record increased the flux rate
from 0.45 mm/yr to 2.0 mm/yr. However, if a decreased wilting point was assumed
together with the increased precipitation record (with anticipated rooting
depth/distribution and 40% vegetation cover) the average flux rate increased from 0.45
mm/yr to 1.2 mm/yr. On the whole, modeling of increased precipitation resulted in a
short-term large magnitude change in infiltration and an asymptotic return to the long-
term average infiltration rate for each scenario modeled following the increased
precipitation in years 17-19 (see Figure G-4). As a comparison, the 10-year period (years
47 through 56 included in both climate records) that received above-average precipitation
resulted in a smaller magnitude change in infiltration; however, the simulations also
predicted a similar asymptotic return to the long-term average infiltration rate (see Figure
G-4).
The model-predicted water flux through the ET cover indicates that the available storage
capacity of the cover should be sufficient to significantly minimize infiltration. The
transport of water below the rooting zone and into the tailings material would occur when
the storage capacity of the overlying soil materials is exceeded; for example, during
multi-consecutive years or longer that receive above average amounts of precipitation.
The lower bound, anticipated, and upper bound long-term average water flux rates
entering the tailings were 0.19 mm/yr, 0.45 mm/yr, and 2.4 mm/yr, respectively, which
would result in approximately 38 mm (0.12 ft), 90 mm (0.3 ft), and 480 mm (1.6 ft) of
water entering the tailings during the 200-year regulatory timeframe, respectively, and
corresponding to an increase in saturated tailings thickness of 67 mm (0.21 ft), 160 mm
(0.53 ft), and 840 mm (2.8 ft), respectively. Therefore a significant build-up of water
(“bathtub effect”) within the cells is not anticipated.

G-9
Moisture Content
Moisture contents at five depths (25, 50, 100, 150, and 284 cm) within and below the
rooting zone of the monolithic ET cover during the anticipated 57-year climate record for
the two scenarios using a decreased wilting point of -30,000 cm and a default wilting
point of -8,000 cm are plotted in Figures G-6, G-7 and G-8. All model results plotted
assumed 40% vegetative cover, the anticipated rooting depth/distribution, and the
anticipated climate record because these conditions would lead to reduced moisture
contents compared to the other scenarios (i.e., 30% cover, reduced rooting
depth/distribution, increased precipitation) which would lead to increased moisture
contents. Within the rooting zone at 25-cm depth (see Figure G-5) and 50-cm depth (see
Figure G-6) the average long-term moisture content at these two depths for the scenario
that assumed the default wilting point was approximately 15%, while for the scenario that
assumed a decreased wilting point was approximately 13%. The average long-term
moisture content, at 100-cm depth, for the scenario that assumed the default wilting point
was approximately 12%, while for the scenario that assumed a decreased wilting point
was approximately 10%. Below the rooting zone at 150-cm and 284-cm depths (see
Figure G-7), there was no significant difference in moisture contents between the two
scenarios, and the average long-term volumetric moisture content for both scenarios was
approximately 15%. Overall, the amount of moisture in the cover predicted by the
infiltration model exceeds the amount of moisture used in the radon attenuation model.
CONCLUSIONS
The results of the sensitivity analysis demonstrate that the design and construction of a
monolithic ET cover will be sufficient to minimize infiltration into the tailings and
prevent the formation of a bathtub effect for a broad range of conditions used to represent
the establishment of vegetation, root water uptake by vegetation, and amount of
precipitation that may occur at the site, thereby meeting closed cell performance
requirements specified in the Ground Water Discharge Permit (Part I.D.8.a and Part
I.D.8.b). The results of the sensitivity analysis, for the broad range of conditions
mentioned above, also demonstrate that the monolithic ET cover will have sufficient

G-10
long-term moisture to attenuate radon fluxes thereby achieving the State of Utah’s long-
term radon emanation standard for uranium mill tailings (Utah Administrative Code, Rule
313-24). Overall, all of the simulations demonstrate that the amount of moisture
predicted with the infiltration model exceeds the amount of moisture used in the radon
attenuation model, which indicates that the predictions of radon emanation at the surface
are conservative.
The results of the sensitivity analysis also demonstrate that establishment of vegetation is
important in reducing infiltration, but that a greater emphasis should be placed on
establishing a mixture of species with a broad range in rooting depths and root density
distributions, rather than the percent vegetative cover to maximize root water uptake.
The establishment of a diverse plant community is supported by the proposed species mix
(see Appendix D).
REFERENCES
Brown, R.W. (1995), The Water Relations of Range Plants: Adaptations to Water Deficits, In: Bedunah, D.J. and R.E. Sosebee (eds.), Wildland Plants: Physiological
Ecology and Developmental Morphology, Society for Range Management, Denver,
Colorado, pp. 291-413.

Burns, J. A., D. L. Flath, and T. W. Clark (1989), On the situation and function of white-tailed prairie dogs burrows, Great Basin Naturalist, 49, 517-524.

Chabot, B. F. and H. A Mooney (1985), Physiological Ecology of North American Plant
Communities. Chapman and Hall. New York, NY. Cheatheam, L. K. (1977), Density and distribution of the black-tailed prairie dog in
Texas, Texas Journal of Science, 29, 33-40.

Denison Mines (USA) Corp. (2009), Reclamation Plan, White Mesa Mill, Blanding,
Utah, Radioactive Materials License No. UT1900479, Revision 4.0, November 2009.
Khire, M.V., C.H. Benson, and P.J. Bosscher (2000), Capillary barriers: design variables
and water balance, Journal of Geotechnical and Geoenvironmental Engineering, 126,
8, 965-708.

G-11

TABLE G‐1. AVERAGE WATER FLUX ENTERING THE TAILINGS CELLS DURING A 57‐YEAR PERIOD USING 
DIFFERENT ASSUMPTIONS REGARDING THE ESTABLISHMENT OF VEGETATION AND ROOT WATER 
UPTAKE.   

Anticipated Climate Record  
& 
40% vegetative cover 
Anticipated Climate Record 
& 
30% vegetative cover 
Anticipated Rooting Depth 
and Root Density Distribution 
& 
Default Wilting Point  
1.2 x10‐4 cm/d  
0.45 mm/yr1 
1.5 x10‐4  cm/d  
0.53 mm/yr 
Reduced Performance Rooting Depth 
and Root Density Distribution 
& 
Default Wilting Point 
6.2 x10‐4 cm/d  
2.3 mm/yr 
6.7 x10‐4  cm/d  
2.4 mm/yr 
Anticipated Rooting Depth 
 and Root Density Distribution 
& 
Decreased Wilting Point 
5.3 x10‐5 cm/d  
0.19 mm/yr 
6.4 x10‐5 cm/d  
0.23 mm/yr 
Notes: 
1. The average water flux for the anticipated case (base case) flux for comparison in the sensitivity analysis is 
1.2 x10‐4 centimeters per day (cm/d) or 0.45 millimeters per year (mm/yr). 

G-12

TABLE G‐2. AVERAGE WATER FLUX ENTERING THE TAILINGS CELLS DURING A 57‐YEAR PERIOD USING 
DIFFERENT ASSUMPTIONS REGARDING ROOT WATER UPTAKE AND AMOUNT OF PRECIPITATION.   

Anticipated Climate Record  
& 
40% vegetative cover 
Increased Precipitation 
Climate Record 
& 
40% vegetative cover 
Anticipated Rooting Depth 
and Root Density Distribution 
& 
Default Wilting Point  
1.2 x10‐4 cm/d1  
0.45 mm/yr 
5.5 x10‐4 cm/d  
2.0 mm/yr 
Anticipated Rooting Depth 
and Root Density Distribution 
& 
Decreased Wilting Point 
5.3 x10‐5 cm/d  
0.19 mm/yr 
3.4 x10‐4 cm/d  
1.2 mm/yr 
Notes: 
1. The average water flux for the anticipated case (base case) flux for comparison in the sensitivity analysis is 
1.2 x10‐4 centimeters per day (cm/d) or 0.45 millimeters per year (mm/yr).

G-13
Figure G-1. Semilog plot of the anticipated and reduced root density distributions and maximum rooting depths.

G-14
Figure G-2. Annual amount of precipitation incorporated into the 57-year anticipated and increased precipitation climate records.

G-15
Figure G-3. Semilog plot of model-predicted rate of water flux through the tailings cell cover during the anticipated 57-year climate record for simulations that assume 30% and
40% vegetative cover combined with the anticipated scenario (AS) and reduced
performance scenario (RPS) for establishment of vegetation (rooting depth/distribution).

G-16
Figure G-4. Semilog plot of model-predicted rate of water flux through the tailings cell cover during 57-year climate record with increased precipitation for simulations that
assume 40% vegetative cover combined with the anticipated scenario (AS) and reduced
performance scenario (RPS) for establishment of vegetation (rooting depth/distribution)
and decreased wilting point.

G-17
Figure G-5. Moisture contents at 25-cm depth within the rooting zone of the monolithic ET cover during the anticipated 57-year climate record for the two scenarios that use a
decreased wilting point of -30,000 cm and a default wilting point of -8,000 cm.

G-18
Figure G-6. Moisture contents at 50-cm (upper figure) and 100-cm (lower figure) depth within the rooting zone of the monolithic ET cover during the anticipated 57-year
climate record for the two scenarios that use a decreased wilting point of -30,000 cm and
a default wilting point of -8,000 cm.

G-19
Figure G-7. Moisture contents at 150-cm and 284-cm depth below the rooting zone of the monolithic ET cover during the anticipated 57-year climate record for the two
scenarios that use a decreased wilting point of -30,000 cm and a default wilting point of -
8,000 cm.

APPENDIX H
RADON EMANATION MODELING FOR THE
EVAPOTRANSPIRATION COVER

H-1
APPENDIX H
RADON EMANATION MODELING FOR THE EVAPOTRANSPIRATION
COVER
This appendix presents the results of modeling the emanation of radon-222 from the top
surface of the proposed monolithic evapotranspiration (ET) cover. The material
thicknesses for the different cover layers were based on the results of radon attenuation
modeling to achieve the State of Utah’s long-term radon emanation standard for uranium
mill tailings (Utah Administrative Code, Rule 313-24). Radon modeling completed for
this appendix supersedes previous radon attenuation modeling (TITAN Environmental,
1996; IUC, 2000) because of the proposed changes to the tailings cover system from a
conventional rock cover design to an ET cover design.
CONCEPTUAL COVER DESIGN
Radon emanation modeling was evaluated for the proposed monolithic ET cover design.
The material and layer thicknesses presented below were based on the radon modeling
results discussed in this appendix. The thickness of the upper platform fill layer (random
fill compacted to 95 percent of Standard Proctor dry density) was optimized in order to
minimize radon fluxes at the surface.
The 2.84-m (9.3-ft) thick cover would consist from top to bottom of:
• 15 cm (0.5 ft) of a gravel-amended topsoil admixture to promote revegetation and
provide for protection against erosion and frost damage
• 107 cm (3.5 ft) of random fill soil placed at 85% of standard Proctor dry density
to serve as a water storage, biointrusion, and radon attenuation layer
• 162 cm (5.3 ft) of random fill soil comprised of 2.8 feet random fill compacted to
95% of standard Proctor dry density over 2.5 feet of random fill placed at 80% of

H-2
standard Proctor dry density, to serve as grading (platform fill) and radon
attenuation layers.
RADON MODEL CONFIGURATION
The thickness of the reclamation cover necessary to limit radon emanation from the
disposal areas was analyzed using the U.S. Nuclear Regulatory Commission (NRC)
RADON model (NRC, 1989). The model utilizes the one-dimensional radon diffusion
equation, which uses the physical and radiological characteristics of the tailings and
overlying materials to calculate the rate of radon emanation from the tailings through the
cover. The model was used to calculate the cover thickness required to limit the radon
emanation rate through the top of the cover to 20 picocuries per square meter per second
(pCi/m2/s), following the guidance presented in NRC publications NUREG/CR-3533
(NRC, 1984) and Regulatory Guide 3.64 (NRC, 1989). This maximum rate of emanation
is an average over the entire surface of the disposal area.
Input Values
Stockpiles of soil will be used to construct the monolithic ET cover. Geotechnical
properties and input data of the soil (also referred to as random/platform fill) and tailings
were based on available site-specific data summarized from previously submitted reports.
Thickness of Tailings. The tailings thickness currently deposited in Cells 2 & 3 is
approximately 30 ft (914 cm), while the anticipated tailings thickness deposited in Cells
4A & 4B will be approximately 42 ft (1,280 cm). However, as documented in NRC
Regulatory Guide 3.64, a tailings thickness greater than 100 to 200 cm is effectively
equivalent to an infinitely thick radon source. Therefore, a thickness of 500 cm may be
used in RADON to represent an equivalent infinitely thick tailings source of radon.
Radium Activity Concentration. The radium-226 activity concentration value for the
tailings in the impoundments was based on laboratory data reported by Rogers &
Associates (1988); their original laboratory report is reproduced here as part of

H-3
Attachment H-1. The radium activity of the random fill (used to construct the frost
barrier, water storage, and platform fill/grading layers for the proposed cover) was
assumed to be zero. The assumption of zero radium activity for the cover soil was based
on guidance in NRC Regulatory Guide 3.64, which states that radium activity in the
cover soils may be neglected for cover design purposes provided the cover soils are
obtained from background materials that are not associated with ore formations or other
radium-enriched materials. The radium activity values used in the model are summarized
in Table H-1.
Radon Emanation Coefficient. The emanation coefficients were based on laboratory
data (see Attachment H-1 for Rogers & Associates 1988). Because site-specific
laboratory data were available, the NRC’s default value of 0.35 is not appropriate. The
radon emanation coefficient values used in the model are summarized in Table H-2.
Specific Gravity, Density, and Porosity. Specific gravity of the tailings was estimated
to be 2.75, corresponding to a solid particle density of 172 pounds per cubic foot (pcf)
(2.75 grams/cm3), while the dry bulk density of the tailings was estimated to be 74.3 pcf
(1.19 grams/cm3), which is 70 percent of standard Proctor dry density. Specific gravity
and maximum dry bulk density were average values determined by laboratory tests (see
Attachment E-1 in Appendix E for Chen and Associates, 1987; see Attachment H-1 in
this Appendix for Western Colorado Testing, 1999b). Porosity was calculated based on
the average specific gravity and dry bulk density based on the following equation:
݊ൌ1െ൬ߩ௕
ߩ௦
൰
where

n = porosity (-)
ρb = dry bulk density of soil or tailings (pcf or grams/cm3)
ρs = density of solids of soil or tailings (pcf or grams/cm3)

H-4
A minimum of 3 feet of random fill has already been placed above Cell 2 on top of the
tailings. It is assumed that the random fill was placed and compacted to 80 percent
standard Proctor compaction by construction traffic. The upper 0.5 feet of this fill will be
compacted by additional passes of compactors to reach 95 percent of standard Proctor
compaction. Subsequent layers of random fill placed above the 0.5-foot zone will also be
compacted to 95 percent until the necessary thickness of this layer is achieved. Porosity
was calculated based on average specific gravity and dry bulk density values determined
by laboratory tests (see Attachment E-1 for Chen and Associates, 1978, 1979, 1987;
Western Colorado Testing, 1999a; Geosyntec, 2006). The uppermost 3.5 feet of random
fill (used to construct the water storage layer) will be placed at 85 percent of standard
Proctor in order to optimize water storage and rooting characteristics for plant growth.
The top 0.5-feet of erosion protection was assumed to consist of a soil-gravel admixture
(e.g., rock mulch) placed by mixing approximately 25 percent gravel into the top layer of
stockpiled topsoil and random fill. Dry bulk density and porosity values for the soil and
tailings materials used as input to the RADON model are summarized in the Table H-3.
Long-term Moisture Content. Long-term moisture content value was assumed to be 6
percent for the tailings. This is a conservative assumption, per NRC Regulatory Guide
3.64, which represents the lower bound for moisture in western soils.
Long-term moisture content for the soil cover corresponding to approximately 15
atmospheres of soil water tension, was estimated using the Rawls and Brakenseik (1982)
equation as presented in NRC Regulatory Guide 3.64 as follows:
ߠ ൌ 0.026 ൅ 0.005ݖ ൅ 0.0158ݕ
where

θ = volumetric water content (-)
z = percent clay in cover soil (%)
y = percent organic matter in cover soil (%).

H-5
Volumetric water content is related to gravimetric water content, w, by the following
equation:
ݓൌ ߠ·ߩ௪
ߩ௕

where

w = gravimetric water content (-)
θ = volumetric water content (-)
ρw = density of water (pcf or grams/cm3)
ρb = dry bulk density of soil or tailings (pcf or grams/cm3).
The percent clay in the random fill was taken from average values measured in laboratory
tests (see Attachment E-1 for Chen and Associates, 1978, 1979, 1987; Western Colorado
Testing, 1999a; Geosyntec, 2006). For sieve analyses in which hydrometer tests were not
conducted, the percent clay (particles finer than 2 μm) was assumed to be 35% of the
portion finer than the No. 200 sieve (0.74 mm), which corresponded to the average clay
fraction from tests in which hydrometer analyses were performed. Organic matter was
considered negligible and assumed to equal zero. Long-term moisture parameters are
summarized in Table H-4.
Diffusion Coefficient. The radon diffusion coefficients used in the RADON model can
either be calculated within the model (based on an empirical relationship dependent upon
porosity and the degree of saturation) or input directly in the model using values
measured from laboratory testing. Although laboratory test data were available for the
tailings and the cover material (see Attachment H-1 for Rogers & Associates 1988), tests
were performed at porosities and water contents different than those estimated to
represent long-term conditions. Therefore, the empirical relationship in RADON was
used to estimate values for use in the model; input values are summarized in Table H-5.

H-6
PROTECTION FROM BIOINTRUSION AND FROST PENETRATION
Biointrusion
Potential impacts to the radon flux at the surface due to biointrusion can be estimated
using weighted average principals by incorporating the bare source flux from the
uncovered tailings into the radon flux through the cover. The bare source flux, or the flux
from the tailings without a cover, from the RADON modeling was approximately 689
pCi/m2/s. The flux from the top of the cover was 20 pCi/m2/s. Root holes (from dead
shrubs) were estimated to have a diameter of 10 mm (0.4 in) and a frequency of one hole
every 10 feet, or 400 root holes per acre of cover. Animal burrow holes were estimated
to have a diameter of 11 cm (4.3 in) and a frequency of 12 holes per acre of cover. Based
on the hole diameters and frequencies, and conservatively assuming that the holes extend
through the cover to the tailings, the cross-sectional area of open hole area to allow radon
emanation at the bare source flux rate was 0.15 m2 (1.57 ft2) per acre. The resulting flux
would be 20.02 pCi/m2/s, or an insignificant increase in average radon flux of 0.1% over
the cover surface. Therefore, it is reasonable to conclude that the radon flux at the
surface would not be affected by root holes and animal burrows.
Frost Penetration
Previously, TITAN Environmental (1996) completed a freeze/thaw evaluation based on
site-specific conditions which indicated that the anticipated maximum depth of frost
penetration was 6.8 inches (0.6 ft). Therefore, the entire soil-gravel admixture layer and
upper few centimeters of the underlying soil layer will provide adequate protection
against frost penetration.
RADON MODEL RESULTS
The material thicknesses supported by the radon modeling, and the long-term model-
predicted radon fluxes at the surface for the monolithic ET cover, are summarized in
Table H-6. The output file for the RADON model is included as part of Attachment H-2.

H-7
CONCLUSIONS
Radon emanation modeling was evaluated for the monolithic ET cover constructed
entirely of sandy clay soil. The results of modeling with RADON indicate this design
attenuates radon to 20 pCi/m2/s or less, which is the stated regulatory criterion (Utah
Administrative Code, Rule 313-24). The actual radon emanation rate is likely to be lower
because the actual moisture content in the cover is likely to be greater than the values
used in the radon emanation modeling.
REFERENCES
Chen and Associates, Inc. (1978), Earth Lined Tailings Cells, White Mesa Uranium
Project, Blanding, Utah, Report prepared for Energy Fuels Nuclear, Inc. on 18 July
1978.

Chen and Associates, Inc. (1979), Soil Property Study, Proposed Tailings Retention
Cells, White Mesa Uranium Project, Blanding, Utah, Report prepared for Energy
Fuels Nuclear, Inc. on 23 January 1979.

Chen and Associates, Inc. (1987), Physical Soil Data, White Mesa Project, Blanding
Utah, Report prepared for Energy Fuels Nuclear, Inc. 1987.

Geosyntec Consultants (2006), Stockpile Evaluation Tailings Cell 4A, White Mesa Mill -
Technical Memo prepared for International Uranium (USA) Corporation on 23
January 2006.

International Uranium (USA) Corporation (2000), Reclamation Plan, White Mesa Mill,
Blanding, Utah, Source Material License No. SUA-1358, Docket No. 40-8681,
Revision 3.0, July 2000.

Rawls, W.J., and Brakensiek, D.L. (1982), Estimating Soil Water Retention from Soil
Properties, Journal of the Irrigation and Drainage Division, American Society of
Civil Engineers, 108, IR2, 166-171.

Rogers & Associates Engineering Corporation (1988), Two separte letters prepared by
Renee Y. Bowser for C.O. Sealy of Umetco Minerals Corporation on 4 March 1988
and 9 May 1988.

TITAN Environmental Corporation (1996), Tailings Cover Design, White Mesa Mill,
Blanding Utah, Report prepared for Energy Fuels Nuclear, Inc. September 1996.

H-8
U.S. Nuclear Regulatory Commission (1984), Radon Attenuation Handbook for Uranium
Mill Tailings Cover Design, NUREG/CR-3533.

U.S. Nuclear Regulatory Commission NRC (1989), Calculation of Radon Flux
Attenuation by Earthen Uranium Mill Tailings Covers, Regulatory Guide 3.64.
Western Colorado Testing, Inc. (1999a), Additional Clarifications to the White Mesa Mill
Reclamation Plan, Report prepared for International Uranium (USA) Corporation and
submitted to the U.S. Nuclear Regulatory Commission 15 April 1999.
Western Colorado Testing, Inc. (1999b), Report of soil sample testing of tailings
collected from Cell 2 and Cell 3, Prepared for International Uranium (USA)
Corporation 4 May 1999.

H-9
Table H-1. Radium activity concentrations used as input to the RADON model.

Material Radium Activity Concentration (pCi/g)
Tailings 981
Random Fill 0
Erosion Protection 0
Note: Units are in picocuries per gram (pCi/g)

Table H-2. Radon emanation coefficients used as input to the RADON model.

Material Radon Emanation Coefficient (-)
Tailings 0.19
Random Fill 0.19
Erosion Protection 0.19

Table H-3. Density and porosity characteristics of soil layers and tailings used as input
to the RADON model.

Material Specific
Gravity
(-)
Maximum
Dry Bulk
Density
(pcf)
Degree of
Compaction
(%)
Placed
Dry Bulk
Density
(pcf)
Porosity
(-)
Tailings 2.75 106.1 70% SP 74.3 0.57
Random Fill (low
compaction used to
construct lower grading
layer, already in place for
Cell 2)
2.67 116.7 80% SP 93.4 0.44
Random Fill (high
compaction used to
construct upper grading
layer)
2.67 116.7 95% SP 110.9 0.33
Random Fill (low
compaction used to
construct water storage
layer)
2.67 116.7 85% SP 99.2 0.40
Erosion Protection 2.67 --- --- 124.2* 0.25
SP = standard Proctor compaction
MP = modified Proctor compaction
pcf = pounds per cubic foot
* Estimated by applying 25% gravel correction factor

H-10
Table H-4. Long term moisture characteristics used as input to the RADON model.

Material Percent
Clay
(%)
Volumetric
Water
Content
(%)
Placed
Dry Bulk
Density
(pcf)
Gravimetric
Water
Content
(%)
Tailings --- --- 74.3 6.0
Random Fill (low
compaction used to
construct lower grading
layer, already in place for
Cell 2)
20 12.6 93.4 8.4
Random Fill (high
compaction used to
construct upper grading
layer)
20 12.6 110.9 7.1
Random Fill (low
compaction used to
construct water storage
layer)
20 12.6 99.2 7.9
Erosion Protection 16 10.6 124.2 5.3

Table H-5. Radon diffusion coefficients calculated using empirical relationships.

Material Saturation
(%)
Diffusion
Coefficient (cm2/s)
Tailings 12.5 0.0499
Random Fill (low compaction
used to construct lower grading
layer, already in place for Cell 2)
28.7 0.0275
Random Fill (high compaction
used to construct upper grading
layer)
37.8 0.0177
Random Fill (low compaction
used to construct water storage
layer)
31.1 0.0244
Erosion Protection 41.5 0.0141

H-11
Table H-6. Summary of RADON model results including material thicknesses (in feet)
and radon emanation at the cover surface for the proposed monolithic ET cover design.

Material Layers Thickness & Radon Exit
Flux
Erosion Protection 0.5
Random Fill (low compaction used to construct water
storage layer)
3.5
Random Fill* (high compaction used to construct
upper grading layer)
2.8
Random Fill (low compaction used to construct lower
grading layer, already in place for Cell 2)
2.5
Total Cover Thickness 9.3
Radon Exit Flux (pCi/m2/s) 20.0
*The thickness of the upper platform fill layer (random fill compacted to 95% of Standard Proctor dry
density) was optimized in order to minimize radon fluxes.

APPENDIX I
TAILINGS HYDRAULIC CONDUCTIVITY EVALUATION

I-1
APPENDIX I
TAILINGS HYDRAULIC CONDUCTIVITY EVALUATION
The purpose of this appendix is to compare grain size distribution data of uranium
tailings at the White Mesa Mill to uranium tailings at the Canon City Mill to determine
whether measurements of hydraulic conductivity at Canon City can be considered as a
representative surrogate for White Mesa. The saturated hydraulic conductivity of the
tailings assumed for White Mesa was based on measured values reported for the Cotter
Corporation’s Canon City Mill tailings impoundment.
GRAIN SIZE ANALYSES
The White Mesa tailings are generally silty sand but heterogeneous due to the placement
process. Based on grain-size analyses performed on the tailings, sand-sized particles are
dominant with the remainder being silt- and clay-sized particles. The average grain size
distribution for White Mesa, based on 13 samples, consists of 57% sand, 26% silt, and
7% clay (see Table I-1). For sieve analyses in which hydrometer tests were not
conducted, the percent clay (particles finer than 2 μm) was assumed to be 17% of the
portion finer than the No. 200 sieve (0.74 mm), which corresponded to the average clay
fraction from tests in which hydrometer analyses were performed.
The Canon City Mill tailings impoundment is operated by Cotter Corporation and is
located near Canon City, Colorado. The tailings are generally silty sand but
heterogeneous due to the placement process. Based on grain-size analyses performed on
the tailings, sand-sized particles are dominant with the remainder being silt- and clay-
sized particles. The average grain size distribution, based on five samples, consists of
61% sand, 33% silt, and 6% clay (see Table I-2) (MFG Inc., 2005). During hydrometer
testing, two of the seven samples experienced flocculation and were not included in the
averaging.

I-2
The mill tailings at Canon City are considered to be representative of the mill tailings at
White Mesa because the average grain size distribution and approximate D5 values
between the two sites are similar:
• Canon City: 61% sand, 33% silt, and 6% clay with a D5 of ~ 2 μm
• White Mesa: 57% sand, 26% silt, and 7% clay with a D5 of ~ 2 μm.
Gradation curves are not available for the Canon City tailings; however, the similarity in
average grain size distribution (% sand, silt, clay) and D5 values between the two sites is
expected to produce tailings materials that would behave in a similar hydrogeologic
manner.
SATURATED HYDRAULIC CONDUCTIVITY TESTING
The saturated hydraulic conductivity of the White Mesa tailings were estimated based on
the results from a multiple well aquifer test completed at the Canon City impoundment
(MFG Inc., 2005) because site-specific measurements were lacking for White Mesa. The
average hydraulic conductivity of the tailings at Canon City ranged from 2.1 ft/day
(7.4 x 10-4 cm/sec) to 8.5 ft/day (3.0 x 10-3 cm/sec) with an average value of 4.8 ft/day
(1.7 x 10-3 cm/sec) (MFG Inc., 2005). A hydraulic conductivity of 4.8 ft/day was
assumed for the White Mesa mill tailings cell dewatering model.
CONCLUSIONS
The saturated hydraulic conductivity of the tailings assumed for White Mesa was based
on measured values reported for the Cotter Corporation’s Canon City Mill tailings
impoundment. The mill tailings at Canon City are considered to be representative of the
mill tailings at White Mesa because the average grain-size distributions between the two
sites are similar.

I-3
REFERENCES
Chen and Associates, Inc., 1987. Physical Soil Data, White Mesa Project, Blanding Utah,
Report prepared for Energy Fuels Nuclear, Inc. 1987.

Colorado School of Mines Research Institute (CSMRI), 1978. Grinding Study, Appendix A.

MFG Inc., 2005. Update of the Mill Decommissioning and Tailings Reclamation Plan
for the Cotter Corporation Canon City Milling Facility. Prepared for Cotter
Corporation. August 2005.
Western Colorado Testing, Inc. (WCT), 1999. Report of soil sample testing of tailings
collected from Cell 2 and Cell 3, Prepared for International Uranium (USA)
Corporation 4 May 1999.

I-4

TABLE I‐1.  GRAIN SIZE DISTRIBUTION OF WHITE MESA MILL TAILINGS.   
Reference Sand 
(%) 
Silt 
(%) 
Clay 
(%) 
CSMRI, 1978 56 37 8 
CSMRI, 1978 53 39 8 
CSMRI, 1978 57 36 7 
CSMRI, 1978 57 35 7 
CSMRI, 1978 57 35 7 
CSMRI, 1978 58 35 7 
Chen & Associates, 1987 54 38 8 
WCT, 1999 76 19 5 
WCT, 1999 17 70 13 
WCT, 1999 67 23 10 
WCT, 1999 68 29 4 
WCT, 1999 39 60 1 
WCT, 1999 77 17 6 
AVERAGE 57 36 7 
Notes: 
1. CSMRI = Colorado School of Mines Research Institute.  Data taken from samples labeled as DSM Screen 
Undersize. 
2. WCT = Western Colorado Testing. 
TABLE I‐2.  GRAIN SIZE DISTRIBUTION OF CANON CITY MILL TAILINGS.   
Reference Sand 
(%) 
Silt 
(%) 
Clay 
(%) 
MFG Inc., 2005 1 ‐Floc 
MFG Inc., 2005 86 10 4 
MFG Inc., 2005 2 ‐Floc 
MFG Inc., 2005 80 15 5 
MFG Inc., 2005 88 7 5 
MFG Inc., 2005 33 59 8 
MFG Inc., 2005 17 73 10 
AVERAGE 61 33 6 
Notes: 
1. Floc = Flocculation occurred during hydrometer testing.  Samples were not included in the averaging. 
2. Clay cutoff varied between 3 to 6 micrometers because hydrometer tests were terminated before 2 
micrometer settling time had been reached.

APPENDIX J
TAILINGS CELL DEWATERING MODELING

J-1
APPENDIX J
TAILINGS CELL DEWATERING MODELING
This appendix describes the dewatering modeling performed with MODFLOW to
estimate the time required to dewater the tailings in Cells 2 & 3 and estimate the residual
saturated thickness of tailings. The model-predicted water levels (saturated thickness of
tailings) are used in the Giroud-Bonaparte Equation to calculate potential flux rates
through the liner into the underlying bedrock vadose zone, as described in Appendix L.
A tailings cell dewatering model was not constructed for Cells 4A & 4B because
analytical solutions presented by Geosyntec Consultants (2007) were deemed adequate
given the uniform distribution of the drain system in those cells.
Tailings Cells 2 & 3 Slimes Drains
To dewater the tailings in Cells 2 & 3, slimes drain networks consisting of perforated
PVC pipe are located across the base of the cells which drain to an extraction sump on
the southern side of each cell. The drains cover an approximately 400-foot by 600-foot
area in the southern part of the cells. The design for the slimes drains is the same for both
cells (D’Appolonia Consulting Engineers, 1982). The drain pipes are situated in nine
alignments spaced 50 feet apart running in an approximately east-west direction. Each
drain is 600 feet long, extending 300 feet in each direction from the central collection
pipe that drains to the sump. The drain pipes are covered by an envelope of sand over the
drains, rather than a continuous layer across the bottom of the tailing cells (“burrito
drains”). Water gravity drains to the sump, whence it is pumped to Cell 1.
METHODOLOGY
Model Code
The computer code MODFLOW was used in this modeling effort with the Department of
Defense Groundwater Modeling System (GMS) pre- and post-processor. MODFLOW is

J-2
a modular three-dimensional finite-difference flow model developed by the United States
Geological Survey (McDonald and Harbaugh, 1988; Harbaugh et al., 2000) to calculate
hydraulic-head distribution and determine flow within a simulated aquifer. This model
was selected because it can adequately represent and simulate the hydrogeologic
conditions necessary and it is well-documented, frequently used, and a versatile program
that is widely accepted by the scientific and regulatory communities (Anderson and
Woessner, 1992).
Model Domain, Layering, and Grid
The domain for the tailings cell model was approximately 3,500 by 1,200 feet,
representing Cells 2 & 3 (see Figure J-1). The finite-difference grid consisted of a
constant spacing of 10 feet. The model included two layers to represent the tailings and
slimes drains. The bottom layer was 1 foot thick so that the drains could be simulated
explicitly (hydraulic conductivity was variable to represent tailings between the drains).
The top layer had a variable thickness that represented the tailings. The water level in the
top layer was allowed to vary spatially and temporally. The bottom elevations were set
based on information presented in the tailings cell construction report (D’Appolonia
Consulting Engineers, 1982).
Boundary Conditions
Boundary conditions define hydraulic constraints at the boundaries of the model domain.
There are three general types of boundary conditions:
1. Specified head or Dirichlet (e.g., constant head)
2. Specified flux or Neumann (e.g., constant flow, areal recharge, extraction
wells, no flow)
3. Head-dependent flux or Cauchy (e.g., drains, evapotranspiration)

J-3
No-flow boundaries are a special case of the specified flux boundary in which the flow is
set to zero.
For the tailings cell model, no-flow boundaries were assumed to surround the domain. A
net flux rate from the cell was assumed across the entire domain. This assumed flux rate
represents the combination of potential fluxes from the cell through the liner and
potential infiltration into the cell through the cover. The net flux rate was calculated
using the average infiltration rate through the cover predicted by the HYDRUS-1D
tailings cover model and the potential flux rate through the bottom of
Cells 2 & 3 (see Appendix L). The resulting average net flux rate for Cells 2 & 3 was
6.9 x 10-4 cm/day (2.27 x 10-5 ft/day). This assumed net flux rate was applied uniformly
across the domain and was simulated with MODFLOW as a negative recharge rate.
The slimes drains were simulated with the Drain package in MODFLOW. Drains are
head-dependent boundary conditions in which flow varies based on the difference in
hydraulic head in the aquifer and the drain: as head in the aquifer declines (tailings in this
case), so does the dewatering rate. Groundwater flow to this array is gravity driven and
dependent on the head difference between the surrounding material and the perforated
pipe. Operation of the slimes drain extraction pump is only necessary to extract the
groundwater driven into this array to maintain a head difference. Essentially, this system
acts as a field drain array. The MODFLOW Drain package was developed specifically to
simulate this sort of gravity driven, head dependent drain system. A thorough
quantitative explanation of the MODFLOW Drain package is presented in A Modular
Three-Dimensional Finite-Difference Ground-Water Flow Model: U.S. Geological
Survey Techniques of Water-Resources Investigations, book 6, chap. A1 (McDonald and
Harbaugh, 1988).
Drain cells were set along nine alignments spaced 50 feet apart. Each drain was 600 feet
long. Drains were set in the model as shown on Figure J-1.

J-4
Hydraulic Properties
The saturated hydraulic conductivity of the tailings assumed for White Mesa was based
on measured values reported for the aquifer testing performed in uranium mill tailings at
Cotter Corporation’s Canon City Mill tailings impoundment (MFG, Inc., 2005). See
Appendix I for details concerning the comparison of tailings grain size for the White
Mesa Mill to those of the Canon City Mill. The average hydraulic conductivity of the
tailings ranged from 2.1 ft/day (7.4 x 10-4 cm/sec) to 8.5 ft/day (3.0 x 10-3 cm/sec) with
an average value of 4.8 ft/day (1.7 x 10-3 cm/sec). A hydraulic conductivity of 4.8 ft/day
was assumed for the tailings (in both model layers). A hydraulic conductivity of
25 ft/day was assumed for the sand adjacent to the slimes drain in the bottom layer of the
model. This was used only in layer 1 in the cells that represent drains. Hydraulic
conductivity values representative of tailings were assumed across the remainder of the
bottom layer.
Calibration
The calibration process involves iterating values for model parameters in sequential
model simulations to produce estimated values that better match field-measured data.
The initial-parameter values were adjusted through calibration until the model produced
results that adequately simulated the known data. The tailings cell model was calibrated
by varying the drain conductance term until the flow rates approximately matched the
2007 dewatering rates (average rate of 12.5 gpm) and average water levels of 20 feet
above the liner.
RESULTS
The MODFLOW dewatering model predicts that the tailings would draindown
nonlinearly through time reaching an average saturated thickness of 3.5 feet (1.07 m)
after 10 years of dewatering (see Figure J-2). The model also predicts that dewatering
rates would decline to approximately 2 gallons per minute (gpm) after 10 years of
pumping. This reduction in pumping rates is caused by the reduction in saturated

J-5
thickness of tailings. Dewatering rates are also controlled by the saturated hydraulic
conductivity of the tailings. If the actual hydraulic conductivity of the tailings is higher
than the value assumed in the model, dewatering rates could be higher and water levels
could be lowered more rapidly. Conversely, if the actual hydraulic conductivity of the
tailings is lower than the value assumed in the model, dewatering rates could be lower
and water levels could require more time to dewater. Mass balance errors for the
MODFLOW model were less than 1%.
A dewatering model was not constructed for Cells 4A & 4B because dewatering rates
were estimated by Geosyntec Consultants (2007). Water levels in Cell 4A were
estimated to decline to less than 1 foot after approximately six years of dewatering. Cells
4A & 4B is estimated to be dewatered significantly faster than Cells 2 & 3 due to the
more extensive slimes drain network. The dewatering system in Cell 4B is assumed to be
designed similarly to Cell 4A, thus dewatering rates were assumed to be similar.
REFERENCES
Anderson, M.P., and W.W. Woessner, 1992. Applied Groundwater Modeling:
Simulation of Flow and Advective Transport. Academic Press, Inc. Harcourt
Brace Jovanovich, Publishers, San Diego, CA. 381p.
D’Appolonia Consulting Engineers, Inc., 1982. Construction Report, Initial Phase – Tailings Management System, White Mesa Uranium Project, Blanding, Utah.
Geosyntec Consultants, 2007. Analysis of Slimes Drains for White Mesa Mill - Cell 4A,
Computations submitted to Denison Mines, 12 May 2007.
Harbaugh, A.W., Banta, E.R., Hill, M.C., and McDonald, M.G., 2000, MODFLOW-
2000, the U.S. Geological Survey modular ground-water model -- User guide to modularization concepts and the Ground-Water Flow Process: U.S. Geological
Survey Open-File Report 00-92, 121 p.
McDonald, M.G., and Harbaugh, A.W., 1988, A Modular Three-Dimensional Finite-
Difference Ground-Water Flow Model: U.S. Geological Survey Techniques of
Water-Resources Investigations, book 6, chap. A1, 586 p.

J-6
MFG, Inc., 2005. Update of the Mill Decommissioning and Tailings Reclamation Plan for the Cotter Corporation Canon City Milling Facility (Appendix A 1999
Tailings Investigation). Prepared for Cotter Corporation. August 2005.

J-7

Figure J-1. MODFLOW tailings cell model domain, grid, and boundary conditions

J-8
Figure J-2. Model-predicted average saturated thickness of tailings in Cells 2 & 3 with dewatering pumping.

APPENDIX K
STATISTICAL EVALUATION OF TAILINGS PORE WATER
CHEMISTRY AND IDENTIFICATION OF SOURCE TERM
CONCENTRATIONS

K-1
APPENDIX K
STATISTICAL EVALUATION OF TAILINGS PORE WATER CHEMISTRY
AND IDENTIFICATION OF SOURCE TERM CONCENTRATIONS
The purpose of this appendix is to present the source term solution chemistry for the
tailings pore water. The source term chemistry is used as input to the predictive vadose
zone geochemical and solute transport models. Descriptive statistics of the source term
chemistry are also presented.
BACKGROUND
Tailings pore water in the slimes drains (i.e., immediately above the tailing cell liners) is
considered to be more representative of solutions that would remain in the tailings cells
during operations and at closure given that these solutions would have had sufficient time
to equilibrate with the tailings. Furthermore, water extracted from the slimes drains, as
opposed to samples grabbed from surface ponds, is not affected as much by
evaporation/evapoconcentration and addition/recirculation of mill process water;
evaporation and recirculation of mill process water would tend to create a variable source
term solution chemistry that is dissimilar to and not representative of the pore water in
the tailings.
As described below, the solution chemistry of the tailings pore water, as represented by
samples collected from the Cell 2 slimes drain, was assumed to be identical for all of the
cells, given the similarities in ores and process solutions used over time. Cell 3 slimes
drain data were not included in the statistical analysis because the analytical results
contained an irregularity: the total dissolved solids (TDS) concentration was identical to
the sulfate concentration, which is chemically impossible. Furthermore, only one sample
has been collected from the Cell 3 slimes drain which limits the evaluation of potential
trends. In other words, this one sample appears to be in error or could represent an
outlier, rather than being representative. Cell 4A was not included in the statistical
analysis because the facility was only constructed recently (as of Fall 2008) and contains

K-2
a minimal amount of tailings. Cell 4B was not included because the facility is currently
being permitted (as of Spring 2010).
RESULTS
The analytical results of samples collected from the slimes drain for Cell 2 are tabulated
in Table K-1. Statistical analyses of the data, including number of data points, minimum
and maximum values, median, arithmetic average (mean), arithmetic standard deviation,
geometric mean, and geometric standard deviation are tabulated in Table K-2. Overall,
the data suggest that the solution chemistry of the tailings pore water is fairly consistent
from year to year.
Four constituents consistently had concentrations detected at the reporting limit: fluoride,
mercury, silver, and tin. Therefore, these analytes were excluded from further discussion
and inclusion as part of the source term chemistry. For each analyte, the arithmetic mean
was greater than the geometric mean, while the mean and median were in general
agreement. The mean plus one-half standard deviation, which corresponds to
approximately 38% of the observations if the data are distributed normally, was less than
the maximum. However, the mean plus one standard deviation, which corresponds to
approximately 68% of the observations, generally exceeded the maximum. Therefore
such a metric (mean plus one standard deviation) should not be used as part of a
sensitivity analysis to yield an end member (e.g., higher concentration) source term
chemistry. In addition, the maximum value generally should not be used because this
metric corresponds to a statistically insignificant percentage of the dataset, and its use
would lead to unrealistic predictions; however, in response to a request by the Utah
Division of Radiation Control, the maximum value was selected as the upper bound. To
determine a range of source term chemistries for the modeled solutes, the maximum
value was selected as an upper bound, the arithmetic average (mean) was selected as the
base case, and the mean minus one-half standard deviation was selected as the lower
bound.

K-3
University of Utah Data
In July 2007, the University of Utah collected samples from the Cell 2 slimes drain
(Hurst and Solomon, 2008). Denison sampled the Cell 2 slimes drain in September 2007.
The values reported by the University of Utah for July 2007 samples are compared to the
September 2007 data reported by Denison below:
• Nitrate + Nitrite as N: 5.19 mg/L (University of Utah) and 30.9 mg/L (Denison)
• Manganese: 139 mg/L (University of Utah) and 117 mg/L (Denison)
• Selenium: <4 mg/L (University of Utah) and 0.422 mg/L (Denison)
• Sulfate: 666,000 mg/L (University of Utah) and 60,600 mg/L (Denison)
• Uranium: 23.7 mg/L (University of Utah) and 23 mg/L (Denison).
Nitrate values reported by the University of Utah were lower than values reported by
Denison. Manganese and uranium values reported by the University of Utah agreed
closely with results reported by Denison. The selenium values could not be compared
because the detection limit reported by the University of Utah was far greater than the
selenium concentration reported by Denison. The sulfate concentration reported by the
University of Utah differed significantly from historical sulfate data reported by Denison
and is considered to be anomalous.
The University of Utah data were not included in the statistical analysis because their
dataset was limited to five analytes and a more comprehensive suite of analytes was
available for data collected two months later by Denison. Furthermore, because the
University of Utah analysis did not include all major ions, a charge balance could not be
performed to check the data.
CONCLUSIONS
The arithmetic average (mean) will be used as the base case to set the initial conditions.
As part of the sensitivity analysis, the modeled solutes will assume the maximum value
for an upper bound and the mean minus one-half standard deviation for the lower bound.

K-4
REFERENCES
Hurst, T.G, and D.K. Solomon (2008), Summary of Work Completed, Data Results,
Interpretations, and Recommendations for the July 2007 Sampling Event at the
Denison Mines, USA, White Mesa Uranium Mill, near Blanding, Utah. Report
prepared for the Utah Division of Radiation Control, May 2008, pp. 62.

K-5

TABLE K-1. Analytical results of samples collected from the Cell 2 slimes drain. All units are in mg/L except for pH which is in standard units.
Analyte
Cell 2 Slimes Drain
Collected by UMETCO
June 1991
Cell 2 Slimes Drain
Collected by Denison
September 2007
Cell 2 Slimes Drain
Collected by Denison
October 2008
Cell 2 Slimes Drain
Collected by Denison
August 2009
MAJOR IONS - - - -
Calcium 390 572 528 508
Chloride 2,573 3,700 3,860 2,750
Fluoride 2.0 3.3 0.055 0.055
Magnesium 2,540 4,100 4,030 3,750
Nitrogen, Ammonia as N 2,240 4,020 3,620 3,240
Nitrogen, Nitrate+Nitrite as N 1002 30.9 20.3 38
Potassium 291 636 560 689
Sodium 2,626 4,050 4,600 4,410
Sulfate 44,586 60,600 74,000 72,200
PHYSICAL PROPERTIES - - - -
pH -3 3.18 3.24 3.11
Total Dissolved Solids (TDS) 67,060 84,300 74,800 84,600
METALS – DISSOLVED1 - - - -
Arsenic 0.54 26.9 19.3 14.2
Beryllium 0.15 0.298 0.245 0.271
Cadmium 1.2 5.50 5.84 5.51
Chromium 1.2 2.75 2.45 2.23
Cobalt 15 46.5 43.8 38.7
Copper 185 106 154 170
Iron 2,420 2,770 3,310 3,230
Lead 0.54 0.566 0.528 0.403
Manganese 178 117 130 160
Mercury 0.000256 0.000256 0.000256 0.000256
Molybdenum 0.54 4.08 3.19 2.24
Nickel 12 123 122 108
Selenium 0.005 0.422 0.647 0.726
Silver 0.9 0.005 0.0057 0.0057
Thallium 1.2 0.361 0.703 0.368
Tin 0.058 0.058 0.058 0.058
Uranium 15 23 29.2 29.9
Vanadium 268 409 463 536
Zinc 53 767 750 582
Notes. (1) June 1991 metals concentrations correspond to total rather than dissolved.(2) June 1991 Nitrate+Nitrite as N not analyzed but original value Nitrate as N 
was <200 mg/L. (3) June 1991 pH not included because original value was not reported. (4) June 1991 Arsenic, Lead, and Molybdenum assumed to equal 0.5 because 
original values were <1 mg/L. (5) Fluoride from 2008 through 2009 assumed to equal one‐half reporting limit or 0.05 mg/L. (6) Mercury assumed to equal one‐half 
lowest reporting limit or 0.00025 mg/L because no detects were reported. (7) Silver from 2007 through 2009 assumed to equal one‐half lowest reporting limit or 0.005 
mg/L. (8) Tin assumed to equal one‐half lowest reporting limit or 0.05 mg/L because no detects were reported. 

K-6

TABLE K-2. Statistical analysis of Cell 2 slimes drain data. All units are in mg/L except for pH which is in standard units.
Analyte 
Number
of
Samples
Minimum Maximum Arithmetic
Mean
Standard
Deviation
Geometric
Mean
Geometric
Standard
Deviation
Median
Mean Minus
0.5 Standard
Deviation
MAJOR IONS - - - - - - - - -
Calcium 4 390 572 500 78 495 1.18 518 461
Chloride 4 2,573 3,860 3,221 653 3,171 1.23 3,225 2,894
Fluoride 4 0.05 3.3 1.4 1.6 0.36 9.81 1.0 0.55
Magnesium 4 2,540 4,100 3,605 726 3,542 1.25 3,890 3,242
Nitrogen, Ammonia as N 4 2,240 4,020 3,280 763 3,206 1.29 3,430 2,899
Nitrogen, Nitrate+Nitrite as N 4 20.3 100 47.3 35.9 39.3 1.96 34.5 29.4
Potassium 4 291 689 544 177 517 1.48 598 456
Sodium 4 2,626 4,600 3,922 893 3,833 1.29 4,230 3,475
Sulfate 4 44,586 74,000 62,847 13,545 61,640 1.26 66,400 56,074
PHYSICAL PROPERTIES - - - - - - - - -
pH 3 3.11 3.24 3.18 0.07 3.18 1.02 3.18 3.14
Total Dissolved Solids (TDS) 3 67,060 84,300 75,387 8,635 75,058 1.12 74,800 71,069
METALS - DISSOLVED - - - - - - - - -
Arsenic 4 0.5 26.9 15.2 11.1 7.8 6.36 16.8 9.7
Beryllium 4 0.15 0.298 0.241 0.064 0.233 1.36 0.258 0.209
Cadmium 4 1.2 5.84 4.51 2.21 3.82 2.16 5.51 3.41
Chromium 4 1.2 2.75 2.16 0.67 2.06 1.45 2.34 1.82
Cobalt 4 15 46.5 36.0 14.4 33.0 1.70 41.3 28.8
Copper 4 106 185 154 34 151 1.28 162 137
Iron 4 2,420 3,310 2,933 416 2,910 1.16 3,000 2,724
Lead 4 0.403 0.566 0.499 0.070 0.495 1.16 0.514 0.464
Manganese 4 117 178 146 28 144 1.21 145 132
Mercury 4 0.00025 0.00025 0.00025 0 0.00025 1.00 0.00025 0.00025
Molybdenum 4 0.5 4.08 2.50 1.53 1.95 2.56 2.72 1.74
Nickel 4 12 123 91 53 66 3.13 115 64.6
Selenium 4 0.005 0.726 0.450 0.323 0.177 10.92 0.535 0.288
Silver 4 0.005 0.9 0.23 0.45 0.018 13.42 0.005 0.005
Thallium 4 0.361 1.2 0.658 0.395 0.579 1.78 0.536 0.460
Tin 4 0.05 0.05 0.05 0 0.05 1.00 0.05 0.05
Uranium 4 15 29.9 24.3 6.9 23.4 1.38 26.1 20.8
Vanadium 4 268 536 419 113 406 1.35 436 362
Zinc 4 53 767 538 334 365 3.64 666 371

APPENDIX L
EVALUATION OF POTENTIAL WATER FLOW THROUGH THE
TAILINGS CELL LINERS

L-1
APPENDIX L
EVALUATION OF POTENTIAL WATER FLOW THROUGH THE TAILINGS
CELL LINERS
The purpose of this appendix is to document the assumptions and methods used to
calculate potential water flux rates through the liners installed beneath the tailings cells at
the White Mesa Site. In principle, a geomembrane liner consists of an impermeable
material that should preclude all leakage. However, the occurrence of a limited number
of installation defects is generally anticipated and is assumed in the assessment of
environmental impacts and sizing of leachate removal systems. Calculated flow rates
through the liners are compared to data published by the U.S. Environmental Protection
Agency (EPA) and data reported for Cell 4A during the first year of operations. There is
evidence to suggest that there has been no leakage of tailings pore water through the liner
systems at White Mesa.
Conservative assumptions were used in the calculations described below. In this
document, the term conservative will generally apply to assumptions that are considered
more protective of the environment, or in this context, will result in greater fluxes (e.g.,
more potential defects). Nevertheless, these conservative estimates of potential water
flux rates through the liners will be used as an upper boundary condition (time-dependent
flux) for the contaminant transport model used to predict flow and solute transport
through the vadose zone to the perched aquifer during the operational, dewatering, and
post-closure steady-state timeframes.
TAILINGS LINER SYSTEMS
Details of the tailings liner systems design and construction are necessary to calculate
potential water flux rates that may migrate into the vadose zone beneath the cells. Details
of the liner systems for Cells 2 & 3 are provided in D’Appolonia Consulting Engineers
(1982), for Cell 4A in Geosyntec Consultants (2006), and for Cell 4B in Geosyntec

L-2
Consultants (2007b). The tailings liner systems are schematically illustrated in Figure L-
1. Leak detection systems were installed beneath the liners and are monitored weekly.
The tailings liner systems for Cells 2 & 3 are identical and consist of a slimes drain
collection system overlying a single liner. The design consists from top to bottom of:
• slimes drain system (cell bottom only)
• liner protective blanket
• 30-mil (0.03-inch) poly vinyl chloride (PVC) flexible membrane liner (FML)
• compacted bedding material
• prepared subgrade with limited leak detection system (i.e., a single pipe at the toe
of the southern dike).
The tailings liner system for Cell 4A is double lined, and consists of a slimes drain
collection system overlying a primary liner, leak detection system, and composite
secondary liner. A composite liner is defined as a geomembrane liner underlain by a
low-permeability soil (e.g., naturally compacted soil or geosynthetic clay layer). The
design for Cell 4B is currently under review but preliminary drawings indicate an
identical design to Cell 4A with minor deviations. The design consists from top to
bottom of:
• slimes drain system (cell bottom only)
• 60-mil (0.06-inch) high-density polyethylene (HDPE) geomembrane (primary
liner)
• geonet drainage layer (leak detection system)

L-3
• 60-mil (0.06-inch) high-density polyethylene (HDPE) geomembrane (secondary
liner)
• geosynthetic clay liner (GCL)
• prepared subgrade.
ASSUMED SIZE OF DEFECTS
Based on forensic analysis of installed landfill liner systems, most defects occur during
installation, and may result from discontinuous seaming and/or accidental puncturing of
the geomembrane during construction activities (Giroud and Bonaparte, 1989). To
estimate potential flux rates through the liners at White Mesa for use in the vadose zone
flow and transport model, defect sizes were assumed based on recommendations
provided in Giroud and Bonaparte (1989). The authors recommend that a circular defect
with a diameter of 2 mm (area of 3.14 x 10-6 m2) should be used for evaluating
performance of the liner, while a circular defect with a diameter of 10 mm (area of 7.85 x
10-5 m2) should be used for sizing leachate removal components and computing
maximum design flows. The smaller diameter defects (small holes) may result from
discontinuous seaming, while the larger diameter defects (large holes) may result from
accidental puncturing of the geomembrane. Generally, large holes in a geomembrane are
included in an analysis to size components of the lining system, and such defects are
generally identified and sealed during routine quality assurance monitoring.
Furthermore, especially in the context of fine-grained tailings, defects that arise may
essentially become sealed by fine-grained slimes during tailings deposition and
consolidation. Additionally, because of the elevated solute concentrations in the tailings
solutions, potential defects may also become sealed by mineral phases that could
precipitate from solution (e.g., gypsum or iron hydroxides). Therefore, for the purposes
of modeling, defects are assumed that would give rise to a potential flux through the
liner, while the actual flux through the liner could be insignificant.

L-4
ASSUMED FREQUENCY OF DEFECTS
To estimate potential flux rates through the liners at White Mesa for use in the vadose
zone flow and transport model, the assumed number of defects was based on
recommendations provided in Giroud and Bonaparte (1989) and Schroeder et al. (1994),
the latter publication being the user’s guide manual for the EPA’s HELP model. General
recommendations follow that one small hole defect is anticipated per acre, while the
number of large hole defects will depend on the quality of the installation. An excellent
installation quality for the liner may have up to one large hole defect per acre, while good
installation quality for the liner may have between one to four large hole defects per acre.
The liner beneath Cells 2 & 3 is comprised of PVC which has a lower puncture resistance
than the HDPE liner beneath Cells 4A & 4B. Therefore, the number of potential large
hole defects beneath Cells 2 & 3 is expected to be greater than the number of potential
large hole defects beneath Cells 4A & 4B. Furthermore, Cells 4A & 4B were installed
more recently (late 2000s versus early 1980s) and as such the liner installation quality is
expected to be better as compared to Cells 2 & 3. The liner installation quality for Cells
2 & 3 and Cells 4A & 4B are assumed to be good and excellent, respectively (these
qualitative ratings are required as part of the calculations).
As part of the sensitivity analysis, a range in potential water flux rates through the liners
as a function of different assumptions regarding the number of defects is presented
below. For Cells 2 & 3, the following scenarios were evaluated:
• base case: 1 small hole and 1 large hole defect per acre
• upper bound: 1 small hole and 3 large hole defects per acre
• lower bound: 1 small hole defect per acre.
For Cells 4A & 4B, only one scenario was calculated:
• base case: 1 small-hole defect per acre.

L-5
The inclusion of only one scenario for Cells 4A & 4B is considered to be representative
because these cells contain a leak detection system installed between the upper primary
and lower secondary liner, have more puncture resistant liners, and the liner installation
quality is considered to be excellent. Furthermore, the head on the secondary liner is
expected to be very small, as explained below.
EQUATION TO DETERMINE POTENTIAL WATER FLOW THROUGH
DEFECTS
The flow of water through a circular defect assuming imperfect contact between the
geomembrane and underlying soil liner can be evaluated using the following equation
Giroud (1997):
ܳൌߚቈ1൅0.1൬݄௪
ܮ௦
൰
଴.ଽହ
቉ ܽ଴.ଵ݄௪଴.ଽܭ௦଴.଻ସ
where:
Q = leakage rate per defect [m3/sec]
β = coefficient of contact between the geomembrane and the underlying
soil liner [dimensionless]
hw = head above geomembrane [m]
Ls = low-permeability soil liner thickness [m]
a = area of defect [m2]
Ks = saturated hydraulic conductivity of the soil liner
[m/sec].
Giroud’s (1997) equation is a semi-empirical equation that must be used with the
identified units. The accuracy of Giroud’s equation was evaluated using numerical
simulations completed by Foose et al. (2001). Overall, leakage rates predicted with
Giroud’s equation were substantially higher and therefore more conservative than those
predicted based on hydraulic theory, especially for facilities constructed with a composite

L-6
liner (a geomembrane overlying a GCL or compacted soil liner). Therefore, Giroud’s
equation is anticipated to over-predict flow rates through the liners for all of the cells.
Furthermore, Giroud’s equation assumes that the material above the liner readily
transmits all available water, which may not be accurate for consolidated fine-grained
tailings that may essentially seal any holes. In reality, the tailings may limit the
transmission of water, thus actual flow rates for a given hole size would be less than the
calculated potential flow rates through the liners. The equations and tests were based on
landfills, which contain materials that have much higher permeability than tailings
slimes.
INPUT VARIABLES USED TO ESTIMATE POTENTIAL WATER FLOW
THROUGH DEFECTS
The total leakage rate per unit area is calculated by summation as a function of the head
above the geomembrane and the total number of defects. The calculated volumetric flux
rates were then divided by the cell area to estimate a flux rate, which was used as the
boundary condition for the bedrock vadose zone flow and transport models.
For Cells 2 & 3, contact between the geomembrane and the underlying soil bedding is
assumed to be good, thus the coefficient of contact (dimensionless empirical coefficient)
is assumed to be 0.21 in the calculations (Khatami et al., 1989). Good contact assumes
that the liner was laid on a well-prepared, smooth soil surface with good wrinkle control.
The thickness of the soil bedding, 0.15 m, is based on as-constructed records
(D’Appolonia Consulting Engineers, 1982). The saturated hydraulic conductivity for the
compacted bedding material was assumed to be 2.0 x 10-7 cm/sec. For simplicity, the
footprint for each cell (Cells 2 & 3) was assumed to equal 70 acres (283,290 m2). The
head on the liner during the operational and dewatering phases is explained below.
For Cells 4A & 4B, the coefficient of contact between the secondary geomembrane and
the underlying GCL was also assumed to equal 0.21 for use in the calculations. The
thickness of the GCL, 0.005 m, was based on material specifications. The saturated
hydraulic conductivity of the GCL was taken as the maximum value reported in

L-7
manufacturer’s specifications (5.0 x 10-9 cm/sec). For simplicity, the footprint for each
cell (Cell 4A & 4B) was assumed equal to 40 acres (161,880 m2). The head on the
secondary liner during the operational and dewatering phases is explained below.
HEAD ON LINERS DURING OPERATIONS AND DEWATERING
The head above the single liner beneath Cells 2 & 3 was used as input to calculate the
potential rate of fluid migration through the liners into the underlying vadose zone. For
Cells 2 & 3, operational data and predictions with the MODFLOW dewatering model
were used to estimate the saturated thickness of the tailings through time (see Appendix J
for details regarding the dewatering model). Cell 2 went into service in 1980 and filling
was completed in 1999, at which time the interim cover was placed, which is equivalent
to a 19-year operational timeframe. Cell 3 went into service in 1983 and completion of
filling and placement of the interim cover is scheduled for 2010, which is equivalent to a
27-year operational timeframe. For simplicity, the average operational period for Cells 2
& 3 (23 years) was used in the flux calculations. For modeling purposes, the head on the
liner was assumed to increase linearly for 13 years from zero to fully saturated
conditions, and then assumed to remain fully saturated for an additional 10 years for a
total operational period of 23 years, at which point active dewatering was assumed to be
initiated. The MODFLOW dewatering model predicted that the tailings would
draindown nonlinearly through time reaching an average saturated thickness of 1.07 m
(3.5 ft) after 10 years (i.e., total operational phase plus dewatering phase equal to 33
years). The maximum saturated tailings thickness during operations was varied as part of
the sensitivity analysis to determine a range of potential liner flux rates for the scenario
that considered the base case number of defects (1 small hole and 1 large hole defect per
acre). For Cells 2 & 3, the following scenarios were evaluated:
• base case: the average saturated thickness across the entirety of the cell (when cell
is entirely full), 5.82 m (19.1 ft), was used as the maximum head during
operations

L-8
• upper bound: the maximum saturated thickness (which only occurs near the
sumps when cell is entirely full), 8.23 m (27 ft), was used as the maximum head
during operations
• lower bound: the average value minus the difference between the upper bound
and base case saturated thicknesses, 3.41 m (11.2 ft), was used as the maximum
head during operations.
For Cells 4A & 4B, the head on the secondary liner during operations and dewatering is a
function of the amount of water that may migrate through the primary liner and remain in
the leak detection system. The head above the secondary liner beneath Cells 4A & 4B
was used as input to predict the potential rate of fluid migration into the underlying
vadose zone. Cell 4A went into service in 2008 and is projected to be filled with tailings
by 2014, which is equivalent to a 6-year operational timeframe. Cell 4B is scheduled to
go into service in 2011 and is assumed to be filled with tailings by 2017, also a 6-year
operational timeframe. As a result, a 6-year operational period was assumed in the
calculations for both Cells 4A & 4B. Potential flow through the primary liner during the
operational phase was evaluated by Geosyntec Consultants (2006, 2007b). A dewatering
model was not constructed for Cells 4A & 4B because dewatering rates were estimated
by Geosyntec Consultants (2007a). Water levels in the tailings for Cell 4A were
estimated to decline to less than 0.3 m (1 ft) after approximately six years of dewatering.
Cell 4A is estimated to be dewatered significantly faster than Cells 2 & 3 due to the
significantly more extensive slimes drain network. The dewatering system in Cell 4B is
assumed to be designed similarly to Cell 4A, thus dewatering rates were assumed to be
similar and the dewatering period was assumed to be approximately six years. The
maximum head on the secondary liner during operations is assumed to equal 0.004 m
(0.01 ft) for Cells 4A & 4B. Significantly reduced head on the secondary liner during
operations for Cells 4A & 4B, as compared to Cells 2 & 3, is due to a more extensive
slimes drain collection system, the upper primary liner, and pumping of the leak detection
system, thus reducing the head on the secondary liner. The maximum head on the
secondary liner was assumed to remain constant throughout the operational and

L-9
dewatering periods (total of 12 years). The actual head on the secondary liner during the
majority of the operational and dewatering periods is expected to be less than 0.004 m.
POTENTIAL FLUX THROUGH LINERS DURING OPERATIONS,
DEWATERING, AND POST-CLOSURE STEADY STATE
Cells 2 & 3 Flux Rate Sensitivity Analysis for Different Saturated Thicknesses
The head on the liner (saturated thickness of tailings) for Cells 2 & 3 during the
operational phase, dewatering phase, and post-closure steady state assuming three
different values of maximum saturated thicknesses during operations (as described
above) are plotted in Figure L-2. The calculated flux of water through the single liner
beneath Cells 2 & 3 as a function of the assumed saturated thicknesses during operations,
dewatering, and post-closure steady state for the base case number of defects (one small
hole and one large hole per acre) is plotted in Figure L-3.
For the base case liner head and base case number of defects, the calculated potential flux
rate for the maximum head during the operational phase is approximately 8.3 mm/yr. For
the upper bound liner head and base case number of defects, the calculated potential flux
rate for the maximum head during the operational phase is approximately 15 mm/yr.
And for the lower bound liner head and base case number of defects the calculated
potential flux rate for the maximum head during the operational phase is approximately
3.6 mm/yr.
During dewatering the calculated potential fluxes decrease from the maximum values
reported above to approximately 0.7 mm/yr at closure (after 10 years of active
dewatering) for all three liner head scenarios assuming the base case number of defects.
As a simplification, excluding early stages of dewatering, the heads were assumed to be
equal for all three liner head scenarios. Therefore, the potential flux rates for all three
scenarios were the same because the post-closure steady state head was predicted to be
the same after dewatering, and the number of assumed defects did not change. The flux

L-10
rates predicted at the end of dewatering are assumed to equal the rate during post-closure
steady state.
Cells 2 & 3 Flux Rate Sensitivity Analysis for Different Liner Defect Frequencies
The calculated flux of water through the single liner beneath Cells 2 & 3 as a function of
the range of assumed number of defects for the base case liner head during operations,
dewatering, and post-closure steady state is plotted in Figure L-4.
For the base case number of defects and base case liner head, the calculated potential flux
rate for the maximum head during the operational phase is approximately 8.3 mm/yr. For
the upper bound number of defects and base case liner head, the calculated potential flux
rate for the maximum head during the operational phase is approximately 18 mm/yr.
And for the lower bound number of defects and base case liner head the calculated
potential flux rate for the maximum head during the operational phase is approximately
3.5 mm/yr.
During dewatering, for the base case line head, the calculated potential flux decreases
from the maximum values reported above to approximately 0.7 mm/yr at closure (after 10
years of active dewatering) for the base case number of defects, 1.5 mm/yr for the upper
bound number of defects, and 0.3 mm/yr for the lower bound number of defects. During
post-closure steady state the flux rates for all three scenarios were slightly different
because the number of assumed defects differed. The flux rates predicted at the end of
dewatering are assumed to equal the rate during post-closure steady state.
Cells 4A & 4B Flux Rate
The calculated flux of water through the secondary liner beneath Cells 4A & 4B for the
maximum head within the leak detection system during the operational and dewatering
periods is approximately 8 x 10-5 mm/yr. The flux rates predicted at the end of
dewatering are assumed to equal the rate during post-closure steady state because the
increase in water levels is anticipated to be minor (see Appendix E).

L-11
DISCUSSION
The EPA (1992) reported measured leakage rates (water pumped from primary leak
detection systems located beneath composite liners) for landfill facilities. The leakage
rates varied according to the level of construction quality assurance (CQA). Liners
installed with excellent CQA had measured leak detection system flow rates that ranged
from <5 to 50 gallons per acre per day (gpad) (<1.7 to 17 mm/yr), while facilities that had
no rigorous CQA had measured leak detection system flow rates that generally were in
excess of 50 gpad (17 mm/yr).
The calculated potential flow rates through the liners beneath Cells 2 & 3 for the
maximum head during the operational phase ranged between 3.5 to 18 mm/yr for the six
sensitivity scenarios presented above. The calculated potential flow rates through the
liners at the end of dewatering and during post-closure steady state, which is more typical
for the conditions evaluated by the EPA, ranged between 0.3 to 1.5 mm/yr for the six
sensitivity scenarios presented above. The calculated values at the end of dewatering are
slightly lower than flux rates from landfill facilities reported in the literature. The
potential difference is assumed to be offset by the difference in hydraulic properties;
leakage rates computed for a tailings facility are expected to be less than the measured
leakage rates for landfills because tailings are likely to have a limited capacity to transmit
all available water and the fine-grained nature of the tailings, coupled with the chemical
nature of the pore water (e.g., precipitation of gypsum), is anticipated to essentially seal
some of the defects.
The measured flow rates within the leak detection system beneath Cell 4A during the first
year of operations (253,955 gallons pumped) for head conditions near capacity averaged
approximately 20 gpad (6.8 mm/yr). The data collected from Cell 4A are not directly
comparable to the EPA data because Cell 4A only has a single liner above the leak
detection system, and the head is near capacity at ~10 meters; the presence of an
underlying compacted soil layer or GCL would significantly reduce these rates. The
measured flow rates through the primary liner of Cell 4A are significantly less than the
value calculated by Geosyntec (2006) by a factor of ~30, which confirms that the

L-12
calculations used to predict flow through the composite liner (secondary liner and
underlying GCL) for Cells 4A & 4B are conservative.
CONCLUSIONS
The potential liner flux rates calculated herein are assumed to be overestimates because
of the conservative nature of the assumptions used. There is strong evidence to suggest
that no significant leakage has occurred through the liner systems beneath Cells 2 & 3
over the past 30 years. Evidence that Cells 2 & 3 are not leaking includes:
• No significant leakage indicated by the leak detection systems
• No leakage indicated by the perched aquifer water table surface
• No observations of contamination (e.g., acid leaching, dissolution of carbonates,
gypsum precipitation, staining) were recorded during drilling of monitoring wells
installed adjacent to the cells during spring 2005
• Total uranium was detected at background levels in bedrock core samples
collected while drilling monitoring wells adjacent to the cells (see Appendix A)
• No contaminants detected in groundwater at levels above natural background
concentrations (INTERA, 2007a; 2007b; 2008), which is corroborated by the
finding that the groundwater age beneath the tailings cells is dominated by water
that is at least 50 years old (Hurst and Solomon, 2008)
• No contaminants detected in groundwater as evaluated through stable isotopes
(Hurst and Solomon, 2008).

L-13
REFERENCES
D’Appolonia Consulting Engineers, Inc., 1982. Construction Report, Initial Phase –
Tailings Management System, White Mesa Uranium Project, Blanding, Utah.
Prepared for Energy Fuels Nuclear, Inc.
EPA, 1992, Action leakage rates for leak detection systems, U.S. Environmental
Protection Agency, Office of Solid Waste, January 1992.

Foose, G.J., C.H. Benson, and T.B. Edil, 2001. Predicting leakage through composite
landfill liners, Journal of Geotechnical and Geoenvironmental Engineering. 127(6): 510-520.

Geosyntec Consultants, 2006. Cell 4A Lining System Design Report for the White Mesa
Mill, Blanding, Utah, prepared for International Uranium (USA) Corporation,
January 2006.
Geosyntec Consultants, 2007a. Analysis of Slimes Drains for White Mesa Mill – Cell 4A,
Computations submitted to Denison Mines, 12 May 2007.

Geosyntec Consultants, 2007b. Cell 4B Design Report, White Mesa Mill, Blanding, Utah, prepared for Denison Mines (USA) Corp., December 2007.

Giroud, J.P., 1997. Equations for calculating the rate of liquid migration through
composite liners due to geomembrane defects. Geosynthetics Int.. Industrial Fabrics
Association International, Minneapolis, 4(3-4):335-348.
Giroud, J.P., and R. Bonaparte, 1989. Leakage through Liners Constructed with
Geomembrane Liners, Parts I and II. Geotextiles and Geomembranes. 8:27-67, 71-
111. Hurst, T.G, and D.K. Solomon, 2008. Summary of Work Completed, Data Results,
Interpretations, and Recommendations for the July 2007 Sampling Event at the
Denison Mines, USA, White Mesa Uranium Mill, near Blanding, Utah. Report
prepared for the Utah Division of Radiation Control, May 2008, pp. 62. INTERA, Inc., 2007a. Revised Background Groundwater Quality Report: Existing Wells
for Denison Mines (USA) Corp.’s White Mesa Uranium Mill Site, San Juan County,
Utah. Prepared for Denison Mines (USA) Corp. October 2007.
INTERA, Inc., 2007b. Revised Addendum Evaluation of Available Pre-Operational and Regional Background Data Background Groundwater Quality Report: Existing Wells
For Denison Mines (USA) Corp.’s White Mesa Mill Site, San Juan County, Utah.
Prepared for Denison Mines (USA) Corp. on 16 November 2007.

L-14
INTERA, Inc., 2008. Revised Addendum Background Groundwater Quality Report: New Wells For Denison Mines (USA) Corp.’s White Mesa Mill Site, San Juan County,
Utah. Prepared for Denison Mines (USA) Corp. on 30 April 2008.

Khatami, A., J. Giroud, and K. Badu-Tweneboah, 1989. Evaluation of the rate of leakage through composite liners. Geotextiles and Geomembranes. 8:337-340.
Schroeder, P.R., N.M. Aziz, C.M. Lloyd, P.A. Zappi, 1994. The hydrologic evaluation of
landfill performance (HELP) model: User’s guide for version 3, EPA/600/R-94/168a,
September 1994, U.S. Environmental Protection Agency Office of Research and Development, Washington, DC.

L-15
Figure L-1. Cross section of tailings liner systems beneath Cells 2 & 3 and Cells 4A & 4B (not to scale).

L-16
Figure L-2. Water levels assumed in the tailings (head on the liner) during the operational period (assumed to be 23 years), dewatering period (assumed to be 10 years),
and post-closure stead state for Cells 2 & 3 for each of the three saturated thickness
scenarios.

L-17
Figure L-3. Calculated potential flux of water through the single liner beneath Cells 2 & 3 for the base case number of defects (one small hole and one large hole per acre) as a
function of three different assumed saturated thicknesses (plotted in Figure L-2) during
the operational phase, dewatering phase, and post-closure steady state.

L-18
Figure L-4. Calculated potential flux of water through the single liner beneath Cells 2 & 3 for the base case saturated thickness (maximum head during operations equal to 5.82
m) as a function of three different assumed defect frequencies during the operational
phase, dewatering phase, and post-closure steady state.

APPENDIX M
GEOCHEMICAL MODEL AND REACTIVE TRANSPORT
MODELING OF FLOW AND TRANSPORT THROUGH THE
BEDROCK VADOSE ZONE
M-1
APPENDIX M
GEOCHEMICAL MODEL AND REACTIVE TRANSPORT MODELING OF
FLOW AND TRANSPORT THROUGH THE BEDROCK VADOSE ZONE
This appendix describes the geochemical and reactive transport model used to predict the
potential transport of conservative and nonconservative solutes through the bedrock
vadose zone beneath the White Mesa Mill tailings cells. Neutralization of the infiltrating
acidic tailings porewater, speciation of solutes, sorption of solutes, and mineral
precipitation/dissolution reactions within the bedrock vadose zone were predicted with
HP1.
The conceptual model describing the potential transport of contaminants through the
bedrock vadose zone is described in Section 2.0 of this report. Details of the
implementation of the conceptual model into the numerical model as well as parameter
values, boundary conditions, and initial conditions used in the modeling are described in
this appendix. A description of the sensitivity analysis and results for the bedrock zone
contaminant transport modeling is also included as part of this appendix.
BACKGROUND
Reactive Transport Modeling
Chemical reactions between dissolved constituents and minerals present within the
vadose zone often dictate spatial and temporal variations in contaminant-plume transport
and mobility. Historically, solute transport processes have been modeled using a
simplified attenuation-based contaminant transport modeling approach, in which a
distribution coefficient (Kd) is calculated and substituted into the advection-dispersion
equation which predicts contaminant fate and transport (e.g., Bethke and Brady, 2000;
Brady and Bethke, 2000; Zhu et al., 2001). However, attenuation-based contaminant
transport models are generally insufficient to describe the complex geochemical reactions
that may occur at mine sites in which acidic solutions containing elevated concentrations
of metals and sulfate may migrate through an unsaturated zone toward a perched aquifer.
M-2
Recent computer and programming advances allow more accurate simulations of reactive
transport processes in the unsaturated zone (e.g., Jacques et al., 2006). In mining and
post-mining environments, neutralization of infiltrating acidic waters coupled with other
geochemical reactions (speciation, sorption, mineral dissolution, mineral precipitation)
control the transport of nonconservative solutes and represent an ideal application of
reactive transport modeling. Therefore, a reactive transport model (e.g., HP1), which
incorporates linkages between flow and reactive transport processes, is preferred and
used in this report over an attenuation-based transport model.
Adsorption of Metals
The sorption of uranium, and other trace elements, onto mineral surfaces is strongly
dependent on the solution pH, initial solute concentration, and mass of adsorbent.
Generally, the amount of sorption increases for decreasing initial concentrations and
increasing mass of adsorbent (Payne, 1999). The variability of sorption as a function of
pH is controlled by the surface charge that develops on the mineral surface. Generally,
cationic metals show maximum sorption at circumneutral pH ranges with decreasing
sorption as the pH changes from neutral to either acidic or basic values. Therefore, a
reactive transport model (e.g., HP1) that accounts for changes in acidity, solution pH, and
solute concentrations during water-rock interactions as solutes are transported through the
unsaturated zone is preferred and used in this report over an attenuation-based transport
model which does not account for these changes.
Reactive transport modeling with HP1 for use in predicting surface complexation
reactions (adsorption/desorption) is preferred instead of laboratory-based Kd tests (using
vadose zone samples exposed to multiple tailings leachate samples with a range of
contaminant concentrations), because:
• HP1 simulations consider the complexity of the tailings solutions, aquifer matrix
chemistry, and potential water-rock reactions that may occur along a flow path in
the subsurface beneath the site.
• Empirical determinations of Kd’s were originally developed to quantify the
sorption of organic compounds and alkali/alkaline earth cations whose speciation
and sorption is nearly insensitive to changes in solution chemistry.
M-3
• Empirical determinations of Kd’s are not an adequate metric for determining the
sorption of uranium and other trace elements because these species are strongly
controlled by the chemical reactions and expected solution chemistry.
• The quantification of potential reactions (required for attenuation-based models) that may occur beneath the facility, during the course of a laboratory experiment,
are very difficult to reproduce, especially taking into account the inherent
variability in the geochemical reactions that may occur during transport through
the vadose zone. Specific examples include:
o Range in uranium concentration and other trace elements
o Range in neutralization potential
o Range in the mass and number of sorbing phases
o Range in alkalinity of (partially) neutralized tailings solutions
o Range in water to rock proportions.
METHODOLOGY
HP1
Reactive transport processes within the bedrock vadose zone, including speciation of
aqueous complexes within the porewater, sorption of aqueous complexes onto mineral
surfaces, dissolution of calcite (acid neutralization), and precipitation of gypsum and
amorphous mineral phases, were predicted using HP1 (Jacques and Simunek, 2005).
HP1 is a reactive transport code that combines the infiltration, unsaturated flow, and
multicomponent contaminant transport modeling capabilities of HYDRUS-1D (Simunek
et al., 2009) with the equilibrium geochemical model PHREEQC (Parkhurst and Appelo,
1999). HYDRUS-1D was used to support the results for flow and transport of a
conservative solute (chloride) predicted by HP1.
The HP1 model was developed by the Belgian Nuclear Research Center in collaboration
with the U.S. Salinity Laboratory and Department of Environmental Sciences at the
University of California at Riverside. The HP1 code retains all of the features
documented in HYDRUS (as described in Section 3.0 of this report) but incorporates
additional modules capable of simulating a broad range of low-temperature geochemical
M-4
reactions in water, soil, and groundwater systems. HP1 can simulate multicomponent
reactive transport, including geochemical interactions with minerals, gases, exchangers,
and sorption surfaces, using thermodynamic equilibrium, kinetics, or mixed equilibrium-
kinetic reactions.
Aqueous Complexation Reactions
The distribution of elements among aqueous species (e.g., uranium as UO2(OH)+,
UO2(SO4)2-2, etc.) and ionic states (e.g., uranium in its +4 oxidation state as the uranous
ion (U+4) or uranium in its +6 oxidation state as the uranyl ion [UO2+2]), has a significant
effect on solution chemistry and contaminant-transport mobility. Generally, dissolved
trace elements in porewater under oxidized conditions will have increased mobility as
compared to reduced conditions. Decreased mobility associated with reduced conditions
often results because minerals with lower solubility tend to precipitate from solution,
sequestering many trace elements including uranium from the dissolved to the solid
phase. Oxidizing geochemical conditions within the vadose zone were assumed.
The speciation of elements and formation of aqueous complexes is governed by mass-
action equations and aqueous-complex formation (stability) constants, and is based on
thermodynamic constraints. Geochemical reactions and formation constants are
contained within the thermodynamic database. For this report, the thermodynamic
database compiled by the U.S. Geological Survey (wateq4f.dat) was updated to be
consistent with uranium thermodynamic data provided by the Nuclear Energy Agency
(Guillaumont et al., 2003) and the U.S. Department of Energy (Bernhard et al., 2000).
Surface Complexation Reactions
Surface complexation modeling with HP1 was used to predict the adsorption of trace
elements using the large body of published literature that has evaluated the sorption of
uranium and other trace elements onto the surfaces of HFO (hydrous ferric oxide)
(Dzombak and Morel, 1990). The diffuse layer (sorption) database developed by
Dzombak and Morel (1990) has also been incorporated by the U.S. EPA into their
geochemical model MINTEQA2 (Allison et al., 1991), by the U.S. Geological Survey
M-5
into their geochemical model PHREEQC (Parkhurst and Appelo, 1999), and into HP1
(Jacques and Simunek, 2005), indicating a general acceptance by the regulatory and
scientific communities.
The Dzombak and Morel (1990) diffuse layer database has been modified slightly to
adjust the sorption coefficients for uranium because the original values tended to
overpredict the amount of adsorption under low-pH conditions and underpredict the
amount of adsorption under high-pH conditions (Mahoney et al., 2009). The agreement
(R2 of ~0.9) between the final model-selected parameters to extensive literature data
(Mahoney et al., 2009), show a consistency that supports the general application of this
revised model in describing uranium adsorption onto HFO.
The surface complexation modeling approach incorporated into this report is functionally
similar to the methodology developed by the U.S. Geological Survey for the U.S. Nuclear
Regulatory Commission, as presented in NUREG/CR-6820 (Davis and Curtis, 2003). As
discussed in NUREG/CR-6820, the use of a surface complexation model that
incorporates linkages between surface and aqueous species is preferable to models that
rely on a constant partition coefficient (i.e., single Kd) or empirical approaches (i.e.,
adsorption isotherms from batch tests).
Surface complexation modeling for the Naturita UMTRA Site suggests that additional
mineral phases (e.g., hematite, montmorillonite/smectite, and quartz) would adsorb
uranium (Davis and Curtis, 2003). Furthermore, adsorption of uranium onto the surfaces
of aluminum and manganese hydroxides is also expected to occur (e.g., Langmuir, 1997).
All of these mineral phases, in addition to HFO, are expected to be present within the
bedrock vadose zone (either as part of the original mineralogy or due to precipitation of
minerals along a flow path during transport), and available to participate in surface
complexation reactions. To remain conservative, these additional mineral phases that
would sorb uranium and other metals were not included in the model. Sorption is only
allowed to occur onto the surface of a single mineral phase (HFO), and the amount of
sorption is limited to the calculated sorbent site densities (i.e., finite number of sorption
sites). As an additional conservative assumption, uranium adsorption was allowed to
M-6
compete with other metals, which would decrease the total amount of uranium that could
adsorb.
Mineral Precipitation/Dissolution Reactions
In addition to sorption, the precipitation/dissolution of minerals during the potential
transport of contaminants through the bedrock vadose zone will control the attenuation
and mobility of nonconservative solutes such as sulfate, aluminum, and iron. The extent
of acid neutralization was based on the measured mass of calcite (acid neutralizing
potential or ANP) of the bedrock vadose zone, while surface complexation reactions
(adsorption) were based on the measured mass of HFO of the bedrock vadose zone (see
Appendix C for a statistical analysis of site-specific geochemical data collected from the
bedrock core samples). Mineral dissolution (calcite) and precipitation (sulfate minerals
and amorphous phases) were based on equilibrium constraints and were mass limited.
The precipitation of aluminum and iron hydroxide phases also introduces additional
mineral acidity (H+ ions), which would consume calcite in addition to the acidic waters
potentially transported through the liners. Coprecipitation reactions were not accounted
for during the geochemical modeling as a simplification and to maintain conservative
assumptions. Coprecipitation of uranium (Abdelouas et al., 1998) and metals onto the
surfaces of precipitating phases was ignored, which could also serve as a sink for metals
and reduce transport mobility.
FLOW AND REACTIVE (CONTAMINANT) TRANSPORT MODEL OF THE
BEDROCK VADOSE ZONE
Reactive transport models were developed for the bedrock vadose zone beneath Cells 2 &
3. Reactive transport models in HP1 were not constructed for Cell 1 (contingency cell
identified for the potential disposal of decommissioning and deconstruction debris), and
Cells 4A & 4B because the model did not predict chloride to exceed 10 mg/L at the
bottom of the bedrock vadose zone, which is the minimum groundwater compliance limit
(GWCL) for monitoring wells located immediately downgradient from the tailings cells
(see Section 4.0 of this report). Considering that chloride is a conservative tracer, and
that transport is not affected by sorption or mineral precipitation reactions, coupled with
M-7
the fact that the model predictions demonstrate nearly zero impact, additional model
predictions of solute transport for nonconservative contaminants (sulfate, uranium, other
trace elements) was considered unnecessary for Cell 1 and Cells 4A & 4B.
Domain
The bedrock vadose zone model extended from the base of the tailings cell liner systems
through the Dakota Sandstone and Burro Canyon Formation to the perched water table
surface (see Figure 3-1). The vadose zone thickness was calculated by taking the
difference between the bottom elevation of the cell and the distance to the water table for
individual monitoring wells. The minimum vadose zone thickness beneath Cells 2 & 3
and Cell 4A was approximately 12.8 m (42 ft) and 12.2 m (40 ft), respectively (based on
2007 water level data). As a comparison, the average vadose zone thickness beneath Cell
2, Cell 3, and Cell 4A are 19.2 m (63 ft), 20.1 m (66 ft), and 17.1 m (56 ft). As a
conservative assumption the minimum vadose zone thickness of 12.8 meters (42 feet)
was assumed for all of the simulations of potential solute transport beneath the cells (see
Appendix C for a discussion of vadose zone thicknesses and a summary table of vadose
zone thickness beneath the tailings cells).
Finite Element Node Spacing
The finite-element nodes were discretized in the vertical direction to simulate layers in
the bedrock vadose zone. The bedrock vadose zone model had a uniform node spacing of
5 cm. In order to reduce numerical errors due to spatial discretization, grid spacing was
based on recommendations provided by Jacques et al. (2006).
Boundary Conditions
Variable specified mass flux rates (flux multiplied by the concentration) were applied to
the upper boundary of the bedrock vadose zone.
Potential water flux rates through the liner systems for Cells 2 & 3 were calculated using
the Giroud-Bonaparte Equation as described in Appendix L. The predicted saturated
M-8
thickness of the tailings during the operational phase, during active dewatering, and
during post-closure steady state was used to calculate the potential flux rate through the
liner for use as an upper boundary condition in the flow and contaminant transport model
of the bedrock vadose zone. Groundwater flow modeling with MODFLOW of Cells 2 &
3 was performed to estimate tailings-dewatering rates through time and average water
levels (saturated thickness) that will remain in the tailings after dewatering (described in
Appendix J). In addition to the maximum saturated thickness of the tailings during
operations, the number of potential liner defects and their impacts on potential water flux
through the liners were evaluated as part of the sensitivity analysis (see Appendix L for
details).
The lower boundary of the domain was assumed to be fully saturated (i.e., water table
conditions with a constant pressure head equal to 0 cm [atmospheric pressure]),
representing the water-table surface of the perched aquifer. A zero concentration
gradient was specified at the lower boundary for solute transport. Because of the one-
dimensional nature of the model, the sides of the domain are implicitly assumed to be
zero-flux boundaries.
Input Parameters
Water Flow. Hydraulic properties required for the vadose zone flow model include
vertical saturated hydraulic conductivity, residual soil water content, saturated soil water
content, and the soil water retention empirical curve-fitting parameters. The saturated
and unsaturated hydraulic properties were measured in core samples collected from the
Dakota Sandstone and Burro Canyon Formation (see Appendix B for original laboratory
report). Bedrock core sample collection methodologies, presentation of soil water
retention and unsaturated hydraulic conductivity curves, and selection of hydrologic units
are discussed in Appendix C. Hydraulic properties used in the model are presented in
Section 3.0 and Table 3-1 of this report.
The vadose zone model assumed a single set of hydraulic properties consistent with the
test results reported for the Dakota Sandstone. This assumption is considered appropriate
because the saturated and unsaturated hydraulic properties of the samples are quite
M-9
similar to one another (see Appendix C). Assignment of a single set of hydrogeologic
properties should not significantly affect the model results given the similarity in
unsaturated hydraulic properties [θ(h)] and [K(h)] for all samples (i.e., there were only
small differences in soil water retention curves or unsaturated hydraulic conductivity
curves for the materials tested). The hydraulic properties (and dry bulk density) from
MW-23 (55.5-56.0 ft) were used as input to the model because the hydraulic functions
are intermediate as compared to the other samples. Unsaturated hydraulic conductivity of
the vadose zone was not included in the sensitivity analysis because the unsaturated
hydraulic conductivities vary to match flux rates under a unit hydraulic gradient.
Contaminants Modeled. The contaminants modeled included pH, major cations and
anions necessary to achieve charge balance (aluminum, calcium, carbonate, chloride,
magnesium, potassium, sodium, and sulfate), and selected trace elements (arsenic,
cadmium, copper, iron, nickel, uranium, vanadium, and zinc). Trace elements included
in the model were based on their elevated concentrations in the tailings slimes drains as
compared to the GWCLs. Aluminum was included and used to obtain charge balance.
These solutes are the most dependable indicators of site water quality and of potential cell
failure due to their predominance (uranium and sulfate) and predominance/mobility
(chloride). In particular, chloride will migrate unretarded and would be expected to be
detected before all other site contaminants. Uranium was included because it is one of
the primary contaminants of concern.
Source Term Concentrations. The average solute concentrations measured in the
tailings slimes drains were used as input to represent the source term solution chemistry
of the tailings pore water (see Appendix K for a discussion of source term chemistries).
The average concentration of chloride, sulfate, and uranium were 3,221 milligrams per
liter (mg/L), 62,847 mg/L, and 24.3 mg/L, respectively. No source degradation,
treatment, or dilution was assumed, that is, concentrations were held constant through
time. As part of the sensitivity analysis, the initial solute concentrations were varied and
ranged between the maximum reported values for the upper bound and the mean minus
one-half standard deviation for the lower bound. The source term concentrations for the
three scenarios are summarized in Table M-1.
M-10
Geochemistry. Geochemical properties of the vadose zone included the mass of ANP
and mass of HFO present in the bedrock vadose zone. The amount of ANP and HFO
were based on measured values obtained from core samples. The sampling methodology,
testing procedures, results, and statistical analysis of the data, in addition to a discussion
regarding the selection of hydrogeochemical units, are summarized in Appendix C, while
the original laboratory data are contained in Appendix A. To simplify the conceptual
model, the geometric mean of the entire population was selected as the base case value
for both ANP and HFO.
The amount of ANP present in the bedrock vadose zone was reported as grams of calcite
(CaCO3) per kilogram of rock and was converted to of moles of calcite per unit volume
of bedrock for input into HP1. As part of the sensitivity analysis, the amount of ANP
was varied with the geometric mean plus one geometric standard deviation for the upper
bound and the geometric mean minus one geometric standard deviation for the lower
bound. The upper bound, base case, and lower bound ANP values were 0.18, 0.11, and
0.04 moles of calcite per unit volume of bedrock. ANP data are considered to be
representative because the test only measures rapidly-reacting carbonate minerals.
The amount of HFO present in the bedrock vadose zone was converted to grams of HFO
per unit volume of bedrock for input into HP1. The mass of HFO per unit volume of
bedrock was 1.8. The total number of sorption sites was based on converting the above
concentration to a molar mass (assuming a molecular weight of 89 grams per mole) and
multiplying by site density recommendations provided by Dzombak and Morel (1990)
(0.2 moles weak sites/mole of iron and 0.005 moles strong sites per mole of iron). The
number of weak and strong sites input into HP1 was 4.0 x 10-3 moles and 9.9 x 10-5
moles, respectively. As suggested by Dzombak and Morel (1990) the surface area of
HFO was input at 600 square meters per gram (m2/g). The amount of HFO did not vary
significantly within the bedrock vadose zone and was not included in the sensitivity
analysis (see Appendix C). HFO is the only solid phase that serves as a potential sorption
site of uranium and other trace elements, which is a conservative assumption because
other phases (e.g., hematite, quartz, clays, etc.) also participate in surface complexation
reactions.
M-11
The partial pressure of oxygen was fixed in the model assuming a dissolved oxygen
concentration in vadose zone porewater equal to 2 mg/L. The partial pressure of carbon
dioxide was fixed in the model assuming 10-2.0 atmospheres of pressure, but was varied
as part of the sensitivity analysis to 10-1.0 atmospheres of pressure used for an upper
bound and 10-3.0 atmospheres of pressure used for the lower bound. Redox conditions
were controlled by the oxygen couple. The following minerals or amorphous phases
were allowed to participate or dissolve, depending on their saturation indices: gypsum,
calcite (ANP), amorphous aluminum hydroxide (HAO), and amorphous iron hydroxide
(HFO). The mass of HFO allowed to participate in surface complexation reactions was
fixed according to measured values in bedrock (geometric mean), and as a conservative
assumption, the HFO that precipitated from solution did not add to the available sorption
sites.
Diffusion. Tortuosity, and its effect on molecular diffusion, was explicitly modeled by
incorporation of a tortuosity factor for the liquid phase (Simunek et al., 2009). Given the
extremely low advective velocity, mechanical dispersion was assumed to be negligible
relative to molecular diffusion (see Section 2.0). Diffusion coefficients for all modeled
solutes were assumed to be equal to 1.75 square centimeters per day (cm2/day) which is
the diffusion coefficient of chloride.
Degradation and Production. No degradation or production of chloride, sulfate,
uranium, or other trace elements was assumed. Radioactive decay of uranium is
considered to be relatively minor due to the slow processes involved (e.g., the half-life
for natural uranium, which is predominantly U-238, is 4.4 x 109 years). Although
uranium and other trace elements can be removed from solution through microbial
processes, to yield more conservative model predictions, these microbial processes were
not simulated.
Initial Conditions
Water Flow. Initial soil water pressure heads within the bedrock vadose zone were
estimated by applying a constant flux boundary using ~1% of average annual amount of
M-12
precipitation. For all HP1 simulations, initial conditions were prescribed as pressure
heads (as opposed to water content) to facilitate model convergence.
Geochemistry. Solution concentrations in the bedrock vadose zone were estimated by
assuming equilibrium of calcite with the HFO. Only calcium and carbonate were
included as aqueous species. While there are naturally-occurring concentrations of
chloride, sulfate, uranium, and other trace elements in the vadose zone initially, the
modeling assumed no initial values for simplicity.
Duration of Simulations and Time Steps
Simulations were run to evaluate potential solute transport during the operational phase,
dewatering phase, and post-closure steady-state timeframes equal to a total duration
simulation of 240 years. The operational and dewatering phases (see Appendix L for
details) were followed by 200 years following closure as required by the Permit.
The minimum and maximum time-steps were 1.04 x 10-2 day (900 seconds) and 180 days
for the HP1 model. The maximum number of iterations per time step was 40. In HP1,
solution efficiency is maximized by incorporating adaptive time-step adjustments based
on criteria described in Simunek et al. (2009). In order to reduce numerical errors due to
temporal discretization, time-step and stability criteria were based on recommendations
provided by Jacques et al. (2006).
Sensitivity Analysis
A sensitivity analysis was performed to quantify the model-prediction uncertainty due to
estimating model input parameters. Three values were selected for each input parameter,
corresponding to a lower bound, base case, and upper bound. Input variables
incorporated into the sensitivity analysis for reactive transport included source term
solution chemistry of the tailings pore water (see Appendix K for details), number of
potential liner defects (see Appendix L for details), acid neutralization potential of the
bedrock (ANP) vadose zone (see Appendix C for details), and partial pressure of carbon
dioxide gas within the bedrock vadose. Results between simulations using different input
M-13
assumptions are compared to evaluate the effect of parameter uncertainty on predictions
of contaminant transport through the bedrock vadose zone. Based on the results for
chloride transport discussed in Section 4.0 of this report, the maximum tailings saturated
thickness was excluded from the sensitivity analysis assessing nonconservative solute
transport. The input variables including ANP and partial pressure of carbon dioxide gas
within the bedrock vadose zone were not included for conservative transport of chloride
because these parameters would only affect the transport of nonconservative solutes.
Because the number of potential liner defects was explicitly simulated in the bedrock
vadose zone model, variability in this parameter was incorporated in the sensitivity
analysis by varying the potential flux rates through the liner.
RESULTS
For the transport of conservative and nonconservative solutes, model-predicted chloride
(conservative) and sulfate (nonconservative) concentrations at the bottom of the vadose
zone (entering the perched aquifer) are presented. In contrast, uranium concentrations
approximately equal to the minimum GWCL are presented at the vadose zone depth at
which they were predicted to occur. Variations in bedrock vadose zone porewater pH
and the depth at which complete calcite dissolution was predicted to occur are also
presented. Sensitivity analysis results are compared to evaluate the effect of parameter
uncertainty on predictions of contaminant transport through the bedrock vadose zone.
Based on the results for conservative transport of chloride (i.e., limited transport distance)
within the bedrock vadose zone beneath Cells 4A & 4B and Cell 1, the sensitivity
analysis was only evaluated for Cells 2 & 3. For all HYDRUS-1D and HP1 simulations
the water and mass balance errors did not exceed 1%. As a general rule-of-thumb, mass
balance errors that do not exceed 3% are considered acceptable.
Chloride Concentrations. The model-predicted chloride concentrations at the bottom of
the bedrock vadose zone beneath Cells 2 & 3 after 240 years of transport was
summarized in Section 4.0 and Table 4-1 of this report. Chloride transport predicted with
HP1 did not differ from simulations predicted with HYDRUS-1D.
M-14
The model-predicted chloride concentration at the bottom of the vadose zone beneath
Cells 2 & 3 for the base case scenario after 240 years of transport is 0.0096 mg/L. The
upper bound model-predicted chloride concentration at the bottom of the vadose zone is
18 mg/L, which is slightly greater than the minimum GWCL of 10 mg/L. However,
considering the extremely low transport rates (0.5 mm/yr), mixing of vadose zone pore
water with groundwater in the perched aquifer system would dilute this concentration
below the minimum GWCL. The lower bound model-predicted chloride concentration at
the bottom of the vadose zone is essentially zero (9.1 x 10-6 mg/L). Assuming all other
variables are equal, the model-predicted chloride concentrations are least sensitive to the
source term chemistry and most sensitive to the number of potential liner defects (which
affects the potential liner flux rate), while the maximum tailing saturated thickness during
operations has an intermediate effect (see Table 4-1, response variable statistic column).
Dissolution of Calcite and pH of Porewater. The bedrock vadose zone depth at which
complete dissolution of calcite was predicted by the model to occur after 240 years of
transport is summarized in Table M-2. The bedrock vadose zone depth for complete
calcite dissolution varied between 0.6 and 2.0 meters; and assuming all other variables
are equal, the results are least sensitive to the partial pressure of carbon dioxide gas and
most sensitive to the mass of ANP within the bedrock followed closely by the number of
potential liner defects (which affects the potential liner flux rate) (see Table M-2,
response variable statistic column). For the base case scenario, complete dissolution of
calcite was predicted to occur at 0.95 meters depth. Complete dissolution of calcite was
correlated with the model-predicted pH of vadose zone porewater (see Figure M-1).
Once complete consumption of calcite occurred within the shallow vadose zone for any
given simulation, variations in subsurface pH of vadose zone porewater were correlated
with equilibrium with HFO (~pH <3.3) followed by HAO (~pH < 5) (see Figure M-2).
The precipitation of gypsum does not affect variations in porewater pH within the
bedrock vadose zone.
Precipitation of Gypsum and Sulfate Concentrations. The model-predicted sulfate
concentration at the bottom of the vadose zone beneath Cells 2 & 3 after 240 years of
M-15
transport was summarized in Section 4.0 and Table 4-2 of this report. Table 4-2 is
reproduced here as Table M-3.
The model-predicted sulfate concentration at the bottom of the vadose zone beneath Cells
2 & 3 for the base case scenario after 240 years of transport is 0.014 mg/L. The upper
bound model-predicted sulfate concentration at the bottom of the vadose zone was 45
mg/L, which is less than the minimum GWCL of 532 mg/L. The lower bound model-
predicted sulfate concentration at the bottom of the vadose zone was essentially zero (1.0
x 10-5 mg/L). Assuming all other variables are equal, the model-predicted sulfate
concentrations at the bottom of the bedrock vadose zone are least sensitive to the ANP of
the bedrock vadose zone and most sensitive to the number of potential liner defects
(which affects the potential liner flux rate), while the source term chemistry and partial
pressure of carbon dioxide gas within the bedrock vadose zone have an intermediate
effect (see Table M-3, response variable statistic column). The distribution of sulfate
within the bedrock vadose zone is primarily controlled by the amount of gypsum that
may precipitate from solution, while below this zone the diffusive transport of sulfate
controls solute concentrations predicted to occur within the bedrock vadose zone (see
Figure M-3).
Uranium Concentrations. The model-predicted bedrock vadose zone depth at which
the uranium concentration approximately equals the minimum GWCL (0.0049 mg/L)
after 240 years of transport beneath Cells 2 & 3 was summarized in Section 4.0 and Table
4-3 of this report. Table 4-3 is reproduced here as Table M-4.
The base case model-predicted depth at which uranium approximately equaled the
minimum GWCL was 2.30 meters (7.5 feet) below the liner system; a minimum of 10.5
meters (34 feet) above the perched water table. The upper bound model-predicted depth
at which uranium approximately equaled the minimum GWCL was 3.9 meters. The
lower bound model-predicted depth at which uranium approximately equaled the
minimum GWCL was 1.3 meters. None of the sensitivity runs predicted that uranium, or
other trace elements (arsenic, cadmium, copper, nickel, vanadium, and zinc), would reach
the perched aquifer in the 240 year period simulated. Assuming all other variables are
M-16
equal, the model-predicted uranium transport depths are least sensitive to the source term
chemistry and most sensitive to the number of potential liner defects (which affects the
potential liner flux rate), while the ANP and partial pressure of carbon dioxide gas within
the bedrock vadose zone have an intermediate effect (see Table M-4, response variable
statistic column). Profile concentrations through time of dissolved uranium within the
bedrock vadose zone are plotted in Figure M-4 for the base case scenario. The
distribution of uranium is primarily controlled by sorption onto the surfaces of HFO
within the bedrock vadose zone, and to a lesser extent the pH of the vadose zone
porewater.
Concentrations of Other Trace Elements. The sorption of uranium was competitive
because additional trace elements were modeled. Solutes included in the model were
based on their elevated concentrations in the tailings pore water as compared to the
GWCLs. Transport of the following trace elements was modeled: arsenic, cadmium,
copper, nickel, vanadium, and zinc. Similar to uranium, these solutes were predicted to
migrate a limited distance below the liner (e.g., a few meters) in the 240 year period
simulated.
CONCLUSIONS
Results of the sensitivity analysis for modeling reactive transport of nonconservative
solutes demonstrates that concentrations of sulfate, uranium, and other trace elements
predicted by the model to potentially enter the perched aquifer will not exceed the
minimum GWCL’s. Model predictions are based on assumptions that are primarily
considered to be conservative.
The transport of uranium and other trace elements are predicted to migrate a limited
distance (a few meters) beneath the liners. The distribution of uranium, and other trace
elements, is primarily controlled by sorption onto the surfaces of HFO within the bedrock
vadose zone, and to a lesser extent the pH of the vadose zone porewater. Sufficient
calcite is present within the bedrock vadose zone to limit the potential transport of acidic
tailings solutions in vadose zone porewater beneath the tailings cells.
M-17
The distribution of sulfate is controlled by the precipitation of gypsum within the bedrock
vadose zone. A significant amount of gypsum, amorphous aluminum hydroxide, and
amorphous iron hydroxide was predicted to precipitate within the shallow bedrock
vadose zone, which would be expected to modify liquid phase saturation and effective
porosities, resulting in decreased water flux rates. It is likely that a layer of mineral
precipitates would act to perch water that could potentially migrate through the liners,
further reducing contaminant transport mobilities and transport distances.
REFERENCES
Abdelouas, A., W. Lutze, and E. Nuttall, 1998. Chemical reactions of uranium in ground
water at a mill tailings site, Journal of Contaminant Hydrology, 34, 343–361.
Allison, J.D., D.S. Brown, and K.J. Novo-Gradac, 1991. MINTEQA2/PRODEFA2, a geochemical assessment model for environmental systems, Version 3.0 User's Manual:
U.S. Environmental Protection Agency Report EPA/600/3-91/021, 106 p.

Bernhard, G., G. Geipel, T. Reich, V. Brendler, S. Amayri, and H. Nitsche, 2000. Uranyl(VI) carbonate complex formation: Validation of the Ca2UO2(CO3)3 (aq) species, Radiochim. Acta, 89, 8, 511-518.

Bethke, C.M. and P.V. Brady, 2000. How the Kd Approach Undermines Ground Water
Clenaup, Ground Water, 38, 3, 435-443. Brady, P.V. and C.M Bethke, 2000. Beyond the Kd Approach, Ground Water, 38, 3,
435-443. 321-322

Davis, J.A. and G.P. Curtis, 2003. Application of Surface Complexation Modeling to Describe Uranium(VI) Adsorption and Retardation at the Uranium Mill Tailings Site
at Naturita, Colorado, Report NUREG CR-6820, U. S. Nuclear Regulatory
Commission, Rockville, MD., pp. 223.

Dzombak D. A. and F.M. Morel, 1990. Surface complexation modeling: hydrous ferric
oxide, John Wiley & Sons, New York, NY, pp. 416.

Guillaumont, R. et al., 2003. Update on the chemical thermodynamics of Uranium,
Neptunium, Plutonium, Americium and Technetium, Elsevier, Amsterdam, The
Netherlands, pp. 919.
Jacques, D., and J. Simunek, 2005. User manual of the multicomponent variably
saturated transport model HP1: Description, verification and examples. Version 1.0.
BLG-998. SCKڄCEN, Mol, Belgium, p. 79.

M-18
Jacques, D., J. Simunek, D. Mallants, and M.Th . van Genuchten, 2006. Operator-splitting errors in coupled reactive transport codes for transient variably saturated
flow and contaminant transport in layered soil profiles. Journal of Contaminant
Hydrology, 88, 197–218.
Langmuir, D., 1997. Aqueous Environmental Geochemistry, Prentice-Hall, Inc., Englewood Cliffs, N.J., pp. 600.

Mahoney, J.J., S.A. Cadle, and R.T. Jakubowski, 2009. Uranyl Adsorption onto Hydrous
Ferric Oxide–A Re-evaluation for the Diffuse Layer Model Database,
Environmental Science and Technology, 43, 9260-9266.
Parkhurst, D.L and C.A.J. Appelo, 1999. User’s guide to PHREEQC User's Guide to
PHREEQC (Version 2)--A Computer Program for Speciation, Batch-Reaction, One-
Dimensional Transport, and Inverse Geochemical Calculations, U.S. Geological
Survey Water-Resources Investigations Report 99-4259, pp. 312. Payne, T.E., 1999. Uranium(VI) interactions with mineral surfaces: controlling factors
and surface complexation modeling, Ph.D. Dissertation, University of New South
Wales, pp. 332.
Simunek, J., M. Sejna, H. Saito, M. Sakai, and M. Th. van Genuchten, 2009. The HYDRUS-1D Software Package for Simulating the Movement of Water, Heat, and
Multiple Solutes in Variably Saturated Media, Version 4.08, HYDRUS Software
Series 3, Department of Environmental Sciences, University of California Riverside,
Riverside, California, USA, pp. 330. Zhu, C., F.Q. Hu, and D.S. Burden, 2001. Multi-component reactive transport modeling
of natural attenuation of an acid groundwater plume at a uranium mill tailings site,
Journal of Contaminant Hydrology, 52, 85–108.

M-19

TABLE M‐1.  SOURCE TERM CONCENTRATIONS OF TAILINGS SLIMES DRAINS USED AS INPUT TO THE 
CONTAMINANT TRANSPORT MODELS.  ALL UNITS ARE IN MG/L EXCEPT FOR PH. 
Analyte Lower Bound 
(Mean Minus 0.5 Standard Deviation) 
Base Case 
(Arithmetic Mean) 
Upper Bound 
(Maximum) 
MAJOR IONS ‐‐‐
Calcium 461 500 572
Chloride 2,894 3,221 3,860
Magnesium 3,242 3,605 4,100
Potassium 456 544 689
Sodium 3,475 3,922 4,600
Sulfate 56,074 62,847 74,000
PHYSICAL PROPERTIES ‐‐‐
pH 3.2 3.2 3.2
Dissolved Oxygen 2.0 2.0 2.0
METALS ‐ DISSOLVED ‐‐‐
Aluminum 3,000 3,000 3,000
Arsenic 9.7 15.2 26.9
Cadmium 3.41 4.51 5.84
Cobalt 28.8 36.0 46.5
Copper 137 154 185
Iron 2,724 2,933 3,310
Nickel 64.6 91 123
Uranium 20.8 24.3 29.9
Vanadium 362 419 536
Zinc 371 538 767
Notes:  A description of the source term concentrations are described in Appendix K.   Aluminum was included and 
used to obtain charge balance.  The partial pressure of oxygen was fixed in the model assuming a dissolved oxygen 
concentration in vadose zone porewater equal to 2 mg/L. 
M-20

TABLE M‐2.  MODEL‐PREDICTED BEDROCK VADOSE ZONE DEPTH AT WHICH ALL CALCITE (ANP) IS CONSUMED 
AFTER 240 YEARS OF TRANSPORT. 
Model 
Run 
Solute Source 
Term 
Concentration 
Partial 
Pressure of 
Carbon 
Dioxide Gas 
Number of 
Potential Liner 
Defects 
Mass of ANP 
Response 
Variable 
Evaluated 
Response
Variable 
Statistic 
Bedrock Vadose 
Zone Depth 
(meters) 
Change in Depth 
(meters) 
1 Base Case Base Case Base Case Base Case 0.95 0
2 Upper Bound Base Case Base Case Base Case 1.10 0.15
3 Lower Bound Base Case Base Case Base Case 0.85 ‐0.10
4 Base Case Upper Bound Base Case Base Case 0.95 0
5 Base Case Lower Bound Base Case Base Case 0.95 0
6 Base Case Base Case Upper Bound Base Case 1.95 1.00
7 Base Case Base Case Lower Bound Base Case 0.40 ‐0.55
8 Base Case Base Case Base Case Upper Bound 0.60 ‐0.35
9 Base Case Base Case Base Case Lower Bound 2.00 1.05
Note:  A description of parameter values used as input to the sensitivity analysis is summarized in the text of this appendix. 
M-21

TABLE M‐3.  MODEL‐PREDICTED SULFATE CONCENTRATIONS AT THE BOTTOM OF THE BEDROCK VADOSE ZONE 
AFTER 240 YEARS OF TRANSPORT. 
Model 
Run 
Solute Source 
Term 
Concentration 
Partial 
Pressure of 
Carbon 
Dioxide Gas 
Number of 
Potential Liner 
Defects 
Mass of ANP 
Response 
Variable 
Evaluated 
Response
Variable 
Statistic 
Sulfate 
Concentration 
(mg/L) 
Change in 
Concentration 
(mg/L) 
1 Base Case Base Case Base Case Base Case 0.014 0
2 Upper Bound Base Case Base Case Base Case 0.017 0.0030
3 Lower Bound Base Case Base Case Base Case 0.012 ‐0.0020
4 Base Case Upper Bound Base Case Base Case 0.034 0.020
5 Base Case Lower Bound Base Case Base Case 0.0085 ‐0.0055
6 Base Case Base Case Upper Bound Base Case 45 45
7 Base Case Base Case Lower Bound Base Case 0.000010 ‐0.014
8 Base Case Base Case Base Case Upper Bound 0.014 0
9 Base Case Base Case Base Case Lower Bound 0.015 0.0010
Note:  A description of parameter values used as input to the sensitivity analysis is summarized in the text of this appendix. 
M-22

TABLE M‐4.  MODEL‐PREDICTED BEDROCK VADOSE ZONE DEPTH AT WHICH URANIUM CONCENTRATIONS
APPROXIMATELY EQUALS THE MINIMUM GWCL AFTER 240 YEARS OF TRANSPORT. 
Model 
Run 
Solute Source 
Term 
Concentration 
Partial 
Pressure of 
Carbon 
Dioxide Gas 
Number of 
Potential Liner 
Defects 
Mass of ANP 
Response 
Variable 
Evaluated 
Response
Variable 
Statistic 
Bedrock Vadose 
Zone Depth 
(meters) 
Change in Depth 
(meters) 
1 Base Case Base Case Base Case Base Case 2.30 0
2 Upper Bound Base Case Base Case Base Case 2.50 0.20
3 Lower Bound Base Case Base Case Base Case 2.15 ‐0.15
4 Base Case Upper Bound Base Case Base Case 3.90 1.60
5 Base Case Lower Bound Base Case Base Case 2.15 ‐0.15
6 Base Case Base Case Upper Bound Base Case 3.70 1.40
7 Base Case Base Case Lower Bound Base Case 1.30 ‐1.00
8 Base Case Base Case Base Case Upper Bound 2.20 ‐0.10
9 Base Case Base Case Base Case Lower Bound 1.55 ‐0.75
Note:  A description of parameter values used as input to the sensitivity analysis is summarized in the text of this appendix.  
 GWCL = groundwater compliance limit. 
M-23
Figure M-1. Profile plots of model-predicted pH of vadose zone porewater through time between 0 and 2.5 meters depth for the base case scenario (upper figure) and scenario
with the highest depth of complete calcite consumption (lower figure), which
corresponded to the simulation that assumed the lower bound mass of calcite.

M-24

Figure M-2. Profile plots of the model-predicted amount of mineral phases dissolved
(calcite) and precipitated (amorphous hydrous ferric oxide [HFO] and amorphous
hydrous aluminum oxide [HAO]) within the vadose zone porewater after 240 years of
transport between 0 and 2.5 meters depth for the base case scenario (upper figure) and scenario with the highest depth of complete calcite consumption (lower figure), which corresponded to the simulation that assumed the lower bound mass of calcite.

M-25
Figure M-3. Profile plots of the model-predicted concentration of sulfate in vadose zone porewater (upper figure) and the amount of gypsum precipitated (lower figure) within the
vadose zone porewater after 240 years of transport between 0 and 8.0 meters depth for
the base case scenario.

M-26
Figure M-4. Profile plots of the model-predicted concentration of uranium in vadose zone porewater through time between 0 and 2.5 meters depth for the base case scenario.

APPENDIX N

PREDICTIVE SIMULATION INPUT AND OUTPUT FILES
IN ELECTRONIC FORMAT ONLY (ON CD)
HYDRUS‐1D & HP1 
Code Version Model Folder Root File Name (all files have a ".h1d" extension)Brief Description
4.13 Cover system Cover/Design Model_1 Infiltration model of cover system for monolithic ET cover.  Anticipated climate record, anticipated root 
density/distribution, and 40% vegetative cover.  Total time 114 years.
4.13 Cover system Cover/Design Model_2 Infiltration model of cover system for ET cover with compacted clay layer.  Anticipated climate record, 
anticipated root density/distribution, and 40% vegetative cover.  Total time 114 years.
4.13 Cover system Cover/Design Model_3 Infiltration model of cover system for ET cover with gravel layer.  Anticipated climate record, anticipated root
density/distribution, and 40% vegetative cover.  Total time 114 years.
4.13 Cover system Cover/Design Model_4 Infiltration model of cover system for rock‐armor cover.  Anticipated climate record.  Only evaporation 
simulated.  Transpiration not simulated.  Total time 114 years.
4.13 Cover system Cover/Veg&Precip Model_1_AC‐P_AC‐root_30%Infiltration model of cover system for monolithic ET cover.  Anticipated climate record, anticipated root 
depth/distribution, and 30% vegetative cover.  Total time 114 years.
4.13 Cover system Cover/Veg&Precip Model_1_AC‐P_AC‐root_30%_WiltP Infiltration model of cover system for monolithic ET cover.  Anticipated climate record, anticipated root 
depth/distribution, and 30% vegetative cover with decreased wilting point and HcritA.  Total time 114 years.
4.13 Cover system Cover/Veg&Precip Model_1_AC‐P_AC‐root_40%_WiltP Infiltration model of cover system for monolithic ET cover.  Anticipated climate record, anticipated root 
depth/distribution, and 40% vegetative cover with decreased wilting point and HcritA.  Total time 114 years.
4.13 Cover system Cover/Veg&Precip Model_1_AC‐P_RP‐root_30%Infiltration model of cover system for monolithic ET cover.  Anticipated climate record, reduced performance 
root depth/distribution, and 30% vegetative cover.  Total time 114 years.
4.13 Cover system Cover/Veg&Precip Model_1_AC‐P_RP‐root_40%Infiltration model of cover system for monolithic ET cover.  Anticipated climate record, reduced performance 
root depth/distribution, and 40% vegetative cover.  Total time 114 years.
4.13 Cover system Cover/Veg&Precip Model_1_INC‐P_AC‐root_40%Infiltration model of cover system for monolithic ET cover.  Increased precipitation, anticipated root 
depth/distribution, and 40% vegetative cover.  Total time 114 years.
4.13 Cover system Cover/Veg&Precip Model_1_INC‐P_AC‐root_40%_WiltP Infiltration model of cover system for monolithic ET cover.  Increased precipitation, anticipated root 
depth/distribution, and 40% vegetative cover with decreased wilting point and HcritA.  Total time 114 years.
4.13 Cover system Cover/StormIntensity/ARI‐100yr Ponding_Daily_ARI‐100yr Infiltration model of cover system for monolithic ET cover.  Daily P and PET input for 10 days.  Anticipated 
root depth/distribution and 40% vegetative cover.  One 100‐yr ARI 1‐hr long storm event simulated.
4.13 Cover system Cover/StormIntensity/ARI‐100yr Ponding_Daily_ARI‐100yr_30d Infiltration model of cover system for monolithic ET cover.  Daily P and PET input for 30 days.  Anticipated 
root depth/distribution and 40% vegetative cover.  One 100‐yr ARI 1‐hr long storm event simulated.
4.13 Cover system Cover/StormIntensity/ARI‐100yr Ponding_Hourly_ARI‐100yr Infiltration model of cover system for monolithic ET cover.  Hourly P and PET input for 10 days.  Anticipated 
root depth/distribution and 40% vegetative cover.  One 100‐yr ARI 1‐hr long storm event simulated.
4.13 Cover system Cover/StormIntensity/Monsoon Ponding_Daily_Sum1987 Infiltration model of cover system for monolithic ET cover.  Daily P and PET input for 1987 monsoon season 
(92 days).  Anticipated root depth/distribution and 40% vegetative cover.
4.13 Cover system Cover/StormIntensity/Monsoon Ponding_Hourly_Sum1987 Infiltration model of cover system for monolithic ET cover.  Hourly P and PET input for 1987 monsoon season 
(92 days).  Anticipated root depth/distribution and 40% vegetative cover.
4.13 Vadose Zone Cell_1/Transport/Base Case VZ‐200yr_BC‐Cl_No‐liner_BC‐Cover Cell 1 transport model for chloride movement through bedrock vadose zone.  Upper boundary based on 
base case long‐term average flux through monolithic ET cover and base case chloride concentration.
4.13 Vadose Zone Cells_4A&4B/Transport/Base Case VZ‐212yr_BC‐Cl Cells 4A & 4B transport model for chloride movement through bedrock vadose zone.  Upper boundary based 
on base case potential flux rate through liner and base case chloride concentration.
4.13 Vadose Zone Cells_2&3/Transport/Base Case VZ‐240yr_BC‐Cl_5.82m_BC‐def Cells 2 & 3 transport model for chloride movement through bedrock vadose zone.  Base case chloride 
concentration, base case number of potential liner defects, and base case tailings saturated thickness.
4.13 Vadose Zone Cells_2&3/Transport/Cl_sens VZ‐240yr_LB‐Cl_5.82m_BC‐def Cells 2 & 3 transport model for chloride movement through bedrock vadose zone.  Lower bound chloride 
concentration, base case number of potential liner defects, and base case tailings saturated thickness.
Table describing tailings cell cover infiltration model and bedrock vadose zone contaminant transport model input files submitted with revised Infiltration and Contaminant Transport 
Modeling (ICTM) Report, March 2010.
4.13 Vadose Zone Cells_2&3/Transport/Cl_sens VZ‐240yr_UB‐Cl_5.82m_BC‐def Cells 2 & 3 transport model for chloride movement through bedrock vadose zone.  Upper bound chloride 
concentration, base case number of potential liner defects, and base case tailings saturated thickness.
4.13 Vadose Zone Cells_2&3/Transport/LinerDefect_sens VZ‐240yr_BC‐Cl_5.82m_LB‐def Cells 2 & 3 transport model for chloride movement through bedrock vadose zone.  Base case chloride 
concentration, lower bound number of potential liner defects, and base case tailings saturated thickness.
4.13 Vadose Zone Cells_2&3/Transport/LinerDefect_sens VZ‐240yr_BC‐Cl_5.82m_UB‐def Cells 2 & 3 transport model for chloride movement through bedrock vadose zone.  Base case chloride 
concentration, upper bound number of potential liner defects, and base case tailings saturated thickness.
4.13 Vadose Zone Cells_2&3/Transport/LinerHead_sens VZ‐240yr_BC‐Cl_3.41m_BC‐def Cells 2 & 3 transport model for chloride movement through bedrock vadose zone.  Base case chloride 
concentration, base case number of potential liner defects, and lower bound tailings saturated thickness.
4.13 Vadose Zone Cells_2&3/Transport/LinerDefect_sens VZ‐240yr_BC‐Cl_8.23m_BC‐def Cells 2 & 3 transport model for chloride movement through bedrock vadose zone.  Base case chloride 
concentration, base case number of potential liner defects, and upper bound tailings saturated thickness.
4.13 (2.1.002 HP1)Vadose Zone Cells_2&3/HP1/BaseCase VZ‐240yr_BC‐sol_C‐2.0_BC‐ANP_5.82m_BC‐def
Cells 2 & 3 reactive transport model for solute transport through bedrock vadose zone.  Base case solute 
concentrations, base case pressure CO2‐gas, base case ANP, base case number of potential liner defects, and 
base case tailings saturated thickness.
4.13 (2.1.002 HP1)Vadose Zone Cells_2&3/HP1/ANP_sens VZ‐240yr_BC‐sol_C‐2.0_LB‐ANP_5.82m_BC‐def
Cells 2 & 3 reactive transport model for solute transport through bedrock vadose zone.  Base case solute 
concentrations, base case pressure CO2‐gas, lower bound ANP, base case number of potential liner defects, 
and base case tailings saturated thickness.
4.13 (2.1.002 HP1)Vadose Zone Cells_2&3/HP1/ANP_sens VZ‐240yr_BC‐sol_C‐2.0_UB‐ANP_5.82m_BC‐def
Cells 2 & 3 reactive transport model for solute transport through bedrock vadose zone.  Base case solute 
concentrations, base case pressure CO2‐gas, upper bound ANP, base case number of potential liner defects, 
and base case tailings saturated thickness.
4.13 (2.1.002 HP1)Vadose Zone Cells_2&3/HP1/CO2_sens VZ‐240yr_BC‐sol_C‐3.0_BC‐ANP_5.82m_BC‐def
Cells 2 & 3 reactive transport model for solute transport through bedrock vadose zone.  Base case solute 
concentrations, lower bound pressure CO2‐gas, base case ANP, base case number of potential liner defects, 
and base case tailings saturated thickness.
4.13 (2.1.002 HP1)Vadose Zone Cells_2&3/HP1/CO2_sens VZ‐240yr_BC‐sol_C‐1.0_BC‐ANP_5.82m_BC‐def
Cells 2 & 3 reactive transport model for solute transport through bedrock vadose zone.  Base case solute 
concentrations, upper bound pressure CO2‐gas, base case ANP, base case number of potential liner defects, 
and base case tailings saturated thickness.
4.13 (2.1.002 HP1)Vadose Zone Cells_2&3/HP1/Conc_sens VZ‐240yr_LB‐sol_C‐2.0_BC‐ANP_5.82m_BC‐def
Cells 2 & 3 reactive transport model for solute transport through bedrock vadose zone.  Lower bound solute 
concentrations, base case pressure CO2‐gas, base case ANP, base case number of potential liner defects, and 
base case tailings saturated thickness.
4.13 (2.1.002 HP1) Vadose Zone Cells_2&3/HP1/Conc_sens VZ‐240yr_UB‐sol_C‐2.0_BC‐ANP_5.82m_BC‐def
Cells 2 & 3 reactive transport model for solute transport through bedrock vadose zone.  Upper bound solute 
concentrations, base case pressure CO2‐gas, base case ANP, base case number of potential liner defects, and 
base case tailings saturated thickness.
4.13 (2.1.002 HP1) Vadose Zone Cells_2&3/HP1/LinerDefect_sens VZ‐240yr_BC‐sol_C‐2.0_BC‐ANP_5.82m_LB‐def
Cells 2 & 3 reactive transport model for solute transport through bedrock vadose zone.  Base case solute 
concentrations, base case pressure CO2‐gas, base case ANP, lower bound number of potential liner defects, 
and base case tailings saturated thickness.
4.13 (2.1.002 HP1)Vadose Zone Cells_2&3/HP1/LinerDefect_sens VZ‐240yr_BC‐sol_C‐2.0_BC‐ANP_5.82m_UB‐def
Cells 2 & 3 reactive transport model for solute transport through bedrock vadose zone.  Base case solute 
concentrations, base case pressure CO2‐gas, base case ANP, upper bound number of potential liner defects, 
and base case tailings saturated thickness.
AC = anticipated case
ARI = average recurrence interval
BC = base case
CO2 = carbon dioxide
ET = evapotranspiration
LB = lower bound
P = precipitation
PET = potential evapotranspiration
RP = reduced performance
UB = upper bound